30 1
Clinical Science ( 1988) 74, 30 1-306
Epithelial cell proliferation and intestinal absorptive function
during starvation and refeeding in the rat
R. A. GOODLAD, J. A. PLUMB AND N. A. WRIGHT
Cancer Research Campaign Cell Proliferation Unit, Department of Histopathology, Royal Postgraduate Medical School, Hammersmith
Hospital, London
(Received 28 July 1987; accepted 8 September 1987)
SUMMARY
1. Intestinal epithelial cell production and intestinal
absorption were measured in fed, starved and refed rats.
2. Four days’ starvation significantly decreased the
crypt cell production rate (CCPR), absorption, small
intestinal length and crypt cell population.
3. There was an immediate increase in absorption 1
day after refeeding, which preceded a slower increase in
CCPR. The absorption rate then decreased progressively
after refeeding, and was significantly lower than control
levels 1 week after refeeding. The CCPR, however,
increased more gradually, reaching control levels after 2
days and then ‘overshooting’control values.
4. There was no significant change in the crypt cell
population immediately after refeeding; thus we propose
that the initial increase in absorption on refeeding is either
due to an accelerated maturation rate of the enterocytes
or to the migration of enterocytes from the base of the
villus to the functional zone.
5. The rapid recruitment of absorptive function
appeared to be a ‘one-off event, the villus compartment
then having to wait for increased cell production in the
crypts to repopulate the villi.
Key words: absorption, cell division, cell migration, food
intake, gastrointestinal tract, intestine, proliferation,
starvation.
Abbreviation: CCPR, crypt cell production rate.
INTRODUCTION
Intestinal epithelial adaption is a very useful model in
which to study epithelial growth control [l].The control
Correspondence: Dr R. A. Goodlad, Department of Histopathology, Royal Postgraduate Medical School, Hammersmith
Hospital, Du Cane Road, London W12 OHS.
of intestinal epithelial cell proliferation is probably
multifactorial,the main factors including the direct and/or
indirect effects of food intake (luminal nutrition), local
and/or systemic humoral factors and negative feedback
from the functional to the reproductive compartments [ 11.
The study of intestinal epithelial cell proliferation
requires the use of robust techniques, capable of accounting for the several changes in proliferative parameters
which may occur. One such method is the metaphase
arrest technique as applied to microdissected crypts [2-41.
Intestinal function (absorptive capacity) can,be measured
in vitro by the single-pass segmented-flow technique of
Fisher & Gardner [5], which minimizes the problems
associated with the susceptibility of the intestine to anoxia
and the problems with delayed absorption due to
unstirred layers [6-81. We have recently described a
method for the application of both of these methods to
the same animal, and have shown that crypt cell production rate (CCPR) is proportional to absorptive function in
animals which are in a relative ‘steady state’ of intestinal
renewal and function [9].
The relationship between cell production and intestinal
function is of some interest as there is disagreement
between those who believe that changes in absorption
cannot occur without previous increases in cell production [2] and those who maintain that changes in absorption can occur which are independent of altered cell
production [lo].The present study describes the response
of these two parameters to starvation and refeeding.
Starvation and refeeding is a useful model for the study of
intestinal epithelial growth control as it avoids resorting to
grossly unphysiological stimuli [l11and can yield valuable
information on the hormonal changes associated with
altered cell proliferation [12] and on the relationship
between crypt and villus cell populations [ 131. Moreover,
there are profound changes in cell production [13],and
the temporal inter-relationship of such changes to the rate
of absorption should give insight into the inter-relationship of these two parameters.
302
R. A. Goodlad et al.
METHODS
CCPR
Eight rats were injected with vincristine sulphate (1
mg/kg, intraperitoneally; Eli Lilly, Basingstoke, U.K.) at
09.00 hours. They were killed at timed intervals, within
3 h of injection. Small pieces of intestine from near the
ligament of Trietz, or from the mid-colon, were fixed in
Carnoy’s fluid and the tissue samples were stored in 70%
(v/v) ethanol. They were later hydrolysed in 1 mol/l HCI
for 6 min and stained with Schiff‘s reagent. The intestinal
crypts were microdissected, the number of arrested
metaphases in 20 crypts counted and the mean values
plotted against time after injection. The slope of the line,
fitted by least squares linear regression, gave the CCPR [4,
141.
