Mutations in the gene encoding UbiA prenyltransferase domain-containing protein-1 (UBIAD1) cause Schnyder corneal dystrophy (SCD), a rare autosomal dominant eye disease characterized by opacification of the cornea (Klintworth, 2009; Weiss, 2009; Orr et al., 2007; Weiss et al., 2007). Although apparent early in life, corneal opacification associated with SCD progresses slowly with age and ultimately leads to reduced visual acuity. The severity of visual impairment is underscored by the frequency in which corneal transplant surgery is utilized for treatment of SCD, which rises from 50% in SCD patients > 50 years of age to more than 70% for those 70 years and older (Weiss, 2007). Biochemical analyses of corneas from SCD patients revealed a marked accumulation of cholesterol (McCarthy et al., 1994; Gaynor et al., 1996; Yamada et al., 1998), suggesting that dysregulation of cholesterol metabolism significantly contributes to the pathogenesis of SCD. Systemic hypercholesterolemia has been reported to be associated with some, but not all cases of the disease (Thiel et al., 1977; Brownstein et al., 1991; Crispin, 2002).

UBIAD1 belongs to the UbiA superfamily of integral membrane prenyltransferases that catalyze transfer of isoprenyl groups to aromatic acceptors, generating a wide variety of molecules ranging from ubiquinones, chlorophylls, and hemes to vitamin E and vitamin K (Li, 2016). UBIAD1 mediates transfer of the 20-carbon geranylgeranyl moiety from geranylgeranyl pyrophosphate (GGpp) to menadione (vitamin K₃) released from plant-derived phylloquinone (vitamin K₁), generating the vitamin K₂ subtype menaquinone-4 (MK-4) (Nakagawa et al., 2010; Hirota et al., 2013). To date, 25 missense mutations that alter 21 amino acids in the UBIAD1 protein have been identified in SCD families. Structural analyses of archaeal UbiA prenyltransferases revealed that residues corresponding to SCD-associated mutations in human UBIAD1 cluster around the active site of the enzyme (Cheng and Li, 2014; Huang et al., 2014). Indeed, all SCD-associated variants of UBIAD1 are defective in mediating synthesis of MK-4 (Hirota et al., 2015; Jun, D.-J. and DeBose-Boyd, R.A., unpublished observations) and for some variants, this defect likely results from reduced affinity for GGpp.

The first link between UBIAD1 and cholesterol metabolism was provided by the discovery of its association with the endoplasmic reticulum (ER)-localized enzyme 3-hydroxy-3-methylglutaryl coenzyme A reductase (HMGCR) (Nickerson et al., 2013). HMGCR catalyzes reduction of HMG CoA to mevalonate, a reaction that constitutes the rate-limiting step in synthesis of cholesterol and the essential nonsterol isoprenoids GGpp and farnesyl pyrophosphate (Fpp) (Goldstein and Brown, 1990; Wang and Casey, 2016). Fpp and GGpp can become transferred to many cellular proteins and are utilized in synthesis of other nonsterol isoprenoids including ubiquinone, heme, dolichol, and MK-4. HMGCR is subjected to tight feedback control through transcriptional and post-transcriptional mechanisms mediated by sterol and nonsterol isoprenoids (Brown and Goldstein, 1980). Sterols mediate the transcriptional effects by inhibiting proteolytic activation of membrane-bound transcription factors called sterol regulatory element-binding proteins (SREBPs). SREBPs enhance transcription of genes encoding HMGCR and other cholesterol biosynthetic enzymes as well as the low density lipoprotein (LDL) receptor that removes cholesterol-rich LDL from circulation (Horton et al., 2003). Post-transcriptional regulation of HMGCR is mediated by sterol and nonsterol isoprenoids, which combine to accelerate the ER-associated degradation (ERAD) of HMGCR through a reaction that involves the 26S proteasome (Nakanishi et al., 1988; Ravid et al., 2000). Together, these feedback regulatory mechanisms coordinate metabolism of mevalonate to assure cells maintain constant production of nonsterol isoprenoids but avoid overaccumulation of cholesterol and other sterols.

Our group discovered that accumulation of sterols in ER membranes triggers binding of HMGCR to ER membrane proteins called Insigs (Sever et al., 2003a; Sever et al., 2003b). Subsequent ubiquitination of HMGCR by Insig-associated ubiquitin ligases (Song et al., 2005; Jo et al., 2011; Jiang et al., 2018) mark the enzyme for extraction across ER membranes and release into the cytosol for proteasome-mediated ERAD (Elsabrouty et al., 2013; Morris et al., 2014). GGpp augments ERAD of ubiquitinated HMGCR by enhancing its membrane extraction (Elsabrouty et al., 2013). Recently, we discovered that sterols also cause a subset of HMGCR molecules to bind to UBIAD1 (32). This binding protects HMGCR from accelerated ERAD, permitting continued synthesis of nonsterol isoprenoids even when cellular sterols are abundant (Schumacher et al., 2018). GGpp triggers release of UBIAD1 from HMGCR, which allows for maximal ERAD of HMGCR and ER-to-Golgi transport of UBIAD1 (34). Eliminating expression of UBIAD1 relieves the GGpp requirement for HMGCR ERAD, indicating the reaction is inhibited by UBIAD1. Further characterization revealed that despite its steady-state Golgi localization in GGpp-replete cells, UBIAD1 continuously cycles between the ER and Golgi. Upon sensing depletion of GGpp in membranes of the ER, UBIAD1 becomes trapped in the organelle and inhibits ERAD of HMGCR to stimulate synthesis of mevalonate for replenishment of GGpp. The physiologic relevance of UBIAD1-mediated sensing of GGpp is highlighted by the observation that the reaction appears to be disrupted in SCD. SCD-associated UBIAD1 resists GGpp-induced release from HMGCR and becomes sequestered in the ER of GGpp-replete cells (Schumacher et al., 2015; Schumacher et al., 2016). The ensuing inhibition of HMGCR ERAD, which occurs in a dominant-negative fashion, leads to a marked increase in synthesis and intracellular accumulation of cholesterol (Schumacher et al., 2018).

The current studies were designed to confirm the role of UBIAD1 in regulation of HMGCR ERAD and cholesterol metabolism in living animals. For this purpose, we generated mice that harbor a knockin mutation that changes asparagine-100 (N100) in UBIAD1 to a serine residue (N100S) (see Figure 1A). The N100S mutation in mouse UBIAD1 corresponds to the SCD-associated UBIAD1 (N102S) mutation in the human enzyme. We show here that knockin mice homozygous for the N100S mutation (designated as Ubiad1^Ki/Ki mice) accumulate HMGCR protein in several tissues, despite a reduction in the amount of Hmgcr mRNA owing to sterol accumulation and reduced activation of SREBPs. The accumulation of HMGCR protein resulted from sequestration of UBIAD1 (N100S) in the ER and inhibition of HMGCR ERAD at a post-ubiquitination step of the reaction. Aged Ubiad1^Ki/Ki mice exhibited signs of opacification of the cornea, which was accompanied by hallmarks of sterol overaccumulation in the tissue. These findings not only indicate that UBIAD1 modulates ERAD of HMGCR in mice through similar mechanisms previously established in cultured cells, but they also establish Ubiad1^Ki/Ki mice as a model for human SCD.

Figure 1 with 1 supplement see all

Download asset Open asset

Ubiad1^WT/Ki heterozygous male and female mice (C57BL/6 × 129 genetic background) were crossed to obtain wild type (WT) and Ubiad1^Ki/Ki littermates. Mice homozygous for the N100S knockin mutation were born at expected Mendelian ratios. WT and Ubiad1^Ki/Ki littermates were externally indistinguishable and had similar body and liver weights (data not shown). Immunoblot analysis revealed that livers of male Ubiad1^WT/Ki and Ubiad1^Ki/Ki mice consuming chow diet ad libitum exhibited a noticeable increase (1.8- and 5.2-fold, respectively) in the amount of HMGCR protein compared to that in WT littermates (Figure 1B, lanes 1–3). However, the amount of Hmgcr mRNA was reduced approximately 40% in knockin mice (Figure 1—figure supplement 1A). UBIAD1 (N100S) protein also accumulated in livers of heterozygous and homozygous Ubiad1 knockin mice (Figure 1B, lanes 1–3); however, this was not accompanied by an increase in hepatic Ubiad1 mRNA (Figure 1—figure supplement 1A). Levels of nuclear SREBP-1 (Figure 1B, lanes 4–6) and SREBP-2 (lanes 7–9) were reduced in livers of Ubiad1^WT/Ki and Ubiad1^Ki/Ki mice, which coincided with reduced expression of mRNAs encoding SREBP target genes (Figure 1—figure supplement 1A). Cholesterol was slightly, but significantly increased in Ubiad1^Ki/Ki livers; however, plasma cholesterol, triglycerides, and non-esterified fatty acids as well as liver triglycerides were not significantly changed between the groups of animals (Figure 1—figure supplement 1B). Similar results were observed in the analysis of female Ubiad1^Ki/Ki mice (data not shown).

To ensure phenotypes associated with the N100S knockin mutation were not influenced by mixed genetic background, we backcrossed BL6/129 Ubiad1^Ki/Ki mice to C57BL/6J mice for at least six generations. For experiments described hereafter, Ubiad1^WT/Ki heterozygous female and male mice on the BL6 background were crossed to obtain WT and Ubiad1^Ki/Ki littermates. The results shown in Figure 2A reveal that male Ubiad1^WT/Ki and Ubiad1^Ki/Ki mice on the BL6 background accumulated hepatic HMGCR and UBIAD1 proteins (lanes 1–3), whereas levels of nuclear SREBP-1 and SREBP-2 were either unchanged (nuclear SREBP-1, lanes 4–6) or reduced (nuclear SREBP-2, lanes 7–9). HMGCR and UBIAD1 proteins accumulated and Hmgcr mRNA was down-regulated to varying degrees in other tissues of the knockin mice (Figure 2B and C). HMGCR and UBIAD1 protein accumulated to a similar extent in livers and eyes of female C57BL/6J Ubiad1^Ki/Ki mice (Figure 2—figure supplement 1).

