Endogenous brain protection: What the cerebral transcriptome teaches us

brain research 1564 (2014) 85–100

Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

Review

Endogenous brain protection: What the cerebral transcriptome teaches us K.E.M. Cox-Limpensa,b,n, A.W.D. Gavilanesb, L.J.I. Zimmermannb, J.S.H. Vlesc a

School for Mental Health and Neuroscience (MHeNS), Maastricht University, Universiteitssingel 50, 6200 MD Maastricht, The Netherlands b Department of Pediatrics, Maastricht University Medical Center (MUMC), postbus 5800, 6202 AZ Maastricht, The Netherlands c Department of Pediatric Neurology, Maastricht University Medical Center (MUMC), P.Debyelaan 25, 6229 HX Maastricht, The Netherlands

art i cle i nfo

ab st rac t

Article history:

Despite efforts to reduce mortality caused by stroke and perinatal asphyxia, these are still

Accepted 1 April 2014

the 2nd largest cause of death worldwide in the age groups they affect. Furthermore,

Available online 5 April 2014

survivors of cerebral hypoxia-ischemia often suffer neurological morbidities. A better understanding of pathophysiological mechanisms in focal and global brain ischemia will

Keywords:

contribute to the development of tailored therapeutic strategies. Similarly, insight into

Stroke

molecular pathways involved in preconditioning-induced brain protection will provide

Perinatal asphyxia Hypoxic/ischemic preconditioning Brain protection

possibilities for future treatment. Microarray technology is a great tool for investigating large scale gene expression, and has been used in many experimental studies of cerebral ischemia and preconditioning

Microarray analysis Whole-genome gene expression

to unravel molecular (patho-) physiology. However, the amount of data across microarray studies can be daunting and hard to interpret which is why we aim to provide a clear overview of available data in experimental rodent models. Findings for both injurious ischemia and preconditioning are reviewed under separate subtopics such as cellular stress, inflammation, cytoskeleton and cell signaling. Finally, we investigated the transcriptome signature of brain protection across preconditioning studies in search of transcripts that were expressed similarly across studies. Strikingly, when comparing genes discovered by single-gene analysis we observed only 15 genes present in two studies or more. We subjected these 15 transcripts to DAVID Annotation Clustering analysis to derive their shared biological meaning. Interestingly, the MAPK signaling pathway and more specifically the ERK1/2 pathway geared toward cell survival/proliferation was significantly

Abbreviations: MCAO,

middle cerebral artery occlusion; ECA,

external carotid artery; CCAO,

common carotid artery occlusion;

BCAO, bilateral common carotid occlusion; DHT, dihydrotestosterone n Corresponding author at: School for Mental Health and Neuroscience (MHeNS), Maastricht University, Universiteitssingel 50, Room 1.152, 6200 MD Maastricht, The Netherlands. Fax: þ31 433876061. E-mail addresses: [email protected], [email protected] (K.E.M. Cox-Limpens), [email protected] (A.W.D. Gavilanes), [email protected] (L.J.I. Zimmermann), [email protected] (J.S.H. Vles). http://dx.doi.org/10.1016/j.brainres.2014.04.001 0006-8993/& 2014 Elsevier B.V. All rights reserved.

86

brain research 1564 (2014) 85–100

enriched. To conclude, we advocate incorporating pathway analysis into all microarray data analysis in order to improve the detection of similarities between independently derived datasets. & 2014 Elsevier B.V. All rights reserved.

Contents 1. 2.

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Focal and global hypoxia-ischemia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Immediate early genes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Stress response genes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. Apoptosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4. Signal transduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5. Neurotransmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.6. Ion channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.7. Inflammation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.8. Cytoskeleton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.9. Ribosome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.10. Neurotrophic factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Preconditioning of the brain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Immediate early Genes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Stress response genes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3. Apoptosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4. Signal transduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5. Neurotransmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6. Ion channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.7. Inflammation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.8. Cytoskeleton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.9. Ribosome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.10. Neuroptrophic factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Focal versus global injurious cerebral ischemia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. From bench to bedside . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3. Is there a transcriptome signature for cerebral preconditioning? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4. Limitations of microarray research. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5. Sex differences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.6. Epigenetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1.

86 87 87 87 89 89 90 90 90 90 91 91 91 92 92 92 92 93 93 93 93 93 94 94 94 94 94 96 96 97 97 97

Introduction

Hypoxic-ischemic brain injury is still a major cause of death in both neonates and adults worldwide. Despite efforts in the last decade to reduce mortality caused by stroke it remains the 2nd largest cause of death worldwide, and in 2011 stroke was responsible for 6.2 million adult deaths. The yearly number of deaths caused by stroke has even grown with 600,000 since the year 2000 (World Health Organization, 2011). In addition, survivors often suffer significant neurological disabilities (Kelly-Hayes et al., 2003; Mukherjee and Patil, 2011). Perinatal asphyxia is a different type of hypoxic-ischemic brain injury affecting neonates. It is the 2nd largest cause of death among neonates and causes over 20% of neonatal deaths (World Health Organization, 2010). Survivors of perinatal asphyxia are at risk of developing hypoxic-ischemic

encephalopathy, which often leads to permanent neurological disabilities such as cerebral palsy (Perez et al., 2013). Both stroke and perinatal asphyxia have high mortality and high neurological morbidity in the age groups they affect. Although these conditions differ in etiology, for example stroke is a local hypoxic-ischemic event whereas perinatal asphyxia is a global hypoxic-ischemic event; it is likely that there are some common denominators in the mechanism of injury. In order to limit mortality and neurological morbidities caused by hypoxic-ischemic brain damage we need a better understanding of the injurious mechanisms. Understanding the molecular pathophysiology of hypoxia-ischemia in the brain will provide possibilities for future therapeutic strategies. Several different rodent models have been developed to study the pathophysiology of both focal and global ischemia.

brain research 1564 (2014) 85–100

Besides studying the pathophysiology of brain injury we can study the mechanisms of endogenous brain protection using rodent models of hypoxic-ischemic preconditioning. Hypoxic-ischemic preconditioning involves a non-lethal hypoxic-ischemic event, the preconditioning stimulus, followed by a second more severe hypoxic-ischemic event. The preconditioning stimulus activates endogenous protective mechanisms which lead to a better outcome in the second event, compared to the outcome of the second event alone (Kitagawa, 2012). Preconditioning in itself is not suitable as a therapeutic measure for several reasons such as the limited time-window of protection, and ethical concerns. However, the experimental study of preconditioning contributes greatly to the knowledge of endogenous neuroprotective mechanisms, which is needed to develop potential pharmaceutical treatment options (Durukan et al., 2008). In a little over a decade it has become possible to investigate gene expression in a large number of genes simultaneously. Gene expression is a key component in biological pathways linking the DNA to production of proteins that function as effector molecules. Microarray technology provides an excellent tool for investigating changes in gene expression for thousands of genes at once, and now offers the ability to investigate the complete genome at once. Mechanisms of damage and repair in cerebral hypoxia-ischemia are very complex and largely dependent on changes in gene expression (Stenzel-Poore et al., 2007; VanGilder et al., 2012). Therefore, analysis of the transcriptome by microarray is a useful tool for studying molecular pathophysiology and transcriptional changes in brain protection. In 2000 Soriano et al. were the first to report on a microarray study in an experimental stroke model, which was performed on a 750 gene oligonucleotide array after left middle cerebral artery occlusion in the rat (Soriano et al., 2000). They compared gene expression on the ipsilateral side in the frontal and parietal cortex, and striatum with the same regions on the contralateral side. Unfortunately, only 70% of the investigated genes on the array were known to be expressed in brain since the array contained mainly muscle and cartilage genes. Still, this study delivered valuable information on gene expression after stroke and paved the way for further studies with better techniques. Numerous microarray studies in focal or global ischemia have been published since, but because different experimental models, array types, and methods of analysis have been used it can be difficult to compare results across studies. In this review we aim to provide a clear overview of studies that investigate the transcriptome of focal hypoxia-ischemia or global hypoxia-ischemia using subheadings such as cellular stress, signal transduction and neurotrophins. Moreover, we focus on studies that unravel mechanisms of endogenous brain protection by studying the transcriptome of hypoxicischemic preconditioning and we aim to find the common denominators identified across these studies.

2.

Focal and global hypoxia-ischemia

Here we summarize results of microarray studies investigating the transcriptome of both focal and global ischemia. An overview of studies and their characteristics can be found in

87

Table 1. Middle cerebral artery occlusion (MCAO) in rodents is a widely used method for modeling stroke. In the first microarray study in an MCAO model by Soriano et al. the observed changes in gene expression were similar to previously reported changes in gene expression using traditional methods, such as in situ hybridization and reverse transcriptase polymerase chain reaction techniques, but provided a more elaborate picture. Models for global hypoxia-ischemia used in a microarray study are less homogenous and range from common carotid occlusion, possibly combined with hypoxia or hypotension, to the 4-vessel occlusion model.

2.1.

Immediate early genes

Upregulation of immediate early genes (IEG's) in response to cerebral ischemia was already well described years before microarray technology was developed (Akins et al., 1996). Recent studies using microarray technology confirm the upregulation of IEG's with an onset at 30 min, and lasting up to 24 h following both focal and global ischemia (Buttner et al., 2009; Lu et al., 2004; Kawahara et al., 2004; SchmidtKastner et al., 2002; Soriano et al., 2000; Tang et al., 2002). C-Fos is the most commonly reported upregulated IEG, and glutamate induced excitotoxicity is known to increase c-Fos transcription due to the calcium-dependent cyclic AMPresponsive element in the c-Fos promoter (Vendrell et al., 1993).

2.2.

