MAP kinase phosphorylation is dispensable for cell division, but required for cell growth in Drosophila

Abstract

Proper activation of the Ras/MAPK pathway is broadly required during development, and in many cases, signal transduction downstream of the receptor is linear. Thus, different mechanisms exist to properly regulate the large number of specific developmental outputs that are required by the activation of this pathway. Previously, we have reported a regulated cytoplasmic sequestration of phosphorylated MAPK (pMAPK) in developing Drosophila compound eyes and wings “called MAPK Cytoplasmic Hold”. In the developing wing, we have shown that cytoplasmic hold promotes the differentiation of wing vein tissue, while pMAPK nuclear translocation regulates growth and division. We had also suggested that the Ras pathway signals for inducing cell growth and cell division split upstream of the nuclear translocation of MAPK itself. Here, we further refine the role of MAPK in Drosophila. We report evidence that suggests, for the first time, that the phosphorylation of MAPK is itself another step in the regulation of cell growth and division in both Drosophila wing and eye cells. We show that inhibition of MAPK phosphorylation, or pMAPK nuclear translocation, is sufficient to block cell growth, but not cell division. These data suggest that non-phosphorylated MAPK is sufficient to induce cell division, but not cell growth, once inside the nucleus of the cell.

Key words: :

• Drosophila
• MAPK
• growth
• division
• proliferation
• phosphorylation

Introduction

Activation of Mitogen Activated Protein Kinase (MAPK, also called Extracellular Signal Regulated Kinase, ERK) is the final cytoplasmic event in the Ras/MAPK signal transduction pathway. Signaling from this pathway begins at the plasma membrane with the activation of a transmembrane Receptor Tyrosine Kinase (RTK, such as the Epidermal Growth Factor Receptor, Egfr) in response to a variety of physiological stimuli. This activation leads to RTK dimerization and autophosphorylation, resulting in the activation of the small GTPase protein Ras.¹ Ras then transduces the signal through a series of phosphorylation events, ultimately leading to the dual phosphorylation and activation of MAPK (MAPK to di-phospho MAPK, or pMAPK) in the cytoplasm.^2-4 Upon activation by phosphorylation, MAPK proteins dimerize,⁵ and can then phosphorylate targets in the cytoplasm, and/or rapidly translocate to the nucleus where they phosphorylate and regulate nuclear target proteins and regulate gene expression.^6-9

Proper activation and regulation of Ras/MAPK is required for a number of different developmental processes including cell division, growth, shape, survival, adhesion, migration, differentiation and death.^2,4,10,11 Even though this one pathway is broadly required throughout development, in the majority of cases, transduction of the signal downstream of the RTK is linear.¹² Thus, a variety of different mechanisms exist in order to properly regulate the large number of specific developmental outputs that are required by the activation of this single pathway. Differences in signal strength and/or duration of the signal, activating or inhibitory ligands, and the subcellular compartmentalization of specific pathway components are all methods by which signal specificity is achieved in vivo.^3,12,13 Determining how each method of pathway regulation correlates to each downstream cellular effect is of critical importance to fully understand how this pathway properly functions throughout development in different tissues.

Previously, we have reported a regulated cytoplasmic -sequestration of pMAPK antigen in the developing Drosophila compound eye and wing “MAPK cytoplasmic hold”.^14-16 We have further shown that this MAPK cytoplasmic hold is functionally regulated in the developing eye by the apical sequestration (and functional inactivation) of the MAPK import cofactor Moleskin (Msk), which is controlled by both the Decepentaplegic (Dpp) and Hedgehog (Hh) signal transduction pathways.¹⁷ In the developing wing, we have shown that pMAPK cytoplasmic hold promotes the differentiation of wing vein tissue, while pMAPK nuclear translocation regulates growth and division¹⁶ (Fig. 6A). This study also suggested that the Ras pathway signals for inducing growth and -division split upstream of the nuclear translocation of the MAPK protein itself.¹⁶ Thus, when we expressed a constitutively nuclear form of MAPK in wing tissue, the overall size of the wing did not change, while the number of cells within this wing increased, suggesting an increase in cell division without a subsequent increase in cell growth.¹⁶ However, when we -specifically increased the nuclear accumulation of the phosphorylated form of MAPK (done by expressing the Msk protein, a nuclear import cofactor which specifically imports the phosphorylated form of MAPK¹⁸) we found that both cell division and cell growth were increased.¹⁶

