A viral FLIP protein, E8, exogenously-expressed in the mesenchymal lineage of mice leads to bone malformations, lipoatrophy, and muscular atrophy

Abstract

The equine herpes virus 2, E8 protein is a member of the viral FLIP family, and as such, it is a potent inhibitor of death receptor-induced apoptosis in cultured cells. To extend our study of the effects of E8 to animals, we generated a mouse model in which the progeny of a cross between two transgenic mice conditionally express E8 under the control of the collagen type I α2 chain (Col1α2) promoter, allowing us to monitor and characterize the effects of E8 expression in the mesenchymal cell lineage. We observed growth defects associated with irregular bone formation during development. In addition, adult animals exhibited both lipoatrophy-like and muscular atrophy-like symptoms. These abnormal phenotypes likely arise from incomplete differentiation of mesenchymal stem cells (MSCs). To examine this hypothesis in more detail, we expressed E8 in the mouse MSC line C3H10T1/2 and performed a microarray analysis. Factors such as Nov/CCN3, STEAP4, and Ankrd1/CARP, which are involved in differentiation from MSCs to osteoblasts, adipocytes and myoblasts were affected. Taken together, our results demonstrate that the constitutive expression of herpesvirus gene products in the mesenchymal progenitors affects differentiation into multiple cell lineages.

Introduction

Protein-protein interactions are essential for many biological processes and one structural motif that can facilitate these interactions is the α-helix [1]. The death domain (DD) superfamily comprises four groups: the DD subfamily, the death effector domain (DED) subfamily, the caspase recruitment domain (CARD) subfamily, and the pyrin domain (PYD) subfamily [2, 3]. These domains form a structure, containing six α-helices, that engages in protein-protein interactions. Many members of this superfamily function in cell death and inflammation [2]. Among them, the DED subfamily includes 7 proteins: PEA-15, FADD, DEDD, DEDD2, cellular FLIP (c-FLIP), caspase-8 (Casp8), and caspase-10 (Casp10) [4]. The latter three proteins possess two DED domains in tandem as a prodomain. Both Casp8 and Casp10 are pro-apoptotic, whereas c-FLIP is a negative regulator of these two caspases via homotypic interactions involving the DEDs [5]. In several viruses, proteins containing tandem DEDs were identified and collectively termed viral FLIP (v-FLIP) to distinguish it from c-FLIP [6, 7]. E8 is a v-FLIP derived from equine herpes virus, and it has been shown to potently inhibit death receptor-mediated apoptosis. It operates by suppressing the pro-apoptotic activity of Casp8 through interactions with their DEDs [8,9,10]. Therefore, it was posited that the ectopic expression of E8 would produce the same effect as the loss of function of Casp8 in the signaling pathway.

We previously showed that Casp8-deficient mice die in utero, with abnormal development of the heart, neural tube, and angiogenesis [11]. In this study, we generated a transgenic (Tg) mouse line conditionally expressing E8 and investigated the extent to which E8-expressing mice resemble Casp8-deficient mice. Unexpectedly, we found that E8 expression affects tissues derived from mesenchymal stem cells, suggesting it may disrupt pathways independent of its known regulation of Casp8.

Materials and Methods

Reagents

Antibodies used for immunodetection were anti-Flag and peroxidase-conjugated anti-Flag (M2, Sigma-Aldrich, St. Louis, MO, USA), anti-GFP (polyclonal, MBL, Nagoya, Japan), anti-actin (MAB1501R, Chemicon International Inc., Temecula, CA, USA), and HRP-conjugated anti-mouse and anti-rabbit IgG (Cell Signaling Technology, Danvers, MA, USA). As a stimulant of apoptotic cell death, an agonistic anti-human Fas antibody (CH11, MBL) was used.

Mice and cultured cells

The Tg mouse line Col1α2-Cre, which was previously generated [12], expresses a Cre recombinase under the control of the promoter/enhancer of the gene encoding the α2 chain of collagen type I (Col1α2). C57BL/6 and C3H mice were purchased from CLEA Japan Inc. (Tokyo, Japan). All procedures were performed in accordance with the ARRIVE guidelines for the care and use of animals and were approved for five years by the Institutional Animal Care and Use Committee at Kyoto University (Lif-K07002, Lif-K08010, Lif-K09010, Lif-K10010, and Lif-K11010). To maintain the health of the animals, mice were housed with a 12 h light-dark cycle, with standard humidity (50%) and temperature (23.5°C), and were fed the GLP-compliant and standard diet (CA-1, CLEA Japan, Inc.) specifically designed for rodents. Anaesthesia and euthanasia of mice were performed in accordance with the AVMA Guidelines 2020. The murine mesenchymal stem cell (MSC) line C3H10T1/2 (RRID: CVCL 0190) [13] and human cervical carcinoma cell line HeLa (RRID: CVCL 0030) were cultured in DMEM with 10% fetal calf serum.

Plasmid constructs

The plasmid construct pCS2-FlagE8 was generated by subcloning a Flag-tagged E8 cassette from pLck-FlagE8 [14] into the pCS2 expression vector. The plasmid construct pCAG-EGFP (enhanced GFP)/FlagE8 was generated with a DNA fragment carrying both EGFP cDNA and an SV40 polyadenylation sequence, which was flanked by two loxP sites, followed by a Flag-E8 DNA fragment from pCS2-FlagE8; the resulting cassette was cloned into the pCAGGS expression vector [15]. For expression of a Cre recombinase in the cultured cells, the plasmid construct pBS185 (Invitrogen, Carlsbad, CA, USA) was used. For detection of transfected cells, the plasmid construct pCAG-mCherry, which was previously generated [16], was used. The plasmid construct pME18S-Neo, which carries a neomycin resistance gene [17], was used for selection of stably transfected cells using G418 disulfate solution (Nacalai Tesque, Kyoto, Japan).

Generation of transgenic (Tg) mice

In a preliminary study, we failed to generate mice constitutively expressing E8 protein. Therefore, as a next approach, we sought to generate mice conditionally expressing E8. A DNA fragment carrying the CAG:EGFP/FlagE8 transgene from the pCAG-EGFP/FlagE8 plasmid was microinjected into F1 (B6C3-F1) heterozygous eggs from a cross between C57BL/6 and C3H strains, and surviving eggs were transferred into the oviducts of recipient pseudopregnant females, as described previously [18]. Three Tg lines expressing EGFP throughout the body were established and designated as “E8-STg”, and these were backcrossed onto C57BL/6 mice to ensure genetic background homogeneity.

Genotyping

The genotype of transgenic mice carrying the cre gene was confirmed by PCR on mouse tail genomic DNA, using primers (5′-TTAGCACCACGGCAGCAGGAGGTT-3′ and 5′-CAGGCCAGATCTCCTGTGCAGCAT-3′). The PCR amplification was performed using KOD-Plus polymerase (TOYOBO, Osaka, Japan) following the manufacturer’s protocol. E8/Cre double-transgenic (DTg) mice were generated by crossing E8-STg mice with Col1α2-Cre Tg mice. To genotype the offspring, genomic PCR was performed to detect the cre gene, while the integration of the transgene derived from E8-STg was verified by screening for EGFP fluorescence on the body surface using a handheld blue LED excitation light.

Histological and immunohistochemical analyses

Tissues isolated from E8-STg and E8/Cre-DTg mice were fixed in 10% PBS-buffered formalin (Nacalai Tesque), embedded in paraffin wax, cut into 3–4 µm sections, and dewaxed in xylene and ethanol. Bone samples were also decalcified with 10% ethylenediaminetetraacetic acid in 0.2M phosphate buffer (pH7.3) for 21 days at 4°C before paraffin embedding. Tissue specimens were then examined by hematoxylin and eosin (HE) staining and Masson trichrome (MT) staining as described previously [19]. For immunohistochemical staining, slides were dipped in methanol containing H₂O₂ (0.3% (v/v)) to quench endogenous peroxidase activities for 30 min. After washing with PBS, slides were incubated with a peroxidase-conjugated anti-Flag antibody (10 µg/ml) with microwave irradiation for 15 min and washed with PBS three times for 5 min [20]. After washing with PBS three times, liquid DAB+ (K3468, DAKO, Carpinteria, CA, USA) was applied, and brown precipitates were visualized. Nuclear counterstaining was done with hematoxylin.

