Research ArticleNeuroscience

Niche-derived laminin-511 promotes midbrain dopaminergic neuron survival and differentiation through YAP

See allHide authors and affiliations

Sci. Signal.  22 Aug 2017:
Vol. 10, Issue 493, eaal4165
DOI: 10.1126/scisignal.aal4165

YAP supports dopaminergic neurons

Parkinson’s disease (PD) is a neurodegenerative disorder marked by progressive loss of dopaminergic neurons and motor control. Various factors promote or inhibit neuronal survival. Zhang et al. found that a prosurvival signal was mediated by the transcription cofactor YAP. YAP was activated in midbrain dopaminergic neurons in culture and in mice through an interaction between an integrin and the extracellular matrix protein laminin-511. YAP then transcriptionally activated dopaminergic neuron differentiation factors and a microRNA that decreased the synthesis of the apoptotic protein PTEN. The findings uncover a new role for YAP in neurons and a pathway that might be explored for the purpose of promoting dopaminergic neuron survival in PD patients.

Abstract

Parkinson’s disease (PD) is a neurodegenerative disorder in which the loss of dopaminergic neurons in the midbrain (mDA neurons) causes progressive loss of motor control and function. Using embryonic and mDA neurons, midbrain tissue from mice, and differentiated human neural stem cells, we investigated the mechanisms controlling the survival of mDA neurons. We found that the extracellular matrix protein laminin-511 (LM511) promoted the survival and differentiation of mDA neurons. LM511 bound to integrin α3β1 and activated the transcriptional cofactor YAP. LM511-YAP signaling enhanced cell survival by inducing the expression of the microRNA miR-130a, which suppressed the synthesis of the cell death–associated protein PTEN. In addition, LM511-YAP signaling increased the expression of transcription factors critical for mDA identity, such as LMX1A and PITX3, and prevented the loss of mDA neurons in response to oxidative stress, a finding that warrants further investigation to assess therapeutic potential for PD patients. We propose that by enhancing LM511-YAP signaling, it may be possible to prevent mDA neuron degeneration in PD or enhance the survival of mDA neurons in cell replacement therapies.

INTRODUCTION

Parkinson’s disease (PD) is a neurodegenerative disorder characterized by the loss of midbrain dopaminergic (mDA) neurons in the substantia nigra, which results in the main motor symptoms of disease (1). Current treatment strategies for PD patients are targeted to symptoms, and thus, there is a need to identify mechanisms that could prevent mDA neuron loss.

Much of the work aiming at identifying mechanisms supporting mDA neuron differentiation and survival has focused on transcription factors and secreted factors such as morphogens and neurotrophic factors (2). However, very little is known about the function of specific extracellular matrix (ECM) molecules on mDA neurons. The laminin family of ECM proteins has gained interest in recent years for their capacity to control not only adhesion, but also several other functions such as stem cell maintenance, survival, differentiation, and migration (3, 4). Laminins are large trimeric proteins formed by three chains (α1−5, β1−3, and γ1−3) that form at least 16 different laminin isoforms in mammals. Laminin-111 (LM111) and laminin-511 (LM511) are among the earliest ECM proteins found during embryogenesis (5, 6). Deletion of Lama1 (encoding laminin α1), Lamb1 (β1), or Lamc1 (γ1) are embryonically lethal (7), whereas deletion of Lama5 cause neural tube closure defects (8). In vitro, LM111 promotes mouse embryonic stem (ES) cell differentiation (9, 10), and LM511 facilitates self-renewal in both mouse (10) and human ES cells (11). LM111 has been long used in cell culture to promote the adhesion of neurons, including mDA neurons (12), and more recently, LM111 and LM511 were shown to support the expansion of early neural cells in human ES cultures (13, 14). However, the function of laminins in mDA neuron development and maintenance, as well as the signaling pathways they activate in mDA neurons remain to be determined.

Our results show that LM511 in the extracellular space is a potent survival and differentiation signal for mDA neurons. We found that LM511 signals via integrin α3β1 to activate the Yes-associated protein 1 (YAP), a transcriptional regulator and a central component of the hippo pathway (15). LM511 enhanced the survival of mDA neurons via a novel pathway involving YAP, miR-130a, and the phosphatase and tensin homolog (PTEN) protein, a phosphatase that induces mDA neuron death (16). Thus, we suggest the LM511-YAP–miR-130a–PTEN pathway as a target for neuroprotective and cell replacement therapies for PD.

RESULTS

Laminin α5 is present in the ECM surrounding mDA neurons

Deep-sequencing analysis of laminin subchain expression in the developing ventral midbrain revealed that several chains, including Lama1, Lama4, Lama5, Lamb1, and Lamb2, as well as Lamc1, are abundant at embryonic day 12 (E12.5) in mice (Fig. 1A). Only Lama5 had significantly increased expression in the midbrain compared to the neighboring hindbrain regions (Fig. 1, A and B, and fig. S1A). Immunohistochemistry for tyrosine hydroxylase (TH), a key enzyme catalyzing formation of the neurotransmitter dopamine, revealed that mDA neurons are present in a LAMA5+ territory at E10.5 and E12.5 (Fig. 1C), the period at which mDA neurons are born. LAMA5 was less abundant in the ventricular zone and gradually increased in the intermediate zone to reach the greatest abundance in the marginal zone, increasing as the cells mature into mDA neurons characterized by the presence of TH (Fig. 1C). Notably, LAMA5 was also present in noradrenergic neurons of the locus coeruleus (fig. S1B), a cell type that degenerates in PD patients along with mDA neurons of the substantia nigra (17).

Fig. 1 LM511 promotes the survival of mDA neurons.

