Transgenic Mice With Pancellular Enhanced Green
Fluorescent Protein Expression in Primitive Hematopoietic
Cells and All Blood Cell Progeny
Massimo Dominici,1Merhdad Tadjali,1Steven Kepes,1Esther R. Allay,1
Kelli Boyd,2Paul A. Ney,3Edwin Horwitz,1and Derek A. Persons1*
1Division of Experimental Hematology, Department of Hematology and Oncology,
St. Jude Children’s Research Hospital, Memphis, Tennessee
2Animal Resource Center, St. Jude Children’s Research Hospital, Memphis, Tennessee
3Department of Biochemistry, St. Jude Children’s Research Hospital, Memphis, Tennessee
Received 21 October 2004; Accepted 15 February 2005
Summary: Transgenic mice homogeneously expressing
enhanced green fluorescence protein (EGFP) in primitive
hematopoietic cells and all blood cell progeny, including
erythrocytes and platelets, have not been reported.
Given previous data indicating H2Kbpromoter activity in
murine hematopoietic stem cells (HSCs), bone marrow
(BM), and lymphocytes, an H2Kbenhancer/promoter
EGFP construct was used to generate transgenic mice.
These mice demonstrated pancellular EGFP expression
in both primitive BM Sca-1þLin?Kitþcells and side pop-
ulation (SP) cells. Additionally, all peripheral blood leu-
kocytes subsets, erythrocytes, and platelets uniformly
expressed EGFP strongly. Competitive BM transplanta-
tion assays established that transgenic H2Kb-EGFP
HSCs had activity equivalent to wildtype HSCs in their
ability to reconstitute hematopoiesis in lethally irradi-
ated mice. In addition, immunohistochemistry revealed
EGFP transgene expression in all tissues examined. This
transgenic strain should be a useful reagent for both
murine hematopoiesis studies and functional studies of
specific cell types from particular tissues. genesis
C 2005 Wiley-Liss, Inc.
Key words: EGFP; transgenic; hematopoiesis
The ability to easily track the fate of transplanted hema-
topoietic stem cells (HSCs) and all their progeny would
facilitate characterization of the hematopoietic pheno-
types of murine gene knockout models. Expression of a
readily detectable marker, such as the enhanced green
fluorescent protein (EGFP), in both HSCs and all blood
cell progeny of a wildtype reference mouse strain would
allow both the HSC number and the lineage differentia-
tion capacity of gene knockout cells to be easily assessed
in competitive bone marrow (BM) transplantation
assays. Previously, we reported that retroviral vectors
encoding EGFP can be used to mark HSCs, allowing
identification and quantitation of both hematopoietic
and nonhematopoietic progeny (Persons et al., 1998;
Priller et al., 2001a,b). However, the ex vivo cell manipu-
lations involved in retroviral gene marking procedures
could affect HSC proliferation and differentiation proper-
ties, thereby possibly influencing posttransplantation
fate. Thus, a transgenic mouse strain with pancellular
EGFP expression in all hematopoietic cells is highly
Although several transgenic mouse lines have been
generated that utilize a chicken b-actin promoter/CMV
immediately early enhancer to express EGFP in hemato-
poietic and nonhematopoietic tissues (Okabe et al.,
1997; Wagers et al., 2002; Wright et al., 2001; Manfra
et al., 1999; Hadjantonakis et al., 1998; Orlic et al.,
2001), detailed analyses of EGFP expression and its pos-
sible effects in both primitive and differentiated hemato-
poietic compartments have not been fully characterized.
In particular, competitive repopulation studies to docu-
ment normal repopulating activity of HSCs expressing
EGFP have not been reported.
