Erythropoietin gene expression in different areas of the developing human central nervous system

Developmental Brain Research 125 (2000) 69–74 www.elsevier.com / locate / bres

Research report

Erythropoietin gene expression in different areas of the developing human central nervous system Christof Dame a , Peter Bartmann a , Eva-Maria Wolber b , Hubert Fahnenstich c , d e, Dietmar Hofmann , Joachim Fandrey * a

Department of Neonatology, University Children’ s Hospital Bonn, Germany b ¨ Institute of Physiology, University of Lubeck , Germany c Department of Neonatology, University Children’ s Hospital Basel, Switzerland d Institute of Paediatric Pathology, University of Bonn, Germany e Institute of Physiology, University of Essen, Germany Accepted 26 September 2000

Abstract Evidence from cell culture and animal experiments suggests a neuroprotective and neurotrophic function of erythropoietin (EPO). We have quantitated the distribution of EPO mRNA expression in the developing human central nervous system (CNS). Patients and Methods: Up to seven biopsies from different areas of the CNS of four preterm fetuses (gestational age 23–37 weeks) were obtained at routine postmortem examinations. EPO mRNA was quantitated by competitive PCR in samples from the CNS, the kidneys, and the liver where the EPO gene is predominantly expressed at this gestational age. Results: EPO mRNA was most abundant in one sample from the cerebellum (0.29 amol / mg total RNA [amol510 218 mol]) and two from the pituitary gland (0.23 amol / mg total RNA), but levels varied considerably. EPO mRNA in the cortex cerebri (median 0.12 amol / mg total RNA; n54) dominated over the expression in the corpora amygdala (median 0.05 amol / mg total RNA; n54), the hippocampus (median 0.03 amol / mg total RNA; n54), or the basal ganglia (median 0.01 amol / mg total RNA; n53). Only little EPO mRNA (,0.01 and 0.06 amol / mg total RNA) was found in the spinal cord. EPO mRNA levels in the cerebellum, pituitary gland, or the cerebral cortex were within the same range as in the liver (0.03–1.67 amol / mg total RNA; n54), or the kidneys (0.06–0.79 amol / mg total RNA; n54). Conclusion: We found the EPO gene expressed throughout the fetal human CNS. Our data provide the basis to discuss a function for EPO in the brain of humans as well.  2000 Elsevier Science B.V. All rights reserved. Theme: Development and regeneration Topic: Hormones and development Keywords: Brain development; Neuroprotection; Neurogenesis; Hematopoiesis; Fetus

1. Introduction Erythropoietin (EPO) is the primary regulator of human erythropoiesis [6]. In the human fetus, the liver is the main site of EPO gene expression [4]. After the 30th week of gestation renal EPO gene expression increases, and this

Abbreviations: CNS, central nervous system; CSF, cerebro-spinal fluid; EPO, erythropoietin; EPO-R, erythropoietin receptor; GAPDH, glycerol aldehyde 3-phosphate dehydrogenase; NIHF, non-immune hydrops fetalis *Corresponding author. Tel.: 149-201-723-4600; fax: 149-201-7234648. E-mail address: [email protected] (J. Fandrey).

may indicate the beginning switch of the EPO production site from the fetal liver to the kidneys in adults [4]. Initially, EPO as a hematopoietic hormone was believed to act exclusively on erythroid progenitor cells [6]. Meanwhile, EPO gene expression has also been found in the bone marrow, spleen, gastrointestinum, heart, lung, testis and ovary, and the central nervous system [4,7,10]. In human adults, EPO mRNA is expressed in the temporal cortex, the corpora amygdala, and the hippocampus, all parts of the telencephalon [15]. In early human fetal life, EPO protein is found in the periventricular germinal matrix zone and the subpial granular layer [11]. Particularly in the thalamus, the hippocampus, the lateral

0165-3806 / 00 / $ – see front matter  2000 Elsevier Science B.V. All rights reserved. PII: S0165-3806( 00 )00118-8

