Reproduction phase-related expression of GnRH-like immunoreactivity in the olfactory receptor neurons,their projections to the olfactory bulb and in the nervus terminalis in the female Indian major carp Cirrhinus mrigala (Ham.)

The reproductive biology of the Indian major carp Cirrhinus mrigala is tightly synchronized with the seasonal changes in the environment. While the ovaries show growth from February through June, the fish spawn in July-August to coincide with the monsoon; thereafter the fish pass into the postspawning and resting phases. We investigated the pattern of GnRH immunoreactivity in the olfactory system at regular intervals extending over a period of 35 months. Although no signal was detected in the olfactory organ of fish collected from April through February following year, distinct GnRH-like immunoreactivity appeared in the fish collected in March. Intense immunoreactivity was noticed in several olfactory receptor neurons (ORNs) and their axonal fibers as they extend over the olfactory nerve, spread in the periphery of the olfactory bulb (OB), and terminate in the glomerular layer. Strong immunoreactivity was seen in some fascicles of the medial olfactory tracts extending from the OB to the telencephalon. Some neurons of the ganglion cells of nervus terminalis showed GnRH immunostaining during March; no immunoreactivity was detected at other times of the year.Plexus of GnRH immunoreactive fibers extending throughout the bulb represented a different component of the olfactory system; the fiber density showed a seasonal pattern that could be related to the status of gonadal maturity. While it was highest in the prespawning phase, significant reduction in the fiber density was noticed in the fish of spawning and the following regressive phases. Taken together the data suggest that the GnRH in the olfactory system of C. mrigala may play a major role in translation of the environmental cues and influence the downstream signals leading to the stimulation of the brain-pituitary-ovary axis.     

Molecular cloning of growth hormone-encoding cDNA of an Indian major carp, Labeo rohita, and its expression in Escherichia coli and zebrafish

The cDNA clones encoding for growth hormone (GH) of an Indian major carp rohu Labeo rohita were isolated from a cDNA library constructed from the poly(A)(+) RNA extracted from the pituitary glands of rohu. Partial GH cDNA of the rohu (3'-end) was amplified by RT-PCR and used as probe to screen the cDNA library. Full-length GH-specific cDNA clones (1180 bp) were isolated and sequenced. The sequence contains 48-bp 5'-noncoding region followed by an ORF of 621 bp and a 3'-noncoding region of 521 bp. The peptide shares about 90% identity with the GH of Cyprinus carpio (Linn.) and >84% identity with GH sequences of other cyprinids. The GH-encoding cDNA of rohu has been cloned into expression vectors and GH protein has been over expressed in Escherichia coli and purified as a soluble protein. The GH cDNA was cloned into a bicistronic vector with EGFP; injection of in vitro transcribed GH-EGFP mRNA into zebrafish embryo has resulted in EGFP expression confirming the cloned GH cDNA is functional in fish and the IRES element could be effectively used in fish for bicistronic expression of foreign genes.     

Seasonal changes in plasma steroid levels in Indian major carp Labeo rohita: influence of homologous pituitary extract on steroid production and development of oocyte maturational competence

Circannual variations in plasma levels of testosterone (T), 17beta-estradiol (E(2)), and 17 alpha, 20 beta-dihydroxy-4-pregnen-3-one (DHP) as well as seasonal fluctuations in ovarian steroid synthetic potential were observed in Indian major carp, Labeo rohita. A study was also conducted to examine the mechanism of the development of gonadotropin-induced maturational competence in oocytes of this fish. The present study recorded the lowest values of plasma E(2) and T in L. rohita during the period from October to January. A mild increase in the levels of these steroids observed in February was followed by their rapid rise reaching peak values in April, when the ovary contained mostly the vitellogenic follicles. In the month of May, as the postvitellogenic follicles predominated in the ovary, there was a decline in plasma concentrations of both T and E(2). Low levels of these steroids in plasma remained until January, except a small elevation detectable during June and July (spawning stage). DHP was not detected in the plasma of this fish collected during the period from August to March. Existence of DHP was first recorded in blood in the month of April (vitellogenic stage) and it quickly reached the peak value in May (postvitellogenic stage), followed by a sudden decline in the month of June. Under stimulation of fish pituitary extract (FPE), as a source of gonadotropin, in vitro production of E(2) and T by the vitellogenic follicles was shown to be highest compared to their production rate in other stages, while the postvitellogenic follicles recorded the highest rate of DHP synthesis. Acquisition of oocyte maturational competence (OMC) was shown to develop either by priming the vitellogenic stage fish with a single dose of FPE or by in vitro addition of FPE in culture. In vitro treatment of trilostane, an inhibitor of 3beta-HSD, blocked the FPE-stimulated steroid production but not the development of OMC. Presence of cycloheximide and actinomycin D in the incubation was shown to inhibit FPE-induced development of OMC, indicating the requirement of de novo protein synthesis for this process.     

