| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Pediatric Surgical Research Laboratories, Department of Surgery, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts 02114
Correspondence: Address all correspondence and requests for reprints to: Jose Teixeira, Pediatric Surgery/WRN1024, 32 Fruit Street, Boston, Massachusetts 02114. E-mail: teixeira{at}helix.mgh.harvard.edu
 |
Abstract |
|---|
 Top Abstract I. IntroductionII. The MIS Ligand...III. MIS ReceptorIV. MIS in the...V. Alternative MIS Signal...VI. Outlook for MISVII. ConclusionReferences |
|---|
B, ß-catenin, and p16 activation.
Finally, we will assess the clinical relevance of studying MIS in
patients with persistent Müllerian duct syndrome and our efforts
to determine the therapeutic value of MIS for patients with ovarian and
other MIS receptor-expressing cancers. I. Introduction
II. The MIS Ligand and Its Biological Effects
A. History of the discovery of MIS
B. Müllerian duct regression
C. Cloning, expression, purification, and cleavage of MIS
D. Diagnostic utility of MIS measurements: ELISA and bioassays
E. Regulation of MIS expression
F. MIS transgenics and knockouts
III. MIS Receptor
A. Signal transduction in the TGFß family
B. MIS type II receptor (MISRII)
C. Candidate MIS type I receptors
IV. MIS in the Adult
A. MIS in the testis
B. MIS in the ovary
V. Alternative MIS Signal Transduction
A. MIS-mediated signal transduction in the ovary
B. MIS-mediated signal transduction in the breast
VI. Outlook for MIS
A. Genetics of MIS and MIS receptor in persistent Müllerian duct syndrome
B. Genetics of MIS receptor in ovarian and breast cancer
C. New delivery systems
D. Other sources of MIS
VII. Conclusion
|
I. Introduction |
|---|
|
TopAbstractI. IntroductionII. The MIS Ligand...III. MIS ReceptorIV. MIS in the...V. Alternative MIS Signal...VI. Outlook for MISVII. ConclusionReferences |
|---|
).
T, another hormone produced by the fetal testis, is required for
Wolffian duct differentiation into the male internal reproductive tract
structures. MIS-mediated regression of the Müllerian duct is a
classic example of tissue resorption, which, if understood and
harnessed, could be instructive in how to control the growth of
Müllerian duct-derived tumors, particularly those that bind the
ligand. Efforts to uncover the molecular mechanisms of MIS-mediated
Müllerian duct regression by several laboratories over the last
few decades and previously reviewed in Refs. 1, 2, 3, 4 have
brought the study of MIS from phenomenology as a biological curiosity
into the molecular era. The genes encoding MIS and its receptors have
been cloned, and the signal transduction pathways that are used by MIS
are being investigated to understand more fully its role in fetal and
adult gonads, extragonadal tissues, and in deregulated tissues
including cancers. The sexually dimorphic regulation of MIS expression
is also a fruitful area of inquiry since it is an early marker in
mammals for the genetic switch that occurs when a bipotential gonad is
instructed to differentiate into a testis in response to the
testis-determining factor, SRY (sex-determining region of the Y
chromosome) (5, 6).
|
|
II. The MIS Ligand and Its Biological Effects |
|---|
|
TopAbstractI. IntroductionII. The MIS Ligand...III. MIS ReceptorIV. MIS in the...V. Alternative MIS Signal...VI. Outlook for MISVII. ConclusionReferences |
|---|
Alfred Jost (8) performed seminal in vivo embryonic experiments that demonstrated the existence of what he called the Müllerian "l’hormone inhibitrice" or the "Müllerian inhibitor." Using embryonic testicular fragments implanted in embryonic female rabbits before the time of sexual differentiation, he found, like the freemartin calf, that the animals were virilized externally and showed stimulation of Wolffian ducts and regression of the Müllerian ducts. When he replaced the testes with T pellets, the female embryos became masculinized, but did not show regression of the Müllerian ducts. These experiments caused him to propose the existence of another testicular product (9) in addition to T, which we now call Müllerian inhibiting substance (MIS). These observations allowed him to explain the phenotypic differences between patients with intersex abnormalities due to congenital adrenal hyperplasia, in which excess adrenal T masculinized the external genitalia of a female, but did not affect the normal Müllerian ducts, and androgen insensitivity syndromes, then referred to as testicular feminization, in which the testes continued to produce functional MIS, causing normal regression of the Müllerian duct (10).
In the absence of research funding after World War II in France, Jost used rabbits indigenous to the Bois de Boulogne to perform his experiments. They were hardy animals that withstood manipulation in utero. The New Zealand White rabbits later developed for research purposes were more susceptible to spontaneous abortion; hence, Jost’s experiments proved difficult to repeat. Thankfully, they were later recapitulated in vitro when Picon (11) in Jost’s laboratory developed an organ culture assay using rat embryonic agonadal urogenital ridges, which showed regression of the Müllerian duct when cocultured with fetal testes.
B. Müllerian duct regression
In the late 1950s the concept of programmed cell death was
introduced, and Hamilton and Teng (12) later described the
regression of the Müllerian duct as a classic example of
apoptosis in tissue remodeling during organogenesis. This process, also
observed during resorption of the interdigital spaces in the formation
of the digits (13, 14) and with branching morphogenesis of
the salivary gland and the lung (15), was characterized by
the formation of dense intracellular particles rich in lysosomal
enzymes, which preceded phagocytosis by adjacent cells. Electron
microscopy of the regressing male Müllerian ducts clearly
visualized the autophagocytotic process (16); similar
observations were made on Müllerian ducts in vitro
under the influence of cocultured testicular fragments
(17). Further analysis using electron microscopy and
markers for membrane-specific proteins demonstrated breakdown of the
basement membrane as one of the earliest morphological signs of
Müllerian duct regression (18, 19). The
fragmentation was accompanied by loss of fibronectin staining
(20) and detachment of the epithelial cells, which were
observed to migrate out of the epithelial compartment to contribute to
a mesenchymal cuff around the disrupted basement membrane (18, 21). Some of the cells undergoing epithelial-mesenchymal
transformation migrated deep into the mesenchyme (18, 22).
When studied in chick-quail chimeras (23), these migrating
cells were observed to take up residence progressively in the
mesonephros, which is itself subsequently resorbed.