70

C. Dame et al. / Developmental Brain Research 125 (2000) 69 – 74

geniculate body, the cortex, and the spinal cord neuronal cytoplasmatic staining for EPO has been noticed which was pronounced with progression of development [11]. The EPO gene is expressed in astrocytes as well as in neuronal precursors and mature neurons [1,8,17]. EPO gene expression in the CNS is inducible by hypoxia up to 20-fold as shown by in vitro as well as in vivo animal experiments [15,17]. EPO protein was detected in the cerebrospinal fluid (CSF) where its concentration seems to decrease during the first months of life [9]. The EPO receptor (EPO-R) is expressed in astrocytes and neurons in various areas of the fetal and adult CNS [7,11,14]. Although there has been some discussion whether the binding characteristics of the EPO-R in the CNS differ from those on erythroid progenitors [19], in vitro studies have provided evidence that the EPO-R on neuronal cells is functional [16,18,22]. Sakanaka and colleagues [22,21] have demonstrated that EPO may protect neuronal precursors and neurons from ischemic and hypoxic damage in vivo. EPO will not pass the intact blood–brain barrier due to its size and highly glycosylated structure, and thus it has to be produced in the CNS [9]. In consequence, EPO in the CNS may play an important role in neurodevelopment and protection which is supported by the identification of a neurotrophic sequence in the EPO protein [2]. EPO is thought to have a role in the persistence of normal pluripotent progenitor cells in the neurogenesis of the adult brain [23]. Herein, we have studied the quantitative distribution of EPO gene expression in various areas of the developing human CNS by use of a competitive PCR. EPO mRNA levels in different brain areas were compared with those in liver and kidney.

2. Materials and methods

2.1. Experimental procedures 2.1.1. Patients /subjects Tissue specimens were taken from four deceased preterm neonates during routine postmortem examination after written parental consent was obtained. The study was done according to the principles of the Declaration of Helsinki and approved by the local Human Research Committee. The gestational age of the patients ranged from 23 to 37 weeks post conception. Only patient D did not die immediately after birth (day 33). The body weight varied from 643 to 3687 g. Patients’ diagnoses were as follows: Nonimmune hydrops fetalis (NIHF) associated with endocardial fibroelastosis (A), NIHF associated with mediastinal teratome (B), skeletal dysplasia and congenital heart failure (C), and severe congenital stenosis of the aortic valve (D). No patient suffered from macroscopically visible malformations of the CNS. Immediately after death, peripheral blood cell counts

were determined when possible. Blood was collected from a central or peripheral vein or the umbilical vein. Serum was removed immediately and stored at 2208C until measurement of the EPO concentration. Tissue specimens were obtained from liver and kidney and from four to seven various areas of the CNS and snap-frozen in liquid nitrogen. To prevent tissue degradation, the following precautions were taken: The bodies of deceased patients were stored at 148C until postmortem examination. All tissue samples from one patient were taken immediately, one after the other to avoid exposure of the organs to room air and room temperature after opening the abdomen or the calotte. The preparation of the brain was always done by the same procedure and the same investigator. Tissue specimens were taken from the following areas of the CNS: Spinal cord representing the myelencephalon; cerebellum representing the metencephalon; pituitary gland representing the diencephalon; basal ganglia, amygdala, hippocampus, and frontal cortex cerebri representing the telencephalon. CSF was collected at the beginning of the preparation of the brain, and stored at 2208C after centrifugation (1000 rpm for 2 min).

