Inter-subspecies mouse F1 hybrid embryonic stem cell lines newly established for studies of allelic imbalance in gene expression

Abstract

Allele-specific monoallelic gene expression is a unique phenomenon and a great resource for analyzing gene regulation. To study this phenomenon, we established new embryonic stem (ES) cell lines derived from F1 hybrid blastocysts from crosses between four mouse subspecies (Mus musculus domesticus, C57BL/6; M. musculus molossinus, MSM/Ms; M. musculus musculus, PWK; M. musculus castaneus, HMI/Ms) and analyzed the expression levels of undifferentiated pluripotent stem cell markers and karyotypes of each line. To demonstrate the utility of our cell lines, we analyzed the allele-specific expression pattern of the Inpp5d gene as an example. The allelic expression depended on the parental alleles; this dependence could be a consequence of differences in compatibility between cis- and trans-elements of the Inpp5d gene from different subspecies. The use of parental mice from four subspecies greatly enhanced genetic polymorphism. The F1 hybrid ES cells retained this polymorphism not only in the Inpp5d gene, but also at a genome-wide level. As we demonstrated for the Inpp5d gene, the established cell lines can contribute to the analysis of allelic expression imbalance based on the incompatibility between cis- and trans-elements and of phenotypes related to this incompatibility.

Introduction

Genomic imprinting is one of the most intriguing phenomena of gene expression regulation in mammals. To understand its mechanisms, research has used F1 offspring between two genetically distant inbred mouse strains; our group has analyzed genomic imprinting in F1 hybrid mice from a cross between the C57BL/6 (B6) and MSM/Ms (MSM) strains [1,2,3]. This experimental system allowed us to discern the maternal and paternal allele expression by using SNPs on each strain’s chromosomes. We have also established an in vitro neurogenesis system based on F1-blastocyst-derived embryonic stem (ES) cells [2, 3].

While analyzing imprinted genes in F1 mice and ES cells, we noticed that monoallelic gene expression differs from genomic imprinting; there are gene loci where only the B6 alleles are expressed in the MSM (mother)/ B6 (father) cross or the B6 (mother)/ MSM (father) cross, but in some cases only MSM alleles are expressed [3]. We called this phenomenon “strain-specific monoallelic gene expression”. We have estimated the number of genes with this type of expression to be one order of magnitude larger than that of imprinted genes. This type of expression is often specific to tissue or developmental stage [4]. This phenomenon was first reported at the genomic scale in 2012 in mouse [5], followed by several other reports also in mouse [4,5,6,7,8,9], including ours [3].

The mechanisms underlying strain-specific monoallelic gene expression in the F1 offspring are largely unknown, but it is thought to be caused by incompatibility between cis-elements such as promoters or enhancers and trans-elements such as transcription factors or mediators [5, 7, 10]. For example, if a promoter on one chromosome has stronger binding ability to a trans-acting factor than that on the other homologous chromosome, the trans-acting factor would bind mainly to the former promoter, resulting in transcription biased toward that allele. This type of monoallelic gene expression is considered to be an extreme form of allelic expression imbalance caused by cis- and trans-element incompatibility. In the process of establishing inbred mouse strains, only the strains with the cis–trans relations adequate for sustaining life survive. When genetically distant strains are crossed, the compatibility between cis- and trans-factors may be compromised, resulting in the expression of only one of the two alleles.

The utility of mouse inbred strains as models for human genetic diseases is occasionally constrained due to the inherent polymorphism of the human genome and the heightened complexity of interactions among loci compared to mouse inbred strains. Notably, allelically biased gene expression observed in F1 individuals resulting from crosses between mouse inbred strains, including strain-specific monoallelic expression, offer valuable models for investigating gene expression mechanisms relevant to human genetic diseases, particularly sporadic ones. The generation of F1 individuals requires breeding inbred strains each time they are required for research, while ES cells, which, once established, can be propagated as needed, present a more accessible alternative. Moreover, various in vitro protocols for differentiating ES cells into desired cell types or tissues are available. Consequently, ES cells derived from F1 hybrid blastocysts emerge as a convenient resource for the initial analysis of gene expression disorders in F1 individuals.

