The age-regulated zinc finger factor ZNF367 is a new modulator of embryonic neurogenesis

Global population aging is one of the major social and economic challenges of contemporary society. During aging the progressive decline in physiological functions has serious consequences for all organs including brain. The age-related incidence of neurodegenerative diseases coincides with the sharp decline of the amount and functionality of adult neural stem cells. Recently, we identified a short list of brain age-regulated genes by means of next-generation sequencing. Among them znf367 codes for a transcription factor that represents a central node in gene coregulation networks during aging but its function, in the central nervous system (CNS), is completely unknown. As proof of concept we analyzed the role of znf367 during neurogenesis. By means of a gene loss of function approach limited to the CNS, we suggested that znf367 might act as a key controller of the neuroblasts cell cycle, particularly in the progression of mitosis and spindle check-point. Using a candidate gene approach, based on a weighted-gene co-expression network analysis, we suggested possible targets of znf367 such as fancd2 and ska3. The age-related decline of znf367 well correlated with its role during embryonic neurogenesis opening new lines of investigation to improved maintenance and even repair of neuronal function.


Introduction
The age-related incidence of many brain diseases coincides with a reduced adult neurogenic potential. The regenerative capability and the amount of adult neural stem cells (aNSCs) decline with age, contributing to the reduced functionality of the aged brain. Despite the great interest in age related diseases, the molecular factors responsible for age-dependent decay of aNSCs function and the transition between stemness and differentiating properties of these precursors are almost unknown. Recently, we identified a set of evolutionarily-conserved genes expressed in aNSCs and age-regulated by RNA-seq analysis in the short-lived fish Nothobranchius furzeri, a well-established animal model in aging studies. Among them, zinc finger protein 367 (znf367) was suggested to occupy a central position in a regulatory network controlling cell cycle progression and DNA replication. We found that znf367 is expressed in the adult brain of N. furzeri, where its RNA level decreases with age, and in neuroblasts and retinoblasts of developing Zebrafish embryos 1 . Znf367 belongs to the Zinc finger (ZNF) transcription factors family that represents a large class of proteins that are encoded by 2 % of human genes 2, 3 . Their functions included DNA recognition, RNA packaging, transcriptional activation, regulation of apoptosis, protein folding and assembly, and lipid binding. Zinc finger proteins have an evolutionarily conserved structure and the ones containing the Cys 2 -His 2 motif constitute the largest family. The function of the majority of ZNF genes is largely unknown, but some of them play a critical role in the development and differentiation of the nervous system. For instance, the Kruppel-like zinc finger transcription factor Zic has multiple roles in the regulation of proliferation and differentiation of neural progenitors in the medial forebrain and cerebellum 4 . The Ikaros family of transcription factors is characterized by two sets of highly conserved Cys 2 His 2-type zinc finger motif and is involved in the maturation and differentiation of striatal medium spiny neurons 5 . Znf367 gene (also known as ZFF29) has been initially isolated in human fetal liver erythroid cells. In the human genome, this gene is on chr 9q and two alterative mRNA splicing products, were identified and designated ZFF29a and ZFF29b.
They both code for nuclear proteins, but only ZFF29b seems to act as an activator factor of erythroid gene promoters 6 . In Human SW13 adrenocortical carcinoma cell line, znf367 is overexpressed and in this cell line Znf367 downregulation caused an increase of cellular proliferation, invasion and migration 7 . Furthermore, znf367 was also identified as a potential tissue-specific biomarker correlated with breast cancer where its expression level is dysregulated influencing cell proliferation, differentiation and metastatic processes 8 . To our knowledge, there are no data available regarding the putative role of znf367 in the Central Nervous System (CNS) during embryonic and adult neurogenesis. In order to characterize the biological function of znf367 in vertebrates CNS, we analyzed its role during Xenopus laevis neurogenesis. The clawed frog Xenopus is the gold standard as animal model to perform functional screening of genes. In Xenopus, it is possible to microinject mRNAs or morpholino oligos in just one side of the early cleaving embryo and compare, in every single embryo, the manipulated side of the embryo with its wild-type counterpart that represent a perfect internal control. This model gave us also the unique possibility among vertebrates, to rapidly perform gene loss of function experiments in a tissue specific manner thanks to the well-defined fate map of each blastomere of the early cleaving embryo. This allowed us to target specific znf367 morpholinos to the central nervous system without interfering with the normal development of all other tissues. In this paper, we showed that znf367 is expressed in the developing CNS in Xenopus and it could have a key role in the primary neurogenesis regulating the neuroblast progression of mitosis. This aspect well correlates with its gene expression decline during CNS aging, suggesting that znf367 could represents a new key piece in the complex mosaic of developmental neurobiology and aging research.

