The haemopoietic growth factor, Flt3L, alters the immune response induced by transcutaneous immunization

Abstract

Topical application of antigen induces antigen-specific humoral and cellular immune responses. In this study we examined whether expansion of dendritic cells (DC) by Flt3 ligand (Flt3L) treatment influences the induction of immune responses following transcutaneous immunization. Mice were treated intraperitoneally with Flt3L or phosphate-buffered saline (PBS) and immunized transcutaneously with hen egg lysozyme (HEL). Flt3L-treated mice developed lower HEL-specific cellular and humoral immune responses than PBS-treated mice. However, in the presence of cholera toxin (CT), a potent adjuvant for mucosal and transcutaneous immunization, Flt3L-treated mice developed significantly higher cellular and humoral immune responses to HEL when compared to PBS-treated mice. We assessed whether the immunomodulatory effects of CT were a result of activation of epidermal dendritic cells (Langerhans' cells; LC). Our results indicate that within 8–12 hr of topical application of CT, epidermal LC cells lose their dendritic morphology and become rounder in appearance. In addition, we observed enhanced expression of major histocompatibility complex (MHC) class II, and of adhesion molecules CD11c and intracellular adhesion molecule-1 (ICAM-1). Our observations support the concept that the state of activation of DC in the skin is central to the regulation of immune responses. This information is relevant to the design of effective transcutaneous vaccination strategies.

Introduction

Non-invasive immunization strategies onto the skin are attractive alternatives for the development of painless and safe vaccines. The epidermis of the skin is populated with a network of dendritic cells (Langerhans' cells; LC), potent antigen-presenting cells (APC) capable of migrating from the epidermis to the draining lymph node where they subsequently present antigen to T cells.^1–3 The skin is therefore an active immunological microenvironment offering a unique potential as a target tissue for immunization. We have shown that topical application of antigen formulated in a lipid-based delivery system promotes the induction of cellular and humoral immune responses with a strong bias towards a T helper 2 (Th2) response.⁴ Models of epicutaneous sensitization to protein antigens are consistent with the observations that the cutaneous microenvironment promotes the induction of Th2 responses.^5,6 However, little is known about the mechanisms that mediate immune responses to proteins delivered to intact skin and consequently, we lack understanding on how the manipulation of the skin microenvironment influences the outcome of these responses.

Treatment of mice with the haemopoietic growth factor Flt3 ligand (Flt3L) increases the numbers of dendritic cells (DC) in lymphoid and non-lymphoid organs, including the skin.^7–9 In addition, Flt3L treatment enhanced mucosal immune responses meditated by cholera toxin (CT)¹⁰ and, paradoxically, the ability of soluble antigen to induce oral tolerance.⁸ These studies demonstrated the important regulatory role of DC on mucosal immunity and tolerance, and provided an in vivo model to assess the role of DC in other systems.

CT has been shown to be an adjuvant for transcutaneous immunization.¹¹ Although the precise mechanism by which CT mediates the induction of immune responses following transdermal immunization is unknown, it is likely that the adjuvant effect is related to the ability of CT to stimulate APC.^12–14 Infectious agents and inflammatory signals induce mobilization of DC from the periphery to lymph node T-cell areas and shift DC from a processing to a presenting stage.¹ Thus, elucidation of the factors that contribute to the regulation of DC should provide us with the ability to improve immunization strategies.

In this study, we examined whether expansion of DC by Flt3L treatment influenced the induction of immune responses following transcutaneous immunization. We showed that transcutaneous immunization of Flt3L-treated mice with hen egg lysozyme (HEL) resulted in lower HEL-specific cellular and humoral responses when compared to immune responses induced in phosphate-buffered saline (PBS)-treated mice. However, in the presence of CT, Flt3L-treated mice develop significantly higher cellular and humoral immune responses to HEL than PBS-treated mice. In addition, we demonstrate that CT induced activation of LC, thus, providing further evidence that the state of activation of DC is central to the regulation of immune responses to antigens delivered through the skin.

Materials and methods

Experimental animals

Six-week-old-female BALB/c mice were purchased from the Animal Resources Center at the University of Saskatchewan. Animals were handled according to the guidelines of the Canadian Council on Animal Care and the University of Saskatchewan Committee on Animal Care and Supply.

