| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Immunology (A.H.), Erasmus University, 3000 DR Rotterdam, The Netherlands; Department of Obstetrics and Gynaecology (J.S.), Academic Hospital of the Vrije Universiteit, Amsterdam, The Netherlands; and Department of Immunology (H.A.D.), Erasmus University, 3000 DR Rotterdam, The Netherlands
 |
Abstract |
|---|
 Top Abstract I. IntroductionII. Definition and Clinical...III. Cells Involved in...IV. Tolerance to SelfV. Autoimmune Endocrine Disease:...VI. POF in Association...VII. Signs of Ovarian...VIII. Animal Models of...IX. SummaryReferences |
|---|
|
I. Introduction |
|---|
|
TopAbstractI. IntroductionII. Definition and Clinical...III. Cells Involved in...IV. Tolerance to SelfV. Autoimmune Endocrine Disease:...VI. POF in Association...VII. Signs of Ovarian...VIII. Animal Models of...IX. SummaryReferences |
|---|
Currently, the distinction between self and nonself is considered to involve a series of complicated and multistage interactions between various cells of the immune system. There is currently accumulating evidence that some cases of premature ovarian failure (POF)1 are due to a faulty recognition of self in the ovary by the immune system.
POF or premature menopause is a syndrome clinically defined by failure of the ovary before the age of 40 yr (1). POF is a heterogeneous disorder with a multicausal pathogenesis, and chromosomal (2, 3, 4, 5, 6, 7, 8, 9), genetic (4, 10, 11, 12), enzymatic (13, 14), iatrogenic (15, 16, 17, 18, 19, 20), or infectious (21, 22) aberrations may all form the basis for the disappearance of ovarian follicles. These aberrations may influence the ovary at any stage of life, including the prepubertal, pubertal, or reproductive stages (23).
This review will primarily focus on the accumulating evidence of an abnormal self-recognition leading to ovarian autoimmunity in a proportion of patients with POF. This places some cases of POF in the group of autoimmune diseases that affect hormone-producing glands, the so-called "autoimmune endocrinopathies," such as thyroiditis, insulin-dependent diabetes mellitus (IDDM), and Addison’s disease.
|
II. Definition and Clinical Presentation of Premature Ovarian Failure (POF) |
|---|
|
TopAbstractI. IntroductionII. Definition and Clinical...III. Cells Involved in...IV. Tolerance to SelfV. Autoimmune Endocrine Disease:...VI. POF in Association...VII. Signs of Ovarian...VIII. Animal Models of...IX. SummaryReferences |
|---|
POF was defined by de Moraes-Ruehsen and Jones in 1967 (1) as an unphysiological cessation of menses before the age of 40 yr and after puberty (hence, in fact, secondary amenorrhea). Women with POF have a hypergonadotropic-hypoestrogenic hormone profile. By 1939, the endocrinological profiles of the syndrome had been recognized on the basis of elevated levels of urinary gonadotropins (31). The clinical picture of POF was first described in detail in 1950 by Atria (32). This author reported 20 young women under the age of 35 yr with secondary amenorrhea, hot flushes, infertility, and an atrophic endometrium. In retrospect, these cases presumably were cases of POF, although at that time confirmatory gonadotropin assays were not routinely performed.
Patients with POF are mainly troubled by infertility due to the cessation of ovarian function. They have a typical menstrual history of normal age at menarche (33, 34) followed by regular periods. The disease thereafter presents either with oligomenorrhea or abrupt amenorrhea. Presently, amenorrhea due to POF is also seen after termination of oral contraception (35, 36, 37). A family history of POF is incidently obtained (10, 11, 12, 37). Fifty percent of patients with POF experience vasomotor symptoms, such as hot flushes and sweating boosts (37, 38, 39) due to the hypoestrogenic status. Other troubling symptoms are atrophy of the vagina and the urological tract, leading to vaginitis, dyspareunia, and cystitis.
The diagnosis of POF rests upon the clinical picture and the demonstration of elevated gonadotropin levels. The level of FSH is disproportionally higher than that of LH (40). Serum levels of FSH greater than 40 IU/liter are the hallmark of the diagnosis. Serum gonadotropin determinations should be repeated at least two or three times to be certain of the diagnosis because serum gonadotropin levels may wax and wane (41, 42, 43). POF presents itself not as an all-or-none phenomenon, and the precise timing of onset is often impossible to determine. The disease may have a fluctuating course, with high gonadotropin levels that later return to normal, and a later regain of ovulatory functions and even pregnancy (44, 45, 46). Nelson et al. (47) examined 65 POF patients by weekly estradiol sampling and sonography. In 50% of the cases, follicular activity could be demonstrated, and 16% of cases regained ovulatory function. Follicle biopsies were carried out in six patients, and these showed luteinized Graafian follicles (47). Alper et al. (44). reported that 7.5% (six of 80) patients were able to conceive after a diagnosis of POF.
The incidence of POF in a population under the age of 40 yr is estimated to be 0.9% (48). The choice of 40 yr as the age that separates premature from normal menopause is arbitrary. If one were to define abnormality as those values less or greater than 2 SD from the mean age of menopause (where ± 95% of the observations of a normally distributed variable are found), then the age of 43 would be the most appropriate lower age limit for the natural cessation of menses.
Kinch et al. (49) were the first to identify two histopathological types of POF: the afollicular and the follicular form. In the afollicular form, there is a total depletion of ovarian follicles and hence a permanent loss of ovarian function. Such total depletion of ovarian follicles is mainly due to gonadal dysgenesis, mixed gonadoblastoma, and hermaphroditism [reviewed by Coulam (23)]. Genetic and chromosomal abnormalities are one of the most well known causes of germ cell maldevelopment and disappearance (2). Such an accelerated loss of oocytes is considered to be the cause of POF in individuals with a 47,XXX and 45,XO and 45,X0 mosaicism (2, 9). Lack of migration of sex cells or faulty differentiation of the gonadal ridges lead to streak ovaries or, in some cases, to POF, depending on the actual number of primordial follicles that result.
In the follicular form, follicular structures are still preserved and hence a possibility of either spontaneous or induced return of ovarian function exists. The follicular form can be subdivided into: 1) oöphoritis (inflammation of follicles); 2) ovaries with a few follicles present; and 3) ovaries in which numerous primordial follicles are present [the so-called resistant ovary syndrome (ROS)] (50).
Although the histological classification suggests a sharp division between the follicular and afollicular forms, there is evidence that some cases of POF that were originally of the follicular type may progress to an afollicular stage. This is particularly the case in blepharophimosis (51, 52), galactosemia (53), and in the animal models of autoimmune oöphoritis (see below).
Recent research suggests that ovarian autoimmunity is a possible cause of both afollicular and follicular forms of POF. This review will list the arguments, pro and con, to such a view. At first, however, a short introduction into the cells involved in the immune response and immunological principles of self- and non-self-recognition will be given. This information will provide the background necessary to understanding these arguments.
|
III. Cells Involved in the Immune Response |
|---|
|
TopAbstractI. IntroductionII. Definition and Clinical...III. Cells Involved in...IV. Tolerance to SelfV. Autoimmune Endocrine Disease:...VI. POF in Association...VII. Signs of Ovarian...VIII. Animal Models of...IX. SummaryReferences |
|---|
(IFN-) and tumor necrosis
factor-
(TNF-) exposition (55). Macrophages, B cells, and
particularly dendritic cells (DC) are, however, the most efficient
(professional) antigen-presenting cells (APC) due to a variety of
factors, among which are the constitutive expression of MHC class II
molecules, the expression of costimulatory and adhesion molecules, and
other characteristics such as motility (56).
The genes encoding for both MHC class I and II molecules are members of
the immunoglobulin supergene family. These genes in the human species
are arranged on chromosome 6 [reviewed by Weetman (57)]. MHC class I
genes encode for the human leukocyte antigens (HLA) A, B, and C. MHC
class II genes encode the HLA DP, DQ, and DR antigens. The encoded HLA
molecules are dimers and comprise an - and ß-chain (Fig. 1
). The -chain of the MHC class I molecules is
encoded in the MHC genes; the ß-chain is termed ß2-microglobulin
and is encoded on a separate chromosome.
|
An additional molecule, known as the invariant chain, is intimately
involved in the biology of HLA class II molecules. The invariant chain
is a membrane glycoprotein, encoded by a non-HLA gene on chromosome 5.
The "invariant" designation stems from the observation that, in
contrast to the extensive polymorphism of some class II - and all
class II ß-chains, the invariant chain is nonpolymorphic. It forms a
trimer with the class II - and ß-chains in the endoplasmic
reticulum during biosynthesis of the MHC class II molecules and directs
the trafficking of the trimer through the posttranslational machinery
of the cell to the endosomal compartment (Fig. 2).
Current evidence indicates that the invariant chain also prevents
peptides from binding in the class II groove until the class II
molecule is delivered to the endosome. The invariant chain then
dissociates from the class II molecule, which can consequently bind
antigenic peptides processed from exogenous antigens taken up by the
APC and degraded in its lysosomal compartment (Fig. 2). The latter
fuses with endosomes (59, 60, 61). The complex of the class II molecule
with its bound peptide is then transported to the cell membrane (Fig. 2); however, the mechanism of transport is still unclear. Due to this
intracellular pathway, exogenous antigens are mainly presented in
association with MHC class II molecules (62). Peptides associated with
MHC class II molecules can only be presented to CD4+ T
cells (63), because the CD4 molecule is a special receptor for the MHC
class II molecules (Fig. 3).
|
|
). After recognition,
CD8+ T cells are able to kill the cell presenting the
endogenous antigen (54, 65). Neither class I nor class II MHC molecules
can distinguish between self and nonself (66, 67). It must also be
noted that the preferential association of exogenous antigens with MHC
class II molecules, and endogenous antigens with MHC class I, is not
absolute (62, 68).
The complex of MHC molecule-antigenic peptide-T cell receptor (TCR) is
insufficient for an adequate activation of the T cell. For full
activation, the interaction of other accessory molecules on APC with
their ligands on T cells is needed, such as the interaction of adhesion
molecules (69, 70, 71). The binding produced by these adhesion molecules
predominantly strengthens the interaction between the MHC-antigenic
peptide-TCR interaction, but also transduces signals that activate the
T cell. Important adhesion molecules are leukocyte function antigen 1
(LFA-1), which interacts with intercellular adhesion molecule 1
(ICAM-1), and leukocyte function antigen 3 (LFA-3), which interacts
with CD2 (70, 71) (Fig. 3). Inhibition in this process by monoclonal
antibodies to either one of these adhesion molecules inhibits the
activation and clonal expansion of T cells (69).
