Dr. Gül is recipient of a Young Scientist award (EA-TUBA-GEBIP/2001-1-1) from the Turkish Academy of Sciences.
Familial Mediterranean fever (FMF) is associated with more than 70 missense mutations in the MEFV gene. The purpose of this study was to investigate the relative expression of messenger RNA (mRNA) for the MEFV gene in peripheral blood leukocytes (PBLs) obtained from patients with FMF during attacks of acute abdominal inflammation as well as during asymptomatic periods.
We studied 16 patients with FMF during an attack of acute peritonitis and 17 otherwise healthy individuals who were undergoing surgery because of acute appendicitis. Blood samples were collected from both groups of patients during both acute inflammatory and asymptomatic periods. Relative levels of MEFV mRNA in PBLs were detected with real-time reverse transcriptase–polymerase chain reaction using LightCycler, with 2 sets of primers for the MEFV gene (exons 7–10 and exons 2–3) and with primers for CIAS1 and PSTPIP1 genes. Expression levels were compared with β2-microglobulin as an internal control.
MEFV expression was reduced in FMF patients during asymptomatic periods as compared with the non-FMF controls (P < 0.001). We observed a further decrease in MEFV expression in FMF patients during periods of inflammation (P = 0.01). Reduced levels of MEFV mRNA were also noted during the preoperative period as compared with asymptomatic periods in control patients with acute appendicitis (P = 0.01). CIAS1 expression in PBLs from patients with FMF was also found to be lower than that in the control patients. However, CIAS1 expression did not change with acute inflammation.
This study confirmed that reduced expression of the MEFV gene is associated with inflammation and that it may be one of the pathogenic mechanisms of the attacks of inflammation in FMF patients, along with disease-associated variations in pyrin.
Familial Mediterranean fever (FMF; OMIM no. #249100), the most common type of hereditary autoinflammatory syndrome, is characterized by recurrent self-limited attacks of fever, inflammation of the serosal membranes, arthritis, and skin rashes, as well as an increased risk of developing AA amyloidosis (1). FMF is seen most frequently in eastern Mediterranean populations, mainly in people of Jewish, Armenian, Turkish, and Arab ancestry, and it is inherited as an autosomal-recessive disorder, with some exceptions (1).
Most FMF patients carry mutations in the MEFV gene, which encodes the pyrin protein (2, 3). Pyrin, a 781–amino acid protein expressed in granulocytes, monocytes, dendritic cells, and synovial, peritoneal, and skin-derived fibroblasts, has been suggested to play an important role in innate immunity through regulation of interleukin-1β (IL-1β) processing, NF-κB activation, and apoptosis (4). More than 70 FMF-associated pyrin mutations have thus far been described (http://fmf.igh.cnrs.fr) (5). Almost all of these genomic variations are conservative missense mutations, and their pathogenic mechanism has yet to be identified. The majority of mutations are accumulated in exon 10, which encodes the carboxy-terminal B30.2, or PRYSPRY, domain (6–9). The recently discovered structure of the SPRY domain suggests that at least some of the pyrin mutations map to the loop regions of the β-sandwich structure and may disrupt a SPRY domain–mediated protein–protein interaction, including an interaction with caspase 1 (7–10).
A mouse model expressing a hypomorphic variant of the pyrin protein, which was developed by targeted disruption of the carboxy-terminal of the MEFV gene, showed increased sensitivity to bacterial lipopolysaccharide (LPS) in homozygous mutants (11). The expression of pyrin in macrophages of homozygous mutant mice was reported to be reduced, as compared with the expression in wild-type littermates, following stimulation with LPS and IL-4 (11). Reduced expression of MEFV messenger RNA (mRNA) in peripheral blood leukocytes (PBLs) from FMF patients obtained during attack-free periods, as compared with the expression in PBLs from healthy individuals, has also been reported (12).