Stress response genes

Another group of genes that are upregulated in numerous microarray studies of focal and global ischemia are the heat shock proteins (Hsps) (Buttner et al., 2009; Gilbert et al., 2003; Jin et al., 2001; Lu et al., 2004; Stenzel-Poore et al., 2003; Kawahara et al., 2004; Ramos-Cejudo et al., 2012; SchmidtKastner et al., 2002; Tang et al., 2002; Yakubov et al., 2004). Hsp70 is most commonly reported, but many other Hsps are found among which are Hspb2, Hspa5, Hspb6, Hspa8, Hspb8, Hspa9a, Hsp25, Hsp27, Hsp40, Hsp60, Hsp86, Hsp32, and Hsp105. The Hsps are known to act as molecular chaperones and they contribute to repair by binding, refolding and trafficking denatured proteins within the cell. Recent literature suggests that Hsps, more specifically Hsp70, also play an important role in the immune response pathways after neurological injury (Kim and Yenari, 2013). Although the role of Hsps in inflammatory processes is versatile, meaning that they can both induce and arrest inflammation, the induction of Hsp70 is believed to be neuroprotective (Zheng et al., 2008, Zheng and Yenari, 2006). The DNA damage-inducible transcript Gadd45 (α, β or γ) is another stress-response gene that is regularly induced by ischemia (Buttner et al., 2009; Schmidt-Kastner et al., 2002; Tang et al., 2002; Wang et al., 2011a, 2011b). Gadd45 is known to increase cellular DNA repair by stimulating DNA nucleotide excision repair (Smith et al., 1994). Several global ischemia studies report upregulation of a different type of the DNA damage-inducible genes such as Gadd153 (Kawahara et al., 2004), and Gadd34 (Buttner et al., 2009; Jin et al., 2001). The different types of DNA damage-inducible genes are likely to

88

Table 1 – Overview of microarray studies in focal and global ischemia. The asterisk mark-up in the ‘array type' column indicates a sub-optimal array type e.g. not wholegenome or an array designed for a different species than the study subjects. Year

Animal

Type

Experimental model

Sex

Age

Time-points

Tissue

Array type

Soriano et al.

2000

Rat

Focal

MCAO permanent

♂

Adult

3 h no reperfusion

Tang et al. SchmidtKastner et al. Lu et al.

2002 2002

Rat Rat

Focal Focal

MCAO permanent MCAO 2h

♂ ♂

Adult Adult

24 h no reperfusion 3 h reperfusion

Striatum, frontal and parietal cortex Parietal cortex Cortex supplied by MCA

*Custom array (designed for bone and cartilage) RG U34A *Mouse UniGene1

2004

Rat

Focal

MCAO 30min/2h

♂

Adult

2007

Rat

Focal

MCAO 2h

Wang et al. RamosCejudo et al. Jin et al.

2011a, 2011b 2012

Mouse

Focal

MCAO 1.5h

♂ (castrated or DHT replaced) ♂

3 h, 12 h, 24 h reperfusion

Hippocampus

*Custom array

Rat

Focal

MCAO permanentþCCAO 60 min.

?

Adult weeks 9–10 12 weeks Adult

Injured hemisphere Periinfarct cortex

*RG U34 Neurobiology

Cheng et al.

30 min, 4 h, 8 h, 24 h, 3days, 7 days. (2 h–7 days reperfusion) 6 h reperfusion

24 h, 3 days

Agilent G4131F

2001

Rat

Global

♂

Adult

4,24, 72 h

Gilbert et al. Yakubov et al. Buttner et al.

2003

Mouse

Global

♂

Adult

12 h

Hippocampus

*Custom array

2004

Rat

Global

♂

Adult

48 h

2009

Rat

Global

Vertebral artery and ECA coagulationþBCAO 15 min. Unilateral CCAOþ8% O2 30 min. BCAOþhypotension 12 min. Unilateral CCAOþhypotension 15 min.

Ischemic core and Periinfarct area Hippocampus

♂

Adult

1 h, 6 h, 24 h

Hippocampus CA1 Whole hemisphere

*DS analysis 2553 transcripts RG 230 2.0

RG 230 2.0

*Custom array (human genes)

brain research 1564 (2014) 85–100

Author

brain research 1564 (2014) 85–100

play similar roles in DNA repair, but the function of Gadd45 is most widely studied. Several studies report differential expression of genes specifically related to oxidative stress. In a focal ischemia study an upregulation of the anti-oxidant glutathione peroxidase 2 was observed, together with a downregulation of glutathione S-transferase which is likely to serve the same purpose because of its function in glutathione degradation (Schmidt-Kastner et al., 2002). Similarly, in a different MCAO study a downregulation of glutathione S-transferase was observed (Tang et al., 2002). In one global ischemia model an upregulation of the anti-oxidative Rac3 was observed (Yakubov et al., 2004), whereas in another global ischemia model an upregulation of the anti-oxidants glutathione peroxidase 4 and superoxide dismutase 1 and 2 was found (Jin et al., 2001).

2.3.

Apoptosis

Several genomic studies found an upregulation of different caspases which could be an indication of the pro-apoptotic state that follows cerebral ischemia. In a mouse model of 1.5 h MCAO an upregulation of caspase 2 and caspase 7 was seen between 3 and 24 h of reperfusion in the hippocampus, and caspase8ap2 was detected after 3 h of reperfusion (Wang et al., 2011a, 2011b). Similarly, in a rat model of 2 h MCAO an upregulation of caspase 2 and caspase 7 was found in the ischemic hemisphere between 8 and 24 h of reperfusion (Lu et al., 2004). Caspase 3 upregulation seems to be specific for global brain ischemia since this was observed in four different studies using a rat model for global brain ischemia 24 h after reperfusion (Buttner et al., 2009; Jin et al., 2001; Kawahara et al., 2004; Yakubov et al., 2004). Differential expression of Bcl2 family members is observed in several studies. In a recent focal ischemia model in mice an upregulation of pro-apoptotic members Bak1, Bad, and Bcl2l11 was found in hippocampus between 3 and 24 h after reperfusion (Wang et al., 2011a, 2011b). A similar study in rats revealed upregulation of pro-apoptotic gene Baxα as a late response (24 h to 3 days), but also an upregulation of the antiapoptotic genes Bcl2, and Bcl2l1 as an early response (30 min to 24 h) to focal ischemia (Lu et al., 2004). Another focal ischemia study in rats showed upregulation of the antiapoptotic gene Bag3 3 h after reperfusion (Schmidt-Kastner et al., 2002). Furthermore, a downregulation of Bcl2 was observed 24 h after ischemia in periinfarct areas (RamosCejudo et al., 2012). In contrast to what was found in most focal ischemia Jin et al. observed a downregulation of Bcl2l1 in rats 4 h after global ischemia, together with an upregulation of the anti-apoptotic Bag1 (Jin et al., 2001). Similarly Gilbert et al. observed a downregulation of the anti-apoptotic Bnip3l in mouse hippocampus 12 h after global ischemia, whereas Yakubov et al. observed an upregulation of the proapoptotic Bax in rat hippocampus after 48 h (Gilbert et al., 2003; Yakubov et al., 2004). In addition, several studies observed an upregulation of tumor suppressor gene p53 (Buttner et al., 2009; Yakubov et al., 2004). While others found upregulation of genes that are closely related to p53, like p53induced gene 6 and 8 (Pig6, Pig8), and p53-responsive gene Dral (Jin et al., 2001; Lu et al., 2004; Kawahara et al., 2004).

89

On the whole, pro-apoptotic gene expression seems to be predominant after cerebral ischemia, despite some reports of upregulation of anti-apoptotic genes in the early response phase up to 24 h.

2.4.

Signal transduction

Genes involved in signal transduction show rather contradictory results in microarray studies of focal and global cerebral ischemia. For example, whereas most report a downregulation in protein kinases such as protein kinase C (PKC) and calcium/calmodulin-dependent kinases (Buttner et al., 2009; Schmidt-Kastner et al., 2002; Soriano et al., 2000; Tang et al., 2002), others report an induction of PKC (Wang et al., 2011a, 2011b), and even others demonstrate a bipolar pattern of both up- and downregulation depending on time-point and the specific protein kinase type (Lu et al., 2004). Protein kinases are known to be activated by redox signaling and regulate most of the signal transduction in eukaryotic cells. However, by phosphorylation of protein substrates they also regulate many other cellular processes such as metabolism, apoptosis and cell cycle progression. When we look at the available evidence for PKC function after ischemia, the majority of studies report a loss of PKC activity and imply that maintaining PKC activity may contribute to brain protection (Bright and Mochly-Rosen, 2005). Similarly contradictory results are observed for regulator of G-protein signaling (Rgs) genes. For example Rgs2 is downregulated in a focal ischemia study by Schmidt-Kastner et al., while it is upregulated in a focal ischemia study by Cheng et al. and a global ischemia model by Gilbert et al. (Cheng et al., 2007; Gilbert et al., 2003; Schmidt-Kastner et al., 2002). Rgs6 was reportedly upregulated in a focal ischemia model (Wang et al., 2011a, 2011b). Other genes related to G-protein signaling that were found to be downregulated are the Gprotein coupled receptor (Soriano et al., 2000), A2 adenosine receptor (Tang et al., 2002), G-protein gamma 3-linked gene (gng3lg) (Schmidt-Kastner et al., 2002), and purinergic receptor P2Y G-protein coupled (p2ry12) (Buttner et al., 2009). Upregulation was observed for GTP-binding protein (Gem) (Zhang et al., 1994), rho guanine nucleotide exchange factor (Gef) (Schmidt-Kastner et al., 2002), and Rad and Gem-like GTP-binding 2 (Rem2) (Buttner et al., 2009; Cheng et al., 2007). Several studies reported an upregulation of ‘dual-specificity phosphatase' (Dusp) genes which belong to the protein tyrosine phosphatase superfamily and have the ability to inactivate members of the MAPK superfamily. Dusp4 and Dusp16 were upregulated in a focal ischemia model (Wang et al., 2011a, 2011b). Dusp5 and Dusp6 were found to be upregulated in dihydrotestosterone (DHT) replaced castrated males compared to castrated males, which indicates the expression of these transcripts is influenced by circulating male sex hormones (Cheng et al., 2007). Interestingly, it has been suggested that Dusp5 is a direct target of p53 with endogenous p53 binding to the promoter region of Dusp5 (Ueda et al., 2003). When considering changes in gene expression related to signaling cascades we need to keep in mind that protein phosphorylation and de-phosphorylation may play a more

90

brain research 1564 (2014) 85–100

significant role then changes in gene expression after brain ischemia (Tang et al., 2002).

