Long-term gene expression changes in the cortex following cortical ischemia revealed by transcriptional profiling

Abstract

Cerebral ischemia evokes changes in gene expression time-dependently after the ischemic event. Most studies on transcriptional changes following ischemia have centered on relatively early postischemic time points, and detected multiple genes relevant to neuronal cell death. However, functional outcome after ischemia depends critically on adaptations of the postischemic brain. Plasticity may derive from network-inherent changes, or from the formation of new nerve cells in the CNS. We have screened for gene expression changes up to 3 weeks following a limited photothrombotic cortical insult in the rat sensorimotor cortex by using the sensitive restriction-mediated differential display (RMDD) technique. A high number of genes were detected as induced at early or intermediate time points in the ipsi- and contralateral cortex (6 and 48 h). Unexpectedly, at the late time point examined (3 weeks), we still detected 40 genes that were changed in their expression. We further characterized the expression of two genes linked to neurogenesis (nestin and stathmin), and two genes likely involved in reconfiguring neuronal networks (semaphorin VIa and synaptotagmin IV). Conclusively, our data highlight the degree of long-term transcriptional changes in the cortex after ischemia, and provide insight into functional pathways of relevance for compensatory recovery mechanisms in neural networks.

Introduction

The cortex has a remarkable ability to adapt to changes in processing requirements throughout life and after injury. In stroke patients suffering from small cortical lesions, a partial or complete recovery of function can often be observed within weeks. This poststroke plasticity is most densely studied in the sensorimotor cortex in humans and in animal models. In animals, landmark studies have shown remapping and expansion to neighboring regions of functional motor cortex areas after injury (Nudo and Milliken, 1996, Nudo et al., 1996a, Nudo et al., 1996b). However, ipsilateral cortical regions distant from the injury also undergo functional reorganization (Frost et al., 2003). In human stroke patients, a common theme is the reduction of laterality in processing as defined by functional imaging (Cramer, 2004). Indeed, the amount of contralateral processing correlates with improved recovery for motor performance (Butefisch et al., 2005). From these and a number of other important studies, it thus appears that the brain can to some extent, and likely use-dependently, relocate functional processing to infarct-adjacent regions, but also to the homotopic contralateral cortex.

What are the structural correlates of this functional plasticity after stroke? Principally, brain plasticity after stroke can originate by changes in existing neurons or networks (e.g. by rewiring and reprogramming networks), or by the generation of new neurons (i.e. neurogenesis). After stroke, increases in dendritic arborization, the number of dendritic spines and synapses, and expansion of the active zone of existing synapses have been described particularly in layer V of the cortex (see Keyvani and Schallert, 2002 for review). These changes also appear partially use-dependent. One important, but very little studied aspect of the reorganization of the poststroke brain refers to reshaping of connections that require axonal outgrowth and pathfinding processes. There are indeed indications that a substantial burst in axonal outgrowth occurs after stroke both for local, intracortical projections as well as long distance, interhemispheric projections (Carmichael, 2003).

The second source for enhanced recovery after cerebral ischemia is neurogenesis. Cerebral ischemia activates adult precursors that start migrating towards lesioned areas, where they may incorporate into neuronal networks (Nakatomi et al., 2002), for review see (Kokaia and Lindvall, 2003). Although causal relationships of increased neurogenesis and behavioral outcome are technically very difficult to establish, an overwhelming amount of literature has shown strong correlations between increased neurogenesis obtained by different means and improved functional parameters. At present, the mechanisms are unclear by which these stem cells contribute to stroke recovery.

Therefore, a number of structural correlates of poststroke have been established. However, there is much less data concerning the question which molecular changes govern these restructuring events, and which might potentially serve as entry points for pharmacological interventions.

The availability of transcriptional profiling technologies has made it possible to search for novel genes involved in disease processes that may uncover novel pathophysiological processes (for review, see Scheel et al., 2002). We and many other groups have studied gene expression changes after focal cerebral ischemia models such as middle cerebral artery occlusion (MCAO), and have defined a large repertoire of genes transcriptionally altered after ischemia (e.g. Schneider et al., 2004b, Soriano et al., 2000, Trendelenburg et al., 2002). A high number of genes that are subject to transcriptional induction after ischemia have been functionally studied and shown to influence neuronal survival and infarct size (e.g. Martin-Villalba et al., 1999, Potrovita et al., 2004, Schneider et al., 1999). While this has proven the overwhelming influence of transcription-dependent processes in determining acute tissue damage, gene expression changes in the recovery phase after stroke with potential effects on cortical plasticity have not been studied in great detail. There are only scarce reports on early induction (or repression) of genes on the contralateral hemisphere, which are thought to be involved in plasticity processes (Stroemer et al., 1995). Only a few genes were characterized that are regulated contralateral to a stroke and are possibly linked to plasticity processes (Keyvani et al., 2000, Witte and Stoll, 1997). Indeed, to our knowledge, there is only one study that systematically searched for transcriptionally altered genes at a longer time interval after cortical stroke (Keyvani et al., 2002). We have therefore undertaken a systematic gene expression profiling study to learn more about molecular mechanisms that might drive functional and structural changes in the cortex after stroke.

We postulated that genes implied in plasticity and recovery processes are characterized by (1) their persistent regulation over long time intervals after ischemia, or their induction in a delayed mode after ischemia, and/or (2) their regulation in the periinfarct area, or in the contralateral homotopic cortex. In order to study these events, we chose the photothrombotic cortical ischemia model with small and defined lesions that guarantee a good postischemic recuperation implying active recovery mechanisms plus a highly predictable lesion size and localization (Markgraf et al., 1993, Wester et al., 1995, Wood et al., 1996). Samples were taken from cortical areas adjacent to the photothrombosis-induced lesion, and corresponding sites from the contralateral cortex, and studied up to 3 weeks following ischemia. As method for transcriptional profiling, we applied a sensitive fragment display technique, restriction-mediated differential display (RMDD) (Schneider et al., 2004b). We report on a number of previously unknown transcriptionally altered genes, and discuss their possible function with regard to regeneration processes after stroke.