Accepted Manuscript Novel human hepatic organoid model enables testing of drug-induced liver fibrosis in vitro Dr. Sofia B. Leite, Tiffany Roosens, Adil El Taghdouini, Inge Mannaerts, Ayla J. Smout, Mustapha Najimi, Etienne Sokal, Fozia Noor, Christophe Chesne, Leo A. van Grunsven, Prof. PII:
S0142-9612(15)00922-9
DOI:
10.1016/j.biomaterials.2015.11.026
Reference:
JBMT 17203
To appear in:
Biomaterials
Received Date: 24 September 2015 Revised Date:
30 October 2015
Accepted Date: 13 November 2015
Please cite this article as: Leite SB, Roosens T, El Taghdouini A, Mannaerts I, Smout AJ, Najimi M, Sokal E, Noor F, Chesne C, van Grunsven LA, Novel human hepatic organoid model enables testing of drug-induced liver fibrosis in vitro, Biomaterials (2015), doi: 10.1016/j.biomaterials.2015.11.026. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
1
Novel human hepatic organoid model enables testing of drug-induced liver
2
fibrosis in vitro
3 4
Sofia B. Leite , Tiffany Roosens , Adil El Taghdouini , Inge Mannaerts , Ayla J. Smout , Mustapha Najimi , 2 3 4 1* Etienne Sokal , Fozia Noor , Christophe Chesne , Leo A. van Grunsven
1
1
1
5
1
6
1
7 8
2
9
3
10
4
11 12 13
*Correspondence to: Prof. Leo van Grunsven (
[email protected]) or Dr. Sofia Batista Leite (
[email protected]). Faculty of Medicine and Pharmacy, Laarbeeklaan 103, 1090 Brussel, Building D Ground Floor Room D022/025B. TEL (+32) 2 477 44 09;
Liver Cell Biology Laboratory, Vrije Universiteit Brussel (VUB), Belgium
Laboratory of Pediatric Hepatology and Cell Therapy, Institute of Experimental and Clinical Research (IREC), Université Catholique de Louvain (UCL), Belgium
SC
Biochemical Engineering Institute, Saarland University, Germany
M AN U
Biopredic International, France
14
EP
TE D
List of abbreviations (by order of appearance): HSC, Hepatic Stellate Cell; Hep, HepaRG; 3D, threedimensional; Hep/HSC, Hep and HSC co-culture, hepatic organoid; PDGFR-β, Platelet derived growth factor receptor β; CYP, cytochrome P450; RIF, Rifampicin; BNF, β-Naphtoflavone; PB, Phenobarbital; MRP2, Multidrug resistance-associated protein 2; CMFDA, 5-chloromethylfluorescein diacetate; 2D, twodimensional; mRNA, messenger Ribonucleic Acid; dCT, deltaCT, gene expression relative to housekeeping gene; TGFβ, Transforming Growth Factor β; LPS, Lipopolysaccharide; APAP, Acetaminophen; ATP, Adenosine Triphosphate; αSMA, alpha Smooth Muscle Actin; HDAC, Histone deacetylase; CK, Cytokine mixture; VPA, Valproic Acid; CCl4, Carbon Tetrachloride; EC50, Half Maximal Effective Concentration; OCR, Oxygen consumption rate; ECAR, Extracellular acidification; Oligo, Oligomycin; FCCP, Trifluoromethoxy carbonylcyanide phenylhydrazone; A+R, Antimycin A and Rotenone; MTX, Methotrexate; ALT, Alanine Aminotransferase; AOP, Adverse Outcome Pathway; DILI, Drug-induced liver injury.
AC C
15 16 17 18 19 20 21 22 23 24 25 26
2
RI PT
1
1
ACCEPTED MANUSCRIPT
ABSTRACT
28
Current models for in vitro fibrosis consist of simple mono-layer cultures of rodent hepatic
29
stellate cells (HSC), ignoring the role of hepatocyte injury. We aimed to develop a method
30
allowing the detection of hepatocyte-mediated and drug-induced liver fibrosis. We used HepaRG
31
(Hep) and primary human HSCs cultured as 3D spheroids in 96-well plates. These resulting
32
scaffold-free organoids were characterized for CYP induction, albumin secretion, and hepatocyte
33
and HSC-specific gene expression by qPCR. The metabolic competence of the organoid over 21
34
days allows activation of HSCs in the organoid in a drug- and hepatocyte-dependent manner.
35
After a single dose or repeated exposure for 14 days to the pro-fibrotic compounds Allyl alcohol
36
and Methotrexate, hepatic organoids display fibrotic features such as HSC activation, collagen
37
secretion and deposition. Acetaminophen was identified by these organoids as an inducer of
38
hepatotoxic-mediated HSC activation which was confirmed in vivo in mice. This novel hepatic
39
organoid culture model is the first that can detect hepatocyte-dependent and compound-induced
40
HSC activation, thereby representing an important step forward towards in vitro compound
41
testing for drug-induced liver fibrosis.
42 43
46 47
Keywords: organoid; liver fibrosis; DILI; in vitro; hepatic stellate cell; Hepatocyte; HepaRG; APAP; Methotrexate; Allyl alcohol
AC C
44 45
EP
TE D
M AN U
SC
RI PT
27
2
ACCEPTED MANUSCRIPT
INTRODUCTION
49
Hepatic stellate cells (HSCs) are the major collagen producing cells during conditions of
50
sustained hepatic injury (either metabolic, cholestatic, viral or toxic), when hepatocyte damage
51
triggers a cascade of events leading to activation of quiescent HSCs into a myofibroblastic
52
(activated) HSC state1. HSC activation is mediated by a plethora of pathways that finally result in
53
increased secretion of extracellular matrix proteins, such as collagens, that accumulate as scar
54
tissue (fibrosis) within the liver parenchyma and to liver cirrhosis in a later stage2. To date, the
55
best in vitro fibrosis models consist of mono-layer cultures of freshly isolated rodent HSCs in
56
regular tissue culture dishes which leads to “spontaneous” HSC activation3. Obvious limitations
57
of these cultures are the rodent background and the un-controlled and hepatocyte damage-
58
independent activation of the HSCs, making these cultures less suitable for pro- and anti-fibrotic
59
compound testing translatable to human.
