Mouse strain specific gene expression differences for illumina microarray expression profiling in embryos
- Petra Kraus†1,
- Xing Xing†1,
- Siew Lan Lim1,
- Max E Fun2,
- V Sivakamasundari1,
- Sook Peng Yap1,
- Haixia Lee2,
- R Krishna Murthy Karuturi2 and
- Thomas Lufkin1Email author
© Kraus et al.; licensee BioMed Central Ltd. 2012
Received: 18 January 2012
Accepted: 5 April 2012
Published: 14 May 2012
In the field of mouse genetics the advent of technologies like microarray based expression profiling dramatically increased data availability and sensitivity, yet these advanced methods are often vulnerable to the unavoidable heterogeneity of in vivo material and might therefore reflect differentially expressed genes between mouse strains of no relevance to a targeted experiment. The aim of this study was not to elaborate on the usefulness of microarray analysis in general, but to expand our knowledge regarding this potential “background noise” for the widely used Illumina microarray platform surpassing existing data which focused primarily on the adult sensory and nervous system, by analyzing patterns of gene expression at different embryonic stages using wild type strains and modern transgenic models of often non-isogenic backgrounds.
Wild type embryos of 11 mouse strains commonly used in transgenic and molecular genetic studies at three developmental time points were subjected to Illumina microarray expression profiling in a strain-by-strain comparison. Our data robustly reflects known gene expression patterns during mid-gestation development. Decreasing diversity of the input tissue and/or increasing strain diversity raised the sensitivity of the array towards the genetic background. Consistent strain sensitivity of some probes was attributed to genetic polymorphisms or probe design related artifacts.
Our study provides an extensive reference list of gene expression profiling background noise of value to anyone in the field of developmental biology and transgenic research performing microarray expression profiling with the widely used Illumina microarray platform. Probes identified as strain specific background noise further allow for microarray expression profiling on its own to be a valuable tool for establishing genealogies of mouse inbred strains.
Mouse models are a fundamental tool in gaining a better understanding of mammalian development in general and human pathology in particular. By studying gene expression patterns in the developing mouse embryo, important genetic pathways and signaling cascades have been revealed, for example in limb patterning (for review see), the central nervous system (for review see)and the digestive system (for review see). However, gene expression profiling techniques have leaped forward in recent years from the classical RNA in situ hybridization analysis to the more detailed and advanced methodologies of microarray analysis[5–8] providing a powerful tool for in depth analysis of genome wide differential gene expression. Yet, by increasing the assay sensitivity and being able to detect more subtle changes in expression profiles, questions arise in how far differences in the mouse genetic background affect the outcome of this advanced type of gene expression profiling. It is of concern, that strain specific differences in gene expression levels might obscure the readout of microarray results for certain genes or tissue types simply by reflecting “background noise” resulting either from true genetic strain dependent differences in expression levels yet, not relevant for a more targeted study or from a “chip artifact” originating from unfavorable hybridization conditions of some probes due to polymorphisms in their DNA sequence[7, 9, 10].
Microarray data is often obtained from cell culture approaches with a fairly homogenous genetic background[11, 12], these however lack the native context provided by in vivo assays analyzing fresh adult tissue or embryos. Historically, the mouse has often been the model organism of choice for in vivo studies and it is well known that different mouse inbred strains differ in their behavioral traits, physiology and anatomy[1, 9, 13–16]. Extensive data has been generated thus far addressing differential gene expression, especially for the Affymetrix array platform, mostly focusing on adult tissue often of the sensory and central nervous system (CNS) type, frequently restricted to only one tissue type or a couple of inbred strains selected for their suitability in behavior studies or within one strain at different time points[7–9, 17–21]. More recent in vivo approaches however combine transgenic models with tissue dissection and microarray based gene expression profiling[5, 22, 23]. Modern genetic engineering often requires crosses between several mouse strains, for example by breeding mice harboring a targeted allele to Flpe- or Cre-deleter strains, yet the production of isogenic strains for each genetically modified allele generated would exceed most funding time frames[24–28].
When studying prenatal development availability of sufficient material can be another limiting factor for expression profiling, hence the breeding advantage of hybrid or outbred strains is often considered[29–32] (http://www.harlan.com). Despite a vast amount of existing data (for a review see), it remains crucial for studies making use of genetically engineered animals to expand our current knowledge of gene expression profiling background noise to additional inbred and even outbred strains and also to a spectrum of embryonic time points, ideally for all microarray platforms as the outcome of expression profiling is clearly dependent on the platform used[7, 33, 34].
With the ultimate aim to complement existing data, using the Illumina microarray platform, we performed a comparative analysis across several commonly used mouse strains in transgenic research (C57BL/6J, 129 S2/SvHsd, FVB/NHanTMHsd and Hsd:ICR(CD-1)®) at three different stages of mid-gestation development and an additional comprehensive comparison across 11 strains [129 S2/SvHsd, FVB/NHanTMHsd, C3H/HeNHs, CBA/JHsd; BALB/cOlaHsd, C.B-17/IcrHanTMHsd-Prkdc scid , C57BL/6 J, B6;SJL-Tg(Col2a1-cre)1Bhr/J, 129 S4/SvJaeSor-Gt(Rosa)26Sor tm1(FLP1)Dym /J, C57BL/6J(Zp3-cre)93Knw/J, Hsd:ICR(CD-1)®] at E12.5 focusing on eviscerated embryos to provide a reference list of gene expression differences, while at the same time observing the impact of a reduction in tissue diversity and increasing genetic strain diversity on differential gene expression levels.
