Perforin, a Ca2+-dependent pore forming protein found in the granules of natural killer (NK) cells and cytotoxic T lymphocytes (CTLs), was once proposed to be “the sole mediator of target cell destruction” (Ewen et al. 2012). However, in the late 1970s perforin was revealed to lack the protease activity required for apoptotic induction (Ewen et al. 2012). To identify other proteases that may be responsible, multiple researchers independently investigated the composition of CTL and NK cell granules (Boivin et al. 2009). Percoll density gradients were used to separate granules from NK cells and gel filtration then removed perforin, leaving an enriched pool of serine proteases. These “granule-associated enzymes” were subsequently named granzymes (Masson and Tschopp 1987). Cation exchange chromatography separated the granzymes into distinct clusters as seen by SDS-PAGE gel electrophoresis (Masson and Tschopp 1987). In 1987, Tschopp et al. went on to purify and characterize multiple granzymes within this pool, identifying their distinct substrate specificities and determining partial amino acid sequences (Masson and Tschopp 1987). One of the most highly expressed of these subtypes, granzyme B (GzmB), was identified as the serine protease directly involved in targeted cell death (Hiebert and Granville 2012). Perforin was required to disrupt the cell membrane of the target cell, thus facilitating GzmB entry into the cytoplasm, thereby providing a mechanistic explanation as to how these two proteins released from intact NK cells and CTLs could induce targeted cell death.
The role of GzmB was traditionally understood to be limited to its intracellular role in apoptosis, but a number of discoveries have challenged this notion. Rather than being exclusively expressed by NK cells and CTLs, GzmB is also expressed by other cell types. Moreover, GzmB can act extracellularly, cleaving substrates including extracellular matrix (ECM) proteins, proteoglycans, cytokines, and cell surface proteins (Boivin et al. 2009). This extracellular role for GzmB is leading scientists into a new era with a focus on defining new functions for GzmB. Critically, extracellular concentrations of GzmB in bodily fluids are elevated in various diseases and it is now apparent GzmB plays a significant role in chronic inflammatory, autoimmune, and degenerative diseases (Hiebert and Granville 2012).
Regulation of GzmB
The human GzmB (EC:18.104.22.168) gene (GZMB) is approximately 3500 base pairs long and consists of 5 exons and 4 introns (Klein et al. 1989). Located on chromosome 14q.11.2 (Klein et al. 1989), GZMB encodes a 247 amino acid polypeptide, with a structure composed of two 6-stranded β sheets and three trans-domain segments (Estébanez-Perpiña et al. 2000). GzmB expression in both immune and nonimmune cell types is regulated at both the transcriptional and translational level, with this influenced by many of the same factors responsible for immune cell activation (Boivin et al. 2009). GzmB is synthesized with a signal sequence that targets the protein to the endoplasmic reticulum, with this then cleaved to produce an inactive proenzyme. In the Golgi, GzmB is tagged with mannose-6-phosphate (M6P), allowing M6P receptor targeting to the acidified lytic granule (Boivin et al. 2009). In the granule, cathepsin C activates GzmB by removing the N-terminal dipeptide, which is stored on a scaffold of the proteoglycan, serglycin (Boivin et al. 2009). GzmB has a neutral pH optima, thus its proteolytic activity remains low until released from the lytic granules (Chowdhury and Lieberman 2008). Once active, GzmB predominantly cleaves peptides immediately adjacent to aspartate (Asp) residues at the P1 position (Waugh et al. 2000). The structure of the active site, which contains an arginine (Arg) residue positioned adjacent to the active site pocket, provides the substrate specificity of GzmB (Waugh et al. 2000).
Role in Apoptosis
Immune cell/GzmB-mediated apoptosis requires the presence of perforin to mediate GzmB entry into target cells (Boivin et al. 2009). Perforin multimerizes into the plasma membrane of target cells, allowing for the trafficking of GzmB into the cytoplasm through a ~ 5–20 nm diameter pore (Boivin et al. 2009). Whether perforin transport is conducted through the immunological synapse or following dynamin-dependent endocytosis is still debated. While the necessity of perforin in CTL-mediated apoptosis has been established, other molecules may also facilitate GzmB internalization (Boivin et al. 2009, 2012).
Extracellular (Perforin-Independent) Role
ECM components cleaved by GzmB include decorin, vitronectin, biglycan, betaglycan, fibronectin, fibulin, aggrecan, and cartilage proteoglycans (Boivin et al. 2012). It is proposed that during chronic inflammation, increased and persistent levels along with uninhibited GzmB activity contribute to ECM proteolysis and loss of tissue integrity and function (Boivin et al. 2009). Besides these indirect effects on cell behavior as a consequence of ECM cleavage, GzmB-mediated activation of membrane receptors such as PAR-1, FGFR, TCRz, notch-1, neuronal glutamate receptor, and acetylcholine receptor may also affect cell phenotype and function (Wang et al. 2012).
Extracellular GzmB is reported to mediate inflammation, effectively cleaving cell-cell junction proteins (VE-cadherin) basement membrane components (laminin, nidogen-2, type IV collagen, and peroxidasin), and inflammatory cytokines (IL-18 and IL-1α) (Boivin et al. 2009). A recent study from our group showed GzmB cleavage of the cell-cell junction protein, VE-cadherin, resulting in increased vascular permeability and further inflammation in a mouse model of cardiac fibrosis (Shen et al. 2016). Similarly, the cleavage of basement membrane proteins may facilitate immune infiltration in certain models by facilitating CTL diapedesis. Conversely, GzmB can promote inflammation more directly through the cleavage of proinflammatory cytokines such as IL-1α, enhancing by several folds their biological activity (Afonina et al. 2011).
Granzyme B in Disease
GzmB is elevated in many autoimmune and/or chronic diseases, including, but not limited to chronic wound healing, asthma, chronic obstructive pulmonary disease, aneurysm, atherosclerosis, and multiple sclerosis. Researchers have successfully identified many of the mechanisms/pathways of GzmB in the pathophysiology of each of the diseases, with the presence of a chronic proinflammatory state being a unifying characteristic. Elevated levels of both intracellular and extracellular GzmB in these diseases implicate both GzmB‘s proapoptotic and extracellular role (Boivin et al. 2009).
GzmB is widely reported to be elevated in a range of skin disorders, including photoaging, diabetic ulcers, burns, alopecia areata, and Stevens-Johnson syndrome/toxic epidermal necrolysis (Reviewed in Boivin et al. 2009). Multiple skin cell types secrete GzmB, including mast cells, macrophages, and keratinocytes (Boivin et al. 2009). The excess of GzmB found in diseased skin ultimately leads to elevated cleavage of ECM proteins, such as collagen, decorin, and fibronectin, thereby interfering with skins reparative processes. Upregulation of GzmB is also reported in photoaged skin, with increased secretion shown in keratinocytes responding to UVA and UVB exposure (Parkinson et al. 2015). This GzmB-mediated ECM degradation contributes to premature skin aging, including wrinkle formation.
A large influx of GzmB secreting lymphocytes is observed in various inflammatory lung disorders (Reviewed in Boivin et al. 2009). Increased expression of GzmB is seen in the bronchoalveolar lavage and lung specimens of chronic obstructive pulmonary disease (COPD) and asthma patients. Common ECM proteins damaged in the lung include fibronectin, vitronectin, and laminin, all of which suggest a role for extracellular GzmB in remodeling and destruction (Boivin et al. 2009; Hendel et al. 2010).
GzmB is elevated in aortic aneurysm, atherosclerosis, allograft vasculopathy (chronic transplant vasculopathy), and cardiac fibrosis, with all of these associated with chronic inflammation (Reviewed in Hendel et al. 2010). With the exception of allograft vasculopathy, pathology is characterized by reduced ECM integrity, thus facilitating impaired tissue function and/or rupture. GzmB expression tends to increase with disease severity and is observed in the extracellular milieu (Hengartner 2000). Elevated GzmB in conjunction with inflammation is also associated with Kawasaki disease and giant cell arteritis (Boivin et al. 2009).
During CTL-mediated apoptosis, GzmB exposes epitopes on intracellular proteins of which would not normally be present in a healthy cell (Darrah and Rosen 2010). These epitopes would not be present during immune cell maturation or development of tolerance, and thus help trigger autoimmune responses. Many of these fragments are specific to GzmB proteolysis and not seen in any other forms of cell death (Darrah and Rosen 2010). As such, autoantibody secretion in response to these fragments leads to the pathogenesis of several autoimmune diseases including, but not limited to, systemic lupus erythematosus, myositis, multiple sclerosis, and rheumatoid arthritis (Darrah and Rosen 2010). In a number of autoimmune skin disorders, including psoriasis, atopic dermatitis, and acne, GzmB is also significantly elevated compared to healthy skin (Bovine et al. 2009).
Therapeutic Inhibition of Granzyme B
GzmB knockout mice exhibit improved healing compared to wild-type controls in variety of disease states, including models of angiotensin II-induced aortic aneurysm (Ang et al. 2011) and cardiac fibrosis (Shen et al. 2016), chronic low-dose ultraviolet light irradiation (Parkinson et al. 2015), diabetic wounds (Hsu et al. 2014), and atherosclerosis (Hiebert et al. 2013). These salient studies suggested that inhibitors of GzmB may have therapeutic value in promoting wound repair in chronic disease states. In support, topical GzmB inhibition can accelerate wound closure and promotion of both granulation tissue maturation and collagen deposition in a mouse model of diabetic wound healing (Hsu et al. 2014). In experimental autoimmune encephalomyelitis mice, used as a model of multiple sclerosis, GzmB inhibition was also shown to reduce axonal and neuronal injury compared to the vehicle-treated control group whilst also maintaining the integrity of myelin (Haile et al. 2015). Additionally, the GzmB inhibitor serpina3n is reported to prevent rupture and increase survival in murine model of abdominal aortic aneurysm (Ang et al. 2011). Together, inhibitors of GzmB are a promising therapeutic for a range of autoimmune and/or chronic inflammatory diseases.
GzmB, a protease with a diverse repertoire of biological activities, is emerging as having a pivotal role in both inflammatory and degenerative disease. Although the pathological roles of GzmB are complex, studies are beginning to reveal this protein as a valuable therapeutic target.
The authors wish to apologize to the many authors involved in studies that were uncited in this article and acknowledge that numerous references were not included due to space/reference limitation. Many of the references therefore refer to review articles where the original studies are cited.