METHOD OF CULTURING PROLIFERATIVE HEPATOCYTES

Abstract

A method of culturing animal cells, preferably primary hepatocytes, including a first step of culturing the animal cells in non-adherent culture vessel, preferably a low or ultra-low attachment culture vessel, a second step of embedding the animal cells in a collagen matrix or in a gelatin matrix, and a third step of culturing the animal cells embedded in the collagen matrix or in the gelatin matrix, thereby obtaining 3D animal cell structures including proliferative animal cells, preferably spheroids including proliferative primary hepatocytes. Also, a spheroid including proliferative primary hepatocytes and the uses thereof for engineering an artificial liver model or an artificial liver organ, and for assessing in vitro the liver toxicity, genotoxicity and/or the effects of a drug or a compound.

Claims

1-15. (canceled)

16. A method of culturing animal cells to obtain 3D animal cell structures comprising proliferative animal cells, said method comprising: a) first culturing the animal cells in a non-adherent culture vessel; b) then transferring the animal cells to a culture medium comprising collagen, or truncated collagen also known as gelatin, thereby embedding the animal cells in a collagen or gelatin matrix; and c) culturing the animal cells embedded in the collagen or gelatin matrix; thereby obtaining 3D animal cell structures comprising proliferative animal cells.

17. The method according to claim 16, said method comprising: a) first culturing the animal cells in a non-adherent culture vessel; b) then transferring the animal cells to a culture medium comprising collagen, thereby embedding the animal cells in a collagen matrix; and c) culturing the animal cells embedded in the collagen matrix; thereby obtaining 3D animal cell structures comprising proliferative animal cells.

18. The method according to claim 17, wherein at step b) the animal cells are transferred to a culture medium comprising collagen at a concentration ranging from about 0.25 mg/mL to about 3 mg/mL.

19. The method according to claim 17, wherein at step b) the animal cells are transferred to a culture medium comprising fibrillar collagen.

20. The method according to claim 17, wherein at step b) the animal cells are transferred to a culture medium comprising fibrillar collagen selected from the group consisting of type I collagen, type II collagen, type III collagen, type V collagen, type VI collagen, type XI collagen, type XXIV collagen, type XXVII collagen and any mixtures thereof.

21. The method according to claim 16, said method comprising: a) first culturing the animal cells in a non-adherent culture vessel; b) then transferring the animal cells to a culture medium comprising methacrylated gelatin (GelMa), thereby embedding the animal cells in a GelMa matrix; and c) culturing the animal cells embedded in the GelMa matrix; thereby obtaining 3D animal cell structures comprising proliferative animal cells

22. The method according to claim 21, wherein at step b) the animal cells are transferred to a culture medium comprising methacrylated gelatin (GelMa) at a concentration ranging from about 1% (w/v) to about 20% (w/v).

23. The method of claim 16, wherein at step a) the non-adherent culture vessel is a low or ultra-low attachment culture vessel.

24. The method according to claim 16, wherein at step a) the animal cells are cultured in a non-adherent culture vessel for a period ranging from about 1 h to about 96 h.

25. The method according to claim 16, further comprising a fourth step of adding an inhibitor of the MAPK MEK1/2-ERK1/2 pathway to the culture of animal cells embedded in the collagen or gelatin matrix, thereby inducing one or more additional wave(s) of proliferation of the animal cells.

26. The method according to claim 16, further comprising a last step of incubating the 3D animal cell structures comprising proliferative animal cells embedded in the collagen or gelatin matrix with a collagenase or an enzyme mixture with collagenolytic activity, thereby isolating the 3D animal cell structures comprising proliferative animal cells from the collagen or gelatin matrix.

27. The method according to claim 16, wherein the animal cells are primary hepatocytes, and the 3D animal cell structures comprising proliferative animal cells are spheroids comprising proliferative primary hepatocytes.

28. The method according to claim 16, wherein the animal cells are primary human hepatocytes, and the 3D animal cell structures comprising proliferative animal cells are spheroids comprising proliferative primary human hepatocytes and having an acinus-like structure with a hollow lumen.

29. A spheroid comprising proliferative primary hepatocytes embedded in a collagen or gelatin matrix, wherein said spheroid has an acinus-like structure with a hollow lumen.

30. The spheroid according to claim 29, wherein said proliferative primary hepatocytes are proliferative primary human hepatocytes.

31. An in vitro method of assessing the liver toxicity, the liver genotoxicity and/or the effects of a drug or a compound, the method comprising: a) obtaining 3D animal cell structures comprising proliferative animal cells according to the method of claim 16; b) contacting the 3D animal cell structures comprising proliferative animal cells with a drug or a compound; and c) assessing the liver toxicity, the liver genotoxicity and/or the effects of the drug or the compound on the 3D animal cell structures comprising proliferative animal cells.

32. The in vitro method according to claim 31, wherein the 3D animal cell structures comprising proliferative animal cells are spheroids comprising proliferative primary hepatocytes.

33. The in vitro method according to claim 31, wherein the 3D animal cell structures comprising proliferative animal cells are spheroids comprising proliferative primary human hepatocytes and having an acinus-like structure with a hollow lumen.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0335] FIG. 1 is a scheme illustrating the method of the invention for culturing animal cells such as hepatocytes. (A) Illustration of the first step of the method of the invention: animal cells are cultured in a non-adherent culture vessel, such as a low or ultra-low attachment plate. (B) Illustration of the second and third steps of the method of the invention: the animal cells first cultured in a non-adherent culture vessel are embedded in a collagen matrix or in a gelatin matrix, in particular a GelMa matrix (second step) and cultured within said matrix (third step).

[0336] FIG. 2 is a group of photographs showing the formation of spheroids of primary human hepatocytes (PHH) in 3D cultures in a collagen matrix. The photographs were taken using two-photon excited fluorescence (TPEF) microscopy, after 2 days (A, B) or after 7 days (C, D, E, F) of 3D cultures established according to the method of the invention. 100μm depth image stack (TPEF stack) were used for 3D reconstruction (B, D, F). Second Harmonic Generation (SHG) microscopy was used after 15 days of 3D cultures established according to the method of the invention in order to analyze collagen signatures (CS) around (CS1) and between (CS2) PHH spheroids (G, H). A, C Bar scale=100 μm; E Bar scale=20 μm, H Bar scale=50 μm.

[0337] FIG. 3 is group of western-blots showing the expression in PHH of (A) pro-apoptotic factors (phospho-BAD, BAX, BAK, BID, BIM and PUMA) and (B) pro-survival factors (MCL1, BCL2-XL, BCL2) after 2, 5, 10 and 15 days of culture in control 2D cultures (2D) or in 3D cultures established according to the method of the invention in a collagen matrix (3D). (β-actin is used as a loading control. Data are representative of at least three independent experiments.

[0338] FIG. 4 is a group of histograms (A) and western-blots (B) showing the expression of factors implicated in the mesenchymal-to-epithelial transition (MET). (A) Histograms showing the mRNA expression of E-cadherin, N-cadherin, vimentin, cytokeratin 8 and cytokeratin 18 in PHH after 4 and 15 days of culture in control 2D cultures (white) or in 3D cultures established according to the method of the invention in a collagen matrix (black). The mRNA expressions correspond to a ratio with respect to the expression level after 4 days of culture in control 2D cultures. (B) Western-blots showing the expression of vimentin, N-cadherin, and E-cadherin in PHH after 2, 5, 10 and 15 days of culture in control 2D cultures (2D) or in 3D cultures established according to the method of the invention in a collagen matrix (3D). β-actin is used as a loading control.

