Novel Majeed Syndrome-Causing LPIN2 Mutations Link Bone Inflammation to Inflammatory M2 Macrophages and Accelerated Osteoclastogenesis

Abstract

Objective: To identify novel heterozygous LPIN2 mutations in a patient with Majeed syndrome and characterize the pathomechanisms that lead to the development of sterile osteomyelitis.

Methods: Targeted genetic analysis and functional studies assessing monocyte responses, macrophage differentiation, and osteoclastogenesis were conducted to compare the pathogenesis of Majeed syndrome to interleukin-1 (IL-1)-mediated diseases including neonatal-onset multisystem inflammatory disease (NOMID) and deficiency of the IL-1 receptor antagonist (DIRA).

Results: A 4-year-old girl of mixed ethnic background presented with sterile osteomyelitis and elevated acute-phase reactants. She had a 17.8-kb deletion on the maternal LPIN2 allele and a splice site mutation, p.R517H, that variably spliced out exons 10 and 11 on the paternal LPIN2 allele. The patient achieved long-lasting remission receiving IL-1 blockade with canakinumab. Compared to controls, monocytes and monocyte-derived M1-like macrophages from the patient with Majeed syndrome and those with NOMID or DIRA had elevated caspase 1 activity and IL-1β secretion. In contrast, lipopolysaccharide-stimulated, monocyte-derived, M2-like macrophages from the patient with Majeed syndrome released higher levels of osteoclastogenic mediators (IL-8, IL-6, tumor necrosis factor, CCL2, macrophage inflammatory protein 1α/β, CXCL8, and CXCL1) compared to NOMID patients and healthy controls. Accelerated osteoclastogenesis in the patient with Majeed syndrome was associated with higher NFATc1 levels, enhanced JNK/MAPK, and reduced Src kinase activation, and partially responded to JNK inhibition and IL-1 (but not IL-6) blockade.

Conclusion: We report 2 novel compound heterozygous disease-causing mutations in LPIN2 in an American patient with Majeed syndrome. LPIN2 deficiency drives differentiation of proinflammatory M2-like macrophages and enhances intrinsic osteoclastogenesis. This provides a model for the pathogenesis of sterile osteomyelitis which differentiates Majeed syndrome from other IL-1-mediated autoinflammatory diseases.

Trial registration: ClinicalTrials.gov NCT02974595.

© 2020 The Authors. Arthritis & Rheumatology published by Wiley Periodicals LLC on behalf of American College of Rheumatology. This article is a U.S. Government work and is in the public domain in the USA.

Figures

Figure 1

Novel mutations in LPIN2 cause a compound heterozygous form of Majeed syndrome. A, The patient with Majeed syndrome was compound heterozygous (Comp Het) for 2 rare variants in LPIN2. B, Magnetic resonance imaging (STIR imaging) of the patient’s knee shows bone marrow enhancement without cortical lesions. C, Genomic structure is shown along with location of the LPIN2 locus and known disease‐causing LPIN2 variants reported in the literature (blue). The patient’s maternal allele is depicted with deleted exons (red), while the paternal allele is depicted with splice site mutations, including spliced out exon 10 (red) or exons 10 and 11 (orange). D, LPIN2 splice junctions were quantified in the patient’s blood using alignment data on whole‐blood RNA‐Seq analysis, yielding predicted transcript levels with each deletion (healthy control [HC] data not shown). Prediction of 56% of the transcript with the maternal deletion suggests that this mutation confers protection against nonsense‐mediated mRNA decay; the 2 alternatively spliced transcripts lacked either exon 10 alone (17.2% of the transcript) or exons 10 and 11 (26.7% of the transcript). E, LPIN2 mRNA expression in human cell subsets and tissues was quantified by real‐time quantitative polymerase chain reaction. Fold changes in LPIN2 expression (relative to 18S) are shown; bars show the mean ± SD of 3 technical replicates. F, Levels of LPIN1, LPIN2, and LPIN3 mRNA were quantified by RNA‐Seq analysis in human monocytes from healthy controls; bars show the mean ± SD of 3 samples. PBMC = peripheral blood mononuclear cell; NK = natural killer; HUVEC = human umbilical vein endothelial cell.

