Relationship of plasma 3‐carboxy‐4‐methyl‐5‐propyl‐2‐furanpropanoic acid concentration with OATP1B activity in patients with chronic kidney disease

Abstract

Organic anion‐transporting polypeptides (OATP)1B are drug transporters mainly expressed in the sinusoidal membrane. Many studies have suggested that OATP1B activity is affected by genetic factor, the uremic toxin 3‐carboxy‐4‐methyl‐5‐propyl‐2‐furanpropanoic acid (CMPF), and inflammatory cytokines, such as tumor necrosis factor‐α (TNF‐α) and interleukin‐6 (IL‐6). Coproporphyrin‐I (CP‐I) is spotlighted as a highly accurate endogenous substrate of OATP1B. We previously reported a positive correlation between plasma CMPF and CP‐I concentrations in patients with chronic kidney disease (CKD). The present study evaluated the impact of genetic polymorphisms, CMPF, IL‐6, TNF‐α, and estimated glomerular filtration rate (eGFR) on individual differences in OATP1B activity in patients with CKD. Seventy‐three patients with CKD who received kidney transplant at least 3 months earlier were analyzed. Plasma CP‐I concentration was higher in OATP1B1*15 carriers than in non‐carriers. In all patients, CP‐I did not correlate significantly with CMPF, IL‐6, TNF‐α, or eGFR. However, when the dataset was cut off at CMPF concentration of 8 and 7 μg/mL, 4 μg/mL, 3 μg/mL or 2 μg/mL, CMPF correlated positively with CP‐I, and correlation coefficient tended to be higher as plasma CMPF concentration was lower. In conclusion, OATP1B1*15 impacted OATP1B activity in patients with CKD, but IL‐6 and TNF‐α did not. However, the impact of CMPF on OATP1B activity was limited to low CMPF concentrations, and the effect could be saturated at high concentrations. When prescribing an OATP1B substrate drug for patients with CKD, the OATP1B1*15 carrier status and plasma CMPF concentration may need to be considered to decide the dose regimen.

Study Highlights.

• WHAT IS THE CURRENT KNOWLEDGE ON THE TOPIC?

A previous in vitro study showed that organic anion transporting polypeptides (OATP)1B polymorphism, the uremic toxin 3‐carboxy‐4‐methyl‐5‐propyl‐2‐furanpropanoic acid (CMPF), and inflammatory cytokines, such as tumor necrosis factor‐α (TNF‐α) and interleukin‐6 (IL‐6), decreased OATP1B activity. Recently, coproporphyrin‐I (CP‐I) is spotlighted as a highly accurate endogenous substrate of OATP1B activity.

• WHAT QUESTION DID THIS STUDY ADDRESS?

Do OATP1B1 polymorphism, CMPF, TNF‐α, and IL‐6 influence plasma CP‐I concentration in patients with chronic kidney disease (CKD)?

• WHAT DOES THIS STUDY ADD TO OUR KNOWLEDGE?

OATP1B1*15 impacted OATP1B activity in patients with CKD, but IL‐6 and TNF‐α did not. On the other hand, the impact of CMPF on OATP1B activity was limited and could be saturated at high concentration of CMPF.

• HOW MIGHT THIS CHANGE CLINICAL PHARMACOLOGY OR TRANSLATIONAL SCIENCE?

When administering an OATP1B substrate drug for patients with CKD, it may not be necessary to consider inflammatory conditions, but the effects of genetic polymorphisms and uremic toxins should be considered to decide the dose regimen.

