Acidotoxicity via ASIC1a Mediates Cell Death during Oxygen Glucose

Jan 24, 2017 - *Mailing address: Biochemistry and Neuroscience Program, Institute of Arctic Biology, University of Alaska Fairbanks, PO Box 757000, Fa...
0 downloads 9 Views 2MB Size
Subscriber access provided by HACETTEPE UNIVERSITESI KUTUPHANESI

Article

Acidotoxicity via Asic1a Mediates Cell Death during Oxygen Glucose Deprivation and Abolishes Excitotoxicity Saurav Bhowmick, Jeanette T. Moore, Daniel L. Kirschner, Mary C. Curry, Emily G. Westbrook, Brian T. Rasley, and Kelly L. Drew ACS Chem. Neurosci., Just Accepted Manuscript • DOI: 10.1021/acschemneuro.6b00355 • Publication Date (Web): 24 Jan 2017 Downloaded from http://pubs.acs.org on January 25, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Chemical Neuroscience is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

1

Acidotoxicity via Asic1a Mediates Cell Death during Oxygen Glucose Deprivation and

2

Abolishes Excitotoxicity

3

Saurav Bhowmick1, 2, Jeanette T Moore2, Daniel L. Kirschner1, Mary C. Curry1, Emily G.

4

Westbrook1, Brian T. Rasley1, Kelly L. Drew1, 2*

5

1

6

USA.

7

2

8

Corresponding author: KL Drew, Biochemistry and Neuroscience Program, Institute of Arctic

9

Biology, University of Alaska Fairbanks, PO Box 757000, Fairbanks, AK 99775-7000, USA.

Department of Chemistry and Biochemistry, University of Alaska Fairbanks, Fairbanks, AK,

Institute of Arctic Biology, University of Alaska Fairbanks, Fairbanks, AK, USA.

10

Email: [email protected], Telephone: 907-474-7190

11

Cover Title: Acidosis Overrides NMDA Mediated Excitotoxicity

12

ABSTRACT

13

Ischemia-reperfusion (I/R) injury is associated with a complex and multifactorial cascade of

14

events involving excitotoxicity, acidotoxicity, and ionic imbalance. While it is known that

15

acidosis occurs concomitantly with glutamate-mediated excitotoxicity during brain ischemia, it

16

remains elusive, how acidosis-mediated acidotoxicity interacts with glutamate-mediated

17

excitotoxicity. Here, we investigated the effect of acidosis on glutamate-mediated excitotoxicity

18

in acute hippocampal slices. We tested the hypothesis that mild acidosis protects against I/R

19

injury via modulation of NMDAR, but produces injury via activation of acid sensing ion

20

channels (ASIC1a). Using a novel microperfusion approach, we monitored time course of injury

21

in acutely prepared, adult hippocampal slices. We varied the duration of insult to delay the return

22

to pre-insult conditions to determine if injury was caused by the primary insult or by the modeled

23

reperfusion phase. We also manipulated pH in presence and absence of oxygen glucose

24

deprivation (OGD). The role of ASIC1a and NMDAR was deciphered by treating the slices with

25

and without an ASIC or NMDAR antagonist. Our results show that injury due to OGD or low pH

26

occurs during the insult rather than the modeled reperfusion phase. Injury mediated by low pH or

27

low pH OGD requires ASIC1a and is independent of NMDAR activation. These findings point

28

to ASIC1a as a mediator of ischemic cell death caused by stroke and cardiac arrest.

29

Keywords: Brain Ischemia, Excitotoxicity, Glutamate, pH, oxygen-glucose deprivation and

30

reperfusion, acute hippocampal slice

31 1 ACS Paragon Plus Environment

ACS Chemical Neuroscience

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 27

1

INTRODUCTION

2

Ischemic reperfusion (I/R) injury associated with stroke and cardiac arrest can lead to severe

3

morbidity and has a mortality rate between 30-100%.1 Several clinical trials aimed at NMDA (N-

4

methyl-d-aspartic acid) mediated excitotoxicity have failed.2,

5

brain injury due to a drop in tissue pH to 6.5 – 6.0.4, 5 However, an intriguing debate has emerged

6

regarding the influence of mild acidosis on cell damage/ protection during cerebral I/R injury.

7

One line of evidence suggests that mild acidosis is neuroprotective6,7 and the neuroprotection is

8

due to modulation of the NMDA receptor mediated ischemic cascade.8,

9

several other works suggest that acidosis activates acid sensing ion channels (ASICs) which

10

results in intracellular Ca2+ overload causing acidotoxicity independent of NMDA mediated

11

excitotoxicity.11, 12

3

Acidosis exacerbates ischemic

9, 10

On the contrary,

12

Although previous in vitro and in vivo studies show that brain ischemia triggers

13

excessive glutamate release,13,14 activation of NMDA receptors15 and concomitant tissue

14

acidosis,16 it remains elusive how the acidotic component interacts with glutamate-mediated

15

excitotoxicity. Challenges in deciphering the interaction between these two mechanisms are

16

overcome in the current study by use of a novel in vitro microperfusion technique in which pH is

17

manipulated independently from oxygen glucose deprivation (OGD).

