Stem cells exhibit a promiscuous gene expression pattern. We show herein that the early embryo and adult MSCs express B-cell receptor component mRNAs. To examine possible bearings of these genes on the expressing cells, we studied immunoglobulin μ chain-deficient mice. Pregnant μ chain-deficient females were found to produce a higher percentage of defective morulae compared with control females. Structure analysis indicated that the μ mRNA species found in embryos and in mesenchyme consist of the constant region of the μ heavy chain that encodes a recombinant 50-kDa protein. In situ hybridization localized the constant μ gene expression to loose mesenchymal tissues within the day-12.5 embryo proper and the yolk sac. In early embryo and in adult mesenchyme from μ-deficient mice, δ replaced μ chain, implying a possible requirement of these alternative molecules for embryo development and mesenchymal functions. Indeed, overexpression of the mesenchymal-truncated μ heavy chain in 293T cells resulted in specific subcellular localization and in G1 growth arrest. The lack of such occurrence following overexpression of a complete, rearranged form of μ chain suggests that the mesenchymal version of this mRNA may possess unique functions.
Lymphocytes of the B lineage are the only cells that synthesize immunoglobulin (Ig) molecules, the effectors of the humoral immune response. Igs comprise of four polypeptide chains, two identical heavy chains (HCs) and two identical light chains (LCs), each containing an antigen-binding variable (V) domain, plus a more conserved constant (C) domain encoded by one to four exons. The products of V(D)J recombination of both immunoglobulin heavy and light chain loci associate covalently to form the ligand recognition domain [1, , , , , –7]. The isotype of the HC molecules determine the subclass of that antibody: IgM (μ), IgD (δ), IgG (γ), IgE (ε), and IgA (α). Ig molecules appear in both secreted and membrane-bound forms. The latter are noncovalently bound to the transmembrane Igα (CD79a) and Igβ (CD79b) proteins to form the B-cell receptor (BCR) (reviewed by ).
In addition to the above orderly synthesis and assembly of Ig molecules, truncated forms of these molecules occur in B cells. The Dμ protein forms in some pre-B cells during normal mouse B-cell development. This protein consists of a short D region composed entirely of a coding sequence without introns or splice sites upstream to a complete C region . The Dμ expressed in pre-B cells blocks further B-cell development [10, , –13] and is thus functional. A second μHC truncation is found in a human B-cell neoplasia, μHC disease. This truncated μHC (Tμ) lacks the variable domain and has its translation initiation site at the fifth amino acid of the first constant region domain [14, 15]. The truncated BCR is expressed on the cell surface in the absence of LC [16, 17]. Although this receptor is lacking an antigen binding domain, overexpression assays showed that this variant is constitutively active . Whereas Dμ protein leads to deletion of the cell that expresses it, Tμ expression allows developmental progression and bone marrow emigration. Whether the mRNA encoding these truncated proteins and the proteins themselves serve a function in normal physiology is unknown. In addition to the above, a truncated form consisting of the C region (Cμ) has been described in T- and pre-B-lymphoid and myeloid cells .
We have reported the expression of truncated forms of T-cell receptor (TCR) mRNAs in mesenchymal cell lines and in primary mouse embryo fibroblasts (MEFs) . The cell lines exhibit MSC functions in promoting hemopoiesis and being capable of differentiation into adipocytes and osteoblast [21, , , , –26]. This MSC nature is shared by MEFs [27, –29]. The unexpected presence of TCR in MSC-like cells raised the question as to whether nonlymphoid cells express also Ig gene products. We show herein that a unique form of truncated μHC is found in the embryo prior to the development of functional B-lineage lymphocytes and is present, along with other components of the pre-B cell receptor, in cultured mesenchymal cells. Because the expression of IgM constitutes a major step in B-cell development, it could be expected that Ig μ chain-deficient mice would have defective B-cell generation pattern. However, in these mice, μHC is replaced by δHC, and the mice exhibit near normal immune status, albeit having a modified antibody repertoire [30, 31]. Our present study shows that in μ-deficient mice, δHC is substituting for μHC in the oocyte, morula, and mesenchyme of the early embryo, as well as in the adult mesenchyme. This unexpected finding implies a role for Ig gene products in the regulation of early embryogenesis and in MSC functions.
