Lmx1a is a master regulator of the cortical hem
Curation statements for this article:-
Curated by eLife
eLife assessment
This important work advances our understanding of the generation of a main organizer of the vertebrate brain, the cortical hem. The authors convincingly show the contribution of multiple downstream effectors, each involved in specific processes regulated by the master gene, Lmx1a. This study has broader implications for how secondary organizers are created in the embryo and will be of interest to a wide readership.
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
Abstract
Development of the nervous system depends on signaling centers – specialized cellular populations that produce secreted molecules to regulate neurogenesis in the neighboring neuroepithelium. In some cases, signaling center cells also differentiate to produce key types of neurons. The formation of a signaling center involves its induction, the maintenance of expression of its secreted molecules, and cell differentiation and migration events. How these distinct processes are coordinated during signaling center development remains unknown. By performing studies in mice, we show that Lmx1a acts as a master regulator to orchestrate the formation and function of the cortical hem (CH), a critical signaling center that controls hippocampus development. Lmx1a co-regulates CH induction, its Wnt signaling, and the differentiation and migration of CH-derived Cajal–Retzius neurons. Combining RNAseq, genetic, and rescue experiments, we identified major downstream genes that mediate distinct Lmx1a-dependent processes. Our work revealed that signaling centers in the mammalian brain employ master regulatory genes and established a framework for analyzing signaling center development.
Article activity feed
-
-
Author Response
Reviewer #1 (Public Review):
Iskusnykh et al. present an elegant and thorough analysis of the role of transcription factor Lmx1a as a master regulator of the cortical hem, which is a secondary organizer in the brain. The authors report that loss of Lmx1a in the hem alters expression levels of Wnts, that Lmx1a is critical for hem progenitors to exit the cell cycle properly, and that Lmx1a loss leads to defects in CR cell differentiation and migration. Furthermore, the authors show that hem-like fate can be induced by overexpressing Lmx1a. This is a fundamental role for a transcription factor that was long used as a hem marker but was never examined for its function in the hem. This study has broader implications for how secondary organizers are created in the embryo and would be of great interest to a wide readership. …
Author Response
Reviewer #1 (Public Review):
Iskusnykh et al. present an elegant and thorough analysis of the role of transcription factor Lmx1a as a master regulator of the cortical hem, which is a secondary organizer in the brain. The authors report that loss of Lmx1a in the hem alters expression levels of Wnts, that Lmx1a is critical for hem progenitors to exit the cell cycle properly, and that Lmx1a loss leads to defects in CR cell differentiation and migration. Furthermore, the authors show that hem-like fate can be induced by overexpressing Lmx1a. This is a fundamental role for a transcription factor that was long used as a hem marker but was never examined for its function in the hem. This study has broader implications for how secondary organizers are created in the embryo and would be of great interest to a wide readership. The conclusions are broadly well supported by the data, though there are a few points of interpretation that need to be addressed.
We appreciate the positive comments and insightful suggestions of Reviewer 1. Please see our response to specific comments below. New text in the revised paper is blue (see our marked up copy of the paper, submitted as related manuscript file). Please note that since we reformatted the paper (re-submitted figures separately rather than embedded them into the text), line numbers changed relative to the original submission.
(1) Figure 3A shows staining intensity in WT and Lmx1a-/- whereas the quantification has Lmx1a+/-. Both genotypes are relevant, -/- and +/-, to test whether the loss of 1 copy of Lmx1a results in a partial diminution of Wnt3a levels. Likewise, it is necessary to examine Wnt3a expression levels in the Wnt3a+/- embryo. Together, these could explain why the Lmx1a+/-; Wnt3a+/- double heterozygote has a DG phenotype, otherwise, it remains an unexplained though interesting observation.
In the original paper, the label in the Wnt3a quantification panel (Fig. 3C) contained a typographical error. The label should read “Lmx1a-/-“, not Lmx1a+/-. (Originally, we did not analyze Lmx1a expression in Lmx1a+/- embryos; we analyzed only wt and Lmx1a-/- embryos.) We apologize for this error and corrected the label typo in the revised manuscript (Fig. 3C).