Evolutionary conservation and embryonic expression analysis of znf367
To verify the evolutionary conservation of znf367 sequence in vertebrates we performed an in silico analysis of the amino acid sequences of ZNF367 in Homo sapiens (both splicing variants: ZFF29a and ZFF29b), Nothobranchius furzeri and Xenopus laevis (both splicing variants: znf367a and znf367b). This approach revealed a high conservation of znf367 with a 66% of identity between the human and Xenopus aminoacidic sequence that reached the 98% at the level of the zinc finger domains (Fig 1) suggesting a conserved putative znf367 function in vertebrates, from fish to tetrapods and primates. To analyze the spatio-temporal gene expression pattern of znf367, whole mount in situ hybridization (WISH) was performed on xenopus embryos at different stages.
Znf367 is expressed maternally in the animal pole in Xenopus embryos at blastula stage ( Fig. 2A-B) when compared to sense control probe treated siblings ( Fig. 2A). At neurula stage znf367 is expressed in the neural tube, in the eye fields, in the pre-placodal territory and in the neural crest cells (NCC) (Fig. 2C). At tadpole stage znf367 is widely expressed in the central nervous system, in the eye, in otic vesicle and in the NCC migrated in the branchial pouches (Fig. 2D). At larval stages of development, znf367 is still widely expressed in the CNS as showed in transverse sections (Fig.

Znf367 knockdown increases proliferation markers in Xenopus laevis embryos
To determine, whether the observed loss of post-mitotic neurons in znf367 morphants was the consequence of impairment in the maintenance of the neuronal progenitors pool, we examined the expression of the stemness genes sox2 and rx1, in the injected embryos. Sox2 and rx1 are involved in maintaining neuroblasts and retinoblasts as cycling precursors in the neural plate 20,21 .
The expression domains of sox2 and rx1 were expanded on the ZNF367-MO injected side of the embryo as compared to either the un-injected and Co-Mo injected sides ( Fig. 4A-B). These data were confirmed also by qRT-PCR analysis that showed a significant increase of both mRNAs in Coherently with the increase in mitotically active cells, a mild loss of p27 expression (phenotype 55%, n=93) was observed in neurula morphants indicating that znf367 depleted neuroblasts are unable to exit cell cycle ( Fig. 4M-N). These data let us to hypothesize that znf367 could be involved in the cell cycle exit and/or for the initiation of maintenance of a differentiated state. Finally, we examined morphants at tailbud stage by performing WHIS using rx1 and elrD (also known as HuD). ElrD labeled post mitotic neurons in the neural tube and the developing cranial ganglia 25 . As stated above, znf367 knockdown, but not control MO, caused an increase in rx1 gene expression (phenotype 52%, n=72) ( Fig. 4O) while inhibited neuronal differentiation affecting the expression of elrD (phenotype 54% n=64) (Fig. 4P). These data showed that the effects of ZNF367 depletion are not recovered even in the late phases of primary neurogenesis.