Additionally, some of these EGFP strains have been
demonstrated to display position effect variegation of
expression of the EGFP transgene in hematopoietic cells,
resulting in significant proportions of nonexpressing
primitive and mature blood cells (Hadjantonakis et al.,
1998; Orlic et al., 2001). Furthermore, these EGFP
mouse strains, as well as the ROSA26 mouse, which
widely expresses b-galactosidase in hematopoietic cells
and tissues, have not been reported to express the trans-
gene in mature red blood cells (RBCs) or in blood plate-
lets (PLTs) (Zambrowicz et al., 1997). In order to gener-
*Correspondence to: Derek A. Persons, M.D., Ph.D., 332 North Lauder-
dale Dr., St. Jude Children’s Research Hospital, Memphis, TN 38105.
Contract grant sponsors: American Lebanese Syrian Associated Charities
(ALSAC), NIH funded Cancer Center Support (Core) grant.
Published online in
Wiley InterScience (www.interscience.wiley.com).
' 2005 Wiley-Liss, Inc. genesis 42:17–22 (2005)
ate a transgenic EGFP mouse strain with improved
expression in all hematopoietic cells that could be used
in competitive repopulation and HSC fate studies, we
utilized a transgenic construct containing the murine
MHC class I H2Kbenhancer/promoter (Domen et al.,
1998). These transcriptional control elements were pre-
viously shown to direct expression of Bcl-2 in all
nucleated murine BM cells, including HSCs and lympho-
cytes (Domen et al., 1998). In addition, nonhemato-
poietic tissues should also express EGFP, given the ubiq-
uitous expression of MHC class I in all nucleated cells.
Our studies document pancellular EGFP expression in
all hematopoietic cells, including RBCs and PLTs, and
demonstrate that expression of EGFP in primitive hema-
topoietic cells does not compromise their functional
activity in competitive transplantation assays. Addition-
ally, expression of EGFP in all of the tissues that were
examined will make this strain potentially useful for
experiments in which nonhematopoietic cell types are
RESULTS AND DISCUSSION
Generation of EGFP Transgenic Mice
FVB/N mice derived from the zygotes injected with the
H2Kb-EGFP construct (Fig. 1, top panel) were screened
by FACS analysis for EGFP expression in PB cells. Three
of 23 offspring demonstrated GFP expression in leuko-
cytes, RBCs, and PLTs. The transgene copy number of
offspring derived from the founder with the strongest
expression (#919) was *12–16 (Fig. 1, bottom panel),
and the configuration of the integrated transgene was in
a head-to-tail fashion, as judged by Southern blot analysis
(data not shown). Offspring derived from founder #919
displayed particularly homogenous, multilineage expres-
sion of GFP (>97% of cells) in all leukocyte subsets,
RBCs, and PLTS (Fig. 2A). This suggested that the integra-
tion site in this founder was particularly conducive to
high-level, pancellular expression, since the other found-
ers and their offspring had variable expression in RBCs
and PLTs. Additionally, GFP-positive offspring with iden-
tical, consistent expression were obtained at a rate of
about 50% in lines established from founder #919 when
GFP-positive mice were bred with wildtype mice. This
observation strongly supports the presence of one inte-
gration site, as often is the case (Voncken, 2003). These
mice were healthy, fertile, had a normal lifespan, and
showed no abnormal organ pathology (data not shown).
Both the FVB/N H2K-GFP animals and transgenic ani-
mals derived from extensive backcrossing to the C57/
Bl6 background demonstrated
(Table 1) and no change in the pattern or level of EGFP
expression out to nearly 1 year. Transgenic animals have
also been backcrossed to the Balb/c strain background
and have maintained the EGFP expression characteristics
of the FVB/N and C57/Bl6 lines.
Primitive BM Cells Express EGFP
Multicolor FACS analysis showed bright EGFP expression
in all of the BM Sca-1þLin?c-Kitþcell population
(Fig. 2B), which is enriched for HSCs. Hoechst 33342
staining also revealed normal numbers of primitive BM
SP cells (0.02–0.04% of nucleated BM cells) in the EGFP
transgenic mice, all of which expressed the EGFP trans-
gene (Fig. 2C). Additionally, all myeloid and erythroid
colonies derived from EGFP transgenic BM cells cultured
in methylcellulose medium supplemented with hemato-
poietic growth factors displayed EGFP expression as
assessed by fluorescence microscopy (data not shown).
