Enhanced Recovery from Thrombocytopenia and Neutropenia in Mice Constitutively Expressing a Placental Hematopoietic Cytokine
http://www.100md.com
内分泌学杂志 2005年第1期
Department of Biochemistry, Molecular Biology, and Cell Biology, Northwestern University, Evanston, Illinois 60208
Address all correspondence and requests for reprints to: Daniel Linzer, Weinberg College of Arts and Sciences Dean’s Office, 1918 Sheridan Road, Evanston, Illinois 60208. E-mail: dlinzer@northwestern.edu.
Abstract
Expression of the placental hormone, prolactin-like protein E (PLP-E), a potent cytokine that acts on multiple myeloid lineages, is normally restricted to pregnancy and certain hematopoietic disease states. We hypothesized that the restricted pattern of PLP-E expression is necessary to avoid hyperstimulation of myelopoiesis. To test this idea, we have produced PLP-E transgenic mice and analyzed their steady-state blood cell levels. We find that blood cell levels remain in the normal range, and thus the constitutive expression of a cytokine of pregnancy fails to overcome the tight control of hematopoietic set points for blood cell levels. In contrast, an effect of constitutive PLP-E expression is detected during the recovery from low blood platelet levels (acute thrombocytopenia) and from low granulocyte levels (acute neutropenia) but not from anemia. Mice producing high circulating concentrations of PLP-E recover more rapidly from both thrombocytopenia and neutropenia, as seen both by an earlier increase of progenitor numbers in the bone marrow and the earlier return to normal circulating blood cell levels.
Introduction
BLOOD PLATELETS ARE the end product of the pathway of megakaryocyte proliferation and differentiation. A broad array of cytokines act on the steps leading from the initial cell committed to the megakaryocyte lineage, colony-forming unit megakaryocyte (CFU-MK), through the terminal differentiation of the mature megakaryocyte. Most of these cytokines are constitutively present and include thrombopoietin, IL-3, IL-6, and stem cell factor. Two other cytokines that act on megakaryocyte progenitors are produced specifically by the placenta in mice and thus are present during pregnancy in the female but otherwise are not present in healthy female or male mice. These two hormones, prolactin-like protein (PLP)-E and PLP-F, by themselves stimulate megakaryocyte differentiation but not CFU-MK proliferation. In contrast, in the presence of other cytokines, PLP-E and PLP-F contribute a synergistic stimulation of CFU-MK growth (1, 2).
The effects of PLP-E are not restricted to the megakaryocyte lineage. This hormone has been found to stimulate proliferation and differentiation of erythroid cell lines (3). In experiments using myeloid progenitors from human bone marrow, mouse PLP-E was shown to act synergistically with thrombopoietin to enhance growth of CFU-MK, burst-forming unit erythroid (BFU-E) progenitors, and the common myeloid progenitor, CFU-granulocyte/erythroid/macrophage/megakaryocyte (CFU-GEMM) (4). Recently, we have found that PLP-E and PLP-F act on all mouse myeloid progenitors, including CFU-granulocyte/macrophage (CFU-GM), BFU-E, and CFU-GEMM (our manuscript in preparation).
PLP-E and PLP-F are closely related: the genes encoding these two proteins are adjacent on mouse chromosome 13 (5, 6); the secreted proteins share 54% identity in amino acid sequence (5); these hormones share the same receptor on megakaryocytes; and they function as agonists (2). Despite these similarities, the patterns of PLP-E and PLP-F synthesis display some striking differences. Transcription of the PLP-E gene initiates several days earlier in gestation than for the PLP-F gene, and the two genes are active in distinct cell types: the polyploid trophoblast giant cells for PLP-E and the diploid spongiotrophoblasts for PLP-F (5). Furthermore, the PLP-E gene can also become active in bone marrow cells in response to severely low platelet levels (thrombocytopenia), whereas the PLP-F gene appears to remain silent under this condition (7).
The production of PLP-E in bone marrow cells during thrombocytopenia suggests a functional role for this hormone in the recovery of platelet levels from the disease state. Indeed, exogenous administration of PLP-E during thrombocytopenia hastens recovery (7). We therefore decided to explore whether mice constitutively expressing high levels of PLP-E would be protected from thrombocytopenia. Our finding that PLP-E acts on all myeloid progenitor populations (manuscript in preparation) suggests that the constitutive expression of this hormone may also hasten recovery from other blood cell diseases, in particular anemia and neutropenia. Furthermore, by expressing PLP-E endogenously, delivery of this cytokine to the bone marrow compartment could be achieved more readily and continuously compared with any injection route.
By producing such mice, we would also be in a position to examine whether the presence of high levels of PLP-E continuously in the adult would alter the hematopoietic set point for platelet levels or for other myeloid lineages. Alterations in blood cell levels would be consistent with the restricted expression of this hormone during pregnancy and hematopoietic disease, both of which represent physiological states of significant stress. Conceivably, restricting PLP-E gene expression in this way would enable this hormone to enhance survival in stress situations while preventing PLP-E from stimulating excessive hematopoiesis under normal conditions.
The approach we took was to create transgenic mice in which transcription of the PLP-E cDNA is driven by the mouse metallothionein I (MTI) gene promoter to achieve expression at moderately high levels in a broad distribution of tissues. Importantly, we had shown that the PLP-E receptor is present on megakaryocytes of nonpregnant female and male mice (7), so myeloid cells should be capable of responding to circulating PLP-E in transgenic mice. We have generated several transgenic lines and have analyzed blood cell levels under normal conditions and after induction of blood cell disease.
Materials and Methods
PLP-E transgenic mice
An expression construct containing the MTI promoter (kindly provided by Kelly Mayo) followed by the full-length PLP-E cDNA and the SV40 splice/polyadenylation signals was generated. The DNA fragment containing the MTI promoter, PLP-E cDNA including its leader sequence, and SV40 splice and polyadenylation signals was released from the vector by restriction enzyme digestion, purified by GenElute Minus EtBr Spin Columns (Sigma Chemical Co., St. Louis, MO) after gel electrophoresis, and injected into fertilized mouse eggs. Embryos were transferred into the oviduct of pseudopregnant foster mothers later that day. Tail DNA from the F0 progeny were examined by PCR for integration of the transgene into mouse genomic DNA, and Western blotting of serum was used to assay for the accumulation of PLP-E protein in the circulation. Transgenic mice and control mice were maintained on 14-hr light, 10-hr dark cycles with food and water freely available. All procedures were conducted with the approval of the Northwestern University Animal Care and Use Committee.
DNA extraction, PCR, and Western blotting
Genomic DNA was extracted from mouse tails using a standard protocol. In brief, mouse tail tissues were digested with proteinase K at 55 C overnight. Samples were then extracted with phenol:chloroform (1:1) followed by two extractions with chloroform alone. Genomic DNA was precipitated in 300 mM sodium acetate, 95% ethanol. Pellets were washed twice with 70% ethanol, resuspended in distilled water or 10 mM Tris-HCl, 1 mM EDTA (pH 7.4), and stored at –20 C.
Primers to detect the insertion of the transgene were designed from the MTI promoter (upstream) and PLP-E (downstream) sequences: primer 1, 5'-GTCCCGCTGTGCACACTGGCGCTC-3'; primer 2, 5'-CCTCAAGAACTGGAGATAAAA-3'.
Genomic DNA from wild-type CD-1 mice was used as negative control and L19 ribosomal RNA as a positive internal control.
For Western analysis, blood samples were collected from the tails of PLP-E transgenic mice and nontransgenic littermates. A rabbit polyclonal antiserum prepared against glutathione-S-transferase-PLP-E was used to detect PLP-E in these samples. Serum from d-10.5 pregnant mice, the sample with peak levels of PLP-E during gestation, was used as the positive control.
