Extracellular signal-regulated kinase (ERK) 5 is necessary and sufficient to specify cortical neuronal fate

Abstract

Multipotent cortical progenitor cells differentiate into neurons and glial cells during development; however, mechanisms governing the specification of progenitors to a neuronal fate are not well understood. Although both extrinsic and intrinsic factors regulate this process, little is known about kinase signaling mechanisms that direct neuronal fate. Here, we report that extracellular signal-regulated kinase (ERK) 5 is expressed and active in proliferating cortical progenitors. Lentiviral gene delivery of a dominant negative ERK5 or dominant negative MAP kinase kinase 5 reduced the number of neurons generated from rat cortical progenitor cells in culture, whereas constitutive activation of ERK5 increased the production of neurons. Furthermore, when cortical progenitor cells were treated with ciliary neurotrophic factor, which induces precocious glial differentiation, ERK5 activation still promoted neuronal fate while suppressing glial differentiation. Our data also indicate that ERK5 does not directly regulate proliferation or apoptosis of cultured cortical progenitors. We conclude that ERK5 is necessary and sufficient to stimulate the generation of neurons from cortical progenitors. These results suggest a previously uncharacterized function for ERK5 signaling during brain development and raise the interesting possibility that extrinsic factors may instruct cortical progenitors to become neurons by activating the ERK5 pathway.

Multipotent neural progenitor cells can differentiate into neurons or glia depending on developmental cues and environmental signals. Neuronal differentiation proceeds by a two-step process: the initial commitment of the progenitors to a neuronal fate, followed by cell cycle exit and terminal differentiation of the committed precursors into mature neurons (1). The specification of cortical progenitors to a neuronal fate requires a coordinated expression of the proneural basic helix–loop–helix (bHLH) transcription factors (2, 3). In addition to this intrinsic program, extrinsic factors also regulate fate choice of neural progenitor cells (4, 5).

ERK5 is a member of the mitogen-activated protein (MAP) kinase family (6). In neurons, ERK5 is activated by neurotrophins including BDNF but not by Ca²⁺ or cAMP (7–9). ERK5 is phosphorylated and activated by MAP kinase kinase (MEK) 5 but not by MEK1 or MEK2 (6). MEK5 is specific for ERK5 and does not phosphorylate or activate other MAP kinases (7, 10).

ERK5 is widely expressed in many tissues with the highest levels in brain (11, 12). In this study, we performed a detailed analysis of ERK5 expression during rat cortical development and discovered that it is expressed as early as embryonic day (E) 11 and is active during cortical neurogenic period. Furthermore, ERK5 is expressed in nestin⁺ and BrdU⁺ cortical progenitor cells in cultures. Its abundant expression in proliferating progenitor cells suggests that ERK5 may play a role in cortical progenitor cell biology that may include proliferation, survival, or cell fate specification. To address these issues, three independent assays were used to evaluate the function of ERK5 in cortical progenitors: a neurosphere-forming assay, a clonal assay, and a standard adherent culture monolayer assay. Our data demonstrate that ERK5 does not directly regulate apoptosis or proliferation of cortical progenitors. However, ERK5 is necessary and sufficient to provide an instructive signal for cortical progenitors to become neurons. These results provide insights concerning molecular mechanisms governing fate choices of neural progenitors and identify a biological function for ERK5.

Results

ERK5 Is Expressed in Proliferating Cortical Progenitor Cells.

Our previous studies using RNA and protein analysis of whole brain showed a higher expression of ERK5 in embryonic brain than in the postnatal brain (12). Here, we performed a more detailed analysis of the developmental expression of ERK5 in the rat cortex. Immunostaining for ERK5 indicated significant expression of ERK in the cortex as early as E11, the earliest time examined (Fig. 1A). High expression of ERK5 in the cortex during embryonic development was confirmed by Western blot analysis (Fig. 1B). Interestingly, endogenous ERK5 in cortical lysates exhibited a reduced electrophoretic mobility (phosphorylation gel shift), indicative of its activation (7, 13). Phospho-ERK5 was readily detectable in the cortex at E12 and peaked at E13 when cortical neurogenesis begins. Furthermore, phospho-ERK5 was elevated during the entire embryonic cortical development and decreased postnatally. Thus, ERK5 is not only highly expressed but also activated in the cortex during the neurogenic period of cortical development.

Fig. 1.