Here we report the establishment, characterization, and availability of hybrid ES cell lines derived from crosses between the laboratory standard B6 and wild-derived strains belonging to subspecies different from B6, in addition to B6/MSM crosses that we have already reported [2, 3]. The use of additional subspecies will increase the phenotypic variety of F1 hybrid ES cells caused by the incompatibilities of cis- and trans-elements and be an excellent strategy for these analyses.

Materials and Methods

Mouse strains

C57BL/6NJcl mice (B6, M. musculus domesticus) were purchased from CLEA Japan Inc. (Tokyo, Japan). MSM/MsRbrc (MSM, RBRC00209; M. musculus molossinus), PWK/RpRbrc (PWK, RBRC00213; M. musculus musculus), and HMI/MsRbrc (HMI, RBRC00657; M. musculus castaneus) were bred and maintained in a specific-pathogen-free state under a 12L:12D light cycle at the barrier facility of RIKEN BRC [11, 12]. All mice were fed with commercial gamma-irradiated food (CE-2, CLEA Japan Inc.). Filtered autoclaved water was freely available. Experimental procedures involving animals were approved by the Institutional Animal Care and Use Committee of the RIKEN Tsukuba Branch under the numbers of T2023-002 and T2023-004.

Cell lines

We used ES cell lines derived from F1 hybrid blastocysts from crosses between B6 and MSM, PWK, or HMI. We used designations for the ES cell lines using uppercase abbreviations of the two parental strains (maternal strain listed first) of the F1 hybrid blastocysts: BM derived from (B6 × MSM) F1 blastocysts [3], MB from (MSM × B6) F1 blastocysts [2], BP from (B6 × PWK) F1 blastocysts, PB from (PWK × B6) F1 blastocysts, BH from (B6 × HMI) F1 blastocysts, and HB from (HMI × B6) F1 blastocysts (Table 1). The cell lines 6NK-7 (B6) [13] and Mol/MSM-1 [14] were used as the control parental ES cell lines for B6 and MSM, respectively.

Table 1. Designations for ES cell lines.

Cell line	Source of F1 hybrid blastocyst
BM	(B6 × MSM) F1
MB	(MSM × B6) F1
BP	(B6 × PWK) F1
PB	(PWK × B6) F1
BH	(B6 × HMI) F1
HB	(HMI × B6) F1

Superovulation

Females were abdominally injected with the appropriate doses of hormones as described previously [11]: C57BL/6N at 13 weeks of age with 0.1 ml anti-inhibin serum (AIS) followed by 7.5 IU human chorionic gonadotropin (hCG), PWK at 7–8 weeks of age with 7.5 IU equine chorionic gonadotropin (eCG) and 7.5 IU hCG, and HMI at 8–9 weeks of age with 0.1 ml AIS and 7.5 IU hCG. AIS/eCG and hCG were injected 48 h apart.

In vitro fertilization (IVF)

