Clinical Science (1994) 86, 723-730 (Printed in Great Britain)
Effects of physiological increments in human a-atrial
natriuretic peptide and human brain natriuretic peptide
in normal male subjects
B. M. Y. CHEUNG, J. E. C. DICKERSON, M. J. ASHBY, M. J. BROWN and 1. BROWN*
Clinical Pharmacology Unit, Department of Medicine. Addenbrooke's Hospital, Cambridge, U.K.,
and *Physiological laboratory, University of Cambridge, Cambridge, U.K.
(Received 19 August/l3 December 1993; accepted 7 January 1994)
~ ~ ~~ ~~
1. Brain natriuretic peptide, closely related to atrial
natriuretic peptide in structure, may be an important
circulating hormone. Its physiological role is unclear.
First, we studied the effects of incremental infusions
of brain natriuretic peptide in six healthy men on
plasma brain natriuretic peptide levels and the phar-
macokinetics of brain natriuretic peptide. Synthetic
human brain natriuretic peptide-32 was infused intra-
venously, at an initial rate of 0.4 pmol min-' kg-',
doubling every 15 min until the dose rate reached
6.4 pmol min-' kg-', at which rate the infusion was
maintained for 30 min.
2. The brain natriuretic peptide infusion raised the
brain natriuretic peptide-like immunoreactivity from
1.4 f 0.5 pmsl/l to 21.4 f 7.6 pmol/l. Brain natriuretic
peptide-like immunoreactivity after the end of infu-
sion was consistent with a bi-exponential decay, with
half-lives of 2.1 min and 37 min.
3. Next, we studied the effects of low-dose infusion of
brain natriuretic peptide to mimic physiological
increments in the circulating levels in comparison
with atrial natriuretic peptide. Six dehydrated male
subjects received intravenous infusions of atrial
natriuretic peptide and brain natriuretic peptide,
separately and in combination, in a randomized
double-blind, placebo-controlled, four-part cross-over
design. Atrial natriuretic peptide and brain natriuretic
peptide were given at the rate of 0.75 and
0.4 pmol min-' kg-', respectively, for 3 h. The
control infusion consisted of the vehicle.
4. Analysis of variance showed that atrial natriuretic
peptide and atrial natriuretic peptide plus brain
natriuretic peptide, but not brain natriuretic peptide
alone, increased urinary flow and decreased urinary
osmolality significantly. However, urinary sodium
excretion was significantly increased by atrial
natriuretic peptide, brain natriuretic peptide and
atrial natriuretic peptide plus brain natriuretic
peptide.
5. None of the four infusates significantly altered the
Key words: atrial natriuretic peptide, brain natriuretic peptide, natriuresis, natriuretic peptides, pharmacokinetics.
Abbreviations: ANP, atrial natriuretic peptide, ANP-ir, atrial natriuretic peptidelike immunoreactivity; BNP, brain natriuretic peptide; BNP-ir, brain natriuretic peptidelike
immunoreactivity; ERPF, effective renal plasma flow; GFR, glomerular filtration rate; ICIp, concentration producing half-maximal inhibition; k.i.u., kallikrein international units;
PAH, paminohippurate; PRA, plasma renin activity; TFA, trifluoroacetic acid.
Correspondence: Dr Bernard Cheung, Barnsley District Hospital. Gawber Road, Barnsley S75 2ED, U.K.
723
blood pressure, heart rate or glomerular filtration
rate.
6. This study showed, for the first time, that physiolo-
gical increments in brain natriuretic peptide, like
those in atrial natriuretic peptide, are natriuretic.
Although atrial natriuretic peptide and brain natriur-
etic peptide do not appear to interact synergistically,
they are likely to act in concert in the physiological
regulation of sodium balance.
INTRODUCTION
Since the discovery by De Bold et al. [l] of a
natriuretic factor in the cardiac atria, atrial natriure-
tic peptide (ANP) has been recognized as one
important factor in the regulation of extracellular
fluid volume and sodium homoeostasis. Exogenous
administration of ANP in uiuo has many effects,
including vasodilatation, natriuresis, diuresis and
suppression of the renin-angiotensin-aldosterone
system [2]. In man, ANP causes natriuresis at doses
which raise the plasma concentration of ANP
within the physiological range, in keeping with its
role as a cardiac hormone [3, 41. These properties
make ANP an attractive pharmacological agent for
use in congestive cardiac failure.
Brain natriuretic peptide (BNP), originally iso-
lated from porcine brain [ 5 ] , is closely related to
ANP in structure and is also secreted from the heart
into the circulation [6, 71. The amino acid sequence
of BNP differs considerably among species, more so
than that of ANP, and the actions and degradation
of BNP can be species-specific [S, 91. The diver-
gence in the amino acid sequence of BNP in
different species sometimes gives rise to anomalous
receptor binding [S]. It is therefore necessary to use
the homologous peptide when attempting to unravel
the functions of BNP. In man, BNP may be of
pathophysiological importance in congestive cardiac
failure. The circulating level of BNP is markedly
724 B. M. Y. Cheung et al.
elevated in this condition and may exceed the level
of ANP [lo]. However, the physiological role of
BNP is still unclear, particularly as the plasma
concentration of immunoreactive BNP is low. This
physiological plasma concentration is increased, for
example, by both dietary sodium intake and
