CYP7A1, NPC1L1, ABCB1, and CD36 Polymorphisms Are Associated with Increased Serum Coenzyme Q10 after Long-Term Supplementation in Women

Coenzyme Q10 (CoQ10), an essential component for energy production that exhibits antioxidant activity, is considered a health-supporting and antiaging supplement. However, intervention-controlled studies have provided variable results on CoQ10 supplementation benefits, which may be attributed to individual CoQ10 bioavailability differences. This study aimed to investigate the relationship between genetic polymorphisms and CoQ10 serum levels after long-term supplementation. CoQ10 levels at baseline and after one year of supplementation (150 mg) were determined, and eight single nucleotide polymorphisms (SNPs) in cholesterol metabolism and CoQ10 absorption, efflux, and cellular uptake related genes were assessed. Rs2032582 (ABCB1) and rs1761667 (CD36) were significantly associated with a higher increase in CoQ10 levels in women. In addition, in women, rs3808607 (CYP7A1) and rs2072183 (NPC1L1) were significantly associated with a higher increase in CoQ10 per total cholesterol levels. Subgroup analyses showed that these four SNPs were useful for classifying high- or low-responder to CoQ10 bioavailability after long-term supplementation among women, but not in men. On the other hand, in men, no SNP was found to be significantly associated with increased serum CoQ10. These results collectively provide novel evidence on the relationship between genetics and CoQ10 bioavailability after long-term supplementation, which may help understand and assess CoQ10 supplementation effects, at least in women.


Introduction
Coenzyme Q 10 (CoQ 10 ) is a vitamin-like molecule that is involved in ATP synthesis in mitochondria and is a crucial antioxidant that protects against oxidative damages of cell membranes and lipoproteins [1][2][3]. CoQ 10 supplementation slowed aging and decreased protein, lipid, and DNA oxidative damage in the senescence-accelerated mouse prone 1 mice [4,5]. Similarly, reduced DNA oxidative damage by CoQ 10 supplementation was observed in male Wistar rats [6]. In addition, CoQ 10 supplementation ameliorated fatigue and promoted exercise performance, possibly by combining enhanced energy production in mitochondria and enhanced antioxidant protection, in male Institute of Cancer Research mice [7]. Human interventional studies also suggest the beneficial effects of CoQ 10 , including anti-inflammatory activities, prevention or improvement of degenerative disorders affecting longevity [8], enhanced vitality of patients undergoing medical treatment as well as on elderly residents of nursing homes [9][10][11], and alleviation of fatigue in patients with chronic fatigue syndrome [12,13]. Hence, CoQ 10 is considered a health-supporting and antiaging supplement. However, meta-analyses of these studies provided inconsistent results related to the effects of CoQ 10 supplementation [14][15][16][17] and even failed to show any beneficial effect of CoQ 10 supplementation [14][15][16]. Hence, further studies are required to produce substantial clinical evidence supporting the benefits of CoQ 10 supplementation [18].
One of the challenges to generate reliable evidence for the beneficial effects of CoQ 10 supplementation is its low bioavailability due to poor water solubility and high molecular weight. Several types of CoQ 10 delivery systems and a stable capsule-based product containing ubiquinol (the reduced form of CoQ 10 ) were developed to increase the bioavailability of supplemental CoQ 10 [19][20][21][22][23][24]. However, considerable variation and high standard derivations were reported for the single-dose pharmacokinetics and bioavailability of CoQ 10 [22,23]. A large variance in the serum CoQ 10 levels was also observed after longterm CoQ 10 supplementation [25,26], suggesting that such variances may be another reason to affect the accurate recognition of the beneficial effects of CoQ 10 .
