Am J Physiol Gastrointest Liver Physiol
279: G587–G596, 2000.
Rotavirus infection impairs intestinal brush-border
membrane Na⫹-solute cotransport activities in young rabbits
NABIL HALAIHEL, VANESSA LIÉVIN, FRANCISCO ALVARADO, AND MONIQUE VASSEUR
Institut National de la Santé et de la Recherche Médicale, Unité 510, Faculté de Pharmacie,
Université de Paris XI, 92296 Châtenay-Malabry, France
Received 19 July 1999; accepted in final form 3 March 2000
of high mortality among
children and young animals. However, in spite of a
great deal of research over several decades, the mechanism(s) underlying these diarrheas remains unclear.
Rotavirus and coronavirus (the transmissible gastroenteritis virus), two entirely distinct viruses, exhibit
similar tropism for the epithelial cells at the tips of the
small intestinal villi (for review, see Ref. 17). Both
viruses can cause extensive enterocyte losses, leading
to flattening of the mucosa, and induce a watery diarrhea that can be fatal if incorrectly treated (17, 23, 30,
33, 34). However, whether or not diarrhea is the result
of malabsorption after tissue damage is open to question. Indeed, the idea is gaining ground that diarrhea
is not necessarily a consequence of any physical lesion
but can precede it, as if cell dysfunction were the cause
rather than the consequence of the histological damage
(2, 4, 7, 31, 38).
Further analogy between rotavirus and coronavirus
includes the fact that both cause a reduced capacity for
intestinal NaCl and nonelectrolyte absorption (9, 17,
22, 23, 28). Both D-glucose and L-alanine absorption are
impaired, although that of L-alanine appears to be
affected only partially (9, 28). By using brush-border
membrane (BBM) vesicles isolated from the jejunum of
14-day-old piglets experimentally infected with coronavirus, Keljo and colleagues (22) showed the existence of
two kinetically distinct D-glucose transport systems,
one of which was selectively abolished after coronavirus infection. The inhibited system was the wellknown, high-affinity Na⫹-D-glucose symporter, presently identified as SGLT1 (20).
Because the Na⫹-D-glucose symporter was suspected
to be present in the BBM of the mature enterocyte but
absent from the crypt cells, Hamilton and colleagues
(9, 17, 22, 23, 28) proposed that virus infection kills off
most of the mature enterocytes, so that crypt cells
invade the villus surface, rendering it unsuitable for
absorption. The resulting malabsorption would be the
direct cause of the diarrhea.
This crypt-cell invasion hypothesis, however, has
never been conclusively demonstrated, and in fact has
been challenged by more recent work from other laboratories (8, 30). The present study concerns an alternative explanation for the mechanism of viral diarrhea, compatible with the original observation by Keljo
and colleagues (22) that viral infection selectively and
strongly inhibits SGLT1 activity.
The existence of a mechanistic relation between the
pathogenesis of viral diarrhea and impaired Na⫹-coupled D-glucose transport becomes apparent when comparing these diarrheas with the human genetic disease
D-glucose and D-galactose malabsorption. The main
symptom of this disease is a watery diarrhea unequivocally attributable to the absence of SGLT1 in the
intestine of babies lacking the corresponding gene (41).
Because both glucose-galactose malabsorption and the
Address for reprint requests and other correspondence: M. Vasseur, Unité 510, INSERM, Faculté de Pharmacie, Université de
Paris XI, 5, rue J.-B. Clément, 92296 Châtenay-Malabry, France
(E-mail:[email protected]).
The costs of publication of this article were defrayed in part by the
payment of page charges. The article must therefore be hereby
marked ‘‘advertisement’’ in accordance with 18 U.S.C. Section 1734
solely to indicate this fact.
viral diarrhea; rabbit intestine; SGLT1; sodium-solute symport
VIRAL DIARRHEAS ARE THE CAUSE
http://www.ajpgi.org
0193-1857/00 $5.00 Copyright © 2000 the American Physiological Society
G587
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.1 on June 18, 2017
Halaihel, Nabil, Vanessa Liévin, Francisco Alvarado,
and Monique Vasseur. Rotavirus infection impairs intestinal brush-border membrane Na⫹-solute cotransport activities in young rabbits. Am J Physiol Gastrointest Liver
Physiol 279: G587–G596, 2000.—The mechanism of rotavirus diarrhea was investigated by infecting young, specific
pathogen-free, New Zealand rabbits with a lapine rotavirus,
strain La/RR510. With 4-wk-old animals, virus shedding into
the intestinal lumen peaked at 72 h postinfection (hpi), and a
mild, watery diarrhea appeared at 124 hpi. No intestinal
lesions were seen up to 144 hpi, indicating that diarrhea does
not follow mucosal damage but can precede it, as if cell
dysfunction were the cause, not the consequence, of the
histological lesions. Kinetic analyses with brush-border
membrane vesicles isolated from infected rabbits revealed
strong inhibition of both Na⫹-D-glucose (SGLT1) and Na⫹-Lleucine symport activities. For both symporters, only maximum velocity decreased with time. The density of phlorizinbinding sites and SGLT1 protein antigen in the membrane
remained unaffected, indicating that the virus effect on this
symporter is direct. Because SGLT1 supports water reabsorption under physiological conditions, the mechanism of
rotavirus diarrhea may involve a generalized inhibition of
Na⫹-solute symport systems, hence, of water reabsorption.
Massive water loss through the intestine may eventually
overwhelm the capacity of the organ for water reabsorption,
thereby helping the diarrhea to get established.
G588
ROTAVIRUS AND INTESTINAL Na⫹-SOLUTE SYMPORT
above-mentioned viral diarrheas have in common a
watery diarrhea accompanying either a genetic or an
acquired dysfunction of SGLT1, in the present study
we have sought to establish whether or not protein
expression and activity of SGLT1 are both similarly
impaired in the course of rotavirus infection. An effect
on SGLT1 activity, unaccompanied by any effect on
SGLT1 protein expression, would constitute compelling evidence that an alternative to the crypt cell invasion hypothesis is necessary.
METHODS
v ⫽ 兵[V max/(K T ⫹ [S])] ⫹ K d其 䡠 [S]
where Vmax (maximum velocity) and KT (the apparent
Michaelis constant) are the capacity and affinity parameters
of classical Michaelis-Menten kinetics, respectively, and Kd is
an apparent diffusion constant (6, 16). To perform each fit,
the procedure of Fletcher and Powell as modified by van
Melle and Robinson (39) was used. Using the commercial
program Stata (Integral Software, Paris, France), we fitted
the nonlinear least-squares regression functions in a single
run to each data set by minimizing the sum of squares of
errors. By comparing “lack of fit” and “pure error” components, we obtained F values providing a quantitative assessment of the goodness of fit (39). All of the fits listed in the
tables were found to be not significant (see Ref. 39), meaning
that, for each given fit, data points did not differ statistically
from the theoretical fit of the equation under study. Statistical comparison between different fits was done by applying
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.1 on June 18, 2017
Rotavirus source. Two 7-wk-old New Zealand rabbits at
the Bergerie Nationale (stock farm) (Rambouillet, France)
were identified as diarrheic. Considered to be naturally infected, presumably by rotavirus, the animals were brought to
our laboratory and immediately killed by cervical dislocation
after stunning. Samples of the intestinal fluid were filtered
through a 0.45-␮m membrane (Analypore, OSI) and used for
virus identification and multiplication in MA-104 monkey
kidney cells in culture (21). By using material from one
animal (LN2; see below) and three passages in MA-104 cells,
a rotavirus stock suspension was thus obtained. Named rotavirus strain “La/RR510” or “RR510” (the lapine rotavirus,
strain La/RR510, has been deposited into the Collection Nationale de Cultures de Microorganismes, Institut Pasteur,
Paris, France), it is stored at ⫺80°C. It has been used as the
source of rotavirus for all the studies that follow.
