QUICK SEARCH:   [advanced] Author: Keyword(s):
Home     Help     Feedback     Subscriptions     Archive     Search     Table of Contents

First published online June 28, 2004
doi: 10.1242/10.1242/dev.01199

This Article Summary Full Text Full Text (PDF) Supplemental Data Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Campos, L. S. Articles by ffrench-Constant, C. Search for Related Content PubMed PubMed Citation Articles by Campos, L. S. Articles by ffrench-Constant, C.

ß1 integrins activate a MAPK signalling pathway in neural stem cells that contributes to their maintenance

¹ Departments of Medical Genetics and Pathology, University of Cambridge, Tennis Court Road, Cambridge CB2 1QP, UK
² Institute of Cell Biology, Department of Biology, Swiss Federal Institute of Technology, ETH Hönggerberg, CH-8093 Zürich, Switzerland
³ Department of Molecular Medicine, Max-Planck-Institut für Biochemie, Am Klopferspitz 18a, D-82152 Martinsried, Germany

View larger version (62K): [in a new window]   Fig. 1. ß1 expression in the mouse and rat germinal neuroepithelium. (A-C) Expression of ß1 (A, red) and 6 (B, red) in the the E12.5 mouse VZ. Proliferative cells are revealed in C by labelling with an antibody against phosphorylated histone H3 (P-H3). Note that these cells are seen immediately adjacent to the ventricle, in the region where ß1 expression is highest. Nuclei in A and B (green) are counterstained with DAPI. (D-G) Proliferation of the ß1-expressing cells is shown directly in a section double labelled for P-H3 (green) and ß1 (red). Nuclei are shown counterstained with DAPI in D (blue), and a merged image is shown in F. (H-K) E12.5 mouse VZ cells adjacent to the ventricle and expressing high levels of ß1 (I) also express nestin (H, merged in J with DAPI shown in K). Note the extensive double labelling, whereas the developing blood vessel (arrowed) labels much more weakly with the anti-nestin antibody. (L-N) Some of the ß1+ cells observed two days after birth (P2) in the mouse VZ/SVZ (L) double label with nestin (M, merged image shown in N, arrow). (O) At P2 ß1+ cells are also prominent in the lateral wall of the ventricle in the rat. Note the elongated morphology and the lack of labelling in the medial wall. All sections are coronal. Scale bar: 15 µm in D-G,L-N; 20 µm in H-K,O; 25 µm in A,B; 30 µm in C.

View larger version (135K): [in a new window]   Fig. 2. Expression of fibronectin, laminin 1 (L1) and laminin 2 chain (L2) in the rat VZ/SVZ. In the rat at E15.5, fibronectin is expressed in a speckled pattern throughout the developing cortex (A), L1 is found around the blood vessels and on the pial surface (B) and laminin 2 shows a condensation in the VZ (C, arrowheads). In the P2 rat brain, fibronectin can be found (D) in the VZ/SVZ, whereas L1 is present only around blood vessels (E) and L2 levels are extremely low (F and inset). L1, L2 and fibronectin are shown in red, and DAPI-counterstained nuclei are shown in green. All sections are coronal. Scale bar: 20 µm.