Blood diseases and blood cell counting
Blood samples were collected from the retroorbital plexus from PLP-E transgenic mice and nontransgenic siblings at several developmental stages. Platelet, erythrocyte, neutrophil, and total white blood cell levels were measured by the Clinical Pathology Facility, Center for Comparative Medicine at the Feinberg School of Medicine, Northwestern University. To induce thrombocytopenia, neuraminidase (Clostridium perfringens type VI, Sigma) in sterile 150 mM NaCl was injected ip at 0.05 U/mouse (8). To induce anemia, mice were injected with phenylhydrazine at 80 mg/kg ip on day 0 and injected 24 h later with 40 mg/kg ip (9, 10, 11). For neutropenia, mice received a single injection of purified monoclonal antibody Rb6-8C5 at 0.15 mg/mouse ip (12).
Bone marrow cell cultures
Femurs from 16-wk-old transgenic mice and wild-type CD-1 control mice were flushed with Iscove’s modified Dulbecco’s medium (IMDM) and 10% fetal bovine serum (Life Technologies, Rockville, MD), and collected material was passed through 21-gauge and then 25-gauge needles to separate cells. After 1 h in IMDM at 37 C, nonadherent cells were collected by centrifugation and washed with PBS and resuspended with IMDM and 1% Nutridoma (Roche Molecular Biochemicals, Indianapolis, IN). These cells were then cultured in MegaCult C medium (Stem Cells Technology, Vancouver, Canada) for CFU-MK colony-forming assays or CollagenCult medium for CFU-GEMM, CFU-GM, and BFU-E colony-forming assays. Initial densities were 5 x 105 nucleated bone marrow cells per well in two-well chamber slides.
For colony formation and differentiation assays, cultures were maintained for 7–10 d for CFU-MK, 10–14 d for CFU-GEMM and CFU-GM, and 8 d for BFU-E. Slides were dried and stained for colony detection and counting. Acetylcholinesterase activity was used as surface marker for MK cells, and MK colonies were defined as clusters of at least three cells staining positive for acetylcholinesterase. To score CFU-GEMM, CFU-GM, and BFU-E colonies, slides were stained with May-Grünwald and Giemsa, and colonies were defined based on their morphological characteristics.
Results
PLP-E transgenic mice
The MTI promoter linked to the PLP-E cDNA was microinjected into fertilized eggs to generate transgenic mice. Preliminary transient transfection studies had demonstrated that this construct results in the synthesis and secretion of mature and functional PLP-E protein (data not shown). Neonates resulting from these eggs were screened by PCR of genomic DNA for integration of the transgene, and Western blotting of serum was used to identify the presence of PLP-E in the circulation. Four PLP-E transgenic lines were established: TG6, TG7, TG14, and TG21. TG6 and TG7 display moderate levels of circulating PLP-E, whereas TG14 and TG21 have high levels of expression of the transgene (Fig. 1).
FIG. 1. Circulating PLP-E in transgenic mice. Serum samples were collected from wild-type and transgenic mice. Equal amounts of protein were fractionated by SDS -PAGE, transferred to a membrane, and probed with anti-PLP-E antiserum. Numbers on the left indicate the position of molecular weight markers. Serum from d-10.5 pregnant mice (d10) was used as a positive control, and serum from wild-type CD-1 mice (wt) was used as a negative control. The four transgenic lines 6, 7, 14, and 21 show increasing levels of PLP-E accumulation.
Hematology of PLP-E transgenic mice
To determine whether the constitutive presence of PLP-E results in the chronic alteration of blood cell concentrations, blood samples of transgenic mice and nontransgenic littermates were assayed from wk 3 to wk 15 of age. No difference was seen at any time point between these groups in terms of platelet, erythrocyte, or total white blood cell levels (Fig. 2). Thus, continuous exposure to PLP-E beyond gestation does not reset steady-state blood cell levels.
FIG. 2. Hematology analysis of transgenic mice. Blood samples were collected from TG6, TG7, TG14, TG21, and wild-type mice (n 6) at different time points: wk 3, 7, 11, and 15 after birth. Platelet, erythrocyte, and white blood cell levels were measured. Platelet levels (A) from transgenic mice show no increase compared with wild-type mice at any time point. Red blood cell (RBC) levels (B) and white blood cell (WBC) numbers (C) from four transgenic mice groups show no significant difference compared with wild-type mice at all time points. Statistical significance was determined by a one-way ANOVA and a post hoc Tukey’s test.
In addition to monitoring blood cells in the circulation, myeloid progenitor populations in the bone marrow of transgenic and nontransgenic controls were also compared. Primary bone marrow cells were cultured in medium containing IL-3 at 10 ng/ml to stimulate proliferation and colony formation. After 7–14 d of incubation, the numbers of CFU-MK, CFU-GM, BFU-E, and CFU-GEMM colonies were counted. At least two mice from each transgenic line were tested (Fig. 3). Again, no differences were found.
FIG. 3. Myeloid progenitor numbers in transgenic mice. Bone marrow was collected from transgenic lines and age-matched wild-type mice. Two mice were selected from each group, and measurements were made in triplicate. Bone marrow cells were grown in semisolid CollagenCult medium for CFU-GEMM (A), BFU-E (C), and CFU-GM (D); MegaCult C medium was used for CFU-MK (B) assays. CFU-MK colonies were identified by staining for acetylcholinesterase activity, and CFU-GEMM, CFU-GM, and BFU-E were stained by May-Grünwald and Giemsa solutions and scored based on their morphological properties. Statistical significance was determined by a one-way ANOVA and a post hoc Tukey’s test.
Recovery from thrombocytopenia
We had previously noted that the PLP-E gene becomes active in the bone marrow a few days after the onset of thrombocytopenia and that injection of PLP-E into thrombocytopenic mice hastens the recovery to normal platelet levels (7). If PLP-E is indeed a limiting factor in the recovery period, these findings suggest that mice constitutively expressing PLP-E would recover faster than wild-type mice. Thrombocytopenia was therefore induced in transgenic and nontransgenic control mice by single injections of neuraminidase, and blood was collected daily for platelet counts (Fig. 4A). For wild-type mice, complete recovery of platelets requires 4–5 d. The recovery period for the two transgenic lines expressing moderate levels of PLP-E, TG6, and TG7, was indistinguishable from that for the wild-type mice. In contrast, the recovery period was significantly faster in the two transgenic lines, TG14 and TG21, that produce high concentrations of PLP-E. Platelet levels are significantly higher in these transgenic mice by the second day of the recovery period and return to normal by d 3. To verify that the effect of this treatment regime was specific for the platelet compartment, erythrocyte and white blood cell levels were also monitored and were found not to change (data not shown).
FIG. 4. Recovery from thrombocytopenia. Neuraminidase was injected at d 0 into groups of transgenic and wild-type mice (n 3 for each group) to induce thrombocytopenia, and blood samples were collected daily for cell counting. A, Platelet levels in two transgenic strains with high PLP-E expression (TG14 and TG21) are significantly above the controls on d 2 (P < 0.001) and d 3 (P < 0.01). By d 3, platelet levels in TG14 and TG21 mice have returned to normal, whereas wild-type mice and the two transgenic mouse lines with low PLP-E expression recover from thrombocytopenia by d 4–5 after injection. Bone marrows were also collected from TG21 and wild-type mice, and equal cell numbers (5 x 105) were plated per well in MegaCult for CFU-MK (B) or in CollagenCult for CFU-GEMM (C) colony-forming assays. CFU-MK numbers are elevated in the transgenic strain on d 2 (P < 0.01), and CFU-GEMM numbers are higher in the TG21 mice on d 1 and 2 (P < 0.01). Statistical significance (marked by asterisks) was determined by a one-way ANOVA and a post hoc Tukey’s test.
The more rapid recovery of platelet levels in TG14 and TG21 mice predicts that the high PLP-E-expressing transgenic mice would also show a rapid elevation in bone marrow myeloid progenitor numbers. Primary bone marrow cultures were prepared each day after neuraminidase injection from TG21 mice and from wild-type mice. Myeloid progenitor numbers were then determined in colony-forming assays (Fig. 4, B and C). As predicted, levels of the platelet progenitor CFU-MK, as well as the multipotential myeloid progenitor CFU-GEMM, increase more rapidly in TG21 mice. CFU-GEMM numbers increase first, by the first day of recovery, followed by an increase in CFU-MK numbers. These transgenic mice and the control mice were identical in terms of the levels of the myeloid progenitors, BFU-E and CFU-GM, that do not lead to platelet formation (data not shown).
Recovery from anemia
Because PLP-E cooperates with other cytokines to stimulate BFU-E growth, we next explored whether constitutive expression of PLP-E can stimulate recovery from acute anemia. Erythrocyte levels in the transgenic mice and in wild-type control mice were rapidly decreased by a single injection of phenylhydrazine. Blood samples were collected and examined every 24 h over the course of the next 9 d. No change in the rate of recovery is seen comparing PLP-E-expressing transgenic with wild-type mice (Fig. 5A). Furthermore, primary bone marrow cell cultures generated from the transgenic and wild-type mice challenged with phenylhydrazine show no difference in CFU-GEMM or BFU-E numbers (Fig. 5, B and C).
FIG. 5. Recovery from anemia. A, Phenylhdrazine was injected on d 0 into transgenic (TG14 and TG21) and wild-type control mice to induce anemia (n = 4 for each group). Blood samples were collected daily for cell counting. No statistically significant difference in erythrocyte levels between transgenic and wild-type mice is seen at any time point. Bone marrows were also collected from transgenic strain TG21 and wild-type mice at each time point. Equal cell numbers (5 x 105) were plated per well in CollagenCult for BFU-E (B) or CFU-GEMM (C) colony-forming assays. No significant differences are detected between transgenics and controls.
Recovery from neutropenia
Finally, we tested the ability of PLP-E to bring about a more rapid recovery from neutropenia. Again, transgenic and wild-type control mice were injected, this time with monoclonal antibody Rb6-8C5. Neutrophil levels declined quickly after injection, and recovery in the wild-type mice required approximately 5 d (Fig. 6A). In the two high-expressing transgenic strains, neutrophil levels rose more rapidly and were significantly increased relative to control mice on d 2 in the TG14 strain and on d 3 in both the TG14 and TG21 strains. Consistent with this rapid recovery in the transgenics, bone marrow levels of the progenitor population CFU-GEMM and GFU-GM were elevated in Rb6-8C5 antibody-injected TG21 transgenics on d 2 and 3, respectively, compared with injected wild-type mice (Fig. 6, B and C).
FIG. 6. Recovery from neutropenia. Transgenic (TG14 and TG21) mice and wild-type controls were injected with monoclonal antibody Rb6-8C5 to induce neutropenia, and blood samples were collected daily for cell counting (n = 4 for each group). Neutrophil levels in the TG14 mice are significantly elevated (asterisks) compared with controls on d 2 and 3 (P < 0.01) and in the TG21 mice on d 3 and 4 (P < 0.01). By d 5, the number of circulating neutrophils in transgenic and wild-type groups are not statistically different from those at d 0. Bone marrows were also collected from the TG21 and control mice, and equal cell numbers (5 x 105) were plated per well in CollagenCult for CFU-GM (B) and CFU-GEMM (C) colony-forming assays. Both CFU-GM and CFU-GEMM numbers are elevated in TG21 compared with wild-type mice on d 2 and 3 (P < 0.01). Statistical significance was determined by a one-way ANOVA and a post hoc Tukey’s test.
Discussion
We have succeeded in generating several transgenic lines of mice that constitutively express the placental hematopoietic hormone PLP-E. Expression levels of PLP-E vary among these lines, in effect providing a means of assessing dose-dependent biological effects. Based on polyacrylamide gel migration, the number, size, and relative amounts of glycoforms of PLP-E derived from the transgene and secreted by nonplacental tissues into the circulation are indistinguishable from the pattern of placental hormone in the circulation during pregnancy. Thus, the transgenic mice appear to be a valid model for testing the biological effects of PLP-E.
The production of these transgenic lines was intended to test the hypothesis that the restricted expression of PLP-E during pregnancy was crucial to eliminate the broad effects of this potent cytokine on the myeloid lineages outside of pregnancy. This hypothesis envisions pregnancy as a unique physiological state that in many respects resembles a pathophysiological state, including anemia and thrombocytopenia caused by blood volume expansion, and that as a result, pregnancy-specific regulatory factors have evolved to superimpose their activities on the homeostatic systems in place in the nonpregnant animal (13). The placenta provides the perfect endocrine gland for this pregnancy-specific restriction of cytokine expression, because that entire endocrine gland is discarded at the end of pregnancy and is not present in the developing newborn. In the transgenics, however, PLP-E expression would continue and myeloid development in the newborn might as a result be altered.
We analyzed the numbers of the various myeloid blood cell types longitudinally in the transgenic mice, but found no change in any of these numbers compared with control, nontransgenic siblings. Thus, the presence of PLP-E outside of pregnancy is insufficient to cause elevated platelet, erythrocyte, granulocyte, or macrophage levels. Two explanations seem most probable. First, hematopoiesis is controlled both by inducers and inhibitors of myeloid cell proliferation and differentiation (14); the presence of PLP-E may trigger a feedback mechanism that results in greater inhibitory activity, thereby maintaining the hematopoietic set points. Second, PLP-E by itself stimulates megakaryocyte differentiation but is dependent on the presence of other cytokines to enhance myeloid progenitor proliferation (1, 4). If the activity of other positive-acting cytokines is limiting, PLP-E may therefore not be able to bring about an expansion of cell numbers.
Although constitutive PLP-E does not result in altered hematopoiesis in the nonstressed state, we do observe an effect when platelet levels fall. Induction of a transient thrombocytopenia by injection of neuraminidase into nontransgenic mice is followed by a gradual recovery of platelet numbers over a 5-d period. In these mice, PLP-E expression is observed in the bone marrow, but expression is detected only after approximately 3 d (7). The same recovery time course is seen in the transgenic lines that express moderate levels of PLP-E. In the two transgenic lines that have high levels of PLP-E, in contrast, platelet levels return to normal more rapidly. Consistent with this observation, bone marrow progenitor cells in the megakaryocyte lineage, first the multipotential myeloid progenitor CFU-GEMM and later the lineage-restricted progenitor CFU-MK, more rapidly increase in numbers in the transgenic high expressers. Importantly, these effects were specific, as no change is seen in the numbers of erythrocytes and their BFU-E progenitors, and no change is seen in total white blood cell numbers. Thus, constitutive expression of PLP-E at high levels eliminates the delay in production of this hormone in response to low platelet levels, and an important consequence is a more rapid recovery. These findings, coupled with the observation that the PLP-E gene is activated in bone marrow cells in response to thrombocytopenia (7), suggest that PLP-E may normally be a limiting factor in platelet production in this disease model. Because PLP-E acts synergistically with other cytokines to enhance progenitor growth, several cytokines may be rate limiting for recovery. Indeed, other cytokines have also been found to enhance recovery from thrombocytopenia in mice (15, 16, 17, 18).
Although the result with the high-expressing lines supports a role for endogenous PLP-E expression in the bone marrow, the finding that the two transgenic lines expressing comparatively moderate levels of PLP-E do not recover more rapidly from acute thrombocytopenia suggests that pharmacological levels of PLP-E may be needed to stimulate recovery. Alternatively, it may be that circulating levels of PLP-E in these transgenic mice are not indicative of bone marrow levels of expression of the transgene, or that the distribution of PLP-E-expressing cells for the transgene and the endogenous gene differ in the bone marrow.
In contrast to the results in recovery from thrombocytopenia, and despite the effects of PLP-E in cell culture on BFU-E growth (our manuscript in preparation), no difference is seen between high-expressing PLP-E transgenics and wild-type mice in the kinetics of recovery of circulating red blood cell levels from phenylhydrazine-induced anemia. The simplest explanation is that some factor other than PLP-E is rate limiting for recovery and that high levels of PLP-E are not able to substitute for the limiting activity that is required for restoration of normal erythrocyte concentrations. Notably, neither of the myeloid progenitors, CFU-GEMM and BFU-E, increases in number in the transgenics compared with wild-type mice under conditions of anemia, even though this hormone does stimulate CFU-GEMM growth during thrombocytopenia. As shown previously (1, 2), the effect of PLP-E on progenitor cell proliferation requires the presence of one or more other cytokines, consistent with the interpretation that another cytokine is limiting during recovery from anemia. It is possible that PLP-E would have a positive effect on the rate of recovery from anemia when present in combination with erythropoietin and other cytokines, as has been seen for erythropoiesis in the mouse in response to other combinatorial treatments (19, 20, 21, 22).
Similar to the recovery from thrombocytopenia (and distinct from the recovery from anemia), the return to normal neutrophil levels after induction of neutropenia occurs more rapidly in PLP-E transgenics than in wild-type mice, as has been seen with other cytokine treatments (23, 24, 25, 26). Preceding and accompanying the faster rise in circulating neutrophils are significantly greater increases in the transgenic bone marrow populations of CFU-GEMM and CFU-GM. Thus, PLP-E can contribute a rate-limiting activity at the level of progenitor expansion during the recovery period in this disease model. In contrast to the observed expression of PLP-E in the bone marrow in response to low blood platelet levels, we have not been able to detect PLP-E mRNA in the bone marrow under conditions of neutropenia. Thus, PLP-E may be able to play a pharmacologically significant role in the recovery from neutropenia and a physiologically significant role in the recovery from thrombocytopenia.
Acknowledgments
We thank Weimin Song for assistance in generating the transgenic mice and Sumit Bhattacharyya for assistance in collecting blood samples.
References
Lin J, Linzer DIH 1999 Induction of megakaryocyte differentiation by a novel pregnancy-specific hormone. J Biol Chem 274:21485–21489
Zhou B, Lum HE, Lin J, Linzer DIH 2002 Two placental hormones are agonists in stimulating megakaryocyte growth and differentiation. Endocrinology 143:4281–4286
Bittorf T, Jaster R, Soares MJ, Seiler J, Brock J, Friese K, Müller H 2000 Induction of erythroid proliferation and differentiation by a trophoblast-specific cytokine involves activation of the JAK/STAT pathway. J Mol Endocrinol 25:253–262
Lefebvre P, Lin J, Linzer DIH, Cohen I 2001 Murine prolactin-like protein E synergizes with human thrombopoietin to stimulate expansion of human megakaryocytes and their precursors. Exp Hematol 29:51–58[CrossRef]
Lin J, Poole JC, Linzer DIH 1997 Two novel members of the PRL/GH family are expressed in the mouse placenta. Endocrinology 138:5535–5540
Wiemers DO, Shao L, Ain R, Dai G, Soares MJ 2003 The mouse prolactin gene family locus. Endocrinology 144:313–325
Bhattacharyya S, Lin J, Linzer DIH 2002 Reactivation of a hematopoietic endocrine program or pregnancy contributes to recovery from thrombocytopenia. Mol Endocrinol 16:1386–1393
Stenberg PE, Levin J, Baker G, Mok Y, Corash L 1991 Neuraminidase-induced thrombocytopenia in mice: effects on thrombopoiesis. J Cell Physiol 147:7–16[CrossRef]
Krystal G 1983 A simple microassay for erythropoietin based on 3H-thymidine incorporation into spleen cells from phenylhydrazine treated mice. Exp Hematol 11:649–660
Canonne-Hergaux F, Zhang AS, Ponka P, Gros P 2001 Characterization of the iron transporter DMT1 (NRAMP2/DCT1) in red blood cells of normal and anemic mk/mk mice. Blood 98:3823–3830
Vannucchi AM, Bianchi L, Cellai C, Paoletti F, Carrai V, Calzolari A, Centurione L, Lorenzini R, Carta C, Alfani E, Sanchez M, Migliaccio G, Migliaccio AR 2001 Accentuated response to phenylhydrazine and erythropoietin in mice genetically impaired for their GATA-1 expression (GATA-1 (low) mice). Blood 97:3040–3050
Czuprynski CJ, Brown JF, Wagner RD, Steinberg H 1994 Administration of antigranulocyte monoclonal antibody RB6–8C5 prevents expression of acquired resistance to Listeria monocytogenes infection in previously immunized mice. J Immunol 152:1836–1846
Linzer DIH, Fisher SJ1999 The placenta and the prolactin family of hormones: regulation of the physiology of pregnancy. Mol Endocrinol 13:837–840.
Ogawa M 1993 Differentiation and proliferation of hematopoietic stem cells. Blood 81:2844–2853
Tsuji-Takayama K, Harashima A, Tahata H, Izumi N, Nishida Y, Namba M, Okuda Y, Ohta T, Kurimoto M 1996 IFN- in combination with IL-3 accelerates platelet recovery in mice with 5-fluorouracil-induced marrow aplasia. J Interferon Cytokine Res 16:447–451
Akahori H, Shibuya K, Ozai M, Ida M, Kabaya K, Kato T, Miyazaki H 1996 Effects of pegylated recombinant human megakaryocyte growth and development factor on thrombocytopenia induced by a new myelosuppressive chemotherapy regimen in mice. Stem Cells 14:678–689
Kovacs CJ, Powell DS, Evans MJ, Thomas-Patterson D, Johnke RM 1996 Enhanced platelet recovery in myelosuppressed mice treated with interleukin-1 and macrophage colony-stimulating factor: potential interactions with cytokines having megakaryocyte colony-stimulating activity. J Interferon Cytokine Res 16:187–194
Akiyama Y, Kikuchi Y, Matsuzaki J, Kajimura N, Ohue C, Yamaguchi K 1997 Protective effect of recombinant human leukemia inhibitory factor against thrombocytopenia in carboplatin-treated mice. Jpn J Cancer Res 88:584–589
de Haan G, Dontje B, Engel C, Loeffler M, Nijhof W 1995 In vivo effects of interleukin-11 and stem cell factor in combination with erythropoietin in the regulation of erythropoiesis. Br J Haematol 90:783–790
de Haan G, Dontje B, Nijhof W, Loeffler M 1995 Effects of continuous stem cell factor administration on normal and erythropoietin-stimulated murine hemopoiesis: experimental results and model analysis. Stem Cells 13:65–76
Brox AG, Zhang F, Guyda H, Gagnon RF 1996 Subtherapeutic erythropoietin and insulin-like growth factor-1 correct the anemia of chronic renal failure in the mouse. Kidney Int 50:937–943
Roeder I, de Haan G, Engel C, Nijhof W, Dontje B, Loeffler M 1998 Interactions of erythropoietin, granulocyte colony-stimulating factor, stem cell factor, and interleukin-11 on murine hematopoiesis during simultaneous administration. Blood 91:3222–3229
Shimamura M, Takigawa T, Urabe A, Okabe T, Souza LM, Takaku F 1988 Synergistic effect of dolichyl phosphate and human recombinant granulocyte colony-stimulating factor on recovery from neutropenia in mice treated with anti-cancer drugs. Exp Hematol 16:681–685
Fibbe WE, van der Meer JW, Falkenburg JH, Hamilton MS, Kluin PM, Dinarello CA 1989 A single low dose of human recombinant interleukin 1 accelerates the recovery of neutrophils in mice with cyclophosphamide-induced neutropenia. Exp Hematol 17:805–808
Sagi O, Witz, IP, Shaked N, Ramot B, Douer D 1989 Effect in vivo of recombinant GM-CSF on neutropenia and survival in mice treated by high-dose melphalan. Eur J Haematol 43:240–244
Grossmann A, Lenox J, Deisher TA, Ren HP, Humes JM, Kaushansky K, Sprugel KH 1996 Synergistic effects of thrombopoietin and granulocyte colony-stimulating factor on neutrophil recovery in myelosuppressed mice. Blood 88:3363–3370(Beiyan Zhou, Xiangrong Ko)
Address all correspondence and requests for reprints to: Daniel Linzer, Weinberg College of Arts and Sciences Dean’s Office, 1918 Sheridan Road, Evanston, Illinois 60208. E-mail: dlinzer@northwestern.edu.
