Co-ordinated Action of DISC1, PDE4B and GSK3β in Modulation of cAMP Signalling
Becky C. Carlyle1,2, Shaun Mackie1, Sheila Christie1, J. Kirsty Millar1*, David J. Porteous1*
1The Centre for Molecular Medicine, Western General Hospital, The University of Edinburgh, Edinburgh EH4 2XU, UK, 2Now at Yale Department of Psychiatry, 300 George Street, New Haven, 06510, USA
* These authors contributed equally to this work
Supplementary Materials & Methods
Western blots
Protein samples were lysed in RIPA buffer, and separated by electrophoresis using the pre-prepared Invitrogen Western blotting system (Invitrogen). The αDISC1 antibody, a kind gift from Professor Akiyama, University of Tokyo [s1], was diluted to 1:10,000 in blocking buffer (PBS-Tween, 5% Marvel), and incubated with membranes overnight. Membranes were washed, then incubated with Swine Anti Rabbit HRP conjugated secondary antibody (Dako). A 100kDa DISC1 immunoreactive band was detected using ECL Plus Western blotting detection reagents (Amersham Biosciences). GAPDH was used as a loading control, for which the primary antibody was the mouse monoclonal MAB374 anti GAPDH antibody (Chemicon, Millipore) diluted 1:80,000, and the secondary Donkey Anti Mouse HRP conjugated antibody (Dako) diluted 1:80,000.
[s1] Ogawa F et al, Biochem and Biophys Res Comm, 2005; 338; 771-776
Quantitative Real Time PCR
Cells were harvested and RNA extracted using QIAshredder and QIAgen RNeasy Minikit (Qiagen) according to manufacturer’s protocols. To remove genomic DNA contamination, we performed on column DNase digestion reactions. cDNA was generated using the 1st Strand cDNA Synthesis Kit (Roche). Real time PCR was performed using iQTMCYBR Green Supermix (Biorad) on the MyIQ iCyler (Biorad). DISC1 transcription levels were normalized to the geometric mean of 3-4 different reference genes [s2]. Primer sequences are as follows:
PDE4A 5’-CCAGAACCTCAGCAAGCGCCA-3’
5’- TCCGGAGGACCTGGATGCGG-3’
PDE4B 5’-CCGATCGCATTCAGGTCCTTCGC-3’
5’-TTTCCATTCCCCTCTCCCGCT-3’
PDE4D 5’-TGCTCAGGTCTTGGCCAGTCTGC-3’
5’-TCCTCCAGGGTCTCGCTGGC-3’
B2M 5’-AAAGATGAGTATGCCTGCCG-3’
5’-AGCAAGCAAGCAGAATTTGG-3’
GAPDH 5’-TCAAGATCATCAGCAATGCC-3’
5’-ATCCACAGTCTTCTGGGTGG-3’
PPIA 5’-AGACTGAGTGGTTGGATGGC-3’
5’-TCGAGTTGTCCACAGTCAGC-3’
RPLPO 5’-AATGGCAGCATCTACAACCC-3’
5’-CCGTTGATGATAGAATGGGG-3’
[s2] Pfaffl, M.W, Nucleic Acids Research, 2001; 29, e49.
Cell treatments
All treatments were performed on serum starved SH-SH5Y cells as detailed in the table:
The IC50 of LiCl for GSK3β inhibition is 2 mM [s3], and for SB216763 the IC50 is 34 nM [s4]. These studies suggest that at 3 mM and 10 mM LiCl, GSK3β activity is reduced by 60% and 85% respectively. At 3 μM SB216763, GSK3β activity is reduced by 95%. After 30 minutes of treatment with 10 μM Forskolin or 10 mM LiCl there were no detectable changes in PDE4B or PDE4D protein levels, as determined by Western blotting of the same lysates used for PDE assays with the pan PDE4B and pan PDE4D antibodies (kind gifts from Professor Miles Houslay, University of Glasgow, Figure S3).
[s3] Ryves & Harwood, Biochem & Biophys Research Comms. 2001; 280:720-725
[s4] Coghlan MP et al, Chemistry & Biology, 2000; 7: 793-803.
Phosphodiesterase activity assays
PDE Assays were performed to a protocol adapted from Marchmont Houslay [s5]. Cell and tissue lysates were prepared using 3T3 buffer plus Complete Protease Inhibitor cocktail tablets (Roche) and Phosphatase Inhibitor II (Calbiochem). A standard amount of protein lysate was diluted in 20 mM pH7.5 Tris solution (Fisher) to a total of 50 μl. To measure the PDE4 component of total activity duplicate samples were incubated with 10 μM Rolipram (Sigma). Significant changes were determined using unpaired two tail Student’s T-tests. Our analysis indicates that PDE4 is the dominant active PDE in SH-SY5Y cells, generally accounting for greater than 75% of the total measureable PDE activity (total measureable PDE activity in our assays ranges from 80-120 pm/mg/min, the addition of 10 μM Rolipram, generally decreases total measureable PDE activity to a range of 20-40 pmoles/min/mg). Evidence from the literature also suggests the presence of PDE1C and PDE2 [s6].