In this study, we investigated the relative mRNA expression of the MEFV gene in PBLs obtained from patients with FMF during attacks of acute abdominal inflammation and during asymptomatic periods. We compared the results with those in PBLs obtained during periods of acute inflammation and asymptomatic periods in a group of non-FMF patients who were undergoing surgery because of acute appendicitis.
PATIENTS AND METHODS
The study group consisted of 16 patients with FMF (8 male and 8 female) who were examined during an attack of acute abdominal inflammation. All FMF patients fulfilled the proposed clinical criteria for classification (13). Their mean ± SD age was 31.6 ± 12.1, and their mean ± SD disease duration was 11.6 ± 9.8 years. All but 1 of the patients was receiving colchicine treatment. The non-FMF patient control group consisted of 17 otherwise healthy individuals (12 male and 5 female) who were undergoing surgery because of acute appendicitis, as well as 3 patients with sepsis (2 male and 1 female). The mean ± SD age of the patients with acute appendicitis was 36.4 ± 17.5 years, and the mean ± SD age of the patients with sepsis was 51.3 ± 11.0 years.
We collected venous blood samples from all study subjects during the period of acute inflammation (inflammation) and following the resolution of all findings of inflammation (asymptomatic). In patients with sepsis, we collected 1 sample during the period when they fulfilled the clinical criteria for sepsis. One of the authors (FS) examined all patients within 48 hours after the onset of the attack of acute abdominal inflammation, during the time when patients had both clinical features of acute abdominal inflammation and laboratory features of an acute-phase response. A standard form was used to record all available clinical and laboratory findings at the time of blood collection. A detailed personal and family medical history specific for FMF was obtained from all non-FMF patients in order to exclude those who themselves or whose first-degree relatives had undiagnosed FMF.
The Ethics Committee of the Istanbul Faculty of Medicine approved the study, and written informed consent was obtained from all subjects prior to blood collection.
Gene expression analyses.
We isolated total RNA from PBLs using the RNeasy Mini kit (Qiagen, Hilden, Germany) and synthesized complementary DNA using 1 μg of total RNA, random hexamer primers, and reverse transcriptase (Fermentas, St. Leon-Rot, Germany). We measured the concentrations of mRNA in PBLs with a real-time reverse transcriptase–polymerase chain reaction (RT-PCR) method using LightCycler 2.0 (Roche Diagnostics, Mannheim, Germany). For amplification of MEFV mRNA, we used the primers described by Notarnicola and colleagues (12), which target exons 7–10 of the MEFV gene, as well as a new set of primers that target exons 2–3 (Table 1). We also analyzed the expression of CIAS1 and PSTPIP1 genes, both of which are involved in the same inflammatory pathway and are associated with other autoinflammatory disorders (1). The primers used for CIAS1 and PSTPIP1 are shown in Table 1. Expression levels of these genes were compared with the expression of B2M, the gene for β2-microglobulin, which was used as an internal control. We calculated relative expression units using the following formula: relative expression units = (expression of the target gene/expression of B2M) × 100. The results are expressed as the mean ± SD relative expression units, and the median values are also given.
Table 1. Primers used in the polymerase chain reaction expression analyses*
MEFV (exons 2–3)
MEFV (exons 7–10)
MEFV mutation analyses.
MEFV mutation analysis results were available for 4 of the FMF patients. All of the remaining patients who consented to undergo mutation analysis were screened for 5 MEFV gene mutations, using a PCR–restriction fragment length polymorphism method for M694V (14), M680I, V726A, and E148Q and an allele-specific oligonucleotide method for M694I.
For each of the genes analyzed, we compared the relative expression units obtained during inflammation with those obtained during the asymptomatic period using the nonparametric Wilcoxon's signed rank test. We used the nonparametric Mann-Whitney U test for comparison of the relative expression units between the patients. P values less than 0.05 were considered significant.
