Diarrheal disease is a major cause of morbidity and mortality in infants and children worldwide. Evidence suggests that the interaction of immature human enterocytes with bacteria and their enterotoxins may account for the increased susceptibility of neonates to diarrheal diseases. However, the precise mechanisms that contribute to the excessive response to cholera toxin by the immature gut are largely unknown. Our aim was to characterize the cellular/molecular changes in Gsα during gut development. In this study, a colonic human epithelial cell line (T84) was used as representative of a mature enterocyte and a human fetal primary small intestinal cell line (H4) as representative of an immature enterocyte. Using our cell culture model of human intestinal development, we provide consistent evidence that cholera toxin (CT)-mediated Gsα activation in fetal enterocytes differs from that of mature enterocytes, and the difference may be related to ADP-ribosylation factor (ARF) interaction with the CT-signaling process. Here we demonstrated that ARF1 may play a critical role in clathrin-mediated CT trafficking through the endoplasmic reticulum and Golgi and that ARF6 may facilitate clathrin-mediated CT endocytosis that leads to enhanced Gsα activation by CT. Collectively, these findings support our hypothesis that there is a developmentally regulated intestinal cellular response to bacterial exotoxins involving complex cellular events that accounts for the increased incidence and severity of toxogenic diarrhea during infancy.
- cholera toxin trafficking
- developmental cholera toxin-Res
- stimulatory guanine nucleotide binding protein-α
cholera toxin (CT) is a pentavalent protein comprised of five binding (B) subunits that interact with an enterocyte microvillus glycolipid ganglioside (GM1) and one active (A) subunit with enzymatic activity that is capable of ADP-ribosylation of the α-subunit of a heterotrimeric guanine nucleotide-binding (G) protein Gs (Gsα) (16). In polarized epithelial cells, the activation of signal transduction by CT requires endocytosis of the toxin-receptor complex into the apical endosome, its translocation by retrograde transport into the Golgi cisternae or endoplasmic reticulum (ER), and finally the movement of the toxin to its site of action in the basolateral membrane (reviewed in Refs. 18, 21, 37).
CT exerts its effect on intestinal cells via ADP-ribosylation of the α-subunit of a heterotrimeric guanine nucleotide binding (G) protein Gs, the stimulatory G protein for adenylate cyclase (AC). After ribosylation, Gsα remains activated and causes a 100-fold increase in cAMP production (37), which can lead to excessive chloride secretion and a severe secretory diarrhea in humans (9).
The discovery of CT and its mechanism of interaction with the enterocyte as a prototypic bacterial exotoxin has yielded vital information about the pathophysiology of toxogenic diarrhea and CT intracellular signal transduction steps, including the role of G proteins in excessive chloride secretion (18, 19, 38). These studies have led to the discovery and characterization of a family of small GTP-binding proteins called ADP-ribosylation factors (ARFs) named for their ability to enhance the CT-catalyzed ADP-ribosylation of Gsα as an allosteric activator of the toxin (29, 30). In addition, ARFs have been recently recognized to play a critical role in vesicular membrane trafficking (30).
There is increasing evidence that an immature enterocyte interaction with bacteria may, in part, explain the increased susceptibility of neonates to infectious diarrhea (4, 14, 40). This laboratory has studied the excessive secretory response to CT in animal models and has concluded that the principal mechanism is mediated by a developmentally regulated inappropriate postreceptor response (5, 20, 41, 43). For example, a previous study has shown that Gsα expression is upregulated in preweaned compared with postweaned and adult rat enterocytes (41). This developmental change in Gsα might be a key factor in mediating the excessive secretory response to CT in the young rat (41).
Recently we have extended these observations to experimental models of human intestinal development (23, 25). In these studies, we have also provided evidence that CT induces an enhanced secretory reaction mediated in part by a developmental upregulation of the cAMP response in immature vs. mature human enterocytes (23), and this enhanced response to CT is largely due to an excessive uptake of CT by a developmentally regulated clathrin-endocytic pathway (25). In this study, we examined the cellular/molecular basis for CT signaling in the developing human gut and provide evidence that ARFs are involved.