Figure 2 with 1 supplement see all

Download asset Open asset

Table 1 shows that WT and Ubiad1^Ki/Ki mice had similar body and liver weights. Plasma levels of triglycerides, cholesterol, and non-esterified fatty acids (NEFAs) were slightly reduced in Ubiad1^Ki/Ki mice; however, these reductions were not significant. The knockin mice exhibited a small but significant increase in the amount of cholesterol in the liver (Table 1). We next used liquid chromatography-tandem mass spectrometry (LC-MS/MS) to measure levels of MK-4 and other nonsterol isoprenoids in livers of WT and Ubiad1^Ki/Ki mice (Figure 3). The results show that levels of MK-4 were reduced 50% relative to those observed in WT animals, despite a 4.2-fold increase in the amount of hepatic UBIAD1 protein. When normalized to the amount of UBIAD1 protein, we estimate the relative level of MK-4 was reduced by more than 80% in livers of Ubiad1^Ki/Ki mice. In contrast to results with MK-4, levels of geranylgeraniol (GGOH; derived from GGpp) and ubiquinone-10, were significantly increased in livers of Ubiad1^Ki/Ki mice. This was accompanied by a small, but significant decrease in plant-derived phylloquinone (vitamin K₁); bacterial-derived vitamin K₂ subtype menaquinone-7 (MK-7) remained unchanged in the knockin livers. MK-4 was reduced, whereas levels of GGOH and/or ubiquinone-10 were increased in kidneys, brains, spleens, and testes of Ubiad1^Ki/Ki mice (Figure 3—figure supplement 1).

Figure 3 with 1 supplement see all

Download asset Open asset

To directly study the regulation of HMGCR in the presence of UBIAD1 (N100S), we established lines of mouse embryonic fibroblasts (MEFs) from WT and Ubiad1^Ki/Ki littermates. A low level of HMGCR and UBIAD1 protein was detected in membrane fractions isolated from WT MEFs cultured in sterol and nonsterol isoprenoid-replete medium containing fetal calf serum (Figure 4A, lane 1). Both proteins markedly accumulated in MEFs derived from Ubiad1^Ki/Ki mice (lane 2). In contrast, levels of nuclear SREBP-1 and SREBP-2 were reduced in Ubiad1^Ki/Ki MEFs (Figure 4A, compare lanes 3 and 4), which is consistent with reduced levels of Hmgcr mRNA and increased levels of intracellular cholesterol (Figure 4B). We next compared sterol-accelerated ERAD of HMGCR in Ubiad1^Ki/Ki MEFs to that in MEFs derived from Hmgcr^Ki/Ki mice, which harbor knockin mutations in HMGCR that prevent its sterol-induced ubiquitination (Hwang et al., 2016). Cells were first depleted of isoprenoids through incubation in medium containing lipoprotein-deficient serum and the HMGCR inhibitor compactin to enhance expression of HMGCR. The cells were subsequently treated in the absence or presence of the oxysterol 25-hydroxycholesterol (25-HC) prior to harvest, subcellular fractionation, and immunoblot analysis. The results show that 25-HC caused the disappearance of HMGCR from membranes of WT MEFs as expected (Figure 4C, lanes 1 and 2; 5 and 6). However, this disappearance was blunted in membranes from either Ubiad1^Ki/Ki or Hmgcr^Ki/Ki MEFs (lanes 3 and 4; 7 and 8). Despite resistance of HMGCR to sterol-accelerated ERAD, the experiment of Figure 4D shows that sterols continued to stimulate HMGCR ubiquitination in Ubiad1^Ki/Ki MEFs. Isoprenoid-depleted cells were treated with the proteasome inhibitor MG-132 (to block degradation of ubiquitinated HMGCR) in the absence or presence of 25-HC. Cells were then harvested for preparation of detergent lysates that were immunoprecipitated with polyclonal anti-HMGCR. 25-HC caused HMGCR to become ubiquitinated in WT and Ubiad1^Ki/Ki MEFs, as indicated by smears of reactivity in anti-ubiquitin immunoblots of the HMGCR immunoprecipitates (Figure 4D, lanes 1–6). As expected, HMGCR resisted 25-HC-induced ubiquitination in MEFs derived from Hmgcr^Ki/Ki mice (compare lanes 7 and 8 with lanes 9–12).

Figure 4

Download asset Open asset

Figure 5A compares expression of HMGCR in WT and Ubiad1^Ki/Ki mice fed a chow diet supplemented with 1% cholesterol. The results show that cholesterol feeding led to reduced expression of HMGCR protein in membranes isolated from livers of WT mice (Figure 5A, lane 2); however, a significant amount of HMGCR protein remained in hepatic membranes of cholesterol-fed Ubiad1^Ki/Ki mice (lane 4). The feeding regimen reduced the amount of Insig-1 protein (lanes 2 and 4) and nuclear SREBP-2 (lanes 6 and 8) in both WT and Ubiad1^Ki/Ki livers. Dietary cholesterol also reduced mRNAs encoding HMGCR and other SREBP targets in livers of WT and Ubiad1^Ki/Ki mice (Figure 5—figure supplement 1). The membrane-bound precursor and nuclear forms of SREBP-1 were induced by cholesterol feeding in livers of both lines of mice (lanes 6 and 8). This induction can be attributed to sterol-mediated activation of liver x receptors (LXRs) that modulate expression of SREBP-1c, the major SREBP-1 isoform in the mouse liver (Repa et al., 2000; Liang et al., 2002). The mRNAs encoding SREBP-1c and two other LXR targets, ABCG5 and ABCG8, were enhanced in WT and Ubiad1^Ki/Ki mice fed cholesterol (Figure 5—figure supplement 1). In contrast to results in the liver, cholesterol-feeding failed to down-regulate levels of HMGCR protein in eyes of Ubiad1^Ki/Ki mice (Figure 5B, lanes 3 and 4); mRNAs encoding both SREBPs and their targets were also unchanged in eyes of cholesterol-fed knockin animals (data not shown).

Figure 5 with 1 supplement see all

Download asset Open asset

In Figure 5C, we compared cholesterol-mediated regulation of hepatic HMGCR in Ubiad1^Ki/Ki and Hmgcr^Ki/Ki mice. As little as 0.1% cholesterol caused a significant decrease in the amount of HMGCR in hepatic membranes from WT controls (Figure 5C, lanes b and j). HMGCR was partially resistant to 0.1% cholesterol in livers of Hmgcr^Ki/Ki mice (compare lanes e and f); however, higher concentrations of cholesterol (0.3% and 1%) caused complete disappearance of HMGCR from membranes (lanes g and h). The resistance of HMGCR to cholesterol feeding was more pronounced in Ubiad1^Ki/Ki mice; levels of the protein persisted when the animals were fed 0.1–1% cholesterol (Figure 5C, compare lane m with lanes n-p). Importantly, cholesterol-feeding continued to suppress levels of nuclear SREBP-2 in livers of Hmgcr^Ki/Ki and Ubiad1^Ki/Ki mice as well their WT littermates (compare lanes a-d with e-h and i-l with m-p). The mRNAs encoding SREBP-2 target genes were also reduced in livers of the cholesterol-fed mice (Figure 5D).

We next evaluated the effect of cholesterol depletion on HMGCR levels in WT and Ubiad1^Ki/Ki mice. Cholesterol depletion using lovastatin, a competitive inhibitor of HMGCR, led to the dose-dependent accumulation of HMGCR protein in livers of WT animals (Figure 6A, lanes a-d). Lovastatin also caused HMGCR to accumulate in livers of Ubiad1^Ki/Ki mice (lanes e-h) (see Figure 6B for quantification). The precursor and nuclear forms of SREBP-2 were induced by lovastatin, whereas those of SREBP-1 were reduced by the treatment in both WT and Ubiad1^Ki/Ki mice (lanes i-p). The mRNAs for SREBP-2 and its target genes (including HMGCR) were elevated in livers of lovastatin-treated animals; mRNA for SREBP-1c was reduced by the inhibitor (Figure 6—figure supplement 1). HMGCR protein was also increased in the eyes of lovastatin-treated WT mice; however, this increase required the highest concentration (0.2%) of the drug (Figure 6C, lanes a-d).

Figure 6 with 1 supplement see all

Download asset Open asset

In the experiment of Figure 6D, we analyzed the subcellular localization of UBIAD1 in WT and Ubiad1^Ki/Ki mice using a fractionation scheme previously utilized to isolate ER membranes from Chinese hamster ovary-K1 cells (Radhakrishnan et al., 2008). Liver homogenates (lysates) were first subjected to centrifugation at 3,000 X g to eliminate unbroken cells and nuclei. The resulting post-nuclear supernatants (PNS) were then applied to discontinuous sucrose gradients and centrifuged at 100,000 X g, generating two distinct membrane layers: a light membrane fraction enriched in Golgi and a heavy membrane fraction enriched in ER. Immunoblot analysis revealed the presence of UBIAD1 and the Golgi membrane protein GM-130 in the light, Golgi-enriched membrane fraction obtained from livers of WT mice fed a chow diet (Figure 6D, lane 4). ER-localized calnexin was observed in the ER-enriched fraction as expected (lane 5). When the mice were fed the chow diet supplemented with lovastatin (0.2%), we observed a shift in the localization of UBIAD1 from the Golgi-enriched fraction to the ER (Figure 6D, compare lanes 9 and 10). GM-130 remained in the Golgi-enriched fraction (lane 9) and calnexin continued to localize to the ER (lane 10) of livers from lovastatin-treated mice. In contrast to results with WT mice, UBIAD1 was concentrated in ER-enriched hepatic membranes of chow-fed Ubiad1^Ki/Ki mice (Figure 6D, compare lanes 4 and 5 with lanes 14 and 15). The ER localization of UBIAD1 did not change when the knockin mice were challenged with lovastatin (lanes 19 and 20). Importantly, calnexin and GM-130 were localized to ER- and Golgi-enriched membranes, respectively, regardless of feeding regimen (compare lanes 14 and 15 with lanes 19 and 20).