Virus identification and quantitation. Viral RNA was extracted directly from the intestinal fluid of both naturally and
experimentally infected rabbits by using a (1:1) phenol-chloroform mixture in the presence of 1% SDS. Electrophoretic
analyses were performed as described previously (36). RNA
bands were stained with the Bio-Rad Silver Stain Plus kit.
Rotavirus present in the intestinal fluid of experimentally
infected rabbits was quantitated by enzyme immunoassay
(IDEIA rotavirus test, DAKO Diagnostics) and by titration
on MA-104 cell monolayers after a complete cytopathic effect
was obtained (21). The titer in infectious particles per milliliter (ip/ml) was calculated according to the method of Wyshak and Detre (42). Virus was also quantitated in MA-104
cell cultures by immunofluorescent assay (21).
Identification of rotaviral proteins in both intestinal homogenates and purified intestinal BBM vesicles was done by
Western immunoblot analysis. Proteins were separated by
SDS-PAGE on 10% polyacrylamide gels, electroblotted onto a
polyvinylidene difluoride membrane (Bio-Rad), and immunostained as described previously (24). The rabbit polyclonal
antirotavirus antibody used (no. 8148) was the gift of Dr. J.
Cohen of INRA (Jouy-en-Josas, France). It was used at a
dilution of 1:5,000.
The presence of intact rotavirus particles in the purified
BBM vesicle preparations was demonstrated by electron microscopy after negative staining, as described by Zeng et al.
(43). All micrographs were taken with a Philips EM-208
electron microscope operating at 80 KV.
Rabbit inoculations and sample collection. Specific pathogen-free (SPF) albino hybrid New Zealand 4- or 7-wk-old
rabbits were obtained from Charles River. Animals were
maintained in positively pressurized rooms at a temperature
of 18–20°C at the animal house of the University of ParisSud (Châtenay-Malabry, France). All animals received a
commercial diet (112 UAR, Villemoisson-sur-Orge, France)
and water ad libitum. Rabbits were orally inoculated with 2
ml of the rotavirus stock suspension (105 ip/ml) by means of
a pediatric feeding tube.
Fecal samples were collected daily and scored for consistency as either normal, pasty, or diarrheic. Diarrhea is defined as characterized by fluid stools accompanied by fecal
staining of the perineum. At appropriate times after infection, animals were killed, and for each animal the entire
contents of the small intestine were collected for virus detection and titration. Both jejunal and ileal samples were taken
for histological examination after fixing in 10% formalin at
room temperature. Longitudinal and transversal sections (4
␮m) of paraffin-embedded tissue were stained with Mayer’s
hematoxylin and eosin. Crypt height and villus depth were
determined in a blinded fashion as described previously (25).
For transport study, either jejunal or ileal segments from
7-wk-old rabbits or the entire small intestine from 4-wk-old
rabbits were removed, rinsed with saline at room temperature, everted, and distributed into plastic bags for storage at
⫺80°C, as described previously (37).
Intestinal BBM vesicle preparation. Intestinal BBM vesicles were prepared by using frozen intestine and the magnesium precipitation method as described previously (18). After
suspension at ⬃40 mg protein/ml in the buffer described
below, vesicles were stored in aliquots of 200 ␮l in liquid
nitrogen until the day of the transport assay, as described
previously (37). The membrane buffer consisted of 20 mM
HEPES and 10 mM Tris 䡠 HCl, supplemented with D-arabinose to a total of 600 mosM and adjusted to pH 7.4 with Tris
base (37). Viral and membrane protein concentrations were
determined by using the Bio-Rad protein assay kit, with
serum globulin as standard.
Transport assays. Transport was assayed by using a rapid
filtration technique (6) and either D-[14C]glucose or L-[14C]leucine
as the substrate. Briefly, 10 ␮l of a BBM vesicle preparation
were used to carry out uptake measurements by mixing with
40 ␮l of transport buffer formed by the membrane buffer
containing variable concentrations of unlabeled substrate
(0.1–150 mM), 14C-labeled substrate as a tracer, and 100 mM
NaSCN (final concentrations in the incubation mixtures).
Initial uptake rate measurements were then carried out
for 2.6 s at 35°C (6). After background subtraction (16),
uptake results were calculated as absolute velocities (in
pmol 䡠 s⫺1 䡠 mg membrane protein⫺1) ⫾ SD. Data were statistically compared by applying a global one-way ANOVA (32).
Kinetic analyses. Uncorrected, initial absolute entry rates
as a function of the substrate concentration ([S]) were fitted
by iteration to an equation involving the sum of one nonsaturable, diffusion-like component and one Michaelian, saturable transport component:
ROTAVIRUS AND INTESTINAL Na⫹-SOLUTE SYMPORT
G589
RESULTS
Identification of group A rotavirus in intestinal contents of naturally infected rabbits. Two young adult,
7-wk-old, naturally infected rabbits, identified because
of their having diarrhea, were used to investigate the
presence of rotavirus in their intestinal contents. A
double-stranded RNA was found, exhibiting a series of
11 bands with a characteristic 4, 2, 3, 2 pattern (results
not shown), which indicated the presence of a typical
group A rotavirus. Further analyses, including electron microscopy, MA-104 cell-culture immunofluorescent assay, IDEIA rotavirus test, and immunoblot
analysis, confirmed that a group A rotavirus was
present (Fig. 1).
On necropsy, a net increase in intestinal fluid volume was observed in the naturally infected animals,
compared with that in healthy rabbits of the same age.
Significant histological lesions were seen in both the
jejunum and ileum of the naturally infected animal
LN2 (Fig. 2). The lamina propria exhibited a moderate,
polymorphous inflammatory state, and there was villus atrophy, with crypt length-to-villi length ratios
(0.34 ⫾ 0.08, n ⫽ 2) significantly increased compared
with the controls (0.13 ⫾ 0.03, n ⫽ 8). However, it
should be noted that even when, in this animal, slight
erosion of the epithelial surface existed, there was no
evidence of any significant enterocyte loss or flattening
of the mucosa (Fig. 2).