Results and discussion
A reference resource for gene expression profiling associated background noise with the illumina mouse WG6 v2.0 microarray in transgenic research and developmental biology
Long lists of up and down regulated genes are the general outcome of microarray based differential gene expression analysis, often with genes of unknown function or misleading gene ontology (GO) terms. Factors like polymorphisms in the genetic background are known to impact on the interpretation of differential expression profiling data sets. This problem is faced when comparing different species like primate and non-primate[36, 37] or different mouse inbred strains and expected when applying differential expression profiling to material from genetically engineered and often still non-isogenic animals or embryonic tissue. In order to generate a reference of gene expression profiling-associated background noise in wild type embryos of strains typically used for gene modification, we chose three Mus musculus musculus inbred strains most commonly used in developmental genetics, gene targeting or transgenic mouse production procedures (C57BL/6J, 129 S2/SvHsd, FVB/NHanTMHsd) along with an outbred mouse strain Hsd:ICR(CD-1)® to address the differential gene expression profile of entire embryos at three mid-gestation developmental stages (E11.5, E12.5 and E13.5), asking the question: Is there a significant strain specific difference for any probe at any given time point? The approach of embryo-pooling according to the experimental design of Korostynski et al. was chosen to generate four biological replicates for each strain and stage analyzed while minimizing the contribution of individual differences or slightest technical variations to the differential expression profile (for details refer to Experimental Design in the Material and Methods section).
According to Illumina’s probe list the Mouse WG-6_V2_0_R3_11278593 array contains a total of 45282 probe sequences and is based on a C57BL/6J genome. Many of the probes on the array are unique, while some loci are represented by multiple probes. Of all these 45282 probes subjected to differential gene expression analysis transcripts of a total of 580 probes (1.28%) at E11.5, 503 probes (1.11%) at E12.5 and 836 probes (1.85%) at E13.5 were found to be significantly differentially regulated when subjecting entire wild type embryos to this strain specific gene expression profiling (for a full list see Additional file1: Tables S1, Additional file2: Table S2 and Additional file3: Table S3). Some of the probes were found ranking in the top 20 for all three stages examined: Fcer1g, Lrrc57, Sspn, Tmem87a, Cap1, Lip1, Gramd4, Ctse, Tm7sf3, Pou6f1, LOC382555 (Hmgb1-rs18) and C920006O11Rik (unclassified RIKEN cDNA) (for details see Additional file1: Tables S1, Additional file2: Table S2 and Additional file3: Table S3, for details regarding the probe ranking refer to Experimental Design in the Material and Methods section.)
List of the top 50 targets ranked by total fold change following a strain by strain comparison after microarray based differential expression analysis for the following three data sets
E12.5 entire embryo/4 strains
E12.5 eviscerated embryos/4 strains
E12.5 eviscerated embryos/11 strains
Based on the use of different platforms, time points and tissues naturally the transcripts identified as differentially expressed are likely to differ for most datasets. Yet, members of the Serpina gene family appear for different types of analyses and tissues: Serpina3n, formerly known as Spi2-2 spi2/eb4 or M64086, was identified as a target using the Affymetrix platform in chondrogenesis and in neural studies[9, 21], while Serpina3h and Serpina1e, appeared in ranking 26 and 11 respectively for our differential expression analysis at E13.5 (see below) using the Illumina platform. All three genes belong to the clade A of serine (or cysteine) peptidase inhibitors according to the MGI databasehttp://www.informatics.jax.org/mgihome/.
Gene expression profiling with the illumina mouse WG6 v2.0 microarray chip on wild type mid-gestation embryos reflects known developmental patterns and has the potential to identify novel candidates
In an attempt to validate and further explore our microarray data we expanded our analysis from the gene expression profiling at each individual time point as described above to the three mid-gestation time points E11.5, E12.5 and E13.5, asking the following questions: Do genes group in a logical fashion according to their known expression patterns during development? And if so, can we discover new targets of potential interest simply based on similarities in their time course heat map to already well established and characterized genes?
