Effects of TRPM2 Inhibition in Neuroprotection following Neonatal Hypoxic-Ischemic Brain Injury

Abstract

Neonatal hypoxic-ischemic (HI) brain injury is a major cause of acute mortality and chronic neurological morbidity in infants and children. Studies indicate that transient receptor potential melastatin 2 or TRPM2 (a non-selective cation channel with high permeability to calcium that can be activated by intracellular adenosine diphosphoribose [ADPR] and H₂O₂) can mediate neuronal death following acute ischemic insults in adult mice as well as HI brain injury in neonatal mice. My study tested the effect of a newly described TRPM2 channel inhibitor AG490 by using a H₂O₂-induced neuronal cell death in vitro model and mouse HI brain injury in vivo model. I found that the inhibition of TRPM2 channels by AG490 demonstrates a neuroprotective effect both in vitro and in vivo. The neuroprotective effect of AG490 following post-injury treatment suggests the potential clinical implications of this drug, including the possible prevention of HI related neurological complications such as hypoxic-ischemic encephalopathy.

Table of Contents

Abstract

Acknowledgements

Table of Contents

List of Abbreviations

List of Figures

Chapter 1  Introduction

1 Neonatal Hypoxic-Ischemic Brain Injury

1.1 Incidence and global impact

1.2 Standard diagnostic criteria for neonatal HI

1.3 Current therapy

2 Failure in Targeting Glutamate Receptors as Pharmacological Targets

3 Targeting Non-Glutamate Channels

3.1 Sodium-calcium exchangers

3.2 Hemichannels

3.3 Volume-regulated anion channels

3.4 Acid-sensing ion channels (ASICs)

3.5 Transient receptor potential melastatin (TRPM) subfamily

4 Transient Receptor Potential Channels (TRPs)

5 TRPM2 Channel

5.1 TRPM2 protein structure, transmembrane topology and distribution

5.2 TRPM2 biophysical properties and gating mechanism

5.3 The physiological and pathophysiological role of TRPM2 channels

6 Pharmacological Interactions

6.1 Flufenamic acid (FFA)

6.2 Anti-fungal agents (clotrimazole and econazole)

6.3 2-APB

6.4 Divalent heavy metal cations

6.5 AG490

Chapter 2  Rationale and Hypothesis

Rationale

Hypothesis

Chapter 3  Aims and Experimental Design

Aims:

Experimental Design Outline:

Chapter 4  Materials and Methods

1 Ethics Approval

2 Animals

3 Reagents

4 Cell Culture

5 In vitro H2O2-induced Neuronal Cell Death Model

6 Cell Viability Assay

7 Electrophysiology (Whole Cell Patch Clamp)

8 Drug Administration

9 In vivo Hypoxic-Ischemic Mouse Model

10 Infarct Volume Measurement, Whole Brain Imaging and Histological Assessments

10.1 TTC staining/Infarct volume measurement

10.2 Whole brain imaging/Nissle staining

11 Neurobehavioral Assessments

11.1 Geotaxis reflex

11.2 Cliff avoidance test

11.3 Grip test

11.4 Passive avoidance test

12 Immunohistochemistry and Confocal Imaging

13 Western Blot

14 Statistics and Data Analysis

Chapter 5  Results

1. The level of TRPM2 mRNA expression in the hippocampus and cortex increases consistently with the developmental age

2 AG490 as a pharmacological inhibitor of the TRPM2 channel

3 AG490 protects neurons from H2O2-induced cell injury in vitro

4 The Effect of AG490 Pre-treatment on Hypoxic-Ischemic Brain Injury in vivo.

4.1  Pre-treatment with TRPM2 inhibitor AG490 reduced the brain infarct volume of hypoxic-ischemic brain injury in vivo.

4.2 Pre-treatment with TRPM2 inhibitor AG490 reduced brain damage following hypoxic-ischemic brain injury.

4.3 Pre-treatment with TRPM2 inhibitor AG490 promotes recovery after HI challenge

4.4 Pre-treatment with TRPM2 inhibitor AG490 improves short-term neurobehavioral performance after HI

4.5 Pre-treatment with TRPM2 inhibitor AG490 also improves long-term neurobehavioral performance after HI

4.6   Pre-treatment with TRPM2 inhibitor AG490 reduces reactive astrocyte activation

4.7 Pre-treatment with TRPM2 inhibitor AG490 may reduce HI brain damage through Akt mediated signaling pathways

5 The Effect of AG490 Post-treatment on Hypoxic-Ischemic Brain Injury in vivo.

5.1  AG490 *post-treatment 1 (30mg/kg, i.p.) reduced brain infarct volume of hypoxic-ischemic brain injury in vivo.

5.2 AG490 *post-treatment 1 (30mg/kg, i.p.) reduced brain damage following hypoxic-ischemic brain injury

5.3 AG490 *post-treatment 1 (30mg/kg, i.p.) improves neurobehavioral performance and general recovery after HI

6 AG490 *post-treatment 2 (30mg/kg, i.p., immediately after HI induction) demonstrate a trend towards neuroprotection following HI brain injury

