Download Efficiency of Utilization of Urea Nitrogen for

Clinical Science and Molecular Medicine (1975) 48, 379-390.
Efficiency of utilization of urea nitrogen for albumin synthesis by
chronically uraemic and normal man
R. VARCOE, D. HALLIDAY, E. R. CARSON,· P. RICHARDSt
AND
A. S. TAVILL
MRC Clinical Research Centre, Harrow, Middlesex,
• Department of Systems Science, City University, London, and. tSt George's Hospital, London
(Received 7 October 1974)
Summary
1. The relation between .endogenous urea metabolism and albumin synthesis has been studied in
ten patients with chronic renal failure and in four
normal subjects, after single intravenous injections
of [14C]urea, [15N]urea and 125I-Iabelled albumin.
2. The rate of urea synthesis was determined from
the dynamics of plasma [14C]urea specific radioactivity and the rate of urea metabolism was estimated from the relative rates of urea synthesis and
urea appearance in urine and body water. Deconvolution analysis of plasma [15N]albumin
enrichment and 125I-Iabelled albumin radioactivity
yielded the cumulative incorporation of 15N
into total exchangeable albumin and enabled
calculation of the absolute rate of urea nitrogen
utilization for albumin synthesis.
3. Although the mean absolute rate of urea degradation in uraemic patients (3'7 mmol/h) was
higher than in normal subjects (2'3 mmol/h) there
was no significant positive correlation between urea
degradation and plasma urea concentration.
4. In uraemic subjects, there was a significant
positive correlation between urea synthetic rate and
urea degradation rate.
5. The rate of utilization of urea nitrogen for
albumin synthesis was low, but was very much
higher in uraemic subjects (mean 83'8 pmol/h)
compared with normal subjects (mean 6'4 pmol/h),
as was the provision by urea of the nitrogen required
for albumin synthesis in uraemic subjects (2'37%)
compared with normal subjects (0,13%).
6. The efficiency of utilization of urea nitrogen
for albumin synthesis was higher in the uraemic
patients (1'3 %) than the normal subjects (0'2 %), and
was higher in those patients with chronic renal
failure who received a 30 g protein diet than those
on 70 g of protein. A significant negative correlation
was noted between efficiency of urea nitrogen utilization and the rate of synthesis of albumin.
7. These studies suggest the presence of a mechanism for the conservation of urea nitrogen in
chronic renal failure which is unrelated to the extent
of urea degradation, and which can only be partly
explained by the higher proportion of intraluminal
gut nitrogen derived from urea.
Key words: albumin synthesis, deconvolution
analysis, intestinal hydrolysis, nitrogen reutilization,
uraemia, urea nitrogen.
Introduction
Since the original demonstration by Walser &
Bodenlos (1959) that approximately 20% of the urea
synthesized by normal man is degraded in the gastrointestinal tract, there have been a number of studies
suggesting that released ammonia nitrogen may
be utilized for protein synthesis. For example.
urea was used to maintain a normal growth rate
in protein-deprived infants, and 15Nfrom this orally
administered p 5N]urea was found in haemoglobin
and plasma protein (Snyderman, Holt, Dancis,
Roitman, Boyer & Balis, 1962). Similarly, urea
played a role in the maintenance of nitrogen balance
in uraemic patients receiving a diet containing
Correspondence: Dr A. S. Tavill, MRC Clinical Research
Centre, Harrow, Middlesex HAl 3UJ.
379
380
R. Varcoe et al.
2 g of essential amino acid nitrogen (Giordano,
1963).
Quantitative assessment of the extent of urea
nitrogen utilization has been based upon isotopic
studies using orally administered (l5N]ammonium
chloride or [15N]urea (Snyderman et al., 1962;
Richards, Metcalfe-Gibson, Ward, Wrong &
Houghton, 1967; Giordano, De Pascale, Balestrieri,
Cittadini & Crescenzi, 1968; Read, McLaren,
Tchalian & Nassar, 1969), or intravenously injected
[15N]urea (Read et al., 1969). However, evaluation
of the efficiency of utilization of endogenous urea
nitrogen has been limited by the effects of unrepresentative precursor ammonia nitrogen enrichment in portal venous blood after oral isotope
administration and the failure to define simultaneously the true dynamics of urea metabolism and protein
turnover by means of non-recycling isotopic preparations.
Recently, it has been possible by means of a
combined intravenous [15N]urea, [14C]urea and
125I-labelled albumin technique to demonstrate
that there was an increase in the rate of urea hydrolysis and an enhanced utilization of urea nitrogen
for albumin synthesis in a patient with the gastrointestinal stagnant-loop syndrome (Varcoe, Halliday
& Tavill, 1974). Since uraemic patients on a lowprotein diet rich in essential amino acids may have
a relative deficiency in exogenous non-essential
amino acids and an increased endogenous supply
of ammonia nitrogen derived from hydrolysed
urea (Giordano, 1963; Richards et al., 1967),
it is relevant to evaluate the efficiency of their hepatic
reutilization mechanisms. We have therefore carried
out studies in patients with chronic renal failure
and normal volunteers on normal and low protein
diets, and have measured the contribution of endogenous urea nitrogen to the synthesis of a specific
liver-produced protein, namely albumin.
Materials and methods
Subjects
The full nature and purpose of the investigation
was explained to patients and volunteers and their
consent obtained before the study. Studies were
performed on six uraemic patients (group A), who
had been on a 30 g protein diet, high in essential
amino acids (Giovanetti & Maggiore, 1964) for
4-18 months before study, and on four uraemic
patients (group B), who were eating a 70 g protein
diet. Studies were also carried out on four normal
volunteers (group C), eating a 70 g protein diet,
two of whom were subsequently studied again
after they had been eating a 30 g protein diet,
rich in essential amino acids, for 9 days. The food
for all subjects during the study was provided and
supervised by a dietitian, and sufficient energy,
usually 10500 kJ/day, was provided to maintain
weight. All patients were in hospital for the study
and the normal volunteers were given meals of six
equal protein portions spaced throughout the day
to avoid large surges in urea synthesis.
Isotopic materials
[14C]Urea (specific radioactivity 58·1 mCi/mmol,
The Radiochemical Centre, Amersham, Bucks.,
U.K.) was prepared as a sterile solution of 10 /lCi
in 10 ml of 154 mmol/I sodium chloride (physiological saline). Approximately 4, 8 or 16 mmol of
[15N]urea (99'6 atoms % enriched, synthesized
free of other labelled nitrogen by Prochem Division
of British Oxygen Company) was weighed and dissolved in the [14C]urea solution (approximately
10 /lCi) and filtered through Millipore filters immediately before reweighing and injection. Purified
human serum albumin was labelled with 1251 by the
iodine monochloride method of McFarlane (1956,
1958).