IVF was performed using epididymal spermatozoa as described previously [15, 16], with slight modifications. At 16–17 h after hCG injection, cumulus-enclosed oocytes were collected from the ampulla region of the oviducts and preincubated for 1 h in 80-µl droplets of HTF medium. These droplets contained 1.25–1.5 mM reduced glutathione to enhance sperm penetration through the zona pellucida [17, 18], except for PWK spermatozoa, which had strong motility. Each droplet contained oocytes collected from the oviduct(s) of one female. Sperm masses collected from the epididymis were suspended in 200 µl of sperm preincubation medium (HTF containing 0.4 mM methyl-β-cyclodextrin [19, 20] and 0.1 mg/ml of polyvinyl alcohol instead of bovine serum albumin) and incubated at 37°C under 5% CO₂ in humidified air for 45–60 min. At the time of insemination, the preincubated spermatozoa were transferred into the droplets containing oocytes at a concentration of 200–400 spermatozoa/µl, and the droplets were incubated for 3–4 h. Oocytes were freed from spermatozoa and cumulus cells by means of a fine glass pipette and transferred into 10 µl droplets of CZB medium [21] containing 5.6 mM glucose, 0.1 mg/ml of polyvinyl alcohol, and 3.0 mg/ml of bovine serum albumin. They were cultured to blastocysts at 37°C under 5% CO₂ in humidified air. To avoid developmental arrest in vitro, some HMI × C57BL/6N and all HMI × HMI one-cell embryos were transferred into the oviducts of day 1 pseudo-pregnant females of the ICR strain and were recovered 3 days later by flushing the uteri.

Establishment of ES cell lines

ES cell lines were established from blastocysts generated by IVF as described previously [22] with some modifications. In brief, the zona pellucida was removed by treatment with acidic Tyrode’s solution, and blastocysts were cultured on mitomycin-C–treated BALB/c mouse embryonic fibroblast cells for 10–14 days. The culture medium consisted of Knockout DMEM (Invitrogen, Waltham, MA, USA), nonessential amino acids (Invitrogen), 0.1 mM β-mercaptoethanol (Sigma, St. Louis, MO, USA), 15% Knockout Serum Replacement (KSR) (Invitrogen), 1,000 U/ml ESGRO (Invitrogen), 3 µM CHIR99021, and 2 µM PD0325901 (both from Stemgent). The outgrowing inner cell mass was dissected mechanically and transferred onto new feeder cells. At the next passage, the expanded colonies were dispersed to single cells with 0.25% trypsin and seeded on a new feeder layer; this process was repeated several times, and putative ES cells were cryopreserved for later use.

ES cell culture

ES cells were cultured as previously reported [23]. KSR (15%), Leukemia inhibitory factor (LIF) (Wako, Osaka, Japan) and 2i condition (supplementation with 3 µM CHIR99021 and 1 µM PD0325901) was used.

Integrated genome browser (IGB) plots

The IGB [23] plots were generated using the RNA-seq datasets we previously published under the accession numbers DRA003816 and DRA003817 [3], and the sgr files to generate expression tracks are available upon request.

DNA extraction

Cells were lysed with lysis buffer (0.1 M Tris·HCl pH8.0, 15 mM EDTA, 0.2 M NaCl, 0.2% SDS, 0.1 mg/ml Proteinase K) overnight at 55°C, the lysates were extracted with phenol/chloroform, and DNA was precipitated with ethanol. The quality and quantity of the purified DNA were evaluated with a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Waltham MA, USA).

RNA isolation and reverse-transcriptase (RT) reaction

Total RNA was isolated with the use of RNAiso (Takara, Kusatsu, Japan) according to the manufacturer’s protocol. RNA quality and quantity were assessed with a NanoDrop ND-1000 spectrophotometer. Total RNA (1 µg) was reverse transcribed in 20 µl with SuperScript IV reverse transcriptase (Invitrogen) using random hexamer primers according to the manufacturer’s protocol. The reaction mixture was diluted with 20 µl of H₂O before PCR.

Analysis of karyotypes

The ES cells were treated with 2 µl/ml colcemid (Gibco, Waltham, MA, USA) for 1 h in the culture medium, centrifuged, and suspended in 5% KCl for 15 min. The inflated ES cells were suspended in Carnoy’s fluid (75% methanol, 25% acetic acid) and washed with it twice. The ES cells were centrifuged and resuspended in 0.5–2 ml of Carnoy’s fluid (depending on the amount of the cells), dripped onto microscope slide glasses, and air-dried. The cells were stained with 5% Giemsa stain solution (Wako).