The absorption of supplemental CoQ 10 into the bloodstream [27] starts with enterocytes absorbing CoQ 10 in the intestine with the aid of carrier molecules/proteins on the cell membrane. The cholesterol transporter Niemann-Pick C1-like 1 (NPC1L1) protein is a candidate molecule to contribute to the absorption process [28,29]. Next, the CoQ 10 molecules are incorporated into chylomicrons, which circulate in the lymph and blood, carrying the CoQ 10 to the liver, where it is attached to the low-density and very-low-density lipoproteins. Then, the CoQ 10 returns to the bloodstream as a lipoprotein complex. Hence, serum CoQ 10 levels correlate strongly and positively with serum cholesterol levels [30]. Some exogenous CoQ 10 can also be incorporated into cells, although the exact mechanism remains unclear [31]. Recently, Anderson et al. reported that the scavenger receptor CD36 drives the uptake of CoQ 10 in brown adipose tissue, where it is required for normal physiological function [32]. In turn, the ATP-binding cassette subfamily B member 1 (ABCB1) expressed in enterocytes may excrete the absorbed CoQ 10 into the intestinal lumen [33]. Thus, several proteins involved in CoQ 10 absorption/incorporation and cholesterol metabolism adjust the individual serum CoQ 10 levels during supplementation.
The involvement of genetic polymorphism in the pharmacokinetic profile, bioavailability, and metabolism of various drugs and nutrients has been reported in the last two decades. For example, single nucleotide polymorphisms (SNPs) and haplotypes of ABCB1 are associated with altered drug disposition, drug response, and toxicity [34]. CD36 SNPs and haplotypes are associated with serum LDL and total cholesterol levels [35]. Recently, Tomei et al. reported that rs731236 in the vitamin D receptor (VDR) gene and rs7116978 in the cytochrome p450 family 2 subfamily R member 1 (CYP2R1) gene were associated with serum 25-hydroxyvitamin D levels, the primary circulating form of vitamin D, after 12 weeks of vitamin D supplementation [36]. Only a few studies have reported an association between genetic polymorphism and basal serum CoQ 10 status in humans [37,38] and increased serum CoQ 10 levels after the two-week supplementation [38], but there is no study reporting the association after long-term supplementation.
In this study, the impact of polymorphisms in genes potentially involved in CoQ 10 bioavailability on serum CoQ 10 levels after long-term supplementation was evaluated in volunteers of the Ubiquinol Health Examination, a prospective intervention trial held in the Ehime Prefecture, Japan [26,39]. The SNPs were assessed in the following genes involved in the cholesterol metabolism: sterol regulatory element-binding protein 2 (SREBP2), 3hydroxy-3-methylglutaryl-coenzyme A reductase (HMGCR), apolipoprotein B (APOB), cytochrome P450 family 7 subfamily A member 1 (CYP7A1), and CoQ 10 absorption (NPC1L1), excretion (ABCB1), and cellular uptake (CD36). We further investigated whether genotyping of SNPs in the above genes is useful for classifying high-or low-responder to CoQ 10 bioavailability after long-term supplementation.

CoQ 10 Supplementation
Two types of reduced CoQ 10 (ubiquinol) supplements (soft-encapsulated and granulated form) were provided by the Kaneka Co. (Osaka, Japan). The soft capsule Kaneka QH TM contains reduced CoQ 10 (50 mg per capsule), rapeseed oil, beeswax, soy lecithin, polyglycerol esters of fatty acids, gelatin, glycerol, and caramel. The granulated form P30 contains 30% (w/w) reduced CoQ 10 (150 mg per sachet), dextrin, gum Arabic, and Lascorbate. The participants could choose either the capsule or the granulated form of the supplement, taking three capsules or one sachet per day, respectively, corresponding to a daily dose of 150 mg CoQ 10 during the intervention trial. The participants also could switch between supplement types every three months.

Study Design
A total of 170 participants who had been participating in the "Verification of health enhancement and QOL improvement effect by continuous ubiquinol ingestion (Ubiquinol Health Examination)" study (UMIN000012612) [26,39] were recruited ( Figure 1). All participants were residents of Kamijima-town of Ehime Prefecture, Japan. Kamijima consists of 25 small islands located in the Seto Inland Sea. Since Kamijima is a super-aging region (population aging rate is more than 40%), the authors considered Kamijima was appropriate to evaluate the effects of antiaging materials such as CoQ 10 . The participants took 150 mg per day of reduced CoQ 10 in the postprandial state (after breakfast or lunch) from November 2016 to November 2017. First blood samples were collected at the time of enrollment in the Kamijima Ubiquinol Health Examination study (from November 2013 to November 2016), which served as a baseline, and another sample was collected in November 2017, after 1 year of 150 mg reduced CoQ 10 supplementation. All participants provided informed consent before they participated in this study. This study was conducted in accordance with the Declaration of Helsinki, and it was approved by the Wayo Women's University Human Research Ethics Committee (No. 1614). Participants were excluded from the analysis if the increased serum CoQ 10 levels were less than 1 µmol/L.

