Establishment of an efficient heterologous expression system for 10.4-14.5K suitable

It remained to be tested whether the observed loss of binding of the mutated proteins to the adaptor proteins in vitro was reflected by an altered trafficking of 10.4K and 14.5K mutant proteins in vivo. To analyze the intracellular distribution of the viral proteins in greater detail, 10.4 and 14.5 had to be expressed in cells more suitable for immunofluorescence than 293 cells. Therefore, plasmid pBS∆X-E3/FLAG-14.5 was transiently transfected into SV80Fas cells, a human fibroblast cell line expressing SV40 T antigen and Fas. At 40h post transfection the intracellular localization of E3 proteins was analyzed. Upon CaPO4-mediated transfection of SV80Fas cells only 25% of the cells stained positive for E3/19K in the endoplasmic reticulum. Moreover, 14.5K could only be detected in one third of these E3 positive cells (Fig. 15). This is consistent with the generally lower abundance of the 10.4-14.5K encoding mRNA species of subgroup C Ads as compared with those encoding E3/19K (Wold et al., 1995). 14.5K localized to a perinuclear compartment, which was stained specifically by rabbit serum anti-14.5K (Rα14.5) and monoclonal anti-FLAG antibody M1 (data not shown). But, the number of positive cells as well as the staining intensity was insufficient for a detailed analysis of the intracellular distribution of 10.4-14.5K proteins by immunofluorescence.

Therefore, an efficient 10.4-14.5 expression system independent of AdE1A products, which upregulate E3 expression in 293 cells had to be established. To this end, Ad2 10.4K and FLAG- 14.5K encoding ORFs were cloned separately into pSG5 expression vectors to drive 10.4K and 14.5K synthesis by the SV40 promoter/enhancer (Fig. 16). A potentially important feature of this vector is the intron II of the rabbit β-globin gene for splicing of the expressed transcript. Inclusion of the intron might help to enhance expression by increasing mRNA half-life and improving the efficiency of RNA processing and transport to the cytoplasm (Kim et al., 2002). As both proteins are known to function as a complex it was also aimed at expressing both proteins from a single vector. The entire Ad2 10.4-14.5K encoding sequence was cloned into the pSG5 vector, and in parallel individual ORFs were introduced into the multigenic expression vector pMG (Fig. 16). This vector provides two different multiple cloning sites, each with a strong promoter for eucaryotic expression, a viral and a housekeeping promoter, thus limiting transcription interference. The 10.4K open reading frame was inserted into the first MCS under the control of the immediate-early HCMV enhancer promoter (HCMV-IA prom) and located downstream of

Fig. 15 Costaining of Ad2 14.5K and E3/19K in SV80Fas cells transiently transfected with plasmid pBS∆∆∆∆X-E3/FLAG-14.5

SV80Fas cells were transiently transfected with plasmid pBS∆X-E3/FLAG-14.5 (calciumphosphate method) for immunofluorescence analysis at 40 hours post-transfection. 14.5K was detected with polyclonal serum Rα14.5 and 19K was stained using mAb Tw1.3 as described in Materials and Methods.

intron A. The 14.5K ORF was cloned into the second MCS with expression driven by the Elongation factor 1 alpha (hEF1) promoter in combination with the Human T-Cell leukemia virus (HTLV) 1 Long terminal Repeat for stabilization of the mRNA followed by an intron sequence (intron 117). The introns preceeding the inserted ORFs are spliced out in mammalian cells. Insert amplification primers for the cloning of individual ORFs were designed to yield a modified sequence 5’ of the ATG that conforms to the Kozak consensus of translation inititation (Kozak, 1987), for optimized heterologous expression. Expression of the proteins was analyzed upon transient transfection of SV80Fas or A549 cells with vector DNA followed by immunofluorescence detection at 40 hours post-transfection. For improved transfection efficiency cells were seeded on the day of transfection and the DNA/CaPO4 mix was added to the cells 4-6

Fig. 16 Circular map of pSG5 (Stratagene, Amsterdam, The Netherlands) and pMG (InvivoGen, San Diego, USA) used for construction of 10.4K and 14.5K expression vectors

Coding sequences of 10.4K, 14.5K or both were inserted into the MCS of pSG5, as described in Materials and Methods. Additionally, the 10.4 ORF was cloned into the BamHI, XbaI sites of pMG and the FLAG- 14.5K CDS into the ClaI, NheI sites of construct pMG10.4. For a description of vector features see text.

Fig. 17 Intracellular localization of 10.4K in SV80Fas cells and FLAG-14.5K in SV80Fas or A549 cells following transient transfection of single expression vectors

SV80Fas cells were transfected with pSG5/F14.5K (A, C, D) or pSG5/10.4K (B) and processed for confocal laser microscopy at 40 h post transfection using polyclonal Ab Bur3 or Rα14.5 to detect 10.4 and 14.5K, respectively (green). Localization of 14.5K was compared with the Golgi marker galactosyltransferase, GLT (red, mAb GTL-2) (C), and ER-resident protein Calnexin (red, mAb AF-8) (D). (E) In A549 cells transfected with pSG5/F14.5K a similar distribution of FLAG-14.5K was detected using Rα14.5 or mAb M1 anti-FLAG (40 h post transfection).

hours after plating, at which time the cells had adhered to the plastic dishes but not yet extended(Marks et al., 1996).

