Abstract
This review examines the role of the neuropeptide substance P within the neuroinflammation that follows traumatic brain injury. It examines it in reference to its preferential receptor, the neurokinin-1 receptor, and explores the evidence for antagonism of this receptor in traumatic brain injury with therapeutic intent. Expression of substance P increases following traumatic brain injury. Subsequent binding to the neurokinin-1 receptor results in neurogenic inflammation, a cause of deleterious secondary effects that include an increased intracranial pressure and poor clinical outcome. In several animal models of TBI, neurokinin-1 receptor antagonism has been shown to reduce brain edema and the resultant rise in intracranial pressure. A brief overview of the history of substance P is presented, alongside an exploration into the chemistry of the neuropeptide with a relevance to its functions within the central nervous system. This review summarizes the scientific and clinical rationale for substance P antagonism as a promising therapy for human TBI.
Introduction
Trauma is a significant cause of death globally, 1 with traumatic brain injury (TBI) a leading cause of death and disability within the spectrum of traumatic injuries worldwide, especially in the younger populations (< 45 years) of both developed and developing nations. 1,2
Approximately 40% of those sustaining severe TBI will die from their injury, and 60% will have an unfavorable outcome (including death) based on the Glasgow Outcome Scale. 3 Mortality increases based on the severity of the injury, 4 and those who survive may be left with debilitating physical and neurocognitive impairments that have a negative impact on the survivor's daily living and ability to work, while also negatively impacting surrounding family and the greater healthcare and socioeconomic system that must cater for post-TBI disability. 5
Despite an inordinate cost to healthy human life and the world economy, no significant breakthroughs in the clinical management for TBI have occurred in the past 40 years. 6 While preventative measures in the form of road laws and mandatory safety devices have proven to be successful in reducing incidence of TBI, 7,8 there has been no progress in developing neuroprotective pharmacological interventions. Well-designed clinical trials in TBI patients are difficult due to the heterogenous nature of the disease and patient cohorts, as well as a lack of appropriate early end-points. 3 This highlights the importance of phase II experimental medicine trials as a means of addressing dosage, timing, and pharmacokinetics of novel agents.
As our understanding of TBI pathophysiology has developed, the scientific literature has moved towards a molecular and mechanistic understanding of the underlying processes that drive cellular injury. In particular, the role of inflammation has been proposed as a key driver of injury and a potentially tractable therapeutic target. Several inflammatory processes, including both innate inflammation and adaptive immunity, have been reported in human TBI. The initiators of inflammation that link directly with tissue trauma or deformation are an area of active study, and this has led the concept of “neurogenic inflammation.”
Neurogenic inflammation occurs as part of this secondary brain injury, a result of the neuropeptide substance P's release from primary afferent neurons that, through a complex chain of events, leads to breakdown of the blood–brain barrier, generation of a classical inflammatory response, and the development of cerebral edema. 9 -11 This cerebral edema results in a rise in intracranial pressure (ICP), a significant cause of mortality and poor outcomes in TBI patients. 12,13
This review will investigate the role of the neuropeptide substance P in the development of cerebral edema following traumatic brain injury and focus on the evidence for inhibition at its preferential receptor, the neurokinin-1 receptor. There has been growing interest in substance P and its potential involvement within neuroinflammation following TBI, with much of the literature accounting studies performed in animal models. This review will consolidate the current literature in animal and man, focusing on the documented uses of neurokinin-1 receptor antagonists that block the action of substance P and its effects.
Secondary Brain Injury Following TBI
From a clinical and therapeutic perspective, it is secondary brain injury that is modifiable and therefore a tractable target for intervention. Secondary brain injury encompasses a range of processes including a combination of neurogenic and classical inflammatory cascades, release of excitotoxic amino acids and neurotransmitters, loss of ionic homeostasis, activation of matrix metalloproteinases (MMPs), and the generation of free radicals. 9 –11,14 These deleterious effects result in oxidative stress, neurotoxicity, and recruitment of inflammatory cells into the site of injury, all furthering cell injury. Hyperpermeability and breakdown of the blood–brain barrier (BBB) lead to the development of vasogenic edema, 9 -11 with the extracellular nature of this fluid resulting in an increased brain volume. 11,15 This results in an increased intracranial pressure (ICP), a major therapy target by neurocritical care due to the mortality and poor outcomes associated with its rise. 12,13 This insidious process begins at the point of the primary injury, and continues to occur in the days to weeks following. 11
Substance P is upregulated following TBI, with increased perivascular SP immunoreactivity. 16,17 The presence of elevated serum substance P levels is associated with an increased mortality and an indicator of severity of injury, 18,19 with the initial opening of the BBB and the associated vasogenic edema thought to invite and accentuate further classical inflammation. 20 -22
