CD155 and EndoA1 mediate growth and tissue invasion downstream of MAP4K4 in medulloblastoma cells

The composition of the plasma membrane (PM)-associated proteome of tumor cells determines cell-cell and cell-matrix interactions and the response to environmental cues. Whether the PM-associated proteome impacts the phenotype of Medulloblastoma (MB) tumor cells and how it adapts in response to growth factor cues is poorly understood. Using a spatial proteomics approach, we observed that hepatocyte growth factor (HGF)-induced activation of the receptor tyrosine kinase c-MET in MB cells changes the abundance of transmembrane and membrane-associated proteins. The depletion of MAP4K4, a pro-migratory effector kinase downstream of c-MET, leads to a specific decrease of the adhesion and immunomodulatory receptor CD155 and of components of the fast-endophilin-mediated endocytosis (FEME) machinery in the PM-associated proteome of HGF-activated MB cells. The decreased surface expression of CD155 or of the FEME effector Endophilin A1 reduces growth and invasiveness of MB tumor cells in the tissue context. These data thus describe a novel function of MAP4K4 in the control of the PM-associated proteome of tumor cells and identified two downstream effector mechanisms controlling proliferation and invasiveness of MB cells. Graphical abstract c-MET activation upon HGF stimulation induces c-MET internalization and induces downstream MAP4K4 activity. (1) MAP4K4 is required downstream of activated c-MET for the maintenance of surface presentation of CD155 in activated cells. CD155 expression is required for MB cell migration, invasion and proliferation in the tissue context. (2) MAP4K4 is required downstream of activated c-MET to maintain membrane depolarization, possibly by regulating the surface localization of several ion channels and transporters. (3) MAP4K4 is required downstream of activated c-MET cause PM-proximal localization of FEME effector CIP4, FBP17 and CIN85. The FEME effector endophilin A is necessary for MB cell migration, invasion and dissemination.


39
The plasma membrane (PM)-associated proteome is a dynamic interface of mammalian cells,

156
We additionally tested drug response in cells maintained first in low serum media for 24h (1% FBS) and 157 then treated the cells for 48h with HGF in combination with Lomustine or Etoposide. Under these 158 conditions, depletion of MAP4K4 caused some reduction in Lomustine sensitivity (Fig. S2E). We 159 observed no impact on the response to Etoposide treatment. Taken together, these results suggest a  (Fig 2A), suggesting that MAP4K4 is either necessary for 172 CLIC1 transfer to or maintenance at the PM in growth factor-stimulated cells. Conversely, the PM-173 association of the acetyl-CoA transporter SLC33A1, the H + /Ca2 + /Mn2 + exchanger TMEM165, the H + /K + 174 transporter ATPase ATP12A, the sodium bicarbonate co-transporter SLC4A7, the acetyl-CoA 175 transporter SLC19A1 and the Na+/K+ transporter ATP1A3 is significantly increased in MAP4K4-176 depleted cells stimulated with HGF (Fig. 2D). Using RT-qPCR we found no evidence that MAP4K4 177 depletion altered the mRNA expression of a selection of proteins identified in our study, indicating that 178 altered PM-association is not the consequence of altered transcriptional control (Fig. S1B).

179
To investigate the significance of this HGF-and MAP4K4-dependent modulation of PM . As we observed the most striking differences in PM association between sgCTL and sgMAP4K4 184 cells after HGF stimulation, we quantified the resting membrane potential (Vmem) under these conditions.

185
We found that HGF stimulation caused hyperpolarization of MAP4K4-depleted compared to control 186 cells (Fig. 2F). This last finding is in accordance with a more depolarized Vmem in disseminating cancer 187 cells, where hyperpolarization leads to a loss of metastatic potential 48 .

188
Taken together, these data indicated the implication of MAP4K4 in modulating PM association 189 of ion transport proteins in growth factor-stimulated cells and that one consequence of this MAP4K4 190 function is to prevent hyperpolarization after HGF stimulation.  To test the requirement of CD155 for HGF-induced collagen invasion, we performed a spheroid 210 invasion assay (SIA) 50,51 (Fig. 3C) using CD155-depleted DAOY cells (Fig. S1H). We found that the 211 reduction of CD155 expression stalled HGF-induced collagen I invasion. To corroborate the implication 9 of CD155 in migration control with another cell line, we tested the effect of CD155 depletion on collagen 213 I invasion in ONS-76 SHH MB cells. These cells invade collagen I independently of exogenous growth 214 factor stimulation. We first confirmed CD155 depletion by siRNA using FACS (Fig. 3D). We then found 215 that CD155 depletion significantly reduced collagen I invasion (Fig. 3E) and that it also reduced 216 proliferation of ONS-76 cells grown as spheroids in 3D cultures (Fig. S3A) without inducing cell death 217 (Fig. S3B). As the inhibition of proliferation by Mitomycin C did not block collagen I invasion in ONS-76 218 cells (Fig. S3C), we concluded that the reduced disseminated cell count in the collagen invasion 219 analysis of CD155-depleted cells is not primarily a consequence of reduced proliferation, but rather of 220 a reduced migratory potential of these cells. To further test this possibility, we assessed the 221 consequence of reduced CD155 expression on motility parameters more directly by single-cell analysis 222 of CD155-depleted ONS-76 cells seeded on collagen I coated plates. Cell movements were recorded 223 by time-lapse video microscopy for five hours. CD155 depletion reduced the average speed of ONS-76 224 cells by 50% ( Fig 3F). Phenotypically, cells with decreased CD155 expression displayed increased 225 circularity, increased area, and increased width to length ratio ( Fig 3F). Taken together, these data  To assess whether CD155 depletion prevented cerebellar tissue infiltration and growth of MB 231 cells, we implanted ONS-76-LA-EGFP tumor spheroids onto cerebellar slices and determined growth 232 and invasion after tumor-cerebellar slice co-culture 52 . After 48 hours of co-culture, ONS-76-LA-EGFP 233 cells transfected with siCTL displayed sheet-like infiltration of the surrounding brain tissue (Fig. 3G). In 234 contrast, spheroids of cells transfected with siCD155 did display a much-reduced invasive behavior and 10 contributes to proliferation in the tissue context (Fig. 3G, S3D). Together, these findings demonstrate 243 that the cell adhesion receptor CD155 contributes to tissue invasion and growth in SHH MB cells.

262
To explore whether MAP4K4 could be involved in PM localization of FEME components such 263 as CIN85/SH3KBP1, CIP4, BIN1 and FBP17, four regulators of FEME priming 55 , we compared their 264 abundance in the different surface proteome samples we generated (Tables 1 & 2). We found that HGF 265 stimulation caused a small increase in PM association of these proteins and of EndoA2 in control cells 266 (Fig. 4B). In contrast, PM association of CIN85/SH3KBP1, CIP4, BIN1 and EndoA2 decreased 267 significantly in HGF-stimulated sgMAP4K4 cells, indicating that MAP4K4 controls membrane 268 association of these proteins in cells with activated c-MET. We could not detect EndoA1 and EndoA3 269 in pour proteomic samples. CIN85/SH3KBP1, CIP4 and FNBP1/FBP17 are considered critical 270 mediators of FEME by priming the PM for endophilin A clustering 55 . PM enrichment of these proteins in 11 a MAP4K4-dependent manner in HGF-stimulated cells could thus indicate a link between MAP4K4 and 272 FEME priming. In cells, CIP4 is detectable associated with dotted and filamentous structures that do 273 not co-localize with F-actin (Fig. 4C, D). The depletion of MAP4K4 alters the subcellular localization of 274 CIP4 and causes CIP4 to accumulate in circular patches near cellular protrusions (Fig. 4D). Depletion 275 of EndoA1 phenocopies CIP4 accumulation in these patches (Fig. 4D). Interestingly, we also observed 276 the early endosome marker EEA1 in similar patches in control but not in MAP4K4-or EndoA1-depleted 277 cells (Fig. S5). Whereas this potential colocalization remains to be further investigated, we observed 278 that depletion of EndoA1 or MAP4K4 led to a 20% reduction of EEA1 puncta per cell in a HGF

304
Endophilin A1 is required for 3D collagen I matrix invasion

305
We next tested whether depletion of EndoA1, DNM1, or clathrin heavy chain (CLTC) also 306 affects 3D collagen I invasiveness of MB cells (Fig. 5B, S6E, F). HGF caused a robust increase in the 307 number of cells disseminating into the collagen I matrix and increased the distance of invasion (~ 6.8-308 fold increase compared to untreated control). Depletion of EndoA1 or DNM1 significantly reduced HGF-309 induced collagen I invasion, whereas depletion of CLTC had no effect, indicating that EndoA1 functions 310 in clathrin-independent endocytosis. Consistent with a conserved role of EndoA1 in migration control, 311 depletion of EndoA1 also reduced (~ 1.6-fold decrease) collagen I invasion of ONS-76 cells (Fig. 5C).

312
Depletion of MAP4K4 caused also some, albeit insignificant, reduction in EndoA1 expression ( Fig.

313
S6F). To determine whether a compensatory mechanism mediated by EndoA2 or EndoA3 24 may be 314 active in our cell model, we depleted all Endophilin-A proteins simultaneously (TKD). We found that 315 endophilin TKD abrogates HGF-induced cell migration in DAOY cells moderately more effectively than 316 siEndoA1 alone (Fig. 5D). We concluded that EndoA2 or EndoA3 contribute to migration control in MB 317 cells, and that EndoA1 is the most relevant Endophilin in our cell model. In conclusion, we found that

381
We observed increased PM-association of potassium channels and potassium transport-associated 382 molecules including the P-type cation transporter ATP12A, the Ca 2+ -dependent channels ATP1A3 and 383 KCNMA1 in MAP4K4-depleted cells after HGF stimulation. Thus, MAP4K4 could maintain optimal 384 polarization in HGF-stimulated cells by controlling potassium efflux rates through regulating PM-385 association of K + transporters. However, whether differential PM association of potassium channels                  Table S1. After 6 hours, the media were changed, and

723
Blocking was performed during 1 hour at RT with 1% FBS. Primary antibody (Table S3) solution was 724 added to the sample and shaken at 100 rpm for 2 hours. Secondary antibody (Table S3)