To study rotavirus infection effects on transport,
BBM vesicles were isolated from the frozen jejunum
and ileum of both healthy and infected rabbits. All the
vesicles were normal (results not shown) according to a
series of previously defined criteria (37). Interestingly,
all BBM vesicles purified from infected, 7-wk-old rabbits contained RR510 antigens and particles, as determined, respectively, by Western immunoblot analysis
and electron microscopy (see Fig. 1). Because VP7 was
among the proteins detected, it can be concluded that
intact triple-layered particles were present, although
the data available do not permit establishing whether
or not double-layered particles were also present.
Fig. 1. Identification of RR510 rotaviral proteins and particles in
intestinal homogenates and brush-border membrane (BBM) vesicles
isolated from the naturally infected rabbit LN1 (see Table 1). A:
Western immunoblot analysis of the jejunal homogenate (lane a),
purified jejunal BBM vesicles (lane b), purified ileal BBM vesicles
(lane c), RR510 virus stock (lane d), and jejunal homogenate (lane e)
with jejunal BBM vesicles from 1 control, healthy rabbit of the same
age (lane f ). Horizontal arrows indicate position of the molecular
mass markers. B: negative staining electron micrograph of the jejunal BBM vesicle preparation shown in lane b. Magnification bar, 100
nm.
Quantitation of RR510 infection in experimentally
infected rabbits. In a preliminary experiment, two SPF
rabbits of the same age as that of the field animals (7
wk) were inoculated with 2 ml of the RR510 virus
stock, a dose found to effectively cause virus replication
in these animals (see below for more detail). However,
the animals did not develop diarrhea and were all
killed at 72 h postinfection (hpi). It was found that, in
contrast with the naturally infected rabbit, no histological damage had occurred, as evinced by the ratios of
crypt length to villi length (0.16 ⫾ 0.03, n ⫽ 8), which
were indistinguishable from those of the controls.
In a second series of experiments, 14 SPF 4-wk-old
animals were used to see if tissue damage and/or diarrhea would be manifested at this age. After inoculation
with the same dose of RR510 virus used above, groups
of three animals were killed at the time intervals
shown in Fig. 3. As indicated in Fig. 3, similar to the
7-wk-old animals, feces had a normal consistency
throughout the first 72 h in 4-wk-old animals (5 of 5
animals). However, a transient, mild, watery diarrhea
appeared at 124 hpi, replaced by pasty feces at 144 hpi.
Rotavirus could be detected in the intestinal lumen
(IDEIA rotavirus test and titration) of each of the
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.1 on June 18, 2017
the F⬘test of van Melle and Robinson (39). All calculations
were done using Apple Macintosh microcomputers.
Phlorizin binding measurements. Quantitation of the
SGLT1 protein present in the isolated BBM vesicles was
performed by measuring phlorizin binding as described by
Garriga et al. (13). The density of specific phlorizin binding
sites was computed as the difference between binding in the
presence of either NaCl or KCl and is expressed as picomoles
bound per milligram of membrane protein at a 50 ␮M phlorizin concentration (B50).
SGLT1 protein identification in isolated BBM vesicles.
BBM proteins (60 ␮g/well) were separated by SDS-PAGE on
8% polyacrylamide gels and immunoblotted as described
previously (14). We used a rabbit polyclonal antibody raised
against a synthetic peptide corresponding to amino acids
564–575 of the adult rabbit SGLT1 sequence (14). The antibody, used at 1:5,000 dilution, was a gift of Dr. M. Kasahara
of Teikyo University (Tokyo, Japan).
G590
ROTAVIRUS AND INTESTINAL Na⫹-SOLUTE SYMPORT
infected animals from the earliest time examined (16
hpi) onward. At 40 and 72 hpi, the excreted viral
particles exceeded the concentration of the input dose
by ⬎12- and 600-fold, respectively. They then dropped
to ⬍4 ⫻ 102 ip/ml by 144 hpi (Fig. 3). It should be
emphasized that, in spite of the large amounts of virus
produced, none of the intestinal samples examined
during this period exhibited any apparent histological
damage. The crypt length-to-villi length ratios for control and infected animals were statistically identical up
to 144 hpi (Fig. 3).
Concerning the identity of the virus shed into the
intestinal fluid of all experimentally infected rabbits
(RNA electropherotype) independently of age, it was
indistinguishable from that originally isolated from
the field animals. Rotaviral proteins (immunoblot
analysis) were detected in BBM vesicles of either all
naturally and experimentally infected, 7-wk-old rabbits or the 4-wk-old rabbits, although here this result
was measurable only at 72 hpi (Fig. 3).
Effect of rotavirus infection on kinetics of D-glucose
and L-leucine uptake. To establish whether or not intestinal Na⫹-organic solute symport activities are affected in the course of rotavirus enteritis, D-glucose and
L-leucine saturation curves were performed by using
BBM vesicles from either normal or infected rabbits.
With both 4- and 7-wk-old animals, a strong inhibition
of the Na⫹-coupled transport of each D-glucose and
L-leucine was apparent in infected rabbits, independent of whether the infection occurred naturally or
experimentally (Tables 1 and 2).
For both symports, Vmax was the only parameter
affected. In contrast, both the KT and the Kd parameters remained practically unchanged under all conditions studied. As shown in Tables 1 and 2, all KT values
fell within the normal limits generally found for this
Fig. 3. Time-dependent viral shedding and characteristics of feces
and intestine of 4-wk-old rabbits infected experimentally with
RR510 rotavirus. Rotavirus excreted in the intestinal fluid was
quantitated by either IDEIA rotavirus test [■, expressed as optical
density (OD) at 450 nm] or titration on MA-104 cell monolayers [Œ,
given as the logarithm of infectious particles/ml (ip/ml)]. a: the
consistency of feces where N ⫽ normal, D ⫽ diarrheic, and P ⫽ pasty.
b: crypt-to-villi length (C/V) ratios where the results were statistically indistinguishable from one another (P ⬍ 0.01). c: group A
rotavirus detected by electrophoretic analyses in intestinal fluid; ⫺
and ⫹ indicate negative and positive results, respectively. d: rotavirus antigens and particles detected in isolated BBM vesicles
(BBMVs) by immunoblot and electron microscopy, respectively.
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.1 on June 18, 2017
Fig. 2. Representative photomicrographs from the jejunum of 7-wkold rabbits (see Table 1). Tissues were stained with hematoxylin and
eosin. Magnification, ⫻300. A: healthy control. B: infected animal
LN2.
parameter in regular intestinal transport studies,
where KT is often found to vary by factors between 1
and 2, occasionally more. Concerning Kd, independent
of age, all the D-glucose (Tables 1 and 2) and L-leucine
(Table 1) results were essentially identical. One set of
Kd results (L-leucine, Table 2) appeared to be double
the rest, but this difference can in principle be attributed to the use of a different batch of filters in this
particular experiment. It is known that different filter
batches can cause Kd measurements to change slightly
but significantly (e.g., see Ref. 27). However, this small
filter effect can be neglected, particularly since it has
no significant impact on the estimation of the critical
Michaelian kinetic parameters, Vmax and KT. As demonstrated by the stability of the Kd parameter within
each given set of experiments (Tables 1 and 2), the
integrity of the BBM vesicles appears not to have been
affected, even at times as long as 144 hpi, a result
indicating that rotavirus infection does not modify the
stability of the membrane. Nevertheless, to verify this
crucial point, additional experiments were performed
to measure the apparent vesicular volume, which, as
defined by Brot-Laroche and Alvarado (5), can be interpreted as a measure of the functional vesicle yield.
