Proteinase 3 and Serpin B1: a novel pathway in the regulation of caspase-3 activation, neutrophil spontaneous apoptosis, and inflammation

Abstract

Neutrophils are the first responders of the inflammatory response. They are characterized by their potent cytotoxic content but also by their limited lifetime. This short half-life is thought to be a self-protecting mechanism for the host, as highlighted by the numerous pathologies associated with imbalanced neutrophil survival. Neutrophil spontaneous death is the prototype of programmed cell death, harboring all the phenotypic hallmarks of apoptosis and dependent on the activation of the effector caspase-3. However, the pathways regulating neutrophil spontaneous death remain ill-defined. In a recent publication, we determined that in aging neutrophils, the cleavage and activation of caspase-3 was mediated by the serine protease Proteinase 3 (PR3), and was independent of the canonical extrinsic and intrinsic apoptosis pathways. In mature neutrophils, PR3 was stored in granules and progressively released to the cytosol during neutrophil aging. The release of PR3 was dependent on lysosomal membrane permeabilization (LMP). Once in the cytosol, PR3 cleaved procaspase-3 at a site upstream of the caspase-9 cleavage site, leading to caspase-3 activation. Inhibition, knockdown or knockout of PR3 delayed neutrophil apoptosis in vitro and in vivo. The adoptive transfer of both WT and PR3-deficient neutrophils to WT mice revealed that the delayed death of neutrophils lacking PR3 in vivo was due to an altered intrinsic apoptosis/survival pathway and not to difference in the inflammatory microenvironment. The cytosolic inhibitor of serine proteases serpin b1 counterbalanced the activity of PR3 in the cytosol of neutrophils, and the deletion of serpinb1 in neutrophils accelerated their spontaneous death. In summary, our results reveal that PR3 and serpinB1 are part of a newly characterized apoptosis pathway, regulating caspase-3 activation and neutrophil spontaneous death and the survival of neutrophils during inflammation.

Neutrophils’ function in the immune response has been often regarded as simple “foot-soldiers”: to find and kill the invading pathogens. It is now clear that besides their classical anti-bacterial and anti-fungi activities, neutrophil also participate in the regulation of the immune response. They interact and process information, are part of immune crosstalks with T-cells, B cells, and even dendritic cells and can secrete a wide variety of cytokines ^[1]. Even in their most basic task, i.e. capturing and killing of invading pathogens, neutrophils surprised researchers with the discovery of their ability to release extracellular traps. Neutrophil extracellular traps are composed of chromatin and cytotoxic granular content that enable them to carry on the fight against invaders even after their death ^[2].

Neutrophils spontaneous apoptosis has been studied for many years, and is regarded as the prototypical form of programmed cell death. It is dependent on the activation of cysteine-aspartic proteases or caspases. Numerous extrinsic factors can regulate this process and have been characterized ^[3]. In non-inflammatory situation, neutrophil apoptosis is thought to be death signal independent ^[4]. Nevertheless it remains unclear how this death program is initiated ^[5]. In our recent publication, “Proteinase 3–dependent caspase-3 cleavage modulates neutrophil death and inflammation”, we took on to better characterize the molecular mechanisms leading to the constitutive neutrophil death ^[6].

We first observed that the survival of neutrophils in culture was significantly delayed in the presence of pan-caspase inhibitor z-VAD-fmk or caspase-3 specific z-DEVD-fmk. Surprisingly, neither inhibition of the extrinsic (caspase-8) or the mitochondrial (caspase-9) pathways could prevent neutrophil spontaneous death. This suggested that caspase-3 was the effector caspase involved in neutrophil apoptosis, but the initiation of the cell death program was dependent on an uncharacteristic pathway. To identify the pathway responsible for the activation of caspase-3 in aging neutrophils, we developed a cell free assay using a N-terminal tagged procaspase-3 as a reporter of its own activation. Pro-caspase-3 does not arbore autocatalytic activity. To get fully activated, the pro-caspase-3 precursor must first be cleaved by caspase-8, caspase-9 or an unidentified protease between its large and small subunits. The presence of a N-terminal His-tag enabled us to visualize this process. As expected, no caspase-3 processing activity was detected in the cytosol of freshly isolated neutrophils. Interestingly, recombinant procaspase-3 was cleaved and activated by the cytosol of neutrophils cultured for 16 hours. Caspase-3 cleavage was not prevented by the inhibition of caspases, cathepsins or calpains. Only Diisopropylfluorophosphate (DFP), a potent inhibitor of serine proteases was able to inhibit the processing of the reporter caspase-3 in vitro. Treatment with DFP also significantly delayed neutrophil apoptosis. Interestingly, death pathways associated with serine protease activity was previously reported in a myeloid leukemia cell line ^[7]. We showed that the serine protease responsible for caspase-3 activation was present in neutrophils granules of freshly isolated cells and in the cytosol of aged neutrophils.

