IRAK1 Duplication in MECP2 Duplication Syndrome Does Not Increase Canonical NF-κB–Induced Inflammation

Purpose Besides their developmental and neurological phenotype, most patients with MECP2/IRAK1 duplication syndrome present with recurrent and severe infections, accompanied by strong inflammation. Respiratory infections are the most common cause of death. Standardized pneumological diagnostics, targeted anti-infectious treatment, and knowledge of the underlying pathomechanism that triggers strong inflammation are unmet clinical needs. We investigated the influence of IRAK1 overexpression on the canonical NF-κB signaling as a possible cause for excessive inflammation in these patients. Methods NF-κB signaling was examined by measuring the production of proinflammatory cytokines and evaluating the IRAK1 phosphorylation and degradation as well as the IκBα degradation upon stimulation with IL-1β and TLR agonists in SV40-immortalized fibroblasts, PBMCs, and whole blood of 9 patients with MECP2/IRAK1 duplication syndrome, respectively. Results Both, MECP2/IRAK1-duplicated patients and healthy controls, showed similar production of IL-6 and IL-8 upon activation with IL-1β and TLR2/6 agonists in immortalized fibroblasts. In PBMCs and whole blood, both patients and controls had a similar response of cytokine production after stimulation with IL-1β and TLR4/2/6 agonists. Patients and controls had equivalent patterns of IRAK1 phosphorylation and degradation as well as IκBα degradation upon stimulation with IL-1β. Conclusion Patients with MECP2/IRAK1 duplication syndrome do not show increased canonical NF-κB signaling in immortalized fibroblasts, PBMCs, and whole blood. Therefore, we assume that these patients do not benefit from a therapeutic suppression of this pathway. Supplementary Information The online version contains supplementary material available at 10.1007/s10875-022-01390-7.


