| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Servicio de Inmunología, Hospital Universitario "Marqués de Valdecilla," Instituto Nacional de la Salud, Santander, Spain; and
Departamento de Biología Celular, Facultad de Medicina, Universidad Complutense, Madrid, Spain
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To explore pathogenetic mechanisms in DS, investigators have produced and characterized animal models. Mouse chromosome 16 is the chromosome most homologous to human chromosome 21, and at least 14 defined genes have been localized on both the human and mouse chromosomes in the critical DS region. Mouse fetuses with trisomy 16, a mouse model for human trisomy 21, also exhibit abnormalities similar to those found in DS: a hypoplastic thymus; and a decreased number of hemopoietic progenitor cells in the liver (4). Abnormal development of thymus and spleen cells in trisomy 16 diploid chimeric mice has also been reported. Unfortunately, trisomy 16 is incompatible with life and postnatal studies could not be performed.
Recently, segmentally trisomic mice, Ts(1716)65Dn (abbreviated Ts65Dn), appear to fulfill the criteria for a DS model. These mice are trisomic for a segment of chromosome 16 that is homologous to human chromosome 21. The segment includes material just proximal to App and extends to Mx (5). Hence, much of the distal end of murine chromosome 16, encoding most of the segment shared with the long arm of human chromosome 21 in the q22 region, is translocated to the centromeric end of murine chromosome 17, forming a small translocation chromosome (6).
Ts65Dn mice survive to adulthood. This made this model particularly valuable for an experimental design of extended duration. Although preliminary characterization of Ts65Dn mice has revealed several consistent phenotypic abnormalities, some of which resemble those seen in DS patients (5, 6, 7, 8), no immunological studies of this model have been published. To determine to what extent elevated levels of normal gene products encoded by genes on the triplicate chromosome (mainly superoxide dismutase-1 gen) could affect the functional integrity of the thymus, we conducted a systematic study of these mice.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Female trisomic mice ((C57BL/6JEi x C3H/HeSnJ)F1-Ts65Dn, abbreviated B6EiC3H-a/A-Ts65Dn) were originally obtained from The Jackson Laboratory, Bar Harbor, ME. Carriers were mated with male B6EiC3H-a/A mice, also from Jackson, to maintain this background. They were bred and maintained in pathogen-free environmental facilities. Sterile trisomic males (6) were used for experimental purposes, whereas trisomic females were used for reproduction. Age- and MHC-matched euploid male progeny were used throughout the study as controls.
General reagents
Salmonella typhimurium LPS was from Difco (Detroit,
MI). Ferricytochrome c, bovine superoxide dismutase (SOD, EC
1.15.1.1), colchicine, PMA,
H2O2,
D-mannitol, 1,3-dimethylurea,
1,3-dimethyl-2-thiourea (DMTU), 3,3',4',5,7-pentahydroxyflavone
(quercetin), IFN-
, HRP (EC 1.11.1.7), dexamethasone (DEX),
3,3'-diaminobenzidine tetrahydrochloride chromogen,
diethyldithiocarbamic acid (DDC),
N-acetyl-L-cysteine (NAC),
2-mercaptoethylamine (cysteamine), N-propyl gallate, NADPH,
phenol red, sulfanilamide, naphthylethylene diamine dihydrochloride,
SDS, and phosphoric acid were from Sigma (St. Louis, MO). PE and
2',7'-dichlorofluorescein diacetate (DCFH-DA) were purchased from
Molecular Probes (Eugene, OR). Hydrazine carboximidamide hemisulfate
(aminoguanidine) was from Calbiochem (San Diego, CA).
Biotin-16-2'-deoxy-UTP (biotin-dUTP), dATP, TdT, and TdT buffer were
purchased from Boehringer Mannheim Biochemicals (Mannheim, Germany).
Culture media were RPMI 1640 (R0) and DMEM (D0), supplemented with 2 mM
D-glutamine and 10% heat-inactivated FCS
(HyClone, Logan, UT) (R10 and D10, respectively). Normal mouse, goat
(NGS), and rabbit sera were used for blocking. All media and reagents
used were confirmed to be endotoxin free (<0.01 ng/ml) by chromogenic
Limulus amebocyte lysate microassay from Whittaker M.A.
Bioproducts (Walkersville, MD).
Antibodies
FITC-labeled rat anti-mouse CD4 (IgG2b, GK1.5), PE-labeled rat anti-mouse CD8 (IgG2a, 3B5) and FITC-labeled hamster anti-mouse CD3 (IgG, S4.1) mAbs were purchased from Caltag Laboratories (San Francisco, CA). The following Abs were obtained from PharMingen (San Diego, CA): biotinylated rat anti-mouse CD35 (IgG2a, 8C12); biotinylated mouse anti-murine Iak (IgG2b, 11-5.2); biotinylated mouse anti-murine Iab (IgG2a, KH74); biotinylated mouse anti-murine Iad (IgG3, 39-10-8); PE-rat anti-mouse CD45R/B220(IgG2a, RA3-6B2); biotinylated rat anti-mouse CD119 (IgG2a, GR20); and purified hamster anti-murine Bcl-2 (IgG, 3F11). Alkaline phosphatase-conjugated goat antiserum to mouse IgG and IgM were obtained from Sigma. Biotinylated F(ab')2 goat anti-hamster IgG (Jackson ImmunoResearch, West Grove, PA), streptavidin-PE, and streptavidin-Red 670 (Tago, Burlingame, CA) were also used as secondary reagents.