Crypt cell population
Tissue.was stained with the Feulgen reaction as above.
Crypts were individually microdissected and placed on a
slide. The length and width of the crypts were then
measured with an eyepiece graticule before being
squashed. The number of epithelial nuclei in 10 crypts
was counted [ 141.
set up for absorption measurement the rat was killed. The
time of death was recorded and taken as the time after
vincristine injection for the calculation of CCPR. While
the first rat gut was being perfused, the second rat was
prepared for the second gut bath. The length of the
perfused segment was recorded immediately after the gut
was removed from the gut bath. The intestine was then
dried at 60°C for 2 days or until a constant weight had
been reached.
Animals
Male Wistar rats were used (Olac Ltd, Blackthorn,
Oxon., U.K.). They were kept in wire-bottomed cages to
minimize co-prophagy and were fed a standard laboratory diet (Labshure P.R.D., Christopher Hill, Poole,
Dorset, U.K.). They were housed four to a cage. Water
was available ad libitum.
Statistics
All results are presented as the m e a n f ~for~ each
~
group of eight rats. Groups were compared by a two
tailed t-test for unpaired data. Parameters were correlated
by the Pearson product moment correlation method.
Intestinal perfusion
RESULTS
The ‘segmented-flow’ technique for single-pass luminal
perfusion of Fisher & Gardner [5] which uses a modified
bicarbonate-saline medium with phenol red and 5 g/1
glucose was used. The perfusion apparatus consisted of
two water-jacketed gut baths mounted on a metal frame,
enclosing a water bath for the gassing reservoirs for the
perfusion medium and ‘saline’. Rats were anaesthetized
with ether, the abdomen opened and a catheter tied into
the jejunum, just below the ligament of Trietz, and into the
ileum (5 cm from the ileo-caecal valve). The intestine was
first rinsed with 0.12 mol/l NaCl 0.025 mol/l NaHCO,
solution (‘saline’) at 38°C. This solution, the medium and
the gut baths were all gassed with Medical grade 95%
co,/5% 0,. The gut was then perfused alternately with
perfusion medium and gas mix (95% C02/5% 02).
The
mesenteric arteries were clamped, the mesentery gently
stripped away and the intestine suspended in the gut bath.
The fluid that came out of the serosal and mesenteric side
of the intestine was collected in a fraction collector. The
water absorption rates were determined directly from the
weight of fluid secreted during three consecutive 5 min
periods (after discarding the first and second fractions).
The rats lost 23.3f2.3% of their body weight after 4
day’s starvation. The mean weight changes for the rats are
given in Fig. 1. There was a linear ( P < 0.001) increase in
weight after refeeding. Body weight was’ restored to the
fed level after 3-4 days.
The food intake before fasting was 27.3 f 1.7 g/day per
rat, on the first day after refeeding it was 30.3 f 2.9 g/day
per rat and on subsequent days was 26.9 f6.4,26.7 f 1.0,
26.3 f 1.0,27.5 k 1.3 and 25.6 f0.4 g/day per rat.
Fig. 1 also shows that the CCPR of the duodenum was
profoundly decreased by the fast and that upon refeeding
it increased gradually. Values near the fed control level
were obtained at 2-3 days after refeeding. There was then
an overshoot at day 4 (P<O.O5). There was no response
to refeeding in the colon at day 1, but after this day there
was a marked and persistent increase in CCPR.
