Evans Ca

The development of a blood-brain barrier mechanism in foetal sheep

1. The penetration of a metabolically inert, small molecular radius lipid insoluble substance ([13C] and [4H]sucrose), from blood into brain and c.s.f., has been studied in developing sheep from 50 days gestation (term, 150 days) through to the new‐born stage. Around 50 days gestation sucrose accumulated rapidly into brain and c.s.f., and reached a steady‐state level in brain of about 12% of the plasma level by 3 hr. By 60 days sucrose penetrated less freely into brain and c.s.f.; the brain steady‐state level was 10% by 4½ hr. A large decrease in sucrose penetration occurred by 70 days gestation, and by 123 days (just before the time when a foetal lamb becomes viable) both the rate of penetration and the brain steady‐state level of sucrose were similar to those of the adult of other species.

Studies of the development of brain barrier systems to lipid insoluble molecules in fetal sheep.

1. The development of the blood‐brain and blood‐c.s.f barriers to lipid insoluble substances of different molecular radii has been studied in fetal sheep, early (60 days) and late (125 days) in gestation, using labelled erythritol (C14), sucrose (3H or 14C), inulin (3H or 14C) and albumin (125I), or albumin and IgG detected by immunoassay. 2. Morphological studies of fetal brain and choroid plexus at the same gestational stages were carried out using thin section electron microscopy and the freeze fracture techniques. 3. Penetration of markers into c.s.f. was substantially greater at 60 days than at 125 days, but at both ages the steady‐state level achieved appeared to be related to molecular size. 4. A simple model describing penetration from blood into c.s.f. at 60 days is proposed. It involves the assumption that c.s.f. and brain extracellular fluid are effectively separate compartments; morphological and permeability data which supports this assumption is presented. The data for c.s.f. at 60 days are consistent with the suggestion that the markers penetrate into c.s.f. by diffusion and are not restricted by small pores in the interface between blood and c.s.f. 5. The reduction in penetration which occurred by 125 days for all markers except erythritol appears to be accounted for by an increase in the sink effect and a decrease in the effective surface area for exchange between blood and c.s.f. 6. Intercellular tight junctions of both cerebral endothelial cells and choroid plexus epithelial cells were well formed at 60 days gestation. There was no change in junctional characteristics previously thought to correlate with transepithelial permeability (tight junction depth and strand number) between the two ages studied, although there were marked changes in permeability. 7. Evidence is advanced in support of the hypothesis that in the fetus much of the penetration of lipid insoluble non‐polar substances across the blood‐c.s.f. barrier and perhaps across the blood‐brain barrier occurs via a transcellular route consisting of a system of tubulo‐cisternal endoplasmic reticulum. Penetration via the choroid plexus appears to be the dominant route for penetration from blood into c.s.f. in the 60 day fetus.

Proteins in cerebrospinal fluid and plasma of fetal rats during development.

The concentrations of total protein, albumin, and α -fetoprotein have been measured in the cerebrospinal fluid (csf) and plasma of fetal (12 to 22 days gestation) and neonatal (0 to 10 days postnatal) rats. Total protein concentration in cisternal csf increased from about 140 mg/100 ml at 12 days to reach a peak of over 300 mg/100 ml around the time of birth. In the postnatal period the total protein concentration declined to about 100 mg/100 ml at 10 days; the adult value was 24±8 mg/100 ml. There was substantially more α -fetoprotein than albumin in csf at 12 days gestation. Both increased in concentration toward the end of gestation; α -fetoprotein reached a plateau of about 100 mg/100 ml at 17–19 days after which it declined markedly to about 5 mg/100 ml at 10 days postnatal; albumin reached a plateau in csf of about 90 mg/100 ml around the time of birth and declined subsequently. Albumin and α -fetoprotein constituted over 50% of the total protein concentration in csf at all fetal ages studied. In plasma these two proteins made up only 35% of the total protein at 17 days gestation but by the time of birth their contribution had doubled. Total protein concentration in plasma increased throughout the developmental period studied as did that of albumin. α -Fetoprotein was at its highest concentration (440 mg/100 ml) at 19 days gestation; it declined markedly in the postnatal period. Other proteins identified in csf and plasma were: transferrin, α 1 -antitrypsin, and IgG. Evidence is discussed which suggests that the high concentration of protein in fetal csf results from specific transfer of plasma protein across the developing choroid plexus rather than from immaturity of the blood-csf barrier.

