| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
FJ
EXPRESS SUMMARY ARTICLE The Full-length version of this article is also available, published online December 13, 2004 as doi:10.1096/fj.04-2834fje. |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Departments of
* Medicine and Physiology, and
Neurological Surgery, University of California, San Francisco, California, USA
1Correspondence: 1246 Health Sciences East Tower, Cardiovascular Research Institute, University of California, San Francisco, San Francisco, CA 94143-0521, USA. E-mail: verkman{at}itsa.ucsf.edu
SPECIFIC AIMS
Extracellular space (ECS) comprises 
20% of brain tissue volume and is important for neuron-glia communication, diffusion of ions, neurotransmitters, and metabolites, as well as drug delivery. Brain edema is expected to alter ECS size, thus profoundly influencing brain function. Using a novel fluorescence-based method, we compared the effects of vasogenic vs. cytotoxic edema on the diffusion of an inert macromolecule in brain ECS in vivo.
PRINCIPAL FINDINGS
1. Macromolecular diffusion in ECS is retarded in and around brain tumor but accelerated in brain away from tumor
Brain ECS was stained with 70 kDa FITC-dextran by incubation after removal of the bony skull but with dura intact. Laser photobleaching was done to measure the diffusion of the FITC-dextran in brain cortex in vivo.
Using a mouse model of brain melanoma, fluorescence recovery after photobleaching was measured within tumor tissue, in peritumoral brain just outside the tumor, in ipsilateral brain away from the tumor margin, and in contralateral brain (Fig. 1
A). Separate studies were done in normal brains from mice without tumors. The site of diffusion measurement had a remarkable effect on fluorescence recovery as shown by representative recovery curves in Fig. 1A
and quantified by recovery half-times (t1/2) and corresponding FITC-dextran diffusion coefficients (Fig. 1B
). Compared with normal brain (t1/2=277±13 ms), diffusion was significantly slowed inside the tumor (t1/2=1.3±0.3 s) and in the peritumoral region (t1/2=1.1±0.2 s). After photobleaching, the recovery of fluorescence for long periods was complete in normal brain (98±2%), but significantly incomplete inside (78±5%) and around (79±4%) the tumor. FITC-dextran diffusion in the hemisphere ipsilateral to the tumor (but >2 mm away from the tumor margin) was significantly faster (t1/2=210±21 ms); diffusion in the contralateral hemisphere (t1/2=294±7 ms) was not different from that in normal brain.
|
To investigate cellular features that may be responsible for the very different diffusion characteristics of tumor vs. brain ECS, GFAP-immunostained brain sections were examined. Tumor and peritumoral tissue showed a higher density of cells and cellular processes compared with normal cerebral cortex.
2. Dexamethasone reduced FITC dextran diffusion in ECS to normal
To test the hypothesis that FITC-dextran diffusion is accelerated in ipsilateral brain tissue because of ECS water accumulation by a vasogenic mechanism, mice with brain tumors were treated with dexamethasone, known to reduce brain tumor edema and inhibit vessel leak. After dexamethasone, FITC-dextran diffusion in ECS of ipsilateral cortex (t1/2=294±10 ms) was reduced to that seen in normal brain; dexamethasone had no effect on FITC-dextran diffusion in the ECS of the contralateral hemisphere (t1/2=293±18 ms).
3. Macromolecular diffusion in ECS is also accelerated after cortical cold injury
Another well-established model of vasogenic edema, cortical freeze injury, was used to verify the generality of the apparent ECS expansion near a site of vascular leak. Transient contact of the cortical surface with a cold metallic probe caused opening of the blood-brain barrier, as evidenced by Evans blue extravasation within the region of contact and transient freeze. Photobleaching measurements showed that FITC-dextran diffusion in the ECS of the hemisphere ipsilateral to the lesion (but >1 mm away from the lesion) was significantly enhanced (t1/2=186±15 ms) compared with normal brain. Diffusion in the contralateral (uninjured) hemisphere (t1/2=296±22 ms) did not differ significantly from that of normal brain. There was complete recovery of fluorescence to the prebleach level in both hemispheres.
4. Cytotoxic edema slowed FITC-dextran diffusion in brain ECS
We also investigated the effect of cytotoxic edema on FITC-dextran diffusion in the ECS (Fig. 2
), testing the hypothesis that acute cell swelling would reduce ECS volume and consequently slow FITC-dextran diffusion. In two models of cerebral hypoxia, a focal model produced by topical cyanide and a global model caused by apnea, FITC-dextran diffusion in the ECS was remarkably slowed by 6.7- and 8.2-fold, respectively. Hypoxia resulted in incomplete recovery of fluorescence after photobleaching (76±8% for cyanide, 60±5% for apnea), suggesting the presence of "dead-space microdomains" in which free FITC-dextran diffusion cannot occur.
|
5. In an in vitro model of brain ECS, cytotoxic edema restricted but vasogenic edema accelerated macromolecular diffusion
We hypothesized that in cytotoxic brain edema, cell swelling produces a reduction in ECS size and creation of dead space microdomains, resulting in restricted macromolecule diffusion in the ECS. In contrast, we postulated that vasogenic brain edema produced ECS expansion, resulting in enhanced macromolecular diffusion. To explore these ideas, we modeled the complex architecture of the ECS using suspensions of silica microbeads in fluorescently stained water/glycerol solutions. The silica particles represent cells and their processes; the aqueous phase represents the ECS. This model allowed us to vary ECS size by changing the density of silica microbeads and ECS viscosity by changing aqueous phase water/glycerol content.
