This paper describes in vitro and in vivo attempts to deplete the 4- to 8-month-old Xenopus laevis (J strain) thymus of its lymphocyte compartment. Gamma irradiation (2–3000 rad) of the excised thymus, followed by two weeks in organ culture, is effective in removing lymphocytes, but causes drastic reduction in size and loss of normal architecture. In contrast, in vivo whole-body irradiation (3000 rad) and subsequent in situ residence for 8–14 days proves successful in providing a lymphocyte-depleted froglet thymus without loss of cortical and medullary zones. In wvo-irradiated thymuses are about half normal size, lack cortical lymphocytes, but still retain some medullary thymocytes; they show no signs of lymphocyte regeneration when subsequently organ cultured for 2 weeks. Light microscopy of 1 μm, plastic-embedded sections and electron microscopy reveal that a range of thymic stromal cell types are retained and that increased numbers of cysts, mucous and myoid cells are found in the thymus following wholebody irradiation. In vivo-irradiated thymuses are therefore suitable for implantation studies exploring the role of thymic stromal cells in tolerance induction of differentiating T lymphocytes.

The transplantation of allogeneic thymus to early-thymectomized Xenopus provides a comparative model system for exploring the role of the thymus in the ‘education’ of host-derived T-lineage lymphocytes (see Du Pasquier & Horton, 1982; Gearing, Cribbin & Horton, 1984; Nagata & Cohen, 1984; Arnall & Horton, 1986). However, in order to explore the role of the thymic stromal cell compartment (see Weiss & Sakai, 1984; Crouse, Turpen & Sharp, 1985), without the complication of influences by donor-derived thymocytes, thymuses need to be depleted of their lymphocyte content, prior to transplantation. In mammals a variety of techniques have been employed to produce a ‘lymphocyte-free’ thymus, with intact stromal elements; these techniques have included in vivo maintenance of thymus in diffusion chambers (Owen & Ritter, 1969), monolayer culture techniques (cited by Jordan & Crouse, 1979), suboptimal culture (Loor & Hägg, 1977; Hong, Schulte-Wissermann, Jarrett-Toth, Horowitz & Manning, 1979) and irradiation (Pritchard & Micklem, 1973; Loor & Hägg, 1977) of postnatal thymus expiants. More recently, low-temperature organ culture alone (Jordan & Crouse, 1979; Jordan, Bentley, Perry & Crouse, 1985), or in combination with high-oxygen environment (Pyke, Bartlett, & Mandel, 1983) has proved effective for the embryonic murine thymus. In vitro exposure of developing thymus rudiments to deoxyguanosine is also successful in depleting lymphocytes and removes other intra-thymic haemopoietic cells (e.g. dendritic cells), while leaving an intact epithelium (Jenkinson, Franchi, Kingston & Owen, 1982).

Both low temperature and deoxyguanosine (DG) treatment of postmetamorphic Xenopus thymus organ cultures have proved ineffective in depleting thymocytes (Russ, 1986). Thymuses cultured at 10°C and 15 °C in fact display an increase in lymphocyte cellularity; this may relate to a reduced mitotic rate, found when Xenopus organs are cultured at lower temperatures (Balls, Simnett & Arthur, 1969), that results in a decrease in numbers of thymic lymphocytes dying. Xenopus lymphocytes can therefore adapt to low environmental temperatures, whereas the mammalian lymphocyte is unable to tolerate temperatures outside the normal physiological range. Russ (1986) has shown that even quite high doses of DG fail to deplete all thymic lymphocytes from postmetamorphic thymus organ cultures, but caused substantial reduction in overall thymus size. In contrast, 1·5 mM-DG-treatment of 3-week-old larval thymus cultures did yield a lymphocyte-depleted organ. DG has an inhibitory effect on proliferating T lymphocytes (Cohen, Lee, Dosch & Gelfand, 1980). There is extensive evidence that DG interferes with DNA synthesis in cultured cells via its phosphorylated product deoxyguanosine triphosphate, which inhibits ribonucleotide reductase and hence interferes with purine and pyrimidine metabolism. DG-sensitivity appears to be related to an enhanced ability of immature cells for uptake and phosphorylation of DG to deoxyGTP, and by their reduced ability to degrade accumulated deoxyGTP. Functional immaturity of larval Xenopus thymocytes (e.g. an inability to respond to PHA under normal culture conditions) compared with postmetamorphic thymocytes has been recorded (Williams et al. 1983).

In this paper our attention has turned to the use of y-irradiation to deplete thymic lymphocytes. The cytoarchitecture of normal and irradiated, in vivo- or in vtiro-maintained thymuses are compared. In a companion paper (Horton, Russ, Aitchison & Horton, 1987) we trace the extent to which lymphoid and lymphocyte-depleted thymuses are colonized by host cells, after their implantation to thymectomized larvae, and assess the levels of persisting implant-derived cells within the thymus and the periphery. Our studies concentrate on the postmetamorphic thymus, since we initially wished to implant foreign thymus stroma expressing both class I and II MHC antigens; thus larval cells appear not to express MHC class I molecules (Flajnik, Kaufman & Du Pasquier, 1985).

(A) Animals

The partially inbred, MHC-compatible J strain (MHC haplotype jj) of Xenopus laevis was used in these experiments. The animals were bred and reared in the laboratory at 23 ± 2°C, under conditions described elsewhere (Horton & Manning, 1972). All animals were postmetamorphic and ranged from 4–8 months of age. Animal and thymus age refers to time postfertilization. The Xenopus used in this study had completed metamorphosis by 2 months of age.

(B) γ-irradiation of thymus or whole animal

Doses from 1000 to 3000 rad were achieved by positioning thymuses (removed from froglets and directly placed in amphibian culture medium in Falcon 12×75 mm tubes) or froglets (placed in 50 mm diameter plastic beakers) at different distances from a Cobalt 60 source. These two procedures for irradiation are referred to subsequently as in vitro and in vivo irradiation, respectively. The total irradiation dose received after 5 min exposure was measured using a ferrous ion chemical dosimeter (Lallone, 1984).

