Elsevier

Molecular Immunology

Volume 45, Issue 9, May 2008, Pages 2660-2666
Molecular Immunology

Characterization of a polymorphism in the coding sequence of FCN3 resulting in a Ficolin-3 (Hakata antigen) deficiency state

https://doi.org/10.1016/j.molimm.2007.12.012Get rights and content

Abstract

Ficolin-3 (Hakata antigen or H-ficolin) is a soluble pattern recognition molecule in the lectin complement pathway. We speculated whether common genetic variations in the FCN3 gene contribute to deficiency of Ficolin-3.

The FCN3 gene was sequenced in 237 healthy Danish Caucasians. The relevance of polymorphisms was assessed with antibodies against Ficolin-3 in a novel ELISA system and by production of recombinant Ficolin-3 variants. Ficolin-3 serum profiles were analyzed by SDS-PAGE and western blotting.

Ficolin-3 serum concentration varied 10-fold (median, 24 μg/ml; range, 3–54 μg/ml). Out of several polymorphisms one FCN3 + 1637delC causing a reading frame shift and a distortion of the C-terminal end of the molecule with an allele frequency of 0.011 was particularly interesting. In individuals heterozygous for the FCN3 + 1637delC deletion lowered Ficolin-3 concentration was observed (P = 0.025). SDS-PAGE and western blotting of serum revealed a weak band corresponding to the truncated molecule in addition to the normal Ficolin-3 pattern. Characterization of recombinant Ficolin-3 derived from FCN3 + 1637delC showed that in the homozygous situation this allelic variant would lead to Ficolin-3 deficiency.

In conclusion an FCN3 + 1637delC deletion variant disrupting the possibility for pattern recognition was detected. Characterization of recombinant variant Ficolin-3 shows that homozygosity for the FCN3 + 1637delC deletion may lead to Ficolin-3 deficiency and may thus be the basis for a novel complement deficiency state.

Introduction

Ficolin-3 (also named Hakata antigen, thermolabile β-2 macroglycoprotein, thermolabile substance and H-ficolin) was initially identified as an autoantigen found in serum from patients with systemic lupus erythematosus (SLE) (Yae et al., 1991). The primary structure of Ficolin-3 has the characteristics of the ficolin protein family with collagen-like and fibrinogen-like domains, which also includes Ficolin-1 (M-Ficolin) and Ficolin-2 (L-Ficolin) (Endo et al., 2007). Like Ficolin-1 and Ficolin-2, Ficolin-3 comprises a short N-terminal region (24 aa), followed by a collagen-like domain (11 Gly-Xaa-Yaa repeats), a neck region (12 aa) and a fibrinogen-like domain (207 aa) in the C-terminus. The homology between Ficolin-1 and Ficolin-2 with Ficolin-3 is overall around 48% (Sugimoto et al., 1998). On SDS-PAGE analysis, Ficolin-3 shows a single band of 35 kDa under reducing conditions and numerous ladder bands under non-reducing conditions indicating that Ficolin-3 exists in serum as a huge homo-polymer.

The FCN3 gene is localized on 1p36.11, however alternative splicing occurs at the FCN3 locus and two variants, each encoding a distinct isoform, have been identified (NM003665 and NM173452, isoform 1 and 2, respectively). Transcript variant 2 lacks exon 4 compared to variant 1, resulting in a shorter protein. Whether splicing variants are present in serum is not clarified. Ficolin-3 mRNA is found in ciliated bronchial epithelial cells, in bile duct epithelial cells and hepatocytes in the liver (Akaiwa et al., 1999). However, the relative FCN3 expression level is even higher in the lung compared with the liver (Hummelshoj et al., 2007a).

The specificity of Ficolin-3 has not yet been fully elucidated, but Ficolin-3 has been shown to have a calcium-independent lectin activity: it binds GlcNAc, GalNAc and fucose (Sugimoto et al., 1998) and although the knowledge about the binding of Ficolin-3 to bacteria is limited it has been shown to bind and inhibit the growth of Aerococcus viridans (Tsujimura et al., 2002, Krarup et al., 2005). Moreover, Ficolin-3 has been shown to mediate sequestration of late apoptotic cells (Honore et al., 2007).