Figure 2

Effect of disease‐causing LPIN2 mutations on interleukin‐1β (IL‐1β) production in monocytes and macrophages. IL‐1β concentration in supernatant was analyzed by enzyme‐linked immunosorbent assay. A, Adherent monocytes from peripheral blood mononuclear cells were stimulated with lipopolysaccharide (LPS; 1 μg/ml) for 2.5 hours and ATP (1 mmole) for 30 minutes. Each symbol represents an individual subject: healthy controls (HCs) (n = 4), patient with Majeed syndrome (n = 1), patients with neonatal‐onset multisystem inflammatory disease (NOMID) (n = 3), and patients with deficiency of the IL‐1 receptor antagonist (DIRA) (n = 3). Bars show the mean ± SD. Values from the patient with Majeed syndrome represent the mean of 3 technical replicates. B, LPS‐ and ATP‐stimulated monocytes were cultured in the presence or absence of anakinra (10 mg/ml). Each symbol represents an individual subject: healthy controls (n = 3) and patient with Majeed syndrome (n = 1). Bars show the mean ± SD. Values from the patient with Majeed syndrome were obtained from 3 independent experiments, using samples from separate visits, with the mean of 2 technical replicates shown. C and D, M1‐like macrophages (C) and M2‐like macrophages (D) were stimulated with LPS (1 μg/ml) and ATP (1 mmole) for 24 hours. Each symbol represents an individual subject: healthy controls (n = 6), patient with Majeed syndrome (n = 1), patients with NOMID (n = 3), and patients with DIRA (n = 4). Bars show the mean ± SD. Values for the patient with Majeed syndrome represent the mean of the technical replicates. * = P < 0.05; ** = P < 0.001, by 2‐sample t‐test assuming equal variances to compare disease groups, or by Student’s unpaired t‐test to compare stimulation with anakinra to no stimulation with anakinra.

Figure 3

Cytokine and chemokine production upon stimulation of M1‐like macrophages (M1‐MDMs) and M2‐MDMs. A and B, M1‐MDMs (A) and M2‐MDMs (B) were stimulated with LPS (1 μg/ml) and ATP (1 mmole) for 24 hours. The relative levels of various cytokines and chemokines in supernatant were analyzed by antibody array, and cytokine density was analyzed using ImageJ software. C, IL‐10 levels were analyzed by enzyme‐linked immunosorbent assay. Each symbol represents an individual subject: healthy controls (n = 4), patient with Majeed syndrome (n = 1), patients with NOMID (n = 3), and patients with DIRA (n = 4). Bars show the mean ± SD. Values from the patient with Majeed syndrome represent the mean of 2 technical replicates.* = P < 0.05; ** = P < 0.001; *** = P < 0.0001, by 2‐sample t‐test assuming equal variances. MCP‐1 = monocyte chemotactic protein 1; GROα = growth‐related oncogene α; TNF = tumor necrosis factor; MIP‐1α/β = macrophage inflammatory protein 1α/β; IP‐10 = interferon‐γ–inducible protein 10 (see Figure 2 for other definitions).

Figure 4

M2‐like macrophages (M2‐MDMs) in Majeed syndrome are more osteoclastogenic, and anakinra could reduce osteoclast fusion. A, MDMs were differentiated with macrophage colony‐stimulating factor (M‐CSF; 100 ng/ml) for 6 days and with a combination of M‐CSF (100 ng/ml) and RANKL (30 ng/ml) for another 6 days, to differentiate them into osteoclasts. Tartrate‐resistant alkaline phosphatase (TRAP)–positive multinucleated cells were quantified in the patient with Majeed syndrome, as well as in DIRA patients, healthy controls, and NOMID patients. B, Pit formation by osteoclasts was measured in macrophage cultures with fluoresceinamine‐labeled chondroitin sulfate/calcium phosphate. In A and B, results are based on 3 independent experiments, with values shown as the mean ± SD (healthy controls, n = 4; patient with Majeed syndrome, n = 1 with 2 visits; patients with NOMID, n = 3; and patients with DIRA, n = 3). C and D, M2‐MDMs were stained for RANK expression on day 7 of culture with M‐CSF (C) or stained for IL‐1 receptor (IL‐1R) (D). Nuclei were stained with DAPI. In C and D, results are based on 3 independent experiments, with values shown as the mean ± SD intensity, quantified using Imaris software (version 9.2.1) (healthy controls, n = 3; patient with Majeed syndrome, n = 1). Arrowheads show the altered surface distribution of IL‐1R in the patient’s M2‐MDMs compared to controls. In A–D, original magnification × 20. * = P < 0.05; ** = P < 0.001, by Student’s unpaired t‐test. NS = not significant (see Figure 2 for other definitions).