Discussion

1. Connection between clinics and the current study

2. Summary of major findings

3. Significance of the current study

4. Differences between neonatal HI brain injury and adult stroke

5. Proposed mechanism of neonatal HI brain injury

6. Pitfalls in the current study and proposed future directions

REFERENCES

List of Abbreviations

ADPR	Adenosine disphosphate ribose
Akt	Protein kinase B
AMP	Adenosine monophosphate
AMPA	α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor
ASICs	Acid-sensing ion channels
ATP	Adenosine triphosphate
BDNF	Brain-derived neurotrophic factor
cAMP	Cyclic adenosine monophosphate
cADPR	Cyclic adenosine diphosphate ribose
Ca2+	Calcium ion
CaM	Calmodulin
Cl–	Chloride ion
CNS	Central nervous system
CTL	Control
Cx43	Connexin 43
DMSO	Dimethyl sulfoxide
EEG	Electroencephalography
E16	Embryonic day 16
FFA	Flufenamic acid
GAPDH	Glyceraldehyde 3-phosphata
GFAP	Glial fibrillary acidic protein
GPCRs	G-protein-coupled receptors
GSK-3α	Glycogen synthase kinase 3 alpha
GSK-3β	Glycogen synthase kinase 3 beta
HEK293	Human embryonic kidney 293
HI	Hypoxic-Ischemic
HIE	Hypoxic-Ischemic Encephalopathy
H2O2	Hydrogen Peroxide
ICC	Immunocytochemistry
IHC	Immunohistochemistry
i.p.	Intraperitoneal
IP3	Inositol 1,4,5-trisphosphate
i.v.	Intravenous
IV	Current-voltage
JAK2	Janus kinase 2
KATP	ATP-sensitive potassium channel
K+	Potassium ion
KO	Knockout
MAPK	Mitogen-activated protein kinases
MCAO	Middle cerebral artery occlusion
MTT	(3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide)
MRI	Magnetic resonance imaging
Na+	Sodium ion
NAD	Nicotinamide adenine dinucleotide
NCX	Na(+)/Ca(2+) exchanger/Sodium-calcium exchanger
NMDA	N-methyl-D-aspartate receptor
NSAIDs	Non-steroidal anti-inflammatory drugs
NUDT9-H	Nudix (Nucleoside Diphosphate Linked Moiety X)-Type Motif 9
OGD	Oxygen glucose deprivation
PARG	Poly (ADP-ribose) glycohydrolase
PARP	Poly (ADP-ribose) polymerase
PCR	Polymerase chain reaction
pH	Potential of Hydrogen
PI3K	Phosphoinositide 3-kinase
PIP2	Phosphatidylinositol 4,5-bisphosphate
P7/P14	Postnatal day 7/14
RCT	Randomized controlled trials
ROS	Reactive oxygen species
siRNA	Small interfering RNA
S5/S6	Segment 5/ Segment 6
TRPs	Transient receptor potential ion channel superfamily
TRPA	Transient receptor potential ankyrin
TRPC	Transient receptor potential canonical
TRPM	Transient receptor potential melastatin
TRPV	Transient receptor potential vanilloid
TRPP	Transient receptor potential polycystin
TRPML	Transient receptor potential mucolipin
TRPM2	Transient receptor potential melastatin 2
TTC	Tetrazolium chloride
VRACs	Volume-regulated anion channels
WT	Wildtype

 

List of Figures

Figure 1. Classical glutamate receptors (NMDA and AMPA receptor) model of neuronal cell death ……………………………………………………………………………………5

Figure 2. Potential mechanisms involved in excitotoxicity following ischemic stress…….9

Figure 3. Family tree of TRP channel superfamily………………………………………11

Figure 4. TRPM2 protein structure and variants………………………………………….13

Figure 5. Proposed mechanisms of TRPM2 channel activation by H₂O₂ and involvement TRPM2 channel activity in physiological and pathophysiological processes……………….16

Figure 6. TRPM2 channel inhibitors………………………………………………………………………19

Figure 7. An outline of the project experimental design……………………………………………21

Figure 8. TRPM2 expression in the cortex and hippocampus increases with development………………………………………………………………………………………………………..31

Figure 9. AG490 efficiently inhibited H₂O₂-induced TRPM2 current on TRPM2 overexpression HEK293 cells………………………………………………………………………………..33

Figure 10. TRPM2 inhibitor AG490 pre-treatment reduced neuronal cell death following H₂O₂-induced cell injury……………………………………………………………………………………….35

Figure 11. Timeline of neonatal hypoxic-ischemic injury and experimental procedures…………………………………………………………………………………………………………..36

Figure 12. Lower dose of AG490 (15 mg/kg) did not show effect on neuroprotection……………………………………………………………………………………………………38

Figure 13. Pretreatment of TRPM2 currents inhibitor AG490 (30 mg/kg) reduced brain infarct volume of hypoxic-ischemic brain injury in vivo……………………………………………39

Figure 14. TRPM2 inhibitor AG490 reduced brain damage following hypoxic-ischemic brain injury………………………………………………………………………………………………………….41

Figure 15. AG490 pre-treatment (20 min before HI injury) improves general health recovery after HI challenge…………………………………………………………………………………..42

Figure 16. AG490 pre-treatment (20 min before HI injury) improves neurobehavioral performance after HI challenge……………………………………………………………………………..44

