Strain-related Differences in Urine Composition of Male Rats of Potential Relevance to Urolithiasis

Abstract

In carcinogenicity studies with PPAR γ and α/γ agonists, urinary bladder tumors have been reported in Harlan Sprague-Dawley (HSD) and Charles River Sprague-Dawley (SD) but not Wistar (WI) rats, with urolithiasis purported to be the inciting event. In two 3-month studies, the authors investigated strain-related differences in urine composition by sampling urine multiple times daily. Urine pH, electrolytes, creatinine, protein, citrate and oxalate levels, and serum citrate were assessed; urine sediment was analyzed by scanning electron microscopy and energy dispersive x-ray spectroscopy. HSD rats had significantly higher urine calcium than SD or WI rats, primarily as calcium phosphate-containing precipitate. When compared to SD rats, HSD rats had lower urine volume, higher urine protein, and a comparable (week 4) to lower (week 13) burden of MgNH₄PO₄ aggregates. Relative to WI rats, HSD rats had higher urine protein and magnesium and lower serum and urine citrate. Overall, the susceptibility to urolithiasis in male rats was HSD > SD > WI; this was likely due to strain-related differences in the amount of urine protein (a nidus for crystal formation), lithogenic ions, citrate (an inhibitor of lithogenesis), and/or volume. Strain-related differences in urine composition need to be considered when interpreting the outcome of studies with compounds that alter urine composition.

Introduction

Strain-related differences in physiology have been demonstrated in numerous species, including rats, and can have a significant impact on the results of both short- and long-term studies. Examples include an increased susceptibility to the nephrotoxicity of hydroquinone in Fischer 344 rats when compared to Sprague-Dawley rats (Boatman et al. 1996), a susceptibility of Dark Agouti rats to adjuvant-induced arthritis while Albino Oxford rats are resistant (Miletic et al. 2007), and a predisposition to addictive or compulsive behavior in Lewis rats (Brimberg et al. 2007). Knowledge of such differences can be critical when choosing a strain of laboratory animal for a particular study.

During standard preclinical testing of the antidiabetic PPAR α/γ (dual) agonist muraglitazar, a dose-related increased incidence of urinary bladder carcinomas in male Harlan Sprague-Dawley (HSD) rats was noted in a two-year carcinogenicity study (Tannehill-Gregg et al. 2007). Subsequent investigative work provided positive support for a male rat-specific nongenotoxic mechanism of urinary bladder carcinogenesis involving chronic mucosal injury and proliferation secondary to muraglitazar-related increases in urinary solids. The increases in urinary solids were accompanied by significant drug-related alterations of the urine milieu including reductions in citrate, increases in oxalate, and an increase in calcium- and magnesium-containing solids in alkaline urine (Dominick et al. 2006). Carcinogenicity bioassays with other PPAR dual agonists have demonstrated increased incidences of urinary bladder tumors in male and female HSD and Charles River Sprague-Dawley (SD), but not Wistar (WI), rats (El Hage 2005). Moreover, Sprague-Dawley rats are more susceptible than WI rats to sodium saccharin-induced urinary bladder tumors (which occur through a similar mechanism as that proposed for muraglitazar; Cohen 1999).

The urinary milieu of rats contains several factors that can be prolithogenic, including high concentrations of protein and various salts, as well as having an extremely high osmolality. Other urine components important to the formation of urine solids include citrate (a key antilithogenic divalent cation chelator) and urine pH. Calcium phosphate-containing solids and magnesium ammonium phosphate (struvite) stones form in alkaline urine (pH ≥ 6.5), while acidic urine favors their dissolution. Urine composition is markedly affected by diurnal variation, so for meaningful evaluation it is critical to collect samples at several time points throughout the day, making sure to at least cover the dark phase (when animals are awake and urine composition is affected by eating and drinking) and the light phase (when animals are asleep; Cohen 1995; Cohen et al. 2007).

In support of our chronic and investigative muraglitazar studies, and in light of findings in chronic studies using other PPAR dual agonists, we were interested in determining strain-related differences in urine composition in male rats. In this study, we investigated differences in urine composition between HSD, SD, and WI rats that may play a role in strain-related susceptibility to urolithiasis.

Materials and Methods

With the exception of urine sediment analysis, studies were compliant with U.S. Food and Drug Administration Good Laboratory Practice Regulations. Furthermore, animal use was in accordance with the Guide for the Care and Use of Laboratory Animals.

Animals

Random-bred, barrier-raised male Hsd:Sprague Dawley^® SD^® [HSD], Crl:CD^®(SD) [SD], and Crl:WI [WI] rats were obtained from Harlan Sprague-Dawley, Inc., Indianapolis, IN (HSD) and Charles River Laboratories, Wilmington, MA (SD, WI). Rats were six to eleven weeks old at study initiation. All animals were housed individually in stainless-steel, wire-bottom cages in environmentally controlled rooms with a 12 hr light/12 hr dark cycle, a humidity range of 30% to 70%, and a temperature range of 18°C to 26°C. Tap water and certified rodent diet (no. 8728C, Harlan Teklad) were available ad libitum.

Experimental Design

In the first study, lasting three months, one group of fifteen HSD rats and one group of fifteen SD rats each received 0.9% Sodium Chloride for injection (NaCl), USP by daily oral gavage at a volume of 5 ml/kg. With respect to freshly voided urine collection, predose samples were taken in the early light phase (approximately 1 hour after the lights were turned on).

In the second three-month study, two groups of fifteen HSD rats each and two groups of fifteen WI rats each were given (1) no test material but were handled daily or (2) no test material with no daily handling. Since no actual dosing occurred in this study, with respect to freshly voided urine collection (see below), “predose and 2.5, 8, and 16 hr postdose” refer to early, mid-, and late light phase and dark phase, respectively.

All experimental samples for both studies were collected from nonfasted rats with drinking water available ad libitum. Individual body weights were collected pretest and at least once weekly thereafter and were used to determine dose volumes (first study), monitor health status, and determine urine volume/body weight ratios. Because neither study included both SD and WI rats, there were no direct comparisons made between these two strains. Although both studies were titled as being three months in duration, due to the complex study design (involving multiple groups of animals, time points, and parameters evaluated), samples were collected up to fifteen weeks.

Urinalyses

Freshly voided urine was collected in a staggered manner across groups to ensure representative sampling from each group. The target for each interval was six to twelve acceptable samples per group. Samples were re-collected, when necessary, in an attempt to obtain samples from the same animals or group at each time point.

Freshly voided urine was collected during weeks 4 to 5 and 13 to 14 at predose and at 2.5, 8, and 16 hr postdose, and urine pH was immediately measured using a MI-411 combination pH microelectrode (Microelectrodes, Inc., Bedford, NH) and SB20 Symphony pH meter (VWR Scientific Products, West Chester, PA). Approximately 75 μl of each urine sample from all time points was removed and centrifuged at 7,000 rpm for 10 min. Most of the supernatant was removed, and the urinary sediment was placed on a 0.22 μm filter and vacuum dried. Filters were transported under ambient conditions to the laboratory of Dr. Samuel Cohen (University of Nebraska Medical Center, Omaha, NE) where they were coated with gold. The total area of each filter was evaluated by scanning electron microscopy to determine the incidence based on morphology of urinary solids present on the filter. Elemental analysis of the solids identified was performed by energy dispersive x-ray spectroscopy to confirm the presence of magnesium (struvite) crystals and/or a calcium phosphate-containing precipitate. Remaining urine from the 2.5- and 16-hr time points was used for determination of the following parameters (in the order presented, sample volume permitting): creatinine, protein, phosphorus (inorganic phosphate), calcium (soluble calcium), acidified calcium (total calcium), magnesium (soluble magnesium), acidified magnesium (total magnesium), sodium, potassium, and chloride using standard methods on the Hitachi 917 (Roche Diagnostic Systems, Inc.).

Urinalyses were also performed after weeks 4 and 13 (first study) or weeks 6 and 15 (second study) on chilled urine collected over an 18-hr period from nonfasted animals that were acclimated to metabolism cages for approximately 48 hr prior to collection. Sodium azide (100 μl of a 1% solution) was added to each collection vial. Measured parameters included urine volume (normalized to body weight), citrate (Spectronic 20D+ spectrophotometer, R-Biopharm), creatinine (Hitachi 917, Roche Diagnostic Systems, Inc.), and oxalate (Trinity Biotech Assay; Hitachi 917, Roche Diagnostic Systems).

With the exception of pH, urine parameters were evaluated as a ratio to the urine creatinine concentration to normalize for volume differences. For urine sediment analysis by scanning electron microscopy, an increased incidence or severity was defined as an increase of ≥ 3 rats with a specific finding or with a higher grade of severity over that recorded in the companion group (HSD vs. SD [first study] or HSD vs. WI [second study]).

