Synergistic effect of uricase blockade plus physiological amounts of fructose-glucose on glomerular hypertension and oxidative stress in rats

Abstract

Fructose in sweetened beverages (SB) increases the risk for metabolic and cardiorenal disorders, and these effects are in part mediated by a secondary increment in uric acid (UA). Rodents have an active uricase, thus requiring large doses of fructose to increase plasma UA and to induce metabolic syndrome and renal hemodynamic changes. We therefore hypothesized that the effects of fructose in rats might be enhanced in the setting of uricase inhibition. Four groups of male Sprague-Dawley rats (n = 7/group) were studied during 8 wk: water + vehicle (V), water + oxonic acid (OA; 750 mg/k BW), sweetened beverage (SB; 11% fructose-glucose combination) + V, and SB + OA. Systemic blood pressure, plasma UA, triglycerides (TG), glucose and insulin, glomerular hemodynamics, renal structural damage, renal cortex and liver UA, TG, markers of oxidative stress, mitDNA, fructokinase, and fatty liver synthase protein expressions were evaluated at the end of the experiment. Chronic hyperuricemia and SB induced features of the metabolic syndrome, including hypertension, hyperuricemia, hyperglycemia, and systemic and hepatic TG accumulation. OA alone also induced glomerular hypertension, and SB alone induced insulin resistance. SB + OA induced a combined phenotype including metabolic and renal alterations induced by SB or OA alone and in addition also acted synergistically on systemic and glomerular pressure, plasma glucose, hepatic TG, and oxidative stress. These findings explain why high concentrations of fructose are required to induce greater metabolic changes and renal disease in rats whereas humans, who lack uricase, appear to be much more sensitive to the effects of fructose.

there is increasing experimental, clinical, and epidemiological evidence that sweetened beverage (SB) consumption increases the risk for obesity, metabolic syndrome, and cardiometabolic and renal consequences (7, 8, 11, 14, 16, 37). Ingestion of SB is also associated with increased risk for nonalcoholic liver disease (2, 4, 23).

Several potential mechanisms have been proposed to explain the mechanisms as how SB may increase the risk for metabolic syndrome, including its high glycemic content and its ability to encourage overnutrition (27, 35). However, some studies have suggested that the fructose component in SB may uniquely increase the risk for metabolic and cardiorenal disorders (26). For instance, data from our group demonstrated that the adverse impact of fructose on renal function and metabolic syndrome was directly related to the amount of fructose administered. (51). In this study, drinking water containing 10% fructose induced mild hypertension in the setting of preserved renal function. In contrast, a diet containing 60% fructose resulted in systemic and glomerular hypertension, renal vasoconstriction, and renal microvascular damage (50). Our group also documented that a high-fructose diet further aggravated renal disease in the 5/6 nephrectomy model of chronic renal damage (19). In addition, mice lacking ketohexokinase (KHK, fructokinase), the first enzyme in fructose metabolism, are also protected from developing metabolic syndrome when exposed to a high-fructose diet (22).

A particular feature of fructose metabolism is the generation of uric acid (UA) as a byproduct (20). While UA has often been viewed as simply a waste product in murine metabolism, there is increasing evidence that UA may have a contributory role in renal damage (29, 40, 49). Experimental studies have shown that, while UA can function as an antioxidant, it also can increase intracellular oxidative stress in various cell types (28, 32, 39). In turn, UA-induced oxidative stress decreases endothelial nitric oxide (NO) biovailability (30, 69, 70). Rats made hyperuricemic by administration of the uricase inhibitor oxonic acid (OA) develop systemic and glomerular hypertension and renal cortical vasoconstriction, which is prevented by blocking oxidative stress or increasing NO substrate, maneuvers that also preserve the renal microvasculature (48, 53, 54).

We therefore hypothesized that the effects of fructose to induce metabolic alterations and renal hemodynamic changes might be enhanced in the setting of uricase inhibition. To test this hypothesis, we administered OA with or without an 11% combination of fructose and glucose in drinking water to rats. The combined use of fructose and glucose was used to be more equivalent with the situation in humans in which fructose is usually delivered either as sucrose or high-fructose corn syrup.

METHODS

Experimental Design

Four groups of male Sprague-Dawley rats (n = 7/group) were studied over a period of 8 wk. Two groups received tap water and the other two groups were offered a SB containing 11% of simple sugars (7.15% fructose and 3.85% glucose, respectively) ad libitum. The proportion of fructose-glucose used in our study was based on recent evidence that major brands of soft drinks use this proportion of fructose-glucose in their products (60). The uricase inhibitor OA was administered by intragastric gavage (750 mg/kg BW, daily) in one tap water (water + OA) and one SB group (SB + OA). In addition, two vehicle (V) groups (water + V and SB + V) were studied (factorial design 2 × 2). For the gastric gavage procedure probes, made of soft polyethylene tube (PE-90, external diameter of 1.7 mm) attached to a 16-G needle and syringe were used, washed, and sterilized daily. It has been showed that the use of soft probes is less stressful for rats and habituation develops (43). Additional rats in all groups (n = 5) were included to better evaluate mitDNA in renal cortex and liver; in these additional animals measurements of fasting plasma glucose and insulin were also performed and included in the analysis for insulin resistance/sensitivity. Experiments were performed in accordance with the Mexican Federal Regulation for Animal Experimentation and Care (NOM-062-ZOO-2001) and were approved by Bioethics and Investigation Committees of the Instituto Nacional de Cardiologia Ignacio Chavez.