hypoxia [ l l , 123, but it remains to be demonstrated
that small increments in plasma BNP levels within
the physiological range have any significant effects.
Hence, we studied the plasma levels of BNP after its
administration in a dose-ranging study. Then we
investigated the effects of human BNP and ANP
infused into normal healthy men at low, physiologi-
cal doses.
METHODS
The protocol of this study was approved by the
district ethics committee.
Doseranging study
Six normal healthy male subjects (age 2&
29 years, weight 75 & 4 kg), on unrestricted diets,
took part in the dose-ranging study with their
informed consent. Each subject passed a screening
medical confirming that he was normotensive, had a
normal electrocardiogram, normal blood indices,
and normal renal and liver function. None was on
any medication. The studies were performed with
the subjects supine. Synthetic human BNP-32
(Peninsula Laboratories, Belmont, CA, U.S.A.) was
dissolved in 0.9% NaCl sterilized by filtration
through a sterile 0.22 pm Millipore filter (Bedford,
MA, U.S.A.), and stored in sterile vials which were
kept at -70C until the time of use. The infusate
consisted of BNP in 0.9% NaCl with 10% Haemac-
cel (Behring, Marburg, Germany), and was delivered
by an infusion pump (Perfusor, Braun, Germany)
into an indwelling intravenous cannula situated in
an antecubital vein at 0.4 pmol min-' kg-'. The
dose rate was doubled every 15 min until
6.4 pmol min-' kg-' was reached. The infusion was
then maintained at this rate for 30 min. Venous
blood (10 ml) was drawn from an intravenous can-
nula in the opposite arm before the infusion and
every 15 min during the infusion, ensuring that
samples were obtained immediately before each dose
increment. Further samples were obtained at 0, 2, 4,
6, 8, 10, 13, 16, 20, 25 and 30 min after the end of
the infusion. Blood pressure and heart rate were
monitored semi-automatically (Datascope 2100
Datascope Corp., Paramus, NJ, U.S.A.) [131. Blood
samples were collected in chilled EDTA polypropy-
lene tubes containing 4000 kallikrein international
units (k.i.u.) of aprotinin (Bayer, Germany) and
50 pg of phosphoramidon (Sigma Chemical Co.,
Poole, Dorset, U.K.), promptly centrifuged at 4C
and stored at -70C before assay.
BNP-like immunoreactivity (BNP-ir) was mea-
sured with a commercial assay (Peninsula Labora-
tories). Briefly, 3 ml plasma samples were acidified
with 0.75 ml of 2 mol/l hydrochloric acid and cen-
trifuged for 10 min at 3000 rev./min. The superna-
tants were loaded on to Sep-Pak C18 cartridges
(Waters Associates, Milford, MA., U.S.A.), which
had been activated with 100% methanol and
double-distilled deionized water. The cartridges were
then washed twice with 5 ml of 0.1% trifluoroacetic
acid (TFA). Elution was performed using 80% meth-
anol, 20% water and 0.1% TFA. The eluates were
dried under vacuum overnight and resuspended in
250 pl of radioimmunoassay buffer. Standard
human BNP-32 or assay sample (100 pl) was incu-
bated overnight at 4C with 100 p1 of rabbit anti-
BNP antiserum (0% cross-reactivity with ANP).
I2'I-BNP-32 (100 pl, 2255 c.p.m.) was added to
each tube and incubated for a further 24 h.
Antibody-bound BNP was precipitated using a goat
anti-rabbit antiserum and counted in a y-counter
(Cobra 5010; Canberra Packard, Pangbourne,
Berks, U.K.). A standard curve was constructed
using serial dilutions of freshly reconstituted human-
BNP-32. The detection limit was 0.5 pmol/l. The
concentration producing half-maximal inhibition
(ICso) was 12 pmol/l. The intra-assay and inter-
assay coefficients of variation were 13% and < 18%,
respectively.
BNP-ir after the end of infusion was first analysed
graphically by plotting the logarithm of BNP-ir
against time. Initial estimates of parameters were
then entered into an iterative non-linear curve-
fitting program (Fig P Version 6.0; Fig P Software
Corporation, Durham, NC, U.S.A.). Half-lives were
calculated from rate constants, k, using the equa-
tion:
t1,2 = 0.693/k
Main study
In the main study, six healthy male subjects (age
22-31 years, weight 7 4 k 6 kg), who had given their
written informed consent, received intravenous infu-
sions of ANP and BNP, separately and in combi-
nation, in a randomized double-blind, placebo-
controlled, four-part cross-over design.
Each subject was studied on four mornings, each
at least 7 days apart. After complete fasting over-
night, each subject arrived at 08.30 hours and
venous cannulae were placed in both forearms
under local anaesthesia. One cannula was used for
the infusion of inulin and p-aminohippurate (PAH),
while the other was used for the infusion of peptide
or placebo. Loading doses and then continuous
infusion of inulin and PAH were given at standard
doses [SO mg/kg followed by 33 mg/min of 10% (w/
v) inulin; Laevosan Gesellschaft, Austria; 8 mg/kg
followed by 12 mg/min of 20% (w/v) PAH; Merck,
Sharp and Dohme, West Point, PA, U.S.A.] in
order to determine the glomerular filtration rate
(GFR) and effective renal plasma flow (ERPF). The
 Effects of atrial and brain natriuretic peptides
subjects remained seated quietly for at least 45 min
to allow the plasma levels of inulin and PAH to
stabilize, before emptying the bladder to start the
urine collection. Urine was collected hourly for 4 h
for the measurement of osmolality, electrolytes, crea-
tinine, inulin and PAH. The peptide or placebo
infusion was given during the last 3 h of urine
collection. The doses of human a-ANP (Bissendorf
Peptides, Germany) and human BNP-32 (Peninsula
Laboratories) were 0.75 and 0.4 pmol min-' kg-',
respectively. The purity of the peptides was con-
firmed by h.p.1.c. as described previously [4]. The
peptides were dissolved in a vehicle of 0.9% NaCl
and 10% Haemaccel and were given at a fixed rate
of 0.5 ml/min (Perfusor; Braun). The control infu-
sion consisted of vehicle only. Subjects remained
seated throughout the study, except when standing
for micturition. Haemodynamic measurements
always preceded blood sampling, which always pre-
ceded urine collection.
Venous samples were obtained from the cannula
in the arm opposite the corresponding infusion.
Venous blood (10 ml) for the measurement of elec-
trolytes, inulin and PAH was drawn at the mid-
point of each urine collection. Venous (37.5 ml)
blood was taken at the end of each hour for the
measurement of plasma renin activity (PRA), ANP
and BNP. Blood samples for ANP and BNP
radioimmunoassays were put into (10 ml) chilled
polypropylene tubes containing 10 mg of potassium
EDTA, 4000 k.i.u. of aprotinin and 50 pg of phos-
phoramidon, centrifuged promptly and the plasma
was frozen in dry ice. Samples for the measurement
of inulin, PAH and PRA were also centrifuged and
stored frozen at -20C or -80C.
Heart rate and rhythm were monitored conti-
nuously. Blood pressure measurements were
recorded in duplicate every 15 min semi-
automatically (Datascope 2100; Datascope Corp.).
Plasma and urine electrolytes were measured with
an autoanalyser (Perspective; American Monitor,
Burgess Hill, Sussex, U.K.). Osmolality was mea-
sured by a vapour pressure osmometer (Westcor,
UT, U.S.A.). Inulin and PAH were measured by an
acid resorcinol method and the Waugh-Beall
method, respectively [14, 151. PRA was measured
by the method of Menard and Catt [16].
ANP-like immunoreactivity (ANP-ir) was mea-
sured by radioimmunoassay. Plasma samples were
first deproteinated with 2 mol/l hydrochloric acid
and loaded on to Sep-Pa4 C18cartridges which had
been activated with methanol and water. The car-
tridges were washed twice with 0.1% TFA and
eluted with 60% acetonitrile in 0.1% TFA. The
eluates were dried under vacuum and each sample
was resuspended in 0.25 ml of radioimmunoassay
buffer (69 mmol/l disodium hydrogen phosphate,
6 mmol/l potassium dihydrogen phosphate,
2.5 mmol/l EDTA, 7 mmol/l sodium azide and 3%
BSA). A series of standards was prepared by serial
dilutions in radioimmunoassay buffer of the same
725
batch of ANP as that infused. Rabbit anti-ANP
antiserum (purchased from Professor S. R. Bloom,
Royal Postgraduate Medical School, London, U.K.)
was added to each sample or standard and incu-
bated overnight at 4C. 1251-ANP (100 pl, 2492
c.p.m.; Amersham), was added to each tube and
incubated at 4C for an additional 24 h. Dextran-
coated charcoal (0.25 ml, Steranti Separex, Steranti
Research Ltd, Herts, U.K.) dissolved in a gelatine-
containing buffer (69 mmol/l disodium hydrogen
phosphate, 6 mmol/l potassium dihydrogen phos-
phate, 2.5 mmol/l EDTA and 2.5 g/l gelatine) was
added to each tube at 4C and promptly centrifuged
at 3000 rev./min for 15 min at 4C. The superna-
tants, which contained the bound fraction of the
label, were removed by suction and the remaining
pellets were counted in a y-counter and counted for
5 min per tube. Cross-reactivity with human BNP-
32 was less than 1%. The minimum detection limit
was 1 pmol/l. The IC,,was 17 pmol/l. The intra-
assay coefficient of variation was 18%. All samples
were assayed together to avoid inter-assay variation.
The recovery of ANP after extraction was 89% as
determined by adding ANP to plasma. All values
are quoted without correcting for this. The concent-
rations of the peptides in infusates were measured
and were 76% and 61% of the calculated concent-
rations of ANP and BNP, respectively.
Clearances of plasma solutes were calculated in
the conventional manner. GFR and ERPF were
measured by inulin and PAH clearance, respectively.
The filtration fraction is the ratio of GFR to ERPF
in percentage. Solute-free water reabsorption (T,,,,,)
was derived from the following equation:
where U,,, and Po,, are respectively urinary and
plasma osmolality and Vis the urinary flow [4].
Results are presented as means and their SEMs,
and were analysed using repeated-measures analysis
of variance (SPSS for Windows, SPSS Inc., Chicago,
IL, U.S.A.). Infusate and time were the repeated
measures. P c 0.05 was considered significant.
RESULTS
Dose-ranging study
Baseline BNP-ir, before the start of the infusion
of BNP, was 1.4k0.5 pmol/l. The BNP infusion
raised the BNP-ir in a dose-dependent manner. At
the end of the infusion, when BNP had been given
at the rate of 6.4 pmol min-l kg-' for 30 min,
BNP-ir reached its highest level, at 21.4 pmol/l (Fig.
1). The decline in BNP-ir after the end of infusion
was biphasic: there was a rapid fall in the first
10min followed by a slower decline. Plotting the
logarithm of BNP-ir against time shows that the
decline in BNP-ir with time is bi-exponential (Fig. 2).
726 B. M. Y. Cheung et al.

BNP infusion 30

25 --I l i

10
>
I

g
I

f
I 1
3
0 15 30 45 60 75 90 92 94 96 100 103 106 I10 115 120
Time (min)

Fig. I . BNP-ir before, during and after infusion of BNP in the

doseranging study. BNP was infused after blood sampling at 0 min, at
0.4 pmol min-' kg-l initially. The dose rate was increased every 15 min I I I I I I I
until it reached 6.4 pmol min-l kg-I, at which rate the infusion was 0 5 10 15 20 25 30
maintained for 30 min. The infusion was discontinued at 90 min. Values are Time (min)
+
means SEM.

Fig. 2. Semi-logarithmic plot of the increment in BNP-ir above

baseline against time in the post-infusion period of the doseranging
study. The data fit the bi-exponential function: y = 10.8e-0.'~'+8.7e-0.09'.
Post-infusion BNP-ir is described by the equation:

Table I. Heart rate and mean blood pressure before infusion

(0 min) and during infusion (&I80 min). The mean blood pressure is
where t is time in min, y is the increment above the diastolic pressure plus one-third of the difference between the systolic
baseline of BNP-ir in pmol/l at time t, and the diastolic pressures. Values are means+SEM.
A = 10.8f1.2 pmol/l, B=8.7f0.9 pmol/l, cc=0.33f 0 min 30 min 60 min 90 min I20 min 150 min 180 min
0.07 min-' and fi=O.O19f.0.005 min-'. The half-
lives derived from the rate constants and are Heart rate (beatslmin)
Vehicle 59+4 57+3 59+3 61+3 59+3 57+4 61+4
2.1 min and 37 min, respectively. Plasma BNP-ir
ANP 58+2 58+3 59+3 61+4 63+4 60+3 62k4
remained higher than the baseline 30 min after the 54+2 54+2 56+2 56+2 56+2 55+2
BNP 56+4
end of infusion. ANPplusBNP 57+3 58+2 56+3 60+3 60+3 58k3 60+3
One subject reported dizziness after the end of the
infusion of BNP. During this episode, which lasted Mean blood pressure (mmHg)
for 5 min, there was a transient fall in the blood Vehicle 92+3 88+2 88+2 91+3 87+2 86+2 92+3
pressure and bradycardia, suggesting vagal overacti- ANP 87+2 87+2 88+3 88+3 90+3 86+3 88+2
vity. BNP-ir in this subject was not different from BNP 91+1 88+2 90+2 92+2 87+2 89+3 88+2
that of the other subjects. None of the other ANPplus BNP 89+2 84+2 85+1 87+2 89+2 89+1 87+2
subjects felt any discomfort, and their blood pres-
sure and heart rate were not significantly altered by
the infusion.
cant changes in the ANP-ir on vehicle and BNP
days.
Main study
Basal plasma BNP-ir was similar in all groups,
with a mean of 2.6f0.2 pmol/l (Fig. 3). BNP levels
There were 110 adverse effects associated with any were unaltered on placebo and ANP days, but rose
of the infusions. Blood pressure and heart rate were within 1 h and remained significantly elevated on
not significantly altered by any of the infusates BNP and ANP plus BNP days, achieving final
(Table 1). There were no significant changes in plasma levels of 4.56 0.66 pmol/l and 5.00 f0.71
serum sodium, potassium, calcium, glucose, creati- pmol/l, respectively.
nine and albumin concentrations or plasma PRA declined during the course of infusion, but
osmolality . the fall during peptide infusion days was not signifi-
Hormonal responses. The mean ANP-ir at base- cantly different from that with vehicle alone (Fig. 3).
line was 8.9k0.4 pmol/l. The infusion of ANP Renal responses. Urine flow declined during
raised this significantly to 12.4+ 1.0 pmol/l (Fig. 3). control infusions (Fig. 4). The infusions of ANP,
The combined infusion of ANP and BNP raised the BNP and ANP plus BNP increased the urine
ANP-ir by a similar amount. There were no signifi- output by 24f. 14%, 11 f.10% and 47 f lo%, respec-
 Effects of atrial and brain natriuretic peptides 121

Baseline hour Baseline hour

1st hour 1st hour
2nd hour 0 2nd hour
3rd hour 3rd hour

*
t5

10
zL

55
n
Vehicle ANP BNP ANP+BNP
Vehicle ANP BNP ANP+BNP
-
g 1200
e l * * *

-
P
L 5
* *
z
5
0
Vehicle ANP BNP ANP+BNP Vehicle ANP BNP ANP+BNP

Vehicle ANP BNP ANP+BNP Vehicle ANP BNP ANP+BNP

Fig. 3. ANP-ir, BNP-ir and PRA before infusion (basline hour) and Fig. 4. Urine flow, osmolality and sodium excretion before infusion
during infusion (1st to 3rd hour). Values are means+SEM. Statistical (baseline hour) and during infusion (1st to 3rd hour). Values are
significance: *P 4 0 5 comparing trend over time with vehicle. Abbreviti means +SEM. Statistical significance: *P <0.05 comparing trend over time
tion: ANG I, angiotensin I. with vehicle.

