Protein tyrosine phosphatase 1B negatively regulates ... - PNAS

Protein tyrosine phosphatase 1B negatively regulates macrophage development through CSF-1 signaling Krista M. Heinonen*†, Nadia Dube´†‡, Annie Bourdeau†‡, Wayne S. Lapp*§, and Michel L. Tremblay*†‡¶ *Division of Experimental Medicine, †McGill Cancer Centre, and Departments of ‡Biochemistry and §Physiology, McGill University, Montreal, QC, Canada H3G 1Y6 Edited by Arthur Weiss, University of California School of Medicine, San Francisco, CA, and approved December 27, 2005 (received for review September 30, 2005)

Protein tyrosine phosphatase 1B (PTP-1B) is a ubiquitously expressed cytosolic phosphatase with the ability to dephosphorylate JAK2 and TYK2, and thereby down-regulate cytokine receptor signaling. Furthermore, PTP-1B levels are up-regulated in certain chronic myelogenous leukemia patients, which points to a potential role for PTP-1B in myeloid development. The results presented here show that the absence of PTP-1B affects murine myelopoiesis by modifying the ratio of monocytes to granulocytes in vivo. This bias toward monocytic development is at least in part due to a decreased threshold of response to CSF-1, because the PTP-1B ⴚ兾ⴚ bone marrow presents no abnormalities at the granulocyte– monocyte progenitor level but produces significantly more monocytic colonies in the presence of CSF-1. This phenomenon is not due to an increase in receptor levels but rather to enhanced phosphorylation of the activation loop tyrosine. PTP-1B ⴚ兾ⴚ cells display increased inflammatory activity in vitro and in vivo through the constitutive up-regulation of activation markers as well as increased sensitivity to endotoxin. Collectively, our data indicate that PTP-1B is an important modulator of myeloid differentiation and macrophage activation in vivo and provide a demonstration of a physiological role for PTP-1B in immune regulation. hematopoiesis 兩 innate immunity 兩 myeloid progenitor cells

P

rotein tyrosine phosphatase 1B (PTP-1B; also known as PTPN1) is a ubiquitously expressed cytosolic phosphatase that is localized to the endoplasmic reticulum (1). PTP-1B ⫺兾⫺ mice were shown to be resistant to diet-induced diabetes and obesity (2–5), and the enzyme has been implicated in several metabolic pathways because of its ability to dephosphorylate and thereby down-regulate the activity of the insulin receptor (2, 3), epidermal growth factor receptor and platelet-derived growth factor receptor (6–8), insulin-like growth factor type I receptor (4, 9), and JAKs associated with cytokine family receptors, including leptin (5, 10) and growth hormone receptors (11). PTP-1B appears to show specificity toward JAK2 and TYK2 (12), and has been shown to regulate IFN signaling in fibroblasts (12). Despite the regulatory role in cytokine signaling, no immunological phenotype has been attributed to the PTP-1Bdeficient mice to date. IL-3, granulocyte colony-stimulating factor (G-CSF), and granulocyte–monocyte colony-stimulating factor (GM-CSF) are cytokines that depend on JAK2 activity for signaling (13–15) and play important roles in the regulation of myeloid development (16). The receptor for another important myeloid cytokine, CSF-1, belongs to the class III receptor tyrosine kinase family together with the platelet-derived growth factor receptor (17), and it also activates TYK2 (18). Furthermore, it has been shown previously that PTP-1B protein levels are up-regulated in cell lines derived from chronic myelogenous leukemia patients (19– 21). These observations led us to investigate the role of PTP-1B in hematopoiesis and, in particular, myelopoiesis. The data presented here show that in the absence of PTP-1B, the monocyte-to-granulocyte ratio is affected in vivo. This defect is likely due to an increased sensitivity to CSF-1, because 2776 –2781 兩 PNAS 兩 February 21, 2006 兩 vol. 103 兩 no. 8

PTP-1B ⫺兾⫺ bone marrow produces more colonies in response to CSF-1, and the receptor was hyperphosphorylated in response to CSF-1 in PTP-1B ⫺兾⫺ bone marrow-derived macrophages or in macrophages expressing the PTP1B aspartic acid-to-alanine substrate-trapping mutant. In addition, PTP-1B-deficient macrophages display increased inflammatory phenotype in vitro and in vivo through the up-regulation of activation markers as well as increased sensitivity to LPS. Together, our results demonstrate that PTP-1B regulates cytokine signaling in the immune system and thereby modulates myelopoiesis and macrophage activation in vivo. Results Augmented Monocytic but Not Granulocytic Development in the Absence of PTP-1B. The mice deficient for PTP-1B are viable and

