Genetic Dissociation of Dacryoadenitis and Sialadenitis in a Sjögren’s Syndrome Mouse Model with Common and Different Susceptibility Gene Loci

Abstract

purpose. Sjögren’s syndrome (SS) is a systemic autoimmune disease in which the main lesions are dacryoadenitis and sialadenitis. It is unclear whether these lesions develop in a common genetic background. A quantitative trait locus (QTL) analysis was performed in the SS mouse model, MRL/MpJ-lpr/lpr (MRL/lpr), to identify the susceptibility loci to dacryoadenitis and sialadenitis and the association with both loci.

methods. MRL/lpr, C3H/HeJ-lpr/lpr (C3H/lpr), (MRL/lpr × C3H/lpr) F1, and (MRL/lpr × C3H/lpr) F2 intercross mice were prepared, and the severity of dacryoadenitis and sialadenitis in individuals was quantified by histopathologic grading. In genomic DNA samples from the F2 mice, the polymorphic microsatellite markers highly associated with each lesion were determined as susceptibility loci.

results. QTLs with significant linkage for dacryoadenitis were mapped on chromosome 1 (the position of maximum logarithm of odds [LOD] score; 64.1 cM), designated Adacm1; chromosome 2 (88.4 cM), Adacm2; and chromosome 5 (63.9 cM), Adacm3. Those for sialadenitis were mapped on chromosome 1 (69.0 cM), Asm3, and chromosome 2 (65.3 cM and 82.1 cM), Asm4 and Asm5. Adacm1/Asm3 and Adacm2/Asm5 seemed to be a common chromosomal region, respectively. MRL-homozygous at Adacm1 and Adacm2 and at Asm3 and Asm5 manifested an additive effect on the development of dacryoadenitis and sialadenitis, respectively, whereas Adacm3 did not.

conclusions. Dacryoadenitis and sialadenitis in MRL/lpr mice are under the control of common and different susceptibility loci, with an allelic combination that leads to regular variations in pathologic phenotypes.

Sjögren’s syndrome (SS) is a systemic autoimmune disease. However, the main lesions in SS are in the lacrimal glands and salivary glands, associated with autoimmune mechanisms. These lesions are commonly characterized by the destruction of the ductules at their beginning, by the accumulation of mononuclear cells followed by destruction of the acinar glands. ¹ However, it is unclear whether these lesions develop in a common genetic background. 

An investigation of the clinical symptoms in patients with SS showed that the incidence of dry eye is 67.5% and that of dry mouth is 93.5%, ² when based on the European Community criteria. ³ Other studies showed that the frequency of both symptoms in SS is 82.5%, ⁴ according to the revised European classification criteria. ^{5 6} These results suggest that almost 20% to 30% of patients with SS do not have either dry eye or dry mouth. Although the laboratory tests in patients with SS revealed that 80.7% of the patients with SS are positive for the ocular component and 94.7% of those are positive for the oral component. ⁴ This result means that almost 20% of patients with SS do not have either dysfunction of the lacrimal glands or salivary glands. Histologic examinations of biopsy specimens from lacrimal glands and minor salivary glands in patients with SS showed that 69.2% of patients with SS met the histopathologic criteria of dacryoadenitis and 88.8% of those met that of sialadenitis, based on the revised Japanese criteria. ⁷ As for this result, the histopathologic manifestations may show the same tendency of the clinical symptoms and laboratory tests. Therefore, the development of dacryoadenitis and sialadenitis in patients with SS may not always be coincidental. 

In general, autoimmune diseases show the complex pathologic manifestations involving various lesions such as vasculitis, glomerulonephritis, arthritis, and/or sialadenitis. It remains controversial whether such a complex disease is a manifestation of advanced disease or of multigenic disease with a different gene combination that might be categorized into distinct disease entities. To clarify whether dacryoadenitis and sialadenitis in SS have common genetic mechanisms, we used a SS model, an MRL/Mp-lpr/lpr strain of mice. ⁸ MRL/lpr mice spontaneously develop progressive dacryoadenitis and sialadenitis and other autoimmune diseases such as lupus-like nephritis, arthritis, and vasculitis, in association with immunologic abnormalities. ^{8 9 10} Dacryoadenitis and sialadenitis in MRL/lpr mice are histopathologically characterized initially by mononuclear cell infiltration into periductular regions, followed by the progressive destruction of ductules and parenchyma, resembling SS, and thus these mice provide a SS model. ^{10 11 12}  

