Phloretin Improves Ultrafiltration and Reduces Glucose Absorption during Peritoneal Dialysis in Rats

Significance Statement

In peritoneal dialysis, ultrafiltration is achieved at the cost of systemic glucose absorption with succeeding treatment-induced metabolic side effects and comorbidities. The transport pathway mechanisms that enable glucose transport and the extent to which they individually contribute to mass transfer are not fully understood. Intraperitoneal administration of phloretin, a nonselective blocker of facilitative glucose transporters (GLUTs), reduced glucose absorption and improved ultrafiltration during peritoneal dialysis in rats. Because GLUTs are also present in the human peritoneal membrane, phloretin or similar acting agents could reduce the metabolic cost of ultrafiltration and improve outcomes of peritoneal dialysis.

Visual Abstract

Abstract

Background

Harmful glucose exposure and absorption remain major limitations of peritoneal dialysis (PD). We previously showed that inhibition of sodium glucose cotransporter 2 did not affect glucose transport during PD in rats. However, more recently, we found that phlorizin, a dual blocker of sodium glucose cotransporters 1 and 2, reduces glucose diffusion in PD. Therefore, either inhibiting sodium glucose cotransporter 1 or blocking facilitative glucose channels by phlorizin metabolite phloretin would reduce glucose transport in PD.

Methods

We tested a selective blocker of sodium glucose cotransporter 1, mizagliflozin, as well as phloretin, a nonselective blocker of facilitative glucose channels, in an anesthetized Sprague–Dawley rat model of PD.

Results

Intraperitoneal phloretin treatment reduced glucose absorption by >30% and resulted in a >50% higher ultrafiltration rate compared with control animals. Sodium removal and sodium clearances were similarly improved, whereas the amount of ultrafiltration per millimole of sodium removed did not differ. Mizagliflozin did not influence glucose transport or osmotic water transport.

Conclusions

Taken together, our results and previous results indicate that blockers of facilitative glucose channels may be a promising target for reducing glucose absorption and improving ultrafiltration efficiency in PD.

Excessive glucose absorption during peritoneal dialysis (PD) contributes to metabolic complications, such as hyperinsulinemia, insulin resistance, dyslipidemia, inflammation, increased visceral fat, and weight gain.^1,2 High dialysate glucose concentrations also have direct adverse effects on peritoneal membrane function and structure.^3,4 Previous studies have demonstrated that some sodium glucose cotransporter (SGLT) and facilitative glucose transporter (GLUT) subtypes are expressed in human mesothelial cells,⁵ suggesting that at least some glucose is transported into or possibly, across cells.^6,7 Peritoneal biopsies from patients with long-term PD and encapsulating peritoneal sclerosis also demonstrated some SGLT and GLUT expression changes in the peritoneal membrane compared with healthy controls.⁸ However, the role of GLUTs during PD is still not fully understood.

Our previous experimental study in a rat model of PD showed that empagliflozin, a selective inhibitor of sodium glucose cotransporter 2 (SGLT2), did not reduce glucose uptake during PD,⁹ which was in contrast to previous findings by Zhou et al.¹⁰ In another study by Balzer et al,¹¹ chronic treatment with an SGLT2 inhibitor dapagliflozin resulted in reduced fibrosis and improved ultrafiltration (UF). Moreover, in a recently published study by our group, we found that intraperitoneal phlorizin, a dual blocker of sodium glucose cotransporter 1 (SGLT1) and SGLT2, was effective in reducing glucose diffusion during PD in rats.¹² Phlorizin, also referred to as phloridzine, and its metabolite phloretin are natural compounds found in fruits, and they have lately gained wide attention in research due to the molecules’ anti-inflammatory, antioxidant, antibacterial, and tumor-suppressive properties.^13,14 Despite widespread applications with promising results in experimental research, little is known about the actual clinical benefits and potential side effects of phloretin.¹⁵ Contrary to phlorizin, which acts as a blocker of SGLT channels,¹⁶ phloretin inhibits glucose transport by blocking GLUTs. Experiments in cells have also shown some weak effects by phloretin on SGLT channels.¹⁷ As phlorizin is a nonselective SGLT1 and SGLT2 blocker,¹⁶ we have been unable to determine whether the observed effects on glucose diffusion capacity (MTAC) during phlorizin containing PD dwells were due to SGLT1 blockade or GLUT blockade by phloretin. We hereby tested intraperitoneal administration of mizagliflozin, a selective SGLT1 blocker, or phloretin during a 1-hour experimental PD dwell in rats. A summarizing drawing of peritoneal GLUT targets and tested drugs in this and our previous rat experiments is found in Figure 1.