That there must be cross-talk between the mesenchymal and epithelial layers for Müllerian duct regression to occur was recognized early (24, 25), an observation strengthened by the expression of the MIS receptor components in the mesenchyme surrounding the duct (see Sections III.B and III.C). More recent analyses of internucleosomal DNA fragmentation, a hallmark of the apoptotic process, have confirmed that MIS can induce apoptosis in the Müllerian duct epithelial cells (26) via a paracrine mechanism from signals derived from the MIS type II receptor-expressing mesenchyme (27) and that the apoptosis is progressive from the cranial to caudal direction and might involve ß-catenin (22). It has been appreciated for some time that the cranial or cephalic part of the Müllerian duct, which becomes the Fallopian tube, was more sensitive to MIS (22, 28). The histology of this Müllerian domain characterizes the most common human ovarian cancers, the cystadenocarcinomas, which are often referred to as Müllerian tumors (29).
C. Cloning, expression, purification, and cleavage of MIS
Evidence that the Sertoli cells produce MIS was first shown by
Josso (30), who separated seminiferous tubules from
interstitial cells and cultured each separately with the agonadal
urogenital ridge from 14-d-old rat fetuses and found that the
Müllerian duct regressed under the influence of only the
seminiferous tubules, even when depleted of germ cells
(31). The Sertoli cell factor was found to be heat
sensitive and thus was likely a protein (30, 32). Rat
testes continued to produce enough MIS after birth to cause some
regression of the Müllerian duct, but the levels fell
precipitously by 3 wk (33), whereas calf testes continued
production for 8–10 postnatal weeks (34). Therefore, both
fetal and newborn calf testes were used as a source for purification of
MIS (34, 35). Various extraction techniques of testes
fragments, followed by dialysis, ammonium sulfate precipitation, and
ion exchange chromatography, were used to purify biologically active
fractions (36) that were further purified by carbohydrate
and dye affinity chromatography (37, 38). Homogeneous
fractions of MIS were first purified using a combination of ammonium
sulfate precipitation, ion exchange chromatography, and iterative
immunoaffinity chromatography that resulted in a single major band at
72 kDa (39).
N-terminal and internal sequencing yielded amino acid sequences that were used to design degenerate oligonucleotide probes to screen a cDNA library constructed from newborn calf testes (40) from which the protein itself had been originally purified (38). A full-length bovine MIS cDNA was cloned and used as a probe to clone human genomic MIS and two overlapping human cDNAs, which covered the predicted full-length human cDNA. Similarly, overlapping cDNAs were also cloned from a bovine fetal testis library (41). Transfection of human genomic MIS in Chinese hamster ovary (CHO) cells resulted in expression and secretion of the full-length human MIS protein (40), which was functional in the organ culture assay of Müllerian duct regression and in antiproliferative assays. The MIS gene and/or cDNA has also been cloned in other species including rat (42), mouse (43), chicken (44, 45), and alligator (46). The overall homology between the human sequence and the other species range from 27% and 33% identity for chicken and alligator to between 74% and 68% for bovine, rat, and mouse.
Monoclonal and polyclonal antibodies were raised to the MIS protein and used effectively for Western blot analysis, immunohistochemistry, and for the design of a specific and sensitive MIS ELISA (47, 48, 49). These monoclonal antibodies were also used to purify MIS from serum-containing media of CHO cells transfected with the MIS gene (50). Recombinant human MIS (rhMIS), purified by immunoaffinity chromatography (50) and found to cause regression of the Müllerian duct, was used in antiproliferation assays where it inhibited growth of Müllerian tumors in vitro (51) as well as growth and metastases of an ocular melanoma cell line in vitro and in vivo (52, 53, 54). Holo-MIS was cleaved into its N- and C-terminal domains by treatment with plasmin (55). The C-terminal domain was purified and found to be the biologically active moiety after cleavage at position 427/428 at a kex-like cleavage site characterized by R-4XXR-1 with a serine in the +1 site, which makes the MIS cleavage site monobasic (56). Addition of the N terminus was shown to enhance the biological activity of the C-terminal moiety (57), an unusual result for the TGFß family. Because its expression has been correlated with MIS and it can efficiently cleave pro-MIS, PC5 has been suggested as the natural proprotein convertase for MIS (58).
More recently, we have employed fast performance liquid chromatography to purify human MIS to homogeneity from serum-free media of transfected CHO cells producing rhMIS. Using serum-free media as the starting material substantially reduces the number of proteins from which the rare MIS product must be separated. The resultant more homogeneous MIS preparation, which is produced at higher yield and is up to 10-fold more potent when tested in bioassays (58A ) is being scaled up to conform to Food and Drug Administration requirements for preclinical in vivo testing against human ovarian tumor xenotransplants, before phase I clinical trials. After fast performance liquid chromatography separation of plasmin-treated holo-MIS from this preparation, the C terminus appears as a single band by gel electrophoresis and is suitable for crystallography (T. Stehle, P. K. Donahoe, and D. T. MacLaughlin, unpublished observations).
D. Diagnostic utility of MIS measurements: ELISA and
bioassays
A mouse monoclonal antibody raised to recombinant human holo-MIS
(48) recognized holo, but not the C-terminal domain of
rhMIS. Purified holo-MIS protein was also used to raise rabbit
polyclonal antibodies that, when combined with the monoclonal 
MIS in
a sandwich assay, were used to constitute an ELISA that was sensitive
to 1 to 2 ng/ml. When normal humans were tested, MIS was found to be
elevated constitutively in serum of prepubertal males. MIS levels were
shown to drop for a short time immediately after birth, peak within
3–6 months, and were maintained throughout infancy and childhood. MIS
then slowly decreased just before puberty when a fall in MIS preceded
the onset of puberty (59, 60). The lower levels were
subsequently maintained throughout male adult life in the 2–5 ng/ml
range. The concentrations of postneonatal serum MIS and T are inversely
related in males (see Fig. 2)
(59, 60, 61, 62). Females, conversely, had undetectable serum
levels by ELISA until the prepubertal period (47, 48, 49),
although recent immunohistochemical evidence indicates that MIS might
be expressed in granulosa cells of preantral follicles
(63). After puberty, serum MIS levels of 2–5 ng/ml (
10
pM) were maintained until menopause when MIS was no longer
be detectable. Thus, in childhood, MIS can serve as a useful and
reliable marker (60) for the presence of testicular tissue
when levels of T are very low. In the adult, when MIS is produced both
in the Sertoli cell in the male and the granulosa cells in the female,
the assay could not differentiate between normal males and females
(60).
|
MIS is particularly helpful in the patient with bilateral nonpalpable gonads (68). In patients with bilateral testicular torsion in utero, MIS levels will be undetectable; the external genitalia is masculine, but the gonads cannot be palpated due to torsion, which presumably occurred late in gestation after external virilization had taken place under the influence of previously normal testes. In cases of bilateral retractile testes, MIS levels would be normal. These gonads can often, but not always, be maneuvered into and palpated in the inguinal canal or scrotum; when this is not possible, normal MIS levels provide reassurance that testes (68) are present, as it is in patients with congenital hypogonadotropic hypogonadism (69), since ultrasound and magnetic resonance imaging may be less reliable at this age. It has not been possible to use the MIS ELISA to help in assessing the unilateral undescended testes since levels are not significantly different from normal.
