Clinical Science ( 1990) 79,117-1 2 1
117
Active and inactive urinary kallikrein in man: effects of diuresis and antidiuresis DEREK G. WALLER*, SATVINDER S. BHATIA, SARA K. CAMPBELL, JANET D. M. ALBANO AND J. GAVIN €3. MILLAR *Clinical PharmacologyGroup, University of Southampton Medical School, Southampton General Hospital, Southampton, and Department of Renal and Endocrine Medicine, University of Southampton Medical School, St Mary's Hospital, Portsmouth, U.K.
(Received 24 July 1989/11 January 1990;accepted 16 March 1990)
SUMMARY 1. The urinary excretion of active and inactive kallikrein was studied in volunteers during diuresis induced by water loading or oral frusemide and during antidiuresisinduced by desamino-Darginine-vasopressin. 2. During acute oral water loading, excretion of active kallikrein was unchanged, despite high urine flow rates and low urine osmolalities being achieved. Excretion of inactive kallikrein correlated with the urine flow rate. 3. After desamino-D-arginine-vasopressin in eight water-loaded and six normally hydrated subjects, excretion of inactive kallikrein also correlated with the urine flow rate. There were no significant changes in the excretion of active kallikrein. 4. After frusemide there was a small transient increase in excretion of active kallikrein 1-2 h after dosing which coincided with the maximum diuresis and natriuresis. Excretion of inactive kallikrein again correlated with urine flow rate but the regression relationshipbetween the two variables was different for water-load-induced and frusemide-induceddiuresis. 5. These studies do not support a role for urinary kallikrein in the modulation of the antidiuretic action of vasopressin, but suggest that it may contribute to the natriuretic action of frusemide. Key words: desamho-D-arginine-vasopressh, kallikrein, urine. Abbreviations: AVP, arginine vasopressin, AUK, active urinary kallikrein, DDAVP, desamino-Dargininevasopressin,IUK, inactive urinary kallikrein.
uncertain. Water diuresis in man has been accompanied by an increase in urinary excretion of kallikrein-like activity in some studies [ 1-31, whereas other investigators have failed to demonstrate a relationship [4-61. Animal studies have indicated a direct relationship between reductions in urine osmolality and uMary kallikrein excretion that is independent of sodium intake [7], whereas other data from the dog and rat suggest that arginine vasopressin (AVP) stimulates urinary kallikrein excretion [81. The consequent generation of kinins may act to antagonize the action of AVP [9]. There is little information about the regulation of urinary kallikrein excretion by AVP in man. Conflicting evidence exists, therefore, that in animals both diuresis after water loading and antidiuresis induced by AVP may stimulate the urinary excretion of urinary kallikrein. Diuresis induced by the loop diuretic frusemide has also been associated with increased urinary excretion of kallikrein [lo, 111. In this situation, however, there are additional profound changesin urine electrolyteexcretion. The majority of urinary kallikrein is present in the inactive form (IUK) [5, 121, but the relationship between IUK and active kallikrein (AUK) has received relatively little attention. We have therefore measured urinary excretion of IUK and AUK during the diuresis accompanying water loading and during antidiuresisinduced by a non-pressor AVP analogue, desamino-margininevasopressin (DDAVP).To determine whether the rate of concurrent electrolyte excretion influences the kallikrein response to diuresis, a comparison between diuresis induced by water loading and by frusemide was also made.
INTRODUCTION
METHODS
The role of the renal kallikrein-kinin system in the regulation of water excretion by the kidney remains
Subjects
Correspondence: Dr D. G. Waller, Clinical Pharmacology Group, Level F,Centre Block, Southampton General Hospital, SouthamptonSO9 4XY, U.K.
Healthy drug-free volunteers aged between 20 and 43 years participated in the studies which were approved by the local ethical sub-committees. All women were in the first half of the menstrual cycle. Four studies were carried
118
D. G. Waller et al.
out, without prior dietary sodium restrictions. A light diet, similar on each study day, was allowed but fluid intake was restricted to that dictated by the protocol. Before each study, two 12 h urine collections were made from 08.00 to 20.00 hours and from 20.00 to 08.00 hours on the morning of the study to establish baseline measurements. We have found a close correlation between excretion of AUK over 4 h and 12 h during unrestricted moderate fluid intake (D. G. Waller ef al., unpublished work). Study 1: acute water loading
Six subjects (three male and three female)each drank 1 litre of water in the first hour, beginning at 08.00 hours. During the subsequent 6 h, they each drank a volume of water equal to the volume of urine passed during the preceding hour. Subsequently, water intake was not standardized. Urine was collected from 08.00 to 12.00, 12.00 to 13.00, 13.00 to 15.00, 15.00 to 20.00 and 20.00 to 08.00 hours. Study 2: antidiuresis after acute water loading
Eight subjects (five male and three female) were water loaded for 4 h beginning at 08.00 hours as for study 1.At 12.00 hours each subject received an intravenous bolus injection of 2 p g of DDAVP. Urine collections were made as for study 1. Study 3: antidiuresis without water loading
Six subjects (four male and two female) were studied as for study 2 but without water loading before DDAVP. Study 4: a comparison of frusemide-induced and waterload-induced diuresis
This study was performed on six male subjects. Before taking frusemide, baseline collections were made for 3 days from 08.00 to 20.00 and 20.00 to 08.00 hours. On the fourth day at 08.00 hours, 40 mg of frusemide was taken orally and urine collections were made at 1 , 2 , 3 , 4 , 5,6,9,12 and 24 h after frusemide. Two baseline urine collections were made from 08.00 to 20.00 and 20.00 to 08.00 hours on the day before water loading. On day 2 fluid intake was increased to 3.5 litres for the first 7 h. Serial urine collections were made throughout the day at the following intervals: 08.00-12.00, 12.00-1 3.00, 13.00-1 5.00, 15.00-1 7.00, 17.00-20.00 and 20.00-08.00 hours. Urine samples for kallikrein assay were kept at 4°C and analysed within 5 days of collection. Kallikrein activity has been shown to be stable over this period [13, 141. Kallikrein-like amidolytic activity was measured using chromogenic substrate S2266 (Kabi Diagnostica, Stockholm, Sweden) [ 151 before and after trypsin activation to obtain active and total kallikrein measurements [16]. IUK represents amidolytic activity after addition of trypsin less the activity before trypsin activation. The assay was standardized using a purified freeze-dried preparation of human urinary kallikrein (Channel
Diagnostics, Walmer, Kent, U.K.). The intra-assay coefficient of variation for this method was 2.3% and the interassay coefficient of variation was 3.1%. Urine osmolality was measured by depression of freezing point, and urine sodium and potassium were determined by flame photometry. Results are expressed as m e a n s f s ~or the median and interquartile range. Two-way analysis of variance and least square linear regression were used in the analysis of results.
RESULTS There were no sigdicant differences between the excretion of AUK and IUK in the three daily collections before frusemide and no diurnal variation was apparent in the 12 h collections.The mean values were therefore used as the baseline for the subsequent study. In all four studies, excretion of IUK showed a close relationship to urine flow rate (Figs. 1-3). Correlations between urine flow rate and excretion of IUK were calculated for each individual study; for the first three studies, the median correlation was 0.958 (interquartile range 0.744-0.981). There was a small non-significant increase in excretion of AUK during the first 4 h of water loading in studies 1and 2. Subsequently, excretion of AUK was not significantly changed from the baseline excretion rate despite large alterations in urine flow rate and osmolality (Table l), particularly in studies 1 and 2 (Table 1). During acute water loading, the urine osmolality fell below the normal range for plasma osmolality for up to 12 h. In contrast, substantial rises in urine osmolality above that of plasma occurred after DDAVP. After frusemide, there was a marked increase in urine flow rate.from 1.1st 0.3 to a maximum of 11.6 f 3.7 ml/ min in the second hour after dosing, which had returned
Time of day (hours)
Fig. 1. Rate of excretion of AUK (0)and IUK (0)and urine flow rate (C- *) after oral water loading. Results are means with bars indicating SD.