(A and B) Analysis of the expression of genes encoding laminin subunits in the ventral midbrain (vMidbrain) (A) or the ventral hindbrain (vHindbrain) (B) at E12.5, as assessed by TruSeq RNA sequencing (RNA-seq). RPKM, reads per kilobase of transcript per million mapped reads. Dotted line represents an arbitrary threshold (1.5-fold). (C) Immunohistochemistry of coronal sections through the ventral midbrain at E10.5 and E12.5 for LAMA5 and the mDA neuron marker TH. Ventricular zone, VZ; intermediate zone, IZ; and marginal zone, MZ. Scale bars, 50 um. (D) Immunocytochemistry of E11.5 midbrain cultures assessing the effect of recombinant laminins on differentiation or survival. LMX1A marks the mDA lineage and TH marks mature mDA neurons. Scale bars, 50 μm. (E and F) Quantification of the percentage of LMX1A+ (E) and TH+ (F) cells in the cultures described in (D). (G) Representative images of immunocytochemical staining for activated caspase-3 (aCASP3; green) and TH (red) in E11.5 primary mouse midbrain neuron cultures treated with the neurotoxin, 6-hydroxydopamine (6-OHDA) and either LM111 or LM511. Scale bars, 10 μm. (H) Percentage of TH+ neurons described in (G) that were also aCASP3+. (I) Representative Western blotting analysis of PTEN in SN4741 cells treated for up to 48 hours with LM511. Data are means ± SEM, n = 3 independent experiments; *P < 0.05, **P < 0.01, and ***P < 0.001, analysis of variance (ANOVA) analysis with Benjamin-Hochberg posttest (B, E, and F) or two-sided t test (H).

LM511 promotes the survival of mDA neurons

The function of laminin proteins whose mRNA chain combinations were enriched in ventral midbrain (LM111, LM411, LM421, LM511, and LM521) was investigated in primary mouse mDA neurons cultivated for 48 hours. We first examined the number of cells positive for LIM homeobox transcription factor 1 α (LMX1A), a transcription factor present in all the cells of the mDA lineage (18). LM511 supported the highest proportion of LMX1A+ cells in the cultures (22%), whereas LM332 supported the lowest proportion (2.5%) (Fig. 1, D and E). LM322 was used as negative control because Lama3 and Lamc2 are expressed at low amounts in the brain (Fig. 1A and fig. S1A). LM521, a laminin similar in composition to LM511, did not support cells in the DA lineage as LM511 did, underlining the specificity of laminin chain composition in trimeric laminins. We next examined the effects of laminins on mDA neurons and found that LM511 increased their numbers by eightfold compared to LM111, the standard laminin used in mDA cultures, and 70-fold compared to control LM322 (Fig. 1, D and F). In addition, LM511 reduced the intensity of staining for the cell death marker, activated caspase-3, by 30% compared to LM111 (fig. S1C). Moreover, LM511 reduced the loss of mDA neurons induced by the DA neurotoxin 6-OHDA by 3.5-fold in the presence of LM111. Whereas 69% of the TH+ neurons cultured with LM111 were activated caspase-3+, only 21% cultured with LM511 were double-positive (Fig. 1, G and H), indicating that LM511 promotes the survival of mDA neurons. We also examined whether laminins regulate the PTEN protein, a phosphatase that antagonizes the survival effect of phosphatidylinositol 3-kinase (PI3K) activity and its deletion, promotes mDA neuron survival (1921). We found that PTEN was present in the ventral midbrain and decreased as mDA neurons mature (fig. S1D). Accordingly, treatment of the substantia nigra mDA cell line SN4741 (22) with LM511 was sufficient to decrease PTEN abundance, in contrast to LM111 (Fig. 1I). Thus, our findings indicate that LM511 promotes survival of mDA neurons by a mechanism involving repression of PTEN.

LM511 increases YAP activity through integrin α3

To investigate how the message of LM511 is conveyed to PTEN, we focused our attention on YAP, a transcriptional coactivator that interacts with different transcription factors (23); regulates key functions, such as proliferation, differentiation, and survival (24); and can transduce mechanical information from the extracellular space, such as stiffness or cell density (25). We first examined whether laminins can activate this pathway and regulate the expression of YAP target genes. Analysis of mouse primary midbrain cultures or the SN4741 cell line revealed that LM111 did not increase the expression of Ctgf and Serpine1 compared to poly-l-lysine (fig. S2, A and B). Moreover, no other laminin tested except for LM511 increased the expression of YAP target genes (Fig. 2, A and B). Similar results were obtained when human SAI2 neuroepithelial stem (hNES) cells (26) differentiated into mDA cells (27) were examined (Fig. 2C and fig. S2C). These results suggested that LM511 selectively activates this pathway in both mouse and human mDA cells. To confirm the activation of YAP by LM511, we examined the translocation of YAP to the nucleus of TH+ mDA neurons in primary midbrain cultures treated with different laminins. YAP accumulated in the nuclei of LM511-treated TH+ cells by 80%, but only half of that was found in TH+ cells treated with other laminins (Fig. 2, D and E). Thus, our results indicate that LM511 selectively activates YAP in mDA neurons.

Fig. 2 LM511 activates YAP via ITGA3B1 in mDA neurons.