Figure 6. Model of MAPK signaling. (A) Previous literature suggested that the subcellular localization of MAPK regulates the difference between growth and division and differentiation in developing fly wing and eye cells. (B) Modified model to incorporate additional regulation of growth and division by the phosphorylation status of the MAPK protein.

Here, we further refine the role of MAP Kinase in Drosophila wing and eye development. We report evidence that suggests, for the first time, that the phosphorylation state of the MAP Kinase protein itself is another step in the regulation of cell growth and division processes in both Drosophila wing and eye cells. We show that inhibition of MAPK phosphorylation in cells either through overexpression of the negative MAPK regulator Argos, or through mutation in the MAP Kinase Kinase MEK (encoded by the Drosophila dSor1 gene) is sufficient to block cell growth, but not cell division when we express a constitutively nuclear form of MAPK in these cells. These data suggest that non-phosphorylated MAPK is sufficient to induce cell division, but not cell growth, once inside the nucleus of the cell.

Results

Expression of Argos leads to increased cell density with decreased compartment size

The Drosophila Argos protein is a secreted Egfr ligand that functions to repress pathway activation by sequestering the Egfr activating ligand sSpi.^30,31 Expression of Argos in various tissues in Drosophila has been a well established, often used, and reproducible means to downregulate pathway activity, and leads to a decrease in the levels of phosphorylated MAPK (pMAPK) in vivo. To determine the effect on cell division and cell growth after downregulation of pMAPK expression, we initially utilized the Gal4/UAS system²⁵ to overexpress Argos specifically in the posterior compartment of the developing Drosophila wing. pMAPK is normally expressed in both the developing wing veins, as well as in two columns of cells on both sides of the Dorsal/Ventral compartment boundary of the wing margin (Fig. 1A and B).³² Expression of Gal4 under the engrailed enhancer limits expression of target responder sequences to the posterior compartment of the developing wing disc only (green expression in Fig. 1A). Expression of Argos in the posterior compartment of wing cells significantly decreases the levels of pMAPK in vivo, compared to the anterior control compartment (Fig. 1C and D). Note that this leads to loss of wing vein structures in the adult wing (Fig. 1F), as previously described.³³

Figure 1A–F. Argos expression increases cell density while decreasing cell growth. (A-D) Third instar larval wing discs, anterior up, dorsal left, genotypes indicated bottom right, stain listed top right. (A) engrailed-Gal4; UAS:GFP (en::GFP) wing disc showing GFP expression (green) in the posterior compartment, and phosphorylated MAPK (pMAPK, red) throughout the wing pouch. (B) pMAPK staining (white) from (A). (C) engrailed-Gal4; UAS:GFP/UAS:Argos (en::argos) wing disc showing GFP expression (green) in the posterior compartment, and pMAPK (red) throughout the wing pouch. Note decreased pMAPK staining in the posterior compartment compared to the anterior compartment. (D) pMAPK staining (white) from (C). (E and F) Adult wings. (E) Normal wing venation pattern in engrailed-Gal4; UAS:GFP wing. (F) Phenotype of engrailed-Gal4; UAS:GFP/UAS:Argos adult wing. Note missing and decreased wing venation (arrow).

