Design

A cyclic AMP protein kinase A-dependent mechanism by which rotavirus impairs the expression and enzyme activity of brush border-associated sucrase-isomaltase in differentiated intestinal Caco-2 cells

Description
A cyclic AMP protein kinase A-dependent mechanism by which rotavirus impairs the expression and enzyme activity of brush border-associated sucrase-isomaltase in differentiated intestinal Caco-2 cells
Categories
Published
of 15
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
     U   N   C   O   R   R   E   C    T   E   D    P   R   O   O   F  Cellular Microbiology (2004) doi:10.1111/j.1462-5822.2004.00396.x  © 2004 Blackwell Publishing Ltd  cmi_396.fm  Blackwell Science, LtdOxford, UKCMICellular Microbiology 1462-5814Blackwell Publishing Ltd, 2004Original Article  S.Martin-Latil et al.Rotavirus-induced sucrase-isomaltase impairment   Received 8 October 2003; revised 18 February, 2004; accepted 20February, 2004. *For correspondence. E-mail alain.servin@cep.u-psud.fr; Tel. (   +   33) 1 46 83 56 61; Fax (   +   33) 1 46 83 58 44.  A cyclic AMP protein kinase A-dependent mechanism by which rotavirus impairs the expression and enzyme activity of brush border-associated sucrase-isomaltase in differentiated intestinal Caco-2 cells  Sandra Martin-Latil, Jacqueline Cotte-Laffitte, Isabelle Beau, Anne-Marie Quéro, Monique Géniteau-Legendre and Alain L. Servin  *  Institut National de la Santé et de la Recherche Médicale, Unité 510, Pathogènes et Fonctions des Cellules Epithéliales Polarisées, Faculté de Pharmacie, Université Paris XI, 92296 Châtenay-Malabry, France.  SummaryWe undertook a study of the mechanism by whichrhesus monkey rotavirus (RRV) impairs the expres-sion and enzyme activity of brush border-associatedsucrase isomaltase (SI) in cultured, human, fully dif-ferentiated, intestinal Caco-2 cells. We provide evi-dence that the RRV-induced defects in the expressionand enzyme activity of SI are not related to the previ-ously observed, RRV-induced, Ca   2    +   -dependent, disas-sembly of the F-actin cytoskeleton. This conclusion isbased on the facts that: (i) the intracellular Ca   2    +   blocker, BAPTA/AM, which antagonizes the RRV-induced increase in [Ca   2    +   ]   i  , fails to inhibit the RRV-induced decrease in SI expression and enzymeactivity; and (ii) Jasplakinolide (JAS) treatment,known to stabilize actin filaments, had no effect onthe RRV-induced decrease in SI expression. Resultsreported here demonstrate that the RRV-inducedimpairment in the expression and enzyme activity ofbrush border-associated SI results from a hithertounknown mechanism involving PKA signalling. Thisconclusion is based on the observations that (i) intra-cellular cAMP was increased in RRV-infected cellsand (ii) treatment of RRV-infected cells with PKAblockers resulted in the reappearance of apical SIexpression, accompanied by the restoration of theenzyme activity at the brush border. In addition, inRRV-infected cells a twofold increase of phosphory-lated form of cytokeratin 18 was observed after immu-nopurification and Western Blot analysis, which wasantagonized by exposing the RRV-infected cells to thePKA blockers.Introduction  Rotaviruses are non-enveloped double-stranded, RNAviruses belonging to the reoviridae   family. They exhibit amarked tropism for the differentiated enterocytes of theintestinal epithelium, and are recognized as being the maincause of infantile viral gastroenteritis worldwide. Duringthe last 10 years, an increasing number of studies haveprovided new insights into the pathophysiological mecha-nisms by which rotaviruses induce diarrhoea (Ciarlet andEstes, 2001; Lundgren and Svensson, 2001; Servin,2003). Recently, using cultures of the enterocyte-likemodel, Caco-2, which spontaneously displays many of themorphologic and biochemical properties of mature entero-cytes, we have gained consistent data about the structuraland functional lesions that result from rotavirus infection(Servin, 2003). In particular, by investigating disacchari-dase expression during