Establishment of a transgenic mouse model with liver-specific expression of secretory immunoglobulin D

Mutation of mevalonate kinase (MVK) is thought to account for most cases of hyperimmunoglobulinemia D syndrome (HIDS) with recurrent fever. However, its mechanism and the relationship between elevated serum immunoglobulin D (IgD) and the clinical features of HIDS are unclear. In this study, we generated by fusion PCR a vector to express high levels of chimeric secretory IgD (csIgD) specifically in the liver. We then generated seven founder lines of transgenic mice by co-microinjection, and verified them using genomic PCR and Southern blotting. We detected the expression of csIgD by reverse transcription PCR, quantitative PCR, western blotting, and enzyme-linked immunosorbent assays. We demonstrated that csIgD could be specifically and stably expressed in the liver. We used flow cytometry to show that overexpression of csIgD in the bone marrow and spleen cells had no effect on B cell development. Morphologic and anatomical observation of the transgenic mice revealed skin damage, hepatosplenomegaly, and nephromegaly in some transgenic mice; in these mice, pathological sections showed high levels of cell necrosis and protein-like sediments in the liver, spleen, and kidney. We demonstrated that the genomic insertion sites of the transgenes did not disrupt the MVK gene on mouse chromosome 5. This transgenic mouse will be useful to explore the pathogenesis of HIDS.

Hyperimmunoglobulinemia D syndrome (HIDS), also called receptor-associated periodic syndrome or etiocholanolone fever, is characterized by high serum levels of immunoglobulin D (IgD) and recurrent febrile attacks, and may also include arthritis, lymphadenopathy, hepatosplenomegaly, and skin rash [1-3]. It was originally described in patients of Dutch ancestry by Van der Meer et al. in 1984 [4] but has now been reported all over the world, including in Japan [5,6]. Its pathogenesis is unknown and there are no effective therapies. High levels of IgA and tumor necrosis factor al-pha (TNF-α) are found in the sera of some HIDS patients [7]. Patients with HIDS also secrete large amounts of the proinflammatory cytokine interleukin 1β (IL-1β), which may be connected with low levels of isoprenoid metabolic end products caused by a lack of mevalonate kinase (MVK) [8].
In 1999, Drenth et al. [9] obtained samples from 34 patients and 44 unaffected members from 16 families with HIDS who originated from the Netherlands (8), France (5), the United Kingdom (1), Spain (1), and the Czech Republic (1). An MVK gene missense mutation resulting in decreased MVK activity was identified. Reduced MVK activity in fibroblasts suggested that the MVK mutation caused HIDS. Since then, other MVK gene mutations or deletions have been found in some HIDS patients [10,11]. MVK is a protease in the cytoplasm that participates in the synthesis of cholesterol and non-sterol isoprenoid. Human MVK is located on chromosome 12q24. The V377I mutation in MVK occurs homozygously in approximately 20% of HIDS patients and heterozygously in most of the others. The second most common mutation is I268T [12,13].
The effect of deletion of a single Mvk allele in mice was reported by Hanger et al. in 2007 [14]. These authors reported significantly increased IgD levels, and a trend towards higher IgA and TNF- levels, in Mvk +/ sera compared with Mvk +/+ littermate sera, which was accompanied by periodic fever and hepatosplenomegaly. Moreover, Ammouri et al. [15] reported that, among fifty HIDS patients, 38 had high serum IgD, even though only 19 were homozygotes or composite heterozygotes for MVK mutations. This indicates that secretory IgD (sIgD) plays an important role in the clinical presentation of HIDS. The function of sIgD is unclear, although it has been reported to combine with antigens of microorganisms and their products. In early inflammatory responses in patients with contact dermatitis, sIgD could induce bone marrow cells-mainly neutrophils, eosinophils, and basophils--to migrate to the skin [16]. sIgD binds to basophils through a calcium-mobilizing receptor that induces antimicrobial and B cell-stimulating factors, after IgD cross-linking. sIgD is an important immune-conditioning molecule; its interaction with its binding partners induces cell activation, organization penetration, immune activation factor release, and enhances other immune defenses. However, excess activation can lead to inflammation and tissue destruction [17]; the effects and levels of increased sIgD vary among HIDS patients [18].
The relationships among sIgD, MVK, and the inflammatory response are unclear; and it is unknown how increased sIgD or MVK mutations could cause inflammation. To explore this question, we constructed a transgenic mouse model with liver-specific sIgD overexpression.

Fusion PCR to generate the en-pAlb fragment
Genomic DNA was extracted from the liver tissue of Kunming white mice (purchased from the Zoology Department of the National Family Planning Institute, Beijing, China). The 1.97-kb albumin (Alb) enhancer [19] and the 412-bp Alb promoter [20,21] were amplified by PCR. The primers for the mouse Alb enhancer were en-Alb-F (5′-AAA AAC  GCG TAT TTG AAT ACA CAC TTT TGT CAT G-3′) and  en-Alb-R (5′-AGG TTG TTC ATC CCA AAG TTA TCA   ATC AAA GCA G-3′). The primers for the mouse Alb  promoter were p-Alb-F (5′-CTG CTT TGA TTG ATA ACT  TTG GGA TGA ACA ACC T-3′) and p-Alb-R (5′-AAA  AGC TAG CTG CCA GAG GCT AGT GGG GTT G-3′). The underlined sequences indicate Mlu I and Nhe I sites, respectively. Four adenine nucleotides were used to protect the restriction enzyme site. The en-Alb-F and p-Alb-R primers were then used to amplify the 2.35-kb en-pAlb fragment by fusion PCR.

Transgene amplification
Total RNA was isolated using Trizol (Tiangen, Beijing, China) from spleen tissue of Kunming white mice; 1 g of total RNA was used for first-strand cDNA synthesis with M-MLV Reverse Transcriptase (Promega, Madison, WI). The coding regions of the sIgD heavy chain constant region (IgDHC) and kappa chain constant region (KC) were amplified from cDNA. The 803-bp IgDHC region was amplified using primers IgDHC-F (5′-CTG GTC ACC GTC TCC TCA GGT GAT AAA AAG GAA CCT GAC-3′) and IgDHC-R (5′-TTT TGC GGC CGC TCA GGC AAG GGC AGG ACC A-3′). The 348-bp KC region was amplified using primers KC-F (5′-CAA GCT CGA GAT CAA ACG GGC TGA TGC TGC ACC AAC-3′) and KC-R (5′-TTT TGC GGC CGC CTA ACA CTC ATT CCT GTT G-3′). The underlined sequences indicate a Not I site. Four thymidines were used to protect the restriction enzyme site.
The coding regions of the heavy chain variable region (IgGHV) and kappa chain variable region (KV) of a human IgG1 monoclonal antibody against Hepatitis A virus (HAV) were amplified by PCR from plasmids pHAVH3 and pHAVL3 (saved by our laboratory). The 450-bp IgGHV region was amplified using primers IgGHV-F (5′-TTT TGA ATT CAG CAT GGG TGA CAA TGA CAT CCA C-3′) and IgGHV-R (5′-GTT CCT TTT TAT CAC CTG AGG AGA CGG TGA CC-3′). The underlined sequence indicates an EcoR I site; the sequence in bold typeface shows the Kozak fragment. Four thymidines were used to protect the restriction enzyme site. The 410-bp KV region was amplified using primers IgGHV-F (as above) and KV-R (5′-GTT GGT GCA GCA TCA GCC CGT TTG ATC TCG AGC TTG-3′).
Chimeric (c) sIgD-H (1214 bp) and cKappa (722 bp) were then amplified by fusion PCR. The primers for csIgD-H were IgGHV-F and IgDHC-R; the primers for cKappa were IgGHV-F and KC-R.

Construction of the en-pAlb-csIgD-H and en-pAlb-cKappa vectors
csIgD-H and cKappa were separately cloned into the expression plasmid pcDNA3.1(+) and digested with EcoR I and Not I. Using Mlu I and Nhe I restriction enzyme digestion of these two expression vectors, the CMV promoter was replaced with en-pALB. The resulting plasmids, designated en-pAlb-csIgD-H and en-pAlb-cKappa, were under control of the mouse Alb regulatory sequences to direct gene expression specifically in the liver.

Generation of transgenic mice
The 4059-bp csIgD-H fragment and the 3567-bp cKappa fragment were released from their plasmids by Mlu I and Nae I digestion, respectively. Each fragment contained the mouse en-pAlb, a T7 promoter, an Ig secretory leader, either the csIgD-H gene or the cKappa gene, followed by the bovine growth hormone (BGH) poly(A) (Figure 1). The two fragments were purified after agarose gel electrophoresis (Omega, Boston, Massachusetts) and subsequently co-microinjected at a 1:1 molar ratio into the pronuclei of fertilized Kunming white ova according to standard protocols. Genomic DNA was extracted from mouse tail biopsies. The presence of the transgenes was verified by PCR with primers p-Alb-F (as above) and en-pAlb-R (5′-AGG AAG GGA AGA AAG CGA AAG-3′). The primers were specific to the Alb promoter and BGH poly(A) sequences, respectively, and so could detect both the heavy and light chain genes simultaneously. Genomic DNA (10 g) was digested with Dra I and analyzed by Southern blotting with digoxigenin-labeled hybridization probes. The probes were the amplified products of the csIgD-H and cKappa genes.