Samples were collected a mean ± SD of 5.6 ± 3.5 hours (range 1–12), 9.0 ± 2.4 hours (range 4–12), and 32 ± 8 hours (range 24–40) after the onset of acute inflammation in patients with FMF, acute appendicitis, and sepsis, respectively. During periods of inflammation, the mean ± SD levels of C-reactive protein were 24.4 ± 12.9 mg/liter, 19.5 ± 10.4 mg/liter, and 60.3 ± 14.6 mg/liter, and the mean ± SD white blood cell counts were 13,250 ± 3,068/mm3, 12,218 ± 2,750/mm3, and 17,667 ± 4,163/mm3 in patients with FMF, acute appendicitis, and sepsis, respectively.
Results of MEFV mutation analyses.
In addition to the 4 patients with known genotypes, screening of patients for 5 MEFV gene mutations revealed positive results in 12 patients with FMF (75%), 2 patients with acute appendicitis (11.8%), and 1 patient with sepsis (33%). Six of the FMF patients carried 2 mutations (4 had M694V/M694V, 1 had M680I/M680I, and 1 had M694V/M680I), and 6 FMF patients (3 had M680I, 1 had M694V, 1 had V726A, and 1 had A744S) and 3 non-FMF patients (2 acute appendicitis patients had M694V and 1 sepsis patient had M680I) carried 1 MEFV mutation.
Results of gene expression analyses.
The relative expression units of the MEFV, CIAS1, and PSTPIP1 genes during asymptomatic periods and during inflammation are given in Figure 1. Expression of the MEFV gene was found to be significantly lower in FMF patients during asymptomatic periods, as compared with control patients with acute appendicitis, using both primer sets that amplified MEFV 2–3 (mean ± SD 2.96 ± 1.06 [median 2.96] versus 6.73 ± 2.87 [median 7.08]; P < 0.001) and MEFV 7–10 (1.30 ± 0.87 [median 1.01] versus 2.84 ± 1.77 [median 2.10]; P = 0.004) (Figure 1). The mean expression units for MEFV 2–3 and MEFV 7–10 in the FMF patients were 44% and 46%, respectively, of the mean expression units in the acute appendicitis controls.
We observed a further decrease in MEFV 2–3 expression in FMF patients during attacks of inflammation (1.87 ± 0.78 [median 1.69]) as compared with asymptomatic periods in these patients (P = 0.01). A reduced expression of MEFV 2–3 was also noted during the preoperative period of inflammation in control patients with acute appendicitis (3.67 ± 3.13 [median 2.29]) as compared with asymptomatic periods in these control patients (P = 0.01). However, the expression of MEFV 7–10 during periods of inflammation in patients with FMF (1.28 ± 0.92 [median 1.08]) and in control patients with acute appendicitis (1.97 ± 2.38 [median 1.11]) was not significantly different from the expression during asymptomatic periods in the same group. In the 3 control patients with sepsis, the mean expression of MEFV 2–3 during periods of inflammation (1.26 ± 0.34 [median 1.29]) and MEFV 7–10 (1.28 ± 0.12 [median 1.28]) was lower than the expression in control patients with acute appendicitis during asymptomatic periods (P = 0.002 and P = 0.054, respectively). However, no statistically significant change in MEFV expression was noted in the sepsis patients as compared with the FMF patients or with the acute appendicitis patients during periods of inflammation.
In patients with FMF, similar patterns of expression were found in those with 2, 1, or no MEFV mutations (Table 2). In control patients with acute appendicitis, the 2 patients who were heterozygous for M694V showed an MEFV expression pattern similar to that in the remaining 15 patients during asymptomatic periods. The decrease of MEFV mRNA, using both MEFV 2–3 and MEFV 7–10 primers, was more prominent in the 2 heterozygous patients during periods of inflammation as compared with the other control patients with acute appendicitis who had no mutation; however, this difference was not statistically significant (Table 2).
Table 2. Relative expression units of the MEFV gene in carriers of 2, 1, or no mutations, as determined during periods of acute inflammation and during asymptomatic periods*
Patient group, sampling period
No. of MEFV mutations
Except where indicated otherwise, values are the mean ± SD.
