Mice with LoxP sites inserted flanking exons 6 and 7 of mouse Orc2 were purchased from Cyagen Biosciences Inc (Figure 1A). Genotyping with primers F2 and R2 distinguished the loxP-marked allele from the WT allele (Figure 1B). In a cross of Orc2^f/+ mice, there was no significant decrease in the yield of Orc2^f/+ or Orc2^f/f compared to Orc2^+/+ mice (Figure 1C), suggesting that the insertion of the loxP sites in the introns of Orc2 did not impair the function of ORC2. Recombination between the loxP sites would delete the coding exons 6 and 7 of Orc2, which removes amino acids L111-V150 (Figure 1D). If a transcript is expressed that skips exons 6 and 7, then the resulting protein mutates A110 to V110 and then throws the protein-coding sequence out of frame (Figure 1D, ΔORC2), so that the bulk of the 576 amino acid ORC2 protein, including the AAA + like and WH domains (K230-A576) are not expressed. The AAA + like domain and the WH domain are key elements for ORC2 assembly into ORC and for ORC function. Even in the remote possibility that a truncated protein is expressed due to alternative splicing and translation initiation from an internal methionine, the next methionine is at M301, so that half of the protein including 70 amino acids of the 239 amino acid AAA+-like domain (K230-G469) would be deleted.

The Sox2-Cre allele expresses active Cre during embryonic development. Crossing the Orc2^f/f mice with Sox2-Cre mice resulted in no Orc2^Δ/Δ embryos at E7.5 days onwards suggesting that embryonic deletion of Orc2 is lethal (Table 1). The near expected number of Orc2^+/Δ embryos suggests that hemizygosity of Orc2 can still support viability.

Mouse embryo fibroblasts (MEF) were obtained from Orc2^f/f E13.5 day embryos and cultured in vitro. Three days after infection with an Adenovirus expressing Cre, effective recombination between the loxP sites is seen, resulting in a genotype showing that exons 6 and 7 have been deleted in most of the MEF (Orc2^Δ) (Figure 1E). Consistent with this, at 5 d after adeno-Cre infection, the ORC2 protein is not detected in the MEF population (Figure 1F). The dilution of the undeleted MEF lysate on the same blot suggests, that if any ORC2 protein is expressed in the MEF after adeno-Cre infection, it is <10% of that in the undeleted MEF. The proliferation rate of the MEF were measured by MTT assays. Even in the Orc2^+/+ MEF, the infection with adeno-Cre decreased proliferation a little (the orange line compared to the blue line in Figure 1G). However, for Orc2^f/f MEF infection with adeno-Cre impairs proliferation even further (yellow line compared to black line in Figure 1G). Furthermore, when the MEF were cultured for 15 d, the surviving cells express significant levels of ORC2 (Figure 1H), suggesting that the MEFs that had not undergone the Cre-mediated deletion take over the culture.

Taken together, these results suggest that Cre-mediated deletion of exons 6 and 7 of Orc2 leads to early embryonic lethality and impairs the proliferation of normal diploid MEFs in culture.

