The developing human embryo. The bulge at the rostral end is the telencephalon. The dorsal telencephalon (the pallium) becomes the cortex (Thyng, 1914).
The developing human embryo. The bulge at the rostral end is the telencephalon. The dorsal telencephalon (the pallium) becomes the cortex (Thyng, 1914).
The neocortex is the youngest part of the brain, and though it is highly organized both functionally and anatomically, the patterning mechanisms that underlie that delineation are not as clear as in the older parts of the brain. The rhombencephalon for example, which develops in a highly segmented manner, forms compartments called rhombomeres which can be correlated with hox gene expression; the patterning of the prosencephalon is very different.[1] The most anterior-dorsal portion of the proencephalon gives rise to the telencephalon, of which only the dorsal telencephalon goes on to develop into the neocortex. This complex compartmentalization is underscored not by gene expression – as is the case in the older part of the brain – but mainly by overlapping morphogen gradients.[2] Gradients themselves however would provide poor definition, but the boundaries seen in the cortex are much sharper. This is due to the presence of signaling centers that outline spatial patterning by the expression of various morphogens, and serve as the boundaries in the developing cortex.[1] The three main cortical signaling centers will be discussed here, their role in development, as well as the morphogens that define them.


1 Cortical Hem
2 Commissural Plate
3 Cortical Anti-Hem
4 References

Cortical Hem

Postion of the signaling centers. The Cortical Hem in blue lies along the dorsal midline, the Commissural Plate in red lie at the anterior pole, and the Cortical Anti-Hem outline the ventral-lateral edge of the pallium (Cognigni, 2007).
Among the signaling centers, the cortical hem is most directly involved in structure formation in the neocortex, in addition to its roles in patterning and cell proliferation. Composed of neuroepithelial tissue, it is best defined as the area of the developing neocortex that acts as the source for the morphogens Wnt and Bmp.[3] In fact, Wnt3a which, in the cortex, is expressed solely in the cortical hem is often used as a marker. Staining reveals that the cortical hem is located in the dorsomedial aspect of the developing neocortex.[4] The position itself suggests involvement in dorsal and medial fates, and indeed the structures upon which the cortical hem has a direct effect – the hippocampal formation and the choroid plexus – are found in the midline. The cortical hem acts as the primary organizer for the hippocampus, its effect mediated through Wnt, while the choroid plexus cells are derived from the cortical hem tissue itself (differentiation mediated by BMP signals).[5] This leads to a decrease in the size of the hem during the course of development; eventually the remnants of the cortical hem are thought to form the radial glia scaffold of the fimbria. Finally, new evidence also implicates the cortical hem in neuronal migration; specifically of the Cajal-Retzius cells, which originate from the hem.[6]

The Lhx Factors

The Lhx class of LIM-homeodomain proteins is extensively involved in cell differentiation and patterning: for example, both Lhx6 and Lhx7 can be found in the medial ganglionic eminence, with the former involved in migration and the latter in differentiation; in the noecortex, Lhx2 and 5 have a prominent role in moderating the development of the cortical hem.[7]

Of the two factors, Lhx5 seems to be most closely tied to cortical hem development. Lhx5 knockouts have shown impaired hippocampal development – the hippocampal progenitor cells are specified, but proliferation is low. In addition, there is a complete loss of cerebral choroid plexus. These are two areas under direct control of the cortical hem, and the loss of function is similar to what is seen when ablatingthe cortical hem.[8]

Lhx2 shows a more sophisticated gradient of expression, and its role in patterning differs substantially from that of Lhx5. Lhx2 expression is high posterior-medially, tapering off toward the anterior-lateral end. However, at the point of the posterior-medial end where the Lhx2 gradient intersects with the cortical hem there is a sharp decline in Lhx2 expression, mainly due to suppression by BMP2 and BMP4.[7] This means that there are high Lhx2 levels directly adjacent the cortical hem, in areas where the choroid plexus and hippocampus will develop, and suggests that this discrete boundary delineates the cortical hem position. Support for this hypothesis comes from Lhx2 knockouts which have shown dramatic expansions of the hem and commensurate decreases in the rest of the neocortex. Lhx2 function has been further elucidated through the use of chimera mice (composed of cells containing both null and wildtype Lhx2); loss of Lhx2 in the lateral cortex showed an expansion of anti-hem fate, suggesting that just as it regulates the cortical hem, so does Lhx2 regulate the cortical anti-hem as well.[9]