fertile with no reported gross abnormalities (2, 3). The reports that PTP-1B regulates cytokine signaling by dephosphorylating JAK2 and TYK2 (5, 10–12) led us to hypothesize that hematopoiesis may be affected by the loss of PTP-1B. Of particular interest was myelopoiesis, both due to the importance of JAK2 (13, 15) and the dysregulation of PTP-1B expression in chronic myelogenous leukemia (19, 20). When total bone marrow from young adult (6–7 weeks old) mice was cultured in the presence of myeloid growth factors (IL-3, IL-6, and SCF), PTP-1B ⫺兾⫺ cells differentiated into monocytic colonies in a larger proportion than the PTP-1B ⫹兾⫹ cells (Fig. 1A). In keeping with the equal numbers of colony-forming unit-GM, the size of the CD34⫹CD16兾CD32⫹ granulocyte–monocyte progenitor (GMP) population was not affected by the loss of PTP-1B (Fig. 1B), suggesting that the enzyme modulates lineage decision below the GMP level. Progenitor cells were first defined as Lin⫺Sca1⫺CD127⫺CD117⫹ and then divided into GMP, common myeloid progenitors, and megakaryocyte-erythrocyte progenitors based on their CD34 and CD16兾CD32 expression (22, 23). The Loss of PTP-1B Sensitizes Cells to CSF-1 and Results in Increased CSF-1 Receptor (CSF-1R) Phosphorylation in Bone Marrow-Derived Macrophages. The major cytokines involved in monocyte devel-

opment and differentiation are CSF-1 and GM-CSF. In single cytokine assays, PTP-1B affected monocytic colony formation only in response to CSF-1 (Fig. 2A). The number of colonies produced in the absence of PTP-1B was approximately three times that obtained with PTP-1B ⫹兾⫹ cells. Conversely, there was no difference in the number of monocytic colonies produced by PTP-1B ⫹兾⫹ vs. ⫺兾⫺ bone marrow in response to a range of concentrations of GM-CSF (0.1–15 ng兾ml), or in the number Conflict of interest statement: No conflicts declared. This paper was submitted directly (Track II) to the PNAS office. Abbreviations: G-CSF, granulocyte colony-stimulating factor; GM-CSF, granulocyte–monocyte colony-stimulating factor; GMP, granulocyte–monocyte progenitor; PTP, protein tyrosine phosphatase; CSF-1R, CSF-1 receptor. ¶To whom correspondence should be addressed at: McGill Cancer Centre, 3655 Promenade

Sir William Osler, Room 701, Montreal, QC, Canada H3G 1Y6. E-mail: michel.tremblay@ mcgill.ca. © 2006 by The National Academy of Sciences of the USA

www.pnas.org兾cgi兾doi兾10.1073兾pnas.0508563103

Fig. 1. Bias toward monocyte over granulocyte development in PTP-1B ⫺兾⫺ bone marrow. (A) Increased monocytic and decreased granulocytic colony development in a mixed colony assay using bone marrow from 6-week-old male mice. The histogram represents the mean ⫾ SE from three separate experiments. n ⫽ 8 for ⫹兾⫹ and 10 for ⫺兾⫺. *, P ⬍ 0.01; **, P ⬍ 0.001. (B) Representative FACS plots from 7-week-old males, demonstrating no significant changes in the myeloid precursor populations. Cells were first selected on Lin⫺Sca1⫺CD127⫺CD117⫹. n ⫽ 10 for both ⫹兾⫹ and ⫺兾⫺.

shown). Therefore, the expansion of the splenic myelocytes could be due to either an increase in lifespan or to in situ hematopoiesis. In the case of PTP-1B ⫺兾⫺ mice, both mechanisms appeared to play a role. Both monocytes and granulocytes exhibited reduced apoptosis as shown by decreased Annexin V staining (Fig. 3B). A mixed colony assay yielded 2-fold more colony-forming unit-GM in the absence of PTP-1B, suggestive of an increase in splenic myeloid progenitor cells (Fig. 3C). In older animals, the relative size of granulocytic population was diminished in the absence of PTP-1B, reflecting the bias in myelopoiesis in favor of monocytes (Fig. 3D). Moreover, the expression levels of both CD11b and Ly6G on the surface of positive cells were decreased as seen by diminished mean fluorescence (Ly6G 343.3 ⫾ 15.7 for ⫹兾⫹ and 308.0 ⫾ 5.5 for ⫺兾⫺, P ⬍ 0.005; CD11b 409.8 ⫾ 102.9 for ⫹兾⫹ and 290.3 ⫾ 13.8 for ⫺兾⫺, P ⫽ 0.06). However, because of the increase in spleen cellularity (data not shown) in PTP-1B ⫺兾⫺ mice, the absolute number of granulocytes was not diminished when compared with ⫹兾⫹ mice. PTP-1B Deficiency Leads to an Enhanced LPS Sensitivity. Macro-

Increase in Monocyte-to-Granulocyte Ratio in the Periphery in PTP-1B ⴚ兾ⴚ Mice. To investigate whether the bias in myelopoiesis was

reflected in the periphery, we performed flow cytometry analysis on the spleen. Both monocytic and granulocytic populations were expanded (by 58% and 48%, respectively) in young adult PTP-1B ⫺兾⫺ mice (Fig. 3A). We had detected no significant increase in bone marrow myelocytes in these same mice (data not Heinonen et al.