CoQ10 Supplementation
Two types of reduced CoQ10 (ubiquinol) supplements (soft-encapsulated and granulated form) were provided by the Kaneka Co. (Osaka, Japan). The soft capsule Kaneka QH TM contains reduced CoQ10 (50 mg per capsule), rapeseed oil, beeswax, soy lecithin, polyglycerol esters of fatty acids, gelatin, glycerol, and caramel. The granulated form P30 contains 30% (w/w) reduced CoQ10 (150 mg per sachet), dextrin, gum Arabic, and L-ascorbate. The participants could choose either the capsule or the granulated form of the supplement, taking three capsules or one sachet per day, respectively, corresponding to a daily dose of 150 mg CoQ10 during the intervention trial. The participants also could switch between supplement types every three months.

Study Design
A total of 170 participants who had been participating in the "Verification of health enhancement and QOL improvement effect by continuous ubiquinol ingestion (Ubiquinol Health Examination)" study (UMIN000012612) [26,39] were recruited ( Figure 1). All participants were residents of Kamijima-town of Ehime Prefecture, Japan. Kamijima consists of 25 small islands located in the Seto Inland Sea. Since Kamijima is a super-aging region (population aging rate is more than 40%), the authors considered Kamijima was appropriate to evaluate the effects of antiaging materials such as CoQ10. The participants took 150 mg per day of reduced CoQ10 in the postprandial state (after breakfast or lunch) from November 2016 to November 2017. First blood samples were collected at the time of enrollment in the Kamijima Ubiquinol Health Examination study (from November 2013 to November 2016), which served as a baseline, and another sample was collected in November 2017, after 1 year of 150 mg reduced CoQ10 supplementation. All participants provided informed consent before they participated in this study. This study was conducted in accordance with the Declaration of Helsinki, and it was approved by the Wayo Women's University Human Research Ethics Committee (No. 1614). Participants were excluded from the analysis if the increased serum CoQ10 levels were less than 1 μmol/L.

Measurements of Serum CoQ 10 and Total Cholesterol
A non-fasting blood sample was drawn between 9:00 and 15:00 on the test day when it was convenient for each participant. Blood serum was prepared by centrifugation at 3000 rpm for 10 min. For quantification of serum CoQ 10 levels, 0.1 mL of serum was mixed with 0.7 mL 2-propanol and stored at −80°C until just before the analysis. Quantitative analysis of serum CoQ 10 concentration was performed by Kaneka Techno Research Co. (Osaka, Japan) using liquid chromatography with tandem mass spectrometry (LC/MS/MS), as described previously [40,41].

Polymerase Chain Reaction (PCR)-Restriction Fragment Length Polymorphism (RFLP) Genotyping
Genomic DNA was isolated from the whole blood samples using the Maxwell RSC Blood DNA kit and the Maxwell RSC Instrument (Promega Corporation. Madison, WI, USA), according to the manufacturer's instructions. DNA concentrations were determined using Qubit dsDNA BR assay kit (Thermo Fisher Scientific, Waltham, MA, USA) and diluted to a working concentration of 10 ng/µL. The PCR mixtures (20 µL) consisted of 0.2 µmol/L forward and reverse primer pairs (Table 1), 0.2 mmol/L dNTP mixtures, 1 ng/µL template DNA, and 0.01 U/µL Hot-start gene Taq NT DNA polymerase (Nippon Gene Co., Toyama, Japan) in (1×) HS Gene Taq NT buffer. The PCR amplification was performed on a TaKaRa PCR thermal cycler dice gradient (TaKaRa Bio, Shiga, Japan), and the amplification conditions were as follows: an initial activation step of 5 min at 95°C, 40 cycles of 30 s at 95°C (denaturation), 30 s at 60°C (annealing), and 1 min at 72°C (extension), followed by a final extension for 10 min at 72°C. Amplification was confirmed on a 2% agarose gel electrophoresis in a tris-borate-EDTA buffer. Then, residual 10 µL of the PCR products were directly digested overnight with an appropriate restriction enzyme ( Table 1). The digested DNA fragments were electrophoretically separated on a 2% agarose 21 (Nippon Gene Co.) gel. The genotypes detected by PCR-RFLP were confirmed by direct DNA sequencing of the PCR products using ABI PRISM 3100 genetic analyzer (Thermo Fisher Scientific) except for rs133291 in SREBP2.