WNT Signaling

The Wnt family of signaling molecules is involved in a variety of processes during development and its role changes both spatially and temporally; in the cortex alone Wnt signals are important for both patterning and expansion (proliferative signals). Within the neocortex, the cortical hem exclusively expresses several members of the Wnt family, specifically Wnt2b, Wnt5, and most importantly Wnt3a.[10] Though the Wnt signal can be transducded through a variety of pathways, the “canonical” Wnt/β-catenin pathway is the one that has been implicated in cortical development. This pathway consists of the Wnt glycoprotein binding to the Frizzled G protein-coupled receptor which triggers an intracellular signal cascade that activates β-catenin mediated transcription.[1]

Wnt genes are expressed throughout the developing embryo. As can be seen, the expression patterns of the genes alone are incredibly complex, but the Wnt signals themselves function in diffusion gradients, and those gradients are modulated and interact with other morphogens, transcription factors - in the brain this intrinsic patterning is then modified by migration of neurons and axons. The embryo in this video is further along in development than what is discussed here, but the midline Wnt signals can still be distinguished.

The β-catenin mediated Wnt signals are required to maintain the identity of the cortex through the specification and maintenance of the dorsal-ventral axis. Since the ventral aspect of the pallium (cerebral cortex) consists of the subpallium (mainly the basal ganglia), Wnt signaling maintains cortical identity by preventing the adoption of a subpallial fate.[1] This role seems to be restricted temporally – Wnt signaling seems to be involved in dorsal-ventral patterning only prior to neurogenesis. When β-catenin was inactivated during this time, there was a downregulation of dorsal markers and an upregulation of ventral markers in the neocortex, indicating a shift in the axis. However this was not seen after neurogenesis had already started. In addition, adding a Wnt signaling source in the subpallium resulted in the repression of ventral cell fate.[11]

Wnt signals are also important for defining the boundaries of cortical hem influence, and accordingly it is an antagonism with Fgf8, the major commissural plate morphogen, that modulates the Wnt gradient. Fgf8 is a very effective suppressor of Wnt, so much so that expanding the Fgf8 gradient has been shown to completely repress Wnt 2b, 3a, and 5 expression in the cortical hem; the expected patterning defects occurred, including hippocampal malformation.[2]

The Hippocampal Formation
The hippocampus (Camillo Golgi).
The hippocampus (Camillo Golgi).

Wnt3a has been shown promote the expression of dorsal markers, but its main role is in hippocampal development, acting as a proliferative signal and enhancing cell survival. Wnt signaling from the cortical hem does not specify hippocampal cell fate; as knockout studies have shown, ablating the cortical hem or suppressing Wnt signaling results in the same phenotype: a small pool of precursors where the hippocampus should be.[12] This conditional knockout can be achieved using Tlc, a secreted form of the extracellular domain of the Frizzled receptor. Wnt binds to Tlc as it would to its receptor, but this prevents Wnt from exerting its normal effects.[13] A Tlc-related Wnt antagonist is also present naturally in the cortical anti-hem. Further evidence for Wnt3a comes from work on the actual Wnt signal cascade. Mice deficient in Lef1 and Tcf transcription factors – two downstream mediators of the canonical Wnt pathway – show similar defects even in the presence of Wnt signals.[14]

BMP Signaling

Bone morphogenetic protein (BMP) signals are especially important during neural tube formation regulating polarity, and then after neural tube closing when BMP4 and BMP2 are secreted from the roof plate to inhibit Lhx2 and establish cortical hem boundaries.[7] However, during cortical patterning BMP activity diminishes. BMP signals differ from Wnt signals in two significant ways: they act in a more local fashion and they signal for differentiationinto non-neural tissue. Acting on the cortical hem itself and the immediate region surrounding it, BMP signaling results in the generation of the most medial neuroepethelial derivative – the choroid plexus.[15]

Both over and under expression of BMPs have demonstrated its role; in the former, a constitutively active BMP receptor (BMPR1a) leads to the development of choroid plexus beyond the midline, encroaching on the neocortex; in the latter, mice without BMPR1a in the pallium demonstrate a dramatic reduction in choroid plexus cells.[15] Therefore, this shows that the cortical hem has a role in medial-lateral patterning, through the combination of Wnt and BMP signals. However, BMP also has a role in anterior-posterior patterning; since Fgf8 represses Wnt, BMP signalling must act to suppress Fgf8 in order to prevent an anterior shift in the neocrotex. The localized BMP gradient also ensures that the Fgf8 suppression does not infringe on the commissural plate but rather helps to define its borders of activity. To this effect, when BMP4 signaling was inhibited by Noggin, Fgf8 activity appeared ectopically across the pallium.[16]

