Acute hyperglycemia during hepatic ischemia amplified the inflammatory response and resulted in elevated transaminase concentrations following I/R. The elevated serum glucose concentration at the start of ischemia seemed to be responsible for the increase in injury, as there was a strong correlation between serum glucose concentrations before ischemia and transaminase concentration after 4 h of reperfusion. We used a transient model of hyperglycemia starting only shortly before ischemia. Serum glucose concentrations were still higher in hyperglycemic animals at the end of the 4 h reperfusion period, albeit the graph (Fig. ) clearly demonstrates a declining trend.
Hepatic I/R has been reported to result in hepatocyte death by two different pathways, necrosis and apoptosis. Whether apoptotic or necrotic cell death predominates following liver I/R has been the subject of debate. Based on terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining, it was suggested that sinusoidal endothelial cells and then subsequently hepatocytes undergo apoptosis but rarely necrosis following 60 min of liver ischemia.15
However, a later study applying a very similar ischemia model found only few apoptotic cells and predominantly necrosis following 60 min of ischemia when combining TUNEL with morphological criteria.16
A subsequent review emphasized that apoptosis and necrosis share features and mechanisms that can make discrimination between both forms of cell death very challenging. In particular, the TUNEL assay is not suited to differentiate between necrosis and apoptosis, since DNA fragmentation was reported in apoptosis as well as necrosis.17
In the present study, histological assessment of the liver samples after 4 h of reperfusion could not reliably distinguish between apoptotic and necrotic cell death. A longer reperfusion could potentially facilitate histological analysis, but 4 h of reperfusion was chosen to enable the detection of inflammatory mediators. As a result, the histological scores were basically identical in both experimental groups, in spite of serologic evidence for increased necrotic cell death in hyperglycemic animals. The higher transaminase concentrations measured in the hyperglycemic group after 4 h of reperfusion reflect increased cellular breakdown due to necrosis with release of intracellular enzymes. Apoptosis maintains the barrier function of the cell membrane and would contribute only to a minor extent to elevated transaminase concentrations. Caspase-3 activation is considered the most reliable method for the detection of apoptosis.14
We assessed caspase-3 activation to quantify the amount of apoptotic cell death and found higher apoptosis scores in the control group. The lower caspase-3 activation in the hyperglycemic group may be further evidence that necrosis, not apoptosis, is the preferential form of cell death in hyperglycemic conditions.
Hyperglycemia per se is known to increase oxidative stress and to cause a proinflammatory state.8
Furthermore, hyperglycemia has been shown to amplify the inflammatory response caused by stressors such as LPS administration.18
Our results support the hypothesis that the mechanisms responsible for increased ischemic injury by hyperglycemia are the amplification of oxidative stress and of the inflammatory response normally seen with I/R. The increased concentration of nitrotyrosine containing protein is an established marker for severe oxidative stress. Reactive oxygen species, such as the superoxide radical, react with NO to form the more potent peroxynitrite species, which then subsequently nitrate tyrosine residues of proteins, leading to inactivation of key cellular proteins, DNA damage, and eventually cell death.19
While Kupffer cell-induced oxidative stress is considered the first step in I/R injury,20
it is followed by a profound inflammatory response that is largely responsible for the extent of I/R injury. This inflammatory response culminates in the hepatic accumulation of neutrophils, which directly damage hepatocytes by releasing oxidants and proteases. The MPO assay confirmed an increased neutrophil infiltration in the liver tissue of hyperglycemic animals. This neutrophil migration and infiltration is initiated by the production and release of cytokines such as tumor necrosis factor alpha and interleukin-6. Earlier studies demonstrated that hyperglycemia enhances cytokine production in response to stress.21
A surprising finding of the present investigation was the downregulation of HSP32 and HSP70 in hyperglycemic animals undergoing I/R. Hepatic I/R normally results in upregulation of HSPs, and the observed effects in livers from hyperglycemic animals differ distinctly from the situation in kidneys7
and the brain,22
where hyperglycemic I/R injury is associated with an increased activation of HSPs. We could demonstrate in sham experiments that hyperglycemia alone was not responsible for the downregulation of HSPs (data not shown) but that the combination of hyperglycemia and I/R is required to block or even suppress HSP activation. Since HSPs are one of the most potent protective mechanisms against I/R injury, it can be assumed that their suppression in hyperglycemic I/R contributes to the increased injury during acute hyperglycemia. Inhibition of HSP activation in response to ischemia has so far not been described in other organs and may represent a liver-specific (mal-)adaptation to hyperglycemia: It has been described before that diabetes does inhibit hepatic HSP70 activation by heat stress,23
although subsequent studies did not confirm this finding.24
The mechanism responsible for the downregulation of HSPs remains to be defined. The expression of the heat shock genes encoding the different HSPs is regulated by heat shock transcription factors (HSFs), which are normally bound to HSPs within the cytosol. When cells are exposed to stress, HSFs are phosphorylated and form trimers that enter the nucleus and bind the heat shock elements located within the promoter of heat shock genes, thus initiating increased expression of HSPs.26
It has been hypothesized that, in diabetes, the activation of HSF is inhibited in insulin-sensitive tissue.27
In type 2 diabetic primates, livers had reduced HSP70 and HSP90 tissue concentrations that were related to 50% lower levels of the transcription factor heat shock factor 1.28
But again, these results are challenged by a study that showed similar heat shock factor 1 content in livers from control and streptozotocin treated rats following heat stress.24
Further interventional studies with activation of HSPs are planned to show whether the suppression of HSP activation is responsible for the worsened injury during hyperglycemia and whether activation of HSPs is capable of reversing such detrimental effects.