Protocol of investigation
Uptake of radioactive iodine by the thyroid
was blocked with oral potassium iodide. At zero
time approximately 10 /lCi of 125I-labelled albumin
was weighed and injected intravenously and, immediately after a heparin-treated blood sample was taken
10 min later a weighed amount of the solution of
combined (l4C]urea and [15N]urea was injected intravenously and an aliquot saved for preparation
of standards. Accurately timed heparin-treated
venous blood samples were taken at 2 h intervals
for 12 h, and then daily for 10 days from normal
subjects, and approximately fifteen times in 10 days
in patients with chronic renal failure.
In normal subjects urine was collected at measured
intervals of approximately 2 h throughout the first
12 h and thereafter as pooled 24 h collections.
Patients with chronic renal failure collected 24 h urine
samples for 7 days.
Utilization of urea nitrogen for albumin synthesis
Measurement of urea and albumin concentrations
Plasma and urine urea concentrations were
measured by the standard diacetyl monoxime
method on a Technicon Autoanalyser II. Plasma
albumin concentrations were measured by an
automated Bromocresol Green dye-binding method
(Doumas, Watson & Biggs, 1971).
Measurement of radioactivity
1251-labelled albumin. Trichloroacetic acid-precipitable radioactivity in 2 ml aliquots of plasma
was measured in a Wallac Decem series auto gamma
spectrometer.
[14C]Urea. Because of overlap in the radiation
spectrum of 14C and 1251, protein-bound iodine
was precipitated from 2 ml plasma samples with
an equal volume of 20% (w/v) trichloroacetic acid
and the free iodine in the supernatant was removed
by exposure to LR.A. 400 (BDH Ltd) resin. The
equivalent of 1 ml of plasma was adjusted to pH
7·0 with sodium hydroxide (1 mol/l), and then made
to a volume of 5 ml with distilled water before
being dissolved in 10 ml of Unisolve (Koch-Light
Laboratories Ltd), and cooled to 17°C before counting in a LKB liquid-scintillation spectrometer.
Aliquots of 1 ml of urine were treated similarly
except for the omission of the deproteinizing step
when there was no proteinuria. Correction for
quenching was made by the external standard
channels-ratio method, which was validated for
14C by use of an internal standard technique.
Measurement of 15 N enrichment in albumin'
Albumin samples, prepared from heparin-treated
plasma by the method of Korner & Debro (1956),
were dialysed against many changes of distilled
water for 7 days to remove all labelled urea. Purity
of the albumin thus obtained was confirmed by
polyacrylamide disc-gel electrophoresis. After Kjeldahl digestion of the samples and distillation into
sulphuric acid, nitrogen was liberated from the
resulting ammonium sulphate solution with lithium
hypobromite and analysed for its 15N content on
an AEI MS.20 mass spectrometer. The 15N enrichment was expressed as atoms % excess compared
with a sample obtained from each patient before
administration of [15N]urea. The reproducibility
of the mass spectrometry was checked on a standard
D
381
enriched sample of ammonium sulphate (SD ± 0'21 %,
n = 13)over a 14 months period, and also on separate purified samples of enriched human albumin
(SD ±002%, n = 13).
Calculations
Plasma and urine [14C]urea specific radioactivity.
This was obtained by dividing the plasma or urine
14C radioactivity (d.p.m.jml) by the urea concentration (mmol/ml). This method of obtaining 14C
specific radioactivity was validated by injecting
[14C]urea and [15N]urea simultaneously in a normal
individual and comparing the decay curves of plasma
14C specific radioactivity and the [15N]urea enrichment based upon measurement of mle 28 and
30 ion currents derived from purified urea extracted
from plasma (Walser, George & Bodenlos, 1954).
These gave almost identical slopes (0 0143 and 0'141).
Likewise, plasma samples analysed directly for
[14C]ureaspecificradioactivity by the urease-vacuum
line method (Tavill, Craigie & Rosenoer, 1968)
gave similar results.
Urea space. This was calculated as the 14C
radioactivity administered (d.p.m.), divided by the
intercept on the ordinate of the semi-logarithmic
decay curve of plasma 14Cradioactivity (d.p.m.jml).
Robson (1964) showed that in renal failure the
amount of labelled urea excreted during the period
of equilibration between plasma and urea space
in patients with chronic renal failure is very small.
We have confirmed this and 'excess excretion'
has therefore been ignored in these patients. However, Walser & Bodenlos (1959) calculated that
excess excretion in normal subjects varied from
o to 19%. Excess excretion is the amount of
radioactive [14C]urea appearing in the urine during
the equilibration period which is the result of the
high, post-injection intravascular urea specific
radioactivity relative to that of urea in the extravascular body water. If equilibration of the isotopic
urea throughout total body water had been instantaneous, the kinetics of urinary [14C]ureaspecific
radioactivity would have been exponential from
zero time. Excess urinary excretion can be determined graphically for each collection period up to
equilibration by multiplying the difference between
the extrapolated urinary [14C]urea specific radioactivity and the observed specific radioactivity
at the midpoint of the collection by the total urea
excreted during that period (Fig. 1). This was sub-
R. Varcoe et al.
382
u
o
~
::>
20
'0
Observed minus expected
specific radiaactivity
a
E
..!:.
10
~ ~
~
:2:u
7
6
o
5
~
4
.2
3
2
I
I
I
I
--..I
I
Urine delay time
2345678
Time (hI
FIG. I. Dynamics of plasma (0) and urinary (6) [14C]urea
specific radioactivity in a normal subject (A.L.) receiving a
70 g protein intake, after a single intravenous injection of
approximately 10 /lCi of [l4C]urea. The graphical determination of excess urinary excretion and of urinary delay times
are illustrated (see the Materials and methods section).
tracted where appropriate from the actual injected
dose in order to give the effective injected dose.
Initial urea pool. This amount (mmol) is the urea
space (I) multiplied by the initial plasma urea concentration (mmol/l).
