Scientists have discovered a new treatment to dramatically reduce swelling after brain and spinal cord injuries, offering hope to 75 million victims worldwide each year.
The breakthrough in treating such injuries – referred to as central nervous system (CNS) edema – is thought to be hugely significant because current options are limited to putting patients in an induced coma or performing risky surgery.
Brain and spinal cord injuries affect all age groups. Older people are more at risk of sustaining them from strokes or falls, while for younger age groups, major causes include road traffic accidents and injuries from sports such as rugby, US-style football and other contact games.
The high-profile example of Formula 1 racing driver Michael Schumacher demonstrates the difficulties physicians currently face in treating such injuries. After falling and hitting his head on a rock while skiing in Switzerland in 2013, Schumacher developed a swelling on his brain from water rushing into the affected cells.
He spent six months in a medically-induced coma and underwent complex surgery, but his rehabilitation continues to this day.
The new treatment, developed by an international team of scientists working at Aston University (UK), Harvard Medical School (US), University of Birmingham (UK), University of Calgary (Canada), Lund University (Sweden), Copenhagen University (Denmark) and University of Wolverhampton (UK), features in the latest edition of the scientific journal Cell.
The researchers used an already-licensed anti-psychotic medicine – trifluoperazine (TFP) – to alter the behaviour of tiny water channel ‘pores’ in cells known as aquaporins.
Testing the treatment on injured rats, they found those animals given a single dose of the drug at the trauma site recovered full movement and sensitivity in as little as two weeks, compared to an untreated group that continued to show motor and sensory impairment beyond six weeks after the injury.
The treatment works by counteracting the cells’ normal reaction to a loss of oxygen in the CNS – the brain and spinal cord-caused by trauma. Under such conditions, cells quickly become ‘saltier’ because of a build-up of ions, causing a rush of water through the aquaporins which makes the cells swell and exerts pressure on the skull and spine.
This build-up of pressure damages fragile brain and spinal cord tissues, disrupting the flow of electrical signals from the brain to the body and vice versa.
But the scientists discovered that TFP can stop this from happening. Focusing their efforts on important star-shaped brain and spinal cord cells called astrocytes, they found TFP prevents a protein called calmodulin from binding with the aquaporins.
Normally, this binding effect sends the aquaporins shooting to the surface of the cell, letting in more water. By halting this action, the permeability of the cells is reduced.
Traditionally, TFP has been used to treat patients with schizophrenia and other mental health conditions. Its long-term use is associated with adverse side effects, but the researchers said their experiments suggested that just a single dose could have a significant long-lasting impact for CNS edema patients.
Since TFP is already licensed for use in humans by the US Federal Drug Administration (FDA) and UK National Institute for Health and Care Excellence (NICE) it could be rapidly deployed as a treatment for brain injuries.
But the researchers stressed that further work would allow them to develop new, even better medicines based on their understanding of TFP’s properties.
According to the World Health Organisation (WHO), each year around 60 million people sustain a traumatic brain or spinal cord injury and a further 15 million people suffer a stroke. These injuries can be fatal or lead to long-term disability, psychiatric disorders, substance abuse or self-harm.
Professor Roslyn Bill of the Biosciences Research Group at Aston University said:
“Every year, millions of people of all ages suffer brain and spinal injuries, whether from falls, accidents, road traffic collisions, sports injuries or stroke. To date, their treatment options have been very limited and, in many cases, very risky.
“This discovery, based on a new understanding of how our cells work at the molecular level, gives injury victims and their doctors hope. By using a drug already licensed for human use, we have shown how it is possible to stop the swelling and pressure build-up in the CNS that is responsible for long-term harm.
“While further research will help us to refine our understanding, the exciting thing is that doctors could soon have an effective, non-invasive way of helping brain and spinal cord injury patients at their disposal.”
Dr. Zubair Ahmed of the University of Birmingham’s Institute of Inflammation and Ageing said:
“This is a significant advance from current therapies, which only treat the symptoms of brain and spinal injuries but do nothing to prevent the neurological deficits that usually occur as a result of swelling. The re-purposed drug offers a real solution to these patients and can be fast-tracked to the clinic.”
Dr. Alex Conner of the University of Birmingham’s Institute of Clinical Sciences said:
“It is amazing that our work studying tiny water channels in the brain can tell us something about traumatic brain swelling that affects millions of people every year.”
Dr. Mootaz Salman, Research Fellow in Cell Biology at Harvard Medical School, said:
“This novel treatment offers new hope for patients with CNS injuries and has huge therapeutic potential. Our findings suggest it could be ready for clinical application at a low cost in the very near future”.
CNS edema is caused by traumatic injuries, infection, tumor growth, and stroke (Jha et al., 2019; Liang et al., 2007). Traumatic injuries are a leading cause of psychiatric disorders, substance abuse, attempted suicide, disability, and early death in adults un- der 45 years of age (Fazel et al., 2014).
The biggest increase in patient numbers is currently in those older than 60 years (so- called ‘‘silver trauma’’). In the United States, this silent epidemic affects more than 1.7 million individuals annually.
Depression, suicidal behavior, and an increased risk of neurodegenerative conditions such as Alzheimer’s disease and Parkinson’s disease are known outcomes of traumatic CNS injury in patients of any age (Chen et al., 2014; Zlokovic, 2011).
Although industry has pursued the development of specific drugs that halt CNS edema progression, all have failed in phase III clinical trials; two recent trials showed that progesterone had no effect on neurological outcome following TBI (Stein, 2015). Notably, few strategies have focused on the primary cause of CNS edema, which is dysregulated flow of water into cells.
Current treatment approaches are therefore limited by an absence of pharmacological interventions and a reliance on alleviating the symptoms of edema and not the causes, using therapies introduced more than 80 years ago (Manley et al., 2000).
AQPs play an essential role in promoting short-term suscepti- bility to the pathological changes in volume that enhance CNS edema formation; consequently, they are established as drug targets (Verkman et al., 2014). All previous strategies to identify AQP inhibitors have focused on blocking the pore of the AQP channel.
Based on this approach, pharmacological inhibition of AQP4 by AER-270 has been suggested to cause a reduction in CNS edema, and AER-271, a pro-drug of AER-270, is the subject of phase I safety trial NCT03804476 in healthy volunteers. How- ever, although 70% maximal inhibition of rat AQP4 and 20% maximal inhibition of mouse and human AQP4 have been re- ported in water transport assays, AER-270 had the same effect on water content in rat and mouse stroke models, suggesting that the effect is not AQP4 dependent (Farr et al., 2019).
Because AER-270 is a known nuclear factor kB (NF-kB) inhibitor, and NF-kB inhibition can reduce CNS water content (Li et al., 2016), it may instead be acting through this modality.
We show a direct mechanistic relationship between inhibition of AQP4 function and a reduction in CNS edema.
We present an in vivo demonstration that targeting the subcellular localization of a membrane channel protein, rather than targeting its activity directly, is a viable therapeutic strategy. Our focus on targeting a fundamental cellular process, rather than trying to block a pore, provides a broadly applicable framework for future drug development. Regulation by vesicular trafficking is a common biological mechanism controlling the function of many membrane protein families (Offringa and Huang, 2013).
It is well- characterized for AQP2 in the renal collecting duct in response to the antidiuretic hormone vasopressin (van Balkom et al., 2002), but current dogma fails to recognize the central role of translocation as a regulatory mechanism for the AQP family as a whole (especially in response to non-hormonal, physiological triggers) and its implications for cell and tissue homeostasis. Here we present pathophysiologically relevant AQP subcellular relocalization and establish it as a regulatory mechanism.
Using a rat model of CNS edema, we show that CaMi or PKAi effectively limits spinal cord water influx 3 dpi and completely abol- ishes spinal cord edema by 7 dpi (Figure 2; Figure S2 shows the same effect after brain edema). Total AQP4 expression and localization at the BSCB increased following DC crush injury, and both were blocked by inhibition of CaM or PKA (Figure 2).
We also demonstrate the central roles of CaM and PKA in increased cell-surface expression of AQP4 and AQP4 subcellular relocalization to the BSCB following injury (Figures 1 and 2). Ablation of edema and AQP4 BSCB relocalization in vivo are accompanied by complete functional recovery by 2 weeks post-injury (Figure 4).
Although CaM has multiple roles in our proposed mechanism, we demonstrate a direct interaction be- tween CaM and AQP4 that is necessary for AQP4 subcellular re- localization (Figure 3).
The anti-psychotic effects of TFP are attributed to its anti- dopaminergic and anti-adrenergic actions (Qin et al., 2009). Dopamine decreases AQP4 expression in cultured astrocytes (Ku¨ ppers et al., 2008); if the anti-dopaminergic effects of TFP were dominant in our experiments, then we would expect increased AQP4 expression following TFP treatment and wors- ening of edema, as seen when the selective D2 antagonist L-741,626 increased spinal cord water content above the DC+vehicle control 3 dpi (Figure 2B).
The fact that we observed the opposite with TFP suggests that the CaMi effects of TFP dominate any anti-dopaminergic effects in CNS edema. This in- crease in edema with an anti-dopaminergic inhibitor along with the increased edema following PKCi is particularly interesting, given that, in the cultured pig kidney cell line LLC-PK1, dopamine signaling via PKC reduced AQP4-dependent plasma membrane water permeability (Zelenina et al., 2002); our data suggest that this pathway may also be active in astrocytes. Anti-adrenergics are associated with delayed onset of edema after intracerebral hemorrhage in humans but have no effect on patient outcome (Sansing et al., 2011).
We see prevention of edema following TFP treatment (rather than just delayed onset) and a strong effect on outcome, whereas the a1 antagonist terazosin had no effect on spinal cord water content following DC crush injury. This sug- gests that the primary modality of TFP in our model is its CaM antagonism (Tanokura and Yamada, 1986).