2.5.

Neurotransmission

Genes related to neurotransmission are downregulated in several studies. Most of these downregulated transcripts are related to glutamate neurotransmission which is not surprising since it is well known that excessive glutamate signaling leads to excitotoxicity and neurological injury in cerebral ischemia (Kostandy, 2012). In two global ischemia studies a downregulation was observed for the metabotropic subtype of glutamate receptor, Grm3 and Grm5 respectively (Buttner et al., 2009; Kawahara et al., 2004). Furthermore, a downregulation of the AMPA-type receptors (Gria1, Gria3), and the NMDA-subtype (Grin1, Grin2a) was observed (Kawahara et al., 2004). Tang et al. studied focal ischemia and found a downregulation of the cholinergic receptor Chrm3, GABA-receptor GABRD, and glutamergic receptors Grin1 and Gria3 (Tang et al., 2002). Similarly, Ramos-Cejudo observed downregulation of Grin1 in periinfarct areas (Ramos-Cejudo et al., 2012). Some studies found differential expression in synaptic genes. In a focal ischemia model in rats both up and downregulation were observed with Snap25a, Snap25b, Synapsin1a, Synapsin 2a, synaptic vesicle SV2, synaptotagmin binding zygin, and synuclein1 being downregulated, while Synaptogyrin, Snap23, Synaptotagmin V, Synaptotagmin VII, Syntaxin4 were upregulated (Lu et al., 2004). In a different focal ischemia study a downregulation of Synaptotagmin XI was found (Schmidt-Kastner et al., 2002). In a global ischemia model upregulation of post synaptic gene Homer1C was found (Kawahara et al., 2004).

2.6.

Ion channels

Several cerebral ischemia microarray studies report differential expression of ion channel genes. Lu et al. reported differential expression of 32 different ion channel genes mainly consisting of potassium channels, followed by sodium and calcium channels of which the majority was downregulated (Lu et al., 2004). Downregulation of potassium channels (Kcnd2, Kcns1) was also observed in another focal ischemia model (Tang et al., 2002). In a global ischemia model downregulation of calcium channels (Cacnb3, Cacnb4) was observed (Jin et al., 2001). As for signal transduction, activation/inactivation of ion channels may play a more significant role than changes in gene expression after brain ischemia.

2.7.

Inflammation

Upregulation of transcripts involved in the inflammatory process is an important part of the response to ischemia (Vexler et al., 2006). CC and CXC chemokines are known to play a vital role in cerebral injury after focal ischemia and this is confirmed with the upregulation of CC and CXC chemokines observed in most focal ischemia microarray studies (Lu et al., 2004; Ramos-Cejudo et al., 2012; Schmidt-Kastner et al., 2002; Soriano et al., 2000; Tang et al., 2002). Targeting these chemokines has even been suggested as a potential therapeutic strategy in stroke (Mirabelli-Badenier et al., 2011).

Toll-like receptors (TLRs) belong to the innate immune system and are also known to play a pivotal role in hypoxicischemic brain injury (Tang et al., 2007). However, just 3 microarray studies of focal ischemia report an upregulation of TLR signaling (Ramos-Cejudo et al., 2012; Schmidt-Kastner et al., 2002; Wang et al., 2011a, 2011b), although several others report upregulation of pro-inflammatory cytokines that are known to be induced by TLR activation such as Il-6 and Tumor Necrosis Factor (TNF) (Lu et al., 2004; Ramos-Cejudo et al., 2012; Schmidt-Kastner et al., 2002; Soriano et al., 2000; Tang et al., 2002). Since it is cellular death that triggers the inflammatory response it is not surprising that we see notably less upregulation of genes related to inflammation in global ischemia models where there is no delineable infarction. In a model of unilateral common carotid artery occlusion combined with hypobaric hypotension an upregulation of a CC chemokine, Ptgs2 and Tnfrsf12a was found (Buttner et al., 2009). Jin et al. observed upregulation of the TNF-factor Litaf, while Yakubov et al. observed upregulation of a CC chemokine (Jin et al., 2001; Yakubov et al., 2004).

2.8.

Cytoskeleton

Cytoskeletal proteins are very important for cellular structure, and transport into and from dendrites. Dissolution of cytoskeletal proteins is known to occur after ischemia and can trigger apoptotic changes (Lipton, 1999). Several microarray studies of cerebral ischemia show differential expression of various cytoskeletal genes. Differential expression of the ‘activity-regulated cytoskeleton-associated protein' (Arc) is observed in several studies. Arc upregulation is mainly observed very early after ischemia (1–3 h), whereas Gilbert et al. observed a downregulation of Arc 12 h after ischemia (Buttner et al., 2009; Gilbert et al., 2003; Soriano et al., 2000). After induction Arc is transported to activated postsynaptic sites on dendrites, and is most likely involved in synaptic plasticity (Rickhag et al., 2007). Furthermore, Arc is also a member of the IEG-family which explains why its upregulation post-ischemia is only observed in the very early phase. In a permanent MCAO model an upregulation of GFAP, Tubulin α6, Vimentin, and Moestin was observed 24 h after the start of ischemia (Tang et al., 2002). Another MCAO study found upregulation of GFAP, Tubulin ß1, Vimentin, Nestin, Neuroglycan, δcatenin, and microtubule-associated protein 2 (MAP2) between 2 h and 7 days of reperfusion, with a peak at 24 h (Lu et al., 2004). At the same time they observed downregulation of microtubule-associated protein 1B and 2C (MAP1B, MAP2C). Schmidt-Kastner et al. also observed downregulation of cytoskeletal genes (Tubulin α2, Tubulin ß5, Tubulin α1) in a MCAO model after 3 h of reperfusion (Schmidt-Kastner et al., 2002). In models of global ischemia Gilbert et al. observed downregulation of Annexin a7, Drebrin, and Spectrin ß2 12 h after ischemia, while Yakubov et al. observed upregulation of karyopherin α2 24 h after ischemia (Gilbert et al., 2003; Yakubov et al., 2004). An upregulation of cytoskeletal genes could be explained as an attempt to compensate for the dissolution of cytoskeletal proteins after ischemia. Furthermore, the expression of GFAP and Vimentin is specific to

91

MG U74Av2 Cortex 1 h and 6 h after start of preconditioning, 6 h after MCAO Permanent MCAO 8% oxygen 6 h Adult ♂ Mouse 2006

8–10 weeks Mouse 2003

♂

MCAO 15 min

60 min MCAO

24 h reperfusion

Cortex

GeneChip Rat Gene 1.0ST MG U74Av1 Whole hem. 96 h – E17 Rat 2013

♂

30min. clamping of uterine circulation.

RG U34A Cortex 0 h, 2 h, 8 h and 24 h – P6 ♂/♀ Rat 2007

Gustavsson et al. CoxLimpens et al. StenzelPoore et al. Tang et al.

Adult Rat

Rat 2004

2007

Rat

Feng et al.

♂

8% oxygen 3 h

RG 230 2.0 Hippo 0 h, 1 h, 4 h, and 24 h

RG U34A Adult ♂

Vertebral artery ligationþBCAO and hypotension, 2 min. Fasted BCAO/ hypotension, 3 min.

Vertebral artery ligationþBCAO and hypotension, 6 min. BCAO/ hypotension, 6minutes

1h,3 h,12 h, 24 h and 48 h

Whole hem. Hippo CA1 0 h, 6 h, 18 h, and 24 h – P6 ♂/♀

Preconditioning Stimulus Age of preconditioning Sex

2002

By studying the transcriptome of preconditioning we can learn about endogenous protective mechanisms in the brain. It has been suggested that preconditioning reprograms the brain transcriptome in order to achieve ischemic tolerance (Stenzel-Poore et al., 2007). So far, seven studies have investigated preconditioning induced gene expression with microarrays of which an overview and characteristics can be found in Table 2. Here we summarize the findings and relate these findings to what we observed in focal/global hypoxiaischemia.

Bernaudin et al. Kawahara et al.

Preconditioning of the brain

Animal

3.

Year

Induction of neurotrophins in cerebral ischemia has been shown to be neuroprotective due to their anti-apoptotic function (Wang et al., 2013). We observed an upregulation of neurotrophic genes in many microarray studies in both focal and global ischemia. Upregulation was most frequently found for brain-derived neurotrophic factor (BDNF) (Lu et al., 2004; Ramos-Cejudo et al., 2012; Schmidt-Kastner et al., 2002; Soriano et al., 2000; Tang et al., 2002), and vascular endothelial growth factor (VEGF) (Jin et al., 2001; Lu et al., 2004; Wang et al., 2011a, 2011b). According to a recent study measuring BDNF and VEGF in human serum, both of these neurotrophins are relevant for clinical prognosis since circulating levels of BDNF and VEGF modify the risk of stroke and transient ischemic attack (TIA) (Pikula et al., 2013). Two studies found members of the transforming growth factor beta (TGF-ß) family to be upregulated such as TGF-ß and the TGF-ß receptor which are known to exert multiple neuroprotective functions (Dobolyi et al., 2012; Lu et al., 2004; RamosCejudo et al., 2012). Others observed upregulation of genes that act in concert with TGF-ß such as Inhibin-A and the RET receptor tyrosine kinase which is a receptor for glial cell linederived neurotrophic factor (Soriano et al., 2000; Tang et al., 2002). Another interesting observation is the upregulation of genes related to the insulin growth factor (IGF) which is a promising neuroprotective agent in oxidative stress (Dobolyi et al., 2012; Zheng et al., 2000). Lu et al. observed upregulation of five genes relates to IGF, whereas Jin et al. observed upregulation of the IGF1 receptor (Jin et al., 2001; Lu et al., 2004).