Mitotic chromosome preparations
Whole blood from the retroorbital sinus was collected in heparinized microhematocrit tubes. Blood (100 µl) was cultured at 37°C in R10 in the presence of 50 µg/ml LPS and 6 µg/ml PHA for 42 h (9). One-half hour before the end of the incubation time, 0.1 ml colchicine (50 µg/ml) was added to each tube. After centrifugation (250 x g, 22°C, 10 min) 2–3 ml of a prewarmed 0.56% KCl solution were added to each tube and incubated for 15 min. They were centrifuged again (480 x g, 22°C, 10 min) and finally fixed (22°C, 30 min) with 3 ml methanol-glacial acetic acid (3:1). Samples were washed three times in freshly prepared fixative before smears were prepared by gentle dropping onto clean slides.
In vivo LPS and DEX administration
In preliminary protocols, mice were injected i.p. with various concentrations of LPS, but the data reported here were from mice injected with 50 µg LPS. Three or four animals were used for each experimental group. A similar procedure was adopted for DEX administration. In this case, 2 mg of the drug were injected i.p. into each mouse. The LPS- and DEX-injected animals were sacrificed after 10 or 48 h, respectively. For some in vivo experiments, parallel sets of both control and trisomic mice were pretreated with NAC or DDC (100 mg/kg body weight i.p. injected daily for 2 weeks) before DEX administration. As controls, both euploid and trisomic mice were injected with endotoxin-free sterile PBS in parallel.
Microscopy and ultrastructural studies
Thymuses were fixed in Bouin’s fixative for 5 h, rinsed, and transferred into 70% ethanol. Tissues were then processed through alcohols and xylene, embedded in paraffin, sectioned at 5 µm, and either stained with hematoxylin and eosin or used for immunohistochemistry. For in situ thymocyte apoptosis detection, 5-µm-thick sections from fixed, paraffin-embedded tissues were processed for 3'-hydroxy-DNA end staining. Cells showing nuclear DNA fragmentation were identified by the TUNEL method (10). Tissue sections were counterstained with methylene green. Samples for electron microscopy were fixed in 4% glutaraldehyde in Sorensen’s buffer and processed routinely.
SOD-1 activity
To measure SOD-1 activity, supernatants of tissue homogenates (thymus, erythrocytes, and liver) were collected and assayed with a commercial kit (Calbiochem), according to the manufacturer’s instructions.
Cell culture
Murine thymuses were placed in sterile PBS, and the tissue was disrupted with the tip of the plunger of a sterile 1-ml syringe. Cells were washed in cold PBS and centrifuged three times (400 x g, 5 min, 4°C) to remove cellular aggregates. Single-cell preparations were counted and viability assayed by trypan blue exclusion. For in vitro experimental procedures, a single-cell suspension (1 x 106/ml) was prepared and cultured with or without LPS (50 µg/ml) or DEX (1 µM) for different periods of time. For some experimental procedures, thymocytes were cultured in the presence of selected scavengers or inhibitors as described. Peritoneal macrophages were obtained as previously described (11) and cultured in appropriate medium for functional studies.
Quantitation of reactive oxygen intermediates (ROI) and determination of nitrite production
To quantitate ROI, O2- and H2O2 production were measured. O2- production was measured by the SOD-inhibitable reduction of ferricytochrome c adapted to a microplate format as previously described (12). Results are expressed as nanomols O2- produced per mg protein. Each result represents the mean of eight determinations. H2O2 production was measured by a technique based on the HRPO-dependent conversion of phenol red by H2O2 into a compound with increased absorbance at 600 nm (12). Results are expressed as nanomols H2O2 produced per mg protein. Each result represents the mean of five determinations. In addition, a highly sensitive flow cytometric method was used to detect intracellular H2O2 formation in thymocytes (13); the method used the H2O2-sensitive fluorescent probe DCFH-DA. Thymocytes were twice washed in PBS, resuspended at 1 x 106 cells/ml in phenol red-free HEPES-buffered D10, and incubated with 5 µM DCFH-DA at 37°C. Interaction of DCFH-DA with peroxides gives rise to the fluorescent DCFH. This oxidation reaction leads to the formation of the highly fluorescent molecule 2',7'-dichlorofluorescein (DCF) detectable by FACS analysis (14). Approximately 104 cells per time point were analyzed (excitation and emission settings were 488 ± 15 and 535 ± 15 nm, respectively). Mean fluorescence peak height was determined during 60 min incubation at 37°C. The generation of nitrite (NO2-) was used to estimate reactive nitrogen intermediates (RNI) indirectly. It was determined by the microplate method of Ding et al. (12, 15). Results of unknown culture fluids were expressed as nanomols NO2- produced per well; derived from a sodium nitrite standard curve.
