The interaction between the uPAR and integrins is highly relevant in the progression of many types of cancer, being these proteins key players in processes such as cell migration, invasion and metastasis. uPAR, a glycosylphosphatidylinositol-anchored cell surface receptor, binds to uPA, activating proteolytic cascades that degrade the ECM, facilitating tumor cell invasion. In parallel, uPAR can trigger intracellular signaling, thereby modulating physiological processes such as wound healing, immune responses and stem cell mobilization, as well as pathological events including inflammation and tumor progression [38, 39]. Integrins and integrin-dependent processes play crucial roles in nearly every phase of cancer progression, actively participating in signaling pathways that promote cell survival, proliferation, and metastasis [40, 41]. Both cell membrane receptors are identified as key contributors to tumor malignancy in GBM.
Interaction between integrin and uPAR has been shown in some other tumors. For instance, in gastrointestinal cancers, the interplay between αVβ6 and uPAR are implicated in the regulation of downstream signaling following uPA binding [30, 32] and also in the induction of MMP9 secretion, thus facilitating the degradation of multiples ECM components [42, 43]. In breast cancer, Annis et al. demonstrated the participation of integrin αVβ3 in tumor invasion via activation of SRC/MAPK signaling and FRA-1 phosphorylation [44].
Although in GBM some reports suggest an interaction between integrins and uPAR, no concrete evidence in this regard has been published up to now. Among the few reports on this are the results published by Veeravalli et al. where they demonstrated a down-regulation of the expression of several integrins such as α1, α2, α6, α7, α9 and αV as well as β1 and β3 in response to the reduction of uPAR mRNA level [45]. Their research underscores the critical roles of integrins, uPAR and MMP9 in glioma tumor biology, and proposes a possible interaction model between them [46].
Under the knowledge that IαV and uPAR are relevant antigens associated with poor clinical prognosis in human GBM, we focused our efforts on understanding this interaction. Employing two high-grade GBM cell lines that express both proteins, the results showed an interaction between them, demonstrated by Co-IP assays and CM. Since no uPAR expression was detected in the low-grade glioma cell line, we attempted to induce its expression by transfection of the uPAR gene sequence, termed PLAUR. Interestingly, no IαV/uPAR interaction was observed in these cells, suggesting that the mere presence of the protein in the low-grade glioma cell is not sufficient to establish a positive interaction. This implies that an additional factor participates in the interaction between these two proteins.
Glycosylation plays a significant role in protein interaction, especially in cell membrane proteins, where a broad spectrum of glycan structures is observed.
Aberrant glycosylation refers to the particular glycophenotype of cancer cells, which involves modification of the glycosylation profile of normal cell progenitors and is concomitant with the multistage process of malignant transformation [36]. The glycophenotype of tumor cells may be based primarily on N- or O-glycans, depending on the biology of the tumor. While in Neuroblastoma the glycophenotype is mostly expressed based on O-glycans [47], the glycosylation profile found in GBM is primarily expressed by N-glycans [48]. An altered glycosylation profile has been associated with cellular features that promote tumor progression, such as adhesion to the extracellular matrix, migration, invasion, inhibition of apoptosis, host immunoregulation and resistance to chemotherapy [49].
The glycosylation of integrins has been demonstrated as a relevant player in cell adhesion, migration, and survival in tumor cells [50–52]. αβ integrin dimers have more than 20 potential N-linked glycosylation sites, as reported [50]. Additionally, integrin αV specifically has 12 putative N-glycosylation sites. The presence of these N-glycan core structures has been demonstrated as crucial for several reasons: they are essential for integrin heterodimerization, stabilization of the conformation, expression at the cell membrane and interaction with ligands [53, 54].
