Acidic polyanion poly(acrylic acid) prevents calcium oxalate crystal deposition

Abstract

Acidic macromolecules inhibit calcium oxalate nucleation, growth, aggregation and attachment to cells in vitro. To test for such an effect in vivo we used osmotic minipumps to continuously infuse several doses of the 5.1 kDa poly(acrylic acid) (pAA_5.1) into rats fed a diet which causes renal calcium oxalate crystal deposition. Although kidneys of rats receiving the saline control contained calcium oxalate crystals, measured by polarized light microscopy, those of animals given pAA_5.1 had significantly lower numbers of crystals in various zones of the kidney. Delivery of pAA_5.1 to urine was confirmed by measuring excretion of infused biotinylated pAA_5.1. Both the derivatized and unlabelled pAA_5.1 had the same effects on crystallization in vitro. Our study shows that acidic polymers hold promise as effective therapies for kidney stones likely through prevention of calcium oxalate crystal aggregate formation.

Urinary macromolecules appear to influence kidney stone formation at several of the steps in their formation. The hypothesis tested in the studies reported in this article is that manipulation of the interaction between urinary macromolecule and crystalline stone material can influence crystal deposition in one experimental model of the initiation of kidney stone formation.

The first step in stone formation is the nucleation of crystalline components. This may occur in the lumens of renal tubules, in the basement membranes of tubule cells, or at both sites, perhaps depending on the type of stone or the underlying conditions that predispose to stone formation.^1,2 Whether growth of the solid phase is required for retention within the kidney depends on several factors These include whether the site of nucleation is within the tubule lumen or outside of it, or whether a specific attachment mechanism to cells or extracellular matrix is present for stone crystals. What is obvious, however, is that such growth must occur for a stone to achieve clinically a significant mass.

This growth appears to be largely a consequence of aggregation of crystals that have nucleated individually in tubular or interstitial fluid, or growth may result from secondary nucleation of new crystals on the surface of those previously formed, regardless of the site of initial crystal fixation in the kidney. In particular, the site of fixation is immaterial to this concept; it may occur either at the apical membranes of tubule cells or within their basement membranes. Inspection of scanning photomicrographs of stones reveals their multicrystalline nature,³ as well as the presence of a matrix that appears to act like a binding agent akin to the function of Portland cement to bind together silica in concrete. Stone matrix contains numerous urinary macromolecules; the most prominent have been highly acidic proteins and glycosaminoglycans.⁴

Acidic urinary macromolecules that have been studied individually interact strongly with crystals and generally inhibit crystal formation, especially those of calcium oxalate (CaOx), in vitro.^5–12 Synthetic polyanions have similar effects on CaOx formation; however, polycations and neutral polymers are, in general, ineffective.^13–17 Among the synthetic polyanions, those with carboxylate side chains are more potent at inhibiting crystal formation than phosphoryl or sulfonyl, in that order. The effects seem to be insensitive to the backbone of the molecule; however, charge group spacing may be important. These synthetic polyanions affect several processes in vitro that may have a bearing on stone formation.

Acidic polyanions inhibit attachment of calcium oxalate monohydrate (COM) to cultured inner medullary collecting duct cells, change crystal habit from COM to calcium oxalate dihydrate (COD) in spontaneous crystallization assays, decrease step velocity of COM growth in constant composition experiments, exhibit strong interaction with multiple crystal surfaces by atomic force microscopy, and disaggregate COM seed crystal aggregates in experiments employing particle sizing.^16,18 Poly(acrylic acid) (pAA) is among the most potent of these molecules¹⁹ (unpublished data). The specific purpose of the current studies was to determine whether infusions of 5.1 kDa pAA (pAA_5.1) are effective in preventing CaOx crystal deposition in rats fed with ethylene glycol.

RESULTS

Inhibition of crystal deposition

Figure 1 shows the results of the pAA_5.1 infusions at various rates into rats on a crystal inducing diet (CID; 0.8% ethylene glycol with 1% NH₄Cl in tap water) compared with controls on the same diet receiving only saline infusions. With saline alone, tissue sections of the rat kidneys showed significant crystal deposition in all three zones of the kidney. Presumably, these were largely comprised of COM, as they could be visualized with partially polarized light. Using similar feeding schedules, we and others^20,21 have induced COM deposition that has been verified by elemental or crystallographic techniques. Rates of pAA_5.1 infusion of 50–300 μg/h inhibited COM deposition significantly. At the rates tested, the degree of inhibition was between 81 and 98%, depending on the anatomical zone (cortex, outer medulla, or inner medulla) of the kidney examined, and was without clear dose dependence. Lower infusion rates of pAA_5.1 did not differ from saline controls, although the 10 and 20 μg/h rates showed a trend toward inhibition (results not shown).