Neurogenic inflammation
The concept of neurogenic inflammation was originally described in 1901 by Bayliss, when he explored vasodilation in the lower limbs of dogs. 23 He theorized that the sensory nerve branches from the dorsal root ganglion that were supplying the blood vessels were responsible for the vasodilatory effect seen. Neurogenic inflammation is now described as an inflammatory response that is neurally elicited through stimulation of capsaicin-sensitive primary sensory neurons, and the subsequent release of neuropeptides that elicit a characteristic inflammatory response. 24,25 Sensory neurons respond to changes in pH and temperature, compounds such as serotonin and histamine, as well as cytokines and direct mechanical injury. 26 Released neuropeptides include neurokinin A and B (NKA and NKB), as well as calcitonin gene-related peptide (CGRP), and of most interest in this review, substance P (SP). 20 The resulting combination of vasodilation, protein extravasation, vascular permeability, and leukocyte adhesion is termed neurogenic inflammation. 24,27 CGRP is a potent vasodilator, while vascular permeability and protein extravasation is a result of NKA, NKB, and chiefly SP, of which all belong to the tachykinin family of neuropeptides. 28 CGRP and SP often coexist within primary sensory neurons. 29 CGRP's release, in addition to vasodilation, causes an upregulation of the neurokinin-1 receptor and competes with SP for catabolism by endopeptidases, resulting in an increased concentration of SP and potentiating its biological effects. 30
Breakdown of the blood–brain barrier occurs as a consequence of increased vascular permeability and protein extravasation of the cerebral arteries. These arteries are surrounded by a dense supply of sensory neurons containing CGRP and SP. 31 This blood–brain barrier dysfunction leads to an influx of water, termed vasogenic edema, and in combination with cerebral vasodilation, a cause of raised ICP and its associated poor outcomes following TBI. Further, SP is involved in augmenting the classical inflammatory cascade. It acts to produce nitric oxide and cytokines, activates mast cells, microglia, and astrocytes, and facilitates the recruitment of peripheral immune cells to the site. There is also evidence SP acts to increase transcytosis through endothelial cells via caveolae transport, allowing shifts of large proteins like albumin, which increase the osmotic gradient and result in the release of inflammatory cytokines through microglial and astrocytic activation (Fig. 1). 20

Diagrammatic representation of neurogenic inflammation at the blood-–brain interface. Substance P binds to the neurokinin-1 receptor, causing neurogenic inflammation. Potentiation of classical inflammation can be seen to occur with the activation of microglia and astrocytes, as well as recruitment of peripheral immune cells. Adapted from “Neurogenic inflammation after traumatic brain injury and its potentiation of classical inflammation,” by Corrigan and colleagues.
20
(Adapted figure created with BioRender.com, used under the Creative Commons Attribution 4.0 International License (
The other distinct family of kinins implicated in the mediation of neurogenic inflammation is the slow-acting kinin family. They consist of the peptides bradykinin and kallidin (Lys-bradykinin), which also are released after central nervous system (CNS) injury. 32 Kinin peptide release occurs in both tissue and plasma. Cell death and the release of lysosomal enzymes in tissue injury activates the low molecular weight precursor protein kininogen, leading to the release of kallidin. In blood, the activation of Hageman Factor XII by damaged surfaces leads to the release of bradykinin from the high molecular weight kininogen. 32,33 Kinin B1 and B2 receptors are expressed throughout the CNS. 34 These receptors are G-protein coupled receptors, where activation results in the conversion of phosphatidylinositol-4-5-bisphosphate to inositol 1,2,5-trisphosphate (IP3) and diacylglycerol (DAG), leading to a rise in intracellular calcium and initiation of secondary messenger systems. 34
Downstream effects of kinin receptor activation include the release of arachidonic acid through the activation of phospholipase A2, which is metabolized into prostaglandins and reactive oxygen species, which contribute to inflammation and oxidative stress. 32,35 Glutamate release is also implicated in bradykinin binding, leading to intraneuronal rises of Ca2+ and activation of enzymes that include phospholipases, endonucleases, and proteases, leading to cellular damage, termed glutamate excitotoxicity. 32 Another consequence is nitric oxide (NO) release, which causes vasodilation and further contributes to oxidative stress. 32 B2 kinin receptor stimulation on vascular endothelial cells by bradykinin increases vascular permeability and plasma extravasation. 35 Additionally, degranulation of mast cells is caused by both bradykinin and SP, resulting in the release of serotonin and histamine, and causing further vascular permeability and plasma extravasation. 32
The kinin and tachykinin systems are believed to act in a synergistic manner in the way they initiate and maintain inflammatory responses, with evidence that endogenous kinins cause inflammatory responses by stimulating the release of CGRP, NKA, and SP. 32,36,37 Despite the strong evidence for kinin involvement in secondary injury, no clear benefit has been identified in clinical trials assessing the use of bradykinin receptor antagonists in patients with traumatic brain injury. 38,39
Substance P
Discovery