The effect of a decrease in tissue diversity and an increase in genetic background diversity on the outcome of microarray based gene expression profiling
While only a small number (<2%) of probes did show significant (FC >1.5) differences in their expression profile at a given time point when subjecting entire embryos to a microarray based strain by strain gene expression profiling, there was concern that the expression profiles of some genes particularly those with multiple roles during development might have been “diluted” by subjecting entire embryos to microarray gene expression analysis without prior enrichment of the target tissue. To address the impact of a decrease in tissue diversity and/or increase in genetic background diversity on the outcome of microarray based gene expression profiling and hence to address the possibility of a “diluted” expression profile when performing a differential gene expression analysis of entire wild type embryos, in a separate study, we subjected E12.5 eviscerated embryos of 11 different commonly used strains in mouse genetics, immunological studies, transgenic and gene targeting approaches [129 S2/SvHsd, FVB/NHan TMHsd, C3H/HeNHs, CBA/JHsd, BALB/cOlaHsd, C.B-17/IcrHanTMHsd-Prkdc scid , C57BL/6J, B6;SJL-Tg(Col2a1-cre)1Bhr/J, 129 S4/SvJaeSor-Gt(Rosa)26Sor tm1(FLP1)Dym /J, C57BL/6-Tg(Zp3-cre)93Knw/J, Hsd:ICR(CD-1)®] to differential gene expression profiling. The E12.5 embryos were staged by the same stringent morphological criteria and then eviscerated, limiting our analysis essentially to the developing neuro/sensory, skeletal and muscular tissue. We followed the same basic experimental design as described earlier with four biological replicates per strain profile (for details refer to Experimental Design in the Material and Methods section). The pooling of three embryos per biological replicate should minimize any expression differences related to the evisceration procedure. We then conducted our analysis in two ways: First, we focused on the four strains common between this study and our previous one on entire embryos, comparing the lists of differentially regulated genes at E12.5 to address to what extent a decrease in tissue diversity impacts on differential gene expression profiling. Second, we included all 11 strains of only the eviscerated embryos at E12.5 in the analysis to see the impact of an increase in strain diversity on differential gene expression profiling. From these comparisons, we made four major observations:
Firstly, a list of 503 targets (1.11% of total probes on the array) with a fold change (FC) >1.5 derived from a strain by strain comparison of the expression profile of entire embryos at E12.5 and a list of 3403 targets (7.5% of total probes on the array) with a FC >1.5 derived from a strain by strain comparison of the expression profile of eviscerated embryos at E12.5 could be identified and ranked by total FC across the four wild type strains analyzed (129 S2/SvHsd, FVB/NHanTMHsd, C57BL/6J, B6;Hsd:ICR(CD-1)®). (For details regarding the probe ranking refer to Experimental Design in the Material and Methods section.) We could identify eight of our top 50 targets established for the expression profile of entire E12.5 wild type embryos in the top ten targets of the list generated for eviscerated E12.5 wild type embryos [See Table1 – (Rank A) Gene Name (Rank B)]: (3) Sspn (9), (8) C920006O11Rik (5), (9) Lrrc57 (7), (10) Fcer1g (2), (11) Ctse (10), (19) Tm7sf3 (1), (21) Pou6f1 (4) and (25) LOC100041516 (3). Similarly we located nine of our top 50 targets for the eviscerated embryos in the top ten ranking targets of a list generated for the expression profile of entire embryos[See Table1 – (Rank B) Gene Name (Rank A)]: (2) Fcer1g (10), (5) C920006O11Rik (8), (7) Lrrc57 (9), (9) Sspn (3), (14) LOC382555 (1), (15) Tmem87a (6), (34) Cap1 (7), (43) Gramd4 (2) and (49) EG384179 (5). For further details see Table1 and Additional file2: Table S2 and Additional file5: Table S4.
Secondly, addressing the aspect of a “diluted” expression profile: Only 50% of the top 50 ranking targets in the list of E12.5 eviscerated embryos could be found to be significantly differentially expressed (FC > 1.5) between the four strains at E12.5 when subjecting the entire embryo to this analysis, while 41 of the 50 top ranking targets (82%) of the 503 targets listed for the entire E12.5 wild type embryos would be detected as significantly differentially expressed in the list of eviscerated E12.5 embryos (limited to the top 500 targets for comparability), suggesting that some differential expression was indeed lost through “dilution” if the tissue is too heterogeneous and/or the target is naturally expressed in multiple tissues of the embryo. Hence the purer the tissue type analyzed, the more likely all relevant targets can be identified with current array based gene expression profiling, making the ideal source a combination of tissue micro dissection along with sorting of gene specific fluorescence labeled cells (Lufkin Lab, work in progress).
Panther gene ontology (GO) analysis showing the percentage of the number (#) of genes for each GO term based on the total number of genes for each of the three data sets analyzed
(A) E12.5 entire embryos/4 strains
(B) E12.5 eviscerated embryos/4 strains
(C) E12.5 eviscerated embryos/11 strains
cell adhesion (GO:0007155)
cell communication (GO:0007154)
cell cycle (GO:0007049)
cellular component organization (GO:0016043)
cellular process (GO:0009987)
developmental process (GO:0032502)
generation of precursor metabolites and energy (GO:0006091)
homeostatic process (GO:0042592)
immune system process (GO:0002376)
metabolic process (GO:0008152)
regulation of biological process (GO:0050789)
response to stimulus (GO:0050896)
system process (GO:0050896)
Total # genes
Total # processed hits
Shows the sequences of the amplification primers used for PCR
PCR primers (5'-3')
Amplicon Size (bp)
In summary, while a decrease in tissue diversity as well as an increase in strain diversity raises the number of probes showing a differential signal, one has to bear in mind that in a typical expression profiling study, such as the comparison of loss-of gene function versus wild type littermates or drug treated versus untreated cohorts, the actual fold changes of the identified true targets might by far outweigh the strain specific signal differences. However, the lists generated here for wild type embryos are meant to serve as guide and resource reference tool for possible gene expression profiling background noise using the Illumina Mouse WG-6 v2.0 and possibly the preceding Mouse Ref-8 v2.0 array platform in gene expression studies performed on mid-gestation embryos for strains classically used in genetic engineering.
Differential gene expression reflects the origin of inbred strains and can serve as valuable tool to establish strain ontology relationships
Most mouse inbred strains available in laboratories today can be traced back to strains established by William Castle, Abbie Lathrop, Clarence Cook Little and Halsey J. Bagg[1, 47] however, making use of polymorphisms and mutations abundant in the genome and the evolving technology, todays available inbred strains can be clearly distinguished not only by coat color but also by their DNA sequence[48–51]. Studies have shown the usefulness of a combination of quantitative trait locus (QTL), single nucleotide polymorphism (SNP) and gene expression data. A paper by Petkov making use of SNPs and QTL analysis displays a mouse family tree with seven distinct groups. While BALB/c, CBA and C3H substrains are all found within group 1, FVB/N in group 2, C57BL/6 in group 4 and 129 substrains in group 5 of his classification.