Materials and Methods
Isolation and Transfer of Embryonic Cells
Balb/c and IgM-deficient mice (on Balb/c background)  were maintained under pathogen-free conditions, crossed, and homozygous IgM−/− as well as IgM+/+ (WT) mice were selected. Superovulation was induced in virgin 5-week old mice by intraperitoneal injection of 5 units of pregnant mare serum gonadotropin (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com) and 5 units of human chorionic gonadotropin (Sigma-Aldrich) 48 hours later. The following day, females were killed, and unfertilized oocytes were collected from the oviduct by flushing M2 medium containing hyaluronidase (300 mg/ml). To isolate morulae, superovulation was induced as above in 4- to 6-week old virgin Balb/c, ICR, and IgM−/− mice. Each superovulated mouse was then placed in a cage overnight with a sexually mature male of the same strain. Successful mating was determined by the presence of a copulation plug on the following day (designated as day 0.5 of gestation). Females were killed on day 2.5, and morulae were collected by flushing M2 medium through the uteri. Embryos cultured in vitro were placed into 30-μl drops of M2 medium with 4 mg/ml bovine serum albumin and covered with light paraffin oil. Embryos were sorted into their respective developmental stages, and defective embryos were microscopically identified. For embryo transfer, recipient female mice were prepared by mating with vasectomized males (ICR) 2.5 days before the embryo transfer. The procedure of embryo transfer was performed by implanting morulae into pseudopregnant recipient females. Fifteen morulae were transferred to each uterine horn (total of 30 per female). The mice were killed 10 days after the embryo transfer. MEF were derived from 12.5-day-old embryos.
BM cells were obtained from 7- to 8-week old female C57BL/6 mice. MSCs were grown in murine Mesencult basal media supplemented with 20% murine mesenchymal supplement (StemCell Technologies, Vancouver, BC, Canada, http://www.stemcell.com) containing 60 μg/ml penicillin and 100 μg/ml streptomycin and incubated at 37°C in a humidified incubator with 10% CO2 in air. Half of the medium was replaced every 3 days to remove the nonadherent cells. Once the adherent cells had reached confluence, the cells were trypsinized, centrifuged, and resuspended in their medium and incubated with antibodies specific to CD45.2 R-phycoerythrin (SouthernBiotech, Birmingham, AL, http://www.southernbiotech.com) and CD11b/ Mac1 fluorescein isothiocyanate (SouthernBiotech) for 1 hour. The cells were sorted using the FACSVantage SE cell sorter (BD Biosciences, San Diego, http://www.bdbiosciences.com). The double-negative cell population was collected and seeded in MSC medium.
Cell Lines and Transfection Procedure
Murine bone marrow-derived stromal cell lines MBA-2.1 and MBA-2.4 endothelial-like, MBA-13 fibroendithelial, MBA-15 osteogenic and 14F1.1 preadipocytes [21, , –24, 32], and the 293T human embryonic kidney cell line were used. These were cultured in Dulbecco's modified Eagle's medium supplemented with 100 μM glutamine and 10% fetal calf serum and containing 60 μg/ml penicillin, 100 μg/ml streptomycin, and 50 mg/L kanamycin and incubated at 37°C in a humidified incubator with 10% CO2 in air. Transient DNA transfections were done as follows: 1.5 × 105 293T cells were plated in each well of a six-well plate (Corning Life Sciences, Acton, MA, http://www.corning.com/lifesciences) a day before transfection. Plasmid DNA (1.5 μg) was transfected to 293T cells by the calcium-phosphate/DNA precipitation method.
Flow Cytometry and Immunohistochemistry
293T cells (1 × 106/100-mm-diameter dish) were transfected and fixed in 100% methanol for 30 minutes, collected by low-speed centrifugation and resuspended in phosphate-buffered saline (PBS) and incubated for 40 minutes with primary antibody anti-IgM (A90-101A, 1:700; Bethyl Laboratories, Montgomery, TX, http://www.bethyl.com/) for 45 minutes, followed by 40-minute incubation with biotin-conjugated donkey anti-goat antibody (AP180B, 1:1,500; Chemicon, Temecula, CA, http://www.chemicon.com) and, finally, by 40-minute staining with Oregon Green 488-conjugated streptavidin (Molecular Probes, Eugene, OR, http://probes.invitrogen.com). Cells were resuspended in PBS containing 50 μg/ml RNAse A (Sigma-Aldrich) and 50 μg/ml propidium iodide (Sigma-Aldrich), incubated in the dark at 37°C for 30 minutes, and 10,000 to 20,000 cells were analyzed for DNA content by FACScan (BD Biosciences). Histograms were prepared using CellQuest software (BD Biosciences). For immunohistochemistry, embryos were fixed in 4% (vol/vol) phosphate-buffered formalin, dehydrated, and embedded in paraffin; sections were prepared, boiled for 10 minutes in 10 mM citrate buffer, pH 6.0, and cooled down for at least 2 hours. Sections were then blocked and permeabilized for 30 minutes at room temperature using a blocking solution (10% normal horse serum and 0.1% Triton X-100, in PBS) and incubated overnight at room temperature with biotinylated goat anti-mouse IgM antibody (115-065-075, 1:2,000; Jackson Immunoresearch Laboratories, West Grove, PA, http://www.jacksonimmuno.com) and then with peroxidase-labeled avidin-biotin complex (ABC complex, K-0377; DAKO, Glostrup, Denmark, http://www.dako.com). Sections were then washed and developed in diamino-benzidine reagent (Sigma-Aldrich), rinsed in water, counterstained with hematoxylin, mounted with mounting reagent Enthellan (Merck & Co., Whitehouse Station, NY, http://www.merck.com), and analyzed using a light microscope (Nikon Eclipse E800; Nikon, Tokyo, http://www.nikon.com/).