Mice carrying one copy of the Albumin promoter-driven Cre gene (Alb-Cre) express the Cre recombinase specifically in hepatocytes (Postic et al., 1999). When the Orc2^f/f mice are crossed with mice carrying Alb-Cre (shortened below as Alb), the Cre recombinase is expected to promote the recombination-mediated deletion of exons 6 and 7 of Orc2 in the hepatocytes. The Alb^+/--Orc2^f/f mice were viable and fertile (Figure 2A and B). Crossing such mice yielded the expected numbers of Alb^+/--Orc2^f/f mice, though there was a drop in the yield of Alb^+/+-Orc2^f/f mice. One allele of Alb-Cre is sufficient to express enough Cre to carry out homozygous deletion of the Orc2 allele in the hepatocytes, so that the results suggest that deletion of Orc2 in the hepatocytes did not impair viability. The partial lethality of the Alb^+/+ (HOM) mice could be due to the toxicity of high dose of Cre recombinase expressed, something that has been noted by other groups (Schmidt et al., 2000; Loonstra et al., 2001; Schmidt-Supprian and Rajewsky, 2007; Naiche and Papaioannou, 2007; Janbandhu and Moik, 2014). Because the Alb-Cre is expressed only in the hepatocytes and not in other cells in the liver, we isolated hepatocytes from these livers by growing them in culture to determine whether the ORC2 protein was decreased (Figure 2C). The ORC2 protein was significantly decreased in all five of the Alb^+/--Orc2^f/f mice, showing consistent and near 100% effect of the deletion in the hepatocytes. Western blots for several other ORC subunits, and CDC6 protein showed that the loss of ORC in the same hepatocytes did not significantly decrease these proteins, though MCM2 and MCM3 (two subunits of the MCM2-7 helicase component) were decreased by 50% (Figure 2D). The body weights of the Alb, Orc2^f/f mice were smaller than in the Orc2^f/f animals of both sexes, though the liver size and liver size normalized to body weight was significantly smaller only in the females (Figure 2E–G). There was some elevation of the circulating liver enzymes in the mice where Alb-Cre is expressed to delete the Orc2 gene, suggesting that there is some impairment of liver function (Figure 2—figure supplement 1).

The livers of mice can grow by proliferation (increasing the number of cells) or by hypertrophy of cells (larger cells with endoreduplicated nuclei). Histological examination of the livers revealed that the Orc2 deletion was accompanied by the presence of fewer (~50% of WT levels), but larger nuclei and cells (Figure 2H and I). The Orc2 deleted hepatocytes had significantly larger nuclei in both males and females (Figure 2J and K).

To determine whether the larger nuclei were generated by endoreduplication, we isolated nuclei from hepatocytes and stained them with the DNA staining dye, DRAQ5, followed by flow cytometry (Figure 3A). Both male and female Alb-Cre, Orc2^f/f hepatocytes showed a decrease in 4 n DNA content and an increase in 16 n DNA content (Figure 3B-D). The expression of Cre would also delete a lox-stop-lox element upstream from EYFP and lead to the expression of EYFP. Indeed, when we gated on EYFP positive cells, the 8 n and 16 N DNA content increase was much more evident relative to the EYFP negative cells, while the 2 N DNA-containing nuclei were decreased (Figure 3E-G). Thus Orc2 deletion promotes endo-reduplication, a result similar to what was noted when Orc1 was deleted in mouse livers (Okano-Uchida et al., 2018).

Primary hepatocytes isolated from 8 to 10-wk-old mouse liver can replicate their DNA and proliferate for a limited time in vitro. We isolated such hepatocytes from Alb-cre^-/- (WT) and Alb-cre^+/-, both from Orc2^f/f mice (Figure 4A). Genotyping reveals that over 90% of the cells from the Alb-cre^+/- have successfully deleted Orc2 exons 6 and 7 (Figure 4B), while immunoblotting shows that ORC2 protein expression is also significantly decreased in these cells (Figure 3C). EdU labeling in vitro for 2 hr, showed that the EYFP positive cells (where Cre recombinase has been active) can incorporate EdU (Figure 4C), although the total number of nuclei that incorporate EdU is decreased to about 30% of that seen in the ORC2 wild-type hepatocytes (Figure 4D). Thus hepatocytes in culture can continue to replicate DNA in the absence of detectable ORC2 protein.

Liver DNA synthesis peaks around 36–48 hr, and the liver regenerates to nearly 50% of its original weight within 2–3 d following partial hepatectomy. Such regeneration involves both normal mitotic DNA replication/cell division and endoreduplication accompanied by hypertrophy of the cells. We therefore tested whether deletion of the Orc2 gene in the hepatocytes adversely affects liver regeneration after partial hepatectomy in 8–14-wk-old mice (Figure 5A). Nearly two-thirds of the liver is removed surgically and the mice allowed to recover for 36–48 hr before harvesting the regenerated liver.

