Fatostatin blocks ER exit of SCAP but inhibits cell growth in a SCAP-independent manner

Sterol regulatory element-binding protein (SREBP) transcription factors are central regulators of cellular lipid homeostasis and activate expression of genes required for fatty acid, triglyceride, and cholesterol synthesis and uptake. SREBP cleavage activating protein (SCAP) plays an essential role in SREBP activation by mediating endoplasmic reticulum (ER)-to-Golgi transport of SREBP. In the Golgi, membrane-bound SREBPs are cleaved sequentially by the site-1 and site-2 proteases. Recent studies have shown a requirement for the SREBP pathway in the development of fatty liver disease and tumor growth, making SCAP a target for drug development. Fatostatin is a chemical inhibitor of the SREBP pathway that directly binds SCAP and blocks its ER-to-Golgi transport. In this study, we determined that fatostatin blocks ER exit of SCAP and showed that inhibition is independent of insulin-induced gene proteins, which function to retain the SCAP-SREBP complex in the ER. Fatostatin potently inhibited cell growth, but unexpectedly exogenous lipids failed to rescue proliferation of fatostatin-treated cells. Furthermore, fatostatin inhibited growth of cells lacking SCAP. Using a vesicular stomatitis virus glycoprotein (VSVG) trafficking assay, we demonstrated that fatostatin delays ER-to-Golgi transport of VSVG. In summary, fatostatin inhibited SREBP activation, but fatostatin additionally inhibited cell proliferation through both lipid-independent and SCAP-independent mechanisms, possibly by general inhibition of ER-to-Golgi transport.


CRISPR/Cas9-mediated knockout of SCAP in HEK293 cells
The SCAP-deficient line, WSC69 (parental line HEK293), was generated by clustered regularly interspaced short palindromic repeat (CRISPR)-Cas9-mediated genome editing (25). The human SCAP gene (GenBank reference number #NM_012235.3) contains 23 exons and is translated into a 1,279 amino acid protein. A CRISPR guide RNA (gRNA) to target sequence 276 to 295 nucleotides (5′-GGCTGCGTGAGAAGATATCT-3′) located in the exon 2 (#ENSE00003728083) of the SCAP mRNA was cloned into the Cas9-gRNA vector PX459 (Addgene #48139) and used to generate the knockout cell line. Transfected HEK293 cells were selected for growth in medium F containing 1.5 g/ml puromycin. Single clones were isolated by dilution cloning. Genomic DNA flanking the gRNA target site was amplified by standard PCR using primers (5′-GGGATTGAGGTCACTAGACC-3′ and 5′-GGT-GAATCAGTAGGTCAGGG-3′) and then sequenced by Sanger sequencing. WSC69 was one of the surviving clones showing two distinct deletions at the gRNA site. Knockout of SCAP was further confirmed by immunoblotting and growth assay under lipoprotein-depleted conditions. with work in breast cancer and prostate cancer (13,14), these studies highlight the critical role for the SREBP pathway in tumor growth. Thus, SREBP inhibitors are potential therapeutics for these diseases. Given that there are three SREBP family members, targeting of a pathway component required for SREBP activation, such as SCAP, S1P, or S2P, is a strategy to inhibit all SREBP function. Among these three essential components, only SCAP is specific to the SREBP pathway, making it the optimal target for SREBP pathway drug development.
Fatostatin (formerly called 125B11) is a diarylthiazole compound that blocks ER-to-Golgi transport of SCAP and inhibits SREBP activation (15,16). Fatostatin was originally identified from a synthetic small molecule library as a chemical inhibitor blocking insulin-induced adipogenesis (15). Mechanistic studies showed that fatostatin binds SCAP directly and blocks Golgi-specific glycosylation on SCAP, suggesting that fatostatin inhibits ER-to-Golgi transport of SCAP (16). Interestingly, fatostatin directly binds to SCAP outside of the sterol-sensing domain where cholesterol binds, and fatostatin does not promote SCAP binding to INSIG (16). These data suggested that fatostatin acts specifically on SCAP through a mechanism different from that of sterols. Fatostatin has been used as a SCAP inhibitor and applied to preclinical models for metabolic diseases and cancer (16)(17)(18). In this study, we further investigated the mechanism of fatostatin action on the SREBP pathway. We provide direct evidence that fatostatin blocks ER exit of SCAP, and that inhibition is independent of INSIG proteins. Both fatostatin and PF-429242, an inhibitor of S1P, potently inhibited cell growth. Exogenous lipids restored cell proliferation in the presence of PF-429242, but unexpectedly failed to rescue fatostatin-treated cells. In addition, fatostatin inhibited growth of cells lacking SCAP. Finally, we show that fatostatin delayed ER-to-Golgi transport of vesicular stomatitis virus glycoprotein (VSVG), a marker for general secretory function. In summary, although fatostatin inhibits SREBP activation, it also inhibits cell growth through both lipid-independent and SCAP-independent mechanisms, potentially through general inhibition of ER-to-Golgi transport.

