(A) NGF diminishes expression of Gata2 protein. PC12 cells were exposed to NGF over 6 days and Western immunoblots of protein lysates were carried out using anti-Gata2. Lysates from Gata2 pCMS-EGFP transfected HEK293 cells were used as a positive control. (B–D) Constitutive over-expression of Gata2 suppresses NGF-promoted neurite outgrowth. Cells were transfected with Gata2 vector and one day later treated with NGF. Cells in B and C were expose to NGF for 8 days. Values in D represent average counts of % of transfected cells with neurites in triplicate replicate cultures ± SEM. Similar results were achieved in five independent experiments. (E) Gata2 over-expression reduces the lengths of NGF-induced neurites in polyclonal PC12-derived cells constitutively expressing Gata2. Gata2 and control vectors were introduced by infection with pQNCX retrovirus. Cells were exposed to NGF for 8 days. Points represent result of individual measurements (n=240–270) of neurite lengths in arbitrary units. Horizontal lines represent mean neurite lengths. (F) Gata2 over-expression does not blunt the effect of NGF on induction of neurofilament light and peripherin proteins. Experiment was carried out in polyclonal cell lines with or without stable expression of retrovirally-introduced Gata2. NGF treatment was for 5 days.

(A) Depiction of Gata2 expression during various stages of rat embryonic midbrain development. Gata2 expression is in red, TUJ1 expression in green. Boxes in diagrams show fields magnified in the images below (B–H). (B; zoom C) The earliest Gata2-expressing cells appear in the ventrolateral midbrain at e12. (D) By e14, expression is widespread in the IZ of the ventral mesencephalon. (E,F) Expression of Gata2 extends to the dorsal mesencephalon where it persists from e15–e17 in the IZ directly beneath the neuronal layer. (G) Expression in zones of ongoing neurogenesis vanishes ventrally at e18, but remains in a layer (arrow) beneath and adjoining the neuronal SC, before (H) finally disappearing from this region by e21. Labels: v = ventricle, MGN = medial geniculate nucleus, RN = Raphe nucleus, ON = olivary nucleus, p = pial surface, SC = superior colliculus. Labels were assigned by reference to a developmental atlas (Altman and Bayer, 1995).

(A) Gata2 immunohistochemistry (red) in e15 rat mesencephalon (dorsal side at the top) labels cells in the IZ around the ventricle, except for the ventral midline (demarcated with dotted lines). The yellow bracket indicates the region where Gata2+ cell distances from the VZ were measured for panel (C). (B) Staining for tyrosine hydroxylase (TH; green) indicates that the ventromedial midbrain is the region generating dopaminergic ventral tegmental area (VTA) and substantia nigral (SN) neurons, but that this region lacks Gata2 expression (green). (C) Vertical scatter plot of three serial brain sections quantifying the distribution of Gata2+ cells within the e15 dorsal mesencephalon where zero equals the ventricular surface and 1 is the pial surface. The average means and average standard deviations of the distributions are 0.55 ± 0.16. Note that the vast majority of the Gata2-expressing population lies between the ventricular zone (VZ) and neuronal layer (NL). The VZ and NL boundaries were determined by average normalized morphological measurements from the ventricle to the edge of the VZ (determined by phase contrast) and the innermost boundary of TUJ1-expressing cells. Labels: v = ventricle, VZ = ventricular zone, p = pial surface, SC = superior colliculus, IC = inferior colliculus, NL = neuronal layer.

Gata2 expression in the e15 rat developing SC is excluded from (A) mitotic progenitor cells immunostaining for phospho-histone H3 (Ser10) and largely from (B) the β-III-tubulin (TUJ1) positive neuronal cell layers. In contrast, Gata2+ cells lie in a zone positive for the markers (C) Msi1 and (D) Sox2. Labels: v = ventricle, p = pial surface. Dotted line in (B) indicates position of ventricular surface.

(A,B) IHC reveals Gata2 co-expression in the outermost population of Pax3+ (A) and Pax7+ cells in the e15 developing SC. Note that Gata2+ cells directly beneath the neuronal layer lack expression of both Pax genes. (C) Many Gata2+ cells also express Ascl1 in the e15 developing SC. (D) Dcx is expressed in the neuronal area of the e15 developing SC and does not co-express with Gata2. A cluster of blood cells with nonspecific, dual-label fluorescence is evident (bright yellow cells in the merge panel) and are indicated by an asterisk in the green channel. Labels: v = ventricle, p = pial surface.