[0339] FIG. 5 is a group of histograms showing the mRNA expression of: hepatocyte differentiation markers (A), drug-metabolizing enzymes of phase I (B), drug-metabolizing enzymes of phase II (C), drug transporters (D), CYP regulators (E), CYP1A2 after 3-MC induction for 24 h (5 μM) (F) and HNF4α (G) in PHH after 4, 15 and 28 days of culture (the latter for HNF4α only) in control 2D cultures (white) or in 3D cultures established according to the method of the invention in a collagen matrix (black). The mRNA expressions correspond to a ratio with respect to the expression level after 4 days of culture in control 2D cultures. The indicated values were obtained from one experiment representative of at least three independent experiments (mean ±SD, *p<0.05, **p<0.01, ***p<0.001). (H) Histograms showing CYP1A (EROD) activity after 3-MC induction for 24 h (5 μM) and basal CYP1A2 (MROD) activity in PHH after 4 and 15 days of culture in control 2D cultures (white) or in 3D cultures established according to the method of the invention in a collagen matrix (black).

[0340] FIG. 6 is a group of photographs obtained with fluorescence microscopy showing the localization of the drug transporters MRP2 and MRP3, of albumin and of nuclei in PHH spheroids after 4, 15 and 28 days in 3D cultures established according to the method of the invention in a collagen matrix. MRP2 functional activity is evaluated using the fluorescein diacetate CDFDA assay showing a clear efflux of the substrate in the central lumen of the PHH spheroids after 4, 15 and 28 days in 3D cultures established according to the method of the invention in a collagen matrix. Bar scale=25 μm.

[0341] FIG. 7A is a group of photographs obtained after HES staining showing PHH spheroids after 2, 4, 6, 10 and 15 days in 3D cultures established according to the method of the invention in a collagen matrix. Bar scale=25 μm. FIGS. 7B-C are a group of graphs showing the evolution of the mean diameter (um) of PHH spheroids as a function of culture time (B) and the number of PHH spheroids with a diameter<60 μm and the number of PHH spheroids with a diameter>60 μm as a function of culture time (C).

[0342] FIG. 8 is a group of graphs showing: (A) the proliferation over time of PHH in a 3D culture established according to the method of the invention in a collagen matrix (expressed as the percentage of nuclei that are both Ki67.sup.+ and Alb.sup.+), based on the quantification of the expression of the proliferation marker Ki67 and albumin in nuclei; (B) the mRNA expression of the proliferation markers Cdk2, CyclinD1, PCNA, P21, P27 in PHH after 2 and 4 days of culture in control 2D cultures (white) or in 3D cultures established according to the method of the invention in a collagen matrix (black). The mRNA expressions correspond to a ratio with respect to the expression level at 2 days of culture in control 2D cultures.

[0343] FIG. 9 is a group of graphs showing: (A) the quantification of the incorporation of BrdU in a 3D culture established according to the method of the invention in a collagen matrix; (B) the proliferation over time of PHH in a 3D culture established according to the method of the invention in a collagen matrix (expressed as the percentage of proliferating cells), based on the simultaneous quantification of the proliferation marker Ki67 (grey boxes) and of the incorporation of BrdU (black line); (C) the representation of the wave of proliferation between days 2 and 7 in eight 3D cultures established according to the method of the invention in a collagen matrix from hepatocytes of different donors as detected through Ki67 staining and BrdU incorporation as indicated.

[0344] FIG. 10 is a group of histograms showing the existence of two waves of proliferation in PHH in 3D culture established according to the method of the invention in a collagen matrix through the (A) quantification of the percentage of Ki67.sup.+/Alb.sup.+ nuclei over 13 days of 3D culture established according to the method of the invention in a collagen matrix; (B) quantification of the percentage of Cyclin D1.sup.+/Alb.sup.+ nuclei over 15 days of 3D culture established according to the method of the invention in a collagen matrix; (C) quantification of the BrdU incorporation over 15 days of 3D culture established according to the method of the invention in a collagen matrix.

[0345] FIG. 11A is a scheme illustrating the waves of proliferation occurring in the 3D culture established according to the method of the invention in a collagen matrix from hepatocytes of different donors during the first week of culture (first wave of proliferation: on average between day 3 and day 7) and during the second week of culture (second wave of proliferation: on average between day 9 and day 14). FIG. 11B is an illustration of an immunostaining of phospho-histone H3, a marker of cells in the M phase of the cell cycle, of albumin and of nuclei in a PHH spheroid after 10 days in 3D culture established according to the method of the invention in a collagen matrix. Bar scale=25 μm. FIG. 11C is a quantification of phospho-histone H3.sup.+/Alb.sup.+ nuclei over 15 days of 3D culture established according to the method of the invention in a collagen matrix.

[0346] FIG. 12 illustrates the proliferation observed in PHH after transient blockage of the MEK-ERK pathway. (A) Western-blot showing the inhibition of ERK phosphorylation after 14 days of culture in control 2D culture (2D) or in 3D culture established according to the method of the invention in a collagen matrix (3D) with addition of U0126 (+) or DMSO (−) between day 12 and day 14. (B) Proliferation between day 12 and day 18 of PHH in 3D culture established according to the method of the invention in a collagen matrix after MEK inhibition with U0126 between day 12 and day 14. Proliferation is assessed through the quantification of the proliferation marker Ki67 (expressed as the percentage of nuclei that are both Ki67.sup.+ and Alb.sup.+). A negative control was carried out using DMSO instead of U0126. (C) Proliferation between day 13 and day 21 of PHH in 3D culture established according to the method of the invention in a collagen matrix after MEK inhibition with U0126 between day 15 and day 17. Proliferation is assessed through the quantification of the proliferation marker cyclin D1 (expressed as the percentage of nuclei that are both cyclinD1.sup.+ and Alb.sup.+). A negative control was carried out using DMSO instead of U0126.

[0347] FIG. 13 is a histogram showing the DNA damage evaluated by the percentage of tail DNA in individual PHH after 9 days of incubation with the indicated genotoxic drugs at the indicated concentrations either in control 2D culture (2D) or in 3D culture established according to the method of the invention in a collagen matrix (3D). Control: no genotoxic drug; APB1 1: 0.1 μM; APB1 2: 0.5 μM; 4-ABP 1: 1 μM; 4-ABP 2: 10 μM.

[0348] FIG. 14 is a histogram showing the quantification of the number of yH2Ax-positive PHH assessed by immunostaining after 9 days of incubation with the indicated genotoxic drugs at the indicated concentrations in 3D culture established according to the method of the invention in a collagen matrix (3D). T: no genotoxic drug; CIS 1: 5 μg/mL cisplatin; CIS 2: 10 μg/mL cisplatin; AFB1 1: 0.1 μM AFB1; APB1 2: 0.5 μM; 4-ABP 1: 1 μM; 4-ABP 2: 10 μM.

[0349] FIG. 15 is a group of photographs is a group of photographs obtained after HES staining showing PHH spheroids after 5, 10 and 15 days in 3D cultures established according to the method of the invention in a methacrylated gelatin (GelMa) matrix. Bar scale=50 μm. Data are representative of three independent experiments.

[0350] FIG. 16 is a group of photographs obtained with fluorescence microscopy showing the localization of N cadherin, of albumin and of nuclei in PHH spheroids after 10 days in 3D cultures established according to the method of the invention in a methacrylated gelatin (GelMa) matrix. Bar scale=50 μm. Data are representative of three independent experiments.

[0351] FIG. 17 is a histogram showing the proliferation over time of PHH in a 3D culture established according to the method of the invention in a methacrylated gelatin (GelMa) matrix (expressed as the percentage of cells that are both cyclin D1.sup.+ and Alb.sup.+), based on the quantification of the expression of the proliferation marker cyclin D1 and albumin in PHH.

EXAMPLES

[0352] The present invention is further illustrated by the following examples.

Example 1

[0353] Materials and Methods

[0354] Material

[0355] Cells

[0356] Human liver samples were obtained from patients undergoing liver resection for primary hepatocellular carcinoma or liver metastases through the Centre de Ressources Biologiques (CRB)-Sante of Rennes (CHRU Pontchaillou, Rennes, France). The research protocol was conducted under French legal guidelines and the local institutional ethics committee. The clinical characteristics of the human liver samples are detailed in Table 1 below.

