Protection Against Cutaneous Leishmaniasis Resulting from Bites of Uninfected Sand Flies

See allHide authors and affiliations

Science  17 Nov 2000:
Vol. 290, Issue 5495, pp. 1351-1354
DOI: 10.1126/science.290.5495.1351


Despite the fact that Leishmania are transmitted exclusively by sand flies, none of the experimental models of leishmaniasis have established infection via sand fly bites. Here we describe a reproducible murine model of Leishmania majorinfection transmitted by Phlebotomus papatasi. Prior exposure of mice to bites of uninfected sand flies conferred powerful protection against Leishmania major that was associated with a strong delayed-type hypersensitivity response and with interferon-γ production at the site of parasite delivery. These results have important implications for the epidemiology of cutaneous leishmaniasis and suggest a vaccination strategy against this and possibly other vector-borne diseases.

The diseases transmitted by arthropod vectors afflict millions of people, particularly in developing countries. Bloodsucking arthropods may be more than just delivery systems for the pathogens they carry, insofar as components in their saliva have been shown to modify the outcome of infection (1,2). Leishmaniasis is a vector-borne disease transmitted exclusively by sand fly bites. Reports of successful laboratory transmission of Leishmania spp. by sand fly bites are few (3–8) and have not addressed the host response to infective bites or considered the effects of prior exposure to uninfected sand fly bites on the outcome of infection. Using the murine ear model, we were able to transmit Leishmania major reproducibly to BALB/c and to C57BL/6 mice (9) by the bite of its natural vector, Phlebotomus papatasi (10). The respective healing and nonhealing phenotypes of C57BL/6 and BALB/c mice, established using high-dose needle inocula, were maintained in fly-transmitted infections. In BALB/c mice, nodular lesions increased steadily in diameter and thickness (Fig. 1, A and B). In C57BL/6 mice, the lesions increased in size up to day 70 when they began to resolve, with complete healing by day 120 (Fig. 1, C and D). Most of the lesions developed small focalized areas of ulceration prior to healing.

Figure 1

The course of lesion development after transmission of L. major by bite of P. papatasi. Transmission by bite to ears of naı̈ve (solid circles) and pre-exposed (solid squares) BALB/c mice (A and B) or of C57BL/6 mice (C and D) or to ears of naı̈ve wild-type (open circles) and naı̈ve IL-4–deficient (open squares) C57BL/6 mice (E and F) is shown. Mice were exposed to 10 infected flies on each ear. Ears were examined for the appearance of lesions at weekly intervals and were measured with a vernier caliper. Because multiple lesions tended to coalesce as they progressed, the values shown at each time point are the sum of the lesion's diameter (left panels) and the largest ear thickness (right panels), mean + 1 SE, 12 to 18 ears per group. Asterisks designate time points with values significantly different between naı̈ve and pre-exposed groups (P < 0.05 by the two-tailed Student'st test) for each time point.

In endemic regions, many individuals are exposed to the bites ofLeishmania-free phlebotomines before being bitten by an infected sand fly. For both BALB/c and C57BL/6 mice, prior exposure toP. papatasi bites (11) resulted in a striking reduction in the severity of the dermal lesions. For BALB/c mice, the difference between naı̈ve and pre-exposed mice was significant from days 31 and 23 onward for lesion diameter and thickness, respectively (P < 0.05) (Fig. 1, A and B). For C57BL/6 mice (Fig. 1, C and D), the pre-exposed animals showed a delay in the appearance of lesions and a dramatic reduction in the peak lesion size that developed before healing. The attenuation in the diameter and/or thickness of lesions was significant during days 28 to 56 (P < 0.05). A second transmission experiment was undertaken in naı̈ve and pre-exposed C57BL/6 mice to compare the parasite loads within the inoculation site. Two weeks after transmission by bite, the pre-exposed mice had over a 1000-fold reduction in the mean number of amastigotes present per infected ear (P < 0.001) (Fig. 2A), indicating that the protection conferred by uninfected fly bites was expressed relatively early after parasite delivery. As a relative measure of tissue parasite burden and as a biologically meaningful estimate of reservoir potential, a third set of C57BL/6 mice was used to investigate the ability of infected ears to transmit infections back to uninfected flies. When exposed to flies at the peak of lesion development, naı̈ve mice provided a significantly better source of parasites than pre-exposed mice, with 11 of 14 ears (79%) able to transmit infections, as compared with 5 of 18 ears (28%) in the pre-exposed group (Fig. 2B). The percentage of blood-fed flies that picked up parasites averaged 27% per ear for the naı̈ve mice and 10% for the pre-exposed mice (P < 0.05).

Figure 2

Comparison of parasite loads and reservoir potential in naı̈ve and pre-exposed C57BL/6 mice. (A) The number of viable amastigotes per ear titrated from complete tissue homogenates, obtained 14 days after transmission by bite. The number of parasites in exposed ears was determined by dilution to extinction of a homogenate of the entire ear dermis, prepared as described previously (16). Statistical significance was determined by the two-tailed Student's t test. (B) The transmissibility of L. major from infected ears to uninfected sand flies 48 days after transmission by bite. About 20 emergent female P. papatasi were allowed to feed for 1 to 2 hours on ears 7 weeks after fly-transmitted infection. Blood-fed females from each vial were separated and maintained in individual pots lined with plaster of Paris and were provided with a 50% sucrose and 5% albumin solution and water. Midguts were dissected 48 to 72 hours later and examined microscopically for the presence of promastigotes. Values represent the percent of blood-fed flies per ear that were positive for promastigotes. Statistical significance was determined by the one-tailed Student's t test.

Prior sensitization to vector saliva could have impaired the feeding behavior of infected flies, compromising their ability to deliver parasites to the site. It is well documented that mammalian hosts develop acquired resistance to tick infestation as a result of previous exposure to tick bites (12, 13), which in turn confers host resistance to tick-transmitted pathogens such as Borreliaand Pasteurella spp. (14). However, for each of the transmission experiments described in the current studies, there was no significant difference in the number of blood-engorged flies recovered from the ears of naı̈ve versus pre-exposed mice [Web table 1 (15)].

A mechanism of protection is suggested by a recent study involving needle inoculations of salivary gland sonicate (SGS) (16), in which the exacerbative effect of SGS onL. major infection, originally described by Titus and Ribeiro (17), was absent in mice that had been previously injected with SGS. The results shown here suggest that infective sand fly bites produce a similar outcome to the intradermal needle injection of a low dose of L. major metacyclics alone that is quite different from that produced by the coinoculation of parasites and SGS. The fly-transmitted infections in C57BL/6 mice in every instance resolved over time, similar to the injection of parasite alone (16, 18), whereas low-dose challenge in the presence of SGS produced nonhealing dermal lesions in C57BL/6 mice (16). In addition, the bites of infected P. papatasi did not elicit the early interleukin-4 (IL-4) response in the skin that was observed after the intradermal inoculation of SGS andL. major, being more comparable to the tissue response observed after the inoculation of parasites alone (16). These differences prompted us to investigate fly-transmitted infections in C57BL/6 IL-4–deficient mice (9), which along with mice treated with antibody to IL-4 have been shown to be refractory to the exacerbative effects of SGS (16, 19). The severity of the dermal lesions (Fig. 1, E and F) and the parasite loads in the site were not reduced in the IL-4–deficient mice as compared with wild-type mice, indicating that IL-4 is not required for, and has little effect on, the evolution ofL. major infections in C57BL/6 mice after transmission by bite. The molecule(s) responsible for eliciting the early IL-4 response after inoculation of P. papatasi SGS may not be present in salivary secretions or may be deposited in the skin in much lower quantities during salivation. Similar discrepancies have been reported for black flies and mosquitoes, where salivary gland extracts elicited immune responses not observed after natural feeding (20,21).

An adaptive immune response to sand fly saliva might modify the tissue environment into which the parasites are introduced in a manner that is directly harmful to the parasite. Phlebotomus papatasi bites produce a long-lasting delayed-type hypersensitivity (DTH) response in humans. Histologic analysis (22) of the ear dermis at 0, 3, 24, 48, and 72 hours after infective bite revealed a small, transient inflammatory reaction in naı̈ve mice. In contrast, a large cellular infiltrate and a 200 to 300% increase in ear thickness that peaked at 24 hours and was sustained for up to 72 hours was seen in pre-exposed mice (Fig. 3A), which is indicative of a DTH response. In naı̈ve mice, the leukocytes recovered from the entire ear dermis (23) 24 hours after fly bites were largely confined to neutrophils, contrasting with pre-exposed mice that showed a two- to fivefold increase in infiltrating neutrophils, eosinophils, macrophages, dendritic cells, and lymphocytes. This response was diminished by intraperitoneal injection of antibodies to CD4 24 hours before sand fly exposure [Web fig. 1 (15)]. The epidermal compartment in the mouse contains three major cell populations: keratinocytes, dendritic epidermal T cells (DETCs), and Langerhans cells, each of which can be activated to secrete a distinct set of cytokines (24). Six hours after exposure to infected flies, the epidermal cells recovered from naı̈ve mice (25) consistently showed a strong up-regulation of IL-2 and IL-3 production (Fig. 3B). Smaller increases above steady-state frequencies were observed in cells producing interferon-γ (IFN-γ). Few cells stained for IL-4. The acute response of epidermal cells from pre-exposed mice was similar to that of naı̈ve mice, with the clear exception of IFN-γ– and IL-12–producing cells (Fig. 3B). The mean number of epidermal cells producing IFN-γ and IL-12 per ear in pre-exposed mice was 1.02 × 105 and 3.2 × 104, which are 9- and 15-fold greater than in naı̈ve mice, respectively. Langerhans cells and keratinocytes are important sources of IL-12 in the skin (24, 26), and each of the major epidermal populations has been shown to produce IFN-γ (24, 27, 28). The early and increased production of IFN-γ and IL-12 in the inoculation site may activate infected macrophages for killing during the initial establishment of infection and may also promote a more rapid and polarizedLeishmania-specific T helper cell type 1 response.

Figure 3

Characterization of the dermal and epidermal responses in naı̈ve C57BL/6 mice and in mice pre-exposed to sand fly bites. (A) Whole ear sections (22) of naı̈ve and pre-exposed mice after infected sand fly bites. (B) Analysis of the epidermal cytokine response (25) 6 hours after infected sand fly bites. The percent of epidermal cells with FL2 signals greater than that of the isotype control for a particular cytokine is shown. The profiles shown are representative of one of four separate experiments.

The strong DTH response elicited at the inoculation site in mice previously exposed to sand fly bites conferred immunity againstL. major infection that was as potent as any achieved by the combination of parasite antigens and adjuvants used to date. These findings imply that the exposure history of individuals to the bites of uninfected sand flies influences the incidence and severity of cutaneous leishmaniasis and that for this and possibly other vector-borne diseases, salivary antigens might be effective components of vaccines directed against transmitted pathogens.

  • * To whom correspondence should be addressed. E-mail: dsacks{at}


Stay Connected to Science

Navigate This Article