Figure 17. Long – term behavioral assessment of functional recovery in the pre-treatment paradigm following hypoxic-ischemic injury…………………………………………………………..46

Figure 18. AG490 pre-treatment restores neuronal cell numbers and reduces reactive astrocyte activation………………………………………………………………………………………………53

Figure 19. Biochemical assessment of signaling pathways affected by hypoxic-ischemic insult on neonatal brain in a pre-treatment paradigm………………………………………………..54

Figure 20. *Post-treatment 1 (30mg/kg, i.p.) reduced brain infarct volume of hypoxic-ischemic brain injury in vivo………………………………………………………………………………….56

Figure 21. TRPM2 inhibitor AG490 reduced brain damage following hypoxic-ischemic brain injury………………………………………………………………………………………………………….57

Figure 22. AG490 post-treatment 1 (immediately after ischemic injury) improves general health and neurobehavioral performance after HI challenge………………………………………59

Figure 23. *Post-treatment 2 (30mg/kg, i.p.) shows a trend of neuroprotective effect following hypoxic-ischemic brain injury…………………………………………………………………60

Figure 24. Propose mechanism for inhibitory effect of AG490………………………………….65

Chapter 1  Introduction

1 Neonatal Hypoxic-Ischemic Brain Injury

When a term neonate’s (defined as 36 gestational weeks or later^1-3) brain does not receive sufficient oxygen (hypoxia) and blood (ischemia), neonatal hypoxic-ischemic encephalopathy may result (or HIE for short⁴).  Neonatal encephalopathy (HIE) can result from diverse conditions, with hypoxic-ischemic brain injury (HI brain injury) being the most common. Therefore, the terms HI brain injury and HIE are usually used synonymously. HI brain injury is a condition which can cause significant mortality and long-term morbidity. The injury can be a clinical consequence of perinatal, birth and/or neonatal asphyxia⁴.

1.1 Incidence and global impact

Clinically, most HI brain injury cases are due to birth asphyxia which causes 840,000 or 23 % of all neonatal deaths worldwide^{5, 6}. The severity and length of oxygen and blood deprivation affects whether HI brain injury occurs and how severe it is. Due to different levels of severity, patients suffer from symptoms ranging from transient behavioral abnormalities, occasional periods of apnea and seizure-related symptoms to even death from cardiorespiratory failure⁵.  50-80 % of the survivors will suffer from severe developmental or cognitive delays, motor impairments and learning disabilities^{5, 7}. Symptoms may worsen as the child continues to develop. The lifetime costs to the health care system have been estimated to be as high as $1.5 million per person in Canada⁵.

1.2 Standard diagnostic criteria for neonatal HI

Based on biomedical markers that correlate to clinical outcomes, there is an established set of predictors for neonatal HI brain injury^{1, 7, 8}:

1. Apgar Score

At birth, doctors and nurses carefully examine the newborn’s condition and give a number rating from 0 to 10. This number is called an Apgar score. The Apgar rates skin color, heart rate, muscle tone, reflexes and breathing effort. An Apgar score of less than or equal to 5 (at 5 and 10 minutes after birth) clearly confers an increased relative risk for HI brain injury^9-11.

2. Fetal Umbilical Artery pH

Fetal Umbilical Artery pHless than 7.0 and/or a base deficit worse than or equal to minus 12 mmol/L from cord/arterial/venous/capillary blood gas obtained within 60 minutes of birth increase the probability of neonatal HI brain injury^{2, 10, 12}.

3. MRI

Magnetic resonance imaging (MRI) is the neuroimaging modality that best defines the nature and extent of neonatal HI brain injury. Distinct patterns of MRI aberrations are recognized in term neonates due to HI brain injury. Recently, electroencephalography (EEG) has also shown helpful information for predicting the clinical outcome of HI brain injury, though some studies reported that EEG is not as reliable as MRI¹³.

4. Presence of Multisystem Organ Failure and Abnormal Neurological Signs

Dysfunction of multiple organs also results in a higher risk for neonatal HI brain injury⁹. The dysfunctional organ systems include hematologic abnormalities, cardiac dysfunction, metabolic derangements and gastrointestinal injury, or a combination of any of above.  Abnormal neurological signs are supplementary predictors of HI brain injury. For example, hypotonic muscles or lack of a sucking reflex may also indicate a probability of HI brain injury⁸.

1.3 Current therapy

There is currently only one available licensed treatment for HI: hypothermia^14-16.

Hypothermia should be offered to term infants with moderate or severe HI brain injury within 6 hours of birth. 6 large randomized controlled trials (RCT) of hypothermia for neonatal HI brain injury have been published, with all involved neonates being ≥36 weeks of gestation (termed infants). The target temperature was 33° to 34 °C, with the cooling duration being 72 hrs^{14, 15, 17}. Rewarming was processed slowly, with an increase of 0.5 °C per hour^{14, 15}. Neonates underwent a set of neurological examinations during the process as well as at the end of cooling. Some clinical trials have demonstrated that either head cooling or whole body cooling reduces mortality or disability either in all the infants or within a certain group of infants between 18 to 24 months of age^{14-16, 18}.

