The protein response
ER-stress induced neuronal cell death has been suggested to play an important role in stroke (DeGracia and Montie 2004; Tajiri et al. 2004). In agreement, agents that counteract the ER stress, such as salubrinal, have been reported to decrease the damage after I/R injury in focal cerebral ischemia (Nakka et al. 2010). The presence of protein aggregates, indicating the presence of ER stress in a similar model to that used in this study, has been reported to appear 1 h after reperfusion (Hu et al. 2000; Truettner et al. 2009). However, the molecules related with the unfolded protein response (UPR) present different responses. Thus, some chaperone molecules (Hsp70, GRP78, Hsp60, GRP94) have been reported to increase progressively in the hippocampus in the first 24 h, while PDI levels remain unchanged (Truettner et al. 2009). Our data on PDI expression in the hippocampus of young animals are consistent with those of Truettner et al. (2009), and suggest that PDI levels are not modified by ischemic challenge. In contrast, PDI level modifications in the cerebral cortex indicated a different UPR in the hippocampus and cerebral cortex, fitting with the differential responses to the ischemia widely described for these structures, as mentioned below.
Our data on GRP78 expression in some structures of young animals are also consistent with the return to normal values at 24 h reported by Truettner et al. (2009). However, we still detected structures with injury-dependent increases in GRP78, indicating that the UPR is still detectable at the protein level 48 h after challenge. The different behavior of both PDI and GRP78 following the insult in young and aged animals reveals age-dependent differences in the I/R-induced UPR and confirms that this response is also detectable in aged animals 48 h after the ischemic insult.
Additional support for the different behavior of PDI and GRP78 expression was provided by treatment with the anti-inflammatory agent meloxicam. This treatment resulted in decreases in the levels of these molecules in the hippocampus and increases in the cerebral cortex. This suggests that the differential vulnerability reported for these structures (Vallet and Charpiot 1994; Jiang et al. 2004; Gee et al. 2006; Kumari Naga et al. 2007; Stanika et al. 2010) could be largely dependent on inflammation.
The mRNA response
Protein and mRNA responses do not necessarily behave in the same way; in fact, discrepancies are usually observed. Different mechanisms have been indicated to explain the relative utilization of a given mRNA, including translational inhibition or activation (Mitchell and Tollervey 2001; Wilusz et al. 2001; Gorospe et al. 2011). Discrepancies between protein and mRNA levels during hypoxia seem to depend mainly on mRNA turnover and translational control (Gorospe et al. 2011). Therefore, mRNA and protein behaviors mirror different processes happening in the cell, explaining many of their discrepancies. In addition, the possibility of amplifying mRNA molecules allows the quantification of molecular changes that are not detected at the protein level.
One of the first effects of the UPR is to block the translation of proteins; however, some mRNAs become preferentially translated (for reviews see Walter and Ron 2011; Korennykh and Walter 2012). Consistently, this report shows increases in the mRNA of some UPR-related gene mRNAs contrasting with the noticeable decrease described in a number of glutamatergic and GABAergic system genes following a similar 48 h I/R insult (Naidoo et al. 2008; Dos-Anjos et al. 2009; Montori et al. 2010a, b, c, 2012; Llorente et al. 2013). These data suggest that I/R results in a decrease in neurotransmission in both excitatory (glutamatergic) and inhibitory (GABAergic) systems, while it activates the UPR to reduce the cellular damage resulting from the insult. In this regards, assays with salubrinal in focal cerebral ischemia, that is, promoting the UPR by inhibiting the dephosphrylation of the eukaryotic translation initiation factor 2 subunit α (eIF2α), significantly decreases I/R-induced damage (Nakka et al. 2010).