The effects of starvation and refeeding on total intestinal absorption are shown in Fig. 2. Starvation significantly decreased total absorption ( P< 0.00 1). After
refeeding there was an immediate increase to near control
levels; however, the absorption rate then decreased with
time so that absorption was significantly ( P < 0.05) lower
than the control level at day 7. Starvation was also
associated with a significant decrease in intestinal length
which persisted until day 4 ( P < O . O l ) . The decrease in
intestinal length means that the increase in absorption per
unit length of the intestine was even greater than that of
the absorption per whole perfused segment, ahd exceeded
that of the control animals on days 1 and 2. Starvation
was also associated with a decrease in intestinal dry
weight (from 1.05k0.13 to 0.80f0.14 8). Dry weight
decreased further on the first day of refeeding (to
Experimental design
Six groups of eight rats were used. The first group was
a control group (fed normally). The second group was
starved for 4 days. The remaining groups were starved for
4 days and refed for 1 , 2 , 4 or 7 days respectively.
Rats were injected with vincristine at 09.00 hours.
Twenty minutes later the first rat was anaesthetized and
the small intestine was cannulated. When the gut had been
Proliferation and absorption in refed rats
2501
303
ftt
c
E
151
:
Y
@
5
5w
10-
L
9)
a
C
.-+0
1
5-
2
3
D
4
“1
40
0-!
1
O!
I
1
1
I
I
1
I
I
1
*
T
-
20-
C
.-0
3
D
0:
Q
I
1
I
I
I
I
I
I
1
2ol
I
I
I
I
1
I
1
1
251
2015105C
O
1
2
3
4
5
Time after refeeding (days)
6
7
Fig. 1. Mean weight changes, CCPR of the duodenum (i.e.
the site 10% of the length of the small intestine) and
CCPR of the mid-colon in rats starved for 4 days and
then refed. Statistical significance: *IJ<0.05, **IJ<0.01,
***IJ<0.001 compared with the fed groups. Abbreviation: C, fed control group.
0.63 f0.30 g) and then gradually increased (0.73k 0.2 on
day 2, 1.09 f0.20 on day 4 and 1.15 f 0.30 g on day 7).
The decrease in intestinal weight after starvation meant
that the effects of starvation on absorption per mg were
minimal, but the initial effects of refeeding were more
pronounced (see Fig. 2).
The effect of starvation and 1 day’s refeeding on intestinal crypt size and cell population is shown in Fig. 3.
There was a highly significant decrease in crypt length,
width and cell population after starvation ( P < 0.001).
O!
I
1
1
I
I
I
I
1
1
c
O
1
2
3
4
5
6
7
Time after refeeding(days)
Fig. 2. Absorption per perfused segment of small intestine
(from the ligament of Trietz to 5 cm above the ileo-caecal
valve) per cm of small intestine and per mg of dry
perfused section in rats starved for 4 days and then refed.
Statistical significance: * I > < 0.05. **I.’< 0.001 compared
with the fed group. Abbreviation: C, fed control group.
Refeeding caused a small but statistically significant
increase in crypt length and width (P<0.05),but had no
effect on cell population size.
DISCUSSION
We have shown that CCPR and absorption are significantly correlated when rats are in a ‘steady state’ of
intestinal cell renewal and function, so that cell produc-
R. A. Goodlad et al.
304
Length
400-
Width
Cell number
h
5
r
400
-300
300-
Y
s
-0
.5
8
0
I-
13
-200
200-
6
5
M
1
-E
-100
100-
0-
c
s
R
C
S
R
C
S
R
-0
Fig. 3. Intestinal crypt size and cell population in the fed control group (C), the starved group (4
days) (S)and the refed group (1 day) (R).Statistical significance: *I>< 0.05, **P< 0.001 compared
with the fed group.
tion will be sufficient to replace cell loss [9]. This was no
longer the case in the present study, as absorption
changed independently of CCPR.
The weight loss after starvation was similar to that seen
in our earlier studies [ 121. There was a far greater weight
gain after refeeding than that seen in a separate investigation where normally fed rats gained an average of 16.5 g
per week per rat over the first 30 weeks of life (R. A.
