Ribosomal biogenesis as an emerging target of neurodevelopmental pathologies.

Abstract

Development of the nervous system is carried out by complex gene expression programs that are regulated at both transcriptional and translational level. In addition, quality control mechanisms such as the TP53-mediated apoptosis or neuronal activity-stimulated survival ensure successful neurogenesis and formation of functional circuitries. In the nucleolus, production of ribosomes is essential for protein synthesis. In addition, it participates in chromatin organization and regulates the TP53 pathway via the ribosomal stress response. Its tight regulation is required for maintenance of genomic integrity. Mutations in several ribosomal components and trans-acting ribosomal biogenesis factors result in neurodevelopmental syndromes that present with microcephaly, autism, intellectual deficits and/or progressive neurodegeneration. Furthermore, ribosomal biogenesis is perturbed by exogenous factors that disrupt neurodevelopment including alcohol or Zika virus. In this review, we present recent literature that argues for a role of dysregulated ribosomal biogenesis in pathogenesis of various neurodevelopmental syndromes. We also discuss potential mechanisms through which such dysregulation may lead to cellular pathologies of the developing nervous system including insufficient proliferation and/or loss of neuroprogenitors cells, apoptosis of immature neurons, altered neuronal morphogenesis and neurodegeneration.

I. Introduction.

Intrinsic or extrinsic factors such as gene mutations, infectious agents or neurotoxins may perturb structural and/or functional development of the nervous system and/or induce neurodegeneration in childhood. The spectrum of neurodevelopmental disruption phenotypes includes brain malformations such as neural tube defects (NTD) or microcephaly, psychiatric neurodevelopmental disorders of brain function (NDD) such as intellectual disability (ID), autism spectrum disorders (ASD), or schizophrenia, various forms of epilepsy as well as cerebral palsy/periventricular leukomalacia (Ernst 2016)(Thapar et al. 2017). While such heterogeneity would dictate different clinical strategies to each group of the broadly defined neurodevelopmental pathologies (Thapar et al. 2017), a reductionistic approach has been proposed to classify them based on cellular pathogenesis including disruptions of neural stem cell (NSC)/neuroprogenitor cell (NPC) proliferation- or altered migration and differentiation of relevant neural cell types such as NSC/NPC, early neuron or glia (Ernst 2016).

For example, congenital microcephaly is usually a manifestation of insufficient proliferation, pre-mature differentiation or death of NPCs and/or death of newly generated neurons (Gilmore & Walsh 2013, Seah et al. 2008, Mao et al. 2015). In turn, aberrant differentiation of post-mitotic neurons that results in too few, too many or misdirected connections in various brain circuitries is believed to be the main cellular mechanism behind functional deficits including NDDs or epilepsy (Ernst 2016). Finally, acute or persistent damage to the newly formed but not completely mature nervous system may trigger neurodegeneration that causes neurodevelopmental regression as exemplified by Rett syndrome (RTT) or spinal muscular atrophy (SMA) (Chahrour & Zoghbi 2007, Lunn & Wang 2008). The reductionistic hypothesis of unifying cellular pathologies would explain often observed co-occurrence of brain malformations and/or neurodegeneration together with functional deficits (Ernst 2016, Chahrour & Zoghbi 2007).

Highly regulated gene expression programs are primary drivers of development including neurodevelopment. Such programs are controlled at both transcriptional and translational level and their disruption results in various neurodevelopmental syndromes (Tebbenkamp et al. 2014, Sahin & Sur 2015). Indeed, mutations that affect translational control of mRNAs result in at least some forms of those pathologies (Sahin & Sur 2015, Sartor et al. 2015).

Translation is carried out by ribosomes. Although traditionally they were viewed as passive executioners of protein synthesis, their regulatory role is increasingly recognized. For instance, ribosomal biogenesis-mediated ribosome diversification supports specialization in specific translational events including those that drive mammalian development (Barna 2013). Furthermore, size of the cellular ribosome pool may have mRNA-specific effects on translation efficiency (Mills & Green 2017).

The nucleolus-based ribosomal biogenesis is a resource-intensive process that includes transcription and processing of large amounts of rRNA as well as transcription, translation and intra-nuclear transport of corresponding quantities of ribosomal proteins (RPs) (Drygin et al. 2010, Henras et al. 2015, Kressler et al. 2017, Sharifi & Bierhoff 2018). Such a complex process that spans a significant portion of cellular biosynthetic activity is monitored by cellular stress signaling including TP53 that can detect its disturbances and activate pro-homeostatic or cytotoxic mechanisms in response (Deisenroth et al. 2016, Pelava et al. 2016). In addition, dysregulated transcription of rRNA is linked to perturbations of genomic integrity and/or activation of the DNA damage response (Lindstrom et al. 2018). Thus, neurodevelopment-disruptive stimuli may engage many effector mechanisms of neuroteratogenicity when targeting ribosomal biogenesis. Aberrant translational regulation may be just one of such mechanisms. Indeed, the potential for patho-mechanistic diversity of dysregulated ribosomal biogenesis may correspond to heterogenous pathogenesis and presentation of neurodevelopmental syndromes.

In this review we will discuss emerging evidence that supports a role of aberrant ribosomal biogenesis in various forms of neurodevelopmental pathologies that range from microcephaly syndromes through NDDs to pediatric neurodegenerative diseases. Although all consequences of dysregulated ribosomal biogenesis may play a pathogenic role throughout that spectrum, current evidence indicates a potential role for ribosomal stress (RS) in pathogenesis of brain malformations such as microcephaly or translational dysregulation in pathogenesis of functional deficits and neurodegeneration. Finally, disruption of genomic integrity due to dysregulated synthesis of ribosomes has been implicated in two pediatric neurodegeneration syndromes.

II. General features of the ribosomal biogenesis pathway

Ribosomes are the principal components of the protein synthesis machinery. Their synthesis is a highly regulated process that requires coordinated synthesis and assembly of two distinct ribosomal subunits made of 4 species of rRNA and 79 ribosomal proteins (RPs) (Henras et al. 2015, Kressler et al. 2017). In human cells, there are also more than 300 ribosomal biogenesis factors (RBFs) that are required for ribosomal biogenesis in trans (Tafforeau et al. 2013). Most steps of ribosomal biogenesis occur in the nucleolus which contains hundreds of copies of the 47S pre-rRNA gene (rDNA) (Sirri et al. 2008). In most cells, about half of these genes are transcriptionally active while the other half is epigenetically silenced (Sharifi & Bierhoff 2018). This transcriptionally inactive rDNA plays a structural role in chromatin organization and maintenance of genomic integrity (Lindstrom et al. 2018). Transcription of active rDNA genes by the RNA-Polymerase-1 (Pol1) initiates ribosome production (Sharifi & Bierhoff 2018). That process culminates with a largely nucleolar assembly of the 40S small ribosomal subunit (SSU) and the 60S large ribosomal subunit (LSU) (Henras et al. 2015, Kressler et al. 2017). Ultrastructurally, the nucleolus has tri-partite organization. While the Pol1-mediated transcription occurs in the fibrillar center (FC) of the nucleolus, rRNA modifications and ribosome assembly take place the dense fibrillary (DFC)- and granular component (GC), respectively (Sirri et al. 2008). As a membraneless structure, the nucleolus is highly dynamic allowing for rapid changes in ribosome synthesis activity and regulation of subcellular localization of various proteins. Being a major cellular RNA center and manufacturing hub for protein synthesis machinery, as well as the most active transcriptional site in the nucleus, the nucleolus has acquired additional functions beyond ribosomal biogenesis. Such extra-ribosomal functions include cell cycle regulation, cellular stress response, generation of non-ribosomal ribonucleoprotein complexes including the signalosome or the splicosome, and, finally, maintenance of chromatin structure and genomic integrity (Tsai & Pederson 2014, Pederson & Tsai 2009, Lindstrom et al. 2018).

Importantly, as structural integrity of the nucleolus is directly related to activity of Pol1 as well as availability of various RBFs and/or RPs, these additional functionalities of the nucleolus do not operate independently of ribosome biogenesis (Nicolas et al. 2016, Langhendries et al. 2016). Instead, they are often linked to ribosomal biogenesis as exemplified by the RS pathway (Deisenroth et al. 2016, Pelava et al. 2016) (Fig 1). In this case, any disturbance in ribosomal biogenesis activity including its inhibition or out-of-context activation of a single component/stage of ribosome production results in stabilization of the stress response transcription factor TP53 (first reports of that pathway include (Pestov et al. 2001), (Rubbi & Milner 2003) and (Zhang et al. 2003)). Importantly, the RS response is unrelated to shortage of ribosomes (Deisenroth et al. 2016, Pelava et al. 2016). Mechanistically, various components of the ribosome that are no longer used for ribosome assembly bind to and inhibit the TP53-destabilizing ubiquitin ligase MDM2/HDM2 (Deisenroth et al. 2016, Pelava et al. 2016) (Fig. 1). The 5S ribonucleoprotein particle (5S RNP) that contains 5S rRNA as well as RPL11 and RPL5 is the most critical mediator of MDM2/HDM2 inhibition. In addition, besides the MDM2/HDM2-TP53 pathway other effectors of RS may include inhibition or pro-growth proteins such as MYC or MTOR (James et al. 2014) as well as activation of autophagy (Katagiri et al. 2015).

Figure 1. The ribosomal stress (RS) pathway and its hypothetical role in neurodevelopmental syndromes.

Dysregulated ribosomal biogenesis results in molar imbalances of ribosomal components. In effect, the 5S ribonucleoprotein (5S RNP) that includes RPL11, RPL5 and 5S rRNA, instead of being used for 60S LSU production, binds to MDM2/HDM2 inhibiting its ubiquitin ligase activity towards TP53. TP53 accumulates activating apoptosis or cell cycle arrest. Besides 5S RNP, also other RPs or RBFs may regulate TP53 stabilization and/or activity. The RS pathway may contribute to neurogenesis failure and microcephaly as well as hypomyelinating leukodystrophies (see text for more details).

