SEARCH

SEARCH BY CITATION

Keywords:

  • Fetuin-A;
  • preterm infant;
  • brain maturation;
  • immunohistochemistry;
  • neuroprotection

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

The serum protein fetuin-A is essential for mineral homeostasis and shows immunomodulatory functions, for example by binding to TGF superfamily proteins. It proved neuroprotective in a rat stroke model and reduced lethality after systemic lipopolysaccharide challenge in mice. Serum fetuin-A concentrations are highest during intrauterine life. Different species show intrauterine cerebral fetuin-A immunoreactivity, suggesting a contribution to brain development. We therefore aimed at specifying fetuin-A immunoreactivity in brains of newborn rats (age P0–P28) and human neonates (20–40 weeks of gestation). In humans and rats, fetuin-A was found in cortex, white matter, subplate, hippocampus, subventricular zone, and ependymal cells which supports a global role for brain function. In rats, overall fetuin-A immunoreactivity decreased with age. At P0 fetuin-A immunoreactivity affected most brain structures. Thereafter, it became increasingly restricted to distinct cells of the hippocampus, cingular gyrus, periventricular stem cell layer, and ependyma. In ependymal cells the staining pattern complied with active transependymal transport from cerebrospinal fluid. Double immunofluorescence studies revealed colocalization with NeuN (mature neurons), beta III tubulin (immature neurons), GFAP (astrocytes), and CD68 (activated microglia). This points to a role of fetuin-A in different brain functional systems. In human neonatal autopsy cases, frequently affected from severe neurological and non-neurological diseases, fetuin-A immunoreactivity was heterogeneous and much less associated with age than in healthy tissues studied earlier, suggesting an impact of exogeneous noxious factors on fetuin-A regulation. Further research on the role of fetuin-A in the neonatal brain during physiological and pathological conditions is recommended. © 2012 Wiley Periodicals, Inc. Develop Neurobiol 73: 354–369, 2013


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Fetuin-A is a serum glycoprotein and member of the cystatin protein family, mainly produced in the liver. The biological functions of fetuin-A include positive and negative interactions with proteases, binding to and inhibition of members of the TGF-beta superfamily (including TGFbeta 1, TGF beta 2, and bone morphogenic proteins (BMP)), binding to Toll-like receptor 4, to growth hormone (GH), nerve growth factor (NGF), platelet-derived growth factor (PDGF), transforming growth factor II and basic fibroblast growth factor (bFGF), and lipid transport (Nie, 1992; Jahnen-Dechent et al., 2011; Pal et al., 2012). Above all, fetuin-A has an important role in mineralization biology which is reflected by its high affinity for calcium phosphate minerals and related compounds and the very high abundance of fetuin-A in mineralized bone matrix (Jahnen-Dechent et al., 2011; Herrmann et al., 2012).

Fetuin-A was initially described as the most abundant globular protein present in fetal calf serum. High intrauterine concentrations have been reported for many animals and humans, and we have recently confirmed this finding in a larger cohort of neonates. We found that fetuin-A serum levels are highest in very low birth weight infants between 24 and 30 completed weeks of gestational age, showing levels twice to three times higher than observed after term birth. Fetuin-A levels slowly decreased with intrauterine as well as extrauterine maturation, and reached adult levels at about 37 completed weeks of gestation (Häusler etal., 2009). Fetuin-A is a negative acute phase protein in that hepatic synthesis is downregulated by proinflammatory cytokines (Daveau et al., 1988; Ohnishi et al., 1994). This downregulation during infections was also seen in our neonatal patients suggesting that this regulatory mechanism is also operative in early intrauterine life. In addition to high intrauterine serum levels, fetuin-A is also abundant in intrauterine cerebrospinal fluid (CSF) (Dziegielewska et al., 1993).

In 2010 Wang et al., showed that fetuin-A may also have a neuroprotective function, as fetuin-A supplementation reduced cerebral infarction size in a rat model of middle cerebral artery occlusion (Wang et al., 2010; Wang and Sama, 2012). This effect was explained by an anti-inflammatory effect of fetuin-A, leading to reduced TNF-alpha synthesis as well as reduced macrophage and microglial activation. The inhibition of TNF-alpha synthesis was related to spermin with fetuin-A as a carrier. Similar to fetuin-A, spermin also shows strong intrauterine expression. It has been assumed to contribute to intrauterine immunotolerance and to reduce TNF-alpha-related cell damage (Chertov et al., 1994). In line with these observations, fetuin-A application has been shown to improve survival after LPS challenge in mice (Li et al., 2011) and in rats (Dziegielewska and Andersen, 1998). This includes reduced synthesis of HMGB1, which is released from activated macrophages and monocytes, and contributes to a late inflammatory response. Lipid-laden fetuin-A may have proinflammatory properties, as fetuin-A may serve as an endogenous ligand for Toll-like receptor 4, mediating adipose tissue inflammation by free fatty acids, thus contributing to insulin resistance and type 2 diabetes (Pal et al., 2012).

The high intrauterine serum and CSF levels of fetuin-A, its anti-inflammatory and neuroprotective effect in a stroke model, and its interactions with growth factors or cytokines point out fetuin-A as a factor involved into brain maturation and/or protection. In line with this hypothesis, fetuin-A has already been detected in different regions of the developing human, sheep, rat and cow brain, including the cortical plate, the subplate, the cingulum, the subventricular zone, ventricular zone, intermediate zone, thalamus and hippocampus (Mollgard et al., 1984; Dziegielewska et al., 1987, 1993, 1997, 2003; Reynolds etal., 1987; Saunders et al., 1992, 1994). Accordingly, during the transition phase from the preterm age (P0 in rats) to the term-born neonate (P8–P14 in rats) cerebral fetuin-A expression seems to show marked time-related changes.

To study the impact of fetuin-A during brain maturation as well as in the context of ischemic and inflammatory brain damage, more detailed knowledge on these age-related changes is needed. We therefore conducted the present study, focusing on two major goals. First, we confirmed fetuin-A immunoreactivity at different ages of gestation in different regions of the newborn human brain. Second, we specified these time-related changes in newborn rats.

METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Tissue Samples

Human brain tissue samples derived from 24 children of 24–41 completed weeks of gestational age who had been treated at RWTH Aachen University Hospital between the years 1990 and 2010. The children had died from different underlying conditions within the first four weeks of life. Clinical details of the patients are summarized in Table 1. In addition to the neonatal cases, brain tissue obtained at autopsy of three randomly selected fetal cases not showing cerebral pathology was also included. All tissues had been prepared as follows. At first, the whole brain was fixed for about 48 h in buffered 10% formaldehyde aqueous solution. After additional fixation in 4% formaldehyde solution for several days, the formaldehyde was removed by rinsing in tap water. Then the brains were dissected and the tissue was dehydrated in an automated tissue processor and manually embedded in paraffin. Tissue sections of 5 μm were used for all staining procedures.