Micro CT analyses

The bone structure integrity and the fat content of E8/Cre-DTg and their littermates were evaluated using a micro computational tomographic scanner (µCT) (R_µCT, Rigaku, Akishima, Japan). The operating conditions of µCT were 90 kV and 120 µA, and the resolution was set to a voxel size range of 20–100 µm for each sample.

Transfection of plasmid DNAs into cultured cells

To express the transgenes in cultured cells, plasmid DNAs were transiently transfected into cells using Lipofectamine 3000 reagent (Invitrogen) according to the manufacturer’s protocol. Typically, two days later, transfectants were examined for cytological and immunoblot analyses. To establish clonal lines, transfected cells were placed under selection with G418 starting two days after transfection.

Cell death assay

For the cell death assay, HeLa cells were transiently transfected with either pCS2, pCS2-FlagE8, or pCAG-EGFP/FlagE8 with or without pBS185 together with pCAG-mCherry. At 48 h post-transfection, transfectants were stimulated with a mixture of 250 ng/ml anti-human Fas antibody CH11 and 5 µg/ml cycloheximide (CHX). Following 8 h incubation, floating and fragile cells were removed, and surviving cells were fixed in PBS containing 3.7% formaldehyde, then washed with PBS. The fixed cells were observed with phase-contrast and fluorescence microscopy, and images were captured (DMIRE2, Leica Microsystems, Wetzlar, Germany).

Bone and cartilage staining

To visualize bone and cartilage in newborn mice, we followed a staining protocol using alizarin red and alcian blue (Nacalai Tesque) [21].

Immunoblot analyses

To detect FlagE8 and EGFP proteins in cultured cells, HeLa cells were transfected with pCS2, pCS2-FlagE8, or pCAG-EGFP/FlagE8 with or without pBS185, and harvested after two days. Briefly, cells were suspended in lysis buffer [50 mM Tris-HCl (pH 7.5), 20 mM MgCl₂, 0.5% Nonidet P-40, 150 mM NaCl supplemented with protease inhibitor cocktail (Nacali Tesque)]. After removing cell debris by centrifugation, cell lysates were combined with Laemmli’s sample buffer, resolved by SDS-PAGE, and analyzed by immunoblotting with anti-Flag, anti-GFP, and anti-actin antibodies, respectively. After incubation with HRP-conjugated anti-mouse or anti-rabbit IgG antibodies, the bound immune complexes were visualized with Immobilon Western Chemiluminescent HRP Substrate (Millipore Corporation, Billerica, MA, USA) using a luminescent image analyzer (LAS-3000, Fujifilm, Tokyo, Japan).

Establishment of E8-expressing C3H10T1/2 cells

To establish mesenchymal stem cells stably expressing E8, C3H10T1/2 cells were co-transfected with pCS2-FlagE8 and pME-Neo plasmids at a 10:1 ratio and selected for G418 resistance. The established G418-resistant bulk cells were termed C3H10T1/2-FlagE8.

Microarray analysis

For the Oligo DNA microarray analysis, parental C3H10T1/2 cells and C3H10T1/2-FlagE8 cells were cultured in parallel, and total RNAs were extracted from these two-type cells using Sepasol solution (Nacalai Tesque), following the manufacturer’s protocol. Microarray analysis was performed with a 3D-Gene Mouse Oligo chip 24k system (Toray Industries Inc., Tokyo, Japan). Briefly, total RNAs were labeled with Cy5 for parental C3H10T1/2 or Cy3 for C3H10T1/2-E8, using an Amino Allyl MessageAMP II aRNA Amplification Kit (Applied Biosystems, Waltham, MA, USA). The Cy5- or Cy3-labeled aRNA pools were mixed with hybridization buffer. The hybridization was performed according to the manufacturer’s protocol (https://www.3d-gene.com). The hybridization signals were obtained with a 3D-Gene Scanner (Toray Industries Inc.), and processed by 3D-Gene Extraction software (Toray Industries Inc.). Detected signals for each gene were normalized by a global normalization method, with the median Cy3/Cy5 ratio adjusted to 1.

As the data availability, the analyzed data was deposited as GSE261470 in the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE261470).

Generation of a heat-map

A heat-map of 20 mfCYPs was constructed by hierarchical cluster analysis using Cluster 2.0 software, and the results were displayed with the TreeView program (https://www.treeview.co.uk/order/).

