| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Walter and Eliza Hall Institute of Medical Research, Melbourne, Australia;
Scripps Clinic and Research Foundation, La Jolla, CA 92037;
Joslin Diabetes Center, Boston, MA 02215;
Oklahoma State University, Stillwater, OK 74078;
¶ Laboratory for Immunological Research, Schering-Plough, Dardilly France
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
homodimer rather than the
heterodimer of T cells. Cell kinetic and other evidence
indicated that these three mature spleen DC subtypes are the products
of separate developmental streams, rather than being maturation steps
within a single DC lineage (2). This is in line with
earlier evidence that the cytokine requirements and transcription
factors involved in DC development differ for the
CD8+ and the CD8- DC
(3, 4, 5, 6). In mouse thymus, the analysis of the DC
populations is complicated by the pickup by the DC of Ags, including
CD8 and CD4, derived from T-lineage thymocytes (1).
If such pickup is eliminated, thymic DC appear to be all
CD4-DEC-205+CD11b-,
but do include major CD8+ and minor
CD8- subsets (1). Our earlier
studies indicated that the CD8+ DC of the thymus
are of lymphoid-precursor rather than myeloid-precursor origin
(reviewed in Refs. 7, 8, 9). Arguing mainly by analogy, we
proposed that the CD8+ DC of the spleen were also
of lymphoid origin (7, 8, 9). However, recent evidence (L.
Wu, unpublished observations, and Ref. 10)
indicates that although some splenic CD8+ DC
derive from a lymphoid precursor, many appear to be of myeloid
precursor origin. Therefore, CD8 expression is a poor indicator of the
lymphoid precursor origin (5, 11, 12, 13). Nevertheless, CD8
marks a group of DC with a developmental history differing from that of
CD8- DC.
The DC populations of the lymph nodes (LN) appeared to be still more
complex than those of the spleen, although we had not assessed LN DC in
the light of the additional populations we now distinguish in spleen
and thymus nor considered the problems of Ag pickup. We previously had
noted the presence in LN of a
CD8-DEC-205+ subtype,
which is largely absent from spleen (11). This may
correspond to the minor DC subtype present in the thymus but obscured
by the pickup of CD8 from T cells. The s.c. LN (CLN), those
draining the skin, would also be expected to include the mature form of
epidermal Langerhans cells (LC; Ref. 14), absent from
thymus, spleen, or mesenteric LN (MLN). It is not clear what proportion
of CLN DC would be mature LC. These should be differentially labeled by
painting the skin surface with a fluorescent dye (15, 16),
but this may label dermal- as well as epidermal-derived DC (14, 17, 18). Studies by Anjuère et al. (16),
Salomon et al. (15), and Ruedl et al. (18)
suggested that the mature LC in unstimulated normal mouse CLN could be
distinguished from other LN DC by their slow turnover, by being larger
cells expressing the highest levels of class II MHC, and by the
surprising marker combination
CD11b+CD8int (12, 16).
In this report, we compare the DC populations of LN with those of the spleen and delineate two additional DC subtypes present in LN, but not evident in spleen. We examine the extent of pickup of T cell-derived markers, which could obscure the characterization of LN DC. We also assess the phenotype of the putative mature LC in CLN, by comparison with the DC in MLN, by tracking skin-derived DC with a fluorescent dye, and by examining the DC that exit from cultured skin explants. The results suggest that more than one type of DC enter CLN from the skin. The mature LC in normal CLN may be distinguished by a much higher expression of a number of typical DC activation markers by a very high level of class II MHC only partially controlled by the class II MHC regulating factor CIITA (19, 20, 21) and by the presence in the cytoplasm of residual langerin, a molecule associated with the Birbeck granules and the Ag-processing system of LC (22, 23).
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Mice were bred under specific pathogen-free conditions at Walter
and Eliza Hall Institute animal breeding facility. Most experiments
used female C57BL/6J Wehi mice at 5–7 wk of age. For constructing bone
marrow chimeras, the recipients were C57BL/6 Ly5.1
Pep3b mice aged 8–12 wk. The bone marrow donors
included the above strains, as well as
CD8-/- C57BL/6 mice and
CD4-/- C57BL/6 mice, the latter two strains
originally obtained from T. Mak (Ontario Cancer Institute, Toronto,
Canada). The CIITA-/- mice were obtained
directly from the Institut de Génétique et de Biologie
Moléculaire et Cellulaire (Strasbourg, France); their generation
is described elsewhere (19, 21).
Bone marrow chimeric mice
Bone marrow chimeric mice were constructed as described previously (1). Briefly, 3 x 106 bone marrow cells from wild-type Ly5.1 mice together with 3 x 106 bone marrow cells from CD8-/- or CD4-/- Ly5.2 mice were injected into lethally irradiated Ly5.1 mice. For each such experimental group, a control group involving transfer of wild-type Ly5.1 bone marrow together with wild-type Ly5.2 bone marrow into the Ly5.1 recipients was set up. Six weeks later, the separate types of LN were taken from the pooled experimental or control groups and the DC isolated for analysis.
Lymphoid organs
The lymphoid organs used for DC isolation and comparative analysis were spleen, MLN, and CLN, the CNL being a pool of axillary, brachial, and inguinal LN. For experiments on the CIITA-/- mice, inguinal LN and brachial LN were studied separately. For experiments on DC migration after labeling the skin of the ears, auricular LN were used.
DC isolation procedure
The DC isolation procedure was based on that described recently
(1, 2). Briefly, LN or spleen fragments from 8–16 mice
were digested for 20 min at room temperature with collagenase-DNase and
then treated for 5 min with EDTA to disrupt T cell-DC complexes. All
subsequent procedures were at 0–4°C in a Ca2+-
and Mg2+-free medium. Light-density cells were
selected by centrifugation in a Nycodenz medium (Nycomed, Oslo,
Norway). For LN DC, a density of 1.082 g/cm3,
4°C, 310 milliosmolar was used, giving a slightly better yield with
equivalent purity compared with the density of
1.077g/cm3 optimal for spleen DC isolation. Cells
not of the DC lineage were then depleted by incubating the
cells with previously optimized amounts of anti-CD3 (KT3),
anti-Thy1 (T24/31.7), anti- B220 (RA3-6B2), anti-Gr-1
(RB6-8C5) and anti-erythrocyte (TER-119) and then removing the
Ab-binding cells with anti-rat Ig-coupled magnetic beads
(Dynabeads; Dynal, Oslo, Norway). Note that the anti-CD4,
anti-F4/80, and anti-CD11b, used in earlier versions of the
procedure (24, 25), were all omitted from the depletion
mix to avoid loss of the
CD4+8- DC population
(1). Note also that anti-B220, omitted in error from a
previous description (1), was included. The LN DC at this
stage were 
90% pure. The preparation then usually was subjected to
presorting to remove 5–10% autofluorescent cells (1), or
used directly for immunofluorescent labeling and analysis by flow
cytometry with gating to eliminate a low level of autofluorescent cells
(1).
For isolation of DC from single LN, the initial light density selection step was omitted, but the immunomagnetic bead depletion was retained. Reduced medium volumes and small-sized tubes were used throughout. Presorting was omitted. The 10% DC in the enriched preparations then were gated as CD11c+ cells.
DC migration from mouse ear skin on culture
The DC migration procedure was modified from that of Schuler and Steinman (26), with exit of DC from the skin enhanced by chemokine following the approach of Kellerman et al. (27). The ears were removed from 10–20 mice, cleared of hair, and briefly washed in 70% ethanol. The ears were placed ventral-side down and split, removing the dorsal skin from the cartilage. The dorsal skin was placed split-side down in 1 ml of modified mouse osmolarity RPMI 1640 culture medium containing 10% FCS (2, 25) for 2–4 h at 37°C in a humidified 10% CO2-in-air incubator to eliminate the many non-DC initially released into the culture. The skin then was transferred onto another 1-ml culture medium, this time containing 0.1 µg of recombinant mouse 6Ckine (R&D Systems, Minneapolis MN) to enhance DC migration. After 24 h of further incubation at 37°C, the cells that had migrated into the culture medium were harvested and kept in cold medium. The skin was transferred onto fresh warm medium containing 6Ckine and then incubated a further 24 h at 37°C. The cells that migrated out of the skin over the first and second 24-h incubations then were pooled. The yield averaged 3–6 x 104 cells per ear. The cells then were immunofluorescent-stained and analyzed as for the LN DC preparation. For direct comparison with skin-derived DC, a CLN DC preparation was incubated 24 h at 37°C in the same 6Ckine-containing culture medium.