MATERIALS AND METHODS
CT was purchased from Calbiochem (EMD, La Jolla, CA). Anti-clathrin heavy-chain (Cla-HC) antibody (mAb) was purchased from BD Transduction Laboratories (Palo Alto, CA). Anti-Gsα and anti-Gβγ-complex rabbit pAb (anti-Gβγ) were obtained from Calbiochem. Anti-ARF1 and 6 mAb were obtained from Abcam (Cambridge, MA). Anti-CT-A and B polyclonal antibodies were obtained from Dr. Wayne Lencer's Laboratory (Children's Hospital, Boston, MA). Horseradish peroxidase-conjugated goat anti-mouse and anti-rabbit antibodies are from Bio-Rad (Hercules, CA). Hemaglutinin (HA)-tagged full-length ARF1 and ARF6 plasmids were kindly provided by Dr. Jim Casanova (Department of Cell Biology, University xof Virginia, Charlottesville, VA). [32P]NAD was obtained from PerkinElmer (Waltham, MA). ARF inhibitory peptide (ARFI) was obtained from Biomol (Plymouth Meeting, PA). Human recombinant EGF, Trizol, SuperScript III Platinum SYBR Green One-Step qRT-PCR kits, Stealth siRNA duplex oligoribonucleotides for ARF1, 3, 4 and 6, and lipofectamine RNAi MAX were obtained from Invitrogen (Carlsbad, CA). All tissue culture media and reagents are obtained from Invitrogen-GIBCO. All other chemicals unless specified were purchased from Sigma-Aldrich (St. Louis, MO).
H4 (fetal) and T84 cells (adult) were used in this study. H4 cells have been used as an immature human intestine model in previous studies (6, 25, 31, 39). H4 cells were routinely maintained in DMEM supplemented with 10% FBS and human recombinant insulin (0.5 U/ml). In selected experiments, H4 cells were preincubated in H4 media containing 1 μM hydrocortisone (HC) for 5–7 days (H4/HC) before further treatment. T84, a well-characterized and highly differentiated human colon carcinoma cell line, has been used as a mature human intestinal model (6, 25, 31, 39). T84 cells (American Type Culture Collection, Rockville, MD) were maintained in DMEM-F12 Ham with 5% FBS. Cells were seeded on either 12-mm (24-well) or 33-mm (6-well) tissue culture plates or permeable inserts. T84 cells, grown on permeable inserts, were used when they became confluent and formed polarized monolayers and tight junctions (e.g., resistance Ω ≥ 1000). In some experiments, T84 cells, grown on permeable inserts, were used before cells reached confluence (subconfluent, Ω < 100). These cells have neither formed tight junctions nor polarized monolayers. CT was used at a 20 nM concentration (dose range 0.2–20 nM). All CT-stimulation experiments were done in serum-free HBSS supplemented with 0.25% bovine albumin.
Determination of the developmental Gsα and ARF gene expression.
The level of Gsα and ARF mRNA was measured in duplicate for each sample by quantitative RT-PCR (qRT-PCR) using a SYBR Green One-Step qRT-PCR kit. The level of Gsα and ARF mRNA expression was normalized to the standard GAPDH level for each sample.
Determination of CT-catalyzed ADP-ribosylation of Gsα.
In vitro [32P]ADP-ribosylation was carried out in H4 and T84 cell lysates using the method described by McKenzie (27). Briefly, H4 and T84 cells were harvested in lysis buffer containing 0.5% SDS. A sample (20 μg) of cell lysates was assayed in a 50-μl volume containing [32P]NAD (3 μM, 4 × 106 counts/min), potassium phosphate buffer (250 mM, pH 7.0), thymidine (20 mM), ATP (1 mM, pH 7.0), arginine hydrochloride (20 mM), and 100 μM GTP. CT was activated by incubation with 50 mM DTT for 60 min at room temperature and then added to the mixture at a final concentration of 50 μg/ml. The assay was carried out in a water bath at 37°C for 60 min and terminated by transfer to ice. Next, after in vitro ADP-ribosylation, the samples were prepared for SDS-PAGE by sodium deoxycholate/ trichloroacetic acid precipitation as detailed elsewhere (27) and visualized by autoradiograph. In selected experiment, 100 μg of fresh rat brain extract was added to the sample before the assay was carried out as described above.
Immunoprecipitation, SDS-PAGE, and Western blot (immunoblot analysis).
