Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2010 Jul 16;285(29):22461-72.
doi: 10.1074/jbc.M109.086587. Epub 2010 Apr 13.

Deregulation of mitochondrial membrane potential by mitochondrial insertion of granzyme B and direct Hax-1 cleavage

Jie Han  1 , Leslie A Goldstein, Wen Hou, Christopher J Froelich, Simon C Watkins, Hannah Rabinowich
Affiliations

Deregulation of mitochondrial membrane potential by mitochondrial insertion of granzyme B and direct Hax-1 cleavage

Jie Han et al. J Biol Chem. .

Abstract

The cytoplasm and the nucleus have been identified as activity sites for granzyme B (GrB) following its delivery from cytotoxic lymphocyte granules into target cells. Here we report on the ability of exogenous GrB to insert into and function within a proteinase K-resistant mitochondrial compartment. We identified Hax-1 (HS-1-associated protein X-1), a mitochondrial protein involved in the maintenance of mitochondrial membrane potential, as a GrB substrate within the mitochondrion. GrB cleaves Hax-1 into two major fragments: an N-terminal fragment that localizes to mitochondria and a C-terminal fragment that localizes to the cytosol after being released from GrB-treated mitochondria. The N-terminal Hax-1 fragment major cellular impact is on the regulation of mitochondrial polarization. Overexpression of wild-type Hax-1 or its uncleavable mutant form protects the mitochondria against GrB or valinomycin-mediated depolarization. The N-terminal Hax-1 fragment functions as a dominant negative form of Hax-1, mediating mitochondrial depolarization in a cyclophilin D-dependent manner. Thus, induced expression of the N-terminal Hax-1 fragment results in mitochondrial depolarization and subsequent lysosomal degradation of such altered mitochondria. This study is the first to demonstrate GrB activity within the mitochondrion and to identify Hax-1 cleavage as a novel mechanism for GrB-mediated mitochondrial depolarization.

PubMed Disclaimer

Figures

FIGURE 1.
FIGURE 1.
Insertion of GrB into a proteinase K-resistant mitochondrial compartment. A and B, mitochondrial localization of GrB. HeLa cells stably transfected with vector control or Bcl-2 were treated with GrB/Ad (33 nm/10 pfu/ml, respectively) for 10 min. The cells were then stained with MitoTracker Deep Red and with anti-GrB Ab. Scale bar, 10 μm. Using the Metamorph colocalization method described under “Experimental Procedures,” it was determined that 52 and 61% of GrB colocalized with MitoTracker in control cells and Bcl-2-HeLa cells, respectively. A higher magnification of this staining is shown in supplemental Fig. S1. The expression level of Bcl-2 in the utilized cells is shown in B. No difference between vector or Bcl-2-transfected cells was detected with regard to the mitochondrial localization of GrB. C, insertion of GrB into purified mitochondria. Purified mitochondria from Jurkat cells were treated with GrB (160 nm) for the indicated time periods. Subsequently, the mitochondria were left untreated (left; lanes 1–6) or treated with proteinase K (30 μg/ml for 10 min followed by 1 μm phenylmethylsulfonyl fluoride to stop the reaction; right, lanes 7–12). The mitochondria were then spun to separate the pellets and their supernatants. All proteins in the supernatants of the proteinase K-treated mitochondria were digested (lanes 10–12). Like Grp75 and prohibitin, mitochondrial GrB remained proteinase K-resistant, whereas Mcl-1, Bid, and Bim were proteinase K-sensitive (lanes 7–9). The arrowhead indicates cleaved Mcl-1 fragments, and the asterisks show unidentified protein bands. The gel loading ratio between the mitochondrial pellet (Mit Pellet) and wash supernatant (Wash Supp) was 1:4, indicating that at 20 min following its application, ∼80% of the GrB was mitochondrially inserted. D, mitochondrial insertion of exogenously applied GrB onto WT or Bcl-XL-overexpressing Jurkat cells. The cells were treated with GrB/Ad (33 nm/10 pfu/ml) for the indicated time periods. The cells were then homogenized in the presence of a GrB inhibitor and subjected to cellular fractionation to obtain the indicated fractions. The mitochondrial fraction was further treated with proteinase K as described in C. The Cox IV and β-actin immunoblots correspond to the membrane GrB application for 30 min. No significant differences were detected between WT and Bcl-XL-overexpressing cells in the levels of mitochondrial association or insertion of GrB. E and F, kinetics of GrB mitochondrial localization following its cellular delivery with Ad. MCF7 (E) and WT Jurkat cells (F) were treated as described in D. A significant fraction of the cell-delivered GrB was associated with the mitochondria within 30 min of its exogenous application, and most of the delivered GrB was inserted into a proteinase K-resistant mitochondrial compartment within 1 h (E, lanes 3 and 4). In Jurkat cells (F), exogenously applied GrB was associated with mitochondria as early as 10 min after treatment, concomitant with the initial detection of the release of mitochondrial cytochrome c but prior to the detection of Bid processing. G and H, kinetics of GrB mitochondrial localization following its cellular delivery with perforin. MCF7 cells were treated with GrB (33 nm) and PFN titrated to kill less than 10% of the cells on its own. Results obtained with perforin (G and H) were similar to those obtained with Ad (E and F). In G, the cells were treated with GrB/PFN for 1 h, and following subcellular fractionation, the mitochondrial fraction was treated with proteinase K, as described in D. In H, GrB was associated with the mitochondria at the earliest time point tested (10 min), whereas cytosolic cytochrome c and SMAC were detected significantly later.
FIGURE 2.
FIGURE 2.
GrB-mediated cleavage and release of mitochondrial Hax-1. A, mitochondrially localized Hax-1 is cleaved and released into the cytoplasm of GrB/Ad-treated cells. Bax/Bak-deficient Jurkat cells were treated with GrB/Ad (33 nm/10 pfu/ml, 2 h) as described in the legend to Fig. 1. The cells were then Dounce homogenized and subjected to subcellular fractionation to obtain mitochondrial and cytosolic fractions. The mitochondria were further treated with proteinase K, as described in the legend to Fig. 1. Subcellular fractions obtained from control or GrB/Ad-treated cells were assessed by immunoblotting for the expression of the indicated proteins. The low release levels of cytochrome c and AIF are probably due to the Bax/Bak-deficiency of these cells. Also, no significant processing of either Bid or caspase-3 was observed in these cells, despite marked processing of Mcl-1. B, GrB-mediated cleavage of Hax-1 within a proteinase K-resistant compartment of mitochondria. Purified mitochondria obtained from WT Jurkat cells were treated with GrB for the indicated time periods (66 nm; lane 2 on ice, lanes 4 and 5 at 37 °C). The mitochondria were then treated with proteinase K (lanes 3–5), as described in the legend to Fig. 1. Hax-1 cleavage product (arrowhead) is detected in mitochondria treated with GrB and subsequently with proteinase K (lanes 4 and 5). C, GrB mediates the release of Hax-1 and its cleaved fragment from mitochondria. Purified mitochondria obtained from WT Jurkat cells were treated with GrB, as described above. The mitochondrial pellet (Mit-Pellet) (30% of input, lanes 1–5) and supernatant (Mit-Sup) (90% of input, lanes 6–10) were separated and assessed by immunoblotting for the presence of the indicated proteins. Loss of mitochondrial full-length Hax-1 in the mitochondrial pellet is accompanied by the detection of the cleaved Hax-1 product (lanes 2–5). Release of Hax-1 and its cleavage fragment is accompanied by the release of cytochrome c and AIF (lanes 7–10). Expression of Cox IV in the mitochondrial pellet serves as a control for equal loading. The asterisks indicate unidentified protein bands.
FIGURE 3.
FIGURE 3.
Direct and preferential cleavage of Hax-1 by GrB. A, loss of Hax-1 expression in GrB/Ad-treated cells is not attenuated in the presence of the caspase inhibitor, Z-VAD-fluoromethyl ketone. Jurkat cells were treated with GrB/Ad (33 nm/10 pfu/ml) in the presence or absence of the pancaspase inhibitor, Z-VAD-fluoromethyl ketone (100 μm). The cells were lysed in the presence of GrB inhibitor, as described above, and assessed by immunoblotting for the expression of Hax-1, caspase-3, and β-actin. B, direct GrB cleavage of immunoprecipitated Hax-1. GrB (66 nm) was applied to the extract of Jurkat cells (lane 2) or a pellet of the protein A·anti-Hax-1·Hax-1 complex immunoprecipitated from the Jurkat cell extract (lane 4). Controls were the untreated extract or the Hax-1 immunoprecipitant (lane 1 and 3, respectively). Reaction products were resolved by SDS-PAGE and assessed by immunoblotting with anti-Hax-1 Ab. GrB acts directly on endogenous Hax-1 because immunoprecipitated (IP) Hax-1 is cleaved by GrB. C, direct GrB activity on in vitro translated Hax-1. In vitro translated Hax-1 was co-incubated with GrB (66 nm) for the indicated time periods (lanes 1–6). Lysates of control or GrB-treated mitochondria (lanes 7 and 8) were run on SDS-PAGE side by side with the in vitro translated Hax-1 to compare migration patterns. The asterisk indicates an unidentified protein band. D, mitochondrial Hax-1 is preferentially cleaved by GrB, although in vitro translated Hax-1 is also a caspase-3 substrate. Mitochondria purified from Jurkat cells (lanes 1–3) or 35S-labeled in vitro translated Hax-1 (lanes 4–6) were left untreated or treated with GrB (66 nm) or recombinant caspase-3 (100 nm). The reaction products were assessed by immunoblotting (IB) for the presence of the indicated proteins and by autoradiography for the 35S-labeled fragment(s) of in vitro translated Hax-1. Mitochondrial Hax-1 was cleaved by GrB but not by recombinant caspase-3 (top), and the GrB cleavage product had a migration pattern similar to that of GrB-treated in vitro translated (IVT) Hax-1 (lanes 2 and 5). The activity of GrB and caspase-3 was confirmed by their activity on mitochondrial Mcl-1 (second panel). Cox IV serves to demonstrate equal loading of mitochondrial proteins (third panel). Autoradiographic assessment of IVT Hax-1 cleavage products suggests that it is cleaved by either GrB or caspase-3, although different cleavage fragments are produced. The arrowheads indicate cleavage products. The asterisk indicates an unidentified protein band present in the in vitro translation mixture.
FIGURE 4.
FIGURE 4.
Mapping of GrB cleavage sites in Hax-1. A, accumulation of a 16-kDa Hax-1 fragment during co-incubation with GrB. 35S-Labeled in vitro translated Hax-1 was co-incubated with GrB (100 nm) for 1–3 h. The reaction products were resolved by SDS-PAGE, and the 35S signal was assessed by autoradiography. B, GrB cleaves Hax-1 at Asp148 and Asp159. In vitro translated products of WT Hax-1 (lanes 1–3), D148A mutant Hax-1 (lanes 4–6), and D159A mutant Hax-1 (lanes 7–9) were treated with GrB, as indicated. The reaction products were resolved by SDS-PAGE and assessed by immunoblotting (IB) (top) or autoradiography (bottom). The two panels show the presence of two distinct bands that are generated by GrB cleavage: a major band that corresponds to the Asp148 cleavage site (lanes 2, 3, 8, and 9) and a minor band that corresponds to the Asp159 cleavage site (lanes 2, 3, 5, and 6). These results are in agreement with the predicted GrB Hax-1 cleavage sites by the GrB Hax-1 fragment analyses described under “Experimental Procedures.” D148 and D159 arrowheads correspond to bands generated by GrB cleavage at Asp148 and Asp159 (D148 and D159), respectively. The asterisks indicate unidentified protein bands. C, cleavage