Since the early comparative studies in animal models (i.e., mouse, chick, and fish), developmental biologists have shed light on the precise choreography of genetically controlled events that shape the vertebrate brain in different domains along the AP (antero-posterior) and D-V (dorso-ventral) axes. These events produce the conserved organization of the forebrain anteriorly, morphologically and functionally distinguished in different domains (Puelles, 1995; Puelles et al., 2000). The pallium originates from the dorsal division of the anterior most part of the forebrain (telencephalon) and gives rise to the neocortex in mammals (Haines and Mihailoff, 2018); responsible for the high computation capacity beyond sensory-motor integration, such as language and abstract thinking skills.

Within the developing pallium, a precise balance and timing of signaling drive neural fate commitment and patterning, as well as the rate of progenitor pools’ expansion (level 1), migratory behavior and differentiation (level 2), and the establishment of connectivity patterns (level 3). Fine changes in local patterning schemes and precursor cells’ behavior across evolution underlie the array of diversity in vertebrate cognitive capabilities and the formation and expansion of the multilayered cytoarchitecture of the mammalian cortex, responsible for high-order functions typical of humans (Striedter, 2005; Van Essen et al., 2018; Stepien et al., 2021). From a comparative point of view, mechanisms of cortex development at the level of repertoire of stem cells and proliferation strategies vary drastically even within closely related species of vertebrates and underly size and folding variation. The molecular roots of these differences are intensely studied by employing lissencephalic (harboring smooth brain) and gyrencephalic mammalians (showing superficial foldings and therefore expansion of the neocortex) (Lui et al., 2011; Dehay et al., 2015; Florio et al., 2015, 2016; Kalebic et al., 2019; Stepien et al., 2021). Furthermore, the presence of a homologous domain giving rise to the neocortex within the pallium territory in non-mammalian vertebrates is equally debated and little consensus exists to this day (Northcutt, 1995; Wullimann and Mueller, 2004; Medina and Abellán, 2009; Nieuwenhuys, 2009). Nevertheless, the basic core of orderly molecular and cellular processes governing anatomical and functional compartmentalization (bauplan) of the forebrain and fundamental principle of neurogenesis are conserved within the vertebrate lineage (Bachy et al., 2002; Puelles and Rubenstein, 2003; Wilson and Houart, 2004; Medina et al., 2005). Interestingly, specific cortical high-order cognitive abilities have also been mapped in teleosts (Portavella et al., 2002; Salas et al., 2003; von Trotha et al., 2014; Messina et al., 2022) and ancestral features can also be found in invertebrates (Williams et al., 2004; Cavallin et al., 2018).

The continuous identification of new genetic lesions affecting these different levels of neurogenesis and resulting in heterogeneous neurodevelopmental disorders with malformation of cortical development (MCD) highlights the importance of these processes during cortex formation. Improving our knowledge of the specific biology and genetics of pallium patterning, regionalization, precursor behavior and cell type specification and its variation on the main vertebrate theme can potentially impact our ability to model, understand and manage MCD. Indeed, these heterogeneous conditions can manifest with various degrees of microcephaly, lissencephaly and cortical dysplasia, originating from impaired proliferation, migration and/or white matter and circuits establishment. They are often associated with cognitive impairments and with at least 40% of non-treatable epileptic conditions (Guerrini and Dobyns, 2014; Barkovich et al., 2015; Represa, 2019). The clinical variability and genetic heterogeneity, as well as the poor knowledge on the underlying patho-mechanisms of the newly described conditions, often make these pediatric disorders not accurately managed. To this aim, non-primate mammalian systems such as rodents are undoubtedly providing an enormous help. Nevertheless, human-specific traits are often not successfully recapitulated in these models (especially related to folding formation) (Wong and Roper, 2016). On the other hand, more recently, the genetically accessible ferrets, gyrencephalic models with an expanded neocortex, are becoming popular to gain insights into neocortex expansion and to model specific traits which characterize human cortex development and disease (Hutsler et al., 2005; Martínez-Cerdeño et al., 2012; Florio et al., 2015, 2016; Fernández et al., 2016; Johnson et al., 2018; Kalebic et al., 2019; Pinson et al., 2019).

Nevertheless, given the current acceleration in the identification of candidate genetic lesions in previously undiagnosed MCD conditions, scientists are exploring alternative systems. These should (i) be more readily implementable than mammalian species in healthcare-oriented institutes, (ii) offer possibilities for fast candidate gene variants validation serving differential diagnosis, while (iii) parallelly allow the search of basic patho-mechanisms, and (iv) via methodologies which reduce the overall burden related to animals experimentation. To this aim, teleost fish can be advantageous (Figure 1 and Table 1). They represent the largest clade of vertebrates and, together with chick and mice, have long served as genetic and developmental models to study the principles of neural induction, and for the search of evolutionary paths to plasticity and adaptation of pallium formation and function within vertebrates (Salas et al., 2003; Menuet et al., 2007; Alié et al., 2018; Dunlap et al., 2019). Indeed, homologous neuronal ensembles, basal circuits and neurogenic modes within the pallium of teleost fish are being identified (Fotowat et al., 2019; Porter and Mueller, 2020). On this ground, and because of a number of factors determining its practical usefulness in developmental genetics and disease modeling, the small and transparent teleost fish zebrafish is effectively used. In many instances, the use of fish can replace less ethically acceptable methodologies required for brain analysis in mammalian embryos and young adults. Along these lines, more teleost models (such as medaka) are recently joining the forces and proving to be crucial to dissect multiple mechanisms of pallium expansion, vertebrate brain growth and rare brain diseases (Indrieri et al., 2013; Peluso et al., 2013; Kirchmaier et al., 2015; Dunlap et al., 2019; Coolen et al., 2020; Nakanishi et al., 2021).

