Adenoviral gene delivery is commonly utilized for functional over-expression and characterization of a gene-of-interest in mammalian cells. However, it is widely recognized that viral infection of human cell lines with homologous adenovirus vectors induces the activation of phosphatidylinositol 3-kinase (PI3-kinase), subsequently leading to the phosphorylation of protein kinase B/Akt on Serine 473 121114. Akt is central for cellular responses mediated by insulin (reviewed in 43) and activation of Akt is hence used as a common marker in insulin signalling. Therefore, while studying the impact of over-expression of various protein kinases on insulin signalling in experimental human cell lines, we investigated the influence of recombinant adenovirus on Akt phosphorylation in SHSY-5Y and C3A cells. Cultured cells were first infected with recombinant adenovirus expressing green fluorescent protein (AdGFP) as a model protein, and phosphorylation of Akt was studied with specific Ser⁴⁷³-phosho-Akt antibodies (ELISA and western blot). Viral titration was performed at multiplicity of infection (MOI) between 10 and 50 and GFP expression was determined 48 hours post infection (p.i.) by flow cytometry gated for intact adenovirus infected cells. Figure 1 displays side scatter against forward scatter for SHSY-5Y and C3A cells infected with AdGFP at MOI 10, 20, and 50, respectively. The resulting transduction efficiencies were determined from expression histograms of the populations of intact cells (illustrated as the upper cell population in respective scatter plot) which are shown in the lower panels of Figure 1 for each respective cell line, and was found to be 26.5, 58.6, and 86.9 % for SHSY-5Y cells, and 26.2, 47.8, and 67.3 % for C3A cells infected at MOI 10, 20, and 50, respectively.

Next, the influence on Akt phosphorylation following adenovirus transduction was investigated. As anticipated, the degree of phosphorylation of Akt increased upon adenovirus infection and this increase was dose-dependent when both SHSY-5Y as well as C3A cells were subjected to adenovirus infection. The degree of phosphorylation increased 4 and 5-fold for SHSY-5Y cells, and 1.8 and 2.4-fold for C3A cells, infected at MOI 10 and 50, respectively, as determined with an ELISA assay measuring phosphoserine 473 in PKB. These results were confirmed under various experimental conditions and are summarized in the upper part of Table 1 for AdGFP. To further validate the results, the data obtained with the commercial ELISA were verified with western blot using specific Ser⁴⁷³-phosho-Akt antibodies. Again, phosphorylation on serine 473 in infected but unstimulated cells was seen, thus confirming the activation of PKB by adenovirus infection measured by two different methods (Figure 2, lane 1 and 3). Taken together, the results from these experiments show that when transduced with adenovirus up to MOI 50, SHSY-5Y and C3A cells respond with an increased phosphorylation of Ser473 Akt indicating that using adenovirus as a gene transfer vehicle in SHSY-5Y and C3A cells may influence the cells responsiveness to insulin treatment.

The use of engineered baculovirus as a gene transfer vehicle in mammalian cells has emerged as a powerful tool for several applications. One advantage of using this heterologous viral system is the virtual absence of endogenous viral gene expression and cytotoxicity in mammalian cells 32. In view of the results obtained above, we therefore investigated the use of baculovirus as a potential tool for gene delivery in human cell lines used to study insulin signalling. High efficiency baculovirus transduction of various liver 4445 and neural cell lines 46 has been established previously. However, the use of the human neuroblastoma SHSY-5Y cells has to our knowledge not been reported and transduction efficiency and expression of reporter protein in SHSY-5Y cells was therefore first confirmed. Similar to that for adenovirus (AdGFP), a viral construct containing GFP under the control of a cytomegalovirus promoter (CMV) was first constructed (BacGFP). Recombinant virus was propagated in Sf9 insect cells and prepared as described in the Methods section. Adherent cultures of SHSY-5Y cells were infected with BacGFP and expression of reporter protein was confirmed by fluorescence microscopy 24 hours p.i. (data not shown). For quantification and titration of transduction ratio, the cells were infected at MOI's between 50 and 10000 and for practical reasons the transduction efficiency (as % GFP expressing cells) and expression level (as relative units, R.U.) were analysed 48 hours p.i. using a chip-based flow cytometry application 53. Without further optimisation of the infection and expression conditions, the transduction ratio increased in a dose dependent manner with increased MOI to a maximum of 35 % (Table 2). However, when butyrate was added to the cells to enhance protein expression at a final concentration of 10 mM, the relative amount of GFP expressing-, and thus apparently transduced cells, was significantly increased from about 35 to 82 %. In addition, the maximum relative mean expression level increased more than 4-fold in the SHSY-5Y cells when butyrate was present (Table 2). Butyrate is a histone deacetylase (HDAC) inhibitor, inducing a hyperacetylation of chromatin and enhancement of transcription. The importance of the chromatin state of the baculovirus genome in the transduced cells for transgene expression and the role of HDAC inhibitors has been well described 3047 and will not be further discussed. Besides the more thorough investigation of SHSY-5Y cells, Chinese Hamster Overy cells (CHO.hIR) stably expressing the human insulin receptor 42, Human Embryonic Kidney cells (HEK293), and rat Fao- and human C3A hepatic-derived cell lines, were screened for the ability to be transduced by BacGFP (Table 3). This screen indicated that the CHO.hIR and the human HEK293 cell lines were easily transduced by BacGFP as previously been reported by others 30. However, under our conditions butyrate was required to confirm efficient transduction in the CHO.hIR cell line. Whether this was conditional or caused by the heterologous IR was not further investigated. The human hepatic-derived C3A cell line was also easily transduced by BacGFP and similarly to HEK293 and CHO.hIR, butyrate enhanced the expression of GFP and thus the apparent transduction ratio. By contrast, only minor GFP expression was detected in the rat hepatic-derived cell line Fao (Table 3) and was essentially unaffected by the addition of butyrate. Thus rat FAO hepatoma cells appear to be relatively resistant to BacMam viral transduction.

Next, the influence of baculovirus infection and recombinant protein expression on the phosphorylation of Akt was investigated. In these studies SHSY-5Y and C3A cells were infected with BacGFP and total and phosphorylated Akt were analysed using an Akt specific ELISA, as previously reported with AdGFP. Baculovirus titres of 1000 and 10000 were chosen to achieve experimentally desired transduction ratios. In contrast to cells infected with adenovirus at MOI 10 and 50, where PKB activation was observed (Table 1), cells transduced with BacMam at MOI 1000 and 10000 showed no apparent increase in PKB phosphorylation. This was true for both SHSY-5Y and C3A cells (Table 1; BacGFP). Taken together, the results from the experiments summarized in Table 1 show that when transduced with adenovirus at MOI 10 and 50, SHSY-5Y and C3A cells respond with an increased phosphorylation of Ser473 Akt (which is not accompanied by an increased expression of Akt) and that this can be avoided by utilizing baculovirus transduction, at MOI's as high as 10000.