SDS-polyacrylamide gel electrophoresis (PAGE) and Western blotting
Blood from the orbital sinus of transgenic and non-transgenic mice was collected in 0.5-mL centrifuge tubes and stored overnight at 4°C. Serum was obtained by centrifugation at 1200×g for 15 min at 4°C, and stored at 80°C. The total serum protein concentration was determined by the BCA method (Beyotime, China). Serum samples containing 8 g of protein were mixed with 5×SDS, denatured for 5 min at 95°C, and then separated on 10% SDS-PAGE gels under reducing conditions, using separate samples for staining with Coomassie brilliant blue and Western blotting. The separated proteins were electrophoretically transferred to nitrocellulose membranes (Amersham Pharmacia, UK), washed twice in Tris-buffered saline/20% Tween 20 (TBST) for 5 min, incubated overnight in blocking buffer (5% defatted milk in TBST) at 4°C, followed by two washes in TBST for 10 min each. The membranes were then incubated with a 1:1000 dilution (in 0.5% blocking reagent) of horseradish peroxidase-conjugated monoclonal anti-mouse IgD antibody (eBioscience, San Diego, California) for 1 h at room temperature, washed three times for 10 min each, and then visualized using enhanced chemiluminescence reagents (Amersham Biosciences, UK) according to the manufacturer's instructions. Mouse GAPDH was used as an internal control when we detected the sIgD expression in transgenic mouse liver with Western blotting.

Enzyme-linked immunosorbent assay (ELISA)
A mouse serum IgD-ELISA kit (BlueGene, China) was brought to room temperature for 30 min before use, according to the manufacturer's instructions. One hundred microliters of serum standard, sample, or distilled water (blanks) was added to each well of a 96-well microtiter plate; 50 L of enzyme-linked marker solution was then added to each well except the blank wells, the wells were covered with adhesive strip and incubated for 1 h at 37°C. The microtiter plate was then washed five times, and patted dry with absorbent paper. Next, 50 L of chromogenic reagents A and B were added to each well except the blank wells, and the plate was incubated in the dark for 15 min at 20-25°C, followed by the addition of 50 L of stop solution to each well. The optical density at 450 nm of each well was determined within 30 min, using an automatic microplate reader (Bio-rad Co., Philadelphia, USA).

Analysis of B cell development by flow cytometry
For the isolation of bone marrow cells and spleen cells, transgenic mice and sex-matched littermate non-transgenic mice were sacrificed by cervical dislocation. The flesh of the mouse legs was stripped, and then 2 mL syringes filled with phosphate-buffered saline (PBS) were inserted into the bone cavities to flush the bone marrow into 50-mL centrifuge tubes that were kept on ice. Spleen tissues were placed onto a 200-mesh sieve (Nanjing Xiaoxiao Instrument and Equipment Co., Ltd.), sprinkled with a small amount of PBS, and gently broken into intact single cells with the core of a syringe; the cells were then flushed with PBS into 50-mL centrifuge tubes kept on ice. Two volumes of red blood cell lysis buffer (eBioscience, San Diego, California) were added to remove red blood cells, mixed and incubated on ice for 10 min. The reaction was then terminated by adding five volumes of PBS. The single-cell suspensions of bone marrow and spleen were centrifuged at 300×g for 10 min at 4°C. The supernatant was discarded and the cells resuspended in an appropriate amount of PBS. Bone marrow and spleen cells were then counted using a blood cell count board.
The cells were divided into five fractions for the addition of antibody: blank (no antibody added); phycoerythrinconjugated Cy5.5 anti-mouse B220; fluorescein isothiocyanate-conjugated anti-mouse IgM; phycoerythrin-conjugated anti-mouse CD43; and all three of the above antibodies. The dilution and dosage of antibodies were according to the manufacturer's instructions (eBioscience, San Diego, California). These ten tubes of cells (five each for bone marrow and spleen) were incubated in the dark for 20 min at 4°C, according to the manufacturer's instructions. The cells were then washed twice with PBS, centrifuged at 300×g for 10 min at 4°C, resuspended in 150 L PBS and kept on ice. The B cells were analyzed by flow cytometry (MOFLO High Performance Cell Sorter, Dako; Fort Collins, Colorado).

Thermal asymmetric interlaced PCR (TAIL-PCR)
Three transgene-specific primers (tgp1: 5′-GAG CAG CAC CCT CAC ATT-3′ or 5′-CAG CAA GAG CCT AGC AAT-3′; tgp2: 5′-TGC CAG CCA TCT GTT GTT-3′; and tgp3: 5′-TCT GAG GCG GAA AGA ACC-3′) and an arbitrary degenerate primer (5′-NGT CGA SWG ANA WGA A-3′, 5′-AGW GNA GWA NCA WAG G-3′, or 5′-WGT GNA GWA NCA NAG A-3′) were used for TAIL-PCR. The primary PCR reaction contained 0.5-1.0 g of genomic DNA. In the secondary or tertiary amplification, 0.65 L of the first or secondary products was used as the template, respectively. The products of tertiary TAIL-PCR were separated on 1.5% agarose gels. A single band in each mouse sample was purified with a TIANgel Midi Purification Kit (Tiangen, China), sequenced, and analyzed by BLAST on the National Center for Biotechnology Information website. In addition, the morphology and anatomy observation and tissue sections (stained hematoxylin-eosin) analysis of the transgenic mice was detected.

Identification of the transgenic mice
Nine transgenic founders (Nos. 3, 4, 5, 6, 7, 12, 15, 24, and 37) were obtained from 37 mice analyzed by PCR ( Figure  2A). All carried both the csIgD-H and cKappa transgenes. In addition, there were another three founders that carried either the csIgD-H or cKappa transgenes; these were not analyzed further. Southern blotting was used to confirm that both transgenes were integrated into the genome of all nine transgenic mice ( Figure 2B). Probes specific for the 1214-bp csIgD-H and 722-bp cKappa fragments were mixed together, allowing both transgenes to be detected simultaneously. Of the founders, the founder 12 died and the founder 15 could not be passaged because of illness; however, 7 were mated with wild-type mice, all of which transmitted the transgenes to their offspring; of these, 26 F 1 transgenic mice with double transgenes were identified among 64 offspring.

Expression of csIgD in the liver tissues of transgenic mice
Transgene mRNA expression in the liver tissues of trans-genic mice was analyzed by RT-PCR ( Figure 2C). Both the csIgD-H and cKappa transgenes were expressed in eight transgenic lines (Nos. 15 and 37 of the F 0 generation and six mice of the F 1 generation). There were no bands in the negative control mice; a band for the housekeeping gene -actin was visible in each at ~109 bp. This confirmed that csIgD-H was expressed from the chimeric mRNA in the liver tissues of all transgenic mice. Quantitative PCR showed differing levels of expression between csIgD-H and cKappa in individual mice and among different mice (Figure 2D).
Serum samples were collected from five transgenic founder mice or F 1 mice (F 0 6-F l 7, F 0 15, F 0 7-F 1 24, F 0 37, and F 0 4-F l 50) that carried both transgenes. The expression of serum sIgD was analyzed by Western blotting. A single band was detected for sIgD-H under reducing conditions with the expected molecular weight of 46 kD ( Figure 3A). The expression of sIgD in the liver and spleen of F 0 37 and two negative littermates was then analyzed. A single 46-kD  band was detected in both tissues of mouse F 0 37; in controls, bands were observed in the spleen but not the liver (sIgD always expressed in spleen but no liver in normal condition) ( Figure 3B). By ELISA, the serum sIgD level in each transgenic line was greater than that in control mice ( Figure  3C).

B cell development in transgenic mice
The bone marrow and spleen cells from a randomly selected transgenic mouse (F 0 4-F 1 81) and a sex-matched littermate negative control were analyzed by flow cytometry. There were no significant differences between them in the content of pro-B (B220 + CD43 + IgM  ), pre-B (B220 + CD43  IgM  ), or immature B (B220 + CD43  IgM + ) bone marrow cells or mature spleen B cells (B220 + IgD + IgM +/ ) (Figure 4). This in- dicates that sIgD overexpression in transgenic mice has no effect on B cell development.

Integration sites of the transgenes
We used genome walking (TAIL-PCR) to detect the site of integration of the transgene fragments in three transgenic mice (F 0 6-F l 7, F 0 7-F 1 18, and F 0 4-F l 50) (data not shown). The transgenes in these three mice integrated into chromosomes 14, 13, and 1, respectively; there were no functional genes near the insertion sites. This indicates that the Mvk locus on mouse chromosome 5 was not disrupted by the transgenes.

The morphology and anatomy observation of the transgenic mice
In observing the morphology and anatomy of four of the transgenic lines (F 0 15, F 0 7, F 0 4, and F 0 37), we found that these mice had skin ulcers, hepatomegaly, splenomegaly, and renomegaly, with livers 2.5-3.3 times, spleens 4.4-6 times, and kidneys 2.7-3.3 times the size of those of non-transgenic littermates ( Figure 5A). Hematoxylin-eosin staining of tissue sections showed that these organs were diseased, with a high degree of cell necrosis and fiber-like protein precipitation ( Figure 5B). These four transgenic founders were mated with wild type mice. F 0 15 and F 0 37 could not be passaged because of illness; F 0 7 had two positive offspring, one of which had skin ulcers, hepatosplenomegaly, and renomegaly, as did two of the three positive offspring from F 0 4.

Discussion
sIgD was once considered to have no function, considering its low levels in serum and short half-life. Recent reports of HIDS [2225] have caused researchers to re-examine sIgD, but its specific role in HIDS is still not clear. To facilitate this research, we generated a mouse model that stably and efficiently expresses sIgD. In this study, we found that csIgD could be expressed in the liver of mice under the control of the Alb enhancer and promoter. The mouse Alb promoter contains six cis-acting elements [20], which bind hepatocyte nuclear factor 1, CCAAT-binding protein, D-binding protein, leucine chlorolactamase, nuclear factor Y, and NF-1, respectively, and thus initiate gene transcription in the liver [26,27]. In addition, enhancers located 10.4-8.5 kb upstream can increase promoter activity 50-fold [28][29][30]. The sIgD expression was increased in different degrees in individual mice. Most mice carried both transgenes and most transgenic founders transmitted both transgenes to their offspring. This is consistent with previous reports that co-microinjection normally leads to the co-integration of two transgenes into a single chromosomal site [31,32]. The expression of sIgD showed that the gene copy number was not directly proportional to the level of gene expression. This is consistent with previous studies indicating that the transgene integration sites reflect gene expression levels better than does the copy number [33]. Deletion of a single Mvk allele in the mouse was able to increase sIgD 9-12-fold compared with negative control mice [14]. We inserted the target gene into the multiple cloning site of pcDNA3.1(+) to construct a high-expression sIgD transgenic mouse model, and expression of the transgenes could be detected by RT-PCR in the heart, liver, spleen, lung, kidney, intestine, brain, and muscle (data not shown). The lack of increase in serum sIgD may be for one of the three following reasons. First, the variable region of csIgD was derived from human rather than mouse IgG1, which may affect the spatial structure of the molecule and result in the affact of its identification by the antibody. Second, while we used the kappa light chain to construct csIgD, sIgD preferentially uses the lambda light chain [34,35], with the kappa light chain used for membrane IgD [36,37]. This might also affect the secretion and detection of csIgD. Finally, sIgD might have increased in serum but then have been rapidly degraded through a protective mechanism, making the increase undetectable. Further studies are needed to explain why we did not achieve stable and efficient expression of sIgD.
sIgD is a sign of early-stage B cell activation; it is also involved in the production and maintenance of memory B cells, and may play an important role in the sensitivity response of B cell tolerance. Therefore, if sIgD is increased, then B cell development is affected. Our flow cytometry analysis showed that there was no change in the proportion of B cells in the various stages of development (pro-B, pre-B, immature B, and mature B) between transgenic and control mice, indicating that the transgenes do not affect B cell development.
By genome walking, we demonstrated that the transgenes were integrated on different chromosomes in different transgenic mice, and that no functional genes (including Mvk) had been disrupted. This indicates that the inflammation in the transgenic mice is caused by increased sIgD rather than by Mvk mutation. sIgD has a proinflammatory function, and can effectively induce antimicrobial, immunostimulating, and proinflammatory factors such as IL-1, TNF-, IL-8, and CXCL10 [38]. Mucosal IgD can enhance local immunity by epithelial transport whereas cir-culating IgD binds to basophils through a calcium-mobilizing receptor. In addition to quickly mediating an innate immune response, IgD also conducts immune signaling when bacteria invade. When IgD-bound basophils encounter antigen, they migrate to the lymphoid system or mucosal immune lymphoid organ driven by chemokines that are released by mast cells under IgD stimulation. Basophils are organized through a calcium-mobilizing receptor that induces antimicrobial, opsonizing, inflammatory, and B cell-stimulating factors to enhance immune protection, and they induce an extreme immune response that leads to lesion destruction. Skin ulcers in transgenic mice with high expression of sIgD may be caused by excessive immune activation. In addition, displayed severe hepatosplenomegaly and renomegaly. These diseased organs had a high degree of cell necrosis and fiber-like protein precipitation--similar to states seen in human HIDS [5,[22][23][24][25]39]. However, further studies are required to prove the usefulness of these transgenic mice as a model of HIDS.