Immunofluorescent labeling and flow cytometric analysis of DC preparations
The mAb, the fluorescent conjugates, the labeling procedure, and the flow cytometric analysis details all have been described previously (1, 2, 24, 25). The mAb HD24 recognizing an intracellular epitope of murine and human langerin was provided by Schering-Plough (Dardilly, France) and was conjugated to FITC in this laboratory. In most experiments, presorting before immunofluorescent labeling was used to eliminate 5–10% autofluorescent non-DC in the enriched DC preparations (1). In some experiments in which autofluorescence was minimal (<5%), autofluorescent cells were gated out along with dead cells in the propidium iodide channel. Propidium iodide was included in the final wash to label and exclude dead cells in all experiments except those involving langerin staining. In the analysis of the DC-enriched preparations, anti-class II MHC or anti-CD11c or both together were used to stain, define, and gate DC, along with gating for the high-side and forward light-scatter characteristics of DC. The other mAb stains then were used to analyze and subdivide these gated DC. To stain for the cytoplasmic domain of langerin, cells were fixed and permeabilized after surface staining as above. Cells were fixed in a 1% formaldehyde-2% glucose-5 mM sodium azide solution for 20 min at room temperature, washed, and then permeabilized with 0.1% saponin-2% FCS-EDTA-balanced salt solution. This permeabilization medium then was used as the subsequent medium for staining. The cells were centrifuged and then blocked with anti-FcR Ab (2.4G2) for 15 min. The FITC-conjugated mAb HD24 or FITC-conjugated isotype-matched control was added and incubated with the cells for 30 min before washing and analysis.
Fluorescent labeling of DC in ear skin
The approach was similar to that of Anjuère et al. (16) and Cumberbatch et al. (28). Tetramethylrhodamine-5- (and-6)-isothiocyanate (TRITC) and FITC were obtained from Molecular Probes (Eugene, OR). A 10% solution of TRITC or FITC was made in DMSO and then diluted to 1% in a solvent of acetone-dibutylphthalate, 1:1. The dorsal side of both ears of C57BL/6 mice was painted with 10 µl of either of these 1% fluorochrome solutions. After 24–48 h, the auricular LN were removed and DC isolated as described above.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In previous studies by us (24, 25) and by others,
selective loss of DC subpopulations and the complications caused by
autofluorescent cells and pickup of material from T cells
(1) would have distorted the immunofluorescent staining
and the analysis of LN DC populations. Therefore, we compared the DC
populations of CLN, MLN, and spleen using isolation and analysis
conditions that avoid these problems (1). Presorting was
used to remove any autofluorescent macrophage-like cells. The
immunomagnetic bead cell depletion procedure adopted avoided the loss
of the CD4+ F4/80+ DC
population (1). Results were basically the same if no
immunodepletion at all was used, but were then less crisp because of
overlap with the fluorescence distribution of non-DC contaminants. The
CD4, CD8, and DEC-205 staining of the LN DC was examined in detail,
because those markers had proved particularly useful for segregating
spleen DC (1, 2), as Salomon et al. (15)
first noted staining for CD4 on murine DC when analyzing LN, and
Anjuère et al. (12, 16) reported CD8 expression by
mature LC in LN. The expression of a series of markers on LN DC is
given in Fig. 1
, and an example of a comparative multiparameter analysis is given in
Fig. 2.
|
|
-DEC-205-
population (Fig. 2
, population 1). All LN, like spleen, contained a
proportion of DC staining strongly for CD8, but not for CD8 (Fig. 2, population 2). However, in contrast to spleen, LN DC and especially
CLN DC included many more cells showing intermediate and low levels of
CD8 staining (Figs. 1 and 2). LN of all types contained some DC
staining brightly for CD4, but the proportion was much less than in
spleen (Fig. 1). However, LN DC included many cells showing
intermediate and low levels of CD4 staining (Fig. 1). The few
CD4high cells in LN were all
CD8-, as in spleen (data not shown). Because
these CD4high DC were a minority population in
LN, as opposed to spleen (1), they are grouped together in
Fig. 2 and elsewhere with the
CD4-8- DC that they most
resemble.
LN of all types contained a higher proportion of DC expressing DEC-205
and high levels of class II MHC than did spleen (Figs. 1 and 2). As we
have noted previously (11), DEC-205 staining did not
correlate with CD8 staining in LN DC, in marked contrast to spleen DC.
Although all CD8high DC were
DEC-205+, as in spleen (Fig. 2, population 2), LN
contained additional major groups of DEC-205+ DC,
which stained negative to intermediate for CD8 (Fig. 2, populations
3 and 4).
Differences between DC in MLN and CLN
It was evident that as well as differences between spleen DC and
LN DC in general, there also were particular differences between the DC
found in CLN compared with MLN. Because the CLN might be expected to
contain skin-derived DC not present in MLN, a more extensive comparison
of the surface markers was undertaken (Figs. 1 and 2).
