The COP9 Signalosome Variant CSN CSN7A Stabilizes the Deubiquitylating Enzyme CYLD Impeding Hepatic Steatosis

: Hepatic steatosis is a consequence of distorted lipid storage and plays a vital role in the pathogenesis of nonalcoholic fatty liver disease (NAFLD). This study aimed to explore the role of the COP9 signalosome (CSN) in the development of hepatic steatosis and its interplay with the deubiquitylating enzyme (DUB) cylindromatosis (CYLD). CSN occurs as CSN CSN7A and CSN CSN7B variants regulating the ubiquitin proteasome system. It is a deneddylating complex and associates with other DUBs. CYLD cleaves Lys63-ubiquitin chains, regulating a signal cascade that mitigates hepatic steatosis. CSN subunits CSN1 and CSN7B, as well as CYLD, were downregulated with speciﬁc siRNA in HepG2 cells and human primary hepatocytes. The same cells were transfected with Flag-CSN7A or Flag-CSN7B for pulldowns. Hepatic steatosis in cell culture was induced by palmitic acid (PA). Downregulation of CSN subunits led to reduced PPAR-γ expression. Flag-pulldowns in both LiSa-2 and HepG2 cells and human primary hepatocytes revealed binding of CYLD preferentially to CSN CSN7A . This was inﬂuenced by PA treatment. Silencing of CSN CSN7B blocked lipid droplet formation caused a compensatory increase of CSN CSN7A stabilizing CYLD. Our results demonstrate that CSN CSN7A -mediated CYLD stabilization impedes hepatic steatosis. Therefore, stabilizing CSN CSN7A -CYLD interaction might be a strategy to retard hepatic steatosis.