MIS levels can be dramatically elevated in females with granulosa cell tumors (70, 71, 72, 73). The ELISA could be used to detect new and recurrent cases of juvenile granulosa cell tumors, which tend to be benign, or the often postmenopausal adult tumors that are highly malignant and for which the best therapy remains early surgical excision. These tumors are often unresponsive to radiotherapy; multiple chemotherapeutic agents are necessary to elicit a response, but subsequent to initial responses they can have a high recurrence and mortality rate. The assay is probably best used in detecting early recurrences, when rising MIS levels have been detected before such tumors can be seen by computerized tomography or magnetic resonance imaging. Hence, we and others have turned to MIS assays to help manage these patients with granulosa cell tumors (70, 71, 72). Although elevated E can be helpful in the differential diagnosis of granulosa cell tumors, it is not informative when a normal ovary is still in place as is often the case in patients with juvenile granulosa cell tumors (74, 75).
E. Regulation of MIS expression
MIS is a critical component in the cascade of mammalian sex
differentiation as an early gene product after the gonad differentiates
into a testis under the influence of SRY (76), and its
embryonic sexually dimorphic expression has been a useful marker in the
identification of factors activating the molecular switch that causes
the indifferent gonad to differentiate into a testis. MIS expression is
conserved among mammalian species and is tightly regulated in a
developmental and tissue-specific manner in fetal, neonatal,
prepubertal, and adult Sertoli cells, as well as prepubertal and adult
granulosa cells. Despite intense study, the factors that determine this
complex MIS expression pattern are still not fully understood.
Transcription factors and promoter elements that affect both basal and
regulated MIS expression during the fetal and neonatal period (as
reviewed by Ref. 6) and perhaps in some cases in the adult
have been elucidated by a number of laboratories. Also, correlative
data indicate that MIS expression might be regulated by androgens
(66), gonadotropins (77, 78), and by meiotic
germ cells (79). There is no direct evidence that MIS
expression is regulated by estrogens. However, it has been observed
that MIS mRNA abundance in the ovary is increased 2-fold in ER and
-ß double knockout mice (ERKOß) (80), but it may
not be reflected at the protein level (81).
Transcription of eukaryotic genes is dependent upon regulatory elements in the enhancer and core promoter region; the latter is usually composed of consensus TATA and CCAAT box elements. Unlike the rodent MIS core promoters, the human MIS core promoter was studied (82) because it had no consensus TATA element. After deletion and mutational analyses, a region of the core promoter (-6 to +10) found to be essential for basal transcription was identified as the human MIS initiator element, although consensus with other known initiator elements of, for example, the adenovirus major late (83) or Vß genes (84), was not high. Transcription of the MIS gene from this region was found to require transcription factor, TFII-I (82), a TATA-associated factor that recognizes selected initiator elements.
The SAP62 gene is only 328 bp upstream of the start site for transcription of mouse MIS and only 762 bp upstream for that of human MIS (85). Hence, it is likely that the region conferring critical control of MIS expression is located relatively close to the transcriptional start site. Moreover, this proximal promoter displays a number of evolutionary conserved elements including those for steroidogenic factor 1 (SF-1), an orphan nuclear receptor critically important for development of the hypothalamic-pituitary-gonadal and adrenal axes (86, 87); GATA-4, a zinc-finger transcription factor that binds to a WGATAR site in the 5'-flanking region of target genes and is highly expressed as early as e15 in the primordial genital ridge (88); and Sox9, an Sry-related homeobox gene that, when mutated, causes campomelic dysplasia and can cause sex reversal (89, 90, 91). A 180-bp proximal promoter region was found to be sufficient to drive expression of mouse MIS in 15-d postnatal Sertoli cells (92), and in that region an SF-1 binding site (93) was required for activation, since a 2-bp mutation abrogated expression. GATA-4 could activate MIS transcription and does so synergistically when coexpressed with SF-1 (88, 94, 95). Mutation of a Sox9 binding site provided convincing evidence that it too was critical for MIS expression both in vitro (96) and in vivo (97). Complicating this interpretation is the observation that, in both chickens (98) and alligators (46), Sox9 expression follows MIS expression temporally. Since MIS is expressed in both male and female embryonic gonads in chickens and an increase in MIS expression in males is coincident with an increase in Sox9 expression (98), these observations may indicate a mechanism for Sox9-mediated up-regulation of MIS expression above basal levels.
We discovered (99), using adenoviral transfection of 2-d postnatal rat Sertoli cells, which on matrigel continued to express MIS for 6 d in culture (100), that multiple combinational interactions were required to maintain MIS expression. By contrast, when postnatal bovine Sertoli cells were cultured on plastic, MIS production fell dramatically and could not be sustained (101). We found that within the -269 region of the human promoter, that proximal and distal SF-1 sites and proximal and distal GATA-4 sites, as well as a Sox9 site, were required for activation of MIS, as measured by luciferase expression. In addition to these functional transcription assays, EMSAs and deoxyribonuclease I footprinting confirmed the importance of these combinatorial sites (99). Strict requirements for a functional Sox9 were further confirmed by "knocking in" mutations at the Sox9 site, which caused the phenotype in the male of retention of the Müllerian ducts, whereas an inserted mutation of the proximal SF-1 site diminished, but did not ablate, MIS expression (97). Given the latter observation in this in vivo model, it is anticipated that mutation of both SF-1 sites would likely produce complete inactivation, since double mutation of the proximal and distal SF-1 sites ablated luciferase activity in vitro in transfection reporter assays in postnatal day 2 Sertoli cells, when compared with the single mutations alone (99). The essential Sox9 site, located between the distal and proximal SF-1 sites, and the GATA-4 binding sites may permit DNA bending and binding to enhance the activation of the MIS promoter (99). While not directly binding to DNA, Dax-1 (dosage-sensitive sex reversal, adrenal hypoplasia congenital, critical region on the X chromosome gene-1) (102, 103), which is mutated in congenital adrenal hypoplasia (104), acts as a repressor of this combinatorial activation by recruiting the nuclear corepressor, NCoR (105), whereas the Wilms’ tumor suppressor, WT1, acts as an enhancer (106) of MIS activation. These observations raise the intriguing possibility that inhibiting an inhibitor of MIS expression might be useful should higher MIS concentrations prove therapeutic.