Based on the above comment, in the revised manuscript, we analyzed the expression of Wnt3a in Lmx1a and Wnt3a single and double heterozygotes, in addition to wt and Lmx1a-/- embryos. To address a comment of Reviewer 2 about a “limited robustness of quantification of in situ hybridization signal”, we isolated CH by LCM and analyzed Lmx1a expression by qRT-PCR (Fig. 3D, E). Interestingly, we found that loss of one copy of either Wnt3a or Lmx1a does not significantly downregulate Wnt3a expression, but loss of one copy of Lmx1a on the Wnt3a+/- background (Lmx1a+/-;Wnt3a+/- mice) reduces Wnt3a expression, providing additional evidence that Lmx1a regulates expression of Wnt3a and explaining the appearance of the DG phenotype only in the double (but not single-gene) heterozygotes. These data are now described in the Results section (page 12, lines 255-260 and Fig. 3D, E). All of our Wnt3a expression data are now properly presented.
(2) Line 309: "to test Wnt3a as a downstream mediator of Lmx1a function in CH/DG development, we performed an analysis of Lmx1a/Wnt3a double heterozygotes rather than Wnt3a overexpression rescue experiments in Lmx1a -/- mice." The authors' reasoning is unclear. The double het experiments do not go on to show that one gene acts via the other. It's entirely possible the two act via parallel pathways. However, since Lmx1a does indeed regulate Wnt3a levels, this is a good argument for suggesting it acts via Wnt3a, even without the overexpression rescue. The authors could reorganize the data and rephrase the definitive "acts via" statement (also in the heading of this section, line 289, and discussion, line 553) to better fit the data.
Thank you for this comment. We reorganized/improved our reasoning as requested. Now we state that we performed an analysis of Lmx1a/Wnt3a double heterozygotes to test “whether Lmx1a and Wnt3a co-regulate hippocampal development” (rather than to test Wnt3a as a downstream mediator of Lmx1a function, as it was stated before) (page 12, lines 271-272). As correctly suggested by the Reviewer, we now conclude that “Although these double heterozygote experiments alone do not necessarily show that one gene acts via the other, as two genes may act via parallel pathways, reduced expression of Wnt3a in Lmx1a-/- embryos and downregulation of Wnt3a expression in Lmx1a+/-;Wnt3a+/- embryos relative to Wnt3a+/- embryos show that Lmx1a acts upstream of Wnt3a, thus, suggesting that Lmx1a promotes DG development, at least partially, by modulating expression of Wnt3a.” (page 13, lines 277-282).
We rephrased the definitive "acts via" statement throughout the text and in the heading of this section. Now we use more balanced phrases. The heading now reads: “Lmx1a regulates expression of Wnt3a to promote DG development.” (Page 11, line 241), while in the Discussion we state that Lmx1a regulates Wnt signaling to promote hippocampal development (page 21, lines 467-468).
(3) In the discussion section, the authors should include that trans-hilar and supragranular scaffold is disrupted in Lrp6 and Lef1 single as well as double mutants, which indicates Wnt signaling has a role to play in the morphogenesis of this scaffold. In this context, the author may discuss how Lmx1a could regulate this process via modulating Wnt signaling.
Now in the Discussion we state: “It has also been previously shown that single and double mutants for Lrp6 and Lef1 genes, which encode components of the Wnt signaling transduction pathway, exhibit disrupted transhilar and supragranular scaffolds (Zhou et al., 2004; Li and Pleasure, 2005), indicating that Wnt signaling has a role in the development of the hippocampal glial scaffold” (Page 20, lines 445-449). Then, we conclude “Our gene expression studies and phenotypic analysis of Lmx1a-/- mutant and Lmx1a+/-;Wnt3a+/- double heterozygous mice identified Lmx1a as a novel regulator of proliferation of DG progenitors, hippocampal glial scaffold formation and electrophysiological properties (input resistance) of DG neurons, which likely, at least partially, promotes hippocampal development by modulating Wnt signaling, particularly expression of its secreted ligand Wnt3a. ” (Page 20, lines 449-454).
(4) Reduction in Tbr2 levels (Fig4B): E13.5, not all Tbr2+ cells in the hem show a visible decrease in Tbr2 levels. The CR cells in the marginal zone show faint Tbr2. It would be useful if the staining intensity within the hem was quantified by dividing the section into three bins along the radial axis: Ventricular Zone, "Intermediate" zone, and Marginal zone to get a sense of the intensity profile. Co-labeling with p73 would identify CR cells and distinguish them from hem progenitors.
