The 423Q polymorphism of the X-linked inhibitor of apoptosis gene influences monocyte function and is associated with periodic fever

Abstract

Objective

Hereditary periodic fever syndromes (HPFs) develop as a result of uncontrolled activation of the inflammatory response, with a substantial contribution from interleukin-1β or tumor necrosis factor α (TNFα). The HPFs include familial Mediterranean fever (FMF), hyperimmunoglobulinemia D with periodic fever syndrome (HIDS), TNF receptor–associated syndrome (TRAPS), and cryopyrinopathies, which are attributable to mutations of the MEFV, MVK, TNFRSF1A, and CIAS1 genes, respectively. However, in many patients, the mutated gene has not been determined; therefore, the condition in these patients with an HPF-like clinical picture is referred to as idiopathic periodic fever (IPF). The aim of this study was to assess involvement of X-linked inhibitor of apoptosis (XIAP), which plays a role in caspase inhibition and NF-κB signaling, both of which are processes that influence the development of inflammatory cells.

Methods

The XIAP gene (X-linked) was sequenced in 87 patients with IPF, 46 patients with HPF (13 with HIDS, 17 with TRAPS, and 16 with FMF), and 182 healthy control subjects. The expression of different alleles was evaluated by sequencing XIAP-specific complementary DNA mini-libraries and by real-time polymerase chain reaction and Western blot analyses. The functional effect of XIAP on caspase 9 activity was assessed by a fluorimetric assay, and cytokine secretion was evaluated by enzyme-linked immunosorbent assay.

Results

Sequencing disclosed a 1268A>C variation that caused a Q423P amino acid substitution. The frequency of 423Q-homozygous female patients and 423Q-hemizygous male patients was significantly higher in the IPF group than in the control group (69% versus 51%; odds ratio 2.17, 95% confidence interval 1.23–3.87, P = 0.007), whereas no significant difference was detected in the HPF group (59%) compared with controls. In primary lymphocytes and transfected cell lines, 423Q, as compared with 423P, was associated with higher XIAP protein and messenger RNA expression and lower caspase 9 activation. In lipopolysaccharide-activated monocytes, 423Q was associated with higher secretion of TNFα.

Conclusion

These results suggest that 423Q is a predisposing factor for IPF development, possibly through its influence on monocyte function.

Hereditary periodic fever syndromes (HPFs) are a family of genetic autoinflammatory diseases that share the trait of recurrent and apparently unexplainable episodes of inflammation characterized by fever and serosal, synovial, and/or cutaneous inflammation (1–3). The HPFs comprise familial Mediterranean fever (FMF), attributable to mutations of the MEFV gene (also known as pyrin or marenostrin) (4), hyperimmunoglobulinemia D with periodic fever syndrome (HIDS), caused by mutations in the gene for mevalonate kinase (MVK) (5, 6), tumor necrosis factor (TNF) receptor–associated syndrome (TRAPS), attributable to mutations in the gene for type I TNF receptor (TNFRI) (7), and 3 cryopyrinopathies, Muckle-Wells syndrome, familial cold-induced autoinflammatory syndrome, and chronic infantile neurologic, cutaneous, articular syndrome, attributable to mutations in the CIAS1 gene encoding cryopyrin (8, 9). Transmission is autosomal recessive in FMF and HIDS, and is dominant in TRAPS and the cryopyrinopathies.

Some patients display an HPF-like syndrome but lack mutations in the HPF genes. These idiopathic periodic fevers (IPFs) could theoretically be caused by mutations in other, unknown causal genes. Alternatively, IPFs may be the outcome of oligogenic variations, each incapable of causing an IPF on its own. These variations might influence the often heterogeneous clinical picture. For instance, symptoms typical of the periodic fever, aphthous stomatitis, pharyngitis, cervical adenitis (PFAPA) syndrome are often found in patients with various conditions characterized by periodic fever (10–12).