Commissural Plate

The commissural plate is the signalling center that develops the earliest. Though all the signaling centers are established after the neural tube closes, the commissural plate develops from the anterior neural ridge (ANR), a prominent source of Fgf signals at the time of neural tube formation, and it is in many ways continuous with the ANR.[17] They are both located at the medial anterior pole of the developing neuroepithelium, and so act as centers for anterior-posterior patterning. Additionally, the commissural plate retains the ANR’s characteristic Fgf signals, with several Fgfs such as Fgf3, 8, 17, and 18 all overlapping in expression at the anterior pole.[17] New evidence using diffusion tensor imaging shows that the commissural plate also functions later in development as a guide for commissural fibres in the forebrain; all three of the major commissural fibres cross at the commissural plate – and it seems that many of the same morphogens and transcription factors that regulate patterning also participate in this role as well.[18]
B) Gradients of Emx2 and COUP-TFI are shifted toward the anterior pole in Fgf8 mutants. C) The addition of sFgfr3 (Fgf8 antagonist) shows a shift in the COUP-TFI expression past the anterior boundary, indicated by the white arrow (Moldrich, R.X. et al. 2010).

Fgf8 Signaling

Fgfs are ubiquitous during development, and as a large family with a variety of members, accordingly they have a diverse set of roles. For example, mice lacking Fgf8 die at gastrulation, before the neural tube even has a chance to develop.[19] However, later in development the large number of Fgf family members presents some redundancies in the gradients – when Fgf1 is knocked out, the shifts in cell fate are minimal due to the presence of other Fgf gradients. These signaling gradients generated by the Fgfs actually exert their effects through the establishment of secondary transcription factor gradients, and it is these that directly affect development.[20]

Conditional knockouts have shown that lack of Fgf8 or Fgf receptor 1 (Fgfr1) lead to midline and commissural plate defects – these defects occur during the transition from ANR to commissural plate, a time during which the neural tube closes and the signaling center becomes concentrated at the midline. After the formation of the commissural plate Fgf8 continues to play an important role in anterior-posterior patterning; if the Fgf8 gradient is shifted toward the posterior end, or Fgf8 expression increased, the posterior cortex adopts an anterior fate.[20] Likewise, using an inhibitor of Fgf8 (such as a secreted form of Fgfr3) will cause loss of anterior cortex; this patterning is accomplished indirectly through repression of the transcription factors Emx2 and COUP-TFI at the anterior pole in a concentration-dependent manner.[21]

Locally at the anterior pole of the forebrain, Fgf8 works indirectly through the expression of the Emx1 transcription factor. Emx1 homozygous null mutants are still viable and show few cortical malformations (which is not the case for Fgf8 knockouts, confirming that the Emx1 path is just one of many pathways Fgf8 can take).[18] However, there is a complete loss of the corpus callosum, indicating a lack of commissural axon crossing, and defects in the fibres of the anterior commissure, which lies further from the commissural plate.[18]

Interaction between Fgfs

Unlike much of the family, Fgf2 is not restricted to the commissural plate and is involved not in patterning but in cell proliferation (much like Wnt3a in the cortical hem). Knockout experiments show that a reduction in Fgf2 expression leads to a decrease in the number of glutamatergic neurons in the frontal and parietal – but not occipital – lobes.[22] However this split between anterior and posterior lobes suggests that Fgf2 is expressed as a gradient and may still be involved in patterning earlier in development. Not all members of the Fgf family are anterior signals - mice with constitutively active Fgf receptor 3 (Fgfr3) resulted in expanded posterior fates, leading to enlarged occipital and temporal areas.[21] The Fgf8 gradient was unaffected in these mutants, so the reason for the posterior expansion most likely lies in the activity of another Fgf (such as Fgf1 or Fgf9) that signals for caudal fates.[21]

The effects of Fgfs are potent, and so the gradient is somewhat restricted not only by opposing morphogen gradients (such as the activity of Bmps from the cortical hem), but also by intrinsic molecular activity: the Fgf gradient is thought to be regulated by the necessary binding of heparin sulfate glycosaminoglycans.[23] Fgfs interact with one another as well; Fgf15 for example, opposes the effects of Fgf8. While Fgf8 increases cell proliferation, Fgf15 suppresses proliferation and increases differentiator.[19] This self-regulation and antagonism within the Fgf family demonstrates how the sharp boundaries of the cortex can be generated with many overlapping gradients.