Determination of lipid peroxidation by-products
Total lipid peroxides (LPO), malonaldehyde (MDA), and 4-hydroxy-2(E)-nonenal (4-HNE) were quantitated as described (16). Briefly, cells were cultured at 1–2 x 106 cells/ml for 1 h with or without the reagents to be tested in R0 medium supplemented with 5% FCS, in the absence of phenol red. For LPO determination, cells were washed twice and solubilized with 0.1% Triton-X and 0.05% deoxycholate in saline. For the detection of MDA and 4-HNE, the cells were resuspended in 20 mM Tris-HCl buffer, pH 7.4, and submitted to three cycles of freeze-thawing and further centrifugation. Commercially available colorimetric assays were used to detect LPO (Kamiya Biomedical, Thousand Oaks, CA) and MDA and 4-HNE (British Biotechnology Products, Abingdon, U.K.) according to the manufacturers’ recommendations.
Flow cytometric analysis
Two- and three-color flow cytometric analyses of single-cell suspensions of thymus were performed with anti-CD4-, anti-CD8-, anti-CD3-, anti-CD45R/B220-, and anti-CD119-conjugated mAbs. In other cases, two-color cytometry was performed with conjugated anti-CD3, unlabeled hamster IgG anti-murine Bcl-2 followed by staining with biotinylated-goat anti-hamster and streptavidin-PE, as previously described (17). Cells were analyzed with a FACScan flow cytometer and a minimum of 3 x 104 events per sample were counted using Lysis software (Becton Dickinson, Mountain View, CA). The apoptosis levels in thymocytes were assessed by propidium iodide (PI) staining as previously described (18). Briefly, a pellet of 5 x 105 cells was resuspended in 600 µl H2O containing PI (50 µg/ml), 0.1% sodium citrate, and 0.1% Triton X-100 and then incubated overnight in the dark at 4°C. Cells were analyzed by flow cytometry. To characterize the phenotype of the apoptotic cells, the detection of a given cell surface marker in TUNEL-positive cells was performed by three-color flow cytometry as described previously (19). Briefly, thymocytes were incubated (30 min, 4°C) with the appropriate anti-CD4 and anti-CD8 Abs in 100 µl PBS. After two washings with PBS, they were fixed with 2% paraformaldehyde for 30 min, washed twice again, and incubated for 2 min with 1% Triton X-100 containing 0.1% sodium citrate. After an additional washing step, the TUNEL reaction was conducted by incubating the cells (1 h, 37°C) with 0.3 nmol biotin-dUTP, 3 nmol dATP, 25 U TdT, and TdT buffer. Cells were washed and incubated with streptavidin-Red 670.
DNA fragmentation
DNA extraction was performed as described (20). In brief, 2 x 106 thymocytes were lysed in 0.5 ml 10 mM Tris-HCl, pH 8, containing 10 mM EDTA and 0.6% SDS. The DNA was then extracted by phenol-chloroform and ethanol precipitated. Samples were electrophoresed in 2% agarose gel.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In both human and previous experimental murine models of DS, the involutive morphological changes and functional abnormalities of the thymus have been described (21, 22, 23, 24, 25, 26, 27). Thymic morphology of Ts65Dn mice was compared with normal littermates at 2, 4 and 7–12 wk of age. No remarkable ultrastructural abnormalities were observed in 2-wk-old Ts65Dn mice (not shown). At 4 wk, trisomic mice exhibited slight abnormalities in both cortical and medullar zones, mostly consisting of increased numbers of TUNEL-positive cells, modified epithelial cells, and macrophages (not shown).
Changes in thymic morphology of Ts65Dn mice were clearly seen when
compared with age-matched normal littermates. TheTs65Dn thymic cortex
was remarkably different. At 7–12 wk of age, the lower number of
subcapsular lymphocytes (*) with hyperchromatic apoptotic-like
clusters (arrows) was evident in Ts65Dn thymuses (see Fig. 1
B vs normal thymus in Fig. 1A). At higher magnification (Fig. 1C),
rosette-like grouped cells (most probably representing degenerative
nuclei) exhibit highly basophilic nuclei of variable size (arrow), in
the subcapsular zone. In Ts65Dn mice, numerous grouped TUNEL-positive
nuclei in small nests close to epithelial cells were seen (Fig. 1E). The number of TUNEL-positive were specially abundant in
areas close to the connective tissue of the capsular and cortical septi
of Ts65Dn thymuses (Fig. 2B).
