Gut 1996; 38: 853-858

Developmental differences in the expression of thecholera toxin sensitive subunit (Gso) of adenylatecyclase in the rat small intestine

I R Sanderson, Z Xu, S W Chu, Q Y Xie, L J Levine, W A Walker

AbstractBackground-The stimulatory guanosinetriphosphate (GTP) binding protein asubunit (Gsa) of adenylate cyclase is thetarget protein for cholera toxin.Aims/methods-The expression of thissignal transducer was analysed in thesmall intestine of developing rats by RNAtransfer (northern blot) analysis byimmunoblotting, and by ADP-ribosyla-tion ofmembrane proteins.Results-Intestinal Gsa mRNA (about 19kb) was increased in the neonate comparedwith the adult rat. Two isoforms of Gsaproteins, a 45 000 and a 52 000 form, wereexpressed in the small intestinal epithelialcell and both were ADP-ribosylated bycholera toxin. A significant increase in thelarger isoform (52 000) and in its ribo-sylation was noted in the 2 week old suck-ling compared with post-weaned olderanimals. The protein content or ribosyla-tion of the smaller form (45 000) did notsignificantly change with age.Conclusion-These data show that adevelopmental decline of intestinal Gsaexpression seems to be, in part, regulatedat the mRNA level. An increased Gsaexpression in the immature intestine mayhelp to explain a previously reported, dosedependent increased adenylate cyclaseresponse and an increase in fluid secretionto cholera toxin in neonates comparedwith adults.(Gut 1996; 38: 853-858)

DevelopmentalGastroenterologyLaboratory, CombinedProgram in PediatricGastroenterology andNutrition, TheMassachusetts GeneralHospital and HarvardMedical School,Boston, USAI R SandersonZXuS W ChuQ Y XieL J LevineWA Walker

Correspondence to:Dr I R Sanderson,DevelopmentalGastroenterologyLaboratory, CombinedProgram in PediatricGastroenterology andNutrition, MassachusettsGeneral Hospital-East,149 13th Street (149-3404),Charlestown, MA02129-2060, USA.

Accepted for publication13 December 1995

Keywords: Gsa subunit, development, cholera toxin,ADP-ribosylation.

The signal transducing guanosine triphosphate(GTP) binding protein (G protein) that regu-lates adenylate cyclase (EC 4.5.1.1) is an

ao3y heterotrimer.1 2 Under physiologicalconditions, the binding of a hormone or neuro-transmitter to the cell surface 3-adrenergicreceptor activates Gs, the stimulatory regulatorof adenylate cyclase. The activation is initiatedby a release ofbound GDP and the subsequentbinding of GTP to the Gsa subunit of Gs.Gsa bound to GTP dissociates from the ry

subunit to yield an active Gsa-GTP form,which is directly responsible for activation ofthe cyclase catalytic unit. Hydrolysis of boundGTP to GDP by the GTPase intrinsic to Gsaterminates the signal.

During infection, cholera toxin (CT) is pro-duced in the upper small intestine by Vibrio

cholerae. The interaction of this toxin with theenterocyte results in a toxigenic diarrhoealdisease that principally affects young children.3CT is an ADP-ribosylating toxin that uses Gsaas its substrate.4 It is an oligomeric protein of84000, composed of one A-subunit (activecomponent) associated non-covalently withfive B-subunits (binding components). The Asubunit contains two polypeptides, denoted A,and A2, linked by a single disulphide bond.The CT B-subunit binds to a GM1 glycolipidreceptor on the enterocyte surface, and theA-subunit disassociates from B and enters intothe cell, presumably via the formation ofendosomes.5 Cleavage of A2 from the A-sub-unit, by a still undefined cellular reaction,activates the ADP-ribosyltransferase of A1.The enzyme activity of A1 is further increasedby another cellular factor, namely ADP-ribosylation factor (ARP).6 The activated A1then catalyses ADP-ribosylation of Gsa,l 4which in turn inhibits GTPase activity that actsto interrupt the conversion of an active a-GTPform to an inactive form. As a result, adenylatecyclase is persistently activated to producecAMP from ATP. Cyclic AMP accumulationsubsequently opens the Cl- channel in intesti-nal crypt cells, and inhibits the NaCl co-trans-porter in intestinal villus cells, resulting in achanged ion flux that causes massive diar-rhoea.7 Recent studies show that the CFTR(cystic fibrosis transmembrane conductanceregulator) itself could be the cAMP-responsivechloride channel8 9 and it is proposed thatpatients with cystic fibrosis may be less suscep-tible to CT induced diarrhoea.'01The human Gsa gene protein on human