Second insult

Neurotrophic factors

Table 2 – Overview of microarray studies of preconditioning.

2.10.

Time-points

Differential regulation of ribosomal proteins is mainly observed in models of global ischemia and most studies report an upregulation of specific (mitochondrial) ribosomal proteins such as Rps20, Rpl35, and Mrp19 (Gilbert et al., 2003; Jin et al., 2001; Tang et al., 2002).

8% oxygen 3h

Tissue

Ribosome

Author

2.9.

Array type

astrocytes, and their upregulation is a clear indicator that glial cells are involved in the transcriptional response to ischemia, possibly in a protective role (Li, 2008).

RG U34A

brain research 1564 (2014) 85–100

92

3.1.

brain research 1564 (2014) 85–100

Immediate early Genes

Upregulation of IEG's, such as c-Fos, is less frequently observed in preconditioning compared to injurious ischemia. Only Kawahara et al. observed an upregulation of c-Fos after preconditioning alone (Kawahara et al., 2004). Some studies report upregulation of other IEG's such as JunB, Egr1, 2 and 4 (Gustavsson et al., 2007; Kawahara et al., 2004).

3.2.

Stress response genes

The marked upregulation of Hsp's that we observed in response to injurious ischemia seems to be less pronounced after preconditioning. Several studies have compared the transcriptome after preconditioning alone, and after ischemia alone. Stenzel-Poore et al. demonstrated that after preconditioning, consisting of 15 min MCAO, Hsp70 was upregulated after 24 h, and Hspb2 was upregulated after 72 h. Furthermore, they show that injurious ischemia, consisting of 60 min MCAO, leads to a more pronounced early phase response with upregulation of Hsp70, Hsp105 and Hsp10 after 24 h (Stenzel-Poore et al., 2003). Similarly, in a global ischemic preconditioning model, consisting of 2 min of global ischemia, an upregulation of Hsp70 was observed after preconditioning alone. In the same experimental model upregulation of Hsp70, Hsp27, Hsp60, Hsp10, Hsp40, and Hsp32 were observed when ischemia lasted for 6 min (Kawahara et al., 2004). From these studies we can conclude that the upregulation of Hsp's is much stronger after injurious ischemia, compared to preconditioning. Furthermore, Feng et al. looked at gene expression after ischemia preceded by preconditioning and he observed an initial upregulation of Hsp70 after 1 h, followed by a downregulation 4 and 24 h after ischemia (Feng, 2007). Contrastingly, Tang et al. observed an upregulation of Hsp40 after ischemia when ischemia was preceded by hypoxic preconditioning. It is important to note that these studies used different paradigms for modeling preconditioning and injurious ischemia (see Table 2). Gustavson et al. observed an upregulation of Hspa1a en Hspa2 in cortex, directly after 8% hypoxia while Bernaudin et al. observed no differential expression in whole hemisphere for any of the Hsps (Bernaudin et al., 2002; Gustavsson et al., 2007). Even though we observe a stronger upregulation of heat shock protein genes in injurious ischemia in microarray studies, we need to keep in mind that these transcripts might not be translated to actual heat shock proteins. In gerbil hippocampus upregulated gene expression of Hsp70 was found after preconditioning and after injurious ischemia, however the Hsp70 proteins were only present in the preconditioned animals (Kanemitsu et al., 1994). The DNA damage-inducible GADD genes are rarely induced by preconditioning. Only Gustavson et al. observed an upregulation of GADD45 after hypoxic preconditioning, while Feng et al. observed a downregulation of GADD45 24 h after ischemia when preceded by preconditioning (Gustavsson et al., 2007; Feng, 2007). The fact that upregulation of DNA damage-inducible genes is more frequently observed in injurious ischemia than after preconditioning is not surprising because it is linked to p53 activation and apoptosis which is believed to be attenuated by preconditioning (Li et al., 1997). In a study investigating neuroprotection in

NMDA-induced excitotoxicity, GADD45 was not upregulated after ischemia when the neuroprotective agent was administered (Laabich et al., 2001). However, GADD45 upregulation has also been observed in surviving neurons which led some to hypothesize it might play a neuroprotective role (Charriaut-Marlangue et al., 1999; Chen et al., 1998). Nevertheless, we need to keep in mind that Gadd45 has numerous effectors and might play a dual role depending on the amount of damage (Salvador et al., 2013). Several preconditioning studies report differential expression of genes related to oxidative stress, for example the upregulation of carbonic anhydrase IV after hypoxic preconditioning alone (Bernaudin et al., 2002). After global injurious ischemia preceded by preconditioning an upregulation of the oxidative phosphorylation pathway was found, which indicates free radical production (Feng, 2007). In a fetal preconditioning study a downregulation of the neuronal nitric oxide synthase 1 (NOS1) pathway was observed after preconditioning alone, which suggest a decrease in the ability to produce nitric oxide in a subsequent ischemic event (Cox-Limpens et al., 2013).

3.3.

Apoptosis

Since part of the protective mechanism of preconditioning is believed to be inhibition of apoptosis, it is not surprising that in preconditioned animals we observe less upregulation of apoptosis-related genes compared to injurious ischemia (Ding et al., 2012). Regarding the differential regulation of caspases we notice that only Gustavson et al. observed upregulation of a caspase gene (caspase 8) after hypoxic preconditioning, while Feng et al. report downregulation of caspase 3 after global ischemia when preceded by preconditioning (Gustavsson et al., 2007; Feng, 2007). Other preconditioning studies did not report differential regulation of caspase genes. An upregulation of Bcl2-related genes was observed in both studies that investigate the transcriptome after 3 h of hypoxic preconditioning. One study focused at the cortex and observed upregulation of anti-apoptotic genes Bag3 and Bnip3, while the other study looking at the whole hemisphere observed upregulation of the pro-apoptotic Bcl2l11 gene (Bernaudin et al., 2002; Gustavsson et al., 2007).

3.4.

Signal transduction

The preconditioned transcriptome shows upregulation of several genes related to the MAPK pathway such as Dusp1 which was previously named MAP Kinase Phosphatase (MKP1), Dusp5, and MAP3K1 (Bernaudin et al., 2002; Gustavsson et al., 2007; Kawahara et al., 2004; Feng et al., 2007). Furthermore, upregulation of PKC family members is observed with contradictive results which might be due to the differences in the experimental models used. In a hypoxic preconditioning an upregulation of protein kinase C-like (PKN1) was observed after preconditioning alone, while a study of MCAO ischemia preceded by MCAO preconditioning reports upregulation of protein kinase inhibitor alpha (PKIA), which is an inhibitor of PKC activity (Gustavsson et al., 2007; Stenzel-Poore et al., 2003). However, previous studies using PKC isoenzyme-selective modulators or PKC selective activators have shown that PKC is required for the

brain research 1564 (2014) 85–100

93

induction of ischemic tolerance (Bright and Mochly-Rosen, 2005). Cardiac ischemic preconditioning models have shown that PKC activates the extracellular signal-regulated kinase (ERK) pathway, which is geared towards cell survival (Downey and Cohen, 2007). Just one study reported a differential expression of g-protein signaling related genes, and they observed an upregulation of GTP-binding protein subunit 8 (Gγ8) (Bernaudin et al., 2002).

et al., 2002), while others observed no differential expression related to ion channels after preconditioning (Gustavsson et al., 2007; Kawahara et al., 2004; Tang et al., 2006). Although acid-sensing ion channels have been implicated in preconditioning induced brain protection (Chu and Xiong, 2012; Pignataro et al., 2011), none of the whole-genome studies in preconditioning observed differential expression for any of the ASICs genes.

3.5.

3.7.

Neurotransmission

Differential expression is observed for genes related to neurotransmission in microarray studies of preconditioning. Similar to studies of injurious ischemia, most studies of preconditioning observe a downregulation with the exception of a hypoxic preconditioning study in the neonatal rat which observed upregulation of glutamate transporter EAAT4 directly after hypoxia (Bernaudin et al., 2002). EAAT4 has been shown to control extrasynaptic glutamate levels following synaptic stimulation which can reduce excitotoxicity (Liang et al., 2008; Tsai et al., 2012). In a model of MCAO ischemia preceded by preconditioning a downregulation of dopamine receptor Drd4, and glutamate receptor Grid2 was observed (Stenzel-Poore et al., 2003). In a model for global ischemia a downregulation of three glutamate receptors (Gria2, Gria3, and Grm3) was found after preconditioning alone. The same article also studied injurious global ischemia where they also observed a downregulation of Gria3, but furthermore of Gria1, Grin1 and Grin2a (Kawahara et al., 2004). In a fetal preconditioning model a downregulation of the serotonin receptor Htr1b was observed while further pathway analysis revealed downregulation of several GO terms related to neurotransmission: GO7268 Synaptic transmission, GO7269 Neurotransmitter secretion, and GO14048 Regulation of glutamate secretion (Cox-Limpens et al., 2013). The observed upregulation of EAAT4, and the downregulation of glutamate receptors and glutamate signaling in preconditioned animals logically leads to decreased vulnerability to excitoxicity in a subsequent hypoxic-ischemic insult.

3.6.