Urea synthetic rate (mmol/h). Distribution of
injected [14C]urea occurred very rapidly and was
followed by exponential decay of plasma urea
carbon specific radioactivity. Linear regression
analysis by the least-mean-square method of the
semi-logarithmic plot of specific radioactivity over
12 h in normal subjects and 7 days in patients gives
a slope (k), which when multiplied by the urea
pool size gives the urea synthetic rate. This method
for measurement of urea synthesis makes the assumption that the synthetic rate remains constant for
the duration of the experiment. Urea half-life
(t i) is calculated from the decay constant of the
slope (k) from the formula tt = Q·693/k.
Urea excretion rate (mmol/h). This is calculated
from the amount of urea excreted in the urine over
the period considered in calculating k.
Ideally, subjects should be in a completely steady
state as evidenced by a constant plasma urea concentration. However, despite run-in periods and
constant protein diets plasma urea concentrations
did change, usually decreasing in uraemic patients.
When calculating the amount of urea metabolized
in the intestine, allowance was made for changes
in body urea pool mass. Thus the rate of change
of urea pool (P, mmol/h) was calculated from the
rate of change of plasma concentration of urea
over the course of the experiment.
Urea metabolized in the gut (M, mmol/h). This
was calculated as the difference between urea synthesized (8) and excreted into urine (U) and accumulating in or disappearing from the pool (P), M =
8 - U - P. This calculation assumes semi-steadystate conditions, the rates of synthesis, metabolism
and excretion being constant, with the urea pool
changing as a result of the constant difference between synthesis and the sum of excretion and metabolism.
Urine delay time. This is the interval between
times at which [14C]urea specific radioactivities
were the same in plasma and urine (Regoeczi,
Irons, Koj & McFarlane, 1965) and is determined
graphically as illustrated in Fig. 1.
Plasma volume (I). This was derived from the
dilution of the injected 125I-Iabelled albumin.
Intravascularalbumin(g). This was calculated from
the product of plasma volume and plasma albumin
concentration (g/l).
Fractional turnover rate of albumin. This, as
~
~
gu
o~
~~::>
.~-
'ij
CZ>-
6
~o
0.0
en E
o E
CZ>"'"
::;E
,......,<1.
u~
s:
'---'
><
5
4
3
2
'i'
'2
1
0
FIG. 2. Dynamics of plasma (0) and urinary (6) [14C]urea
specific radioactivity in a uraemic patient (E.O.) receiving
a 30 g protein intake, after a single intravenous injection of
approximately 10 /lCi of [l4C]urea. The long half-life and
long urinary delay time are illustrated.
Utilization of urea nitrogen for albumin synthesis
ylf)
mit)
Rate of change of
newly synthesized
plasma [~]albumin
12SI-labelled albumin
distribution and
degradation dynanics
Observed plasma
]albumln
enrichment
rSN
FIG. 3. Model for deconvolution of experimentally observed
albumin dynamics.
percentage of intravascular albumin degraded per h,
was calculated by resolution of the plasma 12sI_
labelled albumin radioactivity curves by the method
of Matthews (1957). Under steady-state albumin
metabolic conditions, as defined by a constant plasma
albumin concentration, the absolute synthetic rate
(mgfh) could be calculated from the product of
the fractional catabolic rate and the intravascular
albumin pool. Urinary albumin losses in all patients
were very small ( < 0·7 g/day).
Deconvolution analysis of the
p SNialbumin curve
The observed P sN]albumin enrichments are
based upon a pulse input of pSN]urea as influenced
by the urea dynamics, which include distribution, hydrolysis and recycling and the distribution and
degradation of newly synthesized albumin. If y(t)
is the value of the observed p sN]albumin enrichment
at time t and g(t) is the value at time t of the 12sI_
labelled albumin curve, which defines the albumin
dynamics of distribution and degradation, these
curves are related by the following convolution
integral (Vitek, Bianchi & Donato, 1966):
f
t
y(t) = _1_
g(O)
m(t-'r) . g(-r) • dr
o
where m(t) is the rate of change of newly synthesized
plasma pSN]albumin enrichment (see Fig. 3).
This convolution integral can be solved for m(t)
by numerical inversion (Jones, Craigie, Tavill,
Simon & Rosenoer, 1968). Integrating m(t) with
respect to time enables u(t) to be derived:
t
u(t) =
f
m(t)· dt
o
where u(t) is the plasma p SN]albumin enrichment
383
which would have been obtained had all the newly
synthesized albumin remained intravascularly and
undegraded. Therefore the incorporation of 1sN
into albumin was obtained from the product of
the maximum albumin enrichment resulting from
this deconvolution and the intravascular albumin
pool. Since this can be expressed as a percentage of
the injected dose of pSN]urea, the utilization of
endogenous urea can be obtained from the product
of the 1SN incorporated and the initial urea nitrogen
pool divided by the time taken to reach maximum
labelling. In chronic renal failure, there was delay
in completion of albumin labelling. After the period
of linear labelling which followed the pulse p sN]urea
input, there was a small but continuing rise in albumin enrichment which probably represents the
effects of recycling of nitrogen from several sources
other than solely urea. In the uraemic patients maximum albumin enrichment is defined as the time at
which only these effects remain, namely the end of
linear incorporation. The rate of incorporation of
1sN was obtained from the linear portion of the
deconvolution curve; in practice this was calculated
from the enrichment achieved between the time
of intercept on the abscissa and the time at which the
deconvolution curve starts to reach a plateau (Table
2; Fig. 4).
Results
Urea metabolism (Table 1)
The selection of patients able to tolerate a normal
protein intake was reflected by the generally higher
rate of urea clearance in patients of group B. Only
one (S.S.) with a low urea clearance and a high
plasma urea concentration had later to discontinue
this diet. Each of the normal volunteers had a
lower plasma urea concentration at the start of the
second study after being on a low-protein diet, and
body urea pools were proportionately decreased.
The urea space, as percentage of body weight (Table
1, column 4) calculated from the extrapolated radioactivity regression line, was similar in uraemic
(mean 65'3%) and normal subjects (mean 64'8%).
Excess of excretion of [14C]urea was zero in chronic
renal failure, indicating the rapidity of distribution
compared with renal excretion. In normal subjects,
excess excretion was less than 2 % of the injected
dose.
