PROSTAGLANDINS
URINARY EXCRETION OF IMMUNORFACTIVE PROSTAGLANDIN E: AND THE EFFECT OF POSTURE
A CIRCADIAN RHYTHM
Robert E. Bowden,* James H. Ware, David L. DeMets, and Harry R. Keiser Hypertension and Endocrine Branch, and the Biometrics Research Branch National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland 20014
ABSTRACT The excretion of urinary innnunoreactive prostaglandin E (IPGE), sodium, potassium, creatinine and volume was studied in 4 hr collections in normal women at normal activity. iPGE exhibited a circadian rhythm with an amplitude of 29% and peak excretion at 4:55 P.M. There were also significant circadian rhythms for sodium, potassium, creatinine, and volume, all peaking in late afternoon. There were no significant changes either in the total excretion or in the circadian rhythms of IPGE, potassium, or creatinine excretion when the subjects remained in bed for an entire day while the circadian rhythms of sodium and volume were significantlymodified in amplitude and phase, respectively. Urinary aldosterone excretion decreased significantly when the subjects were at bed rest. iPGE excretion increased 33% when subjects were first recumbent and then erect for consecutive 4 hr periods on the same day (but when subjects were erect 1 day for a 4 hr period, iPGE excretion was lower by 32% than for the same 4 hr period the preceding day when they were recumbent). These data indicate that: 1) the sympathetic nervous system and renin-angiotensin-aldosteronesystem do not affect the circadian rhythm of urinary IPGE, and 2) short-term experiments of prostaglandin E excretion must be designed to avoid misleading results due to the circadian rhythm.
ACKNOWLEDGEMENT The authors are grateful to Dr. Frederic C. Bartter for his critical review of this manuscript and to Mrs. Marian Warner for secretarial help.
*Present address: Division of Cardiology, Stanford Medical School, Stanford, California 93405
JULY 1977 VOL. 14 NO. 1
151
PROSTAGLANDINS
INTRODUCTION Prostaglandins of the E series are vasodilators, produce natriuresis and diuresis when infused into the renal artery (1,2), and antagonize the action of antidiuretic hormone on toad bladder (3). McGiff and his associates have proposed that prostaglandins are involved in autoregulation of renal blood flow (4). Renal blood flow, as well as glomerular filtration rate and the excretion of sodium, potassium, and creatinine exhibit 24 hr periodicity, i.e., circadian rhythms (5-10). Little is known about the etiology of these rhythms. Prostaglandins may be involved in regulating the glomerular filtration rate which in turn could determine the circadian rhythms of sodium and potassium. Me undertook this study to determine if prostaglandin excretion exhibits a circadian rhytbm, and to determine if posture, which has great influence on renal blood flow and renal function, has an effect on prostaglandin E excretion. METHODS Two separate studies were performed. In the first, the influence of the time of day on immunoreactive prostaglandin E excretion was observed. The subjects were 6 normal white women, aged 19-22 years, who ate a diet containing, each day, 109 meq sodium and 100 meq potassium. All subjects ate the diet for at least 3 days before the start of any study, and were confined to the Clinical Center during study. Bach subject was allowed normal activity, and urine was collected by spontaneous voiding at 4 hr intervals for 24 hr. Subjects 3 - 6 were kept in bed for a second day of urine collections, except when they were allowed out of bed to void. In the second study, the influence of posture on immunoreactive prostaglandin E excretion was observed. The subjects were 15 normal women age 19-22 years (14 white, 1 black), who ate a 109 meq sodium and 100 meq potassium diet. The subjects remained recumbent for 4 hr 1 day and remained standing during the same 4 hr interval the next day. The subjects voided at the end of each 4 hr period. To test if circadian variations would modify the effects of posture on prostaglandin excretion, 2 subjects were also studied while they were first recumbent and then erect for consecutive 4 hr periods on the same day. Urine samples were refrigerated until extracted. Urinary sodium and potassium were measured by flame photometry and urinary creatinine was measured by the Jaffe chromogen reaction with picric acid. Urinary aldosterone was determined in subjects 3 - 6 by radioimmunoassay. Immunoreactive prostaglandin E-like material (IPGE) was measured in triplicate aliquots of urine by radioimmunoassayas described by Halushka et al (11). Briefly, 2000 dpm of 3HPGEl (NBN) in 0.02 ml of ethanol was -added to 10 ml of urine. Neutral lipids were extracted with hexane, and prostaglandins extracted with chloroform at pH 3.5. The samples were partially purified by chromatographywith a 13-cm by l-cm (length and diameter) sephadex LPI20column. The PGE fraction was eluted with a mixture of chloroform, heptane, ethanol and acetic acid (100:100:25:2),converted to PGB with 0.1 N methanolic RCH, and then appl -h to a 0.5 g
