APOPTOSIS: A NEW PLAYER IN PD COMPLICATIONS

Alberto Ortiz, MD, PhD

Associate Chief of Nephrology. Unidad de Diálisis.
Fundación Jiménez Díaz, Universidad Autónoma de Madrid y Grupo de Estudios Peritoneales de Madrid del Instituto Reina Sofia de Investigaciones Nefrológicas. Madrid. España

aortiz@fjd.es

Peritoneal dialysis (PD) is a safe and effective form of renal replacement therapy. However, there are complications that limit long-term technique survival. Among them we find peritonitis and peritoneal fibrosis. Both disorders are characterized by changes in cell number.

During acute peritonitis, there are sudden changes in peritoneal fluid cell number, as well as loss of mesothelial cells. Peritoneal fibrosis is characterized by loss of mesothelium, and increased number of fibroblasts and peritoneal vessels. Tissue cell number is carefully regulated through the balance between cell birth (mitosis) and cell death, with occasional participation of cell migration or cell transdifferentiation (Fig. 1)^1,2. An imbalance between these processes can result in disorders of cell number characterized by an excessive (e.g. neoplasia or peritoneal neovascularization) or insufficient cell number (e.g. neurodegenerative diseases or peritoneal demesothelization).

Apoptotic cell death is one of the key processes that regulate cell number. There is evidence that apoptosis participates in peritoneal injury. Stimuli that induce peritoneal cell apoptosis also induce peritoneal injury in vivo, apoptosis has been observed during peritoneal disease, the expression and/or activity of apoptosis modulatory molecules changes in the course of peritoneal disease and therapeutic maneuvers that target apoptosis improved the evolution of al least one complications of PD: peritonitis³.

Thus, understanding the contribution of apoptosis to peritoneal injury offers the opportunity to decrease the negative influence of the complications of the technique.

APOPTOSIS
Apoptosis is an active mode of cell death (cell suicide) under molecular control¹. Apoptosis is characterized by morphological and functional changes (Fig 2).

Apoptosis is an essential process to remove unwanted and harmful cells and to maintain homeostasis of cell number. Teleologically apoptosis is an organized dismantling of cellular structures de×ed to limit tissue damage, while necrosis is an accidental cell death. The intensity of the lethal stimulus, the availability of ATP (energy) or the inactivation of caspases may determine whether the cell dies through apoptosis or necrosis¹.

Tissues depend on well-ordered apoptosis and cell replacement for normal functioning. However, excessive, insufficient or untimely apoptosis may result in pathology. For this reason any therapeutic measure that interferes with apoptosis should be targeted as narrowly as possible to a single cell population.

The de× of appropriate therapeutic strategies requires a correct understanding of the role of apoptosis in peritoneal injury and of the molecular regulation of peritoneal cell apoptosis.

REGULATION OF APOPTOSIS
Cell death is usually a response to the cell microenvironment. The absence of certain factors (survival factors) or the presence of lethal factors promotes apoptosis. Surrounding cells synthesize survival or lethal factors or compete for such factors. A lethal cell microenvironment activates intracellular factors that promote apoptosis and/or inactivate intracellular survival factors. Known intracellular activators of apoptosis include death receptors, proapoptotic members of the Bcl2 family, endoplasmic reticulum stress and mitochondrial injury (Fig. 3). Caspases are a family of intracellular proteases that include both activators and effectors of apoptosis.

ROLE OF APOPTOSIS IN PD
Cell death through apoptosis has been documented in the course of peritoneal injury both in animal models and in clinical PD^4-8. In addition, changes in the expression or activity of apoptosis regulatory factors have been observed in cultured peritoneal cells and in vivo peritoneal injury³.

However, few studies have addressed the role of apoptosis in PD by directly modulating its occurrence⁴. From information gathered in other tissue systems, different roles for apoptosis in peritoneal pathophysiology can be envisioned that depend on the cell type undergoing apoptosis and the amount and timing of apoptosis (Table 1). Apoptosis could be a cause of peritoneal damage, a mode of restoring normal tissue structure and a mechanism for persistence and progression of peritoneal injury.

Apoptosis may be beneficial if it occurs as a compensatory response to cellular proliferation or as a mode of eliminating irreversibly damaged cells. For example, apoptosis could contribute to resolution of fibrosis by eliminating excess fibroblasts, as demonstrated in the skin following the acute phase of scarring⁹, or could limit neovascularization¹⁰.

Under these circumstances, promotion of apoptosis in certain cell types might protect from chronic peritoneal changes associates with peritoneal dysfunction, such as fibrosis and increased vascularization. By contrast, induction of apoptosis by either exogenous factors, such as PD solutions, or endogenous mediators, could contribute to demesothelization both in the setting of acute peritonitis or chronic PD³.

Apoptosis also regulates inflammation and the immune response^{3, 11-15}. Clearance of inflammatory cells by apoptosis contributes to resolution of inflammation. However, excessive or untimely apoptosis also contributes to an impaired host response in the setting of severe infection, such as sepsis^12-14. The disposal of apoptotic cells also influences immunity and inflammation¹⁵.

From this brief glimpse at the diverse potential roles of apoptosis in peritoneal injury, we can infer that both the cell target and the timing of the intervention on apoptosis may have a re§able influence on the ultimate therapeutic consequences. It is necessary to better define the kinetics of apoptosis of different peritoneal cell populations in the course of peritoneal injury and repair.

PECULIARITIES OF PD
PD offers a unique window to access repetitively and unaggresively the protagonists of the inflammatory process in the human being. The daily disposal of peritoneal effluent allows the study of inflammatory cells and the measurement of soluble inflammatory mediators. From this point of view it is tempting to extrapolate data obtained in other systems to the PD situation However, PD has several peculiarities. Both cells and mediators are diluted several times a day by the PD exchange. In addition, the action takes paces in the presence of a uremic milieu and of poorly biocompatible PD solutions. Despite these caveats, data obtained in PD may help unravel the contribution of apoptosis to inflammation and tissue injury in other settings.

PD SOLUTIONS AND THE MODULATION OF APOPTOSIS
There are several types of PD solutions, whose influence on peritoneal cell apoptosis may differ (Table 2). Published studies have only addressed the effects of a handful of these solutions, mostly in the cell culture and animal settings (Table 2).

PD solutions are cytotoxic to leukocytes and mesothelial cells. Several groups have addressed the mode of cell death (apoptosis versus necrosis). Three independent groups have tried to simulate in vivo conditions by exposing leukocytes to undiluted PD solutions for 10 to 30 minutes and then diluting the sample with spent dialysate or cell culture medium and studying apoptosis at 4 to 24h^16-18. Traditional glucose-, high glucose degradation product (GDP)- containing PD solutions buffered with lactate accelerate neutrophil apoptosis in a caspase-dependent manner, especially when glucose concentration (and GDP content) was higher¹⁶.

As a consequence, the antibacterial function of these cells decreases¹⁶. The main proapoptotic component of PD solutions appears to be GDPs, as the toxicity could not be reproduced by adding exogenous glucose to lactate buffered solutions. In addition, glucose and low GDP solutions, such as Balance^® and Bicavera^® (Fresenius Medical Care, Bad Homburg, Germany) did not accelerate neutrophil apoptosis¹⁶ and unpublished observation).

Ongoing studies are characterizing the particular GDP responsible for this lethal effect. PD solutions also promote monocyte apoptosis¹⁷. A third group reported that high glucose, high GDP PD solutions induce monocyte necrosis. However, their techniques did not allow to fully discriminate between primary necrosis and late apoptosis¹⁸.

A higher lethality of high glucose, high GDP solutions, not present in high glucose, low GDP solutions has been observed by other authors¹⁹. In this case neutrophils were exposed to undiluted dialysate for hours. As a consequence necrosis was more prevalent. However the technique used to quantify cell death did not allow to discriminate late apoptosis from necrosis. These authors used electron microcopy to confirm the mode of cell death by morphological features. However, they did not report the results obtained with this technique in samples incubated with high glucose/high GDP PD solution¹⁹.

Glucose/GDP containing PD solutions also induce apoptosis in cultured mesothelial cells. This effect takes 48 to 72 h to develop^20,21 and has been related to glucose/GDP content²⁰ or calcium content²¹. A 1.36% glucose solution did not induce apoptosis at 16h²². The proapoptotic effect of PD solutions on mesothelial cells has been confirmed in experimental models in vivo^6-8.

PERITONITIS
Both apoptosis of mesothelial cells and leukocytes takes place during human PD peritonitis⁵ (Fig. 4). Apoptotic cells (mostly neutrophils) have also been documented in peritoneal effluents of mice with peritonitis in the presence of PD solutions⁴. The percentage of apoptotic cells is low in the early stages of peritonitis, both in the human and the experimental setting. However, given the high number of cells in these early effluents, the absolute number of apoptotic cells is high (around 175 apoptotic cells/µL in humans).

During the evolution of peritonitis, and coinciding with the decrease in total leukocyte number, the percentage of total apoptotic cells increases. In this setting, it has been suggested that mesothelial apoptosis is deleterious, as it may contribute to demesothelization. By contrast, neutrophils are programmed to die by apoptosis at the site of inflammation. This helps limit the inflammatory response. However, accelerated leukocyte apoptosis, such as that promoted by high glucose, high glucose degradation products (GDP) PD fluids or bacterial products may compromise the peritoneal defense.

Evidence to support this concept is derived from three different experimental settings.

A) Maneuvers such as LPS priming result both in decreased neutrophil apoptosis and accelerated bacterial clearance in mice²³.
B) High-glucose, GDP-containing PD solution accelerate neutrophil apoptosis in culture¹⁶ and
C) High-glucose, GDP-containing PD solution accelerate neutrophil apoptosis during experimental peritonitis⁴.

Accelerated neutrophil apoptosis in culture leads to decreased bacterial phagocytosis¹⁶. In vivo, accelerated neutrophil apoptosis is associated to a trend toward a slower bacterial clearance⁴. In both settings the pan-caspase inhibitor zVAD prevents neutrophil apoptosis and accelerates bacterial clearance, suggesting that the accelerated leukocyte apoptosis promoted by PD solutions impairs the peritoneal defense. The contribution of lymphocyte apoptosis to the peritoneal defense in PD has not been addressed yet. Contrary to neutrophils, inflammatory macrophages are able to recirculate and do not die locally²⁴. However, PD solutions accelerate monocyte apoptosis in culture¹⁷ and their possible influence on macrophage apoptosis in vivo has not been addressed either.

During peritonitis there is another potential source of proapoptotic stimuli: lethal cytokines belonging to the TNF superfamily bind to and activate cell membrane death receptors²⁵. TNF, FasL and TRAIL may be synthesized by infiltrating leukocytes and intrinsic peritoneal cells, including mesothelial cells⁵. FasL, but not TNF nor TRAIL, induces apoptosis of mesothelial cells from PD patients⁵, especially upon exposure to other inflammatory cytokines⁵.

PD peritonitis is usually a mild complication characterized by a rapid response to i.p. antibiotics. However, agents such as S. aureus or gram negative bacteria may cause severe and protracted peritonitis even when antibiotic treatment is optimized. For this reason a murine model of peritonitis induced by S. aureus was chosen to study the possible in vivo relevance of apoptosis in PD defense⁴. In the presence of high glucose, high GDP PD solutions the pan-caspase inhibitor zVAD increased the number of viable neutrophils and this was associated with an accelerated clearance of the bacteria.

These results are consistent with those reported in experimental models of septic polymicrobial peritonitis in the absence of PD solutions. In these models, prevention of apoptosis in other cell types may have contributed to the therapeutic effect of caspase inhibitors^12-14. In this regard, maneuvers that prevent lymphocyte or epithelial cell apoptosis also improve the prognosis of septic peritonitis^13,14. One unresolved issue is whether a prolonged survival or leukocytes, given the non-specific nature of the innate responses, may be detrimental to mesothelium. In this regard, mesothelium might also be protected be apoptosis inhibitors.

Once safe caspase inhibitors are available for their use in humans, trials should be de×ed to test whether their administration together with antibiotics can accelerate the recovery of the more severe cases of peritonitis.

ZVAD may also inhibit the processing of proinflammatory cytokines, such as IL-1 and IL-18, by caspases. The possible contribution of this effect to the therapeutic effect is unknown.

The fact that there were no differences in total leukocyte number between animals treated with zVAD or vehicle argues against this possibility⁴.

CHRONIC LOSS OF MESOTHELIUM
The contribution of apoptosis to the chronic loss of mesothelium upon exposure to PD solutions has been documented in experimental animals^6-8. These data support the biological ×ificance of the in vitro observations on the toxicity of PD fluids^20,21. High glucose PD solutions led to a decreased proliferation/apoptosis ratio in mesothelial cells, suggesting a disequilibrium favoring cell depletion⁶.

Chronic ip instillation of glucose solutions led to a ×ificant increase in the number of non-viable mesothelial and apoptotic cells⁸. The lethal effect was higher when lactate-based, high GDP PD solutions were used⁸. Chronic instillation of icodextrin also increases mesothelial cell apoptosis in rats⁷. However, mesothelial cell apoptosis in response to PD solutions has not been studied in human PD in vivo. In addition there is a total lack of information on apoptosis of other cell types in chronic PD and on the relationship of apoptosis to transdifferentiation².

THERAPEUTIC APPROACHES
Understanding the role and regulation of apoptosis in peritoneal disease has the potential to provide the basis for the de× of new therapeutic strategies as well as to improve the biocompatibility of PD fluids. Among the cellular targets, we might be interested in prolonging mesothelial cell survival in chronic PD and during acute peritonitis. Conversely, fibroblasts or vessels may be targeted with proapoptotic maneuvers, in order to limit fibrosis and excessive peritoneal vascularization.

The key to any successful therapeutic manipulation of apoptosis lies in limiting the interference to the cell type we want to manipulate and to a defined time period. Otherwise, we risk unintended side effects derived from interference with physiological apoptosis taking place during peritoneal healing or in the every day function of other organs. For instance, a theoretical risk of promoting neoplasia can be predicted if there are no temporal limits to antagonism of apoptosis. Mutations in apoptosis regulatory proteins that favor cell survival are common in neoplasias.

There are different theoretical approaches to achieve cell specificity. The ideal one would be to define molecular pathways that are activated only under pathological conditions.

However, there is not enough information on this subject. Another would be a cell specific delivery of the drug. Gene therapy using cell specific promoters holds promise in this regard.

Finally if cell-specificity is not achieved, we might choose a non-specific inhibitor of apoptosis administered for a limited period of time. This limited period of time would be chosen according to the time-frame of maximal apoptosis in the target cell population. The fact that the peritoneum is directly accessible during PD can also be used to our advantage. Direct i.p. administration can be used as a means of drug delivery.

Acknowledgments: The author's research has been financed by grants from FIS 98/0637 and 01/0199, Fundación Conchita Rábago, Comunidad de Madrid (08.2/0030/2000), Sociedad Española de Nefrología, Fresenius Medical Care and EU project QLG1-CT-2002-01215.

References