Figure 1. Effect of pAA_5.1 on retention of CaOx in kidneys of rats fed CID.

Values are means ± s.e. of control rats receiving saline (Cont) through minipumps or rats receiving the indicated infusion rates of pAA_5.1. Analysis of variance indicates an F statistic of 3.01 with a P-value of 0.03. The values for the groups receiving pAA_5.1 are all significantly different from Cont P≤0.005. None of the mean values for the different pAA_5.1 infusion groups were significantly different from the others.

The rats infused with pAA and receiving the CID did not differ from saline controls on the same diet in either plasma Ca or creatinine concentrations or urinary excretions of either calcium or oxalate (Table 1).

Table 1.

Plasma and urine creatinine, calcium, and oxalate values of rats on CID

	Plasma		Urine
	Creatinine (mg/100 ml)	Calcium (mg/100 ml)	Calcium/creatinine (mg/mg)	Oxalate/creatinine (mg/mg)
Saline controls	0.69 ± 0.06 (21)	10.3 ± 0.2 (21)	0.18 ± 0.01 (7)	0.71 ± 0.08 (7)
pAA5.1	0.61 ± 0.06 (30)	10.6 ± 0.1 (30)	0.24 ± 0.05 (7)	0.54 ± 0.13 (7)

Values are means ± s.e. for (N); the values for rats receiving saline in minipumps were not different from those receiving pAA_5.1. Plasma creatinine and calcium concentrations did not differ among the various rates of infusion of pAA_5.1 (not shown), so results have been combined. Urine pAA_5.1 calcium and oxalate values are obtained from animals infused with the 300 μg/h dose.

Properties of biotinylated pAA_5.1 in vitro

Bt- pAA_5.1, 1 μg/ml, mimicked the effects of unlabeled pAA_5.1 in promoting crystallization of COD over COM, inducing ~85% COD compared with 100% COM without the polymer, as determined by optical microscopy.²² In addition, in a static aggregation assay, 0.5 μg/ml (equivalent to 100 nM) bt-pAA_5.1 produced the same amount of disaggregation of COM seed crystals (R_D = 0.73, where a value of <1.0 indicates disaggregation and >1.0 indicates aggregation) comparable with unmodified pAA_5.1 (R_D = 0.74).

Recovery and properties of infused bt-pAA_5.1

We administered bt-pAA_5.1 by osmotic minipump at 10 μg/h to four rats. On day 9, the rats had a mean urinary bt-pAA_5.1 excretion of 142 ± 94 μg, which is equivalent to a bt-pAA_5.1 concentration of ~500 nM. Urine from two rats given tap water that were infused with 50 μg/h was added to the COM aggregation assay in an amount normalized to 10 μg of creatinine and proved effective at disaggregating seed crystal. This was the case both in a rat whose urine before infusion promoted aggregation and in a rat whose baseline urine inhibited aggregation (Figure 2). Progressive dilution of the urine collected from the rat whose urine promoted aggregation before the infusion, which provided progressively decreasing concentrations of bt-pAA_5.1, demonstrated significant disaggregation of seed crystal down to somewhat below a concentration of 1 nM (Figure 3).

Figure 2. Effect of urine from two rats infused with Bt-pAA_5.1 at 50 μg/h on COM aggregation.

Urine collected on the day before the infusion (Day 0) as well as on Day 7 and/or Day 9. The urine was added to an in vitro aggregation assay, as described in the Methods section, in amounts containing 10 μg of creatinine. Values are R_D, the ratio of the final average particle diameter to the diameter of the initial seed COM crystal.

Figure 3. Effect of urinary Bt-pAA_5.1 on COM aggregation.

Urine before the Bt-pAA_5.1 infusion in rat 2 (■) shown in the previous figure was tested in an amount containing 10 μg of creatinine; the day-9 urine was diluted from its measured concentration to the concentrations indicated on the abscissa. The trend line remains flat to the highest concentration studied (50 nM), which is not shown. Values are R_D, the ratio of the final average particle diameter to the initial seed COM crystal.

DISCUSSION

The process of formation of kidney stones includes nucleation of stone-constituent crystals, which may occur either within the tubules or in the interstitium. In the case of intratubular nucleation, some investigators have proposed and developed evidence for a specific attachment process to renal tubular cells.^23–27 Alternatively, crystals may enlarge individually through growth or may aggregate with others, resulting in fixation within the kidneys on the basis of size alone.²⁸ In either case, whether fixation occurs at the level of the individual crystal or small crystal aggregates, further enlargement to clinically significant size appears to take place through addition of free-floating crystals or nucleation of crystals secondarily on the surface of aggregates already fixed to tissue.

As was pointed out in the introduction, pAA shares many of the properties of endogenous urinary acidic macromolecules that have been shown to influence crystal formation, either in vitro or, by inference, in animal experiments.^20,29 These include both inhibition of attachment of individual crystals and the aggregation of preformed crystals. If, as proposed above, attachment and aggregation are critical processes that lead to crystal retention, pAA might be expected to prevent crystal retention.

The studies reported in this communication demonstrate that pAA_5.1 infusions into rats fed with a mixture of ethylene glycol and NH₄Cl prevent the accumulation in their kidneys of CaOx crystals. Ethylene glycol administration in rodents has been used as an experimental model for the initiation of kidney stones for a number of decades.²¹ It has been criticized on a number of bases, including that it induces injury in kidneys and that it is more relevant to primary hyperoxaluria rather than idiopathic or hypercalciuric stone disease. However, if given in sufficiently small amounts, ethylene glycol has been shown to induce crystalluria and CaOx deposition in rodent kidneys without inducing hyperoxaluric renal failure, although enzymuria and tubular damage is likely present.^21,30 In the studies reported here, at least with regard to glomerular filtration, there did not appear to be significant renal damage, as the serum creatinine values in both control and experimental groups were not different from those of untreated rats.

Whether pAA_5.1 prevents crystal deposition by inhibiting crystal aggregation or by weakening its attachment to renal epithelial cells cannot be ascertained from these studies; however, in vitro assays show evidence that the excretion of pAA_5.1 reaches a sufficient level to influence aggregation. The studies with pAA_5.1 labeled with biotin suggest that pAA_5.1 is excreted in the urine in amounts sufficient to confer on the urine the capacity to disaggregate COM seed crystals in vitro if not already present, specifically, in a rat whose urine promoted aggregation before the infusion. In a rat that was already excreting urine that disaggregated COM seed crystals, the bt-pAA_5.1 infusion led to excretion of urine with slightly greater disaggregating ability.

The effect does not appear to be explainable by effects of pAA_5.1 on renal delivery of calcium, as there were no differences between saline controls and groups infused with pAA_5.1 that were both on the CID in their plasma creatinine or calcium concentrations. There were also no differences between these controls and experimental groups in urinary excretion rates of calcium or oxalate. Whether pAA_5.1 in urine can antagonize other toxic effects of oxalate on tubule cells that could be responsible for inhibiting CaOx deposition was not examined.

Finally, on the basis of global observation of the animals, as well as the plasma creatinine values, there does not appear to be any overt short-term toxicity to pAA_5.1. On the basis of its chemical structure, it is not likely that it is metabolized to any significant extent in the body. Further studies will have to be performed to assess whether recovery in urine is complete or whether pAA_5.1 is sequestered at various sites in the body. Finally, oral absorption of pAA_5.1 will need to be tested. However, even if pAA_5.1 is not suitable as a therapeutic agent for the prevention of kidney stones, the study results reported here suggest that similar compounds hold such promise.

MATERIALS AND METHODS

Experimental animals

Male Sprague–Dawley rats (Charles River Laboratories Inc., Wilmington, MA, USA), weighing approximately 200 g when delivered, were used in the study. Rats were housed in the Veterinary Medical Unit (Veterans Affairs Medical Center, Milwaukee, WI, USA) and were given regular drinking (tap) water and regular rat chow, Lab Diet 5001 Rodent Diet (PMI Nutrition International, LLC, Brentwood, MO, USA). The rats were acclimatized for approximately 1 week before use. Protocols were approved by the Animal Care Committee.

Surgical techniques

Rats were placed in individual cages and made to fast overnight. Alzet 1 μl/h osmotic minipumps, Model 2001 (Durect Corporation, Cupertino, CA, USA) were filled as per manufacturer’s instructions with pAA_5.1 (Fluka/Sigma, St Louis, MO, USA) or 0.9% saline (Baxter Healthcare Corporation, Deerfield, IL, USA). Rats were anesthetized with 40 mg of pentobarbital (Abbott Laboratories, Abbott Park, IL, USA) per kg body weight by intraperitoneal injection. Employing sterile technique, the filled pumps were subcutaneously implanted along the upper left side of the backs of the rats. The rats were monitored during recovery from anesthesia and housed in regular cages without bedding. Paper towels were placed in the cages for nesting purposes. The next day, the rats were transferred to fresh cages with bedding.

Inhibition of crystal retention

The osmotic minipumps were filled with either varying concentrations (1–300 μg/μl) of pAA_5.1 or saline (experimental rats: n = 3 for 1 and 100 μg/μl pAA_5.1; n = 2 for 10 μg/μl; n = 4 for 20, 50, and 200 μg/μl; and n = 10 for 300 μg/μl; saline control rats: n = 21). As the minipumps deliver fluid at a rate of 1 μl/h, the indicated concentrations are equivalent to delivery of these amounts of this agent per hour. The minipumps are nominally of 1-week duration, but contain enough volume for at least 9 days. On the day following implantation of minipumps, the rats were placed on the CID, 0.8% ethylene glycol (Sigma, St Louis, MO, USA) with 1% NH₄Cl (Sigma) in tap water for 8 days. The rats were fed with regular rat chow. A quantitative urine collection was obtained over 24 h ending on the day of killing.

Tissue harvest and processing

At the end of the experiments, the rats were anesthetized and then killed by pneumothorax/exsanguinations. Blood was collected through cardiac needle stick and transferred to lithium heparin tubes (BD, Franklin Lakes, NJ, USA). The blood was then centrifuged for 15 min at 2800 r.p.m. and 4 °C to isolate plasma, which was then stored at −80 °C until being tested for creatinine and calcium. Kidneys were harvested and then rinsed in chilled phosphate-buffered saline, pH 7.4 (Sigma), that had been saturated with calcium and oxalate (PBS-CaOx) by overnight incubation with 0.1 mg COM crystals/ml and then filtered to remove the crystals. The kidneys were decapsulated, rinsed as before, sliced in 2–3 mm sections, rinsed again, then immersion-fixed for 18–24 h at room temperature in 10% formalin in PBS-CaOx. Kidney slices were then rinsed for three times in chilled PBS-CaOx, kept at 4 °C for 18–72 h in 20% sucrose (Sigma) in 0.1 M PBS-CaOx, placed in 20% gelatin (Fisher Scientific, Pittsburgh, PA, USA) overnight at room temperature, and after removal from the gelatin, the tissue was stored frozen at −80 °C. Cryosections (10 μm thickness) were cut from the tissue samples using an UltraPro 5000 Cryostat (Vibratome, St Louis, MO, USA). The pumps were removed from the animals and the remaining solution was removed and measured to ensure that the pumps had released their contents.

Analysis of blood and urine

Plasma calcium and creatinine testing was performed using the ACE Clinical Chemistry System autoanalyzer (Alfa Wassermann Inc., West Caldwell, NJ, USA) in the Biochemical Core Laboratory, Department of Physiology, Medical College of Wisconsin, Milwaukee, Wisconsin. Creatinine was measured using the Jaffe Reaction (Alkaline Picrate) without Lloyd’s Reagent and reported as mg/100 ml. Calcium was measured using Arsenazo III Dye and reported as mg/100 ml for plasma.

Urinary ions were measured using a dual pump ICS-3000 ion chromatography system equipped with two conductivity detectors, UV detection (ICS-VWD), ion suppressors (CSRS or ASRS Ultra-II), and an autosampler (Dionex, Sunnyvale, CA, USA). Cations were separated on an IonPac CS12A column with guard column using a 20 mN H₂SO₄ mobile phase (1 ml/min, 25 μl injection volumes) over a 20 min period. Creatinine was detected in tandem on this system with UV (A210) detection (Dionex Application Note 107). Oxalate was separated from other urinary anions using an IonPac AS4A column with guard column employing a 1.7 mM sodium bicarbonate, 2.0 mM sodium carbonate eluent system (2 ml/min, 25 μl injection volumes) over a 30 min period (R. Holmes, Wake Forest University, Winston-Salem, NC, USA, personal communication). Detection limits for calcium, creatinine, and oxalate using these methods were 62, 25, and 50 ng, respectively. IC grade reagents were purchased through Sigma and diluted in deionized water (⩾17.8 MΩ cm resistance). Anion and cation standards were purchased from Dionex. Creatinine standards were available from TECO diagnostics (Anaheim, CA, USA).

Crystal quantification

Crystals in the kidney cryosections were observed at a total magnification of 260-fold under partially polarized light using a Nikon Optiphot-2 microscope with a polarizing lens (Nikon, Melville, NY, USA), SIT 66 camera (DAGE-MTI of MC Inc., Michigan City, IN, USA) and Sony Trinitron monitor (Sony Corporation of America, New York, NY, USA). The planar area occupied by the crystals was quantified using Image 1 software (Universal Imaging Corporation, now part of Molecular Devices, Sunnyvale, CA, USA). For each animal, randomly selected fields (n = 6) of each kidney area (cortex, outer medulla, and inner medulla) were imaged. The values were recorded as pixels (object area).

Biotinylation of pAA_5.1

PAA_5.1 was biotinylated at a ratio calculated to provide only one biotin for every molecule of acrylic acid biotin LC-hydrazide according to the manufacturer’s instructions (Pierce, Rockford, IL, USA). The unincorporated label was removed using a PD10 desalting column (Bio-Rad, Hercules, CA, USA) equilibrated in PBS and collected in 1 ml fractions. Positive fractions were assessed using a dot blot with streptavidin-horseradish peroxidase detection (see below) and the recovered amount was quantified by reference to the absorption at 220 nm of known quantities of pAA_5.1.

Biotinylated pAA_5.1 (Bt-PAA_5.1) detection and recovery in rat urine

We administered bt-pAA_5.1 by osmotic minipump at 10 μg/h to four rats and at 50 μg/h in two rats on tap water without CID. Rat urine samples were treated with 10% trichloroacetic acid (30 min on ice, then centrifuged at 4 °C for 10 min at 15,000 r.p.m.) followed with a cold acetone wash to decrease nonspecific sample signal. The precipitates from the urine samples and parallel Bt-pAA_5.1 standard samples were reconstituted in PBS and loaded onto a dot blot manifold using nitrocellulose membranes. The membrane was stained for sample transfer and well integrity using Ponceau S (0.1% Ponceau S, 3% TCA), washed with 0.1% Tween in Tris-buffered saline, and blocked with 2% BSA–Tris-buffered saline for 1 h. Biotin was reacted with streptavidin-horseradish peroxidase (1:10,000 in 2% BSA-Tris-buffered saline for 1 h). The membrane was extensively washed (4 ×, Tris-buffered saline) and horseradish peroxidase was detected by chemiluminescence (Femto-ECL, Pierce). The dot blot image was acquired using an image station (IS2000 MMT, Kodak, New Haven, CT, USA), and Kodak molecular imaging software was used to quantify the signal in pixels.

In vitro COM aggregation assay

The biotinylated and unlabeled pAA_5.1 were tested for their ability to regulate COM aggregation by analysis of the particle size distribution changes of COM seed crystals added to a slightly supersaturated solution of CaOx containing the particular compound, as described previously.¹⁶ Specifically, 300 μg of COM seed crystals are added to 5 ml of a solution of 0.25 mM CaCl₂ and Na₂C₂O₄ in 150 mM NaCl, buffered to pH 7.5 with 10 mM HEPES (supersaturation1.11¹¹), and allowed to mix for 1 h at 37 °C. The slight supersaturation was employed to assure that no dissolution occurred to skew the results. Then, the particle size distribution of an aliquot of this mixture was determined using the AccuSizer 780/SIS (Particle Sizing Systems, Santa Barbara, CA, USA).

To determine the seed crystal particle size distribution, 15 μg of COM crystals were added to 15 ml of sizing buffer (0.225 mM CaCl₂, 0.225 mM Na₂C₂O₄, 10 mM HEPES; 0.2 μm filtered) and measured in duplicate. The sizing buffer was shown to cause neither significant growth nor dissolution of the COM crystals for up to 1 h. Aggregation data from these bulk crystallization methods were expressed as R_D, the ratio of the final weight average particle diameter (D_w) for the particle size distribution to the D_w of the COM seed crystals measured daily.

Statistical analyses

Values are expressed as mean ± s.e. Statistical significance, defined as P-value <0.05, was calculated using the Student’s t-test for unpaired data assuming unequal variances with Microsoft Office Excel 2003 (Microsoft Corporation, Redmond, WA, USA).