To determine relative cell number compared to cell size, we utilized a well established technique to quantify the number of wing hairs in a defined area of the wing (a measure of cell density, or cells per unit area), as each wing cell is known to secrete one hair.³⁴ We then compared this to the measured surface area of the compartments analyzed in the adult wing.¹⁶ We again utilized the engrailed Gal4 (en::Gal4) reagent to restrict the expression of proteins to the posterior compartment of the developing wing disc, allowing the use of the anterior compartment to serve as an internal control among the various samples. In order to compare the effect of cell density and compartment size upon perturbation of Egfr/Ras/MAPK signaling in the developing wing we expressed both activating and inhibitory ligands (UAS:Argos and UAS:secreted Spitz [sSpi] respectively). These ligands will have their effect at the level of the Egfr receptor within these cells. To analyze the effect of increased pathway activation downstream of the receptor, we expressed the activated (phosphorylated) form of MAP Kinase (Rl^SEM) in the posterior compartment of these wings, as well as a constitutively nuclear form of MAPK (Rl.nls). As a control for the Gal4/UAS system employed in these assays, we expressed GFP within the posterior compartment.

Expression of GFP in the posterior compartment of wing discs shows no difference in the cell density between the posterior and anterior compartments, suggesting that the number of cells per unit area is largely the same between these regions in normal adult wings (Fig. 1G). Because the number of cells per unit area is equivalent between these two compartments, the relative size of cells in the wing must be similar between the posterior and anterior compartments. However, the posterior compartment in normal wings is significantly larger than the anterior compartment (Fig. 1H).

Figure 1G and H. Argos expression increases cell density while decreasing cell growth. (G) Cell Density in the posterior and anterior compartments of adult wings that express different proteins in their posterior compartment. Average cell density of the same area in the posterior and anterior compartments of each genotype is shown with Standard Error. All genotypes show engrailed:Gal4 expressing different transgenes, as indicated. (H) Surface area (expressed in pixel number) in the posterior and anterior compartments of adult wings that express different proteins in their posterior compartment. Average pixel number (representative of wing surface area) of the entire posterior and anterior compartments of each genotype is shown with Standard Error. All genotypes show engrailed:Gal4 expressing different transgenes, as indicated. In both graphs * indicates p < 0.05, while ** indicates p < 0.01.

Expression of Argos in the posterior compartment of wings leads to a statistically significant increase in cell density (Fig. 1G), with a concomitant decrease in overall compartment size (Fig. 1H). Because cell density is increased in these wings, these data suggest that the posterior compartment of wings that overexpress Argos have more cells per unit area than wild-type wings. This suggests that the cells in the posterior compartment are smaller in wings that express Argos compared to cells in the posterior compartment of wild-type wings.

Expression of the activating ligand sSpi in the posterior -compartment of these wings has been well established to lead to an increase in the pool of endogenous phosphorylated MAPK (pMPAK) protein, which in turn leads to extra vein formation,^16,35 something that we also see when we overexpress sSpi in the posterior compartment of wings (Fig. 2D). We have shown that the phosphorylated MAPK protein induced upon -activation of Egfr signaling in this fashion remains cytoplasmic when Rhomboid (the protease that activates Spitz) is overexpressed in the wing.¹⁶ We therefore analyzed the subcellular localization of phosphorylated MAPK in the posterior compartment of wings that overexpress sSpi, and find that the pMAPK in this compartment is also predominantly cytoplasmic (Fig. 2E and F). We found that expression of sSpi in this fashion did not significantly alter either cell density or posterior compartment size compared to controls (Fig. 1G and H). These data suggest that though increased cytoplasmic pMAPK can induce alterations in cell differentiation by increasing vein formation, it does not affect either cell density or compartment size in this tissue, presumably because this pMAPK remains cytoplasmic.

Figure 2. Increased pMAPK expression by sSpi expression is predominantly cytoplasmic. All panels show late third instar wing discs, anterior up, dorsal left, genotypes indicated bottom right, stain listed top right. (A) Wild-type wing discs shows GFP expression (green) in posterior compartment driven by the engrailed:Gal4 driver (en::GFP). Normal pMAPK expression (red) is indicated. (B) Expression of pMAPK (white) in engrailed-Gal4; UAS:Rl^SEM wing discs. Note increased expression of pMAPK in normal expression domains in posterior compartment. (C) Expression of pMAPK (white) in engrailed-Gal4; UAS:Rl.nls wing discs. Note no significant alteration of pMAPK expression compared to wild-type. (D) Expression of pMAPK (white) in engrailed-Gal4; UAS:sSpi wing discs. Note increased and expanded expression of pMAPK in posterior compartment. (E) Expression of pMAPK (red) and nuclear GFP (green) from red box in (D). Note that pMAPK is predominantly cytoplasmic. Scale bar indicates 5 μm. (F) pMAPK expression (white) from (E).