Substance P was originally discovered in 193140 by Ulf von Euler, a postdoctoral student in training, and John Gaddum, senior assistant to Henry Dale, when attempting to further describe the distribution of acetylcholine in various organs. They identified an extract in equine brain and intestine that would cause stimulation of smooth muscle and potent hypotension in atropinized rabbit, an effect therefore independent of acetylcholine. It was initially isolated from horse small intestine and was described as “Preparation P,” 40 only becoming publicly termed “Substance P” in 1934 by Gaddum and Schild 41 after developing the private nickname in the laboratory from the purified extracted powder “substance” within the various stored bottles labelled P1, P2, etc. 42
Work on substance P continued after Euler returned to Stockholm to become professor of physiology at the Karolinska Institutet. Bengt Pernow joined Euler's lab as a junior assistant and published his doctoral thesis on the purification of SP in 1953, as well studies on its distribution and biological actions. 43 Not until 40 years later would substance P be isolated, and the undecapeptide amino acid structure identified by Chang and colleagues in 1971. 44 It was successfully synthesized the same year by Tregear and colleagues 45 before being developed as a radioimmunoassay in 1973 46 and facilitating further studies into the distribution of the neuropeptide. 47 –51
The mammalian tachykinin family (of which substance P is an example) would later be established with the discovery of neurokinin A (formerly known as substance K) and neurokinin B (formerly neuromedin K) in 1983. 52,53 The remaining neuropeptide K in 1985, neuropeptide gamma in 1988, and hemokinin-1 in 2000 would be soon to follow. 54 -56 Their designation as “tachykinins” relates to their fast stimulant action on smooth muscle and hypotensive properties, in contrast to the slow-acting kinins—the “true bradykinins.” 57
Structure
Chang and colleagues identified the amino acid structure of substance P in 1971 as the undecapeptide (comprising of 11 amino acids) Arg 1 -Pro 2 -Lys 3 -Pro 4 -Gln 5 -Gln 6 -Phe 7 -Phe 8 -Gly 9 -Leu 10 -Met 11 -NH2. 44 It has a relative molecular mass of 1.3 kDa. 58 All tachykinins are amidated at the carboxyl terminal methionine. Substance P is an amphiphilic peptide, with positively charged residues on the N-terminus and hydrophobic residues on the C-terminus. It carries a net positive charge at physiologic pH. 59
The mammalian tachykinins are closely related, sharing a common C-terminal sequence: Phe 7 -X 8 -Gly 9 -Leu 10 -Met 11 -NH2 (X is Phe or Val). 60 The common carboxyl-terminal domain interacts with the tachykinin receptors, while the amino-terminal sequence dictates receptor specificity. 61 Neuropeptide K and neuropeptide gamma are elongated forms of neurokinin A. 62
Synthesis
Substance P is encoded by the preprotachykinin-1 (PPT1/TAC1) gene, located on the long arm of chromosome 7 between positions 21 and 22 (7q21-q22). 63 TAC1 is a 7 exon gene, with the sequence encoding SP contained within exon 3 (Fig. 2). 64 Other tachykinins that are encoded by the TAC1 gene include neurokinin A, neuropeptide K, and neuropeptide gamma. 63 The remaining neurokinin B and hemokinin-1 are encoded by TAC3 and TAC4, respectively. 60 Four messenger ribonucleic acid (mRNA) isoforms result via alternative RNA splicing of the TAC1 gene transcript: α-PPT, β-PPT, γ-PPT and δ-PPT. All four encode the substance P precursor protein, while only γ-PPT and β-PPT encode for the precursor for neurokinin A, neuropeptide K, and neuropeptide gamma. 65 β-PPT is the predominant mRNA found within human brain tissue, approximately representing 80-85% of the total PPT mRNA. The remaining PPT mRNA is γ-PPT, with no detectable α-PPT. 66 This is in contrast to a study investigating striatal preprotachykinin gene expression within Sprague-Dawley rats, where γ-PPT is found to predominate. 67

A representation of the synthesis of the tachykinins. The TAC1 gene is responsible for the substance P (SP), neurokinin A (NKA), neuropeptide K (NPK) and neuropeptide gamma (NPγ) peptides, with the corresponding sequences contained within the exons involved in encoding labelled. The TAC3 and TAC4 genes are responsible for the peptides neurokinin B (NKB) and hemokinin-1 (HK-1) respectively. Created with
Synthesis of substance P occurs within the ribosomes in the perikaryon. 68 Following synthesis, SP is packed into synaptic vesicles and axonally transported to terminal regions, 69 concentrated within the axon terminals of primary afferent neurons. 70
The expression of PPT mRNA increases following noxious stimulation, as discovered in rats by Noguchi and colleagues. 71 Following injection of formalin into the right hindpaw, dorsal root ganglion neurons showed increased expression of PPT mRNAs. This mirrors previous studies 72,73 that observed an increase in SP immunoreactivity in the dorsal horn following chemogenic nociceptive stimulus and correlates the increased PPT gene expression with SP synthesis. Similar increases in SP-immunoreactivity have been observed following traumatic brain injury and spinal cord injury in humans. 16,17 Prominent immunoreactivity has been observed in the cerebral cortical pyramidal cells, cortical neurons and astrocytes, as well as an extensive perivascular distribution of SP nerve endings demonstrated in the cerebral cortical microvasculature in postmortem tissue of TBI patients. No SP immunoreactivity was observed within the microvascular endothelium. 16
Metabolism
Metabolism of substance P has been documented to occur by a number of enzymes, including: angiotensin-converting enzyme (ACE), 74 neutral endopeptidase (NEP), 74 -76 SP-degrading enzyme (SP-DE), 77 cathepsin-D, 78 cathepsin-E, 79 post-proline endopeptidase (PEP), 80 and dipeptidyl amiopeptide IV (DPIV). 81 All of the above enzymes have been identified to cleave substance P in vitro, but NEP and/or ACE are more likely to result in SP cleavage in vivo due to their cellular localization. 75