2.2. Analysis of EPO mRNA EPO gene expression was investigated by the detection of EPO mRNA in tissue specimen exactly as described previously [4,5]. Total RNA was extracted by the acidic phenol-chloroform method [3]. Five mg of total RNA was reverse transcribed into first strand cDNA using M-MLV reverse transcriptase (GIBCO, Life Technologies, Eggenstein, Germany) with oligo-dT 15 as primer [5]. Initial PCRs for glycerol aldehyde 3-phosphate dehydrogenase (GAPDH) were done to ensure the integrity and comparable amounts of cDNA as described earlier [4,24]. For quantitation of EPO cDNA, competitive PCR was performed in PCR buffer (20 mM Tris–HCl pH 8.4, 50 mM KCl, 1.5 mM MgCl 2 ), 200 mM of each dNTP, 400 nM of each 59primer (59-TCT GGG AGC CCA GAA GGA AGC CAT-39) and 39primer (59-CTG GAG TGT CCA TGG GAC AG-39) and 0.75 units of Taq polymerase (GIBCO, Eggenstein, Germany) in a final volume of 50 ml. PCR was run for 30–35 cycles as described previously [4,5]. Calculation of the EPO mRNA concentration was performed exactly as described earlier [5]. Amplification of potentially contaminating genomic DNA was avoided by the use of intron IV spanning primers. Negative controls (no template cDNA) excluded contamination with PCR products from previous amplifications. The lower limit of reliable quantitation by competitive PCR was 0.03 amol (amol5attomol510 218 mol) of competitor which roughly corresponds to 5000 molecules. We calculated EPO mRNA concentrations per microgram total RNA and per gram tissue. The intra-assay variability of our competitive PCR was 16% in a comparison of RNA samples reverse transcribed and quantitated in parallel during the same

C. Dame et al. / Developmental Brain Research 125 (2000) 69 – 74

experiment. The inter-assay variability of 18% was derived from RNA samples reverse transcribed on different days and quantitated in separate experiments [24].

2.3. Determination of serum EPO levels EPO concentrations in serum and CSF were measured by an enzyme-linked immunoassay (EPO ELISA , Medac, Hamburg, Germany) following the manufacturer’s instructions. Detection limit for EPO was 1.25 U / l.

3. Results

3.1. EPO mRNA expression in the developing human central nervous system EPO mRNA expression was detected in each tissue specimen of the CNS. Fig. 1 shows the result of a representative qualitative RT-PCR for EPO mRNA from

71

two patients. Samples from different areas of the CNS showed different amounts of amplification products from EPO mRNA (Fig. 1A) while corresponding GAPDH mRNA levels in these brain samples appeared equal (Fig. 1B). The results of the quantitative competitive PCR are summarized in Table 1. Of the two samples from the spinal cord in one (Patient A) EPO mRNA was detectable but lower than 0.03 amol / mg total RNA although a sufficient amount of total RNA / g tissue was obtained (see Table 2). EPO mRNA expression in the cerebellum (EPO mRNA 0.09–0.29 amol / mg total RNA) and the pituitary gland (EPO mRNA 0.01–0.23 amol / mg total RNA) showed large variations. However, in two of these samples from areas representing the metencephalon and the diencephalon, respectively, the highest absolute EPO mRNA expression / mg total RNA was found. Due to the small size of the pituitary gland at that gestational stage and the preparation procedure, it was not possible to differentiate between tissue from the adeno- or neurohypophysis. The telencephalon was large enough to

Fig. 1. Qualitative analysis of EPO gene expression in fetal organs of two patients. Representative RT-PCR results for EPO mRNA from patient A (upper half) and patient B (lower half) are shown (Panel A). The GAPDH signal did not differ between organs, confirming constant efficiency of RNA preparation and cDNA synthesis (Panel B; patient A (upper half) and patient B (lower half)). The intensity of the EPO signal varied between different organs. For better visualization gel photographs are presented as negatives of the originals. MWM5molecular weight marker; 100 bp DNA ladder.

C. Dame et al. / Developmental Brain Research 125 (2000) 69 – 74

72

Table 1 EPO mRNA expression in different areas of the developing human CNS in comparison to liver and kidney a Gestational age [week]

A B C D a

23 28 35 37

Myelencephalon Spinal cord ,0.01 0.06 n.a. n.a.