The genes involved in HPF encode for proteins believed to modulate the functions of interleukin-1β (IL-1β) and TNFα. Pyrin and cryopyrin influence formation and/or function of the inflammasome, a multimolecular complex involved in activation of caspase 1 and subsequent cleavage of proIL-1β to active IL-1β (13, 14). The mechanism is less clear for MVK deficiency, which seems to cause isoprenoid end-product shortage, a condition that is possibly responsible for increased IL-1β secretion by mononuclear cells (15). The pathogenesis of TRAPS seems to be related primarily to an altered response of TNFRI to TNFα, which can trigger 2 distinct signaling pathways, leading to either cell activation or apoptosis (16–18). On the one hand, TNFRI recruits TRADD and triggers an activation pathway, leading to nuclear translocation of the NF-κB and activator protein 1 transcription factors, which then stimulates inflammation (18, 19). In TRAPS, this pathway may be hyperactive because of the presence of TNFRI gene mutations affecting TNFRI shedding from the membrane, which serves as a negative feedback mechanism to block the functions of TNFα and thus down-modulate inflammation (20). On the other hand, TNFRI recruits FADD and triggers activation of a caspase cascade, leading to cell apoptosis (20). In TRAPS, defective functioning of this pathway may affect immune cell apoptosis, another mechanism that down-modulates inflammation (21, 22). It is intriguing that a lymphocyte apoptosis defect has also been associated with MVK deficiency, but its mechanism of action has yet to be determined (23).

X-linked inhibitor of apoptosis (XIAP) belongs to the family of inhibitors of apoptosis (IAPs), comprising several proteins that inhibit caspase activity (24–26). The IAPs display 1–3 baculovirus IAP repeat (BIR) domains (27), and several of them also display a RING-finger domain with E3 ubiquitin ligase activity (28). XIAP has 3 BIR domains and the RING domain and inhibits activation of procaspase 9 and activated caspases 3 and 7 (29–31). Inhibition of procaspase 9 activation involves BIR-3, whereas inhibition of caspases 3 and 7 involves the linker region between BIR-1 and BIR-2 (32, 33). A second function of XIAP is to modulate cell activation by inducing NF-κB and MAPK activation and by inhibiting the JNK cascade (34, 35). This function involves the BIR-1 domain interacting with the Tak-1 binding protein 1 adaptor, leading to activation of transforming growth factor β–activated kinase and NF-κB (36, 37). The RING domain also may play a role in cell activation, by inducing proteasome-mediated degradation of key regulatory proteins.

This dual role of XIAP in apoptosis and activation of immune cells prompted our search for variations in the XIAP gene, located in the X chromosome, in patients with IPF. We found a missense polymorphism, Q423P, whose allelic frequency differed significantly between patients with IPF and control subjects. Intriguingly, the IPF-predisposing allele, 423Q, was associated with increased expression of XIAP protein and messenger RNA (mRNA), decreased activation of caspase 9, and increased TNFα secretion by monocytes in response to lipopolysaccharide (LPS) stimulation in healthy control subjects. These results suggest that 423Q is a predisposing factor in the development of IPF, possibly by favoring monocyte activation.

PATIENTS AND METHODS

Patients.

Since 2002, a nationwide laboratory facility established at the G. Gaslini Institute in Genoa, Italy has been in operation for the genetic diagnosis of autoinflammatory disorders in children with periodic fever. Physicians submitting patients' biologic samples are also asked to fill in a form, to provide detailed information on the patient's family and personal history, and to describe the clinical features associated with the patient's episodes of fever. The study enrolled 87 children with IPF and 46 with HPF (comprising 13 with HIDS, being homozygous or compound heterozygous for MVK mutations, 16 with FMF, being homozygous or compound heterozygous for MEFV mutations, and 17 with TRAPS, being heterozygous for TNFRSF1A mutations) (Table 1). The inclusion criteria were as follows: 1) periodic fever attacks of unknown origin (body temperature >38°C, with fever, and symptom-free intervals characterized by normal levels of acute-phase reactants) for a period of more than 6 months, and 2) at least 2 of the following symptoms during fever attacks: lymphadenopathy, splenomegaly, gastrointestinal manifestations, chest pain, mucocutaneous manifestations, and musculoskeletal manifestations. All patients were Italian. Age-matched and ethnically matched control subjects were children without autoimmune or allergic diseases who had visited the general practitioner outpatient clinic at the G. Gaslini Institute in Genoa.

