GFP is frequently used as a marker for tracking donor cells adoptively transplanted into recipient animals. The human ubiquitin C promoter (UBC)–driven-GFP transgenic mouse is a commonly used source of donor cells for this purpose. This mouse was initially generated in the C57BL/6 inbred strain and has been backcrossed into the BALB/cBy strain for over 11 generations. Both the C57BL/6 inbred and BALB/cBy congenic UBC-GFP lines are commercially available and have been widely distributed. These UBC-GFP lines can be a convenient resource for tracking donor cells in both syngenic MHC-matched and in allogenic MHC-mismatched studies as C57BL/6 (H-2b) and BALB/cBy (H-2d) have disparate MHC haplotypes. In this report, we surprisingly discover that the UBC-GFP BALB/cBy congenic mice still retain the H-2b MHC haplotype of their original C57BL/6 founder, suggesting that the UBC-GFP transgene integration site is closely linked to the MHC locus on chromosome 17. Using linear amplification–mediated PCR, we successfully map the UBC-GFP transgene to the MHC locus. This study highlights the importance and urgency of mapping the transgene integration site of transgenic mouse strains used in biomedical research. Furthermore, this study raises the possibility of alternative interpretations of previous studies using congenic UBC-GFP mice and focuses attention on the necessity for rigor and reproducibility in scientific research.
Green fluorescent protein was isolated from the jellyfish Aequorea victoria in 1962 (1), and the GFP gene was cloned 30 years later (2). Within years of its cloning, GFP and many associated derivatives (3–5) became common research tools for the tracking of cells and proteins in vivo.
One such application is the marking of donor hematopoietic cells used for transplantation so that they can be identified hours, days, weeks, or even months later in the recipient animals. Commonly, these donor animals are generated by the microinjection of the GFP gene DNA into fertilized mouse eggs to produce a transgenic mouse (6, 7). Thus far, over 1000 GFP transgenic mice have been created (8), and of them, more than 500 transgenic GFP lines can be accessed through commercial vendors like The Jackson Laboratory (TJL). Ideally, the expression of the GFP transgene is controlled by a ubiquitously expressed promoter so that high levels of GFP will be found in all cells (9). The human ubiquitin C promoter (UBC) is found to direct high-level ubiquitous expression of transgenes in mice (10) and was used by Schaefer et al. (11) to generate UBC-GFP transgenic mice that express high levels of GFP in all leukocytes and RBCs. Importantly, these UBC-GFP transgenic mice were generated directly in inbred C57BL/6 fertilized embryos, and a homozygous UBC-GFP C57BL/6 inbred line was established by backcrossing to C57BL/6.
The chromosomal integration site for most GFP transgenes is unknown as only a few of them have been mapped (12, 13). As transgene integration is usually random, there is always a concern of insertional mutagenesis or chromosomal position effects in which the transgene affects the expression of closely linked endogenous genes or the chromosomal sequences flanking the integration site affect the expression of the transgene, respectively.
To prevent transplant rejection and control for genetic background differences between strains of mice, it has been a common practice for investigators to breed transgenic GFP mice to inbred mouse strains and generate congenic mice. Generally, after four to nine generations of serial backcrossing, genetic background differences are minimized to an acceptable degree such that over 99% of the original donor strain–specific genetic differences are eliminated (14). However, it is well documented that the random insertion of transgenes can disrupt endogenous genes (15–17). Furthermore, chromosomal sequences that flank the transgene integration site will cosegregate and be coinherited along with the transgene. Thus, strain-specific position effects due to random chromosomal integration of a transgene can exert unwanted and many times unknown negative effects on experimental outcomes. Moreover, in studies using both congenic and inbred strains, these remaining genetic differences are often overlooked or simply ignored by most investigators.
In this study, we report that for one of the most widely distributed GFP transgenic mouse lines, UBC-GFP, the GFP transgene is integrated next to the mouse MHC locus. Thus, congenic UBC-GFP mice can retain the MHC haplotype of their original transgenic founder strain, even after many generations of backcrossing. In the absence of this knowledge, congenic UBC-GFP cell transfer experiments designed to be MHC matched are in reality MHC mismatched and vice versa. This study highlights the importance of mapping the transgene integration site of each transgenic mouse line used in biomedical research.