Metencephalon Cerebellum 0.09 0.29 n.a. n.a.

Diencephalon Pituitary gland 0.23 0.01 0.23 n.a.

Telencephalon Basal ganglia

Amygdala

0.09 n.a. 0.01 0.01

0.06 0.03 0.09 0.03

EPO [U / l] Hippocampus

Cortex cerebri

CSF

Serum

0.03 0.09 0.03 0.03

0.12 0.12 0.70 0.12

17 n.a. 7 51

n.a. 9 n.a. 54

Liver

Kidney

0.67 0.06 1.69 0.03

0.13 0.16 0.79 0.06

Values are EPO mRNA concentrations in amol / mg total RNA. n.a., not available.

take samples from four different areas to compare EPO gene expression. EPO mRNA in the cortex cerebri (median 0.12 amol / mg total RNA) was higher than in the corpora amygdala (median 0.05 amol / mg total RNA), the hippocampus (median 0.03 amol / mg total RNA), or the basal ganglia (median 0.01 amol / mg total RNA). EPO gene expression did not correlate with patient’s diagnosis, gestational age, age at death, or any laboratory parameter.

3.2. EPO mRNA expression in liver and kidney EPO mRNA was detectable in each specimen taken from liver and kidney. The median of the EPO mRNA levels expressed in amol / mg total RNA was three-fold higher in the liver than in the kidney. Compared to the amount of EPO mRNA expression in the liver or the kidney of the same patient, the expression of EPO mRNA / mg total RNA in some areas of the CNS, particularly in the cerebellum, the pituitary gland, or the cerebral cortex was within the same range (see Table 1). Of note, the amount of total RNA per gram tissue varied considerably between organs from a single patient (which has to be considered to determine the contribution of the organ to total EPO expression), but also between the same organ from different patients (Table 2). Still, the highest amount of total mRNA / g tissue was isolated from the liver (median 1328 mg total RNA / g tissue) followed by the kidneys (median 592 mg / g). In the developing CNS, in

each case the total amount of mRNA / g tissue was clearly lower than in the liver or kidney. With the exception of the basal ganglia in patient A and the pituitary gland in patient B, the amount of total RNA / g tissue was not too widely ranged, suggesting equal quality of the extraction procedure for the different samples. If one does not consider the two extreme values that may have been caused by the specific anatomic origin of the specimen (e.g. a specific nucleus in the basal ganglia) there may be areas in the developing human CNS where EPO mRNA / g tissue is expressed to a similar or even higher degree than in the kidneys.

3.3. EPO levels in serum and cerebrospinal fluid We found similar EPO concentrations in the CSF (median 17 U / l; range 7–51 U / l) and in serum samples (range 9–54 U / l) (see Table 1).

4. Discussion In the developing human brain EPO mRNA is expressed in the myelencephalon, the metencephalon, the diencephalon, and the telencephalon. For the first time, we were able to detect EPO gene expression in the human fetal cerebellum and the pituitary gland. We also confirmed results from earlier studies in which EPO mRNA has been detected in the developing and adult human brain [7,15].

Table 2 EPO mRNA expression per gram tissue in different areas of the developing human CNS in comparison to liver and kidney a Gestational age [week]

Myelencephalon Spinal cord

Metencephalon Cerebellum

Diencephalon Pituitary gland

2.44 28 9.57 33 n.a. n.a.

6.33 27 0.04 3 24.13 104 n.a.

A

23

B

28

C

35

n.a. 43 1.72 30 n.a.

D

37

n.a.

a

Telencephalon Basal ganglia

Amygdala

Hippocampus

Cortex cerebri

35.32 406 n.a.

9.48 163 0.59 20 4.21 48 0.35 12

1 35 3.47 40 0.38 13 0.31 11

8.47 73 6.45 56 58.08 83 12.93 111

0.79 68 0.48 41

Liver

Kidney

603 1021 314 2707 2670 1629 68 1028

52 890 30 1020 81 116 4 294

Values are EPO mRNA concentrations in amol / g tissue. Values in italic letters represent the total RNA / g tissue. n.a., not available.