Here, we have subjected 11 mouse strains to microarray based differential expression analysis [129 S2/SvHsd, FVB/NHanTMHsd, C3H/HeNHs, CBA/JHsd, BALB/cOlaHsd, C.B-17/IcrHanTMHsd-Prkdc scid , C57BL/6J, B6;SJL-Tg(Col2a1-cre)1Bhr/J, 129 S4/SvJaeSor-Gt(Rosa)26Sor tm1(FLP1)Dym /J, C57BL/6-Tg(Zp3-cre)93Knw/J, Hsd:ICR(CD-1)®]. For the 8805 differentially expressed probes in this 11 strains comparison (19% of total probes on the array) we can find the biological replicates clustering tightly by strain origin when subjecting the data to TreeView analysishttp://rana.lbl.gov/EisenSoftware.htm indicating on one side the accuracy of our assay as well as reiterating the fact that besides polymorphism on the DNA level, clear differences in gene expression have evolved for a small subset of genes in these strains and substrains (Figure6). The clustering observed in our study is supported by the SNP and QTL based study. Based on array clustering, we can define two major groups of mouse strains in our study:
I BALB/cOlaHsd, C.B-17/IcrHanTMHsd-Prkdc scid , CBA/JHsd and C3H/HeNHs.
Within group II we can further subdivide between Group IIa, containing B6;SJL-Tg(Col2a1-cre)1Bhr/J, C57BL/6-Tg(Zp3-cre)93Knw/J, C57BL/6J all comprising a C57BL/6 genetic background, with B6;SJL-Tg(Col2a1-cre)1Bhr/J initially being generated on the SJL background and subsequently mated onto a C57BL/6J background and Group IIb represented by Group IIb-1 129 S2/SvHsd, 129 S4/SvJaeSor-Gt(Rosa)26Sor tm1(FLP1)Dym /J both of 129 genetic background and Group IIb-2 represented by Hsd:ICR(CD-1)® on one side and FVB/NHanTMHsd on the other side of the array clustering branch. ICR mice, an outbred strain, had not been subjected to Perkov’s study.
Unlike previous studies, where strain relationships have been established on the DNA level or combinations of DNA and gene expression analysis[46, 52], we demonstrate here that the function driven analysis of microarray gene expression profiling is sufficient for the accurate confirmation of the genetic ancestry of mouse strains.
Since the results of microarray gene expression profiling can be impacted on by variations in the strain of mouse used, we aimed to provide a resource reference list of probes contributing to strain differences or “noise” when subjecting non-isogenic tissue from any of the frequently used inbred strains in mouse gene targeting or even an outbred strain, to microarray based differential gene expression profiling using the Illumina Mouse WG6 v2.0 microarray chip. We subjected entire embryos as well as eviscerated embryos to this study. This reduction in tissue diversity was reflected in a raised number of significantly differentially expressed genes, a number likely to increase with further reduction of tissue heterogenity. The data retrieved from our extensive expression profiling using the Illumina platform is robust and reflects the anticipated gene expression patterns of known and well-characterized patterning and structural genes during mid-gestation development. For a small number of probes, the data has been impacted on by probe design artifacts (probes not allocated in the actual gene) or natural genetic polymorphisms between mouse strains reflecting “background noise”. This is a problem researchers should be aware of and future array platforms would need to adjust for in order to be a reasonable experimental choice compared to the rapidly evolving RNA-seq technology, which, once affordable, should allow for a more unbiased expression profiling analysis in the near future (see review by).
At the current point in time, owing to the overall robustness of the array, heat map profile similarities make way for the discovery of genes with previously unknown function. Lastly, microarray based differential gene expression analysis on its own can serve as a tool to establish strain ontology relationships similar to SNP and QTL analysis.
All embryos were carefully staged according to morphological criteria and only embryos showing all morphological criteria as displayed for each stage E11.5, E12.5 and E13.5 were included in this study. Three age-matched embryos of undefined gender were combined to form one biological replicate. Altogether a total of four biological replicates (12 embryos) were analyzed for each strain and developmental time point and subjected to expression profiling using the MouseWG-6_V2_0_R3_11278593 array from Illumina. We chose this approach of embryo pooling according to the experimental design by Korostynski et al. to minimize the contribution of individual differences or slight technical variations from embryo dissections/eviscerations to the read out of the differential expression analysis. The fold change (FC) of expression for the four samples per strain was averaged at each of the three time points and subjected to a strain-by-strain comparison for a given developmental stage. The genes were then ranked according to the highest total fold change across all strains. Only genes with a FC >1.5 between any of the strains were considered as significantly differentially expressed.
Ethics statement, mouse husbandry and tissue collection
All animal procedures were performed according to the Singapore A*STAR Biopolis Biological Resource Center (BRC) Institutional Animal Care and Use Committee (IACUC) guidelines and the IACUC protocols employed were reviewed and approved by the aforementioned committee before any animal procedures were undertaken for this study described here (IACUC Protocol No: 080348 and 080377). The mouse strains used in this study were maintained and provided by the A*STAR Biopolis Biological Resource Center (129 S2/SvHsd, FVB/NHanTMHsd, C3H/HeNHsd, CBA/JHsd, BALB/cOlaHsd and C.B-17/IcrHanTMHsd-Prkdc scid , C57BL/6J, Hsd:ICR(CD-1)®) or directly imported from Jackson Laboratories (B6;SJL-Tg(Col2a1-cre)1Bhr/J #003554, 129 S4/SvJaeSor-Gt(Rosa)26Sor tm1(FLP1)Dym /J #003946, C57BL/6-Tg(Zp3-cre)93Knw/J #003651) and then maintained according to Jackson Laboratories guidelines specific for each strain. Males and females of each respective strain were intermated to generate E11.5, E12.5 and E13.5 embryos, with E0.5 being defined as the day the vaginal plug was detected. The mouse embryos were subsequently harvested in ice-cold Leibovitz’s L-15 medium (Gibco) at E11.5, E12.5 or E13.5 and critically staged applying morphological criteria as described. For the study focusing on differential gene expression in eviscerated embryos the E12.5 embryos were dissected free of all internal organs. Embryos were then dissociated for RNA isolation.