Molecular cloning and characterization of a gonadotropin-releasing hormone receptor in the guinea pig, Cavia porcellus

Guinea pig gonadotropin-releasing hormone (gpGnRH) is predicted to have a unique structure among all known forms of GnRH molecule [Endocrinology 138 (1997) 4123] and it is of great interest to determine whether the unique structure of gpGnRH is manifested in the characteristics of the guinea pig GnRH receptor. In the present study, we isolated a full-length cDNA for a GnRH receptor from the pituitary gland of the guinea pig. The putative guinea pig GnRH receptor protein has an amino acid identity of 79-87% with mammalian type I GnRH receptors. The amino acid residues which have been demonstrated to be important for ligand binding and signal transduction were conserved in the guinea pig GnRH receptor. However, there are several specific amino acid substitutions among mammalian type I GnRH receptors. Moreover, though the guinea pig has generally been classified as a rodent, the putative GnRH receptor protein did not have some rodent-specific characteristics. Total IP assays demonstrated that the cloned guinea pig GnRH receptor is a functional GnRH receptor and that it shows different preference of ligand sensitivities from the rat GnRH receptor.     

Ontogeny of adenohypophyseal cells in the pituitary of the American shad (Alosa sapidissima)

The distribution and ontogeny of adenohypophyseal cells have been studied in the pituitary gland of embryos, larvae, and juveniles of the clupeid American shad (Alosa sapidissima) using immunocytochemical techniques. In juvenile specimens, adenohypophysis was composed of rostral pars distalis (RPD), formed by cavities lined by prolactin (PRL), adrenocorticotropic hormone (ACTH), and gonadotropic hormone (GTH) cells; proximal pars distalis (PPD), containing growth hormone (GH), GTH, and putative thyroid stimulating hormone (TSH) cells; and pars intermedia (PI) with somatolactin (SL) and melanophore stimulating hormone (MSH) cells. At 3 days post-fertilization (3 days pre-hatching) the pituitary of embryos consisted of an oval mass of cells, close to the ventral margin of the diencephalon, divided in rostral and caudal regions. At this time PRL and ACTH cells appeared in the rostral region of the adenohypophysis, while SL cells were observed in the caudal region where MSH cells showed reactivity 1 day before hatching. At variance, GH cells showed a weak immunoreactivity in the rostral portion at hatching that increased 2 days latter. GTH cells also showed weak immunoreactivity in the rostral region of the adenohypophysis at hatching time. Two days later GTH cells were located in the rostral and central regions of the adenohypophysis. At hatching, the neurohypophysis was very small and no nerve processes were seen to penetrate the adenohypophysis tissue. After hatching, the pituitary gland elongated and in 7 days old larvae, the RPD showed a small lumen surrounded by a palisade of PRL, ACTH, and GHT cells; the PPD showed GH and GTH cells while the PI contained SL and MSH cells. The adenohypophysis and neural lobe increased in size with development and, in 42 days old larvae, they were similar to those of juvenile specimens.     

Ontogeny and distribution of cholecystokinin-immuno reactive cells in the digestive tract of sharpsnout sea bream, Diplodus puntazzo (Cetti, 1777), during larval development

