Evidence for major alterations in the thymocyte subpopulations in murine models of autoimmune diseases

Journal of Autoimmunity (1990) 3,27 l-288

Evidence for Major Alterations in the Thymocyte Subpopulations in Murine Models of Autoimmune Diseases

V. N. Kakkanaiah, Robert H. Pyle, Mitzi Nagarkatti and Prakash S. Nagarkatti Department Polytechnic

of Biology, Division of Microbiology Institute and State University,

and Immunology,

Blacksburg,

VA 24061,

Virginia USA

Thymocytes can be divided into four major subpopulations: CD4+CDS+ (double-positive), CD4-CDS(double-negative), CD4+CD8- (CD4+) and CD4-CDS+ (CDS+) cells. Recent studies have shown that T-cell development in the thymus progresses as: CD4-CDS-+CD4’CDSf-+CD4+ or CDS+ cells. In the present study we investigated these and other subpopulations of thymocytes in autoimmune MRL- + / + , MRL-lpr/Cpr, C57BL/6-Zpr/Zpr, BXSB and NZB mice before (l-month old) and after (4-6-months old) the onset of lymphadenopathy and autoimmune disease. All the autoimmune strains at one month of age and other H-2, sex and age-matched controls (C3H, DBA/2, and C57BL/6) demonstrated normal proportions of thymocyte subsets with -75% double-positive cells, 5-7% double-negative cells, 11-U% CD4’ cells and 3-S% CDS+ cells. By 4-6 months of age, MRL-+ /+ mice demonstrated a moderate increase in double-negative cells (-13%) and a decrease in double-positive cells ( - 46%). Interestingly, in the presence of the lpr gene, as seen in MRL-lpr/ lpr mice, the double-negative cells increased to -47% and the doublepositive cells decreased to - 16%. In contrast, 4-6-month-old C57BL/6lprjlpr mice failed to demonstrate any alterations in the thymocyte subsets thereby suggesting that background genes, in addition to the lpr gene, played a role in the thymocyte differentiation. BXSB male mice with severe lymphadenopathy behaved very similarly to MRL-Zprflpr mice, inasmuch as their thymus contained - 48% double-negative cells and only -8% double-positive cells. In contrast to MRL-lpr/lpr and BXSB strains, NZB mice at 6 or 10 months of age had normal composition of thymocyte subsets. In MRL and BXSB animals, although there was a significant increase in CD4+ cells (-23-33%), due to a consequent increase in CDS’ cells (- ll%), the ratio of CD4+:CD8+ cells remained 2-3:1, similar to that seen in normal mice. Furthermore, using the Jlld marker expressed by the Correspondence and reprint requests to: Prakash S. Nagarkatti, PhD, Division of Microbiology and Immunology, Department of Biology, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061, USA. 271 0896-841 l/90/030271 + 17 $03.00/O

0 1990 Academic Press Limited

272

V. N. Kakkanaiah et al. majority of the double-negative and all double-positive thymocytes but not by mature functional T cells, we confirmed the above findings and demonstrated further that MRL-Cprllpr mice at 4-6 months of age had an increased percentage of Jl Id- double-negative cells and a decrease in Jlld+ double-negative cells. The present study demonstrates for the first time the existence of a common defect in T-cell differentiation in the thymus of autoimmune strains of mice which is dependent on the genetic background and other accelerating factors responsible for inducing autoimmune disease.

-

Introduction

Studies carried out in several laboratories in the last 4 years have identified four major classes of cells in the adult mouse thymus. These include CD4+CD8+ (doublepositive) cells, CD4_CD8(double-negative) cells, CD4+CD8(CD4+) and CD4_CD8+ (CD8+) (single-positive) cells [l-3]. Double-positive cells constitute about 80% of thymocytes, double-negative cells 4%, CD4+ cells 12% and CD8+ approximately 4% [2]. Usually, the ratio of CD4+ cells to CD8+ cells is about 2:1, similar to that found in the periphery [4]. Recent studies have shown that the immature double-negative cells are the precursors of all other thymocyte subsets [4-61. The double-positive subset is believed to represent an intermediate stage in the differentiation pathway giving rise to the mature CD4+ or CD8+ T cells. Furthermore, it is likely that the T-cell tolerance to self-antigens occurs at the double-positive stage of T-cell development 17-l 11. Jl Id is a differentiation antigen expressed by all cortical double-positive cells [12, 131. It is also expressed by a subpopulation of thymic medullary CD8+ T cells but not by CD4’ T cells [ 131. The J 11 d antigen, however, is not expressed by the mature peripheral CD4+ or CD8+ T cells. Recently, however, we demonstrated that Jl Id was expressed by a significant number of peripheral double-negative T cells in MRL-&r/&r mice [ 141. The J 1 Id marker also subdivides the double-negative thymocytes into Jl Id+ cells expressing CD3 associated y6 type of TCR and Jl Id- cells having CD3 associated c$ type of TCR [15, 161. There are mainly three murine models of autoimmune disease, represented by strains NZB, MRL-&r/&r and BXSB. All these mouse strains spontaneously develop an autoimmune syndrome characterized by autoantibody production and hypergammaglobulinemia [ 17-191. There is evidence that the thymus plays an important role in the regulation of the autoimmune disease. For example, in MRLZprlZpr mice which develop massive lymphadenopathy associated with the autoimmune syndrome [20], neonatal thymectomy prevented both these proeessea{21]. In contrast, neonatal thymectomy led to a marked increase in autoantibody production and lymphadenopathy in BXSB mice [22], whereas in NZB mice, neonatal thymectomy failed to retard the autoimmune disease process [23,24]. These studies suggested that the thymus may play a significant role in regulating autoimmunity and lymphoproliferation in some autoimmune-susceptible mice. In the present study, we therefore investigated various subpopulations of thymocytes based on the expression of CD4, CD8 and Jl Id markers in MRL-Zpr/Zpr,

Thymocyte subsets in autoimmune mice

273

C57BL/64pr/Zpr, BXSB and NZB mice before and after the onset of autoimmunity. Our data suggested major alterations in the thymocyte subsets of MRL-lpr/Epr and BXSB mice after the onset of autoimmunity.

Materials

and methods Mice

The original breeding pairs of MRL-Zpr/Zpr (H-2K) and MRL- + /+ (H-2’) were purchased from the Jackson Laboratory, Bar Harbor, ME, USA. They were bred in our animal facility in a sterile environment (Animal Storage Isolators; NuAire Inc., Plymouth, MN, USA). C57BL/6 (H-2’) normal as well as C57BL/6-Zpr/Zpr (H-2b) mice, NZB (H-2d) and BXSB (H-2’) mice were procured from the Jackson Laboratory. C3H (H-2k) and DBA/2 (H-2d) mice were obtained from the National Cancer Institute, Bethesda, MD, USA.