Introduction
Patients with duplication of methyl CpG binding protein 2 (MECP2) on chromosome Xq28 were first described in 2005 [1,2]. The clinical phenotype is characterized by developmental delay, hypotonia, epileptic seizures, as well as recurrent infections [1,2]. Approximately 1% of severe X-linked intellectual disability in males might be explained by MECP2 duplication syndrome (MDS) [3]. Reviewing the literature, we identified 102 articles describing patients with duplications in Xq28 of varying sizes but encompassing at least the MECP2 and interleukin-1 receptor-associated kinase 1 (IRAK1) gene, 14 of them published before the initial description of MDS (Table 1). From 1987 until now, at least 545 cases with confirmed genotype were published (504 males and 41 symptomatic females) ( Table 1). Additionally, the duplication was suspected in 39 related patients (Table 1). However, the numbers of patients might be underestimated regarding the unevenly distributed origin of publications (43 European,25 North American, 23 East-Asian, 5 rest of Asia including Russia, 3 Australian, 2 South American, 1 African) ( Table 1). Most females with MECP2 duplication are unaffected carriers showing a favorably skewed X chromosome inactivation (XCI) pattern [1,[4][5][6]. However, Table 1 Overview of published patients suffering from MECP2 duplication syndrome. We included all publications describing patients with duplication or triplication of at least MECP2 and IRAK1. Also, we included publications describing patients with duplications in the region which included IRAK1 although this gene was not described by then. Republished cases were only counted once as far as traceable. Note: Throughout the publications, the criteria for intellectual disability and developmental delay differ a lot. ND [7][8][9][10][11][12][13][14][15][16][17][18][19][20]. However, the extent of the symptoms in females with MECP2 duplication cannot be correlated with their XCI pattern, at least as assessed in peripheral blood [21]. Seventy-eight percent of reported patients (376/479) suffer from recurrent or severe infections (Table 1). Most common are respiratory infections with 98% of reported cases (316/324), but patients also present with otitis media, urinary tract infections, and sepsis (Table 1). Early death (defined as < 25 years) is reported with a frequency of 4 to 55% [1,57,92,95]. Among the 67 patients with described cause of early death, 58 (87%) died in the context of a severe infection at the age of 3 weeks to 24 years (median 11 years; data available for 43 patients only) [1, 2, 5-7, 9, 12, 25, 26, 32, 33, 38, 39, 43, 54-57, 63, 67, 70, 75, 76, 80, 84, 86, 87, 90, 92, 93, 95, 97, 99, 104]. Eighty-two percent (328/398) of males suffer from recurrent or severe infections but only 61% (20/33) of the described females. Few studies further examined the detailed infectious and the underlying immunological phenotype of the patients. In contrast to the widespread notion of "recurrent severe infections," information about identified pathogens is only available for 19 patients [7,12,16,54,76,100,105]. Among the 55 isolated pathogens were 45 bacteria (most of all S. pneumoniae, H. influenzae, E. coli, and S. aureus), 6 viruses, and 4 Candida (Table S1). However, as the total viable counts are not stated, it remains unclear if these were the disease-causing pathogens. Bronchoalveolar lavage for the identification of pathogens was only performed in 7 patients [76,105].
Few studies have examined patients for their immunological phenotype [33,39,76,88,100]. The most common characteristic is a poor response to vaccination especially against Streptococcus pneumoniae which was described in 15/26 patients [33,39,76]. Some patients show selective deficiency of immunoglobulin (Ig) A (11/47) and/or IgG2 (7/24) [23,26,30,33,39,46,56,63,75,76,88,100,104]. Moreover, several patients present with episodes of unexplained fever and remarkably high C-reactive protein (CRP) values during non-invasive infections [24,43,54,70,76,105]. In 2015, Bauer et al. suggested the substitution of polyvalent IgG in patients with an IgG2 subclass deficiency and/or low post-vaccination titers against pneumococci who suffer from recurrent infections-eventually combined with prophylactic antibiotics [76,105]. In the 26 studies published since 2016, only three evaluated the immunoglobulin levels, and none mentioned the response to vaccination [88,100,103]. Four patients were mentioned to receive antibiotic prophylaxis [93,100,105]. As infections still limit the quality of life and are the most common cause of death in MDS patients, there seems to be an unmet clinical need regarding pneumological and microbiological diagnostics as well as targeted anti-infectious treatment [92].
It remains unknown whether recurrent fever and strong acute phase response in these patients are rather driven by infections which are difficult to clear and/or by autoinflammation. Throughout the manuscript, we use the term autoinflammation which describes systemic inflammatory processes due to a non-infectious (auto)activation of the innate immune system. Both hypotheses, the one of an "infectious fever" and the one of an "autoinflammatory fever," are not mutually exclusive [31,104,106,107]. In 2009, Kirk et al. suspected a link between IRAK1 duplication and susceptibility to infection [43]. IRAK1 participates in multiple IL-1 and TLR-driven signaling processes that regulate immunity and inflammation [108][109][110][111][112][113][114]. For instance, IRAK1 plays an important role in the regulation of both, the interleukin-1 (IL-1)-mediated and the Toll-like receptor (TLR)-mediated, so-called canonical signaling pathways of NF-κB (nuclear factor "kappa-light-chain-enhancer" of activated B cells) (Fig. 1). Upon binding, IL-1 receptors with their respective cytokine or TLR with their respective ligand recruit the adaptor protein myeloid differentiation primary response 88 (MyD88) which associates with IRAK4 via a homophilic interaction between their death domains. IRAK4 induces the phosphorylation of IRAK1. The hyperphosphorylated IRAK1 then dissociates from the complex and associates with TNF receptor-associated factor 6 (TRAF6) to activate TAK-1/TAB (TGF-β-activated kinase/TAK1-binding proteins). The latter enhances the activity of the IκB kinase (IKK) complex, which in turn leads to phosphorylation and degradation of inhibitors of nuclear factor kappa B (IκB). Thereby, NF-κB dimers comprising p65 (RelA), c-Rel, and p50 are activated and migrate into the nucleus which results in gene transcription and the induction of inflammatory cytokines such as tumor necrosis factor α (TNF-α), IL-1β, IL-6, and IL-12 [108][109][110][111][112][113][114].
Della Mina et al. examined the canonical NF-κB signaling in an IRAK1-null patient [115]. The patient's fibroblasts showed poor responses upon stimulation with TLR2/6 and TLR4 agonists but unimpaired responses to IL-1β. The patient's peripheral blood mononuclear cells (PBMCs) responded normally to IL-1β as well as TLR2/6 and TLR4 agonists [115]. Responses to TLR3 agonist Poly(I:C) were not influenced as it signals via TRIF-dependent pathways [115].
The combination of the clinical phenotype in MDS and the duplication of the IRAK1 gene brings up the question if IRAK1 overexpression causes increased canonical NF-κB signaling and detrimentally increased acute phase responses. Considering the results of Della Mina et al., we hypothesized that patients with MECP2/IRAK1 duplication might show enhanced cytokine production in fibroblasts upon simulation with TLR2/6 and TLR4 agonists. Therefore, we evaluated the production of proinflammatory cytokines as well as the IκBα degradation and IRAK1 phosphorylation upon stimulation with IL-1β and TLR agonists in SV40immortalized fibroblasts of 9 patients with MECP2/IRAK1 duplication syndrome, respectively. Additionally, we investigated the production of proinflammatory cytokines not only in PBMCs but also whole blood.