(A to C) qRT-PCR analysis of the effect of laminins on the expression of YAP-target genes in E11.5 murine mDA neuron cultures (A), murine substantia nigra DA cells (SN4741; B) and human mDA neurons differentiated from hNES-SAI2 cells (C). (D) Representative images of immunocytochemistry assessing the subcellular localization of YAP in primary murine TH+ mDA neurons treated with various laminins, as assessed by immunocytochemistry. Arrowheads indicate cytoplasmic staining of YAP. Scale bars, 10 μm. (E) Quantitative analysis of the ratio of nuclear YAP to total YAP in primary murine TH+ mDA neurons in (D). (F) Effect of blocking antibodies to integrin subunits on the LM511-induced expression of YAP target genes in hNES-SAI2 differentiated into mDA neurons, as assessed by qRT-PCR. (G) Western blot analysis of phosphorylated YAP (Ser127) and PTEN in SN4741 cells treated with LM511 and transfected with Itga3 shRNA (left) or Itga6 shRNA (right) (H) Representative images of immunocytochemistry assessing the subcellular localization of YAP in SN4741 cells transfected with Itga3 shRNA and treated with LM511. Scale bars, 10 μm. (I) Expression of three YAP-target genes in experiments described in (H), as assessed by qRT-PCR. Data are means ± SEM, n = 3 independent experiments; **P < 0.01 and ***P < 0.001, ANOVA analysis with Benjamin-Hochberg posttest (A to C and F) or two-sided t test (E and I).

Because various integrins transduce signaling by laminins (28), we examined which integrin mediates the activation of YAP by LM511 in mDA neurons. Treatment of hNES-derived mDA cells with LM511 and integrin α chain blocking antibodies revealed a significant reduction of CTGF and ANKRD1 by the integrin subunit α3 (ITGA3) antibody (Fig. 2F), suggesting that ITGA3 mediates the activation of YAP by LM511. To confirm this result, we generated SN4741 cell lines with stable knockdown of ITGA3 and also ITGA6, because ITGA6 was previously reported to be an LM511 receptor in other systems (28). ITGA3 or ITGA6 knockdown reduced ITGA3 or ITGA6 protein, respectively, but only ITGA3 knockdown inhibited YAP activation (inferred from increased phosphorylation) and increased the abundance of the cell death protein, PTEN (Fig. 2G). We also observed that ITGA3 knockdown markedly reduced the nuclear translocation of YAP compared with control cells (Fig. 2H). To further test whether blocking ITGA3 could inhibit the activation of YAP by LM511, we seeded these cells on LM511-coated plates and found that ITGA3 knockdown strongly inhibited the expression of the YAP-target genes, Ctgf, Serpine1, and Cyr61 (Fig. 2I). Combined, these observations indicate that ITGA3, but not ITGA6, mediates the activation of YAP by LM511 in mDA neurons. Accordingly, ITGA3 was present in mDA neurons, increasing progressively from the ventricular zone to the marginal zone (fig. S2D), as described for LAMA5 (Fig. 1C). Moreover, analysis of the developing ventral midbrain transcriptome revealed that the two most abundant integrin transcripts were Itgb1 and Itga3 (fig. S2E). Because there are 24 heterodimer αβ integrins and there is only one α3-containing integrin, integrin α3β1 (ITGA3B1) (28), we conclude that the ITGA3B1 mediates LM511-YAP signaling in mDA neurons.

YAP decreases PTEN and promotes the survival of mDA neurons

We reasoned that if YAP is important for the survival of mDA neurons, it should be present in these cells from early development. Immunohistochemical analysis revealed that YAP is present in TH+ mDA neurons at E11.5, including the most lateral TH+ cells moving toward the future substantia nigra (Fig. 3A). At E13.5, YAP remained in TH+ mDA neurons, whereas TH cells present in the tissue were also YAP, arguing for the specificity of YAP presence in mDA neurons (Fig. 3A).

Fig. 3 LM511 regulates miR-130a to repress PTEN via YAP in DA neurons.

(A) Analysis of the ventral midbrain at E11.5 and E13.5 by immunohistochemistry to identify TH+ and YAP+ cells. Arrows indicate YAP cells with very little or no detectable TH. Scale bars, 20 μm. (B and C) Representative Western blots showing PTEN abundance in SN4741 cells after either YAP knockdown with shRNA (B) or YAP overexpression (C). (D) Assessment of the viability of SN4741 cells after YAP knockdown or overexpression. (E) Immunostaining for aCASP3 to examine cell death in primary TH+ neurons after YAP knockdown or overexpression and 6-OHDA treatment. Scale bars, 10 μm. (F) Representative images of immunohistochemistry analysis of active caspase-3 and TH+ in the ventral midbrain of E14.5 mouse embryos electroporated in utero with control or Yap-shRNA at E11.5. Scale bars, 50 μm. (G) Quantification of active caspase-3 in the cells electroporated (GFP+) in (F). (H to K) qRT-PCR analysis of miRNAs predicted to target PTEN in: different embryonic brain regions (H); laminin-treated mouse primary midbrain cultures (I); SN4741 cells after YAP knockdown or overexpression (J); and SN4741 cells treated with LM511, control, or Yap-shRNA (K). (L) Representative Western blot of SN4741 cells stably overexpressing miR-130a. (M) PTEN immunohistochemistry images of the ventral midbrain of E13.5 embryos electroporated in utero with miR-130a–GFP at E11.5. Scale bar, 50 μm. (N) Analysis of the viability of SN4741 overexpressing YAP after treatment with miR-130a inhibitor. Data are means ± SEM, n = 3 independent experiments; *P < 0.05, **P < 0.01, and ***P < 0.001, ANOVA analysis with Benjamin-Hochberg posttest (D, H to J, and N) or two-sided t test (G and K).

