Intradermal Immunization Triggers EpidermalLangerhans Cell Mobilization Required forCD8 T-Cell Immune ResponsesChristelle Liard1, Severine Munier2, Alix Joulin-Giet1, Olivia Bonduelle1, Sabrina Hadam3, Darragh Duffy1,Annika Vogt3, Bernard Verrier2 and Behazine Combadiere1

The potential of the skin immune system for the generation of both powerful humoral and cellular immuneresponses is now well established. However, the mechanisms responsible for the efficacy of skin antigen-presenting cells (APCs) during intradermal (ID) vaccination still remain to be elucidated. We have previouslydemonstrated in clinical trials that preferential targeting of Langerhans cells (LCs) by transcutaneousimmunization shapes the immune response toward vaccine-specific CD8 T cells. Others have shown that IDinoculation of a vaccine, which targets dermal APCs, mobilizes both the cellular and humoral arms of immunity.Here, we investigated the participation of epidermal LCs in response to ID immunization. When human ormouse skin was injected ID with a particle-based vaccine, we observed significant modifications in themorphology of epidermal LCs and their mobilization to the dermis. We further established that this LCrecruitment after ID administration was essential for the induction of antigen-specific CD8 T cells, but was,however, dispensable for the generation of specific CD4 T cells and neutralizing antibodies. Thus, epidermaland dermal APCs shape the outcome of the immune responses to ID vaccination. Their combined potentialprovides new avenues for the development of vaccination strategies against infectious diseases.

Journal of Investigative Dermatology advance online publication, 15 December 2011; doi:10.1038/jid.2011.346

INTRODUCTIONSkin has always been a challenging immune target for vaccinedelivery (Romani et al., 2010b). In particular, the availabilityof functionally active resident and infiltrating antigen-present-ing cells (APCs) at the site of inoculation is crucial for theoptimal capture and presentation of vaccine-derived antigens(Sumida et al., 2004). Among skin vaccine routes, intradermal(ID) delivery has been extensively tested in many clinicaltrials (Lambert and Laurent, 2008; Nicolas and Guy, 2008)including vaccination against rabies (Sabchareon et al., 1998;Charest et al., 2000; Vien et al., 2008), hepatitis B (Egemenet al., 1998; Henderson et al., 2000), and influenza (Frech

et al., 2005; Belshe, 2007). It appears that the ID route iscapable of inducing a humoral immune response that isequivalent or comparable to the one induced by theintramuscular or subcutaneous routes but with a lower doseof antigen, generally a fifth of the standard dose used for anintramuscular vaccine. Nevertheless, the cellular mechanismsinvolved in the induction of T-cell-mediated immuneresponses remain to be studied.

It is now well established that skin APCs can take upvaccine compounds, process them, and present them toT and B cells in the draining lymph node (DLN), resultingin the initiation of cellular and humoral immune responses(Kupper and Fuhlbrigge, 2004; Mahe et al., 2009). However,evidence in the literature suggests that epidermal Langerhanscells (LCs) and dermal dendritic cells (DDCs) exhibitfunctional differences in the induction of cellular andhumoral immunity (Romani et al., 2010b; Ueno et al.,2010a, 2010b). In a recent phase I clinical trial, we haveshown that transcutaneous immunization through openhair follicles, which allows the targeting of follicular LCs,preferentially induces CD8 T cells, whereas the intramuscularroute cannot (Vogt et al., 2006, 2008; Combadiere et al.,2010). Here, we studied the role of epidermis-resident LCsafter the ID injection of particulate vaccine compounds,and elucidated fundamental mechanisms involved in theinduction of cellular immune responses.

& 2011 The Society for Investigative Dermatology www.jidonline.org 1

ORIGINAL ARTICLE

Received 20 March 2011; revised 21 June 2011; accepted 16 July 2011

1Laboratory of Immunity and Infection, Institut National de la Sante et de laRecherche Medicale, INSERM UMR-S 945 and Universite Pierre et MarieCurie (UPMC Univ Paris 06), Paris, France; 2Institut de Biologie et Chimie desProteines, UMR 5086 CNRS/UCBL, Lyon, France and 3Clinical ResearchCenter for Hair and Skin Science, Department of Dermatology and Allergy,Charite-Universitatsmedizin Berlin, Berlin, Germany

Correspondence: Behazine Combadiere, Laboratory of Immunity andInfection, Institut National de la Sante et de la Recherche Medicale, INSERMUMR-S 945, 91 Boulevard de l’Hopital, 75013 Paris, France.E-mail: behazine.combadiere@upmc.fr

Abbreviations: APC, antigen-presenting cell; (D)DC, (dermal) dendritic cell;DLN, draining lymph node; DT, diphtheria toxin; EGFP, enhanced greenfluorescent protein; ID, intradermal; LC, Langerhans cell; MVA, modifiedvaccinia virus Ankara; PLA, poly(lactic) acid

RESULTS

Poly D, L-lactic acid (PLA) or modified vaccinia virus Ankara(MVA) dermal injection induces morphological modificationsof epidermal LCs and their translocation to the dermis

The behavior of epidermal LCs has been monitored afterdermal inoculation of nanoparticles (6-Coumarin-labeledPLA, 6-Coumarin PLA), MVA enhanced green fluorescentprotein (EGFP), or phosphate-buffered saline (PBS; Figure 1and Supplementary Figure S1 online). Quantitative measure-ments of LC-associated dendrite length (Figure 1a), dendritenumber (Figure 1b), LC density (Figure 1c), LC area (mm2;Supplementary Figure S1a online), and percentage of LCs inthe epidermis (Supplementary Figure S1b online) have beenrecorded after immunostaining of 50 randomly chosenepidermal fields, using the CD207 marker. Representativemicroscopic epidermal fields or typical LCs are shown beloweach graph in Figure 1.

We observed a significant decrease in LC dendrite length(Figure 1a), dendrite number (Figure 1c), and LC area(Supplementary Figure S1a online) at 2 hours after ID inocula-tion of PLA (Po0.001) and at 4 hours after ID inoculation ofMVA EGFP (Po0.001) compared with PBS-injected animals.Independently of the type of compound, all morphologicalparameters slowly recovered baseline levels after 16 hoursand remained stable thereafter (Figure 1, SupplementaryFigure S1a online). Simultaneously, the density of LCs andthe percentage of LCs/epidermal sheet were significantlydecreased in epidermal layers 2–3 hours after 6-CoumarinPLA ID injection and 4 hours after MVA-EGFP (Figure 1c,Supplementary Figure S1b online), to recover the baselinelevel at 16 hours post ID injection. Thus, using multiplequantitative parameters, we clearly demonstrated that IDinoculation of either of the two distinct particulate com-pounds is responsible for major modifications in epidermalLCs, including changes in their morphology and densitywithin the epidermis.