Introduction
Nonalcoholic fatty liver disease (NAFLD) has a worldwide prevalence of about 25% [1,2].Many individuals with NAFLD progress from simple hepatic steatosis to nonalcoholic steatohepatitis (NASH) [3], which is characterized by severe lipid accumulation, inflammation, and fibrosis.NASH, accompanied by chronic inflammation, is regarded as a main risk factor for both cirrhosis and hepatocellular carcinoma (HCC) [4].Due to the incomplete understanding of its pathogenesis, there is still no approved pharmacotherapy for NASH [1,5].
Hepatic steatosis results from an imbalance between lipid acquisition and lipid removal, which is regulated by four primary mechanisms: free fatty acid uptake, de novo lipogenesis, fatty acid oxidation, and export from hepatocytes as very-low-density lipoproteins.Disturbance in one or more of these pathways may result in the retention of fat in the liver, leading to hepatic steatosis and, finally, to NASH [6].Primarily, fatty acids are stored as triacylglycerides (TAGs) in lipid droplets (LDs), which are distorted under disease conditions [7][8][9].Hepatic steatosis can be caused by diet, gut microbiota, genetic factors, and de novo lipogenesis via the up-regulation of lipogenic transcription factors such as Livers 2021, 1 117 sterol regulatory binding protein-1c (SREBP1c), carbohydrate-responsive element-binding proteins (ChREBP), and peroxisome proliferator-activated receptor gamma (PPAR-γ) [8,10].
Although many open questions remain, it is known that the ubiquitin (Ub) proteasome system (UPS) plays a key role in the pathogenesis of NASH [11].The ubiquitylation and deubiquitylation of proteins in hepatocytes has been shown to be involved in lipid metabolism [12].In the UPS, a cascade of enzymes activates and transfers Ub to substrates as a signal for degradation.Poly-ubiquitylated proteins are subsequently degraded by the 26S proteasome [13].Substrate specificity of the system is conferred by Ub ligases (E3s).A major representative subfamily of E3s are cullin-RING-ubiquitin ligases (CRLs) [14].CRLs are multiprotein complexes organized around scaffolding cullins (Cullin 1-9, CUL1-9), containing a RING protein (e.g., RBX1 or RBX2) and associated with hundreds of exchangeable substrate-specific receptor (SR) proteins [14].Based on their cognate cullin subunits, CRLs are divided into sub-families (CRL1 to CRL9) [15].CRLs are activated by neddylation of their cullins [16,17].The COP9 signalosome (CSN) is a master regulator of CRLs removing NEDD8 from cullins.Deneddylation inactivates CRLs and allows an exchange of SRs and, thus, adaptation to fluctuations in substrate availability [18][19][20][21][22][23][24][25].The mammalian CSN consists of eight core subunits, CSN1 to CSN8, characterized either by the Proteasome lid-CSN-Initiation factor 3 (PCI) domain (CSN1-CSN4, CSN7, and CSN8) or by the MPR1/PAD1 N-terminal (MPN) domain (CSN5 and CSN6) [26].All CSN subunits are required for the hydrolytic removal of NEDD8 [27,28], catalyzed by the JAMM/MPN+ domain localized to CSN5 [29].Evolutionary diversification led to the occurrence of distinct CSN variants formed by the integration of different subunit paralogs/isoforms.Mammalian CSN variants containing CSN7A or CSN7B paralogs, CSN CSN7A and CSN CSN7B , co-exist in all mammalian cells possessing distinct functions in adipogenesis [30].Recently, we discovered a variant-specific function of CSN CSN7B in DNA repair [31].Because of its deneddylating activity, the CSN belongs to deubiquitylating enzyme (DUB) complexes [32] and is a member of the JAMM subfamily of DUBs.In addition, it is associated with a number of DUBs including USP15 and USP48 forming a multi-DUB complex [32].
Interestingly, the DUB cylindromatosis (CYLD) is involved in the regulation of different cellular processes including inflammation, fibrosis, and cancer and reverses the progression of NASH [33].CYLD specifically affects the disassembly of Lys63-linked ubiquitin polymers [34].It degrades Lys63-chains linked to components of different signaling pathways including NF-κB, Notch, and JNK [35].It directly interacts with the kinase TAK1 and removes its Lys63-linked poly-ubiquitin chains, which blocks downstream activation of the JNK-p38 pathway, thereby retarding NASH [33].CYLD contributes to hepatic homeostasis and restoration upon liver injury [36].It is downregulated during NASH or NAFLD.An increase of CYLD mitigates NASH [33].Interestingly, the stability of CYLD is controlled by CRL1 β-TrCP ubiquitylation upon phosphorylation via IκB kinase [37].On the other hand, during NASH the E3 ligase TRIM47 targets CYLD for degradation [33].However, the exact interactome and stability regulation of CYLD during hepatic steatosis was unclear.
We provided evidence that the two CSN variants, CSN CSN7A and CSN CSN7B , have both general and specific functions during adipogenic differentiation [30,38].Consequently, we were interested to see whether CSN plays a role in hepatic steatosis in human liver cells.For this purpose, an in-vitro model was used in which treatment of HepG2 cells with palmitic acid (PA) caused steatosis in cultured cells characterized by the accumulation of fatty acids in LDs and the overexpression of PPAR-γ [39][40][41].Just as in adipocytes [30], the downregulation of CSN subunits in HepG2 cells and human primary hepatocytes led to the reduced expression of PPAR-γ.Moreover, we discovered that CYLD is associated with the CSN in LiSa-2 preadipocytes as well as hepatocytes.Closer examination revealed that CYLD binds to both CSN CSN7A and CSN CSN7B variants, albeit with preference to CSN CSN7A .PA treatment influences CYLD binding to CSN CSN7A and exacerbated hepatic steatosis.An increase of the CSN CSN7A variant elevated CYLD stability and reduced LD formation raising resistance to hepatic steatosis in HepG2 cells.