EXPERIMENTAL PROCEDURES
Materials 5,000 cells per well (96-well plate) in conditions indicated in figure legends. Seventy-two hours later, MTT (15 l per well) was added and incubated at 37°C for 4 h. Stop solution was added and incubated at room temperature overnight. A 570 was measured. Viability was normalized to the vehicle-treated condition.

VSVG maturation assay
VSVG maturation was measured as previously described (30). CHO-7 cells were infected by VSV (San Juan strain, Indiana serotype) at a multiplicity of infection of 20 for 30 min. Then infection medium was replaced by medium A supplemented with 5% (v/v) FBS and cells were incubated for 1. with 20 g/ml aprotinin, 20 g/ml leupeptin, and 2 g/ml pepstatin A, and VSVG protein was immunoprecipitated with a polyclonal anti-VSV (27). The kinetics of oligosaccharide processing were determined as described previously, using 0.4 mU of endoglycosidase H (Endo H; New England Biolabs, Beverly, MA) (31). Proteins were separated on 10% polyacrylamide gels and detected using a Molecular Imager FX phosphorImager (Bio-Rad). The percent of protein processed at each chase time was determined after quantitation using Quantity One software (Bio-Rad).

Fatostatin inhibits ER exit of SCAP
Fatostatin is reported to block SREBP processing in CHO cells (16). To investigate the mechanism of fatostatin inhibition on SREBP activation, we first confirmed fatostatin inhibition of SREBP in CHO cells by analyzing the expression of SREBP transcriptional target HMG-CoA synthase (HMGCS1) via RT-qPCR. Fatostatin showed dose-dependent inhibition on HMGCS1 expression (Fig. 1A). Consistent with the published results, 20 M fatostatin completely blocked sterol-regulated HMGCS1 expression, and the IC 50 was between 2.5 and 10 M. Next, we conducted a similar experiment in human cells to analyze fatostatin's effect on sterol-regulated SREBP cleavage (Fig. 1B, C). HEK293 cells were incubated without sterols to induce cleavage of SREBPs. Cell nuclei and membranes were assayed for SREBP cleavage products, either the NH 2 -terminal transcription factor domain for SREBP1 or the membranebound COOH-terminal segment for SREBP2. As expected, sterol depletion induced SREBP cleavage and increased levels of the NH 2 -terminal segment of SREBP1 (SREBP1-N) and the COOH-terminal fragment of SREBP2 (SREBP2-C) ( Fig. 1B, C, lanes 1, 2). Similar to sterols, fatostatin inhibited sterol-regulated SREBP processing (Fig. 1B, C, lanes

Protein and RNA preparation and analysis
Mammalian cell fractionation and protein immunoblotting analysis has been described previously (26). Total RNA was isolated from mammalian cells using RNA STAT-60. For RT-quantitative (q)PCR analysis of transcript abundance, total RNA (2 g per sample) was treated with RNase-free DNase I in a total volume of 10 l at room temperature (22°C) for 15 min. Reactions were stopped by the addition of 1 l of 25 mM EDTA. After heating at 65°C for 10 min, each sample received 4 l of dNTPs (2.5 mM), 2 l of 10× RT buffer, 2 l of primers [oligo d(T)23VN for human HEK293 samples and random primer mix for CHO samples], 1 l of RNase inhibitor, and 1 l of M-MuLV reverse transcriptase. Reverse transcription was carried out at 25°C for 5 min followed by 42°C for 60 min and then 80°C for 10 min. cDNAs of the tested genes were quantified by real-time PCR using SYBR Green qPCR master mix. -ACTIN (for CHO cells samples) or 36B4 (for human cell samples) served as the internal control to calculate the relative expression across different samples.