Table 1. Clinical Data of Newborn Children Studied for Cerebral Fetuin-A Immunoreactivity
Patient no.Age at death (corrected weeks of gestation)Age at death (days after birth)Female/maleClinical diagnosis
1240MAsphyxia
2264MIntracerebral hemorrhage
3261MAsphyxia
42712FMeconium ileus, hyperkaliemia, intracerebral hemorrhage
5272MAsphyxia, intracerebral hemorrhage
6271MSyndromal disorder
7280MTeratoma
8291FHyaline membranes, respiratory failure
93014FHyaline membranes, respiratory failure
103013FAsphyxia, circulatory failure
11310FSyndromal disorder, pulmonary hypoplasia
123314MAsphyxia, subependymal hemorrhage
133317FCandida septicemia
143310MStaphylococcal septicemia
15320MPrune belly syndrome
16330MAsphyxia, Prune belly syndrome
17331FSyndromal disorder, ectopia cordis
18340FAsphyxia, syndromal disorder
19341FSudden neonatal death
20372FCongenital ichthyosis
21406MCongenital heart disease
22400MAsphyxia
23403FCongenital heart disease
24412MCongenital heart disease

For animal experiments Wistar rats of 0–28 days of age were used, sacrificing at least three rat pups each on postnatal days (P) 0, 3, 5, 7, 9, 10, 12, 14, 16, 21, 24, and 28. P0 defines the day of birth. Male and female rats were equally distributed between the different age groups. To assure that intravascular plasma could provide an internal positive control for fetuin-A staining, the rats were not perfused. After removal of the brains, the cerebellum and brain stem were discarded whereas the remaining forebrains were immersion-fixed for a minimum of 48 h in Bouins' fixative (75 mL picric acid, 25 mL formaldehyde 37%, 5 mL acetic acid 100%). In some animals, liver tissue was also removed to provide an additional positive control. All brain tissues were cut in a frontal plane and placed into individual cassettes for dehydration by an automated tissue processor (SLEE, Mainz, Germany) and embedding into paraffin (Leica EG1160 embedding system). For histopathological studies 3–5 μm tissue sections were used. All animal tissues were obtained in line with the German laws on animal experiments.

Immunohistochemistry for Fetuin-A

For fetuin-A staining of human brains we used a monoclonal IgG mouse anti-human antibody (clone MAHS-1, dilution 0.5–1 μg/mL), raised against purified human fetuin-A/α2HS-glycoprotein (Behring). Antibody specificity had been confirmed by Western blotting using human serum (Nawratil et al., 1996). A biotinylated goat anti-mouse IgG antibody (Vector, BA-9200; dilution: 1:300) was used as secondary antibody. All antibodies were diluted in Dulbecco's phosphate buffered saline (PBS; Biochrom AG, L 182-50) containing 1% bovine serum albumin (BSA; Sigma-Aldrich, A-9647). For deparaffination, the slides were incubated in xylene (60 minutes, room temperature) followed by incubation in increasing concentrations of ethanol (100%–70%), followed by washing in PBS. This was followed by antigen retrieval (30 minutes heating in a steamer (Tefal) while incubated in 10 mM citrate buffered saline). Peroxidase blocking by hydrogen peroxide was omitted as this had proven unnecessary. After cooling and washing in PBS, blocking of nonspecific binding sites was performed by incubation in 10% normal goat serum in PBS (Dako, X0907) using a moisture chamber (Roth, HA51.1; duration 20 minutes). To minimize nonspecific binding of the streptavidin-labeled horse-radish peroxidase (Sav-HRP) to endogenous biotin or lectins, a streptavidin-biotin blocking kit was used (Vector, SP-2002): Sav was added to the protein blocking solution to mask tissue biotin, and later biotin was added to the diluted primary antibody to saturate Sav binding sites (four drops of Sav or biotin per mL of blocking solution or diluted primary antibody, respectively). Thereafter, the slides were submerged in demineralized water followed by washing in PBS for 5 minutes. Then, the tissue was incubated with 100 μL of diluted MAHS-1 antibody (overnight, moisture chamber, room temperature). After washing in demineralized water and PBS, 100 μL of diluted, biotinylated secondary antibody was added (moisture chamber, 30 minutes). After repeated washing in demineralized water and PBS, 100 μL ready-to-use Sav-HRP solution (BD Pharmingen, 51-75477E) was added (moisture chamber, at least 30 minutes), followed by a further washing step, applying demineralized water and PBS. Thereafter 200 μL staining solution, containing buffer, hydrogen peroxide and DAB was added for 2 minutes (Dako, SK-4100), followed by washing in demineralized water. The slides were then counterstained (Mayer's hematoxylin solution; Roth, T865.1, one minute), washed in demineralized water (9 minutes), dehydrated in ethanol (concentrations from 70% to 100%), mounted (Roth, T160.1), and covered using coverslips.

For fetuin-A staining of rat tissues, the polyclonal rabbit anti-rat fetuin-A antibody EFM2 was diluted 1:3000 and used as described (Terkelsen et al., 1998). A biotinylated goat anti-rabbit IgG antibody (DAKO, E0432; dilution: 1:300) was used as secondary antibody. All antibodies were diluted in Dulbecco's phosphate buffered saline (PBS; Biochrom AG, L 182-50) containing 1% bovine serum albumin (BSA; Sigma-Aldrich, A-9647). The detailed staining procedure was similar to the staining of human tissue with the following exceptions: For deparaffinization the incubation period in xylene could be reduced to 15 minutes. To remove residual yellow Bouins' fixative, the sections were washed with 0.1M Tris buffer (AppliChem; pH 8.5; 3–5 minutes), followed by washing in PBS (three times for 5 minutes). Immunostaining of tissues fixed with Bouins' reagent did not require epitope retrieval procedures. For staining with the primary antibody, the sections were incubated with 50–75 μL of diluted rat fetuin-A antibody in a moist chamber at room temperature. After washing in demineralized water and additional washing in PBS, 50–75 μL of diluted, biotinylated secondary antibody was added for 30 minutes.

Double Immunofluorescence Staining for Colocalization of Fetuin-A with Further Antigens