152
JULY 1977 VOL. 14 NO. 1
PROSTAGLANDINS
silicic acid column. The PGB fraction was eluted with a mixture of benzene and ethyl acetate (60:40) and the collected fraction was radioimmunoassayed with an antibody to PGB (1:75,000 dilution with a PGB standard). The PGB concentration was corrected for recovery and alz assays were performed without knowledge of which time period the sample represented. The presence of a circadian rhythm was determined by the method described by Ralberg -et al (12) with the statistical techniques of Tong (13). This method, known as the "Cosinor" method, involves fitting a sinusoidal curve to the responses of each individual and averaging the estimated amplitudes and phases of individuals. Because our data represented total output in a 4 hr period, the technique was modified slightly to fit the integral of the sine function, as suggested by Fort and Mills (14). Paired analyses were used where possible and a p value of 0.05 or less was considered significant. RESULTS The average excretions of urinary IPGE, sodium, potassium, creatinine and volume in 4 hr periods for the 6 normal subjects during normal activity are shown as percentages of the means for all 6 4 hr periods (Fig. 1). In all subjects urinary iPGE displayed a rhythmic fluctuation with a peak between 11 A.M. and 11 P.M. and a nadir between 3 A.M. and 7 A.M. There were similar rhythms for potassium and creatinine excretion and urine volume. Sodium excretion followed a similar pattern except in 1 subject in whom the lowest sodium excretion occurred between 3 P.M. and 7 P.M. When the integral of a sine function was fitted to the values for the excretion of iPGE as described in Methods, the average amplitude for iPGE excretion for the 6 normal subjects was 25 ng/4 hr (29% change from the mean). The mean acrophase for iPGE excretion, i.e., the time of peak excretion as estimated from the fitted sinusoidal curves, was 4~55 P.M. f 58 min (mean f S.D.). There was a statistically significant circadian rhythm for urinary iPGE excretion (p < .Ol). Similar analyses of the data for percent change from the mean amplitude, phase and time of peak excretion indicated that there were significant circadian rhythms for urinary potassium, sodium, and creatinine (p < .Ol> and volume (p < .05) (Table I). The mean acrophases are all similar and indicate that the excretion of all these substances peaked at close to the same time. iPGE was significantly correlated with urine volume, potassium and creatinine excretion, but not with sodium excretion (Table II). Figure 2 shows the pattern of iPGE excretion for the 4 normal subjects 3 - 6 who were studied first with normal activity and then with bed rest. There was a minor modification of the rhythm of iPGE excretion when the subjects were recumbent manifested by a shift in mean acrophase from 4~17 P.M. during normal activity to 5:47 P.M. when the subjects were at bed rest. This change was not statistically significant. There were also no significant changes in the rhythms of creatinine and potassium excretion between the 2 days. Sodium excretion and urinary volume were significantly different between the 2 days (p < .05); the amplitude of sodium excretion changed from 13 meq to 23 meq without a change in phase,
JULY 1977 VOL. 14 NO. 1
153
PROSTAGLANDINS
w
I++% ______
PERCENT OF MEAN 150 r RELATIVE CHANGE _
61 .ClM
---
z
9
60
y4 __- __ __ --___ z&-I- 100 =,9 Y 850160I++? 200 us
4: 13
--
-
__
--_ -
li)lJ 0 k++i
MEAN6.D. I
1
1
I
I
I
,
TA 3p 11p 7A llA 7P 3A TIME OF DAY
Figure
154
1.
Relative changes in urinary IPGE, creatinine, sodium, potassium, and volume (expressed as percentages of the mean daily excretion) are shown for consecutive 4-hr periods in 6 subjects during normal activity.
JULY 1977 VOL. 14 NO. 1
PROSTAGLANDINS
160 ----AD
LIB ACTIVITY -BED
RESTI-1
145
P
130
z:
115
8
loo
2 e
85
I 70
55 40
I
IA
11”
3P
7p
11”
3”
1
,
,
7”
11A
3P
7P
11P
3”
7A
TIME OF DAY
Figure
2.
change in urinary iPGE (expressed as percentage of the mean daily excretion) is shown for consecutive 4-hr periods in 4 subjects, first during normal activity and then during bed rest.
Relative
JULY 1977 VOL. 14 NO. 1
155
PROSTAGLANDINS
TABLE I: AMPLITUDE, PERCENT CHANGE FROM TRE MEAN AND ACROPEASE FOR IPGE, SODIUM, POTASSIUM, CREATININE AND VOLUME IN 6 SUBJECTS WIT'SNORMAL ACTIVITY
Amplitude
% change frommean 29
4:55 P.M.
7.0 meq/4 hr
38
4:47 P.M.
14.8 meq/4 hr
81
3:50 P.M.
39.9 mgf4 hr
20
6:00 P.M.
152.7 ml/4 hr
40
5:49 P.M.
25.0 ng/4 hr
iPGE* Sodium* Potassium* Creatinine* Volume?