Apart from the adhesion molecule-ligand interaction, the interaction of
so-called "costimulatory molecules" on APCs and T cells is
essential for further T cell activation and T cell clonal expansion
(Fig. 3). If these costimulatory signals are not provided, the result
is T cell anergy [a state of specific nonresponsiveness of T cells
(72)]. Costimulating signals are predominantly provided by the binding
of the B7-1 (CD80) molecule on the APC (73, 74) to the CD28 molecule on
the T cell. Additional binding of T-lymphocytic CTLA-4 (cytolytic T
lymphocyte-associated antigen) to B7-2 (CD86) molecules on the APC also
takes place but it occurs probably later in the process, because CTLA-4
is primarily seen on the T cells after activation (75).
DCs are unique APCs in that they are the only APCs that are able to
effectively stimulate naive (CD45RA+) T cells (76, 77, 78).
Recent investigations led to the idea that the DC population is
heterogeneous with respect to ontogeny (79). It is certainly
heterogeneous with respect to morphology, the expression of adhesion
molecules (80, 81, 82, 83), and in cytokine production. With regard to
ontogeny, part of the DC populations is monocyte-derived and closely
associated with macrophages, whereas other DCs may have a separate
precursor [this subject was extensively reviewed by Kamperdijk
et al. (84)]. DCs are found in virtually all tissues and
organs of the body. The DC of the epidermis and dermis is known as the
Langerhans cell. Langerhans cells contain the peculiar Birbeck granules
that are not seen in DC in other organs, apart from the thymus.
Langerhans cells of the skin and DC of the gut wall are considered as
early (immature) stages in the differentiation of the cell, with a
superb capacity to pick up antigens and to degrade these to antigenic
peptides and place these peptides in the groove of the MHC molecules.
Skin Langerhans cells and gut DCs have been shown to migrate into the
afferent lymph as veiled cells to the skin- and gut-draining lymph
nodes (Fig. 4). These cells can be seen as
interdigitating cells in the T cell areas of these draining lymph
nodes, and these stages of the DC are considered as mature stages of
the cell with a superb capacity to stimulate T cells (85). It seems
likely that DC from other organs, like the heart, kidney, and endocrine
organs, may undergo similar migration and maturation.
|
, DC have been shown to produce
the mRNAs of these cytokines, without a noteworthy production of the
actual products (84, 86, 87). In general, DCs are regarded as poor
producers of cytokines, and their excellent APC function lies probably
in their migratory capacity, their capability to form clusters with T
cells via adhesion molecules (88, 89), and their high expression of
costimulatory molecules such as B7-1 and B7-2 (CD80/CD86) (90).
It is as yet unresolved whether the DC needs other cytokine-producing
accessory cells to guide the generated clonal expansion of naive T
cells in a certain direction of development. IL-12 is known to be
produced by macrophages, and this cytokine is able to push the
development of T cells into T cells that predominantly produce IFN-
(91); however, IL-10 (also produced by macrophages) down-regulates such
development (92, 93).
B. T cells
The ability of the immune system to specifically recognize all
varieties of possible antigens is based on the enormous diversity of
the antigen-specific receptors present on the T cells (TCR) and the
enormous diversity of the surface membrane-bound immunoglobulins
(smIg-receptors) on the B cell (94, 95, 96, 97). When the TCR fits with the
antigenic peptide in the groove of the MHC molecule, lymphocyte
activation and clonal expansion will be initiated (95, 96, 97) provided the
sufficient adhesion molecule and costimulating signals are given. A
specific TCR can bind only one form of an antigenic peptide, and this
will consequently lead to cell division of this type of TCR-specific
lymphocytes. This is referred to as clonal expansion.
The TCR is composed in the majority of cases of an - and ß-chain
(96, 97, 98) or in a minority of cases of a - and 
-chain (99, 100, 101, 102).
The various chains of the TCR are encoded by different gene segments:
variable (V), diversity (D), joining (J), and constant (C) gene
segments. The V, D, and J gene segments form a large repertoire. The
enormous diversity of the TCR is produced by the recombination of the
various V, D, and J genes from this large repertoire during T cell
maturation (103, 104). Thus, the capacity to react with all possible
antigenic peptides is genetically programmed and created by germ line
rearrangements and somatic mutations. The TCR is noncovalently linked
to a series of transmembrane proteins called the CD3 complex (Fig. 3)
(98, 102, 105, 106). Both CD4 and CD8 molecules on the T cells act as
coreceptors for the MHC class II and MHC class I molecules,
respectively, during the interaction of the TCR with the peptide-MHC
complex (107, 108). The entire CD3/TCR/MHC-II/CD4 complex or the entire
CD3/TCR/MHC-I/CD8 complex is involved in signal transduction (Fig. 3).
The identification and classification of various T cells (Table 1) is based on the expression of the CD3 complex, the
coexpression of either CD4 or CD8 molecules, and the composition of the
TCR (genetic makeup of various V, D, and J genes).
|
). One subset predominantly produces IFN-, but also
IL-2; this is the so-called Th1 subset. The other subset, Th2, produces
predominantly IL-4 and IL-5. The functional significance of these
different cytokine production profiles is that they represent different
T cell-regulatory actions. Th1 cells and their cytokine products
stimulate macrophages and hence cell-mediated immunity and
macrophage-mediated cellular destruction. Th2 cells and their cytokine
products stimulate B cells and hence lead to the humoral immune
response. It must be noted, however, that the Th1 and Th2 subtypes
represent extremes. There are many CD4+ T cells clones with
a cytokine production profile intermediate between Th1 and Th2 cells.
The driving of CD4+ T cells (Tho cells, Fig. 5) into either
the direction of Th1 or Th2 is guided in a complicated network by the
cytokines IL-1, IL-12, IL-10, IFN-, IL-4, and products of
arachidonic acid metabolism (91, 92, 93, 111, 112, 113, 114, 115).
|
). When the SmIg-receptor of a B cell
recognizes the antigen against which it is directed, and when
sufficient additional stimulatory signals are provided (see below),
proliferation will occur. The generated B cells will thereafter
differentiate into plasma cells that start to secrete immunoglobulins
with a specificity similar to that of the earlier membrane-bound form
of immunoglobulin.
When antigen-specific B cells and activated T cells recognize the same
antigen or a peptide thereof, a so-called "cognate interaction"
occurs (Fig. 6). The B cell uses its SmIg receptor for
uptake and concentration of the antigen (116). The antigen is processed
and the antigenic peptides are presented on the B cell surface in the
groove of the MHC class II molecules to the antigen-specific T cell.
When the cognate interaction activates the T cell to produce and
release the cytokines IL-4 and IL-5 (Th2 pathway), the B cell is
stimulated to clonally expand and differentiate into a plasma cell to
produce specific antibodies (117).
|
Antibodies are produced in different isotypes (IgA, IgG, IgM, IgE, IgD). The isotype is important in determining whether an antibody will fix complement (118). The majority of B cells in the peripheral blood express IgM and IgD on their cell surface, whereas a few express IgG or IgA. Secreted IgM antibodies are of low affinity and polyspecific. Secreted IgG and IgA antibodies are of high affinity and high specificity and are typical of secondary immune responses. The switch in immunoglobulin heavy chains from IgM and IgD to IgG or IgA is referred to as the isotype switch (119). This isotype switch is mediated by gene rearrangement in which the V region is coupled to another C region (119, 120). Autoantibodies of the IgG isotype are characteristic of certain pathological autoimmune reactions. To induce such antibodies, the help of autoantigen-specific Th2 cells is a prerequisite.
D. Effector cells in immune responses
Effector cells in immune responses are macrophages, natural killer
(NK) cells, and cytotoxic CD8+ T cells. Macrophages form a
heterogeneous population. Apart from immune regulation, certain subsets
of macrophages act as important scavenging cells. They are able to
endocytose and phagocytose microorganisms and cellular debris (121), as
well as exerting cytotoxicity against microorganisms or tumor cells
(122). The main function of macrophages is considered to be
phagocytosis. Phagocytosis of microorganisms and cellular debris is
greatly enhanced by opsonization of the material by specific
antibodies, as well as the capability of the Th1 cytokine IFN- to
strongly activate the cytotoxic properties of macrophages, increasing
their efficiency in killing the microorganism (123). Macrophages also
play a role in wound healing (124, 125) and in the regulation of hemo-
and lymphopoiesis (126, 127). Histologically classic macrophages are
large cells that show lamellapodia and vacuoles and possess an
irregular, indented nucleus (128). They stain for nonspecific esterase
and acid phosphatase throughout the cytoplasm.
Though precursors of macrophages certainly reside in the monocyte pool
the origin of all macrophages is not completely elucidated. A separate
precursor in the bone marrow may exist for some subpopulations of
macrophages (128). The monoblast and promonocyte remain in the bone
marrow very briefly before entering the blood stream as monocytes
(129). These latter cells migrate into the tissues where part of the
cells mature and differentiate into various lines of macrophages such
as the Kupffer cells, the osteoclasts, and the histiocytes. There is
not one monoclonal antibody that recognizes all the lines and
maturation stages of macrophages and that does not show a
cross-reactivity with other hematopoietic cells. The lack of a common
marker for all the subpopulations of macrophages is inherent to their
functional heterogeneity. In human studies, macrophages are normally
identified by specific CD markers (Table 1).
In exerting their various functions, macrophages are able to produce a
variety of signaling molecules, such as the cytokines IL-1, IL-6,
granulocyte-macrophage-colony-stimulating factor (GM-CSF), and TNF-
(128). Metabolites of arachidonic acid metabolism, nitric oxide and
oxygen radicals, are also important products for the regulation of the
immune response and the degradation of ingested material.
NK cells and CD8+ cytotoxic T cells are other important
immune cells in the effector arm of the immune system. NK cells do not
express conventional antigen receptors, such as the TCR or
SmIg-receptors, and the genes for these receptors remain unrearranged
(130, 131). They do express the receptor for the Fc part of the IgG
molecule, the FcRIII (CD16) (132, 133). Other important molecules
expressed by NK cells include CD56, a neural adhesion antigen, and the
ß-chain of the IL-2 receptor. This allows resting NK cells to respond
directly to IL-2 (134). The main function of NK cells is to provide
nonspecific cytotoxic activity toward virally infected cells and tumor
cells (135, 136). They do so by releasing perforin (pore forming) and
serine proteases (137). NK cells, like macrophages, can also kill
specifically if provided with an antibody. The process, known as
antibody-dependent cellular cytotoxicity, occurs via binding of the
antibody to the Fc receptor (CD16). The ontogeny of NK cells is only
partially understood. Although NK cells express a number of membrane
antigens in common with T cells and share functional properties with
some T cell subsets, suggesting a common origin, NK cells are found in
the fetus before the development of T cells or of the thymus. In
addition, NK cells appear to develop normally in nude, athymic mice
(135). Recent studies have indicated that NK cells can arise from
triple negative (CD3-/CD4-/CD8-)
thymocyte precursors that are CD56+ but do not express CD34
or CD5 (138). It must also be noted that NK cells are not only
considered as effector cells in the immune response, but also as
regulator cells. They are sensitive to activation by Il-12, produce
-IFN that activates the TH1 response, and are polyclonal activators
of B cells (139).
