Background

Phosphatase and tensin homolog deleted on chromosome 10 (PTEN) is a well-characterized tumor suppressor gene, and PTEN deficiency is frequently observed in various types of cancers, including brain cancer, breast cancer, prostate cancer, endometrial carcinoma, melanoma, and leukemia. Indeed, PTEN is one of the most commonly mutated genes in human cancers [1–3], and germline mutations in PTEN are considered to be responsible for Cowden syndrome, Bannayan-Zonana syndrome, and Lhermitte-Duclose disease [4]. In addition, PTEN has been shown to regulate a series of fundamental cell behaviors, including cell growth, proliferation, survival, and migration, mainly by suppressing the activity of the PI3K/AKT pathway [5].

Conditional ablation of Pten in the hematopoietic stem cells (HSCs) of adult mice leads to rapid HSC depletion and the formation of leukemia-initiating cells (LICs) [6, 7]. A recent study further showed that the mechanistic target of rapamycin (mTOR) activation upon Pten deletion is critical during rapid HSC depletion [8]. Moreover, mTOR complex 1 (mTORC1) is also required to sustain normal hematopoiesis [9, 10].

CCAAT enhancer-binding protein-α (C/EBPα) is another tumor suppressor that can inhibit cell proliferation [11, 12], and mutations in CEBPA are widely reported in acute myeloid leukemia (AML) patients [13–16]. C/ebpα-deficient mice show a phenotype similar to AML in which the transition from the common myeloid progenitor to the granulocyte/monocyte progenitor is blocked [17, 18]. In fact, C/EBPα plays a vital role in myeloid differentiation by directly enhancing the transcription of many myeloid-specific genes [19–22]. Our previous in vitro study showed some relation between PTEN and CEBPA in HL-60 cell line (derived from a patient with acute promyelocytic leukemia) [23], but their interactive mechanism underlying the early hematopoiesis in vivo still remains elusive.

Over the past two decades, the zebrafish has emerged as an ideal system for studying hematopoiesis owing to its unique advantages, including optical clarity and a high fecundity [24, 25]. Similar to mammals, hematopoiesis in zebrafish consists of two successive waves: primitive and definitive hematopoiesis. Primitive hematopoiesis predominantly produces primitive erythrocytes and macrophages, whereas definitive hematopoiesis gives rise to all mature hematopoietic lineages. In zebrafish, the latter process is initiated in the ventral wall of the dorsal aorta, which is counterpart of the aorta-gonad-mesonephros (AGM) region in mammals. The hematopoietic stem/progenitor cells (HSPCs) derived from the AGM migrate to caudal hematopoietic tissue (CHT), which is similar to the fetal liver in mammals. Lastly, HSPCs colonize the kidney marrow, which is the equivalent of bone marrow in mammals, and the thymus to sustain long-term hematopoiesis throughout adulthood [25–27].

Two pten genes, ptena and ptenb, have been previously identified in zebrafish [28]. Although ptena and ptenb single mutants are able to survive to sexual maturity because of the overlapping functions of these genes [29]. Ptena and ptenb double mutants die approximately 5 days post fertilization (dpf) and exhibit significant hyperplastic-dysplastic phenotypes (enlarged head and heart edema. et al.) [29]. Accordingly, homozygous deletion of Pten in mice is embryonic lethal [30]. Although several elegant reports have established the pivotal role of PTEN in preventing leukemogenesis [6, 7, 10], in vivo evidence for PTEN regulation of hematopoiesis in early development is lacking, and the detailed mechanism underlying this process is still largely unknown.

In this study, we investigated the physiological role of pten signaling in hematopoiesis by utilizing pten mutant zebrafish. We tried to explore the influence of complete loss of pten on primitive hematopoiesis, the developmental process and the innate immune response of myeloid cells in definitive hematopoiesis, and the regulational effects of the PI3K/mTOR pathway involved. Furthermore, we revealed the intriguing function of overexpression of C/ebpα in the hematopoietic defect of pten mutant embryos by acting downstream of the PI3K pathway.

Results

Loss of pten induces abnormal hematopoiesis in zebrafish larvae

To evaluate the role of pten in hematopoiesis, we examined the expression of critical hematopoietic genes in pten-deficient zebrafish by using whole-mount in situ hybridization analysis.