Statistical analysis

Statistical data were presented as bar graphs (mean ± SD) or box-and-whisker plots (median, interquartile range, and full range), depending on the distribution and type of data. Differences were assessed using Student’s t-test. Statistically significant differences (*: P<0.05, **: P<0.01, and ***: P<0.005) are indicated in the figures.

Results

Generation of E8/Cre double transgenic (E8/Cre-DTg) mice

The E8 protein derived from equine herpesvirus contains tandem ‘death effector domains (DEDs)’ termed DED1 and DED2 (Fig. 1A). To establish a transgenic (Tg) mouse line conditionally expressing E8, we generated a plasmid construct, pCAG-EGFP/FlagE8 consisting of the CAG promoter, floxed EGFP and SV40 polyA, Flag-E8 and rabbit β-globin polyA DNA fragments (Fig. 1B). In the presence of Cre recombinase, the EGFP-SV40 polyA cassette should be excised, allowing the expression of the FlagE8 gene (Fig. 1C). We also constructed plasmid pCS2-FlagE8 to investigate the functional activity of E8 in cultured cells (Fig. 1B). To confirm the Cre-mediated recombination and the expressional change of gene products, we transfected these plasmid DNAs into HeLa cells and analyzed protein extracts from the respective cells by immunoblot analysis (Fig. 1C). A shift from EGFP to FlagE8 was detected when Cre was expressed. It has been shown that E8 prevents cell death mediated through death receptors when exogenously expressed [8,9,10]. We further examined whether E8 has an anti-apoptotic activity for death receptor-mediated stimuli (Supplementary Fig. 1). In the cell population expressing FlagE8, more red fluorescent-positive cells survived the cell death stimulus compared to the population of cells not expressing FlagE8.

Fig. 1.

Generation of E8/Cre-DTg mice. (A) Structure of E8 protein. Equine herpesvirus E8 consists of 171 amino acids and carries tandem death effector domains (DEDs). (B) A schematic diagram of plasmid constructs. The plasmid pCS2-FlagE8 carries an expression cassette that consists of the CMV promoter and the transgene, Flag-tag-fused E8 cDNA, with an SV40 polyA site. Another construct, pCAG-EGFP/FlagE8, consists of the CAG promoter, the EGFP cDNA flanked by loxP sequences, and the Flag-tag-fused E8 DNA fragment. (C) Immunoblot analysis. Cell extracts were prepared from cells transfected with pCS2 vector (lane 1), pCS2-FlagE8 (lane 2), pCAG-EGFP/FlagE8 & pCS2 (lane 3), or pCAG-EGFP/FlagE8 & pBS185 (lane 4). Immunoblotting was performed using the indicated antibodies. (D) Immunohistochemical analysis. The heart sections prepared from E8-STg (left) and E8/Cre-DTg (right) fetuses at E16.5 were immunostained using an anti-Flag antibody. Scale bars indicate 200 µm.

To establish a Tg mouse line conditionally expressing FlagE8, a DNA fragment containing the CAG:EGFP/FlagE8 transgene cassette was injected into fertilized heterozygous eggs. Five founder mice carrying the transgene were born and exhibited no developmental abnormalities (data not shown). The established five founders were designated “E8-STg” (E8 single Tg mouse). Among these animals, three mouse sublines, E8-STg (line E), E8-STg (line O), and E8-STg (line U) were selected due to high EGFP expression in the skin of neonatal animals; these were backcrossed onto C57BL/6 mice to ensure genetic background homogeneity (Supplementary Fig. 2). Thus, we have established transgenic mouse lines carrying the CAG:EGFP/FlagE8 construct.

As we observed irregular development of the fetal heart in Casp8-deficient mice [11], we searched for a transgenic mouse line that expresses Cre specifically in tissues including the heart, and chose the Col1α2-Cre Tg mouse line as a candidate. In these animals, as Cre expression is under the control of the promoter/enhancer unit of the col1α2 gene, Cre proteins are synthesized only in the mesenchymal cell lineage in adult animals [12]. To establish the transgenic mice conditionally expressing E8 in the mesenchymal lineage, we crossed E8-STg mice with Col1α2-Cre Tg mice. Progeny from a mating of E8-STg females with Col1α2-Cre Tg males were examined for FlagE8 expression at E16.5 by immunohistochemical analysis with an anti-Flag antibody (Fig. 1D and Supplementary Fig. 3). These staining profiles largely coincided with Cre expression in Col1α2-Cre Tg mice [12]. The resulting E8- and Cre-expressing double transgenic mice are referred to as “E8/Cre-DTg” hereafter.

