Novel liver fibrosis model in Macaca fascicularis induced by thioacetamide

Although transplantation is the only definitive treatment for liver cirrhosis, there remains a shortage of donors, necessitating that novel treatments be developed. We aimed to establish a liver fibrosis model in Macaca fascicularis that can help accelerate preclinical research. Liver fibrosis was induced by administering thioacetamide (TAA) and carbon tetrachloride (CCl4). Analysis of residual liver function and fibrosis progression was based on clinical indices, such as the Child–Pugh score or fibrotic markers, besides histology. TAA-induced marked fibrosis, whereas CCl4 did not induce fibrosis. Concerning residual liver function, both of TAA and CCl4 worsened the indices of the Child–Pugh score, but only the TAA model increased the retention ratio of indocyanine green. The TAA-induced fibrosis model in Macaca fascicularis worsens fibrosis and residual liver function, mimicking Child–Pugh grade B. Given that our model was evaluated by clinical indices, it could be applicable to preclinical research.

In this study, we evaluate liver fibrosis models induced by CCl 4 and TAA in Macaca fascicularis, using clinical indices of both residual liver function and fibrotic progression to establish a model that mimicked Child-Pugh grade B to C.

Results
General appearance and macroscopic appearance. Figure 1 outlines the study protocol and shows the group allocation. Table 1 summarises the general information regarding the allocated monkeys in this study. Monkeys treated by TAA lost appetite and some vomited just after administration during the first week, but their appetites and activities recovered after 2-3 weeks of administration. By contrast, monkeys treated with CCl 4 did not show any change during the first week and lost their activity and appetite gradually over 3-4 weeks after the initial administration. In addition, subcutaneous administration of CCl 4 caused skin erosion.
Monkeys treated by TAA lost body weight over time (Fig. 2a). Macroscopically, monkeys treated by TAA showed coarse granulation formation on their liver surfaces while those treated by CCl 4 showed fatty injury like change (Fig. 2b). Abdominal ultrasound revealed that ascites did not develop in any monkeys and any changes in the biliary tract, indicating biliary obstruction or cholecystitis, did not appear in any monkeys (Fig. 2b). On the other hand, splenomegaly appeared in monkeys receiving TAA and CCl 4 from 2 weeks after initial administration (Fig. 2c).
Residual liver function. Albumin levels decreased over time in monkeys treated with both TAA and CCl 4 from 2 weeks after initial administration. The prothrombin time increased persistently in monkeys receiving TAA, whereas the total bilirubin level increased significantly in monkeys receiving CCl 4 . Only TAA increased ICGR15. Serum NH 3 levels increased in both the TAA and CCl 4 groups, while aspartate transaminase (AST) and alanine aminotransferase (ALT) increased transiently 2-4 weeks after initial administration and recovered over time. Another feature of chronic liver disease, decreasing platelet levels, was also induced by both TAA and CCl 4 . However, neither model affected creatinine levels. Interestingly, white blood cell and C-reactive protein levels significantly increased in monkeys receiving CCl 4 (Fig. 2d).
Histological analysis and progression of liver fibrosis. Histology in control monkeys revealed normal cellular architecture (Fig. 3a). Fibrosis developed significantly in monkeys receiving TAA, but was less marked in monkeys receiving CCl 4 based on the analyses of tissue samples and fibrotic markers. Histology of the biopsy 8 weeks   www.nature.com/scientificreports www.nature.com/scientificreports/ specimens from monkeys receiving TAA at 4 and 8 weeks showed bridging fibrosis, a typical feature of F3 fibrosis, and massive inflammatory infiltration of lymphocytes around sinusoids (arrowhead of Fig. 3a). The histology of monkeys receiving CCl 4 showed centrilobular congestion but not fibrosis formation at 4 weeks. We conducted a second liver biopsy at 7 weeks because the general condition of monkeys receiving CCl 4 became poor. We hypothesised that they would not survive until the eighth week. The histology of the biopsy specimen at 7 weeks showed ballooning that are characteristic of fatty liver injury (arrow, Fig. 3a), but did not show significant fibrosis formation.
Quantitative analysis of fibrosis based on the area positive for Sirius red showed fibrotic progression in monkeys receiving TAA but not in those receiving CCl 4 . The hydroxyproline level tended to increase in monkeys receiving TAA, but this increase was not over time as with the other fibrosis assessments (Fig. 3b). HA levels increased 4 weeks after initial dosing with both TAA and CCl 4 ; however, it only increased persistently to week 8 in monkeys receiving TAA. The levels of P-III-P and Col-IV increased 2 weeks after initial administration and persisted throughout the administration of both TAA and CCl 4 . Finally, levels of HA and Col-IV increased markedly in monkeys receiving TAA (Fig. 3c).

