All animal studies were performed in accordance with local animal welfare regulations, as this project has been approved by the animal experimental committee (Dier ethische commissie, Universiteit van Amsterdam, DEC-UvA), and international guidelines.

The transgenic Mest^tm1Lef strain was a generous gift of Prof. Dr. Azim Surani from the Wellcome Trust/Cancer Research UK Gurdon Institute of the University of Cambridge (Lefebvre et al., 1998). Mest^tm1Lef mouse-line was back-crossed with the C57BL/6 line and KO and wild type (WT) embryos and adults were generated by crossing heterozygous Mest mutant mice. As no differences were expected between males and females, both genders were used in the experiments performed. WT embryos and adults to study normal gene expression patterns were generated by crossing C57BL/6 mice. Pregnant [embryonic day 0.5 (E0.5) was defined as the morning of plug formation] and adult mice were sacrificed by cervical dislocation. Embryos and brains were collected in 1x PBS and immediately frozen on dry-ice, or fixed by immersion for 3–12 h in 4% paraformaldehyde (PFA) at 4°C. After PFA incubation, samples were washed in 1x PBS and cryoprotected O/N at 4°C in 30% sucrose. Embryos and brains were frozen on dry-ice and stored at -80°C. Cryosections were sliced at 16 μm, mounted on Superfrost Plus slides (Thermo Fisher Scientific), air-dried, and stored at -80°C until further use.

E17.5 Lrp6 KO embryonic tissues were a kind gift of Prof. Dr. E. Arenas from the Karolinska Institute in Stockholm, Sweden.

Genotyping of the Mest^tm1Lef strain was performed as described (Lefebvre et al., 1997). In short: PCR was performed in two reactions. The first reaction contains 50 ng of genomic DNA together with a common FP 5′-CTGGCTGCGTACCTGCACATC-3′ and RP1 5′-GCTTCCGACCACACCGACAG-3′ binding at respectively exon 2 and exon 3 of the Mest gene, resulting in a WT product of 1.4 kb. The second reaction contains the common FP and RP3 5′-AGACCGCGAAGAGTTTGTCCTC-3′ binding at respectively exon 2 of the WT gene and the pIFS sequence upstream of the IRES, resulting in a 1.15 kb-product of the mutant allele.

Fluorescence immunohistochemistry was carried out as described previously (Kolk et al., 2009; Fenstermaker et al., 2010). Cryosections were blocked with 4% heat inactivated fetal calf serum (HIFCS) or 5% normal donkey serum (for sheep primary antibodies) in 1x THZT and incubated O/N with a primary antibody [Rb-TH (Pelfreeze 1:1000), Sh-TH (Millipore AB1542, 1:1000)]. The next day sections were incubated with a secondary Alexafluor antibody (anti-rabbit, anti-sheep) diluted 1:1000 in 1x TBS for 2 h at RT (20°C ± 1°C). After extensive washing in 1x PBS, slides were embedded in Fluorsave (Calbiogen) and analyzed with the use of a fluorescent microscope (Leica).

Quantification of TH-expressing cells in the adult and E17.5 midbrain was performed in ImageJ as follows. TH-expressing cells of the SNc, ventral tegmental area (VTA), and both, were counted in 8–12 (adult) (WT n = 3, KO n = 3) and 6 (E17.5) (WT n = 2; KO n = 2) (matching) coronal sections. Cells were counted as TH⁺ cells when TH staining co-localized with a nuclear DAPI-staining. The SNc and VTA were determined based on anatomical landmarks. The SNc was clearly determined in sections rostral from the fasciculus retroflexus, whereas the mdDA area was divided in the VTA and SNc in sections containing, and caudal of, the fasciculus retroflexus. The distinction between the SNc and VTA was made based on the tracts of the medial lemniscus, positioned in between the SNc and VTA. Statistical analysis was performed via a one-tailed student’s t-test.