We co-labeled wt cortical hem with Tbr2 and p73 immunohistochemistry and found that virtually all Tbr2+ cells in the marginal layer (where CR cells accumulate before initiating their tangential migration toward the hippocampal fissure) are p73-positive, while most Tbr2+ cells in the ventricular and intermediate bins are p73-negative (presumably not fully differentiated progenitors) (Figure 4 – figure supplement 2). These data provide further rationale for quantifying Tbr2+ progenitors separately in three different bins, as recommended by the Reviewer, which we now report in Figure 4B, C. This analysis revealed that loss of Lmx1a reduces Tbr2 expression across the three bins in the CH, but most significantly (p<0.001) in the Marginal zone.
These data are now described in the Results section, page 14, lines 308-317.
(5) Are the total number of Prox1+ cells at E14.5 similar between control and Lmx1a-/- ? Might the decrease in Prox1+ cells in the DG of P21 Lmx1a-/- animals occur due to granule cell death or because fewer cells were specified due to lower Wnts from the compromised Lmx1a-/- hem? The authors should examine cell death, labeling with CC3 and Prox1 together to test the cell death angle and discuss if the specification angle applies.
Our new cell counts revealed a reduced number of Prox1+ cells in the DNe of e14.5 Lmx1a-/- mutants (Fig. 1K-M). We also show that proliferation in e14.5 DNe is reduced in Lmx1a mutants (Fig. 1N-Q), which is expected to contribute to the reduced number of Prox1 cells. Since proliferation is diminished in Lmx1a mutants, it is very hard to definitively demonstrate whether (in addition to proliferation) a reduced specification of DG progenitors contributes to the lower number of Prox1+ cells found in the DNe (and later in DG) of Lmx1a mutant mice. However, since Wnt3a is known to both induce DG progenitors and promote their proliferation, it is likely that a reduced specification also contributes to the reduced number of Prox1 cells in Lmx1a -/- mutants. Now we discuss this possibility in the Discussion by stating: “Wnt3a, which is downregulated in the Lmx1a-/- CH, is known to promote not only proliferation but also the specification of DG progenitors (Lee et al., 2000; Mangale et al., 2008; Subramanian and Tole, 2009b). Thus, although not directly tested in the current study, it is likely that the reduced number of Prox1+ DG progenitors in Lmx1a-/- embryos results not only from their reduced proliferation but also because of their decreased specification.” (page 22, lines 497-501).
To study whether increased apoptosis contributes to the reduced number of Lmx1a-/- DG cells, we performed a very detailed analysis of apoptosis with an activated Caspase 3 immunohistochemistry at multiple stages (at e14.5 in the DNe, before DG cells exit the DNe; at e16 and e18.5 in the hippocampal primordium, and at e18.5, P3 and P21 in the DG (when the DG is formed), using Prox1/activated Caspase 3 co-immunostaining). No difference in apoptosis was found at any stage between wt and Lmx1a-/- embryos, indicating that misregulated apoptosis is not a major contributor to the DG phenotype of Lmx1a-/- mutants (Fig. 1R-T; Fig. 1- figure supplement 3).
(6) In figure 6, the authors show that Lmx1a OE is sufficient to induce hem-like features, and identify p73+ cells (CR cell lineage). Is the choroid lineage not induced or was it not examined? A line to this effect would be useful. Also, the validation that it is indeed ectopic hem could be stronger with a few additional markers, since this is a striking finding.
In the original paper, induction of the choroid plexus lineage was not investigated. Now we add two additional markers: Ccdc3 (a marker of CH) and Ttr (a marker of choroid plexus). Lmx1a in utero electroporation into medial telencephalic neuroepithelium induced ectopic expression of Ccdc3 (Fig. 6 – figure supplement 1A-D’) but did not induce expression of Ttr (Fig. 6 – figure supplement 1E-F’), strengthening the conclusion that Lmx1a specifically induces CH features in the medial telencephalon. These data are now described in the Results section, page 17, lines 372-373, 377-379, and 387-389.
Reviewer #2 (Public Review):
The cortical hem is one of the main signaling centers in the vertebrate forebrain, regulating neurogenesis of the medial pallium and the generation of Cajal-Retzius neurons. The authors examine how this signaling center is formed and functions. Previously, transcription factors playing instructive roles in the development of the cortical hem have been identified, but a master regulator had not been found so far. The authors build on their previous work studying the transcription factor Lmx1a which is one of the earliest and most specific cortical hem markers.