chromosome 20 contains 13 exons and 12introns and spans about 20 kb.'2 In the codingregion, the nucleotide sequence homologybetween human and rat Gsa is 95%.13 Thetissue specific expression of the Gsa isoformresults from an alternative splicing of a singlemRNA precursor. In the human brain, Gsaprotein exists predominantly as both a small(45 000) and large (52 000) protein, which canbe ADP-ribosylated.14 In the rabbit intestine,CT catalyses the ADP-ribosylation of proteinsof 40 and 47 000 respectively.9 The proportionof these isoforms varies among tissues and cellsand their functional differences have not yetbeen clarified.

Despite recent extensive studies of G pro-tein, little is known about the gene expressionand regulation of Gsa in the enterocyte, thenatural target cell for CT during develop-ment. Previously, we have reported an increasein adenylate cyclase activity'5 and in fluid

853

on April 9, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.38.6.853 on 1 June 1996. D

ownloaded from

Sanderson, Xu, Chu, Xie, Levine, Walker

secretion16 in the small intestine of sucklingrats compared with older animals. We havehypothesised that a variation in amounts ofGsa protein during intestinal developmentmight be one of the factors that contributes tothis changed host responsiveness. To test thishypothesis, we have studied the ontogeny ofGsa expression in the rat small intestine. Ourresults show that there is indeed an age depen-dent decline in intestinal Gsax mRNA expres-sion, protein concentration, and ribosylationduring development. Accordingly, we haveconcluded that an increase in gene expressionfor Gsox in the immature gut may contribute, inpart, to the increased enterocyte responsive-ness to CT in young mammals. "l

Methods

AnimalsSpraque-Dawley rats (Charles River BreedingLaboratories, Wilmington, MA) were housedin an animal room with a 12 hour light/darkcycle, fed rat chow (Purina, St Louis, MO),and permitted water ad libitum. In all studies,animals remained with their mothers until 2weeks when one half of each litter was removedfor separation of epithelial cells. The remaininganimals were weaned at 3 weeks and examinedat 8 weeks. For mRNA studies, four litters of10 pups each were examined, litters being splitat 2 weeks with five pups removed from eachlitter. Differences in Gsot protein wereexamined in a separate litter and a further litterwas used to study differences in ribosylation.All experiments were approved by the Sub-committee on Research Animal Care at theMassachusetts General Hospital.

Enterocyte isolationEnterocytes were isolated from full length ratsmall intestine according to a modification of amethod described by Weiser.17 A proteaseinhibitor PMSF (1 mM final concentration)was added to all buffers. Intestinal sacs werefilled with the buffer and incubated at 4°C.Enterocytes from postweaned rats were col-lected every 20 minutes for four consecutiveperiods and pooled. The collecting time wasreduced to 15 minutes for neonatal entero-cytes. For studies of Gsox mRNA, isolatedenterocytes were homogenised in guanidiniumsalts. For protein preparations, enterocyteswere transferred to homogenising buffer (seelater).

