Granzyme H destroys the function of critical adenoviral proteins required for viral DNA replication and granzyme B inhibition

Abstract

Granzymes are key components of the immune response that play important roles in eliminating host cells infected by intracellular pathogens. Several granzymes are potent inducers of cell death. However, whether granzymes use additional mechanisms to exert their antipathogen activity remains elusive. Here, we show that in adenovirus-infected cells in which granzyme B (gzmB) and downstream apoptosis pathways are inhibited, granzyme H (gzmH), an orphan granzyme without known function, directly cleaves the adenovirus DNA-binding protein (DBP), a viral component absolutely required for viral DNA replication. We directly addressed the functional consequences of the cleavage of the DBP by gzmH through the generation of a virus that encodes a gzmH-resistant DBP. This virus demonstrated that gzmH directly induces an important decay in viral DNA replication. Interestingly, gzmH also cleaves the adenovirus 100K assembly protein, a major inhibitor of gzmB, and relieves gzmB inhibition. These results provide the first evidence that granzymes can mediate antiviral activity through direct cleavage of viral substrates, and further suggest that different granzymes have synergistic functions to outflank viral defenses that block host antiviral activities.

Figure 1

DBP is cleaved during LAK-mediated cell death. (A) Ad5-infected K652 cells were coincubated without or with LAK cells at increasing E:T ratios, in the absence (lanes 1–3) or presence of 8 mM EGTA (lanes 4–6) for 4 h at 37°C. (B) K562 cells were Ad5-infected (lanes 1 and 2) or pcDNA3.1-DBP vector-transfected (lanes 3–10). At 48 h post-infection or 24 h post-transfection, cells were incubated alone (lanes 1, 3, 5 and 8) or coincubated (lanes 2, 4, 6, 7, 9 and 10) with LAK cells (E:T ratio 8:1) for 4 h at 37°C. In some experiments, LAK and K562 cells were incubated with compound 6 (GzmB inhibitor) (lane 5) or Ac-DEVD-CHO (lane 10) before coincubation. After terminating the reactions, the samples were electrophoresed and immunoblotted with anti-DBP antibody 37-3. The data shown (B, lanes 8–10) were from a single autorad with intervening lanes containing data irrelevant to this figure spliced out. The solid arrow denotes intact DBP, unfilled arrows mark DBP fragments and the solid arrowhead denotes a 46 kDa species that is generated when DBP is expressed by transfection (lanes 3, 5 and 8) (Asselbergs et al, 1983). Note that the 46 kDa species is absent on lanes 4, 6, 7, 9 and 10, suggesting that it is also cleaved during cytotoxic killing.

Figure 2

DBP is directly cleaved by gzmB. (A) DBP is efficiently cleaved by gzmB. [³⁵S]methionine-labeled DBP generated by IVTT was incubated in buffer B with increasing amounts of purified gzmB for 30 min at 37°C. (B) [³⁵S]methionine-labeled DBP was incubated in buffer B with 5 mM DTT and 41 U of caspase-1, 42 pM caspase-3, 2 μM caspase-4, 1.5 U caspase-6, 62.5 nM caspase-7, 52.8 mM caspase-8 or 88 mM caspase-9, for 60 min before SDS–PAGE analysis. (C) GzmB cleavage sites in DBP were defined by mutating P₁ Asp residues to Ala (D → A) and incubating wild-type (wt) and mutated [³⁵S]methionine-labeled products in buffer B with purified gzmB for 30 min at 37°C. The data shown (A, C) were from a single autorad with intervening lanes containing data irrelevant to this figure spliced out. Solid arrows denote intact DBP and unfilled arrows mark prominent gzmB cleavage fragments.

