| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Rheumatology Section and Departments of
Histopathology and
Transplantation Biology, Imperial College School of Medicine, Hammersmith Hospital, London, United Kingdom
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2 = 15.68, p = 7.5 x
10-5), as was D1Mit12 at 63.1 cM
(2 = 20.17, p = 7.1 x
10-6). D1Mit403 at 100 cM was linked to
anti-dsDNA Ab production (2 = 17.28,
p = 3.2 x 10-5). Suggestive
linkages to antinuclear Abs and nephritis were identified on chromosome
3, to splenomegaly on chromosome 4, and to anti-ssDNA Ab production
on chromosome 10. Chromosome 4 and the telomeric region of chromosome 1
have previously been linked to disease in other mouse models of
systemic lupus erythematosus; however, the centromeric regions of
chromosome 1 and chromosomes 3 and 10 are unique to BXSB. This implies
that, though some loci may be common to a number of mouse models of
lupus, different clusters of disease genes confer disease
susceptibility in different strains of mice. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Studies of reciprocal F1 hybrids between BXSB and NZW mice have shown that disease acceleration is linked to a gene on the Y chromosome designated the Y-linked autoimmune accelerating (Yaa) gene (4). Experiments using consomic mice, in which the Y chromosome was transferred to different mouse strains by backcrossing, showed that the Yaa gene was not sufficient to cause an autoimmune response on a putatively nonautoimmune B6 (6) or CBA/J (7) background. In contrast, disease activity was enhanced when the gene was introduced into the SLE-prone MRL (8) and NZW strains (7). Thus the action of the Yaa gene is dependent on the presence of other genes. In the BXSB mouse, these other genes have a dominant effect, since male F1 hybrids produced from B6 female and BXSB male mice (Yaa+) develop splenomegaly and have accelerated mortality, compared with male mice from the reciprocal cross (4). Indeed, the lupus phenotype in BXSB mice has been shown to be influenced by the expression of particular MHC alleles, with BXSB.H-2d congenic male mice displaying prolonged survival (100% at one year) compared with the parental BXSB (H-2b) mice (10% at one year) (9). However, genes other than those of the MHC must also be involved in lupus susceptibility in mice because B6 and B6.Yaa (H-2b) male mice do not develop disease (6).
Major non-MHC loci on chromosomes 1, 4, and 7 have been linked to autoantibody production and nephritis in (NZB x NZW)F1 (10, 11, 12, 13, 14, 15) and NZM/Aeg2410 (16) lupus mouse models. Loci associated with susceptibility to nephritis have also been identified on chromosomes 7 and 12 in MRL/lpr mice (17). In the present study we have investigated the genetic basis of the lupus phenotype in the BXSB mouse. We have analyzed the phenotype of F1 male mice produced by crossing C57BL/10 (B10) nonautoimmune females with BXSB males (F1Yaa+). We have explored the nature of the disease by comparing the phenotype of these F1Yaa+ males with those of the parental strains and with male mice from a (BXSB x (B10 x BXSB)F1) backcross. Using this information, we have conducted a genome-wide linkage analysis of the (BXSB x (B10 x BXSB)F1) backcross mice to identify disease susceptibility loci associated with the disease phenotype.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
All the mice analyzed were male. BXSB mice were obtained from the Jackson Laboratory (Bar Harbor, ME) and bred and maintained at the Hammersmith Hospital. C57BL/10 mice were obtained from Harlan Olac (Bicester, U.K.). Male F1 mice and backcross mice were generated by local breeding and maintained under identical conditions. Mice were free from disease pathogens. B10 mice were selected for this initial analysis for two main reasons: first, to fix the H-2 region (H-2b) since this study was to identify non-MHC genes; second, to keep the analysis as simple as possible by not introducing additional genetic variation from a third strain. This study does not analyze X chromosome effects and therefore (BXSB x (B10 x BXSB)F1) and ((B10 x BXSB)F1 x BXSB) male mice were analyzed together. Genomic DNA for genotypic analysis was extracted from 300 µl whole blood collected in 20 mM EDTA, pH 8.0, using the Puregene DNA isolation kit (Gentra Systems, Minneapolis, MN) according to manufacturer’s instructions.
Serologic analyses
Mice were bled from the tail vein at 2 to 4 monthly intervals, and serum was stored at -70°C before analysis. Serologic disease markers were assessed in cohorts of 35 BXSB, 49 B10, and 42 F1Yaa+ male mice from 4 mo of age (different mice were analyzed at different time points reflecting the amount of sera available at a particular age). Backcross mice were bled every 2 mo from 8 mo of age and sacrificed at 12 mo.
Anti-nuclear Abs (ANA) were measured by indirect immunofluorescence (18) using HEp-2 cells and a fluoroscein-conjugated IgG Fc-specific anti-mouse secondary Ab (Sigma, Poole, U.K.). BXSB, B10, and F1 samples were screened at a 1/40 dilution and backcross samples at 1/160. Positive samples were titrated to end point, and results were expressed as ANA titer.