INTRODUCTION

Organic anion transporting polypeptides (OATP)1B are drug transporters expressed mainly in sinusoidal membranes and mediate uptake of various endogenous and exogenous compounds, such as steroid hormone, bile acid, antihypertensive drugs, and anti‐dyslipidemia drugs. ¹ Because some 3‐hydroxy‐3‐methylglutaryl coenzyme‐A reductase inhibitors are representative substrates of OATP1B, the pharmacokinetics of these drugs have been reported to be affected by in vivo OATP1B activity. ² , ³ In addition, OATP1B substrate drugs have shown an increase in recent years; for example, some recently approved anti‐hepatitis C virus drugs and the anti‐cytomegalovirus drug letermovir have pharmacokinetics influenced by OATP1B activity. ⁴ , ⁵

Genetic, environmental, and physiological factors have been shown to influence OATP1B activity. Genetically, 388A>G (rs2306283; A388G, N130D, exon 5) and 521T>C (rs4149056; T521C, V174A, exon 5) are well known single nucleotide variations with amino acid substitution on OATP1B1. Among the four haplotypes formed by single nucleotide variations of 388A>G and 521T>C (OATP1B1*1a, OATP1B1*1b, OATP1B1*5, and OATP1B1*15), ⁶ liver cells expressing OATP1B1*5 or OATP1B1*15 have been shown to have lower transport activities. ⁷ , ⁸ Environmental factors, including concomitant medications, such as rifampicin and cyclosporin A, as well as exercise significantly increases plasma levels of OATP1B substrates. ⁹ , ¹⁰ As for physiological factors, the uremic toxin 3‐carboxy‐4‐methyl‐5‐propyl‐2‐furanpropanoic acid (CMPF), and inflammatory cytokines, such as interleukin‐6 (IL‐6) and tumor necrosis factor‐α (TNF‐α), have been reported to decrease OATP1B activity in vitro. ¹¹ , ¹² We recently reported that OATP1B activity was influenced by genetic polymorphisms and uremic toxins but not by inflammatory cytokines in patients with rheumatoid arthritis. ¹³ However, although the association of OATP1B activity with drug–drug interactions have been extensively studied, individual variations in OATP1B activity associated with other factors, such as disease and gene polymorphism, in real‐world patients are less well‐documented. Hence, OATP1B substrate drugs are administered at fixed dose in clinical settings unless the patient is taking a drug that may interact with OATP1B, giving rise to individual differences in drug efficacy and adverse effects.

Many patients with chronic kidney disease (CKD) have concurrent hypertension, dyslipidemia, and/or anemia, and take various medications for these diseases. The doses of renally excreted drugs are commonly adjusted based on renal function in individual patients. ¹⁴ However, the pharmacokinetics of drugs mainly eliminated by the liver may also be altered in patients with renal impairment, because drug metabolism and/or transport activity in the liver may change in these patients. A recent review article suggests that elevated plasma levels of uremic toxins and inflammatory cytokines in CKD may dysregulate the activity and/or expression of drug‐metabolizing enzymes, such as CYP2D6, CYP1A2, CYP2C9, and CYP2C19, and transporters such as OATP1B in the liver. ¹⁵ Our study using coproporphyrin‐I (CP‐I) as an endogenous substrate of OATP1B also suggested decreased OATP1B activity in patients with CKD. ¹⁶ In addition, we found a positive correlation between plasma CP‐I and CMPF concentrations, suggesting that CMPF could cause reduced OATP1B activity in patients with CKD. ¹⁷ However, previous studies had relatively small numbers of subjects and did not assess the impact of gene polymorphisms and inflammatory cytokines on OATP1B activity. With this background, the purpose of the present study was to elucidate factors contributing to individual differences in OATP1B activity in patients with CKD, using CP‐I as a biomarker of OATP1B activity.

MATERIALS AND METHODS

Reagents

The CP‐I standard and the stable isotopically labeled internal standard (¹⁵N₄‐CP‐I) were purchased from Frontier Scientific and Toronto Research Chemicals, respectively. CMPF standard and the stable isotopically labeled internal standard (CMPF‐d5) were purchased from Sigma‐Aldrich and Cayman Chemical, respectively. Human serum albumin and all other solvents and reagents (water, methanol, acetonitrile, dimethyl sulfoxide, 85% phosphoric acid, 25% ammonia solution, and formic acid) with the highest analytical quality were purchased from FUJIFILM Wako Pure Chemical Corporation. Ninety‐six‐well solid phase extraction plates (Oasis MAX μElution plate, 30 μm) were purchased from Waters.