[s5] Marchmont RJ, Houslay MD. Biochemical Journal. 1980;187(2):381-92.
[s6] Jang JY, Experimental & Molecular Medicine. 2001; 33:37-45
cAMP AlphaScreen (Perkin Elmer)
Cyclic AMP accumulation assays were performed according to the standard adherent cell protocol described in the AlphaScreen Manual, with the exception that a PDE inhibitor was not added to the wells. Significant changes were determined using unpaired two tail Student’s T-tests.
Inducible fl-DISC1 overexpressing SH-SH5Y line
We used the TRex system (Invitrogen) to inducibly overexpress DISC1. Full length untagged human DISC1 (fl-DISC1, AF222980.1) was subcloned into the pcDNA4TO vector (Invitrogen) using PCR primers which incorporated EcoRI and NotI restriction sites. The EcoR1/NotI digested vector was further treated with Shrimp Alkaline Phosphatase (Roche), before ligation of the DNA ORF insert the Rapid DNA Ligation Kit (Roche). SH-SY5Y cells were cotransfected with a DNA ratio of 6:1 pcDNA6/TR:pcDNA4TODISC1, using a standard Lipofectamine protocol (Invitrogen). Cells transfected with both constructs were dual selected with 50μg/ml Zeocin and 2 μg/ml Blasticidin (both Invitrogen) for 2 weeks before successful colonies were picked and tested. fl-DISC1 expression was induced by treatment for 24 hours with 1μg/ml tetracycline (Invitrogen). Induction of DISC1 mRNA was confirmed by quantitative PCR using the MyIQ iCycler (BioRad) to a level of 5.07 fold (+/-0.21, n=3) compared to uninduced cells (1.1 +/-0.28, n=3). Induction of DISC1 protein was confirmed by immunoblotting (Figure S1a, S1b). Induction of fl-DISC1 expression had no detectable effect upon expression of PDE4A, B or D mRNA in SH-SY5Y within the time-course of the PDE4 assays performed (Figure S2a). Similarly, there were no changes in PDE4 protein expression (data not shown). PDE4C was not detected in this cell type.
5’ fl-DISC1 sense primer incorporating 5’ EcoRI restriction site:
5’-gatcGAATTCGCCGCCATGCCAGGCGGGGGTCCTCAGGGC-3’
3’ fl-DISC1 anti-sense primer incorporating 3’ NotI restriction site:
5’-gatcGCGGCCGCTCAGGCTTGTGCTTCGTGGACACCAGC-3’
Supplementary Figures
Figure S1a: Sample Western blot showing induction of 100 kDa DISC1 protein following 24 hours application of tetracycline to the SH-SY5Y cell line stably transfected with the fl-DISC1 expression construct. At a short exposure time of 30 seconds, overexpression of DISC1 is clearly visible in induced cells using the α-DISC1 antibody.
S1b: An exposure time of over 5 minutes is required to visualise endogenous protein using this antibody.
Figure S2a: Graph depicting PDE4 isoform mRNA expression in SH-SY5Y following induction of fl-DISC1 in comparison to uninduced cells (WT DISC1). Since different primer pairs were used to detect each isoform, no direct comparisons of isoform expression can be made. n=2
Figure S3a: Quantification of western blots of cell lysates used to assay PDE4 activity, probed with a pan PDE4B antibody show no significant changes (P<0.05) in the expression of PDE4B protein after 30 minutes of treatment with 10 μM Forskolin or 10 mM LiCl. The 82 kDa band obtained with this antibody was quantified using ImageJ densitometry software.
Figure S3b: Quantification of western blots of cell lysates used to assay PDE4 activity, probed with a pan PDE4D antibody show no significant changes (P<0.05) in the intensity of a 98 kDa PDE4D species band after 30 minutes of treatment with 10 μM Forskolin or 10 mM LiCl.
Figure S3c: Quantification of western blots of cell lysates used to assay PDE4 activity, probed with a pan PDE4D antibody show no significant changes (P<0.05) in the intensity of an 86 kDa PDE4D species band after 30 minutes of treatment with 10 μM Forskolin or 10 mM LiCl.
Figure S3d: Quantification of western blots of cell lysates used to assay PDE4 activity, probed with a pan PDE4D antibody show no significant changes (P<0.05) in the intensity of an 80 kDa PDE4D species band after 30 minutes of treatment with 10 μM Forskolin or 10 mM LiCl.