QUICK SEARCH:   [advanced] Author: Keyword(s):
Home     Help     Feedback     Subscriptions     Archive     Search     Table of Contents

First published online 2 December 2004
doi: 10.1242/dev.01551

This Article Summary Full Text Full Text (PDF) Supplementary Material Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by McCarty, J. H. Articles by Hynes, R. O. Search for Related Content PubMed PubMed Citation Articles by McCarty, J. H. Articles by Hynes, R. O. Social Bookmarking                         What's this?

Selective ablation of v integrins in the central nervous system leads to cerebral hemorrhage, seizures, axonal degeneration and premature death

Joseph H. McCarty^1,*, Adam Lacy-Hulbert^1,2,3, Alain Charest¹, Roderick T. Bronson³, Denise Crowley^1,*, David Housman¹, John Savill², Jürgen Roes⁴ and Richard O. Hynes^1,*,

¹ Center for Cancer Research, Massachusetts Institute of Technology, 40 Ames Street, E17-227, Cambridge, MA 02139, USA
² Medical Research Council and University of Edinburgh Center for Inflammation Research, Teviot Place, Edinburgh EH8 9AG, UK
³ Harvard Medical School, 220 Longwood Avenue, Boston, MA 02115, USA
⁴ Windeyer Institute for Medical Sciences, University College London, 46 Cleveland Street, London W1T 4JF, UK

View larger version (83K): [in a new window]   Fig. 1. Tie2-Cre conditional v mutants do not develop cerebral vascular defects. (A-F) v integrin is not detected on cerebral endothelial cells or pericytes. Horizontal sections through the diencephalon of an E13.5 v+/- embryo labeled with pre-absorbed anti-v antiserum (A,D), anti-PECAM (B), or anti-smooth muscle -actin (SMA) (E). v protein localizes to neuroepithelial processes (A,D). (C,F) Merged images: arrows indicate close juxtaposition between v-positive neuroepithelial processes with endothelial cells (C) and pericytes (F). Boxed areas are shown as higher magnification insets. (G) Generation of the conditional v mutant allele in endothelial cells. Tie2-Cre-mediated recombination deletes exon four, resulting in a conditional v-null allele (lower panel). Arrows indicate PCR primers for genotyping. (H) Cre-mediated recombination monitored by PCR using DNA isolated from PECAM/Flk1-positive brain endothelial cells (EC), or PECAM/Flk1-negative cells (C). In brain endothelial cells, there is a reduction in the intensity of the 350 bp band owing to recombination of the v-flox allele. We confirmed deletion of exon four using a primer pair that detects the deleted v-flox cassette (data not shown). (I-L) Brains dissected from P5 neonates. No grossly obvious cerebral vascular defects are present in control (I) or mutant (K). Hematoxylin and Eosin-stained coronal sections from control (J) and mutant (L) brains revealed no obvious microscopic neural or vascular defects.

View larger version (60K): [in a new window]   Fig. 2. vß8 integrin is expressed on glial cell processes in the embryonic brain. Horizontal sections through the diencephalon of E13.5 v+/- (A-C) or v-/- embryos (D-F) labeled with pre-absorbed specific anti-v (A,D) and anti-nestin antibodies (B,E). (C,F) Merged images. v integrin and nestin expression co-localize on neuroepithelial cell processes in v+/- embryonic brain sections (C). No v immunoreactivity is present on neuroepithelial cell processes from v-/- brain sections (D,F). Horizontal sections through the diencephalon of E11.5 v+/- (G-I) or v-/- embryos (J-L) labeled with anti-ß8 polyclonal antibody (G,J) and anti-nestin antibodies (H,K). (I,L) Merged images. ß8-integrin and nestin expression co-localize on neuroepithelial cell processes (arrows) in v+/- embryonic brain sections (I). ß8 immunoreactivity is not detectable on neuroepithelial cell processes of v-/- brain sections, but localizes to neuroepithelial cell bodies (arrowheads; J,L). (M,N) Specificity of the affinity-purified anti-ß8 polyclonal antibody is shown using protein lysates prepared from untransfected COS cells (M, lane 1), or from COS cells transfected with cDNA encoding ß8 (M, lane 2). (N) The anti-ß8 antibody immunoprecipitates cell-surface-labeled vß8 from wild-type cultured astrocytes (lane 1), but not from v-/- astrocytes (lane 2).

View larger version (103K): [in a new window]   Fig. 3. Conditional deletion of v integrin in neural cells leads to cerebral hemorrhage. (A-F) The hGFAP-Cre transgene is expressed in central nervous system radial glial cells and post-natal astrocytes. (A,B) Sagittal sections through E15 embryonic neocortex immunostained with anti-Cre (green) and anti-brain lipid-binding protein (BLBP) (red, A) or anti-glutamate transporter (GLAST) (red, B). Expression of Cre is seen in radial glial cells within the embryonic neocortex (arrows). Regions highlighted within the dashed boxes are shown as higher magnification insets. (C) Cre is not expressed by embryonic neurons. E15 sagittal brain sections were immunostained with anti-Cre (green) and anti-ß-tubulin III (red). Cre expression is absent in ß-tubulin III-positive neurons (arrowheads in C). v, ventricle. (D,E) Postnatal astrocytes in the cerebellum (D) and cerebral cortex (E) express Cre. Sagittal sections from postnatal (P7) brains immunostained with anti-Cre (green) and anti-GFAP (red). There is co-localization of GFAP and Cre in cortical astrocytes in the cerebral cortex (arrows in E) as well as in Bergmann glia of the cerebellum (arrows in D). Boxed areas are shown as higher magnification insets. (F) Cultured astrocytes from hGFAP-Cre transgenic mice show mosaic Cre expression. There is co-localization of Cre and GFAP in most cells (arrows); however, some GFAP-positive cells show no detectable Cre protein expression (arrowhead). (G) Strategy for generating the conditional v mutant allele in central nervous system neural cells, particularly glia. (H) hGFAP-Cre-mediated recombination was monitored by PCR using DNA isolated from tails (T) or from mutant cultured astrocytes (A). DNA isolated from astrocytes shows reduced intensity of the 350 bp band, owing to recombination of the v-flox allele. We confirmed deletion of exon four using a primer pair that detects the deleted v-flox cassette (data not shown). (I) Cortical astrocyte lysates were immunoblotted with anti-v antibody. There is a significant reduction in v protein expression in mutants. (J,K) Brains dissected from control (J) and mutant (K) P5 neonates. Arrows in K indicate microhemorrhage in the mutant cerebral cortex. (L,M) Coronal sections from control (L) and mutant (M) mice indicate focal regions of hemorrhage in mutant cerebral cortex (arrows in M). Boxed area shown at higher magnification (inset in M). (N-Q) Gross analysis of adult brains from control and mutant mice. Adult control (N) or mutant (O) brains do not display overt signs of cerebral microhemorrhage. Hematoxylin and Eosin stained coronal sections from adult control (P) and mutant (Q) cerebral cortex reveal normal cytoarchitecture and no microscopic hemorrhage.

View larger version (86K): [in a new window]   Fig. 4. Conditional deletion of v integrin in both CNS neurons and glia leads to cerebral hemorrhage. (A) Generation of the conditional v mutant allele in CNS glia and neurons. (B) PCR performed using DNA from mutant tail (T) or brain (Br) tissue. In a sample from the brain, the intensity of the 350 bp band is significantly reduced, owing to Cre-mediated recombination of the v-flox allele. We confirmed deletion of exon four using a primer pair that detects the deleted v-flox cassette (data not shown). (C) Lysates from the cerebellum and cortex were immunoblotted with anti-v antibody. There is marked reduction in v protein levels in the mutant brains, indicating Cre-mediated deletion of v gene expression. (D-G,I) Control or mutant brains dissected from E17.5 embryos (D,E) and P7 neonates (F,G,I). Unlike the control brains (D,F), there are obvious regions of cerebral hemorrhage present throughout the forebrain and midbrain regions of the mutant brains (arrows in E,G,I). Significantly more severe cerebral hemorrhage is seen in the complete v-nulls (H). (J,K) Horizontal semi-thin sections from the ganglionic eminence region of E14.5 control (J) and mutant (K) brains. Vessels in the control brain are closely juxtaposed to the surrounding parenchyma (arrows in J). There are distended and tortuous vessels that separate from the surrounding neural tissue in the v conditional mutant brain (arrows in K). vz, ventricular zone. (L,M) Horizontal sections from the ganglionic eminence of E14.5 control (L) and mutant (M) embryonic brains stained with anti-PECAM (green) and anti-NG2 (red) antibodies to reveal endothelial cells and pericytes/vSMCs, respectively. The mutant vessels lined with vascular endothelium have a distended appearance and there is surrounding coverage by pericytes (arrows in M). (N-Q) Coronal sections through the cerebral cortex of control (N) or v mutant (O) P7 brains stained with H&E; normal neuronal patterning is observed in both. However, focal regions of hemorrhage are associated with many blood vessels in the mutant cerebral cortex (arrows in O). Coronal sections were stained with an anti-GFAP antibody to visualize astrocytes in control (P) and mutant (Q) cerebral cortex. Elevated numbers of astrocytes are present throughout the cortex of v mutant brain (arrows in Q).

View larger version (57K): [in a new window]   Fig. 5. Nestin-Cre conditional v mutants develop seizures and motor dysfunction, and die prematurely. (A,B) Motor dysfunction in 3-week-old Nestin-Cre conditional v mutants. When lifted by their tails, control mice (A) extend and flail their limbs. Conditional v mutants (B) retract their limbs and remain immobile. The conditional v mutant in B also displayed episodic signs of seizures and associated temporary loss of consciousness (data not shown). (C,D) Hematoxylin and Eosin-stained coronal sections from cerebral cortices of three-week-old control (C) and mutant (D) brains from mice shown in A,B. Regions of cerebral microhemorrhage are present in the mutant brains (arrow in D). (E,F) Cross-sections from control (E) and mutant (F) spinal cords. Focal regions of microhemorrhage are found throughout spinal cord of conditional v mutants (arrow in F). (G,H) Pictures of 7-month-old control (G) and mutant (H) animals. The mutant displays hind-limb spasticity and abnormal posture. (I) Footprint analysis performed using 4- to 5-month-old control (left panel) and mutant (right panel) mice; hind paws painted blue and fore paws painted red. The conditional mutant drags its hindlimb (right). (J) Rotarod analyses of control and Nestin-Cre conditional v mutants as described in the Materials and methods. The times animals remained on the rod rotating at an increasing speed versus the trial number is plotted. A significant reduction (P<0.005 for all trials) in time spent on the rod is observed in mutant (n=7) versus control mice (n=9).

View larger version (102K): [in a new window]   Fig. 6. Adult Nestin-Cre conditional v mutants develop axonal degeneration in the spinal cord and cerebellum. (A,B) Cross-sections through the thoracic region of spinal cords from six-month-old control (A) and mutant (B) mice stained to visualize myelin. The axonal tracts of the white matter (wm) stain intensely blue when compared with the grey matter (gm). (C,D) Higher magnification images of grey matter (dashed boxes in A,B) from control (C) or mutant (D) spinal cords. Motoneurons in both controls and mutants are myelinated and show normal cytoarchitecture (arrows in C,D). (E,F) Higher magnification of fasciculus gracilis white matter region in control (E) and mutant (F) spinal cords (solid boxes in A,B). The myelin pattern is disorganized in the mutant, with obvious regions of macrophage infiltration (arrows in F). (G,H) Caudal cerebellar regions from 6-month-old control (G) and mutant (H) mice were stained to visualize myelin. Representative cerebellar folia are shown, with the cerebellar white matter (wm) surrounded by the granule cell layers (gcl) and molecular layers (ml). Overall cerebellum size and patterning in the control (G) and mutant (H) appear normal. (I,J) Higher magnification images from boxed regions in G,H, respectively. The mutant white matter is pale and has an abnormal pattern of myelination (H,J) when compared with the control (G,I). Macrophage infiltration is also obvious in the mutant white matter, with many macrophages containing Luxol Fast Blue-positive myelin fragments (arrows in J). (K,L) Cerebellar coronal sections from 6-month-old control (K) and mutant (L) mice silver-stained to visualize axonal neurofilaments. When compared with the control (K), there is significant axonal degeneration in the mutant cerebellum (L). Arrowheads in L indicate dystrophic axons. Also obvious in the mutant are macrophages containing silver-positive neurofilament fragments, which are indicative of axonopathy (arrows).

View larger version (144K): [in a new window]   Fig. 7. vß8 integrin is expressed on white matter axons in the cerebellum. Coronal sections through the cerebellum of a five-month-old Nestin-Cre; vflox/+ mouse (A-I) labeled with anti-ß8 antiserum (A,D,G) and anti-neurofilament antibody to visualize axons (B), anti-GFAP antibody to visualize astrocytes (E), or anti-myelin antibody to visualize oligodendrocytes (H). (E,F,I) Merged images. vß8 integrin and neurofilament co-localize on white matter axons (C). In white matter regions of the cerebellum, vß8 is not detected on GFAP-positive astrocytes (F) or myelinating oligodendrocytes (I).

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

Twitter