Acrophaee*
* denotes p < .Ol for test for circadian rhythm
t denotes p < .05 for test for circadian rhythm P
estimated time of peak excretion
TABLE II: MEAN CORRELATION COEFFICIENTS IN SUBJECTS DURING NORHAL ACTIVITY Correlation coefficient with IPGE
P
Sodium
,363 r .690
NS
Potassium
.706 f ,160
p < .Ol
Creatinine
.S25 t ,127
p < .Ol
Volume
.661 f .276
p < .Ol
while the acrophaee of volume excretion changed from 4:59 P.M. to 1:58 P.M. The mean correlation coefficient, period by period between the 2 days, for the excretion of IPGE for the 4 subjects was r - .690 f .06 (p < .Ol>. The mean correlation coefficients of sodium, potassium, and creatinine were all significant at p < ,025 or better, while that for urine volume was less significant (p < .05). Urinary aldosterone had no significant correlation with any parameter measured for either day, nor did it correlate significantly from one day to the next when each period was compared to the same time interval the next day. There was no significant difference in the total excretion of IPGE, creatinine, sodium, potassium or volume between the 2 days, while aldosterone excretion decreased significantly (Table III). The effects of posture on iPGE excretion for 15 subjects are presented in Fig. 3. There was a significant decrease (p ( ,011 in mean iPGE excretion from 124 t 36 ng/4 hr when the subjects were recumbent, to 84 ?r32 ng/4 hr when the subjects were erect. There were also eignifi-
156
JULY 1977VOL. 14NO. 1
PROSTAGLANDINS
TABLE III: EFFECT OF ACTIVITY IN 4 SUBJECTS ON TliE24 RR EXCRETION OF IPGE, SODIUM, POTASSIUM, CREATININE, VOLUME, AND ALDOSTERONE Normal Activity
p
+ 196
580
+ 138
Sodium (meqfd)
75
f
12
112
t
32
NS
Potassium (meq/d)
71
?:
8
72
+
15
NS
+
72
NS
f 148
NS
iPGE (ng/d)
Creatinine (mg/d) Volume (ml/d) Aldosterone (ug/d)
511
Bed Rest
844
f 132
896
1696
+ 388
1652
11.6?
3.6
6.0 f
2.4
NS
c.02
cant decreases in urinary sodium, potassium, creatinine and volume when the subjects were erect (Table IV). When 2 subjects were studied first recumbent and then erect for consecutive 4 hr periods on the same day (Fig. 4), the excretion of iPGE rose 33% instead of falling 32X, as noted above.
TABLE IV: EFFECT OF POSTURR IN 15 SUBJECTS ON TRE EXCRETION OF IPGE, SODIUM, POTASSIUM, CREATININE, AND VOLUME DURING THE SAME TIME INTERVAL ON TWO DIFFWENT DAYS Recumbent iPGE (ng/4 hr)
Erect
124 i
36
84 i:32
< .OOl
Sodium (meq/4 hr
21 f
10
4r
5
< .OOl
Potassium (meqf4 hr)
152
6
8+
3
< .05
Creatinine (mg/4 hr)
212 f
44
142 f 47
< .OOl
Volume (ml/4 hr)
377 ?:278
116 t 92
< .OOl
JULY 1977 VOL. 14 NO. 1
157
PROSTAGLANDINS
f MEAN + S.D.
I
7A - 11A RECUMBENT
Figure 3.
158
I 7A - 11A ERECT
Urinary iPGE in 15 normal subjects studied while recumbent for 4 hrs one day and then while erect during the same 4 hr interval the next day.
JULY 1977 VOL. 14 NO. 1
PROSTAGLANDINS
A’
/ / ,
w’
cl,
/’
7A- llA
RECUMBENT
Figure 4.
llA
- 3P
ERECT
Urinary iPGE in 2 subjects studied first while recumbent and then while erect for consecutive 4 hr periods on the same day. These 2 subjects are identified with dashed lines on Fig. 3.
JULY 1977 VOL. 14 NO. 1
159
PROSTAGLANDINS
DISCUSSION The data presented in this paper clearly indicate the existence of a circadian rhythm for the excretion of immunoreactive prostaglandin E. The time of peak excretion occura in mid-afternoon to early evening. This circadian rhythm follows the well-described pattern for the urinary excretion of many other substances such as sodium and potassium, as well as renal blood flow and glomerular filtration rate (5,7). The etiology of these rhythms is unknown, but it is possible that the renal synthesis of iPGE either influences or is influenced by the control mechanisms for these other rhythms. In this study of circadian rhythms, there were significant correlations between iPGE excretion and urinary creatinine, sodium, potassium and volume. Howwer, in one subject, there were opposite patterns between sodium and iPGE excretion. When subjects remained recumbent for 24 hr urinary iPGE excretion rose 14% and the time of peak excretion of iPGE was postponed by approximately 90 min, but the basic pattern of excretion was unchanged. Thus the sympathetic nervous system, which is markedly affected by posture would appear to have little influence on the circadian rhythm of urinary iPGE. We have also studied iPGE excretion in a woman with a pheochromocytoma, prior to therapy and after treatment first with the a-adrenergic receptor blocker, phenoxybensamine, and then with the catecholamine synThere was no difference between thesis inhibitor, a-methyl-p-tyroeine. her baseline excretion of iPGE and that of normal women, and therapy did not affect iPGE excretion. Thus it appears that urinary iPGE excretion is not affected by endogenous catecholamines. In addition, the failure of a change in posture to alter the rhythm