TABLE-US-00001 TABLE 1 clinical characteristics of the donors from whom the human liver samples were obtained Case Age Gender Liver pathology Smoker Diabetic HL-a 68 F Cholangiocarcinoma yes no HL-b 68 M Cholangiocarcinoma no no HL-c 66 F Liver metastasis from no no colorectal carcinoma HL-d 72 F Liver metastasis from no no colorectal carcinoma HL-e 56 M Cholangiocarcinoma no no HL-f 59 M Liver metastasis from no no colorectal carcinoma HL-g 60 F Liver metastasis from no no colorectal carcinoma HL-h 78 M Liver metastasis from no yes stromal tumor HL-i 75 M Hepatocellular no yes Carcinoma HL-j 60 M Liver metastasis from no no colorectal carcinoma

[0357] Human hepatocytes were isolated by a two-step collagenase perfusion procedure, and parenchymal cells were maintained in a modified William's E medium, called WE HH (human hepatocytes) comprising penicillin (100 U), streptomycin (100 μg/mL), insulin (15 μg/mL), glutamine (2 mM), albumin (0.1% (w/v)), transferrin (5.5 μg/mL), sodium selenite (5 μg/mL), hydrocortisone (1 μM), HGF (hepatocyte growth factor) (2.5 ng/mL), EGF (epidermal growth factor) (0.05 ng/μL) with or without FCS (fetal calf serum) (10% (v/v)).

[0358] Reagents

[0359] The cell proliferation reagent WST-1, Liberase (10 μg/mL) and colcemid (1 μM) were obtained from Roche. Cisplatin was obtained from Mylan. Type I collagen, ethoxyresorufin, methoxyresorufin, Hoechst, 5(6)-Carboxy-2′,7′-dichlorofluorescein diacetate (CDFDA, 10 μM), 1× ITS (insulin, transferrin, selenium) and anti β-actin antibody (A-5441, 1/1000) were obtained from Sigma-Aldrich. BrdU (RPN201V, 1:1000) and the anti-BrdU antibody (RPN202, 1:100) were obtained from GE Healthcare (Chicago, USA). The rhEGF was obtained from Promega (Fitchburg, USA) and the rhHGF from R&D Systems. The anti-HSC-70 antibody (sc-7298, 1:5000) was obtained from Santa-Cruz Biotechnology. The anti-Ki67 (MA5-14520, 1:400) antibody was obtained from Invitrogen. Antibodies against phospho-p44/42 MAPK (Thr202/Tyr204) (9106, 1:1000), E-cadherin (3195, 1:100), pro-apoptosis Bcl-2 family (1:500), and pro-survival Bcl-2 family (1:500) were obtained from Cell Signaling. The anti-N-cadherin antibody (610921, 1:100) was obtained from BD Biosciences. Antibodies against MRP2 (ab3373, 1:100) and Cyclin D1 (ab16663, 1:100) were obtained from Abcam. The anti-vimentin antibody (M0725, 1:1000) was from obtained Dako. The anti phospho-histone H3 (Ser10) antibody (06-570, 1:100) was obtained from Merck and the anti-albumin antibody (A80-229A, 1:100) was obtained from Bethyl Laboratories, Inc. The MEK inhibitor U0126 (50 μM) was obtained from Promega.

[0360] Methods

[0361] Primary Human Hepatocyte 2D Culture

[0362] As a control, aggregates of primary human hepatocytes (PHH) formed in low attachment plates as described below were seeded in standard multi-well plates and cultured in the same medium, i.e., the WE HH medium described hereinabove.

[0363] Primary Human Hepatocyte 3D Culture

[0364] As shown on FIG. 1, primary human hepatocytes (PHH) isolated as indicated above were first cultured in a low attachment plate (FIG. 1A). Then, the aggregates of PHH thus obtained were embedded in a collagen matrix (also referred to as collagen gel) and further cultured in the collagen matrix (FIG. 1B).

[0365] Culture in a low attachment plate: isolated human hepatocytes were incubated at 37° C., 5% CO2, humidity 85-95% during one night (about 15 to 20 h) in 6-well low attachment plates (Corning®, Costar®) at a concentration of 2×10.sup.6 to 2.5×10.sup.6 cells per well, in WE or WE HH. Inclusion in a collagen matrix followed by culture of hepatocytes embedded in the collagen matrix: type I collagen at 3 mg/mL or 6 mg/mL from Sigma-Aldrich was diluted into WE HH culture medium to obtain a collagen solution at the desired concentration. The human hepatocytes were added at a concentration of 3.65×10.sup.5 cells/mL to the WE HH comprising collagen and the pH was adjusted at 7,4 with NaOH 0.1 N. The resulting mix of human hepatocytes and DMEM HH comprising collagen was poured into 96-well plates (100 μl) or 48-well plates (300 μl) and incubated at 37° C., 5% CO2, humidity 85-95%. After at least 2 h, the gels were polymerized and a volume of WE HH medium equal to the volume of gel was added. The hepatocytes embedded in the gel matrix were cultured in the modified WE HH as described above for at least 2 days at 37° C., 5% CO2, humidity 85-95%. The inclusion of the hepatocytes in a collagen matrix thus allowed for a 3D culture of the hepatocytes.

[0366] Hereafter, in Example 1, the term “3D culture” refers to a culture of PHH established as described hereinabove, with a first incubation (or culture) of the PHH in a low attachment plate followed by the inclusion of the PHH in a collagen matrix, according to the method of the invention.

[0367] However, any mention of a time of culture, e.g., day 5 of culture, refers to the time of culture in a collagen matrix (or in a standard multi-well plate for the 2D control condition). In other words, any time of culture mentioned hereafter does not include the period of incubation (or culture) in a low attachment plate.

[0368] Collagen Gels Inclusion in Paraffin

[0369] After fixation in formol (also known as formaldehyde) 4%, collagens gels were dehydrated by successive incubations in alcohol and xylene baths at increasing concentrations before being impregnated with paraffin using EXCELSIOR ES tools (Thermo Fisher Scientific, Waltham, USA). After impregnation, gels were included in paraffin blocs and 4 μm cuts were made.

[0370] Hepatocyte DNA Synthesis

[0371] Incorporation of the thymidine analog BrdU was used as an index of cell proliferation. After 24 h of incorporation, cells in collagen gels were fixed in ethanol-glycine and included in paraffin as indicated above. BrdU positive cells were detected using an anti-BrdU antibody (RPN202, 1:100).

[0372] Immunohistochemistry

[0373] Immuno-histochemical staining was performed on the Discovery Automated IHC stainer using the Discovery Rhodamine kit (Ventana Medical Systems, Tucson, USA). Following deparaffination with Ventana EZ Prep solution at 75° C. for 8 min, antigen retrieval was performed using Tris-based buffer solution CC1 (Ventana Medical Systems, Tucson, USA) at 95° C. to 100° C. for 36 min. Endogen peroxidase was blocked with 3% H.sub.2O.sub.2 (Ventana Medical Systems, Tucson, USA) for 8 min at 37° C. After rinsing with reaction buffer (Roche, Basel, Switzerland), slides were incubated at 37° C. for 60 min with an appropriate dilution of the desired primary antibodies. After rinsing, signal enhancement was performed using the Ventana Rhodamine kit and incubation with the secondary antibody anti-Rabbit HRP (Roche, Basel, Switzerland) was carried out for 16 min After removal from the instrument, slides were manually rinsed, stained with albumin (1:100), with the secondary antibody for albumin detection (Donkey anti-goat 655, 1:250) and with Hoechst (1:1500), and coverslipped.