of in vitro translated C-terminal Hax-1 by GrB. 35S-Labeled in vitro translated full-length Hax-1 (lanes 1 and 2), N-terminal Hax-1 fragment (Met1–Asp148, lanes 3 and 4), C-terminal Hax-1 fragment (Ala149–Arg279, lanes 5 and 6), and D159A mutant C-terminal Hax-1 (lanes 7 and 8) were either untreated or treated with GrB (66 nm for 20 min) and subjected to SDS-PAGE followed by immunoblotting and autoradiography. The asterisks indicate nonspecific proteins. The arrowhead indicates the new C-terminal Hax-1 fragment (Trp160–Arg279) generated by GrB. Because of the close gel migration of the GrB Hax-1 fragments, they are not well distinguishable by immunoblotting or autoradiography. D, double mutant D148A/D159A Hax-1 protein is resistant to cleavage by GrB. 35S-Labeled, in vitro translated double mutant Hax-1 was treated with GrB, as described above. The reaction products of WT Hax-1 and double mutant Hax-1 were run side by side on SDS-PAGE.
FIGURE 5.
FIGURE 5.
Subcellular localization of GrB Hax-1 cleavage products. A, translocation of in vitro translated full-length Hax-1 and the N-terminal Hax-1 fragment, but not the C-terminal Hax-1 fragment, into a proteinase K-resistant compartment of purified mitochondria. Jurkat cell-purified mitochondria were incubated with increasing doses of 35S-labeled IVT Hax-1 (lanes 3–5), 35S-labeled IVT N-terminal Hax-1 fragment (Met1–Asp148, lanes 6–8), 35S-labeled IVT C-terminal Hax-1 fragment (Ala149–Arg279, lanes 9–11), or control reaction lysate with no plasmids (lanes 1 and 2). The mitochondrial supernatant (Mit Sup) was separated, and the mitochondrial pellet (Mit Pellet) was treated with proteinase K. Following the addition of a proteinase K inhibitor and three washes, the mitochondrial pellet fractions were lysed and assessed for the presence of Hax-1-related proteins by immunoblotting and autoradiography. The identified proteins are indicated by numbered arrowheads as follows. 1, endogenous Hax-1 plus exogenous IVT Hax-1; 2 and 3, 35S-labeled IVT Hax-1; 4, endogenous Hax-1; 5–7, exogenous IVT N-terminal Hax-1 fragment; 8, endogenous Hax-1; 9, exogenous IVT C-terminal Hax-1 fragment. The molecular weight for each of the detected proteins is indicated on the left of each panel. B, inducible expression of N- and C-terminal fragments of Hax-1. T-REx-293 clonal cell lines stably transfected with Tet-inducible lacZ, Tet-inducible N-terminal Hax-1 fragment (residues 1–148), or Tet-inducible C-terminal Hax-1 fragment (residues 149–279) were treated with tetracycline (1 μg/ml, 16 h). The cells were lysed and assessed by immunoblotting for the expression of the Hax-1-related proteins or Xpress epitope (used as a tag for lacZ-encoded protein). Of note, The N-terminal Hax-1 fragment consistently demonstrated a faster migration pattern than C-terminal Hax-1 fragment, despite the lower predicted molecular mass of the latter. C and D, mitochondrial localization of the N-terminal, but not the C-terminal, GrB Hax-1 fragment. T-REx-293 clonal cell lines stably transfected with Tet-inducible N-terminal (C) or C-terminal Hax-1 fragment (D) were treated with tetracycline (1 μg/ml, 16 h). Control and treated cells were Dounce homogenized and subjected to subcellular fractionation to obtain cytosolic and mitochondrial fractions. The mitochondria were treated with proteinase K as described under “Experimental Procedures.” β-Actin and Cox IV serve as markers for the subcellular fractions and to demonstrate loading at a proportional cell ratio.
FIGURE 6.
FIGURE 6.
The mitochondrial activity of GrB is enhanced by knockdown of Hax-1 and by the presence of the N-terminal Hax-1 fragment but inhibited by induced expression of WT or the uncleavable form of Hax-1. A and B, increased GrB-mediated mitochondrial depolarization following Hax-1 knockdown. Hct116 cells were treated with Hax-1 siRNA followed by treatment with GrB/Ad (33 nm/10 pfu/ml). Changes in Δψm were assessed by JC-1 flow cytometry (A), and the levels of expression of Hax-1 in the cells were assessed by immunoblotting (B). Because Hax-1 localizes mainly to mitochondria but is released from mitochondria treated with GrB, the mitochondrial and cytosolic fractions were separated to assess for the potential presence of Hax-1 in the cytosol. Cox-IV and β-tubulin serve as equal loading controls and purification markers for the mitochondrial and the cytosolic fractions, respectively. C, inhibition of GrB-mediated mitochondrial depolarization by induced expression of uncleavable double mutant Hax-1 and sensitization to such depolarization by induced expression of N-terminal Hax-1 fragment. T-REx-293 clonal cell lines stably transfected with Tet-inducible lacZ control, full-length Hax-1, double-mutant Hax-1, or the N-terminal Hax-1 fragment (residues 1–148) were treated with tetracycline (1 μg/ml, 16 h) and with GrB/Ad (1.6 or 3.2 nm/10 pfu/ml for 2 h). The cells were then assessed for changes in mitochondrial Δψm as indicated by JC-1 staining. The N-terminal Hax-1-mediated mitochondrial depolarization is shown in j; its sensitization to GrB-mediated depolarization is shown in k and l; and the double mutant (DM) Hax-1 inhibition of GrB-mediated depolarization is shown in w and x. Similar results were obtained in at least four independent experiments. D, GrB-mediated release of mitochondrial apoptogenic proteins is enhanced in cells overexpressing the N-terminal Hax-1 fragment. T-REx-293 cells stably transfected with Tet-inducible N-terminal Hax-1 fragment were kept as controls or treated with tetracycline and subsequently with GrB/Ad (33 nm/10 pfu/ml) for the indicated time periods. Following the addition of a GrB inhibitor, the cells were Dounce homogenized and fractionated to mitochondrial and cytosolic fractions. These subcellular fractions were assessed by immunoblotting for the expression of the indicated proteins.
FIGURE 7.
FIGURE 7.
Mitochondrial alterations following Tet-induction of the N-terminal Hax-1 fragment. A, increased expression of the N-terminal Hax-1 fragment is associated with increased ROS production. T-REx-293 cells stably transfected with Tet-inducible N-terminal Hax-1 fragment were treated with tetracycline (1 μg/ml, 16 h, left) or with H2O2 (0.2 mm 1 h, right) and assessed by flow cytometry for carboxy-H2DCFDA fluorescence. B, increased expression of the N-terminal Hax-1 fragment is associated with ATP depletion. T-REx-293 cells stably transfected with Tet-inducible N-terminal Hax-1 fragment were treated with tetracycline (1 μg/ml) for the indicated time periods and assessed for cellular ATP concentration using the ATPLite Assay System (PerkinElmer Life Sciences). C, enhanced co-localization of mitochondria and lysosomes following Tet induction of the N-terminal Hax-1 fragment. T-REx-293 cells stably transfected with Tet-inducible lacZ or N-terminal Hax-1 fragment were treated with tetracycline (1 μg/ml, 16 h) and assessed by confocal microscopy for a merge between MitoTracker Deep Red and anti-LAMP2 Ab detected by a secondary green fluorescent Ab. Scale bar, 40 μm. Using the Metamorph colocalization method described under “Experimental Procedures,” it was determined that 18% of the MitoTracker labeled mitochondria colocalized with LAMP2 in tetracycline-treated lacZ control cells, and the colocalization level rose to 55% following Tet induction of the N-Hax-1 fragment. A higher magnification of this staining is shown in supplemental Fig. S6. Please note that the MitoTracker is utilized only for mitochondrial identification; this dye is not diagnostic for the level of mitochondrial polarization because its mitochondrial binding depends both on mitochondrial membrane potential and lipophilicity. Each of the panels represents results confirmed by at least three independent experiments. DAPI, 4′,6-diamidino-2-phenylindole.
FIGURE 8.
FIGURE 8.
N-terminal Hax-1 fragment functions as a dominant negative form of Hax-1 in the destabilization of MMP in a Cyp-D-dependent manner. A–C, WT Hax-1 and double mutant Hax-1, but not the N-terminal Hax-1(1–148) fragment, protects mitochondria against valinomycin-mediated depolarization. T-REx-293 cells stably transfected with Tet-inducible WT Hax-1 (A), DM-Hax-1 (B), or N-terminal Hax-1 fragment (C) were treated with tetracycline (1 μg/ml) for 16 h. The cells were then treated with valinomycin (100 nm, 2 h) and assessed by flow cytometry for mitochondrial membrane potential as measured by JC-1 staining. D–F, the mitochondrial depolarization effect of the N-terminal Hax-1 fragment is Cyp-D-dependent. D and E, T-REx-293 cells stably transfected with Tet-inducible N-terminal Hax-1 fragment were treated with Cyp-D or non-target siRNAs (Ambion) for 60 h. The cells were then treated with tetracycline (1 μg/ml) for 6 h and assessed by flow cytometry for JC-1 staining (D) and by immunoblotting for the expression level of Cyp-D (E). Similar results were obtained with three additional non-overlapping Cyp-D siRNAs from Invitrogen (not shown). F, T-REx-293 cells stably transfected with the Tet-inducible N-terminal Hax-1 fragment were treated with tetracycline (1 μg/ml) and CsA (1 μm) for 16 h and then assessed by flow cytometry for JC-1 staining.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Knickelbein J. E., Khanna K. M., Yee M. B., Baty C. J., Kinchington P. R., Hendricks R. L. (2008) Science 322, 268–271 - PMC - PubMed
    1. Lieberman J., Fan Z. (2003) Curr. Opin. Immunol. 15, 553–559 - PubMed
    1. Packard B. Z., Telford W. G., Komoriya A., Henkart P. A. (2007) J. Immunol. 179, 3812–3820 - PubMed
    1. Barry M., Bleackley R. C. (2002) Nat. Rev. Immunol. 2, 401–409 - PubMed
    1. Bots M., Medema J. P. (2006) J. Cell Sci. 119, 5011–5014 - PubMed

Publication types

MeSH terms

LinkOut - more resources