Stereomicroscopic examinations revealed that 8–12 week-old Ubiad1^Ki/Ki mice similar to those analyzed in Figures 2, 3, 5 and 6 failed to exhibit significant corneal opacification that characterizes human SCD (data not shown). However, 46% (11/24) of the knockin mice exhibited signs of corneal opacification at 50 weeks of age (Figure 7—figure supplement 1). One of these animals manifested signs of bilateral opacification of the cornea (Figure 7A). None of the aged WT mice developed corneal opacification; heterozygous knockin mice were not examined (data not shown). Immunohistochemical staining with anti-HMGCR revealed a marked increase in the amount of HMGCR protein in corneas of Ubiad1^Ki/Ki mice compared to their WT littermates (Figure 7B). The accumulation of HMGCR protein in Ubiad1^Ki/Ki corneas was accompanied by reduced levels of mRNAs encoding SREBP-2, HMGCR, and other cholesterol biosynthetic enzymes (Figure 7C). In contrast, expression of mRNAs encoding the LXR targets ABCG5, ABCG8, and ABCA1 was enhanced in Ubiad1^Ki/Ki corneas. Although the amount of total cholesterol remained unchanged in corneas of WT and Ubiad1^Ki/Ki mice, we measured a small, but significant increase in free cholesterol in the knockin mice (Figure 7D). Moreover, significant increases in the amount of several sterol intermediates of cholesterol synthesis including lanosterol, follicular fluid meiosis-activating sterol (FFMAS), 7- and 8-dehydrocholesterol, desmosterol, and 7-dehydrodesmosterol were observed in corneas from Ubiad1^Ki/Ki mice (Figure 7E).

Figure 7 with 1 supplement see all

Download asset Open asset

The current studies provide evidence that inhibition of HMGCR ERAD directly contributes to corneal sterol accumulation and opacification that characterizes the human eye disease SCD. This conclusion was drawn from the analysis of Ubiad1^Ki/Ki mice harboring a knockin mutation (N100S) that corresponds to the SCD-associated N102S mutation in human UBIAD1 (Figure 1A). Consistent with our studies of human UBIAD1 (N102S) in cultured cells (Schumacher et al., 2015; Schumacher et al., 2018), mouse UBIAD1 (N100S) inhibited ERAD of HMGCR in vivo as indicated by accumulation of the protein in livers and other tissues of Ubiad1^Ki/Ki mice (Figures 1B, 2 and 3, and Figure 2—figure supplement 1). These increases in HMGCR protein occurred despite reduced levels of its mRNA (Figures 1 and 2), which was attributable to reduced proteolytic activation of SREBP-2 (Figures 1B and 2A) resulting from accumulation of hepatic cholesterol (Figure 1—figure supplement 1B and Table 1). Corneas from 8 to 12 week-old Ubiad1^Ki/Ki mice failed to exhibit opacification of the cornea (data not shown), despite the marked accumulation of HMGCR protein in the eye (Figure 2B and C and Figure 2—figure supplement 1). Conversely, corneal opacification was observed in aged Ubiad1^Ki/Ki mice (50 weeks) (Figure 7A) that was accompanied by a buildup of HMGCR protein in the tissue (Figure 7B). This opacification occurred in the absence of an increase in total cholesterol (Figure 7D and E). However, Ubiad1^Ki/Ki corneas displayed other hallmarks of sterol overaccumulation including an increase in levels of free cholesterol and sterol intermediates of cholesterol biosynthesis (Figure 7D and E), reduced levels of mRNAs for HMGCR and other cholesterol biosynthetic enzymes, and enhanced expression of mRNAs encoding ABCG5, ABCG8, and ABCA1 (Figure 7C). Based on these observations, we conclude that UBIAD1 (N100S)-mediated inhibition of HMGCR ERAD enhances flux through the cholesterol biosynthetic pathway, thereby initiating sterol accumulation and development of corneal opacification. Our current findings suggest the slow progression of corneal opacification in Ubiad1^Ki/Ki mice results from reduced activation of SREBP-2, which leads to reduced expression of genes encoding cholesterol biosynthetic enzymes. Studies are now underway to monitor progression of corneal opacification in Ubiad1^Ki/Ki mice > 50 weeks of age. We anticipate that as the disorder progresses, free cholesterol will continue to accumulate, initiating formation and accumulation of cholesterol esters, which will further exacerbate opacification of the cornea. Additionally, Ubiad1^Ki/Ki mice will be challenged with high cholesterol/Western diets to determine whether dietary cholesterol contributes to progression of corneal opacification in SCD.

A preliminary analysis of UBIAD1 (N100S) knockin mice has been recently described (Dong et al., 2018). However, this study lacks the molecular characterization of HMGCR and documentation of age-dependent corneal opacification in the knockin animals. Instead, the authors report evidence for mitochondrial damage and accumulation of mitochondrial-localized glycerophosphoglycerols in knockin corneas. The authors conclude that mitochondrial dysfunction is linked to the abnormal deposition of cholesterol in corneas of SCD patients, constrasting our conclusion the response results from inhibition of HMGCR ERAD. The possibility exists that this mitochondrial dysfunction is secondary to cholesterol accumulation and/or impairment of MK-4 synthesis (see Figure 3) in Ubiad1^Ki/Ki mice. Further investigation is required to resolve this discrepancy.

Examination of Ubiad1^Ki/Ki-derived MEFs cultured under isoprenoid-replete conditions yielded results remarkably similar to those obtained with whole mouse tissues – marked accumulation of HMGCR protein, overproduction of cholesterol, reduced levels of Hmgcr mRNA, and reduced activation of SREBPs (Figure 4A and B). HMGCR was also refractory to accelerated ERAD stimulated by the oxysterol 25-HC in Ubiad1^Ki/Ki MEFs and Hmgcr^Ki/Ki MEFs, which were derived from mice expressing ubiquitination-resistant HMGCR (Hwang et al., 2016) (Figure 4C). However, 25-HC continued to stimulate ubiquitination of HMGCR in Ubiad1^Ki/Ki MEFs, but not in those derived from Hmgcr^Ki/Ki mice (Figure 4D). In addition to providing direct evidence that SCD-associated UBIAD1 inhibits ERAD of HMGCR, these results indicate that the inhibition results from a block in a post-ubiquitination step of the reaction (Schumacher et al., 2015).

Our previous characterization of Hmgcr^Ki/Ki mice indicated that dietary cholesterol reduces levels of HMGCR protein primarily by inhibiting activation of SREBP-2 (35); similar results were obtained in the current study (Figure 5C). In contrast, levels of HMGCR protein persisted in livers of Ubiad1^Ki/Ki mice fed cholesterol, even though the feeding regimen continued to block proteolytic activation of SREBP-2 (Figure 5A and C) and reduce levels of mRNAs for HMGCR and other SREBP-2 targets (Figure 5—figure supplement 1). Previous studies found that HMGCR is subjected to sterol-independent ubiquitination and ERAD in cultured cells (Doolman et al., 2004). Thus, we speculate that in livers of Hmgcr^Ki/Ki mice, ubiquitination-resistant HMGCR continues to become degraded through this sterol-independent ERAD pathway. However, both sterol-independent and sterol-dependent pathways for HMGCR ERAD are blocked in Ubiad1^Ki/Ki mice, owing to the ability of UBIAD1 (N100S) to inhibit post-ubiquitination steps of the reaction (see Figure 4D). As a result of this inhibition, levels of HMGCR protein persist in livers of cholesterol-fed Ubiad1^Ki/Ki mice.

Okano and co-workers previously attempted to generate mice lacking Ubiad1; however, Ubiad1-deficient embryos failed to survive past embryonic day 7.5 (40). Ubiad1^-/- embryonic stem cells failed to synthesize MK-4, suggesting the compound plays a pivotal role in development. Taking this and the observation that human UBIAD1 (N102S) is defective in MK-4 synthetic activity (Hirota et al., 2015), we were somewhat surprised that Ubiad1^Ki/Ki mice were born at normal Mendelian ratios and appear grossly normal. Throughout our studies, we noticed that UBIAD1 (N100S) protein accumulated in all tissues of Ubiad1^Ki/Ki mice (see Figures 1–3). Similar results have been observed for SCD-associated variants of human UBIAD1 in transfected cells; stabilization of SCD-associated UBIAD1 has been attributed to ER sequestration and protection from autophagic degradation from the Golgi (Jun, D.-J. and DeBose-Boyd, R.A., unpublished observations). The level of MK-4 was significantly reduced, but not absent, in various tissues of Ubiad1^Ki/Ki mice (Figure 3 and Figure 3—figure supplement 1). We speculate that despite reduced enzymatic activity, accumulated UBIAD1 (N100S) produces sufficient amounts of MK-4 to support development and survival of Ubiad1^Ki/Ki mice. Accumulation of SCD-associated UBIAD1 in the ER has important implications for the pathology of the disease. SCD is an autosomal dominant disorder and our previous studies revealed that SCD-associated variants of UBIAD1 inhibit HMGCR ERAD in a dominant-negative fashion (Schumacher et al., 2015; Schumacher et al., 2016). Enhanced stability of SCD-associated UBIAD1 owing to its ER sequestration helps to explain how cholesterol accumulation occurs in corneas of SCD patients harboring heterozygous UBIAD1 mutations.