Although the RS is often referred to as “nucleolar stress” in the literature and as some inducers of RS disrupt structural integrity of the nucleolus one should note that nucleolar disruption at the morphological level is not required for RS (Fumagalli et al. 2009, Slomnicki et al. 2018). Instead, functional perturbations of ribosomal biogenesis appear to be the most important stimulus for RS. Hence, RS may be accompanied by decreased nucleolar GC as seen in cases when Pol1 activity is blocked, increased GC as found when rRNA processing is perturbed or no obvious changes in nucleolar morphology ((Slomnicki et al. 2018) and references therein).

One should also note that various pathways that regulate cell proliferation and differentiation including the MYC, MTOR or the ERK1/2 MAP kinase directly modulate ribosomal biogenesis at multiple levels (Iadevaia et al. 2014, Kusnadi et al. 2015). Such a regulation may control not only the ribosome number but also lead to production of specialized ribosomes that will help execute a specific aspect of a gene expression program in cells that respond to an extracellular stimulus. Such a ribosomal specialization may originate from differential use of RPs, as demonstrated for RPL38 which facilitates ribosome binding of a homeobox transcription factor mRNA, promotes its translation and supports formation of somites during mouse development (Kondrashov et al. 2011) (Fig. 2 and Fig. 3). However, just changing ribosome quantity rather than quality may have profound consequences on mRNA translation rate (Fig. 2 and Fig. 3). Indeed, as translation initiation rates are different across the mRNA universe reduced ribosomal abundance is expected to produce relatively selective changes in translation with maximal impairment of those mRNAs whose translation is difficult to initiate (Mills & Green 2017)( Fig. 2 and Fig. 3).

Figure 2. Hypothetical effects of ribosomal availability on translation of various neuronal mRNAs.

As many housekeeping proteins are translated from mRNAs with highly efficient translation initiation, their expression is relatively resistant to changes in the size of the cellular ribosome pool. Conversely, those mRNAs, whose translation initiation rates are low due to presence of secondary structures in their 5’UTRs (SS) or upstream open reading frames (uORF), will be translated efficiently only under optimal availability of ribosomes and when appropriate stimuli reduce inhibitory effects of SSs or uORFs. In addition, translation of some mRNAs may require specialized, de-novo produced ribosomes (yellow). Those mechanisms regulate (or, in case of specialized ribosomes, may regulate) expression of proteins that play a role in neuromorphogenesis and/or synaptic functions. Therefore, ribosome availability would be critical for establishment and maintenance of neuronal connectivity (see text for more details).

Figure 3. Translational dysregulation as a hypothetical mechanism of neurodevelopmental defects that are associated with reduced ribosomal biogenesis.

Defects in ribosomal biogenesis decrease the ribosome pool of a cell. Expression of many translationally regulated proteins including those that are required for morphogenesis of dendrites and synapses as well as synaptic function will be reduced. Hence, decreased ribosome synthesis may perturb development, maintenance and/or function of dendrites and synapses leading to ID/ASD-associated changes in neuronal connectivity and/or neurodegeneration (see Fig. 2 and text for more details).

While reduced rDNA transcription is sufficient to instigate DNA damage response (Quin et al. 2016, Scheibye-Knudsen et al. 2016), excessive rDNA transcription may also lead to DNA double strand break (DSB) damage (Ide et al. 2010, Toro et al. 2018) (Fig. 4). Moreover, at least in yeast, Pol1-driven transcription promotes DSBs in rDNA itself (Ide et al. 2010). As rDNA consists of repeated units, DSBs in this region of the genome are highly recombinogenic (Kobayashi 2011). Such a recombination and the associated rDNA loss may be detrimental to chromatin organization on a global scale as inactive (i.e. non-transcribed) copies of rDNA are thought to play an important structural role in chromatin cohesion and heterochromatin maintenance throughout the nucleus (Lindstrom et al. 2018). Indeed, inducing rDNA DSBs in fruit flies resulted in rDNA recombination, rDNA loss, heterochromatin deficits and global changes in gene expression with mitochondria-associated genes being most affected (Paredes et al. 2011, Paredes & Maggert 2009). Thus, both the number of active rDNA copies and their transcription rates must be tightly regulated to ensure integrity of both the genome and the epigenome.

Figure 4. Transcriptional dysregulation of rDNA and activation of the DNA damage response.

Reduced transcription of rDNA such as that observed in deficiencies of ERCC6/8 (Cockayne syndrome) stabilizes G quartet (GQ)structures in the rDNA promoter. Increased presence of GQs initiates DNA damage response including activation of PARP, PARP activity-mediated depletion of its substrate NAD⁺, increased oxidative metabolism in mitochondria, mitochondrial damage and oxidative stress. The resulting oxidative damage would lead to neurodegeneration and loss of OPC/OLs. In addition, decreased translational fidelity (presumably due to reduced ribosome pool renewal and accumulation of dysfunctional ribosomes with various types of oxidative damage as indicated by red and black dots) may worsen oxidative protein damage and lead to ER stress (see text for more detail).

Deficits of ribosome synthesis underlie several human genetic diseases that are known as ribosomopathies. The most common of them is Diamond Blackfan anemia that is caused by heterozygous loss of function mutations in several RP genes (Danilova & Gazda 2015, Ellis 2014). As a result, declining ribosome abundance reduces translation of mRNAs with low translation initiation efficiency including the master regulator of erythropoiesis, GATA1 (Ludwig et al. 2014). In addition the RS pathway contributes to defective erythropoiesis by perturbing hematopoetic stem cell growth and survival in a TP53-dependent manner (Jaako et al. 2011). That mechanism was also shown to play a major role in pathogenesis of skeletal malformations in an animal model of another ribosomopathy, the Treacher Collins syndrome (TCS), where hemizygous loss of function mutations perturb Pol1 activity (Trainor & Merrill 2014).

III. Consequences of dysregulated ribosomal biogenesis in the developing nervous system.

III.1. Dysregulation of ribosomal biogenesis in NPCs.

The nucleolus is prominently present in both NPCs and neurons (Hetman & Pietrzak 2012, Slomnicki et al. 2017). In many types of rapidly dividing cells efficient ribosomal biogenesis is required for proliferation by supplying ribosomes and suppressing RS (Tsai & Pederson 2014, Kusnadi et al. 2015). The transcription factor MYC that is essential for proliferative growth is a positive transcriptional regulator of ribosomal biogenesis genes as well as rRNA synthesis (Grandori et al. 2005, Poortinga et al. 2011). As compared to post-mitotic neurons, NSC/NPCs have relatively highly active ribosomal biogenesis that is likely supported by a MYC relative MYCN, which by itself is required for neurogenesis (Knoepfler et al. 2002, Boon et al. 2001). However, during early neurogenesis a MYC downregulation-mediated shift towards lower rate of ribosome synthesis associates with appearance of NPCs and interfering with that shift resulted in macrocephaly (Chau et al. 2018). Hence, sequential downregulation of ribosomal biogenesis may be required for a transition from pluripotency to increasingly differentiated types of neural cells.

A neurogenic requirement of ribosome synthesis is supported by generation of acephalic mice following NPC-selective knock out of the Pol1 co-factor TIF1A (Parlato et al. 2008). In the ventricular zone of mouse brain embryo Pol1-deficient NPCs displayed structural disruption of the nucleolar GC, accumulation of TP53 and apoptosis. These changes are similar to those observed after the NPC-specific knock out of MDM2 suggesting a role for RS in neurogenesis failure (Xiong et al. 2006) (Fig. 1). Obviously, while complete inhibition of Pol1 led to pro-apoptotic RS, one could expect that partial inhibition of ribosomal biogenesis that would be insufficient to induce RS-mediated apoptosis may also impair NSC/NPC proliferation and/or induce their premature differentiation. Indeed, high rates of ribosome synthesis are required for rapid cell divisions (Tsai & Pederson 2014) while mild RS may be sufficient to induce G1 cell cycle checkpoint without apoptosis (Kumazawa et al. 2015) (Fig. 1). Moreover, inhibition of Pol1 may induce stem cell differentiation (Sharifi & Bierhoff 2018). In either case, insufficient number of neurons would be generated leading to microcephaly. Hence, dependent on intensity/cellular extent of ribosomal biogenesis impairment in NSC/NPCs, anencephaly or microcephaly of varying severity could be expected. Conversely, in those cells, delaying differentiation-associated downregulation of ribosomal biogenesis may result in macrocephaly.

III.2. Dysregulation of ribosomal biogenesis in neurons.

Post-mitotic neurons contain large quantities of ribosomes and rely on regulated translation to establish and maintain their phenotypic characteristics including morphogenesis of neurites and synapses as well as synaptic plasticity ((Slomnicki et al. 2016b) and references therein). Cell soma ribosomes accumulate in large numbers during neuronal maturation (Slomnicki et al. 2016b). Moreover, some ribosomes also appear in extending neurites. Such an expansion of the neuronal ribosome pool occurs despite sharp downregulation of ribosomal biogenesis that is observed after completion of neurogenesis (Slomnicki et al. 2016b). Lack of cell division-associated ribosome dilution combined with a relatively long life of a ribosome (i.e. days-weeks) provide a likely explanation of that apparent paradox (Stoykova et al. 1983, Price et al. 2010).

As in cycling cells, neuronal ribosome synthesis is under surveillance of the RS pathway. Its activation occurs in response to either nucleolus-disrupting inhibition of Pol1 (Kalita et al. 2008, Parlato et al. 2008) or knockdowns of RPs that do not perturb nucleolar structure (Slomnicki et al. 2018). In either case, activation of TP53 follows. In immature neurons that retain ability to launch the TP53 –mediated apoptosis (Wright et al. 2007, Martin et al. 2008, Pietrzak et al. 2011), RS activates this form of programmed cell death (Kalita et al. 2008, Slomnicki et al. 2018). RS-mediated neuronal apoptosis is attenuated when the 5S RNP component RP L11 is knocked down (Slomnicki et al. 2018). Therefore, one could expect that in newly generated neurons, RS and the subsequent apoptosis could contribute to microcephaly (Fig. 1). Noteworthy, in fully matured neurons, in which epigenetic silencing of pro-apoptotic genes prevents TP53 from inducing apoptosis (Wright et al. 2007), TP53-mediated gene expression program may be neuroprotective helping to cope with consequences of insufficient ribosome supply (Kreiner et al. 2013).