tively, by the third hour of infusion compared with
the baseline hour. There was a significant trend for
an increase in urine flow on ANP and ANP plus
BNP days.
Urine osmolality increased on the control day,
but the trends on ANP and ANP plus BNP days
were in the opposite direction, leading respectively
to overall decreases of 140 f86 and 65 & 42 mosmol/
kg compared with basal values (Fig. 4). BNP alone
had no significant effect on urine osmolality com-
pared with the control day. Similarly, there was no
significant difference in the trend of the osmolality
between ANP and ANP plus BNP days, suggesting
that BNP has no additive effect on urinary osmola-
lity compared with ANP alone. Solute-free water
reabsorption was not changed significantly by any
of the infusates (Table 2). ANP, which increased
urine flow, did not increase free water reabsorption.
Urinary sodium excretion rate did not change
significantly on the control day, but incieased signi-
ficantly by 55&28%, 49_+18% and 69225% after
3 h of infusion of ANP, BNP and ANP plus BNP,
respectively, compared with the baseline hour (Fig.
4). The significant increasing trends in sodium excre-
tion due to ANP and BNP were similar and were
significantly smaller than that due to the combined
infusion of ANP and BNP. Urinary excretions of
potassium and phosphate were not changed signifi-
cantly by any of the infusates (Table 2).
There were no significant changes in the clearance
of inulin or creatinine (Table 2). ERPF fell signifi-
cantly by 23 k0.4% regardless of infusate. The filt-
ration fraction tended to rise on the active infusion
days, although this trend was significant only on
ANP plus BNP days (Table 2).
DISCUSSION
We have investigated here, for the first time, the
effects of human BNP in man in comparison with
728 6. M. Y. Cheung et al.