Subjects

Among the patients attending the Department of Urology in Oita University Hospital between November 2020 and April 2021, we recruited 74 patients with CKD with kidney transplant conducted at least 3 months earlier, who were not taking OATP1B inhibitors, such as rifampicin (single dose), letermovir, glecaprevir‐pibrentasvir combination, eltrombopag olamine, cyclosporin A, clarithromycin, roxadustat, vadadustat, and daprodustat, or OATP1B inducers, such as rifampicin (repeated dose). The renal function of the patients was stable. Exclusion criteria were total bilirubin (t‐Bil) greater than 1.5 mg/dL and alanine aminotransferase (ALT) greater than 100 IU/L. One patient was excluded due to t‐Bil greater than or equal to 1.5 mg/dL, and 73 patients satisfied the selection criteria. Blood sample was drawn from a vein into a collection tube containing EDTA‐3K anticoagulant. Four hundred μL of whole blood was dispensed for gene polymorphism analysis, and the remaining sample was centrifuged at 2330 g for 5 min at 4°C. Then, 250 μL of plasma was dispensed into a light shielding tube (ARGOS Technologies) for measuring CP‐I and CMPF, and the remaining plasma into a clear polypropylene tube. Whole blood and plasma samples were stored at −40°C until measurement.

Clinical and laboratory data obtained on the day of blood sampling for CP‐1 and CMPF measurements were extracted from electronic medical records. The extracted data comprised prescription drugs, sex, age, weight, height, white blood cell count, red blood cell count, hemoglobin, hematocrit, C‐reactive protein, albumin, total protein, aspartate aminotransferase, ALT, t‐Bil, γ‐glutamyl transpeptidase, alkaline phosphatase, serum creatinine, blood urea nitrogen, type of renal transplant, primary disease, and immunosuppressive regimen. Only ALP, γ‐GTP, and TP data of a few patients were extracted from a day closest to the date of blood sampling. Body mass index (BMI) was calculated by the following equation: BMI = weight (kg)/[height (m)]². ¹⁸ Estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration equation for Japanese. ¹⁹

The study was performed according to the ethical standards of our institution as well as the Helsinki Declaration of 1975, as revised in 2013, and was reviewed and approved by the Ethics Committees of Oita University (approval numbers: 1433 and P‐13‐14) and Meiji Pharmaceutical University (approval number: 202026). Each patient received prior explanations about this study and gave written informed consent.

OATP1B1 genotyping assay

Total DNA was extracted from 400 μL of whole blood using the Maxwell 16 DNA Purification Kit (Promega). The OATP1B1 allelic variants were determined by analyzing single‐nucleotide mutations 388A>G (rs2306283) and 521T>C (rs4149056). Allele discrimination was determined by TaqMan genotyping assay (Thermo Fisher Scientific) with a LightCycler Nano System (Roche Applied Science). Gene polymorphism analysis identified four haplotypes: OATP1B1*1a, OATP1B1*1b, OATP1B1*5, and OATP1B1*15. Subjects were classified according to genotype into six groups: OATP1B1*1a/*1a, OATP1B1*1a/*1b, OATP1B1*1b/*1b, OATP1B1*1a/*15, OATP1B1*1b/*15, and OATP1B1*15/*15. They were further divided based on OATP1B1*15 carrier status into two groups: OATP1B1*15 allele carrier (OATP1B1*1a/*15, OATP1B1*1b/*15, or OATP1B1*15/*15) and non‐carrier (OATP1B1*1a/*1a, OATP1B1*1a/*1b, or OATP1B1*1b/*1b).

Measurement of plasma CP‐I and CMPF concentrations

Plasma CP‐I and CMPF concentrations were measured simultaneously according to the methods reported by Suzuki et al. ¹⁷ Briefly, Oasis MAX μElution Plate (Waters) was used to pretreat 250 μL of plasma by solid phase extraction. The extract was analyzed by ultra‐high‐performance liquid chromatography coupled to tandem mass spectrometry using the Nexera X₂ LC system coupled to LCMS−8040 Liquid Chromatograph Mass Spectrometer (Shimadzu) equipped with electrospray ionization. The ¹⁵N₄‐CP‐I and CMPF‐d₅ were used as internal standards for CP‐I and CMPF, respectively. CP‐I was measured with (M + H)⁺ signal in positive ion mode, and CMPF was measured with (M – H)⁻ signal in negative ion mode. The tandem mass spectrometry transitions monitored were mass‐to‐charge ratio (m/z) 655.4 → m/z 596.3 for CP‐I, m/z 659.3 → m/z 600.3 for ¹⁵N₄‐CP‐I, m/z 239.0 → m/z 195.2 for CMPF, and m/z 244.2 → m/z 200.2 for CMPF‐d₅. The assays were validated in accordance with the US Food and Drug Administration Guidance for Bioanalytical Method Validation. ²⁰ The lower limit of quantification was 0.1 ng/mL for CP‐I and 50 ng/mL for CMPF. The within‐batch accuracy of the assay ranged from 92.1% to 110.2% for CP‐I, and from 99.1% to 109.3% for CMPF. The within‐batch precision was less than 7.6% for CP‐I and 3.4% for CMPF.

Measurement of plasma inflammatory cytokines

Plasma TNF‐α and IL‐6 concentrations were determined using Human TNF‐alpha Quantikine HS ELISA Kit (R&D Systems, Inc.) and Human IL‐6 Quantikine HS ELISA Kit, respectively. In brief, 50 μL of plasma was placed in a 96‐well microplate coated with a monoclonal antibody for human TNF‐α (or 100 μL of plasma for IL‐6), and incubated for 2 h at room temperature on a shaker at 22 g. Each well was aspirated and washed with buffer; washing was repeated three times. A biotin‐conjugated polyclonal antibody specific for human TNF‐α or IL‐6 (200 μL) was added to each well and incubated for 1 h at room temperature on the shaker. The wells were washed repeatedly, as described above. Then, 200 μL of diluted Streptavidin Polymer‐HRP solution was added, and incubated for 0.5 h at room temperature on the shaker. After repeated washing, 200 μL of extemporaneously prepared Substrate Solution was added and incubated for 0.5 h at room temperature on the benchtop under light shielding. Fifty μL of 2 normal (N) sulfuric acid was added as Stop Solution, and optical density was determined at 450 nm using a microplate reader. According to the manufacturer's product data sheet, the lower limit of quantification was 0.156 pg/mL for both TNF‐α and IL‐6; the intra‐assay precision was below 2.2% for TNF‐α and 4.7% for IL‐6; and the inter‐assay precision was below 6.7% for TNF‐α and 10.8% for IL‐6.

Statistical analysis

Data normality was assessed by Shapiro–Wilk test. Mann–Whitney U test was used to compare plasma CP‐I concentrations between OATP1B1*15 carriers and non‐carriers. Spearman rank correlation coefficient (r _s) was used to evaluate correlation between plasma CP‐I concentration and several physiological factors. All statistical analyses were conducted using the Predictive Analysis Software Statistics version 26.0 (SPSS Inc.). Statistical significance was defined as a p value less than 0.05.

RESULTS

Patient characteristics

Table 1 summarizes the demographic and clinical characteristics. There were 47 men and 26 women, and the mean age was 51.7 ± 13.9 years, body weight was 58.6 ± 12.1 kg, and body mass index was 21.8 ± 3.29 kg/m². The main primary diseases for CKD were unknown (n = 21), diabetic kidney disease (n = 11), chronic glomerulonephritis (n = 11), immunoglobulin A nephropathy (n = 9), and nephrosclerosis (n = 6) in descending order of frequency. Large individual difference in eGFR was observed depending on CKD stage: stage 2 (60 ≤ eGFR ≤ 89), n = 12; stage 3 (30 ≤ eGFR ≤ 30–59), n = 46; stage 4 (15 ≤ eGFR ≤ 29), n = 10; and stage 5D (eGFR ≤ 15), n = 5. All patients had no chronic liver injury and liver function was preserved. The most common immunosuppressive regimen prescribed was tacrolimus, mycophenolate mofetil, and methylprednisolone; followed by tacrolimus, methylprednisolone, and mizoribine.