The significance of the HMGCR regulatory system is evidenced by the widespread use of statins to lower plasma LDL-cholesterol and reduce the incidence of atherosclerotic cardiovascular disease (Stossel, 2008). However, statins block synthesis of sterol and nonsterol isoprenoids that mediate feedback regulation of HMGCR, causing the enzyme’s accumulation in livers of humans and animals (Kita et al., 1980; Reihnér et al., 1990) (Figure 6). This accumulation partially overcomes inhibitory effects of statins, which allows for continued synthesis of cholesterol that limits cholesterol-lowering (Schonewille et al., 2016; Goldberg et al., 1990; Engelking et al., 2006). We showed previously that the statin-induced increase in HMGCR was blunted 5-fold in Hmgcr^Ki/Ki versus WT mice, suggesting inhibition of ERAD significantly contributes to the response (Hwang et al., 2016). Our current findings argue that statin-induced accumulation of HMGCR results in part, from depletion of GGpp from ER membranes. This was indicated by accumulation and ER sequestration of UBIAD1 in livers of lovastatin-treated mice, which coincided with accumulation of HMGCR (Figure 6A and C). UBIAD1 (N100S), which resists GGpp-induced release from HMGCR, accumulated and remained sequestered in the ER, regardless of whether Ubiad1^Ki/Ki mice were treated in the absence or presence of lovastatin (Figure 6A and C). The resultant inhibition of HMGCR ERAD led to accumulation of the protein and overproduction of sterol and nonsterol isoprenoids in various tissues of Ubiad1^Ki/Ki mice (see Figures 3 and 7, and Figure 3—figure supplement 1). Thus, GGpp-regulated, ER-to-Golgi transport of UBIAD1 regulates HMGCR ERAD to coordinate synthesis of sterol and nonsterol isoprenoids in mice through similar mechanisms previously described in cultured cells (Schumacher et al., 2018). The current results indicate that modulating ER-to-Golgi transport of UBIAD1 may have therapeutic value. For example, agents that mimic GGpp in stimulating ER-to-Golgi transport of UBIAD1 should relieve inhibition of ERAD and prevent accumulation of HMGCR that limits the effectiveness of statins in lowering plasma levels of LDL-cholesterol. Agents that restore Golgi localization of SCD-associated UBIAD1 or block its interaction with HMGCR may prevent or retard cholesterol accumulation and corneal opacification associated with SCD. The establishment of a mouse model for SCD described here will prove instrumental in the identification and characterization of such molecules.

All data generated in this study are included in the manuscript.

1. 1. Brown MS
   2. Goldstein JL

   (1980)

   Multivalent feedback regulation of HMG CoA reductase, a control mechanism coordinating isoprenoid synthesis and cell growth

   Journal of Lipid Research 21:505–517.

   • PubMed
   • Google Scholar
2. 1. Brownstein S
   2. Jackson WB
   3. Onerheim RM

   (1991)

   Schnyder's crystalline corneal dystrophy in association with hyperlipoproteinemia: histopathological and ultrastructural findings

   Canadian Journal of Ophthalmology. Journal Canadien d'ophtalmologie 26:273–279.

   • PubMed
   • Google Scholar
3. 1. Cautivo KM
   2. Lizama CO
   3. Tapia PJ
   4. Agarwal AK
   5. Garg A
   6. Horton JD
   7. Cortés VA

   (2016) AGPAT2 is essential for postnatal development and maintenance of white and Brown adipose tissue

   Molecular Metabolism 5:491–505.

   https://doi.org/10.1016/j.molmet.2016.05.004

   • PubMed
   • Google Scholar
4. 1. Cheng W
   2. Li W

   (2014) Structural insights into ubiquinone biosynthesis in membranes

   Science 343:878–881.

   https://doi.org/10.1126/science.1246774

   • PubMed
   • Google Scholar
5. 1. Crispin S

   (2002) Ocular lipid deposition and hyperlipoproteinaemia

   Progress in Retinal and Eye Research 21:169–224.

   https://doi.org/10.1016/S1350-9462(02)00004-6

   • PubMed
   • Google Scholar
6. 1. Dong F
   2. Jin X
   3. Boettler MA
   4. Sciulli H
   5. Abu-Asab M
   6. Del Greco C
   7. Wang S
   8. Hu Y-C
   9. Campos MM
   10. Jackson SN
   11. Muller L
   12. Woods AS
   13. Combs CA
   14. Zhang J
   15. Nickerson ML
   16. Kruth HS
   17. Weiss JS
   18. Kao WW

   (2018) A mouse model of Schnyder corneal dystrophy with the N100S point mutation

   Scientific Reports 8:e10219.

   https://doi.org/10.1038/s41598-018-28545-0

   • Google Scholar
7. 1. Doolman R
   2. Leichner GS
   3. Avner R
   4. Roitelman J

   (2004) Ubiquitin is conjugated by membrane ubiquitin ligase to three sites, including the N Terminus, in transmembrane region of mammalian 3-hydroxy-3-methylglutaryl coenzyme A reductase: implications for sterol-regulated enzyme degradation

   Journal of Biological Chemistry 279:38184–38193.

   https://doi.org/10.1074/jbc.M405935200

   • PubMed
   • Google Scholar
8. 1. Elsabrouty R
   2. Jo Y
   3. Dinh TT
   4. DeBose-Boyd RA

   (2013) Sterol-induced dislocation of 3-hydroxy-3-methylglutaryl coenzyme A reductase from membranes of permeabilized cells

   Molecular Biology of the Cell 24:3300–3308.

   https://doi.org/10.1091/mbc.e13-03-0157

   • PubMed
   • Google Scholar
9. 1. Engelking LJ
   2. Liang G
   3. Hammer RE
   4. Takaishi K
   5. Kuriyama H
   6. Evers BM
   7. Li WP
   8. Horton JD
   9. Goldstein JL
   10. Brown MS

   (2005) Schoenheimer effect explained--feedback regulation of cholesterol synthesis in mice mediated by insig proteins

   Journal of Clinical Investigation 115:2489–2498.

   https://doi.org/10.1172/JCI25614

   • PubMed
   • Google Scholar
10. 1. Engelking LJ
    2. Evers BM
    3. Richardson JA
    4. Goldstein JL
    5. Brown MS
    6. Liang G

    (2006) Severe facial clefting in Insig-deficient mouse embryos caused by sterol accumulation and reversed by lovastatin

    Journal of Clinical Investigation 116:2356–2365.

    https://doi.org/10.1172/JCI28988

    • PubMed
    • Google Scholar
11. 1. Evers BM
    2. Farooqi MS
    3. Shelton JM
    4. Richardson JA
    5. Goldstein JL
    6. Brown MS
    7. Liang G

    (2010) Hair growth defects in Insig-deficient mice caused by cholesterol precursor accumulation and reversed by simvastatin

    Journal of Investigative Dermatology 130:1237–1248.

    https://doi.org/10.1038/jid.2009.442

    • PubMed
    • Google Scholar
12. 1. Gaynor PM
    2. Zhang WY
    3. Weiss JS
    4. Skarlatos SI
    5. Rodrigues MM
    6. Kruth HS

    (1996) Accumulation of HDL apolipoproteins accompanies abnormal cholesterol accumulation in Schnyder's corneal dystrophy

    Arteriosclerosis, Thrombosis, and Vascular Biology 16:992–999.

    https://doi.org/10.1161/01.ATV.16.8.992

    • PubMed
    • Google Scholar
13. 1. Goldberg IJ
    2. Holleran S
    3. Ramakrishnan R
    4. Adams M
    5. Palmer RH
    6. Dell RB
    7. Goodman DS

    (1990) Lack of effect of lovastatin therapy on the parameters of whole-body cholesterol metabolism

    Journal of Clinical Investigation 86:801–808.

    https://doi.org/10.1172/JCI114777

    • PubMed
    • Google Scholar
14. 1. Goldstein JL
    2. Brown MS

    (1990) Regulation of the mevalonate pathway

    Nature 343:425–430.

    https://doi.org/10.1038/343425a0

    • PubMed
    • Google Scholar
15. 1. Hirota Y
    2. Tsugawa N
    3. Nakagawa K
    4. Suhara Y
    5. Tanaka K
    6. Uchino Y
    7. Takeuchi A
    8. Sawada N
    9. Kamao M
    10. Wada A
    11. Okitsu T
    12. Okano T

    (2013) Menadione (vitamin K3) is a catabolic product of oral phylloquinone (vitamin K1) in the intestine and a circulating precursor of tissue menaquinone-4 (vitamin K2) in rats

    The Journal of Biological Chemistry 288:33071–33080.

    https://doi.org/10.1074/jbc.M113.477356

    • PubMed
    • Google Scholar
16. 1. Hirota Y
    2. Nakagawa K
    3. Sawada N
    4. Okuda N
    5. Suhara Y
    6. Uchino Y
    7. Kimoto T
    8. Funahashi N
    9. Kamao M
    10. Tsugawa N
    11. Okano T

    (2015) Functional characterization of the vitamin K2 biosynthetic enzyme UBIAD1

    PLOS ONE 10:e0125737.

    https://doi.org/10.1371/journal.pone.0125737

    • PubMed
    • Google Scholar
17. 1. Horton JD
    2. Shah NA
    3. Warrington JA
    4. Anderson NN
    5. Park SW
    6. Brown MS
    7. Goldstein JL

    (2003) Combined analysis of oligonucleotide microarray data from transgenic and knockout mice identifies direct SREBP target genes

    PNAS 100:12027–12032.

    https://doi.org/10.1073/pnas.1534923100

    • PubMed
    • Google Scholar
18. 1. Huang H
    2. Levin EJ
    3. Liu S
    4. Bai Y
    5. Lockless SW
    6. Zhou M

    (2014) Structure of a membrane-embedded prenyltransferase homologous to UBIAD1

    PLOS Biology 12:e1001911.