Silencing of CSN Subunits Reduces PPAR-γ Expression in Liver Cells
Since CSN is a major player in adipogenesis [30,42], we were interested to see whether the regulator of CRLs is also involved in the pathogenesis of hepatic steatosis.To evaluate the role of CSN in hepatic steatosis, we compared healthy livers and fatty livers from patients by hematoxylin/eosin staining, Oil-Red-O (ORO) staining, and immunohistochemistry staining of both a paraffin-embedded liver and a frozen liver using the anti-CSN3 antibody (Figure 1A).As expected, ORO staining revealed an accumulation of fat in the fatty liver.Since CSN3 is an integrated subunit of the CSN holo-complex [32], staining with anti-CSN3 antibody visualized total CSN in the cells.Immunohistochemical analysis with both paraffin-embedded and frozen tissue revealed an increase of total CSN in fatty liver (Figure 1A, right two panels).
Next, we were interested to see whether the downregulation of CSN subunits had an impact on PPAR-γ, a major player in hepatic steatosis [8].To silence CSN subunits siRNAs against siGFP (control), siCSN1 or siCSN7B were transfected in HepG2 cells with a final concentration of 50 nM.After 24 h, cells were incubated with or without PA for another 24 h.As demonstrated before, treatment of HepG2 cells with PA caused an increase of lipid accumulation and decrease in lipolysis as well as an elevation of PPAR-γ expression [39,40].As shown in Figure 1B, the downregulation of CSN1 (Figure 1C) as well as CSN7B (Figure 1D) by about 40% led to reduction of PPAR-γ expression confirming data obtained in LiSa-2 cells [30,42].Transfection with siCSN1 and siCSN7B and subsequent treatment with PA caused similar downregulation of CSN subunits and PPAR-γ as without PA (Figure 1B).
To optimize the downregulation of CSN7B, a new siRNA was designed called siCSN7B-RD (Figure 2A).Silencing of CSN7B using siCSN7B-RD in HepG2 cells was much more efficien, t reducing its expression by ~70% as revealed by densitometry and quantification of three independent experiments (Figure 2C).Silencing of CSN7B was accompanied by a significant downregulation of PPAR-γ.Most interestingly, under this condition the expression of CSN7A increased (Figure 2B,C).Similar results were obtained after treatment of HepG2 cells with PA (Figure 2B).In human primary hepatocytes, the downregulation of CSN7B by siCSN7B-RD was less efficient.Nevertheless, a significant decrease of PPAR-γ and an increase of CSN7A was measured (Figure 2D,E).After treatment with PA, similar effects were observed (Figure 2D).
In summary, the downregulation of CSN7B caused a reduction of PPAR-γ, the inducer of hepatic steatosis.Simultaneously, CSN7B silencing led to a compensatory increase of CSN7A independent of PA treatment.To optimize the downregulation of CSN7B, a new siRNA was designed called siCSN7B-RD (Figure 2A).Silencing of CSN7B using siCSN7B-RD in HepG2 cells was much more efficien, t reducing its expression by ~70% as revealed by densitometry and quantification of three independent experiments (Figure 2C).Silencing of CSN7B was accompanied by a significant downregulation of PPAR-γ.Most interestingly, under this condition the expression of CSN7A increased (Figure 2B,C).Similar results were obtained after treatment of HepG2 cells with PA (Figure 2B).In human primary hepatocytes, the downregulation of CSN7B by siCSN7B-RD was less efficient.Nevertheless, a significant decrease of PPAR-γ and an increase of CSN7A was measured (Figure 2D,E).After treatment with PA, similar effects were observed (Figure 2D).

The Ub-Specific Protease CYLD Preferentially Binds to CSN CSN7A Variant
For comparison, we studied the expression of CSN subunits in preadipogenic LiSa-2 cells.Recently, we established Lisa-2 cell lines stably expressing either empty Flag-vector, Flag-CSN7A, or Flag-CSN7B [30].The preadipogenic cell lines were stained with ORO (Figure 3A) or lysed for western blots (Figure 3C).As visualized by ORO staining in Figure 3A and as demonstrated before [30], there was an increase of lipids accumulated in LDs in LiSa-2 cells stably expressing Flag-CSN7B (Figure 3B).The Western blot showed a typical protein pattern.While stably transfected Flag-CSN7A was weakly expressed, stably transfected Flag-CSN7B was produced more abundantly.Both Flag-CSN7A and Flag-CSN7B overexpression induced an increase of other CSN subunits compared to control (Figure 3C).

Reciprocal Expression of CYLD and CSN7A as Well as CSN7B
Interestingly, the downregulation of CSN7B using siCSN7B-RD in HepG2 cells led to an increase of CYLD (Figure 4A, input).Since the reduction of CSN7B caused a compensatory increase of CSN7A, we assumed that binding to CSN CSN7A stabilized CYLD protein.
As expected, the subsequent immunoprecipitation of CSN7B revealed low amounts of CYLD associated with the CSN CSN7B complex, indicating that the stimulating effect of Flag-pulldowns were performed to evaluate the interactomes of CSN Flag-CSN7A and CSN Flag-CSN7B in LiSa-2 cells (Figure 3C, right panel).Since the CSN is a multi-DUB complex [32], we were curious to see whether the CSN variants interact with CYLD, a DUB involved in hepatic homeostasis [36].Strikingly, as shown in Figure 3C (left panel), CYLD was induced in Flag-CSN7A-LiSa-2 as well as in Flag-CSN7B-LiSa-2 cells.Flagpulldowns revealed binding of CYLD to both CSN variants (Figure 3C, right panel).However, more CYLD protein was associated with precipitated CSN Flag-CSN7A as compared to CSN Flag-CSN7B , which was precipitated in larger amounts.Based on this data, we concluded that CYLD preferentially binds to the CSN Flag-CSN7A variant.

Reciprocal Expression of CYLD and CSN7A as Well as CSN7B
Interestingly, the downregulation of CSN7B using siCSN7B-RD in HepG2 cells led to an increase of CYLD (Figure 4A, input).Since the reduction of CSN7B caused a compensatory increase of CSN7A, we assumed that binding to CSN CSN7A stabilized CYLD protein.
As expected, the subsequent immunoprecipitation of CSN7B revealed low amounts of CYLD associated with the CSN CSN7B complex, indicating that the stimulating effect of CSN7B silencing on CYLD expression is most likely due to the compensatory increase of CSN7A.
To study the reciprocal effect of CYLD on CSN subunit expression, specific siRNA against CYLD (siCYLD) was transfected in HepG2 cells or human primary hepatocytes with a final concentration of 50 nM.After 48 h, cells were lysed for Western blots.Downregulation of CYLD by about 40-50% (Figure 4C,D) led to an increase of CSN subunits expression including CSN7A and CSN7B in HepG2 cells as well as in human primary hepatocytes.Increased expression of CSN7A and CSN7B induced by CYLD downregulation is more pronounced in human primary hepatocytes and amounts to 1.8-fold for CSN7A and 2.2-fold for CSN7B (Figure 4D).
To see whether the stability of CYLD depends on one of the CSN variants, we used CSN7A and CSN7B knockout cells which we established by CRISPR/Cas9 strategy in HeLa cells [31].In addition, we used CSN5i-3, a specific inhibitor of CSN-mediated deneddylation [43] and the neddylation inhibitor MLN4924 [44], to ask whether degradation of CYLD depends on CSN-CRL axis.Neither the knockout of CSN7A nor that of CSN7B caused complete loss of CYLD, indicating that CSN variants have largely redundant functions in controlling CYLD stability.However, as shown in Figure 4E, CYLD expression in HeLa-CSN7B-KO cells was higher than in HeLa wt or HeLa-CSN7A-KO cells, perhaps induced by overexpression of CSN7A.In all HeLa derived cells MLN4924 (1 µM), completely blocked neddylation of cullin 1 (CUL1) whereas CSN5i-3 (1 µM) inhibited CSN-mediated deneddylation of CUL1.In both cases, prevention of the CSN-CRL pathway led to stabilization of CYLD in the parental HeLa cells, indicating degradation of CYLD at least in part occurred via the CSN-CRL-proteasome pathway.

Overexpression of Flag-CSN7A in Liver Cells Increases CYLD Expression
Flag-CSN7A or Flag-CSN7B were transiently transfected into HepG2 cells.After 24 h, transfected cells were treated with PA using indicated concentrations or control medium.After another 24 h, cells were harvested for Western blot and Flag-pulldowns.In lysates (input), the overexpression of Flag-CSN7A as well as Flag-CSN7B led to an increase of CSN subunit expression (Figure 5A, input), as it has been shown before in LiSa-2 cells stably overexpressing Flag-CSN7A or Flag-CSN7B (Figure 3C).CYLD expression was stimulated predominately by Flag-CSN7A independently of PA treatment.Flag pulldowns revealed that Flag-CSN7A and Flag-CSN7B were integrated into endogenous CSN complexes after 48 h of transfection, because all tested CSN subunits were co-precipitated.As shown before, more CYLD was associated with CSN Flag-CSN7A as compared to CSN Flag-CSN7B .Interestingly, PA treatment led to a reduction of CYLD, which is, however, accompanied by a similar decline of Flag-CSN7A and CSN3 (Figure 5A,B).CSN Flag-CSN7B -CYLD interaction was not affected by PA treatment (Figure 5A,C).Similar results were obtained with human primary hepatocytes (Figure 5D-F).The overexpression of Flag-CSN7A stimulated CYLD expression.Flag-pulldowns confirmed that CYLD preferentially binds to CSN Flag-CSN7A (Figure 5D, Flag-pulldowns).PA treatment caused decrease of CSN Flag-CSN7A accompanied by a decline of CYLD (Figure 5E), whereas CYLD binding to CSN Flag-CSN7B was not affected by PA (Figure 5F).