TABLE 1.

Demographic and clinical data.

Parameter	Value
Gender, male/female	47 (64.4)/26 (35.6)
Age, year	51.7 ± 13.9
Body weight, kg	58.6 ± 12.1
Height, cm	163.8 ± 9.90
White blood cell, ×103/mm3	6.63 [3.32–13.2]
Red blood cell, ×103/mm3	4.22 ± 0.60
Hemoglobin, g/dL	12.6 ± 1.81
Hematocrit, %	38.7 [29.2–50.9]
C‐reactive protein, mg/dL	0.05 [0.01–3.30]
Albumin, g/dL	4.13 ± 0.37
Total protein, g/dL	6.68 [4.99–7.74]
Aspartate aminotransferase, U/L	16.3 [9.40–38.4]
Alanine aminotransferase, U/L	13.0 [4.40–34.8]
Total bilirubin, mg/dL	0.63 [0.32–1.37]
γ‐glutamyl transpeptidase, U/L	24.9 [6.10–130.0]
Alkaline phosphatase, U/L	142.0 [41.0–487.0]
Serum creatinine, mg/dL	1.39 [0.49–14.7]
Blood urea nitrogen, mg/dL	22.7 [0.87–60.6]
Estimated glomerular filtration rate, mL/min/1.73 m2	42.1 ± 17.5
Primary disease
Chronic kidney disease	21 (28.8)
Diabetic kidney disease	11 (15.1)
Chronic glomerulonephritis	11 (15.1)
Immunoglobulin A nephropathy	9 (12.3)
Nephrosclerosis	6 (8.2)
Nephrotic syndrome	3 (4.1)
Polycystic kidney	3 (4.1)
Tubulointerstitial nephritis	2 (2.7)
Benign familial hematuria	2 (2.7)
Reflux nephropathy	1 (1.4)
Gouty nephropathy	1 (1.4)
Pyelonephritis	1 (1.4)
Takayasu arteritis	1 (1.4)
Traffic injury	1 (1.4)
Immunosuppressive regimen
Tacrolimus + mycophenolate mofetil + methylprednisolone	46
Tacrolimus + methylprednisolone + mizoribine	7
Tacrolimus + methylprednisolone + everolimus + mizoribine	4
Tacrolimus + mycophenolate mofetil + methylprednisolone + eerolimus	4
Others	15

Note: Data are expressed as numbers (%) for categorical variables, median [interquartile range] for nonparametric continuous variables, and average ± standard deviation for parametric continuous variables. Data normality was examined by Shapiro–Wilk test.

Comparison of plasma CP‐I concentration between OATP1B1*15 carriers and non‐carriers

Of 73 patients who satisfied the selection criteria, 24 had an OATP1B1*15 allele. Plasma CP‐I concentration was significantly higher in OATP1B1*15 carriers than in non‐carriers (0.76 ± 0.27 vs. 0.59 ± 0.23 ng/mL, p = 0.003; Figure 1a). On the other hand, no significant difference in CMPF concentration was found between OATP1B1*15 carriers and non‐carriers (4.57 ± 5.06 vs. 6.85 ± 10.61 μg/mL, p = 0.874; Figure 1b).

FIGURE 1.

(a) Comparison of plasma (a) CP‐I and (b) CMPF concentrations between OATP1B1*15 carriers and non‐carriers. Plasma CP‐I and CMPF concentrations were analyzed by t‐test. CP‐I, coproporphyrin‐I; CMPF, 3‐carboxy‐4‐methyl‐5‐propyl‐2‐furanpropanoic acid.