A different and complementary approach, quickly gaining popularity and reducing further the burden on animal models, is the establishment of induced pluripotent stem cells (iPSCs) directly from patients’ bioptic material and iPSCs-derived cerebral organoids (COs). The latter are becoming attractive in vitro alternatives in conjunction to the in vivo models to recapitulate species (i.e. human)-specific traits of early neurogenesis, differentiation, connectivity and even cytoarchitecture at small scale (Krefft et al., 2021; Zilova et al., 2021; Figure 1 and Table 1). Direct comparisons of COs development derived from different primates, mammals or other vertebrates are also becoming supportive tools for our understanding of the mechanisms underlying neocortex expansion (Li et al., 2017b; Kanton et al., 2019; Pollen et al., 2019). This review provides an updated overview of the key developmental steps and precursor cells that build the mammalian cortex. We next summarize representative examples of known genetic alterations impacting various molecular and cellular processes and leading to cortical malformations as the cause of intellectual disability and epilepsy. In the context of classifying and dissecting the causes of an increasing number of undiagnosed cortical malformations discovered via patients’ genomic screening, we next examine the possible value of the alternative and complementary fish and dish-based models to the in vivo mammalian systems. We discuss fruitful examples showing case their usefulness in contributing dissecting the fundamentals of neuronal precursors’ behavior and the underlying mechanisms in healthy and diseased brains. We specifically outline the limitations and challenges derived, for example, from the differences in pallium developmental processes and cell type repertoire (teleosts) or from the minimal tissue complexity and lack of whole-organismal context (organoids).

Since the experiments of Spemann and Mangold in the early 20th century, developmental biologists used animal models to demonstrate that brain development is achieved by tightly controlled gene expression schemes that generate specific, recognizable local fingerprints. These events depend on the ability of cells in the early embryological fields to respond to varying doses of inducing and inhibiting molecules (morphogens) secreted by neighboring cells, which sequentially restrict the fate of each previously homogeneous domain, leading to patterning, regionalization, and cell type specification. These events step up the global brain architecture and its functional sub-specialization in different domains, and are globally conserved in evolution (Rallu et al., 2002; De Robertis, 2009). Understanding these processes has represented the ground research for flourishing comparative studies in developmental biology over the last decades that inspired various innovative experimental strategies. As a result, human genetic and brain diseases modeling in vivo have been improved and the breakthrough of organoids – including COs- is revolutionizing the study of the healthy and diseased mammalian brain.

In vertebrate embryos, parallelly to the neural plate folding, signaling activated by retinoic acid, FGF, Wnt and Wnt inhibitors pattern the A-P (anterior-posterior) territories, while a fine-tuning of Bmp, Shh, and Wnt signals from axial mesodermal and non-neural ectodermal cells contribute to define instead the D-V (dorso-ventral) embryo axis (Wilson and Rubenstein, 2000; Houart et al., 2002; Rallu et al., 2002; Wilson and Houart, 2004). In this context, the establishment of the forebrain (Table 1) at the anterior bulge of the developing brain occurs already by the end of gastrulation, via a fine signaling balance of the different morphogens secreted from the organizer and the surrounding tissues (Rubenstein and Beachy, 1998; Stern, 2001; Wilson and Houart, 2004). A nested grid of highly conserved transcription factors is therefore generated and defines the boundaries between the different domains within the forebrain (Fernandez et al., 1998; Puelles et al., 2000). Key signaling events worth mentioning include the activation of Wnt inhibitors, e.g., Dickkopf-1 (Mukhopadhyay et al., 2001), Cerberus (Bouwmeester et al., 1996; Piccolo et al., 1999) and the secreted frizzled-related protein Sfrp3, (Kawano and Kypta, 2003) which counteract “caudalizing factors” and are essential for the vertebrate head formation. Crucial is also the modulatory activity of Fgf8 which globally sets the telencephalon aside from the neighboring diencephalon and contributes to its patterning (Shanmugalingam et al., 2000; Yamaguchi, 2001; Houart et al., 2002). This is practically orchestrated by Bf-1 (Tao and Lai, 1992; Dou et al., 1999; Kitagawa et al., 2004), Emx, and Dlx genes (Fernandez et al., 1998; Puelles et al., 2000). The Emx/Pax6 + dorsal pallium region of the telencephalon gives rise to the cortex, the medial pallium to the hippocampus and the ventral domain to the amygdala. On the ventral side of the telencephalon, the Dlx + subpallium is defined from where the basal ganglia and the hypothalamus originates. Posteriorly to the telencephalon, the diencephalic domain produces the thalamus and pretectum (Rubenstein et al., 1994; Rubenstein and Beachy, 1998; Puelles et al., 2000; Wilson and Rubenstein, 2000; Houart et al., 2002; Wilson and Houart, 2004; Cavodeassi and Houart, 2012).