phages normally need priming by, e.g., IFN-␥ before an LPS stimulus for full activation and NO production (25, 26). CSF-1, but not GM-CSF, can substitute for IFN-␥ to some extent in vitro (27, 28), and the differences in CSF-1 sensitivity in PTP-1B ⫺兾⫺ bone marrow cells led us to hypothesize that the loss of PTP-1B may also lead to an increased sensitivity to LPS. Using spleenderived macrophages, we observed that the PTP-1B⫺兾⫺ cells PNAS 兩 February 21, 2006 兩 vol. 103 兩 no. 8 兩 2777

IMMUNOLOGY

of granulocytic colonies in response to G-CSF. There was no significant colony formation in the presence of serum alone (data not shown). The increased number of colonies pointed to an increased prevalence of CSF-1-responsive cells in the PTP-1B ⫺兾⫺ bone marrow. The colonies also contained more cells (51 ⫾ 10 cells per colony for PTP-1B ⫹兾⫹ vs. 110 ⫾ 13 for PTP-1B ⫺兾⫺), indicating that the loss of PTP-1B may also increase proliferation of monocytic precursors in response to CSF-1. To verify this, we used carboxyfluorescein diacetate succinimidyl ester to trace cell divisions and noted that the PTP-1B ⫺兾⫺ cells divided on average one more time than the control cells over the 5-day incubation period as seen by a decrease in fluorescence due to cell division (Fig. 2B) and increased mitotic index (1.4 ⫾ 0.2 for PTP-1B ⫹兾⫹ vs. 2.5 ⫾ 0.6 for PTP-1B ⫺兾⫺; P ⬍ 0.05) (24). The increased sensitivity to CSF-1 was not due to an increase in receptor levels in total bone marrow (Fig. 2C Left) or in bone marrow depleted of erythrocytes (Ter119⫹) and B lineage cells (CD19⫹). To study CSF-1R phosphorylation, we stimulated bone marrow-derived macrophages with 10 ng兾ml CSF-1 at room temperature to minimize receptor internalization and thus optimize the detection of receptor phosphorylation. Under these conditions, the loss of PTP-1B resulted in hyperphosphorylation of the receptor on the activation loop tyrosine 807 (Fig. 2C Center), indicating that PTP-1B indeed controls CSF-1R signaling at the level of receptor phosphorylation. To confirm these data, we created macrophage cell lines that stably expressed either wild-type or mutant PTP-1B. The expression of wild-type PTP-1B resulted in decreased receptor expression (Fig. 5, which is published as supporting information on the PNAS web site), whereas the substrate-trapping mutant enhanced CSF-1R phosphorylation on tyrosine 807 (Fig. 2C Right) in two separate clones.

Fig. 2. The absence of PTP-1B results in increased responsiveness to CSF-1. (A) Colony assay with CSF-1, G-CSF, or GM-CSF alone, demonstrating sensitivity to CSF-1 in ⫺兾⫺ bone marrow. Data are presented as mean ⫾ SE from three separate experiments. n ⫽ 6 –12 for ⫹兾⫹ and 10 –16 for ⫺兾⫺. *, P ⬍ 0.05; **, P ⬍ 0.001. (B) Representative FACS plot demonstrating enhanced proliferation of bone marrow precursors in response to 10 ng兾ml CSF-1. Cells were gated on CD11b⫹Ly6Glo. Mitotic indexes are shown next to the graphs. (C Left) Western blot showing equal CSF1R expression in ⫹兾⫹ and ⫺兾⫺ total bone marrow. (C Center) Increased CSF1R phosphorylation in ⫺兾⫺ bone marrow-derived macrophages after a 1-min stimulation with 10 ng兾ml CSF-1 at room temperature. Similar results were obtained in three independent experiments. (C Right) Expression of PTP-1B aspartic acid-to-alanine mutant increased CSF-1R phosphorylation after a 1-min stimulation with 100 ng兾ml CSF-1 at room temperature. Similar results were obtained with two independent cell lines.

Fig. 3. Increase in peripheral CD11b⫹Ly6G⫺ cells (monocyte兾macrophages) in the PTP-1B ⫺兾⫺ mice. (A) Representative FACS plots from 6-week-old males, demonstrating an increase in both monocytes and granulocytes in the spleen. Histogram represents the mean ⫾ SE of total number of cells per spleen. n ⫽ 9 for ⫹兾⫹ and 13 for ⫺兾⫺ in four separate experiments. (B) Reduced apoptosis of both Ly6G⫺/lo monocytes and Ly6G⫹ granulocytes in the absence of PTP-1B. n ⫽ 10 for both ⫹兾⫹ and ⫺兾⫺ in three separate experiments. *, P ⬍ 0.002. (C) Extramedullary myelopoiesis in the spleen of 6-week-old PTP-1B ⫺兾⫺ males as shown by mixed colony-forming assay. The histogram represents the mean ⫾ SE from two separate experiments. n ⫽ 6 for both ⫹兾⫹ and ⫺兾⫺. *, P ⬍ 0.01. (D) Representative FACS plots from 14-week-old males, demonstrating a switch in macrophage兾granulocyte ratio. Histogram represents the mean ⫾ SE of total number of cells per spleen. n ⫽ 8 for both ⫹兾⫹ and ⫺兾⫺ in two independent experiments.