18

The present study tested the hypothesis that low pH during OGD blocks NMDA

19

mediated excitotoxicity but produces acidotoxicity via ASIC1a. We utilized acute rat

20

hippocampal slices to ascertain the time course of injury from OGD with and without acidosis as

21

well as acidosis alone. Cell death subsequent to these manipulations was assessed by lactate

22

dehydrogenase (LDH) release.17

23

RESULTS AND DISCUSSION

24

OGD injury via microperfusion approach resembles similar disruption in energy

25

homeostasis and excitatory neurotransmitter efflux that mimics in vivo I/R Injury.

26

During cerebral ischemia in vivo, acidotoxicity and excitotoxicity occur concomitantly

27

and it is unclear how these two mechanisms interact. Moreover, it is not possible to manipulate

28

acidosis to decipher how it interacts with excitotoxicity. Traditional in vitro slice preparations do

29

not produce the same changes in the extracellular milieu as what occurs in vivo because of rapid

30

washing with a highly diluted bath solution.18 The finding of this study involves using a novel

31

microperfusion technique that allows for time resolved study of cell death and manipulation of 2 ACS Paragon Plus Environment

Page 3 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

1

pH with and without OGD thus allowing for study of the interaction between acidosis and OGD-

2

induced injury. The advantages of this model include the opportunity for temporal resolution of

3

injury as well as overflow of metabolites and neurotransmitters and to manipulate each of the

4

components of the ischemic cascade to better understand interactions. With the microperfusion

5

approach we measured only LDH release as a marker of cell death. Cytosolic lactate

6

dehydrogenase release following injury is due to membrane disruption that correlates with cell

7

death. In the present study, for the first time we have monitored continuous LDH release in an

8

acute slice preparation where it allows for observation of the time course of membrane

9

disruption. Here, we investigated the effect of in vitro ischemia like condition (OGD) exposure

10

on energy homeostasis19 and excitotoxic neurotransmitter efflux,13 a hallmark of ischemic injury

11

in acute hippocampal slices from rat. Results show that exposure of acute hippocampal slices to

12

30 min OGD insult resulted in a significant depletion in ATP levels. Slices treated with OGD

13

showed a significant drop in ATP when compared to control slices. Slices with OGD showed no

14

signs of ATP restoration even after 3 hours of reperfusion (Figure 1 A). This indicates that

15

energy homeostasis is highly dysregulated due to OGD exposure. To further validate the

16

disruption of energy homeostasis in correlation to adenosine, we investigated the level of

17

adenosine efflux upon OGD exposure. Slices exposed to 30 min of OGD showed significant

18

efflux in extracellular adenosine when compared to the control slices. The adenosine level

19

returned to baseline within 15 min of the onset of reperfusion (Figure 1 B). From our results, we

20

conclude that when acute hippocampal slices are subjected to 30 min of OGD insult, they release

21

extracellular adenosine produced from rapid degradation of ATP due to ischemic-like conditions.

22

We next investigated the effect of OGD exposure on excitotoxic neurotransmitter efflux

23

such as glutamate and aspartate. Consistent with previous in vivo studies,14,20 perfused slices

24

exposed to OGD injury for 30 min released glutamate (62-fold increase over basal level) (Figure

25

2 A). Similarly, aspartate efflux also increased during OGD (11.4-fold increase over basal level)

26

(Figure 2 B). After the onset of reperfusion, the glutamate levels returned to baseline levels

27

within 60 min, aspartate returned to baseline levels within 45 min.

28

The observations reported herein validate the microperfusion approach. Dysregulation in

29

energy homeostasis and excessive release of excitatory neurotransmitters are decisive for the

30

development of excitotoxicity during cerebral ischemia. With our approach, when hippocampal

31

slices are subjected to OGD, tissue ATP declines and adenosine, glutamate, and aspartate are 3 ACS Paragon Plus Environment

ACS Chemical Neuroscience

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

released into the extracellular space. These observations are in accord with previous in vitro and

2

in vivo studies that supports similar energy dysregulation21,22 and excitatory neurotransmitter

3

release during cerebral ischemia in vivo and OGD in vitro.13, 14, 23, 24, 25

4

OGD or Acidosis Causes Injury during Insult rather than during Reperfusion

5

How much injury occurs during the ischemic and reperfusion phases continues to be