After experiments were completed, cells were rinsed with HBSS. Surface-bound toxin was removed with acidified HBSS (pH 2.5). Cells were lysed in Tris-buffered solution (4°C) containing 1% Triton X-100 (1% TTBS), 0.5% SDS, and protease inhibitors as described previously (25). In some experiments, Triton-soluble proteins were collected first, and the insoluble fraction was then collected in 1% TTBS containing 0.5% SDS. Both fractions (10 μg proteins) were analyzed by SDS-PAGE (4–20% linear gradient gel) and Western blotting. For immunoprecipitation, H4 and H4/HC cells were transfected with HA-tagged ARF1 and ARF6 expression vectors as well as the vector alone as a control. Seventy-two hours after transfection, cells were washed with HBSS(+) and then harvested in lysis buffer containing 50 mM Tris, pH 7.5, 150 mM NaCl, 1.0% Nonidet P-40, 10% glycerol, 0.5 mM PMSF, and a complete protease inhibitor tablet (Roche) and then extracted for 15 min on ice. Cell lysates were then incubated for 2 h (4°C) with monoclonal anti-HA antibody (1 μg/IP) to allow for binding of antibody to antigen. After incubation, 30 μl (∼50% slurry) of protein A Sepharose beads (Santa Cruz Biotechnology, Santa Cruz, CA) to the lysate and incubated on a rotating device at 4°C for 60 min. The protein A beads were collected by spinning briefly and washed with lysis buffer four times before mixing with SDS sample buffer containing reducing agent. The mixture was then boiled for 3–5 min before SDS-PAGE and Western blot analysis.
Briefly, lysates were resolved on a 4–20% linear gradient polyacrylamide gel and transferred onto nitrocellulose membranes. After being blocked with 5% nonfat milk solution, membranes were sequentially probed with appropriate dilution of primary (1:1,000) and secondary antibodies (1:5,000) and visualized by enhanced chemiluminescence (Bio-Rad). GAPDH was used as a housekeeping protein to control for equivalent loading.
Sucrose equilibrium density centrifugation.
Confluent monolayers of T84, H4, and H4/HC cells were used for isolation of detergent-insoluble glycolipid/rafts. All steps were completed at 4°C as described in detail elsewhere (2, 50). Briefly, cells were lysed in 1% Triton X-100 (1% TBS) and homogenized with 15 strokes in a Dounce homogenizer on ice. The subsequent homogenates were adjusted to a 40% sucrose concentration with an equal volume of 80% sucrose in TBS, layered under a linear 5–30% or discontinuous sucrose gradient, and centrifuged at 39,000 revolution/min for 16–20 h in a swinging bucket rotor (model SW 41). Ten fractions were collected sequentially from the top, and 20 μl of each fraction was analyzed by SDS-PAGE and Western blotting.
H4 and H4/HC cells were plated on 6- or 24-well plates at an optimal density and then transfected with small interference RNA (siRNA) using lipofectamine (Invitrogen). ARF1, ARF3, ARF4, and ARF6 (stealth select, Invitrogen) and a scrambled control siRNA (at a dose range 0.1–100 nM for initial study and optimal concentration afterward) were used in selected experiments. After transfection, cells were incubated for 48–72 h before further study.
Expression of HA-ARF1 and 6 DNA.
The expression vector (1 μg DNA) containing full-length HA-tagged ARF1 and ARF6 as well as vector alone were transfected into H4, H4/HC, and T84 cells using lipofectamine agent LTX according to manufacturer's recommendation. Seventy-two hours after transfection, cells were exposed to CT or EGF before cell lysates were harvested for cAMP and Western blot analysis.
We used cAMP levels as the functional indicator for the CT response in this study. Enzyme immunoassays (EIAs) were used to measure cAMP levels as described previously (23).
Data were analyzed by two-sided Student's t-test using Microsoft Excel and expressed as means ± SE (n = 4–6).
Differential expression and activation of Gsα.
