Comparison of two related lines of tauGFP transgenic mice designed for lineage tracing
- Linda Sharp†1,
- Thomas Pratt†1,
- Gillian E. MacKay2, 3,
- Margaret A. Keighren2, 4,
- Jean H. Flockhart2,
- Emma J. Chandler2,
- David J. Price1,
- John O. Mason1 and
- John D. West2Email author
© The Author(s). 2017
Received: 28 February 2017
Accepted: 1 June 2017
Published: 29 June 2017
The tauGFP reporter fusion protein is produced nearly ubiquitously by the TgTP6.3 transgene in TP6.3 mice and its localisation to microtubules offers some advantages over soluble GFP as a lineage marker. However, TgTP6.3 Tg/Tg homozygotes are not viable and TgTP6.3 Tg/− hemizygotes are smaller than wild-type. TP6.4 mice carry the TgTP6.4 transgene, which was produced with the same construct used to generate TgTP6.3, so we investigated whether TgTP6.4 had any advantages over TgTP6.3.
Although TgTP6.4 Tg/Tg homozygotes died before weaning, TgTP6.4 Tg/− hemizygotes were viable and fertile and only males were significantly lighter than wild-type. The TgTP6.4 transgene produced the tauGFP fusion protein by the 2-cell stage and it was widely expressed in adults but tauGFP fluorescence was weak or absent in several tissues, including some neural tissues. The TgTP6.4 transgene expression pattern changed over several years of breeding and mosaic transgene expression became increasingly common in all expressing tissues. This mosaicism was used to visualise clonal lineages in the adrenal cortex of TgTP6.4 Tg/− hemizygotes and these were qualitatively and quantitatively comparable to lineages reported previously for other mosaic transgenic mice, X-inactivation mosaics and chimaeras. Mosaicism occurred less frequently in TP6.3 than TP6.4 mice and was only observed in the corneal epithelium and adrenal cortex.
Mosaic expression makes the TgTP6.4 transgene unsuitable for use as a conventional cell lineage marker but such mosaicism provides a useful system for visualising clonal lineages that arise during development or maintenance of adult tissues. Differences in the occurrence of mosaicism between related transgenic lines, such as that described for lines TP6.3 and TP6.4, might provide a useful system for investigating the mechanism of transgene silencing.
KeywordsMosaic transgene expression Green fluorescent protein tauGFP Adrenal cortex lineages Preimplantation embryo Time-lapse imaging
Two lines of transgenic mice, TP6.3 and TP6.4, express CAG-tauGFP transgenes TgTP6.3 and TgTP6.4 respectively . Both lines TP6.3 and TP6.4 were produced by electroporation of the same construct into embryonic stem (ES) cells, which were then used to generate ES cell chimaeras and founder transgenic mice. Line TP6.3 expresses a tau-green fluorescent protein (tauGFP) fusion protein almost ubiquitously. Its localisation to microtubules has some advantages over soluble GFP as a lineage marker and it is well suited for labelling axons, blood vessels and preimplantation embryos [1, 2]. This TgTP6.3 tauGFP marker has been used to identify cell fusion between ES cells and other cell types  and as a lineage marker for macrophages , ovarian cells , preimplantation embryos  and, in particular, neural tissues [7–12].
TgTP6.4 transgene expression was reported to be strong in many tissues yet only weak in much of the central nervous system but it was not fully characterised . As TgTP6.3 Tg/Tg homozygotes are lethal and TgTP6.3 Tg/− hemizygotes are smaller than wild-type (WT) siblings we evaluated the second tauGFP transgenic line to determine whether TgTP6.4 had any advantages over the TgTP6.3 marker transgene. We investigated growth of TgTP6.4 Tg/− hemizygotes and the viability of TgTP6.4 Tg/Tg homozygotes and we characterised expression of the TgTP6.4 transgene in embryos and adults by confocal microscopy of tauGFP fluorescence. This revealed that TgTP6.4 Tg/− mice showed widespread mosaic expression. As mosaicism can be useful for analysis of clonal lineages that are established during development or by adult stem cells, we also compared mosaic expression in the adrenal cortices of TgTP6.3 Tg/− and TgTP6.4 Tg/− hemizygotes.
C57BL/6 and outbred CD-1 mice were purchased from Bantin and Kingman, Hull, UK and A/J/Ola/Hsd mice were purchased from Harlan Olac Ltd., Bicester, UK. BALB/c/Eumm, CBA/Ca, C57BL/OlaWs (a C57BL/OlaHsd sub-colony), (C57BL × CBA/Ca)F1 hybrids, TP6.3 and TP6.4 mice were bred and maintained under conventional conditions at the University of Edinburgh with a light cycle of 14 h light (05:00 h. - 19:00 h.) and 10 h dark or 12 h light (06:00 h. - 18:00 h.) and 12 h dark. Founder TP6.3 and TP6.4 mice were produced from chimaeras made with E14Tg2a ES cells that had been transfected with the CAG-tauGFP vector, pTP6 [1, 2]. Founder mice were on a largely outbred MF1 genetic background and the transgenes were bred onto other strains as required. Transgenic mice were maintained by crossing TgTP6.3 Tg/− and TgTP6.4 Tg/− hemizygotes to non-transgenic, wild-type (WT) mice. Both TP6.3 and TP6.4 mice were initially crossed to inbred C57BL/6 mice to produce the first colony of each transgenic line, and some mice were also crossed to outbred, albino CD-1 mice to facilitate confocal microscopy of tauGFP expression in eye tissues. These colony I mice were used for the initial investigations shown in Fig. 3 and Additional file 1: Table S1. For later studies, TgTP6.3 Tg/− and TgTP6.4 Tg/− mice from the first colonies were crossed to (C57BL × CBA/Ca)F1 hybrid mice to produce a second colony for each line. These colony II mice were maintained by crossing transgenic mice to (C57BL × CBA/Ca)F1 mice at each generation and were used for the investigations shown in Figs. 1, 2, 4, 5, Additional file 1: Figures S2, S3 and S4 and Table 1. In all genetic crosses the female parental genotype is shown first. TauGFP-positive offspring were identified at weaning by the green fluorescence of their ear punch biopsies, using a fluorescence microscope with a fluorescein isothiocyanate (FITC) filter set. Some tauGFP-positive newborn pups were identified as described elsewhere .
