COMPONENTE URBANO
Capitulo 8 LICENCIAS Artículo 100º DEFINICION.
The bioavailability and disposition of a drug is largely governed by biotransformation mechanisms and transport. These mechanisms occur in the intestine but the liver is clearly a major organ for metabolic clearance. Traditional analyses are completed using established in vitro models but many more are in development and validation stages. Although hepatic cell lines are widely used as they retain many hepatic functions, they have limitations with numerous reports recording lower levels of drug transporters and CYPs (84, 105, 261, 272) than primary hepatic cells, which have very similar gene expression profiles to liver tissue (268, 273). Some studies have reported extremely low levels of CYP3A4 mRNA and protein in HepG2 cells that are not induced upon rifampicin (RIF) treatment (274). Nonetheless, cell lines have been pivotal in defining mechanisms for transcriptional regulation including, constitutive activation of CYP2B6 and CYP3A4 by CAR through its accumulation in the nucleus due to a lack of cytoplasmic CAR retention protein (CCRP, designated DNAJC7 in the NCBI database) (164, 165). Therefore, they provide tools for drug metabolism, toxicity and absorption studies.
Primary human hepatocytes display common hepatic functions, similar levels of gene expression and metabolism observed in vitro and in vivo (271). Analysis can be completed on suspension hepatocytes that retain hepatic function for 4-6 hours (275) or on cultured hepatocytes that display hepatic capabilities for ~24 hours post plating (244), meaning both methods can be utilised in drug metabolism or toxicity studies. Interestingly, cryopreservation of primary human hepatocytes is linked with a selective decrease in gene expression, drug metabolism (275) and a loss of specific sub-types including diploid and tetraploid cells accompanied with decreased cell
viability (271). Although inter-individual variation between human hepatocyte donors is high, the greatest asset of using isolated hepatocytes is the ability to assess active drug transport, metabolism and response to metabolite formation.
This chapter focussed on characterising the gene expression profiles of HepG2 and Huh7 cell lines in comparison with primary human hepatocytes to consequently determine any relationships between NR co-regulators and CYP3A4. It should be noted that the results and conclusions are based on mRNA expression data, which may not always correlate to protein expression or activity. However, strong correlations for efflux transporters and CYPs are well documented in the literature (261, 276). Protein expression analysis of all genes would be required to completely correlate data, but due to challenges in proteomic assessment of nuclear and membrane bound proteins (277) and the number of candidates selected, mRNA analysis was deemed to be the most appropriate alternative.
Although clear similarities were found for some genes, variability was observed between cell lines and primary human hepatocytes. Initially, work focussed on assessing the quantitative differences of key drug transporter and metabolism enzyme gene expression (Figure 2.1). Of the 3 efflux transporters investigated (ABCB1, ABCC1, ABCC2), ABCC2 expression was found to be similar to primary human hepatocytes. The most abundantly expressed efflux transporter in the human liver is ABCC2 (261) suggesting analysis of this transporter may be completed with more confidence in Huh7 cell lines. However, the most abundantly expressed influx transporter of the liver, OATP1B1 was found to have significantly lower levels of expression in the cell lines (~1-20%). Synergistic transport by OATP1B1 and ABCC2
has been identified as a crucial mechanism of both proteins involved in the transport of substrates including, HMG-CoA reductase inhibitors (e.g. simvastatin, rosuvastatin, pravastatin) hence, biliary secretion should be analysed with caution in the two hepatic cell lines investigated. OATP1B1 is an important mediator of CYP3A4 metabolism due to the large overlap in substrate affinity as well as governing access of substrates to the enzyme (30, 278). Hence, it may be a contributing mechanism for the low CYP3A4 metabolic activity reported in these cells.
