in vivo studies of mucosal-serosal transfer in rat jejunum
TRANSCRIPT
Naunyn-Schmiedeberg's Arch Pharrnacol (1985) 329:70-76 Naunyn-Schmiedeberg's
Archives of Pharmacology �9 Springer-Verlag 1985
In vivo studies of mueosal-serosal transfer in rat jejunum D. Winne, H. Giirig, and U. Miiller
Abteilung fiir Molekularpharmakologie, Pharmakologisches Institut der Universit/it, Wilhelmstrasse 56, D-7400 Tfibingen 1, Federal Republic of Germany
Summary. In anesthetized rats, the appearance rates of a series of labeled substances in jejunal venous blood (~B) and serosal bath (~s) were measured in vivo (intestinal blood flow rate 1.5 ml rain-1 g-1) after intraluminal administra- tion of 0.5 ml buffer solution (initial concentration 1 mmol/l or 1 GBq/1 tritiated water) into a closed jejunal segment (length 4 - 5 c m ) . Between 32% (erythritol) and 93% (salicylic acid) of the administered activity (unchanged sub- stance and possible metabolites) appeared in the intestinal venous blood within 60 rain. The fraction recovered from the serosal bath after 15 (60) rain was 11 (6) % for tritiated water, 7 (4) % for aniline, 3 (7)% for aminopyrine, 5 (4) % for butanol, 3 (3) % for benzyl alcohol, 2 (4) % for benzylamine, 1 - 2 % for benzoic acid, theophylline, methyl- e-D-glucopyranoside, L-lysine, antipyrine, and urea, and less than 1% for L-phenylalanine, D-galactose, erythritol, and salicylic acid. During single pass perfusion of a jejunal seg- ment (length 3 - 4 cm) the fraction of serosal transfer 4~s/ (4~B + q~S) was 19% for tritiated water, 4.9% for antipyrine, 0.5% for benzoic acid, and 0.08% for salicylic acid.
Distension of the intestinal wall by administration of I ml buffer solution instead of 0.5ml increased the appearance rate of benzoic acid and antipyrine in intestinal venous blood by a factor of 2 and serosal transfer by a factor of approximately 3. Reduction of blood flow rate from 1.26 to 0.91 ml rain- 1 g- 1 decreased the fraction of antipyrine recovered in the intestinal venous blood within 60 rain from 84% to 72% and increased the fraction transferred into the serosal bath from 2.3% to 9.1%. A high mucosal-serosal transfer in rat small intestine with intact blood supply can be expected for substances with high epithelial permeability and low binding to plasma proteins.
Key words: Intestinal absorption - Serosal transfer - In- testinal blood flow - Distention of intestinal wall
Introduction
In previous experiments (Winne and Remischovsky 1971 a, b; Winne 1972, 1973; Lichtenstein and Winne 1973, 1974) dealing with the relationship between intestinal absorption and blood flow in rat jejunum, it was observed that the appearance rate in the intestinal venous blood is lower than the disappearance rate from the intestinal lumen. Ochsenfahrt (1971, 1979) showed that, in the vascularly
Send offprint requests to D. Winne at the above address
perfused, isolated jejunum of the rat, a fraction of absorbed antipyrine (20%), urea (16%), and salicylic acid (2%) appears on the serosal surface in spite of sufficient blood supply. Thus, the difference between disappearance and appearance rate can be explained, at least partly by serosal transfer.
The following experiments were performed to measure under conditions of natural arterial blood supply the serosal transfer of a series of substances with different properties: lipophilic and hydrophilic as well as passively and actively transported substances, acids, bases and neutral compounds and substances with high protein binding. The substances used were tritiated water, urea, erythritol, butanol, benzyl alcohol, aniline, antipyrine, aminopyrine, benzylamine, theophylline, L-lysine, L-phenylalanine, D-galactose, methyl- e-D-glucopyranoside, benzoic acid and salicylic acid. The results indicate that the fraction leaving the intestinal wall at the serosal side, in spite of undisturbed blood flow, can reach 20%. Moreover, distension of the intestinal wall and reduction of the blood flow rate increases serosal transfer.
Methods
Animals. Conventionally bred male Wistar rats (FW Bi- berach strain, mean weight 350g, SD 15g) were fed Altromin (13244) and water ad libitum. Food was withheld 1 6 - 2 0 h before the experiments.
Preparation of jejunal segment. Full details are given in Winne and Remischovsky (1971a). The rats were an- esthetized with urethane (4.5 ml/kg i.p., 25% solution). The jejunal segment (distance from duodenojejunal flexure 19 em, SD 8 cm; length 4.5 cm, SD 0.9 cm; wet tissue weight 290 mg, SD 70 rag; N = 302) was separated by ligating it from the neighboring segments so that venous blood drained by a "single branch of the superior mesenteric vein. After injection of heparin (2 mg in 0.1 ml isotonic saline solution) into the jugular vein, the jejunal vein was punctured and the outflowing blood collected and weighed. In addition, blood flow rate was monitored by drop recording. Blood loss was compensated by infusion of fresh heparinized rat blood into a jugular vein. The animals were ventilated through a tracheal tube. Blood pressure was continuously measured by means of a cannula in the carotid artery. After injection of 0.5 or 1.0 ml of buffer solution containing the labeled substance into the jejunal segment, the segment was placed into a warmed (38 ~ C) serosal bath filled with 15 ml isotonic saline solution which was recirculated by a pump (rate 10 ml/
71
rain). The evaporated water was substituted at intervals of 10-15 min. In the experiments with tritiated water, butanol, benzyl alcohol, benzylamine, and aniline the solution in the serosal bath (volume 8.5 ml) was mixed by a rotating rod (4,800 rpm, 5 mm diameter), since these substances pen- etrate into the tube wall of the recirculation pump. In the experiments with single pass perfusion an isotonic phosphate buffer (pH 7.4) was added to the serosal bath. At the end of the experiment, the jejunal segment was excised and its length determined. The segment was then opened, gently squeezed out, and weighed. Also the intestinal contents were weighed. In the series with single pass perfusion, cannulae were tied into the jejunal segment. In- traluminal perfusion was started at 10 ml/min to fill the segment and then continued at 0.1 ml/min. The perfusion solution was warmed to 38 ~ C by passing a heat exchanger.