were indeed more sensitive to low (picograms兾milliliter) concentrations of LPS (Fig. 4A). There was no significant difference in NO production in response to IFN-␥ alone (3.5 ⫾ 1.6 ␮M for ⫹兾⫹ vs. 8.2 ⫾ 2.5 ␮M for ⫺兾⫺), or when the cells were maximally activated with IFN-␥ plus LPS (37.2 ⫾ 15.9 ␮M for ⫹兾⫹ vs. 57.3 ⫾ 12.0 ␮M for ⫺兾⫺). Spontaneous NO production was below detection limit (1–2 ␮M) for both PTP-1B ⫹兾⫹ and PTP-1B ⫺兾⫺. The increase in NO production was due to an up-regulation of inducible NO synthase expression (Fig. 4B) despite the absence of differences in JNK or p38 phosphorylation (data not shown). In accordance with Myers et al. (12), STAT1 phosphorylation was also enhanced downstream of IFN-␥ (Fig. 4C). We did not detect any significant differences in baseline inducible NO synthase mRNA (data not shown). The level of CD14 or CD11b expression was not affected by the loss of PTP-1B (data not shown). The loss of PTP-1B resulted in increased macrophage activation in vivo as well, as shown by the expansion of the CD80⫹Ly6G⫺CD11b⫹ population in the spleen (Fig. 4D). The mean fluorescence for CD80 on positive cells was also increased (36.8 ⫾ 4.1 for ⫹兾⫹ vs. 55.8 ⫾ 6.4 for ⫺兾⫺; P ⬍ 0.001) indicating an up-regulation of CD80 surface expression on individual cells. This was specific for CD80 because CD86 was not enhanced either in terms of surface expression levels or the percentage of expressing cells (data not shown). To demonstrate that the macrophage activation in PTP-1B 2778 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0508563103

⫺兾⫺ mice resulted in increased LPS sensitivity in vivo as well, 3-month-old male mice were injected with 0.5 mg兾kg LPS intravenously and bled 2 or 4 h after injection. In PBS-injected control animals, there was no detectable IL-12 or IFN-␥ (Fig. 4E). After 2 h, IL-12 could be detected in most PTP-1B ⫺兾⫺ animals and some ⫹兾⫹ controls, but it was significantly more strongly induced in the absence of PTP-1B (Fig. 4E Upper). Accordingly, IFN-␥, which was detectable after 4 h, was also present at 5-fold higher levels in PTP-1B ⫺兾⫺ mice (Fig. 4E Lower). These data indicate that the increased activation of macrophages can be translated into an increased sensitivity to LPS in vivo. The cytokine production was not only transient but developed into acute shock in the absence of PTP-1B. Mice injected with 4 mg兾kg LPS were followed for 7 days for symptoms of septic shock and killed when they became severely hunched, when their mobility was decreased, or if they lost ⬎20% of their body weight. The majority of PTP-1B ⫺兾⫺ mice became immobile and were killed within 48 h (Fig. 4F). None of these lost ⬎10% of their weight. In comparison, only 20% of control mice were killed over the 7-day course of study, and half of these deaths were due to weight loss. Overall, weight loss was delayed by the absence of PTP-1B, possibly because of increased leptin signaling (29). The mean symptom score was higher in PTP-1B ⫺兾⫺ mice than PTP-1B ⫹兾⫹ controls throughout the course of study. Heinonen et al.

Discussion The data presented above identify PTP-1B as a critical regulator of macrophage and granulocyte development and lineage specification. The absence of PTP-1B in mice resulted in a significant increase in the monocyte兾macrophage population in the spleen and bone marrow (data not shown) as the mice became older. This switch in differentiation was translated into an increased number of CSF-1-responsive cells in the PTP-1B ⫺兾⫺ bone marrow as well as a decreased threshold to both CSF-1 and LPS, because of hyperphosphorylation of the CSF-1R. Collectively, our results suggest that PTP-1B plays a role in lineage commitment after the common granulocyte–monocyte precursor stage. The reason for the enhanced sensitivity of PTP-1B ⫺兾⫺ bone marrow cells to CSF-1 appears to stem from the increased phosphorylation of the receptor upon stimulus at the level of the activation loop tyrosine. The overall level of receptor expression was not affected by the deletion of PTP-1B, although it is possible that the level of receptor on a subset of cells is different, similar to what we detected in cell lines overexpressing PTP-1B. It has been previously published that CSF-1R is not a substrate of PTP-1B in COS cells (6). The authors looked at the total tyrosine phosphorylation of the receptor, however, whereas we concentrated on a specific residue. Moreover, modulation of CSF-1R signaling by PTP-1B may be restricted to a particular cell lineage. Because of its location in the endoplasmic reticulum and contact with the endosomal compartment, PTP-1B may also play an indirect role in receptor signaling by modulating trafficking and recycling. It was recently reported to modulate the maturation of Flt3, a receptor related to CSF-1R (30). Given our Heinonen et al.

