CCR7 is localized in lipid rafts

Figure 4.1. Membrane extraction preparation of CCR7-GFP. HEK293 cells stably transfected with CCR7-GFP were lyzed with hypotonic Tris 50mM solution and centrifuged in order to separate the total membranes from the rest of the cell lysate. Total membrane extraction was loaded on an SDS-PAGE followed by western blot. CCR7-GFP was detected with an anti-GFP antibody. The same procedure was performed with HEK293 WT cells as negative control.

4.3.2 CCR7 is localized in lipid rafts

In order to investigate lipid raft localization of CCR7, we performed a 1%

Triton X-100 extraction at 4°C, where proteins that are situated in these raft compartments remain insoluble (Harder et al., 1998). To differentiate soluble from insoluble fractions, a discontinuous sucrose gradient was performed. After 20 hrs of centrifugation, samples were collected and total proteins were precipitated with 10%

TCA. Then an SDS-PAGE followed by a western blot was performed with an anti-GFP antibody (figure 4.2). However, despite the correct MW for CCR7-anti-GFP, curiously this protein was found only in the middle of the gradient (fraction 7), precisely where the soluble or “non-raft” fractions starts. An explanation for this finding would be the resuspension of total protein after TCA precipitation, since the amount of proteins in fractions 8-12 was very high and to resuspend the proteins in sample buffer was very difficult. So probably proteins in fraction 7 were much more easily to solubilise. Despite the little amount of protein in the insoluble fractions, CCR7-GFP was not found in lipid rafts.

Through this procedure, CCR7-GFP could always be detected with the anti-GFP antibody with the correct MW, however this antibody also recognized specifically another protein of 30KDa. Surprisingly, this band was also described in another western blot with the chemokine receptor CXCR1-GFP (Jiao et al., 2005).

Figure 4.2. Lipid rafts preparation with CCR7-GFP. HEK293 cells stably transfected with CCR7-GFP were lyzed in 1% Triton X-100 MN buffer. Soluble from insoluble proteins were separated as described in “Materials and Methods” by a sucrose gradient. Afterwards samples were collected and proteins were precipitated with 10% TCA. Proteins were loaded on an SDS-PAGE gel and a western blot with anti-GFP antibody was performed. “Raft fractions” should be between fractions 3-5 and

“non-rafts” between 7-12.

To exclude that GFP disturbed CCR7-lipid raft association, we performed the same experiment but using this time CCR7 with an HA tag at the C-terminus. Now we were able to detect CCR7-HA in the soluble fractions (figure 4.3 A) and a considerable portion of CCR7 in the insoluble fractions, clearly indicating that

CCR7-HA is situated in lipid rafts. Unfortunately, in fraction 9 it seems that the resolution of the TCA-protein precipitate was not successful because no CCR7-HA band was found. As a control for lipid raft association, directly coupled CTx-HRP was used to detect the lipid raft ganglioside GM1 in a dot blot (figure 4.3 B). With the same gradients a western blot was performed with anti-calnexin antibody, because calnexin is an ER protein known not be a constituent of lipid rafts (figure 4.3 C).

Figure 4.3. CCR7-HA is in lipid rafts. (A) HEK293 cells stably expressing CCR7-HA were lysed in 1% Triton X-100 and subjected to sucrose density gradient centrifugation. Proteins from equal volumes of collected representative fractions were precipitated with TCA 10%, separated by SDS-PAGE and analyzed by Western blotting using specific antibodies against HA and calnexin (C). To analyze distribution of the ganglioside GM1, 5 µl of each fraction were dot blotted onto a nitrocellulose membrane and detected using CTx^HRP (B). Total protein distribution was monitored by Ponceau red staining. This is one representative figure of four experiments.

Subsequently, we repeated this experiment but in addition incubated the cells with ELC (2µg/ml) for 30 min at 37°C, following the hypothesis that we may find more CCR7 in these signalling platforms compared to unstimulated conditions.

However, no difference in lipid raft partitioning was found (data not shown).

Another approach, to confirm that CCR7 is indeed in lipid rafts, was through immunofluorescence with 1% Tx-100 extraction at 4°C. Proteins situated in plasma membranes or intracellular vesicles containing lipid rafts are resistant to Triton extraction, remaining in the cell, whereas proteins that are not raft resident, would be removed (Damm et al., 2005). HEK293 cells stably expressing CCR7-HA were incubated with ELC for 30 min at 37°C (figure 4.4), cells were fixed and 1% Triton extraction was performed at 4°C for 5 min. Then cells were incubated with anti-CCR7 antibody (red) or with CTx-FITC (green) in order to detect lipid rafts domains. After this procedure cells were not in a very nice shape, nevertheless we were able to detect CCR7 after Triton extraction, mostly in vesicles or endosomes and almost not colocalizing with CTx. Interestingly, cells without Triton extraction, showed almost complete colocalization. As control, Alexa Fluor 546-labeled transferrin, the receptor of which is endocytozed by the classical clathrin-coated pit pathway without any involvement of lipid rafts, was used. With this data we can conclude that despite CCR7 endocytosis, which has been described to occur by clathrin-coated pits, probably a minor share of CCR7 receptors may also be internalized via caveolae/lipid rafts.

Figure 4.4. Association of CCR7 with detergent-insoluble membranes. HEK293 cells stably transfected with vsvCCR7 were incubated with ELC (2µg/ml) (A and B) or Alexa Fluor 546-labeled transferrin (50 µg/ml) (C and D) for 30 min. Then, cells were washed, left untreated (A and C) or extracted for 5 min on ice with 1% Triton X-100 (B and D), and fixed. After fixation, cells were incubated with an anti-CCR7 antibody followed by an anti-rat Cy3 (red) antibody and by CTx-FITC (green) to detect GM1 ganglioside. Images were analyzed by confocal microscopy.

To assess whether lipid rafts play a role in migration, we performed a chemotaxis assay with CEM cells that express CCR7 endogenously, towards ELC in the presence/absence of methyl-β-cyclodextrine (MCD). This drug disrupts lipid rafts by removing cholesterol contained in the plasma membrane. Cells were washed several times in order to remove free cholesterol that is contained in FCS. After MCD treatment for 30 min, lipid free BSA was added. After this treatment most cells were alive and in a good shape. Then, a chemotaxis assay was performed, showing that cells treated with 5mM MCD were not able to migrate (figure 4.5). After migration cells were analyzed with Trypan Blue in order to quantify living cells, revealing that untreated cells were all alive and of the MCD treated cells, 65% were alive and 35%

were dead. This data clearly shows the impact of lipid rafts on chemokine-mediated migration.

Figure 4.5. Effect of MCD on chemotaxis. 300-19 cells stably expressing CCR7-HA were treated/untreated with 5mM MCD for 30 min. Afterwards cells were washed and a chemotaxis assay was performed with a Transwell^TM filter towards ELC (250ng/ml), where indicated. This is a representative figure of three different experiments.

To check the possibility that signal transduction via lipid rafts is blocked after MCD treatment, we analyzed calcium flux triggered after chemokine induction (figure 4.6). Under the same conditions described before, we were not able to find any

response when cells were cholesterol depleted, strongly suggesting the requirement of lipid rafts for signalling and migration.

As a control, TCR (T Cell Receptor) signalling that has been reported to rely on lipid rafts was analyzed following the same procedure as described before. CEM cells containing TCR endogenously were incubated with anti CD3 antibody OKT3 in order to activate the receptor (Burack et al., 2002). As expected, no signal was detected with 5mM MCD treatment. In both cases, when cells were incubated with only 1 mM MCD, receptors were able to activate signalling after ligand binding, showing that probably, this concentration was not high enough to disrupt lipid rafts.

Figure 4.6. Effect of MCD on CCR7 signalling. CEM cells were treated/untreated with 1 mM or 5mM MCD for 30 min. Subsequently, cells were washed and incubated with Ca⁺² buffer with Fluo-3/AM for 30 min. Then, calcium influx was measured in response to ELC (1µg/ml) or to OKT3 antibody (5µg/ml). Ionomycin was used as positive control.

One argument against these experiments is that MCD could affect the conformation of the receptor affecting the binding of its ligand as shown for CCR5 (Nguyen and Taub, 2002). For that reason following the same procedure as before, immunofluorescence microscopy of CEM cells was performed with ELC at 4°C.

Through confocal analysis we could observe that there was no difference between treated/untreated cells (data not shown).