Gliadin peptide P31-43 localises to endocytic vesicles and interferes with their maturation

Abstract

Background: Celiac Disease (CD) is both a frequent disease (1:100) and an interesting model of a disease induced by food. It consists in an immunogenic reaction to wheat gluten and glutenins that has been found to arise in a specific genetic background; however, this reaction is still only partially understood. Activation of innate immunity by gliadin peptides is an important component of the early events of the disease. In particular the so-called "toxic" A-gliadin peptide P31-43 induces several pleiotropic effects including Epidermal Growth Factor Receptor (EGFR)-dependent actin remodelling and proliferation in cultured cell lines and in enterocytes from CD patients. These effects are mediated by delayed EGFR degradation and prolonged EGFR activation in endocytic vesicles. In the present study we investigated the effects of gliadin peptides on the trafficking and maturation of endocytic vesicles.

Methods/principal findings: Both P31-43 and the control P57-68 peptide labelled with fluorochromes were found to enter CaCo-2 cells and interact with the endocytic compartment in pulse and chase, time-lapse, experiments. P31-43 was localised to vesicles carrying early endocytic markers at time points when P57-68-carrying vesicles mature into late endosomes. In time-lapse experiments the trafficking of P31-43-labelled vesicles was delayed, regardless of the cargo they were carrying. Furthermore in celiac enterocytes, from cultured duodenal biopsies, P31-43 trafficking is delayed in early endocytic vesicles. A sequence similarity search revealed that P31-43 is strikingly similar to Hrs, a key molecule regulating endocytic maturation. A-gliadin peptide P31-43 interfered with Hrs correct localisation to early endosomes as revealed by western blot and immunofluorescence microscopy.

Conclusions: P31-43 and P57-68 enter cells by endocytosis. Only P31-43 localises at the endocytic membranes and delays vesicle trafficking by interfering with Hrs-mediated maturation to late endosomes in cells and intestinal biopsies. Consequently, in P31-43-treated cells, Receptor Tyrosine Kinase (RTK) activation is extended. This finding may explain the role played by gliadin peptides in inducing proliferation and other effects in enterocytes from CD biopsies.

Figure 1. Vesicles interacting with P31-43-liss are early endocytic vesicles after a 3 h chase.

In CaCo-2 cells, both after 30 minutes of pulse and 3 hours of chase with P31-43-liss (red), the peptide interacted with vesicles that are positive for EEA1 (green staining, top and middle panel). P31-43-liss did not co- localise with LAMP2-positive vesicles (green staining, bottom panel) after 3 hours of chase. A merge of the red and green panels of the EEA1 or LAMP2 and P31-43-liss or P57-68-liss staining is shown, the yellow/orange colour indicates co- localisation. The zoom panels represent a digital 4× enlargement of the region highlighted by white lines in the merge panels. The co- localisation coefficient was calculated as reported in the “Methods” section. The results are representative of four independent experiments.

Figure 2. Vesicles interacting with P57-68-liss are late endocytic vesicles after a 3 h chase.

At 30 minutes pulse, P57-68-liss co- localised with EEA1-positive vesicles (top panel); however after 3 hours of chase P57-68-liss no longer co- localised with EEA1-positive vesicles (middle panel). After 3 hours of chase, P57-68-liss co- localised with LAMP2-positive vesicles (bottom panel). Merge of the red and green panels of the EEA1 or LAMP2 and P31-43-liss or P57-68-liss staining is shown, yellow/orange colour indicates co- localisation. The zoom panels represent a digital 4× enlargement of the region highlighted by white lines in the merge panels. The co- localisation coefficient was calculated as reported under “Methods”. The results are representative of four independent experiments.

Figure 3. P31-43-liss-carrying vesicles are slower than P57-68-liss-carrying vesicles.

Statistical analysis of three experiments performed resulted as follows: live CaCo2 cells pulsed for 30 minutes and chased for 3 hours with lissamine-labelled peptides were used for time-lapse experiments in which we acquired images every 30 seconds for 10 minutes at the indicated times. The image stacks were assembled to produce a video of vesicles dynamics. In each experiment, the position of at least 25 vesicles per cell was recorded and reconstructed to mark the trajectories of each vesicle during the observation time. The speed of vesicles was calculated by averaging the trajectories produced in 10 minutes at the indicated times. Bars represent mean and standard deviation. Asterisks indicate P<0.05 (Student's t-test). P31-43 carrying vesicles both at 30 minutes of pulse and 3 hours of chase are statistically significantly slower than P57-68 carrying vesicles.

Figure 4. P31-43 delays endocytosis regardless of vesicles cargo.

