Developmentaland ComparativeImmunology,Vol. 14, pp. 335-346, 1990 Printed in the USA. All rights reserved.
0145-305X/90 $3.00 + .00 Copyright © 1990 Pergamon Press plc
ONTOGENY OF LANGERHANS CELLS: PHENOTYPIC DIFFERENTIATION FROM THE BONE MARROW TO THE SKIN Anne de Fraissinette, Colette Dezutter-Dambuyant, Daniel Schmitt, and Jean Thivolet Laboratory of Dermatologyand Immunology,INSERMU 209, Hopital E. Herriot, Lyon, France (Submitted May 1.989;Accepted September 1989) FqAbstractiWe have performed double im- major accessory cells of the Skin Assomunolabelings for cytofluorometric analysis ciated Lymphoid Tissue (SALT) (1,2). and electron microscopy to investigate the The CDla and CDlc antigens, common coexpression of the CDla (OKT6 and DMC1 antigens expressed by LC and cortical monoclonal antibodies) antigen and the prothymocytes, are frequently used as surmonocyte/monocyte differentiation antigens face markers to study LC in human skin CD14 (My4) or CD33 (My9) on putative bone (3,4). LC comprise 2%-4% of all epimarrow and umbilicalcord blood precursors of dermal cells (1). the Langerhans cells (LC) and the epidermal The bone marrow origin of LC was LC. By cytofluorometric analysis, the percentage of CDla+ cells which coexpressed the first demonstrated by Katz et al. (5) in CD33 antigen was different from the bone mice. In humans, the bone marrow ormarrow (5% of CD33+ cells are CDla+), to igin of LC has been established by Volcthe cord blood (3% of the CD33 + are CDla+) Platzer et al. (6) and Perreault et al. (7,8) and to the epidermis (the whole population of after bone marrow transplantation from CD33+ LC are CDla+). The ultrastructural a male to a female. The finding of low morphology of the CDla-expressing bone numbers of LC in untreated aplastic marrow, cord blood cells closelyapproximated anemia is suggestive of a medullary orthat of a promonocyte/monocyte.Only LC epiigin of LC (7). In vitro, the studies of dermal were specificallyrecognized by the inGoordyal (9) and Gothelf (10) demontracytoplasmic Birbeck granules. These CDla+/CD33 + or CD14 + subpopulations strated the presence of a small fraction found in three different locations (epidermis, of CDI positive cells expressing the bone marrow, cord blood) display a similar CD14 antigen among the myelomonoquantitative expression of the CD14 and CD33 cyte lineage. After 8 days in methylcelantigens. lulose, we observed 8% CDla positive cells by indirect immunofluorescence rqKeywords--Langerhans cells; bone marrow; and cytofluorometric analysis. This cord blood; epidermis ontogeny. CDla promonocyte/monocyte subpopulation expressed the following differenIntroduction tiation antigens: CD14, CD33, CD4, HLA-DR, and HLA-DP. No T lymphoid The Langerhans cells (LC) are migratory and thymocyte differentiation antigens bone marrow-derived cells with a spewere observed on these cells and no Bircific intracytoplasmic organelle, the Birbeck granules were found in their cytobeck granule. They express HLA-DR plasm (11). (MHC class II) molecule, and thereby, The presence of 1% CDIa positive antigen-presenting cells represent the cells in normal peripheral blood of adults and infants and 4% in neonates suggest Address correspondence to Pr. J. Thivolet, that LC migrate from the bone marrow INSERM U 209, Pav.R, Hopital E. Herriot, to the epidermis via the blood vessels 69437 Lyon Cedex 03, France. (4,12,13). This number of CD1 + cells in 335
336
A. de Fraissinette et al.
blood increases up to 42% in patients whose cutaneous surface has been partially destroyed by traumatism or burn (10,13,14). Our study was undertaken to establish the phenotype of putative blood and medullar C D l a + precursors of LC and to correlate these results with those obtained on the epidermal LC. We have investigated the phenotype using double immunolabelings which were performed: (i) by immunofluorescence technique and cytofluorometric analysis, and (ii) by immunoelectron microscopy for a quantitative study of three antigens: CDla, CD33, and CD14. This phenotype was confirmed by immunoprecipitation on epidermal LC compared with a promonocyte lineage, U937.
Umbilical Cord Blood U m b i l i c a l c o r d b l o o d from five normal infants delivered at term were isolated by centrifugation on Ficoll-Hypaque gradient.
Epidermal LC Cell suspensions were obtained by trypsinization of seven normal human skin specimens taken during plastic surgery. LC enrichment was obtained by Ficoll-Hypaque sedimentation.
Monoclonal Antibodies All cell suspensions were investigated with two simultaneous monoclonal antibodies (Mab). Their specificities are summarized in Table I.
Materials and Methods Bone Marrow Samples
Bone Marrow Cell Culture Bone marrow samples from 5 normal donors, after informed consent, were obtained from anterior iliac crests and aspirated into heparin-containing flasks. The mononuclear cells were separated on a F i c o l l - H y p a q u e ( P h a r m a c i a , France) density gradient.
Cells were plated at 105 nucleated cells/mL in 0.8% methylcellulose containing Iscove modified Dulbecco's culture m e d i u m (IMDM) (Boeringher, Mannheim, West Germany) supplemented with 30% Fetal Calf Serum
Table 1. Monoclonal antibodies used for the characterization of bone marrow and cord blood progenitors and epidermal LC.
Antibody
Isotype
Cluster of Differentiation (CD)
Fluorescein Conjugated OKT6
IgG1
CDla
Phycoerythrin Conjugated T6
IgG1
CDla
Phycoerythrin or Fluorescein My9
IgG2b
CD33
Myelomonocytes Monocytes
Coulter USA
Phycoerythrin or Fluorescein My4
IgG1 k
CD14
Promonocytes Monocytes
Coulter USA
Specificity
Source
70%-80% human cortical thymocytes and LC
Ortho USA
Coulter USA
Ontogeny of Langerhans cells
337
(FCS) and 100UI/mL Granulocyte-Monocyte Colony Stimulating Factor (GMCSF, Boeringher, Mannheim, West Germany). Cultures were incubated at 37°C in water-saturated air containing 5% CO2 for 8 days. Erythroid bursts (BFU-E) were used as negative controls and were cultured under identical conditions (11).
Cytofluorometric Analysis: Double Immunofluorescence Labeling of Mononuclear Cells and Epidermal Cells Mononuclear cells and epidermal cells were dispersed and stained with a primary specific CDla Mab conjugated with fluorescein (OKT6) (1/20 dilution) and the second specific CD33 (My9) or CD14 (My4) Mab conjugated with phycoerythrin (1/20 dilution)• Negative iso-
BONE
. . . .
~
. . . .
"1
type controls, including incubation with the mouse myeloma IgGs from clarified ascites MOPC21 (IgG1) and MOPCI41 (IgG2b), were carried out in parallel. All i n c u b a t i o n s w e r e p e r f o r m e d in RPMI-1640 supplemented with 10% FCS (FCS-RPMI) at 4°C for 1 h each monoclonal antibody. After each incubation, the pellet was washed twice in cold 10% FCS-RPMI. At the end, the pellet was postfixed with 1% paraformaldehyde in Phosphate Buffered Saline (PBS). Flow cytometry was performed on a cytofluorograph 5OH (Ortho Instruments). The evaluation of percentages of labeled cells was done by setting a threshold on the histogram by comparison with background fluorescence of cells treated with normal mouse serum at a 1/1000 dilution. Determinations were based on 104 cells per sample, selected by the corresponding gate (FAS/RAS).
MARROW CELL
. . . . .
-,t
•
CULTURE
. . m
CD 1 • FITC
CONTROL FITC
A -
CONTROL
PE
CD 14
PE
- - - " I
-
- - v - 1
CO 33
"
-"~'-Ir
. . . .
PE
Figure 1. Cytofluorometric analysis of bone marrow-cultured cells stained with C D l a (OKT6 Mab), CD14 (My4 Mab), and CD33 (My9 Mab). Results are representative of five experiments.
338
A. de Fraissinette et al.
CD1
•,.-- CD33
Figure 2.
Ultrastructural morphology of the bone marrow CD1 a/CD14 + or CD33 + precursors: 2a
and 2b: CD33 (My9) promonocyte differentiation antigen with, 2a: x 8,200 and 2b: x 70,000. 2c and
2d: CD14 (My4) monocyte differentiation antigen with, 2c: x 20,000 and 2d: x 55,000.
Ultrastructural Immunocytochemistry : Double Immunogold Labeling of Mononuclear Cells and Epidermal LC Mononuclear and epidermal cells were h a r v e s t e d and washed in 10% FCS-RPML, then incubated with a primary specific C D l a (DMC1) (15) Mab conjugated with 5nm or 17nm gold granules (dilution 1/5) and the second
specific CD33 (My9) or CD14 (My4) Mab conjugated with 17nm or 5nm gold granules (dilution 1/5) at 4°C for 1 h. After washing with 10% FCS-RPMI, the cells were fixed for 1 h with 2% glutaraldehyde in cacodylate buffer, then for 20 min with 1% osmium tetroxide and emb e d d e d in E p o x y medium. Ultrathin sections were examined, after poststaining with lead citrate and uranyl ace-
Ontogeny of LangerhanS cells UMBILICAL
339 CORD
BLOOD
CELLS
--I CONTROL
PE
I=1 W n~
..I CO
33
PE
I=1 Ld cv --1 CONTROL
FITC
Id I.d c¢ ¢d)
Z CD
la
FITC
W Ld rv
Figure 3. Cytofluorometric analysis of umbilical cord blood monocytes stained with CDla (OK'I'6 Mab) and CD33 (My9 Mab). Results are representative of five experiments.
tate with a Philips EM300 electron microscope. All the solutions were filtered through Millipore filter (0.22 ~m mesh). As a negative isotype control, labelings with the mouse myeloma IgGs from clarified a s c i t e s M O P C 2 1 (IgG1) and MOPC141 (IgG2b) were performed.
lmmunochemical Analysis of Epidermal LC and Mononuclear U937 Cell Line L C - e n r i c h e d epidermal cells and U937 cells were surface labeled with 125I by lactoperoxidase-catalyzed iodination
(16). Cells were washed in PBS and in lysis buffer (150mM NaCI, 30mM TrisHCI p H 7.4, 5mM Naz-EDTA, 1% NP-40, lmM PMSF, and 5mM Iodoacetamide) for 30 min on ice. The lysate supernatant was removed after microfuging (14,000 g) and stored at -70°C. Specific immunoprecipitation was performed using a modification of immunoisolation technique described by Tamura (17). Labeled antigens were electrophoresed with a denaturation electrophoresis sample buffer for one-dimensional analysis. Sodium dodecyl sulfate polyacrylamide electrophoresis (SDS-PAGE) was carried out on discontinuous ver-
340
A. de Fraissinette et al.
4a 4b
4c 4d
~33
E)la
Figure 4. Ultrastructural morphology of the double-labeled CD1a/CD14 + or CD33 + umbilical cord blood cells: 4a and 4b: CD14 (My4) monocyte differentiation antigen with, 4a: x8,200 and 4b: x33,000. 4c and 4d: CD33 (My9) promonocyte differentiation antigen with, 4c: x 6,800 and 4d: x 33,000.
tical slab gels according to a modification of Laemmli's method. Gradient gels were made in 7.5% to 15% acrylamide. Relative mass markers used were 14Cmethylated proteins (Amersham): phos-
phorylase b (Mr 32,500), bovine serum albumin (Mr 69,000), ovalbumin (Mr 46,000), c a r b o n i c a n h y d r a s e (Mr 30,000), trypsin inhibitor (Mr 21,500), and lysozyme (Mr 14,000).
Ontogeny of Langerhans cells
341
Results
Bone Marrow Cell Cultures
After 8 days in culture, bone marrow cells were analyzed by cytofluorometry and different subpopulations. The monocyte-like population was selected with the corresponding gate (FAS/RAS). In these conditions, the cultured cell promonocyte/monocyte population was homogeneous and presented a high granular density in FAS/RAS. In this sub-population 6 __- 1% of cells were C D l a + (OKT6), 68 --- 8% were CD14 (My4), and 73 +-- 5% were CD33 (My9) by immunolabeling. By double immunolabeling, the C D l a + cells were found C D I 4 + and C D 3 3 + , approximately 10% of the CD14+ monocytes coexpressed C D l a antigen and 5% of the CD33 + promonocytes coexpressed the CDla antigen. Data were based on phen o t y p i c analysis f r o m five h e a l t h y donors (Fig. 1). By double immunogold labeling, we observed that 30% of the CD14 or CD33
positive promonocyte/monocyte cells coexpressed the CDla antigen. The increase in percentage of CDla +/CD14 + and C D l a + / C D 3 3 + cells which was noted with immunogold method was due to the higher sensitivity of this method (gold granules of small size-5nm) when compared with the cytofluorometric analysis. This allowed quantitation of double labeled (one-step labeling per Mab) cultured bone marrow cells: CD14 antigen (70 +_ I0 gold-5nm granules/cell section with My4 Mab); CDIa antigen (7 +_3 gold-17nm granules/cell section with DMCI Mab); CD33 antigen (10 ___ 4 gold-17nm granules/cell section with My9 Mab); and CDIa antigen (80 _ 8 gold-5nm granules/cell section with DMC1 Mab) (Fig. 2).
Umbilical Cord Blood
By cytofluorometry and double immunolabeling analysis gating on the monocyteqike population, we obtained
E P I D E R M A L L A N G i ~ I H A N S EELL-ENRKHED SUSPENSION
a
"J
(ONTROLFITc
st I
W I'~
t.
(ONTROL ~C ""
1111. . . .
Ill
"'
'Ill "" Ill rlCHIlm=
CONTROL PE
OCHAH 4?
Ill
""
"ll~ ""
~Ol
TOT.Lm ~
A
P
"' IE32
E K- ~2
CD l a FITC ' ' ~lll
-l
" " |l
....
i~ ....
III
""
"Ill
""
Ill
""
tl
10.
.-J
T O T . L - 42S9 P [ . K m 359 IICHRH 44
d
w ~J~ 4) . . . .
Lo
_J i:=1 W I~
P¢AK= I I I
w w ~ll . . . . . .
till TllTllkm 41loll PEAK" l l l l l l OCHAH 4 l l
LOll " " " I l q
c
TOTAL" | I l l
I~'
W W IP~ (.D
( D 33 F I T (
PE
""
~
=
;
G9 e ~" |¢ -I
(D la
b
TllTALs PEAK= 119? OCHAH 3 1
S~ . . . .
Oil ' ' "ICK ' " ~ +
FICH [ R I
T
' " ~l y++|
CONTROL PE
TOTAL= 1 7 1 4 P[AK= 1 3 1
~J
ICHAH 1 6
w ~
.... tl
- +l.
$~ ....
1++ ' '
w o~
~J+ ''
~60
,0T.~- ,=,, I
..J
CD 14 PE
'15+ ' '
FICHIK • Yll3
3 h
ICHRN tO
:t ....
=. ....
ttA
''
"tE~' : it+
** =E=
Figure 5 (a,b,c,d). Cytofluorometric analysis of epidermal enriched-LC selected by the corresponding gate (FAS/RAS) and stained with CDla (OKT6 Mab), CD33 (My9 Mab), or CD14 (My4 Mab). Results are representative of seven experiments. Labelings (a, b) are observed on one normal skin sample and (c, d) on another normal skin specimen.
A. de Fraissinette et aL
342
Figure 6. Ultrastructural morphology of the double-labeled CD1aJCD14+ or CD33+ epidermal enriched-LC: 6a and 6b: CD33 (My9) promonocyte differentiation antigen with, 6a: x 10,000 and 6b: x70,000. 6c and 6d: CD14 (My4) monocyte differentiation antigen with, 6c: ×8,200 and 6d: x 20,000.
an enrichment of 75 _ 7% CD33 + cells. Approximately less than half of the C D l a + cell population (7 _-_ 3% of Ficoil-enriched cord blood cells) was found to be CD33+. This C D l a + cell subpopulation represented 3% of the CD33 + umbilical cord blood monocytes (Fig. 3). The quantitation of double-labeled umbilical cord blood cells was: C D 1 4 a n t i g e n (30 ___ 9 5 n m - g o l d granules/cell section with My4 Mab), C D l a a n t i g e n (10 _+ 5 17nm-gold
granules/cell section with DMC1 Mab). Similarly, the cells coexpressed CD33 antigen (5 -- 4 17nm-gold granules/cell section with My9 Mab) and CDla antigen (35 __ 10 5nm-gold granules/cell section with DMC 1 Mab) (Fig. 4).
Epidermal LC In epidermal cell suspensions fractioned by Ficoll-Hypaque sedimenta-
Ontogeny of Langerhans cells
343
tion, LC accounted for approximately 30% (Fig. 5). All the normal human epidermal C D l a positive selected by the corresponding gate (FAS/RAS) LC coexpressed the CD33 (My9 Mab) and CD14 (My4 Mab) surface antigens by cytofluorometry and immunoelectron microscopy. The CDIa, CD33, and CD14 antigens were never observed on the epidermal keratinocytes which served as negative cell controls (Fig. 6). The quantitation of double-labeled epidermal cells was: CD14 antigen (30 _ 7 5nm-gold granules per cell section with My4 Mab) and CDla antigen (100 _ 18 17nm-gold granules per cell section with
O tO
kD
92.5
59
45
30
=,
3
QD
"13
-
~ =
kD
21.5-
9 2 . 5 --
69 . - ~ Q
46
30
-Q
~
Im mtBII
w
DMC1 Mab); CD33 antigen (30 _+ 9 17nm-gold granules per cell section with My9 Mab) and CDla antigen (300 ___ 50 5nm-gold granules per cell section with DMC1 Mab) (Figs. 7, 8). The effect of the trypsinization to obtain LC-enriched cell suspensions could alter the native CD14 surface antigen and then prevent its detection.
-g
14.3-
qm~
7. Immunoprecipitation results obtained from 1251-labeled LC-enriched epidermal cells presented the following pattern: the CD33 molecule (Mr 67,000) was precipitated by P 67-5; D 3HL-60; WM-54 Mabs from the epidermal lysate, no specific CD14 molecule (53-55, 000) could be Immunoprecipitated from the epidermal LC. Figure
Figure 8. Immunoprecipitation results obtained from 12Sl-labeled U937 cell line presented the f o l l o w i n g pattern: the CD33 molecule (Mr 67,000) was precipitated according to the same conditions of the epidermal LC lysate, the CD14 molecule (53-55,000) was precipitated by My4, Cris6, MOP9 Mabs from the U937 cell lysate.
Discussion In this study, we demonstrated that within the limits of sensitivity of the three following techniques: cytofluorometry, immuno-electron microscopy, and
344
A. de Fraissinette et al.
SKIN
/:7
D LINEAGE
I
I~::
/c¢ ........... .%°, My LOCYTE LI AG~
I @ ...... I IQBICTTE
.....
MOr~ .
. . . .
i
T Hy~IoCyTE
i~I'AGE
$11PDpILJT~ID
e
,
LINEAGE
Figure 9. Schematic representation of suggested migration of the CDla+ monocyte subpopulation from the bone marrow to the epidermis.
immunobiochemistry, human normal epidermal LC are the only resident epidermal cells which react with monocyte d i f f e r e n t i a t i o n antigens (CD33 and CD14). These data confirm those reported by Wood et al. (1984) on normal Langerhans cells CD14 + (18) and Groh et al. (1988) on histiocytosis X CD14 + cells (19). The CD33 antigen described only on bone marrow and blood mononuclear cells was expressed also on all normal human epidermal LC. The extremely restricted distribution of the CD33 molecule on myeloid cells, particularly progenitors, strongly suggests a role in the regulation of myeloid differentiation and maturation (20). The expression of this CD33 antigen on the C D l a + bone marrow, C D I a + cord blood (putative LC precursors), and the CDla + epidermal LC could implicate a parental relationship b e t w e e n these monocyte lineage subpopulations and the dendritic epidermal LC. Interestingly, culture of Ficoll-selected bone marrow cells in the presence of GM-CSF, enhanced the division of typical monocytic phenotype cells. This favors the CD33+ and CD14+ population where 5% of them coexpressed the C D I a antigen. In GM-CSF cultured bone marrow cells, all the C D l a + expressed both phenotype and ultrastruc-
ture of promonocytic/monocytic cells. In comparison, only 3% of the CD33 + or CD14 + umbilical cord blood cells expressed CDla antigen. This study agrees with the presence of two C D l a + populations in cord blood (one monocyte-like and the other one lymphocyte-like) described by Dezutter-Dambuyant, 1986. The small C D l a + / C D 1 4 + or CD33+ cord cells displayed a typical promonocyte/monocyte ultrastructure. This small subpopulation could follow their own maturation entering the bloodstream and perhaps epidermis where the influence of environment remains to be clarified. The Birbeck granules which are specific markers of epidermal LC appear after their entrance into the epidermis without loss of the promonocyte/monocyte differentiation antigens (CDI4, CD33). When we compared the phenotype and the ultrastructural morphology of bone marrow C D I a + (11), cord blood C D l a + promonocytes/monocytes and the epidermal LC, we observed a similar, semiquantitative expression of the CD14 and CD33 antigens, a comparative monocyte morphology but only a difference concerning the Birbeck granule which is only detected regularly in the LC or experimentally in cord blood (4). There is no doubt that LC are derived from a bone marrow, CDla promono-
Ontogeny of Langerhans cells
345
cyte/monocyte subpopulation (I0,11) which could represent the best candidate for LC putative precursors and then migrate via blood vessels to the skin (Fig. 9). The i m m u n o b i o c h e m i s t r y of LC, when compared with the U937 (promonocyte) cell line, confirms the coexpression of CD33 and CDla on these two cell lineages; the trypsinization of epidermal cells did not affect CD33 antigen immunoprecipitation. Furthermore, the effect of trypsinization to obtain LC-enriched suspensions could alter the native CD14 surface antigen and its detection by immunoprecipitation when compared with an untrypsinized, promonocyte U937 cell population. By the sensitive methods (cytofluorometry, immuno-electron microscopy, immunobiochemistry) we suggest that the epidermal LC belongs to a C D l a +
monocyte cell subpopulation. The presence of Birbeck granules only in the epidermal LC could be induced by the epidermal microenvironment. In mice, after 3 days in c u l t u r e , loss of B i r b e c k granules, the F4/80 (antimacrophage) and 2.4G2 (anti Fc receptors) antigens and their cytochemical activities like nonspecific esterase and membrane ATPase (21) suggests that LC are precursors or immature elements of the dendritic cell system.
Acknowledgements--We thank G. Panaye for technical assistance, Dr. Audra and Dr. Guyotat for supplying human umbilical cord blood and human normal bone marrow samples, and Centre de Microscopie Electronique (CMEABG), Universif6 Lyon I for the electron microscopy processing. This work was partially supported by Universit6 de Biologie Humaine, Universit6 Lyon I.
References 1. Rowden, G. The Langerhans cell. CRC Crit. Rev. Immunol. 3:95-180; 1981. 2. Streilen, J. W. Circuits and signals of the skin associated lymphoid tissue (SALT). J. Invest. Dermatol. 85:10s; 1985. 3. Fithian, E.; Kung, E; Golstein, G.; Rubenfeld, M.; Fenoglio, C.; Edelson, R. Reactivity of Langerhans with hybridoma antibody. Proc. Natl. Acad. Sci. USA 78:2541-2544; 1981. 4. Dezutter-Dambuyant, C.; Schmitt, D.; Thivolet, J. Comparative phenotypic and ultrastructural characteristics of OKT6 positive cells in normal peripheral blood and in epidermis. Dev. Comp. Immunol. 10:571-586; 1986. 5. Katz, I. S.; Tamaki, K.; Sachs, D. Epidermal Langerhans cells are derived from cells originating in bone marrow. Nature. 282:324-326; 1979. 6. Volc-Platzer, B.; Stingl, G.; Wolff, K.; Hinterber, W. Schnedi, W. Cytogenetic identification of allogeneic epidermal Langerhans cells in a bone marrow graft recipient (abs). N. Engl. J. Med. 310:1125; 1984. 7. Perreault, C.; Pelletier, M.; Landry, D.; Gyger, M. Study of Langerhans cells after allogeneic bone marrow transplantation. Blood. 63:807811; 1984. 8. Perreault, C.; Pelletier, M.; Belanger, R.; Montplaisir, S.; Boilean, J.; Bonny, Y.; David, M.; Gyger, M.; Lacombe, M. Persistence of host Langerhans cells (LC) following bone
9.
10.
11.
12.
13.
14.
15.
marrow transplantation (BMT). Blood. 64:219; 1984. Goordyal, P.; Isaacson, P. G. Immunocytochemical characterization of monocyte colonies of Langerhans cells and interdigitating reticulum cells. J. Pathol. 146:189-195; 1985. Gothelf, Y.; Sharon, N.; Gazit, E. Fractionation of human bone marrow mononuclear cells with peanut agglutinin: phenotypic characterization with monoclonal antibodies. Human Immunol. 17:37-44; 1986. de Fraissinette, A.; Schmitt, D.; DezutterDambuyant, C.; Guyotat, D.; Zabot, M. T.; Thivolet, J. Culture of putative Langerhans cell bone marrow precursors: characterization of their phenotype. Exp. Hematol. 16:764768; 1988. Ortaldo, J. R.; Sharrow, S. O.; Timonent, R.; Heberman, R. B. Determination of surface antigens on highly purified human NK cells by flow cytometry with monoclonal antibodies. J. Immunol. 127:2401-2409; 1980. Wood, J.; O'Mahony, J. B.; Palder, S. B.; Rodrick, M. L.; D'eon, P.; Mannick, J. A. Circulating T6 antigen positive cells. J. Invest. Dermatol. 82:387; 1984. Gothelf, Y.; Hanau, D.; Tsur, H.; Sharon, N.; Sahar, E.; Cazenave, J. P.; Gazit, E. T6 positive cells in the peripheral blood of burn patients: are they Langerhans cells precursors? J. Invest. Dermatol. 90:142-148; 1988. Dezutter-Dambuyant, C.; Schmitt, D.; Sta-
346
quet, M. J.; Gaucherand, M.; Cambazard, E; Thivolet, J. DMCI: a monoclonal antibody produced from Histiocytosis X cells which reacts with the native CDIa molecule of human epidermal LC. Hybridoma 8:199-208; 1989. 16. Knowles, R. W.; Bodmer, W. E A monoclonal antibody recognizing a human TL-like antigen. Eur. J. Immunol. 12:676-681; 1982. 17. Tamura, G. S.; Dailey, M. O.; Gallatin, M.; McGrath, M. S.; Weissman, I. L.; Pillemer, E. A. Isolation of molecules recognized by monoclonal antibodies and antisera: the solid phase immunoisotation technique. Anal. Biochem. 136:458-464; 1984. 18. Wood, G. S.; Morhenn, V. B.; Butcher, E. C.; Kosek, J. Langerhans cells react with pan-leu-
A. de Fraissinette et al.
kocyte monoclonal antibody: ultrastructural documentation using a live suspension immunoperoxidase technique. J. Invest. Dermatol. 82:322-325; 1984. 19. Groh, V.; Gadner, H.; Radaszkiwicz, T.; Rappersberger, K.; Konrad, K.; Wolff, K.; Stingl, G. The phenotypic spectrum of Histiocytosis X cells. J. Invest. Dermatol. 90:441-447; 1988. 20. Favaloro, E. J.; Bradstock,. K. E; Kabral, A.; Grimsley, P.; Berndt, M. C. Characterization of monoclonai antibodies to the human myeol~d differentiation antigen, "gp67" (CD33). Disease markers 5:215-225; 1987. 21. Schuler, G.; Steinman, R. M. Murine epidermal Langerhans cells mature into potent immunostimulary dendritic cells in vitro. J. Exp. Med. 161:526-546; 1985.
0145-305X/90 $3.00 + .00 Copyright © 1990 Pergamon Press plc
ONTOGENY OF LANGERHANS CELLS: PHENOTYPIC DIFFERENTIATION FROM THE BONE MARROW TO THE SKIN Anne de Fraissinette, Colette Dezutter-Dambuyant, Daniel Schmitt, and Jean Thivolet Laboratory of Dermatologyand Immunology,INSERMU 209, Hopital E. Herriot, Lyon, France (Submitted May 1.989;Accepted September 1989) FqAbstractiWe have performed double im- major accessory cells of the Skin Assomunolabelings for cytofluorometric analysis ciated Lymphoid Tissue (SALT) (1,2). and electron microscopy to investigate the The CDla and CDlc antigens, common coexpression of the CDla (OKT6 and DMC1 antigens expressed by LC and cortical monoclonal antibodies) antigen and the prothymocytes, are frequently used as surmonocyte/monocyte differentiation antigens face markers to study LC in human skin CD14 (My4) or CD33 (My9) on putative bone (3,4). LC comprise 2%-4% of all epimarrow and umbilicalcord blood precursors of dermal cells (1). the Langerhans cells (LC) and the epidermal The bone marrow origin of LC was LC. By cytofluorometric analysis, the percentage of CDla+ cells which coexpressed the first demonstrated by Katz et al. (5) in CD33 antigen was different from the bone mice. In humans, the bone marrow ormarrow (5% of CD33+ cells are CDla+), to igin of LC has been established by Volcthe cord blood (3% of the CD33 + are CDla+) Platzer et al. (6) and Perreault et al. (7,8) and to the epidermis (the whole population of after bone marrow transplantation from CD33+ LC are CDla+). The ultrastructural a male to a female. The finding of low morphology of the CDla-expressing bone numbers of LC in untreated aplastic marrow, cord blood cells closelyapproximated anemia is suggestive of a medullary orthat of a promonocyte/monocyte.Only LC epiigin of LC (7). In vitro, the studies of dermal were specificallyrecognized by the inGoordyal (9) and Gothelf (10) demontracytoplasmic Birbeck granules. These CDla+/CD33 + or CD14 + subpopulations strated the presence of a small fraction found in three different locations (epidermis, of CDI positive cells expressing the bone marrow, cord blood) display a similar CD14 antigen among the myelomonoquantitative expression of the CD14 and CD33 cyte lineage. After 8 days in methylcelantigens. lulose, we observed 8% CDla positive cells by indirect immunofluorescence rqKeywords--Langerhans cells; bone marrow; and cytofluorometric analysis. This cord blood; epidermis ontogeny. CDla promonocyte/monocyte subpopulation expressed the following differenIntroduction tiation antigens: CD14, CD33, CD4, HLA-DR, and HLA-DP. No T lymphoid The Langerhans cells (LC) are migratory and thymocyte differentiation antigens bone marrow-derived cells with a spewere observed on these cells and no Bircific intracytoplasmic organelle, the Birbeck granules were found in their cytobeck granule. They express HLA-DR plasm (11). (MHC class II) molecule, and thereby, The presence of 1% CDIa positive antigen-presenting cells represent the cells in normal peripheral blood of adults and infants and 4% in neonates suggest Address correspondence to Pr. J. Thivolet, that LC migrate from the bone marrow INSERM U 209, Pav.R, Hopital E. Herriot, to the epidermis via the blood vessels 69437 Lyon Cedex 03, France. (4,12,13). This number of CD1 + cells in 335
336
A. de Fraissinette et al.
blood increases up to 42% in patients whose cutaneous surface has been partially destroyed by traumatism or burn (10,13,14). Our study was undertaken to establish the phenotype of putative blood and medullar C D l a + precursors of LC and to correlate these results with those obtained on the epidermal LC. We have investigated the phenotype using double immunolabelings which were performed: (i) by immunofluorescence technique and cytofluorometric analysis, and (ii) by immunoelectron microscopy for a quantitative study of three antigens: CDla, CD33, and CD14. This phenotype was confirmed by immunoprecipitation on epidermal LC compared with a promonocyte lineage, U937.
Umbilical Cord Blood U m b i l i c a l c o r d b l o o d from five normal infants delivered at term were isolated by centrifugation on Ficoll-Hypaque gradient.
Epidermal LC Cell suspensions were obtained by trypsinization of seven normal human skin specimens taken during plastic surgery. LC enrichment was obtained by Ficoll-Hypaque sedimentation.
Monoclonal Antibodies All cell suspensions were investigated with two simultaneous monoclonal antibodies (Mab). Their specificities are summarized in Table I.
Materials and Methods Bone Marrow Samples
Bone Marrow Cell Culture Bone marrow samples from 5 normal donors, after informed consent, were obtained from anterior iliac crests and aspirated into heparin-containing flasks. The mononuclear cells were separated on a F i c o l l - H y p a q u e ( P h a r m a c i a , France) density gradient.
Cells were plated at 105 nucleated cells/mL in 0.8% methylcellulose containing Iscove modified Dulbecco's culture m e d i u m (IMDM) (Boeringher, Mannheim, West Germany) supplemented with 30% Fetal Calf Serum
Table 1. Monoclonal antibodies used for the characterization of bone marrow and cord blood progenitors and epidermal LC.
Antibody
Isotype
Cluster of Differentiation (CD)
Fluorescein Conjugated OKT6
IgG1
CDla
Phycoerythrin Conjugated T6
IgG1
CDla
Phycoerythrin or Fluorescein My9
IgG2b
CD33
Myelomonocytes Monocytes
Coulter USA
Phycoerythrin or Fluorescein My4
IgG1 k
CD14
Promonocytes Monocytes
Coulter USA
Specificity
Source
70%-80% human cortical thymocytes and LC
Ortho USA
Coulter USA
Ontogeny of Langerhans cells
337
(FCS) and 100UI/mL Granulocyte-Monocyte Colony Stimulating Factor (GMCSF, Boeringher, Mannheim, West Germany). Cultures were incubated at 37°C in water-saturated air containing 5% CO2 for 8 days. Erythroid bursts (BFU-E) were used as negative controls and were cultured under identical conditions (11).
Cytofluorometric Analysis: Double Immunofluorescence Labeling of Mononuclear Cells and Epidermal Cells Mononuclear cells and epidermal cells were dispersed and stained with a primary specific CDla Mab conjugated with fluorescein (OKT6) (1/20 dilution) and the second specific CD33 (My9) or CD14 (My4) Mab conjugated with phycoerythrin (1/20 dilution)• Negative iso-
BONE
. . . .
~
. . . .
"1
type controls, including incubation with the mouse myeloma IgGs from clarified ascites MOPC21 (IgG1) and MOPCI41 (IgG2b), were carried out in parallel. All i n c u b a t i o n s w e r e p e r f o r m e d in RPMI-1640 supplemented with 10% FCS (FCS-RPMI) at 4°C for 1 h each monoclonal antibody. After each incubation, the pellet was washed twice in cold 10% FCS-RPMI. At the end, the pellet was postfixed with 1% paraformaldehyde in Phosphate Buffered Saline (PBS). Flow cytometry was performed on a cytofluorograph 5OH (Ortho Instruments). The evaluation of percentages of labeled cells was done by setting a threshold on the histogram by comparison with background fluorescence of cells treated with normal mouse serum at a 1/1000 dilution. Determinations were based on 104 cells per sample, selected by the corresponding gate (FAS/RAS).
MARROW CELL
. . . . .
-,t
•
CULTURE
. . m
CD 1 • FITC
CONTROL FITC
A -
CONTROL
PE
CD 14
PE
- - - " I
-
- - v - 1
CO 33
"
-"~'-Ir
. . . .
PE
Figure 1. Cytofluorometric analysis of bone marrow-cultured cells stained with C D l a (OKT6 Mab), CD14 (My4 Mab), and CD33 (My9 Mab). Results are representative of five experiments.
338
A. de Fraissinette et al.
CD1
•,.-- CD33
Figure 2.
Ultrastructural morphology of the bone marrow CD1 a/CD14 + or CD33 + precursors: 2a
and 2b: CD33 (My9) promonocyte differentiation antigen with, 2a: x 8,200 and 2b: x 70,000. 2c and
2d: CD14 (My4) monocyte differentiation antigen with, 2c: x 20,000 and 2d: x 55,000.
Ultrastructural Immunocytochemistry : Double Immunogold Labeling of Mononuclear Cells and Epidermal LC Mononuclear and epidermal cells were h a r v e s t e d and washed in 10% FCS-RPML, then incubated with a primary specific C D l a (DMC1) (15) Mab conjugated with 5nm or 17nm gold granules (dilution 1/5) and the second
specific CD33 (My9) or CD14 (My4) Mab conjugated with 17nm or 5nm gold granules (dilution 1/5) at 4°C for 1 h. After washing with 10% FCS-RPMI, the cells were fixed for 1 h with 2% glutaraldehyde in cacodylate buffer, then for 20 min with 1% osmium tetroxide and emb e d d e d in E p o x y medium. Ultrathin sections were examined, after poststaining with lead citrate and uranyl ace-
Ontogeny of LangerhanS cells UMBILICAL
339 CORD
BLOOD
CELLS
--I CONTROL
PE
I=1 W n~
..I CO
33
PE
I=1 Ld cv --1 CONTROL
FITC
Id I.d c¢ ¢d)
Z CD
la
FITC
W Ld rv
Figure 3. Cytofluorometric analysis of umbilical cord blood monocytes stained with CDla (OK'I'6 Mab) and CD33 (My9 Mab). Results are representative of five experiments.
tate with a Philips EM300 electron microscope. All the solutions were filtered through Millipore filter (0.22 ~m mesh). As a negative isotype control, labelings with the mouse myeloma IgGs from clarified a s c i t e s M O P C 2 1 (IgG1) and MOPC141 (IgG2b) were performed.
lmmunochemical Analysis of Epidermal LC and Mononuclear U937 Cell Line L C - e n r i c h e d epidermal cells and U937 cells were surface labeled with 125I by lactoperoxidase-catalyzed iodination
(16). Cells were washed in PBS and in lysis buffer (150mM NaCI, 30mM TrisHCI p H 7.4, 5mM Naz-EDTA, 1% NP-40, lmM PMSF, and 5mM Iodoacetamide) for 30 min on ice. The lysate supernatant was removed after microfuging (14,000 g) and stored at -70°C. Specific immunoprecipitation was performed using a modification of immunoisolation technique described by Tamura (17). Labeled antigens were electrophoresed with a denaturation electrophoresis sample buffer for one-dimensional analysis. Sodium dodecyl sulfate polyacrylamide electrophoresis (SDS-PAGE) was carried out on discontinuous ver-
340
A. de Fraissinette et al.
4a 4b
4c 4d
~33
E)la
Figure 4. Ultrastructural morphology of the double-labeled CD1a/CD14 + or CD33 + umbilical cord blood cells: 4a and 4b: CD14 (My4) monocyte differentiation antigen with, 4a: x8,200 and 4b: x33,000. 4c and 4d: CD33 (My9) promonocyte differentiation antigen with, 4c: x 6,800 and 4d: x 33,000.
tical slab gels according to a modification of Laemmli's method. Gradient gels were made in 7.5% to 15% acrylamide. Relative mass markers used were 14Cmethylated proteins (Amersham): phos-
phorylase b (Mr 32,500), bovine serum albumin (Mr 69,000), ovalbumin (Mr 46,000), c a r b o n i c a n h y d r a s e (Mr 30,000), trypsin inhibitor (Mr 21,500), and lysozyme (Mr 14,000).
Ontogeny of Langerhans cells
341
Results
Bone Marrow Cell Cultures
After 8 days in culture, bone marrow cells were analyzed by cytofluorometry and different subpopulations. The monocyte-like population was selected with the corresponding gate (FAS/RAS). In these conditions, the cultured cell promonocyte/monocyte population was homogeneous and presented a high granular density in FAS/RAS. In this sub-population 6 __- 1% of cells were C D l a + (OKT6), 68 --- 8% were CD14 (My4), and 73 +-- 5% were CD33 (My9) by immunolabeling. By double immunolabeling, the C D l a + cells were found C D I 4 + and C D 3 3 + , approximately 10% of the CD14+ monocytes coexpressed C D l a antigen and 5% of the CD33 + promonocytes coexpressed the CDla antigen. Data were based on phen o t y p i c analysis f r o m five h e a l t h y donors (Fig. 1). By double immunogold labeling, we observed that 30% of the CD14 or CD33
positive promonocyte/monocyte cells coexpressed the CDla antigen. The increase in percentage of CDla +/CD14 + and C D l a + / C D 3 3 + cells which was noted with immunogold method was due to the higher sensitivity of this method (gold granules of small size-5nm) when compared with the cytofluorometric analysis. This allowed quantitation of double labeled (one-step labeling per Mab) cultured bone marrow cells: CD14 antigen (70 +_ I0 gold-5nm granules/cell section with My4 Mab); CDIa antigen (7 +_3 gold-17nm granules/cell section with DMCI Mab); CD33 antigen (10 ___ 4 gold-17nm granules/cell section with My9 Mab); and CDIa antigen (80 _ 8 gold-5nm granules/cell section with DMC1 Mab) (Fig. 2).
Umbilical Cord Blood
By cytofluorometry and double immunolabeling analysis gating on the monocyteqike population, we obtained
E P I D E R M A L L A N G i ~ I H A N S EELL-ENRKHED SUSPENSION
a
"J
(ONTROLFITc
st I
W I'~
t.
(ONTROL ~C ""
1111. . . .
Ill
"'
'Ill "" Ill rlCHIlm=
CONTROL PE
OCHAH 4?
Ill
""
"ll~ ""
~Ol
TOT.Lm ~
A
P
"' IE32
E K- ~2
CD l a FITC ' ' ~lll
-l
" " |l
....
i~ ....
III
""
"Ill
""
Ill
""
tl
10.
.-J
T O T . L - 42S9 P [ . K m 359 IICHRH 44
d
w ~J~ 4) . . . .
Lo
_J i:=1 W I~
P¢AK= I I I
w w ~ll . . . . . .
till TllTllkm 41loll PEAK" l l l l l l OCHAH 4 l l
LOll " " " I l q
c
TOTAL" | I l l
I~'
W W IP~ (.D
( D 33 F I T (
PE
""
~
=
;
G9 e ~" |¢ -I
(D la
b
TllTALs PEAK= 119? OCHAH 3 1
S~ . . . .
Oil ' ' "ICK ' " ~ +
FICH [ R I
T
' " ~l y++|
CONTROL PE
TOTAL= 1 7 1 4 P[AK= 1 3 1
~J
ICHAH 1 6
w ~
.... tl
- +l.
$~ ....
1++ ' '
w o~
~J+ ''
~60
,0T.~- ,=,, I
..J
CD 14 PE
'15+ ' '
FICHIK • Yll3
3 h
ICHRN tO
:t ....
=. ....
ttA
''
"tE~' : it+
** =E=
Figure 5 (a,b,c,d). Cytofluorometric analysis of epidermal enriched-LC selected by the corresponding gate (FAS/RAS) and stained with CDla (OKT6 Mab), CD33 (My9 Mab), or CD14 (My4 Mab). Results are representative of seven experiments. Labelings (a, b) are observed on one normal skin sample and (c, d) on another normal skin specimen.
A. de Fraissinette et aL
342
Figure 6. Ultrastructural morphology of the double-labeled CD1aJCD14+ or CD33+ epidermal enriched-LC: 6a and 6b: CD33 (My9) promonocyte differentiation antigen with, 6a: x 10,000 and 6b: x70,000. 6c and 6d: CD14 (My4) monocyte differentiation antigen with, 6c: ×8,200 and 6d: x 20,000.
an enrichment of 75 _ 7% CD33 + cells. Approximately less than half of the C D l a + cell population (7 _-_ 3% of Ficoil-enriched cord blood cells) was found to be CD33+. This C D l a + cell subpopulation represented 3% of the CD33 + umbilical cord blood monocytes (Fig. 3). The quantitation of double-labeled umbilical cord blood cells was: C D 1 4 a n t i g e n (30 ___ 9 5 n m - g o l d granules/cell section with My4 Mab), C D l a a n t i g e n (10 _+ 5 17nm-gold
granules/cell section with DMC1 Mab). Similarly, the cells coexpressed CD33 antigen (5 -- 4 17nm-gold granules/cell section with My9 Mab) and CDla antigen (35 __ 10 5nm-gold granules/cell section with DMC 1 Mab) (Fig. 4).
Epidermal LC In epidermal cell suspensions fractioned by Ficoll-Hypaque sedimenta-
Ontogeny of Langerhans cells
343
tion, LC accounted for approximately 30% (Fig. 5). All the normal human epidermal C D l a positive selected by the corresponding gate (FAS/RAS) LC coexpressed the CD33 (My9 Mab) and CD14 (My4 Mab) surface antigens by cytofluorometry and immunoelectron microscopy. The CDIa, CD33, and CD14 antigens were never observed on the epidermal keratinocytes which served as negative cell controls (Fig. 6). The quantitation of double-labeled epidermal cells was: CD14 antigen (30 _ 7 5nm-gold granules per cell section with My4 Mab) and CDla antigen (100 _ 18 17nm-gold granules per cell section with
O tO
kD
92.5
59
45
30
=,
3
QD
"13
-
~ =
kD
21.5-
9 2 . 5 --
69 . - ~ Q
46
30
-Q
~
Im mtBII
w
DMC1 Mab); CD33 antigen (30 _+ 9 17nm-gold granules per cell section with My9 Mab) and CDla antigen (300 ___ 50 5nm-gold granules per cell section with DMC1 Mab) (Figs. 7, 8). The effect of the trypsinization to obtain LC-enriched cell suspensions could alter the native CD14 surface antigen and then prevent its detection.
-g
14.3-
qm~
7. Immunoprecipitation results obtained from 1251-labeled LC-enriched epidermal cells presented the following pattern: the CD33 molecule (Mr 67,000) was precipitated by P 67-5; D 3HL-60; WM-54 Mabs from the epidermal lysate, no specific CD14 molecule (53-55, 000) could be Immunoprecipitated from the epidermal LC. Figure
Figure 8. Immunoprecipitation results obtained from 12Sl-labeled U937 cell line presented the f o l l o w i n g pattern: the CD33 molecule (Mr 67,000) was precipitated according to the same conditions of the epidermal LC lysate, the CD14 molecule (53-55,000) was precipitated by My4, Cris6, MOP9 Mabs from the U937 cell lysate.
Discussion In this study, we demonstrated that within the limits of sensitivity of the three following techniques: cytofluorometry, immuno-electron microscopy, and
344
A. de Fraissinette et al.
SKIN
/:7
D LINEAGE
I
I~::
/c¢ ........... .%°, My LOCYTE LI AG~
I @ ...... I IQBICTTE
.....
MOr~ .
. . . .
i
T Hy~IoCyTE
i~I'AGE
$11PDpILJT~ID
e
,
LINEAGE
Figure 9. Schematic representation of suggested migration of the CDla+ monocyte subpopulation from the bone marrow to the epidermis.
immunobiochemistry, human normal epidermal LC are the only resident epidermal cells which react with monocyte d i f f e r e n t i a t i o n antigens (CD33 and CD14). These data confirm those reported by Wood et al. (1984) on normal Langerhans cells CD14 + (18) and Groh et al. (1988) on histiocytosis X CD14 + cells (19). The CD33 antigen described only on bone marrow and blood mononuclear cells was expressed also on all normal human epidermal LC. The extremely restricted distribution of the CD33 molecule on myeloid cells, particularly progenitors, strongly suggests a role in the regulation of myeloid differentiation and maturation (20). The expression of this CD33 antigen on the C D l a + bone marrow, C D I a + cord blood (putative LC precursors), and the CDla + epidermal LC could implicate a parental relationship b e t w e e n these monocyte lineage subpopulations and the dendritic epidermal LC. Interestingly, culture of Ficoll-selected bone marrow cells in the presence of GM-CSF, enhanced the division of typical monocytic phenotype cells. This favors the CD33+ and CD14+ population where 5% of them coexpressed the C D I a antigen. In GM-CSF cultured bone marrow cells, all the C D l a + expressed both phenotype and ultrastruc-
ture of promonocytic/monocytic cells. In comparison, only 3% of the CD33 + or CD14 + umbilical cord blood cells expressed CDla antigen. This study agrees with the presence of two C D l a + populations in cord blood (one monocyte-like and the other one lymphocyte-like) described by Dezutter-Dambuyant, 1986. The small C D l a + / C D 1 4 + or CD33+ cord cells displayed a typical promonocyte/monocyte ultrastructure. This small subpopulation could follow their own maturation entering the bloodstream and perhaps epidermis where the influence of environment remains to be clarified. The Birbeck granules which are specific markers of epidermal LC appear after their entrance into the epidermis without loss of the promonocyte/monocyte differentiation antigens (CDI4, CD33). When we compared the phenotype and the ultrastructural morphology of bone marrow C D I a + (11), cord blood C D l a + promonocytes/monocytes and the epidermal LC, we observed a similar, semiquantitative expression of the CD14 and CD33 antigens, a comparative monocyte morphology but only a difference concerning the Birbeck granule which is only detected regularly in the LC or experimentally in cord blood (4). There is no doubt that LC are derived from a bone marrow, CDla promono-
Ontogeny of Langerhans cells
345
cyte/monocyte subpopulation (I0,11) which could represent the best candidate for LC putative precursors and then migrate via blood vessels to the skin (Fig. 9). The i m m u n o b i o c h e m i s t r y of LC, when compared with the U937 (promonocyte) cell line, confirms the coexpression of CD33 and CDla on these two cell lineages; the trypsinization of epidermal cells did not affect CD33 antigen immunoprecipitation. Furthermore, the effect of trypsinization to obtain LC-enriched suspensions could alter the native CD14 surface antigen and its detection by immunoprecipitation when compared with an untrypsinized, promonocyte U937 cell population. By the sensitive methods (cytofluorometry, immuno-electron microscopy, immunobiochemistry) we suggest that the epidermal LC belongs to a C D l a +
monocyte cell subpopulation. The presence of Birbeck granules only in the epidermal LC could be induced by the epidermal microenvironment. In mice, after 3 days in c u l t u r e , loss of B i r b e c k granules, the F4/80 (antimacrophage) and 2.4G2 (anti Fc receptors) antigens and their cytochemical activities like nonspecific esterase and membrane ATPase (21) suggests that LC are precursors or immature elements of the dendritic cell system.
Acknowledgements--We thank G. Panaye for technical assistance, Dr. Audra and Dr. Guyotat for supplying human umbilical cord blood and human normal bone marrow samples, and Centre de Microscopie Electronique (CMEABG), Universif6 Lyon I for the electron microscopy processing. This work was partially supported by Universit6 de Biologie Humaine, Universit6 Lyon I.
References 1. Rowden, G. The Langerhans cell. CRC Crit. Rev. Immunol. 3:95-180; 1981. 2. Streilen, J. W. Circuits and signals of the skin associated lymphoid tissue (SALT). J. Invest. Dermatol. 85:10s; 1985. 3. Fithian, E.; Kung, E; Golstein, G.; Rubenfeld, M.; Fenoglio, C.; Edelson, R. Reactivity of Langerhans with hybridoma antibody. Proc. Natl. Acad. Sci. USA 78:2541-2544; 1981. 4. Dezutter-Dambuyant, C.; Schmitt, D.; Thivolet, J. Comparative phenotypic and ultrastructural characteristics of OKT6 positive cells in normal peripheral blood and in epidermis. Dev. Comp. Immunol. 10:571-586; 1986. 5. Katz, I. S.; Tamaki, K.; Sachs, D. Epidermal Langerhans cells are derived from cells originating in bone marrow. Nature. 282:324-326; 1979. 6. Volc-Platzer, B.; Stingl, G.; Wolff, K.; Hinterber, W. Schnedi, W. Cytogenetic identification of allogeneic epidermal Langerhans cells in a bone marrow graft recipient (abs). N. Engl. J. Med. 310:1125; 1984. 7. Perreault, C.; Pelletier, M.; Landry, D.; Gyger, M. Study of Langerhans cells after allogeneic bone marrow transplantation. Blood. 63:807811; 1984. 8. Perreault, C.; Pelletier, M.; Belanger, R.; Montplaisir, S.; Boilean, J.; Bonny, Y.; David, M.; Gyger, M.; Lacombe, M. Persistence of host Langerhans cells (LC) following bone
9.
10.
11.
12.
13.
14.
15.
marrow transplantation (BMT). Blood. 64:219; 1984. Goordyal, P.; Isaacson, P. G. Immunocytochemical characterization of monocyte colonies of Langerhans cells and interdigitating reticulum cells. J. Pathol. 146:189-195; 1985. Gothelf, Y.; Sharon, N.; Gazit, E. Fractionation of human bone marrow mononuclear cells with peanut agglutinin: phenotypic characterization with monoclonal antibodies. Human Immunol. 17:37-44; 1986. de Fraissinette, A.; Schmitt, D.; DezutterDambuyant, C.; Guyotat, D.; Zabot, M. T.; Thivolet, J. Culture of putative Langerhans cell bone marrow precursors: characterization of their phenotype. Exp. Hematol. 16:764768; 1988. Ortaldo, J. R.; Sharrow, S. O.; Timonent, R.; Heberman, R. B. Determination of surface antigens on highly purified human NK cells by flow cytometry with monoclonal antibodies. J. Immunol. 127:2401-2409; 1980. Wood, J.; O'Mahony, J. B.; Palder, S. B.; Rodrick, M. L.; D'eon, P.; Mannick, J. A. Circulating T6 antigen positive cells. J. Invest. Dermatol. 82:387; 1984. Gothelf, Y.; Hanau, D.; Tsur, H.; Sharon, N.; Sahar, E.; Cazenave, J. P.; Gazit, E. T6 positive cells in the peripheral blood of burn patients: are they Langerhans cells precursors? J. Invest. Dermatol. 90:142-148; 1988. Dezutter-Dambuyant, C.; Schmitt, D.; Sta-
346
quet, M. J.; Gaucherand, M.; Cambazard, E; Thivolet, J. DMCI: a monoclonal antibody produced from Histiocytosis X cells which reacts with the native CDIa molecule of human epidermal LC. Hybridoma 8:199-208; 1989. 16. Knowles, R. W.; Bodmer, W. E A monoclonal antibody recognizing a human TL-like antigen. Eur. J. Immunol. 12:676-681; 1982. 17. Tamura, G. S.; Dailey, M. O.; Gallatin, M.; McGrath, M. S.; Weissman, I. L.; Pillemer, E. A. Isolation of molecules recognized by monoclonal antibodies and antisera: the solid phase immunoisotation technique. Anal. Biochem. 136:458-464; 1984. 18. Wood, G. S.; Morhenn, V. B.; Butcher, E. C.; Kosek, J. Langerhans cells react with pan-leu-
A. de Fraissinette et al.
kocyte monoclonal antibody: ultrastructural documentation using a live suspension immunoperoxidase technique. J. Invest. Dermatol. 82:322-325; 1984. 19. Groh, V.; Gadner, H.; Radaszkiwicz, T.; Rappersberger, K.; Konrad, K.; Wolff, K.; Stingl, G. The phenotypic spectrum of Histiocytosis X cells. J. Invest. Dermatol. 90:441-447; 1988. 20. Favaloro, E. J.; Bradstock,. K. E; Kabral, A.; Grimsley, P.; Berndt, M. C. Characterization of monoclonai antibodies to the human myeol~d differentiation antigen, "gp67" (CD33). Disease markers 5:215-225; 1987. 21. Schuler, G.; Steinman, R. M. Murine epidermal Langerhans cells mature into potent immunostimulary dendritic cells in vitro. J. Exp. Med. 161:526-546; 1985.