    https://doi.org/10.1371/journal.pbio.1001911

    • PubMed
    • Google Scholar
19. 1. Hwang S
    2. Hartman IZ
    3. Calhoun LN
    4. Garland K
    5. Young GA
    6. Mitsche MA
    7. McDonald J
    8. Xu F
    9. Engelking L
    10. DeBose-Boyd RA

    (2016) Contribution of accelerated degradation to feedback regulation of 3-Hydroxy-3-methylglutaryl coenzyme A reductase and cholesterol metabolism in the liver

    Journal of Biological Chemistry 291:13479–13494.

    https://doi.org/10.1074/jbc.M116.728469

    • PubMed
    • Google Scholar
20. 1. Jiang LY
    2. Jiang W
    3. Tian N
    4. Xiong YN
    5. Liu J
    6. Wei J
    7. Wu KY
    8. Luo J
    9. Shi XJ
    10. Song BL

    (2018) Ring finger protein 145 (RNF145) is a ubiquitin ligase for sterol-induced degradation of HMG-CoA reductase

    Journal of Biological Chemistry 293:4047–4055.

    https://doi.org/10.1074/jbc.RA117.001260

    • PubMed
    • Google Scholar
21. 1. Jo Y
    2. Lee PC
    3. Sguigna PV
    4. DeBose-Boyd RA

    (2011) Sterol-induced degradation of HMG CoA reductase depends on interplay of two insigs and two ubiquitin ligases, gp78 and Trc8

    PNAS 108:20503–20508.

    https://doi.org/10.1073/pnas.1112831108

    • PubMed
    • Google Scholar
22. 1. Jozefczuk J
    2. Drews K
    3. Adjaye J

    (2012) Preparation of mouse embryonic fibroblast cells suitable for culturing human embryonic and induced pluripotent stem cells

    Journal of Visualized Experiments 21:e3854.

    https://doi.org/10.3791/3854

    • Google Scholar
23. 1. Kita T
    2. Brown MS
    3. Goldstein JL

    (1980) Feedback regulation of 3-hydroxy-3-methylglutaryl coenzyme A reductase in livers of mice treated with mevinolin, a competitive inhibitor of the reductase

    Journal of Clinical Investigation 66:1094–1100.

    https://doi.org/10.1172/JCI109938

    • PubMed
    • Google Scholar
24. 1. Klintworth GK

    (2009) Corneal dystrophies

    Orphanet Journal of Rare Diseases 4:7.

    https://doi.org/10.1186/1750-1172-4-7

    • PubMed
    • Google Scholar
25. 1. Li W

    (2016) Bringing bioactive compounds into membranes: the UbiA superfamily of intramembrane aromatic prenyltransferases

    Trends in Biochemical Sciences 41:356–370.

    https://doi.org/10.1016/j.tibs.2016.01.007

    • PubMed
    • Google Scholar
26. 1. Liang G
    2. Yang J
    3. Horton JD
    4. Hammer RE
    5. Goldstein JL
    6. Brown MS

    (2002) Diminished hepatic response to fasting/refeeding and liver X receptor agonists in mice with selective deficiency of sterol regulatory element-binding protein-1c

    Journal of Biological Chemistry 277:9520–9528.

    https://doi.org/10.1074/jbc.M111421200

    • PubMed
    • Google Scholar
27. 1. McCarthy M
    2. Innis S
    3. Dubord P
    4. White V

    (1994) Panstromal schnyder corneal dystrophy. A clinical pathologic report with quantitative analysis of corneal lipid composition

    Ophthalmology 101:895–901.

    https://doi.org/10.1016/S0161-6420(94)31241-3

    • PubMed
    • Google Scholar
28. 1. McDonald JG
    2. Smith DD
    3. Stiles AR
    4. Russell DW

    (2012) A comprehensive method for extraction and quantitative analysis of sterols and secosteroids from human plasma

    Journal of Lipid Research 53:1399–1409.

    https://doi.org/10.1194/jlr.D022285

    • PubMed
    • Google Scholar
29. 1. McFarlane MR
    2. Liang G
    3. Engelking LJ

    (2014) Insig proteins mediate feedback inhibition of cholesterol synthesis in the intestine

    Journal of Biological Chemistry 289:2148–2156.

    https://doi.org/10.1074/jbc.M113.524041

    • PubMed
    • Google Scholar
30. 1. Mitsche MA
    2. McDonald JG
    3. Hobbs HH
    4. Cohen JC

    (2015) Flux analysis of cholesterol biosynthesis in vivo reveals multiple tissue and cell-type specific pathways

    eLife 4:e07999.

    https://doi.org/10.7554/eLife.07999

    • PubMed
    • Google Scholar
31. 1. Morris LL
    2. Hartman IZ
    3. Jun DJ
    4. Seemann J
    5. DeBose-Boyd RA

    (2014) Sequential actions of the AAA-ATPase valosin-containing protein (VCP)/p97 and the proteasome 19 S regulatory particle in sterol-accelerated, endoplasmic reticulum (ER)-associated degradation of 3-hydroxy-3-methylglutaryl-coenzyme A reductase

    Journal of Biological Chemistry 289:19053–19066.

    https://doi.org/10.1074/jbc.M114.576652

    • PubMed
    • Google Scholar
32. 1. Nakagawa K
    2. Hirota Y
    3. Sawada N
    4. Yuge N
    5. Watanabe M
    6. Uchino Y
    7. Okuda N
    8. Shimomura Y
    9. Suhara Y
    10. Okano T

    (2010) Identification of UBIAD1 as a novel human menaquinone-4 biosynthetic enzyme

    Nature 468:117–121.

    https://doi.org/10.1038/nature09464

    • PubMed
    • Google Scholar
33. 1. Nakagawa K
    2. Sawada N
    3. Hirota Y
    4. Uchino Y
    5. Suhara Y
    6. Hasegawa T
    7. Amizuka N
    8. Okamoto T
    9. Tsugawa N
    10. Kamao M
    11. Funahashi N
    12. Okano T

    (2014) Vitamin K2 biosynthetic enzyme, UBIAD1 is essential for embryonic development of mice

    PLOS ONE 9:e104078.

    https://doi.org/10.1371/journal.pone.0104078

    • PubMed
    • Google Scholar
34. 1. Nakanishi M
    2. Goldstein JL
    3. Brown MS

    (1988)

    Multivalent control of 3-hydroxy-3-methylglutaryl coenzyme A reductase. Mevalonate-derived product inhibits translation of mRNA and accelerates degradation of enzyme

    The Journal of Biological Chemistry 263:8929–8937.

    • PubMed
    • Google Scholar
35. 1. Nickerson ML
    2. Bosley AD
    3. Weiss JS
    4. Kostiha BN
    5. Hirota Y
    6. Brandt W
    7. Esposito D
    8. Kinoshita S
    9. Wessjohann L
    10. Morham SG
    11. Andresson T
    12. Kruth HS
    13. Okano T
    14. Dean M

    (2013) The UBIAD1 prenyltransferase links menaquinone-4 [corrected] synthesis to cholesterol metabolic enzymes

    Human Mutation 34:317–329.

    https://doi.org/10.1002/humu.22230

    • PubMed
    • Google Scholar
36. 1. Orr A
    2. Dubé MP
    3. Marcadier J
    4. Jiang H
    5. Federico A
    6. George S
    7. Seamone C
    8. Andrews D
    9. Dubord P
    10. Holland S
    11. Provost S
    12. Mongrain V
    13. Evans S
    14. Higgins B
    15. Bowman S
    16. Guernsey D
    17. Samuels M

    (2007) Mutations in the UBIAD1 gene, encoding a potential prenyltransferase, are causal for Schnyder crystalline corneal dystrophy

    PLOS ONE 2:e685.

    https://doi.org/10.1371/journal.pone.0000685

    • PubMed
    • Google Scholar
37. 1. Radhakrishnan A
    2. Goldstein JL
    3. McDonald JG
    4. Brown MS

    (2008) Switch-like control of SREBP-2 transport triggered by small changes in ER cholesterol: a delicate balance

    Cell Metabolism 8:512–521.

    https://doi.org/10.1016/j.cmet.2008.10.008

    • PubMed
    • Google Scholar
38. 1. Ravid T
    2. Doolman R
    3. Avner R
    4. Harats D
    5. Roitelman J

    (2000) The ubiquitin-proteasome pathway mediates the regulated degradation of mammalian 3-hydroxy-3-methylglutaryl-coenzyme A reductase

    Journal of Biological Chemistry 275:35840–35847.

    https://doi.org/10.1074/jbc.M004793200

    • PubMed
    • Google Scholar
39. 1. Reihnér E
    2. Rudling M
    3. Ståhlberg D
    4. Berglund L
    5. Ewerth S
    6. Björkhem I
    7. Einarsson K
    8. Angelin B

    (1990) Influence of pravastatin, a specific inhibitor of HMG-CoA reductase, on hepatic metabolism of cholesterol

    New England Journal of Medicine 323:224–228.

    https://doi.org/10.1056/NEJM199007263230403

    • PubMed
    • Google Scholar
40. 1. Repa JJ
    2. Liang G
    3. Ou J
    4. Bashmakov Y
    5. Lobaccaro JM
    6. Shimomura I
    7. Shan B
    8. Brown MS
    9. Goldstein JL
    10. Mangelsdorf DJ

    (2000) Regulation of mouse sterol regulatory element-binding protein-1c gene (SREBP-1c) by oxysterol receptors, LXRalpha and LXRbeta

    Genes & Development 14:2819–2830.

    https://doi.org/10.1101/gad.844900

    • PubMed
    • Google Scholar
41. 1. Schonewille M
    2. de Boer JF
    3. Mele L
    4. Wolters H
    5. Bloks VW
    6. Wolters JC
    7. Kuivenhoven JA
    8. Tietge UJ
    9. Brufau G
    10. Groen AK