One should note that pro-microcephalic disruption of ribosomal biogenesis may also involve translation deficits. Thus, many proliferation- or cell survival-associated mRNAs are undergoing translational regulation at the initiation level and, therefore, may be severely affected by reduced abundance of ribosomes in NSC/NPC and/or immature neurons (Chu et al. 2016, Mills & Green 2017).

In the absence of apoptosis, decreased ribosomal biogenesis will result in reduced neuronal ribosome content. In differentiating neurons that increase their ribosome pool, such deficits may arise relatively quickly. For instance, in cultured hippocampal pyramidal neurons from newborn rats, knockdowns of various RPs reduced total ribosomal content by 20–30% in just 3 days (Slomnicki et al. 2016b). Moreover, neuronal ribosome expansion period may be relatively long as a continous increase of ribosomal content was observed in rat cortex from postnatal day 7 to 42 (Slomnicki et al. 2016b). Interestingly, in hippocampal neurons inhibition of ribosomal biogenesis and the resulting deficits in ribosome supply were associated with impairment of dendritic morphogenesis (Gomes et al. 2011, Slomnicki et al. 2016b). In addition, maintenance of immature dendrites was also affected (Slomnicki et al. 2016b). Such defects did not correlate with inhibition of general protein synthesis and were independent of TP53 (Gomes et al. 2011, Slomnicki et al. 2016b). Instead, selective failure to translate mRNAs with inefficient translation initiation (such as GATA1 in hematopoetic cells) may be an important mechanism underlying these growth deficits (Fig. 2 and Fig. 3).

While identity of those mRNAs has to be determined by future experiments, one should mention that many key components of the PI3K/MTOR signaling pathway appear to have highly inefficient translation initiation, at least, in mouse embryonic stem cells (Fujii et al. 2017). Interestingly, reduced activity of the PI3K/MTOR pathway was observed after inhibition of neuronal ribosome synthesis (Slomnicki et al. 2016b) and dysregulation of that pathway is an established cause of both structural as well functional disturbances of neurodevelopment (Sahin & Sur 2015, Switon et al. 2017).

Pol1 is stimulated by the pro-dendritic BDNF-ERK1/2 MAP kinase pathway, and its selective activation accelerates dendritic morphogenesis (Gomes et al. 2011). Consequently, dysregulation of ribosome synthesis in differentiating neurons may affect their potential to form and maintain synapses. Furthermore, as many synaptic functions including synaptic plasticity require regulated translation (Costa-Mattioli et al. 2009), dysregulated ribosome supply may also perturb synapse functionality as demonstrated by impaired long-term potentiation (LTP), learning or epileptic kindling after inhibition of ribosomal biogenesis (Kiryk et al. 2013, Allen et al. 2014, Vashishta et al. 2018). It is tempting to speculate that as reduced number of synapses and/or their impaired function underlies ID or, at least, some forms of ASD (van Spronsen & Hoogenraad 2010), reduced production of ribosomes during and/or after neuronal differentiation could mediate such connectivity phenotypes (Fig. 2 and Fig. 3). In addition, shortage of ribosomes and the subsequent translational defects could result in loss of synapses and neurites leading to neurodegeneration (Fig. 2 and Fig. 3). Finally, overproduction of neuronal ribosomes may lead to excessive connectivity. As a result, excessive ribosomal biogenesis could contribute to epileptogenesis (Vashishta et al. 2018). Furthermore, as at least some forms of ID/ASD may be due to too many synaptic contacts, ribosome oversupply may be a contributing factor to those syndromes especially in cases that are caused by hyperactivation of the ribosomal biogenesis regulator MTOR (Valnegri et al. 2017).

Besides early apoptosis and dysregulated connectivity, aberrant ribosomal biogenesis may lead to chronic neurodegeneration. Such a phenotype has been observed in at least some types of neurons after selective knock out of TIF1A (Rieker et al. 2011, Kreiner et al. 2013). It was accompanied by reduced activity of the MTOR pathway, increased oxidative stress and induction of autophagy (Rieker et al. 2011, Kreiner et al. 2013). It is unclear whether such consequences are primarily caused by deficits in synthesis of specific proteins due to insufficient number of ribosomes and/or lack of specialized ribosomes or by disruption of neuronal proteostasis by general decline in translation fidelity or by loss of other functionalities of the nucleolus. It is interesting to note in this context that in neural cell lines, inhibition of rRNA transcription but not later stages of ribosomal biogenesis activated the DNA damage response, which consequently dysregulated mitochondrial function and led to oxidative stress (Scheibye-Knudsen et al. 2016). Therefore, chronic DNA damage response may contribute to neurodegenerative changes after dysregulation of ribosomal biogenesis in the developing nervous system (Fig. 4).

Taken together, defective ribosomal biogenesis can potentially initiate and/or modulate multiple cyto-pathogenic cascades that are implicated in neurodevelopmental syndromes. Thus, RS-mediated apoptosis may be pro-microcephalic (Fig. 1). Translational dysregulation may underlie pathological changes in neuronal connectivity (Fig. 2 and Fig. 3). Finally, neurodegeneration may involve disrupted chromatin organization and/or the DNA damage response as well as translational deficits (Figs. 3 and 4). The latter mechanism may also play a role in early defects of CNS development including microcephaly.

IV. Neurodevelopmental consequences of genetic defects of ribosomal biogenesis.

Relatively common genetic ribosomopathies including DBA or TCS are usually caused by hemizygous loss of function mutations of RPs or co-factors/components of Pol1, respectively (Danilova & Gazda 2015, Trainor & Merrill 2014). As these conditions are not associated with any major neurological phenotype, neurodevelopment appears to be relatively resistant to partial impairments of ribosomal biogenesis that are sufficient to produce tissue/cell linage-specific phenotypes. Indeed, lowering rate of ribosomal biogenesis during neurodevelopment could explain its relative tolerance of suboptimal supply of ribosomal components due to common hemizygous defects in RP/Pol1 genes (Slomnicki et al. 2016b) (Chau et al. 2018).

Obviously, homozygous loss of function mutations in general ribosomal components or RBFs result in embryonic lethality preventing emergence of neurodevelopmental ribosomopathies (for instance, see (Yuan et al. 2005) or (Hamdane et al. 2014)). However, one could expect neurodevelopmental disruption in a case of homo- or hemizygous loss of function and/or hypomorphic mutations as well as hemizygous gain of function mutations that would dysregulate ribosomal biogenesis/ribosomal function in a relatively tissue-selective manner including the developing nervous system. Such mutations have been discovered in two RP genes as well as several trans-acting RBFs (Table 1). In addition, impaired ribosomal biogenesis in nervous system cells has been proposed as a mechanism underlying neurodevelopmental disruption that is caused by mutations in several genes that encode regulators of transcription and/or RNA metabolism including cohesins, MECP2, LARP7 or SMN1. Many of those mutations share common neurophenotypes suggesting that they perturb ribosomal synthesis in similar cellular targets and at similar developmental stages (Table 2). In most cases, those phenotypes are complex with co-occurrence of various neurodevelopmental abnormalities as well as non-neurological syndromes. In many cases, the complexity of neurological phenotypes would suggest that multiple cytopathogenic cascades are involved. Certainly, as discussed above, ribosomal biogenesis dysregulation could potentially explain such pleiotropic effects on neurodevelopment. Below, we will present some of the more prominent examples of neurodevelopmental syndromes with a likely pathogenic contribution of the dysregulated ribosome synthesis. As some trans-acting RBFs including EMG1, DKC1, EIF4A3 or POLR1C have prominent extra-ribosomal functions which were suggested to participate in pathogenesis of the relevant neurological phenotypes, such factors are not discussed. However, as in those cases, a pathogenic role of deficient ribosome synthesis is not excluded they are listed in Table 2 together with OMIM database ID numbers of the respective syndromes where readers can obtain further information on those gene, their mutations and the resulting phenotypes.

Table 1.

Neurodevelopmental syndromes that are caused by mutations in genes that encode ribosomal proteins or trans-acting RBFs that directly participate in ribosomal biogenesis ¹