Data Analysis
Microsoft Excel 2016 (Microsoft Corporation, Redmond, WA, USA) was used for statistical analyses. The increased serum CoQ 10 (∆CoQ 10 ) and cholesterol-adjusted CoQ 10 (serum CoQ 10 per TC, ∆[CoQ 10 /TC]) levels were calculated by subtracting the baseline value from the value after supplementation. We first classified the participants into two groups by the ∆CoQ 10 and ∆[CoQ 10 /TC] levels to assess the association of each SNP with the response to CoQ 10 supplementation. The participants, whose levels were not less than the average value, were grouped as high-responders (HR), while others were grouped as low-responders (LR). A chi-squared test was performed to investigate the association and to confirm Hardy-Weinberg equilibrium (HWE) of SNP genotypes. All statistical assessments were two-tailed and considered significant at p < 0.05. All HWE p-values were more than 0.9, demonstrating that all SNPs meet HWE criteria. Welch's t-test was performed when comparing the ∆CoQ 10 and ∆[CoQ 10 /TC] values after separating the participants by genotype into two groups.

Results
Although 170 participants agreed to participate in this study, 41 dropped out during the intervention. Moreover, 21 participants whose serum ∆CoQ 10 levels were less than 1 µmol/L were also excluded from the analysis, as they seemed not to follow the daily CoQ 10 supplement program. Therefore, data from 108 participants were analyzed (Figure 1). The baseline characteristics of the participants are listed in Table 2. The ∆CoQ 10 and ∆[CoQ 10 /TC] levels showed wide interindividual variability (Figure 2), as reported previously [26]. The mean values of the ∆CoQ 10 in men and women were 5.52 and 5.05 µmol/L, while those of the ∆[CoQ 10 /TC] were 1110 and 952 µmol/mol, respectively. To investigate the involvement of genetic variations in the ∆CoQ 10 and ∆[CoQ 10 /TC], eight SNPs of seven genes were evaluated (Table 1), and PCR-RFLP was performed for genotyping them in each participant ( Figure S1). The genotypes identified by PCR-RFLP were also confirmed by direct sequencing, except for SREBP2 ( Figure S1).  The bottom of the box is the 25th percentile, the line that intersects the box is the median, the multiplication sign within the box is the mean, and the top of the box is the 75th percentile. Whiskers above and below the box represent the 10th and 90th percentiles, and the points above and below the whiskers indicate the outliers. Differences between the two groups were analyzed using Welch's t-tests (p < 0.05). ΔCoQ10, increased serum coenzyme Q10; Δ[CoQ10/TC], cholesterol-adjusted increased serum CoQ10.