Intensity of TH-levels in the striatum of Mest WT and KO animals was measured with the use of ImageJ, as described previously (McCloy et al., 2014). Intensity was measured in 6 matching coronal slices in dorsal and ventral parts of the striatum (WT n = 3, KO n = 4). After retraction of the background intensity (normalization) we performed statistical analysis via a one-tailed student’s t-test.

In situ hybridization with digoxigenin (DIG)-labeled probes was performed as described previously (Smidt et al., 2004). Fresh frozen sections were fixed in 4% PFA for 30 min and acetylated with 0.25% acetic anhydride in 0.1 M triethanolamine for 10 min. Probe hybridization was carried out at 68°C O/N with a probe concentration of 0.4 ng/μl in a hybridization solution containing 50% deionized formamide, 5x SSC, 5x Denhardt’s solution, 250 μg/ml tRNA Baker’s yeast, and 500 μg/ml sonificated salmon sperm DNA. The following day slides were washed in 0.2x SSC for 2 h at 68°C followed by blocking with 10% HIFCS in buffer 1 (100 mM TricHCl, pH = 7.4 and 150 mM NaCl) for 1 h at RT. DIG-labeled probes were detected by incubating with alkalin-phosphatase-labeled anti-DIG antibody (Roche, Mannheim), using NBT/BCIP as a substrate.

DIG in situ hybridization was performed with the following probes: 491 bp Th fragment bp 252–743 of mouse cDNA, 359 bp Dat fragment bp 762–1127 of rat cDNA, 648 bp Mest fragment bp 648–1298 of mouse cDNA, 993 bp Lrp6 fragment bp 1015–2008 of mouse cDNA. If in situ hybridization was not followed by TH-DAB IHC, slides were washed 2x5 min in T₁₀E₅, dehydrated with ethanol, and embedded in Entellan.

After Mest and Lrp6 DIG in situ hybridization, sections were immunostained for TH. Slides were incubated in 0.3% H₂O₂ in Tris-buffered saline (TBS) for 30 min at RT. Thereafter, blocking was performed with 4% HIFCS in TBS. Slides were incubated O/N with primary antibody Rb-TH (Pelfreeze, 1:1000) in TBS. The following day slides were incubated for 1 h with goat-anti-rabbit biotinylated secondary antibody (Vector, 1:1000) in TBS, followed by incubation with avidine-biotin-peroxidase reagents (ABC elite kit, Vector Laboratories 1:1000) in TBS. The slides were stained with DAB (3,3′-diamino-benzidine) for a maximum of 10 min, dehydrated with ethanol and embedded with Entellan.

Nissl staining was performed as described previously (Veenvliet et al., 2013). Fresh frozen sections of 3- and 8-month old Mest KO and WT littermates were used.

The climbing test was performed as described previously (Costall et al., 1978) Smidt et al., 2004 during the active phase of the mice. Three-months old mice were placed in climbing cages and acclimatized for 30 min. Statistical analysis was performed via a two-tailed student’s t-test.

Mice were sacrificed by cervical dislocation and 250 μm-thick coronal brain slices were prepared on a vibroslicer (Leica VT1000S) using ice-cold modified artificial cerebrospinal fluid (ACSF) containing 120 mM choline chloride, 3.5 mM KCl, 0.5 mM CaCl₂, 6 mM MgSO₄, 1.25 mM NaH₂PO₄, 25 mM NaHCO₃, and 25 mM glucose, continuously bubbled with 95% O₂-5% CO₂ (pH = 7.4). Slices containing the striatum were kept submerged in a recording chamber at RT and superfused with recording ACSF containing 120 mM NaCl, 3.5 mM KCl, 2.5 mM CaCl₂, 1.3 mM MgSO₄, 1.25 mM NaH₂PO₄, 25 mM NaHCO₃, and 25 mM glucose, continuously bubbled with 95%O₂-5%CO₂.

Carbon fibers for recording were made in house. A ∼9 μm thick carbon fiber (generous gift of Dr. Ingo Willuhn, NIN, Amsterdam, The Netherlands) was aspirated in a boroscilicate glass capillary using vacuum suction. The capillary was subsequently pulled on a horizontal electrode puller (P-97, Sutter Instruments, Novato, USA). The exposed carbon–fiber was trimmed to approximately 0.5 mm and the fiber was glued and coated with transparent nail polish, leaving ∼100 μm of the tip exposed. The carbon-fiber electrode was backfilled with 1M KCl and connected to the headstage of an EPC9 patch clamp amplifier controlled by PULSE software (HEKA Electronics, Lambrecht, Germany). The carbon-fiber was placed in either the dorsal of ventral striatum and lowered into the slice (∼50 μm below the surface) at ∼100 μm distance from a bipolar stimulation electrode. DA release was evoked by electrical stimulation (10–500 μA, 1 ms) and recorded by ramping the potential of the carbon-fiber from -500 mV to 1 V and back to -500 mV at 300V/s at 10 Hz vs. an Ag/AgCl reference electrode. Signals were filtered at 3 kHz and samples at 10 kHz. Voltammograms obtained before stimulation were off-line subtracted from those obtained after stimulation using Igor Pro (Wavemetrics, Tigard, OR, USA). Carbon-fibers were calibrated in ACSF containing 1–10 μM DA before and after each experiment. Statistical analysis was performed via a two-tailed student’s t-test.

To investigate which genes could be important during mdDA neuronal birth and later during neuronal differentiation, we performed meta-analysis on the time-line genome-wide expression study performed earlier in our lab (Chakrabarty et al., 2012). We found that, amongst other factors, Mest is highly expressed during the early phase of mdDA neuronal development (Figure 1A). Its high expression early during development suggests that it could be an important factor in the onset of mdDA neuronal birth and initial differentiation.

To validate the data from the time-line genome-wide expression study and define the spatial distribution of Mest in the midbrain area, we analyzed Mest transcript expression by means of in situ hybridization during differentiation (Figure 1B) and in the adult stage (Figure 1C). Mest transcript is detected from E11.5 onward and remains expressed in the adult midbrain. At E11.5 Mest is mainly expressed in the floor plate (FP) and basal plate (BP) of the ventricular zone (VZ) where it remains expressed at least until E14.5. At E12.5 Mest overlaps with a small part of the medial TH-expressing neuronal population in rostral areas (Figure 1B-1). Whereas at E14.5 Mest expression is decreased in the VZ and displays a broad expression, partially overlapping at lateral-caudal areas with TH expression (Figure 1B-2). Later during development, at E16.5, Mest is most prominently expressed in the future SNc region (Figure 1B-3). In the adult mdDA area Mest transcript can be detected in some TH-expressing cells of the SNc (Figure 1C-1 and 3 black arrowheads) and in cells surrounding nigral mdDA neurons (Figure 1C-2). Together, the spatial-temporal expression data confirmed the initial time-line genome-wide expression profile. At early developmental stages Mest is strongly expressed in the midbrain and shows a relatively broad expression pattern. At late developmental stages Mest becomes more restricted to lateral parts of the TH-expressing population and is eventually present in neurons located in the SNc (adult stage).

The presence of Mest in the VZ at early developmental stages and specifically in the SNc at late developmental stages, suggests that it may be involved in mdDA neuronal development and adult function of mdDA neurons located in the SNc. To investigate the role of Mest in the neurochemical and neuroanatomical integrity of the mdDA system, we examined the presence of the mdDA neuronal marker TH in Mest KO animals at E14.5 and E17.5, and both Th and Dat in the adult stage.

At E14.5 the mdDA neuronal pool appears to be smaller and at E17.5 the SNc seems less cell-dense in the KO (Supplementary Figure S1). This initial effect of Mest deletion becomes more clear in adult mice (Figure 2). We examined the presence of mdDA neurons in 3- and 8-months old Mest KO and WT littermate controls. 3-months old Mest KO animals lose expression of Th specifically in the SNc (Figure 2A left panel, black arrowheads). 8-months old Mest KO animals show a more severe effect on Th than 3-months old Mest KO (Figure 2A right panel, black arrowheads), suggesting a progressive loss of Th. Similar defects were observed when Dat expression was analyzed in adjacent sections at these stages in Mest ablated animals (Figure 2B). Nissl staining on adjacent sections shows that in the Mest KO DA cell density is decreased (Figure 2C-1 to 4), indicating that loss of Th and Dat expression represents loss of mdDA neurons.