60
We developed a novel three-dimensional (3D) human co-culture model where both hepatocyte
61
functionality and HSC quiescence can be maintained for at least 21 days. This novel system
62
allows hepatotoxicity testing as well as drug-provoked and hepatocyte-dependent HSC
63
activation and fibrosis.
TE D
EP
64
M AN U
SC
RI PT
48
MATERIAL AND METHODS
66
Cell culture
67
HepaRG/HSC cell culture. Human liver non-parenchymal fractions isolation were obtained from
68
the Agreed Hepatocytes & Hepatic Stem cell Bank (Saint-Luc Hospital and Université Catholique
69
de Louvain). This raw material (obtained after written and signed informed consent) was
70
processed for Hepatic Stellate Cell (HSC) isolation and prepared for culture as previously
71
described4-6. Cells from two different donors were used, both male, healthy and under 13 years
72
old. For a thorough characterization of the cells we refer to El Taghdouini et. al.7 and Coll & El
AC C
65
3
ACCEPTED MANUSCRIPT
Taghdouini et. al.4. For the 3D cultures cells were used between passages 5 and 10.
74
Differentiated cryopreserved HepaRG® cells were obtained from Biopredic. On day 0 of culture,
75
HSCs were trypsinized and HepaRG (Hep) thawed according to the instructions of the provider.
76
For the generation of the cell spheroids, 96-well plates treated with cell-repellent (Greiner) were
77
used. Mono- and co-culture suspensions were prepared and seeded in the cell densities shown
78
below:
Hep seeding 5 (x10 cells/ml)
3D Hep
-
2.00
3D HSC
2.00
-
3D Hep/HSC
1.33
0.67
M AN U
Table 1: 3D cell seeding HSC seeding 5 (x10 cells/ml) Type of culture
SC
RI PT
73
After 15-30 min in the incubator, plates were placed on an orbital shaker, where stirring was kept
80
at 80 rpm for the entire culture period. The outer wells were filled with 100 µl of PBS to minimize
81
liquid evaporation. Culture medium was refreshed (90%) every second day. 3D Hep/HSC co-
82
cultures and control mono-cultures were cultured for 21 days. Cross diameter of the spheroids
83
never exceeded 200 µm.
84
HepaRG/HSC culture medium. Day 0 cells were resuspended in HepaRG thawing medium and
85
from day 1 to 21 cells were kept in the HepaRG culture medium 0% DMSO (Biopredic), unless
86
otherwise stated. In the APAP proof-of-concept studies, medium was replaced by HepaRG
87
Induction/Toxicity medium with 0.1% (v/v) DMSO (Sigma-Aldrich, St. Louis, MO, USA) during
88
compound incubation. In the repeated versus single compound exposure assays, on day 8, the
89
HepaRG culture medium 0% DMSO was replaced by the Serum-free HepaRG maintenance
90
medium described in Klein et. al. 20138 and kept in this medium until the end of culture
91
The remaining material and methods are available in supplementary information.
AC C
EP
TE D
79
4
ACCEPTED MANUSCRIPT
92 RESULTS
94
Characterization of human liver organoid cultures
95
We use cryopreserved differentiated HepaRGs (Heps) as functional hepatocyte-like cells and
96
primary human HSCs for the development of the human 3D hepatic organoids. Differentiated
97
HepaRG, widely characterized in the last decade, show a close correlation with human
98
hepatocyte functions9 which are improved when cultured in 3D10, 11. HSCs are isolated from the
99
non-parenchymal fraction of human livers, expanded in culture and frozen down as previously
100
described4, 6. Liver organoids are generated by mixing the two single-cell suspensions in a ratio
101
of 1Heps:2HSCs in non-attachment round-bottom 96 well-plate wells with orbital stirring. This
102
allows for quick generation of spheroids with an optimal diameter of ≤200 µm, representing the
103
maximum physiological distance between a cell and a blood vessel12 and precludes the
104
formation of necrotic cores13 (Fig. 1A). The diameter variation at day 7 and 21 among the co-
105
culture spheroids is 190µm ± 10 µm (n=100) while only 4% of cells incorporate EdU during a 48h
106
exposure (Suppl. Fig. 1A), illustrating the non-proliferative nature of the cells. The 3D Hep/HSC
107
spheroids demonstrate a segregated organization with a concentration of HSCs in the core of
108
the spheroid (PDGFR-β positive cells, with a small elongated nucleus, Fig. 1B) while
109
hepatocyte-like cells (CYP3A4 positive cells, with large round nuclei, Fig. 1B) accumulate
110
preferentially in the periphery. Phase I hepatocyte metabolic capacity is demonstrated by
111
exposing the cells to prototype cytochrome P450 inducers for CYP3A, CYP1A2 and CYP2C9.
112
Even though there are less hepatocyte-like cells in the co-culture spheroids, the inducibility of
113
the cells shows little variation between mono- and co-cultures of Heps (Fig. 1D). Likewise, CYP-
114
induction values are in agreement with CYP-inducing capacities of Heps maintained in similar
115
conditions (Suppl. Table 4). Additionally, CYP3A4 staining of 3D Hep/HSC spheroids indicates
116
the presence of mature hepatocytes (± 20% of all spheroid cells is CYP3A4 positive, n=4). Note
AC C
EP
TE D
M AN U
SC
RI PT
93
5
ACCEPTED MANUSCRIPT
that differentiated HepaRGs are 50% hepatocyte-like while the other 50% are cholangiocyte-like
118
cells14, thus only 1/6 of the spheroid can be hepatocyte-like cells (Fig. 1B). Excretion of the
119
MRP2 substrate CMFDA15 and its accumulation in canulicular structures is only observed after
120
fixation, underlining the good elimination capacity of the spheroids (Fig. 1C, arrows). Albumin
121
secretion is similar in 3D Hep and 3D Hep/HSC cultures, but is manifold higher than in 2D Hep
122
cultures (Fig. 1E). Finally, mRNA levels of hepatocyte markers such as ALB, CY3A4, GSTA1
123
and SLCO1B1 is maintained in 3D Hep/HSC cultures at comparable levels to Hep mono-
124
cultures and in some cases higher than 2D Hep (Fig. 1F) despite the lower amount of
125
hepatocyte-like cells. In the presence of HepaRG culture medium, human 3D HSCs still activate
126
from day 7 onwards (Fig. 1G) while in 3D Hep/HSC co-cultures they do not, as can be
127
evidenced by the low mRNA levels of ACTA2, COL1A1 and LOX on day 21. Due to the
128
mesenchymal origin HepaRG still express these markers, but in much lower extent when
129
compared with the HSC-containing cultures (Fig. 1G). Nevertheless the HSCs in the co-culture
130
spheroids are still capable of responding to pro-fibrogenic stimuli such as TGFβ and LPS
131
(Suppl. Fig. 1B and C), known to have direct effects on HSCs.
132
Taken together, these experiments demonstrate functionality of both cell types in 3D Hep/HSC
133
cultures and we therefore refer to these cultures as human hepatic organoids. Despite the good
134
cellular profile of these organoids after 7 days in culture, we mainly focus on 21 day culture
135
characterization which perhaps allows long term repeated dose toxicity testing.
136
Drug-induced hepatic injury by acetaminophen
137
- Single exposure of human hepatic organoids to Acetaminophen
138
To show the drug responsiveness of the organoids we used Acetaminophen (APAP) and
139
compared it to the respective mono-cultures for specificity of the response. APAP was selected
140
for its kinetic stability, known effective in vitro concentrations, specific toxicity in hepatocytes and
141
high human relevance16. After a 24h APAP exposure at day 20, we observe only small
AC C
EP
TE D
M AN U
SC
RI PT
117
6
ACCEPTED MANUSCRIPT
differences in toxicity between the 3D organoids and the 3D Hep cultures (Fig. 2A,B),
143
suggesting a good CYP2E1 metabolism in both cultures17. Nevertheless, there is a tendency for
144
higher toxicity in the organoids, more pronounced on day 21 than on day 7 (Suppl. Table 5).
145
Upon APAP exposure, Heps disappear from the periphery of the organoids, probably by
146
necrosis-mediated cell death, while the ones remaining in a ring surrounding the HSC core are
147
caspase 3-positive apoptotic cells (Fig. 2E), representing the second path of APAP-induced
148
hepatocyte injury18. Although not known as a pro-fibrotic agent, metabolization of APAP by
149
functional hepatocytes in the organoid leads to toxicity mainly in hepatocytes and not HSCs, but
150
might lead to hepatocyte-injury dependent HSC activation. Indeed, we detect a strong dose-
151
dependent up-regulation of HSC activation-associated mRNAs COL1A1, COL3A1 and LOXL2
152
only in the hepatic organoids (Fig. 2C), suggesting HSC activation in response to hepatocyte
153
injury. In Hep mono-cultures, APAP exposure also leads to a 2-5 fold increase of collagen
154
mRNA levels, however the basal levels of collagen expression are 10-100 fold lower in 3D Hep
155
cultures when compared to the organoids (Suppl. Fig. 2A), highlighting the activation state of
156
the HSCs, the main producers of collagen in the liver1. Measurements of C-terminal pro-collagen
157
I peptide in the culture supernatant show an APAP-dependent increase of collagen protein only
158
in the organoid cultures (Fig. 2D and Suppl. Fig. 2B). ACTA2 mRNA levels do not change upon
159
APAP exposure in any of the cultures suggesting that the indirect APAP-induced HSC activation
160
is independent of TGFβ, since ACTA2 transcription depends largely on TGFβ signaling19. Yet,
161
we can detect a slight increase in αSMA protein levels in the center of the organoids after APAP
162
exposure (Fig. 2E). Additionally, we confirm collagen production, by immuno-staining and by
163
detection of cross-linked collagens (Sirius-red) in sections of hepatic organoids (Fig. 2E).
164
- Hepatic organoids recapitulate several aspects of liver fibrosis
165
Our results with APAP exposed organoids suggest that besides the known hepatotoxicity, APAP
166
has the capacity to induce HSC activation through hepatocyte injury. This prompted us to
AC C
EP
TE D
M AN U
SC
RI PT
142
7
ACCEPTED MANUSCRIPT
167
investigate drug-dependent HSC activation further by addressing in these organoids three in
168
vivo features of HSC activation: (i) enhanced HSC activation due to inflammation20
169
inhibition of HSC activation by histone deacetylase (HDAC) inhibition; and (iii) recovery from
170
HSC activation. We mimic the cytokine release from immune cells22 by performing the APAP
171
exposure in the presence of a cytokine mixture (CK: TNFα, IL1β, IFNα and IFNγ). While CK
172
alone (0 mM APAP+CK) has no effect on the culture read-outs, when combined with APAP, the
173
toxic effect is potentiated and the up-regulation of COL1A1, COL3A1 and LOXL2 mRNA is
174
increased (Fig. 3A, B). We then used valproic acid (VPA), an HDAC inhibitor that inhibits HSC
175
activation induced by CCl4 in mice23. Similar to in vivo, VPA does not affect APAP toxicity in the
176
organoids while up-regulation of the activation markers at the highest APAP concentrations was
177
inhibited (Fig. 3C, D). Finally, organoids cultured for 3 additional days after APAP washout,
178
show a decrease in HSC activation markers (Fig. 3E) indicative of HSC recovery24.