For all figures, the data are reported as the M ± SD; n = 8–15; and *p < .05, **p < .01 for HSD as compared to SD (first study), or HSD compared to WI (second study) at each interval for each time point and treatment group.

Serum Citrate

Serum citrate analysis (Spectronic 20D+ spectrophotometer, R-Biopharm, South Marshall, MI) was performed after collection of blood (0.5 ml) from the tail vein approximately 0.5 hr postdose on all animals after four and thirteen weeks.

Statistics

Computations were performed using SAS™ (SAS Software, Release 9.1, SAS Institute, Inc., Cary, NC, 2002) PROC TTEST at the .05 level. For each response at each week/day and hour combination, PROC TTEST tested for equal variance and provided t-test results for both the equal and the unequal variance (using Satterthwaite’s approximation) situations. The p values given were associated with the t–test for unequal variance if the test of equality of variance was significant at the .05 level; otherwise, the p value reported was associated with the t-test for equal variance.

Results

Tabular summaries of the percentage change and direction of change between group means of relevant parameters of urine composition are available in Table 1 (HSD vs. SD rats), Table 2 (HSD vs. WI rats handled daily), and Table 3 (HSD vs. WI rats with no daily handling).

Urine Electrolyte, Protein Determinations, and pH

In the first study, the excretion of total calcium into the urine was much greater in HSD rats when compared to SD rats at both intervals and all time points. In the HSD rats, the calcium tended to be in the solid (insoluble) form, which reflected the increase in calcium phosphate-containing solids noted at both intervals in the urine sediment analysis (see below; Figure 1). In the second study, while urine sediment evaluation demonstrated an increase in calcium phosphate-containing solids in HSD rats at both week 5 and week 14 when compared to WI rats (see below), the urine chemistries demonstrated increased urinary excretion of total calcium in HSD rats at the week 5 interval only (Figure 1). In that study, the urine calcium in the HSD was generally in the solid (insoluble) form, which was consistent with the sediment evaluation.

The excretion of total magnesium into the urine was greater in HSD rats when compared to SD rats (week 13) and WI rats (week 5). Regardless of interval or study, the urine magnesium tended to be in the solid (nonsoluble) form in HSD rats during the light phase (2.5 hr postdose), while the soluble magnesium to total magnesium ratios were generally similar between groups during the dark phase (16 hr postdose; Figure 2). While the increase in magnesium-containing (struvite) solids noted in the urine sediment analysis in HSD rats when compared to WI rats (second study; see below) is supported by these urine chemistry findings, the urine chemistry magnesium analysis in the first study was not fully consistent with the urine sediment analysis.

While urine phosphorus levels were decreased during the dark phase (16 hr postdose) in HSD rats when compared to SD and WI rats, comparisons during the light phase (2.5 hr postdose) were more variable with urine phosphorus increased in HSD rats when compared to SD rats but similar when compared to WI rats (Figure 3). The lower urine phosphorus levels in HSD rats during the dark phase likely reflected increased incorporation of phosphorus into magnesium- or calcium-containing solids as determined by urine sediment analysis (see below).

Urine total protein was greatly increased in the HSD rats when compared to SD rats at both intervals and was often greater than that in WI rats (although rarely statistically significant; Figure 4).

While HSD rats tended to have slightly lower urine pH values when compared to SD or WI rats, the interstrain differences were somewhat variable and not large. The urine pH stayed at or above 6.5 at all time points and intervals in all strains and followed the normal diurnal pattern of more alkaline urine associated with the dark phase and more acidic urine associated with the light phase (Figure 5).

Urine volume was mildly decreased in HSD rats when compared to SD or WI rats at the earliest interval (week 5 or 6) and was comparable between groups at the later interval (week 14 or 15; data not shown).

Urinary sodium, potassium, and chloride levels showed no consistent pattern of differences when comparing strains (data not shown).

Urinary Citrate and Oxalate Determinations

Urine citrate levels were generally lower in HSD rats when compared to both SD and WI rats. The greatest interstrain difference was between the HSD and the WI rats (Figure 6).

Urine oxalate levels were generally comparable between strains, although there was a significant increase in urine oxalate in HSD rats by week 14 and when compared to SD rats at the same time point (data not shown).

Urine Sediment Analysis

The urine sediment of male HSD rats exhibited an increased incidence and/or burden of calcium-containing microcrystalline precipitate at both intervals and most time points when compared to the urine of male SD or WI rats (Tables 4 and 5). In a few HSD rats, the precipitate burden was heavy enough to result in piles or mounds on the filter (noted in 3 instances), and one animal had a low burden of calcium phosphate-containing calculi (Figure 7).

When compared to the urine sediment of male WI rats, that of male HSD rats intermittently demonstrated an increased incidence and/or burden of magnesium-containing solids (struvite; most notable at the week 14 interval; Table 5). However, when comparing the urine sediment of male HSD and SD rats, the SD rats exhibited an increased burden and/or incidence of magnesium-containing solids (struvite) at the week 13 interval, notably, in the form of aggregates ranging from < 100 to > 300 μm in size (Figure 8), while the struvite crystals normally found in rat urine were generally similar between the two groups (Table 4).

Serum Citrate

While serum citrate levels were similar between HSD and SD rats at both intervals, HSD rats tended to have lower serum citrate levels than WI rats (Figure 6).

The Effect of Daily Handling

In the second study, daily handling had minimal effects on differences in urine composition when comparing urinalysis and urine sediment analysis between HSD to WI rats and when making intrastrain comparisons. Daily handling was associated with an equivocal decrease in the burden of calcium phosphate-containing precipitate in both strains at week 5 (Table 5), and daily handling was associated with a modest decrease in protein excretion in HSD rats at week 14 (making total protein excretion more equivalent when comparing HSD to WI, although still minimally increased in the HSD; Figure 4).

Discussion

Male HSD rats developed urinary bladder tumors after treatment with the PPAR α/γ (dual) agonist muraglitazar, and male and female HSD and SD, but not WI rats, develop bladder tumors after treatment with other PPAR dual agonists (El Hage 2005; Tannehill-Gregg et al. 2007). Investigative studies with muraglitazar demonstrated a mechanism involving treatment-related alterations in the urine milieu leading to an increase in urine solids, which resulted in mucosal irritation and subsequent regenerative proliferation (Dominick et al. 2006). In support of this investigative work, the current studies were developed to investigate interstrain urine compositional differences in male rats to support that hypothesis that there is a strain-related difference in susceptibility to the formation of urine solids.

Concentrations of various lithogenic ions in the urine can play an important role in the formation of urinary solids, with supersaturation contributing to nucleation (Coe 1992). The major divalent cations in urine are calcium and magnesium, and anions that play an important role in urolithiasis include phosphate and oxalate. Calcium, magnesium, and phosphate are important components of urinary precipitates such as calcium phosphate and magnesium ammonium phosphate (struvite). In general, data from the present studies demonstrated that HSD rat urine contained more total calcium and total magnesium and less phosphorus than that of SD and WI rats, although urine chemistry analyses did not always directly correlate to the urine sediment analysis, and results sometimes varied when comparing different intervals and time points. The most consistent and notable differences in urine composition was for calcium, which was greatly increased in the HSD when compared to the SD and somewhat increased when compared to the WI. The incongruity of urine chemistry data and urine sediment analysis noted for some ions at some time points and/or intervals demonstrates the need to look at the overall pattern based on the collective data.

Although calcium and magnesium concentrations in urine play an important role in the formation of solids, urine of normal rodents often contains supersaturated levels of these major ions, and it is other factors in the milieu that help to control the formation of microcrystalline precipitates. An important factor that prevents nucleation and growth of calcium-containing crystals is the calcium-chelator citrate that is reported to be responsible for up to 50% of the normal inhibitory activity against the precipitation of calcium phosphate in urine (Bisaz et al. 1978). Accordingly, urine levels of citrate are regularly measured in human stone formers, and oral citrate therapy is used in the prevention of calcium-containing stone formation (Renata et al. 2003). The decreased concentration of urine citrate in HSD rats when compared to SD rats, and especially WI rats, likely contributed to a prolithogenic environment, notably, in the presence of higher levels of calcium as seen in the HSD. The demonstration of similarly lower levels of serum citrate in WI rats compared to HSD rats provides additional support that the strain differences in urine concentrations of this cation chelator were real, since excreted citrate is principally derived from the urinary filtrate (Renata et al. 2003).

Urine proteins also play an important role in the formation of urinary precipitates. Rat urine normally has a very high concentration of protein and male rats have approximately ten times the urinary protein as female rats (Lehman-McKeeman 1991). This is predominantly low molecular weight protein (α_2u-globulin) in the young male rat (Hard 1995). It has been suggested that this high concentration of urinary protein is prolithogenic as it predisposes the male rat to the formation of urinary precipitates by acting as a nidus (Cohen 1999). Strain-related differences in urine protein levels are a well-known phenomenon (Hard 1995), and the HSD rat often demonstrates higher levels of urine protein at a younger age due to an earlier onset of chronic progressive nephropathy resulting in albuminuria (Pettersen et al. 1996; Palm 1998). The increased urine protein levels noted in the urine of HSD rats when compared to WI, and especially SD rats, contributes to the more prolithogenic urine in this strain.

The mildly decreased urine volume noted in HSD rats when compared to SD and WI rats at the earliest interval likely also contributed to the prolithogenic environment by contributing to the saturation of lithogenic salts and protein (Cohen 2005).

A urine pH of 6.5 is considered to be the critical limit above which formation of calcium- and magnesium-containing solids is promoted (Cohen 1995). Although urine pH in the normal rodent can range from 5.5 to 7.5 (Cohen 1995), in our studies, the urine pH stayed at or above 6.5 at all time points for all three strains, indicating that differences in urine pH likely do not play a role in strain-related susceptibility to the formation of urinary solids. The normal diurnal variability of urine pH, with the lowest pH seen during the late-light phase (approximately 8 hr after the lights were turned on) and the highest pH noted in and around the dark phase, was noted in all strains. This variability reflects the influence of food consumption on urine pH (rats eat during the dark phase, which alkalinizes their urine) and underscores the importance of sampling urine at various time points throughout the day (Fisher et al. 1989).

Struvite (magnesium ammonium phosphate) crystals and calcium-containing solids can be normal components in the urine of rodents (Schumann 1991). Although the former are typically considered to be nontoxic to the urothelium (Cohen 1995), under certain conditions, struvite crystals may increase in number or may form large aggregates leading to a potential increase in sediment-induced urothelial trauma. The higher incidence and severity of struvite solids (notably, aggregates, some greater than 300 μm) in the urine of SD rats when compared to that of HSD rats suggests that SD rats are moderately susceptible to urolithiasis and to subsequent urothelial damage after treatment with compounds that alter urine composition to promote the formation of solids. Calcium phosphate-containing precipitate is known to be cytotoxic to epithelial cells and, specifically, to urothelial cells both in vitro (Cohen 1999) and in vivo (Cohen et al. 1996). The increased background incidence and burden of calcium phosphate-containing precipitate in the urine of HSD, when compared to SD and WI, rats indicates this strain may be more susceptible to urolithiasis in general and thus to urothelial injury (and subsequent proliferative responses) in response to treatment-related prolithogenic alterations in the urine milieu.

Extensive handling of F-344 rats has been reported to result in an increase in urinary calcium phosphate-containing precipitate, without a corresponding change in urine pH (Cohen et al. 1996). In our study, daily handling of rats had no significant effects on urine pH or calcium phosphate-containing precipitate levels when making both inter- and intrastrain comparisons, although at the earliest interval, there was an equivocal decrease in the burden of calcium phosphate-containing precipitate with handling in both strains. After fourteen weeks, there did appear to be a modest decrease in protein excretion in HSD rats that were handled daily when compared to HSD rats not handled daily. Although a definitive mechanism for this is uncertain, handling-related effects on food consumption may have played a role. Changes in urine composition related to handling may reflect stress-related effects (Cohen et al. 1996); disparate results when comparing the Cohen study to our study likely reflect differences in the strain of rat(s) evaluated and in the experimental design (such as the amount or manner of handling, diet or preparation of urine sample).

In conclusion, in assessment of strain-related differences in urine composition, it is critical to evaluate and integrate multiple endpoints (encompassing parameters evaluated as well as timing of collections) to obtain an accurate overall picture of urine composition in a given strain. Standard urinalysis should be supplemented with urine chemistries, sediment analysis, and evaluation of other parameters important to urolithiasis such as urinary and serum citrate. Daily collection endpoints should include, at a minimum, collection during the light and dark phases. Collection of multiple intervals is desirable to control for age-related changes in urine composition.

Strain-related differences in urine composition likely contribute to differences in susceptibility to urolithiasis. When compared to the HSD, the WI rat is much less predisposed to urolithiasis due to higher levels of urine citrate and lower levels of urine protein and lithogenic ions (notably, calcium). When compared to the HSD, the SD rat is slightly less predisposed to urolithiasis (notably, calcium-containing solids) due to lower levels of urine protein and lithogenic ions (notably, calcium) and higher urine volume. These differences need to be considered when interpreting the outcome of studies using compounds that alter urine composition.