Table 2. Urinary parameters during baseline hour and during We have shown that the endogenous circulating
infusion (1st to 3rd hour.) Values are means fSEM. Abbreviations: C,, form of BNP in rat binds to the ANPR-C receptor
creatinine clearance; Gin, inulin clearance; CPAH, PAH clearance; Twate,, with high affinity [21]. Human BNP has been
solutefree water reabsorption; UKV, urinary excretion of potassium; V,U
.,,
shown to bind to human pulmonary ANPR-C
urinary excretion of phosphate. Statistical significance: *P <0.05 comparing
trend over time with vehicle.
receptors [22]. On the other hand, neutral endopep-
tidase may also play a part in the metabolism of
Baseline hour 1st hour 2nd hour 3rd hour BNP, since BNP-ir immunoreactivity is known to
C, (mlimin) increase after the ingestion of candoxatril, an inhibi-
Vehicle 108k6 109k6 104+6 10524 tor of neutral endopeptidase [23]. It is not clear
ANP +
I03 3 103k6 106+6 10226 which clearance mechanism predominates in vivo,
BNP 10624 107k6 107+8 108k7 but the lower affinity of human BNP for the human
ANP plus BNP 110+5 111+7 10725 122+4 ANPR-C receptor may contribute to the longer
half-life of BNP [22]. The maximum BNP level was
Ci, (ml/min)
probably supraphysiological in our pharmacoklnetic
Vehicle l 0 7 k I4 IOOkIO 101k 18 70k9
ANP 71 k I5 77k11 62+12 76k17
study, and it is possible that the clearance mecha-
BNP 67k7 &+I3 87+7 93f17 nism of BNP may vary at different concentrations.
ANP plus BNP 8 7 k I9 123+28 142+39 118k19 The dose of BNP in the main study was chosen
after the dose-ranging study and approximates the
CPAH (ml/min) smallest increment in BNP level that can be reliably
Vehicle 546 k 78 494246 444k36 425240 detected by radioimmunoassay. While it is difficult
ANP 486 If:43 +
330 20 379 k 37 373 k 33 to compare BNP levels precisely between different
BNP 463+48 423+41 393f36 350+20 protocols, the basal levels of BNP we found in our
ANP plus BNP +
482 46 426+30 397k52 367k51
study correspond to those observed by others, who
Filtration fraction (%) found mean levels ranging from 0.9 pmol/l to
Vehicle 21 * 3 20kl 2324 17+2 6 pmol/l [22, 24, 251. The increments in BNP after
ANP 14+2 24k4 1824 2124 infusion in our main study were similar to those
BNP 16+3 23k4 2323 27+5 found with dietary salt loading [ll], and were
ANP plus BNP 20k6 29+6 41 k I4 39213* substantially less than the pathophysiological levels
encountered in patients with congestive cardiac fail-
Twater(ml/min) ure or left ventricular dysfunction [lo, 251. There-
Vehicle 9.5 k I .6 10.2+ 1.6 10.1 k 2 . 4 I1.5k 1.2
ANP +
10.9 I .3 10.0+ 1.4 15.8f 1.7 14.42 1.5 fore, the plasma BNP levels achieved in our study
BNP 11.9+1.3 10.7+ 1.6 12.3k2.0 13.422.4 may be regarded as physiological. These increments
ANP plus BNP 12.4k 2.1 10.9+1.8 12.6k2.2 13.7k1.3 of BNP exerted a potent natriuretic effect, raising
the urinary sodium excretion by 49%. Thus, small
UKV (pnol/min) changes in the endogenous circulating BNP level
Vehicle +
91 I5 97+7 90+9 87+12 could have an important acute impact on renal
ANP 72+7 77+9 7529 65f7 sodium excretion. The present study therefore con-
BNP 61 + 6 75f7 73k6 70+4
52+ 10 60211 67214
firms the findings of previous studies using higher
ANP plus BNP
doses of human BNP [27-291 and supports the
uphosv (pmol/min)
concept that BNP, a cardiac hormone, takes part in
Vehicle 9.7f1.3 10.5f 1.2 12.0k 1.0 14.62 1.6 the physiological regulation of sodium excretion and
ANP 15.2 f4.0 16.7f3.2 17.5+ 1.8 18.1 k 2 . 2 extracellular fluid volume.
BNP 12.0f2.7 l3.8f 3.0 16.3 f2.6 19.5 k 3.2 This study does not allow any conclusions to be
ANP plus BNP I I .Ok3.0 11.9k1.9 16.0+1.5 18.6k1.6 drawn about the long-term effects of BNP. Increas-
ingly negative sodium balance would entrain
counter-regulatory mechanisms, making it unlikely
ANP. Moreover, this is the first report of the effects that the effects of chronic elevations in plasma BNP
of BNP in its physiological range in any species. would match the effects of acute elevations. In
In the first part of our study, we found that the prolonged infusions of ANP in man [30], and in
half-lives of human BNP in man are 2.1 min (fast transgenic mice [3 13, decreased systemic blood pres-
phase) and 37 min (slow phase), which are longer sure in lieu of natriuresis has been observed. The
than the published values for the half-lives of ANP: absence of changes in the blood pressure within the
1.7-3.1 min (fast phase) and 13.3 min (slow phase) period of infusion in our study does not preclude
[17, 181. Nevertheless, the bi-exponential decline in any long-term depressor effects of BNP.
plasma BNP-ir is strikingly similar to that in ANP- Yoshimura et al. [27] found that infusion of BNP
ir [17, 181, suggesting that BNP and ANP share elevated the plasma level of ANP, possibly by
common clearance mechanisms. ANP is removed interfering with the clearance of ANP. We found no
from the circulation by at least two mechanisms: such increase in plasma ANP-ir when BNP was
binding to the ANPR-C receptor, the so-called infused. The small increases in the plasma levels of
clearance receptor [19], and metabolism by neutral BNP achieved here were unlikely to affect the
endopeptidase (atriopeptidase, EC 3.4.24.11) [20]. clearance of ANP to any significant extent.
 Effects of atrial and brain natriuretic peptides
Infusion of ANP and BNP produced a natriuresis
greater than that produced by either peptide alone.
This may indicate that physiological increments in
ANP and BNP do not saturate the biological
receptor to which they both bind, consistent with
affinity constants of ANP and BNP for the biologi-
cal receptor in the nanomolar range [32]. Both
peptides should now be considered when examining
the natriuresis caused by physiological manoeuvres,
such as head-up water immersion or salt ingestion.
BNP may partly explain the long-standing enigma
that physiological manouevres, such as those men-
tioned above, caused a greater natriuretic response
than the elevation in plasma ANP alone could
suggest [33]. BNP, which has the longer half-life,
may also account for the perplexing phenomenon of
the persistence of the natriuretic response to stimuli
even after plasma ANP levels decline [34].