ABCB1 has a low expression in the human liver but is significantly induced during regeneration and cholestasis (279). ABCB1 and CYP3A4 were significantly under expressed in both hepatic cell lines compared to primary human hepatocytes. In contrast, an up-regulation of CYP3A4 expression has been observed in confluent Huh7 cells but the mechanism by which this occurs is yet to be determined (105). CYP3A4 mRNA expression and activity in HepG2 cells were found to be a poor representation of primary human hepatocytes (Figure 2.1 and 2.3, respectively). Additionally, the passaging process correlated to a decrease in CYP3A4 activity (Figure 2.3), as previously observed (84, 280). CYP3A4 activity was comparable to values noted by others (101). The major drug disposition genes analysed in Figure 2.1 are targets of the NRs CAR and PXR, which regulate their transcription thus it was also important to consider NR expression.
NRs CAR and PXR are strict regulators of drug metabolism, cell growth, differentiation, apoptosis and cell migration. One would therefore predict variable NR expression in immortalised cells, as demonstrated in Figure 2.2. In agreement with
previous studies (111), relative to primary human hepatocytes, HepG2 and Huh7 cells expressed diminished levels of CAR and PXR. CAR and PXR control the expression of the genes analysed in Figure 2.1, hence their reduced expression in hepatoma cell lines (compared to primary human hepatocytes) is a likely consequence of significantly low levels of the NRs (94). Albeit minimal, detection of CYP3A4 and NR mRNA in HepG2 and Huh7 hepatoma cells implies gene expression is not fully eradicated by known molecular mechanisms such as methylation alterations and/or chromatin condensation. CAR and PXR are also regulated by a number of transcriptional modifiers (Table 1.4). Comparative analysis of these transcription factors between HepG2 and Huh7 hepatic cell lines and primary human hepatocytes showed significant disparities in gene expression profiles (Figure 2.4).
The cellular chaperones are essential in coordinating the translocation of proteins. The major chaperone Hsp90 is responsible for not only positively folding DNA but supports and contributes to protein degradation (281), chromatin remodelling (282) and cytoskeleton movement (283). The mechanisms by which Hsp90 orchestrates these functions are currently under investigation. However, due to its diverse abilities and correlation with oncogenic proteins Hsp90 has the potential to be targeted for cancer and immunological dysfunction treatments (282, 284). The chaperones investigated herein (HSPA4, HSPA8, HSP90AA1, PHI1D1, Figure 2.4a) were under expressed in HepG2 cells and significantly under expressed in Huh7 cells when compared to primary human hepatocytes. Within the liver and primary human hepatocytes nuclear translocation of CAR predominantly occurs via activation cascades. However, it has been noted within transformed hepatic cell lines that CAR spontaneously accumulates in the nucleus (285). Hence, the low expression of
chaperones in hepatoma cell lines here may be a result of this spontaneous mechanism.
In contrast to the chaperones, the majority of co-chaperones expressed in Huh7 cells were over-expressed when compared to primary human hepatocytes, whilst the majority were under expressed in HepG2 cells (Figure 2.4b). Co-chaperones serve to stabilise the chaperone conformation and extend the binding region to the NR (205). Inhibition of these co-chaperones (VCP, CDC37) has been shown to result in lethal perinatal outcomes in knockout mice (286). The dissociation of the chaperone/co- chaperone complex from NRs in vivo is highly dependent upon hydrolysis of ATP. In vitro co-chaperones have been found to dynamically exchange on chaperones leading to many behavioural displays by the NR (205). Co-chaperones (FKBP4, FKBP5, PPP5C) primarily attenuate or potentiate receptor activity and binding of hormones involved in cell growth and survival. Therefore, the over-expression of these genes in immortalised cells is not surprising. In addition, the lack of chaperone expression may also contribute to the low expression observed in the HepG2 cells. In contrast to work herein, DNAJC7 has been reported to be significantly over-expressed in HepG2 cells compared to Huh7 cells or primary human hepatocytes and is associated with higher levels of CAR in the cytoplasm (188). Possible reasons for the discrepancy in co- chaperone expression may be due to the maintenance buffer used, media supplements, cell passage number or if the cells have diverged.