The linear regression analysis of the p4C]urea
R. Varcoe et al.
384
.~1i 100
"0
Cl)
98
~
~
0._ 8
~~.s ~8
i! e ~
50
,g.~ ~ 40
::> 0
~~ 30
30
20
o
60 C
~
50
:gc:
'"
e
40 ~ ~
"0 ~
.0 '"
30
.---.'"
if
'0
20
;fl
c:....",
-S
E
£
oS
>
c:
10 8
o'"
40
60
80
100 120
Time (hI
140
160
180 200
220
"o"
FIG. 4. Semilogarithmic plot of the dynamics of exogenous 12sI-labelled albumin (6) and biosynthetically labelled
[lsN]albumin (0) after the simultaneous intravenous administration of 12sI-labelled albumin and [lsNlurea to a
uraemic patient (E.M.) The deconvoluted [lsNlalbumin enrichment (--) is plotted on a linear scale and represents
the dynamics that would have been observed in the absence of distribution and degradation of the newly labelled
albumin. The rate of incorporation of lsN in this patient was calculated from the enrichment achieved during the
period to (i.e. the time of intercept on the abscissa) to the time at which the deconvolution curve deviates from
linearity (- - -).
specific radioactivity curves showed that decay
constants in uraemic patients (Table 1, column 9)
were very much lower than in normal subjects, yielding higher values for urea half-life in patients (groups
A+B,meantt65'5h±9'7hsEM,n = 10)as compared
with normal subjects (group C, mean tt 8·3 h±0'9
h SEM, n = 6). However, when converted into urea
synthetic rate (Table 1, column 10) by multiplying
by the urea pool, it was apparent that uraemic
patients on a 30 g protein diet synthesized urea at
a rate similar to that of normal subjects on this
intake, whereas on a 70 g protein intake urea synthesis was higher but again similar in patients and
normal subjects. Virtually all the patients in groups
A and B showed a fall in body urea pool over the
course of the study, in contrast to those in group C
who showed a rise or no change over the much
shorter time-course of the [14C]urea study. This
may be explained by a normal diurnal rhythm in
the volunteers, whereas in chronic renal failure a
supervised protein intake at either level may have
permitted urea loss by all routes to exceed synthesis.
Urinary urea delay times estimated graphically
can only be approximate. Only one normal subject
(A.L.) had a measurable delay time, which was 1 h
and 1·5 h in the two successive studies. In all three
other volunteers the delay was virtually zero.
In chronic renal failure the delay time was usually
longer and the plasma and urine ['4C]urea specific
radioactivity decay curves occasionally diverged,
the plasma falling faster than the urine with a progressive lengthening of delay times.
The rate of urea metabolism (Table 1, column
13) was greatest in the uraemic patients on a normal
protein intake. However, when expressed as a percentage of the urea synthetic rate (Table 1, column
14) it was highest in uraemic patients on a lowprotein diet and progressively lower in normally
fed uraemic patients, and normal subjects on a
70 g protein intake. Both normal subjects when put
on a low-protein intake increased their rate of urea
breakdown as a percentage of synthesis but not in
absolute terms. There was a significant positive
correlation between the plasma urea concentration
TABLE
I. Urea metabolism data in uraemic patients and normal subjects
F
F
F
F
M
M
Group A
E.M.
E.O.
J.G.
R.C.
M.R.
J.S.
MeanA±sEM
Group C
S.H.
R.V.
A.L. (I)
P.B. (I)
A.L. (2)
P.B. (2)
Mean C±SEM
M
F
M
M
M
M
Group B
J.A.
M
C.P.
M
S.S.
M
F.S.
M
MeanB±sEM
MeanA+B±sEM
Sex
Patient
70
70
70
70
30
30
70
70
70
70
30
30
30
30
30
30
(I)
90'0
56'5
65'0
65'0
65'0
64'0
68'0
67'7
60'7
68'3
47'0
44'0
47'7
50'0
56'0
65'0
(2)
Body
weight
g/24 h) (kg)
Dietary
&rotein
4·0
4·0
3·5
4·5
2·5
3-3
3·6±0·3
20·2
20·2
36·8
13-8
22-8±4'3
25-4± 3-3
26,5
41,5
37·8
15·2
10·7
31·3
27·2±4·6
(3)
(mmol/I)
(5)
Urea
space
(% of
body wt)
=
53·0
59
34,7
62
48·7
74
41·4
64
46·4
71
37·5
59
43·6±2-6 64·8±2·4
52·2
76
41·5
61
41·0
67
47·0
68
45-4±2'3 68'0±2'7
38·2±2·8 65·3±2·5
29·9
60
23-1
53
34·2
72
27·4
54
43·0
76
43-1
66
33'5±3'1 63'5±3'5
(4)
(I)
Urea
space
Plasma urea
concentration
=
212
139
170
186
ll6
124
158±14
1054
838
1509
649
1013±160
932±ll2
792
959
1293
416
460
1349
878±149
(6)
Urea pool
(mmol)
2650
3025
3286
3246
3279
2150
2939±170
571
721
217
742
563±105
362±72
166
136
130
248
446
238
227±44
(7)
0·00
0·00
+2·00
+3095
+0·50
+1·03
-0,37
-0,33
-0,40
+0·30
-0,46
-0'73
-0,22
-0'38
-0,20
+0'50
(8)
pool
(mmol/h)
in urea
(ml/h)
Change
Urea
clearance
=
=
-0·0543±0·0027
-0,0923 ±0·0028
-0·1026±0·0240
-0·ll05±0·0286
-0·1028±0·0161
-0·07l8±0·0071
-0,0891 ±0'0081
-0'0138±0'0005
-0,0257 ±0·0009
-0,0085 ±0·0004
-0·0244±0·0035
-0,0181 ±0·0036
-0·0136±0·0022
-0·0076±0·0003
-0·0084±0·0002
-0·0056±0·0002
-0·0177±0·0015
-0,0161 ±0·0010
-0·0079±0·0003
-0·0106±0·0019
(9)
4·4
4·8
4·8
309
4,6
7·9
5·1±0·5
(ll)
ll·5
10·6
\2·8
\2'1
17·4
ll·5
20·6
14·6
ll·9
8·2
8·9
7·1
1309±\06 10·7±1·0
14,5
12·2
21,5
15·1
\2·8
7·7
15·8
n-z
16·2±1·6 1\06±1·3
IH±I'5 7·7±1·2
6,0
8·1
7·2
7·4
7·4
10·7
7·8±0·6
(10)
92
95
66
71
69
80
79±5
84
70
60
71
7I±4
67±3
73
59
67
53
62
74
65±3
(12)
rate
0·9
0·7
309
2·1
3·2
2-8
2-3±0'5
2·7
6·7
5·5
4·3
4'8±0'7
3·7±0·4
2·1
4·0
2·6
309
3·0
2·3
3·0±0·3
(13)
8
5
22
10
27
31
17±4
19
31
43
27
30±4
36±3
35
49
36
53
41
21
39±4
(14)
rate
Urea metabolized
[I4C)Urea specific
Urea excreted
Urea
radioactivity rate synthetic
Xof
rate
(mmo /h)
constant±sE
%of. (mmol/h)
synthetic
synthetic
(mmol/h)
(fraction/h)
=
Urea clearance values are calculated as mean values of 24 h collection periods throughout the course of study except in normal subjects, where a single urea clearance
was calculated over the course of the [14C]urea study. Group A: uraemic patients on a low protein diet. Group B: uraemic patients on a normal protein diet. Group C:
normal subjects. Column (6)
(3) x (4). Column (10)
(9) x (6). Column (12) 100 x (11)/00). Column (13) (10)-(11)-(8). Column (14) 100 x (13)/(10).
