EGFR trafficking has been demonstrated to be regulated by PKA activity, and inhibition of PKA activity results in internalization of the EGFR in neuroblastoma N2a cells 10. We therefore decided to examine the localization of EGFR in the livers of mice depleted of either PKA Cα or Cβ 2221. Livers from wt and KO Cα and Cβ mice were dissected and the morphology and size examined. As illustrated in Figure 1A, livers from Cα KO mice were smaller compared to wt mice. Frozen sections of liver from wt and KO of Cα and Cβ were stained using anti-EGFR and visualized using confocal immunofluorescence microscopy. We observed a reduction in membrane association as compared to the wild-type littermates for both Cα KO and Cβ KO (Figure 1B). This led us to believe that abolished expression of either Cα or Cβ led to a decrease in total EGFR levels in liver.

To test this hypothesis, cell extracts were prepared from liver and brain of Cα and Cβ KO animals and wt and heterozygous littermates, separated by SDS-PAGE for analysis by Western immunoblotting using anti-EGFR (Figure 1C, upper panel). In liver but not brain, a clear reduction in total EGFR could be observed in both Cα KO and Cβ KO animals, as compared to their wild-type mice (Figure 1C, upper panel). This was also demonstrated by immunohistochemical staining of frozen brain sections (data not shown). To correlate this to total PKA C expression, similar Western blots were analyzed using an anti-PKA C antibody, demonstrating a clear reduction of PKA immunoreactive protein in both Cα KO and Cβ KO animals, as compared to their wild-type and heterozygous littermates (Figure 1C, middle panel). The expression and stability of another EGFR family member, ErbB2, was not affected by PKA depletion (Figure 1C, lower panel), demonstrating specificity of PKA-dependent regulation of the EGFR. EGFR expression levels in heart, kidney, intestine and lung tissue from mice was also analyzed, but due to low EGFR protein levels we were unable to draw conclusions regarding comparison of EGFR protein levels in wt versus knock out animals (data not shown).

Our observation that Cβ KO animals displayed decreased expression of EGFR was unexpected as previous results have indicated that liver mainly expresses the Cα isoform 25. We therefore examined the total PKA activity in cell extracts of livers from both Cα and Cβ KO animals and their wt littermates using a Kemptide based phosphotransferase assay (Figure 1D). Both Cα and Cβ KO animals demonstrated a similar reduction in total PKA activity of more than 70%, indicating that both Cα and Cβ contribute significantly to total PKA activity in mouse liver.

EGFR expression levels may be regulated at several stages including at the level of transcription, mRNA processing, translation as well as post translation. In order to examine how the EGFR level is regulated, we compared the EGFR mRNA levels in liver of wild type animals with Cα and Cβ KO animals, using real-time PCR.

Deficiency of the PKA Cα or Cβ isoforms did not change expression of EGFR mRNA (Figure 2A,B, respectively). This demonstrated that the reduction in EGFR-levels in Cα and Cβ KO mice livers was not due to reduced mRNA expression, and indicates that a modulation of EGFR at post-transcriptional level is responsible for the reduction in EGFR.

Previously it has been reported that Cα KO mice are growth retarded and thus have significantly reduced body size compared to their wild type littermates 22. In this study an early postnatal lethality in the majority of Cα KO mice was reported. However, a small percentage of Cα KO mice survived to adulthood 22. From our preliminary studies with Cα KO mice, we hypothesized that there was a link between body size reduction and a reduced EGFR expression level in these animals. Previous studies did not detect any obvious differences in phenotype of Cβ KO mice 21. In order to examine if there are any decreases in body size for the Cβ mice, we compared body weight of wt, Cα and Cβ mice at different ages using multiple linear regression in SPSS. As expected, the Cα KO mice were significantly smaller (P < 0.0001) than their wt littermates at all ages (Figure 3). The analysis showed a significant correlation between body weight and genotype when adjusted for age (P < 0.0001). This was also the case with the Cβ KO mice (P < 0.0001). It should however be mentioned that this reduction was much less profound when compared to the Cα KO littermates. Based on this we suggest a redundant function of Cα and Cβ in regulating body size which may be associated with reduced expression of EGFR.