To substantiate the mdDA cell-loss and to asses whether the defect is different between SNc and VTA, we performed immunohistochemistry for TH in 3- and 8-months old Mest KO and WT littermates (Figure 3A). TH-immunohistochemistry on 3- (Figure 3A-1 to 4) and 8-months (Figure 3A-5 to 8) old Mest KO animals shows a similar loss of TH-expressing cells as previously show on transcript level. Quantification of the amount of TH⁺ cells in the VTA and SNc showed that 3-months old Mest KO animals (n = 3) (Figure 3B) have significantly less (∼25% loss) TH-expressing cells than WT controls (n = 3; p < 0.01, one-tailed t-test). Separate quantification of the VTA and SNc (see Materials and Methods for definition), shows that loss of mdDA cells is significant for the SNc (∼32% loss p < 0.01, one-tailed t-test). Quantification of TH⁺ neurons in 8-months old Mest ablated animals (n = 3) (Figure 3C) shows a significant loss of TH-expressing cells throughout the mdDA neuronal population (∼32% loss p < 0.01, one-tailed t-test), which is most severe in the SNc (∼42% loss p < 0.01, one-tailed t-test), although we also detected a significant loss of ∼12% in the VTA at this stage (p < 0.01, one-tailed t-test). In comparison of the 3- and 8-months data the progressive loss of TH neurons is only detected in the SNc region (Figure 3D) (∼10% more loss in the 8-months SNc compared to the 3-months SNc p < 0.01, one-tailed t-test). Taken together, these data suggest that ablation of Mest causes a progressive loss of mdDA neurons in the SNc.

We showed that Mest ablation leads to a progressive loss of SNc DA neurons, which are known to project to the striatum in the adult brain (Ikemoto, 2007), suggesting that the dopaminergic output to this area may be affected in Mest KO animals. To confirm this possibility, we analyzed the presence of TH in the striatum in Mest KO animals (Figures 4A,B). In agreement with the above described cell loss, we find a 50% lower signal of TH protein in the striatum of Mest KO brains compared to controls (Figure 4C) (WT n = 3; KO n = 4; p < 0.01 one-tailed t-test). This effect was evenly distributed over the dorsal (50%; p < 0.05) and ventral (49%; p < 0.01) region (Figure 4C).

The observed lower TH protein levels in the striatum indicate that the DA output parameters may be affected in the striatum. To confirm this we analyzed the DA release parameters through cyclic voltammetry on dorsal and ventral regions of the striatum. In both regions WT animals (n = 4) show a DA-release of ∼3 μM after local field stimulation, whereas Mest KO animals (n = 4) display a decreased DA-release toward ∼1.5 μM (Figure 4D). This significant reduction in DA-release (50% less in the KO, p < 0.001 two-tailed t-test) confirms the functional consequence of the observed mdDA cell loss and observed lower TH protein levels in the striatum. Interestingly, although the DA release is impaired in Mest KO animals, the re-uptake machinery seems to be functionally intact, as the DA-decay is similar for both genotypes in both analyzed striatal areas (Figure 4E). General characteristics, like the peak amplitude of the oxidation current and the DA response (example of peak amplitude and DA response in the total and dorsal and ventral striatum in the WT: Supplementary Figures S2A–C) are equally unaltered in Mest KO animals (Supplementary Figure S2D). Taken together, the data show that the initial observed cell loss as a consequence of Mest ablation, is represented by a functional component of a reduction of DA-release in the striatum.

