Effects of isolation and culture of turkey primary follicular oocytes on morphology and germinal vesicle integrity

ELSEVIER

EFFECTS OF ISOLATION AND CULTURE OF TURKEY PRIMARY FOLLICULAR OOCYTES ON MORPHOLOGY AND GERMINAL VESICLE INTEGRITY M.R. Bakst, la D. Gliedt, 1 V. Akuffo, ~ W. Potts, 2 and S.K. Gupta ~ 1Germjolasm and Gamete Physiology Laboratory and "Statistical Consulting Service Laboratory Agricultural Research Service, U.S. Department of Agriculture Beltsville, MD 20705 USA Received for publication: February 3, 1998 Accepted: A p r i l 30, 1998 ABSTRACT A novel approach to the production of transgenic poultry is to use primary follicular oocytes (PFOs). However, fundamental information regarding the impact of isolation and culture procedures on PFO integrity is lacking. This study describes the isolation and culture of PFOs from mature turkeys and the effects of these procedures on PFO morphology and germinal vesicle (GV) integrity. To isolate PFOs, ovarian cortex was incubated in trypsin-EDTA alone or further incubated in collagenase plus hyaluronidase (CH). About 200 to 500 PFOs, ranging in size from less than 100 #m in diameter to 1,000 #m, were recovered from each ovary. The culture of PFOs less than 100 #m in diameter for 4 h resulted in blebbing of the oolemma followed by extrusion of ooplasm. Primary follicular oocytes 100 to 250 #m in diameter survived culture for 24 h whereas larger PFOs survived for up to 7 d. Those PFOs with intact granulosa cell investments survived longer than those fully or partially denuded of granulosa cells with CH. Co-culture of PFOs (100 to 250 ~m in diameter) on a monolayer of granulosa cells derived from mature, yellow-yolk follicles augmented PFO survival rates. The rate of GV breakdown was not influenced by the isolation or culture of the PFO. These data provide the basis for developing procedures for the in vitro maturation and in vitro fertilization of isolated PFOs. PublishedbyElsevierScienceInc. Key words: turkey, IVM, oocytes, morphology, germinal vesicle

Acknowledgments We thank Wayne Smoot for his expert technical assistance. aCorrespondence and reprint requests.

Theriogenology50:1121-1130.1998 Publishedby ElsevierScienceInc.

0093-691X/98/$0.00 PII S0093-691X(98)00213-1

Theriogenology

1122 INTRODUCTION

The potential for the production of transgenic offspring offers a significant tool for the genetic advancement of livestock and poultry. While some successes have been realized in this technology in livestock (11 ), direct gene insertion into chicken or turkey embryos, especially at the pronucleus and early cleavage stages, is difficult to achieve for several reasons: 1) The megalecithal ovum is quite fragile and difficult to access immediately after ovulation. 2) There is no reliable procedure for inducing multiple ovulations of ova capable of developmental competence. 3) The egg mass is relatively rapidly transported through the oviduct, which results in deposition of albumen, shell membranes, and shell around the ovum within 25 h of ovulation. 4) The rapid proliferation of blastodermal cells results in a 40,000-cell turkey (Bakst, unpublished observation) and 60,000 cell chicken embryo (6) 25 h after ovulation and fertilization. Alternative approaches have been employed to produce transgenic poultry. These include injection of retroviral vectors into the blastoderm (1,13) production of chimeras by the transfer of early blastodermal cells to recipient embryos (10), use of primordial germ cells (16), and the microinjection of exogenous DNA directly into the fertilized germinal disc (9, 14). Regardless of which approach, the rate of germ-line transmission of the exogenous DNA is extremely low. Recent work has demonstrated that immature oocytes isolated from the juvenile mouse ovary can be cultured and grown under conditions which would result in in vitro maturation (IVM; 4). Such oocytes were fertilized in vitro (IVF) and transferred to recipient females. These embryos developed to term and resulted in live pups. A unique approach to the production of transgenic poultry may be through the use of primary follicular oocytes (PFOs). These PFOs offer unique opportunities because several hundred can be isolated from a single ovary, they are small in size (less than 1,000/~m in diameter), and they can be subjected to micromanipulation techniques similar to those used with mammalian oocytes. Furthermore, unlike the voluminous megalecithal ovum, PFOs may be candidates for cryopreservation. Avian germplasm preservation in the form of ova or intact embryos has yet to be accomplished. Alternatively, individual blastodermal cells and primordial germ cells can be successfully cryopreserved and introduced into recipient blastoderms (7, 8, 12). The only previous attempts at avian IVF using PFOs was reported by Harwood and Gyles (5). However, neither Petitte (personal communication), nor our laboratory has not been able to reproduce their results. The objective of the current study was to determine if PFO isolation and culture procedures induce spontaneous germinal vesicle breakdown (GVBD). The integrity of the germinal vesicle (GV) and the morphology of PFOs were also critically evaluated and served as the basis of analysis. MATERIALS AND METHODS Large White turkey hens 26 to 46 wk of age were maintained under standard turkey breeder husbandry conditions with feed and water provided ad libitum. Hens were photostimulated at 28 wk (14h light:10h dark) and were in egg production by