Ion channels

In injurious ischemia several studies observed a downregulation of different ion channels. In microarray studies in preconditioning this downregulation seems less extensive and involves mainly potassium and calcium channels. In a study of MCAO ischemia preceded by MCAO preconditioning a downregulation of the Kcna5 potassium channel was observed (Stenzel-Poore et al., 2003). Feng et al. performed a pathway analysis after global injurious ischemia preceded by preconditioning and found a significant enrichment of the KEGG calcium signaling pathway of which most genes were downregulated (Feng, 2007). Similarly, Cox et al. performed pathway analysis after fetal preconditioning alone and observed downregulation of pathways such as GO5216 Ion channel activity, GO5261 Cation channel activity, and GO5244 Voltage gated ion channel activity (Cox-Limpens et al., 2013). On the contrary, one hypoxic preconditioning study observed upregulation of potassium channels Kcna1, and Kcnma1 (Bernaudin

Inflammation

Two studies using different experimental models report a downregulation of prostaglandin synthase Cox2 after injurious ischemia when preceded by preconditioning (Tang et al., 2006; Feng et al., 2007). Others observed differential expression of TNF-related genes. For example, after hypoxic preconditioning alone an upregulation of TNF-receptor superfamily member 12a (Tnfrsf12a) and TNF-receptor superfamily member 1a (Tnfrs1a) was observed, together with a downregulation of the Tnfrsf1a death domain (Tradd) (Gustavsson et al., 2007). Whereas Cox-Limpens et al. observed downregulation of Tnfrsf11b, and upregulation of the Ccl11 chemokine, interleukin 1 receptor accessory protein (Il1rap), and interleukin 5 receptor alpha (Il5ra) after intrauterine fetal preconditioning (Cox-Limpens et al., 2013). Downregulation of Toll-like receptor signaling has been suggested to play a role in preconditioning induced neuroprotection, but similar to our findings in injurious ischemia this cannot be derived from whole-genome gene expression studies. Only one microarray study reports differential tolllike receptor expression (Wang et al., 2011a, 2011b; Feng et al., 2007).

3.8.

Cytoskeleton

After MCAO preconditioning alone a downregulation of betasarcoglycan (Sgcb) and keratin complex 1, together with upregulation of GFAP, Annexin2 and Fibulin2 is observed (Stenzel-Poore et al., 2003). Upregulation of GFAP was also observed after hypoxic preconditioning alone, and after injurious ischemia preceded by preconditioning, similar to the upregulation of GFAP that is seen after injurious ischemia alone (Tang et al., 2006). In a study of fetal hypoxic-ischemic preconditioning a downregulation of the dendritically localized cytoskeletal protein Dendrin (Ddn), and the actin binding FH2 domain containing 1 gene (Fhdc1) was observed together with upregulation of Myosin 3a which plays an important role in actin bundle formation and stability (CoxLimpens et al., 2013; Quintero et al., 2010). Furthermore, a study of global brain hypoxia-ischemia preceded by preconditioning observed upregulation of ßcatenin binding, and cytoskeletal protein binding in pathway analysis (Feng, 2007).

3.9.

Ribosome

Differential regulation of ribosomal proteins is mainly observed in models of global ischemia both with and without preconditioning. Two studies report an upregulation of the KEGG pathway 3010 Ribosome (Cox-Limpens et al., 2013; Feng

94

brain research 1564 (2014) 85–100

et al., 2007). It has been suggested that the observed upregulation of ribosomal proteins after preconditioning or ischemia could facilitate the recovery of protein processing machinery and therefore contribute to preconditioning-induced brain protection (Nakagomi et al., 1993; Feng et al., 2007).

3.10.

Neuroptrophic factors

Subtoxic levels of glutamate are known to stimulate BDNF release and synthesis, and this NMDA induced increase in BDNF has been suggested to play a major role in preconditioning induced brain protection (Marini et al., 2007). However, contrasting to studies of injurious ischemia, none of the microarray studies in preconditioning models report differential expression of BDNF. Several studies did observe differential expression for other neurotrophin or growth factor-related genes. After hypoxic preconditioning alone an upregulation was found for VEGF, Igf1r and TGFa (Bernaudin et al., 2002; Gustavsson et al., 2007). After global hypoxic-ischemic preconditioning an upregulation was observed for TGF-ß inducible early growth response gene (Tieg), which is more strongly upregulated in injurious global hypoxia-ischemia (Kawahara et al., 2004). Furthermore, a fetal preconditioning study found an upregulation of fibroblast growth factor receptor 4 (Fgfr4), and Smad6 which is a negative regulator of TGF-ß signaling (Cox-Limpens et al., 2013). Lastly, two studies observed an upregulation of TGF-beta signaling after pathway analysis (Gustavsson et al., 2007; Feng, 2007). The neuroprotective effects of TGF-beta in ischemia have been well established (Dobolyi et al., 2012).

4.

Discussion

4.1.

Focal versus global injurious cerebral ischemia

The response to focal and global injurious cerebral ischemia is similar regarding the upregulation of IEG's, upregulation of stress-response genes such as Hsps, upregulation of proapoptotic genes, and downregulation of genes involved in glutamergic signaling. However, there are also some interesting differences in the transcriptional response to focal versus global cerebral ischemia. Firstly, several focal ischemia studies report upregulation of Gadd45, global ischemia studies report upregulation of different types of DNA-inducible transcripts such as Gadd 153 and Gadd34. Furthermore, upregulation of transcripts involved in the inflammatory process, such as CC and CXC chemokines, is predominantly observed in response to focal ischemia. Since it is cellular death that triggers this inflammatory response, the observed difference is possibly explained by the pathophysiological difference between focal and global ischemia. More specifically: the fact that there is no delineable infarction in global cerebral ischemia. Finally, upregulation of ribosomal genes is predominantly observed in global ischemia models. The observed upregulation of ribosomal genes has been hypothesized to be a response to post-ischemia protein aggregation leading to ribosomal dysfunction and translational arrest (Ge et al.,

2007; Feng et al., 2007). We hypothesize that the observed difference between focal and global ischemia models results from the fact that most focal ischemia models involve permanent ischemia, and are therefore more likely to induce irreversible translational arrest. On the other hand, global ischemia models involve a transient cerebral ischemic insult and are therefore most likely to induce a reversible translational arrest. Logically, recovery of translation machinery is only useful in cells that are likely to survive, not in infarct areas (DeGracia et al., 2008).

4.2.

From bench to bedside

In many ways rats provide an excellent model for cerebral ischemia in humans but although there are similarities in cerebrovascular anatomy and stroke pathophysiology, there are also important differences (Durukan et al., 2008). Because of ethical and practical concerns it is very difficult to obtain human brain tissue for microarray research and accordingly studying the stroke transcriptome in human brain. Therefore, gene expression in peripheral blood is of great interest considering the ease by which it can be obtained. Furthermore, the circulating immune cells contained in peripheral blood are known to react to organ damage and thus likely to react to ischemic brain damage. Proof of principle was provided by several microarray studies in rat models that showed that gene expression in peripheral blood can distinguish distinct brain injuries such as ischemic stroke, hemorrhagic stroke and brief focal ischemia (Tang et al., 2001; Zhan et al., 2010). The existence of a genomic signature for ischemic stroke has now also been demonstrated in human peripheral blood which provides a promising avenue for developing biomarkers to aid diagnosis and monitoring response to treatment (Grond-Ginsbach et al., 2008; Oh et al., 2012; Sharp et al., 2007).

4.3. Is there a transcriptome signature for cerebral preconditioning? Comparing gene expression across independent studies is crucial in discovering genes that are specific to the research question. This because speculations based on individual studies might not be in agreement with the transcripts that are differentially regulated in common across different studies (Fortunel et al., 2003). In order to find the transcriptome ‘signature' of endogenous brain protection we looked for overlapping findings in all preconditioning microarray studies. We compiled transcripts differentially regulated in seven separate preconditioning studies (see Table 2) and surprisingly, only 15 transcripts, out of a total of 434, were identified as differentially regulated in two studies or more, and only two transcripts are present in three studies (see Fig. 1). The limited amount of overlap in single-gene based analysis could be due to substantial differences in experimental paradigms, such as age, insult type, gender, the investigated brain region, and read out. To give an indication of the diversity among these seven preconditioning studies: six different experimental paradigms, five different brain regions, four different array types, and three different ages were used. The overlapping transcripts comprise mainly transcription factors (Fos, JunB, Egr1, Egr2, Nr4a1), and in order to find a unifying biological theme

95

brain research 1564 (2014) 85–100

Fig. 1 – Heatmap of transcripts that were described in two or more preconditioning studies. Red squares represent upregulated transcripts, green squares represent downregulated transcripts, and white squares represent that the transcript is not described in that particular study. Table 3 – DAVID annotation clustering results. In this table: ‘count' represents the number of genes from our query present in this biological theme. The EASE score is a modified Fisher Exact P-value po0.01 indicates significant enrichment. Bonferroni and FDR (false discovery rate) are different statistical methods to correct for multiple testing, po0.05 is viewed as a significant result after correction for multiple testing. Biological term

Count

Genes

GO:0010033 Response to organic substance GO:0006357 Regulation of transcription from RNA polymerase II promoter KEGG04010:MAPK signaling pathway

9

Nfkbia, Junb, Sgk, Id1, Egr1, Egr2, Cdkn1a, Dusp1, Fos Nfkbia, Junb, Nr4a1, Id1, Egr1, Egr2, Fos Gadd45a, Nr4a1, Dusp1, Ntrk2, Fos, Hspa2