Patients
The study was conducted in accordance with the ethical standards of the 1964 Helsinki declaration and the institutional research committee (Charité-Universitätsmedizin Berlin, Germany, EA2/063/12). Informed consent was obtained from each patient or the patients' parents. Our cohort consists of 9 male patients diagnosed with MDS. We recruited them by contacting patients who participated as well as physicians who cooperated in our previous study [76]. Five of the patients were described before [76,105,116]. A duplication of at least MECP2 and IRAK1 was confirmed in all patients enrolled by array-based Comparative Genomic Hybridization (array-CGH) prior to this study. We standardized the ranges of the duplications to Genome Reference Consortium Human Build 37 (GRCh37) by the NCBI Genome Remapping Service to compare the duplication size of all patients.

Material
Fibroblasts of P1, P2, P3, and P4 as well as of 4 healthy individuals were obtained by skin biopsies and immortalized by simian virus (SV40) as described previously [76,117]. Blood samples of P3, P5, P6, P7, P8, and P9 as well as of healthy controls were acquired in parallel to routine blood tests. As P1 and P2 deceased, and we were not able to contact P4 recently, we were not able to obtain current blood samples from P1, P2, and P4. P5-9 did not donate fibroblasts. We isolated the PBMCs and performed the analysis in our laboratory with the same methods and equipment.

Graphs and Statistical Information
Graphs were created using GraphPad Prism 9 software (GraphPad Software Inc.) and PowerPoint (Microsoft Office). Statistical analyses were performed using SPSS V28.0.1.0 (IBM). Data sets were tested for normal distribution, and statistical comparisons were done using a Mann-Whitney U test. For comparison of multiple groups, Kruskal-Wallis test was used. P values of less than 0.05 after adjusting by Bonferroni method were considered significant. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Patients with MECP2/IRAK1 Duplication Suffer from Recurrent Respiratory Infections
Our 9 patients show duplications of variable sizes at least encompassing the neighboring genes MECP2 and IRAK1 (Fig. 2). In patients 5 and 8, part of the region is triplicated. The exact boundaries and the included genes are shown in the Supplementary Information. All patients suffered from Patient 3 is a 25-year-old patient who is followed-up in our department at least 4 times a year and was clinically and molecularly characterized before [P1 in 76,105]. Array comparative genome hybridization (array-CGH) confirmed a duplication of 1.1 Mb at Xq28. He first presented with global developmental delay, muscular hypotonia, and spastic tetra paresis. He suffers from epilepsy and recurrent severe infections. Of his in total 64 episodes of pneumonia, he had developed 47 until his 14th y/a. After starting an immunoglobulin substitution (at 12 y/a) as well as supportive measures and antibiotic prophylaxis (at 14 y/a), the frequency of infections declined, leading to 3 episodes of pneumonia only in the following 8 years. Despite this treatment, we recorded an increase of hospital admissions due to infections in the last 2 years including 12 episodes of pneumonia and 2 episodes of sepsis (Fig. S1). Throughout the last years, the boy developed chronic aspiration and shows bronchiectasis in his latest CT scans (Fig. S2). We are now detecting opportunistic pathogens such as a multidrug-resistant Citrobacter freundii as well as Candida glabrata and Trichosporon asahii in bronchoalveolar lavages. He is currently under prophylactic anti-infective treatment with cotrimoxazole, penicillin, and fluconazole. In all infectious episodes, our patients presented with fever above 39 °C and high CRP levels, typically above > 100 mg/dl, already during the first 3 days of the infection. The boy shows a normal total immunoglobulin titer but deficiency of IgG2, IgG4, IgA, and IgM. A polysaccharide-specific antibody deficiency persisted despite repeated vaccinations.
The baseline clinical features of all patients in our cohort are summarized in Table 2. Detailed case reports of P1, P2, and P4-P9 are provided in the Supplementary Information.