In this separate set of experiments, equilibrium uptakes were determined after vesicle incubation for 90
min at 22°C. It was found that, under all conditions
studied, this parameter remained constant at 0.6 ⫾ 0.1
␮l/mg protein (n ⫽ 11), confirming that the integrity
ROTAVIRUS AND INTESTINAL Na⫹-SOLUTE SYMPORT
G591
Table 1. Kinetics of D-glucose and L-leucine uptake by BBM vesicles purified from naturally or
experimentally infected, 7-wk-old rabbits
Kinetic Parameters
⫺1
Animal
Vmax, pmol 䡠 s 䡠 mg
membrane protein⫺1
KT, mM
D-Glucose
928 ⫾ 55
265 ⫾ 47
311 ⫾ 41
283 ⫾ 33
60 ⫾ 19
98 ⫾ 20
80 ⫾ 15
0.53 ⫾ 0.08
0.26 ⫾ 0.11
0.45 ⫾ 0.15
0.33 ⫾ 0.09
0.38 ⫾ 0.23
0.36 ⫾ 0.15
0.38 ⫾ 0.13
Control
LN2
248 ⫾ 32
44 ⫾ 14
0.49 ⫾ 0.12
0.32 ⫾ 0.20
F
df
F⬘ (S)
1.08
0.44
0.58
0.51
1.13
0.33
0.86
(13, 133)
(13, 56)
(11, 41)
(13, 111)
(12, 45)
(13, 68)
(13, 128)
(a)
(b)
(b)
(b)
(c)
(c)
(c)
0.34
0.09
(9, 17)
(10, 18)
(a⬘)
(b⬘)
(jejunum ⫹ ileum)
5.8 ⫾ 0.9
9.0 ⫾ 1.9
5.7 ⫾ 1.2
7.6 ⫾ 1.2
7.6 ⫾ 1.4
7.2 ⫾ 0.9
7.4 ⫾ 0.8
L-Leucine
(jejunum)
8.2 ⫾ 1.9
5.1 ⫾ 1.2
Values for kinetic parameters are means ⫾ SD. The uncorrected, initial absolute entry rates as a function of the substrate concentration
were used to estimate the indicated kinetic parameters. Vmax, maximal velocity; KT apparent Michaelis constant; Kd, apparent diffusion
constant; BBM, brush-border membrane. For the F test, degrees of freedom (df) for pure error and lack of fit, respectively, are indicated in
parentheses (39). For D-glucose, no significant difference was found between jejunal and ileal vesicles isolated from the same animal, so the
corresponding data were pooled to obtain the indicated global (jejunum ⫹ ileum) fits. Each line represents a global fit formed by pooling 2–4
saturation curves, all of which were not significant, both individually and within each given set. For L-leucine, each line represents a single
saturation curve involving 1 animal. Fits were further compared by pairs. Identical letters in parenthesis identify results that are
statistically indistinguishable (P ⬍ 0.01). Relevant data have been pooled and fitted again to yield the indicated overall fit values. Control,
2 7-wk-old, noninfected animals; LN1 and LN2, 7-wk-old, naturally infected animals 1 and 2; LE1, LE2, 7-wk-old, experimentally infected
animals 1 and 2. The intestine was retrieved at 72 h postinfection (hpi).
and yield of BBM vesicles did not change in the course
of rotavirus infection.
The entire set of kinetic results obtained (Tables 1
and 2) indicates quite clearly that Vmax is the only
parameter systematically affected by rotavirus infection. With the 7-wk-old animals, there was a certain
degree of scattering in the Vmax results (Table 1), but
this occurred at random and can be attributed to animal variability. Thus, for instance, by applying the
F⬘test and comparing individual animal results by
pairs, the animals studied could be separated into two
groups (see Table 1 and Fig. 4). One group was formed
by animals LN1 and LE2 (overall fit 1), exhibiting an
identical Vmax inhibition of 70%. The other group was
formed by animals LN2 and LE1 (overall fit 2), exhibiting a significantly different inhibition of 91%. The
statistical value of this series of experiments is
strengthened by the fact that no significant difference
was found between jejunal and ileal BBM vesicles
isolated from the same given animal. This is why the
relevant results have been pooled and are shown as
overall fits, jejunum and ileum, in Table 1 and Fig. 4.
Table 2. Evolution with time of the kinetics of D-glucose and L-leucine uptake by intestinal BBM vesicles
purified from experimentally infected, 4-wk-old rabbits
Kinetic Parameters
⫺1
hpi
Vmax, pmol 䡠 s 䡠 mg
membrane protein⫺1
KT, mM
0
16
40
72
144
931 ⫾ 53
508 ⫾ 25
430 ⫾ 24
294 ⫾ 18
159 ⫾ 16
0.40 ⫾ 0.05
0.39 ⫾ 0.04
0.36 ⫾ 0.05
0.36 ⫾ 0.05
0.29 ⫾ 0.07
0
16
40
72
144
919 ⫾ 67
743 ⫾ 43
592 ⫾ 28
352 ⫾ 27
246 ⫾ 23
0.41 ⫾ 0.06
0.51 ⫾ 0.06
0.52 ⫾ 0.05
0.47 ⫾ 0.07
0.46 ⫾ 0.08
Kd, nl 䡠 s⫺1 䡠 mg
membrane protein⫺1
F Tests
F
df
F⬘ (S)
0.78
1.23
0.54
1.07
0.79
(13, 181)
(11, 132)
(11, 134)
(11, 117)
(9, 81)
(a)
(b)
(b)
(c)
(c)
0.97
0.96
1.62
0.54
0.69
(12, 92)
(12, 48)
(9, 20)
(9, 22)
(12, 99)
(a⬘)
(a⬘)
(a⬘)
(b⬘)
(b⬘)
D-Glucose
5.6 ⫾ 0.9
4.9 ⫾ 0.6
5.0 ⫾ 0.7
2.9 ⫾ 0.5
4.4 ⫾ 0.8
L-Leucine
13 ⫾ 2.1
15 ⫾ 1.4
11 ⫾ 0.9
7.4 ⫾ 0.9
14 ⫾ 1.1
Values for kinetic parameters are means ⫾ SD. Conditions are as in Table 1. Each set of data represents the global nonsignificant fit of
3–5 curves involving 3 animals, except for the L-leucine data at 40 and 72 hpi, where each line represents a single saturation curve involving
1 animal.
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.1 on June 18, 2017
Control
LN1
LE2
Overall fit 1
LN2
LE1
Overall fit 2
F Tests
Kd, nl 䡠 s⫺1 䡠 mg
membrane protein⫺1
G592
ROTAVIRUS AND INTESTINAL Na⫹-SOLUTE SYMPORT
Incidentally, this result confirms the existence of a
kinetic homogeneity of sugar transport activities along
both of these segments of the small intestine (29).
