Abstract
Dipeptidyl peptidase 9 (DPP9) is a direct inhibitor of NLRP1, but how it impacts inflammasome regulation in vivo is not yet established. Here, we report two families with immune-associated defects, skin pigmentation abnormalities and neurological deficits that segregate with biallelic DPP9 rare variants. Using patient-derived primary cells and biochemical assays, these variants are shown to behave as hypomorphic or loss-of-function alleles that fail to repress NLRP1. Remarkably, the removal in mice, of a single copy of either Nlrp1a/b/c, Asc, Gsdmd, Il-1r, but not Il-18, was sufficient to rescue the lethality of Dpp9 mutant neonates. These experiments suggest that the deleterious consequences of DPP9 loss are mainly driven by the aberrant activation of the canonical NLRP1 inflammasome and IL-1β signaling. Collectively, our results delineate a Mendelian disorder of DPP9 deficiency driven by increased NLRP1 activity as demonstrated in patient cells and in a mouse model of the disease.
One Sentence Summary Loss of DPP9 activity in humans and mice results in pathogenic NLRP1 overactivation.
The human NLRP1 inflammasome is a multi-protein complex that is triggered by viral infections(1, 2), or danger-associated signals(3), resulting in the release of proinflammatory cytokines, including IL-1 family members, IL-1β and IL-18(4). Dysregulation of the human NLRP1 inflammasome is associated with a range of inflammatory Mendelian diseases such as Multiple Self-healing Palmoplantar Carcinoma (MSPC, MIM615225)(5), Juvenile Recurrent Respiratory Papillomatosis (JRRP, 618803MIM)(6), and Autoinflammation with Arthritis and Dyskeratosis (AIADK, MIM617388)(7). In the absence of any cognate triggers, inflammasome sensors in the family of Nod-like receptors (NLRs) are maintained in an inactive state by a number of regulatory proteins. For example, NLRP1 is bound to, and repressed by, the enzymatic activity and scaffolding function of dipeptidyl peptidase 9 (DPP9) and DPP8(7, 8). The role of DPP9 as a negative regulator of human or mouse NLRP1 has been established through pharmacological inhibition of DPP9 and inactivation of DPP9 in cell lines in vitro, but its function has not been documented in vivo.
As part of a cohort of patients with inborn errors of immunity, we studied a child (F2: II.3) with signs of susceptibility to infections consisting of recurrent facial HSV lesions and airway inflammation, including chronic intermittent asthma and primary pulmonary tuberculosis. Exome sequencing disclosed a germline homozygous stop-gained mutation (c.2551C>T; p.Q851*) in DPP9 (MIM608258) (Fig. 1, A, B and E) which was heterozygous in the parents. A second child (F1: II.1) was identified with help of the GeneMatcher(9) database and carried compound heterozygous DPP9 variants with a paternal missense mutation (c.449G>A, p.G167S) and a maternal early truncating mutation (c.641C>G; p.S214*) (Fig. 1, A to C). Both children shared a triad of symptoms summarized in Table 1 which partially overlapped with NLRP1 gain-of-function mutations causing AIADK(7). This consisted of a general failure to thrive, skin manifestations and blood anomalies, which in the case of F1: II.1 required hematopoietic stem cell transplant. The stop-gain mutations found in the two probands are expected to completely abolish the enzymatic activity of DPP9, as its catalytic domain is situated at the C-terminus. The missense mutation p.(G167S) found in proband F1: II.1 affects a residue in a flexible loop that is located at the periphery of the DPP9 substrate binding site (Fig. 1D). Thus, the DPP9G167S mutation is predicted to diminish the substrate binding and/or catalytic efficiency. To date, no homozygous damaging variants have been reported for DPP9 which according to genomAD is significantly averse to loss-of-function mutations(10).