Gene (protein)	Ribosomal functions	MUTATION-ASSOCIATED NEURODEVELOPMENTAL SYNDROME (mutation: phenotype, syndrome OMIM# or a reference).	Potential neuropathogenic mechanism(s) 2
Ribosomal Proteins
RPL10 (uL16)	Structural protein of the ribosome that mediates SSU/LSU interactions during translation	AUTISM, SUSCEPTIBILITY TO, X-LINKED 5; AUTSX5 (hypomorphic L206M and H213Q: ASD, #300847) X-LINKED SYNDROMIC MENTAL RETARDATION-35 (hypomorphic A64V or G161S- ID: cerebellar hypoplasia, moderate microcephaly; inactivating K78E: severe X-linked microcephaly, growth retardation, and seizures, #300998)	Dysregulated translation for hypomorphic mutants K78E: Neurogenesis failure due to RS and/or dysregulated translation.
RPS23 (uS12)	Structural protein of the ribosome that participates in codon-anti-codon pairing during translation	BRACHYCEPHALY, TRICHOMEGALY, AND DEVELOPMENTAL DELAY (hypomorphic R67K or F120I: microcephaly, hearing loss, growth deficits, and dysmorphic features, ID/ASD /R67K case/, #617412)	Dysregulated translation and RNA metabolism; RS?
RBFs that are directly involved in rRNA processing
LAS1L (LAS1L)	60S LSU RBF	WILSON-TURNER X-LINKED MENTAL RETARDATION SYNDROME (hypomorphic R415W or A269G: ID, dysmorphic face, hypogonadism, short stature, obesity, #309585) SPINAL MUSCULAR ATROPHY WITH RESPIRATORY DISTRESS /SMARD/ (inactivating S477N, neonatal muscle weakness, early respiratory failure (Butterfield et al. 2014))	Translational deficits? RS?
EXOSC3	rRNA processing towards SSU and LSU components	PONTOCEREBELLAR HYPOPLASIA, TYPE 1B /PCH1B/ (hypomorphicD132A: hypotonia, progressive muscular atrophy, and global developmental delay,: spinal motoneuron degeneration, cerebellar atrophy may also occur with ID and progressive microcephaly and global developmental delay; inactivating G31A or G135E: sever hypotonia/areflexia/paresis at birth, variable microcephaly, early lethality, #614678)	Translational deficits? RS?
RBFs that are directly involved in rRNA transcription
UBTF (UBTF1, UBTF2)	rRNA transcription factor	NEURODEGENERATION, CHILDHOOD-ONSET, WITH BRAIN ATROPHY /CONDBA/ (dominant gain-of-function mutation E210K: neurodevelopmental regression, severe ID, dystonia, chorea, spasticity, parkinsonism, progressive, postnatal microcephaly, #617672)	DSB DNA damage response; translational deficits due to dominant negative effects? RS?
POLR1A	Encodes the main subunit of Pol1, RAP194.	HYPOMYELINATING LEUKODYSTROPHY-LIKE SYNDROME (recessive hypomorphic /?/ S394L: ID, cerebellar ataxia, macrocephaly, cerebral/cerebellar atrophy, white matter atrophy (Kara et al. 2017))	DNA damage response? RS? Translational deficits?
ERCC6 CSB, ERCC8 CSA	Supporting rRNA transcription	COCKAYNE SYNDROME B or A /CSB or CSA/ (over 100 hypomorphic or inactivating mutations of ERCC6 or ERCC8 resulting in CS of varying severity; common manifestations include: impaired growth, postnatal microcephaly, progressive ID, atrophy of the brain white matter, loss of neurons in the cerebellum , brainstem, basal ganglia, cerebral cortex, spinal cord, and dorsal root ganglia, #133540 /CSB/, #216400 /CSA/; (Karikkineth et al. 2017))	DNA damage response and subsequent oxidative stress. ER stress due to impaired translational fidelity and protein misfolding.

¹Table 1 does not include neurodevelopmental syndromes that are caused by mutations in RBFs whose extra-ribosomal functions were suggested to participate in pathogenesis of the associated neurological phenotypes. However, as it is not excluded that defective ribosomal biogenesis provides at least partial contribution to those phenotypes, the relevant genes are listed in Table 2.

²See text for more details and relevant references.

Table 2.

Common neurodevelopmental phenotypes that may be associated with dysregulation of ribosomal biogenesis (↓ or ↑ indicate reduced or increased ribosomal biogenesis/ribosomal function).

Phenotype (presumed primary target cells/processes)	Mutated Genes for RPs, RBFs or other regulators of ribosomal biogenesis (Syndrom OMIM# or reference)		Extrinsic neuroteratogenic factors which may involve neuropathogenic dysregulation of ribosomal biogenesis 1
	neuropathogenic role of dysregulated ribosomal biogenesis is highly likely	at least partial neuropathogenic contribution of dysregulated ribosomal biogenesis is possible
Congenital microcephaly (NPCs and immature neurons/ neurogenesis)	RPL10 (#300847, #300998) ↓ RPS23 (#617412) ↓ cohesins (NIPBL, #122470; SMC1A, #300590; SMC3, #610759; ESCO2, #268300; DDX11, #613398) ↓	EMG1 (211180) ↓ DKC1 (305000) ↓ LARP7 (615071) ↓	DNA damaging agents ↓ Ethanol ↓ Zika virus ↓ 5-Fluorouracil ↓
ID and/or ASD without major and/or progressive neurodegeneration (neurons/ neuronal differentiation including morphogenesis and/or synapse function)	RPL10 (#300847, #300998) ↓ RPS23 (#617412) ↓ LAS1L (#309585) ↓ cohesins (NIPBL, #122470; SMC1A, #300590; SMC3, #610759; ESCO2, #268300; DDX11, #613398) ↓	LARP7 (615071) ↓ EIF4A3 (268305) ↓ 15q11-q13 PWS region (176270, (Leung et al. 2009)) ↓ PTEN (605309,158350, 153480)↑ TSC1/2 (# 191100, 613254)↑	ethanol ↓
Hypotonia, muscle atrophy, paralysis (motoneuron survival)	LASL1(#309585) ↓	EXOSC3 (#607596) ↓ SMN1(#253300) ↓	Poliovirus ↓ Zika virus ↓
Progressive degeneration of cortical neuron dendrites without loss of neuron cell bodies, ID/ASD (neurons/dendrite maintenance and synaptic function)	MECP2 (#312750) ↓	not kown	not kown
Progressive grey matter degeneration including postnatal microcephaly (brain neurons) and widespread loss of neurons throughout the nervous system	ERCC6/8 (#216400 and 133540) ↓ RPL10 (K78E, #300998) ↓ UBTF (#617672) ↓	EXOSC3 (#607596) ↓	Zika virus ↓ Cerebral folate deficiency (↑-?)
Hypomyelination and/or leukodystrophy (oligodendrocyte precursor cells and/or mature oligodendrocytes/survival and cell body/myelin maintenance)	ERCC6/8 (#216400 and 133540) ↓ POLR1A (Kara et al. 2017) ↓	POLR1C (#616494) ↓ POLR3A (#607694) ↓	5-Fluorouracil ↓ Cerebral folate deficiency (↑-?) Zika virus ↓

¹See text for references (Section IV)

IV.1. Neurodevelopmental disruption associated with mutations in RPL10 and RPS23.

RPL10 is X-linked and encodes RPL10 (uL16) of the large ribosomal subunit that likely participates in core ribosomal functions such as interactions between the small and the large subunits during late stages of translation initiation as well as activity of the peptidyl transferase center (Ben-Shem et al. 2011, Klinge et al. 2011). Two missense mutations in the C-terminal region of RPL10 (L206M and H213Q) were identified in three families with ASD-affected male children (Klauck et al. 2006). In yeast complementation studies, these mutants were able to rescue loss of yeast rpl10. However, general translation activity was reduced as compared to wt RPL10 suggesting hypomorphic nature of these mutations. Analysis of lymphoblastoid cell lines that were derived from the affected individual revealed very limited changes to their proteomes (Chiocchetti et al. 2014). A hemizygous RPL10 mutation that mapped to the N-terminal region (A64V) was identified in two related male patients with ID and cerebellar hypoplasia (Zanni et al. 2015). That mutation was shown to increase general protein synthesis in the mutant yeast rpl10 complementation assay. Such findings that indicate limited correlation between neurological phenotypes of RPL10 mutants and effects on general protein synthesis may indicate that their main neuropathogenic effects are mediated by selectively dysregulated translation of distinct mRNAs in cell type-specific context(s). Such a possibility is further supported by relatively higher levels of Rpl10 mRNA in the mouse hippocampus or in the developing brain of zebrafish (Allen Brain Atlas) (Brooks et al. 2014).

Another N-terminal mutation of RPL10, K78E was found in three related familial cases of X-linked ID, severe microcephaly with seizures and growth retardation (Brooks et al. 2014). Progressive nature of microcephaly suggests a neurodegenerative component in K78E-associated neurodevelopmental disorder. It has been proposed that the K78E phenotype was a result of complete loss of function of RPL10 as this basic to acidic amino acid residue substitution was predicted to disrupt interactions between RPL10 and 28S rRNA (Brooks et al. 2014). Indeed, in zebrafish, knockdown of RPL10 resulted in microcephaly-like phenotype and reduced general translation specifically in the developing brain (Brooks et al. 2014). Moreover, those effects were rescued by wt RPL10 but not the K78E mutant. Two additional mutations in RPL10 were associated with a syndromic X-linked ID including A64V and G161S (Zanni et al. 2015, Thevenon et al. 2015). Additional phenotypic manifestations included mild microcephaly, cerebellar hypoplasia and delayed neurodevelopment. Less severe phenotypes than those of the K78E mutation indicate hypomorphic character of A64V or G161S RPL10 variants. Indeed, the A64V RPL10 is able to rescue complete loss-of-function (LOF) mutants of yeast RPL10 (Zanni et al. 2015).

It has been proposed that relatively higher expression of RPL10 in the developing brain argues for enrichment of the RPL10-containing ribosomes in that structure and its greater translational dependence on such ribosomes. Indeed, nucleolar enrichment of RPL10 was detected in brain cell nuclei from 7 day old rat pups (Slomnicki et al. 2016a). However, observations of increased apoptosis in brains of RPL10-depleted fish embryos suggest that besides translational deficits, RS may also contribute to the RPL10 deficiency-associated microcephaly (Brooks et al. 2014). Thus, hypomorphic or complete loss of function mutations of RPL10 may produce different neurological phenotypes by presumably different cytopathogenic mechanisms. One can speculate that the hypomorphic mutants lead to translational deficits of specific mRNAs that play a role in development of neuronal connectivity and/or synaptic function. The complete LOF could, in turn, involve more severe and widespread translational deficits that would also affect expression of neurogenesis genes and/or instigate pro-microcephalic RS that would be followed by TP53-mediated apoptosis.

Two distinct missense mutations of RPS23 were found in two unrelated individuals with microcephaly, hearing loss, growth deficits, and dysmorphic features (Paolini et al. 2017). In addition, one patient presented ID/ASD. RPS23 (uS12) participates in biogenesis of the small ribosomal subunit and, as its component, contributes to translation fidelity by monitoring accuracy of the tRNA anti-codon pairing with mRNA (O’Donohue et al. 2010, Stark et al. 2002, Sharma et al. 2007). Indeed, both mutations (R67K and Phe120Ile) affected ribosomal biogenesis reducing supply of the 40S SSU (Paolini et al. 2017). However, at least with one of those mutants (R67K from the ID/ASD case) general protein synthesis was unaffected and there was no deficit in cell proliferation (Paolini et al. 2017). These findings suggest ability of the mutant RPS23 to mediate translation. Furthermore, despite R67K-related deficit in ribosomal biogenesis, RS-associated cell cycle checkpoints did not appear to be activated.