View larger version (105K): [in a new window]   Fig. 3. The three-dimensional structure of the neurospheres compared with the developing neuroepithelium. (A) Expression of ß1 on 14 µm cryostat sections of neurospheres grown in EGF from P0 rat pups. Note that high-ß1-expressing cells are present at the edge of the spheres. (B) Co-labelling of high-ß1-expressing cells (green) with anti-nestin antibodies (red) is seen at the edge of the neurosphere. (C) The cells in the centre of the neurosphere, shown by DAPI staining (blue), do not stain for nestin (red). Expression of GFAP (D, red), nestin/GFAP (E, green and red, respectively) and ß3 tubulin (F, red) in neurospheres grown from P0 rat pups shows that differentiated cells are present mostly in the center. Nuclei are counterstained with DAPI (blue). Proliferation in the developing VZ and in the neurospheres is shown in panels G-K. In the E15.5 developing rat brain, cells divide in the VZ, as shown by the expression of phosphorylated histone H3 (P-H3, red; H,I), whereas ß3 tubulin+ cells (green; H,G) do not express P-H3. (G,I) Higher magnification views of the ß3 tubulin+ region (G) and the phosphorylated histone H3+ region (I) seen in H, as indicated by the boxes. Likewise in the neurospheres BrdU incorporation (green nuclei, J) occurs at the periphery only, and not in the postmitotic ß3 tubulin+ cells in the centre (red cells, K). Note that the yellow (non nuclear) staining seen in the centre of the sphere is due to background. (L) The cells at the edge of the sphere express the EGF receptor. In K and L nuclei are counterstained with DAPI (blue). Scale bar: 15 µm in B; 20 µm in A; 25 µm in C-G,L,I; 30 µm in J,K; 40 µm in H.

View larger version (80K): [in a new window]   Fig. 4. Expression of ECM molecules on neurospheres. (A) Laminin 2 chain (red) is highly expressed at the edge of neurospheres. Nuclei are counterstained with DAPI (green). (B) A model of the neurosphere three-dimensional structure, summarizing the data presented and showing how nestin, EGF receptor (EGFR), ß1 integrin (ß) and laminin 2 (L2) are found in the same region. Note that, for clarity, the GFAP distributed throughout the centre of the sphere is only shown on the left side of the model, whereas the nestin and BrdU labelling around the perimeter of the entire sphere is only shown on the right side. (C,D) In the sectioned neurospheres laminin 1 is found in the center (C), whereas laminin 2 predominates at the edge (A,D). Scale bar: 20 µm.

View larger version (33K): [in a new window]   Fig. 5. Expression of EGF receptor in neurospheres. Sections of mouse neurospheres grown from P0 pups in EGF (A) or FGF2 (B). There is a greater expression of EGF receptor in the neurospheres grown in FGF2 (compare A and B). This is confirmed by western blotting using an anti-EGF receptor antibody, as shown in C. Note also the increased expression of ß1 in spheres grown in EGF (lane E) when compared with spheres grown in FGF2 (lane F). Scale bar: 40 µm for A,B.

View larger version (48K): [in a new window]   Fig. 6. Cells expressing high levels of ß1 generate more neurospheres. Dissociated (A) and intact (B) neurospheres stained for ß1 (red in A, arrow in B). Note the presence of a cell population expressing high levels of ß1. Nuclei in A were counterstained with DAPI (blue). (C) Neurospheres grown in EGF were dissociated, labelled with a monoclonal anti-ß1 antibody conjugated with FITC and sorted into two groups – high expressers (box 1) and moderate expressers (box 2) – as discussed in the text. (D) Neurospheres grown in FGF2 were dissociated, labelled with a monoclonal anti-ß1 antibody conjugated with FITC and sorted into two groups – high expressers (box 3) and moderate expressers (box 4) as above. (E) Graph showing the number of secondary neurospheres formed from the sorted cell populations after one week in culture. The differences between both 1 and 2 (high and moderate ß1 expressers in EGF), and between 3 and 4 (high and moderate ß1 expressers in FGF2), are statistically significant (P<0.001). Scale bar: 20 µm for A,B.