7 6

among these 15 transcripts we subjected them to Functional Annotation Clustering in DAVID Bioinformatic environment v6.7 (Huang et al., 2009). Results are shown in Table 3. Most of these 15 transcripts are significantly enriched in GO: 0010033: “Response to organic substance”, next are seven transcripts which are enriched in GO: 0006357: “Regulation of transcription from RNA polymerase II promoter”, and finally six transcripts are found in the KEGG pathway “MAPK signaling”. Interestingly, the transcripts that were enriched in the KEGG pathway of mitogenactivated protein kinase (MAPK) signaling seem to be geared toward proliferation/differentiation and inhibition of the proapoptotic JNK/p38 MAPK pathway (see Fig. 2). The MAPK family consists of several subfamilies among which p38 and ERK1/2 that are both activated by cellular stress such as ischemia. The difference between these signaling pathways is that the MAPK/ p38 pathway is geared towards apoptosis, whereas the ERK1/2 pathway is geared towards cell survival. In the microarray studies of preconditioning we see enrichment of the MAPK signaling pathway, and looking at the specific transcripts that were regulated it seems that they are geared towards cell survival. The involvement of the ERK1/2 pathway in favor of

Fold enrichment

EASE score

Bonferroni

FDR

7.8

2.3E  06

0.001

0.003

9.3

3.2E  05

0.017

0.047

11.5

4.9E  05

0.001

0.041

the p38 pathway has been previously suggested in preconditioning induced brain protection in a global ischemia model in the rat (Kovalska et al., 2012). Considering the modest amount of single gene overlap between different preconditioning models we feel that it is too early to conclude on the protective signature of the transcriptome, and more research with similar experimental design, end time-points, and read-out is needed. Moreover, we feel that pathway analysis should be incorporated into the analysis of microarray data especially for in vivo experiments. Gene expression of a specific gene in a living organism does not stand alone but is always associated with a multitude of biological pathways. An important advantage of pathway analysis is that it is able to detect smaller changes in expression for groups of genes that share biological function. Although a large increase in expression for a single gene might be interesting, so might a more subtle increase in expression in a group of genes that share biological function (Curtis et al., 2005). Incorporating pathway analysis to microarray analysis has been shown to improve detection of similarities between independently derived

96

brain research 1564 (2014) 85–100

Fig. 2 – KEGG pathway of MAPK signaling. Red squares indicate the transcripts that we identified in at least two microarray studies of preconditioning. datasets and therefore provides greater insight in underlying biology (Subramanian et al., 2005).

4.4.

Limitations of microarray research

Microarray technology is a great tool that allows us to detect changes in mRNA expression on a large scale. However, as with all tools there are some limitations. First of all, it is important to note that not all microarray studies described in this review had access to whole-genome techniques, which results in probes that are limited in number, and therefore biased. Advantages in microarray technology on the one hand and genome annotation on the other lead to regularly emerging updated arrays. For example, transcript recognition was originally limited to the 30 -end whereas now it is possible to also detect alternatively spliced variants and truncated transcripts with microarray technology. Although these new and improved arrays have been shown to correspond quite well to older versions there might be some discrepancies (Okoniewski et al., 2007). Because our knowledge about the structural and functional properties of the genome is constantly improving, especially for provisionally annotated transcripts, the data sources used to build the arrays are continuously changing. For example looking at Affymetrix,

the leading company in oligonucleotide arrays, their U34A rat array was based on UniGene build 34, while the improved 230 2.0 array was based on UniGene build 99, and their latest GeneChip Rat Gene 1.0 array is based on the rat genome sequence of UCSC November 2004 (Baylor 3.4/rn4). Furthermore, we need to be aware that it is mRNA expression we are investigating and although this can give great insight into cellular biology, gene expression does not necessarily translate into protein expression. We need to consider the ischemia induced decrease in protein production (Honkaniemi and Sharp, 1996), and the fact that hypoxic conditions can increase the stability of mRNA for specific transcripts and could therefore lead to false positives (Guhaniyogi and Brewer, 2001; Schmidt-Kastner et al., 2002).

4.5.

Sex differences

Males have since long been known to have an increased sensitivity to cerebral hypoxic-ischemic events (Naeye et al., 1971). This hypothesis was recently reconfirmed for human neonates, while it is has also been demonstrated in adults and other species such as rats (Gibson, 2013; Kirchengast and Hartmann, 2009; Loidl et al., 2000). In experimental studies of cerebral ischemia male subjects are preferred, because of

brain research 1564 (2014) 85–100

their increased sensitivity on the one hand and the absence of hormonal fluctuations present in females on the other hand. Although differential gene expression between the sexes in response to ischemia was demonstrated previously with non whole-genome techniques, Cheng et al. were the first to utilize whole-genome techniques (Cheng et al., 2007; Siegel et al., 2010).They investigated the role of dihydrotestosterone (DHT) in brain ischemia, as a potential culprit responsible for the increased vulnerability in males. A beneficial effect of castration regarding infarct size was demonstrated, which was reversed by dihydrotestosterone (DHT) replacement. Moreover, they observed overall differences in gene expression such as a change in direction for differential expression. While the majority of differentially expressed genes was downregulated in the castrated male, upregulation is predominant in intact and DHT-replaced males after ischemia (Cheng et al., 2007). The latter corresponds to the overall upregulation of gene expression that is reported in microarray studies of cerebral ischemia with exclusively male subjects (Cheng et al., 2007; Lu et al., 2004; Schmidt-Kastner et al., 2002; Soriano et al., 2000; Tang et al., 2002). These findings confirm that DHT mediates, at least in part, the male sensitivity to cerebral ischemia. Furthermore, Cheng et al. were able to identify several genes that are differentially regulated in ischemia in response to DHT. Most of the genes differentially regulated by DHT in ischemia are related to inflammation, and are mostly pro-inflammatory (Cheng et al., 2007).

4.6.

Epigenetics

Across all microarray studies many differentially expressed transcripts have their function in the cell nucleus, interacting with the DNA. It has been suggested that we need to take a look at transcriptional regulation in order to find how differential expression is regulated in cerebral ischemia (SchmidtKastner et al., 2002). We also observed several clear indications of involvement of epigenetic mechanisms across the microarray studies we reviewed. For example, in a microarray study of preconditioning a downregulation of Histone H2b was observed which indicates that chromatin modifications might play a role in preconditioning (Stenzel-Poore et al., 2003). Similarly, in a study of fetal preconditioning upregulation was observed for chromatin-related pathways which contained well known epigenetic players such as Hdac and Mecp2 (Cox-Limpens et al., 2013). Non-microarray studies have also shown that epigenetic mechanisms are important in cerebral ischemia, for example in a stroke study where a Hdac inhibitor was demonstrated to exert anti-inflammatory effects (Kim et al., 2007). Furthermore, Gadd proteins were recently discovered to be key epigenetic regulator (Schäfer, 2013). Gadd45a is one of these Gadd genes, and was identified in the present study as one of the 15 transcripts that overlapped across multiple preconditioning microarray studies. Another recent study demonstrated that chromatin modifications can occur in response to changes in the availability of energy substrates in the cell, and that these modifications might even be inherited transgenerationally (Gut and Verdin, 2013). Considering that the study of epigenetics is a relatively new field, we expect that more research aimed at unraveling

97

epigenetic mechanism in ischemia and brain protection will emerge in the near future.

5.

Conclusion

The amount of data across microarray studies can be daunting and hard to interpret which is why we aimed to provide a clear overview of available data in experimental rodent models. Findings for both injurious ischemia and preconditioning were reviewed under clear subtopics such as cellular stress, inflammation, cytoskeleton and cell signaling. Finally, we researched the transcriptome signature of brain protection across studies of preconditioning. Strikingly, when comparing genes discovered by single-gene analysis we observed only 15 genes found in two studies or more. We subjected these 15 transcripts to DAVID Annotation Clustering analysis in order to derive their shared biological meaning. Interestingly, the MAPK signaling pathway and more specifically the ERK1/2 pathway geared toward cell survival/proliferation was significantly enriched. We note that several studies identified ‘new' transcripts in relation to hypoxia-ischemia and suggest that these transcripts should be studied further. Considering the lack of overlap in better known genes involved in ischemia and brain protection, we wonder if this is the right conclusion, or if further focus on these uniquely identified genes would only leave us sidetracked. We advocate incorporating pathway analysis into all microarray data analysis in order to improve detection of similarities between independently derived datasets.

r e f e r e nc e s

Akins, P.T., Liu, P.K., Chung, Y.H., 1996. Immediate early gene expression in response to cerebral ischemia, friend or foe? Stroke 27, 1682–1687. Bernaudin, M., Tang, Y., Reilly, M., Petit, E., Sharp, F.R., 2002. Brain genomic response following hypoxia and re-oxygenation in the neonatal rat. J. Biol. Chem. 277, 39728–39738. Bright, R., Mochly-Rosen, D., 2005. The role of protein kinase C in cerebral ischemic and reperfusion injury. Stroke 36, 2781–2790. Buttner, F., Cordes, C., Gerlach, F., Heimann, A., et al., 2009. Genomic response of the rat brain to global ischemia and reperfusion. Brain Res. 1252, 1–14. Charriaut-Marlangue, C., Richard, E., Ben-Ari, Y., 1999. DNA damage and DNA damage-inducible protein Gadd45 following ischemia in the P7 neonatal rat. Brain Res. Dev. Brain Res. 116, 133–140. Chen, J., Uchimura, K., Stetler, R.A., Zhu, R.L., Nakayama, M., Jin, K., Graham, S.H., Simon, R.P., 1998. Transient global ischemia triggers expression of the DNA damage-inducible gene GADD45 in the rat brain. J. Cereb. Blood Flow Metab. 18, 646–657. Cheng, J., Alkayed, N.J., Hurn, P.D., 2007. Deleterious effects of dihydrotestosterone on cerebral ischemic injury. J. Cereb. Blood Flow Metab. 27, 1553–1562. Chu, X.P., Xiong, Z.G., 2012. Physiological and pathological functions of acid-sensing ion channels in the central nervous system. Curr. Drug Targets 13, 263–271.