Thus, the EGFP transgene was homogeneously and
highly expressed in the BM cell populations known to
be highly enriched for HSCs.
gene and copy number determi-
nation in GFP strain. A: GFP
cDNA is driven by the H2Kb
enhancer/promoter (box with left
hatching) with RNA processing
signals supplied by noncoding
exons (boxes with right hatching)
with intervening sequences (hori-
zontal lines) and the Moloney
(Mo) leukemia virus long-terminal
repeat (LTR) shown with fine
hatching. B: Southern blot analy-
sis of splenocyte DNA for copy
number in the GFP strain, relative
to copy number controls gener-
amounts of H2K-GFP plasmid
DNA into negative control DNA.
Quantitation was estimated by
phosphoimager densitometry and
the copy number is indicated
above each lane.
Schematic of GFP trans-
DOMINICI ET AL.
ric FACS analysis for EGFP expression (solid line) in PB cells of a representative mouse generated in the line founded by the GFPTG #919.
RBCs and PLTs from whole blood were both identified according to light-scatter characteristics, while granulocytes (Gr1), B cells (B220),
and Tcell subsets (CD4, CD8) were identified by a combination of light-scatter characteristics and specific staining with phycoerythrin (PE)-
conjugated monoclonal antibodies. WT FVB/N PB cells were used as negative control (dotted line) to evaluate the EGFP positivity for each
of the cell types. B: FACS plots for EGFP expression of primitive BM progenitor cells (Sca-1þc-KitþLin?) from an H2K-EGFP mouse. The
dot plot shows cells not staining for lineage markers which were gated on and analyzed for Sca-1 and c-Kit expression using PE- and allo-
phycocyanin-conjugated monoclonal antibodies, respectively. The gate for the Lin?Sca-1þc-Kitþcells that were analyzed for EGFP
expression is shown and is designated R1. The histogram below shows the EGFP expression profile for the gated Lin?Sca-1þc-KitþH2K-
EGFP (solid line) and WT BM cells (dotted line). C: Hoechst 33342 staining of H2K-EGFP nucleated BM cells and gate of SP cells (desig-
nated R1) which was analyzed for forward scatter (FSC) vs. EGFP expression (dot plots to right). The top dot plot shows EGFP fluorescence
of WT BM SP cells, while the lower dot plot shows the EGFP fluorescence of the H2K-EGFP SP cells. D: Analysis of EGFP-expressing PB
cells in mice transplanted with H2K-EGFP BM cells. FACS analysis for EGFP expression is shown for each indicated cell population (as
determined based on light-scatter properties). Left panel: EGFP expression in the PB cells of the H2K-EGFP donor. Middle panel: EGFP
expression in the PB cells of a representative mouse transplanted exclusively with H2K-EGFP BM cells. Right panel: EGFP expression in
the PB cells of a representative mouse transplanted with a 1:1 mixture of WTand H2K-EGFP BM cells. The average percentages (6 stand-
ard deviation) of EGFPþcells in each lineage for all recipient animals in the group (n ¼ 12 for each group) is shown above the histograms.
Pancellular EGFP expression in functional, primitive hematopoietic cells and differentiated, multilineage progeny. A: Multiparamet-
Normal Functional Activity of EGFP
In order to evaluate the functional activity of the HSCs
from this new mouse strain, lethally irradiated, normal
FVB recipient mice were transplanted with EGFP trans-
genic BM cells (Fig. 2D, left panel). Six months posttrans-
plantation, at which time hematopoiesis is fully derived
from donor long-term repopulating HSCs, FACS analysis
of PB cells of recipient animals demonstrated complete
reconstitution of hematopoiesis with EGFP cells (Fig.
2D, middle panel). Essentially all peripheral blood cells
expressed EGFP. In a long-term, competitive repopulat-
ing assay, the proportion of EGFPþcells in the PB of
recipient animals 6 months posttransplantation reflected
the composition of the infused graft, which contained
equal amounts of EGFP transgenic BM and WT BM cells
(Fig. 2D, right panel). Thus, EGFP transgenic HSCs had
equivalent functional activity relative to nontransgenic,
EGFP Expression in Tissues
In order to assess EGFP expression in both hemato-
poietic and nonhematopoietic tissues, immunohisto-
chemical staining for EGFP was performed on formalin-
fixed tissue sections. As expected, expression of the
EGFP transgene was highly expressed throughout the
BM, spleen, and thymus (Fig. 3). In addition, expression
was also strong and homogenous in liver hepatocytes,
cardiac myocytes, skeletal muscle myocytes, kidney tub-
ular cells and glomeruli, lung pneumocytes, testis, small
intestinal crypt and lamina propria cells, and cortical
and cerebellar neurons of the brain (Fig. 4). Interestingly,
in the testis the most homogenous and highest levels of
EGFP expression were noted in the immature basilar,
round spermatid cells.
Similarly, in the small intestine, the most intense stain-
ing was observed in basilar crypt cells, with loss of stain-
ing noted with progression upward along the villus.
Nearly all muscle SP cells, a population containing cells
capable of regenerating myocytes, expressed EGFP as
well (data not shown). Finally, robust EGFP expression
was also noted in the pancreatic islet cells (data not
Here we describe generation of a novel transgenic
strain that uses the H2Kbenhancer/promoter to drive
strong, pancellular EGFP expression in all hematopoietic
compartments as well as in nonhematopoietic tissues.
Furthermore, the expression of EGFP in this strain did
not result in any apparent adverse effects. In contrast to
previously reported EGFP transgenic mouse models, we
also demonstrate using a competitive repopulation assay
that HSC repopulating and differentiation capacity is not
compromised in our model. Although others have
reported transgenic megakaryocytic-erythroid specific
expression of a fluorescent marker gene (Heck et al.,
2003), the unique pancellular expression of EGFP in
mature RBCs and PLTS in our strain is coupled with uni-
form, nonvariegated expression in both the lymphoid
and myeloid lineages. Since two of the founders that
were generated had somewhat poor expression in plate-
lets and RBCs, we believe that a favorable chromosomal
site of integration rather than the specific elements with
the transgenic construct may have been a more impor-
tant factor in the expression properties of the founder
with the strong, nonvariegated expression profile.
Regardless, this particular mouse strain will facilitate
competitive repopulation assays that can easily assess
potential developmental defects of the megakaryocytic
and erythroid lineages in knockout strains of interest.
This characteristic also distinguishes this strain from the
widely used ROSA26 strain, which has not been
reported to display expression in mature RBCs or PLTs
(Zambrowicz et al., 1997). Furthermore, cells derived
Complete Blood Counts of GFP Transgenic Mice
Sample groupn % Hct Hb level, g/dLMCV, fLWBC* LYMPH*GRAN* PLT*
46.6 6 1.1
42.6 6 0.7
44.4 6 1.0
49.5 6 0.3
14.7 6 0.2
13.1 6 0.1
12.6 6 0.3
13.9 6 0.2
49.1 6 0.3
48.8 6 0.5
54.6 6 0.7
55.2 6 0.7
9300 6 1100
10200 6 1500
8084 6 823
10060 6 609
7500 6 900
7600 6 1200
5530 6 533
7868 6 444
1400 6 200
2300 6 300
2304 6 419
1948 6 142
612800 6 31900
594400 6 12700
984000 6 55312
984000 6 91582
Values are represented as means 6 SEMs. C57/Bl6 wildtype and GFP mice were 10 months old, while the FVB/N counterparts were
5 months old.
Hct, hematocrit; Hb, hemoglobin; MCV, mean corpuscular volume.
*Results are expressed in number of cells per mL.
EGFP transgenic mouse. Tissues were stained with a polyclonal
anti-GFP antiserum. Wildtype (WT) animal tissues were used as
controls (original magnification 40?).
EGFP expression in the hematopoietic tissues of an H2K-
DOMINICI ET AL.
from this new EGFP strain, unlike the ROSA26 strain,
require no staining procedure to assess marker expres-
sion and, importantly, EGFP expression allows live cells
to be easily subjected to both FACS analysis and FACS
sorting. This mouse strain should also prove useful in
HSC ‘‘plasticity’’ studies (Graf, 2002), as well as in track-
ing the fate of more differentiated lympho-hematopoietic
cells following cell transfer. Finally, since EGFP is well
expressed in nonhematopoietic tissues, especially in sev-
eral immature cell types, this mouse strain may be prove
to be useful in studies designed to identify tissue-specific
stem cells using cell transfer experiments (Weissman
et al., 2001).
MATERIALS AND METHODS
Generation of Transgenic Mice
The H2Kb-EGFP transgenic construct was generated by
inserting the EGFP cDNA (ClonTech, Palo Alto, CA) into
the H2K-i-LTR plasmid (Domen et al., 1998), which was
a gift from Dr. I. Weissman (Stanford University, Palo
Alto, CA). The EGFP transgenic construct was injected
into fertilized oocytes from FVB/N mice, followed by
transfer to foster mothers. Offspring were assessed for
EGFP expression in peripheral blood (PB) leukocytes,
RBCs, and PLTs using fluorescence activated cell sorting
(FACS) analysis on a FACSCalibur (Becton Dickinson
Immunocytochemistry Systems, San Jose, CA). The anti-
bodies Sca-1, c-Kit, B220, Gr-1, CD4, and CD8, used to
identify leukocyte subsets, were purchased from BD Bio-
sciences (Palo Alto, CA). Hoechst 33342 staining to iden-
tify the side population (SP) cells was performed follow-
ing a previously described protocol (Goodell et al.,
Bone Marrow Transplantation
The 6–8-week-old FVB/N mice were lethally irradiated
(11 Gy) and intravenously injected with either 2 ? 106
EGFP transgenic BM cells (n ¼ 12) or with 2 ? 106cells
of a 1:1 mixture of EGFP and WT BM cells (n ¼ 12).
Engraftment was determined by analyzing the propor-
tion of EGFPþcells in the PB of recipients at 6 months
posttransplantation. BM and PB were collected from ani-
mals according to standard methods.
For determination of copy number, splenocyte DNA
from C57/Bl6 GFP mice was analyzed relative to copy
number control samples that were made by mixing
appropriate amounts of plasmid DNA with nontrans-
genic C57/Bl6 splenocyte DNA. DNAs were digested
with BamHI and NotI, which liberates an *800-bp GFP
fragment. Southern blot analysis was carried out using
standard methods and a radiolabeled GFP probe. The ori-
entation of the concatameric transgenes was determined
by Southern blot analysis of BrsGI digested DNA.
Formalinfixed tissues were paraffin embedded. Dehy-
drated sections were treated with Blocking Reagent
(Roche Diagnostic, Indianapolis, IN) and subsequently
incubated with a rabbit anti-EGFP antibody (Molecular
H2K-EGFP transgenic mouse. Tissues were stained with a polyclo-
nal anti-GFP antiserum. Wildtype (WT) animal tissues were used as
controls (original magnification 40?).
EGFP expression in the nonhematopoietic tissues of an
PANCELLULAR EGFP EXPRESSION
Probes, Eugene, OR). Specific staining for EGFP was
visualized with the ABC kit (Vector Laboratories, Burlin-
game, CA) using NovaRED (Vector Laboratories) as sub-
strate. The slides were counterstained with hematoxylin
(Surgipath Medical Industries, Richmond, IL).
We thank John Raucci and the Transgenic Core facility
directed by Dr. Gerard Grosveld for generation of the
transgenic mice. We also thank the Flow Cytometry Core
Laboratory of Dr. Richard Ashmun for expert analyses of
blood cells for EGFP expression and Mike Straign and
Joseph Emmons for hematologic analysis.
Domen J, Gandy KL, Weissman IL. 1998. Systemic overexpression of
BCL-2 in the hematopoietic system protects transgenic mice from
the consequences of lethal irradiation. Blood 91:2272–2282.
Goodell MA, Brose K, Paradis G, Conner AS, Mulligan RC. 1996. Isola-
tion and functional properties of murine hematopoietic stem cells
that are replicating in vivo. J Exp Med 183:1797–1806.
Graf T. 2002. Differentiation plasticity of hematopoietic cells. Blood
Hadjantonakis AK, Gertsentstein M, Ikawa M, Okabe M, Nagy A. 1998.
Generating green fluorescent mice by germline transmission of
green fluorescent ES cells. Mech Dev 76:79–90.
Heck S, Ermakova O, Iwasaki H, Akashi K, Sun CW, Ryan TM, Townes
T, Graf T. 2003. Distinguishable live erythroid and myeloid cells in
betaglobin ECFP x lysozyme EGFP mice. Blood 101:903–909.
Manfra DJ, Chen SC, Yang TY, Sullivan L, Wiekowski MT, Abbondanzo
S, Vassileva G, Zalamea P, Cook DN, Lira SA. 1999. Leukocytes
expressing green fluorescent protein as novel reagents for adop-
tive cell transfer and bone marrow transplantation studies. Am J
Okabe M, Ikawa M, Kominami K, Nakanishi T, Nishimune Y. 1997.
‘Green mice’ as a source of ubiquitous green cells. FEBS Lett
Orlic D, Kajstura J, Chimenti S, Jakoniuk I, Anderson SM, Li B, Pickel J,
McKay R, Nadal-Ginard B, Bodine DM, Leri A, Anversa P. 2001.
Bone marrow cells regenerate infracted myocardium. Nature
Persons DA, Allay JA, Riberdy JM, Wersto RP, Donahue RE, Sorrentino
BP, Nienhuis AW. 1998. Use of the green fluorescent protein as a
marker to identify and track genetically modified hematopoietic
cells. Nat Med 10:1201–1205.
Priller J, Persons DA, Klett FF, Kempermann G, Kreutzberg GW, Dirnagl
U. 2001a. Neogenesis of cerebellar Purkinje neurons from gene-
marked bone marrow cells in vivo. J Cell Biol 155:733–738.
Priller J, Flugel A, Wehner T, Boentert M, Haas CA, Prinz M, Fernandez-
Klett F, Prass K, Bechmann I, de Boer BA, Frotscher M, Kreutzberg
GW, Persons DA, Dirnagl U. 2001b. Targeting of gene modified
hematopoietic cells to the central nervous system: use of the
green fluorescent protein uncovers microglial engraftment. Nat
Voncken JW. 2003. Transgenic mouse methods and protocols. In:
Hofker MH, van Deursen J, editors. Genetic modification of the
mouse. Totowa, NJ: Humana Press.
Wagers AJ, Sherwood RI, Christensen JL, Weissman IL. 2002. Little evi-
dence for developmental plasticity of adult hematopoietic stem
cells. Science 297:2256–2259.
Weissman IL, Anderson DJ, Gage F. 2001. Stem and progenitor cells: ori-
gins, phenotypes, lineage, commitments, and transdifferentiations.
Annu Rev Cell Dev Biol 17:387–403.
Wright DE, Cheshier SH, Wagers AJ, Randall TD, Christensen JL, Weiss-
man IL. 2001. Cyclophosphamide/granulcyte colony-stimulating
factor causes selective mobilization of bone marrow hemato-
poietic stem cells into the blood after M phase of the cell cycle.
Zambrowicz B, Imanoto A, Fiering S, Herzenberg l, Kerr W, Soriano P.
1997. Disruption of overlapping transcripts in the ROSA b-geo 26
gene trap strain leads to widespread expression of b-galactosidase
in mouse embryos and hematopoietic cells. Proc Natl Acad Sci U S
DOMINICI ET AL.