(C) Thymus histology following in vitro irradiation and/or organ culture

Thymus glands were aseptically removed from MS222-anaesthetized froglets and washed in amphibian strength Leibovitz (L15) culture medium. Thymus organ cultures were then established directly or after in vitro irradiation. Thymuses (one or more) were placed on a 13 mm diameter polycarbonate filter with 0·4 μm pore size (Nuclepore Corporation). Filters were prepared by boiling three times in double-distilled water prior to autoclaving. Individual filters were positioned on a piece of gelatin foam sponge (Sterispon No. 1, Allan & Hanbury) that had been placed in a 35×10 mm plastic Petri dish (Costar), containing approx. 1·5 ml amphibian L15 culture medium. The technique of culturing thymuses in this way was developed from the paper on fetal mouse thymus organ culture by Jenkinson et al. (1982). The amphibian medium consisted of an L15 base (Flow Labs) diluted to 60 % with double-distilled water and supplemented with 10 % heat-inactivated fetal calf serum (Flow) and also with 0 mM-Hepes buffer, 20 mM-sodium bicarbonate, 2 mM-L-glutamine, 50 i.u. ml−1 penicillin, 50 μg ml−1 streptomycin, 2–5 μg ml−1 Fungizone (all from Flow Labs) and 0·05 mM-2-mercaptoethanol. Thymuses were cultured in a water-saturated atmosphere in a 5% CO2 incubator at 26°C±0·5°C and culture medium was changed every 3 or 4 days. Thymuses were organ cultured for various intervals (range 2–28 days) prior to fixing in Bouin’s, then dehydrated and embedded in paraffin wax. 7 μm sections were stained in H & E.

(D) Light and electron microscopy of normal and in vivo-irradiated thymuses

(1) Time-course study

Here, the interval between in vivo irradiation (1000 or 3000 rad) and thymus removal varied from 5–14 days. Thymuses were histologically prepared as in section C above.

(2) Organ culture of in vivo-irradiated thymus

Thymuses were removed from 3000 rad-irradiated froglets at 8 days postirradiation and placed in organ culture for either 7 or 12 days to determine if lymphoid ‘regeneration’ would take place in the absence of any haemopoietic source. At the end of organ culture, thymuses were prepared for light microscopy as above.

(3) Observations-on plastic-embedded thymuses 9 days after 3000 rad total-body irradiation

Thymuses were initially fixed in Karnovsky’s fixative for 1·5 h at 4°C and then postfixed in 1 % osmium tetroxide for 0·5–1 h at 4°C. The organs were continually rotated during fixation to ensure maximum infiltration by each fixative. Tissues were dehydrated at room temperature, passed through three changes of propylene oxide (or acetone)/ Araldite for 30 min at 45°C, and finally embedded in Araldite. 1 μm sections were cut through the thymus at its maximum diameter and stained in toluidine blue for light microscopy. Ultrathin sections were mounted on copper grids and double stained with saturated uranyl acetate and lead citrate. Grids were observed with a Philips 400T electron microscope, operating at 100 kV.

(A) Thymus histology following organ culture and in vitro irradiation

Table 1 shows the outcome of experiments on froglet thymuses (normal or γ-irradiated) after varying periods of organ culture. Nonirradiated thymuses retained a fairly normal structure after 2 weeks of organ culture, as shown by comparing’Fig. 1A with IB. Lymphocyte numbers were only minimally depleted and cortex/medulla differentiation remains. The cortex was observed to develop a ‘follicular-like’ arrangement, with distinct lymphoid-filled follicles, separated by less-lymphoid regions.

Table 1.

Effect of in vitro γ-irradiation on histology of organ-cultured Xenopus thymus

Effect of in vitro γ-irradiation on histology of organ-cultured Xenopus thymus
Effect of in vitro γ-irradiation on histology of organ-cultured Xenopus thymus
Fig. 1.

Representative sections revealing the effects of organ culture and in vitro y-irradiation. 1 μm wax sections, haematoxylin and eosin stain. (A) 8-month in vivo thymus. Thymus shows typical cortex/medulla differentiation. Connective tissue trabeculae penetrate in from the capsule through the lymphoid-rich cortex. Some adipose tissue (a) is seen adjacent to the thymus. Category A (see Table 1). Scale bar, 250 μm. (B) 8-month thymus, organ cultured for 15 days. Lymphocyte numbers have minimally declined during culture, and cortex and medulla are still apparent. Note follicle-like (f) arrangement of cortical lymphocytes, a, adipose tissue; p, pigment. Category B. Scale bar, 250 μm. (C) 5-month thymus, organ cultured for 2 days after receiving 1000 rad in vitro irradiation. Many pyknotic cells are present and are especially visible in the cortical zone (c); lymphocytes are still present in the medulla (m). Clumps of melanin are noticeable. Category >C<. Scale bar, 150 μm. (D) 5-month old thymus, cultured for 14 days, having been treated with a dose of 1000 rad in vitro at the beginning of culture and again on day 6 of culture. Only a few lymphocytes remain in the centre of the small thymus. The organ is now mainly composed of stromal cells. Category >D<. Scale bar, 150 μm. (E) 4-month thymus irradiated in vitro with 2000 rad and cultured for 4 weeks. Thymus is very degenerate and pyknotic cells are evident. The epithelial cells have become arranged into convoluted epithelial layers (el). Large cystic (c) spaces are seen. Pigment and adipose tissue (a) surround the thymic area. Category >E<. Scale bar, 100 μm.

Fig. 1.