Interestingly, both mutants enhanced formation of oxidative stress-induced stress granules and reduced translation fidelity as revealed using reporter constructs for missense codon suppression or stop codon read-through. Hence, RPS23 mutations may be pathogenic by changing mRNA metabolism under stress and/or disrupting translation fidelity. Whether neurodevelopmental phenotypes arise from such changes and whether they are wide spread across mRNAs or relatively mRNA-selective remains to be determined. Also, role of RS in neurodevelopmental manifestations of RPS23 mutations is unclear as lymphoblastoid cells that were used to determine effects of the R67K mutation on RS may have a different threshold for cytotoxic RS than NPCs or immature neurons.

IV.2. Transcriptional dysregulation of ribosomal biogenesis in cohesinopathies and RTT.

Dysregulation of ribosomal biogenesis at the transcriptional level is a common element for several mutation-associated neurodevelopmental syndromes. Similarly to DBA, de-novo heterozygous loss-of-function mutations in the cohesin complex components SMC1A and SMC3 or their regulators NIPBL and ESCO2 lead to most cases of Cornelia DeLange syndrome and/or Roberts syndrome which besides mental retardation/ID with or without microcephaly may also include facial and limb dysmorphia as well as growth restriction (OMIM #122470 and 268300) (Zakari et al. 2015). Mechanistically, pathogenesis of cohesinopathies has been proposed to result from impaired gene expression due to reduced chromatin organization function of cohesins (Zakari et al. 2015). In support of that notion, yeast and/or patient-derived cellular models of cohesinopathies presented structural changes of nucleoli, impaired rRNA transcription, and lower general protein synthesis leading to reduced cell growth (Gard et al. 2009, Bose et al. 2012, Harris et al. 2014). Such deficits were proposed to be caused by disorganization of nucleolar chromatin whose high transcriptional activity requires cohesin function (Gard et al. 2009, Harris et al. 2014). Interestingly, as in other cases of impaired ribosome supply, cellular human- and zebrafish models of cohesinopathies presented with lower MTOR activity and activation of TP53 (Xu et al. 2013, Xu et al. 2015). Treatment with L-leucine increased MTOR signaling leading to improved protein synthesis including RP translation but did not suppress TP53 accumulation (Xu et al. 2013, Xu et al. 2015, Xu et al. 2016). However, it was able to partially rescue ribosomal biogenesis and translation deficits while improving proliferation and reducing apoptosis; defects in fish embryo development were also attenuated (Xu et al. 2013, Xu et al. 2015). These findings indicate that loss of pro-ribosomal activity of cohesins could be at least partially compensated by stimulating other components of the positive regulatory control of ribosomal biogenesis. While it remains to be determined whether cohesins play a similar, pro-ribosomal role in neurons, these proteins were found enriched in the nucleolar proteome from the developing rat brain (Slomnicki et al. 2016a). Therefore, at least some neurological symptoms of cohesinopathies, including ID and/or microcephaly, may be caused by insufficient ribosome supply and translational deficits that affect neuronal connectivity as well as RS-mediated impairment of neurogenesis.

RTT is among the more common neurodevelopmental diseases, and, is usually caused by sporadic LOF mutations in the X-linked gene MECP2 resulting in neonatal male lethality and females that present with postnatal microcephaly, ID/ASD, and, progressive degenerative changes in the cerebral cortex including loss of dendrites (Chahrour & Zoghbi 2007). Hence, neuronal degeneration rather than impaired neurogenesis underlie RTT pathology. As MECP2 binds to methylated CpGs that are recognized as a transcription inhibitory signal it has been believed that transcriptional repression is the main function of MECP2. Surprisingly, global transcriptional impairment has been revealed in RTT human neurons but not NPCs (Li et al. 2013). Progressive depletion of rRNA and tRNA affected total cellular RNA content and was associated with downregulation of many actively transcribed mRNAs including those for ribosomal proteins. As a result, general protein synthesis declined and neurons became atrophic. Of note, as in ribosomal protein-depleted or Pol1-inhibited neurons, there was also reduced activity of PI3K/Akt/mTOR signaling. Its exogenous activation rescued translational deficits and increased size of RTT neurons. Such a cellular pathology is somewhat reminiscent of anti-neuritic effects and reduced mTOR signaling following selective inhibition of ribosomal biogenesis in rat neurons (Gomes et al. 2011, Slomnicki et al. 2016b). Finally, reduced nucleolar size indicative of reduced rRNA transcription was observed in cortical neurons from a mouse model of RTT (Singleton et al. 2011). In summary, while pleiotropic effects of MECP2 deficiency may all contribute to the RTT phenotype, impaired synthesis of ribosomes appears as a major candidate mechanism of protein synthesis deficits, neuronal atrophy and dendritic degeneration.

IV.3. Pediatric motoneuron disease associated with mutations in LASL1, SMN1, and, EXOSC3.

Hemizygous LOF mutation of the X-linked 60S LSU RBF gene LAS1L was found as a likely cause of spinal muscular atrophy with respiratory distress (SMARD) (Butterfield et al. 2014). SMARD, which is manifested by neonatal muscle weakness including respiratory muscles, is a form of motor neuron disease that is associated with degeneration of motoneurons in the brainstem and the spinal cord. Causative role of this mutation in death of developing motoneurons is supported by spinal cord and brain cell apoptosis and disrupted structure of peripheral nerves in zebrafish embryos with a knockdown of LAS1L (Butterfield et al. 2014). While functional consequences for the SMARD-associated mutations of LAS1L are yet to be determined, its knockdown impaired LSU biogenesis and induced the TP53-mediated RS in a human cell line (Butterfield et al. 2014). Of note, two hypomorphic mutations in LAS1L are associated with Wilson-Turner X-linked mental retardation syndrome suggesting that dependent on extent of LAS1L deficiency, mutations in that gene can cause a spectrum of neurodevelopmental symptoms and that LAS1L is critical for ribosome synthesis in various types of neurons (Hu et al. 2016). However, LAS1L may also regulate mRNA transcription by participating in complexes that reduce repressive sumolylation of the transcription factor ZBP89 (Fanis et al. 2012). Hence, neuropathogenic role of deficiency in its pro-ribosomal activity is yet to be established.

Degeneration of developing motor neurons is a common phenotypic manifestation for loss of function mutations that affect RNA metabolism regulators SMN1 and EXOSC3 (Lunn & Wang 2008, Wan et al. 2012). As in both cases, RNA metabolism-associated functionalities of the affected proteins include rRNA processing, impairment of the latter process is a possible neuropathogenic contributor (Wehner et al. 2002, Bernabo et al. 2017, Gillespie et al. 2017). Mutations of SMN1, which is best known for its functions in mRNA splicing, are the most common cause of spinal muscular atrophy (SMA) (Lunn & Wang 2008). Intriguingly, tissue-specific impairment of translation was documented in the brain and the spinal cords of SMN1-deficient mice with the latter structure being most affected (Bernabo et al. 2017). Such deficits were closely correlated with onset of SMA-like symptoms (Bernabo et al. 2017). While reduced ribosomal content was observed in SMA motoneurons (Bernabo et al. 2017, Tapia et al. 2017), translation efficiency for multiple RPs was particularly affected (Bernabo et al. 2017). Mechanistically, SMA-associated ribosomal biogenesis deficits may arise due to loss of SMN1 functions in U3 snoRNA-mediated processing of pre-rRNA and/or in facilitating RP translation at neuronal ribosomes and/or in promoting snoRNA biogenesis in Cajal bodies (Wehner et al. 2002, Bernabo et al. 2017, Tapia et al. 2017).

While motoneuron degeneration-causing mutations in the exosome componetnt EXOSC3 may affect exosome-mediated metabolism of various RNAs, disease severity was shown to correlate with selective effects of mutations on rRNA processing (Gillespie et al. 2017). Therefore, common phenotypic features of mutations in LAS1L, SMN1 and EXOSC3 support a concept that deficient production of ribosomes is toxic to developing motoneurons. A likely mechanism would include compromised translation of proteins that are required for maintenance of these large cells. As defects of LAS1L or EXOSC3 may also lead to ID and/or microcephaly and/or cerebellar neurodegeneration, those cells are not the only type of neurons/CNS cells that may be negatively affected by the associated deficits in ribosomal biogenesis (Table 2).

IV.4. Pediatric neurodegeneration in UBTF- and Cockayne syndromes: dysregulation of nucleolar transcription and genomic integrity.

A pediatric neurodegenerative disease that is likely due to dysregulation of ribosomal biogenesis is caused by a dominant gain-of-function mutation in UBTF (Edvardson et al. 2017, Toro et al. 2018). Protein product of that gene encodes the nucleolar rRNA transcription factor UBTF1 that is required for the transcription-permissive state of rDNA chromatin and for Pol1 activity. In addition, UBTF2 is another splice variant that supports Pol2-driven transcription of many genes that play a role in chromatin organization and genome maintenance including histone gene clusters (Sanij et al. 2015). The UBTF mutation Glu210Lys (E210K) was found in 11 unrelated individuals with neurodevelopmental regression that started at age of 2.5–7 years (Edvardson et al. 2017, Toro et al. 2018). The regression started with motor- or cognitive/language function deficits and progressed through an extrapyramidal movement disorder with chorea, dystonia and parkinsonism to complete loss of ambulation, severe ID, and feeding difficulties. Progressive, postnatal microcephaly was observed in several cases. In addition, MRI revealed signs of primary cortical gray matter atrophy and possibly secondary atrophy of the white matter and the cerebellum. The Glu210 residue is highly conserved and is located in the second high mobility group (HMG) DNA binding domain of UBTF. Its substitution with a positively charged Lys increased rDNA affinity of the mutant UBTF resulting in elevated rRNA transcription (Edvardson et al. 2017, Toro et al. 2018). Importantly, expression of the mutant UBTF1 in flies was neurotoxic indicating sufficiency of this hyperactive rDNA regulator to induce neurodegeneration (Toro et al. 2018).