The observed loss of DA release in the striatum suggests that the animal would be affected in its behavior (Sonnier et al., 2007). In order to substantiate this, we investigated the climbing behavior in WT and Mest KO animals (Costall et al., 1978; Smidt et al., 2004). Adult Mest KO animals (n = 4) show a significantly lower climbing score after 30 min habituation than WT controls (n = 5) (p < 0.01, two-tailed t-test) (Figure 4F). The low climbing activity in the Mest KO mice indicates lower activation of DA receptors in the striatum, which is likely to be caused by the decrease in DA release as shown previously (Protais et al., 1976). Taken together, the reduced DA output is reflected by behavioral deficits in Mest ablated animals.

Thus far we have shown that Mest is involved in the development of the SNc, although the question remains how Mest is involved in this process. Studies performed by Jung et al. (2011) in adipocytes indicate that Mest regulates the balance of non-canonical and canonical WNT-signaling by inhibition of glycosylation and maturation of LRP6. If Mest has a similar function in mdDA neurons, loss of Mest would result in an increase of mature LRP6, thus favoring canonical over non-canonical WNT-signaling. Castelo-Branco et al. (2010) showed that loss of Lrp6 leads to a delay in mdDA neuronal development and they claim that at E17.5 this effect was completely recovered. Based on the function of Mest during adipocytogenesis, we hypothesize that in developing mdDA neurons Mest has a similar function and loss of Mest may lead to an increase of LRP6 on the cell-membrane, eventually resulting in cell loss in the SNc (Supplementary Figure S3). If Mest indeed functions via regulation of LRP6, loss of Lrp6 may lead to an increase of TH⁺ cells in the SNc, and thus an opposite phenotype compared to the Mest KO. To examine whether Mest functions (partially) via regulation of LRP6 maturation, we set out to determine whether Lrp6 is expressed in similar areas as Mest at late developmental stages. To this end we analyzed the expression of Lrp6 compared to the TH-expressing population at E16.5 and E18.5 (Figure 5A). Lrp6 is ubiquitously expressed throughout the brain at both stages (Figure 5A). It is expressed in the midbrain at late developmental stages and overlaps with TH-expressing mdDA neurons at both E16.5 and E18.5.

Since Lrp6 mutants are not viable (Castelo-Branco et al., 2010), we analyzed the mdDA neuronal population in the Lrp6 KO at E17.5. Analysis of TH expression as a readout for mdDA neuronal presence (Figure 5B) shows that the mdDA neuronal pool appears to be broader and more cell-dense in Lrp6 KO embryos at E17.5 compared to WT (Figure 5B blow-ups). The amount of TH⁺ neurons within the midbrain of Lrp6 KO embryos (n = 2) shows a small upward trend compared to WT littermates (n = 2), which can be detected in both the VTA and the SNc (Figure 5B left panel). Taken together, these data show that Lrp6 is expressed in the SNc during late stages of mdDA neuronal development and that loss of this gene possibly results in an increase in TH-expressing neurons at E17.5.

Above we have shown that loss of Mest results in a progressive loss of mdDA neurons in parts of the SNc. Previously it has been shown that loss of Pitx3, a marker for mdDA neurons, similarly results in a specific loss of cells in the SNc, although this loss is not progressive (Smidt et al., 2004, Smits et al., 2005; Maxwell et al., 2005). In order to detect whether the neurons lost in the Mest mutant animals are part of the same population lost in the Pitx3 mutant, we compared the SNc of Pitx3 mutants at 3-months of age to the SNc of Mest mutants at 8-months of age (Figure 6A, anatomically matched). Interestingly, the region of Th-expressing neurons that is lost in the SNc of the Mest KO appears to be present in the Pitx3 mutant (Figure 6A black arrowheads) and vice verse, suggesting that the SNc consists out of different neuronal subsets that are dependent on distinctive signaling for their maintenance (Figure 6B). Noteworthy, Pitx3 is specifically expressed in neurons of the mdDA system (Smidt et al., 2004; Maxwell et al., 2005) and Mest is already expressed in the VZ of the midbrain and later becomes more restricted to SNc neurons. Taken together, these data indicate that the SNc consists out of different subsets that are specified by distinct molecular coding and have a specific vulnerability for neuronal degeneration (Figure 6B).