The development and plasticity of the neocortex, originating from the dorsal pallium in mammals as a late evolutionary innovation (Striedter, 2005; Kelava et al., 2013; Namba et al., 2019) depends on the response of progenitor cells to the above mentioned modulatory signalings via a precise and stereotypical spatial-temporal logic of events. These ultimately control the correct proliferation rate of the initial pool of stem cells, the movement of the newborn neurons to their destination, their maturation in a number of different neuronal cell types and the establishment of their final connectivity pattern. From here, the resulting neural circuitries are distinctively wired, serving functional specialization for higher functions. Each of these events is especially critical to obtain the laminar arrangement of the mammalian pallium (Rakic, 2009a; Rakic et al., 2009b; Fietz et al., 2010; Salomoni and Calegari, 2010; Taverna et al., 2014; Subramanian et al., 2020), ultimately formed by circa 80% of excitatory neurons and the rest of inhibitory interneurons moving from the ventral telencephalon (Parnavelas et al., 2000; Lodato and Arlotta, 2015; Laclef and Métin, 2018). Briefly, in the region lining on the brain ventricle of the developing mammalian cortex (ventricular zone, VZ) the founder neuroepithelial cells with their characteristic apico-basal polarity along the radial axis undergo various rounds of symmetric cell division (“proliferative division,” Table 1). These events produce two new progenitor cells from one and therefore contribute to expanding the pool of progenitor cells. Apical and basal (or outer) (a, b RG cells, respectively) originated from these founder cells and expressing characteristic markers of glia cells (Götz and Huttner, 2005; Kriegstein and Alvarez-Buylla, 2009) maintain the apico-basal polarity and divide mainly by asymmetric cell division (“neurogenic division,” Table 1). This ensures the birth of new RG cells, serving self-renewal purposes (and thereby keeping the pool of stem cells), and differentiating neurons or various types of committed basal progenitors (BP), such as intermediate progenitor cells (IPCs). These cells migrate basally to the subventricular zone (SVZ) and produce neurons only after additional rounds of cell divisions (Noctor et al., 2001, 2004; Kriegstein and Götz, 2003; Haubensak et al., 2004; Kriegstein and Alvarez-Buylla, 2009; Uzquiano et al., 2018; Table 1). A great deal of variability exists across different mammalian species with respect to the fine details regarding the arsenal of progenitors’ cells, their morphology, and behavior, as well as the proliferation potential and the neurogenesis length (Betizeau et al., 2013; Kalebic et al., 2018, 2019; Stepien et al., 2020, 2021).

Importantly, the developing cortex of lissencephalic models, such as mice, have little bRG cells which divide both symmetrically and asymmetrically (Wang et al., 2011) and are instead mostly populated by IPCs-type of cells which have limited proliferative potential (Haubensak et al., 2004; Noctor et al., 2004). On the other hand, gyrencephalic animals like humans have evolved mechanisms to generate foldings which increase the overall surface and therefore host more neural cells in a limited space (Zilles et al., 2013). In these animals, the BP of the SVZ can sustain multiple rounds of divisions to generate other precursor cells before dividing into neurons (Götz and Huttner, 2005; Lui et al., 2011; Florio and Huttner, 2014). Indeed, the increase in the number of the BP outer RG cells (oRG) originating from aRG, and their proliferative potential has been proposed as a contributor for the increase in cortex size (Nonaka-Kinoshita et al., 2013; Fernández et al., 2016), but it does not seem to be sufficient, given the presence of lissencephalic animals with a high number of BP (Kelava et al., 2013; Stepien et al., 2021). Seminal works in the past years have demonstrated that a variety of morphology exists among BP cells even in the same species and that this instead might be a key factor determining neocortical expansion (Betizeau et al., 2013; Kalebic et al., 2019). The different BP morphotypes can vary from cells with monopolar to multipolar processes and it appears as if the increase in the number of processes boosts the proliferative capability, most likely by making the BP cells more available to external signaling. This mechanism might represent one of the major cellular features deciding cortex size expansion and neocortex appearance (Lui et al., 2011; Kalebic et al., 2019; Subramanian et al., 2020). Sustained gliogenesis in the outer SVZ has been implicated in the expansion of the neuropile, and therefore, the neocortex convolutions in primates (Rash et al., 2019). Of note, the genetic bases of such peculiarity of precursor cell types for neocortex expansion are mapped using gyrencephalic models (Florio et al., 2015, 2016; Cárdenas and Borrell, 2020; Heide et al., 2020; Stepien et al., 2021). At the signaling level, a number of pro-proliferative molecules from -and interacting with-the ECM niche of the progenitor pool, likely involving Pi3K-AkT, mTOR, and MAPK signaling pathways, is being recognized as a crucial trigger for the distinctive proliferative capacity of BP cells as well as for their positioning in the developing cortex (Sheppard and Pearlman, 1997; Fietz et al., 2012; Moreno-Layseca and Streuli, 2014; Cavallin et al., 2018; Heinzen et al., 2018; Long et al., 2018; Kalebic et al., 2019; Subramanian et al., 2020; Kyrousi et al., 2021).