of iPGE excretion when urinary aldosterone decreased indicated that aldosterone has little influence on the rhythm of urinary IPGE. We investigated one woman with primary Her exaldosteroniem before and after therapy with spironolactone. cretion of iPGE was normal before therapy and did not change with spironolactone even though the hypokalamia and elwated blood pressure were corrected. Prostaglandin E has been implicated in renin release _in vitro (15), in rabbits (161, and in patients with Bartter’e Syndrome (17). Four hours of erect posture, a potent stimulus for renin and aldosterone release, significantly decreases the urinary excretion of iPGE when controlled for the circadian rhythm. This suggests that iPGE excretion and renin release are not always directly related. The marked fall in urinary iPGE excretion and urine volume that occurs when the subject stands suggests that the excretion of prostaglandin may be dependent upon the rate of urine flow. In order to further investigate this, urine was collected and assayed for urinary iPGE and osmolality in 2 subjects who ate a diet constant for sodium and potassium plus 2500 ml of total dietary water for 5 days, followed by 3 days of 600 ml of total dietary water. There was no change in urinary iPGE excretion even though urine volume decreased from 1530 + 40 to 515 f 30 ml/day and urinary osmolality increased from 350 f 20 m osm/L to 1175 2 35 m o&L. Thus it is unlikely that the change in urinary iPGE excretion with posture or with time of day is secondary to the rate of urine flow. In the studies
160
of posture
and the effect
of time of day on
JULY 1977 VOL. 14 NO. 1
PROSTAGLANDINS prostaglandin
excretion, urinary IPGE excretionmoved In parallelwith urine wlum, sodium,potassiumand creatinineexcretion. Rowever,it remainsunhmm whetherprostaglandin E directlycontrolsthese functions. As noted in Figures3 and 4 when subjectsare recumbentand tha erect for consecutive4 hr periods,urinaryiPGE excretionrises as a reeultof the circadianrhythm,whereaswhen subjectsare recumbent and then e!rect during the same 4 hr periodon consecutive days urinary IPGB excretiondecreases. In futureattemptsto elucidate the role of PGE in renal function,it shouldbe borne in mind that the circadian rhythmmay significantly alter the resultsof short-termexperiments.
RRFRRENCES
1.
2. 3. 4. 5. 6.
7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.
Gross, J.B., and P.C. Bartter. Effectsof prostaglandins Ei, AJ, ti I on renal handlingof salt and water. Am. J. Physio . 2 5: 21g. &73. and A. Melforth, Effectsof PGE2 on Pulgraft,G., G. Brandambusch, rennltubulerfunction. Advan. Biochem.Sci. 9:301,1973. Grloff,J., J.S. Randier,and S. Bergstrom. Effectof prostaglandln ) on the permeability responseof toad bladderto vasopressin, t%z& lline and edenosine3',5'-monophosphate. Nature 205:397,1965. Lomigro,A.J., H.D. Itskovitz,K. Crowshaw, and J.C. McGiff. Depemdeneyof renal blood flow on prostaglandin synthesisin the dog. Circ. Kes. 32:712,1973. Ussson,L.G. Electrolyteexcretionin relationto diurnalcycles of renal function. Medicine43:547,1964. Doa, P.P., J.A. Vennes,and E.B. Plink. Diurnalvariationof 17hydroxycortico-steroids, sodium,potassium,magnesiumand creatininsin normal subjectsand in cases of treatedadrenalinsufficiemcyand Gushing'ssyndrome. J. Clin. Rndocrin.20:253,1960. Uesson,L.G. and D.P. Laular. Diurnalcycle of glomerularflltration rate and sodiumand chlorideexcretionduringresponsesto altered salt and water balancefn man. J. Clin. Invest.40:1967,1977. Hills, J.I. Human circadianrhythms. Physiol.Rev. 46:128,1966. Slrota,J.B., D.S. Baldwin,and B. Villarreal. Diurnalvariations of remal functionin man. J. Clin. Invest.29:1S7,1950. Vagmucci,A.H.. A.P. Shapiro,and R.R. McDonald. Effectof upright postureon renal electrokytecycle. J. Appl. Physiol.26:720,1969. Eelushke,P.V., R.W. Alexander,L.J. Loeffler,H.R. Reiser,and J.J. Pisaao. Radioimmunaassay of urinaryprostaglandln E. Submltted for publication. &kg, P., Y.L. Tong, and E.A. Johnson. Circadiansystemphase amaspectoftemporalmorhpology. & CellularAspectsof Biorhythms. (El. vom Mayersback,Editor)Springer-Verlag, 1967, pp 20-48. Tong, Y.L. Parameterestimationin studyingcircadianrhythms. Biometrics32:85,1976. Fort, A., and J.N. Mills. Fittingsine curves to 24 hour urinary data. Eature 276~657,1970. Dew, XB., A.M. nichelekis, Effectof prostaglandins on renin relesse_In vitro. The Pharmacologist 16:198,1974. Larsom,C., P. Weber, and A. Anggard. Arachiodonic acid increases amd lndomethacin decreasesplasmarenin activityin the rabbit. M. J. Pharmacol. 28:391,1974. Gill, P.R., J.C. Frolich,R.E. Borden,A.A. Taylor,H.B. Reiser, H.W. Seyberth,J.A. Ontes,and F.C. Bartter. Bartter'ssyndrome: a disordercharacterized by hlgh urinaryprostagleodins and s dependencyof hypereninemia on prostaglandin synthesis. Am. J. Ued. 61:43,1976.