Experimental protocol. After injection of approximately 0.5 or 1.0 ml solution (exact volume determined by weighing syringe before and after injection) into the intestinal seg- ment, intestinal venous blood was collected in either three or twelve 5 min periods (total collection time 15 or 60 min). Samples were taken from the serosal bath every 5 or 15 min, respectively. After starting the luminal perfusion, intestinal venous blood was collected in six 5 min periods (total collec- tion time 30 min). Samples were taken from the serosal bath every 5 min.
Determination of thickness of intestinal wall. In one an- esthetized rat two neighboring jejunal segments of equal length (3.5 cm) were ligated and filled with either 0.5 or 1.0 ml Bouin's solution in situ. After 5 rain the segments were excised and placed for 24 h in Bouin's solution. Then the segments were opened, washed several times in 70% ethanol + 3% ammonia. After washing in water for several minutes the segments were cut parallel to their longitunal axis with a razor blade. The slices with thickness of about 0.5 mm were stained with Mayer's hemalum solution for 5 rain, washed in tap water, placed in 50% glycerol, and finally in 89% glycerol. The distances from the serosal surface to the bottom of the intervillous space and to the basal border of the crypts were measured microscopically (magnification 100 x ) at 60 different points.
Substances and solutions. [Dimethylamine- 14C]aminopyrine ' batch 21, 1.11 GBq/mmol; IN-methyl- ~ 4C]antipyrine, batch 9, 1.89 GBq/mmol; benzyl alcohol[carbinol-14C], batch 11, 203.5 MBq/mmol; (7-14C)benzylamine hydrochloride, batch 10, 2.21 GBq/mmol; [carboxyl-14C]benzoic acid, batch 38, 2.07 GBq/mmol; [U-14C]erythritol, batch 36, 3.7 GBq/mmol; D-[1-14C]galactose, batch 17, 111 MBq/ mmol; methyl(e-D-[U-14C]gluco)pyranoside, batch 25, 10.32 GBq/mmol; [carboxyl-14C]salicylic acid, batch 19, 2.18 GBq/mmol; [8-~4C]theophylline, batch 10, 925 MBq/ mmol; [14C]urea, batch 56, 2.01 GBq/mmol; tritiated water, batch 7, 185 MBq/ml (Radiochemical Centre, Amersham, GB). [14C(U)]Anilinehydrochloride, lot 952-098, 396 MBq/ mmol; N-[1-14C]butanol, lot 1757-058, 37 MBq/mmol (New England Nuclear, Dreieichenhain, FRG). L-(U-t4C)Lysine, lot 88-277, 8.55 GBq/mmol; L-[UlgC]phenylalanine, lot 90- 1178, 1.67GBq/mmol (Centre d'Etudes Nucleaires de Saclay, Gif-sur-Yvette, France).
L-Lysinemonohydrochloride, meso-erythritol, L-phenyl- alanine, D(+)galactose, urea for biochemical use; butanol,
benzyl alcohol p.a. (Merck, Darmstadt, FRG). Aniline- hydrochloride p.a., antipyrine purum, methyl-e-D-gtuco- pyranoside purissimum, cyclohexanesulfamic acid sodium salt purum (Fluka, Buchs, Switzerland). Theophylline (Sigma, St. Louis, MO, USA). Aminopyrine DAB 7 (Pharma-Zentrale, Herdecke, FRG). Salicylic acid purissimum (Serva, Heidelberg, FRG). Benzylamine 99%, Ega-Chemie, Steinheim, FRG.
Labeled and unlabeled substances (1 mmol/1) were dis- solved in buffer solutions. The isotonic injection and per- fusion solutions (pH 6.8) contained 5.03 mmol/1 KH2PO4, 4.97 mmol/1 Na2HPO4, 16.8 mmol/1 urethane, 109.5 mmol/1 NaC1, 39.8 mmol/1 cyclohexanesulfamic acid sodium salt with osmolality of 320 mosm/kg, corresponding to the osmolality of the plasma of rats anesthetized with urethane. The isotonic phosphate buffer solution (pH 7.4) contained 1.96 mmol/1 KH2PO4, 8.04 mmol/1 Na2HPO4, 16.8 mmol/1 urethane, and 149.4 mmol/1 NaC1.