C. Dame et al. / Developmental Brain Research 125 (2000) 69 – 74

Apparently, EPO mRNA is widely expressed in the fetal brain and throughout gestation. This may be important in understanding the biological role of EPO production in the central nervous system. In animal experiments recombinant EPO has been found to protect neuronal precursors and differentiated cells from cell death when they were exposed to hypoxia, glutamate, or UV irradiation [7,18,21,22]. From cell culture studies a neurotrophic role of EPO has been proposed that would make the hematopoietic growth factor an important contributor to the development and neurogenesis of the brain [16,23]. Interestingly, an amino acid sequence comparison between EPO, thrombopoietin, nerve growth factor b and neurotrophins revealed areas of high homology [13]. Campana and colleagues [2] identified a 17-mer peptide with neurotrophic function which is identical to that part of the EPO protein that extends from amino acid 29 into the AB-loop of EPO. This region, however, overlaps the region of homology with neurotrophins only by the last four amino acids. In fact, if one compares these amino acids with the corresponding residues in neurotrophin 3 or nerve growth factor b there is no homology even of conserved amino acids. Thus, one may conclude that a lack in structural homology with neurotrophins and EPO peptide does not exclude potential neurotrophic effects [2]. The fact that the EPO receptor is expressed throughout the human fetal brain [7] and that EPO signaling via the EPO receptor can be induced in cells of neural origin [16] further indicates that EPO may play a role in neurogenesis in the developing human brain like it has been suggested from studies in mice [25]. This hypothesis becomes even more appealing in view of a recent report that in the human fetal brain the production of the EPO receptor may be developmentally regulated [11]. For the first time, EPO gene expression was detected in the human cerebellum. In one patient (B), very high EPO mRNA levels were found, higher than in any of the other brain samples (Table 1). During ontogeny of the cerebellum, cerebellar neurons are generated within the neuroepithelium lining the rostral half of the fourth ventricle. The neurons which are determined to form both Purkinje cells and neurons of the deep cerebellar nuclei arise from the subventricular zone [20]. In this area, during early human gestation a strong signal of cells containing EPO was described [11]. Since the calculated amount of EPO mRNA per gram cerebellar tissue was also considerably higher in patient B than in the other cerebellar samples, we may in fact have included some deep nuclei in this tissue specimen which could express high levels of EPO mRNA. We also detected EPO mRNA expression in the pituitary gland. In two out of three patients higher amounts of EPO mRNA than in any other investigated area of the brains were measured. However, we cannot define whether the adenohypophysis and / or the neurohypophysis expresses EPO. No data from animal experiments are available to compare this finding. In the murine brain a high density of

