Oria et al. Cell Death and Disease (2019)10:721 https://doi.org/10.1038/s41419-019-1913-6 Cell Death & Disease ARTICLE Open Access Cell necrosis, intrinsic apoptosis and senescence contribute to the progression of exencephaly to anencephaly in a mice model of congenital chranioschisis Marc Oria1, Soner Duru 1, Rebeca L. Figueira1,2, Federico Scorletti1,3, Lucas E. Turner4,5, Irati Fernandez-Alonso1, Alejandra Fernandez-Martin1,6, Mario Marotta1,6, Lourenco Sbragia1,2, Aimen F. Shaaban4,5 and Jose L. Peiro1 Abstract Exencephaly/anencephaly is one of the leading causes of neonatal mortality and the most extreme open neural tube defect with no current treatments and limited mechanistic understanding. We hypothesized that exencephaly leads to a local neurodegenerative process in the brain exposed to the amniotic fluid as well as diffuse degeneration in other encephalic areas and the spinal cord. To evaluate the consequences of in utero neural tissue exposure, brain and spinal cord samples from E17 exencephalic murine fetuses (maternal intraperitoneal administration of valproic acid at E8) were analyzed and compared to controls and saline-injected shams (n = 11/group). Expression of apoptosis and senescence genes (p53, p21, p16, Rbl2, Casp3, Casp9) was determined by qRT-PCR and protein expression analyzed by western blot. Apoptosis was measured by TUNEL assay and PI/AV flow cytometry. Valproic acid at E8 induced 1234567890():,; 1234567890():,; 1234567890():,; 1234567890():,; exencephaly in 22% of fetuses. At E17 the fetuses exhibited the characteristic absence of cranial bones. The brain structures from exencephalic fetuses demonstrated a loss of layers in cortical regions and a complete loss of structural organization in the olfactory bulb, hippocampus, dental gyrus and septal cortex. E17 fetuses had reduced expression of NeuN, GFAP and Oligodendrocytes in the brain with primed microglia. Intrinsic apoptotic activation (p53, Caspase9 and 3) was upregulated and active Caspase3 localized to the layer of brain exposed to the amniotic fluid. Senescence via p21-Rbl2 was increased in the brain and in the spinal cord at the lamina I-II of the somatosensory dorsal horn. The current study characterizes CNS alterations in murine exencephaly and demonstrates that degeneration due to intrinsic apoptosis and senescence occurs in the directly exposed brain but also remotely in the spinal cord. Introduction of calvaria result in exencephaly. The unprotected fetal Exencephaly is the most severe form of all neural tube brain tissue undergoes progressive damage with advancing defects (NTD) with an incidence of 3 every 10,000 preg- gestational age due to exposure to chemical and mechanical nancies1. The neurulation process occurs during the fourth factors in the intrauterine environment2,3. In this severe week of human gestation and disrupted closure and absence form of neural tube defect the brain is totally disorganized with few defined structures4 and some functions are always present5,6. Exencephaly becomes anencephaly during Correspondence: Marc Oria (Marc.Oria@cchmc.org) 1 gestation due this aggressive environment and the brain Center for Fetal and Placental Research, Cincinnati Children’s Hospital Medical Center (CCHMC), Cincinnati, OH, USA tissue can be destroyed by 8−10 weeks of gestation7. 2 Laboratory of Experimental Fetal Surgery “Michael Harrison”, Division of Acrania-exencephaly-anencephaly sequence theory was Pediatric Surgery, Department of Surgery and Anatomy, Ribeirao Preto Medical proposed by the observation of exencephalic fetuses during School, University of Sao Paulo-USP, Ribeirao Preto, Brazil Full list of author information is available at the end of the article. pregnancy and the later delivered with anencephaly8; this Edited by A. Verkhratsky © The Author(s) 2019 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Official journal of the Cell Death Differentiation Association Oria et al. Cell Death and Disease (2019)10:721 Page 2 of 11 theory suggests a “peeling off” of the neural tissue which has using the RNeasy Plus Mini Kit (Qiagen Science, Hilden, been demonstrated in mice at E18.5 9. Germany) following the manufacturer’s protocol. One This malformation, incompatible with extrauterine life microgram of RNA was reverse transcribed using RT2 and currently untreatable, is poorly investigated and the First strand Kit (Qiagen Sciences, MD, USA) and