Caspases

Apoptosis is orchestrated by a family of proteolytic enzymes called caspases (cysteine-dependent aspartate-specific proteases) which cleave target proteins at aspartate residues. The caspases are synthesized as inactive procaspases and once activated are able to cleave parts of the neuronal cytoskeleton and nuclear DNA.

Many components of the caspase cascade converge on the effector (or executioner) proteins, such as caspase-3, which can be used as a marker of apoptosis. Programmed cell death can be triggered by extrinsic or intrinsic stimuli.

Extrinsic pathway (Fig. 8.2)

Apoptosis can be triggered by cell death ligands in the extracellular fluid. These act on cell-surface receptors belonging to the tumour necrosis factor (TNF) family. After ligand binding, death receptors cluster within the membrane to form trimers and a conformational change in the cytoplasmic part of the protein exposes an ominously named death domain. The intracellular region of the cell death receptor binds apoptotic proteins in the cytoplasm via adaptor proteins. The interaction is mediated by death effector domains (DEDs) that are present in both the adaptor proteins and the procaspases. These interactions lead to formation of a death-inducing signalling complex (DISC) which activates caspase-8 and thereby initiates programmed cell death.

Intrinsic pathway (Fig. 8.3)

Apoptosis can also be triggered from within the cell if it is exposed to excessive pathological stress. Activation of the intrinsic pathway depends upon the balance of pro-apoptotic and anti-apoptotic factors, many of which are members of the BCL (B-cell lymphoma) family of proteins. Members that tend to oppose programmed cell death include Bcl-2 and Bcl-XL; others, such as Bad and Bax are pro-apoptotic. These molecules act as stress sensors in the cytoplasm, responding to pathological stimuli by translocating to mitochondria.

A key event is formation of a permeability transition pore (PTP) in the mitochondrial membrane, which is a large transmembrane pore (or ‘megachannel’). Pore formation allows cytochrome c oxidase and apoptosis inducing factor (AIF) to escape from mitochondria and reach the cytoplasm. Cytochrome c then forms a complex with Apaf-1 (apoptotic protease activating factor 1) which recruits caspase-9 to form a larger multi-protein complex called the apoptosome. Following formation of the apoptosome, caspase-9 is activated and apoptosis is triggered, culminating in the upregulation of cell death genes.

Disposal of the cell

Once a cell is committed to programmed cell death, its DNA and cytoskeleton are dismantled in an orderly manner. This is an active process that expends energy. A key step is activation of the enzyme caspase-activated DNAse (CAD) by effector caspases, which breaks down the DNA into nucleosomal units. Organelles are packaged into membrane-bound apoptotic bodies (see Fig. 8.1) which contain viable mitochondria. These structures express cell-surface markers that trigger their internalization by neighbouring cells. An example is the membrane constituent phosphatidylserine, which translocates from the inner to the outer leaflet of the plasma membrane. Phagocytes recognize and bind these molecules and internalize the apoptotic bodies for degradation. The entire process is carefully orchestrated and, in contrast to necrosis, there is no inflammatory reaction.

Transneuronal degeneration (Fig. 8.4)

Following axonal transection, neuronal degeneration may spread to involve other nerve cells, referred to as transneuronal degeneration. For instance, following interruption of an anatomical pathway consisting of a linear chain of neurons, there may be subsequent loss of nerve cells ‘downstream’ (anterograde) or ‘upstream’ (retrograde) of the original injury, which may take months or years. This reflects the general principle that nerve cells need to be integrated into a functional network and receive trophic signals from other neurons in order to remain viable.

Axonal regrowth

Axons are able to regenerate following peripheral nerve damage and may re-establish connections with muscle fibres or glands. The distal tips of the severed axons form growth cones (Fig. 8.5) which ‘crawl’ along residual Schwann cell basement membranes to reinnervate target structures. This process occurs after peripheral nerve injury (see Clinical Box 8.1) but does not seem to be possible in the brain and spinal cord.