Cyclohexanesulfamate (NaCHS) was added to preserve the luminal volume in the jejunal segment as constant as possible. An isotonic solution of NaCHS introduced into rat jejunum in situ causes a net flux of isotonic fluid (water, sodium, potassium, calcium, chloride, urea) into the lumen (Vogel and Stoeckert 1968). The concentration of NaCHS which gives rise to zero water flux was determined in a pilot study. 0.5 ml of buffer solutions (see above) with different concentrations of sodium cyclohexanesulfamate and sodium chloride (osmolality kept at 320 mosmol/kg) were injected into closed jejunal segments of anesthetized rats. The radio- activity of 14C-polyethylene glycol 4000 was measured in the injection solution and in the intestinal fluid after 60 min. The activity ratio (solution/fluid) amounted to (in paren- theses concentrations of NaCHS and NaC1 in mmol/1) 0.56 + 0.05 (10, 140), 0.79 + 0.06 (30, 120), 0.98.+ 0.04 (40, 110), 1.21 +0.06 (50, 100), 1.46+_0.05 (100, 50); N = 12 each. The water net flux was about zero when the solution with 40 mmol/1 NaCHS and 110 mmol/1 NaC1 was used. The weight of the intestinal contents at the end of the main experiments (Table 1 and 2) was mostly higher than the weight of the injected solution, since the intestinal fluid contained varying amounts of mucus. The partial substitu- tion of NaC1 by NaCHS influences the absorption of L-lysine, but not of antipyrine and benzoic acid. In ex- periments lasting 60 min carried out analogously to the others with administration of 0.5 ml solution into a jejunal segment the substitution did not change significantly the absorption (% B + S + W, % B, % S, %W, time course of appearance rate) of antipyrine and benzoic acid. The appearance rate of L-lysine was initially higher (by 25% in the 5 - 1 0 rain period) and after 20 min lower (by 40% in the 55-60 rain period), when the NaCHS-containing so- lution was used (0.001 < P < 0.01), while % B + S + W, % B, %S, %W were unchanged. Whether the CHS-ion in- terferes with the carrier transport of L-lysine was not inves- tigated further.
Analyses. The radioactivity was determined by liquid scintillation counting after quench correction (external stan- dardization calibrated by internal standardization with 14C-toluene). 10-100 Ixl solution were mixed with 4.5 ml Lumagel (Baker, Deventer, The Netherlands) or with 5 ml of the following scintillation mixture: 2 g PPO + 25 mg dimethyl-POPOP + 400 ml ethanol + 600 ml toluene. 250 ~tl blood were solubilized in 1 ml of a 1 : 1 mixture of
Tab
le 1
. Per
mea
tion
of
seve
ral l
abel
ed s
ubst
ance
s in
to je
juna
l blo
od a
nd s
eros
al b
ath.
Adm
inis
trat
ion
of 0
.5 m
l buf
fer
solu
tion
into
clo
sed
jeju
nal s
egm
ent o
f ra
t at
zer
o ti
me;
init
ial l
umin
al
-o
conc
entr
atio
n 1
reto
ol/1
or
i G
Bq/
1 (t
riti
ated
wat
er).
Sep
arat
e ex
peri
men
ts la
stin
g 15
rai
n (N
= 5
) an
d 6
0 m
in (
N=
10)
, res
pect
ivel
y
%B
+S
+W
%
B
%S
%
W
CB
C
w
Cs
WL
min
15
60
15
60
15
60
15
60
15
60
15
60
15
60
15
60
Ben
zoic
aci
d 79
.8
91.9
77
.9
90.6
1.
0 1.
2 0.
9 0.
13
25.2
0.
6 20
.1
2.0
0.30
0.
44
659
786
_+2.
0 +
0.7
_+
2.1
_+0.
8 -+
0.1
-+0.
i -+
0.2
+0.
02
__+1
.7
_+0.
1 -+
3.
0 _
+0
.2
-+0.
04
_+0.
04
_+38
-+
67
Sa
licyl
ic a
cid
77.5
93
.0
76.0
92
.6
0.5
0.34
1.
0 0.
12
23.7
0.
8 20
.3
1.9
0.15
0.
08
701
716
_+1.
6 +
1.5
_+
1.5
+1
.5
-+0
.1
_+
0.0
3
_+
0.2
-+
0.01
_+
1.2
+_0
.1
_+
1.9
_+
0.3
_
+0
.03
_+
0:01
_+
44
-+ 6
0 T
riti
ated
wat
er
77.4
87
.1
63.9
82
.3
10.8
6.
4 0.
7 0.
03
26.9
1.
0 15
.4
0.9
6.7
4.7
628
815
+2.
1 _+
1.4
__+1
.5
-+1.
6 -+
1.1
-+0.
7 -+
0.1
_+0.
02
-t-1
.7
+-0
.1
_+ 2
.1
___0
.2
_+0.
8 _+
0.3
_+40
_+
119
Ani
line
74
.2
89.3
66
.1
88.6
7.
1 3.
8 1.
2 0.
19
22.4
1.
2 22
.8
3.4
3.6
2.3
693
517
_+1.
9 +
0.3
-+2
,3
-+0
.5
_+0.
8 -+
0.3
+__0
.2
-+0.
01
_+1.
1 +
0.1
+_ 2
.1
+_0
.2
-+0.
3 _+
0.1
_+24
_+
30
Ben
zyl a
lcoh
ol
72.4
93
.2
66.0
89
.8
2.7
3.0
3.7
0.38
25
.5
1.3
58
5.5
1.60
1.
69
510
606
_+2.
6 -+
1.2
-+2
.9
_+1.
4 -+
0.3
-+0.
4 -+
0.3
__
_0.0
5 _+
0.9
__+0
.1
-+ 6
-+
0.5
-+0.
22
_+0.
21
_+34
_+
60
But
anol
68
.4
80.7
56
.9
78.1
4.
6 3.
5 6.
9 2.
0 28
.1
0.9
141
41.5
2.
3 1.
7 65
1 64
7 _+
1.9
-+0.
6 _+
2.0
-+1.
0 -+
0.6
_+0.
4 _+
0.8
_+0.