Intuitively, one would expect that the inverse correlation between
postnatal MIS expression and serum T (Fig. 2
) was due to either
androgen regulation of MIS expression or the regulation of the T
homeostasis by MIS, or both. There is evidence for both activities, and
regulation of the steroidogenic pathway in Leydig cells by MIS is
discussed later in Section IV.A. Analyses of patients with
either central or gonadotropin-independent precocious puberty and
androgen insensitivity syndromes supported the hypothesis that elevated
levels of serum T were concomitant with lower levels of MIS and that
the result of ablation of AR signaling was elevated serum MIS
(62, 66). The serum of normal mice injected daily for
5 d after birth with PMSG showed a significant reduction in MIS
concentration, whereas AR mutant Tfm mice and normal mice injected with
PMSG and FSH, respectively, showed the opposite effect
(107). These results suggest that the LH component of PMSG
could lower MIS expression by stimulating T synthesis.
When LH or T was injected directly into fetuses on day 18–19 of gestation, holo-MIS immunostaining remained the same, but specific immunostaining for the C-terminal or the N-terminal fragment on day 21 increased (108). From these results we speculated that T stimulated the enzyme responsible for cleavage of MIS. When fetuses were injected similarly with FSH, immunostaining for MIS was reduced using all three antibodies, indicating that FSH inhibits MIS synthesis, in support of previous observations using antisera to GnRH (77). No cAMP-responsive elements are evident in the MIS promoter, suggesting that the effect is probably not a direct transcriptional one. A synergy between entry of primary spermatocytes into the meiotic prophase and androgen down-regulation of MIS expression has also been observed in mice (107, 109), which may also be the case in humans even though down-regulation of MIS expression can occur in the absence of spermatogenesis, as is found in patients with germ cell depletion (63). An alternative explanation for the inverse correlation between meiosis and MIS concentration is that higher levels of MIS inhibit completion of the first meiotic division, which has been shown in immature rat ovaries (110). The increased expression of MIS that correlated with the death of germ cells in ectopic gonads (79) leaves open the search for a germ cell inhibitor of MIS expression.
It has been one of our early goals, as yet unattained, to restore MIS expression in aging ovaries in the hope of averting postmenopausal ovarian cancer. However, turning on regulation of MIS is an intricate combinatorial mechanism that has so far been elusive and labyrinthine.
F. MIS transgenics and knockouts
Overexpression of human MIS in mice resulted in the expected
phenotype of females with blunt vaginas and lacking Fallopian tubes and
uteri (111), providing in vivo confirmation
that, in otherwise normal mice, expression of the MIS gene was
sufficient for Müllerian duct regression. Interestingly, the
chronic exposure of these females to MIS under control of the
metallothionein promoter led to other reproductive tract anomalies,
such as loss of germ cells in the ovaries and their reorganization into
seminiferous tubule-like structures in the adult. When this
"freemartin" effect was also produced after ectopic transplantation
of fetal gonads and correlated with the loss of germ cells, MIS
immunostaining in the induced tubules increased, indicating that the
germ cells suppressed MIS expression (79). Subsequent
studies showed that folliculogenesis, aromatase activity, and meiosis
were retarded by MIS in the ovaries (112). Males
expressing the highest levels of MIS were also abnormal. They were
undervirilized externally, had impaired Wolffian duct development, and
had undescended testes, all presumably due to Leydig cell hypoplasia
(111), which was later shown to be due to the effect of
MIS overexpression on Leydig cell development and steroidogenic
capacity (113).
In addition to the identification of mutations in MIS (114) or its receptor (115) in patients with persistent Müllerian duct syndrome (PMDS), the function of MIS and the specificity of the MIS type II receptor (MISRII) were unequivocally established when separate knockouts of the ligand (116) and receptor (117) produced identical phenotypes of viable male offspring with retention of the Müllerian ducts. Older males were noted to have Leydig cell hyperplasia and the rare Leydig cell tumor. Double knockouts of inhibin and MIS had a higher incidence of Leydig cell neoplasias that were seen earlier, were less hemorrhagic, and produced less E2 than the testes of inhibin-deficient controls (118). These findings prompted studies in a number of laboratories of the role of MIS in the regulation of Leydig cell development and T expression, which will be discussed below (see Section IV.A). The specificity of the MISRII for the MIS ligand was demonstrated by producing combined MISRII-null and MIS-overexpressing females that were normal despite very high levels of MIS (119).
|
III. MIS Receptor |
|---|
|
TopAbstractI. IntroductionII. The MIS Ligand...III. MIS ReceptorIV. MIS in the...V. Alternative MIS Signal...VI. Outlook for MISVII. ConclusionReferences |
|---|
), progress with MIS has been
slower. However, the discovery of a mutation in MISRII in a patient
with PMDS (115) and deletion of the MISRII gene by
homologous recombination in mice (117) provide conclusive
evidence of its identity.
|
The signal transduction pathways employed by the TGFß family have fallen into two distinct sets: the TGFß/activin group of type I receptors (ALK5, -1, and ALK4) and the BMP/GDF group (ALK1, -2, -3, and -6). The ligand specificity within the family is determined by the type II receptor, which binds ligand cooperatively with the type I receptor in the case of BMP/GDF or else recruits the appropriate type I receptor into the complex as with TGFß/activin. The ligand-bound type II receptor phosphorylates the type I receptor, activating its latent kinase for subsequent downstream signaling via intracellular Smad proteins that usually translocate to the nucleus to affect gene transcription (123). It has also been observed that all of the type I receptors interact with the immunophilin, FKBP12, which could function to block the activation of its latent kinase activity until the GS domain is phosphorylated by the ligand-bound type II receptor kinase (124, 125).
Smads fall into three different classes, receptor-regulated R-Smads, inhibitory Smads, and the common Smad4 (123). The R-Smads2 and -3 are phosphorylated by the TGFß/activin type I receptors ALK4 and -5, whereas the R-Smads1, -5, and -8 are phosphorylated by the BMP/GDF type I receptors ALK2, -3, and -6. These phosphorylated Smads then dimerize with the common Smad4 to form heteromeric complexes that translocate to the nucleus and effect their respective activities by binding to the Smad-responsive DNA element CAGAC, alone with a relatively loose specificity or in a supercomplex with cofactors that can modulate ligand-specific gene expression (120) such as forkhead activin signal transducer (FAST-1) and the cAMP responsive element-binding protein (CREB)-binding protein (CBP)/p300 (126, 127, 128, 129).
B. MIS type II receptor (MISRII)
The MISRII cDNA was first isolated from a rat Sertoli library by a
differential hybridization screen for T-regulated genes but was
subsequently found to be unregulated by T (130). Genomic
and cDNA clones were also isolated from rabbit fetal ovaries
(131), human postnatal testes (115), rat
urogenital ridges (132), and mouse (133) and
human fetal testes (134). The gene for the human MISRII
was localized to 12q12-q13 and contains 11 exons spanning approximately
8 kb, far smaller than other type II receptors in the TGFß family
(115, 135). However, the predicted amino acid sequence of
MISRII indicates that it is a characteristic type II receptor, which
shares 30% overall homology with other TGFß family type II
receptors.
The mRNA corresponding to MISRII was localized to the Müllerian duct by in situ hybridization at the time when the Müllerian duct is undergoing regression and to fetal Sertoli cells and granulosa cells of both embryonic and adult gonads. Sertoli cells in males and granulosa cells in females surround the germ cells and produce MIS; expression of both ligand and receptor in the same cell suggests an autocrine activity in the fetal male gonad and the postnatal male and female gonad. In the postnatal ovary, expression of MISRII is largely restricted to granulosa cells of preantral and small antral follicles (136). Similarly, in the postnatal testis, developmental expression of the MISRII mRNA in gonads studied by Northern blot analysis indicates that the receptor is highly expressed embryonically, during Müllerian duct regression, and postpubertally, with relatively little or no expression perinatally (132). Expression of MISRII is not uniform in the seminiferous tubules and appears to be lowest at stages XII and highest at stage VII when FSH binding is lowest and MIS expression is highest in these structures (137). An unexplored, but nonetheless intriguing, finding is that MISRII is also expressed in the postnatal uterus (97, 132). Expression of MISRII has also been observed in Leydig cells, which respond to MIS by lowering T production both in vitro and in vivo (see Section IV.A).
As in the MISRII knockout mouse, in which the MISRII gene is inactivated by homologous recombination (117), mutations of MISRII were found in patients with PMDS. For example, at the splice donor site of intron 2 (115), a 27-bp deletion in exon 10, the most common mutation found (138, 139), and a number of other mutations are dispersed throughout the coding region (138, 139). Since the homozygous female MISRII knockout mice have normal fertility and fecundity (117), it would appear that MIS is not required for either blastocyst implantation or fetal and embryonic development.
Regulation of MISRII expression has been studied in vitro, but to date the only transcription factor shown to be essential for receptor expression is SF-1 (140, 141). None of the MISRII promoters, rat, mouse, or human, contains canonical TATA or CCAAT boxes, and the rat promoter contains 150-bp inverted Alu-like repeats, the function of which is unclear. The rat MISRII promoter contains two SF-1 sites, each of which could function in the absence of the other site to promote transcription of a luciferase reporter in R2C cells, a rat Leydig cell line that expresses SF-1. Mutation of the proximal, lower-affinity SF-1 site actually enhanced transcription, suggesting that binding of SF-1 to both sites simultaneously might be inhibitory (141). In a teratocarcinoma cell line, NT2/D1, expression of human MISRII was also dependent on SF-1 binding to its single site in the promoter, and increasing the amount of SF-1 in HeLa cells correlated directly with transcriptional activation of the luciferase reporter (140).
Homozygous deletion in mice of Wnt-7a, a member of the Wg/wnt family of secreted proteins that determine cell fate and polarity and early pattern formation, resulted in the loss of MISRII expression in the Müllerian duct (142). The Wnt-7a-deficient mice were infertile because the retained Müllerian ducts blocked sperm passage in males, and the differentiation of oviducts and uteri was abnormal in females. Wnt proteins signal through the frizzled family of receptors and activate ß-catenin, which translocates to the nucleus and interacts with lymphoid enhancer factor 1 (LEF 1), to function as a transcription factor (reviewed in Ref. 143). The molecular mechanisms of Wnt-7a-regulated expression of MISRII is an area worthy of investigation given the observation that ß-catenin might be involved in Müllerian duct regression (22); this suggests that MIS might regulate the expression of its own receptor even though the level of expression of MISRII in both male and female Müllerian duct mesenchyme appears similar (132). In MA-10 cells, MIS appears to inhibit the expression of the MISRII mRNA, but not an MISRII promoter/luciferase reporter (J. Teixeira, unpublished data).
C. Candidate MIS type I receptors (MISRIs)
While a number of type I receptors for the TGFß family have been
cloned, the identity of the specific MISRI has been more difficult to
deduce. The localization of ALK2 mRNA (first known as R1 or ActR1A) to
the mesenchyme surrounding the embryonic Müllerian duct (See Fig. 4) at the time of Müllerian duct
regression made it a likely candidate (144). Recent
reports suggest that either ALK6 (145) and/or ALK2
(146, 147) might be involved in MIS signaling, and results
with ALK3 conditional knockout mice favor it as the candidate MISRI (S.
Jamin and R. Behringer, personal communication). Ligand-dependent
coimmunoprecipitation of the MISRII with ALK6, but not ALK2, was
observed in cells permanently transfected with MISRII, as was
activation of a downstream signaling cascade that induces the
expression of a BMP-responsive Xvent2 reporter gene in
MISRII-transfected P19 cells (145). However, antisense
oligonucleotides of ALK2 coding sequence, but not ALK6 coding
sequence, specifically blocked regression of the
Müllerian duct in organ culture and diminished the activation of
the BMP-responsive Tlx-2 reporter gene in P19 cells
transfected with MISRII and treated with MIS (146).
Tlx-2 reporter expression was also attenuated by
transfection and expression of dominant-negative ALK2, but not
ALK6 (146). Inhibition of Tlx-2 reporter
expression by ALK2 was also observed in P19 cells transfected with
MISRII using a series of truncated mutants of the type I receptors
(147). Smad6 induction was also observed after addition of
MIS to MA-10 cells, which express endogenous MISRII (147).
Colocalization of the ALK2 mRNA with MISRII in the mesenchyme
surrounding the Müllerian duct by in situ
hybridization further supported the identity of the MISRI as ALK2
(146, 147). ALK2 expression in the urogenital ridge, which
has a short window of responsiveness to MIS, is sexually dimorphic by
in situ hybridization, being highly expressed in the male
ridge that is destined for regression but much lower in the female
ridge (147). The diminished expression of ALK2, as well as
Smad1 and -8, in the female urogenital ridge may explain why higher
concentrations of MIS are required for regression of the agonadal
female ridge than that of males in culture (P. K. Donahoe,
unpublished data), which we speculate might provide the female fetus
with some additional protection against exogenous MIS from a male twin.