6

controversial in both experimental animal and clinical studies.26 Intrinsic to ischemia reperfusion

7

injury is the concept that harmful events are occurring during both ischemia and reperfusion.27, 28

8

With the microperfusion approach, we monitored the timing of injury to decipher whether cell

9

death was due to “insult” (OGD or low pH) or to “reperfusion”. In these experiments,

10

reperfusion was modeled by returning oxygen and glucose or by returning pH to 7.3. We varied

11

the duration of insult between 15, 30, and 45 min with the expectation that delayed return of

12

oxygen and glucose or delayed return to pH 7.3 would delay LDH release if injury was caused

13

by the modeled reperfusion phase. Delaying onset of modeled reperfusion (for 15 min, 30 min,

14

or 45 min) or the return to pH 7.4 from pH 6.4 (for 15 min, 30 min, or 45 min) did not delay the

15

peak time of LDH release. However, the magnitude of injury was increased (Figure 3 A-B). This

16

indicates that injury was caused by the primary OGD or low pH insult and not by modeled

17

reperfusion.

18

One study suggests that low pH causes detrimental effect during reperfusion.29 However,

19

extracellular pH decreases immediately30 after the onset of ischemia and is restored to the

20

normal physiological range upon reoxygenation.31 Given this time course of acidosis, a sustained

21

severe acidosis during the reperfusion phase is unlikely. When acidosis was prolonged during the

22

low pH insult phase, injury increased in magnitude but was not offset by the time of reperfusion.

23

A similar observation was made with OGD insult. Thus, with regards to timing of acidosis, our

24

experimental set-up mimicked the in vivo scenario and shows that injury associated with OGD or

25

low pH occurs during insult rather than during reperfusion.

26

Low pH Delays OGD Injury and Shows Injury Independent of OGD Insult

27

Buffering of the OGD solution during in vitro experiments eliminates the acidotic

28

component that occurs during cerebral ischemia in vivo. To study the interaction between

29

acidosis and OGD on cellular injury, the pH of the OGD solution was adjusted from 7.3 to 6.4.

30

OGD at pH 6.4 decreased and delayed LDH release as compared to OGD at physiological pH

31

7.3. Results show that OGD with low pH shows similar time course of injury as observed with 4 ACS Paragon Plus Environment

Page 4 of 27

Page 5 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

1

low pH alone. With similar time course of injury, no significant statistical differences were

2

observed at any given time point between OGD with low pH and low pH alone (Figure 4 A).

3

This suggests that when low pH is introduced along with OGD, low pH overrides OGD mediated

4

injury. When the total injury represented by the area under the curve was considered, LDH

5

released by pH 6.4 was not different from LDH released by OGD, pH 7.3 or by OGD pH 6.4

6

(Figure 4 B). This indicates that low pH delays OGD injury, but on its own, causes injury that is

7

as deleterious as OGD.

8

Injury Caused by OGD Insult Is Due to NMDAR Activation Whereas Low pH Injury Is

9

ASIC1a mediated

10

We next ask whether the delay in injury caused by acidosis is due to modulation of

11

NMDAR. While application of NMDAR antagonist (MK-801) decreased LDH release in groups

12

subjected to OGD, pH 7.3 (Figure 5 A), no significant effect on LDH release was observed in

13

slices subjected to acidosis, pH 6.4 (Figure 5 B) or OGD with acidosis, pH 6.4 (Figure 5 C).

14

This suggests that OGD injury at physiological pH involves NMDAR activation whereas injury

15

associated with acidosis alone or acidosis in combination with OGD is independent of NMDAR

16

mediated excitotoxicity.

17

Because acidosis contributes to cell death independent of OGD (Figure 4 A-B), we next

18

explored the role of ASICs. We expected that targeting the ASIC1a with specific and non-

19

specific blocker would attenuate pH 6.4-mediated cell death. As illustrated in Figure 6 A-B,

20

continuous perfusion of 100 µM amiloride (nonspecific ASIC blocker which acts by blocking the

21

channel pore) during acidosis, pH 6.4 and OGD pH 6.4 tended to decrease LDH release;

22

however, failed to produce a statistically significant effect. Next, we tested the effect of

23

Psalmotoxin (PcTx), a specific blocker of Ca2+-permeable homomeric ASIC1a. Application of

24

PcTx (50 nM) blocked aCSF pH 6.4 and OGD pH 6.4–induced LDH release (Figure 6 C-D).

25

This result supports the involvement of ASIC1a in low pH mediated injury.

