CD45 negatively regulates monocytic cell differentiation by inhibiting phorbol 12-myristate 13-acetate-dependent activation and tyrosine phosphorylation of protein kinase Cdelta.

The protein-tyrosine phosphatase CD45 is expressed on all monocytic cells, but its function in these cells is not well defined. Here we report that CD45 negatively regulates monocyte differentiation by inhibiting phorbol 12-myristate 13-acetate (PMA)-dependent activation of protein kinase C (PKC) delta. We found that antisense reduction of CD45 in U937 monocytic cells (CD45as cells) increased by 100% the ability of PMA to enlarge cell size, increase cell cytoplasmic process width and length, and induce surface expression of CD11b. In addition, reduction in CD45 expression caused the duration of peak PMA-induced MEK and extracellular signal-regulated kinase (ERK) 1/2 activity to increase from 5 min to 30 min while leading to a 4-fold increase in PMA-dependent PKCdelta activation. Importantly, PMA-dependent tyrosine phosphorylation of PKCdelta was also increased 4-fold in CD45as cells. Finally, inhibitors of MEK (PD98059) and PKCdelta (rottlerin) completely blocked PMA-induced monocytic cell differentiation. Taken together, these data indicate that CD45 inhibits PMA-dependent PKCdelta activation by impeding PMA-dependent PKCdelta tyrosine phosphorylation. Furthermore, this blunting of PKCdelta activation leads to an inhibition of PKCdelta-dependent activation of ERK1/2 and ERK1/2-dependent monocyte differentiation. These findings suggest that CD45 is a critical regulator of monocytic cell development.

CD45 is the dominant leukocyte plasma membrane phosphatase. It is a long single-chain type I transmembrane protein with an extracellular portion that suggests three fibronectin type III domains and an intercytoplasmic tail that contains two phosphatase domains in direct repeat (1). There are currently five expressed isoforms of CD45, and these are generated by differential splicing of exons 4 -6, while exons 1-3 and 7-33 remain constant (2). Presently, the functional variations in these CD45 isoforms appear primarily due to their extracellular motifs and cell of expression as opposed to differences in their catalytic domains (3). To date, the best described bioaction for CD45 is an activator of the Src family kinases Lck and Lyn in T cell and B cell receptor complexes (4 -7). In thymo-cytes lacking CD45, maturation from immature double positive CD4/CD8 cells to single positive CD4 and CD8 cells is blocked (8). In B cells, loss of CD45 inhibits proliferation induced by antigen-dependent cross-linking of surface IgM and IgD (9). However, in monocytic cells, the function of CD45 is less clear. Recent studies in monocytes have shown that cross-linking of CD45 stimulates an increase in the respiratory burst following tumor necrosis factor-␣ treatment (10) and that ligation of CD45 induces both tumor necrosis factor-␣ and granulocyte/ macrophage colony-stimulating factor production in monocytes (11).
Phorbol-12-myristate-13-acetate (PMA) 1 can induce monocytic cells to differentiate by a mechanism dependent on the activation of protein kinase C (PKC). The PKC family currently comprises 10 related isozymes that have been divided into three groups based on their structure and cofactor requirements. The conventional PKC isoforms ␣, ␤I, ␤II, and ␥ require diacylglyerol (DAG), phosphatidylserine, and Ca 2ϩ for activity. The novel PKC isoforms ␦, ⑀, , and do not require Ca 2ϩ as a cofactor but do require DAG and phosphatidylserine. The atypical PKC isoforms and do not require either Ca 2ϩ or DAG but do bind phosphatidylserine when active (reviewed in Refs. [12][13][14][15][16]. PMA activates the atypical PKC isoforms and conventional PKCs because it is similar in structure to DAG (17). PMA, like DAG, binds to a cysteine-rich region contained within the PKC C1 domain producing a contiguous hydrophobic region that allows PKC to associate with the cell membrane (18). This facilitates conventional PKC and novel PKC isoform binding of phosphatidylserine (19). Binding of PKC family members to relevant cofactors initiates removal of the inhibitory pseudosubstrate domain from the PKC core inducing autophosphorylation on COOH-terminal serines and release of activated PKC from the cell membrane to the cytosol (20).