Table 1. Causative mutations in patients with hereditary periodic fevers (HPFs)*

	HIDS (n = 13)	TRAPS (n = 17)	FMF (n = 16)	All HPFs (n = 46)
* HIDS = hyperimmunoglobulinemia D with periodic fever syndrome; TRAPS = tumor necrosis factor receptor–associated syndrome; FMF = familial Mediterranean fever. † By chi-square test.
Sex, no. male/no. female	7/6	10/7	9/7	26/20
XIAP allele, no. (%)
423Q	12 (63.2)	15 (62.5)	16 (69.6)	43 (65.2)
423P	7 (36.8)	9 (37.5)	7 (30.4)	23 (34.8)
P versus healthy controls†	0.87	0.84	0.64	0.77
Target gene	MVK	TNFRSF1A	MEFV	–
Causal mutation	V377I/V377I, I268T/V377I, L265R/V377I, V250I/G376V, G171R/I268T, V132I/V377I, G211E/V377I, H20Q/V377I, I119M/A148T	D12E, C43R, T50M, C52Y, C55Y, C88Y, c.194-14 G>A, R92Q	M680I/M680I, E148Q/M694V, E148Q/A7445, M680I/R761H, M694V/E148Q, E225G/M680I, L110P/L110P, E148Q/R761H, M694V/M680I, M680I/V726A	–

Informed consent was provided by all participants, in accordance with the 1995 Declaration of Helsinki. The research was approved by the ethics committee of the Maggiore Hospital of Novara and G. Gaslini Institute.

Sequence analysis.

The extracellular region of the p55 TNF receptor (exons 1–6) of the TNFRSF1A gene, the 10 coding exons (exons 2–11) of the MVK gene, and exons 2, 3, 5, and 10 of the MEFV gene (i.e., exons known to harbor most mutations) were sequenced as previously described (6, 21). In patients heterozygous for MEFV mutations in these exons, exons 1, 4, 6, 7, 8, and 9 were also sequenced to look for rare mutations. Exons 2–7 of XIAP were amplified by polymerase chain reaction (PCR), and the PCR products were sequenced with the ABI Prism BigDyeTM Terminator kit (Applied Biosystems, Foster City, CA), on an automatic sequencer, using the same primers.

Western blot analysis.

Peripheral blood mononuclear cells (PBMCs) were activated with phytohemagglutinin (PHA) (1 μg/ml) and cultured in RPMI 1640 plus 10% fetal calf serum (FCS) and recombinant IL-2 (10 units/ml) for 6 days; this mixture was found to create the optimal conditions to detect XIAP expression. The cells were then washed twice with phosphate buffered saline and lysed for 20 minutes on ice in AKT buffer (10 mM NaCl, 10 mM MgCl₂, 10 mM Tris HCl, pH 7.5, 1% Triton X-100, 1% sodium deoxycholate, 1 μg/ml aprotinin, 1 μg/ml leupeptin, 1 μg/ml pepstatin A, and 100 μg/ml phenylmethylsulfonyl fluoride). Lysates were then cleared by centrifugation for 20 minutes at 13,000 revolutions per minute at 4°C and separated on 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) gels after denaturation in SDS-PAGE loading buffer (63 mM Tris HCl, pH 6.8, 5% glycerol, 1% SDS, 2.5% bromophenol blue), and transferred to nitrocellulose. Filters were then blocked in Tris buffered saline–Tween (TBST) buffer plus 5% nonfat milk for 1 hour, and then incubated with anti-XIAP monoclonal antibody (Alexis Axxora, San Diego, CA) overnight at 4°C in TBST buffer plus 5% bovine serum albumin. XIAP signals were revealed with a horseradish peroxidase–conjugated anti-mouse Ig secondary antibody (Amersham, Arlington Heights, IL) and detected by enhanced chemiluminescence. Tubulin was detected with mouse anti-tubulin antibody (Sigma-Aldrich, Milan, IT) and the same secondary antibody. Bands were quantified with the GelDoc EQ system.

Plasmid expression vectors and in vitro transfection.

PCR was used to amplify the 2 complementary DNA (cDNA) forms of XIAP, 1268A and 1268C. With the use of gene-cloning techniques, the 2 XIAP forms were inserted into a pEGFP expression vector plasmid to form 2 eukaryotic green fluorescent protein (GFP)–fusion expression vectors (GFP-423Q and GFP-423P). The 2 vectors and an empty pEGFP vector as mock control were transfected into 293T cells by lipofectamine mediation. Twenty-four hours after transfection, expression of the GFP fusion proteins in the transfected cells was examined by confocal microscopy and by Western blotting with an anti-GFP antibody (Sigma-Aldrich), and the lysates were also used for assay of caspase 9 activity.

Analysis of caspase 9 activity.