Thefiogenology

1123

30 to 31 wk. Following cervical dislocation the ovary was quickly removed and placed in warm (39 to 40°C) Earle's balanced salt solution (EBSS). Enzymatic Isolation of PFOs After removing yellow and white follicles, thin strips of ovarian cortex were carefully isolated and minced in Ca÷+-Mg +÷ free EBSS. The minced cortex was incubated for 30 to 120 min in a lO0-mm Petri dish with trypsin-EDTA (T/EDTA; Sigma, St. Louis, MO, USA) at 37°C. Liberation of the PFOs was augmented by gentle pipetting up and down every 10 min. Liberated PFOs were rinsed in Dulbecco's phosphate buffered saline modified by addition of 36 mg sodium pyruvate, 50 mg streptomycin sulfate, 100 mg kanamycin monosulphate, 1000 mg glucose, and 133 mg calcium chloride per liter (DMPBS). The DMPBS was also supplemented with 2% chicken serum. To remove the granulosa cell layer, isolated PFOs were incubated with collagenase (1 mg/mL; Worthington) plus hyaluronidase (0.5 mg/mL; Sigma; CH) at 41°C for 30 to 120 rain. Following the incubation in CH, the PFOs were washed in DMPBS. Culture of Granulosa Cells Obtained from Yellow-Yolk Follicles The granulosa cell layer was isolated from yellow follicles greater than 2.5 cm in diameter, washed 3 times in EBSS to remove adhering yolk, and incubated in 10 mL T/EDTA for 30 min at room temperature with constant stirring. The subsequent tissue suspension was dispersed by gentle pipetting and centrifuged at 60 x g for 5 rain. The pellet was resuspended in a mixture of Dulbecco's minimum essential medium/Ham's F-12 (DME/F-12) containing 2% antibiotic-antimycotic (ABAM) solution and centrifuged as above 2 additional times. The dispersed cells were suspended in DME/F-12 with 2% ABAM and 1% chicken serum at pH 7.0. After undissociated cells were removed by passage through a sieve, the trypan blue exclusion procedure revealed more than 85% of the cells were viable. Granulosa cells were cultured in 24-well plates at 41°C under 5% C02 and 95% humidity. The medium was changed every 2 d. The final concentration of antibiotics supplied by ABAM in the medium was 200 IU penicillin, 0.2 mg streptomycin and 0.5/~g amphotericin B per mL. Culture of PFOs Isolated PFOs were categorized by diameter (the widest outside diameter) into < 100 #m, 101 to 250/~m, 251 to 500/~m, and 501 to 1000 #m. They were also further categorized by the presence or absence (denuded) of a granulosa cell layer and whether the GV was visible. The PFOs were then incubated in DME/F-12 containing 2% ABAM and 10% chicken serum in wells with or without granulosa cell monolayers. Incubation conditions were similar to those described above for granulosa cell cultures. Cultures were monitored by using an Olympus IMT-2 inverted microscope equipped with Hoffman modulation contrast optics. Microscopy Primary follicular oocytes were routinely evaluated by stereomicroscopy and their images recorded using a still video printer system (Sony Medical Systems Division, Montvale, NJ). Morphometric analysis was performed with a video measurement system (VIA-150, Boeckeler, Tucson, AZ).