Abstract
Expression of the placental hormone, prolactin-like protein E (PLP-E), a potent cytokine that acts on multiple myeloid lineages, is normally restricted to pregnancy and certain hematopoietic disease states. We hypothesized that the restricted pattern of PLP-E expression is necessary to avoid hyperstimulation of myelopoiesis. To test this idea, we have produced PLP-E transgenic mice and analyzed their steady-state blood cell levels. We find that blood cell levels remain in the normal range, and thus the constitutive expression of a cytokine of pregnancy fails to overcome the tight control of hematopoietic set points for blood cell levels. In contrast, an effect of constitutive PLP-E expression is detected during the recovery from low blood platelet levels (acute thrombocytopenia) and from low granulocyte levels (acute neutropenia) but not from anemia. Mice producing high circulating concentrations of PLP-E recover more rapidly from both thrombocytopenia and neutropenia, as seen both by an earlier increase of progenitor numbers in the bone marrow and the earlier return to normal circulating blood cell levels.
Introduction
BLOOD PLATELETS ARE the end product of the pathway of megakaryocyte proliferation and differentiation. A broad array of cytokines act on the steps leading from the initial cell committed to the megakaryocyte lineage, colony-forming unit megakaryocyte (CFU-MK), through the terminal differentiation of the mature megakaryocyte. Most of these cytokines are constitutively present and include thrombopoietin, IL-3, IL-6, and stem cell factor. Two other cytokines that act on megakaryocyte progenitors are produced specifically by the placenta in mice and thus are present during pregnancy in the female but otherwise are not present in healthy female or male mice. These two hormones, prolactin-like protein (PLP)-E and PLP-F, by themselves stimulate megakaryocyte differentiation but not CFU-MK proliferation. In contrast, in the presence of other cytokines, PLP-E and PLP-F contribute a synergistic stimulation of CFU-MK growth (1, 2).
The effects of PLP-E are not restricted to the megakaryocyte lineage. This hormone has been found to stimulate proliferation and differentiation of erythroid cell lines (3). In experiments using myeloid progenitors from human bone marrow, mouse PLP-E was shown to act synergistically with thrombopoietin to enhance growth of CFU-MK, burst-forming unit erythroid (BFU-E) progenitors, and the common myeloid progenitor, CFU-granulocyte/erythroid/macrophage/megakaryocyte (CFU-GEMM) (4). Recently, we have found that PLP-E and PLP-F act on all mouse myeloid progenitors, including CFU-granulocyte/macrophage (CFU-GM), BFU-E, and CFU-GEMM (our manuscript in preparation).
PLP-E and PLP-F are closely related: the genes encoding these two proteins are adjacent on mouse chromosome 13 (5, 6); the secreted proteins share 54% identity in amino acid sequence (5); these hormones share the same receptor on megakaryocytes; and they function as agonists (2). Despite these similarities, the patterns of PLP-E and PLP-F synthesis display some striking differences. Transcription of the PLP-E gene initiates several days earlier in gestation than for the PLP-F gene, and the two genes are active in distinct cell types: the polyploid trophoblast giant cells for PLP-E and the diploid spongiotrophoblasts for PLP-F (5). Furthermore, the PLP-E gene can also become active in bone marrow cells in response to severely low platelet levels (thrombocytopenia), whereas the PLP-F gene appears to remain silent under this condition (7).
The production of PLP-E in bone marrow cells during thrombocytopenia suggests a functional role for this hormone in the recovery of platelet levels from the disease state. Indeed, exogenous administration of PLP-E during thrombocytopenia hastens recovery (7). We therefore decided to explore whether mice constitutively expressing high levels of PLP-E would be protected from thrombocytopenia. Our finding that PLP-E acts on all myeloid progenitor populations (manuscript in preparation) suggests that the constitutive expression of this hormone may also hasten recovery from other blood cell diseases, in particular anemia and neutropenia. Furthermore, by expressing PLP-E endogenously, delivery of this cytokine to the bone marrow compartment could be achieved more readily and continuously compared with any injection route.
By producing such mice, we would also be in a position to examine whether the presence of high levels of PLP-E continuously in the adult would alter the hematopoietic set point for platelet levels or for other myeloid lineages. Alterations in blood cell levels would be consistent with the restricted expression of this hormone during pregnancy and hematopoietic disease, both of which represent physiological states of significant stress. Conceivably, restricting PLP-E gene expression in this way would enable this hormone to enhance survival in stress situations while preventing PLP-E from stimulating excessive hematopoiesis under normal conditions.
The approach we took was to create transgenic mice in which transcription of the PLP-E cDNA is driven by the mouse metallothionein I (MTI) gene promoter to achieve expression at moderately high levels in a broad distribution of tissues. Importantly, we had shown that the PLP-E receptor is present on megakaryocytes of nonpregnant female and male mice (7), so myeloid cells should be capable of responding to circulating PLP-E in transgenic mice. We have generated several transgenic lines and have analyzed blood cell levels under normal conditions and after induction of blood cell disease.
Materials and Methods
PLP-E transgenic mice
An expression construct containing the MTI promoter (kindly provided by Kelly Mayo) followed by the full-length PLP-E cDNA and the SV40 splice/polyadenylation signals was generated. The DNA fragment containing the MTI promoter, PLP-E cDNA including its leader sequence, and SV40 splice and polyadenylation signals was released from the vector by restriction enzyme digestion, purified by GenElute Minus EtBr Spin Columns (Sigma Chemical Co., St. Louis, MO) after gel electrophoresis, and injected into fertilized mouse eggs. Embryos were transferred into the oviduct of pseudopregnant foster mothers later that day. Tail DNA from the F0 progeny were examined by PCR for integration of the transgene into mouse genomic DNA, and Western blotting of serum was used to assay for the accumulation of PLP-E protein in the circulation. Transgenic mice and control mice were maintained on 14-hr light, 10-hr dark cycles with food and water freely available. All procedures were conducted with the approval of the Northwestern University Animal Care and Use Committee.
DNA extraction, PCR, and Western blotting
Genomic DNA was extracted from mouse tails using a standard protocol. In brief, mouse tail tissues were digested with proteinase K at 55 C overnight. Samples were then extracted with phenol:chloroform (1:1) followed by two extractions with chloroform alone. Genomic DNA was precipitated in 300 mM sodium acetate, 95% ethanol. Pellets were washed twice with 70% ethanol, resuspended in distilled water or 10 mM Tris-HCl, 1 mM EDTA (pH 7.4), and stored at –20 C.
Primers to detect the insertion of the transgene were designed from the MTI promoter (upstream) and PLP-E (downstream) sequences: primer 1, 5'-GTCCCGCTGTGCACACTGGCGCTC-3'; primer 2, 5'-CCTCAAGAACTGGAGATAAAA-3'.
Genomic DNA from wild-type CD-1 mice was used as negative control and L19 ribosomal RNA as a positive internal control.
For Western analysis, blood samples were collected from the tails of PLP-E transgenic mice and nontransgenic littermates. A rabbit polyclonal antiserum prepared against glutathione-S-transferase-PLP-E was used to detect PLP-E in these samples. Serum from d-10.5 pregnant mice, the sample with peak levels of PLP-E during gestation, was used as the positive control.
Blood diseases and blood cell counting
Blood samples were collected from the retroorbital plexus from PLP-E transgenic mice and nontransgenic siblings at several developmental stages. Platelet, erythrocyte, neutrophil, and total white blood cell levels were measured by the Clinical Pathology Facility, Center for Comparative Medicine at the Feinberg School of Medicine, Northwestern University. To induce thrombocytopenia, neuraminidase (Clostridium perfringens type VI, Sigma) in sterile 150 mM NaCl was injected ip at 0.05 U/mouse (8). To induce anemia, mice were injected with phenylhydrazine at 80 mg/kg ip on day 0 and injected 24 h later with 40 mg/kg ip (9, 10, 11). For neutropenia, mice received a single injection of purified monoclonal antibody Rb6-8C5 at 0.15 mg/mouse ip (12).
Bone marrow cell cultures
Femurs from 16-wk-old transgenic mice and wild-type CD-1 control mice were flushed with Iscove’s modified Dulbecco’s medium (IMDM) and 10% fetal bovine serum (Life Technologies, Rockville, MD), and collected material was passed through 21-gauge and then 25-gauge needles to separate cells. After 1 h in IMDM at 37 C, nonadherent cells were collected by centrifugation and washed with PBS and resuspended with IMDM and 1% Nutridoma (Roche Molecular Biochemicals, Indianapolis, IN). These cells were then cultured in MegaCult C medium (Stem Cells Technology, Vancouver, Canada) for CFU-MK colony-forming assays or CollagenCult medium for CFU-GEMM, CFU-GM, and BFU-E colony-forming assays. Initial densities were 5 x 105 nucleated bone marrow cells per well in two-well chamber slides.
For colony formation and differentiation assays, cultures were maintained for 7–10 d for CFU-MK, 10–14 d for CFU-GEMM and CFU-GM, and 8 d for BFU-E. Slides were dried and stained for colony detection and counting. Acetylcholinesterase activity was used as surface marker for MK cells, and MK colonies were defined as clusters of at least three cells staining positive for acetylcholinesterase. To score CFU-GEMM, CFU-GM, and BFU-E colonies, slides were stained with May-Grünwald and Giemsa, and colonies were defined based on their morphological characteristics.
Results
PLP-E transgenic mice
The MTI promoter linked to the PLP-E cDNA was microinjected into fertilized eggs to generate transgenic mice. Preliminary transient transfection studies had demonstrated that this construct results in the synthesis and secretion of mature and functional PLP-E protein (data not shown). Neonates resulting from these eggs were screened by PCR of genomic DNA for integration of the transgene, and Western blotting of serum was used to identify the presence of PLP-E in the circulation. Four PLP-E transgenic lines were established: TG6, TG7, TG14, and TG21. TG6 and TG7 display moderate levels of circulating PLP-E, whereas TG14 and TG21 have high levels of expression of the transgene (Fig. 1).
FIG. 1. Circulating PLP-E in transgenic mice. Serum samples were collected from wild-type and transgenic mice. Equal amounts of protein were fractionated by SDS -PAGE, transferred to a membrane, and probed with anti-PLP-E antiserum. Numbers on the left indicate the position of molecular weight markers. Serum from d-10.5 pregnant mice (d10) was used as a positive control, and serum from wild-type CD-1 mice (wt) was used as a negative control. The four transgenic lines 6, 7, 14, and 21 show increasing levels of PLP-E accumulation.
Hematology of PLP-E transgenic mice
To determine whether the constitutive presence of PLP-E results in the chronic alteration of blood cell concentrations, blood samples of transgenic mice and nontransgenic littermates were assayed from wk 3 to wk 15 of age. No difference was seen at any time point between these groups in terms of platelet, erythrocyte, or total white blood cell levels (Fig. 2). Thus, continuous exposure to PLP-E beyond gestation does not reset steady-state blood cell levels.
FIG. 2. Hematology analysis of transgenic mice. Blood samples were collected from TG6, TG7, TG14, TG21, and wild-type mice (n 6) at different time points: wk 3, 7, 11, and 15 after birth. Platelet, erythrocyte, and white blood cell levels were measured. Platelet levels (A) from transgenic mice show no increase compared with wild-type mice at any time point. Red blood cell (RBC) levels (B) and white blood cell (WBC) numbers (C) from four transgenic mice groups show no significant difference compared with wild-type mice at all time points. Statistical significance was determined by a one-way ANOVA and a post hoc Tukey’s test.
In addition to monitoring blood cells in the circulation, myeloid progenitor populations in the bone marrow of transgenic and nontransgenic controls were also compared. Primary bone marrow cells were cultured in medium containing IL-3 at 10 ng/ml to stimulate proliferation and colony formation. After 7–14 d of incubation, the numbers of CFU-MK, CFU-GM, BFU-E, and CFU-GEMM colonies were counted. At least two mice from each transgenic line were tested (Fig. 3). Again, no differences were found.
FIG. 3. Myeloid progenitor numbers in transgenic mice. Bone marrow was collected from transgenic lines and age-matched wild-type mice. Two mice were selected from each group, and measurements were made in triplicate. Bone marrow cells were grown in semisolid CollagenCult medium for CFU-GEMM (A), BFU-E (C), and CFU-GM (D); MegaCult C medium was used for CFU-MK (B) assays. CFU-MK colonies were identified by staining for acetylcholinesterase activity, and CFU-GEMM, CFU-GM, and BFU-E were stained by May-Grünwald and Giemsa solutions and scored based on their morphological properties. Statistical significance was determined by a one-way ANOVA and a post hoc Tukey’s test.
Recovery from thrombocytopenia
We had previously noted that the PLP-E gene becomes active in the bone marrow a few days after the onset of thrombocytopenia and that injection of PLP-E into thrombocytopenic mice hastens the recovery to normal platelet levels (7). If PLP-E is indeed a limiting factor in the recovery period, these findings suggest that mice constitutively expressing PLP-E would recover faster than wild-type mice. Thrombocytopenia was therefore induced in transgenic and nontransgenic control mice by single injections of neuraminidase, and blood was collected daily for platelet counts (Fig. 4A). For wild-type mice, complete recovery of platelets requires 4–5 d. The recovery period for the two transgenic lines expressing moderate levels of PLP-E, TG6, and TG7, was indistinguishable from that for the wild-type mice. In contrast, the recovery period was significantly faster in the two transgenic lines, TG14 and TG21, that produce high concentrations of PLP-E. Platelet levels are significantly higher in these transgenic mice by the second day of the recovery period and return to normal by d 3. To verify that the effect of this treatment regime was specific for the platelet compartment, erythrocyte and white blood cell levels were also monitored and were found not to change (data not shown).
FIG. 4. Recovery from thrombocytopenia. Neuraminidase was injected at d 0 into groups of transgenic and wild-type mice (n 3 for each group) to induce thrombocytopenia, and blood samples were collected daily for cell counting. A, Platelet levels in two transgenic strains with high PLP-E expression (TG14 and TG21) are significantly above the controls on d 2 (P < 0.001) and d 3 (P < 0.01). By d 3, platelet levels in TG14 and TG21 mice have returned to normal, whereas wild-type mice and the two transgenic mouse lines with low PLP-E expression recover from thrombocytopenia by d 4–5 after injection. Bone marrows were also collected from TG21 and wild-type mice, and equal cell numbers (5 x 105) were plated per well in MegaCult for CFU-MK (B) or in CollagenCult for CFU-GEMM (C) colony-forming assays. CFU-MK numbers are elevated in the transgenic strain on d 2 (P < 0.01), and CFU-GEMM numbers are higher in the TG21 mice on d 1 and 2 (P < 0.01). Statistical significance (marked by asterisks) was determined by a one-way ANOVA and a post hoc Tukey’s test.
The more rapid recovery of platelet levels in TG14 and TG21 mice predicts that the high PLP-E-expressing transgenic mice would also show a rapid elevation in bone marrow myeloid progenitor numbers. Primary bone marrow cultures were prepared each day after neuraminidase injection from TG21 mice and from wild-type mice. Myeloid progenitor numbers were then determined in colony-forming assays (Fig. 4, B and C). As predicted, levels of the platelet progenitor CFU-MK, as well as the multipotential myeloid progenitor CFU-GEMM, increase more rapidly in TG21 mice. CFU-GEMM numbers increase first, by the first day of recovery, followed by an increase in CFU-MK numbers. These transgenic mice and the control mice were identical in terms of the levels of the myeloid progenitors, BFU-E and CFU-GM, that do not lead to platelet formation (data not shown).