We next analyzed the effect of increasing MAPK -downstream of Egfr within developing wing cells. Expression of exogenous pMAPK protein (Rl^SEM) in the posterior compartment of wings leads to a significant increase in cell density (Fig. 1G). However, it does not significantly alter posterior compartment size (Fig. 1H). Similarly, expression of a constitutively nuclear form of MAPK (MAPK.nls) also leads to an increase in cell density compared to controls (Fig. 1G), with no significant effect on posterior compartment size (Fig. 1H). To determine the effect of this overexpression on MAPK phosphorylation in the wing, we analyzed the expression of pMAPK in both of these genetic backgrounds. In contrast to overexpression of sSpi in this tissue, neither Rl^SEM or Rl.nls expression significantly altered pMAPK expression within intervein cells (Fig. 2B and C). Expression of pMAPK in wing discs overexpressing Rl^SEM appeared more intense in the posterior compartment compared to anterior -controls (Fig. 2B). However, expression of Rl.nls did not significantly alter the expression pattern in either compartment. Taken together, these data suggest that increasing the total amount of MAPK downstream of the EGF receptor, either phosphorylated or nuclear, can increase total cell density without affecting -compartment size.

Argos dependent increases in cell density are due to increased cell division.

Increased cell density could be due to a decrease in cellular apoptosis, an increase in cell division, or both. To determine the nature of the increased cell density we observe in posterior compartments of wings that express Argos, we first analyzed levels of apoptosis in the posterior compartment of these discs by examining the expression of the activated version of Caspase 3, which marks cells undergoing apoptosis. We did not observe a decrease in apoptosis in posterior compartments, compared to anterior controls, in this tissue (Fig. 3A–D), suggesting that the increase in cell density we observe is not due to a general decrease in apoptosis.

Figure 3. Argos expression does not decrease apoptosis in developing wings. All panels show wild-type third instar wing discs stained for activated Caspase 3. Anterior up, dorsal left. (A) Wild-type wing disc. (B) Positive control showing increased Caspase 3 activation (arrow) in the posterior compartment. Genotype is engrailed-Gal4/ UAS:DIM-7 (Drosophila Importin 7 homolog). (C) Caspase 3 staining (blue) in engrailed-Gal4; UAS:GFP/UAS:Argos discs. GFP is colored green. (D) Caspase 3 staining (white) from (C). Note there is no decrease in Caspase 3 activation upon Argos expression in the posterior compartment of wing discs.

To verify if there was an increase in cell division in the posterior compartment of wings that express Argos, we stained these third instar wing discs with an antibody against the -phosphorylated version of Histone H3 (pH3), which specifically marks the cells that are undergoing mitosis. We expressed Argos within the posterior compartment of wing discs with the engrailed Gal4 driver, while simultaneously expressing GFP within this compartment, and counted the number of pH3 positive nuclei within the -posterior compartment (GFP) compared to the anterior compartment (No GFP). In control wing discs expressing only GFP, the ratio of pH3 nuclei in the posterior compartment compared to the anterior compartment was 0.865 ± 0.141 (Fig. 4A). However, in wing discs expressing Argos in their posterior compartment, we observed a significant increase in the ratio of pH3 nuclei in the posterior compartment compared to the anterior compartment, 1.383 ± 0.127, (Fig. 3A, p = 0.029). To determine whether an increase in the phosphorylated form MAPK could also increase this ratio, we analyzed this posterior to anterior ratio in wings expressing Rl^SEM. While the posterior to anterior ratio of these wings were slightly larger, 1.05 ± 0.045, we observed no significant increase in the ratio of pH3 nuclei in the compartments of these wings compared to controls (Fig. 4A, p = 0.657). Taken together, these data suggest that the increased cell density observed in the posterior compartment of adult wings that express Argos is due to an increase in cell division, and not a decrease in apoptosis within these compartments. Further, because an increase in the amount of phosphorylated MAPK does not significantly increase the number of mitoses observed in these wings when compared to controls, these results are consistent with a role for increased unphosphorylated MAPK in promoting the observed increase in cell division in wings that overexpress Argos.