Intracranial haemorrhage in the preterm sheep fetus

The germinal layer in the brain of the sheep fetus at 58--85 days of gestation was found to resemble that of the human infant at 28--30 wk of gestation. Experiments were done on 65 exteriorized fetuses to explore the effect of various combinations of asphyxia and raised intravascular pressures in causing bleeding into the germinal layer, ventricles and other parts of the brain. Asphyxia by itself did not produce an increase in the incidence of intracranial hemorrhages when compared with control fetuses. The combination of asphyxia with intermittent increases in arterial or venous pressure, or both, did cause haemorrhages. Large increases in arterial pressure without asphyxia also caused intracranial haemorrhages, whereas increases in venous pressure without asphyxia did not. The types of haemorrhage observed closely resembled those seen in the preterm human infant, although massive intraventricular haemorrhages (IVHs) were rare. We conclude that: (1) the sheep fetus can be used for investigating factors associated with intracranial haemorrhage in the preterm brain; (2) the most effective method of producing haemorrhages into the germinal layer was by a combination of asphyxia with intermittent increases in cerebral intravascular pressure. Similar mechanisms may be at work in the newborn human infant, and could lead to IVH.

Synthesis of plasma proteins by rat fetal brain and choroid plexus.

Several plasma proteins have previously been demonstrated to be within cells (presumed neurons) in the developing brain of various species. The possibility that the plasma proteins α‐fetoprotein (AFP) albumin and transferrin may be synthesized by developing brain and choroid plexus has been investigated in fetal rats of 18 to 22 days gestation. Samples of these tissues and of liver were incubated in Krebs solution containing [3H]leucine at 37°C for 1 h. Radioactively labelled AFP, albumin and transferrin were extracted and separated by immunoprecipitation. Incorporation of [3H]leucine into the plasma proteins was demonstrated in both fetal brain and choroid plexus. Incorporation was completely blocked by cycloheximide. It is concluded that fetal brain and choroid plexus synthesize AFP, albumin and transferrin and that secretion of these proteins by developing brain and choroid plexus cells probably contributes to the high concentration of plasma proteins in fetal csf.

Plasma proteins in fetal sheep brain: blood-brain barrier and intracerebral distribution.

1. Plasma proteins have been demonstrated to be present in early fetal sheep brain in amounts which cannot be accounted for by blood contamination. 2. The distribution of alpha‐fetoprotein, albumin, fetuin, alpha 1‐antitrypsin and transferrin has been studied by immunoassay of extracts from brain homogenates and by immunoperoxidase histochemistry of fetal brains between 37 and 125 days gestation (term is 150 days). 3. At 35 days gestation fetuin and albumin were quantitatively the most important proteins in fetal brain, both as estimated by extraction and by immunohistochemistry. Both of these proteins, and also alpha‐fetoprotein and alpha 1‐antitrypsin, declined considerably in amount by 60 days gestation. After 60 days the concentrations of albumin and alpha‐fetoprotein were not significantly different from that due to blood contamination and only occasional cells could be demonstrated by immunohistochemistry. Fetuin and alpha 1‐antitrypsin were present in reduced but significant amounts at least up to 125 days gestation. 4. The immunohistochemical results showed that considerable numbers of immature neurones were stained for some plasma proteins. Fetuin positive cells predominated both in terms of the larger number of cells which stained at 35‐40 days gestation and in the persistence of positive cells up to 125 days gestation. Numerous cells in the neuroependymal layer and in several layers of the developing cortical plate were positive, especially early in gestation. Only a few transferrin or alpha 1‐antitrypsin positive cells were observed. 5. Permeability of the blood‐brain barrier to sheep or human serum albumin (labelled with 125I) was tested in 60 day fetal sheep by intravenous injection and estimation of brain radioactivity at 3 or 6 hr with allowance for blood contamination. Only a very small (but significant) penetration of protein was detected. Unlike penetration of protein into c.s.f. at the same age, it did not reach its natural steady state in brain. 6. It is concluded that the blood‐brain barrier to protein is well developed in the immature fetal sheep but that developing neurones probably acquire certain plasma proteins directly from the c.s.f. when they are differentiating in the neuroependyma. The subsequent distribution of plasma protein positive cells in different brain regions is suggested to be due to the migration of developing neurones for the neuroependyma although the possibility of local synthesis of plasma proteins has not been excluded.

CSF-Brain permeability in the immature sheep fetus: A CSF-brain barrier

The permeability of the neuroependyma between CSF and brain extracellular space has been studied in fetal sheep of 60 and 125 days gestation. Both radioactive ([3H]inulin, [14C]sucrose, [125I]albumin) and visible (horseradish peroxidase) markers have been perfused through the ventricular system for periods of up to 5 h in anaesthetized exteriorized fetal sheep whose physiological condition was monitored continuously. A previously undescribed barrier between CSF and brain extracellular fluid has been discovered in the immature (60-day) fetal sheep. Horseradish peroxidase penetration was confined to a limited depth of the neuroependyma and was mainly into the cells lining the cerebral ventricles; in older fetuses there was extracellular penetration to a distance of several millimetres from the ventricular surface, as previously described in adult animals. The volumes of distribution of sucrose and insulin were less in the immature brain than in the more mature brain, which may be a reflection of restricted diffusion across the neuroependyma in the younger brains. The morphological nature of the barrier in fetuses of 60 days and younger appears to be a membrane specialization between the cells of the neuroependyma. It is of a type not previously described; it seems to have the effect of narrowing rather than obliterating the extracellular pathway between CSF and brain. The possible functional significance of this observation is discussed.