The appearance and regional distribution of cholecystokinin-immuno reactive cells (CCK-IR) in the developing gut of larval Diplodus puntazzo were studied by means of immunohistochemistry, with the aim of understanding the role of this peptide hormone in the acquisition of digestive capacity. Immunohistochemical reaction showed CCK-IR cells from 10 days after hatching (DAH), near the pyloric sphincter and past the first bend in the midgut, as well as in the hindgut. At 25 DAH CCK-IR cells were scattered throughout the midgut, as well as in the hindgut. Since gastric glands appeared at 30 DAH, CCK-IR cells were most abundant in the anterior midgut, near and including the pyloric caeca, and just afore the ileo-rectal sphincter in the posterior midgut, as well as in the hindgut. In older larvae (39 DAH), CCK-IR cells were mainly distributed in the anterior midgut, including the pyloric caeca, as well as in the hindgut. No CCK-IR cells were detected in the foregut at any stage. The distribution pattern of CCK-IR cells differed from other species which also possess a rotated gut as D. puntazzo. In fact, although cells were abundant in regions where the ingested food is retained, so that they can be stimulated to modulating the release of digestive enzymes, a large number of cells occurred also in the hindgut.     

The presence and distribution of gonadotropin-releasing hormone-liked factor in the central nervous system of the black tiger shrimp, Penaeus monodon

The distribution and presence of gonadotropin-releasing hormone (GnRH) in the central nervous system (CNS) of Penaeus monodon were examined by immunocytochemistry, high performance liquid chromatography (HPLC), and radioimmunoassay (RIA). We demonstrated the existence of octopus (oct)GnRH-liked immunoreactivity (ir-octGnRH) and lamprey (l)GnRH-III-liked immunoreactivity (ir-lGnRH-III) in cell bodies of medium-sized neurons of the anterior part (protocerebrum) of the supraesophageal ganglion (brain). In addition, only the ir-octGnRH was detected in the nerve fibers located in the brain and segmental ganglia (subesophageal, thoracic, and abdominal ganglia). Moreover, some branches of these fibers also innervated the neurons in the middle (deutrocerebrum), posterior (tritocerebrum) brain and segmental ganglia. There was no ir-lGnRH-I and ir-salmon (s)GnRH detected in the shrimp CNS. The results from HPLC and RIA showed ir-GnRH in the CNS using anti-lGnRH-III, but not with anti-mammalian (m)GnRH. The data from immunocytochemistry, HPLC and RIA suggest that ir-GnRH in shrimp may be more similar to octGnRH and lGnRH-III than the other forms. These findings support the hypothesis that GnRH-liked factor(s) may be an ancient peptide that also exists in this decapod crustacean.     

Pheromonal stimulation of spawning release of gametes by gonadotropin releasing hormone in the chiton,Mopalia sp

The chiton Mopalia sp., a mollusc, was exposed to various dilutions of gonadotropin releasing hormone (GnRH) in sea water to determine whether this peptide is capable of acting as a pheromone that could stimulate release of ripe gametes (spawning). Two of the peptides, lamprey GnRH-1 and tunicate GnRH-2, had this action at a higher concentration (1.0 mg/L) but dilutions to 50 microg/L no longer were effective. Three other GnRHs: lamprey GnRH-3, tunicate GnRH-1, and a modified chicken GnRH-2, had no such action under the same test conditions. Since the spawning response could be produced by some GnRHs and not by others, it would appear that some kind of molecular recognition is involved, possibly by specific binding to a receptor. In earlier preliminary experiments tunicate GnRH-2 rapidly stimulated gamete release in a hemichordate, Saccoglossus. Thus it is suggested that GnRHs, in at least some invertebrates, may function as pheromones, serving to stimulate simultaneous spawning of individuals in a population of animals, and in this way assure more successful fertilization in species that must release their gametes into the water in which they live.     

Malaria transmission dynamics in Niono, Mali: the effect of the irrigation systems

The type of water management and drainage system could be a potential reason for variation in malaria transmission in rice cultivation areas. To investigate this we have compared the population dynamics of Anopheles mosquitoes (Diptera, Culicidae) in rice plots with controlled and uncontrolled water depth, i.e. casiers and hors-casiers, respectively in the Office du Niger, Mali. We also compared malaria transmission in areas with mixed and casiers plots. Larval collection was performed fortnightly with the standard WHO dipping technique. Adult Anopheles were collected both by pyrethrum spray and landing catches. During the dry season rice cultivation cycle, the larval density in the hors-casier was significantly higher than in the casier plots. The larval peak in the casier plots was considerably smaller than the one in the hors-casier. During the rainy season, no significant difference was observed between the two plot types. However, larval densities begin to rise approximately one month earlier in the casier then in the hors-casier plots, and continued to increase trough the rice development phases until the grain filling/maturation phase, declining thereafter. In contrast, in the hors-casier rice plots larval density increased throughout the rice development. This difference was not significantly reflected in the adult vector density and man biting rate. However, high relative frequencies of Anopheles funestus, survival and entomological inoculation rates of An. gambiae s.l. were observed in the mixed plot sector.     

