R.P.E. de Gier, MD, Paediatric Urologist, Paediatric Urology Centre, University Medical Centre St. Radboud, HP 426, PO Box 9101, NL 6500 HB Nijmegen, the Netherlands. e-mail: R.deGier@uro.umcn.nl
To study the pathophysiological changes of renal tubular drug transport mechanisms in congenital renal obstruction, by developing a model for perfusing the isolated kidney (IPK) after neonatal surgical induction of partial ureteric obstruction in Hanover Wistar rats.
MATERIAL AND METHODS
Moderately severe obstruction of the right kidney of male rats was created by burying a segment of the right ureter under the psoas fascia at 5–7 days after birth. Different fluorescent substrates for renal organic anion and cation drug transport systems were added to the IPK, and the concentration of these substances with time analysed in perfusate and urine.
The reproducibility in all groups of the glomerular filtration rate (GFR) and drug excretion was remarkably good. GFR was significantly lower in obstructed kidneys than in unobstructed kidneys. 123Rhodamine, a marker for organic cation and P-glycoprotein transport, had a significantly lower maximum excretion rate in the obstructed than in unobstructed kidneys. Renal fractional clearance (123rhodamine clearance corrected for diminished GFR) was also significantly lower in obstructed kidneys. There was no significant difference in maximum excretion (absolute and corrected GFR) for Lucifer Yellow, a marker for sodium-dependent organic anion transport. The maximum excretion rate of calcein, a marker for sodium-independent organic anion transport, was significantly lower in the obstructed than in the unobstructed kidneys, but significantly higher after correcting for reduced GFR.
The IPK is a good model for studying the effect of neonatal renal obstruction on tubular drug transport. These results show that organic anion and cation transport mechanisms are affected differently by obstruction.
Obstructive uropathy can lead to varying degrees of renal damage; experimental studies in fetal lambs have shown that complete ureteric obstruction results in different patterns of congenital damage, depending on the timing of the ureteric obstruction. Unilateral ligation of the ureter in the first half of gestation results in severe renal dysplasia, but the same obstruction of the ureter in the second half of gestation results in a hydronephrotic but otherwise well recognisable kidney . At a cellular level even brief periods of obstruction during nephrogenesis have been shown to induce apoptosis, aberration in cell proliferation and other histological signs of renal dysplasia .
Partial obstruction of the upper tract, which is of greater clinical significance than complete obstruction, is difficult to reproduce in experimental models such as the fetal lamb. To overcome this problem, incomplete obstruction has been induced in neonatal animals such as the rat or guinea-pig, species in which nephrogenesis continues to occur after birth. Thus, neonatal obstruction of the ureter in these species is a good model for second-trimester obstruction in humans. In these experimental models, renal blood flow and glomerular perfusion are reduced and histological changes of progressive glomerular sclerosis, tubular atrophy and fibrosis are apparent .
Congenital (partial) obstruction of the upper urinary tract, e.g. by PUJ obstruction or vesico-ureteric junction obstruction, is an important clinical entity in (paediatric) urology. Partial obstruction may or may not lead to impairment of kidney function, which is currently usually evaluated with differential function in DMSA or MAG-3 renography .
Many therapeutic drugs, e.g. penicillins, trimethoprim and digoxin, that are used in neonatal infants, are eliminated from the body by renal excretion. Good renal function is important to prevent accumulation, especially of drugs with a small therapeutic index, like digoxin. On the other hand, excretion can also be mandatory for the efficacy of the drug, especially for antibiotics for treating UTIs, when sufficiently high levels of the drug in the urine are necessary for effective treatment.
Different renal tubular transport mechanisms have been identified: the P-glycoprotein transport mechanism (e.g. excretion of cationic and non-ionic drugs like anthracyclines, digoxin, steroids), the organic cation transport mechanism (e.g. excretion of trimethoprim, creatinine, H2-antagonists), and the organic anion transport mechanism (e.g. excretion of penicillins, loop-diuretics and thiazides). At least in vitro, the latter mechanism can be subdivided in sodium-dependent and -independent organic anion transport.