Next, the basal projections from progenitor cells allow newborn neurons and glia produced by asymmetric division to migrate radially to the destination in the cortical plate (CP) by somal translocation or active locomotion (Sheppard and Pearlman, 1997; Nadarajah et al., 2001). The first migrating neurons populate the deep layer of the forming cortex, while newborn ones are located more superficially (Rakic, 1974; Buchsbaum and Cappello, 2019). Again, fine signaling is required for the different phases of migration, with Reelin being the most relevant to control different phases of this migratory activity (Trommsdorff et al., 1999; Kuo et al., 2005; Hirota and Nakajima, 2017). To complete cortex formation, inhibitory interneurons from the ventral telencephalon (the transient ganglionic eminences in humans) migrate long-range moving tangentially to the RG fibers to reach their final correct position and connect with other neurons in the CP (Lavdas et al., 1999; Bellion et al., 2005; Miyoshi et al., 2010; Faux et al., 2012; Barber and Pierani, 2016; Table 1). Their peculiar saltatory migration behavior, characterized by pauses in between fast periods of movement, is obtained via microtubules-dependent nuclear translocation (Bellion et al., 2005; Silva et al., 2018). This mechanism seems to be essential to control the stream of migratory cells populating the cortex and to even influence neurogenesis in the cortex (Silva et al., 2018). Lastly, cortical neurogenesis is nearly completed when gliogenesis starts, the remaining RG cells differentiate into astroglia, and the pool of progenitors is extinguished (Noctor et al., 2008; Ohtsuka and Kageyama, 2019).

The high complexity and specialization of the molecular and cellular events governing neocortex development and function in space and time make them particularly vulnerable to genetic alterations. A large group of malformations of cortical development (MCD) has been described which are characterized by various defects at the level of brain size (decreased or increased), final position of neurons, and formation of cortical layers and foldings (Toi et al., 2009; Guarnieri et al., 2018; Romero et al., 2018; Severino et al., 2020; Subramanian et al., 2020). Different conditions have been classically categorized based on the underlying perturbed cellular event, namely: alteration in the proliferative state of the precursors (a), migratory defects (b), or neural connectivity and circuits formations (c) (Table 2). For instance, microcephaly (reduced brain size) is mainly associated with unbalanced proliferation and/or apoptosis of the neuronal progenitors. Aberrant migratory behavior across developing cortical layers results in ectopic neurons located in the periventricular and subcortical region, disorganization and loss of cortical folding or lamination lead to gyrification defects and result in various heterotopias, polymicrogyria, lissencephaly and dysplasia (Guerrini and Dobyns, 2014). The latest acceleration in the use of genomic sequencing in undiagnosed patients reveals high heterogeneity in the genetic and cellular mechanisms of MCD, which often show shared cellular and morphological features, challenging the boundaries of the original definitions. Deepening our knowledge of the underlying mechanisms of the newly described conditions is particularly relevant to improving the functional classification of these disorders often associated with severe cognitive impairments (Guerrini and Dobyns, 2014) and untreatable epilepsy (Represa, 2019). Many disease-causing mutations have already been identified, which affect the function of key proteins involved in cytoskeleton, chromosomes arrangement machinery, organelle stability and bio-trafficking (Table 2). The fine spatial and temporal balance of these events in the developing pallium controls neurogenesis, migration, and connectivity. Of note, the sustained activity of signaling pathways involved is relevant for cortical malformations (Hong et al., 2000; Hevner, 2015; Jansen et al., 2015; Ossola and Kalebic, 2021; Table 2). We summarize the main features and known alterations of a subset of these conditions. For a more detailed classification, please refer to Table 2 and (Woods, 2004; Guerrini and Parrini, 2010; Barkovich et al., 2015; Watrin et al., 2015; Passemard et al., 2017; Buchsbaum and Cappello, 2019).

Microcephaly is a rare clinical trait defined by the reduced head and brain size, with an occipital-frontal circumference often below the 3rd percentile. It can occur at birth (primary), or postnatally (secondary), as an isolated condition or in combination with other clinical features, i.e., developmental delay, epilepsy, and CNS malformations (Abdel-Salam et al., 2000; Woods, 2004; Marino et al., 2020; Becerra-Solano et al., 2021), as well as intellectual disability (Marino et al., 2020). Aside from environmental factors and viral infections, genetic mutations represent the most frequent underlying cause (Table 2). Indeed, many mutations in conserved genes have been already identified, which negatively impact progenitors’ expansion as well as migration of newborn neurons. Proteins directly controlling microtubules dynamics (Chakraborti et al., 2016) (e.g., KIF14, TBCD) (Filges et al., 2014; Flex et al., 2016), centrosome biogenesis (e.g., STIL) and centrosome proteins regulating microtubules and spindle orientation (e.g., ASPM) (Darvish et al., 2010) are often involved. The involvement of factors controlling chromatin organization (e.g., MCPH1) (Darvish et al., 2010) as well as altered kinase activity (e.g., PTEN) (Dhaliwal et al., 2021) and abnormal substrate phosphorylation of phosphates (e.g., PPP1R15B, (Kernohan et al., 2015) have also been described. The resulting alteration of mitotic spindle formation and orientation, cell cycle, as well as proper chromosome segregation during mitosis often generates a shift between symmetric versus asymmetric cell division, depleting the pool of progenitor cells and causing apoptosis of early precursors. A different group of genes causing microcephaly and linked functionally to Golgi homeostasis and trafficking is also emerging. For instance, secondary microcephaly combined with periventricular heterotopia was linked to impaired Golgi assembly and trafficking affecting membrane proteins, proliferation and cell migration (e.g., ARFGEF2, Banne et al., 2013). Furthermore, fragmented Golgi has also been recently associated with a neurodevelopmental condition showing microcephaly and caused by mutations in the new disease-gene ARF3, encoding a small GTPase involved in post-Golgi trafficking (Fasano et al., 2021). More recently, loss of function of WDR81 which impairs endosomal trafficking of pro-proliferative signal EGFR was functionally linked to microcephaly (Carpentieri et al., 2022).