To identify colocalization of fetuin-A with further cell-specific antigens double immunofluorescence staining was performed using commercially available monoclonal antibodies directed to GFAP (monoclonal mouse IgG anti-rat, clone GA-5, Santa Cruz, sc-58766; dilution: 1:100), NeuN (monoclonal mouse IgG anti-rat, clone 60, Chemicon/ Millipore, MAB377; dilution: 1:100), TUBB3 (monoclonal mouse anti-rat beta3-tubulin, clone Tuj-1, Novus Biologicals, NB100-57388; dilution: 1:100), and CD68 (monoclonal IgG mouse anti-rat, clone ED1, Abcam, ab 31630; dilution: 1:100). A donkey anti-mouse IgG Alexa® 488-conjugated antibody (Invitrogen, 21202; dilution: 1:300) served as secondary antibody. For fetuin-A staining we used the above-mentioned rabbit anti-rat fetuin-A antibody (1:2000), and a goat anti-rabbit Alexa® 546-conjugated secondary antibody (Invitrogen, 11010; dilution: 1:100). For staining procedures all antibodies were diluted in Dulbecco's phosphate buffered saline (PBS; Biochrom AG, L 182-50) containing 1% bovine serum albumin (BSA; Sigma-Aldrich, A-9647). The detailed procedure was as follows: Deparaffination, tissue rehydration, and removal of Bouins' fixative were performed as described above, as was the protein blocking procedure. Because of the simultaneous use of goat- and donkey-derived secondary antibodies, the blocking solution consisted of PBS containing 5% normal goat serum (Dako, X0907) and 5% normal horse serum (Biowest, 0900; incubation time 60 minutes). As no Sav/biotin block was used, the protein blocking solution was removed by tilting the slides only. To achieve double staining, 50–75 μL of diluted rat fetuin-A antibody and the same quantity of the second primary antibody, binding to a further cellular antigen, were pipetted onto the tissue sections. To prevent bleaching of the fluorophores, in all following steps the slides were maintained covered for light protection. The antibodies were allowed to incubate overnight in the moisture chamber at room temperature, after which excessive antibody solution was removed as described above, followed by an extra washing step in PBS. Then, 50–75 μL of the two diluted, fluorescent secondary antibodies were added for 90 minutes. Thereafter the slices were washed, mounted with ImmuMount® (Thermo Scientific, 9990402), and mounted with a coverslip. Coverslips were sealed with nail polish and slides were stored protected from light at 8°C.

Positive and Negative Controls

For all stainings of rat tissues, plasma-filled (and therefore fetuin-A-containing) vascular structures served as internal positive controls. As human tissues did not always display plasma-filled blood vessels, a liver tissue section was usually included as a further positive control. Negative controls for each specimen were also included in all staining experiments, undergoing the same staining processes as regular sections but without the primary antibody which was replaced by 1% BSA in PBS.

Visualization and Data Generation

Visualization of antibody-stained fetuin-A was achieved with a conventional bright field microscope (Leica). Images were acquired using DISKUS software (version 4.80.3922, Techn. Büro Hilgers, Königswinter), via a camera (JVC KY-F75U) directly connected to the microscope. The images were optimized for presentation using Photoshop software (Creative Suite 3, Adobe).

If applicable, the fetuin-A immunoreactivity of distinct anatomical structures was assessed semiquantitatively, using scoring system that differentiated between four different grades of fetuin-A immunoreactivity: 0, negative; 1, very sparse positive cells; 2, diffusely scattered positive cells; 3, focal, patchy or confluent accumulations of fetuin-A-positive cells. This scoring system was applied in regions of high cellular density, such as the ependyma, the stem cell layer or the rat cortex. In regions with low cellular density, such as the rat white matter and in the human hippocampus, a scoring system differentiating between three different grades of fetuin-A immunoreactivity was applied: 0, negative; 1, diffusely scattered positive cells; 2, focal, patchy or confluent accumulations of fetuin-A-positive cells. To assess the maximum diameter of the periventricular stem cell layer a virtual line, originating from the wall of the lateral ventricle and directed at the brain's surface was drawn at the thickest site of the periventricular stem cell layer. Then the number of cell rows crossed by this line was determined.

Visualization of double immunofluorescence staining was performed using a fluorescence microscope (Leica DMRX) equipped with a fluorescence lamp (Leica) and connected to a camera (JVC KY-F75U). Image capturing was performed using DISKUS software (version 4.80.3922, Techn. Büro Hilgers, Königswinter). Visualization parameters such as integration time were adjusted to avoid nonspecific signals, for instance resulting from autofluorescence. Fluorescence image fusion was carried out using PhotoShop software (Creative Suite 3, Adobe). Colocalization was studied in an exemplary fashion staining one animal per day and cell marker: In a first step, the immunoreactivity of additional markers was specified in distinct brain regions at distinct times after birth. Then, colocalization with fetuin-A was quantified in regions where substantial colocalization of fetuin-A and the respective cell marker was detected. When possible, counts were performed bilaterally, at a magnification of 400×. In all regions of interest colocalization was analyzed by applying a semiquantitative score, distinguishing between no, moderate, and marked costaining. For distinct regions we also performed a quantitative analysis, counting the number of marker-positive cells per 10 fetuin-A positive cells, or the number of fetuin-A-positive cells per 10 marker-positive cells. Hereby, depending on the size of the anatomical region 2–10 visual fields were analyzed.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Fetuin-A Immunoreactivity in Human Fetal and Neonatal Brain Tissue

In a first step we retrospectively studied by immunohistochemistry the immunoreactivity of fetuin-A in human fetal and neonatal brain tissues. Samples had been collected during routine autopsies performed during a period of 12 years. Depending on the patient age, the following areas of interest could be distinguished: the ventricular zones/subventricular zones with stem cell layers, the hippocampus, the cortex, the subplate, white matter structures, and occasionally basal ganglia.

Human Ependymal Cells Show Positive Fetuin-A Staining from Fetal Age to Term Birth

Three types of ependymal cells lining the lateral ventricle walls could be identified: flat, round, and prismatic cells, respectively. Whereas prismatic cells were the dominating cell type, all cell types did express fetuin-A. Fetuin-A immunoreactivity was not restricted to early intrauterine life but observed in all age groups. However, fetuin-A immunoreactivity was restricted to distinct ependymal cells sparing tightly adjacent cells [Fig. 1(A,B)].

image

Figure 1. Fetuin-A immunoreactivity in human neonatal brain tissue. Each subfigure displays the positive finding on the left side (*) and the negative control staining on the right side (**). In the preterm (A) and the term (B) infant the ependymal cells and periventricular stem cell layer contain distinct cells staining positive for fetuin-A. Strong fetuin-A immunoreactivity is also present in deep cortical neurons of a preterm child (C), affecting the cell body and axonal processes directed to the superficial cortical layers. Fetuin-A is further detected in distinct neurons of the dentate gyrus granular layer (D), in distinct neurons of the hippocampal CA3 sector (E), and cells within the white matter (F).

Download figure to PowerPoint

The Human Subependymal Gap Contains Fetuin-A-Positive Cells

Occasionally, fetuin-A-positive cell processes were found spanning the subependymal gap zone, which is part of the subventricular zone of the mature organism [Fig. 1(B)]. Throughout all gestational ages, fetuin-A-positive cells were also found within this gap.

From Fetal Age to Term Birth, the Human Periventricular Stem Cell Layer Shows Fetuin-A-Positive Cells

Throughout intrauterine life, periventricular stem cells, located within the SVZ, are considered the major source of neuronal cells. As shown in Figure 2(A), the size of this layer decreases with gestational age but does not completely disappear until birth. Within this layer, fetuin-A-positive cells were consistently found in all weeks of gestation [Figs. 2(C) and 1(A,B)].

image

Figure 2. Time course of fetuin-A immunoreactivity in periventricular stem cell layers and ependymal cells of human and rat brain tissues. The sizes of the human (A) and rat (B) periventricular stem cell layers decrease with age. Within these layers, fetuin-A-positive cells are present throughout the observation period (C, D). Distinct ependymal cells expressing fetuin-A are also found in all rat brains of all ages studied (D).