1 _+
1.9
-+_0
.1
_+12
-I
-2.1
_+
0.3
+_
0.2
+
41
-+ 3
9 T
heop
hylli
ne
67.9
91
.3
65.6
89
.9
0.9
1.2
1.1
0.20
27
.1
2.2
24.5
4.
2 0.
27
0.53
63
4 77
0 _+
2.2
_+0.
3 _+
2,2
+0
.2
-+0.
1 -+
0.1
_+0.
1 -+
0.02
_+
0.4
_+0.
1 _+
1.
7 +
0.4
_
+0
.04
_+
0.07
_+
14
_+
52
n-G
alac
tose
66
.9
92.2
43
.5
66.0
0.
5 0.
6 9.
4 1.
1 22
.5
2.9
188
128
0.21
0.
26
664
524
_+1.
7 +
1.0
-+
1.8
+
3.1
-+0.
1 +
0.1
_+0.
7 _+
0.1
_+_1
.2
-+0.
2 -+
6
-+7
+_0
.03
-+0.
05
-+59
-+
83
L-L
ysin
c 63
.4
86.8
56
.3
80.5
1.
1 0.
5 6.
0 5.
7 29
.6
1.4
116
ll0
0.
32
0.16
66
9 67
7 _+
0.8
-+0
.3
-+0
.5
-+0
.7
-+0.
2 _+
0.1
_+0.
6 _+
0.5
_+0.
8 _+
0.1
_+ 8
__
+7
-+0.
05
_+0.
05
-+24
-+
48
Ant
ipyr
ine
60.6
87
.1
57.6
83
.9
1.7
2.3
1.3
0.8
24.9
1.
5 25
.8
11.9
0.
48
1.0
680
478
-+1
.6
_+1.
2 _+
1,6
_+1.
3 _+
0.2
_+0.
3 _+
0.1
_+0.
1 _+
1.5
_+0.
2 _+
2.1
+
0.9
_
+0
.03
+
_0.1
+_
41
-+
56
Am
inop
yrin
e 58
.8
84.4
54
.7
76.7
2.
8 6.
6 1.
3 0.
52
27.5
3.
2 28
.7
7.4
0.8
2.1
677
791
_+0.
9 _+
1.0
_+0.
8 -+
1.1
+0.
3 -+
0.6
+0
.2
-+0.
05
-+0
.5
-+0
.3
_ 1.
2 -+
0.5
-+
0.1
+0
.2
_+29
_+
50
Met
hyl-
c~-n
-glu
co-
54.9
75
.1
56.2
89
.9
1.3
1.2
10.9
8.
5 29
.3
1.6
147
16.0
0.
65
0.58
56
4 50
0 py
rano
side
_+
1.6
_+3.
3 _+
1.9
_+1.
0 _+
0.1
-+0.
1 _+
0.6
_+0.
6 __
+1.2
_+
0.2
_+10
_
+1
.2
_+
0.0
9
_+0.
05
_+16
-+
_ 43
g-
Phe
nyla
lani
ne
39.5
58
.1
36,8
53
.9
0.6
0.6
2.5
3.6
16.8
1.
3 49
.7
66
0.16
0.
25
674
835
_+0.
8 +
0.4
-+
0.4
-+0.
4 -+
0.1
-+0.
1 +
0.3
-+0.
2 _-
4_-1
.0
+0.
1 _+
2.4
_+
4 _+
0.02
+
0.04
-+
14
__+
64
Ben
zyla
min
e 39
.1
80.7
24
,5
73.6
1.
9 3.
5 12
.7
3.6
22.3
6.
2 16
3 51
0.
97
1.98
49
3 59
2 _+
3.1
_+2.
3 _+
2.3
_+2.
6 -+
0.1
_+0.
3 _+
1.1
_+0.
5 _+
2.1
-+0.
4 -+
10
_+7
_+0.
05
-+0.
11
-+22
-+
48
Ure
a 33
.8
66.2
31
.6
64.2
1.
0 1.
7 1.
1 0.
32
23.6
5.
6 24
.3
7.3
0.28
0.
60
647
754
_+2.
0 +
1.0
+
_1.8
-t
-1.1
-+
0.2
-+0.
2 _+
0.2
_+0.
03
_+0.
8 -+
0.6
_+ 2
.7
__+
0.7
_+
0.0
5
_+0.
07
__+2
6 +
65
E
ryth
rito
l 6.
6 33
.4
5,9
31.5
0.
28
0.9
0.5
0.94
4.
8 6.
4 8.
9 18
.8
0.12
0.
31
699
569
+0
.4
_+2.
2 _+
0.4
+2.
1 +
0.0
4
-+0.
1 -+
0.1
-+0
.23
-+
0.2
_+0.
5 _+
0.6
_
+2
.0
+0.
02
-+0.
04
+2
9
_+ 2
3
% B
, %
S,
% W
=
Perc
enta
ge o
f ad
min
iste
red
acti
vity
rec
over
ed i
n in
test
inal
ven
ous
bloo
d,
conc
entr
atio
n in
int
esti
nal b
lood
, se
rosa
l ba
th,
or i
ntes
tina
l wal
l, kB
q/m
l or
kB
q/g
in t
he c
ase
stan
dard
err
or o
f the
mea
n
sero
sal
bath
, or
int
esti
nal w
all,
% B
+ S
+ W
=
su
m;
Cm
Cs
(nm
ol/m
l),
Cw
(nm
ol/g
) of
tri
tiat
ed w
ater
; W
L (
mg)
=
wei
ght
of in
test
inal
con
tent
at
end
of e
xper
imen
t. M
ean
-+_
0
s
0
,.o
0
<
0
+.a
0 0
,.o eo
, ~ - , 0
.o .~
-='a
>
+1 +1 +1 +1 +1
+1 +1 +1 "~ +1 eq +1
-H +1 +1 +1 +1
+1 ~ +1 ~" -H ~" +[ ~q +1
r t"q " ~ ( ' q '~1" "~"
-H +1 +1 +l +1
ak
+1 ~ +1 ~ +1 ~ +1 ~ +1
+1 +1 +1 +1 -I-I
+1 ~ +1 ~ - H ~ - H ~ - H
~.o I'-~ ( ' q ' , O r 1 6 2 o o k O ~ O o o
8 =
.~ ~ -s .s .~
m r < < <
II .
