Finding a Balance between Protection and Pathology: The Dual Role of Perforin in Human Disease

Abstract

Perforin is critical for controlling viral infection and tumor surveillance. Clinically, mutations in perforin are viewed as unfavorable, as lack of this pore-forming protein results in lethal, childhood disease, familial hemophagocytic lymphohistiocytosis type 2 (FHL 2). However, many mutations in the coding region of PRF1 are not yet associated with disease. Animal models of viral-associated blood-brain barrier (BBB) disruption and experimental cerebral malaria (ECM) have identified perforin as critical for inducing pathologic central nervous system CNS vascular permeability. This review focuses on the role of perforin in both protecting and promoting human disease. It concludes with a novel hypothesis that diversity observed in the PRF1 gene may be an example of selective advantage that protects an individual from perforin-mediated pathology, such as BBB disruption.

Keywords: blood–brain barrier disruption; familial hemophagocytic lymphohistiocytosis type 2; perforin; selective advantage; single nucleotide variants.

Figure 1

Current models of perforin granzyme synergy to induce apoptosis. There are two models to explain the synergistic effect between granzymes and perforin to induce apoptosis in a target cell. Both models are similar in regards to the process by which cytotoxic granules are produced by the effector cell. The difference lies at the target cell membrane surface. In Model 1, perforin forms a pore on the plasma membrane of the target cell allowing granzyme to be delivered to induce apoptosis (A). In model 2, granzyme and perforin are released from the cytotoxic granule, enter the target cell, and are then repackaged within the endosome. Perforin then forms a pore within the endosome, disrupting the membrane integrity, allowing granzyme to escape and induce apoptosis (B).

Figure 2

Proteomic and genomic organization of perforin. Linear representation of matured perforin protein illustrating the various domains and amino acids of this molecule (A). Diagram of PRF1 indicating the nucleotide position of each exon, indicated by number above the diagram. Amino acids encoded by the respective exons are indicated by number below diagram. Mutations found in human populations, as reported in the literature, are listed below each corresponding exon (B). This list is not exhaustive of all PRF1 SNVs identified.

Figure 3

Summary of FHL genetic etiology. A schematic showing the role of the gene causative of FHL subtypes with the listed mutations. Genetic etiology for FHL 1 is not known and is therefore not reported in this Figure. Mutations listed were listed in primary literature. Please use the ExAC Browser database (http://exac.broadinstitute.org/) for more information regarding the mutations in these genes.

Figure 4

2D representation of perforin mediated disruption of the blood–brain barrier. Surrounding the blood vessel is a layer of endothelial cells held together by tight junctions. Encompassing the enodthelial cell layer are various brain cells comprising the neural vascular unit (NVU). NVU cell types include pericytes, astrocytes, microglia, and neurons. Additionally, sinuses run through this dense environment carrying cerebral spinal fluid (CSF) throughout the brain (A). During neuroinflammation, the blood–brain barrier can be disrupted through a perforin dependent process. During this process, CEC tight junctions become disorganized and CNS permeability occurs. Where there was once vascular integrity and sequestration of the CNS, there is now increased non-regulated influx of molecules from the blood (B).

Figure 5

The balance of perforin expression may be advantageous to an individual’s fitness. Hypothetical model of perforin’s contribution to pathogen clearance and vascular permeability. Loss of perforin activity in the human due to single deleterious mutation (red cross) or multiple, compounding mutations (yellow crosses) leads to disease FHL 2. In the mouse model, mice lacking perforin cannot clear virus in the CNS but do not experience BBB disruption. Perforin competent mice can clear virus in the CNS but experience lethal BBB disruption. The dashed line and question mark indicate a level of perforin, found in the human population that may mediate BBB disruption but still allow for CNS pathogen control (A). At full perforin activity, CNS viral control is possible, however there is pathologic BBB disruption. At no perforin expression, pathologic BBB disruption is not present, but there is little of CNS viral infection. Non deleterious perforin SNVs that decrease but do not abolish perforin’s activity may provide a balance between these two pathologies (B). The five SNVs listed are purely there as a frame of reference, we do not conclude that these exact polymorphisms play a role in this balance.