Measurements

Body weight was measured weekly. Mean total caloric intake was calculated from the amount of food and beverage consumed in each group of rats. Systolic blood pressure (SBP) was measured in conscious rats by a validated volume-based tail-cuff method (17) (XBP-1000; Kent Scientific, Torrington, CT). All animals were preconditioned for blood pressure measurements 1 wk before each experiment. Fasting (16–18 h) glucose (Genzyme Diagnostics, Boston, MA), insulin (Chrystal Chem, Downers Grove, IL), nonfasting plasma UA (Amplex red; Life Technologies, Carlsbad, CA), and triglycerides (TG; Genzyme Diagnostics) were measured using commercial kits. Homeostasis model assessment-immunoreactivity (HOMA-IR) and quantitative insulin sensitivity check index (QUICKI) were calculated from fasting glucose and insulin. HOMA-IR was calculated as the product of the fasting plasma glucose (FPG) and fasting plasma insulin (FPI) levels, divided by a constant, assuming that control young adult rats have an average HOMA-IR of 1, analogous to the assumptions applied in the development of HOMA-IR in humans (38). The equation was as follows HOMA-IR (FPG − FPI)/2,430, where FPI was in microunits per milliliter and FPG in milligram per deciliter. QUICKI was calculated according to the original formula (31) as the inverse log sum of fasting insulin in microunits per milliliter and fasting glucose in milligram per deciliter. QUICKI 1/[log(FPG) log − (FPI)]. The equations have been found to be accurate in rats. (9). SBP and biochemical parameters were determined at the end of 8 wk. Proteinuria (Bradford method) and plasma and urinary sodium (Flame photometer; Instrumentation Laboratory, Lexington, MA) were measured at the end of the experiment in 16- to 18-h urine collections in metabolic cages. Fractional sodium excretion (FENa) was calculated using standard formulas.

Renal Outcomes

Micropuncture.

Animals were anesthetized with pentobarbital sodium (30 mg/kg ip) and placed on a thermoregulated table to maintain body temperature at 37°C. Trachea, jugular veins, femoral arteries, and the left ureter were catheterized with polyethylene tubing (PE-240, PE-50, and PE-10). The left kidney was exposed, placed in a Lucite holder, sealed with agar, and covered with Ringer's solution. Mean arterial pressure (MAP) was monitored with a pressure transducer (model p23 db; Gould, San Juan, Puerto Rico) connected to the catheter in the femoral artery and recorded on a polygraph (Grass Instruments, Quincy, MA). Blood samples were taken periodically and replaced with the erythrocytes recovered after centrifugation and mixed with isotonic BSA (5 mg/dl) to substitute the plasma volume. Rats were maintained under euvolemic conditions by infusion of 10 ml/kg of body wt of isotonic BSA (5 mg/dl) during surgery, followed by an infusion of 25% polyfructosan, at 2.2 ml/h (Inutest; Fresenius Kabi, Linz, Austria). After 60 min, five to seven samples of proximal tubular fluid were obtained to determine flow rate and polyfructosan concentrations. Intratubular pressure under free-flow and stop-flow conditions and peritubular capillary pressure were measured in other proximal tubules with a servo-null device (Servo Nulling Pressure System; Instrumentation for Physiology and Medicine, San Diego, CA). Glomerular colloid osmotic pressure was estimated from protein concentrations obtained from blood of the femoral artery and surface efferent arterioles. Polyfructosan was measured in plasma and urine samples by the anthrone-based technique of Davidson and Sackner (15). Total glomerular filtration rate (GFR) was calculated using the following formula: GFR = (U × V)/P, where U is the polyfructosan concentration in urine, V is urine flow rate, and P is the polyfructosan concentration in plasma.

The volume of fluid collected from individual proximal tubules was estimated from the length of the fluid column in a constant bore capillary tube of known internal diameter. The concentration of tubular polyfructosan was measured by the microfluorometric method of Vurek and Pegram (63). Single nephron glomerular filtration rate (SNGFR) was calculated using the formula: SNGFR = (TF/P)_PF × V, where PF is the concentration of polyfructosan in tubular fluid (TF) and plasma (P) and V is the tubular flow rate that is obtained by timing the collection of tubular fluid (2). Protein concentration in afferent and efferent samples was determined according to the method of Viets et al. (61). MAP; GFR; glomerular capillary hydrostatic pressure (PGC); single-nephron plasma flow; afferent arteriole, efferent arteriole, and total resistances; and ultrafiltration coefficient (K_f) were calculated according to equations previously reported (6). After the micropuncture study, the left kidney was fixed with 4% paraformaldehyde and the right kidney was divided in cortex and medulla and snap frozen in liquid nitrogen until posterior processing.