Since infection of SHSY-5Y and C3A cells with baculovirus apparently did not influence the phosphorylation of Akt, the response to insulin following infection was investigated. First the response to insulin of uninfected cells was determined. SHSY-5Y and C3A cells were incubated with 100 nM insulin for 10 and 30 minutes, respectively and the phosphorylation of Akt was determined with phospho-specific ELISA (Figure 3; no virus). The response was most pronounced for SHSY-5Y cells with a transient 4.9-fold increase in Akt phosphorylation 10 minutes after addition of insulin which remained 3.1 – fold elevated after 30 minutes (Figure 3A and 3C, no virus). This data was again verified for SHSY-5Y cells with western blot using specific Ser⁴⁷³-phosho-Akt antibodies (Figure 2, lane 1 and 2). The C3A cells also responded to insulin albeit to a lower degree reaching a sustainable 1.6 – fold increase in Akt phosphorylation after 10 and 30 minutes incubation, respectively (Figure 3B and 3D, no virus). Next, the effect of insulin stimulation of cells infected with adenovirus and baculovirus, respectively, were investigated. As previously found for SHSY-5Y cells infected with adenovirus (Table 1; Figure 2, lane 1 and 3), infection with AdGFP at MOI10 and 50 in the absence of insulin gave a dose dependent increase in Akt phosphorylation (Figure 3A and 3C). In addition, when the cells were incubated in the presence of 100 nM insulin for 10 minutes no additional phosphorylation of Akt was obtained (Figure 3A; Figure 2, lane 3 and 4). After 30 minutes of incubation with 100 nM insulin, Akt was further activated 1.6 and 1.7-fold over the activation already induced by adenovirus infection (Figure 3C). By contrast, SHSY-5Y cells infected with baculovirus showed no sign of basal Akt phosphorylation after 10 or 30 minutes (Figure 3A and 3C). Moreover, upon insulin stimulation of baculovirus infected cells, the response was analogous to that for non infected cells incubated with insulin. At 10 minutes incubation, a 5.7 and 5.4-fold increased phosphorylation was obtained at MOI1000 and 10000 (Figure 3A), and at 30 minutes incubation a 3.6 and 3.8-fold increase in Akt phosphorylation was obtained at MOI1000 and 10000, respectively (Figure 3C). Qualitatively similar data were obtained for C3A cells (Figure 3B and 3D). Thus adenovirus infection increased basal Akt phosphorylation and insulin was not able to increase this further. By contrast, BacMam-transduced cells displayed no elevated basal Akt phosphorylation and insulin was able to induce a 1.6 to 2.1-fold activation of the enzyme, as for control cells. The apparently smaller Akt response in C3A cells compared to SHSY-5Y cells may be related to the higher basal level of Akt phosphorylation in unstimulated cells and is a feature of other hepatoma cells lines in culture including rat FAO and human HepG2 cells (data not shown). Interestingly, when we sought to examine the effects of virus infection on ERK1/2 activity in hepatoma cells, we found elevated basal ERK1/2 phosphorylation such that further increases could not be observed (data not shown). Thus our data show that cells transduced with baculovirus remain responsive to insulin stimulation as measured by phosphorylation of Akt.

All cell culture reagents were from Invitrogen. Hybond-C nitrocellulose membrane and enhanced chemiluminescence (ECL) were from Amersham Life Science. BCA protein analysis kit was from Pierce. Anti-phosphoserine and protein kinase B/Akt anti-bodies and Path Scan™ total Akt and Ser⁴⁷³ Phospho-Akt ELISA kits were from Cell Signalling Technology.

The pADTrack-CMV and pADEasy-1 plasmids, and E. coli strain BJ5183 used to construct recombinant adenovirus were obtained under license agreement from Dr Bert Vogelstein, John Hopkins Oncology center, Howard Hughes Medical Institute (Baltimore, USA) 54. To generate the vector pMB1039, homologous recombination of approximately 1 μg PmeI digested shuttle vector (pADTrack-CMV) and 0.1 μg supercoiled ad5 backbone vector (pAdEasy-1) was performed with 40 μl electrocompetent BJ5183 (endA sbcBC recBC galK met thi-1 bioT hsdR (Str^r)). The BJ5183 cells were grown to OD₅₅₀ 0.8, then collected and washed twice with ice-cold 10% glycerol. Electroporation was performed in 2.0 mm cuvettes at 2500 V, 200 ohms, and 25 μF in a Bio-Rad Gene Pulser electroporator. The mixture was then plated onto agar-kanamycin media. Only the smallest colonies were selected for screening by restriction analysis prepared from mini-preps. Linearized pMB1039 was subsequently transfected into adherent HEK293A cells in MEM containing Earle's salts and glutamine and supplemented with 10 % FCS. Virus was plaque purified and subsequently propagated in HEK293A cells in E-MEM supplemented with 2 % FCS. Amplified virus was purified using CsCl density gradient ultracentrifugation essentially as described previously 52. Isolated virus was dialysed twice in D-PBS/10 % glycerol and finally diluted in storage buffer containing 100 mM Tris-HCl (pH 8), 100 mM NaCl, 0.1 % BSA, and 50 % glycerol. Titer determination was performed by end point dilution on HEK293 cells and adenoviral stocks used for the experiments were typically 10¹⁰ pfu/mL.

The BacMam vector was designed essentially as described by Condreay and coworkers 30. To construct this vector, a fragment containing the CMV-IE promoter/enhancer, multiple cloning site and polyadenylation signal was isolated from the pcDNA3.1+ vector (Invitrogen) by PCR using primers 5'-CGCG

TCCGGA

AAGCTT

Recombinant BacGFP virus was generated using the Bac-to-Bac system (Invitrogen) essentially as described 30. Briefly, propagation of virus was performed in Sf-9 insect cells maintained as suspension cultures in SF900-II SFM (Invitrogen) at 27°C. For amplification to second generation virus used in the experiments, the Sf9 cells were infected at MOI 0.1 at a cell density of 2 × 10⁶ cells/mL. 5 days after infection, the culture medium containing amplified virus was harvested and titres were determined using the BacPAK baculovirus rapid titer kit (BD Biosciences, Clontech, Clontechniques XII(3): 8–9, 1997). In this titering method, a primary monoclonal antibody raised to an AcMNPV envelope glycoprotein (gp64) was used to label infected cells in replicate samples. A secondary HRP-conjugated antibody was used to stain the infected cells so the number of infected foci could be determined under light microscope. Viral titers between 10⁸–10⁹ pfu/mL were routinely obtained. Stocks of virus for experiments in mammalian cells were concentrated by centrifugation at 35000 × g for 60 minutes and pelleted virus was resuspended in Dulbecco's PBS supplemented with FBS (1 vol-%) to a final concentration of 10¹⁰ pfu/mL.

C3A cells were cultured in MEM supplemented with 10% heat inactivated FCS, 1% non-essential amino acids and 1% Na-pyruvate. SHSY-5Y cells were cultured in EMEM: Ham's F12 (1:1) supplemented with 15% heat inactivated FCS and 1% non-essential amino acids. Both cells lines were kept at 37°C in a humidified 5% CO₂ atmosphere. One day prior to experiment, cells were seeded out into 6 well plates (0.25 × 10⁶ cells/well) and cultured over night. Adeno- and Baculo-virus stocks were diluted accordingly and added to the wells. Medium was replaced with fresh serum-free medium after 24 h. Where indicated in Table 2 and 3 for the expression studies, 10 mM butyrate was added to the incubation medium. After additional 24 hour incubation under serum-free conditions, medium was replaced with fresh serum-free medium with or without (control) 100 nM insulin for 10 and 30 minutes, respectively. The plates were then placed on ice, and the cells were washed with ice cold PBS prior to harvest, using lysis buffer included in the ELISA kit (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM Na₂EDTA, 1 mM EGTA, 1% Triton, 2.5 mM Na-pyrophosphate, 1 mM β-glycerophosphate, 1 mM orthovanadate, 1 μg/ml leupeptin). The lysis buffer was supplemented with Complete™ protease inhibitor cocktail (Roche). As indicated in some experiments, cells were stimulated with 100 nM human recombinant insulin (Sigma) for 10 or 30 min prior to harvest. Lysates were then stored at -70°C prior to analysis. Total protein content was determined using Pierce BCA protein analysis kit.

For western blot analyses, samples were matched for protein concentration and heated at 70°C for 10 minutes with reducing sample buffer. Samples were resolved on 4–12 % gradient gels and transferred to nitrocellulose membranes. Membranes were probed for the presence of total and phosphorylated Akt using antibodies obtained from Cell Signaling Technology, and developed using enhanced chemoluminesence. The signal was then quantified by densitometric analysis. Alternatively, total and Ser473 phosphorylated Akt was determined with a commercially available ELISA system for PKB/Akt from Cell signalling Technology.

The flow cytometry data presented in Figure 1 and Table 3 was processed using an EPICS XL-MCL flow cytometer (Beckman-Coulter) essentially as described previously 42. For the experiments in Table 2, protein expression was measured with an Agilent 2100 Bioanalyzer using the chip-based flow cytometry application 53. Briefly, cultures were washed with PBS, harvested with cell dissociation buffer (500 μL for 1 min) and re-suspended in 500 μL OptiMem (Invitrogen) at room temperature. Staining for viable cells was performed by adding carboxynaphtofluorescein (CBNF, Molecular Probes, Inc., Leiden, NL) at a final concentration of 0.5 μM. After 15 minutes, cells were centrifuged at 500 × g for 5 minutes, followed by an additional washing and centrifuge step and finally the pellet was re-suspended in 400 μL Cell buffer (Agilent Technologies). 10 μL cell samples were applied to the chip and analyzed utilizing the GFP cytometry mode. Data was obtained by gating for viable cells and adjusting for background fluorescence. Values were determined for viable GFP expressing cells.