CHEMICAL ENVIRONMENT OF THE CENTRAL NERVOUS SYSTEM 



1871 



fluid is similar to the formation of cerebrospinal 

 fluid. 



Relatively unimpeded exchange between the 

 cerebrospinal fluid and adjacent central nervous 

 system tissue has been demonstrated for a variety of 

 radioisotopes of inorganic ions, as well as other solutes 

 in the cerebrospinal fluid (9). Bakey (9) injected 

 colloidal and diffusible dyes (both acidic and basic) 

 of known particle size into the subarachnoid spaces, 

 and concluded that any barrier between cerebrospinal 

 fluid and central nervous system tissue must have a 

 pore size of at least 20 A in diameter. Similar con- 

 clusions were presented by Tschirgi (146) on the basis 

 of the observation that trypan blue dissolved in 

 plasma did not stain the brain of anesthetized animals 

 even after several hours of application to the cortical 

 pial surface, although the surrounding dura, fascia 

 and muscle which came into contact with the dye- 

 plasma solution were deeply colored. Contrariwise, 

 trypan blue dissolved in buffered saline rapidly 

 diffused into the underlying cortical parenchyma. 

 When the superficial pia-glia mem bra in- was damaged 

 by a slight tear, then the plasma-dye volution readily 

 stained the brain tissue in the area of the defect. 

 Using Bennhold's gelatin diffusion technique for 

 determining the relative amounts of protein-bound 

 and unbound dye present in the solution, it was found 

 that the amount of stain which diffused into the 

 uninjured brain, following topical application of 

 purified plasma albumin-dye solutions, was directly 

 related to the amount of unbound dye present and 

 to the duration of application. Furthermore, the 

 phenomenon could be quite accurately duplicated l>\ 

 a physical system consisting of white opaque gelatin 

 (to represent brain tissue) separated from the various 

 protein-dye solutions by a protein-impermeable 

 'viskin' membrane. Anatomically, this barrier to 

 protein diffusion from the cerebrospinal fluid would 

 appear to be composed of the superficial pia-glia 

 membrane (fig. 4). The epcndymal lining of the 

 ventricular walls does not seem to impose a sig- 

 nificantly greater barrier to the diffusion of inorganic 

 ions from the ventricular cerebrospinal fluid into 

 brain tissue (fig. 4). 



Therefore, on the basis of currently available 

 evidence, it is justified to consider the cerebrospinal 

 fluid as an expanded lacuna of the interstitial fluid 

 compartment, and to consider the neuronal func- 

 tional consequences of cerebrospinal fluid composition 

 as reflecting with some accuracy the state of the 

 cellular milieu. 



THE BLOOD-BRAIN BARRIER 



The question of secretion versus ultrafiltration as 

 the mechanism of formation of cerebrospinal fluid 

 has received considerable attention in the literature. 

 This topic is reviewed in Chapter LXXII in this 

 Handbook and also by Katzenelbogen (88). Insofar 

 as the interstitial fluid of the central nervous system 

 has the same composition as the cerebrospinal fluid, 

 the same question concerning the mechanism of 

 formation must pertain. Flexner (41 ) has attempted 

 to calculate the thermodynamic work required to 

 manufacture a liter of cerebrospinal fluid from a 

 theoretically infinite reservoir of blood plasma. The 

 greatest amount of positive work is needed to account 

 for the relatively higher concentration of sodium 

 and chloride ions in cerebrospinal fluid than in blood 

 plasma. After considering the theoretical Donnan 

 distribution, which would be expected in an ultra- 

 filtrate, and after subtracting the hydrodynamic 

 energy available from the blood pressure, Flexner 

 arrives at an estimated 10.9 cal. per 1. of cerebrospinal 

 fluid which must Ik- provided l>\ the mechanism re- 

 sponsible tor the formation of that fluid. Figure $ gives 

 these data only for [Na + ], [K + ] and [CI J, although 

 the estimate lor total W,,,,,, is based on a considera- 

 tion of all of the significant constituents of cerebro- 

 spinal fluid. 



If the interstitial fluid averages the same composi- 

 tion as the cerebrospinal fluid, then these same 

 thermodynamic considerations must be applied to the 

 formation of the neuronal milieu. Since there is 

 abundant evidence to indicate that the interstitial 

 fluid is not formed by an inward bulk movement of 

 cerebrospinal fluid, but, rather, that the subarachnoid 

 cerebrospinal fluid is formed, in part, by a centrif- 

 ugal flow of interstitial fluid out of the central nervous 

 system (see Chapter LXXII by Davson on cerebro- 

 spinal fluid in this Handbook), then it is necessary to 

 hypothesize some structure coextensive with most 

 of the vasculature of the central nervous system 

 which regulates solute movement and in some aspects 

 actively transports solutes between the plasma and 

 the interstitial fluid. 



The earliest report of this singular 'barrier' between 

 blood plasma and the extravascular fluids of the 

 central nervous system was the discovery by Ehrlich 

 (34) in 1885 that intravenous injection of the acidic 

 dve, coerulein-s, stained most organs in his experi- 

 mental animals, but left the central nervous system 

 relatively uncolored. Subsequently, Roux & Borell 

 (129) in 1898 demonstrated that tetanus toxin was 