rhesus rotavirus (RRV) infection,it has been shown that the activity and expression ofsucrase-isomaltase (SI) at the brush border contrast withthose of brush-border dipeptidyl-peptidase IV, and arespecifically and selectively reduced without any apparentcell destruction (Jourdan et al   ., 1998). The RRV infectiondoes not affect the biosynthesis, maturation or stability ofSI, but does induce a block of the direct transportation ofSI from the trans   -Golgi network (TGN) to the brush-border.The organization and maintenance of the brush border-associated cytoskeleton are crucial for that apical-specificintestinal proteins, such as hydrolases and transporters,function correctly (Heintzelman and Mooseker, 1992;Peterson et al   ., 1993; Peterson and Mooseker, 1993).Moreover, it has been established in enterocyte-like Caco-2 cells that intact cytoskeleton networks appear to becrucial for the final polarity of some apical plasma mem-brane proteins, thus controlling transportation of theseapical membrane proteins to the brush border by bothdirect and indirect routes (Le Bivic et al   ., 1990; Matter  et al   ., 1990; Gilbert et al   ., 1991). In the light of theseobservations, we suggest that one mechanism that couldexplain how RRV-induced impairment of SI expression     U   N   C   O   R   R   E   C    T   E   D    P   R   O   O   F  2  S. Martin-Latil et al.    © 2004 Blackwell Publishing Ltd, Cellular Microbiology   cmi_396.fm  occurs at the brush border of infected Caco-2 cells isrelated to the previously reported, RRV-induced rearrange-ments in cytoskeletal proteins (Brunet et al   ., 2000a,b).Indeed, it has been previously been reported that RRVinfection induces microvillar F-actin disassembly in humanintestinal epithelial Caco-2 cells (Jourdan et al   ., 1998).Intracellular Ca   2   +  is known to be a determinant factor incytoskeleton regulation, and we have demonstrated thatRRV-induced; apical; F-actin cytoskeleton disassembly inCaco-2 cells results from a mechanism involving RRV-induced changes in the intracellular concentration of cal-cium (Brunet et al   ., 2000a). We conducted experiments inorder to find out whether the Ca   2   +  -dependent, apical, F-actin disassembly in RRV infected Caco-2 cells in turnpromotes impaired expression of brush-border associatedSI. We show here that SI-targeting changes in RRV-infected Caco-2 cells occur in a Ca   2   +  -independent manner.Using protein kinase A (PKA) inhibitors, we demonstratedthat the RRV-induced impairment in the expression andenzyme activity of SI is a cAMP-protein kinase A-dependent event. We observed that a cAMP-protein kinaseA-dependent increase in cytokeratin (CK) 18 phosphory-lation develops in RRV-infected cells. These data combinedwith our previous reports provide consistent insights intoa new cAMP-dependent PKA mechanism that is developedduring rotavirus infection of fully differentiated Caco-2 cells,and which controls the rotavirus-induced impairment inthe expression and enzyme activity of the brush border-associated hydrolase SI.  Results  Impairment of SI expression in RRV-infected Caco-2 cells is independent of RRV-induced apical F-actin disassembly   It has been clearly established that the brush bordercytoskeleton controls the expression and activity of spe-cific intestinal functional proteins, such as SI (Peterson  et al   ., 1993; Peterson and Mooseker, 1993). We havepreviously demonstrated in differentiated Caco-2 cells thatthe RRV infection is accompanied by Ca   2   +  -dependentrearrangements in the F-actin apical cytoskeleton (Brunet  et al   ., 2000a). In order to find out whether the decreasein SI expression induced by RRV reported above is relatedto the RRV-induced injury of the apical F-actin cytoskele-ton, we conducted experiments in which the F-actincytoskeleton in differentiated Caco-2 cells was stabilizedduring RRV infection. To do this, we used Jasplakinolide(JAS), a monocyclic peptide isolated from the sea sponge  Jaspis johnstoni,  which is known to stabilize actin fila-ments by inducing actin polymerization in polarized epi-thelial cells and/or stabilizing existing actin filaments(Zabriskie et al   ., 1986). Jasplakinolide and phalloidin areknown to bind to the same site on F-actin (Bubb et al   .,1994), which makes it impossible to observe the stabili-zation of the F-actin cytoskeleton by JAS (Le et al   ., 2002).This means that the F-actin changes in RRV-infected cellstreated with JAS could not be demonstrated in fluores-cein-phalloidin stained cells. Together with F-actin, villin(Athman et al   ., 2002), plays a pivotal role in the architec-ture of the intestinal brush border (Mooseker, 1985;Bretscher, 1986; 1991). In order to show that the apicalcytoskeleton had been stabilized by the JAS treatment,we investigated the expression of villin in control and RVV-infected, fully differentiated Caco-2 cells which had or hadnot been treated with JAS. As observed in Fig. 1, villinexpression in the control cells displayed the typical patternof brush-border associated proteins in intestinalcells (Costa de Beauregard et al   ., 1995; Peterson andMooseker, 1992; 1993). This distribution results from thepatterns of the microvilli (MV) on the apical surface of thecells; some cells have well-ordered MV, and others displaydensely packed MV (Zweibaum et al   ., 1991). Villin expres-sion in RRV-infected, fully differentiated cells 22 h postinfection (p.i.), was dramatically altered and was either   Fig. 1.   Effect of Jasplakinolide (JAS) treatment on the RRV-induced alteration in distribution of brush border-associated villin in fully dif-ferentiated Caco-2 cells. Experimental conditions and villin or RRV proteins indirect immunofluorescence labelling were described in Experimental procedures    . En face    micrographs (horizontal x–y optical section) obtained by confocal laser scanning microscopy show expression of villin or RRV proteins in unpermeabilized control, control JAS-treated (1 m   M), RRV-infected cells at 22 h p.i., and RRV-infected +   JAS-treated (1 m   M) cells at 22 h p.i. Compared with expres-sion in control cells, villin expression in RRV-infected cells is dramatically disorganized showing large translucent zones in cells. In contrast, the apical expression of villin is restored in RRV-infected +   JAS cells resembling the expression found in control +   JAS cells. En face    micrographs show the JAS treatment did not affect the level of RRV proteins in RRV-infected +   JAS-treated cells at 22 h p.i. as compared with untreated RRV-infected cells at 22 h p.i. The micro-graphs are representative of three experiments.     U   N   C   O   R   R   E   C    T   E   D    P   R   O   O   F  Rotavirus-induced sucrase-isomaltase impairment   3    © 2004 Blackwell Publishing Ltd, Cellular Microbiology   cmi_396.fm  decreased or totally abolished, localizing only within thecell-to-cell contacts. Treating the RRV-infected cells withJAS resulted in a villin expression pattern similar to thatobserved in control and control, JAS-treated cells. Takenall round, these findings indicate that JAS treatment sta-bilized the brush border cytoskeleton in RRV-infectedCaco-2 cells.We investigated the expression of SI in control, andRRV-infected cells that had or had not been treated withJAS (Fig. 2). SI expression was investigated by indirectimmunofluorescence labelling, using an mAb anti-SI anti-body applied to fixed and non-permeabilized cells, thusallowing us to detect SI present within the apical cellmembrane. The En Face   micrograph shows the typicalmosaic pattern of SI distribution in control cells. Themosaic pattern results from the differing patterns of MV inCaco-2 cells (Peterson and Mooseker, 1992; 1993) andthe fact that the level of brush-border associated, func-tional-protein expression can vary from one cell to another(Peterson et al   ., 1993). When the cells were infected byRRV, this mosaic pattern of SI distribution disappeared,and large translucent zones were observed in the centreof the cells with clumped SI around them. By examiningthe apical expression of SI in JAS-treated, RRV-infectedcells 22 h post infection (p.i.), we found that the RRV-induced impairment of the apical expression of SI had notbeen modified. Indeed, SI immunolabelling of JAS-treated, RRV-infected cells did not display the typicalmosaic pattern of distribution of MV-associated SI that isobserved in JAS-treated, uninfected cells. In conclusion,the fact that stabilizing the F-actin cytoskeleton by JAShad no effect on the RRV-induced reduction of SI expres-sion indicates that this defect develops independently ofthe RRV-induced rearrangements of the apical F-actincytoskeleton.  