Tissue-engineered follicles produce live, fertile offspring

Abstract

Oocytes grown in vitro are of low quality and yield few live births, thus limiting the ability to store or bank the ova of women wishing to preserve their fertility. We applied tissue engineering principles to the culture of immature mouse follicles by designing an alginate hydrogel matrix to maintain the oocyte's 3- dimensional (3D) architecture and cell-cell interactions in vitro. A 3D culture mimics the in vivo follicle environment, and hydrogel-encapsulated follicles develop mature oocytes within the capacity for fertilization similar to that of oocytes matured in vivo. Embryos derived from cultured oocytes fertilized in vitro and transferred to pseudopregnant female mice were viable, and both male and female offspring were fertile. Our results demonstrate that alginate hydrogel-based 3D in vitro culture of follicles permits normal growth and development of follicles and oocytes. This system creates new opportunities for discovery in follicle biology and establishes a core technology for human egg banks for preservation of fertility.

FIG. 1

Development and differentiation of a representative multilayered secondary follicle for 8 days in alginate scaffold. (A) At day 0, a multilayered secondary follicle with a centrally located immature oocyte and some attached theca cells around was isolated and encapsulated in alginate gel. (B) Granulosa cells clearly proliferated after 4 days culture. (C) Follicle maintained their 3-dimensional structure and formed antrums at day 8 (follicles have been removed from alginate gel). (D) The average size of follicles increased 123% from day 0 to day 8. Bar = 100 µm. Color images available online at www.liebertpub.com/ten.

FIG. 2

(A) After 8 days of culture, follicle displayed an in vivo preovulatory phenotype, spherical shape with a central fluid-filled antral cavity, an oocyte within tightly compacted cumulus cells, layers of mural granulosa cells outside, and (B) an intact outer theca cell layer shown by 3β-hydroxysteroid dehydrogenase stain. (C) Oocytes maintained their meiotic arrest during culture (granulosa cells have been removed), and (D) resumed meiosis and extruded first polar body (white arrow) after exogenous human chorionic gonadotropin stimulation. (E) Metaphase II oocytes can be fertilized normally in vitro (black arrow indicates 2 pronuclei), and (F) resulted in the live birth after being transferred back into the oviduct of pseudopregnant mouse. Bar = 100 µm (A, B), 50 µm (C–E).

FIG. 3

(A) Basal steroid concentrations were measured from the condition media collected every other day (A: androstenedione, E2: estradiol, P: progesterone). Data represented as mean±SD (n = 4). Androstenedione increases represent theca cells maintaining normal physiologic function. Estradiol and progesterone both increased, and the ratio of P to E2 was less than 0.5. (B) Oocytes from in vitro cultured multilayered secondary follicles increased significantly from day 0 to day 8 (n = 30; p <.05). In addition, the final size of in vitro growth oocytes was not significantly different in size from oocytes grown in vivo (n = 30; p =.078). (C) Most fully grown oocytes (n = 99) underwent germinal vesicle breakdown and progressed to metaphase II, although at slightly lower rates than oocytes developed in vivo (n = 93). (D) The fertilization rate, determined by the appearance of 2 pronuclei, was 68.2%± 14.5% for metaphase II oocytes (n = 86) from in vitro culture follicles, and 81.7%±5.0% for metaphase II oocytes (n = 65) from in vivo control. Statistical significance was noted between groups with different letters. DG = degeneration; GV = germinal vesicle.