Expression of a novel brain specific isoform of C3G is regulated during development

Mice lacking C3G (RapGEF1), a ubiquitously expressed protein essential for neuronal differentiation, show multiple defects in brain development. Function of C3G in neurogenesis is poorly defined. Here, we identify brain specific expression of a novel C3G isoform in mice and humans. This isoform has an insert in the Crk-binding region, generating a polypeptide of 175 kDa, unlike the previously known 140 kDa form expressed in all other tissues. In the adult mouse brain, C3G expression is seen in neurons, but was not detectable in GFAP-positive cells. C3G levels were high in the CA3 region of hippocampus and in mitral cells of olfactory bulb. Neural progenitor cells positive for Doublecortin and Nestin, show expression of C3G. During development, C3G is expressed in precursor cells prior to their differentiation into mature neurons or astrocytes. The 175 kDa as well as 140 kDa forms are seen in embryonic mouse brain, while only the 175 kDa variant is seen in post-natal brain. Human cerebral organoids generated from induced pluripotent stem cells predominantly expressed the 140 kDa polypeptides, and the 175 kDa isoform appeared upon maturation. This study describes developmental regulation and neuronal expression of a brain specific isoform of C3G, a molecule essential for normal development of the mammalian brain.


Results
C3G expression in human and mouse brain. Studies from our lab as well as that of others have shown the expression of C3G RNA and protein in a variety of tissues and cell types 15,16,28,38,56,57 . Prior to examining the expression and localization of C3G in the mouse brain by immunohistochemistry, we analyzed C3G expression in brain tissues obtained from mice of different age groups by western blotting. Whole brain lysates showed the presence of two closely running polypeptides of molecular weight around 175 kDa when examined using commercial polyclonal antibodies, H300, and C-19, that recognize the N-terminal and C-terminal of C3G, respectively (Fig. 1A, Supplementary figure 1A). Their specificity to detect C3G in mouse brain lysate by western blot Figure 1. Expression of C3G in mouse brain tissue. (A) Lysates of brains obtained from 3-days-old (3d), 3-months-old (3 m) and 1-year-old (1y) mice were subjected to western blotting and probed with C3G (H300) antibody to detect C3G. Lysates of testis tissue from 3m-old mouse and Neuro2A cells were also used for comparison. After probing for C3G, the lower part of the blot was cut between 63 and 48 kDa markers, and probed for β-actin as a loading control. (B) Expression of C3G in spinal cord. Western blotting of tissues obtained from 3m-old mouse brain, spinal cord and testis were probed with H300 antibody to detect C3G. β-Actin was used as loading control. (C) Schematic representation of mouse C3G gene and exons. Alternately spliced transcripts, mouse isoforms 3 and 1 with their NCBI-Nucleotide ID's are shown below. Exon numbers are indicted below the transcripts. The longer isoforms have additional exons (13,14) in the region that encodes the CBR domain of C3G protein (shown in the boxed region). m_C3G_ISO primers were designed to give products of different lengths based on presence or absence of exons 12-15. Expected products from isoforms 1 and 3 are shown, along with their sizes. (D) Isoform specific amplification of C3G transcripts. RNA was extracted and cDNA prepared from 14-day mouse embryo brain (E14) and 3-month-old mouse brain tissue. Semi-quantitative PCR was performed using primers designed to differentiate between the C3G isoforms. cDNA prepared from C2C12 cells was used as a control for expression of NM_054050.2 isoform. β-Actin was used as internal control. NC indicates no-template control for the PCR. Arrow and arrowhead indicate the 2 kb and 1.5 kb PCR products, respectively. (E) Identification of peptides from C3G expressed in brain by mass spectrometry. Schematic shows the domains present in the 140 kDa and 175 kDa isoforms of C3G. Position of the exons, poly-proline tracts (P1-P5) and residues from the additional exon (black box) are indicated. The 175 kDa polypeptide of C3G immunoprecipitated from adult mouse brain was subjected to mass spectrometry. Identified peptides (underlined) corresponding to sequences present in the 138 amino acid insert (pink letters) of the longer isoform are shown. (F) Expression of C3G in adult human brain tissue. Lysates from adult mouse testis, human brain and adult mouse brain were subjected to western blotting using three C3G antibodies as indicated. The upper part of the blot was cut at the 100 kDa marker to probe for C3G. The blot was deprobed after probing with each of the antibodies. β-Actin was used as a loading control. The vertical dotted line indicates removal of three intervening lanes. (G) C3G localization in mouse brain cortex. Immunohistochemistry of cortical sections was carried out using the two indicated antibodies raised against non-overlapping domains of C3G. Images shown were captured using the Axioimager Z1 microscope. Blank image shows section processed similarly without addition of primary antibody and only the anti-rabbit Alexa-488 (R-488) secondary antibody. Scale bar, 50 µm. www.nature.com/scientificreports/ is shown in Supplementary figure 1A. These antibodies have been used extensively to study cellular C3G and specifically recognize C3G, and no other proteins from cell lysates 10,13,22,25,27,31,38,39,[57][58][59][60][61][62][63][64][65][66] . As an alternate tissue type and cell line for comparison, lysates prepared from 3-months-old mouse testis and Neuro2A cells were used, which showed C3G-specific polypeptides only at 140 kDa ( Fig. 1A and Supplementary figure 1C) as described earlier 28,29,57 . Mouse spinal cord also showed expression of higher molecular weight polypeptides similar to that of the brain when probed using C3G H300 or C-19 antibodies ( Fig. 1B and Supplementary figure 1B). The expression of the higher molecular weight C3G polypeptides in the brain, was validated using an alternate monoclonal antibody, 3F6mAb, that recognizes epitopes in the CBR. This antibody recognizes 140 kDa polypeptides of C3G in a variety of cell lines 57 , but detected 175 kDa polypeptide in the brain (Supplementary figure 1C).