-6*
~A
G A
~ A
A
II * II
f
= 8
JJ
73
isopropanol and Lumosolve (Baker, Deventer, The Nether- lands) and decolourized with 0.4 ml of 30% hydrogen perox- ide. After 15 min 10 ml HC1-Lumagel (0.5 1 of 0.5 mol/1HC1 + 4.5 1 Lumagel) were added. Approximately 200 mg tissue were solubilized with 100 ~tl water in 1 ml of the isopropanol- lumosolve mixture at 50~ After 1 6 - 2 0 h 0.1 ml of 30% hydrogen peroxide and, after further 15 min, 10 ml HC1- Lumagel were added.
Representation of data. Blood flow rate (ml min- 1 g- 1) was calculated from the weight of the collected blood, its specific gravity of 1.054 g/ml (Spector 1956), wet tissue weight of the jejunal segment, and duration of the sampling period. The specific activity of the substance in the injected or perfused solution was derived from the measured radioactivity and the amount of labeled and unlabeled substance added. The concentration in blood CB and serosal bath Cs (nmol/ml) as well as in intestinal wall Cw (nmol/g) was determined from the measured radioactivity and the specific activity of the injected or perfused solution. Since blood, intestinal wall, and serosal solution were not analyzed for metabolites, the concentrations and amounts may include metabolites. The data obtained with tritiated water are given inBq and were not converted into tool. The appearance rate in intestinal venous blood (nmol rain- 1 g - 1) was calculated on the basis of blood flow rate and blood concentration. The amount recovered in blood nB, serosal bath ns, and intestinal wall nw was calculated from volume or weight and concentration. The sum nB + ns + nw is taken as the amount of substance passing the mucosal surface, nB, ns and nw were related to the amount administered: % B, % S, %W. The standard error of the mean is generally designated by " + " and the standard deviation by SD.
R e s u l t s
A. Single administration into a jejunal segment
1. Penetration through mucosal surface and appearance in intestinal venous blood. 0.5 ~tmol of a series of labeled sub- stances listed in Table 1 or 0.5 MBq of tritiated water dis- solved in 0.5 ml buffer solution was injected into a closed jejunal segment of a rat. The appearance rate of the activity (unchanged substance and possible metabolites) in intestinal venous blood reached its maximal value within the first 5 min for all substances except urea, benzylamine and erythritol which peaked between 5 and 10 rain (Fig. 1). For the sake of clarity only a few examples were plotted in Fig. I. The benzylamine curve is situated somewhat below the urea curve. The aminopyrine, galactose, and L-lysine curves are situated between the theophylline and aniline curves, the curves of the other substances between the aniline and benzoic acid curves. The appearance rates declined in accordance with the absorbability of the substances. In the final period (55 -60 rain), the appearance rate of benzoic acid was only 0.6% of its maximal value. The maximal appearance rates of urea, benzylamine and erythritol, which were absorbed slowly, were lower than the rates of the other substances and the slopes of their curves were less steep. The curves in the semilogarithmic plot (Fig. I) did not decline linearly, an indication that the time dependence of the appearance rate cannot be described by a simple mono- exponential function.
The good absorbability of most of the investigated sub- stances is demonstrated by the fact that 5 5 - 8 0 % of the
74
100'
~,=t- ~~= c"5
re" "O o
~- u3
.< >.
Mr Er
Th
i o i . . . . , , .
10 2 0 3 0 4 0 5 0 6 0
Time(mini
so i-
(- 40
0 10 20 30 40 50 60
T ime ( r a i n )
Fig. 1. Permeation of several labeled substances into jejunal blood and serosal bath. Administration of 0.5 ml buffer solution into closed jejunal segment of rat at zero time; initial luminal concentra- tion 1 mmol/1 or 1 GBq/1 (tritiated water); blood flow rate 1.5 ml min- 1 g- 1. Radioactivity in blood and serosal bath converted into mol irrespective of presence of possible metabolites. Left panel: Data from 60-rain experiments (N= 10); right panel: Data from separate experiments 0 -15 rain (N= 5) and 0--60 rain (N= 10). Mean _+ standard error of the mean. A antipyrine, An aniline, B benzoic acid, Bu butanol, Er erythritol, HTO tritiated water, Sa salicylic acid, Th theophylline, Ur urea
administered activity passed the mucosal surface after 15 min and 7 5 - 9 3 % after 60 rain (column %B + S + W, Table 1). The values of L-phenylalanine, benzylamine, urea, and particularly erythritol are lower. 8 8 - 9 9 % of the activity entering the intestinal wall was washed out by the blood within 60 rain independent of epithelial permeability.
