TspanC8 Tetraspanins and A Disintegrin and Metalloprotease 10 (ADAM10) Interact via Their Extracellular Regions: EVIDENCE FOR DISTINCT BINDING MECHANISMS FOR DIFFERENT TspanC8 PROTEINS

Abstract

A disintegrin and metalloprotease 10 (ADAM10) is a ubiquitously expressed transmembrane metalloprotease that cleaves the extracellular regions from its transmembrane substrates. ADAM10 is essential for embryonic development and is implicated in cancer, Alzheimer, and inflammatory diseases. The tetraspanins are a superfamily of 33 four-transmembrane proteins in mammals, of which the TspanC8 subgroup (Tspan5, 10, 14, 15, 17, and 33) promote ADAM10 intracellular trafficking and enzymatic maturation. However, the interaction between TspanC8s and ADAM10 has only been demonstrated in overexpression systems and the interaction mechanism remains undefined. To address these issues, an antibody was developed to Tspan14, which was used to show co-immunoprecipitation of Tspan14 with ADAM10 in primary human cells. Chimeric Tspan14 constructs demonstrated that the large extracellular loop of Tspan14 mediated its co-immunoprecipitation with ADAM10, and promoted ADAM10 maturation and trafficking to the cell surface. Chimeric ADAM10 constructs showed that membrane-proximal stalk, cysteine-rich, and disintegrin domains of ADAM10 mediated its co-immunoprecipitation with Tspan14 and other TspanC8s. This TspanC8-interacting region was required for ADAM10 exit from the endoplasmic reticulum. Truncated ADAM10 constructs revealed differential TspanC8 binding requirements for the stalk, cysteine-rich, and disintegrin domains. Moreover, Tspan15 was the only TspanC8 to promote cleavage of the ADAM10 substrate N-cadherin, whereas Tspan14 was unique in reducing cleavage of the platelet collagen receptor GPVI. These findings suggest that ADAM10 may adopt distinct conformations in complex with different TspanC8s, which could impact on substrate selectivity. Furthermore, this study identifies regions of TspanC8s and ADAM10 for potential interaction-disrupting therapeutic targeting.

Keywords: ADAM; GPVI; N-cadherin; TspanC8; cell surface enzyme; endothelial cell; metalloprotease; platelet; shedding; tetraspanin.

FIGURE 1.

Endogenous ADAM10 and Tspan14 interact in platelets and primary endothelial cells. A, HEK-293T cells were mock transfected (−) or transfected with a FLAG-tagged human Tspan14 expression construct (+). The cells were lysed in 1% Triton X-100 lysis buffer and subjected to anti-Tspan14 (top panel) and anti-FLAG (lower panel) Western blotting. The Tspan14 antibody was raised in goat against a C-terminal cytoplasmic peptide, in collaboration with Everest Biotech. B, washed human platelets; C, washed mouse platelets and D, human umbilical vein endothelial cells were lysed in 1% digitonin lysis buffer, and proteins were immunoprecipitated with an antibody against ADAM10 or an isotype-matched control. Precipitates were then run on non-reducing gels, Western blotted, and probed with Tspan14 (top panels), ADAM10 (middle panels), and CD9 (lower panels) antibodies. Arrows indicate the positions of the predominant mature form of ADAM10 (A10) and the signal from the immunoprecipitating antibodies (IgG).

FIGURE 2.

The large extracellular loop (LEL) of Tspan14 is the region that interacts with ADAM10 and is required for ADAM10 maturation. A, schematic of Tspan14 and CD9 chimeras. The large extracellular loop (LEL) and variable (var) region of CD9 (black) and Tspan14 (gray) were interchanged; the N-linked glycosylation site of Tspan14 is indicated by a filled oval. B, HEK-293T cells were mock transfected (−) or transfected with expression constructs containing the FLAG-tagged human tetraspanin chimeras with Myc-tagged human ADAM10 (+). Cell lysates were produced using 1% digitonin lysis buffer and immunoprecipitated with an anti-FLAG antibody. Immunoprecipitated proteins were blotted with anti-Myc tag antibody (top panel) or anti-FLAG antibody (lower panel). Whole cell lysates were probed with the anti-Myc tag antibody (middle panel). Data are representative of three independent experiments. C, quantitation of immunoprecipitated ADAM10. Data in panel B (upper panel) were quantitated using the Odyssey Infrared Imaging System (LI-COR), and the amount of ADAM10 immunoprecipitated was shown relative to immunoprecipitated Tspan14, which was arbitrarily set at 100. Data were normalized by log transformation and statistically analyzed using a one-way ANOVA with a Dunnett's multiple comparison test compared with the mock (****, p < 0.0001). Error bars represent standard error of the mean from three experiments. D, data in panel B (middle panel) were quantitated, the percentage of mature ADAM10 calculated, and the data log transformed and statistically analyzed as described for panel C (***, p < 0.001).

FIGURE 3.

The large extracellular loop (LEL) of Tspan14 is critical for its ability to increase ADAM10 cell surface accumulation. A, HeLa cells were transfected with the indicated Tspan14-CD9 chimeras (see Fig. 2A) and GFP to identify transfected cells. Cells were stained with an APC-conjugated ADAM10 antibody and analyzed by flow cytometry. Dot plots are representative of three independent experiments. The bottom left panel shows isotope control staining. B, average geometric mean fluorescent intensities for ADAM10 staining, gated on live and GFP-positive cells, were compared statistically using a one-way ANOVA with a Dunnett's multiple comparison test, compared with the CD9 control (***, p < 0.001; **, p < 0.01). Error bars represent standard error of the mean from three experiments.

FIGURE 4.

All Tspan14-CD9 chimeras partially co-localize with ADAM10 and so have access to the metalloprotease. HeLa cells were transfected with the indicated Tspan14-CD9 chimeras (see Fig. 2A) and HA-tagged mouse ADAM10. Cells were fixed and stained with an anti-HA antibody (green) and an anti-FLAG antibody (red). Confocal microscopy images are representative of three independent experiments and at least 15 fields of view.

FIGURE 5.

The region of ADAM10 comprising the disintegrin domain (D), the cysteine-rich (C), and stalk (S) regions mediates the interaction with Tspan14. A, schematic of ADAM10 and ADAM17 chimeras. The extracellular disintegrin (D), cysteine-rich (C), and stalk (S) regions of ADAM10 (black) and ADAM17 (gray) were interchanged together (DCS) or individually. B, HEK-293T cells were mock transfected (−) or transfected with FLAG-tagged mouse Tspan14 (+) in addition to either HA-tagged mouse ADAM10, ADAM17, ADAM17 10DCS, or ADAM10 17DCS. Cells were lysed in 1% digitonin lysis buffer and immunoprecipitated with an anti-FLAG antibody. Immunoprecipitated proteins were blotted with anti-HA tag antibody (top panel) or anti-FLAG antibody (lower panel). Whole cell lysates were probed with the anti-HA tag antibody (middle panel). The blots are representative of three independent experiments. C, HEK-293T cells were co-transfected with (+) or without (−) FLAG-tagged mouse Tspan14 and either HA-mouse ADAM10, ADAM17, ADAM17 10DCS, ADAM17 10D, ADAM17 10C, or ADAM17 10S. Cells were treated as in B. D, data from panels B and C were quantitated and presented as the relative amount of each ADAM10/17 construct immunoprecipitated with Tspan14, having arbitrarily set wild-type ADAM10 to 100. Data were normalized by log transformation and statistically analyzed using a one-way ANOVA with a Dunnett's multiple comparison test, compared with the ADAM17 control (*, p < 0.05). Error bars represent standard errors of the mean from 3–6 experiments.

FIGURE 6.

All TspanC8s interact with the region of ADAM10 comprising the disintegrin (D), cysteine-rich domain (C), and stalk (S). A, HEK-293T cells were transfected with expression constructs for the HA-tagged mouse ADAM17 10DCS chimera and FLAG-tagged mouse TspanC8s, CD9 or negative control (−). Lysates were extracted in 1% digitonin lysis buffer and proteins immunoprecipitated with an anti-FLAG antibody. Immunoprecipitates were blotted with anti-HA tag antibody (top panel) or anti-FLAG antibody (lower panel). Whole cell lysates were probed with the anti-HA tag antibody (middle panel). B, data in panel A (upper panel) were quantitated, and the amount of ADAM17 10DCS immunoprecipitated was normalized for the amount in the whole cell lysate. Data are shown relative to immunoprecipitated Tspan14, which was arbitrarily set at 100. Data were normalized by log transformation and statistically analyzed using a one-way ANOVA with a Dunnett's multiple comparison test compared with the mock. All TspanC8s bound significantly to ADAM17 10DCS (p < 0.0001). Error bars represent standard error of the mean from three experiments. C, ADAM17 10DCS whole cell lysate data in panel A were quantitated, and the amount of ADAM17 10DCS expressed was normalized to the expression in the first lane, which was arbitrarily set at 100. Error bars represent standard error of the mean from three experiments.

FIGURE 7.

The disintegrin (D), cysteine-rich (C), and stalk (S) regions of ADAM10 are essential for Tspan14-mediated exit from the ER. A, HeLa cells were transfected with combinations of FLAG-tagged Tspan14 and HA-tagged mouse ADAM10 wild-type or ADAM10 17DCS. Cells were fixed and stained with an anti-HA antibody (green), an anti-FLAG antibody (red) and WGA to visualize the plasma membrane and internal cellular structures by confocal microscopy. B, HeLa cells were transfected and stained as in panel A except an anti-calnexin antibody was used instead of WGA to define the limits of the ER (images not shown). The HA signal was quantitated across the whole cell and within the mask of the calnexin staining, and presented as a percentage of HA-ADAM10 or HA-ADAM10 17DCS signal localized in the ER. Data are representative of three independent experiments and at least 15 fields of view. A two-way ANOVA statistical analysis was performed with a Bonferroni's multiple comparisons test (ns, non-significant, ****, p < 0.0001). C, HEK-293T cells were mock transfected (−), or transfected with HA-tagged mouse ADAM10 wild-type or ADAM10 17DCS. Cells were surface biotinylated, lysed, and immunoprecipitated with an anti-HA antibody. Immunoprecipitates were stained with neutravidin (top panel) or an anti-HA antibody (bottom panel). Whole cell lysates were stained with an anti-HA antibody (middle panel).

FIGURE 8.

The combined cysteine-rich (C) and stalk (S) region of ADAM10 without the disintegrin (D) is sufficient to interact with Tspan14. A, HEK-293T cells were mock transfected (−) or transfected with FLAG-tagged human CD9 or Tspan14, with co-transfection of Myc-tagged human ADAM10, or pDisplay constructs containing ADAM10DCS or ADAM10CS, which also possessed Myc tags. Cells were lysed in 1% digitonin lysis buffer and immunoprecipitated with an anti-FLAG antibody. Immunoprecipitated proteins were blotted with anti-Myc tag antibody (top panel) or anti-FLAG antibody (lower panel). Whole cell lysates were probed with the anti-Myc tag antibody (middle panel). B, data in panel A (upper panel) were quantitated from three experiments. Data were log transformed and compared statistically with a one-way ANOVA with a Dunnett's multiple comparison test against the mock. Tspan14 bound significantly to ADAM10DCS (p < 0.0001) and ADAM10CS (p < 0.0001). A diagrammatic representation of the ADAM10 constructs is shown below the graph. C, HEK-293T cells were mock transfected (−) or transfected with FLAG-tagged human CD9 or Tspan14, with co-transfection of pDisplay ADAM10CS or ADAM10S. Cells were treated as in panel A. D, data in panel C were quantitated from three experiments. Data were log transformed and compared statistically with a one-way ANOVA with a Dunnett's multiple comparison test against the mock. Tspan14 bound significantly to ADAM10CS (p < 0.0001) and ADAM10S (p < 0.001).