then run degenerative mechanisms in the brain and the spinal cord on specific TaqManR Gene Expression Assays for: TPR53, are not clearly defined. CDKN2A, CDKN1A, RBL2, Caspase3 and Caspase9 The use of an exencephaly mice model, induced fol- (Applied Biosystems, Foster City, CA, USA) (Supple- lowing maternal administration of the antiepileptic drug mentary Table 1) and the 7500 Fast Real-Time PCR Valproic Acid (VPA)10,11, in the current study allows System. Samples were performed in triplicate. The relative characterization of the neural damage present in the brain expression fold change of mRNAs was calculated by the and distal spinal cord of exencephalic fetuses to elucidate 2−ΔΔCt method on eight fetuses per group (n = 8). the degenerative mechanisms involved in the pathogen- esis of exencephaly before anencephaly occurs. Protein extraction Proteins were extracted from total fetal brain and spinal Material and methods cord samples at E17 by sonication (Fisher Scientific, All experimental protocols were approved by the Pittsburgh, PA, USA) in N-PER buffer (Thermo Fisher Institutional Animal Care and Use Committee at The Scientific, Rockford, IL, USA) + Proteinase K + Phos- Children’s Hospital of Cincinnati and followed guidelines phoSTOP (Roche Diagnostics GmbH, Mannheim, Ger- set forth in the National Institutes of Health Guide for many) and centrifuged at 10,000 rpm for 5 min. Protein Care and Use of Laboratory Animals (IACUC 2013-0293). concentration was assessed using the BCA-mini kit (Thermo Scientific, Rockford, IL, USA). Animal preparation and valproic acid (VPA) exposure Timed-pregnant CD1 mice weighing 30−35 g (Charles Western blot River Laboratories, Inc, Wilmington, MA), were individually Brain and spinal cord homogenates on six animals per housed with a standard dark:light schedule (10:14; 22 °C) group (n = 6) were loaded onto 4−12% acrylamide gels and access to water and standard chow ad libitum. Valproic and transferred to PVDF membranes (Life Technologies, acid (VPA; Sigma- Aldrich Chemical, St. Louis, MO) was CA, USA). After blocking with BSA (5%) for 1 h, mem- dissolved at room temperature in sterile water and used branes were hybridized with primary antibody: anti-NeuN within 1 h of preparation. Pregnant mice were injected IP (1:1000, Abcam, Cambridge, MA, USA), anti-GFAP with 600 mg/kg VPA on E8 at 10:00 hours (n = 22), mating (1:1000, Dako, Carpinteria, CA, USA), anti- date was defined as E-1. Control (Sham) animals were oligodendrocyte (1:250, Abcam, Cambridge, MA, USA), injected with the same volume of sterile water (n = 11)12. anti-Iba1 (1:1000, Wako, Richmond, VA, USA), anti-p21 (1:500, Abcam, Cambridge, MA, USA) and anti-ß-Actin Tissue samples: brain, spinal cord (1:10,000, Abcam, Cambridge, MA, USA) overnight at 4 °C. On embryonic day 17 fetuses were collected into three Membranes were washed and hybridized with appropriate groups; exencephalic, internal control (nonexencephalic secondary antibody (1:10,000, Abcam, Cambridge, MA, littermates from VPA-treated dams), and Sham USA) for 1 h at RT and visualized using SuperSignal (untreated dams). Whole brain and spinal cord samples WestPico Chemiluminescent substrate (Thermo Scien- dissected for RNA and protein expression analysis were tific, Rockford, USA). Protein expression was quantified dissected, snap frozen and stored at −80 °C. For histology by densitometry using ImageJ (National Institutes of and immunofluorescence, fetuses were fixed for 24 h in Healt, Bethesda, MD) and normalized to ß-Actin. 4% paraformaldehyde, processed and embedded in par- affin. Samples for flow cytometry were dissected and Histological analysis processed following the standard protocol for the neural Sagittal sections of the brain and spinal cord cross tissue dissociation kit (Miltenyi Biotec, Auburn, CA, sections were fixed in formalin for 48 h and processed and USA); briefly, neural tissue was digested with papain at embedded in paraffin blocks. Histology was performed on 37 °C with rotation using a gentle MACS Dissociator 5 μm sections from sagittal brain and spinal cord cross (Miltenyi Biotec, Auburn, CA, USA). Cell suspension was sections including the periventricular region, brain stem passed through 70 µm cell strainer (BD) and labeled for and cerebellum. Hematoxylin and eosin (H&E) staining flow cytometry as described below. was performed using a Varistain Gemini ES autostainer (Thermo Scientific). A 1:4 mixture of Mayer’s Hematox- RNA extraction and RT-qPCR ylin (Lillie’s Modification) histological Staining Reagent Tissues were homogenized (IkaT10 basic Ultra-Turrax (Dako Carpinteria, CA, USA) and Automation Hema- homogenizer, USA) in RLT buffer and RNA extracted toxylin Staining Reagent (Dako Carpinteria, CA, USA) Official journal of the Cell Death Differentiation Association Oria et al. Cell Death and Disease (2019)10:721 Page 3 of 11 was used with a 0.25% Eosin-Y (Richard Allen Scientific) Results for contrast. Loss of layer organization in VPA-induced exencephalic fetuses Immunofluorescence Twenty-two percent of the fetuses developed exence- Sections (5 µm) of fetal brain and spinal cord were phaly following maternal administration of VPA. We dried, permeabilized with 0.1% Triton X-100 (Sigma observed typical alterations seen in exencephalic fetuses Aldrich, St. Louis, MO, USA) in PBS and blocked for 1 h with absent frontal, parietal and occipital bones (Fig. 1b). in 5% BSA at RT. Sections were incubated with primary At E17 exencephalic fetuses demonstrated a lack of antibody; anti-GFAP (1:1000, DAKO, Carpinteria, CA, structural organization in areas such as the olfactory bulb, USA), anti-Iba1 (1:1000, Wako, Richmond, VA, USA), hippocampus, dental gyrus, and septal complex, with anti-p21 (1:500, Abcam, Cambridge, MA, USA), anti- evidence of collapsed ventricles (Fig. 1b, f). The brain Cleaved Caspase 3 (1:500, Cell Signalling, Danvers, MA, stem was hypertrophic (Fig. 1b, f) and the cerebellum USA), overnight at 4 °C in a humid chamber. Slides were atrophied with no circumvolutions nor cortical layer washed and incubated for 1 h with Alexa Fluor 488 or 568 organization (Fig. 1d, h). Control fetuses from nonaffected labeled secondary antibodies (1:1000, Life Technologies, littermates and those from sham mothers demonstrated Eugene, OR, USA) at RT in the dark. Slides were washed normal brain development and structural organization and mounted in mounting media with DAPI (South- (Fig. 1a–g). ernBiotech, Birminghan, AL, USA). The exencephalic brain cortex showed no layered cell organization (marginal zone, cortical plate, sub plate and Neurotracer assay intermediate zone) (Fig. 1l–n) compared to littermate Neurotracer assay (Life Technologies, Eugene, OR, controls and sham (Fig. 2a–c)14. Derangement of neurons USA) was performed on 5 µm sections of fetal brain and was observed in the gray matter with neuronal cells ran- spinal cord following the manufacturers’ instructions to domly arranged from the superficial molecular layer to identify and visualize neurons. the deeper cortical layer, without laminar organization (Fig. 1i, k). TUNEL assay TUNEL assay (Roche, Diagnostics GmbH, Mannheim, Reduction of neuronal and glial protein expression in Germany) was performed following the manufacturers’ exencephalic fetuses instructions on brain and spinal cord at E17 to analyze The brains of the exencephalic fetuses showed lower apoptosis/cell death13. protein expression of NeuN, GFAP, Oligodendrocyte compared to control and sham group (p < 0.005); in Flow cytometry with Annexin V (AV) and Propidium Iodide contrast, exencephalic fetuses showed increased Iba1 (PI) protein expression in the brain compared to sham and Phosphatidyl-serine is located on the inner part of the littermate controls (Fig. 2a). Immunofluorescence for plasma membrane facing the cytosol in normal, live cells, Neurotracer in the brain corroborates these results but during apoptosis it is translocated to the cell surface (Fig. 2b–e). The exencephalic fetuses showed dispersed and can be labeled with Annexin V (AV). The vital dye neuronal distribution with no structural organization or propidium iodide (PI) was used to determine cell mem- visible cortical distribution (Fig. 2d, e) compared with brane injury in the single-cell suspension from neural control and sham fetuses at E17 in which the outline of tissue. The percentage of cells labeled as PI + AV+ (late the dentate gyrus, including the hippocampal fissure, was apoptosis), PI + AV− (necrosis), PI − AV+ (early apop- clearly identifiable (Fig. 2b) along with the cortical pro- tosis) and PI − AV− (live cells) were quantified following jections (Fig. 2c). flow on an LSR III (BD Biosciences, San Jose, CA, USA) At embryonic day 17 in normal brain development, and FlowJo (FlowJo LLC, Ashland OR, USA) analysis GFAP expression is limited to cells in the ventricular software. zone (VZ), suprafimbrial region, subpial region between the fimbria and the dentate gyrus (Fig. 2f, g)15. These Statistical analysis astrocytes (GFAP positive cells) are the migrating cell Results are expressed as means ± standard error (SE) for population that contribute to the formation of the relative gene expression (2-ΔΔCt) and means ± standard dentate granule cell layer, but not the pyramidal cell deviation (SD) for protein expression. Inter-group com- layer as Seki et al. previously described16. In exence- parisons were performed with the unpaired Student’s t phalic fetuses at E17 the astrocytes (GPAP positive test considering a P value of <0.05 significant. GraphPad cells) are expressed in the layers in contact with the Prism 7 (GraphPad Software Inc., La Jolla, CA, USA) ventricles as a potential VZ but lack a well-defined package was used for statistical calculations. pattern (Fig. 2h, i). Official journal of the Cell Death Differentiation Association Oria et al. Cell Death and Disease (2019)10:721 Page 4 of 11 Fig. 1 Brain maldevelopment in VPA-induced exencephaly. Sagittal section of normal (a) and exencephalic (b) fetal head at E17 (4× H&E); Cerebellum and brain stem in normal fetuses (c) and hypertrophic brain stem in exencephaly fetuses (d) (10×, H&E); normal structured cortex (e) and brain cortex with no structures in exencephaly (f) (20×, HE) and cerebellum with cortical layers and circumbulutions in normal fetus (g) and nonstructured mass in exencephalic fetuses at E17 (h) (20×, HE). C cortex, CB cerebellum, BS brain stem. Normal cortical layer organization in Sham (i–k) and loss of any cellular layer organization in exencephaly (l–n) (10×, 20× and 40×, HE). MZ marginal zone, CP cortical plate, SP sub plate, IZ intermediate zone In control littermates and sham fetuses, microglial the exencephalic fetuses there was an increase in numbers expression of Iba1 is limited to the brain cortex at E17 of microglia demonstrating activated morphology (Fig. 2l, m) (Fig. 2j) and Iba1-positive cells showed a resting pheno- and these were mainly expressed in the layers exposed to type, with extended projections (Fig. 2k). In contrast, in amniotic fluid. Official journal of the Cell Death Differentiation Association Oria et al. Cell Death and Disease (2019)10:721 Page 5 of 11 Fig. 2 (See legend on next page.) Official journal of the Cell Death Differentiation Association Oria et al. Cell Death and Disease (2019)10:721 Page 6 of 11 (see figure on previous page) Fig. 2 Downregulation of neurons, astrocytes, oligodendrocytes and upregulation of microglia in the exencephalic brains. Western blot analysis showed a clear reduction in protein expression levels in the brain of a neurons (NeuN), astrocytes (GFAP) and oligodendrocytes. Also, an increase of the protein expression of microglia (Iba1) in brain lysates from exencephaly vs. normal and vs. Sham group (*p < 0.05) (n = 6−8). Neuronal marker staining (Neurotracer) showed normal structured brain (b) with normal layer organization in the cortex (c) at E17 in normal fetuses at E17 compared with no neuronal organization in exencephalic fetuses (d, e). GFAP+ cells expressed in the ventricular zone, dentate gyrus and fimbria in normal brain at E17 (f, g). GFAP+ cells are restricted around the ventricle in exencephalic mice (h, i). Neurotracer (green), GFAP (red) and DAPI (blue) (20× and 40×). Microgliosis at the external layers of the brain exposed to the amniotic fluid with activated morphology in exencephalic brain with activated morphology (l, m) compared with less expression in normal fetuses with resting phenotype (j, k). Iba1 staining (green) (20× and 100×) (n = 6–8) Detection of necrosis, apoptosis and senescence in VPA- Discussion induced exencephalic fetuses In this study, we described the activated mechanisms of Following VPA exposure and nonclosure of cranial the central nervous system in the valproic acid-induced bones, neural tissue remains exposed to the amniotic murine model of exencephaly. We analyzed changes on fluid. Exencephalic fetuses at E17 exhibited necrosis/cell day 17 of gestation in the brain and the spinal cord to death in the cell layer exposed to amniotic fluid (Fig. 3c, d). characterize the effect of direct exposure and indirect In controls, no cells in the inner brain and minimal exposure to amniotic fluid. Focusing