~
00
Vl
V>
0:;'
~
~
sS·
::::
5=
l:l
:::
~
....
::s
::;
~
~
-.
l:l
~
::::
....
~
o'
;::
l:l
-
s-.
Group C
S.H.
R.Y.
A.L. (I)
P.B. (1)
A.L. (2)
P.B. (2)
Mean C±SEM
GroupB
J.A.
C.P.
S.S.
F.S.
Mean B±SEM
Mean A+B±SEM
Group A
E.M.
E.O.
J.G.
RC.
M.R
J.S.
MeanA±sEM
umn
(9)
(6)
=
=
70
70
70
70
30
30
70
70
70
70
30
30
30
30
30
30
(I)
45
48
47
48
46
47
47±0·4
30
32
30
40
33±2·1
35±H
38
33
33
35
38
38
36±0·9
(2)
117
106
127
106
124
106
96
83
72
104
106
69
83
77
95
125
(3)
80
80
80
50
70
50
100
90
90
90
90
80
70
60
60
100
(4)
(h)
Protein Concn, of Intravenous
Linear
albumin
content plasma
pool
incorporation
of diet albumin
(g)
interval
(g/I)
(g/24 h)
Col
umn
m
(9)
100
(13) x IoiiO
=
Column (14)
6'0
17-8
9.1
7'1
18'7
13'7
17-3
13'7
36·7
14'0
35·2
261'0
27·0
84·5
12'3
8·8
8·0
21·5
13·2
8·6
26·5
16·6
18·9
13·0
30·2
16·7
42-8
206'7
25-4
74·4
13-3
12-6
8·0
14'1
16'0
7·9
14·7
7'8
8'1
7·3
7·7
7-7
7·5
30'8
7·3
7·6
8·4
7·9
0·10
0'15
0·08
0·11
0·18
0·21
0·14±0·02
0·23
0·18
0·39
0·22
0·26±0·04
0·39±0·08
0·57
0·67
0·35
0·98
0·16
0'16
0'48±0'12
5·3
5'2
3·4
8·2
6'0
10·4
6·4±0·93
48·5
33·5
130·8
31-7
6H±20'4
83-8±15'7
100·3
160·6
129·3
135·9
24·5
43'2
99·0±20·2
604
235
450
403
654
472
470±56
456
468
380
364
417±23
369±29
440
281
194
294
316
496
337±41
3-6
6'8
2-8
5'2
4·6
7'5
5'4
5·4±0·6
5'2
5'4
4'3
4'2
4·8±0·3
4·2±0·3
5'6
3-8±0'5
5'0
3'2
2'2
3'4
0·08
0'19
0·07
0·18
0·08
0·19
0·13±0·02
0·94
0·62
3'04
0'75
1'34±0'49
2'37±0'60
2·01
5·02
5·88
4·00
0·68
0·77
3'06±0'83
eo
1·8
1·4
7.8
4·2
6·4
5·6
4·5±1·1
5·4
13·4
11·0
8·6
9·6±1·5
7·4±0·9
4·2
8·0
5·2
7-8
6·0
4·6
6·0±0·6
0'29
0'37
0·04
0·20
0·09
0'19
0·20±0·05
0·90
0·25
1·19
0·37
0·68±0·19
1·27±0·25
2·39
2'01
2'49
1·74
0·41
0·94
l-66±0'31
104 X
Total ISN Injected Proport on
Rate of
Albumin N required Proportion N available Efficiency of
Albumin incorporated dose of ofi njected
urea N
synthetic
for
of total N
from urea
utilization
enrichment
(Jlmol)
ISN
ISN
utilization
rate
synthesis
required
degradation of available
achieved
(mmol) incorporated for albumin (mgfh)
(mmol/h) provided by (mmol/h)
N
(atoms %
into albumin synthesis
urea N
co
excess)
(%)
(pmol/h)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
(13)
(14)
Utilization of urea N for albumin synthesis
2xcolumn (13) in Table I
(9)
(11) x 10
=
=
(10)
= 6'25 x 14
Column (13)
Column (12)
Column (11)
Deconvoluted [lsN]albumin data
(8) urea N pool x 1000
100 x
(4)
(3) X (5)
6'25x14
(6)
Column (8) =
x 100
Col
Calculations
Group A: uraemic patients on a low-protein diet. Group B: uraemic patients on a normal protein diet. Group C: normal subjects.
2. Albumin metabolism data in uraemic patients and normal subjects: utilization of urea nitrogen for albumin synthesis based upon incorporation of lSN after
intravenous administration of [ 1 SN lurea
Patient
TABLE
w
r-
'"...'"
l::l
<;)
~
~
~
00
0\
Utilization of urea nitrogen for albumin synthesis
(Table 1, column 3) and the urea metabolized as a
percentage of the urea synthetic rate (Table 1,
column 14) (r = 0'56, P<0'05, n = 16). However,
the absolute amount of urea metabolized (Table
1, column 13) failed to show a significant positive
correlation with the plasma urea concentration
(Table 1, column 3) (r = 0'33, P>O·I, n = 16).
Finally, the only other significant correlation noted
was in chronic renal failure between urea synthetic
rate (Table 1, column 10) and urea metabolic rate
(Table 1, column 13)(r = 0'76, P < 0,05, n = 10).
Albumin metabolism (Table 2)
Plasma albumin concentration was lower in
both groups of uraemic patients than in the normal
subjects (Table 2, column 2). Since all subjects
were in a steady albumin metabolic state, the rate
of 12'I-Iabelled albumin degradation could be
equated with synthesis, which was generally lower
in uraemic patients, but not invariably so even on
a reduced protein intake (Table 2, column 10).