ERK5 is expressed in proliferating cortical progenitor cells. (A) Confocal images of ERK5 immunostaining in rat cortex. Control, anti-ERK5 antibody preabsorbed with recombinant ERK5 protein. (B) Western blot analysis of ERK5 expression in rat cortex from E12, to postnatal day (P) 0, to adult. β-Actin was used as a loading control. ERK5 is expressed in BrdU⁺ (C) and nestin⁺ (D) cortical progenitor cells in culture. E13 cortical progenitor cells were plated as monolayer cultures. On the second day in culture (DIV2), cells were pulsed with 10 μM BrdU for 4 h to label S phase cells.

Because virtually all cells in the E11 rat cortex are proliferating progenitor cells that have not committed to a specific lineage, the data in Fig. 1A suggested that ERK5 is expressed in proliferating cortical progenitor cells. To directly confirm the expression of ERK5 in proliferating progenitors, we prepared cortical progenitor cell cultures from E13 rats. These cultured cells were treated with BrdU for 4 h and stained with an anti-BrdU antibody to identify actively proliferating S phase cells. In addition, nestin staining was used as a marker to identify progenitor/precursor cells. ERK5 was expressed in both BrdU⁺ and nestin⁺ cells (Fig. 1 C and D), confirming that ERK5 is expressed in actively proliferating progenitor cells.

Effect of ERK5 on the Specification of Neuronal Fate by a Neurosphere-Forming Assay.

The expression profile of ERK5 suggests that it might play a role in neuronal cell fate determination. Neural stem/progenitor cells, but not cell fate-committed precursor cells, have the capacity to form neurospheres in suspension culture (14). Cortical progenitor cells were isolated from E13 rats and plated at a clonal density to allow neurosphere formation (15). Although these cells propagate in the presence of bFGF and/or EGF and can be expanded in vitro, we only used freshly prepared cultures to better mimic progenitor cell conditions in vivo. Because early born cortical progenitors respond to bFGF, whereas later born progenitors respond to EGF (16), we added bFGF to the culture medium to select for a more homogeneous population of cortical progenitor cells.

The freshly dissociated progenitor cells were infected with various lentiviral stocks encoding activating or dominant interfering forms of constructs for the ERK5 signaling pathway. These genes were coupled to internal ribosomal entry site (IRES)-GFP so that lentivirus-infected cells could be easily identified by GFP expression. The corresponding biochemical activities of these viral stocks were confirmed in primary cortical neurons (see Supporting Text and Fig. 8, which are published as supporting information on the PNAS web site). Lentiviral infection was carried out 3 h after initial plating while the cells were still at the single-cell level in suspension. Neurospheres infected with wild-type (wt) ERK5 together with constitutive active (ca) MEK5, which activates ERK5 in infected cells, contained many neurons (βIII-tubulin⁺) with elaborate processes (Fig. 2A). In contrast, neurospheres expressing dominant negative (dn) MEK5, which inhibits endogenous ERK5 signaling, had significantly fewer βIII-tubulin⁺ cells. We defined those neurospheres with fewer than 10 βIII-tubulin⁺ cells per neurosphere as nonneuronal spheres. Quantitation of the data demonstrated that activation of ERK5 signaling substantially decreases, whereas blocking ERK5 signaling increases, the number of nonneuron spheres (Fig. 2B).

Fig. 2.

ERK5 is necessary and sufficient to instruct cortical progenitors to a neuronal fate in a neurosphere-forming assay. Freshly dissociated E13 cortical progenitor cells were infected with lentiviruses encoding GFP control (control), caMEK5 plus wtERK5, or dominant negative dnMEK5 where indicated. The dissociated single cells were allowed to form neurospheres for 5 days in the presence of bFGF and then transferred to coated dishes to allow differentiation for 2–3 days in the absence of bFGF. (A) Representative neurospheres infected with lentiviruses and double immunostained with an anti-GFP antibody (green) to identify viral infected neurospheres, and an anti-βIII-tubulin antibody (red) to identify newly generated neurons. (B and C) Quantification of the data. Nonneuron spheres in B were defined as those with fewer than 10 βIII-tubulin⁺ cells.

