Study participants
Patients with suspected MSMD, i.e. presenting with severe, persistent, unusual and/or recurrent (SPUR) TB, mycobacterial infections or other clinical MSMD-defining infections such as Salmonella were recruited to the study over a period of five years (September 2013–July 2018). Patients were recruited through the paediatric immunology clinic at Tygerberg academic hospital in the Western Cape of South Africa. A severe infection was defined as an infection that was uncontrolled or complicated—e.g. TB empyema, TB pericarditis, TB meningitis, TB spine, multiple osteoarticular sites of TB, other disseminated forms of TB including miliary TB. Persistent infections were defined as infections that required longer duration of therapy (> 6 or > 9 months depending on severity) or had a lack of response to appropriate treatment regimen for > 2 months. Unusual was defined as infections at an unusual or unexpected site, e.g. dissemination of TB to ears, spine/brain, liver, osteoarticular TB, etc., or an unusual organism, e.g. M. avium or BCG which only cause disease in patients with compromised immune systems; and recurrence was defined as two or more occurrences of the same infection (at least one year apart) despite completion of therapy and subsequent elimination of previous infections.
Only individuals under the age of 15 years and older individuals who had their first relevant clinical presentation before the age of 15 years were included in this study. All the patients underwent standard clinical examinations (including analysis of a medical and family history, vaccination and infection history, physical examination with a focus on assessment of any dysmorphic features, evaluation of lymph nodes and rashes, etc.) and routine laboratory testing (including HIV tests, conventional TB testing including the Qiagen QuantiFERON TB IFN-γ release assay (QFT), full blood counts and differential subsets enumeration, enumeration of immunoglobulins, Neutrophil burst assay, etc.) to exclude other illnesses or causes of immunodeficiency.
Blood was collected by venepuncture into ethylenediaminetetraacetic acid (EDTA) tubes and peripheral blood mononuclear cells (PBMCs) were isolated using the density-gradient centrifugation method [31]. PBMCs were cryopreserved in liquid nitrogen. Blood plasma was also stored at – 80 °C.
An additional MSMD patient from our setting presenting with both pulmonary TB and disseminated M. avium infection, with a well described heterozygous, partial dominant mutation in IFNGR1 (c.818del4), was included as a control for the functional assays [32]. PBMCs were also collected from 10 self-reported healthy volunteers (ages: 22–45 years) to serve as controls for this study.
Cytokine-induced cytokine production assay
The assay used in this study to assess the levels of IL-12-induced IFN-γ production and vice versa was derived from the IFN-γ and IL-12 production assays with BCG/Phytohaemagglutinin (PHA) co-stimulation developed by Feinberg et al. [33] and Dorman and Holland [30, 34].
All recombinant human cytokines used in this study were procured from BD Biosciences (CA, USA). PHA, PenStrep and Foetal Bovine Serum (FBS) was procured from Sigma-Aldrich® (MO, USA).
Cryopreserved PBMCs were thawed, washed with Roswell Park Memorial Institute medium (RPMI-1640 with 1% PenStrep and 10% FBS) and counted on a Bio-Rad TC20™ cell counter (Bio-Rad Laboratories, CA, USA) to determine cell concentration and viability. Four wells in a 24-well cell culture plate were prepared for each patient: (1) no activation (NIL), (2) PHA activation (5 µg/mL), (3) PHA activation and IL-12 stimulation (20 ng/mL), and (4) PHA activation and IFN-γ stimulation (50 ng/mL). Between 150,000 and 500,000 cells were added per well (with > 90% viability), depending on availability. The exact number of cells added into each well was recorded in order to adjust cytokine concentrations to represent cytokine production per 105 PBMCs. The plate was then incubated for 48 h at 37 °C with 5% CO2. After incubation, the contents of each well were centrifuged at 2,500 RCF for 10 min and the supernatants were harvested.