In order to further substantiate the correlation between depletion of PKA catalytic isoforms and expression of the EGFR protein, we studied HeLa cells treated with siRNA directed against Cα and Cβ. Following 24 h incubation with either Cα or Cβ-specific siRNA PKA specific kinase activity was measured by a Kemptide based phosphotransferase assay. This demonstrated a 30% reduction in PKA kinase activity (Figure 4A). When EGFR mRNA levels in the same samples were analyzed by RT-PCR, it was clear that the reduced activity of PKA did not result in a down-regulation of EGFR mRNA (Figure 4B). In contrast the protein level of EGFR was significantly decreased as measured by immunoreactive protein in cell extracts of control cells and cells treated with siRNA against Cα and Cβ (C, Cα and Cβ, respectively, Figure 4C).

Mice ablated (knock-out, KO) for the Cα, and Cβ genes were kindly contributed by Professor G. Stanley MckKnight (Department of Pharmacology, University of Washington School of Medicine, Seattle, 98195, USA.) 2221. All use of animals has been approved and registered by the Norwegian Animal Research Authority. The mice strains were both on a mixed C57BL/6 × 129 background and were treated and bread identically. Animals were housed in a temperature controlled (22°C) facility with a strict 12 hour light/dark cycle, and allowed the RM1 diet (Special Diet Services Ltd, Witham, Essex, UK) and water ad libitum before euthanasia. Heterozygote animals were crossed and offspring genotyped by PCR. The Cα and Cβ KO allele were detected as previously described 2238.

Immunohistochemical analysis of EGFR distribution and expression in liver sections from 6 weeks old wild type and PKA Cα and Cβ KO mice was performed on 8 μm thick sections fixed in ethanol as described previously 39. Sections were incubated with a sheep anti-EGFR antibody (Fitzgerald) over night, following detection with a Cy2-conjugated donkey anti-sheep IgG antibody (Jackson Immunoresearch). Sections were examined using a Leica TSC XP confocal microscope (Leica Microsystems) equipped with an Ar (488 nm) and two He/Ne (543 and 633 nm) lasers. A Plan apochromat 100×/1.4 oil objective was used.

Liver and brain from 6 weeks old mice, and HeLa cells, were lysed in Tris lysis buffer, pH 7.4 (50 mM Tris-HCl, 150 mM NaCl, 1 mM EDTA, 1% NP-40, 1 mM Na₃VO₄, 20 mM NaF, 1 μg/ml chymostatin, leupeptin and antipain. Liver and brain were homogenized using a Mini beadbeater (Biospec products) with 1 mm silica beads for 20 sec at 4°C. Lysates were incubated on ice for minimum 15 min, and cell debris and nuclei removed by centrifugation at 8000 rpm at 4°C for 10 min. The Dc protein detection kit was applied to measure and adjust total protein concentrations. A 5× protein sample buffer stock solution (2% SDS, 10% glycerol, 0.02% bromophenol blue and 2% β-mercaptoethanol, final concentrations) was mixed with the lysates. The samples were boiled for 5 min, and proteins separated by SDS-PAGE in 6% and 10% gels. Proteins were transferred wet to nitrocellulose membranes, rinsed in ice-cold Tris-buffered saline [TBS; 10 mM Tris, pH 8.0, and 150 mM NaCl] and incubated in blocking buffer (TBS containing 5% dry milk) for 30 min at RT. The membranes were incubated over night at 4°C with donkey anti-EGFR (Fitzgerald), rabbit anti-pan C (anti-PKAαcat, Santa Cruz biotechnology) and mouse anti-PKA C (BD Biosciences), and anti-erbB2 (Zymed). Immunoreactive proteins were recognized using anti-sheep IgG and rabbit IgG conjugated to HRP (Jackson Immunoresearch and Sigma, respectively). Proteins were detected using HRP-conjugated donkey anti-sheep IgG and anti-mouse IgG (Jackson Immunoresearch), and goat anti-rabbit IgG (Sigma) for 90 min at RT. All antibodies were diluted in TBS containing 1% w/v dry milk and 0.01% thimerosal. The filters were washed in TBS, and antigens were visualized by the enhanced chemiluminescence (ECL) method with Hyperfilm MP (Amersham Biosciences).

The PKA kinase activity in liver or HeLa cell was assayed by phosphorylation of the PKA-specific substrate Kemptide (Peninsula Laboratories) using γ-[³²P]ATP (Amersham Biosciences). The assay was performed in the presence of cAMP. Cells were washed in PBS, harvested by scraping, solubilised in 200 μl homogenising buffer (5 mM KH₂PO₄, 5 mM K₂HPO₄, 1 mM EDTA, 250 mM sucrose and protease inhibitor cocktail (Roche Diagnostics), pH 6.8) and sonicated for 3 × 5 s. The homogenate was cleared by centrifugation at 16,000 × g for 10 min at 4°C. Total protein amount was estimated by Dc protein assay (BioRad). Ten-micro liter homogenate was used in an assay mixture earlier described 40. Reactions were performed at 30°C, and stopped after 9 min by transfer to P81 phosphocellulose paper (Whatman). Filters were washed in 75 mM phosphorus acid for 1 h at room temperature, incubated with 96% ethanol for 10 min and dried. Activity was measured by liquid scintillation in 3 ml Opti-fluor™ (Packard BioScience).

Mouse liver from 6 weeks old mice was homogenized as described above. HeLa cells were washed in PBS, harvested by scraping and subjected to total RNA isolation using the RNeasy Minikit (Qiagen). RNA (1 μg) was used to make first-strand cDNA by Reverse Transcription system (Promega). Titration with cDNA was performed demonstrating that the method was able to detect small changes in sample size. The cDNA was used as templates in PCRs with primer combinations specific for mouse EGFR (5'-TCT TCA AGG ATG TGA AGT GTG-3', 5'-TGT ACG CTT TCG AAC AAT GT-3'). Human EGFR Primers used for EGFR amplification in HeLa cells were: 5'-GCC AAG GCA CGA GTA ACA AGC-3', 5'-AGG GCA ATG AGG ACA TAA CC-3'. The PCR reaction was carried out in a 20 μl final volume containing the following: H₂O up to 20 μl; 2.4 μl 25 mmol/l MgCl₂; 1 μl 10 pmol sense primer, 1 μl 10 pmol antisense primer, 2 μl LC-Faststart master mix, and 4 μl cDNA. PCR cycles were as follows: 94°C for 120 s (initial denaturation), 95°C for 5 s, 58°C for 5 s, and 72°C for 13 s (45 cycles).

HeLa cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (Gibco-BRL), 2 mM l-glutamine and maintained at 37°C and 5% CO₂. The cells were seeded in 60-mm dishes and all transfections and experiments were initiated two days after plating. Two siRNA duplexes protected by two 3'-overhanged (2'-deoxy) thymidines (dT) were synthesized by Dharmacon Research. These oligonucleotides are: PKA Cα: 5'-AAG CUC CCU UCA UAC CAA AGU-3', PKA Cβ: 5'-AAG GUC CGA UUC CCA UCC CAC-3'. HeLa cells were transiently transfected with siRNA against PKA Cα and Cβ using lipofectamine 2000 (Invitrogen) in Optimem medium (Gibco) at the final concentration of 115 nM. Nontargeting (scrambled) siRNA pool (Dharmacon) at the same concentration was used as a control. After 24 h transfection, cell lysates were collected with Tris lysis buffer and total protein concentration was estimated and adjusted by Dc protein assay (Bio-Rad).