Even though hypothermia has shown impressive clinical outcomes and is currently well established as a standard treatment for neonates suffering from moderate to severe HI brain injury, it is stated to be “partially effective”^{19, 20}. Therefore, there is an urgent need for novel therapeutic opportunities beyond this current form of treatment.

2 Failure in Targeting Glutamate Receptors as Pharmacological Targets

Although the exact pathophysiology of HI brain injury is not completely understood, it is commonly accepted that a lack of sufficient blood flow in conjunction with decreased blood oxygen content leads to loss of normal cerebral autoregulation and diffuse brain injury^{21, 22}. Glutamate is an important excitatory neurotransmitter at excitatory synapses in the CNS^22-24. Based on the theory of excitotoxicity, it has long been accepted that a lack of blood flow can lead to high concentrations of glutamate release and eventually HI brain injury²⁵. Therefore, glutamate receptors have been extensively investigated as potential therapeutic targets for neuroprotection. The biochemical cascade of the theory of excitotoxicity is summarized as follows²²: Lack of cerebral blood flow triggers energy failure and neuronal depolarization which then releases large amounts of glutamate into the extracellular space. Excessive glutamate in the extracellular space overactivates NMDA (N-methyl-d-aspartic acid) and AMPA (dl-α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid) glutamate receptors, leading to an increased influx of calcium. Subsequently, intracellular calcium overload can be observed within neurons., which is thought to induce neuronal cell death. Popular strategies against excitotoxicity have targeted the pharmacological blockage of NMDA receptors. Whereas several compounds including dezocilipine maleate (MK-801), aptiganel hydrochloride (Cerestat), dexthrometorphan (DMX) and CGS 19755 (Selfotel) demonstrated promising neuroprotective effects in rodent models^{21, 26}, all clinical trials aimed at using NMDA and AMPA glutamate receptors as pharmacological targets failed to yield the expected protective outcomes^{22, 27}.

One potential explanation for this failure is that the injured brain may attempt to recruit endogenous recovery mechanisms^{22, 28}. While glutamate signals may be truly neurotoxic to neurons, this may suggest that some aspects of the signaling process may still have beneficial effects. It is already known that synaptic NMDA receptors and extrasynaptic NMDA receptors may have opposite effects following activation^{21, 29}. The activation of synaptic receptors could promote cell survival while the activation of extrasynaptic receptors can downregulate BDNF and ultimately lead to cell death^{22, 30}. Another potential limitation of this standard NMDA-AMPA model is that it does not consider the roles of cells other than neurons, for example, astrocytes and oligodendrocytes^{23, 31, 32}. As astrocytes and oligodendrocytes also express NMDA and AMPA receptors^{23, 33}, they too are vulnerable to excessive glutamate and play important roles in glutamate regulation. Additionally, glutamate is an essential neurotransmitter that is necessary for important physiological processes. Therefore, the blockage of glutamate receptors during the treatment of ischemic injury may lead to unwanted side effects. Studies have shown that the hypofunction of NMDA receptors may be partially responsible for the memory loss associated with aging³⁴.  Schizophrenia has also been reported in association with NMDA receptor dysfunction³⁵.

Figure 1. Classical glutamate receptor (NMDA and AMPA receptor) model of neuronal cell death. This classical glutamate receptor driven model indicates the potential roles of NMDA and AMPA receptors involved in inducing neuronal cell death through excessive extracellular glutamate. Overactivation of these two channels leads to intracellular calcium imbalance, consequently inducing several pathways that eventually lead to neuronal cell death. This figure was modified from Elaine Besancon et al., 2008.

Taken together, the traditional model of excitotoxicity emphasized treatment at the level of glutamate receptor channels to curtail such events. However, the failure of targeting glutamate receptors in all clinical trials indicated that new therapeutic targets needed to be identified for HI brain injury.

 

3 Targeting Non-Glutamate Channels

Several non-glutamate ion channels, including transient receptor potential channels, acid-sensing channels²², hemichannels³⁶, volume-regulated anion channels³⁷ and sodium-calcium exchangers (NCX)^{38, 39} etc. have been implicated and thus identified as potential novel therapeutic targets for ischemic brain injury.

3.1 Sodium-calcium exchangers

The Na(+)/Ca(2+) exchangers (NCX) are bi-directional transmembrane proteins that express widely in the brain³⁸. Under normal physiological conditions, the NCXs exchange one calcium ion out of the cell with three sodium ions going into the cell³⁸. Under pathophysiological conditions like ischemia, the activity of NCXs can be reversed^{38, 40-42}. Instead of transporting calcium ions out of cells, they may alternatively transport them  into cells. Since dysregulation of sodium and calcium homeostasis is a fundamental hallmark following ischemic brain injury, the role of NCXs under ischemia has been studied both in vitro and in vivo^41-44. However, the conclusions remain controversial. Some studies have shown that using NCXs blockers may reduce brain infarction in in vivo stroke models^43-45,while other reports brought up conflicting results that inhibition of NCXs can lead to even worse infarction outcomes^{22, 42}. These variations may in part be due to the diverse responses of NCXs to differences in the severity of ischemic brain injury²². Under mild ischemic brain injury conditions, the NCXs operate in transporting calcium ions out of cells. Therefore, blockage of NCXs reduces calcium extrusion and ends up worsening calcium-mediated cell injury⁴⁵. On the other hand, severe ischemic brain injury conditions involving an overload of intracellular sodium leads to the reverse where NCXs conduct calcium into the cells^{41, 46}.Hence, the blockage of NCXs under these severe ischemic brain injury conditions may potentially be neuroprotective. Therefore, the targeting of these differential responses and the resulting beneficial effects needs to be further elucidated.

3.2 Hemichannels

Hemichannels are proteins that are involved in forming gap junctions. One gap junction channel is composed of two hemichannels, and each hemichannel consists of a hexamer from the connexins transmembrane protein family⁴⁷. The gating of gap junction channels is regulated by the phosphorylation status of connexin proteins^{47, 48}. Under normal conditions, the two hemichannel components stay in a closed state while forming an open state for gap junction channels. Under ischemic conditions, the active opening of hemichannels leads to a reduction in the opening of gap junction channels which consequently results in decreased cell-cell communication^{36, 48, 49}. Research on hemichannels has elucidated their clear involvement in the response to ischemia brain injury. However, their precise role still remains elusive and controversial. Some studies have shown that knocking out connexin 43 (Cx43) in mice results in an increased level of brain infarct volume and apoptosis following stroke^{47, 50-54}. In contrast, other studies report that under specific conditions, gap junctions exacerbate ischemic brain injury by spreading cytotoxic substances into cells^{22, 50, 55}. Further validation on the role of hemichannels is warranted before we can truly assess them as therapeutic targets for stroke and brain injury.

3.3 Volume-regulated anion channels

Chloride (Cl^–) permeates the cell membrane through several types of Cl^– channels. An important class of Cl– channels is the volume-regulated anion channel (VRAC). VRACs are responsible for mediating the swelling-induced Cl– current^{37, 56}, which also plays essential role in the regulatory mechanism in cells for balancing cell volume during osmotic perturbations^57-59. Under normal physiological conditions, VRACs stay in a closed state^58-60. Under pathophysiological conditions like ischemia, VRACs are abnormally overactivated⁵⁸. Once overactivated, subsequent pathological mechanisms can be triggered, including the inhibition of NCXs due to ATP depletion²². VRACs inhibitors have shown to provide a neuroprotective effect following stroke in rats^61-64, which supports the hypothesis that the activation of VRACs can be neurotoxic. DCPIB, a VRACs specific inhibitor, was recently shown to have a neuroprotective effect in several rodent models of hypoxic-ischemic brain injury⁶².

3.4 Acid-sensing ion channels (ASICs)

Acidosis, which worsens neurotoxicity, is a featured outcome that always follows ischemia^{65, 66}. Acid-sensing ion channel 1a (ASIC1a) is highly expressed in the brain and is believed to be involved in acidosis induced brain injury^{65, 67}. During ischemic conditions, the accumulation of lactic acid rapidly decreases the pH level of the brain to 6.2 or even lower, and subsequently activates ASICs^{68, 69}. In vitro and in vivo studies have shown that the blockage of ASICs results in neuroprotective effects. In particular, an in vitro study showed that ASIC currents and ASIC desensitization could be amplified following oxygen-glucose deprivation (OGD), which increased the length of calcium influx⁷⁰. Other in vivo studies have shown that knockouts of ASIC1 in mice protected animals from acidosis-reduced brain injury, and the administration of ASIC blockers reduced brain infarct volume following MCAO brain injury^{68, 71}. Hence, evidence suggests ASICs as potential therapeutic targets for the inhibition of ischemic neuronal death.

3.5 Transient receptor potential melastatin (TRPM) subfamily

The transient receptor potential melastatin (TRPM) protein family is one of 6 subfamilies among the TRP channels superfamily⁷². Two members of the family, TRPM7 and TRPM2, have been implicated in mediating neuronal cell death^73-78. In vitro pharmacological blockage⁷⁹ and in vivo siRNA suppression of TRPM7 both resulted in neuroprotective effects⁸⁰. Recently, in vivo studies have investigated the role of TRPM2 channels in hypoxic-ischemic brain injury and the results showed that knockouts of TRPM2 channels in rodent models have neuroprotective effects^81-83. The underlying mechanism of TRPM family-mediated ischemic brain injury still remains unknown. Since TRPM proteins are calcium permeable ion channels, dysregulation of calcium levels and overload of intracellular calcium levels are the most acceptable mechanisms^{78, 83, 84}. However, TRPM ion channels as TRPM7 are also permeable to other ions such as zinc⁸⁵, which has also been illustrated to play a role during the hypoxic ischemic cascade. In this study, our focus is on the role of TRPM2 channels during neonatal hypoxic ischemic brain injury.

 

 

 

 

 

 

 

Figure 2. Potential mechanisms involved in excitotoxicity following ischemic stress. The left half of the figure shows the traditional glutamate driven model of excitotoxicity. The right half of the figure shows that more and more evidence nowadays suggest that despite the traditional model, non-glutamate driven channels including NCXs, hemichannels, VRACs, ASICs, TRPs may play important roles in mediating the excitotoxicity following ischemic stress. This figure was modified from Elaine Besancon et al., 2008.

