|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Ovary |
Andrus Gerontology Center and Department of Biological Sciences,3 University of Southern California, Los Angeles, California 90089
Department of Endocrinology and Fertility,4 University Medical Centre Utrecht, 3584 CX Utrecht, The Netherlands
The Jones Institute for Reproductive Medicine,5 Eastern Virginia Medical School, Norfolk, Virginia 23507
International Pacific Halibut Commission,6 Seattle, Washington 98145
| ABSTRACT |
|---|
|
|
|---|
aging, oocyte development, ovary, ovulatory cycle, seasonal reproduction
| INTRODUCTION |
|---|
|
|
|---|
Much less is known about de novo oogenesis in adults of nonmammalian vertebrates. Histological studies of two relatively short-lived birds, the domestic quail and chicken, do not indicate de novo oogenesis in young adults, and there is evidence of declining ovulatory function with age [7, 8]. In contrast, oocytes in germinal beds showed extensive thymidine incorporation in sexually mature turtles aged 1 yr, which is a strong indicator of de novo oogenesis [7]. The oldest frogs examined (13 yr, which is near the recorded maximum life span of 15 yr) still produced full-size clutches of ova in response to gonadotropin treatment [9]. These examples suggest that reproductive aging and de novo oogenesis in adults may vary widely by species. Evidence of longer reproductive life span (in both absolute years and relative to total life span) in the more primitive vertebrate groups therefore presents an evolutionary sequence of interest. Fish are particularly interesting for comparative studies of reproductive aging because of their huge range of life spans and reproductive schedules, ranging from the semelparous Pacific salmon, with life spans in the upper range of laboratory rodents [2], to many deep-living Pacific Ocean fish, which can live up to or beyond the upper limit of human life span [2, 1021]. Individual ages in feral fish populations can be assessed accurately by the annual otolith rings, as calibrated by natural radioisotopic decay, e.g., in various rockfish species of the genus Sebastes [10, 11, 14, 15, 18], from the northeast Pacific, and in the orange roughy [19, 20], and warty oreo [21] from the southwest Pacific. The genus Sebastes includes, where the populations have not been exploited, many species in which individual life spans far exceed those of most mammals. For example, specimens of S. aleutianus have lived up to at least 205 yr [12, 13, 18], which greatly exceeds the human 122-yr record life span of Jeanne Calment, who lived 70 yr after menopause [22].
The Sebastes we examine in this paper are very long-lived teleosts, and their reproduction in relation to age has not been detailed previously. In many teleosts, the ovary is considered to generate a new crop of oocytes each year though continuing de novo oogenesis in adults [2128]. Seasonally recurrent oogenesis is well studied in trout (Salmo), which mature much earlier and have a shorter average life span than Sebastes [26, 27]. In contrast, lampreys (Agnatha) are thought to complete oogenesis before their only season of spawning [2, 24]. Other iteroparous fish may also lack de novo oogenesis as adults, according to limited histological data [24, 25].
For further insights into the diversity of reproductive aging manifested in mammals, it is essential to evaluate ovaries of Sebastes at advanced ages. As a first approach, we examined oocyte and follicle size classes in samples of S. aleutianus and S. alutus of determined ages. We describe the size distribution of oocytes and follicles present in the ovaries of mature fish of different ages sampled simultaneously. The specimens were divided into three age groups in which the oldest (>50 yr) represent advanced ages, as expected from previous studies [12].
These deep-caught specimens died during removal to the surface because, like most other physoclistous fish, Sebastes cannot equilibrate its swim bladder under decompression from depths. Thus, it is impossible to trace the sizes of a developmental series from primary oocyte to mature follicle or to describe the ontogenetic progression of oocyte or follicle sizes for individual fish of these species as they age. Whereas the available fish differed according to the season of sampling for our study (see Methods), both species have the same seasonal cycles of ovarian oocyte maturation and follicle growth, with the maximum follicular development in OctoberNovember, followed by spawning in FebruaryMarch [28]. Spawning is followed by a brief period of ovarian recovery (AprilMay) during which any unspawned larvae or atretic follicles are resorbed, and which precedes the development of a new crop of oocytes beginning in June. Seasonal timing is important with regard to the interpretation of the characteristics of oocytes and follicles examined in this first report of age-specific reproductive development in long-lived fish species.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Sebastes aleutianus (rougheye rockfish) and S. alutus (Pacific ocean perch) were collected by bottom trawls offshore of British Columbia, Canada, from a Canadian Department of Fisheries and Oceans research vessel. All fish were treated in accordance with approved departmental policies on the care and treatment of animals during scientific research. These policies conform to both Canadian National Research Council and U.S. National Academy of Sciences policies on animal care. Animals were killed as rapidly as possible to minimize suffering.
