Ex vivo characterization of human thymic dendritic cell subsets

ARTICLE IN PRESS

Immunobiology 212 (2007) 167–177 www.elsevier.de/imbio

Ex vivo characterization of human thymic dendritic cell subsets Nathalie Schmitta,, Marie-Christine Cumontb, Marie-The´re`se Nugeyrea, Bruno Hurtrelb, Franc¸oise Barre´-Sinoussia,, Daniel Scott-Algaraa, Nicole Israe¨la,1 a

Unite´ de Re´gulation des Infections Re´trovirales, Institut Pasteur, 25 rue du Dr Roux, 75724 Paris Cedex 15, France Unite´ de Recherche et d’Expertise Physiopathologie des Infections Lentivirales, Institut Pasteur, 25 rue du Dr Roux, 75724 Paris Cedex 15, France

b

Received 24 July 2006; received in revised form 20 December 2006; accepted 18 January 2007

Abstract Interactions between thymic dendritic cells (DC) and thymocytes are critical for proper development of T-cells. We identified human thymic DC populations on the basis of CD123, CD11c and CD14 expression. High levels of CD123 (IL-3R) and CD45RA defined the plasmacytoid DC (pDC) subset. Human thymic CD11c+ DC expressed CD45RO and myeloid-related markers (CD13, CD33 and CD11b). CD11c+ DC could be separated into two main subsets based on differential expression of CD14: CD11c+ CD14 and CD11c+ CD14+ cells. Spontaneous production of IL-10 and IFNg without exogenous stimulation, was observed in the three DC subsets. Important phenotype modifications were observed in pDC cultures supplemented with IL-3. A down-regulation of CD123 and appearance of myeloid markers such as CD11b and CD11c on CD45RA+ cells was noticed within the first 48 h; at a later time there was a shift from CD45RA to CD45RO expression, as well as appearance of CD14 expression. CD11c+ cells emerging in pDC culture did not express high levels of HLA-DR, CD83 and co-stimulatory molecules. This suggests an in vitro evolution of human thymic pDC toward a myeloid phenotype found in the CD11c+ subset of thymic DC. r 2007 Elsevier GmbH. All rights reserved. Keywords: Cytokines; Dendritic cells; Human; Thymus; Plasmacytoid DC

Introduction Thymic dendritic cells (DC) have specific functions compared to blood DC. Indeed they can present selfantigens and induce negative selection, the induction of apoptotic death in potentially self-reacting developing T-cells (Brocker et al., 1997). They may also be involved in additional tolerogenic mechanisms like the induction Corresponding authors. Tel.: +33 1 4568 8733; fax: +33 1 4568 8957. E-mail addresses: [email protected] (N. Schmitt), [email protected] (F. Barre´-Sinoussi). 1 This work is dedicated to the memory of Nicole Israel.

0171-2985/$ - see front matter r 2007 Elsevier GmbH. All rights reserved. doi:10.1016/j.imbio.2007.01.008

of regulatory T-cells. Human thymus contains distinct thymic DC populations. Human thymic CD11c+ DC are defined phenotypically as HLA-DR+ CD11c+ CD4+ CD45RAlow cells. However, the description of the different human thymic CD11c+ DC subsets varies depending on the isolation procedure used (BendrissVermare et al., 2001; Res et al., 1999; Schmitt et al., 2000; Vandenabeele et al., 2001). Two populations of CD11c+ CD14 DC with morphological and phenotypical differences have been described: one HLA-DRint CD83 CD86low DC-Lamp and the other HLADRhigh CD83+ CD86high DC-Lamp+ (Bendriss-Vermare et al., 2001). Thymic CD14+ DC were also described according to their phenotypes and functions

ARTICLE IN PRESS 168

N. Schmitt et al. / Immunobiology 212 (2007) 167–177

(Vandenabeele et al., 2001). This paper described two populations of CD11c+ DC: CD11b CD14 and CD11b+ CD14+. Studies on human thymic DC development suggest that, as in the mouse, many CD11c+ DC derive from an intrathymic lymphoidrestricted pathway (Marquez et al., 1998; Res et al., 1996). Some human thymic DC may also develop through a myeloid pathway (de Yebenes et al., 2002). Human thymus also contains plasmacytoid DC (pDC) sharing common functional properties with blood and secondary lymphoid organs pDC (BendrissVermare et al., 2001; Res et al., 1999; Schmitt et al., 2000; Vandenabeele et al., 2001). However, thymic pDC express higher levels of CD2, CD5 and CD7 than their peripheral counterparts (Res et al., 1999). Thymic pDC can be distinguished from CD11c+ DC as being HLADR+ CD45RA+ CD11c and CD123high. Thymic pDC also express lymphoid-related RNA transcripts, including pre-Ta and Spi-B (Bendriss-Vermare et al., 2001; Res et al., 1999). Upon activation with IL-3 and CD40L, human thymic pDC have been shown to differentiate in culture into mature CD11c+ CD14 DC expressing CD83 and co-stimulatory molecules CD80 and CD86 at high levels (Bendriss-Vermare et al., 2001; Res et al., 1999). However, these mature DC have not yet been identified in vivo. In the absence of CD40 stimulation, thymic pDC also acquire CD11c but express CD86 at low levels (Bendriss-Vermare et al., 2001). In the present study, we have used an approach that allows the isolation of a sufficient number of human thymic DC, permitting thorough characterization of their phenotypes and further study of their functions. We discriminated three major thymic DC populations based on CD123, CD11c and CD14 expression. In thymic pDC cultures supplemented with IL-3, the development of CD11c+ cells expressing low level of co-stimulatory molecules could be observed. Interestingly, these culture conditions led to the appearance of cells expressing CD11c as well as CD14, properties similar to freshly isolated human thymic CD11c+ DC.

Isolation of thymic DC subpopulations Cells were dissociated by gentle teasing. Non-dispersed cells were incubated with collagenase IV (5 mg/ ml, Sigma, St. Louis, MO) and DNase (150 U/ml, Sigma) for 1 h at 37 1C, with 5 mM EDTA being added in the last 10 min. Cells were subjected to Percoll gradient centrifugation (52%) and the low-density fraction, containing DC, was collected. The samples were immunodepleted using CD3, CD34, CD19 and CD56 mAbs (all from Beckman-Coulter) and antimouse IgG-coated magnetic beads (Dynal-Biotech, Oslo, Norway). CD11c+ DC were isolated using biotinylated CD11c mAb (BU15, Beckman-Coulter), anti-biotin-coated magnetic microbeads (Miltenyi-Biotec, Bergisch-Gladbach, Germany) and passage through a Miltenyi column. Residual CD14+ cells were then removed from the negative fraction using CD14 microbeads (Miltenyi-Biotec). pDC were isolated by positive selection for CD123 expression on CD11c CD14 cells using a biotinylated CD123 mAb (BDPharmingen, Mountain View, CA) and anti-biotincoated magnetic microbeads. The purity of DC populations obtained was about 80%.

Phenotypic analysis of thymic DC subsets Cells were incubated with a mixture of IgG1 (Beckman-Coulter), IgG2a (Beckman-Coulter) and hIgG (Tegeline, LFB, Courtaboeuf, France) for 30 min at 4 1C. Cells were then washed and labeled with mouse mAbs for 30 min at 4 1C. FITC-conjugated CD45RO, CD13, CD33; PE-conjugated CD2, CD4, CD11b, CD11c, CD14, CD34, CD45RA, CD83; Pc5-conjugated CD11c and CD14 mAbs were from Beckman-Coulter. FITC-conjugated CD14, PE-conjugated CD80, CD83, CD86, CD123, HLA-DR and PE-Cy5 conjugated CD45RA mAbs were from BD-Pharmingen. PE-conjugated HLA-ABC mAb was from Dako (Glostrup, Denmark). Immunostaining was analyzed with an XL4C cytofluorometer (Beckman-Coulter).

Immunohistochemistry

Materials and methods Thymuses Fresh thymus fragments were obtained during elective cardiac surgery (Hoˆpital Necker, Paris, France and Marie Lannelongue, Le Plessis Robinsson, France) on children (age range, 6 days–24 months). As human thymi are surgical wastes according to the French law regulating the usage of human tissues, informed consent from individuals was not required.

Sections (4 mm) of cryopreserved OCT-embedded thymus fragments were analyzed by immunohistochemistry. The sections were incubated in blocking buffer for 15 min at room temperature and then with IgG1 or mAbs specific for CD11c, CD14, CD83, DC-Lamp, CD34 (Beckman Coulter), CD123 (BD-Pharmingen) for 90 min at room temperature. After rinsing in PBS, the sections were exposed to biotinylated horse anti-mouse immunoglobulin (Vector Laboratories, Burlingame, CA) in blocking buffer for 30 min at room temperature and then incubated with ABC-AP (Vector Laboratories)

ARTICLE IN PRESS N. Schmitt et al. / Immunobiology 212 (2007) 167–177

for 30 min at room temperature. Immunostaining was revealed using Fast-Red in the presence of Naphthol and Levimasole (Sigma). Finally, slides were counterstained in Mayer’s Hematoxylin, rinsed in water and mounted in Aquamount.

Cell culture DC subsets were cultured in 96-well plates (2.5  105 cells per well in 200 ml) in RPMI 1640, 10% FCS, 1 mM L-glutamine, 10 mM HEPES, antibiotics and 20 ng/ml IL-3 (R&D Systems, Minneapolis, MN). Half of the supernatants were collected every 2–3 days and wells were replenished with fresh medium.

Cell stimulation and measurement of cytokine production Thymic DC were stimulated for 16 h with 5 mg/ml CpG-C ODN, 5 mg/ml control ODN (Duramad et al., 2003; Marshall et al., 2003) (Invitrogen, Paisley, UK) or with 100 ng/ml dsRNA complex polyinosinic:polycytidylic acid (poly(I:C)) (Fujimoto et al., 2004) (Sigma). Brefeldin A (Sigma) was added to culture 1 h after stimulation at a low dose (2 mg/ml) as previously described (Scott-Algara et al., 2003). The cells were then harvested, stained with mAbs against CD11c, CD14 or CD123, fixed and permeabilized with the Cytofix/Cytoperm kit (BD-Pharmingen). Finally, they were stained with PE-conjugated control Ig or PEconjugated mAbs (BD-Pharmingen) directed against IL10, IL-12, TNFa or IFNg. Cells were analyzed with an XL-4C cytofluorometer (Beckman-Coulter).

A

B

103 102

169

Results Identification of three thymic DC populations To identify different thymic DC populations, we first studied the expression of DC markers on thymic cells. Percoll cell fractions were analyzed for surface expression of the thymic DC markers (Bendriss-Vermare et al., 2001; Vandenabeele et al., 2001). Cells were subject to triple staining for CD11c, CD123 and CD14 and then analyzed by flow cytometry. Distinct subsets expressing these molecules could be detected in the low-density cell fraction (Fig. 1). Cells expressing high levels of CD123 (CD123high) were CD11c and CD14 (Fig. 1A and B), and were thus considered to be pDC. Cells expressing CD11c were CD123low/ (Fig. 1A) and could be divided into two main subpopulations on the basis of CD14 expression: CD11c+ CD14 DC and CD11c+ CD14+ DC (Fig. 1C) (Vandenabeele et al., 2001). All cells expressing CD14 were CD11c+ CD123low/ (Fig. 1B and C).

Detailed phenotypic characterization of thymic DC subsets To further characterize these populations, we developed a multistep cell isolation method based on CD11c, CD14 and CD123 expression. After depletion of B cells, NK cells, CD34 progenitors and CD3+ thymocyte subsets from the low-density cell fraction, we isolated CD11c+ cells by positive selection yielding CD11c+ CD14 and CD11c+ CD14+ DC. In a second step, pDC (CD123high) were obtained from the CD11c CD14 fraction by positive selection of CD123 cells.

C

103

103 102

102

101 2.8%

100 10-1 10-1

100

101

CD123 – PE

102

103

101

5%

2.8%

100 10-1 10-1

CD11c – PC5

CD14 – FITC

CD11c – PC5

6.8%

0

1

10 10 CD123 – PE

10

2

10

3

1.8%

5%

101 100 10-1 10-1

100 101 102 CD14 – FITC

103

Fig. 1. Human thymus contains pDC and CD11c+ DC subsets. Distinct thymic DC subsets can be detected in the low-density percoll cell fraction. Thymic pDC express a high level of CD123 and are negative for CD11c and CD14 staining. Thymic CD11c+ DC express CD123 at a low level and can be separated into two main subsets on the basis of CD14 expression: CD11c+ CD14 and CD11c+ CD14+. Results from one experiment representative of four performed with cells from different donors are shown.

ARTICLE IN PRESS 170

N. Schmitt et al. / Immunobiology 212 (2007) 167–177

Due to the previous selection/depletion of CD11c+ CD14+/ CD123low/ cells, only cells expressing high levels of CD123 were selected during the second step. The purity of the DC populations obtained was about 80% as reported by others (Schmitt et al., 2000). We then analyzed in detail the surface phenotype of the three freshly isolated thymic DC subsets (Fig. 2). HLADR and HLA-ABC were expressed at the same moderate level on pDC and CD11c+ CD14+ DC and strongly expressed on CD11c+ CD14 DC. pDC expressed high levels of CD45RA but not CD45RO, in contrast to CD11c+ DC, which expressed CD45RO and little to no CD45RA. CD11b could not be detected on pDC but was expressed heterogeneously on CD11c+ CD14 DC and strongly on CD11c+ CD14+ DC.

CD13 and CD33 were only expressed on CD11c+ DC. CD86 was expressed on CD11c+ CD14+ DC and on pDC but at a lower level on pDC than on CD11c+ CD14+ DC. By contrast, CD86 could only be detected slightly on CD11c+ CD14 DC. CD80 and CD83 were either weakly expressed or could not be detected on thymic DC, depending on the donors. All thymic DC subsets expressed high levels of CD4.

Localization of DC populations in thymus We explored the localization of thymic DC by immunohistochemistry on thymus sections (Fig. 3). CD123, CD11c and CD14 cells were mainly detected

HLA-DR

HLA-ABC

CD45RA

CD45RO

CD11b

CD13

CD83

CD86

CD80

CD4

CD2

CD34

CD33

pDC

CD11c+ CD14-DC

CD11c+ CD14+DC

pDC

CD11c+ CD14-DC

CD11c+ CD14+DC

Fig. 2. Thymic DC subsets express low levels of co-stimulatory molecules and differ by their expression of myeloid-related markers. Thymic pDC express CD45RA and are negative for myeloid-related markers expression. CD11c+ CD14 and CD11c+ CD14+ DC express CD45RO but differ by the expression level of myeloid-related markers and HLA molecules. All three populations express low levels of CD83 and co-stimulatory molecules. The black area represents expression of the specified marker and the gray area represents control Ig. Similar results were obtained in seven independent experiments.

ARTICLE IN PRESS N. Schmitt et al. / Immunobiology 212 (2007) 167–177

IgG

CD123

171

CD11c

CD14 C

M

x40

M

M

C

M

C C

HC M

HC

M

C

x100

HC

C

C

M

M

DC-Lamp

C

HC

CD34

CD83

C

M

C

M

x40

M C

M

M HC

x100

HC

HC

M

C C

C

Fig. 3. Human thymic DC subsets are mainly located in the medulla. The medulla (M) is identified by the presence of Hassall’s corpuscles (HC) and a lower cell density than the cortex (C). Original magnification:  40 and  100 (as indicated).

in the medulla, although we also detected a few cells in the cortex, capsule and septum regions. This cell distribution is consistent with the expression of the specific DC markers DC-LAMP and CD83, which were detected only in the medulla. We detected a weak expression of CD83 by immunohistochemistry which is consistent with the flow cytometry data on isolated thymic DC. As expected, we found CD34 progenitors principally in the capsule and septum regions.

Evolution of the thymic pDC phenotype into CD11c+ CD14+ cells in vitro We next investigated the evolution of the thymic DC populations in vitro by monitoring phenotypic changes over time during culture in IL-3 supplemented medium. As previously reported (Bendriss-Vermare et al., 2001; Grouard et al., 1997; Kohrgruber et al., 1999), IL-3 allowed a substantial proportion (25–30%) of pDC to be maintained in culture (Fig. 4A). The same culture

conditions allowed both CD11c+ CD14 and CD11c+ CD14+ DC subsets to survive in culture (not shown). This confirmed differences in culture requirements between thymic CD11c+ DC and monocyte-derived DC (Saunders et al., 1996). The spontaneous production of cytokines, primarily by CD11c+ CD14+ DC (see below and Fig. 7), may contribute to this survival without the need for exogenous cytokines (except IL-3). We observed large changes of cell phenotype in pDC cultures (Fig. 4B). At day 8 of culture, 70–80% of the cells expressed markers such as CD11c, CD11b and CD45RO and about 40% of the cells were CD14+. By contrast, pDC markers such as CD123 and CD45RA were down-regulated. The decrease in CD123 expression was not related to the presence of IL-3 in the culture, as the CD123 antibody used binds to an epitope that remains available after the binding of IL-3 to its receptor (Sun et al., 1996). The surface phenotypes of CD11c+ CD14 DC and CD11c+ CD14+ DC subsets did not significantly change after 8 days of culture (not shown).

ARTICLE IN PRESS 172

N. Schmitt et al. / Immunobiology 212 (2007) 167–177

A 2.0E+05 1.8E+05 1.6E+05

Cell count

1.4E+05 1.2E+05 1.0E+05 8.0E+04 6.0E+04 4.0E+04 2.0E+04 0.0E+00 0

1

2

3

4 Days of culture

5

6

7

8

B CD123

CD14

CD11c

CD11b

CD45RA

CD45RO

CD13

Day 0

Day 8

Fig. 4. Cells expressing myeloid-related markers can be detected after day 8 in pDC cultures supplemented with IL-3. (A) Culture of thymic pDC in IL-3 supplemented medium allows the survival of a substantial proportion of cells. Results are expressed as mean7SD of values from four independent replicates. (B) Important cell phenotype modifications can be detected in pDC cultures supplemented with IL-3. At day 8, an important proportion of cells expresses CD11b, CD11c, CD14, CD13 and CD45RO while CD123 and CD45RA expression is strongly decreased. The black area represents expression of the specified marker and the gray area is control Ig. Similar results were obtained in four independent experiments.

To confirm that cells expressing CD11c and/or CD14 derived from pDC and not from contaminant cells such as CD34+ progenitors, we analyzed the markers expressed by cultured pDC at various times between 0 and 8 days (Fig. 5). On day 0, 77% of the cells exhibited a typical pDC phenotype (CD123high CD45RA+ CD11b CD11c CD14 CD45RO ) (Fig. 5A). Contaminating cells expressed CD123, CD11b or CD11c at different levels, suggesting that they were a heterogeneous cell population; most of these cells were undetectable after 1 day of culture. CD34 was detected on less than 5% of cells during the kinetic period of culture. On day 1, the CD45RA positive cells began to downregulate CD123 (Fig. 5B). Of the cells still expressing CD45RA, 19% expressed a lower level of CD123. On day 2, down-regulation of CD123 increased (occurring in 52% of cells) (Fig. 5C). Moreover, CD11b and CD11c were detected on 48% and 27%, respectively, of the CD45RA+ cells which are down-regulating CD123. At this time, we observed an isolated population of cells expressing CD45RO, high levels of CD2 and negative

for CD11b, CD11c, CD14, CD45RA and CD34 staining. These cells, which may be a thymocyte subset proliferating in culture (Fukuhara et al., 2002), did not appear to be derived from pDC and could not be detected at later times. On day 4, 94% of cells expressed low levels of CD123 whereas CD11b and CD11c were expressed by 66% and 59% of the cells, respectively (Fig. 5D). At this time, we detected a switch from CD45RA to CD45RO expression on CD123low/ CD11b+ CD11c+ cells. We also observed CD13 appearance on the CD45RA cells (not shown). CD14 (about 7%) appeared on CD123low/ CD11b+ CD11c+ CD45RA CD45RO+ CD13+ cells. On day 6, we observed a greater increase of cells expressing CD11b (84%), CD11c (70%), CD45RO (52%), CD13 (61%) and CD14 (25%) (Fig. 5E). Finally, after 8 days, we could detect 37% of CD14+ CD11b+ CD11c+ CD123low/ CD45RO+ CD45RA CD13+ cells (Fig. 5F). These data suggest that during in vitro culture with IL-3, human thymic CD123high CD45RA+ cells under-

ARTICLE IN PRESS

13

34

0.8

42

32

CD45RA-PC5

67

26

6.3

1.1

67

25

31

26

39

4.0

CD14-PE +

23

4.7

12

CD45RA-PC5 55

29

16

0

1.1

60

38

CD45RA-PC5

60

9.8

11

20

23

27

3.0

CD14-FITC 31

21

26

21

CD45RA-PC5

67

8.4

0.7

19

72

0.3

84

9.0

6.8

69

0.7

3.7

77

17

19

47

4.0

30

61

5.8

34

0.1

0.2

4.8

21

74

12

6.9

12

5.1

12

25

CD45RA-PC5 37

1.1

42

20

CD45RA-PC5

3.7

1.7

58

37

CD14-FITC 30

38

29

2.7

CD14-PE

38

43

19

0.1

0.6

71

25

CD2-PE

2.3

47

2.7

48

19

40

4.4

37

54

5.2

40

0.3

9.9

37

11

43

CD45RA-PC5

69

6.6

9.7

14

CD45RA-PC5 36

41

20

3.9

CD14-FITC

CD14-FITC 3.1

71

CD14-FITC

CD45RA-PC5 CD45RA-PC5 58

24

CD45RA-PC5

CD45RA-PC5

69

1.4

CD45RA-PC5

CD14-FITC

CD14-PE

1.7

92

CD45RA-PC5

CD14-FITC 24

CD11c-PE

CD11b-PE CD34-PE

CD123-PE

F

6.2

4.3

CD11c-PE

66

3.4

CD11c-PE

28

0.2

3.0

CD123-FITC

CD2-PE

5.9

CD11b-PE

7.4

CD45RA-PC5

Day 8

48

0.6

68

CD45RA-PC5

CD11b-PE

CD123-PE 11

CD45RA-PC5

CD14-FITC 0.6

CD45RO-FITC CD123-PE

15

CD45RA-PC5

61

77

CD123-PE

CD34-PE

CD45RA-PC5

21

CD45RA-PC5 CD123-PE

73

0.1

23

CD123-FITC

CD11c-PE

5.8

CD14-FITC CD45RO-FITC

25

2.1

2.3

4.9

CD11c-PE

7.2

D

CD45RA-PC5 2.6

19

CD11b-PE

1.5

0.1

Day 4

22

4.3

4.3

CD2-PE

45

CD45RA-PC5

CD14-FITC

39

CD45RA-PC5 CD123-PE

CD45RO-FITC

Day 6

2.8

CD14-FITC

52

76

CD11b-PE

E

38

19

23

CD14-FITC

CD45RA-PC5

CD123-PE

CD45RA-PC5

21

0.3

27

0.7

CD34-PE

74

CD123-FITC

80

2.7

0.3

CD45RO-FITC

21

25

47

77

CD123-PE

5.0

CD2-PE

CD34-PE

0.2

45

0.1

4.1

CD45RA-PC5

CD45RA-PC5 20

38

CD14-FITC

18

10

31

11

173

CD45RA-PC5

CD45RO-FITC

31

CD123-FITC

19

18

Day 1

CD14-FITC

15

CD123-PE

36

CD11c-PE

CD11b-PE

CD45RA-PC5

48

86

CD123-PE

52

CD11c-PE

CD45RA-PC5

2.8

10

B

CD45RO-FITC

28

88

0.8

CD45RO-FITC

19

9.0

CD11c-PE

Day2

0.6

0.5

CD11c-PE

C

81

CD123-PE

CD123-PE

CD14-FITC

18

2.4

3.2

CD45RA-PC5

CD2-PE

0

0.4

CD34-PE

17

0.9

CD11b-PE

0.1

87

CD45RA-PC5

CD11b-PE

83

CD45RO-FITC

CD123-PE

CD45RA-PC5

7.3

CD2-PE

4.0

1.1

CD11c-PE

13

4.7

CD45RO-FITC

77

CD11b-PE

Day0

6.2

CD34-PE

A

CD11b-PE

CD123-PE

N. Schmitt et al. / Immunobiology 212 (2007) 167–177

CD45RA-PC5

43

13

32

13

CD45RA-PC5

+

Fig. 5. Apparition of CD14 CD11c cells in pDC culture supplemented with IL-3 is characterized by a progressive and sequential acquisition of myeloid-related markers. After 1 day of culture, pDC start down-regulating CD123. At day 2, CD11b and CD11c expression can be detected on CD45RA+ cells down-regulating CD123. From day 4, cells start shifting from CD45RA to CD45RO expression and expressing CD14. A representative experiment is shown.

went progressive and sequential phenotype modifications. These modifications included a down-regulation of typical plasmacytoid markers CD123 and CD45RA and appearance of myeloid-related markers CD11b, CD11c, CD14 and CD13 like CD45RO. Previous studies showed that differentiation of thymic pDC into CD11c+ DC following activation with

CD40L and IL-3 is associated with HLA-DR, CD80, CD83 and CD86 up-regulation (Bendriss-Vermare et al., 2001; Res et al., 1999). We thus determined expression levels of these molecules on CD11c+ CD14+ cells detected in pDC cultured with IL-3 alone. We did not detect a significant increase of HLA-DR, CD80, CD83 and CD86 on CD11c+ CD14+ cells present in pDC

ARTICLE IN PRESS 174

N. Schmitt et al. / Immunobiology 212 (2007) 167–177

HLA-DR

CD86

CD80

CD83

Day 0

Day 7

Day 14

Fig. 6. Apparition of CD11c+ CD14+ cells in pDC culture is not associated with up-regulation of classical maturation markers. HLA-DR, CD86, CD80 or CD83 are not strongly expressed on CD14+ cells emerging in pDC cultures supplemented with IL-3. The black area represents expression of the specified marker on pDC (day 0) and on CD14+ cells from pDC cultures (day 7 and 14). Gray area represents control Ig. Similar results were obtained in three independent experiments.

cultures on day 7 compared to freshly isolated pDC (Fig. 6). On day 14, we detected a moderate increase in CD86 expression on these cells, whereas the expression of the other markers was not significantly altered. We also conducted experiments in order to identify intermediate phenotypes between pDC and CD11c+ CD14+ cells in thymic tissue ex vivo. Hence, we positively selected CD11b+ cells, as this molecule is absent on freshly isolated pDC and is one of the earlier markers of the phenotypic switch detected during their in vitro culture (Fig. 5C). As expected, CD11b+ HLADRint CD14 cells co-expressed CD4, CD11c, CD45RA and CD123low, and did not express CD34, CD56, CD20 or CD10 (not shown). The identification of CD14 HLA-DR+ CD4+ cells expressing CD45RA, CD11b and CD11c in human thymic tissue ex vivo may support the hypothesis of an evolution of pDC toward more mature CD11c+ CD14+ cells in vivo.

Flow cytometry profiles showed that, in the presence of control ODN, thymic DC spontaneously produced cytokines only at low levels, as shown by staining intensities (Fig. 7). CD11c+ CD14+ DC were found to produce IL-10 and IFN-g while pDC and CD11c+ CD14 DC produced these cytokines only marginally. IL-12 was produced by CD11c+ CD14+ DC but not by pDC or CD11c+ CD14 DC. TNFa was not produced from any DC subsets. After CpG-C or poly(I:C) stimulation, we observed different levels of cytokine production depending on the donor (not shown). However, these exogenous stimuli did not consistently increase the amount of cytokine-producing DC compared to unstimulated cells (data not shown), unlike for blood DC (Fujimoto et al., 2004; Lore et al., 2003; Marshall et al., 2003).

Discussion Cytokine production by thymic DC We next investigated the ability of each thymic DC subset to produce cytokines with or without stimulation.

In this study, we used a new method for the isolation of whole DC populations to define three thymic DC subsets on the basis of CD123, CD11c and CD14

ARTICLE IN PRESS N. Schmitt et al. / Immunobiology 212 (2007) 167–177

pDC

CD11c+CD14- DC

175

CD11c+CD14+ DC

IL-10

IL-12

TNFα

IFNγ

Fig. 7. Thymic DC spontaneously produce IL-10 and IFN-g Flow cytometer profiles obtained for unstimulated thymic DC are shown. Histograms are gated on CD123high (pDC), CD11c+ CD14 or CD11c+ CD14+ cells; gray lines represent Ig staining and black lines represent positive staining. Similar results were obtained in three independent experiments.

expression (pDC, CD11c+ CD14 DC and CD11c+ CD14+ DC) and determine their phenotypes and anatomical locations. We have also shown that CD123low/ CD11c+ CD14+ cells could be detected in pDC cultures supplemented with IL-3. Interestingly, these cells expressed low levels of maturation molecules and presented a phenotype similar to that of freshly isolated thymic CD11c+ CD14+ DC. Our results confirm and extend previous descriptions of thymic DC subsets based on different technical approaches (Bendriss-Vermare et al., 2001; Res et al., 1999; Schmitt et al., 2000; Vandenabeele et al., 2001). In contrast to these studies, CD83 was not detected on freshly isolated or cultured CD11c+ CD14 DC although it could be weakly detected by immunohistochemistry in thymic tissue. CD86 expression on these cells was also lower than previously reported (BendrissVermare et al., 2001; Vandenabeele et al., 2001). These variations in the expression of maturation molecules may reflect differences in the thymic DC isolation procedures since CD86 is known to be constitutively expressed at low levels on DC and rapidly up-regulated following activation. Relatively low CD86 levels observed on thymic DC are consistent with their physiological role, as thymic DC have been shown to induce negative selection and are not involved in specific response processes. CD11c+ DC had a broad range of staining intensity with CD11c and CD14 mAbs. The CD11c expression level on CD11c+ CD14 DC correlated to CD13, CD33 and CD45RO expression levels, while CD14+ cells expressing higher CD11c and

CD14 levels also expressed slightly higher CD13, CD33 and HLA-DR levels (not shown). This may suggest that higher CD11c expression on thymic DC is associated with a more differentiated state. We found that expression of the specific DC markers DC-LAMP and CD83, as well CD11c, CD14 and CD123, was restricted to thymic medulla. This is consistent with previous studies showing that DC-LAMP, CD86 and CD40 cells are mainly located in thymic medulla (Bendriss-Vermare et al., 2001; Vandenabeele et al., 2001). We observed that freshly isolated thymic DC and especially CD11c+ CD14+ DC produced IL-10 and IFN-g in the absence of specific stimulation. Also, factors able to stimulate blood DC (Lore et al., 2003) did not further enhance cytokine production by thymic DC. Our results suggest that thymic DC are permanently stimulated by factors present in the thymic microenvironment. It is reasonable to assume that freshly isolated thymic DC can maintain their functional characteristics acquired in the thymic micro-environment during short-term cultures. Indeed, it has been shown that cytokines act as factors or cofactors of thymocyte proliferation (Zlotnik and Moore, 1995), differentiation and maturation in the thymus (Suda and Zlotnik, 1992; Waanders et al., 1992; Zuniga-Pflucker et al., 1995). Thus, a permanent production of cytokines would be consistent with the physiological function of thymic DC. The lack of response by thymic DC to classic stimuli such as CpG-C and poly(I:C) seems consistent with the lack of physiological involvement of these cells in inflammation and ‘‘specific’’ response

ARTICLE IN PRESS 176

N. Schmitt et al. / Immunobiology 212 (2007) 167–177

processes. The capacity of thymic DC to produce IL-10 may suggest that these cells play a role in establishing thymic T-cell tolerance. However, we cannot exclude the possibility that the isolation procedure and/or the disruption of the thymic DC microenvironment alters their functional properties. In this study, we cultured pDC in the presence of IL3, which has been shown to rescue pDC from spontaneous apoptosis (Bendriss-Vermare et al., 2001; Grouard et al., 1997; Kohrgruber et al., 1999). We avoided the use of exogenous stimuli such as GM-CSF or CD40L, which are known to accelerate the maturation of blood, tonsil or thymic pDC without significantly increasing their viability (Bendriss-Vermare et al., 2001; Grouard et al., 1997; Kohrgruber et al., 1999; Vandenabeele et al., 2001). In our conditions, the pDC phenotype gradually evolved over 8 days toward a phenotype similar to that of CD11c+ CD14+ DC (Fig. 5). The gradual loss of CD123 expression and the acquisition of several myeloid markers by CD45RA+ cells before the down-regulation of CD45RA are consistent with a shift from pDC toward CD11c+ CD14+ cells. However, we cannot definitively exclude that the CD11c+ CD14+ cells may derive from a small population of contaminating cells. A follow-up cell count did not reveal significant cell expansion in pDC cultures. CD34 expression was only detected at low level on a small proportion of cells (o5%), and this expression remained stable over time (Fig. 5). Moreover, DC generation from thymic CD34+ CD1a progenitors has been shown to require SCF and IL-7 or TNFa while IL-3 does not increase significantly the yield of DC (Dalloul et al., 1999; Res et al., 1996). Generation of CD14+ cells from CD34+ CD1a thymic progenitors can only be detected upon exogenous M-CSF addition (Dalloul et al., 1999). These CD34-derived DC are characterized by high HLA-DR, CD83 and CD86 expression. Thymic CD34+ CD1a+ progenitors do not generate DC in culture. It is thus unlikely that CD11c+ CD14+ cells emerging in IL-3 supplemented pDC cultures had derived from CD34+ precursors. We have previously shown that IFNa secretion following HIV-1 infection reaches maximum levels in pDC cultures at day 6 while no significant secretion of IFNa can be detected in CD11c+ DC cultures (Schmitt et al., 2006). This suggests that pDC undergoing phenotypic modifications can transiently preserve some of their functional characteristics. Previous studies on thymic pDC cultured in the presence of IL-3, IL-3 and CD40L, or IL-3, CD40L and GM-CSF (Bendriss-Vermare et al., 2001; Res et al., 1999; Vandenabeele et al., 2001) also report a decrease in CD123 and CD45RA expression and the appearance of CD11c and CD13. CD83, CD80 and CD86 are strongly up-regulated on pDC after culturing with CD40L. However, cells co-expressing CD83 and

CD123 cannot be detected in thymic tissue (Res et al., 1999). After culturing with IL-3 alone, we observed a decrease of CD123 and CD45RA expression as well as appearance of myeloid-related markers CD11c, CD11b, CD13, CD33 and CD45RO previously described upon CD40L and IL-3 stimulation. Moreover, we observed that culture of thymic pDC only with IL-3 led to appearance of CD14 on CD11c+ CD11b+ CD45RO+ CD123low cells. Interestingly, these cells did not express high levels of classical maturation markers. This suggests that, upon culture with IL-3, pDC acquires a phenotype similar to that of CD11c+ CD14+ DC isolated from thymic tissue. This hypothesis is supported by detection in thymic tissue of cells with phenotype similar to the one observed during in vitro culture of pDC (i.e. CD123low/ CD11b+ CD11c+ CD45RA+ CD14 ) (data not shown). Interestingly, a significant proportion of CD11c+ cells emerging in pDC culture did not express CD14 at day 8; whether these cells are destined to acquire CD14 later remains to be determined. In conclusion, we have established robust isolation and characterization procedures for fresh ex vivo thymic DC subsets. Further studies are necessary to define the underlying mechanisms of human thymic pDC differentiation as well as to determine the function of each thymic DC subset within the thymic microenvironment and, in particular, their implication in establishing T-cell tolerance.

Acknowledgments We thank Dr. Sonia Berrih-Aknin (Hoˆpital Marie Lannelongue, Le Plessis-Robinson, France) and Professor Leca (Hoˆpital Necker, Paris, France) for providing us with thymi. We thank Gianfranco Pancino, Anne Hosmalin and Matthew Albert for the critical reading of the manuscript and helpful discussion, and Francine Brie`re for useful information. We thank Gerard Zurawski and Kimberly Gehlbach for proofreading the manuscript. N. Schmitt was successively the recipient of fellowships from the French Ministry of Education and Research (MENESR) and from Sidaction.

References Bendriss-Vermare, N., Barthelemy, C., Durand, I., Bruand, C., Dezutter-Dambuyant, C., Moulian, N., Berrih-Aknin, S., Caux, C., Trinchieri, G., Briere, F., 2001. Human thymus contains IFN-alpha-producing CD11c( ), myeloid CD11c(+), and mature interdigitating dendritic cells. J. Clin. Invest. 107, 835–844. Brocker, T., Riedinger, M., Karjalainen, K., 1997. Targeted expression of major histocompatibility complex (MHC)

ARTICLE IN PRESS N. Schmitt et al. / Immunobiology 212 (2007) 167–177

class II molecules demonstrates that dendritic cells can induce negative but not positive selection of thymocytes in vivo. J. Exp. Med. 185, 541–550. Dalloul, A.H., Patry, C., Salamero, J., Canque, B., Grassi, F., Schmitt, C., 1999. Functional and phenotypic analysis of thymic CD34+CD1a- progenitor-derived dendritic cells: predominance of CD1a+ differentiation pathway. J. Immunol. 162, 5821–5828. de Yebenes, V.G., Carrasco, Y.R., Ramiro, A.R., Toribio, M.L., 2002. Identification of a myeloid intrathymic pathway of dendritic cell development marked by expression of the granulocyte macrophage-colony-stimulating factor receptor. Blood 99, 2948–2956. Duramad, O., Fearon, K.L., Chan, J.H., Kanzler, H., Marshall, J.D., Coffman, R.L., Barrat, F.J., 2003. IL-10 regulates plasmacytoid dendritic cell response to CpGcontaining immunostimulatory sequences. Blood 102, 4487–4492. Fujimoto, C., Nakagawa, Y., Ohara, K., Takahashi, H., 2004. Polyriboinosinic polyribocytidylic acid [poly(I:C)]/TLR3 signaling allows class I processing of exogenous protein and induction of HIV-specific CD8+ cytotoxic T lymphocytes. Int. Immunol. 16, 55–63. Fukuhara, K., Okumura, M., Shiono, H., Inoue, M., Kadota, Y., Miyoshi, S., Matsuda, H., 2002. A study on CD45 isoform expression during T-cell development and selection events in the human thymus. Hum. Immunol. 63, 394–404. Grouard, G., Rissoan, M.C., Filgueira, L., Durand, I., Banchereau, J., Liu, Y.J., 1997. The enigmatic plasmacytoid T cells develop into dendritic cells with interleukin (IL)-3 and CD40-ligand. J. Exp. Med. 185, 1101–1111. Kohrgruber, N., Halanek, N., Groger, M., Winter, D., Rappersberger, K., Schmitt-Egenolf, M., Stingl, G., Maurer, D., 1999. Survival, maturation, and function of CD11c- and CD11c+ peripheral blood dendritic cells are differentially regulated by cytokines. J. Immunol. 163, 3250–3259. Lore, K., Betts, M.R., Brenchley, J.M., Kuruppu, J., Khojasteh, S., Perfetto, S., Roederer, M., Seder, R.A., Koup, R.A., 2003. Toll-like receptor ligands modulate dendritic cells to augment cytomegalovirus- and HIV-1specific T cell responses. J. Immunol. 171, 4320–4328. Marquez, C., Trigueros, C., Franco, J.M., Ramiro, A.R., Carrasco, Y.R., Lopez-Botet, M., Toribio, M.L., 1998. Identification of a common developmental pathway for thymic natural killer cells and dendritic cells. Blood 91, 2760–2771. Marshall, J.D., Fearon, K., Abbate, C., Subramanian, S., Yee, P., Gregorio, J., Coffman, R.L., Van Nest, G., 2003. Identification of a novel CpG DNA class and motif that optimally stimulate B cell and plasmacytoid dendritic cell functions. J. Leukoc. Biol. 73, 781–792.

177

Res, P., Martinez-Caceres, E., Cristina Jaleco, A., Staal, F., Noteboom, E., Weijer, K., Spits, H., 1996. CD34+CD38dim cells in the human thymus can differentiate into T, natural killer, and dendritic cells but are distinct from pluripotent stem cells. Blood 87, 5196–5206. Res, P.C., Couwenberg, F., Vyth-Dreese, F.A., Spits, H., 1999. Expression of pTalpha mRNA in a committed dendritic cell precursor in the human thymus. Blood 94, 2647–2657. Saunders, D., Lucas, K., Ismaili, J., Wu, L., Maraskovsky, E., Dunn, A., Shortman, K., 1996. Dendritic cell development in culture from thymic precursor cells in the absence of granulocyte/macrophage colony-stimulating factor. J. Exp. Med. 184, 2185–2196. Schmitt, C., Fohrer, H., Beaudet, S., Palmer, P., Alpha, M.J., Canque, B., Gluckman, J.C., Dalloul, A.H., 2000. Identification of mature and immature human thymic dendritic cells that differentially express HLA-DR and interleukin-3 receptor in vivo. J. Leukoc. Biol. 68, 836–844. Schmitt, N., Nugeyre, M.T., Scott-Algara, D., Cumont, M.C., Barre-Sinoussi, F., Pancino, G., Israel, N., 2006. Differential susceptibility of human thymic dendritic cell subsets to X4 and R5 HIV-1 infection. AIDS 20, 533–542. Scott-Algara, D., Truong, L.X., Versmisse, P., David, A., Luong, T.T., Nguyen, N.V., Theodorou, I., Barre-Sinoussi, F., Pancino, G., 2003. Cutting edge: increased NK cell activity in HIV-1-exposed but uninfected Vietnamese intravascular drug users. J. Immunol. 171, 5663–5667. Suda, T., Zlotnik, A., 1992. In vitro induction of CD8 expression on thymic pre-T cells. II. Characterization of CD3-CD4-CD8 alpha+cells generated in vitro by culturing CD25+CD3-CD4-CD8- thymocytes with T cell growth factor-beta and tumor necrosis factor-alpha. J. Immunol. 149, 71–76. Sun, Q., Woodcock, J.M., Rapoport, A., Stomski, F.C., Korpelainen, E.I., Bagley, C.J., Goodall, G.J., Smith, W.B., Gamble, J.R., Vadas, M.A., Lopez, A.F., 1996. Monoclonal antibody 7G3 recognizes the N-terminal domain of the human interleukin-3 (IL-3) receptor alphachain and functions as a specific IL-3 receptor antagonist. Blood 87, 83–92. Vandenabeele, S., Hochrein, H., Mavaddat, N., Winkel, K., Shortman, K., 2001. Human thymus contains 2 distinct dendritic cell populations. Blood 97, 1733–1741. Waanders, G.A., Hugo, P., Boyd, R.L., 1992. Treatment of fetal thymic organ culture with IL-1 leads to accelerated differentiation of subsets of CD4-CD8- cells. Cell Immunol. 144, 237–245. Zlotnik, A., Moore, T.A., 1995. Cytokine production and requirements during T-cell development. Curr. Opin. Immunol. 7, 206–213. Zuniga-Pflucker, J.C., Di, J., Lenardo, M.J., 1995. Requirement for TNF-alpha and IL-1 alpha in fetal thymocyte commitment and differentiation. Science 268, 1906–1909.