At this point one can only speculate how UBTF gain-of-function results in neurodegeneration. One possibility is that unbalanced production of pre-rRNA without simultaneous increase in availability of other ribosomal components and RBFs will produce dominant negative effects, thus blocking ribosomal biogenesis and/or leading to synthesis of dysfunctional ribosomes and/or activating RS (Edvardson et al. 2017). Neurons could be relatively more sensitive to such effects as the mutation-linked dys-coordination of ribosomal biogenesis could be magnified by downregulation of most components of the ribosome synthesis machinery during neuronal differentiation (Slomnicki et al. 2016b). In addition, hyperactive UBTF1 could lead to rDNA loci loss and/or a neurotoxic DNA damage response as increased rRNA transcription is associated with rDNA DSBs, recombinational instability of rDNA and rDNA damage (Kobayashi & Ganley 2005, Tchurikov et al. 2016, Xie et al. 2017). Indeed, increased occurrence of DSBs, mitotic deficits and apoptosis were observed in fibroblasts from patients with the E210K mutation (Toro et al. 2018). However, localization of E210K-associated DSBs at the gene level remains to be determined. Finally, such mechanisms could also interact with the apparent reduction in transcriptional activity of E210K UBTF2 at at least some of its target genes, many of which support genomic integrity (Toro et al. 2018, Sanij et al. 2015).

A potential pathogenic link between dysregulation of rDNA transcription, rDNA damage and activation of the neurotoxic DNA damage response has been recently proposed for Cockayne syndrome (CS), a congenital neurodegenerative disorder that affects the developing nervous system (Karikkineth et al. 2017). Most cases are caused by recessive loss of function mutations in ERCC6 (encoding the ERCC6/CSB protein) or ERCC8 (encoding the ERCC8/CSA protein). Clinical manifestations include impaired growth, postnatal microcephaly and progressive ID. Neuropathological findings include atrophy of the brain white matter as well as loss of neurons, especially in the cerebellum and the basal ganglia, but also in the brainstem, spinal cord, dorsal root ganglia and the cerebral cortex (Weidenheim et al. 2009). In addition, other organs display pathological changes consistent with premature aging. Both CSA and CSB play a role in the transcription-coupled DNA repair (the nucleotide excision pathway) which may account for progressive accumulation of oxidative damage in actively transcribed genes, leading to gene expression perturbations and triggering neurodegeneration (Karikkineth et al. 2017). In addition, mitochondrial malfunction has been extensively documented in CS (Karikkineth et al. 2017). Moreover, CSA and CSB contribute to Pol1-mediated transcription of rDNA (Koch et al. 2014, Bradsher et al. 2002).

Interestingly, in CS, transcriptional dysregulation of rDNA and mitochondrial impairment may be connected by activation of the DNA damage response mediator PARP1 (Scheibye-Knudsen et al. 2016). Namely, in CSB or CSA deficient cell lines, CS patient fibroblasts or CSA/B-depleted rat primary cortical neurons, increases in mitochondrial oxygen consumption, mitochondrial potential and superoxide production were observed together with impaired transcription near G quadruplex (GQ) structures in DNA templates (Scheibye-Knudsen et al. 2016). As such structures are highly abundant in rDNA, this likely accounted for a relatively selective impairment of rDNA expression that was also present in CSA or CSB-deficient cells. Interestingly, inhibiting rDNA transcription using a pharmacological inhibitor of Pol1 or a chemical stabilizer of GQs was sufficient to trigger CS-like changes in mitochondrial physiology. Mechanistically, increased occurrence of GQs has been proposed to activate PARP1, which by depleting cellular NAD⁺ to ribozylate DNA damage response proteins, altered mitochondrial function to promote oxidative stress (Scheibye-Knudsen et al. 2016) (Fig. 4). Inhibition of other RNA polymerases or downstream steps of ribosomal biogenesis was relatively less efficient inducing such changes (Scheibye-Knudsen et al. 2016). Interestingly, CSB was directly implicated in melting GQs and its nucleolar targeting was sufficient to rescue mitochondrial phenotypes in CSA cells (Scheibye-Knudsen et al. 2016). Finally, genetic or pharmacological inhibition of PARP1 improved mitochondrial physiology while supplementation of the NAD⁺ precursor, attenuated Pol1 inhibition-induced shortening of lifespan in C.elegans (Scheibye-Knudsen et al. 2016). These findings suggest that interference with rDNA transcription leads to mitochondrial dysfunction by stabilization of rDNA GQs, activation of PARP1 and the resulting depletion of NAD⁺. It remains to be tested if neurodevelopmental consequences of other mutations that interfere with rDNA transcription trigger a similar pathological cascade. In the meantime, gene expression analysis and mitochondrial potential assays suggest that neurodegeneration-associated defects in at least two transcriptional regulators including POLR3A and TBP may activate this pathway (Scheibye-Knudsen et al. 2016). Moreover, at least some neurological components of the CS phenotype including degeneration of the grey and/or white matter throughout the cerebrum and the cerebellum are similar to those in patients with disease-causing mutations in POLR1A or POLR1C that encode the largest Pol1 subunit, RPA194, or a common component of Pol1 and Pol3, respectively (Kara et al. 2017, Thiffault et al. 2015).

In addition, in human CS cells, translational fidelity is reduced (Alupei et al. 2018). Such a deficit has been proposed to contribute to increased errors of protein folding that resulted in endoplasmic reticulum (ER) stress (Fig. 4). Such a defective translation occurred despite normal ribosome content suggesting a possibility that the cellular ribosome pool contained more damaged ribosomes as ribosome replacement was slow due to reduced rRNA synthesis (Alupei et al. 2018).

Interestingly, besides neuronal degeneration, Cockayne syndrome also includes white matter pathology (Weidenheim et al. 2009). Similar changes are observed in cases of hypomyelinating leukodystrophies that are associated with mutations in such transcriptional regulators of ribosomal biogenesis as POLR1A, or POLR1C (Kara et al. 2017, Thiffault et al. 2015) (Table 2). Therefore, it is possible that, in addition to neurons, OL/OPCs are also sensitive to the DNA damage response/ER stress triggered by transcriptional inhibition of ribosomal synthesis. Indeed, OPCs are extremely sensitive to disruption of ER homeostasis (Lin & Popko 2009).

Taken together, genetic evidence supports the role of dysregulated ribosomal biogenesis in pathogenesis of human neurodevelopmental disorders with underlying phenotypes that include congenital or postnatal microcephaly, a spectrum of ID/ASD manifestations with or without changes in nervous system anatomy, and, progressive neurodegeneration. While more research is needed to define neuropathogenic mechanisms that underlie such phenotypes, likely candidate contributors include (i) RS-induced apoptosis of NSC/NPC and/or early post-mitotic neurons (Fig. 1), (ii) cell type- and/or mRNA-specific impairment of translation, which may compromise gene expression programs that are important for successful neurogenesis and/or neuronal/glia morphogenesis/maintenance/function (Fig. 2 and Fig. 3), and, finally, (iii) the neurotoxic DNA damage response and ER stress (Fig. 4). It is possible that these mechanisms coincide and/or co-operate to produce the relevant phenotypes. In addition, as several of the mutation-affected RBFs have pleiotropic functions beyond ribosomal synthesis, the neurodevelopmental phenotypes may also arise through interactions of ribosomal biogenesis dysregulation with other altered functionalities of the respective RBFs.

V. Neuroteratogenic factors that target ribosomal biogenesis.

In addition to gene mutation-mediated impairments of ribosome synthesis, extrinsic factors may also have negative effects on neurodevelopment due to dysregulation of ribosomal biogenesis. Those factors include alcohol, folate deficiency and Zika virus (ZIKV).

V.1. Folate deficiency.

In early pregnancy, folate deficiency is associated with neural tube defects (NTD) (Greene & Copp 2014). In addition, infantile neurodegenerative disease including disturbed myelination is associated with cerebral folate deficiency due to mutations that affect the cerebral folate transport (Hyland et al. 2010). A recent report that used an unbiased approach to determine hot spots for DNA double strand breaks in folate-deficient mouse embryonic stem cells identified the intergenic spacer region of rDNA as being particularly sensitive to damage (Xie et al. 2017). That damage was associated with altered epigenetic landscape of rDNA, increased recruitment of UBTF1 and higher expression of rRNA. The latter effect was also found in samples from human fetuses with NTD (Xie et al. 2017). Thus, RS-mediated impairment of neuroepithelium proliferation or RS-induced apoptosis could contribute to NTD. Likewise, rDNA strand break-driven RS and/or chronic activation of the DNA damage response and/or rDNA instability could underlie neurodegeneration in cerebral folate deficiency.

V.2. Alcohol.

Alcohol exposure during pregnancy is the most common cause of neurodevelopmental disruption that leads to fetal alcohol spectrum disorder (FASD) (Murawski et al. 2015). Such condition may combine CNS abnormalities with general growth deficits and facial dysmorphia. The components of the CNS phenotype may include ID, microcephaly as well as hippocampal, cerebro-cortical and/or cerebellar hypoplasia/atrophy. While multiple cellular mechanisms may underlie pathogenesis of FASD, increased apoptosis has been proposed to contribute to abnormalities in the CNS as well as facial morphogenesis (Ikonomidou et al. 2000, Smith et al. 2014).

Interestingly, comparative gene expression analysis of chicken NPCs that were isolated from strains with differential sensitivity to alcohol-induced apoptosis revealed ribosomal proteins as the most enriched gene ontology cluster with differential gene expression (Garic et al. 2014). In the sensitive chicken strain, both increased and decreased levels of RP mRNAs were observed. It has been proposed that the associated differences in ribosomal biogenesis activity could increase sensitivity to alcohol-mediated ribosomal stress. A support for that notion was provided by another whole transcriptome analysis that focused on alcohol effects on cranial gene expression in chicken embryos (Berres et al. 2017). In that system, alcohol reduced expression of 29 RP mRNAs and several ribosomal biogenesis associated genes. Moreover, in zebrafish embryos, sublethal knockdowns of several RPs enhanced craniofacial deficits and apoptosis in response to alcohol exposure (Berres et al. 2017). Therefore, a possibility exists that alcohol perturbs ribosome synthesis, and its pro-apoptotic effects in various sensitive cell populations including neural crest cells, NPCs or differentiating neurons are mediated by RS and/or translational deficits of pro-survival proteins. In addition, genetic background-dependent differences in activity of ribosomal biogenesis may determine resistance to neuroteratogenic effects of alcohol.