179
APAP seems to affect HSC mono-cultures only at the highest concentrations (EC50 = 31.4
180
±11mM). Instead of the CYP-mediated metabolization of APAP into NAPQI which is the main
181
toxic pathway in hepatocytes, this effect could also be linked to mitochondrial oxidative stress25,
182
26
183
assessed the mitochondrial functions and glycolytic activity by measuring the oxygen
184
consumption rate (OCR) and the extracellular acidification rates (ECAR) respectively (Fig. 4A).
185
Comparison of the basal respiration, ATP production and spare respiratory capacity in the
186
untreated control samples (Fig. 4B), suggests that the hepatic organoid cultures are more stable
187
than the equivalent mono-cultures. Furthermore, when comparing Hep-containing cultures with
188
the HSC mono-cultures, all parameters are roughly ten times higher. Especially in the case of
189
ATP, this seems to indicate that Heps are the major production source in the organoids and
190
suggests that the observed ATP-drop after APAP incubation in the hepatic organoids (Fig. 2B)
191
is mainly a consequence of Hep toxicity. Additionally, at 24h, a shift towards a more glycolytic
; (ii)
TE D
M AN U
SC
RI PT
21
AC C
EP
. In order to get a more complete picture of the cellular metabolism of the hepatic organoids we
8
ACCEPTED MANUSCRIPT
metabolism (higher ECAR and lower OCR) is only seen in 3D Hep mono-cultures, which in the
193
HSC monocultures was already high, but not in the organoids. In the hepatic organoids, a
194
general OCR decrease is observed, only after 24 hours of APAP exposure (Fig. 4C), while in the
195
respective mono-cultures this is observed already after 12 hours. Together, this shows that there
196
is little influence of APAP on mitochondrial functions in the hepatic organoids in contrast to the
197
respective mono-cultures. These findings are in line with the mitochondrial resistance observed
198
in hepatocytes during early stages of fibrosis in vivo27.
199
Single and repeated exposure of hepatic organoids to Methotrexate and Allyl alcohol.
200
To further explore the hepatic organoids as a read-out for drug-induced liver fibrosis we exposed
201
the cultures to the reference pro-fibrotic compounds Allyl alcohol, known to induce fibrosis in
202
rat28, and Methotrexate (MTX) which is classified to induce liver fibrosis in human subjects after
203
prolonged exposure29. Since the fibrotic effect of these compounds in vivo is only observed upon
204
long-term exposure, both compounds were tested in single and repeated doses (and compared
205
to APAP for HSC activation (Suppl. Fig. 4A-C)). Serum-free culture conditions were established
206
to avoid interference of the serum during repeated compound exposure30,8 (Suppl. Fig. 5A). The
207
two exposure profiles are compared by assessing the cells on day 21, 24h after the last
208
compound dose (Fig. 5A). Single MTX exposure does not affect cell viability at any of the tested
209
concentrations (Fig. 5B). However, upon repeated exposure, the percentage of ATP production,
210
even at the lowest concentration, decreases to less than 20%. Although repeated exposure to
211
1.9 µM MTX leads to smaller spheroids, the amount of disaggregating cells around the spheroid
212
is less than in organoids repeatedly exposed to 5 mM APAP (Fig. 5I and Suppl. Fig. 4D),
213
suggesting that the lower ATP levels not necessarily represent cell death. While single exposure
214
to MTX does not lead to an upregulation of activation markers at the mRNA level in the
215
organoids, repeated exposure to MTX clearly induces HSC activation (Fig. 5C and Suppl. Fig.
216
6B). Positive Collagen1 and αSMA staining in cross sections of the hepatic organoid however
AC C
EP
TE D
M AN U
SC
RI PT
192
9
ACCEPTED MANUSCRIPT
demonstrates that HSCs in the core of the spheroids are activated after both single and
218
repeated exposures (Fig. 5J). Sirius-red stainings are more in line with the mRNA levels of
219
activation showing a much stronger deposition of cross-linked collagen in the repeated dose
220
MTX setting than after a single exposure.
221
As for the case of Allyl alcohol, the clear dose-response observed at single exposure almost
222
disappears upon repeated exposure, letting us to presume induced toxicity resistance of the
223
cells (Fig. 5E). This can be explained by the inhibition of its metabolism to acrolein31 observed
224
during rat and mice exposures32. In single exposure, despite the drop in ATP at high
225
concentrations it does not affect the spheroid integrity (Fig. 5I,J). Both operational exposures
226
show an increased tendency of mRNA levels of HSC activation (Fig. 5F and Suppl. Fig. 6C).
227
While at the mRNA level the difference between single and repeated exposure to Allyl alcohol is
228
not as outspoken as for MTX, at the protein level, there is a strong fibrotic response in the
229
repeated exposures leading to a ∼100 fold increase of secreted collagen at day 14 (Fig. 5G) and
230
a strong accumulation and cross-linked collagen at the end of the experiment, shown by Sirius-
231
red and collagen stainings (Fig. 5J).
232
For both Allyl alcohol and MTX, repeated exposure shows high collagen secretion at day 14, 48h
233
after the 3rd exposure (Fig. 5D,G). At the end of the culture, the 24h pro-collagen secretion is
234
comparable to the single exposure in the case of Allyl alcohol, but lower in the case of MTX (Fig.
235
5D). For both compounds, nearly no collagen secretion is detected in the 3D mono-cultures (Fig.
236
5H).
237
Finally, APAP caused toxicity and HSC activation both in single and repeated-exposures, with a
238
shift to lower APAP concentrations for the repeated dose (EC50 from 10.4 ±2.3 mM to 2.25 ±0.35
239
mM APAP; Suppl. Fig. 4A). HSC activation followed with a peak at around 5-10 mM for the
240
repeated exposure instead of 20-40 mM (Suppl. Fig. 4B and Suppl. Fig. 6A).
241
Once again, the capacity of the cultures to recover from the insult was analyzed 3 days after
242
compound washout, however recovery was only observed in APAP single exposures and 63 µM
AC C
EP
TE D
M AN U
SC
RI PT
217
10
ACCEPTED MANUSCRIPT
MTX repeated exposures (Suppl. Fig. 7) which most likely reflects the mechanism by which
244
these compounds induce HSC activation.
245
Pro-fibrotic effect of APAP in vivo
246
Acute liver failure as a consequence of exposure to high doses of APAP has been widely
247
reported in animals and humans33, but so far there are no reports describing a fibrotic response
248
of APAP. Since our results with hepatic organoids indicated that APAP can indirectly induce
249
HSC activation, we next investigated the in vivo fibrotic potential of APAP in mice. Blood
250
analysis 24h post-injection demonstrates increased alanine transaminase (ALT) levels indicating
251
hepatocyte toxicity upon exposure of regular chow-fed mice to 300-500 mg/kg APAP (Fig. 6A).
252
300mg/kg APAP is already sufficient for HSC activation within 24 hours, demonstrated by
253
increased mRNA levels of Acta2, Col1a1 and Lox in freshly isolated HSCs (Fig. 6B). Bi-weekly
254
injections of APAP for 4 and 8 weeks (300 mg/Kg), results in the formation of scar tissue and
255
increased αSMA expression in BALB/c and C57BL/6 mice, but has the tendency to decline over
256
time (Fig. 6C and Suppl. Fig. 8). Indeed, when compared to CCl4-induced liver injury, APAP
257
treatment appears to be less fibrogenic (Fig. 6C, D).
258
TE D
M AN U
SC
RI PT
243
DISCUSSION
260
The microenvironmental factors that can affect a cell’s performance in culture are countless,
261
varying from chemical medium composition, mechanical factors or even diffusion rates. In the
262
case of HSCs, rigidity of the matrix to which they are attached at least partly determines their
263
activation state3. We report, for the first time, metabolically active human hepatic organoids in
264
which HSCs maintain a quiescent-like state for 21 days while retaining their capacity to respond
265
to pro-fibrotic compounds directly and in a hepatocyte-dependent manner.
266
In the last years, simplified models for the use in chemical risk assessment have been
267
developed, the so-called Adverse Outcome Pathways - AOP (https://aopkb.org) which should
AC C
EP
259
11
ACCEPTED MANUSCRIPT
finally lead to an integration of all AOPs into one large network34. A good predictive in vitro
269
method should at least recapitulate some of the key events of such AOP (Suppl. Fig. 9). The
270
developed human hepatic organoid consisting of differentiated HepaRGs and primary HSCs is to
271
our knowledge the first in vitro method that mimics drug-induced liver fibrosis and could be used
272
to further optimize the AOP of liver fibrosis.
273
The HepaRG to HSC cell-ratio of 1:2 and the hepatocyte-specific medium guarantees HSC
274
quiescense and keeps the hepatocytes functional. Although the used ratio is far from the
275
physiological ratio, we believe that in these organoids the HSCs work as feeder cells to support
276
hepatocyte functions35. The central accumulation of HSCs in the organoids is most likely due to
277
their stronger capacity to aggregate at the time of spheroid formation; HSCs are cultured in 2D
278
beforehand while the HepaRG cells are freshly thawn and immediately put in 3D co-culture. We
279
did question the use of primary HSCs and differentiated HepaRG cells by culturing organoids
280
consisting of the widely used HepG2 and LX-2 cells, hepatocyte and HSC lines respectively.
281
Besides the fact that cells in these organoids keep proliferating, the LX-2 cells activate over time
282
without injury and the organoids are much less sensitive to APAP, probably reflecting the low
283
hepatocyte functionality of the HepG2 cells (Suppl. Fig. 10).
284
The use of Allyl alcohol and MTX confirms the applicability of the method by showing HSC
285
activation only in the organoids and not in 3D mono-cultures. The upregulation of some of the
286
activated HSC-associated mRNAs in 3D Hep mono-cultures (albeit remaining at relatively low
287
levels) might be due to an EMT-like process due to the treatment with MTX or Allyl alcohol.
288
While EMT-like differentiation of hepatocytes is readily observed in vitro36, this does not
289
contribute to liver fibrosis formation in vivo1.
290
In contrast, repeated-dose exposure of the 3 tested compounds resulted in a significantly
291
higher, and sometimes unique up-regulation of COL1A1, COL1A3 and LOXL2 in the organoids
292
(Suppl. Fig. 11). These results demonstrate that only the hepatic organoids, and not their
AC C
EP
TE D
M AN U
SC
RI PT
268
12
ACCEPTED MANUSCRIPT
respective 3D mono-cultures, have the capacity to respond positively to the reference fibrotic
294
compounds Methotrexate and Allyl alcohol.
295
Even the hepatotoxic-independent HSC activation observed after repeated exposure to Allyl
296
alcohol is hepatocyte-dependent since the effect is not observed in the HSC mono-cultures
297
(Suppl. Fig. 11C). This highlights the potential of the hepatic organoids to not only “just” identify
298
hepatotoxic compounds that might cause indirect HSC activation (i.e. APAP), but also by
299
compounds that are not hepatotoxic but need the presence of functional hepatocytes to
300
indirectly activate HSCs (i.e. Allyl alcohol at the tested concentrations). The in vivo confirmation
301
of the unexpected fibrotic potential of APAP identified in the organoid cultures, highlights the
302
potential to identify thus far unknown fibrotic compounds in vitro that can actually cause fibrosis
303
in vivo. The lack of literature describing this fibrotic potential of APAP could be due to the much
304
lower profibrotic capacity than for instance CCl4 and the decline of this effect over time. The
305
report from Watelet and co-workers on preliminary APAP fibrosis evidence in humans37
306
suggests that this assessment should perhaps be pursued.
307
In conclusion, we have developed functional human hepatic organoids that can identify
308
compounds that induce fibrosis, a drug-induced liver injury (DILI) rarely addressed in vitro unlike
309
steatosis, choleastatis and phospholipidosis. This is a big step forward from the regular 2D HSC
310
cultures that are generally used. The hepatic organoids are suitable for repeated dosage and are
311
sensitive to the nature of the compounds by displaying differential toxicity and HSC activation
312
profiles. These hepatic organoids represent a substantial improvement when screening for drug-
313
induced liver fibrosis in terms of cost, animal use and prediction of liver fibrosis in human.
314
Furthermore, this technology could stimulate the development of culture models representative
315
of fibrosis in other organs such as lung and kidney, since these share common mechanisms38, 39.
AC C
EP
TE D
M AN U
SC
RI PT
293
316 317 318 13
ACCEPTED MANUSCRIPT
ACKNOWLEDGMENTS
320
First we would like to acknowledge the finantial support from different entities: T. Roosens and A. El
321
Taghdouini are funded by the Institute for the Promotion of Innovation through Science and Technology in
322
Flanders (IWT-Vlaanderen; SB/13170 and SB/111008). I. Mannaerts is supported by a Fund of Scientific
323
Research Flanders FWO-V post-doctoral fellowships (12N5415N LV). This work was supported by the
324
HeMiBio consortium funded by the European Commission and Cosmetics Europe as part of the SEURAT-
325
1 cluster (N° HEALTH-F5-2010-266777) and by the IWT project HILIM-3D (SBO 140045).
326
Furthermore, we thank P. Antony (Université Du Luxembourg) for the possibility to use the XFe96
327
Analyser, S. Kalaydjiev and D. Buurman (Seahorse Bioscience) for the help in the performance and
328
interpretation of the Mito Stress test, A. P. Batista (ITQB) for the help in the interpretation of the ECAR and
329
OCR results.
SC
RI PT
319
330
REFERENCES
332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359
1.
5.
6. 7. 8.
9.
10. 11. 12. 13.
TE D
4.
EP
3.
Mederacke, I. et al. Fate tracing reveals hepatic stellate cells as dominant contributors to liver fibrosis independent of its aetiology. Nat Commun 4, 2823 (2013). Friedman, S.L. Hepatic stellate cells: protean, multifunctional, and enigmatic cells of the liver. Physiol Rev 88, 125-172 (2008). Olsen, A.L. et al. Hepatic stellate cells require a stiff environment for myofibroblastic differentiation. Am J Physiol Gastrointest Liver Physiol 301, G110-118 (2011). Coll, M. et al. Integrative miRNA and Gene Expression Profiling Analysis of Human Quiescent Hepatic Stellate Cells. Sci Rep 5, 11549 (2015). Dusabineza, A.C. et al. Hepatic stellate cells improve engraftment of human primary hepatocytes: A pre-clinical transplantation study in animal model. Cell Transplant, doi: 10.3727/096368915X096686788 (2015). Thoen, L.F. et al. A role for autophagy during hepatic stellate cell activation. J Hepatol 55, 13531360 (2011). EL Taghdouini, A. et al. Genome-wide analysis of DNA methylation and gene expression patterns in purified, uncultured human liver cells and activated hepatic stellate cells. Oncotarget (2015). Klein, S., Mueller, D., Schevchenko, V. & Noor, F. Long-term maintenance of HepaRG cells in serum-free conditions and application in a repeated dose study. J Appl Toxicol 34, 1078-1086 (2014). Lubberstedt, M. et al. HepaRG human hepatic cell line utility as a surrogate for primary human hepatocytes in drug metabolism assessment in vitro. J Pharmacol Toxicol Methods 63, 59-68 (2011). Leite, S.B. et al. Three-dimensional HepaRG model as an attractive tool for toxicity testing. Toxicol Sci 130, 106-116 (2012). Gunness, P. et al. 3D organotypic cultures of human HepaRG cells: a tool for in vitro toxicity studies. Toxicol Sci 133, 67-78 (2013). Carmeliet, P. & Jain, R.K. Angiogenesis in cancer and other diseases. Nature 407, 249-257 (2000). Folkman, J. & Hochberg, M. Self-regulation of growth in three dimensions. J Exp Med 138, 745753 (1973).
AC C
2.
M AN U
331
14
ACCEPTED MANUSCRIPT
19. 20. 21.
22. 23. 24. 25.
26.
27. 28. 29. 30. 31. 32. 33. 34.
RI PT
18.
SC
17.
M AN U
16.
TE D
15.