Neutral endopeptidase has been shown to be the predominant peptidase in tachykinin metabolism in the brain, as well as acting with a significant role within the spinal cord. 82,83 It preferentially cleaves hydrophobic amino acids, catalyzing the hydrolysis of the peptide bonds Gln 6 -Phe 7 , Phe 7 -Phe 8 and Gly 9 -Leu 10 within the amino acid structure of substance P. This results in fragments that lack the carboxyl terminal region that are necessary for its binding to the tachykinin receptor, therefore inactivating it. 84 NEP's role in the degradation of SP has been further explored through the use of NEP -/- (knockout) mice, where the NEP gene was deleted by homologous recombination. Widespread plasma extravasation in postcapillary endothelia was observed and reversed by recombinant NEP and neurokinin-1 receptor (SP) antagonists. 85
Angiotensin-converting enzyme has been shown to inactivate substance P within plasma in vivo. 86,87 ACE has been shown to catalyze the hydrolysis of the Phe 8 -Gly 9 bond, leaving the peptide lacking its carboxyl terminal region required for binding to the tachykinin receptor. 87 Following ACE-inhibitor administration in mice, significant increases in plasma extravasation of Evans blue dye in the airways, gut, and pancreas was noted. 88 Pre-treatment with a neurokinin-1 receptor antagonist prior to administration of the ACE-inhibitor resulted in markedly reduced plasma extravasation of the Evans blue dye, indicating the contribution of ACE to substance P's effect of plasma extravasation. 88
SP is rapidly degraded, with its half-life reported to be in the range of seconds to tens of minutes in blood and tissues. 59,74,89,90 It is able to interact with higher molecular weight binding proteins such as fibronectin due to structural similarities between regions of the NK-1R and the glycoprotein. The resulting complex allows SP to evade degradation and provides stability. 91 This suggests a shorter half-life in tissues, as unbound substance P within the tissues can be degraded by the cell surface metallopeptide neutral endopeptidase. This has been evidenced by stability noted of endogenous substance P compared with an added exogenous substance P, and the majority of SP within the plasma identified as being associated with high molecular weight material in a bound complex. 89
Localization
Substance P can be found widely distributed throughout the central, peripheral, and enteric nervous systems. Within the human nervous system, SP immunoreactivity has been demonstrated by immunohistochemistry 92,93 and radioimmunoassay 94 in the telencephalon, diencephalon, mesencephalon, metencephalon, myelencephalon, and spinal cord.
Within the telencephalon of the brain, the greatest immunoreactivity was found within the basal ganglia (striatum, putamen, and internal globus pallidus), hippocampus, and septum. Lower concentrations were identified in the cortex, external globus pallidus, and amygdala. The hypothalamus within the diencephalon showed immunoreactivity for SP, while none was detected in the thalamus. Within the mesencephalon, the substantia nigra (both pars reticular and compacta) revealed the highest immunoreactivity of SP within the CNS, and reduced levels within the hypothalamus. The pons and medulla oblongata of the metencephalon and myelencephalon show immunoreactivity to SP (Fig. 3). 92,93,95 Edvinsson and colleagues 96 investigated the distribution of substance P-like immunoreactivity in the cerebral arteries, choroid plexus, and dura mater. SP fibers were numerous in pial vessels, with few identified in the choroid plexus and dura mater (in man—compared with high levels found in cat, rabbit, and guinea pig). Those SP-containing nerve fibers that were found were closely associated with blood vessels. 96 Mai and colleagues 97 examined the distribution of SP immunoreactivity within the human brain using immunohistochemistry. They observed SP-immunoreactive “tubules” that were assumed to correspond with dendrites, and single-stranded or beaded fibers with were assumed to correspond to varicosities and synaptic boutons along dendrites. Grouped and diffuse beaded processes were observed in association with perikarya and proximal dendrites. 97

A representation of the distribution of substance P immunoreactivity within the different regions of the brain, from higher concentrations to lower concentrations. Created with
Within the spine, the substantia gelatinosa contains the highest levels of immunoreactive SP, with high levels continuing in extension to the nucleus proprius and into the dorsal roots. The spinal ganglia contain moderately high levels of SP immunoreactivity, with levels remaining similar across segments. SP immunoreactivity drops significantly in the ventral root, ventral horn, and other areas of gray matter, and is undetectable within the white matter. 94,98
On a cellular level, substance P is expressed by a plethora of cell types including neurons, 99 microglia, 100 astrocytes, 101 endothelial, 102 and epithelial 103 cells. Many cells of the immune system also express high levels of SP, including T cells, 104 dendritic cells, 105 macrophages, 106 and eosinophils. 107 They are also expressed by stem cells 108 and mesenchymal stem cells. 109 Sensory nerve fibers positive for SP are found surrounding most blood vessels throughout the body. Cerebral arteries have a rich supply of sensory neurons, with SP-immunoreactive nerve fibers in close association with blood vessels in the dura mater and choroid plexus. 96 With the widespread expression of substance P throughout various cell types, especially neurones, it is no surprise that substance P is implicated in a wide range of physiological and pathophysiological functions.
The neurokinin-1 receptor
Three tachykinin receptors (TACR; also referred to as neurokinin receptors) of differing types according to ligand affinity are described: the neurokinin-1 receptor (NK-1R), neurokinin-2 receptor (NK2R), and neurokinin-3 receptor (NK3R). Their preferential ligand of affinity is substance P, neurokinin A, and neurokinin B, respectively, though they are not exclusive in their ability to bind to their respective receptors—especially at high ligand concentration. 61,110 SP's receptor binding affinity to the NK-1R is 100- and 500-fold greater than NKA and NKB, respectively. 111 The neurokinin-1 receptor is encoded by the 5-exon TACR1 gene localized on the short arm of chromosome 2 (2p13.1-p12). 111 There is significant sequence homology between species, as cloned neurokinin 1 receptors from human, mouse, rat, and guinea pig reveal a difference in amino acid structure of less than 10%. 112
There are two isoforms of the neurokinin-1 receptor: a 407-amino acid residue full length version (NK-1R-F; long isoform), and a 311-amino acid residue truncated version (NK-1R-Tr; short isoform). The length of the carboxyl-terminus tail is the only difference between the two, with the short isoform boasting a relative molecular mass of 37 kDa and the long isoform a relative molecular mass of 46 kDa. It is the structural basis for the difference in functional properties between the two isoforms. 113 -115 The long isoform is significantly more abundant within select regions of the brain, while the short isoform appears more prevalently throughout the central nervous system and peripheral tissues. 116 The short isoform receptor responds to lower concentrations of SP and is resistant to desensitization. 114 However, it has been shown to have a binding affinity 10-fold less than the long isoform full length version of the NK-1R. 113
Within the brain, NK-1R mRNA expression was found to be high in the striatum, moderate in the cortex, hippocampus, and amygdala, and low in the thalamus and cerebellum. Peripheral tissues that expressed high levels of NK-1R mRNA expression included adipose tissue, the pituitary, the prostate, bone, spleen, skeletal muscle, intestine, heart, and lungs. 116 On a cellular level, the NK-1R is expressed by neurons, 117 endothelial cells, 118 epithelial cells, 119 fibroblasts, 120 smooth muscle cells, 121 and immune cells such as lymphocytes, 122 natural killer cells, 123 dendrites, 124 monocytes/macrophages, 106 microglia, 100,122,125 astrocytes, 125 eosinophils, and mast cells. 126
The tachykinin receptor family belong to family 1 (rhodopsin-like) of G protein-coupled receptors (GPCRs). These receptors have seven hydrophobic transmembrane domains (TM I-VII), with three extracellular loops (EL 1-3), three intracellular loops (IL 1-3), an intracellular carboxyl-terminus, and an extracellular amino-terminus. 63 The binding site of NK-1R's agonists involves the residues of the first and second extracellular loop (EL 1 and 2), as well as the second and seventh transmembrane domain (TM II and VII). 127 The third intracellular loop (IL 3) is responsible for binding to protein G, and the intracellular carboxyl-terminus dictates desensitization of the receptor following phosphorylation of its serine and threonine residues in response to repeated binding of its agonist. 61 SP enters into the hydrophobic ligand binding pocket (TM II and VII) between the extracellular surface, lipid bilayer and transmembrane domain. The remainder of the molecule interacts with the amino acids on the extracellular surface. 24
Following binding of substance P to the neurokinin-1 receptor, translocation of G-protein-coupled-receptor kinases (GRKs) and β-arrestins from the cell's cytosol to the plasma membrane occurs. GRKs phosphorylate the SP/NK-1R complex, while β-arrestins interact with the phosphorylated SP/NK-1R complex. 128,129 The SP/NK-1R-β-arrestin complex is internalized via the process of endocytosis, contributing to desensitization of the cell to SP-signaling. Upon internalization of the SP/NK-1R-β-arrestin complex, substance P is detached from the receptor and metabolized. The NK-1R is recycled to the cell surface, leading to resensitization of the cell. 130
The formation of the SP/NK-1R complex activates phospholipase C and adenylate cyclase. 131 Phospholipase C catalyzes the hydrolysis of phosphoinositides into inositol 1,4,5-trisphosphate (IP3) and DAG. 132 Adenylate cyclase activation results in the formation of cyclic adenosine monophosphate accumulation (cAMP). 133 These secondary messengers signal to a PKC-Raf-MEKs/PKA-MEKs pathway that activate extracellular signal-related kinases 1/2 (ERK1/2), which translocate into the cell's nucleus and mediates the expression of cytokines. 134 –137 This is achieved through the mammalian target of rapamycin (mTOR) signaling pathway, serine/threonine-specific protein kinases, and transcription factors such as nuclear factor kappa B (NF-κB) 134,138 and activator protein 1 (AP-1). 133 Activation of ERK was brisk (peak within 1-2 min) when the full length isoform NK-1R was bound, compared with a slower (peak within 20-30 min) activation at the truncated short isoform NK-1R. 115 Interestingly, cells that expressed the truncated isoform of the NK-1R did not activate the NF-κB pathway when NK-1R-Tr was activated, and they appear to have an alternate downstream signal pathway. 115
Functions
The widespread expression of substance P throughout the body and in a variety of differing cell types affirms its role in a diverse array of biological functions. These include neurotransmission, 139 plasma extravasation, 140 vasodilation, 140 nociception, 141 smooth muscle contraction and relaxation, 142,143 memory modulation and reinforcement, 144 emesis, 145 respiration, 146 neurogenesis, 147 stimulation of cell growth, 148 and inflammation. 61,149 In terms of pathophysiology, SP has been implicated in asthma, 150 eczema, 151 cancer, 152 HIV/AIDS, 153 rheumatoid arthritis, 154 inflammatory bowel disease, 155 sickle cell disease, 156 affective disorders, 157 chemotherapy-induced emesis, 158 cardiovascular disease, 159 and the focus of this review, traumatic brain injury via neurogenic inflammation.