Cytotoxic T lymphocytes consist of mature T cells that are usually, but not always, CD8+. They exert cell contact-dependent cytotoxic functions through a perforin-dependent pathway (140). The cells also release the cytotoxic cytokine TNF. The perforin-dependent pathway is largely responsible for the T cell-dependent cytotoxic clearance of virus-infected cells and for rejection of tissue grafts and tumors (141).
|
IV. Tolerance to Self |
|---|
|
TopAbstractI. IntroductionII. Definition and Clinical...III. Cells Involved in...IV. Tolerance to SelfV. Autoimmune Endocrine Disease:...VI. POF in Association...VII. Signs of Ovarian...VIII. Animal Models of...IX. SummaryReferences |
|---|
A. Clonal deletion
T cells mature in the thymus from prothymocytes to mature T cells.
Because TCR rearrangement is random, self-reactive T cells are
generated in this process. However, the vast majority of self-reactive
T cells are deleted during further maturation in the thymus, the
so-called "clonal deletion" (143, 144). DCs occurring in great
numbers at the corticomedullary junction of the thymus express
self-antigens and are responsible for this deletion. Clonal deletion
depends upon recognition of the self-antigenic peptides by not fully
matured T cells, which, upon the antigenic recognition signal, do not
proliferate but go into apoptosis (145). A similar mechanism of
deletion may exist for self-reactive B cells in the bone marrow (146).
This mechanism is, however, not so well studied as clonal T cell
deletion in the thymus.
Clonal deletion for T and B cells is, however, incomplete, and T and B cells with a specificity for autoantigenic peptides and autoantigens can easily be found in the circulation (147). This is partly explained by the fact that not all self-antigens are expressed in the thymus and bone marrow. Some self-antigens, such as ocular lens antigens, are sequestrated from the immune system. Other self-antigens, such as sperm-antigens, are only expressed during late fetal life or only in adult life. Some autoantigens probably never reach the thymus or bone marrow and are never expressed there. This particularly applies to cryptic epitopes. Cryptic epitopes are de novo expressed epitopes on self-antigens that are caused by changes in the antigen after, for instance, an inflammatory process.
B. Clonal anergy
When autoreactive B and T cells have escaped clonal deletion, a
second control mechanism, namely the process of clonal anergy, should
come into operation. This process takes place predominantly in the
periphery (148, 149, 150). The induction of an immunologically anergic state
of the T cell is supposed to be due to a lack of provision of
sufficient second signals by APCs. Hence, when antigen is presented to
T cells by nonprofessional APC, such as MHC class II-positive
epithelial cells, clonal anergy will occur. Late in organ-specific
autoimmune diseases (e.g. in autoimmune thyroiditis and
IDDM) there is an aberrant expression of MHC class II molecules on the
epithelial cells of the endocrine tissues (151). Initially, this
aberrant expression of MHC class II molecules was interpreted as an
impetus for the increased self-reactivity (152). However, this aberrant
MHC class II expression late in organ-specific autoimmune disease can
also be considered as a sign of induction of clonal anergy (151, 153).
C. Active immunosuppression
When clonal deletion and clonal anergy have failed, yet another
down-regulating mechanism should come into operation, namely active
immunosuppression exerted by so-called "suppressor" immune cells.
These suppressor immune cells do not only include CD4+ and
CD8+ T cells (154, 155), but also suppressor macrophages.
The cells involved in immune suppression may be antigen-specific or
non-antigen-specific. They may also operate in an idiotype-antiidiotype
network (155). How active immune suppression is regulated remains
unclear. Taken together, earlier and recent evidence suggests that in
each individual a balance exists between autoreactive effector and
suppressor immune cells. In the healthy state this balance tips over in
favor of the suppressor forces, whereas in the autoimmune diseased
state the balance is in favor of the self-reactive effector forces (see
also later animal models of autoimmune oöphoritis).
D. Balance between Th1 and Th2 pathways
A recently developed theory approaches the problem of the control
of self-reactivity from yet another angle. Endocrine autoimmune
diseases with an ultimate failure of the target gland, such as IDDM,
are predominantly caused by Th1-mediated pathways in which the
endocrine cells are destroyed by -IFN-activated scavenger
macrophages. The recently developed theory emphasizes the reciprocal
relation between the Th1 and Th2 pathways (109, 156) and suggests that
if the Th1 pathway is diverted into the Th2 pathway that the
Th1-mediated autoimmune reactivity is dampened. In essence, tolerance
to self is not restored, but the harmful autoimmune reaction is
diverted to a less harmful one. There are indeed reports on cytokine
treatments that are able to induce such a switch from Th1 to Th2
pathways, resulting in an amelioration of the endocrine autoimmune
disease. Circulating antibodies, whose production is switched on by the
stimulation of the Th2 cells, apparently contribute little to the
damage of the target cells. It is known that endocrine autoantibodies
may exist for years in the circulation before endocrine autoimmune
disease develops (157, 158).
|
V. Autoimmune Endocrine Disease: Developmental Stages and Genetic Predisposition |
|---|
|
TopAbstractI. IntroductionII. Definition and Clinical...III. Cells Involved in...IV. Tolerance to SelfV. Autoimmune Endocrine Disease:...VI. POF in Association...VII. Signs of Ovarian...VIII. Animal Models of...IX. SummaryReferences |
|---|
The animal models for autoimmune disease of the thyroid are the Obese Strain of chicken (OS chicken) (159), the Bio Breeding (BB) rat (160), and certain strains of the Non Obese Diabetic (NOD) mouse. The BB rat and NOD mouse also suffer from an autoimmune insulitis leading to IDDM. Animal models for spontaneous autoimmune adrenalitis and oöphoritis are lacking. Only manipulations of normal mice (immunization with crude adrenal and ovarian extracts, and thymectomy plus cyclosporin A treatment) will lead to autoimmune adrenalitis and/or oöphoritis (161). A word of caution is necessary when trying to extrapolate data obtained in the animal models to the human situation: the animal models clearly show exaggerated and extreme forms of thyroiditis and insulitis, which already differ between the models themselves (let alone from patients), indicating a heterogeneity of the disease process. Hence, general conclusions drawn on the basis of studies in one animal model should always be verified in other animal models and certainly in human patients.
The animal models of insulitis and thyroiditis indicate that the
pathogenesis of the autoimmune failure of an endocrine gland is a
multistep process, requiring several genetic and environmental
abnormalities to come together before full-blown autoimmune thyroiditis
and/or insulitis develops. The following phases in the disease process
can be discerned (Fig. 7): 1) An initial phase of early
accumulation of APC and accessory cells (DCs, subclasses of
macrophages) in the endocrine tissue; 2) A later phase of an apparently
uncontrolled production of autoreactive CD4+ and
CD8+ T cells and of autoantibodies of the IgG class in the
draining lymph nodes; 3) A last phase where the target endocrine tissue
becomes susceptible for the autoimmune attack by the generated
autoreactive T cells and autoantibodies; this finally results in the
destruction of the glandular tissue.
|
The initial phase of glandular accumulation of macrophages and APCs is followed by a phase of an apparently uncontrolled clonal expansion of autoreactive T cells and B cells and the production of autoantibodies in the draining lymph nodes. In both the BB rat and the NOD mouse, there are strong indications for a genetically linked systemic immunodysregulation leading to the local exaggerated production of T cells, B cells, and IgG antibodies to various self-antigens. This systemic immune abnormality is partly associated with the presence of particular MHC class I and class II haplotypes (see below) and apparently leads to abnormalities in the stimulation and differentiation of cells involved in tolerance induction, such as the APCs, macrophages, and/or T cells (169). Indeed, APCs of NOD mice (169) and BB rats (our unpublished observations) have defects in their capability to generate T suppressor cells. With regard to such abnormal maturation of immunoregulatory T cells, the BB rat is special in that it lacks a regulator population of T cells (the RT6 cells). BB rats also show a rapid thymic involution (170). The OS strain of chickens has inborn defects in its suppressor cell system (159). Whether there are similar inborn defects in immunoregulatory cells in the human that lead to an endocrine autoimmune disease needs to be established. There are, however, numerous reports on both numerical and functional deficits in the suppressor cell system of patients with thyroiditis and IDDM (171).
Deficits in immunoregulatory cells do not only exist on an inheritable, genetic basis. They can also be acquired by (fetal) viral infections. In chickens, Avian Leucosis Virus has proven to exert a detrimental effect on thymus and bursa development, which disturbs delicate immune regulatory systems, leading to thyroid autoimmunity (172). Whether similar viruses or retroviruses with an affinity for immune cells are operative in human endocrine autoimmune diseases has been speculated upon but has not yet been proven. Experiments to detect virions and/or retroviral antigens have not been conclusive in showing the involvement of infectious viruses in human endocrine autoimmune disease (172).
After the stage of the excessive generation of autoreactive T cells and IgG autoantibodies, yet another factor or factors, at least in the OS chicken, determine whether or not a full blown autoimmune disease will develop (173). A prerequisite for clinical thyroid failure in this bird is a susceptibility of the target, the thyrocyte, for an autoimmune attack by the generated autoreactive T cells and IgG autoantibodies. Experiments have shown that this susceptibility factor is genetically determined, and it has been speculated that this factor might be an abnormal susceptibility of the thyrocytes for the cytokines produced by the autoreactive immune cells after infiltration. Whether such susceptibility factors are also important in the other animal models and in human disease needs further investigation, although in the BB rat a high susceptibility of pancreatic islet cells for IL-1 has been established.
Population, family, and twin studies have clearly shown that genetic factors exert a significant influence on the predisposition for an autoimmune endocrine disease. It is also clear that environmental factors (diet, infections, etc) contribute to disease expression because concordance rates in monozygotic twins and inbred animals are often imperfect. Since endocrine autoimmunity can be transferred by lymphocytes and bone marrow precursor cells into recipients, genes associated with the immune system have received prime attention. At present, convincing evidence does not exist for a relationship between the predisposition to an endocrine autoimmune disease and particular TCR haplotypes or polymorphisms, immunoglobulin allotypes and idiotypes, or cytokine genes (174). However, there is a clearly established genetic association with genes encoding for the MHC. In the BB rat this is the Rt1 haplotype (167), and in the NOD mouse the H2 g7 haplotype (168). The human thyroid, islet, and adrenal autoimmune diseases are predominantly associated with HLA-DR3, DR4, and DR5 haplotypes (57). The MHC may affect predisposition to endocrine autoimmune disease by several mechanisms that are not mutually exclusive. Autoantigenic peptides of glandular autoantigens may combine more easily with these particular MHC molecules than with others. It is, however, also possible that the disease association with these MHC haplotypes is due to a specific MHC-controlled shaping of the T cell repertoire.