2. Appearance in serosai bath. In spite of an adequate blood flow rate (1.5 ml min -a g-a , SD 0.4ml min - j g-a , N = 240) a relatively large amount of tritiated water, aniline, benzyl alcohol, butanol, benzylamine and aminopyrine pen- etrated through the entire intestinal wall. 3 - 1 1 % of the administered activity was recovered in the serosal bath (column %S, Table 1). As expected, serosal transfer of erythritol was very low. But when the penetration into the serosal bath is examined in relation to the amount that passed the mucosal surface, the fraction of erythritol appearing in the serosal bath ( 3 - 4 % ) corresponds to the fraction of urea (3 %) and of other substances. Serosal trans- fer of salicylic acid was also extremely low even though its absorption rate was high. Less than 2% of salicylic and benzoic acid, theophylline, galactose, methyl-e-D-ghico- pyranoside, h-lysine, and g-phenylalanine were transferred into the serosal bath.
The highly permeable substances reached their maximal value in the serosal bath within 15 min, urea after 45 rain (Fig. 1). The concentration of erythritol in the serosal bath increased steadily up to 60 min. The decline of the tritiated water, benzyl alcohol, butanol and aniline curves is due partly to evaporation from the serosal bath as experiments without intestinal segments have demonstrated (data not given in detail). Fifteen minutes after starting the experi- ment, the concentration of all substances in the serosal bath was lower than in the blood (columns CB and Cs, Table 1). After 60 rain, however, the concentration of tritiated water, aniline, benzyl alcohol, and butanol in intestinal venous blood felI below the serosaI concentration.
_ ~ 100
C"T ~.E_
c ~
oe- 7o
c 113
"" 40.
rnl ~ 30. A1D O
Ik 1%_'~k,~ _A0.5 l
3, x,, 'g~bf ~ 20- 3 . "x,j. 8 '~ =
\r i,o " ~ a0.5 <
' 1'o 2'0 3'o 4'o 5'o 6'o ~ o
Time ( m i n )
A 0.5 Lb.f
10 20 30 4 0 50 60
Time (rain)
Fig. 2. Influence of injected volume (distension of intestinal wall) and blood flow rate on intestinal permeation. Administration of 0.5 or 1.0 ml buffer solution into closed jejunal segment of rat at zero time; initial luminal concentration 1 retool/1. Mean -t-_ standard error of the mean (N = 10). A antipyrine, B benzoic acid; l.bf. low intestinal blood flow rate of 0.91 _+ 0.09 ml min- 1 g- 1 (control A 0.5: 1.26 __ 0.06 m! rain- 1 g- 1)
3. Intestinal wall uptake. To compare the radioactivity in the intestinal wall at the end of the experiments with the concentration in blood (column Cw, Table 1) this activity was converted into nmol/g. Fifteen minutes after administra- tion, tissue concentration was roughly equal to or somewhat lower than blood concentration; the concentrations of butanol, benzyl alcohol, benzylamine, g-lysine, e-phenyl- alanine, galactose, and methyl-c~-o-glucopyranoside were higher. Sixty minutes after beginning the experiment, the concentration in blood was usually lower than in tissue. Uptake of butanol, h-lysine, galactose, and L-phenylalanine in tissue reached factors higher than 40. The relatively high concentration of erythritol in the ifitestinal wall was due to the high luminal concentration which only gradually decreased by absorption.
4. Effect of intestinal wall distension and reduced blood flow rate. The intestinal wall was distended by injecting 1 ml of solution with the same concentration instead of 0.5 ml into a jejunal segment of comparable length. Wet tissue weight and blood flow rate did not differ significantly (Table 2). The initial appearance rates of benzoic acid and antipyrine in intestinal venous blood of the distended loops were higher and the slopes of the curves less steep (Fig. 2). The amount transferred into the serosal bath was also higher. The dis- tension of the intestinal wall resulted in a twofold increase of the appearance rate in intestinal venous blood and an almost threefold increase of penetration into the serosal bath (Table 2). The change of intestinal permeation is clearly demonstrated by the absolute values (columns nB and ns, Table 2), the percentages, however, remain more or less unchanged. The effect of distension on wall thickness was measured tentatively in two neighboring jejunal segments of equal length fixed in situ. The distance from the serosal surface to the bottom of the intervillous space was reduced from 228 _+ 2 to 165_ 2 ~tm and the distance from the seros~[1 surface to the basal border of the crypts from 79 _+ 1 to 50 _+ 1 pan (60 measurements each).
Reduction of the blood flow rate from 1.26 to 0.91 ml rain- a g - 1 resulted in a 12% decrease in the appearance rate
75
17" m
o 600 E I--
..c 500.
m 400
t/1 o L. 300 u r )
.--q 200 c
o 100 E
<
~ 0
jA 0
0 0 10 20 30
Time (rain)
Fig. 3. Permeation of some labeled substances into serosal bath. Single pass perfusion (0.1 ml/min) of rat jejunal segment; inflow concentration I mmol/1 or I GBq/I (tritiated water); btood flow rate 1.5 (SD 0.30, N = 32) ml min-1 g-1. Radioactivity in serosal bath converted into mol irrespective of the presence of possible metab- olites. Insert: Note magnified scale of ordinate. Mean + standard error of the mean (N= 8). A antipyrine, B benzoic acid, HTO tritiated water, Sa salicylic acid
of antipyrine; penetration into the serosal bath increased by a factor of 4.
