Figure 1.
Expression of PAG during brain development.
A) Total RNA was isolated from whole C57Bl6 mouse brains (head or whole embryo for E12) of different ages as indicated. The expression of PAG-mRNA and functionally related genes was analyzed following an RT-reaction and PCR using gene-specific primers. CHK-RT-PCR produced a band at the expected molecular weight (460 bp) as well as a much weaker band at 560 bp. Sequencing revealed that the latter band represented an incompletely spliced product including intron 10. One representative out of three similar experiments is shown. B) Expression of PAG and functionally related proteins was analyzed by Western blot from whole brain lysate of postnatal C57Bl6 mice of different ages as indicated. PAG was detected here with the antiserum against the whole cytoplasmic domain. Detection with the anti-PAG peptide antiserum Ig452 resulted in a similar signal. As all structural proteins tested showed differential expression during development, equal loading was assured by determination of protein concentration (50 µg/lane) and controlled for by Coomassie staining of the gels (not shown). A double band for CHK at P1 was consistently shown with both a polyclonal anti-CHK-antiserum and a monoclonal antibody against this protein and may represent two different isoforms or phosphorylated forms. A similar double band has previously been shown in Western blot [55]. One representative out of four similar experiments is shown.
Figure 2.
In situ distribution of PAG transcripts.
Film autoradiographs of horizontal mouse brain sections showing PAG mRNA distribution in A) adult and B) postnatal (P1) mice. The respective sense controls are shown below (C, D). In the adult brain most prominent mRNA expression was detectable in septum (S), medial habenula (MHb), thalamic nuclei (Th), hippocampus (Hi), and cerebellum (Cb). One representative out of two similar experiments is shown.
Figure 3.
Localization of PAG transcripts in specific regions of adult brain.
Representative dark-field micrographs (A, C, E, G, I, L) illustrating the hybridization and the respective bright-field (B, D, F, H, K, M) micrographs in adult mouse brain. Pictures are taken from horizontal sections, with the exception of L and M, which are photographed from frontal sections. A, B) Septal nuclei (SepN), C, D) habenula (MHb), E, F) hippocampus: CA2 and 3 and dentate gyrus (DG), G, H) cerebellum (ML: molecular layer, Pu: Purkinje cell layer, GL: granular cell layer), I, K) thalamus (Th), L, M) pineal gland (PG). Scale bar 200 µm.
Figure 4.
Binding partners of PAG in postnatal and adult brain.
A) PAG immunoprecipitates from whole brain lysates of newborn, 6 month old, and adult mice were analyzed by Western blot. Each association was shown in at least three independent experiments. B) Csk and CHK were immunoprecipitated from lysates of adult mouse brain. Western blot analysis with CHK and Csk-antibodies, respectively, reveals no cross-reactivity. C) HEK-293 cells were transfected with a mouse-Csk-CFP construct versus vector containing CFP alone. Over-expression is demonstrated with the anti-Csk-antibody. There is no cross-reactivity of the anti-CHK-antibody with the over-expressed Csk. Csk-CFP is indicated with an arrow, endogenous Csk with an asterisk.
Figure 5.
Reduced Csk-recruitment to lipid rafts in P1- Pag1-/- mice.
Whole brains from A) P1 wild type C57Bl6 (WT) and Pag1-/- mice or B) 3 month old WT/ Pag1-/- littermate pairs were lysed in GEM-lysis buffer and subjected to sucrose-density centrifugation. Fractions were analyzed by Western blotting. PAG was detected using the Ig452 antiserum. Lipid raft fractions were identified by binding of cholera toxin subunit B (Ctx) binding to GM1-lipids in a dot blot from the fractions. * Arrowheads indicate non-specific bands in 3 months old Pag1-/- and wild type brain extracts, see also Fig. 6A.
Figure 6.
Quantification of kinases in GEM gradients.
