Cross-Platform Validation




Yüklə 1 Mb.
səhifə1/9
tarix20.04.2016
ölçüsü1 Mb.
  1   2   3   4   5   6   7   8   9






Cross-Platform Validation
In order to validate the accuracy of the data generated by the ABA a systematic comparison was made with other publicly available ISH data sources. These sources included the Brain Gene Expression Map (BGEM, http://www.stjudebgem.org/web/mainPage/mainPage.php) and radioactive ISH data generated by Dr. Ed Lein and Dr. Fred Gage (Salk Institute)1. BGEM is a publicly accessible database using high throughput radioactive ISH to map the expression pattern of selected genes in the C57Bl/6J mouse brain across multiple developmental time points including E11.5, E15.5, P7, and P42. Lein et al. (2004) analyzed the expression of more than 100 genes with unique expression patterns in the hippocampus in 10-11 week old C57Bl/6J mice. Expression patterns within the hippocampus were fully annotated and the full data set was available for review. These two sources provided an ideal data set with which to validate the ABA data set.
A systematic comparison of the three data sets was performed. Expression patterns were annotated for all genes represented in Lein et al. (2004) for which coronal images were available in the ABA database (72 genes). Of these genes, 25 had ISH data available in BGEM at the P42 (adult) time point. Annotation of gene expression in the hippocampal subregions was performed for the available data. A four-point scale was used to score intensity of expression relative to other brain regions. For each data set, an observer recorded a relative intensity score for the primary cell type within each hippocampal brain region including: CA1, CA2, and CA3, the dentate gyrus (DG), the hilus, the subiculum, the fimbria, and the choroid plexus. A general summary of the expression pattern within the hippocampus was then generated based on these scores. This analysis, in addition to a direct visual comparison of all available expression data, was used to form a conclusion as to whether the expression patterns of the different data sets match. These data are recorded in Table 1. Comparisons of plane-matched images from the ABA and Lein et al. (2004) are presented in the accompanying figures (Figures 1 and 2).
Figure 1 shows matching coronal sections for Calbindin 1 (Calb1), Nephroblastoma overexpressed gene (Nov), Connective tissue growth factor (Ctgf) and Parvalbumin (Pvalb). Darkfield-illuminated radioactive ISH data from Lein et al. (2004) is on the left, and DIG-based non-isotopic data from the ABA is presented as the mirrored hemisphere. Probe lengths for each gene were approximately 500 nt for 35S-based ISH, and approximately 900 nt for DIG-based ISH. The patterns and relative expression levels within a section between these two methodologies are, qualitatively, nearly identical for each of these genes. At this low magnification the autoradiographic signal is easier to visualize than the non-isotopic purple reaction product; however, at higher magnification the cellular morphology and ability to resolve individual cells in densely packed structures is superior in the non-isotopic methodology. Figure 2 shows additional images of plane-matched coronal sections comparing ISH data from Lein et al. (2004) to data from the ABA for all genes examined in the cross platform validation effort.
Overall there was very strong concordance across all three data sets. 71/72 genes available in the ABA and Lein et al. (2004) have qualitatively matching expression patterns. 24/25 genes available in all 3 data sets show similar expression patterns. One gene (minichromosome maintenance deficient 6 [Mcm6]) for which the expression pattern of the ABA data did not match that of the data from Lein et al. (2004), (BGEM data was not available), and one gene (Zinc finger protein 312 [Zfp312]) for which the expression pattern of the BGEM data did not match that from the ABA or from Lein et al. (2004). Neither of these genes was represented in GenePaint (http://genepaint.org), another database which includes ISH expression data for mouse brain. In order to better understand the cause of this discrepancy, we examined the probes that were used by the different databases for these genes. For Mcm6 the ABA used an 839 nt probe spanning exons 7 through 12. Lein et al. (2004) used a 1515 nt probe spanning exons 10-17. This suggests that alternate transcripts may account for the differences in expression pattern that were noted for this gene. In addition, the ABA has generated three sets of ISH data for Mcm6 using the same probe sequence, making it unlikely that a processing error led to this discrepancy. For Zfp312, the probe used by the ABA (830 nt) showed no overlap with the probe used by BGEM (382 nt). The probe used by Lein et al. (2004) spanned nearly the entire coding sequence of the gene (~2000 nt). This also leaves open the possibility that the distinct expression patterns that were observed may represent alternate transcripts. Overall the data generated for the ABA showed strong concordance with these two independent sources of ISH data.

1. Lein ES, Zhao X, Gage FH. Defining a molecular atlas of the hippocampus using DNA microarrays and high-throughput in situ hybridization. J Neurosci. 2004 Apr 14;24(15):3879-89.