Immunohistochemical localization of gonadotropin releasing hormones in the brain and pituitary gland of the Nile perch,Lates niloticus (Teleostei,Centropomidae)

In the present study we investigated the distribution of gonadotropin-releasing hormones (GnRH) in the brain of Lates niloticus and their association with different pituitary cell types using immunohistochemical techniques. We found immunoreactive (ir) chicken GnRH-II (cGnRH-II) and mammalian GnRH (mGnRH) as the main components of the GnRH-ir system within the brain of the Nile perch. The results indicate that mGnRH and cGnRH are localized in different neurons: mGnRH-ir perikaria were observed in the preoptic region particularly in the organum vasculosum laminae terminalis (OVLT) and in the nucleus lateralis tuberis pars posterior (NLTP) of the mediobasal hypothalamus. These cell bodies are located along a continuum of ir-fibers that could be traced from the olfactory nerve to the pituitary. mGnRH-ir fibers were detected in many parts of the brain (olfactory bulbs, ventral telencephalon, hypothalamus, and mesencephalon) and in the pituitary. cGnRH-ir cell bodies are restricted to the optic tract, but few scattered fibers could be detected in different parts of the brain. The pituitary exhibited very few cGnRH-II ir fibers, contrasting with an extensive mGnRH innervation. Moreover, mGnRH-ir fibers were targeting the three areas of the pituitary gland: rostral pars distalis (RPD), proximal pars distalis (PPD), and pars intermedia (PI). Double immunolabeling studies showed GnRH-ir fibers in close proximity with prolactin (PRL)- and adrenocorticotropic hormone (ACTH)-producing cells in the RPD, growth hormone (GH)-producing cells in the PPD, gonadotropins (GTHs)-producing cells in the PPD in the external border of the PI, and with somatolactin (SL)- and alpha-melanocyte stimulating hormone (alpha-MSH)-producing cells in the PI. Our results showed direct morphological evidence for a close association of GnRH-ir fibers with the different adenohypophysial cell types. These results suggest a multiple role of GnRH in the regulation of various pituitary hormones' release.     

Steroid-induced changes in the morphology of GnRH neurons in the male leopard frog, Rana pipiens: correlation with plasma gonadotropin and gonadal size

Previously, we reported that hypothalamic explants isolated from male leopard frogs (Rana pipiens) implanted with 17beta-estradiol (E2), but not 5alpha-dihydrotestosterone (DHT), released significantly higher levels of gonadotropin-releasing hormone (GnRH) in response to a veratridine challenge. In this study, we measured changes in GnRH soma size, circulating luteinizing hormone (LH), and gonadosomatic index (GSI) in response to these two steroid hormones to further assess the impact of these hormones on the hypothalamic-pituitary-gonadal axis. Sexually mature male R. pipiens were implanted with silastic capsules containing cholesterol (Ch; control), E2, or DHT for 20 days. GnRH immunocytochemistry (ICC) revealed that both E2 and DHT significantly enlarged GnRH soma size without affecting the total number of GnRH neurons in the forebrain. The effects of E2 and DHT were specific, since neither hormone altered the soma size of tyrosine hydroxylase neurons in the dorsomedial posterior tuberculum. Circulating LH levels were significantly reduced in animals treated with both steroid hormones, with E2 exerting the most potent inhibitory effect. A significant inverse correlation was observed between the GSI and GnRH soma size in Ch controls, suggesting animals with larger GnRH neurons tended to have smaller gonads. Overall, our results showed that both steroid hormones induced the accumulation of GnRH and ultimately the swelling of the GnRH soma. Further, larger GnRH neurons were associated with smaller gonads and lower circulating levels of LH, suggesting a link between enlarged GnRH neurons and an overall decrease in the reproductive activity of R. pipiens.     

Molecular characterization of the GnRH system in zebrafish (Danio rerio): cloning of chicken GnRH-II,adult brain expression patterns and pituitary content of salmon GnRH and chicken GnRH-II

The zebrafish has proven to be a model system with unparalleled utility in vertebrate genetic and developmental studies. Substantially less attention has been paid to the potential role that zebrafish can play in answering important questions of vertebrate reproductive endocrinology. As an initial step towards exploiting the advantages that the zebrafish model offers, we have characterized their gonadotropin-releasing hormone (GnRH) system at the molecular level. GnRHs comprise a family of highly conserved decapeptide neurohormones widely recognized to orchestrate the hormonal control of reproduction in all vertebrates. We have isolated the gene and cDNA encoding chicken GnRH-II (cGnRH-II) from zebrafish, as well as several kilobases of upstream promoter sequence for this gene. As the gene encoding salmon GnRH (sGnRH) has been previously isolated (Torgersen et al, 2002), this is the second GnRH gene isolated from zebrafish to date. We have localized expression of these two genes in the brains of reproductively mature zebrafish using in situ hybridization. sGnRH is localized to the olfactory bulb-terminal nerve region (OB-TN), the ventral telencephalon-preoptic area (VT-POA) and, as we report here for the first time in any teleost species, the hindbrain. cGnRH-II is expressed exclusively in the midbrain, as has been found in all other jawed vertebrate species examined. Finally, the levels of both GnRH peptides in pituitaries of reproductively mature zebrafish were quantified using specific ELISAs. sGnRH pituitary peptide levels were shown to be 3- to 4-fold higher than cGnRH-II pituitary peptide. The cumulative results of these experiments allow us to conclude that zebrafish express just two forms of GnRH in a site-specific manner within the brain, and that sGnRH is the hypophysiotropic GnRH form. This work lays the foundation for further research into the control of reproduction in zebrafish, such as the functional significance of multiple GnRHs in vertebrates, and the molecular mechanisms controlling tissue-specific GnRH expression.     

Changes of gonadal CB1 cannabinoid receptor mRNA in the gilthead seabream, Sparus aurata, during sex reversal

Two cannabinoid receptor-like genes (CB1-like), named CB1A and CB1B, have been isolated in teleost fish, specifically in the puffer fish, Fugu rubripes. However, information on the physiological roles, such as the control of reproduction and development in fish is still scarce. Therefore, the aim of the present study was to investigate the presence of CB1-like mRNA in the gonads of a marine teleost species, the gilthead seabream, Sparus aurata, a hermaphrodite species in which the gonadal tissues first develop as testes, and then as functional ovary. We isolated an 890 bp fragment (GenBank accession number ); that corresponded to the open reading frame of the teleost CB1 receptor gene, encoding for the central portion of the protein, which was aligned with the other bony fish sequence. Using "in situ" hybridization, CB1-like mRNA was localized in both mature and sex-reversing gonads, and relative changes in CB1-like expression levels were detected through semi-quantitative RT-PCR. In the mature testis and in the testicular part of the sex-reversing gonad, CB1 expression levels were found to be much higher compared to the ovarian portion. This suggests that the CB1 signaling is likely involved in the process of testicular regression of the S. aurata, but its actual role has yet to be determined.     

Nucleotide sequence, tissue expression patterns and phylogenetic analysis of estrogen receptor one mRNA in the Murray rainbowfish (Melanotaenia fluviatilis) (Atheriniformes, Actinopterygii)

Estrogens are steroidal hormones that control many physiological processes in both female and male vertebrates. Like other vertebrates, fish have two distinct estrogen receptors (Esr) subtypes, Esr1 and Esr2a that have been isolated in a number of species, as well as a third subtype, Esr2b. The mRNA encoding the Esr1 was isolated from the female liver of an Australian freshwater fish, the Murray rainbowfish, Melanotaenia fluviatilis. The rainbowfish esr1 cDNA was 2569 bp in length and with an open reading frame to encode a protein of 611 amino acids. Phylogenetic analysis and multiple amino acid sequence alignment indicated close relationship and high similarity with killifish (Fundulus heteroclitus) and gilthead sea bream (Sparus aurata). Expression of rainbowfish esr1 mRNA was abundant in the liver, gonads and intestine of adult female and male rainbowfish. This is the first isolation of the full-length nucleotide sequence of an estrogen receptor from rainbowfish. This sequence provides a valuable molecular tool that can be used in future studies investigating estrogen mechanisms, actions and tissue-specific expression in juvenile and adult rainbowfish.     

A physiological role for androgen actions in the absence of androgen receptor DNA binding activity

We tested the hypothesis that androgens have physiological actions via non-DNA binding-dependent androgen receptor (AR) signaling pathways in males, using our genetically modified mice that express a mutant AR with deletion of the 2nd zinc finger of the DNA binding domain (AR^ΔZF2) that cannot bind DNA. In cultured genital skin fibroblasts, the mutant AR^ΔZF2 has normal ligand binding ability, phosphorylates ERK-1/2 in response to 1 min DHT treatment (blocked by the AR antagonist bicalutamide), but has reduced androgen-dependent nuclear localization compared to wildtype (WT). AR^ΔZF2 males have normal baseline ERK-1/2 phosphorylation, with a 1.5-fold increase in Akt phosphorylation in AR^ΔZF2 muscle vs WT. To identify physiological actions of non-DNA binding-dependent AR signaling, AR^ΔZF2 males were treated for 6 weeks with dihydrotestosterone (DHT). Cortical bone growth was suppressed by DHT in AR^ΔZF2 mice (6% decrease in periosteal and 7% decrease in medullary circumference vs untreated AR^ΔZF2 males). In conclusion, these data suggest that non-DNA binding dependent AR actions suppress cortical bone growth, which may provide a mechanism to fine-tune the response to androgens in bone.     

Molecular and developmental analyses of thyroid hormone receptor function in Xenopus laevis, the African clawed frog

The current review focuses on the molecular mechanisms and developmental roles of thyroid hormone receptors (TRs) in gene regulation and metamorphosis in Xenopus laevis and discusses implications for TR function in vertebrate development and diversity. Questions addressed are: (1) what are the molecular mechanisms of gene regulation by TR, (2) what are the developmental roles of TR in mediating the thyroid hormone (TH) signal, (3) what are the roles of the different TR isoforms, and (4) how do changes in these molecular and developmental mechanisms affect evolution? Even though detailed knowledge of molecular mechanisms of TR-mediated gene regulation is available from in vitro studies, relatively little is known about how TR functions in development in vivo. Studies on TR function during frog metamorphosis are leading the way toward bridging the gap between in vitro and in vivo studies. In particular, a dual function model for the role of TR in metamorphosis has been proposed and investigated. In this model, TRs repress genes allowing tadpole growth in the absence of TH during premetamorphosis and activate genes important for metamorphosis when TH is present. Despite the lack of metamorphosis in most other vertebrates, TR has important functions in development across vertebrates. The underlying molecular mechanisms of TR in gene regulation are conserved through evolution, so other mechanisms involving TH-target genes and TH tissue-sensitivity and dependence underlie differences in role of TR across vertebrates. Continued analysis of molecular and developmental roles of TR in X. laevis will provide the basis for understanding how TR functions in gene regulation in vivo across vertebrates and how TR is involved in the generation of evolutionary diversity.     

Chagas' disease: TCRBV9 over-representation and sequence oligoclonality in the fine specificity of T lymphocytes in target tissues of damage

Using the same mouse strain and two Trypanosoma cruzi sub-populations (CA-I and RA) it is possible to induce pathology in different target tissues: skeletal muscle (CA-I) or sciatic nerve and spinal cord (RA). On the other hand, T cells are directly involved in tissue injury in a strain-dependent way, resembling the abnormalities of chronic Chagas' disease. In the present work, we examined the TCRBV repertoire and the CDR3 sequence polymorphism of T cells infiltrating spinal cord, sciatic nerve and skeletal muscle in chronically infected mice. The TCRBV9 segment was systematically over-represented in the target tissues for each T. cruzi strain: sciatic nerve and spinal cord in RA and skeletal muscle in CA-I-infected mice. The analysis of CDR3 sequence polymorphism in the same tissues showed a high proportion of identical TCRBV9 clones in RA-infected mice: 66.6% of the TCRBV9 clones found in sciatic nerve and spinal cord expressed one out of four major CDR3 rearrangements. Sequence identity was shared among clones from sciatic nerve and spinal cord, tissues that are also damaged by passive transfer of CD8 + TL. Those observations are consistent with an antigen driven T-cell expansion sequestered at the inflammation site and demonstrate -- for the first time -- the presence of an oligoclonal repertoire in the antigen recognition site of over-represented T cells in nervous system tissues in chronic Chagas' disease.