SEARCH

SEARCH BY CITATION

Introduction

  1. Top of page
  2. Introduction
  3. Basic Biology of Sex
  4. Organ-Specific Responses
  5. Pregnancy
  6. Hormones
  7. Epidemiology
  8. Hypothesis
  9. Implications and Summary
  10. REFERENCES

The Institute of Medicine and other organizations have recently addressed sex inequality in all human illness, identifying gaps in knowledge and suggesting research agendas, but to date few authors have offered hypotheses that might explain sex discrepancy in systemic lupus erythematosus (SLE) (1, 2). In order to readdress these questions with specific reference to SLE, The Mary Kirkland Center for Lupus Research convened a meeting of experts in biology, genetics, endocrinology, immunology, epidemiology, and clinical medicine. Persons listed in the footnotes were the invited speakers (3). This article summarizes the presentations at the conference and suggests an hypothesis to answer the questions above.

Basic Biology of Sex

  1. Top of page
  2. Introduction
  3. Basic Biology of Sex
  4. Organ-Specific Responses
  5. Pregnancy
  6. Hormones
  7. Epidemiology
  8. Hypothesis
  9. Implications and Summary
  10. REFERENCES

Casually considered to comprise 2 discrete entities, biological sex is in fact a continuum (2). In water animals, in contrast to land animals, sex chromosomes are not dichotomous. Although a single gene determines chromosomal sex in vertebrates, environmental variables can override the gene. In birds and mammals, sex chromosomes generate a hormone cascade that dictates secondary sex characteristics. All tissues are sex-steroid responsive, with evidence for this assertion being sex differences in body size, color, and shapes. Immunocompetent cells, those most discussed in SLE, represent a minority of cell lines worthy of study.

Bizarre natural anomalies challenge basic definitions of sex. Gynandromorphs (animals divided left and right, one side having female sex chromosomes and the other male, with corresponding color, shape, and size) occur in insects, crustaceans, and birds. How cells on the male or the female side respond to blood-borne sex hormones and what hormone profiles these animals have are mostly unknown. In one well-studied animal, the song pattern of a gynandromorphic finch was dictated by neural rather than by hormonal sex (4).

On a scale of biologic impact, chromosomal sexes carry unequal weights. Germline base substitution mutations occur in sperm more than ova; deletions and chromosome replications occur more in ova, probably due to different replication rates of ova and sperm. Telomeric recombination errors are lethal in female germline cells, therefore such errors occur mostly in males. Imprinting of genes occurs mostly in regions near the centromere and many distal X chromosome genes escape inactivation. X inactivation may be skewed to the paternal or maternal X (5, 6). Imprinting affects diverse body systems, including intelligence (7, 8). A male X chromosome produces less gene product than does a female X chromosome; in females a maternal gene diminishes the activity of a paternal gene. These genetic differences do not speak directly to influencing female:male ratios of illness, but they illustrate genetic ways in which sexes may have different susceptibilities to disease (9).

Methylation of DNA CG pairs, a mechanism for suppressing gene expression, is a process that can be influenced by drugs and by diet (10). Patients with SLE have abnormally low total T cell DNA methylation (more activated genes). Because women have 2 X chromosomes, one of which has genes that are mostly inactivated, failure to inactivate affects women more than men. Demethylation of sites on an inactive X could contribute to female susceptibility to lupus.

An apparent high incidence of lupus in men with Klinefelter's (XXY) syndrome and the confirmed high incidence of antithyroid antibodies in patients with Turner's (XO) syndrome suggest that the relationship between the X chromosome and autoimmunity is important and complex (11). The Y chromosome may also be implicated. In BXSB mice, which experience male-predominant lupus, B cells containing a Y-linked autoimmune accelerator gene make antibodies to nucleolar antigens because of increased expression of toll-like receptor 7, an innate immune response receptor; the increased expression is due to a duplication of a Y chromosome gene (12). The ligand for the same receptor induces interferon production in females (13), likely relevant to the high interferon signature of active SLE (14).

Organ-Specific Responses

  1. Top of page
  2. Introduction
  3. Basic Biology of Sex
  4. Organ-Specific Responses
  5. Pregnancy
  6. Hormones
  7. Epidemiology
  8. Hypothesis
  9. Implications and Summary
  10. REFERENCES

Male and female embryonic cells in primary culture respond differently to experimental insults in an organ-specific manner (ref.15, and Zakeri Z: unpublished observations). Addition of estrogen or testosterone can modify responses in some cases. Microarrays demonstrate that gene activation of whole embryos shows differences between males and females that are quantitatively small but frequent. What determines these sex-specific different responses remains to be learned.

In healthy women, blood vessels, interrogated by vasodilation measures, behave differently from those of men, probably because of estrogen (16). Whether postmenopausal estrogen supplements do or do not prevent progression of atherosclerosis may be a function of timing: blood vessels affected by early atherosclerosis (or early postmenopause) may respond to estrogen, whereas those affected by later atherosclerosis (or late menopause) may not. Male blood vessels may not respond at all. The timing of in vivo exposure to estrogen suggests an important consideration: different tissues may respond differently at different times of a life or of a menstrual cycle. Angiogenesis, a critical process in both inflammation and repair, is sex discrepant, probably because the function of the intermediary P-selectin is sex restricted and estrogen dependent. This conclusion derives from knockout mouse models, subjected to experimental inflammation (Koch AE: unpublished observation). Thus in many senses the female vascular bed differs greatly from the male.

Pregnancy

  1. Top of page
  2. Introduction
  3. Basic Biology of Sex
  4. Organ-Specific Responses
  5. Pregnancy
  6. Hormones
  7. Epidemiology
  8. Hypothesis
  9. Implications and Summary
  10. REFERENCES

A blastocyst burrows into and parasitizes the maternal circulation within the first few weeks of conception. One of its first activities is to open widely the uterine spiral arteries to create a low-pressure vascular lake. Toxemia, which results if this process fails, predicts a young woman's future development of cardiovascular disease. Conversely, poor fetal nutrition (because of maternal toxemia or other causes) has long-lasting effects on the child that may result in disease when that child is fully grown. In both mother and child, pregnancy events may recur as illness after many decades (17).

Trophoblast normally circulates in a pregnant woman's blood; apoptotic bodies from trophoblastic syncytial knots circulate in the blood of toxemic women to the extent that up to 6% of plasma DNA may be fetal. Maternal blood contains fetal cells and fetal blood contains maternal cells (microchimerism) for as many as 40 years after a woman has given birth. It is possible that microchimerism can pass to a third generation. Microchimeric cells appear in lupus nephritis kidneys (18). HLA relationships in which a mother and child have excess sharing of HLA class II alleles occur with scleroderma in women and with SLE in men, suggesting both an abnormality in those few men who develop SLE and a role for maternal-to-fetal transmission in this disease (19). HLA class II mismatched fetal cells offer a possible explanation for pregnancy-induced amelioration of rheumatoid arthritis. Fetal cells from prior pregnancies are prominent in thyroid nodules and in the affected skin and spleen of patients with scleroderma; maternal cells are prominent in the intraventricular node of fetuses affected with congenital heart block (20).

Microchimeric cells may participate in tissue repair or may initiate injury. Circulating trophoblast, microchimeric cells, pregnancy itself, and toxemia represent a unique biology that males cannot share and of which the biologic implications are just beginning to be understood.

Hormones

  1. Top of page
  2. Introduction
  3. Basic Biology of Sex
  4. Organ-Specific Responses
  5. Pregnancy
  6. Hormones
  7. Epidemiology
  8. Hypothesis
  9. Implications and Summary
  10. REFERENCES

Gonadal hormones have well-known effects on isolated immunocytes and on whole animals; in general, estrogen up-regulates and testosterone down-regulates immune function (21). However, most of these hormone activities would explain sex differences in severity of disease rather than sex differences in incidence (22). In human illness, by contrast, exogenous estrogen has very little effect on the courses of patients with SLE (23, 24). Both animal models and human experience indicate that brain programming, adult sexual behavior, and personality are affected by the maternal hormone levels that the fetus sees. The details regarding what levels, which hormones, what time during the pregnancy, and which specific effects are unclear, but the ability of pregnancy variables to condition a fetus (perhaps to develop later disease) is not in question.

Estrogen levels are high enough and testosterone levels low enough in rheumatoid synovium and fluid to influence cell proliferation and function (25). According to data presented at the conference, estrogen receptors are highly expressed in Sjögren's syndrome salivary glands. Estradiol up-regulates 5% of genes in tissue culture cells (Cutolo M: unpublished observations). Castration/replacement studies in whole animals have critical windows, usually early in life, for their effects to occur. By ignoring these windows, microarray and castration/replacement experiments leave open the possibility that in vitro tests of hormone effect do not apply to intact humans. Nonetheless, physiologic hormone levels (including progesterone and prolactin) likely do have effects in processes relevant to SLE. The hypothesis that genomic or chromosomal female sex predisposes to the beginning of the illness and estradiol perpetuates it is plausible.

Cortisol response to inflammation and urinary 2-hydroxylated estrogen levels are low in both sexes in inflammatory disease, whereas 16-hydroxylated estrogen levels are normal in both sexes. Physiologic levels of estradiol rescue and activate anti-DNA–secreting B cells in mouse spleens (26). Estradiol allows maturation of high-affinity antibody-secreting cells, which then outcompete low-affinity antibody-secreting cells, resulting in persistence of autoreactive cells and autoimmunity. In a mouse model, the degree to which estradiol can modulate immune response is genetically determined, raising the possibility that some women may be genetically more susceptible to estrogen-influenced immune dysregulation than others (27).

Epidemiology

  1. Top of page
  2. Introduction
  3. Basic Biology of Sex
  4. Organ-Specific Responses
  5. Pregnancy
  6. Hormones
  7. Epidemiology
  8. Hypothesis
  9. Implications and Summary
  10. REFERENCES

The Georgia Lupus Registry is a population-based effort to capture the full spectrum of SLE, including more accurate definition of the female:male ratio across age groups, allowing for better hypothesis generation and testing (Lim SS: unpublished observations). An earlier study in North Carolina found that, compared with controls, patients with lupus are older at menarche, have fewer pregnancies, earlier menopause, less use of oral contraceptives, less use of postmenopausal estrogen, less exposure to environmental estrogen, and fewer breast cancers, suggesting that patients with SLE may be hormonally abnormal prior to the clinical onset of their illnesses (28). A search for a cause of lupus might therefore look to the pregnancies that produced the patients and to those patients' early childhoods.

Hypothesis

  1. Top of page
  2. Introduction
  3. Basic Biology of Sex
  4. Organ-Specific Responses
  5. Pregnancy
  6. Hormones
  7. Epidemiology
  8. Hypothesis
  9. Implications and Summary
  10. REFERENCES

Reports presented at the conference suggest the following hypothesis: a child is conditioned by X chromosomes or by an in utero or early childhood event to be susceptible to SLE. Unmasking of susceptibility may require exposure to one or many environmental insults, such as a virus. At female puberty (but not male puberty), high levels of estradiol may be permissive (or testosterone may be suppressive), allowing clinical disease to occur. The presence of a necessary but not sufficient checkpoint at female puberty may explain both the striking female predominance of SLE and its occurrence in young adulthood.

Implications and Summary

  1. Top of page
  2. Introduction
  3. Basic Biology of Sex
  4. Organ-Specific Responses
  5. Pregnancy
  6. Hormones
  7. Epidemiology
  8. Hypothesis
  9. Implications and Summary
  10. REFERENCES

Several conclusions follow from these discussions. First, hormone interventions for treatment of clinically apparent lupus are unlikely to help individual patients. Second, combining all autoimmune diseases in a single mechanistic construct—those that are highly female predominant (such as SLE and autoimmune thyroid disease) with those that are not (such as type I diabetes, multiple sclerosis, inflammatory bowel disease, ankylosing spondylitis, and Goodpasture's syndrome)—is probably wrong, because autoimmune diseases differ widely in sex dimorphism. Third, there may be critical, and differing, windows for intervention to induce, prevent, or treat SLE. New initiatives may direct attention to early identification of susceptible girls and to avoidance of premorbid triggers. Postmorbid interventions will continue to be directed against downstream components of inflammation. Focusing on the sex-dependent components of disease susceptibility, initiation, and progression is likely to be more productive than is considering lupus to result from a sex-independent single inflammatory or immune pathology.

REFERENCES

  1. Top of page
  2. Introduction
  3. Basic Biology of Sex
  4. Organ-Specific Responses
  5. Pregnancy
  6. Hormones
  7. Epidemiology
  8. Hypothesis
  9. Implications and Summary
  10. REFERENCES