The CCR7 C-terminus is important for migration but not for endocytosis 54

Once the possibility that a putative signal for endocytosis is ubiquitylation was refuted, we generated three different deletions within the C-terminal region in order to detect, which amino acids play crucial roles in endocytosis and/or signalling. Figure 3.8 illustrates the three different mutants, all tagged with HA at the C-terminus. In MT-1 most of the C-terminal tail has been removed and the remaining lysine replaced by an arginine as shown in red. This has been made with the purpose of avoiding any possibility of ubiquitylation at the C-terminal part and to confirm our previous data.

The last 34 amino acids have been removed in the MT-2 mutant, whereas in MT-3 only the last 24 amino acids were deleted. Within these last 24 amino acids, the threonines and serines (shown in black) were reported to be phosphorylated after ligand binding (Kohout et al., 2004), indicating a possible role in receptor signalling and/or trafficking for these amino acids.

Figure 3.8. Scheme of the three CCR7 mutants. Arrows show the different sites of deletion within the C-terminus and where an HA tag has been inserted. In case of MT-1 a lysine has been replaced by an arginine, shown in red. Dark circles show possible sites of phosphorylation. Adapted from Kohout et al. (Kohout et al., 2004).

To characterize the properties of these mutants, 300-19 and HEK293 cells were stably transfected with each of the constructs. With 300-19 cells a chemotaxis assay towards both chemokines was performed (figure 3.9 A). MT-2 and MT-3 trasnfected cells migrated comparable to CCR7wt. Interestingly, MT-1 transfected cells did not migrate. After this finding, to investigate whether the receptor is still functional, calcium influx together with ERK activation after chemokine triggering was analyzed, showing that MT-1 was the only mutant where neither calcium influx nor ERK activation was found (data not shown). This indicates that amino acids between MT-1 and MT-2 are crucial for CCR7 signalling. Further site-directed mutations are required in order to localize precisely the amino acids responsible for this process.

Through microscopical and flow cytometric analysis, CCR7 expression at the plasma membrane was analyzed in the three mutants. Data showed no differences in the expression. The mutated receptors were able to reach the plasma membrane without being retained in the ER (data not shown). With this fact, trafficking of these mutants was analyzed through endocytosis/recycling assays. 300-19 cells stably expressing the three mutants were incubated with ELC or SLC for 30 min to internalize CCR7, and were then incubated with anti-CCR7 antibody or were left another hour at 37°C after washing in order to allow the receptor to recycle (figure 3.9 B). Upon ELC binding all CCR7 mutants were internalized showing the same pattern as CCR7wt. Similar result were obtained in recycling assays, where no difference between CCR7 mutants and CCR7wt was found (figure 3.9 B). To confirm this data the same experiment was performed in HEK293 cells stably transfected, which showed the same result as described before (data not shown).

To conclude these results, two important points can be made: firstly, endocytosis/recycling seems to be an independent process of chemotaxis. Secondly the signals for endocytosis must be situated in a intracellular loop, probably where the DRY motif is located, as has been reported for other chemokine receptors and other GPCRs (Arora et al., 1997; Hawtin, 2005).

Figure 3.9. Characterization of CCR7 mutants. (A) 300-19 cells stably expressing one of the mutated receptors or the wt respectively were subject to a Transwell^TM chemotaxis assay. Cells migrated towards ELC and SLC (0.25µg/ml) as well as cells in the input were counted by flow cytometry during an interval of 120s. As a control for spontaneous migration, 300-19 CCR7-HA cells were assayed without chemokine. Values represent the mean of three independent experiments ± SEM. (B) Receptor surface staining of 300-19 cells stably expressing CCR7-HA and the mutants respectively with an anti-CCR7 antibody were performed after endocytosis of either ligand (2µg/ml) or after allowing 1h recycling of the internalized receptor at 37ºC. Total fluorescence was detected by flow cytometry. Non-transfected cells were used as a control to show receptor independent effects. Values represent the mean of three independent experiments ± SEM. (nc: no chemokine).