We undertook to identify the serine protease involved in caspase-3 cleavage. Neutrophils express 4 serine proteases: Neutrophil Elastase (NE), Cathepsin G (CathG), Proteinase3 (PR3) and the recently characterized NSP4 ^[8, ^9]. A fifth protein called Azurocidin, which is related to the NSP family and is expressed in neutrophils, lacks protease activity. These NSP are expressed simultaneously during neutrophil differentiation ^[10]. Using purified NE, Cath G and PR3, we showed that PR3 was the only NSP capable of cleaving caspase-3. The cutting site we identified was different from the canonical cleavage site of caspase-9, but enzymatic activity toward a specific colorimetric substrate of the PR3-cleaved procaspase-3 was similar to the enzymatic activity of caspase-9 activated procaspase-3.

PR3 is stored in all neutrophil granules ^[11]. Nevertheless, we showed by immunofluorescence and by biochemical fractionation that PR3 was released from the granules during neutrophil constitutive death. This release was due tolyzosomal membrane permeabilization (LMP) as inhibition of LMP prevented the release of PR3. Inhibition of LMP also resulted in inhibition of PMN constitutive apoptosis drastically. These results showed that the release of PR3 to the cytosol was essential for the regulation of apoptosis. Interestingly, PR3 optimal activity toward caspase-3 was at neutral pH (cytosolic) and not at acidic pH (in azurophilic granules).

We also noted that the PR3 activity was inhibited by the cytosolic fraction of freshly isolated PMN, suggesting the presence of an inhibitor of serine proteases in the cytoplasm of neutrophils, inhibitor that could counterbalance PR3 activity and prevent inappropriate caspase-3 activation. Serpins are a family of suicidal inhibitors of serine proteases. Someserpins are secreted (clade A, such as serpin A1 also called A1-AT), while others are intracellular inhibitors (clade B) ^[12]. Recent reports showed that neutrophils lacking serpin b1 had an increased death rate ^[13]. We were able to show that during neutrophil constitutive death, PR3 was forming a complex with SerpinB1. We also confirmed that the survival of serpin b1 deficient neutrophils was impaired. PR3 knockdown by siRNA in mouse bone marrow neutrophils led to an increased survival of the cells. Therefore, our data suggested that PR3 is causally related to the constitutive neutrophil death ex vivo.

To elucidate the role of PR3 in the regulation of neutrophil survival in vivo in normal and inflammatory situation, we generated PR3 null mice. PR3^−/−mice were previously generated but as a Neutrophil Elastase (NE)/PR3 double knockout ^[14], making it difficult to discriminate the specific functions of PR3 from NE functions. The intraperitoneal injection of live E. coli induced similar recruitment of neutrophils in WT and PR3^−/− mice. Nevertheless, we could observe a prolonged presence of neutrophils in the peritoneal cavity of the PR3 null mice, suggesting an increased survival of the cells. Indeed, the percentage of Annexin V positive cells was higher for WT cells than for PR3^−/−. To circumvent the possible effect of PR3 knockout on the multiple steps leading to the recruitment of neutrophils at the site of inflammation, we performed an adoptive transfer experiment, in which the same number of WT and PR3^−/−neutrophils were injected in the peritoneal cavity of a WT mouse. More viable transplanted PR3-deficient neutrophils than WT neutrophils were detected in the inflamed peritoneal cavity, directly supporting the role of PR3 as an intrinsic regulator of neutrophil survival in vivo.

Concluding remarks and future directions

Just as neutrophil function is not limited to bacteria killing, neutrophil serine proteases (NSP) are more than anti-bacterial enzymes. It is known that NSP are also important regulators of inflammatory process and granulopoiesis ^[15]. Numerous physiological substrates are converted, inactivated or activated by PR3 ^[15, ^16]. PR3 is also present in the secretory vesicles, and that secreted PR3 is sufficient to induce apoptosis of epithelial cell through the inactivation of p21 ^[17]. Understanding how PR3 regulates neutrophil survival during infection will shed new light on the role of NSP beyond direct bacteria killing. Characterizing the multiple facets of NSP may hopefully lead to new strategies to treat inflammatory disorders, acute and chronic, in which neutrophils are involved.