Goodlad, unpublished work). Body weight gain per unit
energy intake after refeeding is usually greater than that
seen in normally fed animals as starvation-induced energy
conservation persists [15]. The refed rats in this study
should have reached the unfasted body weight after
approximately 10 days of refeeding [15]. There was no
sign of post starvation anorexia [ 15,161.
CCPR fell markedly after 4 day’s starvation. Refeeding
caused a progressive increase in CCPR which returned to
near control levels 2 days after refeeding. The CCPR then
‘overshot’ fed levels, with values close to control levels
being seen at the last time point studied (7 days). The
main changes in proliferative indices observed in starvation and refeeding can be attributed to alterations in the
duration of the phases of the cell renewal cycle [ 17, 181.
Starvation is associated with a slow and gradual decline in
labelling index, mitotic index and crypt cell population,
while the compensatory changes in proliferation seen
after refeeding are more rapid [18]. The increase in
absorption noted in the present study peaked before the
increase in CCPR, thus the relationship between CCPR
and absorption seen in ‘steady state’ models of intestinal
adaptation [9] no longer applies when the intestinal
epithelium is in a state of flux; in fact there was an inverse
relationship between CCPR and absorption in the refed
groups ( r = - 0.703). The increase in absorptive capacity
after refeeding could be the result of an increase in
enterocyte number. We have previously demonstrated
that there can be an influx of crypt cells on to the villus
within 3 h of refeeding starved mice, and that this
precedes increased crypt cell production [ 131. Although
cell migration was originally attributed to ‘mitotic
pressure’ [ 11, intestinal cell migration without cell proli-
feration has been observed [ 19, 201. Villus cell population
can thus be increased by cell migration from the crypts
before cell production increases. However, no change in
crypt cell population was observed in this experiment.
increases in villus height and cell loss within a few hours
of refeeding have been previously described [211;
however, villus height measurements can be confounded
by changes in three-dimensional structure [3, 4, 221. Rat
villi are larger than mouse villi and their shape is less
regular, thus (in contrast to the mouse studies) it is not
practicable to determine cell population by the ideal
method of microdissection and direct counting [4].
Stereological techniques [23] could be used to estimate
villus volume fraction and surface area, but can only give
a relative estimate and must have a stable reference point
with which to relate. A newer stereological method
measures villus surface area and a ‘shape factor’ and thus
is not dependent on a reference point. This method also
confirms that while area and shape may be sensitive indicators of the effects of fasting in the rat, estimates of villus
height are not [24].
An alternative explanation for the increased absorption
seen after refeeding could be that it is the result of
increases in the metabolic function of enterocytes, or the
result of an increase in the maturation of enterocytes as
they migrate up the villus. The expression of absorption in
terms of intestinal dry weight also strongly suggests that
while the decline in absorption after starvation is due to
reduced intestinal mass, the increase after refeeding is not.
The absorption measure used in the present study can
only indicate the number of functional cells not the total
number of villus cells. Only the top third of the villus is
thought to be fully functional [25], and therefore function
is only an approximate indicator of villus cell population.
The subsequent loss of absorptive capacity after day 1
shows that the initial increase was a ‘one-off event and
that wherever the extra capacity came from it could not be
readily replaced.
There are two distinct problems in the quantification
of the effects of dietary change (or restriction) on intestinal absorption, namely the choice of an appropriate
Proliferation and absorption in refed rats
method, finding a suitable manner of normalizing the data
[25] and finding a suitable denominator [26].Absorption
per unit of gut is sufficient for a gross description, but,
although absorption per cell would be better, one is still
left with the problems caused by only a small proportion
of all the cells of the villus actually being active in absorption [25]. The study of the functional maturation of
enterocytes as they migrate from crypt to villus is now
possible [27],and should prove illuminating.
The relationship between villus cell population and
crypt cell population is finely regulated, so that there is an
almost exact equivalence between the cell population of
the two compartments in ‘steady-state’ systems [ 141.
When the system is perturbed by the chemically induced
destruction of crypt cells, crypt cell production is only
stimulated once the villus cell population has also been
reduced [28], suggesting the presence of a negative feedback mechanism from the villus to the crypts [l].The
results of the present study are not incompatible with such
a mechanism, as once the initial, reserve, absorptive
capacity of the villi was used, absorption and CCPR were
inversely related.
Feedback control can be either from the maturation/
functional compartment or from the proliferative
compartment [29].Most of the evidence for negative feedback mechanisms in the control of intestinal epithelial cell
proliferation comes from models when tissue damage
leads to villus cell population depletion. Although
negative feedback may have a role to play in the response
to starvation and refeeding, several other factors are also
likely to b e involved. The presence of food in the gut is
one of the most important stimulators of intestinal
epithelial cell proliferation [l,301. The trophic effects of
food intake can either act directly (luminal nutrition or
luminal workload) or be mediated via systemic and/or
local humoral factors.
In conclusion, we have shown that cell production and
absorption are no longer equivalent in the perturbed
intestine or refed rats. Absorption increases rapidly after
refeeding and then declines as cell production increases,
suggesting that there may be a rapid migration of
immature enterocytes from the base of the villus to the
functional zone or an enhanced maturation of the cells in
situ. This rapid mobilization of absorptive function is a
‘one-off event and the villi then have to wait for increased
cell production from the crypts to refurbish them.
ACKNOWLEDGMENTS
We thank the Cancer Research Campaign for financial
assistance.
REFERENCES
1. Wright, N.A. & Alison, M. (1984) The Biology of Epithelial
Cell Populations, vol. 2. Clarendon Press, Oxford.
2. Clarke, RM. (1973) Progress in measuring epithelial
turnover in the villus of the small intestine. Digestion, 8,
161-175.
305
3. AI-Mucktar, M.Y.T., Polak, J.M., Bloom, S.R. & Wright,
N.A. (1982) The search for appropriate measurements of
proliferative and morphological status in studies on intestinal adaptation. In: Intestinal Adaptation, vol. 11, pp. 3-25.
Ed. Robinson, J.W.L., Dowling, R.H. & Reicken, J.W.L.
MTP Press Ltd, Lancaster.
4. Goodlad, R.A. & Wright, N.A. (1982) Quantitative studies
on epithelial replacement in the gut. In: Techniques in the
Life Sciences, vol. P 2, Techniques in Digestive Physiology,
pp. 21211-212/23. Ed. Titchen, T.A. Elsevier Biomedical
Press, Dublin.
5 . Fisher, R.B. & Gardner, M.L.G. (1974) A kinetic approach
to the study of absorption of solutes by isolated perfused
small intestine. Journal of Physiology (London), 241,
21 1-234.
6. Lerner, J.A. & Miller, D.S. (1982) Techniques in intestinal
absorption studies. In: Techniques in the Life Sciences, vol.
P2, Techniques in Digestive Physiology, pp. 20711-207125.
Ed. Titchen, T.A. Elsevier Biomedical Press, Dublin.
7. Gardner, M.G.L. (1978)The absorptive viability of isolated
intestine prepared from dead animals. Quarterly Journal of
Experimental Physiology, 6 3 , 9 3-9 5 .
8. Gardner, M.G.L. & Plumb, J.A. (1979)Release of dipeptide
hydrolase activities from rat small intestine perfused in vitro
and in vivo. Clinical Science, 57,529-534.
9. Goodlad, R.A., Plumb,.J.A. & Wright, N.A. (1986) The
relationship between intestinal crypt cell production and
water absorption measured in vitro in the rat. Clinical
Science, 72,297-304.
10. Scott, J., Batt, R.M. & Wright, N.A. (1979) Enhancement of
ileal adaptation by prednisolone after proximal small bowel
resection in the rat. Gut, 20,858-864.
1 1. Lipscomb, H.L. & Sharp, J.G. (1982)Effect of reduced food
intake on morphometry and cell production in the small
intestine of the rat. Virchows Archiv B: Cell Pathology, 41,
285-292.