    (2016) Statins increase hepatic cholesterol synthesis and stimulate fecal cholesterol elimination in mice

    Journal of Lipid Research 57:1455–1464.

    https://doi.org/10.1194/jlr.M067488

    • PubMed
    • Google Scholar
42. 1. Schumacher MM
    2. Elsabrouty R
    3. Seemann J
    4. Jo Y
    5. DeBose-Boyd RA

    (2015) The prenyltransferase UBIAD1 is the target of geranylgeraniol in degradation of HMG CoA reductase

    eLife 4:e05560.

    https://doi.org/10.7554/eLife.05560

    • Google Scholar
43. 1. Schumacher MM
    2. Jun DJ
    3. Jo Y
    4. Seemann J
    5. DeBose-Boyd RA

    (2016) Geranylgeranyl-regulated transport of the prenyltransferase UBIAD1 between membranes of the ER and golgi

    Journal of Lipid Research 57:1286–1299.

    https://doi.org/10.1194/jlr.M068759

    • PubMed
    • Google Scholar
44. 1. Schumacher MM
    2. Jun DJ
    3. Johnson BM
    4. DeBose-Boyd RA

    (2018) UbiA prenyltransferase domain-containing protein-1 modulates HMG-CoA reductase degradation to coordinate synthesis of sterol and nonsterol isoprenoids

    Journal of Biological Chemistry 293:312–323.

    https://doi.org/10.1074/jbc.RA117.000423

    • PubMed
    • Google Scholar
45. 1. Sever N
    2. Yang T
    3. Brown MS
    4. Goldstein JL
    5. DeBose-Boyd RA

    (2003a) Accelerated degradation of HMG CoA reductase mediated by binding of insig-1 to its sterol-sensing domain

    Molecular Cell 11:25–33.

    https://doi.org/10.1016/S1097-2765(02)00822-5

    • PubMed
    • Google Scholar
46. 1. Sever N
    2. Song BL
    3. Yabe D
    4. Goldstein JL
    5. Brown MS
    6. DeBose-Boyd RA

    (2003b) Insig-dependent ubiquitination and degradation of mammalian 3-hydroxy-3-methylglutaryl-CoA reductase stimulated by sterols and geranylgeraniol

    Journal of Biological Chemistry 278:52479–52490.

    https://doi.org/10.1074/jbc.M310053200

    • PubMed
    • Google Scholar
47. 1. Song BL
    2. Sever N
    3. DeBose-Boyd RA

    (2005) Gp78, a membrane-anchored ubiquitin ligase, associates with Insig-1 and couples sterol-regulated ubiquitination to degradation of HMG CoA reductase

    Molecular Cell 19:829–840.

    https://doi.org/10.1016/j.molcel.2005.08.009

    • PubMed
    • Google Scholar
48. 1. Stossel TP

    (2008) The discovery of statins

    Cell 134:903–905.

    https://doi.org/10.1016/j.cell.2008.09.008

    • PubMed
    • Google Scholar
49. 1. Thiel HJ
    2. Voigt GJ
    3. Parwaresch MR

    (1977)

    [Crystalline corneal dystrophy (Schnyder) in the presence of familial type IIa hyperlipoproteinaemia (author's transl)]

    Klinische Monatsblatter Fur Augenheilkunde 171:678–684.

    • PubMed
    • Google Scholar
50. 1. Wang M
    2. Casey PJ

    (2016) Protein prenylation: unique fats make their mark on biology

    Nature Reviews Molecular Cell Biology 17:110–122.

    https://doi.org/10.1038/nrm.2015.11

    • PubMed
    • Google Scholar
51. 1. Weiss JS
    2. Kruth HS
    3. Kuivaniemi H
    4. Tromp G
    5. White PS
    6. Winters RS
    7. Lisch W
    8. Henn W
    9. Denninger E
    10. Krause M
    11. Wasson P
    12. Ebenezer N
    13. Mahurkar S
    14. Nickerson ML

    (2007) Mutations in the UBIAD1 gene on chromosome short arm 1, region 36, cause schnyder crystalline corneal dystrophy

    Investigative Opthalmology & Visual Science 48:5007–5012.

    https://doi.org/10.1167/iovs.07-0845

    • Google Scholar
52. 1. Weiss JS

    (2007)

    Visual morbidity in thirty-four families with Schnyder crystalline corneal dystrophy (an american ophthalmological society thesis)

    Transactions of the American Ophthalmological Society 105:616–648.

    • PubMed
    • Google Scholar
53. 1. Weiss JS

    (2009) Schnyder corneal dystrophy

    Current Opinion in Ophthalmology 20:292–298.

    https://doi.org/10.1097/ICU.0b013e32832b753e

    • PubMed
    • Google Scholar
54. 1. Yamada M
    2. Mochizuki H
    3. Kamata Y
    4. Nakamura Y
    5. Mashima Y

    (1998) Quantitative analysis of lipid deposits from Schnyder's corneal dystrophy

    British Journal of Ophthalmology 82:444–447.

    https://doi.org/10.1136/bjo.82.4.444

    • PubMed
    • Google Scholar

Thank you for submitting your article "Schnyder corneal dystrophy-associated UBIAD1 Inhibits ER-associated degradation of HMG CoA reductase in mice" for consideration by eLife. Your article has been reviewed by three peer reviewers, one of whom is a member of our Board of Reviewing Editors, and the evaluation has been overseen by Randy Schekman as the Senior Editor. The following individual involved in review of your submission has agreed to reveal his identity: Randy Y Hampton (Reviewer #3).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

The referees all agreed that the quality of the data is high and that the generation and initial characterization of a mouse model for SCD warrants publication as a Research Advance linked to your prior paper. They further judged that the study provides important in vivo support for prior mechanistic models on the link between UBIAD1, HMG CoA reductase, and cholesterol homeostasis established from cell culture. One referee raised a number of potential additional mechanistic links that might be important to explore. During our discussion, we decided that these extensions of the project are beyond the scope of the current paper. Although we have appended the full referee comments for your interest, we expect you to address only the minor suggestions for improvement provided below. anticipate that addressing these comments will probably only need clarifications to the text, perhaps some re-organisation of figure panels, and possibly answering queries on unclear points.

Reviewer #1:

This Research Advance contribution from the Debose-Boyd lab follows up an earlier study from this group analysing the function of UBIAD1 as it relates to HMGCR regulation. The earlier study reported that UBIAD1 stimulates HMGCR degradation in a ppGG-dependent manner, and that dominant human SCD disease mutants in UBIAD1 interfere with full HMGCR degradation under cholesterol-replete conditions. The hypothesis from the earlier work was that the disease mutants may work by causing elevated HMGCR, resulting in elevated cholesterol which eventually accumulates in the eye to cause corneal opacities. The current study examines this idea in mice by generating a knock-in of one such SCD disease mutant (N100S). Analysis of the mouse reveals clear evidence of dysregulated HMGCR including its elevated steady state levels, various compensatory sequelae, and elevated levels of sterol-related molecules. Importantly, the mouse appears to also show age-dependent opacities in the eye. Analysis of MEFs from the mutant mice verified altered cholesterol-mediated HMGCR degradation. Considered together with their earlier work, the study makes a strong case that the cholesterol-containing deposits observed in SCD result from altered HMGCR regulation caused by its excessively strong interaction with mutant UBIAD1. It is important because it provides a nice mouse model for SCD, and more speculatively, provides additional in vivo insights into the intricacies of cholesterol homeostasis. The paper is generally well organized and written.

1) As I am sure the authors are aware, another group recently published a preliminary analysis of a N100S UBIAD1 knock-in mouse (PMID 29977031). This earlier study lacks the molecular analysis of HMGCR, age-dependent phenotypes, and mechanistic interpretations of the current work. Nevertheless, it is important that the authors provide a careful discussion of these earlier findings and how they relate to the current study.

2) In Figure 1B, it would be useful to show some indicator of equal loading (perhaps the stained blot showing total protein or a blot for an unrelated protein). In the graph, it would be useful to have error bars on the protein levels if available. I realise the next figure provides a more thorough analysis with suitable controls, but a reader is immediately faced with a rather sparse and somewhat incomplete observation right at the outset. Perhaps one option is to simply omit Figure 1 and start with Figure 2 since this is better characterized anyway.

3) Figure 1B would be easier to follow in the text and legend if the three sub-panels were labelled as panels B, C, and D. This would make it easier to call out the individual experiments. The same applies to other figure panels where more than one type of experiment is shown together (e.g., the top and bottom parts of Figure 2A. In general, it is easier for the reader if each panel displays a single experiment.

4) The increase in UBIAD1 in Ki livers (and other tissues) is based on increased levels in the membrane fraction. Is it possible that the total levels are the same, and the difference is due to efficiency of recovery of certain subcellular membranes? Figure 6 shows that mutant UBIAD1 is preferentially in the ER whereas wild type UBIAD1 is preferentially in the Golgi. So if the ER is more effectively recovered than Golgi in the membrane fraction, then one might get the impression that total UBIAD1 is increased. Can a blot also be done on total lysate to clarify this issue? Total lysate is shown in Figure 6C, but it is unclear if the two blots are directly comparable to infer anything about total UBIAD1 levels.

5) I think it is more useful for a reader to directly state the absolute level of increase in HMGCR rather than the value after normalization to mRNA levels. The absolute level of protein is what is important for its enzymatic activity, which is the key parameter for the disease (i.e., the amount of cholesterol that accumulates) that forms the focus of this paper. The normalized values inform more about the relative rates of synthesis/degradation, which is also interesting but less central to the phenotype. Thus, one could simply report total values in the figures and Results section, then in the discussion state how the levels are increased despite much lower mRNA levels, suggesting that either synthesis rate is higher and/or degradation rate is lower.