It was apparent from the sequential measurements
of [1 'N]albumin enrichment (Fig. 4) that biosynthetically labelled albumin could be detected in
plasma as early as 2 h in normal subjects and 3 h
in chronic renal failure. In all groups there were
examplesoflatent periods of 3-4 h, and in one patient
(J.A.) there was a long latent period of 23 h before
[1'N]albumin was detected (samples at 4 and 8 h
being below the limits of detection, i.e, < 0·0001
atoms % excess).
As indicated in the Materials and methods section,
both the dynamics and magnitude of l'N incorporation into albumin are derived by deconvolution of
the [1'N]albumin enrichment curve, thereby permitting direct comparison between the rate of release
of urea nitrogen by metabolism and the rate of
utilization of this nitrogen for albumin synthesis.
The cumulative albumin enrichment achieved over
the linear period of labelling is shown in column
5 of Table 2. Since deconvolution provides the
plasma albumin enrichment that would have been
achieved had there been no extravascular distribution
of labelled albumin, the total incorporation of
l'N within this period (Table 2, column 6) can be
obtained from the product of enrichment (column
5) and the plasma albumin pool (column 3). Since
the injected dose of l'N is known (Table 2, column
7), the percentage of this dose which appeared in
albumin can be derived (column 8). On the assump-
387
tion that after distribution through body water the
injected [1'N]urea behaves identically with endogenous urea, the absolute rate of urea nitrogen
utilization for albumin synthesis can be calculated
(Table 2, column 9). This can then be expressed
as the percentage of the total nitrogen requirements
for the measured rate of albumin synthesis (Table 2,
column 12) and as a percentage of the rate of urea
nitrogen released by metabolism, the latter being
an indication of the efficiency of utilization of
available urea nitrogen by the liver (column 14).
In spite of the greatly expanded urea pool in
groups A and B, with its resultant dilution of enrichment of precursor [1'N]urea, the percentage of
the injected dose incorporated was greater in most
cases than in group C (Table 2, column 8). When
this factor is taken into account in calculation of
the rate of urea nitrogen utilization (Table 2, column
9) the differences between patients with chronic
renal failure and normal subjects become more
apparent, with uraemic patients using very much
more (mean 83·8 pmol/h±sEM 15'7, n = 10) than
normal subjects (mean 6·4 Jlffiol/h ± SEM 0,93,
n = 6). On average, in uraemic subjects a lowprotein diet seemed to promote utilization of
urea nitrogen. In both normal subjects there was a
higher utilization on the 30 g than on the 70 g protein
diet.
The contribution of urea nitrogen to the nitrogen
required for albumin synthesis (Table 2, column
12) in normal subjects was small. However, urea
nitrogen forms a consistently higher fraction of
albumin synthetic requirements in uraemic patients
(up to 80-fold higher), the highest values being seen
in uraemic patients on protein restriction. However,
it is noteworthy that protein restriction produced
no effect on the provision of these requirements
in the two normal subjects who were studied twice.
The highest efficiency of utilization of available
urea nitrogen was seen in group A, and although
there were individual exceptions, efficiency was
higher in groups A and B than group C. Of the two
subjects studied twice, one (P.B.) showed no change
in efficiency of utilization after a short period of
protein restriction; the other (A.L.) doubled the
efficiency but still showed the lowest value seen
in any study.
No significant positive correlation was seen between the rate of urea nitrogen utilization for albumin synthesis (Table 2, column 9) and the amount of
nitrogen available from urea degradation (Table
388
R. Varcoe et al.
2, column 13) (r = 0'28, P>O'I, n = 16), in spite of
the previous finding of a significant correlation
between urea synthesis and urea degradation. When
expressed in terms of efficiency of utilization of
released nitrogen, the highest positive correlation
existed between efficiency (Table 2, column 14)
and the absolute rate of utilization (Table 2, column
9) (r =0'90, P < 0'001, n = 16). In contrast, negative
correlations were apparent between both efficiency
and the absolute rate of utilization of urea nitrogen
and the total nitrogen utilized for albumin synthesis (Table 2, column 11) (r = -0,53 and -0'54
respectively, P < 0'05, n = 16).
Discussion
Urea metabolism
Calculation of the parameters of urea metabolism
from the dynamics of intravenously administered
isotopically labelled urea has been well established
in previous studies (Walser & Bodenlos, 1959;
Jones, Smallwood, Craigie & Rosenoer, 1969;
Robson, Kerr & Ashcroft, 1968; Deane, Desir &
Umeda, 1968; Scholz, 1968; Walser, 1970). Calculations in the present study were based upon
observations over at least one half-life of the injected
urea. In all subjects the urea space was consistent
with the expected total body water and with the
data of Scholz (1968), who demonstrated that there
was no increase in the volume of urea distribution
in renal failure and that the urea space was nearly
identical with the volume of total body water
measured with tritiated water. As expected the
['4C]urea plasma half-life was considerably longer
than normal in chronic renal failure (mean tt 65'5 h).
The values in normal subjects (mean 8·3 h) were
similar to the 7 h mean value observed by Walser
& Bodenlos (1959) and the 6·7 hand 9·2 h values
recorded by Jones et al. (1969). Likewise, the mean
percentage of urea synthesized which was metabolized by uraemic patients (39%) is similar to the rate
of urea degradation reported by Robson (1964) and
Deane et al. (1968), but considerably less than the
average of 70% for urea degradation as percentage
of synthesis recently reported by Walser (1974) in
uraemic patients receiving essential amino acids or
keto acid analogues of five essential amino acids.
The higher percentage degradation seen in patients
and normal subjects receiving a low-protein intake is
predominantly the result of a relative reduction in
urea synthesis. Nevertheless, in absolute terms there
is more urea degraded in chronic renal failure than in
normal subjects regardless of the dietary protein,
although the difference was not as great as that described by Jones et al. (1969), who claimed that the
enterohepatic circulation of urea nitrogen was at least
twice that of normal subjects. Although uraemic
patients on a normal protein intake tend to degrade
more urea than those on dietary protein restriction,
no consistent effect was noted in the two normal
subjects when placed on 30 g of protein for 9 days.
This phenomenon was evident in chronic renal
failure in spite of the fact that those able to tolerate
a normal protein intake had on average a lower
plasma urea concentration.