Further, we investigated the translocation of LCs from theepidermis to the dermis. Epidermal or dermal cell suspen-sions (n¼ 6 mice, three independent experiments) wereprepared 6 hours after MVA ID injection. We used a multi-parametric flow cytometry analysis to distinguish epidermalLCs from other DDCs (Merad et al., 2008; Figure 2). In theepidermis, one could detect epidermal CD11cþCD207þ

CD103� resident LCs; three different cell populations wereidentified in the dermis (Figure 2a): (1) CD11cþCD103þ

CD207� cells representing CD103þ DDCs, (2) CD11cþ

CD103þCD207þ cells representing CD207þ DDCs (Gin-houx et al., 2007), and (3) CD11cþCD103�CD207þ cellsdefined as transmigrating epidermal LCs (Merad et al., 2008).We found a 2-fold decrease in the percentage ofCD11cþCD207þCD103� resident LCs in the epidermisafter ID injection of MVA compared with the percentage ofcontrol PBS-injected mice. On the contrary, we found a2-fold increase in the percentage of LCs present in thedermis after MVA ID administration compared with PBScontrol mice (Figure 2b). The percentage of CD103þ andCD207þ DDCs, which are essentially resident dermal cells,was not modified.

We also observed similar results on skin cryosections6 hours after MVA-EGFP ID injection (MVA but expressingthe green fluorescent protein EGFP). We showed thepresence of numerous LCs (CD207high) in the dermis at theID injection site (visualization of MVA-EGFP green fluores-cence), but not at the PBS injection site (Figure 2c). Together,these results strongly demonstrate that the dermal deliveryof a particle-based vaccine induces a rapid translocation ofepidermal LCs to the site of antigen delivery (dermis) within afew hours following the injection.

Epidermal LCs shuttle MVA antigens to the DLNs after IDinoculation

We then asked whether these transmigrating LCs werecapable of antigen capture before potential homing to theDLNs. At 6 hours after MVA-EGFP ID injection, dermal cellsuspensions were prepared (n¼5 mice, four independentexperiments). The dermal CD11cþCD103�CD207þ trans-migrating LCs were sorted by flow cytometry and assessed forMVA-EGFP detection (Figure 2d and e). MVA-EGFP wasdetectable in 0.62% of LCs isolated from the dermis,compared with PBS control mice (Figure 2d and e). Inaddition, MVA-EGFP particles were detected in sortedLCs (Figure 2d). These results clearly demonstrate that LCstraveling from the epidermis to the dermis were capable oftaking up vaccine compounds in vivo.

To investigate further the transport of antigens, DLNswere subsequently monitored for CD11cþCD207þCD103�

(resident and migrating) LCs at several time points after MVAdermal immunization (n¼ 7 mice, two independent experi-ments). First, a significant increase in the absolute number ofLCs was observed at 24 hours after MVA injection whencompared with PBS-treated mice (Po0.05; Figure 3a). TheseLCs expressed high levels of the activation marker CD8624 hours after MVA injection, as compared with the controlgroup (Po0.01; Figure 3b). More importantly, the percentageof MVA-EGFPþ LCs in the DLNs was also increased in MVA-EGFP-injected animals compared with control mice (n¼ 7mice, two independent experiments; Po0.05; Figure 3c).MVA-EGFPþ LCs from the DLNs were sorted by flowcytometry at 24 hours for microscopic analyses and showedthe intracellular presence of the green antigens (Figure 3d).These results are similar to previous observations in LCssorted from dermal suspensions (Figure 2d). Similarly,high levels of activation marker CD86 were detected inMVA-EGFPþ LCs from DLNs 24 hours after ID injectionwhen compared with control cells (Figure 5e).

Thus, epidermal LCs actively participate in antigenshuttling from the dermis to the DLNs within the first 24 hoursafter the dermal inoculation of a particulate vaccine.

Epidermal LCs are essential for the induction of MVA-specificCD8 T cells but not for MVA-specific CD4 T cells or thehumoral response after ID vaccination

To investigate the role of epidermal LCs in the inductionof immune responses following the dermal administration ofMVA, we used the Langerin–diphtheria toxin receptor (DTR)transgenic mice (Kissenpfennig et al., 2005). Figure 4a shows

2 Journal of Investigative Dermatology

C Liard et al.Epidermal LC Mobilization After ID Immunization

20 12

10

8

6

4

15

Mea

n de

ndrit

ele

ngth

(μm

)

10

*** ***5

0

2 Hours 48 Hours 4 Hours 24 Hours

2 Hours4

33

4

3

3

3

4 51

22

22 2

1

1 1

1

3,000

2,500

2,000

*** ***

Num

bers

of

LCs

per

mm

2

1,500

3,000

2,500

2,000

1,500

48 Hours 4 Hours 24 Hours

2 Hours 48 Hours 4 Hours 24 Hours

10

4.5

4.0

3.5

3.0Num

bers

of

dend

rites

per

cel

l

2.5

4.5

4.0

3.5

3.0

PBS PLA MVA

PBS PLA MVA

PBS PLA MVA

*** ***

20 30 40 50

0 10 20 30 40 50

0 10 20 30 40 50 0 10 20 30 40 50

0 10 20 30 40 50

0 10 20 30 40 50Time (hours) after ID injection

Time (hours) after ID injection

Time (hours) after ID injection

Figure 1. Morphological modifications and decrease in density of epidermal Langerhans cells (LCs) after intradermal (ID) injection of particle-based vaccines.