FIGURE 9.

The TspanC8s bind differentially to the disintegrin (D), cysteine-rich (C), and stalk (S) regions of ADAM10. A, HEK-293T cells were mock transfected (−) or transfected with FLAG-tagged mouse TspanC8s or CD9, and co-transfected with the pDisplay vector containing HA-tagged human ADAM10DCS. Cell lysates were produced in 1% digitonin lysis buffer and immunoprecipitated with an anti-FLAG antibody. Immunoprecipitated proteins were blotted with anti-HA tag antibody (top panel) or anti-FLAG antibody (lower panel). Whole cell lysates were probed with the anti-Myc tag antibody (middle panel). B, data from panel A (upper panel) were quantitated and presented as the amount of immunoprecipitated ADAM10DCS relative to the Tspan14 immunoprecipitation, which was arbitrarily set to 100. Data were normalized by log transformation and statistically analyzed using a one-way ANOVA with a Dunnett's multiple comparison test compared with the CD9 control. All TspanC8s bound significantly to ADAM10DCS (p < 0.001). Error bars represent the standard error of the mean from three experiments. C and D, these experiments were carried out as described for panels A and B except using HA-tagged human ADAM10CS. All TspanC8s bound significantly to ADAM10DCS (p < 0.0001). E and F, these experiments were carried out as for panels A and B except using HA-tagged human ADAM10S (****, p < 0.0001; **, p < 0.01; *, p < 0.05).

FIGURE 10.

Evidence that different TspanC8s interact with ADAM10 by distinct mechanisms. A, comparison of TspanC8 co-immunoprecipitations with ADAM10 truncation constructs. Quantitation of the co-immunoprecipitations of ADAM10DCS, ADAM10CS, and ADAM10S with each tetraspanin from Fig. 9 were compared. Values were normalized using Tspan14 data from Fig. 8. All data were relative to the co-immunoprecipitation of ADAM10DCS with Tspan14, which was arbitrarily set to 100. Data were log transformed and statistical analysis was performed using a one-way ANOVA with a Dunnett's multiple comparison test comparing ADAM10CS (#, p < 0.01) or ADAM10S (*, p < 0.01) to the ADAM10DCS for each tetraspanin. Error bars represent the standard error of the mean from three experiments. B, schematic of the potential differential modes of interaction of the TspanC8s with ADAM10. Bold regions of ADAM10 represent those required for a strong interaction with the corresponding TspanC8. Note that Tspan15 has 3 N-linked glycosylation sites and Tspan17 has 2, whereas Tspan5, 10, 14, and 33 have 3, 0, 1, and 2, respectively; for the latter, Tspan14 is depicted as an example.

FIGURE 11.

Differential effects of TspanC8s on ADAM10 substrate cleavage: Tspan15 promotes cleavage of N-cadherin and Tspan14 reduces cleavage of GPVI. A, HEK-293T cells were mock transfected (−) or transfected with FLAG-tagged mouse TspanC8s. The cells were lysed in 1% Triton X-100 lysis buffer and subjected to Western blotting with an antibody to the C-terminal cytoplasmic tail of N-cadherin (upper panel) or with an antibody to the FLAG epitope (lower panel). B, data from A (upper panel) were quantitated and the lower, cleaved band given as a percentage of the total (upper and lower band combined). Data were normalized by log transformation and statistically analyzed using a one-way ANOVA with a Dunnett's multiple comparison test compared with the mock control. Error bars represent the standard error of the mean from three experiments (*, p < 0.05). C, HEK-293T cells were co-transfected with GPVI and FcRγ and one of each of the FLAG-tagged mouse TspanC8s or without a tetraspanin (−) or with the addition of the ADAM10 inhibitor GI254023X at 10 μm. Cells were treated as in panel A, except lysates were subjected to an anti-GFP antibody (upper panel) instead of an anti-N-cadherin antibody. D, data from panel C (upper panel) were quantitated as described in panel A (***, p < 0.001).