Download figure to PowerPoint

The Human Dentate Gyrus and CA3 Sector Are Major Sites of Hippocampal Fetuin-A Immunoreactivity

Similar to the stem cell and ependymal cell layers, hippocampal fetuin-A immunoreactivity was present at preterm age and in term born children. The highest number of fetuin-A expressing cells was found in the dentate gyrus and in the hippocampal CA3 sector (3A). In the latter most fetuin-A-positive cells were found adjacent to the dentate gyrus [Fig. 1(D,E)]. Hippocampal fetuin-A immunoreactivity varied in children of the same age group and showed interindividual differences.

Fetuin-A-Positive Cells Are also Present in the Human Cerebral Cortex, Subplate, and White Matter

Cortex and white matter could be studied in tissues deriving from 20 neonates age 24–41 completed weeks of gestation. About 50% of the samples showed fetuin-A immunoreactivity, affecting all regions studied and all age groups. In cortical areas, fetuin-A immunoreactivity showed a more focal, in white matter a more diffuse distribution pattern. Children of the same age group showed interindividual differences with regard to their fetuin-A staining patterns. Figure 1(C,F) shows examples for cortical and white matter fetuin-A immunoreactivity, respectively.

Fetuin-A Is also Present in Human Basal Nuclei and Thalamus

Thalamus and basal nuclei, including the caudate nucleus, putamen, and pallidum, were available from few cases only, occasionally also containing groups of large fetuin-A expressing neurons.

Fetuin-A Immunoreactivity in Rat Brains Age P0 to P28

The data summarized above strongly suggest a role of fetuin-A during brain development. As all human tissues were derived from gravely ill children, additional factors, including infections or hypoxia, may have influenced the cerebral fetuin-A immunoreactivity pattern. We therefore compared this immunoreactivity pattern to brains of healthy Wistar rats between days P0 to P28. Whereas P0 may represent a very premature human infant (Sheldon et al., 1996), the period between P8 and P14 is assumed to serve as a model of a term born human brain (Dobbing and Sands, 1979; Romijn et al., 1991).

Fetuin-A Positive Cells Are also Found in the Rat Subventricular Zone and Ependymal Layer

Like in human brain, the size of the periventricular stem cell layer decreased with age but did not completely disappear until P28 [Fig. 2(B)]. Similar to human tissue, ependymal cells of rat lateral ventricles and cells within the adjacent stem cell layers expressed fetuin-A. Whereas this was only observed for a short period in ependymal cells adjacent to the hippocampus, ependymal cells adjacent to the stem cell layer and the brain cortex showed fetuin-A immunoreactivity from days P0 to P28. Similar to human brains, fetuin-A immunoreactivity was restricted to distinct ependymal cells suggesting a specific process confined to specific ependymal cells rather than a global phenomenon [Figs. 2(D) and 4(A,B)].

image

Figure 3. Time course of fetuin-A immunoreactivity in hippocampus, white matter and cortex of human and rat brain tissues. In human (A) and rat (B) brains, strong fetuin-A immunoreactivity is found in dentate gyrus granular layers and in CA sectors. Overall, in rat brains a marked decrease of fetuin-A immunoreactivity occurred between P14 and P16, which corresponds with the age of a term-born human neonate. Whereas widespread fetuin-A staining of rat white matter is found at P0 only, prolonged fetuin-A staining is observed in cingular cells and fibres (C). From P3 to P7, the cingular tissue shows a spongy or softened appearance. Similar to white matter, wide spread fetuin-A staining of cortical neurons is present at P0 only. Thereafter it becomes restricted to the cingular interhemispheric cortex (D). Even at P0 fetuin-A staining does not affect all cortical cells but is restricted to deep cortical and subplate layers. DGGL,: dentate gyrus granular layer; DGSGL, dentate gyrus subgranular layer; DGH, dentate gyrus hilus; CA3, CA3 sector hippocampus; CA1, CA1 sector hippocampus; SVZ, subventricular zone; SUB, subiculum.

Download figure to PowerPoint

image

Figure 4. Fetuin-A immunoreactivity in rat brain tissue. Within the ependyma (A), fetuin-A is not found in all ependymal cells but is restricted to distinct cells (arrows). Frequently, prominent fetuin-A staining is observed in ependymal cells adjacent to the chorioid plexus (asterisk). Fetuin-A-positive cells (arrows) are also frequent in the periventricular stem cell layer (B; asterisk: blood vessel). Hippocampal cells showed strong fetuin-A staining from P0 to P14, being present in the outer layer of the dentate gyrus granular layer (C, arrows), subgranular layer (C, asterisk), and the CA3 sector (D). In the cortex, fetuin-A-positive cells were abundant at P0 but restricted to deep cortical layers (E). At older ages (F), cortical fetuin-A-positive cells were restricted to the interhemispheric or cingular cortex (arrows), adjacent to the central callosum (asterisk). At older ages (G), within white matter fetuin-A positive cells and fibres were restricted to the cingular white matter (cells: arrows; asterisk: blood vessels). Distinct fetuin-A-positive cells were also found in deep nuclear structures (H, arrows: cells; asterisk: blood vessels). Bar = 50 μm.

Download figure to PowerPoint

The Dentate Gyrus and CA3 Sector Are Major Sites also of Rat Hippocampal Fetuin-A Immunoreactivity

Similar to the human hippocampus, strong fetuin-A immunoreactivity was also observed in different rat hippocampal areas. Overall, fetuin-A immunoreactivity decreased between P14 to 28, which represents the human postnatal life period. As already seen in humans, fetuin-A immunoreactivity was more prominent in the dentate gyrus granular layer and CA3 sector than in the CA1 sector [Figs. 3(B) and 4(C,D)].

Fetuin-A Immunoreactivity in the Rat Dentate Gyrus Granular Layer and CA3 Sector Show Time-Dependent Quantitative and Qualitatitive Changes

Figures 5 and 6 illustrate the time course and spatial distribution of fetuin-A immunoreactivity in the rat hippocampus in a semiquantitative scheme, distinguishing between the cranial, apical, and caudal sectors of the dentate gyrus granular layer as well as between the medial, intermediate and lateral parts of the CA3 sector, respectively.

image

Figure 5. Fetuin-A immunoreactivity in the rat dentate gyrus granular layer. Two corresponding horizontal bars displayed in the left (layer 1) and right (layer 2) box, respectively, represent the two opposing dentate gyrus granular layers (DGGL) of one animal. Each granular layer is further divided into its cranial, apical/medial and caudal segments. Figure 5 therefore displays the time course of fetuin-A immunoreactivity in all DGGL of all animals studied between P0 and P28. It shows that fetuin-A immunoreactivity was strongest from P0 to P10 without being symmetrical or homogenous. From P10, fetuin-A immunoreactivity decreased, showing prolonged immunoreactivity in cranial layers first, than in apical segments.

Download figure to PowerPoint

image

Figure 6. Fetuin-A immunoreactivity in the rat CA3 sector. Two corresponding horizontal bars, displayed in the left (side 1) and right (side 2) box, respectively, represent two opposing CA3 sectors of one animal. Each CA3 sector is further divided into its medial, intermediate and lateral segments. Figure 6 therefore displays the time course of fetuin-A immunoreactivity in all CA3 sectors of all rats studied between P0 and P28. It shows that fetuin-A immunoreactivity was strong between P3 and P14 without being symmetric or homogenous. From P16, fetuin-A immunoreactivity was restricted to the medial aspects of the CA3 layers, adjacent to the dentate gyrus.

Download figure to PowerPoint

As depicted in Figure 5, fetuin-A immunoreactivity in the dentate gyrus granular layer was inhomogeneous involving the caudal, apical, and cranial parts in an isolated or combined fashion. Staining was asymmetrical as the two hippocampi of one animal differed from each other, and the staining patterns of animals derived from the same age group varied. Thus fetuin-A immunoreactivity did not follow a simple time-dependent kinetics as it was strong at P0 and between P5 and P9. Overall, fetuin-A immunoreactivity was found to decrease from P10 and dropped to background levels at P28. The difference in fetuin-A immunoreactivity between the outer and inner layers of the rat dentate gyrus granular layer is also depicted in Figure 4(C).

In the CA3 sector fetuin-A immunoreactivity was also asymmetrical, and the staining patterns of animals derived from the same age group varied, too. From P16, fetuin-A immunoreactivity decreased markedly, being more pronounced in the medial parts, adjacent to the dentate gyrus (Fig. 6). Again, not all cells of the CA3 sector were fetuin-A positive, indicating a differentiated rather than a random and nonspecific process [Fig. 4(D)]

In the Rat Neocortex and White Matter, Wide Spread Fetuin-A Immunoreactivity Is Present at P0 Only

In contrast to the hippocampus or periventricular stem cell layer, cerebral cortical cells showed strong fetuin-A immunoreactivity at P0 only. Fetuin-A immunoreactivity was restricted to deep cortical neurons and subplate neurons whereas superficial cortical neurons showed no staining [Figs. 3(D) and 4(E)]. After P0 fetuin-A staining of cortical neurons was restricted to cortical neurons located at the base of the interhemispheric cleft, adjacent to the corpus callosum [Fig. 4(F)]. These neurons contribute to the cingulate cortex.

Strong fetuin-A staining of callosal cells, in turn, was present from P5 to P14 only, and most pronounced in the medial corpus callosum [Fig. 3(D)].

Widespread fetuin-A staining of subcortical white matter tissue located between the corpus callosum and the subplate layer was restricted to P0 [Fig. 3(C)] delineating the majority of fibres and few cell bodies. After P0, fetuin-A-positive fibers and/or cell bodies were restricted to the cingular white matter [Fig. 4(G)]. Transiently, from P3 to P7, the cingular white matter showed a spongy appearance. From P16 neither cells nor fibres of the cingular white matter stained positive for fetuin-A.

Fetuin-A Immunoreactivity in Rat Basal Nuclei and Thalamus

Similar to human brain tissue, rat basal nuclei and thalamus were also found to contain fetuin-A-positive cells, peaking between P9 and P16, and frequently showing a neuronal phenotype with multiple processes [Fig. 4(H)].

Colocalization of Fetuin-A with Cell-Specific Markers in Rat Brains

To further specify the cell types showing fetuin-A immunoreactivity, we performed double antibody staining for fetuin-A and macrophages, neuronal precursors, mature neurons and astrocytes by staining with CD68, beta3-Tubulin (TUBB3), NeuN or GFAP antibodies, respectively, at P0, P3, P9, P16, and P24.

Colocalization of Fetuin-A and TUBB-Positive Cells in the Rat Brain

TUBB3 is considered to stain neuronal progenitor cells. As seen from Figure 7(A), TUBB3-positive cells were found in nearly all brain areas from P0 to P24. From P0 to P3 TUBB3 staining was chiefly restricted to cell bodies. From P9, with increased brain maturation, there was an increasing number of fibres with positive TUBB3 staining. Strong colocalization of cell bodies with TUBB3 and fetuin-A was observed from P0 to P9 only and restricted to the inner cortical layers, subplate, subventricular zone, and the hippocampus. Most TUBB3-positive cells of the hippocampal hilus, the majority of TUBB3-positive cells within the CA3 hippocampal sector, about 50% of the TUBB3-positive cells of the inner neocortex and about 25% of the TUBB3-positive cells within the subplate were also fetuin-A positive. Large numbers of TUBB3/fetuin-A-colocalizing cells were found in the inner cortical layers at P0 but rapidly disappeared until P3 [Figs. 7(B,C) and 8(F)].

image

Figure 7. Colocalization of fetuin-A and TUBB3 (beta tubulin, neuronal progenitor cells). TUBB3 immunoreactivity was found in all brain areas at all time points studied. Strong general coimmunoreactivity with fetuin-A, in contrast, was found at P0 only. Prolonged coimmunoreactivity of TUBB3 and fetuin-A, until P9 was seen in the subventricular zone and hippocampal hilus (A). In the hippocampal hilus nearly all TUBB3-positive cells were found to express fetuin-A (B). The number of colocalizing cells, however, decreased from P0 to P9 (C).

Download figure to PowerPoint

image

Figure 8. Coimmunoreactivity of fetuin-A with lineage specific brain markers. Colocalization of fetuin-A with NeuN (A – C; C = overlay) in outer layers of the dentate gyrus granular layer (arrow) and the hippocampal hilus (asterisk). Colocalization with beta tubulin (TUBB3) in cortical cell processes (D–F; F = overlay), with GFAP within the cingular white matter (G–I; I = overlay) and with CD68 within basal nuclear areas (J–M; M = overlay).

Download figure to PowerPoint

Colocalization of Fetuin-A with NeuN in the Rat Brain

Figure 9 shows that NeuN immunoreactivity in mature neurons could first be observed at P9. It was restricted to the interhemispheric cortex and hippocampus. Strong NeuN/fetuin-A colocalization was only seen in hippocampal cells from P9 to P16. The majority of NeuN-positive cells in the hilus, subgranular zone, and CA3 sector also stained positive for fetuin-A. In the rat hippocampus, the number of fetuin-A/Neu-N colocalizing cells markedly increased from P3 to P9 and decreased after P16 [Figs. 8(C) and 9(B)]. At P9 more than 70% of the NeuN-positive cells in the hippocampal subgranular zone, in the CA3 sector and in the DG hilus, as well as 25%–50% of the cells in the DG granular zone expressed fetuin-A.

image

Figure 9. Colocalization of fetuin-A and NeuN. In contrast to TUBB3, NeuN (mature neurons) immunoreactivity was only observed from P9 (A). Strong colocalization was restricted to the hippocampus and rapidly decreased over time. This is also mirrored by the number of NeuN/Fetuin-A colocalizing cells in the hippocampus (B).