Representative sections revealing the effects of organ culture and in vitro y-irradiation. 1 μm wax sections, haematoxylin and eosin stain. (A) 8-month in vivo thymus. Thymus shows typical cortex/medulla differentiation. Connective tissue trabeculae penetrate in from the capsule through the lymphoid-rich cortex. Some adipose tissue (a) is seen adjacent to the thymus. Category A (see Table 1). Scale bar, 250 μm. (B) 8-month thymus, organ cultured for 15 days. Lymphocyte numbers have minimally declined during culture, and cortex and medulla are still apparent. Note follicle-like (f) arrangement of cortical lymphocytes, a, adipose tissue; p, pigment. Category B. Scale bar, 250 μm. (C) 5-month thymus, organ cultured for 2 days after receiving 1000 rad in vitro irradiation. Many pyknotic cells are present and are especially visible in the cortical zone (c); lymphocytes are still present in the medulla (m). Clumps of melanin are noticeable. Category >C<. Scale bar, 150 μm. (D) 5-month old thymus, cultured for 14 days, having been treated with a dose of 1000 rad in vitro at the beginning of culture and again on day 6 of culture. Only a few lymphocytes remain in the centre of the small thymus. The organ is now mainly composed of stromal cells. Category >D<. Scale bar, 150 μm. (E) 4-month thymus irradiated in vitro with 2000 rad and cultured for 4 weeks. Thymus is very degenerate and pyknotic cells are evident. The epithelial cells have become arranged into convoluted epithelial layers (el). Large cystic (c) spaces are seen. Pigment and adipose tissue (a) surround the thymic area. Category >E<. Scale bar, 100 μm.

Within 2 days of irradiation, organ-cultured thymuses contained many pyknotic cells, particularly in the cortex (Fig. 1C). By 14 days, 1000rad-irradiated thymuses became reduced in size and displayed significantly reduced lymphocyte numbers. The effect of 2–3000 rad was similar, but more dramatic; now very few lymphocytes were seen at 14 days of culture, and thymuses were extremely small. Thymuses given 1000 rad on day 1 and again on day 6 of culture were also virtually devoid of lymphocytes by 2 weeks, but reduction in their size was less marked than those given a single 2000rad dose (see Fig. ID). Irradiated thymuses cultured for 4 weeks became degenerate (Fig. IE) and surviving epithelial cells had become arranged into convoluted epithelial layers, with cystic spaces being prominent. The observation of epithelial outgrowths at this time supports the concept that epithelial cells are a major radiation-resistant cell type in the thymus.

(B) Light and electron microscopy of normal and in NiNO-irradiated thymus

Animals generally remained healthy during the first 7–10 days after whole-body irradiation, but after this time froglets often became sluggish and deaths were frequent from 10 days onwards, particularly in the 3000 rad-irradiated group.

(1) Time-course study

The results are summarized in Table 2. 5 to 8 days after 1000 rad irradiation, thymuses appeared healthy, with only a small decrease in overall size; however, they contained very reduced lymphocyte numbers, the latter mainly being found in the medulla. Pyknotic cells were scarce. By 12 days postirradiation, lymphocytes were now also very evident in the cortex of two thirds of the thymuses examined. Presumably, in these thymuses, 1000 radresistant lymphocytes have proliferated and have partially repopulated the organ.

Table 2.

Effect of in vivo γ-irradiation on histology of Xenopus thymus

Effect of in vivo γ-irradiation on histology of Xenopus thymus
Effect of in vivo γ-irradiation on histology of Xenopus thymus

Thymuses removed 5–8 days after 3000 rad totalbody irradiation showed clearly the initial destruction of cortical lymphocytes, the medulla containing only scattered lymphocyte-rich foci. By 10–14 days post-3000 rad, thymuses consisted mainly of stromal cells, although a few scattered lymphoid cells remained in the medulla. The cortex appeared to have become reduced in size and was virtually lymphocyte-free. Cystic spaces, mucous and myoid cells were noticeable in these irradiated thymuses.

(2) Organ culture of in vivo-irradiated thymus

Five thymuses were taken from froglets 8 days after 3000 rad whole-body irradiation and placed in organ culture. The two thymuses organ cultured for 8 days displayed loss of cortex/medulla differentiation. However, some lymphocytes were still seen scattered in the organs. After 12 days of organ culture, the irradiated thymuses had become small and looked rather abnormal in structure. They were still composed mainly of stromal cells, with very few lymphocytes.

(3) Light microscopic observations on 1 μm plastic-embedded thymuses, 9 days post-3000 rad total-body irradiation

Three control and three irradiated thymuses were examined in this and the subsequent electron micrographic study. Irradiated thymuses were about one third to one half the size of the normal thymus.

Irradiated thymuses were, as expected, extremely depleted of lymphocytes, although cortex and medulla could still be recognized (Fig. 2B). Large fat cells were seen around both normal and irradiated thymuses, but were particularly obvious around the irradiated organ. Pigment became very noticeable around and within irradiated thymuses.

Fig. 2.

Toluidine-blue-stained, 1 μm plastic-embedded sections of control and in vivo (3000 rad)-irradiated thymuses (9 days postirradiation). Thymuses taken from 8-month-old froglets. In A and C zone 1, subcapsular cortex; zone 2, deep cortex; 3, cortico/medullary junction; 4, medulla. (A) Montage through a segment of the cortex and part of medulla of a control thymus, showing extensive, lymphoid cortex, be, blood capillaries; c, capsule; ep, epithelial cells; gr, granulocyte; z, putative ‘interdigitating dendritic cells’; l, lymphocytes; lb, lymphoblasts; my, myoid cell. Scale bar, 30 μm. (B) Low-power view of an irradiated thymus. Note dramatic loss of lymphocytes from cortex (c). Some lymphocytes (l) retained in medulla (m). Thymus is reduced in size, but cortex/medulla differentiation is still evident, a, adipose tissue. Scale bar, 120pm. (C) Montage through a segment of cortex and medulla of an irradiated thymus, be, blood capillary; c, capsule; cy, cystic spaces; ep, epithelial cells; I, lymphocytes; Igc, large granulocyte; li, lipid inclusions of mucous cell; me, mast cell; m.o, macrophages; p, pigment. Scale bar, 30 μm.

Fig. 2.