Growth retardation of E8/Cre-DTg mice

Growth of E8/Cre-DTg mice was monitored. They were able to survive beyond 2 years (data not shown). In addition, male and female animals have normal reproductive capacities (data not shown). However, they exhibited growth defects (Fig. 2A). E8/Cre-DTg mice were significantly smaller than E8-STg mice. Body weight of E8/Cre-DTg mice was dramatically decreased compared with that of E8-STg mice (Fig. 2B). Histological analysis revealed no abnormalities in the endoderm-derived tissues such as liver (data not shown). Thus, it is likely that only mesenchymal tissues specifically manifest abnormalities. Taken together, when E8 is expressed in the mesenchymal cell lineage, transgenic animals can survive and reproduce but display some morphological defects.

Fig. 2.

Generation and analysis of E8/Cre-DTg mice. (A) Observation of growth rate of E8/Cre-DTg transgenic mice. E8-STg (left side) and E8/Cre-DTg (right side) mice at 5 days (a), 11 days (b), 15 days (c) and 10 months (d). Scale bars indicate 10 mm. (B) Statistical analysis of the body weight. The body weight of E8/Cre-DTg mice and their littermates was measured at 5 months, 10 months, and 20 months after birth. The number in the bracket indicates the number of individuals examined. Data are presented as the means and standard deviations of samples. Double asterisks indicate P<0.01 (Student’s t-test).

Skeletal defects in E8/Cre-DTg mice

To investigate in more detail the cause of growth defects in E8/Cre-DTg mice, we focused on bone formation. We stained bone and cartilage of E8/Cre-DTg newborn mice and their littermates (Supplementary Fig. 4, panels a–d). Skeletal elements, including bone and cartilage, of E8/Cre-DTg mice were weaker and shorter than those of their single-transgenic littermates, suggesting that the abnormal bone formation has already occurred in neonatal animals. To further examine whether expression of E8 affects postnatal bone formation, we monitored the skeletal architecture of E8/Cre-DTg female mice by micro-computed tomography (µCT). It has been shown that skeletal elements form by two distinct processes: endochondral ossification, during which a cartilage template is replaced by bone, and intramembranous ossification, whereby mesenchymal cells differentiate directly into osteoblasts [22]. The bones of the skull are primarily formed by intramembranous ossification. µCT analysis of the skull revealed that the heads of E8/Cre-DTg female mice were significantly smaller than those of E8-STg littermates at 20 months of age (Figs. 3A and B). This difference was similarly observed when comparing younger mice, such as those aged 5 months or 10 months (Fig. 3B). Thus, the length from the tip of the nose to the back of the head in E8/Cre-DTg mice was clearly shorter than that in E8-STg mice. The difference did not change from neonatal mice to aged mice (Figs. 3A, B and Supplementary Fig. 4). Additionally, incomplete closure of the fontanelle at the cranial vertex (Supplementary Fig. 4, panels e–f) and a curved nasal structure of the anterior head (data not shown) were also observed in some adult E8/Cre-DTg mice. Since the fontanelle is a key site of cranial development, this morphological irregularity structure likely stems from impaired cartilage and bone development [23].

Fig. 3.

Analysis of bone formation in E8/Cre-DTg mice. (A) Micro-CT (µCT) images of the skull from 20-month-old E8-STg (a) and E8/Cre-DTg (b) mice. Images of both a representative cross section (left) and the 3D structural model (right) are shown. The voxel size of µCT images was set to 50 µm. Scale bars indicate 5 mm. (B) Statistical analysis of skull length in E8-STg and E8/Cre-DTg mice. Five-, 10-, and 20-old month female mice were evaluated for their skull using µCT. Triple asterisks indicate P<0.005 (Student’s t-test). (C) µCT images of the tibia from 20-month-old E8-STg (a) and E8/Cre-DTg (b) mice. Cross-sectional views of the XZ-plane are shown. The voxel size of µCT images was set to 20 µm. Scale bars indicate 2 mm. (D) Statistical analyses of bone volume (a) and bone density (b). Both E8-STg and E8/Cre-DTg mice were evaluated for the articular site of the proximal tibia-indicated by the area enclosed by the orange line in (C)-at 5, 10, and 20 months of age. The single asterisk indicates P<0.05, and the double asterisks indicate P<0.01 (Student’s t-test). (E) Histological analysis. The trabecular bone area of femurs from 10-month-old E8-STg (a, c) and E8/Cre-DTg (b, d) mice, were stained with HE. Arrows indicate the trabecula. Scale bars indicate 500 µm (a, b) and 100 µm (c, d), respectively.