The Blood–Brain Barrier
Physiological functions of the BBB
A barrier between the CNS and peripheral circulation is necessary to maintain a strict homeostatic environment required for normal brain function. This environment is formed by three physical interfaces: the blood–brain barrier, the blood–CSF barrier, and the arachnoid epithelium that separates the blood from the subarachnoid CSF. 160 The current focus will be on the BBB separating the blood circulation and the brain's extracellular fluid and how its dysfunction is involved and exacerbates the secondary injury process in TBI.
The BBB is comprised of a monolayer of specialized endothelial cells termed “brain microvascular endothelial cells,” joined together by tight junctions (TJs) and adherens junctions. 161 These tight junctions are comprised of numerous important proteins with differing functions, including claudins, occludins, junctional adhesion molecules and the zona occludens. 162 The adherens junctions contain protein cadherins, vascular endothelial cadherins (VE-cadherins) in high levels and lower levels of neural (N-) and epithelial (E-) cadherins. 163 Together, they result in limited paracellular diffusion of hydrophilic molecules, sealing adjacent endothelial cells and mediating their association with surrounding pericytes. 164 In addition, there is reduced transcytosis and restricted vesicle-mediated transcellular transport, necessitating additional and polarized transporters that vary between the luminal and abluminal sides of the endothelial cells. 165 This gives tight control over movement of molecules into and out of the brain. Transporters on the luminal side of the endothelial cell monolayer transport lipophilic molecules and specific nutrients (e.g., glucose, amino acids, nucleosides) into the CNS, while transporters on the abluminal side remove specified waste products. 166,167
Astrocytic end-feet nearly completely ensheathe the vascular tube, regulating water flow through colocalized aquaporin-4 channels and ionic control of extracellular K+ via Kir4.1 K+ channels. 164 Through intracellular Ca2+ levels they regulate neuronal activity and cerebral blood flow. In addition, they have been shown to communicate via gap junctions, and it has been proposed their regulation of vasodilation and vasoconstriction is transmitted through this method. 168 Finally, they are involved in strengthening the tight junctions between endothelial cells, modulating the polarized endothelial cell transporters, and promoting enzymatic systems. 169
Between the astrocytic end-feet and the endothelial cells sit pericytes embedded within the basal membrane. 166 They offer maintenance and stability to the endothelial cells via regulation of angiogenesis and the extracellular matrix, with BBB permeability seen to increase with a decrease in pericyte coverage. 164,165 In response to neuronal activity, they modulate cerebral blood flow and vessel diameter. 164
The BBB serves as a barrier isolating the brain's environment, restricting access to pathogens and xenobiotics that may cause harm. Unfortunately, this same defense creates challenging barriers for drug delivery to the CNS. 170
BBB dysfunction and vasogenic edema
Vasogenic edema is a result of increased BBB permeability, leading to an increase of fluid in the extracellular space that originates from the vasculature. An oncotic gradient forms as extravasated vascular components enter the extracellular space, and water follows this gradient to add to the brain water content. 171 This in turn results in increased total brain volume and a raised ICP. 172,173 Numerous contributing factors result in the increased BBB permeability seen after injury, many having been discussed above. These include mechanical disruption of the TJ proteins between endothelial cells, recruitment of local and peripheral inflammatory cells, release of proinflammatory cytokines and chemokines from glial cells, and upregulation of MMPs. 174 -176 Disruption of the TJs allows increased paracellular transport of molecules that were initially restricted, including inflammatory cells. In addition, increased transcytosis through endothelial cells of proteins such as albumin and other large molecules results in further osmotic shifts and water movement. 177
Blood–brain barrier disruption in TBI appears to be biphasic, initially occurring acutely after injury with increased permeability lasting for a few hours before declining to more normal levels. 178,179 A second disruption occurs again at around Day 3, although no additional brain edema was noted with this later opening on an Evans blue study performed on Sprague-Dawley rats. 178,180
Neurokinin-1 Receptor Antagonists
Animal models
Repeated capsaicin administration results in substance P depletion from primary sensory neurons. 181 In animals pre-treated with capsaicin that underwent TBI, no significant edema development was noted in comparison to controls. 182 This was associated with a reduction in BBB permeability assessed by Evans blue dye. Further, animals treated with capsaicin prior to injury suffered a reduction in post-traumatic functional deficit. 182 While this is unreasonable as a potential treatment, it offered insight into what antagonism of the NK-1R might achieve.