Urinary kallikrein and water excretion almost to baseline values by 3 h. Sodium and potassium excretion rose from 7.8 f 1.4 to 80.0 f 40.9 mmol/h and 3.2f0.6 to 13.2f6.5 mmol/h, respectively. The peak excretion rates coincided with the peak urine flow rate. The excretion of AUK showed a small but significant rise of 35% (P
117
Active and inactive urinary kallikrein in man: effects of diuresis and antidiuresis DEREK G. WALLER*, SATVINDER S. BHATIA, SARA K. CAMPBELL, JANET D. M. ALBANO AND J. GAVIN €3. MILLAR *Clinical PharmacologyGroup, University of Southampton Medical School, Southampton General Hospital, Southampton, and Department of Renal and Endocrine Medicine, University of Southampton Medical School, St Mary's Hospital, Portsmouth, U.K.
(Received 24 July 1989/11 January 1990;accepted 16 March 1990)
SUMMARY 1. The urinary excretion of active and inactive kallikrein was studied in volunteers during diuresis induced by water loading or oral frusemide and during antidiuresisinduced by desamino-Darginine-vasopressin. 2. During acute oral water loading, excretion of active kallikrein was unchanged, despite high urine flow rates and low urine osmolalities being achieved. Excretion of inactive kallikrein correlated with the urine flow rate. 3. After desamino-D-arginine-vasopressin in eight water-loaded and six normally hydrated subjects, excretion of inactive kallikrein also correlated with the urine flow rate. There were no significant changes in the excretion of active kallikrein. 4. After frusemide there was a small transient increase in excretion of active kallikrein 1-2 h after dosing which coincided with the maximum diuresis and natriuresis. Excretion of inactive kallikrein again correlated with urine flow rate but the regression relationshipbetween the two variables was different for water-load-induced and frusemide-induceddiuresis. 5. These studies do not support a role for urinary kallikrein in the modulation of the antidiuretic action of vasopressin, but suggest that it may contribute to the natriuretic action of frusemide. Key words: desamho-D-arginine-vasopressh, kallikrein, urine. Abbreviations: AVP, arginine vasopressin, AUK, active urinary kallikrein, DDAVP, desamino-Dargininevasopressin,IUK, inactive urinary kallikrein.
uncertain. Water diuresis in man has been accompanied by an increase in urinary excretion of kallikrein-like activity in some studies [ 1-31, whereas other investigators have failed to demonstrate a relationship [4-61. Animal studies have indicated a direct relationship between reductions in urine osmolality and uMary kallikrein excretion that is independent of sodium intake [7], whereas other data from the dog and rat suggest that arginine vasopressin (AVP) stimulates urinary kallikrein excretion [81. The consequent generation of kinins may act to antagonize the action of AVP [9]. There is little information about the regulation of urinary kallikrein excretion by AVP in man. Conflicting evidence exists, therefore, that in animals both diuresis after water loading and antidiuresis induced by AVP may stimulate the urinary excretion of urinary kallikrein. Diuresis induced by the loop diuretic frusemide has also been associated with increased urinary excretion of kallikrein [lo, 111. In this situation, however, there are additional profound changesin urine electrolyteexcretion. The majority of urinary kallikrein is present in the inactive form (IUK) [5, 121, but the relationship between IUK and active kallikrein (AUK) has received relatively little attention. We have therefore measured urinary excretion of IUK and AUK during the diuresis accompanying water loading and during antidiuresisinduced by a non-pressor AVP analogue, desamino-margininevasopressin (DDAVP).To determine whether the rate of concurrent electrolyte excretion influences the kallikrein response to diuresis, a comparison between diuresis induced by water loading and by frusemide was also made.
INTRODUCTION
METHODS
The role of the renal kallikrein-kinin system in the regulation of water excretion by the kidney remains
Subjects
Correspondence: Dr D. G. Waller, Clinical Pharmacology Group, Level F,Centre Block, Southampton General Hospital, SouthamptonSO9 4XY, U.K.
Healthy drug-free volunteers aged between 20 and 43 years participated in the studies which were approved by the local ethical sub-committees. All women were in the first half of the menstrual cycle. Four studies were carried
118
D. G. Waller et al.
out, without prior dietary sodium restrictions. A light diet, similar on each study day, was allowed but fluid intake was restricted to that dictated by the protocol. Before each study, two 12 h urine collections were made from 08.00 to 20.00 hours and from 20.00 to 08.00 hours on the morning of the study to establish baseline measurements. We have found a close correlation between excretion of AUK over 4 h and 12 h during unrestricted moderate fluid intake (D. G. Waller ef al., unpublished work). Study 1: acute water loading
Six subjects (three male and three female)each drank 1 litre of water in the first hour, beginning at 08.00 hours. During the subsequent 6 h, they each drank a volume of water equal to the volume of urine passed during the preceding hour. Subsequently, water intake was not standardized. Urine was collected from 08.00 to 12.00, 12.00 to 13.00, 13.00 to 15.00, 15.00 to 20.00 and 20.00 to 08.00 hours. Study 2: antidiuresis after acute water loading
Eight subjects (five male and three female) were water loaded for 4 h beginning at 08.00 hours as for study 1.At 12.00 hours each subject received an intravenous bolus injection of 2 p g of DDAVP. Urine collections were made as for study 1. Study 3: antidiuresis without water loading
Six subjects (four male and two female) were studied as for study 2 but without water loading before DDAVP. Study 4: a comparison of frusemide-induced and waterload-induced diuresis
This study was performed on six male subjects. Before taking frusemide, baseline collections were made for 3 days from 08.00 to 20.00 and 20.00 to 08.00 hours. On the fourth day at 08.00 hours, 40 mg of frusemide was taken orally and urine collections were made at 1 , 2 , 3 , 4 , 5,6,9,12 and 24 h after frusemide. Two baseline urine collections were made from 08.00 to 20.00 and 20.00 to 08.00 hours on the day before water loading. On day 2 fluid intake was increased to 3.5 litres for the first 7 h. Serial urine collections were made throughout the day at the following intervals: 08.00-12.00, 12.00-1 3.00, 13.00-1 5.00, 15.00-1 7.00, 17.00-20.00 and 20.00-08.00 hours. Urine samples for kallikrein assay were kept at 4°C and analysed within 5 days of collection. Kallikrein activity has been shown to be stable over this period [13, 141. Kallikrein-like amidolytic activity was measured using chromogenic substrate S2266 (Kabi Diagnostica, Stockholm, Sweden) [ 151 before and after trypsin activation to obtain active and total kallikrein measurements [16]. IUK represents amidolytic activity after addition of trypsin less the activity before trypsin activation. The assay was standardized using a purified freeze-dried preparation of human urinary kallikrein (Channel
Diagnostics, Walmer, Kent, U.K.). The intra-assay coefficient of variation for this method was 2.3% and the interassay coefficient of variation was 3.1%. Urine osmolality was measured by depression of freezing point, and urine sodium and potassium were determined by flame photometry. Results are expressed as m e a n s f s ~or the median and interquartile range. Two-way analysis of variance and least square linear regression were used in the analysis of results.
RESULTS There were no sigdicant differences between the excretion of AUK and IUK in the three daily collections before frusemide and no diurnal variation was apparent in the 12 h collections.The mean values were therefore used as the baseline for the subsequent study. In all four studies, excretion of IUK showed a close relationship to urine flow rate (Figs. 1-3). Correlations between urine flow rate and excretion of IUK were calculated for each individual study; for the first three studies, the median correlation was 0.958 (interquartile range 0.744-0.981). There was a small non-significant increase in excretion of AUK during the first 4 h of water loading in studies 1and 2. Subsequently, excretion of AUK was not significantly changed from the baseline excretion rate despite large alterations in urine flow rate and osmolality (Table l), particularly in studies 1 and 2 (Table 1). During acute water loading, the urine osmolality fell below the normal range for plasma osmolality for up to 12 h. In contrast, substantial rises in urine osmolality above that of plasma occurred after DDAVP. After frusemide, there was a marked increase in urine flow rate.from 1.1st 0.3 to a maximum of 11.6 f 3.7 ml/ min in the second hour after dosing, which had returned
Time of day (hours)
Fig. 1. Rate of excretion of AUK (0)and IUK (0)and urine flow rate (C- *) after oral water loading. Results are means with bars indicating SD.
Urinary kallikrein and water excretion almost to baseline values by 3 h. Sodium and potassium excretion rose from 7.8 f 1.4 to 80.0 f 40.9 mmol/h and 3.2f0.6 to 13.2f6.5 mmol/h, respectively. The peak excretion rates coincided with the peak urine flow rate. The excretion of AUK showed a small but significant rise of 35% (P