Therefore, it appears that mouse brain expresses alternate isoforms of C3G that give rise to larger polypeptides of molecular weight 175 kDa, unlike the 140 kDa polypeptide, which is seen in other tissues and cell types. The mouse C3G gene has 28 exons, and the widely expressed 140 kDa protein arises from a 5837 bp C3G transcript, described as mouse C3G isoform 3 (NM_054050.2), which has 24 exons (Fig. 1C). RefSeq database search showed the presence of two longer isoforms of C3G that could produce polypeptides of 175 kDa-C3G isoform 1(NM_001039087.1; 6251 bp), and isoform 4, (NM_001362702.1; 6134 bp) obtained from a cDNA library made from adult mouse brain (FANTOM Consortium, RIKEN Genome Exploration Research Group). These isoforms 1 and 4, differed from the shorter isoform 3 (NM_054050.2), in having additional exons 13 and 14, in the CBR coding sequence (Supplementary figure 2A) 67,68 . Exons of other sizes have also been seen at the same splice site of C3G 68,69 (mouse C3G isoform 2, NM_001039086.1 found in adult mouse testis cDNA RIKEN library and RIKEN full-length enriched mouse cDNA library from pooled tissues). Isoform 2 has an insert of 511 bp, corresponding to exons 13, 14 and 15 (Supplementary figure 2C). Isoforms 2 and 4 differ from isoforms 1 and 3 at the N-terminal (Supplementary figure 2A).
The size of the predicted polypeptide of the longer C3G isoforms, corresponded to that of the larger polypeptide (175 kDa) expressed in mouse brain. In order to determine which of the longer isoforms was expressed in the mouse brain, we prepared cDNA to carry out PCR with specific primers. Primers positioned on either side of the exon positions 12 and 16 were designed to give rise to amplicons of different lengths based on presence or absence of exons 13, 14 and 15 (Fig. 1C, Supplementary figure 2C). While a product of about 1.5 kb was amplified from C2C12 cells, brain tissue from embryonic-day 14, and adult (3-months-old) mouse, showed a major amplicon of about 2 kb (Fig. 1D). The 2 kb and 1.5 kb amplicons were sequenced using m_C3G_ISO forward and reverse primers. The 2 kb fragment showed the presence of nucleotides seen in the longer isoforms from the brain cDNA library that matched the transcripts having additional exons 13 and 14 (Fig. 1C, Supplementary  figure 3A). In the protein sequence, this insertion is located at an exon junction in the sequence following the residues that code for the fifth poly-proline tract in CBR (Fig. 1E). Sequencing using reverse primer, showed that the 2 kb amplicon lacks the extra base pairs seen in isoform 4 with exon 15 (data not shown). The 1.5 kb insert lacked any of these residues and had contiguous sequence corresponding to isoform 3, with exon 12 spliced to exon 16 (Supplementary figure 3B). To determine the presence of exon 4 in the major brain transcript, we used m_N-term primers, specific to this region, which would give rise to a 550 bp product if exon 4 was present (Supplementary figure 2B). Thus, we confirmed that the brain specific isoform was C3G isoform 1, as it has exons 4, 13 and 14 apart from those seen in Isoform 3.