Activated PBMCs (6 × 10⁶) were left untreated or were treated with etoposide (5 μg/ml) on ice for 30 minutes, and then moved to an atmosphere of 37°C for 6 hours and centrifuged. Caspase 9 activity was assessed in cell lysates using a fluorimetric assay (MBL, Watertown, MA). At least 3 control samples from healthy donors were always run in parallel. Results were expressed as the relative caspase activity (in %), calculated as (activity of etoposide-stimulated cells/activity of unstimulated cells) × 100.

Quantitative reverse transcription–PCR (RT-PCR).

Total mRNA was isolated from the PBMCs and then cultured for 6 days in RPMI plus 1 μg/ml PHA and 10 units/ml IL-2 using NucleoSpin RNAII (Macherey-Nagel, Duren, Germany). Five hundred nanograms of total mRNA was reverse transcribed using the ThermoScript RT-PCR System kit (Invitrogen, Carlsbad, CA). The TaqMan probe for XIAP (assay no. Hs00236913_m1) and the probe for GUSB (assay no. Hs99999908_m1) were purchased from Applied Biosystems. All RT-PCR steps were performed on an ABI Prism 7000 Sequence Detector (Applied Biosystems), and relative mRNA expression was quantified using the cycle threshold method, normalized to the values for GUSB mRNA.

Cytokine assay.

Monocytes were isolated from PBMCs with the Monocyte Isolation Kit II (Miltenyi Biotec, Sunnyvale, CA) and cultured (5 × 10⁴ cells/100 μl) for 3 hours in RPMI 1640 plus 10% FCS in the presence or absence of LPS (1 μg/ml); to induce IL-1β secretion, ATP (1 mM) was added to the cultures in the last 15 minutes (38). Supernatants were then collected to measure the secretion of TNFα and IL-1β by enzyme-linked immunosorbent assay (ELISA) (Quantikine; R&D Systems, Minneapolis, MN).

Statistical analysis.

To study the relationship between clinical condition and genotype, subjects were classified into 2 groups: Q-genotypes (QQ females and Q males), and P-genotypes (PP and PQ females, and P males). The chi-square test was used to assess genotype heterogeneity among clinical groups and to make post hoc comparisons with the healthy subjects. Functional data were analyzed with the nonparametric Mann-Whitney U test.

RESULTS

XIAP gene expression in periodic fever syndromes.

The exons and the exon–intron boundaries of XIAP were sequenced in 182 healthy control subjects, 87 patients with IPF, and 46 patients with HPF (13 with HIDS, 17 with TRAPS, and 16 with FMF) (see Table 1 for causative mutations). The criteria for study inclusion are described in Patients and Methods. All patients with IPF displayed recurrent episodes of inflammatory symptoms, including fever, abdominal pain, myalgia with migratory erythematous macular rashes, conjunctivitis and periorbital edema, chest pain, and arthralgia or monarticular synovitis, but none of these patients had mutations in the MEFV, MVK, and TNFRSF1A genes involved in HIDS, TRAPS, and FMF, respectively. All patients and control subjects were Italian, and their geographic origin was similarly distributed in all groups.

We found only 1 missense variation, 1268A>C (rs5956583) located in exon 6 (ATG = +1) (starting code ATG + 1, where underline indicates the target amino acid), that was predicted to cause the Q423P amino acid substitution. The chi-square test for genotype heterogeneity showed that the distribution of the genotypes was unbalanced between patients with IPF, patients with HPF, and control subjects (P = 0.016 between groups) (Table 2). Moreover, 423Q-homozygous females and 423Q-hemizygous males (Q-genotypes) were significantly more frequent in the IPF group than in the control group (69% versus 51%; odds ratio [OR] 2.17, 95% confidence interval [95% CI] 1.23–3.87, P = 0.007), whereas no significant difference was detected in the HPF group as compared with controls (59% versus 51%; OR 1.39, 95% CI 0.69–2.82, P = 0.4). The genotype distribution did not deviate significantly from Hardy-Weinberg equilibrium in any group.