Rapid transport of acetylcholine in rat sciatic nerve proximal and distal to a lesion

1. Changes in distribution of acetylcholine (ACh) along the length of the sciatic nerve of the rat have been studied up to 24 hours after axotomy produced usually by crushing but sometimes by cutting or ligating the nerves. The ACh was extracted with trichloracetic acid from 5 mm lengths of nerve cut relative to the lesion and estimated by bio-assay techniques. 2. Axotomy usually produced an increase in the ACh content in the 5 mm on either side of the lesion within 1–2 minutes of operation. Subsequently there was a continued marked increase in ACh content in the 5 mmproximal to the lesion up to nearly three times the control level by 12 hours, with no further increase by 24 hours. In the 5 mmdistal to the lesion there was a further slight increase up to 60% above control by 6 hours with a subsequent decline to about the control level by 24 hours. In the more distal parts of the nerve (i.e. 5–20 mm distal to the lesion) there was a decline in ACh content of about 20% by 6–12 hours after operation. The three types of axotomy produced similar changes. 3. The initial small increase of ACh on both sides of the lesion was probably due to local synthesis of ACh, as previously described by other authors. It is suggested that the marked proximal increase in ACh was due to interruption of a proximo-distal transport of ACh and that the small decline in the more distal parts of the nerve was due to continued transport of ACh out of that segment of the nerve following operation; the size of this decline has been taken as an estimate of the proportion of axonal ACh which is rapidly transportable (20% of the total). The rate of transport of this fraction of the axonal ACh has been estimated as about 5 mm/hour. The rest of the axonal ACh is thought to be either stationary or moving slowly with bulk proximo-distal movement of the axoplasm. Changes in distribution of acetylcholine (ACh) along the length of the sciatic nerve of the rat have been studied up to 24 hours after axotomy produced usually by crushing but sometimes by cutting or ligating the nerves. The ACh was extracted with trichloracetic acid from 5 mm lengths of nerve cut relative to the lesion and estimated by bio-assay techniques. Axotomy usually produced an increase in the ACh content in the 5 mm on either side of the lesion within 1–2 minutes of operation. Subsequently there was a continued marked increase in ACh content in the 5 mmproximal to the lesion up to nearly three times the control level by 12 hours, with no further increase by 24 hours. In the 5 mmdistal to the lesion there was a further slight increase up to 60% above control by 6 hours with a subsequent decline to about the control level by 24 hours. In the more distal parts of the nerve (i.e. 5–20 mm distal to the lesion) there was a decline in ACh content of about 20% by 6–12 hours after operation. The three types of axotomy produced similar changes. The initial small increase of ACh on both sides of the lesion was probably due to local synthesis of ACh, as previously described by other authors. It is suggested that the marked proximal increase in ACh was due to interruption of a proximo-distal transport of ACh and that the small decline in the more distal parts of the nerve was due to continued transport of ACh out of that segment of the nerve following operation; the size of this decline has been taken as an estimate of the proportion of axonal ACh which is rapidly transportable (20% of the total). The rate of transport of this fraction of the axonal ACh has been estimated as about 5 mm/hour. The rest of the axonal ACh is thought to be either stationary or moving slowly with bulk proximo-distal movement of the axoplasm.

The distribution of acetylcholine in normal and in regenerating nerves

1. The distribution of acetylcholine (ACh) in various nerves which had been regenerating for different periods after crushing has been compared with that in uncrushed nerves.

Comparison of proteins in CSF of lateral and IVth ventricles during early development of fetal sheep

This study examines the relationship between plasma proteins in blood and in CSF in the developing brain of sheep fetuses between 30 and 60 days gestation. Five proteins account for the very high concentration of protein in fetal CSF (over 1000 mg/100 mg/100 ml at 30 days): alpha-feto-protein, fetuin, albumin, alpha 1-antitrypsin and transferrin; the concentration of each protein is similar in lateral and IVth ventricular CSF at 30 days. By 40 days there is considerable decrease in protein concentration in lateral ventricular CSF. At this age in the IVth ventricle the overall total was unchanged, although there were changes in concentration of individual proteins. At 60 days the concentration of each protein in both compartments had fallen below that at 40 days; the marked concentration difference between lateral and IVth ventricular CSF was still present. Experiments using i.v. [125I]- or [3H] labeled plasma proteins in 30-40-day fetuses showed that very little protein penetrated into lateral ventricular CSF by 3-5 h after injection; in the same experiments [125I]albumin reached a CSF/plasma ratio of about 15% in the IVth ventricle (compared with 55% for the natural steady state). Autoradiographic studies carried out on material from the same animals did not give evidence for transfer of labeled protein across the choroid plexuses although any such penetration may have been below the threshold of the method. Other explanations for the high concentration of protein in CSF that were considered include penetration via cerebral vessels and synthesis of plasma proteins by choroid plexus epithelial cells or neurons within the brain.