Anti-dsDNA and anti-ssDNA Abs were individually measured by modifications of an ELISA (19). All reagents were obtained from Sigma, unless otherwise stated. Nunc Maxisorp immuno-plates (Nunc, Roskilde, Denmark) were sensitized with 50 µl methylated BSA at 10 µg/ml in PBS for 1 h at 37°C. Plates were then coated overnight at 4°C with 50 µl calf thymus dsDNA at 10 µg/ml in PBS or with 50 µl ssDNA, produced by boiling the dsDNA for 15 min. Contaminating ssDNA was removed from the dsDNA Ag by an S1 nuclease digest (5 U/µg DNA) at 37°C for 30 min. Plates were blocked with PBS containing 2% BSA, and sera were assayed in 0.1 M Tris HCl, pH 7.2, containing 2% bovine gamma globulin (Cohn fraction II,III) and 1% BSA, at a dilution of 1/25 for anti-dsDNA and 1/75 for anti-ssDNA. The alkaline phosphatase-conjugated anti-mouse IgG was diluted 1/2000 in PBS containing 0.1% Tween and 0.5% BSA. Results were expressed in ELISA Units (EU) relative to a standard positive sample of BXSB serum, which was assigned a value of 100, such that (results in EU) = (100 x OD of sample)/(OD of positive control).
IgG3, IgG2b, and C3 levels were individually measured by radial immunodiffusion (20) using the respective polyclonal anti-mouse IgG3 or IgG2b or C3 (The Binding Site, Birmingham, U.K.) in a 1.2% PBS-agarose gel. Results were calculated in mg/ml from a standard curve generated by serial dilution of Ig or C3 calibrator (The Binding Site).
Hemolytic complement activity was determined by a fluid phase 51Cr release assay using rabbit erythrocytes (Tissue Culture Services, Aylesbury, U.K.) as previously described (21). Complement activity was then expressed in hemolytic units (HU) with respect to a standard B10 serum pool.
Autopsy analysis
BXSB and F1 Yaa+ animals were sacrificed when unwell (8–12 mo and 20 mo, respectively), with control B10 mice sacrificed at 8 to 14 mo. Backcross mice were sacrificed at 12 mo to obtain age-matched autopsy specimens. Spleen and lymph nodes (inguinal, axillary, and cervical) were weighed (wet weight). Spleen weight and average lymph node weight were expressed as a percentage of total body weight. Kidneys were fixed in 10% formal saline and stained with hematoxylin and eosin. Coded sections were scored for glomerulonephritis according to the intensity and extent of the histopathologic changes, as described previously (22). Glomerular hypercellularity was graded on a 0 to 4 scale where 0 = no involvement, and grades 1–4 = <25%, 25 to 50%, 51 to 90%, and >90% abnormal glomeruli, respectively. Mesangial matrix increase was graded on a 0 to 3 scale where 0 = no increase, and grades 1 to 3 = mild, moderate, and severe increase, respectively.
Genotyping of backcross mice
Microsatellite primer pairs were synthesized by OSWEL, Southampton, U.K. One oligonucleotide from each marker pair was fluorescently labeled with the dyes 6-FAM, TET, and HEX, and the labeled oligonucleotide was subsequently HPLC purified. Microsatellite markers were amplified by the PCR, and all reaction volumes were 15 µl. Multiplex PCR was conducted with between two and four microsatellites being coamplified in the same PCR reaction. Two buffer systems were used in multiplex amplifications. System A contained 1x Taq buffer (Life Technologies, Paisley, Scotland), 0.02% W-1 detergent, 250 µM each of four dNTPs, 1.5 mM MgCl2, 0.17 µM each oligonucleotide, and 1.9 U Taq DNA polymerase (Life Technologies). System B contained 1x Taq buffer, 10% DMSO, 1.0 mM each of four dNTPs, 4.5 mM MgCl2, 0.33 µM each oligonucleotide, and 1.9 U Taq DNA polymerase. In both cases, 40 ng mouse genomic DNA was added. PCR reactions were conducted in a 96-well thin-walled microtiter plate. The PCR protocol was 3 min at 94°C followed by 26 cycles of 94°C for 45 s, 55°C or 50°C for 1 min, and 72°C for 2 min. PCR reactions were pooled, and 0.8 µl was added to 1 µl deionized formamide and 0.2 µl TAMRA-labeled GS-350 size standard (Perkin-Elmer, Foster City, CA). Samples were analyzed on an ABI 377 Automatic DNA sequencer according to manufacturer’s instructions (Perkin-Elmer). Genotypes were stored in the Excel 5 spreadsheet package (Microsoft, Redmond, WA).
Statistical methods
B10 mice were used as controls. The mean value and SD obtained
for B10 mice at all ages were used to define normal and abnormal values
for the disease markers analyzed. Nonparametric methods of analysis
were otherwise used throughout, and results are expressed as median
(range) values. Differences between groups were analyzed using the
Mann-Whitney U test. The log rank test was used to compare
survival curves between groups. The correlation between genotypes and
phenotypes (2) was calculated using a standard 2 x
2 contingency table with one degree of freedom. Only mice that were
categorized as positive or negative for a trait were included in the
linkage analysis. The other mice, categorized as intermediate, were
excluded. Since the disease traits used to analyze the backcross mice
were highly correlated, it would be excessively conservative to apply a
standard correction factor of one additional degree of freedom per
trait analyzed. We therefore analyzed the results on the basis that the
study involves the equivalent of a single genome-wide screen and that,
hence, the p values were determined by the look-up function
in the Microsoft Excel 5 program (Microsoft) and were not
corrected for multiple testing.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Mice were not assessed for any other phenotypic marker than life span, and only males were analyzed. Median mortality was 11 mo in 17 BXSB mice, 25 mo in 11 B10 mice, and 19 mo in 17 F1Yaa+ mice. Mortality in BXSB mice was delayed compared with previous reports, and this may reflect local environmental factors (1, 3, 4). Mortality in BXSB and F1Yaa+ mice was significantly accelerated compared with B10 mice, hazard ratios (95% C.I.) 9 (3.2, 25.8), p < 0.001 and 2.96 (1.43, 6.13), p = 0.01 respectively. However, mortality was significantly delayed in F1 Yaa+ mice compared with BXSB mice, hazard ratio 0.33 (0.11, 0.93), p < 0.001.