Figure 4. Phosphorylated MAPK, but not nuclear MAPK rescues Argos phenotype. (A) Shows ratio of posterior to anterior compartment of phosphorylated Histone H3 (pH3) staining in engrailed-Gal4; UAS:GFP third instar wing discs as compared to engrailed-Gal4; UAS:GFP/UAS:Argos and engrailed-Gal4; UAS:GFP/UAS:Rl^SEM wing discs. Note increased pH3 staining in engrailed-Gal4; UAS:GFP/UAS:Argos, indicative of increased mitotic events in the posterior compartment of these discs. (B) Cell Density in the posterior and anterior compartments of adult wings that express different proteins in their posterior compartment. Average cell density of the same area in the posterior and anterior compartments of each genotype is shown with Standard Error. Genotypes are indicated. (C) Surface area (expressed in pixel number) in the posterior and anterior compartments of adult wings that express different proteins in their posterior compartment. Average pixel number (representative of wing surface area) of the entire posterior and anterior compartments of each genotype is shown with Standard Error. Genotypes are indicated. In both graphs * indicates p < 0.05, while ** indicates p < 0.01.

Argos dependent increases in cell density are due to increased nuclear, unphosphorylated MAPK

MAPK functions downstream of Egfr. Because we are suppressing the MAPK pathway upstream at the level of receptor activation, we wanted to -determine the effect of expressing different downstream MAPKs on the Argos phenotypes. To accomplish this, we expressed both phosphorylated MAPK (Rl^SEM), and nuclear MAPK (Rl.nls) while simultaneously expressing Argos in the posterior -compartment of wing discs. Expression of phosphorylated MAPK (Rl^SEM) within an Argos background showed no significant difference in cell density compared to flies expressing only Rl^SEM alone (Fig. 4B). These flies did, however, show a significant decrease in cell density compared to flies that only express Argos within their posterior compartments (Fig. 4B). These data, while not significant, do suggest a trend towards a smaller cell density (larger cells) in flies that express both Argos and Rl^SEM. In contrast to these results, increasing the amount of nuclear MAPK in wings that simultaneously express Argos did not show a significant increase or decrease in cell density (Fig. 4B), suggesting that the cell density phenotype observed in Argos wings may be saturated, and unable to be enhanced. Taken together, these data suggest that the trend towards decreased cell density in wings that overexpress Argos is specific for expression of the -phosphorylated form of MAPK.

We next analyzed posterior compartment size in these flies. Expression of phosphorylated MAPK (Rl^SEM) showed a significant increase in the posterior compartment size of wings expressing both Argos and Rl^SEM (Fig. 4C). However, expression of nuclear MAPK (Rl.nls) did not significantly increase posterior compartment size when co-expressed with Argos (Fig. 4C). Taken together, these data suggest that by specifically increasing the phosphorylated form of MAPK within the posterior compartment of these wings (as opposed to nuclear MAPK), we are partially rescuing the defective cell growth and division phenotypes normally observed when we express Argos.