Renal histology and quantification of morphology.

Fixed renal tissue was embedded in paraffin and processed accordingly. Evaluation and quantifications were performed blinded.

microvascular damage.

Two-micrometer sections of fixed tissue were stained with periodic acid Schiff reagent. Arteriolar morphology was assessed by indirect peroxidase immunostaining for α-smooth-muscle actin (DAKO, Carpinteria, CA) (59). Sections of kidney tissue incubated with normal rabbit serum were used as negative controls for immunostaining against α-smooth-muscle actin. For each arteriole, the outline of the vessel and its internal lumen (excluding the endothelium) were generated using computer analysis (Image Pro Plus 7.0; Media Cybernetics) to calculate the total arteriolar medial area (outline − inline) in 30 arterioles per biopsy. The media/lumen (M/L) ratio was calculated by the outline/inline relationship.

tubulointerstitial fibrosis.

Sections were stained with Masson's trichrome. Ten noncrossed fields of cortex (640 × 477 mm, 10×) per biopsy were analyzed by light microscopy (Olympus BX51; Olympus American, Melville, NY) and captured with a digital camera (VF Evolution; Media Cybernetics, Silver Spring, MD). Positive blue color areas (excluding glomeruli and vessels) were analyzed in Image Pro Plus (Media Cybernetics).

glomerulosclerosis.

Masson's trichrome-stained renal cortical sections were divided in four quadrants. Segmental and global sclerosed glomeruli were reported as a percentage of the total number of glomeruli counted in one quadrant.

glomerular volume.

In periodic acid Schiff-stained sections the cross-sectional area (A) of 30 representative superficial glomeruli was analyzed; only glomeruli in which a vascular pole was present were included. The individual radius (r) of the glomeruli was determined by r = (A/π)^1/2. The mean glomerular volume (V) was estimated by the following formula: V = 4/3πr³ (36).

Liver Outcomes

At the end of micropuncture experiment, the right lobe of the liver was excised and snap frozen in liquid nitrogen until subsequent analysis for UA, TG, markers of oxidative stress, and Western blotting.

Evaluation of tissue UA, TG, markers of oxidative stress, mitDNA quantification, KHK, and fatty acid synthase protein expressions.

tissue UA.

UA was extracted as previously described (10). In brief, 20 mg of renal cortex or liver were homogenized in buffer (25 mM HEPES, 100 mM KCl, 1 mM DTT, and 0.1 mM EDTA, pH 7.1). Homogenates were frozen in liquid N₂ and unfrozen three times. UA was measured in supernatants obtained after centrifuge using Amplex red assay kit (Life Technologies). Fluorescence was measured on a Synergy H1 hybrid multimode microplate reader using Gen5 analysis software (Biotek Instruments, Winooski, VT). Values of UA were normalized by protein concentration.

tissue TG.

Twenty milligrams of renal cortex or liver were homogenized in a 1-ml solution containing 5% Triton X-100 in water, then heated slowly to 90°C in a water bath for 2–5 min or until Triton X-100 became cloudy, and then slowly cooled down to room temperature (34). This procedure was repeated one more time. Samples were centrifuged for 5 min (top speed in microcentrifuge) to remove insoluble materials. Samples were diluted 10-fold with water before assay using a lipase based colorimetric kit (Genzyme Diagnostics). Absorbance was measured on a Synergy H1 hybrid multimode microplate reader using Gen5 analysis software (Biotek Instruments). TG concentration was calculated by interpolating the values of samples absorbance in a standard curve and normalized by protein concentration.

markers of oxidative stress.

Tissue homogenates were prepared in potassium phosphate buffer (20 mM) containing BHT (0.5 M) and a proteases cocktail (halt protease inhibitor; Thermo Scientific, Waltham, MA), pH 7.0. For protein carbonyl content, homogenates were incubated overnight with streptomycin sulfate to remove nucleic acids. Later, homogenates were treated with 2,4-dinitrophenylhydrazine and HCl and finally with guanidine hydrochloride. Assessment of carbonyl formation was done on the basis of formation of protein hydrazone by reaction with 2,4-dinitrophenylhydrazine. The absorbance was measured at 370 nm (46). Protein carbonyl content was expressed as nanomoles of carbonyl per milligrams of protein. Lipid peroxidation was assessed by measuring 4-hydroxynonenal (4-HNE) using a standard curve of tetramethoxypropane. A solution of 1-methyl-2-phenylindole in a mixture of acetonitrile:methanol (3:1) was added to the homogenates, and the reaction was started by adding 37% HCl. Optical density was measured at 586 nm after 1 h of incubation at 45°C (18). Data were expressed as nanomoles per milligrams of protein.

assay for relative mitDNA copy number.

We evaluated the effect of the different treatments in renal cortex and liver mitDNA. Micropuncture and additional animals (n = 7 + 5 = 12 each group) were included for this analysis. Extra rats were anesthetized with sodium pentobarbital after 16–18 h fasting. Blood samples were taken from abdominal aorta for glucose and insulin measurements, and renal cortex and liver were quickly excised and snap frozen in liquid nitrogen and stored at −86°C until mitDNA analysis. Total DNA was isolated using Quick-gDNA MiniPrep (Zymo Research, Irvine CA). Primers and probes for the rat D-loop (mitDNA copy number) were previously reported (55). mitDNA was compared against the nuclear gene for 18S rRNA. Each sample was assessed in triplicate, and fluoresce spectra were continuously monitored by the ABI-Prism 7300 Sequence Detection System (Applied Biosystems, Carlsbad, CA) with sequence detection software version 1.3.1. Data analysis was based on measurement of the cycle threshold (C_T). As a measure of the relative expression of D-Loop mitDNA among groups we took the ΔC_T mean value of control group as calibrator and obtained the ΔΔC_T values. Results are reported as the relative mitDNA copy number (calculated as 2^−ΔΔCT) (42).

KHK and fatty acid synthase protein expression by Western blot analysis.

Hepatic and renal cortex proteins were extracted using a MAP kinase lysis buffer, as previously described (47). Thirty to forty micrograms of liver homogenate samples were resolved on Ready Gel Tris-HCl precast gels (12% or 4 to 15%; Bio-Rad Laboratories, Hercules, CA) and transferred to polyvinylidenedifluoride membranes by electroblotting. Each primary antibody was incubated at 4°C overnight. Anti-KHK (GeneTex, Irvine, CA; liver and renal cortex), anti-fatty acid synthase (anti-FAS; Cell Signaling, Danvers, MA; liver), and anti-β-actin antibody (Cell Signaling; liver and renal cortex). After being washed with Tween-TBS, the membrane was rocked with secondary antibody (anti-rabbit IgG, horseradish peroxidase-linked antibody; Cell Signaling). Blots were then developed using the Immun-Star HRP chemiluminiscence kit (Bio-Rad Laboratories). Chemiluminescence was recorded and quantified using the ID image-analysis system software (Kodak Digital Science, Rochester, NY).

Statistical Analysis

Values are expressed as means ± SD. Significant differences between treatment groups were determined by two-way ANOVA. When the ANOVA P value was <0.05, posttest comparisons were made using a Bonferroni multiple-comparison test. The relationship between variables was assessed by correlation analysis. Statistical analysis was performed with Prism version 5.04 (Graph Pad Software, San Diego, CA).

RESULTS

Effect of Uricase Inhibition With or Without SB on Metabolic Alterations

OA and SB treatments induced the development of metabolic alterations.

As we reported previously, OA induced a significant increment of plasma UA in water groups (+400%). The effect of SB alone on plasma UA was also significant but milder compared with OA (+173%). The cotreatment of SB + OA also significantly increased UA plasma levels (+400%); this rise was comparable to the W + OA group (Table 1).

Table 1.

Effect of the single treatments and in combination on metabolic syndrome traits

	W + V	W + OA	SB + V	SB + OA	Treatment	Beverage	Interaction
Uric acid, mg/dl	0.19 ± 0.02	1.03 ± 0.12‡	0.52 ± 0.17	1.01 ± 0.08‡	<0.0001	<0.01	<0.0001
TG, mg/dl	78 ± 24	109 ± 42	125 ± 35	156 ± 43	<0.05	<0.01	NS
Glucose, mg/dl	79 ± 23	88 ± 28	105 ± 26	151 ± 45†	<0.001	<0.0001	<0.02
Insulin, ng/ml	0.47 ± 21	0.48 ± 0.86	0.65 ± 0.33	0.58 ± 0.34	NS	<0.03	NS
HOMA-IR	2.2 ± 1.3	2.6 ± 1.4	4.2 ± 2.7	4.6 ± 1.7	NS	<0.0001	NS
QUICKI	0.69 ± 0.13	0.66 ± 0.13	0.58 ± 0.9	0.58 ± 0.34	NS	<0.0001	NS
SBP, mmHg	122 ± 8	141 ± 3†	137 ± 5	147 ± 2*	<0.0001	<0.0001	<0.03
BW, g	342 ± 17	327 ± 22	337 ± 19	353 ± 33	NS	NS	NS
Uprot, mg/day	7 ± 2	9 ± 3	4 ± 4	9 ± 10	NS	NS	NS
FENa, %	0.49 ± 0.17	0.72 ± 0.34	0.26 ± 0.13	0.29 ± 0.26	NS	<0.0001	NS

Values are means ± SE. W, water; V, vehicle; OA, oxonic acid; SB, sweetened beverage; TG, triglycerides; HOMA-I, homeostasis model assessment-immunoreactivity; QUICKI, quantitative insulin sensitivity check index; SBP, systolic blood pressure; FENa, fractional sodium excretion; BW, body weight; Uprot, urinary protein; NS, nonsignificant.

^*P < 0.01 vs. V.

^†P < 0.001 vs. V.

^‡P < 0.0001 vs. V.

Rats with OA-induced hyperuricemia and rats with chronic SB consumption both demonstrated a significant and similar increment of postprandial plasma TG (OA +39% and SB +60%) at 8 wk. The combination of both treatments had a tendency for a further increase plasma TG with respect to control animals (+200%) but was not significant different from SB alone (+25%; Table 1). We found a significant effect of treatment and beverage in fasting (16–18 h) glucose (Table 1) In addition, fasting plasma glucose was significantly raised by the cotreatment inducing a synergistic effect and causing a further increase in glucose compared with SB alone (+44%). We found a significant effect of SB alone on plasma insulin concentration, HOMA-IR, and QUICKI compared with both groups receiving tap water (Table 1).

Both OA treatment and SB were associated with a significant increment in systolic blood pressure compared with control (OA +16% and SB +12%); combination of both treatments induced a further and significant increase of 7% respect to SB alone and a slight increment compared with OA alone (+4%). Finally, the mean total caloric intake was not different among the groups (W + V: 91.1 ± 4.5 Kcal; W + OA: 91.7 ± 5.4 Kcal; SB + V 91.1 ± 6.2 Kcal; and SB + OA: 92.5 ± 14.8 Kcal; P = NS); this was mirrored by a similar body weight gain during the follow-up (Fig. 1). Rats on SB consumed less solid food in compensation for the calories added in drinking water; in those groups SB contributed 26.5 ± 1.4% of the total caloric intake, with the fructose component providing 17.2 ± 0.9% of the overall energy intake.

Fig. 1.

Weekly body weight gain in rats receiving tap water with or without the uricase inhibitor oxonic acid during 8 wk. Body weight gain was similar along the follow-up in the 4 experimental groups. W, water; Veh, vehicle; OA, oxonic acid; SB, sweetened beverage.

Renal Outcomes

Twenty-four hour urinary protein excretion was not modified by OA treatment nor SB after 8 wk of follow-up. On the other hand, SB severely reduced urinary sodium excretion in V- and OA-treated animals. Both SB groups (V or OA treated) excreted ∼30% of the sodium relative to V- or OA-treated rats receiving water (Table. 1).

OA-induced hyperuricemia has a primary role in causing renal vasoconstriction.

MAP measured during micropuncture in anesthetized rats receiving OA was increased compared with water-treated rats (+20%) and mirrored the changes noted by SBP that was measured using a validated tail-cuff method (Fig. 2). In contrast, SB alone had a mild, but significant, effect on MAP compared with control rats (+8%). The coadministration of SB + OA further increased MAP compared with SB alone (+18%) and vs. OA alone (+6%). OA-treated rats showed a 17% reduction in whole GFR, while in SB + V animals GFR was similar to control group. The concomitant treatment of SB + OA significantly reduced GFR vs. SB + V (−27%).

Fig. 2.

Rats receiving tap water with or without the uricase inhibitor oxonic acid during 8 wk were submitted to renal hemodynamics evaluation at the end of the experiment. Mean arterial pressure, whole glomerular filtration rate (GFR), glomerular pressure, single nephron GFR, glomerular plasma flow, ultrafiltration coefficient, and afferent and efferent resistances were evaluated after 8 wk of follow-up. *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 2 depicts the glomerular hemodynamic data. We confirmed a significant renal vasoconstrictive effect mediated by OA; therefore, single nephron GFR (−10%), glomerular plasma flow (−20%) and ultrafiltration coefficient (−47%) were significantly reduced and coupled with higher afferent (+61%) and efferent (+72%) arteriolar resistances compared with the control group receiving water alone. In contrast, administration of a SB + V for 8 wk showed minimal changes in cortical glomerular function as evaluated by renal micropuncture, single nephron GFR, glomerular plasma flow, ultrafiltration filtration coefficient, glomerular pressure, as well as afferent and efferent resistances were comparable to control rats (Fig. 2). The cotreatment of SB + OA did not cause further changes in those parameters compared with W + OA (Fig. 2).

SB + OA accelerates glomerular hypertension.

Earlier our group showed that 5 wk of OA-induced hyperuricemia produced a significant increment of PGC associated with renal vasoconstriction (Fig. 2) (49). The present study confirmed that this effect is maintained after 8 wk of OA administration and was associated with a 21% greater PGC compared with rats that received tap water alone. On the other hand, SB alone had a mild, yet significant, effect on PGC compared with control rats (+4%), and coadministration of SB + OA further increased PGC vs. SB alone (+33%) and vs. OA alone (+14%).

SB is associated with renal hypertrophy and more structural lesions.

OA treatment resulted in a slight increment in glomerular volume compared with control animals (+13%), similar to what we previously reported (41) (Table 2). However, both V- and OA-treated rats receiving SB demonstrated marked glomerular hypertrophy (increment of glomerular volume of 77 and 67%, respectively, vs. their counterparts) and arteriolar wall area (+54 and +65%, respectively, vs. their counterparts). In addition, tubulointerstitial fibrosis and glomerulosclerosis were increased by SB (Table 2). Neither treatment nor SB induced alterations in renal cortex relative mitDNA copy number (W + V 1.08 ± 0.39; W + OA 1.28 ± 0.38; SB + V. 1.18 ± 0.55; and SB + OA 1.28 ± 0.47).

Table 2.

Effect of the single treatments and in combination on renal histology

	W + V	W + OA	SB + V	SB + OA	Treatment	Beverage	Interaction
Glomerular volume, μm3	319 ± 13	363 ± 38	565 ± 32	605 ± 53	<0.05	<0.0001	NS
Arteriolar wall area, μm2	197 ± 21	193 ± 38	302 ± 39	320 ± 41	NS	<0.0001	NS
M/L ratio	1.52 ± 0.16	2.59 ± 0.60*	2.62 ± 0.31	3.41 ± 0.91	<0.001	<0.001	NS
Glomerulosclerosis, %	0.68 ± 0.79	3.90 ± 2.18	6.23 ± 5.99	7.04 ± 3.69	NS	<0.01	NS
TI fibrosis, %	2.50 ± 2.43	4.71 ± 4.61	8.67±.20	11.86 ± 6.57	NS	<0.001	NS

Values are means ± SE. M/L, media/lumen; TI, tubulointerstitial.

^*P < 0.01 vs. V.

Microvascular damage, evaluated as the M/L ratio, was also increased in OA-treated rats compared with rats receiving water (+71%). SB + V also induced a similar increment of the M/L ratio compared with control (+73%). OA administered with SB increased the M/L ratio by 30% more vs. SB alone, but this difference did not reach statistical significance (Table 2).

OA-induced hyperuricemia results in higher renal cortex UA levels and increased oxidative stress.

Treatment with OA significantly increased renal cortical UA concentrations in both W and SB groups (+250 and +61% compared with their counterparts, respectively; Fig. 3). Compared with SB alone, SB + OA increased renal UA content by 85%. SB in tap water drinking rats had a milder and nonsignificant effect on renal UA content and increased it by 80% compared with W + V. Renal UA levels correlated with values of plasma UA (r = 0.84; P < 0.0001).

Fig. 3.

Uric acid was extracted from renal cortical tissue and measured using the Amplex red kit. The markers of oxidative stress protein carbonyls and 4-hydroxynonenal (4-HNE) were also measured in renal cortex. Hepatic fructokinase (KHK) protein expression was evaluated in renal cortex using β-actin as load control. ***P < 0.001; ****P < 0.0001.

Both OA treatment alone and SB treatment alone increased protein carbonylation and lipid peroxidation (4-HNE) in the renal cortex compared with the W + V group. In addition, a significant synergistic effect in lipid peroxidation was observed in rats administered OA and SB.

SB increases KHK protein expression.

Administration of SB in V- and OA-treated groups induced a significant and similar increment in kidney cortex KHK (Fig. 3).

Liver Effects

OA and SB increase hepatic UA and TG.

Both OA treatment alone, and SB alone, increased liver UA content, with OA causing a greater increment than SB (Fig. 4). Coadministration of SB + OA did not further increased UA hepatic concentration compared with W + OA animals. Likewise, OA and SB alone significantly increased liver TG; however, OA alone had a more robust effect compared with SB alone. Combination of both treatments had a synergistic effect in the SB + OA group, and liver TG doubled their concentration compared with OA animals receiving tap water.

Fig. 4.

Liver outcomes. Hepatic uric acid was extracted and measured with Amplex red kit. In addition hepatic triglycerides, markers of oxidative stress and ketohexokinase and fatty acid synthase protein expressions were also evaluated after 8 wk of follow-up with the different treatments. *P < 0.05; **P <0.01; ***P < 0.001; ****P < 0.0001.

OA and SB increase liver oxidative stress.

OA and SB treatments increased hepatic protein carbonylation and lipid peroxidation in tap water drinking groups (Fig. 4). SB + OA-treated rats showed a greater effect in oxidized proteins than SB treatment alone. Lipid peroxidation was also increased in OA- and SB-treated rats and the combination showed a synergistic effect. In addition, we found significant correlations between: liver UA content and protein carbonyls (r = 0.72; P < 0.0001), liver UA and 4-HNE (r = 0.69; P < 0.0001), and liver TG and 4-HNE (r = 0.92; P < 0.0001). Neither treatment nor SB induced alterations in hepatic mitDNA relative copy number (W + V 1.02 ± 0.50; W + OA 1.20 ± 0.82; SB + V 1.45 ± 0.79; and SB + OA 1.46 ± 0.75).

SB increases KHK and FAS protein expression.

Administration of SB in V- and OA-treated groups induced a significant and similar increment in liver KHK and FAS expressions (Fig. 4).

DISCUSSION

The present study has several important findings. First, we demonstrate that chronic hyperuricemia alone extended for 8 wk can induce renal and metabolic alterations, including hyperuricemia, postprandial hypertriglyceridemia, systemic and glomerular hypertension, and hepatic TG accumulation. Second, we found that SB containing clinically relevant concentrations of fructose and glucose could also induce mild hyperuricemia and similar metabolic manifestations as animals receiving OA alone. Finally, we demonstrated that SB plus uricase inhibition induced a combined phenotype including all the deleterious changes induced by OA and SB independently and, in addition, also acted synergistically on systemic and glomerular pressure, plasma glucose, hepatic TG, and renal and liver lipid peroxidation. There was also a tendency for the combination of SB and OA to induce a greater rise in plasma TG. Thus the novelty of our study relies on the demonstration that inhibiting uricase increases the effect of fructose to enhance hepatic fat stores, increase serum glucose, increase systemic blood pressure, and produce renal damage. We believe the clinical implication may be explained by the fact that humans who lack uricase are likely more sensitive to the effects of fructose than mammals that do not express uricase. In turn, this can explain why high concentrations of fructose are required to induce greater metabolic changes and renal disease in rats whereas humans appear to be much more sensitive to the effects of fructose (25). In this regard, from the beverage intake, the mean fructose consumption in SB groups was 17% of total caloric intake; this level of fructose consumption is observed in the 90th percentile of American general population (62). Nevertheless, we are aware that care should be taken in extrapolating experimental data to humans.

Renal Alterations Associated with SB Intake With or Without Uricase Inhibition

A key finding in these studies was the observation that SB + OA produced a significant rise in glomerular hypertension to levels comparable to what is observed in models of more extensive chronic renal damage such as the 5/6 nephrectomy model (59). Indeed, the levels of glomerular pressure induced by SB + OA were significantly higher vs. SB alone (+33%) and vs. OA alone (+14%). In contrast, in SB + V animals we observed minimal changes in cortical glomerular hemodynamics, as evaluated by renal micropuncture; thus single nephron GFR, glomerular plasma flow, ultrafiltration filtration coefficient, glomerular pressure, as well as afferent and efferent resistances were comparable to control rats (Fig. 2). In addition, oxidative stress (lipid peroxidation) was significantly increased by the combination of SB + OA compared with each treatment alone. On the other hand, despite no evident glomerular hemodynamics alterations, SB per se did produce glomerular hypertrophy, arteriolopathy, glomerulosclerosis, and tubulointerstitial fibrosis. It is likely that additional mechanisms stimulated by SB, besides UA increment, act to induce mild structural renal damage. For example, it has been reported that bosentan (12) and lacidipine (13) treatments protect against renal damage induced by high fructose, implicating a role for endothelin and L-type calcium channels in this process. Moreover, in SB animals a lower increase in oxidative stress compared with OA groups was observed. In summary, the rise in systemic BP likely interacted with impaired autoregulation induced by the renal oxidative stress (67) and microvascular arteriolopathy (52) to lead to the increased glomerular hypertension in SB + OA animals. Therefore, the combination of SB and OA treatment significantly accelerates renal damage progression, compared with each treatment given alone.

An additional intriguing finding was that FENa was found to be lower in SB groups compared with tap water drinking groups. This result is partially explained by the fact that SB animals had a reduction of food consumption to compensate for the extra calories ingested in the SB. There are other factors that also may contribute to reduce Na excretion and FENa: 1) for example, dietary fructose has been associated with a reduction in renal medullary blood flow and this mechanism was associated with decreased FENa (44). 2) An additional reduction in sodium excretion and FENa in SB groups might be explained by the increased expression of the apical chloride/base exchanger Slc26a6 (PAT1) and the Na⁺/H⁺ exchanger NHE3 (sodium hydrogen exchanger-3) in the jejunum and renal proximal tubule induced by fructose (5, 56), thereby increasing sodium and water absorption in small intestine and kidney.

Effect of SB With or Without Uricase Inhibition on Metabolic and Liver Alterations

An important finding in this study was that 8 wk of SB containing clinically relevant concentrations of fructose and glucose could induce mild hyperuricemia and mild postprandial hypertriglyceridemia, insulin resistance, and hepatic TG accumulation. These studies are consistent with the increasing literature linking SB intake with metabolic syndrome traits in humans (21). However, a more interesting finding was that rats receiving OA alone also developed features of metabolic syndrome, with higher fasting glucose, postprandial hypertriglyceridemia, and higher hepatic TG stores. The latter effect could be the consequence of increased oxidative stress we observed in the liver. UA has also been described as conditional prooxidant that can induce peroxidation of low density lipoproteins (1). This report is consistent with this study in which we found increased lipid peroxidation and protein carboxylation in the livers of OA-treated rats and which correlated closely with hepatic UA levels. We recently found that UA can directly increase TG accumulation in liver (HepG2) cells via a mechanism that involves mitochondrial oxidative stress (33). In the present studies we did not observe an increment on hepatic fructokinase (KHK) and FAS levels. However, we cannot rule out that an increase in enzyme activities and likely other mechanisms are involved in the liver and plasma increase of TG observed in these animals. Many years ago Wexler (65) and Wexler and Greenberg (66) also reported that chronic uricase inhibition could induce features of metabolic syndrome, but these findings were largely ignored because of the high doses of uricase inhibitor used. This may need to be revised in light of recent studies showing that high serum UA levels can predict the development of metabolic syndrome in humans, and the increasing experimental evidence that UA may have a role in metabolic syndrome (25).

An intriguing finding was that we did not demonstrate an effect on insulin levels or insulin resistance (HOMA-IR and QUICKI) despite mild but significant effects on fasting glucose levels. In this regard, we have found that a high-sucrose diet induced pancreatic inflammation and mild islet injury and reduced insulin levels with the development of type 2 diabetes mellitus, effects that were partially mediated by UA (47). Although speculative, we believe that in SB + OA rats mild pancreatic damage could also occur with increased susceptibility to UA-mediated damage, therefore diminishing insulin synthesis and secretion, allowing fasting plasma glucose to increase despite almost normal insulin levels.

Some synergistic metabolic effects were observed when SB was administered to rats receiving OA. The combined treatment resulted in higher plasma glucose, hepatic TG, and hepatic lipid peroxidation. In this respect, chronic exposure to diets enriched with sucrose or fructose induces hepatic insulin resistance and increased hepatic gluconeogenesis (45, 64), effects partially attributed to an increased activation of JNK, which in turn interferes with proximal steps in the insulin-signaling pathway (3). Chronic oxidative stress and 4-HNE act synergistically to trigger JNK overactivation in hepatocytes (57). In this study we observed that OA substantially enhanced 4-HNE formation in SB drinking rats; therefore, we speculate that in liver this effect might contribute to increase liver insulin resistance, gluconeogenesis, and glycogenolysis, as suggested by the marked increment in fasting plasma glucose in these animals.

An intriguing finding was that SB + OA did not further increase UA concentrations in plasma or tissue compared with W + OA animals. This effect likely relates to the timing of the UA measurement. Thus Stavric et al. (58) reported that the acute administration of OA could dramatically enhance an acute rise in UA in response to fructose. It is known that the effects of OA induce a rise in UA that peaks at 6 h and then decreases over time; we dosed OA to rats during the morning while majority of SB consumption is at night, due to the nocturnal rat feeding habits, so it is possible that we missed the time point when this synergy would have been maximally demonstrated. In addition, it is also possible that the synergy observed with UA and fructose may decrease over time, either due to a compensatory increase in uricase that reduces the efficacy of the uricase inhibitor (68) or by the fact that UA and its precursors may feedback to block xanthine oxidase activity (9–10). Despite our inability to show a synergy of fructose and OA on UA levels in our study, we did show a synergy for oxidative stress in the kidney and liver (lipid peroxidation). In addition, we also found a synergy of fructose and OA on glomerular hydrostatic pressure.

In summary, we document the contribution of hyperuricemia to the deleterious effects induced by the increased consumption of added sugars on diet. These additive effects are likely fully operative in humans, who lack the expression of uricase. The documentation that uricase inhibition can increase hepatic fat content and serum TG is consistent with our hypothesis that the uricase mutation occurred in the Miocene provided a survival advantage to apes starving in Europe during a period of climatic cooling that resulted in food shortage due to a loss of fruits (24). Today, with a dramatic increase in fructose, the loss of uricase may enhance the risk for humans to develop metabolic syndrome and cardiorenal disease.

GRANTS

Funding was supported in part by National Heart, Lung, and Blood Institute Grant HL-068607-10 and National Council of Science and Technology (CONACyT) Mexico Grants 133232 and No. 167949.

DISCLOSURES

R. J. Johnson is listed as an inventor on a patent from the University of Washington for use of allopurinol to treat hypertension and is also an inventor on patent applications from the University of Florida, Takeda, and University of Colorado for use of UA-lowering agents or agents that block fructose metabolism in the management of metabolic and renal disorders. R. J. Johnson is the author of two lay books on fructose, The Fat Switch (mercola.com, 2012) and The Sugar Fix (Rodale, 2008). R. J. Johnson has also consulted for Ardea, Astellas, Biocryst, Danone, and Novartis and is on the Scientific Advisory Board for Amway.

AUTHOR CONTRIBUTIONS

Author contributions: E.T., M.A.L., C.A.R.-J., T.I., M.M., R.J.J., and L.G.S.-L. conception and design of research; E.T., M.C., F.E.G.-A., V.S., F.M.-S., U.P., D.C.-R., and L.G.S.-L. performed experiments; E.T., V.S., M.A.L., C.A.R.-J., T.I., M.M., R.J.J., and L.G.S.-L. analyzed data; E.T., M.C., F.E.G.-A., V.S., F.M.-S., U.P., M.A.L., C.A.R.-J., D.C.-R., T.I., M.M., R.J.J., and L.G.S.-L. edited and revised manuscript; E.T., R.J.J., and L.G.S.-L. approved final version of manuscript; M.C., F.E.G.-A., V.S., F.M.-S., U.P., D.C.-R., R.J.J., and L.G.S.-L. interpreted results of experiments; R.J.J. and L.G.S.-L. drafted manuscript; L.G.S.-L. prepared figures.