73

blood capillaries builds an ideal region in this area for sensing hypoxia or intrauterine stress as has been already shown for neurohypophyseal hormones [12]. These peptides are secreted into the CSF, and it is of interest whether EPO secretion into the CSF is regulated by similar mechanisms. This would explain the relatively high EPO protein concentrations in the CSF that do not correlate with the serum concentration [9] (and own unpublished results). The capacity of the diencephalon and the metencephalon to strongly express EPO mRNA correlates well with intense staining for EPO protein in early gestation in these brain areas [11]. At 5–6 weeks post conceptionem EPO reactivity was most prominent in the periventricular germinal matrix zone localized immediately ventrolateral to the lateral ventricles. This area serves as the source of cerebral neuroblasts between 10 and 20 weeks of gestation; in the third trimester of pregnancy it provides glioblasts that become cerebral oligodendroglia and astrocytes. Many thin-walled vessels are localized in this area. The matrix undergoes a progressive decrease in size during later gestation and a nearly complete involution by approximately 36 weeks of gestation. During development and also during neurogenesis in the adult brain, neurons migrate from the germinal matrix zone to deeper brain areas [20]. Thus, if EPO contributes to neuronal development, EPO mRNA expressing cells may originate from here. Our data allow us to discuss EPO gene expression in the telencephalon in detail. Except for the basal ganglia in one patient (A), we found similar amounts of total RNA / g tissue in each investigated area of the telencephalon. As mentioned above, the most plausible cause for the exceptional high total RNA in the basal ganglia may be caused by a higher expression in a specific, not yet identified nucleus. Whereas the amount of EPO mRNA / mg total RNA was similar in basal ganglia, amygdala, and hippocampus, EPO gene expression was higher in the cortex cerebri. Juul et al. [11] reported EPO immunoreactivity throughout the cortical wall, with the most intense staining in the ventricular and subventricular zones at 10 weeks post conception. Basal ganglia originate from this area which may be the reason for a higher capacity of expressing the EPO gene. Marti et al. [15] found hypoxic inducibility of EPO gene expression in the monkey brain. This may explain the differences we found in EPO gene expression within various areas of the telencephalon, since prefinal perfusion of the brain is different. Herein, we show quantitative data with respect to EPO mRNA distribution in the central nervous system. This allowed us to compare the EPO mRNA concentrations in CNS tissues to EPO mRNA levels in the liver and kidney which are the predominant sites of EPO gene expression during fetal life in humans [4]. Surprisingly, despite a significantly higher expression of EPO mRNA / g tissue in the liver than in the brain, as described in a previous study [4], the amount of EPO mRNA / mg total RNA in the liver

74

C. Dame et al. / Developmental Brain Research 125 (2000) 69 – 74

was not always higher than in the areas of the CNS of the same patient. The finding that EPO is expressed to a degree comparable to that in liver and kidneys which are the most relevant sites for circulating EPO, extends previous data on EPO in the CNS, since EPO may be sufficiently expressed to exert a function of biological significance in the brain.

[12]

[13] [14]

Acknowledgements

[15]

We thank Patricia Freitag and Gabi Strackbein for excellent technical assistance.

[16]

References [17] [1] M. Bernaudin, A. Bellail, H.H. Marti, A. Yvon, D. Vivien, I. Duchatelle, E.T. Mackenzie, E. Petit, Neurons and astrocytes express EPO mRNA: oxygen-sensing mechanisms that involve the redox-state of the brain, GLIA 30 (2000) 271–278. [2] W.M. Campana, R. Misari, J.S. O’Brien, Identification of a neurotrophic sequence in erythropoietin, Int. J. Mol. Med. 1 (1998) 235–241. [3] P. Chomczynski, N. Sacchi, Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction, Anal. Biochem. 162 (1987) 156–159. [4] C. Dame, H. Fahnenstich, P. Freitag, D. Hofmann, T. Abdul-Nour, P. Bartmann, J. Fandrey, Erythropoietin mRNA expression in human fetal and neonatal tissue, Blood 92 (1998) 3218–3225. [5] J. Fandrey, H.F. Bunn, In vivo and in vitro regulation of erythropoietin mRNA: Measurement by competitive polymerase chain reaction, Blood 81 (1993) 617–623. [6] W. Jelkmann, Erythropoietin: Structure, control of production and function, Physiol. Rev. 72 (1992) 449–489. [7] S.E. Juul, D.K. Anderson, Y. Li, R.D. Christensen, Erythropoietin and erythropoietin receptor in the developing human central nervous system, Pediatr. Res. 43 (1998) 40–49. [8] S.E. Juul, R.J. Ferguson, R.D. Christensen, Erythropoietin induces differential gene expression in NT2 cells, Pediatr. Res 45 (Suppl. 1) (1999) 263 A. [9] S.E. Juul, J. Harcum, Y. Li, R.D. Christensen, Erythropoietin is present in the cerebrospinal fluid of neonates, J. Pediatr. 130 (1997) 428–430. [10] S.E. Juul, A.T. Yachnis, R.D. Christensen, Tissue distribution of erythropoietin and erythropoietin receptor in the developing human foetus, Early Hum. Dev. 52 (1998) 235–249. [11] S.E. Juul, A.T. Yachnis, A.M. Rojiani, R.D. Christensen, Immuno-