Another way to answer the question whether endocytosis/recycling is required for chemotaxis was the transfection of dominant negative proteins of Rab GTPases (Etienne-Manneville and Hall, 2002). In particular, we wanted to know whether the recycling of CCR7 is necessary for migration. For this approach we used a dominant negative form of Rab11, which blocks the receptor from going back to the plasma

membrane (Ren et al., 1998). CEM cells, which express CCR7 endogenously, were transiently transfected with Rab11 DN-GFP. The transfection efficiency, which was relatively low, was monitored with GFP expression by FACS. Cells were then subjected to a Transwell^TM chemotaxis assay towards ELC (figure 3.10). After 3 hrs, cells were analyzed by FACS. Almost 10% of the migrated cells (8% from total cells) expressed Rab11 DN-GFP. This clearly shows that cells containing Rab11DN were able to migrate, indicating that probably CCR7 recycling is not necessary for migration. To confirm that transiently transfected cells are able to migrate, the same experiment was performed with YFP alone.

Figure 3.10. CCR7 recycling is not necessary for chemotaxis. Migration assay of CEM cells transiently transfected with Rab11 DN-GFP and YFP, towards ELC (250ng/ml). On the left panel the number of live cells is indicated and on the right panel the number of transfected cells. This is a representative result of three different experiments.

3.3.4 Generation of recombinant SLC-Flag and SLC-Fc

To elucidate which proteins are involved in the CCR7 signalling complex after ligand binding, we generated a recombinant SLC tagged with a Flag tag or with the Fc part of IgG in order to precipitate the whole signalling complex.

Both recombinant proteins were stably transfected into HEK293 cells and after cloning by limiting dilution, the cell supernatant was collected and analyzed by western blot (figure 3.11) or by immunofluorescence with the respective antibodies (data not shown). In the case of SLC-Fc, the purification was much easier (only protein A column was required) and functional analyses (Chapter 1) were performed, showing that this recombinant protein was fully functional.

Figure 3.11. Purification of SLC-Flag. Supernatant of HEK293 cells stably transfected with SLC-Flag or HEK293 WT were immunoprecipitated with anti-Flag M2 agarose beads. Then a western blot with anti-Flag antibody was performed.

Purity of SLC-Fc was analyzed by SDS-PAGE followed by Coomassie staining (figure 3.12). SLC-Fc was found as a monomer and a dimer with the expected molecular masses. The recombinant protein was enriched in fraction 4, with a concentration of 1 mg/ml.

.

Figure 3.12. SLC-Fc purification. A 500 ml volume of supernatant contained SLC-Fc, obtained from HEK293 cells stably transfected with SLC-Fc was bound overnight onto a Protein A HiTrap column.

Then the column was washed and SLC-Fc was eluted with a glycine acidic buffer. Different fractions were collected and an SDS-PAGE followed by Coomasie staining was performed. SLC-Fc was found in a larger amount in fraction 4 with a concentration of 1 mg/ml.

Using SLC-Fc we next investigated whether CCR7 can be immunoprecipitated by its ligand. To this end, HEK293 cells grown on ³⁵S Cys/Met medium stably expressing vsvCCR7 were incubated with the recombinant chemokine.

Afterwards, the complex was precipitated with protein A beads (figure 3.13).

Unfortunately, just a slight smear with a higher MW (~55 KD) has been found compared to untransfected cells. More experiments will be required to establish a better technique in order to facilitate the comparison with ELC-Fc.

Figure 3.13. Precipitation of vsvCCR7 complex with SLC-Fc. HEK293 WT cells and stably transfected with vsvCCR7 were incubated with ³⁵S Cys/Met for 2 hrs at 37°C. Then the radioactivity was removed and the cells were left for 4 hrs at 37°C with fresh medium. Cells were then incubated for 30 min with SLC-Fc. An overnight precipitation from cell lysis was performed with protein A. The day after, beads were washed three times with wash buffer and the protein was eluted with Laemmli Sample Buffer. To resolve the proteins, an SDS-PAGE was performed, and then the gel was dried followed by exposure to an X ray film (1 week).

3.4. Discussion

In this study we demonstrated that the C-terminal tail of CCR7 is neither required for its plasma membrane localization nor for receptor endocytosis. However this region is extremely important for a proper function of CCR7 in cellular migration.