2. Goodlad, R.A., Al-Mukhtar, M.Y.T., Ghatei, M.A., Bloom,
S.R. & Wright, N.A. (1983) Cell proliferation, plasma
enteroglucagon and plasma gastrin levels in starved and
refed rats. VirchowsArchiv B: Cell Pathology, 43,55-62.
3. Goodlad, R.A. & Wright, N.A. (1984) The effects of starvation and refeeding on intestinal cell proliferation in the
mouse. VirchowsArchiv B: Cell Pathology, 45,63-73.
4. Wright, N.A. & Irwin, M. (1982) The kinetics of villus cell
populations in the mouse small intestine. I. Normal villi, the
steady state requirement. Cell and Tissue Kinetics, 15,
595-609.
15. Kotler, D.P., Kral, J.G. & Bjorntorp, P. (1982) Refeeding
after fasting in the rat: effects on body composition and food
efficiency. American Journal of Clinical Nutrition, 36,
444-449.
16. Hamilton, C.L. (1969) Problems of refeeding after starvation in the rat. Annals of New York Academy of Science,
157,1004-1017.
17. Rose, P.M.,Hopper,A.F. & Wannemacher,R.W.(1971) Cell
population changes in the intestinal mucosa of protein
depleted or starved rats. 1. Changes in mitotic cycle time.
Journal of Cell Biology, 50,887-892.
18. Al-Dewachi, H.S., Wright, N.A., Appleton, D.R. & Watson,
A.J. (1975) The effects of starvation and refeeding on cell
population kinetics in the rat. Journal of Anatomy, 119,
105-116.
19. Altmann, G.G. (1972) Influence of starvation and refeeding
on mucosal structure and epithelial renewal in the rat small
intestine. American Journal ofAnatomy, 133,391-400.
20. Potten, C.S., Chwalinski, S., Swindell, R & Palmer, M.
( 1982)The spatial organisation of the hierarchical proliferative cells of the crypts of the small intestine into clusters of
‘synchronised’cells. Celland Tissue Kinetics, 15,35 1-370.
21. Clarke, R.M. (1975)The time course of changes in mucosal
architecture and epithelial cell production and cell shedding
in the small intestine of the rat fed after fasting. Journal of
Anatomy, 120,321-327.
306
R. A. Goodlad et al.
22. Wright, N.A. & Alison, M. (1984) The Biology of Epithelial
Cell Populations, vol. 1. Clarendon Press, Oxford.
23. Aheme, W.A. & Dunnill, M.S. (1982) Morphometry.
Edward Arnold, London.
24. Ross, G.A. & Mayhew, T.M. (1984) Effects of fasting on villi
along the small intestine, a stereological approach to the
problem of quantifying villus 'shape'. Experientia, 40,
856-857.
25. Levin, R.J. (1984) Intestinal adaptation to dietary change as
exemplified by dietary restriction studies. In: Function and
Dysfitnciion of the Small Intestine in Animals, pp. 167-198.
Ed. Blatt, R. & Lawrence T.L.J. Liverpool University Press,
Liverpool.
26. Levin, R.J. (1982) Assessing small intestinal function in
health and disease in vivo and in vitro. Scnndanavian
Jorcrnal of Gastroenterology, 17 (Suppl. 74), 3 1-5 1.
27. Smith, M.W. (1985) Expression of digestive and absorptive
function in differentiating enterocytes. Annual Review of
Physiology, 47,247-260.
28. Wright, N.A. & Al-Nafussi, A. (1982) The kinetics of villus
cell populations in the mouse small intestine. 11. Studies on
growth control after death of proliferative cells induced by
cytosine arabinoside, with special reference to negative
feedback mechanisms. Cell and Tissue Kinetics, 15,
6 11-622.
29. Britton, N.F., Wright, N.A. & Murray, J.D. (1982) A
mathematical model for cell population kinetics in the
intestine. Journalof Theoretical Biology, 98,531-542.
30. Clarke, R.M.(1976) Luminal nutrition versus functional
workload as controllers of mucosal morphometry and
epithelial replacement in the rat small intestine. Digestion,
15,411-429.