IRAK1 Duplication Leads to Increased Protein Levels in Patient-Derived Fibroblast Cell Lines
First, we characterized the SV40-immortalized fibroblasts cell lines of both patients and healthy controls for their expression of IRAK1, IRAK4, and GAPDH (Fig. 3). We used an IRAK1-deficient and an IRAK4-deficient cell line as negative controls. The patients' cells (P1-P4) contained at least twice as much IRAK1 as the cells of the healthy controls (C1-C4) (Fig. 3). The calculated ratios are stated in the Supplementary Information.

IRAK1 Duplication Does Promote Excessive Cytokine Production Neither in Fibroblasts Nor in PBMC Nor in Whole Blood
We hypothesized that the susceptibility to infection could be caused by a hyperinflammatory immune response due  to increased canonical NF-κB signaling because of IRAK1 overexpression. Therefore, we determined the impact of the MECP2 and IRAK1 duplication on the canonical NF-κB signaling. Hence, we performed an ELISA to measure the cytokine production in the cell culture supernatants of SV40immortalized fibroblasts, PBMCs, and whole blood upon stimulation with IL-1β as well as the TLR agonists LPS (TLR4), PAM 2 CSK 4 (TLR2/6), and Poly(I:C) (TLR3). We used TNF-α and PMA/Ionomycin as NF-κB-independent intra-assay controls. Production of IL-6 and IL-8 upon stimulation with IL-1β or TLR2/6 agonist PAM 2 CSK 4 was increased in fibroblasts of both healthy controls and MECP2/IRAK1-duplicated patients, but we did not see a difference between the two groups ( Fig. 4a and 4b). Interestingly and in contrast to our hypothesis, the data suggests that the cytokine production upon stimulation with TLR4 agonist (LPS) in immortalized fibroblasts of MECP2/IRAK1-duplicated patients is lower than in healthy controls. In IRAK4-deficient fibroblasts, cytokine production upon stimulation with IL-1β, TLR4 agonist LPS, and TLR2/6 agonist PAM 2 CSK 4 was absent ( Fig. 4a and 4b).
These results indicate that MECP2/IRAK1 duplication does not lead to a higher amount of inflammatory cytokines upon stimulation in immortalized fibroblasts, PBMCs, and whole blood.