overexpression results, it might be interesting to further study the role of PTP-1B in this context. PTP-1B not only regulates the lineage specification but also activation of monocyte兾macrophage lineage cells. Tissue macrophages in the absence of PTP-1B display an activated phenotype, with a significant percentage of them expressing CD80. Exclusive up-regulation of CD80 but not CD86 has previously been reported in pulmonary alveolar macrophages (31, 32) from asthmatic patients as well as in spleen-derived dendritic cells (33) in a mouse model of allergic asthma. CD80 was proposed to play a role in the maintenance of inflammation, whereas CD86 was important for its induction. Together with the decreased level of CD11b expression on individual cells, the increase in CD80 levels points to the development of a variant type of macrophages with a predisposition to an inflammatory type of immune response. Lower levels of CD11b on the monocyte surface are also an indication of a potentially decreased phagocytic capacity. It has been previously shown that integrin signaling is affected by the loss of PTP-1B (34, 35), which may have an impact on exocytosis and phagocytosis. CSF-1 stimulation is known to enhance LPS sensitivity (27, 28). Although the loss of PTP-1B is not lethal in mice under normal, controlled conditions, challenge with LPS resulted in increased systemic IFN-␥ production as well as decreased survival of PTP-1B ⫺兾⫺ mice. This hypersensitivity is most likely due to a decreased threshold of PTP-1B ⫺兾⫺ macrophages to LPS, resulting in enhanced IL-12 secretion, which would in turn activate IFN-␥ production by T cells and natural killer cells (36). Recent findings indicate that PTP-1B deficiency confers protection from endoplasmic reticulum stress-induced cell death PNAS 兩 February 21, 2006 兩 vol. 103 兩 no. 8 兩 2779

IMMUNOLOGY

Fig. 4. Hypersensitivity to LPS in the absence of PTP-1B. (A) Highly increased LPS sensitivity as demonstrated by strong NO production by ⫺兾⫺ (■) as compared with ⫹兾⫹ (E) spleen-derived macrophages. The results are representative of three separate experiments with a total of six mice per group. (B) Increase in inducible NO synthase expression in response to 10 ng兾ml LPS in ⫺兾⫺ as compared with ⫹兾⫹ cells. Similar results were obtained in three separate experiments. (C) Increased STAT1 phosphorylation in response to 100 units兾ml IFN-␥ in the absence of PTP-1B. Similar results were obtained in three separate experiments. (D) Representative FACS plots showing increased CD80 surface expression on CD11b⫹Ly6G⫺/lo cells in the absence of PTP-1B, expressed as the average percentage of CD80⫹ cells from three independent experiments. n ⫽ 12 for both ⫹兾⫹ and ⫺兾⫺. (E) Increased serum IL-12 (p70) (Upper) and IFN-␥ (Lower) 2 and 4 h after an i.v. injection with 0.5 mg兾kg LPS in the absence of PTP-1B. The data are presented as the mean ⫾ SE from three separate experiments. n ⫽ 7 for both groups at 2 h; n ⫽ 10 for both groups at 4 h. *, P ⫽ 0.05; **, P ⬍ 0.005 (using the Wilcoxon rank sum test). (F) The loss of PTP-1B decreases survival after an i.v. injection with 4 mg兾kg LPS. x, control (PBS-injected) PTP-1B ⫹兾⫹; ‚, control (PBS-injected) PTP-1B ⫺兾⫺; E, LPS-injected PTP-1B ⫹兾⫹; ■, LPS-injected PTP-1B ⫺兾⫺. n ⫽ 9 and 4 for LPS-injected animals and PBS-injected animals, respectively.

(37) and Fas-mediated liver damage (38). PTP-1B ⫺兾⫺ macrophages could therefore be more resistant to activation-induced apoptosis and show prolonged potential for cytokine and nitrogen radical secretion. We have also detected an increase in the DX5⫹TCR␤⫹ population in the spleen of PTP-1B ⫺兾⫺ mice (K.M.H. and M.L.T., unpublished data), suggesting that part of the in vivo response to LPS may also be directly due to splenic natural killer T cells (39). We did not detect any difference in the number of GMPs in the bone marrow, or any impairment in granulocyte survival, suggesting that the decrease in the percentage of splenic granulocytes may be directly due to the increased numbers of monocytes. Indeed, when one accounts for the increase in spleen cellularity, there is no actual decrease in the absolute number of granulocytes in PTP-1B ⫺兾⫺ mice (2.3 ⫾ 0.8 million for ⫹兾⫹ vs. 2.1 ⫾ 0.6 million for ⫺兾⫺). The increased sensitivity of PTP-1B-deficient cells to CSF-1 may in this setting play a role in the lineage commitment. Sustained expression of transcription factors, such as Pu.1 (40–42) and early growth response 1 (Egr-1) (43), also favors monocytic differentiation. Egr-1 activity suppresses PTP-1B expression (44), and the loss of PTP-1B could in this context mimic sustained Egr-1 expression. The increased sensitivity to IFN-␥ may also increase the activity of IFN regulatory factor 8 and thus modulate cell fate toward monocyte兾macrophage lineage (45). Interestingly, we did not see increased colony formation by PTP-1B ⫺兾⫺ bone marrow in response to either GM-CSF or G-CSF, both of which signal through JAK2, clearly indicating that PTP-1B controls myeloid lineage commitment independent of cytokine signaling in addition to its role downstream of CSF-1R. PTP-1B inhibitors are currently under study as treatment modalities for obesity and type II diabetes (46, 47). In our study, heterozygous animals did not show appreciable intermediate phenotype. Also, the development of increased LPS sensitivity most likely requires changes at the level of macrophage development, because short-term treatment with suramin, a broadspectrum PTP inhibitor, was shown to be protective against liver damage due to endotoxic shock (48). Therefore, although PTP-1B undoubtedly plays an important role in myeloid development and activation, the short-term use of inhibitors would most likely not have deleterious effects on macrophage function in humans. Conversely, the results presented here identify PTP-1B as a potential target for treating myeloid malignancies, because the level of PTP-1B expression affects myeloid commitment. The increase in PTP-1B expression in chronic myelogenous leukemia patients has been documented (19–21), but the actual consequences of this increase are not known. It is therefore possible that decreasing the level of PTP-1B in these cells could influence their differentiation and IFN sensitivity, and consequently their response to treatment (49). In conclusion, our data indicate a previously uncharacterized role for PTP-1B in myeloid development and CSF-1 signaling in macrophages and identify it as a potential target for the treatment of myeloid malignancies. Materials and Methods Mice. The generation of ptpn1⫺/⫺ mice was described in ref. 2.

The mice were kept in specific pathogen-free housing and used between 6 and 15 weeks of age as specified. All animal work was carried out in accordance with the regulations of the Canadian Council of Animal Care and approved by the McGill Animal Care Committee. Reagents, Antibodies, and Media. Cell culture reagents were pur-

chased from Invitrogen unless otherwise specified. LPS (Escherichia coli O111:B4) was obtained from Calbiochem and prepared as described in ref. 50. Serum for liquid cultures was endotoxin-free and purchased from HyClone. Recombinant 2780 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0508563103

murine (rm) IFN-␥ was purchased from Invitrogen, and rm CSF-1, rm G-CSF, and rm GM-CSF were purchased from StemCell Technologies. Antibodies for flow cytometry were purchased from BD Biosciences, Serotec, BioSource International, and BioLegend. Antibodies for Western blotting were obtained from Cell Signaling Technology, Upstate Biotechnology, and Santa Cruz Biotechnology. Flow Cytometry. Fresh cell suspensions from bone marrow or

spleen of age-matched male PTP-1B ⫹兾⫹ and ⫺兾⫺ mice were prepared in PBS plus 2% FBS. Nonspecific binding was blocked with purified anti-CD16兾CD32 antibody (BD Biosciences), and cells were labeled with a combination of f luorochromeconjugated substrate-specific antibodies. Data acquisition and analysis was conducted on a FACSCalibur machine (Becton Dickinson) by using CELLQUEST software.

Methylcellulose Colony Assays. All colony assays were performed

on young-adult male mice (6–7 weeks of age). M3434 (StemCell Technologies) was used for mixed colony assays, and colonies were scored after 14 days based on morphology according to the manufacturer’s instructions. For single cytokine colony assays, cells were cultured in M3231 (StemCell Technologies) and either CSF-1, G-CSF, or GM-CSF as indicated. Colonies were scored on day 11. Proliferation Assays. Bone marrow cells from 7-week-old male

mice were labeled with carboxyfluorescein diacetate succinimidyl ester (Molecular Probes) in DMSO according to the manufacturer’s instructions. The cells were cultured for 5 days in Iscove’s modified Dulbecco’s medium supplemented with 10% FBS and 10 ng兾ml CSF-1. The CSF-1 supplemented medium was refreshed on day 4. On day 5, adherent cells were collected by scraping and analyzed by flow cytometry. Mitotic index was calculated as described in ref. 24. Generation of Stable Cell Lines. Retroviral vectors for PTP-1B were

described in ref. 37. Raw 264.7 cells were exposed to the virus for 24 h and then selected by using 140 ␮g兾ml hygromycin. Two independent cell lines were used for either PTP-1B wild-type or aspartic acid-to-alanine mutant. Spleen Cultures. Spleen-derived macrophages were obtained es-

sentially as described in ref. 51 by using medium supplemented with 30% L-929 conditioned medium and 10% FBS. The L-929 medium was replaced with unsupplemented medium for 24 h before stimulation of the adherent cells with IFN-␥ and兾or LPS. Macrophage lineage was confirmed by CD11b and CD14 stain. Measurement of NO production in supernatant was performed essentially as described in ref. 52. Determining LPS Sensitivity in Vivo. Twelve- to 14-week-old male

mice were injected i.v. with 200 ␮l of sterile PBS containing 0, 0.5, or 4 mg兾kg LPS. Mice were followed for 7 days after injection and scored for symptoms (body weight, eyes, fur, mobility, posture, and diarrhea) twice daily. Mice were killed if they lost 20% of their body weight or became hunched and immobile. Blood samples for serum cytokine levels were collected by cardiac puncture 4 h postinjection and analyzed with OptEIA mouse IFN-␥ and IL-12 (p70) ELISA kits (BD Biosciences). Statistical Analysis. All statistical differences were determined by

a two-tailed, unpaired, Student t test analysis unless otherwise specified. We thank Rosemary Siegrist-Johnstone and Ailsa Lee Loy for technical assistance, Malcolm Baines and Sofi Julien for helpful comments on the Heinonen et al.