By combining loss- and gain-of-function studies, RNA sequencing, histology, and analysis of downstream factors, the authors rigorously show Lmx1a is required for the expression of signaling molecules in the hem, the proliferation and functionality of dentate gyrus neurons, the cell cycle exit and differentiation (and also migration) of cajal-retzius cells and this by activating different downstream regulators.
They use golden standard experiments in the field such as BrdU-Ki67 cell-cycle exit measurements, RNA sequencing, and patch clamping; combined with state-of-the-art techniques such as RNAscope and laser capture microdissection. These convincingly show that Lmx1a regulates the proliferation of dentate gyrus progenitor cells and a malformation of the transhilar scaffold.
We appreciate the positive comments and insightful suggestions of Reviewer 2. Please see our response to specific comments below (see our marked up copy of the paper, submitted as related manuscript file). New text in the revised paper is blue. Please note that since we reformatted the paper (re-submitted figures separately rather than embedded them into the text), line numbers changed relative to the original submission. The authors also claim a migration deficit for dentate gyrus progenitors, but they do not consider apoptosis or show direct evidence for migration abnormalities.
Now we provide additional in vivo data to support migration abnormalities from the DNe (Fig. 1 – supplement 2) and modified the Discussion related to migratory defects from the DNe as recommended by the Editors. Also, by performing a very detailed analysis of apoptosis, we provide strong evidence that apoptosis is not altered in Lmx1a-/- mutants at multiple stages (Fig. 1 – supplement 3). These results are described in detail below, in our response to the first specific comment of Reviewer 2.
In the hem, the authors report normal proliferation and apoptosis in the Lmx1a mutants, but aberrant cell-cycle-exit, from which the authors conclude a problem in differentiation. However, this could be a cell cycle progression problem too (stuck in a certain cell cycle phase?), as the RNAseq data suggest. The authors should acknowledge this possibility.
The possibility of a cell cycle progression problem in Lmx1a -/- CH is now acknowledged in the Discussion. Specifically, we state: “Finally, in Lmx1a mutants, we linked a decreased number of CR cells with a reduced exit of CH progenitors from the cell cycle. However, our data do not exclude a possibility that loss of Lmx1a also causes a cell cycle progression defect (resulting in CH progenitors being delayed in a certain phase of the cell cycle). This hypothesis remains to be tested.” (page 22, lines 501-505).
The RNAseq dataset provides candidate downstream regulators of the observed phenotypes and the authors test the functionality of Wnt3a, Tbr2, and Cdkn1a, showing they are involved in distinct processes.
Strikingly, Wnt3a is not significantly downregulated in the RNAseq data in the Lmx1a mutant, but quantification of in situ hybridization signal (which is less robust) did reveal a significant difference. Is this a splice variant issue? A timing issue or specificity of the RNAscope probe? The authors should look into this more carefully.
Our Wnt3a RNAscope in situ hybridization recapitulates known Wnt3a expression pattern (specific expression in the CH), indicating that this probe is specific. A splice variant issue is also unlikely because, according to the Genome Browser and the NCBI Gene Bank, only one Wnt3a splice variant exists in the mouse. It can be a timing issue (e13.5 for RNAseq versus e14 for RNascope analysis). But, please, note that in our RNAseq experiment, the FDR for Wnt3a downregulation was 0.13, which is close to significance.
To further address the downregulation of Wnt3a expression in Lmx1a-/- CH, we performed additional experiments using a complementary technical approach. We isolated the CH from e14 wt and Lmx1a-/- mutants by laser capture microdissection (LCM) and analyzed Wnt3a expression by qRT-PCR with already published/validated primers for Wnt3a (Watanabe et al., 2016, Biol Open 5, 1834-1843). We focused on e14 because it is closer to e14.5 when we observed a reduced proliferation in the DNe in Lmx1a-/- embryos. Our new LCM/qRT-PCR analysis confirmed Wnt3a downregulation (Fig. 3D, E) that we initially observed in our in situ hybridization experiments (Fig. 3A-C), increasing our confidence that Lmx1a regulates Wnt3a expression in the CH.
To study the role of Cdkn1a, the authors performed rescue experiments using in utero electroporation, which is a standard in the field. However, they argued before that "CR cell migration and DG morphogenesis are complex processes that require precise expression levels of key genes" when studying downstream factors Wnt3a and Tbr2. Why is this no longer an issue studying Cdkn1a?