The estimation of the quantity of urea nitrogen
recycled is dependent on the assumption that
synthesized urea not appearing in the urine has
been hydrolysed in the gut and not lost by other
routes. Although there is no detectable urea in the
faeces of normal individuals it might be expected
that there would be an increased faecal loss in
uraemia. However, Wilson, lng, Metcalfe-Gibson
& Wrong (1968) have shown that there was no
detectable faecal urea in two-thirds of their patients
with renal failure, with a mean faecal dialysate
urea concentration of only 6% of the blood urea
concentration in the whole group. Evidence that
urea hydrolysis in the gut is the main extra-renal
source of urea clearance has been provided by the
observation that antibiotic treatment can virtually
eliminate degradation (Walser & Bodenlos, 1959).
Although urea diffuses readily through most body
tissues, the wall of the colon is relatively impermeable
(Billich & Levitan, 1969), and colonic bacterial
ureases would therefore have to act upon urea
delivered intraluminally from the small intestine.
However, Gibson, Sladen & Dawson (1973),
on the basis of studies in patients with ileostomies
or excluded colonic loops, suggest that a significant
proportion of urea breakdown can occur in the
small bowel. Since urea is readily diffusible into the
small intestine, and faecal bacterial urease activity
is increased in uraemia (Brown, Hill & Richards,
1971), and as mentioned uraemic patients tend to
have on average a higher rate of urea metabolism,
it would be expected that the latter would correlate
with plasma urea concentration. The fact that our
results, like those of Robson (1964) and Walser
(1974), fail to show such a correlation emphasizes
the conclusion that urea degradation must be in-
Utilization 0/ urea nitrogen for albumin synthesis
dependent of diffusion and that some component
of the chronic uraemic state leads to a relative
reduction in gastrointestinal urea hydrolysis. The
finding of a significant correlation between plasma
urea concentration and the fraction of synthesized
urea which is metabolized must be interpreted with
caution because of the apparent over-riding importance of dietary protein in regulating the urea synthetic rate. There is no evidence that factors exist in
chronic renal failure which simultaneously increase
urea breakdown and decrease urea synthesis.
Urea nitrogen reutilization
The extent of reutilization of ammonia nitrogen
released by urea degradation is indicated by two
sets of data. First, the correlation between the
rate of urea synthesis and the rate of urea degradation suggests that urea synthesis may be promoted
by ammonia returning to the liver in portal blood.
Secondly, a direct indication of the extent of utilization of urea nitrogen for albumin synthesis is obtained by the present method for provision and
analysis of [1sN]albumin incorporation data.
Although 1SN recycling may be perpetuated by
exchange phenomena, e.g. transamination reactions
as suggested by Jones, Craigie, Koj & McFarlane,
(1967), the pathways of initial fixation of ammonia
nitrogen are probably those which are available
for reutilization of endogenous urea nitrogen in the
process of net nitrogen gain. Fixation of ammonia
nitrogen into arginine, glutamic acid and glutamine
and aspartic acid represents a potential net contribution to nitrogen balance.
The advantages of the present approach to the
quantification of the extent of urea nitrogen reutilization are twofold. First, intravenous administration of [lsN]urea together with [l4C]urea as a
single injection permits equilibration with endogenous urea. Thereafter, the fate of both isotopic
forms of urea is a valid indication of endogenous
metabolism, and in particular of intestinal hydrolysis (Wolpert, Phillips & SummerskiII, 1971).
Secondly, the simultaneous intravenous administration of 12sI-labelled albumin, followed by frequent blood sampling, permits deconvolution of an
accurately defined plasma [l sN]albumin enrichment
curve, thereby establishing with a precision that
has not been possible previously the time-course
of incorporation of 1 sN into total exchangeable
albumin. In this way it has been possible to make
389
a quantitative assessment of the fate of endogenous
urea which has left the circulation to be hydrolysed
in the lumen or mucosa of the intestine. From measurements of the rate of turnover of albumin the
extent to which recycled urea nitrogen fulfils the
requirements for albumin synthesis has been determined, and from the rate of metabolism of [l4C]urea
the efficiency of utilization of urea nitrogen for
this purpose has been calculated.
As with the extent of urea hydrolysis, the incorporation of urea nitrogen into albumin does not
appear to be dependent upon the urea concentration
iii. body water. Although the provision of nitrogen
requirements for albumin synthesis and the efficiency
of utilization of released urea nitrogen for this
purpose was on average greater in patients with
renal failure than normal, the reason for these
observations is not immediately evident. Although
all measured albumin synthetic rates (expressed per
unit body weight) were within the normal range
of 135-260 mg day" ' kg- 1 (Dykes, 1968), the significant inverse relationship between nitrogen requirements for albumin synthesis and the efficiency of
utilization of available urea nitrogen are evidence
for. a possible hepatic mechanism for adjusting
the reclamation of urea nitrogen. The level of
dietary protein may be a major factor in such a
mechanism since in dietary protein restriction
urea nitrogen will form a greater proportion of
total intraluminal nitrogen. Any suggested mechanisms must take into account the fact that the patients
with the lowest urea clearance were those requiring
restriction of dietary nitrogen and as such were
also the ones whose fractional urea turnover was
the slowest. These two factors, namely prolonged
protein restriction and the fact that urea has an
enhanced survival time in body water, may be the
key mediators of long-term adaptation in nitrogen
economy in chronic renal failure. Evaluation of
their relative importance will require further observations of the effects of each studied in isolation.
Whichever factor is predominant it appears that
in chronic uraemia under the circumstances studied
the utilization of urea nitrogen for albumin synthesis
is of minor nutritional significance. Because of the
relatively high contribution of albumin to overall
hepatic protein synthesis and the predominant
enterohepatic route of ammonia nitrogen fixation,
it is unlikely that the utilization of urea nitrogen
for the synthesis of other hepatic or extrahepatic
proteins is of any greater proportional significance.
390
R. Varcoe et al.
Acknowledgments
This project was approved by Northwick Park
Hospital Ethical Committee (Project E. 142). R.V.
was supported at the Clinical Research Centre
by a Fellowship from The Wellcome Trust. We
thank Dr G. Lewis of the Edgware General Hospital
and Dr J. Moorhead of the Royal Free Hospital,
who allowed us to study individual patients under
their care, Miss P. Hulme for expert dietetic supervision, our colleagues at Northwick Park Hospital
who volunteered to act as control subjects for these
studies and Dr A. S. McFarlane for assistance in
the preparation of radioiodine-labelled albumin
and for valuable discussion.
References
BILLICH, C.O. & LEVITAN, R. (1969) Effects ofsodium concentration and osmolality on water and electrolyte absorption
from the intact human colon. Journal of Clinical Investigation, 48, 1336-1347.