Treatment/immunizations

Mice were treated intraperitoneally with 10 µg of Flt3L (containing 0·1% normal mouse serum as a carrier protein; a generous gift from Immunex, Seattle, WA) once daily for 10 consecutive days. Control animals received 0·1% normal mouse serum in PBS. Mice were immunized on day 9 of Flt3L treatment. Transcutaneous immunizations were performed as previously described.⁴ Briefly, animals were anaesthetized by inhalation of nitrous oxide/halothane (MTC Pharmaceuticals, Ontario, Canada) and shaved on the back with an electric blade. Twenty-four hr later, mice were transcutaneously immunized with 60 mg of 12·5% methylcelullose gel (Aldrich, Oakville, Ontario, Canada) containing either 50 µg of HEL (Sigma Chemical Co., St. Louis, MO) or 1 µg of CT (List Biologicals, Campbell CA) or a combination of the two. Patches were applied and secured with an adhesive membrane (Opsite, Smith and Nephew Medical Ltd., London, UK) and tape. The patches were carefully removed 48 hr later and the site of application was thoroughly cleaned. In some experiments the immunization procedure was repeated 14 days later. Animals were killed 10 days after the last immunization to evaluate the induction of antigen-specific cellular and humoral immune responses.

Assessment of antigen-specific cytokine secreting cells

Cells from the draining lymph nodes (axillary and inguinal) were aseptically removed following euthanasia, and teased through a 100-µm nylon mesh (Becton Dickinson, Franklin Lakes, NJ). Cells were washed twice with MEM (Gibco, Life Technologies, Grand Island NY) and resuspended in AIM-V medium (Gibco, Life Technologies) supplemented with 100 µm non-essential amino acids (Gibco, Life Technologies), 1 mm sodium pyruvate (Gibco, Life Technologies), 10 mm HEPES (Gibco, Life Technologies), and 50 µm 2-mercaptoethanol (Sigma).

Cytokine-specific enzyme-linked immunospot (ELISPOT) assay was used as described in our previous study¹⁵ with the addition of some modifications. Briefly, cells were incubated in culture medium at 37° and 5% CO₂ for 20 hr in the presence or absence of CT (0·5 µg/ml) or HEL (20 µg/ml). As a positive control, cells were cultured with concanavalin A (Con A; 0·5 µg/ml). The cells were washed and resuspended in supplemented culture medium (see above) and transferred to nitrocellulose plates (Millipore Multiscreen-HA; Millipore, Bedford, MA) coated with purified anti-interleukin (IL)-4 or anti-interferon-γ (IFN-γ) as capture antibodies (PharMingen, San Diego, CA). Biotinylated antibodies specific to mouse IL-4 or IFN-γ (PharMingen) were used as detecting antibodies at a concentration of followed by the addition of streptavidin–alkaline phosphatase (Jackson Immunoresearch Laboratories Inc., West Grove, PA). All antibodies were used in a volume of 100 µl/well with a concentration of 2 µg/ml. The substrate was prepared from 5-bromo-4-chloro-3-indolyl phosphate/nitroblue tetrazolium (BCIP/NBT) tablets (Sigma) and added to the wells in 100 µl volumes to detect the amount of alkaline phosphatase bound to the plate. After 10–25 min the wells were washed in distilled water and subsequently air-dried. Spots representing individual cytokine-secreting cells (SC) were enumerated in each well using a stereo microscope. Values are expressed as the number of positive, stained spots per 1×10⁶ or 2×10⁶ cells.

Antigen-specific proliferative responses

Triplicate cultures with 2×10⁵ spleen cells/well (96-well plate, Nunc, Roskilde, Denmark) were assayed in the presence of 100 µg/ml of HEL or 0·5 µg/ml of Con A as a positive control. After three days in culture, 0·4 µCi [methyl-³H]thymidine (Amersham) was added for the last 8 hr of culture. Incorporation of [methyl-³H]thymidine was determined using a standard method for liquid scintillation counting. Proliferative responses were expressed as a stimulation index (SI; counts per minute in the presence of antigen/counts per minute in the absence of antigen).