PCR

Takara Ex Taq (Takara) or Dream Taq DNA polymerase (Thermo Fisher Scientific) was used according to the manufacturer’s protocol, using 5% or 10% (final concentration) dimethyl sulfoxide and adjusting the Mg²⁺ concentration as necessary. The following primer pairs were used to detect SNPs in RT-PCR products: Inpp5d (B6 and PWK) 5′-CTCCTGGATTCCGACTTTTTGA-3′ and 5′-CTTCTCCGTCTCCACCAAAATC-3′ (for SNPs between B6 and PWK). The following primer pairs were used to detect SNPs in the genome sequences: Inpp5d (B6 and MSM, B6 and HMI) 5′-TCTGGTGTCTCCTCTGCAGCTA-3′ and 5′-TGAGGAGCTGGCTAGGGACTAA-3′ (for SNPs between B6 and MSM; and B6 and HMI), and Inpp5d (B6 and PWK) 5′-CTCCTGGATTCCGACTTTTTGA-3′ (the same primer as for RT-PCR products) and 5′-GATGGCAGTGTCTTCCACCTTT-3′ (for SNPs between B6 and PWK).

The strain-specific SNPs are not included in these primer sequences.

Direct sequencing to estimate allelic expression ratio

To estimate the expression difference between two alleles of F1 hybrid ES cells, we directly sequenced RT-PCR products (Sanger Sequencing) and measured signal intensity on the electropherogram at the SNP positions [24]. RT-PCR products were purified with a Nucleo Spin^®︎ Gel and PCR Clean-up (Macherey-Nagel, Düren, Germany) following the manufacturer’s protocol. Sanger sequencing was performed at Azenta Life Science (Tokyo, Japan). Information on the SNPs of PWK and HMI was obtained from the UCSC Genome Browser [25] and Mouse Genome database with high added value (MoG+) [26].

Quantitative reverse transcription–PCR (qPCR)

Each reaction mixture (10 µl) contained 500 nM of each primer, 0.5 µl cDNA, and 5 µl SYBR Premix Ex Taq III (Takara). The qPCR was performed with 40 cycles of 5 s at 95°C and 30 s at 60°C in a Thermal Cycler Dice^®︎ Real Time System III (Takara). The relative expression of each gene was calculated by ΔΔCt analysis with the instrument software (Takara). The following primers were used: Gapdh 5′-GTGTTCCTACCCCCAATGTGTC-3′ and 5′-GGTCCTCAGTGTAGCCCAAGAT-3′, Oct4 5′-TAGCATTGAGAACCGTGTGAGG-3′ and 5′-CGCCGGTTACAGAACCATACTC-3′, Nanog 5′-TGTGCACTCAAGGACAGGTTTC-3′ and 5′-GCTTGCACTTCATCCTTTGGTT-3′, Sox2 5′-AGAACCCCAAGATGCACAACTC-3′ and 5′-TTATAATCCGGGTGCTCCTTCA-3′, Rex1 5′-ATGGACTAAGAGCTGGGACACG-3′ and 5′-GGTATTTGGGGACAACACTTGG-3′, Klf4 5′-GATGCAGTCACAAGTCCCCTCT-3′ and 5′-CCACGACCTTCTTCCCCTCT-3′, Tbx3 5′-CTTTGGGACCAGTTTCACAAGC-3′ and 5′-TGGCCTTTTTATCCAGTCCAGA-3′, Tfcp2l1 5′-AGTCGTATGAAATCCGGCTGCT-3′ and 5′-GTCATGGAAAACGACACGATG-3′, Gbx2 5′-CTCGCTGCTCGCTTTCTCTG-3′ and 5′-CTTCGGGTCATCTTCCACCTTT-3′, Esrrb 5′-AGAACAGCCCCTACCTGAACCT-3′ and 5′-AGCATACAGCTTGTCCTGCTCA-3′, Inpp5d 5′-CTGGCTTGGGGATCTCAACTAC-3′ and 5′-GGCCAGAAGGTCTGAATACTGC-3′.