Determination of autoimmune phenotypes
The analysis of BXSB, B10, and F1Yaa+ male mice was used to identify markers of disease activity for mapping disease susceptibility genes segregating in (BXSB x (B10 x BXSB)F1) backcross mice. The results obtained in B10 and BXSB mice were used to establish the criteria defining positivity and negativity for each trait, and these criteria were initially applied to the F1Yaa+ mice. Absolute values, based on the mean value and SD recorded in B10 mice were set for each marker defining normal (B10) and abnormal (BXSB) values. These values were selected so that the vast majority of B10 analyses were classified as normal and the vast majority of BXSB analyses were classified as abnormal. Values that overlapped between BXSB and B10 mice were categorized as intermediate. Serial longitudinal analysis of mice was undertaken to categorize each mouse as positive or negative for each serologic disease trait, as discussed below. All phenotypic classification was conducted before genotypes were assigned to the animals to avoid bias.
Serologic analysis
None of the B10 mice developed ANA at titers of 1/640 or above
(Fig. 1
a) and ANA was detected
at 1/320 in only five mice. One of these had ANA at 1/320 at both 18
and 20 mo, but this was not sustained at 22 to 26 mo. In contrast,
three of the six BXSB mice analyzed at 12 mo had ANA 
1/640, as
did one of the three surviving mice at 14 mo (Fig. 1b); only
one of the BXSB mice analyzed at 12 mo had ANA 
1/160. An ANA
titer 1/640 was therefore designated as abnormal or elevated,
and a titer 1/160 as normal. ANA developed later in F1Yaa+ than in BXSB mice, and, at 8 to 9
mo, only one of 33 F1 Yaa+ mice had
elevated ANA compared with four of 13 BXSB. However, ANA titers in
F1 Yaa+ mice exceeded those in BXSB
(Fig. 1, c and b), and titers of 1/1280 or
greater were demonstrated for at least one time point in 23 of the 37
F1 Yaa+ mice analyzed for ANA but
only three of the 33 BXSB mice. Analysis of (BXSB x (B10 x
BXSB)F1) mice revealed that the trait of high ANA titers
had segregated, such that 45 of 185 (24%) mice had abnormal ANA at 8
to 9 mo, while 104 (56%) had normal values (Fig. 1d). Thus,
elevated ANA titers developed most rapidly in BXSB mice, followed by
backcross animals and then F1 Yaa+.
|
a). Two of the B10 mice had
Ab levels exceeding 4.4 EU (mean + 3 SD). This was detected before
sacrifice in one of the mice and was not sustained on subsequent
analysis in the other. None of the B10 mice had anti-dsDNA Abs
6.6 EU (mean + 5S D) whereas, at 10 mo, seven of eight
BXSB mice had anti-dsDNA Ab levels exceeding 4.4 EU and six had
levels exceeding 6.6 EU (Fig. 2b). Hence, anti-dsDNA Ab
levels 6.6 EU were designated abnormal or elevated. Eight B10
mice had anti-dsDNA Ab levels > 2 EU (mean + 1 SD), although this
was sustained in only three mice. None of the BXSB mice aged 10 mo had
levels 2 EU (Fig. 2b). Anti-dsDNA Ab
levels 2 EU were therefore designated normal. F1Yaa+ mice developed abnormal
anti-dsDNA Abs although the age of onset was markedly delayed
compared with BXSB mice (Fig. 2c), and, at ten mo, only two
of 11 mice had elevated Ab levels. In contrast, Abs developed earlier
in a proportion of (BXSB x (B10 x BXSB)F1)
backcross mice, and, at 8 to 9 mo, 37 of 188 (20%) had abnormal levels
of anti-dsDNA Abs (Fig. 2d).
|
28 EU
(mean + 3 SD) were designated abnormal or elevated and levels
12 EU normal (mean + 1 SD). Mean IgG3 level in B10 mice was 0.85 mg/ml
(±0.58) and IgG3 levels 2.59 mg/ml (mean + 3 SD) were
designated abnormal or elevated and levels 2.01 mg/ml (mean + 2
SD) normal. A summary of the findings on levels of IgG3 in all groups
is presented in Table I.
|
Mean hemolytic complement activity in B10 mice was 1.03 HU (±0.6)
(Fig. 3a). Hemolytic
complement activity decreased with age in BXSB mice. At 6 mo none of
the eight BXSB mice had hemolytic complement activity 0.4 HU
(mean -1 SD): all had activity 0.3 HU (Fig. 3b). A
value 0.4 HU was therefore designated normal, and a value
0.3 HU designated abnormal for hemolytic complement activity.
Comparison of F1 Yaa+ and BXSB mice
revealed that F1 Yaa+ mice did not
develop a comparable reduction in complement activity with age, and
only four of 13 mice had abnormal hemolytic complement activity at 20
mo (Fig. 3c). However, median hemolytic complement activity
in F1 Yaa+ mice was significantly
lower than that in B10 mice at all ages (Fig. 3c). For
example, median complement activity in F1Yaa+ mice aged 4 mo was 1.0 HU (0.7–1.5)
compared with 1.7 HU (1.3–1.9) in B10 mice (U = 73,
p < 0.001). From data shown in Figure 3d,
it is clear that this trait was segregating in the backcross, although
the distribution of activity meant some animals were intermediate
(Table I
).