Familial Mediterranean fever
No. of patients
MEFV exons 2–3
2.30 ± 1.20
3.33 ± 0.75
3.02 ± 1.18
1.54 ± 0.62
2.03 ± 0.81
1.92 ± 0.90
MEFV exons 7–10
1.02 ± 0.91
1.72 ± 1.00
1.08 ± 0.64
0.89 ± 0.28
1.50 ± 1.31
1.34 ± 0.76
No. of patients
MEFV exons 2–3
6.67 ± 3.06
7.20 ± 0.17
3.96 ± 3.23
1.53 ± 0.42
MEFV exons 7–10
2.91 ± 1.87
2.34 ± 0.73
2.20 ± 2.45
0.31 ± 0.33
CIAS1 expression in PBLs from patients with FMF was found to be significantly lower than that in control patients with acute appendicitis during asymptomatic periods (mean ± SD 0.96 ± 0.47 [median 0.86] versus 2.11 ± 1.08 [median 2.32]; P = 0.001). The difference in CIAS1 mRNA levels between patients with FMF and those with acute appendicitis was not significant during periods of inflammation (0.79 ± 0.60 [median 0.68] versus 1.86 ± 2.54 [median 0.81]; P = 0.14) (Figure 1). In addition, we observed no significant change in CIAS1 expression in PBLs between periods of inflammation and asymptomatic periods in both the FMF and acute appendicitis groups (Figure 1). In the only FMF patient who was not taking colchicine, CIAS1 expression was higher than the mean value in the remaining 15 FMF patients during both assessment periods (1.77 versus 0.90 ± 0.43 during asymptomatic periods, and 1.27 versus 0.76 ± 0.61 during periods of inflammation). However, this difference was not statistically significant.
No significant difference was observed in the PSTPIP1 expression in PBLs from FMF patients (mean ± SD 1.81 ± 0.90 [median 1.58]) and acute appendicitis control patients (2.37 ± 1.31 [median 2.20]) during asymptomatic periods. During periods of inflammation, a lower expression of PSTPIP1 in FMF patients (1.05 ± 0.44 [median 1.07]) was noted, despite a nonsignificant increase in expression in control patients with acute appendicitis (3.96 ± 7.01 [median 1.68]) (Figure 1). No significant effect of colchicine on PSTPIP1 expression was noted in the FMF patients (data not shown).
In this study, we found decreased levels of MEFV mRNA in PBLs obtained from FMF patients during asymptomatic periods, confirming the findings reported by Notarnicola and colleagues in 92 asymptomatic patients with FMF (12), and we observed a further decrease in the levels of MEFV mRNA in these patients during an attack of acute abdominal inflammation as compared with their asymptomatic period. We also detected decreased MEFV mRNA levels in PBLs obtained from patients with acute appendicitis during the period of inflammation, as well as similar MEFV mRNA levels in patients with sepsis.
These observations support the idea that reduced MEFV expression in PBLs is associated with acute inflammation. In a recent study using double-stranded MEFV small interfering RNA in the monocytic THP-1 cell line, Chae et al (10) showed that pyrin production can be markedly reduced and that these cells produced significantly higher levels of IL-1β following LPS stimulation. These results confirm that the main function of the pyrin protein is inhibitory and that it acts as a negative regulator of the inflammation. It has been demonstrated that pyrin can inhibit cryopyrin-mediated inflammation and apoptosis by interfering with the interaction of cryopyrin with ASC (15). Reduced expression of pyrin may result in a defect in the regulation of ASC-dependent pathways of inflammation, which are also called inflammasomes (4). Chae and colleagues' studies (10) also revealed an ASC-independent role of pyrin in the regulation of IL-1β activation. They demonstrated that the B30.2 domain of the pyrin protein interacts with caspase 1, and crystal-structure models imply an impaired interaction in subjects with M694V and M680I mutations, which are located at the putative binding site interface. These findings suggest that at least some of the conservative missense mutations can contribute to the disease phenotype by ASC-independent means, through defective binding to caspase 1.