[0374] RT-qPCR Analysis

[0375] Cells were extracted from the collagen gels by the action of the Liberase enzyme blend 10 μg/mL (Roche, Basel, Switzerland) for 15 min at 37° C. Then, total RNA was extracted from cell pellets using NucleoSpin RNA (Macherey-Nagel, Hoerdt, France) and the concentration of total RNA was measured with a NanoDrop ND-1000 (NanoDrop Technologies, Wilmington, USA). Total RNA was used for cDNA synthesis using the High Capacity cDNA reverse transcription kit (Applied Biosystems, Saint Aubin, France). Real-time PCR for all genes was performed using SYBR green technology with the Power SYBR Green PCR master mix (Applied Biosystems, Saint Aubin, France) and the CFX384 Real-Time System (Biorad) according to the manufacturer's recommendations. Primer sequences used are described in Tables 2 and 3 below.

[0376] The amplification curves were analyzed with the Biorad CFX Manager software using the comparative regression method. GAPDH was used for the normalization of expression data. The relative amount of measured mRNA in samples was determined using the 2-ΔΔCT method where ΔΔCT=(CTtarget-CTGAPDH).sub.sample−(CTtarget-CTGAPDH).sub.calibrator. Final results were expressed as the n-fold differences in target gene expression in tested samples when compared with the mean expression value of calibrator. Values are given for one experiment representative of at least three independent experiments.

TABLE-US-00002 TABLE 2 forward primer sequences Gene Forward primer SEQ ID NO GAPDH GTGGACCTGACCTGCCGTCT SEQ ID NO: 1 E-CADHERIN TGCTCTTGCTGTTTCTTCGG SEQ ID NO: 2 N-CADHERIN GTGCATGAAGGACAGCCTCT SEQ ID NO: 3 VIMENTIN CCAAACTTTTCCTCCCTGAA SEQ ID NO: 4 CC CYTOKERATIN 8 AGCTGGAGTCTCGCCTGGAA SEQ ID NO: 5 CYTOKERATIN 18 GAGCCTGGAGACCGAGAAC SEQ ID NO: 6 α-FETOPROTEIN TGCAGCCAAAGTGAAGAGGG SEQ ID NO: 7 AAGA ALBUMIN GGGCATGTTTTTGTATGAAT SEQ ID NO: 8 ALDOLASE B AGAATGGACTGGTACCTATT SEQ ID NO: 9 G CYP1A2 TGGAGACCTTCCGACACTCC SEQ ID NO: 10 T CYP3A4 CTTCATCCAATGGACTGCAT SEQ ID NO: 11 AAAT CYP2E1 CCTGCTCGTGGAAATGGAGA SEQ ID NO: 12 GSTA1/2 CGCTACTTCCCTGCCTTTGA SEQ ID NO: 13 UGT1A1 TGACGCCTCGTTGTACATCA SEQ ID NO: 14 G NAT2 AGGAACAAATTGGACTTGGA SEQ ID NO: 15 MRP2 TGAGCAAGTTTGAAACGCAC SEQ ID NO: 16 AT MRP3 GTCCGCAGAATGGACTTGAT SEQ ID NO: 17 OCT1 TAATGGACCACATCGCTCAA SEQ ID NO: 18 CAR TGATCAGCTGCAAGAGGAGA SEQ ID NO: 19 PXR CCAGGACATACACCCCTTTG SEQ ID NO: 20 AhR CTTCCAAGCGGCATAGAGAC SEQ ID NO: 21 CDK2 CATTCCTCTTCCCCTCATCA SEQ ID NO: 22 CYCLIN D1 CCGTCCATGCGGAAGATC SEQ ID NO: 23 PCNA ATCATTACACTAAGGGCCGA SEQ ID NO: 24 AGATAAC P21 TGAGCCGCGACTGTGATG SEQ ID NO: 25 P27 CATTTGGTGGACCCAAAGAC SEQ ID NO: 26 HNF4a CAGAAGGCACCAACCTCAAC SEQ ID NO: 27

TABLE-US-00003 TABLE 3 reverse primer sequences Gene Reverse primer SEQ ID NO GAPDH GGAGGAGTGGGTGTCGCTGT SEQ ID NO: 28 E-CADHERIN TGCCCCATTCGTTCAAGTAG SEQ ID NO: 29 N-CADHERIN ATGCCATCTTCATCCACCTT SEQ ID NO: 30 VIMENTIN GTGATGCTGAGAAGTTTCGT SEQ ID NO: 31 TGA CYTOKERATIN 8 TGTGCCTTGACCTCAGCAAT SEQ ID NO: 32 G CYTOKERATIN 18 TTGCGAAGATCTGAGCCC SEQ ID NO: 33 α-FETOPROTEIN CATAGCGAGCAGCCCAAAGA SEQ ID NO: 34 AGAA ALBUMIN CCCACTTTTCCTAGGTTTCT SEQ ID NO: 35 ALDOLASE B GTAACATACTGGCAGTGTTC SEQ ID NO: 36 CYP1A2 CGTTGTGTCCCTTGTTGTGC SEQ ID NO: 37 CYP3A4 TCCCAAGTATAACACTCTAC SEQ ID NO: 38 ACAGACAA CYP2E1 TCTCTGTCCCCGCAAAGAAC SEQ ID NO: 39 GSTA1/2 AGTCAAGCTCCTCGACGTAG SEQ ID NO: 40 UGT1A1 CCTCCCTTTGGAATGGCAC SEQ ID NO: 41 NAT2 ACCCCGGTTTCTTCTTACAA SEQ ID NO: 42 MRP2 AGCTCTTCTCCTGCCGTCTC SEQ ID NO: 43 T MRP3 TCACCACTTGGGGATCATTT SEQ ID NO: 44 OCT1 AGCCCCTGATAGAGCACAGA SEQ ID NO: 45 CAR AGGCCTAGCAACTTCGCATA SEQ ID NO: 46 PXR CTACCTGTGATGCCGAACAA SEQ ID NO: 47 AhR AGTTATCCTGGCCTCCGTTT SEQ ID NO: 48 CDK2 TTTAAGGTCTCGGTGGAGGA SEQ ID NO: 49 CYCLIN D1 GAAGACCTCCTCCTCGCACT SEQ ID NO: 50 PCNA TCATTTCATAGTCTGAAACT SEQ ID NO: 51 TTCTCCTG P21 GTCTCGGTGACAAAGTCGAA SEQ ID NO: 52 GTT P27 AGAAGAATCGTCGGTTGCAG SEQ ID NO: 53 HNF4a CTCGAGGCACCGTAGTGTTT SEQ ID NO: 54

[0377] Immunoblotting Analysis

[0378] Cells were extracted from the collagen gels by the action of the Liberase enzyme blend 10 μg/mL (Roche, Basel, Switzerland) for 15 min at 37° C. Protein samples were extracted from cell pellets and dosed. Protein samples were then separated using SDS-PAGE and transferred onto nitrocellulose membranes in a transfer buffer (25 mM Tris, 200 mM glycine, Ethanol 20%). The blots were blocked with 5% low fat milk in Tris-buffer saline (TBS) (65 mM Tris pH 7.4, 150 mM NaCl) at room temperature for 1 h. The blots were incubated overnight with the desired primary antibodies at 4° C. The blots were washed with TBS and incubated for 1 h with a mouse-IgG HRP or a rabbit-IgG HRP secondary antibody (1:1000) in 5% low fat milk in TBS at room temperature. The blots were then washed with TBS. Immunocomplexes were visualized with a Fujifilm LAS-3000 imager (Fujifilm, Tokyo, Japan) after a chemiluminescent reaction using the Immobilon Western Chemiluminescent HRP substrate (Millipore, Merck, Darmstadt, Germany). Densitometric analyses of the bands were carried out with MultiGauge software (Fujifilm, Tokyo, Japan).

[0379] MRP2 Transporter Activity

[0380] A fluorescence-based efflux assay was used to investigate MRP2 transporter activity in 3D primary human hepatocyte (PHH) cultures. The membrane permeable and non-fluorescent substrate 5-(and-6)-carboxy-2′,7′-dichlorofluorescein diacetate (CDFDA) enters into hepatocytes where its hydrolysis by intracellular non-specific esterases results in a fluorescent product. This product, which is a substrate of the membrane transporter MRP-2, is effluxed from hepatocytes into the bile canaliculi.

[0381] The 3D cultures were incubated 10 min with CDFDA (10 μM). The dye-solution was then replaced by serum-free medium for 2 h. The cultures were washed twice with PBS solution before the fluorescence was assessed using microscopy.

[0382] CYP Activities Measurement

[0383] Ethoxyresorufin O-deethylation (EROD) and methoxyresorufin O-deethylation (MROD) associated with CYP 1A1/2 and 1A2 activities, respectively, were measured in cultured hepatocytes as described by Burke and Mayer (Burke and Mayer, 1983). Briefly, cells were washed with phosphate-buffer saline at 37° C. before being incubated with salicylamide (1.5 mM) to block phase II-conjugation enzymes. 7-Ethoxyresorufine or 7-Methoxyresorufine was added 1 min later and the oxidation of the substrates was measured by fluorescence detection every 2 min during 20 min at 37° C. The reaction rate was linear with time. After the assessment of P450 1A1/2 and 1A2 activities, the colorimetric WST-1 based assay was performed to normalize said activities with the quantity of viable cells. Values (pmol/min/OD) are correspond to the mean values±standard deviation of triplicate measurements.

[0384] Mitotic Index

[0385] To measure the mitotic index of PHH cultivated in 3D cultures established according to the method of the invention, a microtubule-depolymerizing drug, colcemid, was used to arrest cells in metaphase. Cells were treated during 24 h with colcemid before being fixed in formol 4%. The quantification of the mitotic index was performed by immunostaining of the phosphorylated histone H3.

[0386] Imaging

[0387] Observation and imaging of the cultured hepatocytes were carried out using two-photon excited fluorescence (TPEF) microscopy (also referred to as non-linear, multiphoton, or two-photon laser scanning microscopy). TPEF microscopy is an alternative to confocal and deconvolution microscopy that is particularly suited for deep and high-resolution three-dimensional imaging. TPEF enables observation of endogenous auto-fluorescent unstained or stained samples. In tandem with Second Harmonic Generation (SHG) which allows observation of hyper polarizable fibrillar proteins such as type 1 or type 3 collagen, these methods provide images of spatially resolved 3-dimensionnal structures of cells and collagen matrix. The SHG imaging system consisted of a confocal SP5 scanning head (Leica Microsystems, Mannheim, Germany) mounted on a DMI 6000 inverted microscope (Leica Microsystems) and equipped with a MAITAI femtosecond laser (Spectra Physics, Santa Clara, Calif.). A high NA water immersion objective (LUMFL 60 W×1.1 NA; Olympus, Tokyo, Japan) was used. The SHG signal was collected in the forward direction using a water immersion condenser (S1, NA=0.9; Leica Microsystems). A 405-20 bandpass filter was placed before the photomultiplier tube and oil immersion objective 10×/0.4 HC PL APO oil or 20×/0.7 HC PL APO oil.

[0388] Statistical Analysis

[0389] Results are expressed as mean values±standard deviation. Data were analyzed with two-tailed Student's t-test. Differences were considered significant when p<0.05 (*), p<0.01 (**), or p<0.001 (***). All experiments were performed at least three times.

[0390] Results

[0391] 3D culture in collagen matrix after pre-culture in LAP allows long term survival of adult primary human hepatocytes

[0392] As described hereinabove, primary human hepatocytes were first cultured in a low attachment plate for 15 h before being embedded in collagen gels in the presence of an appropriate growth factors/cytokines cocktail. As shown on FIGS. 2A-B, two days after inclusion in collagen gels (d2), primary human hepatocytes appeared as isolated cells or as small clusters of round cells without any membrane extension and without any particular organization (see FIG. 2B for 3D reconstruction). These clusters had very heterogeneous sizes and were not well organized. However, as shown on FIGS. 2C-D, seven days after inclusion in collagen gels (d7), the number of clusters of primary human hepatocytes was higher and said clusters were bigger, thus showing that the collagen matrix could influence cluster formation over culture time. As illustrated on

[0393] FIGS. 2E-F, TPEF imaging and Z stack reconstruction performed at day 7 showed that the clusters of primary human hepatocytes had organized into acini-like structures of different sizes with a hollow lumen. These acini-like structures, also referred to as hepatospheres or spheroids, were characterized by the presence of one layer of well-organized hepatocytes embedded in the collagen matrix. Fifteen days after inclusion in collagen gels (d15), SHG microscopy showed that, at a distance from the spheroids, collagen fibers appeared relaxed, with smooth randomly oriented nanofibers (FIG. 2G). These collagen fibers appeared more concentrated in the close microenvironment of the spheroids and also stretched, thereby constraining the spheroid volume (see collagen signature CS-1 on FIG. 2H). Remarkably, parallel bundles of collagen fibers were distributed perpendicular to close spheroids (see collagen signature CS-2 on FIG. 2H).

[0394] All these observations thus demonstrate that adult primary human hepatocytes can form acini-like structures, i.e., spheroids, when embedded in a collagen gel. This process appears to be dynamic and spheroids seem able to remodel the collagen matrix over culture time.

[0395] Then, long-term survival of the primary human hepatocytes (PHH) was evaluated by analyzing a panel of pro-apoptotic and pro-survival factors. As shown on FIG. 3, the expression over culture time of pro-apoptotic factors (phospho-BAD, BAX, BAK, BID, BIM, PUMA) (FIG. 3A) and pro-survival factors (MCL1, BCL2-XL, BCL2) (FIG. 3B) was assessed in PHH either in control 2D culture or in 3D culture established as described hereinabove. The expression of most pro-apoptotic proteins assessed appeared similar in PHH in control 2D culture and in PHH in 3D culture. Only P-BAD was less phosphorylated in PHH in 3D culture than in PHH in 2D culture. By contrast, the expression of pro-survival proteins decreased rapidly in PHH in 2D cultures whereas their expression was maintained, and for some actually up-regulated (BCL2-XL, BCL2), in PHH in 3D culture. The difference in pro-survival protein expressions began to be clearly seen at day 5 and increased thereafter until at least day 15 of culture. Moreover, no cleaved-caspase 3 could be detected in PHH in 3D culture but caspase-dependent apoptosis could be induced in PHH in 3D culture by cisplatin treatment (data not shown).

[0396] Taken all together, these data demonstrate that PHH in 3D culture do not undergo apoptosis.

[0397] 3D culture in collagen matrix after pre-culture in LAP allows stable differentiation of adult primary human hepatocytes

[0398] As shown on FIG. 4, analysis by RT-qPCR and western blotting showed that the long-term survival of the PHH in 3D culture was associated with an increase in the expression of epithelial markers (E-cadherin) and a decrease in the expression of mesenchymal markers (N-cadherin, vimentin, cytokeratin 8, cytokeratin 18). The expression of N-cadherin and vimentin was significantly lower in PHH in 3D cultures than in PHH in control 2D cultures, stimulated by the same growth factors/cytokines, both at the mRNA level (FIG. 4A) and at the protein level (FIG. 4B). Notably, the expression of N-cadherin increased rapidly in PHH in control 2D culture, from day 2 of culture and thereafter. E-cadherin at the mRNA level was highly expressed in PHH in 3D culture and this expression further increased over culture time. Moreover, immuno-detection in PHH spheroids showed a distinct localization of N- and E-cadherin at the apical, lateral and basal membranes. More precisely, increased E-cadherin labeling could be clearly seen between day 5 and 15 at the apical and lateral membranes, whereas N-cadherin was located preferentially at the lateral and basal membranes in agreement with their specific localizations in hepatocytes in a human liver (data not shown). Interestingly, MKL1/MRTF-A transcription factor (megakaryoblastic leukemia 1/myocardin-related transcription factor), which has been described to be force-mediated dependent (Willer et al., 2017; Flouriot et al. 2014), showed a nuclear localization only in PHH in 3D cultures. The transcription factor could not be detected in PHH in 2D cultures, neither in the nucleus nor in the cytoplasm (data not shown).

[0399] Taken all together, these data indicate that the protein expression in PHH in 3D culture established as described hereinabove is significantly different from that in PHH in control 2D culture and closely resembles that of in situ hepatocytes.

[0400] Adult primary human hepatocytes in 3D culture in collagen matrix after a pre-culture in LAP display hepatic functions

[0401] To ascertain the impact of 3D culture on human hepatocyte differentiation, analyses by RT-qPCR were carried out to investigate the mRNA expression levels of alpha-fetoprotein, albumin, aldolase B, and phase xenobiotic metabolism enzymes (also referred to as detoxifying enzymes). As shown on FIG. 5A, the mRNA expression of alpha-fetoprotein, albumin and aldolase B was significantly higher in PHH in 3D cultures than in PHH in control 2D culture, stimulated by the same growth factors/cytokines cocktail. As shown on FIGS. 5B-D, phase I (CYP1A2, CYP3A4, CYP2E1), phase II (GSTA1/2, UGT1A1, NAT2) and phase III (MRP2, OCT1) detoxifying enzymes were also highly expressed in PHH in 3D culture.

[0402] RT-qPCR analysis also showed a strong induction of the expression of the transcription factors involved in the regulation of detoxifying pathways (CAR, PXR and AhR) in the PHH in 3D culture as compared to that in PHH in control 2D cultures (FIG. 5E). CYP1A2 (MROD) activity was significantly increased in PHH in 3D culture and, interestingly, a potentialization of 3-methylcholanthrene (3-MC) induction of CYP1A (EROD) activity was only observed in PHH in 3D culture (FIG. 5H). Said potentialization demonstrates the high detoxification capacities of primary human hepatocytes embedded in collagen matrix according to the method described hereinabove. The marked 3-MC induction of CYP1A activities in PHH in 3D culture was confirmed with the observation of a significantly increased CYP1A2 mRNA expression in PHH in 3D culture as compared to that in PHH in control 2D culture, stimulated by the same growth factors/cytokines cocktail (FIG. 5F). The mRNA expression of the nuclear receptor HNF4 alpha was also higher in PHH in 3D culture than that in PHH in 2D culture at days 4, 15 and 28 (FIG. 5G).

[0403] As shown on FIG. 6, drug-transporter MRP2 localized exclusively at the apico-lateral region and the apical/bile canalicular domain. Moreover, evaluation of MRP2 functional activity using the fluorescein diacetate CDFDA assay showed a clear efflux in the central lumen of the PHH spheroids at days 4, 10, 28. Drug-transporter MRP3 localized on both apical and lateral domains.

[0404] Formation of PHH spheroids in 3D culture in collagen matrix after a pre-culture in LAP

[0405] The size and morphology of the hepato-spheroids were assessed over two weeks of culture (FIG. 7). The cells were observed after hematoxylin, eosin and saffron (HES) staining and the mean diameter of the spheroids quantified. Over the two weeks of culture, all the spheroids showed only one layer of cells and displayed an acinus-like structure, with a hollow lumen (see FIGS. 6 and 7A).

[0406] As indicated above, the PHH forming the spheroids were polarized, as could be observed notably with the exclusive localization of the drug-transporter MRP2 at the apico-lateral region and the apical/bile canalicular domain Moreover, as indicated above, the PHH forming the spheroids retained their differentiated state and their hepatic functions.

[0407] Expression of hypoxia and/or apoptosis markers was not detected in the spheroids obtained with the method of the invention. Moreover, the size of the spheroids did not decrease over time during the 3D culture. On the contrary, as shown on FIGS. 7A-B, the size of the spheroids gradually increased over time.

[0408] Proliferation of adult primary human hepatocytes in 3D culture in collagen matrix after a pre-culture in LAP

[0409] As mentioned above, the size of the spheroids gradually increased over time. Notably, the number of hepatocyte spheroids with a diameter inferior to 60 μm decreased whereas the number of spheroids with a diameter superior to 60 μm progressively increased (FIG. 7C). As shown in Table 4 below, the size quantifications indicated an increase of the mean diameter of spheroids from 47.62 μm±1.61 μm at day 2 to 72.04 μm±6.92 μm at day 10 corresponding to a 3-fold increase of the spheroid volume, approximatively. At day 15, the spheroid size stabilized or decreased slightly, staying constant thereafter (Table 4 and results not shown).

TABLE-US-00004 TABLE 4 quantification of the size of PHH spheroids in 3 D cultures Time from the start Mean diameter of the spheroids of the 3 D culture (μm) ± standard deviation Day 2 47.62 ± 1.61 Day 3 51.24 ± 3.30 Day 4 53.45 ± 8.65 Day 5 58.36 ± 3.80 Day 6 64.83 ± 1.84 Day 7 68.83 ± 4.90 Day 8 66.72 ± 3.52 Day 9 71.31 ± 6.12 Day 10 72.04 ± 6.92 Day 15 65.47 ± 7.33

[0410] Next, the number of Ki67 and BrdU positive cells was quantified over the first 8 days of culture. Both Ki67 and BrdU are makers of cell proliferation, and Ki67 and BrdU positive cells are cells undergoing S phase. Hepatocytes were isolated from 8 different patients (see Table 1) and Ki67 staining or BrdU labeling were analyzed both in 3D and 2D cultures. Albumin, a marker of mature hepatocyte function, was detected as a positive control to quantify only Alb.sup.+ hepatocytes. As shown on FIGS. 8A and 9A-B, the high numbers of Ki67.sup.+/Alb.sup.+ and/or

[0411] BrdU±/Alb+hepatocytes observed between days 2 and 7 demonstrated the presence of proliferative primary human hepatocytes in the 3D cultures established as described hereinabove. No Ki67.sup.+ hepatocytes could be detected in the control 2D cultures, stimulated by the same growth factors/cytokines cocktail. RT-qPCR analysis of the expressions of the cell cycle markers Cdk2, cyclin D1, PCNA, P21 and P27 showed that said cell cycle markers were highest expressed in 3D cultures at day 4 as compared to 3D day 2 and to 2D days 2 and 4 (FIG. 8B). In summary, the analysis of 3D cultures of primary human hepatocytes from 8 different patients gave highly similar results (FIG. 9C), demonstrating the presence of a high number of Ki67.sup.+/Alb.sup.+ and BrdU.sup.+/Alb.sup.+ hepatocytes with only slight variations in the kinetics of the marker expression. By contrast, the Alb+hepatocytes in control 2D cultures were always Ki67.sup.− and BrdU.sup.− (data not shown). A mean of the cumulative/additive index of Ki67.sup.+/Alb.sup.+ hepatocytes from all the experiments reached 280% (+/−60%) indicating that primary human hepatocytes could undergo at least two cell cycles during the first week of the 3D culture established as described hereinabove.

[0412] The results clearly demonstrate for the first time that 3D cultures in collagen gels established as described hereinabove provide a microenvironment enabling adult primary human hepatocytes to proliferate, in particular in the presence of a growth factor cocktail. By contrast, PHH in control 2D cultures did not proliferate, even in the presence of the same growth factor cocktail.