Results

We established six types of novel mouse ES cell lines to analyze strain-specific monoallelic gene expression in the F1 hybrids obtained by crossing different inbred strains (Fig. 1); BP (B6 × PWK), PB (PWK × B6), BH (B6 × HMI), HB (HMI × B6), PWK, and HMI (Fig. 1). When characterizing these new cell lines, we used MB (MSM x B6) as a reference [2, 3, 24]. We established seven male BP lines, nine female BP lines, ten male PB lines, five male BH lines, nine female BH lines, nine male HB lines, six female HB lines, four male PWK lines, one female PWK line, and two male HMI lines from blastocysts by IVF; and four male HB lines and four female HB lines by natural mating. The quality of all cell lines was evaluated in several categories and is described below. The sex of each line was determined by using Ube1 gene polymorphism between the X and Y chromosomes [27].

Fig. 1.

Mouse strains used in this study. In the abbreviated names, the first letter stands for the mother and the second one for the father.

Analysis of karyotypes

The karyotype of newly established ES cell lines (male) was analyzed by the standard procedure. The cell lines with 40 chromosomes at the ratio of 72–84% in 50 metaphase plates were selected and used in this study (Table 2).

Table 2. Karyotypes of ES cell lines.

Strain	chromosome number
	≤37	38	39	40	41	42	≥43
BP12	2%	8%	6%	78%	2%	4%	-
BP21	-	10%	6%	76%	8%	-	-
PB7	-	4%	-	78%	18%	-	-
PB8	4%	10%	-	78%	6%	2%	-
BH3	-	2%	2%	76%	16%	4%	-
BH9	2%	-	12%	74%	8%	4%	-
HB2	-	2%	14%	84%	-	-	-
HB16	4%	10%	6%	74%	4%	2%	-
PWK2	2%	10%	14%	74%	-	-	-
PWK3	4%	8%	12%	72%	-	4%	-
HMI2	4%	10%	6%	74%	4%	2%	-
HMI9	8%	6%	-	78%	4%	4%	-

N=50.

Expression of gene markers for undifferentiated status

We analyzed the expression of Oct4, Sox2, Nanog, Rex1, Klf4, Tbx3, Tfcp2l1, Gbx2, and Esrrb as indicators of the undifferentiated status at naïve pluripotency (Fig. 2) [28]. Their expression in most of the lines was comparable to that in MB4. The expression of Tfcp2l1 and Gbx2 varied among the lines. The expression of Tfcp2l1 was lower in HB2 than in MB4 and was higher in PWK2. The expression of Gbx2 was lower in MB4 than in all lines except BP12 and BP21.

Fig. 2.

Analysis of undifferentiated stem cell markers in newly established embryonic stem (ES) cell lines. Relative expression levels of selected undifferentiated stem cell markers (Oct4, Nanog, Sox2, Rex1, Klf4, Tbx3, Tfcp2l1, Gbx2, and Esrrb) were determined by the DDCt method and normalized to the expression of Gapdh in each cell and to the expression in MB4. (*) P<0.01; assessed by Student’s t-test.

Analysis of strain-specific gene expression in F1 hybrid ES cell lines

The strain-specific monoallelic gene expression found in the F1 hybrid cells or tissues is thought to be caused by the incompatibilities between cis- and trans-elements brought from two different, parental inbred strains. The origins of these incompatibilities can be the SNPs existing in cis- and/or trans-elements.