(A) Gata2 siRNA knocks down rat, but not human Gata2. HEK293 cells were transfected with either rat Gata2 (rGata2) or human Gata2 (hGata2) and either siCon (control), siGata2-1, or siGata2-2 as indicated. In a comparable experiment, siGata2-2 also lacked hGata2 knockdown activity (data not shown). (B) Knockdown of Gata2, in the developing SC affects cell migration. The indicated eGFP-expressing constructs were electroporated into the dorsal midbrains of e16 rat embryos and brains were harvested at e21. The normalized distances of electroporated cell migration were measured and sorted into 3 zones – Sector A: 0–30% of the distance between the ventricle and pial surface, Sector B: 31–60%, Sector C: 61–100%. Values indicate mean % of electroporated cells in each sector ± SEM (n=3 brains). In Sector A, the proportions of siGata2-1 and siGata2-2 electroporated cells present were significantly greater from those expressing siCon control vector (t-test, p=0.003 and p=0.03, respectively). Similarly, the proportions of siGata2-1 and siGata2-2 electroporated cells in Sector C were markedly reduced compared to siCon control (t-test, p=0.014 and p=0.037, respectively). There was no statistical significance between the proportions of cells found in any condition in Sector B. (C–F) Representative images of eGFP expression in brains electroporated with (C) pCMS-EGFP, (D) siCon pCMS-EGFP control, or Gata2 knockdown constructs (E) siGata2-1 or (F) siGata2-2. Yellow lines in (C) indicate positions of sectors A,B and C. Ventricle is at lower left corner of images. Labels: v = ventricle, p = pial surface.

(A) Co-electroporation of hGata2 with siGata2-1 allows cells to overcome knockdown of endogenous Gata2 and migrate to the SC outer laminae. Verticle black line to the left of the panel indicates the region where cell migration is rescued and where data were collected for panels C and. White scale bar = 75 μm. Electroporation e16, harvest e21 here and in subsequent panels. (B) Immunofluorescent staining for Gata2 and GFP expression detects Gata2 in cells electroporated with hGata2 and siGata2-1. Area shown is within sector A where endogenous Gata2 in not expressed at e21. Dotted lines show position of ventricular surface. (C,D) Electroporation with hGata2 rescues migration defect in Sectors B/C caused by knockdown of endogenous Gata2. Panel (C) shows GFP+ cells in Sector B/C. Cells electroporated with siCon or siGata2-1 and hGata2 have a neuronal morphology while a brain receiving only siGata2 contains only radial glial fibers. Panel (D) shows numbers of electroporated cell bodies per field in each condition, and values are means ± SEM (n=3; Student’s t-test values of siGata2-1/hGata2 and siGata2-2/hGata2 compared with siGata2-1 or siGata2-2, respectively were p= 0.048 and p= 0.003, respectively). There were no statistical differences between the values for siCon, siGata2-1/hGata2 and siGata2-2/hGata2. (E) hGata2-1-rescued cells present in Sectors B/C include NeuN+ neurons. Arrows indicate GFP+/NeuN+ cells. Labels: v = ventricle, p = pial surface.

Dorsal midbrain cells electroporated with siCon or siGata2-1 at e16 and harvested at e21 lack expression of (A) β-III-tubulin and (B) NeuN in the area between the VZ and deep layer SC neurons in Sector A. The ventricle is oriented to the top left in all images. The scale bar denotes 40 μm. (C) Electroporated cells present in midbrain Sector A outside the VZ are Ki67-, irrespective of receiving siCon or siGata2-1. (D) Expression of radial glial marker BLBP in dorsal midbrain is unaffected by shRNA-mediated knockdown of Gata2, since cells in the VZ (arrows), but not outside of the VZ, express high levels of BLBP after electroporation with either siCon or siGata2. (E) Knockdown of Gata2 in dorsal midbrain causes accumulation of Ascl1+ cells at the interface of the VZ and the deepest neuron layer of the SC. Arrows show locations of eGFP+/Ascl1+ cells. (F) Knockdown of Gata2 increases the proportion of Pax3+ cells in dorsal midbrain Sector A outside of the VZ. Arrows show locations of eGFP+/Pax3+ cells. (G) Developing SC cells electroporated with siGata2-2 retain expression of Ascl1. Graph shows proportions of eGFP+ cells in Sector A expressing Ascl1 after electroporation with the indicated constructs. A total of approximately 100–300 electroporated cells were evaluated/brain (n=3 brains). Values are means ± SEM (n=3–5 brains; ANOVA p=0.005; Student’s t-test of siGata2-1 and siGata2-2 compared to siCon control are p=0.03 and p=0.004, respectively). (H). Cells subjected to knockdown of Gata2 in dorsal midbrain retain expression of Pax3. Experimental details and analysis as in (G) except that staining was for Pax3 (ANOVA p=0.002; Student’s t-test of siGata2-1 and siGata2-2 compared to siCon control are p=0.005 and p=0.006, respectively). In this figure, where the label siGata2 is used, the cells shown are expressing siGata2-1 but the results shown are also representative of results with siGata2-2.

Animals electroporated with siGata2-1 at e16 were harvested at p5. Similar to observations in e21, the majority of labeled cells at this stage were situated in Sector A outside the VZ and still exhibited unipolar or bipolar morphology. Immunofluorescence labeling also shows that the majority of knockdown cells still express (A) Pax3 and (B) Sox2. The ventricular surface is about 100 μm out of view in the direction of the lower right of the panel, and the pial surface is several hundred μm past the top left of the panel.

(A). During embryonic dorsal midbrain neurogenesis from e15–e18, cells pass from a radial glial stem cell identity through an increasingly refined set of neuronal precursors before becoming mature neurons. Each stage shows a characteristic set of markers as indicated. (B) Disruption of Gata2 induction impedes the neuronal maturation and arrests cells in a state characteristic of an immatu