4 Transient Receptor Potential Channels (TRPs)

As mentioned above, the TRPM channel subfamily is one of 6 subfamilies among the TRP channel superfamily. The TRP channel was first identified as a protein in Drosophila melanogaster⁸⁶. TRP channels are non-selective cation channels that can be grouped into 6 families named TRPC (canonical), TRPM (melastatin), TRPV (vanilloid), TRPP (polycystin), TRPML (mucolipin) and TRPA (ankyrin), ^{72, 86, 87}. There are 28 mammalian TRP channels that have been identified to date⁸⁸. All TRP channels consist of six transmembrane domains arranged in a tetrameric structure, and they are widely expressed in various cell types in the body including neurons^{87, 88}.

5 TRPM2 Channel

5.1 TRPM2 protein structure, transmembrane topology and distribution

TRPM2 is a calcium-mediated nonselective cation channel. TRPM2 channels are expressed in many tissues including the brain (high expression), lung, liver and heart^{89, 90}. TRPM2 also expresses in various cell types, including neurons, microglial cells, immune cells and pancreatic β-cells^{78, 89}.

TRPM2 proteins are encoded by TRPM2 genes. In rodents, the TRPM2 gene consists of 34 exons and spans around 61 kb. In human, the TRPM2 gene consists of 32 exons and spans about 90 kb, with the location of the gene being on chromosome 21q22.3⁹⁰. There is an additional exon located at the 5’ terminus with a CgG island in the human TRPM gene. The full length transcript of TRPM2 is approximately 6.5 kb and encodes TRPM2 protein comprising of 1503 amino acids

Figure 3. The TRP channel superfamily. The TRP channel superfamily comprises of 6 subfamilies including TRPC (canonical), TRPM (melastatin), TRPV (vanilloid), , TRPP (polycystin), TRPML (mucolipin). and TRPA (ankyrin).

with a molecular weight of 170 kDa⁹⁰. In addition to full length TRPM2 transcripts, four splice variants of TRPM2 have been identified: TRPM2-ΔN, TRPM2-ΔC, TRPM2 –S and TRPM2-SSF^89-91. Consistent with their names, TRPM2-ΔN is loss of amino acids 538–557 in the N-terminus; TRPM2-ΔC is loss of amino acids 1292–1325 in the C-terminus, particularly the CAP domain of the NUDT9-H domain; TRPM2-S (short) is loss of the entire C terminus including the channel pore; TRPM2-SSF (striatum short form) is loss of the first 214 amino acids of the N-terminal and has been found to uniquely express within the striatum^89-91.

The TRPM2 protein structure consists of six transmembrane segments (S1-S6) with a pore loop region located between S5 and S6⁹⁰. To form a channel, TRPM proteins typically assemble into homotetramers with both N- and C- termini flanking the intracellular sides⁹⁰. At the N-terminus, there are 4 homologous regions and a calmodulin (CaM) binding domain, which is a region that plays an important role in regulating the channel activation property⁹⁰. At the C-terminus, there is a TRP box as well as a coiled-coil domain, both of which are assumed to be essential for TRPM2 homogenous tetrameric assembly⁹⁰.

Figure 4. TRPM2 protein structure and variants. The upper panel of the figure shows a representative structure of the TRPM2 channel and its topology. The lower panel of the figure shows different forms of TRPM2, including the full-length long form TRPM2 (TRPM2-FL), TRPM2 cleavage of N terminus K538-Q557 (TRPM2-ΔN), TRPM2 cleavage of C terminus T-1292-L1325 (TRPM2-ΔC), TRPM2 short striatum variant that has 214 residues missing from the C terminus (TRPM2-SSF) and TRPM2 cleavage of the entire C terminus (TRPM2-S). This figure was modified from Lin-Hua Jiang et.al, 2010.

5.2 TRPM2 biophysical properties and gating mechanism

The TRPM2 channel exhibits a linear I/V curve, which suggests that channel activity is independent from voltage-gating^{87, 92}. Instead, TRPM2 is a ligand-gated channel that can be activated by several intracellular and extracellular features, among which ADPr and hydrogen peroxide are the most potent activators^{89, 90}.

Hydrogen peroxide can directly and indirectly activate TRPM2 channels. The ability of H₂O₂ to activate TRPM2 channels has attracted significant scientific interest. Studies have used the H₂O₂ driven mechanism for explaining the pathological processes that are mediated by elevation of the oxidative microenvironment^{79, 93}. Such pathological processes include hypoxic ischemic (HI) brain injury, diabetes, inflammation and other neurodegenerative disorders like bipolar diseases.^{94, 95} Endogenously, H₂O₂ is initially generated from mitochondria following oxidative phosphorylation⁹⁶. Exogenously, generation of H₂O₂ is induced as a result of responding to external factors such as certain drugs, heavy metals, visible light or even heat, in consistence with other reactive oxygen species (ROS) like hydroxyl radicals (OH.)^97-99. Overgeneration of ROS results in dramatic damage to biological molecules such as DNA and proteins, or any molecule that is involved in the chain reaction cascade producing cellular damage and disease.

In terms of activating the TRPM2 channel indirectly, hydrogen peroxide activates the TRPM2 channel through regulation of the metabolic pathway that produces ADPR¹⁰⁰. ADPR can bind to TRPM2 at the active site in the NUDT9-H region at the C-terminus^{90, 100}. The extracellular stimulation of hydrogen peroxide leads to an intracellular increase in hydrolase activity, which thereby hydrolyzes more NAD+ and cADPR to produce additional ADPR⁹⁰. Another source of ADPR is the action combination of poly (ADPR) polymerases (PARPs, PARP enzymes) and poly (ADPR) glycohydrolases (PARG enzymes)^{89, 90, 101}. This source indirectly generates ADPR through the formation and hydrolysis of poly-ADPR when it is overactivated in response to DNA damage^{89, 90, 100, 101}. At the TRPM2 N-terminus, calcium can bind to the CaM-binding motif, which is another mechanism of gating that is independent of ADPR and hydrogen peroxide^{89, 90, 102}.

5.3 The physiological and pathophysiological role of TRPM2 channels

As mentioned above, TRPM2 has been identified in several different cell types including neurons, immune cells and pancreatic β-cells. Therefore, it is not surprising that TRPM2 is associated with ailments such as CNS diseases and type II diabetes^{78, 103, 104}. However, the precise mechanisms of these pathologies still require further investigation. In this case, oxidative stress^{78, 105-107} and amyloid beta^108-110 mediated pathological activation of TRPM2 channels seem the most likely mechanisms. TRPM2 is highly permeable to calcium, and can mobilize calcium ions from both the extracellular and intracellular spaces. Hence, its biological significance is strongly associated with the intracellular calcium level. Under normal physiological conditions, the calcium-mediated activity of the TRPM2 channel has been reported to be involved in several physiological processes, including inflammation¹¹¹, synaptic transmission¹¹², microglial activation¹¹³ and insulin secretion¹¹⁴. Under pathophysiological conditions, the abnormal overactivation of the TRPM2 channel may lead to intracellular calcium overload, which can subsequently lead to various diseases.

In the CNS, TRPM2 is most abundantly expressed in the brain where it has been implicated in triggering numerous physiological and pathophysiological processes. For example, TRPM2 is involved in mediating neuronal cell death that can consequently lead to CNS diseases including stroke, Alzheimer’s disease and bipolar disorder. One study has shown that patients with bipolar disorders have relatively higher levels of basal intracellular calcium ions and that the TRPM2 gene sites are located within chromosome region 21q22.3, conferring increased susceptibility to this pathology⁹⁵. Another study showed that knocking out TRPM2 reduced ischemic brain damage in MCAO model of adult mice⁸³. Our lab also recently demonstrated that TRPM2 knockouts provide a neuronal protective effect following HI brain injury in neonatal mice¹¹⁵. Together, these previous studies indicate that TRPM2 is a promising therapeutic target for the treatment of HI brain injury. Therefore, my project will examine the effects of TRPM2 inhibition on neuroprotection that may lead to potential drug development for neonatal hypoxic-ischemic brain injury.

In addition to the CNS, TRPM2 has also been identified in pancreatic β-cells¹¹⁴. Activation of TRPM2 channels has also been linked to insulin secretion and H₂O₂-induced apoptosis of insulin-secreting cells, implicating a potential role of TRPM2 in diabetes. A recent study using the TRPM2 knockout mouse model revealed the involvement of the channel in insulin secretion from pancreatic β-cells. In this case, TRPM2 knockout mice demonstrated relatively higher basal blood glucose levels in comparison to WT mice, while plasma insulin levels remained similar¹¹⁴.

TRPM2 has also been identified in some cell types in the immune system, including macrophages, neutrophils and lymphocytes, suggesting a possible association with inflammatory diseases^{90, 116}.

 

Figure 5. Proposed mechanisms of TRPM2 channel activation by H₂O₂ and involvement TRPM2 channel activity in physiological and pathophysiological processes. H₂O₂can activate the opening of TRPM2 channels directly and indirectly. Activation of the TRPM2 channel leads to extracellular calcium influx, which may further facilitate the channel opening. Influx of calcium leads to intracellular calcium imbalance, which suggests the involvement of TRPM2 channels in physiological processes such as insulin release, cytokine production, increased endothelial permeability and cell death. Therefore, the actions of TRPM2 may contribute to pathologies or disease conditions. +: activation. This figure was modified from Lin-Hua Jiang et.al, 2010.

6 Pharmacological Interactions

As there is significant interest in the exploration of TRPM2 as a potential target for neurodegenerative diseases, the pharmacology of TRPM2 has also started to receive considerable attention in the field of research. Through patch clamp electrophysiological techniques and the availability of HEK293 cells, different TRPM2 channel inhibitors have been studied. Thus far, none have demonstrated a desirable specificity with compounds that have been reported to have an inhibitory effect on the TRPM2 channel also affecting other TRP channels⁸⁹.