To assess the association between each SNP and the response to CoQ10 supplementation, the participants were divided into two groups (HR and LR) according to the average ΔCoQ10 and Δ[CoQ10/TC] levels. Among women, rs2032582 (ABCB1) and rs1761667 (CD36) were associated significantly with the HR/LR classification of ΔCoQ10 (chi-squared test, p = 0.049 and p = 0.022, respectively; Table 3). In addition, in women, rs3808607 (CYP7A1) and rs2072183 (NPC1L1) showed a significant association with the HR/LR classification of Δ[CoQ10/TC] (chi-squared test, p = 0.037 and p = 0.027, respectively;  The bottom of the box is the 25th percentile, the line that intersects the box is the median, the multiplication sign within the box is the mean, and the top of the box is the 75th percentile. Whiskers above and below the box represent the 10th and 90th percentiles, and the points above and below the whiskers indicate the outliers. Differences between the two groups were analyzed using Welch's t-tests (p < 0.05). ∆CoQ 10 , increased serum coenzyme Q 10 ; ∆[CoQ 10 /TC], cholesterol-adjusted increased serum CoQ 10 . To assess the association between each SNP and the response to CoQ 10 supplementation, the participants were divided into two groups (HR and LR) according to the average ∆CoQ 10 and ∆[CoQ 10 /TC] levels. Among women, rs2032582 (ABCB1) and rs1761667 (CD36) were associated significantly with the HR/LR classification of ∆CoQ 10 (chi-squared test, p = 0.049 and p = 0.022, respectively; Table 3). In addition, in women, rs3808607 (CYP7A1) and rs2072183 (NPC1L1) showed a significant association with the HR/LR classification of ∆[CoQ 10 /TC] (chi-squared test, p = 0.037 and p = 0.027, respectively; Table 3). In contrast, none of the SNPs were associated with the HR/LR classification of ∆CoQ 10 or ∆[CoQ 10 /TC] in men. Next, the SNPs identified with significantly low p-values of ∆CoQ 10 and ∆[CoQ 10 /TC] were further evaluated in women. Figure 3 shows the frequency of the different genotypes according to the HR/LR classification in the women. The rs2032582 GG and rs1761667 AA were associated with a "∆CoQ 10 HR" status, and rs3808607 GT/TT and rs2072183 CC were associated with a "∆[CoQ 10 /TC] HR" status.
Next, the women participants were divided into two groups by combining the four identified SNPs in women, and their serum levels of CoQ 10 , ∆CoQ 10 , and ∆[CoQ 10 /TC] were compared (Figure 4). Group 1 consisted of the participants who had four or more of rs3808607 T, rs2072183 C, rs2032582 G, and rs1761667 A alleles, whereas group 2 consisted of the participants who had three or less of these alleles. The serum CoQ 10 levels before and after supplementation and the ∆CoQ 10 of group 1 were significantly higher than that of group 2 (Welch's t-test, p = 0.021, p = 0.010, and p = 0.025, respectively; Figure 4). The ∆[CoQ 10 /TC] was of borderline significance between the two groups (p = 0.051), but group 1 showed a tendency for higher values than group 2. Each group of women was further subdivided into HR or LR according to ∆CoQ 10 and ∆[CoQ 10 /TC] average values to assess potential associations with the response to CoQ 10 supplementation ( Table 4). The group with the above indicated four SNPs showed a significant association with the HR/LR classification of the ∆[CoQ 10 /TC] (chi-squared test, p = 0.003). The association between the group and the HR/LR classification of the ∆CoQ 10 was borderline significant (chi-squared test, p = 0.063). A similar subgroup analysis was performed with the above four SNPs in the men, but no statistical differences were detected between the two groups ( Figure S2A). These results suggest that genotyping of the above four SNPs is useful for predicting the HR/LR classification of the serum ∆CoQ 10 and ∆[CoQ 10 /TC] levels prior to the long-term CoQ 10 supplementation at least in women. increased serum CoQ10; HR, high-responder; LR, low-responder; SNP, single nucleotide polymorphism.