Regardless of the mechanisms, it is apparent that the MHC haplotype per se is insufficient for the development of an endocrine autoimmune disease, as shown by the fact that autoimmunity-associated HLA-DR haplotypes are also found in perfectly normal individuals. Also, the H2 g7 haplotype of NOD mice in congenic strains does not lead to IDDM in these animals. The genetic analysis of endocrine autoimmune diseases evidently requires an approach other than detailed typing of the MHC encoding genes. Such an approach has been found in the study on microsatellites.
Microsatellites or single-sequence length polymorphisms (SSLPs) are
repeat sequences [usually dinucleotides, e.g.
(CA)n] that exhibit high degrees of polymorphism both
between individuals and in the number of repeats at a given chromosomal
site. SSLPs are abundant (>100.000) and are randomly dispersed
throughout the mammalian genome, thereby providing an enormous pool
from which to derive markers (175). Several thousand microsatellite
markers have thus far been identified and mapped to the mouse/rat and
human genomes, respectively. Specific SSLP loci can easily be defined
by PCR using oligonucleotide primers specific for conserved sequences
flanking the individual repeats, and length polymorphisms among
individuals are identified by electrophoresis of the amplified products
on agarose or polyacrylamide gels. Todd et al. (168)
pioneered the use of microsatellites and other informative markers to
define broadly the genes associated with diabetes in NOD mice. Scanning
the entire genome of the NOD mouse, they obtained evidence of linkage
with ten distinct loci, termed Idd-1 to -10,
distributed on at least nine different chromosomes and affecting
different immunopathological features (Table 2). With
the exception of Idd-1, which is linked with the MHC locus
on chromosome 17, no individual locus appears to be absolutely
essential for disease onset.
|
In human IDDM, previous intrafamilial association studies and limited
chromosomal marker analyses have shown linkage to the MHC (IDDM1) on
chromosome 6, and the insulin locus (IDDM2) on chromosome 11. Two
recent studies using dense microsatellite maps (
300 markers at an
average spacing of 11 centimorgans), reconfirmed the major
importance of IDDM1, but provided limited, if any, support for IDDM2.
Both studies also identified new susceptability loci (Table 2).
|
VI. POF in Association with Adrenal Autoimmunity and/or Addison’s Disease |
|---|
|
TopAbstractI. IntroductionII. Definition and Clinical...III. Cells Involved in...IV. Tolerance to SelfV. Autoimmune Endocrine Disease:...VI. POF in Association...VII. Signs of Ovarian...VIII. Animal Models of...IX. SummaryReferences |
|---|
Addison’s disease is an uncommon disorder (10–20 per million) caused by a deficiency of adrenocortical hormones. The prevalence is highest in the fourth decade of life, and there is a marked female preponderance (2.5:1). The nature of idiopathic Addison’s disease in the majority of patients in developed countries is now regarded as autoimmune (178), in contrast to the nature of the disease in developing countries, which is still mainly due to tuberculosis (179). Autoimmune Addison’s disease seldom develops in isolation, and several other endocrine glands and organs are generally affected (180), leading to an autoimmune polyglandular syndrome (APGS). Two main forms of APGS can be clinically discerned. APGS type I mainly affects children and is characterized by the association of mucocutaneous candidiasis, hypoparathyroidism, and Addison’s disease. Ovarian failure is often part of the syndrome (in approximately 60% of cases). APGS type 1 is also termed APECED (autoimmune polyendocrinopathy-candidosis-ectodermal dystrophy). APGS type II is characterized by adrenal failure in association with hypothyroidism. The latter mainly occurs in the fourth decade of life and has a female preponderance. In this syndrome only 25% of women have amenorrhea and 10% have a classic POF (181, 182).
With regard to POF, the literature indicates that 2–10% is associated with Addison’s disease and/or adrenal autoimmunity (183).
A. Antibodies in POF patients with adrenal autoimmunity and/or
Addison’s disease
The discovery in the 1970s of autoantibodies to the adrenal cortex
(adrenal cytoplasmatic antibodies, Cy-Ad-Abs) formed an important
impetus for the studies on the autoimmune nature of idiopathic
Addison’s disease. Two varieties of adrenal antibodies were
subsequently recognized in the sera of patients with Addison’s disease
using indirect immunofluorescence (IIF) and cryostat sections of human
or monkey adrenal glands. One variety demonstrated reactivity with the
three layers of the adrenal cortex only, whereas the other variety also
reacted with cytoplasmic antigens of other steroid-producing cells
present in the ovary, testis, and placenta (184, 185). This latter
subvariety of adrenal cytoplasmic antibodies was called steroid-cell
antibodies (St-C-Abs), and its reactivity could be absorbed by adrenal
homogenates, thus confirming the cross-reactivity with the adrenal
cytoplasmic antibodies (186). There is an absolute association between
the presence of St-C-Ab and that of Cy-Ad-Ab, the former being
detectable only when the latter is also present. St-C-Ab are of the IgG
type and bind within the ovary to the hilar cells, the cells of a
developing follicle, such as theca and granulosa cells, and to the
corpus luteum cells.
Almost all patients with a primary amenorrhea and Addison’s disease
have a detectable serum titer of St-C-Ab; 60% of patients with a
secondary amenorrhea and Addison’s disease show these antibodies
(Table 3). In the absence of clinically overt gonadal
failure, St-C-Ab have been described in about 15–20% of patients with
clinical or latent Addison’s disease (181). In the follow-up of the
St-C-Ab-positive addisonian patients, about 40% of females developed
ovarian failure in a period of 10–15 yr, whereas in males the St-C-Abs
did not herald gonadal failure (however, numbers of studied patients
were small).
|
): 60–80% of patients with hypoparathyroidism and
Addison’s disease (type I APGS) and 25–40% of patients with type II
APGS show these antibodies. In type 1 APGS without Addison’s disease,
10% of patients show St-C-Abs. The high prevalence of St-C-Ab in
patients with APGS type I probably explains the common association with
gonadal failure seen in this group, and the appearance of the St-C-Abs
in a female patient with APGS type I without adrenocortical or ovarian
failure signals a high risk of their development (185, 187, 188). The
sensitivities/specificities/predictive values for St-C-Abs in females
with type 1 APGS who initially had normal adrenocortical and ovarian
function were 1.0/0.56/0.50 in predicting ovarian failure and
0.86/0.83/0.86 for St-C-Abs in predicting adrenocortical failure (188). The mere presence of an autoantibody in the serum of a patient is certainly not evidence for the pathogenic significance of the antibody; the autoantibody may also be the consequence of cellular destruction, such as is seen after the destruction of cardiac muscle cells in myocardial infarction, giving rise to anti-heart cell antibodies. It has been shown, however,that sera of patients with APGS type I and Addison’s disease, positive for Cy-Ad-Ab and St-C-Ab, are cytotoxic for cultured granulosa cells in the presence of complement, when high titers of these antibodies were demonstrated in nine of 23 cases (189). Complement-dependent cytotoxicity of the St-C-Abs might indeed be one of the mechanisms leading to destruction of steroid-producing cells in vivo and thus to ovarian failure.
In recent years, considerable progress has been made with regard to the
identification of the target antigens of Cy-Ad-Abs and possibly of
St-C-Abs (190). It has been found that the adrenal cytochrome p450
enzyme 21 hydroxylase (which converts 17--progesterone and
progesterone into 11-deoxycortisol and deoxycorticosterone), is the
major autoantigen recognized by autoantibodies present in patients with
Addison’s disease (191, 192), either in the form of isolated adrenal
failure or associated with hypothyroidism (type II APGS).
In type I APGS it is thought that autoantibodies are directed to other
members of the cytochrome p450 enzyme family, namely to the p450
side-chain cleavage enzyme (p450-scc) and to 17--hydroxylase
(17--OH) (192, 193, 194, 195, 196), and to an ill-defined 51-kDa protein (197).
However, there is some confusion on this subject, and not all
investigators could confirm the presence of these autoantibodies in
type 1 APGS [negative results: p450-scc (198); 17--OH (198, 199)].
Of the steroidogenic p450 enzymes 21-hydroxylase is adrenal-specific,
17--OH is expressed in both adrenals and gonads, whereas p450-scc is
present in adrenal, gonads, and placenta. The 51-kDa protein is present
in islets, granulosa cells, and placenta.
Possible targets of the St-C-Abs in POF patients not belonging to the
groups of APGS I or II are thus 17--OH and the p450-scc enzyme.
However, in the one such patient with St-C-Abs, 17--OH was not
recognized (191). To the authors’ knowledge, studies have not been
published on correlations between the presence and activity of St-C-Abs
and autoantibodies to either 17--OH or p450-scc in patients without
APGS type 1. Also, studies in which St-C-Ab activity would be adsorbed
with the enzymes 17--OH or p450-scc would further illuminate the
subject.
Apart from the autoantibodies, another strong argument for considering St-C-Ab-positive ovarian failure as an autoimmune disease is the histology of the ovarian lesions.
B. Histology of the ovaries in patients with POF in combination
with adrenal autoimmunity and/or Addison’s disease
Table 4 gives an overview of the reported histology
of POF, including the reported cases of histologically confirmed
oöphoritis. All St-C-Ab-positive cases had lymphocytic
oöphoritis, and of all lymphocytic oöphoritis cases
reported, 78% had St-C-Abs.
|
Concerning the pattern of microscopical infiltration, there is a marked similarity in the reported cases of lymphocytic oöphoritis in the different compartments of the ovary. In most cases the primordial follicles are unaffected, as well as the cortex of the ovary. It is the developing follicle that is predominantly infiltrated by mononuclear inflammatory cells. There is a clear pattern of increasing density of the infiltration with more mature follicles. Preantral follicles are surrounded by small rims of lymphocytes and plasma cells, whereas larger follicles have a progressively more dense infiltrate usually in the external and internal theca. The granulosa layer is usually spared in this process until luteinization of the degenerating follicle occurs. When cysts are present, they are luteinized with a marked leukocytic infiltration in the cyst wall and destruction of the lining cells. Atretic follicles and, when present, corpora lutea or corpora albicantia are infiltrated as well. This pattern of infiltration confirms that steroid-producing cells are a main target for the autoimmune attack. Mild infiltration might be seen in the medulla and hilar region of the ovaries. There is a perivascular and, surprisingly, a perineural infiltration in the hilus of the ovary (200).
Immunohistochemical analysis of the lymphocytic oöphoritis
reveals that the inflammatory cells are mainly formed by T lymphocytes
(CD4+ and CD8+) with a few B cells, together
with large numbers of plasma cells. Macrophages and NK cells can also
be found. The plasma cells mainly secrete IgG, but also IgA or IgM
(201, 202), which probably indicates the local production of ovarian
autoantibodies. That T cells are important in the ovarian destructive
autoimmune reaction is mainly supported by data generated in the animal
models of autoimmune lymphocytic oöphoritis (see below). The
involvement of T cells also in human oöphoritis is suggested by a
case report on a patient with autoimmune thyroiditis, adrenalitis, and
POF in whom migration-inhibiting factor (MIF) production by peripheral
T cells toward ovarian as well as testicular antigens was found (203).