Protein kinase activation is a critical step in PMA-induced differentiation of promyelocytic cells (21) and is associated with activation of the Raf-1/MAP kinase-signaling pathway (22). The MAP kinase family can be loosely divided into three arms that end in the activation of Jun amino-terminal kinases, p38/ HOG1 kinases, or extracellular signal-regulated kinases (ERKs) (23)(24)(25). PMA-induced ERK activation is through a kinase cascade that requires Raf-1-dependent phosphorylation of MEK1/2 on serines 217 and 221 and MEK1/2-dependent phosphorylation of ERK1/2 on threonine 202 and tyrosine 204 (26,27). PKC-dependent activation of Raf-1 appears to be by direct phosphorylation of serines 43, 259, and/or 499 (28,29). Importantly, inhibitors of PKCs like H-7, GF 109203X, and staurosporine (30 -32) and of MEK (PD98059) block PMA-induced monocytic cell differentiation as measured by a variety of methods including morphometric analysis and surface marker studies (33). Here we report that that CD45 negatively regulates PMA-induced monocytic cell differentiation by impeding PMA-dependent PKC␦ tyrosine phosphorylation, PKC␦ activation, and PKC␦-dependent activation of ERK1/2. These findings indicate that CD45 is a critical regulator of monocytic cell development.
Cell Culture-U937 and U266 cells were grown in growth media (RPMI 1640 media supplemented with 10% fetal bovine serum, 2.0 g/liter sodium bicarbonate, 2.5 g/liter glucose, 100,000 units/liter penicillin, 100 mg/liter streptomycin, 1 mM sodium pyruvate, and 10 mM HEPES pH 7.4). Cells were passaged 1:1 with fresh media every 3 days. For PMA treatment, cells were washed twice and resuspended in growth media supplemented with 100 nM PMA.
Vector Construction-The episomal CD45 antisense (CD45as) vector was created to target the expressed 5Ј sequence in exons 1, 2, and 3 that is shared by all known CD45 isoforms (1,2). In brief, total cellular RNA from 5 ϫ 10 6 U266 cells was isolated in Trizol reagent and the target sequence amplified by polymerase chain reaction from random hexamer-primed cDNA using the forward primer (5Ј-GCGGATCCG-GAAATTGTTCCTCGTCTGA-3Ј) and the reverse primer (5Ј-GCAAGCTTCAGTGGGGGAAGGTGTTGG-3Ј) by methods we have described (34). The resultant 196-base pair polymerase chain reaction product was introduced into the pCEP4 vector using the BamHI and HindIII restrictions sites included in the forward and reverse primer sequences and the construct verified by sequence analysis.
Vector Transfection-The U937 cell lines U937-CD45as and U937-pCEP4 were created by introducing the vectors CD45as and pCEP4, respectively, by electroporation. 20 ϫ 10 6 U937 cells were washed twice in and resuspended in 800 l of phosphate-buffered saline (80 mM Na 2 HPO 4 , 20 mM NaH 2 PO 4 ⅐H 2 O, and 100 mM NaCl, pH 7.4). Cells were added to 4-mm gap electroporation cuvettes with either 20 g of CD45as or pCEP4 vector and electroporated at 400 V in an EC 100 electroporator (Fisher, Pittsburgh, PA). Cells were recovered in growth media for 24 h and then removed to growth media supplemented with 250 g/ml hygromycin B. After 2 weeks, CD45as cells were labeled with FITC-conjugated CD45 antibody (35) and sorted for low CD45 expression on a MoFlo flow cytometer (Cytomation, Fort Collins, CO).
Flow Cytometry-Immunolabeling of cells was performed as previously described (36). In brief, after indicated treatments, cells were incubated in growth media supplemented with 5 mM EDTA for 1 h at 37°C and then washed once in 0.5% bovine serum albumin containing phosphate-buffered saline. FITC-conjugated antibodies at 7 g/ml/test were added to 1 ϫ 10 6 cells and incubated on ice for 15 min. Fluorescence was detected at an excitation of 480 nm on an Epics XL flow cytometer (Beckman Coulter, Fullerton, CA) quantifying 1 ϫ 10 4 events using gates to exclude nonviable cells as determined by propidium iodide staining.
