Ploidy reductions in murine fusion-derived hepatocytes

Abstract

We previously showed that fusion between hepatocytes lacking a crucial liver enzyme, fumarylacetoacetate hydrolase (FAH), and wild-type blood cells resulted in hepatocyte reprogramming. FAH expression was restored in hybrid hepatocytes and, upon in vivo expansion, ameliorated the effects of FAH deficiency. Here, we show that fusion-derived polyploid hepatocytes can undergo ploidy reductions to generate daughter cells with one-half chromosomal content. Fusion hybrids are, by definition, at least tetraploid. We demonstrate reduction to diploid chromosome content by multiple methods. First, cytogenetic analysis of fusion-derived hepatocytes reveals a population of diploid cells. Secondly, we demonstrate marker segregation using ss-galactosidase and the Y-chromosome. Approximately 2-5% of fusion-derived FAH-positive nodules were negative for one or more markers, as expected during ploidy reduction. Next, using a reporter system in which ss-galactosidase is expressed exclusively in fusion-derived hepatocytes, we identify a subpopulation of diploid cells expressing ss-galactosidase and FAH. Finally, we track marker segregation specifically in fusion-derived hepatocytes with diploid DNA content. Hemizygous markers were lost by >or=50% of Fah-positive cells. Since fusion-derived hepatocytes are minimally tetraploid, the existence of diploid hepatocytes demonstrates that fusion-derived cells can undergo ploidy reduction. Moreover, the high degree of marker loss in diploid daughter cells suggests that chromosomes/markers are lost in a non-random fashion. Thus, we propose that ploidy reductions lead to the generation of genetically diverse daughter cells with about 50% reduction in nuclear content. The generation of such daughter cells increases liver diversity, which may increase the likelihood of oncogenesis.

Figure 1. Hepatocyte-blood fusion generates fusion-derived hepatocytes and associated daughter cells.

(A) Serial transplantation scheme for expansion of fusion-derived hepatocytes. KLS cells from female Fah^+/+ mice were transplanted into lethally irradiated Fah^{− /−} male mice. Following liver repopulation, hepatocytes were isolated from primary hosts and serially transplanted into Fah^{− /−} female mice (F>M>F). Alternatively, male KLS cells from ROSA26 mice (lacZ^Tg/0) were transplanted into female recipients and subsequently serially transplanted into Fah^{− /−} female mice (M>F>F). (B and C) Hepatocytes from serially transplanted mice (F>M>F) were karyotyped (n = 62 karyotypes from 4 transplanted mice). The distribution of Y-chromosome positive hepatocytes is shown (B). Fusion events between hepatocytes and blood cells gave rise to fusion-derived hepatocytes, which, in turn, underwent ploidy reductions (C). A fraction of reduced-ploidy daughter cells polyploidized. Parentheses indicate the percentage of metaphases scored with the indicated chromosome content.

Figure 2. Marker loss in livers repopulated with fusion-derived hepatocytes.

(A–D) Marker analysis was performed on sequential liver sections from serially transplanted mice. The transplantation scheme (F>M>F or M>F>F) is indicated. FAH+ nodules (brown) either co-expressed the Y-chromosome (black dots) (A) or were devoid of Y-chromosome staining (B and C). Double asterisks (**) indicate double positive nodules (FAH+ and Y-chromosome+) whereas single asterisks (*) indicate single positive nodules (FAH+ and Y-chromosome−). (D) ß-gal (blue staining) is expressed in a single FAH+ nodule but is absent from peripheral FAH+ tissue. Double diamonds (◊◊) indicate double positive tissue (FAH+ and ß-gal+) and single diamonds (◊) indicate a single positive nodule (FAH+ and ß-gal−). Scale bars are 200 µm. (E) Percentage of transplanted mice containing FAH+ fusion-derived nodules that lack Y-chromosome or ß-gal is shown.

Figure 3. Diploid hepatocytes express markers of cell fusion.

(A) Transplantation scheme for detection of fusion-derived hepatocytes. Donor bone marrow cells (from female transgenic mice hemizygous for R26R) were transplanted into lethally irradiated male Fah^−/− mice containing the Alb-Cre transgene. Following fusion between donor hematopoietic cells and host hepatocytes, loxP sites were excised by Cre recombinase, leading to ß-gal expression via the ROSA26 promoter. (B and C) Hepatocytes from repopulated mice expressed ß-gal (blue, as seen by X-Gal staining) (B) and FAH (red) (C). Scale bars are 100 µm (n = 5). (D and E) Hepatocyte suspensions were loaded with Hoechst 33342 and FDG. Livers from repopulated mice were comprised of diploid and polyploid hepatocytes (D). FACS analysis shows the ploidy distribution from a representative mouse (n = 8). ß-gal was expressed by a subset of each hepatocyte ploidy population (E). FACS analysis shows FDG expression by diploid, polyploid and unfractionated hepatocytes from a representative liver (n = 5). (F) FAH was expressed by unfractionated hepatocytes as well as FACS-sorted diploid hepatocytes. Data represent percentage (average±SEM) of cells expressing FAH (n = 4).

Figure 4. Fusion markers segregate independently in diploid daughter hepatocytes.

FACS-isolated cell populations from mice repopulated via the scheme in Figure 3A were genotyped by single cell PCR for donor (Fah and R26R) and host (Cre and Y-chromosome) markers. Two primer sets per marker were utilized (i.e., primer sets “a” and “b”). (A) Representative data for splenocytes is shown (n = 100 cells from 4 repopulated mice). Cells 1 and 2 were positive for host markers and cells 3–5 were positive for donor markers. Negative control is marked “neg.” (B) Representative data for diploid hepatocytes is shown (n = 157 cells from 2 repopulated mice). Cell 1 was positive for all markers, whereas cell 7 was positive for only recipient markers. Cells 2 and 3 were negative for the Y-chromosome and Cre transgene, respectively. An example of PCR failure is shown for cell 2 (R26R). Although this cell was negative for R26R-a, it was positive for R26R-b. Therefore, cell 2 was scored “positive” for R26R. Loss of two or more markers was seen in cell 4 (R26R/Y-chromosome), cell 5 (Cre/Y-chromosome) and cell 6 (R26R/Cre/Y-chromosome). (C and D) Fah+ diploid hepatocytes were genotyped (n = 44 cells from two mice). The percentage of single cells that lost R26R, Cre and the Y-chromosome is compared to the calculated PCR failure rate for each marker (C). *, P = 0.006; ** P<0.0001. The presence (shaded boxes) and absence (white boxes) of donor and host markers is shown for each subpopulation (D). The percentage of each subpopulation is indicated.

Figure 5. Potential mechanisms for diploid hepatocyte formation from polyploid fusion-derived hepatocytes.

(A) Cytokinesis without mitosis. A binucleated cell undergoes cytokinesis before entering the next cell cycle. This process would not produce marker loss or aneuploidy. (B) Multiple spindles, followed by multipolar (in this case tetrapolar) mitosis. Extreme aneuploidy would result. (C) Mitosis without S-phase with chromosome pairing. This would ensure proper chromosome segregation and would facilitate distribution of hemizygous markers between daughter cells. (D) Horizontal gene transfer. A diploid cell engulfs a neighboring cell undergoing apoptosis. Single chromosomes and/or chromosome fragments would be incorporated into the nucleus while maintaining a nearly diploid karyotype. The parental cell is shown at left, mitotic spindle(s) (when necessary) in the middle and resulting diploid cells on the right. Open circles represent centromeres. Black circles represent centrosomes. Chromosomes are shown in different colors to indicate their lineage. Host hepatocyte chromosomes are red, and donor chromosomes are blue.