Furthermore, the hindlimbs of E8-STg and E8/Cre-DTg mice were examined (Figs. 3C–E). Compared to E8-STg mice, E8/Cre-DTg mice exhibited reduced trabecular bone and fewer osteophyte formations in the distal femur and proximal tibia at 20 months of age, as assessed by µCT (Fig. 3C). Both bone volume and density at the articular site of the proximal tibia (the area enclosed by the orange line in Fig. 3C) were analyzed in 5-, 10-, and 20-month-old mice, revealing significant differences between E8-STg and E8/Cre-DTg mice (Fig. 3D). In addition, histological analysis of the femurs showed that the trabecular area in the epiphysis was markedly lower in 10-month-old E8/Cre-DTg mice compared with E8-STg mice (Fig. 3E). Thus, µCT and histological analyses suggest that the expression of E8 partially inhibits bone formation and cartilage development.

Abnormal phenotype in the fat tissues of E8/Cre-DTg mice

E8/Cre-DTg mice are lean as well as small. Lipoatrophy is a known manifestation of a lack of fat accumulation [24]. By µCT analysis, we found that E8/Cre-DTg mice exhibit less fat accumulation compared with their E8-STg littermates (Figs. 4A and B). Histological examinations verified the small size of adipocytes in the visceral and subcutaneous fat, indicating a lower uptake of fat (Figs. 4C and D). As adipocytes as well as osteoblasts are derived from mesenchymal stem cells, the lipoatrophy-like phenotype of E8/Cre-DTg mice might be caused by aberrant adipocyte differentiation from stem cells.

Fig. 4.

Detection of the lipoatrophy-like phenotype in E8/Cre-DTg mice. (A) Images of the abdominal area by µCT scan. The fat content in the abdominal area, which contains the fourth lumber vertebra (white), from 20-month-old E8-STg (a) and E8/Cre-DTg (b) mice was analyzed. The arrow shows the area containing fat tissue and the scale bar indicates 10 mm. The voxel size of µCT images was set to 50 µm. (B) Statistical analysis of the fat content. The fat content in the abdominal area containing the fourth lumber vertebra at indicated months is summarized. The number in the bracket indicates the number of individuals examined. Data are presented as the means and standard deviations of samples. The single asterisk indicates P<0.05, and the double asterisks indicate P<0.01 (Student’s t-test). (C) Histological analysis of the visceral fat. Sections prepared from the visceral fat of E8-STg (a, c) and E8/Cre-DTg (b, d) mice at 5 months (a, b) and 10 months (c, d) after birth were stained with HE. The scale bar indicates 100 µm. (D) Histological analysis of the skin. The skins prepared from 10-month-old E8-STg (a) and E8/Cre-DTg (b) mice were stained with HE. The asterisks indicate the adipocytes infiltrating into the area of the dermis from the subcutaneous layer. Abbreviation: SL, subcutaneous layer. The scale bar indicates 100 µm.

Abnormal phenotype in skeletal and cardiac muscles of E8/Cre-DTg mice

We observed a Kyphosis-like symptom in E8/Cre-DTg mice, especially in aged females (Supplementary Fig. 5, shown by an arrow). By µCT scanning, we noticed that the curve of the spine was due to underdevelopment of the muscles of the back (Fig. 5A, shown by a circle). Histological examination confirmed this irregular phenotype (Fig. 5B). The skeletal muscles in the femur of E8/Cre-DTg mice showed patchy degeneration and atrophy of fascicles. The Masson trichrome staining clearly highlights the abnormality (Fig. 5B, panel d, shown by arrows). Furthermore, cardiomyocytes also showed degeneration and focal fibrosis (Figs. 6e and f, shown by arrows and an ellipse). Thus, both skeletal and cardiac muscles, which are differentiated from mesenchymal stem cells, showed abnormalities in E8/Cre-DTg mice.