K.M.H.), a CIHR Doctoral Research Award (to N.D.), and a Fellowship from the Cancer Research Institute (to A.B.). M.L.T. is a Chercheur National of Fonds de la Recherche en Sante´ du Que´bec and holds a Jeanne and J-Louis Le´vesque Chair in Cancer Research.

1. Frangioni, J. V., Beahm, P. H., Shifrin, V., Jost, C. A. & Neel, B. G. (1992) Cell 68, 545–560. 2. Elchebly, M., Payette, P., Michaliszyn, E., Cromlish, W., Collins, S., Loy, A. L., Normandin, D., Cheng, A., Himms-Hagen, J., Chan, C. C., et al. (1999) Science 283, 1544–1548. 3. Klaman, L. D., Boss, O., Peroni, O. D., Kim, J. K., Martino, J. L., Zabolotny, J. M., Moghal, N., Lubkin, M., Kim, Y. B., Sharpe, A. H., et al. (2000) Mol. Cell. Biol. 20, 5479–5489. 4. Buckley, D. A., Cheng, A., Kiely, P. A., Tremblay, M. L. & O’Connor, R. (2002) Mol. Cell. Biol. 22, 1998–2010. 5. Zabolotny, J. M., Bence-Hanulec, K. K., Stricker-Krongrad, A., Haj, F., Wang, Y., Minokoshi, Y., Kim, Y. B., Elmquist, J. K., Tartaglia, L. A., Kahn, B. B. & Neel, B. G. (2002) Dev. Cell 2, 489–495. 6. Liu, F. & Chernoff, J. (1997) Biochem. J. 327, 139–145. 7. Haj, F. G., Markova, B., Klaman, L. D., Bohmer, F. D. & Neel, B. G. (2003) J. Biol. Chem. 278, 739–744. 8. Dube, N., Cheng, A. & Tremblay, M. L. (2004) Proc. Natl. Acad. Sci. USA 101, 1834–1839. 9. Buckley, D. A., Loughran, G., Murphy, G., Fennelly, C. & O’Connor, R. (2002) Mol. Pathol. 55, 46–54. 10. Cheng, A., Uetani, N., Simoncic, P. D., Chaubey, V. P., Lee-Loy, A., McGlade, C. J., Kennedy, B. P. & Tremblay, M. L. (2002) Dev. Cell 2, 497–503. 11. Gu, F., Dube, N., Kim, J. W., Cheng, A., Ibarra-Sanchez Mde, J., Tremblay, M. L. & Boisclair, Y. R. (2003) Mol. Cell. Biol. 23, 3753–3762. 12. Myers, M. P., Andersen, J. N., Cheng, A., Tremblay, M. L., Horvath, C. M., Parisien, J. P., Salmeen, A., Barford, D. & Tonks, N. K. (2001) J. Biol. Chem. 276, 47771–47774. 13. Silvennoinen, O., Witthuhn, B. A., Quelle, F. W., Cleveland, J. L., Yi, T. & Ihle, J. N. (1993) Proc. Natl. Acad. Sci. USA 90, 8429–8433. 14. Shimoda, K., Iwasaki, H., Okamura, S., Ohno, Y., Kubota, A., Arima, F., Otsuka, T. & Niho, Y. (1994) Biochem. Biophys. Res. Commun. 203, 922–928. 15. Quelle, F. W., Sato, N., Witthuhn, B. A., Inhorn, R. C., Eder, M., Miyajima, A., Griffin, J. D. & Ihle, J. N. (1994) Mol. Cell. Biol. 14, 4335–4341. 16. Friedman, A. D. (2002) Oncogene 21, 3377–3390. 17. Reilly, J. T. (2002) Br. J. Haematol. 116, 744–757. 18. Novak, U., Harpur, A. G., Paradiso, L., Kanagasundaram, V., Jaworowski, A., Wilks, A. F. & Hamilton, J. A. (1995) Blood 86, 2948–2956. 19. LaMontagne, K. R., Jr., Flint, A. J., Franza, B. R., Jr., Pandergast, A. M. & Tonks, N. K. (1998) Mol. Cell. Biol. 18, 2965–2975. 20. LaMontagne, K. R., Jr., Hannon, G. & Tonks, N. K. (1998) Proc. Natl. Acad. Sci. USA 95, 14094–14099. 21. Donato, N. J., Wu, J. Y., Zhang, L., Kantarjian, H. & Talpaz, M. (2001) Blood 97, 2846–2853. 22. Akashi, K., Traver, D., Miyamoto, T. & Weissman, I. L. (2000) Nature 404, 193–197. 23. Passegue, E., Wagner, E. F. & Weissman, I. L. (2004) Cell 119, 431–443. 24. Angulo, R. & Fulcher, D. A. (1998) Cytometry 34, 143–151. 25. Gifford, G. E. & Lohmann-Matthes, M. L. (1987) J. Natl. Cancer Inst. 78, 121–124. 26. Stuehr, D. J. & Marletta, M. A. (1987) J. Immunol. 139, 518–525.