Co-repressors bind to unliganded, antagonist or inverse agonist bound NRs to prevent the interactions and actions of co-activators (184). Similar to all the co-regulators analysed, Huh7 cells poorly expressed co-repressors in comparison to HepG2 cells
(Figure 2.4c). However, TMTC3 strongly competes with the co-activators NCOA2 and PGC1α (212) and therefore, the significant over expression of TMTC3 in Huh7 cells may contribute to the significantly reduced expression of the co-activators (Figure 2.4c and 2.4d). TMTC3 recruits HDAC1 and HDAC2, thus the expression of the co-repressors may be a consequence of this signalling cascade. The same trend was not observed for HepG2 cells. Expression of SREBF1 is associated with decreased xenobiotic clearance and regulation of cholesterol and triglyceride synthesis (208). Expression of this co-repressor was significantly lower in Huh7 cells but HepG2 cells expressed similar levels of the gene to primary human hepatocytes. Regulation of this gene is crucial to ensure optimum lipid metabolism suggesting HepG2 cells may be more reliable in this context.
The NCOA family of co-activators function as intermediate signalling genes to enhance transcription and it has also been demonstrated they play a key part in general metabolism (287). Conflicting data exist for NR co-activator expression in hepatoma cell lines (192, 207). We found NCOA1, NCOA2, NCOA3, NCOA5 and NCOA6 were under expressed in both cell lines compared to primary human hepatocytes suggesting the transformed cell lines may utilise alternative pathways. To restore CYP3A4 activity and inducibility, studies have also found transfection of NCOA1 and PGC1α genes increases PXR, hepatocyte nuclear factor 4 α (HNF4α) and CYP3A4 gene expression in HepG2 cells. However, CYP3A4 activity and induction was not restored (96, 288). Gadd45β is known to mediate numerous cellular functions including: apoptosis, cell growth and DNA repair by interacting with cyclin and cyclin-dependent kinases (203). Hence, one may predict increased expression of the co-activator in immortalised cells but this was not apparent (Figure 2.4d).
Activation of PXR by Gadd45β has no effect on cell growth in vitro, rather HepG2 cells have been found to differentiate into epithelial-mesenchyma (202). In addition, an over expression of CAR is correlated to an upregulation of Gadd45β (203); the converse may also be expected which is demonstrated herein (Figure 2.1 and 2.4d). PGC1α regulates adaptive thermogenesis, gluconeogenesis, mitochondrial activity and oxidative metabolism, making it a crucial hepatospecific transcription factor. In comparison to primary human hepatocytes, PGC1α was significantly under-expressed in Huh7 and HepG2 cell lines, as observed previously (207, 288).
In comparison to the environment in vivo, cultured hepatoma cell lines experience different stimuli in terms of cell stress, growth factors, tissue specific functions and xenobiotic exposure which may contribute to the significantly altered phenotype observed here. It is feasible to hypothesise that the hepatoma cell may adapt to its artificial environment with its primary aim to preserve life in the most efficient way possible rather than preservation of hepatic functions, since selective pressures will be different. The analysis completed clearly outlines the extent of gene expression disparities between hepatoma cells and primary human hepatocytes and the caution that should be applied when utilising the cells for drug metabolism and transport studies.
Cell passage can alter the phenotype and morphology of cell lines, which are particularly apparent at high passage number (67, 289). CYP3A4 activity (Figure 2.3) and NR co-activator Gadd45β and PGC1α expression were significantly reduced as HepG2 cell passage number increased (Figure 2.5). Linear correlations between Gadd45β/PGC1α expression and CYP3A4 activity in HepG2 cells were observed
(Figure 2.6). Whilst the correlation for Gadd45β and CYP3A4 is novel, the association of PGC1α and CYP3A4 is consistent with previous work (288).
In summary, we have determined the extent of the disparities between hepatic cell lines (HepG2 and Huh7) and primary human hepatocytes by comprehensively analysing the expression of key drug disposition genes, PXR, CAR and their co- regulators, which has not previously been assessed. Potential reasons for altered gene profiles in HepG2 and Huh7 cells were elucidated and Gadd45β and PGC1 were identified as potential mediators for the lack of CYP3A4 activity and expression in HepG2 cells and were therefore selected for work aimed at restoration of metabolic function in Chapter 3.