View larger version (36K): [in a new window]   Fig. 7. Effects of blocking MAPK on neural stem cell maintenance. (A) Phosphorylated MAPK expression and DAPI labelling of a sectioned neurosphere grown from P0 rat brain. Note the peripheral location of the P-MAPK labelling. Scale bar: 40 µm. (B-G) A decrease in neurosphere formation is seen in serial dilution assays in the presence of blockers for MAPK (PD98059). B shows the entire dilution series in a single experiment for cells grown in EGF plotted on a log-log scale, whereas E shows the same dilution series with linear scales and the regression lines from the data. C and F show the same data for cells grown in FGF. Note that, as expected, the EGF receptor inhibitor AG1478 also inhibits neurosphere formation at all dilutions in EGF. Blockers of MAPK38 (SB203580) and PI3 kinase (wortmannin and LY294002) do not show any effect on neurosphere formation. PD, PD98059; AG, AG1478; WM, wortmanin; Ly, Ly294002; SB, SB203580. D and G show the effect of the MEK inhibitor U0126 on cells grown in EGF, as compared with the control U0124 or untreated cells. Note that inhibition of sphere formation is also seen with this inhibitor. The efficacy of the inhibitors is shown in the western blots (H). Lanes 1-6 and 8-13 show cells grown in EGF and FGF, respectively, and exposed to LY294002 (lanes 1, 8), wortmannin (lanes 2, 9), U0124 (lanes 3, 10), U0126 (lanes 4, 11) or PD98059 (lanes 5, 12), or grown without inhibitors (lanes 6, 13). Lanes 7 and 14 show cells starved of growth factors prior to analysis. Note that growth factor deprivation reduces phosphorylation of both MAPK and Akt (compare lanes 6, 7 and 13, 14). PD98059 and U0126 reduce only MAPK phosphorylation, whereas wortmannin and LY294002 reduce only Akt phosphorylation.

View larger version (44K): [in a new window]   Fig. 8. MAPK is regulated by ß1 integrin. (A) Western blots show that activation of MAPK by both EGF and FGF2 is reduced by a monoclonal anti-ß1 blocking antibody (Ha2/5). Addition of the antibody is indicated by `+', absence by `–'. The lower panel shows expression of total MAPK (ERK1 and ERK2), confirming equivalent levels in the different lysates. (B-E) ß1 Staining of 14 µm cryostat sections of floxed ß1/wt (B,C) and floxed ß1/null (D,E) neurospheres exposed to cre recombinase. Note that cells at the edge of the sphere still express high levels of ß1 in the floxed ß1/wt sphere (C, arrow) in contrast to the floxed ß1/null spheres (D,E). Note also that the negative controls (omitting the anti-ß1 antibody) show no staining (data not shown). The weaker labelling in the center of the spheres therefore most likely represents low levels of ß1 expression on more differentiated and non-mitotic cells, which is also reduced in the floxed ß1/null neurospheres exposed to cre recombinase. The decrease in ß1 expression was confirmed in the neurosphere cells by flow cytometry (F,G), which reveals a shift to the left (decrease) of the ß1 levels in the floxed ß1/null spheres (G) when compared with the floxed ß1/wt (F), following cre exposure. (H) Western blot of lysates from floxed ß1/wt (left) and floxed ß1/null (right) neurospheres showing the decrease of ß1 in the null cells (top lanes). The middle panels show levels of actin, confirming equal loading, whereas the lower panels show expression of ß-galactosidase, thus confirming excision of the floxed allele as discussed in the text. (I) In cre-exposed neurospheres, MAPK phosphorylation is considerably reduced in the floxed ß1/null (right) neurospheres, whereas total MAPK is maintained. After several passages, however, MAPK phosphorylation is the same in the floxed ß1/wt (left) and floxed ß1/null (right) neurospheres (J). WT, floxed ß1/wt; Null, floxed ß1/null. Scale bar: 20 µm for B-E.

View larger version (70K): [in a new window]   Fig. 9. (A) Western blot of lysates from floxed ß1/wt (right lane) and floxed ß1/null (left two lanes) neurospheres after more than ten passages, showing that MAPK is phosphorylated equally in both groups. (B) Western blot of lysates from floxed ß1/wt (right) and floxed ß1/null (left) neurospheres (the same cultures as in A) exposed to AG1478 (20 µM), an inhibitor of the EGF receptor. This panel shows that MAPK phosphorylation is reduced to a greater extent in the null spheres (floxed ß1/null) than in the floxed ß1/wt in the presence of the inhibitor. AG, AG1478. WT represents floxed ß1/wt and Null represents floxed ß1/null spheres following exposure to cre recombinase.