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

histochemical localisation of erythropoietin and its receptor in the developing human brain, Pediatr. Dev. Pathol. 2 (1999) 148–158. R.D. Leake, Foetal and neonatal neurohypophyseal hormones, in: R.A. Polin, W.W. Fox (Eds.), Foetal and Neonatal Physiology, W.B. Saunders, Philadelphia, 1992, pp. 1815–1819. B. Li, W. Dai, Thrombopoietin and neurotrophins share a common domain, Blood 86 (1995) 1643–1644. Y. Li, S.E. Juul, J.A. Morris-Wiman, D.A. Calhoun, R.D. Christensen, Erythropoietin receptors are expressed in the central nervous system of mid-trimester human foetuses, Pediatr. Res. 40 (1996) 376–380. H.H. Marti, R.H. Wenger, L.A. Rivas, U. Straumann, M. Digicaylioglu, V. Henn, Y. Yonekawa, C. Bauer, M. Gassmann, Erythropoietin gene expression in human, monkey and murine brain, Eur. J. Neurosci. 8 (1996) 666–676. S. Masuda, M. Nagao, K. Takahata, Y. Konishi, F. Gallyas Jr., T. Tabira, R. Sasaki, Functional erythropoietin receptor of the cells with neural characteristics. Comparison with receptor properties of erythroid cells, J. Biol. Chem. 268 (1993) 11208–11216. S. Masuda, M. Okano, K. Yamagishi, M. Nagao, M. Ueda, R. Sasaki, A novel site of erythropoietin production, J. Biol. Chem. 269 (1994) 19488–19493. E. Morishita, S. Masuda, M. Nagao, Y. Yasuda, R. Sasaki, Erythropoietin receptor is expressed in rat hippocampal and cerebral cortical neurons, and erythropoietin prevents in vitro glutamate-induced neuronal death, Neuroscience 76 (1997) 105–116. K.M. Moritz, G.B. Lim, E.M. Wintour, Developmental regulation of erythropoietin and erythropoiesis, Am. J. Physiol. 273 (1997) R1829–R1844. P. Rakic, R.L. Sidman, Histogenesis of cortical layers in human cerebellum, particularly the lamina dissecans, J. Comp. Neurol. 139 (1970) 473–500. Y. Sadamoto, K. Igase, M. Sakanaka, K. Sato, H. Otsuka, S. Sakaki, S. Masuda, R. Sasaki, Erythropoietin prevents place navigation disability and cortical infarction in rats with permanent occlusion of the middle cerebral artery, Biochem. Biophys. Res. Com. 253 (1998) 26–32. M. Sakanaka, T.C. Wen, S. Matsuda, S. Masuda, E. Morishita, M. Nagao, R. Sasaki, In vivo evidence that erythropoietin protects neurons from ischaemic damage, Proc. Natl. Acad. Sci. 95 (1998) 4635–4640. B. Scheffler, M. Horn, I. Blumcke, E.D. Laywell, D. Coomes, V.G. Kukekov, D.A. Steindler, Marrow-mindedness: a perspective on neuropoiesis, Trends Neurosci. 2 (1999) 348–357. E.-M. Wolber, C. Dame, H. Fahnenstich, D. Hofmann, P. Bartmann, W. Jelkmann, J. Fandrey, Expression of the thrombopoietin gene in human fetal and neonatal tissues, Blood 94 (1999) 97–105. Y. Yasuda, M. Nagao, M. Okano, S. Masuda, R. Sasaki, H. Konishi, T. Tanimura, Localisation of erythropoietin and erythropoietinreceptor in postimplantation mouse embryos, Develop. Growth Diff. 35 (1993) 711–722.