Figure 5

Effect of IL‐1 blockade on osteoclastogenesis and proinflammatory cytokine/chemokine production by M2‐MDMs. A, Monocyte‐derived M2‐MDMs were stimulated with LPS (1 μg/ml) and ATP (1 mmole) for 24 hours, and LPS‐ and ATP‐stimulated monocytes were cultured in the presence or absence of anakinra (10 μg/ml). The relative levels of various cytokines and chemokines in supernatant were analyzed by antibody array, and intensities of cytokine expression were analyzed using ImageJ software. Bars show the mean ± SD in cells from healthy controls (n = 2) (solid bars) and the patient with Majeed syndrome (n = 1 with 2 duplicates) (shaded bars). B, Monocytes from the patient with Majeed syndrome, patients with NOMID, and healthy controls were cultured with macrophage colony‐stimulating factor (M‐CSF) for 6 days, then RANKL was added to the M‐CSF, and, simultaneously, M2‐MDMs were cultured in the presence or absence of anakinra (1 mg/ml or 10 mg/ml). After 6 days, small and large tartrate‐resistant alkaline phosphatase–positive cells with >3 nuclei were quantified. Data are shown as box plots, where the boxes represent the 25th to 75th percentiles, the lines within the boxes represent the median, and the lines outside the boxes represent the minimum and maximum values (healthy controls [n = 3], NOMID patients [n = 3], patient with Majeed syndrome [n = 1 with 2 duplicates]). Original magnification × 20. * = P < 0.05; ** = P < 0.001. TNF = tumor necrosis factor; IP‐10 = interferon‐γ–inducible protein 10; GROα = growth‐related oncogene α; MCP‐1 = monocyte chemotactic protein 1; NS = not significant (see Figure 2 for other definitions).

Figure 6

IL‐1 and macrophage migration inhibitory factor (MIF) blockade inhibit osteoclastogenesis, and mutant lipin 2 regulates proinflammatory M2‐like macrophages (M2‐MDMs) by altering JNK/MAPK expression. A, IL‐1 inhibition with anakinra and MIF inhibition with ISO‐1 inhibited osteoclastogenesis most effectively. Macrophages were differentiated into osteoclasts. Cells were left untreated or treated with a MIF inhibitor (ISO‐1; 100 mM), an IL‐1 inhibitor (anakinra; 1 mg/ml), an IL‐6 inhibitor (tocilizumab; 10 mg/ml), a tumor necrosis factor inhibitor (SPD304; 5 mM), or a JNK inhibitor (SP600125; 10 mM). All experiments were performed in cells from healthy controls. Tartrate‐resistant alkaline phosphatase (TRAP)–positive multinucleated cells with >3 nuclei were quantified. Bars show the mean of triplicate assays. Original magnification × 100. B, The Human Phospho‐Kinase Array was used to detect multiple phosphorylated kinases in macrophages in the sample from the patient with Majeed syndrome (shaded bars) and healthy controls (solid bars), differentiated with macrophage colony‐stimulating factor (100 ng/ml) for 6 days. Cell lysate (60 μl) was analyzed to detect the phosphorylation state of M2‐MDMs at the basal level. Spot densities of phosphoproteins were quantified using ImageJ software. Bars show the mean ± SD of 2 independent experiments, with duplicates for the patient with Majeed syndrome. * = P < 0.05; ** = P < 0.001, by Student’s unpaired t‐test. GSK‐3 α/β = glycogen synthase kinase 3 α/β; TOR = mechanistic target of rapamycin (see Figure 2 for other definitions).