The strong expression of ALK2 observed in the coelomic epithelium in
these studies might also explain why human ovarian cancers, which
predominantly originate from the coelomic epithelium, might be targets
for MIS (134).
|
All three reports provide evidence that Smad1, -5, and -8, which are known to transduce signals for a number of BMP/GDF ligands (120), also transduce the signals from MIS (145, 146, 147). Smad1 phosphorylation in Sertoli- and Leydig-derived cell lines was observed in response to MIS (145), and a Smad1/Gal4 DNA binding domain fusion construct can activate a Gal4-responsive promoter in P19 cells with added MIS (147). Of note was the observation that, in the mesenchyme surrounding the Müllerian duct, Smad8 expression appears more robust than Smad1, while Smad5 is barely detectable (147). However, in P19 cells, overexpression of Smad5 was able to induce MIS-mediated activation of an ALK2-responsive luciferase reporter, and overexpression of a mutant Smad5 attenuated that response (146). Thus, it is likely that the MIS receptor-regulated Smad will be in the BMP/GDF pathway but that the choice of the Smad1, -5, or -8 may be tissue- or cell-specific. It is important to understand the signal transduction pathway for MIS in tissue-, development-, and tumor-specific contexts that could lead to a small molecule screen to uncover possible leading candidates to replace or enhance the ligand as a therapeutic.
|
IV. MIS in the Adult |
|---|
|
TopAbstractI. IntroductionII. The MIS Ligand...III. MIS ReceptorIV. MIS in the...V. Alternative MIS Signal...VI. Outlook for MISVII. ConclusionReferences |
|---|
A. MIS in the testis
Expression of MISRII has been detected in Sertoli cells,
indicating an autocrine role for MIS, which was first suggested by the
inhibition of aromatase activity by MIS in cultured primary Sertoli
cells (151). Inhibition of aromatase activity by MIS had
also previously been observed in fetal ovary cultures
(152). MIS is also present in seminal fluid of various
species (153, 154, 155, 156, 157) at concentrations that may be
significantly higher than that observed in adult serum. Normal adult
human males had a mean seminal fluid concentration of MIS of 150
pM vs. 11 pM in serum, and
higher seminal fluid concentrations of MIS were associated with
hypospermatogenesis in patients with nonobstructive azoospermia
(157). Although MISRII expression is not observed in round
spermatids (130), MIS can be detected bound to sperm by
immunohistochemistry with MIS-specific antibodies where it might be
involved in sperm motility (155, 158).
The paracrine effects of MIS on Leydig cell development and function
are better understood. Except for a short period after birth in human
males (see Fig. 2), there is a reciprocal relationship between the
concentration of serum MIS and T such that when MIS is at its highest,
T is at its lowest and when MIS falls at puberty, T rises. After
puberty, the expression of MIS is greatly reduced in males to the low
levels observed in females (48, 59, 60, 62). The low level
of T observed in childhood correlates with the paucity of remaining
fetal Leydig cells after birth when this population of cells appears to
die off. Another population of Leydig cells differentiates from
mesenchymal precursors to progenitor Leydig cells before puberty, then
to immature Leydig cells, and finally to adult Leydig cells at puberty.
These mature Leydig cells are the source of T in the adult
(159).
A paracrine role for MIS had been postulated for Leydig cell differentiation and function because of the phenotypes observed in mice that were manipulated to either overexpress MIS or lacked functional MIS or its receptor. Male mice chronically overexpressing MIS had lower levels of T (112) and Leydig cell hypoplasia: in those mice expressing the highest levels of MIS, undervirilized external genitalia were observed (111). Conversely, mice with null mutations in either MIS or MISRII had Leydig cell hyperplasia and other gonadal abnormalities (116, 117).
Subsequent analyses showed that normal Leydig cells express MISRII by
RT-PCR (113) and that differentiation of the Leydig cells
in the MIS-overexpressing mice was blocked, leading to a decrease in
both immature and mature Leydig cell number. Steroidogenic function was
also repressed with decreased testicular mRNAs for the T
biosynthetic pathway enzymes, cytochrome P450, cholesterol
side-chain cleavage (P450SCC), 3ß-hydroxysteroid
dehydrogenase/
5-4-isomerase (3ßHSD),
and most significantly, cytochrome P450C17–20
hydroxylase/lyase (Cyp17), which catalyzes the committed step in
T synthesis (113). In contrast, as one might expect,
analysis of Cyp17 in MIS-deficient mice was higher than normal
(113). Incubation of primary Sertoli cell and Leydig cell
cultures with MIS showed a dramatic effect on aromatase activity and T
secretion, respectively (151). Using a combination of
Northern blot analyses and flow cytometry with highly purified
populations of developmental stage-specific Leydig cells, MISRII
expression was shown in both progenitor and immature cells, but MIS
binding was only observed with progenitor cells, suggesting that MIS
might be involved in the differentiation of the progenitor to immature
Leydig cells (160). MIS was found to inhibit T synthesis
by 10-fold when added to MA-10 cells, a mouse Leydig cell line, and to
do so, in part, at the transcriptional level of Cyp17
(161). A single ip injection of MIS in mice was shown to
lower serum and testicular T by 9-fold and Cyp17 mRNA by 2-fold after
24 h (162). Despite dramatic reduction in T
concentration, there was little affect on 17-hydroxyprogesterone
concentration, which supported that MIS might also be regulating the
lyase activity of Cyp17, but not its hydroxylase activity, at least in
the mouse. It is tempting to speculate that MIS might play a role in
controlling androgen production in syndromes such as precocious puberty
or may be an adjuvant to suppress T synthesis for benign prostatic
hypertrophy, prostate cancer, and polycystic ovarian syndrome.
B. MIS in the ovary
Although MIS is difficult to detect in the serum of normal females
before puberty, MIS expression in the ovaries has been observed as
early as 32 wk gestation in humans (63). In rodents, MIS
expression in the ovary has been detected by postnatal day 3
(163, 164). MISRII has also been detected in theca cells
or maturing follicles by in situ hybridization
(165). The theca cells in the ovary are the T-producing
counterpart of the Leydig cells in females. Expression of MISRII in
those cells suggests that MIS might also regulate the steroidogenic
capacity of theca cells to produce T as it does with Leydig cells.
Although homozygous deletion of the gene for either MIS ligand or its
type II receptor in mice resulted in overtly normal females (116, 117), subsequent analyses showed that the MIS-null mice had more
advanced follicle recruitment for age than wild-type mice, resulting in
an earlier depletion of the primordial follicles (166).
Heterozygous mice had an intermediate number of follicles, suggesting
that follicle recruitment is inversely proportional to MIS
concentration. Complicating this interpretation are the results shown
with follicles isolated from 12-d-old rats, which indicate that
addition of MIS induces preantral follicle growth in culture
(167). Since primordial follicles do not appear to express
MISRII (136), a secondary effect of MIS has been proposed.
These results may be explained by the observation that MIS inhibits
both aromatase activity and LH receptor number in FSH-stimulated
granulosa cells of postnatal rats and pigs, respectively
(168).
Earlier studies have shown that MIS inhibits the first meiotic division of diplotene oocytes in immature rats (110), but this has not been confirmed by others (169). In addition, MIS blocked epidermal growth factor-induced steroidogenesis, and proliferation of human granulosa-luteal cells in vitro (170) and the concentration of MIS in follicular fluid were inversely proportional to mitotic indices of granulosa cells in vivo (171). Thus, MIS appears to have an autocrine, as well as an endocrine, role in maturation of normal follicles.
|
V. Alternative MIS Signal Transduction |
|---|
|
TopAbstractI. IntroductionII. The MIS Ligand...III. MIS ReceptorIV. MIS in the...V. Alternative MIS Signal...VI. Outlook for MISVII. ConclusionReferences |
|---|
B (NFB) pathway (173), as well as the
ß-catenin/lymphoid enhancer factor pathway (22).
Antiproliferative activities of MIS in normal ovarian and ovarian
cancer cell lines, for example, correlated with changes in
cyclin-dependent kinase (CDK)/cyclin, and in normal breast and breast
cancer cell lines, required activation of NFB. As with other
pathways, including cAMP (174), MIS signaling might not
occur in a cell in a simple linear progression through Smads. It is
likely that interaction with other intracellular pathways such as Ras
or Jak/Stat might occur to create a complex combinatorial matrix of
signaling networks that may or may not involve the entire MISRII/RI
receptor complex. While the candidate MISRI has not been investigated
in these settings, studies of the effect of MIS on ovarian and breast
cancers indicate that they at least express the MISRII.
A. MIS-mediated signal transduction in the ovary
In addition to its well established role in the regression of the
Müllerian duct in male embryos, MIS has been shown also to
inhibit the proliferation of several transformed human cell lines
in vitro, including human epithelial ovarian cancer cells
and a cell line derived from the normal ovarian surface epithelium, the
origin of human epithelial ovarian cancers (51, 134, 172, 175, 176, 177) and other tumors of Müllerian duct origin, as
well as ocular melanomas in vivo (51, 52, 54, 178). Although early studies with rhMIS were only minimally
effective against a series of gynecological tumor cell lines in
vitro (179), later studies of its efficacy with
highly purified recombinant material (172) or rhMIS
produced ectopically (180) have been more convincing.
Since female mice that overproduce MIS demonstrate ablation of the
ovary, a role for MIS in the postnatal development of the ovary can be
inferred. Using Northern and Western blot analyses, MISRII has been
detected (172) in both ovarian cancer cell lines and in
HOSE6–3 (human ovarian surface epithelium), an immortalized cell line
derived from laser dissection of the coelomic epithelium covering the
normal ovary (181).
MIS-mediated inhibition of ovarian cancer cell proliferation correlated with interference with cell cycle progression and induction of apoptosis (172). Cell cultures treated with MIS demonstrated a 10–15% increase in the G1 phase fraction compared with untreated cells. Cell cycle progression through the G1 phase of the cell cycle is regulated through a complex mechanism that involves the phosphorylation of the retinoblastoma (Rb) family of pocket proteins. The members of this family (Rb, p107, and p130) are phosphorylated by the cyclin/CDK complexes, the kinase activity of which is negatively regulated by the CDK inhibitors (CDKI). In vitro inhibition of proliferation by MIS of the human ovarian cancer cell line, OVCAR 8, and of the immortalized ovarian surface epithelium cell line, HOSE 6–3, correlated with up-regulation of the CDKI p16 protein (172), a member of the INK4a family of CDKIs, which specifically inhibit the kinase activity of cdk4 and cdk6. Up-regulation of p16 protein expression correlated with increased protein stability and possibly enhanced translation, since p16 mRNA expression remained unchanged. Up-regulation of p16 is necessary for MIS-mediated growth inhibition since expression of antisense p16, which impedes p16 translation, blocks MIS effects on ovarian cancer cells. In addition to atrial natriuretic peptide, which inhibits cell cycle progression in astrocytes through stimulation of p16 mRNA (182), MIS is the only other peptide hormone known to induce the expression of p16. Precedent is further provided by the fact that growth inhibition by TGFß, activin, BMP2, and BMP4 can be mediated through induction of the CDKIs, p15, p21, and p27 in many cells lines (183, 184, 185, 186). The ability of MIS to increase p16 expression selectively indicates that TGFß family members inhibit cell proliferation using distinct molecular pathways.
The most well known mechanism by which CDKIs interfere with cell cycle progression is by preventing the phosphorylation of the Rb family of proteins, which includes Rb, p107, and p130 (187). The lack of Rb protein expression in OVCAR 8 and HOSE 6–3 cells suggests that MIS-mediated inhibition of ovarian epithelial cell proliferation occurs through an Rb-independent pathway, possibly by employing another protein of the Rb family. In support of this latter mechanism, we found that MIS treatment selectively suppressed expression of the p130 protein in OVCAR-8 cells after 4 d of MIS treatment, but did not suppress expression of p107. The delay required to observe p130 suppression reflects an indirect mechanism that may involve decreased transcription, translation, or the stability of the p130 protein. The decrease in p130 protein expression in MIS-treated ovarian cancer cells correlated with an increase in E2F1, a transcriptional regulator that, in addition to facilitating cell cycle progression into the S phase, is a potent inducer of apoptosis. E2F1 null mice develop many tumors, including tumors of the ovary and uterine horn (188), tissues that are known to express high levels of MISRII mRNA (132).
Overexpression of p16 can regulate apoptosis in many cell types. Whereas induction of p16 protein is an early event, the apoptosis observed after treatment of OVCAR 8 with MIS is delayed and probably indirect. Delayed induction of apoptosis after overexpression of p16, p18, and p27 has been demonstrated in A549, HeLa, SKOV-3, and MTA1A2 cells (189). Apoptosis of OVCAR-8 cells may result from the induction of E2F1, the growth-inhibitory effect of which is inhibited by Rb family proteins. Expression of p16 is low or absent in one-third of primary epithelial ovarian cancers due to homozygous deletions, missense mutations, or hypermethylation of the promoter (190, 191, 192, 193). Thus, up-regulation of p16 protein by either gene transfer techniques or MIS treatment could offer therapeutic benefit in the treatment of ovarian cancer patients.
B. MIS-mediated signal transduction in the breast
While the MIS-overexpressing mice indicate that the ovary and
ovarian cancer might be targets for MIS, knockouts and overexpressing
mice give no clear evidence for the breast as a target. MISRII
expression, however, was demonstrated in recent studies in normal
breast, breast fibroadenomas and adenocarcinomas, and breast cancer
cell lines using several different techniques, suggesting that the
breast might be a potential target for the action of MIS
(173). In addition to blocking the growth of ovarian
cancer cells, MIS also inhibited the growth of both ER-positive and
-negative human breast cancer cells in vitro. As with
inhibition of ovarian cancer cell proliferation, MIS-mediated
inhibition of breast cancer cell proliferation resulted from an
increase in the fraction of cells in the G1 phase
of the cell cycle and induction of apoptosis as measured by increased
levels of caspase-3 activity and annexin V staining. MIS treatment of
breast cancer cell lines induced the DNA binding activity of the NFB
family of transcription factors (173) initially shown to
play a key role in mediating inflammatory responses. Abrogation of
MIS-induced growth inhibition of breast cancer cells by expression of
the dominant negative IB demonstrates that activated NFB is
required for this process.
A role for NFB activation in breast epithelial cell growth and
differentiation is becoming increasingly evident. 1) NFB DNA binding
activity is strongly induced during postlactational involution (194), a
period of extensive apoptosis in the breast. 2) NFB p50/p65
heterodimers inhibit ß-casein expression by interfering with
phosphorylation of STAT 5 (195). 3) NFkB DNA binding activity and
expression of transactivators and inhibitors of NFkB are misregulated
in primary breast cancer and cancer cell lines
(196, 197, 198).
MIS selectively up-regulated IEX-1S, the short form of IEX-1, via an
NFB-dependent pathway both in vitro (173)
and in vivo (199). Overexpression of IEX-1S inhibited
growth, suggesting that it might be a putative suppressor of breast
cancer cell proliferation (173). Testing the effects of
MIS on the growth of ER-positive and -negative mammary carcinoma cell
growth in vivo will be important to determine whether MIS
would be of potential therapeutic benefit in the treatment of breast
cancer, thereby expanding the potential tumor targets of MIS as a
therapeutic.
The majority of mammary gland growth and differentiation occurs in the adult animal during puberty and pregnancy. We have found that MISRII expression varies inversely with breast growth. Treatment of adult virginal female mice with MIS results in increased apoptosis of mammary duct epithelium (199), suggesting MIS may play a significant physiological role in controlling involution of the normal breast.
|
VI. Outlook for MIS |
|---|
|
TopAbstractI. IntroductionII. The MIS Ligand...III. MIS ReceptorIV. MIS in the...V. Alternative MIS Signal...VI. Outlook for MISVII. ConclusionReferences |
|---|
). Mutations were almost equally divided
between the ligand and the receptor. The fact that patients with the
retained Müllerian duct phenotype may have gene abnormalities
either in the MIS ligand or the MISRII proves that they are in the same
pathway. Most of the mutations were located in the N-terminal domain of
MIS, which resulted in low or absent levels of MIS in the serum, which
can be detected by ELISA. Patients lacking MIS by ELISA are further
screened by single-stranded conformational polymorphism of the GC-rich
MIS gene to detect the mutated region. After the MISRII was identified,
a mutation in the 5-splice donor site for intron 2 was found that led
to aberrant splicing of the gene product in mRNA prepared from a biopsy
of testicular tissue in a patient with PMDS (115). The
receptor gene is now first studied by PCR for a defect in exon 10 of
the receptor because analysis of 69 families of patients with PMDS
indicated that a 27-bp deletion in exon 10 was the most common defect
observed in patients with a mutated receptor (139).
Sequencing is necessary to detect other defects that can span the
entire 8 kb of the MISRII gene.
|
C. New delivery systems
When one searches for a potential anticancer therapeutic it is
important that the net be spread broadly. In the case of MIS, one can
focus on the efficacious MIS ligand, which we have purified to
homogeneity and are now producing for extensive preclinical testing.
The downstream molecular pathway found in the course of eliciting
apoptosis in these target tissues is also being studied systematically.
Knowledge of this pathway will give clues to potential small molecules
that can be developed to modulate the pathway. These therapeutics could
potentially be more useful than MIS itself if engineered to be orally
active. Equally important is the continued search for target genes that
are activated or inhibited by treatment of target cells with MIS. An
intense study of the transcription factors that regulate those genes,
as well as the transcription factors that activate the expression of
MIS, is justified since they may conceivably be substituted for or used
synergistically with the ligand. Knowledge of the complete pathway
raises the possibility of discovering molecules that can superactivate
the pathway. The ideal candidate would be a nontoxic small molecule,
which can be made to usurp the pathway in a cell- or tumor- specific
manner.
If, however, the therapeutic must be the complex ligand, which has the advantage of receptor specificity, then delivery systems take on increased importance. We have employed the principles of tissue engineering using a biodegradable mesh as a scaffold for the creation of neoorgans (202) made from cells transfected with the MIS gene for continuous production of the MIS protein in vivo. This strategy bypasses the need for pharmaceutical scaleup and production of complex molecules, which, at present, fail to make it into the rate-limiting "incubator queue."
As a proof of principle we have transfected CHO cells with genomic MIS and confirmed that they produce and secrete MIS into the media, where MIS can be quantified by ELISA. After the cells are seeded onto a biodegradable mesh, MIS secretion is first confirmed in vitro, after which the mesh is implanted into immunosuppressed mice. Increasing levels of MIS production can be detected in the serum of the animals. When human ovarian tumors were implanted beneath the renal capsule of these animals, growth of the tumor was significantly suppressed (180). It is the ultimate goal of these experiments to seed a patient’s own cells onto the mesh after in vitro transfection of the cells with the MIS gene. It will be interesting to determine whether an implanted MIS-producing neoorgan can achieve sufficiently high levels to suppress tumor growth in vivo in humans, as was observed in immunosuppressed mice (180). If continuous production is found to down-regulate receptor-mediated effects, then systems employing induction constructs may need to be engineered.
D. Other sources of MIS
If one is to use MIS for therapeutic purposes, an inexpensive and
reliable source must be found. To achieve this, we are investigating
the production of MIS in plants. A novel promoter strategy based on the
plant’s response to pressure or touch has been developed and is being
employed to express rhMIS in tobacco plants. Since MIS is a
glycoprotein, concern whether plant glycosylation will produce a
product that is immunogenic in humans, led us to produce only the
bioactive nonglycosylated C-terminal domain of the MIS gene. We can now
detect secreted C-terminal human MIS from transfected plants by Western
blotting and have observed bioactivity in the Müllerian duct
regression assay (K. Oishi, D. T. MacLaughlin, and P. K.
Donahoe, unpublished data). Production of MIS is being scaled up
and will be tested in a number of other functional assays for MIS
bioactivity.
|
VII. Conclusion |
|---|
|
TopAbstractI. IntroductionII. The MIS Ligand...III. MIS ReceptorIV. MIS in the...V. Alternative MIS Signal...VI. Outlook for MISVII. ConclusionReferences |
|---|