Downregulation of CSN7B Induces CYLD and Hinders Steatosis in HepG2 Cells
Mild steatosis in HepG2 cells can be induced by endogenous lipids after five to eight days.Steatosis in HepG2 cells was monitored for two and eight days by ORO staining (Figure 6A,B).In the upper panel, the entire well was directly photographed from a six-well plate (25-fold amplification), whereas the lower panel is a 200-fold amplification.On day two, the number of lipid loaded LDs did not differ between control and siCSN7B-RD sample.However, on day eight steatosis was developed as indicated by lipid loaded LDs, and the downregulation of CSN7B significantly reduced the number of LDs formed indicating retardation of steatosis.This result was confirmed by quantification of ORO staining (Figure 6B).In Figure 6C-E, using the above model, siGFP (control) or siCSN7B-RD were transfected in HepG2 cells and analyzed by Western blot after day two or day eight of steatosis.On day two, as expected, CSN7B was downregulated by about 60% causing an increase of CSN7A and of CYLD by about 30-40% and a concomitant decrease of PPAR-γ by more than 30% (Figure 6C,D).Although CSN7B downregulation on day eight was not as efficient anymore as on day two, there was still a significant increase of CSN7A as well as of CYLD and a considerable reduction of PPAR-γ (Figure 6E).
Collectively, our data demonstrate that steatosis can be retarded by CSN7B silencing causing a compensatory increase of CSN7A accompanied with CYLD stabilization and with the reduction of PPAR-γ.

Discussion
Data show that the CSN variants and their associated CYLD have a role in the pathogenesis of human fatty liver tissue.The CSN is a deneddylating enzyme complex and a major regulator of CRLs, conferring specifically to the UPS.As shown before in [31] and here, CSN CSN7A and CSN CSN7B variants substitute each other in terms of CSN-mediated deneddylating activity and binding to CRLs.Beyond that, the CSN variants are associated with additional proteins, which might account for specific functions in different locations, as shown in plants [45].The CSN is a DUB complex belonging to the JAMM sub-family and it is associated with additional DUBs [32].We have shown that the CSN binds to USP15 [46] and silencing of the CSN is accompanied by a reduction of USP15-mediated protection of CRL3-bound KEAP1 leading to inhibition of PPAR-γ expression and a blockade of adipogenesis [42].In addition, USP48 is a predominantly nuclear CSN-associated DUB, stabilizing the nuclear pool of NF-κB transcription factor RelA upon TNF stimulation [47].
Here we demonstrate, for the first time, the association of the DUB CYLD with the DUB CSN.It has been demonstrated before that CYLD suppresses NASH in mice and monkeys [33].Reduced CYLD expression in fatty liver tissue is associated with the severity of NAFLD.Under these conditions, it is believed that tripartite-motif-containing protein 47 (TRIM47) ubiquitylates CYLD following its degradation via 26S proteasome [33].Supplementing CYLD represents a promising therapeutic strategy for NASH [33,48].CYLD directly interacts with the kinase TAK1 and removes Lys63-ubiquitin chains.This impedes hyperactivation of JNK-p38 signaling mitigating NASH [33].
We observed a stable pool of CYLD binding to both CSN CSN7A as well as CSN CSN7B variants.On the other hand, there is a mobile CYLD pool, which binds exclusively to CSN CSN7A .Downregulation of CSN7B and knockout of CSN7B in HeLa cells is accompanied by a compensatory increase of CSN7A complex stabilizing CYLD.Likewise, the overexpression of CSN7A in liver cells protects CYLD.We conclude that CYLD preferentially interacts with the CSN CSN7A variant, which stabilizes the DUB.In our cell model of hepatic steatosis, downregulation of CSN7B in HepG2 cells after two and eight days revealed an increase of CSN7A that stabilized CYLD.Moreover, PPAR-γ was reduced.Interestingly, in other tissues increased PPAR-γ protects from triglyceride accumulation and insulin resistance [49].In HepG2 cells, the stabilization of CYLD and decline of PPAR-γ led to a retardation of hepatic steatosis demonstrated by significant reduction of LD formation after eight days.
Our data shows that an increase of CSN7A, be it by downregulation of CSN7B or by overexpression of CSN7A, stabilizes CYLD and prevents development of hepatic steatosis.To our knowledge, there is currently no pharmacological treatment approved for NAFLD/NASH and weight loss.Following a diet and lifestyle intervention is the only therapy to reduce liver fat and insulin resistance [50,51].Therefore, stabilizing CSN CSN7A and its interaction with CYLD might be a strategy to retard the pathogenesis of human fatty liver tissue.

Figure 1 .
Figure 1.Silencing of CSN1 or CSN7B reduces the expression of PPAR-γ, a regulator of hepatic steatosis.(A) Immunohistochemical staining using anti-CSN3 antibody in human normal and fatty liver tissue.Hematoxylin/eosin staining visualized nuclei, whereas ORO staining shows the number of LDs.Staining was visualized by light microscopy at 200-fold magnification.(B) Levels of the indicated proteins were assessed by Western blot analysis of HepG2 cells without (left panel) or with treatment using 0.2 mM PA (right panel) For 24 h.Control siRNA (siGFP) or siRNA directed against CSN1 (siCSN1) or CSN7B (siCSN7B) was transiently transfected into HepG2 cells 24 h before treatment with PA. (C) and (D) Blots as shown in (B) (left panel) were quantified by densitometry to measure the abundance of CSN1, CSN7B and PPARγ after transfection with siCSN1 (C) or siCSN7B (D).Data were normalized to γ-tubulin and are expressed as means of relative amounts plus/minus standard deviations (SD, n = 3).Unpaired Student's t-test was used for statistical analysis.* p < 0.05, ** p < 0.01.

Figure 1 .
Figure 1.Silencing of CSN1 or CSN7B reduces the expression of PPAR-γ, a regulator of hepatic steatosis.(A) Immunohistochemical staining using anti-CSN3 antibody in human normal and fatty liver tissue.Hematoxylin/eosin staining visualized nuclei, whereas ORO staining shows the number of LDs.Staining was visualized by light microscopy at 200-fold magnification.(B) Levels of the indicated proteins were assessed by Western blot analysis of HepG2 cells without (left panel) or with treatment using 0.2 mM PA (right panel) For 24 h.Control siRNA (siGFP) or siRNA directed against CSN1 (siCSN1) or CSN7B (siCSN7B) was transiently transfected into HepG2 cells 24 h before treatment with PA. (C,D) Blots as shown in (B) (left panel) were quantified by densitometry to measure the abundance of CSN1, CSN7B and PPAR-γ after transfection with siCSN1 (C) or siCSN7B (D).Data were normalized to γ-tubulin and are expressed as means of relative amounts plus/minus standard deviations (SD, n = 3).Unpaired Student's t-test was used for statistical analysis.* p < 0.05, ** p < 0.01.

Figure 2 .
Figure 2. Silencing by siCSN7B-RD increases CSN7A expression and reduces PPAR-γ in HepG2 cells and in human primary hepatocytes.(A) The sequence of siCSN7B-RD.(B) Western blot of HepG2 cells transfected with siGFP or siCSN7B-RD for 24 h followed by treatment without (left panel) or with 0.2 mM of PA (right panel) for 24 h.Antibodies were used as indicated.(C) Blots as shown in (B) (left panel) were quantified by densitometry to measure the abundance of CSN7B, CSN7A and PPAR-γ after transfection with siGFP (control) or siCSN7B-RD.Data were normalized to γ-tubulin and expressed as means of relative amounts plus/minus standard deviations (SD, n = 3).Unpaired Student's t-test was used for statistical analysis.** p < 0.01.(D) Western blot of human primary hepatocytes transfected with siGFP or siCSN7B-RD for 24 h, followed by treatment without (left panel) or with 0.2 mM of PA (right panel) for 24 h.Antibodies were used as indicated.(E) Blots as shown in (D) (left panel) were quantified by densitometry to measure the abundance of CSN7B,

Figure 2 .
Figure 2. Silencing by siCSN7B-RD increases CSN7A expression and reduces PPAR-γ in HepG2 cells and in human primary hepatocytes.(A) The sequence of siCSN7B-RD.(B) Western blot of HepG2 cells transfected with siGFP or siCSN7B-RD for 24 h followed by treatment without (left panel) or with 0.2 mM of PA (right panel) for 24 h.Antibodies were used as indicated.(C) Blots as shown in (B) (left panel) were quantified by densitometry to measure the abundance of CSN7B, CSN7A and PPAR-γ after transfection with siGFP (control) or siCSN7B-RD.Data were normalized to γ-tubulin and expressed as means of relative amounts plus/minus standard deviations (SD, n = 3).Unpaired Student's t-test was used for statistical analysis.** p < 0.01.(D) Western blot of human primary hepatocytes transfected with siGFP or siCSN7B-RD for 24 h, followed by treatment without (left panel) or with 0.2 mM of PA (right panel) for 24 h.Antibodies were used as indicated.(E) Blots as shown in (D) (left panel) were quantified by densitometry to measure the abundance of CSN7B, CSN7A and PPAR-γ after transfection with siGFP (control) or siCSN7B-RD.Data were normalized to γ-tubulin and expressed as means of relative amounts plus/minus standard deviations (SD, n = 3).Unpaired Student's t-test was used for statistical analysis.* p < 0.05, ** p < 0.01.