Although the natriuretic effects of ANP and BNP
were comparable, the two peptides may differ in
their effects on urine osmolality. Whereas there was
a significant fall in urine osmolality due to ANP,
BNP did not significantly change urine osmolality,
although small changes in urine osmolality might be
obscured by the errors introduced by the urinary
dead space in dehydrated subjects. ANP has been
shown to decrease urine osmolality in man by
antagonizing the effects of vasopressin [35], and this
would be consistent with the failure of free water
reabsorption to rise during ANP infusions in the
present study despite increases in urinary sodium
excretion. The significant effect of ANP on urine
osmolality observed in this study confirms previous
results [4, 351. It is also relevant that pathological
levels of BNP do not suppress urine osmolality
significantly [28]. These results, which imply differ-
ential control of water excretion by the natriuretic
peptides, may be presaged in the effects of oxidized
human u-ANP in the rat. Oxidation in the methio-
nine residue of u-ANP (met-0-ANP) results in a
peptide which increases water excretion but not
sodium excretion [36]. Consequently, met-0-ANP
has effects which are the inverse of those described
here for BNP. Furthermore, it is possible that
oxidation of a-ANP in vivo may account for the
increased water excretion caused by ANP but not
by BNP.
That ANP and BNP should have dissimilar
effects may, at first sight, seem surprising in view of
the shared receptors, ANPR-A and ANPR-C [21,
371. One possibility is the existence of further
receptor subtypes. Differential degradation of ANP
and BNP by neutral endopeptidase may also
explain differences in their actions [9]. Human BNP
has a lower affinity for ANPR-C than ANP [22].
The local concentration of ANPR-C might therefore
alter the local concentration of BNP compared with
ANP available for binding to ANPR-A. Moreover,
there is accumulating evidence that ANPR-C may
be coupled to intracellular second messengers in
addition to its function as a clearance receptor,
729
although this is controversial [38]. It is therefore
possible that BNP may bind poorly to a physiologi-
cally relevant receptor that ANP binds with high
affinity. The differences between the diuretic and
natriuretic effects of BNP, u-ANP and met-0-ANP
also raise the possibility of receptor heterogeneity.
High doses of ANP and BNP cause systemic
hypotension and tachycardia [27]. Thus, the lack of
alterations in the blood pressure and heart rate may
indicate that, at physiological levels of ANP and
BNP, any vasomotor effects of the natriuretic pep-
tides are balanced by other neuronal and hormonal
factors. At higher doses, the hypotensive effects of
the natriuretic peptides may overcome the normal
physiological mechanisms regulating blood pressure
and heart rate. We have not found any significant
fall in the PRA nor any rise in plasma proteins
compared with control. These potentially important
actions of ANP and BNP, of suppressing the renin-
angiotensin-aldosterone system [39] and shifting
fluid from the intravascular to the extravascular
compartment [40], may likewise be inoperative with
small increments in ANP or BNP. Interestingly,
porcine and human BNP, given at doses higher
than ours, suppressed plasma aldosterone but not
renin [28, 29, 411.
GFR was not significantly altered by any of the
infusates, although, because of dead space errors,
the present method of measuring GFR would not
be accurate enough to detect small but consequen-
tial changes. As most of the filtered load of sodium
is reabsorbed under normal circumstances, a minute
alteration in the GFR could account for the
observed effects of the natriuretic peptides on
sodium excretion. Interestingly, the filtration frac-
tion increased on the peptide infusion days,
although the increase was significant only on ANP
plus BNP days when compared with control. Dead
space errors in the measurement of GFR and ERPF
cancel out in the filtration fraction [4]. The present
observations therefore raise the possibility that glo-
merular actions may be a component of the renal
effect of even physiological levels of the natriuretic
peptides.
The natriuretic effect of BNP suggests that BNP
may be useful therapeutically in congestive cardiac
failure. Indeed, pharmacological doses of BNP have
an impressive natriuretic effect in patients with
congestive cardiac failure, coupled with improve-
ments in the pulmonary capillary wedge pressure
and the cardiac index [27]. Candoxatril, an orally
active neutral endopeptidase inhibitor, may be an
effective alternative to parenteral administration of
BNP, as candoxatril raises both plasma ANP and
BNP concentrations [23].
In conclusion, the present study shows that BNP
is natriuretic at physiological levels. BNP may
therefore take part in sodium homoeostasis and
regulation of extracellular fluid volume in conjunc-
tion with ANP. However, BNP may act differently
from ANP on water excretion. Further studies are
730 8. M. Y. Cheung et al.
necessary to refine our understanding of the roles of
ANP and BNP in different physiological and patho-
physiological states.
ACKNOWLEDGMENT
The support of the British Heart Foundation is
gratefully acknowledged.
REFERENCES
I, De Bold AJ, Borenstein HB, Veress AT, Sonnenberg H. A rapid and potent
natriuretic response to intravenous injection of atrial myocardial extracts in
rats. Life Sci 1981; 28: 89-94.
2. Struthers AD. Atrial natriuretic factor. Oxford: Blackwell Scientific
Publications, 1990.
3. Anderson JV, Donckier J, Payne NN, Beacham J, Slater JDH, Bloom SR. Atrial
natriuretic peptide: evidence of action as a natriuretic hormone at
physiological plasma concentrations in man. Clin Sci 1987; 72: 305-12.
4. Brown J. O'Flynn. Acute effects of physiological increments of @atrial
natriuretic peptide in man. Kidney Int 1989 36 645-52.
5. Sudoh T, Kangawa K, Minamino N, Matsuo H. A new natriuretic peptide in
porcine brain. Nature (London) 1988; 332 76-81.
6. Saito Y, Nakao K, ltoh H. Brain natriuretic peptide is a novel cardiac
hormone. Biochem Biophy Res Commun 1989; 158: 360-8.
7. Kambayashi Y, Nakao K, Mukoyama M, et al. Isolation and sequence