To determine the biological function of YAP in DA neurons, we generated substantia nigra SN4741 cells that had either stable knockdown or overexpression of YAP. Whereas YAP knockdown increased PTEN abundance, YAP overexpression reduced it (Fig. 3, B and C). Next, we examined whether the survival of these cells was affected in neuronal media (N2) but without serum or survival factors. Whereas control SN4741 cells survived for 96 hours, YAP knockdown accelerated cell death, reducing it to 72 hours, and YAP overexpression prolonged survival up to 120 hours (Fig. 3D). Notably, surviving cells in the YAP overexpression condition exhibited more neuron-like morphologies (fig. S3A). Moreover, exposure of YAP-depleted SN4741 cells to oxidative stress (H2O2, 50 μm for 24 hours) induced substantial cell death, as assessed by the presence of activated caspase-3 compared to control cultures (fig. S3, B and C). To further corroborate the survival-promoting effects of YAP and its capacity to promote neuroprotection, as shown for LM511 (Fig. 1), we treated mouse primary midbrain cultures with the DA neurotoxin, 6-OHDA (20 μm), and performed YAP knockdown or overexpression using lentiviruses. We found that YAP knockdown increased the loss of mDA neurons, as inferred from a decrease in TH+ cells, an increase in active caspase-3 staining and the presence of condensed and fragmented nuclei, a hallmark of apoptosis (Fig. 3E). YAP overexpression reduced apoptosis (nuclear condensation and active caspase-3) and maintained the number of TH+ neurons with healthy neuronal morphology (Fig. 3E). Combined, these results indicated that YAP is at least partially required and sufficient for the survival of mDA neurons in culture.

To determine the significance of these findings in vivo, we intracerebroventricularly injected E11.5 embryos and electroporated with control or Yap-targeted short hairpin RNA (shRNA), along with enhanced green fluorescent protein (EGFP) to identify the electroporated area. Analysis of the ventral midbrain at E14.5 revealed very low amounts of activated caspase-3 in control shRNA-treated EGFP+ cells. YAP knockdown decreased the abundance of YAP in EGFP+ cells (fig. S3D) and increased the proportion of EGFP+ cells with activated caspase-3 (Fig. 3, F and G). These included caspase-3+ cell fragments and TH+ cells in vivo (Fig. 3F). Thus, our results indicate that YAP is also at least partially required for the survival of mDA neurons in vivo.

mir-130a mediates the down-regulation of PTEN in cultured cells and in vivo

We next sought to identify the mechanism by which YAP reduces PTEN (Fig. 3C). Because YAP is a transcriptional coactivator and can stimulate microRNA (miRNA) biogenesis (29), we examined whether the inhibition of PTEN is mediated by miRNAs (30). miRNAs potentially targeting the 3′ untranslated region of PTEN mRNA were predicted with five different softwares. Four miRNAs with high prediction score were selected for further study (fig. S3E). Analysis of their expression in three regions of the embryonic brain showed that miR-130a was the most enriched in midbrain, followed by miR-107, whereas miR-148a and miR-29a were not enriched compared to forebrain and hindbrain (Fig. 3H). We then examined whether laminins regulate any of these miRNAs in mouse primary cultures. Notably, LM511 was the only laminin capable of regulating the expression of any of the miRNAs targeting PTEN (Fig. 3I). In addition, miR-130a was the only miRNA increased by LM511 in primary cultures (2.7-fold; Fig. 3I) and regulated by both YAP overexpression and knockdown (2.1-fold increase and 53% decrease, respectively) in SN4741 cells (Fig. 3J). Having established that both YAP and LM511 increase the expression of miR-130a, we then investigated whether YAP mediates the effect of LM511 on miR-130a. Notably, we found that Yap shRNA blocked the induction of miR-130a by LM511 in SN4741 cells (Fig. 3K), indicating that LM511 increases miR-130a through YAP. Finally, we examined whether miR-130a regulates PTEN in mDA neurons. We first performed in situ hybridization and found that miR-130a was expressed in the mDA domain (fig. S3F), following a similar distribution as LM511 (Fig. 1C). We then generated a stable SN4741 cell line overexpressing miR-130a and found that miR-130a overexpression reduced PTEN abundance (Fig. 3L). Similarly, expression of a miR-130a–GFP construct in the developing ventral midbrain in vivo reduced PTEN abundance (Fig. 3M), and no activated caspase-3 was detected (fig. S3E). Moreover, the effect of YAP overexpression on the survival of substantia nigra cells was blocked by a miR-130a inhibitor (Fig. 3N). Finally, to determine whether PTEN decreases the survival of mDA neurons by inhibiting PI3K, we then used a PTEN inhibitor, SF1670, and a PI3K inhibitor, wortmannin, and found that they respectively increased and decreased the survival of cultured substantia nigra cells (fig. S3, H and I). Combined, our findings suggest that LM511 activates a pathway in which a YAP-mediated up-regulation of miR-130a reduces the abundance of PTEN, thus increasing the survival of mDA cells.