Epidermal sheets were immunostained with anti-CD207 (Texas Red) at different time points following ID injection of phosphate-buffered saline (PBS;

discontinued line), 6-Coumarin poly D, L-lactic acid (PLA; continued line, left panels), or MVA-EGFP (continued line, right panels). (a) LC dendrite length

in micrometers (mean±SD; n¼10–20 fields), (b) the number of primary dendrites/LC (mean±SD; n¼ 15–60 fields), or (c) the number of CD207þ cells per mm2

of epidermis (mean±SD; n¼9 mice per time point) were monitored at different time points. Mann–Whitney U-test: ***Po0.001. Representative LCs (a, b)

or epidermal sheets (c) are shown below each graph. Bar¼10 mm.

www.jidonline.org 3

C Liard et al.Epidermal LC Mobilization After ID Immunization

the protocol used for conditional depletion of CD207þ cellsin the skin, based on the previous work by the group ofMalissen (Bursch et al., 2007; Ginhoux et al., 2007). Asshown in Figure 4a, a first group of mice received diphtheriatoxin (DT) 2 days before ID inoculation of MVA (condition 1),lacking both LCs and CD207þ DDCs in the skin at the timeof vaccination. A second group of mice received DT 13 daysbefore ID injection, lacking only epidermal LCs at the timeof vaccination. These data were verified by skin histologystudies before our experiment and were compared with

control PBS-injected mice (Supplementary Figure S2 online).Indeed, we confirmed the results obtained by the group ofMalissen, which stipulate that 4 days after DT treatment theCD207þ DDCs have reconstituted the dermis, whereas ittakes 18 days for LCs to reconstitute the epidermis (Ginhouxet al., 2007).

At 7 days post ID immunization with MVA or PBS (controlgroup), we measured both antigen-specific cellular andhumoral responses. First, we assessed the absolute numbersand percentages of MVA-specific IFN-g-producing effector

Epidermis(gated on CD103–)

CD

207-

PE

CD11c-APC

100Epidermis

Gated on LCs CD11c+CD103–CD207+cells in the dermis

Dermis250

200

150

100

50

0

80

60

Per

cent

con

trol

LC

s

40

20

PBS MVA PBS MVA

0 %

MVA-EGFP

CD

207-

PE

0.62 %

0

CD103-PerCpCy5.5

23.9 %105

104

103

102

0 102 103 104 105

0 102 103 104 105 0 102 103 104 105

0102 103 104 105

0

105

104

103

102

0

105

104

103

102

0

105

104

103

102

0

1.34 %

1.14 %

1.49%

12.8 % 3.27% 0.15%

Dermis(gated on CD11c+)

PBS

HF

HF

D

E

HF

HF

CD207-TRDAPIMVA-EGFP (anti-MVA-A488)

CD207-TRDAPIMVA-EGFP(anti-MVA-A488)

D

E

PBS

MVA-EGFP MVA-EGFP

0.067 %

Figure 2. Migration of Langerhans cells (LCs) to the dermis and capture of modified virus Ankara (MVA)-enhanced green fluorescent protein (EGFP) after

intradermal (ID) immunization. (a) Flow cytometry analyses of epidermal or dermal cell suspensions 6 hours after phosphate-buffered saline (PBS) or MVA ID

immunization (n¼ 6 mice, four independent experiments). In the epidermis, resident LCs are CD11cþCD207þCD103�. In the dermis, cells are gated on

the CD11c marker: (1) resident CD207þ dermal dendritic cells (DDCs) (CD11cþCD207þCD103þ ), (2) resident CD103þ DDCs (CD11cþCD207�CD103þ ),

and (3) migrating epidermal LCs (CD11cþCD207þCD103� cells). (b) Percent control of LCs in the epidermis or dermis 6 hours after PBS or MVA ID

immunization. (c) Representative ear cryosections 6 hours after MVA-EGFP ID immunization. Bar¼ 25 mm. D, dermis; E, epidermis; HF, hair follicle.

(d) EGFPþ and EGFP� LCs sorted from dermal suspensions and immunostained as indicated, 6 hours after ID immunization. Bar¼5 mm (n¼2 independent

experiments). DAPI, 40,6-diamidino-2-phenylindole.

4 Journal of Investigative Dermatology

C Liard et al.Epidermal LC Mobilization After ID Immunization

CD4 and CD8 T cells in auricular DLNs, using anintracellular cytokine assay (Figure 4b and c). MVA-specificeffector CD8 T cells in auricular DLNs were severely reducedwhen both epidermal LCs and CD207þ DDCs were absent atthe time of ID injection (Figure 4b, condition 1; Po0.001),and also in mice that lacked only LCs (Figure 4b, condition 2;Po0.001) as compared with control PBS-treated mice. Incontrast, DT treatment did not affect MVA-specific IFN-g-producing effector CD4 T cells (Figure 4c, condition 1 or 2).Results are shown for neutralizing antibody titers (Figure 4d)in the serum of mice that lacked LCs only or both LCsand CD207þ DCs at the time of MVA ID injection.MVA-specific neutralizing antibodies remained unchangedin all conditions.

Thus, we have demonstrated the crucial role playedin vivo by epidermal LCs for the generation of MVA-specificCD8 T cells but not for MVA-specific CD4 T cells andhumoral responses after dermal administration of a vaccine.

Interlayer movement of LCs in human skin explants afterID injection of particle-based compounds

LC translocation was also investigated in human skin explantsafter ID injection of non-fluorescent MVA (1.5�107 plaque-forming unit (PFU)) or PBS. Similar to the experimentsconducted in the mouse model, we monitored LC morphol-ogy (as defined by dendrite length) and epidermal LC numberswith microscopic analyses, using anti-human CD1a in50 randomly chosen epidermal sheet layers (Figure 5b).

20,000 *60

**

40

20

% a

DLN

LC

s C

D86

high

0PBS MVA

15,000A

bsol

ute

no. o

f LC

s /a

DLN

s

10,000

5,000

0

PBS

0.5100

Gated on EGFP+LCs

80

60%

Eve

nts

40

20 CD86-PE-Cy7Isotype

00 103 104 105

*

0.4

0.3

% L

Cs

e-G

FP

+/2

aDLN

s

0.2

0.1

0.0PBS MVA-EGFP

CD207-A594

MVA-EGFP (anti-MVA-488)

DAPI

4Hours

8Hours

16Hours

24Hours

48Hours

Figure 3. Epidermal Langerhans cells (LCs) shuttle modified virus Ankara (MVA) antigens to the auricular draining lymph node (DLN). DLNs were analyzed by

flow cytometry after MVA-enhanced green fluorescent protein (EGFP) or phosphate-buffered saline (PBS) intradermal (ID) injection. (a) Kinetics of the absolute

number (mean±SD) of LCs and auricular DLNs, (b) percentage of total auricular DLN LCs (mean±SD) expressing high levels of CD86 marker (24 hours) and

(c) absolute number of auricular DLN LCs MVA-EGFPþ (mean±SD; 24 hours). (d) Auricular DLN LCs MVA-EGFPþ and MVA-EGFP� were sorted by flow

cytometry at 24 hours post immunization. A representative LC illustrates the intracellular localization of the green MVA-EGFP compared with its absence

in the PBS group. Bar¼5 mm. (e) CD86 expression in MVA-EGFPþLCs, 24 hours after ID injection (red line), compared with isotype control (black line;

n¼ 7 mice per time point). *Po0.05, **Po0.05; Mann–Whitney U-test. DAPI, 40,6-diamidino-2-phenylindole.

www.jidonline.org 5

C Liard et al.Epidermal LC Mobilization After ID Immunization

We observed a significant decrease in LC-associated den-drite length on epidermal sheets 4 hours after MVA ID injec-tion when compared with an ID injection of PBS (Po0.001;Figure 5a).