Nucleic acid probesHybridisation experiments were performedusing (1) Gscx: cDNA for Gsot was made fromthe polymerase chain reaction (PCR) productsof reverse transcription of rat epithelial cellmRNA using Gsa-specific oligonucleotides asprimers. Primers were synthesised from pub-lished sequences Genbank M12673.'2 13 Afterreverse transcription using 50 U/,lI of reversetranscriptase, cDNA was amplified in fourseparate reaction tubes with each combination

of the following upstream primers (Gsa 331 s:CAGCTGCAGA AGGACAAGCA; Gsa483s: TGCAAGGAGC AACAGCGATG)and downstream primers (Gsa 849as: CGTC-CTGACC TCTGGAATCT; Gas 931as:GATGAACGCC GCAAGTGGAT). Thesize of four PCR products was as predicated bythe distance between the primers. Confirma-tion of their identity was obtained by transfer-ring DNA to nylon membranes (Southernblot) and hybridising with a separately synthe-sised 39-base oligonucleotide Gsax probe(DuPont/NEN, Boston, MA). (2) y-actin:cDNA as a BamHI/HindIII insert in a pSP64vector (a gift of Dr Herman Eisen,Massachusetts Institute of Technology,Cambridge, MA).'8 DNA was labelled with[32P]dCTP (3000 Ci/mmol, DuPont/NewEngland Nuclear, Boston, MA) with Klenowfragment after random hexanucleotidepriming'9 using Prime-It II labelling system(Stratagene, Anaheim, CA).

RNA transfer (northern blot analysis)The epithelial cell pellet was dissolved into atleast 20 vol GIT buffer (4 M guanidineisothiocyanate, 50 mM TRIS (pH 7.6), 2%Sarkosyl and 100 mM 2-mercaptoethanol).The RNAs were deproteinised by selective pre-cipitation from GIT buffer and by extractionwith phenol, chloroform, and isoamyl alco-hol.2022 RNA was quantified by absorbance at260 nm and stored at -20°C. For RNA trans-fer blots (northern blots), RNAs (5 Kug per well)were separated in morpholinopropane sul-phonic acid-formaldehyde agarose gels23 24 andtransferred to GeneScreen Plus membranes(DuPont/NEN) by capillarity. Blots werehybridised and washed according to manufac-turer's recommendations (DuPont/NEN).Washed blots were autoradiographed betweenintensifying screens at -70°C. The amount ofhybridisation of labelled Gsax chain cDNA toRNA transfer blots was determined by measur-ing the optical density of the Gsa mRNAby scanning the autoradiographs utilisingMolecular Dynamics (Molecular Dynamics,laser densitometer Sunnyvale, CA) as des-cribed later. Once Gsax had been measuredblots were reprobed with y-actin whose expres-sion does not change with development inepithelial cells. The amount of Gsa. mRNA ofthe intestinal epithelium of each individual ratwas expressed as the ratio of Gsa mRNAoptical density to Py-actin mRNA opticaldensity. RNA samples from any one litter werealways electrophoresed, transferred, probed,and measured together as a single unit. RNAfrom different litters were examined on dif-ferent membranes. It is possible therefore toexamine the changes in Gsa within a litter, butabsolute values cannot be compared betweenlitters.

Enterocyte protein preparationsAll procedures were performed at 4'C.Isolated enterocytes were homogenised in ahomogenising buffer (0. 1 M TRIS-HCl, pH

854

on April 9, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.38.6.853 on 1 June 1996. D

ownloaded from

Ontogeny of intestinal Gsa expression

bands were quantified using a laser densito-meter (Molecular Dynamics, Sunnyvale, CA).The densitometry programme was establishedto subtract for background, which was recordedbefore other measurements. The optical densityof bands relative to protein content was verifiedin other experiments using radioactive probes,where direct measurements of radioactivitycorrelated with densitometry.27

J L4

ADP-ribosylation ofproteins with CTMembrane fractions prepared from isolatedneonatal and adult enterocytes and adult brainhomogenates were used. CT catalysed ADP-ribosylation was performed as described by Gilland Woolkalis.28 The reaction mixturecontained 100 ,ug of membrane, 20 jig/ml ofactivated CT, 20 ,uM [32P]NAD (20 000cpm/pmol), 20 m.1M isoniazid, 1 mM 3-acetylpyridine adenine nucleotide, 10 mnM