The NK-1 receptor antagonist N-acetyl-L-tryptophan (NAT) was administered to male Sprague-Dawley rats 30 min after injury using the acceleration-induced impact TBI model and compared with vehicle controls. 27 Donkin and colleagues found significantly reduced edema as assessed by brain water content and Evans blue dye extravasation. MRI showed evidence of an attenuation of edema using diffusion-weighted imaging. Additionally, NAT-treated rats showed significantly reduced functional deficits using the rotarod test and Barnes maze for motor function and cognitive outcome respectively. 27 The study was repeated in female Sprague-Dawley rats by Corrigan and colleagues to ensure results did not differ depending on gender, identifying a similar attenuation of BBB breakdown seen in males. 183 Donkin and colleagues repeated the experiment 2 years later, looking to assess what occurs when administration is delayed to 12 h. 179 They found that despite requiring administration of a BBB-permeable form of NAT (L-732,138) beyond 5 h, inhibition of the NK-1 receptor resulted in functional improvement compared with vehicle-controls. In addition, attenuation of axonal injury was found. 179 Finally, Li and colleagues examined the NK-1 receptor antagonist L-733,060 in TBI in adult male wild type C57BL/6J mice and compared with vehicle, sham, and SP knock out (SP-KO) mice. 184 Functional deficits, lesion volume, brain water content, and BBB breakdown were attenuated in SP-KO and NK-1R antagonist-treated mice. Additionally, SP-KO and NK-1R antagonist-treated mice showed a reduction in cytochrome c release (associated with apoptosis), reduced inflammatory cytokines and indicators of oxidative stress, as well as reactive oxygen species production. 184
NK-1R antagonism has also been explored within an ovine model. Sheep have a gyrencephalic brain akin to humans, unlike the rodent lissencephalic brain. 185 Gabrielian and colleagues examined the effect of NK-1 receptor antagonists on ICP in sheep following TBI. A humane stunner was used to induce impact-acceleration injury, and then an ICP monitor is inserted through a burr hole to measure ICP. An NK-1 receptor antagonist was administered 30 min after injury, while other sheep received vehicle or were sham controls. After 4 h, there was a significant increase in ICP (> 30 mm Hg) of sheep that received vehicle, while those that received a NK-1 receptor antagonist had an ICP comparable to the sham group. 186
Hyperphosphorylation of tau occurs following traumatic brain injury, with evidence that suggests it results in an increased risk of later developing dementia. 187 The postulated mechanism is an increase in kinase activity following substance P's binding to the neurokinin-1 receptor, notably kinases associated with tau phosphorylation. These include the serine/threonine-specific protein kinase Akt, extracellular signal-related kinases ERK1/2 and c-Jun N-terminal (JNK). In a study by Corrigan and colleagues, the neurokinin-1 receptor antagonist EU-C-001 was administrated to male Sprague-Dawley rats following sham procedures, single moderate-to-severe TBI, or repeated mild TBI. The NK-1RA EU-C-001 significantly attenuated tau phosphorylation following both single moderate-to-severe TBI and repeated mild TBI. 188 It is interesting to note that this was not previously seen following the administration of the NAT in a prior study, suggesting EU-C-001's ability to cross the intact blood–brain barrier to bind to the neuronal NK-1 receptor is necessary for an effect on tau phosphorylation. 179,188
Finally, an intriguing study exploring the use of angiotensin-converting enzyme (ACE) inhibitors in mice following TBI was performed by Harford-Wright and colleagues. 189 ACE is integral to the metabolism of substance P, and by inhibiting ACE, one would expect to see increased levels of substance P and potentiation of its effects as a result. The study team identified as expected increased immunoreactivity of SP compared with controls, and the ACE inhibitor–treated mice experienced a worsening of motor deficits. 189 While not a study on neurokinin-1 receptor antagonism, it offers further insights into the benefit that might be gained from blocking the action of substance P following TBI.
Previous human studies of substance P inhibition
The use of neurokinin-1 receptor antagonists has not yet been explored within humans in the context of traumatic brain injury. However, there are currently licensed uses within clinical practice, as well as much evidence in clinical trials investigating the use of NK-1R antagonists in a variety of disease processes.
Currently, post-operative and chemotherapy-induced nausea and vomiting (CINV) are the only licensed uses for neurokinin-1 receptor antagonists. Aprepitant was the first oral NK-1R antagonist approved for clinical use, when it was found to significantly improve chemotherapy-induced nausea and vomiting following cisplatin therapy in a clinical trial performed by Navari and colleagues in 1999. 190 Other NK-1R antagonists have since been developed, such as rolapitant, 191 netupitant, 192 fosaprepitant, 193 and fosnetupitant 194 in a combination of oral and IV preparations. Fosaprepitant and fosnetupitant are prodrugs of aprepitant and netupitant, respectively, where metabolism converts the drug into its pharmacologically active agent. 195 NK-1R antagonists are often given in combination anti-emetic therapy in addition to dexamethasone and serotonin type 3 receptor antagonists (5HT3RAs), as NK-1R antagonists have been thought to play a role in both acute and delayed CINV through antagonism of SP in the vomiting center of the brain. 196,197 An NK-1R antagonist was similarly investigated in the treatment of post-operative nausea and vomiting (PONV) in 1999 by Diemunsch and colleagues 198 and found to be efficacious, though the drug first approved for PONV would be aprepitant and not the investigated vofopitant.
