The elucidation of the cell death pathway triggered by the protein granzyme A (GzmA) has prompted investigators to re-evaluate their concept of programmed cell death. It is no longer considered exclusively mediated by caspases, sometimes thought of as executioner enzymes, which can be evaded by tumor cells. Instead, GzmA activates a death cascade that has many of the hallmarks of apoptosis but without engaging the caspases.
Granzyme proteases are deployed by cytotoxic T lymphocytes (CTLs) and natural killer (NK) cells to destroy tumors and cells infected with viruses. They are released from cytotoxic granules together with perforin, which facilitates delivery of granzymes into target cells, where they activate apoptosis. GzmA is the most abundant of the five known human granzymes, but its mechanism of action has only recently been described.
The lab of Judy Lieberman, HMS professor of pediatrics at the Immune Disease Institute, has been unraveling the GzmA cell death machinery, identifying some of its molecular targets. The researchers have now uncovered a new GzmA-activated mitochondrial cell death pathway, the first stage in triggering subsequent DNA damage and cell death. Mitochondrial damage is triggered when GzmA surprisingly traffics across two membranes into the heart of the mitochondrion, the mitochondrial matrix. As described in the May 16 Cell, they used a proteomics approach to identify potential GzmA mitochondrial cleavage substrates. Of the 307 mitochondrial proteins they resolved, NDUFS3, a component of the mitochondrial electron transport chain complex I, proved to be the critical GzmA target.
A better understanding of the granzyme’s basis of action may offer novel therapeutic targets for treatment of autoimmune diseases and alternate therapies for killing tumor cells that have found ways to escape cell death.
Mitochondrial dysfunction is a hallmark of apoptosis. In the caspase pathway, mitochondrial outer membrane permeabilization permits the release of pro-apoptotic factors like cytochrome c (cyt c) that activate the caspases. Membrane permeabilization was thought to be the critical apoptotic event until a study published by the Lieberman lab three years ago. Denis Martinvalet, a research fellow in the Lieberman lab, noticed that GzmA caused generation of reactive oxygen species (ROS) and the destruction of the mitochondrial membrane potential, but did not disrupt the outer mitochondrial membrane or activate caspases. Moreover, he found that ROS generation was critical for both GzmA and caspase-dependent cell death, since treating target cells with superoxide scavengers blocked both pathways completely.
“Most apoptosis research has been focused on the caspases, the pathway that is important in development and remodeling. For a long time people couldn’t believe that we had another pathway of apoptosis,” said Lieberman.
Part of what makes the GzmA cell death pathway unique is that DNA damage is caused by single-stranded cuts, resulting in fragments that are much larger than the double-strand breaks found in the classical caspase-mediated apoptotic pathways. These DNA fragments are too large to be released from the nucleus until hours later, causing GzmA to have been incorrectly tied to a slow, non-apoptotic death.
ROS generation by GzmA in the mitochondria is an essential first step for this caspase-independent apoptotic pathway, triggering the nuclear translocation of the GzmA-activated DNases that destroy DNA. These DNases are inhibited if superoxide is scavenged, GzmA is blocked from entering mitochondria, or a GzmA-noncleavable variant of its mitochondrial substrate NDUFS3 is expressed in a target cell. While the researchers previously showed that GzmA acted directly on mitochondria, their current study unveils the mitochondrial substrate NDUFS3.
The mitochondrial electron transport chain consists of a series of metalloproteins, complexes I–V, bound to the mitochondrial inner membrane. It behaves like a molecular wire that transfers electrons to O2. During transport, energy is pumped from electrons and transiently stored as a proton gradient across the inner membrane, which is used to produce ATP. GzmA cleavage of NDUFS3 in electron transport complex I interferes with the start of electron transfer. As a consequence, the misdirected electrons generate ROS, the mitochondrial transmembrane potential dissipates, and ATP generation is compromised.
Immunoelectron microscopy confirmed GzmA was internalized into mitochondria, raising the question of how GzmA gets inside. GzmA neither contains a mitochondrial import sequence, nor disrupts the mitochondrial membrane; however, it was found to bind with high affinity to heat shock proteins that chaperone their cargo to specialized transport machinery embedded in the mitochondrial double membranes.
The NDUFS3 cleavage site was identified by mass spectrometry analysis and was confirmed by expressing the cleavage site mutant in target cells, where it prevented GzmA-mediated ROS generation and cell death, underscoring the importance of ROS in GzmA-mediated cell death.
In contrast to what scientists used to believe about free-radical production during cell death, “this study confirms the fact that there is a pathway leading to free radical production. ROS are not byproducts, they are produced for a purpose. There seems to be a really organized way that GzmA reaches its mitochondrial target that is focused on producing free radicals,” said Martinvalet.
R. Chris Bleackley, professor of biochemistry at the University of Alberta, who has made significant advances in GzmB-mediated apoptosis, which activates both caspase-dependent and caspase-independent apoptosis, believes that a better understanding of the molecular basis of the action of GzmA and GzmB will be helpful in determining how to downregulate them when they go awry in autoimmune disease. Moreover, caspase-independent pathways are important for killing tumor cells because many tumor cells and some viruses are resistant to caspase cell death pathways. “In these particular cases, Judy’s new pathway is incredibly important,” said Bleackley, who was not an author on the paper. “That’s the way GzmA will really come into prominence.”