6 thymidine, 10 mM dithiothrietol, 0 1% TritonX-100, and 0.5 mM Gpp(NH)p. The reaction

5 es was incubated at 25°C for one hour. The mem-Z brane was recovered and subjected to SDS-

4 E PAGE on 10% gel. The gel was dried and. subjected to autoradiography. Ribosylation of

3 0 each isoform was determined by measuring1 densitometry of bands at the appropriate mole-

2 C cular weight on the autoradiograph.

1

0

Figure 1: RNA transfer blot analysis of Gsa mRNA expression in the rat small intestineduring perinatal development. (A) Northern blot of epithelial cell RNA probe with Gsaand y-actin cDNA showing a reduction in Gsa mRNA ratio in small intestinal epithelialcells of rats between 2 weeks and 8 weeks (from a single litter). Five ,g ofRNA from eachanimal was electrophoresed and transferred to nylon membranes. (B) The blots werehybridised with Gsa and y-actin cDNAs sequentially and the degree of hybridisationmeasured by densitometry. Each individual transfer blot contained all the specimens ofasingle liner at both time points. The ratio of the Gsa and y-actin for each rat wascalculated. The data are expressed as mean (SEM) of the Gsa mRNA to y-actin mRNAratios. Gsa/actin mRNA was significantly greater (p<O 05) in suckling animals than in 8week rats.

7.4) containing 1 mM PMSF, 20 [ug/ml apro-tinin, and 10 pug/ml leupeptin as proteaseinhibitors. After centrifugation of the homo-genate at 3000 g for 15 minutes, the supernatantwas centrifuged at 105 000 g for one hour. Theresulting pellet was resuspended in homogenis-ing buffer and saved for immunoblotting andADP-ribosylation experiments. A particulatefraction was also prepared from adult rat brainand used as a positive control.

Gel electrophoresis and immunoblottingMembrane proteins were subjected to SDS-PAGE on a 1/0% gel,25 transferred to a nitro-cellulose membrane,26 and probed with a rabbitanti-G protein (RM/1 Gsa) polyclonal antibody.The blot was subsequently incubated with a

horseradish peroxidase conjugated donkeyantirabbit immunoglobulin polyclonal anti-body. The target proteins were then detectedwith an enhanced chemiluminescence westernblotting system as described by the supplier(Amersham, Arlington Heights, IL). Protein

StatisticsmRNA expression was examined by analysis ofvariance (ANOVA). The effects of age andlitter on Gsoractin mRNA were examined asindependent variables. Differences wereregarded as significant at p<0 05. Gso- proteinexperiments were all performed within onelitter and the results of the change between 2weeks and 8 weeks were compared usingStudent's t test. Similarly ribosylation was alsoexamined within a single litter. In these lattercases each point represents the mean andstandard deviation of three experiments.

Results

Developmental decline in Gsax mRNA expressionTo examine the effect of age on Gso mRNA,four litters (40 pups) were examined. Twentypups were killed at 2 weeks of age and theintestinal epithelial cell fraction of each animalwas examined separately. The remaining 20were weaned at 3 weeks and examined at 8weeks. Hybridisation was observed at a bandthat corresponded to 1.9 kb, the size of GsoxmRNA (Fig 1A). Northern blot analysis ofGs(x mRNAPy-actin mRNA from epithelialcells showed a significant decrease with agebetween 2 and 8 weeks (p<0-05) (Fig 1).Although each litter showed a reduction inGsa RNA in relation to that of y-actin, therewas a variation in the degree of this effectbetween litters, being 74%, 38%, 20%, and54% in litters 1 to 4, respectively. The amountof -y-actin mRNA per cell is known to remainconstant in intestinal epithelial cells duringdevelopment.24

A

2 weeks 8 weeks

GeA

'tactin

0.6

z

E

c

1.