Interestingly, we observed differential expression of isoforms 1 (175 kDa) and 3 (140 kDa) in brain tissue derived from the embryo, and adult mouse. Level of the longer transcript (isoform 1) was higher in the adult brain, while the shorter transcript (isoform 3) was higher in the embryonic brain (Fig. 1D). The predominantly expressed isoform 1 in mouse brain tissue, therefore, could be giving rise to larger polypeptides of 175 kDa seen in western blotting. This was confirmed by carrying out mass spectrometric analysis of C3G protein immunoprecipitated from adult mouse brain. The 175 kDa polypeptide analyzed is shown in Supplementary figure 2E. Analysis showed the presence of two peptides, with sequence that corresponds to fragments translated from exons 13 and 14 of the mouse C3G isoform 1 (Fig. 1E). The absence of exon 15 in this isoform was evidenced by the presence of a peptide with sequence generated from exon 14 spliced to exon 16 (peptide 2 in Fig. 1E). Amplicons of sizes between 1.5 kb and 2 kb are also seen weakly expressed in the brain tissue, which may correspond to predicted isoforms that have different number of exons at the splice site in the CBR region. (Supplementary figure 2C). Sequencing of these minor amplicons confirmed that they are generated from variants with either exon 13 or 14 corresponding to products of about 1.7 and 1.8 kb (Supplementary figure 2D). The size and sequence of 1.7 kb product shows a match to a predicted mouse C3G isoform X8 (XM_030246488.1), that has 13th exon but lacks 14th exon 68 .
The 138 amino acids arising from exons 13 and 14 are well conserved in predicted C3G transcripts from various species (Supplementary figure 4A) suggesting that tissue specific splicing at this site may be an essential feature. Adult human brain tissue obtained during surgical resection from an epilepsy patient was examined for C3G expression using three different antibodies that detect C3G. Polypeptide of 175 kDa was seen in human brain sample similar to that seen in adult mouse brain (Fig. 1F). Two human cell lines K562, and IMR-32 showed expression of only the 140 kDa C3G isoform (Supplementary figure 1D). RNA from human brain tissue showed the presence of additional exons similar to those seen in mouse C3G, as confirmed by PCR using hu_C3G_ISO primers which showed a 2 kb amplicon (Supplementary figure 4B). Sequencing of 2 kb amplicon showed the presence of basepairs corresponding to sequences present in mouse C3G isoform 1 (Supplementary figure 4C). Various transcripts were found for human RapGEF1 in RefSeq database, derived by automated computational analysis using gene prediction method: Gnomon. One of the predicted human cDNAs, transcript variant X10, mRNA: XM_011518575.3 showed insertion of 414 bp similar to mouse isoform 1 (Supplementary figure 4D) 68 . Our results therefore confirm brain specific expression of a human C3G isoform with additional exons in the CBR region.
Scientific Reports | (2020) 10:18838 | https://doi.org/10.1038/s41598-020-75813-z www.nature.com/scientificreports/ The expression and localization of C3G in cells of the adult mouse brain was examined by immunohistochemistry. Specific staining for C3G was predominantly seen in the cytoplasm of cortical neurons when examined using H300 or C-19 antibodies. It was observed that a subset of cells was negative for C3G staining. Similar section processed without addition of primary antibody (Blank) does not show signals specific to cells (Fig. 1G). We examined the expression of C3G in neuronal and glial cell lines derived from human or mouse, by western blotting as well as immunofluorescence. All cell lines examined, irrespective of whether they were of neuronal or glial cell origin, showed C3G polypeptides of 140 kDa, similar to that seen in a mouse myoblast cell line, C2C12 (Supplementary figure 5A). Polypeptides of 175 kDa seen in mouse and human brain tissue were not expressed in these cells. The subcellular localization of C3G was also examined in some of these cells by indirect immunofluorescence. C3G staining was seen in the nucleus as well as cytoplasm, though the relative intensities in these compartments differed in the various lines (Supplementary figure 5B). These results suggest that mouse and human brain tissue express the 175 kDa C3G protein generated by alternately spliced mRNA, but its expression is lost in neuronal and glial cell lines. Expression of C3G is high in CA3 region of hippocampus. Transverse brain sections through the hippocampal region were stained for C3G, and NeuN, or GFAP. Images taken using a 5 × objective were reconstructed by photomerging to show the architecture of the whole region (Fig. 3A). Examination using higher objective showed the expression of C3G in all neurons of the hippocampus (Fig. 3B). Expression was seen in the granular cells of dentate gyrus (DG) as well as pyramidal cells of Cornu Ammonis (CA) regions. C3G fluorescence intensity quantitation showed highest expression in CA3 region of the hippocampus (Fig. 3E). Granular cells and pyramidal cells were identified by their morphological characteristics 72 . Granular cells of DG are smaller and closely packed, having polarized morphology, with the cell body projecting into dentate molecular layer and axons projecting towards hilus and CA3 pyramidal cell layer. Pyramidal cells in the CA sub-region have a large cell body, and are seen as a densely packed layer curving into a 'U' shape. Sections processed without addition of primary antibodies are shown as blanks (Fig. 3D). Neurons of the hippocampus showed predominantly cytosolic localization. NeuN was seen in both nucleus and cytosol. C3G expression was not seen in any of the GFAP positive cells in this region (Fig. 3C).
Expression of C3G in the cerebral cortex and other brain regions. We examined the expression of C3G in cells of different sub-regions of cerebral cortex using a transverse section from the mid brain. Supplementary figure 6A shows a reconstructed image indicating various regions. Neurons in the retrospenial (RSP) area, somatomotor, auditory and olfactory regions showed positivity for C3G, when probed with H300 antibody along with NeuN (Supplementary figure 6B). C3G was also present in NeuN negative neurons of the thalamus. Most of the neurons in the regions of the fiber tract such as corpus collosum (CC), fimbria (fi) and corticospinal tract (cpd and int) that were negative for NeuN, showed positivity for C3G. It was seen that a subset of cells in these areas, particularly the internal capsule, which were NeuN negative, did not also show C3G staining. Neurons belonging to different regions are known to serve varied functions in the mammalian brain 70,[73][74][75][76] . It therefore appears that C3G is expressed in all neuronal sub-types, irrespective of their distinct functions. Expression of C3G in mouse brain cortical sections was also validated using an alternate C3G antibody, C-19 that targets residues in the C-terminal domain of C3G (Supplementary figure 7A). C3G co-expression with NeuN was seen in the cortical neurons. In other brain regions also, C3G detected using the C-19 antibody showed similar pattern of expression to that seen with H300 antibody (Supplementary figure 7B).
Age dependent expression of C3G in hippocampus and olfactory bulb. We examined the expression of C3G in hippocampal sections of brains collected from neonatal (2 days-old), adult (3-months-old) and aged (1-year-old) mice. Western blotting of whole brain lysates obtained from these age groups did not show significant difference in C3G levels (Fig. 1A). The hippocampal region is poorly differentiated in neonatal mice and showed weak expression of C3G and NeuN in the DG (Fig. 4A). Distinct expression of C3G was seen in the neurons of CA2 and CA3 regions of 3-months old mice (Fig. 4B). NeuN expression was primarily restricted to neurons present in the CA2 and CA3 regions of the neonatal brain, possibly due to differential maturation of hippocampal neurons 77,78 . Similar to brains from 3-months-old mice, 1-year-old mice showed prominent C3G expression along with NeuN in DG as well as CA2 and CA3 regions (Figs. 3B, 4C). Though whole brain lysates from mice of these age groups did not show significant differences in total C3G protein expression, quantitation of fluorescence intensity in sub-regions showed weak expression in neonatal hippocampal neurons compared to corresponding regions in 3-months-old and 1-year-old mice (Fig. 4D). Age-related differences in C3G expres-Scientific Reports | (2020) 10:18838 | https://doi.org/10.1038/s41598-020-75813-z www.nature.com/scientificreports/   C3G expression was also examined in the olfactory bulb of mice from different age groups. Low magnification images showed nearly uniform expression of C3G in the neonatal (1-day-old) olfactory bulb, which is poorly differentiated (Supplementary figure 8A). Examination of cells co-labelled for expression of C3G and NeuN at higher magnification showed C3G expression in all olfactory bulb cells of the neonatal brain while NeuN was weakly seen in a subset of cells (Supplementary figure 8B). Cells of the mitral layer (ML) showed high levels of C3G expression in 3-months-old as well as 1-year-old mice. Neurons of the granule cell layer (GCL) showed strong expression of NeuN and moderate expression of C3G in 3-months-old and 1-year-old mice.
Expression of C3G in embryonic mouse brain and human cerebral organoids. To understand the role of C3G in mouse brain development, we examined its expression in mouse brain lysates from embryonicday 14 and 18 (E14 and E18). Western blotting showed the presence of the 175 kDa C3G protein in brain tissue from all age groups (Fig. 5A). Unlike adult brain tissue, embryonic brain samples expressed the 140 kDa C3G polypeptide, found predominantly in cell lines and testis tissue. Similar to the differential expression of the two isoforms detected by RT-PCR (Fig. 1D), E14 brain lysate showed higher expression of the 140 kDa polypeptides relative to the 175 kDa products. At these developmental stages, neuronal precursor cell marker DCX is expressed, while mature neuron marker is weakly expressed. GFAP expression was not observed (Fig. 5A).
To examine C3G expression during human brain development, we generated cerebral organoids from iPSCs (schematic shown in Supplementary figure 9). Brain organoids grown in vitro serve as good models to recapitulate development process and organization of human brain 45 . Figure 5B shows the predominant expression of the 140 kDa C3G isoform in iPSCs as well as 55-day organoids. By 90 days of maturation, the cerebral organoids show expression of the 175 kDa isoform, concomitant with the expression of NeuN, a marker of mature neurons. Pax6, Nestin, and DCX were examined as markers of neuroectoderm, and neural precursor cells. Their expression is seen in the organoid samples, but not in adult brain tissue. These results show that expression of C3G isoforms in the human brain is developmentally regulated, with the 175 kDa isoform expressing as neurons mature.
C3G expression in neural progenitor cells. C3G is widely expressed in the developing brain collected from E14 and E18 mice as seen from transverse sections. This pattern of expression is similar to that described earlier 36 . Most of the cells at this stage show expression of C3G, and very weak expression of NeuN, as observed by immunohistochemistry (data not shown) and by western blotting (Fig. 5A). The expression of C3G in embryonic brain neurons prior to expression of NeuN suggested that C3G is expressed in NPCs. This was examined in the hippocampal and olfactory brain sections of 3-months-old mouse, by co-staining C3G with NeuN, GFAP or DCX. DCX is a microtubule associated protein expressed in adult NPCs, and used as an established marker 49 . As seen from Fig. 6A,B, C3G showed co-expression with DCX in a subset of cells in the olfactory bulb, as well as hippocampus. C3G also showed good co-localization with DCX in precursor cells of the olfactory bulb. The expression of C3G in NPCs was also confirmed by examining its expression in Nestin positive cells of brain sections obtained from a transgenic mouse expressing GFP-Nestin (Fig. 6C).

Discussion
The spatial and temporal pattern of C3G expression in the developing and mature mammalian brain have been examined in this study. C3G polypeptides are seen in western blots as a doublet of nearly 140 kDa in a variety of cell lines when probed with commercial or in-house generated antibodies that detect the C3G protein specifically 10,20,22,28,57 . Examination of mouse and human brain tissue showed the presence of bands with slower mobility compared to the C3G polypeptides seen in adult mouse testis or other cell lines derived from neuroblastomas or gliomas. Using three antibodies that target different epitopes in C3G, we demonstrate that the predominant polypeptides expressed in the brain are larger, with a molecular weight of 175 kDa. The expression of this form in spinal cord, and brain stem (data not shown) suggested that it is specific to the CNS. Similar to the 140 kDa polypeptide, which is known to be phosphorylated resulting in mobility shift 22 , the higher molecular weight polypeptides also ran as a doublet. This is indicative of post-translational modifications of this form, or the presence of a variant with additional residues in the brain. C3G is a substrate of c-Abl, SFKs and Cdk5, kinases expressed in the CNS 14,20,[79][80][81][82] . These kinases may also be phosphorylating the 175 kDa form to cause a mobility shift.
A cDNA with an insertion of 414 base pairs in the CBR has been obtained from an adult mouse brain library. Using RNA isolated from the brain, and mass spectrometry of C3G polypeptides immunoprecipitated from brain tissue, we were able to show the expression of this longer form, which gives rise to a protein of 175 kDa. Multiple sequence alignment using BLASTp and CLUSTAL Omega tools showed that the insert sequence is highly Figure 4. C3G expression in hippocampus of mice of different age groups. Panels show images of hippocampal sections of (A) 2-day-old mouse, (B) 3-months-old mouse and (C) 1-year-old mouse, stained with C3G (H300) (green) and NeuN (red). Sub-regions of hippocampus are labelled as DG, dentate gyrus; CA1, CA2 and CA3, Cornu Ammonis areas. Inset images are blanks of the same magnification. Secondary antibodies anti-rabbit Alexa-488 and anti-mouse Alexa-594 were used. Images were taken using Nikon ECLIPSE microscope. Scale bar, 50 µm. (D) Quantitation of C3G expression in hippocampal regions of mice of different age groups. Scatter plot shows fluorescence intensity of C3G in the indicated regions from mice belonging to different age-groups. Horizontal line, and error bar represent mean ± standard deviation. p values are shown to indicate significance of difference in intensities at 95% confidence limit between the indicated regions. www.nature.com/scientificreports/ conserved among the predicted mRNAs of various species, though there are differences in the lengths of the inserts [83][84][85][86] . The splice site is the same as that shown earlier, where an insert of 153 base pairs has been observed in rat testes 15 . Splicing in this region of CBR, appears as a hotspot generating expression of tissue specific variants. Multiple isoforms of C3G are known to exist in various species, but isoform specific functions are largely unknown. An N-ter truncated variant, p87C3G is overexpressed in chronic myeloid leukemia 87 . The insertion of additional residues may confer unique properties to this isoform, required in the brain neurons, like controlling neuronal morphogenesis and structural plasticity. Interestingly, unique phosphorylation sites for Protein kinase G, casein kinase I, and insulin receptor were found among the 138 amino acids, that were absent in the 1086 isoform, when the sequences were analyzed through NetPhos3.1 (https ://www.cbs.dtu.dk/servi ces/NetPh os/) 88 . In fact, mass spectrometry data collected from the mouse brain 89 identified phosphorylation at amino acid S675 in a peptide present within the brain specific 175 kDa form. We propose to clone this brain specific isoform to carry out further studies and understand its unique functions. Despite major differences in the microenvironment, gene regulatory networks, developmental processes, and 3D organization are preserved in human brain organoid models 45,46,53 . Our results from embryonic mouse brain, human cortical organoids and adult mouse and human brain indicate similarity in developmental regulation of the C3G isoforms between mouse and humans. Our results also validate the utilization of brain organoids grown in culture for studying splicing that occurs during development.
C3G immunostaining was seen only in neurons and not detected in GFAP positive cells. Though we cannot rule out the possibility of very low expression below detection limits in astrocytes, our findings are consistent with data obtained by proteomic analysis of different cell types in the mouse brain 90 . Many other proteins like Serine racemase, Eps8, and HRD1 are known to be expressed only in neurons [91][92][93] . During development of the nervous system, neurogenesis precedes gliogenesis 94 . It is known that inducers of neuronal fate, like, Neurogenin, suppress glial differentiation 95 . As a promoter of neuronal differentiation, C3G may be functioning as a suppressor of glial cell differentiation, and its expression is shut off in glial cells of the adult mammalian brain. C3G is predominantly expressed in pyramidal neurons in the cerebral cortex and hippocampal CA regions. Its presence in granule cells of the olfactory bulb as well as hippocampus suggested that it is expressed in both GABAergic as well as glutamatergic neurons. In neonatal olfactory bulb, all cells show C3G expression irrespective of their differentiation. Adult brain neurogenic areas, the subependymal zone of the olfactory bulb and the dentate gyrus of the hippocampus, which are positive for DCX and Nestin, also showed expression of C3G indicating that its expression is not restricted to differentiated neurons. But, identification of the isoforms expressed in these cells will require generation of antibodies that specifically recognize the 175 kDa isoform. Embryonic brain, largely composed of undifferentiated cells, and lacking expression of either NeuN or GFAP showed the presence of C3G, suggesting its possible function in proliferating precursor cells as well as differentiated neurons.
The 175 kDa protein predominantly seen in western blots of brain tissue is the form primarily expressed in neurons of the adult brain, which are in a non-proliferating state. In contrast, the 140 kDa protein is the only form detected in neuronal as well as glial cell lines derived from tumors. Many miRNAs show altered expression in Figure 5. C3G expression in embryonic mouse brain and human cortical organoids. (A) 14th (E.14) and 18th (E.18) day mouse embryo, 3-days-old (3d), 3-months-old (3 m) and 1-year-old (1y) mouse brain lysates were subjected to western blotting by probing the same membrane for C3G (H300), NeuN, GFAP and DCX. The upper part of the membrane, cut above 75 kDa, was probed for C3G (H300 antibody). The lower part between markers 75 kDa and 25 kDa was probed for NeuN, GFAP and DCX (blots were deprobed after each antibody probing). Neuro2A cell lysate was used as positive control for expression of 140 kDa isoform of C3G (Q91ZZ2). β-Actin was used as loading control. (B) Expression of C3G in human cerebral organoids. Lysates prepared from iPSCs, 55-day (55d), 90-day (90d) organoids and human brain lysate were subjected to western blotting using indicated antibodies. The blots were cut and processed as indicated in (A). www.nature.com/scientificreports/ gliomas 96,97 , and it is known that C3G is one of the targets of miR-124 98 . Therefore, repression of miR-124 could explain the increased expression of C3G seen in glioma cell lines. We also observed that the cell lines derived from tumors show the presence of significant amounts of C3G in the nucleus, unlike normal brain neurons and NPCs which show predominant cytoplasmic staining. We have earlier observed that compared to cells derived from other tissues, IMR-32 cells, derived from a neuroblastoma show significantly higher levels of nuclear C3G 22 . It therefore appears that external cues and alterations leading to tumorigenicity regulate expression and localization of C3G. It is possible that immortalized neuronal cell lines capable of continued proliferation, lose expression of the 175 kDa isoform and gain expression of the 140 kDa C3G isoform. The 175 kDa isoform seen in adult brain, may be inhibitory to cell proliferation. Since C3G plays an important role in neuronal differentiation, and mice with C3G deficiency show defects in neuronal maturation and migration, it was not surprising to see that C3G is expressed in embryonic brain and also in adult NPCs. C3G is required for multipolar to bipolar transitioning in developing neurons 36 , and the activity of C3G targets like Rap1, regulate synaptic function, spine morphology, and stability [99][100][101][102] . Neurons in the CA3 region of the hippocampus show a higher degree of connectivity and play specific role in memory 103 . The higher expression of C3G in this region is suggestive of an important role for C3G in maintaining structural and functional properties of these neurons. During development, the program regulating division and self-renewal, or exit from the cell cycle and differentiation is tightly controlled, and imbalances can lead to severe functional consequences. This is evident from the phenotype of mice lacking normal levels of C3G. at 37 °C in a CO 2 incubator. DMEM supplemented with 20% FBS was used to maintain C2C12 cells. All experiments were carried out with exponentially growing cells. Human induced pluripotent stem cells (hiPSC) F2-3F1 used for generation of cerebral organoids have been described 104 . IEC sanction no. IEC-71/2019, was obtained for work on human iPSCs. They were maintained in E8 medium and were differentiated into cerebral organoids using the protocol described earlier 105 . The organoids were allowed to mature for 3 months. 55-day and 90-day organoids were used for examining C3G protein. RRID:AB_2223041] was from Millipore. All fluorescent labelled secondary antibodies were obtained from Thermo Fisher Scientific. C3G 3F6mAb, mouse monoclonal antibody, was raised against recombinant CBR in the lab, and characterized 57 . C3G-GFP plasmid has been described earlier 12 .

Materials. Anti
Protein molecular weight marker prism ultra was purchased from Abcam. DNA ladder, Gene Ruler (100 bp-5 kb) was obtained from Thermo Fisher Scientific. 1 kb ladder was obtained from NEB. Protein A/G agarose beads (Cat no: sc-2003) were purchased from Santa Cruz Biotechnology. Protease inhibitor cocktail (Cat no: 11873580001) was obtained from Roche. Trypsin used for in-gel digestion was from Roche. Acetonitrile and formic acid were from Biosolve (Mass spectrometry grade).
Animals and tissue processing. The work was carried out after approval by the Institutional animal ethics committee (IAEC) of CCMB, no. IAEC 08/1017, and the protocols strictly follow the guidelines of Committee for the Purpose of Control and Supervision of Experiments on Animals (CPCSEA), Government of India for animal welfare. Institutional ethics committee (IEC) of CCMB approval no. IEC 57/2017 was obtained for procuring human brain tissue. Experiments were undertaken after obtaining the informed and written consent of each subject. None of them were below 18 years of age. The study conforms with The Code of Ethics of the World Medical Association.
Wild-type C57 black mice (RRID: IMSR_JAX:000664) and Nestin-GFP transgenic mice, expressing GFP, under the control of nestin gene regulatory regions, were used in this study 106 . Animals were deeply anesthetized with ether and cardiac perfusion performed with normal saline followed by 4% paraformaldehyde (PFA) in PBS. Brains were collected after decapitation, fixed in 4% PFA overnight at 4 °C and transferred to 30% sucrose solution until they settled down. For cryosections, fixed brains were sliced from the mouse brain matrix and mounted in OCT (Optimal cutting temperature) freezing medium, and 10 µm transverse sections taken using a Leica CM1950 cryostat. Sections were then collected on slides and used for immunohistochemistry. All results were validated by examining 3 or more sections from each brain region, obtained from 3-5 animals of each age group. Western blotting and transcript analysis was carried out using tissues prepared from three animals of each group.
Indirect immunofluorescence and microscopy. Immunohistochemistry was carried out using tissue sections on slides. Slides were rehydrated by sequential immersion in gradient alcohol (100%, 90% and 70%), for 5 min each, followed by a 5 min wash in distilled water. Antigen retrieval was done by immersing the slides in trisodium citrate buffer (pH 6) for 24 min, in a boiling water bath, and distilled water wash for 5 min at room Scientific Reports | (2020) 10:18838 | https://doi.org/10.1038/s41598-020-75813-z www.nature.com/scientificreports/ temperature. Sections were then blocked using 2% BSA for 1 h at room temperature followed by primary antibody incubation for 2 h at room temperature or overnight at 4 °C. The slides were washed and incubated with the secondary antibody for 1 h at RT. After washes, the slides were mounted with glycerol containing anti-fade and DAPI. Adjoining tissue sections, in all cases were processed similarly, but without addition of primary antibody, to serve as blanks for secondary antibodies.
Dual labelling was carried out by sequential incubations of the two primary and corresponding secondary antibodies. Anti-rabbit Alexa-488 (green) secondary antibody was used for C3G staining, except in case of Nestin-GFP mouse brain section (Fig. 6C), where anti-rabbit Alexa 546 secondary antibody was used for C3G staining. H300 antibody was used to detect C3G in all experiments, unless otherwise mentioned. Anti-mouse Alexa-594 (red) secondary antibody was used for NeuN staining. Anti-goat Alexa-546 (red) secondary antibody was used for GFAP and DCX staining. For examining endogenous C3G in various murine and human-origin cell lines, cells were cultured on 18 mm coverslips, fixed with cold methanol and indirect immunofluorescence carried out as described earlier 64 .
Image analysis and statistics. Images were captured using Nikon ECLIPSE microscope, Leica TCS SP8 Confocal microscope (Leica Microsystems, Germany) or Axioimager Z1, Carl Zeiss (Germany). Images from the Confocal Z stacks were merged and processed using the Leica Application Suite. Constant image acquisition parameters were used for capturing images of all samples from a given experiment. Image reconstruction was carried out by capturing multiple images of various mouse brain sections using a 5 × objective of Axioimager Z1 fluorescence microscope (Carl Zeiss). Complete image was reconstructed using Photomerge tool in Adobe Photoshop CS3 software. Fluorescence intensity quantitation was carried out using ImageJ software. Several 25 μm 2 areas in the DG and CA regions from multiple fields/sections of the hippocampus were marked for C3G fluorescence intensity quantitation, and adjusted with average background intensity in areas of the same section lacking cells. Data obtained from multiple sections of indicated brain regions belonging to 3 animals were averaged to assess differential expression of C3G. We used unpaired t-tests between pair of regions to determine the statistical significance of the intensities between the regions. The p-values at 95% confidence interval and number of samples (n) are shown in the figures.
Western blotting. Whole cell lysates were prepared from cells in Laemmli buffer, western blotting carried out as described earlier 20 . Brain and testis lysates were prepared by homogenizing in modified RIPA buffer (50 mM Tris-HCl, pH 7.4; 150 mM NaCl; 1 mM EDTA; 1% NP-40; 0.5% sodium deoxycholate, 0.1% SDS and 1× Protease inhibitor cocktail). Subsequently, samples were boiled in 1× Laemmli buffer and subjected to SDS-PAGE, as described earlier 20 . 20 µg of total protein was loaded in each lane after quantitation 107 . After transfer, the membranes were cut based on size of expected bands of various antigens, and probed simultaneously, whenever commercial and previously characterized primary antibodies were used. The images of the membranes (some full length, and some cut to size to save primary antibodies/avoid deprobing multiple times) used are shown in the supplementary information provided with original scans. Deprobing was carried out when the same part of the membrane was used to probe for more than one antigen. Bands were detected using Vilber-Lourmat Chemiluminescence system (Germany). Western blot panels shown in each figure are those obtained from the same gel.
Semi-quantitative RT-PCR and sequencing. Total RNA was prepared from mouse tissues, adult human brain and C2C12 cell line using Trizol (Invitrogen) reagent. RNA was converted to cDNA using reverse transcriptase (Invitrogen SuperScript III First-Strand Synthesis System kit). PCR was carried out using the prepared cDNA for amplification of C3G isoforms and β-actin. Mouse specific primers m_C3G_ISO-F (5′-GAA CAA GCA CAT GCT GGC CTA-3′) and m_C3G_ISO-R (5′-CCT AGG CTT GAT CTT CAG AGAG-3′) for distinguishing isoforms with and without the additional exons, were used to amplify C3G from cDNA of mouse tissues and C2C12. Actin-F (5′-GGC TGT ATT CCC CTC CAT CG-3′) and Actin-R (5′-CCA GTT GGT AAC AAT GCC ATGT-3′) were used to amplify actin, used as an internal control. PCR was performed in Eppendorf Master Cycler with, Phusion High-Fidelity DNA Polymerase (Thermo Fisher Scientific). The following PCR conditions were used for amplification of C3G: initial denaturation was at 98 °C for 30 s, followed by 30 cycles of denaturation at 98 °C for 10 s; annealing at 65 °C for 30 s, and extension at 72 °C for 2 min. Final extension was for 5 min at 72 °C. Amplified PCR products were examined and pictures captured using Vilber-Lourmat Gel Documentation system (Germany). The PCR fragments were eluted and sequenced to confirm that the amplicons are generated from alternate splice variants. Mouse specific m_N-term_F primer (5′-CGT TCT CAT CTC TCC TCC TTC-3′) and m_N-term_R primers (5′-ACA GTT GTC ACC ATC TCC TTATC-3′) were used to determine presence or absence of exon 4 in mouse brain cDNA.PCR conditions used are similar as above, except annealing temperature used was 48 °C. Human specific primers hu_C3G_ISO-F (5′-CAA ACA CAT GCT GGC CTA CAT-3′) and hu_C3G_ISO-R (5′-CTG GGT TTA ATT TTC AGA GACAG-3′) were used to amplify C3G from human brain cDNA, in order to distinguishing isoforms with and without additional exons. PCR was performed as mentioned above. PCR conditions used was: initial denaturation was at 98 °C for 30 s, followed by 30 cycles of denaturation at 98 °C for 10 s; annealing at 48 °C for 30 s, and extension at 72 °C for 3 min. Final extension was for 10 min at 72 °C.
For peptide identification, raw MS data files were loaded onto computational platform Proteome Discoverer Received: 28 January 2020; Accepted: 1 October 2020