[0413] Analyses of Ki67 (FIG. 10A) and cyclin D1 (FIG. 10B) expressions and of BrdU incorporation (FIG. 10C) were performed during the first fifteen days after seeding in 3D cultures of primary human hepatocytes established as described hereinabove from seven different donors. As shown in FIG. 11A, primary human hepatocytes in said 3D cultures could undergo a second wave of proliferation during the second week of culture, between day 8 and 15 depending on the donor. These two successive waves of proliferation were also detected by a phospho-histone H3 immunostaining (FIG. 11B) allowing the quantification of the proportion of cells blocked in the M phase of the cell cycle after 24 h of treatment with colcemid. Cumulative mitotic index (FIG. 11C) showed that about 300% of cells are in M phase during the first week of culture and about 200% during the second one. Thereafter, no more Ki67, cyclin D1 and BrdU positive cells could be detected (results not shown).

[0414] Transient MEK/ERK pathway inhibition induces a new wave of proliferation

[0415] As described hereinabove, analysis of all the 3D cultures established according to the method of the invention from hepatocytes isolated from different donors (HL-a to HL-j, see Table 1) demonstrated the proliferation capability of adult human hepatocytes between day 2 and 15, with small variations in the proliferation response depending on the donor. No further proliferation could be detected after two weeks of culture. It was previously shown that transient activation of the MEK/ERK pathway stimulates DNA replication in rodent hepatocytes, and that inhibition of the MEK/ERK pathway blocks rodent hepatocyte proliferation. It was also shown that sustained activation of the MEK/ERK pathway can have a negative role on rat hepatocyte proliferation (Fremin et al., 2009) and that overactivation of the cascade can also inhibit cell replication in the human hepatocarcinoma cell line HuH-7 (Guegan et al., 2015). Proliferation of the primary human hepatocytes was assessed after the MAPK MEK1/2-ERK1/2 pathway was transiently inhibited for 48 h with the MEK inhibitor U0126 in 3D cultures established as described hereinabove at day 12 or 15. Control of ERK1/2 phosphorylation confirmed a strong inhibition of ERK1/2 activity at the end of the U0126 treatment in all culture conditions (FIG. 12A). Following removal of the MEK inhibitor, cumulative Ki67 (FIG. 12B) and cyclin D1 (FIG. 12C) positive cells indicated that 50% to 70% of the cells could undergo a new cell cycle (Ki67.sup.+/Alb.sup.+ and cyclinD1.sup.+/Alb.sup.+). No positive cells could be detected in DMSO control experiments or in the control 2D cultures after a similar transient MEK/ERK inhibition (data not shown).

Example 2

[0416] Materials and Methods

[0417] Material

[0418] Cells

[0419] Primary human hepatocytes (PHH) were obtained as described hereinabove.

[0420] Methods

[0421] Primary Human Hepatocyte Culture

[0422] Primary human hepatocytes 3D cultures were set up as described hereinabove (see Example 1). Briefly, PHH were first cultured in a low attachment plate. Then, the aggregates of PHH thus obtained were embedded in a collagen matrix and further cultured in the collagen matrix.

[0423] Culture Medium

[0424] The PHH were cultured in WE HH medium, defined hereinabove (see Example 1).

[0425] Immunohistochemistry

[0426] Immuno-histochemical staining was performed as described hereinabove to assess the proliferation of the PHH.

[0427] Results

[0428] The importance of the growth factor stimulation for the proliferation of PHH in the 3D culture according to the invention was further studied. PHH in collagen matrix were cultured as described hereinabove in WE HH medium depleted of EGF, HGF, and/or ITS (insulin, transferrin and sodium selenite). PHH in collagen matrix were also cultured as described hereinabove in WE HH medium depleted of FCS. As shown in Table 5 below, the proliferation of PHH cultured in the different culture media was assessed by quantification of the number of Cyclin D1.sup.+/Alb.sup.+ cells with respect to the percentage of positive cells in the culture control (WE HH medium) and related to the number of cells detected.

TABLE-US-00005 TABLE 5 Percentage of proliferating PHH in different culture media with respect to the control culture condition (WE HH medium) Culture conditions % of proliferation Control (WE HH medium) 100 Hydro N 99.9 −EGF 73.2 −HGF 86.8 −ITS 70.2 −EGF/−HGF/−ITS 66.4 −FCS 58 Data are from one experiment representative of three.

[0429] When grown in WE HH medium comprising hydrocortisone at a concentration of 50 μM (“Hydro N”), i.e., a concentration 50-fold higher than that in the control condition, PPH displayed a percentage of proliferation similar to that observed when they were grown in the control condition. Single depletion of each of the growth factor present in the cocktail (“-EGF”, “-HGF” or “-ITS”) only resulted in a partial decrease in PHH proliferation (at most a 30% decrease). Moreover, neither the concomitant depletion of EGF, HGF and ITS (“-EGF/-HGF/-ITS”) nor the single depletion of fetal calf serum (“-FCS”) abolished the proliferation of PHH. These results thus demonstrate that the PHH can proliferate in 3D cultures according to the invention even in absence of EGF, in the absence of HGF, and/or in the absence of ITS. Strikingly, the observed results also demonstrate that the PHH can proliferate in 3D cultures according to the invention even in absence of fetal calf serum (FCS).

[0430] The effect of collagen stiffness on the proliferation of PHH was also assessed. As shown in Table 6 below, the proliferation of PHH cultured in the indicated collagen matrices was assessed by quantification of the number of Cyclin D1.sup.+/Alb.sup.+ cells with respect to the percentage of positive cells in the collagen matrix control (1.5 mg/mL) and related to the number of cells detected. Increasing type I collagen concentration from 1.5 mg/mL to 3 mg/mL or 4 mg/mL induced a decrease in PHH proliferation of about 50% and 80%, respectively. By contrast, decreasing type I collagen concentration from 1.5 mg/mL to 0.75 mg/mL did not induce a decrease in PHH proliferation. These results thus demonstrate that increasing the collagen concentration of the 3D matrix, and thus the matrix stiffness, inhibits the proliferation of hepatocytes in a concentration-dependent manner

TABLE-US-00006 TABLE 6 Percentage of proliferating PHH in different culture conditions with respect to the control culture condition (1.5 mg/mL collagen) % of Culture conditions proliferation 1.5 mg/mL collagen 100 0.75 mg/mL collagen 100 3 mg/mL collagen 49.9 4 mg/mL collagen 17.6 No incubation in a LAP 16.9 Incubation in a LAP not followed 11.7 by 3 D culture in collagen matrix

[0431] Strikingly, PHH embedded directly in 3D collagen gels and cultured in WE HH medium without a prior incubation in a low attachment plate displayed a very low rate of proliferation (16.9%, see Table 6). Similarly, PHH cultured in a low attachment plate in WE HH medium without ever being embedded in a 3D collagen matrix displayed a very low rate of proliferation (11.7%, see Table 6).

[0432] Taken all together, these results demonstrate that the 3D microenvironment and controlled collagen stiffness are major factors for inducing human adult primary hepatocyte proliferation. Moreover, these results show that both a first incubation (or culture) of the PHH in a non-adherent culture vessel, e.g., a low attachment plate, and the subsequent inclusion of the PHH in a collagen matrix are required to induce the proliferation of adult PHH.

Example 3

[0433] Materials and Methods

[0434] Material

[0435] Cells

[0436] Primary human hepatocytes (PHH) were obtained as described hereinabove (see Example 1).

[0437] Methods

[0438] Primary Human Hepatocyte Culture

[0439] Primary human hepatocytes 2D and 3D cultures in a collagen matrix were set up as described hereinabove (see Example 1). PHH were cultured in WE HH medium as described hereinabove.

[0440] Comet Assay

[0441] PHH either in control 2D culture (2D) or in 3D culture established according to the method of the invention in a collagen matrix (3D) were incubated with the indicated genotoxic drugs at the indicated concentrations (see Table 7) from the start of the culture. After 9 days, PHH were extracted from the collagen matrix by the action of purified collagenase. The cell pellets were resuspended in 0.5% low-melting point agarose and laid on conventional microscope slides covered with regular agarose. The electrophoresis migration was processed and after staining with propidium iodide, at least 100 images were acquired using a fluorescence microscope. The images were analyzed using the Comet Assay IV software. The extent of DNA damage in individual cells was evaluated by the percentage of tail DNA.

TABLE-US-00007 TABLE 7 genotoxic drugs Concentration 1 Concentration 2 Cisplatin (CIS) 5 μg/mL 10 μg/mL AFB1 0.1 μM 0.5 μM 4-ABP 1 μM 10 μM

[0442] γH2Ax Immunostaining after Treatment of PHH

[0443] PHH in 3D culture in a collagen matrix established according to the method of the invention (3D) were incubated with the indicated genotoxic drugs at the indicated concentrations (see Table 7) from the start of the culture. After 9 days, immuno-histochemical staining was performed as described hereinabove to detect the presence of phosphorylated H2AX (i.e., γH2Ax), a marker of DNA double-strand break, in the PHH.

[0444] Results

[0445] To assess the sensitivity of the spheroids comprising proliferative PHH as a model for testing drug-induced genotoxicity, the effects of two classes of compounds on said spheroids were studied: alkylating agents, i.e., cisplatin (phosphorylated histone γH2Ax assay, see FIG. 14), that exert direct effect on DNA without being metabolized; 2 proven carcinogens, i.e., 4-ABP (4-aminobiphenyl) and AFB-1 (aflatoxin B1) that become genotoxic only after metabolic activation (FIGS. 13 and 14). The primary human hepatocytes were incubated for 9 days with the genotoxic drugs from the start of the 3D culture, during a period when they proliferate. To compare the induced genotoxicity of these compounds in control 2D culture (2D) or in 3D culture established according to the method of the invention in a collagen matrix (3D), comet assay was performed. In 2D cultures, 4-ABP and AFB1 caused only very low levels of damages (about 3%). By contrast, in 3D culture established according to the method of the invention in a collagen matrix, AFB-1 caused high and dose-dependent levels of DNA damages (up to about 25%) and 4-ABP 1 caused higher (about the double) and dose-dependent levels of DNA damages (FIG. 13).

[0446] To confirm these results, the number of yH2Ax positive cells in 3D cultures established according to the method of the invention in a collagen matrix was quantified by immunostaining. Very high levels of DNA damages were thus observed with the 3 tested drugs: cisplatin induced DNA damages in about 90% of the PHH, 4-ABP in about 50% of the PHH, and AFB1 in about 35% of the PHH (FIG. 14).

[0447] These results thus demonstrate that, with their highly differentiated and proliferating status, the spheroid comprising proliferative PHH constitute a relevant model for testing the genotoxicity of environmental and chemical compounds.

Example 4

[0448] Materials and Methods

[0449] Material

[0450] Cells

[0451] Primary human hepatocytes (PHH) were obtained as described hereinabove (see Example 1).

[0452] Reagents

[0453] Methacrylated gelatin was obtained from the ART Bio-encres (Accelerateur de Recherches Technologiques de l'Inserm, Bordeaux). Lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP) was obtained from TCI.

[0454] Methods

[0455] Primary Human Hepatocyte Culture in Gelatin Methacrylate (GelMa)

[0456] In a manner similar to that shown on FIG. 1, primary human hepatocytes (PHH) isolated as indicated above (see Example 1) were first cultured in a low attachment plate (FIG. 1A). Then, the aggregates of PHH thus obtained were embedded in a GelMa matrix (also referred to as GelMa hydrogel) and further cultured in the GelMa matrix (FIG. 1B).

[0457] Culture in a low attachment plate: as indicated above (see Example 1), isolated human hepatocytes were incubated at 37° C., 5% CO2, humidity 100% during two days (about 48 h) in 6-well low attachment plates (Corning®, Costar®) at a concentration of about 3×10.sup.6 cells well, in a modified William's E medium comprising gentamicin (50 μg/mL), penicillin (100 U/mL), streptomycin (100 μg/mL), insulin (5 μg/mL), L-glutamine (2 mM), albumin (0.1% (w/v)) and with or without FCS (fetal calf serum) (10% (v/v)).

[0458] Preparation of medium comprising 5% GelMa: small crushed fragments of freeze-dried GelMa (up to 2 mm in length) were dissolved at 37° C. for a minimum of 8 h in a modified William's E medium comprising gentamicin (50 μg/mL), penicillin (100 U/mL), streptomycin (100 mg/mL), insulin (15 μg/mL), L-glutamine (2 mM), albumin (0.1% (w/v)), transferrin (5.5 μg/mL), sodium selenite (5 μg/mL), hydrocortisone (1 μM), HGF (hepatocyte growth factor) (2.5 ng/mL), EGF (epidermal growth factor) (0.05 ng/μL) and with or without FCS (fetal calf serum) (10% (v/v)). A volume of lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP) at a concentration of 10 mg/mL was added to obtain a modified William's E medium comprising concentration of GelMa of 5 g/100 mL, i.e., a final concentration of 5% GelMa and a final concentration of lithium phenyl-2,4,6-trimethylbenzoylphosphinate of 100 mg/mL, i.e., a final concentration of 0.1% LAP. Before use, the medium comprising 5% GelMa and 0.1% LAP was kept at 37° C., protected from light.

[0459] Inclusion in a GelMA matrix followed by culture of hepatocytes embedded in the collagen matrix: human hepatocytes were transferred to a suitable container such as an Eppendorf tube (1.5 or 2 mL) or a Falcon tube (15 mL) and centrifuged at 200 g for 2 min. The supernatant was removed and medium comprising 5% GelMa and 0.1% LAP was added to obtain a concentration of about 10.sup.6 human hepatocytes per mL of GelMa matrix, i.e., of medium comprising 5% GelMa and 0.1% LAP. 100 μL of medium comprising 5% GelMa, 0.1% LAP and human hepatocytes were added to the wells of a 96-well plate. Alternatively, 300 μL of medium comprising 5% GelMa, 0.1% LAP and human hepatocytes were added to the wells of a 48-well plate. Polymerization of the GelMa matrix was induced by illumination for 30 seconds to 10 min with a 365, 405 nm or 530 nm LED, preferably by illumination for 60 seconds with a 405 nm LED. After polymerization of the GelMa matrix, a volume of the same modified William's E medium equal to the volume of GelMa matrix was added. The hepatocytes embedded in the GelMa matrix were cultured in the modified William's E medium as described above for at least 2 days at 37° C., 5% CO2, humidity 100%. The medium was changed every 48 h.

[0460] Results

[0461] Formation of PHH spheroids in 3D culture in GelMa matrix after a pre-culture in LAP

[0462] The size and morphology of the hepato-spheroids was assessed over two weeks of culture (FIG. 15). The cells were observed after hematoxylin, eosin and saffron (HES) staining. Over the two weeks of culture, all the spheroids showed only one layer of cells and displayed an acinus-like structure, with a hollow lumen (see FIGS. 15 and 16).

[0463] As with the PHH spheroids cultured in a collagen matrix, the PHH forming the spheroids cultured in a GelMa matrix were polarized, as could be observed notably with the exclusive localization of the drug-transporter MRP2 at the apico-lateral region and the apical/bile canalicular domain.

[0464] Proliferation of adult primary human hepatocytes in GelMa matrix after a pre-culture in LAP

[0465] Proliferation of the primary human hepatocytes in GelMa matrix was assessed through the analyses of cyclin D1 and albumin (Alb) expression over 30 days after seeding in 3D cultures established according to the method of the invention in a GelMa matrix (FIG. 17).

[0466] Alb.sup.+/cyclin D1.sup.+ PHH could be detected from day 2, demonstrating that PHH were able to proliferate in 3D cultures established according to the method of the invention in a GelMa matrix. As shown on FIG. 17, during the wave of proliferation, a maximum level of proliferation of about 60% was observed.