H1299 cells (ATCC no. CRL-5803; American Type Culture Collection; Manassas, VA, USA) and baby rat kidney (BRK) cells (Nevels et al., 1997; Speiseder et al., 2014) were kept in incubators at 37°C in a 5% CO₂ atmosphere and maintained in Dulbecco’s modified Eagle medium (DMEM; Gibco; Carlsbad, CA, USA) with 10% fetal bovine serum (PAN Biotech; Aidenbach, Germany) and antibiotics (100 U penicillin/100 µg streptomycin per ml, PAN Biotech). All cell lines were regularly monitored for mycoplasma contamination.

All recombinant plasmids were generated by site-directed mutagenesis using the vectors pcDNA3 and pCI-FLAG. SENP-encoding plasmids (kindly provided by Dr. Stefan Müller), plasmids encoding 6x His SUMO 3 (Tatham et al., 2009), SUMO 3 Q90P (this work, forward primer 5’-GAT GTG TTC CAA CAG CCG ACG GGA GGT TAG-3’, reverse primer 5’-CTA ACC TCC CGT CGG CTG TTG GAA CAC ATC-3’), HAdV-C5 E1B-55K (Nevels et al., 2001), the E1B-55K K101R and K104R mutants (Kolbe et al., 2022) and the HAdV-C5 E4orf6 (Querido et al., 2001) were used for transient polyethylenimine (PEI; Polysciences; Warrington, PA, USA) transfections (Schreiner et al., 2010).

Transfected cell pellets were lysed on ice using radioimmunoprecipitation assay (RIPA) buffer (50 mM Tris/HCl pH 8, 150 mM NaCl, 5 mM EDTA, 1% P-40, 0.1% SDS, and 0.5% sodium deoxycholate). The lysates were then sonicated and centrifuged to remove cell debris. To investigate protein-protein interactions, one part of the resulting supernatant was immunoprecipitated for 2 h at 4°C followed by centrifugation at 600 × g for 5 min at 4°C exactly as described previously (Berscheminski et al., 2013). The remaining supernatant was boiled at 95°C for 5 minutes in 5% Laemmli buffer (input). Samples were stored at -20°C until further analyses by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS PAGE) and immunoblotting.

For nickel-nitrilotriacetic acid (Ni-NTA) SUMO pulldown analyses, cells were harvested, washed with pre-cooled PBS, and lysed in guanidine hydrochloride (GuHCl) buffer (6 M Guanidinium-HCl, 10 mM Tris and 100 mM sodium phosphate buffer pH 8.0). His-SUMO modified proteins were coupled to the Ni-NTA agarose beads (Thermo Scientific) by incubation overnight at 4°C. Next, His-SUMO conjugates coupled to the Ni-NTA agarose were pelleted by centrifugation, washed, and His-SUMO conjugated proteins were eluted from the beads and stored at -20°C until further use (Tatham et al., 2009).

For immunoblotting, SDS PAGE-separated proteins were transferred to nitrocellulose membranes with a pore size of 45 μm (GE Healthcare; Chicago, IL, USA) via wet electroblotting using the TransBlot Electrophoretic Transfer Cell System (BioRad; Hercules, CA, USA). Next, membranes were incubated in 5% non-fat dry milk-PBS solution for 1 h at 4°C to saturate non-specific antibody binding sites. The membranes were then washed 3 times with PBS-tween and incubated with the respective primary antibody at 4°C ( Table 1 ). After 3 hours of incubation, membranes were washed and incubated with the respective horseradish peroxidase (HRP)-conjugated secondary antibody for 2 h at 4°C. Membranes were washed again and proteins were visualized on X-ray films using the SuperSignal West Pico Chemiluminescent Substrate (Thermo Scientific; Waltham, MA, USA) according to the manufacturer’s instructions. X-ray films were developed using the GBX Developer (Kodak; Rochester, NY, USA) and digitalized.

BRK cells were seeded in 6-well plates and transfected with the respective plasmids with calcium phosphate following the manufacturer’s protocol (ProFection Mammalian Transfection System, Promega; Madison, WI, USA). Transfected cells were cultivated for 4–8 weeks with weekly media changes until multilayered cell accumulations (foci) were visible. Foci were visualized by crystal violet staining and quantified in the same way as described previously (Speiseder et al., 2014). These experiments were repeated as four independent replicates.

Statistical analyses were performed using Graph-Pad Prism v9 (GraphPad Software, Inc.; La Jolla, CA, USA). A one-way analysis of variance with Dunnett’s multiple comparisons test was used to analyze focus formation in BRK cells and data were considered significant if p values were ≤ 0.05. Data are presented as means and error bars indicate standard deviations (SDs).