Another conclusion of interest is that the observed
variability in the Vmax results takes place irrespective
of any difference between natural and experimental
infection, as well as between the presence or absence of
intestinal lesions. It can be attributed exclusively to
variability between animals within this age group (7
wk). We have no explanation for the observed difference between animals, but, in any case, these differences do not affect in the least the general conclusion
reached in this study, that Vmax is the only kinetic
parameter affected by rotavirus infection.
Because of the small size of 4-wk-old rabbits, transport assays were performed here by using BBM vesicles isolated from the entire small intestine of these
animals. As depicted in Fig. 5, the inhibition of Na⫹D-glucose symport increased as a function of the time
after infection and indicated the existence of a selective, progressive fall in the Vmax parameter, which
dropped by ⬃45% in 16 h and 83% at 144 hpi (Table 2).
The different groups of infected animals were all statistically different from the control group. In contrast
with the 7-wk-old animals (Table 1), there were no
significant differences between individual animals
within each given time period.
Concerning the Na⫹-L-leucine symporter, the Vmax
parameter also decreased as a function of the time
after infection, but significant inhibition was evident
only at 72 h, compared with 16 h for D-glucose. There
seems to be a parallelism in the Vmax parameter decay
observed for either D-glucose or L-leucine after rotavi-
Fig. 5. D-glucose uptake by intestinal BBM vesicles purified from
experimentally infected 4-wk-old rabbits. Saturation curves were
performed under standard conditions in presence of an out/in ⫽
100/0 mM NaSCN gradient, using vesicles from either noninfected
controls (■) or infected animals at 16 (●), 72 (Œ), or 144 ({) h
postinfection (hpi). Results are expressed as the absolute initial
D-glucose uptake rates ⫾ SD (n ⫽ 12–15 determinations/point). Solid
lines represent theoretical fits computed by using the equation given
in METHODS and the kinetic parameters listed in Table 2. As in Fig 4,
only the data from 0.1 to 10 mM D-glucose have been depicted.
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.1 on June 18, 2017
Fig. 4. D-glucose uptake by intestinal BBM vesicles purified from
naturally and experimentally infected 7-wk-old rabbits. Saturation
curves were performed under standard conditions in presence of an
out/in ⫽ 100/0 mM NaSCN gradient, using vesicles from either
noninfected controls (■), naturally infected animals LN1 (●) and LN2
(Œ), or experimentally infected animals LE1 (E) and LE2 (‚). Results
are expressed as the absolute initial D-glucose uptake rates ⫾ SD
(n ⫽ 9–11 or 4–6 determinations/point for controls and infected
animals, respectively). Because no significant difference was found
between animals LN1 and LE2 or between animals LN2 and LE1,
the corresponding results have been pooled by pairs and fitted again
to obtain, respectively, the overall fits 1 and 2 (see Table 1). From
here were then computed the theoretical fits illustrated by the lowest
two curves. All fits were performed by using whole sets of available
results from 0.1 to 150 mM substrate concentration, but only the
data from 0.1 to 10 mM D-glucose are depicted.
rus infection at long time periods (1 to 6 days). However, a quantitative analysis of the Vmax results (not
illustrated graphically) revealed important differences.
The L-leucine results followed a single exponential
with a decay time constant of 0.22 day⫺1. In contrast,
the D-glucose data involved two exponentials, namely,
a rapid one with a decay time constant of 0.91 day⫺1
plus a second one equal to that characterizing L-leucine
(0.22 day⫺1). We conclude that rotavirus infection
causes a strong inhibition of the Vmax parameters respectively characterizing the symport with Na⫹ of both
D-glucose and L-leucine. Both of these inhibitions took
place in the absence of any apparent histological damage.
We have also investigated whether the Vmax effect
was due to a rotavirus-induced drop in the size of the
NaSCN gradient responsible for uphill D-glucose transport. The results shown in Fig. 6 reveal the existence of
clear overshoots, as demonstrated by the 60 s/equilibrium D-glucose concentration ratios, r60, that are
greater than unity, even at times as long as 144 hpi
(namely, r60 ⫽ 1.4). The size of these overshoots diminishes with the time after infection, indicating that the
rate of substrate influx was inhibited but the size of the
underlying Na⫹ gradient was not affected by the infection. This is the result to be expected if rotavirus
infection selectively impairs the turnover rate of the
symporter.
Finally, to detect possible changes in the total mass
of the Na⫹-D-glucose symporter per unit weight of
membrane protein, specific phlorizin binding and the
amount of SGLT1 antigen present in the purified BBM
vesicles were investigated. First, phlorizin binding was
found not to change as a function of time after rotavirus infection, the B50 values remaining constant at
39 ⫾ 6 pmol/mg membrane protein (n ⫽ 10), independent of time. Second, as determined by densitometry,
the amount of SGLT1 antigen present in the purified
ROTAVIRUS AND INTESTINAL Na⫹-SOLUTE SYMPORT
BBM vesicles was also found to be practically the same
at 144 hpi and in the controls (relative densities of 125
and 100, respectively; Fig. 7). It is worth emphasizing
that the signal given by the young rabbit BBM vesicles
exhibited a single band with a molecular mass of 75
kDa, identical to that of the adult rabbit (14).
DISCUSSION
This study examines whether rotavirus infection
specifically impairs intestinal Na⫹-and-solute symport
systems, which, the results suggest, may play a mechanistic role in the pathogenesis of rotavirus diarrhea.
Kinetic analyses with BBM vesicles isolated from rabbits infected either naturally or experimentally with
the RR510 rotavirus indicated strong inhibition of both
Na⫹-D-glucose (SGLT1) and Na⫹-L-leucine symport activities. For both symporters, inhibition affected only
the Vmax parameter and no correlation was found with
any intestinal lesion. In fact, in contrast with naturally
infected animals, no changes in mucosal histology have
been detected in any of the experimentally infected
rabbits.
With 7-wk-old animals, no evidence of diarrhea occurred during the first 3 days after inoculation, which
was the longest time used. Aside from the unknown
length of time elapsed after infection, the presence of
histological lesions only in naturally infected animals
of the same age may in principle be explained in terms
of the superposition of factors that are absent in the
better-protected experimentally infected laboratory
animals (we used SPF animals). Rotavirus disease is
known to be multifactorial, and it seems logical to
expect that the symptoms exhibited by field animals
are more complex, and graver, than those observable in
laboratory animals.
With 4-wk-old rabbits, mild diarrhea did occur, but
only after 124 hpi. Even when the symptomatology was
mild, it is clear that, in all of our experiments, the
animals were successfully infected, as indicated by the
fact that the virus titer in the intestinal lumen strongly
increased with time (Fig. 3). The possibility exists that
the virus had been attenuated during its passage
through the MA104 cell-culture multiplication procedure (see Refs. 4, 7, and 38). The fact that the progressive inhibition of both the Na⫹-D-glucose and Na⫹-Lleucine symporters took place in the absence of any
apparent histological damage, up to at least 6 days
after infection, suggests that these inhibitions can be
important, although not necessarily the only, determinants in the pathogenesis of rotavirus diarrhea.