Table 2. Frequency distribution of the XIAP 1268A>C polymorphism (Q423P) in patients with IPF, patients with HPF, and controls*

Group				Genotype
	Allele			Females			Males		All
	Q	P	P versus controls†	QQ	QP	PP	Q	P	QQ + Q	QP + PP + P	P versus controls‡
* Values are the number (%) of subjects. The proportion of male subjects and female subjects in each group was as follows: in controls, 56% and 44%, respectively; in patients with idiopathic periodic fever (IPF), 56% and 44%, respectively; in patients with hereditary periodic fever (HPF), 57% and 43%, respectively. † For comparison of total allele frequencies in the patient groups versus control group, by chi-square test. ‡ For comparison of total genotypes in the patient groups versus control group, by chi-square test.
Controls (n = 182)	163 (62)	99 (38)		30 (38)	41 (51)	9 (11)	62 (61)	40 (39)	92 (51)	90 (49)
IPF (n = 87)	97 (78)	28 (22)	0.004	22 (58)	15 (40)	1 (2)	38 (78)	11 (22)	60 (69)	27 (31)	0.007
HPF (n = 46)	43 (65)	23 (35)	0.8	8 (40)	8 (40)	4 (20)	19 (73)	7 (27)	27 (59)	19 (41)	0.4

IPF is likely to comprise multiple pathologic entities that share the characteristic of periodic or recurrent fever, but these entities are heterogeneous in terms of pathogenesis and clinical manifestations. To assess whether the Q423P genotypes are a marker of more homogeneous clinical subgroups, we evaluated whether the Q-genotypes displayed an association with the PFAPA syndrome, since this is a well-known clinical picture and was observed in 47 of 87 of our patients with IPF whose symptoms were in accordance with current clinical criteria (11). Our analyses did not detect any association with the PFAPA syndrome, since the Q-genotypes were carried by 17 (36%) of 47 patients with PFAPA and 10 (25%) of 40 patients without PFAPA (P = 0.26).

Association of Q423P with XIAP levels.

To assess the functional consequences of being a carrier of Q423P, we analyzed the expression of XIAP and activation of caspase 9, which is highly susceptible to XIAP-mediated inhibition, in healthy subjects carrying different Q423P genotypes. PHA-activated PBMCs obtained from these subjects were cultured for 6 days in the presence of IL-2.

The expression of XIAP was compared between 423Q- and 423P-hemizygous males, by Western blot analysis of whole cell lysates from each group. The results showed that the 423Q carriers (n = 9) displayed higher XIAP levels than did the 423P carriers (n = 8). Densitometric analysis of the XIAP bands, normalized to the corresponding tubulin bands, showed that the difference in XIAP expression between 423Q and 423P carriers was significant (P = 0.004), and expression of the 423Q variant was almost 2-fold higher than that of the 423P variant (Figures 1A and B).

Figure 1. X-linked inhibitor of apoptosis (XIAP) expression and caspase 9 activity in phytohemagglutinin-activated peripheral blood mononuclear cells derived from healthy subjects carrying 423Q or 423P. A, XIAP expression was analyzed by Western blotting of cell lysates from 9 male 423Q carriers and 8 male 423P carriers; results were normalized to those for tubulin. Lanes marked 1–17 correspond to samples from each of the 17 different donors; note that some subjects were analyzed twice. B, The results of Western blotting for XIAP expression were further analyzed by densitometry, where XIAP expression was normalized to the values for tubulin; the median XIAP expression detected in all samples run in each experiment was 100%. C, The relative amount of 423Q and 423P XIAP mRNA in 4 heterozygous females was quantified by sequencing 153 cDNA clones from XIAP-specific cDNA mini-libraries; XIAP mRNA expression in the total number of clones (range 20–66) analyzed in each subject was 100%. D, The amount of 423Q and 423P XIAP mRNA was quantified in 5 male 423Q carriers and 6 male 423P carriers by reverse transcription–polymerase chain reaction; the median XIAP mRNA expression level detected in 423P carriers was 100%. E, Etoposide-induced activation of caspase 9 was assessed in 5 male 423Q carriers and 5 male 423P carriers; the median caspase 9 activity induced in all samples run in each experiment was 100%. In B–E, horizontal lines show the median values, and boxes represent the 25th and 75th percentiles. Statistical analyses were performed with the nonparametric Mann-Whitney U test.

Expression of XIAP mRNA was initially assessed by producing XIAP-specific mini-libraries from the cDNA of Q423P-heterozygous females (n = 4). Sequencing of 153 independent cDNA clones (range for each mini-library, 20–66 cDNA clones) showed that the proportion of 423Q clones was higher (P = 0.004) than that of 423P clones in all subjects (Figure 1C). Since this difference could be influenced by cell selection, we quantified the XIAP mRNA by RT-PCR in 423Q- and 423P-hemizygous males. The results showed that 423Q carriers (n = 5) displayed higher levels of XIAP mRNA than did 423P carriers (n = 6) (P = 0.05) (Figure 1D).