293T cells (1.5 × 105) were seeded on glass coverslips (13 mm in diameter). Twenty-four hours after transfection, cells were fixed, permeabilized, and incubated with goat anti-IgM (A90-101A, 1:700; Bethyl) for 45 minutes, washed in PBS, incubated 40 minutes with biotin-conjugated donkey anti-goat antibody (AP180B, 1:1,500; Chemicon) and then stained 40 minutes with Oregon Green 488-conjugated streptavidin (Molecular Probes), and washed in PBS. Cells were viewed and photographed using Nikon E 1000 and the Openlab 4.0.1 software (Improvision, Lexington, MA, https://www.improvision.com).
The expression constructs of cytosolic and transmembrane mesenchymal Ig μHC was generated as follows: transcripts were cloned from the MBA-2.1 cDNA library into pCANmycA vector (Stratagene, LA Jolla, CA, http://www.stratagene.com). The cytosolic mesenchymal Ig μHC was amplified using the sense primer 5′-CCGGAATTCGGCTGCCTAGCCCGGGACTTCC-3′ and the antisense primer 5′-CGGCTCGAGTCAATAGCAGGTGCCGCCTGTGTC-3′; the transmembrane mesenchymal Ig μHC was amplified using the sense primer 5′-CCGGAATTCGGCTGCCTAGCCCGGGACTTCC-3′ and the antisense primer 5′-CGGCTCGAGTCATTTCACCTTGAACAGGGTGACG-3′. Both fragments were digested with XhoI and EcoRI and ligated into pCANmycA vector. Another construct of the cytosolic mesenchymal Ig μHC was designed for the cell-free transcription/translation assay described. The insert was cloned into the vector pBluescript II KS (±) (Stratagene) using the same primers (only the XhoI restriction site was modified by NotI). The vector containing the cytosolic form of the full-length Ig μHC was a kind gift from Dr. Yair Argon (University of Chicago, Chicago, www.uchicago.edu).
The transcription/ translation experiment was performed by means of the TNT quick-coupled transcription/translation system (Promega, Madison, WI, http://www.promega.com) according to the instructions of the producer.
Total RNA was extracted using TriReagent (Molecular Research Center, Cincinnati, OH, http://www.mrcgene.com) and 2- to 40-μg samples were hybridized with the μHC constant region probe. Separation of RNA samples by electrophoresis was performed on 1% agarose, 5.2% formaldehyde (37% solution), and 1× 4-morpholinepropanesulfonic acid (MOPS) gels. RNA was transferred to a Hybond-N membrane (Amersham Biosciences, Piscataway, NJ, http://www.amersham.com). The blot was hybridized at 68°C for 60 minutes in express hybridization solution containing α-[32P]-labeled probe. After washing, the blot was exposed to x-ray film (Kodak, Rochester, NY, http://www.kodak.com).
Cells were harvested in 400 μl ice-cold radioimmunoprecipitation assay lysis buffer (50 mM Tris, 150 mM NaCl, 1% Nonidet P-40, 0.5% deoxycholate, 0.1% SDS, 10% glycerol, and 1 mM EDTA, pH 8, plus 1/100 protease Inhibitor Cocktail [Sigma-Aldrich]) followed by centrifugation (15,000g, 15 minutes, 4°C). The supernatants were boiled after the addition of SDS sample buffer (5% glycerol, 2% SDS, 62.5 mM Tris-HCl, pH 6.8, 2% 2-mercaptoethanol, and 0.01% bromphenol blue), separated on 10% SDS-polyacrylamide gel, and transferred to nitrocellulose membranes (Whatman Schleicher and Schuell, Keene, NH, http://www.whatman.com/). The membranes were incubated for 1 hour in TBS-T (25 mmol/l Tris-base, 150 mmol/l NaCl, and 0.05% Tween 20, pH 7.4) containing 5% (wt/vol) nonfat dry milk to block nonspecific antibody binding and then incubated with horseradish peroxidase-goat anti-mouse IgM antibody (115-035-020, 1:5,000; Jackson). Antibody-labeled proteins were detected by enhanced chemiluminescence (ECL) substrate on Kodak film.