on day 17 we char- cells in the outer layers stained positive with TUNEL acterized the alterations in the CNS prior to total (Fig. 3a, b) destruction of the brain tissue (anencephaly) to under- The percentage of necrotic and late apoptotic cells, stand the mechanisms involved in the pathophysiology of (AV−/PI+ and AV+/PI+) were significantly increased in this model9. This point in gestation is crucial to under- exencephalic brains (Fig. 3f, j) compared to control and stand the neurodegenerative and maldevelopment sham fetuses (Fig. 3f, i) (*p < 0.05). No changes were seen mechanisms of the brain exposed to amniotic fluid as well in the number of live, AV−PI−, and early apoptosis, as the indirect effect on neurodegeneration in other fur- AV+/PI events (Fig. 3e). Furthermore, no differences were ther areas of the CNS such as the spinal cord, which is not seen in the cell populations of the spinal cord (Fig. 3g, h). exposed to the chemical and mechanical actions of the amniotic fluid. Similar effects occur in other open neural Overexpression of intrinsic apoptosis pathway in the brain tube defects like open spina bifida in which the spinal and spinal cord of VPA-induced exencephalic fetuses cord exposure to amniotic fluid induced progressive Exencephalic fetuses had increased brain expression of neurodegeneration with neuroinflammatory response3. the intrinsic apoptosis genes p53, Caspase 9 and 3 but VPA causes lack of skull protection to the developing only showed increased Caspase 3 expression in the brain, which generates severe disruption to the laminar spinal cord (*p < 0.05, Fig. 4a). Activated Caspase 3 was architecture of the cerebral cortex17. We found no expressed in the layers of the neural tissue exposed to structural tissue organization in exencephalic fetuses, and amniotic fluid in exencephalic fetuses (Fig. 4d) and in the brain protrudes like a mass of tissue when exposed to the nonexposed spinal cord (Fig. 4e) when compared the aggressive AF environment. Neurons are not orga- with normal and sham neural tissue (Fig. 4b, c). nized in the expected layers and axonal projections are not identifiable. There is variability among fetuses in the Local and diffuse overexpression of cdkn1a (p21)- shape of the exposed brain and the degree of neural tissue Retinoblastoma (Rbl2) in the brain and spinal cord of VPA- loss. Furthermore, architecture also is disrupted in regions induced exencephalic fetuses with reduced cellular density. Exencephalic fetuses also exhibited increased gene We found a reduction in the total protein expression of expression of senescence-related genes p21, p16 and Rbl2 neurons, oligodendrocytes and astrocytes that corrobo- in the brain and p21 and Rbl2 in the spinal cord compared rate with previous studies using the same teratogen18. In to the control and sham group (*p < 0.05, n = 6, Fig. 5a). our model, GFAP expression in migrating astrocytes is This increase was reflected at the protein level (Fig. 5h) restricted to the area around the ventricles in the exen- and localized by immunofluorescence to scattered neu- cephalic fetuses. GFAP expression in the developing rons in the brain (Fig. 5e). P21-positive neurons were also mouse brain follows the basic progression of developing identified in the lamina I and II in the posterior horn of radial glia and astrocytes. GFAP expression can be the spinal cord (Fig. 5f, g). No p21 expression was found detected earlier in normal mice with the radial glia genesis in the brain (Fig. 5b) nor the spinal cord of control and around the telencephalic VZ19 and then in the fim- sham fetuses (Fig. 5c, d). bria20,21. At E17 in normal mice we observed astrogenesis Official journal of the Cell Death Differentiation Association Oria et al. Cell Death and Disease (2019)10:721 Page 7 of 11 Fig. 3 Increased necrosis/apoptosis in exencephaly brains VPA-induced mice. Brain sections of control fetuses at E17 (a, b) with minimal TUNEL-positive cells (green) and exencephalic brains (c, d) with TUNEL-positive cells in exposed brain to the amniotic fluid (a and c 40×) (b and d 40×). The apoptosis/necrosis analysis by flow cytometry with PI and Annexin V in the brain (e, f) and spinal spinal cord (g, h), % population mean (±SD) (n = 6). Exencephalic brains show higher % of cells PI + AV+ (necrosis) and PI − AV+ (late apoptosis) (f, j) (*p < 0.05, **p < 0.001) than control and sham fetuses (f, i). No differences seen in the spinal cord (g, h) (n = 6–8) Official journal of the Cell Death Differentiation Association Oria et al. Cell Death and Disease (2019)10:721 Page 8 of 11 Fig. 4 Exencephaly activates apoptotic genes in the brain and spinal cord. Relative gene expression for intrinsic apoptotic genes (p53, Caspase3 and Caspase9) in brain and spinal cord (a). Exencephalic fetuses at E17 showed upregulation of p53, Caspase3 and Caspase9 in the brain for and Caspase3 in the spinal cord vs. Sham and Control group (*p < 0.05, **p < 0.001). Caspase 3-positive cells (green) in the region exposed to the amniotic fluid in exencephalic brains (d) and scattered staining of active Caspase 3 in the spinal cord (e). No Caspase3 is seen in the brains nor spinal cords from normal fetuses at E17 (b, c) (n = 6–8) which was not seen in the exencephalic fetuses. Exence- Using flow cytometry, we evaluated necrosis and phalic mice at E17 resembled normal mice at E9.5−E11 apoptosis in the neural tissue of the brain directly exposed with expression of only the first radial glia22. to the amniotic fluid and indirectly in the spinal cord. Official journal of the Cell Death Differentiation Association Oria et al. Cell Death and Disease (2019)10:721 Page 9 of 11 Fig. 5 Senescence via p21 in VPA-induced exencephalic fetuses at E17. The relative expression of senescence genes (p21, p16 and Rbl2) in brain and spinal cord (a) of exencephalic fetuses at E17 showed upregulation of p21-p16-Rbl2 senescence pathway in the brain and p21-Rbl2 in the spinal cord. Exencephaly vs. Sham and Control groups (*p < 0.05). p21-positive cells (green) in the brain are scattered in clustered cell groups in exencephalic brains (e) and expressed in the lamina I and II of the spinal cord (f, g). No expression of p21 was seen in normal brains nor spinal cords at E17 (b–d). Relative protein expression of p21 by western blot in brain and spinal cord tissue from normal and exencephalic fetuses at E17 of gestation (h) (*p < 0.05) (n = 6–8) Official journal of the Cell Death Differentiation Association Oria et al. Cell Death and Disease (2019)10:721 Page 10 of 11 Analyzing the early, late apoptotic and necrotic cells from Conclusion the CNS tissue by flow cytometry and the histological Exencephaly is one of the leading causes of fetal and images demonstrate the effect of the in utero environment neonatal mortality as the most extreme open neural tube to the exposed cell layers2,23. This exponential chemical defect. There are currently no treatments, and the etiology action due to enzymes present in the amniotic fluid and progression of this lethal malformation are poorly increases during late gestation24 and physical action by investigated. The current study analyzing CNS disrup- friction with the amniotic membranes2 could be the tions in exencephalic fetal mice demonstrates neurode- mechanism inducing apoptosis in the external layers of generation due to extrinsic local cell death, but also by the brain but not in the protected areas, such as the spinal intrinsic apoptosis and senescence in both the brain and cord. In those areas directly exposed to the amniotic fluid, spinal cord. This local and diffuse cell degeneration may we observed a neuroinflammatory response involving contribute to the devastating neurological alterations seen microglia activation. While microglia are the resident at birth. macrophage population of the CNS, their function is more than being the first immune sentinel cell, they also Acknowledgements contribute to CNS homeostasis and cell proliferation. We greatly appreciate Dr. Helen N. Jones for invaluable help and advice during the manuscript editing and Holly Poling for her help with sample preparation. However, over reactive microglia leads to neural degen- This work was supported by Prof. Jose L. Peiro internal Cincinnati Children’s eration in neural tube defect injured tissues3. Hospital funding. The continued cell death in the brain layers exposed to amniotic fluid lead to an indirect disruption to the spinal Author details 1 cord. These alterations were detected by upregulation of Center for Fetal and Placental Research, Cincinnati Children’s Hospital Medical Center (CCHMC), Cincinnati, OH, USA. 2Laboratory of Experimental Fetal the intrinsic apoptotic genes p53 and Caspase 9 and 3 Surgery “Michael Harrison”, Division of Pediatric Surgery, Department of genes in the exposed brain tissue but also in the spinal Surgery and Anatomy, Ribeirao Preto Medical School, University of Sao Paulo- cord of the fetuses with exencephaly. Apoptotic cells were USP, Ribeirao Preto, Brazil. 3Department of Pediatric Surgery, Hospital Bambino Gesu, Rome, Italy. 4The Chicago Institute for Fetal Health, Ann & Robert H. Lurie detected in the external layers of the brain exposed to the Children’s Hospital of Chicago, Chicago, IL, USA. 5Department of Pediatric amniotic fluid. This apoptosis could underlie the func- Surgery, Northwestern University, Feinberg School of Medicine, Chicago, IL, tional locomotor alterations with abnormal motor beha- USA. 6Bioengineering, Cell Therapy and Surgery in Congenital Malformations Laboratory, Vall d’Hebron Research Institute (VHIR), Universitat Autònoma de vior that anencephalic fetuses show after birth25,26. Also, Barcelona, Barcelona, Spain the activation of these apoptotic markers in the spinal cord could determine diffuse progression of the cortico- Authors’ contributions motor deficits and the loss of reflex5. Study concept and design: M.O., S.D., J.L.P.; Acquisition of data: M.O., R.L.F., A.F.-M., Both growth arrest and apoptosis are associated with L.S., F.S., M.M., L.E.T., I.F.-A.; Analysis and interpretation of data: M.O., J.L.P.; Drafting of the manuscript: M.O., A.F. S., J.L.P.; Critical revision of the manuscript the induction of p53, particularly in response to DNA for important intellectual content: A.F.S., M.M., L.S., J.L.P; Statistical analysis: M.O.; damage or other modes of cellular stress27. We observed Obtained funding: J.L.P.; Technical or material support: M.O., S.D., A.F.-M., I.F.-A., upregulation of CDKN1A (p21) and retinoblastomas R.L.F., L.S., F.S., M.M., L.E.T.; Study supervision: J.L.P. protein (Rbl2) in the exencephalic mice in the brain; however, there was no specific pattern of cells expressing Data availability The authors declare that the data supporting the findings of this study are p21 in the brain which may be due to the lack of available within the article and its supplementary information. cytoarchitecture and structural organization. It is possible these p21-positive neurons are from the hippocampus as Conflict of interest has previously been described28. The authors declare that they have no conflict of interest. Interestingly, p21-positive cells were found in the lamina I and II at the Subtantia Gelatinosa of the dorsal Ethics approval and consent to participate horn in the absence of activation of apoptotic genes (p53, All experimental protocols were approved by the Institutional Animal Care and Use Committee at The Children’s Hospital of Cincinnati and followed Caspase 3 or Caspase 9) in that region. These results guidelines set forth in the National Institutes of Health Guide for Care and Use suggest that the somatotopically areas are in a senescent of Laboratory Animals (IACUC 2013-0293). state probably due to the loss of the somatosensory cir- cuits in the brain. The loss of the neurons in the brain, Consent for publication including the somatosensory brain areas, and the altera- Authors read and approved the final manuscript tion of motor responses by the progressive and continued abrasion of the brain layers by the amniotic fluid plus the Publisher’s note internal disruption affect the sensory neurons in the Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. dorsal horn29. Further experiments with neuronal tracers are needed to reveal if the p21-positive cells in the brain Supplementary Information accompanies this paper at (https://doi.org/ and the spinal cord are part of the same neuronal circuits. 10.1038/s41419-019-1913-6). Official journal of the Cell Death Differentiation Association Oria et al. Cell Death and Disease (2019)10:721 Page 11 of 11 Received: 22 April 2019 Revised: 8 July 2019 Accepted: 26 August 2019 mice (Mus musculus): an immunohistochemical analysis. Pak. J. Biol. Sci. 33, 86–9 (2013). 15. Mamber, C. et al. GFAPd expression in glia of the developmental and ado- lescent mouse brain. PLoS ONE 7, e52659 (2012). 16. Seki, T. et al. Distinctive population of Gfap-expressing neural progenitors References arising around the dentate notch migrate and form the granule cell layer in 1. Papp, Z., Csécsei, K., Tóth, Z., Polgár, K. & Szeifert, G. T. Exencephaly in human the developing hippocampus. J. Comp. Neurol. 522, 261–83 (2014). fetuses. Clin. Genet. 30, 440–4 (1986). 17. Copp, A. J. Neurulation in the cranial region—normal and abnormal. J. Anat. 2. Copp, A. J. & Greene, N. D. Neural tube defects–disorders of neurulation and 207, 623–635 (2005). related embryonic processes. Wiley Interdiscip. Rev. Dev. Biol. 2, 213–27 (2013). 18. Semmler, A. et al. Intrauterine valproate exposure is associated with alterations 3. Oria, M. et al. CD200-CD200R imbalance correlates with microglia and pro- in hippocampal cell numbers and folate metabolism in a rat model of inflammatory activation in rat spinal cords exposed to amniotic fluid in reti- valproate teratogenicity. Seizure 46, 7–12 (2017). noic acid-induced spina bifida. Sci. Rep. 8, 10638 (2018). 19. Fox, I. et al. Developmental expression of glial fibrillary acidic protein mRNA in 4. Trenouth, M. J. Craniofacial shape in the anencephalic human fetus. J. Anat. mouse forebrain germinal zones implications for stem cell biology. Dev. Brain 165, 215–224 (1989). Res. 153, 121–125 (2004). 5. Funayama, C. A. et al. Three-year-old child with meroacrania—neurological 20. Woodhams, P. L., Basco, E., Hajos, F., Csillag, A. & Balazs, R. Radial glia in the signs. Brain Dev. 33, 86–9 (2011). developing mouse cerebral cortex and hippocampus. Anat. Embryol. 163, 6. Melnick, M. & Myrianthopoulos, N. C. Studies in neural tube defects. II. 331–343 (1981). Pathologic findings in a prospectively collected series of anencephalics. Am. J. 21. Brenner, M., Kisseberth, W. C., Su, Y., Besnard, F. & Messing, A. GFAP promoter Med. Genet. 26, 797–810 (1987). directs astrocyte-specific expression in transgenic mice. J. Neurosci. 14, 7. Wood, L. R. & Smith, M. T. Generation of anencephaly: 1. Aberrant neurulation 1030–1037 (1994). and 2. Conversion of exencephaly to anencephaly. J. Neuropath. Exp. Neurol. 22. Gotz, M. & Barde, Y. A. Radial glial cells defined and major intermediates 43, 620–33 (1984). between embryonic stem cells and CNS neurons. Neuron 46, 369–372 (2005). 8. Timor-Tritsch, I. E., Greenebaum, E., Monteagudo, A. & Baxi, L. Exencephaly- 23. Copp, A. J., Stanier, P. & Greene, N. D. Neural tube defects: recent advances, anencephaly sequence: proof by ultrasound imaging and amniotic fluid unsolved questions, and controversies. Lancet Neurol. 12, 799–810 (2013). cytology. J. Matern. Fetal Med. 5, 182–5 (1996). 24. Olguner, M. et al. Urinary trypsin inhibitor present in fetal urine prevents 9. Matsumoto, A., Hatta, T., Moriyama, K. & Otani, H. Sequential observations of intraamniotic meconium-induced intestinal damage in gastroschisis. J. Pediatr. exencephaly and subsequent morphological changes by mouse exo utero Surg. 41, 1407–12 (2006). development system: analysis of the mechanism of transformation from 25. Andonotopo, W., Kurjak, A. & Kosuta, M. I. Behavior of an anencephalic exencephaly to anencephaly. Anat. Embryol. 205, 7–18 (2002). fetus studied by 4D sonography. J. Matern. Fetal Neonatal Med. 17, 165–8 10. Lammer, E. J., Sever, L. E. & Oakley, G. P. Teratogen update: valproic acid. (2005). Teratology 35, 465–473 (1987). 26. Visser, G. H., Laurini, R. N., de Vries, J. I., Bekedam, D. J. & Prechtl, H. F. Abnormal 11. Ornoy, A. Valproic acid in pregnancy: how much are we endangering the motor behaviour in anencephalic fetuses. Early Hum. Dev. 12, 173–82 (1985). embryo and fetus? Reprod. Toxicol. 28, 1–10 (2009). 27. Culmsee, C. & Mattson, M. P. p53 in neuronal apoptosis. Biochem. Biophys. Res. 12. Greene, N. D. & Copp, A. J. Mouse models of neural tube defects: investigating Commun. 331, 761–77 (2005). preventive mechanisms. Am. J. Med. Genet. C Semin. Med. Genet. 15, 31–41 (2005). 28. Aizawa, S. & Yamamuro, Y. Valproate administration to mice increases hip- 13. Gavrieli, Y., Sherman, Y. & Ben-Sasson, S. A. Identification of programmed cell pocampal p21 expression by altering genomic DNA methylation. Neuroreport death in situ via specific labeling of nuclear DNA fragmentation. J. Cell Biol. 26, 915–20 (2015). 119, 493–501 (1992). 29. Warsof, S. L., Abramowicz, J. S., Sayegh, S. K. & Levy, D. L. Lower limb move- 14. Irnidayanti, Y. Existence of vimentin and GFAP protein expressions as a result ments and urologic function in fetuses with neural tube and other central of 2-Methoxyethanol administration in cerebral cortex tissue of Swiss Webste nervous system defects. Fetal Ther. 3, 129–34 (1988). Official journal of the Cell Death Differentiation Association