Notarnicola et al (12) reported a correlation between the MEFV genotype and mRNA levels, and M694V was found to be significantly associated with the lowest level of mRNA. In the same study, the MEFV RNA level was significantly lower as compared with that in healthy controls, even in FMF patients with no coding-region mutations (12). In the current study, we screened our patients only for the 5 most frequent mutations, and 2 of the patients also had exon 10 sequencing results. It is possible that at least some of the 6 patients with 1 mutation and the 4 patients with no mutations may have other rare MEFV variants associated with the FMF phenotype. However, we did not observe any significant difference in the relative expression of MEFV mRNA in PBLs from patients with 0, 1, or 2 mutations. We also did not find a significant difference between patients with the M694V mutation (n = 5) and patients with other or no mutations (data not shown), although this is probably because the study group was rather small.
These findings support the hypothesis that there may be some FMF patients who have no MEFV coding-region mutations but have a reduced expression of MEFV. Other genetic variations in the regulatory regions of the MEFV gene or variations in the other genes that interfere with the expression of the MEFV gene may result in a similar FMF phenotype. It is still unclear whether reduced transcription activity or increased mRNA turnover is associated with decreased levels of MEFV mRNA during inflammation in patients with FMF and in patients with other inflammatory conditions. Clarification of MEFV mRNA metabolism would be helpful in understanding the pyrin-related regulatory networks.
We observed differences between the expression levels of MEFV 7–10 and MEFV 2–3 in FMF patients as well as in control patients. This variation in expression may result from a methodologic problem, such as the efficiency of the RT-PCR reaction, which may be affected by the length (180 bp for MEFV 2–3, and 215 bp for MEFV 7–10) or the nucleotide sequence of the amplified product. But, these findings may also suggest the different expression patterns for splice variants. Papin et al (16) and Diaz et al (17) reported various MEFV splice variants, including one that lacks all of exon 2 and an isoform with an extension of exon 8, which causes a frameshift and encodes a protein that lacks the highly mutated C-terminal B30.2 domain. The changing subcellular distribution of these isoforms in different cell types could imply diverse functions for each one, and the role of splice variants in the expression of MEFV needs to be studied further.
Another important finding of this study was the reduced expression of CIAS1 mRNA in PBLs from patients with FMF. CIAS1 has previously been shown to be expressed in PBLs and especially in polymorphonuclear cells (18). Decreased CIAS1 mRNA levels in FMF patients may indicate a new homeostatic mechanism in ASC-dependent pathways that compensates for the lower MEFV expression even during asymptomatic periods. However, Ben-Chetrit et al (19) recently showed that long-term administration of high doses of colchicine suppresses endothelial cell expression of many proinflammatory genes at the RNA level. Although CIAS1 was not named among the genes suppressed in endothelial cells by colchicine therapy, the higher level of CIAS1 expression in the only FMF patient who was not taking colchicine suggests that drug therapy may have contributed to the reduced level of CIAS1 in PBLs from the FMF patients.
In conclusion, our findings confirmed that reduced expression of the MEFV gene is associated with inflammation and that, along with disease-associated variations in pyrin, it may be one of the pathogenic mechanisms of the attacks of inflammation in FMF patients. Reduced levels of MEFV expression even during asymptomatic periods may have caused a new homeostatic order in the regulation of inflammation, and these regulatory changes may include a reduced expression of CIAS1 in PBLs from patients with FMF.
Dr. Gül 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 design. Dr. Gül.
Acquisition of data. Drs. Üstek and Selçukbiricik, Mr. Cakiris, Ms Oku, and Drs. Yanar and Taviloglu.
Analysis and interpretation of data. Drs. Üstek, Ekmekci, Mr. Cakiris, Ms Oku, and Drs. Vural, Özbek, and Gül.