To investigate the impact of the DPP9 variants on mRNA expression and enzymatic function, primary skin cells were isolated from affected patients of both families (Fig. 2A). When compared with control primary fibroblasts, and those derived from the heterozygous parents of family 2 (F2: I.3 and I.4), a significant decrease in DPP9 protein was found in the proband’s (F1: II.1) fibroblasts. This was consistent with one allele not encoding the full-length protein due to an early stop mutation p.S214* (Fig. 2A, lanes 1 and 2). Notably, we could not observe any endogenous DPP9 protein in the proband of family 2 (F2: II.3), indicating that the stop-gain mutant (p.Q851*) was degraded via the nonsense-mediated decay pathway (Fig. 2A, lanes 1 and 5). Thus, the p.Q851* DPP9 variant probably yields a loss-of-expression, resulting in a protein-null allele. We next sought to assess the enzymatic activity of the DPP9G167S variant towards a model substrate Gly-Pro-AMC (Fig. 2B and fig. S1A). The p.G167S mutation was found to reduce the enzymatic activity to ∼15% of that of wild-type DPP9.
Given that DPP9 has been identified as a specific interacting partner and inhibitor of human and mouse NLRP1(7, 8), we measured the ability of DPP9WT, catalytically-dead DPP9S759A or the patient DPP9G167S mutant to repress NLRP1 when re-expressed in DPP8/DPP9 double KO 293T-ASC-GFP reporter cells (Fig. 2C and fig. S1, B and C). Exogenous DPP9WT repressed NLRP1-dependent ASC-GFP speck formation. In contrast, both DPP9S759A and DPP9G167S were less effective in inhibiting ASC speck formation. Taken together, these experiments suggest that these DPP9 patient variants are less able to maintain NLRP1 in the inactive state.
We and others have demonstrated that NLRP1 is the most prominent inflammasome sensor in human keratinocytes and that these cells readily undergo pyroptotic cell death in response to NLRP1 activation(5, 11). Using primary keratinocytes isolated from family 2 and the Luminex platform, we found that the cytokine and chemokine profile of the proband differed from those of his heterozygous parents (fig. S1D). We demonstrated that the proband keratinocytes, as do keratinocytes from patients with gain-of-function NLRP1 mutations(5), spontaneously secreted active IL-1β and IL-18 into the media without any exogenous stimulus. Despite measurable IL-1 release visualised by ELISA and western blot (Fig. 2, D and E respectively), the patient’s keratinocytes were viable in culture and could be serially cultured (fig. S1E) for at least 4 passages. This cytokine release was also accompanied by ASC oligomerization, as demonstrated using a DSS crosslinking assay, and GSDMD cleavage (Fig. 2E). These data are consistent with a small fraction of the cells undergoing spontaneous inflammasome activation, without overt pyroptosis. A similar phenomenon has been documented in murine macrophages(12).
To dissect the pathogenicity of the loss of DPP9 in vivo, we took advantage of the DPP9 mutant mouse line which carries the catalytic-dead DPP9 p.S729A mutation. Homozygous Dpp9S729A/S729A knock-in mice die within 24 hours of birth(13, 14). Based on our results obtained in patient-derived primary cells, we examined whether this lethality might be associated with overactivation of the NLRP1 inflammasome. Indeed, homozygous loss of all Nlrp1 alleles (Nlrp1a/b/c, herein referred to as Nlrp1) protect Dpp9S729A/S729A mice from neonatal lethality. Furthermore, deletion of even a single Nlrp1 allele was also sufficient to generate viable animals (Fig. 3A and fig. S2, A and B). Dpp9S729A/S729ANlrp1−/− mice were healthy and proved to be fertile, although runted compared to Dpp9S729A/+Nlrp1−/− siblings (Fig. 3, A and B). Short stature of Dpp9S729A/S729ANlrp1−/− was associated with decreased hip width but relatively conserved limb width (Fig. 3, C and D).