98

brain research 1564 (2014) 85–100

Cox-Limpens, K.E.M., Vles, J.S.H., Schlechter, J., Zimmermann, L.J. I., Strackx, E., Gavilanes, A.W.D., 2013. Fetal brain genomic reprogramming following asphyctic preconditioning. BMC Neurosci. 14, 61. Curtis, R.K., Oresic, M., Vidal-Puig, A., 2005. Pathways to the analysis of microarray data. Trends Biotechnol. 23, 429–435. DeGracia, D.J., Szymanski, J.J., Marshall MK, J.R., 2008. Translation arrest and ribonomics in post-ischemic brain: layers and layers of players. J. Neurochem. 106, 2288–2301. Ding, Z.M., Wu, B., Zhang, W.Q., Lu, X.J., Lin, Y.C., Geng, Y.J., Miao, Y.F., 2012. Neuroprotective effects of ischemic preconditioning and postconditioning on global brain ischemia in rats through the same effect on inhibition of apoptosis. Int. J. Mol. Sci. 13, 6089–6101. Dobolyi, A., Vincze, C., Pal, G., Lovas, G., 2012. The neuroprotective functions of transforming growth factor beta proteins. Int. J. Mol. Sci. 13, 8219–8258. Downey, JMDA, Cohen, M.V., 2007. Signaling pathways in ischemic preconditioning. Heart Fail Rev. 12, 181–188. Durukan, A., Stribian, D., Tatlisumak, T., 2008. Rodent models of ischemic stroke: a useful tool for stroke drug development. Curr. Pharm. Des. 14, 359–370. Feng, Z., Sasik, D.D.R., Patel, H.H., Drummond, J.C., Patel, P.M., 2007. Pathway and gene ontology based analysis of gene expression in a rat model of cerebral ischemic tolerance. Brain Res. 1177, 103–123. Fortunel, N.O., Otu, H.H., Ng, H.H., Chen, J., Mu, X., Chevassut, T., Li, X., Joseph, M., Bailey, C., Hatzfeld, J.A., 2003. Comment on “ ‘Stemness’: transcriptional profiling of embryonic and adult stem cells”. Science 302, 393. Ge, P., Luo, Y., Liu, C.L., Hu, B., 2007. Protein aggregation and proteasome dysfunction after brain ischemia. Stroke 38, 3230–3236. Gibson, C.L., 2013. Cerebral ischemic stroke: is gender important? J. Cereb. Blood Flow Metab. 33, 1355–1361. Gilbert, R.W., Blanchard, C.W., Mullen, M.E., Currie, K.L., Robertson HA, R.W., 2003. DNA microarray analysis of hippocampal gene expression measured twelve hours after hypoxia-ischemia in the mouse. J. Cereb. Blood Flow Metab. 23, 1195–1211. Grond-Ginsbach, C., Hummel, M., Wiest, T., Horstmann, S., Pfleger, K., Hergenhahn, M., Hollstein, M., Mansmann, U., Grau, A.J., Wagner, S., 2008. Gene expression in human peripheral blood mononuclear cells upon acute ischemic stroke. J. Neurol. 225, 723–731. Guhaniyogi, J., Brewer, G., 2001. Regulation of mRNA stability in mammalian cells. Gene 265, 11–23. Gustavsson, M., Wilson, M.A., Mallard, C., Rousset, C., Johnston, M.V., Hagberg, H., 2007. Global gene expression in the developing rat brain after hypoxic preconditioning: involvement of apoptotic mechanisms. Pediatr. Res. 61, 444–450. Gut, P., Verdin, E., 2013. The nexus of chromatin regulation and intermediary metabolism. Nature 502, 489–498. Honkaniemi, J., Sharp, F.R., 1996. Global ischemia induces immediate-early genes encoding zinc finger transcription factors. J. Cereb. Blood Flow Metab. 16, 557–565. Huang, D.W., Sherman, B.T., Lempicki, R.A., 2009. Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res. 37, 1–13. Jin, K.L., Mao, X.O., Eshoo, M.W., Nagayama, T., Minami, M., et al., 2001. Microarray analysis of hippocampal gene expression in global cerebral ischemia. Ann. Neurol. 50, 93–103. Kanemitsu, H., Kirino, T., Nakagomi, T., Tsujita, Y., Iwamoto, T., Tomich, J.M., Tamura, A., 1994. Key of induced tolerance to ischaemia in gerbil hippocampal CA1 is not at transcriptional

level of hsp70 gene: in situ hybridization of hsp70 mRNA. Neurol. Res. 16, 209–212. Kawahara, N., Wang, Y., Mukasa, A., Furuya, K., Shimizu, T., Hamakubo, T., Aburatani, H., Kodama, T., Kirino, T., 2004. Genome-wide gene expression analysis for induced ischemic tolerance and delayed neuronal death following transient global ischemia in rats. J. Cereb. Blood Flow Metab. 24, 212–223. Kelly-Hayes, M., Beiser, A., Kase, C.S., Scaramucci, A., D’Agostino, R.B., Wolf, P.A., 2003. The influence of gender and age on disability following ischemic stroke: the Framingham study. J. Stroke Cerebrovasc. Dis. 12, 119–126. Kim, H.J., Rowe, M., Ren, M., Hong, J.S., Chen, P.S., Chuang, D.M., 2007. Histone deacetylase inhibitors exhibit antiinflammatory and neuroprotective effects in a rat permanent ischemic model of stroke: multiple mechanisms of action. J. Pharmacol. Exp. Ther. 321, 892–901. Kim, J.Y., Yenari, M.A., 2013. The immune modulating properties of the heat shock proteins after brain injury. Anat. Cell Biol. 46, 1–7. Kirchengast, S., Hartmann, B., 2009. The male disadvantage hypothesis reconsidered: is there really a weaker sex? An analysis of gender differences in newborn somatometrics and vital parameters. J. Life Sci. 1, 63–71. Kitagawa, K., 2012. Ischemic tolerance in the brain: endogenous adaptive machinery against ischemic stress. J. Neurosci. Res. 90, 1043–1054. Kostandy, B.B., 2012. The role of glutamate in neuronal ischemic injury: the role of spark in fire. Neurol. Sci. 33, 223–237. Kovalska, M., Kovalska, L., Pavlikova, M., Janickova, M., Mikuskova, K., Adamkov, M., Kaplan, P., Tatarkova, Z., Lehotsky, J., 2012. Intracellular signaling MAPK pathway after cerebral ischemia-reperfusion injury. Neurochem. Res. 37, 1568–1577. Laabich, A., Li, G., Cooper, N.G., 2001. Characterization of apoptosis-genes associated with NMDA mediated cell death in the adult rat retina. Brain Res. Mol. Brain Res. 91, 34–42. Li, L., Lundkvist, A., Andersson, D., Wilhelmsson, U., Nagai, N., Pardo, A.C., Nodin, C., Stahlberg, A., Aprico, K., Larsson, K., Yabe, T., Moons, L., Fotheringham, A., Davies, I., Carmeliet, P., Schwartz, J.P., Pekna, M., Kubista, M., Blomstrand, F., Maragakis, N., Nilsson, M., Pekny, M., 2008. Protective role of reactive astrocytes in brain ischemia. Protect. Role React. Astrocyt. Brain Ischem. 28, 468–481. Li, Y., Chopp, M., Powers, C., Jiang, N., 1997. Apoptosis and protein expression after focal cerebral ischemia in rat. Brain Res. 765, 301–312. Liang, J.T.H., Doi, Y., Kawanokuchi, J., Sonobe, Y., Jin, S., Yawata, I., Li, H., Yasuoka, S., Mizuno, T., Suzumura, A., 2008. Excitatory amino acid transporter expression by astrocytes is neuroprotective against microglial excitotoxicity. Brain Res. 1210, 11–19. Lipton, P., 1999. Ischemic cell death in brain neurons. Physiol. Rev. 79, 1431–1568. Loidl, C.F., Gavilanes, A.W.D., Van Dijk, E.H.J., Vreuls, W., Blokland, A., Vles, J.S., Steinbusch, H.W., Blanco, C.E., 2000. Effects of hypothermia and gender on survival and behavior after perinatal asphyxia in rats. Physiol. Behav. 68, 263–269. Lu, X.C.M., Williams, A.J., Yao, C., Berti, R., Hartings, J.A., et al., 2004. Microarray analysis of acute and delayed gene expression profile in rats after focal ischemic brain injury and reperfusion. J. Neurosci. Res. 77, 843–857. Marini, A.M., Jiang, X., Wu, X., Guo, Z., Mattson, M.P., Blondeau, N., Novelli, A., Lipsky, R.H., 2007. Preconditioning and neurotrophins: a model for brain adaptation to seizures, ischemia and other stressful stimuli. Amino Acids 32, 299–304. Mirabelli-Badenier, M., Braunersreuther, V., Viviani, G.L., Dallegri, F., Quercioli, A., Veneselli, E., Mach, F., Montecucco, F., 2011. CC