Analysis of postnatal growth
Newborn pups from crosses between (C57BL × CBA/Ca)F1 females and hemizygous TgTP6.4 Tg/− males were sexed and weighed on postnatal days (P) 1 and 7 and then at weekly intervals thereafter until P84. Individual pups were marked with a marker pen until they were old enough for an ear punch biopsy, which was used both for identification and classification of their tauGFP phenotype.
Analysis of preimplantation embryos
For collection of preimplantation embryos, females were superovulated by intraperitoneal (i.p.) injections of 5 international units (I.U.) of pregnant mare’s serum gonadotrophin (PMSG; Folligon, Intervet) at approximately 12 noon followed 48 h later by 5 I.U. of human chorionic gonadotrophin (hCG; Chorulon, Intervet). After hCG injections, each female was caged with a male overnight and mating was verified by the presence of a vaginal copulation plug the following morning, which was designated embryonic day (E) 0.5. Preimplantation embryos were flushed from the reproductive tract with KSOM-H handling medium  and their ages in hours post coitum (p.c.) were timed from the mid-point of the dark period (midnight) following the hCG injection. For collection of 1-cell stage embryos, mice were killed at approximately 10:00 h. on the morning after the hCG injection, cumulus masses were released from the oviducts and cumulus cells were dispersed in a solution of 100 I.U. hyaluronidase (Sigma-Aldrich, Gillingham, UK) per ml of phosphate-buffered saline (PBS). Two cell-stage embryos were flushed from the oviducts at approximately 10:00 h. on the following day (34 h.).
For time-lapse monitoring, preimplantation embryos were cultured for 24 h in drops of KSOM culture medium under mineral oil in thin glass-bottomed dishes (WillCo HBSt 3522, Intracel Ltd. Royston, UK). The dishes were placed in an environmental chamber on top of the heated stage (THD 60, Linkam Scientific Instruments Ltd., Tadworth, UK) of an inverted Leica DMIRB/E confocal microscope, and the atmosphere within the chamber was maintained at 37 °C, 5% CO2 in air (Additional file 1: Fig. S1). Images were acquired every 15 or 30 min for approximately 24 h. in both fluorescent (FITC) and transmitted light mode, using the Leica TCS NT confocal system. The time-lapse files contained both FITC and transmitted light channels from each time point. These were merged to provide photographs showing GFP fluorescence overlaid on a transmitted light image for the figures. Each image from the FITC channel is a pseudocoloured greyscale representation of the intensity of the emission signal from the sample, rendered on a 0–255 pixel intensity scale, which represents the strength of fluorescence. For quantitative analysis, files for each time point were opened in the freeware image analysis program, Scion Image (http://downloads.informer.com/scion-image/), each embryo was outlined as a region of interest (ROI) and an average pixel intensity reading obtained. If embryos moved between frames (time points), the ROI was either moved or redrawn for analysis of the next frame. Embryos that were cultured from the 2-cell stage but failed to cleave were excluded from the analysis. The embryos sometimes drifted out of focus so, typically, a short (1–2 h.) culture period was recorded at the beginning of the experiment and examined, so the focus could be adjusted if required. The focus was also checked prior to overnight recording. These interruptions meant that each time-lapse experiment comprised several separate files, which sometimes differed slightly in pixel intensity because of fluctuations in laser power that occurred while scanning was interrupted.
Analysis of transgenic fetuses
Fetuses were produced by natural matings and staged from the date of the copulation plug, as described above. Females were killed by cervical dislocation, and the numbers of implantation sites, moles (resorbing conceptuses), dead fetuses and live fetuses were noted. Conceptuses were removed from the uterus and the fetuses dissected free of their placentas and extraembryonic membranes with fine forceps. Fetuses were classified for tauGFP expression with a Leica M2FLIII, fluorescence dissecting microscope using a GFP filter set. For comparison of fetal sizes, E14.5 fetuses were blotted dry, weighed and their crown-rump lengths were measured.
Analysis of tauGFP expression in tissue sections from fetal and adult transgenic mice
Hemizygous adults and E13.5 and E17.5 fetuses were produced as described above. Fetuses were decapitated and fixed in ice cold 4% paraformaldehyde (PFA) in phosphate buffered saline (PBS) overnight at 4 °C. Some small tissue samples from adults were also fixed by this immersion method. Other adult samples were perfusion fixed, while the mouse was deeply anaesthetised with a 0.4 ml i.p. injection of 25% urethane in saline. The vascular system was then flushed with ice-cold 0.9% saline via cardiac puncture, followed by perfusion with ice-cold 4% PFA in PBS. After perfusion, the mouse was killed and various organs, including brain, heart, liver, kidney and eyes, were removed and fixed overnight in fresh ice-cold 4% PFA in PBS.
Fixed samples were washed in two changes of PBS and embedded in 4% agarose (LMP Agarose, Bethesda Research Laboratories, Life Tech inc., USA). Blocks were trimmed, mounted on glass with adhesive and 200 μm sections were cut with a vibratome. Prior to mounting, some vibratome tissue sections were counterstained with propidium iodide (PI) solution (0.1% PI, 0.05% RNase, 0.1% Triton X-100 in PBS) for 5 min then rinsed in PBS, to label nuclear DNA fluorescent red. In some cases RNase was omitted, so cytoplasmic RNA was also counterstained. Sections were washed in PBS, immersed in a mixture of glycerol and PBS (1:1 v/v) at 4 °C until saturated and mounted on glass slides in a 1:1 mixture of glycerol and PBS containing 1% Vectashield antifade reagent (H-100, Vector Labs Inc., USA). The coverslips were sealed with Aquamount. Sections were viewed with an upright Leica TCS NT confocal microscope (Leica Microsystems, Germany). Bright-field images were collected in the transmitted light channel and GFP was detected in the FITC channel. For sections that were counterstained with propidium iodide, optical sections were acquired simultaneously in the FITC (green) and tetramethylrhodamine isothiocyanate (TRITC; red) channels. Single optical sections and stacks of sections were acquired with a minimum of 4 averages per optical section.