Download figure to PowerPoint

Colocalization of Fetuin-A and CD68 in the Rat Brain

As outlined above, one major putative role of fetuin-A refers to an immunomodulatory function. Microglia represents the resident immune cell of the CNS. Microglia is involved in infection defence triggering inflammatory processes, thus contributing to and enhancing neuronal death. In adult and neonatal brain microglia also modulates stem cell function, promoting neurogenesis, neuronal differentiation and healing. We therefore studied activated microglia (CD68 positive) for colocalization with fetuin-A. As depicted in Figure 10, CD68-positive cells were only found in the cingulum, in the periventricular germinal layer, and in the basal nuclear areas from P3 to P9. This was associated with strong fetuin-A / CD68 colocalization; about 70% to 90% of all CD68-positive cells within the germinal layer and the cingulum also stained positive for fetuin-A [Figures 10 and 8(M)].

image

Figure 10. Colocalization of CD68 and fetuin-A. CD68-positive cells (activated macrophages, microglia) were observed from P3 to P9, always showing strong colocalization with fetuin-A.

Download figure to PowerPoint

Colocalization of GFAP and Fetuin-A in the Rat Brain

GFAP colocalization characterizes astrocytes, which represent one major glial cell type of the brain but also contribute to the adult stem cell pool located in the subventricular zone. GFAP was detected in various brain areas from P3 [Figs. 8(I) and 11]. Strongest colocalization with fetuin-A was found in the cingular white matter with stronger immunoreactivity in occipital than in rostral cingular areas. Overall, fetuin-A immunoreactivity in GFAP-positive cells was much weaker when compared to CD68+ or TUBB3+ cells.

image

Figure 11. Colocalization of GFAP with fetuin-A. The overall strength of GFAP immunoreactivity increased from P9, affecting large brain areas. Strong colocalization with fetuin-A was observed from P9 to P16 only and confined to the cingular white matter.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

We studied fetuin-A immunoreactivity in neonatal brains focusing on time-related changes of fetuin-A immunoreactivity in different brain areas and cell populations. First, we investigated the brains of human newborns, born between 24 and 41 completed weeks of gestational age, who had died from various pathological conditions, including asphyxia, septicaemia, pulmonary failure after premature birth or congenital heart disease. In the brains of these children, fetuin-A was found widely expressed, for instance being present in cells of the periventricular stem cell layer, white matter, cerebral cortex, basal ganglia and hippocampus. Although fetuin-A immunoreactivity seemed to decrease with gestational age, it was also found in cortices of term-born neonates, extending previous studies (Dziegielewska et al., 1987). On the one hand this might emphasize the importance of fetuin-A for the developing human brain. On the other hand, this long lasting immunoreactivity and the heterogeneous fetuin-A immunoreactivity observed among patients of the same age group might also have been induced by exogeneous and endogenous noxious factors, and therefore these findings may specifically represent a pathological setting.

For clarification we analyzed fetuin-A immunoreactivity in brains of healthy rats aged P0 to P28. In line with the literature, fetuin-A immunoreactivity was present in hippocampal, ventricular zone, subplate, callosal and cingular areas. We confirmed time-related changes, such as its presence in subplate and deep cortical areas of the newborn rat, and its disappearance during aging (Dziegielewska et al., 2000). Simultaneously, we developed a more detailed view on time- and space- related changes of cerebral fetuin-A distribution, showing that fetuin-A immunoreactivity decreases with age in the rat hippocampus, cortex and white matter whereas it is expressed until P28 within the ependymal layer and within the subventricular zone; anatomical regions that harbour the brains stem cell pool (Tramontin et al., 2003). This suggests that fetuin-A may be more important for maturing or developing neurons than for mature brain cells.

Different cortical areas, in turn, did not show similar temporal immunoreactivity patterns. Soon after birth, neocortical fetuin-A immunoreactivity became restricted to cells adjacent to the interhemispheric cleft, corresponding to the cingulate cortex. Simultaneously, prolonged fetuin-A staining among white matter fibres and cells became restricted to the adjacent cingular white matter. Persistent neuronal fetuin-A immunoreactivity was also observed in hippocampal neurons until P16, which corresponds to the age of term-born human neonates.

Together with the hippocampus, the cingulate cortex is a major component of the limbic system, so that limbic structures were characterized by increased and prolonged fetuin-A immunoreactivity. Within the hippocampus, fetuin-A immunoreactivity was not uniformly found in all cell types, but was restricted to distinct anatomical structures, suggesting that the function of fetuin-A is cell-specific. The hippocampal CA3 sector showed stronger fetuin-A staining than the CA1 sector (Fig. 3), with increasing age, those cells of the CA3 sector adjacent to the dentate gyrus showed stronger fetuin-A immunoreactivity as did the cranial parts and the outer layers of the dentate gyrus (Figs. 5 and 6). After P14, fetuin-A immunoreactivity sharply decreased in the CA3 sector and from P28 in the dentate gyrus. Metabolic, biochemical and or functional differences between adjacent hippocampal structures are well known. In adult rats, the CA3 sector is less vulnerable to exogeneous noxes such as hypoxia or ischemia than the CA1 sector, which is mirrored by different gene expression patterns of these hippocampal structures (Greene et al., 2009; Jackson and Foster, 2009; Zhang et al., 2011a). Like fetuin-A immunoreactivity, these differences are not static but change with time. In rats, within the first four weeks of life, the susceptibility of the CA3 and CA1 sector to hypoxemia undergoes marked changes (Towfighi et al., 1997). The susceptibility of newborn rats towards kainic acid-induced seizures in general is considerably lower than in adults (Haas et al., 2001). Whereas newborn rats (P9) show little hippocampal damage after kainic acid-induced seizures, damage occurs in the CA3 sector and subiculum at P15, and additionally in the CA1 sector at P21 (Rizzi et al., 2003). In adults increased susceptibility towards seizures may be linked to low glucose transporter 1 expression (Zhang et al., 2011a). A link between these different susceptibilies and fetuin-A expression in newborn animals remains to be proven.

Sex may be another factor influencing fetuin-A regulation. In the present study we did not detect sex-related differences (data not shown) which, however, might well be explained by the small sample sizes and therefore remains to be addressed in further studies.

The results of our double immunostaining experiments may provide further insight into the role of fetuin-A in the neonatal brain: Immediately after birth, immature neurons of many brain regions stained positive for fetuin-A including inner cortical layers, the subventricular zone, basal nuclear areas, white matter and the hippocampus (Fig. 7). In the rat CA3 sector and hippocampal hilus, fetuin-A staining affected all immature neurons, whereas only 50% of the immature subplate neurons stained fetuin-A positive. In the hippocampal hilus this pattern persisted until P9. In the hippocampus even mature neurons still stained positive for fetuin-A as observed at P9 for the hippocampal hilus and at P6 for the CA3 sector. Thereafter neuronal fetuin-A immunoreactivity ceased. These findings comply with the fact that fetuin-A antagonizes TGFbeta and BMP, factors that drive developing brain cells into a mature, resting state, for example promoting differentiation to dopaminergic or acetylcholinergic neurons, but suppress ongoing cell division or renewal of the stem cell pool (Mira et al., 2010; Schnitzler etal., 2010; Azari et al., 2011; Ordway et al., 2011; Zhang et al., a).