YAP promotes mDA neuron differentiation

In addition to the effects of the LM511-YAP pathway on mDA neuron survival, we also found that LM511 induced the morphological differentiation and maturation of primary mDA neurons, which displayed long neurites and axons with large growth cones compared to LM111 (Fig. 4A). It has been previously reported that axonal outgrowth and growth cone formation are prerequisites for proper functionality of mDA neurons and a requirement for the clinical application of stem cell–derived dopamine neurons (31). Thus, we decided to investigate whether YAP can improve the DA differentiation of human NES stem cells. We first examined whether LM511 increased the DA differentiation of two different hNES lines, SAI2 and AF22 (26, 32). We found that LM511 increased the expression of genes required for mDA neuron differentiation, such as those encoding the LIM homeodomain transcription factors LMX1A (18) and LMX1B (33) or the paired-like homeodomain transcription factor PITX3 (34) as well as the expression of TH, as assessed by quantitative reverse transcription polymerase chain reaction (qRT-PCR) and the number of TH+ neurons after 8 days (Fig. 4B and fig. S4, A and B). We next investigated whether YAP overexpression also improved the differentiation of hNES cells into mDA neurons and found a 2.5-fold up-regulation of LMX1A and LMX1B, as well as a twofold increase in PITX3, and a 40% increase in TH compared to EGFP-expressing hNES cells (Fig. 4C). Western blot analysis showed that PITX3, a protein enriched in substantia nigra DA neurons (35), was abundant in hNES cells overexpressing YAP, but not in control GFP-hNES cells, after DA differentiation for 8 days (fig. S4D). We then performed immunocytochemistry to examine whether YAP lead to improved acquisition of mDA markers in TH+ neurons. We found that the LMX1A, PITX3, LMX1B, DAT, and ALDH1A1 were induced or increased to a greater extent in TH+ neurons derived from YAP-hNES cells than from GFP-hNES cells (Fig. 4, D and E, and fig. S4, E and F). Moreover, we also found that inhibition of miR-130a with an inhibitor reduced both the number of mDA neurons derived from hNES cells and the expression of both TH and LMX1A (fig. S4, G to I). Finally, to determine the functional relevance of YAP in mDA neurons in vivo, we performed in utero electroporation of the ventral midbrain at E11.5 and examined the impact of YAP overexpression on midbrain development. We found that ectopic YAP expression in the basal plate induced PITX3 in the basal plate (Fig. 4F), but not TH, indicating that additional factors in the floor plate are required to make mDA neurons. In agreement with this hypothesis, overexpression of YAP in the midbrain floor plate increased the number of TH+ mDA neurons, which were then abundant in the ventricular zone, without affecting proliferation, as assessed by 5-ethynyl-2′-deoxyuridine (EdU) incorporation (Fig. 4, G and H). Thus, our results indicate that YAP promotes mDA neuron development not only in rodent cells in vitro, but also in human stem cells and in the rodent brain in vivo.

Fig. 4 LM511 and YAP improve DA neuron differentiation.

(A) Effects of LM511, compared to LM111, on the numbers of mDA neurons and their neurite extension as assessed by immunohistochemistry for NR4A2 and TH. Scale bars, 50 μm. (B and C) qRT-PCR analysis of the effects of LM511, compared to LM111, on the expression of specific DA markers in hNES-AF22 cells (B) or hNES-SAI2 cells differentiated into DA neurons (C). (D and E) Immunocytochemistry analysis of the effect of YAP overexpression on the number of TH+ human DA neurons derived from hNES cells and the levels of PITX3 (D) and LMX1A (E) in TH+ cells. Scale bars, 10 μm. (F) Immunohistochemistry for PITX3 in the midbrain basal plate of E13.5 embryos electroporated in utero with YAP-IRES-GFP or GFP at E11.5. Boxes in the upper right are magnified below. Arrowheads point to ectopic PITX3+ cells. Scale bars, 50 μm. (G) Effects of YAP overexpression on the number of TH+ cells and EdU+ cells as assessed by immunohistochemistry in E13.5 embryos electroporated in utero with YAP-IRES-GFP or GFP in the midbrain floor plate at E11.5. Scale bars, 50 μm. (H) Quantitative analysis of TH+ and EdU+ cells in (G). (I) Schematic diagram of the proposed mechanism by which the LM511-ITGA3B1-YAP axis enhances mDA neuron survival (via suppression of PTEN by miR-130a) and differentiation (by up-regulation of mDA neuron transcription factors, such as LMX1A, LMX1B, and PITX3). Data are means ± SEM, n = 3 independent experiments; *P < 0.05, **P < 0.01, and ***P < 0.001, two-sided t test (B, C, and H).

DISCUSSION

Here, we identify the ECM protein, LM511, as a specific activator of YAP. Moreover, we identify the LM511-YAP pathway as a novel signaling cascade leading to mDA neuron survival and differentiation. LM211 combined with mechanical stimuli has been previously found to activate YAP in Schwann cells (36). However, we found that LM211 had no effect on the survival of mDA neurons, suggesting that the effects of laminins are cell type–specific, although signaling mechanisms might be shared. The activation of YAP by LM511 in mDA neurons was mediated by ITGA3. Upon nuclear translocation, YAP increased miR-130a, which inhibited PTEN, resulting in reduced loss of mDA neurons. Reduced PTEN has been found to activate the Akt pathway and lead to mDA neuron survival (37, 38). A second function of YAP in mDA neurons was to increase the expression of genes controlling mDA identity, such as Lmx1a, Lmx1b, and Pitx3, as well as mDA neuron differentiation (Fig. 4I). Our results thus indicate that YAP activation by LM511 in the ECM orchestrates a signaling program that allows mDA neurons to sense their environment and thrive in their niche. This is very interesting because the low stiffness of the brain and cell-to-cell contact inhibit YAP signaling (25, 39, 40). Thus, the activation of YAP by LM511 provides a mechanism by which YAP inhibitory signals can be counteracted in specific cellular niches, such as that occupied by mDA neurons in the ventral midbrain. In a broader context, our results indicate that YAP plays not only a key role in the transduction of the physical properties of the ECM, such as stiffness (25), but also to transduce the chemical and molecular properties of ECM proteins, such as the chain composition of laminins. We thus suggest that YAP plays a much wider role than previously anticipated, as integrator and coordinator of cellular responses to diverse niche-derived ECM signals.

Our study also shows that the effects of different laminins on mDA neurons are highly specific, underlining the remarkable signaling and functional specificity of laminin chains and proteins in the ECM. LM322, a laminin absent in the midbrain, did not promote the survival of mDA neurons. However, laminins present in the midbrain, such as LM111 and LM511, supported mDA neurons. Accordingly, the β2 chain of laminin has been reported to inhibit neuronal regeneration, whereas β1 laminin promotes neurite outgrowth (41). However, we also found that the activities of β1-containing laminins, such as LM111 and LM511, differed to a large extent. LM511, but not LM111, reduced PTEN, prevented the loss of mDA neurons by oxidative stress, promoted axonal growth, and induced the acquisition of an authentic mDA neuron phenotype in hNES cells. Thus, our results indicate that the α5 laminin chain plays a key additional role in mDA neuron survival and differentiation.