Recovery from anemia
Because PLP-E cooperates with other cytokines to stimulate BFU-E growth, we next explored whether constitutive expression of PLP-E can stimulate recovery from acute anemia. Erythrocyte levels in the transgenic mice and in wild-type control mice were rapidly decreased by a single injection of phenylhydrazine. Blood samples were collected and examined every 24 h over the course of the next 9 d. No change in the rate of recovery is seen comparing PLP-E-expressing transgenic with wild-type mice (Fig. 5A). Furthermore, primary bone marrow cell cultures generated from the transgenic and wild-type mice challenged with phenylhydrazine show no difference in CFU-GEMM or BFU-E numbers (Fig. 5, B and C).
FIG. 5. Recovery from anemia. A, Phenylhdrazine was injected on d 0 into transgenic (TG14 and TG21) and wild-type control mice to induce anemia (n = 4 for each group). Blood samples were collected daily for cell counting. No statistically significant difference in erythrocyte levels between transgenic and wild-type mice is seen at any time point. Bone marrows were also collected from transgenic strain TG21 and wild-type mice at each time point. Equal cell numbers (5 x 105) were plated per well in CollagenCult for BFU-E (B) or CFU-GEMM (C) colony-forming assays. No significant differences are detected between transgenics and controls.
Recovery from neutropenia
Finally, we tested the ability of PLP-E to bring about a more rapid recovery from neutropenia. Again, transgenic and wild-type control mice were injected, this time with monoclonal antibody Rb6-8C5. Neutrophil levels declined quickly after injection, and recovery in the wild-type mice required approximately 5 d (Fig. 6A). In the two high-expressing transgenic strains, neutrophil levels rose more rapidly and were significantly increased relative to control mice on d 2 in the TG14 strain and on d 3 in both the TG14 and TG21 strains. Consistent with this rapid recovery in the transgenics, bone marrow levels of the progenitor population CFU-GEMM and GFU-GM were elevated in Rb6-8C5 antibody-injected TG21 transgenics on d 2 and 3, respectively, compared with injected wild-type mice (Fig. 6, B and C).
FIG. 6. Recovery from neutropenia. Transgenic (TG14 and TG21) mice and wild-type controls were injected with monoclonal antibody Rb6-8C5 to induce neutropenia, and blood samples were collected daily for cell counting (n = 4 for each group). Neutrophil levels in the TG14 mice are significantly elevated (asterisks) compared with controls on d 2 and 3 (P < 0.01) and in the TG21 mice on d 3 and 4 (P < 0.01). By d 5, the number of circulating neutrophils in transgenic and wild-type groups are not statistically different from those at d 0. Bone marrows were also collected from the TG21 and control mice, and equal cell numbers (5 x 105) were plated per well in CollagenCult for CFU-GM (B) and CFU-GEMM (C) colony-forming assays. Both CFU-GM and CFU-GEMM numbers are elevated in TG21 compared with wild-type mice on d 2 and 3 (P < 0.01). Statistical significance was determined by a one-way ANOVA and a post hoc Tukey’s test.
Discussion
We have succeeded in generating several transgenic lines of mice that constitutively express the placental hematopoietic hormone PLP-E. Expression levels of PLP-E vary among these lines, in effect providing a means of assessing dose-dependent biological effects. Based on polyacrylamide gel migration, the number, size, and relative amounts of glycoforms of PLP-E derived from the transgene and secreted by nonplacental tissues into the circulation are indistinguishable from the pattern of placental hormone in the circulation during pregnancy. Thus, the transgenic mice appear to be a valid model for testing the biological effects of PLP-E.
The production of these transgenic lines was intended to test the hypothesis that the restricted expression of PLP-E during pregnancy was crucial to eliminate the broad effects of this potent cytokine on the myeloid lineages outside of pregnancy. This hypothesis envisions pregnancy as a unique physiological state that in many respects resembles a pathophysiological state, including anemia and thrombocytopenia caused by blood volume expansion, and that as a result, pregnancy-specific regulatory factors have evolved to superimpose their activities on the homeostatic systems in place in the nonpregnant animal (13). The placenta provides the perfect endocrine gland for this pregnancy-specific restriction of cytokine expression, because that entire endocrine gland is discarded at the end of pregnancy and is not present in the developing newborn. In the transgenics, however, PLP-E expression would continue and myeloid development in the newborn might as a result be altered.
We analyzed the numbers of the various myeloid blood cell types longitudinally in the transgenic mice, but found no change in any of these numbers compared with control, nontransgenic siblings. Thus, the presence of PLP-E outside of pregnancy is insufficient to cause elevated platelet, erythrocyte, granulocyte, or macrophage levels. Two explanations seem most probable. First, hematopoiesis is controlled both by inducers and inhibitors of myeloid cell proliferation and differentiation (14); the presence of PLP-E may trigger a feedback mechanism that results in greater inhibitory activity, thereby maintaining the hematopoietic set points. Second, PLP-E by itself stimulates megakaryocyte differentiation but is dependent on the presence of other cytokines to enhance myeloid progenitor proliferation (1, 4). If the activity of other positive-acting cytokines is limiting, PLP-E may therefore not be able to bring about an expansion of cell numbers.
Although constitutive PLP-E does not result in altered hematopoiesis in the nonstressed state, we do observe an effect when platelet levels fall. Induction of a transient thrombocytopenia by injection of neuraminidase into nontransgenic mice is followed by a gradual recovery of platelet numbers over a 5-d period. In these mice, PLP-E expression is observed in the bone marrow, but expression is detected only after approximately 3 d (7). The same recovery time course is seen in the transgenic lines that express moderate levels of PLP-E. In the two transgenic lines that have high levels of PLP-E, in contrast, platelet levels return to normal more rapidly. Consistent with this observation, bone marrow progenitor cells in the megakaryocyte lineage, first the multipotential myeloid progenitor CFU-GEMM and later the lineage-restricted progenitor CFU-MK, more rapidly increase in numbers in the transgenic high expressers. Importantly, these effects were specific, as no change is seen in the numbers of erythrocytes and their BFU-E progenitors, and no change is seen in total white blood cell numbers. Thus, constitutive expression of PLP-E at high levels eliminates the delay in production of this hormone in response to low platelet levels, and an important consequence is a more rapid recovery. These findings, coupled with the observation that the PLP-E gene is activated in bone marrow cells in response to thrombocytopenia (7), suggest that PLP-E may normally be a limiting factor in platelet production in this disease model. Because PLP-E acts synergistically with other cytokines to enhance progenitor growth, several cytokines may be rate limiting for recovery. Indeed, other cytokines have also been found to enhance recovery from thrombocytopenia in mice (15, 16, 17, 18).
Although the result with the high-expressing lines supports a role for endogenous PLP-E expression in the bone marrow, the finding that the two transgenic lines expressing comparatively moderate levels of PLP-E do not recover more rapidly from acute thrombocytopenia suggests that pharmacological levels of PLP-E may be needed to stimulate recovery. Alternatively, it may be that circulating levels of PLP-E in these transgenic mice are not indicative of bone marrow levels of expression of the transgene, or that the distribution of PLP-E-expressing cells for the transgene and the endogenous gene differ in the bone marrow.
