Mechanisms of Ageing and Development 96 (1997) 59 – 73 Premature thymic involution, observed at the ultrastructural level, in two lineages of human-SOD-1 transgenic mice B. Nabarra a,*, M. Casanova b, D. Paris b, E. Paly b, K. Toyoma b, I. Ceballos b, J. London b a b U 345 INSERM, Institut Necker, 156 rue de Vaugirard, 75730 Paris Cedex 15, France URA 1335 CNRS, Institut Necker, 156 rue de Vaugirard, 75730 Paris Cedex 15, France Received 7 November 1996 Abstract The human Cu/Zn superoxide dismutase (hSOD-1) gene, catalyses the dismutation of O2 to H2O2 and O2. It is located on chromosome 21 in q22.1 and is overexpressed in Down’s syndrome (DS) patients. These patients present various abnormalities including mental retardation, congenital heart disease, immunological deficits and premature aging. In order to explore the potential role of SOD-1 overexpression in DS, we have generated two lineages of transgenic mice for the hSOD-1 gene and studied, at the ultrastructural level, the effect of hSOD-1 overexpression on the thymic microenvironment. Modification of the cellular architecture and morphology associated with a lipidic invasion, signs of a premature involution of the thymus, were observed in both lineages. A rupture of the filamentous network in the extracellular and probably also in the intracellular matrix was first observed. These results correlate the thymic alterations vizualized in light microscopy, on the thymus from DS patients, and raise the question of the relationship between the SOD-1 overexpression and the different morphological alterations associated with the premature thymic involution observed in SOD-1 transgenic mice. They suggest that thymic and immunological impairments present in DS patients may be related to the SOD-1 gene dosage effect. © 1997 Elsevier Science Ireland Ltd. * Corresponding author. Present address: Clinique Nephrologique, Hopital Necker, 161 rue de Sevres, ´ ˆ ` 75743 Paris Cedex 15, France. 0047-6374/97/$17.00 © 1997 Elsevier Science Ireland Ltd. All rights reserved. PII S 0 0 4 7 - 6 3 7 4 ( 9 7 ) 0 1 8 9 2 - 7 60 B. Nabarra et al. / Mechanisms of Ageing and De6elopment 96 (1997) 59–73 Keywords: hSOD-1 Transgenic mice; Ultrastructural study; Premature thymic involution; Down’s syndrome 1. Introduction Thymic damages (morphological, physiological, biochemical) associated with aging, in human as well as in mice, are gathered under the denomination of involution. This phenomenon corresponds to a suitable background involving morphological irreversible alterations of the thymic microenvironment followed by the general decline of immunocompetent functions. Indeed, thymic involution occurs before the decline in the ability of the immune system and may thus be responsible for this effect. Thus, the thymus has been considered to be implicated as a ‘clock’ in the aging process of the immune system and of the general biological functions of the organism [1 –9]. It is well known that oxygen free radicals (OFR) are involved in protein modifications and membrane alterations [10,11]. Cu/Zn superoxide dismutase (SOD-1), which catalyses the dismutation of superoxide radicals (O2−) into oxygen and hydrogen peroxides (H2O2), is a key enzyme of ORF metabolism [12]. SOD-1 is also able to catalyse surrogate reactions such as the production of OH− [13] and the nitrosylation of the tyrosine residues, present in proteins, by peroxynitrite [14]. Several reports have shown that an increase in Cu/Zn SOD activity may cause oxidative damage [15,16]. Down’s syndrome (DS), the most frequent chromosomal abnormality in man, results from the presence of three copies of chromosome 21. The SOD-1 gene located in 21q22.1 is overexpressed by a factor of 1.5 in DS patients [17,18]. Apart from symptoms including impairment in brain function and skeletal abnormalities, DS patients show many signs of premature aging [19–23] including thymic premature involution observed in light microscopy [24] and various types of impairment of T and B cell distribution and functions [25–28]. In an effort to explore the role of SOD-1 overexpression in DS patients, transgenic mice for the human Cu/Zn SOD gene (hSOD-1) have been bred in several laboratories [29 – 33] and analyzed for physiological consequences [34–37]. However, the general immunological status of these mice has not been investigated except in two recent studies on some immunological aspects of the thymus and bone marrow [38,39]. The primordial role of the thymus in the differentiation selection and maturation of T cells and the necessity of an intact thymic microenvironment for the acquisition and the maintenance of immune functions [40–43] is well documented. Thus, it appeared interesting to perform thymic morphological studies on these hSOD-1 transgenic mice, as a potential model of DS immunodeficiency. Two different transgenic lines for the hSOD-1 gene have been bred in our laboratory, one on the hybrid C57/B6-DBA/2 background (B6/D2) [30,31] and another on the inbred FVB/N background [32,33]. B. Nabarra et al. / Mechanisms of Ageing and De6elopment 96 (1997) 59–73 61 The thymic aspect of these hSOD-1 transgenic mice has been studied by electron microscopy, the best technique for a true identification of cellular modifications in most tissues. The morphological aspect of the thymus from transgenic mice as well as controls have been compared at different ages. The controls are either the murine strains used for transgenesis (C57Bl/6, DBA, B6/D2 and FVB/N) which have been extensively studied for their thymic microenvironment morphology [44 – 48] or non transgenic mice from the littermates of transgenic hSOD-1 lineages. In the first hSOD-1 transgenic lineage (PB line) which overexpresses SOD-1 in the thymus by a factor of 1.5, we have previously described a premature thymic involution (as early as 7 – 8 months-old) similar in morphological aspect to that observed in the thymus of control aged mice (18–20 months) [39]. In this report, we described a similar morphology of premature thymic involution in a second hSOD-1 transgenic lineage (KT line) obtained on a FVB/N background, and which overexpresses SOD-1 by a factor of 5 in blood [49] and by a factor of 5.8 in the thymus [50]. The general involution phenomenon is similar in both transgenic strains, but several differences in the onset of the involution process and differences in the morphological aspects of the thymus in both lines is observed. 2. Materials and methods The entire hSOD-1 gene under the control of its own promoter was microinjected into the ovocytes of two murine strains B6/D2 and FVB/N for obtaining hSOD-1 transgenic animals [30 – 33]. The first lineage (named PB line in our nomenclature) has been obtained on a B6/D2 background and has been described previously [30,31]. The animals were shown to have integrated only one copy of the transgene and to have overexpressed the SOD-1 enzyme by a factor of 1.5 in the thymus. The second lineage (named KT line), was obtained by microinjecting zygotes of FVB/N mice with a 11.5 kb EcoRI–BamH1 linear fragment of human genomic DNA containing the entire hSOD-1 gene. The genome of the KT line carries 13/14 copies of the transgene integrated in tandem head to tail at a single autosomal locus [32]. Overexpression of the SOD-1 was evaluated by enzymatic activity using the classic test of the inhibition of nitroblue tetrazolium reduction by SOD-1 and automatically assayed using a Ransod kit (Randox laboratories) and also at the RNA level. The SOD-1 activity was increased by a factor of 5 in the blood and of 5 – 8 in the thymus as a function of age [49,50]. For the ultrastructural studies, two types of controls have been considered. First, the B6/D2 and FVB/N strains used as transgenesis background at different ages (15 days, 2, 3, 5, 8, 10, 12, 14, 18, 20, 25, 30 months) and called ‘naive controls’. Secondly, the thymus of non transgenic mice were obtained from crosses between heterozygous SOD-1 transgenic mice and B6/D2 or FVB/N mice, these control being called ‘transgenesis controls’. 62 B. Nabarra et al. / Mechanisms of Ageing and De6elopment 96 (1997) 59–73 For electron microscopy preparations, a classic technique was used. Immediately after sampling a first fixation of thymic fragments was performed for 1 h in 2% glutaraldehyde and phosphate buffer pH 7.4, followed by a postfixation in 2% osmium tetroxide in the same buffer for 1 h. Dehydration in alcohol series and embedding in Epon follow the fixation. Ultrathin sections impregnated with uranyl acetate and lead citrate were observed under a Philips EM300 electron microscope. For each age, six mice of each strain were studied in the experimental and control groups. Thymuses from 16 days to 12 months have been used for both transgenic lines. For each group of mice, and for each age ten blocks of different thymic areas (cortex and medulla) were analyzed. 3. Experimental results 3.1. Thymic microen6ironment morphology in control mice The thymic morphology observed either in ‘naive’ control mice (C57/Bl6, DBA2, B2/D6, FVB/N) or in ‘transgenesis control’ mice (issued from the same littermate as transgenic animals) was similar according to the age considered. 3.1.1. Young mice The ultrastructural aspect of the normal murine thymic microenvironment has been largely described previously [44–48]. The basic structure of the thymus appeared as a network organization of different stroma cells distributed in two zones and forming a cellular microenvironment enclosing numerous lymphocytes. The cortical area contained one type of epithelial cells (Type I) whereas the medullary area contained three types of epithelial cells morphologically different (type I, II and III) (Fig. 1aFig. 2a and b). These cells associated with macrophages (in both zones) and interdigitated cells (in the cortico–medullary junction and in the medulla) form the thymic microenvironment (Fig. 1bFig. 2c). This organization, surrounded by fibrous capsula, was maintained by fibrous trabeculae carrying vessels and innervations. Some differences in the medullary zone were observed between the two murine strains used in this work for transgenesis. Thus, FVB/N mice seemed to have less thymocytes in the medullary zone than other murine strains. This allows a better observation of the stroma and the identification of an undifferentiated cell which may be a precursor of the epithelial lineage [48]. 3.1.2. In6olution and aged thymuses A detailed description has been previously reported [8]. Briefly, we observed progressive lesions beginning at about 10–12 months characterized by a vacuolisation of a few epithelial cells, scattered in normal zones, which progressively formed large altered zones infiltrated by lipofuscin pigment and small lipidic globules. Then the number of thymocytes decreased and the architecture of the organ was altered with the loss of separation between the cortical and medullary zones. The whole B. Nabarra et al. / Mechanisms of Ageing and De6elopment 96 (1997) 59–73 63 microenvironment appeared disorganized. Many cells lost their morphological characteristics and appeared as undifferentiated cells, with rupture of plasma membranes. They appeared to form a large clear cytoplasmic syncytium, containing Fig. 1. Control: thymic cortical microenvironment in 3 month old normal mouse. (a) Type I epithelial cell showing its morphological characteristics (arrow heads: few ‘clear intracytoplasmic vacuoles’ containing dense granulations) surrounded by numerous lymphocytes. (b) Macrophage with numerous dense, homogeneous primary and secondary lysosomes. E: type I epithelial cell, L: lymphocyte. (a) × 8000; (b) × 15 000. 64 B. Nabarra et al. / Mechanisms of Ageing and De6elopment 96 (1997) 59–73 Fig. 2. Control: thymic medullary microenvironment in 5 month-old normal mouse. (a) Type II epithelial cell with its morphological characteristic: an intracytoplasmic ‘alveolar’ network (arrow heads). (b) Type III epithelial cell showing a large intracytoplasmic, ciliated cavity (arrows heads). (c) Interdigitated cell with clear cytoplasm and numerous compressed digitations on the peripheral area (arrow heads). The section does not involve the nucleus. (a) ×16 500; (b) × 18 000; (c) ×7500. few cellular organelles and lipidic globules and to engulf the few remaining thymocytes (Fig. 3). In the terminal stage of involution, at around 18–20 months, the thymus was quite reduced in cell number and in weight and was practically B. Nabarra et al. / Mechanisms of Ageing and De6elopment 96 (1997) 59–73 65 transformed into a mass of adipose tissue containing scattered islands of more or less normal parenchyma in small cellular nests and/or bordering cystic cavities and very few lymphocytes. Fig. 3. Control: involuted thymic microenvironment in 16 month-old mice. (a) Epithelial cell (arrow heads: tonofilaments) showing disrupted part of limiting membrane ( ) and a clarified cytoplasm. Few thymocytes, plasma cell ( [ ) and lipofuscin pigment (Lip.) are present. (b) Disorganized cytoplasmic clear layer with a few scattered cellular organites and few thymocytes engolfed in thymus of 20 month-old mice. Large lipidic globules (arrow heads). (a, b) × 7500. 66 B. Nabarra et al. / Mechanisms of Ageing and De6elopment 96 (1997) 59–73 Differences in the beginning of the involution exist between the FVB/N mice and the other murine strains considered here. Indeed, in C57/Bl6, DBA2 and B6/D2 mice, the morphological involution of the thymus starts at about 10–12 months while the more complete involuted thymus can be visualized at about 18–20 months. In FVB/N mice, the involuted process appeared at about 7–8 months and was complete by 12 months. In spite of this difference, the evolutive steps of the microenvironmental morphology modifications and the final picture were similar in FVB/N and other murine strains, and correspond to the previous description of the aged thymus [8]. 3.2. Thymic microen6ironment morphology in transgenic mice o6erexpressing hSOD-1 Thymuses from both lineages of hSOD-1 transgenic mice exhibit a premature involution in comparison with controls. The main difference between the two types of transgenic mice was the onset of the premature involuting process. In the PB transgenic line (bred on the hybrid B6/D2 background and overexpressing SOD-1 by a factor of 1.5 in the thymus), the microenvironment involution started at about 7 – 8 months [44] whereas in the KT transgenic line (bred on the inbred FVB/N background and overexpressing SOD-1 by a factor of 5–8 in the thymus), these morphological modifications appeared earlier, at 4–5 months. The architectural modifications as well as the cellular alterations were the same at the final stage of the involution process in both transgenic lineages and corresponded to the pictures observed in control aged thymus. The primary event was a large vacuolisation of epithelial cells (Fig. 4a) followed by an invasion of the epithelial cells by a few lipofuscin and lipidic globules. Thymuses from 2 – 3 month-old transgenic mice exhibited a peculiar aspect in their cellular network. The architectural organization of the thymus in two zones was still conserved, but in some areas, the microenvironment network appeared to be completely retracted and often broken. The lymphocytes were completely separated from the epithelial cells as if they were ‘floating’ in large, clear, empty ‘holes’. Numerous cytoplasmic finger-like processes which seemed to originate from the epithelial cells were observed (Fig. 4b). This aspect of epithelial network retraction appeared less extended in the PB transgenic line than in the KT line. Around 3 – 4 months, hSOD-1 transgenic mice presented thymic areas modified in different ways: some of these areas displayed ‘holes’ in the stroma, some vacuolized epithelial cells and others areas presented a cellular involuted aspect with a clarified cytoplasm and few clusters of dense lipofuscin material or small lipidic globules (Fig. 5a). Finally few areas showed normal cells. Few adipocytes invaded the capsule and fibrous septae. At 5 months, in the KT transgenic mice, and at 7 months in the PB mice, the architecture was already modified and the separation in cortical and medulla zones was mostly blurred. The epithelial cells had lost part of their limiting membranes and formed a continuous layer of clear poorly differentiated cytoplasm with few organelles and with inclusions of lipids and lipofuscin. A small number of lymphocytes were present in this cytoplasmic layer (Fig. 5). B. Nabarra et al. / Mechanisms of Ageing and De6elopment 96 (1997) 59–73 67 Fig. 4. Premature involution in the thymus of 3 month-old hSOD-1 transgenic mice. (a) Type I epithelial cell showing cytoplasm with numerous vacuoles and dilated perinuclear endoplasmic reticulum. (b) Aspect of ‘hole’in the thymic stroma with disrupted epithelial cells (arrow heads), numerous finger-like processes ( ) and lymphocytes. (a) ×9500; (b) ×8000. Alterations similar to those usually present in the involuted thymus of ‘naive’ mice have been observed in the PB transgenic mice from the age of 7–8 months and from 5 – 6 months in the KT line. However, the lipidic invasion, was the same in the KT mice and in aged control mice, and was less extant in the PB mice. 68 B. Nabarra et al. / Mechanisms of Ageing and De6elopment 96 (1997) 59–73 4. Discussion Thymic involution is a normal process which occurs during aging. Electron Fig. 5. Premature involution in the thymus of hSOD-1 transgenic mice. The reticulum appears as a large, poorly differentiated, unlimited cytoplasmic layer containing lipofuscin material and small lipidic globules (arrow heads). Few resting lymphocytes are engulfed in this cytoplasmic syncitium. (a) 8 month-old transgenic mouse (PB lineage) × 7500; (b) 6 month-old transgenic mouse (KT lineage) × 7500. B. Nabarra et al. / Mechanisms of Ageing and De6elopment 96 (1997) 59–73 69 microscopy is a powerful tool to identify the numerous cellular modifications occurring from the onset through the various stages of the involution process. The most striking ultrastructural morphological changes observed in the aging thymus ([8] and references in this publication) are related to the distribution, architectural arrangement, and cytologic appearance of the different cellular components of the thymic microenvironment. After a stage of vacuolisation, epithelial cells lose their limiting cellular membranes and a large, clear, poorly differentiated cellular syncytium is formed, followed by a progressive replacement of this thymic tissue by adipose tissue. At the terminal stage, lipidic invasion by fat cells with enormous globules appeared surrounding few areas of epithelial cell islands organized in small nests and bordering cystic cavities [8]. In this thymic morphological study from two different transgenic lines expressing the hSOD-1 gene, ultrastructural modifications in comparison with various control thymus is observed. They showed an aspect which appeared, for a great part, similar to that observed in aged mice. These modifications are specific of the hSOD-1 transgenic mice since they are not present in different controls of the same age and are observed in the two transgenic lines obtained with the same construct but on different murine backgrounds with different levels of both transgene copy numbers and SOD-1 overexpression activity. Three important differences are visualized in the hSOD-1 transgenic mice compared to the aged control mice: first, the beginning of the thymic involution occurred around 7 – 8 months in the PB transgenic line and around 4–5 months in the KT one. A second difference has to do with the rupture of the cellular and extracellular filamentous network and the formation of ‘holes’ by retraction of the cellular network which formed the stroma. The cellular network appeared retracted with broken epithelium leaving large ‘holes’ containing fragments of cytoplasm in finger-like cellular protrusions scattered in empty zones. These alterations, not observed in aged controls, appeared early in the thymic involution process of the transgenic mice and are then followed by more ‘classic’ involution modifications. This aspect is rarely described and perhaps corresponds to the description, in light microscopy, of the empty areas in diabetic rodents strains [51,52]. We suggest that such morphological alterations are perhaps due to the rupture of the cytoskeleton network of epithelial cells and extracellular matrix as a consequence of a SOD-1 overexpression which is more drastic in the KT transgenic line than in the PB transgenic line. It is interesting to correlate these observations on thymic involution in hSOD-1 transgenic mice with the modifications in the extracellular matrix and cellular microfilament network in DS patients thymus. Indeed, in observation of DS thymus morphology by immunofluorescence [53], a rupture and retraction of the cellular skeleton are also reported. The third difference concerns the lipidic invasion always observed in the involuted thymic microenvironment. Contrary to the very large lipidic invasion throughout the thymic tissue in aged mice, we noticed in the thymus of PB transgenic mice only a moderate amount of lipofuscin pigment and lipidic globules 70 B. Nabarra et al. / Mechanisms of Ageing and De6elopment 96 (1997) 59–73 [39]. However in the thymus of the KT lineage the lipidic invasion is more extent and similar to the thymus of aged control mice. The lipidic invasion of the thymus and the organ transformations are poorly understood. All aging phenomena are associated with increase of lipidic components in the tissues and it is suggested that in the thymus lipidic material may reflect storage of different components after drastic thymocyte lysis occurs in the aged thymus [54,55]. Our observations raise the question of a possible relationship between the overexpression of the hSOD-1 transgene and the premature involution observed in the thymus of these transgenic mice. As our results have been obtained in two different lineages of transgenic mice for the hSOD-1 gene and on two different backgrounds, the thymic involution described here can be correlated with the transgene expression and not with a possible modification of the murine genome at the transgene insertion site. The present results are more or less in agreement with the recent results published by Peled-Karman et al. [38] showing that hSOD-1 transgenic mice injected with lipopolysaccharide exhibit a loss of thymic architecture and an increase in thymocyte apoptosis. Unfortunately, this study has been performed only with light microscopy which does not allow precise identification of the cellular and chromatin alterations. Nevertheless, it is possible that these observations are related to an involution process. In conclusion, the overexpression of SOD-1 gene appears involved in the process of premature thymic involution observed in the two different transgenic lines, but we cannot overlook the fact that the enzyme effect may be either direct or indirect, as it is known that the SOD-1 gene may have alternatively protective or deleterious functions in different tissues and physiological conditions [56 – 58]. Furthermore, the morphological and cellular modifications observed in the thymus during aging both in control mice and in hSOD-1 transgenic mice, suggest the possible implication of the superoxide dismutase overexpression in the etiology of premature thymus involution and related immunological impairments observed in DS patients [50]. As the thymus is considered as a putative ‘clock’ for numerous physiological phenomena we suggest, in thymus involution and in various more general aging processes, the involvement of the level of SOD-1 or the implication of reactive oxygen species (ROS). Furthermore the balance between the levels of the main enzymes involved in ROS elimination (Cu/Zn SOD, Mn/SOD, catalase and glutathione peroxydase) might also be very important. Further studies on the expression of these enzymes in the different areas of thymus should give more information for a better understanding of SOD-1 transgenic mice as a model for DS and suggest the necessity of similar approaches to understand the general mechanism of aging in DS patients. B. Nabarra et al. / Mechanisms of Ageing and De6elopment 96 (1997) 59–73 71 Acknowledgements We wish to thank I. Andrianarison and A. Nicol for their technical work, T. Sacksick and P. Lemoine for their photographic work, and Sonia Hamon for help in preparing the manuscript. References [1] K. Hirokawa, The thymus and aging. In T. Makinodan and E. Yunis (eds.), Immunology and Aging, Plenum, New York 1977, pp. 51 – 72. [2] M. Kay and T. Makinodan, Relationship between aging and the immune system. Prog. Allergy, 29 (1981) 134–181. [3] T. Manikodan, Immunity and ageing. In C. Finck and L. Hayflick (eds.), Handbook of the Biology of Aging, Van Nostrand Reinhold, New York, 1977, pp. 379 – 408. [4] H. Schuurman, L. Nagelkerken, R. deWeger and J. Rozing, Age-associated involution: significance of a physiological process. Thymus, 18, (1992) 1 – 56. [5] G. Steinmann, Changes in the human thymus during aging. In H.K. Muller – Hermelink (ed.), Human Thymus, Springer Verlag, Berlin, 1986, pp. 43 – 89. [6] A. Clarke, The many facets of thymic involution. Immunol. Today, 7 (1986) 204 – 207. [7] R. Miller, Aging and immune function. Int. Re6. Cytol., 124 (1991) 187 – 215. [8] B. Nabarra, and I. Andrianarison, Ultrastructural studies of involuted thymic microenvironment in aged mice. Exp. Gerontol., 31 (1995) 489 – 506. [9] A. Georges and M.A. Ritter, Thymic involution with aging. Obsolescence or good housekeeping. Immunol. Today 17 (1996) 267 – 272. [10] B. Halliwell and J.M. Gutteridge, The importance of free radicals and catalytic metal ions in human diseases. Mol. Asp. Med., 8 (1985) 89 – 193. [11] I. Fridovich, In E. Meister (ed.), Ad6ances in Enzymology and Related Areas of Molecular Biology, Wiley, New York 58 1986, pp. 61 – 97. [12] M.B. Yim, P.B. Chock and E.R. Stadman, Copper – Zinc superoxide dismutase catalyses hydroxyl radicals production from hydrogene peroxide. Proc. Natl. Acad. Sci. USA, 87 (1990) 5006 – 5010. [13] J.S. Beckman, M. Carson, C.D. Smith and W.H. Koppenol, ALS, SOD and peronitrite. Nature, 364 (1993) 584–587. [14] Pal Yu Byung, Cellular defenses against damage from reactive oxygen species. Physiol. Re6., 74 (1994) 139–162. [15] P. Amstad, A. Peskin, G. Shah et al., The balance between Cu/Zn superoxide dismutase and catalase affects the sensitivity of mouse epidermal cells to oxidative stress. Biochemistry, 30 (1991) 9305–9313. [16] P.M. Sinet and P. Garber, Inactivation of the human Cu/Zn superoxide dismutase during exposure to O2 and H2O2. Arch Biochem Biophys., 212 (1981) 411 – 416. [17] P.M. Sinet, Metabolism of oxygen derivatives in Down’s syndrome. Ann. NY Acad. Sci., 396 (1982) 83–94. [18] Y. Groner, J. Lieman-Hurwitch and N. Dafni, Molecular structure and expression of the gene locus on chromosome 21 encoding the Cu/Zn superoxide dismutase and its relevance to Down’s syndrome. Ann. NY Acad. Sci., 450 (1985) 133 – 156. [19] S. Pueschel, Clinical aspects of Down’s syndrome from infancy to adult hood. Am. J. Med. Genet. (Suppl.), 7 (1990) 52–56. [20] L. Liss, C. Shim and M. Thase, The relationship between Down’s syndrome and dementia Alzheimer’s type. J. Neuropathol. Exp. Neurol., 39 (1980) 371 – 375. [21] R. Walford, T. Gesset, F. Marim and C. Tam, Immunological and biochemical studies of Down’s syndrome as a model of accelerated aging. In D. Segre and L. Smith (eds.), Immunological Aspects of Aging, Marcel Dekker, New York, 1981, pp. 479 – 484. 72 B. Nabarra et al. / Mechanisms of Ageing and De6elopment 96 (1997) 59–73 [22] Y. Groner, N. Dafni, L. Sherman et al., Down’s syndrome and Alzheimer’s disease. In M. Fisher, A. Hanin and L. Lachman (eds.), Alzheimer’s and Parkinson’s Diseases, Plenum, New York, 1986, pp. 271–285. [23] S. Levin, The immune system and susceptibility to infection in Down’s syndrome. In E. McCoy and I. Epstein (eds.), Oncology and Immunology of Down’s Syndrome, A.R. Liss, New York, 1987, pp. 143–146. [24] C. Benda and G. Strassman, The thymus in mongolism. J. Ment. Defic. Res., 9 (1965) 109 – 117. [25] G.R. Burgio, A. Ugazio, L. Nespoli and R. Maccario, Down syndrome: a model of immunodeficiency. Birth Defects Orig. Artic. Ser., 19 (1983) 325 – 327. [26] A. Bertotto, R. Gerli, G. Fabietti, C.D.F Arcangeli, S. Bianchi and R. Vaccaro, Immunodeficiency in Down’s syndrome. Analysis by OKT monoclonal antibodies. Thymus, 6 (1984) 365 – 367. [27] R. Noble, and R. Warren, Altered T-cell subsets and defective T-cell function in young children with Down’s syndrome. Immunol. In6est., 16 (1987) 371 – 382. [28] M. Murphy, and L.B. Epstein, Down’s syndrome (Trisomy 21) thymus has a decreased proportion of cells expressing high levels of TCR ab and CD3. Clin. Immunol. Immunopathol., 55 (1990) 453–467. [29] C.J. Epstein, K.B. Avraham, M. Lovett et al. Transgenic mice with increased Cu/Zn-superoxyde dismutase activity: an animal model of dosage effects in Down’s syndrome. Proc. Natl. Acad. Sci. USA, 84 (1987) 8044–8048. [30] I. Ceballos-Picot, A. Nicole, M. Clement, J.M. Bourre and P.M. Sinet, Age-related changes in antioxidant enzymes and lipid peroxidation in brains of control and transgenic mice overexpressing copper–zinc superoxide dismutase. Mutat. Res., 275 (1992) 281 – 293. [31] I. Ceballos, A. Nicole, P. Briand et al. Expression of human Cu/Zn superoxide dismutase gene in transgenic mice: model for gene dosage effect in Down’s syndrome. Free Rad. Res. Commun., 12 (1991) 581–589. [32] J. London, D. Paris, K. Toyama, E. Paly, M. Casanova and P.M. Sinet, The use of FVB/N inbred strain to generate and study of transgenic mice expressing high level of Cu/Zn SOD-1. Biol. Cell, 77 (1993) 335 [33] D. Paris, K. Toyama, P.M. Sinet, P. Kamoun and J. London, A comparison between an inbred strain and hybrid lines to generate transgenic mice. Mouse Genome, 93 (1996) 1038 – 1040. [34] P.H. Chan, G.Y. Yang, S.F. Chen, E. Carson and C.J. Epstein, Cold-induced brain aedema and infarction are reduced in transgenic mice overexpressing Cu/Zn superoxide dismutase. Ann. Neurol., 29 (1990) 482–486. [35] S. Przedborski, V. Kostic, V. Jackson-Lewis and A. Naini, Transgenic mice with increased Cu/Zn superoxidase dismutase activity are resistant to N-methyl-4-phenyl-1,2,3,6-tetrahydropyrine induced neurotoxicity. J. Neurosci., 12 (1992) 1658 – 1667. [36] K. Avraham, M. Schickler, R. Sapoznikow, R. Yarom and Y. Groner, Down’s syndrome: abnormal neuromuscular junction in tongue of transgenic mice with levels of human Cu/Zn superoxide dismutase. Cell, 54 (1988) 823 – 829. [37] O. Elroy-Stein, Y. Berstein and Y. Gronen, Overproduction of human Cu/Zn superoxyde dismutase in transfected cells: extenuation of paraquat-mediated cytotoxicity and enhancement of lipid peroxidation. EMBO J., 5 (1986) 615 – 622. [38] M. Peled-Karman, J. Lotem, E. Okson, L. Sachs and L. Groner, Thymic abnormalities and enhanced apoptosis of thymocytes and bone marrow cells in transgenic mice overexpressing Cu/Zn superoxidase dismutase. EMBO J., 14 (1995) 4985 – 4993. [39] B. Nabarra, M. Casanova, D. Paris et al., Transgenic mice overexpressing the human Cu/Zn-SOD gene: ultrastructural studies of a premature thymic involution model of Down’s syndrome (Trisomy 21). Lab. In6est., 74 (1996) 1 – 10. [40] W. Van Ewijk, T-cell differentiation is influenced by thymic microenvironment. Annu. Re6. Immunol., 9 (1991) 591–615. [41] H. Von Boehmer, The developmental biology of T lymphocytes. Ann. Re6. Immunol., 6 (1988) 306–326. [42] P. Marrack, D. Lo, R. Brunster and R. Palmiter, The effect of thymus microenvironment on T cells development and tolerance. Cell, 53 (1988) 627 – 632. B. Nabarra et al. / Mechanisms of Ageing and De6elopment 96 (1997) 59–73 73 [43] R. Boyd, C.L. Tucek, D. Godfrey and D. Izon, The thymic microenvironment. Immunol. Today, 14 (1993) 445–459. [44] B. Nabarra and I. Andrianarison, Ultrastructural studies of thymic reticulum. I. Epithelial components. Thymus, 9 (1987) 95 – 100. [45] B. Nabarra and I. Andrianarison, Ultrastructural studies of thymic reticulum. II. Non-epithelial components. Thymus, 17 (1991) 39 – 43. [46] B. Nabarra and I. Andrianarison, Ultrastructural studies of thymic reticulum. III. Connective components. Tissue Cells, 27 (1995) 249 – 261. [47] B. Nabarra and M. Papiernik, Phenotype of thymic stromal cells. An immunoelectron microscopy study. Lab. In6est., 48 (1988) 524 – 527. [48] B. Nabarra, Ultrastructural studies of FVB/N thymic microenvironment. Presence of a precursor of medullary epithelial cells. (submitted for publication in Cells and Tissue Research). [49] D. Paris, K. Toyama, A. Magarbane, M. Casanova, P.M. Sinet and J. London, Rapid fluorescence in situ hybridization on interphasic nuclei to discriminate between homozygous and heterozygous transgenic mice. Transgenic Res., 5 (1996) 397 – 403. [50] J. London, D. Paris, B. Nabarra and M. Casanova, Human SOD-1 transgenic mice: a murine model for studying aging and trisomy 21. (submitted for publication in Mechanisms of Ageing and Dev.). [51] J. Rozing, C. Cooken, F. Tielen and J. Weegenaar, Defects in the thymic epithelial stroma of diabetes prone BB rats. Thymus, 14 (1989) 125 – 135. [52] W. Savino, C. Boitard, J.-F. Bach and M. Dardenne, Studies on the thymus in nonobese diabetic mouse. I. Changes in microenvironmental compartments. Lab. In6est., 64 (1991) 405 – 417. [53] E. Fonseca, J. Lannes-Viera, D. Villa-Verde and W. Savino, Thymic extracellular matrix in DS. Bras. J. Med. Biol. Res., 22 (1989) 971 – 974. [54] N. Dourov, Cytoplasmic lipids in thymic involution. Pathol. Eur., 9 (1974) 43 – 46. [55] A. Oksanen, Multilocular fat in thymuses of rat and mice associated with thymus involution. J. Pathol., 105 (1971) 223–227. [56] S. Przedborski, V. Jackson-Lewis, V. Kostic, E. Carlson, C.J. Epstein and J.L. Cadet, Superoxide dismutase, catalase and glutathione peroxidase activities in copper/zinc – superoxide dismutase transgenic mice. J. Neurochem., 58 (1992) 1760 – 1767. [57] G. Rao, E. Xia and A. Richardson, Effect of age on the expression of antioxidant enzymes in male Fischer (F344) rats. Mech. Aging De6., 53 (1990) 49 – 60. [58] E. Kellogg and I. Fridovich, SOD in the rat and mouse as a function of age and longevity. J. Gerontol., 31 (1976) 405–408. .