Immunofluorescence microscopy
GFP-SCAP cells were seeded on day 0 at a density of 2 × 10 5 cells per well (6-well plate, 22 × 22 mm coverslip per well) in medium A supplemented with 5% (v/v) FBS. On day 1, cells were washed twice with PBS and then incubated in DMEM/F12 medium containing 1% HPCD to deplete cholesterol for 1 h. Then cells were washed with PBS and refed with medium C containing sterols or different concentrations of fatostatin for another 2 h. Cells were fixed, permeabilized, and stained as previously described (27). Briefly, cells were fixed in 3% paraformaldehyde in PBS at room temperature for 10 min and then permeabilized by 0.5% Triton X-100/PBS/glycine for 3 min at room temperature. Primary antibodies (anti-GFP, 1:500 or anti-GM130, 1:250) and secondary antibodies (Alexa-488 goat anti-rabbit IgG or Alexa-594 goat antimouse IgG, 1:250) were incubated for 30 min, respectively. Coverslips were mounted to slides and dried in the dark overnight before visualization by the Zeiss AXIO Imager-M2 microscope. Images were captured by Zeiss Plan-Neofluar 100×/1.30 oil objective and processed by iVision software. Quantitative colocalization analysis was conducted using Image J with JACoP plug-in (28). Pearson's correlation coefficient was calculated by the equation: where R is the red channel (GM130) and G is the green channel (GFP-SCAP).

Cell growth and viability assays
Crystal violet growth assay used for CHO-7 and other stable cell lines has been described previously (29). Briefly, cells were seeded on day 0 at a density of 3 × 10 4 cells per well (6-well plate) in medium A supplemented with 5% (v/v) FBS. On day 1, cells were refed as indicated in the figure legends. Cells were refed every 2 days. On day 14, cells were washed with PBS once, fixed in cold methanol at 20°C for 10 min, and stained with 0.05% crystal violet at room temperature for 10 min. Plate images were acquired using a flat bed scanner in transmitted light mode at a resolution of 300 dpi. For quantification, crystal violet-stained cells were dissolved by freshly-prepared 10% acetic acid solution (1 ml per well) at room temperature for 15 min. Then 100 l of each sample was transferred to a 96-well plate and measured at A 590 . MTT growth assay was performed using CellTiter 96 NonRadioactive cell proliferation assay kit. Briefly, cells were seeded at a density of SREBP ER-to-Golgi transport, rescued fatostatin inhibition of SREBP processing (16,32). These data suggested that fatostatin blocked ER-to-Golgi transport of SCAP (16). To rule out indirect effects of fatostatin on Golgi glycosylation machinery or brefeldin A on fatostatin action, we directly analyzed SCAP trafficking using a GFP-SCAP fusion protein to test fatostatin action on the ER-to-Golgi transport of SCAP ( Fig. 2A) (22). GFP-SCAP cells were sterol-depleted using hydroxypropyl--cyclodextrin for 1 h (33), and then incubated for an additional 2 h in the absence of sterols. [3][4][5][6], and the inhibition was dose-dependent. Fatostatin inhibition of SREBP activation was further validated by analyzing the expression of SREBP target HMGCS1 using RT-qPCR. Fatostatin showed dose-dependent inhibition on HMGCS1 expression (Fig. 1D). Collectively, fatostatin blocked sterol-regulated SREBP activation in mammalian cells, and IC 50 was between 2.5 and 10 M.
Previously, Kamisuki et al. demonstrated that fatostatin blocked Golgi-dependent glycosylation of SCAP and that brefeldin A, an agent that bypasses the requirement for contrast, when cells were incubated in the presence of sterols, GFP-SCAP localized to the ER ( Fig. 2A, panel 1). Similar to sterols, fatostatin treatment blocked ER exit of GFP-SCAP in a dose-dependent fashion ( Fig. 2A, panels 9, 13, 17, and 21). Quantitative colocalization analysis showed On day 1, cells were washed twice with PBS and then incubated in medium A containing 1% HPCD for 1 h. Cells were then washed with PBS and refed medium C with addition of sterols (1 g/ml 25-HC, 10 g/ml cholesterol) or fatostatin as indicated for another 2 h. Cells were fixed in 3% paraformaldehyde for indirect immunofluorescence analysis. GFP-SCAP was imaged using rabbit polyclonal anti-GFP antibody followed by Alexa-488 secondary antibody; Golgi was visualized by mouse anti-GM130 followed by Alexa-594 secondary antibody; DAPI was used for nuclear staining. B: Quantification of SCAP localization in Golgi using Pearson's correlation coefficient.

