Activity profiles of prostaglandin 15- and 9-hydroxydehydrogenase and 13-reductase in the developing rat kidney.

Three prostaglandin F2alpha-catabolizing enzyme activities have been demonstrated in kidneys from adult rats. Activity of each of the enzymes varied with animal age. Whereas 15-hydroxydehydrogenase and delta13-reductase appeared important to the early developing kidney (prior to 4 weeks of age), 9-hydroxydehydrogenase appeared to be characteristic of the adult kidney. Prostaglandin 15-hydroxydehydrogenase rose sharply after birth to a maximal value at 19 days (59-fold relative to the adult) decreasing to adult values by Day 40. Prostaglandin delta13-reductase followed a similar pattern rising about 20-fold at Day 19. Prostaglandin 9-hydroxydehydrogenase, on the other hand, was undetectable up to Day 19, rising gradually to adult values by Day 50. Prostaglandin biosynthesis in whole kidney and renal papilla at the peak period of 15-hydroxydehydrogenase activity, i.e. 19, 22, and 24 days, did not vary significally from adult values. The dramatic rise in 15-hydroxydehydrogenase activity, reflecting an important requirement for prostaglandin inactivation during the first 3 weeks after birth, appears to correlate well with the increase during this period in the number of glomeruli, cortical tubules, and redistribution of blood flow to the cortex. These results suggest for the first time an important relationship between prostaglandin catabolizing activites and nephrogenesis.


SUMMARY
Three prostaglandin Fz,-catabolizing enzyme activities have been demonstrated in kidneys from adult rats. Activity of each of these enzymes varied with animal age. Whereas 15-hydroxydehydrogenase and A13-reductase appeared important to the early developing kidney (prior to 4 weeks of age), 9-hydroxydehydrogenase appeared to be characteristic of the adult kidney. Prostaglandin 15-hydroxydehydrogenase rose sharply after birth to a maximal value at 19 days (59-fold relative to the adult) decreasing to adult values by Day 40. Prostaglandin A13-reductase followed a similar pattern rising about ZO-fold at Day 19. Prostaglandin O-hydroxydehydrogenase, on the other hand, was undetectable up to Day 19, rising gradually to adult values by Day 50. Prostaglandin biosynthesis in whole kidney and renal papilla at the peak period of 15-hydroxydehydrogenase activity, i.e. 19,22, and 24 days, did not vary significantly from adult values. The dramatic rise in 15-hydroxydehydrogenase activity, reflecting an important requirement for prostaglandin inactivation during the first 3 weeks after birth, appears to correlate well with the increase during this period in the number of glomeruli, cortical tubules, and redistribution of blood flow to the cortex. These results suggest for the first time an important relationship between prostaglandin catabolizing activities and nephrogenesis.
The biosynthesis and catabolism of prostaglandins in kidneys of several species have been the subject of considerable attention (2)(3)(4)(5)(6)(7)(8)(9)(10)(11). In the rabbit, a dissociation between the prostaglandin synthetase and the major catabolizing enzyme 15-hydroxydehyclrogenase has been observed, the former predominantly located in the papilla and the latter in the cortex (12). This finding has led to considerable speculation regarding the site of action of the naturally occurring and predominant renal prostaglandin types, Az, Ez and Fz,l (12)(13)(14).
* This work was supported by Grant MA4181 from the Medical Research Council of Canada. This is Paper IV of the series, "Prostaglandins During Development." Part of this work was presented at the 58th Annual Meeting of the Federation of American Societies for Experimental Biology and appeared in abstract form (1). 1 The abbreviations used are: PGF,, prostaglandin Fa (9cu,lla, 15(S)-trihydroxyprostad,13-dienoic acid); 15K-PGFh,, 15-keto-Little work has been reported on developmental changes in the prostaglandin system (biosynthesis and catabolism). Our first study with rat lung (15) showed interesting variations in product types and enzyme activities between rats of various ages from birth to the adult. High prostaglandin inactivation appears to be an important feature of the early developing tissue (15).2 Since the swine kidney has been reported to contain the highest 15hydroxydehydrogenase activity of all tissues tested in this species (6), we decided to study next the kidney from developing rats in order to determine whether our findings of high prostaglandin inactivation during early development would also be applicable to this tissue, and to investigate whether a correlation might exist between changes in prostaglandin inactivation with nephrogenesis.
EXPERIMENTAL PROCEDURE Materials Prostaglandin standards were generously supplied by Dr. J. E. Pike and Dr. U. AXen, The Upjohn Company, Kalamazoo. Tritium-labeled prostaglandin metabolites were prepared by the methods described below using only tracer as substrate.
Animals-Pregnant rats (Wistar strain) were purchased from a local supplier and housed in our animal quarters. Birth time was approximated to within a few hours by checking the animals periodically.
Male Wistar rats weighing 250 to 300 g are referred to as adult in the subsequent sections.
Diethvl ether was fresblv distilled before use in a Buchler rotary evaporator. Water was deionized and then distilled from six male adult rats (250 to 300 g) ping small portions of N-methyl-N'-nitro-N-nitrosoguanidine were used in this section since this product was best formed from (1 g) in a mixture of ice-cold KOH (10 g) in water (10 ml) and adult kidneys. Homogenates were prepared in 20 volumes of buffer diethyl ether (10 ml). The ethereal diaaomethane was subse-and incubated 20min at 30" in two Erlenmeyer flasks each containquently distilled.
ing NAD+ (4 mM final concentration), 3  The right kidney from one adult was routinely used as control with every assay of renal catabolism in newborn rats. Usually four newborn rats were used except at the early ages (fetal, 1 and 6 days after birth) when the right kidneys from eight rats were pooled.

Incubations
were terminated by addition of 5 volumes of absolute ethanol. The precipitated protein was centrifuged at 1800 rpm for 15 min (Sorvall GLC-1 operated at ambient temperature), and the clear supernatant solution was transferred to small pearshaped flasks (25 ml) and evaporated to complete dryness. The residue was resuspended in ethanol (5 ml) and aliquots (l/10) were withdrawn for measurement of radioactivity. The samples were taken to dryness again, resuspended in 300 ~1 of chloroformmethanol (1:l (v/v)), and an aliquot (l/4) was spott.ed on Silica Gel G thin layer &at& (5 X 20 cm). A mixture ofstandards (5 rg each) of arachidonic acid. 12-hvdroxvstearic acid, PGAZ. 15KD-PGF'b, PGE,, and PGFk'was &so spotted. The plates were developed in a freshly prepared solvent mixture of chloroformmethanol-acetic acid-water (90:9: 1:0.65 by volume) for 2 hours at room temperature.
After development the radioactive portion of the plate was covered with a glass plate and the portion where standards were spotted was sprayed with an ethanolic solution of phosphomolybdic acid (1O~c). Spots were visualized by gentle heating with hot air. The radioactive zones were marked after scanning the plate on a Panax thin layer radioscanner and quantitated by scraping zones of silica gel corresponding to peaks of radioactivity into scintillation vials, addition of water (1 ml) and Aquasol (10 ml), and counting in a liquid scintillation spectrophotometer (Beckman LS-255) at ambient temperature.

Large Scale Preparation of Metabolifes
16K-PGF% and 16KD-PGFk-Kidneys from seven newborn rats (Day 20) were prepared as described above, homogenized in 20 volumes of buffer, and incubated 20 min at 30" in an Erlenmeyer flask (500 ml) containing NAD+ (final concentration 4 mM), 1 X 107 dpm [9&8H1]PGFb diluted with 1 mg of unlabeled PGFb. Incubation was terminated with 5 volumes of absolute ethanol. After suction filtrat.ion through Whatman No. 42 filter paper, the resulting clear filtrate was evaporated to complete dryness. Recovery of radioactivity in the residue was 98%. Radio thin layer chromatographic analysis of an aliquot of the residue indicated the following distribution of compounds: unreacted PGFp, (lye), 15K-PGF* (48%), and 15KD-PGFk (47%). The metabolites were purified by preparative thin layer chromatography (10 plates, 20 X 20 cm, 0.250 pm thickness) using tbe above mentioned solvent system and as previously described (15). up as above. Recovery of radioactivity was better than 95yc. Radio thin layer cbromatographic analysis showed two peaks of radioactivity, one migrating with Rp 0.40 as in the reference starting material, and a less polar product, RF 0.45. Complete resolution of both products was achieved by silicic acid column chromatography (Bio-Sil HA-minus 325 mesh, Calbiochem). The mixture was introduced to a column (1 X 15 cm) of silicic acid in benzene and eluted with the following proportions of ethyl acetate in benzene: O:lO, 2:8, 4:6, 6:4, and 10:O. The starting material (15KD-PGFk) was found in ethyl acetate-benzene 6:4 fraction and the metabolite (15KD-PGE,) was obtained in the 4:6 fraction. Each product was over 90% pure as judged by radio thin layer chromatography.
Further purification of the metabolite was required and was achieved by preparative thin layer chromatography as in previous section.

Biosynthesis Experiments
The conversion of tritiated arachidonic acid (['Hn] 20:4, 2.2 X 10' dpm) into PGE, and PGFb was measured in homogenates of whole kidney from one adult and one 22-day-old rat and in homogenates of papilla from kidneys from adult, 19-, and 24-dayold rats. Papilla from kidneys from three adult and seven newborn rats were removed, -shed with cold 0.05 M KHZPOI-NaOH, pH 7.4, buffer, and homogenized in 20 volumes of buffer in a Polytron tissue homogenizer (10 s, 0", top speed). Aliquots (1.5 ml) were incubated (10 min, 37", 01) in flasks (25 ml) containing tracer arachidonic acid. Reaction was terminated by the addition of water (5 ml), freshly distilled diethyl ether (10 ml), and enough 0.05 N HCl to bring the aqueous phase to pH 3.0. The aqueous phase was extracted twice with diethyl ether (30 ml) and the diethyl ether layers were washed to neutrality with several small portions of water (5 ml). The combined diethyl ether phases were taken to dryness in vacua and analyzed for distribution of radioactivity by radio thin layer chromatography (Table II).

Sodium Borohydride and Borodeuteride Reduction
The sample (20 to 50 pg) was dissolved in 2 ml of methanol and placed in a glass-stoppered flask maintained at 0". Twenty milligrams of NaBlH, or NaB*Hd were added slowly to the stirred solution. After 30 min, the mixture was diluted with 10 ml of diethyl ether, 10 ml of water, and enough N HCl to bring the pH to 3. The contents were then transferred to a separatory funnel, diluted with diethyl ether and water, and extracted. The ether extract was subsequently washed to neutrality with water and evaporated to complete dryness. An aliquot was assayed for completion of reduction by radio thin layer chromatography and the sample was used without further purification for gas chromatography and mass spectrometry. Usually complete reduction was observed under these conditions.

Detivatives for Gas Chromatography-Mass Spectrometry
Samples (20 to 50 pg) were converted to methyl esters in 50-~1 glass vials by the addition of 2 ~1 of methanol and 18 ~1 of freshly DreDared and distilled solution of ethereal diazomethane.
After 60 bin at 23", the solvent was blown off with a fine stream of nitrogen and the residue was converted to trimethylsilyl ethers by the addition of 20 d of Tri Sil Z (Pierce Chemical Co.) and heating at 60" for 5 min. Gas chromatography retention data of the isol;ted metabolite prior to reduction with sodium borohydride was obtained on the methyl ester methoxime trimethylsilyl ether derivatives. The methyl esters were first converted to methoxime derivatives by reacting overnight (23") in 15 ~1 of MOX reagent (Pierce Chemical Co.). The solvent was blown off with nitrogen and the residue converted to trimethylsilyl ethers as described above.

Gas Chromatography and Mass Spectrometry
Gas chromatograph retention data were obtained on methyl ester trimethylsilyl ether derivatives of the metabolites before and after sodium borohydride or sodium borodeuteride reduction (Table 1) as well as on methyl ester methoxime trimethylsilyl ethers. Analyses were performed on a Hewlett-Packard model 5711 equipped with flame ionization detector (300") and a stream splitter attached to a Packard model 892 radioactivity proportional counter. The heated inlet to the proportional counter was maintained at 300". The gas chromatograph contained a glass column filled with 3% SE-30 on Gas Chrom Q (Applied Science) maintained at 250". Mass spectra were recorded on a Varian MAT CH-5 mass spectrometer coupled to a gas chromatograph and computer (Varian PDP-8) assembly. The same column packing was used as in the gas chromatographic analyses and this column was maintained at 250". The electron energy was 70 e.v.

Protein Measurement
Protein in homogenates was assayed by the method of Lowry et al. (17).

Method of Expression of Data
Each assay for 15-hydroxydehydrogenase and A13-reductase activity was performed on a fixed quantity of tracer substrate (PGFh) containing at least four different concentrations of unlabeled substrate. In this way, the amount of product formed could be determined by the percentage radioactivity in the various products since the specific activity of the precursor in each tube was known. This allowed construction of saturation curves for each homogenate (Fig. 4). Maximal activity was calculated from these curves and is expressed as picomoles/min/mg of protein. Since 15-hydroxydehydrogenase activity is the first step in the catabolism of PGFk and the product, 15K-PGFp,, is further catabolized by A13-reductase into 15KD-PGFh, which in turn is converted into 15KD-PGEa (16), total 15-hydroxydehydrogenase activity is represented by the sum of IBK-PGFb, 15KD-PGFb, and 15KD-PGE2 formed. Similarly A13-reductase activity is represented by the sum of 15KD-PGE* and 15KD-PGF?,, formed. Q-Hydroxydehydrogenase activity is expressed as per cent of activity observed in the adult using tracer 15KD-PGFb as substrate (16). An incubation period was chosen such that only 66% of substrate was converted by the adult,.

Incubation
of homogenates of adult rat kidney with tritiumlabeled prostaglandin Fz, gave three main metabolites (Fig. 1). Sufficient material for mass spectrometric characterization was obtained from large scale incubations. The thin layer chromatographic mobility of these compounds suggested that one or two keto groups had been introduced into PGF2,. Direct evidence for this as well as the localization of the keto groups was obtained by sodium borodeuteride and sodium borohydride reduction of the metabolites with subsequent comparison of the mass spectra of these reduced derivatives with that of authentic PGFz, (18). Sodium borohydride reduction of Metabolite I (RI 0.32) gave two more polar radioactive products (RF 0.22 and 0.15, Table I, Fig. 2) whose mass spectrum was identical with that of authentic PGF2, (RF 0.15); The mass spectra of derivatives after treatment of Metabolite I with sodium borodeuteride indicated that one deuterium atom had been incorporated during chemical reduction. The location of the deuterium atom was placed at carbon-15 since the fragment at m/e 173, due to C&CH=&Si(CHJa in PGFz,, was retained in the sodium borohydride reduced derivative of Metabolite I and shifted to m/e 174 in the sodium borodeuteride reduced derivative. Metaboiite I was identified as 15-keto-prostaglandin Fz, and the two products after chemical reduction represented isomers around carbon 15. Reduction of Metaboiite II (RF 0.40) with sodium borohydride gave only one radioactive product (Rp 0.23, Table I, Fig. 2) as determined by thin layer chromatography. Its mass spectrum showed a molecular ion two mass units higher than PGFz,, Incubation (10 min at 30") was terminated with the addition of absolute ethanol (2.5 ml). The resulting precipitate was centrifuged and the supernatant was transferred and evaporated to complete dryness in uocuo. The residue was resuspended in 300 ~1 of chloroformmethanol (2:l) and an aliquot (l/4) was spotted on thin layer plates (5 X 20 cm) and developed with chloroform-methanol-acetic acid-water (90:9:1:0.65, (v/v)). Standards (5 pg each) were spotted on the side and visualized after development by spraying with an ethanolic solution of phosphomolybdic acid (w/v, 1:lO) and gentle heating. Radioactivity profile was obtained by scanning the plate with a Panax radiochromatogram scanner.
suggesting that a biological reduction of one of the two double bonds of the substrate had taken place during incubation. The reduced double bond was located at positions 13,14 since the relative intensity of the fragment at m/e 173 was much greater than that from PGFz, as would be expected after saturation of the double bond at A13. Reduction of Metabolite II with sodium borodeuteride gave a product with a molecular ion three mass units higher than prostaglandin Fk,, indicating that Metabolite II had only one double bond and one keto group. The location of the deuterium atom at position 15 was established in a similar way to Metabolite I. Metabolite II was therefore 15keto-13,14dihydroprostaglandin Fh and the two isomers after chemical reduction had the same chromatographic mobility. Metabolite III (RF 0.45) gave two radioactive products after sodium borohydride reduction (8~ 0.23, 0.15; Table I, Fig. 2) as expected of 15-keto-13,14dihydroprostaglandin Es. Mass spectral analysis of the reduced compound showed a molecular ion two units higher than PGFz,, with a pattern very similar to the reduced product of Metabolite II. Reduction of Metabolite III with sodium borodeuteride and subsequent mass spectral analysis showed a displacement of the molecular ion by two mass unite from the sodium borohydride reduced product indicating that two deuterium atoms had been incorporated during chemical reduction. Comparison of the mass spectra of the deuterated and protonated reduced derivatives of Metabolite III and with PGFpor indicated that both fragments at m/e 173 and 217 in the 2798 (A to C) treated in a similar way. A, 15K-PGFk; B, 15KD-PGF2.; C, 15KD-PGEi. Reference compounds are shown on the extreme left and right of the plate. Spots were visualized by spraying the plate with an ethanolic solution of phosphomolybdic acid (w/v 1:lO) and gentle heating. Samples to be reduced (20 to 50 pg) were dissolved in 2 ml of distilled and dry methanol, and sodium borohydride (20 mg) was added slowly to the stirred solution maintained at 0". After 30 min at O", the mixture was diluted with distilled diethyl ether (10 ml), water (5 ml), and enough N HCl to adjust the pH to 3.0. The mixture was extracted in a separatory funnel and the diethyl ether phase was washed to neutrality with water. Evaporation of the diethyl ether phase to complete dryness in vacua gave a residue which was analyzed by thin layer chromatography using the same developing system as in Fig. 1.
mass spectra of the protonated derivatives were shifted to m/e 174 and m/e 218 in the deuterated derivative. Since these fragments are derived from parts of the molecule containing carbon 15 and carbon 9, respectively, i.e. CbHiiCD=&-Si(CH,), and (CH,)$i-;)=CD-CH=CH-0-Si(CH,)3, then the deuterium atoms (and therefore the original keto groups) are located at the 9 and 15 positions. Metabolite III is therefore 15-keto-13,14dihydro prostaglandin Ez and the two products obtained after chemical reduction constituted isomers about the 9 and 15 positions.
That the keto group was not at position 11 instead of position 9 was ruled out by a separate set of experiments in which prostaglandin Fzo specifically labeled at the 90 position was incubated with adult kidney homogenates. A time-dependent elimination of tritium from the S&position was shown in the preceding paper (16). No such loss of radioactivity would have occurred if the 11-hydroxyl group were oxidized to a keto group. Prostaglandin 9-and 15-hydroxydehydrogenase and A13reductase activities were linear up to 10 min at 30" (Fig. 3). The effect of increasing substrate concentration on prostaglandin 15-hydroxydehydrogenase and Ala-reductase is shown in Fig. 4.
V ,,,&= for adult kidney was 6.1 f 3.5 S.D. pmol/min/mg of protein for 15-hydroxydehydrogenase and 3.8 Z!G 1.6 SD. pmol/ min/mg of protein for A13-reductase, and varied significantly with the age of the animal. Fig. 5 shows the activity profiles for the three enzymes measured as a function of increasing animal age. At gestational age 20 days, 15-hydroxydehydrogenase activity was 1Bfold greater than the adult and rose linearly to a peak at 19 days postnatally, 59-fold higher than the adult, decreasing thereafter to adult levels by Day 40. A similar but less spectacular increase occurred with Ala-reductase activity (6-fold higher than adult at gestational and A13-reductase activity (O-O) in adult rat kidney homogenates. Each point represents the mean from 10 experiments. Incubation conditions are the same as in Fig. 1    Papillae from three adult and seven newborn rats were removed, homogenized in 20 volumes of 0.05 Y KHzPO,-NaOH, pH 7.4, and aliquots (1.5 ml, approximately 7 mg of protein) were incubated (10 min, 37') in an oxygen atmosphere in flasks containing 2.2 X lo6 dpm tracer tritiated arachidonic acid. Biosynthesis was also measured in whole kidney from an adult and a 22-day-old rat. Incubations were stopped by the addition of water (5 ml), freshly distilled diethyl ether (10 ml), and enough 0.05 N HCI to bring the aqueous phase to pH 3. The ether phase was subsequently washed to neutrality and evaporated to complete dryness. Aliquots (x) were spotted on thin layer chromatoplates (Brinkmann, 5 X 20 cm) and developed in chloroform-methanol-acetic acid-water (90:9:1:0.65 (v/v)). The plates were scanned on a Panax radiochromatogram scanner and the radioactivity in zones along the plate was quantitated in a liquid scintillation spectrophotometer after scraping the zones into scintillation vials, and the addition of water (1 ml) and Aquasol (10 ml). o Fifty micrograms of unlabeled PGEz and PGFz,, were added at the start of the incubation to minimize degradation of the tritiated prostaglandin formed during incubation.

DISCUSSION
The activity of three enzymes involved in the conversion of PGF2, into 15KD-PGEz have been measured in the rat kidney (Fig. 6) and found to vary with animal development. In an earlier study (15) we also observed that prostaglandin catabolism in developing rat lung varied with animal age. In that study it The cortical region of the rat kidney is very immature at birth. Its morphology develops almost completely within a brief 3-week postnatal period. The main thrust of nephrogenesis and functional differentiation of tubules occurs within 4 weeks after birth. During this period the number of glomeruli increases almost 3-fold and a marked redistribution of blood flow to the cortex takes place (20)(21)(22)(23)(24)(25)(26).
Prostaglandin Ez and Fz. are potent vasoconstrictors in the rat kidney (27). Although no studies are available describing the effects of these compounds on kidneys from newborn rats, this tissue probably behaves in a similar fashion to that from an adult rat although the extent of its sensitivity to prostaglandins might be different from that of the adult animal. Prostaglandins, unless eliminated, would therefore be expected to interfere with and likely inhibit blood flow to the cortex during this critical period (1 to 4 weeks). Our observations show that the developing kidney is capable of catabolizing prostaglandins more efficiently (59-fold greater than the adult) during the first 3 weeks after birth (Fig. 5).
The rapid postnatal rise in prostaglandin catabolism does not reflect an over-all increase in the turnover of prostaglandins since prostaglandin biosynthesis in the whole kidney as well as in the papilla, the main site of the prostaglandin biosynthetic enzymes (12), does not vary appreciably between the peak period of prostaglandin catabolism (Day 19 to 24) and the adult (Table II). In fact, our observation of a higher catabolism in the newborn (59-fold) with similar biosynthetic activity relative to the adult suggests that during the first 4 weeks of postnatal development, the rat kidney is in a state of local prostaglandin deficiency. This may be an important requirement to permit the proper renal blood flow needed for cortical development. Interference with prostaglandin catabolism during this period may be expected to result in abnormal renal corticogenesis. Studies along these lines are presently in progress.