Livers 2021, 1 , 7 Figure 3 .
Figure 3.The deubiquitylating enzyme CYLD is associated with the CSN and preferentially binds to CSN CSN7A variant.(A) Increased formation of LDs in LiSa-2 preadipocytes stably expressing Flag-CSN7B compared to control (Flag-vector) and Flag-CSN7A expressing cells.The three cell lines were stained with ORO and LD formation was visualized by light microscopy at 200-fold magnification.(B) Quantification of ORO stain as shown in (A).Cells were lysed and supernatants were measured spectrophotometrically at 520 nm.Results are expressed as means plus/minus standard deviations (SD, n = 3).Unpaired Student's t-test was used for statistical analysis.** p < 0.01.(C) Left panel: Western blot of LiSa-2-Flag, LiSa-2-Flag-CSN7A, and LiSa-2-Flag-CSN7B cells.Antibodies were used as indicated.Right panel: Flag-pulldowns of LiSa-2-Flag, LiSa-2-Flag-CSN7A, and LiSa-2-Flag-CSN7B cells.

Figure 3 .
Figure 3.The deubiquitylating enzyme CYLD is associated with the CSN and preferentially binds to CSN CSN7A variant.(A) Increased formation of LDs in LiSa-2 preadipocytes stably expressing Flag-CSN7B compared to control (Flag-vector) and Flag-CSN7A expressing cells.The three cell lines were stained with ORO and LD formation was visualized by light microscopy at 200-fold magnification.(B) Quantification of ORO stain as shown in (A).Cells were lysed and supernatants were measured spectrophotometrically at 520 nm.Results are expressed as means plus/minus standard deviations (SD, n = 3).Unpaired Student's t-test was used for statistical analysis.** p < 0.01.(C) Left panel: Western blot of LiSa-2-Flag, LiSa-2-Flag-CSN7A, and LiSa-2-Flag-CSN7B cells.Antibodies were used as indicated.Right panel: Flag-pulldowns of LiSa-2-Flag, LiSa-2-Flag-CSN7A, and LiSa-2-Flag-CSN7B cells.

Figure 4 .
Figure 4. Silencing of CYLD stabilizes CSN7A as well as CSN7B in HepG2 cells and in human primary hepatocytes.(A) Immunoprecipitation of CSN7B.Western blot of HepG2 cells transfected with siGFP or siCSN7B-RD for 48 h (input, left panel) and IP of CSN7B (middle and right panel).Antibodies were used as indicated.(B) Western blot of HepG2 cells (left panel) or human primary hepatocytes (right panel) transiently transfected with siGFP or siCYLD for 48 h.Antibodies were used as in (A).(C,D) Western blots as shown in (B) left panel, (C) or in (B) right panel, (D) were quantified by densitometry

Figure 4 .
Figure 4. Silencing of CYLD stabilizes CSN7A as well as CSN7B in HepG2 cells and in human primary hepatocytes.(A) Immunoprecipitation of CSN7B.Western blot of HepG2 cells transfected with siGFP or siCSN7B-RD for 48 h (input, left panel) and IP of CSN7B (middle and right panel).Antibodies were used as indicated.(B) Western blot of HepG2 cells (left panel) or human primary hepatocytes (right panel) transiently transfected with siGFP or siCYLD for 48 h.Antibodies were used as in (A).(C,D) Western blots as shown in (B) left panel, (C) or in (B) right panel, (D) were quantified by densitometry to measure the abundance of CYLD, CSN7A and CSN7B after transfection with siGFP (control) or siCYLD.Data were normalized to γ-tubulin and expressed as means of relative amounts plus/minus standard deviations (SD, n = 4).Unpaired Student's t-test was used for statistical analysis.* p < 0.05, ** p < 0.01.(E) Western blots of HeLa wildtype (wt), HeLa-CSN7A-KO and HeLa-CSN7B-KO cells treated without or with MLN4924 (1 µM) or CSN5i-3 (1 µM) for 16 h.Antibodies were used as indicated.

Figure 5 .Figure 5 .
Figure 5. Overexpression of CSN7A, but not of CSN7B, stabilizes CYLD.(A) Western blots of 1.2 × 10 7 HepG2 cells transiently transfected with 60 μg Flag-vector (control), Flag-CSN7A or Flag-CSN7B.24 h after transfection, cells were treated with 0, 0.04 or 0.2 mM PA for another 24 h.Cells were harvested for Western blots (left three panels) or Flag pulldowns (right three panels).Antibodies were used as indicated.(B,C) Blots of Flag-pulldowns as shown in (A) (right three panels) were quantified by densitometry to measure the abundance of CSN7A, CSN3, and CYLD in Flag-CSN7A pulldowns (B) Figure 5. Overexpression of CSN7A, but not of CSN7B, stabilizes CYLD.(A) Western blots of 1.2 × 10 7 HepG2 cells transiently transfected with 60 µg Flag-vector (control), Flag-CSN7A or Flag-CSN7B.24 h after transfection, cells were treated with 0, 0.04 or 0.2 mM PA for another 24 h.Cells were harvested for Western blots (left three panels) or Flag pulldowns (right three panels).Antibodies were used as indicated.(B,C) Blots of Flag-pulldowns as shown in (A) (right three panels) were quantified by densitometry to measure the abundance of CSN7A, CSN3, and CYLD in Flag-CSN7A pulldowns (B) or of CSN7B, CSN3 and CYLD in Flag-CSN7B pulldowns (C) after treatment with 0, 0.04 and 0.2 mM PA. Data are expressed as means of relative density plus/minus standard deviations (SD, n = 3).Unpaired Student's t-test was used for statistical analysis.* p < 0.05, ** p < 0.01.(D) Western blots of human primary hepatocytes transiently transfected with Flag-vector (control), Flag-CSN7A or Flag-CSN7B as in (A).Immediately after transfection, cells were treated with 0 or 0.2 mM PA for 24 h.Cells were harvested for Western blots (left two panels) or Flag pulldowns (right two panels).Antibodies were used as in (A).(E,F) Blots of Flag-pulldowns as shown in (D) (right two panels) were quantified by densitometry to measure the abundance of CSN7A, CSN3 and CYLD in Flag-CSN7A pulldowns (E) or CSN7B, CSN3 and CYLD in Flag-CSN7B pulldowns (F) after treatment with 0 and 0.2 mM PA. Data are expressed as means of relative density plus/minus standard deviations (SD, n = 3).Unpaired Student's t-test was used for statistical analysis.* p < 0.05, ** p < 0.01.

Livers 2021, 1 , 13 Figure 6 .
Figure 6.Silencing of CSN7B retards hepatic steatosis in HepG2 cells by elevated CSN CSN7A stabilizing CYLD.(A) siGFP (control) or siCSN7B-RD were transiently transfected into HepG2 cells.Cells were harvested on day 2 and day 8 of hepatic steatosis and stained with ORO to visualize LD formation by light microscopy with a magnification of 25-fold (upper panels) or 200-fold (lower panels).(B) Quantification of ORO stain as shown in (A).Cells were lysed and supernatants were measured spectrophotometrically at 520 nm.Results are expressed as means plus/minus standard deviations (SD, n = 3).Unpaired Student's t-test was used for statistical analysis.** p < 0.01.(C) Western blot of HepG2 cells transfected with siGFP (control) or siCSN7B-RD and harvested after 2 days (left panel) or 8 days (right panel) of hepatic steatosis.Antibodies were used as indicated.(D,E) Blots of HepG2 cells transfected with siGFP (control) or siCSN7B-RD after two days of steatosis (D) or 8 days of steatosis (E) were quantified by densitometry to measure the abundance of CSN7A, CSN7B,

Figure 6 .
Figure 6.Silencing of CSN7B retards hepatic steatosis in HepG2 cells by elevated CSN CSN7A stabilizing CYLD.(A) siGFP (control) or siCSN7B-RD were transiently transfected into HepG2 cells.Cells were harvested on day 2 and day 8 of hepatic steatosis and stained with ORO to visualize LD formation by light microscopy with a magnification of 25-fold (upper panels) or 200-fold (lower panels).(B) Quantification of ORO stain as shown in (A).Cells were lysed and supernatants were measured spectrophotometrically at 520 nm.Results are expressed as means plus/minus standard deviations (SD, n = 3).Unpaired Student's t-test was used for statistical analysis.** p < 0.01.(C) Western blot of HepG2 cells transfected with siGFP (control) or siCSN7B-RD and harvested after 2 days (left panel) or 8 days (right panel) of hepatic steatosis.Antibodies were used as indicated.(D,E) Blots of HepG2 cells transfected with siGFP (control) or siCSN7B-RD after two days of steatosis (D) or 8 days of steatosis (E) were quantified by densitometry to measure the abundance of CSN7A, CSN7B, PPAR-γ and CYLD.Data were normalized to γ-tubulin and expressed as means of relative amounts plus/minus standard deviations (SD, n = 3).Unpaired Student's t-test was used for statistical analysis.* p < 0.05, ** p < 0.01.