Endofin is required for HD-PTP and ESCRT-0 interdependent endosomal sorting of ubiquitinated transmembrane cargoes

Endofin forms a complex with ESCRTs and EGF receptor

Endofin directly binds Tom1 and HD-PTP, both of which have direct roles in receptor trafficking (Doyotte et al., 2008; Gahloth et al., 2017a; Katoh et al., 2004; Kharitidi et al., 2015; Seet et al., 2004; Yamakami et al., 2003). Endofin’s interactome is likely to reflect its function; however, it remains poorly characterized. To investigate the binding partners of HD-PTP and Endofin in further detail, two proteomic workflows were designed. In the first, we performed proximity-dependent biotin tagging (BioID), in which a biotin ligase (BirA∗) fused to the N- or C-terminus of either HD-PTP and Endofin are used as ‘baits’ to covalently tag nearby proteins (‘preys’) with biotin, which can then be purified and identified by MS (Lambert et al., 2015). Subsequently, an affinity-purification mass spectrometry (AP-MS) approach was employed, with an inducible expression of Flag-tagged HD-PTP and Endofin as baits. Our findings reveal Endofin as a robust proximity interactor of HD-PTP by BioID and AP-MS (Figures 1A and S1B), and that HD-PTP forms a complex with a variety of ESCRT proteins from different ESCRT complexes (ESCRT-0, -I and -III) (Figure 1A). Likewise, HD-PTP was identified as a proximity interactor of Endofin-BirA∗ (Figure 1A), and while interactions between Endofin and ESCRT proteins were not detected by BioID (Figure 1A), AP-MS revealed that various ESCRT proteins (HD-PTP, ESCRT-I and -III) complex with Endofin (Figure S1B). Further, our results demonstrate HD-PTP as a robust interactor of clathrin heavy chain 1 (CLTC) and, more broadly, endocytic and trafficking machinery within the cell (Figure S1B). Collectively, these proteomic assays demonstrate that Endofin and HD-PTP form a complex with endocytic machinery and ESCRTs (further annotated in Figure S1A). To better understand the role of Endofin in relation to HD-PTP, BioID data were further analyzed using the STRING database, by which networks of interaction were designed and clustered into different cellular processes. In support of our observations, Endofin’s network is linked to receptor trafficking, a finding that paralleled that of HD-PTP (Figures S2 and S3) (Doyotte et al., 2008).

To validate that Endofin, HD-PTP, and ESCRTs are forming a complex on early endosomes, isopycnic sucrose gradient centrifugation was performed on 293T cell organelles. While Endofin and HD-PTP were absent in cellular fractions containing LAMP1 and Tom20, the lysosome and mitochondria markers, respectively (Figure S1C), they were present together within cellular fractions containing ESCRT proteins (Hrs, Tsg101, and UBAP1). Tom1 and the early endosomal marker EEA1 were also present in these fractions. In addition, subcellular fractionation followed by size-exclusion chromatography (HPLC) showed Endofin, HD-PTP and ESCRT-0 components (Hrs and STAM2) as part of large protein complexes ranging from 525 to 927 kDa, however the ESCRT-I component Tsg101 was found mostly within complexes of lower molecular weight (Figure S1D). As a further validation of complex formation with ESCRT proteins, co-immunoprecipitation was performed on transiently expressed myc-Endofin. Indeed, Endofin showed to form a complex with HD-PTP, ESCRT-0 (Hrs and STAM2) and ESCRT-I (Tsg101 and UBAP1) (Figure 1B), where STAM2, Tsg101, and UBAP1 directly interact with HD-PTP (Ali et al., 2013; Gahloth et al., 2016; Ichioka et al., 2007). This data further supports that Endofin is part of a complex containing HD-PTP and ESCRTs on early endosomes.

Finally, we probed whether EGFR activation elicits Endofin recruitment using MiniTurbo BioID, which allows for shortened labeling time as compared to the BirA∗ enzyme. Results showed that upon EGF stimulation Endofin and HD-PTP are rapidly recruited to EGFR along with common endocytic machineries (Figure 1C). Altogether, this data demonstrates that Endofin forms a complex with HD-PTP, ESCRTs, and EGFR, suggesting that Endofin might have a role in regulating receptor trafficking.