The kinase bands in each fraction represented in Fig. 5 were quantified from three independent experiments for A) the P1 and B) the 3 month time point. The summative signal from GEM fractions in relation to the total signal is presented as ratio to wildtype for each kinase. GEM containing fractions were determined by choleratoxin staining. For phosphotyrosine 529 the total signal in GEMs and the total signal on the blot are given in separate graphs, as most of the signal localizes to GEM fractions and the total amount of pY529 is reduced in PAG1-/- mice (see also Fig. 6). Given are the means +/− SEM. •p<0.05, Student's t-test, all other pairs did not show significant differences.
Figure 7.
Decreased phosphorylation of Src-kinase negative-regulatory tyrosine 529 in P1- Pag1-/- mice.
For P1-material, protein extracts were prepared from separate newborn litters of time-mated Pag1-/- mice and wild type C57Bl6 mice. For 3 months-material, protein extracts were prepared from adult Pag1-/- and wild type littermates of Pag1+/- crossings. A) Incubation with anti-PAG-antiserum Ig452 detects a band at 75-80 kDa in brain extracts of 3 month old Pag1-/- mice. Co-incubation with the antigenic peptide abolished the specific PAG-band in WT mice, but left the additional band unchanged in both WT and Pag1-/- extracts, indicating that this band is not peptide-specific. B) Equal concentrations of protein from each sample were analyzed by immunoblotting for PAG (Ig452), Csk, CHK, pY529, Fyn, and Src. Actin staining served as loading control. C) Phosphorylation of the negative regulatory tyrosine (pY529) was quantified in different experiments and is presented as the ratio to wild type. Mean and SEM are given for n = 4 (P1) and n = 6 (3 months) independent experiments. ** p<0.01, n.s. not significant, Student's t-test.
Figure 8.
Src kinase activity is altered in both postnatal and adult Pag1-/- brain.
GEM-containing fractions (identified via cholera toxin subunit B-binding) of A) postnatal or B) adult wild type and Pag1-/- brains were pooled, 10% was retained for a Western blot control, the rest was split into two samples, from which Fyn and Src, respectively, were precipitated. Precipitates were again split in halves and subjected to either in vitro kinase assay (upper panels) or Western blot (lower panels) which controlled for equal amounts of kinase precipitated. C) and D) Kinase activities are presented as ratio to wild type. E) and F) Phosphorylated Y529 was quantified in the Western blot controls of Fyn and Src immunoprecipitates and normalized to the amount of precipitated SFK. Given are the means +/- SEM, n = number of independent experiments. p<0.05, ** p<0.01, n.s. not significant, Student's t-test.
Figure 9.
Developmental expression and Csk-binding of PSD93.
A) Whole brain lysates from WT mice of the indicated ages were Western blotted and stained for PSD93. B) Brains of P1 wild type and Pag1-/- mice and 3 month old WT/ Pag1-/- littermate pairs were lysed at concentrations of 3 ml/g and 4.5 ml/g, respectively, in GEM-lysis buffer and subjected to sucrose-density centrifugation. Fractions were analyzed by Western blotting and staining for PSD93. Lipid raft fractions were identified by binding of cholera toxin subunit B (Ctx) binding to GM1-lipids in a dot blot from the fractions. C) Densitometric quantification of the PSD93 signal in GEM-fractions as percentage of the total PSD93-signal. For both time points, the PSD93 signal in fractions 2-5 was expressed as a ratio to control. This most likely overestimates the lipid raft localized material in P1, as only fraction 2 and 3 contain rafts (n = 3 independent experiments for each time point) D) Csk-immunoprecipitates from whole brain lysates of WT mice (ages as indicated), stained for PSD93 and Csk. Densitometric quantification of the relative amount of PSD93 bound to Csk in brains of P1 and 3 month old mice (n = 3 independent experiments) E) Csk-immunoprecipitates from whole brain lysates of 3 month old WT and Pag1-/- mice stained for PSD93 and Csk. Two exposures are given for the PSD93-signal in the left panel. Comparison of whole brain lysates of 3 month old WT and Pag1-/- to P1. Given are the means +/- SEM, n.s. not significant, Student's t-test.