What Is Granzyme A and What Are Its Functions?

Granzyme A is a protein-cutting enzyme produced by the immune system’s killer cells, and it serves as one of the body’s primary weapons for destroying infected or abnormal cells. It belongs to a family of enzymes called granzymes, all stored in tiny granules inside cytotoxic T cells and natural killer (NK) cells. But what makes Granzyme A particularly interesting is that it kills through a pathway entirely separate from the one most immunologists focused on for decades, and more recent research has revealed that killing cells is only part of what it does.

Where Granzyme A Comes From and How It Reaches Its Target

Your immune system has specialized cells whose entire job is to kill other cells that have been infected by viruses or that have turned cancerous. Cytotoxic T cells and NK cells are the main ones, and both produce granzymes as part of their arsenal. Granzymes belong to a broader class of enzymes called serine proteinases, meaning they cut other proteins apart at specific sites. The granzyme family includes several members (A through M in humans, though not all letters are used), and Granzyme A is one of the most abundant.¹

Getting inside a target cell is the critical first step. Granzyme A cannot simply waltz through a cell membrane on its own. It relies on a partner molecule called perforin, a calcium-dependent protein that punches pores into the target cell’s outer membrane. When a killer cell recognizes something it needs to destroy, it releases both perforin and granzymes from its granules into the narrow gap between itself and its target. Perforin assembles into ring-shaped pores in the target cell’s membrane, and granzymes slip through these openings into the cell’s interior. There is also evidence that the target cell itself actively participates in this process, essentially helping to internalize the granzymes in a way that steers the cell toward a controlled death rather than simply bursting open.²

What Granzyme A Cuts and Why That Matters

Like all enzymes in this family, Granzyme A is a protease: it cuts proteins. But it is selective about where it makes those cuts. Detailed analyses of its cleavage preferences show that Granzyme A strongly favors cutting after arginine, a positively charged amino acid, with high selectivity for arginine over other basic residues.³ That preference narrows the field of possible targets, but Granzyme A is still remarkably versatile. Substrate profiling studies have identified more than 260 cleavage sites across roughly 200 different protein targets, including DNA repair enzymes and histone proteins that package chromosomal DNA.⁴

The enzyme also uses mechanisms beyond simple recognition of a single cutting site. Structural features called exosites and its ability to form dimers expand what it can interact with, allowing it to recognize substrates that a simpler enzyme would miss entirely.¹ This versatility is a big part of why Granzyme A can carry out so many different functions inside a doomed cell.

Killing Cells Without Caspases

For years, the star of granzyme research was Granzyme B, which kills target cells by activating caspases, a well-characterized family of cell-suicide enzymes. Granzyme A takes a completely different route. It triggers cell death through a pathway that does not involve caspases at all, which matters because some viruses and cancers have evolved ways to block caspase-dependent death. Cells that are resistant to Granzyme B because they overexpress anti-apoptotic proteins or caspase inhibitors remain vulnerable to Granzyme A.⁵

The cell-death program that Granzyme A sets off unfolds in two stages, starting in the mitochondria and finishing in the nucleus.

Mitochondrial Damage and Reactive Oxygen Species

Once inside a target cell, Granzyme A heads for the mitochondria, the organelles responsible for generating the cell’s energy. There, it cleaves a specific component of the electron transport chain called NDUFS3, a protein within complex I. Cutting NDUFS3 disrupts normal energy production and causes electrons to leak out, generating highly reactive molecules called superoxide anions. These reactive oxygen species flood the cell and act as a signal to trigger the next phase of destruction.⁶ When researchers engineered target cells with a mutant version of NDUFS3 that Granzyme A could not cut, those cells resisted Granzyme A-mediated killing but were still susceptible to Granzyme B, confirming that this mitochondrial step is essential and specific to the Granzyme A pathway.

The mitochondrial damage happens rapidly and directly. Granzyme A causes a spike in reactive oxygen species and a collapse of the electrical potential across the mitochondrial membrane, but unlike Granzyme B, it does not cause the release of the usual apoptotic signaling molecules from mitochondria. The entire mitochondrial effect is insensitive to both bcl-2 (a major anti-apoptotic protein) and caspase inhibitors, reinforcing that this is a fundamentally different death mechanism.⁷

Nuclear Destruction Through the SET Complex

The reactive oxygen species generated by mitochondrial damage serve as the trigger for the second phase. They cause a group of proteins called the SET complex, normally found associated with the endoplasmic reticulum, to move into the nucleus. The SET complex includes a protein called NM23-H1, which can nick DNA, but under normal circumstances it is held in check by a protein called SET. Granzyme A cuts SET, releasing NM23-H1 from inhibition and allowing it to begin slicing up chromosomal DNA.⁸ This produces single-stranded DNA nicks rather than the double-stranded breaks typically associated with caspase-dependent death.

Granzyme A does not stop with the SET complex. It also degrades histones (the protein spools around which DNA is wound), lamins (structural proteins that hold the nuclear envelope together), and key DNA damage repair proteins like Ku70 and PARP-1. By destroying both the DNA and the cell’s ability to repair it, Granzyme A makes the damage irreversible.⁵ The movement of the SET complex into the nucleus happens within minutes and depends on superoxide generation, since scavengers that mop up superoxide block the translocation entirely.⁷

Beyond Killing Cells: Driving Inflammation

For a long time, researchers treated Granzyme A primarily as a cell killer. But a growing body of work shows that it has potent functions outside target cells, particularly in ramping up inflammation. When human immune cells were treated with purified Granzyme A in the absence of any delivery agent, the enzyme triggered substantial secretion of the pro-inflammatory cytokines IL-1β, IL-6, and TNF-α, along with IL-8. Only the active form of the enzyme produced this response; an inactive version had no effect, ruling out simple contamination as the cause.⁹

This cytokine-boosting activity means Granzyme A acts as a bridge between the adaptive immune system (the killer T cells that produce it) and the innate immune system (the broader inflammatory response). By triggering secretion of these powerful signaling molecules, Granzyme A can amplify the immune response against infection even without directly killing the pathogen-hosting cell. It essentially calls in reinforcements.

Fighting Viruses and Bacteria

The cytotoxic and inflammatory properties of Granzyme A converge in its role during infections. Against viruses, granzymes can directly break down viral proteins or host-cell proteins that the virus needs for entry, replication, or trafficking inside cells. They can also boost pro-inflammatory antiviral cytokine responses, mounting a multipronged defense.¹⁰ In some cases, this noncytotoxic activity may be as important as actual cell killing, since destroying an infected cell is only useful if you can also prevent viral particles from spreading.

Against bacteria, the story takes a surprising turn. Research has shown that another immune molecule, granulysin, can deliver granzymes directly into bacteria. Once inside, granzymes attack the bacterium’s own electron transport chain and oxidative stress defense proteins, generating reactive oxygen species that kill the microbe. This was demonstrated across diverse bacterial strains, including Escherichia coli, where granzyme-mediated cleavage of complex I components mirrored what happens to complex I inside human target cells.¹¹ The parallel is striking: Granzyme A uses the same basic strategy (sabotaging energy production and unleashing oxidative damage) against both human cells and bacteria.

When Granzyme A Turns Harmful: Sepsis and Autoimmunity

An immune weapon this powerful can cause serious collateral damage if its activity gets out of control. Sepsis is one of the clearest examples. In patients with abdominal sepsis, circulating Granzyme A levels are markedly elevated compared to healthy controls, roughly 2.3-fold higher at baseline. Granzyme A also stays elevated over at least 48 hours, and patients who died had higher baseline levels than those who survived.¹² In experimental models, inhibiting Granzyme A reduced inflammation and improved survival during abdominal sepsis, suggesting the enzyme actively contributes to the runaway inflammatory response rather than simply being a bystander marker of immune activation.¹³

Autoimmune diseases present a similar picture. In rheumatoid arthritis, Granzyme A levels in both blood plasma and the fluid inside inflamed joints are dramatically elevated compared to healthy individuals or to patients with non-inflammatory joint conditions like osteoarthritis.¹⁴ The combined cell-killing and pro-inflammatory activities of Granzyme A make it a plausible contributor to joint destruction and chronic inflammation in these patients, though disentangling cause from consequence remains an active area of research.

How the Body Keeps Granzyme A in Check

Given the damage Granzyme A can inflict, the body needs ways to rein it in. Several natural inhibitors exist. In blood plasma, antithrombin III and alpha-2-macroglobulin are major inhibitors that can neutralize free Granzyme A.¹⁵ However, there is an important wrinkle: Granzyme A that is bound to proteoglycans (large sugar-protein molecules found in the spaces between cells) is shielded from these circulating inhibitors. This means that in inflamed tissues, where proteoglycans are abundant, Granzyme A can persist in an active form for longer than you might expect, potentially sustaining inflammatory damage even after the initial immune signal has passed.

At the cellular level, a member of the serpin family of protease inhibitors called SERPINB12 can form a covalent bond with Granzyme A and inactivate it using the slow-binding trap mechanism characteristic of serpins. SERPINB12 may function as an intracellular safety catch, preventing Granzyme A from causing friendly fire inside the very immune cells that produce it.¹⁶ The balance between Granzyme A activity and these inhibitory mechanisms is one of the factors determining whether an immune response stays productive or tips into pathological territory.

An Ancient Immune Weapon

Granzyme A is not a recent evolutionary invention. Comparative studies of granzyme-like sequences across vertebrates indicate that the lineage containing Granzyme A and its close relative Granzyme K diverged as a distinct group before tetrapods (four-limbed land animals) split from ray-finned fish, roughly 420 million years ago.¹⁷ The fact that this enzyme has been conserved across such a vast span of evolutionary time underscores how fundamental it is to immune defense. Vertebrates that lost effective granzyme function would have been at a severe disadvantage against intracellular pathogens, and natural selection has kept this molecular weaponry remarkably intact.

Granzyme A as a Clinical Biomarker

The fact that Granzyme A levels rise in the bloodstream during inflammation, infection, and cancer has made it an attractive candidate as a clinical biomarker. In sepsis, as discussed earlier, elevated levels track with disease severity and mortality risk. In cancer immunotherapy, researchers have started incorporating Granzyme A into predictive models. One exploratory study in patients with non-small-cell lung cancer treated with immune checkpoint inhibitors found that post-treatment increases in Granzyme A were associated with the occurrence of immune-related adverse events, and the enzyme was included as one of eight plasma proteins in a predictive scoring model for treatment resistance.¹⁸

The appeal of Granzyme A as a biomarker lies in its dual nature. Because it reflects both cytotoxic immune activity and pro-inflammatory signaling, a single measurement potentially captures two dimensions of the immune response at once. That said, the research is still early-stage. Most studies so far have identified associations rather than establishing that Granzyme A measurements actually improve clinical decision-making compared to existing tools. Whether Granzyme A testing will move from research labs into routine clinical use depends on whether those associations hold up in larger, prospective studies and whether they add value beyond cheaper and simpler markers of inflammation.

How Granzyme A Differs from Granzyme B

Readers encountering Granzyme A for the first time often wonder how it compares to the more famous Granzyme B. The two enzymes come from the same killer cells and are delivered by the same perforin mechanism, but they diverge from there. Granzyme B cleaves after aspartic acid residues and activates caspases to trigger a fast, well-characterized apoptotic death. Granzyme A cleaves after arginine and drives the caspase-independent, mitochondria-and-nucleus pathway outlined above. This is not redundancy. The two pathways serve as complementary arms of the immune response: if a virus or tumor evolves resistance to one, the other remains effective. The sepsis data illustrate this distinction clearly, since in patients with abdominal sepsis, Granzyme A levels were significantly elevated while Granzyme B levels were not different from healthy controls.¹²

The pro-inflammatory role also appears more prominent for Granzyme A than for Granzyme B. While Granzyme B has been studied extensively in the context of extracellular matrix breakdown and wound healing, Granzyme A’s ability to directly trigger cytokine secretion from immune cells gives it a unique niche as an amplifier of innate immunity. Understanding that these are two functionally distinct enzymes, not interchangeable versions of the same weapon, is essential for anyone trying to make sense of granzyme biology.