RNA extraction and microarray analysis
Fresh mouse embryonic tissues were rapidly dissociated into small clumps in L-Leibovitz medium by repeated pipetting. The small tissue clumps were collected by centrifugation at 2000 rpm for 5 minutes. Applying the TRIzol/RNeasy hybrid method, TRIzol (Invitrogen) was added to the pelleted tissues at approximately 1 ml per 50 mg tissue for homogenization. The homogenate was stored in −80°C for no longer than 3 months before RNA extraction. During the RNA extraction, 0.2 ml chloroform was added per 1 ml of homogenate and the top aqueous phase was gained after centrifugation at 12,000 g for 15 minutes at 4°C. The aqueous phase was loaded onto a gDNA Eliminator spin column based on the DNA removal capacity of the column. Subsequent steps were done according to the RNeasy Plus Mini kit (Qiagen) following the manufacturer’s instruction. Total RNA extracted from fresh mouse embryonic tissues was quantified by a NanoDrop ND-1000 Spectrophotometer. For quality control, RNA was diluted to the working concentration of the Agilent RNA 6000 Nano Kit and 1 ul of the diluted RNA sample was run on the Nano chip using an Agilent 2100 electrophoresis Bioanlyzer. The Nano chip assay was performed according to the manufacturer’s instructions. The quality of total RNA was assessed primarily via the profile of the electropherogram and secondarily by RNA integrity number (RIN) generated by the Bioanalyzer software, only samples with a RIN > 9.4 were included in the study (see Additional file5: Table S4). The RNA concentration given by Nanodrop and Nano chip coincided. For each biological replicate 50 ng of high quality total RNA was labeled using Illumina TotalPrep RNA Amplification kit from Ambion and hybridized on Illumina’s MouseWG-6_V2_0_R3_11278593 array according to the manufacturer’s instructions. Microarray data was normalized using GenomeStudio (background subtraction, rank invariant normalization). Any negative values were replaced by the value “1” for fold change calculation and then all signals were Log2 transformed. Linear modeling of the transformed data was performed using Limma in R with the Benjamini and Hochberg correction. The model used included developmental stage, strain and batch factors, where appropriate. P value and FDR were obtained for coefficient of each factor depending on the comparison, ie. coefficients for developmental stage or strain. Fold difference was calculated by taking the ratio of the individual signals with higher expressing value over the signals with the lower expression value. Only expression levels with at least 1.5× fold difference and a false discovery rate (FDR) below 5% were considered as significantly differentially regulated. Microarray data was hierarchical clustered by average linkage clustering with uncentered correlation using Cluster and the heatmap was generated with R. Strain ontology relationships were established with TreeViewhttp://rana.lbl.gov/EisenSoftware.htm.
Genomic DNA extraction and sequencing over illumina probes
To extract genomic DNA embryos were removed from the yolk sac, briefly washed in 1x PBS and placed in 500 ul PKDB digestion buffer (50 mM Tris–HCl (pH 8.0), 200 mM NaCl, 5 mM EDTA, 1% SDS) containing 1 mg/ml proteinase K and incubated at 55°C overnight. Digested samples were extracted with an equal volume of phenol-chloroform, DNA was precipitated with ethanol and washed with 70% ethanol. DNA pellets were air-dried and resuspended in DNase free water.
Sequences 500 nucleotide upstream and downstream of each Illumina probe sequence were obtained from UCSC using BLAST-Like Alignment Tool (BLAT) at (http://genome.ucsc.edu/). For primer design, sequences obtained from BLAT were imported into the Vector NTI Advance 10 (Invitrogen, CA, USA) software and primer pairs flanking the Illumina probe sequences were designed carefully avoiding similarities with repetitive sequences or other loci in the genome. Table3 shows the sequences of the amplification primers used for PCR.
PCR products were generated using Platinum® Pfx DNA polymerase (Invitrogen) and PCR products were purified on a MinElute PCR purification spin column (Qiagen, Hagen, Germany) following the manufacturer's instructions. The DNA was eluted in 30 ul of elution buffer and sent for sequencing. Sequences were aligned against and compared with the sequences of respective Illumina probes using Vector NTI .
cDNA synthesis and real-time qPCR analysis
Total RNA from mouse embryos was isolated, assessed and quantified as described above. First strand cDNA was synthesized from 5ug of total RNA by reverse transcription PCR at 50°C for 30 min in the presence of 200 ng/ul random hexamers and 10 mM each of dNTPs and RevertAid™ Premium Enzyme mix (Fermentas). The synthesized cDNAs were adjusted to 50 ng/ul of which 100 ng was used in a final volume of 20 ul. Each sample was run in triplicate on an Applied Biosystems 7500 Real-Time PCR systems using Maxima™ SYBR Green/ROX qPCR master mix (Fermentas). The Hprt gene had no significant variation in expression across the four mouse strains and therefore was used as endogenous control for normalization . Expression level was evaluated relative to a calibrator according to the 2-ΔΔCt method for quantitation.
