Recombinant rabbit leukemia inhibitory factor and rabbit embryonic fibroblasts support the derivation and maintenance of rabbit embryonic stem cells

Abstract

The rabbit is a classical experimental animal species. A major limitation in using rabbits for biomedical research is the lack of germ-line-competent rabbit embryonic stem cells (rbESCs). We hypothesized that the use of homologous feeder cells and recombinant rabbit leukemia inhibitory factor (rbLIF) might improve the chance in deriving germ-line-competent rbES cells. In the present study, we established rabbit embryonic fibroblast (REF) feeder layers and synthesized recombinant rbLIF. We derived a total of seven putative rbESC lines, of which two lines (M5 and M23) were from culture Condition I using mouse embryonic fibroblasts (MEFs) as feeders supplemented with human LIF (hLIF) (MEF+hLIF). Another five lines (R4, R9, R15, R21, and R31) were derived from Condition II using REFs as feeder cells supplemented with rbLIF (REF+rbLIF). Similar derivation efficiency was observed between these two conditions (8.7% vs. 10.2%). In a separate experiment with 2×3 factorial design, we examined the effects of feeder cells (MEF vs. REF) and LIFs (mLIF, hLIF vs. rbLIF) on rbESC culture. Both Conditions I and II supported satisfactory rbESC culture, with similar or better population doubling time and colony-forming efficiency than other combinations of feeder cells with LIFs. Rabbit ESCs derived and maintained on both conditions displayed typical ESC characteristics, including ESC pluripotency marker expression (AP, Oct4, Sox2, Nanog, and SSEA4) and gene expression (Oct4, Sox2, Nanog, c-Myc, Klf4, and Dppa5), and the capacity to differentiate into three primary germ layers in vitro. The present work is the first attempt to establish rbESC lines using homologous feeder cells and recombinant rbLIF, by which the rbESCs were derived and maintained normally. These cell lines are unique resources and may facilitate the derivation of germ-line-competent rbESCs.

FIG. 1.

The cDNA and suggested protein sequence of rabbit pLIF7.2b. The complete sequence of rabbit pLIF7.2b cDNA (A, between the brackets) was cloned and sequenced to synthesize functional recombinant rbLIF. The italic and underlined sequences are revealed by our work. The recombinant LIF protein composition suggested by the pLIF7.2b sequence (B, between brackets) was compared with that of hLIF and mLIF, revealing higher homology between rbLIF and hLIF than between mLIF and hLIF. In particular, rbLIF and hLIF are homologous in five (D⁸¹, H¹³⁶, S¹³⁷, V¹⁷⁹, and K¹⁸²) of the six critical residues (B, indicated by arrowheads) for cross-species reactivity of LIF with LIF receptors. The only exception is at S¹³¹, where rbLIF is homologous to mLIF (T), but different from hLIF (S).

FIG. 1.

The cDNA and suggested protein sequence of rabbit pLIF7.2b. The complete sequence of rabbit pLIF7.2b cDNA (A, between the brackets) was cloned and sequenced to synthesize functional recombinant rbLIF. The italic and underlined sequences are revealed by our work. The recombinant LIF protein composition suggested by the pLIF7.2b sequence (B, between brackets) was compared with that of hLIF and mLIF, revealing higher homology between rbLIF and hLIF than between mLIF and hLIF. In particular, rbLIF and hLIF are homologous in five (D⁸¹, H¹³⁶, S¹³⁷, V¹⁷⁹, and K¹⁸²) of the six critical residues (B, indicated by arrowheads) for cross-species reactivity of LIF with LIF receptors. The only exception is at S¹³¹, where rbLIF is homologous to mLIF (T), but different from hLIF (S).

FIG. 2.

Production of recombinant rbLIF. (A) The functional region of rbLIF was introduced into pGEX-6P-1 vector. Colony #3 was selected for the expression of recombinant protein after confirming the size and sequence. (B) The GST–rbLIF fusion protein was identified after western blotting by anti-human LIF antibody at a molecular weight of 46 kDa (upper arrow) in both the total input protein (before purification) and the purified fusion protein (before thrombin cleavage). After thrombin cleavage, the r-rbLIF was found at the similar molecular weight (19 kDa, lower arrow) with human LIF (positive control). Input and fusion proteins did not show signals at 19-kDa size.

FIG. 3.

Evaluation of REF cells and derivation of rbES cells on REF cells. REF cells were treated with mitomycin-C, followed by calcein AM assay to evaluate their viability and BrdU assay to evaluate their proliferation status. In the calcein AM assay, the presence of light fluorescence indicated viable cells (A), which correlated well with the 4′,6-diamidino-2-phenylindole (DAPI) staining of the same population (B), indicating that almost all cells were still viable. In the BrdU assay, non-mitomycin C–treated REF cells consisted of a large population of dividing cells (dark staining) (C), whereas most mitomycin C–treated cells stopped proliferating, showing no dark staining (D). After seeding blastocysts (E) on the mitomycin C–treated REF cells (F), several putative cell lines were successfully derived and maintained (line R4, passage 10, H), displaying a similar ESC morphology with rbESCs derived and maintained on MEF cells (line M5, passage 10, G). Scale bars, 200 μm.

FIG. 4.

Specific ESC markers, gene expression, and karyotyping results of putative rbESCs derived and maintained under Condition II. Putative rbESC lines (R4, R9, R15, R21, and R31) derived and maintained under Condition II (REF+rbLIF) are positive for AP (A, top panel) and other pluripotency markers, including Oct-4, Nanog, Sox2, and SSEA4 (B, lower panels). (A) Representative staining images of R4 (passage 7). RT-PCR assays revealed that these cell lines (R4, passage 7; R9, passage 6; R15, passage 4; R21, passage 5; and R31, passage 4) expressed pluripotency genes such as Oct-4, Sox-2, Nanog, c-Myc, Klf4, and Dppa5 (B). GAPDH was used as the internal control. Karyotyping analysis on line R4 (passage 15) and line M5 (passage 15) revealed 57.5% and 60.0% normal ploidy rates, respectively (C). Both R4 (D and E) and M5 (images not shown) were identified as male cell lines. Scale bars, 200 μm.

FIG. 4.

Specific ESC markers, gene expression, and karyotyping results of putative rbESCs derived and maintained under Condition II. Putative rbESC lines (R4, R9, R15, R21, and R31) derived and maintained under Condition II (REF+rbLIF) are positive for AP (A, top panel) and other pluripotency markers, including Oct-4, Nanog, Sox2, and SSEA4 (B, lower panels). (A) Representative staining images of R4 (passage 7). RT-PCR assays revealed that these cell lines (R4, passage 7; R9, passage 6; R15, passage 4; R21, passage 5; and R31, passage 4) expressed pluripotency genes such as Oct-4, Sox-2, Nanog, c-Myc, Klf4, and Dppa5 (B). GAPDH was used as the internal control. Karyotyping analysis on line R4 (passage 15) and line M5 (passage 15) revealed 57.5% and 60.0% normal ploidy rates, respectively (C). Both R4 (D and E) and M5 (images not shown) were identified as male cell lines. Scale bars, 200 μm.

FIG. 5.

In vitro differentiation of rbESC line R4. Rabbit ESC line R4 (passage 10) was used for EB formation (A, upper panel, D0). When these cells were cultured in specialized nonadherent dishes, we observed EB formation starting at D1. The number and the size of the EBs continued to increase throughout EB culture (A, lower panel, D5). The EBs expressed all three primary germ layer markers (Gata4 for endoderm, Desmin and Brachyury for mesoderm, Nestin and Pax6 for ectoderm), but not the pluripotency gene Oct4 (B). GAPDH was used as the internal control. Scale bars, 200 μm.