The lymphoproliferative gene, lpr, is a mutation of the Fas gene, ¹³ which causes an insufficiency of Fas-mediated apoptosis in T and B cells and activated macrophages. ^{14 15} Therefore, this mutant gene plays an important role in the mechanisms of development of autoimmune diseases in MRL/lpr mice. MRL/lpr mice develop autoimmune diseases, whereas other strains of mice carrying the lpr gene, such as C3H/HeJ and C57BL/6J, do not. ¹⁶ Moreover, MRL mice have the genes that make them susceptible to each lesion. That is, lupus-like nephritis, arthritis, vasculitis, and sialadenitis in MRL/lpr mice were genetically dissociated in crossing with non–autoimmune-prone lpr mice. ^{17 18 19 20 21}  

In the past, many susceptibility loci associated with SS have been demonstrated with various mouse models. ^{21 22 23 24} However, all these studies focused on the salivary glands. There has so far been no study in which the susceptibility loci for dacryoadenitis were identified. Of course, the commonality of susceptibility loci of dacryoadenitis and sialadenitis has not yet been determined. 

Our study confirmed the difference in the severity of dacryoadenitis and sialadenitis between MRL/lpr and C3H/lpr mice, independently from the MHC, since both strains have an H-2^k haplotype. Therefore, using F2 intercross with MRL/lpr and C3H/lpr mice, we clarified the mode of inheritance of dacryoadenitis and sialadenitis in the MRL, background and the non-MHC gene loci susceptible to them were identified. This is the first report to our knowledge to identify whether autoimmune dacryoadenitis and sialadenitis in MRL mice have common susceptibility loci. 

Materials and Methods

Mice

MRL/lpr and C3H/lpr mice were originally purchased from The Jackson Laboratory (Bar Harbor, ME). Both strains were bred and housed under specific pathogen-free conditions in the Animal Research Institute of Tohoku University Graduate School of Medicine, Sendai, and the Department of Biological Resources, the Integrated Center for Science, Ehime University. From these strains, 71 MRL/lpr, 80 C3H/lpr, 40 (MRL/lpr × C3H/lpr) F1 (MCF1) and 527 (MRL/lpr × C3H/lpr) F2 (MCF2) mice were prepared in the same institutes. All procedures involving mice adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and the guidelines for laboratory animal experiments. 

Histopathologic Evaluation of Dacryoadenitis and Sialadenitis

At 16 to 20 weeks of age, the mice were dissected while under ether anesthesia. The lacrimal glands and submandibular glands were removed, fixed with 10% formalin in 0.01 M phosphate buffer (pH 7.2) and embedded in paraffin. Each section of lacrimal glands and submandibular glands, bilaterally, was stained with hematoxylin-eosin (HE) and histopathologically examined by light microscopy for dacryoadenitis and sialadenitis, respectively, in which 10 or more ductules in each organ were randomly selected for grading the lesion. MCF2 mice developed a variety of phenotypes for dacryoadenitis and sialadenitis. Some of them showed both dacryoadenitis and sialadenitis, and others showed dacryoadenitis or sialadenitis alone (Fig. 1) . 

The histopathologic examinations for dacryoadenitis and sialadenitis were performed as described elsewhere. ²¹ The findings were graded on a scale of 0 to 3, where 0 is normal; 1 is mononuclear cell infiltration localized in periductular regions, but no destruction of the parenchyma or ductules; 2 is cell infiltration into duct epithelium, with or without extending to acinar parenchyma; and 3 is grade 2 changes with fibrosis and/or granulation and/or regenerative ductal hyperplasia (Fig. 2) . The grading score of each mouse is represented by the index, which was the mean of the individual grading score of 10 or more ductules randomly selected in the histologic section. 

Genotyping of MCF2 Mice

Genomic DNA of 527 MCF2 mice was extracted from the tail or liver. The genotyping of MCF2 was determined by simple sequence-length polymorphism analysis with 119 polymorphic microsatellite markers that were purchased from Research Genetics (Huntsville, AL). The map positions of the microsatellite loci were based on Mouse Genome Informatics (MGI, http://www.informatics.jax.org/ provided in the public domain by the Jackson Laboratory). ²⁵ They were an average of 12 cM apart, with a maximum distance between any two markers of 37 cM. Briefly, a total volume of 20 μL of polymerase chain reaction (PCR) mixture (200–300 ng of genomic DNA, 2 μL of 10× PCR buffer [Takara, Tokyo, Japan], 1.6 μL of dNTP mixture [2.5 mM each; Takara], 0.4 units of Taq polymerase [Ex Taq; Takara], and 0.16 μM of each primer) was prepared for PCR and amplified (GeneAmp PCR System 9600; Perkin Elmer, Wellesley, MA), by using the following thermocycling protocol: 94°C for 2 minutes, followed by 40 cycles of 94°C for 30 seconds, 55 to 60°C for 30 seconds, 72°C for 30 seconds, and 72°C for 5 minutes. 

The PCR products were separated by electrophoresis on 2% to 5% agarose gels or standard denaturing 10% polyacrylamide gels. The genotypes of the PCR products from the MCF2 intercross mice were designated MRL/MRL (M) and C3H/C3H (C) homozygotes or MRL/C3H (H) heterozygotes based on size differences. 

Genome-Wide Screening and QTL Analysis

Initially, the genotyping of each microsatellite marker and phenotyping of each histologic index was performed in the 266 MCF2 mice. Linkage studies on 266 MCF2 mice were performed. Markers with P ≤ 0.01 were selected (selective genotyping). 

Next, linkage studies on another 261 MCF2 mice were performed on the chromosomes bearing the marker showing P < 0.01 by using the markers compiled in the first selection. The linkage map for the MCF2 mice was created with Mapmaker3. ²⁶ A linkage analysis was performed on each candidate chromosome with WinQTL Cartographer (ver. 2.5), ^{27 28} based on each histologic index as the indicators of phenotype. A logarithm of odds (LOD) was calculated using the LRmapqtl program. Composite interval mapping was performed using model 6 of the ZMapqtl program with options set at 2-dM intervals and a 10-cM window size, considering background loci that included unlinked and linked loci-positioned forward parameters. ^{28 29} The experiment-wise significance level for each index was determined by analyzing 10,000 random shuffling permutations of the actual data and the results of chromosomes with significant linkages ³⁰ (α < 0.01) for either index are shown in the Results section. 

Statistical Analysis

Comparisons of the histologic indexes between parental strains of mice and other generations of mice and interactions among three loci were performed with the Tukey-Kramer test and between females and males in MRL/lpr and MCF2 progeny with the unpaired t-test. 

Results

Phenotypic Analyses of MRL/lpr, C3H/lpr, MCF1, and MCF2 Mice

The phenotypic analyses for dacryoadenitis and sialadenitis were represented by the index (Fig. 3) . The indexes of MRL parents for both lesions were females, (mean score of index ± SD, dacryoadenitis, 1.63 ± 0.57; sialadenitis; 1.15 ± 0.50; Fig. 3B ), males (dacryoadenitis, 0.51 ± 0.37; sialadenitis, 0.71 ± 0.29; Fig. 3C ) and both of them (dacryoadenitis, 1.28 ± 0.73; sialadenitis, 0.97 ± 0.47; Fig. 3A ). Conversely, C3H parents of females (dacryoadenitis, 0.28 ± 0.26; sialadenitis, 0.17 ± 0.24), males (dacryoadenitis, 0.01 ± 0.04; sialadenitis, 0.03 ± 0.07), and both (dacryoadenitis, 0.14 ± 0.23; sialadenitis, 0.10 ± 0.19) hardly developed these lesions. 

MCF1 hybrids had a phenotype similar to that of their C3H parents in the females (dacryoadenitis, 0.41 ± 0.25; sialadenitis, 0.20 ± 0.15), the males (dacryoadenitis, 0.05 ± 0.07; sialadenitis, 0.26 ± 0.26), and both (dacryoadenitis, 0.22 ± 0.26; sialadenitis, 0.24 ± 0.23). Scores for both lesions in MRL parents were significantly higher than those in C3H parents and MCF1 hybrids (P < 0.001, respectively), which means that both dacryoadenitis and sialadenitis are inherited recessively in mice with an MRL background. Both lesions in females and dacryoadenitis in males showed the same results, and so these were inherited recessively in mice with an MRL background as well. Sialadenitis in males showed an additive manner of inheritance because the MCF1 hybrids had significantly higher indexes than did the C3H parents (P < 0.001). 