ami-
4/l/77 - Approved 6/25/77
JULY 1977 VOL. 14 NO. 1
161
URINARY EXCRETION OF IMMUNORFACTIVE PROSTAGLANDIN E: AND THE EFFECT OF POSTURE
A CIRCADIAN RHYTHM
Robert E. Bowden,* James H. Ware, David L. DeMets, and Harry R. Keiser Hypertension and Endocrine Branch, and the Biometrics Research Branch National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland 20014
ABSTRACT The excretion of urinary innnunoreactive prostaglandin E (IPGE), sodium, potassium, creatinine and volume was studied in 4 hr collections in normal women at normal activity. iPGE exhibited a circadian rhythm with an amplitude of 29% and peak excretion at 4:55 P.M. There were also significant circadian rhythms for sodium, potassium, creatinine, and volume, all peaking in late afternoon. There were no significant changes either in the total excretion or in the circadian rhythms of IPGE, potassium, or creatinine excretion when the subjects remained in bed for an entire day while the circadian rhythms of sodium and volume were significantlymodified in amplitude and phase, respectively. Urinary aldosterone excretion decreased significantly when the subjects were at bed rest. iPGE excretion increased 33% when subjects were first recumbent and then erect for consecutive 4 hr periods on the same day (but when subjects were erect 1 day for a 4 hr period, iPGE excretion was lower by 32% than for the same 4 hr period the preceding day when they were recumbent). These data indicate that: 1) the sympathetic nervous system and renin-angiotensin-aldosteronesystem do not affect the circadian rhythm of urinary IPGE, and 2) short-term experiments of prostaglandin E excretion must be designed to avoid misleading results due to the circadian rhythm.
ACKNOWLEDGEMENT The authors are grateful to Dr. Frederic C. Bartter for his critical review of this manuscript and to Mrs. Marian Warner for secretarial help.
*Present address: Division of Cardiology, Stanford Medical School, Stanford, California 93405
JULY 1977 VOL. 14 NO. 1
151
PROSTAGLANDINS
INTRODUCTION Prostaglandins of the E series are vasodilators, produce natriuresis and diuresis when infused into the renal artery (1,2), and antagonize the action of antidiuretic hormone on toad bladder (3). McGiff and his associates have proposed that prostaglandins are involved in autoregulation of renal blood flow (4). Renal blood flow, as well as glomerular filtration rate and the excretion of sodium, potassium, and creatinine exhibit 24 hr periodicity, i.e., circadian rhythms (5-10). Little is known about the etiology of these rhythms. Prostaglandins may be involved in regulating the glomerular filtration rate which in turn could determine the circadian rhythms of sodium and potassium. Me undertook this study to determine if prostaglandin excretion exhibits a circadian rhytbm, and to determine if posture, which has great influence on renal blood flow and renal function, has an effect on prostaglandin E excretion. METHODS Two separate studies were performed. In the first, the influence of the time of day on immunoreactive prostaglandin E excretion was observed. The subjects were 6 normal white women, aged 19-22 years, who ate a diet containing, each day, 109 meq sodium and 100 meq potassium. All subjects ate the diet for at least 3 days before the start of any study, and were confined to the Clinical Center during study. Bach subject was allowed normal activity, and urine was collected by spontaneous voiding at 4 hr intervals for 24 hr. Subjects 3 - 6 were kept in bed for a second day of urine collections, except when they were allowed out of bed to void. In the second study, the influence of posture on immunoreactive prostaglandin E excretion was observed. The subjects were 15 normal women age 19-22 years (14 white, 1 black), who ate a 109 meq sodium and 100 meq potassium diet. The subjects remained recumbent for 4 hr 1 day and remained standing during the same 4 hr interval the next day. The subjects voided at the end of each 4 hr period. To test if circadian variations would modify the effects of posture on prostaglandin excretion, 2 subjects were also studied while they were first recumbent and then erect for consecutive 4 hr periods on the same day. Urine samples were refrigerated until extracted. Urinary sodium and potassium were measured by flame photometry and urinary creatinine was measured by the Jaffe chromogen reaction with picric acid. Urinary aldosterone was determined in subjects 3 - 6 by radioimmunoassay. Immunoreactive prostaglandin E-like material (IPGE) was measured in triplicate aliquots of urine by radioimmunoassayas described by Halushka et al (11). Briefly, 2000 dpm of 3HPGEl (NBN) in 0.02 ml of ethanol was -added to 10 ml of urine. Neutral lipids were extracted with hexane, and prostaglandins extracted with chloroform at pH 3.5. The samples were partially purified by chromatographywith a 13-cm by l-cm (length and diameter) sephadex LPI20column. The PGE fraction was eluted with a mixture of chloroform, heptane, ethanol and acetic acid (100:100:25:2),converted to PGB with 0.1 N methanolic RCH, and then appl -h to a 0.5 g