As depicted in Figure 5c, we found a significantaccumulation of LCs (high expression of CD1a) in the dermis4 hours following FITC nanoparticles ID inoculation whencompared with PBS-injected skin controls (Po0.001). Inter-estingly, after the ID injection of FITC nanoparticles,numerous CD1aþ LCs were distributed outside the baseof the epidermal layer in contrast to PBS-injected skin(Figure 5d). These results further support the relevanceof interlayer LCs movement after ID immunization ofparticle-based compounds in the human skin.

DISCUSSIONFollowing the dermal inoculation of vaccine compounds,epidermal LCs rapidly translocate into the dermis and

vigorously participate in CD8 but not CD4 and humoralanti-MVA immune responses. Morphological changes,trans-layer movement, uptake of antigens, and migration ofLCs from the skin to the DLN were the main features of theseearly events.

A report by Pearton et al. (2010) also demonstratesmorphological changes of LCs in human skin explants afterthe dermal injection of virus-like particles. We have,however, noted differences in the kinetics of LC modifica-tions between the two studies. In the work of Pearton et al.,there is a marked lack of monitoring of changes in both themorphology and number of LCs at time points earlier than24 hours following ID injection. Indeed, we found thatthe presence of LCs in the dermis was transient (only a fewhours). It has been shown that under inflammatory condi-tions, epidermal LCs derived from blood-borne monocytesrapidly replenished the epidermis (Merad et al., 2002). Theexpression of CCR6 on the surface of these blood-borne

** Condition 2

* Condition 1

Day 0Day 2Day 13

DT1

0.2050,000

40,000

30,000

Abs

olut

e no

. of I

FN

-γ+

CD

8 T

cel

ls/2

DLN

s

Abs

olut

e no

. of I

FN

-γ+

CD

4 T

cel

ls/2

DLN

s20,000

10,000 NS

NS

*********

0

******

NS

0.15

0.10

0.05NS

0.00

0.4 1,00,000

80,000

60,000

40,000

20,000

100

PBSPBS + DTMVAMVA + DT1

MVA + DT275

50

% O

f MV

Ase

rone

utra

lizat

ion

25

00 1

Log (dilution)

2 3 4 50

***

***

***

0.3

0.2

% IF

N-γ

+ C

D4

T c

ells

0.1NS

NS

NS

NS

NS

NS

NS

0.0

PBS PBS MVA MVA MVA– –DT DT1 DT2

PBS PBS MVA MVA MVA– –DT DT1 DT2

PBS PBS MVA MVA MVA– –DT DT1 DT2

PBS PBS MVA MVA MVA– –DT DT1 DT2

DT2 MVA MVA-specific T cells

Day 7 Day 21

% IF

N-γ

+ C

D8

T c

ells

MVA-specific Ab titers

Figure 4. Conditional depletion of epidermal Langerhans cells (LCs) showed their contribution to the induction of modified virus Ankara (MVA)-specific

IFN-c-secreting CD8þ T cells, but not of CD4 T cells nor humoral responses after intradermal (ID) immunization. (a) Langerin–DTR mice were injected by

intraperitoneal route with 1mg diphtheria toxin (DT) on day 2 (condition 1: DT treatment 2 days before ID vaccination for depletion of both CD207þ LC and

CD207þDDC) or day 13 (condition 2: DT treatment 13 days before ID vaccination for depletion of both CD207þ LC and CD207þDDC and reconstitution

of the pool of CD207þDDC) before ID vaccination with either phosphate-buffered saline (PBS) or MVA. The PBS DT group is a pool of mice from conditions 1

and 2. Auricular draining lymph nodes (DLNs) were recovered 7 days post injection and stimulated in vitro for 16 hours with MVA (0.1 PFU per cell).

MVA-specific IFN-g-producing CD8 (b) and CD4 T cells (c) were analyzed by flow cytometry. Percentage (mean±SD; left panels) or absolute numbers

(mean±SD; right panels) of MVA-specific IFN-gþ CD8 T cells or MVA-specific IFN-gþ CD4 T cells are shown. Sera were collected 21 days after ID

vaccination. MVA-specific antibody titers (d) were measured in vitro by an MVA seroneutralization assay. Data are representative of three distinct experiments,

n¼ 6–10 mice. ***Po0.01, NS, nonsignificant; Mann–Whitney U-test.

6 Journal of Investigative Dermatology

C Liard et al.Epidermal LC Mobilization After ID Immunization

monocytes then allows them to differentiate into LCs in theepidermis in response to the expression of the chemokineMIP3a, which is secreted by resident keratinocytes (Ginhouxet al., 2006). This could explain why, in our study, LCnumbers in the epidermis were already replenished 24 hours

after ID injection. The different molecular mechanismsfeaturing LC reconstitution in the skin after an ID inoculationneed to be further studied.

In addition to the previous work of Pearton et al. (2010), weprovide the first evidence for the involvement of epidermal LCs

201,500

1,000

No.

of C

D1a

+ c

ells

per

mm

2

abdo

men

epi

derm

is

500

0

*** ***

15M

ean

dend

rite

leng

th(μ

m)/

CD

1a+ c

ell

10

5

0PBS

PBS

MVA

MVA

400***

300

No.

of C

D1a

+ c

ells

per

mm

2

inje

cted

der

mis

200

100

E

D

E

E

D

D

PBS CD1a-Alexa594FITC-NPDAPI

H

0PBS NP

FITC-NPs

CD1a-Alexa488

PBS MVA

PBS MVA

Figure 5. Intradermal (ID) injection of nanoparticulate compounds influences the morphological features and translocation of human epidermal

CD1aþ Langerhans cells. Human epidermal sheets immunostained with anti-CD1a 4 hours following phosphate-buffered saline (PBS) (a) or modified virus

Ankara (MVA) (b) ID injection (n¼ 2 independent experiments). (a) Langerhans cells (LC dendrite length (mm; mean±SD; n¼10–20 fields). Representative