27. Rutherford, M. S. & Schook, L. B. (1992) J. Leukocyte Biol. 51, 69–76. 28. Sweet, M. J., Campbell, C. C., Sester, D. P., Xu, D., McDonald, R. C., Stacey, K. J., Hume, D. A. & Liew, F. Y. (2002) J. Immunol. 168, 392–399. 29. Faggioni, R., Moser, A., Feingold, K. R. & Grunfeld, C. (2000) Am. J. Pathol. 156, 1781–1787. 30. Schmidt-Arras, D. E., Bohmer, A., Markova, B., Choudhary, C., Serve, H. & Bohmer, F. D. (2005) Mol. Cell. Biol. 25, 3690–3703. 31. Balbo, P., Silvestri, M., Rossi, G. A., Crimi, E. & Burastero, S. E. (2001) Clin. Exp. Allergy 31, 625–636. 32. Burastero, S. E., Magnani, Z., Confetti, C., Abbruzzese, L., Oddera, S., Balbo, P., Rossi, G. A. & Crimi, E. (1999) J. Allergy Clin. Immunol. 103, 1136–1142. 33. Cheng, X., Wang, C., Qian, G. & Zhu, B. (2003) Immunol. Lett. 89, 31–38. 34. Cheng, A., Bal, G. S., Kennedy, B. P. & Tremblay, M. L. (2001) J. Biol. Chem. 276, 25848–25855. 35. Liu, F., Sells, M. A. & Chernoff, J. (1998) Curr. Biol. 8, 173–176. 36. Schoenhaut, D. S., Chua, A. O., Wolitzky, A. G., Quinn, P. M., Dwyer, C. M., McComas, W., Familletti, P. C., Gately, M. K. & Gubler, U. (1992) J. Immunol. 148, 3433–3440. 37. Gu, F., Nguyen, D. T., Stuible, M., Dube, N., Tremblay, M. L. & Chevet, E. (2004) J. Biol. Chem. 279, 49689–49693. 38. Sangwan, V., Paliouras, G. N., Cheng, A., Dube, N., Tremblay, M. L. & Park, M. (2005) J. Biol. Chem. 281, 221–228. 39. Dieli, F., Sireci, G., Russo, D., Taniguchi, M., Ivanyi, J., Fernandez, C., Troye-Blomberg, M., De Leo, G. & Salerno, A. (2000) J. Exp. Med. 192, 1645–1652. 40. DeKoter, R. P., Walsh, J. C. & Singh, H. (1998) EMBO J. 17, 4456–4468. 41. Anderson, K. L., Smith, K. A., Conners, K., McKercher, S. R., Maki, R. A. & Torbett, B. E. (1998) Blood 91, 3702–3710. 42. McIvor, Z., Hein, S., Fiegler, H., Schroeder, T., Stocking, C., Just, U. & Cross, M. (2003) Exp. Hematol. 31, 39–47. 43. Kharbanda, S., Nakamura, T., Stone, R., Hass, R., Bernstein, S., Datta, R., Sukhatme, V. P. & Kufe, D. (1991) J. Clin. Invest. 88, 571–577. 44. Fukada, T. & Tonks, N. K. (2001) J. Biol. Chem. 276, 25512–25519. 45. Tsujimura, H., Nagamura-Inoue, T., Tamura, T. & Ozato, K. (2002) J. Immunol. 169, 1261–1269. 46. Johnson, T. O., Ermolieff, J. & Jirousek, M. R. (2002) Nat. Rev. Drug Discov. 1, 696–709. 47. Harley, E. A. & Levens, N. (2003) Curr. Opin. Investig. Drugs 4, 1179–1189. 48. Eichhorst, S. T., Krueger, A., Muerkoster, S., Fas, S. C., Golks, A., Gruetzner, U., Schubert, L., Opelz, C., Bilzer, M., Gerbes, A. L. & Krammer, P. H. (2004) Nat. Med. 10, 602–609. 49. Shimizu, T., Miyakawa, Y., Oda, A., Kizaki, M. & Ikeda, Y. (2003) Exp. Hematol. 31, 601–608. 50. Nestel, F. P., Price, K. S., Seemayer, T. A. & Lapp, W. S. (1992) J. Exp. Med. 175, 405–413. 51. McCormack, J. M., Leenen, P. J. & Walker, W. S. (1993) J. Immunol. 151, 6389–6398. 52. Nestel, F. P., Greene, R. N., Kichian, K., Ponka, P. & Lapp, W. S. (2000) Blood 96, 1836–1843.

Heinonen et al.

PNAS 兩 February 21, 2006 兩 vol. 103 兩 no. 8 兩 2781

IMMUNOLOGY

manuscript, and Eric Massicotte and Martine Dupuis for technical expertise on flow cytometry. This work was supported by the Canadian Institutes of Health Research (CIHR) Operating Grant MOP-62887 (to M.L.T.), a CIHR Canada Graduate Scholarship Doctoral Award (to