6) Results section – It states that the Lovastatin effect on HMGCR is blunted in Ki mice liver relative to wild type, but the result in Figure 6A shows that accumulation of HMGCR is enhanced in Ki mice liver. Perhaps the sentence is worded incorrectly?

7) Results section – It says Lovastatin led to a slight elevation of HMGCR in Ki mice; However, it was not easy for me to judge from the blot if the increase was any more or less than in wild type mice. Perhaps it is useful to quantify the result and be more precise in the wording with something like "led to XX-fold elevation above the already high levels of HMGCR in Ki mice…" to emphasize that the levels at baseline are already very high.

8) It is worth stating how frequently (if at all) corneal inclusions are observed in wild type and heterozygous mice of similar age. This is important to make sure that the inclusions are not simply a consequence of old age.

Reviewer #2:

This manuscript follows up on earlier work from the deBose-Boyd laboratory on the regulation of HMG-CoA-reductase (HMGCR) by UBIAD1, a protein mutated in Schnyder Corneal Dystrophy (SCD) patients. Here a SCD-associated UBIAD1 mutation (N100S) knock-in mouse (Ubiad1ki/ki) is described. It is convincingly shown that Ubiad1ki/ki mouse recapitulates several aspects of SCD pathology, including corneal opacification in aged mice. Analysis of tissues and mouse embryonic fibroblast (MEF) derived from the knock-in mouse largely confirm UBIAD1 effects on HMGCR levels described earlier in cell lines and further characterize its consequences on SREBP transcription program. The Ubiad1ki/ki mouse model appears a useful tool and may prime further research into SCD disease mechanism and treatment. However, I am not convinced that at this stage this work provides a significant advance on the HMGCR regulation by UBIAD1 or on SCD pathophysiology. This study touches on several questions but leaves them mostly unanswered. For example:

1) In Ubiad1ki/ki mice, both HMGCR and UBIAD1 protein levels are increased, however no mechanistic insight is provided to explain these observations, particularly the increase in UBIAD1 protein levels by the N100S mutation.

2) Previous work indicate that UBIAD1 inhibits HMGCR degradation by acting in a post ubiquitination state. This study does not clarify the mechanism by which UBIAD1 prevents HMGCR degradation and how this is impacted by the N100S mutation. Does this mutant bind tighter to HMGCR or the effects are simply a result of its higher protein levels? It is assumed that its lower capacity to bind to GGPP is responsible for tighter binding to HMGCR but this is not tested.

3) In Ubiad1ki/ki mice, HMGCR levels are disturbed in all tissues analyzed. However, it is unclear why the disease manifest specifically in the eye. Is HMGCR more susceptible to UBIAD1 mediated control in the eye while other mechanisms dominate in the remaining organs? Or does the overall impact of UBIAD1 in cholesterol homeostasis insufficient to lead to other complications such as atherosclerosis?

4) Why have the authors chosen to backcross their C57BL/6 x 129 Ubiad1ki/ki mice to C57BL/6 pure genetic backgrounds? This may be not be trivial for non-experts.

5) The quality of the SREBP-2 blots are poor throughout the manuscript.

Figure 2A: Poor quality SREBP-2 blots.

Figure 5A; Lane 5 and 6 seem to be reversed regarding the SREBP-2 protein levels both in membrane fraction and nuclear extract. While the authors provided some explanation for this in the main text, it would be more convincing if the authors could comment on this more clearly or repeat the western blot.

Figure 5C; Poor quality of SREBP-2 blot and the SREBP-2 blotting for membrane fraction is missing.

6) Overexposed HMGCR blots

Figure 4C; While it is expected that HMGCR protein level are lower in WT and specifically increased in Ubiad1ki/ki mouse tissues (and MEF), due to overexpression, it is difficult to assess if this is the case in the HMGCR blot in Figure 4C.

Figure 6A; The authors claim that lovastatin caused HMGCR accumulation in Ubiad1ki/ki mouse liver. However, this is difficult to assess when the HMGCR blot is overexposed and no quantification of protein levels is provided.

7) In Figure 7D, the author showed there is no significant increase of total cholesterol in the range of 300-400ng sterol/ug DNA and similar total cholesterol measurement in Figure 7E (for measurement of intermediates of cholesterol synthesis) presented in the range of 100-200ng sterol/ug DNA.

8) Textual changes

Last paragraph, last line should refer to Figure 7E, not 7D.

The legend of Figure 4C and 4D is confusing. Figure references do not seem to be correct. Also, the use of MG132 is mentioned, but nowhere in the figures it is annotated. Please amend.

Reviewer #3:

The manuscript "Schnyder corneal dystrophy-associated UBIAD1 Inhibits ER-associated degradation of HMG CoA reductase in mice" is a beautiful addition to the regulated ERAD and sterol physiology literature, with the thoroughness and quality that I have come to expect from Dr. Debose-Boyd and his intrepid group. It is very pleasing to see such a thoughtful and well-rendered test of a disease model; this work most certainly should be published in eLife.

The only thing I am not totally clear on is the experiment in Figure 5C. In that experiment, the highly stabilized K→R form of HMGCR (shoot, my whole career I have called it HMGR…) is used as an excellent pre-stabilized control. But if UBIAD only works by allowing enhanced degradation of HMGCR, why are the effects of UBIAD mutant on wild-type HMGCR, which presumably still undergo some degradation, more pronounced than the degradation behavior of the cis highly stabilized mutant?

If this were one of those more evil journals that care not about a PIs resources, or the time of their hard working underlings, I would suggest an experiment with the double mutant, to see if there are regulatory effects that transcend control of degradation. Fortunately, the eLife ethos (which I enthusiastically endorse after spending 50,000 dollars for an extraneous, confirmatory biochemical test of an air-tight genetic study (true story)) encourages reviewers to find merit in the extant work and resist requesting more experiments. So, maybe an explanation or even a mention of this feature of the Figure 5C results would be useful.

Another question that might be interesting to explore in this brief study is the connection (or lack thereto) between this opaquitopathy (I just made that up; not bad!)-meaning opacity-causing disease- and the recent work on cataracts that appear to be relieved by lanosterol, since HMGCR would be expected to promote synthesis of more of this alleviating agent in the disease modeled herein. Different mechanism? Different pathology? No relation? Just an interesting thing to reference in the discussion maybe. Or maybe not.

Otherwise, full speed ahead. Fantastic work!

Reviewer #1:

[…] The paper is generally well organized and written.

1) As I am sure the authors are aware, another group recently published a preliminary analysis of a N100S UBIAD1 knock-in mouse (PMID 29977031). This earlier study lacks the molecular analysis of HMGCR, age-dependent phenotypes, and mechanistic interpretations of the current work. Nevertheless, it is important that the authors provide a careful discussion of these earlier findings and how they relate to the current study.

We are aware that another group published a preliminary analysis of UBIAD1 (N100S) knockin mice, which lacks molecular characterization of HMGCR and documentation of age-dependent opacification of knockin corneas. The authors report evidence of mitochondrial damage and accumulation of glycerophosphoglycerol in corneas of the knockin mice. Based on this finding, the authors propose that mitochondrial dysfunction is linked to the abnormal cholesterol deposition that occurs in SCD patients. We have now included a discussion of these results in the revised manuscript.

2) In Figure 1B, it would be useful to show some indicator of equal loading (perhaps the stained blot showing total protein or a blot for an unrelated protein). In the graph, it would be useful to have error bars on the protein levels if available. I realise the next figure provides a more thorough analysis with suitable controls, but a reader is immediately faced with a rather sparse and somewhat incomplete observation right at the outset. Perhaps one option is to simply omit Figure 1 and start with Figure 2 since this is better characterized anyway.

3) Figure 1B would be easier to follow in the text and legend if the three sub-panels were labelled as panels B, C, and D. This would make it easier to call out the individual experiments. The same applies to other figure panels where more than one type of experiment is shown together (e.g., the top and bottom parts of Figure 2A. In general, it is easier for the reader if each panel displays a single experiment.

It is important to note that although separated into different panels, membrane and nuclear extract fractions immunoblotted in Figure 1B were obtained from the same group of livers. We noted in the figure legend that “although shown in separate panels, calnexin and LSD-1 serves as loading controls for the HMGCR and nuclear SREBP immunoblots.” The remainder of experiments (e.g., Figure 2A) were conducted similarly and the fact that loading controls are shown in separate panels was indicated in the figure legend. To improve clarity of Figure 1B, we have removed the HMGCR mRNA data (which appears in Figure 1—figure supplement 1A) and estimation of the relative amount of HMGCR protein. We now report the absolute level of HMGCR protein under the immunoblot as suggested by this reviewer (point 5). We have also indicated in the figure legend that “although shown in a separate panel, LSD-1 serves as a loading control for the nuclear SREBP immunoblots.”

4) The increase in UBIAD1 in Ki livers (and other tissues) is based on increased levels in the membrane fraction. Is it possible that the total levels are the same, and the difference is due to efficiency of recovery of certain subcellular membranes? Figure 6 shows that mutant UBIAD1 is preferentially in the ER whereas wild type UBIAD1 is preferentially in the Golgi. So if the ER is more effectively recovered than Golgi in the membrane fraction, then one might get the impression that total UBIAD1 is increased. Can a blot also be done on total lysate to clarify this issue? Total lysate is shown in Figure 6C, but it is unclear if the two blots are directly comparable to infer anything about total UBIAD1 levels.

Throughout the studies presented in this manuscript, we have immunoblotted membrane fractions for ERlocalized calnexin or UBXD8 (Figure 3) to control for loading. Even though levels of UBIAD1 and HMGCR accumulate in livers and other tissues of knockin mice, the levels of calnexin and UBXD8 remain constant. If the ER was more efficiently recovered than Golgi in the membrane fractions of the knockin mice, it would be anticipated that levels of both calnexin and UBXD8 would also be increased.

