Ken D. McCarthy
Investigating the Role of Astrocyte Signaling in Brain Function
Approximately 50% of the mammalian brain is composed of astrocytes. These cells are present in every region of brain, are always closely associated with neuronal elements, and exhibit a wide variety of morphological phenotypes and neurotransmitter receptors. It is striking that while these cells constitute the largest single population of cells in brain, we know very little concerning their role in brain function. Many neurobiologists believe that astrocytes play a critical role in buffering extracellular potassium levels within the narrow range required for neuronal activity. Similarly, astrocytes are thought important in removing glutamate following its release at neuronal synapses. Modulation of either potassium buffering or glutamate uptake into astrocytes would markedly affect neuronal excitability. Our lack of understanding the role of astrocytes in brain function stems, in large part, from the difficulty in studying these cells in situ . Astrocytes do not exhibit the electrical properties or span the distances that have enabled neurobiologists to study neurons in situ so effectively. Furthermore, astrocytes frequently exhibit a very fine, branching network of velate processes that envelope neurons. To date, it has been extremely difficult to study dynamic events occurring within the fine velate processes of astrocytes. The velate processes of astrocytes are frequently connected to other astrocytic processes via gap junctions to form an astrocytic syncytium through which intercellular signaling may propagate. It seems likely that discrete microdomains within the astrocytic syncytium may interact autonomously with neurons. A primary goal of our laboratory is to determine how astrocytes and neurons are signaling one another and the functional outcome of neuron astrocyte conversation.
Techniques became available in the 1970s for preparing primary cultures of purified astroglia (astrocytes in vitro ) from neonatal rodent brain. A large number of laboratories have used cultures of astroglia to study the signaling systems expressed by these cells in vitro . It is clear from these studies that astroglia express a very wide variety of neuroligand receptors (both ligand-gated ion channels and G-protein linked receptors) and that these receptors are coupled to most of the known intracellular signaling cascades. Further, astroglia are heterogeneous with respect to their expression of different neuroligand receptors and individual astroglia often express multiple types of neuroligand receptors. In vitro, these receptor systems appear to regulate a wide variety of cellular processes including 1) the uptake and release of neurotransmitters, 2) the synthesis and release of neurotrophic factors, 3) proliferation, 4) apoptosis, 5) intracellular volume, 6) the conductance of potassium channels, and 7) the opening and closing of the gap junctions that normally connect astroglia (and astrocytes in situ ) to form a syncytium. If astrocytes exhibit similar responses to neurotransmitters in vivo , it is likely that these glial cells play an important role in regulating neuronal excitability.
In the mid 1990s we became convinced that in order to understand the role of astrocyte signaling in brain function, it was critical to study astrocytes in situ where their complex morphology and intimate association with neurons remains intact. To accomplish this we developed two complementary lines of investigation, both aimed at elucidating the role of astrocytes in neurophysiology and neuropathology. One approach is based on the analysis of astrocyte signaling within the CA1 stratum radiatum region of the hippocampus using confocal calcium imaging and electrophysiological methods. The second approach is based on molecular strategies designed to disrupt astrocytic signaling systems in transgenic and conditional knockout (cKO) mice. The long term goal of these studies is to develop an understanding of the manner and conditions under which astrocytes and neurons communicate in situ and determine the outcome of disrupting their communication on neuronal excitability and animal behavior.
Astrocytes exhibit a number of properties which suggest that they modulate neuronal activity in vivo. These properties include their ability to 1) buffer extracellular potassium, 2) rapidly take up neurotransmitters following release at neuronal synapses, 3) release neurotransmitters and neuromodulators (including glutamate, ATP and D-serine), 4) release neurotrophic factors, and 5) regulate the volume of the extracellular space. The complex morphology of astrocytes and their ability to form a syncytium connected by gap junctions suggest that their interactions with neurons may occur in discrete microdomains that function autonomously. Our working hypothesis is that there are microdomains within astrocytic syncytium that interact with neuronal synapses to facilitate or to dampen neuronal excitability and/or neurotransmission. These microdomains may be derived from a single astrocyte or from multiple astrocytes and may function independently from one another or in unison depending on the level of neuronal activity. Further, the ability of signals to travel within the astrocytic syncytium is likely to be modulated by second messengers which regulate the opening and closing of gap junctions. Ultimately, we believe that the complexity of signaling within the astrocytic syncytium will be as complex as that occurring between neurons and will function to regulate neuronal excitability.
Our current efforts in this area are directed at examining signaling between astrocytes and neurons in the region of the Schaffer collateral, CA1 pyramidal neuron synapse (SC-CA1 synapse) within acutely isolated mouse hippocampal brain slices. This synaptic field has been a favorite among electrophysiologists for many years because of its laminar organization, well-defined synaptic connections and data suggesting that it is important in the process of learning and memory. Our experiments are carried out using electrophysiological methods to stimulate and record neuronal activity while monitoring astrocytic responses using calcium imaging dyes and our confocal microscope based detection system. These experiments were initially carried out using astrocytes that were loaded by incubation of hippocampal slices with a permeable calcium indicator dye (a technique known as "bulk loading") . Once the astrocytes were loaded with the calcium indicator dye, their responses to bath applied neuroligands or electrical stimulation of the Schaffer collateral pathway were recorded. The results of these studies were the first to demonstrate that astrocytes within dendritic fields in situ respond to synaptically-released neurotransmitter. Additional studies demonstrated that astrocytes in the region of SC-CA1 synapses express a number of different receptor systems linked to calcium mobilization. Overall, these early studies demonstrate that astrocytes in situ express a number of different neuroligand receptors whose activation increases intracellular calcium. Most importantly, these experiments indicate that when neurons are talking to one another, astrocytes are listening.
Recently, we have modified our experimental approach in two important ways. First, we have switched from rat to mouse as a tissue source to take advantage of mouse genetics and the mouse lines we are generating (see below). Second, we have developed methods that enable us to load individual astrocytes with calcium indicator dyes using patch clamp electrodes. The advantage of patch clamp dye loading is that the signal to noise is greatly improved enabling us to study calcium responses within the fine processes of astrocytes. Recent experiments using mouse hippocampal brain slices indicate that astrocytes exhibit spontaneous oscillations in calcium that occur independent of neuronal action potentials. Further, there are multiple microdomains of calcium oscillations within the fine processes of an individual astrocyte that oscillate independently from one another. The spontaneous calcium oscillations observed in astrocytes are dependent on phospholipase C activity and the release of calcium from internal stores. Current efforts in this area are aimed at elucidating the mechanisms mediating astrocytic calcium oscillations and determining if oscillations in astrocytic calcium affect neuronal excitability.
A number of laboratories have reported that increases in astroglial calcium leads to the release of glutamate in vitro that is sufficient to stimulate neuronal glutamatergic receptors and depolarize these neurons. Further, data suggests that astrocytes in situ may also release glutamate in a calcium dependent manner. To directly address this issue, we developed a laser-based uncaging system that enables us to release caged signaling molecules (e.g. IP3) within hippocampal brain slices. We used this system to load individual astrocytes with caged second messengers via patch clamp electrodes such that the uncaged the second messenger is restricted to the "patched" astrocyte. Our uncaging system has the capability of uncaging (i.e., activating) the second messenger in msec time intervals within a region 2-3 microns in diameter. Using this approach, we recently demonstrated that increases in astrocytic calcium in situ lead to increases in the frequency and amplitude of mEPSPs recorded in CA1 pyramidal neurons (J. Neuroscience 24:722, 2004).
These finding led to the model provided below of how astrocytes might modulate synaptic transmission at the Schaffer collateral CA1 synapse; studies in progress are designed to test different predictions of this model.
Molecular strategies are being used in our laboratory to study the functional consequences of signaling between astrocytes and neurons in vivo. We see these studies critical in our effort to move beyond characterizing astrocyte signaling to developing an understanding of the functional role of astrocyte signaling in neurophysiology and animal behavior. Two different molecular approaches are being used to disrupt astrocyte signaling processes in vivo. First, we are making transgenic mouse lines that express RASSLs (Receptors Activated Soley by Synthetic Ligands). An astrocyte specific promoter linked to an inducible gene expression system is being used to direct the expression of RASSLs to astrocytes in vivo. Using an inducible gene expression system enables analysis of mice expressing "physiological" and non "physiological" levels of G-protein coupled receptors (GPCRs) selectively expressed in astrocytes. We are currently studing the role of astrocytic Gi-linked GPCRs in neurophysiology and behavior using RASSL mice. Interestingly, overexpression of this RASSL leads to hydrocephalus and premature death. This is the first model of hydrocephalus linked to a GPCR signaling system. Experiments are being carried out to define the site of the lesion and the signaling pathways involved in the development of hydrocephalus in these mice. We are also using mice expressing "physiological" levels of this RASSL to determine the role of astrocytic Gi-linked GPCRs in neurophysiology. We are currently making transgenic lines expressing Gq- and Gs-linked GPCRs in astrocytes.These transgenic mouse models will enable us to carry out electrophysiological, confocal imaging and behavioral studies to determine the consequence of activating astrocytic GPCRs in situ and in vivo.
Conditional gene knockout (cKO) mice are being made as a second approach to perturbing astrocytic signaling systems in vivo . Conditional gene knockouts enable cell specific, timed gene knockouts and have several advantages over the use of inhibitory transgenes or traditional gene knockouts. The primary advantages of cKOs over inhibitory molecules is that inhibitory molecules quantitatively interfere with signal transduction molecules whereas gene knockouts qualitatively eliminate signal transduction molecules. Advantages of cKOs over traditional gene knockouts include the ability to limit the gene deletion to a subset of cells and to time that knockout to specific stages of maturation. Conditional knockouts are based on the ability of a phage recombinase to efficiently excise DNA segments flanked by recognition sites known as loxP sites; genes containing loxP sites are referred to as floxed genes. Two separate lines of mice must be generated to prepare cKO mice where the deletion is targeted to a specific cell population. First, a transgenic mouse line expressing Cre recombinase driven by a cell specific promoter. Second, a mouse line prepared through homologous recombination in which a fragment of the targeted endogenous gene is replaced with the same gene fragment but surrounded with loxP sites. The goal is to have the floxed gene function normally by placing the loxP sites in regions unimportant in gene expression but surrounding a fragment of the gene critical for its function. When a transgenic line expressing Cre recombinase in a specific cell population is crossed with a mouse line containing a floxed gene, the gene fragment surrounded by loxP sites is excised and the gene inactivated in the cell population expressing Cre recombinase. The use of cKOs in neurobiology is particularly attractive given the enormous cellular and functional micro heterogeneity within the CNS.
A cKO of connexin43 (Cx43) in astrocytes leads to aberrant neuronal placement. Cx43 is the predominant gap junction protein expressed in astrocytes and thought critical for the formation of the astrocyte syncytium which enables astrocytes to directly communicate with one another. We prepared a cKO of Cx43 in astrocytes to test the hypothesis that gap junction communication between astrocytes is important in the regulation of neuronal excitability. Interestingly, the deletion of Cx43 in astrocytes leads to a striking phenotype that includes disruption of neuronal placement during development. Our studies suggest that Cx43 is important in the correct orientation of radial glial processes and that this, in turn, leads to aberrant neuronal placement. Further, our results suggest that Cx43 may exert it influence on neuronal migration through mechanisms other than gap junctional communication, e.g. through tethered signaling pathways. Experiments in progress are designed to elucidate the mechanisms underlying the developmental phenotype displayed by Cx43 cKO mice as well as use inducible cKOs to study the role of astrocytic Cx43 postdevelopmentally.
A cKO of Kir4.1, an inward rectifying potassium channel, leads to seizure activity and loss of myelination. Kir4.1 is the predominant inwardly rectifying potassium channel expressed by astrocytes and thought important in the maintenance of potassium homeostasis during neuronal activity. A cKO of astrocytic Kir4.1 was prepared to test the hypothesis that this channel is necessary to maintain [K+]o within the normal limits required for neuronal activity. Kir4.1 cKO mice exhibit seizure activity and abnormal myelination within white matter tracts. Further, astrocytes are markedly depolarized from their resting membrane potential of ~-85 mV to ~ -30 mV. Interestingly, the current patterns observed in Kir4.1 cKO astrocytes are similar to those observed in wild type astrocytes. Experiments in progress are designed to identify the channels responsible for the passive currents expressed in Kir4.1 cKO astrocytes and determine if potassium homeostasis is perturbed in these mice.
Currently, we are preparing cKOs that will target astrocyte signaling systems (e.g., GPCRs). Our current view is that conditional gene knockouts will be required to sort out the role of astrocytic signaling systems in brain function and a great deal of our effort is directed along these lines of investigation.
Recent advances in glial biology indicate that astrocytes may play a much more dynamic role in modulating neuronal activity than previously thought. Through a combination of 2-photon confocal imaging, electrophysiological and molecular studies we hope to begin unraveling the role of these cells in brain function and animal behavior.
- Agulhon, C., Petravicz, J., McMullen, A.B., Sweger, E.J., Minton, S.K., Taves, S.R., Casper, K.B., Fiacco, T.A., and McCarthy, K.D. (2008) What is the role of astrocyte calcium in neurophysiology? Neuron 59:932-46.
- Petravicz, J., Fiacco, T.A., and McCarthy, K.D. Loss of IP3 receptor-dependent Ca2+ increases in hippocampal astrocytes does not affect baseline CA1 pyramidal neuron synaptic activity. J. Neurosci. 28:4967-73.
- Djukic, B., Casper, K.B., Philpot, B.D., Chin, L.S., and McCarthy, K.D. (2007) Conditonal knockout of Kir4.1 leads to glial membrane depolarization, inhibition of postassium and glutamate uptake, and enhanced short-term synaptic potentation. J. Neurosci. 27:11354-65.
- Fiacco, T.A., Agulhon, C., Taves, S.R., Petravicz, J., Casper, K.B., Dong, X., Chen, J., McCarthy, K.D. Selective stimulation of astrocyte calcium in situ does not affect neuronal excitatory synaptic activity. Neuron 54: 611-26.
- Sweger, E.J., Casper, K.B., Scearce-Levie, K., Conklin, B.R., McCarthy, K.D. (2007) Development of hydrocephalus in mice expressing the G(i) coupled GPCR Ro1 RASSL receptor in astrocytes. J. Neurosci. 27:2309-17.
- Wienchken-Barger, A.E., Djukic, B., Casper, K.B., McCarthy, K.D. (2007) A role for Connexin43 during development. Glia, 55:675-86.
- Fiacco, T.A., and McCarthy, K.D. Intracellular astrocyte calcium waves in situ increase the frequency of spontaneous AMPA receptor currents in CA1 pyramidal neurons. J. Neurosci. 24: 722-732, 2004.
- Nett, W.J., Oloff, S.H., and McCarthy, K.D. Hippocampal astrocytes in situ exhibit calcium oscillations that occur independently of neuronal activity and are restricted to subcellular components. J. Neurophysiology, 87:528-537,2002.
- Howe DG and McCarthy KD. Retrovial inhibition of cAMP-dependent protein kinase inhibits myelination but not Schwann cell mitosis stimulated by interaction with neurons. J. Neurosci. 20: 3513-3521, 2000.