We performed a more detailed analysis of the effect of ERK5 activation on the cell fate of cortical progenitor cells by scoring the percentage of neurons within each sphere. The majority (72%) of control GFP viral-infected spheres contained ≤10% neurons/sphere (Fig. 2C). Expression of caMEK5 with wtERK5 significantly increased the number of spheres containing a higher percentage of neurons and decreased the number of spheres with fewer neurons. For example, the number of spheres with ≤10% neurons dropped from 72% in the controls to 6%. Accordingly, whereas none of the control-infected spheres had ≥30% neurons, more than half of the spheres infected with caMEK5+wtERK5 did. This result suggests that ERK5 may play a critical role in the specification of cortical progenitors to a neuronal fate.

Effect of ERK5 on the Specification of Cortical Progenitors to a Neuronal Fate in an Adherent Culture Clonal Assay.

In addition to the sphere-forming assay, we developed a clonal assay under adherent culture conditions to examine the effect of ERK5 on neuronal differentiation at the single progenitor cell level. LeX is a carbohydrate found on the cell surface of pluripotent neural stem cells (17). Temple and colleagues (17) demonstrated that neurosphere-generating neural stem cells can be highly enriched by labeling cells with an anti-LeX antibody coupled with flow cytometry analysis (FACS). We modified this procedure and used magnetic activated cell sorting to highly enrich the LeX⁺ progenitor population in the cultures prepared from E13 rat cortex (see Supporting Text and Fig. 9, which are published as supporting information on the PNAS web site).

Fresh LeX-sorted cells were infected with lentivirus for 20 h. The cells were then dissociated into single cells and mixed at a 1:200 ratio with noninfected E13 cortical progenitor cells. The mixture was plated as a monolayer at low density in coated tissue culture dishes in bFGF-containing culture medium. Under these conditions, each single progenitor cell proliferated to form a “clone” of daughter cells in cluster. The clones from virus-infected progenitor cells expressed GFP and were well separated from each other by noninfected cells. Thus, the GFP⁺ clones, each originating from a single LeX⁺ progenitor cell, could be easily identified and scored (see Fig. 10, which is published as supporting information on the PNAS web site). This assay allowed us to specifically follow the cell fate of a single LeX⁺ cortical progenitor cell in culture. Because these progenitor cells were prepared from E13 rat cortex and were only in culture for several days, most of the clones only differentiated into neuron-containing clones, and we seldom observed clones with GFAP⁺ astrocytes. This finding is consistent with the notion that neurogenesis precedes gliogenesis and cortical progenitors primarily differentiate into neurons at E13 both in vivo and in culture (18).

Using this clonal assay, we quantified the percentage of neurons in each progenitor clone, which were defined as those clones having both neurons (βIII-tubulin⁺ or MAP2⁺, markers for immature and mature neurons, respectively) and nonneurons (Fig. 3A). Lentiviral expression of dnMEK5 or dnERK5 significantly decreased the number of MAP2⁺ or βIII-tubulin⁺ neurons (Fig. 3 B and C). In contrast, activation of ERK5 by lentiviral expression of caMEK5 alone or together with wtERK5 increased the number of neurons. We also examined the influence of ERK5 on neuron differentiation using NeuN staining, another marker for mature neurons (Fig. 3D). In control GFP virus-infected clones, ≈25% of the clones contained one or more NeuN⁺ cells (NeuN⁺ clones). Almost all clones (96%) infected with dnMEK5 lentivirus had no NeuN⁺ cells (NeuN⁻ clones). In contrast, lentiviral expression of caMEK5+wtERK5 increased the number of NeuN⁺ clones from 25% in the controls to ≈50%. These data are consistent with those obtained with the neurosphere-forming assay and demonstrate that endogenous ERK5 activity is required for the specification of cortical progenitors to a neuronal fate. Furthermore, ectopic ERK5 activation is sufficient to stimulate cortical progenitors to differentiate into neurons.

Fig. 3.

ERK5 is necessary and sufficient to instruct cortical progenitors to a neuronal fate in an adherent culture clonal assay. (A) A representative precursor cell clone that contains both neurons (MAP2⁺; arrows) and nonneurons (MAP2⁻). (B–D) Effect of ERK5 on neuronal differentiation of cortical progenitors using three independent markers for neurons: MAP2 (B), βIII-tubulin (C), and NeuN (D). NeuN⁺ clones, clones that have at least one NeuN⁺ cell; NeuN⁻ clones, clones that have no NeuN⁺ cells.

Effect of ERK5 on the Differentiation of Cortical Progenitors to Neurons in an Adherent Culture Monolayer Assay.

We also used a standard adherent culture monolayer assay by directly plating E13 cortical progenitor cells as a monolayer on coated dishes. Infection with wtERK5 alone or together with caMEK5 was sufficient to increase the number of cells staining positive for βIII-tubulin in GFP⁺ virus-infected cells (Fig. 4A). In contrast, expression of either dnERK5 or dnMEK5 reduced the number of βIII-tubulin⁺ cells. Because the infection rate was high, we also quantified the percentage of βIII-tubulin⁺ neurons in the entire population and came to the same conclusion as when only infected cells were scored (Fig. 4B). Similar results were obtained when neurons were identified with markers for mature neurons NeuN⁺ (Fig. 4C) or MAP2⁺ (Fig. 4D).

Fig. 4.

Effect of ERK5 on neuronal differentiation of E13 cortical progenitors in a standard adherent culture monolayer assay. (A) Percentage of βIII-tubulin⁺ neurons in infected cell population. (B) Percentage of neurons in the entire population. (C) Percentage of NeuN⁺ neurons in the entire population. (D) Percentage of MAP2⁺ neurons in infected cell population. The cells in this experiment were allowed to differentiate for 6 days instead of the usual 3–4 days to better investigate the inhibitory effect on neuronal differentiation. Hence, the percentage of neurons in control virus-infected culture was higher.

Cells That Do Not Become Neurons When ERK5 Is Inhibited Remain in the Proliferative Stage.

To begin to map the fate of cortical progenitors in which neuronal differentiation was inhibited by blocking ERK5 signaling, we stained cells for GFAP, an astroglial marker. We did not observe any GFAP⁺ cells using the adherent culture clonal assay or the monolayer assay (data not shown). Furthermore, Western blot analysis indicated that there was no detectable GFAP expression in E13 cortical progenitor cells with or without control virus infection (Fig. 5A), consistent with the notion that there is very little spontaneous gliogenesis under these conditions. Expression of dnMEK5 did not increase GFAP protein. Thus, inhibition of ERK5 signaling under conditions permissive for neurogenesis does not seem to cause precocious astroglial differentiation. Furthermore, after cortical progenitor cells were differentiated for 3 days in the clonal assay, clones infected with dnMEK5 lentivirus had a higher number of BrdU⁺ cells than control virus-infected clones (Fig. 5B). These data suggest that cortical progenitor cells that did not become neurons when ERK5 was inhibited kept proliferating. In contrast, clones infected with caMEK5+ERK5 lentiviruses had a dramatic decline in the number of BrdU⁺ cells as a result of differentiation.

Fig. 5.

Cortical progenitors that do not become neurons when ERK5 is inhibited stay proliferative. (A) Expression of dnMEK5 does not increase GFAP expression as analyzed by Western blotting. LeX-sorted E13 cortical progenitor cells were infected with GFP control or dnMEK5 lentivirus and grown in the standard monolayer culture for 4 days. Cells treated with CNTF (50 ng/ml for 3 days) were used as a positive control for GFAP expression. (B) BrdU incorporation at the end of 3-day differentiation in a clonal assay. After 3 days of proliferation followed by 3 days of differentiation in a clonal assay, cells were pulsed with 20 μM BrdU for 4.5 h before fixing and immunostaining for BrdU and GFP.

ERK5 Instructs Neuronal Cell Fate Determination at the Expense of Glial Differentiation.

The experiments described above indicate that activation of ERK5 promotes neuronal cell fate under neurogenic conditions. Can ERK5 specify a neuronal fate when cortical progenitors are placed under gliogenic conditions? To address this issue, we applied ciliary neurotrophic factor (CNTF) to the culture when bFGF was removed from the medium in the adherent culture clonal assay. CNTF is a growth factor that promotes astroglial differentiation. Astroglial differentiation was monitored by immunostaining for GFAP. As anticipated, the number of GFAP⁺ clones increased from undetectable in the absence of CNTF to ≈50% after treatment with CNTF for 4 days (Fig. 6A). In addition to an increase in GFAP⁺ cells, a large majority of the clones in the presence of CNTF contained some cells expressing nestin (data not shown) and some cells expressing βIII-tubulin (Fig. 6B). This finding suggests that these clones originated from actively proliferating, multipotent progenitor cells that have the potential to become either neurons or glia under appropriate conditions.

Fig. 6.