B

T

I0.5F-

0.4

0.3

:0

0.l1

oI ... . z.--- ..I ... , 4

timer

855

E

91%

4

11.

15

M2.WooksM 8 Weeks

on April 9, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.38.6.853 on 1 June 1996. D

ownloaded from

Sanderson, Xu, Chu, Xie, Levine, Walker

Expression patternsdeveloping rat intesTo identify intestthere was a declir

A

Anti-Gsoc +

Age (weeks) 2

Figure 2: Gsa proteinexpression in wholeintestine and isolatedenterocytes. (A) Themembrane fraction forwhole intestine (SO ,gprotein/lane) of2 week and8 week old and brain(C) from adult rats weresubjected to SDS-PAGE,transferred to nitrocellulosefilter, andprobed withanti-G protein (RM1Gsa) polyclonal antibody.(B) Densitometric analysisof western blot (A) forwhole intestine performed inthree separate experiments,as expressed as mean (SD).(C) Densitometric analysisof Gsa protein fromenterocytes isolatedfor 2and 8 week old rats (seemethods). Membranefractions were examined asfor whole intestine. Resultsshow mean (SD) of threeseparate experiments.*Significantly different atthe p<0 05 levelfrom thevalue in the 8 week rat.

toD

._

N

0)

C

c

E0

00

B

500 _

s of Gsa isoforms in the as noted with Gsa mRNA during develop-stine ment, intestinal membrane samples wereinal Gsa and to determine if analysed by immunoblotting using an antibodyne in Gsa protein expression that detected Gsa isoforms with a single litter.

Two predominant Gsa isoforms were detectedin whole intestine (Fig 2A) as well as in iso-lated enterocytes, one a 45 000 (small form)and the other a 52 000 (large form). These two

+ + - - -. isoforms were also detected in rat brain, usedhere as a positive control (Fig 2A). An agerelated decline in the expression of the large(52 000) but not the small (45 000) Gsoaisoform was noted in membrane samplesprepared from both whole intestine and fromisolated enterocytes. Densitometric analysisshowed approximately a 45-fold decrease inthe abundance of the 52000 Gsoa isoform

*kDa expressed in whole intestine from 2 week oldkDa preweaned compared with the 8 week old post-- 52 weaned rat (Fig 2B). A similar decline was- 45 seen in Gsa isolated from isolated enterocytes

(Fig 2C) (p<0 05).

CT catalysed ADP-ribosylation ofmembraneproteins in rat enterocytesTo determine ifthe two immunodetectable Gsaoisoforms could serve as substrates for CT inADP-ribosylation, CT catalysed [32P]ADP-ribosylation was performed using membraneproteins prepared from isolated enterocytes. CTcatalysed the ADP-ribosylation of both the45 000 and 52 000 Gsa isoforms. The expres-sion of the large isoform (52 000) declined from

8 C 2 8 C 2 to 8 weeks of age (p<0 025); however, therewas no change in the smaller protein (45 000).In accordance with the data examining Gsoaprotein expression (discussed earlier), there wasa significant fall in ribosylation activity inenterocytes from 2 weeks old compared with 8weeks old rats (Fig 3). These experiments

f250 _-

o

0Q

Ue

C

*C

04.0

c

0

" 3000

to-

i. .

811

8~~

o1000 -

500

o2 8

Age (Weeks)

T

2Age (weeks)

Figure 3: Cholera toxin catalysed P22P] ADP-ribosylationof enterocyte membrane proteins in the pre and postweanedrat. Densitometry analysis of the ribosylated 52 000 bandand standard deviation of three experiments. Theribosylation reaction was performed using 100 jigmembrane, 20 ug/ml activated cholera toxin, 20 pMM 2IyNAD (20 000 cpm/pmol), 20 mM isoniazid, 1 mM 3-acetylpyridine adenine nucleotide, 10 mM thymidine, 10mM thymidine, 10 mM dithiothrietol, 0 1% TritonX-100, and 05 mM Gpp(H)p. The reaction wasincubated at 25°Cfor one hour. The membrane wasrecovered and subject to SDS-PAGE on a 10% gel.*Significant difference atp<0 05 levelfrom correspondingvalue of8 week old rat.