Next, the SNPs identified with significantly low p-values of ΔCoQ10 and Δ[CoQ10/TC] were further evaluated in women. Figure 3 shows the frequency of the different genotypes according to the HR/LR classification in the women. The rs2032582 GG and rs1761667 AA were associated with a "ΔCoQ10 HR" status, and rs3808607 GT/TT and rs2072183 CC were associated with a "Δ[CoQ10/TC] HR" status.  Next, the women participants were divided into two groups by combining the four identified SNPs in women, and their serum levels of CoQ10, ΔCoQ10, and Δ[CoQ10/TC] were compared (Figure 4). Group 1 consisted of the participants who had four or more of rs3808607 T, rs2072183 C, rs2032582 G, and rs1761667 A alleles, whereas group 2 consisted of the participants who had three or less of these alleles. The serum CoQ10 levels before and after supplementation and the ΔCoQ10 of group 1 were significantly higher than that of group 2 (Welch's t-test, p = 0.021, p = 0.010, and p = 0.025, respectively; Figure 4). The Δ[CoQ10/TC] was of borderline significance between the two groups (p = 0.051), but group 1 showed a tendency for higher values than group 2. Each group of women was further subdivided into HR or LR according to ΔCoQ10 and Δ[CoQ10/TC] average values to assess potential associations with the response to CoQ10 supplementation ( Table 4). The group with the above indicated four SNPs showed a significant association with the HR/LR classification of the Δ[CoQ10/TC] (chi-squared test, p = 0.003). The association between the group and the HR/LR classification of the ΔCoQ10 was borderline significant (chi-squared test, p = 0.063). A similar subgroup analysis was performed with the above four SNPs in the men, but no statistical differences were detected between the two groups ( Figure S2A). These results suggest that genotyping of the above four SNPs is useful for predicting the HR/LR classification of the serum ΔCoQ10 and Δ[CoQ10/TC] levels prior to the long-term CoQ10 supplementation at least in women. The bottom of the box is the 25th percentile, the line that intersects the box is the median, the multiplication sign within the box is the mean, and the top of the box is the 75th percentile. Whiskers above and below the box represent the 10th and 90th percentiles, and the points above and below the whiskers indicate the outliers. Differences between the two groups were analyzed using Welch's t-tests (p < 0.05). ΔCoQ10, serum coenzyme Q10; Δ[CoQ10/TC], cholesterol-adjusted serum CoQ10.  The bottom of the box is the 25th percentile, the line that intersects the box is the median, the multiplication sign within the box is the mean, and the top of the box is the 75th percentile. Whiskers above and below the box represent the 10th and 90th percentiles, and the points above and below the whiskers indicate the outliers. Differences between the two groups were analyzed using Welch's t-tests (p < 0.05). ∆CoQ 10 , serum coenzyme Q 10 ; ∆[CoQ 10 /TC], cholesterol-adjusted serum CoQ 10 .
CYP7A1 encodes the rate-limiting enzyme, cholesterol 7α-hydroxylase, which regulates bile acids synthesis. Participants with rs3808607 GG show lower cholesterol levels with a higher level of bile acid synthesis after ingesting high-molecular-weight β-glucan [44]. Moreover, the GT/TT individuals have higher serum cholesterol levels than those harboring the GG genotype. NPC1L1 plays a pivotal role in intestinal cholesterol absorption in addition to its involvement in CoQ 10 absorption. The hepatic expression of NPC1L1 was higher in rs2072183 CC individuals with gallstone disease than in those with CG/GG [45]. ABCB1 encodes the P-glycoprotein pump involved in drug efflux. In patients with colorectal cancer, rs2032582 GG genotype correlates with the highest P-glycoprotein expression in the tumor tissue [46]. In accordance with the ABCB1 expression, its rs2032582 polymorphism was associated with irinotecan-induced severe mucositis in metastatic colorectal cancer patients [47]. CD36 is a multi-ligand scavenger receptor whose primary function is to take up fatty acids and oxidized lipoproteins into cells [48]. The rs1761667 A allele reduces CD36 expression in monocytes [49] and lowers sensitivity to fatty acid molecules [50]. Based on these reports, it is reasonable to hypothesize that the T allele of rs3808607 is involved in raising serum CoQ 10 levels accompanied by serum TC levels, while the C allele of rs2072183 is related to increased CoQ 10 transport from the intestinal epithelium via increased NPC1L1 expression, and the A allele of rs1761667 decreases CoQ 10 uptake from blood/tissue fluid into cells by downregulating CD36 expression; thus predicting the HR to CoQ 10 supplementation. Notably, the G allele selection of rs2032582 showed a contrasting effect because it may increase the expression of ABCB1, resulting in accelerated efflux of molecules from the intestinal epithelium. Rs2032582 (G > T) is in linkage disequilibrium with rs1045642 (C > T) and rs1128503 (C > T), which may alter the higher-order structure of substrate/inhibitor interaction sites, resulting drug-specificity changes [51]. Thus, it may be inferred that ABCB1 translated from the T allele of rs2032582-containing mRNA can efflux more CoQ 10 molecules than that from the G allele-containing mRNA. Further investigations should address the reason of this discrepancy.