5) I think it is more useful for a reader to directly state the absolute level of increase in HMGCR rather than the value after normalization to mRNA levels. The absolute level of protein is what is important for its enzymatic activity, which is the key parameter for the disease (i.e., the amount of cholesterol that accumulates) that forms the focus of this paper. The normalized values inform more about the relative rates of synthesis/degradation, which is also interesting but less central to the phenotype. Thus, one could simply report total values in the figures and Results section, then in the discussion state how the levels are increased despite much lower mRNA levels, suggesting that either synthesis rate is higher and/or degradation rate is lower.

We now report the absolute levels of HMGCR rather than values after normalization. These values are included under the HMGCR immunoblots in Figure 1B and Figure 2B.

6) Results section – It states that the Lovastatin effect on HMGCR is blunted in Ki mice liver relative to wild type, but the result in Figure 6A shows that accumulation of HMGCR is enhanced in Ki mice liver. Perhaps the sentence is worded incorrectly?

The sentence to which the reviewer refers is confusing and has been removed.

7) Results section – It says Lovastatin led to a slight elevation of HMGCR in Ki mice; However, it was not easy for me to judge from the blot if the increase was any more or less than in wild type mice. Perhaps it is useful to quantify the result and be more precise in the wording with something like "led to XX-fold elevation above the already high levels of HMGCR in Ki mice…" to emphasize that the levels at baseline are already very high.

We have removed the sentence that states lovastatin caused a slight increase in the amount of HMGCR protein in eyes of the knockin mice.

8) It is worth stating how frequently (if at all) corneal inclusions are observed in wild type and heterozygous mice of similar age. This is important to make sure that the inclusions are not simply a consequence of old age.

We have now indicated in the Results section that none of the wild type UBIAD1 (N100S) knockin mice developed corneal opacification and the heterozygous animals were not examined.

Reviewer #2:

[…] This study touches on several questions but leaves them mostly unanswered. For example:

1) In Ubiad1ki/ki mice, both HMGCR and UBIAD1 protein levels are increased, however no mechanistic insight is provided to explain these observations, particularly the increase in UBIAD1 protein levels by the N100S mutation.

2) Previous work indicates that UBIAD1 inhibits HMGCR degradation by acting in a post ubiquitination state. This study does not clarify the mechanism by which UBIAD1 prevents HMGCR degradation and how this is impacted by the N100S mutation. Does this mutant bind tighter to HMGCR or the effects are simply a result of its higher protein levels? It is assumed that its lower capacity to bind to GGPP is responsible for tighter binding to HMGCR but this is not tested.

3) In Ubiad1ki/ki mice, HMGCR levels are disturbed in all tissues analyzed. However, it is unclear why the disease manifest specifically in the eye. Is HMGCR more susceptible to UBIAD1 mediated control in the eye while other mechanisms dominate in the remaining organs? Or does the overall impact of UBIAD1 in cholesterol homeostasis insufficient to lead to other complications such as atherosclerosis?

4) Why have the authors chosen to backcross their C57BL/6 x 129 Ubiad1ki/ki mice to C57BL/6 pure genetic backgrounds? This may be not be trivial for non-experts.

We backcrossed the knockin mice to pure C57BL/6 genetic background to ensure observed phenotypes are not influenced by the mixed genetic background. This rationale is now included in the revised manuscript.

5) The quality of the SREBP-2 blots are poor throughout the manuscript.

Figure 2A: Poor quality SREBP-2 blots.

Figure 5A; Lane 5 and 6 seem to be reversed regarding the SREBP-2 protein levels both in membrane fraction and nuclear extract. While the authors provided some explanation for this in the main text, it would be more convincing if the authors could comment on this more clearly or repeat the western blot.

Figure 5C; Poor quality of SREBP-2 blot and the SREBP-2 blotting for membrane fraction is missing.

In the revised manuscript, we have provided darker exposures of the nuclear SREBP-2 blots for Figure 2A and Figure 5C. We have observed some variation in the quality of SREBP-2 nuclear immunoblots over the years and do not understand the basis for this. It should be noted; however, that the immunoblots show a decrease in nuclear SREBP-2 in UBIAD1 (N100S) knockin mice fed chow or cholesterol-containing diets, which correlates with reduction in mRNAs encoding its target genes. In Figure 5C, we only immunoblotted for nuclear SREBP-2 as a control for cholesterol feeding. Figure 5 now includes mRNA data showing that cholesterol feeding downregulates mRNAs encoding SREBP target genes in livers of both wild type and knockin mice.

Lanes 5 and 6 of Figure 5A were not reversed. The result shows that cholesterol feeding causes an increase in the precursor of SREBP-2, which likely results from inhibition of its cleavage to the mature nuclear form.

6) Overexposed HMGCR blots

Figure 4C; While it is expected that HMGCR protein level are lower in WT and specifically increased in Ubiad1ki/ki mouse tissues (and MEF), due to overexpression, it is difficult to assess if this is the case in the HMGCR blot in Figure 4C.

Figure 6A; The authors claim that lovastatin caused HMGCR accumulation in Ubiad1ki/ki mouse liver. However, this is difficult to assess when the HMGCR blot is overexposed and no quantification of protein levels is provided.

Figure 4C shows an experiment in which MEFs were first depleted of isoprenoids through incubation in medium containing lipoprotein-deficient serum and the HMGCR inhibitor compactin. As a result of this depletion, HMGCR transcription, translation, and stability are increased. These increases result in the accumulation of HMGCR protein to levels that approach those observed in the knockin MEFs. In the isoprenoid-replete conditions shown in Figure 4A, levels of HMGCR protein are low because of sufficient levels of sterol and nonsterol isoprenoids; HMGCR protein accumulates in Ubiad1^Ki cells because of its resistance to degradation.

7) In Figure 7D, the author showed there is no significant increase of total cholesterol in the range of 300-400ng sterol/ug DNA and similar total cholesterol measurement in Figure 7E (for measurement of intermediates of cholesterol synthesis) presented in the range of 100-200ng sterol/ug DNA.

Corneas from different sets of mice were analyzed in Figure 7E and 7D, which may explain differences in total cholesterol levels.

8) Textual changes

Last paragraph, last line should refer to Figure 7E, not 7D.

The legend of Figure 4C and 4D is confusing. Figure references do not seem to be correct. Also, the use of MG132 is mentioned, but nowhere in the figures it is annotated. Please amend.

We now changed Figure 7D to Figure 7E in the last paragraph as indicated by this reviewer.

The reviewer is correct, the legend to Figure 4 is wrong. We should have indicated that all of the cells in D received MG-132. This has now been corrected in the revised manuscript.

Reviewer #3:

[…] Another question that might be interesting to explore in this brief study is the connection (or lack thereto) between this opaquitopathy (I just made that up; not bad!)-meaning opacity-causing disease- and the recent work on cataracts that appear to be relieved by lanosterol, since HMGCR would be expected to promote synthesis of more of this alleviating agent in the disease modeled herein. Different mechanism? Different pathology? No relation? Just an interesting thing to reference in the discussion maybe. Or maybe not.

Otherwise, full speed ahead. Fantastic work!

We thank the reviewer for the supportive comments!

Author details

1. Youngah Jo

   Departments of Molecular Genetics, Center for Human Nutrition, University of Texas Southwestern Medical Center, Dallas, United States

   Contribution

   Conceptualization, Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing—original draft, Writing—review and editing

   Competing interests

   No competing interests declared

2. Jason S Hamilton

   Departments of Molecular Genetics, Center for Human Nutrition, University of Texas Southwestern Medical Center, Dallas, United States

   Contribution

   Data curation, Validation, Investigation, Visualization, Methodology

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-3291-3125

3. Seonghwan Hwang

   Departments of Molecular Genetics, Center for Human Nutrition, University of Texas Southwestern Medical Center, Dallas, United States

   Contribution

   Resources, Validation, Investigation, Methodology

   Competing interests

   No competing interests declared

4. Kristina Garland

   Departments of Molecular Genetics, Center for Human Nutrition, University of Texas Southwestern Medical Center, Dallas, United States

   Contribution

   Investigation, Methodology

   Competing interests

   No competing interests declared

5. Gennipher A Smith

   Departments of Molecular Genetics, Center for Human Nutrition, University of Texas Southwestern Medical Center, Dallas, United States

   Contribution

   Investigation, Methodology

   Competing interests

   No competing interests declared

6. Shan Su

   Departments of Molecular Genetics, Center for Human Nutrition, University of Texas Southwestern Medical Center, Dallas, United States

   Contribution

   Investigation, Methodology

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-9369-1642

7. Iris Fuentes

   Departments of Molecular Genetics, Center for Human Nutrition, University of Texas Southwestern Medical Center, Dallas, United States

   Contribution

   Investigation, Methodology

   Competing interests

   No competing interests declared

8. Sudha Neelam

   Department of Ophthalmology, Center for Human Nutrition, University of Texas Southwestern Medical Center, Dallas, United States

   Contribution

   Investigation, Methodology

   Competing interests

   No competing interests declared

9. Bonne M Thompson

   Center for Human Nutrition, University of Texas Southwestern Medical Center, Dallas, United States

   Contribution

   Investigation, Methodology

   Competing interests

   No competing interests declared

10. Jeffrey G McDonald

    Center for Human Nutrition, University of Texas Southwestern Medical Center, Dallas, United States

    Contribution

    Investigation, Methodology

    Competing interests

    No competing interests declared

11. Russell A DeBose-Boyd

    Departments of Molecular Genetics, Center for Human Nutrition, University of Texas Southwestern Medical Center, Dallas, United States

    Contribution

    Conceptualization, Supervision, Funding acquisition, Visualization, Writing—original draft, Project administration, Writing—review and editing

    For correspondence

    Russell.Debose-Boyd@utsouthwestern.edu

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0002-7452-5227

• 1,480

  Page views

• 266

  Downloads

• 13

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.