brain research 1543 (2014) 271–279

Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

Research Report

Selective vulnerability of hippocampal sub-fields to oxygen–glucose deprivation is a function of animal age Crystal C. Lalonde, John G. Mielken Neuroplasticity Research Group, School of Public Health and Health Systems, University of Waterloo, 200 University Avenue West, Waterloo, Ontario, Canada N2L 3G1

art i cle i nfo

ab st rac t

Article history:

For more than a century, the hippocampal sub-fields have been recognized as being

Accepted 30 October 2013

differentially vulnerable to injury. While the cause remains unknown, the explanations

Available online 7 November 2013

generally considered have involved either vascular differences, or innate variability among cells. To examine the latter possibility, we prepared acute hippocampal slices from

Keywords:

Sprague-Dawley rats, applied a brief period of oxygen–glucose deprivation (OGD; an

Ischemia

in vitro model of ischemia), and assessed the viability of dissected sub-fields (CA1, CA3,

Stroke

dentate gyrus) by measuring mitochondrial 2,3,5-triphenyltetrazolium chloride (TTC)

Rat

metabolism. In slices from young animals (15 weeks of age), post-OGD TTC metabolism

Hippocampus

was significantly reduced in the CA sub-fields relative to the dentate gyrus. Since previous

Glutamate receptor Triphenyltetrazolium chloride

studies found increasing age may worsen ischemic injury, we completed the same experiment using tissue from animals at 52 weeks of age, and found no differences in TTC metabolism across sub-fields. Given the established role of glutamate receptors in ischemic cell death, we examined two key subunit proteins (GluN1, found in all NMDA receptors, and GluA2, found in most AMPA receptors) across sub-fields and age to determine whether their expression complemented our viability data. We found that, relative to the CA1, the DG displayed greater GluN1 expression and lower GluA2 expression in both young and old animals. Our results confirm that regional vulnerability can be shown in a slice model, that the property is not intransigent, and that these features are likely not attributable to the expression pattern of key glutamate receptor subunits, but another molecular variable that changes over the lifespan. & 2013 Elsevier B.V. All rights reserved.

Abbreviations: ACSF, DG,

artificial cerebrospinal fluid; BSA,

dentate gyrus; DMSO,

dimethyl sulphoxide; HRP,

bovine serum albumin; CA,

horseradish peroxidase; LDH,

deprivation; PVDF, polyvinylidene fluoride; SEM, standard error of the mean; TBST, n Corresponding author. E-mail address: [email protected] (J.G. Mielke). 0006-8993/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.brainres.2013.10.056

cornu ammonis; CI,

confidence interval;

lactate dehydrogenase; OGD,

oxygen–glucose

tris-buffered saline with tween-20

272

1.

brain research 1543 (2014) 271–279

Introduction

Although the entire brain may be inadequately perfused during global ischemia, the pattern of injury that develops is not homogeneous (Pulsinelli et al., 1982). One brain region that has received considerable attention for its selective vulnerability to ischemic injury is the hippocampus, which has also been found to display damage in a heterogeneous fashion (Schmidt-Kastner and Freund, 1991). Histological analyses of hippocampal damage following models of global ischemia in both gerbils (Kirino, 1982) and rats (Pulsinelli et al., 1982; Smith et al., 1986; Schmidt-Kastner, Hossman, 1988; Kadar et al., 1998) have revealed a relatively consistent pattern: the cornu Ammonis 1 (CA1) sub-field appears to be the most vulnerable region, while the dentate gyrus (DG) appears to be the most resistant region. Although the unique susceptibility of certain areas of the hippocampus to injury was first described in the late nineteenth century (Sommer, 1880), the cause of the selective vulnerability remained a heavily debated topic for decades. The majority of discussion initially focused upon whether variation in the hippocampal vasculature might be responsible for the regional differences in injury (Spielmeyer, 1927), and then gradually gave way to the view that intrinsic differences in physicochemical properties of neurons within the hippocampus were responsible (Vogt and Vogt, 1937). Over time, the idea that the differences were attributable to innate variability in the cells comprising the hippocampus (that is, the notion of pathoclisis) began to prevail, in part due to experiments conducted with brain slice preparations. In hippocampal slices acutely prepared from adult rats, field potentials recorded from the CA1 sub-field have been consistently abolished by a length of hypoxic insult that had a less pronounced effect upon responses evoked from the CA3 sub-field (Cherubini et al., 1989) and no permanent effect upon responses evoked from the DG (Aitken and Schiff, 1986; Kass and Lipton, 1986). As well, relative to other sub-fields, the CA1 region in hippocampal slice cultures has exhibited the greatest degree of propidium iodide uptake (a commonly used assay of plasma membrane integrity that is taken as a measure of cellular viability) following a number of insults that reflect various elements of ischemic injury, including oxygen–glucose deprivation (Gee et al., 2006), NMDA-mediated excitotoxicity (Bonner et al., 2010; Butler et al., 2010; Stanika et al., 2010), proteasomal inhibition (Bonner et al., 2010), and chemicallyinduced superoxide generation (Wilde et al., 1997). Along with regional differences in the response to insult, a further general characteristic of ischemic injury that has been suggested by a number of studies is the relationship between animal age and the degree of damage. For example, the magnitude of insult following focal ischemia (as determined by infarct size, histological analysis, or neurological deficit) was consistently greater in both aged rats (Futrell et al., 1991; Yao et al., 1991; Sutherland et al., 1996; Kharlamov et al., 2000) and mice (Fuentes-Vargas et al., 2002) relative to young adult animals. As well, when hippocampal slices were exposed to either anoxia, or oxygen–glucose deprivation, tissue harvested from older rats was found to experience anoxic depolarization sooner (Roberts et al., 1990), to display a greater degree of

injury-related disruption in pH regulation (Roberts and Chih, 1997), and to show a more significant degree of plasma membrane damage (Siqueira et al., 2004) relative to slices prepared from younger animals. Given the two general trends that hippocampal tissue appears to display following ischemic injury – a tendency toward regional variation in the effect of the insult and a tendency for the degree of effect to be greater with age – the present study sought to determine how these two patterns might intersect. In particular, we wanted to examine whether differences in regional vulnerability to injury observed in younger hippocampal tissue would remain across the life span, or if there would be an increased disparity with age. To address the differences, the study employed acutely prepared hippocampal slices from rats at different ages, and then examined the effect that a period of oxygen–glucose deprivation had upon their ability to metabolize triphenyltetrazolium chloride (a commonly employed assay of mitochondrial function, Mielke et al., 2007).

2.

Results

2.1. Extracted formazan provides a sensitive, high-throughput measure for the viability of brain slices and dissected hippocampal sub-fields Given that TTC is metabolized within mitochondria, we considered the possibility that variation in the number of mitochondria present (that is, the amount of tissue analyzed) might influence our assay. As a result, we compared formazan values, alone and normalized to slice weight, collected from slices challenged with a moderate length of OGD (15 min). Relative to control slices (i.e., SHAM), the degree of change caused by OGD was similar regardless of whether slice weight was considered (percent of SHAM slices: formazan alone, 48.9%73.4%; formazan relative to slice weight, 44.1%74.3%; n¼ 3 slices from each of N¼8 animals; Fig. 1B). Since normalizing to slice weight did not seem to appreciably alter our measurement, subsequent experiments compared formazan values alone. We next determined whether our assay would be able to detect changes in slice viability that arose as a result of increasing lengths of OGD. Slices were challenged for 5, 15, or 30 min, with the expectation that formazan levels would be reduced in a fashion inversely related to OGD length. As expected, greater lengths of OGD lead to clear reductions in TTC metabolism (formazan absorbance, arbitrary units: 5 min OGD, 0.1170.015, N¼5, p¼ 0.077 relative to SHAM; 15 min OGD, 0.07870.011, N¼8, p¼ 0.0063; 30 min OGD, 0.03270.010, N ¼5, p ¼0.0007; Fig. 1C). Since a 15 min challenge provided a consistently moderate degree of reduction in formazan levels, this length of OGD was adopted as the standard for the remaining experiments. To confirm the measure of slice viability provided by the TTC assay, which assesses mitochondrial function, we completed a set of experiments where a companion measure of viability/injury was examined. Lactate dehydrogenase (LDH) is a cytoplasmic enzyme released from cells due to injuryinduced loss of cell membrane integrity, and has been readily detected in bath medium of acute hippocampal slices exposed

273

brain research 1543 (2014) 271–279

~18h

remove aliquots & measure absorbance

80 60

*

15 10 5 0

40

SHAM

OGD

20 0

Formazan Formazan ABS/ ABS slice weight 15 min OGD

SHAM

Formazan Absorbance (arbitrary units)

20

Released LDH (nmoles)

transfer slices to extraction buffer

1 h TTC wash incubation

OGD (15 min) or SHAM

remove ACSF aliquots

slices 3h in 2 mL recovery ACSF

slices in TTC solution

OGD

0.20

LDH assay

formazan extraction & measurement

0.15

*

0.10

*

0.05 0.00 5 min

15 min

30 min

Length of OGD

Fig. 1 – Metabolism of TTC allows for the sensitive, highthroughput measurement of hippocampal slice viability. (A) The schematic presents the general experimental procedure that was followed to apply oxygen–glucose deprivation (OGD) to hippocampal slices, and then to assess their viability using TTC metabolism. (B) Normalizing formazan values to slice weight does not significantly alter the magnitude of effect that OGD is observed to have upon TTC metabolism (percent of SHAM: formazan absorbance alone, 48.9%73.4%; formazan absorbance relative to slice weight, 44.1%74.3%; n¼ 3 slices, N ¼8 rats). (C) Increasing lengths of OGD cause progressively lower levels of TTC metabolism (formazan absorbance, arbitrary units: 5 min OGD, 0.1170.015, N ¼5, p ¼0.077; 15 min OGD, 0.07870.011, N¼ 8, p¼0.0063; 30 min OGD, 0.03270.010, N¼ 5, p¼ 0.0007; n significance determined relative to SHAM with one-tailed, unpaired Student's t test). Each bar represents the mean7SEM. ABS¼ absorbance.

to excitotoxic stress (Zhou and Baudry, 2006). Following treatment, and the usual recovery period, LDH content in the bathing medium was assessed, and was observed to be significantly greater in those slices challenged with OGD (released LDH, nmoles: SHAM, 9.2970.73; OGD, 14.4171.29; n¼ 3, N¼7; p¼ 0.0023; Fig. 2A). To gauge the degree of correlation between TTC metabolism and LDH efflux, each viability measure was collected from slices during several separate experiments (n¼ 3, N¼7; Fig. 2B), and the nature of their correlation examined; a significant, negative relationship was found between the two measures of slice viability (Pearson r¼  0.5906; 95% CI¼ 0.8537 to 0.08720; p¼ 0.026; Fig. 2C). Next, we assessed whether the viability of each of the major sub-fields contained within our hippocampal slices could be measured with the TTC assay. The slices were subdissected into the three key hippocampal sub-fields (DG, CA3 and CA1; Fig. 3A), and matched sub-fields from each of three slices were then pooled and examined for their degree of TTC

Released LDH (nmoles)

Percent of SHAM Value

3h OGD (5-30min) or SHAM recovery

25 20 15 10 5 0 0.00

0.05

0.10

0.15

Formazan Absorbance (arbitrary values)

Fig. 2 – Changes in TTC metabolism caused by OGD are significantly correlated with challenge-induced changes in LDH efflux. (A) A moderate length of OGD (15 min) leads to a significant increase in LDH efflux (released LDH, nmoles: SHAM, 9.2970.73; OGD, 14.471.29; n¼ 3, N¼ 7; np¼ 0.0023, one-tailed, unpaired Student's t test). Each bar represents the mean7SEM. (B) The schematic outlines the procedure used to assess both TTC metabolism and LDH efflux following OGD. (C) Formazan absorbance and LDH efflux were assessed (n¼ 3, N¼ 7), and a significant, negative relationship was found between the two measures of slice viability (Pearson r¼  0.5906; 95% CI¼  0.8537 to  0.08720; p¼0.026).

metabolism. The formazan absorbance was similar between the CA1 and DG sub-fields, but was significantly lower in the CA3 region (formazan absorbance, arbitrary units: CA1, 0.1970.023; CA3, 0.1270.011; DG, 0.2470.036; n¼ 3, N¼6; F (2, 15)¼3.960; p¼ 0.042; Fig. 3B). To determine whether the differences in TTC metabolism were related to variability in protein content among the sub-fields, formazan production and protein levels were measured in sub-fields dissected from alternating slices within each hippocampus (that is, slices 1, 3, and 5 were used for TTC metabolism, while slices 2, 4, and 6 were used for protein analysis; N¼ 4). A significant, positive relationship was observed between the degree of TTC metabolism and protein content (Pearson r¼0.6688; 95% CI¼0.2934 to 0.8655; p¼0.0024; Fig. 3C).

2.2. Hippocampal region and animal age influence post-OGD metabolism of TTC Previous work strongly suggests that the hippocampal subfields are differentially susceptible to both hypoxia and ischemia (Schmidt-Kastner and Freund, 1991). To evaluate whether

274

brain research 1543 (2014) 271–279

SHAM

Formazan Absorbance (arbitrary units)

CA3 CA1

0.25

0.15

0.15

*

0.10

# 0.05

CA1

CA3

DG

Hippocampal Sub-Field

*

0.10 0.05 0.00

CA1

CA3

DG

Hippocampal Sub-Field 2.5

Protein Concentration (µg/µL)

ns

0.20

0.00

0.20

Formazan Absorbance (% SHAM)

Formazan Absorbance (arbitrary units)

DG

OGD

0.25

2.0 1.5

100

*

80 60 40 20 0

CA1

CA3

DG

Hippocampal Sub-Field

1.0 0.5 0.0 0.0

0.1

0.2

0.3

0.4

Formazan Absorbance (arbitrary units) Fig. 3 – The viability of hippocampal sub-fields can be assessed by measuring TTC metabolism. (A) The schematic illustrates how the three hippocampal sub-fields were dissected from each slice. (B) Upon dissection, matched subfields from each of 3 slices were pooled and the degree of TTC metabolism assessed. Formazan absorbance was similar between the CA1 and DG sub-fields, but was significantly lower in the CA3 region (formazan absorbance, arbitrary units: CA1, 0.1970.023; CA3, 0.1270.011; DG, 0.2470.036; n ¼3, N¼ 6; F (2, 15) ¼3.960; np ¼0.042, one-way ANOVA followed by Tukey's post-hoc test). Each bar represents the mean7SEM. (C) The degree of TTC metabolism and protein content in hippocampal sub-fields were assessed (n ¼ 3, N ¼6), and a significant, positive relationship was found (Pearson r¼ 0.6688; 95% CI¼ 0.2934 to 0.8655; p ¼0.0024).

OGD-induced changes in TTC metabolism would differ across sub-fields, slices were challenged with a 15 min insult, allowed to recover for 3 h, and then sub-dissected prior to placement in TTC solution (the method for sub-dissection follows Coultrap et al. and is illustrated in Fig. 3A). In agreement with the trend described in earlier studies, the effect of OGD was greatest in the CA sub-fields (formazan absorbance, percent of SHAM: CA1, 54.1%73.3%; CA3, 56.1%78.5%; DG, 80.0%73.0%; n¼ 3, N¼ 6; F (2, 15)¼7.255; p¼0.006; Fig. 4).

Fig. 4 – Hippocampal sub-fields display a heterogeneous reduction in TTC metabolism following OGD. (A) Formazan absorbance was significantly reduced in each CA sub-field following 15 min OGD and 3 h recovery, but was not altered to a significant degree in the DG (np¼ 0.0024, #p¼ 0.0046, ns p ¼0.24; one-tailed, unpaired Student's t test; n ¼3, N¼ 6). (B) The graph presents the degree of change seen among the three sub-fields following the insult and recovery (formazan absorbance, percent of SHAM: CA1, 54.1%73.3%; CA3, 56.1% 78.5%; DG, 80.0%73.0%; n ¼3, N¼ 6; F (2, 15)¼ 7.255; n p¼ 0.006, one-way ANOVA followed by Tukey's post-hoc test). Each graph presents the mean7SEM. CA¼cornu ammonis, DG¼ dentate gyrus, ns¼ non-significant.

A number of pre-clinical studies have suggested that animal age may influence outcome from ischemia-related injury (Yager and Thornhill, 1997). Given that regional variation in the effects of OGD was observed within slices from younger animals, we wanted to determine whether the difference among hippocampal sub-fields would be the same when slices were prepared from older animals. We found that the greater post-OGD TTC metabolism observed in the DG from younger animals was absent in older animals (DG formazan absorbance, percent of SHAM: mean animal age¼ 15 wks, 80.0%73.0%, N¼6; mean animal age¼ 52 wks, 48.6%76.8%, N¼ 6; p¼0.003; Fig. 5).

2.3. Relative sub-field expression of key glutamate receptor subunits does not change with age The excessive activation of glutamate receptors has been clearly identified as one of the earliest points in the cascade of events leading to ischemic cell death (Lai et al., 2011); in

275

brain research 1543 (2014) 271–279

x age = 15 wk

x age= 52 wks

duplicate

Formazan Absorbance (%SHAM)

100 80

Ponceau S

60

*

40

GluN1 20 0 CA1

CA3

DG

GluA2

Hippocampal Sub-Field

particular, the NMDA sub-type of glutamate receptor has received considerable attention that is largely attributable to its ability to permit Ca2þ influx (Traynelis et al., 2010). As a result, we considered the possibility that the enhanced vulnerability of the CA1 region in younger animals was attributable to a greater expression of NMDA receptors. Since all functional NMDA receptors possess the GluN1 subunit (Forrest et al., 1994), we decided to measure its expression across the different hippocampal sub-fields at points in the lifespan that coincided with our OGD experiments. We observed that the DG sub-field in slices taken from younger animals had greater levels of GluN1, and that the relatively greater amount of the subunit was also seen in slices from older animals (GluN1 expression, percent of age-matched CA1 sub-field: young-DG, 147.4% 729.0%; old-DG, 148.5%718.8%; N¼5; Fig. 6). The AMPA receptor, another sub-type of glutamate receptor, mediates the vast majority of fast excitatory transmission within the mammalian CNS (Isaac et al., 2007), and has also been shown to play a critical role in the process of ischemic cell death (Liu and Zukin, 2007). Since the GluA2 subunit is found in a very large majority of AMPA receptors and determines many of the receptor's properties (Isaac et al., 2007), we also chose to examine its expression in our slices. In contrast to the pattern observed with the GluN1 subunit, we found that the DG sub-field in slices taken from younger animals had lower levels of GluA2, and that the lower relative level remained unchanged between the two ages examined (GluA2 expression, percent of age-matched CA1 sub-field: young-DG, 61.9%715.9%; old-DG, 67.8%716.5%; N ¼5; Fig. 6).

3.

Discussion

The different susceptibility of hippocampal sub-fields to injury has been recognized for more than a century, but the cellular features that account for the variability remain unclear. To build upon earlier work in the area, we wanted to confirm that regional variability following oxygen–glucose

Young

Old

duplicate

200

Percent of Age-Matched CA1 Sub-Field

Fig. 5 – Animal age influences the effect that OGD has upon TTC metabolism across hippocampal sub-fields. Among sub-fields, only the DG displayed a clear effect of age on post- OGD TTC metabolism (formazan absorbance, percent of SHAM: 15 wks, 80.0%73.0%, N¼ 6; 52 wks, 48.6%76.8%, N¼ 6; np¼ 0.003, two-tailed, unpaired Student's t test). The graph presents the mean7SEM. CA¼cornu ammonis, DG¼ dentate gyrus.

150

100

50

0

Y-DG

O-DG

GluN1

Y-DG

O-DG

GluA2

Fig. 6 – Glutamate receptor subunit proteins are differentially expressed across sub-fields, and the difference remains consistent with age. (A) Representative images of immunoblots wherein 10 lg of homogenates from either the CA1, or DG sub-fields (dissected from each of 3 hippocampal slices per animal) were loaded in duplicate. To control for loading errors, optical densities from the immunolabelling were normalized to selected bands from Ponceau S staining (denoted by the boxed area). The arrow indicates the band that was used for densitometry of GluA2 immunolabelling. Average age of young (Y) animals¼15 weeks; average age of old (O) animals¼52 weeks. (B) The graph presents the amount of either GluN1, or GluA2 protein found in the DG relative to the CA1 region at two points across the lifespan (percent of age-matched CA1: GluN1 Y-DG, 147.4%729.0%; GluN1 O-DG, 148.5%718.8%; GluA2 Y-DG, 61.9%715.9%; GluA2 O-DG, 67.8%716.5%). Each bar presents the mean7SEM of N¼5 animals. CA¼ cornu ammonis, DG¼dentate gyrus.

deprivation could be detected in hippocampal slices, a model wherein an equivalent degree of ischemic-like stress could be reasonably applied to different neuronal populations. While our findings do indicate that hippocampal sub-fields vary in their ability to metabolize TTC to formazan following OGD, this pattern changes with animal age; in particular, a difference was present in slices harvested from young adults, but was absent in slices harvested from mature adults. Studies using both animal (Isayama et al., 1991; Watson et al., 1994; Preston and Webster, 2000) and hippocampal slice

276

brain research 1543 (2014) 271–279

(Watson et al., 1994; Mielke et al., 2007) models of ischemia have shown that changes in the level of TTC reduction by the mitochondrial electron transport chain can be used to measure insult severity. An earlier report by our group illustrated that the effects of OGD on TTC metabolism and synaptic transmission were positively correlated (Mielke et al., 2007). As an extension of our earlier work, we examined the relationship between post-OGD TTC metabolism and the loss of plasma membrane integrity (a frequently examined effect of cellular injury) within the same group of slices. Following a brief period of OGD, a clear inverse relationship could be seen between the two measures (that is, a reduced level of TTC metabolism was accompanied by an increased level of released LDH). In agreement with our previous study, the present result supports the use of TTC metabolism as a sensitive and high-throughput method to assess slice viability; as well, the observations suggest that, within this particular model system, loss of membrane integrity and impairment of mitochondrial function begin to develop along a similar time course. As noted earlier, a substantial body of literature has illustrated that the various regions of the hippocampus respond to ischemic injury in a heterogeneous fashion. For example, two of the first reports in the area suggested that transient ischemia in either the gerbil (Kirino, 1982), or the rat (Pulsinelli et al., 1982) caused the greatest degree of damage in the CA region, particularly CA1. Subsequent studies that employed different models of ischemia in the rat consistently showed that, relative to the DG, the CA1 sub-field displayed a greater degree of progressive damage (Smith et al., 1986; SchmidtKastner and Hossman, 1988; Schmidt-Kastner and Freund, 1991; Kadar et al., 1998). As well, a similar pattern of regional susceptibility to injury has been found following various degrees of hypoxia in acutely prepared hippocampal slices (Aitken and Schiff, 1986; Ashton et al., 1989; Kass and Lipton, 1986). In agreement with the trend shown in previous work, we found that a brief period of OGD reduced mitochondrial TTC metabolism nearly two and a half times more in the CA1 region than in the DG. A number of studies have found that the way in which the brain responds to ischemia changes over the lifespan (Yager and Thornhill, 1997; Schaller, 2007). As well, differences in response to either OGD, or anoxia have been observed between hippocampal slices prepared from animals of different ages; in particular, slices from older animals were found to undergo a greater loss of membrane integrity (Siqueira et al., 2004), a longer lasting rise in the level of cytosolic Ca2þ (Tonkikh and Carlen, 2009), and a more significant impairment of synaptic transmission (Roberts et al., 1990, 1998). As a result, we considered the possibility that the resilience of the DG might lessen when slices were taken from older animals. Notably, the DG in slices from animals that were about one year of age showed a change in TTC metabolism similar to that observed in other hippocampal regions, which suggests that the injury tolerance displayed by the sub-field is not an intransigent property. One of the earliest explanations proposed for regional variability following injury was that the CA1 area experienced a more profound loss of ATP; however, significant differences in ATP levels across sub-fields were not apparent after either global

ischemia in the gerbil (Munekata and Hossmann, 1986), or anoxia in the rat hippocampal slice (Kass and Lipton, 1986). As a result, alternative explanations began to be explored, and the one that, arguably, received the most attention was variation in the degree to which injury mediated a change in the concentration of intracellular Ca2þ. In support of this possibility, the greatest change in cytosolic Ca2þ observed in slice models of anoxia (Kass and Lipton, 1986), OGD (Mitani et al., 1993; Shimizu et al., 1996; Yamashima et al., 1996), and NMDA (Stanika et al., 2010) was shown to occur within the CA1 sub-field. Although the definitive cause of the elevated internal Ca2þ level remains unclear, evidence has suggested that either reduced extrusion mechanisms (Kass and Lipton, 1986), or increased release from internal stores (Mitani et al., 1993) may be partly responsible. Given that activation of the NMDA sub-type of glutamate receptor permits Ca2þ influx (Traynelis et al., 2010), and that excessive activation of glutamate receptors is a critical upstream event in the ischemic cell death cascade (Lai et al., 2011), we considered the possibility that the change in regional vulnerability over the lifespan was attributable to age-related differences in the expression of NMDA receptors. Since all functional NMDA receptors possess the GluN1 subunit (Forrest et al., 1994), we decided to measure its expression across the different hippocampal sub-fields in younger versus older animals. Given the fashion in which OGD affected TTC metabolism across the sub-fields, we had expected to find less of the GluN1 subunit in the DG of younger animals and a similar amount across subfields in older animals; however, we were surprised to find that GluN1 protein expression in the DG was greater at both of the ages examined. Notably, our observations contrast with another report that found no difference in GluN1 expression between CA1 and DG (Coultrap et al., 2005); although an explanation for the discordant findings is not apparent, one possible explanation may rest in the difference in ages examined across the two studies (animals were 6–9 weeks of age in the Coultrap et al. study, while our animals were approximately 15 and 52 weeks of age). A further point that should be considered is that alternative splicing of the GluN1 gene leads to the production of eight possible splice variants, which can convey unique properties upon the NMDA receptors that they help to form (Traynelis et al., 2010). Given that the splice variants have been shown to differentially contribute to excitotoxicity (albeit in a cell line model; Rameau et al., 2000), there exists the possibility that overall levels of the GluN1 subunit might have been higher in the DG, but that levels of variants that would render the receptor more likely to cause ischemic injury were greater in the CA1 region. Along with the NMDA receptor, another sub-type of glutamate receptor, the AMPA receptor, has also been shown to play a critical role in the process of ischemic cell death (Liu and Zukin, 2007) as a result, we considered the alternate possibility that age-related differences in the expression of AMPA receptors may have contributed to changes in regional vulnerability over the lifespan. Given that a very large proportion of AMPA receptors contain the GluA2 subunit, which influences many of the biophysical properties of the receptor (Isaac et al., 2007), we decided to measure its expression. Relative to the CA1 sub-field, levels of the GluA2 subunit in the DG were noticeably reduced in younger animals, which agrees with the pattern shown by Coultrap

brain research 1543 (2014) 271–279

et al. (2005) with an antibody directed against GluA2/3. Since AMPA receptor activation permits the change in membrane depolarization needed to remove the Mg2þ blockade of the NMDA receptor pore, the reduced amount of GluA2 in the DG may have provided cells in the region with an indirect means to limit NMDA receptor activation. However, as with the GluN1 subunit, there was no change in the expression of the GluA2 subunit observed between the two ages studied; as a result, some other cellular change is likely responsible for normalizing the vulnerability of the sub-fields with age. The microcirculation of the CA1 region is such that the area is supplied with less blood than other regions of the forebrain, and that there is less blood flow to the area upon reperfusion (Imdahl and Hossmann, 1986); taken together, both of these features provide support for Spielmeyer's assertion that regional susceptibility within the hippocampus is attributable to vascular factors. However, a number of studies (Aitken and Schiff, 1986; Kass and Lipton, 1986; Cherubini et al., 1989), along with the present report, have indicated that innate differences among hippocampal cell populations may also contribute to regional variability. In addition to providing further evidence for the notion of pathoclisis, our study also shows that the differences may not be stable across the lifespan; notably, a loss of regional variability within the hippocampus may be an additional feature associated with brain aging. Although the current study was not able to identify the molecular variable underlying the age-dependent shift in how the sub-fields responded to injury, given that certain splice variants of the GluN1 subunit might alter sensitivity to excitotoxicity (Rameau et al., 2000) and that the expression of certain GluN1 splice variants does change with age (Magnusson, 2012), future work should closely examine regional and age-related changes in their expression.

4.

Experimental procedures

4.1.

Preparation of acute hippocampal slices

Male Sprague-Dawley rats were anesthetized with CO2 and decapitated in accordance with procedures approved by the University of Waterloo animal care committee. Brains were rapidly removed ( 60 s) and immediately placed in cooled (o4 1C) artificial cerebrospinal fluid (ACSF) containing (in mM): 127.0 NaCl (Sigma, Oakville, ON, Canada; all subsequent reagents from Sigma, unless otherwise noted), 26.0 NaHCO3, 10.0 glucose, 2.0 CaCl2, 2.0 KCl, 2.0 MgSO4, 1.2 KH2PO4 and equilibrated with carbogen (95% O2/5% CO2), pH 7.37–7.43, 310– 320 mOsm. Depending upon the experiment, one or two hippocampi were removed from each hemisphere and placed on the platform of a McIlwain Tissue chopper (Mickle Laboratory Engineering Co., Surrey, UK); after this, 350 μm thick slices were cut and placed in turn upon mesh platforms rested within separate compartments of an interface incubation chamber (2–4 slices per platform). The ACSF was continuously gassed with carbogen, the incubation chamber was kept at 35.070.5 1C, and 60–90 min were allowed prior to the beginning of experiments.

4.2.

277

Oxygen–glucose deprivation

Oxygen–glucose deprivation (OGD) was induced by transferring platforms containing slices to a separate incubation chamber filled with ACSF (kept at 35.070.5 1C) in which sucrose had been substituted for glucose, and which had been saturated with 95% N2/5% CO2. Following OGD of varying lengths, slice platforms were transferred back to their original compartment in the control incubation chamber, and a 3 h recovery period allowed.

4.3.

2,3,5-Triphenyltetrazolium chloride (TTC) assay

Following recovery, slices were transferred to small glass vials containing a 2% (w/v) TTC solution in ACSF aerated vigorously for 10–15 min with anoxic gas (Mielke et al., 2007), and incubated for 1 h at 3570.5 1C. After this, slices were washed twice with carbogenated ACSF before being placed in extraction buffer (DMSO and 100% ethanol prepared 1:1) overnight (approximately 18 h) at room temperature in the dark. The amount of extracted formazan was then measured via spectrophotometry at 485 nm. In a subset of experiments, slices were subsequently placed overnight in an oven at 60 1C prior to being weighed.

4.4.

Lactate dehydrogenase (LDH) release assay

A subset of slices was transferred to glass vials containing 2 mL carbogenated ACSF, and maintained at 3570.5 1C for 3 h after OGD (Fig. 2B). For these slices, TTC was assessed as described, while LDH release was measured using an established fluorometric assay (Chi et al., 1983), with slight modification for use with neural tissue. Aliquots (1 mL) of ACSF were removed from each vial and diluted 1:2 with a solution containing 200 mM Imidazole-HCl and 0.1% (w/v) bovine serum albumin (BSA; pH 7.0). The diluted sample was then combined with the first reaction reagent, which contained (in mM): 100 Imidazole-HCl, 2.0 sodium pyruvate, 0.3 NADH, and 0.05% (w/v) BSA (pH 7.0). After 1 h at room temperature (protected from light), the reaction was stopped by the addition of 1.0 N HCl, and the mixture allowed to incubate at room temperature for another 10 min. Next, 6.0 M NaOH was added to samples, which were then incubated at 60 1C for 20 min. Fluorescence of sample aliquots was measured using an RF 1501, Version 2 fluorometer with excitation at 365 nm and emission at 455 nm.

4.5.

SDS-PAGE and immunoblotting

Sample homogenates were loaded in duplicate (10 μg protein/ lane), resolved using SDS-PAGE, and transferred to PVDF membranes (Bio-Rad Laboratories, Mississauga, ON, Canada). Membranes were blocked for 1 h with 5% skim milk powder (w/v) prepared in TBST [20 mM Tris, 140 mM NaCl, 0.1% Tween-20 (v/v), pH 7.6], and immunoblotting performed by overnight incubation (4 1C) with anti-GluN1 (1:1000, mouse monoclonal, #05-432; Millipore, Billerica, MA, USA) or antiGluA2 (1:1000, mouse monoclonal, MAB397; Millipore) antibodies as appropriate. Membranes were washed and placed into an HRP-linked anti-mouse secondary antibody (1:5000)

278

brain research 1543 (2014) 271–279

for 1 h. All antibody solutions were prepared in blocking buffer. Bands were visualized using enhanced chemiluminescence reagents and the ChemiGenius 2 Bio-Imaging System (Syngene, Frederick, MD, USA). The approximate molecular weight for each protein was estimated using Precision Plus Protein WesternC Standards and Precision Protein StrepTactin HRP Conjugate (Bio-Rad Laboratories). The relative density of each band of interest, from within the linear range of exposures, was measured, background-subtracted, and normalized to the average density of selected bands observed with Ponceau S staining. As appropriate, membranes were placed into stripping buffer (ReBlot Plus; Millipore) for 15–20 min at room temperature with constant agitation before being re-probed.

4.6.

Statistical analyses

All data in figures are presented as mean7SEM. Either unpaired Student's t tests, or one-way ANOVA followed by the Tukey's test was used to examine differences between groups, with significance set at po0.05. Correlational analyses were performed using Pearson's correlation coefficient with significance set at po0.05.

Acknowledgments The work reported in this article was generously supported through a Discovery Grant awarded by the Natural Sciences and Engineering Research Council (Canada) to JGM.

references

Aitken, P., Schiff, S., 1986. Selective vulnerability to hypoxia in vitro. Neurosci. Lett. 67, 92–96. Ashton, D., Willems, R., Wauquier, A., 1989. Selective vulnerability in hippocampus to ex-vivo ischemia: effects of extracellular ionic substitution in the postischemic period. Brain Res. 487, 402–406. Bonner, H., Concannon, C., Bonner, C., Woods, I., Ward, M., Prehn, J., 2010. Differential expression patterns of Puma and Hsp70 following proteasomal stress in the hippocampus are key determinants of neuronal vulnerability. J. Neurochem. 114, 606–616. Butler, T., Self, R., Smith, K., Sharrett-Field, L., Berry, J., Littleton, J., Pauly, J., Mulholland, P., Prendergast, M., 2010. Selective vulnerability of hippocampal cornu ammonis 1 pyramidal cells to excitotoxic insult is associated with the expression of polyamine-sensitive N-methyl-D-aspartate-type glutamate receptors. Neuroscience 165, 525–534. Cherubini, E., Ben-Ari, Y., Krnjevic, K., 1989. Anoxia produces smaller changes in synaptic transmission, membrane potential, and input resistance in immature rat hippocampus. J. Neurophys. 62 (4), 882–895. Chi, M., Hintz, C., Coyle, E., Martin, W., Ivy, J., Nemeth, P., Holloszy, J., Lowry, O., 1983. Effects of detraining on enzymes of energy metabolism in individual human muscle fibers Am. J. Physiol 244, C276–C287. Coultrap, S., Nixon, K., Alvestad, R., Valenzuela, C., Browning, M., 2005. Differential expression of NMDA receptor subunits and splice variants among the CA1, CA3, and dentate gyrus of the adult rat. Mol. Brain Res. 135, 104–111.

Forrest, D., Yuzaki, M., Soares, H., Ng, L., Luk, D., Sheng, M., Stewart, C., Morgan, J., Connor, J., Curran, T., 1994. Targeted disruption of NMDA receptor 1 gene abolishes NMDA response and results in neonatal death. Neuron 13 (2), 325–338. Fuentes-Vargas, M., Santiago-Mejia, J., Pinzon, E., Rodriguez, R., 2002. Age-related susceptibility to brain ischemia in mice. Drug. Dev. Res. 57, 161–166. Futrell, N., Garcia, J., Peterson, E., Millikan, C., 1991. Embolic stroke in aging rats. Stroke 22, 1582–1591. Gee, C., Benquet, P., Raineteau, O., Rietschin, L., Kirbach, S., Gerber, U., 2006. NMDA receptors and the differential ischemic vulnerability of hippocampal neurons. Eur. J. Neurosci. 23, 2595–2603. Isayama, K., Pitts, L., Nishimura, M., 1991. Evaluation of 2,3,5triphenyltetrazolium chloride staining to delineate rat brain infarcts. Stroke 22 (11), 1394–1398. Imdahl, A., Hossmann, K., 1986. Morphometric evaluation of postischemic capillary perfusion in selectively vulnerable areas of gerbil brain. Acta. Neuropath. 69, 267–271. Isaac, J., Ashby, M., McBain, C., 2007. The role of the GluR2 subunit in AMPA receptor function and synaptic plasticity. Neuron 54, 859–871. Kadar, T., Dachir, S., Shukitt-Hale, B., Levy, A., 1998. Sub-regional hippocampal vulnerability in various animal models leading to cognitive dysfunction. J. Neural Transm. 105, 987–1004. Kass, I., Lipton, P., 1986. Calcium and long-term transmission damage following anoxia in dentate gyrus and CA1 regions of the rat hippocampal slice. J. Physiol. 378, 313–334. Kharlamov, A., Kharlamov, E., Armstrong, D., 2000. Agedependent increase in infarct volume following photochemically induced cerebral infarction: putative role of astroglia. J. Gerontol A Biol. Sci. Med. Sci. 55 (3), B135–B141. Kirino, T., 1982. Delayed neuronal death in the gerbil hippocampus following ischemia. Brain Res. 239, 57–69. Lai, T., Shyu, W., Wang, Y., 2011. Stroke intervention pathways: NMDA receptors and beyond. Trends Mol. Med. 17 (5), 266–275. Liu, S., Zukin, R., 2007. Ca2þ-permeable AMPA receptors in synaptic plasticity and neuronal death. Trends Neurosci. 30 (3), 126–134. Magnusson, K., 2012. Aging of the NMDA receptor: from a mouse's point of view. Future Neurol. 7 (5), 627–637. Mielke, J., Comas, T., Ahuja, T., Preston, E., Mealing, G., 2007. Synaptic activity and triphenyltetrazolium chloride metabolism are correlated after oxygen–glucose deprivation in acute, but not cultured, hippocampal slices. Brain Res. 1176, 113–123. Mitani, A., Yanase, H., Sakai, K., Wake, Y., Kataoka, K., 1993. Origin of intracellular Ca2þ elevation induced by in vitro ischemia-like condition in hippocampal slices. Brain Res. 601, 103–110. Munekata, K., Hossmann, K., 1986. Effect of 5-minute ischemia on regional pH and energy state of the gerbil brain: relation to selective vulnerability of the hippocampus. Stroke 18, 412–417. Pulsinelli, W., Brierley, J., Plum, F., 1982. Temporal profile of neuronal damage in a model of transient forebrain ischemia. Ann. Neurol. 11, 491–498. Rameau, G., Akaneya, Y., Chiu, L., Ziff, E., 2000. Role of NMDA receptor functional domains in excitatory cell death. Neuropharmacology 39, 2255–2266. Roberts, E., Jr., Chih, C., 1997. The influence of age on pH regulation in hippocampal slices prepared before, during, and after anoxia. J. Cereb. Blood Flow Metab 17 (5), 560–566. Roberts Jr., E., Rosenthal, M., Sick, T., 1990. Age-related modifications of potassium homeostasis and synaptic transmission during and after anoxia in rat hippocampal slices. Brain Res. 514, 111–118.

brain research 1543 (2014) 271–279

Roberts Jr., E., Wisotzky, D., Chih, C., 1998. Aging and the effects of MK-801 on anoxic damage in rat hippocampal slices. Brain Res. 791, 321–324. Schaller, B., 2007. Influence of age on stroke and preconditioninginduced ischemic tolerance in the brain. Exp. Neurol. 205, 9–19. Schmidt-Kastner, R., Hossman, K., 1988. Distribution of ischemic neuronal damage in the dorsal hippocampus of rat. Acta. Neuropathol. 76, 411–421. Schmidt-Kastner, R., Freund, T., 1991. Selective vulnerability of the hippocampus in brain ischemia. Neuroscience 40 (3), 599–636. Shimizu, H., Mizuguchi, A., Aoki, M., 1996. Differential responses between CA1 pyramidal cells and granule cells to ischemic insult in rat hippocampal slices. Neurosci. Lett. 203, 195–198. Siqueira, I., Cimarosti, H., Fochesatto, C., Salbego, C., Netto, C., 2004. Age-related susceptibility to oxygen and glucose deprivation damage in rat hippocampal slices. Brain Res. 1025, 226–230. Smith, M., Auer, R., Siesjo, B., 1986. The density and distribution of ischemic brain injury in the rat following 2–10 min forebrain ischemia. Acta Neuropathol. 64, 319–332. Sommer, W., 1880. Erkrankung des Ammonshorns als aetiologisches Moment der Epilepsie. Arch. Psychiatr Nervenkr. 10, 631–675. Spielmeyer, W., 1927. Die Pathogenese des epileptischen Krampfes. Z Dtsch. Ges. Neurol Psychiatr 109, 501–520. Stanika, R., Winters, C., Pivovarova, N., Andrews, S., 2010. Differential NMDA receptor-dependent calcium loading and mitochondrial dysfunction in CA1 vs. CA3 hippocampal neurons. Neurobiol. Dis. 37, 403–411. Sutherland, G., Dix, G., Auer, R., 1996. Effect of age in rodent models of focal and forebrain ischemia. Stroke 27 (9), 1663–1667. Tonkikh, A., Carlen, P., 2009. Impaired presynaptic cytosolic calcium and mitochondrial calcium dynamics in aged

279

compared to young adult hippocampal CA1 synapses ameliorated by calcium chelation. Neuroscience 159 (4), 1300–1308. Traynelis, S., Wollmuth, L., McBain, C., Menniti, F., Vance, K., Ogden, K., Hansen, K., Yuan, H., Myers, S., Dingledine, R., 2010. Glutamate receptor ion channels: structure, regulation, and function. Pharmacol. Rev. 62 (3), 405–496. Vogt, C., Vogt, O., 1937. Sitz und Wesen der Krankheiten im Lichte der topistischen Hirnforschung und des Variierens der Tiere. J. Psychol. Neurol. 47, 237–457. Watson, G., Lopez, O., Charles, V., Lanthorn, T., 1994. Assessment of long-term effects of transient anoxia on metabolic activity of rat hippocampal slices using triphenyltetrazolium chloride. J. Neurosci. Meth 53, 203–208. Wilde, G., Pringle, A., Wright, P., Iannotti, F., 1997. Differential vulnerability of the CA1 and CA3 subfields of the hippocampus to superoxide and hydroxyl radicals in vitro. J. Neurochem. 69, 883–886. Yager, J., Thornhill, J., 1997. The effect of age on susceptibility to hypoxic-ischemic brain damage. Neurosci. Biobehav. Rev. 21, 167–174. Yamashima, T., Saido, T., Takita, M., Miyazawa, A., Yamano, J., Miyakawa, A., Nishijyo, H., Yamashita, J., Kawashima, S., Ono, T., Yoshioka, T., 1996. Transient brain ischaemia provokes Ca2þ, PIP2, and calpain responses prior to delayed neuronal death in monkeys. Eur. J. Neurosci. 8, 1932–1944. Yao, H., Sadoshima, S., Ooboshi, H., Sato, Y., Uchimura, H., Fujishima, M., 1991. Age-related vulnerability to cerebral ischemia in spontaneously hypertensive rats. Stroke 22, 1414–1418. Zhou, M., Baudry, M., 2006. Developmental changes in NMDA neurotoxicity reflect developmental changes in subunit composition of NMDA receptors. J. Neurosci. 26 (11), 2956–2963.

Selective vulnerability of hippocampal sub-fields to oxygen-glucose deprivation is a function of animal age.

For more than a century, the hippocampal sub-fields have been recognized as being differentially vulnerable to injury. While the cause remains unknown...
842KB Sizes 0 Downloads 0 Views