Statistical Analysis-Where indicated, experimental data was analyzed by Student's t test for comparison of medians using Excel (Microsoft, Redmond, WA).

CD45 Negatively Regulates PMA-induced Monocytic Cell
Differentiation-PMA is a potent activator of monocytic cells increasing surface expression of CD45 and CD11b and inducing phenotypic maturation (21). To determine the role of CD45 in PMA-dependent monocytic cell activation, CD45, CD11b, and cell morphology were examined in U937 cells transfected with the CD45 antisense vector (U937-CD45as cells) and U937 cells transfected with the vector backbone (U937-pCEP4 cells) treated with 100 nM PMA for 2 days. Fig. 1A shows that PMA increased CD45 surface expression in U937-pCEP4 cells by 80% at 2 days from a median fluorescence of 22.8 Ϯ 2.3 to 41.7 Ϯ 7.2. In contrast, CD45 surface expression in U937-CD45as cells was 50% that of U937-pCEP4 cells basally (day 0) at a median fluorescence of 11.9 Ϯ 0.9 and after 2 days of PMA treatment at a median fluorescence of 20.9 Ϯ 2.1. Fig. 1B demonstrates that in U937-pCEP4 cells PMA induced a 3-fold elevation in surface CD11b expression at 2 days as median fluorescence increased from 0.31 Ϯ 0.13 to 1.03 Ϯ 0.06. However, in U937-CD45as cells, PMA exposure increased CD11b expression by 14-fold as median fluorescence increased from 0.28 Ϯ 0.01 to 4.05 Ϯ 0.79. To confirm that the CD45as vector did not down-regulate all surface makers, CD33 expression was examined. As Fig. 1C shows, CD33 surface expression in U937-pCEP4 and U937-CD45as cells was similar both basally (median fluorescence, 1.72 Ϯ 0.11 versus 1.60 Ϯ 0.07) and after PMA treatment (median fluorescence, 1.93 Ϯ 0.03 versus 1.91 Ϯ 0.07). Finally, Fig. 1D demonstrates that PMA increased U937-pCEP4 cell nuclear size, cell size, and cell process length by 2-fold after 2 days. In U937-CD45as cells, PMA induced a 2-fold increase in nuclear size but a 4-fold increase in cell size and cell process length. Additionally, when cells were pre-treated with the MEK inhibitor PD98059 (25 M) for 15 min prior to PMA treatment, PMA-induced morphologic differentiation was blocked. Taken together, these findings demonstrate that transfection of the episomal CD45as vector in U937 cells reduces by 50% the surface expression of CD45. Importantly, reduction in CD45 expression induces at least a doubling of PMA-dependent CD11b expression, cell size, and cell process length when compared with control (U937-pCEP4) cells and that these phenotypic changes are sensitive to MEK inhibition.
PMA-dependent Activation of ERK1/2 Is Regulated by CD45-PMA mediates its effect on monocytic cell differentiation through activation of the MAP kinases (22). To determine whether CD45 affected PMA-dependent MAP kinase activation, MEK and ERK1/2 activation were examined by Western analysis. Fig. 2A shows that in U937-pCEP4 cells PMA induced a 5-fold activation of MEK at 5 min and that by 30 min MEK activation had decreased to 2-fold over basal. In U937-CD45as cells, PMA treatment also caused a 5-fold activation of MEK at 5 min but at 30 min MEK activation still remained 5-fold over basal. Likewise, in Fig. 2B, PMA induced a 5-fold increase in ERK1/2 activation at 5 min and by 30 min this increase had returned to basal levels. In U937-CD45as cells, however, ERK1/2 activation was 5-fold over basal at 5 min and remained elevated at 4-fold over basal at 30 min. Importantly, in split samples, mass of MEK and ERK1/2 was unchanged by antisense expression (Fig. 2, A and B, lower panels). Taken together, these findings demonstrate that antisense reduction of CD45 leads to an extension of peak MEK and MAP kinase activation from 5 to 30 min, indicating that CD45 attenuates MAP kinase activation in PMA activated monocytic cells.