This moth-eaten or empty aspect of the Ts65Dn thymic cortex could be
observed through all their thickness but was more remarkable in areas
close to the connective septi (Fig. 2C, arrows).
|
B). Isolated hyperchromatic
nuclei (arrows) could also be seen homogenously distributed in the
vicinity of Hassall’s corpuscles (*) and perivascular spaces (Fig. 3C). In conclusion, the majority of apoptotic thymocytes in
situ were detected at the corticomedullary junction, although they were
also found scattered throughout the cortex.
The ultrastructural morphology of the cellular organization showed
changes and with evident signs of hypocellularity. Grouped
hyperchromatic nuclei correspond to apoptotic lymphocytes (retracted
hyperchromatic nuclei and degenerative signs in the cytoplasm) and
macrophages with evidence of active phagocytosis (Fig. 4B vs normal thymus in Fig. 4A). In the bosom of large and empty spaces, epithelial
cells showed numerous filiform cytoplasmic processes, and there were
few contacts with normal lymphocytes (Fig. 4B).
|
B). The
loss of cellular contact between both cell populations became more
evident in the deep cortex of Ts65Dn mice, where epithelial cells show
irregular cytoplasm with dendritic processes and no contacts with
neighboring floating lymphocytes (Fig. 5D). The nuclear
membrane as well as the saccules of the endoplasmic reticulum of these
rounded and vacuolated epithelial cells was abnormally enlarged (Fig. 5F vs normal thymus in Fig. 5E).
No dramatic alterations in the microvasculature at the level of the
corticomedullar junction and the medulla itself were observed. However,
cellular retraction, a certain degree of intracellular edema, and
residual apoptotic cells were frequently seen in similar areas of the
trisomic thymus (Fig. 6B). At
the level of the Ts65Dn thymic medulla, the stromal populations showed
little activity, with epithelial cells exhibiting large numbers of
tonofilaments in their prolongations, macrophages had residual
phagocytosed bodies and numerous apoptotic lymphocytes (Fig. 6D).
When Ts65Dn mice were treated in vivo with apoptotic drugs (LPS or DEX), these changes increased accordingly as were expected (not shown). Similar observations were obtained in LPS-treated mice made transgenic for human SOD-1 (28, 29). Older animals (>20 wk of age) were not evaluated because of the thymic involution observed in aged mice.
Phenotypic analysis of thymocytes was performed by two-color flow
cytometry using cells labeled simultaneously with FITC-conjugated
anti-CD4 and PE-conjugated anti-CD8 Abs (Table I). These studies were performed in
thymuses from Ts65Dn mice as well as in control littermates of
different ages. In general, the proportion of
CD4+CD8+ subpopulation was
clearly diminished in trisomic mice at both 5–7 and 12–15 wk of age.
By three-color cytometry, the larger population of apoptotic thymocytes
corresponded to the
CD4+CD8+ phenotype (see
below and Fig. 9).
|
|
It is well known that both Gram-negative and Gram-positive bacteria induce thymic atrophy via apoptosis in mice (30, 31). LPS of Gram-negative bacteria has been reported to be a major factor in the pathogenesis of Gram-negative septic shock (32). In vivo injection of LPS also induces loss of thymic weight and a decrease in thymic lymphocytes (33). On the other hand, it is also well established that immature murine thymocytes undergo apoptosis in response to DEX (34). Specifically, apoptosis in both control and trisomic thymocytes was evaluated after i.p. injection of LPS and the synthetic glucocorticoid agonist, DEX.
Thymuses from 8-wk-old control mice were removed 12 and 24 h after
i.p. injection of various doses of LPS (0–50 µg) or DEX (0–2 mg).
Thymic cells were collected and cultured for up to 24 h. Cell
viability was recorded, and the appearance of fragmented DNA was
examined by gel electrophoresis. There was a dose-dependent increase in
fragmented DNA by injection with more than 20 mg LPS or 500 mg DEX.
Maximal DNA fragmentation was obtained with mice injected with 50 µg
and 2 mg LPS and DEX, respectively (these amounts were chosen for
additional experiments). A typical experiment is shown in Fig. 7. Both agents accelerated the induction
of apoptosis in thymuses of both euploid and trisomic mice, but this
effect was, as expected, specially significant in the latter. Apoptosis
was demonstrated by counting viable thymocytes (Fig. 7, left) as well as by visualization of fragmented DNA (Fig. 7, right).
|
). Thus, thymocytes from
both control and Ts65Dn mice were phenotypically evaluated by flow
cytometry. As is described below, the apoptotic subset
(TUNEL+ cells) was that corresponding to the
CD4+CD8+ phenotype in both
populations. Thymocytes of Ts65Dn mice are more susceptible to DEX-induced PCD in vitro
Glucocorticoid hormones cause rapid depletion of thymocytes by apoptosis in vivo as well as in vitro (34). However, LPS is a well-established PCD inducer agent in vivo only (33). Because glucocorticoid and bacterial products appear to induce thymocyte PCD by different pathways, the sensitivity to DEX- and LPS-induced cell death was evaluated in thymocyte cultures.
To evaluate thymocyte apoptosis sensitivity of Ts65Dn mice to
glucocorticoid-induced and LPS-induced DNA fragmentation, thymocytes
were treated with DEX (1 µM) or LPS (50 mg/ml) in culture for up to
18 h. In Fig. 8
(left), the results of a typical experiment are
shown. Untreated thymocytes from Ts65Dn showed a higher proportion of
apoptotic cells. DEX increased the percentage of apoptotic cells in
both thymocyte populations, being significantly higher in Ts65Dn cells,
indicating an enhanced sensitivity of these trisomic cells to undergo
DEX-driven PCD (Fig. 8, lower left) Flow cytometric analysis
revealed that DEX-treated thymocytes displayed a
sub-G0 DNA peak, which is known to correspond to
fragmented DNA in thymocytes (35). As expected, no
significant differences were observed between LPS-treated and untreated
thymocytes from both mice (Fig. 8, upper left). In Fig. 8
(right), the DNA fragmentation of both DEX-treated and
untreated cells is illustrated.
|
). Again, quantitative analysis showed
that the subset of TUNEL-positive cells corresponded to those stained
simultaneously for both CD4 and CD8 molecules. Bcl-2 expression in thymocytes
An important factor in shaping the TCR repertoire during thymocyte development is the susceptibility of double-positive (CD4+CD8+) thymocytes to enter in a program of cell death (negative selection) when the TCR is engaged by self-Ags. Recent evidence has suggested that this susceptibility to PCD might be influenced by the expression of Bcl-2 (see Ref. 36 for a review). For this reason, we considered whether or not Ts65Dn thymocyte cell death could be related to Bcl-2 expression. Approximately 28% of normal thymocytes expressed Bcl-2. A similar percentage occurred in thymocytes from Ts65Dn mice. Bcl-2 was expressed by nearly all CD4+CD8- and CD4-CD8+ subpopulations in both control and Ts65Dn mice, but by only 5–7% of CD4+CD8+ thymocytes in both mice. In the CD4-CD8- subpopulation, 60–65% of thymocytes were Bcl-2+ (not shown). Moreover, Bcl-2 was found to be expressed throughout the thymic medulla, but only by scattered cells in the thymic cortex (not shown).
SOD-1 levels and generation and kinetics of ROI generation by Ts65Dn thymocytes
Ts65Dn mice overexpress the homologues of multiple genes located
on human chromosome 21 (5, 6). We have measured SOD-1
protein levels in several tissues of Ts65Dn mice. As expected, SOD-1
sp. act. in thymus of trisomic mices was 
1.8 times control values
(2.7 ± 0.3 vs 1.48 ± 0.3 SOD525 U/ml
in Ts65Dn and euploid mice, respectively).
The high SOD-1 levels in tissues of Ts65Dn mice may lead to increased production of H2O2 and thus to oxidative stress; and a higher susceptibility of the Ts65Dn mouse thymocytes to PCD. As a preliminary approach we correlated SOD-1 levels with ROI and RNI production in activated macrophages (in these cells oxygen consumption can be easily measured). The steady-state level of oxygen consumption depends on the level of NADPH oxidoreductase activity and on the rate at which O2- and H2O2 are utilized in the cell. Thus, an increased intracellular amount of SOD-1 in cells dismutates a higher amount of O2- to O2. This was observed in macrophages from trisomic mice, which released 4 times less O2- than control cells (7 ± 2 vs 30 ± 4 nmol/mg in Ts65Dn and euploid cells, respectively). It would be reasonable to expect that overexpression of SOD-1 would also lead to the accumulation of higher concentrations of H2O2. Indeed, Ts65Dn macrophages released 3 times more H2O2 than macrophages from control littermates (94 ± 14 vs 29 ± 6 nmol/mg in Ts65Dn and euploid cells, respectively). RNI production was also evaluated in macrophages. There was no significant difference in terms of NO2- production up to 1 µg/ml LPS (8.7 ± 1 vs 10 ± 1 nmol/105 cells in euploid and Ts65Dn cells, respectively).
To examine whether the production of intracellular
H2O2 is increased in
thymocytes from Ts65Dn mice, the level of intracellular peroxides in
the thymocytes was determined. This was studied in vivo in thymocytes
from mice injected or not with LPS or DEX and in vitro by culturing
thymocytes with these apoptosis-triggering agents. ROI production is
difficult to measure in thymocytes by conventional procedures. For this
reason, intracellular H2O2
levels were assessed using the sensitive DCFH-DA probe
(13): Cells from Ts65Dn and control mice before and after
in vivo injection of LPS (50 µg) or DEX (1 mg) were treated with
DCFH-DA at different periods, and fluorescent DCF was analyzed by
flow cytometry as described in Materials and Methods.
Fluorescence values of thymocytes from PBS-injected Ts65Dn mice were
always higher than in control euploid cells, although these differences
were not significant (Fig. 10). On the
other hand, statistically significant differences were observed in the
fluorescence values between control and Ts65Dn thymocytes from animals
injected with LPS (Fig. 10A). Similar results were obtained
when H2O2 generation was
recorded in thymocytes from DEX-injected mice (Fig. 10C). or
when quantitated in thymocytes treated with DEX in vitro (Fig. 10D). No differences in
H2O2 production were
recorded between Ts65Dn and euploid thymocytes, neither when cultured
in medium alone nor after LPS addition (Fig. 10B). These
results, together with those of susceptibility to apoptosis, indicate
that: 1) thymocytes from Ts65Dn mice produce higher levels of peroxides
in comparison with euploid mice; 2) this difference is enhanced by
treatment with DEX both in vivo and in vitro; 3) LPS only increase this
effect in vivo; and 4) these observations matched very well with
susceptibility to apoptosis by these agents in vivo and in vitro.
|
Lipid peroxidation is an autocatalytic free radical chain reaction
stimulated by the highly reactive hydroxyl radical
·OH (37). Certain
membrane-associated structures can be considered targets of
·OH-induced peroxidative damage: mitochondria,
cytoplasmic membranes and lysosomes (for a review, see Ref.
37). To correlate apoptosis by LPS and DEX with the
effects of these substances on the oxidative burst, we quantitated
total lipid peroxides produced by cells treated or not with these
apoptosis-inducing reagents. Results of two experiments performed
indicated: first, that treatment with increasing amounts of DEX
generates detectable and significant levels of peroxides in a
dose-dependent manner in thymocyte extracts, whereas untreated cells do
not form lipid peroxides (Table II). In
addition, because of several problems and the lack of specificity of
many available methods for detecting lipid peroxidation (discussed in
Ref. 38), 4-HNE and MDA, two major products of
6-polyunsaturated fatty acids, were also measured (Table II). Almost
identical results were obtained. Production of 4-HNE and MDA followed a
pattern similar to that of total LPO and was DEX induced. This effect,
although to a lesser extent, was also observed with LPS. MDA and 4-HNE
are only two of a great number of carbonyl compounds formed and often
represent only a small percentage of the total products formed, so
quantitation of total lipid peroxides largely exceeded the quantitated
4-HNE and MDA together.
|
According to the results generated, oxidative stress probably
contributed to enhanced apoptosis in the thymus of Ts65Dn mice. To
further study the contribution of both ROI and RNI generation in
thymocyte apoptosis, Ts65Dn and control thymocytes, treated or not with
DEX, were cultured in the presence or absence of a large array of
scavengers and inhibitors and dead cell counted as an estimate of
apoptosis (Fig. 11). Radical scavengers
can be used to abolish intracellular redox processes in intact cell
preparations or to block a defined molecule. Some inhibitors which were
generally effective at low concentrations and nontoxic for intact cells
were selected. DDC (at 5 µM), NAC (1–10 mM), vitamin E (only the 250
µM concentration was tested), N-propyl gallate (at 5
µM), and mannitol (5–20 mM) inhibited the DEX-induced apoptosis in
Ts65Dn thymocytes (this effect was to some extent also observed in
euploid thymocytes except for mannitol). Only marginal inhibition was
observed with DMTU (10–100 µM). Similarly, quercetin (up to 100
µM) and low concentrations of cysteamine (1 µM) did not protect.
Similarly, antioxidants enzymes, catalase, or SOD (at 100 and 250
µg/ml, respectively) did not affect the percentages of apoptotic
thymocytes. Finally, the addition of aminoguanidine (an inducible NO
synthase inhibitor) had no effect, probably indicating that NO
production was not involved in DEX-induced thymocyte apoptosis.
|
. It seems that NAC and DDC were able
to protect thymocytes from PCD and that this in vivo effect is
equivalent to that observed in vitro.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
PCD plays a crucial role in defining the T cell repertoire during T cell development in the thymus. Several studies in humans (42, 43) and mice (42, 44), concluded that deletion occurs at the CD4+CD8+ stage, that both CD4+CD8+ and CD4-CD8- thymocytes die in situ via a PCD process, and that expression of Bcl-2 does not protect immature thymocytes from PCD. In Ts65Dn thymuses, the architecture is maintained. However, TUNEL staining demonstrated increased apoptosis in both cortex and medulla of such thymuses. Although apoptosis was always increased in Ts65Dn thymuses in comparison with age-matched normal controls, percentages of single- and double-positive and -negative subpopulations were maintained in constant ratios.
Thymocyte subpopulations were screened for expression of Bcl-2 in our model, because some reports claimed that Bcl-2 could act as an endogenous antioxidant (45), and it seems that Bcl-2 antagonizes a relatively late step that could be defining an irreversible commitment point (46) However, others claimed that Bcl-2 function is an oxidative-independent process (47, 48). Within the apoptotic thymocytes, both control and Ts65Dn mice express similar percentages of the apoptotic regulatory protein Bcl-2. In addition, because >75–80% of the apoptotic thymocytes are CD4+CD8+ and <10% of this subpopulation express Bcl-2, our data could indicate that Bcl-2 is down-regulated at the CD4+CD8+ stage of development in both Ts65Dn and control thymocytes (42, 43, 44, 49).
The question that arises when extrapolating in vitro findings (DEX
induces thymocyte apoptosis but LPS does not) to an in vivo situation
(both agents induce apoptosis) is, "What could be the targets of LPS
effects in the context of thymocyte apoptosis?" Available evidence
suggests that the hypothalamic-pituitary-adrenal (HPA) axis,
circulating cytokines, and activated macrophages are directly linked:
1) LPS injection enhanced in vivo production of various cytokines
(i.e., IL-1, TNF-
, IL-6) via macrophage activation (J.-E.
Paz-Miguel, R. Flores, L. López de Rego, and F.
Leyva-Cobián, manuscript in preparation); 2) these cytokines
stimulated the HPA axis, increasing circulating levels of
adrenocorticotrophic hormone and corticosteroids (reviewed in Refs.
50, 51, 52); 3) the rapid return of circulating cytokine
concentrations to basal levels appears to be regulated by negative
glucocorticoid feedback; 4) on the other hand, it is well established
that glucocorticoid hormones modulate T cell maturation in vivo.; 5) it
seems that a precise balance of hormone concentration is crucial for T
cell homeostasis in the thymus: low levels of hormones are required for
appropriate T cell development, whereas higher glucocorticoid
concentrations allow immature T cells to undergo increased apoptosis
during systemic stress (53); and 6) it is considered that
the glucocorticoids induce apoptosis in thymocytes by binding to a
specific receptor belonging to the erbA oncogene-related
steroid hormone receptor superfamily (54) and at least two
glucocorticoid-inducible genes (they encode two receptors that function
as ATP- and inositol 1,4,5-triphosphate-gated calcium channels) have
been implicated in thymocyte apoptosis (reviewed in Ref.
55).
Also, homeostatic control by the HPA axis is compromised with aging and in DS, leading to an increase in plasma adrenocorticotrophic hormone and corticosteroid levels (1, 52). This situation is theoretically operative in Ts65Dn mice. Probably under pathological situations (such as bacterial and viral infections), HPA axis-linked thymus processes (e.g., PCD) could be exacerbated (as reported here by the experimental administration of a bacterial product, LPS, or a corticosteroid analogue, DEX). The absence of these complex influences on cultured thymocytes allows us to explain the virtual absence of LPS-induced apoptosis in vitro.
During oxidative metabolism, harmful ROI are generated. These species are physiologically neutralized by antioxidant enzymes. First, SOD converts O2- to H2O2. Thereafter, catalase and glutathione peroxidase independently convert this to H2O (catalase, besides this activity, oxidizes formic acid and formate (56, 57)). However, the existence of increased SOD-1 activity in trisomic cells produce higher amounts of H2O2, and because this molecule is highly diffusible and convertible (via metal ion-driven Haber-Weiss or Fenton reactions) to the most highly reactive ·OH radical, oxidative damage and further lipid peroxidation of cell structures occur (28).
Several ROI scavengers and inhibitors such as NAC, N-propyl gallate, and DDC significantly enhanced in vitro thymocyte survival from Ts65Dn mice and normal controls. NAC and DDC were also able to impair in vivo DEX-induced thymic apoptosis. Theoretically, the antioxidant effects of NAC could be exerted at two levels: 1) by direct reduction of ROI; and 2) by deacetylation, forming cysteine and consequently increasing glutathione production (12). This, in turn, acts as a substrate in the reduction of H2O2, catalyzed by glutathione peroxidase, to be converted to H2O. Finally, glutathione is an endogenous antioxidant that scavenges ·OH radicals. However, DMTU, cysteamine, mannitol, and quercetin did not contribute to thymocyte rescue from PCD in Ts65Dn thymocytes as well as those from normal mice. DMTU reacts with H2O2 and HOCl (58) and although it could have contributed to the rescue of cells from PCD, the possibility that it may exert other biochemical effects cannot be ruled out. On the other hand, the observation that low concentrations of cysteamine do not protect against apoptosis was not surprising. It could indicate that processes where O2- participate are not dependent on triggering apoptosis. Mannitol is also an active scavenger of ·OH radicals (12) but does not protect thymocytes. Similarly, quercetin, a competitive inhibitor of the ATP-binding site of phosphatidylinositol-3-kinase, does not inhibit H2O2 production although it inhibits O2- formation (12). Finally, catalase and SOD had no effect, probably because they could not easily diffuse into the cells. The conclusion derived from studies with inhibitors is that results must be interpreted with caution, because the specificity of these scavengers is not absolute and varies according to cell type and apoptotic agent. For example: 1) staurosporine-induced apoptosis in cortical neurons from DS is not reversed by antioxidants (59); 2) NAC does not protect 7-M12 myeloid leukemic cells (which show a high intrinsic level of H2O2 production) from cycloheximide-induced apoptosis (60); 3) antioxidants show dual effects on oxidative-linked processing of proteins (12); and 4) catalase and SOD did not affect intracellular ROI in rat thymocytes, unlike NAC (61). Finally, it has been recently shown that ROI regulate signals involved in caspase activation and apoptosis (62). These authors elegantly demonstrated that T cells are rescued from O2- generation and cell death with a SOD mimetic, Mn(III)-tetrakis(5,10,15,20-benzoic acid)porphyrin (62). Our results support this observation.
NO has also been associated with PCD of fresh thymocytes (63). Although the role of RNI in our model was not specifically addressed in this paper, it is probable that they do not contribute significantly to thymocyte apoptosis in Ts65Dn mice because of the following indirect observations: 1) NO2- concentration in peritoneal macrophages from Ts65Dn mice and controls was similar; and 2) aminoguanidine (an inducible NO synthase inhibitor) did not protect thymocytes from DEX-induced apoptosis. In addition, it has been reported that S-methylisothiourea (another inhibitor of NO production) had no effect on the PCD of activated T cells (62).
Many studies have addressed the toxic role of
·OH on mitochondria, cytoplasmic membranes, and
lysosomes (38). It is probable that in our model,
H2O2 could be participating
in free radical formation, allowing the NADPH-dependent peroxidation to
occur in subcellular structures through many possible interactions
operating sequentially or simultaneously (37, 64, 65, 66, 67).
These mechanisms can induce protein and lipid peroxidation of key
structural targets. 4-HNE and MDA are produced as major products of the
peroxidative decomposition of 6-polyunsaturated fatty acids, the
former being much more toxic than the latter (see Ref. 38
for a review). Generation of lipid peroxidation by-products clearly
correlates with our observation that thymocyte apoptosis is induced by
these compounds. Lipid peroxidation in thymocytes from Ts65Dn mice was
greater than that in normal thymocytes, and DEX plays a collaborative
role in vitro in the generation of harmful oxygen derivatives. As
expected, DEX stimulates lipid peroxidation more than LPS. In
consequence, lipid peroxidation could lead to increased membrane
rigidity and other alterations in subcellular membranes that could
contribute to apoptotic mechanisms. These observations agree with those
indicating ROI generation also correlates with enhanced thymocyte
apoptosis in the Ts65Dn model.
Despite the fact that this issue is controversial, it is being
generally accepted that mitochondria exert a decisive role in PCD,
mainly for the following reasons: 1) inhibition of caspases does not
always prevent irreversible apoptotic changes (62, 68); 2)
PCD can be induced in cytoplasts (68); and 3) caspase
inhibitors block DNA degradation but not PCD in activated T cells
(62). Consequently, at least two pathways can operate
simultaneously, separately, or sequentially to converge in irreversible
T cell (and probably in cells from other lineages) death
(62): 1) a direct caspase-dependent activation (e.g.,
through TNF- or Fas signaling) and ROI-independent route; and 2) a
ROI-dependent route together with a caspase-dependent activation needed
for DNA degradation but not for PCD.
Our data suggest that ·OH plays a role in thymocyte apoptosis. These results also strongly indicate that ROI derived from H2O2 mediates the degeneration of thymocytes in Ts65Dn mice. These observations as well as those reporting severe early thymic involution in SOD-1 transgenic mice (28, 29) stress the relationship between apoptosis in both T cells and those of the thymic microenvironment and intracellular redox balance. However, their main molecular targets and downstream pathways are still undefined. In addition, Ts65Dn mice represent an interesting model with which to study regulatory pathways in the thymus in relation to: 1) the aging processes in general; 2) DS pathogenesis in particular; and 3) certain immunopathological processes (e.g., autoimmune disorders and leukemogenesis) associated with these conditions. Such studies are currently under way.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Francisco Leyva-Cobián, Servicio de Inmunología, Hospital Universitario "Marqués de Valdecilla," Instituto Nacional de la Salud, 39008-Santander, Spain. E-mail address:
3 Abbreviations used in this paper: DS, Down’s syndrome; DCFH-DA, 2',7'-dichlorofluorescein diacetate; DEX, dexamethasone; PCD, programmed cell death; RNI, reactive nitrogen intermediates; ROI, reactive oxygen intermediates; SOD, superoxide dismutase; DMTU, 1,3-dimethyl-2-thiourea; DDC, diethyldithiocarbamic acid; NAC, N-acetyl-L-cysteine; NGS, normal goat serum; DCF, 2',7'-dichlorofluorescein; LPO, lipid peroxides; MDA, malonaldehyde; 4-HNE, 4-hydroxy-2(E)-nonenal; PI, propidium iodide; HPA, hypothalamic-pituitary-adrenal.
Received for publication June 21, 1999. Accepted for publication September 7, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
factor and IFN- expression in human thymus: localization and overexpression in Down’s syndrome (trisomy 21). J. Immunol. 149:2506.[Abstract]