Identification of putative Znf367 targets: a candidate gene approach
Our previous results suggested that znf367 could represent a hub in the control of gene expression in the N. furzeri brain. In order to test the conservation of this co-regulation across species, we analyzed CORTECON 18 a public dataset of RNA-seq during cortical differentiation of human embryonic stem cells (hESCs) using weighted-gene co-expression network analysis (WGCNA) 17 . WGCNA constructs co-expression networks based on topological criteria, it was shown to be more robust than simple correlation and it has become the method of choice for gene expression studies in the nervous system. We therefore tested the conservation of gene coexpression networks between N. furzeri brain and human neuronal differentiation in vitro. We identified a conserved module that contains znf367 (Fig. 5). Then, we tested whether znf367 can be considered a hub in both species by computing its connectivity. Znf367 was among the top connected genes in the gene module in both species (98% percentile in N. furzeri and 92% percentile in human cells). Gene Ontology overrepresentation analysis revealed that cell-cycle related terms are highly overrepresented in this module. It should be also noted that all these genes have high expression in the hESCs, they are down-regulated during early differentiation and show a peak of expression around 12 days of differentiation in vitro that correspond to the period of cortical specification 18 . Among the genes that showed the highest topological overlap, we particularly noted an enrichment in genes known to be involved in the progression of mitosis and in the mitotic spindle check point (Fig.5B). This corroborate the idea that znf367 had a role in the control of cell cycle and it could be preeminent in mitosis, when the diving cell has the fundamental task to correctly arrange the genetic content in the two daughter cells. To verify our hypothesis, we decided to test by qRT-PCR the expression of three of the closest genes to znf367 showed in the network (Fig. 6). We analyzed the expression level of smc2, ska3 and fancd2 in   Among these, we decided to closely analyze the relation between znf367 and smc2, ska3 and fancd2. Smc2 is part of the condensing complex required for the structural and functional organization of chromosomes 36 . Its role is crucial in the chromatin compaction in the prophase 26 .
In our functional study the loss of znf367 seemed to interfere with smc2 mRNA level but even if smc2 mRNA seemed to be more abundant in znf367 morphants in respect to controls, the results are suggestive of a trend but not statistically significant. Fancd2 is essential during zebrafish CNS development to prevent neural cell apoptosis during neuroblasts proliferative expansion 29 . The loss of znf367 caused a differentiation failure keeping an enlarged number of neuroblasts in mitosis. In this condition, neuroblasts did not activate an apoptotic pathway that can be prevented when fancd2 is present 29 . The gene network analysis also suggested a possible function of znf367 in the regulation of the spindle checkpoint during the metaphase acting on the expression level of ska3.
In conclusion, we unveiled a role for znf367 during neurogenesis in vertebrates. In particular,

Embryo preparation
Animal handling and care were performed in strict compliance with protocols approved by Italian

In situ hybridization (ISH) experiments
Whole mount in situ hybridization (WISH) approaches was performed as described 11

TUNEL and PH3 staining in Xenopus
TUNEL (TdT-mediated dUTP-dig nick end labeling) and PH3 (phosho histone 3) staining was performed at neurula stage according to established protocols 13 10 . TUNEL and PH3 positive cells were counted within defined areas in control and injected side of each manipulated embryo.

Quantitative Reverse Transcription Polymerase Chain Reaction (qRT-PCR)
Total RNA was extracted from 15 Xenopus morphants using Nucleospin® RNA (Macherey-Nagel) according to the manufacturer's instruction. cDNA was prepared by using iScript™ cDNA Synthesis Kit (Bio-Rad) and quantitative real-time PCR was performed using GoTaq®qPCR master mix

WGCNA (Weighted Gene Co-expression Network Analysis)
Network analysis was performed using WGCNA method 17 . Samples used for the workflow were derived from two independent datasets, one from Nothobranchius furzeri's brain, comprehensive of two strains (MZM-04010 and GRZ), six different time points and five replicates per time point 1 and the other one from human embryonic stem cells. In particular the second one was obtained from a cerebro-cortical developmental experiment performed on hESC with 9 different time points 18 .
Network analysis was performed through different steps: -Setting of the soft threshold, coefficient necessary for the adjacency matrix construction, as shown in the formula: -Adjacency matrix and TOM (Topological Overlap Matrix), defined as: -Hierarchical clustering and modules detection after measuring the module eigengenes; every module is characterized by a color (as the module which has been studied for the analysis, defined by the turquoise color) -Module-trait relationship table construction, as correlation between single gene expression and external trait (in this case aging/development) -Module membership plot (as correlation between single gene expression and module eigengene): this was done for both the N. furzeri and the H. sapiens datasets, as described in Figure 5A -Visualization with Cytoscape software.
Network construction was done in two independent analyses: the first one only on Nothobranchius furzeri dataset, the second one using a consensus network obtained matching the two datasets. As soft threshold we chose β=6 for both the analyses to obtain the correspondent adiacency matrix and TOM, and significant modules negatively correlated with N. furzeri brain aging were selected. The genes contained in the selected modules were then tested for GO analysis using WebGestalt software, and then visualized using Cytoscape. Finally, the overall module membership of the genes contained in the "turquoise" module (as specified above, and only for the second analysis) was plotted on the ranked genes for both the killifish and the human data. Network analysis was performed using WGCNA method 17 . Samples used for the workflow were derived from two independent datasets, one from Nothobranchius furzeri's brain,