Exogenous lipid does not rescue cell growth inhibition by fatostatin
Fatostatin inhibits cell proliferation, and it has been proposed that fatostatin limits lipid supply by blocking SREBP activation (36,37). We next designed experiments to test this hypothesis. Cultured mammalian cells acquire lipids through both de novo synthesis and uptake from the culture medium. When cultured in lipoprotein-deficient medium, cells rely on SREBP-dependent transcription to upregulate de novo lipid synthesis and to provide lipids required for cell growth. Consistent with this, wild-type CHO cells grew in lipoprotein-deficient medium, but cells lacking functional SREBP (SCAP null and S1P null) failed to grow (Fig. 4A, well 1). Addition of exogenous lipids (cholesterol, oleate, and the isoprenoid precursor, mevalonate) rescued the growth of SCAP-null and S1P-null cells (Fig. 4A, well 4), indicating that SCAP-null and S1P-null cells failed to grow due to an inadequate supply of lipids. PF-429242, a specific S1P inhibitor (35), inhibited wildtype cell growth in lipoprotein-deficient medium, and growth was restored when lipids were provided (Fig. 4A,  wells 3 and 6). In addition, PF-429242 failed to inhibit growth in cells lacking either S1P or SCAP when lipids were supplemented. These data indicated that PF-429242 inhibited cell growth by limiting lipid supply through blocking SREBP activation. Unexpectedly, fatostatin showed distinct effects on growth compared with the S1P inhibitor, PF-429242. Fatostatin inhibited wild-type cell growth in lipoprotein-deficient medium, but lipid supplementation failed to rescue cell growth in the presence of fatostatin (Fig. 4A, wells 2 and 5). We further determined the minimum dose of fatostatin required for growth inhibition. Fatostatin inhibited growth at 2.5 M in lipoprotein-deficient medium and at 5 M when supplemented with exogenous lipids (Fig. 4B). In addition, we tested the effects of fatostatin on cell growth using a MTT cell proliferation assay. Consistent with the crystal violet stain assay results (Fig. 4A, B), PF-429242 inhibited cell growth and lipid supplement rescued growth (Fig. 4C), while lipid supplement had no effect on cell growth in the presence of fatostatin (Fig. 4D). Collectively, these results demonstrate that growth inhibition by fatostatin cannot be bypassed by exogenous lipid, indicating that fatostatin blocks cell growth through an SREBPindependent mechanism.

Fatostatin inhibits cell growth in a SCAP-independent manner
Fatostatin inhibited growth of SCAP-null CHO cells even when supplemented with exogenous lipids, suggesting that the growth inhibition by fatostatin was SCAP independent that 20 M fatostatin was as effective as sterols in preventing GFP-SCAP ER-to-Golgi transport (Fig. 2B), consistent with the gene expression data in CHO cells (Fig. 1A). These data demonstrate directly that fatostatin prevents sterol-regulated ER exit of SCAP.

Fatostatin inhibition on SREBP activation does not require INSIGs
Cholesterol and oxysterols block ER exit of SCAP by promoting the interaction between SCAP and the ERresident INSIG proteins (34). Next, we investigated whether fatostatin inhibition of SREBP activation also required INSIG proteins. We treated wild-type CHO-7 cells or mutant cells lacking both INSIG1 and INSIG2 with fatostatin or 25-hydroxycholesterol, which inhibits SREBP processing by promoting the interaction between SCAP and INSIGs (4). In wild-type CHO-7 cells, SREBP1-N accumulated upon sterol depletion and both fatostatin and 25-hydroxycholesterol blocked SREBP1-N accumulation (Fig. 3A,  lanes 1-3). Consistent with these protein results, expression of the SREBP target gene, HMGCS1, was suppressed by treatment with fatostatin, 25-hydroxycholesterol, and PF-429242, a specific inhibitor of S1P (Fig. 3B) (35). As expected, in cells lacking INSIGs, PF-429242, but not effects on cell growth, indicating that fatostatin has additional molecular targets in both CHO and human cells.