Endofin regulates plasma membrane density, stability, and lysosomal degradation of integrin α5 and EGFR

HD-PTP is known to have a direct role in regulating EGFR, integrin α5, and PDGFRβ endo-lysosomal membrane trafficking (Belle et al., 2015; Doyotte et al., 2008; Kharitidi et al., 2015). Since Endofin directly interacts with HD-PTP and forms a complex with different ESCRTs (Gahloth et al., 2017b) and EGFR, we thought to investigate the role of Endofin in receptor trafficking. We used HD-PTP-depleted HeLa cells as a positive control to disrupt receptor trafficking (Figure S4A). HeLa cells were Endofin-depleted by shRNA and infected with a non-target shRNA (NT) as a negative control for depletion (Figure S4A). Western blotting of total cell lysates from Endofin- and HD-PTP-depleted cells revealed at least 2.5-fold increase in cellular expression levels of integrin α5 (Figure S4B), and ∼1.2–1.6-fold increase in EGFR levels (Figure S4C).

Several studies have demonstrated that ligand activation, ubiquitination, and internalization of both integrin α5 and EGFR facilitate their sorting toward lysosomal degradation (Alwan et al., 2003; Kharitidi et al., 2015; Lobert et al., 2010). Next, we assessed the effect of Endofin depletion on integrin α5 and EGFR lysosomal degradation kinetics in comparison to HD-PTP depletion. Serum-starved cells were stimulated with fibronectin (3 and 6 h) or EGF (2 and 4 h) in the presence of cycloheximide (CHX) to inhibit protein synthesis. To inhibit lysosomal degradation, cells were pretreated with Bafilomycin A1 (1 h) in the presence of CHX. Western blots revealed that upon Endofin and HD-PTP depletion, lysosomal degradation of integrin α5 and EGFR was delayed by >60% and ∼40% after 6 and 2 h chase, respectively (Figure 2A), suggesting that Endofin plays a role in integrin α5 and EGFR lysosomal degradation.

Considering the crosstalk between ESCRT and ubiquitinated transmembrane proteins, which are preferentially targeted toward lysosomal degradation over recycling back to the PM, we asked whether Endofin depletion can delay the PM turnover of activated receptors. Cell surface stability of activated integrin α5 and EGFR were measured by cell surface ELISA (cs-ELISA) upon fibronectin (4 h) and EGF (20 min) stimulation, respectively. Compared to NT cells, ∼ 20% of integrin α5 was stabilized at the PM resembling the effect of an integrin α5β1 blocking antibody, which interferes with the activation and internalization of the receptor (Figure 2B). Similarly, ∼ 20% of EGFR was stabilized at the PM upon Endofin depletion. A comparable effect was shown for Gefitinib, an EGFR inhibitor that blocks receptor activation, ubiquitination, and internalization (Figure 2B). Notably, Endofin and HD-PTP depletion had no effect on the internalization rate of EGFR upon EGF stimulation (5 min) (Figure S4D). Moreover, using cs-ELISA we measured steady state PM receptor density upon Endofin and HD-PTP depletion. Integrin α5 and EGFR cell surface densities increased by at least 1.5-fold compared to NT cells (Figure 2C). Thus, Endofin depletion likely delays the lysosomal degradation of both integrin α5 and EGFR and increases their PM density and stability by facilitating recycling at the early endosome level.