V.3. Zika virus.

Zika virus (ZIKV) is an RNA virus with a single-strand, positive-sense RNA as its genome. It is classified as a flavivirus and together with such relatives as Japanese encephalitis virus (JEV), West Nile virus (WNV) or Dengue virus (DENV) is spread by mosquitos (Baud et al. 2017). It was originally isolated in Africa and has been usually associated with a mild and transient infection in humans. In 2015 a ZIKV epidemic appeared in South America that was associated with increased incidence of severe microcephaly. Based on data from the US, estimated risk of ZIKV-induced microcephaly following infection with the South American strains of ZIKV during pregnancy is 6% and increases to 11% with possible infections during the first trimester (Honein et al. 2017). In addition to microcephaly at birth, the congenital ZIKV syndrome may also include other brain malformations such as lisencephaly, cortical thinning, hydrocephalus, cerebellar hypoplasia, atrophy of the cerebral white matter as well as damage to the retina and motoneuron loss in the spinal cord (Baud et al. 2017, Schwartz 2017, Chimelli et al. 2017). The accompanying neurological symptoms may include psychomotor delay, seizures, spasticity, blindness, akinesia and neurogenic muscle loss. Cases of postnatal microcephaly were also reported following ZIKV exposure in utero (Baud et al. 2017). Recent studies in ZIKV-infected human fetal brains revealed infection of intermediate cortical neuroprogenitor cells and immature cortical neurons (Ho et al. 2017, Lin et al. 2017). In the latter population, increased apoptosis was also observed. In addition, extensive apoptosis of non-infected neurons was reported. Neuronal apoptosis was found in rodent models of ZIKV infection in the developing brain (Cugola et al. 2016, Li et al. 2016, Shao et al. 2016, Rosenfeld et al. 2017). Moreover, in ZIKV-infected NPCs, reduced proliferation, increased differentiation and elevated apoptosis were reported in whole rodent- as well as human- and rodent cell culture models (Tang et al. 2016, Miner et al. 2016, Cugola et al. 2016, Wu et al. 2016, Li et al. 2016). However, molecular mechanisms that underlie these diverse cytotoxic effects of ZIKV are just beginning to emerge. In this context, some identified cytotoxic targets/effectors of ZIKV include the pro-growth/pro-survival AKT signaling pathway (Liang et al. 2016), the cytotoxic arm of the interferon/cytokine signaling (Yockey et al. 2018) (Bayless et al. 2016) , the ER stress pathway (Gladwyn-Ng et al. 2018), centromere and mitochondria (Onorati et al. 2016), Toll-like receptor 3 (Dang et al. 2016), TP53 (Zhang et al. 2016, Ghouzzi et al. 2016, Devhare et al. 2017) or NMDA receptors (Costa et al. 2017). While most of these mechanistic studies have focused on cytotoxic response in proliferating NPCs, mechanisms of ZIKV-induced neuronal apoptosis may involve similar or other mediators that act either in a cell autonomous or cell non-autonomous manner.

Interestingly, although the positive strand RNA viruses including flaviviruses replicate at the ER membrane, some of their proteins appear in the host cell nucleus (Salvetti & Greco 2014). For instance, capsid protein (protein C) of JEV, WNW or DENV accumulates in the nucleolus (Urbanowski et al. 2008, Byk & Gamarnik 2016). Moreover, these proteins interact with nucleolar proteins of the host including such RBFs as nucleolin (NCL), DDX56, or, nucleophosmin-1 (NMP1) (Fraser et al. 2016, Tsuda et al. 2006, Balinsky et al. 2013, Xu & Hobman 2012). As in most cases, knockdowns of such interactors reduced viral titers, flaviviral capsid proteins may serve as recruiters of host cell RBFs to support viral replication and/or viral assembly. In fact, various activities of host cell RBFs that are critical for ribosome synthesis may be useful in efficient production of other ribonucleoprotein complexes such as viral particles (Salvetti & Greco 2014). Regardless of functionality of the capsid protein-RBF interactions for the viral life cycle, they may lead to reduced RBF availability and compromised ribosome synthesis. Hence, RS and/or translational deficits could follow cell infection with various flaviviruses, and contribute to their cytotoxic effects including neuronal apoptosis.

Indeed, RS, as revealed by disruption of nucleolar NPM1 staining, preceded ZIKV-induced apoptosis of the infected cultured cortical neurons from rat embryos (Slomnicki et al. 2017). Both responses were observed after infection with either the prototype ZIKV strain (African strain MR766) or a contemporary, Asian lineage ZIKV isolate (the 2015 Puerto Rico isolate PRVABC59) that has an established causative link to human microcephaly unlikely the African lineage (Lanciotti et al. 2016).

When the capsid protein of ZIKV (ZIKV-C) was expressed in cultured neurons, its fraction accumulated in nucleoli (Slomnicki et al. 2017). In addition, overexpression of ZIKV-C produced similar disruptive effects on nucleolar morphology as ZIKV infection. Interestingly, despite the strong displacement of nucleolar NPM1 out to the nucleoplasm that resembled effects of inhibiting Pol1, rRNA transcription was only moderately reduced in ZIKV-C-overexpressing neurons (Slomnicki et al. 2017). As NPM1 enrichment in the nucleolus requires interactions with newly synthesized rRNA as well as other components of the ribosomal biogenesis pathway (Mitrea et al. 2016), ZIKV-C may have perturbed such interactions without a major decrease in rRNA synthesis. One possibility could be that ZIKV-C sequesters newly synthesized rRNA preventing rRNA-NPM1 interactions. In support of that notion, ZIKV-C was displaced from neuronal nucleoli when Pol1 was inhibited (Slomnicki et al. 2017). In addition, both nucleolar accumulation of ZIKV-C and NPM1 displacement were abolished by deletions of ZIKV-C C-terminal amino acid residues that in a closely related capsid protein of DENV are necessary for RNA binding (Byk & Gamarnik 2016). However, regardless of its mechanism, ZIKV-C-mediated displacement of nucleolar NPM1 is expected to have negative effects on ribosomal biogenesis as NPM1 provides the poly-valent scaffold for the GC component of the nucleolus where ribosomal assembly takes place (Mitrea et al. 2016). Such conclusion was further supported by observations of ZIKV-C-induced neuronal apoptosis that was attenuated by inhibition of the ribosomal stress effector TP53 or the ribosomal stress mediator RPL11 (Slomnicki et al. 2017).

Compared to primary neurons, rat or human NPCs displayed relatively milder changes in nucleolar morphology, when they were infected with ZIKV. The nucleolar changes were mostly limited to reduced nucleolar expression rather than nucleoplasmic displacement of GC components NMP1 or PES1 (Slomnicki et al. 2017). Consistent with such observations, ZIKV-C overexpression in human NPCs did not induce RS despite nucleolar enrichment of ZIKV-C. These data suggest that postmitotic neurons are more sensitive to ZIKV-induced RS including its downstream consequences such as TP53-mediated apoptosis. Such a differential sensitivity could explain high levels of neuronal apoptosis that were reported in ZIKV-infected fetal brain (Ho et al. 2017, Lin et al. 2017).

While mechanisms of neuronal sensitivity to ZIKV-mediated RS are not clear, one can speculate that different levels of ribosomal biogenesis activity in neurons and NPCs can be an important factor that determines ZIKV response. Namely, NPCs are in a proliferative state that is associated with increased rates of ribosome synthesis and elevated expression of ribosomal components and RBFs (Drygin et al. 2010, Pelletier et al. 2018). In these cells, high expression of those genes is maintained by the c-Myc family of transcription factors including N-Myc (Knoepfler et al. 2002, Boon et al. 2001). Conversely, when neurons emerge from neurogenesis as post-mitotic cells and start to differentiate, N-Myc expression goes down together with sharp downregulation of ribosomal biogenesis including the expression of NCL or NMP1, which are the target RBFs for flaviviral capsid proteins (Slomnicki et al. 2016b). Hence, in NPCs, ZIKV-C interactions with rRNA/RBFs may be insufficient to compromise ribosomal biogenesis as enough reserve capacity exists in that pathway. This is no longer the case in neurons where ZIKV-C would easily sequester critical amounts of ribosomal components/RBFs, including NCL and NPM1, triggering RS and TP53-mediated apoptosis. Obviously, verifying that concept requires further mechanistic studies such as identification of ZIKV-C host cell interactome or determination of cellular sensitivity to ZIKV-C after direct stimulation or inhibition of ribosomal biogenesis gene expression.

However, it is tempting to speculate that combination of neuronal differentiation-associated lowering of ribosomal biogenesis, together with high potential for apoptosis in early post-mitotic neurons may underlie sensitivity of the developing brain not only to ZIKV but also other neurotropic viruses that target the nucleolus. In fact, capsid protein of the DENV, which is closely related to ZIKV, displayed nucleolar enrichment, disrupted nucleolar NPM1, activated TP53, and induced apoptosis in rat embryonic cortical neurons (Slomnicki et al. 2017). Importantly, DENV becomes a leading cause of viral encephalitis in such countries as Brazil, Thailand or India (Puccioni-Sohler & Rosadas 2015). It is also capable of infecting cortical neurons in mouse neonates and triggering neuronal apoptosis (Despres et al. 1998).

One should note here that the pro-RS potential of flaviviral capsid proteins differs as no nucleolar disruption was induced by mature WNV-C despite its enrichment in cortical neuron nucleoli (Slomnicki et al. 2017). However, the less abundant immature WNV-C (i.e. corresponding to the membrane-anchored precursor form of ZIKV-C) has been shown to activate TP53 by directly binding to and inhibiting MDM2/HDM2 (Yang et al. 2008). Hence, it is likely that the flavivirus species-specific spectrum of host cell interactors underlies differences in the anti-nucleolar activity of flaviviral capsid proteins.