1124

Theriogenology

Fixation, sectioning, and staining of the ovarian cortical tissue were described in detail elsewhere (3). Briefly, specimens were fixed in buffered 2% paraformaldehyde (PF) plus 2.5% glutaraldehyde (GA) for a minimum of 2 h at 4°C. All specimens were then rinsed in 0.1M cacodylate buffer or Millonig's phosphate buffer plus 5% sucrose for 30 min, followed by two 10-min rinses in buffer, then stored in buffer overnight at 4°C. Post-fixation for 30 to 60 rain in 1% osmium tetroxide in 0.1 M cacodylate buffer was followed by two 10-rain washings in 0.1M cacodylate buffer or Millonig's phosphate buffer. After the last rinse, specimens were stored in buffer overnight at 4~C and then dehydrated in ethanol (5 min each in 50, 70, 85 and 95%, and then three 30-sec changes in 100% ethanol). Dehydrated samples were placed in 1:1 mixture of a low viscosity embedding medium and ethanol for 10 min, followed by 2:1 mixture of embedding medium and ethanol for 15 min, and then 100% embedding medium for 15 min. Infiltrating specimens were placed under low vacuum at room temperature for 4 h to remove any air bubbles. The resin was then cured at 68°C for 48 h. One micrometer thick sections were cut on an ultramicrotome with glass knives. Sections were mounted on slides and stained with toluidine blue for examination by light microscopy (LM). Each PFO was sectioned until the central region of the GV was reached. Visualization of the GV Isolated PFOs (with granulosa cell layer) were randomly distributed into 2 groups (minimum 50 oocytes/group). Group 1 PFOs were fixed within 0.5 h of the final isolation step in methanol:acetic acid (MAA) fixative (3:1 ratio, made fresh daily). Group 2 PFOs were transferred into M2 incubation medium within 0.5 h of the final isolation step and cultured as described above for 18 to 22 h before fixation in MAA. After 1 h of fixation, the presence or absence of the GV was determined by stereomicroscopy. Visualization of the GV was enhanced by clearing the PFO in glycerol (1:1, glycerol:MAA) for 60 min before examination. To augment visualization of the GV in viable, unfixed PFOs, polarity of the ooplasm was induced by centrifugation. Thin strips of ovarian cortex were pooled and centrifuged at 14,500 x g at 4°C, for 10 min. Primary follicular oocytes were then isolated as described previously and the polarity was assessed. For histology, specimens were fixed as described above after each step in the centrifugation procedure. Primary follicular oocyte morphology (normal vs abnormal), outside diameter (widest diameter), and the polarity of the GV and dense material in the ooplasm were also evaluated. Polarity was assigned Value 1 when all the dense material in the ooplasm was aggregated to one pole of the PFO. Value 3 was assigned when the dense material was uniformly distributed throughout the ooplasm. Value 2 was assigned when the PFO could not be assigned a value of either 1 or 3.