Correlation of plasma CP‐I concentration with physiological factors

The correlation between plasma CP‐I concentration and several physiological factors (CMPF, IL‐6, TNF‐α, and eGFR) was analyzed in all patients, and no significant correlation was observed (Figure 2). Next, we repeated the correlation analysis between plasma CP‐I and CMPF concentrations when the dataset was cut off at plasma CMPF concentration of (a) 11 μg/mL, (b) 10 and 9 μg/mL, (c) 8 and 7 μg/mL, (d) 6 and 5 μg/mL, (e) 4 μg/mL, (f) 3 μg/mL, or (g) 2 μg/mL (Figure 3). The results showed a significant positive correlation in (c) and (e) to (g). Furthermore, the correlation coefficient tended to be higher as plasma CMPF concentration was lower (a) r _s  = 0.141; (b) r _s  = 0.149; (c) r _s  = 0.264; (d) r _s  = 0.228; (e) r _s  = 0.344; (f) r _s  = 0.342; and (g) r _s  = 0.409. On the other hand, there was no significant correlation between CP‐1 and eGFR when the dataset was cut off at the same CMPF concentrations as in Figure 3 (Figure 4).

FIGURE 2.

Correlation of plasma CP‐I concentration with (a) CMPF, (b) IL‐6, (c) TNF‐α, and (d) eGFR. All correlation analyses were conducted using Spearman rank correlation test. CP‐I, coproporphyrin‐I; CMPF, 3‐carboxy‐4‐methyl‐5‐propyl‐2‐furanpropanoic acid; eGFR, estimated glomerular filtration rate; IL‐6, interleukin‐6; TNF, tumor necrosis factor.

FIGURE 3.

Correlation between plasma CP‐I and CMPF concentrations when the dataset was cut off at plasma CMPF concentration of (a) 11 μg/mL, (b) 10 and 9 μg/mL, (c) 8 and 7 μg/mL, (d) 6 and 5 μg/mL, (e) 4 μg/mL, (f) 3 μg/mL, or (g) 2 μg/mL. Correlation was tested using Spearman rank correlation coefficient. CP‐I, coproporphyrin‐I; CMPF, 3‐carboxy‐4‐methyl‐5‐propyl‐2‐furanpropanoic acid.

FIGURE 4.

Correlation between plasma CP‐I concentration and eGFR when the dataset was cut off at plasma CMPF concentration of (a) 11 μg/mL, (b) 10 and 9 μg/mL, (c) 8 and 7 μg/mL, (d) 6 and 5 μg/mL, (e) 4 μg/mL, (f) 3 μg/mL, or (g) 2 μg/mL. Correlation was tested using Spearman rank correlation coefficient. CP‐I, coproporphyrin‐I; eGFR, estimated glomerular filtration rate.

DISCUSSION

This study in patients with CKD evaluated the impact of genetic polymorphisms and CMPF, a uremic toxin, on individual differences in OATP1B activity using CP‐I as the endogenous biomarker. The results revealed the following: (1) plasma CP‐I concentration was higher in OATP1B1*15 carriers than in non‐carriers; (2) no significant correlation of plasma CP‐I with physiological factors (CMPF, IL‐6, TNF‐α, and eGFR) was found in all patients; (3) plasma CP‐I concentration correlated positively with CMPF concentration in patient groups with CMPF concentrations less than 8 and 7 μg/mL, less than 4 μg/mL, less than 3 μg/mL, and less than or equal to 2 μg/mL, and the correlation coefficient tended to be higher as CMPF concentration was lower; whereas no significant correlation was found between CP‐I and eGFR at all the cutoff CMPF concentrations.