Toluidine-blue-stained, 1 μm plastic-embedded sections of control and in vivo (3000 rad)-irradiated thymuses (9 days postirradiation). Thymuses taken from 8-month-old froglets. In A and C zone 1, subcapsular cortex; zone 2, deep cortex; 3, cortico/medullary junction; 4, medulla. (A) Montage through a segment of the cortex and part of medulla of a control thymus, showing extensive, lymphoid cortex, be, blood capillaries; c, capsule; ep, epithelial cells; gr, granulocyte; z, putative ‘interdigitating dendritic cells’; l, lymphocytes; lb, lymphoblasts; my, myoid cell. Scale bar, 30 μm. (B) Low-power view of an irradiated thymus. Note dramatic loss of lymphocytes from cortex (c). Some lymphocytes (l) retained in medulla (m). Thymus is reduced in size, but cortex/medulla differentiation is still evident, a, adipose tissue. Scale bar, 120pm. (C) Montage through a segment of cortex and medulla of an irradiated thymus, be, blood capillary; c, capsule; cy, cystic spaces; ep, epithelial cells; I, lymphocytes; Igc, large granulocyte; li, lipid inclusions of mucous cell; me, mast cell; m.o, macrophages; p, pigment. Scale bar, 30 μm.

Fig. 2A reveals the cytoarchitecture of cortex and medulla of a control thymus. Blood capillaries were noticeable at the edge of the thymus and extended down through the cortex; they were also prominent at the cortico/medullary junction. Penetrating blood capillaries were associated with connective tissue trabeculae. In the subcapsular cortex were found relatively pale-staining, large lymphoid cells. These cells are the lymphoblasts and were frequently seen in mitosis. Epithelial cells were also evident in the subcapsular cortex. The inner (or deep) cortex was very rich in small lymphocytes, that contain densely stained nuclear chromatin and have minimal cytoplasm. Within this deep cortex (and also in the medulla) were seen some distinct cells with large, pale-staining nuclei and voluminous, pale-staining cytoplasm, that appear to interdigitate with surrounding lymphoid cells (see Fig. 2A). These cells resemble those described in the Rana temporaria thymus medulla (Bigaj & Plytycz, 1984) and could represent the amphibian equivalent of interdigitating antigenpresenting (dendritic) cells, that are found in the mammalian thymic medulla (see Crouse et al. 1985).

The medulla was rich in epithelial cells, although it still contained large numbers of lymphocytes. The latter are less-densely stained than the small lymphocytes of the deep cortex. Lymphocytes could be distinguished from adjacent epithelial cells by their distinctive chromatin pattern. Thus hetero- and eu-chromatin patches with distinct nuclear membrane staining are lymphocyte characteristics, whereas epithelial cell types have paler-staining nuclei, often prominent nucleoli and a more extensive cytoplasm (Fig. 2A). Myoid cells, mucous cells and cystic spaces were occasionally seen in the medulla, as were granulocytes (Fig. 2A). Macrophages were also found, but were not particularly common in control thymuses.

Fig. 2C illustrates the changed cellularity of an irradiated thymus. In the virtual absence of lymphocytes, the epithelial cells in the subcapsular cortex were very noticeable and contained prominent nucleoli; some cells were rounded and others were elongated. Both pale- and darker-staining, elongate epithelial cells were found. The deeper cortex was very reduced in size and was composed of epithelial cells and noticeable cystic spaces. The medulla contained a scattering of lymphoid cells (Fig. 2C), but was mainly composed of epithelial cells and a variety of other stromal cell types. Thus myoid cells, mucous cells and cystic structures were conspicuous, and various types of granulocyte were seen. Deposits of pigment were noticeable in the medulla and lipid inclusions were commonly found. Macrophages, containing phagocytosed material, were frequently seen in irradiated thymuses (Fig. 2C). The granulocyte population included mast cells (containing relatively round granules) and large granulocytic cells that are comparable to the granular glands found in frog skin – these contain ellipsoid granules, typical of eosinophils (Fig. 2C). Irradiated thymuses were still well vascularized.

(4) Electron microscopic observations on normal and 3000 rad-irradiated thymuses

Fig. 3A,B are representative views of the subcapsular cortex and medulla of a control thymus. The subcapsular cortical layer contains many lymphoblasts, two of which are shown in stages of mitosis. Fig. 3B shows the cellular nature of the medulla. Lymphocytes are seen interspersed between the cytoplasmic extensions of much paler-staining cells. In many instances it was uncertain whether these cytoplasmic processes belong to interdigitating cells (known to originate extrathymically – see Discussion) or epithelial reticular cells. The putative interdigitating cells referred to here (see Fig. 3B) were not readily apparent in the irradiated thymuses. Medullary myoid cells were identified in control thymuses and possessed concentrically arranged striated myofibrils.

Fig. 3.

(A,B) Electron micrographs of thymuses from normal, 8-month-old Xenopus. (A) Subcapsular cortex, revealing many lymphoblasts, two of which are in mitosis (m). c, capsule; p, pigment. Scale bar, 5 μm. (B) Medullary region showing lymphocytes (l) interspersed between cytoplasmic extensions of paler-staining cells - the latter being either epithelial reticular cells or interdigitating cells, i, putative interdigitating cell with extensive cytoplasmic processes. Scale bar, 5 μm.

(C–F) Electron micrographs of thymuses taken from 8-month-old frogs, 9 days post-3000 rad in vivo irradiation.

(C) Subcapsular cortex showing epithelial cells (ep) with cytoplasmic processes. N.B. Intercellular spaces are also seen in the normal thymic cortex. Scale bar, 5 μm. (D) Myoid cell (my) in medulla. Some lymphocytes (l) still remain. Scale bar, 5 pm. (E) Two cysts (cy) in the medulla, (ep) epithelial cell with an abundance of rough endoplasmic reticulum. Cellular debris is seen within the cyst. Scale bar, 5 μm. (F) High-magnification view of E showing cilia (ci) in cyst interior and villi projecting into cyst. Scale bar, 1 μm.

Fig. 3.

(A,B) Electron micrographs of thymuses from normal, 8-month-old Xenopus. (A) Subcapsular cortex, revealing many lymphoblasts, two of which are in mitosis (m). c, capsule; p, pigment. Scale bar, 5 μm. (B) Medullary region showing lymphocytes (l) interspersed between cytoplasmic extensions of paler-staining cells - the latter being either epithelial reticular cells or interdigitating cells, i, putative interdigitating cell with extensive cytoplasmic processes. Scale bar, 5 μm.

(C–F) Electron micrographs of thymuses taken from 8-month-old frogs, 9 days post-3000 rad in vivo irradiation.

(C) Subcapsular cortex showing epithelial cells (ep) with cytoplasmic processes. N.B. Intercellular spaces are also seen in the normal thymic cortex. Scale bar, 5 μm. (D) Myoid cell (my) in medulla. Some lymphocytes (l) still remain. Scale bar, 5 pm. (E) Two cysts (cy) in the medulla, (ep) epithelial cell with an abundance of rough endoplasmic reticulum. Cellular debris is seen within the cyst. Scale bar, 5 μm. (F) High-magnification view of E showing cilia (ci) in cyst interior and villi projecting into cyst. Scale bar, 1 μm.

Fig. 3C-F are representative micrographs of irradiated thymuses. Fig. 3C shows the epithelial cells that predominate in the cortical region. Their nuclei are generally pale staining, except for the nucleolus and nuclear membrane. Their cytoplasm is rich in ribosomes, mitochondria and some rough endoplasmic reticulum is also seen. Fig. 3E,F show a portion of the medulla and reveal the nature of cystic structures that are frequently seen following irradiation. Numerous microvilli and cilia (with 9+2 arrangement of microtubules), project into the lumen of the cyst, which contains much debris. The cytoplasm of the cells associated with the cysts was sometimes rich in granular material. Adjacent epithelial cells had an abundance of rough endoplasmic reticulum, indicating active protein synthesis. Myoid cells of normal appearance (Fig. 3D) and mucous cells were frequently seen in micrographs of irradiated thymuses.

Changes in cytoarchitecture of the Xenopus thyrmis after y-irradiation were initially examined here by an in vitro approach, which involved irradiation of the excised thymus followed by varying periods of organ culture. This study revealed that irradiation damage first becomes evident within the cortex, which is the zone rich in rapidly dividing thymocytes. Two weeks after 2–3000rad irradiation, thymus organ cultures became very reduced in size, contained few lymphocytes overall and lacked the cortex/medulla differentiation evident in organ cultures of nonirradiated froglet thymuses. Although a variety of tissues from X. laevis have been successfully organ cultured by workers in other laboratories (see review by Mon-nickendam & Balls, 1973), culture of whole Xenopus thymus was first attempted only relatively recently by Williams (1981), whose studies were limited to the larval thymus.

This paper has more critically assessed (through light microscopy of 1 μm sections and by electron microscopy) the effects of whole-body irradiation on thymus structure. These experiments showed that in the second week following 3000rad irradiation, Xenopus thymuses are predominantly composed of stromal cells (e.g. thymus epithelial cells) and display loss of cortical lymphocytes and depletion of medullary lymphocytes, although some lymphocytes still persist in this latter zone. Many workers have previously noticed the relatively radioresistant population of lymphocytes in the (mammalian) thymic medulla compared with those in the cortex (e.g. Trowell, 1961; Bartel, 1984). Regaud & Crémieu (1912) first described the phenomenon of thymus inversion following irradiation, i.e. the atrophy of cortical lymphocytes and the persistence of epithelial cells, making the cortex resemble the medulla. Today, it is generally thought (Sharp & Watkins, 1981) that the difference in radiobiological properties of cortical and medullary lymphocytes represents the existence of two different subpopulations of thymocytes – immature and proliferating lymphocytes (about 85 % of the total) in the cortex, while the remaining 15 % of thymic lymphocytes found in the medulla are a functionally mature population and may resemble fully differentiated helper, cytotoxic and suppressor T cells.

The effect of in vivo γ-irradiation on the cytoarchitecture of the Xenopus thymus has a marked similarity to that achieved by the cytotoxic agent methylnitrosourea (NMU) which alkylates DNA. Thus Clothier and others (Clothier, Balls, Hodgson & Horn, 1980; Clothier, Knowles, James, Whittle & Balls, 1982) revealed that a single dose of NMU injected peritoneally into immature and mature X. laevis caused loss of lymphocytes from the thymus cortex within 3–5 days: this cortical depletion proved to be permanent provided a sufficiently high dose of NMU was given to immature animals, low doses being followed by cortex regeneration. In contrast, most of the medullary cell types (e.g. myoid cells, reticular cells, cystic cells, pigment cells, mast cells, granular cells and mucous cells) were apparently not morphologically affected by NMU treatment, although a temporary depletion of medullary lymphocytes was noted. NMU-treated Xenopus permanently fail to reject skin allografts and do not respond to haptens conjugated to erythrocyte carriers (Clothier et al. 1980; James, Clothier, Ferrer & Balls, 1982).

The nature of the thymic lymphocyte subsets that can survive 3000 rad irradiation in the frog is not known. Certainly the radiation-sensitive B cells found in the Xenopus thymus (Williams, Cribbin, Zettergren & Horton, 1983; Hsu, Julius & Du Pasquier, 1983) should be effectively removed (see discussion by Gearing et al. 1984). T suppressor cells are also known to be irradiation-sensitive in both mammals (see Sharp & Watkins, 1981) and in Xenopus (Ruben, Buenafe, Oliver, Malley, Barr & Lukas, 1985 – where 1000 rad eliminates suppressor-inducer function). That Xenopus T helper cells from the spleen may be able to withstand 3000 rad is suggested by Cribbin (1984) from her in vivo reconstitution studies of primary antibody responses and by Lallone (see Lallone, 1984) in his experiments on in vitro reconstitution. However, Ruben et al. (1985) show that 2000 rad prevents thymus cells from helping cocultured spleen fragments to produce antibody.

Interestingly, a preliminary experiment with a pool of 12 thymuses (needed to obtain sufficient lymphocyte numbers for culture), taken from Xenopus froglets 7 days post-3000 rad irradiation and subsequently organ cultured for 10 days, showed they still retained a thymocyte population that could respond to PHA (Russ, 1986). These radioresistant, PHA-reactive, medullary T-lineage lymphocytes, although few in number, must therefore not be excluded from playing a role in restoration of the immune system of thymec-tomized animals given irradiated thymus implants (see discussion in Arnall & Horton, 1986 and Russ, 1986). In this respect, host T-lineage cell development appears to be able to take place in nude mice injected with irradiated human T cells (Dosch, White & Grant, 1985).

The studies reported here have illustrated that the (normal) amphibian thymus has distinct subcapsular and deep cortical zones. This situation is also found in the mammalian thymus, where the subcapsular region houses a population of blasts (∼10 % of the total thymic lymphocyte population), whereas the deep cortex contains lymphocytes of smaller size (some 75 % of the total), many of which are destined to die within the thymus (Scollay, 1983). Another feature concerning froglet thymic structure suggested in this paper is the presence of dendritic interdigitating cells (IDC) in the normal thymus, but their possible absence following irradiation, although further studies are required to confirm the identity and radiosensitivity of these cells. The putative IDC noted here appear morphologically similar to mammalian IDC (Kaiserling, Stein & Müller-Hermelink, 1974); however, functional characterization is required before we can determine whether the large, pale-staining dendritic cells in the Xenopus thymus represent a population distinct from epithelial reticular cells, which also have extensive cytoplasmic processes (Weiss & Sakai, 1984). Along with thymic epithelial cells, IDC and macrophages are believed to play a critical role in T cell selection, education and maturation. Turpen & Smith (1986) have studied the ontogeny of both phagocytic and nonphagocytic (dendritic) accessory cells in the amphibian thymus, by grafting haemopoietic stem cells into cytogenetically distinct Xenopus embryos (2n or 3n) prior to thymic colonization. Their experiments provide circumstantial evidence suggesting thymocytes and thymic accessory cells could arise from a bipotential precursor, that diverges into these separate lineages after colonization of the epithelial thymic rudiment, during early development. Wood (1985) believes that thymic interdigitating/dendritic cells do not exist as more than morphologic variants of macrophages and that the terms ‘interdigitating’ or ‘dendritic’ cell are misleading and useful for descriptive purposes only.

In the present study on irradiated Xenopus thymus, macrophages were frequently seen, illustrating the fact that these cells must be considered as part of the radiation-resistant stromal population.

The light microscopic observations on 1 μm sections reported here are in agreement with the detailed ultrastructural observations on the developing Xenopus thymus recently made by Clothier & Balls (1985), who described the presence of pale- and dark-staining epithelial cells. These two populations were clearly seen here in the 3000 rad-irradiated thymuses. Such differential staining of epithelial cells may indicate varying states of metabolic activity of a single epithelial cell type (Singh, 1981) or may reflect their different embryologic origins (Farr & Nakane, 1983; Crouse, Perry & Jordan, 1984). From studies on lymphoid-free, low-temperature-cultured, embryonic mouse thymic epithelium, Crouse et al. (1984) propose that the lightly stained epithelial cells are ectodermal in origin and are responsible for inducing stem cell immigration and their early proliferation, whereas the endodermally derived epithelium (dark epithelial cells) is responsible for the emigration and/or maturation of the T-cell population found in the periphery.

As noted in some detail by Clothier & Balls (1985), a number of secretory cell types containing distinctive granules are also found within the Xenopus thymus. The present experiments reveal that 3000 rad wholebody irradiation does not remove these cells; indeed mucous cells, myoid cells, mast cells and large granular cells with ellipsoid cytoplasmic granules (the latter resembling skin granular glands) appeared in increased numbers in irradiated thymuses. Myoid cells are epithelial derivatives (Törö, Olah, Röhlich & Vigrach, 1969) and contain actin and myosin myofilaments arranged as in striated muscle. It is thought that contraction of these cells may facilitate thymic fluid circulation and/or the cells may act as a source of muscle-specific self-antigens (Törö et al. 1969; Rimmer, 1980). Another distinct effect of irradiation on the thymus gland, apart from the obvious reduction in lymphocytes, was the increase in number of cysts. Cysts may well act as sites for the deposition of dead cellular material; cells lining the cyst contain secretory-looking granules and microvilli and may be involved in the release of secretory material into the ciliated cyst lumen, or may reabsorb this material for transport to the vicinity of thymocytes or into the bloodstream (Curtis, Cowden & Nagel, 1979). It is possible that material resulting from lymphocyte breakdown is salvaged (Clothier & Balls, 1985); this could partly explain the increase in ‘cystic activity’ within the irradiated thymus.

In conclusion, the 1 pm light and electron microscopic observations presented here reveal that, in the short term, a radiation dose of 3000 rad given in vivo achieves a dramatic reduction in lymphocyte numbers, particularly from the cortex. However, such irradiation does not deplete the variety of stromal cells present within the thymus (with the possible exception of ‘dendritic cells’), and indeed causes an increase in thymic cysts and certain other cell types. Whether 3000 rad irradiation affects MHC antigen expression of thymic stromal cells is not known, although reagents are now available (see, for example, Flajnik, Kaufman, Reigert & Du Pasquier, 1984) to resolve this important issue. The functional capabilities of the γ-irradiated Xenopus thymus are currently being examined in this laboratory (see companion paper by Horton et al. 1987). The ability of an irradiated thymus to restore T cell-dependent antibody production and alloreactivity to early-thy-mectomized Xenopus and induce in vivo allotolerance towards thymus donor strain skin grafts (Gearing et al. 1984; Arnall & Horton, 1986) indicates that γ-irradiation provides an interesting model system for exploring thymus stromal cell functions.

This work was supported by research grants from the Science and Engineering Research Council and the North of England Cancer Research Campaign (to J.D.H.). A SERC Research Studentship to J.H.R. is also gratefully acknowledged. We thank Pamela Aitchison and Christine Richardson for technical assistance, and Dr Mike Stacey and Dr Bob Banks for help with microscopy. We are also grateful to David Hutchinson and Trudy Horton for help with manuscript preparation.

Arnall
,
J. C.
&
Horton
,
J. D.
(
1986
).
Impaired rejection of minor-histocompatibility-antigen-disparate skin grafts and acquisition of tolerance to thymus donor antigens in allothymus-implanted, thymectomized Xenopus
.
Transplantation
41
,
766
776
.
Balls
,
M.
,
Simnett
,
J. D.
&
Arthur
,
E.
(
1969
).
Organ cultures of normal and neoplastic amphibian tissues
.
In Biology of Amphibian Tumors
(ed.
M. M.
Mizell
), pp.
385
398
.
New York
:
Springer Verlag
.
Bartel
,
H.
(
1984
).
Ionizing irradiation effect on the thymus development
.
Folia Morphologica
1
,
1
17
.
Bigaj
,
J.
&
Plytycz
,
B.
(
1984
).
Cytoarchitecture of the thymus gland of the adult frog (Rana temporaria)
.
Folia Histochemica et Cytobiologica
22
,
63
70
.
Clothier
,
R. H.
&
Balls
,
M.
(
1985
).
Structural changes in the thymus glands of Xenopus laevis during development
.
In Metamorphosis
(ed.
M.
Balls
&
M.
Bownes
), pp.
332
359
.
Oxford
:
Clarendon Press
.
Clothier
,
R. H.
,
Balls
,
M.
,
Hodgson
,
R. M.
&
Horn
,
N.J.
(
1980
).
Effects of N-methyl-N-nitrosourea on the lymphoid tissues and skin allograft response of Xenopus laevis
.
Devi. comp. Immunol
.
4
,
265
272
.
Clothier
,
R. H.
,
Knowles
,
K. R.
,
James
,
H. S.
,
Whittle
,
D. J.
&
Balls
,
M.
(
1982
).
The pathological effects of N-methyl-N-nitrosourea in Xenopus laevis
.
In First International Colloquium on Pathology of Reptiles and Amphibians
(ed.
C.
Vago
&
G.
Matz
), pp.
147
156
.
Angers
:
University of Angers Press
.
Cohen
,
A.
,
Lee
,
J. W.
,
Dosch
,
H. M.
&
Gelfand
,
E. W.
(
1980
).
The expression of deoxyguanosine toxicity in T lymphocytes at different stages of maturation
.
J. Immunol
.
125
,
1578
1582
.
Cribbin
,
F. A.
(
1984
).
Reconstitution of a T celldependent antibody response in the thymectomized clawed toad Xenopus laevis
.
Ph.D. Thesis, University of Durham, UK
.
Crouse
,
D. A.
,
Perry
,
G. A.
&
Jordan
,
R. K.
(
1984
).
Transplantation of lymphoid free thymic epithelium in syngeneic and allogeneic immunodeficient mice
.
In Regulation of the Immune System
pp.
231
242
.
New York
:
A. R. Liss
.
Crouse
,
D. A.
,
Turpén
,
J. B.
&
Sharp
,
J. G.
(
1985
).
Thymic non-lymphoid cells
.
Surv. Immunol. Res
.
4
,
120
134
.
Curtis
,
S. K.
,
Cowden
,
R. R.
&
Nagel
,
J. W.
(
1979
).
Ultrastructural and histochemical features of the thymus glands of the adult lungless salamander, Plethodon glutinosas (Caudata: plethodontidae)
.
J. Morph
.
160
,
241
274
.
Dosch
,
H-M.
,
White
,
D.
, &
Grant
,
C.
(
1985
).
Reconstitution of nude mouse T cell function in vivo: IL2-independent effect of human T cells
.
J. Immunol
.
134
,
336
342
.
Du Pasquier
,
L.
&
Horton
,
J. D.
(
1982
).
Restoration of antibody responsiveness in early thymectomized Xenopus by implantation of major histocompatibility complex-mismatched larval thymus
.
Eur. J. Immunol
.
12
,
546
551
.
Farr
,
A. G.
&
Nakane
,
P. K.
(
1983
).
Cells bearing la antigens in the murine thymus. An ultrastructural study
.
Am. J. Path
.
III
,
88
97
.
Flajnik
,
M. F.
,
Kaufman
,
J. F.
&
Du Pasquier
,
L.
(
1985
).
Studies on the Xenopus major histocompatibility complex
.
Devi. comp. Immunol
.
9
,
777
781
.
Flajnik
,
M. F.
,
Kaufman
,
J. F.
,
Reigert
,
P.
&
Du Pasquier
,
L.
(
1984
).
Identification of class I major histocompatibility complex encoded molecules in the amphibian Xenopus
.
Immunogenetics
20
,
433
442
.
Gearing
,
A. J. H.
,
Cribbin
,
F. A.
&
Horton
,
J. D.
(
1984
).
Restoration of the antibody response to sheep erythrocytes in thymectomized Xenopus implanted with MHC-compatible or MHC-incompatible thymus
.
J. Embryol. exp. Morph
.
84
,
287
302
.
Hong
,
R.
,
Schulte-Wissermann
,
H.
,
Jarrett-Toth
,
E.
,
Horowitz
,
S. D.
&
Manning
,
D. D.
(
1979
).
Transplantation of cultured thymic fragments: II. Results in nude mice
.
J. exp. Med
.
149
,
398
415
.
Horton
,
J. D.
&
Manning
,
M. J.
(
1972
).
Response to skin allografts in Xenopus laevis following thymectomy at early stages of lymphoid organ maturation
.
Transplantation
14
,
141
153
.
Horton
,
J. D.
,
Russ
,
J. H.
,
Aitchison
,
P.
&
Horton
,
T. L.
(
1987
).
Thymocyte/stomal cell chimaerism in allothymus-grafted Xenopus: developmental studies using the X. borealis fluorescence marker
.
Development
100
,
107
117
.
Hsu
,
E.
,
Julius
,
M. H.
&
Du Pasquier
,
L.
(
1983
).
Effector and regulator functions of splenic and thymic lymphocytes in the clawed toad Xenopus
.
Annis Immun. Inst. Pasteur
134D
,
277
292
.
James
,
H. S.
,
Clothier
,
R. H.
,
Ferrer
,
I. R.
&
Balls
,
M.
(
1982
).
Effects of Mmethyl-N-nitrosourea on carrier-primed anti-hapten responses in Xenopus laevis
.
Devl. comp. Immunol
.
6
,
499
507
.
Jenkinson
,
E. J.
,
Franchi
,
L. L.
,
Kingston
,
R.
&
Owen
,
J. J. T.
(
1982
).
Effect of deoxyguanosine on lymphopoiesis in the developing thymus rudiment in vitro: application in the production of chimeric thymus rudiments
.
Eur. J. Immunol
.
12
,
583
587
.
Jordan
,
R. K.
,
Bentley
,
A.
,
Perry
,
G.
&
Crouse
,
D.
(
1985
).
Thymic epithelium. I. Lymphoid-free organ cultures grafted in syngeneic intact mice
.
J. Immunol
.
134
,
2155
2160
.
Jordan
,
R. K.
&
Crouse
,
D. A.
(
1979
).
Non-lymphoid cells and intrathymic T cell differentiation
.
In Development and Differentiation of Vertebrate Lymphocytes
(ed.
J. D.
Horton
), pp.
47
63
.
Amsterdam
:
Elsevier/North-Holland
.
Kaiserling
,
E.
,
Stein
,
H.
&
Müller-Hermelink
,
H. K.
(
1974
).
Interdigitating reticulum cells in the human thymus
.
Cell Tiss. Res
.
155
,
47
55
.
Lallone
,
R. L.
(
1984
).
Histocompatibility recognition in effector and helper T cell responses of Xenopus
.
Ph.D. Thesis, University of Durham, UK
.
Loor
,
F.
&
Hägg
,
L. B.
(
1977
).
The restoration of the T-lymphoid system of nude mice: Lower efficiency of non-lymphoid epithelial thymus grafts
.
Cell. Immunol
.
29
,
200
209
.
Monnickendam
,
M. A.
&
Balls
,
M.
(
1973
).
Amphibian organ culture
.
Experientia
29
,
1
17
.
Nagata
,
S.
& COHEN, (
1984
).
Induction of T-cell differentiation in early-thymectomized Xenopus by grafting adult thymuses from either MHC-matched or from partially or totally MHC-mismatched donor
.
Thymus
6
,
89
104
.
Owen
,
J. J. T.
&
Ritter
,
M. A.
(
1969
).
Tissue interaction in the development of thymus lymphocytes
.
J. exp. Med
.
129
,
431
437
.
Pritchard
,
H.
&
Micklem
,
H. S.
(
1973
).
Haemopoietic stem cells and progenitors of functional T-lymphocytes in the bone marrow of ‘nude’ mice
.
Clin. exp. Immunol
.
14
,
597
607
.
Pyke
,
K. W.
,
Bartlett
,
P. F.
&
Mandel
,
T. E.
(
1983
).
Lymphocyte depletion of murine fetal thymus and subsequent chimeric recolonization in vitro
.
Thymus
5
,
95
104
.
Regaud
,
C.
&
Crémieu
,
R.
(
1912
).
Thymus inversion after irradiation
.
C. r. hebd. Soc. Biol. Paris
72
,
253
523
.
Rimmer
,
J. J.
(
1980
).
Myoid cells and myasthenia gravis: a phylogenetic overview
.
Devl. comp. Immunol
.
4
,
385
394
.
Ruben
,
L. N.
,
Buenafe
,
A.
,
Oliver
,
S.
,
Malley
,
A.
,
Barr
,
K.
&
Lukas
,
D.
(
1985
).
Suppression in Xenopus laevis: thymus inducer, spleen effector cells
.
Immunology
54
,
65
71
.
Russ
,
J. H.
(
1986
).
Use of organ-culture, irradiation and adoptive-transfer to investigate the role of the Xenopus thymus in T lymphocyte development
.
Ph.D. Thesis, University of Durham, UK
.
Scollay
,
R.
(
1983
).
Intrathymic events in the differentiation of T lymphocytes: a continuing enigma
.
Immunol. Today
4
,
282
283
.
Sharp
,
J. G.
&
Watkins
,
E. B.
(
1981
).
Cellular and immunologic consequences of thymic irradiation
.
In Immunopharmacologic Effects of Radiation Therapy
, (ed.
J. B.
Dubois
), pp.
137
179
.
New York
:
Raven Press
.
Singh
,
J.
(
1981
).
The ultrastructure of epithelial reticular cells
.
In The Thymus Gland
, (ed.
M. D.
Kendall
), pp.
133
151
.
New York
:
Academic Press
.
Tôrô
,
J.
,
Olah
,
J.
,
Rôhlich
,
P.
&
Vigrach
,
S. Z.
(
1969
).
Electron microscopic observations on myoid cells of the frog’s thymus
.
Anat. Rec
.
165
,
329
342
.
Trowell
,
O. A.
(
1961
).
Radiosensitivity of the cortical and medullary lymphocytes in the thymus
.
Int. J. Rad. Biol
.
4
,
163
173
.
Turpén
,
J. B.
&
Smith
,
P. B.
(
1986
).
Analysis of hemopoietic lineage of accessory cells in the developing thymus of Xenopus laevis
.
I. Immunol
.
136
,
412
421
.
Weiss
,
L.
&
Sakai
,
H.
(
1984
).
The hematopoietic stroma
.
Am. J. Anat
.
170
,
447
463
.
Williams
,
N. H.
(
1981
).
Studies on the thymus and ontogeny of lymphocyte heterogeneity in the clawed toad: Xenopus laevis (Daudin)
.
Ph.D. Thesis, University of Durham, UK
.
Williams
,
N. H.
,
Cribbin
,
F. A.
,
Zettergren
,
L. D.
&
Horton
,
J. D.
(
1983
).
Ontogeny and characterization of mitogen-reactive lymphocytes in the thymus and spleen of the amphibian, Xenopus laevis
.
Immunology
49
,
301
309
.
Wood
,
G. W.
(
1985
).
Macrophages in the thymus
.
Surv. Immunol. Res
.
4
,
179
191
.
JOHN D.
Horton
,
JANE H.
Russ
,
PAMELA AITCHISON and TRUDY L. HORTON Thymocyte/stromal cell chimaerism in allothymus-grafted Xenopus: developmental studies using the X. borealis fluorescence marker