The mouse HerC6 RLD confers efficient ISG15 conjugation to human HerC6
In order to assess what moiety of the mouse HerC6 protein confers ISG15 E3 ligase activity in comparison to human HerC6, which doesn't possess ISG15 E3 activity, we constructed plasmids expressing chimeric human/mouse HerC6 proteins. Either the N-terminal RLD or the C-terminal HECT domain was exchanged with the equivalent of the alternate species (). Analysis of the chimeric proteins by Western blot demonstrated that the chimeras express to similar levels as their wild-type equivalents ().
The mouse HerC6 RLD confers efficient ISG15 conjugation.
Next, these constructs were tested for their ability to conjugate ISG15 to cellular targets in the presence of over-expressed E1 and E2 enzymes. As previously described, both hHerC5 and mHerC6 effectively conjugated both human and mouse ISG15, whereas no ISG15 conjugates were detected in cells overexpressing hHerC6 (, lanes hH6, hH5 and mH6). Replacement of the human HerC6 N-terminal RLD by its mouse equivalent almost completely restored ISGylation (, compare lanes mH6 and MNHC), suggesting that mainly differences in the RLD between hHerC6 and mHerC6 allow mHerC6 to conjugate ISG15. However, exchange of the mouse HerC6 C-terminus partially restored ISGylation by hHerC6 (, compare lanes hH6 and HNMC), indicating that the ability of mHerC6 to function as an E3 ligase compared to hHerC6 also partially relies on differences in the protein's C-terminal half that contains the HECT domain. The same constructs were analyzed in the presence of the mouse E1, E2 and ISG15. Despite the fact that mouse ISG15 was much more efficiently conjugated in the presence of its native conjugation machinery, similar results were obtained with the HerC6 chimeras (data not shown).
The results described in suggested that mainly the mHerC6 N-terminal half is required for efficient ISG15 conjugation, irrespective of the species origin of ISG15. Conjugation of ISG15 chimeric proteins harboring either the N- or C-terminal Ub-like domain of mouse or human ISG15 confirmed and strengthened this observation. The mHerC6 N-terminal half almost completely restored E3 activity in hHerC6 for both chimeric ISG15 proteins (, compare lanes mH6 and MNHC), although the mHerC6 C-terminal half also partially restored ISG15 conjugation (, compare lanes hH6 and HNMC). Moreover, the mISG15 and HNMC ISG15 molecules were consistently conjugated with higher efficiency than hISG15 or the MNHC ISG15 chimera. Together with the observation that conjugation of endogenous ISG15 upon IFN treatment is consistently more robust in mouse cells than in human cells (data not shown), these latter data seem to indicate that mISG15 may be more efficiently conjugated than hISG15 and that this property seems to reside in its C-terminal Ub-like domain (., compare panel c and d).
In conclusion, these data demonstrate that mouse HerC6 has ISG15 E3 ligase activity and that this activity mainly depends on features in the N-terminal half of the protein compared to human HerC6. Next we investigated if expression of endogenous mHerC6 is also stimulated by type I IFN and localizes similar to human HerC5 and HerC6.
mHerC6 is induced by type I interferon and localizes exclusively in the cytoplasm
Although human HerC5 and HerC6 differ in their ability to conjugate ISG15, they share their inducibility by type I IFN and their localization in the cytoplasm 
. We investigated if mouse HerC6 is also IFN-inducible and localizes to a sub-cellular site similar to its human relatives. To this end, HerC5 and hHerC6 mRNA upregulation was measured in human Hep2 and HeLa cells stimulated with IFNβ and compared to mHerC6 mRNA regulation in several IFN-stimulated mouse cell lines ().
mHerC6 is induced by type I interferon and localizes exclusively in the cytoplasm.
IFN stimulation upregulated human HerC5 and HerC6 mRNA in both human cell lines, albeit to a lesser extent in Hep2 than in HeLa cells (). HerC5 and HerC6 were both upregulated 9-fold in Hep2 cells, whereas HerC5 mRNA increased slightly more (20-fold) compared to HerC6 mRNA (14-fold) in HeLa cells (, left panel). The same amount of IFN induced an average 12-fold upregulation of mHerC6 mRNA levels in the six mouse cell lines tested (, right panel), varying from 7-fold in NIH-3T3 cells to 25-fold in the macrophage cell line RAW 264.7.
Subsequently, localization of HerC proteins was examined by immuno-fluorescence assays of HeLa cells transfected with HerC5/6 expression plasmids. Since all tested HerC5/6 proteins are IFN-inducible, we investigated sub-cellular localization both under non-induced conditions, as well as during infection with SeV, a potent inducer of the type I IFN system 
Infection with SeV translocated IFN-regulatory factor 3 (IRF-3), a transcription factor critical for efficient IFNβ induction 
, to the nucleus (). These results indicated that the assay conditions were correct for detecting sub-cellular changes of molecules important for the IFN-response. In agreement with previously published observations, human HerC5 and HerC6 localized exclusively to the cytoplasm in mock-induced cells (). Despite the nuclear translocation of IRF-3, which is considered a hallmark of productive IFN induction and hence SeV infection, human HerC5 and HerC6 localization was unchanged during SeV infection (). Similarly, mouse HerC6 localized in the cytoplasm and its localization remained unchanged during SeV infection (). Together these data show that mouse HerC6 IFN-inducibility and localization in mouse cells are very similar to that of hHerC5/6 in human cells.
mHerC6 is essential for global ISG15 conjugation in mouse cells
Since mouse HerC6 was able to facilitate ISG15 conjugation in an overexpression system, we next set out to determine if HerC6 is the main E3 ligase mediating global ISGylation in mouse cells. To that end mouse L-929 cells were transfected with siRNAs specifically targeting mHerC6. As a control siRNAs targeting hHerC6 (but not mHerC6) were used.