We previously detected strain-specific monoallelic gene expression in the MB and BM F1 hybrid ES cells and adult brains by analyzing RNA-seq data [3]. Here, we confirmed one of these findings by RT-PCR and further analyzed the expression status of the same genes in selected newly established ES cell lines. Inpp5d showed clear strain-specific monoallelic expression between B6 allele and MSM allele in RNA-seq data (Fig. 3A and the left panel of Fig. 3B). Inpp5d is considered to be related to pluripotency in the stem cells [29]. Inpp5d was also analyzed for the allelic expression balance in the F1 hybrid ES cell lines with B6 and PWK, or B6 and HMI alleles (the middle and right panels of Fig. 3B). At the Inpp5d locus, the expression of the B6 allele was higher than that of the MSM allele. Although that of the PWK allele was higher than that of the B6 allele in both BP and PB ES cells, that of the HMI allele was almost the same as that of the B6 allele in both BH and HB ES cells (Fig. 3B). Thus, instances of differential allelic expression were observed, depending on the strain origin of the alleles. We also confirmed that the Inpp5d gene was expressed in the parental strains of the F1 hybrid ES cell line (Fig. 4). The expression of Inpp5d in MB4 was doubled to that in MSM or B6, and much lower than that in PWK, and HMI. In addition, there were differences in the expression level between B6 ES cells and HMI ES cells (Fig. 4), however, the gene expression from the B6 allele or HMI allele in the BH/HB cells was similar to each other (Fig. 3B). Thus, the loss of allelic expression in the F1 hybrid ES cell lines was not caused by the loss of the expression in the parental strains.

Fig. 3.

Analysis of strain-specific monoallelic gene expression. (A) Inpp5d gene structure (from the UCSC Genome Browser) and a screenshot of RNA-seq data at the Inpp5d locus in the BM and MB F1 hybrid ES cells from the IGB [23]. The alleles of the transcripts in the BM and MB cells can be distinguished by using SNPs in B6 and MSM. Yellow, maternal expression; red, paternal expression. Expression only from B6-derived alleles was observed in BM23 and MB4. We used the datasets previously published for this plot under the accession numbers DRA003816 and DRA003817 [3]. (B) Strain-specific monoallelic gene expression at the Inpp5d locus. The relative expression ratio of each allele in each cell strain is shown in the Sanger sequencing electropherograms. Green, A; red, T; black, G; blue, C. In BM23 and MB4, only B6-derived alleles were expressed (left panel), in agreement with the RNA-seq data shown in Fig. 3A. In BP12, BP21, PB3, and PB8, mostly PWK-derived alleles were expressed (middle panel). In BH3, BH9, HB2, and HB16, expression was bi-allelic (right panel). In each panel, Sanger sequencing of genomic DNA of F1 hybrid cells is shown on the right (“Genome”).

Fig. 4.

Expression of Inpp5d in parental embryonic stem (ES) cell lines of the F1 hybrid ES cell lines. Relative expression level of Inpp5d in the 6NK-7 (B6), Mol/MSM-1, PWK2, PWK3, HMI2, HMI9, BM23, MB4, 5BP12, BP21, PB7, PB8, BH3, BH9, HB2, HB16 lines compared to the MB4 was determined by the ΔΔCt method. The relative expression in 6NK-7 (B6) is presented as one. The expression was normalized to that of Gapdh in each line.

Discussion

In diploid organisms, allele-specific monoallelic gene expression is an important biological phenomenon and is sometimes related to the health of individuals. It is in some cases intrinsic to developmental processes or activities of cells and tissues, such as those found in genomic imprinting, or monoallelic gene expression in immune and olfactory sensory cells. Allele-specific monoallelic gene expression may also be accidentally caused by incompatibility between cis- and trans-elements in the same nucleus. In naturally crossbreeding animals, accidental monoallelic gene expression at a specific gene locus is possible because, for example, there is always a possibility that a maternal cis-element is incompatible with a paternal trans-element, and vice versa. Aberrant gene expression caused by cis-and trans-element incompatibility could be one of the factors in sporadic genetic diseases in humans [30]. In our previous study, we used inbred mouse strains MSM and B6, and observed genome-wide monoallelic expression in the F1 hybrid cells, which was possibly caused by incompatibility between cis- and trans-elements. In this study, we added two new mouse strains, PWK and HMI, and crossed them to B6 to obtain F1 hybrid ES cell lines with different genetic backgrounds. Adding other mouse strains from genetically distant subspecies is expected to increase the numbers of phenotypes caused by incompatibility of cis- and trans-elements. The wild-derived inbred strains are known to possess unique characteristics such as resistance to diseases [31]. The ES-cell-based experiments have advantages over animal-based experiments, especially at the initial stage of research, because repeated mating is not needed, and occasionally wild-derived inbred strains are difficult to handle.

