This article is based on a presentation at the Workshop on Linking Environmental Agents and Autoimmune Diseases held 1-3 September 1998 in Research Triangle Park, North Carolina.
Address correspondence to D.A. Lawrence, Wadsworth Center, New York State Dept. of Health, Empire State Plaza, Albany, NY 12201-0509. Telephone: (518) 402-5684 Fax: (518) 474-1412. E-mail: david.lawrence@wadsworth.org
Animal use: protocols for the use of mice were approved by institutional review committees in accordance with National Institutes of Health guidelines.
Received 15 January 1999; accepted 20 April 1999.
Systemic lupus erythematosus (SLE) was first considered a multiphenotypic disease syndrome by Moritz Kaposi of Vienna in 1872 (
1). At the beginning of this century SLE was already believed to be, at least in part, an immune disorder on the basis of observations of false-positive syphilis serology in some patients. SLE is the prototypic autoimmune disease, characterized by a large variety of autoantibodies and a wide range of clinical manifestations and diverse organ involvement. Once thought of as a rare and invariably fatal disease, improved diagnosis shows that 1:1000 women may be at risk to develop certain SLE symptoms during their lifetime (
2). SLE is predominantly a disease of women of chilbearing age. It is estimated that in the United States alone 400,000 to 1,000,000 suffer from various forms of SLE. Clinical manifestations range from transient skin lesions or polyarthritis to renal failure (
3,4). This wide spectrum of clinical pictures, and thus phenotypes, presents difficulties in early diagnosis and treatment and harnessing modern genetic analysis to search for susceptibility loci and candidate genes. Genome mapping depends on high penetrance of an unequivocal phenotype, i.e., the full expression in most individuals with the same genotype. This is a problem with the varied clinical features of SLE; furthermore, SLE may not be a single disease entity.
It was recognized early that SLE is a great mimic of other diseases such as syphilis, rheumatoid arthritis, hemolytic anemia, and tuberculosis. Additionally, it was proposed that there are genetic as well as environmental components that contribute to this disease (1,5). Exposure to ultraviolet (UV) radiation, infection, diet, therapeutic drugs, physical and mental stress, and hormonal status (pregnancy) can have profound effects on the immune system and thus may contribute to the pathogenesis of SLE (3-5). Family and ethnic studies and investigations of identical twins strongly suggest a genetic basis for this syndrome as well as for the influence of the environment on genetic susceptibility. Because of the wide range of the SLE phenotype penetrance, it has been impossible to resolve whether the environmental influences on this disease initiate pathogenesis or enhance the penetrance of a low-grade, previously unrecognized phenotype.
The pathogenesis and etiology of SLE are only partially understood, but the hallmark is an inappropriate immune response to several autoantigens, particularly native DNA, chromatin, nucleoproteins, nucleosomes, histones, phospholipids, myeloperoxidase, thyroglobulin, and many other subcellular organelles (6). The autoimmune response in SLE does not appear to be an indiscriminate polyclonal activation of silent clones but several specific antigen-driven responses, each under separate genetic control. This implies that environmental effects may also act in a specific fashion. It is noteworthy that in SLE and diabetes mouse models, evidence is accumulating that each of these autoimmune humoral responses is influenced or directed by separate and diverse non-major histocompatability complex (MHC) loci (discussed later). These autoimmune responses per se may not be pathogenic. The formation of DNA-anti-DNA autoantibody complexes in the circulation or in situ, the generation of organ- or tissue-specific cytotoxic T cells, and the generation of inflammatory cascades and the release of cytokines combine to cause severe chronic or acute illness. These pathogenic immune reactivities are modulated by neurologic and endocrine functions. Thus, SLE is the result of complex abnormalities of the neuro-endocrine-immune networks.
Environmental insults on genetically susceptible backgrounds are rational candidates for the induction, recurrence, or exacerbation of SLE and other autoimmune diseases. Flare-ups of SLE are triggered in certain patients in remission as a result of infection or exposure to UV. However, other factors that modulate immune responses such as diet and stresses of a physical, mental, or hormonal, nature should also be considered (5). This is ample evidence that environmental factors play an important part in the pathogenesis of SLE (5); moreover, they may act as direct modulators of a specific immunologic function or affect the immune system in a more complex interaction within circuits of the neuroendocrine-immune network. Such factors could act as substitutes for epistatic genes that may potentiate, unmask, or suppress a potential susceptibility gene or immunoregulatory gene. Environmental factors may therefore act on certain susceptibility genes such as interleukins to modulate the autoimmune responses (see "Effects of Environment on Murine SLE").
We propose (using a new mouse model of SLE) that an environmental influence on SLE may not be observable in individuals with the full complement of SLE genes, but that individuals with a number of susceptibility loci below a certain threshold level will respond to such an insult. Thus, an effect may be more readily observed in a moderately lupus-prone strain rather than in strains with already extremely high disease penetrance. This is apparent in some patients with transient SLE symptoms when they have an adverse reaction to a drug such as procainamide (7). Thus, it may be a cumulative insult on the immune system that triggers SLE in individuals with a partial genotype. This may also account for the unusual discordance of SLE in some sets of identical twins in that the partial genotype results in environmentally dependent phenotypic expression, whereas the full genotype alone is sufficient for disease in both identical individuals (8).
Genetic analyses of the murine SLE models provide support for the emerging concept that SLE susceptibility involves a complex interaction between SLE disease-specific as well as pleiotropic autoimmune genes. Such complexities make the genetic analysis of heterozygous families and ethnic groups difficult until loci and candidate genes are mapped in an appropriate mouse model such as the female NZB 
NZWF1 (BWF1) (9). The genetic and phenotypic diversity of the NZM (New Zealand mixed) recombinant inbred mouse strains (10,11) and the congenic strains derived from them (12-16) resemble human patients more closely. They offer the advantage of providing numerous identical replicates with a defined susceptibility genotype and phenotype. Particularly informative may be the inbred strains with partial phenotypes and low penetrance.
This present article describes the origin of the SLE-prone NZM recombinant strains and their unique phenotypic and genotypic characteristics. The NZM recombinant strains greatly extend the usefulness of the original classical BWF1. We also review the effect of lead (Pb) on the immune system and the potential effects of this heavy metal on SLE induction and exacerbation in genetic variants of SLE-prone mice.
Spontaneous SLE models [reviewed in Theofilopoulos and Dixon (
17)] in F
1 hybrids and inbred strains (BWF
1, NZB
SWRF
1, MRL/
lpr/
lpr/Mp, BXSB/Mp) have offered an appropriate starting point to investigate the effects of environmental agents on genetic susceptibility and autoimmune disease progression. However, such studies are hampered by some inherent genetic and phenotypic problems:
a) these lupus models involve accelerating genes such as lymphoproliferation (
lpr) of MRL mice or the male autoimmune accelerator (
Yaa) of the BXSB strain; and
b) the SLE mice are already at the extreme high end of susceptibility and disease penetrance, making it difficult to observe further significant environmental influences with them. Conversely, normal inbred strains lacking SLE or other autoimmune loci altogether may not be advantageous for assessment of the effect of an adverse environmental agent on autoimmune disease induction or progression.
Sets of recombinant inbred strains as defined by Bailey (
18) are an indispensable tool of genetics and permit an analysis of polygenic traits and diseases that is not possible in heterogeneous backcross or F
2 progeny. However, breeding of recombinants from lupus progenitors by the standard method has been difficult because maintaining the large numbers of mice required to obtain at least one line/strain with the full phenotype is an overwhelming enterprise. Even when serendipity and an observant mind played a critical role, the discovery, selection, and propagation of the NZM recombinant inbred strains became a costly and labor-intensive undertaking that involved the propagation of more than 100,000 mice. More than 10,000 kidneys were examined, and hundreds of mice were phenotyped for disease course and anti-DNA and anti-red blood cell autoantibodies. Genotyping for polymorphic markers was done on several hundred mice.
Closely related SLE-prone, H2 identical inbred recombinant strains with a wide range of autoimmune phenotypic traits were not available until the derivation of the NZM strains (10). The NZM strains were bred from NZB/BlUmc and NZW/OuUmc progenitors in an unconventional manner, i.e., not directly from pairs of F2 progeny without selection. While crossing NZB females with NZW males to produce F1 hybrid females, an unintentional cross occurred in 1980 among F1 hybrids or an F1 hybrid female with its NZW father, the latter being more likely. Recombination and reassortment of the ancestral genomes resulted in a litter of mixed coat colors. Mice with a grey coat color and red eyes [an example of epistasis of the pink-eyed dilution (p) gene on the black (B) gene] were selected further for this trait by pen-breeding and color selection (1980-1981). Inbreeding was begun by A. Gabrielsen at the Wadsworth Center (New York State Department of Health, Albany, NY) in late 1981 (10) when it became apparent that some of the noninbred grey descendents of NZB/NZW ancestry succumbed to early renal failure. Thus, the NZM strains are the descendents of a pair of non-inbred littermates from a grey nephritic mother. The selection of lines was based on the presence and patterns of renal disease and other criteria deemed important by Gabrielsen (10), such as the disappearance of the BWF1 hybrid gender bias of nephritis in one line (NZM 2410) or lack of or late onset of renal disease in other lines (NZM 64 and NZM 2758). A large number of lines was started; some lines were split at about F10 to retain a duplicate or parallel back-up line. About 28 lines attained strain status (F20) and survived until 1995 (F35-F40). At that time it became necessary to focus further propagation on six strains with different penetrance (NZM 2410, 2328, 88, 391, 2758, and 64).
To date, diverse phenotypic traits have been observed in all or certain NZM strains as delineated herein. Differences in nephritis, neurologic deficit, Coombs positive anemia, splenomegaly, CD5
+ B cells, autoantibodies to single-stranded DNA (ssDNA), double-stranded DNA (dsDNA), chromatin, IgG, histone, thyroglobulin, and myeloperoxidase exist within the NZM strains. Furthermore, the penetrance, degree, and onset of some of these traits are important variables among the strains, as is gender.
Nephritis Phenotypes
The renal disease phenotype of the NZM strains varies by time course and penetrance (10). As shown in Figure 1, several strain-specific patterns have emerged: a) rapid onset of severe glomerulonephritis is present in males and females of strain NZM 2410; b) early onset in females, and delayed onset in males are seen in strain NZM 2328; c) some strains resemble the BWF1 or intermediate patterns with early onset in females and late onset in males (NZM 88 and 2754); and d) intermediate incidence or absence of nephritis were observed NZM 2758 (not shown) and 64.
Figure 1. Percent cumulative glomerulonephritis in NZM strains. The differences in the patterns of onset and penetrance of renal disease vary among the strains. Note that NZM 2410 female and male mice show a more rapid time course than BWF1 females. Other strains resemble BWF1 or partial phenotypes; NZM 64 is the most normal strain.