This is because, in Cdkn1a rescue experiments, we test a much simpler (binary) output: whether electroporated (GFP+ cells) are Ki67 positive (cycling progenitors) or Ki67 negative (exited the cell cycle). In contrast, Wnt3a or Tbr2-related experiments require the evaluation of either DG formation (the number of Prox1+ cells in the DG) or the location of CR cells in the HF, both of which are very complex outputs. (DG formation relies on the correct proliferation, glial scaffold formation, migration and differentiated events, while CR location involves long-range migration). Both DG morphogenesis and CR migration are highly sensitive to the expression level of their essential developmental genes (Zhou et al., 2004; Arredondo et al., 2020; Gil et al., 2014; Ha et al., 2020; Hevner, 2016 in the paper reference list). As in utero electroporation does not easily allow precise control of gene expression level, such an approach would likely produce higher levels of Wnt3a and Tbr2 in at least some cells of Lmx1a-/- embryos relative to endogenous levels of Wnt3a/Tbr2 in wild type mice. Higher than physiological levels of expression of these proteins may cause additional abnormalities, complicating the interpretation of results of Wnt3a and Tbr2 electroporation experiments aimed to rescue Lmx1a-/- hippocampal phenotypes.
As mentioned above, because in the case of Cdkn1a, we test a much simpler output (the presence or absence of Ki67 expression), we do not expect Cdkn1a overexpression to complicate the interpretation of the results: some electroporated Lmx1a-/- cells could exit the cell cycle “too fast”, but it still does not complicate the interpretation of the Ki67 expression readout.
We provide additional explanations for the Cdkn1a rescue experiment in the paper. We state: “To study whether decreased Cdkn1a expression mediates a reduced cell cycle exit of CH progenitors in Lmx1a-/- embryos (Fig. 2A-C), we used immunohistochemistry with antibodies specific for Ki67, which labels cycling progenitors. As the presence/absence of Ki67 expression is a simpler output than complex DG morphogenesis and long-range migration of CR cells, we performed Cdkn1a overexpression rescue studies using in utero electroporation of the CH at e11.” (Pages 15-16, lines 344-347).
To study cell-cycle exit in this model, the authors quantified GFP and Ki67. Since electroporation not only targets the progenitor cells (see e.g. Govindan et al. 2018, Nature protocols), the authors should confirm these results with a BrdU/Ki67 quantification as in previous experiments, or confirm electroporation only targeted progenitor cells in their model.
Now we experimentally demonstrated that electroporation targets progenitor cells in our model. Thus, we confirmed that our approach is appropriate for the analysis of progenitor differentiation in the CH.
Specifically, we in utero electroporated a GFP expressing plasmid into the CH of e11 embryos and imaged the GFP signal 15 hrs later (to identify electroporated cells) together with Ki67 immunolabeling (to identify progenitors). We reasoned that 15 hrs would be sufficient to produce GFP protein from the plasmid but also short enough to avoid differentiation of progenitors that received the plasmid. We found that in both wt and Lmx1a-/- embryos, almost all GFP+ cells in the CH were Ki67+ (e.g., progenitors). There was no difference between wt and Lmx1a-/- embryos at this early time point (Fig 5 – supplement 1). (GFP+/Ki67- cells were extremely rare in both genotypes. These cells may be either differentiated cells that took the plasmid during electroporation or electroporated progenitors that exited the cell cycle during the 15-hr interval after electroporation.)
In the Results section, we now state: “The ventricular layer of the CH that borders the lateral ventricles consists of progenitor cells, so it is expected that plasmids injected into the lateral ventricles and electroporated into the CH will target such progenitors. However, since electroporation can also target differentiated cells (Govindan et al. 2018), we first injected a GFP-encoding plasmid into the lateral ventricles, electroporated it in utero into the CH of e11 embryos and analyzed GFP+ cells after a short (15 hrs) time period. This analysis revealed that virtually all (~95%) GFP+ cells were Ki67+ (progenitors) in both wild type and Lmx1a-/-embryos (Fig. 5 – figure supplement 1), confirming that this system is appropriate to target progenitors.” (Page 16, lines 348-355).
Lastly, the authors ectopically expressed Lmx1a and convincingly show its ability to generate a hem-like structure. Could the authors elaborate on the necessity for a medial signature? Can the hem be ectopically induced in the lateral pallium?
To address this question, we electroporated Lmx1a into the lateral cortex and found that laterally, it could not induce a major cortical hem marker Wnt3a (Fig. 6 – supplement 2). Thus, a medial identity is required for Lmx1a to induce the cortical hem, the finding which is now presented in the Results section (page 17, lines 388-389).
