| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Technical Advance |
From the Departments of Pathology*
and Medical
Microbiology and Immunology,
University of
Ulm, Ulm, Germany
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Green fluorescent protein (GFP) is a bioluminescent protein from the jellyfish Aequorea victoria.3,4 Mutant variants of this protein have been developed that show enhanced fluorescence. A commonly used variant with enhanced fluorescence is GFPmut1 (designated eGFP in this article).5 eGFP has been used as a reporter gene in prokaryotic and eukaryotic cells in vitro. A transgenic C57BL/6 mouse line has been constructed that expresses eGFP under the chicken ß-actin promoter control.6-10 These mice have been used to study hemopoietic and lymphoid development,11-13 spermatogenesis,14 and maternal cell transmission to offspring through placenta during pregnancy and by breast-feeding after birth.15 These studies required the isolation of viable cells from the host and their subsequent flow cytometric identification and characterization. Because of the lack of optimized techniques to detect eGFP-expressing cells in situ, an analysis of cell migration, proliferation, and differentiation by immunohistological techniques was difficult to perform in this attractive model.
We transferred eGFP+ cells of well-defined lymphoid or myeloid subsets, ie, splenic CD4+ T cells or bone marrow-derived dendritic cells (DCs), from eGFP-transgenic (eGFP-tg) C57/BL6 (B6) donor mice into congenic, immunodeficient RAG1-/- B6 hosts. We used the stable expression of eGFP as a sensitive and highly specific marker for the in situ detection of donor cells in repopulated tissues of the host. We describe a method for processing tissue that permits the simple, sensitive, and specific detection of eGFP+ donor cells in histological sections of transplanted hosts. In addition, we demonstrate that this technique can be combined with three other labeling techniques to analyze in situ the fate of the engrafted eGFP+ cells. It can be used with 1) fluorescent antibody staining to determine in situ the heterogeneity and distribution of marker profile expression in engrafted and autochthonous cell populations indicative of their differentiation and/or activation; 2) terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) staining to detect apoptotic death of engrafted and resident cells; and 3) immunolabeling of incorporated bromodeoxyuridine (BrdU) to measure the fraction of proliferating graft and host cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
eGFP-tg C57BL/6J mice and C57BL/6J-Rag1tm1Mom (RAG1-/-) mice16 (stock no. 2216; Jackson Laboratories, Bar Harbor, ME) were used. Dr. Masaru Okabe, Osaka University Research Institute for Microbial Diseases (Osaka, Japan) kindly provided eGFP-tg B6 mice.6,11 These mice carry copies of the eGFP-transgene cloned into a pCAGGS expression vector. Breeding colonies of RAG1-/- mice were established under specific pathogen-free conditions in the animal colony at the University of Ulm.
Isolation and Adoptive Transfer of eGFPbright CD4+ Cells and DCs
Spleen cells were obtained from of eGFP-tg B6 mice. CD8+ T cells were depleted from these spleen cells by treatment with anti-CD8 antibody and low toxicity rabbit complement (catalog no. CL3051; Cedarlane, Homby, Canada) following the manufacturer’s instructions. CD4+ T cells were enriched to >98% purity by positive selection using the magnetic-activated cell separation (MACS) system (Miltenyi Biotec, Bergisch-Gladbach, Germany). Briefly, 107 cells were incubated with CD4(L3T4) Micro Beads (1:10 dilution, catalog no. 130-049-201; Miltenyi Biotec) in 100 µl of MACS buffer [phosphate-buffered saline (PBS) supplemented with 2 mmol/L ethylenediaminetetraacetic acid (EDTA) and 0.5% bovine serum albumin] for 30 minutes at 8°C. Cells were washed twice in MACS buffer, resuspended in 500 µl of MACS buffer, and pipetted onto a prerinsed LS separation column (catalog no. 130-042-401, Miltenyi Biotec) attached to a MidiMACS separation unit (catalog no. 130-042-302, Miltenyi Biotec). After washing three times with 3 ml of MACS buffer, columns were removed from the separation unit. To elute bound cells, 6 ml of MACS buffer were passed through the columns using a plunger supplied with the columns. Cells from eGFP+-tg B6 donors were always selected for high eGFP expression (eGFPbright cells) by fluorescence-activated cell sorting before transfer into the eGFP- host. The purity of the positively separated CD4+ population was always >98%. We injected intraperitoneally into RAG1-/- B6 hosts 3 x 105 CD4+ eGFPbright T cells per mouse.
