The authors established the first 'double humanised' lupus mouse model (DhuSLE) combining a human immune system and faecal microbiota from SLE patients, which induced proteinuria and elevated human IgM when combined with pristane, though limitations including insufficient antibody class switching remain.
Key Findings
Results
FMT in general transiently suppressed the development of human T cells in NSG-hu mice regardless of donor type.
NSG immunodeficient mice were engrafted with human CD34+ haematopoietic stem and progenitor cells to create NSG-hu mice
Faecal microbiota transplantation (FMT) from both SLE and non-SLE donors caused transient suppression of human T cell development
The suppression was described as transient, suggesting T cell reconstitution recovered over time
This effect was observed as a general response to FMT rather than being specific to SLE-derived microbiota
Results
SLE-FMT, but not FMT from non-SLE donors, promoted superficial skin lesions in NSG-hu mice.
Skin lesions were only observed following transplantation of faecal microbiota from patients with SLE
Non-SLE donor FMT did not produce this phenotype, indicating disease-specificity of the SLE gut microbiota effect
The skin lesions were described as 'superficial'
This finding suggests SLE gut microbiota carries disease-promoting properties independent of a fully humanised immune context
Results
The combination of SLE-FMT and pristane in NSG-hu mice (DhuSLE-P model) induced proteinuria.
Pristane is a chemical adjuvant known to induce lupus-like disease in mice
Neither SLE-FMT alone nor pristane alone in NSG-hu mice was sufficient to drive this phenotype based on the study design requiring combination
The authors noted that the proteinuria observed in mice 'did not reflect that of the microbiota donors,' indicating a discordance between mouse phenotype and human donor disease status
Proteinuria is a key clinical sign of lupus nephritis, making this a relevant disease manifestation
Results
DhuSLE-P mice exhibited higher levels of human IgM in circulation compared to NSG-hu mice, positively correlated with splenic plasma cell frequency.
Elevated circulating human IgM was a distinguishing feature of DhuSLE-P mice relative to control NSG-hu mice
The level of human IgM was 'positively correlated with the frequency of plasma cells in the spleen'
Nuclear BCL6 was 'minimally detected' in splenic sections of DhuSLE-P mice, indicating limited germinal center activity
CD138 expression was evident in splenic sections, confirming plasma cell presence
These findings together suggest most plasma cells generated were not class-switched and predominantly produced IgM rather than IgG
Results
Some human IgG was detected in the kidneys of DhuSLE-P mice with a trend towards increased total IgG in serum.
Human IgG deposition was detected in kidney tissue of DhuSLE-P mice
There was a 'trend towards increased total IgG in the serum,' though this did not reach statistical significance
IgG kidney deposition is consistent with immune complex-mediated nephritis seen in human SLE
The limited IgG production was identified as a model limitation, with the authors suggesting methods to promote antibody class switching could improve the model
Results
Gut microbiota composition differed between DhuSLE-P mice and NSG-hu mice due to SLE-FMT but not pristane injection.
Faecal microbiota analysis revealed distinct compositional differences between DhuSLE-P and NSG-hu mice
The microbiota differences were attributable specifically to SLE-FMT rather than to pristane treatment
Pristane injection alone did not alter gut microbiota composition
This result confirms successful engraftment of SLE-derived microbiota and separates the contributions of microbiota versus chemical adjuvant to the model phenotype
Conclusions
The DhuSLE model represents the first humanised lupus mouse model combining both a human immune system and gut microbiota from SLE patients.
Prior humanised lupus models lacked the dual humanisation of both immune system and gut microbiota
The model was built on NSG immunodeficient mice engrafted with human CD34+ haematopoietic stem and progenitor cells
SLE-FMT was performed to introduce patient-derived gut microbiota
The authors identify the model as a platform for 'elucidating mechanisms and/or evaluating SLE treatments' with further development
Key limitations include insufficient antibody class switching and discordance between mouse phenotype and donor disease status
What This Means
This research describes the creation of a new type of mouse model for lupus (SLE), a chronic autoimmune disease. The researchers engineered mice to have a human immune system by transplanting human blood stem cells, and then also gave the mice gut bacteria (microbiome) from lupus patients through a procedure called fecal microbiota transplantation (FMT). This 'double humanised' approach is novel because previous lupus mouse models have not combined both of these human elements. When the mice also received a chemical called pristane that promotes autoimmunity, they developed protein in their urine (proteinuria), a hallmark sign of kidney damage seen in lupus patients, and showed elevated levels of a human antibody called IgM. Notably, only the lupus patient microbiota — not microbiota from healthy donors — caused skin lesions in the mice, pointing to a specific disease-promoting role for the SLE gut microbiome.
The study also found that while the mice produced some human immune proteins and showed lupus-like features, the antibody responses were incomplete — the immune cells largely produced IgM rather than the class-switched IgG antibodies more typically associated with lupus tissue damage in humans. The gut bacteria successfully engrafted from the lupus donors into the mice, and these changes in the microbiome were driven by the transplant itself rather than the chemical adjuvant pristane. Interestingly, the proteinuria seen in the mice did not mirror the kidney disease status of the human donors whose gut bacteria were used, suggesting the model does not yet fully replicate the human disease relationship between microbiome and clinical outcomes.
This research suggests that combining a humanised immune system with lupus patient gut microbiota in mice is feasible and can partially reproduce lupus-like features, opening a path for studying how gut bacteria interact with the immune system in lupus. The model could eventually be used to test new treatments. However, the authors acknowledge limitations, particularly the need for better antibody class switching, and note that further refinements are needed before the model can reliably mirror human lupus disease.
Xu T, Lu R, Oakland D, Estaleen R, Rawlings A, Montano H, et al.. (2026). Double humanised lupus mouse model with human immune system and faecal microbiota from patients with SLE.. Lupus science & medicine. https://doi.org/10.1136/lupus-2026-001982