Impairment in SI expression and enzyme activity in RRV-infected Caco-2 cells is independent of the RRV-induced increase in [Ca    2    +    ]    i   We have previously reported that RRV infection in differ-entiated Caco-2 cells is also followed by events unrelatedto the level of Ca   2   +  (Brunet et al   ., 2000b). This promptedus to try to find out whether the RRV-induced decrease inSI expression and enzyme activity is a Ca   2   +  -independentevent. We used the intracellular Ca   2   +  blocker BAPTA/AMto do this. The intracellular calcium concentration wasdetermined in control and RRV-infected, differentiatedCaco-2 cells, and in RRV-infected cells treated withBAPTA/AM (Fig. 3). In uninfected Caco-2 cells, the aver-age basal [Ca   2   +  ]   i  was 131 ±  12 nM. A significant increasein [Ca   2   +  ]   i  was found in RRV-infected cells, because at 18 hp.i., the [Ca   2   +  ]   i  reached a maximum value of the order of259 ±  10 nM. In RRV-infected cells treated with BAPTA/ AM, [Ca   2   +  ]   i  was of the order of 105 ±  6 nM, a value that   Fig. 2.   Effect of JAS treatment on the RRV-induced alteration in the distribution of brush-border associated SI in fully differentiated Caco-2 cells. Experimental conditions and indirect SI or RRV protein immu-nofluorescence labelling are described in the Experimental proce- dures    section. En face    micrographs (horizontal x–y optical section) obtained by confocal laser scanning microscopy show expression of SI in unpermeabilized controls, JAS-treated (1 m   M) controls, RRV-infected cells at 22 h p.i., and RRV-infected +   JAS-treated (1 m   M) cells at 22 h p.i. Compared to the expression of SI in control cells, SI expression in RRV-infected cells is dramatically disorganized, show-ing large translucent zones in cells expressing RRV proteins. The apical expression of SI in RRV-infected +   JAS-treated cells remains dramatically disorganized, displaying large translucent zones in con-trast to control and control +   JAS cells. The micrographs are repre-sentative of three experiments.   Fig. 3.   Effect of the intracellular Ca   2   +   blocker, BAPTA/AM (25 m   M), on the RRV-induced increase in [Ca   2   +   ]   i   in fully differentiated Caco-2 cells. Experimental conditions and determination of the intracellular Ca   2   +   concentration were described in Experimental procedures    . Results are means ±   SD of five independent experiments conducted with successive cell passages. P    <   0.01 for RRV-infected cells versus con-trol. P    <   0.01 for RRV +   BAPTA/AM cells versus RRV-infected cells. Not significant for RRV +   BAPTA/AM cells versus control cells.     U   N   C   O   R   R   E   C    T   E   D    P   R   O   O   F  4  S. Martin-Latil et al.    © 2004 Blackwell Publishing Ltd, Cellular Microbiology   cmi_396.fm  was not significantly different from that in control, unin-fected and untreated cells.As shown in Fig. 4A, the expression of SI in controlCaco-2 cells treated with BAPTA/AM, displayed the sametypical mosaic pattern observed in control, untreated cells.At 22 h p.i., RRV-infected cells treated with BAPTA/AMshowed the same dramatic decrease in SI expressionobserved in untreated RRV-infected, cells.The sucrase activity was determined in the cell homo-genate and in a brush-border enriched membrane fraction(BBM). The enzyme activity of the cell homogenate ofRRV-infected, differentiated Caco-2 cells was the same asthat of the