The in vitro generation of DCs from murine bone marrow has been described.17 Briefly, bone marrow cells prepared from femurs were depleted of CD4+ and CD8+ T cells, B220+ B cells, and MHC-class-II+ maturing myeloid cells (antibody catalog no. 492-01, 494-01, 495-01, 524-01; Miltenyi Biotec) by MACS sorting. These progenitor-enriched marrow cells were cultured at a density of 106 cells/ml in serum-free UItraCulture medium (BioWhittaker, Verviers, Belgium) supplemented with 5 ng/ml GM-CSF and 10 ng/ml Flt3 ligand (catalog nos. 315-03 and 300-19; PeproTech, Rocky Hill, NJ), 2 mmol/L glutamine, and antibiotics. Cultures were incubated at 37°C in humidified air with 5% CO2. Cells were fed by medium exchange on day 3 and day 5 of culture. DCs were harvested at day 8 of culture. We injected intraperitoneally into RAG1-/- B6 hosts 1 x 107 DCs per mouse.
Only 30 to 45% of the homozygous transgenic B6 mice used in this study express the transgene. The expression pattern of the eGFP-encoding transgene was heterogeneous. In some organs, eg, liver and kidney, strongly eGFP+ epithelial cells were found beside cells without detectable eGFP expression. Tissues from eGFP+-tg B6 mice were used to study tissue processing for the presence of eGFP fluorescence.
Tissue Processing
Tissue specimens obtained from eGFP-tg B6 mice or transplanted RAG1-/- mice included the mesenteric and peripancreatic lymph nodes, spleen, thymus, liver, kidney, small and large intestine, femur, and vertebral column. The tissues were either snap-frozen in liquid nitrogen and stored at -70°C, or treated with a 4% phosphate-buffered paraformaldehyde solution (pH 7.2) for 16 hours at 8°C. For bone tissue, a prolonged fixation in paraformaldehyde for 2 days was chosen followed by decalcification in an EDTA solution (200 g/L, pH 7.2) at 37°C for 1 week.
Tissue samples submitted to chemical fixation and paraffin embedding were further processed in a Tissue-Tek VIP automate (Sakura, Torrance, CA) set to the following program: formalin-fixation at 35°C (1x 30 minutes), dehydration in an ascending ethanol row at 40°C (1x 75%, 1x 80%, 2x 96%, 2x 100%; 30 minutes per each step), equilibration in xylene at 40°C (1x 30 minutes plus 1x 1 hour) and soaking in paraffin at 56°C (3x 30 minutes plus 1x 1 hour). After embedding in paraffin, tissue was cut into 2-µm sections that were subsequently deparaffinized in xylene, rehydrated by decreasing ethanol concentrations, and embedded in Vectashield mounting medium (catalog no. H1000; Vector, Burlingame, CA). As an alternative to paraffin embedding, tissue samples were incubated for 48 hours at 8°C in 4% phosphate-buffered paraformaldehyde solution (pH 7.2) containing sucrose at various concentrations (0 to 60%), snap-frozen in liquid nitrogen and processed as described below.
Paraformaldehyde fixed or untreated cryopreserved tissue was embedded in OCT compound (catalog no. 4583; Miles Inc., Naperville, IL). Two-µm cryosections were cut with a cryostat (Leica, Wetzlar, Germany). The sections were mounted on poly-L-lysine-coated microscopic slides. The mounted sections were air-dried for 12 hours at room temperature. They were then left untreated, or were immersed for 30 minutes at room temperature in either paraformaldehyde solution, or acetone, or 70% ethanol. Fixed sections were subsequently rinsed in PBS buffer for 5 minutes at room temperature. They were thereafter incubated for 1 minute with 0.3 µg/ml 4,6-diamidino-2-phenylindole (catalog no. D9542; Sigma-Aldrich, Deisenhofen, Germany) for fluorescence nuclear counterstaining and embedded in Vectashield medium. In addition, hematoxylin and eosin (H&E) staining was performed.
In Situ Detection of eGFP+ Cells that Have Bound Fluorochrome-Conjugated Antibodies
For immunohistology, 2-µm cryosections of paraformaldehyde-fixed and subsequently snap-frozen tissue were mounted on poly-L-lysine-coated microscopic slides and air-dried for 12 hours at room temperature. These cryosections were first incubated with a biotin-conjugated anti-CD45R/B220 clone RA3–6132 (dilution 1:100, catalog no. 01122D; Pharmingen, Hamburg, Germany) or anti-CD45 clone 30-F11 (dilution 1:30, catalog no. 01112D; Pharmingen) mAb for 1 hour, washed twice in PBS buffer, and then incubated with Cy3-conjugated streptavidin (catalog no. 016-160-084; Dianova, Hamburg, Germany) for 30 minutes at a dilution of 1:1000 to detect bound mAb. All incubation steps were performed at room temperature in PBS. To test if eGFP fluorescence can withstand demasking of formalin-resistant epitopes, sections were immersed in 200 ml of citrate-buffer (0.01 mol/L, pH 6.0) and heat-treated in a microwave oven at 800 W for 20 minutes.
In Situ Detection of in Vivo Proliferating eGFP+ Cells by BrdU Labeling
Proliferating cells labeled with BrdU in vivo were detected in situ. Animals were intraperitoneally injected 1 hour before sacrifice with 1 mg/mouse BrdU (catalog no. 85002, Sigma-Aldrich) dissolved in 0.2 ml of PBS. Cryosections of paraformaldehyde-fixed intestinal tissue were denatured in various concentrations of hydrochloric acid (0.002 to 2.0 mol/L) and neutralized in borate buffer (0.1 mol/L, pH 8.0) as described.18 BrdU-positive nuclei were detected by the biotin-conjugated anti-BrdU mAb clone BR-3 (catalog no. MD5215; Caltag, Burlingame, CA) and Cy3-labeled streptavidin. Immediately after staining and embedding in Vectashield mounting medium, sections were evaluated by fluorescent microscopy.
In Situ Detection of Apoptotic eGFP+ Cells by TUNEL Staining
Cells undergoing apoptosis were detected in situ by labeling DNA strand breaks using TUNEL.19 Briefly, cryosections of paraformaldehyde-fixed mesenteric lymph nodes were digested with proteinase K (20 µg/ml, Sigma-Aldrich) for 15 minutes at 37°C. The labeling reaction was performed using 10 U of TdT (catalog no. M1871; Promega, Mannheim, Germany) and 20 µmol/L of biotin-16-dUTP (catalog no. 1093070, Boehringer-Mannheim) in 50 µl of TdT buffer (0.5 mol/L cacodylic acid, sodium salt, pH 6.8; 1 mmol/L CoCl2, 0.5 mmol/L dithiothreitol, 0.05% w/v bovine serum albumin, 0.15 mol/L NaCl). Labeled cells were detected using Cy3-conjugated streptavidin (catalog no. 016-160-084, Dianova) diluted 1:5000.
Fluorescence Microscopy
Slides were examined under an Axioskop fluorescence microscope (Zeiss, Jena, Germany) using a mercury-vapor light source and an appropriate filter set. Images were recorded by a video camera and processed on a computer using the ISIS3-software (version 3.02; Metasystems, Heidelberg, Germany).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We obtained tissues from either eGFP-tg B6 mice (selected for eGFP expression), or RAG1-/- B6 mice transplanted with purified eGFP+ cells. We tested different protocols for tissue processing to optimize detection of eGFP+ cells in situ.
eGFP fluorescence was almost completely eliminated in paraformaldehyde-fixed, paraffin-embedded tissue. Only minimal eGFP fluorescence could be observed in organs with high levels of eGFP expression, such as liver and kidney (data not shown).
We tested if eGFP+ cells can be detected in snap-frozen, cryopreserved tissue. eGFP fluorescence was virtually undetectable on cryosections from liver and kidney that were counterstained and embedded in Vectashield mounting medium. This effect is certainly because of the high solubility of eGFP in aqueous media that has been described before.9 We therefore treated air-dried cryosections with fixative agents for 30 minutes at room temperature. Among these, acetone had no effect whereas 4% paraformaldehyde solution and 70% ethanol rescued eGFP fluorescence only incompletely which moreover was faint and ill circumscribed. The low effectiveness of ethanol treatment can be ascribed to its limited ability to inhibit diffusion of eGFP, whereas cross-linking by paraformaldehyde fixation was probably too slow to counteract the diffusion process.
We fixed cryosections from tissues in paraformaldehyde solution before
snap-freezing in liquid nitrogen. This procedure maintained bright eGFP
fluorescence readily detectable by fluorescent microscopy. Individual
eGFP+ cells or eGFP+ cell
groups could be clearly identified in different tissues (Figure 1, A and B)
. The high intensity of eGFP
fluorescence permitted sensitive detection of positive cells even in
organs displaying a high autofluorescent background, such as liver
tissue (Figure 1A)
.20
The eGFP signal was remarkably
stable as it was detectable without significant loss of signal strength
in sections after storage for several weeks at room temperature. When
H&E stainings from cryosections of paraformaldehyde-fixed tissue were
examined, the histomorphology was found to be impaired by intercellular
retraction artifacts and cytoplasmic clumping (Figure 1C)
. This was
overcome by adding sucrose to the paraformaldehyde solution. Of the
different sucrose concentrations tested, 50% seemed to be optimum and
resulted in improved histomorphology (Figure 1D)
.
|
To detect eGFP+ cells in bone marrow,
paraformaldehyde-fixed femur and vertebrae were decalcified in the
presence of EDTA and subsequently snap-frozen in liquid nitrogen. This
protocol allowed us to readily detect eGFP+ cells
on frozen sections containing bone tissue that were cut with a
conventional cryostat (Figure 2, A and B)
. The prolonged treatment with EDTA had no detrimental influence on
eGFP fluorescence, as can be demonstrated by the brightly fluorescent
chondrocytes in Figure 2A
.
|
Because visualization of eGFP in unfixed cryosections was
virtually impossible because of quick diffusion in aqueous media,
immunohistology had to be performed on paraformaldehyde-fixed tissue
and was therefore restricted to the use of antibodies against
formalin-resistant epitopes. We used the anti-CD45R/B220 (clone 16A)
and anti-CD45 (clone 30-F11) mAbs that are able to recognize
formalin-fixed antigen (BD Pharmingen, technical information). With
CD45R/B220, it was possible to label B lymphocytes in lymph node tissue
while simultaneously detecting eGFP+ cells
(Figure 3A)
. CD45 displayed a moderately
strong staining of leukocytes, including eGFP+
adoptively transferred T cells (Figure 3B)
. Recognition of antigens in
tissue fixed by cross-linking agents is often improved by target
retrieval using heat.21
Following established protocols,
we tried boiling the sections while immersed in citrate buffer in a
microwave oven. This procedure led to a nearly complete loss of eGFP
fluorescence that can probably be ascribed to the applied heat (the
Tm of wild-type GFP is 70°C) but not to the
mildly acidic buffer (pH 6.0) because eGFP is reported to retain 50%
of its fluorescence at this pH.22
|
For the detection of proliferating cells, mice were injected
with BrdU 1 hour before sacrifice. Cryosections of
paraformaldehyde-fixed tissue were treated with hydrochloric acid to
denature the double-stranded DNA thus exposing incorporated BrdU to the
mAb. As shown in Figure 3, C and D
, nuclei of proliferating cells at
the base of intestinal crypts were specifically labeled. At the same
time, eGFP fluorescence was preserved to an extent that readily
permitted unambiguous microscopic detection of
GFP+ cells in eGFP-tg B6 mice (Figure 3C)
and
RAG1-/- hosts transplanted with
eGFP+ T cells (Figure 3D)
. Titration of
hydrochloric acid revealed that a concentration of 0.2 mol/L was
optimum to sufficiently retain both eGFP+
fluorescence and BrdU staining. These data demonstrate that a low pH of
0.7 does not completely (or not irreversibly) abolish the fluorescence
activity of eGFP. It was crucial to evaluate the BrdU stainings
immediately after their processing and labeling because the intensity
of the fluorescent signals markedly deteriorated within a few hours.
In Situ Detection of Apoptotic eGFP+ Cells and Host Cells
Apoptotic cell death was detected by TUNEL staining of DNA strand
breaks performed at cryosections of paraformaldehyde-fixed tissue.
Numerous apoptotic cells could be found in the mesenteric lymph nodes
of eGFP-tg B6 mice (Figure 3E)
and RAG1-/-
hosts transplanted with eGFP+ T cells (Figure 3F)
. Some of the apoptotic bodies proved to be remnants of
eGFP+ cells, as demonstrated by their yellow
color resulting from the addition of green and red fluorescence.
Staining was specific as no apoptosis was visible in negative control
reactions on the omission of TdT. eGFP fluorescence was not
significantly impaired in TUNEL reactions. Note that proteolytic
digestion with proteinase K, a step in the TUNEL-staining protocol, had
no obvious influence on eGFP fluorescence.
Two Instructive Examples to Demonstrate the Power of the System and the Protocols
Transfer of eGFP+ DCs into Congenic RAG1-/- hosts
In vitro cultured bone marrow-derived DCs from eGFP-tg
B6 donor mice were transferred into congenic
RAG1-/- hosts. We found low numbers of
eGFP+ DCs in the thymus of the transplanted host
2 to 5 days after transfer (Figure 4A)
.
Higher numbers of eGFP+ DCs were found in the
mesenteric and peripancreatic lymph nodes of the adoptive host (Figure 4B)
. The level of eGFP fluorescence of DCs was similar to the high
fluorescence intensity of engrafted CD4+ T cells
and permitted the sensitive detection of the in situ
distribution of these potent antigen-presenting cells.
|
We transferred purified eGFP+
CD4+ T cells from eGFP-tg B6 donor mice into
congenic, immunodeficient RAG1-/- hosts.
CD4+ T cells were selected for the
eGFPbright phenotype on the
fluorescence-activated cell sorting because eGFP is heterogeneously
expressed from the transgene in the CD4+ T cell
population of tg mice. Transferred eGFP+ cells
were easily detected on cryosections of paraformaldehyde-fixed
tissue from transplanted mice. Repopulation of the host with
eGFP+ donor T cells could be readily monitored in
any target organ of interest, such as liver, mucosal tissues, bone
marrow, spleen, or lymph nodes (Figure 5, A–E)
even if only low numbers of eGFP+ cells
were present (Figure 5
; B, E, and F). All CD4+
lymphoid cells found in transplanted RAG1-/-
host 3 to 5 weeks after transfer were eGFP+
(Figure 6)
. Thus, within the adoptive
host, eGFP expression by T cells was stable.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The in situ detection of donor eGFP+
cells is readily achieved on histological sections processed according
to the simple technical guidelines summarized in Table 1
. Unlike paraformaldehyde-fixed frozen
tissue, paraffin-embedded tissue showed a nearly complete loss of
eGFP-fluorescence. In contrast to published data and the results shown
in this article, Walter and colleagues26
found only a
minimal reduction of eGFP fluorescence after a paraffin-embedding
procedure when they examined tissue from mice that express
transgene-encoded eGFP under CMV promotor control. The difference in
these findings is difficult to explain. Loss of eGFP fluorescence may
be because of the high background fluorescence present in
paraffin-embedded material.9
Alternatively, eGFP
fluorescence may be destroyed during dehydration, paraffin embedding,
and rehydration. The method we describe also offers the advantage to be
suitable for direct and technically simple identification of donor-type
cells in EDTA-treated bone marrow, thereby circumventing cumbersome
procedures required for producing cryosections of bone tissue that has
not been previously decalcified.27
This is of considerable
practical significance because in situ evaluation of the
bone marrow is likely to play a pivotal role in many adoptive transfer
experiments.
|
Taken together, the usage of eGFP+ donor cells for adoptive transfer experiments combined with the novel histological tissue processing we describe facilitates many studies on the in situ development of differentiating cell systems, eg, in lymphoid or myeloid cell lineages. In addition, eGFP+ cells can be recovered at different time points after transfer from the adoptive host and assayed in vitro for expression of cell surface markers or for functional reactivities.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by grant no. I.A07.22 from the Interdisziplinäres Zentrum für Klinische Forschung, Medical University of Ulm (to F. L. and J. R.), and grant no. DFG Re549/9-1 from the Deutsche Forschungsgemeinschaft (to J. R.).
Accepted for publication February 26, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  F. Leithauser, T. Meinhardt-Krajina, K. Fink, B. Wotschke, P. Moller, and J. Reimann Foxp3-Expressing CD103+ Regulatory T Cells Accumulate in Dendritic Cell Aggregates of the Colonic Mucosa in Murine Transfer Colitis Am. J. Pathol., June 1, 2006; 168(6): 1898 - 1909. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Moldenhauer, S. W. Popov, B. Wotschke, S. Bruderlein, P. Riedl, N. Fissolo, R. Schirmbeck, O. Ritz, P. Moller, and F. Leithauser AID expression identifies interfollicular large B cells as putative precursors of mature B-cell malignancies Blood, March 15, 2006; 107(6): 2470 - 2473. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. L. Kusser and T. D. Randall Simultaneous Detection of EGFP and Cell Surface Markers by Fluorescence Microscopy in Lymphoid Tissues J. Histochem. Cytochem., January 1, 2003; 51(1): 5 - 14. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