Impaired migration of newborn neurons from the ventricle to the CP along the RG scaffold causes different forms of neuronal migration disorders (NMDs). This is a group of highly heterogeneous conditions, which can show mixed features, including microcephaly or other cerebral malformations (Sheen et al., 2006; Lui et al., 2011; Buchsbaum and Cappello, 2019; Table 2). Among NMDs, lissencephaly (also known as “smooth brain”) refers to a large number of heterogeneous conditions, mostly characterized by an aberrant and reduced cerebral folding, in combination with other cerebral malformations; while gray matter heterotopia specifically refers to abnormally located neuronal clusters (Watrin et al., 2015; Di Donato et al., 2017; Buchsbaum and Cappello, 2019; Buchsbaum et al., 2020; Koenig et al., 2021). Furthermore, neuronal migration errors (Guerrini and Parrini, 2010) or aberrant proliferation of glia precursors (Rash et al., 2019) also lead to polymicrogyria, characterized by architectural anomalies and cytological variations involving the number of cortical convolutions. Abnormal focal migration of the newborn neurons causing impaired cortical lamination has also been linked to focal cortical dysplasia, which is among the most frequent causes of refractory epilepsy in children (Tassi et al., 2002; Guerrini and Dobyns, 2014). As previously discussed, interneurons are essential for correct cortex formation and to develop healthy connectivity and signal transmission (André et al., 2010). Impaired behavior during their long migratory tour from the ventral telencephalon to the cortex and their arborization can cause a disorganized lamination with impaired circuits activity within the cortex, leading to epilepsy and cognitive dysfunction (Pancoast et al., 2005; Carabalona et al., 2012; Nakagawa et al., 2017; Buchsbaum and Cappello, 2019; Subramanian et al., 2020). Furthermore, studies exist that functionally correlate the alterations in interneurons’ location and connectivity patterns to seizures occurrence in cells derived from patients with focal cortical dysplasia (Calcagnotto et al., 2005; Cepeda et al., 2005; Subramanian et al., 2020).

At the molecular level, these conditions are associated with mutations in genes encoding proteins involved in microtubules dynamics and transport cytoskeleton-cell matrix interactions, e.g., LIS1, DCX, KIF2A, TUB1A, (Chakraborti et al., 2016), MAP1B, (Heinzen et al., 2018), ARX, FLNA (Parrini et al., 2006), ARFGEF2, (Sheen et al., 2004). Mutations in these genes hinder neuronal polarity and cause migratory impairment and unbalanced proliferative potential of the precursor cells (Carabalona et al., 2012; Buchsbaum and Cappello, 2019). In addition, impaired signaling, caused by mutations in key molecules of the mTOR, FGF and REELIN signaling and ECM components, can impair proliferation, migratory behavior or neurites formation during cortex development leading to severe forms of MCD. Among these, lissencephaly (Hong et al., 2000) and polymicrogyria besides characteristic forms of muscular dystrophy (Devisme et al., 2012) and polymicrogyria (Godfrey et al., 2007).

As discussed, new disease genes and candidate pathogenic variants involved in MCD are continuously described, which may underlie the poorly unexplored heterogeneity in the clinical, genetic, and mechanistic aspects of cortical malformations. Yet, our little understanding challenges the utility of the existing disease classification and the development of tailored therapies. Classically, mouse models have provided -and continue to provide- considerable experimental evidence into the fundamental mechanisms of forebrain development and cortex formation in health and disease (Buchsbaum and Cappello, 2019; Reinert et al., 2021). For instance, mutant studies in mice models have predicted the involvement of Reelin signaling in complex forms of cortical malformation before the actual discovery of patients harboring mutations in genes belonging to the signaling cascade (D’Arcangelo et al., 1995; Trommsdorff et al., 1999; Hong et al., 2000). On the other side, successful models for deciphering the mechanisms underlying differences in neocortex size are primates, and more recently ferrets, gyrencephalic mammals, which harbor an expanded neocortex likewise humans and for which in utero genetic manipulation of embryonic brain is now possible (Hutsler et al., 2005; Florio et al., 2015, 2016; Gilardi and Kalebic, 2021). These models are particularly beneficial for recapitulating specific patho/physiological traits of MCD, especially when addressing cortical growth and folding defects, which were not previously observed in mammals with smooth cortex such as mice (Masuda et al., 2015; Kalebic et al., 2019; Pinson et al., 2019).