Although the liver weight after regeneration was smaller than that of WT livers, the liver:body weight ratio was similar in the livers with Orc2 deletion and Orc2 WT (Figure 5B-D). The H&E stain of the regenerated liver shows that the liver cells are larger and have larger nuclei in the liver with Orc2 deletion (Figure 5E, similar to Figure 2). In these experiments, both before and after regeneration, the livers with the Orc2 deletion have ~50% nuclei compared to the wild-type livers (Figure 5F). Finally, labeling the livers by injection of EdU in the mice 3–4 hr before harvesting, shows that the EYFP positive cells (indicating activity of Cre recombinase) were proficient in synthesizing DNA and incorporating EdU (Figure 5G and H). Nearly 100% of the hepatocytes were positive for EYFP, suggesting that the Cre recombinase was active in the vast majority of the hepatocytes (Figure 5G), and consistent with the complete depletion of ORC2 protein from the hepatocytes of all mice with that genotype (Figure 3C). Thus the 30% of the liver cells that were incorporating EdU in the Cre expressing livers (Figure 5H) were doing so in the absence of ORC2 protein. Also, we noted several pairs of EdU positive nuclei (marked by arrows in Figure 5G) whose relative positions suggest that they are sisters born from the same mitosis, suggesting that some EdU positive cells can go through mitosis in the EYFP positive cells. Finally, the nuclei were significantly larger in the Cre active Orc2^f/f hepatocytes than in the hepatocytes without Cre, both before and after regeneration (Figure 5I).

We did not explore why the EYFP protein is mostly nuclear in hepatocytes in culture (Figure 4C) and mostly cytoplasmic in hepatocytes in the liver tissue (Figure 5G), but speculate that differences in signaling pathways or fixation techniques between the two conditions contribute to this difference.

We have reported that conditional deletion of Orc1 in developing mouse livers still allowed livers to develop and induced premature endoreduplication, suggesting that significant DNA synthesis can occur in liver cells that are genetically deleted of Orc1. We bred the Orc1^f/f mice with Orc2^f/f mice to obtain mice where both Orc1 and Orc2 are floxed but Alb-cre is not present (Figure 6A and B). Even though Cre was not expressed in these livers, for unknown reasons there was a decrease in the percentage of progeny when Orc2 was floxed in the liver (with or without Orc1 being floxed), but not when just Orc1 was floxed (Figure 6C). These mice were then bred with Orc2^f/f Alb-cre^+/+ mice, and the resulting Orc1^f/+, Orc2^f/f, Alb-cre^+/- intercrossed to get Orc1^f/f, Orc2^f/f, Alb-Cre^+/-mice where the Orc1 and Orc2 alleles are conditionally deleted in hepatocytes, and the deletion is accompanied by activation of EYFP expression. Immunoblotting of isolated hepatocytes showed that in four out of four mice expressing the Cre recombinase the ORC1 and ORC2 proteins were decreased significantly (Figure 6D). Immunoblotting of some of the other replication proteins showed no decrease in other ORC subunits, CDC6, and two of the MCM2-7 subunits (Figure 6E). The ORC3 protein, which did not change when Orc2 was deleted (Figure 2D) was surprisingly elevated in Figure 6E when Orc1 and Orc2 were both deleted. We do not know why this is the case, and it was not seen consistently in all four animals. Similarly, the 50% decrease of MCM2 and MCM3 that was seen in the absence of ORC2 (Figure 2D), disappeared when both Orc1 and Orc2 are deleted.