The ability of ZIKV-C to disrupt a membraneless organelle such as the nucleolus may suggest a potential to interfere with other membraneless structures including RNA stress granules, Cajal’s bodies or nuclear pore complexes. Indeed, recent work suggests that formation of all these organelles requires liquid-liquid phase transition (LLPS) (Lee et al. 2016). That process can be disrupted by arginine-containing dipeptide repeats (DPRs) that are produced from the ALS/FTLD-associated intronic expansions of the C9ORF72 gene (Lee et al. 2016). DRPs reduce the dynamics of membraneless organelles and compromise their functions by disrupting protein-protein interactions that involve basic amino acids (Lee et al. 2016). Therefore, a possibility exists that the highly basic ZIKV-C may perturb the function of not just nucleoli but also other membraneless organelles.

One should note that the ZIKV-C-induced RS may only be one of many factors that determine ability of ZIKV to disrupt neurodevelopment. First, the RS itself may represent one of several aforementioned effector mechanisms of ZIKV-mediated cell damage in the nervous system. Those effector mechanisms could be set off at various stages of neurodevelopment and/or in distinct neural cell populations dependent on the ability of a ZIKV strain to evade antiviral response, penetrate into the developing and/or mature CNS and infect specific types of neural cells. Thus, a recent comparison of early and contemporary ZIKV strains revealed similar neurovirulence in organotypic mouse brain slice cultures suggesting that differential ability to perturb neurodevelopment may be dependent on distinct potential for transplacental transfer and/or neuroinvasion of the fetal brain (Rosenfeld et al. 2017).

RS may be a contributing factor to the neurovirulence of other RNA viruses beyond the flavivirus family as many of these organisms produce proteins that accumulate in nucleoli. Some of these viruses are also known to perturb neurodevelopment. For instance, the Schmallenberg virus (SBV) belongs to the Orthobunyaviridae family and has documented neuroteratogenic effects in ruminants. The SBV non-structural protein (NS) localizes to the nucleolus leading to ZIKV-like nucleoplasmic translocation of NPM1 (Gouzil et al. 2017).

Another example of a neurovirulent RNA virus that disrupts ribosomal biogenesis is the poliovirus that in 1–2% of infected individuals can penetrate into the nervous system infecting and eventually killing motoneurons (Racaniello 2006). In this case, at least 2 pro-ribosomal stress mechanisms can be considered. First, the viral 3Cpro protease may inactivate Pol1 co-activators including UBTF1 and selectivity factor-1 (SL1) (Banerjee et al. 2005). Importantly, the nucleolar enrichment of 3CPro has been documented in a case of a related rhinovirus (Amineva et al. 2004). In addition, the 65 nucleotide long 3’ non-coding region of the poliovirus genome interacts with NCL, trapping the RBF in cytoplasm, thereby promoting viral replication (Waggoner & Sarnow 1998). Interestingly, similar RNA-mediated trapping of NCL has been shown to induce ribosomal stress and compromise neuronal survival in models of neurodegenerative diseases (Tsoi et al. 2012, Haeusler et al. 2014). These mechanisms may underlie brainstem neuron damage that sometimes results from infections of neonates/young children with the enterovirus 71, which is closely related to the poliovirus (Lee 2016). Lastly, nucleoplasmic translocation of NPM1 was also reported with the Newcastle virus which causes wide spread brain cell apoptosis in infected chickens (Duan et al. 2014, Venkata Subbaiah et al. 2015). In this case, however, no potential mechanism for RS has been yet indentified.

Clearly, RS emerges as a potential cytotoxic mechanism for various viruses that affect the developing nervous system. When neural cells are confronted with the viral infection-associated RS, apoptosis may follow, at least at those developmental stages, at which the apoptosis potential of given cell type is high. Thus, early post-mitotic neurons are among likely victims of the infection-associated RS.

VI. The unknowns.

While several lines of evidence support a role of dysregulated ribosomal biogenesis in neurodevelopmental pathologies, there are many outstanding issues that need to be addressed to fully evaluate its significance in that context. First, it is unclear what is the relative neuropathogenic contribution of the RS pathway in whole animal models of early brain damage that underlies microcephaly. Such a question could be answered by combining mouse models of these conditions with mutants that specifically perturb RS response including mouse knock-in of the MDM2 point mutation that prevents its interaction with the 5S RNP (Macias et al. 2010).

The second key question is whether translational deficits also play a role in neurodevelopmental manifestations of ribosomal biogenesis dysregulation. To address that question, neural cell type-specific analysis of individual mRNA translation efficiency across the entire transcriptome could be determined in relevant cell culture- or animal models using such technologies as Ribo-Tag or the translating ribosome affinity purification (TRAP) (Sanz et al. 2009, Heiman et al. 2014). An intriguing possibility is that such deficits may be dependent on types of 5’UTR regions that may be expressed in a cell type specific manner (Leppek et al. 2018). In this way, ribosomal deficits could make certain types of neurons more prone to connectivity problems by promoting dysmorphognesis and/or early neurodegeneration. Indeed, ribosome biogenesis deficits that were caused by a schizophrenia-associated fusion protein DISC1-Boymaw had most dramatic inhibitory effects on translation of the GABA synthesis enzyme GAD67 (Ji et al. 2014). Such anti-translational selectivity could contribute to defective inhibitory neurotransmission in Rett syndrome (Chao et al. 2010).

The third key issue is whether loss of other (i.e. non-ribosomal, non-RS) functionalities of the nucleolus underlie potential neuropathogenic consequences of dysregulated ribosomal biogenesis. In particular, role of the DNA damage response, rDNA instability and presence of global changes in chromatin organization could be interrogated on a transcriptomic/genomic scale in models of neurodevelopmental disease and compared to effects of experimental dysregulation of ribosomal biogenesis in whole animal brains.

Fourth, if ribosomal biogenesis dysregulation is indeed a significant player in neurodevelopmental pathologies, can it be targeted therapeutically? While a need for tight regulation of ribosomal biogenesis that is enforced by the RS and/or rDNA recombination may limit use of interventions focusing on any single step in that process, one could consider a possibility to harness endogenous signaling pathways such as MTOR that stimulate ribosome production at all levels (Iadevaia et al. 2014). In addition, if neuropathogenic consequences would be primarily due to specific events such as defective translation of a protein or a group of proteins, therapeutic development efforts could focus on such effector mechanisms rather than the ribosomal biogenesis per se. In a similar manner, deficient expression of GATA1 may be a better target than ribosomal biogenesis to restore erythropoiesis in DBA patients (Ludwig et al. 2014). Lastly, inhibition of ribosomal biogenesis including selective pharmacological targeting of Pol1 could be tested as a potential treatment for the pediatric neurodegenerative disease in patients with mutant UBTF that triggers excessive transcription of rDNA and the subsequent DNA damage (Bywater et al. 2012, Peltonen et al. 2014).

Taken together, advent of unbiased research approaches to search for genetic causes and/or pathological cascades that underlie neurodevelopmental syndromes resulted in rapidly accumulating evidence that dysregulation of ribosomal biogenesis is neuropathogenic in the context of the developing nervous system. Dependent on the developmental period/cell type in which ribosomal biogenesis defects are expressed their pathological consequences can vary from microcephaly, through functional deficits such ID or ASD to neurodegeneration. Given the scope of its potential involvement that ranges from rare genetic diseases through neuroteratogenic effects of alcohol or RNA viruses such as ZIKV, dysregulation of ribosomal biogenesis deserves further evaluation as a disruptor of neurodevelopment. In addition, pediatric neurodegeneration that is observed in at least some cases of perturbed ribosome production may shed light on pathogenesis of age-related neurodegenerative diseases including ALS/fronto-temporal dementia, or Huntington’s disease in which defective ribosomal biogenesis is well documented (Herrmann & Parlato 2018, Lee et al. 2014). Last but not least, uncovering effector mechanisms that underlie neurodevelopmental phenotypes of dysregulated ribosomal biogenesis will likely improve understanding of gene expression regulation that drives formation, function and maintenance of the nervous system.

Abbreviations:

5S RNP

5S ribonucleoprotein particle

ALS

amyotrophic lateral sclerosis

ASD

autism spectrum disorders

BDNF

brain-derived neurotrophic factor

CSA

Cockayne syndrome A

CSB

Cockayne syndrome B

CNS

central nervous system

DBA

Diamond-Blackfan anemia

DDX56

ATP-dependent RNA helicase 56

DENV

Dengue virus

DFC

dense fibrillar component

DSB

double strand break

ER

endoplasmic reticulum

FC

fibrillar center

FTD

fronto-temporal dementia

GC

granular component

GQ

G quadruplex

ID

intellectual disability

JEV

Japanese encephalitis virus

LOF

loss-of-function

LSU

60S large ribosomal subunit

LTP

long-term potentiation

MTOR

mechanistic target of rapamycin

NCL

nucleolin

NDD

neurodevelopmental disorders

NPC

neuroprogenitor cell

NPM1

nucleophosmin -1

NSC

neural stem cell

NTD

neural tube defects

OL

oligodendrocyte

OPC

oligodendrocyte precursor cell

PI3K

Phosphoinositide 3-kinase

Pol1

RNA-Polymerase-1

Pol3

RNA-Polymerase-3

RBF

ribosomal biogenesis factor

rDNA

47S pre-rRNA gene

RP

ribosomal protein

RS

ribosomal stress

RTT

Rett syndrome

SMA

spinal muscular atrophy

SMARD

spinal muscular atrophy with respiratory distress

SS

secondary structure

SSU

40S small ribosomal subunit

TCS

Treacher-Collins syndrome

TIF1A

Transcription initiation factor 1A of RNA-Polymerase-1

TP53

tumor protein 53

UBTF

Upstream transcription binding factor

uORF

upstream open reading frame

5’UTR

5’ untranslated region

WNV

West Nile virus

WNV-C

capsid protein of West Nile virus

ZIKV

Zika virus

ZIKV-C

capsid protein of Zika virus