Previous in vitro assays revealed lower transporting activities of hepatic cells expressing OATP1B1*5 or OATP1B1*15. ⁷ , ⁸ Neuvonen et al. ²¹ reported that plasma CP‐I concentration was 68% higher in healthy Finnish volunteers with c.521CC than with wild type. Mori et al. ²² reported that healthy Japanese individuals with OATP1B1*15/*15 alleles had higher CP‐I levels than those with OATP1B1*1b/*1b or OATP1B1*1b/*15 alleles. Suzuki et al. ²³ demonstrated significantly higher plasma CP‐I concentrations in OATP1B*1b/*15, OATP1B1*1a/*15, and OATP1B1*15/*15 carriers compared to OATP1B1*1b/*1b carriers, suggesting that individual variability in OATP1B activity depends on whether or not a person carries OATP1B1*15. In that study, the average CP‐I concentration in healthy individuals (all genotypes) was 0.48 ± 0.17 ng/mL, and that in OATP1B1*15 carriers was 0.53 ± 0.16 ng/mL. On the other hand, the average CP‐I concentration in the present study was 0.65 ± 0.26 ng/mL, and that in OATP1B1*15 carriers was 0.76 ± 0.27 ng/mL, suggesting reduced OATP1B activity in patients with CKD compared to healthy volunteers, regardless of OATP1B1*15 carrier status.

Figure 2b,c indicate no significant correlation between plasma CP‐I concentration and plasma IL‐6 or TNF‐α concentration. In a previous in vitro study, exposure of hepatic cells to IL‐6 or TNF‐α was found to downregulate OATP1B messenger RNA level. ¹² However, IL‐6 and TNF‐α concentrations used in that in vitro study were 10 and 100 ng/mL, respectively, ~1000 times higher than the plasma levels in patients with CKD of this study (maximum plasma concentration: IL‐6, 15.05 pg/mL; TNF‐α, 3.68 pg/mL). Therefore, the discrepancy between the present findings and previous in vitro results may be due to the differences in IL‐6 and TNF‐α concentrations. In our previous study conducted in patients with rheumatoid arthritis (RA), the median (range) plasma IL‐6 and TNF‐α concentrations were 8.73 (range: 0.78–219.82) pg/mL and 8.10 (range, 0.51–217.02) pg/mL, respectively, and both did not correlate significantly with CP‐I concentration, which agree with the present study findings. Hence, inflammatory cytokines may have little effect on OATP1B activity, at least at plasma concentrations observed in patients with CKD.

Figure 2a,d show no correlation between plasma CP‐I concentration and CMPF concentration, and between plasma CP‐I concentration and eGFR in all subjects. In a previous study by Suzuki et al., ²³ there was no significant correlation between CP‐I and CMPF concentrations in healthy volunteers with eGFR greater than 60 mL/min/1.73 m². However, they evaluated the influence of OATP1B polymorphisms on plasma CP‐I, and CMPF concentrations in a study population with CMPF clustered at lower concentrations than in patients with CKD. In addition, CP‐I tended to correlate positively with CMPF in the population with genetically high transporter activity (r _s = 0.19, p = 0.055). On the other hand, previous reports suggested reduced OATP activity by uremic toxins in patients with CKD. ²⁴ , ²⁵ We recently demonstrated a positive correlation between plasma CMPF and CP‐I concentrations in patients with CKD or RA, ¹³ , ¹⁷ and the result contradicts with the finding in the present study. This study included patients with higher plasma CMPF concentrations compared to previous reports. In a previous study in patients with CKD, only two patients had high CMPF concentrations above 15 μg/mL. ¹⁷ In our previous study in patients with RA, the median (range) CMPF concentration was 1.88 (range: 0.47–13.44) μg/mL. ¹³ On the other hand, the median (range) CMPF concentration in this study was high at 2.67 (range, 0.22–58.46) μg/mL, and six patients had concentrations above 15 μg/mL. A previous in vitro study by Fujita et al. ¹¹ indicated that CMPF inhibited the uptake of OATP1B substrate drugs into hepatocytes according to the Michaelis–Menten kinetics. From our results and the above previous evidence, we hypothesized that the inhibitory effect of CMPF may saturate at high concentrations. Because our previous study showed plasma CMPF concentrations of 11.02 ± 6.22 μg/mL in patients with stage 5D CKD, ¹⁷ we subsequently evaluated the correlation between CP‐I and CMPF concentrations when the dataset was cut off at serial CMPF concentrations from 11 μg/mL in decrements of 1 or 2 μg/mL. Interestingly, we found a significant positive correlation between CP‐I and CMPF concentrations only in patient groups with plasma CMPF concentrations less than 8 and 7, less than 4, less than 3, and less than or equal to 2 μg/mL, and the correlation coefficient tended to be higher as the plasma CMPF concentration was lower. These results suggest a concentration‐dependent inhibitory effect of CMPF on OATP1B1 activity, and that this effect could be saturated at high CMPF concentrations. On the other hand, there was no significant correlation between CP‐1 and eGFR at all the cutoff CMPF concentrations. Ichimura et al. ²⁶ reported that plasma indoxyl sulfate concentration correlated negatively with eGFR in 20 patients on hemodialysis, whereas plasma CMPF concentration did not. Given the above results and evidence, plasma CMPF concentration may be a more prominent factor influencing OATP1B activity than eGFR.