Assessment of antigen-specific antibody responses

Antibodies specific for HEL in serum were determined by enzyme-linked immunosorbent assay (ELISA). Ninety-six well plates (Immulon 2; Dynatech Laboratories Inc., Alexandria, VA) were coated with HEL (2 µg/ml) or mouse immunoglobulin G (IgG) standards (Southern Biotechnology, Birmingham AL) in a carbonate/bicarbonate buffer (pH 9·6). The plates were incubated overnight at 4° and then washed four times in PBS containing 0·05% Tween (PBS-T). Four-fold dilutions of mouse sera were prepared in diluent (PBS-T containing 0·5% gelatin) and dispensed in 100 µl volumes. The plates were incubated for 2 hr and washed. Affinity-purified antimouse biotin conjugates (IgG, IgG1 or IgG2a; Caltag, Toronto, Ontario, Canada) were used as the detecting antibodies at a dilution of 1 : 10 000. After incubation for 1 hr and four subsequent washes streptavidin-alkaline phosphatase was added for 1 hr. Di(Tris) p-nitrophenyl phosphate (PNPP, Sigma) was used as the chromogenic substrate. The absorbance was read after 15–20 min at 405 nm (BIO-RAD, Richmond, CA). The amount of HEL-specific IgG1 and IgG2a in serum was calculated by extrapolation to standard curves for IgG1 and IgG2a.

Preparation of epidermal sheets and immunohistochemical staining

Animals were anaesthetized for eight hours by a subcutaneous injection of 20 µg of Atravet (acepromazine; Ayerst Laboratories, Montreal Quebec, Canada). Two hundred and fifty mg 12·5% methylcelullose gel (Aldrich) containing 1 µg of CT/60 mg of a gel was topically applied to both sides of the ear and left for 8–12 hr. Control animals received formulations with no antigen. Animals were euthanized and the ears were collected and split ventrally using fine forceps. Tissues were then incubated, epidermis side up, in 20 mm ethylenediaminetetraacetic acid (EDTA)/PBS solution for 1 hr at 37°. Epidermal sheets were gently peeled away from the dermis using fine forceps and then fixed in acetone for 10 min at room temperature. Sheets were blocked with 5% goat serum and 5% normal mouse serum for 30 min at room temperature. To block non-specific binding we used an avidin/biotin blocking step (Vector Laboratories, Peterborough, UK). Epidermal sheets were incubated at 4° overnight with anti-I-A^d/I-E^d, anti-CD11c, anti-ICAM-1 and the appropriate antibody isotype controls (all antibodies were purchased from PharMingen). Biotinylated antibodies were detected using ABC kits (Vector Laboratories) and substrates DAB (Vector Laboratories). For immunofluorescence staining of anti-major histocompatibility complex (MHC) class II, tissues were stained with anti-I-A^d/I-E^d antibody as above followed by a goat anti-rat–fluoroscein isothiocyanate (FITC) conjugate (Zymed Laboratories Inc. San Francisco CA). All steps were followed by thorough washes in PBS.

Scanning confocal laser microscopy images were obtained using a Bio-Rad MRC-600 lasersharp scanning microscope (Bio-Rad Microscience, Mississauga, Ontario, Canada), mounted on a Nikon FXA microscope. The laser lightsource used for excitation of FITC-labelled antibodies was an argon lamp (maximum emission lines at 488 and 514 nm). Optical thin sections of labelled tissue were collected using a 60× plan-apochromatic, 1.4 numerical aperture oil immersion objective lens (Nikon Corp., Chiyoda-ku, Tokyo). Images were collected through the xy plane at 1·0 µm intervals, and processed using a Northgate 480 desktop computer equipped with Bio-Rad software.

Statistical analyses

Differences in immune responses among vaccine groups were analysed using InStat™ statistical software. To assess differences between experimental groups we used one-way anova.

Results

Flt3L influences immune responses to antigens delivered through the skin

The effectiveness of transcutaneous immunization appears to be related to the specialized function of LC in antigen uptake and migration to the lymph node. We were interested to determine whether expansion of DC with the haemopoietic growth factor, Flt3L, resulted in modulation of immune responses to antigens delivered through the skin. BALB/c mice were treated with Flt3L over a period of 10 days and then immunized topically with 50 µg of HEL. Cytokine production by HEL-specific T cells from draining lymph nodes was assessed by ELISPOT assay. Figure 1 shows the frequency of IL-4 and IFN-γ secreting cells in the draining lymph nodes following immunization. Consistent with our previous results, transcutaneous immunization with HEL induced a predominant Th2 response⁴ with a significant increase in antigen-specific IL-4 secreting cells over IFN-γ secreting cells. There were no significant differences in the frequency of IL-4 secreting cells between the Flt3L-treated and untreated groups (Fig. 1).

Figure 1.