In contrast, in men, none of the eight SNPs was associated with the HR/LR classification (Table 3). This could be attributed to the difference in the number of participants in each sex group; men participants (n = 38) were almost half of women participants (n = 70). Moreover, the serum CoQ 10 levels after supplementation and ∆CoQ 10 values in three men participants (arrowheads in Figure S2) were too high, which were identified as outlier values by the Smirnov-Grubbs test (data not shown). Even when those three men participants, who belonged to both group 1 and group 2, were excluded from the analysis, the serum CoQ 10 levels before and after supplementation, ∆CoQ 10 , and ∆[CoQ 10 /TC] remained statistically insignificant between the two groups (p = 0.21, 0.068, 0.15, and 0.50, respectively; Figure S2B). However, those average values were higher in group 1 than in group 2, giving a possibility that classification based on the four SNPs might be useful for HR/LR prediction in both sexes. Alternatively, gender-specific effects due to estrogenregulated gene expression could be responsible for the differences observed in response to CoQ 10 supplementation. Indeed, previous reports showed the gender-related modulation by rs3808607 [52,53], rs2072183 [54,55], rs2032582 [56], and rs1761667 [57]. In either case, whether the classification by the above four SNPs is useful would be clarified by performing the same study with a larger number of men participants.
However, this study has some limitations. First, only eight SNPs from seven genes were evaluated due to throughput limitations and the research budget. These SNPs were carefully selected based on reliable documentation from the PubMed/PMC database from the National Center for Biotechnology Information, USA. However, there are 20, 20, 36, 12, 18, and 178 PubMed-cited SNPs in SREBP2, HMGCR, APOB, CYP7A1, NPC1L1, and ABCB1, respectively (https://www.ncbi.nlm.nih.gov/snp, accessed on 3 January 2021), which may indicate that potentially important SNPs useful for the determination of HR/LR may have been left out form the analysis. Additionally, more suitable SNPs for HR/LR prediction may also exist in other genes. A genome-wide association study with the human SNP array may help find SNPs affecting serum CoQ 10 levels after long-term supplementation. Second, two types of reduced CoQ 10 supplements (a soft-encapsulated and a granulated form) were used according to the preferences of the participants. The participants also could switch between supplement types during the intervention. To increase the number of participants and ask them to follow the daily intake protocol of CoQ 10 supplementation for a long-term period, it was not possible to standardize the administered supplement form. Indeed, around 45% of the participants switched the supplement types during the intervention. Third, it was challenging to monitor the daily supplement intake of the participants for one year. We also provided the participants with a self-check calendar to record the daily intake of the CoQ 10 supplement when providing the supplements every three months. However, the participants sometimes forgot to record, which made it hard to monitor appropriately. Therefore, all we could do for relevant analyses was that the participants, whose serum ∆CoQ 10 levels were less than 1 µmol/L, were excluded from the analysis.
Currently, whether these SNPs and the HR/LR prediction are meaningful or useful to obtain conclusive evidence for the beneficial effects of CoQ 10 remains to be investigated in further detail. A combination of intervention studies with CoQ 10 and SNPs genotyping would help clarify this issue.

Conclusions
This study identified four SNPs in CYP7A1, NPC1L1, ABCB1, and CD36 involved in the regulation of serum CoQ 10 status after long-term supplementation of reduced CoQ 10 in women. Classification according to four SNP genotypes could help predict the response to CoQ 10 supplementations, thereby helping find reliable clinical evidence of the beneficial effects of CoQ 10 in future interventional studies.
Institutional Review Board Statement: The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Wayo Women's University Human Research Ethics Committee (Ethical approval number: 1614, Approval date: 14 October 2016).
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data that support the findings of this study are available from the corresponding author upon reasonable request.