The DC showing high levels of CD8 staining did not differ
significantly between MLN and CLN, when many samples were compared. The
DC showing intermediate levels of CD8 staining were variable in
incidence, but usually more numerous in CLN. Importantly, neither LN
group showed bright or intermediate staining for CD8, even though
the anti-CD8 mAb used gave bright staining of T cells; thus,
both CD8high and CD8int
cells were CD8, as for spleen DC (1, 24). However,
low-level staining was seen for both CD8 and CD8, suggesting
these weakly staining DC had CD8 on their surface.
The level of class II MHC expressed by LN DC was in general higher than
that on spleen DC, but CLN contained in addition a group of around 25%
of DC with exceptionally high surface class II MHC levels, higher than
on MLN (Figs. 1 and 2b). Note that staining for class II MHC
deliberately used a low fluorochrome-to-protein conjugation ratio on
the mAb to keep the fluorescence distribution on-scale and allow
accurate color compensation; thus, the actual level compared with other
markers is much higher than indicated. These class II
MHChigh DC also were relatively large in size, as
judged by the forward light-scatter profiles (Fig. 2b). Such
a high expression of class II MHC on larger-sized cells has been
considered a marker of LC-derived DC in CLN (15).
CLN also contained a distinct population of DC showing only moderate
staining with anti-CD8, but expressing DEC-205 at levels as high or
higher than on CD8high DC. This population was
seen only at marginal levels in MLN and not detected in spleen (Figs. 1 and 2, population 4). This distinct group of
DEC-205high CD8low DC
corresponded to the large-sized MHC class IIhigh
cells and so were putative LC-derived DC. CLN also contained some DC
expressing very high levels of the activation markers CD40, CD80, and
CD86, levels higher than seen on spleen or MLN DC (Fig. 1). In
agreement with the results of Anjuère et al. (16) a
high expression of CD11a also was observed (Fig. 1). Cross-correlation
studies (data not shown) indicated that these DC also corresponded to
the DEC-205high CD8low
class II MHChigh group (Fig. 2, population
4).
Not all markers characteristic of CLN mapped to the same DC population.
CLN contained more DC staining especially strongly for the low-affinity
Fc receptor CD16/32 (Fig. 1). However, only part of these
CD16/32high cells overlapped population 4 (Fig. 2), the remainder being DC expressing lower levels of class II MHC and
DEC-205. CLN also contained more DC expressing intermediate rather than
high levels of CD11c (Fig. 1). However, these did not correlate with
population 4 (Fig. 2), and such DC have been suggested to be
monocyte-derived rather than LC-derived (17).
Overall, these results suggested that the mature forms of skin-derived
DC, including mature LC, are likely to be within the CLN DC expressing
especially high levels of class II MHC, DEC-205, CD40, CD80, CD86, and
CD11a but staining only weakly for CD8 and CD4.
Tests for authentic CD4 and CD8 expression by LN DC
The low to moderate levels of staining for CD4 and CD8 among LN DC
recalled the situation in the thymus where such staining was
attributable to pickup of Ags from T lineage thymocytes
(1). The finding that at least the low-level staining for
CD8 involved both CD8 and CD8 on the DC surface was indirect
evidence that this was attributable to pickup from CD8 T cells.
To distinguish authentic CD4 or CD8 expression by the DC themselves
from staining attributable to pickup of Ag from associated T cells,
bone marrow chimeras were constructed. Irradiated mice were
reconstituted with mixes of wild-type and
CD4-/- bone marrow, or mixes of wild-type and
CD8-/- bone marrow, using a Ly5 allotype
difference to distinguish DC derived from the normal or from the
gene-knockout bone marrow. If staining was lost from the gene-knockout
DC, despite the presence of T cells expressing CD4 and CD8, the
staining of the wild-type DC was considered to be authentic; if
staining persisted on DC lacking the relevant gene, it must have been
attributable to material picked up from wild-type T cells.
As shown in Fig. 3a, staining for both high and intermediate levels of CD8
disappeared from CLN and MLN DC when they were derived from
CD8-/- mice, indicating authentic expression
by the DC themselves. However, the shoulder of low-level CD8
staining persisted in the CD8-/- DC, so this
lowest level of staining appeared to reflect Ag pickup. This pickup was
more than previously seen from spleen DC, but less than seen with
thymic DC (1). It was observed consistently that the DC
expressing intermediate levels of CD8 were less frequent in the
CLN of these reconstituted animals than in normal mice, suggesting that
these DC were slow to be replenished from bone marrow.
|
b, the highest level of CD4 staining on a
small population of LN DC disappeared from the DC derived from
CD4-/- mice, indicating these few DC displayed
authentic CD4 expression. However, much of the CD4 staining, at
intermediate to low levels above the background, persisted in the DC
derived from CD4-/- mice, and so had been
derived from associated T cells. This level of pickup of CD4 was far
more than seen with splenic DC, was greater than for CD8 on the LN
DC, and it resembled the CD4 pickup among thymic DC
(1). The effect of the CIITA regulatory factor on class II MHC expression by LN DC
Because the surface levels of class II MHC seemed particularly
high among CLN DC, the role of CIITA, a regulator of class II MHC
expression, was examined (19). In a recent reassessment,
such CIITA-/- mice were found to have a
population of DC in certain LN, but not spleen, which retained class II
MHC expression, albeit at a lower level (19). To determine
which DC retained the class II MHC expression in
CIITA-/- mice, DC in the inguinal LN, brachial
LN, and MLN were isolated and analyzed. Because only a few
CIITA-/- mice were available, the isolation
procedures were modified to allow DC isolation from single LN. The
expression of CD11c was used as the criterion to gate for DC in the
partially enriched LN preparations, although this still allowed some
non-DC contaminants to appear in the analysis. The gated DC then were
analyzed for CD8, DEC-205, and class II MHC, as shown in Fig. 4, as well as for CD40 (data not shown).
|
) and at a slightly reduced incidence in brachial LN (not shown), but
were not present at significant levels in MLN (Fig. 4). This agreed
with earlier immunofluorescent staining of tissue sections (data not
shown). The class II MHC+ DC in
CIITA-/- mouse inguinal (and brachial) LN were
among the DC in these nodes staining brightest for DEC-205 (Fig. 4) and
for CD40 (data not shown). Most of these class II
MHC+ DC also showed intermediate staining for
CD8 but none showed high staining for CD8 (data not shown).
Although these class II MHC+ DC were those
showing the highest staining for DEC-205 and CD40 in these
CIITA-/- mice, the actual level of DEC-205
staining and of CD40 staining was below that of the corresponding
bright staining DC in the control
CIITA+/- mice. This suggested that the
surface expression of many markers was (directly or indirectly)
influenced by the absence of the CIITA gene. In addition, the DC
numbers were 2-fold lower in the CIITA-/-
mice. Despite this difference in numbers and absolute staining levels,
the results suggested that the DC retaining class II MHC expression
corresponded to the class II MHChigh
DEC-205high CD40high DC in
CLN (Fig. 2, population 4), the DC suspected of being skin-derived
LC. The nature of the DC that migrate out of cultured skin explants
To help distinguish the LC and other skin-derived DC in CLN, the
cells that migrated out of the mouse ear skin explants in culture were
isolated, immunofluorescent stained, and analyzed. These DC recently
derived from skin might be expected to be initially less mature than
their equivalents in CLN. However, the procedure involves incubation of
migrating cells in culture medium containing 6Ckine at 37°C, which
could itself induce further maturation. Accordingly, these skin-derived
DC were compared with a side-by-side sample of CLN DC, which had been
incubated in the same culture medium for the same average time as the
cells exiting from the skin. The comparison is shown in Fig. 5.
|
) had, on incubation, changed to
expressing high levels on the cell surface. Likewise, CLN DC expressing
these lower levels of CD40, CD80, and CD86 (Fig. 1) had all
up-regulated these markers to a higher level on incubation. However,
the migrating DC expressed still higher levels of these activation
markers, indicating that the skin-derived DC already possess or readily
acquire a very activated phenotype.
The DC migrating from skin expressed CD11b, F4/80 and CD32/16 at
moderate levels, although much lower than the levels typical of
macrophages. They were all clearly negative for CD4 and CD8. Thus
they expressed low levels of characteristic myeloid markers, but did
not express characteristic lymphoid markers.
The DC migrating out of the skin expressed high levels of CD24, higher
than on most incubated CLN DC. The migrating DC also were all positive
for DEC-205, but always displayed two peaks of low and high DEC-205
expression. Two distinct peaks of DEC-205 fluorescence also were seen
if the cells were fixed and permeabilized before staining, showing that
the distinction reflected total DEC-205 levels and not just a
difference in the proportion on the DC surface (data not shown). The
migrating DC that were DEC-205high
surface-stained more brightly than the majority of incubated CLN DC,
with only 20% of incubated CLN DC showing this high level of staining.
This reinforces the picture from Figs. 1 and 2, suggesting that very
high DEC-205 expression is a useful marker of one type of
skin-derived DC.
Tracking skin-derived DC into LN
To mark in a more direct way LN DC that originated from skin,
mouse ear skin was painted with FITC or TRITC and the fluorescent DC
was tracked in the draining auricular LN. A discrete population of FITC
or TRITC-labeled DC could be sorted from unlabeled LN DC and from
autofluorescent contaminants by using two fluorescence channels. An
example with TRITC labeling is shown in Fig. 6. The sorted fluorescence positive DC and the negative DC then were
stained in the other fluorescence colors for other DC markers and
analyzed. The results were similar from days 1–3 after painting the
ears, after which the FITC or TRITC fluorescence declined. Fig. 6 shows
the results after day 1, when the label is more likely to be associated
with the original migrating DC. These results were the same when only
the cells with the brightest FITC or TRITC fluorescence were selected,
rather than the total fluorescent population being gated as in Fig. 6.
Again this is an argument against the label being passed from the
original migrating DC into secondary LN DC. The label was found only in
the draining auricular LN, with the adjacent cervical LN being
negative. The TRITC-positive DC were analyzed and compared with the
total DC of the auricular LN as shown in Fig. 6. Note that the
unlabeled DC also should contain unlabeled skin-derived DC that arrived
in the LN before painting.
|
, even 3 days after painting the ears. A very
low-level CD8 staining on most cells was obtained, similar to the
pickup level of Fig. 3a, together with a small
population of the TRITC+ DC showing intermediate
CD8 staining. The TRITC+ DC differed somewhat
from the DC from skin explant cultures (Fig. 5) in showing reduced
levels of DC with high expression of CD24 and CD11a. It is not clear
whether this was attributable to down-regulation after the migrating DC
reached the LN or to an artificial up-regulation of these markers
during the culture required to collect the migrating DC. Overall, the
putative skin-derived TRITC+ LN DC represented a
restricted subgroup of LN DC, but were not homogeneous, resembling a
mix of populations 3 and 4 of CLN (Fig. 2). They resembled by most
markers the DC migrating out of cultured skin explants. Detecting LC-derived DC with langerin as a marker
It seemed likely that the DC migrating into LN from skin consisted
of maturing forms of LC derived from the epidermis and dermal DC
resembling the interstitial DC common to many tissues. To selectively
label LC-derived DC, the DC were permeabilized and stained for
cytoplasmic langerin, a marker of the Ag-processing system of LC
(22, 23). Because Ag processing decreases as LC mature
into T cell-stimulating DC, there was no guarantee that the langerin
marker would persist. However, DC containing high levels of langerin
were detected in both the DC migrating out of ear skin and in CLN as
shown in Fig. 7. Surprisingly, a shoulder of moderate staining for langerin was
obtained in DC from other tissues, including spleen and MLN (Fig. 7),
and an above background shoulder of staining was seen in B cells (Fig. 7) and T cells (data not shown). The basis and specificity of this
low-level staining was not established. It was concluded that only
high-level expression of langerin could be used as a LC marker.
|
, marking population 4
of Fig. 2 as LC-derived or LC-related DC (Fig. 7), in confirmation of
the previous conclusions. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, but which clearly express
DEC-205 and express high surface levels of class II MHC. One subset,
common to all LN, expresses moderate levels of DEC-205 and may
represent interstitial tissue-derived DC, monocyte-derived DC, or an
activated version of the splenic
CD4-8-DC. These are not
mutually exclusive descriptions. A second more distinctive DC subtype,
largely restricted to CLN, expresses higher levels of DEC-205 and
appears to represent mature forms of LC. Both of these DC subtypes are
present among the DC that migrate out of the skin into LN, as revealed
either by catching the DC that emerge from skin explants before
locating in LN or by tracking skin-derived DC after painting the skin
with a fluorescent dye. As others have suggested (16, 17, 18),
these two DC subtypes may derive from dermal DC and epidermal LC,
respectively. All five DC subsets found in LN are capable of
stimulating naive T cells in a mixed leukocyte culture system
(our unpublished data). Other biological properties such as
cytokine production are currently being investigated. One objective of this study was to identify the mature form of LC among CLN DC and to determine what proportion of CLN DC are LC-derived in a normal mouse with noninflamed skin. A DC population representing at most 25% of CLN DC, and absent from spleen, thymus, or MLN, appeared to be LC derived. This points to a normal steady-state migration of some epidermal LC into CLN, even in the absence of obvious "danger" signals. This migration would be enhanced by inflammatory stimuli such as painting the skin with a solvent containing a fluorescent dye (18). An alternative explanation is that putative LC that fail to find a vacant niche in the epidermis lodge in the draining LN without ever serving as sentinel LC in the epidermis.
The putative LC-derived DC were identified in our study by a range of
surface markers, none LC specific but differing sufficiently in level
of expression to distinguish this population. Several of these marker
combinations have been used previously by others to segregate
LC-derived DC (15, 16, 17, 18). Some of the LC-derived DC markers,
such as large size, high CD40, and very high surface class II MHC, are
markers of fully activated DC. Because other DC in the same LN are not
in this extreme state of activation, the LC lineage must either be
hypersensitive to activation stimuli or inherently of this phenotype.
The persistence of some class II MHC on these DC even in CIITA-null
mice suggests the regulation of class II MHC expression differs between
the LC lineage and other DC. However, an alternative view is that CIITA
simply acts as a 10- to 30-fold amplifier of class II MHC expression,
and on its removal, only cells with an initially very high class II MHC
expression are then detectable by immunofluorescence. We have found the
high level of expression of DEC-205 to be an especially useful marker
of the LC DC lineage, provided the CD8high DC
that also show high DEC-205 expression are first gated out. These
CD8+ DC, common to all lymphoid tissues, cannot
be confused with LC, because in addition to the much higher CD8
expression, they lack the myeloid markers CD11b, F4/80, and
CD16/32.
The one marker we expected to be specific for the LC DC lineage was langerin, because it is associated with the Ag-processing system and the Birbeck granules of LC (22, 23). Indeed, if high level cytoplasmic staining was the criterion, this appeared to mark the LC-derived DC among skin emigrant DC and within CLN and confirmed the conclusions based on surface staining. The significance of lower-level cytoplasmic staining of DC in spleen, MLN, and other tissues is uncertain at present.
The possible expression of CD4 and CD8 on LN DC and on the LC lineage
has been a source of confusion and controversy. We find no significant
expression of CD4 on LC-derived DC and only a low incidence of true
high CD4-expressing DC in LN generally. This is ironic because it was
the study of Salomon et al. (15), who found that
CD4+ DC in LN, which prompted our finding of a
major CD4+ DC population in spleen
(1). We now show that most of the low and medium staining
of CD4 on LN DC is attributable to pickup of this Ag from T
lymphocytes. The much higher pickup of CD4 compared with CD8 is
presumably because CD4 T cells outnumber CD8 T cells in peripheral
lymphoid organs. The lowest level of staining for CD8 on LN DC,
visualized as a shift from the background rather than a clear
population, also seems to be attributable to low-level pickup from T
cells. However, both the high-level CD8 staining seen on a subset of
LN and spleen DC and the intermediate CD8 staining characteristic of
LN DC including some LC-derived DC appears to represent authentic
expression by the DC themselves.
Various interpretations have been made of the expression of CD8 on
DC derived from LC. Both Anjuère et al. (12, 16) and
Merad et al. (13) have shown that although LC in the
epidermis or those migrating from skin do not express CD8, they can
express this marker after entry into LN, at least if activated. We also
find that LC exiting the skin do not express CD8, but then find only
general low level "pickup" staining and only a modest proportion of
cells with medium level expression among the putative DC of LC origin
in the LN. The few LC-related DC that could be classed as moderate
expressors of CD8 are clearly distinguished from the
CD8high DC population found in spleen, thymus,
and LN. Therefore, our results are in line with those of Ruedl et al.
(18), who found little CD8 expression by putative
LC-derived DC. But given that mature LC can, under some circumstances,
express CD8, does this imply that they are of lymphoid origin, as
originally proposed by Anjuère et al. (12, 16, 29)?
Or does it imply that CD8 is a poor marker of precursor origin, as
proposed by Merad et al. (13)? Recent studies have rather
suggested that populations of bone marrow precursors considered as
lymphoid-committed or myeloid-committed retain considerable
developmental flexibility and that both are capable of forming
CD8+ DC (L. Wu, unpublished observations, and
Ref. 10). Conversely, in contrast to our initial
conclusions (30), we (5, 11) and Martin et
al. (31) have found that thymic lymphoid precursors can
generate both CD8+ and
CD8- DC. It appears CD8 is induced on
developing DC at some stage downstream of early hemopoietic precursors.
Induction of CD8 on DC generated in vitro can be achieved, although
only on those DC generated by particular cytokines (32).
Although CD8 no longer seems a reliable marker of the original
precursor origin, its expression at high levels nevertheless seems to
mark a functionally distinct DC state that in all lymphoid tissues has
the highest capacity to produce the bioactive form of IL-12
(33). At least in spleen, it also marks a subtype of DC
originating from a separate developmental stream (2). The
branch point of this stream and the precise factors inducing CD8
expression on DC remain to be determined.
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Sandrine Henri, Walter and Eliza Hall Institute of Medical Research, P. O. Royal Melbourne Hospital, Victoria, 3050, Australia. E-mail address: henri{at}wehi.edu.au
3 Abbreviations used in this paper: DC, dendritic cells; LN, lymph nodes; CLN, s.c. LN; MLN, mesenteric LN; LC, Langerhans cells; TRITC, tetramethylrhodamine-5- (and-6)-isothiocyanate; int, intermediate.
Received for publication February 13, 2001. Accepted for publication May 8, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
- dendritic cells but not of lymphoid-related CD8+ dendritic cells. Immunity 9:839.[Medline]
-positive dendritic cells from a common myeloid progenitor. Science 290:2152.-positive dendritic cells in vivo. Blood 96:1865. are potent chemoattractants for in vitro- and in vivo-derived dendritic cells. J. Immunol. 162:3859.- and CD8+ dendritic cells are generated from CD4low lymphoid-committed precursors. Blood 96:2511., and IFN-
by mouse dendritic cell subsets. J. Immunol. 166:5448.This article has been cited by other articles:
|
|
 |
|
  E. Jaensson, H. Uronen-Hansson, O. Pabst, B. Eksteen, J. Tian, J. L. Coombes, P.-L. Berg, T. Davidsson, F. Powrie, B. Johansson-Lindbom, et al. Small intestinal CD103+ dendritic cells display unique functional properties that are conserved between mice and humans J. Exp. Med., September 1, 2008; 205(9): 2139 - 2149. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Shklovskaya, B. Roediger, and B. Fazekas de St. Groth Epidermal and Dermal Dendritic Cells Display Differential Activation and Migratory Behavior While Sharing the Ability to Stimulate CD4+ T Cell Proliferation In Vivo J. Immunol., July 1, 2008; 181(1): 418 - 430. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Wang, L. S. Bursch, A. Kissenpfennig, B. Malissen, S. C. Jameson, and K. A. Hogquist Langerin Expressing Cells Promote Skin Immune Responses under Defined Conditions J. Immunol., April 1, 2008; 180(7): 4722 - 4727. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. De Trez, K. Schneider, K. Potter, N. Droin, J. Fulton, P. S. Norris, S.-w. Ha, Y.-X. Fu, T. Murphy, K. M. Murphy, et al. The Inhibitory HVEM-BTLA Pathway Counter Regulates Lymphotoxin Receptor Signaling to Achieve Homeostasis of Dendritic Cells J. Immunol., January 1, 2008; 180(1): 238 - 248. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. F. Poulin, S. Henri, B. de Bovis, E. Devilard, A. Kissenpfennig, and B. Malissen The dermis contains langerin+ dendritic cells that develop and function independently of epidermal Langerhans cells J. Exp. Med., December 24, 2007; 204(13): 3119 - 3131. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Ginhoux, M. P. Collin, M. Bogunovic, M. Abel, M. Leboeuf, J. Helft, J. Ochando, A. Kissenpfennig, B. Malissen, M. Grisotto, et al. Blood-derived dermal langerin+ dendritic cells survey the skin in the steady state J. Exp. Med., December 24, 2007; 204(13): 3133 - 3146. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Moser, D. D. Desai, M. P. Downie, Y. Chen, S. F. Yan, K. Herold, A. M. Schmidt, and R. Clynes Receptor for Advanced Glycation End Products Expression on T Cells Contributes to Antigen-Specific Cellular Expansion In Vivo J. Immunol., December 15, 2007; 179(12): 8051 - 8058. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Cools, P. Ponsaerts, V. F. I. Van Tendeloo, and Z. N. Berneman Balancing between immunity and tolerance: an interplay between dendritic cells, regulatory T cells, and effector T cells J. Leukoc. Biol., December 1, 2007; 82(6): 1365 - 1374. [Abstract] [Full Text] [PDF]
|
|