The structural association between integrins and uPAR is extensively reported. Simon et al. and Zhang et al. described that the surface loop of the β-propeller domain of integrin α-chain in α3β1 and αMβ2 heterodimers functionally associates with uPAR [55, 56]. Subsequently, Chaurasia et al. further showed that integrin α5β1 interacts with the domain III region of uPAR [57]. Also, Ahn et al. suggested that the interaction with uPAR requires the expression of the complete αβ heterodimer (e.g. αVβ6) rather than individual subunits [33]. Additional insights through docking simulations of integrin αVβ3 and uPAR showed that the β-propeller region of αV from αVβ3 interacts with domains II and III of uPAR [58].
Although Integrin/uPAR interaction is well described, the role of glycans in it has been scarcely evaluated. To the best of our knowledge, only one report shows that treatment of melanoma cells with SW inhibits the interaction of αV and α3 with uPAR, suggesting an important role of N-glycosylation in this interaction [34]. In the same line, our results demonstrate that N-glycosylation plays a key role in integrin αV/uPAR interaction since treatment with both SW and PNGase inhibits it as demonstrated by Co-IP and CM. Of the 12 reported potential glycosylated positions of IαV, the evaluation in the GBM cell line LN229 showed six glycosylated positions (N74, N554, N615, N704, N874 and N945) and only two in the low-grade glioma SW1088 (N74 and N874). Interestingly, the glycan profiles between these cells showed substantial differences. The glycosylation profile of IαV obtained from the GBM cell line exhibited a high proportion of complex and hybrid glycans, with a limited presence of oligomannose-type glycans, except at position N945, which showed 100% of oligomannose. In contrast, IαV from low-grade glioma glycosylation was characterized exclusively by oligomannose-type glycans at the two glycosylated positions. As mentioned before, the β-propeller domain of integrin is, according to the literature, involved in the interaction with uPAR with N74 being the only glycosylation site within this domain. Although position N74 of IαV is glycosylated in both cell lines, the glycan profiles differ significantly: in the GBM cell line predominantly exhibits complex-type glycans, while the low-grade glioma cell line primarily shows oligomannose-type glycans. To identify the participating glycan structures, we performed treatments with NeuA which severely affects the IαV/uPAR interaction in both GBM cell lines A172 and LN229, indicating a significant involvement of sialic acids which were found in all glycosylated positions. Moreover, incubation with PHA-L that binds to β1–6 branched N-glycans also inhibits the IαV/uPAR interaction. Conversely incubation with ConA that recognizes oligomannose structures does not inhibits the IαV/uPAR interaction. Our results postulate that β1–6 branched glycans could be present at positions N74, N554, and N874. The first site is located within the β-propeller domain, the second adjacent to it, and the third near the transmembrane domain, forming part of Ig Domain 3. Considering the role of the integrin β-propeller domain in contacting uPAR, the glycan at N554 is unlikely to participate in the interaction, and the involvement of N874 is even less expected. Therefore, we speculate that the predicted branched β1–6 glycan at N74 plays an important role in this interaction. Further studies should be conducted to evaluate the participation of the glycosite N554.
To assess the impact of integrin-related glycosylation on cell response, we evaluated the phosphorylation of AKT, a known Integrin/uPAR-derived signal transducer [59, 60]. Treatment with SW in both GBM cell lines inhibits AKT phosphorylation by more than 50%, indicating that N-glycans are involved in AKT activating signaling. Similar results were observed after treatment with NeuA, pointing to the presence of sialic acid as a relevant player in this process. We performed the same experiments in cells transfected with PLAUR, showing an increase in the phosphorylation of AKT most likely due to overexpression of uPAR in both cell lines and an inhibition of it as a consequence of SW and NeuA treatment.
In conclusion, our findings demonstrate for the first time the interaction of IαV and uPAR in GBM cells and the major role played by N-glycans in it, suggesting an essential participation of β1–6 branched N-glycans as well as sialic acids, both structures found at glycosylation position N74 of IαV within the β-propeller domain. This work provides novel and valuable insights into the interaction between IαV and uPAR, two significant proteins in GBM tumor biology, highlighting the crucial role of glycosylation.