Previous studies from our laboratory demonstrated that there is an increase in Gsα mRNA expression and an excessive Gsα ribosylation in preweaned compared with postweaned rat enterocytes (41). However, little is known about the gene expression and regulation of Gsα in human enterocytes. Before investigating the cellular mechanism for CT activation of Gsα, we examined its expression in H4 vs. T84 cells. No significant difference was noted in Gsα mRNA expression between H4 and T84 cells. Real-time, semi-quantitative PCR indicated that Gsα mRNA is highly expressed in both cells, but the expression level is slightly higher in T84 compared with H4 (Fig. 1A). Next, we examined the cellular distribution of Gsα. In H4 cells, Gsα is present principally in the detergent-soluble membrane (non-raft) fraction, whereas in T84 cells Gsα is largely present in the detergent-insoluble and lipid raft-associated membrane fractions (Fig. 1B). In addition, Gsα protein levels were higher in T84 compared with H4 cells in both cellular fractions (Fig. 1C). However, after CT exposure, Western blot analysis demonstrated a higher CT-induced Gsα activation in H4 compared with T84 cells, as indicated by a slower migration of the 45-kDa Gsα isoform on a 4–20% linear gradient SDS-PAGE gel attributable to the ADP-ribosylation of Gsα (Fig. 1C, arrow). It is hypothesized that Gsα activation is necessarily accompanied by subunit dissociation in situ. We next compared the Gβγ cellular distribution after CT intoxication in H4 vs. T84 cells. In H4 and T84 cells, Gβγ was principally localized in the detergent-soluble membrane fraction. Furthermore, upon CT exposure, there was a decrease of membrane-associated Gβγ in H4 cells but not in T84 cells (Fig. 1D).
Previously, we have shown a CT-induced excessive cAMP response in immature human enterocytes, indicating an excessive activation of AC (23). We have also shown that immature enterocytes have a significantly higher rate of CT endocytosis compared with mature enterocytes, and this may contribute to the excessive CT-catalyzed Gsα activation in these cells (25). We next investigated in T84 cells the CT dose responsiveness under subconfluent (nonpolarized) and confluent (polarized) conditions and compared this response to that in H4 cells with or without HC. Polarized T84 cells can form tight junctions and thus separate the apical from basolateral surface, whereas subconfluent T84 cells are not polarized and are thus without apical and basolateral segregation. This is similar to the condition of cultured H4 cells. As shown in Fig. 2, subconfluent, nonpolarized T84 cells demonstrated a significantly higher CT uptake than confluent, polarized T84 cells. However, there was no significant difference in the CT-induced cAMP response between these two cell types. This result is consistent with a previous report that, when CT is applied basolaterally, T84 cells demonstrated a higher CT uptake but not a greater secretory response compared with apical application of CT (19). In contrast, HC-treated H4 cells demonstrated a lower rate of CT uptake and a significantly lower cAMP response than H4 cells alone.
Other factors that modulate CT activation of Gsα.
First, using an in vitro ADP-ribosylation assay, we examined CT-catalyzed Gsα ribosylation in H4 and T84 cells compared with that reported for pre- and postweaned rat small intestinal enterocytes. Consistent with the previous report (41), we found significantly more Gsα ribosylation in pre- compared with postweaned and adult rat enterocytes (data not shown). However, using cell lysates in a cell-free ribosylation assay, we were unable to detect Gsα ribosylation in H4 cells, whereas we were able to observe a ribosylation of Gsα in T84 cells. These observations suggested that the CT-A subunit is capable of ADP-ribosylation of substrate analogs in vitro. However, it has been known that a number of additional membrane proteins and intracellular factors are needed to promote or augment the enzymatic activity in vivo (35). Accordingly, we next investigated whether the addition of rat brain extract that is enriched with ADP-ribosylation factors could facilitate ADP-ribosylation of Gsα as described in Ref. 29. Increased levels of Gsα-ribosylation in H4 and T84 (to a lesser degree) cells were noted when rat brain extract was added to the assay mixture (Fig 3A). This result suggested that cytosolic/membrane factors in rat brain extract facilitated the CT-mediated ADP-ribosylation of Gsα in immature human enterocytes. Since newly identified ARFs have been known to facilitate ADP-ribosylation of Gsα by CT in live cells (35), we next began to examine the role of ARFs in the CT response in H4 vs. T84 cells.
To examine whether ARFs were involved in the CT response, we used an ARFI to block its function. We observed that the ARF inhibitor significantly diminished the CT-induced cAMP response in H4 cells but not in T84 cells, suggesting a possible role of ARFs in mediating the CT response in immature enterocytes (Fig. 3B). In mammalian cells, there are six known ARF isoforms, which belong to three classes: ARF1, 2, and 3; ARF 4 and 5; and ARF6. To identify which class/classes ARFs were involved in CT activation of Gsα, we transfected H4 and T84 cells with siRNAs targeting human ARF genes 1, 3, 4, and 6 before exposure to CT. As shown in Fig. 3C, a knockdown of ARF1 and ARF6 significantly inhibited the CT-induced cAMP response in H4 but not in T84 cells. In contrast, neither ARF3 nor ARF4 knockdowns exhibited an effect on the CT response in either cell (Fig. 3C). Our RT-PCR data indicated that there is little ARF3 and ARF4 expression compared with ARF1 and 6 expression in H4 and T84 cells (Fig. 3D), and this might be the reason that we were not able to observe any difference in ARF3/4 knockdown experiments.