MCF2 progeny developed various severities of dacryoadenitis and sialadenitis in the females (dacryoadenitis, 0.85 ± 0.54; sialadenitis, 0.41 ± 0.37), the males (dacryoadenitis, 0.19 ± 0.27; sialadenitis, 0.49 ± 0.37), and both (dacryoadenitis, 0.49 ± 0.53; sialadenitis, 0.45 ± 0.37). In the MCF2 progeny, the same tendency was seen—namely, that the dacryoadenitis and sialadenitis indexes were higher than in C3H parents and MCF1 progeny, but they presented a wide range of the indexes covering those of MRL parents. Dacryoadenitis scores of MRL/lpr and MCF2 progeny and sialadenitis scores of MRL/lpr in females were significantly higher than those in the males (P < 0.001, respectively). The penetrance of dacryoadenitis was higher in the females than in the males in the MCF2 intercrossed mice, similar to their MRL parents. 

Distribution of Indexes of Both Lesions in MCF2 Progeny

In the next series of experiments, we investigated whether dacryoadenitis and sialadenitis developed associatively in MCF2 progeny (Fig. 4) . The individual index of both lesions was discretely distributed. Some of the individuals showed higher indexes in both lesions, some of them were higher in either dacryoadenitis or sialadenitis, and others were lower in both lesions. Only a small positive correlation was observed between dacryoadenitis and sialadenitis (R ² = 0.06411, P < 0.001). 

Mapping of Dacryoadenitis and Sialadenitis Susceptibility Loci in MCF2 Progeny

A genome-wide scan was performed as described in the Materials and Methods section on the 266 MCF2 progeny with a total of 119 microsatellite markers, and an association study was conducted. In chromosomes 1, 2, 5, and 9, there was at least one microsatellite marker on one chromosome that fulfilled the standard of P < 0.01. 

A QTL analysis of these candidate chromosomes using all 527 MCF2 progeny demonstrated that chromosomes 1 and 2 had the highest LOD scores with significant linkage for both dacryoadenitis and sialadenitis. The highest peak on chromosome 5 was limited to one with significant linkage only for dacryoadenitis (Fig. 5) . The highest LOD scores for chromosome 1 were 4.0 (map position 64.1 cM) for dacryoadenitis and 16.8 (69.0 cM) for sialadenitis. These positions were susceptibility loci in the region of an MRL allele and were designated Adacm1 and Asm3. On chromosome 2, the highest LOD score for dacryoadenitis was 3.7 (88.4 cM) and sialadenitis was 5.6 (65.3 cM). These positions were also susceptibility loci in the region of an MRL allele, designated Adacm2 and Asm4, respectively. A LOD score of the second highest peak on chromosome 2 was 5.4 (82.1 cM) for sialadenitis, close to Asm4, also a susceptibility locus in the region of an MRL allele, designated Asm5. On chromosome 5, the highest LOD score for dacryoadenitis was 3.1 (63.9 cM), also a susceptibility locus in the region of an MRL allele, designated Adacm3. Regarding chromosome 9, there were no loci with a significant linkage, but only suggestive (LOD score; 2.6) for dacryoadenitis in the females. 

The best model to fit the inheritance of these susceptibility loci was determined with the WinQTL Cartographer program. The results are summarized in Table 1 . Adacm1 and Asm3, -4, and -5 corresponded to an additive manner of inheritance, whereas Adacm1 and -2 corresponded to recessive. These adjustments were confirmed by nonparametric tests. None of these loci was protective. As shown in Figure 5 , Adacm1 was located very close to Asm3 and Adacm2 was to Asm5. Each of the two loci seemed to be located in the common chromosomal region, whereas Adacm3 and Asm4 were independent from any other loci. 

Interaction Analysis of the Loci

The interactions of the two loci for dacryoadenitis and sialadenitis were investigated to clarify the relationships among the multiple loci which were identified (Table 2) . Since Adacm2 on chromosome 2 and Adacm3 on chromosome 5 were recessive as shown in Table 1 , mice with H and C genotypes at the D2Mit285 and D5Mit136 markers, respectively, were grouped. The other four loci, that is to say Asm3, -4, and -5 and Adacm1, showed an additive mode in inheritance (Table 1) . It disclosed a complex pattern of interaction, so a comparison of M to C was examined at the related markers, respectively. 

First, each of the two loci, which seemed to be located in a common chromosomal region (Adacm1 and -2, or Asm3 and -5), were examined. As shown in Table 2A , the M-M group was significantly higher than the other three groups in dacryoadenitis and sialadenitis, and so each of the two loci has an additive effect. Furthermore, the M-H/C group was significantly higher than the C-H/C group in dacryoadenitis and the M-C group was significantly higher than the C-C group in sialadenitis, so Adacm1 and Asm3 were hierarchical against Adacm2 and Asm5, respectively. 

Next, the interactions of Adacm3 and the other two loci, Adacm1 and -2, were examined. As shown in Table 2B , there was no significant additive effect, which indicates that Adacm1 and -2 have a significant interaction, but Adacm3 is independent from these two loci. 

Discussion

The completion of the human and mouse genome projects has allowed genetic analysis to become a standard approach for identifying alleles for susceptibility to various diseases with multiple phenotypes. ³² A cumulative effect of such susceptibility genes resulting from genome crossing may be responsible for the diversity of pathologic manifestations of diseases. In this study, we analyzed the mode of inheritance of dacryoadenitis and sialadenitis in MRL/lpr mice, a mouse model of human SS, ^{10 11 12 33} since the lacrimal and submandibular glands in MRL/lpr mice are histopathologically characterized by mononuclear cell infiltration consisting of T cells, B cells, dendritic cells, and macrophages, into periductular regions, ^{34 35} followed by the progressive destruction of ductules and parenchyma, eventually leading to parenchymal atrophy and replacement of the glandular structures by fibrotic scar tissue. ³⁶ As a result, three susceptibility loci were identified that affect the development of dacryoadenitis and sialadenitis in MRL/lpr mice. These loci were independent of the susceptibility loci to other lupus phenotypes such as glomerulonephritis, vasculitis, and arthritis, which have been identified in the backcrosses and intercrosses of MRL/lpr and C3H/lpr mice. ^{17 18 19 20 21} Inheritance of dacryoadenitis and sialadenitis in MRL/lpr mice was polygenic, partly characterized by an additive effect and was hierarchical. Each of the polygenes only by itself is not remarkably efficient for disease development. It is likely that all these candidate genes are polymorphic genes in the MRL/Mp strain, different from those in the C3H/HeJ strain. These may quantitatively regulate the cascade reactions extending to the development of dacryoadenitis and/or sialadenitis, with a regular variation of the pathologic phenotypes based on their combinations. 

To our knowledge, this is the first study in which the susceptibility loci for dacryoadenitis have been examined. For sialadenitis, two susceptibility loci were identified previously: Asm1 on chromosome 10 and Asm2 on chromosome 4 in the MRL genetic background. ²¹ However, in the present study, we did not detect them in the same chromosomal regions. There were several differences between the earlier study and this one in the experimental methods. For example, MRL/lpr × (MRL/lpr × C3H/lpr) F1 (N2 backcross) was evaluated in the earlier study. However, the MCF2 progeny were examined in this study. In addition, an association study was conducted in the earlier study, but QTL was used in the present study. These distinctions may account for the differences in the positions of the susceptibility loci in both studies. Similar results were seen in the past study for sialadenitis. ²⁴ Namely, QTL was performed on the (NOD × C57BL/6(B6)) F2 cross and ((NOD × B6) × NOD) backcross and no common chromosomal region was identified except for that in chromosome 1. It is likely that in an N2 backcross generation, only recessive loci can be detected, but in an F2 cross, all loci with a recessive, dominant, or incomplete dominant (additive) mode of inheritance can be detected. This difference may lead to the result that all susceptibility to sialadenitis in the present study was due to a additive mode of inheritance, whereas all loci in our previous study were recessive. That is, the difference may depend on the mode of inheritance. 

In this study, Adacm1 was located very close to Asm3, whereas Adacm2 was close to Asm5. And Adacm2 and Asm5 as well as Adacm1 and Asm3 were overlapped in confidence intervals as shown in Figure 5 . Therefore, the two susceptibility loci to both lesions are a common genetic basis for the development of these two diseases, although the other was independent. In addition, each of the two loci has an additive effect. Therefore, dacryoadenitis and sialadenitis in MRL/lpr are under the control of different susceptibility loci with an allelic combination. 

In addition, a similar tendency is shown in human autoimmune diseases in terms of the partial commonality of susceptibility loci. The co-occurrence of autoimmune diseases is a common phenomenon. Monozygotic twins develop autoimmune disease, such as rheumatoid arthritis and systemic lupus erythematosus, ^{37 38} with a high concordance rate. As a result, genetic predisposition is a dominant factor in the development of autoimmune disease. Association and linkage studies of autoimmune diseases have been conducted in different populations and have demonstrated that several susceptibility loci overlap and the genetic factor consists of two types: one is common to many autoimmune diseases, and the other is specific to a given disorder. ³⁹  

In the present study, the common features for both lesions demonstrated that the susceptibility locus involving Adacm1 and Asm3 on chromosome 1 had the highest LOD score among each of the three loci (Table 1) . Furthermore, Adacm1 and Asm3 on chromosome 1 were hierarchical against Adacm2 and Asm5 on chromosome 2, respectively (Table 2) . Therefore, the locus on chromosome 1 is most important for development of both lesions, at least in the MRL/lpr allele. This locus was also seen as one with a suggestive linkage to sialadenitis in MRL/lpr × MCF1 backcross. ²¹ Moreover, a susceptibility locus to sialadenitis on chromosome 1 was also detected in a (NOD × B6) F2 cross, a (NOD × NZW) F2 cross, and a ((NOD × B6) × NOD) backcross. ²⁴ Therefore, this chromosomal region has a significant role, common to sialadenitis and dacryoadenitis. 

In contrast, Adacm3 on chromosome 5 was independent from the other two susceptibility loci to dacryoadenitis. Dacryoadenitis and sialadenitis were observed with varied severity in the MCF2 intercross, but the association of the two lesions was weak (Fig. 4) , thus suggesting that the two phenotypes could be separated in an MCF2 intercross. In fact, as a result of the QTL analysis of MCF2, Adacm3 had no association with either Adacm1 or -2 (Table 2) . These data suggest that Adacm3 may be associated with a specific mechanism in the development of dacryoadenitis. 

At the region with the highest peak of Adacm3 there were seven genes according to MGI; Evi25 (ecotropic viral integration site 25), Lhx5 (LIM homeobox protein 5), Mvk (mevalonate kinase), Selplg (selectin, platelet ligand), Tesc (tescalcin), Ubc (ubiquitin C), and Vps33a (vacuolar protein sorting 33A) on chromosome 5 (map position 64 cM), ²⁵ some of which may be responsible for inflammation and cell proliferation processes. Evi25 is an integration site of the murine leukemia retrovirus in murine leukemias. ⁴⁰ Lhx5 is a member of the LIM homeobox gene family that encodes a transcription factor ⁴¹ and Mvk codes for the mevalonate kinase (MK) which is the first enzyme in the isoprenoid biosynthesis pathway and the decrease in MK activity leads to inflammation. ⁴² Selplg codes for the P-selectin glycoprotein ligand-1 (PSGL-1), a mucin-like glycoprotein which is expressed on the surface membrane of all leukocyte that binds to P- and L-selectin and promotes cell adhesion in the inflammatory response associated with leukocyte rolling and extravasation. ^{43 44} UbC is one of two stress-inducible polyubiquitin genes and constitutes an essential source of Ub in maintaining cellular Ub during cell proliferation and stress and Vps33a is homologous to the yeast vacuolar protein-sorting vps33 which codes the protein that is required for trafficking newly synthesized proteins from the late Golgi network to the vacuole. ⁴⁵ Further studies should be performed to determine any allelic polymorphism in these genes between the two strains, at least in the coding and promoter regions. 

A marked sex bias of dacryoadenitis and sialadenitis in the MRL parent was observed in Figure 3 . A similar tendency was seen for dacryoadenitis in the MCF2 progeny, that is, females demonstrated higher severity than males. Therefore, QTL analysis was conducted separately in male and female mice. As a result, the similar findings to those of the both sexes were obtained. There was no new locus with a significant linkage on any chromosomes (data not shown). 

In conclusion, the present study demonstrated that the development of dacryoadenitis and sialadenitis were under the control of polygenes partly with an additive effect and a hierarchical manner. Furthermore, two susceptibility loci to dacryoadenitis seemed to be common to those to sialadenitis, whereas the other was independent from them. Therefore, dacryoadenitis and sialadenitis in MRL/lpr are considered to be under the control of common and different susceptibility genes, with an allelic combination that leads to regular variations in the clinical phenotypes. 

 Supported by Grant-in-Aid for Scientific Research (B) 18390123 of the Ministry of Education, Science and Culture of Japan (MN).

 Submitted for publication November 10, 2008; revised December 23, 2008, and January 18, 2009; accepted May 1, 2009.

 Disclosure: T. Kamao, None; T. Miyazaki, None; Y. Soga, None; H. Komori, None; M. Terada, None; Y. Ohashi, None; M. Nose, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Dissociation of histopathologic phenotypes of dacryoadenitis and sialadenitis in 5-month-old MCF2 intercrossed female mice. (HE stain) (A) Dacryoadenitis and sialadenitis, showing marked parenchymal destruction associated with mononuclear cell infiltration into the ductules and extending to the parenchyma. (B) Dacryoadenitis without sialadenitis. (C) Sialadenitis without dacryoadenitis.

Figure 2.

 

Histopathologic grades of dacryoadenitis (A) and sialadenitis (B) (HE stain). (1) Grade 0, no inflammatory cell infiltration; (2) grade 1, mononuclear cell infiltration with limited localization in the periductular regions, but no destruction of the parenchyma or ductules; (3) grade 2, cell infiltration into the duct epithelium extending to the acinar parenchyma; and (4) grade 3, significant parenchymal destruction with regenerative ductal hyperplasia and fibrotic change.

Figure 3.

 

A quantitative analysis of dacryoadenitis and sialadenitis indexes (mean and SD) in MRL/lpr, C3H/lpr, their F1 and F2 progenies in both (A), females (B), and males (C). The grade score of each mouse is presented by an index that is the mean of the individual grade of each ductule randomly selected on the histologic section (circle). A comparison of the means was performed among the four groups by Tukey-Kramer test. *P < 0.001, and between females and males in MRL/lpr and MCF2 progeny using unpaired t-test, †P < 0.001.

Figure 4.

 

Correlation between dacryoadenitis and sialadenitis in MCF2 (n = 527). The indexes of each mouse are presented in the x-axis (sialadenitis) and y-axis (dacryoadenitis). Their correlation was estimated by using the Pearson product-moment correlation coefficient. There was only a slight positive correlation between both (R ² = 0.06411; P < 0.001). Line of regression (heavy line) and confidence interval (thin line) are shown on the graph.

Figure 5.

 

Plots of the LOD scores for QTLs that control lacrimal and submandibular lesions in MCF2 progeny. Indexes for chromosomes with significant linkages for dacryoadenitis (chromosome 1, 2, and 5) and sialadenitis (chromosomes 1 and 2) are shown. Indexes for composite interval mapping for dacryoadenitis (dotted line) and sialadenitis (solid line) are shown in the highest mode for the LOD score on each chromosome (a free mode for all chromosomes). Horizontal lines and their values (dacryoadenitis and sialadenitis index, respectively) show the thresholds of significance levels (α = 0.01), with the dotted line representing the dacryoadenitis index and the solid line representing the sialadenitis index, all obtained by analyzing 10,000 random shuffling permutations of the actual data. Triangles on the x-axis indicate the data for the following microsatellite markers (positions based on information from the Mouse Genome Informatics maintained by The Jackson Laboratory, Bar Harbor, ME; online at http://www.informatics.jax.org) for chromosome 1: D1Mit276, D1Mit18, D1Mit22, D1Mit46, D1Mit49, D1Mit187, D1Mit286, D1Mit394, D1Mit202, D1Mit143, D1Mit291, and D1Mit293; for chromosome 2: D2Mit355, D2Mit83, D2Mit522, D2Mit323, D2Mit380, D2Mit37, D2Mit102, D2Mit441, D2Mit395, D2Mit304, D2Mit258, D2Mit22, D2Mit285, D2Mit50, and D2Mit200; and for chromosome 5; D5Mit145, D5Mit74, D5Mit149, D5Mit233, D5Mit134, D5Mit259, D5Mit23, D5Mit115, D5Mit136, D5Mit33, and D5Mit223. The 95% confidence intervals of the QTL span were calculated according to the LOD drop-off method described by Lander and Botstein ³¹ and represented as a solid bar for each QTL.

Table 1.

 

Summary of the Susceptibility Loci for Dacryoadenitis and Sialadenitis in MCF2 Progeny

Phenotype	Chromosome	QTL											Related Marker (Position, cM), ∥	Genotype, ¶	Mice (n)	Index (Mean ± SD)	Comparison, #	P , **
				Designation	Map Position (cM)	LOD Score*	P , †	Inheritance, ‡	Confidence Interval (cM), §
Dacryoadenitis	1	Adacm1	64.1	4.0	0.0002	Additive	31.0	D1Mit187 (62.0)	M	120	0.66 ± 0.60	M vs. H	0.004
									H	255	0.48 ± 0.52	M vs. C	3 × 10−5
									C	136	0.37 ± 0.45	H vs. C	0.04
	2	Adacm2	88.4	3.7	0.0002	Recessive	17.3	D2Mit285 (86.0)	M	134	0.65 ± 0.59	M vs. H/C	6 × 10−5
									H	256	0.44 ± 0.50
									C	121	0.37 ± 0.45
	5	Adacm3	63.9	3.1	0.03	Recessive	6.8	D5Mit136 (65.0)	M	120	0.62 ± 0.58	M vs. H/C	4 × 10−5
									H	292	0.47 ± 0.53
									C	99	0.41 ± 0.45
Sialadenitis	1	Asm3	69.0	16.8	3 × 10−17	Additive	6.2	D1Mit286 (67.0)	M	126	0.66 ± 0.41	M vs. H	2 × 10−7
									H	260	0.45 ± 0.34	M vs. C	3 × 10−18
									C	130	0.26 ± 0.27	H vs. C	3 × 10−8
	2	Asm4	65.3	5.6	2 × 10−6	Additive	18.6	D2Mit395 (66.9)	M	141	0.56 ± 0.41	M vs. H	0.008
									H	259	0.45 ± 0.36	M vs C	1 × 10−7
									C	116	0.31 ± 0.30	H vs C	0.0002
	2	Asm5	82.1	5.4	2 × 10−5	Additive	8.0	D2Mit22 (84.0)	M	134	0.57 ± 0.41	M vs. H	0.0007
									H	261	0.44 ± 0.36	M vs. C	1 × 10−7
									C	121	0.34 ± 0.32	H vs. C	0.02

*  Determined with WinQTL Cartographer ver 2.5 by composite interval mapping (model 6). Genome-wide significance levels were determined after 10,000 permutations.

†  Nonparametric P (Kruskall-Wallis test).

‡  Model of inheritance determined with WinQTL Cartographer ver 2.5.

§  95% confidence intervals of the QTL span were calculated according to the LOD drop-off method described by Lander and Botstein. ³¹

∥  Closest marker to the QTL. Map position of the marker was based on the Mouse Genome Informatics (MGI) database.

¶  MRL/MRL and C3H/C3H homozygotes and MRL/C3H heterozygotes were designated M, C, and H, respectively.

#  An additive mode of inheritance disclosed a complex pattern of interaction, and so all combinations of possible genotypes were considered. H and C genotypes at the D2Mit285 and D5Mit136 markers, which were a recessive mode of inheritance, were grouped, respectively. Then, the group of M genotype and the group of H and C genotypes were analyzed.

**  Nonparametric P (Mann-Whitney U test).

Table 2.

 

Interactions among Each of the Two Loci for Dacryoadenitis and Sialadenitis in MCF2 Progeny

A. Interactions among Each of the Two Loci Located on Chromosome 1 and 2
Phenotype	Genotype of the Marker*			Mice (n)	Index (Mean ± SD)	P , †
Dacryoadenitis	D1Mit187 (Adacm1)	D2Mit285 (Adacm2)
	M	M	32	(0.86 ± 0.63)	, ‡	, §		, ∥
	M	H/C	88	(0.58 ± 0.57)			, ‡
	C	M	35	(0.42 ± 0.47)
	C	H/C	101	(0.36 ± 0.45)
Sialadenitis	D1Mit286 (Asm3)	D2Mit22 (Asm5)
	M	M	29	(0.85 ± 0.45)	, §	, ∥		, ∥
	M	C	33	(0.55 ± 0.32)			, ∥
	C	M	32	(0.32 ± 0.27)	, ‡
	C	C	30	(0.11 ± 0.18)
B. Interactions of Adacm3 and the Other Two Loci
Phenotype	Genotype of the Marker*			Mice (n)	Index (Mean ± SD)	P , †
Dacryoadenitis	D1Mit187 (Adacm1)	D5Mit136 (Adacm3)
	M	M	28	(0.83 ± 0.64)		, ∥
	M	H/C	92	(0.60 ± 0.58)	, ∥
	C	M	38	(0.50 ± 0.56)
	C	H/C	98	(0.32 ± 0.40)
Dacryoadenitis	D2Mit285 (Adacm2)	D5Mit136 (Adacm3)
	M	M	38	(0.77 ± 0.57)		, ∥
	M	H/C	96	(0.60 ± 0.59)	, §
	H/C	M	82	(0.55 ± 0.57)
	H/C	H/C	295	(0.41 ± 0.47)

*  Genotype of the marker closest to the QTL was presented. MRL/MRL and C3H/C3H homozygotes and MRL/C3H heterozygotes were designated M, C and H, respectively.

†  A comparison of the means was performed using the Tukey-Kramer test.

‡   P < 0.05;

§   P < 0.01;

∥   P < 0.001.

The authors thank the medical students of Ehime University; and Aoi Sukeda, Yuko Uenaka, and Mami Kobayashi for technical help with the simple sequence-length polymorphism analysis.