control, uninfected cells (Fig. 4Ba). In contrast,there was a significant decrease in the enzyme activity inthe BBM of RRV-infected, Caco-2 cells (Fig. 4Bb). Indeed,22 h p.i., the level of sucrase activity was 140 mU mg   -   1  protein, corresponding to a decrease of 50% in their activ-ity, compared to that of control cells. The observation thatsucrase activity was not modified in the cell homogenateof RRV-infected cells, but was dramatically reduced in theBBM suggests that the enzyme is displaced from thebrush border during infection. Sucrase activity determinedin the BBM (Fig. 4Bb), showed that enzyme activity inRRV-infected cells treated with BAPTA/AM reached160 mU mg   -   1  protein. This value was not significantly dif-ferent from the reduced enzyme activity found in RRV-infected, untreated cells. Overall, these results demon-strate that the RRV-induced impairment in SI expressionand activity is not related to the RRV-induced changes in[Ca   2   +  ]   i  .  The impaired expression and enzyme activity of brush border-associated SI in RRV-infected Caco-2 cells were both blocked by protein kinase A inhibitors   It has been shown that a dramatic change occurs in theexpression of brush border-associated hydrolases devel-ops in fully differentiated Caco-2 cells treated with forsko-lin (FK), a drug known to stimulate cAMP adenylyl cyclase(Rousset et al   ., 1989; Baricault et al   ., 1993; 1995). Weinvestigated whether there is any increase in cAMP intra-cellular content in RRV-infected, fully differentiated Caco-2 cells. As shown in Fig. 5, there was a highly significant,   Fig. 4.   Effect of BAPTA/AM treatment (25 m   M) on the RRV-induced change in the distribution and enzyme activity of brush-border asso-ciated SI in fully differentiated Caco-2 cells. Experimental conditions, indirect SI or RRV protein-immunofluorescence labelling and deter-mination of the enzyme activity of sucrase, are described in the Experimental procedures    .A. En face    micrographs obtained by CLSM in double-immunofluores-cence labelled cells (green fluorescence for SI, red fluorescence for RRV proteins, yellow fluorescence for the superimposition of both). Compared to control cells treated with BAPTA/AM, the absence of superimposed green and red fluorescence in RRV-infected +   BAPTA/ AM cells shows that the apical expression of SI remains dramatically disorganized in cells containing high levels of RRV proteins. Micro-graphs showing untreated control and RRV-infected cells are in Fig. 1. The micrographs are representative of three experiments.B. The sucrase enzyme activity determined in the cell homogenate (Ba) and brush-border enriched fraction (Bb). Results are means ±   SD of five independent experiments conducted with successive cell passages. There was no significant difference between control and RRV-infected cells for sucrase enzyme activity in the cell homogenate in either the untreated and BAPTA/AM-treated conditions. Sucrase activity in the brush-border enriched fraction (Bb) of RRV-infected cells treated with BAPTA/AM remains dramatically reduced, as in untreated, RRV-infected cells. P    <   0.01 between control and RRV infected cells (Bb). P    <   0.01 between control +   BAPTA/AM and RRV infected +   BAPTA/AM cells (Bb).   Fig. 5.   Increase in the intracellular cAMP content in RRV-infected, fully differentiated Caco-2 cells. Experimental conditions and the determination of the intracellular cAMP concentration are described in Experimental procedures    . Results are means ±   SD of three inde-pendent experiments conducted with successive cell passages. P    <   0.01 between control and RRV infected cells.     