Target innervation and LGN/colliculus development
Download
Report
Transcript Target innervation and LGN/colliculus development
LGN, superior colliculus, and
optic tectum development and
innervation
Chris Strang Ph.D.
Vision Sciences
WORB 308
[email protected]
975-7222
Learning Objectives
• LGN specification
– Zona limitans- Shh
– Timing of LGN lamination
• Superior colliculus and tectum specification
– Isthmus
– FGF8 and transcription factors Pax2, En1, and Otx2
– Timing of colliculus lamination
• Target innervation- gradients of ligands; receptors
– Ephrin-A2, Ephrin-A5; ephA3
– Ephrin-B; ephB
– Wnt3; Ryk
– EphrinB1; ephB2, ephB3
Once axons extending from ganglion cells have found their way
to the targets they must invade the target and find the proper
location either in the LGN and superior colliculus (mammals), or
optic tectum (non-mammals). The axons must maintain
retinotopic mapping, and segregate into eye specific and celltype specific layers. But, first, where and how do these targets
arise?
The LGN of the thalamus is the main
target of retinal projections in
mammals. The thalamus arises in
the diencephalon. It is induced by
Shh from the zona limitans, the
organizing region at the p2-p3
border. The zona limitans becomes
the boundary between the
prethalamus and thalamus. LGN
cells are born at the wall of the third
ventricle at the midline and migrate
to positions at the lateral margin of
the thalamus, and are then arranged
in layers.
At the time that the retinal
afferents arrive, the LGN is
incompletely laminated. Glial
cells are distributed in layers
after retinal afferents arrive
but before LGN neurons show
lamination. The adult pattern
of cell rich layers interleaved
with cell free layers arises
after innervation and glial
redistribution.
The colliculus and tectum are analogous
midbrain structures. In mammals, the
superior colliculus is involved in eye
movements, including saccades. It also
receives and integrates non-visual inputs.
It is homologous to the non-mammalian
optic tectum, but the tectum, rather than
the LGN, is major the visual path in nonmammals. For mammals and nonmammals, the posterior midbrain
boundary is defined by expression
patterns of Wnt1, Otx2, Gbx2, and FGF8
at the isthmus. Otx2 (pink) is expressed
from the forebrain to the isthmus at the midbrain-hindbrain border.
Fgf8 (orange) is expressed at the isthmus. Gbx2 (blue) is expressed
in the anterior hindbrain region. En1, En2 and Pax2 (hatched red)
are expressed in the midbrain and in the anterior hindbrain. Wnt1
(purple) is expressed at the posterior midbrain and at the dorsal
midline.
Pax2, En1, and FGF8 interact in specification
of the tectum. Overexpression of any of these
factors at the level of the diencephalon results
in ectopic tectum formation. Interestingly,
overexpression of Otx2 in the hindbrain also
results in ectopic tectum formation, indicating
that Otx2 is also a required factor. Shh may
be instrumental in limiting tectum
development to the dorsal neural tube.
Overexpression of Shh on one side results in
elimination of tectum on that side.
Nakamura, (2001) TINS, 24:32-39
Engrailed 1 and 2 are also
important in retinal targeting
to superior colliculus. The En
gradients are high at the
isthmus, and decrease further
anteriorly (En1) and
posteriorly (En2). Reversal of
the isthmus can reverse
engrailed gradients and
subsequent axon innervation.
Lamination and innervation of the
superior colliculus. In humans at
embryonic week 8, the
presumptive colliculus consists
of one layer of proliferating
progenitor cells at the ventricular
zone. Cell rich areas begin to
appear around embryonic week
11. At approximately 14 weeks
fiber layers begin to appear. The
characteristic alternation of white
and grey matter layers seen in
the mature superior colliculus is
evident by embryonic week 16.
Retinal axons begin to invade the
superficial layers by week 12, and
penetrate into the deeper layers
by week 13.
Qu et al, (2006) Experimental Eye Research, 82:300-310
However, while prenatal retinal
inputs have adult-like
organization, the SC cells
themselves are functionally
immature until sometime after
birth. Studies in cat have
demonstrated that maturity of
the superior colliculus
projections to upstream brain
areas may develop from center
to periphery. For example, the
ability to orient to visual cues is
dependent on superior
colliculus, and the ability to
orient to cues in the center of the
visual field develops before the
orienting response to cues in the
periphery.
Qu et al, (2006) Experimental Eye Research, 82:300-310
Retinotopic mapping
must be maintained
when axons invade the
target to form synaptic
connections. Studies in
the frog first gave solid
evidence for the
possibility that there are
specific molecular cues
that determine how
axons invade the target
tissue. How do axons
from anterior retina
choose posterior
tectum, and axons from
the posterior retina
choose anterior optic
tectum?