The results illustrated in the fly wing suggest that a decrease in the phosphorylated form of MAPK (due to increased expression of the inhibitory ligand Argos) leads to an increase in cell density with a concomitant decrease in cell and organ size. We suggest the data presented in the fly wing is consistent with the hypothesis that it is an increase in the nuclear non-phosphorylated form of MAPK that leads to the observed increase in cell density and decrease in posterior compartment size in these wings. These results are consistent with previous results, suggesting that the forced nuclear translocation of MAPK could also lead to an increase in cell -division with a concomitant decrease in cell growth.¹⁶

Nuclear, unphosphorylated MAPK can induce cell division in eye tissue

To test whether increased nuclear MAPK could induce cell division in tissues outside the Drosophila wing, we turned to the developing Drosophila retina. We used the eyeless-Flip reagent to create clones of cells mutant for the Drosophila MAP Kinase Kinase (MEK, encoded by the dSor1 gene in flies)^36,37 in the developing Drosophila retina. Clones of dSor1 mutants have been previously shown to have severe defects in the rate of cell proliferation, showing only one or two doubling events compared to their wild-type twin clones.³⁸ When we used the null mutant dSor1^LH110 to create clones in the developing retina, we also observed very small clones (data not shown). However, to generate clones of sufficient size in which to analyze cell proliferation, we utilized the loss-of-function mutation dSor1^S-1221,²³ which generated clones large enough for analysis of individual markers of cell growth and division within these clones (Fig. 5). Loss of dSor1 function within these mutant clones is expected to generate a loss of pMAPK staining, with a concomitant decrease in photoreceptor differentiation, cell growth and cell division. In order to determine decreased cell growth, we analyzed the expression of the Capicua protein, which has been shown to restrict cell growth, and whose expression is decreased in response to Ras/MAPK signaling.³⁹ In clones of both Egfr and Ras in the developing retina, Capicua protein levels are increased, indicating a decrease in cell growth in these cells.³⁹ Similarly, Capicua levels are increased in clones of dSor1^S-1221, confirming a decrease in cell growth within these clones (Fig. 5A and B). Photoreceptor differentiation is decreased in these clones (Fig. 5C and D), and the phosphorylated form of MAPK (pMAPK) is absent within these clones as well (Fig. 5E and F). Cell division, as measured by counting the number of mitotic events stained with pH3 within clones was also decreased (Fig. 5G and H, clones within the second mitotic wave in the developing eye), and averaged roughly 0.03 mitoses per clone (Fig. 5I).

Figure 5. Nuclear, unphosphorylated MAPK increases cell division in Drosophila retinas. (A–H) Show third instar retinas, anterior up (A–D), anterior right (E–H). All retinas show clones of dSor1^S-1221 marked by the absence of GFP (green). (A) Shows Capicua staining (blue) is increased in dSor1^S-1221 clones. (B) Shows Capicua (white) from (A). (C) Shows ELAV (red, a marker of photoreceptor differentiation) is decreased in dSor1^S-1221 clones. (D) Shows ELAV (white) from (C). (E) Shows phosphorylated MAPK (pMAPK, red) is absent in dSor1^S-1221 clones. (F) Shows pMAPK (white) from (D). (G) Shows phosphorylated Histone H3 (pH3, blue) is absent from dSor1^S-1221 clones in the second mitotic wave. (H) Shows pH3 (white) from (G). (I) Shows the average number of pH3 positive nuclei in dSor1^S-1221 clones alone (first column), in dSor1^S-1221 clones that also express full length MAPK within the clone (hs-Rl, second column), and in dSor1^S-1221 clones that also express full length MAPK with a strong nuclear localization signal attached to it (hs-Rl.nls, third column). Error bars represent Standard Error. ** Indicates p < 0.01.

In order to determine if increased nuclear MAPK could -activate cell division even in the absence (or severe diminishment) of its phosphorylation by MEK, we generated dSor1^S-1221 mutant clones in the developing retina, and expressed both the normal form of the MAPK protein, as well as a constitutively nuclear form of the MAPK protein within these clones (hs-MAPK and hs-MAPK.nls respectively^15,16). We found that though the markers of cell growth, differentiation and phosphorylated MAPK were unchanged in these clones, clones in which we expressed the nuclear form of MAPK showed a significant increase in the number of mitoses (pH3 nuclei) as compared to clones in which we expressed normal MAPK protein, or clones in which we did not express either (Fig. 5I). When taken together with both our previous results,¹⁶ and the results described above in the fly wing, these results suggest that the phosphorylation state of the MAPK protein is dispensable for cell division in developing Drosophila retina and wing tissue, but is required for cell growth and -differentiation in these tissues.

Discussion

Understanding how the same canonical pathway is capable of eliciting multiple distinct cellular processes in both a tissue specific and temporally distinct manner remains an important problem in biology. The results we present in this study reveal a novel method of regulating the Ras/MAPK pathway, and suggest that the non-phosphorylated form of MAPK is sufficient to induce cell division, but not cell growth, once inside the nucleus of the cell (Fig. 6B).

What might be the mechanism for nuclear import of -non-phosphorylated MAPK? Monomers of MAPK have been previously shown to passively diffuse into the nucleus in Xenopus.⁴⁰ Further, literature points out that the non-phosphorylated form of MAPK can enter the nucleus in an energy-independent fashion by directly interacting with nuclear pore complex proteins.^41,42 Though phosphorylation of MAPK was inhibited successfully in our model system in both the eye and the wing, if the normal synthesis and degradation of MAPK is unperturbed in these cells, this might cause an abnormally high concentration of non-phosphorylated MAPK in the cytoplasm. This accumulation may, in turn, cause an increase in the concentration gradient of non-phosphorylated MAPK across the nucleus, facilitating the passive nuclear diffusion of -non-phosphorylated MAPK in these cells.

Perhaps the most interesting results of our present study are that once the non-phosphorylated MAPK does move into the nucleus, it appears to have a critical cellular function in promoting cell division there. How could this non-phosphorylated MAPK cause the activation of cell division in Drosophila -tissues? Results in both flies and mammals suggest that both cell division and cell growth are regulated by MAPK activation and nuclear localization.^6,16,43,44 Our results show that cell growth was impaired in both wings and eyes on inhibition of MAPK phosphorylation, consistent with the idea that MAPK phosphorylation is necessary to activate cell growth-related processes in developing wings and eyes of Drosophila. However, our results also suggest that phosphorylation of MAPK is not required to activate cell division. Once in the nucleus, MAPK is normally thought to phosphorylate transcription factors that subsequently promote cell division and growth. Non-phosphorylated MAPK could not facilitate transcription factor activation through phosphorylation however. Thus, this form of MAPK must be able to activate cell division by another mechanism. Recent studies have shown that ERK2 can directly bind to DNA and function as a transcriptional repressor for interferon gamma induced genes.⁴⁵ Thus, it may be that Drosophila MAPK, once inside the nucleus, is able to directly bind to DNA and regulate the expression of gene(s) critical for cell division.

Targeted inactivation of the Egf Receptor is an effective and promising therapy for many cancers,⁴⁶ but their efficacy in some solid tumors, and in brain tumor glioblastoma multiforme (the most prevalent tumor in the central nervous system of human adults) has been low.^46,47 Previous research has shown that co-activation of multiple different RTKs within some of these tumors is responsible for some of the inefficacy of single RTK inhibitor treatment in these tumors.⁴⁷ Our results shown here suggest an additional level of regulation, whereby simply inactivating the Ras/MAPK pathway is not sufficient to limit MAPK‑mediated cell division.

Materials and Methods

Drosophila stocks and genetics

Unless otherwise noted, all crosses were carried out at 25°C on standard cornmeal-molasses-agar medium. BL# refers to Bloomington Stock Center stock number. Stocks used are described: en:Gal4; UAS:GFP,¹⁹ UAS:Argos,²⁰ UAS:sSpi,²¹ UAS:Rl-SEM;²² dSor1^S-1221,²³ -eyeless-FLP (ey:FLP),²⁴ UAS:Rl.nls (MAPK overexpression under -control of the Gal4 system), hs-Rl (MAPK overexpression under heat shock control) and hs-Rl.nls (nuclear MAPK expression under heat shock control).^15,16

The Gal4/UAS system was used for the overexpression of Argos, sSpi and Rl-SEM as described.²⁵ Mosaic clones were generated using ey:FLP,²⁴ as described.²⁶ Flip stock y^-, w^-, Ubi-GFP, P{neoFRT}19A; ey:FLP (on 2^nd Chromosome) were crossed to the following stocks as appropriate to generate mutant clones:

(1) dSor1^S-1221, P{neoFRT}19A/FM7C

(2) dSor1^S-1221, P{neoFRT}19A/FM7C; hs-Rl/hs-Rl

(3) dSor1^S-1221, P{neoFRT}19A/FM7C; hs-Rl.nls/hs-Rl.nls.

In order to express both MAPK and nuclear MAPK within dSor1 clones, late third instar larvae were heat shocked at 37°C for 1 hour, were then allowed to recover for 1 hour, and eye imaginal discs were then dissected and immediately fixed and stained for mitotic markers. A minimum of 10 larval retinas were examined for each genotype, and a minimum of 50 different clones were examined total for each genotype.

Immunohistochemistry

Wing and eye imaginal discs were dissected, fixed and prepared as described earlier.²⁷ The wing discs were mounted in vectashield (Vector Labs, H-1000). Immunohistochemistry imaging was done using a Nikon TE-2000 laser scanning confocal microscope. The primary antibodies used were mouse pMAPK (Sigma #M8159, 1:200 for wing imaginal discs and 1:250 for eye imaginal discs), rabbit anti-phosophorylated histone-3 (pH3, Cell Signaling Technologies # 9701, 1:1,000), Rabbit anti-activated caspase3 (BD Biosciences #551150, 1:200), Rat anti-ELAV (Developmental Studies Hybridoma Bank #7E8A10, 1:1,000), and guinea pig anti-capicua (a gift from Iswar Hariharan, University of California, Berkeley, 1:300, as described²⁸).

Secondary antibodies used were goat anti-mouse TRITC (for wing imaginal discs, # 115-116-072, 1:150), goat anti-rabbit TRITC (for eye imaginal discs, # 111-116-144, 1:250), goat -anti-rabbit Cy5 (for wing imaginal discs, #111-176-144, 1:1,000), goat anti-rat Cy5 (for eye imaginal discs, # 712-176-150, 1:200), goat anti-mouse Cy5 (eye imaginal discs, # 115-176-072, 1:500). All above mentioned secondary antibodies were from Jackson ImmunoResearch. HRP antibodies used were Goat anti-rabbit HRP (General Bioscience # &nbspGR2201, 1:100), and goat anti-mouse HRP (General Bioscience # &nbspGM5201, 1:40).

Adult wing analysis

Adult wings were dissected and dehydration was done in 70% ethanol for one minute. Mounting was done using DPX mountant (Fluka corporation, Sigma-Aldrich # 44581). Cell count assays were performed using phase contrast microscopy, and wing surface area was measured using pixel counts through ImageJ software, as previously described.²⁹ For each genotype a minimum of five wings were independently analyzed. For surface area compartment measurements, posterior compartments were defined as the area below wing vein L3, while anterior was the area above for all genotypes. For cell density, wing hair counts were analyzed in an area distal to posterior cross-vein and below vein L5, while anterior measurements were above wing vein L2 in all genotypes.

Statistical analysis

All statistical analyses were performed on SPSS version 17. To determine significance between multiple different genotypes, a one-way ANOVA analysis was performed with Tukey posthoc analysis. Genotype is the independent variable while actual wing measurements or hair counts were the dependent variable. To determine significance between two different genotypes, a two-tailed Student’s t-test was performed between the two groups. Significance was determined at the 95% confidence interval.

Acknowledgements

We would like to thank Iswar Hariharan for the Capicua antibody, the Bloomington Stock Center, and the Iowa Hybridoma Bank for reagents. We would additionally like to thank members of the Marenda lab, and two anonymous reviewers for helpful suggestions. This work was supported by a grant from the NIH, R15EY018431-01A1 (to Daniel R. Marenda).