BROWN, C.L., HILL, M.J. & RICHARDS, P. (1971) Bacterial
ureases in uraemic men. Lancet, ii, 406408.
DEANE, N., DESIR, W. & UMEDA, T. (1968) The production
and extra-renal metabolism of urea in patients with chronic
renal failure treated with diet and dialysis. In: Dialysis and
Renal Transplantation. Proceedings of the 4th Conference
of the European Dialysis and Transplant Association, pp.
244-249. Ed. Kerr, D.N.S. Excerpta Medica Foundation,
New York.
DOUMAS, B.T., WATSON, W.A. & BIGGS, H.M. (1971)
Albumin standards and the measurement ofserum albumin
with bromcresol green. Clinica Chimica Acta, 31, 87-96.
DYKES, P.W. (1968) The rates of distribution and catabolism
of albumin in normal subjects and in patients with cirrhosis
of the liver. Clinical Science, 34, 161-183.
GIBSON, J.A., SLADEN, G.E. & DAWSON, A.M. (1973) Studies
on the role of the colon in urea metabolism. (Abstract).
Gut, 14, 816.
GIORDANO, C. (1963) Use of exogenous and endogenous
urea for protein synthesis in normal and uraemic subjects.
Journal ofLaboratory and Clinical Medicine, 62,231-246.
GIORDANO, C., DE PASCALE, C., BALESTRIERI, C., OTTADlNI,
D. & CRESCENZI, A. (1968) Incorporation of urea lsN
in amino acids of patients with chronic renal failure on low
nitrogen diets. American Journal of Clinical Nutrition,
21, 394-404.
GIOVANETTI, S. & MAGGIORE, Q. (1964) A low nitrogen diet
with proteins of high biological value for severe chronic
uraemia. Lancet, i, 1000-1003.
JONES, E.A., CRAIGIE, A., KOJ, A. & McFARLANE, A.S.
(1967) Ammonia nitrogen and protein synthesis. (Letter),
Lancet, ii, 989.
JONES, E.A., CRAIGIE, A., TAVlLL, A.S., SIMON, W. & ROSENOER, V.M. (1968) Urea kinetics and the direct measurement of the synthetic rate of albumin utilising [14C]
carbonate. Clinical Science, 35, 553-564.
JONES, E.A., SMALLWOOD, R.A., CRAIGIE, A. & ROSENOER,
V.M. (1969) The enterohepatic circulation of urea nitrogen.
Clinical Science, 37, 825-836.
KORNER, A. & DEBRO, J.R. (1956) Solubility of albumin in
alcohol after precipitation by trichloroacetic acid: a
simplified procedure for separation of albumin. Nature
(London), 178, 1067.
McFARLANE, A.S. (1956) Labelling of plasma proteins with
radioactive iodine. Biochemical Journal, 62, 135-143.
McFARLANE, A.S. (1958) Efficient trace labelling of proteins
with iodine. Nature (London), 182, 53.
MATTHEWS, C.M.E. (1957) The theory of tracer experiments
with 1311-labelled plasma proteins. Physics in Medicine
and Biology, 2, 36-53.
READ, W.W.C., McLAREN, D.S., TCHALIAN, M. & NASSAR,
S. (1969) Studies with lsN labeled ammonia and urea
in the malnourished child. Journal of Clinical Investigation,
48, 1143-1149.
REGOECZI, E., IRONS, L., KOJ, A. & McFARLANE, A.S.
(1965) Isotopic studies of urea metabolism in rabbits.
Biochemical Journal, 95, 521-532.
RICHARDS, P., METCALFE-GIBSON, A., WARD, E.E., WRONG,
O.. & HOUGHTON, BJ. (1967) Utilisation of ammonia
nitrogen for protein synthesis in man, and the effect of
protein restriction and uraemia. Lancet, ii, 845-849.
ROBSON, A.M. (1964) Urea metabolism in chronic renal
failure. M.D. thesis, University of Newcastle-upon-Tyne.
ROBSON, A.M., KERR, D.N.S. & ASHCROFT, R. (1968) Urea
metabolism in chronic uraemia. In: Nutrition in Renal
Disease, pp. 71-85. Ed. Berlyne, G.M. Livingstone,
Edinburgh.
SCHOLZ, A. (1968) Investigations on the distribution and
turnover rate of 14C-urea and tritiated water in renal
failure. In: Proceedings of the 4th Conference of the European Dialysis and Transplant Association, pp. 240-244.
Ed. Kerr, D.N.S. Excerpta Medica Foundation, New York.
SNYDERMAN, S.E., HOLT, L.E., DANCIS, J., ROITMAN, E.,
BOYER, A. & BALIS, M.E. (1962) 'Unessential' nitrogen:
a limiting factor for human growth. Journal of Nutrition,
78,57-72.
TAVlLL, A.S., CRAIGIE, A. & ROSENOER, V.M. (1968) The
measurement of the synthetic rate of albumin in man.
Clinical Science, 34, 1-28.
VARCOE, R., HALLIDAY, D. & TAVILL, A.S. (1974) Utilization
of urea nitrogen for albumin synthesis in the stagnant
loop syndrome. Gut, 15, 898-902.
VITEK, F., BIANCHI, R. & DONATO, L. (1966) The study of
distribution and catabolism of labelled serum albumin
by means of analog computer technique. Journal of
Nuclear Biology and Medicine, 10, 121.
WALSER, M. (1970) Use of isotopic urea to study the distribution and degradation of urea in man. In: Urea and the Kidney, pp, 421-429. Ed. SChmidt-Nielsen, B. Excerpta Medica,
Amsterdam.
WALSER, M. (1974) Urea metabolism in chronic renal failure.
Journal of Clinical Investigation, 53, 1385-1392.
WALSER, M. & BODENLOS, L.J. (1959) Urea metabolism in
man. Journal of Clinical Investigation, 38, 1617-1626.
WALSER, M., GEORGE, J. & BODENLOS, L.J. (1954) Altered
proportions of isotopes of molecular nitrogen as evidence
for a monomolecular reaction. Journal of Chemical
Physics, 22, 1146.
WILSON, D.R., lNG, T.S., METCALFE-GIBSON, A. & WRONG,
O.M. (1968) The chemical composiion of faeces in uraemia as revealed by in vivo faecal dialysis. Clinical Science,
35, 197-209.
WOLPERT, E., PHILLIPS, S. & SUMMERSKILL, W.H.J. (1971)
Transport of urea and ammonia production in the human
colon. Lancet, ii, 1387-1390.