The female Alb-Orc1^f/f Orc2^f/f mice were significantly smaller in size, with smaller livers and decreased liver to body weight ratio (Figure 6F, Figure 6—figure supplement 1A, Figure 6—figure supplement 2A), but the male mice were relatively less affected (Figure 6F). We have also observed ~50% lethality of double knock-out female mice within the first month of life, but the remaining 50% survive beyond 4 mo of age (Figure 6—figure supplement 1C). H&E staining showed that the double knockout (DKO) male livers have fewer but larger cells with significantly larger nuclei (Figure 6G and H, Figure 6—figure supplement 2C). Interestingly, this phenotype is even more marked in the female liver (Figure 6—figure supplement 1B). By flow cytometry, the EYFP positive cells as evidence of Cre recombinase activity had distinctly more polyploidization than the EYFP negative cells (Figure 6I). Taken together the results reveal that hepatocytes can synthesize DNA by endoreduplication to produce very large nuclei and very large cells so that the liver size is not hugely decreased, even in the absence of two subunits of ORC. Female mice that have a deletion in two subunits of ORC suffer more morbidity and mortality than male mice.

Finally, partial hepatectomy was performed in the livers of male mice with DKO of Orc1 and Orc2. As can be seen even before partial hepatectomy the body weights were not significantly different from WT mice (Figure 7A). Post regeneration, the liver weights and liver/body weight ratios were not decreased in the DKO mice compared to the WT mice (Figure 7B and C). H&E sections showed that the hepatocyte nuclei were larger and hepatocyte nuclear density lower in the DKO livers both pre-hepatectomy and in the regenerated livers 36 hr post-hepatectomy (Figure 7D–F). EdU labeling for 3–4 hr before harvesting the livers showed that despite the DKO, the EYFP positive cells showed extensive DNA synthesis (Figure 7G). Even though nearly 100% of the hepatocytes are EYFP positive (Figure 7G) and ORC1 and ORC2 proteins depleted in the hepatocytes (Figure 6D), 15% of nuclei stained with EdU (Figure 7H), a percentage that was higher in the DKO regenerating livers than in the WT livers, suggesting that the endoreduplication that accompanies liver regeneration can occur even after deletion of two of the six-subunits of ORC.

Work involving mice adhered to the guidelines of the Institutional Animal Care and Use Committees (IACUC) at the University of Virginia (protocol number 4198), the University of Alabama at Birmingham (protocol number 22335), the Ohio State University and the Medical University of South Carolina. Orc1 ^f/f ROSA26^stop-EYFP animals were previously reported (Okano-Uchida et al., 2018). Orc2 ^f/f mice were generated by Cyagen Biosciences Inc Exons 6–7 (amino acids L111-V150) was selected as conditional knockout region (cKO). Mouse genomic fragments containing homology arms (HAs) and cKO region were amplified from BAC clone by using high fidelity Taq DNA polymerase and were sequentially assembled into a targeting vector together with recombination sites and selection marker, Neo cassette, flanked by SDA (self-deletion anchor) sites. The linearized vector was electroporated into C57BL/6 ES cells that were subject to G418 selection (200 μg/mL) after 24 hr. 188 G418 resistant clones were picked and amplified in 96-well plates. The PCR screening identified 29 potential targeted clones, from among which 12 were expanded and further characterized by Southern blot analysis. Eleven of the twelve expanded clones were confirmed to be correctly targeted. Targeted ES cell clone N-1F10 was injected into C57BL/6 albino embryos, which were then re-implanted into CD-1 pseudo-pregnant females. Founder animals were identified by their coat color, their germline transmission was confirmed by breeding with C57BL/6 females and subsequent genotyping of the offspring. Three male and five female heterozygous targeted mice were generated from clone N-1F10. Floxed Orc2 mice were crossed to the Orc1 ^f/f ROSA26^stop-EYFP animals to generate Orc2 ^f/f ROSA26^stop-EYFP and Orc1 ^f/f Orc2 ^f/f ROSA26^stop-EYFP strains for usage of EYFP expression as a reporter for Cre recombinase activity and Orc2 or Orc1 and Orc2 deletion. Those were further bred into tissue-specific Sox2-Cre (Hayashi et al., 2002) or Alb-Cre mice to obtain Orc2 or Orc1 and Orc2 knockout in all cells of the inner cell mass following implantation or in the liver, respectively. We used two independent Alb-Cre lines to introduce the gene. At UVA/UAB we used the Alb-Cre mice from Postic et al., 1999. The Alb-Cre mice used at the University of Wisconsin are from Schuler et al., 2004. All the mice used in this study were maintained in a mixed background (C57BL/6x129 x FVB/N).

Genomic DNA from ear punches was isolated using Quick Extract DNA Extraction Solution (Lucigen., Cat# QE09050). All genotyping PCRs were carried out using MyTaq Red Mix (Bioline, Cat# BIO-25043) according to the manufacturer’s instructions. Orc1 genotyping was carried out with primers F1 forward (common to both alleles; 5′-GCTGCTTCAGTGTGGCAATA-3′), R1 reverse (specific for the WT allele; 5′-CTCCAATTGTTCCCCAGCTA-3’), and R2 reverse (specific for deleted allele, 5′-CACCTGTCACTGGACCACAC-3′). The PCR parameters were 95 °C for 30 s, 45x (95 °C for 15 s, 60 °C for 20 s, 72°for 60 s), 72 °C for 5 min. PCR product was run on 2% agarose gel and WT band was detected at 439 bp, transgenic band at 310 bp, and deleted at 677 bp. Orc2 genotyping was carried out with primers F1 forward (common to both alleles; 5′- GAGGTTGTGGCTGTAATATACGTGATC –3′), and R1 reverse (common to both alleles; 5′- CTGAGCCATCTAACTCCTTCCTAGC –3’), or F2 forward (common to both alleles; 5′- TGGGTAGGTTCATTCCAGTTTAGCC –3′), and R2 reverse (common to both alleles; 5′- ACCTTGGTATTGGACGTCTCTATTC –3’). The PCR parameters were 95 °C for 30 s, 35x (95 °C for 15 s, 55 °C for 20 s, 72°for 60 s), 72 °C for 5 min. To detect constitutive KO allele combination of F1, F2, and R2 was used. PCR product was run on 2% agarose gel and following bands were detected: for F1 + R1 – WT at 342 bp, and transgenic band at 398 bp; for F2 +R2 WT at 254 bp, and transgenic band at 367 bp; for F1+F2+R2 WT at 254 bp, transgenic band at 367 bp, and deleted at 306 bp. ROSA26 genotyping was carried out with primers F1 forward (common to both alleles; 5′-AAAGTCGCTCTGAGTTGTTAT-3′), R1 reverse (specific for the WT allele; 5′- GGAGCGGGAGAAATGGATAT-3’), and R2 reverse (specific for deleted allele, 5′- GCGAAGAGTTTGTCCTCAACC-3′). The PCR parameters were 95 °C for 30 s, 45x (95 °C for 15 s, 60 °C for 20 s, 72°for 60 s), 72 °C for 5 min. PCR product was run on 2% agarose gel and WT band was detected at 650 bp, and transgenic band at 340 bp. Sox2-Cre genotyping was carried out with primers F1 forward (common to both alleles; 5′- ATGCTTCTGTCCGTTTGCCG –3′) and R1 reverse (common to both alleles; 5′- CCTGTTTTGCACGTTCACCG –3’), with Orc1 primers F1 and R1 for an internal control. The PCR parameters were 94 °C for 3 min, 38x (94 °C for 30 s, 60 °C for 30 s, 72 °C for 40 s), 72 °C for 3 min. PCR product was run on 3% agarose gel and the transgenic band was detected at 875 bp with an internal control band at 439 bp. Alb-cre genotyping was carried out with primers Alb-cre-20239-F (specific for the WT allele; 5’-TGCAAACATCACATGCACAC-3’), Alb-cre-olMR5374-F (specific for the transgenic allele; 5’-GAAGCAGAAGCTTAGGAAGATGG-3’), and Alb-cre-20240-R (common to both alleles; 5’-TTGGCCCCTTACCATAACTG-3’). The PCR parameters were 95 °C for 30 s, 35x (95 °C for 15 s, 55 °C for 20 s, 72°for 60 s), 72 °C for 5 min. PCR product was run on 4% agarose gel and WT band was detected at 351 bp, and transgenic band at 390 bp. To establish sex of embryos isolated for MEFs, SRY genotyping was carried out with chromosome Y specific forward (5’-TTGTCTAGAGAGCATGGAGGGCCATGT-3’) and reverse primers (5’-CTCCTCTGTGACACTTTAGCCCTCCGA-3’). The PCR parameters were 95 °C for 30 s, 35x (95 °C for 15 s, 55 °C for 20 s, 72°C for 60 s), 72 °C for 5 min. PCR product was run on 2% agarose gel and Y-chromosome positive band was detected at 270 bp.