We analyzed the sex, karyotypes, major undifferentiated stem cell markers, and differentiation potential of the newly established ES cell lines. Of 70 such lines, 41 were male. Most of the female ES cell lines had abnormal karyotypes (data not shown), while the karyotypes of the male ones were mostly within the normal ranges (Table 2). In the new ES cell lines, we analyzed Inpp5d, which showed strain-specific monoallelic expression in our previous study in the MB and BM ES cell lines, to look for a similar tendency. Its allele-specific monoallelic gene expression patterns in the PB, BP, HB, and BH F1 hybrid ES cell lines differed from those in the MB and BM lines. The overall level of allelic expression in the F1 hybrid ES cell lines was the highest in PWK, similar between B6 and HMI, and the lowest in MSM (Fig. 3B). This result might reflect sequence differences in the cis-elements among the parental mouse strains affecting the affinity to the trans-elements and consequently resulting in strain-dependent allele-specific expression imbalance. The actual mechanism of allele-specific expression in the F1 hybrid ES cell lines, however, is unknown until detailed analysis is performed at each locus. Many genes other than Inpp5d exhibit allele-specific monoallelic expression in the MB and BM F1 hybrid ES cell lines [3], and such genes might also show different allelic expression imbalance in the F1 hybrid ES cell lines, depending on the parental strain used.

The F1 animals from crosses between two genetically distant strains are excellent human disease models. Breeding such mice is able to introduce genome-wide genetic polymorphisms into the F1 offspring, resulting in allele-specific expression imbalance at many gene loci, possibly leading to distinct phenotypes. Gene expression itself is accepted as a “molecular phenotype”. Allele-specific monoallelic gene expression in F1 organisms can be a useful indication for studying disease phenotypes. The use of F1 hybrid ES cell lines can be a convenient initial step toward such phenotypic analysis, as we have identified hundreds of instances of allele-specific monoallelic gene expression in the BM and MB F1 hybrid ES cell lines and in in vitro differentiated cells derived from those ES cells [3]. In this study, we added two genetically distant mouse strains to produce F1 hybrid ES cell lines. A more variable genetic background could facilitate the analysis of cis-elements affecting target gene expression, for example, in cases where the allele-specific monoallelic gene expression pattern depends on the parental mouse strains used.

The metastabilities observed among mouse ES cell lines of different genetic backgrounds must also be considered when using the cell lines presented in this study. Generally, mouse ES cells are in a naïve state of pluripotency. However, genetic backgrounds influence the stability of pluripotency, especially in its naïve-to-primed spectrum [32]. The observed fluctuations of the naïve pluripotency markers (Tfcp2l1, Gbx2) observed in different F1 hybrid ES cell lines might indicate such metastabilities. Of note, a previous report showed the differential expression of Inpp5d in D3-ES cells versus iPSCs of C57BL/6 origin and pointed toward its influence in the pluripotency maintenance through the PI3K/Akt signaling pathway [29]. Therefore, the battery of the F1 hybrid ES cell lines could also offer a fresh resource to study pluripotency metastabilities.

Gene expression aberration similar to that found in the mouse F1 hybrid ES cells was also reported in yeast, where most of the cases were due to cis-element differences [33]. Many common human disease–associated variations identified by Genome-Wide Association Studies were found in the regulatory regions of the genome, that is, mostly in the cis-element regions [30]. Our newly established ES cell lines would be of great utility in understanding cis-element variation leading to the phenotypes associated with diseases in mammals.