Analysis of tauGFP expression by fluorescence activated cell sorting
TauGFP-positive and negative E14.5 fetuses, from crosses to outbred CD-1 strain mice, were dissected in ice-cold Earle’s Balanced Salt Solution (EBSS), pre-equilibrated in 5% CO2. Whole brains, hippocampus, cerebral cortex, ventral telencephalon, central and dorsal thalamus, and midbrain and hindbrain tissues were dissected and dissociated using the Worthington Papain Dissociation System, (Worthington Biochemical Corporation, USA). Dissociated cells were sorted for green fluorescence using a Fluorescence Activated Cell Sorter (FACS; Becton Dickinson, Rutherford, NJ, USA) to generate histograms of green fluorescence intensity (FL1 channel) versus cell number for 10,000 cells per sample. Gates identifying cell populations with different fluorescent intensities were defined using a tauGFP-positive TP6.3 brain sample, considered to contain 100% fluorescent cells and the same gates were applied to all samples. After gating to exclude outliers (predominantly dead cells), the percentage of cells in three gated regions, corresponding to different fluorescent intensities, were compared: region M1 included all fluorescent cells, M2 included cells with low fluorescence and M3 included cells with high fluorescence.
Analysis of mosaicism in adrenal glands
In TgTP6.4 Tg/− and many TgTP6.3 Tg/− hemizygous mice, the adrenal cortex was not uniformly tauGFP-positive but, in sections, showed a radial pattern of tauGFP-positive and tauGFP-negative stripes, most of which extended continuously across the full width of the adrenal cortex. This 2-dimensional radial stripe pattern in sections was analysed as a 1-dimensional pattern of stripe widths by measuring around the adrenocortical circumference as described previously . The width of each tauGFP-negative and tauGFP-positive stripe was measured in calibrated images of single sections near the middle of the adrenal gland using UTHSCSA Image Tool software for Microsoft Windows (http://compdent.uthscsa.edu/dig/itdesc.html). Measurements were made around the complete circumference of the adrenal cortex at a similar distance from the outer edge of the adrenal gland (equivalent to 20% of the distance from the capsule to the inner margin of cortex). This provided the numbers of tauGFP-negative and tauGFP-positive stripes, the observed mean stripe widths, the total measured circumference and the proportions of tauGFP-negative and tauGFP-positive cells at the depth of the measurements for each adrenal cortex analysed. To allow for differences in stripe widths at different radial positions, the stripe width was expressed as a proportion of the circumference.
The radial stripes are believed to represent coherent clones of cells derived from stem cells at the periphery of the adrenal cortex [15, 16]. However, it is misleading to compare observed stripe numbers directly because each stripe may comprise several adjacent coherent clones and the number of clones per stripe is likely to vary according to the proportions of tauGFP-positive and tauGFP-negative cells in the tissue. We, therefore, converted the observed number of tauGFP-positive stripes plus tauGFP-negative stripes to a ‘corrected stripe number’ for each section, which adjusts for the expected number of clones per stripe . First, the observed mean tauGFP-negative stripe width was divided by the correction factor 1/(1-p), where p is the proportion of tauGFP-negative cells around the circumference [15, 17, 18]. As the radial stripes in sections of the adrenal cortex form a complete circle, the calculated ‘corrected mean stripe width’ will be the same for both the positive and negative stripes, as explained previously [15, 17, 18]. Thus, correction of mean tauGFP-positive stripe widths, using the proportion of tauGFP-positive cells as p, produced identical results. As the ‘corrected mean stripe width’ applies to both tauGFP-negative and tauGFP-positive stripes, the ‘corrected stripe number’ for the positive plus negative stripes is the reciprocal of the corrected mean stripe width, expressed as the proportion of the measured circumference. The calculated corrected stripe number adjusts for differences in proportions of tauGFP-positive cells among adrenals and is likely to be proportional to the number of active coherent clones of stem cells, so was used to compare different groups of adrenal glands as described elsewhere .
Minimum group sizes were guided by previous experience and power calculations. For comparison of postnatal growth, TP6.4 group sizes were chosen to ensure sufficient power to detect, as significant (P < 0.05), mean body mass differences that were smaller than those previously published for TP6.3 [1, 2]. For quantitative analysis of adrenal mosaicism, results for only one adrenal gland per mouse were included. For mice where both adrenal glands were imaged, the choice of including the left or right adrenal was randomised by tossing a coin. The choice of parametric or non-parametric tests was guided, in part, by normality tests. GraphPad Prism versions 5.0c and 7 (GraphPad Software Inc., La Jolla, CA) were used for most statistical tests including 2-way analysis of variance (ANOVA) followed by Bonferroni multiple comparison post-tests, repeated measures 2-way ANOVA followed by Bonferroni multiple comparison post-tests, 1-way ANOVA, the paired t-test, the Mann–Whitney U-test and the Kruskal-Wallis test followed by Dunn’s multiple comparison post-tests. An online statistical calculator (http://vassarstats.net/index.html) was used for the chi-square goodness of fit test and Fisher’s exact test. Error bars shown are 95% confidence intervals (CI).