26

Results reported herein demonstrate that delayed injury was solely due to ASIC1a

27

activation since NMDAR blockade had no effect on OGD-induced cell death in the presence of

28

proton load. No decrease in LDH release with NMDAR blockade in presence of proton load

29

could also be due to high concentration of MK-801(100 µM). The higher concentration might

30

have compromised the specificity of MK-801 on the NMDAR receptor favoring non-specific

31

binding that masked the neuroprotective effect of NMDAR blockade. Since, with the same 5 ACS Paragon Plus Environment

ACS Chemical Neuroscience

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 27

1

concentration of MK-801, we observed decreased LDH release in groups subjected to OGD at

2

pH 7.3, we believe that the non-specificity associated with high concentration of MK-801 did not

3

confound the results of our study. Blocking the NMDAR with MK-801 will block the NR2B-

4

NMDARs/CaMKII cascade described previously to enhance ASIC1a-mediated current32. The

5

present work cannot rule out a NR2B-specific pathway, however, injury caused by low pH alone

6

or with OGD did not depend on a NR2B/CaMKII pathway since MK-801 blocks all NMDARs

7

including those that contain NR2B.

8

Preclinical in vivo studies and clinical trials have also failed to show neuroprotective effects of

9

MK-801.33-35 The interpretation that low pH abolishes NMDAR mediated excitotoxicity

10

contrasts with other studies where authors have suggested that OGD-induced cell death is due to

11

NMDAR, but delayed by pH dependent inhibition of NMDA.7, 8, 36 Here we show that delayed

12

injury is solely due to ASIC1a, that is unmasked by complete attenuation of excitotoxicity.

13

This interpretation is consistent with prior work showing that mild acidosis during brain

14

ischemia protects against excitotoxic injury by modulating glutamate-mediated excitotoxicity.6, 7

15

These results also support prior work showing that increased load of proton due to acidosis4, 37

16

activates a distinct family of ligand-gated channels, the acid-sensing ion channels, ASICs

17

(ENaC/Deg superfamily)38, 39, 40 and mediates injury unrelated to NMDAR. The role of ASIC1a

18

activation during acidotoxicity has been previously confirmed in in vivo studies. Furthermore,

19

ASIC1a knock-out or pharmacological blockade of ASIC1a attenuates injury associated with

20

brain ischemia. Our present results with selective antagonism of ASIC1a with PcTX1 are

21

consistent with the role of ASIC1a in ischemic damage as previously reported. The less robust

22

ability of amiloride to attenuate the effect of low pH could be due to dose. The IC50 of amiloride

23

to inhibit ASICs is between 5 and 100 µM.41 In this dose range, amiloride is not specific for

24

ASIC1a. For example, at 100 µmol/L, the concentration used in this study, amiloride enhances

25

the sustained current evoked in cells transfected with ASIC3.42,

26

concentrations amiloride affects other targets including other ion channels (e.g., epithelial Na+

27

channel and calcium channels) and exchangers (e.g., Na+/H+ and Na+/Ca2+).44, 45 Thus, it was not

28

feasible to test a higher dose of amiloride without confounding results by nonspecific effects. For

29

this reason, we tested the specific ASIC1a blocker PcTX1. This specific blocker of ASIC1a

30

completely abolished LDH release caused by both low pH alone and low pH with OGD.

31

Both activation of NMDAR and ASIC1a causes high influx of intercellular Ca2+ that causes cell 6 ACS Paragon Plus Environment

43

At slightly higher

Page 7 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

1

damage. Previous studies provide strong evidence that ASIC1-a mediated increase in Ca2+ influx

2

could contribute to Ca2+ overload, an important mechanism of cell damage and death in

3

ischemia.11, 12 Others have shown that Ca2+ is not required for ASIC1a-mediated cell death.46 Our

4

study of the interaction of OGD and pH was enabled through the use of a novel microperfusion

5

technique. This in vitro preparation allowed for independent manipulation of OGD and pH and

6

for time resolved assessment of cell death. The approach was validated by demonstrating that it

7

mimics an in vivo cerebral ischemia scenario in terms of dysregulation in energy homeostasis

8

and excessive release of excitatory neurotransmitters, hallmarks of ischemic injury. With the

9

microperfusion approach, we measured only LDH release as a marker of cell death throughout

10

the slice and cannot distinguish between sub regions of the hippocampus or between cell type.

11

LDH release is a standard approach to monitor cell death47 which is released from the cytosol

12

due to membrane disruption and not due to metabolic changes related to OGD. In the present

13

study, for the first time we have monitored continuous LDH release in an acute slice preparation

14

where it allows for observation of the time course of membrane disruption from the entire slice.

15

We interpret membrane disruption as an early indicator of cell death. Other markers of cell death

16

such as propidium iodide (PI) staining produce high background due damage of the cells at the

17

surface during acute slice preparation. Results here, based on time resolved release of LDH are

18

consistent with prior work from this laboratory, based on PI staining in acutely prepared or short-

19

term cultured slices.48, 49 Although slices are incubated until high baseline levels of LDH reach a

20

stable minimum, there is a possibility that acutely injured cells on the surface of the slice are

21

affected more than cells at the center of the slice. Thus, the novel microperfusion method

22

developed in our laboratory50 for study of cell death in acute, adult slices overcomes the

23

limitations of methods used in prior studies where high baseline cell death may compromise

24

interpretation.48,

25

trauma at the surface of the slice by capturing LDH release from the entire slice. The

26

microperfusion approach also monitors the time course of injury as well as the overflow of

27

metabolites and neurotransmitters. Despite these advantages over other in vitro approaches to

28

monitor cell death, we acknowledge that the slice preparation differs from the in vivo scenario.51,

29

52

30

environment that could not be achieved in vivo. Here we show that LDH release during OGD or

31

low pH insult reaches a maximum peak and then subsides. We suggest that the decay in the LDH

49

The microperfusion approach minimizes background cell death caused by

However, the microperfusion approach gives superior control over the extracellular

7 ACS Paragon Plus Environment

ACS Chemical Neuroscience

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

signal peak is not a consequence of dilution because flow rate and sampling rate are constant

2

throughout the experiment. Furthermore, the waning in LDH release is not due to depletion of

3

LDH since no difference in total LDH content was observed in slices subjected to aCSF, OGD or

4

low pH treatment (Data not shown). Further study will be required to understand why LDH

5

release subsides over time. Previous in vitro studies performed in cell culture or in cultured post-natal brain tissue.8, 9,

6 7

29, 53

8

shows that 30 minutes of insult is sufficient to cause immediate membrane disruption marked by

9

LDH release. It is to be noted that there is a difference in how each model responds to a

10

commonly studied ischemic-like paradigm, oxygen-glucose deprivation. Previous studies show

11

that acute slice viability is affected by OGD lasting as little as 10 min54, 55 whereas, organotypic

12

hippocampal slice cultures tolerate a much longer duration of challenge.53, 56, 57. We chose to use

13

acute slices, because it allows for study of cell death in adult tissue not possible with organotypic

14

hippocampal slice culture. Organotypic cultures differ from adult brain slices in several ways

15

such as influence of age (adult vs. neonatal) which may influence the mechanisms of resistance

16

to injury, the proportion of neurons to glia, synaptic connections between neurons, neural

17

circuitry that influences metabolic demand of the tissue and maturity of the neurons studied58-60.

18

Thus, study in cultured cells or post-natal tissue may not reflect mechanisms operative in adult

19

brain tissue.

20

CONCLUSION

suggest that cell death typically occurs 4 hours after OGD treatment. By contrast, our study

21

This paper reports several important findings using a novel microperfusion technique that

22

for the first time provides time-resolved observations of excitotoxicity and acidotoxicity in an in

23

vitro model of cerebral ischemia/reperfusion in adult hippocampus. Firstly, our results

24

demonstrate that injury associated with acidosis (pH 6.4) or OGD (pH 7.3) occurs during insult

25

rather than during the reperfusion phase. Secondly, we show that acidosis combined with OGD

26

delays OGD injury. Finally, we show that when OGD is combined with acidosis as occurs in

27

vivo, acidotoxicity mediated via ASIC1a occurs but low pH abolishes NMDAR mediated

28

excitotoxicity.

29

In conclusion, we show that acidosis eliminates the NMDAR contribution to OGD-

30

induced cell death but causes injury through an ASIC1a mediated mechanism. Although this

31

study resolves controversy surrounding the interplay between NMDA and ASIC1a, the 8 ACS Paragon Plus Environment

Page 8 of 27

Page 9 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

1

translational potential of ASIC1a may be limited. Acidosis may have longer-term effects not

2

addressed here. Pignataro et al., 2007 observed protective effects when an ASIC1a antagonist

3

was applied as late as 5 h after occlusion in vivo.61 Other mechanisms such as uncoupling of

4

NADPH oxidase from NMDAR activation by intracellular acidosis62 or acidosis mediated

5

reorganization of mitochondrial efficiency63 could also contribute to longer-lasting intercellular

6

acidosis mediated protective effects. Identification and understanding of the interaction between

7

acidotoxicity and NMDAR mediated excitotoxicity point towards a better mechanistic approach

8

in design of an effective therapeutic strategy for brain ischemia caused by stroke and cardiac

9

arrest.

10

METHODS

11

Animals

12

Adult Male Sprague Dawley rats (Simonsen labs, Gilroy, CA; age 3-4 months, 250-400 g) were

13

used for all experiments. The animals used in this study were housed in University of Alaska

14

Fairbanks Animal Resource Center (ARC), fed ad libitum, and kept under a 12:12 light: dark

15

cycle with lights on at 0700 and off at 1900. During housing, animals were monitored once daily

16

for health status. No adverse events were observed during housing. All experimental procedures

17

were approved by the University of Alaska Fairbanks Institutional Animal Care and use

18

Committee (IACUC), performed in accordance with the strict guidelines of the 8th edition of the

19

Guide for the Care and Use of Laboratory Animals published by the US National Institutes of

20

Health. Research were conducted in accordance with ARRIVE guidelines for reporting animal

21

research.64

22

Acute Hippocampus slices preparation

23

Slices were prepared between 0800 and 1030. Animals were anesthetized using 5% (v/v)

24

isoflurane with medical graded O2 at a constant flow rate of 1.5 L/min. Once unresponsive, the

25

animals were euthanized via rapid decapitation and their brains quickly removed. The whole

26

brain was then placed in ice chilled, oxygenated HEPES buffered artificial cerebral spinal fluid

27

(HEPES-aCSF) containing 120 mM NaCl, 20 mM NaHCO3, 6.68 mM HEPES acid, 3.3 mM

28

HEPES sodium salt, 5.0 mM KCl, 2.0 mM MgSO4 (pH 7.3 - 7.4) to attenuate edema during the

29

course of slicing and incubation. Rapidly dissected hippocampi were embedded in 2.5 % agar

30

and transverse hippocampal slices. Approximately 10 mm were discarded from both ends and

31

400 µm thick slices from each hippocampus were cut at approximately 2°C in oxygenated 9 ACS Paragon Plus Environment

ACS Chemical Neuroscience

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 27

1

HEPES-aCSF using a Vibratome® 1000plus sectioning system (The Vibratome Company, St.

2

Louis, MO). The slices were then transferred to a brain slice keeper (Scientific Systems Design

3

Inc., Mississauga, Ontario, CA) and allowed to recover for 1-1.5 h at room temperature in

4

HEPES-aCSF bubbled continuously with 95% O2/ 5% CO2 before transferring to experimental

5

conditions as indicated.

6

Microperfusion Chamber

7

To address the time course of injury, treatment was applied using an in vitro microperfusion

8

technique described previously.50 Briefly, after 1-1.5 h recovery as described above, individual

9

slices were transferred gently to microperfusion chambers and lids sealed. The 4-8 parallel

10

chambers were perfused with artificial cerebrospinal fluid (aCSF), pH 7.3 containing 120 mM

11

NaCl, 45 mM NaHCO3, 10 mM glucose, 3.3 mM KCl, 1.2 mM NaH2PO4, 2.4 mM MgSO4, 1.8

12

mM CaCl2 bubbled with 95% O2 /5% CO2 and submerged in aCSF bath at 36°C (±0.2°C) at a

13

flow rate of 7 µL/min. Sampling began 75 min after submerging the sealed chambers to allow

14

adequate time for stabilization of neurochemical efflux. At 15 min interval, perfusates (7 µL /min

15

X 15 min =105 µL) was collected and analyzed as indicated. The volume of each chamber was

16

estimated at 35µL based on dimensions of 9×5×0.7 mm (L×W×H) with 0.3-mm deep

17

microchannels. Approximately half of this volume was displaced by the 400µm thick slice.

18

Treatment / Injury Conditions

19

To model ischemia-induced alterations in the ionic microenvironment in the hippocampus, acute

20

hippocampal slices were exposed to 30 min of one of the four treatments: (1) aCSF, pH 7.3, (2)

21

OGD, pH 7.3 (glucose-oxygen free aCSF), (3) aCSF, pH 6.4 (120 mM NaCl, 7.03 mM NaHCO3,

22

10 mM glucose, 3.3 mM KCl, 1.2 mM NaH2PO4, 2.4 mM MgSO4 1.8 mM CaCl2, (4) OGD, pH

23

6.4 (glucose-oxygen free low pH aCSF). Acidosis was induced with a pH of 6.4 based on

24

previous studies that suggest hippocampal homomeric ASIC1a peak current showed a high

25

sensitivity to pH 6.4 (pH0.5 = 6.4 and nH = 1.65 for ASIC1a).65 Ischemia-induced alteration was

26

made by switching the solution (from aCSF pH 7.3 to OGD pH 7.3, aCSF, pH 6.4 and OGD, pH

27

6.4) 8 min prior to the start of insult, the time it takes to completely replace the solution in the

28

chamber with a flow rate of 7 µL/min. All aCSF solutions were equilibrated with 95% O2 and

29

5% CO2 whereas the OGD solutions were equilibrated with 5% CO2 and 95% N2, for a minimum

30

of 1 h until pH stabilized in the desired range. The osmolarity of these solutions was between

31

290 and 300 mOsm. The PO2 in OGD solution varied from 0-2.9 mmHg with an average for 6 10 ACS Paragon Plus Environment

Page 11 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

1

determinations of 1.1 mm-Hg as measured using a miniature Clark-style electrode (Instech

2

Laboratories, Plymouth Meeting, PA). Collected fractions were analyzed for cell death (LDH

3

release) on the day of collection. Remaining volume was kept at -80°C for subsequent analysis.

4

For evaluating the effect of NMDAR and ASIC1a on OGD or acidosis mediated cellular injury

5

in acute hippocampal slices, the respective inhibitors were added throughout the experiment

6

(during pretreatment, the insult phase and reperfusion). MK-801 (noncompetitive NMDAR

7

antagonist, 100 µM, Sigma-Aldrich, St Louis, MO), Amiloride (non-specific ASIC inhibitor, 100

8

µM, Cayman chemical, Ann Arbor, MI), Psalmotoxin (PcTX1 - specific ASIC1a inhibitor, 50

9

nM, Abcam, Cambridge, MA) were used for NMDAR and ASIC manipulations.

10

Assessment of Cellular Injury- LDH Measurement

11

Cell injury was assessed by the measurement of LDH release.17 Following collection of

12

perfusates every 15 min during the course of the experiment, 50 µl of perfusate was transferred

13

to 384-well plates and mixed with 50 µl reaction solution provided by the LDH assay kit

14

(Cayman chemical, Ann Arbor, MI). Optical density was measured at 490 nm 60 min later, using

15

a microplate reader (Epoch Microplate Spectrophotometer, BioTek, Winooski, VT). Background

16

absorbance at 620nm was subtracted. The LDH release was quantified as an arbitrary unit per mg

17

of protein (Bio-Rad protein assay kit, Hercules, CA) and represented as a fraction of baseline

18

LDH release. The maximal releasable LDH was also obtained by treating the slices with 1% NP-

19

40 lysis buffer (50 mM Tris-HCl, 0.5% sodium deoxycholate, 0.1% SDS, 1% Igepal, 150 mM

20

NaCl, 0.05% Triton X-100) at the end of each experiment.

21

Assessment of Metabolic Efflux

22

Glutamate and aspartate were analyzed by capillary electrophoresis. Perfusates were derivatized

23

as described previously50, 66 to form highly fluorescent cyanobenz[f]isoindole amino acids (CBI-

24

amino acids). Briefly, 2 µL thawed perfusate was reacted with 2 µL 2 mM naphthalene-2,3-

25

dicarboxaldehyde (NDA) in methanol and 2 µL NaCN (5.5 mM) in 60 mM sodium tetraborate.

26

Cyanide solution contained 3µM D-amino adipic acid as internal standard. Samples were reacted

27

at room temperature for approximately 20 min prior to analysis. CE-LIF was performed on a

28

custom in-house-built instrument using a 420-nm diode laser, 490-nm bandpass filter, and PMT

29

for LIF detection. Samples were injected onto a bare fused silica capillary of dimensions 25 cm x

30

25 µm (22.5 cm to detection window) for 1-3 sec at 380-mbar vacuum and separated by using

31

positive polarity (20 kV). Separation buffer optimized for analysis of L-glutamate and L11 ACS Paragon Plus Environment

ACS Chemical Neuroscience

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 27

1

aspartate is a 50 mM sodium tetraborate background electrolyte (BGE) adjusted to pH 9.3. BGE

2

additionally contained 1.0 mM β-cyclodextrin to decrease sample run time for glutamate and

3

aspartate analysis. The addition of β-cyclodextrin allowed for baseline resolution of glutamate,

4

aspartate, and the internal standard in less than 2 min. PeakFit software (SeaSolve Software Inc.,

5

Framingham, MA) was used to process raw data and quantify peak areas in all experiments

6

based on a Gaussian peak shape for each analyte. Linear calibration curves were constructed for

7

glutamate and aspartate as a function of concentration verses the peak area ratio (analytes

8

area/internal standard area). Efflux of analytes was quantified as a concentration of glutamate or

9

aspartate normalized to total protein for the slice used in the efflux experiment and represented

10

as a fraction of baseline glutamate or aspartate release.

11

Determination of ATP and Adenosine

12

ATP was determined after bath application of OGD or normoxic conditions as described

13

previously.48 In brief, slices were incubated in either aCSF or OGD solution and collected at the

14

time point indicated. From each treatment conditions, 3 hippocampal slices were transferred to a

15

flat-bottomed microfuge tube containing 200 µL of ice-cold 5% trichloroacetic acid, sonicated (8

16

pulses), centrifuged at 16,000 g for 1 min at 4°C. Supernatant was then collected and assayed for

17

ATP using ENLITEN ATP assay (Promega, Madison, WI). The pellets were further

18

reconstituted with 200 µL 1M NaOH and protein concentration was analyzed (Bio-Rad protein

19

assay kit, Hercules, CA). ATP level was represented as a concentration of ATP normalized to

20

total protein for the slices.