Cortical Anti-Hem

Shows the precise definition of the anti-hem when using Sfrp2 as a marker (Assimacopoulos, 2003).
The anti-hem is the most loosely defined of the three signaling centers, and its role in patterning the least understood.[22] As the name suggests, the cortical anti-hem is positioned as the mirror-image of the cortical hem, found in the ventral-lateral portion of the pallium. It is therefore assumed that the anti-hem participates in medial-lateral patterning, and findings suggest that it has a role in ventral patterning as well.[24] Like the other signaling centers, delineating borders is also one of its functions; in this case, the anti-hem, positioned on the future dividing line between pallium and subpallium, is suggested to maintain the boundary between the two.[25] Interestingly, the cortical anti-hem also gives rise to Cajal-Retzius cells, just like the hem – this might be an indication of a larger role for Wnt and associated morphogens in neuronal migration. The anti-hem is characterized by Sfrp2 (a Wnt antagonist related to Tlc) and three members of the epidermal growth factor (EGF) family.[23]

Sfrp2 Antagonism

Members of the EGF family: notice the diffuse gradients (Assimacopoulos, 2003).

The secreted frizzled receptor protein (Sfrp2) best delineates the anti-hem as a boundary and is the best marker for its position. It is secreted from the anti-hem to counteract Wnt signaling from both the cortical hem and ventrally from the subpallium.[24]
Though the cortical hem is the major source of Wnt signaling in the pallium, Wnts are ubiquitous throughout the developing embryo, and as part of the pallial-subpallial boundary, it is the job of the anti-hem to create a neutral zone separating the dorsal and ventral telencephalon. The Sfrp2 gradient will suppress Wnt signals from the cortical hem to allow for the expression of ventral markers; in addition, Sfrp2 will also act to suppress the Wnt7b gradient spilling over from the subpallial side.[24] A localized Sfrp2 signal prevents the spread of a subpallial fate and also acts to sequester Wnt7b, which is thought to act as a signal for radial glia scaffolds and so participates in neuron migration.[22]

EGF Signaling

Though most of the members of the EGF family (including Egf itself) are expressed predominantly in the ventral aspect of the developing pallium, they are not concentrated in the ventral-lateral stripe that defines the anti-hem. Tgfα and Neuroligins (Nrg) 1 and 3 however, do concentrate in the anti-hem during the period of cortical patterning when the other signaling centers are active (approximately embryonic days 9-13).[23] The activity of these EGFs is still disputed, but evidence from Drosophila analogs suggest the specification of ventral fates. Another possible role is the patterning of the cortical limbic system; the limbic system-specific membrane protein LAMP can be induced in non-limbic cells upon exposure to EGF ligands.[23] These are only putative roles, and further investigation should elucidate the functions particular to the anti-hem.