The highly penetrant expression of nephritis in NZM 2410 males and females is more rapid than in BWF1 females; male BWF1 mice (data not shown) become nephritic after 12-18 months of age. Nephritic NZM mice develop proteinuria, azotemia, and edema (Figure 2). Immune complex deposits in the skin (19) and hypertension phenotypes have not been determined (20). Glomerular lesions progress from infiltration by neutrophils to accumulations of mononuclear cells with concomitant mesangial matrix expansion and cellular proliferation. The mice succumb with florid crescentic glomerulonephritis or diffuse end-stage glomerulosclerosis (Figure 3A-C). Tubulo-interstitial abnormalities show diverse histologic features such as hyaline casts, tubular atrophy and dilitation, and interstitial mononuclear cell infiltrates. Arteritis or accumulations of mononuclear cells around small arteries is often seen. There may actually be several histopathologic renal disease phenotypes that have not been systematically investigated; however, Morel et al. (11) mapped the acute form of nephritis (renal failure before 44 weeks of age). The phenotype of delayed-onset type nephritis could not mapped. Some aging mice are found dead with relatively normal renal histology but massive diffuse glomerular hyaline thrombus formation; whether this is a subphenotype of murine SLE is not known. Immunofluorescent patterns of immune complex deposits in peripheral capillary loops and mesangial areas of glomeruli (IgG, IgM, and C3) are quite similar to those seen in BWF1 female mice.
|
Figure 2. Clinical features of systemic lupus erythematosus (SLE) in strain NZM 2410. Overt renal failure results in edema (left mouse), proteinuria, and elevated blood urea nitrogen.
|
Figure 3. Examples of renal histopathology in NZM mice. Kidney sections from a male mouse 179 days of age (A) and from a female mouse 221 days of age (B), both NZM 2410, have typical diffuse lupus glomerulonephritis showing sclerosis, "wire loops," hyaline thrombi, inflammatory infiltrates, and karyorhexis. Both mice were edematous. (C) Representative normal glomerulus of a male mouse of strain NZM 64, 180 days of age. Magnification 380 . Periodic acid-Schiff-Alcian Blue stain. Reproduced from Rudofsky et al. (10) with permission of the US-Canadian Academy of Pathology.
Neurologic Deficits
Postural neurologic deficits were described in BWF1 hybrids by Kier (21). The phenotype is present in some of the NZM strains (Figure 4A-D). This phenotype segregates independently from nephritis (Figure 5) and is also variably penetrant (Figure 6), suggesting that this trait may also have a genetic threshold mode of inheritance (discussed later). Brain capillaries contain immune-complex-like deposits of IgG and IgM; however, an immunopathologic mechanism for this postural abnormality has not been established. Electron-dense deposits in brain capillaries resembling immune complexes have been described in BWF1 (22).
Figure 4. Postural responses of NZM mice at 6 months of age. (A) Strain NZM 2328: the normal posture reflex with spreading of hind limbs and toes, raising of head. (B) Strain NZM 64: the partial phenotype with moderate hind limb adduction and curling inward of toes, lack of head orientation. (C,D) Strain NZM 88 with the severe postural deficit phenotype with clasping of fore and hind limbs.
Figure 5. Differential segregation of nephritis and neuropathy. Assessment of the nephritic and neurologic conditions of NZM strains. Note that each strain has a different combination of two segregating phenotypes. Scores of 0 to 4+ were assigned to the overall expression of abnormalities.
|
Figure 6. Percent cumulative postural response deficit in NZM strains. Abbreviations: F, female; M, male. Variations in the penetrance and time course are apparent. (A) NZB mice show high penetrance of pronounced impairment compared to NZW and NZM 2410. (B) NZM 391 has an intermediate phenotype, whereas NZM 2328 appears to be normal. (C) NZM 88 is comparable to NZB and the impairment is severe. NZM 64 has an intermediate phenotype with moderate clinical features (as shown in Figure 4B).
|
Other SLE Phenotypes
Anti-dsDNA autoantibodies, the hallmark of SLE, are present early in life in most NZM mice, a notable exception being strains 64 and 2758. These strains have low titers late in life, if ever (Table 1). Titers of anti-dsDNA do not consistently correlate with the nephritic phenotype (11). The most comprehensive data on serologic testing for autoantibodies of NZM 2410 have been recently reported by Morel et al. (16). There was a high incidence of several autoantibodies: IgM anti-ssDNA and IgG anti-dsDNA, IgM anti-IgG, and antibodies to chromatin, histone, nuclear proteins, nucleosomes, thyroglobulin, and myeloperoxidase. These findings are remarkably similar to findings in patients with SLE.
The main autoimmune phenotype of the NZB mouse strain is Coombs positivity (antierythrocyte autoantibodies). Most mice succumb to anemia. Fulminating crescentic glomerulonephritis with edema, as seen in BWF1 females, was a rare phenotype in the NZB/BlUmc strain. Aging male and female NZB and NZW mice (>12-18 months of age) develop varying degrees of focal or diffuse glomerulosclerosis. Direct Coombs tests show that this phenotype is expressed in some NZM strains with variable penetrance and at low titers (Table 1). Nevertheless, in strain NZM 2410 this trait appears to exceed the prevalence in BWF1 hybrids (30% vs 5%). Liver and spleen histology of affected mice show an increase in erythropoiesis.
Recently, Mohan et al. (23) reported the presence of expanded populations of B1a cells (CD5+ B lymphocytes) in strain NZM 2410. However, this NZB trait of CD5+ hyperdiploid B cells has also been identified in nonnephritic or low-penetrance strains NZM 64 and 2421 (24). The time of their appearance and their quantity in spleen and peritoneal cavity were not assessed. Splenomegaly that is attributable to lymphoproliferation was seen in NZM mice.
Notably, the role of the complement cascades has recently received little attention in the murine SLE equation. In the BWF1 model, C1 activity declines to virtually undetectable levels (25). C1q knockout mice develop SLE symptoms (26). Furthermore, C1q is near a strong susceptibility locus (Nba1) (discussed below). It is of interest that all NZM strains except the least nephritic NZM 64 strain are C5 deficient, a NZB trait on chromosome 2. Furthermore, all NZM strains are C4 (C4-Slp) deficient, a NZW trait on chromosome 17 (10). C5 deficiency was proposed as a possible susceptibility locus by Lanier et al. (27). Complement component deficiencies play an important role in human SLE (3,4). As already noted, some susceptibility loci in NZM 2410 are near complement component and receptor genes.
To deploy the NZM strains as an excellent model of the effect of environmental toxicants on susceptibility to SLE and autoimmunity, a brief review of their genetic characterization is necessary. The analysis of NZ mice and other SLE strains [reviewed in Vyse and Kotzin (
28)] uncovered a complex pattern of the genetic controls of SLE. As described earlier, the NZM strains were derived from the NZB and NZW strains that are the basis for the original classical murine model of SLE in the F
1 female hybrids (
9). The NZM strains are
H2 identical (
H2z of NZW), recombinant inbred strains containing portions of their ancestral genomes through reassortment and recombination (Tables 2, 3) (
10). It was proposed many years ago that both strains contribute one or two MHC-linked lupus genes and that heterozygosity at
H2 is an essential component (
29). Despite MHC homozygosity (
H2z), NZM 2410 has severe early nephritis (Figures 1, 2, and 3).
Conventional analysis of polymorphic markers (Table 3) did not uncover any SLE loci, but it provided important insights into the diverse genetic makeup and integrity of the strains. It is significant that the original selection of grey mice, i.e., for the NZW recessive pink-eyed dilution (p) on a recombinant NZW/NZB chromosome 7 (retention of pink-eyed-dilute, loss of albino), provided the serendipitous basis for selection that resulted in the discovery of two strong NZW SLE loci more than 15 years later: Sle3 is linked closely to p and Sle 5 is near the NZW centromere (Tables 3-5) (11,16). Extensive genetic studies were done to map SLE loci using NZM 2410 with highly penetrant renal disease in both sexes. Wakeland and colleagues (11-16,30) analyzed crosses of NZM 2410 (NZM) and C57BL/ 6 (B6) mice: (NZM B6)F1 hybrids, (NZM B6) NZM backcrosses, and (NZM B6)F2 hybrids. Furthermore, congenic B6 strains are being constructed that contain one or more of the mapped SLE loci or smaller intervals of a relevant mapped genomic segment to more closely identify candidate genes and the effects of these loci on the SLE composite phenotypes. These congenic strains may greatly enhance the usefulness of the NZM model for assessing environmental effects on autoimmunity. Conversely, these congenic and subcongenic strains will also provide an excellent tool to examine therapeutic interventions. For example, nephritis in NZM 2410 can be prevented with anti-CD40L monoclonal antibodies. This prophylactic treatment could be analyzed in detail in the congenic and subcongenic strains of NZM 2410 (31).
It is important to consider the genetic contributions of normal strains used in the analysis of such crosses. For example, NZM 2410 B6F1 crosses develop anti-DNA antibodies, albeit slowly (11). B6 also contributed genomic segments to the lupus-prone MRL and BXSB mice and may harbor some autoimmune genes. To analyze this phenomenon further, we made backcrosses of NZM 2410 to the ancestral NZB and NZW strains and F1 crosses with BALB/c ByJ and SM/J (32). The preliminary results provide presumptive evidence for genetic complementation or epistasis. Unlike B6, BALB/c mice did not provide a permissive genetic environment for NZM genes to induce detectable titers of anti-DNA antibodies or nephritis. However, similar to B6, one BALB/c mouse had histologic renal lesions after 477 days of observations; whether this was SLE nephritis could not be determined. Both NZB and NZW supported the expression of severe nephritis with 70-75% penetrance, but the time courses differed. The NZW NZM 2410 cross provided homozygosity at the three mapped Sle loci (Table 4), but this resulted in the delayed onset of severe nephritis (day 379 ± 142). The rapid onset of nephritis (day 218 ± 111) in the NZM 2410 NZB cross may have been due to H2 heterozygosity (Sle4) (11) or other accelerating, epistatic, or complementing genes present in NZB. The presumably normal SM/J strain provided genes to promote SLE in the F1 hybrids (40%). Unexpectedly, one SM/J breeder female mouse also died with severe lupuslike nephritis, suggesting that this strain may indeed contain some SLE susceptibility loci. These findings may complicate the use of the NZB SM recombinant inbred strains NXSM/Ei for SLE genetic analysis (28,33).
These and other complex breeding and mapping strategies have uncovered a substantial number of SLE loci [reviewed in Vyse and Kotzin (28)]. Some of these susceptibility loci (named Sle, Lbw, Sbw, Nba) are summarized in Table 4 along with their phenotype expressions. Several other NZ recently mapped non-MHC loci contribute to the control of autoantibody formation to DNA, nucleoproteins, myeloperoxidase, and thyroglobulin (16). Previously, murine SLE in BWF1 hybrids was thought to be controlled by one or two dominant loci derived from the MHC region of each parent (28). SLE-promoting effects of H2 heterozygosity (Sle4) can be seen in several strain combinations. However, a complex interaction of MHC loci and non-MHC loci on several chromosomes, the apparent involvement of complementation, and epistasis genes make the genetic analysis of SLE unexpectedly difficult even in an inbred mouse system. At present there are more than a dozen loci that map with varying degrees of statistical confidence to 15 of the 19 autosomal chromosomes [reviewed in Vyse and Kotzin (28)]. This list is expected to grow considerably.
Although NZM 2410 mice are being phenotyped and mapped in ever greater detail (11-16,30), the genetic makeup of the other NZM strains remains tentative. Table 5 summarizes a hypothetical view of the distribution of some of the mapped and proposed loci deduced from several nearby known polymorphisms such as coat color genes on chromosomes 2, 4, and 7 (agouti, brown, pink-eyed- dilute, and albino, respectively). Sle1 of NZW on chromosome 1 appears to be present in all NZM strains. The NZB loci may be absent. Chromosome 2 is recombinant only in NZM 64 (C5+/nonagouti); this provides a possible protective (C5-sufficient) locus (27). On chromosome 4, the close linkage of Sle2 with b (brown) of NZW shows that the two pink-eyed-dilute/brown (tan) strains (2410 and 391) contain this locus; however, 2410 also has most probably the additional NZB loci Nba1, Lbw2, and Sbw2. Chromosome 7 is recombinant and retained p (pink-eyed-dilute) and lost c (albino). This fortuitous mixing of the NZB/NZW genome on chromosome 7 suggests that all these mapped loci (Table 5) are present in all six strains, particularly Sle3 that is closely linked to p. The all-important MHC SLE loci are only partially, but uniformly, present in the NZM strains. All are homozygous for the NZW H2 complex (Ku, Iu, Sz, Dz), which implies that the direct or epistatic effects of Sle4/Lbw1 (heterozygosity at H2) are absent in all strains. The mechanism by which this heterozyotic element imparts its influence on renal disease is unknown. However, the possible predisposing loci, C4 (10) and tumor necrosis factor (TNF)-
(34) deficiencies, are present in all NZM strains. Thus, this hypothetical as well as established genetic diversity of the strains may provide further important information. Table 5 suggests that genetic differences on chromosomes 2 and 4 could be responsible for the substantial penetrance differences among the six NZM strains.
There is now ample evidence that murine lupus and many of the associated autoimmune phenomena are under polygenic control and that the penetrance, time course, and severity of disease are influenced by a large number of susceptibility loci. It is important to recognize that some of these SLE loci colocalize with autoimmune insulin-dependent diabetes (Idd) loci of nonobese diabetic (NOD) mice (28,35). These findings support the concept that disease-specific loci as well as pleiotropic autoimmune loci must be considered. These loci interact in various combinations to fit a threshold liability model (Figure 7) (11). The linkage studies lead to the conclusion that the accumulation of multiple component phenotypes results in the diverse forms of the SLE syndrome. These traits may have to exceed several thresholds for the disease to become overtly manifest. There may also be epistatic loci that suppress or alter the level of expression. It is significant that many of these SLE susceptibility loci map in or near regions that code for interleukins, interferons, tumor necrosis factor, complement components, and their receptors, i.e., candidate genes that are intimately involved with immunoregulation, inflammation, and the clearance or dissolution of immune complexes. Furthermore, there must be genetic controls of the synthesis of autoantigens in a pathogenic configuration or at an abnormal rate. Such loci have not been identified.
|
Figure 7. Penetrance of acute glomerulonephritis as a function of the number of Sle alleles. GN, glomerulo-nephritis. The frequency of nephritis (diamonds) demonstrates disease as a function of the number of susceptibility alleles. Various combinations of the recessive NZW Sle loci gave similar results. Note the potentiating effect of the fourth locus (Sle4) of heterozygosity at H2. Vertical intervals indicate 95% confidence intervals. Reproduced from Morel et al. (11) with permission of Cell Press.
|
We propose that some of the interactions of susceptibility loci or epistatic genetic effects may be affected or substituted by environmental factors.
The new NZM model represents a significant advance in studying environmental influences on SLE and autoimmunity in general. The main features of this model are the high, partial, and low expression of several autoimmune and disease phenotypes in MHC-identical strains that also have inherited mixed proportions of susceptibility (Table 5) and resistance genes from the NZB and NZW progenitors (hence, New Zealand mixed), making it possible to add a novel genetic dimension for assessing the effects of the environment on autoimmunity. In addition, the generation of congenic and subcongenic strains with the single or multiple SLE intervals of NZM 2410 may permit the dissection of environmental effects on specific autoimmune loci or genes in greater detail (
30).
Operationally, these loci can be categorized as controlling: a) loss of tolerance for specific self-antigens; b) amplification or dysregulation of humoral and cellular immune responses and functions; c) availability of antigens for immune complex formation; d) induction or amplification of target tissue injury; and e) interference with removal of immune complexes. The genetic control for the production and release of altered or immunogenic self-antigens has not been addressed. All these can be potentially manipulated directly or indirectly by the environment.
Use of Pb as a prototypic environmental effector in a genetic model of SLE is based on the well-documented specific impacts that Pb has on the immune system [reviewed in McCabe and Lawrence (36), Lawrence (37,38), and Lawrence and McCabe (39)] and is provided as an example of posited means by which an environmental agent could interface with genetics to affect autoimmune disease incidence. On the basis of previous observations in BWF1 mice (Tables 6, 7) and the curious fact that Pb intoxication of normal mice for more than 1 year does not cause renal lesion or autoimmunity or mortality (40,41), it is suggested that without particular and defined genetic backgrounds, Pb does not induce the SLE phenotype or autoimmunity in general.
The effects of Pb on humoral and cellular immunity of mice have been well characterized, and Pb can be considered an experimental paradigm for environmental influences on the immune system. It is therefore possible that alterations in immunoregulatory mech-anisms could initiate or exacerbate SLE in individuals with a sufficient genetic predisposition, i.e., mice with subthreshold numbers of SLE loci (see above). Unlike other heavy metals (Hg, Au, Cd), Pb does not induce toxic or autoimmune renal diseases in normal CBA/J mice exposed for 16-24 months to 2-10 mM PbCl in drinking water; there were no increases in immune complex deposition in the glomeruli (38). In one such exposure, parents were treated prior to conception, with offspring exposed throughout fetal, newborn, and adult life (37,39). Pb was found to be nonnephritogenic in these mice. As discussed in a recent review (36), BCF1 mice also display no signs of ill health and actually have a lower accumulated mortality than control mice. However, if Pb-exposed mice are given any infectious agent, their health, i.e., immunity, is compromised to a greater extent than that of control mice regardless of strain. In BALB/c or BCF1 mice exposed to Pb in vivo, cytokine levels are modified (42). Interleukin (IL)-4 levels are elevated and interferon (IFN)-
levels are lowered. If it can be demonstrated that an environmental agent can influence SLE in an NZM strain, we may be able to determine the degree of alteration of penetrance. These data would allow evaluation of whether an environmental agent can initiate autoimmune disease or exacerbate the disease process (heighten the pathology or cause it to occur earlier).
Unlike normal strains of mice, BWF1 lupus-prone mice showed an unusual response and gender difference during long-term exposure to Pb (Tables 6, 7). When mice were left untreated, mortality was as predicted: 50% at 34 weeks for females, 0% at 60 weeks for males; 2 mM Pb resulted in loss of male resistance: 50% mortality by 60 weeks. As BWF1 females are at the peak expression of the SLE phenotype, additionally induced changes were more difficult to observe. On the basis of these observations, it follows that it is possible to effect temporal, quantitative, and qualitative shifts in the expression of murine SLE by exposure to Pb.
Several immunologic mechanisms are affected by Pb administration that could accelerate or modify SLE. Pb is known to amplify humoral immune responses. It modifies or reduces the number of available macrophages, thus diminishing the efficient removal of immune complexes, which are known to accumulate in various tissues in SLE-prone mice. It is possible that Pb could alter the clearance of immune complexes by the glomerular mesangial cells (related to macrophages) of the kidney. It has also been reported that Pb increases T-cell responses by upregulating MHC class II expression and that there can be a general enhancement of both B- and T-cell functions (43,44). Furthermore, Pb modifies cytokine patterns throughout the body including the central nervous system, a significant partner of the immune and endocrine systems.
Pb modifies the cytokine networks (45) in a fashion similar to that observed in SLE-prone mice (46), that is, with elevated levels of the T helper (Th)2-promoting cytokines and depressed levels of Th1-promoting cytokines. In fact, IL-10 (cytokine synthesis inhibitory factor) is produced by macrophages, B cells, and Th2 cells, and is able to block IL-12 production, which promotes development of Th1 cells. It is of interest, therefore, that SLE susceptibility loci Sle1 and Nba2 on chromosome 1 are near IL-10 and a number of complement proteins and Fc receptor genes (28,35). In general, cytokines play a central role in the pathogenesis of human and murine SLE (47-49). These observations are supported by the fact that some of the susceptibility loci are closely linked to cytokine genes. Sle1 is near IL-10 (see above and Sle2 maps near IFN- and IFN-ß on chromosome 4 (11). Sle5 is virtually identical with the locus for IL-11 on chromosome 7 (16). Furthermore, a new locus identified by Morel et al. (16) on chromosome 11 is near a cluster of IL genes for IL-3, 4, 5, and 13. Some of these loci co-localize with susceptibility intervals of the NOD mouse (28,35) discussed above.
IL-6 positively correlates with active disease (47,48), and IL-6 has been reported to promote SLE in BWF1 mice (49). One environmental promoter of SLE is sunlight (UV exposure), which exacerbates SLE (the photosensitivity of SLE ) by enhancing IL-6 generation in SLE patients (50). The generation of abnormal amounts of interleukins may contribute to the potentiation of the pathogenic activity displayed by immune complex deposits in the epidermal-dermal junction of sun-exposed skin.
Another well-documented Pb-affected cytokine that enters the SLE equation is TNF- (34). Here again, murine lupus susceptibility loci in NZM strains (Table 5) provide genetic clues: the NZW TNF- defect maps to H2 (chromosome 17) and Nba1 maps near the NZB Tnfr2 (chromosome 4). Some NZM strains may contain these genes (Table 5). Recent studies by Fujimura and colleagues (51) in H2-congenic NZB and NZWF1 crosses demonstrate that the TNF- and MHC class II polymorphisms of NZW are distinct susceptibility loci that modulate NZB alleles directly or epistatically.
Because Pb and other environmental toxicants may have profound effects on the up- or downregulation of IL, IFN, and other chemokines, vasoactive substances, adhesion molecules, and too many others to elaborate on, the role of environmental agents must be assessed in a set of genetically defined variants of lupus-prone mice. As demonstrated herein, the NZM strains resemble the human spectrum of SLE by virtue of their varied phenotype and complex patterns of inheritance. Using genetically defined mice with identical immune response loci (MHC class I and class II genes) but with different degrees of SLE penetrance should allow more definitive analysis of the ability of an environmental agent to initiate or exacerbate an autoimmune disease. Further, it may be possible to delineate the loci most influenced by the environmental agent.
REFERENCES AND NOTES
1. Graininger W, Smolen JS, Zielinski CC. Description of systemic lupus erythematosus: a historical perspective. In: Systemic Lupus Erythematosus, Clinical and Experimental Aspects (Smolen JS, Zielinski CC, eds). New York:Springer-Verlag, 1987;2-20.
2. Albert DA. Health services research and systemic lupus erythematosus: a reciprocal relationship. Perspect Biol Med 41:327-340 (1998).
3. Boumpas DT, Austin HAI, Fessler BJ, Balow JE, Klippel JH, Lockshin MD. Systemic lupus erythematosus: emerging concepts. Part 1: Renal, neuropsychiatric, cardiovascular, pulmonary, and hematologic disease. Ann Intern Med 122:940-950 (1995).
4. Boumpas DT, Fessler BJ, Austin HAI, Balow JE, Klippel JH, Lockshin MD. Systemic lupus erythematosus: emerging concepts. Part 2: Dermatologic and joint disease, the anti-phospholipid antibody syndrome, pregnancy and hormonal therapy, morbidity and mortality, and pathogenesis. Ann Intern Med 123:42-53 (1995).
5. Lahita RG. The clinical presentation of systemic lupus erythematosus. In: Systemic Lupus Erythematosus (Lahita RG, ed). New York:Academic Press, 1999;325-336.
6. Tan EM. Antinuclear antibodies: diagnostic markers for autoimmune diseases and probes for cell biology. Adv Immunol 44:93-151 (1989).
7. Adams LE, Hess EV. Drug-related lupus. Incidence, mechanisms and clinical implications. Drug Safety 6:431-449 (1991).
8. Reichlin M, Harley JB, Lockshin MD. Serologic studies of monozygotic twins with systemic lupus erythematosus. Arthritis Rheum 35:457-464 (1992).
9. Helyer BJ, Howie JB. Renal disease associated with positive lupus erythematosus tests in crossbred strains of mice. Nature 191:197 (1963).
10. Rudofsky UH, Evans BD, Balaban SL, Mottironi VD, Gabrielsen AE. Differences in the expression of lupus nephritis in New Zealand Mixed (NZM) H-2z homozygous inbred strains of mice derived from New Zealand Black (NZB) and New Zealand White (NZW) mice. Origins and initial characterization. Lab Invest 68:419-426 (1993).
11. Morel L, Rudofsky UH, Longmate JA, Schiffenbauer J, Wakeland EK. Polygenic control of susceptibility to murine systemic lupus erythematosus. Immunity 1:219-229 (1994).
12. Morel L, Yu Y, Blenman KR, Caldwell RA, Wakeland EK. Production of congenic mouse strains carrying SLE-susceptibility genes derived from SLE-prone NZM/Aeg2410 strain. Mamm Genome 7:335-339 (1996).
13. Morel L, Mohan C, Coker BP, Tian X-H, Wakeland EK. Functional dissection of systemic lupus erythematosus using congenic mouse strains. J Immunol 158:6019-6028 (1997).
14. Mohan C, Morel L, Yang P, Wakeland EK. Genetic dissection of SLE pathogenesis: Sle2 on murine chromosome 4 leads to B-cell hyperactivity. J Immunol 159:454-465 (1997).
15. Mohan C, Alas E, Morel L, Yang P, Wakeland EK. Genetic dissection of SLE pathogenesis: Sle1 on chromosome 1 leads to a selective loss of tolerance to H2A/H2B/DNA subnucleosomes. J Clin Invest 101:1362-1372 (1998).
16. Morel L, Mohan C, Yu Y, Schiffenbauer J, Rudofsky UH, Tian N, Longmate JA, Wakeland EK. Multiplex inheritance of component phenotypes in a murine model of lupus. Mamm Genome, #10:171-186 (1999).
17. Theofilopoulos AN, Dixon FJ. Murine models of systemic lupus erythematosus. Adv Immunol 37:269-379 (1985).
18. Bailey DW. Recombinant inbred strains. Transplantation 11:325 (1971).
19. Somer VM, Rudofsky UH, Gabrielsen AE. Immune complexes in skin of NZB/NZW mice. Clin Exp Immunol 22:461-467 (1975).
20. Rudofsky UH, Dilwith RL, Roths JB, Lawrence DA, Kelly VE, Magro AM. Differences in the occurrence of hypertension in NZBxNZW F1, MRL/lpr, and BXSB mice with lupus nephritis. Am J Pathol 116:107-114 (1984).
21. Kier AB. Clinical neurology and brain histopathology in NZB/NZW F1 lupus mice. Comp Pathol 102:165-177 (1990).
22. Rudick RA, Eskin TA. Neuropathological features of a lupus-like disorder in autoimmune mice. Ann Neurol 14:325-322 (1983).
23. Mohan C, Morel L, Yang P, Wakeland EK. Accumulation of splenic B1a cells with potent antigen-presenting capability in NZM lupus-prone mice. Arthritis Rheum 41:1652-1662 (1998).
24. Raveche ES, Rodufsky UH. Unpublished.
25. Olsen CT, Gabrielsen AE. Early complement components in NZB/NZW mice. I: The first component. J Immunol 122:133-135 (1979).
26. Walport MJ, Davies KA, Botto M. C1q and systemic lupus erythematosus. Immunobiology 199:265-285 (1998).
27. Lanier BG, McDuffie FC, Holley KE. Role of C5 in the nephritis of NZB/W mice. J Immunol 106:740-446 (1971).
28. Vyse TJ, Kotzin BL. Genetic susceptibility to systemic lupus erythematosus. Annu Rev Immunol 16:261-292 (1998).
29. Knight JG, Adams DD. Three genes for lupus nephritis in NZB NZW mice. J Exp Med 147:1653-1660 (1978).
30. Wakeland EK, Morel L, Mohan C, Yui M. Genetic dissection of lupus nephritis in murine models of SLE. J Clin Immunol 17:272-281 (1997).
31. Mohan C, Shi Y, Laman JD, Datta SK. Interaction between CD40 and its ligand gp39 in the development of murine lupus nephritis. J Immunol 154:1470-1480 (1995).
32. Rudofsky UH, Plitnick LM, Gabrielsen AE. Unpublished.
33. Drake CG, Rozzo SJ, Hirschfeld HF, Smarnworawong NP, Palmer E, Kotzin BL. Analysis of New Zealand Black contribution to lupus-like renal disease. Multiple genes that operate in a threshold manner. J Immunol 154:2441-2447 (1995).
34. Jacob CO. Studies on the role of tumor necrosis factor in murine and human autoimmunity. J Autoimmunity 5:133-143 (1992).
35. Vyse TJ, Todd, JA. Genetic analysis of autoimmune disease. Cell 85:311-318 (1996).
36. McCabe MJ, Lawrence DA. Immune system: metals. In: Xenobiotic Induced Inflammation: Roles of Cytokines and Growth Factors (Schock LB, Laskin D, eds). New York:Academic Press, 1994;193-216.
37. Lawrence DA. Posited mechanisms of metal immunotoxicity. Hum Exp Toxicol 14:114-116 (1995).
38. Lawrence DA. Metal ions in autoimmune and infectious diseases. In: Handbook of Metal-Ligand Interactions in Biological Fluids, Vol 2 (Berthon G, ed). New York:Marcel Dekker, 1995;904-913.
39. Lawrence DA, McCabe MJ. Immune modulation by toxic metals. In: Organ-specific Metal Toxicology (Waalkes M, Goyer R, Klaasen C, eds). New York:Academic Press, 1995;305-337.
40. Lawrence DA, Mitchell D, Rudofsky U. Heavy metal modulation of lymphocyte and macrophage activity. In: Proceedings of the Thirteenth Conference on Environmental Toxicology, 16-18 November 1983, Dayton, Ohio. Air Force Aerospace Medical Research Laboratory TR-82-101. U.S. Air Force, 1983;63-79.
41. Lawrence DA, Mudzinski S, Rudofsky U, Warner G. Mechanisms of metal-induced immunotoxicity. In: Immunotoxicology (Berlin A, Dean J, Draper MH, Smith EMB, Spreafico F, eds). New York:Martinus Nijhoff, 1987;293-307.
42. Kishikawa H, Song RJ, Lawrence DA. Interleukin-12 promotes enhanced resistance to Listeria monocytogenes infection of lead-exposed mice. Toxicol Appl Pharmacol 147:180-189 (1997).
43. Warner GL, Lawrence DA. Stimulation of murine lymphocyte responses by cations. Cell Immunol 101:425-439 (1987).
44. Warner GL, Lawrence DA. Cell surface and cell cycle analysis of cation induced murine T cell proliferation. Eur J Immunol 16:1337-1342 (1987).
45. Heo Y, Parsons PJ, Lawrence DA. Lead differentially modifies cytokine production in vitro and in vivo. Toxicol Appl Pharmacol 138:149-157 (1996).
46. Hagiwara E, Gourley MF, Lee s, Klinman DK. Disease severity in patients with systemic lupus erythematosus correlates with an increased ration of interleukin-10: interferon gamma-secreting cells in the peripheral blood. Arthritis Rheum 39:379-385 (1996).
47. Stuart RA, Littlewood AJ, Maddison PJ, Hall ND. Elevated serum interleukin-6 levels associated with active disease in systemic connective tissue disorders. Clin Exp Rheumatol 13:17-22 (1995).
48. Linker-Israeli M, Denas RJ, Wallace DJ, Prehn J, Ozeri-Chen T, Klinenberg JR. Elevated levels of endogenous IL-6 in systemic lupus erythematosus. A putative role in pathogenesis. J Immunol 147:117-123 (1991).
49. Finck BK, Chan B, Wofsy D. Interleukin 6 promotes murine lupus in NZB/NZW F1 mice. J Clin Invest 94:585-591 (1994).
50. Pelton BK, Hylton W, Denman AM. Activation of IL-6 production by UV irradiation of blood mononuclear cells from patients with systemic lupus erythematosus. Clin Exp Immunol 89:251-254 (1992).
51. Fujimura T, Hirose S, Jiang Y, Kodera S, Ohmuro H, Zhang DQ, Hamano Y, Ishida H, Furukawa S, Shirai T. Dissection of the effects of tumor necrosis factor-alpha and class II gene polymorphisms within the MHC on murine systemic lupus erythematosus (SLE). Int Immunol 10:1467-1472 (1998).
Last Updated: September 21, 1999