Materials and Methods
All animal procedures were approved by the University of Alabama at Birmingham’s Institutional Animal Care and Use Committee. Humanized Cooley anemia mice were derived by targeted gene replacement in ES cells (18–20). C57BL/6 UBC-GFP transgenic mice [C57BL/6-Tg(UBC-GFP)30Scha/J, JAX stock number 004353], BALB/cBy congenic UBC-GFP transgenic mice [CByJ.B6-Tg(UBC-GFP)30Scha/J, JAX stock number 007076], wild type BALB/cBy mice (BALB/cByJ, JAX stock number 001026), and wild type C57BL/6 mice (JAX stock number 000664) were all obtained from TJL.
Total PBMCs were prepared by lysing red cells in ACK lysis buffer (A1049201; Thermo Fisher Scientific). Bone marrow cells were harvested from tibias and femurs of euthanized mice by flushing the bones with Dulbecco PBS (Thermo Fisher Scientific). Cells were washed, counted, stained with Abs specific to H-2 Kd (H-2 Kd-APC, SF1-1.1.1; eBioscience) and H-2 Kb (H-2 Kb-PE, AF6-22.214.171.124; eBioscience), and analyzed by flow cytometry on a FACS Calibur (Becton Dickinson).
Linear amplification–mediated PCR
Linear amplification–mediated (LAM)-PCR was performed by following the protocol described by Schmidt et al. (21) with modifications. In brief, a 5′ biotinylated primer (bovine growth hormone [bGH] F1, 5′-/Biosg/CATCGCATTGTCTGAGTAGGTGT-3′) complementary to the 3′ end of the bGH sequence of the UBC-GFP transgene was used for linear extension into the 3′ chromosomal flanking sequence. One hundred cycles of linear amplification with 100 ng of BALB/cBy UBC-GFP genomic DNA as templates were carried out with 0.1 μl Ex Taq (Takara Ex Tag, 5 U/μl, 3P RR001A; Takara Bio). An additional fresh 0.1 μl Ex Taq was added after 50 cycles. The amplification product was incubated with 20 μl streptavidin beads (Dynabeads M280, PN 112.05D; Dynal) at room temperature for 30 min to allow the linkage buildup between the biotinylated ssDNA and the streptavidin beads. After a brief incubation at 90°C for 1 min, the beads bound to DNA were pulled down using Promega magnetic particle concentrator, and supernatant was removed. After removal from the magnet, the pelleted bead–bound DNA was washed with buffer, pulled down again using Promega magnetic particle concentrator, and supernatant removed. The linear-amplified ssDNA was converted into dsDNA by using Klenow polymerase (PN BP3201-1; Thermo Fisher Scientific) and random hexamer primers (Random Hexamers, PN C1181; Promega) at 37°C for 1 h. After one wash/pull down, the dsDNA was subjected to digestion by the four-cutter restriction endonuclease Tsp509I for 1 h. The digested product was washed and pulled down and then ligated to 100 pmol double-stranded asymmetric linker cassettes (LC) using T4 DNA ligase (PN M0202S; New England BioLabs) at 16°C overnight. The double-stranded LC was made by annealing two oligos LC Tsp509I forward (F) and LC Tsp509I reverse (R) (5′-pAATTGCATGTGTCAGAGGTTTTCACCGTCATCACCGAAACGC-3′ and 5′-GCGTTTCGGTGATGACGGTGAAAACCTCTGACACATGC-3′, respectively) together. The two oligos were heated to 94°C and then slowly cooled to room temperature to form the annealed double-stranded LC. After ligation to the linker, the ligated product was washed, pulled down, resuspended in H2O, and was used as the template for nested PCR. The first nested PCR primers were LC P1 (5′-GCGTTTCGGTGATGACGGTG-3′) and bGH nestPCR F1 (5′-AGGACAGCAAGGGGGAGGATT-3′), the second nested PCR primers were LC P2 (5′-GGTGATGACGGTGAAAACCTC-3′) and bGH nestPCR F2 (5′-GGGAGGATTGGGAAGACAAT-3′), and the third nested PCR primers were LC P3 (5′-GTGAAAACCTCTGACACATGC-3′) and bGH nestPCR F3 (5′-GATTGGGAAGACAATAGCAG-3′). Then, 0.2% of the first or second nested PCR product was used as a template for the second or third rounds of nested PCR, respectively. The second and third PCR products were separated on 2% agarose gels, and the third nested PCR product was excised, purified with Omega Bio-tek E.Z.N.A. Gel Extraction Kit (PN D2500-01), and sequenced by Sanger sequencing at the University of Alabama at Birmingham Heflin Center for Genomics Sciences. All DNA oligos were purchased from Integrated DNA Technologies.
UBC-GFP 5′ end junction PCR
The UBC-GFP transgene 5′ end flanking sequences were mapped through PCR amplification of the junction between the transgene and endogenous flanking genomic sequence. The primers used for this PCR were H-2 F (5′-CACACACACACACGTCCTTG-3′) and UBC R (5′-TCCATTCAAGACTCGGGAAC-3′) that bind to endogenous flanking sequence and UBC, respectively. The 1059-bp PCR product was electrophoresed on a 0.5% agarose gel, the band excised, purified, and then sequenced by Sanger sequencing.
We recently demonstrated that a humanized Cooley anemia mouse model could be rescued from lethal anemia by bone marrow transplantation without cytoreductive conditioning of the recipient (18). The donor bone marrow cells used for this study were isolated from UBC-GFP transgenic mice [C57BL/6-Tg(UBC-GFP)30Scha/J, TJL stock number 004353] (11) and have the H-2b MHC haplotype. The recipient humanized Cooley anemia mice are a mixture of C57BL/6 and 129 strains, which are also H-2b MHC haplotypes. Although the donor and recipient mice shared the H-2b MHC Ags, they differ with respect to the GFP transgene and murine hemoglobin that is only present in the donor versus human hemoglobin and minor histocompatibility Ags from the 129 strain found only in the recipient. In the absence of cytoreductive conditioning of the recipient, donor cells were able to establish low-level stable hematopoietic chimeras and reconstitute the entire erythron with GFP-positive donor-derived erythrocytes (18).
To increase the clinical significance of our findings, we wished to extend these studies using allogenic bone marrow from animals with a disparate MHC haplotype. Fortuitously, congenic UBC-GFP transgenic mice that were backcrossed to BALB/cBy inbred mice for 11 generations were available [CByJ.B6-Tg(UBC-GFP)30Scha/J, TJL stock number 007076]. Because BALB/cBy mice have the H-2d MHC haplotype, repeating our earlier transplantation experiments using the BALB/cBy congenic UBC-GFP mice as donors should model a strict allogenic MHC-mismatched study. Twenty weeks posttransplant into humanized Cooley anemia recipients without cytoreductive conditioning, the BALB/cBy UBC-GFP donor cells had established a low-level hematopoietic chimera, reconstituted the erythron, and rescued the animals from lethal anemia similar to our earlier findings (manuscript in preparation). To rigorously verify the disparate MHC haplotypes of donor and recipient animals used in our experiment, we examined the H-2 K allele of the UBC-GFP donor mice by flow cytometry. The BALB/cBy congenic UBC-GFP mice still expressed the H-2 Kb Ag instead of the expected H-2 Kd Ag normally present on cells of the BALB/cBy strain (Fig. 1). The UBC-GFP transgenic mice were initially generated by microinjection of DNA into C57BL/6 fertilized pronuclei (11). After 11 generations of backcrossing to BALB/cBy, 99.95% of the genome in UBC-GFP BALB/cBy mice should have been congenic with BALB/cBy. Only the C57BL/6 chromosomal sequence that flanks the transgene insertion site should be coinherited with the UBC-GFP transgene after extensive backcrossing to another strain.
Thus, we reasoned that the UBC-GFP transgene must be inserted in or close to the H-2 locus so that the H-2b haplotype of C57BL/6 would be coselected with GFP after each backcross. To prove this hypothesis, we mapped the UBC-GFP chromosomal insertion site by LAM-PCR (Fig. 2A) (21). After several rounds of nested PCR with primers complementary to the bGH polyadenylation (polyA) sequence of the UBC-GFP transgene and 3′ linker sequence, respectively, a specific LAM-PCR product was obtained, gel purified, and sequenced (Fig. 2B, left and center panels). As expected, the sequencing result indicated that the UBC-GFP transgene was inserted into chromosome 17 next to the H-2 locus ∼4.6 Mbp from the H-2 K1 gene (Fig. 3). To confirm the chromosomal location of the transgene integration site, we successfully PCR amplified and sequenced the 5′ end of the UBC-GFP transgene together with the 5′ flanking genomic sequence (Fig. 2B, right panel). Thus, the UBC-GFP transgene integration site is located on chromosome 17 and is linked to the MHC locus.
Cells isolated from UBC-GFP mice are useful for both in vitro assays and in vivo studies as they are easily identified and tracked by virtue of their fluorescence. The UBC-GFP transgene is expressed in all tissues, and all hematopoietic lineages can be tracked in adoptively transplanted recipients. The use of UBC-GFP transgenic mice is demonstrated by their broad distribution by TJL and by more than 200 published papers that cite the use of this mouse line. These mice can be bred to other inbred strains of mice to generate congenic mouse strains, and the best-known example is the CByJ.B6-Tg(UBC-GFP)30Scha/J line that has now been backcrossed to inbred BALB/cBy mice for more than 12 generations. Many publications report using these BALB/cBy UBC-GFP mice in their studies (22–31). Most of these studies transfer various GFP-positive donor cells into wild type BALB/cBy mice in what were presumed to be MHC-matched, syngenic studies. Ironically, in reality, these studies were just the opposite; the donor and recipient are MHC mismatched, and these were allogenic studies. It is possible that many more studies were initiated that transferred cells from donor UBC-GFP congenic mice into BALB/cBy recipients under the assumption of an MHC match but ultimately were abandoned and unreported because of a false negative experimental result after the rejection of the donor cells. The results and conclusions of any experiments that use the UBC-GFP BALB/cBy congenic mice may need to be carefully reexamined and evaluated in light of our new findings. The above publications highlight the importance and urgency for the mapping of transgene integration sites, especially those widely distributed lines used in biomedical research.
We mapped the UBC-GFP transgene insertion site into a noncoding region 4.6 million bases from the H-2 K1 gene on mouse chromosome 17, in which the recombination rate is ∼0.57 cM/million bases (32). At this recombination rate, the UBC-GFP transgene should segregate from the MHC locus approximately once in every 38 meioses. Careful immunophenotyping by flow cytometry (Fig. 1) of dozens of offspring from a UBC-GFP BALB/cBy backcross to BALB/cBy should be able to identify a UBC-GFP transgenic mouse with a cross-over event between the transgene integration site and the MHC locus, resulting in an H-2 Kb/d heterozygous phenotype. Such a mouse could then be used to ultimately generate homozygous congenic UBC-GFP BALB/cBy that are also homozygous for H-2d MHC haplotype of inbred BALB/cBy mice.
Transgene mapping can be accomplished through many different PCR-based strategies, including inverse PCR (33), T-linker PCR (34), ligation-mediated PCR (35), fusion primer and nested integrated PCR (36), and single-specific primer PCR (37, 38) as well as next-generation sequencing–based strategies, including whole genome sequencing (39) and targeted locus amplification (12, 40, 41). In this report, we used LAM-PCR (21, 42), an efficient and cost-effective method for mapping transgene integration sites that can easily be used to map the transgenes in other transgenic lines used in research.
Finally, although we demonstrate that congenic UBC-GFP BALB/cBy mice are currently not an MHC-matched strain for syngenic studies in BALB/cBy mice, we serendipitously have discovered their usefulness as a new BALB/cBy congenic UBC-GFP transgenic mouse line with an H-2b haplotype useful for stringent MHC-mismatched allogenic studies.
We thank Joseph Ruisi, UNICO National, and the Thalassemia-Cooley’s Anemia Group at the University of Alabama at Birmingham for support.
This work was supported by National Institutes of Health Grants R01 HL072351, R01 HL073440, R56 HL072351, G20RR022807-1, F31 HL120614, P30 AR048311, and P30 AI027667. This work was supported by a Research Fellowship grant from the Cooley’s Anemia Foundation. This work was sponsored in part by the Alabama Institute of Medicine Stem Cell Research Grants Program. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the State of Alabama or the Department of Public Health of the State of Alabama.
Abbreviations used in this article:
bovine growth hormone
The Jackson Laboratory
ubiquitin C promoter.
The authors have no financial conflicts of interest.