Fetuin-A immunoreactivity is not restricted to neuronal cells but also observed in cerebral white matter. However, diffuse fetuin-A immunoreactivity in rat white matter fibres was present at P0 only, suggesting that fetuin-A has a minor role if any in oligodendrocyte maturation. After P0 fetuin-A white matter immunoreactivity was restricted to the cingular white matter where it colocalized with the astrocyte marker GAFP as well as with CD68, characterizing macrophages and activated microglia. Colocalization of fetuin-A and CD68 was also present in the subventricular zone and in the basal nuclear areas. Intrauterine fetuin-A colocalization with macrophages has already been reported in the literature. In contrast to adults, where fetuin-A-positive monocytes are restricted to the bone marrow, in fetal sheep fetuin-A is associated with monocytes resident in lymph nodes, thymus, spleen, and liver (Dziegielewska et al., 1996). Our findings demonstrate for the first time that activated brain microglia also shows fetuin-A immunoreactivity. This finding offers an important link to the immune system, to neuroprotection as well as to inflammatory cell damage as reported by Wang et al. in their stroke model (Wang et al., 2010).

One further question refers to the mechanisms by which fetuin-A immunoreactivity in distinct cell populations is regulated. Fetuin-A synthesis in neuronal cells and especially in the choroid plexus has been previously reported by Dziegielewska and colleagues (1993) (Dziegielewska et al., 1993). This, however, could not be explored further with the methods used in this current study. In line with earlier observations by other researchers, we could not detect any positive staining in perivascular cells or cell processes around immunoreactive, plasma-filled blood vessels. This renders a direct transfer of fetuin-A across the blood-brain-barrier rather improbable. In contrast, secretion of fetuin-A into the cerebrospinal fluid by choroid plexus cells and reuptake by ependymal cells is fully supported by the findings reported here. The choroid plexus is an essential component of the blood-CSF barrier, of major impact for the synthesis of CSF and it has already been shown to transfer fetuin-A from the blood to the CSF (Liddelow et al., 2009). In this study choroid plexus cells of all ages studied showed fetuin-A immunoreactivity rendering local synthesis and/or transport of fetuin-A across the plexus highly likely. In all age groups, distinct ependymal cells stained positive for fetuin-A, some showing the morphological characteristics of tanycytes with processes protruding into the subventricular zone. Tanycytes are known to extract different substances from the CSF, including glucose, glutamate, neuropeptides, NGF, and transforming growth factors. After intracellular transport, these substances can be released from tanycyte processes to adjacent neurons, thus contributing to intercellular signaling (Ugrumov and Mitskevich, 1980; Rodriguez et al., 2005; Feng et al., 2012). The fact that not all ependymal cells stained positive for fetuin-A further supports an active and controlled process, rather than nonspecific transport of fetuin-A over the CSF brain barrier.

In conclusion, whereas overall cerebral fetuin-A immunoreactivity decreased with gestational age, cerebral fetuin-A immunoreactivity appeared highly regulated underscoring its importance in brain physiology. Transependymal fetuin-A transport, for example, was confined to distinct ependymal cells, suggesting a controlled rather than a random process. Within the parenchyma, proliferating rather than mature neuronal cells were the major sites of fetuin-A immunoreactivity which is compatible with a regulatory role for fetuin-A during brain development. Its presence in activated microglia offers a link to the immune system which may be of special importance during pathological conditions, such as infections and hypoxic neonatal brain damage. Prolonged immunoreactivity within the limbic system could be of ontogenetic origin, but might also signify the unique susceptibility of the limbic system towards noxious agents.

Future functional studies should focus on forced expression of fetuin-A in pathological conditions including neonatal hypoxic-ischemic brain damage, on downregulation of fetuin-A expression in stuctures with physiologically high fetuin-A content (limbic system), on its potential anti-inflammatory activities, for instance mediated by actions on macrophages and microglial cells, and on its ability to bind and inhibit intracellular calcium, one potent trigger of cell death.

ACKNOWLEDGMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

The authors want to thank S. Gräber, J. Wüffel, and A. Knischewski for their excellent technical assistance, and A. Goetzenich and S. Dunda for kindly providing the newborn rats.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES
  • Azari H, Osborne GW, Yasuda T, Golmohammadi MG, Rahman M, Deleyrolle LP, Esfandiari E, et al. 2011. Purification of immature neuronal cells from neural stem cell progeny. PLoS One 6:e20941.
  • Chertov O, Ermolaeva MV, Satpaev DK, Saschenko LP, Kabanova OD, Lukanidin EM, Lukjianova TI, et al. 1994. Inhibitory effect of calf fetuin on the cytotoxic activity of LAK cell-derived factors and tumor necrosis factor. Immunol Lett 42:97100.
  • Daveau M, Christian D, Julen N, Hiron M, Arnaud P, Lebreton JP. 1988. The synthesis of human alpha-2-HS glycoprotein is down-regulated by cytokines in hepatoma HepG2 cells. FEBS Lett 241:191194.
  • Dobbing J, Sands J. 1979. Comparative aspects of the brain growth spurt. Early Hum Dev 3:7983.
  • Dziegielewska K, Brown WM, Deal A, Foster KA, Fry EJ, Saunders NR. 1996. The expression of fetuin in the development and maturation of the hemopoietic and immune systems. Histochem Cell Biol 106:319330.
  • Dziegielewska KM, Andersen NA. 1998. The fetal glycoprotein, fetuin, counteracts ill-effects of the bacterial endotoxin, lipopolysaccharide, in pregnancy. Biol Neonate 74:372375.
  • Dziegielewska KM, Daikuhara Y, Ohnishi T, Waite MP, Ek J, Habgood MD, Lane MA, et al. 2000. Fetuin in the developing neocortex of the rat: Distribution and origin. J Comp Neurol 423:373388.
  • Dziegielewska KM, Matthews N, Saunders NR, Wilkinson G. 1993. Alpha 2HS-glycoprotein is expressed at high concentration in human fetal plasma and cerebrospinal fluid. Fetal Diagn Ther 8:2227.
  • Dziegielewska KM, Mollgard K, Reynolds ML, Saunders NR. 1987. A fetuin-related glycoprotein (alpha 2HS) in human embryonic and fetal development. Cell Tissue Res 248:3341.
  • Dziegielewska KM, Reader M, Matthews N, Brown WM, Mollgard K, Saunders NR. 1993. Synthesis of the foetal protein fetuin by early developing neurons in the immature neocortex. J Neurocytol 22:266272.
  • Dziegielewska KM, Williams RT, Knott GW, Kitchener PD, Monk SE, Potter A, Saunders NR. 1997. TGF-beta receptor type II and fetuin in the developing sheep neocortex. Cell Tissue Res 290:515524.
  • Feng CY, Wiggins LM, von Bartheld CS. 2012. The locus ceruleus responds to signaling molecules obtained from the CSF by transfer through tanycytes. J Neurosci 31:91479158.
  • Greene JG, Borges K, Dingledine R. 2009. Quantitative transcriptional neuroanatomy of the rat hippocampus: Evidence for wide-ranging, pathway-specific heterogeneity among three principal cell layers. Hippocampus 19:253264.
  • Haas KZ, Sperber EF, Opanashuk LA, Stanton PK, Moshe SL. 2001. Resistance of immature hippocampus to morphologic and physiologic alterations following status epilepticus or kindling. Hippocampus 11:615625.
  • Häusler M, Schafer C, Osterwinter C, Jahnen-Dechent W. 2009. The physiologic development of fetuin-A serum concentrations in children. Pediatr Res 66:660664.
  • Herrmann M, Kinkeldey A, Jahnen-Dechent W. 2012. Fetuin-A function in systemic mineral metabolism. Trends Cardiovasc Med, epub ahead of print.
  • Jackson TC, Foster TC. 2009. Regional health and function in the hippocampus: Evolutionary compromises for a critical brain region. Biosci Hypotheses 2:245251.
  • Jahnen-Dechent W, Heiss A, Schafer C, Ketteler M. 2011. Fetuin-A regulation of calcified matrix metabolism. Circ Res 108:14941509.
  • Li W, Zhu S, Li J, Huang Y, Zhou R, Fan X, Yang H, Gong X, Eissa NT, Jahnen-Dechent W, Wang P, Tracey KJ, Sama AE, Wang H. 2011. A hepatic protein,fetuin-A, occupies a protective role in lethal systemic inflammation. PLoS One 6:e16945.
  • Liddelow SA, Dziegielewska KM, Ek CJ, Johansson PA, Potter AM, Saunders NR. 2009. Cellular transfer of macromolecules across the developing choroid plexus of Monodelphis domestica. Eur J Neurosci 29:253266.
  • Mira H, Andreu Z, Suh H, Lie DC, Jessberger S, Consiglio A, San Emeterio J, et al. 2010. Signaling through BMPR-IA regulates quiescence and long-term activity of neural stem cells in the adult hippocampus. Cell Stem Cell 7:7889.
  • Mollgard K, Reynolds ML, Jacobsen M, Dziegielewska KM, Saunders NR. 1984. Differential immunocytochemical staining for fetuin and transferrin in the developing cortical plate. J Neurocytol 13:497502.
  • Nawratil P, Lenzen S, Kellermann J, Haupt H, Schinke T, Muller-Esterl W, Jahnen-Dechent W. 1996. Limited proteolysis of human alpha2-HS glycoprotein/fetuin. Evidence that a chymotryptic activity can release the connecting peptide. J Biol Chem 271:3173531741.
  • Nie Z. 1992. Fetuin: Its enigmatic property of growth promotion. Am J Physiol 263:C551C562.
  • Ohnishi T, Nakamura O, Arakaki N, Miyazaki H, Daikuhara Y. 1994. Effects of cytokines and growth factors on phosphorylated fetuin biosynthesis by adult rat hepatocytes in primary culture. Biochem Biophys Res Commun 200:855868.
  • Ordway GA, Szebeni A, Chandley MJ, Stockmeier CA, Xiang L, Newton SS, Turecki G, et al. 2012. Low gene expression of bone morphogenetic protein 7 in brainstem astrocytes in major depression. Int J Neuropsychopharmacol:15:865868.
  • Pal D, Dasgupta S, Kundu R, Maitra S, Das G, Mukhopadhyay S, Ray S, et al. 2012. Fetuin-A acts as an endogenous ligand of TLR4 to promote lipid-induced insulin resistance. Nat Med 18:1279–1285.
  • Reynolds ML, Sarantis ME, Lorscheider FL, Saunders NR. 1987. Fetuin as a marker of cortical plate cells in the fetal cow neocortex: A comparison of the distribution of fetuin, alpha 2HS-glycoprotein, alpha-fetoprotein and albumin during early development. Anat Embryol (Berl) 175:355363.
  • Rizzi M, Perego C, Aliprandi M, Richichi C, Ravizza T, Colella D, Veliskova J, et al. 2003. Glia activation and cytokine increase in rat hippocampus by kainic acid-induced status epilepticus during postnatal development. Neurobiol Dis 14:494503.
  • Rodriguez EM, Blazquez JL, Pastor FE, Pelaez B, Pena P, Peruzzo B, Amat P. 2005. Hypothalamic tanycytes: A key component of brain-endocrine interaction. Int Rev Cytol 247:89164.
  • Romijn HJ, Hofman MA, Gramsbergen A. 1991. At what age is the developing cerebral cortex of the rat comparable to that of the full-term newborn human baby?Early Hum Dev 26:6167.
  • Saunders NR, Habgood MD, Ward RA, Reynolds ML. 1992. Origin and fate of fetuin-containing neurons in the developing neocortex of the fetal sheep. Anat Embryol (Berl) 186:477486.
  • Saunders NR, Sheardown SA, Deal A, Mollgard K, Reader M, Dziegielewska KM. 1994. Expression and distribution of fetuin in the developing sheep fetus. Histochemistry 102:457475.
  • Schnitzler AC, Mellott TJ, Lopez-Coviella I, Tallini YN, Kotlikoff MI, Follettie MT, Blusztajn JK. 2010. BMP9 (bone morphogenetic protein 9) induces NGF as an autocrine/paracrine cholinergic trophic factor in developing basal forebrain neurons. J Neurosci 30:82218228.
  • Sheldon RA, Chuai J, Ferriero DM. 1996. A rat model for hypoxic-ischemic brain damage in very premature infants. Biol Neonate 69:327341.
  • Terkelsen OB, Jahnen-Dechent W, Nielsen H, Moos T, Fink E, Nawratil P, Muller-Esterl W, et al. 1998. Rat fetuin: Distribution of protein and mRNA in embryonic and neonatal rat tissues. Anat Embryol (Berl) 197:125133.
  • Towfighi J, Mauger D, Vannucci RC, Vannucci SJ. 1997. Influence of age on the cerebral lesions in an immature rat model of cerebral hypoxia-ischemia: A light microscopic study. Brain Res Dev Brain Res 100:149160.
  • Tramontin AD, Garcia-Verdugo JM, Lim DA, Alvarez-Buylla A. 2003. Postnatal development of radial glia and the ventricular zone (VZ): A continuum of the neural stem cell compartment. Cereb Cortex 13:580587.
  • Ugrumov MV, Mitskevich MS. 1980. The adsorptive and transport capacity of tanycytes during the perinatal period of the rat. Cell Tissue Res 211:493501.
  • Wang H, Li W, Zhu S, Li J, D'Amore J, Ward MF, Yang H, et al. 2010. Peripheral administration of fetuin-A attenuates early cerebral ischemic injury in rats. J Cereb Blood Flow Metab 30:493504.
  • Wang H, Sama AE. 2012. Anti-inflammatory role of fetuin-A in injury and infection. Curr Mol Med 12:625633.
  • Zhang M, Li WB, Liu YX, Liang CJ, Liu LZ, Cui X, Gong JX, et al. 2011a. High expression of GLT-1 in hippocampal CA3 and dentate gyrus subfields contributes to their inherent resistance to ischemia in rats. Neurochem Int 59:10191028.
  • Zhang X, Santuccione A, Leung C, Marino S. 2011b. Differentiation of postnatal cerebellar glial progenitors is controlled by Bmi1 through BMP pathway inhibition. Glia 59:11181131.