152
JULY 1977 VOL. 14 NO. 1
PROSTAGLANDINS
silicic acid column. The PGB fraction was eluted with a mixture of benzene and ethyl acetate (60:40) and the collected fraction was radioimmunoassayed with an antibody to PGB (1:75,000 dilution with a PGB standard). The PGB concentration was corrected for recovery and alz assays were performed without knowledge of which time period the sample represented. The presence of a circadian rhythm was determined by the method described by Ralberg -et al (12) with the statistical techniques of Tong (13). This method, known as the "Cosinor" method, involves fitting a sinusoidal curve to the responses of each individual and averaging the estimated amplitudes and phases of individuals. Because our data represented total output in a 4 hr period, the technique was modified slightly to fit the integral of the sine function, as suggested by Fort and Mills (14). Paired analyses were used where possible and a p value of 0.05 or less was considered significant. RESULTS The average excretions of urinary IPGE, sodium, potassium, creatinine and volume in 4 hr periods for the 6 normal subjects during normal activity are shown as percentages of the means for all 6 4 hr periods (Fig. 1). In all subjects urinary iPGE displayed a rhythmic fluctuation with a peak between 11 A.M. and 11 P.M. and a nadir between 3 A.M. and 7 A.M. There were similar rhythms for potassium and creatinine excretion and urine volume. Sodium excretion followed a similar pattern except in 1 subject in whom the lowest sodium excretion occurred between 3 P.M. and 7 P.M. When the integral of a sine function was fitted to the values for the excretion of iPGE as described in Methods, the average amplitude for iPGE excretion for the 6 normal subjects was 25 ng/4 hr (29% change from the mean). The mean acrophase for iPGE excretion, i.e., the time of peak excretion as estimated from the fitted sinusoidal curves, was 4~55 P.M. f 58 min (mean f S.D.). There was a statistically significant circadian rhythm for urinary iPGE excretion (p < .Ol). Similar analyses of the data for percent change from the mean amplitude, phase and time of peak excretion indicated that there were significant circadian rhythms for urinary potassium, sodium, and creatinine (p < .Ol> and volume (p < .05) (Table I). The mean acrophases are all similar and indicate that the excretion of all these substances peaked at close to the same time. iPGE was significantly correlated with urine volume, potassium and creatinine excretion, but not with sodium excretion (Table II). Figure 2 shows the pattern of iPGE excretion for the 4 normal subjects 3 - 6 who were studied first with normal activity and then with bed rest. There was a minor modification of the rhythm of iPGE excretion when the subjects were recumbent manifested by a shift in mean acrophase from 4~17 P.M. during normal activity to 5:47 P.M. when the subjects were at bed rest. This change was not statistically significant. There were also no significant changes in the rhythms of creatinine and potassium excretion between the 2 days. Sodium excretion and urinary volume were significantly different between the 2 days (p < .05); the amplitude of sodium excretion changed from 13 meq to 23 meq without a change in phase,
JULY 1977 VOL. 14 NO. 1
153
PROSTAGLANDINS
w
I++% ______
PERCENT OF MEAN 150 r RELATIVE CHANGE _
61 .ClM
---
z
9
60
y4 __- __ __ --___ z&-I- 100 =,9 Y 850160I++? 200 us
4: 13
--
-
__
--_ -
li)lJ 0 k++i
MEAN6.D. I
1
1
I
I
I
,
TA 3p 11p 7A llA 7P 3A TIME OF DAY
Figure
154
1.
Relative changes in urinary IPGE, creatinine, sodium, potassium, and volume (expressed as percentages of the mean daily excretion) are shown for consecutive 4-hr periods in 6 subjects during normal activity.
JULY 1977 VOL. 14 NO. 1
PROSTAGLANDINS
160 ----AD
LIB ACTIVITY -BED
RESTI-1
145
P
130
z:
115
8
loo
2 e
85
I 70
55 40
I
IA
11”
3P
7p
11”
3”
1
,
,
7”
11A
3P
7P
11P
3”
7A
TIME OF DAY
Figure
2.
change in urinary iPGE (expressed as percentage of the mean daily excretion) is shown for consecutive 4-hr periods in 4 subjects, first during normal activity and then during bed rest.
Relative
JULY 1977 VOL. 14 NO. 1
155
PROSTAGLANDINS
TABLE I: AMPLITUDE, PERCENT CHANGE FROM TRE MEAN AND ACROPEASE FOR IPGE, SODIUM, POTASSIUM, CREATININE AND VOLUME IN 6 SUBJECTS WIT'SNORMAL ACTIVITY
Amplitude
% change frommean 29
4:55 P.M.
7.0 meq/4 hr
38
4:47 P.M.
14.8 meq/4 hr
81
3:50 P.M.
39.9 mgf4 hr
20
6:00 P.M.
152.7 ml/4 hr
40
5:49 P.M.
25.0 ng/4 hr
iPGE* Sodium* Potassium* Creatinine* Volume?
Acrophaee*
* denotes p < .Ol for test for circadian rhythm
t denotes p < .05 for test for circadian rhythm P
estimated time of peak excretion
TABLE II: MEAN CORRELATION COEFFICIENTS IN SUBJECTS DURING NORHAL ACTIVITY Correlation coefficient with IPGE
P
Sodium
,363 r .690
NS
Potassium
.706 f ,160
p < .Ol
Creatinine
.S25 t ,127
p < .Ol
Volume
.661 f .276
p < .Ol
while the acrophaee of volume excretion changed from 4:59 P.M. to 1:58 P.M. The mean correlation coefficient, period by period between the 2 days, for the excretion of IPGE for the 4 subjects was r - .690 f .06 (p < .Ol>. The mean correlation coefficients of sodium, potassium, and creatinine were all significant at p < ,025 or better, while that for urine volume was less significant (p < .05). Urinary aldosterone had no significant correlation with any parameter measured for either day, nor did it correlate significantly from one day to the next when each period was compared to the same time interval the next day. There was no significant difference in the total excretion of IPGE, creatinine, sodium, potassium or volume between the 2 days, while aldosterone excretion decreased significantly (Table III). The effects of posture on iPGE excretion for 15 subjects are presented in Fig. 3. There was a significant decrease (p ( ,011 in mean iPGE excretion from 124 t 36 ng/4 hr when the subjects were recumbent, to 84 ?r32 ng/4 hr when the subjects were erect. There were also eignifi-
156
JULY 1977VOL. 14NO. 1
PROSTAGLANDINS
TABLE III: EFFECT OF ACTIVITY IN 4 SUBJECTS ON TliE24 RR EXCRETION OF IPGE, SODIUM, POTASSIUM, CREATININE, VOLUME, AND ALDOSTERONE Normal Activity
p
+ 196
580
+ 138
Sodium (meqfd)
75
f
12
112
t
32
NS
Potassium (meq/d)
71
?:
8
72
+
15
NS
+
72
NS
f 148
NS
iPGE (ng/d)
Creatinine (mg/d) Volume (ml/d) Aldosterone (ug/d)
511
Bed Rest
844
f 132
896
1696
+ 388
1652
11.6?
3.6
6.0 f
2.4
NS
c.02
cant decreases in urinary sodium, potassium, creatinine and volume when the subjects were erect (Table IV). When 2 subjects were studied first recumbent and then erect for consecutive 4 hr periods on the same day (Fig. 4), the excretion of iPGE rose 33% instead of falling 32X, as noted above.
TABLE IV: EFFECT OF POSTURR IN 15 SUBJECTS ON TRE EXCRETION OF IPGE, SODIUM, POTASSIUM, CREATININE, AND VOLUME DURING THE SAME TIME INTERVAL ON TWO DIFFWENT DAYS Recumbent iPGE (ng/4 hr)
Erect
124 i
36
84 i:32
< .OOl
Sodium (meq/4 hr
21 f
10
4r
5
< .OOl
Potassium (meqf4 hr)
152
6
8+
3
< .05
Creatinine (mg/4 hr)
212 f
44
142 f 47
< .OOl
Volume (ml/4 hr)
377 ?:278
116 t 92
< .OOl
JULY 1977 VOL. 14 NO. 1
157
PROSTAGLANDINS
f MEAN + S.D.
I
7A - 11A RECUMBENT
Figure 3.
158
I 7A - 11A ERECT
Urinary iPGE in 15 normal subjects studied while recumbent for 4 hrs one day and then while erect during the same 4 hr interval the next day.
JULY 1977 VOL. 14 NO. 1
PROSTAGLANDINS
A’
/ / ,
w’
cl,
/’
7A- llA
RECUMBENT
Figure 4.
llA
- 3P
ERECT
Urinary iPGE in 2 subjects studied first while recumbent and then while erect for consecutive 4 hr periods on the same day. These 2 subjects are identified with dashed lines on Fig. 3.
JULY 1977 VOL. 14 NO. 1
159
PROSTAGLANDINS
DISCUSSION The data presented in this paper clearly indicate the existence of a circadian rhythm for the excretion of immunoreactive prostaglandin E. The time of peak excretion occura in mid-afternoon to early evening. This circadian rhythm follows the well-described pattern for the urinary excretion of many other substances such as sodium and potassium, as well as renal blood flow and glomerular filtration rate (5,7). The etiology of these rhythms is unknown, but it is possible that the renal synthesis of iPGE either influences or is influenced by the control mechanisms for these other rhythms. In this study of circadian rhythms, there were significant correlations between iPGE excretion and urinary creatinine, sodium, potassium and volume. Howwer, in one subject, there were opposite patterns between sodium and iPGE excretion. When subjects remained recumbent for 24 hr urinary iPGE excretion rose 14% and the time of peak excretion of iPGE was postponed by approximately 90 min, but the basic pattern of excretion was unchanged. Thus the sympathetic nervous system, which is markedly affected by posture would appear to have little influence on the circadian rhythm of urinary iPGE. We have also studied iPGE excretion in a woman with a pheochromocytoma, prior to therapy and after treatment first with the a-adrenergic receptor blocker, phenoxybensamine, and then with the catecholamine synThere was no difference between thesis inhibitor, a-methyl-p-tyroeine. her baseline excretion of iPGE and that of normal women, and therapy did not affect iPGE excretion. Thus it appears that urinary iPGE excretion is not affected by endogenous catecholamines. In addition, the failure of a change in posture to alter the rhythm of iPGE excretion when urinary