MEFs were isolated from E12.5 Orc2^+/+ or Orc2^f/f embryos, transformed with SV40 large T antigen, and infected with adenovirus Cre-eGFP (#VVC-U of Iowa-1174, University of Iowa). The infected MEFs were cultured for indicated days in DMEM with 10% FBS medium. To measure the cell proliferation, 24 hr. after the Adenovirus Cre-eGFP transduction, SV40 transformed MEF cells were plated in 96 well plates.

The cell viability was measured every 24 hrs using CellTiter 96 Non-Radioactive Cell Proliferation Assay (Promega, #G4100) according to the manufacturer’s instructions. All experiments were conducted in triplicate and absorbance relative to that on day 1 was expressed.

Control (Orc2^f/f ROSA26^stop-EYFP or Orc1^f/f Orc2^f/f ROSA26^stop-EYFP) and experimental animals (Alb-Orc2^f/f ROSA26^stop-EYFP or Alb-Orc1^f/f Orc2^f/f ROSA26^stop-EYFP) were euthanized using CO₂. Blood was collected was further metabolic studies. Animals were perfused with prewarmed (39 °C) Hank’s buffered salt solution (HBSS) containing EDTA, MgCl₂, and HEPES. The livers were dissected and weighted. Each liver was divided for the following experiments: ploidy analysis (right lobe; fresh processing), EYFP flow cytometry (median lobe; fresh processing), immunoblotting and histology (left lobe; half for LN₂ flash freeze and second half into 10% formalin), and genotyping (caudate lobe; LN₂ flash freeze).

SV40 transformed MEF with or without adenovirus Cre-eGFP infection were directly lysed in 2 x Laemmli Sample Buffer and sonicated. Liver was lysed in modified RIPA buffer (150 mM Sodium Chloride, 50 mM Tris-HCl, pH 7.4, 1 mM EDTA, 1 mM PMSF, 1% Triton X-100, 1% Sodium Deoxycholic Acid, 0.1% SDS), sonicated, and lysate was clarified by centrifugation. Mouse ORC2 antibody was raised against His tagged full length of ORC2 recombinant protein in Rabbit (Pacific Immunology). The antibodies used in this study are listed: ORC1 (Santa Cruz; sc-28741); ORC3 (Santa Cruz; sc-374231); ORC5 (Boster Biological technology; A03676-1); ORC6 (Santa Cruz; sc-390490); CDC6 (Santa Cruz; sc-9964); MCM2 (Abcam; ab4461); MCM3 (Santa Cruz; sc-9850).

All formalin-fixed paraffin-embedded (FFPE) sections and H&E staining were performed by Research Histology Cores at UVA and UAB. Nuclei size was measured using ImageJ 1.50i (Java 1.6.0_24) (56) (Schneider et al., 2012). The number of analyzed animals is annotated at each figure. The statistical method used for comparison between experimental groups was a two-tailed homoscedastic Student’s t-test. Statistical significance was expressed as a p-value. We captured images of 5–10 fields per liver and measured nuclear size relative to scale bar and nuclear density by counting the number of nuclei per field (at a fixed scale to compare between mice). Hepatocyte nuclei can be easily distinguished from stromal nuclei by their roundness and relative de-condensation.

Blood was collected from control Orc2^f/f ROSA26^stop-EYFP and experimental Alb-Orc2^f/f ROSA26^stop-EYFP animals and centrifuged at 1000 g for 10 min at 4 °C. To measure Alanine Transaminase (ALT) and Aspartame Aminotransferase (AST) activities EnzyChromTM Alanine Transaminase Assay Kit (BioAssay Systems, Cat# EALT-100) and Liquid AST (SGOT) Reagent Set (Pointe Scientific, Cat# A7561150) were used respectively according to manufacturers’ instructions. The number of analyzed animals is annotated at the figure. The statistical method used for comparison between experimental groups was a two-tailed homoscedastic Student’s t-test. Statistical significance was expressed as a p-value.

Hepatocytes were isolated according to STAR protocols (Charni-Natan and Goldstein, 2020). In brief, the liver perfused with perfusion Buffer for 10 min followed by digestion buffer for 10 min was dissected out and hepatocytes were released into plating medium containing dish. A single-cell suspension was obtained by filtering through a 100 µm cell strainer. Percoll centrifugation (Cytiva #17089102) was performed to remove dead cells. Isolated live Hepatocyte were suspended in William’s medium (WEM, GIBCO A1217601) with Plating Supplement (GIBCO #CM3000) and plated on collagen-coated cover glass (5×10^5 cells/6-well). After 3 hr, the medium was exchanged for maintenance medium (William’s E Medium (WEM, GIBCO A1217601) supplemented with GIBCO #CM4000).

Nuclei Isolation Medium (NIM; 250 mM Sucrose, 25 mM KCl, 5 mM MgCl₂, 10 mM Tris-Cl, 1 mM DTT, 1 x Protease inhibitor) with 2% paraformaldehyde and 0.1% Triton X-100 was added to the liver pieces that were subsequently homogenized with Pestle A Dounce homogenizer (25 x times). After all samples were processed, they were centrifuged at 1000 g for 10 min at 4 °C. Pellet was resuspended in NIM and equal volume of 50% iodixanol was added. The mixed sample was carefully layered on the top of 29% Iodixanol solution in ultracentrifuge tube and spun at 10,300 rpm for 10 min at 4 °C in ultracentrifuge. The nuclei pellet was resuspended in FACS buffer (1 x PBS, 2.5% (v/v) BSA, 2 mM EDTA, 2 mM NaN₃) with 100 µg/mL RNase A. DRAQ5-stained liver nuclei samples were processed using Attune NxT flow cytometer (Life Technologies). Flow cytometry data from 10,000 nuclei were analyzed with FCS express software. The bottom 40% of nuclei on the EYFP axis in the FACS profiles was considered as low EYFP and the top 20% as high EYFP. The EYFP low nuclei are mostly from non-hepatocytes (Kupffer cells, endothelial cells, bile duct cells, contaminating blood cells) and some hepatocytes that have not yet expressed sufficient levels of EYFP and they all serve as a control population. The EYFP high nuclei are exclusively from hepatocytes that have undergone the Cre-mediated recombination a sufficient time back to express high levels of EYFP and are the experimental population with ORC subunit deletion. The statistical method used for comparison between experimental groups was a two-tailed homoscedastic Student’s t-test. Statistical significance was expressed as a p-value.

Two days after plating, Hepatocyte was labeled with 20 μM of EdU (Lumiprobe #10540) for 2 hrs and fixed with 4% paraformaldehyde for 10 min followed by permeabilized with 0.25% Tritonx-100 for 5 min. The fixed cells were incubated with label mix [8 µM Sulfo-Cy3-Azide (Lumiprobe #B1330), 2 mM CuSO4*5H2O, 113 mM Ascorbic Acid] for 30 min. Anti-GFP antibody (Abcam #ab6556) was used to detect EYFP signal after the EdU staining.

The partial hepatectomy experiment was conducted following a standardized protocol (Mitchell and Willenbring, 2008). Briefly, 8–14 wk-old mice were utilized. After inducing anesthesia with 2% isoflurane and maintaining at 0.2%, the mice were subcutaneously injected with Buprenorphine-SR (0.6 mg/kg) and Carprofen (5 mg/kg). The abdominal wall was shaved and prepared aseptically. A 3 cm long transverse incision was made to expose the xiphoid. The left and middle lobes were tied using 4–0 thread and cut. The incision site was double-checked for bleeding following washed with 0.9% sodium chloride and sutured. Mice were placed on a warm pad for recovery. To detect DNA synthesis, 0.2 mg EDU (Lumiprobe #10540) was injected via the tail vein 3–4 hr before sacrifice. At 36–48 hr post-surgery, samples from the regenerating right liver were collected and weighed. For IHC and IF staining, samples were fixed with 4% formaldehyde, dehydrated, cleared, embedded in paraffin, and sectioned at 5–8 μm thickness. Tissue sections were deparaffinized and rehydrated, and heat antigen retrieval methods were applied in 10 mM sodium citrate buffer (pH = 6.0). Permeabilization with 0.25% Triton X-100 followed EDU incorporation by Click-in reaction for 40 min at room temperature, protected from light. After washing and blocking, an anti-GFP antibody (Abcam, cat# ab6556, 1:500 dilution) and Alexa-488 fluorescent-conjugated secondary antibody (A11029; Life Technologies) were used to detect EYFP signaling. Finally, nuclei were stained with DAPI, and images were captured using Zeiss Confocal microscope and processed using ImageJ and GraphPad software. To improve the EYFP signal, we used frozen sections for the liver regeneration experiments in the Orc1, Orc2 double knockout mice in Figure 7.

To estimate the minimal number of cell divisions occurring in the hepatoblasts after the appearance of albumin-driven Cre, we first estimated how many hepatocyte nuclei populate an adult liver. Given that there are 125 million liver cells/gram of tissue (Sohlenius-Sternbeck, 2006), we estimate that there are 162.5 (female) to 218 (male) million hepatocytes for the 1.3 gram (female) or 1.75 gram (male) livers (all estimates of liver weight and liver nuclear density are from Figures 5 and 6, and Figure 6—figure supplement 1).

We next estimated the number of albumin-positive cells early in embryogeneis, from published scRNAseq results that suggest that ~1% of the embryonic cells are of hepatocyte lineage when all embryonic cells are harvested from E9.5-E13.5 embryos (Cao et al., 2019). Given that there are ~200,000 cells in E9.5 embryos (Cao et al., 2019), we, therefore, estimate that there are 2000 albumin positive cells at the stage when albumin mRNA expression is readily detected. Thus, normal development of the liver requires at least 20 rounds of diploid cell division from the 2000 hepatoblasts/hepatocytes seen in E9.5 embryos to produce the ~200 million hepatocyte nuclei in the adult liver (Table 2, bottom row).

From the relative weights of the adult livers (row 1) and the relative densities of the hepatocyte nuclei (row 2) we can estimate the number of hepatocytes in the mutant livers relative to wild-type livers (row 3). The relative deficit of hepatocytes in the mutant livers allows us to estimate how many fewer cell divisions the hepatocytes underwent in the mutant livers relative to the WT livers during development (row 4). Since WT hepatocytes undergo ~20 duplications, we thus estimate that the Orc2-/- hepatocytes and the male DKO hepatocytes undergo at least 18 divisions from the 2000 hepatoblast-stage seen in E9.5 embryo, while the female DKO hepatocyte undergo at least 15 divisions (Table 2, bottom row). This underestimates the number of divisions because we do not take into account any apoptosis that may be occurring or any endoreduplication cycles.