Endofin is required for the efficient lysosomal delivery of integrin α5 and EGFR

If the lysosomal degradation of integrin α5 and EGFR is delayed upon Endofin depletion, it is plausible that Endofin has a role in ESCRT-dependent receptor sorting from early endosomes toward MVBs and lysosomes. Hence, to better understand the role of Endofin in the endo-lysosomal transfer kinetics, receptor delivery to lysosomes was determined by monitoring the pH of receptor-containing vesicles (pH_v) by fluorescent ratiometric image analysis (FRIA) (Figure 2D) (Barriere and Lukacs, 2008). HD-PTP- and Hrs-depleted HeLa cells (Figure S4A) were used as positive controls for the disruption of receptor trafficking (Kharitidi et al., 2015; Wenzel et al., 2018). First, serum-starved cells were labeled on ice with anti-EGFR or anti-integrin α5 followed by F(ab’)₂ secondary antibody coupled to the pH-sensitive fluorescein isothiocyanate (FITC). Next, to induce synchronized receptor internalization, cells were stimulated with EGF (30- and 60-min chase, 37°C) or with fibronectin (4 h chase, 37°C) and pH_v was measured by FRIA. As expected, EGFR trafficking was delayed after a 30- and 60-min chase in HD-PTP- and Hrs-depleted cells (pH_v ∼ 5.7 ± 0.09 and pH_v ∼ 5.7 ± 0.1, respectively) compared to NT cells (pH_v ∼ 5.2 ± 0.07) (Figures 2D and S4E). Strikingly, after 30 min EGFR chase in Endofin-depleted cells, the receptor is confined to early endosomes (pH_v ∼ 6 ± 0.17), however, in NT cells EGFR is delivered to late endosomes (pH_v ∼ 5.4 ± 0.07) (Figures 2D and S4E). In addition, after 60 min chase, EGFR is delivered to late endosomes in Endofin-depleted cells (pH_v ∼ 5.4 ± 0.23), whereas in NT cells EGFR is delivered to lysosomes (pH_v ∼ 4.5 ± 0.1) (Figures 2D and S4E). Endofin depletion also inhibited the egress of integrin α5 from early endosomes (pH_v ∼ 6.4 ± 0.1) while in control NT cells the receptor was delivered to late endosomes (pH_v ∼ 5.7 ± 0.1) (Figure 2D). The FRIA results clearly demonstrate that Endofin has a crucial role in the transport of receptors from early endosomes toward MVBs and lysosomes. This is exemplified by the identical effects of HD-PTP and Hrs depletion on EGFR trafficking.

Endofin is required for the efficient lysosomal delivery of poly-ubiquitinated model cargo

The recognition of ubiquitinated receptors by ESCRTs is a key step in the sorting process of internalized receptors toward MVBs and lysosomes. To examine whether Endofin also contributes to the endo-lysosomal sorting of ubiquitinated cargoes incapable of receptor signaling, we used the established CD4-Ub chimera which undergoes constitutive poly-ubiquitination. For this, two different model receptors were used. One was truncated CD4 (CD4Tl), in which the cytoplasmic tail was deleted and replaced with a linker, thus lacking a ubiquitin acceptor site and sorting signal (Barriere et al., 2006). In addition, we used CD4Tl-Ub, a chimera where the linker has been fused to a ubiquitin (Ub) moiety. The fusion of a single Ub moiety to the linker is sufficient to induce constitutive poly-ubiquitination of the CD4Tl-Ub, which in turn accelerates its ESCRT-dependent lysosomal delivery (Apaja et al., 2010; Barriere et al., 2007).

Endofin- and HD-PTP-depleted cells were transiently transfected with CD4Tl and CD4Tl-Ub constructs to assess their degradation kinetics by CHX chase (2 h). Western blotting demonstrated that the lysosomal degradation of ∼40% CD4Tl-Ub was delayed upon Endofin and HD-PTP depletion compared to NT cells (Figure 2E). As expected, Endofin and HD-PTP depletion had no significant effect on the levels of CD4Tl (Figure 2E). Furthermore, FRIA showed that Endofin depletion confined CD4Tl-Ub to recycling endosomes (pH_v ∼ 6.4 ± 0.07) compared with NT cells where it is delivered to lysosomes (pH_v ∼ 4.8 ± 0.07) after a 2-h chase (Figure 2F). Hence, Endofin is required for the efficient lysosomal degradation of poly-ubiquitinated cargoes, even in the absence of second messenger signaling activity. To assess whether Endofin-dependent endo-lysosomal sorting has a preference toward specific ubiquitinated cargo, lysosomal delivery of the tetrameric mono-ubiquitinated cargo (CD4TCC-UbAllRΔG) was also determined by FRIA upon Endofin depletion. In CD4TCC-UbAllRΔG, a tetramerization signal was inserted before the Ub moiety and all seven Lys residues of the Ub moiety were replaced with Arg and the carboxy-terminal Gly residues were deleted (Barriere et al., 2006). Consistent with the effect of HD-PTP depletion on CD4TCC-UbAllRΔG trafficking (Kharitidi et al., 2015), Endofin depletion did not affect the trafficking of the tetrameric mono-ubiquitinated cargo (Figure 2G), even though it was previously shown that the tetrameric mono-ubiquitin can serve as an efficient endocytic signal (Barriere et al., 2006). These data show that Endofin does not only play a role in EGFR and integrin α5 trafficking, but it regulates the specific sorting of poly-ubiquitinated cargo toward lysosomal degradation.