We next examined whether ARFs play a role in either CT uptake or Gsα activation in H4 cells. Western blot analysis demonstrated that knockdown of ARF1 or ARF6 using siRNA blocked CT-catalyzed ADP-ribosylation of Gsα compared with cells either maintained in media alone or transfected with control (scrambled) siRNA (Fig. 4). In these experiments, we demonstrated that knockdown of ARF1 and 6 proteins by siRNA (shown in Fig. 4A) downregulated CT-mediated Gsα activation. This is indicated by an absence of retarded migration of the 45-kDa Gsα isoform on the Western blot as shown in Fig. 4B. In addition, knockdown of ARF6 also decreased CT uptake in H4 cells. In contrast, transfection of ARF1 siRNA had no effect on CT uptake (Fig. 4C).
ARFs have been shown to be involved in the formation of clathrin-coated vesicles (44) and vesicular transport from the ER to the Golgi (49). Accordingly, we next examined the interaction of ARF1/ARF6 with the CT-endocytosis machinery. H4 cells were transfected with a HA-tagged full-length ARF1/ARF6 for 72 h before exposure to CT. Total cell lysates were immunoprecipitated with an anti-HA antibody before being fractionated on a SDS-PAGE and probed with an anti-Cla-HC antibody. Western blot analysis showed a coprecipitation of clathrin and ARFs in H4 cells that have been exposed to CT at 37°C (Fig. 5A). Since the EGF receptor (EGFR) is known to be internalized via a clathrin-mediated endocytosis upon ligand binding, we used EGFR endocytosis to examine the specificity of ARFs in regulating CT endocytosis and trafficking. As shown in Fig. 5B, there was no coprecipitation of clathrin and ARFs in H4 cells that have been incubated with EGF, suggesting that the ARF effect is specific for the CT response.
Collectively, these data suggest that ARFs might affect the CT response through complementary mechanisms, e.g., ARF1 had its effect principally on CT trafficking and signaling (Figs. 4B and 5A), whereas ARF6 plays a major role in CT endocytosis (Figs. 4C and 5A).
In summary, the CT response via Gsα signaling in fetal enterocytes may differ from that of mature enterocytes, and the difference may be related to ARF interaction with the CT-signaling process.
HC affects CT activation of Gsα in H4 cells.
In a previous study, we have shown that a known breast milk trophic factor, HC, can alter the excessive secretory response to CT in immature human fetal enterocytes (23) and can induce a maturation change in endocytosis (25). We concluded that this change in CT endocytosis was partially attributable to a HC-induced CT-GM1 lipid raft association through maturational changes in the plasma membrane phospholipid composition of immature enterocytes (24). In this study, we investigated whether HC affected the developmental control of CT-induced Gsα activation. H4 cells were preincubated with HC (1 μM) for 24 h or 5 days before CT exposure. As shown in Fig. 6A, pretreatment of H4 cells in HC (H4/HC) induced a Gsα lipid raft association when fractionated on a 5–30% sucrose gradient column. After H4/HC cells were challenged with CT, Western blot analysis demonstrated a decrease in Gsα activation only in cells that have been pretreated with HC for 5 days compared with H4 cells grown either in media or in HC for 24 h (Fig. 6B). This lack of activation was not due to a decrease in Gsα mRNA or protein expression. The gene expression analysis and real-time RT-PCR in H4 cells preincubated with HC for various lengths of time (12, 24, and 48 h to 7 days) demonstrated a slight increase of mRNA expression after HC treatment (Fig. 1A).
Next, we examined whether pretreatment of H4 cells with HC (5–7 days) could affect the role of ARFs in the CT response. We transfected H4/HC cells with siRNAs targeting either ARF1 or ARF6. In contrast to data shown in Fig. 3B, HC pretreatment diminished the ARF-dependent CT response measured as CT-induced cAMP production in these cells (Fig. 6C). Western blot analysis demonstrated a sufficient knockdown of ARF1/6 proteins in H4/HC cells (Fig. 6D).