856

on April 9, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.38.6.853 on 1 June 1996. D

ownloaded from

Ontogeny of intestinal Gsa expression 857

therefore show that Gsa mRNA, protein levels,and ribosylation activity are under developmen-tal regulation.

DiscussionCT induced secretory diarrhoea occurs morecommonly in young infants than in olderchildren and adults.3 11 It is the best charac-terised disease mediated by a GsoacAMP sig-nalling pathway. Previous studies from ourlaboratory have shown an increased sensitivityin the adenylate cyclase response to CT in thesmall intestine of suckling rats.15 This wasaccompanied by an increased responsivenessin toxin induced fluid secretion.16 In thisstudy, we have examined these findings andprovide evidence to suggest that the molecularbasis for this changed host responsivenessmight be related to a developmental variationin GsaL gene expression.

Several lines of evidence show that Gsa geneexpression may be modulated by developmen-tal and hormonal regulation as well as bydisease states. For example, an age dependentdecrease in the expression of the 1.9 kb Gsa-mRNA transcript was noted in rat brain29 andrat testes.30 Treatment of F9 mouse terato-carcinoma cells with retinoic acid for five daysresulted in a 20-fold increase in steady statelevels of a 2.0 kb Gsax mRNA, accompanied byan increase in the expression of the 45 000 and52 000 Gsa polypeptides.31 In addition, adecreased Gsoa mRNA level was associatedwith a fall in Gs protein and in adenylatecyclase activity in compensated left ventricularhypertrophy.32 In a neuroblastoma a gliomahybrid, NG108-15 cells, the mechanisms of anethanol induced change in adenylate cyclaseentails a decrease in gene expression of Gsax,resulting in a decrease in the quantity of GsamRNA and Gsa protein in those cellmembranes exposed longterm to ethanol.33

In the small intestine, our results have shownan increased level of Gsax mRNA expression inthe suckling compared with the mature ratalthough the exact degree of decrease variedwith the litter examined (Fig 1). A dependenceon litter is a feature of a number of develop-mentally regulated genes.24 The 52 000, butnot the 45 000, Gsa protein isoform concentra-tion was also significantly greater in sucklingcompared with post-weaned rats. This wasreflected in a reduction of ADP-ribosylation ofthe 52 000 isoform with development althoughboth isoforms were ADP-ribosylated by CT invitro.Under physiological conditions, intestinal

Gsat is used by vasoactive intestinal peptide asa signal transducer, coupling the ,-adrenergicreceptor to adenylate cyclase. Chastre et a134reported that intestinal cells are more sensitiveto a vasoactive intestinal peptide inducedadenylate cyclase response in 17 and 19 dayold fetuses than in adult rats. They furthershowed that the difference might result from adevelopmental variation at the level of vaso-active intestinal peptide receptors.

It is noteworthy that in enterocytes, -the cat-alytic unit of adenylate cyclase is located on the

basolateral membrane,7 while depending onspecies, the distribution of the enzyme'sregulatory unit is not necessarily restricted tothe basolateral membrane. In rat enterocytes,the CT targeted Gsa is found on the basolat-eral membrane.35 In contrast, in the smallintestines ofrabbits, CT catalyses ADP-ribosy-lation of Gsoc proteins in the microvillus mem-brane and these proteins then migrate to thebasolateral membrane where they activate thecatalytic component of adenylate cyclase.36 37The basis for this species specific sorting differ-ence ofGsa proteins in polarised enterocytes iscurrently unknown.

In summary, this study suggests that a post-natal decline in Gsa gene expression in ratenterocytes may in part contribute to an agedependent difference in adenylate cyclaseresponsiveness to stimulation by bacterial tox-ins.We wish to thank Ms Suzzette McCarron and Ms Sally Burkefor their expert typing of this manuscript. We also are indebtedto David Schoenfeld, PhD for statistical advice.