The acceleration of large patients’ sequencing efforts made possible by advances and increased affordability of the NGS (next-generation sequencing) on one side and the consensus on minimizing the animal distress associated with the invasive nature of experimentation in these large animals, quest for possible alternative models. These might be used for fast and large-scale functional validation of variants of unknown impact. Ideally, valid model systems successfully complementing the established and emerging mammalian ones should: (i) be rapidly implementable and manageable in medical research institutes, (ii) bring to a fast validation of the candidate genetic variants, (iii) while reducing the reliance on invasive animal procedures and still (iv) allowing scientific and translational gains in terms of mechanistic understanding of the disease. In this sense, despite not harboring the multi-layered cortex and given the conservation of fundamental mechanisms, the arsenal of genetic, imaging, and behavioral tools, teleost fish (Figure 1 and Table 1) are actively being used. Many studies now provide a critical comparison of telencephalon development and the specific neurogenic programs in teleosts. This knowledge is beneficial for data interpretation, given the increased use of these models (especially zebrafish) for validating MCD-causing mutations and deciphering the underlying mechanism (Table 2).

Parallelly, a different in vitro approach is also gaining increasing popularity, which is the use of iPSCs and COs derived from patients’ fibroblasts (Figure 1 and Table 1). Despite many drawbacks associated with the in vitro nature of the system, these models are useful to directly assess certain mechanisms of human cortex neurogenesis in health and disease or to draw hypotheses that can help strategize in vivo experiments. This section examines the ground similarities and differences between teleost and mammalian pallium and the use of zebrafish, iPSCs and COs for brain development and disease, highlighting representative examples. We discuss the main benefits and limitations, which still need to be overcome to improve the translational value of these models.

The genetic and molecular paradigms of the ground patterning and regionalization of the forebrain are conserved within various vertebrates (Puelles et al., 2000; Wilson and Houart, 2004; Medina et al., 2005; Murakami et al., 2005) and even in more distantly amniotes vertebrates, i.e., teleost fish (Heisenberg et al., 1996; Fernandez et al., 1998; Shinya et al., 2001; Puelles and Rubenstein, 2003; Kitagawa et al., 2004; Ganz et al., 2012). These processes have deep roots in evolution such that the molecular topography of a “paleopallium” can be traced back even to basal Bilateria (Medina and Abellán, 2009; Tomer et al., 2010).

Teleost fish models are a successful radiation of vertebrates and account for at least half of existing ones. They have populated developmental biology and neurobiology laboratories since the first studies on body plan formation and neural induction. These fish have been instrumental in establishing the first embryological events patterning the vertebrate telencephalon (e.g., see Houart et al., 1998; Kitagawa et al., 2004; Hashiguchi and Mullins, 2013), and complementing findings in mouse and other species. However, it should be noted that, conversely to mammals, teleost fish, show a “non-laminar” pallium structure that does not generate a bona fide cortex (Ito and Yamamoto, 2009). Instead of forming by invaginating cells as in mammals, the teleostean pallium domain is generated by an eversion process, whose underlying mechanisms are still under debate. This process ultimately generates a medial T-shaped ventricle and two lobes, in net contrast with the overall mature morphology of the mammalian counterpart (Wullimann et al., 1996; Wullimann and Mueller, 2004; Nieuwenhuys, 2009; Yamamoto, 2009). To this date, there is a lack of a solid conclusion on the existence of homologous forebrain and pallium sub-domains between teleosts and mammals. Hence, the teleostean pallium has long been considered rather “simple” because of these morphological differences and the uncertainty on assigning homology. This has precluded teleost models for a long time from further exploitation in neuroscience and brain disease modeling and limits the use and data interpretation of teleost fish in the field of MCD. Nevertheless, it is largely accepted by a wide community that, to a certain extend, an embryological origin, anatomical and functional homology exists between mammalian and teleost forebrain (Wullimann et al., 1996; Wullimann and Mueller, 2004; Nieuwenhuys, 2009; Yamamoto, 2009). Anatomical and functional components homologous to a number of pallial domains which originated from the dorsal portion of the telencephalon have already been proposed. In the course of the last decade, researchers have found in teleost fish regions homologous to the amygdala driving motivated behavior (von Trotha et al., 2014; Porter and Mueller, 2020), a domain resembling the hippocampus, which mediates memory and spatial information (Portavella et al., 2002; Salas et al., 2003; Rodríguez et al., 2021) and one responsible for cortex-type of high cognitive functions, such as logical and inference to determine social status (Grosenick et al., 2007) or numerosity tuning (Messina et al., 2022). On the other side, differences and plastic changes of the brain and pallium development at the level of morphology, connectivity and function exist across Bilateria and within vertebrates (Northcutt, 1995). One could say that the pallium itself underwent multiple events of sophistication and diversification in aquatic and land vertebrates, likely due to the elaboration of early stem cells independently from allometric relationships (Striedter, 2005; Noreikiene et al., 2015; Montgomery et al., 2016). Such phenotypic plasticity independently of other brain regions is manifest in teleosts, which results in considerable variations in the forebrain topology, neuroanatomy, and functional organization, likely underlying their ability to adapt to a variety of environments (Ito et al., 2007; Sylvester et al., 2013). Therefore, it should be mentioned that teleosts also represent an experimental advantage to showcase fundamental principles on how brain modularity and diversification within and across species is obtained. As shown indeed, this can occur by genetic changes (adaptation), which trigger mosaic responses in basal and shared neurogenic programs and thereby underly anatomical changes (Rétaux et al., 2013; Hall and Tropepe, 2020), as also demonstrated within mammals (Florio et al., 2015, 2016; Kalebic et al., 2018, 2019; Stepien et al., 2021). These changes can affect early patterning and sub-domain fate decision, therefore proportioning brain regions via directly modulating, the timing of local neurogenesis, cell cycle state and migratory behavior. Examples of the relevance of these shifts for forebrain plasticity in domain size and function were accumulated from comparative studies in teleosts. For instance, even small heterochronic and heterotopic variations in Shh and Fgf signaling doses can determine morphological and functional changes in the development of various forebrain domains in different teleosts. Namely, the expansion of Shh secreted from the midline was causatively linked to the expansion of the diencephalic domain (i.e., of the ventral hypothalamic region) in cave-dwelling cavefish as compared to the surface conspecifics adapted to surface. Of note, species differences also at the level of neuropeptidergic cell clusters (e.g., an increase in orexigenic NPY or Hcrt producing cells), controlling behavioral adaptations such as locomotion likely for foraging needs, were sustained in cave-adapted fish (Menuet et al., 2007; Alié et al., 2018). Again, evidence exists also in sand and rock-dwelling cichlids that similar variations in competing Shh and Wnt signaling expand or restrict the dorsal telencephalon (pallium). Specifically, an early function of Wnt inhibitors in development produces a larger pallium in sand dwellers fish as compared to the rock counterparts. In the latter on the other side expansion of Shh signaling and blocking Wnt inhibitors produce an extended subpallium. The potential modulatory effect of changes in these signalings on telencephalon development was further demonstrated by manipulating them in zebrafish (Rétaux et al., 2013; Sylvester et al., 2013).