The MIF test is a sensitive antigen-specific test for the production of
a cytokine, MIF, by peripheral blood T-lymphocytes when cultured in the
presence of specific antigens. It must also be noted in this respect
that granulosa cells of POF patients are more sensitive to -IFN,
another T cell cytokine, than normal granulosa cells (55).
C. Immunogenetic aspects of POF in association with adrenal
autoimmunity and/or Addison’s disease
POF in association with adrenal autoimmunity and/or Addison’s
disease has not been analyzed for any separate immunogenetic
susceptibility for the ovarian component. Autoimmune Addison’s disease
itself is associated with the haplotype HLA-B8/DR3, and in particular
with the DR B11 0301 allele (204).
Ovarian failure in the context of the APGS type I syndrome has been studied in more detail, and APGS type I does not display an HLA-B8/DR3 association. The only association of APGS type I and HLA haplotypes reported so far has been with HLA-A28 (205). Positive associations were found between the presence of HLA-A28 and hypoparathyroidism, adrenocortical failure, and IDDM within the APGS type I syndrome, but not with ovarian failure (205), which is important for this review. Interestingly, in the APGS type I patients with ovarian failure, HLA-A3 was more frequent while HLA-A9 was less frequent than in those with normal ovarian function (205). Using the microsatellite approach, the responsible gene for APGS type I has recently been mapped to the long arm of chromosome 21 (206). The Unverricht-Lundborg type of progressive epilepsy EPM1 has been mapped to the same locus, viz 21q,22.3, and a candidate gene (EHOC-1) for APGS type I, but in particular for EPM1, has been identified as a gene coding for a protein with partial homologies to transmembrane proteins including sodium channel proteins (207).
D. Conclusions
If the above observations are correct, POF in the presence of
Addison’s disease and/or adrenal autoimmunity (only 2–10% of cases)
is almost certainly an endocrine autoimmune disorder. This view is
supported by: 1) the presence of autoantibodies to steroid-producing
cells in the patients, 2) the characterization of shared autoantigens
between adrenal and ovarian steroid-producing cells, and 3) the
histological picture of ovaries of such cases (lymphoplasmacellular
infiltrate particularly around steroid-producing cells).
The existence of an animal model for the autoimmune syndrome of adrenalitis/oöphoritis (see below) lends additional support to this view. It is clear that further genetic studies need to be performed to analyze whether there is a separate (immuno)genetic susceptibility for the ovarian component within the syndrome oöphoritis/adrenalitis.
|
VII. Signs of Ovarian Autoimmunity in Patients with Idiopathic POF in the Absence of Adrenal Autoimmunity and/or Addison’s Disease |
|---|
|
TopAbstractI. IntroductionII. Definition and Clinical...III. Cells Involved in...IV. Tolerance to SelfV. Autoimmune Endocrine Disease:...VI. POF in Association...VII. Signs of Ovarian...VIII. Animal Models of...IX. SummaryReferences |
|---|
. Approximately
60% of such cases of POF lack ovarian follicles, and in these cases
fibrotic ovaries are found. In 40% of the cases, ovarian follicles are
detectable and numbers vary from few to numerous. About 10% of such
follicular cases have numerous follicles, and these latter cases
probably belong to the ROS (208). In 1969, Jones and de Moraes-Ruehsen (50) were the first to report on three patients with ROS; they called it the "Savage" syndrome after the name of their first patient. The syndrome is defined by the presence of numerous primordial follicles in the ovaries, a hypergonadotropic hypoestrogenic hormone profile, and a hyporeceptivity for high dosages of exogenous gonadotropins given for ovulation induction in patients with either primary (49, 50) or secondary amenorrhea (51, 209, 210, 211, 212, 213, 214, 215, 216). ROS patients with a secondary amenorrhea clinically present as POF patients. The etiology of the syndrome is unknown, although several hypotheses have been put forward. These range from a lack of gonadotropin or estrogen receptors, postreceptor pathway disturbances, gonadotropins with inadequate bioactivity (217), serum factors modulating the action of FSH (218), and immune factors such as antibodies to gonadotropin receptors and thymus pathology, which is relevant for this review (216).
It is important to note that cases of lymphocytic oöphoritis can
hardly be found in POF patients in the absence of adrenal
autoimmunity/Addison’s disease [six of 215 cases (Table 4)].
Muechler et al. (219), however, showed the presence of
immunoglobulins in such non-oöphoritis-affected ovaries using
direct immunofluorescence: in 50% of his cases he found vascular wall
staining (IgA, IgM, or IgG), and in 30% the stroma and follicular
cells were positive for immunoglobulins. Hypothetically, autoantibodies
to the ovary may have been present in the ovary without reaching
detectable levels in the serum or inducing a local inflammation. It
must also be noted that Muechler’s data have not been confirmed by
others, and in fact the histology of POF in the absence of adrenal
autoimmunity/Addison’s disease is not helpful in supporting an immune
pathogenesis of the disease. This also applies to the atrophy found in
the majority of cases. This phenomenon may represent the endstage of an
autoimmune process directed against ovarian structures (as is seen in
animal models, see below), but it may also represent a final depletion
of oocytes due to genetic or environmental factors.
More positive evidence of isolated POF representing an endocrine
autoimmune disease is the reported higher than normal frequency of some
other endocrine and neurological autoimmune diseases in POF patients
(Table 5).
|
). Thyroid
autoimmunity is the most prevalent (14%) associated endocrine
autoimmune abnormality reported in POF patients without an adrenal
autoimmune involvement, followed by the presence of parietal cell
antibodies (4%), IDDM (2%), and myasthenia gravis or positivity for
acetylcholine receptor antibodies (2%) (Table 5). However, the general
prevalence of positivity for thyroid antibodies and gastric parietal
cell antibodies is only slightly above the prevalence found in normal
populations. It is, however, remarkable that IDDM and myasthenia
gravis, both relatively uncommon autoimmune diseases (<<1%) are
found relatively frequently in POF patients (2–4%). Whether this high
frequency is due to publication bias or to shared underlying immune
dysregulating factors remains to be established. Systemic lupus
erythematosus (SLE), antinuclear antibodies, and rheumatoid factors
have also been reported with a higher frequency than normal in POF
patients (Table 5). A relationship of POF with SLE is further
strengthened by the finding of the presence of anti-ovarian antibodies
in 84% of the cases of female SLE patients by Moncayo-Naveda et
al. (220).
2. Ovarian autoantibodies. Strong support for an autoimmune
character of isolated POF would be the presence of antibodies to
ovarian structures in the serum of these patients. However, the major
conclusion drawn from several investigations using IIF on gonadal
tissue (animal or human) is that patients are negative for St-C-Abs
(186, 187, 188, 221, 222, 223, 224, 225). It is worthwile to note that positive results
regarding anti-ovarian antibodies have been obtained using assay
methods other than routine IIF (see Table 6), but
control subjects without idiopathic ovarian failure, postmenopausal
women, and patients with iatrogenic ovarian failure were also found
positive in these assays. It has become gradually clear from these
studies that the presence and clinical activity of POF does not
correlate with the presence of these antibodies in serum. Moreover, the
results indicate that although antibodies to ovarian antigens are
common in POF, their pathogenic role remains questionable. They may be
the consequence rather than the cause of the disease.
|
Thus, it is easily understood that receptors such as the LH and FSH
receptors might become targets for blocking antibodies (Fig. 8), and such hypothetical antibodies could be a cause of
ovarian failure. Experiments showing an interaction of antibodies of
POF patients with FSH and LH receptor (function) have been described by
various authors (Table 7). However, inconsistent data
were generated with regard to the prevalence and the exact target of
these antibodies; moreover, receptor antibodies were also found in
patients with iatrogenic ovarian failure.
|
|
In conclusion, the data on receptor antibodies in POF are not conclusive; antibodies to the LH and FSH receptors may exist, but their precise role and prevalence require further studies.
4. Antibodies to zona pellucida (ZP). Yet another specific set of ovarian antibodies playing a role in POF might be the antibodies to the ZP. The ZP is the acellular matrix that surrounds developing and ovulated oocytes and is also detectable in atretic follicles. Autoantibodies to ZP have been described as a cause of infertility in women. In women with unexplained infertility, these antibodies were seen in 5.6% of the cases, whereas in the normal controls positivity was seen in only 1.7% (242). ZP antibodies were thought to interfere with the sperm-oocyte interaction, thus inducing infertility. Animal models have demonstrated, however, that the ZP antibodies interfere with follicular development, and the presence of these antibodies in the experimental animals leads to follicular depletion and amenorrhea (see below). Grootenhuis et al. (unpublished observations), using an enzyme-linked immunosorbent assay, recently found three of 34 POF patients positive for IgG antibodies toward human recombinant ZP3. However, three of six postmenopausal women were also positive, and it is thus likely that the antibodies toward ZP3 are the result of ovarian follicle damage, rather than their cause (in analogy to the other antiovarian antibodies described earlier).
C. Cellular immune abnormalities in patients with idiopathic POF in
the absence of adrenal autoimmunity and/or Addison’s disease
Recent literature (243) in the field of thyroid autoimmunity and
IDDM indicates that immune cells, such as CD4+ Th1
lymphocytes, macrophages, and CD8+ T cytotoxic cells, are
more important in the destruction of endocrine cells in endocrine
autoimmunity as compared with the autoantibodies. So what is the
evidence of such immune cell involvement in POF in the absence of
adrenal autoimmunity and/or Addison’s disease?
Table 8 gives an overview of studies on the numerical
presence of various lymphocyte subsets present in the peripheral blood
of patients with idiopathic POF. Although the data on the numbers of
CD3+, CD4+, and CD8+ T cells vary
between the reported studies, a consistent pattern of an increased
number of activated T cells (as defined by MHC-class II+ or
IL-2R+) is evident in the majority of the studies. Similar
increased numbers of activated peripheral blood T cells have been
described in other autoimmune endocrinopathies, such as recent onset
Graves’ disease (249), IDDM (250), and Addison’s disease (251). A
word of caution is needed, however, because we recently observed that
postmenopausal women may also show raised numbers of activated
peripheral T cells (248). Estrogen substitution lowered the number of
activated peripheral T cells in women with POF, although not to
completely normal levels. Ho et al. (252) also demonstrated
the importance of the estrogen status for the number of peripheral
blood lymphocyte subsets. We therefore consider the hypergonadotropic
hypoestrogenic hormone status present in POF patients and
postmenopausal women as partly responsible for the raised numbers of
activated blood T cells. Another more direct indication of the
involvement of the T cell system in the pathogenesis of POF is given in
the experiments of Pekonen et al. (224). They detected in
several cases of POF a positive MIF test toward gonadal antigens.
|
With regard to the peripheral B cell numbers, two of three studies reported an increase in the number of peripheral blood B cells (245, 248). One (245) correlated the raised numbers of B cells to the presence of various auto-antibodies; we were unable to confirm this correlation (248). A similar increase in the number of peripheral B cells has been observed in other autoimmune endocrinopathies. It is therefore not unreasonable to interpret the raised numbers of peripheral blood B cells as a sign of activation of the humoral immune system crucial for autoantibody production, especially because estrogen substitution in POF women did not lower the raised number of peripheral B cells (248).
With regard to the number and activity of peripheral NK cells in POF, two reports have been published. We showed a decrease in the number of peripheral CD56+/CD16+/CD3- NK cells (248). Pekonen (224) showed decreased activity (lysis of K562 cells) of normal numbers of peripheral blood NK cells in 30% of POF women. A lowered activity of NK cells has also been described in patients with Graves’ disease (253). Because NK cells play a role in immunoregulation, it has been hypothesized that these lowered numbers of NK cells or the lowered activity of the cells might influence B and T cells, resulting in the production of autoantibodies. On the other hand, it has been hypothesized that a decreased activity of the peripheral blood NK cells indicates a susceptibility for viral infection, thus increasing the chance for a viral oöphoritis. However, there is little clinical or histopathological evidence for a viral infection in POF.
An interesting new avenue is the study on the number and functions of monocytes and monocyte-derived DCs in endocrine autoimmune disease. In IDDM (254) and Graves’ disease (255), an abnormal function of monocytes and monocyte-derived DCs (abnormal polarization, abnormal interaction with T cells) has been found by our group. Recently, studies were extended to POF, and similar disturbances were found that were not correctable by estrogen substitution (256).
The abnormalities in the function of peripheral monocytes, monocyte-derived DCs, T cells, and B cells in patients with POF seem to be part of a more complex cell-mediated immune abnormality, including defects in the delayed type hypersensitivity (DTH) reactivity to Candida antigen (256) and the MIF production of peripheral T cells toward this commensal antigen (246). Although we do not understand the clinical significance of these general defects and abnormalities in cell-mediated immunity in POF patients (patients did not show recurrent infections), they might be related to an immunodysregulation leading to endocrine autoimmunity. It must be noted in this respect that patients with chronic mucocutaneous candidiasis (part of the APGS type I) and patients with recurrent vaginal candidiasis (who do show DTH abnormalities to Candida) also show a raised incidence of autoantibodies toward ovarian antigens (257). Whether the APGS type 1 syndrome (where there is a connection between candidiasis and oöphoritis) represents the extreme of a spectrum of disorders combining T cellular deficiencies with ovarian autoimmunity requires further investigation.
D. Conclusions
In conclusion, there is some, albeit debatable, evidence that some
cases of idiopathic POF in the absence of Addison’s disease/adrenal
autoimmunity may belong to the group of endocrine autoimmune diseases.
The positive evidence is formed by the fact that these cases of POF show similar cellular immune abnormalities to other endocrine autoimmune diseases such as IDDM, Graves’ disease, and Addison’s disease. These cellular immune abnormalities include abnormalities in the numbers and/or function of peripheral monocytes, monocyte-derived DC, and subsets of T cells and B cells. Another point of positive evidence might be the more than normal association of POF with IDDM and myasthenia gravis. However, data need to be confirmed and their relevance investigated. Moreover, data on anti-ovarian antibodies and anti-receptor antibodies are not conclusive, because these antibodies, although found by the majority of authors, might be the consequence rather than the cause of the disease. Another point of doubt is that the histology of POF in the absence of Addison’s disease/adrenal autoimmunity hardly shows oöphoritis (<3%).
With regard to immunogenetic studies, one report (258) gives support for a concept that POF belongs to the endocrine autoimmune disorders: POF was associated with HLA-DR3 (258). This study, however, involved only 22 POF patients without adrenal autoimmunity. It must be noted that others were unable to confirm the association in later studies, using larger groups of POF patients of whom only a neglectable number had associated Addison’s disease (259, 260).
|
VIII. Animal Models of Autoimmune Oöphoritis |
|---|
|
TopAbstractI. IntroductionII. Definition and Clinical...III. Cells Involved in...IV. Tolerance to SelfV. Autoimmune Endocrine Disease:...VI. POF in Association...VII. Signs of Ovarian...VIII. Animal Models of...IX. SummaryReferences |
|---|
A. Immunization with crude ovarian antigens.
B. Immunization with well defined ovarian antigens, such as ZP3, or peptides thereof.
C. Neonatal thymectomy in certain strains of mice.
D. Transfer of normal T cells into syngeneic athymic nude (nu/nu) mice.
These four approaches will be described in detail in this review. Other approaches (variants of C. and D.) are the neonatal treatment of mice with cyclosporin A (261), engrafting of fetal rat thymic grafts to nu/nu mice (262), transplantation of neonatal thymic grafts to nu/nu mice (263), and transfer of normal neonatal spleen cells, neonatal thymocytes, and adult thymocytes to syngeneic nu/nu recipients (264).
A. Immunization with crude ovarian antigens
Experimental autoimmune oöphoritis can be induced in
animals, such as the rat and the BALB/c mouse, using immunization with
bovine or rat ovarian extracts in complete Freund’s adjuvant
(265, 266, 267). The immunization establishes an autoimmune allergic
oöphoritis as early as day 14 after immunization, with
infiltration of the ovaries by immune cells. The autoimmune nature of
the oöphoritis is underlined by a positive DTH reaction toward
the injected ovarian antigens by day 14, illustrating the existence of
a cell-mediated immune reaction toward the ovarian antigens. The
appearance of germinal centers in the thymus and increased T cell
activity and B cell stimulation in the spleen indicates that this
experimental oöphoritis involves both T and B cells (267). The
experimental autoimmune oöphoritis could also be induced by
passive transfer of peripheral blood lymphocytes, spleen cells, and
enriched T and B cell suspensions from ovarian antigen-immunized rats
to naive recipients, indicating that T cells and B cells are important
in the pathogenesis of the disease.
Anti-ovarian antibodies in the serum of the affected animals were not detectable before day 28 (265). The reproductive capacity of the rats, measured by litter size, could be correlated to the titer of the anti-ovarian antibodies. Moreover, passive immunization of rats with rabbit anti-rat ovarian serum resulted in a temporary dose-dependent reduction in litter size (268), indicating a role of the antibodies in the pathogenesis of the disorder. The antibodies produced in this experimental oöphoritis animal model are thought to interfere with ZP antigens inhibiting fertilization, and/or to disturb ovulation (269).
Histological examination of the ovarian tissue at day 14 after immunization showed characteristic perivenous accumulations of lymphocytes and macrophages as well as plasma cells (265, 266). The infiltrate was found beneath the tunica albuginea and in the interfollicular tissue, as well as in the granulosa layer of follicles. Occasionally, cell infiltrates were found in the external theca. The large secondary follicles and corpora lutea seemed unaffected, in contrast to the primordial and small secondary follicles. While the number of follicles and corpora lutea were decreased, the number of atretic follicles was increased. It is evident, in comparing the histology of this experimental oöphoritis rat model to the known cases of human autoimmune oöphoritis, that there are major differences. In human autoimmune oöphoritis, the main targets are the steroid-producing cells of the theca of maturing follicles and the corpus luteum, and not the interfollicular space and the secondary and primordial follicles such as seen in this animal model. This implies that the model may have only limited value in the study of human autoimmune oöphoritis and POF.
B. Immunization with heterologous ZP antigens or purified ZP3
antigens
Immunization of New Zealand white rabbits with heterologous ZP
antigens shows an induction of ovarian failure due to follicle
depletion. Immunization experiments with porcine ZP in rabbits showed
the development of ZP antibodies in the immunized animals. It was
demonstrated that rabbits actively immunized with ZP proteins ceased to
ovulate in response to hCG administration (269). The immunization of
the rabbits induced a marked reduction in follicles and an atretic
appearance of primary follicles. Growing follicles disappeared
completely by 30 weeks post immunization. The reduction in number of
normal follicles was accompanied by a striking increase in the number
of oocyte-free cell clusters. An oöphoritis such as that seen in
the immunization experiments with crude ovarian extracts was not
detected (270).
The alteration in ovarian function and histology in the rabbits could be correlated with the presence of serum antibodies to ZP glycoproteins. These studies and the histological pictures indicate first that the antibodies to ZP alter ovarian function and histology by interfering with cells during the stage of follicle differentiation at which ZP proteins are being synthesized (270), and second that the model might be of relevance in the study of human POF in the absence of adrenal autoimmunity. It has been hypothesized that ROS or premature depletion of ovarian follicles might represent the human counterpart of this animal model. However, Starup and Sele (33) showed that in the ovaries of ROS patients there was a normal ultrastructural appearance of the early follicles and no "oocyt-empty" follicle remnants, such as described by Skinner et al. (270) in the rabbit model. On the other hand, two cases of ROS have been reported in whom hyalinization of preantral follicles was described (271, 272).
The proteins of the ZP are conserved among mammals (273). ZP3 is a
major ZP glycoprotein that functions as a sperm receptor (274), and
mouse and human ZP3 proteins are 67% identical. It has been shown that
a 15-amino-acid peptide of ZP3 was able to induce oöphoritis in
(C57BL/6xA/J)F1(B6AF1) mice after immunization in Freund’s complete
or incomplete adjuvant. The histology of the lesion was reminiscent of
the picture of human oöphoritis. ZP3-specific T cell responses
and antibodies directed to ZP3 were detectable in these ZP3-immunized
animals (273). In an adoptive transfer experiment to naive mice,
ZP3-specific CD4+ T cells were sufficient for induction of
the oöphoritis without observable antibody production to the ZP.
The ZP3-specific CD4+ T cells mainly produced IL-2,
IFN-, and TNF, but not IL-4, indicating that the disease-specific T
cells belonged to the TH1 subset of CD4+ T cells.
Subsequently, Lou and Tung (275) very elegantly showed that a transfer of T cells that were directed to the small T cell epitope of ZP3 (15-amino acid peptide) alone differed from the adoptive transfer of T cells to whole ZP3. The former transfer could already result in a full-blown autoimmune oöphoritis, and, apart from a T cell response to the self-peptide and histomorphologically confirmed oöphoritis, serum antibodies to native ZP3 and preferential binding of the antibody to the ZP in vivo were found. Crucial in the experiments was the presence of the ovaries during the antigen-specific CD4+ T cell transfer. The phenomenon shows that B cells autoreactive to ovarian antigens can be generated after a T cell transfer, and that these cells can be activated by ZP3-specific CD4+ T cells to produce antibodies that are directed to and bind ZP3 in vivo. It is thought that the ovarian antigen required for antibody production in this model is provided by the normal ovaries, since the ZP antigens may be generated through a process of follicular atresia (epitope spreading).
Another mechanism by which the ZP autoantibodies can be induced is by
idiotype mimicry of autoantigens in the absence of the antigen itself.
Tung and colleagues (276) also investigated whether a nonovarian
peptide could be recognized by ZP3-specific T cells. The author
detected, by searching the protein sequence library, nonovarian
peptides sharing sufficient residues with ZP3. Interestingly, the
-chain of the murine acetylcholine receptor and the ZP3 peptide had
certain homology. The ZP3 peptide derivate and the -chain of the
acetylcholine receptor both elicited severe oöphoritis and also
stimulated the ZP3-specific T cell clone to proliferate. Through the
mechanism of T cell peptide mimicry, using a -chain of the murine
acetylcholine receptor, autoimmune oöphoritis could be elicited
by clonal activation of ZP3-specific pathogenic T cells. Hence, T cell
epitope mimicry as autoimmune disease mechanism was detected in the
murine model, and this mimicry may explain the clinical association
between POF and myasthenia gravis. Unfortunately, however, in the human
there does not exist a homology between ZP3 and the acetylcholine
receptor (277). Still, it remains remarkable that there is a marked
coincidence between POF and myasthenia gravis (Table 5). The fact that
in human oöphoritis a clear perineural infiltration of the
ovarian hilus nerves is seen might also suggest a shared pathogenic
mechanism between ovarian and neuronal diseases.
C. Neonatal thymectomy models
Neonatal thymectomy in BALB/c mice (and also some other strains,
see below) at day 3 after birth results in oöphoritis, among
other organ-specific autoimmune manifestations such as thyroiditis,
gastritis, and parotitis (278, 279, 280). The inflammations are
characterized by the presence of T cell infiltrates in the affected
organs and the development of organ-specific antibodies in the serum.
There is a strict temporal relationship between the development of the
autoimmune syndrome and the day of thymectomy, which has to occur
between the second and the fifth day after birth (280, 281). An
explanation for the phenomenon has been proposed and is based on the
premise that self-reactive CD4+ T cells are generated in
the thymus throughout life and exported to the periphery. In euthymic
animals, autoimmune disease is not observed because these autoreactive
CD4+ T cells are controlled by CD4+ T cells
with regulatory or suppressor activity. These cells are also generated
in the thymus, but only after the first week of life. Hence thymectomy
at day 3, restricting the T cell repertoire to only effector autoimmune
CD4+ T cells, explains the spontaneously occurring
autoimmune diseases, because the balance between self-reactive T cells
and regulatory T cells tips over to the former.
That animal oöphoritis is directly due to autoimmune T cells is shown by transfer experiments of CD4+ T cells of thymectomized donors to young recipients, which causes an oöphoritis in these recipients (282, 283). This transfer of oöphoritis could be prevented by infusion of CD4+ CD5+ T cells from normal adult mice in an early stage after the transfer of the CD4+ cells of the thymectomized donors.
The histopathological events of the oöphoritis in the thymectomized mice occur in an orderly manner. Initially the oöphoritis is evident as a patchy or diffuse infiltration of lymphocytes; later, developing follicles are clearly affected and monocytes, macrophages, neutrophils, and plasma cells are found between and within ovarian follicles. The onset of puberty markedly potentiates the oöphoritis, indicating that probably a change in antigen profile due to the gonadotropin stimulation is important. The oöphoritis is most severe between 4–14 weeks after thymectomy. This is accompanied by loss of ova and collapse of ovarian follicles. Autoantibodies are detected in the circulation by week 4, with a peak between weeks 7–9. Autoantibodies are directed toward oocytes, ZP, and in lower titers also to steroid-producing cells such as the granulosa cells, the theca cells, and the luteal cells. The inflammation subsides after 14 weeks, and the ovary becomes atrophic (278, 279, 280). IgG-producing plasma cells are found, but not frequently. The overall picture of the oöphoritis is one of a cell-mediated autoimmune reaction.
With regard to the genetics of this ovarian autoimmunity model, certain strains of mice are susceptible, such as the BALB/c and A/J mice, whereas other strains (C57bl/6J, DBA/2 mice) are resistant. Since susceptability and resistance are not associated with the MHC haplotype (H2) of the mice, these molecules are apparently of minor importance. Using the susceptible and resistant mice strains and backcrosses of these strains in combination with a microsatellite approach, a locus has been found on chromosome 16, controlling the abrogation of the tolerance due to neonatal thymectomy day 3 (284). This so-called Aod1 locus was associated with the presence of oöphoritis in the mice. Interestingly, the markers on chromosome 16 failed to exhibit a significant linkage to the concomitant ovarian atrophy in this mouse oöphoritis model. Rather, this atrophy exhibited an association with markers on mouse chromosome 3 (284).
With regard to another experimental mouse model of gonadal
autoimmunity, viz the male counterpart of allergic oöphoritis,
experimental allergic orchitis, studies have shown a similar complexity
of gene involvement and the recognition of various susceptibility and
resistance loci (the Orch genes) (285). In this model the H2 locus is
of importance as a susceptibility locus, and the so-called Orch 1 gene
has been mapped to the Hsp 70.3/G7 interval within the H-2S/H-2D
region. Genes controlling resistance have also been identified: Orch 3
maps centrally on chromosome 11, while Orch 4 maps on chromosome 1.
Orch 5, also on chromosome 1, is probably a gene governing the extent
of the inflammatory lesions seen in susceptable mice. Most significant
is the linkage of Orch 3 to Idd-4 and Orch 5 to
Idd-5, two susceptibility genes that play a role in IDDM of
the NOD mouse model (see above and Table 2).
The histological and serological manifestations of the murine autoimmune oöphoritis are comparable to the histological and serological picture of human autoimmune oöphoritis in association with Addison’s disease. It is remarkable, however, that the adrenal glands are unaffected in the neonatally thymectomized mice, even in the presence of antibodies to steroid-producing cells. However, modifications of the model, viz immunomodulation using Cyclosporin A after birth, does affect the adrenals (261).
As the inflammation of the ovaries subsides, serum anti-oocyte and anti-zona antibodies also decrease to sometimes undetectable levels at day 150–360, when oocytes have completely disappeared from the atrophic ovary (278, 279, 280). Therefore, the absence of serum autoantibodies does not exclude an autoimmune etiology of the ovarian disease. This finding may be of relevance in patients with adrenalitis and/or amenorrhea; detection of St-C-Abs may not always be expected unless they are looked for in an early stage of the disease.
D. Transfer of normal T cells to athymic (nu/nu) mice
Yet another mechanism for the induction of an autoimmune
oöphoritis is the transfer of T cells to athymic nude mice.
The nude mouse model is characterized by a deficient T cell function because the most important function of the thymus, education of T cells to properly recognize self and nonself, cannot take place. When CD4+CD8-thymocytes from normal neonatal or adult BALB/c mice are transferred to athymic mice, approximately 50–75% of the recipients develop an autoimmune oöphoritis and/or gastritis (264). Neonatal CD4+ splenocytes are also able to transfer the autoimmune diseases, whereas T cells from adult spleen do not (286). However, a fraction of adult spleen CD4+ with a low expression of CD5+ can induce oöphoritis in athymic recipients. The disease-generating CD4+ T cells are of the Th1 type. Regulatory T cells that down-regulate self-reactive T cells in this animal model are also present, as studied by the combined infusion of neonatal spleen cells that enhance autoimmune oöphoritis and adult spleen cells that inhibit this process (264). The exact nature of these regulatory cells in this animal model is not yet elucidated; however, it is hypothesized that these belong to the Th2 subset of the CD4+ T cells.
The ovarian histopathology of day 3 thymectomized animals and nude mice that develop an oöphoritis after an adoptive transfer experiment are indistinguishable; hence it might be a good model for POF in the presence of adrenal autoimmunity/Addison’s disease.
In the human situation, experiments of nature show us that athymic girls show evidence of dysgenetic, atrophic ovaries, devoid of follicles (287). Whether this is the ultimate consequence of an autoimmune process that began very early on remains hypothetical.
E. Conclusions
The murine animal oöphoritis models of neonatal thymectomy
or T cell transfer in the nude animal clearly show that pathogenic
self-reactive T cells exist in the normal neonatal and adult repertoire
of at least the mouse, and that these autoreactive effector T cells are
controlled by regulatory T cells also existing in the normal adult
repertoire. There evidently exists a genetically controlled balance
between these two T cell populations that ensures tolerance to ovarian
and other self-antigens. When the balance tips over in favor of
effector T cell activity, autoimmune oöphoritis develops (often
in association with thyroiditis and other endocrine autoimmune
disorders). This type of murine autoimmune oöphoritis is
histologically and serologically similar to the human autoimmune
oöphoritis occurring in association with adrenal autoimmunity and
Addison’s disease (2-10% of POF cases). The autoantibodies and the
histology of this type of POF shows that steroid-producing cells are
one of the main targets of the autoimmune response, and autoantigens of
these steroid-producing cells are gradually identified (important
enzymes in steroid synthesis). When all the data are considered
together, POF in association with adrenal autoimmunity and/or
Addison’s disease and positive for St-C-Abs can now with certainty be
considered as an endocrine autoimmune disease.
The animal models of immunization with crude and purified ZP antigens demonstrate that autoimmune ovarian failure can also be reached via other histological pictures and mechanisms. First, there is the follicular infiltration by various immune cells in the ZP3-peptide immunization model, similar to the thymectomy model and the T cell transfer model in athymic nude mice. This picture is again reminiscent of the autoimmune oöphoritis seen in POF patients with St-C-Abs and associated with Addison’s disease. The infiltration ultimately leads to follicular depletion and fibrosis. Second, there is the simple depletion of follicles after immunization with crude ZP antigens in the absence of a clear lymphocytic infiltration, but with the production of antibodies. This model might be most relevant for POF cases without adrenal involvement and with fibrotic ovaries. Also in the human, ZP autoantibodies have been detected and there is some, though weak, evidence that such POF cases belong to the group of autoimmune diseases (associations with other autoimmune diseases, abnormal numbers and functions of peripheral lymphocytes and monocytes; see earlier discussions in this review). However, further experiments are required to establish or refute such a view.
|
IX. Summary |
|---|
|
TopAbstractI. IntroductionII. Definition and Clinical...III. Cells Involved in...IV. Tolerance to SelfV. Autoimmune Endocrine Disease:...VI. POF in Association...VII. Signs of Ovarian...VIII. Animal Models of...IX. SummaryReferences |
|---|
It is concluded in this review that POF in association with adrenal autoimmunity and/or Addison’s disease (2–10% of the idiopathic POF patients) is indeed an autoimmune disease. The following evidence warrants this view: 1) The presence of autoantibodies to steroid-producing cells in these patients; 2) The characterization of shared autoantigens between adrenal and ovarian steroid-producing cells; 3) The histological picture of the ovaries of such cases (lymphoplasmacellular infiltrate around steroid-producing cells); 4) The existence of various autoimmune animal models for this syndrome, which underlines the autoimmune nature of the disease.
There is some circumstantial evidence for an autoimmune pathogenesis in idiopathic POF patients in the absence of adrenal autoimmunity or Addison’s disease. Arguments in support of this are: 1) The presence of cellular immune abnormalities in this POF patient group reminiscent of endocrine autoimmune diseases such as IDDM, Graves’ disease, and Addison’s disease; 2) The more than normal association with IDDM and myasthenia gravis. Data on the presence of various ovarian autoantibodies and anti-receptor antibodies in these patients are, however, inconclusive and need further evaluation.
A strong argument against an autoimmune pathogenesis of POF in these patients is the nearly absent histological confirmation (the presence of an oöphoritis) in these cases (<3%). However, in animal models using ZP immunization, similar follicular depletion and fibrosis (as in the POF women) can be detected.
Accepting the concept that POF is a heterogenous disorder in which some of the idiopathic forms are based on an abnormal self-recognition by the immune system will lead to new approaches in the treatment of infertility of these patients. There are already a few reports on a successful ovulation-inducing treatment of selected POF patients (those with other autoimmune phenomena) with immunomodulating therapies, such as high dosages of corticosteroids (288, 289, 290, 291, 292).
|
Acknowledgments |
|---|
|
Footnotes |
|---|
1 Work in our laboratory is funded by several grants of NWO-Health
Sciences, the Dutch Diabetic Fund, and the Prevention Fund. 
|
References |
|---|
|
TopAbstractI. IntroductionII. Definition and Clinical...III. Cells Involved in...IV. Tolerance to SelfV. Autoimmune Endocrine Disease:...VI. POF in Association...VII. Signs of Ovarian...VIII. Animal Models of...IX. SummaryReferences |
|---|
-hydroxylase. Lancet 339:770–773[CrossRef][Medline]
This article has been cited by other articles:
 |
|
 |
|
  A. La Marca, S. Marzotti, A. Brozzetti, G. Stabile, A. C. Artenisio, V. Bini, R. Giordano, A. De Bellis, A. Volpe, A. Falorni, et al. Primary Ovarian Insufficiency due to Steroidogenic Cell Autoimmunity Is Associated with a Preserved Pool of Functioning Follicles J. Clin. Endocrinol. Metab., October 1, 2009; 94(10): 3816 - 3823. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Gleicher, A. Weghofer, K. Oktay, and D. Barad Do etiologies of premature ovarian aging (POA) mimic those of premature ovarian failure (POF)? Hum. Reprod., October 1, 2009; 24(10): 2395 - 2400. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E H Dickerson, A S Raghunath, and S L Atkin Initial investigation of amenorrhoea BMJ, August 4, 2009; 339(aug04_1): b2184 - b2184. [Full Text]
|
|
|
|
 |
|
 L. M. Nelson Primary Ovarian Insufficiency N. Engl. J. Med., February 5, 2009; 360(6): 606 - 614. [Full Text] [PDF]
|
|
|
|
 |
|
 S. O'Leary, M. L. Lloyd, G. R. Shellam, and S. A. Robertson Immunization with Recombinant Murine Cytomegalovirus Expressing Murine Zona Pellucida 3 Causes Permanent Infertility in BALB/c Mice Due to Follicle Depletion and Ovulation Failure Biol Reprod, November 1, 2008; 79(5): 849 - 860. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Cervera and J. Balasch Bidirectional effects on autoimmunity and reproduction Hum. Reprod. Update, May 22, 2008; (2008) dmn013v1. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Tsigkou, S. Marzotti, L. Borges, A. Brozzetti, F. Reis, P. Candeloro, M. Luisa Bacosi, V. Bini, F. Petraglia, and A. Falorni High Serum Inhibin Concentration Discriminates Autoimmune Oophoritis from Other Forms of Primary Ovarian Insufficiency J. Clin. Endocrinol. Metab., April 1, 2008; 93(4): 1263 - 1269. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Z. Altuntas, J. M. Johnson, and V. K. Tuohy Autoimmune Targeted Disruption of the Pituitary-Ovarian Axis Causes Premature Ovarian Failure J. Immunol., August 1, 2006; 177(3): 1988 - 1996. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Sundblad, V. A. Chiauzzi, L. Andreone, S. Campo, E. H. Charreau, and L. Dain Controversial role of inhibin {alpha}-subunit gene in the aetiology of premature ovarian failure Hum. Reprod., May 1, 2006; 21(5): 1154 - 1160. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Goswami and G. S. Conway Premature ovarian failure Hum. Reprod. Update, July 1, 2005; 11(4): 391 - 410. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Steckler, J. Wang, F. F. Bartol, S. K. Roy, and V. Padmanabhan Fetal Programming: Prenatal Testosterone Treatment Causes Intrauterine Growth Retardation, Reduces Ovarian Reserve and Increases Ovarian Follicular Recruitment Endocrinology, July 1, 2005; 146(7): 3185 - 3193. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. K. Welt, A. Falorni, A. E. Taylor, K. A. Martin, and J. E. Hall Selective Theca Cell Dysfunction in Autoimmune Oophoritis Results in Multifollicular Development, Decreased Estradiol, and Elevated Inhibin B Levels J. Clin. Endocrinol. Metab., May 1, 2005; 90(5): 3069 - 3076. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Tauchmanova, R. Carrano, M. Sabbatini, M. De Rosa, F. Orio, S. Palomba, T. Cascella, G. Lombardi, S. Federico, and A. Colao Hypothalamic-pituitary-gonadal axis function after successful kidney transplantation in men and women Hum. Reprod., April 1, 2004; 19(4): 867 - 873. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Wu, K. H. Van der Hoek, N. K. Ryan, R. J. Norman, and R. L. Robker Macrophage contributions to ovarian function Hum. Reprod. Update, March 1, 2004; 10(2): 119 - 133. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Forges, P. Monnier-Barbarino, G.C. Faure, and M.C. Bene Autoimmunity and antigenic targets in ovarian pathology Hum. Reprod. Update, March 1, 2004; 10(2): 163 - 175. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Tauchmanova, C. Selleri, G. D. Rosa, M. Esposito, F. Orio Jr, S. Palomba, G. Bifulco, C. Nappi, G. Lombardi, B. Rotoli, et al. Gonadal status in reproductive age women after haematopoietic stem cell transplantation for haematological malignancies Hum. Reprod., July 1, 2003; 18(7): 1410 - 1416. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.L. Luborsky, P. Meyer, M.F. Sowers, E.B. Gold, and N. Santoro Premature menopause in a multi-ethnic population study of the menopause transition Hum. Reprod., January 1, 2003; 18(1): 199 - 206. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V.K. Bakalov, V.H. Vanderhoof, C.A. Bondy, and L.M. Nelson Adrenal antibodies detect asymptomatic auto-immune adrenal insufficiency in young women with spontaneous premature ovarian failure Hum. Reprod., August 1, 2002; 17(8): 2096 - 2100. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Betterle, C. Dal Pra, F. Mantero, and R. Zanchetta Autoimmune Adrenal Insufficiency and Autoimmune Polyendocrine Syndromes: Autoantibodies, Autoantigens, and Their Applicability in Diagnosis and Disease Prediction Endocr. Rev., June 1, 2002; 23(3): 327 - 364. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z.-B. Tong, C. A. Bondy, J. Zhou, and L. M. Nelson A human homologue of mouse Mater, a maternal effect gene essential for early embryonic development Hum. Reprod., April 1, 2002; 17(4): 903 - 911. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Bagavant, C. Sharp, B. Kurth, and K. S. K. Tung Induction and Immunohistology of Autoimmune Ovarian Disease in Cynomolgus Macaques (Macaca fascicularis) Am. J. Pathol., January 1, 2002; 160(1): 141 - 149. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. F. Muller, H. A. Drexhage, and A. Berghout Postpartum Thyroiditis and Autoimmune Thyroiditis in Women of Childbearing Age: Recent Insights and Consequences for Antenatal and Postnatal Care Endocr. Rev., October 1, 2001; 22(5): 605 - 630. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Reimand, P. Peterson, H. Hyöty, R. Uibo, I. Cooke, A. P. Weetman, and K. J. E. Krohn 3{beta}-Hydroxysteroid Dehydrogenase Autoantibodies Are Rare in Premature Ovarian Failure J. Clin. Endocrinol. Metab., June 1, 2000; 85(6): 2324 - 2326. [Abstract] [Full Text]
|
|
|
|
|
|
J. Luborsky, B. Llanes, R. Roussev, and C. Coulam Ovarian antibodies, FSH and inhibin B: independent markers associated with unexplained infertility Hum. Reprod., May 1, 2000; 15(5): 1046 - 1051. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y.M. van Kasteren, R.D.L. Hundscheid, A.P.T. Smits, F.P.M. Cremers, P. van Zonneveld, and D.D.M. Braat Familial idiopathic premature ovarian failure: an overrated and underestimated genetic disease? Hum. Reprod., October 1, 1999; 14(10): 2455 - 2459. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. H. Melner and F. A. Feltus Editorial: AutoImmune Premature Ovarian Failure--Endocrine Aspects of a T Cell Disease Endocrinology, August 1, 1999; 140(8): 3401 - 3403. [Abstract] [Full Text]
|
|
|
|
|
|
Z.-B. Tong and L. M. Nelson A Mouse Gene Encoding an Oocyte Antigen Associated with Autoimmune Premature Ovarian Failure Endocrinology, August 1, 1999; 140(8): 3720 - 3726. [Abstract] [Full Text]
|
|
|
|
|
|
Philip.E. Castle and J. Dean Manipulating the genome to study reproduction: Mice with `humanized' zonae pellucidae Hum. Reprod., August 1, 1999; 14(8): 1927 - 1939. [Full Text] [PDF]
|
|
|
|
|
|
S.N. Kalantaridou, D.T. Braddock, N.J. Patronas, and L.M. Nelson Treatment of autoimmune premature ovarian failure: Case report Hum. Reprod., July 1, 1999; 14(7): 1777 - 1782. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Ishizuka, Y. Kudo, A. Amemiya, H. Yamada, T. Matsuda, and T. Ogata Anti-nuclear antibodies in patients with premature ovarian failure Hum. Reprod., January 1, 1999; 14(1): 70 - 75. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. L. Kuijpens, M. de Haan-Meulman, H. L. Vader, V. J. Pop, W. M. Wiersinga, and H. A. Drexhage Cell-Mediated Immunity and Postpartum Thyroid Dysfunction: A Possibility for the Prediction of Disease? J. Clin. Endocrinol. Metab., June 1, 1998; 83(6): 1959 - 1966. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