ERK5 promotes neurogenesis and inhibits gliogenesis in the presence of CNTF. E13 cortical progenitor cells were subjected to the adherent culture clonal assay. Freshly isolated, Lex-sorted cells were infected with lentiviruses for 20 h, mixed with noninfected cells, cultured for 1 more day in bFGF-containing medium, and treated with or without 50 ng/ml CNTF for an additional 4 days in bFGF-free medium. (A) Effect of ERK5 activation on astroglial differentiation. (B) Effect of ERK5 activation on neuronal differentiation.

Interestingly, the CNTF-induced precocious glial differentiation was suppressed by ectopic ERK5 activation through lentiviral delivery of caMEK5+wtERK5 (Fig. 6A). Similar results were obtained when S100-β was used as a marker for glial precursor cells and immature astrocytes (data not shown). In contrast, inhibition of endogenous ERK5 activity by dnMEK5 enhanced CNTF-induced gliogenesis. Ectopic ERK5 activation was sufficient to increase the number of βIII-tubulin⁺ clones even in the presence of CNTF (Fig. 6B). These data illustrate that ERK5 signaling promotes neuronal fate commitment, rather than simply enhancing or accelerating terminal differentiation of precursors that have already committed to a neuronal fate. Furthermore, ERK5 seems to instruct neuronal fate specification at the expense of glial differentiation.

ERK5 Activation Does Not Immediately Affect Cell Cycle Exit or Survival of Cultured Cortical Progenitor Cells.

To rule out the possibility that the effects of ERK5 activation on neuronal differentiation results indirectly from an increase in immediate cell cycle exit or cell survival, rather than from a direct effect on neuronal fate specification, we examined the effect of ERK5 activity on progenitor cell proliferation and survival. The LeX-sorted, virus-infected cells in the adherent culture clonal assay were extremely healthy with very few apoptotic cells (Table 1). We also examined the effect of ERK5 on proliferation or apoptosis when progenitor cells were directly plated as a monolayer (Fig. 7). Similar to the clonal assay, <1.2% of the cells were apoptotic, and no significant differences were observed among cells infected with various lentiviruses (Fig. 7A).

Table 1.

Percentage of apoptotic cells in infected clones in the adherent culture clonal assay

Infected clones	% Apoptosis
Control	<0.3
caMEK5 + wtERK5	<0.2
dnMEK5	<0.3
dnERK5	<0.2

Fig. 7.

ERK5 does not directly affect proliferation or survival of E13 cortical progenitor cells. E13 cortical progenitors were cultured as a standard monolayer, infected on DIV1 with lentivirus for 18 h, and placed in fresh media. (A) Percentage of apoptotic cells in infected cell population (GFP⁺). (B) Percentage of BrdU⁺ cells in infected cell population (GFP⁺). Cells were pulsed on DIV2 with 10 μM BrdU for 6 h in bFGF-containing proliferation medium. (C) Effect of ERK5 on [³H]thymidine incorporation, measured as described in ref. 27, in the presence or absence of mitogen (10 ng/ml each of bFGF and EGF).

To determine whether ERK5 directly regulates cell cycle exit, we measured BrdU incorporation in cultured cortical progenitor cells at day in vitro (DIV) 2 under proliferative conditions. Neither activating nor blocking ERK5 signaling caused a significant change in BrdU labeling (Fig. 7B). Because of the high infection rate with lentivirus, we also measured the effect of ERK5 on cell proliferation using [³H]thymidine incorporation both in the presence and absence of mitogen (bFGF/EGF). Expression of wtERK5 or dnERK5 had no effect on thymidine incorporation under either condition (Fig. 7C). These data suggest that ERK5 signaling does not directly regulate cell proliferation or modulate survival of E13 cortical progenitor cells.

Discussion

The objective of this study was to examine the role of ERK5 signaling in the regulation of cell fate specification of cortical progenitor cells. Although the brain shows the highest levels of ERK5 expression, its function(s) in the brain have not been completely defined (9, 12). In this study, we documented abundant expression of ERK5 in proliferating cortical progenitor cells at E11 in rats, before the peak of cortical neurogenesis. Furthermore, endogenous ERK5 is activated during the entire period of cortical neurogenesis, starting from E13. This suggested that ERK5 may play a critical role during cortical neurogenesis. Indeed, we report evidence from three independent assays that ERK5 is both necessary and sufficient to promote the commitment of cortical progenitor cells toward a neuronal fate. This study defines a function for ERK5 in the specification of neuronal fate.