The correlation between the UPR and delayed cell death has been previously indicated. In this regard, ischemic preconditioning has been reported to be a very effective way of preventing delayed neuronal death (Kato et al. 1994; Shamloo et al. 1999; Kirino 2002), and this has been reported to decrease protein aggregation (Liu et al. 2005). A post-conditioning treatment, with a series of mechanical interruptions of the reperfusion, has also been reported to reduce ischemic/reperfusion damage, with decreases in caspase 12 and CHOP expression, but increased expression of GRP78 (Yuan et al. 2011). These data support the notion of the UPR is a crucial factor in neuronal vulnerability. In this regard, our mRNA results suggest that the UPR and delayed cell death in hippocampus are greater than in cerebral cortex 48 h after challenge. These data suggest that the UPR may overcome the reticulum stress in some areas but not in others, which might be a clue to understanding the differential vulnerability of different brain regions to ischemia.
Previous in situ hybridization studies in young animals following global cerebral ischemia have shown increased mRNA transcripts in a number of chaperone proteins (GRP78, HERP and GRP94) following ischemic damage. The peak of these increases appears at 24 h and return to similar values to sham-operated animals at 48 h (Truettner et al. 2009). However, our results in young animals indicate that 48 h after challenge, there is still a strong increase in the mRNA levels of all genes analyzed, including CHOP, a widely used marker of apoptosis (Tajiri et al. 2004; Chan et al. 2011).
The role of CHOP has been analyzed in some ischemia studies. Thus, CHOP-knockout mice have been reported to present a decreased loss of neurons following ischemia compared to wild-type mice in bilateral common carotid occlusion. Primary hippocampal neurons from CHOP−/− mice also showed greater resistance to hypoxia/reoxygenation-induced cell death (Tajiri et al. 2004). Ischemia-associated ER stress has been hypothesized to be induced predominantly through the CHOP-dependent signaling pathway in neurons of the hippocampus (Osada et al. 2010). Our data are in agreement with this idea, and also support the hypothesis that CHOP gene expression and ER stress are involved in differences in neuronal vulnerability. In this regard, the CHOP response observed in this study is parallel to that observed for GRP78 and GRP94, corroborating the differential response between the cerebral cortex and hippocampus in young ischemic animals.
What is the effect of anti-inflammatory agents? CD11b and GFAP mRNA levels are used as microglial and astroglial markers, respectively, as well as indicators of inflammation (Massaro et al. 1990; McGeer and McGeer 1995; Kim and de Vellis 2005; Yatsiv et al. 2005; Giovannoni 2006; Hamby et al. 2007). Meloxicam treatment, following the ischemic insult, has been reported to result in a different response in the hippocampus and the CX when mRNA levels for CD11b and GFAP are analyzed. Thus, treatment with meloxicam results in a normalization of the levels of these inflammatory markers in the hippocampus (i.e., they are similar to those in sham-operated animals), although in the CX, they are still significantly higher than in sham-operated controls (Montori et al. 2010a). Meloxicam treatment in this model has also been shown to lessen the ischemia-induced decreases in mRNA levels in a number of glutamatergic and GABAergic genes (Montori et al. 2010a, b, c, 2012; Llorente et al. 2013). Although more studies are needed using different anti-inflammatory agents, different time points of analysis and different doses, our data show that meloxicam treatment decreases CHOP and chaperone increases following ischemic insult, suggesting that reducing inflammation could result in a decrease in mortality. In our hands, treatment with meloxicam revealed that UPR gene transcripts decreased in a similar way in the hippocampus and the cerebral cortex. These data suggest that the differential vulnerability to ischemic damage correlates with the extent of inflammation, but it seems that UPR gene response is independent of this different vulnerability.
Our study also reveals that UPR gene transcripts levels were similar in young and aged sham-operated animals; however, the ischemic insult elicited noticeable age-dependent differences in the UPR. In addition, differences observed between structures in young animals were less noticeable in aged animals, and the response to I/R insult was lower. Thus, our CHOP data support an age-dependent decrease in delayed cell death, and this response correlates with a dampened UPR. This idea is also supported by our data on the chaperone protein levels. In fact, our results support the idea that aged animals have a noticeable decrease in their ability to activate the ischemia-induced UPR.