U   N   C   O   R   R   E   C    T   E   D    P   R   O   O   F  Rotavirus-induced sucrase-isomaltase impairment   5    © 2004 Blackwell Publishing Ltd, Cellular Microbiology   cmi_396.fm  1.5-fold increase in the intracellular cAMP content in RRV-infected cells as compared to uninfected control cells.This finding prompted us to find out whether the RRV-induced reduction in the expression and enzyme activityof brush-border associated SI in Caco-2 cells was blockedby treating the RRV-infected cells with the PKA blockersH-89 and KT-5720 (Fig. 6). As observed in Fig. 6A, indi-rect immunofluorescence of SI shows that the reductionin the expression of SI in RRV-infected cells was inhibitedby treating the cells with the PKA blockers, as a typicalmosaic pattern of SI expression was found, that resem-bled the expression of SI in the uninfected, PKA blocker-treated cells. In contrast, in the RRV-infected cells treatedwith K-252c, a PKC inhibitor related to KT-5720, but whichhas no PKA blocking activity, the large translucent zonescharacteristic of the RRV-induced decrease in the apicalexpression of SI, remained present. A restoration of thesucrase activity in RRV-infected cells treated with PKAblocker was observed in the BBM fraction of RRV-infected,and RRV-infected +  H-89-treated cells. The results inFig. 6B, show that enzyme activity reached 200 mU mg   -   1  protein in RRV-infected +  H-89-treated cells, a value notsignificantly different from the value found in uninfectedcells.The observation described above that the RRV-inducedimpairment in expression and enzyme activity of SI couldbe antagonized by PKA blockers prompted us to try andfind out whether the observed RRV-induced alteration inthe apical expression of the brush-border associated pro-tein villin resulted from cAMP-PKA-dependent signalling.As observed in Fig. 8, the expression of villin was consis-tently restored in H89-treated, RRV-infected cells as com-pared with that in H89-treated, uninfected cells and inuntreated, RRV-infected cells.  Appearance of the phosphorylated form of CK18 in RRV-infected, differentiated Caco-2 cells accompanied the RRV-induced SI impairment, a phenomenon inhibited by a PKA blocker   It has been previously reported that RRV infection of dif-ferentiated Caco-2 cells affects the organization of boththe epithelial intermediate filament (IF) proteins and theCKs by a Ca 2 + -independent mechanism (Brunet et al  .,2000b). The appearance of CK18-positive, vesicularstructures in RRV-infected, Caco-2 cells (Brunet et al  .,2000b) resembles the vesicular form of phosphorylatedCKs observed in FK-stimulated, Caco-2 cells (Baricault et al  ., 1994). Interestingly, phosphorylated CKs have beenobserved in rotavirus-infected, undifferentiated HT-29cells (Liao et al  ., 1995). This prompted us to find outwhether CKs are phosphorylated following RRV infectionof differentiated Caco-2 cells (Fig. 7). Immunopurificationand Western Blot analysis using an anti-CK18 antibody Fig. 6. Effect of the PKA blockers, H-89 and KT-5720, and the PKC blocker, K-252c, on the RRV-induced change in the distribution and enzyme activity of brush border-associated SI in fully differentiated Caco-2 cells. The experimental conditions, determination of the enzyme activity of sucrase and indirect SI immunofluorescence or RRV-protein labelling are described in the Experimental procedures.  In A, En face   micrographs obtained by CLSM show that the apical expression of SI is restored in RRV-infected +  H-89 (10 m M) and RRV-infected +  KT-5720 (10 m M) cells. The RRV-infected +  H-89, and RRV-infected +  KT-5720 cells expressed an the same pattern of SI expression as the control +  H-89 and control +  KT-5720 cells, respec-tively. In contrast, the RRV-infected +  K-252c cells (10 m M) show a dramatic loss in SI expression in cells expressing RRV proteins. SI expression in control, untreated and RRV-infected cells is shown in Fig. 1. The micrographs are representative of three experiments.In B, the sucrase activity determined in the brush-border enriched fraction results are expressed as mean ±  SD of four independent experiments conducted with successive cell passages. P    <  0.01 between control, untreated cells and RRV-infected, untreated cells. No significant difference between control +  H-89 (10 m M) cells and RRV-infected +  H-89 (10 m M) cells.
Search
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks