| The major interneurons in the olfactory bulb(OB), the periglomerular and granule cells, are derived from neural stem cells (NSCs) that migrate from the lateral ganglionic eminences(LGE) in embryo. Postnatally, they are derived from NSCs that migrate in the rostral migratory stream(RMS) from the anterior subvenricular zone (SVZa) in the lateral ventricles of the forebrain. The SVZa-derived cells, after arriving at the OB, disperse radially and ultimately differentiate into the granule and periglomerular cells of the OB. This form of migration is classified as chain migration. It was found that about 70% and 30% of periglomerular interneurons were immunoreactive for GABA and tyrosine hydroxylase (TH) respectively. Local environments have an important role in the proliferation, migration and differentiation of the SVZa-derived cells and numerous genes may be involved in this process. DNA microarray technique is an advanced molecular biological platform which can profile the dynamic changes of more than ten thousands of genes. In this study, a total of 12 RNA samples from different regions of OB, RMS and SVZa in the four different stages( Embryonic day 16 (E16), Postnatal day 1(P1), P7 and P21) were screened by large-scale Oligo microarray to identify genes related to chain migration during mice forebrain development. The main results were summed up as follows. 1.DNA microarray experiment We used oligo microarrays that were purchased from Capital Biochip Corporation National Engineering Research Center for Beijing Biochip Technology and contain probes for 16,844 genes to profile comprehensive molecular and genetic programs underlying the developmental mice forebrain. Microarray experiment design involved the preparation of the total RNA samples. Twelve experimental samples were obtained and pooled from the same regions and time points in different animals. A common reference sample was obtained and pooled from whole brain tissues of the four different stages. We also designed dye-swap pairs to minimize the effects of any gene specific dye-bias. Lowess normalization method was used to adjust for scale differences among slide and comparisons of expression levels cross slides. Thus, we could obtain two ratio and then calculate the geometric mean for each gene in dye-swap pairs which must have the similar expression patterns. 2. Microarray data verification Distal-less homebox 2 ( DLX2) gene was selected from 12 tissues identified in prior studies by DNA microarray technique. Expression of DLX2 gene was calculated by real-time revers tanscription PCR (RT-PCR) using SYBR Green I and compared the relative gene expression with DNA microarray. Our results showed the similarity of DLX2 gene expression patterns between RT-PCR and DNA microarray data which can be used for advanced analysis. 3. Initial statistical anaysis The gene expression patterns and functions of the thousands of interested genes were initially analyzed by using powerful analytical approaches, such as hierarchical clustering, relevance hetworks analysis and some conventional statistical methods. 3.1 The genes were identified for the subsequent analysis if they are classified as "active"and "differentiated". The "active"gene must qualified as good spot in the four time points. The good spot includes no significant change, up-regulated, down-regulated, turn on and off gene. The "differentiated"gene must be "active", and the difference between the maximum and minimum of the four time points is larger than 2-fold change (≥2 or ≤0.5). 3.2 The "active"genes were obtained from the developmental mice OB, RMS and SVZa by the criteria above. Some of the "active"genes were uniquely confined in an individual region (OB:64; RMS: 223; SVZa: 687). Some of them were observed to appear in two of the three regions (OB-RMS:64; OB-SVZa:423; RMS-SVZa: 372). In addition, there were 1,847 "active"genes that were overlapped in all three regions. Among the 1,847 overlapped "active"genes, they were classified into 13 functional groups based on their gene ontologies. Our results showed that Metabolism,Transport and Signal transduction functional groups contained the greatest number except the Unclassified group.The 1,847 genes were hierarchical clustered into 6 gene clusters which consist of different list of genes: C(1)-541,C(2)-194,C(3)-421,C(4)-114,C(5)-343,C(6)-324. Gene distribution analysis (x2) showed the distribution of the number of gene in Transport functional group was non-random. The 1,847 genes were analyzed spatially and gene clusters of OB, RMS and SVZa identify by hierarchical dendrograms were compared. It was found that 6 pairs of gene clusters expressed with high similarity which indicated these genes had the similar expression patterns in different regions during the proliferation, migration and differentiation of SVZa-derived NSCs. We also found that there were more genes between RMS and OB or SVZa than OB and SVZa. By using gene distribution analysis, the 1,847 genes were observed non-randomly distributed in all three regions. By using the hierarchical clustering of the 1,847 genes with the functionally clustered pattern, The candidate genes were identified from the three regions. The candidate genes were listed for each brain region in the three functional groups: Cell differentiation(OB:2,RMS:2,SVZa:2), Response to stress(OB:2,RMS:2,SVZa:2), Developmen(tOB:16,RMS:14,SVZa:15).There were the most number of candidate genes in Development functional group. An approach based on time-shift analysis was used to demonstrate the closeness of expression patterns among the 1,847 genes. A number of positively and negatively correlated genes were identified from pair-wised comparison which was classified as high correlation pair and statistically significant if its correlation coefficient is larger +0.99 or smaller than-0.99. 3.3 The "differentiated"genes were obtained from the development mice OB, RMS and SVZa by the criteria above. Some of the "differentiated"genes were uniquely confined in an individual region (OB:149;RMS:616;SVZa:342). Some of them were observed to appear in two of the three regions (OB_RMS:118; OB_SVZa:89; SVZa_RMS:252). In addition, there 267 "differentiated"genes that were overlapped in all three regions. Among the 267 overlapped "differentiated"genes, they were classified into 13 functional groups based on their gene ontologies. Our results showed that Metabolism, Transport and Development functional groups contained the greatest number except the Unclassified group.The 267 genes were hierarchical clustered into 8 gene clusters which consist of different list of genes: C(1)-15,C(2)-64,C(3)-26,C(4)-45,C(5)-33,C(6)-21,C(7)-38,C(8)-25. Gene distribution analysis showed the distribution of the number of gene in all of the 13 functional groups was non-random. The 267 genes were analyzed spatially and gene clusters of OB, RMS and SVZa identify by hierarchical dendrograms were compared. We found that there were more genes between RMS and OB or SVZa that OB and SVZa. By using gene distribution analysis, the 267 genes were observed non-randomly distributed in all three regions. By using the hierarchical clustering of the 267 genes with the functionally clustered pattern, The candidate genes were identified from the three regions. The candidate genes were listed for each brain region in the three functional groups. There were the most number of candidate genes in Development functional group. The 267 genes were analyzed by relevance networks algorithm and showed the 0.99 correlation coefficient (>+0.99 or <-0.99). We obtain 11 relevance-networked groups with 26 genes. 3.4 The differentiated genes were clustered in the three brain regions separately (OB-623, RMS-1253, SVZa-950). According to the dendrograms in the clustering, the differentiated genes were classified into different gene clusters in each brain region--OB:(C1-215,C2-158,C3-124,C4-126),RMS:(C1-334,C2-93,C3-312,C4-301,C5-215),SVZa:(C1-132,C2-268,C3-114,C4-292,C5-144). The uniquely differentiated genes were clustered in the three brain region separately (OB-149,RMS-616,SVZa-342). According to the dendrograms in the Clustering, the uniquely differentiated genes were classified into different gene clusters in each brain region—OB:(C1-44,C2-20,C3-39,C4-46),RMS:(C1-30,C2-165,C3-48,C4-39,C5-103,C6-129,C7-102),SVZa:(C1-50,C2-119,C3-16,C4-50,C5-99,C6-8). 4. Biological analysis 4.1 To study TH relative genes in developmental mice OB and discuss the molecular mechanism of differentiation of the SVZa-derived NSCs. TH relative genes were Screened using genetic relevance networks analysis. We had observed 623 differentially expressed genes with differential ratio ≥2 and 18 TH relative genes were screened at the 0.95 correlation coefficient level.There were 8 classified genes which had functions inMetabolism,Development and Transport and 10 Unclassified genes in these 18 genes. The 18 genes have a role in TH expression during OB development.The 18 genes were observed in hierarchical clustering groups of OB differentiated genes. We obtained the list of them: C1-7,C2-3,C3-2,C4-6. These data information should prove valuable in advancing our understanding of the molecular mechanisms underlying differentiation of SVZa-derived NSCs in OB. 4.2 To study DLX2 relative genes in developmental mice OB,RMS and SVZa. DLX2 relative genes were Screened using genetic relevance networks analysis. We observed 623 differentially expressed genes with differential ratio ≥2 and 29 DLX2 relative genes were screened at the 0.95 correlation coefficient level in OB.The 29 genes were observed in hierarchical clustering groups of OB differentiated genes.We obtained the list of them: C1-5,C2-4,C3-8,C4-12.In addition, we observed 1253 differentially expressed genes with differential ratio ≥2 and 43 DLX2 relative genes were screened at the 0.99 correlation coefficient level in RMS.The 43 genes were observed in hierarchical clustering groups of RMS differentiated genes.We obtained the list of them:C1-29,C2-0,C3-13,C4-0,C5-1.We also observed 950 differentially expressed genes with differential ratio ≥2 and 34 DLX2 relative genes were screened at the 0.995 correlation coefficient level in SVZa. The 34 genes were observed in hierarchical clustering groups of SVZa differentiated genes.We obtained the list of them: C1-0,C2-23,C3-0,C4-11,C5-0. These genes have a role in DLX2 expression during OB,RMS and SVZa development. |
PDF Full Text Request