We used freshly dissociated cortical cells from E13 rats in this study because they are actively proliferating, bFGF-responsive, early progenitor cells. In addition, we infected freshly dissociated cells with various lentiviruses while the cells were still well separated at the single-cell level. The infected cells were plated at a clonal density in the neurosphere-forming and the clonal assays to study the cell fate of single progenitor cells. The clones in these two assays originated from actively proliferating, multipotent progenitors. Thus, the observed effect of ERK5 on neurogenesis is most likely due to its action on progenitors.

We expressed wtERK5 alone, caMEK5 alone, or the two together to ectopically activate ERK5 signaling. Inhibition of ERK5 signaling was achieved by expression of dnERK5 or dnMEK5. Neurons were identified by three independent markers to label both newly generated, immature neurons (βIII-tubulin⁺) and mature neurons (MAP2⁺ and NeuN⁺). Using the neurosphere-forming and adherent culture clonal assays, we found that ectopic activation of ERK5 increases the number of clones containing a high percentage of neurons and decreases the number of clones containing few neurons. In contrast, inhibition of endogenous ERK5 greatly increased the number of clones with no or few neurons, indicating suppression of spontaneous neuronal differentiation. The effect of ERK5 on neuronal differentiation was also demonstrated by using a standard adherent culture monolayer assay. The proneural effect of ERK5 was also apparent even when cortical progenitor cells were challenged under precocious glial differentiation conditions. Activation of ERK5 was sufficient to induce neuronal differentiation at the expense of glial differentiation, even under CNTF treatment; inhibition of ERK5 enhanced CNTF-stimulated gliogenesis. Together, these data strongly suggest that ERK5 activation plays an instructive role in neuron fate specification of cortical progenitor cells.

MEK5 and ERK5 null mice have a very similar phenotype; both die around E9.5–E10.5 because of cardiovascular defects (11, 19–21). Although high levels of apoptosis (TUNEL⁺ cells) are observed in their heads, there is no evidence that apoptosis occurred in neurons or neural progenitors. In fact, increased apoptosis in the heads of ERK5 null mice is attributed to apoptosis in the cephalic mesenchyme tissue (11). Our in vitro data suggest that ERK5 does not affect the survival of E13 cortical progenitor cells.

We considered the possibility that the proneural effect of ERK5 is due to its ability to directly promote cell cycle exit and terminal differentiation of the precursor cells that have already committed to a neuronal lineage. ERK5 has been implicated in cellular proliferation in cultured cell lines (22). However, no significant difference has been observed in the ability of mouse embryonic fibroblasts derived from MEK5^−/− and its wt littermates to progress from G₁ through S phase (21). Thus, the role for ERK5 signaling in proliferation may be cell type-dependent. In this study, we show that neither activating nor inhibiting ERK5 has an immediate effect on DNA synthesis of E13 cortical cultures. However, as a result of changes in cell fate and differentiation, clones with ectopic ERK5 activation had largely exited cell cycle by the end of a 3-d differentiation program. Thus, the effect of ERK5 on neuronal differentiation is most likely due to its direct action on progenitor cells, driving them toward a neuronal fate. When ERK5 is inhibited, cortical progenitor cells seem to stay in the proliferative stage, rather than undergoing gliogenesis.

In addition to the developmentally programmed intrinsic factors, a large body of evidence suggests that environmental cues and extrinsic factors also play an important role in cell fate determination of neural progenitors (23). In neurons, ERK5 is activated by neurotrophins including BDNF, neurotrophin 3, and neurotrophin 4 (7–9). Thus, extrinsic factors may instruct cortical progenitors to become neurons by activating the ERK5 pathway.

In summary, we have discovered a function of the ERK5 signaling pathway in the regulation of neuronal cell fate specification of cortical progenitor cells. We hypothesize that ERK5 promotes neuronal fate specification and may thereby contribute to cortical development.

Materials and Methods

Reagents.

The lentiviral transfer vector pRRL-cPPT-CMV-X-PRE-SIN is described in ref. 24. The polyclonal anti-ERK5 antibody is described in ref. 7. The following antibodies for immunostaining were purchased commercially: a monoclonal (M1) anti-flag antibody (Sigma), monoclonal anti-nestin (Becton Dickinson), monoclonal anti-BrdU (Sigma), monoclonal NeuN (Chemicon), polyclonal anti-GFP (Molecular Probes), monoclonal anti-βIII-tubulin (Promega), monoclonal anti-MAP 2 (Sigma), monoclonal anti-LeX (Becton Dickinson), monoclonal anti-Ki67 (NovoCastra), and monoclonal anti-GFAP (Sigma). BDNF and Lipofectamine 2000 were purchased from Invitrogen.