Both Sebastes species have the same annual reproductive cycle, with spring spawning. However, it was not feasible to collect both species in the same season. Sebastes aleutianus was collected off the west coast of the Queen Charlotte Islands from a single trawl haul at depths of 341 360 m in June 1995. Sebastes alutus was collected from single hauls off the west coast of the Queen Charlotte Islands, Vancouver Island, and southern Hecate Strait at 260357 m in November 1985. Note that these different seasons of collection represent different stages of the annual cycle of oocyte and follicle development (dates are given because of possible historical trends). S. aleutianus was sampled several months after spawning, whereas S. alutus was sampled before spawning, when preovulatory follicles are at maximum development.
The age of each specimen was estimated by counting the annular growth zones on cross sections of the sagittal otoliths, a method that has been validated by tetracycline marking and by natural radioisotopic ratios [10, 11]. Otoliths were sectioned transversely through the center, the transverse face burned over an alcohol flame, and swabbed with vegetable oil to enhance the contrast between the seasonal growth zones (summer is translucent; winter, opaque) [11, 12].
Specimens were classified into three age groups: young (030 yr), middle (3149 yr), and old (
50 yr) (Table 1). The youngest age group represented the age of sexual maturity to 30 yr. For S. alutus, this grouping included ages from 930 yr; for S. aleutianus, ages from 1930 yr. The youngest age range represented for S. alutus is lower than for S. aleutianus because the latter matures at a later age and we include only mature fish in the analysis [28]. The maximum ages in these samples were 77 yr for S. alutus and 80 yr for S. aleutianus.
|
Histology
Histological preparations differed as appropriate for the different oocyte/follicle maturational stages at the time each species was sampled. Ovaries of S. alutus were fixed first in Smith formal dichromate and then transferred to buffered formalin for storage. The initial fixation method was chosen to achieve optimal fixation on the larger follicles in these samples and not distort measurements for comparison with the S. aleutianus samples. Sebastes aleutianus ovaries, at an earlier stage, were fixed in buffered formalin. Samples from each species were embedded in paraffin wax and sectioned at 6 µm. Samples were taken from 610 locations across each ovary to obtain representative data of follicle profiles. For both species, sections were stained with Harris hematoxylin, counterstained with alcoholic eosin, and mounted on slides for examination. Ovaries of S. alutus were analyzed by B.M.L.; S. aleutianus specimens were analyzed using an image analyzer by J.-P. de B. and R.G.G.
For examination of S. alutus ovaries, we used the primary rectilinear axes of the slides as a common standard, and all mature follicles and immature oocytes were counted along the axes. The diameters of 20 each mature follicles and immature oocytes along the axes, which were sectioned through the nucleus, were measured using a calibrated ocular micrometer on a Leitz compound microscope (Leica Microsystems, Wetzlar, Germany). The seasonal cycle of oocyte and follicular development has been described previously in detail for this genus [16, 29, 30]. and we follow the characterizations and classification scheme of these earlier studies.
For S. aleutianus, 10 random microscopic fields were chosen for image analysis using the IBAS system (Kontron Elektronik, Dusseldorf, Germany). This method was chosen over standard ocular micrometer measurements to obtain sufficient precision with the generally smaller size of follicles in the S. aleutianus samples. The slides were coded to avoid observer bias. The images obtained by a Zeiss photomicroscope (Carl Zeiss, Oberkochen, Germany) at 50x magnification were captured on a 512 x 512 frame. All identifiable follicles and oocytes were selected on the basis of their contrast and morphology. The numbers and diameters deduced from area measurements were estimated for both follicles and oocytes and assigned to a database.
Statistics
Descriptive statistics and analysis of oocyte and follicle distributions were conducted with standard analytical software (SYSTAT v10.0.; SPSS Inc., Chicago, IL). Differences in follicle and oocyte frequency among ages and age groups were examined with one-way ANOVA. The same software was used for linear and nonlinear regressions, fitted to data sets using distance-weighted least squares algorithms, to examine follicle and oocyte relationships with age.
| RESULTS |
|---|
|
|
|---|
50 yr). S. aleutianus (Rougheye Rockfish)
Most follicles and oocytes were in smaller size classes (
75 µm) (Figs. 1 and 2), consistent with the time of year and prior spawning (see Methods). Age up to 80 yr did not significantly alter the percentage distribution of either oocytes or follicles (Fig. 2, A and B). There was no significant correlation between the density of oocytes or follicles (number per 2-mm optical field) and age (R2 = 0.08 for oocytes and 0.13 for follicles) (Tables 1 and 2). Nor was there any histological indication of follicular atresia or ovarian pathology. Atresia would be evidenced by hypertrophy of the follicular granulosa as well as dissolution and resorption of the follicle lipids.
|
|
|
S. alutus (Pacific Ocean Perch)
The larger sample size for S. alutus (248 fish) permitted more detailed analysis of potential age effects on ovarian characteristics (Table 3 and Figure 3). Regression of mean follicle abundance and diameter with age indicated significant relationships (P < 0.0005 for both). Similarly, mean oocyte frequency was significantly related to age (P < 0.0005). However, the mean diameter of oocytes was not significantly dependent on age (P = 0.1308). We examined age effects further by looking at the relationship of oocyte and follicle size classes with grouped ages. The oldest age group (
50 yr) lacked the largest and smallest size classes of oocytes (4060 µm, >80 µm), as well as the smallest size class of follicles (<350 µm) (Figs. 1 and 4, A and B). For the two younger age groups, the mean frequency of these largest oocytes was greatest in the group aged 3150 yr, while the youngest group had the highest frequency of the smaller two oocyte size classes. Analysis of variance of the relationship of size classes of oocytes and follicles using age groups as the independent variable also revealed significant positive relationships for frequencies but not mean sizes (Table 3). Follicular atresia was detected in about 0.1% of all follicles examined (53/51 000) [16].
|
|
|
The relationship between both round weight (body weight with viscera included) and ovary weight with age in S. alutus is nonlinear (Fig. 5). Ovarian weight showed an inverted-U distribution with age, such that younger and older fish had smaller ovarian weight than middle-aged fish (Fig. 5A). This relationship resembles, but is slightly more accentuated, than the relationship of age and body length (Fig. 5C) and round weight (body weight including the viscera) (Fig. 5B). Thus, the older fish were slightly smaller, consistent with smaller ovarian mass. Fecundity (number of ova) showed an inverted-U distribution with age (Fig. 5D), similar to that for length and age. Whereas fecundity as a function of weight was essentially linear, it showed marginally greater R2 values when described by a power function (0.431 vs. 0.462, respectively).
|
| DISCUSSION |
|---|
|
|
|---|
The most striking finding in this study was that the oldest fish in both species had active ovaries with follicular and oocyte enlargement. No female mammal is known to retain fecundity beyond 70 yr because the fixed store of ovarian follicles becomes depleted [14]. In S. aleutianus, the normal range of oocyte and follicle size classes was found in all adult ages. All ages had small-sized follicles and oocytes
75 µm (Fig. 1), as expected from the summer time of collection when the next crop of eggs is still growing. However, in the larger sample of S. alutus, the older fish conspicuously lacked the largest and smallest oocyte size class and also the smallest follicle size class. It is believed that this apparent deficit is due to the schedule of sampling relative to seasonal ovarian development, as discussed below.
Follicular atresia was uncommon in S. alutus (about 0.1% of follicles). Although no atresia was found in S. aleutianus, the smaller sample size does not allow firm conclusions to be drawn. Because follicular atresia is observed in other fish, e.g., rainbow trout [31] and electric ray [32], S. aleutianus might also show atresia with larger samples but the relatively low level of atresia in S. alutus, with adequate sampling, suggests atresia may be uncommon in these species. In mammals, the incidence of atresia in antral follicles can be reduced experimentally by unilateral ovariectomy or by administration of FSH, which increase the numbers of mature oocytes produced by the remaining ovary [33, 34]. In rainbow trout, unilateral ovariectomy also increased the crop of mature oocytes [27], but it is unclear whether this was due to decreased atresia or increased de novo oogenesis.
The present data cannot resolve whether the apparent ovarian deficit in the largest size group of oocytes in S. alutus is due to senescence. An alternate and more logical explanation than senescence is that the effect can be accounted for simply by the stage of ovarian development of the fish during November, when the specimens were collected. This period is late in the development cycle, immediately prior to ovulation. Because only a single batch of follicles is ripened per year and older/larger fish in this genus and others mature earlier in a given year [3537], it is probable that all follicles had passed beyond this smallest follicle size class at the time of sampling for these older fish. Because older fish begin to mature oocytes and follicles earlier than young ones, the absence of both the largest oocytes and smallest follicles in the oldest fish is consistent with a relatively later stage of development. We note that the frequency of the middle size class of oocytes in the oldest age group is not significantly different from that in the middle age group. Because only a single follicle crop is produced each year, it is not possible for the later stages of follicle development to be present had the requisite larger oocyte size class not been present at an earlier seasonal period. Resolution of this question could be achieved with serial sampling during the annual reproductive cycle in fish of different ages; however, this process is not possible for the species reported here because of the barotrauma associated with capture from great depths, which causes death from swim bladder rupture.
Rainbow trout also show substantial individual variations in oocyte diameters when sampled 34 mo before spawning [26]. Further sampling of different individuals of S. alutus during spring spawning may resolve whether the absence of the largest oocytes observed here is merely due to a seasonal effect or to ovarian senescence.
Other reports indicate that at least some species of poikilothermic vertebrates retain a capacity for oogenesis at advanced ages, in contrast with virtually all mammals (see Introduction). Feral turtles provide further examples. In longitudinal studies of Blanding turtle [38, 39], the fecundity at 60 yr is at least as high as at 2030 yr; remarkably, the oldest turtles continued to produce viable eggs, as judged by the number hatched and the body size at hatching [38]. Two other turtle species also retain full egg production capacity at advanced ages [40, 41]. Measurement of hormones during the reproductive season can be informative about fertility potential [42].
For those vertebrates with reproduction extending into advanced decades, demographic data are insufficient to evaluate whether mortality rates accelerate at later ages, which is a criterion for senescence at the population level [2, 4346]. Assessment of senescent mortality is also complicated by the large differences in abundance of individual cohorts in these species, where 20-fold differences in cohort strengths are common. The presence or absence of older individuals of specific ages is therefore dependent on the relative strength of the cohorts present at the time of sampling. The oldest S. alutus tend to be smaller (round and ovarian weights, and length) in association with an inverse relation of survival and growth rates [15]. We do not know if genetic diversity within the populations contributes to these different growth schedules. In the guppy, a freshwater fish, genetic influences on the age and size at maturation and fecundity in natural populations arise from size-class predation [47]. Many species of Sebastes share similar mean follicle sizes across a wide range of ages, which suggests strong selection for optimum packaging of oocytes/ follicles in this genus. However, it is also significant that older S. alutus, despite being shorter and lighter than their middle-aged counterparts, have higher proportions of larger follicles. Previous studies of this species have shown that these older but smaller individuals also have slightly higher egg dry weight than their larger, middle-aged counterparts [16]. This relationship implies an increased probability of reproductive success at advanced ages (assuming gamete quality is equivalent), which is contrary to the trends for reproductive senescence in mammals [24].
These data should encourage further study of how some species achieve great longevity with minimal, or even negligible, reproductive senescence. Direct experimental observations on de novo oogenesis in Sebastes may be feasible through thymidine labeling of adult ovaries in short-term organ culture; however, it will be more difficult to evaluate the viability of eggs at advanced ages.
Investigations of the cytogenetics and molecular development of follicles in Sebastes species could give insights on the causes of aneuploidy and follicular atresia in mammalian ovaries. It has long been suspected that cumulative damage in long-lived oocytes is responsible for the age-related loss of oocyte quality, for the decline of fertility, and for increased Down syndrome with maternal age in our species. In contrast to Sebastes, formation of new oocytes in mammals has been thought to be confined to prenatal ages, although this doctrine has recently been challenged by new data [6], and a lively debate is ensuing [4850]. In consequence, ovarian biology of lower vertebrates may take on new significance if germline stem cells are present and active in postnatal mammalian ovaries. Whereas laboratory study of Sebastes spp. is impractical due to death from swim bladder rupture, nonetheless, the gonadal tissue from these species should be viable for study of oogenesis and meiosis in vitro [48, 49]. Comparative biology will then have to address the question of why human ovaries are peculiarly vulnerable to age changes, both in respect of the depletion of oocytes by midlife and the early deterioration of oocyte quality and rising incidence of aneuploidy [51].
| FOOTNOTES |
|---|
2 Correspondence: Caleb E. Finch, Andrus Gerontology Center, 3715 McClintock Ave., University of Southern California, Los Angeles, CA 90089. FAX: 213 740 0853; cefinch{at}usc.edu ![]()
Received: 5 August 2003.
First decision: 21 August 2003.
Accepted: 28 April 2004.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
F Mathews, P Johnson, and A Neil Reply to Comment by Young et al. Proc R Soc B, April 7, 2009; 276(1660): 1213 - 1214. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |