SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Objective

Cutaneous neonatal lupus resembles subacute cutaneous lupus erythematosus (SCLE), and photosensitivity is a common symptom. Tumor necrosis factor α (TNFα) release by ultraviolet light–exposed keratinocytes may be exaggerated in SCLE patients who have the haplotype TNFα −308A;DRB1*03. Accordingly, this study was undertaken to seek genetic and histologic evidence for a role of TNFα in the pathogenesis of cutaneous neonatal lupus.

Methods

DNA was isolated from 83 children (22 with rash, 35 with congenital heart block [CHB], 26 unaffected siblings) and 58 mothers from the Research Registry for Neonatal Lupus.

Results

The −308A allele (associated with higher TNFα production), HLA–DRQB1*02, and HLA–DRB1*03 were each present in the majority of children with rash (64%, 68%, and 64%, respectively). The frequency of all 3 6p alleles occurring together in 1 individual was greater in children with rash than in children who had either CHB or no manifestation of neonatal lupus (59% versus 30%; P = 0.02). This association with neonatal lupus rash was equivalent to published findings in a cohort of patients with SCLE, but significantly greater than the association in patients with discoid lupus erythematosus. Prominent TNFα staining in the epidermis was observed in lesional skin from 3 children with rash, but not in skin from a healthy neonate.

Conclusion

Taken together, the finding of a genetic predisposition to generate increased levels of TNFα following tissue injury and the histologic demonstration of TNFα in the target organ support the notion that this inflammatory cytokine plays a role in the pathogenesis of cutaneous neonatal lupus. Furthermore, the results of these studies provide evidence of a biologic link between neonatal lupus and the rash of SCLE.

Nearly a quarter of a century ago, Franco and colleagues reported that maternal anti-SSA/Ro antibodies were associated with a neonatal skin rash that resembled subacute cutaneous lupus erythematosus (SCLE) (1), an immunologic finding that would later prove important in relation to congenital heart block (CHB) as well. It is now established that rash and CHB, despite the former being transient and the latter permanent, are each major manifestations of the so-called neonatal lupus syndromes. The disparate timing and clinical implications of neonatal lupus rash and CHB suggest that both fetal and environmental factors contribute to the specific tissue injury evoked by the maternal autoantibodies. In cutaneous disease, a major clue to elucidation of the pathogenesis may be the photoinductive nature of the rash. The notion that photosensitivity has an influence is supported by the findings of 2 major studies, one of which showed that in 45 of 48 cases (94%), rashes appeared after or were exacerbated by exposure to ultraviolet (UV) light (2). In the other study, photosensitivity was a prominent feature, observed in 12 of 18 children (67%) (3).

UVB light (290–320 nm) is the exogenous trigger that is most often responsible for the skin lesions of cutaneous LE (4). There are several pathologic mechanisms that may be relevant to the induction and/or potentiation of these lesions in the setting of autoimmunity. UV light promotes tumor necrosis factor α (TNFα)–induced apoptosis in cultured keratinocytes. Although this has not been formally demonstrated in keratinocytes, apoptosis in cardiocytes redistributes nuclear antigens to the surface membrane, which can result in binding of nonpermeabilized cells by cognate extracellular antibodies, e.g., anti-SSA/Ro and SSB/La. Macrophages that phagocytose these opsonized apoptotic cells can release TNFα, favoring an inflammatory response (5). Of relevance, UVB can also directly trigger release of TNFα from keratinocytes and dermal fibroblasts (6).

There is variability in the UV-induced TNFα secretion by different transformed keratinocyte lines, suggesting that human polymorphisms may also be contributory. The gene encoding TNFα is highly polymorphic, and a substitution of G to A at position 308 in the promoter region (TNF2) has been associated with increased production of this cytokine. The common (wild-type) allele, 308G (TNF1), has a frequency of ∼80% in Caucasians and 92% in African Americans (7). Werth and colleagues recently demonstrated that treatment with UVB plus interleukin-1α caused a 300-fold increase in chloramphenicol acetyl transferase transcription over baseline with the 308A promoter, compared with <15% over baseline with the 308G promoter (8). However, it is acknowledged that specific cell type and stimuli may be influential in the differential expression of the 308 promoter polymorphisms. Of clinical relevance, an increased prevalence of 308A has been observed in SCLE, an extremely photosensitive form of cutaneous LE, and in adult dermatomyositis (DM) (8), another photosensitive disease, but not in discoid LE (DLE), a less photosensitive form of lupus.

The link between anti-SSA/Ro antibodies, photosensitivity, and TNFα promoter polymorphisms extends to HLA class II molecules. In Caucasians, there is strong linkage disequilibrium between the 308A allele and HLA–DRB1*03 (9). The presence of DRB1*03, as well as DQB1*02, is also common in individuals who are positive for anti-SSA/Ro and SSB/La antibodies (10). Accordingly, it is of note that in the recent studies by Werth and coworkers, the proportion of Caucasians with at least 1 308A allele who also carried the HLA–DRB1*03 allele was 100% among SCLE patients (who are often anti-SSA/Ro positive) but only 60% among DM patients (who are infrequently anti-SSA/Ro positive) and 56% in the control group (8).

Given these recent insights into molecular and genetic features of 2 photosensitive autoimmune rashes, this investigation was initiated to explore the role of TNFα in the pathogenesis of cutaneous neonatal lupus. Extensive studies were performed to address whether the frequency of the TNFα 308A allele in association with HLA–DQB1*02 and DRB1*03 is increased in children with neonatal rash compared with either children with CHB alone or children who are completely healthy but have been exposed in utero to maternal anti-SSA/Ro antibodies. In addition, immunohistologic studies were performed on available lesional skin from 3 infants born to mothers with anti-SSA/Ro antibodies, in order to establish the protein expression of TNFα.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Study subjects.

The family members included in the genetic study were enrolled in the Research Registry for Neonatal Lupus, which was established in September 1994 by the National Institute for Arthritis and Musculoskeletal and Skin Diseases and has been extensively described (11). For inclusion in the present study, a child was considered to have cutaneous manifestations of neonatal lupus if the following 2 criteria were met: 1) photograph demonstrating the characteristic annular lesions, and/or clear description of the rash in the medical records, and/or skin biopsy result characteristic of neonatal lupus (hydropic degeneration of the basal cell layer and a mononuclear cell infiltrate); and 2) presence of antibodies to 52-kd SSA/Ro, 60-kd SSA/Ro, 48-kd SSB/La, or U1 RNP in the maternal serum, as determined by enzyme-linked immunosorbent assay (ELISA; Diamedix, Miami, FL), ELISA with recombinant proteins, and sodium dodecyl sulfate immunoblot to identify the fine specificity of the autoantibody response. The fetus, neonate, or child was considered to have CHB if the following 2 criteria were met: 1) heart block (first-, second-, or third-degree) documented by electrocardiogram, echocardiogram, history of pacemaker, or statement in medical record; and 2) maternal autoantibodies, documented as in the criteria for cutaneous manifestations listed above. A child was considered to be a healthy sibling if 1) the child had no cardiac, cutaneous, hepatic, or hematologic manifestations of neonatal lupus but had a sibling with neonatal lupus; and 2) maternal autoantibodies were documented, as in the criteria for cutaneous manifestations listed above. A mother was included in the cohort if she had 1) at least 1 child with any manifestation of neonatal lupus; and 2) autoantibodies as described above.

Although members of the neonatal lupus families are of diverse ethnicity, only Caucasians were included in this study, given the rarity of the –308A allele in African Americans. Specifically, the subjects included 58 mothers, 22 children with rash, 35 children with CHB, and 26 unaffected siblings exposed to maternal anti-SSA/Ro antibodies.

DNA analysis.

DNA was isolated from anticoagulated blood using a kit from Qiagen (Valencia, CA) according to the manufacturer's instructions. Polymerase chain reaction (PCR)–restriction length fragment polymorphism was used to genotype the TNFα 308 region in each patient, as previously reported (12). Low-resolution microtyping (DQB1*02/DQB1*06 and DRB1*01–*14) was performed by PCR–sequence-specific oligonucleotide (SSO) reverse line-blot assay using the RELITM SSO HLA-DRB1 Test and RELITM SSO HLA-DRQ1 Test kits (no. 810.45 and 810.01; Dynal, Wirral, UK). Genotypes of HLA–DRB1 and HLA–DQB1 alleles were interpreted using the Dynal FELITM SSO Pattern Matching Computer Program. HLA–DRB1 and HLA–DQB1 alleles were determined by direct counting. Carrier frequency (%) was calculated as follows: % = n/N × 100, where % = carrier frequency, n = number of individuals positive for each genotype, and N = total number of individuals tested.

Skin biopsies.

Skin biopsy specimens were obtained from the lesional skin of 3 infants with neonatal lupus. In addition, a section of normal cutaneous tissue from a newborn male, obtained at the time of circumcision, was studied. Slides (formalin- or paraformaldehyde-fixed paraffin sections) were processed and stained as previously described (12), using rabbit anti-human TNFα (R&D Systems, Minneapolis, MN) or rabbit IgG, and then incubated with goat anti-rabbit alkaline phosphatase–conjugated anti-IgG.

Statistical analysis.

The frequencies of independent alleles were determined by direct counting. Statistical significance of the differences in frequencies of –308A, DQB1*02, and DRB1*03 alleles, and of –308A/DQB1*02/DRB1*03 associations, was estimated by Fisher's exact test with the aid of INSTAT software (GraphPad, San Diego, CA). P values less than 0.05 were considered significant.

RESULTS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Association of the HLA molecules DQB1*02 and DRB1*03 and the TNFα –308A polymorphism with neonatal lupus rash.

The presence of DQ2/DR3 has frequently been reported to be associated with autoantibody responses to both SSA/Ro and SSB/La autoantigens (10). In support of this observation, evaluation of the HLA class II molecules DQB1*02 and DRB1*03 revealed an identical frequency of 79% (46 of 58) for DQB1*02 and for DRB1*03 in the neonatal lupus mothers (Table 1). Although the frequency of DQB1*02 was higher in children with rash (15 of 22 [68%]) than in children without rash (CHB or unaffected) (30 of 61 [49%]), the difference did not reach statistical significance. A similar distribution for DRB1*03 was observed in these 2 groups of children (14 of 22 [64%] in those with rash; 27 of 61 [44%] in those without rash). The polymorphism at TNFα –308A, which is also at chromosomal location 6p, was evaluated. Consistent with our previous findings of a higher allelic frequency of the TNFα 308A allele in children and mothers from families with at least 1 neonatal lupus–affected individual compared with controls (12), the carrier frequency of –308A observed in mothers, children with rash, and children without rash in the present cohort was 78% (45 of 58), 64% (14 of 22), and 41% (25 of 61), respectively.

Table 1. Occurrence of the TNFα −308A polymorphism, HLA–DRB1*03, HLA–DQB1*02, and the combination of −308A with 1 or both HLA alleles in members of neonatal lupus families*
 SubjectsP
Children with neonatal lupus rash (n = 22), %Anti-Ro–exposed children without rash (n = 61), %Mothers (n = 58), %Children with neonatal lupus rash versus children without rashChildren with neonatal lupus rash versus mothersChildren without rash versus mothers
  • *

    TNFα = tumor necrosis factor α; NS = not significant.

  • Includes both children with complete heart block and unaffected children born to mothers with anti-Ro antibodies.

DQB1*02684979NSNS0.0022
DRB1*03644479NSNS0.0001
−308A644178NSNS<0.0001
−308A; DQB1*025933600.0393NS0.0018
−308A; DRB1*035930670.0204NS<0.0001
−308A; DQB1*02; DRB1*035930600.0204NS0.0009

In children with rash, the prevalence of the –308A allele paralleled the prevalence of HLA–DQB1*02 and DRB1*03. Specifically, 59% of the children with rash had –308A as well as either DQB1*02 or DRB1*03. The presence of –308A in association with either of these HLA molecules was significantly more frequent in the children with rash compared with children without rash (P = 0.0393 and P = 0.0204 for DQB1*02 and DRB1*03, respectively). In a subgroup analysis, 32% of children with CHB expressed both 308A and DRB1*03 (P = 0.057 compared with those with rash), as did 26% of unaffected children (P = 0.023 compared with children with rash). The frequency of all 3 6p alleles occurring together in 1 individual was significantly greater among children with rash than among those without rash (59% versus 30%; P = 0.0204). Children without rash, but not children with rash, differed significantly regarding the frequency of each of the separate alleles when compared with the neonatal lupus mothers (P = 0.0022, P = 0.0001, and P < 0.0001 for DQB1*02, DRB1*03, and –308A, respectively, and P = 0.0009 for the 3 alleles occurring in combination).

In Table 2, children with rash are compared with children without rash and with adults with other autoimmune cutaneous diseases for which data on the presence of –308A and DRB1*03 were available. The frequency of the combined presence of –308A and DRB1*03 in children with rash was equivalent to that in a published cohort of patients with SCLE (13), but was higher than that in patients with DM (P = 0.0565) or DLE (P = 0.0496).

Table 2. Occurrence of the combination of the TNFα −308A polymorphism and HLA–DRB1*03 in children with neonatal lupus rash versus anti-Ro–exposed children without rash and patients with SCLE, DLE, or DM*
Subjects (n)Frequency of −308A; DRB1*03, %P versus children with neonatal lupus rash
  • *

    SCLE = subacute cutaneous lupus erythematosus; DLE = discoid lupus erythematosus; DM = dermatomyositis (see Table 1 for other definitions).

  • Includes both children with complete heart block (CHB) and unaffected children born to mothers with anti-Ro antibodies (P = 0.057 and P = 0.023 for children with CHB and unaffected children, respectively, versus children with neonatal lupus rash).

  • Ref. 13.

Children with neonatal lupus rash (22)59
Anti-Ro–exposed children without rash (61)300.0204
SCLE (19)470.5379
DLE (17)240.0496
DM (39)310.0565

Evaluation of TNFα in neonatal lupus lesional skin.

In an infant with neonatal lupus diagnosed at 8 weeks of age, lesional skin from the left thigh exhibited intense staining by anti-TNFα in both the epidermis and the dermis. In the epidermis, TNFα appeared to be intracellular (Figure 1A). Isotype controls (Figure 1B) stained appropriately. Similar staining was observed in the lesional skin of 2 other infants with neonatal lupus (results not shown). TNFα was not detected in the foreskin of an otherwise healthy neonate (Figure 1C).

thumbnail image

Figure 1. Tumor necrosis factor α (TNFα) immunoreactivity in the lesional skin of an infant with neonatal lupus. Sections were stained with either anti-TNFα (A and C) or rabbit IgG (isotype control) (B and D) followed by alkaline phosphatase–conjugated antibody (goat anti-rabbit IgG) and were counterstained with hematoxylin. Lesional skin was obtained from the left thigh of an 8-week-old girl (A and B); normal skin was obtained from the foreskin of a 1-day-old boy (C and D). Individual cells within the epidermis of neonatal lupus skin, but not normal skin, stained intracellularly with anti-TNFα (original magnification × 20).

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Since rash occurs in only a minority of children exposed to maternal anti-SSA/Ro and SSB/La antibodies in utero, it is likely that a genetic predisposition is operative. Given the clinical resemblance of SCLE to the neonatal lupus rash, the TNFα –308A promoter polymorphism (associated with high cytokine production) may be one important determinant contributing to the pathogenesis of the cutaneous disease. The following observations support this hypothesis: 1) the majority of children with rash had the TNFα –308A promoter allele; 2) this allele was associated with the presence of both HLA–DQB1*02 and HLA–DRB1*03 in children with rash significantly more often than in anti-SSA/Ro–exposed children without rash; and 3) TNFα was demonstrated in lesional skin.

The proposed pathologic cascade to rash may be initiated by UVB light or another unknown exogenous stimulus that results in increased dermal fibroblast and keratinocyte secretion of TNFα. Cells from neonates with the –308A allele may have increased secretion of this inflammatory cytokine (perhaps even obviating the need for photoinduction), given the preferential binding of specific transcription factors to this promoter polymorphism. The secreted TNFα may augment keratinocyte (6) and dermal fibroblast apoptosis. In turn, apoptotic cells are a source of intracellular antigens, such as SSA/Ro and SSB/La, that may be bound by circulating maternal autoantibodies, thus rendering the cells “opsonized.” Keratinocytes express CD16 (Fcγ receptor III [FcγRIII]), and FcγR-mediated uptake of opsonized apoptotic cells represents an important pathologic pathway involving clearance of opsonized apoptotic cells, in addition to contributing to the synthesis/release of TNFα by FcγR-dependent signaling. For example, Miranda-Carus et al demonstrated that macrophages release TNFα during the clearance of opsonized apoptotic cardiocytes (5). Accordingly, the extensive staining for TNFα may be due to both UVB induction and secretion by inflammatory cells.

It is tempting to speculate that the association of –308A, DQB1*02, and DRB1*03 in children with neonatal lupus rash has functional consequences with regard to pathogenesis. The frequency of these 3 alleles alone and/or together was significantly higher than in a previously reported healthy control group of 93 Caucasians, in which the prevalence of DRB1*03 was only 17% and that of DQB1*02 was 21% (13), as well as in another group of 210 healthy controls (93% Caucasian, 7% African American), in which the prevalence of both DRB1*03 and –308A was 16% (13). It is readily acknowledged that use of a larger sample, association tests such as transmission disequilibrium testing, and ethnically and geographically matched controls for each family would strengthen the genetic findings presented herein. However, these limitations notwithstanding, the most compelling genetic association was that the children with rash had a significantly higher prevalence of the combination of –308A, DQB1*02, and DRB1*03 than did children without rash who were also exposed to maternal anti-SSA/Ro antibodies. It is also of note that the prevalence of the 308A;DQB1*02;DRB1*03 combination was greatest in the neonatal lupus mothers. The high prevalence of DRB1*03 confirms and extends previous findings (13) and is not surprising given the well-established association of anti-Ro/La antibodies with the presence of DQB1*02 and DRB1*03 (10, 14).

Taken together, these findings indicate that the same extended haplotype might contribute to a “double hit,” one in the mother and one in her offspring. The HLA portion of the extended haplotype DQB1*02;DRB1*03 provides the genetic predisposition for the generation of the candidate autoantibodies, which cross the placenta and bind apoptotic neonatal keratinocytes and fetal cardiocytes, thus initiating an inflammatory response. The TNFα portion of the extended haplotype in the children may further contribute to tissue injury by amplifying the inflammatory cascade, a notion supported by the limited immunohistologic results presented. The availability of nonlesional skin from neonates exposed to maternal anti-Ro/La, and demonstration that TNFα was not expressed therein, would further support the pathologic relevance of the TNFα staining in the lesional skin.

In terms of the morphology of the photoeruption and the histology of the skin lesions, the rash of neonatal lupus is remarkably similar to that of SCLE, another photosensitive autoimmune skin disease. In SCLE there is also basal cell damage in the epidermis and a superficial mononuclear cell infiltrate in the upper dermis (15). The notion of a similarity in pathogenesis between the rash of neonatal lupus and SCLE is supported by genetic evidence from the present study and that by Werth et al (8). While a high frequency of –308A has been noted in patients with DM, DRB1*03 was not increased, likely due to the absence of anti-Ro/La antibodies in that disease. Neither the –308A allele nor DRB1*03 was associated with DLE, a less photosensitive rash in which anti-Ro/La antibodies are infrequently observed. Furthermore, the characteristic lesions of DLE, such as follicular plugging, dermal atrophy, and scarring, are not typical of neonatal lupus or SCLE.

In summary, through the use of a translational approach to evaluate genetic factors in the context of histologic features, this study has provided insights into the pathogenesis of the cutaneous manifestation of neonatal lupus,. The finding that the TNFα –308A polymorphism in association with HLA–DQB1*02 and DRB1*03 influences disease expression in the setting of a common exposure to a risk factor, in this case maternal antibodies to Ro/La, is of biologic interest and complements the histologic findings. Although proof of concept must await application in an animal model, our results indicate that TNFα is an important neonatal susceptibility factor, and a component of its synthesis/secretion may be a target in future strategies for treatment.

REFERENCES

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES
  • 1
    Franco HL, Weston WL, Peebles C, Forstot SL, Phanuphak P. Autoantibodies directed against sicca syndrome antigens in the neonatal lupus syndrome. J Am Acad Dermatol 1981; 4: 6772.
  • 2
    Neiman AR, Lee LA, Weston WL, Buyon JP. Cutaneous manifestations of neonatal lupus without heart block: characteristics of mothers and children enrolled in a national registry. J Pediatr 2000; 137: 67480.
  • 3
    Weston WL, Morelli JG, Lee LA. The clinical spectrum of anti-Ro-positive cutaneous neonatal lupus erythematosus. J Am Acad Dermatol 1999; 40: 67581.
  • 4
    Werth VP, Zhang W, Dortzbach K, Sullivan K. Association of a promoter polymorphism of tumor necrosis factor-α with subacute cutaneous lupus erythematosus and distinct photoregulation of transcription. J Invest Dermatol 2000; 115: 72630.
  • 5
    Miranda-Carus ME, Askanase AD, Clancy RM, Di Donato F, Chou TM, Libera MR, et al. Anti-SSA/Ro and anti-SSB/La autoantibodies bind the surface of apoptotic fetal cardiocytes and promote secretion of TNF-α by macrophages. J Immunol 2000; 165: 534551.
  • 6
    Kock A, Schwarz T, Kirnbauer R, Urbanski A, Perry P, Ansel JC, et al. Human keratinocytes are a source for tumor necrosis factor α: evidence for synthesis and release upon stimulation with endotoxin or ultraviolet light. J Exp Med 1990; 172: 160914.
  • 7
    Sullivan KE, Wooten C, Schmeckpeper BJ, Goldman D, Petri MA. A promoter polymorphism of tumor necrosis factor α associated with systemic lupus erythematosus in African-Americans. Arthritis Rheum 1997; 40: 220711.
  • 8
    Werth VP, Callen JP, Ang G, Sullivan KE. Associations of tumor necrosis factor α and HLA polymorphisms with adult dermatomyositis: implications for a unique pathogenesis. J Invest Dermatol 2002; 119: 61720.
  • 9
    Wilson AG, de Vries N, Pociot F, di Giovine FS, van der Putte LB, Duff GW. An allelic polymorphism within the human tumor necrosis factor α promoter region is strongly associated with HLA A1, B8, and DR3 alleles. J Exp Med 1993; 177: 55760.
  • 10
    Gottenberg JE, Busson M, Loiseau P, Cohen-Solal J, Lepage V, Charron D, et al. In primary Sjögren's syndrome, HLA class II is associated exclusively with autoantibody production and spreading of the autoimmune response. Arthritis Rheum 2003; 48: 22405.
  • 11
    Buyon JP, Hiebert R, Copel J, Craft J, Friedman D, Katholi M, et al. Autoimmune-associated congenital heart block: demographics, mortality, morbidity and recurrence rates obtained from a national neonatal lupus registry. J Am Coll Cardiol 1998; 31: 165866.
  • 12
    Clancy RM, Backer CB, Yin X, Kapur RP, Molad Y, Buyon JP. Cytokine polymorphisms and histologic expression in autopsy studies: contribution of TNF-α and TGF-β1 to the pathogenesis of autoimmune-associated congenital heart block. J Immunol 2003; 171: 325361.
  • 13
    Siren MK, Julkunen H, Kaaja R, Kurki P, Koskimies S. Role of HLA in congenital heart block: susceptibility alleles in mothers. Lupus 1999; 8: 529.
  • 14
    Harley JB, Alexander EL, Bias WB, Fox OF, Provost TT, Reichlin M, et al. Anti–Ro (SS-A) and anti–La (SS-B) in patients with Sjögren's syndrome. Arthritis Rheum 1986; 29: 196206.
  • 15
    Bangert JL, Freeman RG, Sontheimer RD, Gilliam JN. Subacute cutaneous lupus erythematosus and discoid lupus erythematosus: comparative histopathologic findings. Arch Dermatol 1984; 120: 3327.