Endofin depletion decreases clathrin recruitment to early endosomes

Endofin recruits clathrin via Tom1 to early endosomes (Seet and Hong, 2005), and Tom1 directly interacts with Endofin and with clathrin heavy chain (Seet and Hong, 2005; Seet et al., 2004) (Figure S5C). Our data demonstrates that clathrin colocalization with EGFR reached a peak (∼0.65 ± 0.09) after a 15-min EGFR chase (Figures S5A and S5B). To better understand why receptor trafficking is delayed upon Endofin depletion, we first assessed the effect of Endofin depletion on clathrin recruitment to early endosomes by immunofluorescence. Endofin-depleted HeLa cells were transiently transfected with Rab5Q79L, a GTPase deficient mutant, to induce the formation of enlarged endosomes (Raiborg et al., 2001; Wegener et al., 2010). Upon Endofin depletion, clathrin recruitment decreased by 33% ± 3% compared to NT cells (Figures S5D and S5E). In addition, we assessed Tom1 interaction with clathrin heavy chain in Endofin-depleted cells, and co-IP showed that clathrin interaction with Tom1 decreased upon Endofin depletion (Figure S5F). Since clathrin is indispensable for receptor clustering on the endosomal surface and their accumulation in ILVs during MVB biogenesis (Wenzel et al., 2018), this might contribute to the delay in EGFR lysosomal delivery. To mimic the recruitment of Tom1 protein to early endosomes via Endofin, as previously described by Seet et al., we used a 2xFYVE-Tom1 chimera construct to restore Tom1/clathrin interaction in Endofin-depleted cells (Seet and Hong, 2005). Of note, 2xFYVE-Tom1 chimera in Endofin-depleted cells lacks the ability to form a complex with Endofin binding partners once it is expressed, including HD-PTP. Our results showed that the complementation of Endofin-depleted HeLa cells with 2xFYVE-Tom1 chimera failed to restore EGFR lysosomal delivery (Figure S5G). Hence, although Endofin depletion reduced clathrin interaction with Tom1 and decreased clathrin recruitment to early endosomes, the delay in EGFR lysosomal delivery upon Endofin depletion is not directly linked to the loss of interaction between Endofin and Tom1 or to the role of Tom1 in receptor trafficking.

Endofin expression enhances HD-PTP colocalization with EGFR, Hrs and CHMP4B

HD-PTP directly interacts with ESCRT-0, -I, and -III proteins facilitating EGFR sorting toward MVBs and lysosomal degradation (Doyotte et al., 2008; Kharitidi et al., 2015). Endofin binds to the Bro1 domain of HD-PTP sharing the same binding site with STAM2 (ESCRT-0) and CHMP4B (ESCRT-III), however, Endofin binds HD-PTP with a higher affinity than STAM2 and CHMP4B (Gahloth et al., 2017b). In addition, STAM2 also binds to a motif in the proline rich region of HD-PTP, which serves as a second site for STAM2 interaction (Ali et al., 2013). To investigate whether Endofin could regulate HD-PTP complex formation with ESCRTs, endogenous HD-PTP was immunoprecipitated in Endofin-depleted cells and the co-IP efficiency of the ESCRT-0 protein Hrs was assessed. The indirect interaction of HD-PTP with Hrs was reduced by ∼ 80% upon Endofin depletion (Figure 3A).

Since HD-PTP interaction with ESCRT-0 and -III on early endosomes is essential for the sorting of activated EGFR toward MVBs and lysosomes (Ali et al., 2013) (Doyotte et al., 2008) (Wenzel et al., 2018), we assessed in Endofin-depleted cells the colocalization of EGFR, red fluorescence protein-tagged Hrs (RFP-Hrs) and mCherry-tagged CHMP4B (mCherry-CHMP4B) with HD-PTP by pulse-chase experiments. EGFR colocalization with HD-PTP showed a significant decrease upon Endofin depletion (0.4 ± 0.03), after 5 min EGF stimulation and 10 min EGFR chase, compared to NT cells (0.54 ± 0.03) (Figure 3B). Consistent with the co-IP result, Hrs colocalization with HD-PTP, after 5 min EGF stimulation and 0 min EGFR chase, was significantly reduced upon Endofin depletion (0.44 ± 0.01) compared to NT cells (0.71 ± 0.03) (Figure 3C). Moreover, CHMP4B colocalization with HD-PTP also decreased upon Endofin depletion (0.44 ± 0.06) compared to NT cells (0.62 ± 0.05), after 5 min EGF stimulation and 20 min EGFR chase (Figure 3D). Therefore, this data suggests that EGFR, ESCRT-0 (Hrs), and ESCRT-III (CHMP4B) colocalizations and/or complex formation with HD-PTP are promoted by Endofin, which facilitates EGFR sorting toward MVBs and lysosomal degradation via ESCRT-dependent mechanisms.

Next, the colocalization of EGFR and RFP-Hrs with Endofin was also assessed in HD-PTP-depleted cells by pulse-chase experiments. We found that the colocalizations of EGFR and Hrs with Endofin are stabilized upon HD-PTP depletion after 5 min EGF stimulation and until 20 min EGFR chase. However, EGFR and Hrs showed dynamic colocalizations with Endofin in NT cells (Figures S6A and S6B). This infers that the dynamic colocalizations of EGFR and Hrs with Endofin are regulated by HD-PTP. This dynamic colocalization might reflect the egress of cargoes from early endosomes toward MVBs and lysosomes, which is indeed regulated by HD-PTP (Doyotte et al., 2008). On the other hand, the stabilized colocalizations of EGFR and Hrs with Endofin upon HD-PTP depletion might reflect the delay in cargo sorting from early endosomes toward MVBs and lysosomes.

Endofin/HD-PTP interaction is permissive for an efficient lysosomal delivery of activated EGFR

Endofin directly binds and recruits HD-PTP to early endosomes (Gahloth et al., 2017b). The double L202D/I206D-HD-PTP mutation, which abrogates CHMP4B and Endofin binding to HD-PTP, prevents L202D/I206D-HD-PTP recruitment to early endosomes even upon its co-expression with myc-Endofin (Gahloth et al., 2017b). Our data showed that EGFR and ESCRT-0 colocalizations with HD-PTP are stabilized by Endofin. To assess the functional aspect of Endofin/HD-PTP interaction in regulating cargo sorting toward MVBs and lysosomes, we generated shRNA resistant wild-type (WT) and mutant Endofin constructs.

Transient overexpression of the FYVE-domain C753S-Endofin mutant displayed an impaired localization to early endosomes, a finding we reciprocated here (Figure S6C) (Seet and Hong, 2001). To weaken Endofin’s interaction with the Bro1 domain of HD-PTP (Gahloth et al., 2017a), we employed the L15P-Endofin mutant.

To investigate the effect of Endofin complementation on EGFR endo-lysosomal trafficking, pH_v of EGFR-containing vesicles was monitored by FRIA. Endofin-depleted HeLa cells were transiently transfected with WT- or mutated-Endofin (Figure S6D), and the mean pH_v was assessed after 30 min of EGF stimulation. WT-Endofin complementation restored EGFR lysosomal delivery (mean pH_v ∼ 5.1 ± 0.03) compared to mock transfected Endofin-depleted cells (mean pH_v ∼ 5.9 ± 0.09) and NT cells (mean pH_v ∼ 5.3 ± 0.04) (Figure 4A). L15P-Endofin complementation partially restored EGFR lysosomal delivery (mean pH_v ∼ 5.5 ± 0.02) compared to WT-Endofin (Figure 4A), in line with the partially disrupted biochemical interaction of L15P-Endofin with HD-PTP (Figure S6E). Conversely, C753S-Endofin complementation completely failed to rescue EGFR lysosomal delivery (mean pH_v ∼ 5.9 ± 0.05) (Figure 4A) by profoundly reducing Endofin recruitment to early endosomes.

Importantly, complementing HD-PTP-depleted HeLa cells with WT-HD-PTP (Figure S6D) restored EGFR lysosomal delivery (mean pH_v ∼ 5.4 ± 0.15) (Figure 4B). T145K-HD-PTP, which partially disrupted HD-PTP/Endofin interaction (Figure S6F) (Gahloth et al., 2017c), failed to rescue EGFR lysosomal delivery (mean pH_v ∼ 5.9 ± 0.05) compared to mock transfected HD-PTP-depleted (mean pH_v ∼ 6.1 ± 0.08) and NT (mean pH_v ∼ 5.4 ± 0.09) cells (Figure 4B).

To better understand the mechanism by which Endofin’s FYVE domain influences HD-PTP’s recruitment to early endosomes, we extended our BioID investigation to C753S-Endofin N-terminally BirA∗ tagged baits. Consistent with our findings, disruption of the FYVE domain did not reduce Endofin’s interaction with HD-PTP but did disrupt its proximity interaction with ubiquitin and several early endosome components, providing an insight into a potential mechanism by which Endofin supports HD-PTP’s recruitment to ubiquitinated cargoes on endocytic vesicles (Figure 4C). These results strongly suggest that both Endofin localization on early endosomes via its FYVE domain and its interaction with HD-PTP are necessary for an efficient endo-lysosomal transfer of activated EGFR.

Endofin depletion sustains integrin α5 and EGFR downstream signaling

We previously showed that the heterozygous loss of PTPN23 is frequent in several types of human cancers (Manteghi et al., 2016). Using the cBioPortal database, we show here that the heterozygous loss of ZFYVE16 (Endofin gene) is also frequent in several types of human cancers (Figure 5A). In addition, the PRECOG database, which measures associations between mRNA expression profiles and cancer prognosis by calculating a Z-score, revealed a negative Z-score of ZFYVE16 mRNA in several types of human cancers (Figure 5B).

Cell surface receptors are competent of signaling from early endosomes or MVBs before the formation of ILVs (Mamińska et al., 2016; Rodahl et al., 2009; Wegner et al., 2011). Furthermore, HD-PTP depletion was shown to delay integrin α5 lysosomal delivery, thereby increasing its downstream signaling (Kharitidi et al., 2015). To investigate whether Endofin could also influence integrin α5 signaling, Endofin- and HD-PTP-depleted HeLa cells were stimulated with fibronectin (7 h) to assess levels of phosphorylated FAK, Src and Erk1/2 by Western blotting. In agreement with our previous study, HD-PTP depletion in serum-starved cells increased the levels of pFAK, pSrc, and pErk1/2 compared to NT cells, both at steady state and upon fibronectin stimulation (Figure 5C). Remarkably, the same effect was observed upon Endofin depletion (Figure 5C). This result clearly demonstrates that in addition to receptor stabilization, Endofin depletion further increased integrin α5 signaling.

Next, the effect of Endofin and HD-PTP depletion on EGFR activation and the kinetics of its downstream signaling were evaluated. Interestingly, Endofin and HD-PTP depletion had no significant impact on steady-state EGFR phosphorylation and the activation of its downstream kinase MEK (Figure 5D). However, EGFR and MEK activation were both sustained after 30 min of EGF stimulation upon Endofin and HD-PTP depletion. The levels of pEGFR also remained higher in these cells, even after 60 min of EGF stimulation (Figure 5D). Altogether, these results indicate that Endofin and HD-PTP do not only regulate EGFR and integrin α5 plasma membrane stability, but also control receptor activation and its downstream signaling.