We also showed that CCR7 is not ubiquitylated even upon ligand binding as it has been reported for the chemokine receptor CXCR4. Nevertheless, using HA-CXCR4 as a positive control we were not able to reproduce this data following exactly the same protocol as described by Marchese et al (Marchese and Benovic, 2001) (data not shown). Immunoprecipitation and transfection efficiency were controlled, performing the same experiment but using an anti-HA antibody during the western blot in order to detect the receptor under these conditions. In both cases the receptor was easily detected as is shown in figure 3.7 B. We still have no plausible explanation why we were not able to detect an ubiquitylation of CXCR4. In the case of CCR7 one could have expected to find such a result, due to the fact that no phenotype was found with the CCR7-3K3R mutants (figure 3.4, A and B). Moreover, transient transfection experiments of an E3 ligase mutant, the enzyme that is responsible for the ubiquitylation of CXCR4 (Marchese et al., 2003), in cells expressing vsvCCR7 showed no effect on CCR7 trafficking as examined by confocal microscopy (data not shown).

Nevertheless there are four more lysines, which are possible sites for ubiquitylation of CCR7, two in the first intracellular loop and two in the third one, as shown in figure 3.2. Site directed mutagenesis would be required to exclude any possible site for ubiquitylation in order to confirm our postulate.

It is still a matter of debate whether endocytosis/recycling and chemotaxis are related or not, because of many reports describing both alternatives. Our data showed that these processes are independent. In the case of recycling we could observe, that cells that endogenously express CCR7, when transfected with a dominant negative form of Rab11, were still able to migrate (figure 3.10). This result indicates that CCR7 recycling is not necessary for migration. If recycling is completely blocked, it could well be, that after ligand induced receptor endocytosis there is still some

receptor left at the plasma membrane. The remaining receptor could bind to the ligand and may transduce the signal, resulting in migration (shown in chapter 1). Another explanation would be that recycling of the receptor is not required for chemotaxis at all. Nevertheless, a proper positive control to show that this dominant negative form of Rab11 is working in our experiments is missing.

Two main endosomal recycling pathways have been described: a slow rate recycling, where Rab11 is involved and a rapid recycling process dependent on Rab4 and PI3K (Sonnichsen et al., 2000). In the case of CXCR2 it has been described that this receptor recycles via Rab11 (Ullrich et al., 1996). However, for CCR7 we still do not know whether this receptor employs this pathway. If this is not the case, a dominant negative Rab11 will have never an effect on the recycling of CCR7. Further experiments will be required to clarify this point.

Nevertheless, studies on other chemokine receptors have demonstrated that recycling is important for chemotaxis. This is the case for CXCR2 and CXCR4, where the expression of the protein myosin Vb, which interacts with Rab11, in a mutant form that lacks the motor domain, impaired migration mediated by these two chemokine receptors (Fan et al., 2003). This study demonstrates that recycling could indeed have an important role on the chemokine receptor function. However, on the other hand, experiments on CXCR1 and CXCR2 have revealed that endocytosis is not required for chemotaxis (Richardson et al., 2003; Rose et al., 2004).

Of interesting is our mutant CCR7 MT-1, where the whole C-terminal region was removed (figure 3.8). Unexpectedly, this mutant could easily reach the plasma membrane (data not shown) being endocytozed and recycled like CCR7wt. This fact is remarkable because for most chemokine receptors it is known that when their cytoplasmatic tail was removed they showed impaired endocytosis behaviour as it has been demonstrated for CCR5 (Kraft et al., 2001), CXCR4 (Roland et al., 2003), CXCR1 and CXCR2 (Richardson et al., 2003). Despite the normal endocytosis behaviour of MT-1 transfected cells, they were not able to migrate as shown in figure 3.9 A. This data clearly showed that the signals for endocytosis and recycling are not contained in the C-terminal region of the receptor, where the signals for signal transduction are localized.