B. Single pass perfusion of a jejunal segment
When an intestinal segment is perfused in the single pass manner, luminal inflow concentration is constant and the absorption reaches a steady state after approximately 5 min. Absorption and penetration into the serosal bath of the following substances was investigated with this method: tritiated water (HTO), antipyrine (A), benzoic (B) and salicylic acid (Sa); luminal concentration 1 GBq/1 and lmmol/1, respectively. The appearance rate in intestinal venous blood was 4 5 4 _ 4 3 k B q min -1 g ~ (HTO), 63.7_+4.1 (A), 92.0_+4.6 (B), 79.4_+5.5 (Sa) nmol min -a g- 1. Activity in the serosal bath increased with time (Fig. 3). The appearance rates in the serosal bath were calculated from the slope of the curves (initial slope of HTO curve): 106_+15kBq rain-1 g 1 (HTO), 3.25_0.34 (A), 0.47 _+ 0.07 (B), 0.062 -t- 0.036 (Sa) nmol min- 1 g - 1. Taking the sum of the two appearance rates as total absorption rate, the fraction penetrating into the serosal bath was 18.9 (HTO), 4.9 (A), 0.51 (B), and 0.078% (Sa).
Discussion
Our experiments have shown that, in the rat small intestine, a fraction of a permeating substance is transferred to the serosal side in spite of intact blood supply. The small bowel wall of the rat is so thin that molecules entering the subepi- thelial tissue cannot be completely washed out by blood. When the intestine is situated inside the abdominal cavity, the fraction appearing on the serosal side is taken up by neighboring loops or by the peritoneum. Under experimen- tal conditions with the intestinal segment placed outside the
abdominal cavity, this fraction diffuses into the moist tissue covering the segment or into the serosal bath as in the experiments reported here. Under conditions of natural blood supply, the difference between the appearance rate in intestinal venous blood and the disappearance rate from the intestinal lumen in the case of substances not accumulated in the intestinal wall (Winne and Remischovsky 1971 a, b; Winne 1972) can be explained by transfer to the serosal surface. If: necessary, this second pathway of intestinal absorption can be introduced into the models describing the relationship between intestinal absorption and blood flow (Winne 1971, 1978). In humans and animals with thick in- testinal walls (e.g. cats and dogs) a mucosal-serosal transfer probably does not take place when the intestinal blood flow is undisturbed. In the case of ischemia a fraction may reach the serosal surface, but clinical implications are improbable.
One requirement for a high serosal appearance rate during undisturbed blood flow is high permeability of the epithelium. Since urea and particularly erythritol are ab- sorbed slowly, their appearance rate on the serosal surface is low. High epithelial permeability alone, however, is not sufficient. Whilst a relative large fraction of tritiated water, aniline, and butanol penetrated into the serosal bath, the fraction of benzoic and salicylic acid which penetrated was small, even though these two substances had the highest absorption rates. Their binding to plasma proteins (Davison and Smith 1961; Kucera and Bullock 1969) could explain the efficiency of washout by blood such that the appearance rate in the serosal bath is low. Although L-lysine, L- phenylalanine, galactose, and methyl-~-D-glucopyranoside are markedly taken up in the intestinal wall, serosal transfer of these actively transported substances is low. This observa- tion indicates that the compartments of accumulation are not in direct contact with the serosal surface and that the washout by blood is efficient. A lack of effect of blood washout is the reason why the accumulation of amino acids and sugars in the intestinal wall and serosal bath above the luminal concentration can be demonstrated in vitro (Kimmich 1981; Munck 1981). The different concentrations in the subepithelial space (low in the presence, high in the absence of blood supply) may be the reason for why in our experiments L-lysine was absorbed at a higher rate than L- phenylalanine, while in vitro the reverse order has been observed (Larsen et al. 1964). Butanol and benzyl alcohol are presumably taken up into lipophilic compartments of the intestinal wall. This process is not necessary for a high serosal appearance rate since the concentration of tritiated water in the intestinal wall is low but its serosal transfer is extremely high. The data obtained in the single pass perfusion experiments demonstrate the broad range of pene- tration to the serosal surface in spite of intact blood flow. The ratios of the appearance rates of tritiated water, anti- pyrine, as well as benzoic and salicylic acid in the serosal bath are 100:3:0.4:0.06.
Serosal transfer depends on the effective mucosal surface area and the thickness of the intestinal wall. Intestinal dis- tension achieved by doubling the injected volume increased serosal transfer of antipyrine and benzoic acid by a factor of 3. Concurrently, the appearance rate in intestinal venous blood was doubled. The blood flow rate remained un- changed. Enlargement of effective mucosal surface area in- creased penetration into tissue and, subsequently, the ap- pearance rate in intestinal venous blood and in serosal bath. Reduction of wall thickness contributed to the increase of
76
serosal transfer. The increase of pressure against the mucosal surface may also trigger intramural nerve reflexes to increase the mucosal permeability. An increase of the appearance rate in intestinal venous blood due to distension of the intestinal wall has also been observed in earlier experiments with single pass perfusion of rat jejunum (Winne 1979a), where the enlargement of the intraluminal radius f rom 1.6 to 3.1 m m increased the absorption rate by a factor of 1.2 for urea to 2.0 for butanol.
Reduction of intestinal blood flow rate diminishes the washout effect and therefore leads to an increase of serosal transfer (Fig. 2). In his experiments with vaseularly perfused jejunum (see Winne 1979b, Fig. 9; Winne 1980, Fig. 4a), Ochsenfahrt showed that the reduction of blood flow lowers the disappearance rate f rom the intestinal lumen and the appearance rate in intestinal venous blood but increases the appearance rate in the serosal bath. When the blood flow rate was reduced to zero, the disappearance rate f rom the lumen and the appearance rate in the serosal bath were equal.