CT is responsible for the devastating diarrheal syndrome characteristic of cholera. CT-induced secretory diarrhea occurs more commonly and severely in young infants than in older children and adults. It is the best-characterized disease mediated by the Gsα/cAMP signaling pathway. Previous studies from our laboratory have shown an increased cAMP response to CT in the small intestine of suckling compared with mature rats (43). This increased sensitivity of AC to CT is largely due to higher Gsα mRNA expression and increased ADP-ribosylation of one or both Gsα isoforms in the preweaned rat small intestine (41). We previously reported that immature human enterocytes exhibit an excessive secretory response to CT, and this response is, in part, due to a developmental difference in CT endocytosis (25). However, little is known about the developmental regulation of the CT signaling leading to Gsα activation in immature human enterocytes. In this study, we have examined CT signaling via Gsα activation and provide evidence to suggest that the molecular basis for this developmental change in host responsiveness might be related to a difference in the CT-endocytosis machinery utilized by immature human enterocytes. In addition, we provide evidence for a major role that ARF1/6 play in CT trafficking and signaling in immature human enterocytes.
Gsα is ubiquitously expressed in various tissues, and its cellular distribution changes under a wide range of metabolic conditions such as cellular differentiation, ontogenic development, aging, and various adaptive processes (10, 11, 45). Here, our results have shown a higher level of Gsα protein expression as well as in vitro CT-catalyzed ADP-ribosylation of Gsα in mature compared with immature enterocytes. However, this observation was in contrast with the observed increase of Gsα ADP-ribosylation in intact H4 vs. T84 cells. This inconsistency between cell-free and intact cell observations suggests that additional factor(s) maybe involved in CT signaling leading to Gsα activation. Furthermore, we compared the difference in the CT response in subconfluent vs. confluent (nonpolarized vs. polarized) T84 cells. Interestingly, subconfluent T84 cells showed an increase in CT uptake, yet no difference in the CT-induced cAMP response compared with confluent cells. This observation is in contrast to the observations made in H4 cells in which pretreatment with HC significantly decreased CT uptake (25) and diminished the cAMP response, respectively. Taken together, these data suggest that the observed difference in the CT response between H4 and T84 cells is likely attributable to a posttranscriptional modification of Gsα that leads to its excessive activation.
Recent developments have provided interesting speculations about the possible cellular or molecular basis for regulating CT induced Gsα signaling in developing human enterocytes. First, compartmentalized (segregated) Gsα may prevent CT access. A striking enrichment of Gsα in the Triton X-100-insoluble and lipid-raft fraction of the cell membrane was found in T84 and H4/HC cells, which was absent in H4 cells. Secondly, the involvement of Gβγ subunits in Gsα activation may differ. It has been proposed that the Gβγ subunits are a membrane anchor for Gsα, and dissociation/release of Gβγ from Gsα leads to activation of AC and elevation of cAMP (15, 48). In a cyc-s49 system, both Gsα and Gβγ had to be present in the membrane, and they had to be able to form a heterotrimer in order for CT to ADP-ribosylate Gsα (15). Moreover, it was reported that Gβγ plays an important role in targeting Gsα to the plasma membrane and is essential for binding and activating of the heterotrimeric G proteins by receptors in HEK293 cells (12, 13). In H4 cells, Gβγ subunits were mostly present in the Triton X-100-soluble membrane fraction similar to the distribution of Gsα (Fig. 1). We provide evidence that, upon CT exposure, there was a decrease in membrane Gβγ and an increase in CT-induced Gsα activation in H4 cells. However, in T84 cells Gβγ subunits were largely present in the Triton X-100-soluble membrane fraction, a location segregated from the lipid raft-associated Gsα. In this manner, pretreatment of H4 cells with HC can induce a maturational change in the plasma membrane (24) and an increased lipid raft association with Gsα that results in a decreased CT response. These observations suggest that in human fetal enterocytes the lack of lipid-raft association might provide an easier access for CTA1 to interact with Gsα. Furthermore, the association with Gβγ subunits in the plasma membrane may facilitate the CT-catalyzed ADP-ribosylation of Gsα, leading to AC activation and cAMP elevation. In contrast, in T84 cells, Gsα proteins are enriched in lipid rafts and segregated from Gβγ. This structure might restrict the access of CTA1 to Gsα, resulting in a reduced CT response. However, further studies are needed to determine whether interaction between Gsα and Gβγ at the plasma membrane and subsequent dissociation of Gβγ is important for Gsα activation and the CT response in human enterocytes.
CT-catalyzed ADP-ribosylation of Gsα and activation of AC are enhanced by a family of soluble or membrane factors termed ARFs (35, 46). Previous studies have supported the hypothesis that ARFs are allosteric activators of CTA1 (35). However, recent studies have demonstrated that ARF proteins are not only activators of CT but also critical components of intracellular vesicular transport processes (30, 36). Here, we provided evidence that ARF1 and as ARF6 are required for CT uptake and signaling in human fetal enterocytes (Fig. 4). Both ARF1 and ARF6 have been shown to participate in clathrin-mediated vesicle trafficking events, but they function at distinctly different steps in membrane trafficking (3, 44). For instance, ARF1 can function through the assembly of clathrin-coated proteins in the ER/Golgi membrane, in ER-Golgi and intra-Golgi transport, or in transport vesicles from the trans-Golgi network (28, 42). In contrast, ARF6 can play an important role in clathrin-mediated endocytosis (1, 33). Their function in membrane trafficking is the recruitment of membrane coat proteins including coatomer protein (22), clathrin, and clathrin adaptor proteins such as AP1, AP2, AP3, and AP4 complexes (17, 26, 32, 34). Some studies have shown that ARFs can also activate phosphatidylinositol 4-phosphate 5-kinase (PIP 5-kinase), an enzyme that generates PtdIns(4,5)P2 (3). Phosphoinositols [PtdIns(4,5)P2 and PtdIns(3,4,5)P3] have been shown to affect clathrin-mediated membrane trafficking (7, 8, 47). However, further studies are needed to clarify the possible interactions between ARF6 and clathrin-coated vesicles to elucidate the mechanism by which ARF6 regulates clathrin-mediated vesicular trafficking either by being directly involved in coat formation, by recruitment of clathrin adaptor protein (such as AP2), or by interaction with phospholipids particularly, PtdIns(4,5)P2 and PtdIns(3,4,5)P3.
ARF1 is required in both clathrin-dependent and -independent vesicular transport in the Golgi (26). We have shown in our preliminary results that ARF1 is required for CT-mediated AC activation in H4 and to a lesser degree in T84 cells (Fig. 2) because knocking down ARF1 expression in H4 cells resulted in a significantly diminished CT-induced cAMP response without affecting CT uptake. In contrast, there is only a slight decrease of cAMP response in T84 cells. It has been reported that ARF1 is required for the formation of clathrin-coated vesicles at the Golgi level in cells, whereas ARF6 has been involved in plasma membrane trafficking. This is consistent with our findings that, in human fetal immature enterocyte, ARF1 and ARF6 function at different levels in regulating clathrin-dependent CT trafficking or endocytosis, respectively (Figs. 4 and 5).
In this study, we show that an immature Gsα response exists in immature human fetal enterocytes leading to an excessive AC activation and cAMP response. We provide evidence that membrane/cellular factors, particularly ARF1 and ARF6, contribute to a critical role in regulating clathrin-mediated CT uptake, intracellular trafficking, and/or CT signaling in these immature enterocytes. Moreover, we provide preliminary evidence that trophic factors may modulate the immature Gsα response toward a more mature response. It is noteworthy that in human enterocytes there is a developmental alteration in the plasma membrane phospholipid content (24). A maturation of plasma membrane lipid composition may result in compartmentalization and distribution changes in many cellular factors such as heterotrimeric G proteins Gsα and βγ subunits, as well as ARF family proteins, leading to changes in CT endocytosis, intracellular trafficking, and eventually in signal transduction and effector responses. Collectively, these findings support our hypothesis that there is a developmentally controlled intestinal cellular response to bacterial exotoxins involving complex cellular events. One caveat in interpreting our studies is that T84 cells, which are an adult human colonic carcinoma epithelial cell line, do not perfectly model normal mature enterocytes. Therefore, it would be inappropriate to use T84 cells as a comparison for all factors that influence intestinal immaturity. However, previous studies comparing intestinal development in H4 cells and T84 cells have been validated in multiple systems including suckling rats, xenografted human fetal small intestine, and primary cultures of freshly isolated human small intestinal epithelial cells. We believe these studies will further our understanding of the development of host defense during postpartum gut maturation. A better understanding of these mechanisms may help to plan strategies for prevention.
This work was supported by National Institutes of Health Grants RO1-DK70260, R37-HD12437, PO1-DK33506, and P-30-DK 40561 to W. A. Walker.
- Copyright © 2009 the American Physiological Society