Fig. 1

Negligible osmotic contribution to CSF secretion. a Demonstration of dextran levels after an ended ventriculo-cisternal perfusion experiment in a rat. Representative sagittal cryostat sections of brains displaying the injected lateral ventricle (dotted line) and DAPI stained choroid plexus and parenchymal cells. A droplet of dextran with the concentration present within the ventricles during the experimental procedure was placed on the coverslip (red circle) and served as reference intensity. LV: lateral ventricle, dex: dextran. It is evident that the dextran had not diffused into the brain parenchyma to any detectable extent during the experimental procedure. b Representative graph of the ratio of dextran (outflow/inflow) during a 1 h control ventriculo-cisternal perfusion performed in a rat. Inset denotes the average CSF production rate in control rats, n = 28. c CSF production rate in rats subjected to varied trans-choroidal osmotic gradients by ventricular infusion of aCSF of osmolalities of ± 100 mOsm (compared to aCSF), expected to be diluted 1:1 with newly formed CSF. Each dot represents one rat, n = 11. The slope was determined from linear regression analysis (R² = 0.48) on the slope obtained based on the 11 rat experiments (slope significantly different from 0, P < 0.05). d Mean arterial pressure (AP) in rats subjected to ventriculo-cisternal perfusion upon ventricular infusion of aCSF with ± 100 mOsm (mannitol), providing an expected ventricular osmolarity of ± 50 mOsm. Data are presented as the AP (in %) at the end of the experiment compared to the initial AP measurement, n = 3, paired t-test. e CSF secretion rates obtained from ventriculo-cisternal perfusion of rats upon elevated (~ 50 mOsm) ventricular osmolarity with either mannitol or NaCl as the osmotic agent, n = 5. One-way ANOVA followed by Tukey’s multiple comparisons test, NS = not significant. f Required osmotic gradient to sustain a CSF production rate of 6.8 μl min⁻¹ given the osmotic water permeability of the choroid plexus: Osmolality of CSF should exceed that of blood with ~ 280 mOsm. g–i, Osmolality of blood (plasma or serum) and ventricular (V) or lumbar (L) cerebrospinal fluid from rats, n = 5 (g), pigs, n = 7 (h) and humans, n = 14 (i). Paired two-tailed t-test, NS = not significant