Postnatal growth and viability of hemizygous TgTP6.4 Tg/− transgenic mice
Viability of TgTP6.4 Tg/Tg homozygotes
We were unable to identify any adult TgTP6.4 Tg/Tg homozygotes by overt differences in GFP fluorescence and, as the TgTP6.4 insertion site is unknown, we had no simple PCR method for distinguishing TgTP6.4 Tg/Tg homozygotes from TgTP6.4 Tg/− hemizygotes. In the absence of an easy method for identifying homozygotes, we used a statistical analysis of offspring phenotype frequencies from a genetic test cross to determine whether homozygotes were viable. Thirty-nine tauGFP-positive progeny, from TgTP6.4 Tg/− × TgTP6.4 Tg/− crosses, were crossed to non-transgenic mice to determine whether any only produced tauGFP-positive offspring and so were likely to be homozygotes. Thirty-six mice were classified as hemizygotes because they produced both tauGFP-positive and tauGFP-negative offspring in either their first viable litter (35 cases) or their second litter (one case). One other mouse produced three litters with 9/9, 7/8 and 9/12 tauGFP-positive offspring, respectively. As the overall frequency (25/29) was significantly higher than the expected 50% tauGFP-positive offspring (P = 0.0002), this mouse might have been a germline homozygous/hemizygous mosaic, rather than a hemizygote but it died before further investigations could be performed. Five of the other 36 mice produced equal numbers of tauGFP-positive and tauGFP-negative offspring, 16 produced more tauGFP-positive and 15 produced more tauGFP-negative offspring. As the 16:15 ratio did not differ significantly from the 1:1 ratio expected if all 36 mice were TgTP6.4 Tg/− hemizygotes (P = 1.0000), there was no evidence that any of these were homozygous/hemizygous mosaics. This is also supported by the overall frequency of tauGFP-positive offspring from these 36 mice (177/345; 51.3%), which was not significantly different from the expected 50% (P = 0.6714) for hemizygotes. The remaining two males failed to sire any offspring in three-months and we cannot exclude the possibility that these were infertile homozygotes. One third of the tauGFP-positive offspring from TgTP6.4 Tg/− × TgTP6.4 Tg/− crosses are expected to be homozygous but none of 37 fertile offspring bred as homozygotes. The probability of this occurring by chance is only (2/3)37 (P = 0.0000003), if homozygotes are actually viable and fertile. These results imply that homozygous TgTP6.4 Tg/Tg mice do not survive to reproduce.
Viability of TgTP6.4 hemizygotes and homozygotes assessed by frequencies of tauGFP-positive mice produced in different crosses
Genetic cross (female × male)
Tg/− × Tg/−
Tg/− × −/−
−/− × Tg/−
A. Mice classified for tauGFP at ~3 weeks
Dead before weaning
Dead after weaning (most unclassified)
Viable at weaning
TauGFP not classified
Pre-weaning death attributable to crossj
B. Fetuses classified for tauGFP at E14.5
Moles (early deaths)
Total live fetuses
Death attributable to crossk
To investigate whether homozygous, TgTP6.4 Tg/Tg E14.5 fetuses were viable, we compared observed and expected frequencies of tauGFP-positive fetuses in two crosses (Table 1). The control and experimental crosses produced close to 50 and 75% tauGFP positive E14.5 fetuses respectively, as expected if all hemizygotes and homozygotes survived to E14.5. Survival of most homozygotes to E14.5 is also suggested by the relatively low frequency of dead embryos, which was not significantly elevated in the experimental cross. However, there was some evidence that homozygous tauGFP fetuses were slightly smaller than normal by E14.5. After allowing for variation among litters, tauGFP-positive fetuses in the experimental cross were significantly lighter than tauGFP-negative fetuses (P < 0.0001 by a 2-way ANOVA) and had shorter crown-rump lengths (P < 0.0001; Additional file 1: Figure S2). In the control cross, fetal mass differences were less pronounced (P = 0.0343) and crown-rump length differences were not significant (P = 0.0597), implying that hemizygous TgTP6.4 Tg/− fetuses were more normal in size at E14.5. In summary, the breeding experiments indicate that most TgTP6.4 Tg/Tg homozygotes survive to E14.5 but some may already be retarded and they probably all die between E14.5 and weaning.
TauGFP expression in TgTP6.4 Tg/− preimplantation embryos
Preimplantation embryos from the reciprocal cross between TgTP6.4 Tg/− females and (C57BL × CBA/Ca)F1 males, initially all retained oocyte-encoded tauGFP but TgTP6.4 Tg/− and TgTP6.4 −/− embryos could be distinguished by their tauGFP levels by the 4-cell stage (Fig. 2d-f). There was no overlap in pixel intensity from 45 h. p.c. for the experiment shown in Fig. 2h and a repeated measures 2-way ANOVA with Bonferroni multiple comparison post-tests showed that pixel intensity differed significantly between the two genotypes from 42.5 h., consistent with the onset of embryo-coded TgTP6.4 expression by the late 2-cell stage. Comparisons of pixel intensity between TgTP6.4 −/− embryos, with some residual tauGFP, and control WT embryos, with no tauGFP, showed no overlap in pixel intensity at any ages analysed (34.6 to 57.6 h.) and a repeated measures 2-way ANOVA with Bonferroni multiple comparison post-tests confirmed that pixel intensities remained significantly different at 57.6 h. This indicates that residual oocyte-encoded tauGFP remained at least until the 4-cell stage.
TauGFP expression in fetal and adult TgTP6.4 Tg/− tissues
Fluorescence activated cell sorting (FACS) analysis of cell suspensions prepared from whole E14.5 fetal brains confirmed that tauGFP fluorescence was present in fewer TgTP6.4 Tg/− cells than TgTP6.3 Tg/− cells (Fig. 3q-t). The analysis using three gated regions (Fig. 3q) is described in the Methods. Fig 3r shows that 95% of all the cells from whole TgTP6.3 Tg/− brains were fluorescent (M1 gate in Fig. 3q), of which 87% were highly fluorescent (Fig. 3r). In contrast, only 25% of cells from whole TgTP6.4 Tg/− brains were fluorescent, 10% of which were highly fluorescent (Fig. 3s). Nevertheless, this profile shows more fluorescence than the TgTP6.4 −/− negative controls (Fig. 3t). Similar results were obtained for cells from separate brain regions although the hippocampus showed more fluorescence than other TgTP6.4 Tg/− brain regions (Additional file 1: Table S1).
Comparison of mosaic tauGFP expression in TgTP6.3 Tg/− and TgTP6.4 Tg/− adrenal cortices
The pattern of radial stripes seen in the adrenal cortex of some TgTP6.3 Tg/− and TgTP6.4 Tg/− mice (Fig. 4m-p,t) has been reported for several other experimental systems and is believed to represent coherent clones of cells, which are derived from stem cells at the periphery of the adrenal cortex, and move towards the medulla (see Discussion). After correcting for the expected number of multiple adjacent clones in a single stripe, the corrected stripe number can provide an indirect estimate of the number of clones of active stem cells as described elsewhere [15, 17, 18] so this can be used to compare stem cell function in different groups. The uncorrected mean stripe width varies with the percentage of tauGFP-positive cells in the adrenal cortex, which will affect the number of clones per stripe, but the corrected mean stripe width adjusts for this, as shown in Additional file 1: Figure S3. The corrected mean stripe width was used to calculate a corrected stripe number, as explained in the Methods, to provide a quantitative comparison of TgTP6.3 Tg/− and TgTP6.4 Tg/− stripe patterns.
The aim of the present study was to determine whether the TgTP6.4 marker transgene offered any advantages over TgTP6.3. Although tauGFP expression in TgTP6.3 Tg/− cells provides an excellent marker, TgTP6.3 Tg/Tg homozygotes do not survive and TgTP6.3 Tg/− hemizygotes are slightly smaller than normal. TgTP6.4 Tg/− growth was less affected than TgTP6.3 Tg/− hemizygotes but TgTP6.4 Tg/Tg homozygotes died between E14.5 and weaning age, as reported for TgTP6.3 Tg/Tg homozygotes  and the widespread mosaicism means that TgTP6.4 is unsuitable as a conventional cell lineage marker. The reason why both TgTP6.3 Tg/Tg and TgTP6.4 Tg/Tg homozygotes die was not investigated but high levels of GFP or tau protein are neurotoxic . Other tauGFP transgenic mice have been produced by inserting a tauGFP transgene into a specific locus [20–22] or randomly inserting a CAG-tauGFP construct into the genome to produce another CAG-tauGFP transgenic line , but no mosaic phenotypes were described for these mice. While mosaicism in TgTP6.3 Tg/− mice was only identified in the cornea and adrenal cortex, mosaicism became increasingly common in TgTP6.4 Tg/− mice during our investigation and eventually affected all tissues examined and all TgTP6.4 Tg/− mice.
Mosaic patterns produced by mouse chimaeras and X-inactivation mosaics have been useful for visualising clonal lineages and for analysing the extent of cell mixing and movement during development, growth and maintenance of adult tissues [24–29]. Mosaic transgene expression can provide a useful alternative for these types of investigations , particularly when mosaicism is widespread, as in TgTP6.4 Tg/− mice. We used this approach to investigate whether there were any overt differences in maintenance of the adrenal cortex in TgTP6.3 Tg/− and TgTP6.4 Tg/− adults.
Mosaic TgTP6.3 Tg/− and TgTP6.4 Tg/− adrenals showed qualitatively similar radial stripes across the adrenal cortex. This pattern has previously been reported for other experimental systems, including mouse and rat chimaeras [24, 28, 30–33], mouse X-inactivation mosaics  and some other mouse lines that show mosaic transgene expression [15, 34–36]. The stripes are believed to represent coherent clones of cells that are derived from stem cells at the periphery of the adrenal cortex and they maintain the tissue as they move slowly across the cortex before undergoing apoptosis [15, 16, 37–40]. The corrected stripe numbers were similar in the two genotypes and were comparable to results reported previously for mouse chimaeras, X-inactivation mosaics and other mosaic transgenic mice [15, 33], suggesting that maintenance of the adrenal cortex involved a similar number of active clones of stem cells in all these groups.
Mosaic transgene expression is thought to involve position effect variegation caused by stochastic transgene silencing during development [34, 41–43]. Indirect evidence suggested that mosaic expression of a different transgene in the mouse adrenal cortex involved stochastic silencing at the level of the chromosome or transgene, rather than at the cellular level  but this may not apply to all mosaic transgenic lines. Transgene silencing can occur if the transgene integrates near an endogenous heterochromatic site , which may prevent the transcriptional machinery from accessing the transgene, or if it triggers the formation of heterochromatin by integrating as a series of repeated copies [45, 46]. Repeat-copy transgenes are also associated with methylation changes that could mediate stochastic transgene silencing if transgene methylation levels vary among cells . Moreover, several genes that alter the extent of transgene silencing have been identified in a mutagenesis screen and some have been characterised [48, 49]. However, the copy number is not known for either of the transgenic lines that we studied and the mechanism(s) that caused widespread mosaic silencing of TgTP6.4 and more restricted silencing of TgTP6.3 is unknown.
TauGFP expression in TgTP6.3 Tg/− preimplantation embryos was characterised previously [1, 2] and was similar in TgTP6.4 Tg/− embryos. TauGFP fluorescence was ubiquitous in preimplantation embryos of both transgenic lines, so mosaic expression presumably involves transgene silencing in some cells during development. As there was no evidence for an age-effect on the percentage of cells expressing tauGFP in either TgTP6.3 Tg/− or TgTP6.4 Tg/− adrenals, we have no evidence that silencing is a continuous process. The simplest model assumes that transgene silencing involves a single stochastic event that occurs during development and the silent or active state is then stably inherited by daughter cells. As the frequency of TgTP6.4 Tg/− mosaic expression changed over a number of generations of breeding and ultimately affected all tissues examined and the entire TP6.4 colony, it seems likely that the genetic background somehow affects the probability of transgene silencing but, again, the mechanism remains unknown.
In contrast to TgTP6.4 Tg/− mice, mosaic expression did not occur in all TgTP6.3 Tg/− mice, nor did it occur in so many tissues. Of the TgTP6.3 Tg/− tissues examined, mosaicism was only identified in the adrenal cortex and cornea. Moreover, when mosaicism occurred in the adrenal cortex of TgTP6.3 Tg/− mice, there were significantly more tauGFP-positive cells than in TgTP6.4 Tg/− adrenals. This implies that silencing of the TgTP6.3 transgene affects fewer cells per tissue as well as fewer tissues and fewer mice. In both TgTP6.3 Tg/− and TgTP6.4 Tg/− mice, the mosaic pattern in the cornea was in the form of radial stripes, strongly suggesting it was in the corneal epithelium layer, as described for studies of the corneal epithelium using other mosaic systems [17, 18, 50–57].
Mosaicism in TgTP6.3 Tg/− mice is clearly more restricted than in TgTP6.4 Tg/− hemizygotes so it might arise later in development but this was not tested directly. Investigation of why the extent of mosaicism differs between these two transgenic lines would probably require breeding both transgenes onto the same inbred genetic background, as well as knowledge of the insertion sites, copy numbers and epigenetic modifications around the insertion sites. Stem cells at the tissue’s periphery are thought to maintain both the adult corneal epithelium [55–60] and the adrenal cortex [15, 16, 38–40]. It may also, therefore, be worth investigating whether transgene silencing can occur in tissue-specific stem cells as well as earlier in development.
Although widespread mosaic expression means that the TgTP6.4 transgene is not a suitable alternative to TgTP6.3 as a standard lineage marker, it provides a useful system for identifying and analysing clonal lineages. We exploited this mosaicism to demonstrate that similar clonal lineages maintained the adult adrenal cortex in different groups of mice. Differences, in the frequency and tissue distribution of mosaicism, between related transgenic lines, such as TP6.3 and TP6.4, may also provide a useful experimental system for investigating transgene silencing.
We thank William Seldon for technical help during his undergraduate summer project, Andrew Sanderson for assistance with the FACS analysis, Katy Gillies, Denis Doogan, Maureen Ross and Jim Macdonald for other technical help and Steven Morley for helpful discussion.
This work was supported by the Wellcome Trust (research grant 059904/JS to JDW JOM and DJP and equipment grant 052578/Z/97/Z to JBL Bard, JA Davies, DJP, N Spears and JDW), the University of Edinburgh Moray Endowment Fund and PhD studentship funding for GEM from the College of Medicine and Veterinary Medicine, University of Edinburgh. The funding bodies played no role in the design of the study, in the collection, analysis and interpretation of data or in writing the manuscript.
Availability of data and materials
Raw numerical data have been deposited in Edinburgh DataShare, the University of Edinburgh’s data repository (https://doi.org/10.7488/ds/2049). TP6.3 and TP6.4 embryos have been frozen by the College of Medicine and Veterinary Medicine Cryopreservation Service, University of Edinburgh. For availability, email email@example.com or JOM
TP produced the transgenic mice and participated in lab work, LS & GEM carried out lab work, participated in data analysis, produced some figures and helped draft the manuscript, MAK, JHF and EJC carried out lab work, DJP and JOM designed and supervised parts of the study, JDW designed and supervised part of the study, analysed the data, produced some figures and drafted the manuscript. LS and TP contributed equally to this work. All authors contributed to the preparation of the final manuscript and gave final approval for publication.
The authors declare that they have no competing interests.
Consent for publication
Animal work was performed in accordance with institutional guidelines and UK regulations and was authorised by the UK Home Office (licences PPL 60/1989, 60/2374 and 60/2887).
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Pratt T, Sharp L, Nichols T, Price DJ, Mason JO. Embryonic stem cells and transgenic mice ubiquitously expressing a tau-tagged green fluorescent protein. Dev Biol. 2000;228(1):19–28.View ArticlePubMedGoogle Scholar
- MacKay GE, Keighren MA, Wilson L, Pratt T, Flockhart JH, Mason JO, et al. Evaluation of the mouse TgTP6.3 tauGFP transgene as a lineage marker in chimeras. J Anat. 2005;206(1):79–92.View ArticlePubMedPubMed CentralGoogle Scholar
- Ying QL, Nichols J, Evans EP, Smith AG. Changing potency by spontaneous fusion. Nature. 2002;416(6880):545–8.View ArticlePubMedGoogle Scholar
- Thomas JA, Pope C, Wojtacha D, Robson AJ, Gordon-Walker TT, Hartland S, et al. Macrophage therapy for murine liver fibrosis recruits host effector cells improving fibrosis, regeneration, and function. Hepatology. 2011;53(6):2003–15.View ArticlePubMedGoogle Scholar
- Campbell L, Trendell J, Spears N. Identification of cells migrating from the thecal layer of ovarian follicles. Cell Tissue Res. 2013;353(1):189–94.View ArticlePubMedGoogle Scholar
- MacKay GE, West JD. Fate of tetraploid cells in 4n↔2n chimeric mouse blastocysts. Mech Dev. 2005;122(12):1266–81.View ArticlePubMedGoogle Scholar
- Edgar JM, Price DJ. Radial migration in the cerebral cortex is enhanced by signals from thalamus. Eur J Neurosci. 2001;13(9):1745–54.View ArticlePubMedGoogle Scholar
- Fink AJ, Englund C, Daza RAM, Pham D, Lau C, Nivison M, et al. Development of the deep cerebellar nuclei: transcription factors and cell migration from the rhombic lip. J Neurosci. 2006;26(11):3066–76.View ArticlePubMedGoogle Scholar
- Englund C, Kowalczyk T, Daza RAM, Dagan A, Lau C, Rose MF, et al. Unipolar brush cells of the cerebellum are produced in the rhombic lip and migrate through developing white matter. J Neurosci. 2006;26(36):9184–95.View ArticlePubMedGoogle Scholar
- Magnani D, Hasenpusch-Theil K, Jacobs EC, Campagnoni AT, Price DJ, Theil T. The Gli3 hypomorphic mutation Pdn causes selective impairment in the growth, patterning, and axon guidance capability of the lateral ganglionic eminence. J Neurosci. 2010;30(41):13883–94.View ArticlePubMedGoogle Scholar
- Magnani D, Hasenpusch-Theil K, Benadiba C, Yu T, Basson MA, Price DJ, et al. Gli3 controls corpus callosum formation by positioning midline guideposts during telencephalic patterning. Cereb Cortex. 2014;24(1):186–98.Google Scholar
- Clegg JM, Li Z, Molinek M, Caballero IM, Manuel MN, Price DJ. Pax6 is required intrinsically by thalamic progenitors for the normal molecular patterning of thalamic neurons but not the growth and guidance of their axons. Neural Dev. 2015;10:26.View ArticlePubMedPubMed CentralGoogle Scholar
- Tyas DA, Pratt T, Simpson TI, Mason JO, Price DJ. Identifying GFP-transgenic animals by flashlight. Biotechniques. 2003;34(3):474–6.PubMedGoogle Scholar
- Summers MC, Bhatnagar PR, Lawitts JA, Biggers JD. Fertilization in vitro of mouse ova from inbred and outbred strains: complete preimplantation embryo development in glucose-supplemented KSOM. Biol Reprod. 1995;53(2):431–7.View ArticlePubMedGoogle Scholar
- Chang S-P, Mullins JJ, Morley SD, West JD. Transition from organogenesis to stem cell maintenance in the mouse adrenal cortex. Organogenesis. 2011;7(4):267–80.View ArticlePubMedPubMed CentralGoogle Scholar
- Chang S-P, Morrison HD, Nilsson F, Kenyon CJ, West JD, Morley SD. Cell proliferation, movement and differentiation during maintenance of the adult mouse adrenal cortex. PLoS One. 2013;8(12):e81865.View ArticlePubMedPubMed CentralGoogle Scholar
- Collinson JM, Morris L, Reid AI, Ramaesh T, Keighren MA, Flockhart JH, et al. Clonal analysis of patterns of growth, stem cell activity, and cell movement during the development and maintenance of the murine corneal epithelium. Dev Dyn. 2002;224(4):432–40.Google Scholar
- Mort RL, Ramaesh T, Kleinjan DA, Morley SD, West JD. Mosaic analysis of stem cell function and wound healing in the mouse corneal epithelium. BMC Dev Biol. 2009;9:4.View ArticlePubMedPubMed CentralGoogle Scholar
- Klein RL, Dayton RD, Leidenheimer NJ, Jansen K, Golde TE, Zweig RM. Efficient neuronal gene transfer with AAV8 leads to neurotoxic levels of tau or green fluorescent proteins. Mol Ther. 2006;13(3):517–27.View ArticlePubMedGoogle Scholar
- Rodriguez I, Feinstein P, Mombaerts P. Variable patterns of axonal projections of sensory neurons in the mouse vomeronasal system. Cell. 1999;97(2):199–208.View ArticlePubMedGoogle Scholar
- Potter SM, Zheng C, Koos DS, Feinstein P, Fraser SE, Mombaerts P. Structure and emergence of specific olfactory glomeruli in the mouse. J Neurosci. 2001;21(24):9713–23.PubMedPubMed CentralGoogle Scholar
- Tyas DA, Simpson TI, Carr CB, Kleinjan DA, van Heyningen V, Mason JO, et al. Functional conservation of Pax6 regulatory elements in humans and mice demonstrated with a novel transgenic reporter mouse. BMC Dev Biol. 2006;6:21.View ArticlePubMedPubMed CentralGoogle Scholar
- Hadjantonakis AK, Pisano E, Papaioannou VE. Tbx6 regulates left/right patterning in mouse embryos through effects on nodal cilia and perinodal signaling. PLoS One. 2008;3(6):e2511.Google Scholar
- Iannaccone PM. The study of mammalian organogenesis by mosaic pattern analysis. Cell Differ. 1987;21(2):79–91.View ArticlePubMedGoogle Scholar
- Reese BE, Harvey AR, Tan SS. Radial and tangential dispersion patterns in the mouse retina are cell-class specific. Proc Natl Acad Sci U S A. 1995;92(7):2494–8.View ArticlePubMedPubMed CentralGoogle Scholar
- West JD. Insights into development and genetics from mouse chimeras. Curr Top Dev Biol. 1999;44:21–66.View ArticlePubMedGoogle Scholar
- Collinson JM, Hill RE, West JD. Analysis of mouse eye development with chimeras and mosaics. Int J Dev Biol. 2004;48(8–9):793–804.View ArticlePubMedGoogle Scholar
- Ma DF, Sudo K, Tezuka H, Kondo T, Nakazawa T, Niu DF, et al. Polyclonal origin of hormone-producing cell populations evaluated as a direct in situ demonstration in EGFP/BALB/C chimeric mice. J Endocrinol. 2010;207(1):17–25.Google Scholar
- Eckardt S, McLaughlin KJ, Willenbring H. Mouse chimeras as a system to investigate development, cell and tissue function, disease mechanisms and organ regeneration. Cell Cycle. 2011;10(13):2091–9.View ArticlePubMedPubMed CentralGoogle Scholar
- Weinberg WC, Howard JC, Iannaccone PM. Histological demonstration of mosaicism in a series of chimeric rats produced between congenic strains. Science. 1985;227(4686):524–7.View ArticlePubMedGoogle Scholar
- West JD. Genetic studies with mouse chimaeras. In: Reeve ECR, editor. Encyclopedia of genetics. London and Chicago: Fitzroy Dearborn; 2001. p. 293–302.Google Scholar
- Iannaccone P, Morley S, Skimina T, Mullins J, Landini G. Cord-like mosaic patches in the adrenal cortex are fractal: implications for growth and development. FASEB J. 2003;17(1):41–3.PubMedGoogle Scholar
- Morley SD, Chang SP, Tan SS, West JD. Validity of the 21-OH/LacZ transgenic mouse as a model for studying adrenocortical cell lineage. Endocr Res. 2004;30(4):513–9.View ArticlePubMedGoogle Scholar
- Morley SD, Viard I, Chung BC, Ikeda Y, Parker KL, Mullins JJ. Variegated expression of a mouse steroid 21-hydroxylase/β-galactosidase transgene suggests centripetal migration of adrenocortical cells. Mol Endocrinol. 1996;10(5):585–98.PubMedGoogle Scholar
- Hu MC, Chou SJ, Huang YY, Hsu NC, Li H, Chung BC. Tissue-specific, hormonal, and developmental regulation of SCC-LacZ expression in transgenic mice leads to adrenocortical zone characterization. Endocrinology. 1999;140(12):5609–18.View ArticlePubMedGoogle Scholar
- Ma DF, Tezuka H, Kondo T, Sudo K, Niu DF, Nakazawa T, et al. Differential tissue expression of enhanced green fluorescent protein in 'Green mice'. Histol Histopathol. 2010;25(6):749–54.PubMedGoogle Scholar
- Iannaccone PM, Weinberg WC. The histogenesis of the rat adrenal cortex - a study based on histologic analysis of mosaic pattern in chimeras. J Exp Zool. 1987;243(2):217–23.View ArticlePubMedGoogle Scholar
- Kim AC, Barlaskar FM, Heaton JH, Else T, Kelly VR, Krill KT, et al. In search of adrenocortical stem and progenitor cells. Endocr Rev. 2009;30(3):241–63.View ArticlePubMedPubMed CentralGoogle Scholar
- King P, Paul A, Laufer E. Shh signaling regulates adrenocortical development and identifies progenitors of steroidogenic lineages. Proc Natl Acad Sci U S A. 2009;106(50):21185–90.View ArticlePubMedPubMed CentralGoogle Scholar
- Wood MA, Acharya A, Finco I, Swonger JM, Elston MJ, Tallquist MD, et al. Fetal adrenal capsular cells serve as progenitor cells for steroidogenic and stromal adrenocortical cell lineages in M. musculus. Development. 2013;140(22):4522–32.Google Scholar
- Dobie KW, Lee M, Fantes JA, Graham E, Clark AJ, Springbett A, et al. Variegated transgene expression in mouse mammary-gland is determined by the transgene integration locus. Proc Natl Acad Sci U S A. 1996;93(13):6659–64.Google Scholar
- Alami R, Greally JM, Tanimoto K, Hwang S, Feng YQ, Engel JD, et al. β-globin YAC transgenes exhibit uniform expression levels but position effect variegation in mice. Hum Mol Genet. 2000;9(4):631–6.Google Scholar
- Opsahl ML, Springbett A, Lathe R, Colman A, McClenaghan M, Whitelaw CBA. Mono-allelic expression of variegating transgene locus in the mouse. Transgenic Res. 2003;12(6):661–9.View ArticlePubMedGoogle Scholar
- Chang S-P, Opsahl ML, Whitelaw CBA, Morley SD, West JD. Relative transgene expression frequencies in homozygous versus hemizygous transgenic mice. Transgenic Res. 2013;22(6):1143–54.View ArticlePubMedGoogle Scholar
- Dorer DR. Do transgene arrays form heterochromatin in vertebrates? Transgenic Res. 1997;6(1):3–10.View ArticlePubMedGoogle Scholar
- Garrick D, Fiering S, Martin DIK, Whitelaw E. Repeat-induced gene silencing in mammals. Nat Genet. 1998;18(1):56–9.View ArticlePubMedGoogle Scholar
- Rosser JM, An W. Repeat-induced gene silencing of L1 transgenes is correlated with differential promoter methylation. Gene. 2010;456(1–2):15–23.View ArticlePubMedPubMed CentralGoogle Scholar
- Blewitt ME, Vickaryous NK, Hemley SJ, Ashe A, Bruxner TJ, Preis JI, et al. An N-ethyl-N-nitrosourea screen for genes involved in variegation in the mouse. Proc Natl Acad Sci U S A. 2005;102(21):7629–34.Google Scholar
- Ashe A, Morgan DK, Whitelaw NC, Bruxner TJ, Vickaryous NK, Cox LL, et al. A genome-wide screen for modifiers of transgene variegation identifies genes with critical roles in development. Genome Biol. 2008;9(12):R182.View ArticlePubMedPubMed CentralGoogle Scholar
- Nagasaki T, Zhao J. Centripetal movement of corneal epithelial cells in the normal adult mouse. Invest Ophthalmol Vis Sci. 2003;44(2):558–66.View ArticlePubMedGoogle Scholar
- Endo M, Zoltick PW, Chung DC, Bennett J, Radu A, Muvarak N, et al. Gene transfer to ocular stem cells by early gestational intraamniotic injection of lentiviral vector. Mol Ther. 2007;15(3):579–87.View ArticlePubMedGoogle Scholar
- Zhang W, Zhao J, Chen L, Urbanowicz MM, Nagasaki T. Abnormal epithelial homeostasis in the cornea of mice with a destrin deletion. Mol Vis. 2008;14:1929–39.PubMedPubMed CentralGoogle Scholar
- Hayashi Y, Watanabe N, Ohashi Y. The “replacement hypothesis”: corneal stem cell origin epithelia are replaced by limbal stem cell origin epithelia in mouse cornea during maturation. Cornea. 2012;31(Suppl 1(11)):S68–73.View ArticlePubMedGoogle Scholar
- Iannaccone S, Zhou Y, Walterhouse D, Taborn G, Landini G, Iannaccone P. Three dimensional visualization and fractal analysis of mosaic patches in rat chimeras: cell assortment in liver, adrenal cortex and cornea. PLoS One. 2012;7(2):e31609.View ArticlePubMedPubMed CentralGoogle Scholar
- Di Girolamo N, Bobba S, Raviraj V, Delic NC, Slapetova I, Nicovich PR, et al. Tracing the fate of limbal epithelial progenitor cells in the murine cornea. Stem Cells. 2015;33(1):157–69.View ArticlePubMedGoogle Scholar
- Amitai-Lange A, Altshuler A, Bubley J, Dbayat N, Tiosano B, Shalom-Feuerstein R. Lineage tracing of stem and progenitor cells of the murine corneal epithelium. Stem Cells. 2015;33(1):230–9.View ArticlePubMedGoogle Scholar
- Dorà NJ, Hill RE, Collinson JM, West JD. Lineage tracing in the adult mouse corneal epithelium supports the limbal epithelial stem cell hypothesis with intermittent periods of stem cell quiescence. Stem Cell Res. 2015;15(3):665–77.View ArticlePubMedPubMed CentralGoogle Scholar
- Davanger M, Evensen A. Role of the pericorneal papillary structure in renewal of corneal epithelium. Nature. 1971;229:560–1.View ArticlePubMedGoogle Scholar
- Schermer A, Galvin S, Sun TT. Differentiation-related expression of a major 64K corneal keratin in vivo and in culture suggests limbal location of corneal epithelial stem cells. J Cell Biol. 1986;103(1):49–62.View ArticlePubMedGoogle Scholar
- West JD, Dorà NJ, Collinson JM. Evaluating alternative stem cell hypotheses for adult corneal epithelial maintenance. World J Stem Cells. 2015;7(2):281–99.View ArticlePubMedPubMed CentralGoogle Scholar
- Lyon MF. X-ray induced dominant lethal mutations in male guinea-pigs, hamsters and rabbits. Mutat Res. 1970;10:133–40.View ArticlePubMedGoogle Scholar