| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Albert Einstein College of Medicine, Bronx, NY
2 University of Illinois College of Medicine, Chicago, IL
* Corresponding author: harrripp{at}uic.edu
The vertebrate retina is a highly specialized sheet of neural tissue derived from an undifferentiated population of neural progenitor cells, which, upon completion of their final division, migrate to their laminar positions and differentiate (1). In the adult retina, almost every class of neuron and glial cell is linked to its neighbors by gap junctions—aqueous channels that allow the intercellular exchange of ions, second messengers, and other small molecules (
1 kDa). Thus, this communication pathway aids in the synchronization of cellular activity, and plays a significant role in maintaining cellular homeostasis (2). During development, however, many retinal cells undergo programmed cell death, or apoptosis, and both intracellular and intercellular communication are known to regulate the process (3). Indeed, we recently reported that gap junctions mediate a form of "bystander" cell death in the developing retina (4). Although this process is largely arrested in the mature retina, differentiated neurons and glia undergo apoptosis in neurodegenerative diseases (5), as well as in ischemia and trauma. In all of these cases, the spread of cell death from one dying cell to its otherwise unaffected neighbors (bystanders) may increase the total number of cells that enter the apoptotic pathway.
Because the retina is a complex tissue, some studies of bystander cell death are technically unfeasible at this time. We have developed a model system for the study of gap junction-mediated cell death using Xenopus oocytes which express an endogenous gap-junctional protein (connexin38, [Cx38]). Oocytes are paired at their vegetal poles following removal of their vitelline membranes and become electrically coupled via gap junctions. Because the Xenopus oocyte can serve to express many different connexins, the system should enable us to identify which gap junctional channels remain open under apoptotic conditions, to study the dynamics of gap junctional coupling during cell death, and to determine which molecules may pass from a dying cell to its neighbors to trigger the apoptotic process. Conversely, it is important to recognize that gap-junction-mediated cell death may result from the depletion of essential molecules (e.g., ATP) passing from the healthy cell to its dying neighbor.
Previous studies of single Xenopus oocytes have shown that microinjection of cytochrome c induces apoptotic cell death, accompanied by a progressive loss of membrane potential, activation of caspase 3, and DNA fragmentation (6). In the present study we have shown that injection of cytochrome c into one oocyte of a Cx38-coupled pair causes death in both the injected and noninjected cells over a period of 2–3 h (Fig. 1a,b). Pairs of oocytes preinjected with an antisense oligonucleotide to Cx38 did not exhibit bystander cell killing following cytochrome c injection; although the cell into which cytochrome c had been injected did die, its paired neighbor did not. Similarly, in cases where the cells were electrically coupled, but where the cytochrome c injected cell lysed in less than 40 min, the noninjected cell did not die (Fig. 1a,b, left pair). This result suggests that the intercellular channels joining the cytoplasm of the coupled cells must remain intact for longer than 40 min for bystander cell death to occur. It also indicates that bystander killing is not mediated by contact or by extracellular toxins, since the surviving cell continued to be in close apposition to the dying cell but remained intact. Vehicle injections failed to induce cell death in either injected or noninjected cells of electrically coupled pairs.
|
Although we have induced cell death by intracellular injection of cytochrome c, it is evident that the molecular mass of this apoptotic agent (13 kDa) is too great to pass through gap junctions (8). Clearly, cytochrome c itself cannot be the toxic substance carrying the death signal to bystander cells. Likewise, bystander killing is unlikely to be mediated by contact or diffusible substances, since antisense to Cx38 prevented bystander cell death, but not primary cell death. Of particular interest is the fact that gap junctional coupling persists during the apoptotic process, encouraging us to use the Xenopus oocyte and other expression systems in future experiments to identify the intercellular signals that pass between a dying cell and its coupled partners to induce bystander cell death.
These studies were conducted at the Marine Biological Laboratory, Woods Hole, Massachusetts, and were supported by grants from the NIH (HR: EY-06516 and EY-01792; KC: HL-07675); a Grass Foundation Fellowship (KC); an unrestricted award to the UIC Department of Ophthalmology and Visual Sciences from Research to Prevent Blindness, Inc.; a Senior Research Investigator Award from the RPB (HR); and an Award of Merit from the Alcon Research Institute (HR).
Literature Cited
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