Fatostatin delays ER-to-Golgi transport of VSVG
Fatostatin blocks ER exit of GFP-SCAP (Fig. 2). Given the fact that fatostatin has SCAP-independent effects on cell growth, we next asked whether fatostatin broadly inhibits ER-to-Golgi transport. For these studies, we examined the transport of the VSVG, which is constitutively transported from the ER to the Golgi by COPII-coated vesicles and has been widely used to study membrane transport (38,39). VSVG is translated in the ER and receives core N-linked glycosylation, which can be removed by the endoglycosidase, Endo H, to yield a faster migrating species denoted as G S . Deglycosylation of VSVG by Golgi-resident enzymes makes the VSVG carbohydrate linkages Endo H-resistant, leading to a slower migrating species denoted (Fig. 4A, well 5). To test whether the SCAP-independent nature of fatostatin inhibition is specific to CHO cells, we examined the effects of fatostatin on growth of both kidney-and pancreas-derived human cells lacking SCAP. We knocked out SCAP in HEK293, a human embryonic kidney line, and Pa03c, a human pancreatic carcinoma line, using CRISPR-Cas9-mediated genome editing (25). We then assayed cell growth of these paired cell lines in medium supplemented with exogenous lipids. If SCAP is the molecular target of fatostatin, cells lacking SCAP may exhibit cell growth defects, but we would expect fatostatin to have no additional effect on cells lacking the target. Indeed, fatostatin inhibited cell growth of HEK293 wild-type and SCAP-null cells equally with an IC 50 of 10 M (Fig. 5A). The similar result was observed for Pa03c cells and the corresponding SCAP-null mutant (Fig. 5B). Collectively, these observations show that fatostatin has SCAP-independent lacking both INSIG1 and INSIG2 (Fig. 3), demonstrating that fatostatin blocks ER exit of SCAP through a distinct INSIGindependent mechanism. Fatostatin might retain SCAP in the ER by promoting its interaction with a different ER-resident protein. Alternatively, fatostatin might specifically block SCAP interaction with COPII components, thereby preventing incorporation into COPII vesicles.
as G R . Therefore, the rate of VSVG ER-to-Golgi transport can be measured by acquisition of Endo H-resistant glycosylation in a pulse-chase experiment (31). Compared with vehicle-treated cells, fatostatin inhibited the ER-to-Golgi transport of VSVG (Fig. 6A, B). These data demonstrate that fatostatin has general effects on ER-to-Golgi transport, providing a potential mechanism for the SCAPindependent effects on cell growth.