This simple in vitro model reproduced semiquantitatively the changes in macromolecular diffusion in the ECS observed in vivo during cytotoxic and vasogenic brain swelling (Fig. 3
). An increase in the size of the space between particles produced by reduction in silica density (as seen in vasogenic edema) caused a reduction in t1/2. Increased silica density with corresponding reduction of aqueous phase volume (as seen in cytotoxic edema) not only slowed diffusion, but prevented complete recovery of fluorescence. A threshold phenomenon was seen as found in brain in vivo, in which diffusion becomes greatly slowed and restricted when aqueous phase volume is reduced over a relatively small range. To investigate the interdependence of ECS geometric and diffusive properties, photobleaching measurements were carried out at constant silica density but different aqueous phase viscosities. We found increased t1/2 with increasing aqueous phase viscosity, as expected. However, the ratio of t1/2 in the presence of the silica microbeads to t1/2 in the absence of the microbeads was constant, indicating that overall diffusion is slowed independently by geometric constraints and aqueous phase viscosity.
|
CONCLUSIONS AND SIGNIFICANCE
Using a novel, minimally invasive technique, we found that vasogenic brain edema enhanced the diffusion of FITC-dextran in the ECS, whereas cytotoxic edema slowed its diffusion and prevented complete recovery of signal. These findings suggest that the presence of excess fluid in brain ECS in vasogenic edema expands the ECS, whereas cell swelling in cytotoxic edema contracts the ECS. Measurement of macromolecule diffusion in the ECS by cortical photobleaching should be useful in delineating, spatially and temporally, relative roles of these opposing mechanisms in vivo.
Implantation of melanoma cells and rapid growth of a solid melanoma tumor mass produced marked changes in FITC-dextran diffusion in brain within, nearby, and at a distance away from the tumor. The increased ECS diffusion in normal brain distant from the tumor probably reflects ECS expansion caused by vasogenic edema. We found remarkably slowed FITC-dextran diffusion in the ECS within the tumor mass and peritumoral astrogliotic rim probably resulting from a combination of hypercellularity, reduced ECS volume, and overproduction of extracellular matrix components.
Cortical photobleaching revealed reduced FITC-dextran diffusion in the ECS in cytotoxic brain edema produced by chemical hypoxia and apnea. The increased t1/2 was associated with incomplete recovery of fluorescence, probably due to micro-compartmentalization of the ECS. Cortical photobleaching is likely to be more sensitive in detecting cell swelling than intracranial pressure, which rises only when the functional compliance of intracranial compartments has been exceeded. In some conditions, such as early hypoxia or seizures, there is a redistribution of water from the ECS into the intracellular compartment without a net increase in brain water content. In such cases, ICP and total brain water content may remain normal, but macromolecule diffusion in the ECS would be slowed.
Diffusion of macromolecules in brain ECS (Decs) is slower than in solution (Do). The degree of slowing has been expressed as "tortuosity," defined as (Do/Decs)1/2. To better understand the molecular determinants of tortuosity, we modeled the complex geometry of brain ECS with silica particles suspended in a fluorescent water-glycerol matrix. Tortuosity depended on the size and viscosity of the ECS. Below a threshold aqueous phase volume fraction, ECS abruptly lost its continuity and became micro-compartmentalized, associated with failure of the extracellular fluorescence to recover after photobleaching. In the brain, micro-compartmentalization is a catastrophic event in which macromolecule diffusion in the ECS becomes highly restricted. Consequently, clearance of metabolites and toxic components such as potassium and glutamate from the ECS is greatly reduced, potentially leading to cell death. Although data in vivo and in the silica model were qualitatively similar, micro-compartmentalization occurred at a higher ECS volume in the silica model than in brain. We suggest that the brain has evolved mechanisms to resist micro-compartmentalization. For example, ECS matrix components may maintain cell-cell spacing during cell swelling, a feature not reproduced in the silica microbead model where direct bead-bead contact was not prevented. Another conclusion from the silica microbead experiments was the independence of the two principal determinants of ECS diffusion: geometric restriction and intrinsic matrix diffusivity.
Macromolecular diffusion in brain ECS is important in drug delivery to swollen brain associated with brain tumor, head injury, and stroke. In general, drug delivery by ECS diffusion is expected to be slow in regions primarily affected by cytotoxic edema but fast in regions where vasogenic edema predominates. Investigations of the diffusion of DNA, protein and liposome macromolecules in models of brain swelling will be needed to validate these ideas.
FOOTNOTES
To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.04-2834fje;
This article has been cited by other articles:
|
|
 |
|
  Z. Zador, M. Magzoub, S. Jin, G. T. Manley, M. C. Papadopoulos, and A. S. Verkman Microfiberoptic fluorescence photobleaching reveals size-dependent macromolecule diffusion in extracellular space deep in brain FASEB J, March 1, 2008; 22(3): 870 - 879. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. C. Papadopoulos, J. K. Kim, and A. S. Verkman Extracellular Space Diffusion in Central Nervous System: Anisotropic Diffusion Measured by Elliptical Surface Photobleaching Biophys. J., November 1, 2005; 89(5): 3660 - 3668. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. C. Papadopoulos and A. S. Verkman Aquaporin-4 Gene Disruption in Mice Reduces Brain Swelling and Mortality in Pneumococcal Meningitis J. Biol. Chem., April 8, 2005; 280(14): 13906 - 13912. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