The non-linear decline of the appearance rates in the semilogarithmic plot (Fig. 1) indicates that the penetration o f the substances into the blood after bolus administration into an intestinal segment did not follow first order kinetics. The following processes might have contributed to this phenomenon: small increase of luminal volume, increase o f the pre-epithelial diffusion resistance by secretion of mucus, distribution into and redistribution out of further com- partments (e.g. mucus, cells). The less steep curves of the appearance rates after distention of the intestinal wall (Fig. 2) seem to contradict the observation of higher absorp- tion rates. But the rate constant (time constant) k characterizing the initial slope o f the curves depends on rnucosal permeability (P), effective absorbing area (A), and intraluminal volume (V): k=P*A/V (Winne 1978). In- troducing the cylindrical surface A = 2RnL (R = radius, L = length), and using the volume of a cylinder V = RZ~zL we obtain finally k = 2P(nL/V)1/2. The rate constant for the absorption from a cylindrical compartment is approximately inversely proport ional to the square root o f the volume. An increase o f the volume diminishes the rate constant in spite of an enlarged area.
Since only the radioactivity has been measured in blood, intestinal wall, and serosal bath, it is possible that metab- olites contributed to the data presented in Table 1 and 2. The corresponding values for the original substances would be smaller, but the general conclusions are not affected.
In summary, the wall of rat small intestine is so thin that, in spite of normal blood flow rate, a relatively large fraction (up to 20% in the case of tritiated water) is not washed out by blood but penetrates to the serosal surface. This effect can be expected especially for highly permeable substances with low protein binding. Distension of the intestinal wall and reduction of blood flow rate increases serosal transfer.
Acknowledgements. The authors are indebted to the Zentrum ffir Datenverarbeitung in Tiibingen for carrying out some of the calculations.
References
Davison C, Smith PK (1961) The binding of salicylic acid and related substances to purified proteins. J Pharmacol Exp Ther 133:161 - 170
Kimmich GA (1981) Intestinal absorption of sugar. In: Johnson LR (ed) Physiology of the gastrointestinal tract, vol. 2. Raven Press, New York, pp 1035-1061
Kucera JL, Bullock FJ (1969) The binding of salicylate to plasma protein from several animal species. J Pharm Pharmacol 21:293 -296
Larsen PR, Ross JE, Tapley DF (1964) Transport of neutral, dibasic and N-methyl-substituted amino acids by rat intestine. Biochim Biophys Acta 88 : 570- 577
Lichtenstein B, Winne D (1973) The influence of blood flow on the absorption of 3-O-methylglucose from the jejunum of the rat. Naunyn-Schmiedeberg's Arch Pharmacol 279:153-172
Lichtenstein B, Winne D (1974) The influence of blood flow on the phlorizine-insensitive and sensitive galactose absorption in rat jejunum. Naunyn-Schmiedeberg's Arch Pharmacol 282:195- 212
Munck BG (1981) Intestinal absorption of amino acids. In: Johnson LR (ed) Physiology of the gastrointestinal tract, vol 2. Raven Press, New York, pp 1097-1122
Ochsenfahrt H (1971) The mucosal-serosal transfer of drugs in the rat jejunum with and without blood flow. Naunyn-Schmiede- berg's Arch Pharmacol 270 (Suppl): R 102
Ochsenfahrt H (1979) The relevance of blood flow for the absorp- tion of drugs in the vascularly perfused, isolated intestine of the rat. Naunyn-Schmiedeberg's Arch Pharmacol 306:105-112
Spector WS (1956) Handbook of biological data. Saunders, Philadelphia, p 51
Vogel G, Stoeckert I (1968) Fliissigkeits- und Substanzbewegungen durch die Mucosa verschiedener Abschnitte des Darmes von Ratten bei Angebot isotoner und hypertoner Lrsungen von Natriumcyclohexansulfamat-einem Na-Salz geringer Penetrabi- lit fit. Pfliigers Arch 301:76-90
Winne D (1971) Die Pharmakokinetik der Resorption bei Perfusion einer Darmschlinge mit variabler Durchblutung. Naunyn-Schmiedeberg's Arch Pharmacol 268: 417- 433
Winne D (1972) The influence of blood flow and water net flux on the absorption of tritiated water from the jejunum of the rat. Naunyn-Schmiedeberg's Arch Pharmacol 272:417-436
Winne D (1973) The influence of blood flow on the absorption of L- and D-phenylalanine from the jejunum of the rat. Naunyn- Schmiedeberg's Arch Pharmacol 277:113-138 "
Winne D (1978) Blood flow in intestinal absorption models. J Pharmacokin Biopharm 6: 55-- 78
Winne D (1979 a) Rat jejunum perfused in situ: Effect of perfusion rate and intraluminal radius on absorption rate and effective unstirred layer thickness. Naunyn-Schmiedeberg's Arch Pharmacol 307:265 -274
Winne D (1979b) Influence of blood flow on intestinal absorption of drugs and nutrients. Pharmac Ther 6 : 333 - 393
Winne D (1980) Influence of blood flow on intestinal absorption of xenobiotics. Pharmacology 21 : 1 - 15
Winne D, Remischovsky J (1971a) Der EinfluB der Durchblutung auf die Resorption von Harnstoff, Methanol und ,~thanol aus dem Jejunum der Ratte. Naunyn-Schmiedeberg's Arch Pharma- col 268:392-416
Winne D, Remischovsky J (1971b) Der EinfluB der Durchblutung auf die Resorption yon Polyalkoholen aus dem Jejunum der Ratte. Naunyn-Schmiedeberg's Arch Pharmacol 270:22--40
Received July 20, 1984/Accepted November 26, 1984