Theriogenology

1125

Statistical Methods The data consisted of a binary response variable, presence or absence of the GV, and the set of explanatory variables: diameter, polarity, morphology, centrifugation status, and incubation status. The relationship between the response and explanatory variables was modeled with a class probability tree using the CART (classification and regression trees) statistical software module (15). A class probability tree is a nonparametric, computationally-intensive alternative for logistic regression type data (2). A binary tree was constructed by first splitting the oocytes into 2 groups (nodes) corresponding to a split of an explanatory variable. For example, all the PFOs that were centrifuged might be in one node and those not centrifuged in the other node. All possible splits of every explanatory variable were considered; the split that resulted in the greatest homogeneity with respect to the presence or absence of the GV was selected. The same process was then repeated at each node. After construction, a maximal tree is pruned back using cross validation. The terminal nodes give the predicted probability of the presence of the GV. A class probability tree was built separately for the cases with and without the use of a fixative. RESULTS AND DISCUSSION Normal Morphology When viewed by stereomicroscopy, the morphological appearance of fresh, intact PFOs varied with its outside diameter. Those PFOs less than 250/~m in diameter were spherical and possessed a more centrally located GV, which was discernible by stereomicroscopy (Figure 1). Here the distribution of the dense ooplasm, which was predominantly composed of lipid droplets, was polarized. The eccentrically situated GV was generally observed in a zone between the dense and less dense ooplasm. Primary follicular oocytes greater than 300 #m in diameter were spherical to oblong and uniformly dense. Unless highly polarized, the dense material obstructed visualization of the GV. The MAA fixative rendered PFOs less than 300/~m in diameter transparent and a profile of the GV was discernible (Figure 2). In contrast, due to the increased volume of the larger oocytes (greater than 300 #m) the GV was not readily apparent. Initially, the larger fixed PFOs were cracked open using fine glass probes to expose the ooplasm with or without the GV. Centifugation resulted in polarization of the dense material and subsequent visualization of the GV in PFOs less than 350 #m in diameter (Figure 2) but not in larger PFOs. Alternatively, a 1:1 dilution of glycerol and M A A "clears" fixed PFOs as large as 800 #m in diameter (Figure 3) rendering the GV visible. Regardless of diameter, all POFs possessed a clearly distinguishable granulosa cell investment ranging from 8 to 15 #m thick (Figures 1, 2). The histological appearance of PFOs were consistent with the appearance of the PFO in situ, as described by Carlson et al. (3). In general, the Balbiani stage (Stage II) was evident in isolated PFOs less than 200/~m. Lipid droplets arranged in a ring-like configuration in the cortical aspect of PFOs larger than 350 #m were evident Stage IV. Likewise, in PFOs larger than 500/~m, a zonation of organelles and lipid droplets were observed (Figure 4). Therefore, the staging system devised by Carlson et al. (3) to describe the early morphological development of PFOs in the turkey ovary were applicable to the isolated PFOs.

1126

Theriogenology

Figure 1. A light micrograph of numerous isolated PFOs after incubation in T/EDTA. The narrow peripheral outer ring is the granulosa cell layer and the clear smaller round central zone is the intact GV. (Bar = 1,000/~m) Figure 2. Several isolated PFOs from centrifuged ovarian cortex are observed. The smaller PFOs possess polarized dense cytoplasmic material revealing the GV (arrow) in one POF. The dense cytoplasmic material in PFOs 350/~m in diameter (seen here) and greater generally does not polarize. Morphological anomalies were rare in freshly isolated PFOs either before (Figure 1) or after (Figure 2) centrifugation. If observed, anomalies primarily consisted of a loosening and a shedding of the granulosa cell layer or an irregular condensation and segmentation of the ooplasm. This segmentation may resemble a mammalian embryo in the cleavage stages of development. Effects of Time and Duration of Enzymatic Incubation The yield of normal PFOs was highest with incubation in T/EDTA. At 41°C, release of the PFOs from strips of ovarian cortex was observed after 45 rain in T/EDTA. The number of liberated PFOs increased with duration of the incubation and occasionally exceeded 500 PFOs, ranging from less than100 #m to about 1,000 /~m in diameter. After 60 min of incubation, PFOs initially liberated from the ovarian stroma but not removed form the T/EDTA began to show signs of granulosa cell shedding and degeneration (Figure 5). Further incubation in T/EDTA resulted in extrusion of ooplasmic contents, as seen in Figure 6. Thus, 45 to 60 min was optimum time for incubation in T/EDTA. Morphologically normal PFOs were then transferred to enzyme-free medium to minimize the inimical effects of the T/EDTA. For partial or complete removal of the granulosa cell layer, PFOs were further incubated in CH for 30 min at 41°C (Figure 4). Granulosa cells that remained associated with the vitellus after CH incubation lacked an adherent basement membrane. Occasionally, single or clusters of granulosa cells were observed in vacuoles in the cortical ooplasm of the PFOs (Figure 4). It is assumed that the granulosa cells were phagocytized by the vitellus. After 60 rain in CH, blebbing and rupture of the oolemma was observed. Finally, it should also be noted that the total number of POFs released from the ovarian cortical strips tendened to be greatest in hens in egg production for less than 6 wk. We concluded that 45 to 60 rain of incubation in either T/EDTA or CH appeared to be optimum for the release of PFOs

Theriogenology

1127

from the ovarian cortical strips. Furthermore, PFOs larger than 250/~m in diameter were released mostly by T/EDTA whereas PFOs smaller than 250 #m were released mostly by CH. Incubation in Ca+÷-Mg +÷ free solutions of CH liberated PFOs which

Figure 3. A 660 #m diameter (long axis) PFO which had been fixed and cleared in glycerol. The GV (arrow), which would other wise have been masked by the dense ooplasm, is partially discernible. Figure 4. In this histological section, a denuded PFO (about 800 #m) is observed following a final incubation in CH. Note the vacuoles in the cortical ooplasm, one (arrow) containing granulosa cells. The adjacent PFO has retained some of the granulosa cell layer, but the basement membrane is conspicuously absent. Figure 5. Partially denuded PFOs after prolonged incubation in T/EDTA are observed. The granulosa cell layer appears to be peeling away from the vitellus in three PFOs (arrows). The fuzzy appearance of the top PFO, which is about 550/~m in diameter, is due to swelling and disruption of individual granulosa cells. The outlined arrow highlights a GV. Figure 6. As in Fig. 5, partially denuded PFOs after prolonged incubation in T/EDTA are observed with the granulosa cell layer appearing to be peeling away from the vitellus. Note the ballooning of the oolemma (arrowhead) and some partially extruded dense ooplasmic material (outlined arrow). Figure 7. After 24 h culture, at least two abnormal PFOs, about 300 and 600 #m in diameter (arrows), are observed with several morphologically normall PFOs.

1128

Theriogenology

were partially or completely denuded of granulosa cells. Alternatively, the granulosa investment was not affected by incubations in CH solutions which contained Ca ++ and Mg ++ ions. Culture of PFO The capacity to survive in culture varied with the diameter of the PFO. Culturing PFOs less than 100/~m in diameter for 4 h resulted in blebbing of the oolemma and ultimately extrusion of ooplasmic contents. Primary follicular oocytes ranging from 100 to 250 #m in diameter survived in culture for up to 24 h (Figure 6) with or without the granulosa cell investment. Larger PFOs (up to 700 #m) with the granulosa cell investment survived for up to 7 days. Co-culturing of PFOs up to 250/~m in diameter and partially denuded of granulosa cells directly on a monolayer of cultured granulosa cells derived from F1 or F2 yellow-yolk follicles improved PFO survival rates. In addition to the blebbing and extrusion of ooplasmic contents, PFOs were considered abnormal if the dense ooplasm appeared to condense and had an asymmetrical appearance within the vitellus (Figures 6, 7). Analysis of PFO Morphology and GV Integrity Unfixed PFOs less than 381 #m in diameter (Figure 8a; n - 1 0 5 9 ) with abnormal morphology had a 0.82 probability of the visible presence of the GV (n = 580). Such PFOs, with normal morphology and demonstrating polarity (n = 70), had a 0.66 probability of the visible presence of the GV. Nonpolar PFOs (n = 122) had only a 0.34 probability. The PFOs greater than 381 #m in diameter that were polar had a 0.61 (n = 46) probability of the visible presence of the GV. Nonpolar PFOs (n = 241) had only a 0.19 probability. Incubation and centrifuge status had little explanatory power. Alternatively, after fixation (n = 546), polar (n = 222) as well as nonpolar abnormal PFOs (n = 241 ) had a high probability for the visible presence of the GV (0.90 and 0.80, respectively; Figure 8b). This can be explained by the fact that the condensed "polarized" dark ooplasmic material resulted in the GV being more visible in abnormal PFOs. Primary follicular oocyte diameter, whether incubated or centrifuged, had very little explanatory power, and consequently these categories did not appear in the final tree. To summarize, a substantial number of PFOs can be isolated from the ovary of the mature turkey. Because of the large number of PFOs recovered, polar PFOs with a clearly discernible GV can be selected for further culture and/or manipulation. The isolation procedures and the culture of the PFOs do not induce a spontaneous GVBD and have a minimal disruptive effect on PFO morphology. If PFO culture is necessary, we recommend that PFOs greater than 250/~m in diameter and with an intact granulosa cell layer be used because these will survive for at least 7 d. Possibly due to the absence of the perivitelline layer, an investment homologous to the mammalian zona pellucida, removal or even partial disruption of the granulosa cell layer resulted the degradation of the PFO within 24 h. This would present a problem for those interested in determining the role, if any, of granulosa cells in maintaining meiotic arrest.

1129

Theriogenology

A

,dk'~

Diameter< 381

/ Abnormal ~ I ~ ~

/Morpho,ogy/," n=lS2

n=580

n=70 P(GV:I): .66

P(GV=I)= .82

" ~ I k ~

"~

n=122 P(GV=I) = .34

n=46

n = 241 P(GV=I): .19

P(GV=I) = .61

B

/ n=222 P(GV=I)= .90

Yes

Yes . Abn.°r.mal'~o

~bk~

Morpn°l°gy

~

n=241

n=83

P(GV=I): .80

P(GV:I): .53

Figure 8. A class probability tree for unfixed (A) and fixed (B) PFOs is presented. Within each node the sample size (n) and the estimated probability of the presence of the GV (P(GV = 1)) is given. The terminal nodes and final probabilities are presented as rectangles.

Theriogenology

1130 REFERENCES

1. Bosselman RA, Hsu RY, Boggs T, Hu S, Bruszewski Ou S, Kozar L, Martin F, Green C, Jacobson F, Nicholson M, Schulz JA, Semon KM, Rishell W, Stewart RG. Germline transmission of exogenous genes in the chicken. Science 1989;243:533-535. 2. Breiman L, Friedman J, Olshen R, Stone C. Classification and Regression Trees. Pacific Grove CA: Wadsworth, 1984. 3. Carlson JL, Bakst MR, Ottinger MA. Developmental stages of primary oocytes in turkeys. Poult Sci 1996;75:1569-1578. 4. Eppig J J, Schroeder AC. Capacity of mouse oocytes from preantral follicles to undergo embryogenesis and develop into live young after growth, maturation, and fertilization in vitro. Biol Reprod 1989;41:268-276. 5. Harwood DE, Gyles NR. Genetic variation for in vitro fertilization of avian oocytes. Poult Sci 1987;66 (Supplement 1):112. abstr. 6. Kochav S, Ginsburg M, EyaI-Giladi, H. From cleavage to primitive streak formation: a complementary normal table and a new look at the first stages of development of the chick. II. Microscopic anatomy and cell population dynamics. Dev Biol 1980; 296-308. 7. Naito M. Preservation of chick primordial germ cells in liquid nitrogen and subsequent production of viable offspring. J Reprod Fertil 1994;102:321-325. 8. Naito M, Nirasawa K, Oishi T. Preservation of quail blastoderm cells in liquid nitrogen. Brit Poult Sci 1992;33: 449-453. 9. Naito M, Sasaki E, Ohtaki M, Sakurai M. Introduction of exogenous DNA into somatic and germ cells of chickens by microinjection into the germinal disc of fertilized ova. Molec Reprod Dev 1994;37:167-171 10. Petitte JN, Clark ME, Liu G, Verrinder Gibbins AM, Etches RJ. Production of somatic and germline chimeras in the chicken by transfer of early blastodermal cells. Development 1990;108:185-189. 11. Pursel VG, Rexroad CE. Status of research with transgenic farm animals. J Anim Sci 1993; 71 (Suppl. 3): 10-19. 12. Reedy SE, Leibo SP, Clark ME, Etches RJ. Beyond Freezing Semen. In Proc 1st Int Symp Artif Insem Poult 1995;251-261. 13. Salter DW, Smith EJ, Hughes SH, Wright SE, Crittenden LB. Transgenic chickens: Insertion of retroviral genes into the chicken germ line. Virology 1987; 157:236-240. 14. Sang H, Perry MM. Episomal replication of cloned DNA injected into the fertilized ovum of the hen, Gallus domesticus. Molec Reprod Dev 1989;1:98-106. 15. Steinberg D. Colla P. CART: Tree-Structured Non-Parametric Data Analysis. San Diego CA: Salford Systems, 1995. 16. Wentworth BC, Tsai H, Hallett JH, Gonzales, DS, Rajcic-Spasojevic, G. Manipulation of avian primordial germ cells and gonadal differentiation. Poult Sci 1989; 68:999-1010.