21

Adenosine level was determined using HPLC as previously reported with slight modifications. In

22

brief, 2mM acetic acid (pH 4.0) in 7% acetonitrile constitute the mobile phase having a flow rate

23

of 20 µL/min produced by a PrimeLine pump (ASI, Analytical Scientific Instruments, El

24

Sobrante, CA). The process of separation was achieved by a BAS microbore column (MF-8949;

25

1x 100 mm and 1x14mm guard column, with C18 packing of 3-µm particle size, BASi, West

26

Lafayette, IN), which was attached directly to the injector (Rheodyne 9125, Rohnert Park, CA)

27

and to the UV detector (Waters 2487 UV detector, outfitted with a Waters microbore cell kit,

28

Waters Corporation, Milford, MA). Adenosine was detected at a wavelength of 258 nm.

29

Chromatographic data were recorded with a Peak Simple Chromatography data system (SRI,

30

model 203, SRI Instruments, Torrance, CA) and the peak areas of microperfusion samples were

31

compared to the peak areas of adenosine standards (10µL of 50nM) for quantification. The 12 ACS Paragon Plus Environment

Page 13 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

1

detection limit of the assay was 33 fmol (based on a signal-to-noise ratio of 3:1). Concentration

2

is expressed as nM and retention time for adenosine was between 11.0 and 11.3min.

3

Statistical Analysis

4

A priori power analysis (using G*Power software) was performed to estimate sample size

5

needed to yield 80% power for detecting a therapeutically significant effect of treatment. No

6

animal was excluded from the study. Slices from animals were randomly assigned to treatment

7

and control groups without prior knowledge to the treatment conditions. Data processing and

8

statistical analyses were performed using Microsoft Office Excel 2010 and Graph PadPrism6.

9

Data were analyzed with one-way or two-way ANOVA with repeated measures. Significant

10

main effects or interactions were followed by t-tests with Bonferroni's correction. Data are

11

expressed as mean ± SEM, and P < 0.05 was considered as statistically significant.

12

ACKNOWLEDGEMENTS

13

Research reported in this publication was supported by the US Army Medical research

14

and Material Command, No 0517800; the National Institute of Neurological Disorder and

15

Stroke, Nos. NS041069-06 and R15NS070779; Institutional Development Award (IDeA) from

16

the National Institute of General Medical Sciences of the National Institutes of Health under

17

grant number P20GM103395. The funders had no role in study design, data collection and

18

analysis, decision to publish, or preparation of the manuscript. The authors thank Dr. Tom Kuhn

19

for helpful discussions and Ian Herriott, technician, UAF DNA core facility for assisting with the

20

microplate reader.

21

AUTHOR CONTRIBUTION STATEMENT

22

SB, JTM and KLD contributed to development of study design, performance of research,

23

writing of manuscript. DLK, MCC, EGW and BTR contributed to performance of research. KLD

24

contributed to financial support. All authors revised the manuscript for important intellectual

25

content and approved the final version.

26

DISCLOSURES

27

The author(s) declared no potential conflicts of interest with respect to the research,

28

authorship, and/or publication of this article. Figure 2 A was published in preliminary form in a

29

book chapter 67.

30

FIGURE LEGENDS 13 ACS Paragon Plus Environment

ACS Chemical Neuroscience

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 27

1

Figure 1: Integrity of acute slices is validated from observation that OGD injury causes

2

disruption in energy homeostasis. (A) Levels of whole tissue ATP were determined over a

3

course of time using bath application in normoxic slices or slices treated with OGD. Grey bar

4

indicates 30 min insult period. *p< 0.05 OGD, n= 4 versus aCSF, n= 4. Triplicate slices were

5

pooled per treatment per time point, n= 4 animals. (B) Level of adenosine efflux monitored in

6

slices subjected to 30 min aCSF or OGD using the microperfusion approach. Grey bar indicates

7

30 min insult period. *p< 0.05 represents adenosine efflux normalized to baseline (n= 7 control

8

slices, 4 animals; n= 7 OGD slices, 4 animals).

9

Figure 2. Oxygen glucose deprivation triggers metabolic efflux similar to in vivo brain

10

ischemia. Time-dependent metabolite efflux in rat hippocampal slices induced by 30 min OGD

11

insult. (A) and (B) illustrates glutamate and aspartate efflux in hippocampal slices during OGD

12

insult (n=17 slices from 6 animals) and during aCSF (aCSF, pH=7.3 - 7.4, n= 16 slices from 6

13

animals). Grey bar indicates insult period. Data shown are normalized to baseline *p