Figure 1.

Summary of peritoneal glucose transporter targets and tested drugs. Solid symbols indicate inhibitory effects.

Methods

PD was performed in 9- to 10-week-old male Sprague–Dawley rats of an average body weight of 382 g (interquartile range [IQR], 333 − 402; n=32) with free access to water and chow (Special Diets Services RM1(P) IRR.25 no. 801157). The rats were treated in accordance with the guidelines of the National Institutes of Health for the care and use of laboratory animals. The Ethics Committee for Animal Research at Lund University approved the experiments (Dnr 5.8.18–05699). Figure 2 describes the schematic setup of experiments. For induction of anesthesia, the rat was gently placed in a covered glass container, to which a continuous supply of 5% isoflurane in air (Isoban; Abbot, Stockholm, Sweden) was connected. When fully anesthetized, the animal was removed from the container, and anesthesia was maintained with 1.6%–1.8% isoflurane in air delivered via mask. After tracheostomy, the animals were connected to a ventilator (Ugo Basile; Biologic Research Apparatus, Comerio, Italy) and ventilated in a volume-controlled mode using a positive-end expiratory pressure of 4 cm H₂O. End-tidal pCO₂ was monitored continuously and kept between 4.8 and 5.5 kPa (Capstar-100; CWE, Ardmore, PA). Body temperature was kept between 37.1°C and 37.3°C via a feedback-controlled heating pad. The left femoral artery was cannulated for monitoring of mean arterial pressure and heart rate and to obtain blood samples (95 μl) for measurement of glucose, creatinine, urea, electrolytes, hemoglobin, and hematocrit (i-STAT1; Abbot, Abbot Park, IL) before and after dialysis. The i-STAT1 device is calibrated yearly by staff from Abbot. The right femoral vein was cannulated and used for continuous saline infusion of 50 μl/min. The right internal jugular vein was cannulated for drug infusion. Access to the peritoneal cavity was established percutaneously via a multifenestrated silastic catheter (Venflon; BOC Ohmeda AB, Helsingborg, Sweden; outer diameter of 1.7 mm) secured to the skin using cyanoacrylate (Histoacryl; B. Braun Surgical, Rubi, Spain). After 60 minutes, the dialysate was recovered from the peritoneal cavity first by using a syringe and thereafter, by carefully retrieving the rest of the fluid using preweighed gauze tissues. Washout was thereafter performed using 15 ml of PD fluid with no tracer to retrieve residual amounts of the tracer in the peritoneal cavity. Radio-labeled ¹²⁵I human serum albumin (RISA), ⁵¹Cr-EDTA, and creatinine (∼0.3 mmol/L) were added to the dialysis fluid prior to infusion. Free unbound RISA iodine was <1% as measured after trichloroacetic acid precipitation. Hematocrit was determined by centrifuging thin capillary glass tubes. All PD solutions were prewarmed to 37°C before instillation. After the experiment, the rats were euthanized with an intravenous bolus injection of potassium chloride. Prior to dialysate sampling, 1 ml was flushed back and forth several times to ensure a valid sample from the dialysate.

Figure 2.

Schematic of the experimental setup. PD with 1.5% glucose fluid with or without intraperitoneal phloretin or mizagliflozin was performed in anesthetized Sprague–Dawley rats using a fill volume of 20 ml.