This work was supported by grants from the NationalInstitutes of Health (RO1 HD31852 and RO1 HD12437).

1 Gilman AG. G proteins: transducers of receptor-generatedsignals. Ann Rev Biochem 1987; 56: 615-49.

2 Kaziro Y, Itoh H, Kozasa T, NaKafuky M, Satoh T.Structure and function of signal-transducing GTP-bind-ing proteins. Ann Rev Biochem 1991; 60: 349-400.

3 Kelly MT. Cholera: a worldwide perspective. Pediatr InfectDis 1986; 5: 5101-5.

4 Moss J, Vaughan M. ADP-ribosylation of guanylnucleotide-binding regulatory proteins by bacterial toxins.Adv Enzymnol 1988; 61: 303-79.

5 Janicot M, Fouque F, Desbuquois B. Activation of rat liveradenylate cyclase by cholera toxin requires toxin internal-ization and processing in endosomes. Jf Biol Chem 1991;266: 12858-65.

6 Lencer WI, Delp C, Neutra MR, Madara JL. Mechanism ofcholera toxin action on a polarized human intestinalepithelial cell line: role of vesicular traffic. J Cell Biol 1992;117: 1197-209.

7 Field M, Rad MC, Chang EB. Intestinal electrolyte transportand diarrheal disease. N Engi J7 Med 1989; 321: 879-83.

8 Anderson MP, Gregory RJ, Thompson S, Souza DW, PaulS, Mulligan RC, et al. Demonstration that CFTR is achloride channel by alteration of its anion selectivity.Science 1991; 253: 202-5.

9 Dominguez P, Barros F, Lazo PF. The activation of adeny-late cyclase from small intestinal epithelium by choleratoxin. EurJBiochem 1985; 146: 533-8.

10 Hansson GC. Cystic fibrosis and chloride-secreting diar-rhoea. Nature 1988; 333: 711.

11 Chu SW, Walker WA. Bacterial toxin interaction with thedeveloping intestine: a possible explanation for toxigenicdiarrhea of infancy. Gastroenterology 1993; 104: 916-25.

12 Itoh H, Kozasa T, Nagata S, Nakamura S, Katada T, Ui M,et al. Molecular cloning and sequence determination ofcDNAs for a subunits of the guanine nucleotide-bindingproteins G,, G,, and G. from rat brain. Proc Natl Acad SciUSA 1986; 83: 3776-80.

13 Kozasa T, Itoh H, Tsukamoto T, Kaziro Y. Isolation andcharacterization of the human Gsa gene. Proc Nad AcadSci USA 1988; 85: 2081-5.

14 Gill DM, Meren R. ADP-ribosylation of membraneproteins catalyzed by cholera toxin: basis of the activationof adenylate cyclase. Proc Nad Acad Sci USA 1978; 75:3050-4.

15 Seo JK, Chu SW, Walker WA. Development of intestinalhost defense: an increased sensitivity in the adenylatecyclase response to cholera toxin in suckling rats. PediatrRes 1989; 25: 225-7.

16 Chu SW, Ely IG, Walker WA. Age and cortisone alter hostresponsiveness to cholera toxin in the developing gut. AmJtPhysiol 1989; 256: G220-5.

17 Weiser MM. Intestinal epithelial cell surface membrane gly-coprotein synthesis. I. An indicator of cellular differentia-tion. J Biol Chem 1973; 248: 2536-4 1.

18 Enoch T, Zinn K, Maniatis T. Activation of the human ,B-interferon gene requires an interferon-inducible factor.Mol Cell Biol 1986; 6: 801-10.

19 Feinburg AP, Vogelstein B. A technique for radioactivelylabeling DNA restriction endonuclease fragmnents to highspecific activity. Anal Biochem 1984; 137: 266-7.

20 Chirgwin JM, Przybyla AE, MacDonald RI, Rutter WJ.Isolation of biologically-active ribonucleic acid fromsources enriched in ribonuclease. Biochemistry 1979; 18:5294-9.