The role of calcium

The free calcium ion concentration in the cytoplasm is normally kept very low by several mechanisms including sequestration by calcium-binding proteins (e.g. parvalbumin and calbindin) and export from the cell. The high calcium influxes generated by excitotoxic stimulation overwhelm buffering and extrusion mechanisms, leading to activation of harmful calcium-dependent enzymes, including:

Several pro-apoptotic genes are also upregulated by calcium-mediated cascades, which promotes degradation of the cytoskeleton and may initiate programmed cell death. Calcium-mediated activation of xanthine oxidase and nitric oxide synthase may also lead to oxidative stress.

Nitric oxide

Nitric oxide gas is synthesized from L-arginine by isoforms of the enzyme nitric oxide synthase (NOS). It is a free radical species with a number of important physiological roles (e.g. as a vasodilator, transmitter substance and regulator of inflammatory and immune responses). Its synthesis is induced by glutamate signalling via the calcium-permeable NMDA receptor and excessive production of nitric oxide is a feature of excitotoxicity.

Nitric oxide reacts with superoxide anion to produce peroxynitrite (ONOO⁻). This is a reactive nitrogen species that may damage proteins by interacting with cysteine and tyrosine residues. Although nitric oxide has an anti-apoptotic effect in many cell types, excessive production contributes to cell death in neurodegenerative diseases, multiple sclerosis and stroke.

Astrocytosis (Fig. 8.7)

Following brain injury, nearby astrocytes enlarge, multiply and increase their expression of glial fibrillary acid protein (GFAP). Proliferation of astrocytes may be sufficient to fill in a small tissue defect, but larger areas of damage (e.g. following a major stroke, see Ch. 10) are transformed into a fluid-filled cystic cavity lined by a glial scar. Astrocytes secrete (i) cytokines that recruit inflammatory cells from the blood and (ii) various trophic factors, including:

These are chemical mediators that promote neuronal survival and axon sprouting. They are released into the extracellular fluid, but can also be delivered directly to the neuronal cytoplasm via intercellular gap junctions (see Ch. 7).

Microgliosis (Fig. 8.8)

Microglia are the resident phagocytes (scavengers) of the brain. They normally exist in a ramified, quiescent state, but following tissue injury they become activated in response to inflammatory cytokines and growth factors. Activated microglia migrate towards injured tissues by following chemotactic gradients. They differentiate into macrophages and internalize cellular debris and microorganisms. Those that have ingested myelin debris form lipid-laden foam cells.

Activated microglia are immunocompetent cells that express MHC class II (major histocompatibility) proteins and are antigen-presenting cells. They may therefore contribute to T-cell-mediated immune responses and have been implicated in the inflammatory demyelinating disease multiple sclerosis (Ch. 14). Microglial activation is also a component of most neurodegenerative disorders such as Alzheimer’s disease and Parkinson’s disease (Chs 12 and 13).

Inflammation in the CNS

A small focus of inflammation in the brain is referred to as cerebritis. More extensive and diffuse brain inflammation is termed encephalitis. Encephalitic processes are further subdivided into three main types:

The term myelitis indicates inflammation of the spinal cord, such as the inflammatory spinal cord disease poliomyelitis (caused by infection with a poliovirus); whereas combined inflammation of the brain and spinal cord is referred to as encephalomyelitis. It should be emphasized that these terms are descriptive and do not indicate the underlying cause of the inflammation.

Neuronal loss and gliosis

The affected brain regions show neuronal loss and reactive gliosis in a disease-specific pattern, affecting particular groups of neurons, but sparing others. Importantly, the clinical features in each case are determined by the anatomical distribution of the pathological changes rather than the particular protein involved. In most cases it is not clear why specific populations of neurons are susceptible whilst others are resistant (referred to as selective vulnerability).

Normal protein folding

Nuclear DNA encodes the primary structure of proteins, consisting of the basic amino acid sequence. Attainment of the correct three-dimensional conformation requires protein folding (Fig. 8.11). This transforms the linear amino acid sequence into more complex spatial arrangements with alpha-helices and beta-pleated sheets that make up the secondary structure. Further folding gives rise to a three-dimensional globular protein with a particular tertiary structure. Association with other proteins may occur, to form a multi-protein complex with its own quaternary structure. Protein folding relies on physical and chemical properties of the constituent amino acid residues (i.e. attraction and repulsion by hydrogen bonds, electrostatic forces and hydrophobic interactions).