At birth (day 0), Dpp9S729A/S729A mice appeared normal and were suckling, as evidenced by a milk spot (Fig. 3E). We used flow cytometry to characterise immune cell populations in the blood of Dpp9S729A/S729A mutant mice sufficient, heterozygous or deficient for NLRP1, along with relevant controls. Numbers of monocytes, dendritic cells, T cells, B cells, NK cells and neutrophils in the bone marrow were comparable between all genotypes assessed (Fig. 3F and fig. S2D). This is consistent with viable fetal liver chimeras that have previously been generated from this line(15, 16). In contrast, we observed a dramatic and specific monocytopenia affecting Ly6Cmed and dendritic cells in the blood (Fig. 3G and fig. S2, C and D). This corresponds with cell types that have the highest expression levels of Nlrp1a and Nlrp1b(17) (fig. S2E) and indeed, monocytopenia is rescued by the heterozygous or homozygous loss of Nlrp1 (Fig 3G). Thus, the haematopoietic abnormality and lethality caused by germline loss of DPP9 activity in mice is largely driven by overactive NLRP1 signaling.
In order to determine NLRP1-independent effects of DPP9 loss of function we further interrogated the Dpp9S729A/S729A Nlrp1−/− line. Aside from being runted, Dpp9S729A/S729A Nlrp1−/− mice had a slight decrease in life expectancy (Fig. 3H) compared to control lines. Dpp9S729A/S729A Nlrp1−/− mice also displayed immune infiltrate in the lungs which contained foamy macrophages (F4/80+), B cells (B220+) and T cells (CD3+) post 6 months of age (Fig. 3I and fig. S3, A and B). While histological analysis of other organs and automated blood cell analysis showed no overt differences between genotypes (fig. S3, C to F). Overall, these results suggest inflammasome-independent roles of DPP9 relating to stature and lung inflammation, which is consistent with the broad range of processes regulated by DPP9. Decreased stature and lung inflammation were also recorded in DPP9-deficient patients (Table 1). Notably, two GWAS studies have implicated germline DPP9 variants in the development of pulmonary fibrosis(18, 19) and severe COVID19 pathology(20).
To further dissect the contribution of downstream components of the NLRP1 signalling axis, we next crossed Dpp9S729A/S729A mice with strains deficient for effectors of the NLRP1 inflammasome. Both heterozygous and homozygous loss of Asc, Gsdmd and Il-1r rescued Dpp9S729A/S729A neonatal lethality to a similar extent as Nlrp1 deficiency (Fig. 3J). Interestingly, the biallelic loss of Il-18 did not rescue DPP9 loss of function (Fig. 3J) suggesting that Il-1 may be the main driver of the disease. Rescued mice exhibited a similar phenotype to Dpp9S729A/S729A Nlrp1−/− mice, appearing healthy yet runted. For example, Dpp9S729A/S729A Gsdmd−/− mice did not gain weight at the same rate as controls, and developed the same immune cell infiltrate in their lungs as Dpp9S729A/S729A Nlrp1−/− mice at 6 months of age, despite normal blood cell counts (fig. S4). The number of rescued Dpp9S729A/S729A mice that survived to weaning age fell significantly below Mendelian ratios (Fig. 3J). However, these numbers might be confounded by potential cannibalisation of the runted pups. Overall it is clear that the deletion of a single allele of either Asc, Gsdmd or Il-1r is sufficient to protect Dpp9S729A/S729A mice. These results raise the possibility that even partial inhibition of the NLRP1-inflammasome pathway, with anti-IL-1β antibodies for example, would provide some therapeutic benefit for DPP9 deficient patients. In further support of this, anti-IL-1β therapy was partially successful in a patient harbouring a germline NLRP1 p.P1214R mutation which abrogated binding and therefore inhibition by DPP9(21).