There were several limitations in this study. First, the sample size was small, with a total of 73 patients with CKD, 24 of whom were OATP1B1*15 carriers. A positive correlation between CP‐I and CMPF concentrations was found when patients with low CMPF concentrations were analyzed. However, the impact of OATP1B1*15 on this correlation was not ruled out, because we were not able to conduct correlation analysis in OATP1B1*15 non‐carriers only, due to the small sample. Second, some recruited patients had diseases other than CKD. There is no report on the association of CP‐1 level with physiological factors such as OATP1B1*15, CMPF, IL‐6, and TNF‐α in diseases other than CKD and RA. Therefore, the impact of other diseases on OATP1B activity cannot be ruled out. Third, the immunosuppressive regimen prescribed varied among patients. Some immunosuppressive drugs can downregulate the activity and/or expression of inflammatory cytokines, such as IL‐6 and TNF‐α in blood, and the magnitude depends on the drug. ²⁷ Therefore, systemic cytokine levels may not reflect hepatic levels, which may in part result in underestimation of the effect of inflammatory cytokines. Finally, the correlation coefficients between CP‐I and CMPF concentrations were weak even after the dataset was cut off at various CMPF concentrations. This may be attributed in part to several potential confounding factors, such as genetic background and underlying disease. However, this weak correlation is comparable with those of our previous study ¹³ in patients with RA (r _s  = 0.382 to 0.383) and the study by Fujita et al. ²⁸ that evaluated the correlation between CMPF and an OATP1B substrate SN‐38 (r _s = 0.302).

To the best of our knowledge, this is the first study in patients with CKD which evaluated the impact of several genetic and physiological factors, including OATP1B1*15, CMPF, IL‐6, TNF‐α, and eGFR on OATP1B1 activity measured using an endogenous substrate. In conclusion, OATP1B1*15 impacted OATP1B1 activity in patients with CKD, but IL‐6 and TNF‐α did not. On the other hand, the impact of CMPF on OATP1B1 activity was limited to low CMPF concentrations, and the effect may be saturated at high concentrations. When prescribing an OATP1B1 substrate drug for patients with CKD, the OATP1B1*15 carrier status and plasma CMPF concentration may need to be considered to decide the dose regimen.

AUTHOR CONTRIBUTIONS

H.O., R.Tanaka, and Y.S., wrote the manuscript. R.Tanaka, Y.S., K.O., T.S., and H.I. designed the research. H.O., R.Tanaka, Y.S., A.O., H.S., R.Tatsuta, and T.A. performed the research. H.O., R.Tanaka, and Y.S. analyzed the data. H.O., R.Tanaka, Y.S., A.O., and K.O. contributed new reagents/analytical tools.

FUNDING INFORMATION

This work was supported in part by Grant‐in‐Aid for Japan Research Foundation for Clinical Pharmacology and Grant‐in‐Aid for JSPS KAKENHI Grant Number JP19K16455.

CONFLICT OF INTEREST STATEMENT

The authors declared no competing interests for this work.

Ono H, Tanaka R, Suzuki Y, et al. Relationship of plasma 3‐carboxy‐4‐methyl‐5‐propyl‐2‐furanpropanoic acid concentration with OATP1B activity in patients with chronic kidney disease. Clin Transl Sci. 2024;17:e13731. doi: 10.1111/cts.13731