Firstly, RNA was extracted from siRNA transfected cells and specific knock-down was determined by real-time RT-PCR with species-specific HerC6 primers. We compared IFN treated samples to mock samples to assure that IFN induction would not abolish knock-down. Efficiency of knock-down was determined by normalizing the observed knock-down to the human specific siRNA controls. Four different commercial siRNAs targeting mHerC6 were tested. Efficient and specific knock-down was observed for all of them, whereas controls had no effect on mHerC6 mRNA levels (data not shown). The assay was repeated with the two most efficient siRNAs (mHerC6a and –b) which specifically reduced mHerC6 mRNA in mouse cells by >85% compared to the control ().
mHerC6 is essential for global cellular ISG15 conjugation in mouse cells.
Finally, global ISGylation was analyzed in the context of mHerC6 knockdown. We hypothesized that if mHerC6 were the functional equivalent of human HerC5, knock-down of mHerC6 in L-929 would abolish global ISGylation. We transfected siRNAs targeting either mHerC6, hHerC6, mouse ISG15 or a scrambled control in L-929 cells and analyzed ISGylation 48 h after IFN stimulation. ISG15 expression was only upregulated in IFN-treated cells. Moreover, ISGylation was detected when a non-targeting control siRNA was transfected, whereas ISGylation was specifically abolished by knockdown of mHerC6, indicating that mHerC6 is essential for global ISGylation in mouse cells. Two different mHerC6-specific siRNAs demonstrated the same effect (), thus making off-target effects by the siRNAs unlikely. The increase of unconjugated ISG15 upon knockdown of mHerC6 is likely due to accumulation of ISG15 that would otherwise be conjugated.
To validate these results in a transfection system in which ISG15 expression is controlled, we subsequently co-transfected a V5-tagged ISG15 expression plasmid together with a mHerC6-specific or control siRNA and stimulated the cells with IFN. ISG15 was expressed from a plasmid and thus at a constant level in all samples (). However, global ISGylation was only detected upon IFN treatment (), indicating that upregulation of the endogenous E1, E2 and E3 enzymes was IFN-dependent as expected. Moreover, transfection of four different siRNAs targeting mouse HerC6 abolished global ISGylation in comparison with the control siRNA (, IFN treated lanes and Fig. S1
). Taken together these data confirm that HerC6 is the main ISG15 E3-ligase in mouse cells and in that respect functionally parallels human HerC5, even though genetically it is more closely related to hHerC6. The use of four different mHerC6-specific siRNAs strongly demonstrated that the loss of global ISGylation upon mHerC6 knock-down is specific and not resulting from off-target effects.
mHerC6 enhances IFNβ promoter induction and confers antiviral activity in mouse cells
Previous studies have demonstrated that hHerC5 positively regulates the IFNβ promoter and confers antiviral activity 
. To investigate if the HerC5/6 proteins could enhance an innate immune signal initiated by the RNA sensor RIG-I, HEK-293T cells were transfected with an IFNβ reporter construct controlling firefly luciferase, a constitutively active renilla luciferase internal control plasmid, a limiting amount of a plasmid expressing a constitutively active form of RIG-I (RIG-I(2CARD)) and the indicated HerC5/6 proteins. TRIM25, a well-established activator of RIG-I, was used as a positive control 
. As previously reported, hHerC5 enhanced IFNβ promoter activity (by 8.7 fold; ) over the unspecific control protein GST, whereas its hHerC6 relative without ISG15 E3 ligase activity did not further enhance promoter activity. In contrast, mHerC6 also further enhanced IFNβ activity by 8.5 fold, similar to its hHerC5 counterpart ().
mHerC6 stimulates the IFNβ promoter and confers antiviral activity against VSV and NDV.
To address the antiviral properties of mHerC6 in mouse cells, we determined if HerC5/6 expression would confer antiviral resistance to VSV-GFP and NDV-GFP infection. We transfected L-929 cells with the indicated constructs and subsequently infected the cells with either vesicular stomatitis virus (VSV) or Newcastle disease virus (NDV) expressing GFP, and analyzed the produced virus secreted into the supernatant by TCID50 assay. In correlation with what has been previously reported, hHerC5 significantly reduced both VSV and NDV virus production by 100-fold, whereas its hHerC6 relative without E3 activity did not (). Moreover, mHerC6 also reduced the titers of both viruses, highly similar to its human counterpart, hHerC5.
Taken together, these results indicate that mHerC6 parallels hHerC5 in its ability to enhance the IFNβ promoter and confer antiviral activity. In contrast, the hHerC6 protein did not have these activities, which correlates with its lack of E3 ligase activity. Moreover, the fact that the hHerC5 and mHerC6 proteins had activity in both human and mouse cells, is in line with our previous observation that hHerC5 and mHerC6 can mediate global ISGylation using ISG15 and the conjugation machinery of the opposite species.