Abstract
Density of the mesothelial cell populations in vivo, exposed to high glucose or to 7.5% polyglucose for periods of 2 hours, 15 days, and 30 days. Additional samples were taken after 7 days and 30 days of recovery, counted after the last intraperitoneal injection.
These changes were linked to the concentration of glucose, as indicated by the dose-dependent effect, and were unrelated either to the low pH and high osmolarity of the solution, or to the presence of lactate or bicarbonate buffer.
This view of the problem was confirmed in an additional study using a 4.25% glucose concentration in filter-sterilized fluid, prepared using the much more physiological and compatible Hanks Balanced Salt Solution (HBSS) (2). The new investigation corroborated the concept that sustained in vivo exposure of the monolayer to high concentrations of glucose resulted in a depopulated mesothelium composed mostly of senescent cells near the end of their life cycle. Senescence was defined on the basis of hypertrophy, multi-nucleation, low density of the population, substantially reduced prevalence of mitosis (Figure 2), and low viability (Figure 3), as well as on an increased number of cells in apoptosis.
Prevalence of mitosis in the mesothelial monolayer of mice treated with high glucose dialysis fluid or with 7.5% polyglucose. Notice the peak level at 2 hours in the high glucose group, and the nil levels seen in both groups of animals after 15 days and 30 days of exposure.
Cell viability evaluated using trypan blue staining. The prevalence of stained cells (no viable cells) is significantly higher in the polyglucose group than in mice treated with high glucose at each time interval, during the period of exposure as well as after a recovery period of 7 days counted after the last intraperitoneal injection.
Methods and Results
How can high glucose per se induce the observed changes?
We hypothesized that the answer could be found only in investigating some critical points in the population cell cycle (Shostak A, Wajsbrot V, Gotloib L. High glucose accelerates the life cycle of the in vivo exposed mesothelium. In preparation):
The G1 check point (expression of proliferative cell nuclear antigen [PCNA]).
The S-phase (tritiated thymidine incorporation).
Prevalence of mitosis.
Prevalence of apoptosis (death being the end-point of the life cycle).
Approximately 1% of cells seen in imprints taken from intact, unexposed mouse mesothelium showed PCNA expression. In samples of the monolayer taken from animals treated with HBSS for up to 24 hours, the corresponding values were not significantly different from those seen in the unexposed group of mice. On the other hand, intraperitoneal injections of 4.25% glucose solution induced a peak at 2 hours, 3 hours, and 4 hours after the injection. At those times, the prevalence of cells expressing PCNA activity was above 6%, significantly higher than the values seen in both control groups of mice (unexposed and injected with HBSS). Twenty-four hours after the first injection, as well as after 30 daily intraperitoneal instillations of the high glucose solution, PCNA expression was again low, and observed in even less than 1% of cells.
A similar pattern was seen when the incorporation of tritiated thymidine was evaluated. A significantly higher peak was detected between 2 hours and 4 hours after one intraperitoneal injection of the high glucose fluid. This peak was sustained at 24 hours, but reached nil levels after 15 and 30 daily injections of the same solution.
Prevalence of mitosis followed a pattern similar to that observed in previous studies of high glucose fluids by our laboratory: a peak at 2 hours after the injection, and a sharp reduction at 4 hours and at 15 days and 30 days, with values near zero (Figure 2). This part of the investigation was completed in an additional experiment evaluating the prevalence of mitosis during five consecutive days, 2 hours after one intraperitoneal injection of the experimental solution. The 2-hour peak was observed during the first three days, after which the replication capability of the exposed mesothelium became exhausted.
Prevalence of cells in apoptosis was substantially higher after 30 days of injections, as well as after a recovery period of 7 days free from exposure to the high glucose fluid.
So far, the data indicate that in vivo exposure of mesothelium to high glucose fluid induces a substantial acceleration of the cells’ life cycle. The result is a depopulated monolayer composed of hypertrophic and senescent cells near the end of their life span and showing significantly reduced viability (Figure 3). These changes may well derive from sustained oxidative stress, effected by the presence of early products of the Maillard reaction in peritoneal dialysate, or the formation of Amadori and advanced glycosylation end-products (AGEs), or the autoxidation of glucose, or any combination of these circumstances.
This working hypothesis is based on several lines of investigation showing that various cultured cell types exposed to high glucose concentrations (but lower than 20 mmol/L) reacted with substantial stimulation of cell proliferation for 24 – 72 hours, followed by a significant inhibitory effect when cells were maintained in a high-glucose medium for longer periods of time (3,4).
Discussion
Why this early acceleration and premature senescence of cells exposed to high concentration of glucose? How can glucose, apparently the most innocent component of dialysis solutions, become the determining factor behind the alterations detected in the life cycle of the mesothelial cell population?
Various cell types exposed to H2O2 display multi-nucleation, cellular hypertrophy, premature senescence, a reduced rate of proliferation, and a higher prevalence of apoptosis (5,6).
Low levels of oxidants potentiate growth signals and enhance cell proliferation; higher oxidant concentrations can block cell proliferation, which in turn results in premature senescence and the activation of the mechanisms leading to apoptotic cell death (7).
Mesothelial cells in culture significantly increased generation of H2O2 when exposed to high glucose concentrations (8).
A 24 - 27 mmol/L concentration of D-glucose—but not an equimolar concentration of L-glucose or raffinose—induces a substantial impairment in H2O2 degradation by human endothelial and mesothelial cells (9,10). Consequently, the D-glucose-specific metabolism and the subsequently increased generation of reactive oxygen species (ROS) are more likely responsible for the altered mesothelial cell cycle observed in our experiments.
Long-term (30 days) in vivo exposure of the mesothelial monolayer to a high-concentration (4.25%), mannitol-enriched (mannitol is a known oxygen radical scavenger), filter-sterilized dialysis fluid failed to induce any change in the life cycle of the cell population (11). A more recent study showed that pyruvate preserved the viability of cultured mesothelial cells exposed to H2O2 (12). Therefore, it may be hypothesized that D-glucose-dependent cell damage is a function of both the reduced production of scavengers and the increased generation of ROS not only by the mesothelial cells but also by the peritoneal phagocytes 13).
Additional sources of oxidative stress include:
Amadori products and AGEs already detected in mesothelium, blood vessels, and connective tissue of continuous ambulatory peritoneal dialysis (CAPD) patients (14).
Rapid glycation of the peritoneal tissue (2 hours).
Early products of the Maillard reaction in peritoneal dialysis fluid (15).
A similar experimental protocol was used to evaluate the effects of a commercially available 7.5% polyglucose solution (Icodial: Baxter Healthcare SA, Castlebar, Ireland) upon the exposed mesothelium. Under the effect of this osmotic agent, density of the cell population was substantially reduced during the period of exposure and failed to reach full recovery, even after 30 days (Figure 1).
Although observations made in the two groups of animals were not significantly different, mice exposed to polyglucose showed numerous areas of peritoneum—tens of square microns in size—that were deprived of mesothelial dressing. The polyglucose-treated animals, like the high glucose-treated animals, showed increased mean cell surface area, multi-nucleation, reduced prevalence of mitosis (Figure 2), and substantially reduced cell viability—this last even lower than that seen in mice exposed to high glucose (Figure 3). The nuclear cytoplasmic index was substantially higher in the polyglucose group of mice than in the control and glucose-treated mice. This finding, which could possibly derive from a dysplastic effect upon the exposed cell population, raises some questions about the safety of polyglucose and requires further investigation (16).
All the changes derived from exposure to poly-glucose can be explained, at least in theory, on the basis of severe oxidative stress. This question is currently being investigated in our laboratory.
Conclusions
So far, our evidence indicates that the osmotic agents are the more problematic component of PD solutions. At the local level, both glucose and polyglucose appear to substantially modify the life cycle of the mesothelial cell population. Our preliminary hypothesis is that the changes can be traced to varying degrees of oxidative stress that result directly from the osmotic agent, or from its metabolically derived substances, or from both.