There has been interest in the use of NK-1R antagonists within neuropsychiatric disease processes such as depression, 199 –204 generalized anxiety, 205 post-traumatic stress disorder, 206,207 migraines, 208,209 and pain. 210 –215 Unfortunately, in comparison to PONV and CINV, the evidence for their efficacy is less forthcoming. Kramer and colleagues 202 investigated MK-869 (aprepitant) in 1998 in patients with major depressive disorder. They observed a similar antidepressant profile compared with paroxetine, with improvements of insomnia and genital symptoms observed when using the NK-1R antagonist. Another clinical trial by Kramer and colleagues performed in 2003 observed antidepressant effects of the NK-1R antagonist L-759274 versus placebo based on clinical depression scales. 201 However, in a clinical trial by Keller and colleagues 199 comparing aprepitant, paroxetine, and placebo in major depressive disorder, the NK-1R antagonist showed no significant effect over placebo, despite positron emission tomography imaging indicating appropriate high level receptor blockade. Another study by Ball and colleagues examining aprepitant against paroxetine and placebo achieved the same no efficacy result. 200 The theory as to why results appeared inconclusive was investigated by Ratti and colleagues, 203 whose hypothesis was that antidepressant effects would only be observed if full central neurokinin-1 receptor blockade is achieved. This was investigated using positron emission tomography, where >99% receptor occupancy was achieved, and efficacy was noted in one of two studies (one not meeting statistical significance based on the Hamilton Depression Rating scale). Further discussion on the topic is well summarized in a review article by Rupniak and Kramer. 204
There has been investigation into NK-1R antagonists in the respiratory system, with studies investigating chronic cough, asthma, and chronic obstructive pulmonary disease (COPD). Smith and colleagues 216 observed significant improvement of chronic cough following orvepitant administration, as defined by objective cough frequency, cough severity visual analog scale score, and quality of life. The recovery of exercised-induced airway narrowing in asthma patients following administration of the NK-1R antagonist FK-888 was documented by Ichinose and colleagues. 217 A single inhaled dose of the NK-1R antagonist AVE5883 protected against neurokinin A-induced bronchoconstriction in asthma patients, although no effect was seen with 7 days of pre-treatment by Boot and colleagues. 218 The perception of breathlessness in patients with COPD was not observed with NK-1R antagonism. 219
Efficacy of NK-1R antagonists has also been explored within disorders causing pruritis. Clinical trials investigating their use in chronic refractory pruritis, 220 atopic dermatitis, 221,222 Epidermal growth factor receptor inhibitor (EGFRI)-induced pruiritis, 223 and epidermolysis bullosa 224 have been performed with mixed, generally unfavorable results.
Lastly, studies that explored treatment of overactive bladder, 225 tinnitus, 226 gastric motor function, 227 and the prevention of post-endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis 228 showed no significant effect with NK-1R antagonism.
Human studies in TBI
Studies exploring the effects of neurokinin-1 receptor antagonists in animal TBI models have shown promise. They have been observed to reduce post-TBI edema development, reduce ICP, attenuate BBB breakdown, as well improve functional outcomes. 27,179,182,184,186 The safety profiles of NK-1R antagonists have been shown to be generally well tolerated in both human and animal studies, without significant reported adverse effects. Assessment of these drugs in a clinical trial, performed in human TBI patients, would be the next logical step. One would hypothesize that neurokinin-1 receptor antagonism in human TBI patients, akin to animal models, would result in an attenuation of BBB breakdown and therefore a reduction in post-TBI cerebral edema, mitigating the resultant intracranial pressure rise and poor clinical outcomes that follow.
A multi-center phase II clinical trial of an investigational medicinal product (CTIMP) is planned, assessing the efficacy, tolerability, and safety of EU-C-001 (a novel NK-1R antagonist) in intensive care unit patients with moderate to severe traumatic brain injury. It includes both open-label and randomized double-blind placebo-controlled phases to address both pharmacokinetics and putative efficacy. Efficacy will be evaluated through assessment of ICP and Therapy Intensity Level (TIL) and other surrogate outcomes including computed tomography and magnetic resonance imaging, biomarker, and cytokine data, as well as functional outcome measures. Once safety is examined in a smaller sample size, a larger phase III randomized-control trial is planned to better assess drug efficacy.
Conclusion
Traumatic brain injury remains a considerable global health issue. Despite significant morbidity and mortality, no noteworthy pharmacological advances have been made towards its management since the introduction of hypertonic saline for ICP control in 1988. 229 The growing evidence for substance P's role in neuroinflammation and resultant development of cerebral edema following TBI provides a promising novel therapeutic target. Animal studies provide promising results when the action of substance P is blocked at its preferential receptor, the neurokinin-1 receptor.
Despite the current lack of evidence of NK-1R antagonist use in human TBI in the literature, the numerous clinical trials that have been performed in humans to date show NK-1R antagonists to be safe and well tolerated. No significant adverse events were noted in the previously described studies. Considering the apparent safety of NK-1R antagonists in human populations and the promising results seen in animal models of TBI, it would be prudent to investigate NK-1R antagonism in human TBI patients.
Transparency, Rigor, and Reproducibility Summary
This review paper is not comprised of original research and has appropriately researched sources using PubMed's search engine for MEDLINE. Sourced papers are appropriately cited in the references section.
Footnotes
Acknowledgments
Adel Helmy: Medical Research Council/Royal College of Surgeons of England Clinical Research Training Fellowship (Grant no. G0802251), the NIHR Biomedical Research Centre and the NIHR Brain Injury MedTech Co-operative.
Authors' Contributions
Adam Safwat: investigation, writing—original draft, visualization.
Adel Helmy: conceptualization (equal), supervision (equal), writing—review & editing (equal), funding acquisition (equal).
Arun Gupta: conceptualization (equal), supervision (equal), writing—review & editing (equal), funding acquisition (equal).
Funding Information
No additional funding to declare.
Author Disclosure Statement
The authors would like to disclose that PresSura Neuro (Melbourne, Australia) are paying to cover staff and study costs for the PANGEA drug study.