We first examined hematopoiesis in ptena and ptenb single-mutant embryos and found no obvious alteration in hematopoiesis (data not shown). We then examined the hematopoietic phenotypes in ptena and ptenb double-mutant (ptena−/−ptenb−/−, hereafter referred to as pten−/−) embryos, which were derived from an incross of ptena+/−ptenb−/− (hereafter referred to as pten+/−) zebrafish. Both primitive hematopoiesis at 22 hours post-fertilization (hpf) (Additional file 1: Figure S1A-H) and definitive hematopoiesis at 36 hpf and 48 hpf (Additional file 1: Figure S1I-P) were normal in the pten−/− embryos.

However, in comparison to the control embryos, the pten−/− embryos at 90 hpf showed an obvious increase in the number of cmyb- expressing and scl-expressing HSPCs in the CHT (Additional file 1: Figure S2A-D). Of note, a significant increase in the number of α-E1 globin-expressing mature erythrocytes that were ectopically dispersed in the head and yolk sac of the pten−/− embryos was observed, although only subtle expansion was noted in the CHT (Additional file 1: Figure S2E-F), the major region colonized by mature erythrocytes in control embryos.

We observed that definitive myelopoiesis in the pten−/− embryos was significantly increased at 90 hpf, as indicated by the enhanced expression of lyz (a myeloid marker) in both the CHT and kidney region (Figure 1A-B and E), and we also confirmed expansive myelopoiesis in the pten−/−;lyz:EGFP embryos [31] (Figure 1C-D and F). Furthermore, we examined other markers of myeloid cells (l-plastin, mpo, and nephrosin) in addition to lyz and found that their expression levels were also increased in the CHT at 90 hpf (Figure 1G-M), indicating that pten signaling is essential for normal definitive myelopoiesis. To further characterize the developmental process of abnormal myelopoiesis in pten−/− fish, we performed whole-mount in situ hybridization for lyz every 6 hours, from 72 to 108 hpf (Additional file 1: Figure S3A-N and Q). The whole-mount in situ hybridization results showed that abnormal myelopoiesis appeared as early as 78 hpf (Additional file 1: Figure S3A-D) and lasted for approximately 15 hours until the myeloid cells became scattered along the upper edge of the yolk sac at approximately 96 hpf (Additional file 1: Figure S3C-J). Lastly, the myeloid cells of the pten−/− embryos became much more dispersed, and their number decreased to a level comparable to that of the controls (Additional file 1: Figure S3K-N), which is likely due to enhanced apoptosis (Additional file 1: Figure S3O-P and R). However, it should be noticed that apoptosis is not limited to lyz-positive cells (Additional file 1: Figure S3P). This may be due to the rapidly deteriorated health of pten−/− fish with faint heart beating, smaller eyes and enlarged head from 4 to 5 dpf. In fact, pten−/− embryos only had slightly visible abnormality (subtle heart edema) at 90 hpf when most of our assays were performed. Thus, the marked changes at late stage hardly impacted on our observation in myelopoiesis.

Figure 1
figure 1

Definitive myelopoiesis is strongly expansive in pten- deficient embryos. (A-B) Whole-mount in situ hybridization analyses of definitive myelopoiesis with riboprobes for lyz in the indicated embryos at 90 hpf. The lower panels are high-magnification views of the indicated region in the upper panel. The red arrowheads indicate myeloid cells in the CHT and kidney/thymus region. (C-D) The pten−/−;lyz:EGFP embryos show increased EGFP-positive myeloid cells. The red arrowheads indicate EGFP-positive cells in the CHT and kidney region. The embryos are shown in lateral views with the anterior to the left. e: eye. (E-F) Statistical result for a, b and c, d respectively. (G-M) Whole-mount in situ hybridization analyses for other myeloid markers (l-plastin, mpo, nephrosin) in the CHT. The red arrowheads indicate the myeloid cells marked by each probe. The data shown in (E-F) and (M) are the means ± SEM of at least 30 embryos; ***p < 0.001 versus the control. Scale bar: 100 um.

Full size image

Taken together, these results demonstrate that a deficiency in pten signaling induces abnormal hematopoiesis, particularly the abnormal expansion of definitive myeloid cells, in zebrafish.

Expansive myelopoiesis in pten−/− embryos is due to reduced apoptosis and block in myeloid cell maturation

To determine whether the relatively expansive myelopoiesis in pten−/− embryos can be attributed to the enhanced myeloid cell proliferation, we performed double staining in pten−/−;lyz:EGFP embryos using antibodies against phospho-histone 3 (PH3) and EGFP. There was no significant difference in the proliferation of lyz-positive myeloid cells between the pten−/− and control embryos (Additional file 1: Figure S4A-F and M). In addition, no obvious changes in the proliferation of HSPCs was observed in the pten−/− embryos, as evidenced by the co-staining of PH3 and EGFP in pten−/−;cmyb:EGFP larvae (Additional file 1: Figure S4G-L and N).

Notably, the number of myeloid cells in the control embryos at 90 hpf was significantly lower than that at 80 hpf; however, the myeloid cell number in the pten−/− embryos at 90 hpf was almost the same with that at 80 hpf (Additional file 1: Figure S3, the second and the forth value on the dashed line, both of which is about 170). We hypothesized that the expansion of myeloid cells was due to the relative reduction in apoptosis in the pten−/− embryos. To test this possibility, a terminal transferase UTP nick end-labeling (TUNEL) assay was performed at 80 hpf when the lyz-positive cells in pten−/− zebrafish just started to increase. Indeed, fewer apoptotic myeloid cells were observed in the pten−/− embryos compared to the control embryos, suggesting that reduced apoptosis in the pten−/− embryos may partially contribute to the observed expansive myelopoiesis.

Considering that the block in maturation could also contribute to an increased number of myeloid cells, we next examined the morphology of the myeloid cells by performing Wright-Giemsa staining of FACS-purified EGFP-positive cells from pten−/−;lyz:EGFP and control embryos at 90 hpf. The nuclei of the control myeloid cells exhibited a typical kidney shape, which is indicative of mature neutrophils (Figure 2H). However, the nuclei of the pten−/− myeloid cells exhibited a high nucleus-to-cytoplasm ratio and loose chromatin, which are features of metamyelocytes and promonocytes (Figure 2I). These data suggest that neutrophil maturation is blocked at the metamyleocytic stage in the pten−/− embryos.

Figure 2
figure 2

Reduced apoptosis and block in myeloid cell maturation underlie pten deficiency-induced expansive myelopoiesis. (A-G) Double staining of lyz-driven EGFP protein and TUNEL assay in the CHT of control and pten−/− embryos at 80 hpf. Although relative more EGFP-positive myeloid cells were observed in the pten−/− embryos (A and B, all green cells were manually counted as positive), few cells were simultaneously apoptotic, as indicated by merging with TUNEL staining (E and F, white arrowheads). (H and I) Wright-Giemsa staining and morphological characterization of FACS-purified EGFP-positive cells from 90 hpf pten−/−;lyz:EGFP and control embryos. The black arrowheads indicate the nuclear shape, which was typically kidney-like and condensed in control embryos (H) but remained in a loose state in the pten−/− embryos (I). (J-L) Sudan Black staining of neutrophils at 90 hpf. The red arrowheads indicate neutrophils in the CHT. The data shown in (G) and (L) are the means ± SEM of at least 15 and 30 embryos; ***p < 0.001 versus the control. Scale bar: h and i, 5 um; others, 100 um.

Full size image

To further confirm the reduction in mature neutrophils, we performed whole-mount in situ staining utilizing Sudan Black B, which specifically stains the granules of granulocytes [32, 33]. The control embryos typically showed heavily stained neutrophils in the CHT, whereas only lightly stained cells were detected in the pten−/− embryos (Figure 2J-L). This result indicated that the maturation of neutrophils was severely blocked upon pten loss during definitive myelopoiesis.

Collectively, our results suggest that the observed expansive myelopoiesis in pten−/− embryos is due to the inhibition of myeloid maturation and a modest blockage of apoptosis.

Loss of pten impairs the immune response of myeloid cells

Because the immune response of myeloid cells is important in the early development of zebrafish [31, 34–37], we evaluated the ability of these cells to respond to inflammation in pten−/− embryos. Tail transections near the caudal circulatory system of pten−/−;lyz:EGFP and control embryos at 84 hpf were performed as previously described [31, 37]. Six hours later, the directional migration of EGFP-positive cells toward the acute injury induced by the tail transections was observed by fluorescence microscopy. Compared to the control embryos, the number of EGFP-positive cells in the pten−/− embryos that migrated to the injury site was remarkably reduced (Figure 3A-C), particularly in the region close to the wound (Figure 3A-B, rightmost areas of the lower panels). Additionally, more EGFP-positive cells in the region far from the wound were unable to migrate to the injury site (Figure 3B, leftmost areas of the lower panels, Additional file 2: Movie S1 and Additional file 3: Movie S2). Furthermore, the increased total number of EGFP-positive myeloid cells in pten−/− embryos (Figure 3D) enhanced the immune-deficiency of the myeloid cells in pten−/− embryos, with a lower percentage of migration neutrophils (Figure 3E).