Caspase 9 activity was evaluated by a fluorimetric enzyme assay on whole lysates of cells that were left untreated or were treated with etoposide to activate caspase 9 in 423Q- and 423P-hemizygous males. As shown in Figure 1E, 423Q carriers (n = 5) displayed significantly lower levels of etoposide-induced caspase 9 activity than did 423P carriers (n = 5) (P = 0.02).

These findings suggest that the Q423P variation influences XIAP expression and thus affects caspase 9 activity. To confirm this possibility, we cloned the 423Q and 423P XIAP cDNA into the pEGFP expression vector fused to GFP, to produce the GFP-423Q and GFP-423P constructs, which were then transfected into 293T cells; the pEGFP empty (mock) vector was used as a control. Twenty-four hours after transfection, construct expression was evaluated on whole cells by confocal microscopy to assess GFP fluorescence, and on cell lysates by Western blotting with an anti-GFP antibody. Moreover, caspase 9 activity was assessed on the cell lysates by a fluorimetric enzyme assay. Both confocal microscopy and Western blotting showed that expression of GFP-423Q was higher than that of GFP-423P. Moreover, expression of GFP-423Q was similar to that of GFP-mock (Figures 2A and B). Analysis of caspase 9 activity showed that GFP-423P–transfected cells displayed a significantly higher activity than did cells transfected with GFP-423Q (P = 0.021), and the caspase 9 activity in the GFP-423Q–transfected cells was similar to that displayed by cells transfected with GFP-mock (Figure 2C).

Figure 2. Expression of the green fluorescent protein (GFP)–fusion constructs with XIAP-Q and XIAP-P, and extent of caspase 9 activity after transfection of 293T cells with the GFP-XIAP-Q or GFP-XIAP-P constructs. A, GFP expression in the cells transfected with the XIAP-Q and XIAP-P plasmids was analyzed by Western blotting, with results normalized to tubulin expression. Representative results from 1 of 4 experiments are shown. B, GFP expression by 293T cells after transfection was analyzed by confocal microscopy. Representative results from 1 of 2 experiments are shown. In A and B, a GFP-mock construct was used as the control. C, Caspase 9 activity was evaluated 12 hours after transfection in 293T cells transfected with the XIAP-Q or XIAP-P plasmids. Bars show the mean and SEM caspase 9 activity relative to that displayed by mock-transfected cells in 8 experiments. The horizontal line indicates the activity detected in mock-transfected cells (100%). Statistical analyses were performed with the nonparametric Mann-Whitney U test.

Correlation of Q423P with monocyte function.

Since monocytes are key inflammatory cells and are likely to play an important role in the systemic inflammatory response of periodic fevers, we assessed whether the Q423P variation influences the function of these cells by assessing secretion of IL-1β and TNFα in monocytes from healthy males who were hemizygous for 423Q (n = 9) or 423P (n = 9). To induce TNFα secretion, monocytes were stimulated with LPS for 3 hours, whereas to induce IL-1β secretion, these LPS-activated monocytes were treated with ATP for a further 15 minutes (38). TNFα and IL-1β levels in the culture supernatants were then measured by ELISA. The results showed that LPS-activated monocytes from 423Q donors produced higher levels of TNFα (2-fold increase) compared with LPS-activated monocytes from 423P donors (P = 0.03). In contrast, IL-1β secretion was not different between the 2 donor groups in any condition (Figure 3).

Figure 3. Tumor necrosis factor α (TNFα) and interleukin-1β (IL-1β) secretion by monocytes derived from healthy males carrying either 423Q (XIAP-Q) or 423P (XIAP-P). Left, Secretion of TNFα by monocytes in medium alone or stimulated with lipopolysaccharide (LPS) for 3 hours. Right, Secretion of IL-1β by monocytes stimulated with LPS for 3 hours in the presence or absence of ATP in the last 15 minutes of culture. Bars show the mean and SEM results. Statistical analyses were performed with the nonparametric Mann-Whitney U test.