Equalization of ERK1/2 Activation in U937-pCEP4 and U937-CD45as Cells Results in Equivalent PMA-induced CD11b
Expression-As demonstrated in Fig. 1D, the MEK kinase inhibitor PD98059 blocks PMA-induced phenotypic maturation. To demonstrate that loss of CD45-dependent regulation of MAP kinase activation was the cause for PMA-dependent increased CD11b expression in U937-CD45as cells when compared with U937-pCEP4 cells, MAP kinase normalization studies were performed. Fig. 3A shows that, as in  Fig. 2B, PMA-induced ERK1/2 activation is reduced at 15 min by 50% in U937-pCEP4 cells when compared with U937-CD45as cells. However, when U937-CD45as cells are treated with 3 M PD98059 for 15 min prior to PMA exposure, U937-CD45as cell ERK1/2 activation at 5 and 15 min is equivalent to that of U937-pCEP4 cells. Taken together, these findings demonstrate that activation of ERK1/2 is required for PMA-dependent CD11b expression and indicate that regulation of ERK1/2 activation by CD45 controls CD11b surface levels.
PMA-dependent Activation and Tyrosine Phosphorylation of PKC␦ Is Negatively Regulated by CD45-PMA-dependent initiation of the MAP kinases cascade requires activation of PKCs (21,22). Therefore, to determine whether CD45 impacted PKC activation Western analysis was performed using a phospho-PKC antibody to isoforms ␣, ␤I, ␤II, ␥, and ␦. Fig. 4A shows that in U937-pCEP4 cells 100 nM PMA induced a maximal 3-fold PKC autophosphorylation on serine 660 at 5 min. In U937-CD45as cells, however, PMA led to a 12-fold PKC autophosphorylation at the same time point. In addition, masses of PKC␣, ␤I, ␤II, ␥, and ␦, measured in split samples, were not altered by antisense expression (Fig. 4A, lower panels). Since PKC␦ has been shown to be tyrosine-phosphorylated in the myeloid progenitor line 32D in response to PMA (38), the next step was to determine its role in monocytic cell differentiation and its regulation by CD45. Fig. 4B demonstrates that the PKC␦ inhibitor, rottlerin (39), at 2.5 M blocked by 66% PMA induced surface expression of CD11b in U937 cells at 2 days reducing median fluorescence from 7.20 Ϯ 0.30 in PMA-treated cells to 2.4 Ϯ 0.32 in rottlerin/PMA-treated cells. Importantly, rottlerin did not markedly alter cell viability, as measured by propidium iodide staining in that control cells were 94% viable, rottlerin-treated cells 88% viable, PMA-treated cells 88% viable, and rottlerin/PMA-treated cells 73% viable after 2 days of treatment. To confirm that PMA activated PKC␦ and that PKC␦ autophosphorylation was regulated by CD45, Western analysis was performed on using a PKC␦-specific immunoprecipitating antibody. Fig. 4C (top panel) shows that, in U937-pCEP4 cells, 100 nM PMA induced a 3-fold autophosphorylation of PKC␦ at 5 min but that, in U937-CD45as cells, PMA caused a 12-fold autophosphorylation of PKC␦. To determine the potential mechanism by which CD45 modulates PMA-dependent PKC␦ activity, phosphorylation of PKC␦ was examined in split immunoprecipitates. Fig. 4C (upper panel) demonstrates that PMA induced a 3-fold increase in PKC␦ serine 660 phosphorylation in U937-pCEP4 cells and that CD45 antisense expression quadrupled this effect. The middle panel demonstrates that PMA also induced a 3-fold increase in PKC␦ tyrosine phosphorylation in U937-pCEP4 cells and that CD45 antisense expression quadrupled this effect. The lower panel of Fig. 4C shows that in split immunoprecipitates the mass of PKC␦ was unaffected by CD45 expression or PMA treatment. Finally, to examine PKC␣, -␤I, -␤II, and -␥ activation and tyrosine phosphorylation, Western analysis was again performed on split immunoprecipitates. Fig. 4D demonstrates that the PKC isoforms ␣, ␤I, ␤II and ␥ are neither phosphorylated on serine 660 (top panel) nor phosphorylated on tyrosine (middle panel) in response to PMA. The lower panel of Fig. 4D shows that PKC␣, -␤I, -␤II, and -␥ were present in the immunoprecipitates from PMA-treated cells and that PKC␣, -␤I, -␤II