|
The criteria for defining positivity and negativity for a trait were selected such that nearly all B10 mice were categorized as negative while the majority of BXSB mice were positive. Mice that were difficult to assign to either grouping with confidence were classified as intermediate. Serial analysis of individual BXSB and F1Yaa+ mice at different time points revealed that the age of onset of the development of abnormal values varied in individual mice, and values waxed and waned with disease as previously reported (23). Therefore, no single time point could be used to establish positivity and negativity for each serologic disease trait. Consequently, backcross mice were analyzed longitudinally, at a number of time points. Mice were categorized as negative for a particular trait if normal values were demonstrated at two or more time points with no intermediate or positive values, whereas they were categorized as positive if an abnormal value was demonstrated at any time point. These criteria were modified slightly for hemolytic complement activity because a relatively high proportion of B10 mice (9 of 50) had abnormal complement activity, although this was transient in eight of the mice. Hence, mice with transient abnormal complement activity were categorized as intermediate rather than positive.
Using these criteria, the majority of B10 mice were categorized as
negative for each disease trait, with 12.5% positive (Table I).
In contrast, the majority of BXSB mice were categorized as positive for
each trait, with 19% negative (Table I). Less than 28% of BXSB
and B10 mice were categorized as intermediate for any trait. Analysis
of older F1 Yaa+ mice revealed that
the majority were categorized as positive for ANA, anti-DNA, and
IgG3 (Table I). More than 90% were categorized as positive for ANA and
IgG3, although only 57% were positive for anti-ssDNA Abs, 60%
were positive for anti-dsDNA Abs, and 30% were positive for
reduced hemolytic complement activity.
Sufficient data were available to categorize at least 199 (BXSB x
(B10 x BXSB)F1) mice for all of the serologic disease
markers. Additional mice were analyzed for autoantibodies and IgG3
levels depending on serum availability. (BXSB x (B10 x
BXSB)F1) mice segregated in terms of positivity and
negativity for each of the disease markers with 37 to 62% of mice
being categorized as positive and 29 to 44% being categorized as
negative for each trait (Table I).
Histologic analysis
Spleen and lymph node size did not alter with age in B10 mice, and
mean relative spleen weight in 19 B10 mice aged 8 to 20 mo was 0.38%
(±0.18) total body weight, and mean relative lymph node weight was
0.017% (±0.009) (data not shown). Splenomegaly was therefore defined
as spleen weight 0.92% total body weight (mean + 3 SD) and
lymphadenopathy as weight 0.04% (mean + 3 SD). Normal spleen
weight was defined as weight 0.56% (mean + 1 SD) and normal
lymph node weight as weight 0.026% (mean + 1 SD). Using these
criteria, all of the five BXSB mice and none of the 19 B10 mice had
splenomegaly or lymphadenopathy (Table I). Seventeen of the B10 mice
had normal spleen weight, and 16 had normal lymph node weight. At 20
mo, only four of the eight F1 Yaa+
mice had splenomegaly, and one had normal spleen weight, although three
of the mice had lymphadenopathy, and four had normal lymph node weight
(Table I). In contrast, there was a clear segregation of both spleen
and lymph node weight in (BXSB x (B10 x
BXSB)F1) mice: 49% of 181 (BXSB x (B10
x BXSB)F1) mice had splenomegaly while 26% had normal
spleen weight, whereas 40% of 180 (BXSB x (B10 x
BXSB)F1) mice had lymphadenopathy, while 42% had normal
lymph node weight (Table I).
Kidney sections from 11 BXSB mice aged 8 to 14 mo and 14 B10 mice
aged 8 to 12 mo were analyzed. All of the BXSB mice had
hypercellularity affecting more than 25% of the glomeruli (
grade 2) and, in nine, more than 50% of the glomeruli were
affected ( grade 3) (Table II). In
contrast, 11 of the B10 mice aged 8 to 12 mo had no glomerular
hypercellularity, and the remaining three had hypercellularity
affecting less than 25% of the glomeruli only (Table II). Glomerular
hypercellularity grade 2 was therefore considered to be
indicative of nephritis, and a grade 1 indicative of normal
kidneys. Using these criteria, 10 of 11 BXSB mice had nephritis, and
all of the B10 mice had normal kidneys. One of the BXSB mice with
borderline grade 1/2 changes was categorized as intermediate (Table I).
Mesangial matrix increase was associated with hypercellularity in the
BXSB mice, and all but one of the mice had moderate or severe matrix
increase ( grade 2) (Table II). In contrast, none of the young B10
mice demonstrated moderate or severe matrix increase. In 7% (14/188)
of the (BXSB x (B10 x BXSB)F1) mice and in one
of the 14 F1 Yaa+ mice at 20
mo, mesangial matrix increase in the absence of hypercellularity was
observed (Table II). These mice were categorized as intermediate, as
were mice with borderline grade 1/2 changes. Using these criteria, four
F1 Yaa+ mice aged 20 mo had
nephritis while five had normal kidneys, whereas 45% of 188 backcross
mice aged 12 mo had nephritis and only 30% had normal kidneys (Table I).