Like mannose-binding lectin (MBL), Ficolin-1 and Ficolin-2 also Ficolin-3 activates the lectin complement pathway (Matsushita et al., 2002). Recently it has been shown that Ficolin-3 indeed has the highest complement activating capacity among the lectin pathway initiator molecules (Hummelshoj et al., 2007a). Thus, Ficolin-3 along with the other structural related molecules is now regarded as the recognition molecules in the lectin complement pathway analogous to C1q from the classical complement pathway.

There is compelling evidence that complete deficiency of any of the early complement components C1q, C1r, C1s, C2 and C4 is strong genetic risk factors SLE development and increased tendency for infection (Pickering et al., 2000). However, homozygous deficiency of these factors is very rare seen in less than 1% of the patients. Thus, other disease causing factors with related or overlapping functions may probably add to susceptibility in SLE patients. MBL (MBL2 gene locus) has also been shown to weakly contribute to SLE manifestations and infectious tendency (Garred et al., 2001). However, defects in the Ficolin system has not been reported. We have previously shown that the promoter and coding regions of both FCN1 and FCN2 encoding Ficolin-1 and Ficolin-2, respectively, are highly polymorphic and that polymorphisms in the promoter region of FCN2 is associated with serum variations of Ficolin-2 (Hummelshoj et al., 2005). Further, we have shown polymorphisms in FCN2 situated within exons encoding the fibrinogen-like domain affect the binding of Ficolin-2 to GlcNAc. In the same study we screened 60 healthy individuals for variation in the coding part of the FCN3 gene and we only detected one mutation in a single individual; a deletion (FCN3 + 1637delC, rs28357092) in exon 5 which leads to a reading frame shift (fs) and consequently in an early termination of the transcript (Hummelshoj et al., 2005). This frame shift causes a altered amino acid composition of the C-terminal end of the protein spanning from the amino acid position 117 to an early stop codon at position 180 and gives rise to a truncated Ficolin-3 variant (Leu117fs) lacking 182 of the correct amino acid sequence equivalent of the entire fibrinogen-like domain.

In this study we asked the question whether the frame shift mutation is really an established polymorphism in the Caucasian population and investigated its putative functional and structural consequences. Moreover, we investigated the promoter and coding sequences of FCN3 gene for polymorphisms in order to determine to which degree the Ficolin-3 serum concentration may be determined by FCN3 polymorphisms.

Section snippets

Donor samples

Peripheral venous blood samples were obtained from 237 unrelated adult Danish blood donors with consent. Genomic DNA was successively prepared from 237 of blood samples using the method described by Miller (Miller et al., 1988) and parallel serum samples. The local ethical committee approved the study.

Recombinant ficolins

Recombinant Ficolin-3 (rFicolin-3) was produced. Briefly, rFicolin-3 was amplified from liver cDNA and tagged C-terminally with or without a penta-His sequence. This construct was cloned into an

Identification of polymorphisms in the promoter region and exon 5 of FCN3

In 237 Danish blood donors the FCN3 promoter region, spanning from approximately −1000 to +100 bp (with regard to the first ATG site) was PCR amplified in one fragment and sequenced using intern primers, due to location of both Alu and L2 elements. Apart from one known SNP FCN3-608 C > T (0.981) (rs28385648) two novel SNPs were found in the promoter: FCN3-954 G > A 0.994 (rs41465749) and -325 C > T (0.888) (rs41415450). In exon 1 a novel SNP was identified + 34 C > G (0.992) (rs41366544) which leads to an

Discussion

Based on sequence analysis of all eight FCN3 exons and the intron exon boundaries we have previously observed in a single individual out of 60 investigated individuals the existence of a deletion (FCN3 + 1637delC, rs28357092) which leads to a reading frame shift and consequently in an early termination of the transcript (Hummelshoj et al., 2005). However, whether this was a random mutation or an established polymorphism with structural and functional relevance was until the present study unknown.

Acknowledgements

The authors wish to thank Mss Vibeke Witved and Vibeke Weirup for excellent technical assistance. This work was supported by grants from the Danish Medical Research Council, Novo Nordisk Fonden, Danish Rheumatism Association, The Lundbeck Foundation, Copenhagen University and Rigshospitalet, Dr. Lea Munthe-Fog is a Rigshospitalet research fellow.

References (17)

There are more references available in the full text version of this article.

Cited by (112)

View all citing articles on Scopus
View full text