PD patients have been found to exhibit shrinkage of putamen, degradation of the ECM, and alteration of ECM enzymes (42, 43). It remains to be determined whether the loss of ECM may contribute to the disease by reducing the size of the mDA niche. Our work identifies LM511-YAP as a key pathway by which niche signals control the survival and differentiation of mDA neurons. These results warrant further investigation on the role of this previously unknown pathway in PD samples and in PD models. We suggest that strategies aiming at increasing the size of the mDA niche, by local delivery of LM511 or by activation of the LM511-YAP pathway, may be useful for neuroprotective or regenerative therapies in PD.

MATERIALS AND METHODS

Mice

Wild-type CD-1 mice (Charles River Laboratories) were mated overnight, and by noon of the day, the plug was considered E0.5. Mice were housed, bred, and treated according to local ethical committees: Stockholm Norra Djurförsöksetisks Nämnd (ethical permit number N273/11, N370/09, N486/12, and N40/15).

TruSeq

E12.5 mouse brains were quickly dissected out under the microscope (44). RNA was extracted from dissected tissues (ventral midbrain, ventral hindbrain, ventral forebrain, dorsal midbrain, and lateral midbrain) according to the standard TRizol RNA isolation protocol (Life Technologies). RNA (200 ng) from each sample was used to establish cDNA libraries, which were prepared according to Illumina’s TruSeq DNA sample preparation guide. Ten enriched libraries (duplicate samples from five different parts of the brain) were subjected to Illumina sequencing in one lane of HiSeq 2000. The sequencing reads were sorted by bar code and then mapped into the mouse genome (mm9) using Bowtie short read aligner. The data were analyzed using Qlucore Omics Explorer. RNA-seq data are deposited in the public National Center for Biotechnology Information Gene Expression Omnibus database with accession number GSE82099.

Cryosection of midbrain and antigen retrieval

Embryos were dissected out of the uterine horns in ice-cold phosphate-buffered saline (PBS), and their brains were carefully dissected. Embryos and brains were fixed in 4% (w/w) paraformaldehyde (PFA) for 4 hours to overnight at 4°C, cryoprotected in 15% d-sucrose (catalog no. 27480.294, VWR International) for overnight at 4°C, then changed to 30% d-sucrose for overnight at 4°C, and frozen in Tissue-Tek cryomold 15 mm by 15 mm by 5 mm (reference no. 4566, SAKURA) using Tissue-Tek Optimum Cutting Cryomount Compound (catalog no. 45830, SAKURA) in dry ice. Serial coronal 14-μm sections of the brain were obtained on a cryostat. After dehydration at 37°C for 30 min, sections were exposed to heat-induced epitope retrieval in target retrieval solution (catalog no. S1699, DAKO) in a microwave for 5 min and were then cooled down for 30 min at room temperature.

Immunohistochemistry

After antigen retrieval and blocking with blocking buffer [0.1% Triton X-100, 0.1% Tween 20, and 10% normal donkey serum (#017-000-121, Jackson ImmunoResearch)] for 1 hour, sections were probed with various primary antibodies (prepared in 10% blocking buffer) at 4°C overnight, washed, and then incubated with appropriate secondary antibodies. Sections were visualized with Zeiss LSM 700 confocal microscope and collected with an AxioCam digital microscope camera facilitated with intuitive ZEN software. Images were processed with Photoshop (Adobe) and/or ImageJ. Figures were assembled in Illustrator (Adobe). All the antibodies used are listed in table S1.

Primary midbrain culture

E11.5 mouse ventral midbrain tissue was dissected in ice-cold PBS/0.2% glucose. The dissected tissue was triturated through flame-narrowed glass Pasteur pipettes for three times, with each time, a narrower Pasteur pipette was used. Trituration was performed in N2 medium [50% minimum essential medium (MEM), catalog no. 21090-021, Gibco Invitrogen; 50% F-12 medium, catalog no. 21765-029, Gibco Invitrogen; 0.6% d-glucose, G8270, Sigma-Aldrich; 0.5% glutamine, catalog no. 25030-081, Gibco Invitrogen; 1.5% Hepes, catalog no. 15630, Gibco Invitrogen; and 1% N2 supplement, catalog no. 17502-048, Gibco Invitrogen]. The cell suspension was plated at 1 × 105 cells/cm2 and cultured in N2 medium. Primary cultures were differentiated for 3 days in vitro with appropriate drugs, vehicle, or plasmids, fixed with 4% PFA, and processed for immunostaining using appropriate antibodies as described above.

Human stem cell cultures

A human induced pluripotent stem cell–derived long-term neuroepithelial stem cell line (hNES) line (AF22) and a human hindbrain NES (SAI2) (26, 32, 45) were cultured as previously described, with some modifications. Briefly, hNES cells were maintained in the Dulbecco’s modified Eagle’s medium (DMEM)/F-12 GlutaMAX (catalog no. 31331, Gibco Invitrogen) supplemented with recombinant human (rh) basic fibroblast growth factor (10 ng/ml) (catalog no. 233-FB, R&D Systems), rhEGF (10 ng/ml) (catalog no. 236-EG, R&D Systems), B-27 supplement (1:1000; catalog no. 17504-044, Gibco Invitrogen), and N-2 supplement (1:100; catalog no. 17502-048, Gibco Invitrogen). hNES cells were passaged at a ratio of 1:3 every second to third day using TrypLE Express (catalog no. 12604, Gibco Invitrogen) and defined trypsin inhibitor (catalog no. R-007-100, Gibco Invitrogen). hNES cells were cultured on double-coated plates with 0.002% of poly-l-ornithine (catalog no. P4957, Sigma-Aldrich) and laminin (2 μg/ml) (catalog no. L2020, Sigma-Aldrich). DA differentiation of hNES cells was induced through sequential treatment with growth factors for 8 days (27). Briefly, hNES cells were seeded on poly-l-ornithine/laminin-coated dishes and treated with N2 (1:100; Gibco Invitrogen), B-27 (1:100; Gibco Invitrogen), sonic hedgehog (200 ng/ml; R&D Systems), glycogen synthase kinase 3b inhibitor CT99021 (1 μM), brain-derived neurotrophic factor (20 ng/ml; R&D Systems), and glial cell line–derived neurotrophic factor (20 ng/ml, R&G).