DISCUSSION
Fatostatin is a chemical inhibitor of SCAP and the SREBP pathway that is being used in preclinical disease models (17,18). Previous work by Kamisuki et al. (16) demonstrated that fatostatin blocked SREBP cleavage and target gene expression. In addition, fatostatin prevented Golgi modification of SCAP N-linked glycans, suggesting that fatostatin blocked SREBP activation by inhibiting the ER-to-Golgi transport of SCAP. Here, we confirmed that fatostatin blocked SREBP cleavage and SREBP-dependent transcription using human kidney cells (Fig. 1). Using a direct assay of SCAP trafficking (Fig. 2), we observed that fatostatin blocked ER exit of GFP-SCAP at doses that prevent SREBP cleavage. Collectively, these data demonstrate that fatostatin inhibits SREBP cleavage by blocking ER exit of SCAP.
Studies using biotinylated derivatives of fatostatin indicate that fatostatin binds directly to SCAP (16). Notably, fatostatin binds SCAP outside of the INSIG binding domain (SCAP amino acids 1-448), which also binds to cholesterol. In addition, unlike sterols, fatostatin failed to promote binding between SCAP and INSIG (16). These data suggested that fatostatin and sterols inhibit SCAP trafficking through distinct mechanisms. To examine this directly, we tested whether fatostatin inhibition of SREBP cleavage requires INSIGs using CHO cells that lack both INSIG1 and INSIG2. INSIGs are absolutely required to mediate sterol inhibition of SREBP. In INSIG-null cells, sterols fail to block the ER exit of SCAP, resulting in constitutive SREBP activation (21). Fatostatin inhibited SREBP cleavage and target gene expression in cells Fatostatin inhibits growth in SCAP knockout cells. A: Wild-type HEK293 and SCAP-null WSC69 cells were seeded on at a density of 5,000 cells per well (96-well plate) in medium D supplemented with 10% (v/v) FBS containing the indicated concentration of fatostatin, respectively. Seventy-two hours later, MTT (15 l per well) was added and incubated at 37°C for 4 h. Stop solution was added and incubated at room temperature overnight, and was measured at A 570 . Viability was normalized to vehicle-treated condition. Error bars represent the SD from five technical replicates (mean ± SD). B: Wild-type Pa03c and SCAP-null WSC55 cells treated and assayed for viability as in (A). Error bars represent the SD from five technical replicates (mean ± SD).  (20 M). Cells were lysed and VSVG was immunoprecipitated using anti-VSVG antibody. Endo Htreated samples were resolved on 10% SDS-PAGE and exposed to film overnight. B: Data are the average of two biological replicates.
Genetic loss-of-function studies demonstrate a role for the SREBP pathway in supporting cancer cell proliferation and tumor growth (14,17). Consequently, the anticancer potential of fatostatin has been investigated. In prostate cancer cell lines, fatostatin inhibits cell proliferation and insulin-stimulated protein and DNA synthesis (18,40,41). These effects have been attributed to fatostatin inhibition of SCAP-and SREBP-dependent transcription. CHO cells lacking S1P or SCAP fail to grow in the absence of exogenous lipid, but cell growth can be rescued by lipid addition, demonstrating that exogenous lipid can bypass essential SREBP pathway functions (Fig.  4A) (2,20). Likewise, chemical inhibition of S1P using PF-429242 blocked cell growth, and addition of exogenous lipids rescued cell growth (Fig. 4C). In contrast, fatostatin inhibited cell proliferation both in the absence and presence of exogenous lipids (Fig. 4A, B, D), demonstrating that fatostatin has lipid-independent and SREBPindependent effects on cell proliferation. Furthermore, fatostatin inhibited growth of SCAP-null CHO cells grown in the presence of exogenous lipids, suggesting that the effects of fatostatin on cell growth are SCAP-independent. Finally, using human kidney-derived cells and pancreatic cancer cells, fatostatin inhibited cell growth equally in wild-type and SCAP-null cells supplemented with exogenous lipids (Fig. 5). Collectively, these results indicate that fatostatin has SCAP-independent and off-target effects on cell proliferation.
As a first step to investigate the SCAP-independent effects of fatostatin, we tested whether fatostatin generally inhibits protein transport for the ER to the Golgi by examining the model secretory protein, VSVG (30,38,39). Using a kinetic pulse-chase assay, fatostatin delayed ERto-Golgi trafficking of VSVG (Fig. 6). At 20 min of chase, fatostatin treatment resulted in an 50% decrease of Endo H-resistant VSVG. By comparison, the VSVG del 4-29 mutant, which completely lacks the tyrosine and diacidic motifs required for export (42), showed an 66% decrease of Endo H-resistant VSVG at 20 min (43). Thus, fatostatin severely affects ER-to-Golgi transport of a model secretory protein, providing a potential mechanism for the SCAP-independent effects of fatostatin on cell growth.
Accumulating evidence highlights the importance of the SREBP pathway in the development of metabolic diseases and cancer. Consequently, SCAP inhibitors are potential therapeutics. In this study, we showed that fatostatin inhibits SCAP ER exit and SREBP-dependent transcription. However, fatostatin nonspecifically blocks ER-to-Golgi transport and has SCAP-independent effects on cell proliferation. Thus, fatostatin has additional cellular targets and is not a specific inhibitor for SCAP. Given the clinical importance of the SREBP pathway, efforts to develop a specific SCAP inhibitor should continue.
The authors thank members of the Espenshade laboratory for excellent advice and discussion; Jiwon Hwang for critical reading of the manuscript; Shan Zhao for outstanding technical support; Russell DeBose-Boyd, Michael Brown, and Joseph