Fig. 5.

Analysis of muscular atrophy in E8/Cre-DTg mice. (A) µCT images of the backbone. Images of both a representative cross section (left) and the 3D structural model (right) are shown. The circle indicates the dorsal skeletal muscle. “P” indicates a supporting plate. The voxel size of µCT images was set to 50 µm. Scale bars indicate 5 mm. (B) Histological analysis. Sections prepared from femoral muscle of 10-month-old E8-STg (a, c) and E8/Cre-DTg (b, d) mice, visualized by HE (a, b) or Masson trichrome (MT) (c, d) staining. Arrows indicate degenerated and atrophied muscle fibers.

Fig. 6.

Histological analysis of the heart in E8/Cre-DTg mice. Histological analyses. Cross sections prepared from hearts of E8-STg (a-c) and E8/Cre-DTg (d-f) mice, sacrificed at 5 months (a, b, d, e) and 10 months (c, f), stained with HE (a, c, d, f) or MT (b, e). Arrows indicate degenerated and atrophied muscle fibers (e) and the region within the ellipse contains many degenerated muscle fibers (f).

Establishment and characterization of C3H10T1/2 transfectants expressing E8

Mesenchymal stem cells (MSCs) are highly proliferative and clonogenic cells that are capable of differentiating into a variety of cell types, including osteoblasts, chondrocytes, adipocytes, fibroblasts, and myoblasts [25]. The murine cell line C3H10T1/2 has been shown to retain these properties of MSCs [26]. Therefore, this cell line is useful for analyzing the effect of E8 expression in cells cultured in vitro. We co-transfected plasmid DNAs which carry the FlagE8 transgene and the neomycin-resistance gene into C3H10T1/2 cells and cultured them in the presence of G418 to remove non-transfected cells. Using microarray analysis, we examined the expression profiles of transcripts from transfected C3H10T1/2-FlagE8 cells, and compared those to transcript profiles of parental C3H10T1/2 cells. The analyzed data was deposited as GSE261470 in the GEO database. Among 23,471 transcripts examined, 52 transcripts were decreased by more than 4-fold, and 9 transcripts were increased by more than 4-fold in C3H10T1/2-FlagE8 cells (Supplementary Tables 1 and 2). We focused on STEAP4 (Six-Transmembrane Epithelial Antigen of Prostate4), ANKRD1 (Ankyrin repeat domain-containing protein 1)/CARP (cardiac ankyrin repeat protein), and NOV (Nephroblastoma Overexpressed)/CCN3 (cellular communication network factor 3) among the transcripts that changed significantly (Fig. 7). STEAP4 is known as an inducer of adipogenesis [27], and ANKRD1/CARP is involved in cardiogenesis and myogenesis [28]. The expression levels of these two transcripts were decreased in C3H10T1/2-FlagE8 cells (Steap4: normalized counts from 61 to 4, log₂ fold change=−3.98; Ankrd1: normalized counts from 474 to 114, log₂ fold change=−2.06). In contrast, NOV/CCN3, which is a negative regulator of osteoblastogenesis [29], was highly expressed in C3H10T1/2-FlagE8 cells (normalized counts from 4,202 to 18,186, log₂ fold change=4.33). Based on a heat map analysis, we suggest that overall, the expression of genes involved in the induction of differentiation are suppressed by E8 expression in mesenchymal stem cells; and by the same token, the expression of suppressive factors is increased by E8 expression (Supplementary Fig. 6).

Fig. 7.

Assessment of the transcriptional profile on microarray analysis. Based on Supplementary Tables 1 and 2, transcripts with differences in expression of 4-fold or more between C3H10T1/2-FlagE8 cells and parental C3H10T1/2 cells are plotted. Among them, three transcripts of interest, Steap4, Ankrd1, and Nov are marked.

Discussion

We observed that equine herpes virus E8 protein causes growth retardation associated with abnormal bone development, as well as lipoatrophy-like and muscular atrophy-like symptoms when it is expressed in the mesenchymal cell lineages of mice (Figs. 2–6 and Supplementary Figs. 4 and 5). These phenotypes were observed in all three established sublines (Data not shown). Furthermore, when E8 is expressed in C3H10T1/2 mesenchymal stem cells (MSCs), the expression levels of differentiation-inducing factors are reduced (Fig. 7, Supplementary Fig. 6, Supplementary Tables 1, and 2). Taken together, these results suggest that E8 acts in MSCs or progenitor cells to weaken, but not completely suppress, differentiation.

In our previous study, we generated Casp8-deficient mice [11]. Homozygous Casp8-deficient mice are embryonic lethal and show abnormalities in the heart and neural tube, as well as defective angiogenesis. Other groups also verified the same phenotype in Casp8-defcient mice [30]. However, when a necroptosis-inducing protein RIPK3 (Receptor Interacting Protein Kinase 3) was further deleted, Casp8-deficient mice were viable [30, 31]. It is worth noting that Casp8-Ripk3 double-deficient mice are normal with regard to bones, fat, and muscles, even in adults. This suggests that Casp8 is unrelated to bone formation, adipose formation, and muscle formation. In this study, we revealed that E8, which was previously identified as an inhibitor of Casp8 activation, affects the induction of differentiation from MSCs. Consequently, we propose that E8 interacts with a factor (s) different from Casp8 in MSCs.

E8 is a member of the virus DED-containing protein family, and other members in this family have the ability to activate NF-κB (Nuclear Factor-kappa B) [32, 33]. Although our results are still preliminary, we detected the activation of NF-κB in E8-expressing C3H10T1/2 cells. Therefore, we explored the possibility that NF-κB is a regulator of multiple lineage differentiation from MSCs. However, it has been shown that suppression of NF-κB activation promotes differentiation of osteoblasts and has an antianabolic effect on bone formation [34]. In addition, it has been suggested that NF-κB is not involved in myogenesis and adipogenesis [35, 36]. Therefore, NF-κB is unlikely a main regulator. So far, Runx2 for osteogenesis, Sox9 for chondrogenesis, PPARγ for adipogenesis, and MyoD for myogenesis from MSCs have been identified as representative differentiation factors [37]. They are cell-lineage specific regulators but not a universal factor. Conceivably, E8 may interact with a master regulator controlling multiple lineage differentiation from MSCs. Binding of E8 to such a factor would then suppress the differentiation-inducing ability.

From a different perspective, E8 might activate or stabilize a factor that suppresses differentiation. Interestingly, E8 has been shown to enhance the Wnt signaling pathway [38]. Wnt is known to act as an inhibitor of cell fate determination for osteogenesis, chondrogenesis, adipogenesis, and myogenesis [39]. It will be important to identify E8-interacting proteins or molecules to gain insights on the molecular mechanisms governing the regulation of MSC differentiation. Additionally, our transgenic mice, which have normal lifespans despite these growth abnormalities, are unique and thus useful as a model animal.

Conclusion: While E8 is a known viral FLIP protein that antagonizes Casp8-mediated apoptosis, this study demonstrates that its exogenous expression in murine mesenchymal lineages leads to bone malformations, lipoatrophy, and muscular atrophy. Consistently, the exogenous expression of E8 in the C3H10T1/2 cell line alters the expression profile of genes involved in mesenchymal stem cell (MSC) differentiation. As shown in Fig. 8, these results provide strong evidence that E8 modulates the multi-lineage differentiation potential of mesenchymal stem and progenitor cells.

Fig. 8.

E8 influences mesenchymal lineage commitment and tissue development. Exogenous expression of E8 in the mesenchymal lineages, the resulting transgenic mice exhibited bone malformations, lipoatrophy, and muscular atrophy. Furthermore, exogenous expression of E8 in the mesenchymal stem cell line alters the expression profile of genes involved in MSC differentiation. Based on these findings, it strongly suggests that E8 influences the induction of multi-lineage differentiation from mesenchymal stem or progenitor cells.

Author Contributions

K.S. established the double transgenic mouse line, prepared figures and wrote the main manuscript text, N.Y. has generated the E8-carrying transgenic mouse line, Y.O. and S.T. performed histological and immunohistological analyses, T.T. and S.T. analyzed the bone histologically, Y.O. examined mice with a µCT. All authors reviewed the manuscript.

Conflicts of Interest

The authors declare no conflict of interest.

Funding Statement

This work was supported in part by JSPS KAKENHI (JP16H06276 [AdAMS (Aa210038)]) to S.T.