Experimental Protocol

This study consisted of four groups of animals treated without (n=8) or with (n=8) mizagliflozin (Merck, Darmstadt, Germany) 56.5 mg/L (1 mmol/L) or treated without (n=8) or with (n=8) phloretin (Merck, Darmstadt, Germany) 50 mg/L (0.18 mmol/L) in 20 ml of 1.5% glucose PD fluid (Balance; Fresenius, Bad Homburg, Germany). Mizagliflozin was dissolved in 200 μl MilliQ water, and phloretin was dissolved in 100 μl DMSO (Merck, Darmstadt, Germany). The corresponding volume of solvent was not added to sham animals’ dialysis fluids. Dialysate samples for analysis of glucose concentration and electrolyte content (i-STAT1 CHEM8) were obtained directly after instillation of the PD-fluid, at 30 minutes, and after 60 minutes of dwell time. The plasma oncotic pressure was determined before dialysis with a colloid osmometer (Osmomat 050 Colloid Osmometer; Gonotec, Berlin, Germany). Samples for radioactivity measurements (25 μl) were obtained from the dialysis fluid before infusion; from the dialysate at 1, 10, 20, 30, 40, 50, and 60 minutes; and from blood serum at 5, 15, 25, 35, 45, and 60 minutes and analyzed on a gamma counter (Wizard 1480; Wallac Oy, Turku, Finland). The intraperitoneal volume was determined from the dilution of RISA using a modified method described by Zakaria and Rippe,¹⁸ as described in the Supplemental Material. Urine was collected during dialysis and analyzed for glucose content (i-STAT1) and radioactivity. Diffusion capacity (MTAC) of glucose and urea was estimated using an isocratic model.⁹ Isocratic diffusion capacity assumes that solute diffusion occurs during a constant UF rate (equal to the average UF rate), of which 50% is free water transport.⁹ MTAC was also estimated using an isovolumetric model¹⁹ and/or a three-pore model (TPM).^9,20 The UF coefficient (LpS) was also estimated by using the TPM. Analysis also included comparison of pre- and postdialysis plasma concentrations, sodium transport parameters, urea dialysate to plasma concentration ratio, and D/D₀ (intra-treatment to initial dialysate concentration ratio) of glucose and creatinine.

Statistical Methods

Data are shown as medians (IQRs), and effect sizes are shown as 95% confidence intervals (95% CIs; wilcox.test in R) unless otherwise stated. Outliers in the intraperitoneal volume versus dwell time analysis were identified using boxplot.stats in R. Significant differences were assessed using a Wilcoxon–Mann–Whitney test. P values less than 5% were considered significant. Calculations were performed using R for Mac version 4.1.1.

Results

Glucose Absorption Was Reduced, whereas UF Rates Were Greatly Improved by Phloretin

Glucose absorption was markedly lower in rats treated with intraperitoneal phloretin, being 75 mg (IQR, 63–87) compared with the control group of 112 mg (IQR, 102–120; P=0.01) (95% CI, −58 to −12 mg) (Table 1). The UF rate was greatly improved in phloretin-treated animals, being 68 μl/min (IQR, 56–73) compared with the control group of 44 μl/min (IQR, 34–48; P=0.007) (95% CI, 10 to 41 μl/min) (Figure 3, Table 1). Intraperitoneal volume versus dwell time curves are found in Figure 4A. The UF efficiency (UF per unit glucose absorbed) was improved by phloretin (P=0.005), with an observed 95% CI for the treatment effect of 11–50 μl/mg glucose. No phloretin effect was observed regarding glucose plasma concentrations before or after dialysis (Supplemental Table 1) or urine production or urinary glucose excretion (Table 2).

Table 1.

