Humanized mice in triple deficient NOD/scid
(NSG) mice develop all cellular components of the human immune system such as T, B, and NK cells as well as myeloid cells and are able to mount an adaptive and innate immune response to antigens 
. Based on this innovative concept to create a human immune system in a murine environment, scientific issues regarding virus-caused infectious diseases 
, therapy and vaccine studies 
, graft versus host disease (GvHD) 
, and tumor diseases 
were already studied.
To characterize the interactions between human immune cells and obligatory intracellular protozoan parasites under in vivo
conditions we infected humanized mice with L. major
parasites. In general these parasites induce cutaneous leishmaniasis. However, exceptions of this classic form with mild visceral tendency have been reported in patients before 
. This capacity for visceralization was also described in L. tropica
infected patients 
The L. major
strain (MHOM/IL/81/FE/BNI] used in our experiments was isolated from a LCL patient 
. Of note, under experimental conditions this L. major
strain can result in visceral manifestations in resistant C57BL/6 and susceptible BALB/c mice as well 
. Thus, it cannot be generally excluded hat L. major
parasites might show the tendency of mild visceralization in parallel to the cutaneous manifestation.
Here we questioned for the first time whether L. major parasites can replicate within humanized mice. Furthermore, we tested this novel experimental model - sharing murine and human components – for its capacity to characterize human adaptive and innate immune response to L. major and the efficiency of treatments like such as hexadecylphosphocholine Miltefosine.
Initial experiments revealed that cutaneous inoculation of L. major
parasites resulted in an inflammation at the site of infection. Additionally the severity of disease correlates with the dose of parasites injected. Thus, humanized mice develop a local inflammation caused by the protozoan parasites. Based on the reconstitution of NSG mice with naïve human leucocytes it is most likely that human cells are involved in the parasite-induced inflammatory response. Thus, we further investigated the phenotype and origin of potential host cells at the site of infection of humanized mice. It is important to mention that humanized mice still possess mouse myeloid cells such as macrophages and granulocytes. And indeed, our data revealed that Leishmania
-antigens can be detected in human as well as murine myeloid cells indicating that mouse myeloid cells represent target cells for Leishmania
in humanized mice. As suggested by several findings infected mouse myeloid cells might function as a niche for Leishmania
allowing immune evasion. First, the human adaptive immune system is not able to control or stimulate mouse immune cells due to mismatched MHC expression and the species-specificity of human IFN-γ 
. Second, most of those myeloid mouse cells show functional weaknesses as for instance insufficient differentiation of fully functional DC from NSG bone marrow cells 
. And third, it has been shown that in NOD-scid
mice the remaining mouse macrophages show an impaired functionality 
However, we detected a 10-fold increase in the expression of murine iNOS mRNA in the footpad tissue of humanized mice three weeks after infection with L. major
whereas no increase of iNOS mRNA was found in the spleen. This clearly indicates that mouse macrophages somehow get activated during infection, although the upregulation of iNOS mRNA in humanized mice is ~factor 100 lower than that seen in infected BALB/c mice. In accordance, no iNOS mRNA increase was visible in the spleen of BALB/c mice upon infection [unpublished data, U. Schleicher and C. Bogdan, Erlangen). With regard to parasite control this lack of iNOS increase in the spleen is of minor importance as rather NADPH-oxidase-dependent than iNOS-dependent mechanism are required for the resolution of the parasites in this organ 
. In contrast to mouse iNOS mRNA no induction of human iNOS mRNA was detectable in infected humanized mice. Since we demonstrated that human macrophages get infected in humanized mice this result rather supports the idea that iNOS expression by macrophages or other cells in response to infections does not occur in the human immune system 
. That this statement might not hold true in human cutaneous leishmaniasis is supported by results demonstrating iNOS mRNA and protein expression in skin lesions of CL patients 
. Together, it can be therefore speculated that the atypical visceralisation of L. major
parasites within humanized mice might be at least partially due to the insufficient production of murine derived leishmanicidal molecules and the lack of human iNOS induction.
Considering that infected macrophages need to be activated by an efficient Th1-cell response 
the human-derived T cell response was further analyzed. Inflammatory human T cells could be detected at the site of infection. Most of them represent the phenotype of a memory T cells and the majority of these cells belong to the CD4+
T helper cell population, known to play a major role in the activation of infected macrophages and subsequent killing of the intracellular Leishmania
. The fact that CD4+
as well as CD8+
T cells were efficiently primed in vivo
indicates that human antigen presenting cells are functional and that the induced human T cell response is Leishmania
-specific. Based on the fact that splenocytes of infected humanized mice respond with a pronounced human IFN-γ production after restimulation with soluble Leishmania
antigen (SLA) we conclude that a human Th1-like immune response was induced.
From the experimental model of leishmaniasis it is known that resistant C57BL/6 mice develop a delayed hypersensitivity (DTH] response after s.c. injection of SLA whereas susceptible BALB/c mice do not show any signs of a DTH response 
. This lack of DTH reaction was also detectable in infected humanized mice and might represent the dysfunction of the induced Th1-type effector cells to get recruited to the site of antigen inoculation.
Based on the fact that humanized mice do not show signs of a DTH reaction we can conclude that Th1 cells are efficiently primed within the spleen but failed to migrate to the site of the infection. This might also explain the relative low parasite density within the spleen compared to the parasite load detectable at the site of infection.
To investigate the versatility of our model for future drug studies, we treated infected humanized mice and BALB/c mice with Miltefosine. In BALB/c mice the application of Miltefosine induced a reduction of the parasite load predominantly in visceral organs such as liver and spleen in the absence of liver damage and weight loss. In contrast humanized mice respond different to the Miltefosine therapy. They show signs of liver damage and significant weight loss after treamtent. Thus the Miltefosine-derived side effect must be the consequence of the presence of human immune cells. This is in agreement with published data 
demonstrating signs of liver toxicity in patients after Miltefosine therapy. The observed Miltefosine-based side in humanized nice might be due to the fact that Miltefosine can activate human immune cells to release pro inflammatory cytokines 
that in turn cause the observed liver damage. Thus, humanized might be a useful tool to test possible side effects which are not detectable in classical animal models such as BALB/c, before starting clinical trials.
The leishmanicidal effect of Miltefosine in humanized mice is visible but not comparable to control BALB/c mice. This might be due to the lack of mouse IFN-γ which in turn can not activate infected mouse macrophages.
In conclusion, we have generated a novel Leishmania major infection model using humanized mice which develop a systemic non-self-healing disease. Humanized mice might therefore be especially interesting for studies on visceral leishmaniases offering an additional challenge for new treatment strategies: the elimination of resting (“hiding”) Leishmania parasites within the mouse immune cells. In addition, side effects induced by activated human immune cells (cytokine release) which do not occur in wild type mice can now be illuminated in a small animal model which might help to predict possible side effects before the drugs get tested in clinical trials.