  1. ^ Kiecker, C. & Lumsden, A. Compartments and their boundaries in vertebrate brain development. Nature Neuroscience 6, 553-564 (2005)
  2. ^ Fotaki, V., Price, D. J., & Mason, J. O. Wnt/b-Catenin Signaling is Disrupted in the Extra-toes (Gli3Xt/Xt) Mutant from Early Stages of Forebrain Development, Concomitant with Anterior Neural Plate Patterning Defects. The Journal of Comparative Neurology 519, 1640–1657 (2011)
  3. ^ Backmana, M. et al. Effects of canonical Wnt signaling on dorso-ventral specification of the mouse telencephalon. Developmental Biology 279, 155– 168 (2005)
  4. ^ Assimacopoulos, S., Grove, E.A., & Ragsdale, C.W.. Identification of a Pax6-dependent epidermal growth factor family signaling source at the lateral edge of the embryonic cerebral cortex. Journal of Neuroscience 23, 6399–6403 (2003).
  5. ^ O’Leary, D., Chou, S., & Sahara, S. Area Patterning of the Mammalian Cortex. Neuron 56, 252-269 (2008).
  6. ^ Cho, G., Lim, Y., & Golden, J. A. XLMR candidate mouse gene, Zcchc12 (Sizn1) is a novel marker of Cajal–Retzius cells. Gene Expression Patterns 11, 216–220 (2011).
  7. ^ Chou, S., Perez-Garcia, S., Kroll, T.T. & O’Leary., D. Lhx2 specifies regional fate in Emx1 lineage of telencephalic progenitors generating cerebral cortex. Nature Neuroscience 12, 1381-1390 (2009).
  8. ^ Fragkouli, A., Verhey van Wijk, N., Lopes, R., Kessaris, N. & Pachnis, V. LIM homeodomain transcription factor-dependent specification of bipotential MGE progenitors into cholinergic and GABAergic striatal interneurons. Development 136, 3841-3851 (2009).
  9. ^ Vishakha S. Mangale et al. Lhx2 Selector Activity Specifies Cortical Identity and Suppresses Hippocampal Organizer Fate. Science 319, 304-309 (2009).
  10. ^ Hasenpusch-Theil, K. et al. Transcriptional Analysis of Gli3 Mutants Identifies Wnt Target Genes in the Developing Hippocampus. Cerebral Cortex (2012).
  11. ^ Lorenza Ciani and Patricia C. Salinas. WNTs in the Vertebrate Nervous System: from Pattering to Connectivity. Nature 6, 351-362 (2005).
  12. ^ Wei, L.C. et al. Low-dose Radiation Stimulates Wnt/β-catenin Signaling, Neural Stem Cell Proliferation and Neurogenesis of the Mouse Hippocampus in vitro and in vivo. Current Alzheimer’s Research (2012).
  13. ^ Kim, H. et al. Downregulation of Wnt/β-catenin signaling causes degeneration of hippocampal neurons in vivo. Neurobiology of Aging 32, 2316-2331 (2011).
  14. ^ Galceran, J., Miyashita-Lin, E.M., Devaney, E., Rubenstein, J.L. & Grossched, R. Hippocampus development and generation of dentate gyrus granule cells is regulated by LEF1. Development 127, 469-482 (2000).
  15. ^ Liu, A. & Niswander, L.A. Bone Morphogenetic protein signaling and nervous system development. Nature Reviews Neuroscience 6, 945-954 (2005).
  16. ^ Shimogori, J. et al. Embryonic signaling centers expressing BMP, WNT and FGF proteins interact to pattern the cerebral cortex. Development 131, 5639-5647 (2004).
  17. ^ Cajal, M. et al. Clonal and molecular analysis of the prospective anterior neural boundary in the mouse embryo. Development 139, 423-436 (2012).
  18. ^ Moldrich, R.X. et al. Molecular regulation of the developing commissural plate. Journal of Comparative Neurology 18, 3645–3661 (2010).
  19. ^ Borello, U., Cobos, I., Long, J.E., Murre, C. & Rubenstein, J. FGF15 promotes neurogenesis and opposes FGF8 function during neocortical development. Neural Development 3 (2008).
  20. ^ Toyoda, R. et al. FGF8 acts as a classic diffusible morphogen to pattern the neocortex. Development 137, 3439-3448 (2010).
  21. ^ Thompson, R.E. et al. Fgf receptor 3 activation promotes selective growth and expansion of occipitotemporal cortex. Neural Development 4 (2009).
  22. ^ Tiberi, L., Vanderhaeghen, P. & Van den Ameele, J. Cortical neurogenesis and morphogens: diversity of cues, sources and functions. Current Opinion in Cell Biology 24, 269–276 (2012).
  23. ^ Faedo, A., Borello, U. & Rubenstein, J. Repression of Fgf Signaling by Sprouty1-2 Regulates Cortical Patterning in Two Distinct Regions and Times. The Journal of Neuroscience 30, 4015– 4023 (2010).
  24. ^ Assimacopoulos, S., Grove, E.A. & Ragsdale, C.W. Identification of a Pax6-Dependent Epidermal Growth Factor Family Signaling Source at the Lateral Edge of the Embryonic Cerebral Cortex. The Journal of Neuroscience 23, 6399–6403 (2003).
  25. ^ Hansen, D.V., Rubenstein, J. & Kriegstein, A.R. Deriving Excitatory Neurons of the Neocortex from Pluripotent Stem Cells. Neuron 70, 645–660 (2011).