Discussion
Although research on MECP2/IRAK1 duplication syndrome has increased, a comprehensive pathophysiological mechanism that explains the frequency and severity of infections, the most common cause of death, remains unknown. Numerous publications describe patients who repeatedly require hospitalization, invasive ventilation, and intensive care admission [63,75,76,103,105]. In P3, pneumococcal immunization and antibiotic prophylaxis reduced the number of infections per year drastically for many years (Fig. S1). However, after successful long-term prophylaxis on antibiotics and IgG, he has been presenting multiple times with pneumonia caused by multidrug-resistant and rare pathogens since the age of 23. Patients like these show that the control of infections clearly is still an unmet clinical need.
IRAK1 participates in multiple IL-1 and TLR-driven signaling processes that regulate immunity and inflammation [108][109][110][111][112][113]. Therefore, we hypothesized that the infections may be triggered by a strong acute phase response due to IRAK1 overexpression and subsequently increased canonical NF-κB signaling. However, in our cohort, we did not see any evidence of increased IRAK1-dependent degradation of IκBα. We demonstrated that the production of proinflammatory cytokines IL-6 and IL-8 upon stimulation with IL-1β and TLR2/6 agonist PAM 2 CSK 4 is similar in immortalized fibroblasts as well as PBMCs and whole blood of patients with MECP2/IRAK1 duplication and healthy controls. Also, we did not see an enhanced response upon stimulation with TLR4 agonist LPS in PBMCs and whole blood of patients compared to healthy controls. The results in our healthy controls as well as our IRAK1-and IRAK4-deficient controls were similar to the results of Della Mina et al. [115]. Response to IL-1β and TLR agonists seems to be normal not only in PBMCs but also whole blood which suggests that canonical NF-κB signaling is also neither increased nor impaired in neutrophilic granulocytes of patients with MECP2/IRAK1 duplication.
The NF-κB signaling in fibroblasts and blood seems to be unimpaired. However, this might be different in other tissues such as lung epithelia. On the one hand, IRAK-1 was shown to be essential for IL-8 production in human airway epithelial cells [120]. On the other hand, IRAK-1 is necessary for the rhinovirus-stimulated induction of CXCL-10 in airway epithelial cells and macrophages [121]. Both excessive production of IL-8 and CXCL-10 could contribute to lung inflammation leading to the clinical phenotype of MDS patients. From a scientific point of view, it seems interesting to study the cytokine production and CXCL-10 induction in airway epithelial cells of patients with MECP2/IRAK1 duplication. However, it seems almost impossible to obtain sufficient amounts of primary lung tissue from children with such a rare disease in a standardized way, let alone enough to culture lung epithelia. An alternative strategy to investigate the role of IRAK1 in lung epithelia might be to differentiate human-induced pluripotent stem cells (hiPSCs) to lung epithelial cells [122].
Yang et al. proposed that severe infections in MDS patients occur due to the lack of TH1 response and subsequently low IFN-γ activity [106]. However, a generally impaired IFN-γ secretion could not be reproduced by Bauer et al. [76]. Furthermore, complete IFN-γ deficiency is characterized by a selective predisposition to infections caused by mycobacteria, Salmonella, or Candida species [123,124]. This does not correlate with the clinical phenotype of MDS patients who typically show purulent bronchitis caused by bacteria which are capable of building a capsule such as Streptococcus pneumoniae or Haemophilus influenzae [76]. In the so far published cases of MDS, an infection with mycobacteria was only described once [76].
Besides its role for canonical NF-κB signaling, IRAK1 controls the induction of interferons via interferon regulatory factor 7 (IRF7) [109,111,114]. In human IRF7 deficiency, individuals are selectively susceptible to severe infections by influenza and SARS-CoV-2 and show an impaired type I IFN signature [125,126]. In vitro, IRAK-1 regulates the transcriptional activation of IRF7 by directly binding and phosphorylating it. TLR7-and TLR9-mediated IFNα production is abolished in IRAK1-deficient mice, whereas inflammatory cytokine production is not impaired [111]. This brings up the question whether duplication of the IRAK1 gene and thus IRAK1 overexpression causes an increased activation of the TLR7-and TLR9-mediated interferon-α induction pathway leading to an increased release of interferons and consequently to a hyperinflammatory immune response. However, CD169 expression on monocytes, which is correlated with systemic type I IFN levels, was normal in P3 both while he suffered from an infection and when he was free of infections [127,128]. Further, MECP2-overexpressing mice had been described as particularly susceptible for severe influenza A infection. During infection, they show neutrophilia, increased cytokine production, excessive corticosterone levels, defective adaptive immunity, and vascular pathology. This raises the question if the inflammation-underlying pathomechanism in humans suffering from MECP2 duplication syndrome is rather caused by the overexpression of MECP2 than the overexpression of IRAK1 [107]. In a humanized mouse model of MDS, intracerebroventricular antisense oligonucleotide (ASO) therapy was shown to decrease MECP2 expression in the brain and to reduce behavioral deficits as well as to restore/correct reduced IFN-γ mRNA levels in the blood [129]. If inflammation in MDS is rather caused by the duplication of MECP2 itself, than by duplication of IRAK1, ASO against MECP2 might be a feasible treatment option for these patients. The effects of such ASO therapy, applied in compartments such as the blood and lungs, may also warrant further investigation.
In summary, patients with MECP2 duplication syndrome do not show increased canonical NF-κB signaling in whole blood, PBMCs, or SV40-immortalized fibroblasts. Therefore, we assume that these patients do not benefit from a therapeutic suppression of this pathway.

Consent for Publication
The authors affirm that human research participants' legal guardians provided informed consent for publication of the data of their children.

Competing Interests The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.