SN4741 cells

SN4741 cells were cultured in DMEM (catalog no. 41965-039, Gibco Invitrogen) supplemented with 10% fetal bovine serum (FBS) (catalog no. 10270, Gibco Invitrogen), 1% d-glucose (catalog no. G8270, Sigma-Aldrich), 1% penicillin-streptomycin (catalog no. 15140-122, Gibco Invitrogen), and 1% l-glutamine (catalog no. 25030-024, Gibco Invitrogen).

Treatments

Human recombinant laminins (from BioLamina AB) were used at a concentration of 1 μg/cm2 (in Dulbecco’s PBS with Ca++/Mg++ overnight at 4°C) to coat culture dishes (BD Falcon) previously coated with 0.002% poly-l-ornithine (catalog no. P4957, Sigma-Aldrich) in PBS overnight. 6-OHDA (catalog no. H4381, Sigma-Aldrich) was dissolved in 0.01% (w/v) ascorbic acid was used at 20 uM. Hydrogen peroxide (H2O2; catalog no. H1009, Sigma-Aldrich) was used at 50 μM.

Cell counts

The number of TH+, LMX1A+, active caspase-3, and 4′,6-diamidino-2-phenylindole (DAPI+) cells in primary mDA neuron cultures was manually counted in 8 to 10 fields per condition in 3 independent experiments. The ratio of TH+DAPI+ over DAPI+ and LMX1A+DAPI+ over DAPI+ were generated from these counts. SN4741 cells were counted manually in nine fields and three independent experiments. The number of TH+ and EdU+ cells was also counted in coronal sections though the embryonic midbrain in at least three separate animals.

Image analysis

Mean fluorescence intensity of activated caspase-3 staining was measured in nine fields and three independent experiments using ImageJ program and mean intensity measurement. Subcellular YAP distribution was measured in cultures labeled with fluorescent antibodies against YAP and TH and counterstained with DAPI to outline the nuclei. The ratio of YAP fluorescence in nuclei over its fluorescence in the whole cell was quantified using the ImageJ program. Mean fluorescence intensity was measured in nine fields and three independent experiments.

Western blot

Cells were lysed, and protein was extracted in radioimmunoprecipitation assay buffer [50 mM tris-HCl (pH 8.0), 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, and cocktail protease inhibitors (catalog no. 04693132001, Roche)]. Lysates were then sonicated (instrument settings: duty cycle, constant; output control, 2; and six impulses per sample; Sonifier 450). Twenty micrograms of protein-containing lysate was separated on an SDS–polyacrylamide gel electrophoresis 5 to 12% gel and transferred onto polyvinylidene difluoride membrane (catalog no. 162-0177, Bio-Rad). Then, membranes were blocked with 5% milk for 30 min and were probed with appropriate primary antibodies overnight at 4°C. Next day, after washing out unbound primary antibodies, membranes were incubated with secondary antibodies for 1 hour at room temperature and visualized with enhanced chemiluminescence reagent (RPN2236, GE Healthcare). Antibodies are listed in table S1

Integrin blocking assay

Integrin blocking assays were performed as described elsewhere (34), to determine the integrin receptors to LM511. Briefly, plates were coated with LM511. hNES cell suspension was incubated with function-blocking antibodies to integrin (concentration as recommended by supplier) for 30 min, plated on LM511-coated plates, and the next day, RNA were extracted for qRT-PCR analysis.

qRT-PCR for mRNA

Total RNA was isolated using RNeasy Mini kit (catalog no. 74106, Qiagen), and cDNA was made with cDNA SuperScript II reverse transcriptase kit (#18064, Invitrogen). Specific genes were amplified using Fast SYBR Green Master mix kit (part no. 4385614, ABI). Real-time PCR was performed using standard protocols on a 7500 fast real-time PCR system (Applied Biosystems). Glyceraldehyde-3-phosphate dehydrogenase was used to normalize the expression of mRNA. Primers used are listed in table S2.

Cloning

shRNA targeting mYAP1 was cloned into lenti–pLKO.1-puro cloning vector (Addgene #10878) digested with AgeI/EcoRI. shRNA targeting mYAP1 was cloned into pSilencer H1 vector digested with BamHI/HindIII (Ambion). shRNA targeting mITGA3 was cloned into lenti–pLKO-puro vector digested with AgeI/EcoRI. shRNA targeting mITGA6 was cloned into lenti–pLKO-puro vector digested with AgeI/EcoRI. pLKO.1-control (Addgene #10879) was used as control. Activated form of YAP (S127A) cDNA was amplified from its template (Addgene #27370) and a 1541 base pairs (bp) long insert was cloned into the lenti-pLJM1 vector (Addgene #19319) with digestion sites of AgeI/EcoRI. To clone YAP (S127A) cDNA into pCAGIG-IRES (internal ribosomal entry site)-GFP, pCAGIG-IRES-GFP was modified at the multiple cloning site (MCS) to have the same cloning sites as in pLJM1. With the new MCS site, YAP (S127A) cDNA was subcloned into pCAGIG-IRES-GFP. miR-130a was amplified from genomic DNA and was cloned into pCAGIG-IRES-GFP vector and lenti-pLJM1 vector, respectively. The genomic fragment (583 bp in length) contained the miR-130a precursor and flanking hairpin sequences. For shRNA cloning, appropriate oligonucleotides pairs were designed, annealed, and cloned into vectors. For amplification of cDNA or miRNA, PCR was performed with phusion DNA polymerase (M0530L, New England Biolabs) according to the manufacturer’s instructions. Sequences of primers used are listed in table S3.

