obophenotype / uberon

An ontology of gross anatomy covering metazoa. Works in concert with https://github.com/obophenotype/cell-ontology
http://obophenotype.github.io/uberon/
Other
131 stars 29 forks source link

What is the relationship between the meninges and the brain? #1261

Open cmungall opened 8 years ago

cmungall commented 8 years ago

It seems that the most uncontroversial way to represent this is to have the meningeal cluster be surrounding yet spatially disjoint from the the brain itself.

(as an aside, this resolves an awkward modeling decision to place periosteal dura mater as outside the meningeal dura mater, to avoid making the brain and the skeletal system overlap)

we should first review some existing resources and see what statements they make

MP

MP (cc @sbello) implicitly classifies brain meninges as part of the NS, but not the CNS (or brain).

  is_a MP:0003631 ! nervous system phenotype
   is_a MP:0003632 ! abnormal nervous system morphology
    is_a MP:0005623 ! abnormal meninges morphology
     is_a MP:0009022 ! abnormal brain meninges morphology *** 

The implicit suggestion is that NS = CNS + PNS + M, but this contradicts MP's own definitions. I can see why the above structure may be desirable from a developmental biology perspective but it seems slightly odd. It is perhaps justifiable however if we define morphology of X to be inclusive of any immediately surrounding layers. It would be interesting to see how MP models the periosteal layer, which is histologically more aligned with bone than nervous system tissue.

ZFA

Zebrafish have a single meninx. The ZFA the text def states this surrounds the CNS, but places it as part-of the NS

[Term]
id: ZFA:0001355
name: primitive meninx
namespace: zebrafish_anatomy
def: "The vascularized fibrous sheath that surrounds the central nervous system." [http://www.archive.org/stream/contributionstoe04carn#page/n3/mode/2up]
synonym: "meninges" EXACT []
synonym: "meninx primitiva" EXACT []
xref: TAO:0001355
is_a: ZFA:0001488  ! multi-tissue structure
relationship: develops_from ZFA:0000045 ! neural crest
relationship: end ZFS:0000044 ! Adult
relationship: part_of ZFA:0000396 ! nervous system
relationship: start ZFS:0000000 ! Unknown

MA

MA has this as part_of the CNS (with subclasses that are part-of brain regions, such as the midbrain)

[Term]
id: MA:0001113
name: meninges
namespace: adult_mouse_anatomy.gxd
relationship: part_of MA:0000167 ! central nervous system

FMA

The meningeal cluster (ie mereological sum of each meninx) has no part-of axioms

 FMA:231572 ! meningeal cluster [subclass: "FMA:64989"] [subclass: "heterogeneous anatomical cluster"]
 FMA:321869 ! cranial meningeal cluster [subclass: "FMA:231572"] [subclass: "meningeal cluster"]
 FMA:321870 ! spinal meningeal cluster [subclass: "FMA:231572"] [subclass: "meningeal cluster"]

When we examine the parts and look at their partonomy we see that they are part-of the "cerebral part of head" but not the NS itself.

FMA relates the innermost layer to the CNS using adheres to. This seems clear in its ontological commitment that this is not part of the CNS:

         po FMA:317760 ! cerebral part of head
          po FMA:269079 ! cranial compartment
           po FMA:269081 ! cerebral part of leptomeningeal sac
            po FMA:78443 ! cranial pia mater ***  [adheres_to: "brain"]
            po FMA:83981 ! cranial arachnoid mater ***  [adheres_to: "cranial dura mater"]
           po FMA:71236 ! cranial dura mater ***  [adheres_to: "cranial arachnoid mater"]

What next?

I would advocate for consistency with FMA in this case (ie meninges surround and spatially disjoint from brain). However, note that this will lead to some odd axiomatization in the composite-metazoan and collected uberon releases; it will seem as if the relationships between meninges and CNS are species-specific when in fact it is really ontology-specific.

The ideal situation is we jointly decide on a modeling pattern. Cc @tfhayamizu @ybradford @cerivs

cmungall commented 8 years ago

This is related to the discussion of whether lymphatics are present in the CNS

This was claimed in a Nature paper by Louveau et al (doi:10.1038/nature14432 http://www.nature.com/nature/journal/v523/n7560/full/nature14432.html), which claimed discovery of lymph vessels in the dura mater (and inferring from this presence in the CNS). Both the CNS inference and the claims to a new discovery are disputed by Bucchieri et al (doi:10.1111/joa.12381, http://onlinelibrary.wiley.com/doi/10.1111/joa.12381/epdf):

The most recent studies (Aspelund et al. 2015; Lou- veau et al. 2015) did not confute the dogma that lymphatics in the central nervous system (CNS) do not exist. In fact, dura mater is not a component of the neuraxis but covers, along with the other meningeal layers, both the brain and the spinal cord

cmungall commented 3 years ago

Updating this ticket so it doesn't get auto-closed

Also noting the FMA still lacks part relations for these

NCIT also places meninges in the NS