Reagents The hybridomas GK 1.5 (rat IgG2b, anti-CD4), 3.155 (rat IgM, anti-CD8) and Jl ld2 (rat IgM) were grown in vitro and were used as antibody concentrated from supernatant as described elsewhere [25, 261. Fluorescein isothiocyanate (FITC) conjugated goat anti-rat IgM, FITC-conjugated anti-rat IgG F(ab’), and FITCconjugated rabbit anti-goat IgG were purchased from Cappel Laboratories, Cooper Biomedical Inc., Melvern, PA, USA. Affinity purified, biotin-conjugated anti-CD8 (53-6.7, rat IgG) and anti-Thy-l .2 (30-H12) antibodies, phycoerythrin conjugated anti-CD4, and phycoerythrin-streptavidin (PE-avidin) were purchased from Becton Dickinson, Mountain View, CA, USA. Affinity-purified normal rat gammaglobulin and goat anti-mouse IgM were obtained from Jackson Immuno Research Laboratories, West Grove, PA, USA.

Cell isozation

Single cell suspension of thymocytes was prepared using a laboratory blender (Stomacher, Tekmar Co., Cincinnati, OH, USA) in RPMI-1640 supplemented with 10% fetal calf serum (FCS) (Gibco Laboratories, Grand Island, NY, USA), 10 mM HEPES, 1 mM glutamine and 40 ug/ml gentamycin sulphate. The red cells were lysed using 0.83% ammonium chloride and were resuspended in the required volume after two washings [26].

Purijication

of double-negative

cells

To the cell pellet, containing 10 x lo7 thymocytes, 100 ul of anti-CD4 and 100 ul of anti-CD8 hybridoma supernatants were added and incubated on ice for 45 min. After two washings, 2ml of 1:lO diluted rabbit complement (Low-Tox-M; Cederlane Laboratory, Hornby, Canada) was added and incubated at 37°C for 30 min. The same process was repeated once to remove any contaminating cells

274

V. N. Kakkanaiah et al.

bearing CD4+ and CD8+ antigens, spared during the first treatment with antibody and C. Finally, the dead cells were removed by gradient centrifugation over histopaque (Sigma Chemicals Co., St. Louis, MO, USA).

Two-colourjhorescent

staining

Two-colour fluorescent staining was done as described in detail elsewhere [ 14,271. Briefly, staining for CD4 vs CD8 and Jl Id vs CD4 was performed as follows: one million thymocytes were incubated in 100 ~1 of anti-CD8 or Jl Id culture supernatants on ice for 30 min. The cells were washed twice with phosphate buffered saline (PBS) containing 0.1 y0 sodium azide, followed by staining with 1:lO diluted fluorescein isothiocyanate (FITC)-conjugated goat anti-rat IgM for 30 min on ice. The cells were then washed twice and incubated on ice for 30 min with rat Ig in some experiments. The cells were washed again and stained with 1:20 diluted PEconjugated anti-CD4 for 30 min on ice. Finally, the cells were washed three times and resuspended in 300 ~1 of PBS. In some experiments, to stain for CD4 vs CD8, cells were first stained with 1:20 diluted anti-CD4 antibodies (rat IgG) followed by 1:lO diluted FITC-conjugated anti-rat IgG F(ab’), antibody. Further, before the addition of biotin-conjugated anti-CD8 antibody, the cells were preincubated with rat Ig to block any residual binding sites on the FITC-conjugated anti-rat IgG F(ab’), antibody. Then the cells were stained for CD8 with 1:20 diluted biotinconjugated anti-CD8 antibodies, followed by PE-avidin. It should be noted that when we stained the thymocytes first with anti-CD8, which is rat-IgM, followed by FITC-conjugated anti-rat IgM, we observed that there was no need to add rat Ig to block any residual binding sites because the second antibody used (anti-CD4) was rat IgG. Addition of rat Ig was, however, essential when we used anti-CD4 (rat IgG) first, followed by FITC-conjugated anti-rat IgG. To stain for Jl ld vs Thy-l .2 and Jl Id vs CD8, staining for Jl Id was performed as above with Jlld culture supernatant followed by FITC-conjugated goat anti-rat IgM. After two washings, the cells were incubated in 1:20 diluted biotin-conjugated anti-Thy-l .2 or anti-CD8 for 30 min on ice. After two washings, they were stained with 15 diluted PE-avidin for 15 min on ice. Finally, the cells were washed three times and resuspended in 300 ul of PBS. Negative controls consisted of cells incubated with normal rat gammaglobulin (50 pg/ml) in place of various antibodies. Further, they were stained with FITCconjugated goat anti-rat IgM followed by normal rat Ig and lastly with PE-avidin. TO stain for sIg+ cells in the thymus, the cells were incubated with affinity purified goat anti-mouse IgM followed by 1:lO diluted FITC-conjugated rabbit anti-goat IgG. The cells were washed and analysed using a flow cytometer.

Flow cytometry

analysis

The stained cells were scanned on an Epics V model 752 (Coulter Electronics, Hialeah, FL, USA), laser flow cytometer and cell sorter. The four parameters studied per cell were; forward angle light scatter, 90” light scatter, green fluorescence (GFL) and red fluorescence (RFL). Laser excitation was normally 300 mV at

Thymocyte subsets in autoimmune mice

275

488 nm using a 5-W Innova 90 Argon Laser (Coherent

Inc., Palo Alto, CA, USA). The scanned data were analyzed with the multiparameter data acquisition and display system and the Easy 88 microcomputer system of Coulter Electronics. Forward angle light scatter was collected by log integral. Histograms showing cell number per channel as a function of fluorescence were collected at a resolution of 256 channels and gated on forward angle/90” light scatter and GFL/RFL dual parameter histograms of 64 x 64 channel resolution, defining the cell population of interest. The total number of counts for each sample was 10,000 cells.