Although it is well established that congenital renal obstruction can lead to a general impairment of renal function, very little is known about the direct effect of obstruction on the different transport mechanisms and whether these mechanisms are affected to the same extent. Such knowledge could provide important clinical information about the administration and dosage of different therapeutic drugs in children with obstructive uropathy.
To study the effect of congenital renal obstruction on the different tubular transport mechanisms, we developed an experimental model, in which we combined neonatal obstruction in rats with isolated perfused kidney (IPK) experiments.
MATERIALS AND METHODS
Drugs used were sodium pentobarbital (Apharmo, Arnhem, the Netherlands), Pluronic F108 (BASF, Arnhem, the Netherlands), and heparin, aldosterone and inulin (Organon, Oss, the Netherlands). Lucifer yellow (LY-CH dilithium salt), tetramethylrhodamine isothiocyanate-dextran (TRITC-dextran, molecular weight 4400), and 123rhodamine were obtained from Sigma (St. Louis, MO, USA). Lysine-vasopressin was obtained from Sandoz Pharma Ltd. (Basel, Switzerland), angiotensin II from Beckman (Palo Alto, USA) and Synthamin 14 from Travenol (Thetford, Norfolk, UK). Calcein-acetoxymethylester (calcein-AM), and calcein were from Molecular Probes (Eugene, OR, USA). All other chemicals were of analytical grade and purchased from Merck (Darmstadt, Germany) or Sigma.
The experimental procedures for creating ureteric obstruction in newborn rats and of isolated perfusion of the rat kidneys were reviewed and approved by the local Ethics Committee on Animal Experimentation.
Experiments were conducted on specific pathogen-free male Hanover Wistar rats, maintained in the central animal laboratory of our institution under standard laboratory husbandry conditions in closed systems, i.e. newborn rats together with their mother and siblings in one cage, and older rats with two to three rats in another. The temperature was regulated at 22–24 °C and humidity at 50–60%, with artificial lighting from 08.00 to 20.00 hours and dark from 20.00 to 08.00 hours. Water and rat food were freely available and the air changed 15 times/h.
Pregnant female rats (Harlan Nederland, Horst, the Netherlands) were obtained ≥ 7 days before the expected date of delivery. Moderately severe obstruction of the right ureter was created as described by Josephson et al.. Male new-born rats were operated 5–7 days after birth under general halothane inhalation anaesthesia, by an experienced biotechnician using a dissection microscope. After fixing the anaesthetized animal supine on a heated surface, to prevent hypothermia, the abdomen was opened in the midline and the right ureter identified where it runs over the psoas fascia. The psoas fascia was opened longitudinally, parallel to the ureter, and a segment of ≈ 0.5 cm of the ureter then buried under the psoas fascia, using two or three nonabsorbable sutures. After closing the abdomen the animal was returned to the mother and siblings. The mean operative duration time was < 10 min.
At 10–12 weeks old the isolated right kidney was perfused as described in detail by Cox et al.. Briefly, after anaesthesia with intraperitoneal pentobarbital (45 mg/100 g) and opening the peritoneal cavity, the right kidney was exposed and freed from adhering tissues. All obstructed right kidneys showed marked dilatation of the collecting system (hydronephrosis). After cannulating the ureter (proximal to the segment buried in the psoas fascia) with a PE-50 polyethylene tube, the renal artery was cannulated via the mesenteric artery. The kidney was then excised and placed in a fluid bath at a constant temperature of 37.5 °C. Renal perfusion was maintained by a roller pump and renal pressure measured with a pressure transducer. The remainder of the perfusate which was flowing from the renal vein into the fluid bath was collected and returned to the arterial perfusate (closed circuit system).
The perfusion fluid had the following composition (mmol/L, except if indicated otherwise): Na 114; KCl 5.2; CaCl2 1.8; MgCl2 1.0; NaHCO3 22.5; Na2HPO4 0.84; KH2PO4 0.28; glucose 5.0; urea 4.0; pluronic F108 25.0 g/L; cyanocobalamin 0.014; glutathione 0.33; inositol 0.083; cysteine 0.50; glycine 2.30; Na-pyruvate 2.00; Na-acetate 1.22; Na-propionate 0.208; inosine 1.00; alanine 5.00; glutamine 0.106; l-glutamine acid 2.00; ascorbic acid 0.01; Na-lactate 1.0; choline chloride 1.0 mg/L; insulin 4 IU/L; aldosterone 2.0 µg/L; lysine-vasopressin 0.01 IU/L; and angiotensin II 15.0 ng/L. To this solution 1.0% Synthamin 14 (a mixture of 15 amino acids) was added.
The perfusion fluids were filtered before and during the experiment (0.22 and 5.0 µm pore radius, respectively). The solution was oxygenated with a flow-through oxygenator using 95% O2 and 5% CO2. The urine was led through a bubble flowmeter to determine the urine flow rate.
After starting perfusion, the isolated kidney was allowed to stabilize for 30 min, during which the intraperitoneally injected frusemide was washed out completely and no urine or perfusate samples taken. The subsequent experimental time was 150 min, during which urine was collected over 10-min periods. The perfusion pressure was set to 80 mmHg, which resulted in a flow of ≈ 12 mL/min. To measure GFR, cyanocobalamin or TRITC-dextran was added to the perfusate . The cyanocobalamin concentration in the perfusate was measured spectrophotometrically, and TRITC-dextran by fluorometry. The experiment was started by adding fluorescent substrates for renal organic anion and cation drug transport systems, and the concentration of these substances with time analysed in perfusate and urine. Studies were conducted with 123rhodamine (a marker for organic cation and P-glycoprotein transport)  Lucifer Yellow (a marker substance for sodium-dependent organic anion transport) , nonfluorescent calcein-AM (which is hydrolysed in proximal tubule cells to fluorescent calcein, a marker for sodium-independent organic anion transport)  and controls with no marker added. Statistical differences between means were determined by Student's t-test and the level of significance set at P < 0.05.
Urine and perfusate samples were analysed for glucose, using the Gluco-Quant kit (Boehringer), and alkaline phosphatase, to assess proximal tubule function . The concentration of calcein, 123rhodamine, Lucifer Yellow or TRITC-dextran in perfusate and urine samples were determined using fluorescence spectrophotometry. For this, an aliquot of 50 µL of the perfusate sample was adjusted to 600 µL with analysis buffer (Sörensen buffer, pH 7.36). Urine samples were diluted 10 times with buffer, from which an aliquot of 25 µL was adjusted to 600 µL with 50 µL blank perfusion fluid and 525 µL buffer. Fluorescence in these prepared samples was measured using a luminescence spectrophotometer (Perkin Elmer, LS50, Beaconsfield, Bucks, UK). To measure 123rhodamine the excitation and emission wavelengths were 501 and 531 nm, for calcein 488 and 518 nm, for Lucifer Yellow 425 and 525 nm, and for TRITC-dextran 550 and 573 nm, respectively. In all cases, a bandwidth of 5 nm was used. Concentrations were calculated by comparing fluorescence intensity (in photomultiplier units) with a calibration curve of samples of blank perfusion fluid spiked with different concentrations of 123rhodamine, calcein, Lucifer Yellow or TRITC-dextran.
All animals in which hydronephrosis was induced, except one, survived this procedure. In eight animals the IPK was not feasible, either because the right kidney was destroyed by massive hydronephrosis (six) or because it was technically impossible to cannulate the renal artery (two).
The viability of the perfused rat kidneys was assessed by following the fractional excretion of glucose, cumulative excretion of alkaline phosphatase, fractional reabsorption of water, urine flow and pH, GFR and renal perfusate pressure. Figure 1 shows that the measured GFR was 49% lower in obstructed kidneys than in unobstructed kidneys. All other functional variables are shown in Table 1. Kidneys from untreated rats showed good function over the 2 h course of the clearance experiments (Table 1), but obstructed kidneys had a markedly higher diuresis and alkaline phosphatase excretion, and decreased fractional reabsorption of water and glucose than untreated kidneys.
Table 1. Renal functional variables (over 30–120 min) of the isolated perfused rat kidney from four normal (control) and four obstructed rat kidneys during control perfusions
For 123rhodamine the maximum excretion rate in the obstructed was significantly lower than in the unobstructed kidney (Fig. 1b). Even after correcting for the lower GFR in obstructed kidneys, clearance/GFR was still significantly lower in obstructed than in unobstructed kidneys (Fig. 1c), indicating that the organic cation/P-glycoprotein transport mechanism is more severely reduced by obstruction than is GFR.
For Lucifer Yellow the outcome differed; there was no significant difference in maximum excretion rate, with a mean (sd) of 4400 (1400) for obstructed kidneys and 5400 (700) pmol/min for unobstructed kidneys (Fig. 1d). After correcting for the reduced GFR in the obstructed kidneys, clearance/GFR in the obstructed kidneys was even slightly higher (although not statistically significant) than in the unobstructed kidneys, at 1.8 (0.3) and 1.5 (0.2), respectively (Fig. 1e).
Results for calcein-AM again differed from both the results for 123rhodamine and Lucifer Yellow. The maximum excretion rate of calcein (which is formed inside the tubular cells) was significantly lower in the obstructed, at 19 (4) pmol/min, than in unobstructed kidneys, at 24 (7) pmol/min (Fig. 1f) but the maximum excretion rate corrected for GFR in obstructed was significantly higher than in unobstructed kidneys, at 64 (15) vs 41 (12) ng/mL (Fig. 1g).
Because the concentration of calcein in tubular cells is independent of calcein concentrations in the perfusion medium, but depends on calcein-AM concentration in perfusion medium, renal calcein clearance could not be calculated. Therefore, we corrected renal calcein excretion rate for GFR.
Obstruction beyond the kidney is an important cause of impairment of renal function. Obstruction can be either unilateral with a normal contralateral kidney, unilateral in a single functional kidney (contralateral kidney absent or not functioning) or bilateral. Impairment of renal function may vary, depending on the degree and duration of obstruction. Furthermore, the impairment of the function of the obstructed kidney and of total renal function (combined function of both kidneys in case of bilateral functioning kidneys) can be distinguished.
Obstruction can be caused by many different disorders. Acute obstruction, which is the minority of paediatric cases, may be caused by obstructive urolithiasis or external compression on the urinary tract, e.g. obstruction by tumour. Chronic obstruction, as seen in most paediatric cases and most often congenital, can be caused by PUJ obstruction, vesico-ureteric junction obstruction, (ectopic) ureterocele, ectopic ureter, neuropathic bladder or infravesical obstruction (PUV). Most of these conditions may occur unilaterally or bilaterally, while infravesical obstruction usually affects both kidneys, although not necessarily both kidneys equally.
Clinically the impairment of renal function is usually evaluated by serum creatinine level or creatinine clearance (total renal function) or by nuclear renography with either 99mTc-MAG3, -DMSA or -DTPA nuclear renal scintigraphy. Nuclear renal scans are especially suitable for evaluating differential renal function and less suitable for evaluating total renal function.
Other factors than GFR that influence clearance of different substances or radionuclides are differences in serum protein (predominantly albumin) binding, differences in active tubular reabsorption and passive backward diffusion .
Creatinine clearance is an inaccurate estimate of true GFR because of the tubular secretion of creatinine by the organic cation transport mechanism . Particularly when renal function is compromised, creatinine clearance will significantly overestimate GFR.
99mTc-DTPA is only filtered by the glomerulus and not actively excreted or reabsorbed by any of the tubular transport mechanisms, and therefore reflects true GFR, like inulin clearance [12,14,15]. However, theoretically there are still some limitations because Tc-DTPA shows varying degrees of serum protein binding , resulting in 99mTc-DTPA-based GFR estimates being slightly lower than inulin-based GFR estimates. However, in different commercially available 99mTc-DTPA kits, protein binding is relatively low, at only 1–2% .
99mTc-MAG3 is filtered and strongly actively excreted in the renal tubular system by the organic anion transport mechanism, resulting in very efficient excretion by the kidney. Glomerular filtration is responsible for only 11% of the total renal clearance of 99mTc-MAG3 in the rat kidney (glomerular filtration being limited by the high serum protein binding of 99mTc-MAG3), the remaining 89% being realised by active (organic anion) tubular transport . Most probably, 99mTc-MAG3 is unaffected by tubular reabsorption .
99mTc-DMSA is largely accumulated by proximal tubular cells, where it is concentrated in the cytoplasm . It is unclear which tubular transport mechanism is responsible for the uptake of 99mTc-DMSA in the tubular cells, although it is apparent that this is not the same mechanism that is responsible for the transport of 99mTc-MAG3 [18–20]. The renal uptake of DMSA is strongly influenced by dehydration, urinary pH and other factors . The GFR of 99mTc-DMSA is very low, and the main extraction route of 99mTc-DMSA from the blood being by tubular extraction. Evidence for the tubular reabsorption of 99mTc-DMSA is controversial .
Thus, as the substances used in the methods to evaluate impairment of renal function are handled differently by the kidney, it is questionable whether all methods will always indicate the same degree of impairment of renal function. The present data show that if the impairment of renal function is caused by partial obstruction, different renal tubular transport mechanisms are affected differently.
Creatinine is partly transported by the organic cation transport mechanism which is, according to the present results, the tubular transport mechanism that is affected the most by partial obstruction. Thus, the serum creatinine level will to some extent overestimate renal damage compared with the reduction of GFR.
99mTc-DTPA is only filtered by the glomerulus; therefore the result of DTPA-based nuclear renal scintigraphy will not be influenced by the differences in renal tubular transport mechanisms in obstructed kidneys, and thus will reflect accurately the reduction in GFR. However, 99mTc-MAG3 is mainly excreted actively in the renal tubules by the organic anion transport mechanism, and thus will be influenced by the differences in damage to the different renal tubular transport mechanisms.
Creatinine serum levels or creatinine clearance will tend to overestimate impairment of renal function (as creatinine is transported by the transport mechanism that is affected most clearly), while MAG3, which is transported by the less affected organic anion transport mechanisms, will show a somewhat lesser degree of impairment of renal function.
Clearly, more information is needed on the molecular identity of the transporters by which the various diagnostic agents are excreted in the kidney, and how these transporters are regulated in response to partial ureteric obstruction. We are currently examining the expression of organic anion and cation transporters in obstructed kidneys.
In conclusion, the IPK is a good model for studying the effect of neonatal renal obstruction on tubular drug transport mechanisms. GFR is diminished by almost half in the obstructed kidney. The present data clearly show that organic anion and cation transport mechanisms are affected differently by obstruction, with the most pronounced influence on the organic cation/P-glycoprotein transport mechanism. The sodium-dependent anion transport mechanism is affected significantly less, while in the sodium-independent transport mechanism there are indications for a partial compensatory mechanism, caused by the obstruction. Although the total excretion ratio in obstructed was significantly lower than in unobstructed kidneys, after correcting for the diminished GFR in obstructed kidneys, excretion was relatively higher in obstructed than in unobstructed kidneys.
This study was presented at the 13th annual meeting of the ESPU, April 11–13, 2002, Budapest, Hungary .