Preprint Club
A cross-institutional Journal Club Initiative

The authors find that:

Keywords

• Tuberculosis in mice, non-human primates and humans

• Interleukin-1 signaling and antagonism

• Immunosuppressive myeloid cells

Main Findings

Tuberculosis is a severe lung disease caused by infection with Mycobacterium tuberculosis (Mtb). In humans, progressive infection with Mtb is characterized by exacerbated inflammatory responses and myeloid cell infiltration with a high abundance of neutrophils. The pathology of human tuberculosis is not fully recapitulated by standard mouse models (C57BL/6, BALB/c) and requires the use of genetically modified mice (f.e. Sp140-/-, Nos2-/-, Acod1-/-), which are more susceptible to Mtb infection. The present study investigates a possible common feature of Mtb susceptibility in these different mouse strains and its reproducibility across other species using single cell RNA-sequencing (scRNAseq) datasets from mice, non-human primates and humans, and validates their results with several mouse experiments.

• Differentially expressed genes in myeloid cells from the three susceptible mouse models (Sp140-/-, Nos2-/-, Acod1-/-) infected with Mtb can more effectively discriminate human patients with active tuberculosis from healthy subjects than genes from Mtb-infected C57BL/6 mice.

• While a deficient interferon (IFN) type I signaling reduces bacterial burdens in Sp140-/- mice, this is not the case in other susceptible mouse models (Nos2-/-, Acod1-/-), indicating that type I IFN signaling does not play a common causative role for Mtb disease progression in susceptible mice.

• Increased recruitment of myeloid cells with a predominance of Spp1+ macrophages is common in all susceptible mouse strains but is largely absent in Mtb-restrictive C57BL/6 mice.

• Spp1+ macrophages express genes for immunosuppressive molecules such as IL-1rn (coding for IL-1Ra) and are also found in scRNAseq from Mtb-infected non-human primate granulomas. In the lungs of infected Sp140-/-mice, Spp1+ macrophages are preferentially detected at Mtb disease sites.

• IL-1rn expression is found in active human Mtb signatures together with expression of other immunosuppressive molecules.

• Global IL-1rn deletion in susceptible mouse strains decreases bacterial burden to wild-type levels. Specific deletion of IL-1rn in myeloid cells of Sp140-/- also decreases bacterial load, underlining a myeloid specific role of this immunosuppressive molecule.

• IL-1 mediates protection against Mtb infection by signaling in hematopoietic and non-hematopoietic bystander cells and not in myeloid cells which express little to no IL-1R

• IL-1 induced ligands produced by bystander cells in Mtb infection can act on myeloid cells. Most likely identified ligands are IFNγ, TNF, IL-6, CCL2 and IL-17.

• IL-1 mediated increase in IL-17/22 expression is redundant for IL-1 driven protection in Mtb infection

Limitations

• While the study uses three relevant mouse models susceptible to Mtb-infection, it might be helpful for readers to describe why Sp140-/-, Nos2-/-, Acod1-/- and not other susceptible mouse models (f.e. Atg5-/-, Irg1-/-, Card9-/-, see Ravesloot-Chávez MM, et al. 2021, Annu. Rev. Immunol., https://doi.org/10.1146/annurev-immunol-093019-010426) were used in this study, as well as interesting to discuss if the depleted genes Nos2 and Acod1themselves have any described effects on IL-1 signaling.

• While the figures are mostly very neat and structured, it is not fully clear why controls for Sp140-/- are labelled as B6 mice (Figure 1E), but other controls are “littermate control” (f.e. Figure 1E-G). Does this mean that the background is not exactly B6 background? It is also not fully clear from the figure or text why there is no Acod1-/-Ifnar1+/+ in Figure 1F. In Figure 3A, there is a significant difference marked with a star between AM but not IM numbers, however it seems on the figure as if it is the opposite.

• The Spp1+ macrophages are a very interesting population, which could be described more extensively in this study, also in regard as potential therapeutic targets in tuberculosis. Based on the CITEseq data in this study, these cells are monocyte derived and highly infected with Mtb (compared to B6 Spp1+ IM, Trem2+ IM, well as to infected monocytes), while monocytes themselves are very lowly infected with Mtb (Suppl. Fig 4D). Are these macrophages truly monocyte derived? What makes these macrophages differentiate into Spp1+ in comparison toTrem2+ activation state, which are more infected in C57BL/6 (Figure 2D)? Is the bacterial copy number higher in these cells compared to other infected cells? Is the induction of immunosuppressive molecules such as IL-1rnpossibly mediated by the bacteria themselves?

• While it is evident from the study that Spp1+ macrophages highly express IL-1rn in Mtb-infection on the scRNAseq data, the claim that Spp1+ macrophages are the “dominant macrophage source of IL-1Ra in mice” is not fully proven in this study. It would add great value to the study if Il-1Ra production by Spp1+ macrophages or generally Mtb+ macrophages could be measured, maybe by sorting these cells and measuring Il-1Ra production and other cytokines compared to other cell types in vitro (f.e. the Trem2+ macrophage population), or by infecting cultured tissue resident or bone marrow derived macrophages from naïve animals in vitro with Mtb and measuring the produced cytokines in the supernatant.

• Generally, while the exploration of scRNAseq datasets is carefully and extensively done, cytokine measurements in plasma or lung supernatants from infected animals would be a great addition to this study, such as the described reduced IFNγ or TNF production or high IL-1 levels in the susceptible mice, which are only cited from other studies.

• While it is discussed in the study, that Ifnγ and TNF would be the most promising ligands to mediate protective IL-1 effects, a mixed BMT with IL17-/- IL22-/- bone marrow is used and it is not fully clear, why the authors focused on this pathway instead of IFNγ and TNF and could be discussed more in the text.

• While it is great how this study links their results to human data, it would be more convincing if the authors additionally included a human subset with another infectious (f.e. streptococcal pneumonia) or non-infectious lung disease for the discrimination of human Mtb disease from healthy humans with the mouse signatures:  Since the Mtb specific genes especially in the susceptible mice are characterized by a strong inflammatory response and the Mtb specificity is identified by comparison to naïve mice (Method section, Gene signature and ROC curve analysis), it might otherwise also seem that these genes will always predict a diseased state rather than a healthy patient, which could be validated by including another human lung disease cohort as a comparison.

Significance/Novelty

Using scRNAseq and mouse infection experiments, the preprint elegantly explains the so far paradoxical effects of increased IL-1 signaling in exacerbated tuberculosis and the protective effects of IL-1 by describing the role of IL-1Ra expressing immunosuppressive macrophages and IL-1Ra signaling to limit these protective IL-1 effects. The authors characterize Mtb susceptible mouse models and find a conserved role for IL-1 inhibition in all three mouse models, as well as expression of IL-1Ra and other immunosuppressive molecules in non-human primates and humans. While their extensive characterization of Mtb-susceptible mice is an important contribution for scientists working with these genetically deficient mice, their overall findings fit well with the recently proposed tipping point model of tuberculosis by Mayer-Barber KD (2023, Current Opinion in Immunology, https://doi.org/10.1016/j.coi.2023.102365) and describe a fascinating role for immunosuppressive myeloid cells in infectious disease. Especially in tuberculosis, where antibiotic treatment is usually proceeding over months, these immunosuppressive cells could be a very interesting targets for therapeutic interventions reversing the immunosuppression.

Credit

Reviewed by Lisabeth Pimenov as part of a cross-institutional journal club between the Vanderbilt University Medical Center (VUMC), the Max-Delbrück Center Berlin, the Medical University of Vienna and other life science institutes in Vienna.

The author declares no conflict of interests in relation to their involvement in the review.