When considering teleost fish as neurodevelopmental and disease models, one should also consider that the enrichment of progenitor cell types, their proliferative potential, the migratory routes of the newborn neurons resulting in the peculiar laminar organization of the neocortex are distinctive features of mammals, further elaborated in primates and in some gyrencephalic animals. Nevertheless, despite the differences discussed before, conservation of the founder precursor cells and the core mechanisms of neurogenesis exists in teleost telencephalon, as well as the functional competence derived from it (Portavella et al., 2002; Salas et al., 2003; Wilson and Houart, 2004; Nieuwenhuys, 2009; Rakic, 2009a; von Trotha et al., 2014; Porter and Mueller, 2020). Specifically, when it comes to the compendium of cell types, RG cells with proliferative potentials exist in various teleost’s brain regions, including the telencephalon. Debate exists on whether radial migration occurs during pallium development in these fish. Interestingly, (Mueller et al., 2011) described proliferative (Brdu +) cells migrating from the medial proliferative area of the dorsal telencephalon to the lateral and posterior area of the pial surface in developing zebrafish forebrain between 3 and 8 days post-fertilization. Similarly, photoconversion experiments in zebrafish provide evidence of a migratory cell population moving from the proliferative VZ of the developing pallium to the lateral wall during expansion of the telencephalic domain between 2 and 5 days of development (Folgueira et al., 2012). However, this behavior was not confirmed by later birth-dating experiments (Furlan et al., 2017); while findings of potentially migratory RG-like cells have been recently proposed also in cichlids (Mack et al., 2021). Noteworthily, conversely to mammalian species, teleost RG cells conserve the ability to proliferate and give rise to newborn neurons that integrate into existing circuits, in various neurogenic areas of the brain also in adulthood and in response to injury and disease. This feature defines the distinctive regenerative potential of the teleosts brain (Zupanc and Clint, 2003; Strobl-Mazzulla et al., 2010; Baumgart et al., 2012; Lange et al., 2020) and was also described for other non-mammalian vertebrates (García-Verdugo et al., 2002). Of note, the use of transgenic lines in zebrafish allowed also the real time visualization of migratory interneurons from the developing ventral to the dorsal telencephalon (Mione et al., 2008), resembling the saltatory migratory behavior observed in mammals and other vertebrates (Bellion et al., 2005; Laclef and Métin, 2018). In addition, other newborn neurons migrating out of the proliferative zones have been found in various sites of the adult zebrafish brain. Tangential migration of interneurons from ventral telencephalon was documented, which resembles rodent and tetrapods rostral migratory stream populating the anterior-most part of the telencephalon, the olfactory bulb. Altogether the available data seem to support a possible conserved mechanism of dual origin and migratory behaviors contributing to telencephalon cell types, and highlight zebrafish as an interesting model to study and visualize the mechanisms of these processes not invasively (Doetsch et al., 1997; Byrd and Brunjes, 2001; Grandel et al., 2006; Alvarez-Buylla et al., 2008; Mione et al., 2008; Ganz et al., 2012). Recent work demonstrates also the utility of alternative teleost models (such as killifish) besides zebrafish to dissect the impact of divergent and adaptive mechanisms at the level of precursor cells and neurogenesis and their function in sustaining pallium growth in vertebrates (Coolen et al., 2020).