Histology and RNA in situ hybridization
Mouse embryos were processed by fixation with 4% paraformaldehyde (PFA), overnight at 4°C, then washed with 1x PBS, dehydrated in graded ethanol and embedded in paraffin. A Leica RM 2165 microtome was used to make 10 um thick paraffin sections. Sectioned in situ hybridization was performed as described in. The cDNA of 0.8 kb Prl3b1 (IMAGE clone: 30787415) linearized with EcoRV and Ninl cDNA of 4.2 kb (IMAGE clone: 30615484) linearized with EcoRI were used as templates for synthesizing antisense DIG-labeled Prl3b1 and Ninl RNA probes (DIG RNA labeling kit, Roche).
We are grateful to all A*STAR/BRC staff, in particular to Sharon Heng Yee Choy and all members of the Lufkin Lab for support and interesting discussions in particular Song Jie and Sumantra Chatterjee. This work was supported by the Agency for Science Technology and Research (A*STAR) Singapore.
- Green EL: Biology of the laboratory mouse. 1966, New York: Dover PublicationGoogle Scholar
- Niswander L: Interplay between the molecular signals that control vertebrate limb development. Int J Dev Biol. 2002, 46: 877-881.PubMedGoogle Scholar
- Zervas M, Blaess S, Joyner AL: Classical embryological studies and modern genetic analysis of midbrain and cerebellum development. Curr Top Dev Biol. 2005, 69: 101-138.PubMedView ArticleGoogle Scholar
- Roberts DJ: Molecular mechanisms of development of the gastrointestinal tract. Dev Dyn. 2000, 219: 109-120. 10.1002/1097-0177(2000)9999:9999<::AID-DVDY1047>3.3.CO;2-Y.PubMedView ArticleGoogle Scholar
- Vokes SA, Ji H, Wong WH, McMahon AP: A genome-scale analysis of the cis-regulatory circuitry underlying sonic hedgehog-mediated patterning of the mammalian limb. Genes Dev. 2008, 22: 2651-2663. 10.1101/gad.1693008.PubMedPubMed CentralView ArticleGoogle Scholar
- Amit G, Shukha K, Gavriely N, Intrator N: Respiratory modulation of heart sound morphology. Am J Physiol Heart Circ Physiol. 2009, 296: H796-H805. 10.1152/ajpheart.00806.2008.PubMedView ArticleGoogle Scholar
- Bottomly D, Walter NA, Hunter JE, Darakjian P, Kawane S, Buck KJ, Searles RP, Mooney M, McWeeney SK, Hitzemann R: Evaluating gene expression in C57BL/6 J and DBA/2 J mouse striatum using RNA-Seq and microarrays. PLoS One. 2011, 6: e17820-10.1371/journal.pone.0017820.PubMedPubMed CentralView ArticleGoogle Scholar
- Jelcick AS, Yuan Y, Leehy BD, Cox LC, Silveira AC, Qiu F, Schenk S, Sachs AJ, Morrison MA, Nystuen AM, et al: Genetic variations strongly influence phenotypic outcome in the mouse retina. PLoS One. 2011, 6: e21858-10.1371/journal.pone.0021858.PubMedPubMed CentralView ArticleGoogle Scholar
- Sandberg R, Yasuda R, Pankratz DG, Carter TA, Del Rio JA, Wodicka L, Mayford M, Lockhart DJ, Barlow C: Regional and strain-specific gene expression mapping in the adult mouse brain. Proc Natl Acad Sci USA. 2000, 97: 11038-11043. 10.1073/pnas.97.20.11038.PubMedPubMed CentralView ArticleGoogle Scholar
- Malone JH, Oliver B: Microarrays, deep sequencing and the true measure of the transcriptome. BMC Biol. 2011, 9: 34-10.1186/1741-7007-9-34.PubMedPubMed CentralView ArticleGoogle Scholar
- Boimel PJ, Cruz C, Segall JE: A functional in vivo screen for regulators of tumor progression identifies HOXB2 as a regulator of tumor growth in breast cancer. Genomics. 2011, 98: 164-172. 10.1016/j.ygeno.2011.05.011.PubMedPubMed CentralView ArticleGoogle Scholar
- Zhang L, Ju X, Cheng Y, Guo X, Wen T: Identifying Tmem59 related gene regulatory network of mouse neural stem cell from a compendium of expression profiles. BMC Syst Biol. 2011, 5: 152-10.1186/1752-0509-5-152.PubMedPubMed CentralView ArticleGoogle Scholar
- Crawley JN, Belknap JK, Collins A, Crabbe JC, Frankel W, Henderson N, Hitzemann RJ, Maxson SC, Miner LL, Silva AJ, et al: Behavioral phenotypes of inbred mouse strains: implications and recommendations for molecular studies. Psychopharmacology (Berl). 1997, 132: 107-124. 10.1007/s002130050327.View ArticleGoogle Scholar
- Gerlai R: Gene-targeting studies of mammalian behavior: is it the mutation or the background genotype?. Trends Neurosci. 1996, 19: 177-181. 10.1016/S0166-2236(96)20020-7.PubMedView ArticleGoogle Scholar
- Peirce JL, Derr R, Shendure J, Kolata T, Silver LM: A major influence of sex-specific loci on alcohol preference in C57Bl/6 and DBA/2 inbred mice. Mamm Genome. 1998, 9: 942-948. 10.1007/s003359900904.PubMedView ArticleGoogle Scholar
- Pavlidis P, Noble WS: Analysis of strain and regional variation in gene expression in mouse brain. Genome Biol. 2001, 2: 1-15. RESEARCH0042View ArticleGoogle Scholar
- Fernandes C, Paya-Cano JL, Sluyter F, D'Souza U, Plomin R, Schalkwyk LC: Hippocampal gene expression profiling across eight mouse inbred strains: towards understanding the molecular basis for behaviour. Eur J Neurosci. 2004, 19: 2576-2582. 10.1111/j.0953-816X.2004.03358.x.PubMedView ArticleGoogle Scholar
- de Jong S, Fuller TF, Janson E, Strengman E, Horvath S, Kas MJ, Ophoff RA: Gene expression profiling in C57BL/6 J and A/J mouse inbred strains reveals gene networks specific for brain regions independent of genetic background. BMC Genomics. 2010, 11: 20-10.1186/1471-2164-11-20.PubMedView ArticleGoogle Scholar
- Hoffman BG, Zavaglia B, Witzsche J, Ruiz De Algara T, Beach M, Hoodless PA, Jones SJ, Marra MA, Helgason CD: Identification of transcripts with enriched expression in the developing and adult pancreas. Genome Biol. 2008, 9: 99-10.1186/gb-2008-9-6-r99.View ArticleGoogle Scholar
- Korostynski M, Kaminska-Chowaniec D, Piechota M, Przewlocki R: Gene expression profiling in the striatum of inbred mouse strains with distinct opioid-related phenotypes. BMC Genomics. 2006, 7: 146-10.1186/1471-2164-7-146.PubMedPubMed CentralView ArticleGoogle Scholar
- Suzuki Y, Nakayama M: Differential profiles of genes expressed in neonatal brain of 129X1/SvJ and C57BL/6 J mice: A database to aid in analyzing DNA microarrays using nonisogenic gene-targeted mice. DNA Res. 2003, 10: 263-275. 10.1093/dnares/10.6.263.PubMedView ArticleGoogle Scholar
- Sansom SN, Griffiths DS, Faedo A, Kleinjan DJ, Ruan Y, Smith J, van Heyningen V, Rubenstein JL, Livesey FJ: The level of the transcription factor Pax6 is essential for controlling the balance between neural stem cell self-renewal and neurogenesis. PLoS Genet. 2009, 5: e1000511-10.1371/journal.pgen.1000511.PubMedPubMed CentralView ArticleGoogle Scholar
- Cameron TL, Belluoccio D, Farlie PG, Brachvogel B, Bateman JF: Global comparative transcriptome analysis of cartilage formation in vivo. BMC Dev Biol. 2009, 9: 20-10.1186/1471-213X-9-20.PubMedPubMed CentralView ArticleGoogle Scholar
- Wright S: Systems of mating. V. General considerations. Genetics. 1921, 6: 167-178.PubMedPubMed CentralGoogle Scholar
- Wright S: Systems of mating IV. The effects of selection. Genetics. 1921, 6: 162-166.PubMedPubMed CentralGoogle Scholar
- Wright S: Systems of mating III. Assortative mating based on somatic resemblance. Genetics. 1921, 6: 144-161.PubMedPubMed CentralGoogle Scholar
- Wright S: Systems of mating II. The effects of inbreeding on the genetic composition of a population. Genetics. 1921, 6: 124-143.PubMedPubMed CentralGoogle Scholar
- Wright S: Systems of mating I. The biometric relations between parent and offspring. Genetics. 1921, 6: 111-123.PubMedPubMed CentralGoogle Scholar
- Roberts RC: The effects on litter size of crossing lines of mice inbred without selection. Genet Res. 1960, 1: 239-252. 10.1017/S0016672300000227.View ArticleGoogle Scholar
- Barnett SAaC E: 'Heterosis' in F1 mice in a cold environment. Genet Res. 1960, 1: 25-38. 10.1017/S0016672300000045.View ArticleGoogle Scholar
- McCarthy JC: The effect of litter size of crossing inbred strains of mice. Genetics. 1965, 51: 217-222.PubMedPubMed CentralGoogle Scholar
- Silver LM: Mouse genetics. 1995, Oxford: New YorkGoogle Scholar
- McCall MN, Murakami PN, Lukk M, Huber W, Irizarry RA: Assessing affymetrix GeneChip microarray quality. BMC Bioinformatics. 2011, 12: 137-10.1186/1471-2105-12-137.PubMedPubMed CentralView ArticleGoogle Scholar
- Edwards YJ, Bryson K, Jones DT: A meta-analysis of microarray gene expression in mouse stem cells: redefining stemness. PLoS One. 2008, 3: e2712-10.1371/journal.pone.0002712.PubMedPubMed CentralView ArticleGoogle Scholar
- Greenhall JA, Zapala MA, Caceres M, Libiger O, Barlow C, Schork NJ, Lockhart DJ: Detecting genetic variation in microarray expression data. Genome Res. 2007, 17: 1228-1235. 10.1101/gr.6307307.PubMedPubMed CentralView ArticleGoogle Scholar
- Caceres M, Lachuer J, Zapala MA, Redmond JC, Kudo L, Geschwind DH, Lockhart DJ, Preuss TM, Barlow C: Elevated gene expression levels distinguish human from non-human primate brains. Proc Natl Acad Sci USA. 2003, 100: 13030-13035. 10.1073/pnas.2135499100.PubMedPubMed CentralView ArticleGoogle Scholar
- Karaman MW, Houck ML, Chemnick LG, Nagpal S, Chawannakul D, Sudano D, Pike BL, Ho VV, Ryder OA, Hacia JG: Comparative analysis of gene-expression patterns in human and African great ape cultured fibroblasts. Genome Res. 2003, 13: 1619-1630. 10.1101/gr.1289803.PubMedPubMed CentralView ArticleGoogle Scholar
- Zhou D, Wang J, Zapala MA, Xue J, Schork NJ, Haddad GG: Gene expression in mouse brain following chronic hypoxia: role of sarcospan in glial cell death. Physiol Genomics. 2008, 32: 370-379.PubMedView ArticleGoogle Scholar
- James CG, Stanton LA, Agoston H, Ulici V, Underhill TM, Beier F: Genome-wide analyses of gene expression during mouse endochondral ossification. PLoS One. 2010, 5: e8693-10.1371/journal.pone.0008693.PubMedPubMed CentralView ArticleGoogle Scholar
- Vlaeminck-Guillem V, Carrere S, Dewitte F, Stehelin D, Desbiens X, Duterque-Coquillaud M: The Ets family member Erg gene is expressed in mesodermal tissues and neural crests at fundamental steps during mouse embryogenesis. Mech Dev. 2000, 91: 331-335. 10.1016/S0925-4773(99)00272-5.PubMedView ArticleGoogle Scholar
- McMahon AP, Aronow BJ, Davidson DR, Davies JA, Gaido KW, Grimmond S, Lessard JL, Little MH, Potter SS, Wilder EL, Zhang P: GUDMAP: the genitourinary developmental molecular anatomy project. J Am Soc Nephrol. 2008, 19: 667-671. 10.1681/ASN.2007101078.PubMedView ArticleGoogle Scholar
- Simmons DG, Rawn S, Davies A, Hughes M, Cross JC: Spatial and temporal expression of the 23 murine Prolactin/Placental Lactogen-related genes is not associated with their position in the locus. BMC Genomics. 2008, 9: 352-10.1186/1471-2164-9-352.PubMedPubMed CentralView ArticleGoogle Scholar
- Satokata I, Ma L, Ohshima H, Bei M, Woo I, Nishizawa K, Maeda T, Takano Y, Uchiyama M, Heaney S, et al: Msx2 deficiency in mice causes pleiotropic defects in bone growth and ectodermal organ formation. Nat Genet. 2000, 24: 391-395. 10.1038/74231.PubMedView ArticleGoogle Scholar
- Shimogori T, Lee DA, Miranda-Angulo A, Yang Y, Wang H, Jiang L, Yoshida AC, Kataoka A, Mashiko H, Avetisyan M, et al: A genomic atlas of mouse hypothalamic development. Nat Neurosci. 2010, 13: 767-775. 10.1038/nn.2545.PubMedPubMed CentralView ArticleGoogle Scholar
- Gu W, Wells AL, Pan F, Singer RH: Feedback regulation between zipcode binding protein 1 and beta-catenin mRNAs in breast cancer cells. Mol Cell Biol. 2008, 28: 4963-4974. 10.1128/MCB.00266-08.PubMedPubMed CentralView ArticleGoogle Scholar
- Hitzemann R, Malmanger B, Reed C, Lawler M, Hitzemann B, Coulombe S, Buck K, Rademacher B, Walter N, Polyakov Y, et al: A strategy for the integration of QTL, gene expression, and sequence analyses. Mamm Genome. 2003, 14: 733-747. 10.1007/s00335-003-2277-9.PubMedView ArticleGoogle Scholar
- Silver LM: Mouse Genetics - Concepts and Applications. 1995, New York, USA: Oxford University PressGoogle Scholar
- Beck JA, Lloyd S, Hafezparast M, Lennon-Pierce M, Eppig JT, Festing MF, Fisher EM: Genealogies of mouse inbred strains. Nat Genet. 2000, 24: 23-25. 10.1038/71641.PubMedView ArticleGoogle Scholar
- Wade CM, Daly MJ: Genetic variation in laboratory mice. Nat Genet. 2005, 37: 1175-1180. 10.1038/ng1666.PubMedView ArticleGoogle Scholar
- Yang H, Wang JR, Didion JP, Buus RJ, Bell TA, Welsh CE, Bonhomme F, Yu AH, Nachman MW, Pialek J, et al: Subspecific origin and haplotype diversity in the laboratory mouse. Nat Genet. 2011, 43: 648-655. 10.1038/ng.847.PubMedPubMed CentralView ArticleGoogle Scholar
- Keane TM, Goodstadt L, Danecek P, White MA, Wong K, Yalcin B, Heger A, Agam A, Slater G, Goodson M, et al: Mouse genomic variation and its effect on phenotypes and gene regulation. Nature. 2011, 477: 289-294. 10.1038/nature10413.PubMedPubMed CentralView ArticleGoogle Scholar
- Petkov PM, Ding Y, Cassell MA, Zhang W, Wagner G, Sargent EE, Asquith S, Crew V, Johnson KA, Robinson P, et al: An efficient SNP system for mouse genome scanning and elucidating strain relationships. Genome Res. 2004, 14: 1806-1811. 10.1101/gr.2825804.PubMedPubMed CentralView ArticleGoogle Scholar
- Kaufman MH: The Atlas of Mouse Development. 1992, Academic PressGoogle Scholar
- Smyth GK: Linear models and empirical bayes methods for assessing differential expression in microarray experiments. Stat Appl Genet Mol Biol. 2004, 3: 1-25. Article3Google Scholar
- Eisen MB, Spellman PT, Brown PO, Botstein D: Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci USA. 1998, 95: 14863-14868. 10.1073/pnas.95.25.14863.PubMedPubMed CentralView ArticleGoogle Scholar
- Yap SP, Xing X, Kraus P, Sivakamasundari V, Chan HY, Lufkin T: Generation of mice with a novel conditional null allele of the Sox9 gene. Biotechnol Lett. 2011, 33: 1551-1558. 10.1007/s10529-011-0608-6.PubMedView ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License(http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.