Although diarrhea is characteristic of rotavirus enteritis, it is not an obligatory symptom (e.g., see Ref. 7).
Appearance of clear-cut diarrhea varies widely among
animals, depending on factors that include the animal
species, age, and virus strain. Although severe or moderate diarrhea has been found to occur occasionally in
rabbits infected with homologous rotavirus strains
(35), virus shedding and diarrhea do not necessarily
correlate (7, 12). In this study, we will apply the criterion that virulence should be evaluated in terms of the
replication efficiency of the virus, not on its ability to
induce the macroscopic symptom of diarrhea (12).
As shown here, rotavirus infection strongly inhibits
both D-glucose and L-leucine transport into BBM vesicles from young rabbits, although the effect on amino
acid transport takes place more slowly (Table 2). Such
results agree with previous observations showing that,
in piglets infected with either coronavirus (22, 23, 28)
or rotavirus (9), both intestinal D-glucose and L-alanine
transport are inhibited, although the amino acid effect
is quantitatively smaller. For either symporter, Vmax
was the only kinetic parameter affected.
There is only one instance in the literature (25) in
which rotavirus has been found to stimulate rather
than inhibit glucose and amino acid uptake. For instance, Leichus and colleagues (25) failed to find any
evidence of NaCl and/or nonelectrolyte transport inhibition but instead observed an increase in short-circuit
current as induced by either D-glucose or L-alanine. It
is doubtful, however, that use of this particular meth-
Fig. 7. Western blot identification of the SGLT1 antigen in intestinal
BBM vesicles from 4-wk-old rabbits. Lane a: noninfected controls;
lane b: infected animals at 144 hpi. Horizontal arrow indicates
position of the molecular mass marker.
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.1 on June 18, 2017
Fig. 6. Time course of D-glucose uptake into intestinal BBM vesicles
purified from experimentally infected 4-wk-old rabbits. The uptake
of 0.1 mM D-glucose was measured at 22°C in presence of an out/in ⫽
100/0 mM NaSCN gradient, using vesicles from either noninfected
controls (■) or infected animals at either 16 (●), 40 (E), 72 (Œ), or 144
(‚) hpi. Results are expressed as the absolute uptakes, with n ⫽ 3
determinations/point. SD is smaller than the symbols. Inset: the
scale has been reduced to better illustrate the difference between the
lowest 2 curves.
G593
G594
ROTAVIRUS AND INTESTINAL Na⫹-SOLUTE SYMPORT
diminution in the size of the Na⫹ gradient to occur.
Second, the transport measurements are based on use
of isolated BBM vesicles in the presence of 100/0 (out/
in) mM NaSCN gradient (16). Under these conditions,
the membrane potential is likely to approach that
imposed by the thiocyanate gradient because the permeability of this anion is ⬃50 times larger than that of
Na⫹ (see Refs. 3 and 15). Even when, in the present
study, the membrane potential was not measured, it
seems unlikely that a rotavirus-induced increase in
Na⫹ permeability could significantly modify this permeability ratio and, hence, the membrane potential.
Third, an analysis of the D-glucose uptake rate as a
function of time indicates that transport continues to
occur uphill for long periods of time after infection (see
Fig. 6), indicating that the Na⫹ electrochemical gradient energizing this uptake dissipates slowly. Such an
observation agrees with published evidence (9, 17, 30)
indicating that, after rotavirus infection, Na⫹ and Cl⫺
fluxes are unaltered, whereas D-glucose-mediated Na⫹
absorption is diminished. The simplest explanation for
the above-mentioned set of results is that K is the
parameter selectively impaired after rotavirus infection, affecting both the D-glucose and L-leucine Vmax
parameters.
In view of these facts, we conclude that either the
virion itself or, most probably, some product thereof
(see below for the case of NSP4), has such a direct,
specific effect that K drops in the absence of any significant decrease in [CT]. A similar, direct mechanism
has been recently proposed to explain the interaction of
protein kinase C with rabbit intestinal SGLT1 expressed in COS-7 cells (40).
Possible role of NSP4 in pathogenesis of rotavirus
diarrhea. A rotavirus nonstructural glycoprotein,
NSP4, and a synthetic peptide corresponding to residues 114 to 135 [NSP4-(114–135)] both have been
shown to cause age- and dose-dependent diarrhea in
young rodents (2). The induction of diarrhea was specific, unaccompanied by any histological damage, and
occurred within a period of ⬃3 h, similar to that characterizing the heat-stable toxin of Escherichia coli.
Because of these analogies, it was proposed (10) that
NSP4 acts as a viral enterotoxin, activating a Ca2⫹dependent signal transduction pathway that alters intestinal epithelial transport. However, it is known that
ionic concentrations in the stools of rotavirus- and
coronavirus-infected animals, although high, are considerably less than those occurring in the secretory
diarrheas caused by secretagogues such as the enterotoxins of Vibrio cholerae and E. coli (17).
Experiments now underway in our laboratory (1)
show that the NSP4-(114–135) peptide has direct, in
vitro inhibitory effects on the Na⫹-D-glucose symport
activity of BBM vesicles isolated from rabbit intestine,
similar to those evinced by the present in vivo study.
These results (1) strongly suggest that NSP4 is at least
one among other effectors directly causing SGLT1 inhibition during rotavirus infection in vivo. The participation of NSP4-(114–135) in these inhibitions may
explain why the effects of rotavirus infection on trans-
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.1 on June 18, 2017
odological approach could explain the lack of accord
between the results obtained by Leichus et al. (25) and
practically all other laboratories.
Mechanism of rotavirus inhibitory effects on Na⫹and-organic solute symport systems. Two main interpretations can be proposed to explain a selective rotavirus effect on the Vmax parameter. First, it can be
envisaged that the absolute concentration of transporter molecules per unit mass of BBM vesicles ([CT])
drops as a result of the infection. Second, the apparent
rate constant governing substrate translocation under
saturating conditions (the turnover rate; K) could be
the affected parameter.
An effect on [CT] could result from a drop in the
number of functional vesicles as a result of the infection, for instance, due to vesicle lysis. However, this
interpretation can be ruled out in view of the Kd results
demonstrating that rotavirus infection neither interferes with vesicle yield nor causes vesicle lysis. Alternatively, there might be an interference with the correct insertion of newly synthesized symporters into the
BBM, as has been proposed to occur, for instance, with
the enzyme sucrase isomaltase in rotavirus-infected
Caco-2 cells (21). At present, however, no direct information is available concerning this interesting but
highly speculative proposal.
In accord with Hamilton’s crypt cell invasion hypothesis (9, 17, 22, 23), a [CT] effect could result from tissue
damage with substitution of the mature enterocytes by
nontransporting crypt cells. Apart from the fact that,
in the present study, no evidence of massive enterocyte
loss has been found, this interpretation is directly
negated by the following observations. First, if phlorizin binding gives a measure of the SGLT1 protein
concentration present in the membrane (11, 40), our
results indicate that this parameter remains constant
throughout the entire duration of the animals’ contact
with the virus, 144 h (no data shown). Second, the
concentration of SGLT1 antigen revealed in the vesicles by Western blot analysis remains constant for up
to 144 hpi (see Fig. 7). Together, these results demonstrate that the total mass of SGLT1, a specific marker
of the mature enterocyte population (20), is unaffected
by rotavirus infection and, consequently, the observed
Vmax drop cannot be attributed to any significant diminution in the size of the enterocyte population of the
mucosa. Moreover, strong D-glucose transport inhibition occurs at times as short as 16 hpi, which are too
short when compared with the 2–4 days that would be
needed for any substantial replacement of the mature
enterocyte layer by crypt cells (see Refs. 19 and 33).
We conclude that it is the K parameter that is impaired after rotavirus infection. To explain this effect,
it could perhaps be proposed that, directly or indirectly, rotavirus infection modifies ion gradients across
the brush border and, hence, the membrane potential,
thereby causing rheogenic Na⫹-coupled substrate uptake to be inhibited. This possibility, however, can be
rejected for several reasons. First, the short incubation
times (2.6 s) used to measure the initial rates of substrate uptake are too short to permit any significant
ROTAVIRUS AND INTESTINAL Na⫹-SOLUTE SYMPORT
We thank Dr. Joana Planas (Barcelona, Spain) for help with the
detection of SGLT1 protein in the isolated brush-border membrane
vesicles; Dr. Michel Lemullois (Orsay, France) for help with electron
microscopy; Dr. Jean-François Flejou (Paris, France) for interpretation of the histopathologic findings; and our colleagues Catherine
Linxe and Jacqueline Cotte-Laffite for assistance in the viral quantitation.
This work was supported in part by the Institut National de la
Santé et de la Recherche Médicale (INSERM), the Fondation pour la
Recherche Médicale (Paris, France), the Association Française de
Lutte contre la Mucoviscidose (AFLM), and the INCO Program of the
European Economic Community (Grant ERB 3514 PL 950019).
Present address of F. Alvarado: Dpto. Bioquimica y Biologia Molecular, Facultad de Farmacia, Universidad de Salamanca, Salamanca,
E-37007 Spain.
REFERENCES
1. Alvarado F, Halaihel N, Ball JM, Estes MK, and Vasseur
M. Rotavirus nonstructural glycoprotein NSP4 abolishes SGLT1
activity of brush border membrane vesicles from rabbit jejunum
(Abstract). Z Gastroenterol 36: 326, 1998.
2. Ball JM, Tian P, Zeng CQ-Y, Morris AP, and Estes MK.
Age-dependent diarrhea induced by a rotaviral nonstructural
glycoprotein. Science 272: 101–104, 1996.
3. Berteloot A. Highly permeant anions and glucose uptake as an
alternative for quantative generation and estimation of membrane potential differences in brush border membrane vesicles.
Biochim Biophys Acta 857: 180–188, 1986.
4. Bridger JC, Hall GA, and Parsons KR. A study of the basis of
virulence variation of bovine rotaviruses. Vet Microbiol 33: 169–
174, 1992.
5. Brot-Laroche E and Alvarado F. Disaccharide uptake by
brush-border membrane vesicles lacking the corresponding hydrolases. Biochim Biophys Acta 775: 175–181, 1984.
6. Brot-Laroche E, Serrano MA, Delhomme B, and Alvarado
F. Temperature sensitivity and substrate specificity of two distinct Na⫹-activated D-glucose transport systems in guinea-pig
jejunal brush border membrane vesicles. J Biol Chem 261: 6168–
6176, 1986.
7. Ciarlet M, Gilger AG, Barone C, McAuthur M, Estes MK,
and Conner ME. Rotavirus disease, but not infection and development of intestinal histopathlogical lesions, is age restricted
in rabbits. Virology 251: 343–360, 1998.
8. Collins J, Candy DCA, Starkey WG, Spencer AJ, Osborne
MP, and Stephen J. Disaccharidase activities in small intestine of rotavirus-infected suckling mice: a histochemical study.
J Pediatr Gastroenterol Nutr 11: 395–403, 1990.
9. Davidson GP, Gall DG, Petric M, Butler DG, and Hamilton
JR. Human rotavirus enteritis induced in conventional piglets.
J Clin Invest 60: 1402–1409, 1977.
10. Dong Y, Zeng CQ, Ball JM, Estes MK, and Morris AP. The
rotavirus enterotoxin NSP4 mobilizes intracellular calcium in
human intestinal cells by stimulating phospholipase C-mediated
inositol 1,4,5-trisphosphate production. Proc Natl Acad Sci USA
94: 3960–3965, 1997.
11. Ferraris RP and Diamond JM. A method for measuring
apical glucose transporter site density in intact intestinal mucosa by means of phlorizin binding. J Membr Biol 94: 65–75,
1986.
12. Franco MA, Feng N, and Greenberg HB. Molecular determinants of immunity and pathogenecity of rotavirus infection in
the mouse model. J Infect Dis 174, Suppl 1: S47–S50, 1996.
13. Garriga C, Moreto M, and Planas JM. Hexose transport in
the apical and basolateral membranes of enterocytes in chickens
adapted to high and low NaCl intakes. J Physiol (Lond) 514:
189–199, 1999.
14. Garriga C, Rovira N, Moreto M, and Planas JM. Expression
of Na⫹/D-glucose cotransporter in brush-border membrane of the
chicken intestine. Am J Physiol Regulatory Integrative Comp
Physiol 276: R627–R631, 1999.
15. Gunther RD, Schell RE, and Wright EM. Ion permeability of
rabbit intestinal brush border membrane vesicles. J Membr Biol
78: 119–127, 1984.
16. Halaihel N, Gerbaud D, Vasseur M, and Alvarado F. Heterogeneity of pig intestinal D-glucose transport systems. Am
J Physiol Cell Physiol 277: C1130–C1141, 1999.
17. Hamilton JR. Viral enteritis. Pediatr Clin North Am 35: 89–
101, 1988.
18. Hauser H, Howell K, Dawson RMC, and Boyer DE. Rabbit
small intestinal brush border membrane preparation and lipid
composition. Biochim Biophys Acta 602: 567–577, 1980.
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.1 on June 18, 2017
port are not instantaneous. In vitro, NSP4-(114–135)
inhibits SGLT1 practically instantaneously (1), but the
situation in vivo is different, probably because NSP4 is
a nonstructural protein. Being absent from the mature,
infective virion particle, NSP4 needs to be synthesized
after infection, and time will be required for the newly
synthesized protein to be released into the cytoplasm
and/or the intestinal lumen and find its way to its
target, the symporter molecule.
Role of specific inhibition of solute-driven water reabsorption mechanisms in pathogenesis of diarrhea. By
using the Na⫹-D-glucose symporter, SGLT1, expressed
in Xenopus oocytes, Loo et al. (26) have shown that 260
water molecules are directly coupled to each sugar
molecule transported; and they have estimated that
this coupling can account for the net absorption of 5 l of
water/day in the intestine of normal humans. Such
observations suggest strongly that the interference
with water reabsorption, resulting from inhibition of
the Na⫹-D-glucose symporter during rotavirus infection, might be a determining factor in the watery
diarrheas caused by rotavirus in the absence of any
histological damage to the intestinal mucosa.
The question whether, similar to that of D-glucose,
L-leucine transport inhibition plays a role in the pathogenesis of rotavirus diarrhea is also worthy of consideration. It agrees with the well-known fact that organic
solute, Na⫹, and water cotransport is a general phenomenon. Symports underlie the absorption of many
nutrients, not only of glucose and amino acids (19).
Consequently, it seems possible to conceive the mechanism of rotavirus diarrhea as involving a generalized
inhibition of intestinal solute-dependent water reabsorption. Individual effects on the transport of several
common nutrients would act additively, the result being a massive loss of water into the intestinal lumen.
This massive water loss could eventually overwhelm
the intestinal capacity for water reabsorption, thereby
permitting the main symptom of enteritis, diarrhea, to
get established.
Nevertheless, it should be recognized that the evidence provided in this study does not prove that the
mechanism proposed here, inhibition of water reabsorption, explains fully the voluminous diarrhea
caused by rotavirus. As mentioned, rotavirus disease is
multifactorial and additional factors might be needed.
We would propose that the inhibition of water reabsorption described here is a conditioning factor that
paves the way for additional agents (e.g., bacterial
superinfection and the enterotoxin effect of NSP4) to
move in. Most probably, it is the sum of all of these
elements that is responsible for the onset of full-blown
diarrhea.
G595
G596
ROTAVIRUS AND INTESTINAL Na⫹-SOLUTE SYMPORT
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
tion during rotavirus infection in neonatal rats. Gut 36: 231–
238, 1995.
Shaw RD, Hempson SJ, and Mackow ER. Rotavirus diarrhea is caused by nonreplicating viral particles. J Virol 69:
5946–5950, 1995.
Snedecor GW and Cochran WG. Statistical Methods (6th ed.).
Ames, IA: Iowa State Press, 1967.
Snodgrass DR, Angus KW, and Gray EW. Rotavirus infection
in lambs: pathogenesis and pathology. Arch Virol 55: 263–274,
1977.
Starkey WG, Collins J, Candy DCA, Spencer AJ, Osborne
MP, and Stephen J. Transport of water and electrolytes by
rotavirus-infected mouse intestine: a time course study. J Pediatr Gastroenterol Nutr 11: 254–260, 1990.
Thouless ME, DiGiacomo RF, Deeb BJ, and Howard H.
Pathogenecity of rotavirus in rabbits. J Clin Microbiol 26: 943–
947, 1988.
Thouless ME, Digiacomo RF, and Neuman DS. Isolation of
two lapine rotaviruses: characterization of their subgroup, serotype and RNA electropherotypes. Arch Virol 89: 161–170, 1986.
Touzani K, Caüzac M, Vasseur M, and Alvarado F. Rheogenic Cl⫺ conductance and Cl⫺-Cl⫺ exchange activities in guinea
pig jejunal basolateral membrane vesicles. Am J Physiol Gastrointest Liver Physiol 266: G271–G281, 1994.
Tzipori S, Unicomb L, Bishop R, Montenaro J, and Vælioja
LM. Studies on attenuation of rotavirus. A comparison in piglets
between virulent virus and its attenuated derivative. Arch Virol
109: 197–205, 1989.
Van Melle G and Robinson JWL. Systematic approach to the
analysis of intestinal transport kinetics. J Physiol (Paris) 77:
1011–1016, 1981.
Vayro S and Silverman M. PKC regulates turnover rate of
rabbit intestinal Na⫹-glucose transporter expressed in COS-7
cells. Am J Physiol Cell Physiol 276: C1053–C1060, 1999.
Wright EM. Genetic disorders of membrane transport. I. Glucose and galactose malabsorption. Am J Physiol Gastrointest
Liver Physiol 275: G879–G882, 1998.
Wyshak G and Detre K. Estimating the number of organisms
in quantal assays. Appl Microbiol 23: 784–790, 1972.
Zeng CQ-Y, Wentz MJ, Cohen J, Estes MK, and Ramig RF.
Characterization and replicase activity of double-layered and
single-layered rotavirus-like particles expressed from baculovirus recombinants. J Virol 70: 2736–2742, 1996.
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.1 on June 18, 2017
19. Hediger MA and Rhoads DB. Molecular physiology of sodiumglucose cotransporters. Physiol Rev 74: 993–1026, 1994.
20. Hwang ES, Hirayama BA, and Wright EM. Distribution of
the SGLT1 Na⫹/glucose cotransporter and mRNA along the
crypt-villus axis of rabbit small intestine. Biochem Biophys Res
Commun 181: 1208–1217, 1991.
21. Jourdan N, Brunet JP, Sapin C, Blais A, Cotte-Laffitte J,
Forestier F, Quero AM, Trugnan G, and Servin AL. Rotavirus infection reduces sucrase-isomaltase expression in human
intestinal epithelial cells by perturbing protein targeting and
organization of microvillar cytoskeleton. J Virol 72: 7228–7236,
1998.
22. Keljo DJ, Macleod RJ, Perdue MH, Burler DG, and Hamilton JR. D-Glucose transport in piglet jejunal brush-border
membranes: insights from a disease model. Am J Physiol Gastrointest Liver Physiol 249: G751–G760, 1985.
23. Kerzner B, Kelly MH, Gall DG, Butler DG, and Hamilton
JR. Transmissible gastroenteritis: sodium transport and the
intestinal epithelium during the course of viral enteritis. Gastroenterology 72: 457–461, 1977.
24. Labbé M, Charpilienne A, Crawford SE, Estes MK, and
Cohen J. Expression of rotavirus VP2 produces empty corelike
particles. J Virol 65: 2946–2952, 1991.
25. Leichus LS, Goldhill JM, Long WH, Percy RD, Shaw V,
Donovan R, and Burakoff R. Effects of rotavirus on epithelial
transport in rabbit small intestine. Dig Dis Sci 39: 2202–2208,
1994.
26. Loo DDF, Zeuthen T, Chandy G, and Wright EM. Cotransport of water by the Na⫹/glucose cotransporter. Proc Natl Acad
Sci USA 93: 13367–13370, 1996.
27. Malo C and Berteloot A. Analysis of kinetic data in transport
studies: new insights from kinetic studies of Na⫹-D-glucose cotransport in human intestinal brush-border membrane vesicles
using a fast sampling, rapid filtration apparatus. J Membr Biol
122: 127–141, 1991.
28. Rhoads JM, MacLeod RJ, and Hamilton JR. Diminished
brush border membrane Na-dependent L-alanine transport in
acute viral enteritis in piglets. J Pediatr Gastroenterol Nutr 9:
225–231, 1989.
29. Robinson JWL and van Melle G. Kinetics of the sodium/betamethyl-D-glucoside cotransport system in the guinea pig.
J Physiol (Lond) 344: 177–187, 1983.
30. Salim AF, Phillips AD, Walker-Smith JA, and Farthing
MJG. Sequential changes in small intestinal structure and func-