A) Dextran-Alexa-488-carrying vesicles moved faster than dextran-Alexa-488- and P31-43-liss-carrying vesicles. Live CaCo2 cells were pulsed for 30 minutes with dextran-Alexa-488 with and without P31-43-liss or P57-68-liss then chased for 3 hours in time lapse experiments. Only the 30 minutes pulse experiments are shown. The histogram shows the statistical analysis of three experiments. We calculated the speed of vesicles by averaging the trajectories produced in 10 minutes as in Figure 2. Bars represent mean and standard deviation. P31-43-liss and Dextran-Alexa-488 co-localised in the vesicular compartment. B) Live CaCo2 were pulsed for 30 minutes and chased for 3 hours (not shown) with EGF-Alexa-488 conjugated with and without P31-43-liss or P57-68-liss. Only the 30 minutes pulse experiments are shown. The histogram on the right side shows the statistical analysis of three experiments. We calculated the speed of vesicles by averaging the trajectories produced in 10 minutes as in Figure 2. Bars represent mean and standard deviation. P31-43-liss co- localises with EGF-Alexa 488 in the vesicular compartment. Asterisk indicates P<0.05 (student's t-test). The results show that P31-43 carrying vesicles are slower regardless of the cargo they are carrying.

Figure 5. P31-43 is similar to Hrs and is localised at the membrane vesicles.

A) Multiple alignments of gliadin peptides P31-43 and P31-49 are shown, with Hrs from mouse, rat, human and Drosophila. Numbers on the right side of the figure represent the terminal amino acid position of each sequence shown. Of the 13 residues in P31-43, 7 are identical (red) and 2 similar (green) to the corresponding residues of human Hrs. B) Enlarged vesicles visible by multiple digital enlargement (8×) of the cytosol of a single cell. Hrs-EGFP was transfected for 48 hours in CaCo2 cells. P31-43 or P57-68 was added for 15 minutes. On the right side is the profile of the red and green channels along the white arrows that run across a single vesicle in the figure. The figure is representative of four similar independent experiments. Enlargement of the endocytic vesicles allow the observation of the vesicle membrane. In a short time (15 minutes) only P31-43 but not P57-68 co- localises with Hrs-EGFP at the vesicle membrane.

Figure 6. P31-43 competes with Hrs localisation to the endocytic vesicles.

A) Western blot analysis of endogenous Hrs after separation by ultracentrifugation of cytosolic and membrane proteins. EGFR was used as a control for membrane fraction and tubulin as a control for cytosol. The last line on the right is a lysate of CaCo-2 cells transfected with Hrs as a size control for the protein. The results are representative of three independent experiments. B) Densitometric analysis. The Hrs concentration was normalised to a control protein, namely tubulin in the cytosolic fraction and EGFR in the membrane fraction. Mean and SD of three independent experiments is shown. Asterisk indicates P<0.05 (student's t-test). After P31-43 treatment, Hrs increases in the cytosolic fraction and decreases in the membrane fraction in comparison to the not treated sample in a statistically significant way. C) Confocal analysis of CaCo-2 cells transfected for 24 h with Hrs-EGFP, not treated and treated with P57-68 and P31-43 for the last 3 h of transfection. The results are representative of four independent experiments.

Figure 7. Hrs competes with the effects of P31-43 on the G0 >S transition.

Bromodeoxyuridine (BrdU) incorporation of CaCo-2 cells transfected or not with Hrs-EGFP cDNA and treated as indicated 24 h after transfection. The bars represent the fraction of BrdU incorporating cells as a percent of total cells and are the mean ± SD of three independent experiments. Asterisks indicate P<0.05 (Student's t-test) with respect to the 0,1% FCS. These results indicate that Hrs compete with the effects exerted by P31-43 on cell proliferation.

Figure 8. After a 24-h chase, gliadin peptide P31-43-CY3 is still present in epithelial cells of crypts of celiac disease patients, but not of controls.

Intestinal biopsies from control subjects (A) and celiac disease patients on a gluten-containing diet (B) were cultivated with CY3-labelled P31-43 for 3 h and then chased for 24 h as indicated. Thin sections of the cultivated biopsies were then stained with anti-EEA1 antibodies. In control sections (A), P31-43-CY3 was visible only after 3-h pulse, but not after a 24 h chase. In the section of a celiac disease patient, P31-43-CY3 was present at both 3 h and 24 h in the epithelial cells of crypts. Overlay panels show that in cultivated biopsies from celiac patients at any time and in controls only at 3 h pulse, the P31-43 peptide co- localised with EEA1. White arrows indicates localisation of peptide in the early endocytic vesicles. Representative results from 3 independent experiments are shown.

Figure 9. Overview of the effects of P31-43 on the endocytic pathway.

Endocytosis has many effects on signalling: in fact, signalling pathways and endocytic pathways are regulated in a reciprocal manner. It is now widely accepted that the “Endocytic Matrix” is a master organiser of signalling, governing the resolution of signals in space and time. Consequently endocytosis affects several cell functions that range from proliferation to cell motility . Growing evidences [22, 33, 34 and the present paper], point to an effect of certain gliadin peptides (i.e. P31-43) on the endocytic compartment. By interfering with Hrs localisation to the endocytic membranes, P31-43 induces two important effects: a) it delays endocytic maturation, and b) it alters the recycling pathway. By delaying the maturation of endocytic vesicles P31-43 reduces EGFR and other RTK degradation and prolongs their activation which in turn results in increased proliferation, actin modification and other biological effects. The alteration of the recycling pathway is able to direct more transferrin receptor and likely other recycling receptors such as IL15 to the membranes.