Also, in the Discussion, we elaborate on the necessity for a medial signature: “Interestingly, while Lmx1a induced CH features in the medial telencephalon, Lmx1a overexpression in the lateral cortex failed to induce ectopic expression of Wnt3a, indicating that medially expressed competence factors (permissive genes) are needed to maintain the CH-inducing activity of Lmx1a. Such factors are likely to include Gli3 and Dmrt3/4/5, loss of which compromises the development of the endogenous CH (Grove et al., 1998; Kikkawa and Osumi, 2021; Quinn et al., 2009; Subramanian et al., 2009a; Subramanian and Tole, 2009b) (page 19, lines 424-430).
-
eLife assessment
This important work advances our understanding of the generation of a main organizer of the vertebrate brain, the cortical hem. The authors convincingly show the contribution of multiple downstream effectors, each involved in specific processes regulated by the master gene, Lmx1a. This study has broader implications for how secondary organizers are created in the embryo and will be of interest to a wide readership.
-
Reviewer #1 (Public Review):
Iskusnykh et al. present an elegant and thorough analysis of the role of transcription factor Lmx1a as a master regulator of the cortical hem, which is a secondary organizer in the brain. The authors report that loss of Lmx1a in the hem alters expression levels of Wnts, that Lmx1a is critical for hem progenitors to exit the cell cycle properly, and that Lmx1a loss leads to defects in CR cell differentiation and migration. Furthermore, the authors show that hem-like fate can be induced by overexpressing Lmx1a. This is a fundamental role for a transcription factor that was long used as a hem marker but was never examined for its function in the hem. This study has broader implications for how secondary organizers are created in the embryo and would be of great interest to a wide readership. The conclusions are …
Reviewer #1 (Public Review):
Iskusnykh et al. present an elegant and thorough analysis of the role of transcription factor Lmx1a as a master regulator of the cortical hem, which is a secondary organizer in the brain. The authors report that loss of Lmx1a in the hem alters expression levels of Wnts, that Lmx1a is critical for hem progenitors to exit the cell cycle properly, and that Lmx1a loss leads to defects in CR cell differentiation and migration. Furthermore, the authors show that hem-like fate can be induced by overexpressing Lmx1a. This is a fundamental role for a transcription factor that was long used as a hem marker but was never examined for its function in the hem. This study has broader implications for how secondary organizers are created in the embryo and would be of great interest to a wide readership. The conclusions are broadly well supported by the data, though there are a few points of interpretation that need to be addressed.
(1) Figure 3A shows staining intensity in WT and Lmx1a-/- whereas the quantification has Lmx1a+/-. Both genotypes are relevant, -/- and +/-, to test whether the loss of 1 copy of Lmx1a results in a partial diminution of Wnt3a levels. Likewise, it is necessary to examine Wnt3a expression levels in the Wnt3a+/- embryo. Together, these could explain why the Lmx1a+/-; Wnt3a+/- double heterozygote has a DG phenotype, otherwise, it remains an unexplained though interesting observation.
(2) Line 309: "to test Wnt3a as a downstream mediator of Lmx1a function in CH/DG development, we performed an analysis of Lmx1a/Wnt3a double heterozygotes rather than Wnt3a overexpression rescue experiments in Lmx1a -/- mice."
The authors' reasoning is unclear. The double het experiments do not go on to show that one gene acts via the other. It's entirely possible the two act via parallel pathways.
However, since Lmx1a does indeed regulate Wnt3a levels, this is a good argument for suggesting it acts via Wnt3a, even without the overexpression rescue. The authors could reorganize the data and rephrase the definitive "acts via" statement (also in the heading of this section, line 289, and discussion, line 553) to better fit the data.(3) In the discussion section, the authors should include that trans-hilar and supragranular scaffold is disrupted in Lrp6 and Lef1 single as well as double mutants, which indicates Wnt signaling has a role to play in the morphogenesis of this scaffold. In this context, the author may discuss how Lmx1a could regulate this process via modulating Wnt signaling.
(4) Reduction in Tbr2 levels (Fig4B): E 13.5, not all Tbr2+ cells in the hem show a visible decrease in Tbr2 levels. The CR cells in the marginal zone show faint Tbr2. It would be useful if the staining intensity within the hem was quantified by dividing the section into three bins along the radial axis: Ventricular Zone, "Intermediate" zone, and Marginal zone to get a sense of the intensity profile. Co-labeling with p73 would identify CR cells and distinguish them from hem progenitors.