IFN-γ and IL-12p70 concentration was determined through Enzyme-linked Immunosorbent assay (ELISA) using Quantikine® ELISA for human IFN-γ and Quantikine® HS ELISA for human IL-12p70 (Qiagen, Hilden, Germany), according to manufacturer’s instructions. Concentrations of each cytokine were expressed as pg/mL cytokine produced by the equivalent of 105 PBMCs:
$$[\text{cytokine produced per }{10}^{5}\text{ cells}\left(\frac{{\text{pg}}}{{\text{mL}}}\right)=\frac{{\text{total cytokine concentration }\left(\frac{{\text{pg}}}{{\text{mL}}}\right)\text{measured by ELISA}}}{{\text{number of cells per well}}} \times {10}^{5}]$$
Net IL-12 production induced by IFN-γ and net IFN-γ production induced by IL-12 was determined for each individual using the calculated concentrations of cytokines produced per 105 PBMCs in each of the respective wells.
IFN-γ concentration present in blood plasma at the time of PBMC isolation was also determined for each patient using the same Quantikine® ELISA kit for human IFN-γ.
Assessment of cytokine receptor expression and signalling through Flow Cytometry
All monoclonal antibodies, recombinant human cytokines, and Phosflow™ reagents used in this study were procured from BD Biosciences (CA, USA). Flow cytometric data acquisition was performed on a BD LSR II (BD Biosciences, CA, USA) at the BD-CAF Flow Cytometry Centre at Stellenbosch University, Tygerberg Medical Campus.
Cytokine receptor detection through standard surface flow cytometry
Assessment of IFN-γR1 (CD119) and IL-12Rβ1 (CD212) expression has previously been used to confirm several cases of MSMD and the method used in this study was adapted from these previous studies [30, 35,36,37,38]. For cytokine receptor detection, cryopreserved PBMCs were thawed and washed with RPMI (with 1% PenStrep and 10% FBS), and then split into two separate polypropylene tubes, with each tube containing at least 5 × 105 cells (with > 90% viability). One of the tubes were left unstimulated, for IFN-γR1 detection, and 5 µg/ mL PHA was added to the other, for upregulation of IL-12Rβ1 [39]. Both tubes were placed in a 5% CO2 incubator at 37 °C for 24 h. After the incubation period, the PBMCs were washed twice, first with RPMI (with 1% PenStrep and 10% FBS) and then with staining buffer (Phosphate buffered saline [PBS; Sigma-Aldrich®, MO, USA] with 2% FBS).
Surface expression of IFN-γR1 (CD119) and IL-12Rβ1 (CD212) on lymphocyte subsets, NK cells and monocytes were then evaluated by standard surface flow cytometry, whereby cells were stained with the following fluorochrome-conjugated monoclonal antibodies at the specified dilutions: 1:20 CD3-FITC (UCHT1), 1:160 CD4-APC (SK3), 1:40 CD8-PE-Cy5 (HIT8a), 1:40 CD14-PE-Cy7 (M5E2), 1:20 CD56-BV510 (R19-760), 1:160 CD19-BB700 (HIB19), 1:20 CD119-PE (GIR-208), 1:20 CD212-BV421 (2-4E6)] and 1:2000 FVS575V (a fixable, amine reactive viability dye).
Assessment of cytokine signalling through phospho-specific intracellular flow cytometry
IFN-γ and IL-12 signalling was assessed by measuring the intracellular phosphorylation of the respective cytokine receptor-associated signal transducer molecules STAT1 and STAT4. This was achieved by an intracellular flow cytometric-based phosphorylation assay—Phosflow™—where, after PBMC stimulation with either recombinant human IFN-γ or IL-12 cytokines, levels of phosphorylated STAT1 and STAT4 (pSTAT1 and pSTAT4) were measured intracellularly by detecting the bound phospho-specific fluorochrome-conjugated antibodies. Similar methods for the detection of pSTAT1 and pSTAT4 have previously been used to assess IFN-γ and IL-12 signalling in several MSMD cases [30, 36, 40].
Cryopreserved PBMCs were thawed and washed with RPMI (with 1% PenStrep and 10% FBS), and split into three separate polypropylene tubes (labelled unstimulated, IFN-γ stimulated, and IL-12 stimulated), with each containing at least 5 × 105 cells (with > 90% viability). PHA (5 µg/mL) was added to the IL-12 stimulated tube—to upregulate IL-12Rβ1 expression for optimal pSTAT4 detection following IL-12 stimulation. The samples were then placed in a 5% CO2 incubator at 37 °C for ± 18 h. After the incubation period, the PBMCs were washed with RPMI (with 1% PenStrep and 10% FBS). The cells were subsequently stained with the fixable viability marker, FVS575V.
Cells were then stimulated with 500 µL of either RPMI (unstimulated), 100 ng/mL human recombinant IFN-γ (IFN-γ stimulated) or 100 ng/mL human recombinant IL-12 (PHA pre-stimulated tube) for 15 min at 37 °C, 5% CO2. Immediately after stimulation, the cells were fixed using BD Cytofix™. Thereafter, BD Perm IV (1X) was used for permeabilisation of cells.
Each tube was then stained with an antibody cocktail containing the following antibodies at the specified dilutions: 1:20 CD3-FITC (UCHT1), 1:160 CD4-APC (SK3), 1:40 CD8-PE-Cy5 (HIT8a), 1:40 CD14-PE-Cy7 (M5E2), 1:20 CD56-BV510 (R19-760), 1:10 CD20-PerCP-Cy5.5 (H1), 1:20 pSTAT1-BV421 (4a), and 1:5 pSTAT4-PE (38-pSTAT4).
Data analysis
For the receptor panel, the frequency (%) of CD119 and CD212 expression for each of the parent populations were measured as well as the density or median fluorescence intensity (MFI) of expression on each cell subset. For the signaling panel, the fold change in pSTAT detection between the cytokine stimulated and unstimulated samples for each patient was determined for both pSTAT1 and pSTAT4. Fold change in pSTAT detection was calculated as the MFI of cytokine stimulated sample divided by the MFI of unstimulated sample.
The ELISA data and the flow cytometric data [analysed in FlowJo® version 10.6.2 (©Becton Dickinson & Company)] were exported into Microsoft Excel files for statistical analysis. Medians and 95% confidence intervals (CI) were used to describe the non-parametric data. The 95% CI was used as a ‘normal’ range against which patient data could be compared to on an individual basis.
Genetic testing
Following the functional assays, the patients were referred for whole exome sequencing (WES) through the Primary Immunodeficiencies Genetics Network (PIDDGEN) research group at Stellenbosch University. Sequencing was carried out on the Ion Proton™ (Thermo Fisher, Carlsbad, California, United States) at the Central Analytical Facility (CAF) at Stellenbosch University, Stellenbosch, South Africa. Sequences were aligned to the human reference genome, hg19 (https://www.ncbi.nlm.nih.gov/assembly/GCF_000001405.13/) using TMA (version 5.10) in the ion-analysis workflow on the Torrent Suite (version 5.10). The variant caller (version 5.10) plugin on the Torrent suite was used for base quality score recalibration, indel realignment and variant calling. TAPER™, a custom-designed, in-house method was used for variant prioritization and variants in a homozygous recessive state or compound heterozygous state were ranked as potential candidate variants [27].
Various in-silico prediction tools were used to estimate the likelihood of pathogenicity of the identified variants. Sorting intolerant from tolerant (SIFT) [41], PolyPhen [42], MutationTaster [43] and Combined Annotation Dependant Depletion (CADD) Scores [44] were all used. Thereafter The American College of Medical Genetics and Genomics (ACMG) criteria [45] and the newer Sherloc criteria, which is a refined version of the ACMG criteria, [46] were used to make a final call on the pathogenicity of each variant.
Candidate variants identified through WES were confirmed by Sanger sequencing using the BigDye® Terminator v3.1 Cycle Sequencing Kit (Perkin-Elmer, Applied Biosystems Inc., CA, USA.), followed by electrophoresis on an ABI 3130XL Genetic Analyzer (Perkin-Elmer, Applied Biosystems Inc., CA, USA). All automated DNA sequencing reactions were performed at CAF at Stellenbosch University, Stellenbosch, South Africa.