6.1 Flufenamic acid (FFA)

Flufenamic acid (FFA) was the first TRPM2 blocker identified¹¹⁷. It belongs to class of non-steroidal anti-inflammatory drugs (NSAIDs). Such fenamates are capable of producing anti-inflammatory effects in the CNS. Studies have been conducted on TRPM2-overexpression HEK293 cells, where FFA evoked a pH-dependent inhibition of ADPR- or H₂O₂-induced cation currents¹¹⁷. However, the inhibitory effect of FFA is not limited to the TRPM2 channel; it can similarly affect other channels in the TRP family including TRPM4, TRPM5, TRPC3 and TRPC5^{90, 117}. Furthermore, it also has an activation effect on TRPC6 and transient receptor potential ankyrin 1 (TRPA1)^{90, 117}. Therefore, fenamates such as FFA are hardly satisfactory tools in clarifying the role of the TRPM2 channel.

6.2 Anti-fungal agents (clotrimazole and econazole)

Anti-fungal compounds block TRPM2 channels activated by ADPR in TRPM2-overexpressing HEK293 cells, but the inhibitory effect is irreversible^{89, 118}.

6.3 2-APB

2-APB was first identified as an inositol 1,4,5-trisphosphate(IP3) receptor antagonist. It has also been reported to exert an inhibitory effect on certain TRPC and TRPM channels while having an activation effect on some TRPV channels^{24, 117, 119}.

6.4 Divalent heavy metal cations

Some heavy metal ions, La³⁺ and Gd³⁺ for instance, are known to have an inhibitory effect on most TRP channels, though TRPM2 seems to be an exception. A recent study shows that divalent copper (Cu^{2 +} ) may be a potent TRPM2 channel blocker⁹⁸.

6.5 AG490

Recently, AG490 was identified to have an inhibitory effect on the TRPM2 channel¹²⁰. AG490 was shown to almost completely block H₂O₂-induced intracellular Ca²⁺ increase and significantly reduce H₂O₂-induced TRPM2 currents¹²⁰. While H₂O₂ can also activate the TRPA1 channel, AG490 has no significant effect on the H₂O₂-induced Ca²⁺ influx mediated by TRPA1 channels. ADPR is another endogenous activator for TRPM2. However, in a patch clamp study on TRPM2-overexpressing HEK293 cells, AG490 only inhibited H₂O₂-induced but not ADPR- induced TRPM2 inward current¹²⁰. Structurally, AG490 belongs to the tyrphostin family and was synthesized in the early 1990s. Since AG490 was initially found to be a JAK2 inhibitor, the same study examined whether the inhibitory effect of AG490 on TRPM2 activity was dependent on the inhibition of JAK2. The study tested the effects of other JAK2 inhibitors, none of which had an effect on H₂O₂-induced Ca²⁺ increase by TRPM2 channels. Thus, the results suggested that the inhibitory effect of AG490 on TRPM2 activation is independent from JAK2 inhibition¹²⁰. Hence, AG490 may manifest its inhibitory effect by scavenging intracellular hydroxyl radicals.    

 

 

     

 

Figure 6. TRPM2 channel inhibitors. Shown above are the chemical structures of compounds that have been indicated to have an inhibitory effect on TRPM2 channels. FFA, anti-fungal agents (clotrimazole and econazole) and 2-APB are the classical TRPM2 inhibitors. However, studies have shown their inhibitory effects to be unsatisfactory. AG490 is a newly discovered TRPM2 inhibitor which inhibits activation by efficiently blocking TRPM2 currents.

Chapter 2  Rationale and Hypothesis

Rationale: HI brain injury is a severe public health issue with no effective pharmacological method of prevention thus far. It has been previously confirmed that TRPM2 plays a neuroprotective role in adult cerebral ischemia using TRPM2 KO mice⁸³. Our lab has also recently shown that knocking out TRPM2 provided a neuroprotective effect following HI brain injury in neonates¹¹⁵. Therefore, I proposed to test the effects of a TRPM2 channel inhibitor in HI brain damage in relation to potential drug development for HI brain injury.

AG490 is a recently identified TRPM2 current inhibitor that efficiently blocks H₂O₂ induced TRPM2 current in HEK cells¹²⁰. AG490 was initially discovered as a JAK2 inhibitor but could inhibit TRPM2 in a JAK2 independent manner¹²⁰. In this case, AG490 may act as a ROS scavenger and reduce the concentration of hydrogen peroxide: . It was previously reported that silencing TRPM2 using siRNA reduced H₂O₂-induced neuronal cell death in vitro⁷⁹. It is also reported that following hypoxic-ischemic injury, there is an accumulation of H₂O₂ in the neonatal brain at postnatal day 7 (neonatal mice brains) which is not been seen in the postnatal day 42 mouse brain (adult mice brains)¹²¹. Thus, it is expected that TRPM2 may play a greater role in neonates compared to adult, and AG490 could be a suitable pharmacological tool for my project

Hypothesis: I hypothesize that inhibition of TRPM2 channels by TRPM2 inhibitor AG490 provides neuroprotection following hypoxic-ischemic brain injury in neonates.

Chapter 3  Aims and Experimental Design

Aims:

AIM1: Investigate the effect of TRPM2 inhibition on H₂O₂-induced neuronal cell death in vitro.

AIM2: Investigate the effect of TRPM2 inhibition on neonatal hypoxic-ischemic brain injury in vivo.

Experimental Design Outline:

 

Figure 7. An outline of the project experimental design.