In Situ Hybridization
WT and IgM−/− embryos (12.5 days post coitum [dpc]) and their extraembryonic tissues were fixed in buffered 10% formalin at 4°C for 16 hours and processed for paraffin embedding. The 5-μm thick paraffin sections were prepared and mounted on TESPA-subbed SuperFrost Plus slides (Menzel-Glaser, http://www.menzel.de). To generate the probes, a consensus fragment of either mesenchymal Ig μHC or mesenchymal Ig δHC cDNAs was cloned into pCDNA3 vector (Stratagene) and pGEM-T (Promega), respectively. Sense and antisense riboprobes were transcribed in vitro (Promega kit) using 35S-labeled UTP. Radioactive in situ hybridization was performed according to a previously published protocol  with slight modifications. In brief, deparaffinized sections were heated in 2× standard saline citrate (SSC) at 70°C for 30 minutes, rinsed in distilled water, and incubated with 10 μg/ml proteinase K in 0.2 M Tris-HCl, pH 7.4, and 0.05 M EDTA at 37°C for 20 minutes. After proteinase digestion, slides were postfixed in 10% formalin in PBS (20 minutes), quenched in 0.2% glycine (5 minutes), rinsed in distilled water, rapidly dehydrated through graded ethanols, and air-dried. The hybridization mixture contained 50% formamide, 4× SSC, pH 8.0, 1× Denhardt's solution, 0.5 mg/ml herring sperm DNA, 0.25 mg/ml yeast RNA, 10 mM dithiothreitol (DTT), 10% dextran sulfate, and 3 × 104 cpm/μl 35S-UTP-labeled riboprobe. After application of the hybridization mixture sections were covered with sheets of polypropylene film cut from autoclavable disposable bags and incubated in humidified chamber at 65°C overnight. After hybridization, covering film was floated off in 5× SSC with 10 mM DTT at 65°C, and slides were washed at high stringency (2× SSC, 50% formamide, and 10 mM DTT at 65°C for 30 minutes) and treated with RNAse A (10 μg/ml) for 30 minutes at 37°C. Slides were next washed in 2× SSC and 0.1× SSC (15 minutes each) at 37°C. Then slides were rapidly dehydrated through ascending ethanols and air-dried. For autoradiography, slides were dipped in Kodak NTB-2 nuclear track emulsion diluted 1:1 with double-distilled water and exposed for 3 weeks in light-tight box containing desiccant at 4°C. Exposed slides were developed in a Kodak D-19 developer, fixed in a Kodak fixer, and counterstained with hematoxylin-eosin. Microphotographs were taken using Zeiss Axioscop-2 microscope (Carl Zeiss, Jena, Germany, http://www.zeiss.com) equipped with Diagnostic Instruments Spot RT CCD camera (Sterling Heights, MI, http://www.diaginc.com/).
Reverse transcription-polymerase chain reaction (RT-PCR) was performed on cDNAs obtained from the indicated cells and tissues. Total RNA was isolated from the above cells or tissues using either TriReagent (Molecular Research Center) or RNeasy Mini Kit (Qiagen, Valencia, CA, http://www1.qiagen.com) in accordance with the manufacturer's instructions. To prevent genomic DNA contamination, samples were treated with DNase (Roche Diagnostics, Basel, Switzerland, http://www.roche-applied-science.com). Single-strand cDNAs were then prepared using SuperScript reverse transcriptase (Invitrogen, Carlsbad, CA, http://www.invitrogen.com). Analysis of gene expression was done using polymerase chain reaction (PCR) with ReadyMix PCR Master Mix (ABgene, Surrey, U.K., http://www.abgene.com). The primers that were used are summarized below. Primers generated for heavy and light chains were designed to the constant region of the specific chain mentioned. The oligonucleotide primers examined are shown as follows:
μHC sense (s) 5′-TAGGTTCAGTTGCTCACGAG-3′, antisense (as) 5′-GTGACCATCGAGAACAAAGG-3′; δHC s 5′-CTCCTCTCAGAGTGCAAAGCC-3′, as 5′-GGATGTTCACAGTGAGGTTGC-3′; αHC s 5′-CATGAGCAGCCAGTTAACCCTG-3′, as 5′-ATGCAGCCATCGCACCAGCAC-3′; εHC s 5′-GACTCCCTGAACATGAGCACTG-3′, as 5′-GGTACTGTGCTGGCTGTTTGAG-3′; γ1HC s 5′-CTGGAGTCTGACCTCTACACTCTG-3′, as 5′-CAGGTCAGACTGACTTTATCCTTG-3′; γ2AHC s 5′-GATGTCTGTGCTGAGGCCCAGG-3′, as 5′-GGAAGCTCTTCTGATCCCAGAG-3′; γ2BHC s 5′-GAGTCAGTGACTGTGACTTGGAAC-3′, as 5′-ACCAGGCAAGTGAGACTGAC-3′; γ3 HC s 5′-CTGGCTGCAGTGACACATCT-3′, as 5′-GGTGGTTATGGAGAGCCTCA-3′; κLC s 5′-CTTGCAGATCTAGTCAGAGCC-3′, as 5′-CAATGGGTGAAGTTGATGTCTTG-3′; λLC s 5′-CCAAGTCTTCGCCATCAGTCAC-3′, as 5′-GAACAGTCAGCACGGGACAAAC-3′, VH deg s 5′-SARGTNMAGCTGSAGSAGTCWGG-3′ adopted from ,VH J558 s 5′-ATAGCAGGTGTCCACTCC-3′ adopted from , λ5 s 5′-TGGGGTTTGGCTACACAGAT-3′, as 5′-CCCACCACCAAAGACATACC-3′; VpreB s 5′-GTACCCTGAGCAACGACCAT-3′, as 5′-GTACCCTGAGCAACGACCAT-3′; Igα s 5′-TGCCTCTCCTCCTCTTCTTG-3′, as 5′-TGATGATGCGGTTCTTGGTA-3′; Igβ s 5′-TCAGAAGAGGGACGCATTGTG-3′, as 5′-TTCAAGCCCTCATAGGTGTGA-3′; and B220 s 5′-CAAAGTGACCCCTTACCTGCT-3′, as 5′-CTGACATTGGAGGTGTGTGT-3′.
Rapid Amplification of cDNA Ends
The 5′-end of the mesenchymal Ig μHC or mesenchymal Ig δHC transcripts were mapped using the FirstChoice RNA Ligase-Mediated Rapid amplification of cDNA Ends kit (Ambion, Austin, TX, http://www.ambion.com) in accordance with manufacturer's instructions. RNA was isolated from either MBA-2.1 cells or IgM−/− MEFs using the RNeasy Mini Kit (Qiagen) according to manufacturer's instructions. Nested PCRs were used to amplify the 5′-end of the mesenchymal Ig μHC transcript. The 5′-rapid amplification of cDNA ends (RACE) outer primer provided was used for the outer PCR reaction, together with the specific primer 5′-CACGGCAGGTGTACACATTCAGGTTC-3′, whereas the 5′-RACE inner primer provided, together with the specific primer 5′-CGTGGCCTCGCAGATGAGTTTAGACTTG-3′, was used for the inner PCR reaction. Nested PCRs were then used to amplify the 5′-end of the mesenchymal Ig δHC transcript. The 5′ RACE outer primer provided was used for the outer PCR reaction together with the specific primer 5′-GGATGTTCACAGTGAGGTTGC-3′, whereas the 5′-RACE inner primer provided, together with the specific primer 5′-AGTGACCTGGAGGACCATTG-3′, was used for the inner PCR reaction. The 3′-end of the mesenchymal Ig δHC transcript was mapped using the same total RNAs. First-strand cDNA was generated using a tagged oligo(dT) primer (Gibco-BRL, Gaithersburg, MD, http://www.gibcobrl.com) followed by RNAse-H reaction. The cDNA was then used as a template for PCR performed with the universal amplification primer (UAP) provided and the specific primer 5′-GCAACCTCACTGTGAACATCCTG-3′. A second PCR was obtained with the same UAP primer and the specific primer 5′-GCTTAATGCCAGCAAGAGCCTAG-3′. The resultant PCR products were cloned into pGEM-T (Promega) and sequenced.
Student's paired t-test was used to evaluate the significance of differences between experimental groups.
Expression of Pre-BCR/BCR Components in Primary and Long-Term Cultured Mesenchyme: IgM Deficiency Results in Upregulation of δ Chain mRNA
We set out to examine the possibility that the mesenchyme expresses BCR components. RT-PCR detected expression of Ig μHC mRNA in primary MEF cell strains from 12.5-dpc embryos and in a cloned mouse bone marrow stromal cell line, MBA-2.1  (Fig. 1Ai). In contrast, MEFs from IgM−/− that serve as a negative control had no such transcript. No expression of light chains was detected in WT MEF or MBA-2.1 cells, indicating that the μHC expression is not as a result of contamination of the mesenchymal cell cultures with lymphocytes. Northern blot analysis of mRNA from MBA-2.1 cells with a probe for μHC revealed a short transcript (∼2 kilobases) (Fig. 1B). Analysis of additional Ig isotypes indicated that δ chain is not found. Surprisingly, δ substituted for μ isotype in the IgM−/− MEF (Fig. 1Ai). We further identified VpreB expression in the three cell types under study (Fig. 1Aii), as well as Igα, Igβ, and λ5 that were, however, detected inconsistently in WT and IgM−/− MEFs and were not expressed in the MBA-2.1 cell line (Fig. 1C). Neither γ nor ε Ig isotypes were expressed in MEFs, nor were κ and λ LCs (Fig. 1Ai). The μHC transcript was further detected in several murine mesenchymal cells lines that exhibit MSC functions [22, 24, 25] as well as in primary bone marrow-derived MSCs (Fig. 1D).
Expression of Pre-BCR/BCR Components in the Early Embryo
To ascertain that the detection of Ig gene products in cultured mesenchyme was not an in vitro-restricted phenomenon, we turned to the examination of these mRNAs in embryonic tissues. Two different sets of primers were designed to enable RT-PCR detection of, and differentiation between, germline versus rearranged Ig μHC transcripts. Unfertilized oocytes were found to express an unrearranged Ig μHC transcript, whereas δHC was not detected (Fig. 2). In addition, light chains expression was not observed (data not shown). Because the μHC transcripts were found in cells and tissues that were not expected to harbor such mRNAs, we verified the validity of the analysis by examining tissues from μ chain-deficient mice in which such transcripts were indeed absent as expected (Fig. 2). In the IgM−/− mouse oocytes, the unrearranged Ig μHC transcript was replaced by an unrearranged Ig δHC transcript. Likewise, morulae from WT mice expressed unrearranged Ig μHC mRNA, whereas no δHC was detectable. The reverse was found in the IgM−/− mice (Fig. 2). Although these results could imply that the expression of μ and δ chains is mutually exclusive, analysis of 11.5-dpc heterozygous (IgM+/−) embryos revealed that both the unrearranged Ig μHC and the Ig δHC transcripts were concomitantly detectable (Fig. 2). We subsequently investigated the expression of the Ig μHC mRNA in older 12.5-dpc WT and IgM−/− embryos. The Ig μHC mRNA transcript was expressed both in the embryo proper and in the yolk sac, whereas no Ig δHC expression was observed. In the IgM−/− embryos and yolk sacs, expression of only the δHC mRNA was observed. The lack of expression of the B-cell marker B220 (data not shown) or transcripts derived from Ig μ rearrangements further supports the inference that the Ig μHC gene in WT embryos and Ig δHC in IgM−/− embryos are being expressed by nonlymphoid cells. To further assure that maternal lymphocytes do not account for detection of Ig transcripts in the embryonic tissues, RT-PCR analysis was performed using 12.5-dpc WT embryos that were transplanted, at the morulae stage, into IgM−/− pseudopregnant recipient mothers. These embryos did express μHC (Fig. 2), thus providing strong evidence that Ig HC mRNAs that were detected are endogenous to the embryo.
Although IgM−/− mice exhibit normal B-cell development and maturation , the antibody repertoire in these animals is altered . The question was therefore raised as to whether the lack of μHC mRNA would impact mesenchymal cell functions and early development. In our animal stock, IgM−/− mice had smaller litter sizes than their WT counterparts (Fig. 3Ai). To examine early stages of development, we obtained 2.5-dpc morulae from both IgM−/− and WT mice. Four independent experiments were performed; in each experiment, morulae were harvested from 12 to 20 female mice per group. The results of the average total number of morulae per female are shown in Figure 3Aii. IgM−/− mice had an average number of 9.66 ± 0.26 total morulae per female compared with 16.2 ± 0.28 per WT female. Furthermore, morulae were scored as having good developmental potential (being “intact”) if compacted and containing at least four cells and up to 16 cells. IgM−/− mice had an average number of 2.9 ± 0.14 intact morulae per female compared with 7.2 ± 0.14 per WT female (Fig. 3Aiii). The reduced frequency of intact morulae imply a role for μ chain mRNA or protein in early development. Western blot analysis of protein extracts from 12.5-dpc embryo proper versus the yolk sac detected protein bands of 75 and 50 kDa only in WT yolk sac (Fig. 3Bi) visceral layer (Fig. 3Bii). This protein was maternally derived; 2.5-dpc WT morulae were transplanted into IgM−/− pseudopregnant recipient mothers and vice versa. Subsequently, embryos were collected at 12.5 dpc. Both Western (Fig. 3Ci) and immunohistochemical (Fig. 3Cii, 3Ciii) analysis of tissues indicate that only yolk sacs derived from IgM−/− embryos transplanted into WT pseudopregnant recipient mothers were IgM-positive (Fig. 3Ci, 3Ciii).
Identification of the μHC and δHC mRNA-Expressing Cells Within Mid-Gestation Mouse Embryo
We next examined the nature of cells in mid-gestation that express BCR components. Sections from both WT and IgM−/− embryos were hybridized in situ with 35S-labeled antisense RNA probes derived from the constant regions of either μHC (Fig. 4) or δHC (Fig. 5). In 12.5-dpc WT embryos, the positive cells expressing μHC were mesenchymal cells located in the loosely packed mesenchyme adjacent to the spinal cord (Fig. 4A–4C), attached to the yolk sac (Fig. 4D), or similar cells in the proximity of blood vessels (Fig. 4E, 4F). No signal for μHC was detected in IgM−/− embryos (Fig. 4G, 4H). In 12.5-dpc IgM−/− embryos, δHC-positive cells were observed located in the proximity of blood vessels (Fig. 5A–5C) or embedded within loose mesenchymal tissue (Fig. 5D). Thus, the in vivo identification of the Ig HC mRNA-expressing cells in mid-gestation embryos corroborates the in vitro detection of these mRNAs in mesenchyme.
Cloning and Structure Analysis of Ig μ and δHC Transcripts from Mesenchymal Cells
RACE using RNA derived from MBA-2.1 cells indicated that the mesenchymal Ig μHC transcript is an unrearranged truncated form (Fig. 6B). A unique 5′-UTR is found in the mRNA that is homologous to a part of the μ switch region D-q52. Downstream to this 5′-UTR, the clone encodes the complete four exons of the Ig μHC constant region. Thus, this mRNA is a hybrid transcript that includes some exons from previously characterized genes (supplemental online Fig. 1). Both cytosolic and membrane type of transcript were cloned from the stromal cell line (Fig. 6B). The mesenchymal from of δHC lacks the variable segments that are upstream to the μHC constant region in the Ig locus (Fig. 6A). The MEF form of δHC consists of only the C region of the lymphoid form. The DNA sequence is composed of two C-region domains, Cδ1 and Cδ3, separated by the CδH hinge domain. A 5′-UTR stretch of 39 bases is present upstream to the described C region (Fig. 6D and supplemental online Fig. 1), which is homologous to a part of the μ switch region D-q52. Four distinctive 3′ ends that generate four mRNA isoforms of the mesenchymal truncated δ were isolated (Fig. 6D). Thus, mesenchyme expresses truncated forms of μ and δ HCs that consist of the C region only, that is, Cμ and Cδ. Because both transcripts contained in-frame ATGs (Fig. 6B, 6D), proteins could potentially be encoded.
Cμ mRNA Encodes a 50-kDa Protein That Assumes an Intracellular Localization and Causes Growth Arrest
To examine whether the mesenchyme derived Cμ does encode a protein, we first examined the cDNA in an in vitro transcription/translation system. Figure 7A shows that this mRNA encodes a newly synthesized protein of approximately 50 kDa. An expression vector containing the mesenchymal Cμ transcript and the full-length Ig μHC were used to transfected 293T cells (Fig. 7B). Western blot analysis of extracts from Cμ-transfected cells showed a protein band at about 50 kDa (Figure Aiii). Whereas mesenchymal Cμ was found in a diffuse cytoplasmic staining, the full-length μHC chain was observed in punctate structures scattered throughout the cells (Fig. 7B). To get an insight as to the possible function of this truncated protein, we studied the effects of overexpression in cultured cell lines. The mesenchymal Cμ and the full-length μHC-form B lymphocytes were compared following transfection of 293T cells. The overexpression of mesenchymal Cμ results in morphological changes in the cultured cells that were not seen with the full-length μHC (Fig. 7Cii). Flow cytometric analysis showed that cells expressing the mesenchymal Cμ exhibit a pronounced G1 arrest (Fig. 7Di, right panel). Cells negative for IgM expression (Fig. 7Di, left panel) or cells transfected with empty plasmid (Fig. 7Dii) have normal cell-cycle distribution. In contrast, overexpression of full-length Ig μHC did not affect the cell cycle in a similar manner (data not shown).
The precise gene expression pattern of mesenchymal cells has thus far not been determined. We have previously demonstrated that these poorly defined cell populations express TCR mRNA components that are of germ-line nature . It is shown herein that unfertilized mouse oocytes, as well as morulae, express components of the BCR. This expression continues later in embryogenesis in at least a subset of mesenchymal cells, as well as in cultured adult mesenchymal cells. Because the mesenchyme is widely represented in most organs and tissues, it is possible that the presence of the mesenchymal Ig Cμ transcripts may be similarly frequent. This premise is corroborated by the report of expression of truncated TCR forms in neuronal cells [36, 37].
Various forms of truncated Ig μHC transcripts have previously been characterized, such as the Dμ protein encoded by a DJH rearrangement [9, 38] and the heavy chain disease  μ chain protein, which lacks the entire variable region . The Ig HC D to JH rearrangement occurs frequently in developing T cells , but the V to DJH rearrangement is limited to developing B cells . All of these occurrences of truncated forms of Ig HCs were reported to occur in lymphoid cells, whereas we report herein that Ig μHC and δHC transcripts are expressed in nonlymphoid, mesenchymal cells as well as in unfertilized oocytes and early embryos.
The sequence analysis and expression of the recombinant cDNAs clearly show that the truncated mesenchymal Cμ and Cδ mRNAs may encode proteins. The calculated predicted molecular size of the putative mesenchymal Cμ protein is 50 kDa, whereas the complete form is 75 kDa. Although a 50-kDa band reactive with anti-IgM antibodies was detected in the yolk sac, this protein was shown in our experiments to originate from a maternal source. Thus, the yolk sac seems to be a sequestration site for maternal IgM. The fact that this IgM is found both in a high and a low mol. wt. form may indicate that it is in a process of degradation. It is possible therefore that the yolk sac serves, in this respect, as a barrier that adsorbs and degrades IgM to minimize the transfer of the molecule into the embryo. Indeed, IgG and small amounts of IgM are placentally transferred from the mother to fetus during gestation in the mouse [42, 43]. Light chains from maternal sources were found to be involved in cell-cell interactions during cerebral cortical development in brain from 12-dpc embryos . These maternal light chains are involved in immunoglobulin-like immunoreactivity. The degradation machinery may be more effective within the embryo proper and could explain our lack of ability to demonstrate the presence of significant amount of μ chain protein despite the abundance of the corresponding mRNA.
Several lines of evidence suggest that the expression of Ig μ chain mRNAs may have functional significance in mesenchymal cells. First, the fact that δ substituted for μ suggests that the former may compensate for the lack of μ chain. One possibility is that the presence of μ in WT cells suppresses δ expression, which is derepressed in the IgM−/− mice. However, because both chains are expressed concomitantly in mesenchyme in heterozygous mice, and also in B lymphocytes, a more feasible interpretation is that δ compensates for the lack of μ. The deletion of μHC positions the δ gene closer to a putative promoter upstream to the Cμ region and may result in the observed expression of δ in the IgM−/− mice. Second, overexpression of the mesenchymal μ chain in 293T cells caused a morphological change that was not observed following overexpression of the full-length form of this molecule. The functions of other truncated Ig μHC proteins, which have been characterized in B cells, were also shown to be different from the functions of full-length rearranged Ig μHC [9, 14, 40, 45, , , , , –51]. Although Dμ is functional [11, 52, 53], its activity is aberrant because it blocks further B-cell development [10, 12, 13]. Third, overexpression of the Cμ in 293T cells resulted in growth arrest, suggesting possible involvement in cell growth control. Fourth, μHC deficiency was associated with a poor number of intact morulae that could be derived from pregnant females. These findings point to the possible requirement of Ig gene expression in early development. In the IgM−/− animals used in this study, a deletion of the Cμ and μ-δ intron including the regulatory cis-elements for Cδ expression has been performed . However, the animals exhibit a normal compartment of immature B cells. IgM+ cells were absent but were replaced by IgD-positive cells. Consequently, IgM−/− animals were capable of mounting specific B-cell responses. In view of this normal capability, it is expected that any defect that IgM−/− animals may exhibit, including the reduced incidence of intact morulae, would not be secondary to immune deficiency .
We did not observe truncated IgM proteins in the embryo corresponding to the mRNAs that we detected. It should be therefore considered that either the truncated Ig proteins are expressed at a very low level, below the power of detection of the methods used due to, for example, a high-level of degradation. On the other hand, the transcripts may not encode a significant amount of protein, and thus, the implied functions could be mediated by the mRNA itself. Our findings imply that this mRNA or the encoded proteins may have more than one function: the one could be related to early embryo development and the other to mesenchymal cell functions. The mesenchymal undifferentiated appearance, taken together with the capacity of MSCs to differentiate to a variety of cell types, suggests that these cells may in fact be in a stem state, expressing a vast range of genes [54, 55]. The same holds for the early embryo that harbors bona fide stem cells. The findings reported herein might reflect the promiscuous gene expression of stem cells but further suggest that the expression of BCR mRNAs serves specific functions.
The authors indicate no potential conflicts of interest.
D.Z. is an incumbent of the Joe and Celia Weinstein Professorial Chair at the Weizmann Institute of Science. We are indebted to Varda Segal for excellent technical assistance. This work was supported by The Charles and David Wolfson Charitable Trust and Ruth Zeigler Trust grants for Stem Cell Research at the Weizmann Institute of Science and by the Gabrielle Rich Center for Transplantation Biology.