Unfolded protein response

The unfolded protein response (UPR) is triggered by the presence of misfolded proteins in the endoplasmic reticulum or as a result of errors in post-translational modification. It is a type of cell stress response characterized by upregulation of molecular chaperone proteins that attempt to refold abnormally configured proteins. In the event of an overwhelming pathological stimulus or major dysfunction of the normal mechanisms for protein disposal, large numbers of misfolded proteins may aggregate in the cytoplasm. When this happens they tend to precipitate out of solution as insoluble clusters called micelles, since hydrophobic groups that would normally be buried are exposed to the aqueous environment of the cell.

Disposal of abnormal proteins

Abnormal proteins are earmarked for destruction by tagging them with the 8.5 kDa protein ubiquitin (Fig. 8.12). This is attached in a series of enzyme-catalysed steps involving activating (E1), conjugating (E2) and ligating (E3) enzymes. These add a polyubiquitin chain that is composed of multiple ubiquitin monomers. Once ubiquitinated, the abnormal protein is targeted to the proteasome. This is a large multi-subunit protein complex that digests proteins into peptide fragments and amino acids in an active (ATP-dependent) process.

The proteasome consists of a cylindrical core region (the 20S core particle) composed of four rings stacked around a central pore. This is capped at either end by regulatory particles (19S or 11S) which also have a central pore that allows access to a hollow degradation chamber. The entire assembly is the 26S proteasome. Proteins are fed into the proteasome and digested, whilst the polyubiquitin chain is recycled. Accumulation of ubiquitin-tagged proteins is a feature of many neurodegenerative disorders, sometimes due to disturbance or overload of the ubiquitin-proteasome system.

Amyloid

Many proteins and peptides that form pathological aggregates have a beta-pleated sheet structure. This enables monomers to stack together to form elongating protofibrils and fibrils of around 10 nm in diameter, stabilized by hydrogen bonds. Deposits of these insoluble protein fibrils are referred to as amyloid. It is important to emphasize that the term ‘amyloid’ does not refer to one particular protein and that many different peptides with a beta-sheet structure can form ‘amyloid deposits’. The name derives from the Greek, meaning starch-like.

All forms of amyloid take up certain tissue stains such as Congo red and thioflavin S. Due to the regular, crystalline arrangement of the amyloid fibrils, the deposits also have the ability to rotate the plane of polarized light, termed birefringence. As a result, amyloid deposits stained with Congo red have a characteristic apple green colour when viewed under polarized light (Fig. 8.14).

Amyloid fibril formation

The process of amyloid formation is illustrated in Figure 8.15. Fibrillogenesis is the process by which a peptide forms insoluble aggregates of amyloid. It involves three sequential steps. A peptide with a beta-pleated sheet structure must first be produced in sufficient quantities. The second step is nucleation, which starts the process of fibril formation. It requires a supportive microenvironment with a sufficiently high protein concentration, together with various permissive factors including appropriate acidity (pH), temperature or the presence of certain metallic ions. Finally, the phase of fibril growth involves the sequential addition of monomeric peptide units (each with a beta-sheet structure) to form an extending chain. This leads to the gradual assembly of oligomeric species (or protofibrils) which associate to form mature amyloid fibrils.

Amyloid toxicity

It is not certain how abnormal protein aggregates cause neuronal damage, but factors that have been implicated include inflammation, gliosis, oxidative stress and production of toxic intermediates during fibrillogenesis. There is growing evidence that oligomeric intermediates generated during amyloid fibril formation may be primarily responsible for neurotoxicity in a number of neurodegenerative disorders (including Alzheimer’s disease and Parkinson’s disease).

In particular, it has been shown in cultured cells that oligomeric prefibrillary species of amyloid beta are able to form a membrane attack complex that can perforate the neuronal cell membrane. This leads to cellular swelling, loss of transmembrane gradients and influx of free calcium ions – all of which promote neuronal cell death. A similar phenomenon has been described with oligomers of alpha-synuclein protein (the main pathological species implicated in Parkinson’s disease and related synucleinopathies).

Infectivity

In addition to sporadic and inherited forms, prion diseases have the unique property (among degenerative diseases) of infectivity. They are therefore also referred to as transmissible spongiform encephalopathies (TSEs). In the 1990s in the UK and Europe, bovine spongiform encephalopathy entered the human food chain via contaminated beef and led to a new variant of CJD (Clinical Box 8.4). Prion diseases have also been transmitted iatrogenically (Greek: iatros, doctor) by blood transfusions and growth hormone supplements (obtained from human donors) and various neurosurgical procedures (via contaminated instruments or dural grafts).

 Clinical Box 8.4:   New variant CJD

The new variant of CJD was first identified by its microscopic appearance. This differs from that of classical CJD and includes characteristic florid plaques, named for their resemblance to flowers (Fig. 8.17). The new variant affects younger people (usually below the age of 30) and has a longer duration, typically a year or more. Psychiatric features and cerebellar ataxia are also common and myoclonic jerks are absent. MRI scanning characteristically shows T2-hyperintensity in the posterior thalamus (the pulvinar sign). As with classical CJD, the condition is incurable and fatal, but it is much rarer, with fewer than 180 cases reported in the UK (representing 80% of the worldwide total).

Codon 129

There is a common polymorphism in the general population that affects codon 129 of the prion gene (PRNP, located on chromosome 20) which codes either for methionine (M) or valine (V). Since each gene has two alleles, there are three possible genotypes with different population frequencies:

All pathologically confirmed cases of variant CJD have occurred in people who are homozygous for methionine at codon 129 (MM). This has therefore been referred to as the susceptibility genotype, which implies that the other forms confer resistance to infection. In keeping with this idea, mice that are homozygous for valine (VV) are virtually immune to inoculation with abnormal prion protein whereas heterozygotes (MV) show intermediate susceptibility. However, since the incubation period for prion diseases is sometimes measured in decades, the possibility remains that new cases of variant CJD will eventually emerge in people who do not have the susceptibility genotype.

Conversion of PrP^c to PrP^Sc

In sporadic disease the abnormal prion protein is assumed to arise by spontaneous transformation of native prion protein to the abnormal (scrapie) form. This is thermodynamically unfavourable and is an extremely unlikely spontaneous event, which presumably explains why sporadic disease is so rare.

Animal models have shown that propagation of abnormal prion protein can only take place if the normal cellular protein is present. For instance, prion-knockout mice are completely resistant to infection, but infectivity can be restored if native prion protein is reintroduced.

Conversion models

There are two main models that attempt to describe how cellular prion protein may be converted to the abnormal form. The first is the template-directed refolding (heterodimer) model (Fig. 8.19). This suggests that abnormal prion protein is able to recruit the cellular form and form a heterodimer with it, acting as a template to catalyse its conversion to the scrapie form. The converted prion can then recruit more cellular protein or polymerise to form amyloid fibrils.

Another possibility is described by the nucleation–polymerization model which suggests that the critical event is the formation of a nucleus (which consists of an oligomeric aggregate of prion protein). This initial ‘seeding’ event is highly unlikely since it is thermodynamically unfavourable – but once the nucleus has formed, polymerization and fibril elongation proceed rapidly.

Key Points

 The pathogenic agent in transmissible spongiform encephalopathies such as CJD and variant CJD is the prion (proteinaceous infectious particle). This is a unique pathogen that appears to be composed entirely of protein.

 The pathogenic (scrapie) form of prion has the same primary structure as the native protein, but a different secondary structure that is rich in beta-pleated sheets.

 Animal models show that infection and propagation can only take place if the normal cellular protein is present, since prion-knockout mice are resistant.

 The template-directed refolding and nucleation-polymerization models are attempts to explain how abnormal prion protein might recruit and convert native (cellular) prion protein.