(a) (b) (c) (d)

Figure 1 - Comparison of expression patterns in plane-matched coronal mouse brain sections processed for radioactive ISH (a,c) and non-isotopic DIG ISH (b,d). ISH data is shown for Calbindin 1 (Calb1, upper panels in a,b), Nephroblastoma overexpressed gene (Nov, upper panels in c,d), Connective tissue growth factor (Ctgf, lower panels in a,b) and Parvalbumin (Pva, lower panels in c,d).




MGI Gene Name

Genbank Number

Data Source

CA1


CA2

CA3

DG Granule Cells

Hilus

Subiculum

Fimbria (Oligodend-rocytes)

Choroid Plexus

Final Classification-Hippocampus

Patterns Match?

Anxa11


annexin A11

U65986

ABA

++


+++

+++

++++

-

-

-

++

DG+CA2+CA3 enriched








Lein et al., 2004

++


++++

++++

++++

-

-

-

++

DG+CA2+CA3 enriched

yes






BGEM

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

Atp1a1

ATPase, Na+/K+ transporting, alpha 1 polypeptide


Atp1a1

AA063844

(also W12919)



ABA

++++

++++

++++

++++

+

+++

-

+++

Highly expressed in all excitatory cells of the hippocampus








Lein et al., 2004

++

++

++

++++

+

+

-

++

DG enriched

yes






BGEM

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

Atp2a2


ATPase, Ca++ transporting, cardiac muscle, slow twitch 2

SERCA2a


AA153484

ABA

++++


++++

++++

+++

+++

+++

-

++

CA1-CA3 enriched








Lein et al., 2004

+++


++++

++++

++

++

++

-

-

CA1-CA3 enriched

yes






BGEM

++


+++

+++

+

N/A

N/A

-

-

CA2+CA3 enriched

Yes with lower expression

Bok


bcl-2-related ovarian killer protein

Mtd

AA537500

ABA

-


-

++++

-

++

+

-

-

CA3 restricted








Lein et al., 2004

-


-

++++

-

++

-

-

-

CA3 restricted

yes






BGEM

+


+

++++

+

++

-

-

+

CA3 enriched

yes

Btg1


B-cell translocation gene 1, anti-proliferative

L16846

ABA

++

-

-

++++

-

-

-

+

DG + CA1 restricted








Lein et al., 2004

++

-

-

++++

-

-

-

-

DG+CA1 restricted

yes






BGEM

+

-

-

++

-

-

-

-

DG+CA1 restricted

Yes

Cadps


Ca2+dependent activator protein for secretion

Caps1

D86214

ABA

++


++++

++++

+

+++

+++

-

-

CA2+CA3 enriched








Lein et al., 2004

++


++++

++++

+

+++

+

-

-

CA2+CA3 enriched

Yes






BGEM

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

Calb1


calbindin-28K

CB

M21531

ABA

+++

-

-

++++

-

-

-

-

DG+CA1 restricted








Lein et al., 2004

+++

-

-

++++

-

-

-

-

DG+CA1 restricted

yes






BGEM

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

Calb2


calbindin 2

calretinin (CR)

X73985

ABA

-

-

-

-

++

-

-

-

Dentate hilus restricted








Lein et al., 2004

-

-

-

-

+++

-

-

-

Dentate hilus restricted

yes






BGEM

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

Camk2g


calcium/calmodulin -dependent protein kinase II gamma

Camkg


W12204

(also AA068735)



ABA

++++

+++

+++

++++

++

+++

-

+

DG+CA1 enriched








Lein et al., 2004

++

++

++

++++

++

++

-

-

DG enriched

yes






BGEM

++

+

+

+++

+

++

-

+

DG+CA1 enriched

yes

Car4


carbonic anhydrase 4

U37091

ABA

+++


++

++++

++

++

++

-

-

CA1+CA3 enriched








Lein et al., 2004

+++


++

++++

++

+

+

-

-

CA1+CA3 enriched

yes






BGEM

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

Ccnd2


cyclin D2

M83749

ABA

+


++

++++

++

+++

+

-

++++

CA3 enriched








Lein et al., 2004

++


++++

++++

++

++

+

-

+++

CA2+CA3 enriched

yes






BGEM

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

D0H4S114


DNA segment, human D4S114

P311, Sez17, Ptz17


X70398

ABA

-


-

-

+++

+

++++

(pre-subiculum)



-

+

DG restricted








Lein et al., 2004

-


-

-

+++

+

++++

(pre-subiculum)



-

+

DG restricted

yes






BGEM

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

Dcn


Decorin

Dspg2, Pg40, Pgs2, Slrr1b




X53929

ABA

++

(caudal CA1 only)



-

-

-

-

++++

ventral


-

+++

Caudal CA1+Subiculum only



  1   2   3   4   5   6   7   8   9


Verilənlər bazası müəlliflik hüququ ilə müdafiə olunur ©azrefs.org 2016
rəhbərliyinə müraciət

    Ana səhifə