To fully exploit the unique features of zebrafish (and more generally teleosts) for functional validation and classification of rapidly emerging heterogeneous MCD in humans, experimental designs, questions, and data interpretation in these models should be carefully confined by the limitations discussed. In translational workflows dedicated to solving the underlying biology of MCD in vivo zebrafish models should be complementary to other feasible alternatives closer to large mammals and humans (Figure 1). Of particular interest in this sense and complementing the popularity of gyrencephalic mammals, is the establishment of in vitro human models of neural proliferation and differentiation using induced pluripotent stem cells (iPSCs), as well as iPSCs-derived COs for better recapitulating the structural and functional complexity of the human brain (Figure 1 and Table 1). These advances have been possible thanks to the studies of the early 20th century on neural induction from tissues of animal models and capitalize on the cell-autonomous competence of progenitor stem cells, especially in organizing anterior neural tissue (Lancaster et al., 2013; Paşca et al., 2015). iPSCs can be obtained from somatic cells by induction of known factors that reprogram the adult tissue toward different cell/tissue fate, including cortical neurons. COs have been more recently established (Lancaster et al., 2013) as 3D-self-oganizing cerebral tissue cultures derived from stem cells (i.e., iPSCs) and can be useful to study all the different processes of cortex formation, including interneurons migration (Lancaster et al., 2013).

The ontogeny, morphogenesis, and structural/functional identity of the forebrain domains, and especially of the pallium, from which the cortex and neocortex originated, arose via divergent paths in vertebrates which likely represent key events for multiple adaption strategies throughout evolution. Clearly, the neocortex and its expansion are only visible in a subset of mammals. Its radial construction is achieved via a delicate series of events and precursor cell types, whose proliferative potential and neurogenic timing most likely shapes the differences observed in size and complexity across mammals. The alteration of the fine-tuning of these processes leads to rare conditions with heterogeneous causes and severe clinical consequences. Rapid protocols for effective modeling of these conditions, which are increasingly being reported, are pivotal to tackling the underlying mechanisms and prompt patient care. On the other hand, increasing experimental evidence from the field of comparative neurobiology shows that the teleost brain, considered “simple” for long time, exhibits high cognitive functions, some of the underlying core cell types, development strategies and circuits which are closely related to those build within mammalian pallium derivatives. However, analysis of the extent of similarity and shared ancestry between the pallium ontogeny, morphology, and function in teleosts and amniotes (and gyrencephalic) mammals just began. From a technical point of view, compared to classical and innovative mammalian models, in both the context of brain EvoDevo and disease, teleosts offer a distinctive readiness for generating affordable models with a broad and fast interrogation capacity of nano-scale dynamics within precursor cells and up to neuronal circuits logics of entire brain territories at the whole-organism scale. Furthermore, quantifiable behavioral readouts can be performed simultaneously, creating an ideal setting for scalable preclinical drug testing. Specifically, for disease modeling, limitations exist, however, when researching biological processes which could be altered in MCD using zebrafish and other teleosts, given the differences discussed so far and the clear lack of a bona fide cortex in this species. Not with understanding these differences, fish have indeed already demonstrated their usefulness in speeding up functional validations of new MCD-causing variants and disease genes. In this context, they can often model basal defects in brain/forebrain development, at the level of proliferation, migration and connectivity, which recapitulate those found in patients (Table 2). It is without a doubt that zebrafish work has already benefited our understanding of rare neurodevelopmental diseases. However, the speed by which teleosts are being used, calls for a careful interpretation of the data obtained in these models, which should be assessed against complementary systems more closely resembling human cortical features. Mammalian 2D and 3D neuronal cultures -especially if patient-derived- begin to provide interesting possibilities. We are at the beginning of a new era for brain disorder modeling. Understanding the extent to which small vertebrate fish and 3D organoid cultures can be used as high-fidelity/informative systems recapitulating human brain development is compulsory. As shown, examples of successful in vivo fish models for MCD exist, which expand the possibilities offered by COs and validate the use of teleost fish in the translational research of MCD. Complementary studies combining “fish and dish”-based brain models are accumulating, which demonstrate how they may be extremely compelling to dissect different aspects of human rare disorders (Ayala-Nunez et al., 2019; Pini et al., 2020) and to counterbalance possible pitfalls of the single models (Hengel et al., 2020). A robust integration of patients’ genomic sequencing and functional investigations in zebrafish, iPSCs and organoids exhibiting diverse and unbeatable advantages promises to be a simplified yet efficient multimodal paradigm of contemporary translational research on MCD.

AL conceived the manuscript and generated the illustration. AL, GF, and CC wrote the manuscript. MT and BD revised the final manuscript. GF and AL conceived and generated the tables. All authors contributed to the article and approved the submitted version.