|
We genotyped 226 (BXSB x (B10 x BXSB)F1)
mice to identify BXSB disease susceptibility loci. All backcross mice
analyzed were male (Yaa+), and H-2
was fixed since both B10 and BXSB are H-2b. Many other
regions of the genome are identical between the two strains since the
recombinant inbred strain, BXSB, derives approximately 50% of its
genome from C57BL/6, a strain very closely related to B10 (24). Hence,
434 microsatellite markers were analyzed for size polymorphism between
B10 and BXSB, and only 20% were informative. Table III describes the 64 polymorphic
microsatellite markers used in this study. The microsatellites are
distributed throughout the autosomes such that 85% of the genome is
within 20 cM of an informative genetic marker. This is probably an
underestimate of the proportion of segregating polymorphisms that are
within 20 cM of an informative marker in this backcross, due to the
close genetic relationship between BXSB and B10. For example,
preliminary data (M. Maibaum and B. J. Morley, unpublished
observations) indicate that the region between the centromere and
D3Mit137 (35.2 cM) on chromosome 3 is derived
predominantly from the B6 parental strain (eight microsatellite markers
covering the region between the centromere and D3Mit137 are
B6, not SB/Le, derived). Consequently, this region, like the H-2
region, is fixed in this cross. We are currently examining the BXSB
genome in detail, to determine the percentage of B6-derived genomic
material.
|
2 analysis (2 x 2 contingency table with 1
degree of freedom) was conducted to determine the degree of correlation
between individual disease traits within the backcross mice (data not
shown). Each of the disease markers was highly correlated to all other
markers (p = 9.6 x 10-3 to
3.3 x 10-15), except for the associations of
lymphadenopathy, firstly with hypocomplementemia
(p = 0.14) and also with elevated IgG3 levels
(p = 0.051), which were less well correlated.
All phenotypic classification was conducted before genotypes were
assigned to the animals to avoid bias, and mice categorized as
intermediate were excluded from the analysis. A summary of the
genome-wide linkage analysis for each of the phenotypic traits is shown
in Table IV. Only suggestive
(p < 3.4 x 10-3) or
significant linkages (p < 1 x
10-4) as defined by Lander and Kruglyak (25) are reported.
|
2 = 15.68, p = 7.5 x
10-5), as was D1Mit12 at 63.1 cM
(2 = 20.17, p = 7.1 x
10-6). D1Mit403 at 100 cM was linked to
anti-dsDNA Ab production (2 = 17.28,
p = 3.2 x 10-5). These three loci
have been designated Bxs1, Bxs2 and
Bxs3 respectively. Suggestive linkage of these three loci to
other disease traits was also observed. All chromosome 1 microsatellite
markers tested showed suggestive or significant linkage to at least one
of the disease traits. This suggests that multiple loci on chromosome 1
contribute to the production of the autoimmune disease observed in the
backcross mice.
Two markers on chromosome 3 also showed suggestive linkage to disease
traits. BXSB homozygosity at D3Mit217 (63.1 cM) was linked
to the development of ANA and nephritis (2 = 11.67 and
11.3, respectively). However, ANA production was also linked to BXSB
heterozygosity at D3Mit137 (35.2 cM) on chromosome 3
(2 = 10.27). This would suggest the presence of two
distinct susceptibility loci on chromosome 3 associated with production
of ANA, one of which may be a BXSB protective allele or a B10 disease
susceptibility allele. On chromosome 10, anti-ssDNA Ab production
was linked to D10Mit20 (2 = 12.31), with a
marked increase of BXSB homozygosity in unaffected backcross mice,
again implying that this locus may represent a BXSB protective or a B10
disease effect.
The central portion of chromosome 4 has been strongly linked to
nephritis (16) in the NZM/Aeg2410 strain (Sle2) and to both
nephritis and mortality (Lbw2, Nba1) in the
(NZB x NZW)F1 model (10, 12, 13). We tested markers
D4Mit199 and D4Mit72, which are located in this
implicated region of chromosome 4 and found suggestive linkage only
with lymphadenopathy (2 = 11.11 and 8.63, respectively).
No other phenotypic marker demonstrated linkage (2 < 2)
to either of these microsatellites. Chromosome 7 has also been strongly
linked to the development of nephritis and autoantibodies in both the
NZB/W (14) and MRL/lpr (17) mouse models of lupus; however,
we found no evidence of linkage to chromosome 7 in this analysis.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We have used the phenotype data taken from the study of BXSB, B10, and F1 mice to define criteria for selecting diseased and nondiseased mice in the backcross (BXSB x (B10 x BXSB)F1). For all the phenotypic characteristics used, backcross mice segregated into the two groups, with a proportion that were difficult to assign to either grouping with confidence. This third group were not included in the linkage analysis. A number of other studies have used similar methodology, studying individuals at the low and high ends of the distribution of a trait and excluding from the genetic analysis those individuals whose values were intermediate (26). One notable feature of this linkage analysis was the use of a number of distinct subphenotypes of lupus-like disease, to identify chromosomal regions that might be involved in specific aspects of the overall lupus phenotype. The presence of each of these phenotypes was highly correlated with each other in the (BXSB x (B10 x BXSB)F1) backcross mice, although lymphadenopathy was less correlated to the other lupus traits and was not correlated with hemolytic complement activity or IgG3. The phenotypes analyzed in this study are therefore related but nonidentical. This analysis was therefore used to dissect the genetic basis of individual characteristics of the lupus syndrome.
The genome-wide linkage analysis identified susceptibility loci for the
development of SLE in BXSB mice on chromosomes 1, 3, 4, and 10.
Chromosome 1 showed a broad interval of linkage covering most of the
chromosome, but three microsatellites showed significant peaks of
linkage, two with nephritis (D1Mit5 and D1Mit12)
and one (D1Mit403) with anti-dsDNA Ab production. This
would suggest that multiple loci are present on chromosome 1, and we
have termed these loci Bxs1, Bxs2, and
Bxs3, respectively. The interval defined as Bxs3
has previously been identified in studies of the lupus-prone mouse
strain (NZB x NZW)F1 (Lbw7 (10),
Nba2 (13), and Sle1 (16)). A number of candidate
genes lie in the region of Bxs3, including the Fc
receptor II and III genes, the genes encoding E-, L- and P-selectin,
and the Tgf-ß2 gene. The human FcRIIA gene has
been linked to the development of glomerulonephritis in both Caucasians
(27) and African-Americans (28), and we have identified coding
polymorphisms in FcRII between BXSB and B10 (29). This
remains a potential candidate disease susceptibility gene. More
recently, the region of human chromosome 1 syntenic with
Bxs3 (1q41-q42) has been associated with SLE in humans and
with the production of anti-chromatin Abs (30), in a study designed
following the results of these murine genome-wide linkage analyses.
However, physical distance seems to rule out the human Fc genes as
responsible for linkage to this interval since they lie in 1q23-q24.
D1Mit5 (Bxs1) and D1Mit12 (Bxs2) are located 30 cM apart on chromosome 1 and, therefore, it is possible that Bxs1 and Bxs2 represent two distinct susceptibility alleles. Bxs1 is unique to the BXSB mouse model, while linkages within 15 cM of Bxs2 have been identified in the (NZB x NZW)F1 model, Sbw1 (10) and D1Mit48 (15). The microsatellite D1Mit12 showed suggestive linkage to ANA production, high levels of IgG3, splenomegaly, low hemolytic complement activity, and anti-dsDNA Ab production in addition to significant linkage to nephritis. Therefore, Bxs2 would appear to have pleiotropic effects within these backcross mice. The issue of whether this study has identified the presence of one or two nephritis susceptibility loci on chromosome 1 is difficult to address, given the broad linkages observed across chromosome 1. The imperfect correlation between genotype and phenotype in all complex disease analysis reduces the ability to define a candidate region below a 15- to 20-cM region (31). The nature of Bxs1 and Bxs2 could be further investigated by analysis of the (BXSB x (B10 x BXSB)F1) backcross mice with additional informative microsatellites that are located between D1Mit5 and D1Mit12. However, the construction of congenic strains containing different portions of chromosome 1 from BXSB on a B10-derived genetic background is the only reliable method for identification of whether more than one nephritis-susceptibility gene resides on chromosome 1 of the BXSB mouse. This study is underway.
Chromosome 3 was suggestively linked to the development of ANA at D3Mit137 and ANA and nephritis at D3Mit217. Disease phenotype was linked to BXSB homozygosity for the microsatellite D3Mit217, as with all the loci on chromosome 1. For D3Mit137, however, the development of ANA was linked to BXSB/B10 heterozygosity. The skewing in the genotype distribution was seen in both trait positive and negative backcross mice, but there was no trend for linkage in the other phenotypic markers. This genotype distribution suggests one of two things: the presence of a protective gene derived from BXSB or the presence of a disease susceptibility gene derived from B10. This idea is supported by the fact that both ANA titers and incidence of positivity were higher in (B10 x BXSB)F1 animals than in the parental BXSB strain, i.e., the loss of a protective allele or gain of a disease allele in the F1 mice results in the observed, more severe development of ANA. This type of allele, where the "low susceptibility" strain (B10) provides a disease susceptibility allele or the "high susceptibility" strain (BXSB) a protective allele, is known as a transgressive allele (31). An allele of this type has previously been identified in the nonobese diabetic (NOD) mouse model where the locus, Idd7, is a disease susceptibility allele derived from the "low" susceptibility nonobese, nondiabetic (NON) mouse strain (32). The fact that the genotype distributions were different for the two loci on chromosome 3 and that they lie 28 cM apart suggests that two susceptibility alleles are present on this chromosome. The single suggestive linkage observed on chromosome 10 may represent a second transgressive trait locus.
The linkages identified in this study at D1Mit5 (Bxs1) and D1Mit3 toward the centromeric portion of chromosome 1 are the first linkages of lupus phenotypes to this region of chromosome 1. These may represent BXSB/B10-specific susceptibility alleles, whereas Bxs2 and Bxs3 may represent susceptibility alleles shared with other lupus models. No linkages have been previously identified on chromosome 3, and therefore these, presumably, are also BXSB/B10-specific effects. A major difference between the pattern of disease susceptibility in the BXSB mouse and the (NZB x NZW)F1 model of lupus is the lack of significant linkages of autoantibody production and glomerulonephritis to chromosomes 4 and 7 in this linkage analysis. The region between D4Mit9 (44.5 cM) and D4Mit70 (65.4 cM) has been strongly linked to the development of glomerulonephritis in a number of studies (10, 12, 14, 16). Chromosome 7 loci have also been strongly linked to lupus traits in a number of studies (10, 11, 14, 16, 17). The three chromosome 7 markers used in this study showed no indication of linkage to the lupus traits analyzed, and therefore the chromosome 7 susceptibility allele(s) present in the (NZB x NZW)F1-derived models do not appear to be present in the backcross population analyzed here. It is possible that, within the same intervals identified in linkage studies using different mouse models, Bxs3 (Sle1, Lbw7, Nba2) on chromosome 1 and Nba1 (Lbw2, Sle2) on chromosome 4, some genetic factors may be common. Others, e.g. Bxs1, may be unique to a particular mouse strain. This is consistent with the phenotypic differences displayed by the different strains and has important implications for the genetic analysis of human lupus. In SLE in humans, different patients demonstrate clinical symptoms as diverse as those seen in the different mouse models. Linkage analysis in humans may therefore be complicated by a different array of genes conferring susceptibility in each individual patient. It is possible that some genes may be common within patient groups, while others may be much more restricted in their distribution and therefore difficult to identify in a genome-wide linkage analysis. The identification of the genes involved in mice, and the characterization of their defects first at the molecular level and thereafter at the cellular level, will contribute to understanding the disease process and facilitate future human studies.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 These authors contributed equally to this work.
3 Address correspondence and reprint requests to Dr. Bernard J. Morley, Rheumatology Section, Division of Medicine, Imperial College School of Medicine, Hammersmith Campus, Du Cane Road, London, W12 0NN, U.K.
4 Abbreviations used in this paper: SLE, systemic lupus erythematosus; Yaa, Y-linked autoimmune accelerator; ANA, anti-nuclear Abs; B10, C57BL/10; B6, C57BL/6; EU, ELISA units; HU, hemolytic units; MGD, Mouse Genome Database.
Received for publication February 5, 1998. Accepted for publication May 18, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
RII gene. Immunogenetics 46:361.[Medline]
This article has been cited by other articles:
|
|
 |
|
  M. J. Ryan Young Investigator Award Lecture of the APS Water and Electrolyte Homeostasis Section, 2008: The pathophysiology of hypertension in systemic lupus erythematosus Am J Physiol Regulatory Integrative Comp Physiol, April 1, 2009; 296(4): R1258 - R1267. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. J. Rogers, M. J. Lees, L. Gabriel, E. Maniati, S. J. Rose, P. K. Potter, and B. J. Morley A Defect in Marco Expression Contributes to Systemic Lupus Erythematosus Development via Failure to Clear Apoptotic Cells J. Immunol., February 15, 2009; 182(4): 1982 - 1990. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Xie, Z. S. M. Rahman, S. Xie, J. Zhu, Y. Du, X. Qin, H. Zhou, X. J. Zhou, and C. Mohan Strain distribution pattern of immune nephritis--a follow-up study Int. Immunol., June 1, 2008; 20(6): 719 - 728. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Tarasenko, H. K. Kole, and S. Bolland A Lupus-Suppressor BALB/c Locus Restricts IgG2 Autoantibodies without Altering Intrinsic B Cell-Tolerance Mechanisms J. Immunol., March 15, 2008; 180(6): 3807 - 3814. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Wermeling, Y. Chen, T. Pikkarainen, A. Scheynius, O. Winqvist, S. Izui, J. V. Ravetch, K. Tryggvason, and M. C.I. Karlsson Class A scavenger receptors regulate tolerance against apoptotic cells, and autoantibodies against these receptors are predictive of systemic lupus J. Exp. Med., October 1, 2007; 204(10): 2259 - 2265. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. E. K. Haywood, L. Gabriel, S. J. Rose, N. J. Rogers, S. Izui, and B. J. Morley BXSB/long-lived Is a Recombinant Inbred Strain Containing Powerful Disease Suppressor Loci J. Immunol., August 15, 2007; 179(4): 2428 - 2434. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Rankin, J. J. Boyle, S. J. Rose, L. Gabriel, M. Lewis, V. Thiruudaian, N. J. Rogers, S. Izui, and B. J. Morley The Bxs6 Locus of BXSB Mice Is Sufficient for High-Level Expression of gp70 and the Production of gp70 Immune Complexes J. Immunol., April 1, 2007; 178(7): 4395 - 4401. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Carlucci, J. Cortes-Hernandez, L. Fossati-Jimack, A. E. Bygrave, M. J. Walport, T. J. Vyse, H. T. Cook, and M. Botto Genetic Dissection of Spontaneous Autoimmunity Driven by 129-Derived Chromosome 1 Loci When Expressed on C57BL/6 Mice J. Immunol., February 15, 2007; 178(4): 2352 - 2360. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D.S. C. Graham, A.K. Wong, N.J. McHugh, J.C. Whittaker, and T. J. Vyse Evidence for unique association signals in SLE at the CD28-CTLA4-ICOS locus in a family-based study Hum. Mol. Genet., November 1, 2006; 15(21): 3195 - 3205. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M-C Zhang, N Misu, H Furukawa, Y Watanabe, M Terada, H Komori, T Miyazaki, M Nose, and M Ono An epistatic effect of the female specific loci on the development of autoimmune vasculitis and antinuclear autoantibody in murine lupus Ann Rheum Dis, April 1, 2006; 65(4): 495 - 500. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Hamano, K. Tsukamoto, M. Abe, G. D. Sun, D. Zhang, H. Fujii, S. Matsuoka, M. Tanaka, A. Ishida-Okawara, H. Tachikawa, et al. Genetic Dissection of Vasculitis, Myeloperoxidase-Specific Antineutrophil Cytoplasmic Autoantibody Production, and Related Traits in Spontaneous Crescentic Glomerulonephritis-Forming/Kinjoh Mice J. Immunol., March 15, 2006; 176(6): 3662 - 3673. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B R Lauwerys and E K Wakeland Genetics of lupus nephritis Lupus, January 1, 2005; 14(1): 2 - 12. [Abstract] [PDF]
|
|
|
|
|
|
M. E. K. Haywood, N. J. Rogers, S. J. Rose, J. Boyle, A. McDermott, J. M. Rankin, V. Thiruudaian, M. R. Lewis, L. Fossati-Jimack, S. Izui, et al. Dissection of BXSB Lupus Phenotype Using Mice Congenic for Chromosome 1 Demonstrates That Separate Intervals Direct Different Aspects of Disease J. Immunol., October 1, 2004; 173(7): 4277 - 4285. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. H. Kono, M. S. Park, and A. N. Theofilopoulos Genetic Complementation in Female (BXSB x NZW)F2 Mice J. Immunol., December 15, 2003; 171(12): 6442 - 6447. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. E. Wither, G. Lajoie, S. Heinrichs, Y.-C. Cai, N. Chang, A. Ciofani, Y.-H. Cheung, and R. MacLeod Functional Dissection of Lupus Susceptibility Loci on the New Zealand Black Mouse Chromosome 1: Evidence for Independent Genetic Loci Affecting T and B Cell Activation J. Immunol., August 15, 2003; 171(4): 1697 - 1706. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. A. Kelly, K. Thompson, J. Kilpatrick, T. Lam, S. K. Nath, C. Gray-McGuire, J. Reid, B. Namjou, C. E. Aston, G. R. Bruner, et al. Evidence for a susceptibility gene (SLEH1) on chromosome 11q14 for systemic lupus erythematosus (SLE) families with hemolytic anemia PNAS, September 3, 2002; 99(18): 11766 - 11771. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Fu, M. E. K. Haywood, and B. J. Morley Representational difference analysis in a lupus-prone mouse strain results in the identification of an unstable region of the genome on chromosome 11 Nucleic Acids Res., March 15, 2002; 30(6): 1394 - 1400. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Xie, S. Chang, P. Yang, C. Jacob, A. Kaliyaperumal, S. K. Datta, and C. Mohan Genetic Contributions of Nonautoimmune SWR Mice Toward Lupus Nephritis J. Immunol., December 15, 2001; 167(12): 7141 - 7149. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. H. Kono, M. S. Park, A. Szydlik, K. M. Haraldsson, J. D. Kuan, D. L. Pearson, P. Hultman, and K. M. Pollard Resistance to Xenobiotic-Induced Autoimmunity Maps to Chromosome 1 J. Immunol., August 15, 2001; 167(4): 2396 - 2403. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. E. K. Haywood, T. J. Vyse, A. McDermott, E. M. Thompson, A. Ida, M. J. Walport, S. Izui, and B. J. Morley Autoantigen Glycoprotein 70 Expression Is Regulated by a Single Locus, Which Acts as a Checkpoint for Pathogenic Anti-Glycoprotein 70 Autoantibody Production and Hence for the Corresponding Development of Severe Nephritis, in Lupus-Prone BXSB Mice1 J. Immunol., August 1, 2001; 167(3): 1728 - 1733. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X. F. Li, D. S. Charnock-Jones, E. Zhang, S. Hiby, S. Malik, K. Day, D. Licence, J. M. Bowen, L. Gardner, A. King, et al. Angiogenic Growth Factor Messenger Ribonucleic Acids in Uterine Natural Killer Cells J. Clin. Endocrinol. Metab., April 1, 2001; 86(4): 1823 - 1834. [Abstract] [Full Text]
|
|
|
|
 |
|
 T. R. Merriman, H. J. Cordell, I. A. Eaves, P. A. Danoy, F. Coraddu, R. Barber, F. Cucca, S. Broadley, S. Sawcer, A. Compston, et al. Suggestive Evidence for Association of Human Chromosome 18q12-q21 and Its Orthologue on Rat and Mouse Chromosome 18 With Several Autoimmune Diseases Diabetes, January 1, 2001; 50(1): 184 - 194. [Abstract] [Full Text]
|
|
|
|
|
|
M. A. Jordan, P. A. Silveira, D. P. Shepherd, C. Chu, S. J. Kinder, J. Chen, L. J. Palmisano, L. D. Poulton, and A. G. Baxter Linkage Analysis of Systemic Lupus Erythematosus Induced in Diabetes-Prone Nonobese Diabetic Mice by Mycobacterium bovis J. Immunol., August 1, 2000; 165(3): 1673 - 1684. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Morel, B. P. Croker, K. R. Blenman, C. Mohan, G. Huang, G. Gilkeson, and E. K. Wakeland Genetic reconstitution of systemic lupus erythematosus immunopathology with polycongenic murine strains PNAS, June 6, 2000; 97(12): 6670 - 6675. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. L. O'Keefe, G. T. Williams, F. D. Batista, and M. S. Neuberger Deficiency in CD22, a B Cell-specific Inhibitory Receptor, Is Sufficient to Predispose to Development of High Affinity Autoantibodies J. Exp. Med., April 19, 1999; 189(8): 1307 - 1313. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Bolland, Y.-S. Yim, K. Tus, E. K. Wakeland, and J. V. Ravetch Genetic Modifiers of Systemic Lupus Erythematosus in Fc{gamma}RIIB-/- Mice J. Exp. Med., May 6, 2002; 195(9): 1167 - 1174. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