Discussion
Liver cirrhosis is characterised by devastation of residual liver function and replacement of the normal parenchyma by fibrosis. Transplantation is the only definitive, albeit radical, treatment for liver cirrhosis, but donor shortages remain an issue. Several novel treatments have been developed, such as cellular transplantation 16 and molecular targeted drugs 17,18 ; however, none of these treatments have proven to be suitable to replace liver transplantation. Therefore, the development of viable alternates continues to be an unmet public health need. To accelerate the development and clinical application of novel treatments or procedures via effective preclinical research, we must first establish reliable models of liver fibrosis in NHPs (e.g., Macaca fascicularis), which are more closely related to humans 6 .
Clinical applicability requires that any usable model of liver fibrosis model in NHPs be formulated based on clinically relevant indices. Otherwise, the models may not truly simulate liver cirrhosis in clinical settings. Quantitative analysis of histological data is not always used clinically, but it can be used to establish liver fibrosis in animal models objectively. To date, no model of fibrosis in NHPs has been established based on clinical indices and quantitative analyses of fibrosis. In this study, based on these relevant clinical indices, we showed that TAA and CCl 4 induced quantitative fibrosis progression in Macaca fascicularis.
Liver damage induced by TAA more closely reflected the clinical features of liver cirrhosis than that induced by CCl 4 in terms of both residual liver function and fibrotic progression in our protocol. Fibrosis progression was much more significant in monkeys receiving TAA than in monkeys receiving CCl 4 . Although histological analysis by biopsy is the most accurate assessment of fibrosis, it is invasive. Therefore, fibrotic markers such as HA, P-III-P and Col-IV are generally used as supplemental markers of fibrosis progression in clinical settings. In this study, each of these markers predicted fibrogenesis and correlated well with the histological findings. Changes in the fibrotic markers showed that TAA administration for 4 weeks induced liver fibrosis and that administration for an additional 4 weeks worsened that fibrosis. Hydroxyproline assays are widely used as a quantitative analysis of fibrosis in animal models, but we did not show a correlation with either the histological results or the fibrotic markers. Given that fibrogenesis is not homogeneous in the liver, hydroxyproline assays of small specimens may not always reflect overall fibrosis formation.
The blood chemistry results indicate that liver damage began 2 weeks after the reagents were initially administered. We considered that these changes indicated the onset of acute liver damage because AST and ALT transiently increased 2-4 weeks after initial administration and recovered over time. Because of this acute injury, a monkey (TAA-2) died just 5 days after administration. To prevent acute death by TAA administration, we should start with a low dose and then gradually increase the dose to the effective dose. Additional administration 2-4 weeks after initial administration induced chronic liver damage and resulted in persistent exacerbation of the clinical indices Child-Pugh score and Liver damage grade. According to the indices of Child-Pugh score, such as albumin, total bilirubin, NH 3 and prothrombin time, both reagents induced liver damage that mimicked grade B severity by the eighth week. However, only TAA administration worsened ICG clearance function. The ICG clearance test is widely used clinically for dynamic assessments of the capacity of the liver to metabolise and eliminate drugs and it indicated whether liver damage is present. Our study is the first we know of in the published literature to show ICG clearance data in NHPs.
TAA was superior to CCl 4 in our study because it both induced liver fibrosis progression and worsened residual liver function. Notably, the protocol took only 8 weeks to establish a fibrosis model, which was the shortest time required among the existing models in NHPs, which took more than 16 weeks at least. This could contribute to the accomplishment of the three R's principle 19 ; Replacement, Reduction and Refinement. Furthermore, research with this model could predict effects in humans with greater accuracy because the results were based on clinical indices and objective data specific to fibrosis.
Our study has several limitations. First, we found individual differences in the effects of the reagents. One monkey that received TAA showed neither significant fibrosis formation nor a reduction in liver function, as was apparent in the other monkeys treated by TAA; it may be that the response to reagents is much less predictable in NHPs than in rodents. Individual modifications to reagent administration may be necessary to establish usable models of fibrosis in all NHPs, as previously reported 12 . Second, we could not assess what changes might have occurred after stopping reagent use. Previous reports have indicated that fibrosis can reverse after stopping the reagents in rodents 20 and NHPs 11,12 . Therefore, it will be essential to determine the effects of stopping TAA and CCl 4 in future research. Because we strictly complied with the three R's principle, we designed this study with minimum required monkeys. Additional studies are necessary to verify our results to establish the preclinical model.  www.nature.com/scientificreports www.nature.com/scientificreports/ In this study, we showed that TAA administration for 8 weeks in Macaca fascicularis induces the clinical changes of chronic liver damage consistent with Child-Pugh grade B and class F3 fibrosis (Inuyama classification or Metavir scoring system). This model could prove to be of great benefit in preclinical research for the development of novel treatments for liver cirrhosis.

Animals. All animal studies were approved and performed according to the Tsukuba Primate Research
Center. All experiments were performed under the guidelines of the Tsukuba Primate Research Center as well as under the guiding principles for animal experiments using NHPs set out by the Primate Society of Japan. Tsukuba Primate Research Center provided 11 healthy Macaca fascicularis (6-18 years old, 3.38 ± 0.86 kg). All monkeys were caged individually in a ventilated room with a 12-h cycle of artificial light from 7 AM to 7 PM. Temperature and humidity were maintained at 23 °C-27 °C and at 50%-70%, respectively. Reagent administration. Figure 1 outlines the study protocol and shows the group allocation. TAA (Sigma-Aldrich, St. Louis, MO, USA) was dissolved in normal saline (Otsuka Pharmaceutical, Tokyo, Japan) and administrated three times a week at a dose of 100 mg/kg (n = 5). CCl 4 (WAKO) was diluted with corn oil (WAKO) and administrated twice a week at a dose of 1000 mg /kg (n = 3). The control group received normal saline at a dose of 1 ml/kg three times a week (n = 3). All reagents were administrated subcutaneously for 8 weeks.

Assessment of residual liver function.
To assess the Child-Pugh score, MELD score and Liver damage grade, blood tests were conducted to determine the levels of albumin, total bilirubin, NH 3 , and creatinine, indocyanine green (ICG) clearance, and prothrombin time to evaluate residual liver function. The ICG clearance test was determined by the retention of ICG at 15 minutes (ICGR15). In addition, AST and ALT levels were measured as general indices of liver damage. Blood was drawn from femoral veins every 2 weeks. We also performed abdominal ultrasonography (AUS) every 2 weeks to determine the presence of ascites and splenomegaly. Spleen index was calculated with short diameter by long diameter of spleen on AUS.