A series of crucial
experiments showed that
there are specific
molecules in the tectum
itself that inhibit axon
growth. Concentration
varied with tectal
position, and the
sensitivity of retinal
axons to the factor varied
with retinal position.
These factors turned out
to be the receptor
tyrosine kinases called
the eph kinases, and
their membrane
associated ligands, the
ephrins.
These studies lead to the identification of 2
RTK ligands, ephrin-A2 and ephrin-A5. These
are expressed in low to high gradients in the
anterior direction in the tectum. The
complimentary Eph receptors are expressed
by chick retinal ganglion cells in a low to high
nasal to temporal gradient. These receptors,
ephA3, bind to both ephrin-A2 and A5. The
ephrins inhibit temporal retina axon growth so
the axons stop in the anterior tectum. Axons
from anterior (nasal) retina express less
ephA3 and so can continue to grow into
posterior tectum. There is a second gradient
of Wnt3 which is highest in medial tectum and
lowest laterally. The Wnt receptor is highest in
ventral retina. Thus each axon has a
“preference” for specific locale in tectum. This
preference is probably manifested by a
difference in adhesion, with signaling making
the substrate more or less permissive for axon
growth.
As in frog tectum, GCs in mouse nasal
retina express low EphA and project to
posterior superior colliculus, which has
a high concentration of ephrin-A. GCs
in temporal retina have high expression
of EphA and project to anterior
colliculus.
Dorsal and ventral GC mapping to lateral and medial SC is mediated
by an ephrin-B gradient and EphB receptors. Low expression of
EphB receptors by dorsal GCs maps to the lateral (ventral) SC. GCs
in ventral retina have high expression of EphB and project to the
medial (dorsal) SC.
Topographical organization of retinogeniculate projections are also
specified by ephrin gradients. GCs in nasal retina project to dorsal
contralateral LGN. GCs in temporal retina project to ipsilateral ventral
LGN. RGCs in ventral retina project to the medial LGN, while GCs in
dorsal retina project to lateral LGN.
Wnt3 is expressed in a medial to lateral gradient in chick tectum and
mouse colliculus. The Wnt3 receptor, Ryk, is expressed by retinal
ganglion cells in a gradient that decreases from ventral to dorsal in
GCs. GC axon termination zones are repelled by Wnt3 expression.
Wnt3 activation of different receptor, Frizzled, mediates attraction.
These cues interact with the gradient established by Eph receptors
to provide a mechanism for mapping and axon branching in the
tectum.
Schmitt et al., (2006) Nature 493(5):31-36
Wnt3 signaling mediated by Ryk appears to provide for lateral
mapping. Ventral axons are repelled by Wnt3, while dorsal axons are
attracted by low Wnt3 and repelled by high Wnt3. EphrinB1 activation
of EphB2 and EphB3 receptors provides medial mapping by attracting
axon branches. When Wnt3 repulsion is blocked, ephrinB1–EphB
signaling causes branches to project medially. When ephrinB1–EphB
signaling is eliminated, interstitial branches are repelled from the
medial side by Wnt3 and project laterally.
In many species, axon innervation of the target
must be refined. As in humans, the segregation
of the cat colliculus into eye specific bands takes
place prior to birth. Direct retinal input is initially
only to the superficial layers and axons from both
eyes invade along the entire colliculus. The
ipsilateral projection retracts later.
In the LGN of cats and primates, GCs project to
eye-specific LGN layers (LGN projects to layer IV
of visual cortex forming ocular dominance
columns). However, althought GC axons
projecting to the LGN invade the correct target
region, they initally extend a little beyond and then
pull back. It turns out that the axons actually make
two kinds of wrong projections. Prenatal axons
branch in the wrong layer and the terminal
arborization in the correct layer is too wide.
Improper connections in the LGN and SC are
pruned back postnatally. Axon innervation and
map refinement are activity dependent but begin
before visual experience.
Learning Objectives
• LGN specification
– Zona limitans- Shh
– Timing of LGN innervation and lamination
• Superior colliculus and tectum specification
– Isthmus
– FGF8 and transcription factors Pax2, En1, and Otx2
– Timing of colliculus innervation and lamination
• Retinotopic mapping- gradients of ligands; receptors
– Ephrin-A2, Ephrin-A5; ephA3
– Ephrin-B; ephB
– Wnt3; Ryk
– EphrinB1; ephB2, ephB3