Antherieu, S., Rogue, A., Fromenty, B., Guillouzo, A. & Robin, M.A. Induction of vesicular steatosis by amiodarone and tetracycline is associated with up-regulation of lipogenic genes in HepaRG cells. Hepatology 53, 1895-1905 (2011). Bogman, K., Erne-Brand, F., Alsenz, J. & Drewe, J. The role of surfactants in the reversal of active transport mediated by multidrug resistance proteins. Journal of pharmaceutical sciences 92, 1250-1261 (2003). Antoine, D.J. et al. Mechanistic biomarkers provide early and sensitive detection of acetaminophen-induced acute liver injury at first presentation to hospital. Hepatology 58, 777787 (2013). Lee, S.S., Buters, J.T., Pineau, T., Fernandez-Salguero, P. & Gonzalez, F.J. Role of CYP2E1 in the hepatotoxicity of acetaminophen. J Biol Chem 271, 12063-12067 (1996). Nakagawa, H. et al. Deletion of apoptosis signal-regulating kinase 1 attenuates acetaminopheninduced liver injury by inhibiting c-Jun N-terminal kinase activation. Gastroenterology 135, 13111321 (2008). Herrmann, J., Haas, U., Gressner, A.M. & Weiskirchen, R. TGF-beta up-regulates serum response factor in activated hepatic stellate cells. Biochim Biophys Acta 1772, 1250-1257 (2007). Marques, P.E. et al. Chemokines and mitochondrial products activate neutrophils to amplify organ injury during mouse acute liver failure. Hepatology 56, 1971-1982 (2012). Pradere, J.-P. et al. Hepatic macrophages but not dendritic cells contribute to liver fibrosis by promoting the survival of activated hepatic stellate cells in mice. Hepatology 58, 1461-1473 (2013). Aoyama, T., Paik, Y.H. & Seki, E. Toll-like receptor signaling and liver fibrosis. Gastroenterol Res Pract 2010, 1-8 (2010). Mannaerts, I. et al. Chronic administration of valproic acid inhibits activation of mouse hepatic stellate cells in vitro and in vivo. Hepatology 51, 603-614 (2010). Kisseleva, T. et al. Myofibroblasts revert to an inactive phenotype during regression of liver fibrosis. Proc Natl Acad Sci U S A 109, 9448-9453 (2012). Chen, C., Krausz, K.W., Idle, J.R. & Gonzalez, F.J. Identification of novel toxicity-associated metabolites by metabolomics and mass isotopomer analysis of acetaminophen metabolism in wild-type and Cyp2e1-null mice. J Biol Chem 283, 4543-4559 (2008). Prill, S. et al. Real-time monitoring of oxygen uptake in hepatic bioreactor shows CYP450independent mitochondrial toxicity of acetaminophen and amiodarone. Arch Toxicol, doi:10.1007/s00204-00015-01537-00202 (2015). Mitchell, C. et al. Protection against hepatocyte mitochondrial dysfunction delays fibrosis progression in mice. Am J Pathol 175, 1929-1937 (2009). Jung, S.A. et al. Experimental model of hepatic fibrosis following repeated periportal necrosis induced by allylalcohol. Scand J Gastroenterol 35, 969-975 (2000). Lindsay, K. et al. Liver fibrosis in patients with psoriasis and psoriatic arthritis on long-term, high cumulative dose methotrexate therapy. Rheumatology (Oxford) 48, 569-572 (2009). Pomponio, G. et al. In vitro kinetics of amiodarone and its major metabolite in two human liver cell models after acute and repeated treatments. Toxicol In Vitro (2014). Covaci, O.I., Bucur, B. & Radu, G.L. Acrolein detection based on alcohol dehydrogenase inhibition. International Journal of Environmental Analytical Chemistry 93, 325-334 (2013). Auerbach, S.S., Mahler, J., Travlos, G.S. & Irwin, R.D. A comparative 90-day toxicity study of allyl acetate, allyl alcohol and acrolein. Toxicology 253, 79-88 (2008). Kaplowitz, N. Idiosyncratic drug hepatotoxicity. Nat Rev Drug Discov 4, 489-499 (2005). Vinken, M. The adverse outcome pathway concept: a pragmatic tool in toxicology. Toxicology 312, 158-165 (2013).
EP
14.
AC C
360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407
15
ACCEPTED MANUSCRIPT
35.
36. 37. 38. 39.
Leite, S.B. et al. Merging bioreactor technology with 3D hepatocyte-fibroblast culturing approaches: Improved in vitro models for toxicological applications. Toxicol In Vitro 25, 825-832 (2011). Nambotin, S.B., Tomimaru, Y., Merle, P., Wands, J.R. & Kim, M. Functional consequences of WNT3/Frizzled7-mediated signaling in non-transformed hepatic cells. Oncogenesis 1, e31 (2012). Watelet, J. et al. Toxicity of chronic paracetamol ingestion. Aliment Pharmacol Ther 26, 15431544; author reply 1545-1546 (2007). Friedman, S.L., Sheppard, D., Duffield, J.S. & Violette, S. Therapy for fibrotic diseases: nearing the starting line. Sci Transl Med 5, 167sr161 (2013). Henderson, N.C. et al. Targeting of alphav integrin identifies a core molecular pathway that regulates fibrosis in several organs. Nat Med 19, 1617-1624 (2013).
RI PT
408 409 410 411 412 413 414 415 416 417 418 419
FIGURE LEGENDS
421
Figure 1: Human 3D Hep/HSC co-culture characterization. (A) Spheroid formation and maintenance
422
over 21 days. White scale bar represents 100 µm; upon 100 measurements, co-culture spheroid size was
423
determined to be 180 ± 20µm. (B) Merged and separate pictures of CYP3A4, PDGFR-β and DAPI of 3D
424
Hep/HSC paraffin sections on day 21; white scale bar represents 50 µm. Pictures are representative of
425
the culture spheroids in the 3 experimental repeats. (C-F) Hepatocyte functionality of Hep-containing
426
cultures; (C) Confocal image of CMFDA bile accumulation (arrows) in the Hep/HSC spheroids on day 17;
427
(D) Day 21 CYP induction upon exposure to prototype CYP inducers (Rifampicin – RIF; β-Naphtoflavone –
428
BNF; Phenobarbital – PB) for 48 hours. (E) Albumin secretion rate on days 7 and 21; (F) Gene expression
429
of reference hepatocyte markers (Albumin; Phase I – CYP3A4; Phase II –GSTa1; transporter –
430
SLCO1B1) on days 7 and 21 of culture. dCt over GAPDH values are displayed. (G) mRNA levels of HSC
431
activation markers using HepaRG medium on days 7 and 21. dCt over GAPDH values are displayed.
432
Error bars represent standard deviations (N=3 assays, pool of 6 spheroids/time point); *p<0.05 vs 3D
433
Hep/HSC. Expression of these markers on 2D and 3D Hep cultures vary from 2.8-6x, 800-14000x and
434
1000-2400x lower than in HSCs respectively for ACTA2, COL1A1 and LOX.
435
Figure 2: Single exposure of human hepatic organoids to Acetaminophen. (A) Time-line of APAP
436
exposure to the 3D cultures. (B) Cell viability in 3D Hep mono- and co-cultures after 24 hours incubation
437
with different concentrations of APAP. Each graphs shows 3 independent experiments where each point
438
represents the average % ATP over control of n=4-5 spheroids. (C) Relative mRNA gene expression of
439
ACTA2, COL1A1, COL3A1 and LOXL2 in the three 3D cultures showing APAP dose-dependent gene
440
transcription in 3D Hep/HSC. Each point represents the pool of 5 spheroids. *p<0.05 3D HSC vs 3D
441
Hep/HSC; $p<0.05 3D Hep vs 3D Hep/HSC; N≥3 independent assays (4-6 spheroids/condition). (D)
442
Procollagen measured in the supernatant 24h after incubation with APAP/solvent. N= 3 assays
443
(supernatant
444
Immunohistochemistry for Caspase 3, αSMA, Collagen and cross-linked collagen (Sirius-red) on paraffin
AC C
EP
TE D
M AN U
SC
420
pooled
from
6
spheroids).
*p<0.05
vs
solvent
control
(0
mM
APAP).
(E)
16
ACCEPTED MANUSCRIPT
embedded hepatic organoids exposed to 0 and 20 mM APAP. Black bars represent 50 µm. All error bars
446
represent standard deviations.
447
Figure 3: Hepatic organoids recapitulate several aspects of liver fibrosis. (A, C) Cell viability of the
448
hepatic organoids exposed to APAP alone or together with (A) cytokine/inflammatory mixture (CK) or (C)
449
Valproic Acid (VPA); (B, D, E) Relative mRNA levels of COL1A1, COL3A1 and LOXL2 from control (0 mM
450
APAP) and APAP-exposed Hep/HSC organoids (5-40 mM) and the effect of co-treatment with the (C) CK
451
mixture or (D) VPA and (e) APAP wash out. *p<0.05; test vs APAP alone; p<0.05 vs 0 mM APAP; N≥2
452
independent assays (n=4-6 spheroids/condition).
453
Figure 4: Oxidative phosphorylation and glycolysis rates in 3D human hepatic cultures. (A)
454
Seahorse Mito Stress test profile with key parameters of mitochondrial function. (B) Basal respiration, ATP
455
production and Spare respiration capacity of the 3D Hep/HSC co-culture and respective mono-cultures
456
after 12 and 24h in low buffer capacity assay medium (0 mM APAP). (C) Oxygen consumption rate
457
(OCR) and extracellular acidification rate (ECAR) of 3D cultures exposed to 10 mM APAP and respective
458
controls at 12h and 24h incubation. Error bars correspond to the standard deviations of measurements of
459
3-4 independent spheroids. p<0.05 vs. 12h; p<0.05 control vs 10 mM APAP (12h), *p<0.05 control vs 10
460
mM APAP (24h). FCCP - Trifluoromethoxy carbonylcyanide phenylhydrazone, A+R – Antimycin A &
461
Rotenone, Oligo – Oligomycin.
462
Figure 5: Single- and repeated-exposure of Methotrexate and Allyl alcohol to hepatic organoids.
463
(A) Schematic representation of the compound exposure experimental setup. (B, E) Cell viability
464
represented by %ATP (EC50(AA-single)= 57.5 ± 8.6 µM); (C, F) mRNA levels of LOXL2 and COL1A1; (D,
465
G) Procollagen secretion in culture supernatant (day 14 and 21); Error bars represent the standard
466
deviations between independent assays (N=2), each condition represents a pool of 5-6 spheroids. Graph
467
inserts in (C) and (F) represent the second repeat of the assay.
468
Procollagen levels in supernatant of the hepatic organoids and 3D HSC or Hep mono-cultures upon 24h of
469
single exposure at day 20. (I) Brigthfield images of Hep/HSC organoids at day 21 after single or repeated
470
exposure to the different compounds. (J) Immunohistochemical detection of cross-linked collagen (Sirius-
471
red), Collagen1 and αSMA in sections of day 21 Hep/HSC organoids (single and repeated exposure).
472
Black bar represents 100 µm.
473
Figure 6: In vivo data obtained from BALB/c mice after single and repeated exposure to APAP. (A)
474
ALT blood levels 24h after a single exposure to 300 mg/kg APAP or vehicle control (B) mRNA levels of
475
Acta2, Col1a1 and Lox in mouse HSCs isolated 24h after a single injection of 300 mg APAP/kg and the
476
vehicle control; (C, D) Sirius-red images and quantification of the percentage of red stained surface in
477
mice livers after 4 weeks of repeated exposure to APAP or CCl4.
RI PT
445
#
*
p<0.05 vs control (no compound). (H)
AC C
EP
TE D
$
M AN U
SC
$
17
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT