Acta physiol. scand. 1978. 103. 100-106 From the Department of Pharmacology, Karolinska Institutet, Stockholm, Sweden
The influence of dietary sodium on urinary prostaglandin excretion BY
D. DAVILA,’T. DAVILA,’ E. OLIWand E. ANGGARD Received 30 November 1977
Abstract DACILA, D.. T. DAVILA, E. OLIWand E. ANGGARD. The influence of dietary sodiirrn on urinary prostuglandin escretioii. Acta physiol. scand. 1978. 103. 100-106. The influence of dietary sodium chloride on the urinary excretion of prostaglandins (PGs) was studied in unanesthetired female rabbits housed in metabolic cages. Urinary PG levels were determined by radioinmunoassay. bioassay and gas chromatography-mass spectrometry. In the first experiment rabbits were fed at high ( 2 . 5 “ , ) and later a low ( O . l S ~ ’ , ) sodium chloride diet ad libitrtnr. A 2-5 fold increase in excretion of immunoreactive PGFyz (iPGF2x) and iPGE, was noticed when animals were given the low salt diet. Since it could not be excluded that dietary factors other than sodium chloride contributed to the changes a second, more controlled, experiment was undertaken. Rabbits were fed 30 g/kg per day of diets differing only in the content of sodium chloride, 2 and 0.37 uI, respectively. On the high salt diet the rabbits excreted 0.1 fO.O4pg/day of PGE, and 2.0&0.5/rg/day of iPGF.L1. After equilibration on the low salt diet the PGE: excretion rate increased to 1 . 5 1 0 . 3 pg/day (p-:O.OOI) and that of iPGFzm to 3.4?0.4pg/day ( p c 0.01). These results thus point to an inverse relationship between renal sodium excretion and the activity of the renal prostaglandin system.
The role of the prostaglandins (PGs) in the renal handling of sodium is unsettled. Early studies using infusion of E and A types of PG demonstrated natriuretic and diuretic effects (Herzog et a/. 1967, Vander 1968, Lee 1973). Similarly a n infusion of arachidonic acid. a PG precursor, into the renal artery caused natriuresis in doses which did not alter renal blood flow o r glomerular filtration rate (Chang et a / . 1975, Tannenbaum et ul. 1975, Bolger et a / . 1976). On the other hand meclofenamate increases renal sodium excretion both in the conscious and in the anesthetized dog in response to volume expansion (Kirschenbaum and Stein 1976, 1977, cf. Oliw rt a / . 1978a). Furthermore, Tobian et al. (1974) and Tobian and O’Donnell ( 1976) have shown that sodium loading in rats depresses intrarenal concentration of PGE, whereas deprivation has the opposite effect. Quite recently Weber ~t u / . (1977) reported that a high sodium diet depresses the renal excretion of PGE, in the rabbit. I n the present study we have attempted to evaluate the effect of the dietary sodium on the renal excretion of PGs in the conscious rabbit. As has been shown earlier the primary
’ Permanent
address: Department of Pharmacology. Medical School, Pristina, Yugoslavia.
DIETARY SODIUM ON PG EXCRETION
101
TABLE I. Description of rabbit diets. Diet
Sodium chloride content ( 7 6 )
Source
1
0.25
2
2.5
3
0.37
4
2.0
5
0.43
Altromin GmbH Lage, Lippe, West Germany Altromin GmbH Lage, Lippe, West Germany Astra-Ewos AB, Sodertalje, Sweden Astra-Ewos AB, Sodertalje, Sweden Astra-Ewos AB, Sodertalje, Sweden
urinary PGs originate within the kidney (Frolich et al. 1975, Williams et al. 1977) and their measurement therefore constitute an index of intrarenal biosynthetic activity. Our results show that sodium deprivation markedly enhances renal excretion of PGs pointing to the possibility of an antinatriuretic role of the PGs.
Material and Methods The rabbits were nonpregnant albinos weighing 2-2.5 kg at the start of the study. They were maintained on either of the diets shown in Table I. The animals were kept in metabolic cages with one rabbit in each cage. The 24-hour urines from each rabbit were collected separately. Special care was taken to avoid degradation of the PGs by collecting the urine in traps surrounded by dry ice. The urine was thus frozen very soon after being voided and kept frozen until analyzed. E~perinzentalprotocol:Two experimental designs were used. In the first the rabbits were fed the normal diet (5). high sodium diet (2) and low sodium diet (1) ad libitum. On each diet animals were allowed to equilibrate for one week and data were then collected for 4 days. It was observed that intake of both food and water varied considerably between the high sodium diet and the other dietary regimens. With the low sodium diets the food intake was 3-4 times that of the high sodium diet. In a second series of expts. we therefore decided to give the animals different amounts of sodium while keeping the intake of food constant. The diets were given in the following order: high sodium diet (4) for 6 days, normal diet ( 5 ) for 6 days and low sodium diet (3) for 6 days. The first 3 days were allowed for equilibration and urine was collected daily for another 3 days. Analysis: Urinary sodium and potassium was measured using flame photometry (Instrumentation Laboratory Inc., Model 143). PGFz, in urine was measured by radioimmunoassay (RIA). This procedure has been described in detail elsewhere (Oliw ef al. 1978 b). In brief, the precision of the method was evaluated by replicate (n = 6) determination of the same samples. The coefficient of variation was determined on 4 separate occasions and found to be 10.7”/6, 13.4%, 12.7% and 18.5% (mean 13.8%). The validity of the method was assessed by co-determination of 6 different samples with gas chromatography-mass spectrometry (GC-MS) using deuterated PGFeor as an internal standard. The results are shown in Table 11. It is seen that the values obtained by the two methods are in close agreement, thus demonstrating the validity of the RIA method. In experimental design 1 PGE, was reduced by NaBH, to PGFza and PGFtP and PGF,B measured by RIA using a PGF,B-antiserum (Oliw et al. 1978b). In experimental design 2 PGE, was determined by bioassay on the rat fundus strip essentially according to Larsson and Anggird (1974). To 20 ml of rabbit urine was added 30.000 cpm of 3H-labelled PGE, (100-200 Ci/mmole, New England Nuclear, Darmstadt, West Germany). After adjusting the pH to 3 using 30-40 drops of 5 0 % formic acid, the sample was extracted twice with 20 ml of CHCI,. The solvent was evaporated to dryness and applied on a 1 g
I02
u.
D A V I L A , T. DAVILA, E. OLIW AND E. ANGGARD
TABLE 11. Comparison of P G F L ' determination ~ by radioimmunoassay and gas chromatography-mass spectrometry in urine. RIA ng,'ml
GC-MS ng;ml
36 21
36 ?4
0 38
3
37
4
20 27 32
31 '3 28 39
I? 4 18
2917
33x6
12214
Sample
I
3
5 6
Mean
-
S.D.
Difference o0
0
silicic acid (Uiiisil, 100-200 mesh, Clarkson Chem. Comp., Inc., Williamsport, Pa., USA) column in 3 0 : ; ethyl acetate in toluene. After 30 ml has been allowed to pass through the column the PGEJike material *as eluted using 60 ml of 60'>,,ethyl acetate in toluene. This fraction was evaporated to dryness, dissolved in 1.5 ml of 96"" ethanol and kept at 20-C until assayed. The recovery of 3H-PGE, through this isolation procedure was about 30"0 (range 25-57 Yb). The recovery of added 50 ng of PGE, was assessed in 1 expt. and after correction of losses of 3H-PGEI found to be 98"". Prior to bioassay the samples were reduced t o dryness and reconstituted in 1 ml of Tyrode solution. The assay of the rat fundus strip was carried out according to Weeks c v al. (1968). The validity of the assay was further checked by treatment of the samples with dilutc NaOH. This procedure, which coverts all PGE-compounds to the respective PGB-compounds, resulted in complete biological inactivation. The level of the PGE2 present in the original sample was obtained using the data from the bioassay and the recovery of radioactive PGE, in each sample. Statistical significance was determined by Student's t-test. ~
Results In the first experiment rabbits were fed a normal diet, a high (2.5";) and a low (0.25y0) sodium chloride containing diet ad libitum. The excretion of prostaglandins, sodium and water is shown in Fig. 1 . An inverse relationship was found to exist between the excretion of sodium on the one hand and of iPGF,, on the other hand. When the rabbits were fed the normal rabbit chow diet (5) containing 0.43 3'; NaCl a 3fold interindividual variation in output of iPGF,, was seen. However, the intraindividual variation was relatively sniall (cf. Fig. 1). The mean excretion of iPGF,, was 2.9 i- 1.7 pg/24 h (mean I S.D.). Change to the higher sodium diet led to a decrease in the renal output of iPGF,, to a mean value of 0.69 20.3 pg/24 h but increased both sodium and water excretion. When, during the last stage of the expts., the rabbits were eating the diet, with the lowest content of sodium the urinary PGF,, excretion again increased to a mean level of 3.5 :0.7 pg/24 h. The mean level of iPGF,, and iPGE, for the two later diets are shown in Table 111. The differences on PG excretion between the two diets were statistically significant (both p .:0.001). It thus appeared that there might be an inverse relationship between the renal excretion of sodium and the excretion of iPGF,, and iPGE,. However, two factors led us to undertake a second more controlled experiment. Firstly, the rabbits ate 3-4 times more of the low sodium diet as compared to the high sodium diet. Secondly, the rabbits drank roughly 3 times more during the high salt diet as compared to the low salt diet. These differences
103
DIETARY SODIUM O N PG EXCRETION
Fig. 1. The effect of dietary sodium on the daily excretion of immunoreactive PGFz,, sodium and water in 6 different rabbits fed ad libitirnz. The registrations (mean of 4 days* S.E.) of variables on each diet were made after an equilibration period of one week. Figures under the bars denote each individual rabbit.
I23456
I23456
I23456
Lzz./
could possibly alter the renal output of PGs by mechanisms unrelated to renal handling of sodium. In the second series of expts. the rabbits were restricted to a food intake of 30 g/kg/day. Water was offered adlibiturn. The animals were first given the high sodium diet (4) and then low sodium diet (3) and measurements were made after three days of equilibration on each diet. The results are shown in Fig. 2. Again an inverse relationship was found between PG excretion and sodium output. The values for urinary volume, pH and output of potassium were constant between the two dietary regimens. The changes in iPGF,, and PGE, excretion were significant (p
The influence of dietary sodium on urinary prostaglandin excretion BY
D. DAVILA,’T. DAVILA,’ E. OLIWand E. ANGGARD Received 30 November 1977
Abstract DACILA, D.. T. DAVILA, E. OLIWand E. ANGGARD. The influence of dietary sodiirrn on urinary prostuglandin escretioii. Acta physiol. scand. 1978. 103. 100-106. The influence of dietary sodium chloride on the urinary excretion of prostaglandins (PGs) was studied in unanesthetired female rabbits housed in metabolic cages. Urinary PG levels were determined by radioinmunoassay. bioassay and gas chromatography-mass spectrometry. In the first experiment rabbits were fed at high ( 2 . 5 “ , ) and later a low ( O . l S ~ ’ , ) sodium chloride diet ad libitrtnr. A 2-5 fold increase in excretion of immunoreactive PGFyz (iPGF2x) and iPGE, was noticed when animals were given the low salt diet. Since it could not be excluded that dietary factors other than sodium chloride contributed to the changes a second, more controlled, experiment was undertaken. Rabbits were fed 30 g/kg per day of diets differing only in the content of sodium chloride, 2 and 0.37 uI, respectively. On the high salt diet the rabbits excreted 0.1 fO.O4pg/day of PGE, and 2.0&0.5/rg/day of iPGF.L1. After equilibration on the low salt diet the PGE: excretion rate increased to 1 . 5 1 0 . 3 pg/day (p-:O.OOI) and that of iPGFzm to 3.4?0.4pg/day ( p c 0.01). These results thus point to an inverse relationship between renal sodium excretion and the activity of the renal prostaglandin system.
The role of the prostaglandins (PGs) in the renal handling of sodium is unsettled. Early studies using infusion of E and A types of PG demonstrated natriuretic and diuretic effects (Herzog et a/. 1967, Vander 1968, Lee 1973). Similarly a n infusion of arachidonic acid. a PG precursor, into the renal artery caused natriuresis in doses which did not alter renal blood flow o r glomerular filtration rate (Chang et a / . 1975, Tannenbaum et ul. 1975, Bolger et a / . 1976). On the other hand meclofenamate increases renal sodium excretion both in the conscious and in the anesthetized dog in response to volume expansion (Kirschenbaum and Stein 1976, 1977, cf. Oliw rt a / . 1978a). Furthermore, Tobian et al. (1974) and Tobian and O’Donnell ( 1976) have shown that sodium loading in rats depresses intrarenal concentration of PGE, whereas deprivation has the opposite effect. Quite recently Weber ~t u / . (1977) reported that a high sodium diet depresses the renal excretion of PGE, in the rabbit. I n the present study we have attempted to evaluate the effect of the dietary sodium on the renal excretion of PGs in the conscious rabbit. As has been shown earlier the primary
’ Permanent
address: Department of Pharmacology. Medical School, Pristina, Yugoslavia.
DIETARY SODIUM ON PG EXCRETION
101
TABLE I. Description of rabbit diets. Diet
Sodium chloride content ( 7 6 )
Source
1
0.25
2
2.5
3
0.37
4
2.0
5
0.43
Altromin GmbH Lage, Lippe, West Germany Altromin GmbH Lage, Lippe, West Germany Astra-Ewos AB, Sodertalje, Sweden Astra-Ewos AB, Sodertalje, Sweden Astra-Ewos AB, Sodertalje, Sweden
urinary PGs originate within the kidney (Frolich et al. 1975, Williams et al. 1977) and their measurement therefore constitute an index of intrarenal biosynthetic activity. Our results show that sodium deprivation markedly enhances renal excretion of PGs pointing to the possibility of an antinatriuretic role of the PGs.
Material and Methods The rabbits were nonpregnant albinos weighing 2-2.5 kg at the start of the study. They were maintained on either of the diets shown in Table I. The animals were kept in metabolic cages with one rabbit in each cage. The 24-hour urines from each rabbit were collected separately. Special care was taken to avoid degradation of the PGs by collecting the urine in traps surrounded by dry ice. The urine was thus frozen very soon after being voided and kept frozen until analyzed. E~perinzentalprotocol:Two experimental designs were used. In the first the rabbits were fed the normal diet (5). high sodium diet (2) and low sodium diet (1) ad libitum. On each diet animals were allowed to equilibrate for one week and data were then collected for 4 days. It was observed that intake of both food and water varied considerably between the high sodium diet and the other dietary regimens. With the low sodium diets the food intake was 3-4 times that of the high sodium diet. In a second series of expts. we therefore decided to give the animals different amounts of sodium while keeping the intake of food constant. The diets were given in the following order: high sodium diet (4) for 6 days, normal diet ( 5 ) for 6 days and low sodium diet (3) for 6 days. The first 3 days were allowed for equilibration and urine was collected daily for another 3 days. Analysis: Urinary sodium and potassium was measured using flame photometry (Instrumentation Laboratory Inc., Model 143). PGFz, in urine was measured by radioimmunoassay (RIA). This procedure has been described in detail elsewhere (Oliw ef al. 1978 b). In brief, the precision of the method was evaluated by replicate (n = 6) determination of the same samples. The coefficient of variation was determined on 4 separate occasions and found to be 10.7”/6, 13.4%, 12.7% and 18.5% (mean 13.8%). The validity of the method was assessed by co-determination of 6 different samples with gas chromatography-mass spectrometry (GC-MS) using deuterated PGFeor as an internal standard. The results are shown in Table 11. It is seen that the values obtained by the two methods are in close agreement, thus demonstrating the validity of the RIA method. In experimental design 1 PGE, was reduced by NaBH, to PGFza and PGFtP and PGF,B measured by RIA using a PGF,B-antiserum (Oliw et al. 1978b). In experimental design 2 PGE, was determined by bioassay on the rat fundus strip essentially according to Larsson and Anggird (1974). To 20 ml of rabbit urine was added 30.000 cpm of 3H-labelled PGE, (100-200 Ci/mmole, New England Nuclear, Darmstadt, West Germany). After adjusting the pH to 3 using 30-40 drops of 5 0 % formic acid, the sample was extracted twice with 20 ml of CHCI,. The solvent was evaporated to dryness and applied on a 1 g
I02
u.
D A V I L A , T. DAVILA, E. OLIW AND E. ANGGARD
TABLE 11. Comparison of P G F L ' determination ~ by radioimmunoassay and gas chromatography-mass spectrometry in urine. RIA ng,'ml
GC-MS ng;ml
36 21
36 ?4
0 38
3
37
4
20 27 32
31 '3 28 39
I? 4 18
2917
33x6
12214
Sample
I
3
5 6
Mean
-
S.D.
Difference o0
0
silicic acid (Uiiisil, 100-200 mesh, Clarkson Chem. Comp., Inc., Williamsport, Pa., USA) column in 3 0 : ; ethyl acetate in toluene. After 30 ml has been allowed to pass through the column the PGEJike material *as eluted using 60 ml of 60'>,,ethyl acetate in toluene. This fraction was evaporated to dryness, dissolved in 1.5 ml of 96"" ethanol and kept at 20-C until assayed. The recovery of 3H-PGE, through this isolation procedure was about 30"0 (range 25-57 Yb). The recovery of added 50 ng of PGE, was assessed in 1 expt. and after correction of losses of 3H-PGEI found to be 98"". Prior to bioassay the samples were reduced t o dryness and reconstituted in 1 ml of Tyrode solution. The assay of the rat fundus strip was carried out according to Weeks c v al. (1968). The validity of the assay was further checked by treatment of the samples with dilutc NaOH. This procedure, which coverts all PGE-compounds to the respective PGB-compounds, resulted in complete biological inactivation. The level of the PGE2 present in the original sample was obtained using the data from the bioassay and the recovery of radioactive PGE, in each sample. Statistical significance was determined by Student's t-test. ~
Results In the first experiment rabbits were fed a normal diet, a high (2.5";) and a low (0.25y0) sodium chloride containing diet ad libitum. The excretion of prostaglandins, sodium and water is shown in Fig. 1 . An inverse relationship was found to exist between the excretion of sodium on the one hand and of iPGF,, on the other hand. When the rabbits were fed the normal rabbit chow diet (5) containing 0.43 3'; NaCl a 3fold interindividual variation in output of iPGF,, was seen. However, the intraindividual variation was relatively sniall (cf. Fig. 1). The mean excretion of iPGF,, was 2.9 i- 1.7 pg/24 h (mean I S.D.). Change to the higher sodium diet led to a decrease in the renal output of iPGF,, to a mean value of 0.69 20.3 pg/24 h but increased both sodium and water excretion. When, during the last stage of the expts., the rabbits were eating the diet, with the lowest content of sodium the urinary PGF,, excretion again increased to a mean level of 3.5 :0.7 pg/24 h. The mean level of iPGF,, and iPGE, for the two later diets are shown in Table 111. The differences on PG excretion between the two diets were statistically significant (both p .:0.001). It thus appeared that there might be an inverse relationship between the renal excretion of sodium and the excretion of iPGF,, and iPGE,. However, two factors led us to undertake a second more controlled experiment. Firstly, the rabbits ate 3-4 times more of the low sodium diet as compared to the high sodium diet. Secondly, the rabbits drank roughly 3 times more during the high salt diet as compared to the low salt diet. These differences
103
DIETARY SODIUM O N PG EXCRETION
Fig. 1. The effect of dietary sodium on the daily excretion of immunoreactive PGFz,, sodium and water in 6 different rabbits fed ad libitirnz. The registrations (mean of 4 days* S.E.) of variables on each diet were made after an equilibration period of one week. Figures under the bars denote each individual rabbit.
I23456
I23456
I23456
Lzz./
could possibly alter the renal output of PGs by mechanisms unrelated to renal handling of sodium. In the second series of expts. the rabbits were restricted to a food intake of 30 g/kg/day. Water was offered adlibiturn. The animals were first given the high sodium diet (4) and then low sodium diet (3) and measurements were made after three days of equilibration on each diet. The results are shown in Fig. 2. Again an inverse relationship was found between PG excretion and sodium output. The values for urinary volume, pH and output of potassium were constant between the two dietary regimens. The changes in iPGF,, and PGE, excretion were significant (p