Figure 3
figure 3

The immune response of expansive myeloid cells is impaired in pten −/− embryos. (A and B) Tail transection analyses of myeloid cell movement in response to inflammation. The myeloid cells of the control and pten−/− embryos moved to the wound site at 6 hours post-transection (hpt) (90 hpf) (A and B, white boxes). The red arrowheads indicate the myeloid cells away from the wound (A and B, left region of the lower panels) and the myeloid cells that arrived at the wound (A and B, right region of the lower panels). A white asterisk indicates the heart edema of pten−/− embryos. (C-E) Quantification of myeloid cells in the region to the right of the white dashed line (A and B, lower panel) (C) and the total number of myeloid cells (D) in each embryo. (E) The percentage of myeloid cells that migrated to the wound. (F and G) Neutral Red staining of macrophages at 90 hpf. The black arrowheads indicate macrophages in the CHT. (H and I) Whole-mount in situ hybridization analyses of macrophages with riboprobe for c-fms. The black arrowheads indicate macrophages in the CHT. (J and K) Statistical results of F, G and H, I respectively. The data shown are the means ± SEM of at least 30 embryos; ***p < 0.001 versus the control. Scale bar: 100 um.

Full size image

To specifically assess the immune response of macrophages in the pten−/− embryos, we performed Neutral Red staining by incubating zebrafish larvae in staining buffer from 80 to 90 hpf. Intriguingly, more heavily stained macrophages were observed in the CHT of pten−/− embryos compared to the control embryos, suggesting that the pten-deficient macrophages take up relatively more Neutral Red (Figure 3F-G and J). We also performed whole-mount in situ hybridization with c-fms, a specific marker for macrophages. Our result indicated that the number of macrophages was greatly increased in pten−/− embryos compared to controls (Figure 3H-I and K), which is consistent with our Neutral Red result. These data suggested that relatively more macrophages were produced upon pten loss.

Collectively, these results demonstrate that the normal immune response of myeloid cells is severely impaired during definitive myelopoiesis in pten−/− zebrafish embryos.

The PI3K/mTOR pathway contributes to pten deficiency-induced dysmyelopoiesis

It is well known that PTEN plays a major role in the regulation of cell survival, metabolism, and migration by negatively regulating PI3K signaling [4]. We therefore hypothesized that the expansive myelopoiesis in pten−/− fish may be rescued by treatment with specific inhibitors of PI3K signaling, such as LY294002. To test this hypothesis, pten−/− embryos were treated with LY294002 from 78 hpf to 90 hpf at concentrations of 5, 15, and 30 μM. Whole-mount in situ hybridization analyses showed that LY294002 rescued expansive myelopoiesis in both the CHT and kidney region in a dose-dependent manner (Figure 4A-F and M). Moreover, the pten−/− embryos treated with 30 μM LY294002 displayed fewer myeloid cells than the controls (Figure 4A and F).

Figure 4
figure 4

The PI3K/mTOR pathway contributes to the expansion of definitive myelopoiesis in pten −/− embryos. (A-L) Whole-mount in situ hybridization analyses of lyz- positive myeloid cells in embryos treated with increasing doses of LY294002 (LY) or rapamycin (Ra) at 90 hpf. Untreated and mock-treated pten−/− embryos (B, C and H, I) showing expansive myelopoiesis compared to the controls (A and G), as mentioned above. The pten−/− embryos treated with stepwise increases in LY294002 or rapamycin showed an obvious rescue effect, as indicated by the decrease in myeloid cells (D-F and I-L). The red arrowheads indicate the main positions of the myeloid cells. (M and N) Quantification of myeloid cells in each embryo (a-l). The data shown are the means ± SEM of at least 30 embryos, *p < 0.05, **p < 0.01, ***p < 0.001 versus the corresponding controls. Scale bar: 100 um.