E13 Cortical Progenitor Cell Cultures.

Unless otherwise specified, the culture medium consisted of Neurobasal medium (Invitrogen) supplemented with 2% B27 (Invitrogen) and 10 ng/ml bFGF (Invitrogen). Cortices were isolated from E13 Sprague–Dawley rats. We used magnetic activated cell sorting to enrich neural progenitor cells after labeling with an anti-LeX antibody.

For neurosphere-forming assays, the freshly dissociated E13 cortical progenitor cells were plated at a density of 2,000 cells per ml in Petri dishes without coating. The cells were cultured in the regular culture medium plus 1% N2 supplement. Three hours after initial plating, the single-cell suspension cultures were infected with various lentiviral stocks, and the cells were left to grow for 5 days to form neurospheres. At this time, neurospheres were transferred to culture dishes coated with laminin and poly(d-lysine) (Becton Dickinson). Two hours later, the medium was changed and replaced with fresh media without bFGF to allow neurosphere differentiation. Three days later, the cultures were fixed and stained with an anti-βIII-tubulin antibody to identify newly generated neurons.

For the adherent culture clonal assay, the freshly dissociated, LeX-sorted cells were plated on coated plates at a density of 1 × 10⁴ cells per well of a 24-well plate. Three hours after plating, cells were infected with various lentiviral stocks. Twenty hours later, the culture medium was removed, and cells were washed with Neurobasal medium and trypsinized to detach them from the culture plates. The single-cell suspensions of virally infected cells (GFP-expressing) were then mixed at a 1:200 ratio with noninfected E13 cortical progenitor cells, and the mixture was plated as monolayer culture at low density (1 × 10⁴ cells per well of a 24-well plate) on coated plates. After 1–3 days in culture in the presence of bFGF, the cells were switched to bFGF-free medium to allow spontaneous differentiation for another 3–5 days. CNTF (50 ng/ml) was added in bFGF-free medium to stimulate glial differentiation.

For the standard adherent culture monolayer assay, freshly isolated E13 cortical progenitor cells were plated on coated plates at a density of 1 × 10⁴ cells per well of a 24-well plate. At the day of plating (DIV0), the cultures were infected with lentivirus overnight. The medium was then changed and replaced with fresh media. The cells were cultured for 2 more days in the presence of bFGF and then allowed to differentiate in bFGF-free medium for 3–4 days.

Preparation of Lentiviral Stocks.

We constructed lentiviral transfer vectors with genes inserted into a multiple cloning site upstream from an IRES-directed marker protein eGFP (enhanced GFP). The IRES-GFP sequence was PCR-amplified from the pIRES2-EGFP vector (Clontech) and ligated into the XhoI–NheI site of transfer vector pRRL-cPPT-CMV-X-PRE-SIN (24). The multiple cloning site was modified by adding ApaI, EcoRV, and XbaI between the XhoI site and IRES sequence. This modified transfer vector was designated pRRL-cPPT-CMV-X-IRES-GFP-PRE-SIN. cDNAs encoding rat MEK5, human ERK5, or the mutants were subcloned into XhoI–XbaI sites of this transfer vector. All of these constructs also contain an N-terminal flag-epitope tag to allow verification of transgene expression. The replication-incompetent, VSV-G-pseudotyped lentiviral stocks were prepared by calcium phosphate-mediated three-plasmid transfection of 293T cells as described in ref. 25, the transfer vector, second generation gag-pol packaging construct pCMVΔR8.74, and VSV-G expression construct pMD.G.

Apoptosis.

Apoptotic cells were identified as those with nuclear fragmentation and/or condensation visualized after Hoechst staining (26).

Data analysis.

Results are either representative or the average of three or more independent experiments. Error bars represent SEM.

Abbreviations

ca

constitutive active

CNTF

ciliary neurotrophic factor

DIVn

day in vitro n

dn

dominant negative

En

embryonic day n

ERK

extracellular signal-regulated kinase

IRES

internal ribosomal entry site

MAP

mitogen-activated protein

MEK

MAP kinase kinase

wt

wild type