aldosterone decreased indicated that aldosterone has little influence on the rhythm of urinary IPGE. We investigated one woman with primary Her exaldosteroniem before and after therapy with spironolactone. cretion of iPGE was normal before therapy and did not change with spironolactone even though the hypokalamia and elwated blood pressure were corrected. Prostaglandin E has been implicated in renin release _in vitro (15), in rabbits (161, and in patients with Bartter’e Syndrome (17). Four hours of erect posture, a potent stimulus for renin and aldosterone release, significantly decreases the urinary excretion of iPGE when controlled for the circadian rhythm. This suggests that iPGE excretion and renin release are not always directly related. The marked fall in urinary iPGE excretion and urine volume that occurs when the subject stands suggests that the excretion of prostaglandin may be dependent upon the rate of urine flow. In order to further investigate this, urine was collected and assayed for urinary iPGE and osmolality in 2 subjects who ate a diet constant for sodium and potassium plus 2500 ml of total dietary water for 5 days, followed by 3 days of 600 ml of total dietary water. There was no change in urinary iPGE excretion even though urine volume decreased from 1530 + 40 to 515 f 30 ml/day and urinary osmolality increased from 350 f 20 m osm/L to 1175 2 35 m o&L. Thus it is unlikely that the change in urinary iPGE excretion with posture or with time of day is secondary to the rate of urine flow. In the studies
160
of posture
and the effect
of time of day on
JULY 1977 VOL. 14 NO. 1
PROSTAGLANDINS prostaglandin
excretion, urinary IPGE excretionmoved In parallelwith urine wlum, sodium,potassiumand creatinineexcretion. Rowever,it remainsunhmm whetherprostaglandin E directlycontrolsthese functions. As noted in Figures3 and 4 when subjectsare recumbentand tha erect for consecutive4 hr periods,urinaryiPGE excretionrises as a reeultof the circadianrhythm,whereaswhen subjectsare recumbent and then e!rect during the same 4 hr periodon consecutive days urinary IPGB excretiondecreases. In futureattemptsto elucidate the role of PGE in renal function,it shouldbe borne in mind that the circadian rhythmmay significantly alter the resultsof short-termexperiments.
RRFRRENCES
1.
2. 3. 4. 5. 6.
7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.
Gross, J.B., and P.C. Bartter. Effectsof prostaglandins Ei, AJ, ti I on renal handlingof salt and water. Am. J. Physio . 2 5: 21g. &73. and A. Melforth, Effectsof PGE2 on Pulgraft,G., G. Brandambusch, rennltubulerfunction. Advan. Biochem.Sci. 9:301,1973. Grloff,J., J.S. Randier,and S. Bergstrom. Effectof prostaglandln ) on the permeability responseof toad bladderto vasopressin, t%z& lline and edenosine3',5'-monophosphate. Nature 205:397,1965. Lomigro,A.J., H.D. Itskovitz,K. Crowshaw, and J.C. McGiff. Depemdeneyof renal blood flow on prostaglandin synthesisin the dog. Circ. Kes. 32:712,1973. Ussson,L.G. Electrolyteexcretionin relationto diurnalcycles of renal function. Medicine43:547,1964. Doa, P.P., J.A. Vennes,and E.B. Plink. Diurnalvariationof 17hydroxycortico-steroids, sodium,potassium,magnesiumand creatininsin normal subjectsand in cases of treatedadrenalinsufficiemcyand Gushing'ssyndrome. J. Clin. Rndocrin.20:253,1960. Uesson,L.G. and D.P. Laular. Diurnalcycle of glomerularflltration rate and sodiumand chlorideexcretionduringresponsesto altered salt and water balancefn man. J. Clin. Invest.40:1967,1977. Hills, J.I. Human circadianrhythms. Physiol.Rev. 46:128,1966. Slrota,J.B., D.S. Baldwin,and B. Villarreal. Diurnalvariations of remal functionin man. J. Clin. Invest.29:1S7,1950. Vagmucci,A.H.. A.P. Shapiro,and R.R. McDonald. Effectof upright postureon renal electrokytecycle. J. Appl. Physiol.26:720,1969. Eelushke,P.V., R.W. Alexander,L.J. Loeffler,H.R. Reiser,and J.J. Pisaao. Radioimmunaassay of urinaryprostaglandln E. Submltted for publication. &kg, P., Y.L. Tong, and E.A. Johnson. Circadiansystemphase amaspectoftemporalmorhpology. & CellularAspectsof Biorhythms. (El. vom Mayersback,Editor)Springer-Verlag, 1967, pp 20-48. Tong, Y.L. Parameterestimationin studyingcircadianrhythms. Biometrics32:85,1976. Fort, A., and J.N. Mills. Fittingsine curves to 24 hour urinary data. Eature 276~657,1970. Dew, XB., A.M. nichelekis, Effectof prostaglandins on renin relesse_In vitro. The Pharmacologist 16:198,1974. Larsom,C., P. Weber, and A. Anggard. Arachiodonic acid increases amd lndomethacin decreasesplasmarenin activityin the rabbit. M. J. Pharmacol. 28:391,1974. Gill, P.R., J.C. Frolich,R.E. Borden,A.A. Taylor,H.B. Reiser, H.W. Seyberth,J.A. Ontes,and F.C. Bartter. Bartter'ssyndrome: a disordercharacterized by hlgh urinaryprostagleodins and s dependencyof hypereninemia on prostaglandin synthesis. Am. J. Ued. 61:43,1976.
ami-
4/l/77 - Approved 6/25/77
JULY 1977 VOL. 14 NO. 1
161