Studies on the CCR7 C-terminus showed that serines and threonines, as shown in figure 3.8, are phosphorylated upon ligand binding, confirming that this region is

extremely important for the function of CCR7 (Kohout et al., 2004). However, this report showed that these amino acids are situated at the very end of the C-terminal tail. These data are in contradiction with our results, which showed that mutant MT-3 that has no possible sites for phosphorylation, is fully functional, as well as MT-2 where even more C-terminal amino acids were removed. In conclusion, our data clearly indicate that the pivotal amino acids are situated between MT-1 and MT-2 mutants, where no possible phosphorylation sites are located.

Our data was also confirmed by calcium flux experiments that revealed that CCR7 MT-1 was not functional any more (data not shown). Preliminary experiments with β-arrestin 2-GFP, performed with confocal microscopy showed that there was no colocalization of MT-1 with this protein (data not shown). β-arrestin 2 in this case seems to have no role on receptor endocytosis, but on signal transduction through C-terminal binding, as has been reported by other authors as well (Huttenrauch et al., 2002; Huttenrauch et al., 2005). Moreover, experiments on different mutants of formyl peptide receptor (FPR) showed that β- arrestin 2 played no role in the ERK1/2 activation through FPR, whereas the G protein pathway was essential (Gripentrog and Miettinen, 2005).

Concerning the classical motif described for receptor endocytosis, studies on different chemokine receptors have shown that some of them contain these motives.

In the case of the chemokine receptor CXCR2, which contains a leucine-based LLKIL motif in its C-terminus, this motif was identified as a determinant for endocytosis upon ligand binding (Fan et al., 2001). In the C-terminal region of CCR5 and CXCR4 receptors dileucine motifs are also present. This sequence was shown to be essential for the direct coupling of AP-2 adaptor, thus being important for the internalization process via clathrin coated pits (Kraft et al., 2001; Orsini et al., 1999). The C-terminal tail of CCR7 contains 5 leucines but they are not arranged in pairs, not representing a known motif that could be responsible for the internalization of CCR7. Moreover CCR7 MT-1 was efficiently endocytosed upon ligand binding although none of these leucines were present.

Further studies will be necessary to identify amino acids responsible for receptor endocytosis. At least now we know that these amino acids are not present in the C-terminal tail and that endocytosis/ recycling can occur independently of

chemotaxis. However, we still cannot tell whether recycling is necessary for migration or not.

Similar studies on chemokine receptors like CCR5 and CXCR3 showed abrogated surface expression when the C-terminal region was removed. On the other hand CXCR4 and CXCR1, like CCR7, did not require their cytoplasmatic tail to be expressed at the cell surface (Venkatesan et al., 2001). Although all chemokine receptors have the same function their behaviour and structures vary markedly (Colvin et al., 2004). Furthermore it is known that chemokine receptors can modulate the actin cytoskeleton in a way that the cell is able to migrate. Like all chemokine receptors, CCR7 also contains the DRY motif that is present in most GPCR at the cytoplasmatic side of the second intracellular loop, which is indispensable for G-protein coupling. Concerning this, it will be very interesting to observe via site directed mutations whether there is an effect on migration and endocytosis of CCR7, as is the case for CCR5 and CXCR4 which require the DRY motif for directed chemotaxis (Lagane et al., 2005; Roland et al., 2003). Moreover, G protein activation may not be required for endocytosis as has been shown for CXCR2 (Feniger-Barish et al., 2000). Presumably, CCR7-mediated chemotaxis would also be regulated by this conserved domain, whether it is essential for endocytosis remains unclear.

In order to study CCR7 signal transduction we also tagged CCR7 at the N-terminus with a vsv tag, as is shown in figure 3.1 A. Our results showed that CCR7 was functional and normally expressed at the plasma membrane (data not shown) but for an unknown reason the tag was not detectable. In addition to HA-CXCR4, another example of an N-terminally tagged chemokine receptor is GFP-CCR5, which is fully functional and the GFP is detectable (Gomez-Mouton et al., 2004). Further experiments will be required to elucidate whether there is any N-terminal processing.

If this is the case, this possible N-terminal processing will add to the uniqueness of CCR7.

Nevertheless, using HEK293 cells stably expressing vsvCCR7 and recombinant SLC-Fc we were able to perform receptor precipitation experiments.

However, our results were not conclusive. Additional experiments will be required to improve this technique.