Ectopic overexpression of c-mpl by retroviral-mediated gene transfer suppressed megakaryopoiesis but enhanced erythropoiesis in mice

Experimental Hematology 27 (1999) 1409–1417

Ectopic overexpression of c-mpl by retroviral-mediated gene transfer suppressed megakaryopoiesis but enhanced erythropoiesis in mice Xiao-Qiang Yana, David L. Laceya, Chris Sarisb, Sharon Mub, David Hilla, Robert G. Hawleyc, and Frederick A. Fletchera a Department of Pathology, bDepartment of Cell Biology, Amgen Inc., Thousand Oaks, CA; cHolland Laboratory, American Red Cross, Rockville, MD

(Received 11 January 1999; revised 27 April 1999; accepted 28 April 1999)

In this report, we tested whether ectopic overexpression of a cell surface receptor cDNA could be used to explore the physiological roles of that receptor. We generated c-mpl overexpressing animals by reconstituting mice with retroviral vectortransduced bone marrow (BM) cells. We observed that platelet counts in the c-mpl overexpressing mice failed to recover to normal levels and remained at ,200 3 106/mL post-transplantation, while platelet numbers in the control mice returned to . 800 3 106/mL by 4 weeks post-transplantation. However, platelet counts in the c-mpl overexpressing mice could be stimulated to normal levels after administration of rhMGDF. No significant changes in peripheral leukocyte counts were observed, although the number of CFU-E, GM-CFC, and CFCmulti were reduced two- to threefold in the BM of the c-mpl overexpressing mice. In addition, enhanced erythropoiesis was observed in the c-mpl overexpressing mice. The mpl receptors on erythroid cells were functional as demonstrated by tyrosine-phosphorylation of mpl receptor on RBC and by in vitro erythroid colony-formation in response to MGDF stimulation, respectively. These results suggested that ectopically expressed mpl receptors competed for ligand in vivo leading to an insufficient amount of circulating thrombopoietin (Tpo) for the development of megakaryocytic lineage. These results further suggest that, in addition to sequestering circulating Tpo, overexpression of the mpl receptor on erythroid progenitors may directly contribute to enhanced erythropoiesis in vivo. Our studies demonstrate that ectopic overexpression of a receptor by retroviral-mediated gene transfer provides an approach to explore the biological roles of novel receptors. © 1999 International Society for Experimental Hematology. Published by Elsevier Science Inc. Keywords: C-mpl—Gene transfer—Megakaryopoiesis—Erythropoiesis

Offprint requests to: Xiao-Qiang Yan, Ph.D., Amgen Inc., 15-2-B, Thousand Oaks, CA 91320-1789; E-mail: [email protected]

Introduction Recent interest in the human genome project has accelerated the discovery of novel cDNAs. It is estimated that most of the estimated 100,000 genes in the human genome will be known within the next decade. Elucidation of the function of these cDNAs will provide the potential to develop novel therapeutics to treat human diseases. However, to understand the physiologic role of newly identified genes remains challenging. Current in vivo methods for identifying the function of novel cDNAs mainly include targeted gene deletion (“knock out”) and transgenic overexpression of a cDNA. The “knock out” and conventional transgenic techniques have revealed the functions of many genes in mice, but these techniques are very time-consuming. In addition, conventional transgenic methods of determining in vivo gene function suffer from expression during embryogenesis where developmental effects may mask or prevent the detection of function in otherwise normal adult animals. Retroviral vector-mediated delivery of novel cDNAs to mice in vivo has become an attractive alternative. Overexpression of cDNA in animals is accomplished by transplantation of bone marrow (BM) cells transduced with recombinant retrovirus encoding the cDNA of interest into lethally irradiated recipient mice. Overexpression of hematopoietic factors such as Tpo [1] and Flt-3 ligand (FL) [2] or nonhematopoietic factors such as platelet-derived growth factor (PDGF) [3] has led to significant phenotypic changes in mice. While all of these cDNAs encode soluble proteins, we hypothesized that receptor function could also be studied in this system. The proposed mechanism is that the overexpressed receptor could sequester its ligand from circulation, leaving an insufficient amount for the development of normal target cells. Since the biology of the c-mpl receptor and its ligand is well characterized, we selected it as a model system for this study. The mpl receptor is mainly found on normal megakaryocytes and platelets, although c-mpl mRNA expression is de-

0301-472X/99 $–see front matter. Copyright © 1999 International Society for Experimental Hematology. Published by Elsevier Science Inc. PII S0301-472X(99)0 0 0 6 9 - 7

1410

X.-Q. Yan et al./Experimental Hematology 27 (1999) 1409–1417

tected in CD341 cells, erythroid cells, and erythroid/megakaryocytic leukemia cell lines [4–7] (review [8]). The interaction of mpl receptor with its ligand, Tpo, plays a major role in megakaryopoiesis and thrombopoiesis. It is believed that circulating Tpo produced in the liver acts on megakaryocytic precursors in the BM, stimulating their development into mature megakaryocytes that shed platelets into the circulation. It has been shown that both c-mpl -/and Tpo -/- mice generated by gene knockout [9–11] have severely reduced BM megakaryocytic precursors, megakaryocytes and circulating platelets. Peripheral red blood cell (RBC) and white blood cell (WBC) counts in those animals were not affected. In addition, the number of BM precursors including erythroid colony-forming cells (CFU-E and BFU-E), granulocyte/macrophage colony-forming cells (GM-CFC), and multilineage colony-forming cells (CFUmix) were also significantly reduced in the c-mpl -/- and Tpo -/- mice, suggesting that the interaction of mpl with Tpo is also important for the development of more primitive hematopoietic precursors including hematopoietic stem cells and non-megakaryocytic lineages [10,12,13]. We report here that mice overexpressing full-length c-mpl via retroviral vector-mediated gene transfer had decreased megakaryopoiesis, similar to that described in the c-mpl -/and Tpo -/- knock out mice. In addition, ectopic overexpression of c-mpl resulted in enhanced erythropoiesis in mice. Severe thrombocytopenia led to recurrent bleeding episodes resulting in anemia and, ultimately, death in these animals. These results demonstrate that in vivo overexpression of a receptor cDNA can be used to explore the potential roles of that receptor.

Materials and methods Mice Eight- to 12-week-old (C57BL/6J 3 DBA/2J) F1 (BDF1) were used as donors for BM harvest and viral infection and as recipients for BM transplantation. Recipient mice were g-irradiated 12 Gy, Cs137 (at split dose of 2 3 6.0 Gy, 4 hours apart) and transplanted with virus transduced BM cells. All mice were purchased from Charles River Laboratories (Wilmington, MA) and housed at Amgen vivarium under pathogen-free conditions and supplied with acidified water post-irradiation. The c-mpl retrovirus A full length of murine c-mpl cDNA was obtained from plasmid pSP73-mpl digested with Not I [14]. The 2.9-kb fragment was cloned into the Hpa I site of MSCV 2.2 [15] by blunt end ligation resulting in plasmid pmpl-MSCV. DNA sequence of plasmid pmpl-MSCV was confirmed by sequencing analysis of cloning junctions. The c-mpl gene was under the transcriptional control of viral LTR promoter, while a neomycin phosphotransferase (neo) gene was driven by internal phosphoglycerate kinase (PGK) promoter (Fig. 1). Parental vector MSCV2.2 and pmpl-MSCV were linearized by digestion with Afl III and independently electroporated into GP 1 E86 cells [16] (kindly supplied by Dr. A. Bank,

Columbia University, New York, NY). A dozen G418 resistant (G418r) clones were expanded and tested for viral production. Viral supernatants harvested from individual clones of GP 1 E86 cells were titrated on NIH3T3 cells as described [1]. Clones with highest G418 resistant titer were expanded and frozen as aliquots. The aliquots from the same viral producing clones were used for the whole experiments. BM infection and transplantation The methods for BM infection have been described previously [1]. Briefly, BM cells harvested from femurs of mice post 4 days of 5-FU treatment (150 mg/kg) were incubated (approximately at 5 3 105/mL) in 150-mm tissue culture dishes containing fresh viral supernatant, 10% fetal bovine serum (FBS), 0.1 % Bovine serum album, 2.5 ng/mL recombinant murine interleukin-3 (rmIL-3), 100 ng/mL each of recombinant human interleukin-6 (rhIL-6), recombinant human interleukin-11 (rhIL-11) and recombinant rat stem cell factor (rrSCF), and 6 mg/mL polybrene, at 378C, 5% CO2. Culture media were replaced four times with fresh viral supernatant and growth factors at intervals of 12 to 16 hours during the 3-day infection period. BM cultures were further selected in media containing 0.4 mg/mL (active) G418 (Life Technologies, Gaithersburg, MD) for 48 hours. Total nonadherent and adherent cells were harvested, washed, and resuspended in 1% BSA saline. Each lethally irradiated animal received 0.5 mL of cell suspension by intravenous injection. Following transplantation, mice were sampled by retro-orbital bleeding. Peripheral blood (PB) was analyzed on a Technicon H1E blood analyzing system (Miles Inc., Tarrytown, NY). Colony assay BM cells (2.0-5.0 3 104/mL) and spleen cells (1.0-3.0 3 105/mL) were plated in 1.0-mL IMDM plus methyl cellulose containing 10% FBS, 1% BSA, lipids, transferrin, insulin, 2.5 ng/mL rmIL-3, 2.5 ng/mL rhIL-1b, 100 ng/mL rrSCF, 100 ng/mL rhIL-6, 100 ng/ mL rhIL-11, and 2 U/mL recombinant human erythropoietin (rhEpo). In cultures where was indicated, G418 was included at 0.5 mg/mL (active). Cultures were scored at day 2 for CFU-E, and at day 9 to 11 under inverted microscope for myeloid colonies including GM-CFC, MK-CFC, CFCmulti, incubated at 378C, 5% CO2. All growth factors used in the current studies were supplied by Amgen Inc. (Thousand Oaks, CA). Recombinant human MGDF (rhMGDF) and recombinant murine MGDF (rmMGDF) were the shorter forms of native Tpo protein [17]. Analysis of the provirus The presence of provirus in the hematopoietic precursors of transplanted animals was determined by single colony PCR analysis with neo specific primers. Granulocyte/macrophage and multilin-

Figure 1. Schematic representation of the MSCV (control) and mplMSCV (mpl overexpressing) retroviral vectors. A full length of murine c-mpl cDNA (2.9 kb) was inserted into Hpa I site of the MSCV vector.

X.-Q. Yan et al./Experimental Hematology 27 (1999) 1409–1417

1411

eage colonies were picked from 9- to 11-day methylcellulose cultures and processed individually. DNA was then amplified by PCR with two sets of primers, neo primers (59-tga atg aac tgc cgg acg ag-39, and 59-gtc aag aag gcg ata gaa gg-39), and platelet-derived growth factor B receptor (PDGF BR) specific primers (59-cat tgg ctc cat cct gca ta-39, and 59-gga taa gcc tcg aac acc ac-39) resulting in a 610 bp and a 750 bp fragment, respectively. PDGF BR acted as an internal control for PCR amplification. Conditions for PCR amplification have been described previously [18].

dard histologic methods to yield 3-mm paraffin sections that were subsequently stained with hematoxylin and eosin. To enhance the identification of megakaryocytes and platelets, immunohistochemistry for von Willebrand’s factor (Dako, Carpinteria, CA) was also performed using an automated immunostainer (Biotek, Santa Barbara, CA). To determine the bleeding episodes in the transplanted mice, various tissue sections were stained for iron deposition as described [21].

Immunoprecipitation and Western blot analysis Immunoprecipitation procedure was the same as described [19]. Briefly, RBC were isolated from reconstituted mice by density gradient separation of whole PB. Cells in the density fraction . 1.080 g/mL were collected and washed twice with IMDM. RBC, 4 3 108, were resuspended in 200mL IMDM and incubated in the presence and absence of 1 mg/mL rmMGDF or rhEpo for 10 minutes. Five milliliters of incubation was stopped by adding an equal volume of 23 sample buffer to generate total cell lysate (TCL). TCL was subjected to SDS-PAGE, Western blotting, and probed with anti-phosphotyrosine antibody (Upstate Biotechnology Inc., Lake Placid, NY). The remainder of each incubation (195 mL) was stopped by adding 200 mL 23 lysis buffer and immunoprecipitated with rabbit anti-mouse mpl antibody (Amgen Inc.). Immunoprecipitates were run on SDS-PAGE gels and probed with anti-mpl and anti- phosphotyrosine antibodies, respectively.

Results The c-mpl overexpressing animals were generated by transplantation of BM cells infected with mpl-MSCV retrovirus, while control animals were produced under parallel conditions with BM cells transduced with the parental MSCV vector. Three independent experiments were performed (Table 1). The frequency of viral-infected hematopoietic precursors prior to and post-G418 selection were . 50% and . 90%, respectively, measured as G418r CFCs. A total of 36 control and 58 c-mpl overexpressing mice were generated. Both groups of animals were transplanted with approximately the same number of cells harvested from cultures that had been initiated with the same number of donor BM cells. At 10.5 weeks post-transplantation, 98% of the BM-derived myeloid colonies were neo-positive as determined by PCR analysis with neo-specific primers (Table 1). PB was analyzed weekly in the transplanted animals. Figure 2A shows platelet counts in the transplanted animals. Platelet counts in the control animals declined post-irradiation and transplantation and recovered to . 800 3 106/mL at 5-week post-transplantation and were maintained at similar levels for the duration of the experiments. However, the platelet numbers in the c-mpl overexpressing mice did not recover to a normal level remaining at , 200 3 106/mL except for the 5-week point (Fig. 2A). At 16 weeks post-transplantation, the platelet numbers in the c-mpl overexpressing mice were 52 6 18 3 106/mL. In every one of the 58 c-mpl overexpressing mice transplanted, the platelet counts failed to recover to normal levels post-transplantation throughout the observation period. On the other hand, the recovery of WBC (Fig. 2B) in c-mpl overexpressing mice post-transplantation was not significantly different from control animals. The levels of lymphocytes, neutrophils, and monocytes were the same as those in control animals (data not shown). As shown in Fig. 2C by immunohistochemical

Detection of soluble mpl receptor The levels of soluble mpl receptor in the platelet-poor plasma (PPP) of reconstituted animals was measured by enzyme-linked immunoabsorbent assay (ELISA) using microwell plates coated with affinity-purified anti-mpl antibodies as described [20]. Recombinant murine soluble mpl receptor (Amgen Inc.) was used as a standard in the ELISA. After incubation with PPP or standard at room temperature for 18 to 24 hours, binding was detected with anti-mpl antibodies coupled to horseradish peroxidase. Histological analysis For histological analysis, c-mpl overexpressing mice (n 5 13), sacrificed at 10.5 and 15.5 weeks after bone marrow transplantation were assessed. The control groups were transplanted with bone marrow cells transfected with the vector sequence alone (n 5 11). For standard organ histology, tissues (brain, lung, heart, thymus, spleen, liver, pancreas, trachea, esophagus, stomach, small and large intestine, ovaries, uterus, adrenal gland, kidney, bladder, bone, bone marrow, skin, and skeletal muscle) were fixed in zinc formalin. For bone and bone marrow histology, femurs were decalcified using a 8N formic acid: 1N sodium formate (1:1) solution after formalin fixation. The tissues were then processed using stan-

Table 1. Reconstituting mice with BM cells transduced with MSCV or mpl-MSCV retrovirus

Constructs

Number of experiments

Viral titer G418r CFU 3 105/mL

Number of mice transplanted (n)

Number of cells/recipient 3 106/mL

% Neo1 CFC BM (16 weeks) (n)

MSCV mpl-MSCV

3 3

10–30 3.0–6.0

36 58

8.2–19.1 5.5–22

.98% (6) .98% (5)

The percentage of neo1 CFC was determined by single colony PCR analysis with neo- and PDGF BR-specific primers. Eight to ten colonies were analyzed for each animal; n indicates the number of mice.

1412

X.-Q. Yan et al./Experimental Hematology 27 (1999) 1409–1417

Figure 2. Peripheral blood analysis of mice transplanted with retroviral transduced bone marrow (BM) cells. Panels A and B shows platelet and white blooc cell (WBC) counts, respectively. Dashed and solid lines represent control and c-mpl overexpressing mice, respectively. Results are mean 6 SEM (n 5 10). Panel C shows immunohistochemical analysis of BM stained with Factor VIII from control (left) and c-mpl overexpressing mice (right) sacrificed at 16 weeks post-transplantation.

analysis of bone sections stained using Factor VIII antibodies, the numbers of megakaryocytes in c-mpl overexpressing mice were significantly decreased. In addition to a reduction in numbers, the megakaryocyte in the c-mpl overexpressing mice had dysplastic features characterized by asynchronous between cytoplasm and nuclei, although both small and large megakaryocytes were present. A group of reconstituted mice were analyzed 10.5 weeks post-transplantation (Table 2). PB analysis showed that the c-mpl overexpressing mice had 40.9%, 49.6%, and 52.5% elevated RBC, HGB content and HCT, respectively, but 82% decrease of platelet counts, compared to the control group. The number of CFU-E, BFU-E, CFU-GM, CFCmulti, CFC-ME, and CFC-MK in the femur of c-mpl overexpressing mice decreased 2.0-, 1.6-, 2.6-, 3.4-, 6.6-, and 4.3-fold compared to the control group, respectively, al-

though the cellularity in the marrow of c-mpl overexpressing mice was the same as that in the control animals. However, the cellularity in the spleens of c-mpl overexpressing mice was significantly elevated. The total number and frequency of splenic CFU-E in the c-mpl overexpressing mice increased ten- and sixfold, respectively, resulting in a 49.5% increase in the total number of CFU-E in c-mpl overexpressing mice, assuming that a femur represents 5% of total marrow cellularity [22]. Although a twofold increase of splenic GM-CFC was observed in the c-mpl overexpressing mice, the total number of GM-CFC decreased 56% in these animals analyzed at 10.5 weeks post-transplantation. Similar results were observed in the c-mpl overexpressing mice analyzed at 16 weeks post-transplantation (data not shown). We next examined the responsiveness of c-mpl overexpressing mice to exogenous administration of rhMGDF.

X.-Q. Yan et al./Experimental Hematology 27 (1999) 1409–1417

1413

Table 2. Effects of over-expression of c-mpl in mice on hematopoiesis

Items Blood WBC 3 106/mL RBC 3 109/mL HGB g/dL HCT % PL 3 106/mL Bone marrow Cell Number 3 106 /F CFU-E /F BFU-E /F GM-CFC / F CFCmulti / F CFC-ME / F MK-CFC /F G418r CFC % Spleen Cell Number 3 107 /spl CFU-E / spl *GM-CFC / spl G418r CFC %

MSCV (n 5 11)

mpl-MSCV (10.5 weeks n 5 6 )

7.7 6 3.4 9.3 6 0.8 13.3 6 1.0 42.1 6 4.0 642 6 275

9.9 6 2.7 13.1 6 1.6 19.9 6 2.4 64.2 6 7.7 117 6 42

13.1 6 0.4 23700 6 5000 9900 6 1300 28600 6 3600 2400 6 600 3300 6 300 6000 6 800 44.6 6 2.5

13.5 6 0.4 11400 6 2400 5800 6 1700 11000 6 2000 700 6 100 500 6 200 1400 6 400 21.2 6 6.3

15.8 6 5.0 58000 6 20000 16100 6 4800 39.0 6 19

26.0 6 6.2 568000 6 83000 32800 6 4800 19.5 6 5.5

Results from the control (MSCV) mice were pooled from 11 and 16 weeks post-transplantation. The numbers of colony-forming cells are mean 6 SEM and the rest were mean 6 SD. *GM-CFC in spleen represents total number of CFCs.

Ten control and 10 c-mpl overexpressing mice were implanted 5 days post-transplantation with osmotic pumps delivering rhMGDF. Both groups received rhMGDF at a dose of 3 mg/kg/day for 15 days. The osmotic pumps were replaced on day 20 with pumps delivering rhMGDF at 10 mg/ kg/day for an additional 17 days. Figure 3 shows that the platelet counts in the c-mpl overexpressing mice responded to rhMGDF at 10 mg/kg/day, but did not respond to the dose of 3 mg/kg/day. Control mice responded to both doses of rhMGDF but not to carrier (data not shown). After withdrawal of rhMGDF, the platelet numbers in the control group returned to the same levels as those in the untreated control mice (Fig. 2A). The platelet numbers in the c-mpl overexpressing mice declined to ≈200 3 106/mL. The peripheral WBC, and RBC counts in the rhMGDF-treated animals did not change significantly compared to carriertreated mice during administration (data not shown). These results suggest that overexpression of c-mpl receptors on hematopoietic cells (mainly on RBC) sequestered circulating Tpo, leaving an insufficient amount for the normal development of megakaryocytes and platelets. In contrast to suppressed megakaryopoiesis, overexpression of c-mpl enhanced erythropoiesis in mice. Figure 4 shows the peripheral RBC counts in the transplanted animals. Increased RBC counts in the c-mpl overexpressing mice were observed as early as 10 days post-transplantation. The overall RBC levels in the c-mpl overexpressing mice

Figure 3. Response of control (s) and c-mpl overexpressing (d) mice to exogenous pegylated (PEG)-rhMGDF administration. The PEG-rhMGDF was delivered by osmotic pumps starting on day 5 post-transplantation at 3 mg/kg/day for 15 days. The osmotic pumps were replaced with new pumps on day 20 for additional 17 days of treatment at a dose of 10 mg/kg/day. Results were mean 6 SEM (n 5 10).

increased 15% to 40% compared with control mice. At approximately 10 weeks post-transplantation, the RBC number in the c-mpl overexpressing mice started to decline. By 16 weeks, all the surviving mice were severely anemic, with RBC counts at 4.9 6 2.2 3 109/mL. Long-term overexpressing c-mpl was lethal to the animals shown in Fig. 4B. A significant loss of c-mpl overexpressing mice occurred at 10 to 11 weeks post-transplantation. Approximately 75% of c-mpl overexpressing mice were dead by 16 weeks post-transplantation. All the long-term survivors were severely anemic and thrombocytopenic. Histological analysis was performed on 11 control, and 13 c-mpl overexpressing mice at 10.5 and 16 weeks post-transplantation. A moderate splenomegly was observed in the c-mpl overexpressing mice with a 43% and 102% increase in weight at 10.5 weeks and 16 weeks post-transplantation, respectively. No changes in the size of the liver or other organs were observed. At 16 weeks post-transplantation, the bone marrow of c-mpl overexpressing mice showed a marked reduced number of megakaryocytes and dysplastic megakaryocytes with hyper-condensed nuclei, scant cytoplasm, and megakaryocytes at different stages of maturation. In addition, a left shift of myeloid maturation was also seen. No morphological changes in marrow matrix structures were found in these mice. Massive internal bleeding was seen in 3/5 c-mpl overexpressing mice sacrificed at 16 weeks post-transplantation. Pathological analysis of sections from c-mpl overexpressing mice including heart, lymph node, and lung revealed previous bleeding episodes by staining for iron deposition, which occurs after local hemorrhage (data not shown). In addition, we did not see any erythroid blastic morphological alterations in PB smears or in histology sections including spleen, liver,

1414

X.-Q. Yan et al./Experimental Hematology 27 (1999) 1409–1417

Figure 4. (A) shows peripheral RBC counts in the control (s) and c-mpl overexpressing (d) mice. Data was pooled from two experiments. Results were mean 6 SEM (n 5 12 to 26 in both groups, except for the 16-week point representing 6 control and 4 c-mpl overexpressing mice). (B) shows the survival of mice transplanted with bone marrow (BM) cells infected with control (dashed line, n 5 27) or mpl-MSCV (solid line, n 5 29) retrovirus.

and lymph node in mpl-overexpressing mice at both 10 and 16 weeks post-transplantation. The elevated RBC counts in the c-mpl overexpressing mice were unexpected findings. To examine whether erythroid precursors in the c-mpl overexpressing mice could respond to MGDF in vitro, spleen cells from control or c-mpl overexpressing mice at 12 to 13 weeks post-transplantation were seeded in methylcellulose cultures containing no growth factor, 2 U/mL rhEpo, or 100 ng/mL rHuMGDF. Figure 5A shows the numbers of CFU-E colonies. In the spleen cells from control mice, rhEpo stimulated the growth

Figure 5. Response of erythroid cells to MGDF in vitro. (A) Shows CFUE numbers in the spleen of control (h) and c-mpl overexpressing (j) mice obtained from methylcellulose culture containing no growth factor, or 2 U/mL rhEpo or 100 ng/mL rHuMGDF indicated as none, E, T, respectively. Results were mean 6 SEM (n 5 6 control and 7 c-mpl overexpressing mice). (B) Shows MGDF-induced tyrosine phosphorylation of mpl receptor on red blood cells (RBC). The blot a was probed with anti-phosphotyrosine antibodies and shows phosphorylated proteins in RBC total cell lysate (TCL) after stimulation with rmMGDF. The blot b and c show Western Blots of RBC lysates immunoprecipitated with anti-mpl antibody and probed with anti-phosphotyrosine antibody (b) and anti-mpl antibody (c), respectively; c and m indicate control and mpl overexpressing mice, respectively.

of CFU-E, but rHuMGDF did not, as expected. In the spleen cells from c-mpl overexpressing mice, both rhEpo and rHuMGDF stimulated the growth of CFU-E, although higher numbers of CFU-E were observed in cultures con-

X.-Q. Yan et al./Experimental Hematology 27 (1999) 1409–1417

taining rhEpo. Similar results were observed in the BM CFU-E assays (data not shown). To document the presence and activation of mpl receptors on mature RBC, RBC were incubated in the presence and absence of 1 mg/mL rMuMGDF (shown in Fig. 5B). RBC lysates from c-mpl overexpressing mice showed an increased family of tyrosine-phosphorylated proteins in response to rHuMGDF stimulation (Fig. 5B, panel a). Panels b and c show the presence of mpl receptor proteins in RBC lysate from c-mpl overexpressing mice, and that mpl receptors became tyrosine-phosphorylated in response to rmMGDF stimulation, respectively. Stimulation of RBC with rmEpo under the same conditions did not result in detectable levels of increased tyrosine-phosphorylated proteins including the mpl receptor (data not shown). These results suggest that the mpl receptor on erythroid cells might directly contribute to the enhanced erythropoiesis in the c-mpl overexpressing mice. We also observed elevated levels of soluble mpl receptor (670 pg/mL and 1090 pg/mL, n 5 7, measured at day 32 and day 46 post-transplantation, respectively) in plasma of mpl overexpressing mice. The levels of soluble mpl receptor in plasma from the control mice were undetectable (, 80 pg/mL, n 5 9).

Discussion We generated mice overexpressing c-mpl in the hematopoietic compartment using the full-length murine c-mpl cDNA encoded in the MSCV retroviral vector. Animals developed a phenotype remarkably similar to that reported in the c-mpl -/- and Tpo -/- mice generated by gene knock out [9,11] including decreased number of megakaryocyte colony-forming cells (MK-CFC), megakaryocytes, platelets and marrow CFCs, except for the erythroid lineage. We believe that overexpression of mpl receptor on non-megakaryocytic cells as well as increased levels of circulating soluble mpl competed with megakaryocytic precursors for circulating Tpo, leaving insufficient amounts of Tpo for physiologic megakaryocyte development. RBC in the c-mpl overexpressing mice could play a major role in competing for and sequestering circulating Tpo, because the ratio of total number of RBC (2 3 1010) to that of MK-CFC (2 3 105) in a mouse is estimated to be 105:1. The results of the rhMGDF infusion study (Fig. 3) support our hypothesis that an insufficient amount of Tpo in the c-mpl over-expressing mice was likely responsible for the decreased numbers of megakaryocytes and platelets. It is intriguing that overexpression of c-mpl in mice led to enhanced erythropoiesis. Increasing evidence suggests that Epo promotes megakaryopoiesis, and that Tpo promotes erythropoiesis both in vitro and in vivo, to some extent [23–28]. The expression of Epo receptor (Epo R) on megakaryocytes and of mpl receptor on erythroid cells suggests that the stimulatory effects of Epo on megakaryopoiesis and Tpo on erythropoiesis were direct [6,29]. However,

1415

the Epo R on megakaryocytes and the mpl receptor on erythroid cells may be only able to generate moderate responses, which were insufficient to reach physiological significance. Mice infected with a recombinant spleen focusforming virus expressing an oncogenic Epo R (but not wild type Epo R), induced megakaryopoiesis and erythropoiesis, demonstrating that enhanced thrombopoiesis could be mediated through the Epo R signing pathway [30]. We observed enhanced erythropoiesis in the c-mpl overexpressing mice, although the decreased RBC counts observed in the late phase of transplantation likely resulted from internal bleeding. We suggest that the c-mpl receptor overexpressed on erythroid cells directly contributed to the enhanced erythropoiesis in vivo. This is based on two observations: first, the erythroid precursors in the BM (data not shown) and spleens (Fig. 5A) of c-mpl overexpressing mice were able to form CFU-E colonies in vitro in the presence of a single growth factor, rhMGDF; second, the c-mpl receptor on mature RBC from c-mpl overexpressing mice could be activated by rmMGDF. This is shown by tyrosine-phosphorylation of a set of proteins including c-mpl receptor in response to rmMGDF stimulation (Fig. 5B). Taken together, our results demonstrate that the mpl receptors on erythroid cells were able to respond to Tpo or MGDF, and generate proliferative signals leading to elevated RBC counts in the c-mpl overexpressing mice. In addition, the interaction of the mpl receptor with its ligand, Tpo, on RBC might generate additional signals resulting in a longer half-life of RBC that remains to be further examined [31]. However, no evidence was found to indicate the development of erythroblastic proliferation in the c-mpl overexpressing mice, demonstrating that ectopic overexpression of a normal receptor did not lead to cellular transformation. Two similar studies of overexpression of c-mpl in mice have been reported and outlined in Table 3. In the first study by Cocault et al. [32], the mpl receptor was expressed by inoculating adult mice with a mixture of replication-competent recombinant Harvey retrovirus containing murine c-mpl cDNA and helper Friend Murine Leukemia Virus (FMuLV). Although gradually reduced platelet levels were observed, all animals developed myeloproliferative syndrome characterized by hepato-splenomegaly with hyperproliferation of erythroid blasts, and died within 9 to 12 weeks after infection. In the second study by Goncalves et al. [33], human c-mpl was overexpressed in mice using a similar approach described in our study. Except for a 50% reduction of circulating platelet number, no significant effects were observed in c-mpl overexpressing animals up to 7 months post-transplantation. The discrepancy between the observations made in our studies and that described by Gonclaves et al. [33] may reflect the different experimental approaches. Most significant, perhaps, is the low gene transfer efficiency described in Goncalves’ study showing that 54% and 28% of the transplanted mice were neo-positive determined PCR analysis at 1 and 7 months post-transplantation, respectively.

1416

X.-Q. Yan et al./Experimental Hematology 27 (1999) 1409–1417

Table 3. Comparison of ectopic overexpression of c-mpl in mice Cocault et al. [32] Species Delivery methods

murine c-mpl inoculated competent virus into adult mice

Levels of expression in vivo

c-mpl RNA was shown in BFU-E and CFU-GM by RT-PCR erythroidblastic transformation no effect gradually decreased lethal myloproliferative syndrome

Effects on erythroid cells Effects on megakaryocytes Effect on platelets Pathological lesions

Conclusions

promoted proliferation of hematopoietic progenitors and could play a role in leukemogenesis

In addition, the expression levels of c-mpl were not documented in their overexpressing animals. Although both studies reported by Cocault et al. [32] and by Goncalves et al. [33] showed decreased platelet counts, no significant changes were observed in the numbers of marrow MK-CFC and megakaryocytes in their c-mpl overexpressing animals, suggesting different mechanisms responsible for the decreased platelet counts. We demonstrated that the long-term overexpression of c-mpl in mice led to the severe thrombocytopenia, anemia, internal bleeding, and death. It is not known whether the function of platelets in the c-mpl overexpressing mice was also altered. In addition, elevated levels of soluble mpl receptor were also observed in the c-mpl overexpressing mice, although a full-length c-mpl was overexpressed. The roles of elevated soluble mpl receptors in circulation remain to be further examined. In conclusion, overexpressing a cell-surface receptor cDNA in the hematopoietic system by retroviral-mediated gene transfer can be used to explore the biological functions of novel receptors in vivo. The mechanism involves the ability of ectopically overexpressed receptors to act as a sink for circulating ligand, thus affecting the normal development of target cells, yet avoiding the effects of embryo development compared to knock-out and transgenic techniques.

Acknowledgments The authors would like to thank Y. Chen and Dr. Del Castillo for technique assistance.

Goncalves et al. [33] human c-mpl lethally irradiated mice were reconstituted with retroviral-transduced BM cells not shown

no effect no effect 50% decrease no significant observation did not affect the differentiation of stem cell and did not result in preferential commitment toward megakaryocytic lineage

Current studies murine c-mpl same as in Goncalves’ report

mpl protein was shown in RBC lysate and plasma by Western analysis, ELISA, respectively increased erythropoiesis decreased megakaryopoiesis sever thrombocytopenia lethal due to sever thrombocytopenia, anemia, and internal bleeding suppressed megakaryopoiesis but enhanced erythropoiesis, with no evidence of leukemic transformation; an alternative approach to explore the biological roles of novel receptors

References 1. Yan XQ, Lacey D, Fletcher F, et al. (1995) Chronic exposure to retroviral vector encoded MGDF (mpl-ligand) induces lineage-specific growth and differentiation of megakaryocytes in mice. Blood 86:4025 2. Juan TS, McNiece IK, Van G, et al. (1997) Chronic expression of murine flt3 ligand in mice results in increased circulating white blood cell levels and abnormal cellular infiltrates associated with splenic fibrosis. Blood 90:76 3. Yan XQ, Brady G, Iscove NN (1994) Overexpression of PDGF-B in murine hematopoietic cells induces a lethal myeloproliferative syndrome in vivo. Oncogene 9:163 4. Vigon I, Mornon JP, Cocault L, et al. (1992) Molecular cloning and characterization of MPL, the human homolog of the v-mpl oncogene: identification of a membrane of the hematopoietic growth factor receptor superfamily. Proc Natl Acad Sci U S A 89:5640 5. Vigon I, Dreyfus F, Melle J, et al. (1993) Expression of the c-mpl proto-oncogene in human hematologic malignancies. Blood 82:877 6. Broudy VC, Lin NL, Sabath DF, Papayannopoulou T, Kaushansky K. (1997) Human platelets display high-affinity receptors for thrombopoietin. Blood 89:1896 7. Graf G, Dehmel U, Drexler HG (1996) Expression of thrombopoietin and thrombopoietin receptor MPL in human leukemia-lymphoma and solid tumor cell lines. Leuk Res 20:831 8. Drexler HG, Quentmeier H (1996) Thrombopoietin: expression of its receptor MPL and proliferative effects on leukemic cells. Leukemia 10:1405 9. Gurney AL, Carver-Moore K, de-Sauvage FJ, Moore MW (1994) Thrombocytopenia in c-mpl-deficient mice. Science 265:1445 10. Alexander WS, Roberts AW, Nicola NA, Li R, Metcalf D (1996) Deficiencies in progenitor cells of multiple hematopoietic lineages and defective megakaryocytopoiesis in mice lacking the thrombopoietic receptor c-Mpl. Blood 87:2162 11. de-Sauvage FJ, Carver-Moore K, Luoh SM, et al. (1996) Physiological regulation of early and late stages of megakaryocytopoiesis by thrombopoietin. J Exp Med 183:651 12. Carver-Moore K, Broxmeyer HE, Luoh SM, et al. (1996) Low levels

X.-Q. Yan et al./Experimental Hematology 27 (1999) 1409–1417

13.

14.

15. 16. 17.

18.

19.

20.

21. 22.

23.

of erythroid and myeloid progenitors in thrombopoietin-and c-mpldeficient mice. Blood 88:803 Kimura S, Roberts AW, Metcalf D, Alexander WS (1998) Hematopoietic stem cell deficiencies in mice lacking c-mpl, the receptor for thrombopoietin. Proc Natl Acad Sci U S A 95:1195 Skoda RC, Seldin DC, Chiang MK, Peichel CL, Vogt TF, Leder P (1993) Murine c-mpl: a member of the hematopoietic growth factor receptor superfamily that transduces a proliferative signal. EMBO J 12:2645 Hawley R, Lieu F, Fong A, Hawley T (1994) Versatile retroviral vectors for potential use in gene therapy. Gene Ther 1:136 Markowitz D, Goff S, Bank A (1988) A safe packaging line for gene transfer: separating vial genes on two different plasmids. J Virol 62:1120 Bartley T, Bogenberger J, Hunt P, et al. (1994) Identification and cloning of a megakaryocyte growth and development factor that is a ligand for the cytokine receptor mpl. Cell 77:1117 Yan XQ, Briddell R, Hartley C, Stoney G, Samal B, McNiece I. (1994) Mobilization of long-term hematopoietic reconstituting cells in mice by the combination of stem cell factor plus granulocyte colony-stimulating factor. Blood 84:795 Mu SX, Xia M, Elliott G, et al. (1995) Megakaryocyte growth and development factor and interleukin-3 induce patterns of protein-tyrosine phosphorylation that correlate with dominant differentiation over proliferation of mpl-transfected 32D cells. Blood 86:4532 Ulich TR, del-Castillo J, Yin S, et al. (1995) Megakaryocyte growth and development factor ameliorates carboplatin-induced thrombocytopenia in mice. Blood 86:971 Bancroft JD, Stevens A (1990) Theory and practice of histological techniques. New York: Churchill Livingstone Schofield R, Cole LJ (1968) An erythrocyte defect in spenectomized X-irradiated mice restored with spleen colony-forming cells. Br J Hematol 14:131 Berridge MV, Fraser JK, Carter JM, Lin FK (1988) Effects of recombinant human erythropoietin on megakaryocytes and on platelet production in the rat. Blood 72:970

1417

24. McDonald TP, Cottrell MB, Clift RE, Cullen WC, Lin FK (1987) High doses of recombinant erythropoietin stimulate platelet production in mice. Exp Hematol 15:719 25. Era T, Takahashi T, Sakai K, Kawamura K, Nakano T (1997) Thrombopoietin enhances proliferation and differentiation of murine yolk sac erythroid progenitors. Blood 89:1207 26. Kieran MW, Perkins AC, Orkin SH, Zon LI (1996) Thrombopoietin rescues in vitro erythroid colony formation from mouse embryos lacking the erythropoietin receptor. Proc Natl Acad Sci U S A 93:9126 27. Papayannopoulou T, Brice M, Farrer D, Kaushansky K (1996) Insights into the cellular mechanisms of erythropoietin-thrombopoietin synergy. Exp Hematol 24:660 28. Kaushansky K, Broudy VC, Grossmann A, et al. (1995) Thrombopoietin expands erythroid progenitors, increases red cell production, and enhances erythroid recovery after myelosuppressive therapy. J Clin Invest 96:1683 29. Fraser JK, Tan AS, Lin FK, Berridge MV (1989) Expression of specific high-affinity binding sites for erythropoietin on rat and mouse megakaryocytes. Exp Hematol 17:10 30. Longmore GD, Pharr P, Neumann D, Lodish HF (1993) Both megakaryocytopoiesis and erythropoiesis are induced in mice infected with a retrovirus expressing an oncogenic erythropoietin receptor. Blood 82:2386 31. Ratajczak MZ, Ratajczak J, Marlicz W, et al. (1997) Recombinant human thrombopoietin (TPO) stimulates erythropoiesis by inhibiting erythroid progenitor cell apoptosis. Br J Haematol 98:8 32. Cocault L, Bouscary D, Le-Bousse-Kerdiles C, et al. (1996) Ectopic expression of murine TPO receptor (c-mpl) in mice is pathogenic and induces erythroblastic proliferation. Blood 88:1656 33. Goncalves F, Lacout C, Villeval JL, Wendling F, Vainchenker W, Dumenil D (1997) Thrombopoietin does not induce lineage-restricted commitment of Mpl-R expressing pluripotent progenitors but permits their complete erythroid and megakaryocytic differentiation. Blood 89:3544