and -␥ mass were unaffected by CD45 expression. These findings show that PKC␦ is required for PMA-induced CD11b surface expression and that CD45 negatively regulates the autophosphorylation and tyrosine phosphorylation of PKC␦. Overall, these data indicate that control of PKC␦ tyrosine phosphorylation by CD45 regulates PMA-induced PKC␦ activity and subsequent monocyte differentiation. DISCUSSION These data establish that CD45 negatively regulates PMAdependent monocytic cell differentiation by impairing PKC␦ activation and tyrosine phosphorylation. Flow cytometry demonstrated that antisense reduction of CD45 reduces PMA-dependent expression of CD45 while quadrupling that of CD11b and not affecting that of CD33 (Fig. 1, A-C). Morphologic studies showed that phenotypic characteristics of monocyte differentiation such as increased cell size and cell process length were doubled in U937-CD45as cells when compared with that of U937-pCEP4 (Fig. 1D). In addition, the MEK inhibitor PD98059 blocked all morphologic changes induced by PMA in both cell lines (Fig. 1D). Antisense reduction of CD45 increased the duration of peak PMA-induced MEK and ERK1/2 activation from 5 min to nearly 30 min (Fig. 2, A and B). Furthermore, normalization of PMA-induced ERK1/2 activity in U937-CD45as cells to that in U937-pCEP4 cells resulted in equivalent PMA-dependent CD11B surface expression in these two cell lines (Fig. 3, A and B). Finally, PMA activation in U937-CD45as cells was 4-fold greater than that in U937-pCEP4 cells (Fig. 4A). Importantly, the PKC␦ inhibitor rot- tlerin reduced by 66% PMA-dependent CD11b expression (Fig.  4B) while CD45 antisense expression quadrupled PMAdependent PKC␦ activation and tyrosine phosphorylation (Fig.  4C). Notably, PKC isoforms ␣, ␤I, ␤II, and ␥ were neither activated nor tyrosine-phosphorylated by PMA (Fig. 4D). Taken together, these findings indicate that CD45 blunts PKC␦ activation by reducing PMA-dependent PKC␦ tyrosine phosphorylation and that this inhibition in PKC␦ autophosphorylation results in reduced activation of ERK1/2 and ERK1/2-dependent monocyte differentiation.
PMA-dependent monocytic cell differentiation requires PKC activation and ERK1/2 activation (21,22), but the PKC isoforms involved in this process have only recently become better defined. In the interleukin-3-dependent myeloid cell line 32D, transfection studies initially demonstrated that PKC␣ and PKC␦ were essential for PMA-induced macrophage differentiation while PKC␤II, -⑀, -, and -were not required (40). Subsequent studies in the same cell system demonstrated that only PKC␦ appeared to be required for phorbol ester-induced macrophage differentiation (41). Our data support these findings in that we found that PKC␦ but not PKC␣, -␤I, -␤II, and -␥ were autophosphorylated on serine 660 after PMA treatment and autophosphorylation of serine 660 has been shown to be required for PKC activation (42). In addition, we found that rottlerin a competitive inhibitor of the PKC ATP-binding site, which preferentially inhibits the ␦ isoform at concentrations between 3 and 6 M (39), blocked by 66% PMA-induced monocyte differentiation. Since phorbol esters lead to PKC autophosphorylation (42) and activated PKC initiates the MAP kinase cascade (21,22), the increased serine 660 phosphorylation of PKC␦ we observed in CD45 antisense cells is the likely cause of the increased MEK and ERK1/2 activity we also observed in these cells. In addition, since ERK1/2 activation is required for PMA-dependent monocyte differentiation as shown by us here with MEK inhibition studies and by others (33), we conclude that PKC␦ is the critical PKC responsible for PMA-induced monocyte differentiation in U937 cells and that its autophosphorylation on serine 660 probably controls this process. Furthermore, from the 32D cell studies sited above, PKC␦ would be expected to play an important role, generally, in the phorbol ester activation pathway (43) in monocytes.