Figure 3

DBP is cleaved by gzmH in vitro. (A) Ad5-infected K562 cells were incubated alone (lane 1) or coincubated with LAK cells (lane 2) for 4 h at 37°C. Cell lysates from Ad5-infected K562 cells were incubated in the absence (lane 3) or presence (lane 4) of purified gzmH for 30 min at 37°C. After terminating the reactions, the samples were electrophoresed and immunoblotted with anti-DBP antibody 37-3. (B) LAK cells express gzmH. Cell lysates from LAK cells were electrophoresed and immunoblotted with a rabbit polyclonal against gzmB (lane 1) or the anti-gzmH monoclonal antibody 4G5 (lane 2). (C) [³⁵S]methionine-labeled DBP generated by IVTT was incubated in buffer A in the presence of increasing amounts of purified gzmH for 60 min at 37^oC. (D) GzmH cleavage sites in DBP were defined by mutating P1 residues Phe¹²¹ → Gly, Met¹¹⁸ → Val and Phe¹²¹ → Gly/Met¹¹⁸ → Val, and incubating wild-type and mutated [³⁵S]methionine-labeled products in buffer A with or without gzmH for 60 min at 37°C. The solid arrow denotes intact proteins, the unfilled arrow marks the prominent gzmH cleavage fragment and the unfilled arrowheads mark minor DBP fragments.

Figure 4

Adenovirus DNA replication and gzmH cleavage of DBP during cytotoxic-mediated cell death. (A) Schematic representation of the functional domains and the gzmB and gzmH cleavage sites in DBP. The protein is shown as consisting of an N-terminal domain (NH₂) containing residues 1–173 and a C-terminal domain (COOH) consisting of residues 174–529. The N-terminal domain contains two nuclear localization sequences (N) located at positions 42–46 and 84–89, and the host range (hr) region encompassing residues 130–148. The anti-DBP monoclonal antibody 37-3 (Ab 37-7) recognizes residues 131–174 (Cleghon et al, 1993). The gzmB and gzmH cleavage sites are marked above or below the diagram, respectively. (B) Ad5wt- or Ad5^dbp-infected K562 cells were coincubated with or without LAK cells at increasing E:T ratios. After 4 h, the samples were electrophoresed and immunoblotted with anti-DBP antibody 37-3. The solid arrow denotes intact DBP and the open arrow marks the gzmH cleavage fragment. (C) Ad5wt- or Ad5^dbp-infected K562 cells were coincubated in the absence (control) or presence of LAK cells at E:T ratio 8:1. After 12 and 24 h, genomic DNA was purified and the DNA viral load was quantified by real-time PCR as described in Materials and methods. Values are shown as percentages in which 100% represents the total viral DNA load in control non-killed cells at 12 or 24 h. The error bars represent the s.d.

Figure 5

Adenovirus 100K assembly protein is cleaved during cytotoxic-mediated cell death. Ad5wt- or Ad5^dbp-infected K562 cells were coincubated in the absence or presence of LAK cells at increasing E:T ratios. After 24 h, the samples were electrophoresed and immunoblotted with anti-DBP antibody 37-3 (A) or anti-100K antibody (B). Solid arrows denote intact proteins, the open arrow marks the gzmH cleavage fragment, the solid arrowheads denote gzmB-induced fragments and the unfilled arrowhead marks a novel fragment likely generated from the co-cleavage of DBP by gzmH and gzmB.

Figure 6

100K is directly cleaved by GzmH. (A) [³⁵S]methionine-labeled 100K generated by IVTT was incubated in buffer A in the presence of increasing amounts of purified gzmH for 60 min at 37°C. Samples were analyzed by electrophoresis on 10% SDS–polyacrylamide gels. (B) GzmH cleavage of 100K and gzmB reactivation. Purified gzmH (150 nM) or purified 100K (150 mM) were incubated alone or coincubated for 15 min at 37°C in buffer A. Then, aliquots containing 50 nM gzmH (lanes 3, 8 and 13), 100K (lanes 4, 9 and 14) or gzmH plus 100K (lanes 5, 10 and 15) were preincubated in the absence (lanes 3, 8 and 13) or presence (lanes 4, 5, 9, 10, 14 and 15) of 50 nM gzmB for 15 min at 4°C, before adding [³⁵S]methionine-labeled procaspase-3. As controls, [³⁵S]methionine-labeled procaspase-3 was incubated alone (lanes 1, 6 and 11) or in the presence of 50 nM gzmB (lanes 2, 7 and 12). After 1 h (lanes 1–5), 2 h (lanes 6–10) or 3 h (lanes 11–15) at 37°C, the reactions were terminated and gel samples electrophoresed on 12% SDS–polyacrylamide gels. Intact proteins and their cleaved fragments were visualized by fluorography. The solid arrows denote intact proteins, the open arrow marks the gzmH cleavage fragment and the unfilled arrowhead and the solid arrowhead denote the 21 and 17 kDa procaspase-3 fragments, respectively.