(5) Are the total number of Prox1+ cells at E14.5 similar between control and Lmx1a-/- ? Might the decrease in Prox1+ cells in the DG of P21 Lmx1a-/- animals occur due to granule cell death or because fewer cells were specified due to lower Wnts from the compromised Lmx1a-/- hem? The authors should examine cell death, labeling with CC3 and Prox1 together to test the cell death angle and discuss if the specification angle applies.
(6) In figure 6, the authors show that Lmx1a OE is sufficient to induce hem-like features, and identify p73+ cells (CR cell lineage). Is the choroid lineage not induced or was it not examined? A line to this effect would be useful. Also, the validation that it is indeed ectopic hem could be stronger with a few additional markers, since this is a striking finding.
-
Reviewer #2 (Public Review):
The cortical hem is one of the main signaling centers in the vertebrate forebrain, regulating neurogenesis of the medial pallium and the generation of Cajal-Retzius neurons. The authors examine how this signaling center is formed and functions. Previously, transcription factors playing instructive roles in the development of the cortical hem have been identified, but a master regulator had not been found so far. The authors build on their previous work studying the transcription factor Lmx1a which is one of the earliest and most specific cortical hem markers.
By combining loss- and gain-of-function studies, RNA sequencing, histology, and analysis of downstream factors, the authors rigorously show Lmx1a is required for the expression of signaling molecules in the hem, the proliferation and functionality of …
Reviewer #2 (Public Review):
The cortical hem is one of the main signaling centers in the vertebrate forebrain, regulating neurogenesis of the medial pallium and the generation of Cajal-Retzius neurons. The authors examine how this signaling center is formed and functions. Previously, transcription factors playing instructive roles in the development of the cortical hem have been identified, but a master regulator had not been found so far. The authors build on their previous work studying the transcription factor Lmx1a which is one of the earliest and most specific cortical hem markers.
By combining loss- and gain-of-function studies, RNA sequencing, histology, and analysis of downstream factors, the authors rigorously show Lmx1a is required for the expression of signaling molecules in the hem, the proliferation and functionality of dentate gyrus neurons, the cell cycle exit and differentiation (and also migration) of cajal-retzius cells and this by activating different downstream regulators.
They use golden standard experiments in the field such as BrdU-Ki67 cell-cycle exit measurements, RNA sequencing, and patch clamping; combined with state-of-the-art techniques such as RNAscope and laser capture microdissection. These convincingly show that Lmx1a regulates the proliferation of dentate gyrus progenitor cells and a malformation of the transhilar scaffold. The authors also claim a migration deficit for dentate gyrus progenitors, but they do not consider apoptosis or show direct evidence for migration abnormalities.
In the hem, the authors report normal proliferation and apoptosis in the Lmx1a mutants, but aberrant cell-cycle-exit, from which the authors conclude a problem in differentiation. However, this could be a cell cycle progression problem too (stuck in a certain cell cycle phase?), as the RNAseq data suggest. The authors should acknowledge this possibility.The RNAseq dataset provides candidate downstream regulators of the observed phenotypes and the authors test the functionality of Wnt3a, Tbr2, and Cdkn1a, showing they are involved in distinct processes.
Strikingly, Wnt3a is not significantly downregulated in the RNAseq data in the Lmx1a mutant, but quantification of in situ hybridization signal (which is less robust) did reveal a significant difference. Is this a splice variant issue? A timing issue or specificity of the RNAscope probe? The authors should look into this more carefully.To study the role of Cdkn1a, the authors performed rescue experiments using in utero electroporation, which is a standard in the field. However, they argued before that "CR cell migration and DG morphogenesis are complex processes that require precise expression levels of key genes" when studying downstream factors Wnt3a and Tbr2. Why is this no longer an issue studying Cdkn1a?
To study cell-cycle exit in this model, the authors quantified GFP and Ki67. Since electroporation not only targets the progenitor cells (see e.g. Govindan et al. 2018, Nature protocols), the authors should confirm these results with a BrdU/Ki67 quantification as in previous experiments, or confirm electroporation only targeted progenitor cells in their model.Lastly, the authors ectopically expressed Lmx1a and convincingly show its ability to generate a hem-like structure. Could the authors elaborate on the necessity for a medial signature? Can the hem be ectopically induced in the lateral pallium?
-