CD45 is expressed in nearly all cells of hematopoietic origin (44), but its function in cells of the myeloid lineage is especially ill defined. In vitro, the tyrosine phosphatase activity of CD45 is relatively substrate nonspecific consistent with what is found with many of the transmembrane and intracytoplasmic tyrosine phosphatases (45). This has made knockout and antisense approaches critical to defining the role of CD45 in hematopoietic cells. Our stable episomal antisense vector to the conserved exons 1, 2, and 3 led to a 50% reduction in CD45 expression in unstimulated and PMA-stimulated U937 cells. Surprisingly, this led to an increase in differentiation as measured by CD11b expression and morphologic examination. Expected results were that, as in the T and B cells systems, loss of CD45 would be associated with decreased maturation and reduced ability to respond to activating stimuli. The reason for this incongruity may be due to the role of the Scr family kinases in PMAinduced differentiation.
Unique to PKC␦ is that it is phosphorylated on tyrosine residues after cells are PMA-stimulated (38). Although the kinases involved in this process are not fully delineated, v-Src, Lyn, and Fyn have been shown to associate and/or phosphorylate PKC␦ (46 -48). Currently, the role of PKC␦ tyrosine phosphorylation is not clear. Studies have demonstrated that PKC␦ tyrosine phosphorylation can both increase (38) and reduce (49) PCK␦ activity. Here we show that PKC␦ is tyrosine-phosphorylated in response to PMA and that antisense reduction in CD45 expression increases PMA-dependent PKC␦ tyrosine phosphorylation. Importantly, the increase in PKC␦ tyrosine phosphorylation seen with loss of CD45 expression is associated with increased PKC␦ autophosphorylation, ERK1/2 activation, and differentiation, indicating that CD45-regulated tyrosine phosphorylation is likely responsible for reducing PKC␦ serine 660 phosphorylation and, hence, its autophosphorylation and activation. It is then this reduction in PKC␦ activity that results in reduced ERK1/2 activation and reduced CD11b expression.
Unclear, however, is how CD45 might regulate PKC␦ tyrosine phosphorylation. In terms of the Src family kinases, CD45 is generally thought to dephosphorylate an inhibitory tyrosine thereby activating the kinase (50). In our system, loss of CD45 expression would be expected to lead to reduced Src family kinase activity and decreased PKC␦ tyrosine phosphorylation contrary to what we observed. Interestingly, CD45 has also been implicated in dephosphorylating the activating tyrosine 394 on Lyn and reducing Lyn kinase activity (6,50). In CD45 knockout macrophage precursors, Lyn activity is increased nearly 5-fold (51), implicating Lyn as potentially important to increased PKC␦ tyrosine phosphorylation in CD45-deficient monocytic cells. Finally, CD45 may directly dephosphorylate PKC␦ controlling PKC␦ activity at the inner membrane surface. Preliminary data from our laboratory show that CD45 can dephosphorylate in vitro tyrosine-phosphorylated PKC␦, but, as indicated above, tyrosine phosphatases tend to have little in vitro substrate specificity. In summary, we found that CD45 negatively regulates PMA-induced PKC␦ autophosphorylation, PKC␦ tyrosine phosphorylation, and monocytic cell differentiation. We conclude that CD45 inhibits PMA-dependent PKC␦ tyrosine phosphorylation, thereby blunting PKC␦ activation and PKC␦-dependent activation of ERK1/2. Reduced ERK1/2 activity then limits monocyte differentiation.