Effects of mizagliflozin and phloretin on glucose transport and UF outcomes

Outcome	Mizagliflozin, Median (IQR)	Control, Median (IQR)	P Value	95% CI for Drug Effect	Phloretin, Median (IQR)	Control, Median (IQR)	P Value	95% CI for Drug Effect
Glucose outcomes
Diffusion capacity (MTAC), μl/min
30-min isocratic	221 (198–255)	225 (208–332)	NS	−199 to 66	240 (201–275)	273 (250–304)	NS	−87 to 18
Isocratic	234 (207–280)	217 (206–243)	NS	−45 to 98	163 (138–174)	252 (209–263)	**	−119 to −40
Isovolumetric	258 (227–301)	249 (238–259)	NS	−46 to 97	203 (188–218)	277 (243–294)	**	−102 to −31
TPM	224 (208–266)	218 (211–234)	NS	−76 to 59	170 (152–198)	238 (214–265)	*	−107 to −12
Absorption, mg	105 (100–122)	103 (92–111)	NS	−19 to 48	75 (63–87)	112 (102–120)	*	−58 to −12
D60/D0	0.52 (0.48–0.55)	0.52 (0.5–0.54)	NS	−0.06 to 0.05	0.55 (0.52–0.58)	0.49 (0.47–0.5)	**	0.02 to 0.1
Clearance,a μl/min	174 (161–207)	162 (154–182)	NS	−37 to 77	118 (99–136)	191 (162–199)	**	−102 to −28
UF outcomes
UF rate, μl/min	28 (25–30)	28 (15–42)	NS	−16 to 16	68 (56–73)	44 (34–48)	*	10 to 41
UF efficiency, μl/mg	15 (11–17)	21 (12–31)	NS	−24 to 9	53 (39–66)	23 (17–28)	**	11 to 50

Values refer to a 60-minute evaluation time unless otherwise stated. *, P<0.05; **, P<0.01. NS, not significant.

^aDialysate to plasma clearance.

Figure 3.

Forest plot of phloretin treatment effects. 95% CIs of the phloretin effect after 30 and 60 minutes on glucose and urea diffusion capacity (MTAC), UF rate, and sodium removal compared with the sham group. Interval markers represent the medians of the difference.

Figure 4.

Phloretin effect on volume and small solute transport. (A) Intraperitoneal volume as a function of dwell time estimated in phloretin-treated animals compared with sham. The solid line represents nonlinear regression in drug-exposed animals, whereas the dashed line represents the control group. (B and C) D/D₀ ratios of glucose and creatinine at 1, 30, and 60 minutes. (D) Urea dialysate to plasma ratios at 30 and 60 minutes in phloretin-exposed animals and controls. The solid line represents nonlinear regression in drug-exposed animals, whereas the dashed line represents the control group. **, P<0.01.

NS, not significant.

Table 2.

Effects of mizagliflozin and phloretin on urea diffusion capacity, sodium transport, urine parameters, and UF capacity

Outcome	Mizagliflozin, Median (IQR)	Control, Median (IQR)	P Value	95% CI for Drug Effect	Phloretin, Median (IQR)	Control, Median (IQR)	P Value	95% CI for Drug Effect
Urea diffusion capacity (MTAC), μl/min
30 min isocratic,	352 (333–405)	404 (353–413)	NS	−107 to 54	297 (268–339)	394 (327–427)	NS	−188 to 22
Isocratic	484 (444–594)	511 (468–556)	NS	−165 to 111	350 (318–365)	498 (446–529)	***	−220 to −77
TPM	325 (297–390)	390 (343–426)	NS	−119 to 118	252 (235–294)	308 (250–498)	NS	−271 to 15
Sodium outcomes
30-min removal, μmol	69 (11–127)	38 (−96–116)	NS	−141 to 218	262 (198–315)	226 (127–357)	NS	−239 to 213
30-min removal/UF, mmol/L	61 (45–83)	81 (39–174)	NS	−251 to 151	85 (77–91)	84 (67–101)	NS	−35 to 42
Removal, μmol	144 (131–163)	119 (13–245)	NS	−160 to 197	381 (280–475)	261 (162–298)	*	5 to 293
Removal/UF, mmol/L	87 (85–88)	70 (18–94)	NS	−35 to 105	90 (83–106)	97 (79–105)	NS	−18 to 28
Sodium dip,a mmol/L	4 (3–4)	3 (3–3)	NS	−1 to 1	7 (6–7)	3 (3–4)	**	3 to 4
Clearance,b μl/min	18 (16–20)	15 (2–30)	NS	−24 to 19	47 (34–57)	32 (20–36)	*	1 to 36
Urine outcomes
GFR, ml/min	1.7 (1.3–1.9)	1.6 (1.2–2.1)	NS	−0.9 to 0.7	2.0 (1.8–2.6)	1.8 (1.5–1.9)	NS	−0.9 to 0.9
Urine glucose, mmol/L	22 (17–86)	22 (14–122)	NS	−157 to 127	15 (3–31)	15 (5–31)	NS	−20 to 20
Urine volume, ml	1.5 (1–1.5)	0.8 (0.6–0.9)	NS	−0.02 to 0.9	1.5 (0.7–1.6)	1.4 (0.9–1.6)	NS	−0.8 to 0.8
UF capacity, μl/min per mm Hg
TPM	0.8 (0.6–0.8)	0.7 (0.4–1.3)	NS	−0.7 to 0.6	2.2 (1.8–2.5)	1.3 (1.2–1.9)	NS	−0.1 to 1.3
Solute clearance, μl/min
Chlorideb	39 (34–43)	45 (27–59)	NS	−31 to 19	62 (54–75)	52 (40–65)	NS	−12 to 34

Values refer to a 60-minute evaluation time unless otherwise stated. *, P<0.05; **, P<0.01; ***, P<0.001. NS, not significant.

^a134 mmol/L (fresh fluid contents according to the producer) minus dialysate sodium at 60 minutes.

^bPlasma to dialysate clearance.

Phloretin Reduced Small Solute Diffusion Capacities for Glucose and Urea

Phloretin treatment resulted in reduced isocratic 60-minute diffusion capacity of glucose, being 163 μl/min (IQR, 138–174) in the phloretin group compared with 252 μl/min (IQR, 209–263) in sham animals (P=0.001) (Figure 3, Table 1). Isovolumetric and TPM estimation methods revealed similar results, which are also shown in Figure 3 and Table 1. At 30 minutes, no phloretin effect on glucose diffusion capacity was observed. Similarly, there was no difference in glucose D/D₀ at 30 minutes, yet D/D₀ was higher in the phloretin treatment group at 60 minutes (P=0.003) (Figure 4B).

Phloretin treatment reduced isocratic urea diffusion capacity after a 60-minute dwell time; the phloretin treatment group median was 350 μl/min (IQR, 318–365) compared with control group (498 μl/min; IQR, 446–529), with a 95% CI for the treatment effect of −220 to −77 μl/min (P<0.001). Isocratic MTAC analysis after 30 minutes did not significantly show reduction, nor did the 60-minute analysis using TPM. Despite nonsignificant reductions in urea diffusion capacity, both isocratic MTAC at 30 minutes and TPM MTAC at 60 minutes had asymmetric CIs toward MTAC reduction (Figure 3, Table 2). There was no statistically supported difference in phloretin-treated animals regarding urea dialysate to plasma ratio after 30 minutes, whereas the ratio was lower compared with in the control group at 60 minutes (P<0.001) (Figure 4D). Phloretin-exposed animals also demonstrated improved sodium removal and clearance and a larger sodium dip, yet there was no difference in plasma sodium or removal per liter UF (Table 2).

Selective SGLT1 Inhibition Did Not Affect Glucose or Water Transport during PD

No effect on glucose transport–related parameters was observed using dialysis fluids with the addition of mizagliflozin (Table 1). No significant effects on UF rate, plasma, or dialysate sodium kinetics or small solute transport were observed during mizagliflozin-containing treatment compared with controls (Tables 1 and 2, Supplemental Figure 1, Supplemental Material, Supplemental Table 1).

Discussion

These data show markedly reduced glucose absorption, lower glucose diffusion capacity, and >50% improvement in median UF rate following intraperitoneal phloretin treatment in a rat model of PD. The effects demonstrated herein connect with two key clinical issues in PD: fluid overload and excessive glucose absorption. Volume management is nowadays recognized as a key dimension in the prescription of high-quality PD, whereas small solute clearance is given less attention.²¹ Similarly, increasing evidence shows that glucose absorption is associated with worse outcomes in PD³ and is responsible for significant weight gain and metabolic side effects.² Other than pharmacologic interventions, several strategies exist to reduce glucose absorption and increase UF efficiency in PD: for example, by the use of icodextrin or by altering the prescription.^3,22,23 However, several studies have shown that sodium removal in relation to the UF volume (i.e., millimoles of sodium removed per liter of UF volume) is lower with automated PD,^24,25 which could potentially limit net sodium removal in clinical practice. Phloretin exposure here improved sodium removal, yet there was no effect on sodium removal per liter UF, indicating that the observed improvements are due to enhanced UF rates because sodium transport in PD occurs mainly via convection. Phloretin-treated rats also displayed a larger dip in dialysate sodium, supporting the presence of enhanced UF during intraperitoneal phloretin exposure.

With regard to the mechanisms at work in these experiments, we conclude that taken together with our previous results on SGLT inhibition,^9,12 these results imply a role for GLUT channel–mediated glucose transport during PD and that SGLT1 and -2 have little or no role in these experiments and previous acute experiments.^9,12 Although the presence of GLUT channels in peritoneal tissues was demonstrated earlier, these results are, to our knowledge, the first to suggest their participation in glucose transport during experimental PD. Also, results imply a phloretin effect on urea diffusion capacity, which could possibly be explained by the fact that phloretin is also a nonselective blocker of urea transporters,²⁶ implying a slightly lower urea removal following phloretin treatment.

This study has several important limitations; the limited sample size inflates type 2 error rates, which might limit the ability to detect significant differences between groups: for example, in the mizagliflozin group where no differences were detected. Additionally, the acute nature of the experiments means that no knowledge regarding long-term effects could be obtained. A short dwell time of 60 minutes was chosen as previous experiments with phlorizin indicated maximal effects after 1 hour.¹² The nonselective nature of phloretin makes it difficult to draw conclusions regarding which GLUT channels are involved in peritoneal glucose transport and affected by phloretin exposure. Complementary to previous studies using combined SGLT1 and -2 channel inhibition (phlorizin) or SGLT2 inhibition alone (empagliflozin),^9,12 we here explored glucose transport under the influence of SGLT1 inhibition by mizagliflozin. No effect on any parameter reflecting glucose transport was observed under these experimental conditions. On the basis of our results, a previously observed reduction in glucose kinetics by phlorizin appears as a consequence of phlorizin conversion to phloretin and GLUT blockage. Although results should be assessed with respect to the limitations in the experimental setup and a limited sample size, these results imply little effect of selective SGLT channel inhibition.

Phloretin reduced glucose absorption and improved UF during PD in rats. Future studies should investigate phloretin or similar agents as potential glucose-sparing treatments in clinical PD.

Disclosures

K. Bergling has pursued two master thesis projects with Gambro Lundia AB (unrelated to this work). K. Bergling reports research funding from Baxter Healthcare and patents or royalties with Baxter Healthcare. K. Bergling and C. Öberg are inventors of a pending patent filed by Gambro Lundia AB (Baxter; unrelated to this work). C. Öberg reports research grants (unrelated to this work) from Baxter Healthcare and Fresenius Medical Care and speaker’s honoraria from Baxter Healthcare. C. Öberg reports a consultancy agreement with Baxter Healthcare and an advisory or leadership role with the Peritoneal Dialysis International editorial board. The remaining author has nothing to disclose.

Funding

This work was funded by Lund University Medical Faculty Foundation grant YF 2020-YF0056.

Author Contributions

G. Martus and C.M. Öberg conceptualized the study; K. Bergling was responsible for formal analysis; G. Martus and C.M. Öberg were responsible for methodology; C.M. Öberg was responsible for project administration; K. Bergling and C.M. Öberg were responsible for software; K. Bergling was responsible for visualization; C.M. Öberg was responsible for funding acquisition; C.M. Öberg provided supervision; K. Bergling and G. Martus wrote the original draft; and K. Bergling, G. Martus, and C.M. Öberg reviewed and edited the manuscript.

Data Sharing Statement

Original experimental data reported in this paper have been deposited in Dryad (doi: 10.5061/dryad.x69p8czm8).²⁷

Supplemental Material

This article contains the following supplemental material online at http://jasn.asnjournals.org/lookup/suppl/doi:10.1681/ASN.2022040474/-/DCSupplemental.

Supplemental Material. Intraperitoneal volume and GFR estimation methodology.

Supplemental Table 1. Complementary results.

Supplemental Figure 1. Mizagliflozin effect on volume and small solute transport.