Despite the presence of strong promoters transfection of the bicistronic vector pMG yielded only a low percentage of positive SV80Fas cells. 10.4K was barely detectable whereas about 20% of the cells showed a specific staining for FLAG-14.5K in a perinuclear compartment (data not shown). Even a smaller number of positive cells were observed upon transfection of the pMG

Fig.18 Intracellular localization of 10.4K and F14.5K expressed in A549 (A, C) or SV80Fas (B,D) cells, upon transfection of pSG5/10.4-F14.5

At 40 hours post-transfection with pSG5/10.4-F14.5 vector DNA cells were processed for immunofluorescence analysis. Transfected A549 cells (A) or SV80Fas cells (B) were costained for 14.5K with Rα14.5K (green) and mAb M1 against FLAG (red). For detection of 10.4K a single stain with Bur3 was performed on transfected A549 (C) or SV80Fas cells (D).

vector into A549 cells (~10%). By contrast, a satisfactory expression level suitable for immunofluorescence analysis of intracellular protein localization could be achieved upon transfection of pSG5 vectors. Up to 60% of the cells exhibited a strong staining for 10.4K and 14.5K in SV80Fas cells transfected with pSG5/10.4K or pSG5/F14.5K (Fig. 17A, 17B). A similar pattern was observed for 14.5K expression in A549 cells. 14.5K specific rabbit serum and monoclonal antibody M1 directed against the FLAG-tag revealed an identical staining pattern (Fig. 17E). The 14.5K positive perinuclear structure corresponded to the Golgi/trans-Golgi network as it costained with galactosyltransferase (Fig. 17C, red) and human TGN46 (data not shown), which are cellular marker proteins for this compartment. Furthermore, 14.5K colocalized with Calnexin in the endoplasmic reticulum (Fig. 17D). In cells transfected with pSG5/10.4K 10.4K could be specifically stained using polyclonal rabbit serum Bur3 (Fig. 17B), or antiserum R59 directed against the entire cytoplasmic tail of 10.4K (data not shown). In pSG5/10.4K positive SV80Fas cells 10.4K localized to the ER and Golgi/TGN.

In cells transfected with the pSG5/10.4-F14.5 vector 14.5K was detected in a perinuclear compartment identified to correspond to the Golgi/TGN, which also stained positive for 10.4K. Additionally, 14.5K localized to a few dot-like structures surrounding the perinuclear compartment (Fig. 18A, 18B, 14.5K). 10.4K (Fig. 18C, 18D, 10.4K) localized to the endoplasmic reticulum and the Golgi/TGN, as observed in cells transfected with pSG5/10.4K (Fig. 17B). Interestingly, 14.5K specific staining of the endoplasmic reticulum was significantly reduced as

compared to cells transfected with pSG5/F14.5K (compare Fig.17A, 17E and Fig. 18A, 18B, 14.5K). Thus, coexpression of 10.4K in the transfected cells caused a shift in 14.5K steady-state localization, as evidenced by a reduced ER staining. This correlates with biochemical evidence presented for the Ad5 14.5K protein. Ad5 14.5K can form a complex with Ad2 10.4K and it has been reported that the efficiency and site of cleavage of the Ad5 14.5K signal sequence depends on the presence of 10.4K (Krajcsi et al., 1992b). Moreover, the extent of glycosylation and phosphorylation of Ad5 14.5K depends on expression of 10.4K (Krajcsi et al., 1992c; Krajcsi and Wold, 1992). Therefore, the 10.4K-induced loss of 14.5K ER staining might be the consequence of association of 10.4K with 14.5K and efficient signal sequence cleavage in the 14.5K protein, which is then no longer retained in the ER.

In SV80Fas cells the number of cells bearing 14.5K at the cell surface was higher than in A549 cells, whereas in about 5% of the transfected A549 cells 14.5K was also detected in the ER. The increase in 14.5K signal intensity at the cell surface of SV80Fas cells suggested enhanced expression of 10.4-14.5K in these cells. Given that transfection efficiency was very similar in both cell types increased expression in SV80Fas cells might be due to SV40 T antigen-mediated amplification.

In sum, upon transient transfection of pSG5 expression constructs into SV80Fas cells a high number of cells could be efficiently transfected yielding expression levels that were suitable for immunofluorescence analysis of 10.4K and 14.5K intracellular localization. Therefore, for most studies the pSG5 expression constructs were introduced into SV80Fas cells.

4.2. Functional activity of 10.4 and 14.5K proteins encoded by single expression vectors