brain research 1564 (2014) 85–100

and CXC chemokines are pivotal mediators of cerebral injury in ischaemic stroke. Thromb. Haemost. 105, 409–420. Mukherjee, D., Patil, C.G., 2011. Epidemiology and the global burden of stroke. World Neurosurg. 76, S85–S90. Naeye, R.L., Wright, B.L., Blanc, D.L., Tatter DL, W.A., 1971. Neonatal mortality, the male disadvantage. Pediatrics, 48. Nakagomi, T., Kirino, T., Kanemitsu, H., Tsujita, Y., Tamura, A., 1993. Early recovery of protein synthesis following ischemia in hippocampal neurons with induced tolerance in the gerbil. Acta Neuropathol. 86, 10–15. Oh, S.H., Kim, O.J., Shin, D.A., Song, J., Yoo, H., Kim, Y.K., Kim, J.K., 2012. Alteration of immunolohic responses on peripheral blood in the acute phase of ischemic stroke: a blood genomic profiling study. J. Neuroimmunol. 249, 60–65. Okoniewski, M.J., Hey, Y., Pepper, S.D., Miller, C.J., 2007. High correspondence between Affymetrix exon and standard expression arrays. BioTechniques 42, 181–185. Perez, A., Ritter, S., Brotschi, B., Werner, H., Caflisch, J., Martin, E., Latal, B., 2013. Long-term neurodevelopmental outcome with hypoxic-ischemic encephalopathy. J. Pediatr. 163, 454–459. Pignataro, G.C.O., Esposito, E., Sirabella, R., Di Renzo, G., Annunziato, L., 2011. ASIC1a contributes to neuroprotection elicited by ischemic preconditioning and postconditioning. Int. J. Physiol. Pathopgysiol. Pharmacol. 3, 1–8. Pikula, A., Beiser, A.S., Chen, T.C., Preis, S.R., Vorgias, D., Decarli, C., Au, R., Kelly-Haves, M., Kase, C.S., Wolf, P.A., Vasan, R.S., Seshadri, S., 2013. Serum brain-derived neurotrophic factor and vascular endothelial growth factor levels are associated with risk of strok and vascular brain injury: Framingham study. Stroke 44, 2768–2775. Quintero, O.A., Moore, J.E., Unrath, W.C., Manor, U., Salles, F.T., Grati, M., Kachar, B., Yengo, C.M., 2010. Intermolecular autophosphorylation regulates myosin IIIa activity and localization in parallel actin bundles. J. Biol. Chem. 285, 35770–35782. Ramos-Cejudo, J., Gutie´rrez-Ferna´ndez, M., Rodrı´guez-Frutos, B., Alcaide, M.E., Sa´nchez-Cabo, F., Dopazo, A., Dı´ez-Tejedor, E., 2012. Spatial and temporal gene expression differences in core and periinfarct areas in experimental stroke: a microarray analysis. PLoS One 7, e52121. Rickhag, M., Teilum, M., Wieloch, T., 2007. Rapid and long-term induction of effector immediate early genes (BDNF, Neuritin and Arc) in peri-infarct cortex and dentate gyrus after ischemic injury in rat brain. Brain Res. 1151, 203–210. Salvador, J.M., Brown-Clay, J.D., Fornace Jr., A.J., 2013. Gadd45 in stress signaling, cell cycle control, and apoptosis. Adv. Exp. Med. Biol., 793. Scha¨fer, A., 2013. Gadd45 proteins: key players of repair-mediated DNA methylation. Adv. Exp. Med. Biol. 793, 35–50. Schmidt-Kastner, R., Zhang, B., Belayev, L., Khoutorova, L., et al., 2002. DNA microarray analysis of cortical gene expression during early recirculation after focal brain ischemia in the rat. Mol. Brain Res. 108, 81–93. Sharp, F.R., Xu, H., Lit, L., Walker, W., Pinter, J., Apperson, M., Verro, P., 2007. Genomic profiles of stroke in blood. Stroke 38, 691–693. Siegel, C., Turtzo, C., McCullough, L.D., 2010. Sex differences in cerebral ischemia: possible molecular mechanisms. J. Neurosci. Res. 88, 2765–2774. Smith, M.L., Chen, I.T., Zhan, Q., Bae, I., Chen, C.Y., Gilmer, T.M., Kastan, M.B., Fornace, A.J., 1994. Interaction of the p53-regulated protein Gadd45 with proliferating cell nuclear antigen. Science 266, 1376–1380. Soriano, M.A., Tessier, M., Certa, U., Gill, R., et al., 2000. Parallel gene expression monitoring using oligonucleotide probe arrays of multiple transcripts with an animal model of focal ischemia. J. Cereb. Blood Flow Metab. 20, 1045–1055.

99

Stenzel-Poore, M.P., Stevens, S.L., Xiong, Z., Lessov, N.S., Harrington, C.A., Mori, M., Meller, R., et al., 2003. Effect of ischaemic preconditioning on genomic response to cerebral ischaemia: similarity to neuroprotective strategies in hibernation and hypoxia-tolerant states. Lancet 362, 1028–1037. Stenzel-Poore, M.P., Stevens, S.L., King, J.S., Simon, R.P., 2007. Preconditioning reprograms the response to ischemic injury and primes the emergence of unique endogenous neuroprotective phenotypes: a speculative synthesis. Stroke 38, 680–685. Subramanian, A., Tamayo, P., Mootha, V.K., Mukherjee, S., Ebert, B.L., Gilette, M.A., Paulovich, A., Pomeroy, S.L., Golub, T.R., Lander, E.S., Mesirov, J.P., 2005. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl. Acad. Sci. 102, 155545–155550. Tang, S.C., Arumugam, T.V., Xu, X., Cheng, A., Mughal, M.R., Jo, D.G., Lathia, J.D., Siler, D.A., Chigurupati, S., Ouyang, X., Magnus, T., Camandola, S., Mattson, M.P., 2007. Pivotal role for neuronal Toll-like receptors in ischemic brain injury and functional deficits. Proc. Natl. Acad. Sci. U. S. A. 104, 13798–13803. Tang, Y., Lu, A., Aronow, B.J., Sharp, F.R., 2001. Blood genomic responses differ after stroke, seizures, hypoglycemia, and hypoxia: blood genomic fingerprints of disease. Ann. Neurol. 50, 699–707. Tang, Y., Lu, A.G., Aronow, B.J., Wagner, K.R., Sharp, F.R., 2002. Genomic responses of the brain to ischemic stroke, intracerebral haemorrhage, kainate seizures, hypoglycemia and hypoxia. Eur. J. Neurosci. 15, 1937–1952. Tang, Y., Pacary, E., Freret, T., Divoux, D., Petit, E., SchumannBard, P., Bernaudin, M., 2006. Effect of hypoxic preconditioning on brain genomic response before and following ischemia in the adult mouse: identification of potentials neuroprotective candidates for stroke. Neurobiol. Dis. 21, 18–28. Tsai, M.C.T.K., Overstreet-Wadiche, L., Wadiche, J.I., 2012. Neuronal glutamate transporters regulate glial excitatory transmission. J. Neurosci. 32, 1528–1535. Ueda, K., Arakawa, H., Nakamura, Y., 2003. Dual-specificity phosphatase 5 (Dusp5) as a direct transcriptional target of tumor suppressor. Oncogene 22. VanGilder, R.L., Huber, J.D., Rosen, C.L., Barr, T.L., 2012. The transcriptome of cerebral ischemia. Brain Res. Bull. 88, 313–319. Vendrell, M., Curran, T., Morgan, J.I., 1993. Glutamate, immediateearly genes, and cell death in the nervous system. Ann. N. Y. Acad. Sci. 679, 132–141. Vexler, Z.S., Tang, X.N., Yenari, M.A., 2006. Inflammation in adult and neonatal stroke. Clin. Neurosci. Res. 6, 293–313. Wang, L., Zhou, C., Wang, Z., Liu, J., Jing, Z., Zhang, Z., Wang, Y., 2011a. Dynamic variation of genes profiles and pathways in the hippocampus of ischemic mice: a genomic study. Brain Res., 1372. Wang, Y., Cao, M., Liu, A., Di, W., Zhao, F., Tian, Y., Jia, J., 2013. Changes of inflammatory cytokines and neurotrophins emphasized their roles in hypoxic-ischemic brain damage. Int. J. Neurosci. 123, 191–195. Wang, Y.C., Lins, S., Yang, Q.W., 2011b. Toll-like receptors in cerebral ischemic inflammatory injury. J. Neuroinflammation http://dxdoi.org/10.1186/1742-2094-8-134. World Health Organization, WHO factsheet n310, The top 10 causes of death, 2011, updated July 2013. World Health Organization, Partnership for Maternal Newborn and Child Health, 2010. Countdown to 2015 Decade Report (2000-2010). In: Countdown to 2015. Vol., ed.\widehateds., Washington D.C. Yakubov, E., Gottlieb, M., Gil, S., Dinerman, P., Fuchs, P., Yavin, E., 2004. Overexpression of genes in the CA1 hippocampus region

100

brain research 1564 (2014) 85–100

of adult rat follwing episodes of global ischemia. Mol. Brain Res. 127, 10–26. Zhan, X., Ander, B.P., Jickling, G., Turner, R., Stamova, B., Xu, H., Liu, D., Davis, R.R., Sharp, F.R., 2010. Brief focal cerebral ischemia that simulates transient ischemic attacks in humans regulates gene expression in rat peripheral blood. J. Cereb. Blood Flow Metab. 30, 110–118. Zhang, J.Z.G., Chopp, M., Gautam, S., Zaloga, C., Zhang, R.L., HHHW, Schmidt, Pollock, J.S., Forstemann, U., 1994. Upregulation of neuronal nitric-oxide synthase and messengerRNA, and selective sparing of nitric-oxide synthase-

containing neurons after focal cerebral-ischemia in rat. Brain Res. 654, 85–95. Zheng, W.H., Kar, S., Dore¨, S., Quirion, R., 2000. Insulin-like growth factor-1 (IGF-1): a neuroprotective trophic factor acting via the Akt kinase pathway. J. Neural Transm. 60 (Suppl), S261–S272. Zheng, Z., Kim, J.Y., Ma, H., Lee, J.E., Yenari, M.A., 2008. Antiinflammatory effects of the 70kDa heat shock protein in experimental stroke. J. Cereb. Blood Flow Metab. 28, 53–63. Zheng, Z., Yenari, M.A., 2006. The application of Hsp70 as a target for gene therapy. Front Biosci. 11, 699–707.