In contrast to the results in recovery from thrombocytopenia, and despite the effects of PLP-E in cell culture on BFU-E growth (our manuscript in preparation), no difference is seen between high-expressing PLP-E transgenics and wild-type mice in the kinetics of recovery of circulating red blood cell levels from phenylhydrazine-induced anemia. The simplest explanation is that some factor other than PLP-E is rate limiting for recovery and that high levels of PLP-E are not able to substitute for the limiting activity that is required for restoration of normal erythrocyte concentrations. Notably, neither of the myeloid progenitors, CFU-GEMM and BFU-E, increases in number in the transgenics compared with wild-type mice under conditions of anemia, even though this hormone does stimulate CFU-GEMM growth during thrombocytopenia. As shown previously (1, 2), the effect of PLP-E on progenitor cell proliferation requires the presence of one or more other cytokines, consistent with the interpretation that another cytokine is limiting during recovery from anemia. It is possible that PLP-E would have a positive effect on the rate of recovery from anemia when present in combination with erythropoietin and other cytokines, as has been seen for erythropoiesis in the mouse in response to other combinatorial treatments (19, 20, 21, 22).
Similar to the recovery from thrombocytopenia (and distinct from the recovery from anemia), the return to normal neutrophil levels after induction of neutropenia occurs more rapidly in PLP-E transgenics than in wild-type mice, as has been seen with other cytokine treatments (23, 24, 25, 26). Preceding and accompanying the faster rise in circulating neutrophils are significantly greater increases in the transgenic bone marrow populations of CFU-GEMM and CFU-GM. Thus, PLP-E can contribute a rate-limiting activity at the level of progenitor expansion during the recovery period in this disease model. In contrast to the observed expression of PLP-E in the bone marrow in response to low blood platelet levels, we have not been able to detect PLP-E mRNA in the bone marrow under conditions of neutropenia. Thus, PLP-E may be able to play a pharmacologically significant role in the recovery from neutropenia and a physiologically significant role in the recovery from thrombocytopenia.
Acknowledgments
We thank Weimin Song for assistance in generating the transgenic mice and Sumit Bhattacharyya for assistance in collecting blood samples.
References
Lin J, Linzer DIH 1999 Induction of megakaryocyte differentiation by a novel pregnancy-specific hormone. J Biol Chem 274:21485–21489
Zhou B, Lum HE, Lin J, Linzer DIH 2002 Two placental hormones are agonists in stimulating megakaryocyte growth and differentiation. Endocrinology 143:4281–4286
Bittorf T, Jaster R, Soares MJ, Seiler J, Brock J, Friese K, Müller H 2000 Induction of erythroid proliferation and differentiation by a trophoblast-specific cytokine involves activation of the JAK/STAT pathway. J Mol Endocrinol 25:253–262
Lefebvre P, Lin J, Linzer DIH, Cohen I 2001 Murine prolactin-like protein E synergizes with human thrombopoietin to stimulate expansion of human megakaryocytes and their precursors. Exp Hematol 29:51–58[CrossRef]
Lin J, Poole JC, Linzer DIH 1997 Two novel members of the PRL/GH family are expressed in the mouse placenta. Endocrinology 138:5535–5540
Wiemers DO, Shao L, Ain R, Dai G, Soares MJ 2003 The mouse prolactin gene family locus. Endocrinology 144:313–325
Bhattacharyya S, Lin J, Linzer DIH 2002 Reactivation of a hematopoietic endocrine program or pregnancy contributes to recovery from thrombocytopenia. Mol Endocrinol 16:1386–1393
Stenberg PE, Levin J, Baker G, Mok Y, Corash L 1991 Neuraminidase-induced thrombocytopenia in mice: effects on thrombopoiesis. J Cell Physiol 147:7–16[CrossRef]
Krystal G 1983 A simple microassay for erythropoietin based on 3H-thymidine incorporation into spleen cells from phenylhydrazine treated mice. Exp Hematol 11:649–660
Canonne-Hergaux F, Zhang AS, Ponka P, Gros P 2001 Characterization of the iron transporter DMT1 (NRAMP2/DCT1) in red blood cells of normal and anemic mk/mk mice. Blood 98:3823–3830
Vannucchi AM, Bianchi L, Cellai C, Paoletti F, Carrai V, Calzolari A, Centurione L, Lorenzini R, Carta C, Alfani E, Sanchez M, Migliaccio G, Migliaccio AR 2001 Accentuated response to phenylhydrazine and erythropoietin in mice genetically impaired for their GATA-1 expression (GATA-1 (low) mice). Blood 97:3040–3050
Czuprynski CJ, Brown JF, Wagner RD, Steinberg H 1994 Administration of antigranulocyte monoclonal antibody RB6–8C5 prevents expression of acquired resistance to Listeria monocytogenes infection in previously immunized mice. J Immunol 152:1836–1846
Linzer DIH, Fisher SJ1999 The placenta and the prolactin family of hormones: regulation of the physiology of pregnancy. Mol Endocrinol 13:837–840.
Ogawa M 1993 Differentiation and proliferation of hematopoietic stem cells. Blood 81:2844–2853
Tsuji-Takayama K, Harashima A, Tahata H, Izumi N, Nishida Y, Namba M, Okuda Y, Ohta T, Kurimoto M 1996 IFN- in combination with IL-3 accelerates platelet recovery in mice with 5-fluorouracil-induced marrow aplasia. J Interferon Cytokine Res 16:447–451
Akahori H, Shibuya K, Ozai M, Ida M, Kabaya K, Kato T, Miyazaki H 1996 Effects of pegylated recombinant human megakaryocyte growth and development factor on thrombocytopenia induced by a new myelosuppressive chemotherapy regimen in mice. Stem Cells 14:678–689
Kovacs CJ, Powell DS, Evans MJ, Thomas-Patterson D, Johnke RM 1996 Enhanced platelet recovery in myelosuppressed mice treated with interleukin-1 and macrophage colony-stimulating factor: potential interactions with cytokines having megakaryocyte colony-stimulating activity. J Interferon Cytokine Res 16:187–194
Akiyama Y, Kikuchi Y, Matsuzaki J, Kajimura N, Ohue C, Yamaguchi K 1997 Protective effect of recombinant human leukemia inhibitory factor against thrombocytopenia in carboplatin-treated mice. Jpn J Cancer Res 88:584–589
de Haan G, Dontje B, Engel C, Loeffler M, Nijhof W 1995 In vivo effects of interleukin-11 and stem cell factor in combination with erythropoietin in the regulation of erythropoiesis. Br J Haematol 90:783–790
de Haan G, Dontje B, Nijhof W, Loeffler M 1995 Effects of continuous stem cell factor administration on normal and erythropoietin-stimulated murine hemopoiesis: experimental results and model analysis. Stem Cells 13:65–76
Brox AG, Zhang F, Guyda H, Gagnon RF 1996 Subtherapeutic erythropoietin and insulin-like growth factor-1 correct the anemia of chronic renal failure in the mouse. Kidney Int 50:937–943
Roeder I, de Haan G, Engel C, Nijhof W, Dontje B, Loeffler M 1998 Interactions of erythropoietin, granulocyte colony-stimulating factor, stem cell factor, and interleukin-11 on murine hematopoiesis during simultaneous administration. Blood 91:3222–3229
Shimamura M, Takigawa T, Urabe A, Okabe T, Souza LM, Takaku F 1988 Synergistic effect of dolichyl phosphate and human recombinant granulocyte colony-stimulating factor on recovery from neutropenia in mice treated with anti-cancer drugs. Exp Hematol 16:681–685
Fibbe WE, van der Meer JW, Falkenburg JH, Hamilton MS, Kluin PM, Dinarello CA 1989 A single low dose of human recombinant interleukin 1 accelerates the recovery of neutrophils in mice with cyclophosphamide-induced neutropenia. Exp Hematol 17:805–808
Sagi O, Witz, IP, Shaked N, Ramot B, Douer D 1989 Effect in vivo of recombinant GM-CSF on neutropenia and survival in mice treated by high-dose melphalan. Eur J Haematol 43:240–244
Grossmann A, Lenox J, Deisher TA, Ren HP, Humes JM, Kaushansky K, Sprugel KH 1996 Synergistic effects of thrombopoietin and granulocyte colony-stimulating factor on neutrophil recovery in myelosuppressed mice. Blood 88:3363–3370(Beiyan Zhou, Xiangrong Ko)