DISCUSSION

The results of the present study show that a common missense polymorphism of the XIAP gene (423Q) located in the linker region between the BIR-3 and RING domains may be a predisposing factor for the development of periodic fever in the absence of mutations of the genes known to cause HPFs. Moreover, the findings suggest that this predisposing effect is due to increased expression of XIAP, which may influence secretion of TNFα by monocytes.

The first finding was that the frequency of 423Q-homozygous females and 423Q-hemizygous males was increased in the group of patients with IPF, and that the presence of the Q-genotypes increased the risk of IPF by ∼2-fold. In contrast, these genotypes did not seem to be associated with FMF, HIDS, and TRAPS, but these results are not conclusive, since the number of patients was small in each group, and genotype frequencies were only moderately different between patients with IPF and the control group.

IPF accounts for the majority of periodic fevers, since fewer than 20% of these patients display mutations of the MVK, TNFRSF1A, pyrin, or cryopyrin genes. IPF is likely to comprise a heterogeneous group of diseases that share periodic or recurrent fever as a feature, but the disease development may be attributable to partly different pathogenic mechanisms. A possibility is that 423Q acts in collaboration with causal mutations of unknown genes driving IPF development. Alternatively, 423Q may participate in an oligogenic background in which each predisposing gene has a small effect on IPF development.

Q423P was not associated with the PFAPA syndrome as defined by the current diagnostic criteria (39), but this is not surprising, since the ability to identify a homogeneous subgroup of patients has been questioned. Moreover, PFAPA is also displayed by a substantial proportion of patients with HPF (especially those with MVK mutations) (40).

The second finding was that 423Q was associated with higher levels of XIAP than was 423P, which was supported by a series of observations. First, expression levels of the XIAP protein and mRNA were higher in healthy males who were hemizygous for 423Q than in healthy males hemizygous for 423P. Second, expression of 423Q mRNA was higher than that of 423P mRNA in healthy heterozygous females. Third, 293T cells transfected with 423Q cDNA expressed higher XIAP levels than did those transfected with 423P cDNA. These data suggest that different RNA stability or processing may be responsible for the different expression of the 2 variants. This significantly influenced XIAP activity, since 423Q hemizygotes and 423Q-transfected cell lines displayed lower etoposide-induced activation of caspase 9, the main target of the anticaspase activity of XIAP, as compared with their 423P counterparts. Moreover, in silico analysis with Polyphen software (http://genetics.bwh.harvard.edu/pph/) suggested that Q423P also affects XIAP function by altering its structure (results not shown).

The third finding was that these differences may influence the inflammatory response, since LPS-activated monocytes from male 423Q carriers produced 2-fold higher levels of TNFα than did those from male 423P carriers. The possibility that TNFα plays a role in the pathogenesis of IPF is in line with the key role ascribed to this cytokine in TRAPS.

In 423Q carriers, increased XIAP expression may favor TNFα secretion in response to activating stimuli in 2 non–mutually exclusive manners. On the one hand, it may inhibit caspase activity and thus diminish immune cell apoptosis, which is a mechanism involved in switching off the inflammatory response. On the other hand, it may support the NF-κB and mitogen-activated protein kinase pathways and thus favor inflammatory cell activation. It is noteworthy that in both TRAPS and HIDS, mutations of the causal gene have also been suggested to exert a dual effect on apoptosis and activation of inflammatory cells.

In summary, this study shows that increased expression of XIAP associated with a common polymorphism of the XIAP gene may favor development of an autoinflammatory syndrome, possibly by modulating secretion of TNFα. It is noteworthy that this is the second report that indicates an association of XIAP variations influencing the expression of XIAP with an inherited immunologic disorder. Previously, Rigaud et al (41) described deleterious XIAP mutations that were found to severely affect XIAP expression in male patients displaying an X-linked lymphoproliferative syndrome, which was characterized by a defective response to Epstein-Barr virus infection, lymphohistiocytosis, and hepato- and/or splenomegaly, as well as decreased counts of natural killer T cells, with or without hypogammaglobulinemia. It is intriguing that defective expression of XIAP can cause immunodeficiency, whereas in the present study, we have shown that increased expression of XIAP may be a predisposing factor in the development of a hyperimmune disease.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Dianzani had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Gattorno, Chiocchetti, Ceccherini, Martini, Dianzani.

Acquisition of data. Ferretti, Mesturini, Orilieri, Bensi, Cappellano, Cerutti, Biava, Bardelli.

Analysis and interpretation of data. Sormani, Nicola, Federici, Baldi, Santoro, Dianzani.