Results Expression of CD4 and CD8 antigens by MRL-

+ I+,

MRL-lprllpr

thymocytes

Thymocytes from female MRL-Zpr/Zpr mice before the onset of autoimmune disease and lymphoproliferation (l-month old) and after the onset of the disease (4-6months old), were screened for the expression of CD4 and CD8 antigens using twocolour fluorescent staining technique. A minimum of five individual mice were screened in each group, and the various groups were compared for statistical significance using Student’s t-test. Age, sex and H-2 matched MRL- + /+ and C3H mice were also included as controls. A representative experiment of the two-colour flow cytometric analysis of different cell populations is shown in Figure 1 and the data of several similar experiments are summarized in Table 1. In Figure 1, the left panels demonstrate representative negative controls in which the cells from l-month-old mice had been incubated with affinity-purified rat Ig instead of the specific antibodies, followed by staining with FITC-conjugated anti-rat IgM and PE-avidin. In most experiments, the negative controls showed > 98% of the cells negative for green and red fluorescence. For this reason, the negative controls in subsequent experiments have not been depicted in Figures. As seen in Figure 1 (top row) and Table 1, C3H thymocytes demonstrated -7576 CD4+CD8+,5% CD4-CD8-, N 14% CD4+CD8and ~6% CD4_CD8+ T cells. These data were consistent with the proportion of different subsets of thymocytes studied by other investigators in the normal thymus [2]. MRL-+ /+ mice (l-month old) demonstrated similar proportions of thymocyte subsets (Figure 1, middle row). However, 4-6-month-old MRL-+ /+ mice showed a significant decrease in double-positive cells (46 IfI8.3%), a slight but statistically significant increase in double-negative cells (13rt2.9%) and an increase in CD4+ T cells (33 + 8.9%). MRL-Zpr/Zpr mice before the onset of lymphadenopathy (l-month old) behaved (Figure 1, bottom row) like normal mice, but following the onset of the disease and lymphoproliferation (4-6-months old), demonstrated very low percentages of double-positive cells (16 _lz6.4%), high double-negative cells (47 &-8.5%) and a minor increase in both CD4+ (23 + 5 %) and CD8’ (11 rfr6.4%) cells. Although there was an increase in both CD4+ and CD8+ cells, the ratio was N 2: 1 which was the normal ratio found in control mice. The above data suggested that by 4-6 months of age, even MRL-+ /+ mice demonstrated a slight but significant decrease in double-positive cells and an increase in double-negative T cells, and that this process was perhaps enhanced by the presence of the Zpr gene, since in the MRL-Zpr/Zpr mouse this effect was more pronounced.

276

V. N. Kakkanaiah

et al. 6 months

I month

C3H 1

13yo,

77%

1

1

I

3 Green fluorescence-Rot

lg

I month 14% 73%

6%

7%

/

76%

,

6 months

I month

MR L-lpdlpr

11%

2

I

6 months 26% 17%

2

4

3

42%

13%

4

Green fluorescence-CD6

Figure 1. Subpopulations of thymocytes based on the expression of CD4 and CD8 antigens using twocolour flow cytometry in C3H (top row), MRL-+/+ (middle row) and MRL-Zpr/Zpr (bottom row) strains. These strains were studied at one month or at 4-6 months of age. Cells were stained sequentially with anti-CD8 and FITC-anti-rat IgM followed by PE-anti-CD4. Negative controls consisted of cells incubated with normal rat Ig in place of antibodies, followed by FITC-anti-rat IgM, normal rat Ig and PE-avidin. The left panels represent negative controls using thymocytes from one-month-old mice. These panels showed > 98% of the cells negative for green and red fluorescence. The negative controls for 4-6-month-old strains were similar to the above and therefore not depicted.

CD4+ and CD8+ subpopulations

in the thymus of C57BL/6-lprllpr, mice

BXSB

and NZB

Since we observed phenotypic changes in the subpopulations of thymocytes from 4-6-month-old MRL mice bearing the lpr gene, we next determined whether these alterations were attributable only to the Zpr gene or whether other factors and the background genes also played a role. For this purpose, we studied the C57BL/6-Zpr/ lpr mouse with lymphadenopathy and compared the subpopulations with normal, age- and sex-matched C57BL/6 mice. The data shown in Figure 2 and summarized in Table 1 demonstrated that the proportions of the subpopulations studied in onemonth-old or 4-6-month-old C57BL/6-lpr/Zpr mice were normal and similar to the control C57BL/6 mice (Figure 2, Table 1). In contrast, when we investigated the BXSB strain, it was observed that l-month-old male mice had normal levels of

Thymocyte subsets in autoimmune mice

277

Table 1. Thymocyte subsets based on the expression of CD4 and CD8 antigens in various

strains of mice’ “/bof cells expressing

Strain

MHC (H-Z)

C3H C3H MRL-+/+ MRL-+/+ MRL-lpr/lpr MRL-lprjlpr C57BL/6 C57BL/6 C57BL/6-Zpr/Zpr C57BL/6-lpr/lpr BXSB BXSB’ BXSB’ DBA/2 DBA/2 NZB NZB

k k k k k k b b b b b b b d d d d

Sex Female Female Female Female Female Female Female Female Female Female Male Male Male Female Female Female Female

Age (months) 1 4-6 1 4-6 1 4-6 1 8 1 4-6 1 4-6 4-6 1 4-6 1 4-6

CD4+ CD8-

CD4+ CD8+

CD4CD8-

CD4CDS+

14.0f 1.4 15.5kO.7 15.Ok3.3 33.0 + 8.9 17.8k2.5 23.Ok5.1 11.0*2.8 13.5 + 7.8 9.OkO.O 10.3+ 1.5 12.1 f 1.5 13.0+ 11.3 32.0 f 7.6 16.5k2.1 16.Ok 1.4 9.0+ 1.4 13.Ok2.8

75.Ok2.8 74.0+ 1.4 71.5k5.8 46.0f8.3 68.8f4.5 16.0+6.4 77.Ok4.2 73.5k6.4 81.0+ 1.4 77.7k5.5 78.Ok4.4 55.Ok2.8 8.0f2.9 74.5k6.4 73.5 f 3.5 81.5k2.1 75.5f3.5

5.0& 1.4 7.5kO.7 4.5k1.6 13.Ok2.9 7.0+ 1.5 47.Ok8.5 7.0+ 1.4 9.Ok1.4 7.5k2.1 9.3k3.5 6.7k2.1 15.5kO.7 48.5 + 13.0 7.5k3.5 8.5 +0.7 5.OkO.O 7.oIfIo.o

5.5kO.7 3.5kO.7 8.Ok5.6 3.Ok2.2 6.252.3 ll.Ok6.4 5.OkO.O 4.OkO.O 2.5kO.7 2.0,t 1.0 2.7kO.6 6.5kO.7 11.5k5.6 4.OkO.O 4.OkO.7 4.5$0.7 5.Okl.4

‘Thymocytes were analyzed for surface markers CD4 and CD8 using two-colour immunofluorescent staining techniqueas described in Figure 1 and at length in Materials and methods. The data represent the percentage mean+ SD of a minimum of three experiments using C3H, C57BL/6 and DBA/2 mice and > 5 experiments in all other strains. *BXSB mice with moderate lymphadenopathy. 3BXSB mice with severe lymphadenopathy.

thymocyte subpopulations. However, by 4-6 months of age, correlating with the degree of lymphadenopathy, the thymocytes showed a decrease in double-positive cells and an increase in double-negative cells (Figure 2, Table 1). It was interesting to note that some BXSB mice at 4-6 months of age had minor lymphadenopathy. Such mice demonstrated a less dramatic increase in double-negative cells (15.5 +0.79/,) and a decrease in double-positive cells (55 + 2.8) (Table 1). In contrast, similar aged mice but with severe lymphadenopathy demonstrated a dramatic increase in doublenegative cells (48.5 f 13) and a decrease in double-positive cells (8 f 2.9) (Table 1). The percentages of CD4+ (32 + 7.6%) and CD8+ (11.5 + 5.6%) cells also increased in these mice, although the ratio between CD4+ to CD8+ cells remained 3:l (Table 1). In contrast to 4-6-months-old MRL-Zpr/Zpr and BXSB mice, 4-6-month-old NZB female mice had normal phenotypic composition when compared to young NZB mice for H-2 and age matched DBAj2 mice (Figure 3, Table 1). Furthermore, NZB mice, even at 10 months of age, demonstrated normal composition of thymocyte subsets (data not shown). It should be noted that while staining for CD4 and CD8 markers, we also stained the thymocytes for Thy-l antigens and in most experiments, Thy-l + cell content was > 95 %.

278 V. N. Kakkanaiah et al. I month 1)

3

/

II%

6%

11

5%

4

I month

6%

,

I[

3

9%

1

77%

4%

4

6 months

6 months

BXSB

62%

/

10%

C57BL/6-lpr//nr

I month j

6 months

C5?BL/6 76%

Ii

3

36%

43% Orebn fluorescence

/

9%

12%

-CD8

Figure 2. Thymocyte subpopulations based on the expression of CD4 and CD8 antigens in l- or 6-8month-old C57BL/6, C57BL/6-&r/&r and BXSB strains. The negative controls were included in each experiment as described in Figure 1 and were > 98% negative for green and red fluorescence, similar to those depicted in Figure 1 and therefore not depicted. The thymocytes were stained for CD4 and CD8 Ag as described earlier in Figure 1.

Expression

of J11 d and CD4

Jl Id is expressed by all cortical double-positive T cells [12]. It is also expressed by a majority of the double-negative T cells [ 131. In the medulla, Jl 1d is expressed only by a proportion of CD8+ T cells but not by CD4’ T cells [ 131. Recently we reported that Jl Id was expressed by a significant number of peripheral double-negative T

Thymocyte subsets in autoimmune mice DBA/2

I month

6

monlhs

6

months

279

e

z I

z 5 g

3

7 %

4 % I month

4

3

8 %

NZB

3 Green

fluorescence

7 %

4 %

6 %

4

- CO8

Figure 3. Subpopulations of thymocytes based on the expression of CD4 and CD8 markers in DBA/2 and NZB mice at 1 or at 6 months of age. The negative controls, included in all experiments, as depicted in Figure 1, were z 98% negative for green and red fluorescence.

cells in MRL-Zpr/Zpr mice [ 141. We therefore carried out further studies to investigate whether autoimmune susceptible mice have similar or different composition of thymocyte subpopulations when compared to normal mice based on two-colour staining with Jlld and anti-CD4 or Jlld and anti-CD8 antibodies. It should be noted that in experiments using Jl Id ~1 CD4, the Jl ldfCD4+ cells represented the double-positive cortical thymocytes; the J 1 ld_CD4population consisted of a mixture of double-negative cells and CD8+Jlldcells; and Jl ld_CD4+ cells represented the CD4+ T cells. This assumption was based on our observation in MRL-Zpr/Epr mice that purified CD4+ thymocytes failed to express Jl Id (data not shown), consistent with the observation in normal mice that Jl Id was not associated with CD4+ thymocytes [13,14]. A representative experiment of thymocyte staining for Jl Id and CD4 antigens in 4-6-month-old C3H, MRL- + /+, C57BL/6-Zpr/Zpr, MRL-lpr/Zpr, NZB and BXSB mice has been depicted in Figure 4. Several experiments carried out before and after the onset of autoimmunity in different strains, and in age and sex matched controls, has been summarized in Table 2. The Jl Id+CD4’ subpopulation representing the double-positive cortical cells was decreased in 4-6-month-old MRL-

280

V. N. Kakkanaiah et al.

3

3

32%

6%

4

2%

6% NZB

Green

fluorescence-Jlld

Figure 4. Thymocyte subpopulations based on the expression of Jl Id and CD4 antigens. Data from 46-month-old strains, C3H, MRL- + /+ , C57BL/6-Zpr/lpt, MRL-lpr/lpr, NZB and BXSB with severe lymphadenopathy have been depicted. The thymocytes from these mice were stained sequentially with Jl Id Ab, FITC-conjugated goat anti-rat IgM, rat Ig, and PE-anti-CD4. The negative controls consisted of cells incubated with normal rat Ig in the place of various Ab and stained with FITC-anti rat-IgM and PE-avidin. The negative controls in most experiments showed > 98% of the cells negative for both green and red fluorescence and have not been depicted.

+ /+, in older MRL-lpr/lpr and BXSB mice but not in C57BL/6-Zpr/Zpr and NZB mice, confirming earlier results (Table 1). Similarly, only 4-6-month-old MRL+ /+ , MRL-@r/&r and BXSB mice contained higher percentages of Jl ld_CD4cells, which is partially due to the increase in double-negative thymocytes. When the Jlld’CD4subpopulation was compared in different strains of mice, it was

Thymocyte subsets in autoimmune mice

281

Table 2. Expression of Jl Id and CD4 antigens on the thymocytes of various strains of

mice’ “/d of cells expressing

Strain

C3H C3H MRL-+/+

MRL-+/+ MRL-lpr/lpr MRL-lpr/lpr C57BL/6 C57BL/6 C57BL/6-lpr/lpr C57BL/6-lpr/lpr BXSB BXSB’ NZB NZB

Sex

Age (months)

JlldCD4+

Female Female Female Female Female Female Female Female Female Female Male Male Female Female

1 4-6 1 4-6 1 4-6 1 8 1 4-6 1 4-6 1 4-6

9.5f3.5 10.4kO.5 11.5k3.0 30.5k5.1 11.0+3.2 19.Ok6.0 8.5kO.7 11.3k6.0 6.050.0 7.7k2.3 5.5k2.3 41.Ok4.2 9.0* 1.0 11.5k4.9

Jlld+ CD4+

81.025.7 81.5kO.9 74.Ok2.8 32.5fll.O 76.Ok8.2 23.0f8.3 86.Ok2.8 81.7k8.6 85.5f0.7 83.7*5.5 91.5k2.1 12.1k1.2 86.5k3.5 80.5k3.5

JlldCD4-

5.0-t 1.4 6.2kO.6 8.5k2.4 30.5k6.7 10.0k4.6 34.Ok7.6 3.0+ 1.4 5.0f2.0 4.0fO.O 6.Ok3.6 1.5kO.7 36.2kl.O 3.0f 1.0 7.0+ 1.0

Jlld+ CD4-

5.OkO.l 1.7f 1.7 6.Oi5.5 2.Ok3.1 3.0+0.7 24.Ok6.9 2.0+ 1.4 2.3+ 1.2 5.0+ 1.4 3.0* 1.0 1.5kO.7 10.5f2.1 1.5kO.7 1.orto.o

lThymocytes were analyzed for surface markers J 1 Id and CD4, using a two-colour immunofluorescenr staining technique as described fi Figure 4 and at length in Materials and methods. The data represent the percentage mean of 3-5 experiments & SD. ‘BXSB mice with severe lymphadenopathy.

interesting to note that all stkains carried - 2-6% of these cells except 4-6-monthold MRL-lpr/lpr mice which had - 24% of these cells (Table 2). As discussed earlier, since Jl ld+CD4cells predominantly represent the double-negative cells and since 4-6-month-old MRL-lpr/lpr had - 47% double-negative cells (Table 1) together, this suggested that MRL-Zpr/Zpr double-negative cells contained less than 50% J 11 d+ cells. To address this question directly, double-negative cells were purified by treating thymocytes with anti-CD4 and anti-CD8+C. These cells were next strained for Jl Id ~1sThy-l or CD4 vs CD& It was observed that in l-month-old MRL-Zpr/Epr mice, -90% of the double-negative cells were Jlld+. In contrast, in 4-6-month-old MRL-Zpr/Zpr mice, only about 50% of the double-negative cells were Jlld+ (data not shown). These double-negative cells were over 98% Thy-l+CD4_CD8-. Lastly, the composition of the Jlld_CD4+ subpopulation, which represented the CD4+ thymocytes, was comparable to the data reported earlier (Table 1).

Expression of Jl 1d and CD8 When thymocytes were stained for Jlld vs CD8, four subpopulations were obtained: Jl ldfCD8+ which represented the double-positive cortical thymocytes and CD8+ Jlld+ cells; JlId_CD8cells which were double-negative and CD4’Jl Id- cells; Jl ld+CD8cells which represented only the double-negative

282

V. N. Kakkanaiah et al. C3H D

MRL-+/+ 16%

30%

42 %

12%

!r -r

3 C57BL/6-lpr/lpr I

3%

76 %

MRL-fpr/lpr

I

13%

3

56%

NZB 5%

17%

14% BXSB

79%

13%

3. 67% Green fluorescence-Jlld

II%

9%

4

Figure 5. Subpopulations of thymocytes based on the phenotypic expression of CDS and Jlld. Data from 4-6-month-old strains, C3H, MRL- + /+, C57BL/6-lpr/lpr, MRL-lpr/lpr, NZB and BXSB with severe lymphadenopathy have been depicted. The thymocytes were stained first with J 11 d Ab followed by FITC-anti rat IgM, biotin-conjugated anti-CD8 and PE-avidin. The negative controls were included in each experiment and consisted of cells incubated with normal rat-Ig in the place of various antibodies followed by FITC-anti rat IgM and PE-avidin. The negative controls were > 98”/, negative for both green and red fluorescence and have not been depicted.

Jlld+ cells; and Jlld_CD8+ cells which were CD8+ cells that did not express Jl Id. The expression of Jlld OS CD8 in 4-6-month-old C3H, MRL-+/+, C57BL/6-lpr/lpr, MRL-lpr/lpr, NZB and BXSB mice with severe lymphadenopathy has been depicted in Figure 5 and the data from several experiments has been summarized in Table 3. The data showed a significant decrease in Jl ld+CD8+ and

Thymocyte subsets in autoimmune mice

283

Table 3. Thymocyte subpopulations based on the expression ofJl1 d and CD8 antigens in

various strains of mice’ % of cells expressing

Strain

C3H C3H MRL-+/+ MRL-+/+ MRL-lpr/lpr MRL-lpr/lpr C57BL/6 C57BL/6 C57BL/6-lpr/lpr C57BL/6-lpr/Zpr BXSB BXSB’ NZB NZB

Sex

Age (months)

Female Female Female Female Female Female Female Female Female Female Male Male Female Female

1 4-6 1 4-6 1 4-6 1 8 1 4-6 1 4-6 1 4-6

JlldCD8+

4.Ok2.8 4.liO.6 10.053.3 18.8k4.4 ll.Ok4.2 ll.Ok1.6 9.5 k6.4 4.0f 1.0 2.OkO.O 3.Okl.O 2.121.4 10.0+4.1 3.5kO.7 5.5kO.7

Jlld+ CD8+

71.Ok9.9 71.8k1.9 67.Oi5.9 32.Ok6.5 69.0+9.9 18.8k5.9 77.5 kO.7 73.3+ 12.0 82.OkO.O 78.Ok8.2 86.5k2.1 lO.Of 1.4 82.5f3.5 73.Ok2.8

JlldCD8_

13.5k6.4 13.7kl.l ll.Ok5.5 39.Ok8.6 15.Ok5.5 44.4211.4 7.Ok2.8 13.7f6.8 7.5kO.7 12.7k6.7 4.0+0.4 70.9k3.9 9.5k2.1 16.5kO.7

Jlld+ CD8-

11.5kO.7 10.51to.3 10.0+4.9 10.2k4.1 9.Ok2.0 26.2k8.7 6.0 +4.2 9.3k4.6 8.5 +0.7 7.0* 1.0 7.5f0.7 9.0+0.0 4.5kO.7 4.5kO.7

‘Thymocytes were analyzed for Jl Id and CD8 antigens using a two-colour immunofluorescent staining technique as described in Figure 5 and at length in Materials and methods. The data represent the percentage mean of 3-5 experiments k SD. 2BXSB mice with severe lymphadenopathy.

in Jl ld_CD8cells in 4-6-month-old MRL-+ /+, MRL-Zpr/Zpr and BXSB mice as described earlier (Table 1). In contrast, NZB and C57BL/6-Zpr/Zpr mice failed to show any significant alterations.

increase

Alterations in the thymocyte subsets is not due to injiltration of lymph node cells into thymus Since the lymph nodes in close proximity to the thymus are enlarged in MRL-Zpr/Zpr mice, extreme care was taken to avoid harvesting lymph nodes at the time of collection of the thymus. To further substantiate, histopathological sections of the harvested thymus were studied and these failed to demonstrate any contaminating lymph node tissue (data not shown). An additional possibility was that the lymph node cells were actively infiltrating the thymus from the adjoining lymph nodes. To address this possibility, the presence of sIg+ B cells in the thymus was investigated. MRL- + /+ and MRL-Zpr/Zpr mouse (3-months old) thymocytes were stained for sIgM and also for CD4 and CD8 antigens. It was observed that MRL-Zpr/Zpr thymocytes contained -2% sIg+ cells, similar to findings in MRL- + /+ mice (w-3%, data not depicted). This minor percentage of B cells was probably due to contaminating blood cells at the time of harvesting the thymus, as also observed by others [28]. The MRL-Zpr/Zpr thymus, however, demonstrated a significant increase in the double-negative cells and decrease in double-positive cells, as observed in Table 1.

284 V. N. Kakkanaiah et al.

These results therefore demonstrated that the changes seen in the thymocyte subsets in MRL-&r/&r mice were not due to active infiltration of adjoining lymph node cells but probably to an intrinsic defect in thymocyte differentiation. Discussion

In the present study we investigated different subpopulations of thymocytes based on the expression of CD4, CD8 or Jl Id antigens in several strains of mice genetically susceptible to autoimmunity before and after the onset of the disease process. The data suggested that in MRL- + /+ mice, by 4-6 months of age, there was a tendency toward a decline in the percentage of double-positive cells and an increase in the percentage of double-negative cells. This effect was very pronounced in the presence of the Zpr gene, since MRL-lpr/lpr mice demonstrated a severe decrease in doublepositive cells and increase in double-negative T cells. Interestingly, however, C57BL/6-Zpr/Zpr mice exhibited no alterations in the thymocyte subpopulations. BXSB mice with severe lymphadenopathy behaved very much like MRL-Zpr/Zpr mice in demonstrating increased double-negative cells and decreased doublepositive cells, while NZB mice at 4-6 months of age had a normal composition of thymocyte subsets. Recently, we observed that C57BL/6-Zpr/Zpr and NZB mice even at 8 and 12 months of age respectively, did not demonstrate any significant alterations in the thymocyte subsets. The alterations in the proportions of thymocyte subpopulations were predominantly restricted to double-positive and double-negative subpopulations. In some strains, such as 4-6-month-old MRL-Zpr/Zpr and BXSB, although there was a significant increase in CD4+ cells, due to a consequent increase in CDS+ cells as well, the ratio of CD4+:CD8+ cells remained 2-3:l. This ratio is commonly found to vary in that range, depending on mouse strains, age, and sex [4]. Surface staining with Jl Id and CD4 or Jl Id and CD8 confirmed the above findings and further demonstrated in MRL-Zpr/Zpr mice that the percentage of double-negative Jl Id+ cells decreases with the onset of lymphadenopathy. Several recent studies have suggested that double-negative thymocytes are the precursors of all other thymocyte subsets [4-61, perhaps giving rise to the doublepositive cells that in turn mature into single positive CD4+ or CD8+ cells [7-111. Furthermore, tolerance to self-antigens may be achieved at the double-positive stage of T-cell development [7-l 11. In our study we observed, particularly in MRL-Zpr/Zpr and BXSB mice with lymphadenopathy, an increase in the double-negative cell population and a decrease in double-positive T cells, which may have been caused by failure of double-negative T cells to differentiate into double-positive cells. Alternatively, these two events may be unrelated and the decrease in double-positive cells may have been caused by increased elimination of self-antigen reactive doublepositive cells, since it is known that the intrathymic lifespan of double-positive cells is under 3 days and normally a majority of them die in the thymus [2]. Recently, Kotzin et al. [29] reported that the Zpr double-negative T cells had undergone repertoire modification resulting in elimination of their potential self-reactive VP specificities. These authors proposed that the double-negative T cells in Zpr mice were derived from a thymocyte population that expressed CD4 and CD8 antigens at one time in ontogeny. However, whether the repertoire selection may occur at the double-negative cell-stage in the Zpr strain remains an interesting possibility.

Thymocyte subsets in autoimmune mice

285

The fact that the altered thymocyte subpopulations were restricted to MRL-Zpr/ Zpr mice and not observed in C57BL/6-Zpr/Zpr animals suggested that background genes may play an important role. It has been well established that the onset and manifestations of the autoimmune disease induced by the lpr gene in mice having different background genes differs both in quality and severity (reviewed in [20]). Recently Budd et al. [30], studying the T-cell lineages in the thymus of C57BL/6-Zpr/ Zpr mice, observed that 4-6-month-old mice had similar or slightly altered proportions of the different subpopulations. The double-positive cells in these mice were 55 + 18% (mean + SD), double-negative 11 f 7%, CD4+ 28 + 10% and CDB’ cells 6 f 3%. These numbers fall within the range that we observed in the present study with C57BL/6-Zpr/Zpr mice. Also, earlier studies have revealed cells having abnormal surface phenotype, similar to those found in lymph nodes of MRL-Zpr/Zpr mice, to be present in the thymus of S JL-Zpr/lpr and MRL-Zpr/Zpr mice but not in C57BL/6-Zpr/Zpr animals [30]. These observations suggest that the C57BL/6-Zpr/Zpr strain may not represent a model strain to study the effect of the Zprgene and that the outcome of abnormalities in the thymus and the degree and severity of the disease may depend on the interaction of background genes and the Zprgene. It should be noted that, while removing the thymus from old Zpr mice, we took extreme care to avoid harvesting lymph nodes present in close proximity to the thymus. Thus, it was unlikely that the increase in the double-negative T-cell population in the present study was caused by lymph node T cells contaminating the thymocyte cell preparation. Histopathological sections of thymus from the MRLZpr/Zpr strain in the present study revealed no contaminating lymph node tissue (data not shown). A second possibility was that the double-negative lymph node cells were actively infiltrating the thymus from adjoining lymph nodes. Several investigators have demonstrated the presence of B220+ and Pgp-1 + cells in the thymus of MRLZpr/Zpr mice [28] and in the purified double-negative population of C57BL/6-Zpr/Zpr thymocytes [30]. However, the fact that such cells originate in the thymus and migrate to the periphery rather than infiltrate from the lymph node back to the thymus was suggested by the fact that the ‘abnormal’ markers, such as B220, Pgp-1 and Ly-6C present on the peripheral Zpr double-negative T cells, are also expressed by a minor subpopulation of normal double-negative thymocytes 130,311. Based on these observations, it has been proposed that the abnormal Zprperipheral T cell may actually represent a stage of normal T-cell development which expands enormously and is exported to the periphery [30,31]. Several observations support the view that the alterations in the thymocyte subsets seen in different autoimmune mice were not due to infiltration of the lymph node cells into the thymus: (a) we observed that there was no increase in the sIg+ cells in MRL-Zpr/Zpr thymus compared to MRL- + /+ thymus. (b) The C57BL/6-Zpr/Zpr mouse, in spite of moderate lymphadenopathy, failed to demonstrate any alterations in the thymocyte subsets. (c) In BXSB mice, the lymphadenopathy is caused by B-cell hyperplasia but still we detected increased percentages of double-negative T cells. (d) Infiltration of lymph node cells into the thymus could lead to thymic enlargement. In contrast, there is thymic atrophy, particularly of the cortex, and decrease in double-positive cells. (e) In old MRL- + /+ mice, in spite of the absence of abnormal T cells in the periphery, there was a significant increase in double-negative thymocytes. (f) Recent preliminary studies from our lab demonstrate that the double-negative T cells in the thymus of Zpr

286

V. N. Kakkanaiahet al.

mice can be activated by cross-linked anti-CD3 MoAbs, unlike the double-negative T cells in the lymph nodes which fail to respond to such a stimulation (unpublished data). Collectively, these observations suggest that the alterations in the thymocyte subpopulations in different autoimmune strains was caused by an intrinsic defect in thymocyte differentiation rather than by lymph node infiltration. The antibody Jlld has been shown to define a differentiation antigen found on immature thymocytes but not on mature functional peripheral CD4+ or CD8+ T cells [ 12, 131. Normal double-negative thymocytes contain J 11 d+ and J 11 d ~ subpopulations, of which J 11 d+ thymocytes express CD3y6 type of TCR and Jl 1dthymocytes express CD3aP TCR [15, 161. Normally the majority of the doublenegative cells are Jlld+ (-90%) and a small percentage (- 10%) are Jlld[13]. In the present study we observed that thymocytes of young MRL-lpr/lpr mice had normal distribution of these subpopulations. In contrast, in 4-6-month-old MRLlpr/lpr mice, the percentage of Jl Id- cells rose to - 50% and the proportion of Jl 1d+ to Jlldcells was - 1:l. In an earlier study, we observed that in the periphery, MRL-lpr/lpr double-negative cells are -35% Jlld’ [14]. This suggested that the Jl 1d- double-negative thymocyte subpopulation may be selectively expanded in old MRL-lpr/lpr mice and that this population may migrate along with the Jl Id+ subpopulation to the periphery. It is interesting to note that Budd et al. [30] made a similar observation using B2A2, i.e., that normal double-negative thymocytes contained - 90% B2A2+ cells, and in C57BL/6-lpr/lpr double-negative thymocytes, this proportion decreased to 45% with a consequent increase in B2A2- cells. Since C57BL/6-lpr/lpr peripheral double-negative T cells are B2A2-, these authors suggested that the double-negative B2A2- population may selectively expand and migrate to the periphery. Recently we have observed that in contrast to B2A2, J 11 d is expressed by a subpopulation of C57BL/6-lpr/lpr peripheral double-negative T cells (unpublished data). Similarly, Davidson et al. [32] reported the presence of Ml/69 on a significant number of old lpr lymph node cells. These findings, along with the observation that B2A2 and Ml/69 but not Jl Id, can identify distinct subsets among CD4+CD8thymocytes, suggests that the antibodies B2A2, Ml/69 and Jlld, although sharing many common properties, may exhibit distinct reactivity patterns (reviewed in [ 131). In conclusion, our data suggest that in several strains of autoimmune-susceptible mice there is a consistent alteration of T-cell ontogeny correlating with the onset and degree of lymphoid hyperplasia and autoimmune disease. Studies using several other thymocyte differentiation markers and interleukin receptors should provide additional insights into abnormal T-cell maturation in autoimmune mice. Acknowledgements We thank Ms Judith McCord for excellent and expert assistance with the breeding and maintenance of the MRL- + I+ and MRL-lpr/lpr mice colony, and Dr Sponenberg for the histological studies of thymus and lymph nodes. This work was supported in part by grants CA45009 and CA45010 from the National Institute of Health. References 1. Ceredig, R., D. P. Dialynas, F. W. Fitch, and H. R. MacDonald. 1983. Precursors of T cell growth factor producing cells in the thymus: ontogeny, frequency, and quantitative

Thymocyte

2.

3.

4.

5.

6.

7. 8. 9. 10.

Il.

12. 13. 14. 15.

16.

17. 18. 19.

20. 21.

22.

subsets

in autoimmune

mice

287

recovery in a subpopulation of phenotypically mature thymocytes defined by monoclonal antibody GK 1.5.J. Exp. Med. 158: 1654-1671 Scollay, R., P. Bartlett, and K. Shortman. 1984. T cell development in the adult murine thymus: Changes in the expression of the surface antigens Ly2, L3T4 and B2A2 during development from early precursor cells to emigrants. Zmmunol. Rev. 82: 79-103 Mathieson, B. J. and B. J. Fowlkes. 1984. Cell surface antigen expression on thymocytes: Development and phenotypic differentiation of intrathymic subsets. Zmmunol. Rev. 82: 141-173 Scollay, R., A. Wilson, A. D’Amico, K. Kelly, M. Egerton, M. Pearse, L. Wu, and K. Shortman. 1988. Developmental status and reconstitution of subpopulations of murine thymocytes. Zmmunol. Rev. 104: 81-120 Fowlkes, B. J., L. Edison, B. J. Mathieson, and T. M. Chused. 1985. Early T lymphocytes. Differentiation in vivo of adult intrathymic precursor cells. J. Exp. Med. 162: 802-822 Kingston, R., E. J. Jenkinson, and J. J. T. Owen. 1985. A single stem cell can colonize an embryonic thymus, producing phenotypically distinct T cell populations. Nature 317: 811-813 Smith, L. 1987. CD4+ murine T cells develop from CD8’ precursors in vivo. Nature 326: 798-800 Kappler, J. W., N. Roehm, and P. Marrack. 1987. T cell tolerance by clonal elimination in the thymus. Cell 49: 273-280 MacDonald, H. R., H. Hengarmer, and T. Pedrazzini. 1988. Intrathymic deletion of selfreactive cells prevented by neonatal anti-CD4 antibody treatment. Nature 335: 174-176 Kisielow, P., H. Bluthmann, U. D. Staerz, M. Steinmetz, and H. von Boehmer. 1988. Tolerance in T-cell-receptor trangenic mice involves deletion of nonmature CD4+8” thymocytes. Nature 333: 742-746 Teh, H. S., P. Kisielow, B. Scott, H. Kishi, Y. Uematsu, H. Bluthmann, and H. von Boehmer. 1988. Thymic major histocompatibility complex antigens and the c$ T-cell receptor determine the CD4/CD8 phenotype of T cells. Nature 335: 229-233 Bruce, J., F. W. Symington, T. J. McKearn, and J. Sprent. 1981. A monoclonal antibody discriminating between subsets of T and B cells. J. Zmmunol. 127: 2496-2501 Crispe, N. and M. J. Bevan. 1987. Expression and functional significance of the Jlld marker on mouse thym0cytes.J. Zmmunol. 138: 2013-2018 Seth, A., R. H. Pyle, M. Nagarkatti, and P. S. Nagarkatti. 1988. Expression of the Jl Id marker on peripheral T lymphocytes of MRL-lpr/Zpr mice. J. Zmmunol. 141: 1120-l 125 Ceredig, R., F. Lynch, and P. Newman. 1987. Phenotypic properties, interleukin-2 production, and developmental origin of a “mature” subpopulation of Lyt2-L3T4mouse thymocytes. Proc. Natl. Acad. Sci. USA 84: 8578-8582 Crispe, I. N., M. W. Moore, L. A. Hausmann, L. Smith, M. J. Bevan, and R. P. Shimonkevitz. 1987. Differentiation potential of subsets of CD4_CD8thymocytes. Nature 329: 336-339 Murphy, E. D. and J. B. Roths. 1977. A single-gene model for massive lymphoproliferation with autoimmunity in new mouse strain MRL. Fed. hoc. 36: 1246 Eisenberg, R. A., E. M. Tan, and F. J. Dixon. 1978. Presence of anti-Sm reactivity in autoimmune mouse strains. J. Exp. Med. 147: 582-587 Datta, S. K. and R. S. Schwartz. 1978. Genetic, viral, and immunologic aspects of autoimmune disease in NZB mice. In General control of Autoimmune Disease. N. R. Rose, P. E. Bigazzi, and N. L. Warner, eds. Elsevier/North-Holland, NY, pp. 193 Theofilopoulos, A. N. and F. J. Dixon. 1981. Etiopathogenesis of murine SLE. ZmmunoZ. Rev. 55: 179-216 Steinberg, A. D., J. B. Roths, E. D. Murphy, R. T. Steinberg, and E. S. Raveche. 1980. Effects of thymectomy or androgen administration upon the autoimmune disease of MRL/MP-lpr/lpr mice.J. Zmmunol. 125: 871-873 Smith, H. R., T. M. Chused, P. A. Smathers, A. D. Steinberg. 1983. Evidence for thymic regulation of autoimmunity in BXSB mice: Acceleration of disease by neonatal thymect0my.g. Zmmunol. 130: 1200-1204

288

V. N. Kakkanaiah et al.

23. East, J., M. A. B. DeSousa, and H. Jaquet. 1967. Consequences of neonatal thymectomy in NZB mice. Clin. Exp. Immunol. 2: 203-215 24. Roubinian, J. R., R. Papoian, and N. Talal. 1977. Effect of neonatal thymectomy and spleenectomy on survival and regulation of autoantibody formation NZB/NZW Fl mice. J. Immunol. 118: 1524-1529 25. Nagarkatti, P. S., M. Nagarkatti, and A. M. Kaplan. 1985. Normal Lytl+2- T cells have the unique capacity to respond to syngeneic autoreactive T cells: demonstration of a T cell network. J. Exp. Med. 162: 375-380 26. Nagarkatti, I?. S., M. Nagarkatti, L. W. Mann, L. A. Jones, and A. M. Kaplan. 1988. Characterization of an endogenous Lyt2+ T-suppressor-cell population regulating autoreactive T cells in vitro and in viva. Cell. Immunol. 112:64-77 27. Nagarkatti, M., A. Seth, and I’. S. Nagarkatti. 1988. Chemotherapy of mice bearing syngeneic tumors with 1,3-Bis(2-chloroethyl)-1-nitrosourea is effective only in normal but not in irradiated or nude mice: Role of L3T4+(CD4+) and Lytl’(CD8’) T cells. Cell Immunol. 115:383-392 28. Morse, H. C., H. W. Davidson, R. A. Yetter, E. D. Murphy, J. B. Roths, and R. L. Coffman. 1982. Abnormalities induced by the mutant gene lpr: expansion of a unique lymphocyte subset. J. Immunol. 129: 2612-2615 29. Kotzin, B., S. K. Babcock, and L. R. Herron. 1988. Deletion of potentially self-reactive T cell receptor specificities in L3T4- Lyt2- T cells of lpr mice.g. Exp. Med. 168: 2221-2229 30. Budd, R. C., M. Schreyer, G. C. Meischer, andH. R. MacDonald. 1987. T cell lineages in the thymus of Zpr/Zpr mice: evidence for parallel pathways of normal and abnormal T cell deve1opment.g. Zmmunol. 139: 2200-2210 31. Gause, W. C., R. D. Mounts, and A. D. Steinberg. 1988. Characterization and differentiation of CD4_CD8thymocytes sorted with the Ly-24 marker. J. Immunol. 140: l-7 32. Davidson, W. F., F. J. Dumont, H. G. Bedigian, B. J. Fowlkes, and H. C. Morse III. 1986. Phenotypic functional and molecular genetic comparisons of the abnormal lymphoid cells of C3H-lpr/lpr and C3H-gld/gld mice. 3. Immunol. 136: 4075-4084