Background and Purpose-After focal cerebral ischemia, depending on its localization and extent, secondary neuronal damage may occur that is remote from the initial lesion. In this study differences in secondary damage of the ventroposterior thalamic nucleus (VPN) and the reticular thalamic nucleus (RTN) were investigated with the use of different ischemia models. Methods-Transient middle cerebral artery occlusion (MCAO) leads to cortical infarction, including parts of the basal ganglia such as the globus pallidus, and to widespread edema. Photothrombotic ischemia generates pure cortical infarcts sparing the basal ganglia and with only minor edema. Neuronal degeneration was quantified within the ipsilateral RTN and VPN 14 days after ischemia. Glial reactions were studied with the use of immunohistochemistry. Results-MCAO resulted in delayed neuronal cell loss of the ipsilateral VPN and RTN. Glial activation occurred in both nuclei beginning after 24 hours. Photothrombotic ischemia resulted in delayed neuronal cell loss only within the VPN. Even 2 weeks after photothrombotic ischemia, glial activation could only be seen within the VPN. Conclusions-Pure cortical infarcts after photothrombotic ischemia, without major edema and without effects on the globus pallidus of the basal ganglia, only lead to secondary VPN damage that is possibly due to retrograde degeneration. MCAO, which results in infarction of cortex and globus pallidus and which causes widespread edema, leads to secondary damage in the VPN and RTN. Thus, additional RTN damage may be due to loss of protective GABAergic input from the globus pallidus to the RTN or due to the extensive edema. Retrograde degeneration is not possible because the RTN, in contrast to the VPN, has no efferents to the cortex.

In brains of the rabbit, pig, and human, expression of the high‐affinity Na+‐d‐glucose cotransporter SGLT1 and of the protein RS1, which alters the activity of SGLT1, was demonstrated. In situ hybridization showed that SGLT1 and RS1 are transcribed in pyramidal cells of brain cortex and hippocampus and in Purkinje cells of cerebellum. In neurons of pig brain SGLT1 protein was demonstrated by western blotting with synaptosomal membranes and by immunohistochemistry, which showed SGLT1 in pyramidal and Purkinje cells. To test whether SGLT1 in neurons may be activated during increased d‐glucose consumption, an epileptic seizure was induced in rat brain, and the uptake of specific nonmetabolized substrates of SGLT1 {[14C]methyl‐α‐d‐glucopyranoside ([14C]AMG)} and of Na+‐independent transporters {2‐deoxy‐d‐[14C]glucose([14C]2‐DG)} was analyzed by autoradiography. During the seizure the uptake of AMG and 2‐DG was increased in the focus. Within two hours after the seizure 2‐DG uptake in the focus returned to normal. In contrast, the AMG uptake in the focus area was still increased 1 day later. The data show that the high‐affinity Na+‐d‐glucose cotransporter SGLT1 is expressed in neurons and can be up‐regulated.

Immunoreactivity of the Na+-D-glucose cotransporter SGLT1 was demonstrated in intracerebral capillaries of rat and pig. Immunostaining suggested that SGLT1 is located in the luminal membrane of the endothelial cells and in intracellular vesicles. Using in situ hybridization, SGLT1 mRNA was not detectable in intracerebral capillaries of non-treated or sham-operated Wistar rats. However, 1 day after a transient occlusion of the right middle cerebral artery, SGLT1 mRNA was detected in capillaries of both brain hemispheres. Expression of SGLT1 was also demonstrated in primary cultures of capillary endothelial cells from pig using polymerase chain reaction after reverse transcription and western blotting. The data suggest that SGLT1 participates in transport of D-glucose across the blood-brain barrier and is upregulated after brain ischemia and reperfusion.