HEL-specific cytokine secreting cells (SC) following transcutaneous immunization. Flt3L-treated and control mice (PBS-treated) were immunized with 50 µg of HEL. Cells from the draining lymph nodes were isolated and stimulated in vitro with 20 µg/ml of HEL and the frequency of IL-4 and IFN-γ-SC was assessed by ELISPOT assay. Data represent the mean cytokine-SC±(SEM) of individual samples from seven mice.

Serum was analysed for the presence of HEL-specific IgG and IgG subclasses. After one immunization no detectable antibody responses were found (data not shown). However, 10 days after the second immunization, Flt3L-treated mice had significantly lower levels of anti-HEL IgG in serum when compared to nontreated group (Fig. 2a). The humoral immune responses were also evaluated by assessing the frequency of HEL-specific IgG secreting cells in draining lymph nodes (Fig. 2b). Although there were no statistical differences, the group that received the Flt3L treatment had lower frequency of IgG secreting cells (Fig. 2b). The anti-HEL IgG response was characterized by the preferential stimulation of IgG1 over IgG2a (Table 1).

Figure 2.

HEL-specific IgG response in mice immunized transcutaneously. Flt3L-treated and control mice were immunized twice with 50 µg of HEL and 10 days after the last immunization the levels of anti-HEL IgG in serum were assessed by ELISA (a) and the frequency of HEL-specific IgG SC in the draining lymph nodes was assessed by ELISPOT (b). Data represent the mean titre±(SEM) of seven mice. *Statistically significant (P < 0·05) against all other groups.

Table 1. HEL-specific IgG subclasses following transcutaneous immunization of Flt3L-treated mice.

	HEL-specific antibody (ng/ml)
Treatment	IgG1	IgG2a
Naive	10±7*	35±4
HEL	18000±7000	56±18
HEL + Flt3L	4000±2000	47±9

^*SEM.

Treatment with Flt3L enhances humoral immune responses induced by transcutaneous immunization with HEL in the presence of CT

As indicated in the previous section, treatment with Flt3L did not result in an enhancement of immune responses to antigen delivered through the skin, but rather inhibited the induction of antibody responses. Because CT has been shown to be a potent adjuvant for transcutaneous vaccines¹¹ we assessed whether topical administration of CT could influence the immune response to HEL in Flt3L-treated mice. Control mice were treated with PBS and immunized with either HEL (50 µg) alone or with HEL (50 µg) + CT (1 µg). Incorporation of CT to the vaccine formulation resulted in a significant increase in the IgG anti-HEL antibody titres in Flt3L-treated mice when compared to Flt3L-treated mice immunized with HEL alone (Fig. 3). This response was characterized by a significant increase in HEL-specific IgG1 (Table 2). Although it is clear that 1 µg of CT is suboptimal for exerting adjuvant effect when applied to control mice (PBS-treated), immunization with much higher doses of CT (50 µg) significantly enhanced the level of IgG anti-HEL in serum of PBS-treated control mice (data not shown), confirming the adjuvant effect of CT for transcutaneous immunization.¹¹

Figure 3.

HEL-specific IgG response following transcutaneous immunization with or without CT. Flt3L-treated and control mice were immunized twice with 50 µg of HEL alone or with CT. The level of anti-HEL IgG in serum was assessed by ELISA 10 days after the last immunization. Data represent the mean titre±(SEM) of seven mice. *Statistically significant (P < 0·05) against all other groups.

Table 2. Effect of CT on HEL-specific IgG subclasses following transcutaneous immunization of Flt3L-treated mice.

		HEL-specific antibody (ng/ml)
Treatment	Immunization	IgG1	IgG2a
PBS	HEL	11400±62*	54±7
	HEL + CT	27800±62	94±36
Flt3L	HEL	3300±8	38±7
	HEL + CT	82600±207	104±34

^*SEM.

Effect of Flt3L treatment on cellular immune responses induced by transcutaneous immunization with HEL and CT

To examine the effects of CT on the stimulation of cellular immune responses to HEL in mice treated with Flt3L, we determined the cytokine profile and the proliferative response of spleen and lymph node cells following antigen stimulation in vitro. Flt3L treatment significantly enhanced the adjuvant effect of CT for cellular immune responses (Figs 4 and 5). In vitro lymphocyte blastogenesis assay of spleen cells showed that Flt3L treated mice immunized with HEL (50 µg) + CT (1 µg) had significant proliferative responses to HEL as compared to mice immunized with HEL alone (Fig. 4). Interestingly, CT together with Flt3L treatment enhanced both the number of IL-4 (Fig. 5a) and IFN-γ (Fig. 5b) secreting cells. These findings show that Flt3L treatment influences the stimulation of cellular immune responses induced by transcutaneous immunization in the presence of CT. Moreover, immunization of Flt3L treated mice with CT exhibited a more balanced T helper response as both IL-4 and IFN-γ secreting cells were stimulated as compared to the predominant Th2 response found in Flt3L-treated mice that received HEL alone.

Figure 4.

HEL-specific proliferative responses following transcutaneous immunization. Mice were immunized twice by the transcutaneous route with 50 µg of HEL. Ten days after the second immunization, cells from the spleen were collected and restimulated with 100 µg/ml of HEL. HEL-specific proliferative responses were assessed by [³H]thymidine incorporation and the results represent the mean±(SEM) stimulation index (SI) of seven mice.

Figure 5.

HEL-specific cytokine secreting cells (SC) following two transcutaneous immunizations. Flt3L-treated and control mice were immunized with 50 µg of HEL alone or with CT. Cells from the draining lymph nodes were isolated and stimulated in vitro with 20 µg/ml of HEL. The frequency of IL-4 (a) and IFN-γ-SC (b) was assessed by ELISPOT assay. Data represent the mean cytokine-SC (SEM) of individual samples from seven mice.

CT activates LC in vivo

To examine whether the adjuvant effect of CT could be associated with stimulation of inflammatory signals and activation of LC in the skin, we assessed the expression of cell surface molecules by immunohistochemical staining of CT-treated skin. CT (1 µg CT/60 mg of formulation) was applied to the ears of anesthetized mice, control mice received formulation alone. After 8–12 hr epidermal sheets were obtained and stained with anti-CD11c and anti-ICAM-1 antibodies (PharMingen). Staining of epidermal sheets revealed that topical application of CT results in increased staining intensity for both CD11c and ICAM-1 compared to control epidermal sheets (Fig. 6).

Figure 6.

Detection of CD11c and ICAM-1 expression by immunohistochemical staining of CT-treated skin. One µg of CT in a methylcellulose gel was applied to the ears of anesthetized mice. Control mice received gel alone; 12 hr later epidermal sheets were obtained and stained with anti-CD11c and anti-ICAM-1.

Evaluation of MHC class II⁺ cells in epidermal sheets was assessed by fluorescent antibody staining and imaging by scanning confocal laser microscopy. Staining revealed that CT induced pronounced morphological changes in LC (Fig. 7) and increased MHC class II expression (Fig. 8). CT induced LC to become rounder in appearance with the loss of the dendrites. This change in morphology suggests LC are possible preparing for migration from the epidermis. CT did not induce MHC class II expression in keratinocytes and there was no evidence of cell infiltration (data not shown). To compare levels of MHC class II staining of LC between the CT-treated and the control tissues, the images were captured using the same settings and were not digitally compensated to enhance brightness or contrast of the observed fluorescence (Fig. 8). Treatment with Flt3L did not influence the changes induced by CT (Fig. 7) or the number of MHC class II⁺ cells in the epidermis (Fig. 9).

Figure 7.

Effect of topical application of CT on LC morphology. Following Flt3L treatment, 1 µg of CT in a methylcelullose gel was applied to the ears of mice. Control mice received gel alone; 8 hr later epidermal sheets were obtained and stained with anti-MHC class II followed by antirat FITC-labeled antibody. Images were obtained using a scanning confocal laser microscope (×60).

Figure 8.

Effect of topical application of CT on MHC class II expression in LC. The ears of mice were treated with 1 µg of CT in a methylcelullose gel. Control mice received gel alone; 8 hr later epidermal sheets were obtained and stained with anti-MHC class II followed by anti-rat FITC-labelled antibody. Images were obtained using a scanning confocal laser microscope (×60) using the same settings; images were not subjected to any further digital enhancement.

Figure 9.

MHC class II positive DC in epidermal sheets. Mice treated with Flt3L or PBS received 1 µg of CT in a methylcelullose gel. Control mice received gel alone; 12 hr later epidermal sheets were obtained and stained with rat anti-MHC class II followed by anti-rat HRP-labelled antibody.

Discussion

In this study, we assessed whether expansion of DC by intraperitoneal administration of Flt3L could influence the immune response induced by transcutaneous immunization. Our results demonstrated that increasing the number of DC by administration of Flt3L was not sufficient to enhance immune responses to antigen delivered through the skin. In fact, our findings indicate that mice treated with Flt3L developed lower antigen-specific cellular and humoral responses than PBS-treated mice (Figs 4 and 5). These observations extend the findings of Viney et al.⁸ by showing that similar to the gut, transcutaneous immunization of Flt3L-treated mice in the absence of adjuvants does not enhance but rather suppresses the induction of immune responses to antigen delivered to the epidermis. These data support the concept that the skin and the mucosa share similar regulatory mechanisms where DC play a central role on the induction of active immunity while controlling and limiting harmful responses.^8,10

The results of this study confirmed our previous observations that the immune responses induced by transcutaneous immunization are characterized by the induction of a Th2 response.⁴ Although CT has been considered a potent mediator of Th2 responses^12,16,17 in the present study, we found that CT promoted the stimulation of HEL-specific IFN-γ secreting cells in Flt3L-treated mice and consequently revealed a more balanced T helper response (Fig. 5). It is important to note that the significant increase in IFN-γ secretion was only observed in Flt3L-treated mice that were immunized in the presence of CT (Fig. 5). These results suggest that activation and migration of large numbers of DC from the skin to the lymph nodes may influence the regulation of T-cells by increasing the antigen load and the dynamics of T-cell priming, as proposed by Lanzavecchia and Sallusto.¹⁸ Our previous study demonstrated that similar to the effects of Flt3L, incorporation of IL-12 to the vaccine formulation resulted in a decreased humoral immune response to antigens delivered by the transcutaneous route.⁴ Together this information suggests that administration of antigen through the skin can result in either induction of active immunity or suppression of immune responses.

In the present study we did not find evidence that Flt3L induces significant accumulation of LC as assessed by quantification of MHC class II⁺ cells in epidermal sheets. These observations are in contrast with previously reported studies where a slight but significant increase in DC, labelled with the NLDC 145 antibody, was found in the epidermis of Flt3L treated mice.⁹ However, these apparent discrepancies may be explained by the fact that the antibodies used in the two studies label different molecules, and therefore, it is possible that they identify different subpopulations of LC. Furthermore, the effect of Flt3L treatment on the modulation of immune responses induced by transcutaneous immunization may be regulated by the expansion of DC in the draining lymph nodes. Indeed, Flt3L-treated mice contained a significant proportion of CD11c⁺ cells when compared to PBS-treated mice (data not shown). These results are consistent with the observations that Flt3L increases the numbers of mature DC in vivo.^7,8

The morphological changes induced by CT on LC suggest that activation of these cells results in increased mobilization from the epidermis. Indeed, CT induces increased migration of cells to the lymph node after epicutaneous FITC painting (unpublished observations). These observations are similar to the in vivo effects that immunostimulatory oligonucleotides¹⁹ and haptens²⁰ have on LC morphology and migration. The regulation of LC activation and migration from the epidermis involves a complex series of events. Cytokines, in particular IL-1 and tumour necrosis factor-α (TNF-α), have been shown to induce mobilization of LC by regulating the expression of chemokine receptors and adhesion molecules (reviewed in ²¹). It is possible that the immunomodulatory effects of CT in the skin are mediated by a direct stimulation of resident cells (keratinocytes and LC) to secrete cytokines such as IL-1 and TNF-α. Interestingly, numerous reports have shown that CT induces the secretion of IL-1¹³ and that the ability of CT to enhance mucosal immune responses in Flt3L-treated mice is mediated in part by IL-1.¹⁰ TNF-α has been demonstrated to induce rearrangement of the cytoskeletal network in DC resulting in increased cell motility in vitro.²²In vivo, administration of TNF-α induces changes in LC morphology similar to those induced by CT and enhances migration of DC from the skin to the lymph node.²³ Furthermore, TNF-α is known to regulate ICAM-1 expression by keratinocytes²⁴ and it is known that ICAM-1 plays an important role in mediating the migration of LC to the lymph node.²⁵ Future studies will focus on elucidating the role of these and other cytokines in mediating the adjuvant activity of CT by the transcutaneous route.

Our observations support the concept that the state of activation of DC in the skin is central to the regulation of immune responses. Therefore, elucidation of the mechanisms that influence DC recruitment, maintenance and migration from the skin to lymph nodes, is critical to the design of effective transcutaneous vaccination strategies.

Abbreviations

HEL

hen egg lysozyme

CT

cholera toxin

DC

dendritic cells

LC

Langerhans' cells

SC

secreting cells

PNPP

Di(Tris) p-nitrophenyl phosphate

SI

stimulation index