Previously published data using a rat model of stroke demonstrate that treatment with TFP prevents onset of brain edema, which has been proposed to be via CaMi stabilizing the integrity of the BBB. We suggest that the beneficial effects are due to a reduction in AQP4 peri-endothelial localization (Figure 2), which was not measured in that study (Sato et al., 2003).
Our data show that CaM has at least two distinct roles in translocation of AQP4 in astrocytes. First, activation of CaM following opening of TRPV4 (which has been suggested to interact with AQP4; Benfenati et al., 2011; Jo et al., 2015) leads to activation of an adenylyl cyclase, production of cAMP, and activation of PKA (Figure 1).
Second, CaM binds directly to AQP4 (Figure 3), and the strength of this binding is modulated by phosphorylation of AQP4 at a single PKA consensus site, S276. We have shown previously (Kitchen et al., 2015) that hypotonicity-mediated AQP4 relocalization is blocked by several PKA inhibitors (hypericin, H-89, and myr-PKI, in increasing or- der of specificity), that a non-phosphorylatable AQP4 mutant (S276A) does not relocalize, and that a phospho-mimetic mutation (S276D) removes the PKA dependence of the relocal- ization.
Furthermore, multiple phosphoproteomics datasets (retrieved via dbPAF; Ullah et al., 2016; available at http:// dbpaf.biocuckoo.org) demonstrate that AQP4-S276 is phos- phorylated in human, rat, and mouse tissue samples.
Together, these data suggest a model of AQP4 translocation whereby an influx of calcium ions activates CaM; this activates PKA via a CaM-activated adenylyl cyclase (e.g., AC1, AC3, or AC8; all three are expressed in rat and mouse brain; Sanabra and Mengod, 2011). PKA phosphorylates AQP4, and CaM binds to AQP4 to facilitate its translocation to the plasma membrane (Figure 5).
Our NMR data suggest that this binding causes a conformational change in the AQP4 carboxyl terminus, which becomes more structured, with increased a-helical content (Figure 3C). An increased affinity toward the AQP4-S276E phospho-mimetic mutant suggests that CaM preferentially binds phosphorylated AQP4, with a 2-fold decrease in Kd.
Although this increase in affinity is modest, it is possible that the additional charge found on an actual phosphoserine residue compared with the mimetic glutamate residue leads to a more enhanced difference in vivo.
CaM and PKA also appear to be involved in the post-injury increase in AQP4 protein expression. A previous study using a mouse TBI model found that direct activation of the Aqp4 gene by the transcription factor Foxo3a is responsible for increased AQP4 expression (Kapoor et al., 2013); our data suggest that this response is conserved in rats (Figures 2 and S3).
We demonstrate that astrocyte swelling initiates Foxo3a nuclear translocation via PKA-mediated inactivation of Akt. The Akt agonist SC79 has been shown recently to have a neuroprotective effect in a rat middle cerebral artery occlusion (MCAO) model of stroke (Luan et al., 2018).
This protective effect was attributed to inhibition of apoptosis; our data suggest that an alternative interpretation of these findings is inhibition of Aqp4 upregulation in astrocytes following MCAO. Previous studies showing increases in AQP4 surface localization used a model of astrocytic cell swelling based on hypotonic treatment of astrocytes (Kitchen et al., 2015; Salman et al., 2017a).
Although this creates a tonicity gradient similar to that seen during cytotoxic edema formation after stroke or traumatic injury, physiologically, astrocytes actually experience an intracellular increase in tonicity rela- tive to the extracellular fluid caused by hypoxia-driven effects on ion channels and transporters (Lafrenaye and Simard, 2019).
Our data in Figures 1A–1D provide evidence of acute hypoxia-mediated AQP4 translocation and are the basis of a simple and easy- to-implement in vitro model to study CNS injury and edema, which reproduces the AQP4 translocation response observed in vivo.
Following injury, increased AQP4 immunoreactivity (Fig- ure 2H) might be influenced by astrogliosis. Hypertrophy and migration of reactive astrocytes (which have increased AQP4 expression; Vizuete et al., 1999) into the injury epicenter aid tissue repair and cause glial scarring.
Astrocyte migration depends on AQP4 at the leading edge (Saadoun et al., 2005); inhibiting AQP4 relocalization may reduce the number of infil- trating reactive astrocytes. An additional benefit of inhibiting
AQP4 relocalization may therefore be a reduction in the number of invading, reactive astrocytes and in the extent of glial scarring. This may facilitate axon sprouting and could account for the improved electrophysiological outcomes reported in our study.
TFP is licensed as a drug for human use (NICE, 2019) that we administered in rats at a dose approximately equivalent to its licensed dose in humans.
This treatment resulted in functional recovery 2 weeks after DC crush injury; animals treated with TFP could walk normally after 2 weeks, whereas untreated ani- mals had still not recovered 6 weeks post-injury.
The future socio-economic impact of this work is enormous; our data pro- vide a molecular mechanistic understanding of water channel regulation (Figure 5) that has the potential to define a therapeutic framework for the tens of millions of CNS edema patients annu- ally, worldwide, for whom there is still no pharmacological intervention.
More information:Cell (2020). DOI: 10.1016/j.cell.2020.03.037