21 Quellette AJ, Cordell B. Accumulation of abundant mes-senger ribonucleic acids during postnatal development ofmouse small intestine. Gastroenterology 1988; 94: 114-21.

on April 9, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.38.6.853 on 1 June 1996. D

ownloaded from

858 Sanderson, Xu, Chu, Xie, Levine, Walker

22 Perry RP, LaTorre J, Kelley DE, Greenburg JR. On thelability of poly(a) sequences during extractions of messen-ger RNA from polyribosomes. Biochem Biophys Acta 1979;262: 220-6.

23 Sambrook J, Fritsch EE, Maniatis EF. Molecular cloning.2nd ed. New York: Cold Spring Harbor Laboratory Press,1989.

24 Sanderson IR, Ouellette AJ, Carter EA, Harmatz PR.Ontogeny of class II MHC mRNA in the mouse intestinalepithelium is modulated by age of weaning and diet.Gastroenterology 1993; 105: 974-80.

25 Laemmli UK. Cleavage of structural protein during assem-bly of the head of bacteriophage T4. Nature (London)1970; 227: 680-5.

26 Towbin H, Staehelin T, Gordon J. Electrophoretic transferof proteins from polyacrylamide gels to nitrocellulosesheets: procedure and some applications. Proc Natl AcadSci USA 1979; 76: 4350-4.

27 Oguchi S, Walker WA, Sanderson IR. Insulin-like growthfactor binding protein profile secreted by human intestinalepithelial cells varies with polarity. Biochem Biophys ResCommun 1993; 196: 789-93.

28 Gill DM, Woolkalis MJ. Cholera toxin-catalyzed[32P]ADP-ribosylation of proteins. Methods Enzymol1991; 195: 267-80.

29 Duman RS, Saito N, Tallman JF. Development of the 1B-adrenergic receptor and G protein messenger RNA in ratbrain. Mol Brain Res 1989; 5: 289-96.

30 Haugen TB, Paulssen EJ, Gordeladze JO, Hansson V. Aunique mRNA species for the a subunit ofGs is present in

rat haploid germ cells. Biochem Biophys Res Commun 1990;168: 91-8.

31 Chan SDH, Strewler GJ, Nissenson RA. Transcriptionalactivation of Gsoa expression by retinoic acid and parathy-roid hormone-related protein in F9 teratocarcinoma cells.J Biol Chem 1990; 265: 20081-4.

32 Chen L, Vatner ED, Vatner SF, Hittinger L, Homcy CJ.Decreased Gsa mRNA levels accompany the fall in Gsand adenyl cyclase activities in compensated left ventricu-lar hypertrophy: in heart failure, only the impairment ofadenylyl cyclase activation progresses. J Clin Invest 1991;87: 293-8.

33 Mochly-Rosen D, Chang F-H, Cheever L, Kim M, DiamonI, Gordon AS. Chronic ethanol causes heterologousdesensitization of receptors by reducing oas messengerRNA. Nature 1988; 333: 848-50.

34 Chastre E, Emami S, Rosselin G, Gespach C. Ontogenicdevelopment of vasoactive intestinal peptide receptors inrat intestinal cells and liver. Endocrinology 1987; 121:2211-21.

35 Pamukcu R, Yamashita R, Chang EB. Immunological andbiochemical identification of GTP-binding protein distri-bution in rat enterocyte plasma membranes.Gastroenterology 1989; 96: A382.

36 Dominquez P, Velasco G, Barros F, Lazo PS. Intestinalbrush border membranes contain regulatory subunits ofadenyl cyclase. Proc NadAcad Sci USA 1987; 84: 6965-9.

37 Longbottom DS, van Heyningen DS. The activation ofrabbit intestional adenylate cyclase by cholera toxin.Biochem BiophysActa 1987; 1014: 289-97.

on April 9, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.38.6.853 on 1 June 1996. D

ownloaded from