Full size image

Because mTOR signaling could be activated by the activation of the PI3K pathway [23, 38, 39], we also treated pten−/− fish with rapamycin, a specific inhibitor of mTOR. A similar dose-dependent rescue of expansive myelopoiesis was also observed (Figure 4G-L and N).

To further explore whether the suppression of PI3K/mTOR pathway could also abolish the block of neutrophil maturation upon pten loss, we performed Sudan black assay on LY294002- and rapamycin-treated pten−/− embryos (Figure 5A-F). As we speculated, the maturation of neutrophils was largely recovered in LY294002- and rapamycin-treated pten−/− embryos (Figure 5D and F) compared to mock-treated ones (Figure 5A and G).

Figure 5
figure 5

The PI3K/mTOR pathway contributes to the blockage of myeloid cell maturation in pten −/− embryos. (A-F) Sudan Black staining of neutrophils at 90 hpf. Control and pten−/− embryos were treated with DMSO (A-B) or LY294002/Rapamycin (C-D and E-F) at indicated concentration. The red arrowheads indicate neutrophils in the CHT. (G) statistical results of A-F. The data shown in (G) are the means ± SEM of at least 30 embryos; ***p < 0.001 versus the control. Scale bar: 100 um.

Full size image

We have shown that the maturation of neutrophils was blocked upon pten loss, which may imply that the rescue effects of LY294002 and rapamycin occurred in a cell-autonomous manner, a question need to be further addressed. To selectively antagonize PI3K/mTOR signaling in the myeloid cells of pten−/− embryos, we transiently expressed zebrafish pten (ptena and ptenb) by Tol2-mediated gene transfer using lyz promoter (Additional file 1: Figure S5). As predicted, pten−/− embryos with transiently expressed myeloid pten show less lyz- positive myeloid cells (Additional file 1: Figure S5D) compared to control embryos (Additional file 1: Figure S5C and E).

Taken together, our data indicate that the abnormal activation of the PI3K/mTOR pathway resulting from the deficiency of pten contributes to the expansive dysmyelopoiesis observed in pten−/− zebrafish embryos.

C/ebpα plays a vital role in myelopoiesis downstream of pten

Our previous study indicated that C/EBPα could regulate myeloid development downstream of PTEN. Elevated expression of PTEN could promote the expression of CEBPA in HL-60 cell line [23]. Moreover, C/EBPα is a well-known master regulator of myelopoiesis [40]. To further elucidate the role of C/ebpα in pten deficiency-induced expansive myelopoiesis, we performed real-time quantitative PCR to evaluate the expression level of cebpa in EGFP-positive myeloid cells isolated from pten−/−;lyz:EGFP embryos at every 6 hours between 72 and 102 hpf. In contrast to the gradually increasing number of lyz-positive myeloid cells in the pten−/− embryos (Additional file 1: Figure S3A-L) from 72 to 102 hpf (Figure 6A, dashed line and the right Y axes), the expression level of cebpa in the EGFP-positive myeloid cells gradually decreased (Figure 6A, solid line and the left Y axes).

Figure 6
figure 6

The overexpression of C/ebpα ameliorates expansive myelopoiesis in pten- deficient embryos. (A) The expression level of cebpa was negatively related to the number of myeloid cells. Real-time quantitative PCR of cebpa in EGFP-positive myeloid cells from pten−/−;lyz:EGFP and control embryos was performed every 6 hours from 72 to 102 hpf (solid line and left Y-axis). The lyz- positive myeloid cells in each embryo were counted via the whole-mount in situ hybridization analyses of embryos at 72 to 102 hpf (dashed line and right Y-axis). The indicated value is the ratio of pten−/− to control myeloid cells. (B-E) Whole-mount in situ hybridization analyses for lyz in embryos injected with in vitro- synthesized control (B and C) or cebpa (D and E) mRNA at the one-cell stage. The overexpression of cebpa mRNA largely rescues expansive myelopoiesis in pten−/− embryos. The red arrowheads indicate expansive myelopoiesis in the mock-injected pten−/− embryos and the decrease in myeloid cells in the cebpa- injected pten−/− embryos. (F-I) Sudan Black staining of neutrophils in embryos injected with control (F and H) or cebpa (G and I) mRNA. The red arrowheads indicate neutrophils in the CHT. (J and K) Quantification of myeloid cells in each embryo (B-E and F-I). The data shown are the means ± SEM of at least 30 embryos; *p < 0.05, **p < 0.01, ***p < 0.001 versus the corresponding controls. Scale bar: 100 um.

Full size image

The downregulation of cebpa may contribute to expansive myelopoiesis in the pten−/− embryos. We therefore microinjected cebpa mRNA into one-cell-stage pten−/− embryos and then performed whole-mount in situ hybridization analyses on the injected embryos using a probe for lyz at 90 hpf. Compared to the mock-injected pten−/− embryos (Figure 6D), the number of lyz-positive myeloid cells significantly decreased in the pten−/− embryos injected with cebpa mRNA (Figure 6E and J), suggesting that C/ebpα functions downstream of pten to regulate the homeostasis of definitive myelopoiesis.

To further determine the functional consequences of cebpa overexpression, we performed Sudan black staining on cebpa mRNA- injected pten−/− embryos (Figure 6F-I). Our result showed that the number of mature neutrophils was obviously rescued in pten−/− embryos injected with cebpa mRNA (Figure 6I), in contrast to the mock-injected ones (Figure 6H and K). These data suggested that overexpression of cebpa mRNA could abolish the block in neutrophil maturation induced by pten-dificiency.

We next examined whether the PI3K/mTOR pathway was involved in the contribution of C/ebpα to expansive myelopoiesis. To test this possibility, real-time quantitative PCR was performed to detect the expression level of cebpa in lyz- positive myeloid cells treated with LY294002 or rapamycin. Surprisingly, the elevated expression of cebpa was observed only in the LY294002-treated embryos (Additional file 1: Figure S6), suggesting that C/ebpα functions downstream of pten in a PI3K-dependent, mTOR-independent manner.

Collectively, our results demonstrate that the downregulation of cebpa by the PI3K pathway contributes to pten deficiency-induced expansive myelopoiesis.

Discussion

More than a decade of work has been performed to examine the role of PTEN, which is a frequently mutated tumor suppressor gene [41]. In this study, we investigated the role of pten in the developmental process of hematopoietic lineages, which has not been fully characterized. Our data revealed an essential role for pten in definitive hematopoiesis, particularly myelopoiesis. Importantly, we determined that immune-deficient cells are produced during the expansive myelopoiesis caused by the loss of pten. Moreover, this expansive myelopoiesis could be completely rescued by treatment with inhibitors of the PI3K/mTOR pathway. Additionally, we found that the expression of cebpa is regulated by PTEN/PI3K signaling and that cebpa downregulation contributes to the expansive myelopoiesis induced by pten deficiency. Collectively, our results help to elucidate the previously obscure role of pten in zebrafish myelopoiesis and the innate immune response in addition to the critical role of C/ebpα, which functions downstream of pten in the regulation of myelopoiesis.

Our previous report indicated that ptenb morphants develop myelodysplasia during primitive hematopoiesis [23]; however, this was not observed in our ptenb and pten−/− mutants (data not shown and Additional file 1: Figure S1). This discrepancy may be due to the toxicity of the morpholino and/or cross-reactions of the ptenb morpholino with other unknown transcripts. Nonetheless, we did observe abnormal hematopoiesis at a later stage (90 hpf) in the pten−/− embryos, which further confirmed the indispensability of pten in hematopoiesis. The most striking phenotype that we observed in the pten−/− embryos was expansive myelopoiesis (Figure 1A-B and E), which was specific to the developmental stage and could not be detected at 72 hpf (Additional file 1: Figure S3A-B), as indicated by the whole-mount in situ hybridization staining of lyz. Considering that our pten−/− embryos were obtained by incrossing ptena+/−ptenb−/− fish, the maternal contribution of ptena[42] might have contributed to the developmental stage-specific occurrence of expansive myelopoiesis, which might no longer be observable, when the maternal influence of ptena eventually dissipates.

Despite the relative increased staining of HSPCs (Additional file 1: Figure S2A-D), we observed dispersed erythrocytes in the head region (Additional file 1: Figure S2E-F), which might have been caused by vascular and/or circulation defects [43].

Similar to mammals, zebrafish myeloid cells (mainly neutrophils and macrophages) are important partners of the innate immune system [33, 36, 44]. However, the exact function of pten in myeloid cells has remained unclear. In mice, Pten was first highlighted as a suppressor of myeloid cell migration [45]. Conversely, another study indicated that Pten functioned to promote neutrophil influx during inflammation [46]. Our results clearly show that pten is indispensable for the maturation of neutrophils, ensuring normal migration during the inflammation induced by acute injury (Figure 2J-L and Figure 3A-C). Interestingly, a recent study indicated that Pten loss could arrest differentiation, which may help to explain impaired immune function of myelocytes [47]. Moreover, our data also suggest that the means by which pten regulates macrophage function are largely different from the means by which it regulates neutrophils, as more Neutral Red absorption was observed in the relatively more macrophages of the pten deficient embryos but almost no mature neutrophils was observed in these embryos (Figure 3F-G and Figure 2J-L). Although the phagocytosis ability of pten−/− macrophages might be greatly elevated, which was previously indicated in ex vivo murine peritoneal and alveolar macrophages [48–50], other possibilities should not be excluded. For example, pten−/− macrophages might be less intact and in turn could easily be saturated by Neutral Red. A more critical functional assay of pten−/− macrophages may help to clarify this issue.

In addition to the characterization of expansive myelopoiesis and the impaired immune response of myeloid cells induced by pten deficiency, this study also indicated the contribution of C/ebpα and PI3K/mTOR signaling in pten-regulated myelopoiesis (Figures 4, 5 and 6). Indeed, the contribution of PI3K/mTOR signaling was further clarified in our pten-deficient zebrafish model showing expansive myelopoiesis (Figures 4 and 5), though the possibility a PI3K-independent function of pten was responsible for this phenotype could not be completely excluded. Although non-specificity of LY294002 at high doses was reported before [51] and this non-specificity was not easy to assess, the increasing rescue effect of dysmyelopoiesis with increasing dose of LY294002 could still support our conclusion. In fact, no additional obvious abnormality was observed in pten−/− fish treated with 30 μM LY294002 compared to mock treated ones. Our previous study indicated that C/EBPα may act as a downstream effector of PTEN in hematopoiesis [23]. Moreover, other study shown that PTEN/PI3K signaling (through regulating eIF4E and eIF2) could influence the expression of CEBPA [52]. Interestingly, our data strongly suggest that the elevated expression of cebpa could significantly ameliorate pten deficiency-induced expansive myelopoiesis (Figure 6); however, more research is still needed to determine the detailed mechanism of C/ebpα regulation via the pten pathway.

Our results also indicated that the regulation of C/ebpα by pten is PI3K dependent and mTOR independent though inhibition of the mTOR pathway could rescue the expansive myelopoiesis induced by the loss of pten (Additional file 1: Figure S6 and Figure 4G-L). This result is conceivable considering that the phosphorylation of Akt at Ser 473 is also regulated in a PI3K-dependent, mTOR-independent manner [53]. We propose that some unknown factors that underlie pten deficiency-induced expansive myelopoiesis might exist downstream of mTOR. Nevertheless, our results suggest that both mTOR-dependent and -independent mechanisms may underlie pten-regulated myelopoiesis; clearly, these mechanisms warrant further investigation.

Conclusions

Our study reveals the developmental function of pten in myelopoiesis and further dissects its role in the immune response of myeloid cells. Our finding that C/ebpα acts as a downstream effector of pten provides more informative clues regarding the intrinsic nature of pten-regulated hematopoiesis. In addition, pten−/−;lyz:EGFP fish are an ideal model for performing large-scale whole-animal small molecule screening to identify drug candidates that can ameliorate defects in myelopoiesis.

Methods

Zebrafish maintenance and embryo production

Zebrafish maintenance, breeding, and staging were performed as described previously [54]. The zebrafish facility and zebrafish study were approved by the Institutional Review Board of the Institute of Health Sciences.

Genomic DNA isolation and genotyping

Embryos or tail fins were incubated in lysis buffer (1 M Tris–HCl [pH 8.3], 1 M KCl, 10% Tween 20, and 10% NP40) and subsequently treated with proteinase K (10 mg/ml) at 55°C overnight. After centrifuging at 12, 000 × g for 10 minutes at 4°C, the supernatant was subjected to genomic PCR using primers specific for the indicated genes.

Riboprobe synthesis and whole-mount mRNA in situ hybridization

Digoxigenin-labeled antisense RNA probes were transcribed from linearized constructs using T3, T7, or SP6 polymerase (Roche Applied Science, Indianapolis, IN, USA). Whole-mount mRNA in situ hybridization was performed as described previously [55]. The probes were detected using an alkaline phosphatase-coupled anti-digoxigenin Fab fragment antibody (Roche Diagnostics, Indianapolis, IN, USA) with 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium (BCIP/NBT) staining (Vector Laboratories, Burlingame, CA).

Phosphorylated histone H3 labeling and TUNEL assay

TUNEL assays were performed using the In Situ Cell Death Detection Kit and TMR Red (Roche Diagnostics) according to the manufacturer’s recommendations. PH3 labeling of fixed embryos was performed by overnight incubation with a rabbit anti-phosphohistone H3 antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA) at 4°C, followed by incubation with an Alexa Fluor 488 donkey anti-rabbit secondary antibody (Invitrogen, Carlsbad, CA, USA).

Flow cytometry analysis and cytology

lyz:EGFP and cmyb:EGFP transgenic embryos were dissected and digested with 0.5% trypsin (GIBCO, Grand Island, NY, USA) for 30 minutes at 37°C. A single-cell suspension was obtained by centrifugation at 400 × g for 5 minutes, washing twice with PBS, and passing through a 40-μM nylon mesh filter. Fluorescence-activated cell sorting (FACS) was performed using the MoFlo system (DakoCytomation, Carpinteria, CA, USA) to obtain homogenous EGFP + cells, which were subsequently subjected to cytospinning at 40 × g for 5 minutes, followed by Wright-Giemsa staining. The staining was performed according to the manufacturer’s instructions (Sigma-Aldrich, UK). The micrographs were obtained using a microscope (Nikon ECLIPSE 80i) with a 100× oil immersion objective and Nikon ACT-1 software (Nikon Corporation, JP).

Neutral Red and Sudan Black staining

For Neutral Red staining, embryos were collected and incubated in egg water with 2.5 μg/mL Neutral Red (Sigma-Aldrich) from 80 to 90 hpf [56]. For Sudan Black staining, fixed embryos were treated with a Sudan Black (Sigma-Aldrich) solution, as previously described [32]. Staining was then observed under a microscope.

Treatment of embryos with small molecule inhibitors

Small molecule inhibitors dissolved in dimethyl sulfoxide (DMSO) were added to egg water as previously described [29, 57]. The concentrations of LY294002 and rapamycin were 5–30 μM and 1–20 μM, respectively. Both inhibitors were added at 78 hpf.

mRNA synthesis and microinjection

mRNA was synthesized using the mMessage mMachine kit (Ambion, Austin, TX, USA). For the microinjection experiment, a volume of 1 nL (30 ng/μL) was injected into embryos at the one-cell stage.

Tol2-mediated gene transfer

The zebrafish lyz promoter was cloned into the EGFP-2A-pDestTol2 plasmid at the XhoI and ClaI sites to get the lyz-pDestTol2 construct. The zebrafish ptena and ptenb were obtained by PCR from the cDNA of wild-type embryos respectively, and then utilizing overlapping PCR to get EGFP-2A-ptena/ptenb fragments which were cloned into the lyz-pDestTol2 construct at the ClaI and SalI sites to get the lyz-EGFP-2A-ptena/ptenb-pDestTol2 constructs. Then the lyz-EGFP-2A-pDestTol2 and lyz-EGFP-2A-ptena/ptenb-pDestTol2 transgenic plasmids (25 ng/μL) were individually co-injected with Tol2 transposase mRNA (25 ng/μL) as previously reported [58] into the pten−/− or control embryos at one cell stage.

Real-time quantitative PCR

Total RNA was extracted from at least 5,000 FACS-sorted cells using the Trizol reagent (Sigma-Aldrich). RNA was reverse-transcribed using oligo (dT) and Superscript™ III reverse transcriptase (RT) (Invitrogen). The subsequent PCR assay was performed with the SYBR Premix ExTaq kit (Takara, JP) (sequence data are available upon request). The relative expression values were normalized to the internal control (β-actin). The cDNA amplification was performed using an ABI Prism 7900 HT cycler (Applied Biosystems, Foster City, CA, USA). Primers for cebpa, 5′-CAAGCAAGAGAAGCTCAAAC, 5′-ACCGTGGTGGTAGTCGTAG; for β-actin, 5′-TGCTGTTTTCCCCTCCATTG, 5′-TTCTGTCCCATGCCAACCA.

Statistical analysis

The quantitative data are expressed as the mean ± SEM. The statistical significance was determined by a two-tailed Student’s t-test (*p < 0.05, **p < 0.01, ***p < 0.001).