determination of human brain natriuretic peptide in human atrium. FEBS Lett
1990; 213: 341-5.
8. Kambayashi Y, Nakao K, Kimura H, et al. Biological characterization of human
BNP and rat BNP-species-specific actions of BNP. Biochem Biophys Res
Commun 1989; 173: 599-605.
9. Norman ]A, Little D, Bolgar M, di Donato G. Degradation of brain natriuretic
peptide by neutral endopeptiiase: species specific sites of proteolysis
determined by mass spectrometry. Bimhem Biophys Res Commun 1991; 175:
22-30.
10. Mukoyama M, Nakao K. Saito Y, et al. Increased human brain natriuretic
peptide in congestive cardiac failure. N Engl J Med 1990 323: 757-8.
I I. Lang CC, Coutie WJ. Khong TK, Choy AM], Struthers AD. Dietary salt
loading increases plasma brain natriuretic peptide levels in man. J Hypertens
1991; 9 779-8882,
12. Lang CC, Coutie WJ, Struthers AD, Dhillon DP, Winter JH, Lipworth BJ.
Elevated levels of brain natriuretic peptide in acute hypoxaemic chronic
obstructive pulmonary disease. Clin Sci 1992; 83: 529-33.
13. Morice A, Pepke-Zaba J, Loyren E. et al. Low dose infusion of atrial
natriuretic peptide causes salt and water excretion in normal man. Clin Sci
1988; 74: 359-63.
14. Schreiner GE. Determination of inulin by means of rerorcinol. Proc Soc Exp
Biol Med 1950 74: 117-20.
IS. Waugh WH, Beall PT. Simplified measurement of paminohippurate and other
arylamines in plasma and urine. Kidney Int 1974; 5: 429-36.
16. Menard J, Catt K. Measurement of renin activity, concentration and zubstrate
in rat plasma by radioimmunoassay of angiotensin I. Endocrinolcgy (Baltimore)
1972; W: 422-30.
17. Yandle TG, Richards AM, Nicholls MG, Cuneo R, Espiner EA, Livesey JH.
Metabolic clearance rate and plasma half-life of alpha-human atrial natriuretic
peptide in man. Life Sci 1986; 38: 1827-33.
18. Nakao K, Sugawara A, Morii N. et al. The pharmacokinetics of alpha-human
atrial natriuretic peptide in healthy subjects. Eur J Clin Pharmacol 1986; 31:
101-3.
19. Maack T, Suzuki M, Almeida FA, et al. Physiological role of silent receptors of
atrial natriuretic factor. Science (Washington D.C.) 1987; UB: 675-8.
10. Kenny AJ, Stephenson SL. Role of endopeptidase 24. I I in the inactivation of
atrial natriuretic peptide. FEBS Lett 1988; 232: 1-8.
21. Brown J, Cheung B. Binding of rat brain natriuretic peptide (BNP) t o atrial
natriuretic peptide (ANP) receptors in rat kidney. J Phyriol (London) 1992;
446: 91P.
22. Mukoyama M, Nakao K. Hosoda K, et al. Brain natriuretic peptide as a novel
cardiac hormone in humans. J Clin Invest 1991; 8 7 1402-12.
23. Lang CC, Motwani JG, Coutie W, Struthers AD. Influence of candoxatril on
plasma brain natriuretic peptide in heart failure [Letter]. Lancet 1991; ii: 255.
24. Buckley MG. Sethi D, Markandu ND, Sagnella GA, Singer DGJ, MacGregor
GA. Plasma concentrations and comparisons of brain natriuretic peptide and
atrial natriuretic peptide in normal subjects, cardiac transplant recipients and
patients with dialysis-independent or dialysisdependent chronic renal failure.
Clin Sci 1992; 83 437-44.
25. Motwani J, McAlpine H, Kennedy N, Struthers AD. Plasma brain natriuretic
peptide as an indicator for angiotensin-converting-enzyme inhibition after
myocardial infarction. Lancet 1993; MI: 1109-13.
26. Yandle TG, Richards AM, Gilbert A, Fisher S, Holmes S , Espiner EA. Assay of
brain natriuretic peptide (BNP) in human plasma: evidence for high molecular
weight BNP as a major plasma component in heart failure. J Clin Endocrinol
Metab 1993; 7 6 832-8.
27. Yoshimura M, Yasue H, Morita E, et al. Haemodynamic, renal, and hormonal
responses to brain natriuretic peptide infusion in patients with congestive
cardiac failure. Circulation 1991; 84: 1581-88.17.
28. Holmes SJ, Espiner EA, Richards AM, Yandle TG, Frampton C. Renal,
endocrine, and haemodynamic effects of human brain natriuretic peptide in
normal man. J Clin Endocrinol Metab 1993; 76 91-6.
29. Richards AM, Crozier IG, Holmes SJ, Espiner EA, Yandle TG, Frampton C.
Brain natriuretic peptide: natriuretic and endocrine effects in essential
hypertension. J Hypertens 1993; II: 163-70.
30. Singer DGJ, Markandu ND, Buckley MG, et al. Prolonged decrease in blood
pressure after atrial natriuretic peptide infusion in essential hypertension: a
new anti-pressor mechanism. Clin Sci 1989 77: 253-8.
31. Field LJ, Veress AT, Steinhelper ME, Cochrane K, Sonnenberg H. Kidney
function in ANF-transgenic mice: effect of blood volume expansion. Am J
Physiol 1991; 260: RI-5.
32. Suga S, Nakao K, Hosoda. K, et al. Receptor selectivity of natriuretic peptide
family, atrial natriuretic peptide, brain natriuretic peptide and C-type
natriuretic pepride. Endocrinology (Baltimore) 1992; 130 229-39.
33. Goetz KL. Physiology and pathophysiology of atrial peptides. Am J Physiol
1988; 254: El-IS.
34. Goetz KL. Evidence that atriopeptin is not a physiological regulator of sodium
excretion. Hypertension 1990; 15: 9-19.
35. Brown J, Forsling ML, Valdes G, Slater ID, Dollery CT. Antagonism of
V,-receptor effect of antidiuretic hormone by atrial natriuretic peptide in man.
Experientia 1988; 44: 513-16.
36. Willenbrock RC, Tremblay J, Garcia R, Hamet P. Dissociation of natriuresis
and diuresis and heterogeneity of the effector system of atrial natriuretic
factor in rats. J Clin Invest 1989; 83: 482-9.
37. Koller KJ, Lowe DG, Bennett GL. et al. Selective activation of the B
natriuretic peptide receptor by C-type natriuretic peptide (CNP). Science
(Washington, D.C.) 1991; 251 120-3.
38. Levin ER. Natriuretic peptide C-receptor: more than a clearance receptor. Am
J Physiol 1993; 164: E483-9.
39. Cuneo RC, Espiner EA, Nicholls MG, Yandle TG, Livesey JH. Effects of
physiological levels of atrial natriuretic peptide on hormone secretion:
inhibition of angiotensin-induced aldosterone secretion and renin release in
normal man. J Clin Endocrinol Metab 1987; 65: 765-2.
40.Williamson JR, Holmberg SW. Chang K, Marvel J, Sutera S, Needleman P.
Mechanisms underlying atriopeptin-induced increases in haematocrit and
vascular permeation in rats. Circ Res 1989 M 890-9.
41. McGregor A, Richards M, Espiner E, Yandle T, lkram H. Brain natriuretic
peptide administered to man: actions and metabolism. J Clin Endocrinol Metab
1990; 7 0 1103-7.