CD1aþ LCs are shown. Scale bar¼ 10mm. (b) Number of CD1aþ cells per mm2 (mean±SD). Representative human epidermal sheets are shown (140 mm2,

n¼ 10–20 fields). Bar¼ 25mm. (c) Number of CD1ahigh cells per mm2 epidermis (mean±SD) 4 hours after FITC nanoparticles (NPs) or PBS ID injection

(n¼50 fields). (d) Representative human skin cryosections after FITC-NPs or PBS ID are shown (n¼30 fields). Bar¼100 mm. D, dermis; DAPI, 40,6-diamidino-2-

phenylindole; E, epidermis; HF, hair follicle.

www.jidonline.org 7

C Liard et al.Epidermal LC Mobilization After ID Immunization

in particulate-vaccine shuttling from the dermis to draininglymphoid tissues. We and others have shown that in addition toskin-resident APCs, other inflammatory cells are recruited tothe site of injection after the ID administration of a vaccine;e.g., neutrophils in the case of the Bacillus Calmette-Guerin(Abadie et al., 2005) or blood-borne monocytes in the MVAmodel (Abadie et al., 2009). These cell subsets also participatein antigen transport from the skin to the DLN.

Allan et al. (2003, 2006) have suggested that in the case ofan Herpes simplex virus skin infection, LCs do not directlyprime T cells in the DLNs but function as ‘‘antigen carriers’’for CD8aþ DLN-resident DCs. Regardless of the scenarioinvolved, i.e., direct T-cell priming by skin migrating LCs orindirect cross-priming through CD8aþ DLN-resident DCsthat obtain antigens from ‘‘cargo LCs’’, our results show thatepidermal LCs are indispensable for the induction of CD8T-cell responses. However, LCs were expendable for thegeneration of MVA-specific CD4 T cells, as well as specificneutralizing antibodies. Our findings are consistent withrecent data in the literature, which suggest that humoral andcellular responses are regulated by different subsets of skinand DLN DCs with distinct intrinsic properties (Ueno et al.,2010a). In addition, several in vitro studies have shown thatCD34þ -derived LCs and human skin LCs were more potentat priming IFN-g-secreting CD8þ T cells than DDCs(Caux et al., 1997; Ratzinger et al., 2004; Klechevsky et al.,2008). Moreover, Kissenpfennig et al. (2005) have demon-strated, in the Langerin–DTR murine model, that after topicalimmunization (TRITC-painting), epidermal LCs were prefer-entially localized in the inner paracortex of the DLN, in theT-cell zone, in close proximity to the CD8aþ -resident DCs.LCs are also capable of inducing cytotoxic cellular immuneresponses for protection against tumors in mice (Stoitzneret al., 2008) and also in humans (Fay et al., 2006). The factthat transcutaneous vaccination targeting LCs surroundinghair follicles (Vogt et al., 2006) induces preferentiallyCD8-mediated cellular immunity (Combadiere et al., 2010)confirms this unique feature of LCs.

Langerin–DTR mouse model allows to explore the roleof CD207þ skin DCs (DDCs and/or LCs) in the initiation ofimmune responses to different types of vaccine administra-tions or infections (Romani et al., 2010a). Recent studies havehighlighted the importance of CD207þ DCs for the inductionof specific CD8 T-cell responses after an ID plasmid DNAimmunization (Elnekave et al., 2010) or after an ID injectionof a lentiviral vector (Furmanov et al., 2010), as well as in thecontext of gene gun vaccination (Stoecklinger et al., 2011).LCs are major factors of the immunity against skin virusessuch as Herpes simplex virus and Varicella, which functionby transporting the virus to the DLN (Bedoui et al., 2009;Cunningham et al., 2010). In addition, both LCs and CD207þ

DDCs would be involved in contact hypersensitivity depend-ing on the hapten dose (Kaplan et al., 2005; Bennett et al.,2007; Honda et al., 2010; Noordegraaf et al., 2010).Eventually, Nudel et al. (2011) proved the crucial role ofLCs in the induction of CD8 T cells after an oral mucosalimmunization. Indeed, the nature of the vaccine preparationor of the pathogen, as well as the antigen dose and the route

of administration, are important parameters that couldinfluence the differential involvement of LCs or CD207þ

DCs for the induction of immune responses. LCs areundeniably of major importance for the induction of Th1immune responses; i.e., to fight Borrelia burdorferi in theLyme disease (Vesely et al., 2009), or for the generation ofalloreactive cytotoxic T lymphocytes in the skin after bonemarrow stem cell transplantation (graft-versus-host disease;Merad et al., 2004).

Contrary to transcutaneous or topical application ofvaccine compounds, the ID route does not preferentiallytarget LCs, but instead target resident DDCs, which aremore involved in the induction of T-follicular helper (Tfh)cells in vitro. This specific T-cell subset contributes to theswitching and proliferation of B cells within the germinalcenters, leading to the production of large amounts ofdifferent subtypes of immunoglobulins (Klechevsky et al.,2008). The ID route thus allows the mobilization of both armsof the immune response by harnessing the unique plasticity ofcutaneous APCs distributed among the different skin com-partments. Our work is of major importance for the under-standing of cellular mechanisms related to an ID inoculation.It provides new evidence for the role of LCs in the inductionof immune responses to particle-based vaccines injected intothe skin. Moreover, it paves the way for other studies that willaim at fully understanding the primary immune eventstriggered by ID administered compounds.

MATERIALS AND METHODSMice

Female Balb/c mice (6- to 8-week old) were purchased from Charles

River Laboratories (L’Arbresle, Orleans, France) and Langerin–DTR

mice (Kissenpfennig et al., 2005) from the Transgenesis, Archiving

and Animal Models (CNRS laboratory, Orleans, France). Langer-

in–DTR mice received 1 mg of diphtheria toxin (Sigma-Aldrich, Saint-

Quentin Fallavier, France) intraperitoneally 2 or 13 days before ID

vaccination. All animals were housed at the specific pathogen-free

animal facility of the Pitie-Salpetriere hospital (Paris, France). All

experiments complied with the local animal experimentation and

ethics committee guidelines.

Human skin explants

Fresh skin samples were obtained from healthy volunteers under-

going plastic surgery: mandibular region of the face or retroareolar

region of the breast (Charite Hospital, Berlin, Germany). All skin

explants were taken after informed consent obtained by the

Institutional Ethics Committee of the Charite Hospital (Berlin,

Germany) according to the ethical rules stated in the Declaration

of Helsinki Principles.

A few hours after surgical excision, intact skin samples were injected

ID (Mantoux injection technique) with 6� 105 FluoSpheres (40nm

yellow-green FITC nanoparticles; Molecular Probes, Life Technologies,

Carlsbad, CA), 1.5� 107 PFU MVA, or PBS, using U-100 insulin needles

(29G� 1/200�0.33� 12 mm; Terumo, Interleuvenlaan, Belgium).

Mice immunization

Mice received a single ID injection at the back of the ear as

previously described (Abadie et al., 2009). A volume of 30 ml of

8 Journal of Investigative Dermatology

C Liard et al.Epidermal LC Mobilization After ID Immunization

vaccine was used; i.e., 3.0� 1010 6-Coumarin PLA (Lamalle-Bernard

et al., 2006; Primard et al., 2010) or 1.5� 107 PFU MVA-EGFP

(fluorescent virus particles) or MVA (non-fluorescent). Control

animals were injected ID with PBS.

MVA-EGFP neutralization assay

Blood was collected in dry Eppendorf tubes before ID immunization

(pre-immune samples) and at 21 days post priming. Centrifuged sera

were collected and stored at �80 1C.

Neutralizing antibody titers were determined using a neutraliza-

tion assay as previously described (Abadie et al., 2009). The

percentage of neutralization was calculated as follows: (1�[percen-

tage of GFP-expressing cells in treated samples/percentage of

GFP-expressing cells in untreated controls])� 100.

Preparation of murine cell suspensions

DLNs were treated for 30 minutes at 37 1C with collagenase IV

(Sigma-Aldrich), 400 U ml�1 in RPMI 1640 medium (Invitrogen

Europe, Paisley, UK), pressed through a 70-mm nylon mesh

(cell strainer, BD Falcon, Franklin Lakes, NJ), and washed with

PBS supplemented with fetal calf serum (FCS).

For skin cell preparation, epidermis was separated from the

dermis as follows: the dorsal halves of the ears were incubated,

epidermal side down, for 30 minutes at 37 1C in ammonium

thiocyanate phosphate buffer (0.5 M). The dermal layer was

mechanically separated by gently peeling off the epidermis with

tweezers. Small pieces of dermis and epidermis were pooled

separately in 2.5 U Dispase-II (Sigma) and dissociated overnight

with gentle shaking at 4 1C. Digested tissues were pressed through a

70-mm nylon mesh (cell strainer) and then filtered (Miltenyi, Auburn,

CA) in order to remove most aggregates.

Flow cytometry

Cell suspensions were stained for surface markers (20 minutes at 4 1C

in PBS 1� , 2% FCS) with biotinylated anti-mouse CD103 (Integrin

a-IEL chain; clone M290) and anti-mouse CD11c-APC (clone HL3).

Streptavidin-PerCp was used for detection. All antibodies were

purchased from BD Biosciences (Franklin Lakes, NJ). Anti-mouse

CD86-FITC (clone GL1) or anti-mouse CD86-PE-Cy7 allowed the

detection of activated cells. Cells were sequentially fixed with 4%

paraformaldehyde for 20 minutes at 4 1C, rinsed, and permeabilized

with 2% FCS—0.1% saponin (Sigma)—PBS 1� . Intracellular

staining was performed using PE-conjugated anti-mouse CD207/

Langerin (clone eBioL31). Fluorescence was analyzed on a total of

2� 105 cells of each population per sample with a LSRII flow

cytometer or a FACSCalibur and analyzed using the FlowJo software

(Becton Dickinson, San Jose, CA).

Intracellular cytokine assay

DLN cells were cultured for 16 hours at 37 1C with 0.1 PFU MVA per

cell, RPMI (negative control), or 2.5 mg concanavalin A (positive

control). This was incubated for 4 hours at 37 1C with 5 mg ml�1

brefeldin A to block cytokine secretion. Cellular suspensions were

washed and surface stained for 20 minutes at 4 1C with anti-mouse

CD8a (Ly-2)–PercP-Cy5.5 (clone 53-6.7) and anti-mouse CD4-PE

(BD Biosciences), respectively, in PBS-5% FCS. Cells were fixed with

4% paraformaldehyde before intracellular staining with a mix of

anti-mouse IL2-FITC and anti-mouse IFN-g-APC (20 minutes at 4 1C).

Fluorescence was monitored on a minimum of 50,000 CD8þ /CD4þ

cells with a FACSCalibur (BD Biosciences) and analyzed using

the FlowJo software.

Immunofluorescence

Mice ear skin (ID injection sites) or auricular DLNs were removed

after ID injection at indicated time points. Cell suspensions

(epidermis, dermis, or auricular DLN) were surface stained using

anti-mouse CD103-biotin (streptavidin-PerCp-Cy5.5) and anti-

mouse CD11c-APC. CD11cþCD103� cells with or without MVA-

EGFP were sorted using a FACSVantage cell sorter (Becton

Diskinson) and coated onto Poly-L-lysine (Sigma-Aldrich) coverslips

in RPMI with 30% FCS. Cell staining was performed directly on

the coverslips with slight shaking, using purified biotinylated rat

anti-CD207 (Euromedex, Soufellweyersheim, France).

Murine skin was removed at indicated time points, embedded

in optimum cutting temperature medium (Tissue-Tek, Sakura

Finetek, Alphen aan den Rijn, The Netherlands), frozen, and

sequentially cryosectioned (5mm) with a Microm HM550 cryostat

(Thermo Scientific, Courtaboeuf, France). Immunostainings were

performed as described previously (Mahe et al., 2009) using

biotinylated rabbit-anti-MVA (AbCys SA), Streptavidin Alexa 488

(Molecular Probes), purified biotinylated rat anti-CD207 (Eurome-

dex), and chicken anti-rat IgG-Alexa Fluor 594 (Invitrogen Europe).

Slides were coverslipped with Vectashield mounting medium

containing 40,6-diamidino-2-phenylindole (Vector Laboratories,

Abscys sa, Paris, France). The dermal layer of mouse ears was

mechanically separated from the epidermis as described above. After

fixation in 1% paraformaldehyde (20 minutes at 4 1C), epidermal

sheets were incubated by gently shaking with, in sequence, a

biotinylated anti-mCD207 mAb (Clone 929F3-01; Euromedex) and

Texas Red-conjugated streptavidin (Zymco, Invitrogen, Carlslab,

CA). They were mounted for microscopy using the Fluoromount-G

medium (Southern-Biotechnology Associates, Birmingham, UK).

Human skin explants were incubated at 37 1C in a Petri dish with

DMEM-10% FCS for 4 hours after ID injection. Human skin samples

were frozen and cryosectioned as described for mouse skin. Sections

were immunostained as described previously (Vogt et al., 2006),

using, in sequence, mouse anti-hCD1a (Dako, Glostrup, Denmark)

and a secondary anti-mouse Alexa-594 antibody (Molecular Probes)

in PBS 1� containing 1% BSA. Slides were coverslipped with

Fluoromount-G supplemented with 40,6-diamidino-2-phenylindole.

Human epidermal sheets were prepared as follows: skin tissues

were digested overnight at 4 1C as described previously (Kitano and

Okada, 1983). Epidermal sheets were separated and fixed in 1%

paraformaldehyde in PBS 1� for 20 minutes at 4 1C. Immunostain-

ing was performed using, in sequence, mouse anti-human CD1a

antibodies (Dako; Glostrup, Denmark, Biotest, Dreieich, Germany)

and secondary anti-mouse-FITC antibodies diluted in human serum

(Vector laboratories, Burlingame, CA), and mounted for microscopy

using Fluoromount-G medium (Southern-Biotechnology Associates).

Microscopy analyses

All tissues were analyzed with a fluorescence microscope (BX51;

Olympus, Rungis, France) or a BX60F3 microscope (Olympus,

Hamburg, Germany) equipped with an image processing and

analysis system (Qimaging; Media Cybernetics, Bethesda, MD).

All measurements were recorded using the software ImageJ

www.jidonline.org 9

C Liard et al.Epidermal LC Mobilization After ID Immunization

(NIH, Bethesda, MD). Quantitative parameters for LCs morphology

and density in skin tissues have been chosen according to the

literature ((Pearton et al., 2010, skin LCs, or (Gervaz et al., 1995,

anal mucosa LCs)).

Statistical analyses

Data are presented as the mean±standard deviation. Statistical

analyses were performed with the Graphpad software. P-values of

o0.05 were considered statistically significant.

CONFLICT OF INTERESTThe authors state no conflict of interest.

ACKNOWLEDGMENTSThis work was supported by the EU-FP6 health program MuNanoVac‘‘Mucosal HIV Vaccines’’ and the EU-FP7 health program CUT’HIVAC‘‘Cutaneous and Mucosal HIV Vaccination’’. C. Liard was supported by agrant from the Fondation pour la Recherche Medicale (FRM). B. Combadiereis an awardee of the INSERM-Interface AP/HP program. Darragh Duffy was arecipient of a grant from the Agence National de la Recherche Contre le SIDA(ANRS). We thank the Imaging Facility of the Pitie-Salpetriere Hospital, andDr Aurelien Dauphin in particular, for helpful advice.

SUPPLEMENTARY MATERIAL

Supplementary material is linked to the online version of the paper at http://www.nature.com/jid

REFERENCES

Abadie V, Badell E, Douillard P et al. (2005) Neutrophils rapidly migrate vialymphatics after Mycobacterium bovis BCG intradermal vaccination andshuttle live bacilli to the draining lymph nodes. Blood 106:1843–50

Abadie V, Bonduelle O, Duffy D et al. (2009) Original encounter withantigen determines antigen-presenting cell imprinting of the quality ofthe immune response in mice. PLoS One 4:e8159

Allan RS, Smith CM, Belz GT et al. (2003) Epidermal viral immunity inducedby CD8alpha+ dendritic cells but not by Langerhans cells. Science301:1925–8

Allan RS, Waithman J, Bedoui S et al. (2006) Migratory dendritic cells transferantigen to a lymph node-resident dendritic cell population for efficientCTL priming. Immunity 25:153–62

Bedoui S, Whitney PG, Waithman J et al. (2009) Cross-presentation of viraland self antigens by skin-derived CD103+ dendritic cells. Nat Immunol10:488–95

Belshe RB (2007) The burden of influenza and strategies for prevention.Manag Care 16:2–6

Bennett CL, Noordegraaf M, Martina CA et al. (2007) Langerhans cells arerequired for efficient presentation of topically applied hapten to T cells.J Immunol 179:6830–5

Bursch LS, Wang L, Igyarto B et al. (2007) Identification of a novel populationof Langerin+ dendritic cells. J Exp Med 204:3147–56

Caux C, Massacrier C, Vanbervliet B et al. (1997) CD34+ hematopoieticprogenitors from human cord blood differentiate along two independentdendritic cell pathways in response to granulocyte-macrophagecolony-stimulating factor plus tumor necrosis factor alpha: II. Functionalanalysis. Blood 90:1458–70

Charest AF, McDougall J, Goldstein MB (2000) A randomized comparisonof intradermal and intramuscular vaccination against hepatitis B virusin incident chronic hemodialysis patients. Am J Kidney Dis 36:976–82

Combadiere B, Vogt A, Mahe B et al. (2010) Preferential amplification of CD8effector-T cells after transcutaneous application of an inactivatedinfluenza vaccine: a randomized phase I trial. PLoS One 5:e10818

Cunningham AL, Abendroth A, Jones C et al. (2010) Viruses and Langerhanscells. Immunol Cell Biol 88:416–23

Egemen A, Aksit S, Kurugol Z et al. (1998) Low-dose intradermal versusintramuscular administration of recombinant hepatitis B vaccine: acomparison of immunogenicity in infants and preschool children.Vaccine 16:1511–5

Elnekave M, Furmanov K, Nudel I et al. (2010) Directly transfected langerin+dermal dendritic cells potentiate CD8+ T cell responses followingintradermal plasmid DNA immunization. J Immunol 185:3463–71

Fay JW, Palucka AK, Paczesny S et al. (2006) Long-term outcomes in patientswith metastatic melanoma vaccinated with melanoma peptide-pulsed CD34(+) progenitor-derived dendritic cells. Cancer ImmunolImmunother 55:1209–18

Frech SA, Kenney RT, Spyr CA et al. (2005) Improved immune responsesto influenza vaccination in the elderly using an immunostimulant patch.Vaccine 23:946–50

Furmanov K, Elnekave M, Lehmann D et al. (2010) The role of skin-deriveddendritic cells in CD8+ T cell priming following immunization withlentivectors. J Immunol 184:4889–97

Gervaz E, Dauge-Geffroy MD, Sobhani I et al. (1995) Quantitative analysisof the immune cells in the anal mucosa. Pathol Res Pract 191:1067–71

Ginhoux F, Collin MP, Bogunovic M et al. (2007) Blood-derived dermallangerin+ dendritic cells survey the skin in the steady state. J Exp Med204:3133–46

Ginhoux F, Tacke F, Angeli V et al. (2006) Langerhans cells arise frommonocytes in vivo. Nat Immunol 7:265–73

Henderson EA, Louie TJ, Ramotar K et al. (2000) Comparison of higher-doseintradermal hepatitis B vaccination to standard intramuscular vaccinationof healthcare workers. Infect Control Hosp Epidemiol 21:264–9

Honda T, Miyachi Y, Kabashima K (2010) The role of regulatory T cells incontact hypersensitivity. Recent Pat Inflamm Allergy Drug Discov 4:85–9

Kaplan DH, Jenison MC, Saeland S et al. (2005) Epidermal langerhans cell-deficient mice develop enhanced contact hypersensitivity. Immunity23:611–20

Kissenpfennig A, Henri S, Dubois B et al. (2005) Dynamics and function ofLangerhans cells in vivo: dermal dendritic cells colonize lymph node areasdistinct from slower migrating Langerhans cells. Immunity 22:643–54

Kitano Y, Okada N (1983) Separation of the epidermal sheet by dispase.Br J Dermatol 108:555–60

Klechevsky E, Morita R, Liu M et al. (2008) Functional specializations ofhuman epidermal Langerhans cells and CD14+ dermal dendritic cells.Immunity 29:497–510

Kupper TS, Fuhlbrigge RC (2004) Immune surveillance in the skin:mechanisms and clinical consequences. Nat Rev Immunol 4:211–22

Lamalle-Bernard D, Munier S, Compagnon C et al. (2006) Coadsorption ofHIV-1 p24 and gp120 proteins to surfactant-free anionic PLA nanopar-ticles preserves antigenicity and immunogenicity. J Control Release115:57–67

Lambert PH, Laurent PE (2008) Intradermal vaccine delivery: will new deliverysystems transform vaccine administration? Vaccine 26:3197–208

Mahe B, Vogt A, Liard C et al. (2009) Nanoparticle-based targeting of vaccinecompounds to skin antigen-presenting cells by hair follicles and theirtransport in mice. J Invest Dermatol 129:1156–64

Merad M, Ginhoux F, Collin M (2008) Origin, homeostasis and function ofLangerhans cells and other langerin-expressing dendritic cells. Nat RevImmunol 8:935–47

Merad M, Hoffmann P, Ranheim E et al. (2004) Depletion of host Langerhanscells before transplantation of donor alloreactive T cells prevents skingraft-versus-host disease. Nat Med 10:510–7

Merad M, Manz MG, Karsunky H et al. (2002) Langerhans cells renew inthe skin throughout life under steady-state conditions. Nat Immunol3:1135–41

Nicolas JF, Guy B (2008) Intradermal, epidermal and transcutaneousvaccination: from immunology to clinical practice. Expert Rev Vaccines7:1201–14

Noordegraaf M, Flacher V, Stoitzner P et al. (2010) Functional redundancy ofLangerhans cells and Langerin+ dermal dendritic cells in contacthypersensitivity. J Invest Dermatol 130:2752–9

10 Journal of Investigative Dermatology

C Liard et al.Epidermal LC Mobilization After ID Immunization

Nudel I, Elnekave M, Furmanov K et al. (2011) Dendritic cells in distinct oralmucosal tissues engage different mechanisms to prime CD8+ T cells.J Immunol 186:891–900

Pearton M, Kang SM, Song JM et al. (2010) Changes in human Langerhanscells following intradermal injection of influenza virus-like particlevaccines. PLoS One 5:e12410

Primard C, Rochereau N, Luciani E et al. (2010) Traffic of poly(lactic acid)nanoparticulate vaccine vehicle from intestinal mucus to sub-epithelialimmune competent cells. Biomaterials 31:6060–8

Ratzinger G, Baggers J, de Cos MA et al. (2004) Mature human Langerhans cellsderived from CD34+ hematopoietic progenitors stimulate greater cytolytic Tlymphocyte activity in the absence of bioactive IL-12p70, by either singlepeptide presentation or cross-priming, than do dermal-interstitial ormonocyte-derived dendritic cells. J Immunol 173:2780–91

Romani N, Clausen BE, Stoitzner P (2010a) Langerhans cells and more:langerin-expressing dendritic cell subsets in the skin. Immunol Rev234:120–41

Romani N, Thurnher M, Idoyaga J et al. (2010b) Targeting of antigens to skindendritic cells: possibilities to enhance vaccine efficacy. Immunol CellBiol 88:424–30

Sabchareon A, Chantavanich P, Pasuralertsakul S et al. (1998) Persistence ofantibodies in children after intradermal or intramuscular administrationof preexposure primary and booster immunizations with purified Verocell rabies vaccine. Pediatr Infect Dis J 17:1001–7

Stoecklinger A, Eticha TD, Mesdaghi M et al. (2011) Langerin+ dermal dendriticcells are critical for CD8+ T cell activation and IgH gamma-1 classswitching in response to gene gun vaccines. J Immunol 186:1377–83

Stoitzner P, Green LK, Jung JY et al. (2008) Tumor immunotherapy by epicu-taneous immunization requires langerhans cells. J Immunol 180:1991–8

Sumida SM, McKay PF, Truitt DM et al. (2004) Recruitment and expansionof dendritic cells in vivo potentiate the immunogenicity of plasmidDNA vaccines. J Clin Invest 114:1334–42

Ueno H, Palucka AK, Banchereau J (2010a) The expanding family of dendriticcell subsets. Nat Biotechnol 28:813–5

Ueno H, Schmitt N, Klechevsky E et al. (2010b) Harnessing human dendriticcell subsets for medicine. Immunol Rev 234:199–212

Vesely DL, Fish D, Shlomchik MJ et al. (2009) Langerhans cell deficiencyimpairs Ixodes scapularis suppression of Th1 responses in mice. InfectImmun 77:1881–7

Vien NC, Feroldi E, Lang J (2008) Long-term anti-rabies antibody persistencefollowing intramuscular or low-dose intradermal vaccination of youngVietnamese children. Trans R Soc Trop Med Hyg 102:294–6

Vogt A, Combadiere B, Hadam S et al. (2006) 40 nm, but not 750 or1,500 nm, nanoparticles enter epidermal CD1a+ cells after transcuta-neous application on human skin. J Invest Dermatol 126:1316–22

Vogt A, Mahe B, Costagliola D et al. (2008) Transcutaneous anti-influenzavaccination promotes both CD4 and CD8 T cell immune responses inhumans. J Immunol 180:1482–9

www.jidonline.org 11

C Liard et al.Epidermal LC Mobilization After ID Immunization