In summary, we have identified two unrelated families with autosomal recessive DPP9 deficiency, which is characterised by immune-associated defects, dermatological anomalies and neurological deficits. This Mendelian disease bears some degree of resemblance with auto-immune disorders caused by NLRP1 activating mutations. In particular the AIADK-causing NLRP1 p.P1214R mutation which entirely abrogates DPP9 binding and inhibition of NLRP1, is expected to yield comparable phenotypes(7). Both diseases appear to manifest with recurrent episodes of fevers, auto-immune anemia and a general failure to thrive (Table 1). The milder skin manifestations seen in DPP9 patients may be explained by the presence of active DPP8 which we have shown can compensate for DPP9 deficiency in human cells(7).
Using biochemical and cellular assays, we demonstrated that all patient-derived DPP9 alleles had reduced catalytic activity and were defective in suppressing the NLRP1 inflammasome in vitro. These results suggest that the patients’ symptoms are at least in part due to the aberrant activation of the NLRP1 inflammasome. This notion was validated by in vivo experiments: the neonatal lethality of Dpp9 mutant mice was rescued by the deletion of Nlrp1 and downstream inflammasome signaling components Asc, Gsdmd and Il-r1. Thus IL-1β inhibition might be a viable therapeutic strategy for these patients. Recent studies have shown that human DPP9 has many cellular substrates and also represses an orthologous inflammasome sensor, CARD8, which is absent in rodents(22–25). These findings could account for the broad spectrum of symptoms in patients with DPP9 mutations, some of which might not be modelled in mice. Overall, our results delineate a new human Mendelian disease of germline DPP9 deficiency, whose pathogenesis is largely driven by the aberrant activation of the NLRP1 inflammasome.
Data Availability
The data that support the findings of this study are available from the corresponding authors upon request.
Funding
The Laboratory of Human Genetics of Infectious Diseases is supported by the National Center for Research Resources and the National Center for Advancing Sciences (NCATS) of the National Institutes of Health (NIH) Clinical and Translational Science Award (CTSA) program (UL1TR001866), the French National Research Agency (ANR) under the “Investments for the Future” ANR program (ANR-10-IAHU-01), the Integrative Biology of Emerging Infectious Diseases Laboratory of Excellence (ANR-10-LABX-62-IBEID), and the “PNEUMOPID” project (Grant ANR 14-CE15-0009-01), the French Foundation for Medical Research (FRM) (EQU201903007798), the Howard Hughes Medical Institute, the Rockefeller University, the St. Giles Foundation, Institut National de la Santé et de la Recherche Médicale (INSERM), and the “Université de Paris”. K.L. is supported by a NMRC Open Fund - Young Individual Research Grant (OF-YIRG19MAY-0039). F.L.Z. is an Nanyang Assistant Professor and a recipient of National Research Foundation (NRF, Singapore) fellowship. B.R. is an investigator of the National Research Foundation (NRF, Singapore) and Branco Weiss Foundation (Switzerland) and an EMBO Young Investigator. This work was funded by a Strategic Positioning Fund for Genetic Orphan Diseases and an inaugural A*STAR Investigatorship from the Agency for Science, Technology and Research in Singapore.
Author contributions
F.L.Z, S.L.M, and B.R designed the study. J.B., D.B., H.K., and N.H. made clinical diagnoses and collected clinical data and samples. C.H, P.L, S.D, C-H.Y and S.L.M performed and supervised the mouse experiments. K.R. and K.L. performed biochemical and cell-based experiments and analyzed the data. C.H, K.R, K. L, F.L.Z, S.L.M and B.R wrote the manuscript with input from all authors.
Competing interests
S.L.M. is a Scientific Advisor for IFM Therapeutics.
Data and materials availability
All data is available in the main text or the supplementary materials.
Supplementary Materials
Materials and Methods
Figures S1-S4
Acknowledgments
We are grateful to all members of the Masters, Zhong and Reversade laboratories for support. We thank Prof. Anthony Oro (Stanford University) for generous help in fibroblast derivation from family 1. We thank E. Kravets for study coordination and M.W. Allain for data collection (Stanford University). We thank S. Russo for outstanding animal husbandry.