Endofin depletion increases cell migration

Integrin α5β1’s central role in cell migration and its ESCRT-dependent lysosomal desensitization are well established (Kharitidi et al., 2015; Lobert et al., 2010). The migration rate of Endofin-depleted HeLa cells was measured by real time cell analysis (RTCA) using fetal bovine serum as a chemoattractant. Endofin depletion revealed the same effect on cell migration rate as HD-PTP depletion with ∼ 2-fold increase compared to NT cells. Migration of Endofin-depleted cells was suppressed by integrin α5β1 blocking antibody (Figure 5E). This shows that the increase in cell migration upon Endofin depletion is linked to the increase of PM and total integrin α5 levels (Figures 2C and S4B).

Since Endofin/HD-PTP interaction is permissive for an efficient lysosomal delivery of activated EGFR (Figures 4B and 4C), we next addressed whether simultaneous depletion of Endofin and HD-PTP could further impair receptor trafficking. Thus, we compared total EGFR and integrin α5 levels between NT and Endofin-depleted cells, transiently transfected either with a non-targeting shRNA or an shRNA targeting HD-PTP. We found that depleting Endofin or HD-PTP increased total EGFR levels by ∼ 2-fold and ∼1.5-fold, respectively, whereas EGFR levels in double knock-down cells increased by ∼ 2.7-fold in comparison to NT cells (Figure 5F). Interestingly, the impact was even more pronounced for integrin α5. Indeed, the individual depletion of Endofin or HD-PTP increased integrin α5 levels by ∼ 3.4-fold and ∼2.7-fold, respectively, whereas double knocked-down cells exhibited ∼7.7-fold increase in comparison to NT cells (Figure 5F). To address the functional relevance of these findings, we also performed migration assays using fetal bovine serum as a chemoattractant. Results showed that the migration rate increased in double knock-down cells by ∼ 1.5-fold in comparison to Endofin-depleted cells (Figure 5G). Altogether, these results indicate that the simultaneous depletion of Endofin and HD-PTP has an additive effect on total receptor levels, which in turn increase their migration rate in comparison to Endofin-depleted cells.

In certain types of human cancers, integrin α5 and EGFR are well-known drivers of tumorigenesis. Here, we show that Endofin depletion increased integrin α5 and EGFR signaling and cell migration, which are two hallmarks of cancer progression. Hence, loss of Endofin expression delays receptor lysosomal degradation and increases receptor signaling and cell migration, which could ultimately contribute to tumorigenesis and cancer progression.

Limitations of the study

In this study, we have characterized the role of Endofin in targeting integrin α5 and EGFR toward lysosomal degradation, thus regulating receptor signaling and cell migration. We have also shown that Endofin is required for HD-PTP association with ESCRT-0/-III for an efficient EGFR lysosomal delivery. However, the exact mechanism of Endofin association/dissociation with/from HD-PTP, ESCRTs and clathrin complex on early endosomes is still unknown. This will be an interesting direction for future research requiring advanced imaging techniques and protein structural analysis.

Furthermore, it is not fully known whether Endofin regulates the endo/lysosomal delivery of all cell surface receptors undergoing poly-ubiquitination upon activation. Further investigation would be required to address the effect of Endofin depletion on the global cell surface proteome, which might further influence downstream transcriptional pathways.

Key resources table

Resource availability

Experimental model and subject details

Methods details

Quantification and statistical analysis

Western blots and immunofluorescence images were quantified using imageJ software. FRIA data was quantified using MetaFluor software, and pHv peaks were drawn and further analyzed using Origin software. Statistical analysis was performed using Microsoft Excel software. Significance was determined by paired two-tailed Student's t test, and data with p values: ∗<0.05, ∗∗<0.01 and ∗∗∗<0.001 are considered to be significant.