To test if E1B-55K interacts with SENPs, we performed co-immunoprecipitation (co-IP) assays using plasmids expressing E1B-55K and catalytically inactive (CAT) FLAG-tagged SENP isoforms from all three SENP groups (SENPs 1, 2, 3, and 6). These catalytically inactive SENPs were used to elongate interactions between SENPs and E1B-55K as deSUMOylation is a quick process and consequently, interactions between SENPs and their targets are rather unstable (Kunz et al., 2018). E1B-55K precipitation and subsequent FLAG-staining revealed an interaction between SENP 1-CAT and E1B-55K, indicated by a SENP 1-CAT-specific band at 70 kDa ( Figure 1A ). Moreover, a rather faint band in SENP 3-CAT/E1B-55K co-transfections indicates their interaction. These results suggest that E1B-55K could be targeted for deSUMOylation by SENP 1 and 3, although SENP 1 likely is the main SENP that deSUMOylates E1B-55K, given the much stronger staining ( Figure 1A ). In fact, SENP 6-CAT expression was very weak, and steady-state concentrations were not detectable ( Figure 1A ). On the other hand, SENP 1-CAT steady-state concentrations were also rather low, while the co-IP showed a solid enrichment of the protein ( Figure 1A ). These observations suggest that an interaction between E1B-55K and SENP6 is unlikely, but we have to note that it is difficult to make a definitive statement regarding the binding of E1B-55K to SENP 6. Collectively, however, our investigations on the interactions of E1B-55K with SENPs identified E1B-55K as a specific interaction partner of SENPs 1 and 3 ( Figure 1A ).

To assess the deSUMOylation of E1B-55K by SENP 1, we performed Ni-NTA pulldown analyses using H1299 cells that overexpress SUMO. We transfected these cells with different combinations of plasmids that express E1B-55K, SENP 1, and a catalytically inactive SENP 1 (SENP 1-CAT) ( Figure 1B ). Here, the His-SUMO pulldowns showed that SENP 1 reduced E1B-55K SUMO modification. At the same time, co-expression of the catalytically inactive mutant SENP 1-CAT did not affect SUMO modification of E1B-55K ( Figure 1B ).

To investigate the effect of SENP 1 on E1A/E1B-dependent transformation of rodent cells in vitro, we transfected BRK cells with E1A and E1B-expressing plasmids alone or in combination with plasmids expressing active and inactive SENP 1. As controls for E1B-55K SUMOylation, either K104R (negative, mutation abrogates E1B-55K SUMOylation) or K101R (positive, mutation increases E1B-55K SUMOylation) (Kolbe et al., 2022) were transfected in combination with E1A. Our transformation experiments confirmed that E1A and E1B expression result in efficient transformation of BRK cells ( Figures 2A, B ). E1A and K101R co-transfections induced a remarkably increased transformation rate, while E1B-55K almost lost its transforming capabilities upon introduction of the K104R mutation ( Figure 2 ). Co-transfections of E1A and E1B-expressing plasmids with SENP 1 revealed that expression of SENP 1 causes a significant decrease in the foci numbers, indicating that SENP 1 inhibits E1A- and E1B-dependent transformation of BRK cells through efficient deSUMOylation of E1B-55K ( Figure 2 ). Concomitantly, co-transfections of E1A and E1B with the catalytically inactive SENP 1 (SENP 1-CAT) restored focus formation to wild-type (wt) levels ( Figure 2 ). In sum, these data confirm that deSUMOylation of E1B-55K is mediated by SENP 1, leading to strongly reduced proliferation-promoting features of this adenoviral multifactorial oncoprotein.

To determine if E4orf6 influences the interaction of E1B-55K with SENPs, both proteins were transfected into H1299 cells in combination with the SENP CAT isoforms 1 to 3 and 6, followed by immunoprecipitation assays ( Figure 3A ). These experiments showed that the binding of E1B-55K to SENPs was not abrogated by E4orf6 because E1B-55K still efficiently bound to SENP 1-CAT and rather weakly to SENP 3-CAT both in the presence ( Figure 3A ) or absence of E4orf6 ( Figure 1 ). Further, our results suggest that E4orf6 does not recruit additional SENPs for E1B-55K deSUMOylation. However, it cannot be excluded that E4orf6 deSUMOylates E1B-55K independently of SENP 1. To test this, we used a SUMO mutant that harbors a glutamine (Q) to proline (P) mutation at its SENP-cleavage site at position 90 (Q90P) leading to structural changes within SUMO that prevent deSUMOylation by SENPs ( Figure 3B ) (Owerbach et al., 2005; Bekes et al., 2011). Upon transfection with wt E1B-55K and the hyper-SUMOylation mutant K101R without and with E4orf6, and after subsequent Ni-NTA pulldowns, we found that co-transfection of E4orf6 resulted in reduced SUMO levels of E1B-55K and K101R ( Figure 3B ). Similar to wt SUMO, Q90P was conjugated to E1B-55K and K101R ( Figure 3B ), and Q90P-conjugation to E1B-55K and K101R were also decreased in presence of E4orf6 ( Figure 3B ). Thus, we confirmed that SENP 1 reduces SUMO on E1B-55K and demonstrate that this process can be, in part, compensated by E4orf6.