Lentivirus production

Lipofectamine 2000 (catalog no. 52887, Invitrogen) was used as transfection reagent in a ratio 3:2 with the plasmid cocktail. Lentiviral expression plasmids, psPAX2 packaging plasmid, and pMD2.G envelope plasmid were mixed at a ratio of 4:3:1, respectively. The mixture was added dropwise to human embryonic kidney–293FT cells (catalog no. R700-07, Invitrogen) and cultured less than 10 passages in DMEM supplemented with 10% FBS, 1% MEM non-essential amino acids (catalog no. 11140-035, Gibco Invitrogen), 1% l-glutamine, 1% MEM sodium pyruvate (catalog no. 11360-039, Gibco Invitrogen), 1% penicillin-streptomycin, and geneticin (500 μg/ml) (catalog no. 10131-027, Gibco Invitrogen). The medium was harvested at 24 and 48 hours, combined, and centrifugated at 1500 rpm for 5 min at 4°C. The supernatant was transferred to a centrifuge tube (reference no. 326823, Beckman Coulter) and centrifugated at 20,000g at 4°C for 2 hours with an Avanti J-30I ultracentrifuge (Beckman Coulter). The supernatant was then carefully removed, and the pellet containing the virus, was resuspended in PBS, prechilled overnight at 4°C, and kept at −80°C until use.

Stable cell lines

Lentivirus particles harboring the corresponding genes were added to the cells. Selection and maintenance of stably transduced cells was performed in puromycin (4 μg/ml for SN474 and 1250 ng/ml for hNES cells).

qRT-PCR for miRNA

Total RNA was isolated using TRIzol reagent (#15596, Ambion) and chloroform, precipitated with isopropanol, and purified with 75% ethanol. miRNA cDNA was synthesized with NCode VILO miRNA cDNA synthesis kit (Invitrogen), and specific mature miRNA expression was determined with EXPRESS SYBR GreenER miRNA qRT-PCR kits (Invitrogen). U6 was used to normalize the expression of miRNA. Primers used are listed in table S2.

In utero electroporation

E11.5 pregnant female mice were deeply anesthetized using isofluorane (IsoFlo, Abbott Laboratories), and the uterine horns were accessed through an abdominal incision. pCAGIG-IRES-EGFP or pCAGIG-YAP-IRES-EGFP or pCAGIG-IRES-EGFP–miR-130a were injected into the mesencephalic ventricle for gene overexpression. pCAGIG-IRES-EGFP mixed with H1-shRNA-CTRL or pCAGIG-IRES-EGFP mixed with H1-shRNA-YAP1 were injected into the mesencephalic ventricle for gene silencing. Plasmids were used at 1 μg/μl in PBS containing 10% of Fast Green (Sigma-Aldrich). Square electric pulses of 30 V and 50 ms were passed through the uterus five times, spaced by 950 ms, using a square pulse electroporator (CUY21, Nepa Gene). The uterine horns were placed back in the abdominal cavity, which was then closed with sutures. EdU (50 mg/kg body weight; Invitrogen) was injected 16 hours after electroporation. Embryos were collected 48 hours after electroporation (E13.5).

Statistical analysis

Multiple comparison was performed with ANOVA analysis followed by Benjamin-Hochberg was used for statistical analysis. Data were assumed to be normally distributed. Two-group comparison test was performed by Student’s t test; *P < 0.05, **P < 0.01, and ***P < 0.001. Data are represented as means ± SEM, n = 3, unless specified in the figure legends.

SUPPLEMENTARY MATERIALS

www.sciencesignaling.org/cgi/content/full/10/493/eaal4165/DC1

Fig. S1. Expression and function of laminins in the developing mouse brain.

Fig. S2. Expression of YAP target genes and integrins in ventral midbrain tissue and cell lines.

Fig. S3. Analysis of the YAP–miR-130a–PTEN pathway in ventral midbrain tissue and cell lines.

Fig. S4. Effects of the LM511-YAP pathway on human NES cells differentiated into mDA neurons.

Table S1. Antibodies.

Table S2. qRT-PCR primers.

Table S3. Oligonucleotides and cloning primers.

REFERENCES AND NOTES

Acknowledgments: We thank K. Tryggvason and C. ffrench-Constant for critical reading of the manuscript, BioLamina for laminins and support to this work, members of the Arenas laboratory for their help and suggestions, and J. Söderlund and A. Nanni for the technical and secretarial assistance. Funding: This work was supported by grants from Swedish Research Council (VR projects: Developmental Biology for Regenerative Medicine, 2011-3116, 2011-3318, and 2016-01526), Swedish Foundation for Strategic Research (SRL program), European Commission (NeuroStemcellRepair), Cancerfonden (CAN 2016/572), Hjärnfonden (FO2015:0202), and Parkinsonfonden and Karolinska Institutet (SFO Thematic Center in Stem cells and Regenerative Medicine) to E.A. Author contributions: D.Z. performed most of the experiments, prepared the figures, and wrote the manuscript. S.Y. performed the electroporation experiments. E.M.T. performed the RNA-seq analysis. D.G. performed the primary cultures. C.S. and J.C.V. performed the hNES cultures and differentiation. E.A. designed the experiments, supervised the project, and wrote the manuscript. Competing interests: The authors declare that they have no competing interests.
View Abstract

Navigate This Article