Chorionic gonadotropin stimulates maternal hepatocyte proliferation during pregnancy

Abstract

The liver increases its size during pregnancy to adapt to metabolic demand associated with pregnancy. Our previous study showed that proliferation of maternal hepatocytes are increased during pregnancy in mice and that estradiol (E2) is one of the candidate hormones responsible for maternal hepatocyte proliferation. Here, we discovered that chorionic gonadotropin (CG) induces maternal hepatocyte proliferation during pregnancy. CG administration was sufficient to stimulate hepatocyte proliferation in non-pregnant mice as well as in cell culture system. We conclude that CG stimulates proliferation in the early pregnancy of maternal hepatocytes. In contrast, estrogen stimulates hepatocyte proliferation in the late pregnancy.

Graphic abstract

Introduction

There is a close association between pregnancy and the liver. The liver has a role as a reproductive organ in female fish [1, 2]. Fish liver expresses zona pellucida sperm-binding protein 3 (ZP3), a one of the components of extracellular glycoprotein layer surrounding the plasma membrane of mammalian oocytes. In mammals, the liver increases its size to adapt to multitude of changes associated with pregnancy and support fetal growth [3]. We previous unmasked that estrogen, a female hormone that steadily rises during pregnancy, stimulates hepatocyte proliferation through one of the estrogen receptors, estrogen receptor alpha (ERα, ESR1, [4]).

It is well known that the liver has high capacity for regeneration. In rodents, the liver proliferates to restore the size in few weeks after 70% partial hepatectomy (PH) [5, 6]. Francavilla et. al., (1986 and 1989) reported that endogenous β-estradiol (E2) production is induced within a few hours after 70% PH in humans [7, 8]. Using ESR1 KO mice and hepatocyte-specific ESR1 KO mice, we showed PH-induced E2 binds to ESR1 in hepatocyte and stimulates liver regeneration [4, 9]. Interestingly, we found that E2 induced hepatocyte proliferation is also observed in estrus cycle [9]. However, it is difficult to study the effect of ESR1 on the liver during pregnancy using whole body ESR1 KO mice that are infertile [4, 10].

Importantly, hormonal changes associated with pregnancy is not limited to the rise in E2. Chorionic gonadotropin (CG) is a reproductive hormone secreted from placenta, especially at high level during the first trimester of pregnancy [11]. CG supports the maintenance of the corpus luteum at the beginning of pregnancy to promote progesterone secretion necessary for the establishment of pregnancy. Interestingly, serum CG was elevated in approximately 22% of male subjects affected by hepatocellular carcinoma implicating a potential role of CG in proliferation of hepatocytes [12]. However, it is unknown whether CG plays a role in the increase in liver size during pregnancy. Here, we found that CG promotes maternal hepatocyte proliferation.

Materials and methods

Reagents:

β-Estradiol (E2, Wako 052-04041), and progesterone (P, Wako 632-23141) were purchased from Wako Chemicals (Osaka, Japan). The estradiol enzyme immunoassay (EIA) kit (No. 582251) was purchased from Funakoshi (Tokyo, Japan). Human chorionic gonadotropin (CG) was purchased from Mochida Pharmaceutical Company (HCG Mochida for intramuscular injection, [13, 14]). Recombinant human placental lactogen (PL) was purchased from Sigma (SRP4869). RNase A was purchased from Nacalai (30142-04, Kyoto Japan).

Cell culture:

The Hep G2 cell line from human hepatocellular carcinoma was purchased from the American Type Culture Collection (HB-8065, ATCC). The Hep G2 cell line was maintained in alpha-minimal essential medium (α-MEM, 11900-073, Gibco, Tokyo, Japan) supplemented with 10% fetal calf serum (FCS, CC3008-504, Cell Culture Technology, Tokyo, Japan). The cells were plated at a density of 1.5–3 × 10⁶ cells/60-mm dish containing α-MEM supplemented with 10% FCS, and after 48 h, the culture medium was exchanged for α-MEM without phenol red but supplemented with 0.5% charcoal-treated FCS. Phenol red was eliminated because it is a known phytoestrogen. The charcoal treatment was used to reduce the high concentration of endogenous estrogen in FCS. After 24 h, the cells were treated with β-estradiol or an equal volume [0.1% (v/v)] of vehicle (ethanol/EtOH, 14712-34, Nacalai Tesque, Kyoto, Japan).

The HEK293T cells was purchased from the American Type Culture Collection (CRL-1573, ATCC). The HEK293T cell line was maintained in α-MEM (11900-073, Gibco, Tokyo, Japan) supplemented with 10% fetal calf serum (FCS, CC3008-504, Cell Culture Technology, Tokyo, Japan). The cells were plated at a density of 1.5–3 × 10⁶ cells/60-mm dish containing α-MEM supplemented with 10% FCS, and after 48 h, the culture medium was exchanged for α-MEM without phenol red but supplemented with 0.5% charcoal-treated FCS. Phenol red was eliminated because it is a known phytoestrogen. The charcoal treatment was used to reduce the high concentration of endogenous estrogen in FCS. After 24 h, the cells were treated with β-estradiol or an equal volume [0.1% (v/v)] of vehicle (ethanol/EtOH, 14712-34, Nacalai Tesque, Kyoto, Japan).

Mouse hepatocyte primary culture:

Primary hepatocytes were isolated C57BL/6J mouse (4-week-old or 1-year-old, Charles River Japan) and prepared using the two-step collagenase perfusion method [15]. The primary culture of mouse hepatocyte from C57BL/6J was maintained with CM4000 (Thermo Fisher) changing every two days. The medium was contained with 0.1 μM dexamethasone, 6.25 μg/mL insulin, 6.25 μg/mL transferrin, 6.25 ng/mL selenous acid, 1,25 ng/mL BSA, 5.35 μg/mL linoleic acid, 2mM GlutaMAX and 15 mM HEPES 7.4 [13].

Mice:

C57BL/6J and CD1 WT mice were obtained from Charles River Japan. After vaginal plug observation in the morning, pregnant females were identified as 0.5 DPC. Number of embryo/placenta was counted after dissection at indicated time [9].

Maternal hepatocyte proliferation in mice:

The hepatocyte proliferation rates of female mice were analyzed by counting bromodeoxyuridine (BrdU, M0744, Dako, Tokyo, Japan) positive nuclei by immunohistochemistry (IHC, SK-4105, Vector). The mice received 50 mg/kg of BrdU intraperitoneally 2 h before euthanasia, and the livers were then removed, rinsed, and embedded in Tissue-Tek OCT cryo embedding compound (Sakura Finetek Japan, Tokyo, Japan). Subsequently, 10-μm cryosections of the liver were fixed with 4% paraformaldehyde, incubated with an anti-BrdU antibody (No. 11170376001, Roche Diagnostics Japan, Tokyo, Japan) diluted 50-fold in 0.1% bovine serum albumin/phosphate-buffered saline, and then labeled with CY3-conjugated donkey anti-rabbit IgG antibody. The sections were then mounted on slides in Vectorshield medium (Vector Laboratories), and the number of BrdU-positive hepatocyte nuclei in each sample containing approximately 2000 hepatocytes were counted in at least five low-magnification microscopic fields. Ki67 immunohistochemistry was performed with 10 μm cryosections of the liver using anti Ki67 antibody (ab15580, abcam, [9]). Mitotic body was counted with HE-stained paraffin section [9]. E2, CG, and PL were intravenously injected with indicated doses in the figure legends.

Radiolabeled thymidine (methyl-³H, NET027A) was injected to mice for 7 days (1 MBq/mouse/day). At day 8, the liver was removed, and homogenized

mRNA analyses:

Total liver RNA was extracted by the guanidium-thiocyanate-phenol-chloroform method. The reverse-transcribed cDNA was synthesized from 1 μg total liver RNA mixed with three mice. qPCR was performed with using Applied Biosystems 7300 Real-Time PCR System [9]. PCR primers are LHCGR; 5’- cacattccatttctgaaaacttttccaaac-3’ and 5’-cttgggtaagcagtttggatattcatagtc -3’. As internal control β-actin; 5’-catgtacccaggcattgctgacaggatgca-3’ and 5’-tccacacagagtacttgcgctcaggaggag-3’.

Animal study compliance:

All experiments were performed in accordance with the ethical guidelines for animal care established by the National Center for Geriatrics and Gerontology (NCGG), and the study was approved by the Animal Care Committee.

Statistical and reproducibility:

The values are reported as the means ± standard error (S. E.). Statistical significances (single-sided Student’s t-test) are indicated in figure legends as follows: *, p < 0.05; **, p < 0.005; ***, p < 0.0001. For reproducibility of key experiments, we performed experiments more than 10 times in total including experiments performed for conditioning. Also, we employed multiple approaches to confirm one result [16, 17].

Results

Proliferation of maternal hepatocyte is up-regulated in an embryo/placenta number dependent manner during early and middle pregnancy. (Figure 1).

Figure 1. Proliferation of maternal hepatocyte is up-regulated in an embryo/placenta number dependent manner during early and middle pregnancy.

(A) BrdU-positive hepatocytes were counted after immunohistochemical staining. Higher maternal hepatocyte proliferation was detected in CD-1 (approximately 12 embryos) and C57/BL6J (approximately 6 embryos) during pregnancy. Data represent the mean ± S.E. (n = 6). **, p < 0.005; ***, p<0.0001. 3.5dpc = 0.0006470694, 4.5 dpc = 0.0003908772. Peak p = 0.0000800.

(B and C) Maternal hepatocyte proliferation were increased in an embryo/placenta number dependent manner. Hepatocyte proliferation was analyzed as BrdU incorporation (B), and the number of mitotic body (%, C) respectively. Proliferating hepatocyte number was increased in an embryo/placenta number dependent manner at 6.5-8.5 dpc. Data represent the mean ± S.E. (n = 6).

To identify a factor that increases proliferation of maternal hepatocyte during pregnancy, we compared maternal hepatocytes of prolific CD-1 mice (approximately 12 embryos/pregnancy) and those of C57/BL6J mice (approximately 6 embryos/pregnancy, Fig. 1A). Analysis revealed that the proportion of proliferating hepatocytes was higher in pregnant CD-1 female mice compared with C57/BL6J mice. The proliferation peak was in the middle of pregnancy (Fig. 1A). The proliferation of maternal hepatocyte positively correlated with the number of embryo/placenta (Figs. 1B and 1C). These data suggest that embryo/placenta secretes factors that stimulate proliferation of maternal hepatocyte in the beginning or middle pregnancy.

Candidates that regulate maternal hepatocyte proliferation in early pregnancy (Figure 2)

Figure 2. Candidates that regulate maternal hepatocyte proliferation in early pregnancy.

Estradiol (E2, A), placental lactogen (PL, B) and chorionic gonadotropin (CG, C) levels in plasma were measured. All the three hormones are in embryo/placenta number dependent manner. E2 and PL were mainly secreted in the late pregnancy, although CG secreted rather in middle of pregnancy. Data represent the mean ± S.E. (n = 6).

The first candidate for a maternal hepatocyte proliferator secreted from embryo/placenta is E2 based on our previous study showing that E2 stimulates liver regeneration after partial hepatectomy and pregnancy [4, 9]. We demonstrated that E2 induces ESR1 expression in the liver and stimulates hepatocyte proliferation. ESR1 mediates the E2 signaling in hepatocyte based on our study in ESR1 KO mice [4, 9]. In agreement with our previous study, E2 is secreted in the late pregnancy (Fig. 2A). Placental lactogen (PL) is also induced in the late pregnancy (Fig. 2B). On the other hands, chorionic gonadotropin (CG) is secreted before middle pregnancy (Fig. 2C). As shown in Figs. 1B and 1C, hepatocyte proliferation during pregnancy peaks before 10 dpc when determined as BrdU incorporation and mitotic body. In agreement, both BrdU incorporation and mitotic body counting indicated that the increase in hepatocyte proliferation that depends on the number of embryo/placenta occurs from 6.5 to 8.5 dpc. During the same period of pregnancy, E2 (Fig. 2A) did not differ between mice with low and high numbers of placenta/embryo. A number of hormones are secreted from the embryo/placenta that may regulate hepatocyte proliferation. (Fig. 1 and [11]). CG was higher during early pregnancy in a group of mice with a high number of placenta/embryo making CG a strong candidate for maternal hepatocyte proliferator during pregnancy (Fig. 2C).

CG stimulates hepatocyte proliferation (Figures 3, 4).

Figure 3. Chorionic gonadotropin (CG) administration stimulates the proliferation of mouse hepatocytes, hepatocyte primary culture and Hep G2 cells.

(A-C) CG administration to mice induces liver weight and hepatocyte proliferation. CG was injected for 7 days, and then mice liver was analyzed with liver weight (A), hepatocyte proliferation with BrdU immunohistochemistry (A), thymidine incorporation (B), Ki67 immunohistochemistry (B), and mitotic body (C).

(D) CG-administrated primary hepatocyte culture. CG stimulates hepatocyte primary culture in a dose-dependent manner.

(E) CG-administrated Hep G2 culture. CG stimulates Hep G2 culture in a dose-dependent manner.

(F) CG administration to HEK293T cells had no effect on cell numbers. Data represent the mean ± S.E. (n = 6).

Figure 4. Chorionic gonadotropin (CG) stimulates mice hepatocyte proliferation in male and female mice.

(A-C) CG induced liver weight gain (A) and thymidine incorporation in the liver (B) in both sexes of C57/BL6J and CD-1 mice, but not induce LHCGR expression (C).

(D) Higher CG-induced hepatocyte primary culture in younger mice (4-weeks old, filled rectangle). No difference was observed without CG between 4 weeks and 12 month mice, although CG administration induces hepatocyte proliferation in younger mice.

(E) Aging reduced LHCGR expression.

(F) Possible model of aging-reduced regenerative activity.

CG administration to non-pregnant CD1 mice induced liver weight gain (Fig. 3A), BrdU incorporation (Fig. 3A), thymidine incorporation (Fig. 3B), Ki67-immunocytochemistry (Fig. 3B), and mitotic body counting (Fig. 3C) in a dose-dependent manner. CG administration induced the proliferation of primary hepatocyte culture (Fig. 3D) and hepatocellular carcinoma cell line Hep G2 (Fig. 3E), but not non-hepatocyte cell line HEK293T (Fig. 3F). Moreover, CG administration stimulated hepatocyte proliferation in both male and female CD-1 and C57BL/6J mice (Figs. 4A and 4B) without changing LHCGR expression determined by qPCR (Fig. 4C). Collectively, CG is a likely mediator that simulates hepatocyte proliferation during early pregnancy. Taken together, we identify a CG as maternal hepatocyte proliferator during pregnancy.

CG administered to CD-1 and C57BL6J mice, and CG stimulates hepatocyte proliferation in both sex of CD-1 and C57BL6J mice (Figs. 4A and 4B). Although CG does not change LHCGR expression (Fig. 4C).

Aging reduces regenerative activity [18], so we analyzed hepatocyte primary culture from young (4-week-old) and aged (12-month-old) mice with/without CG (Fig. 4D). There is no difference between young and aged hepatocyte number without CG at 7 and 14 days (p=0.4260710 and p=0.8618488), although significant differences were observed with CG. We analyzed LHCGR expression in young and aged mice liver (Fig. 4E). LHCGR expressions were declined by age in both sex (Fig. 4E).

Conclusion

The proliferation of maternal hepatocyte during pregnancy is regulated by maternal hormone of CG in the early phase of pregnancy, a hormone that also induced polyploidy and apoptosis in hepatocytes (Graphic abstract).

Discussion

We identified CG as hepatocyte-specific proliferative factor during pregnancy that regulate hepatic proliferation in the pregnancy and apoptosis. It is well established that the maternal liver increases its size during pregnancy. However, the biological significance of proliferation of maternal hepatocytes is not fully understood. One possible benefit is the detoxification of embryo derived substances. Waste products from the embryo are increased especially in late pregnancy as the size of embryos increase. There also is increased demand for synthesis of protein and other macromolecules by the maternal liver during fetal development. It is plausible that proliferation of maternal hepatocytes driven by CG during early pregnancy may prepare the maternal liver ahead of the time for growth of embryo. Considering that the liver receives waste derived from embryos, the increase in maternal hepatocyte number by CG during early pregnancy may be important to protect material liver for the increased demand for detoxification from the growth of embryos.

In summary, we report that CG as a novel mediator of hepatocyte remodeling during early pregnancy that increases hepatocyte proliferation.

Highlights.

• CG-LHCGR signaling is maternal hepatocyte proliferator during pregnancy.

• Impaired CG-induced aged hepatocyte proliferation.

Abbreviation

BrdU

Bromodeoxyuridine

CG

Chorionic gonadotropin

LHCG-R

Luteinizing hormone/chorionic gonadotropin receptor

DPC

Day post coitum

E2

Estradiol

ESR

Estrogen Receptor

FCS

Fetal calf serum

ISH

In situ hybridization

P

Progesterone

PPARα

Peroxisome proliferator-activated receptor α

PI

Propidium iodide

PL

Placental lactogen

RT-PCR

Reverse transcriptase polymerase chain reaction

WB

Western blot

WT

Wild type

Data and materials availability

The authors declare that all reported data in the main and supplementary files will be provided to other investigators as requested should be addressed to TI (timai@ncgg.go.jp).