Printer Friendly

Use of the combination index to determine interactions between plant-derived phenolic acids on hepatotoxicity endpoints in human and rat hepatoma cells.

ARTICLE INFO

Keywords:

Hepatotoxicity

Dietary phenolic acids

Mixtures

Interactions

Species differences

ABSTRACT

The beneficial or adverse effects of isolated phytochemicals are not always concordant with effects of the botanical dietary supplements from which they were derived. This disparity could be due to interactions between the various phytochemicals present in the whole plant. The phenolic acids, rosmarinic acid (RA), caffeic acid (CA) and ferulic acid (FA) are widely present in foods and dietary supplements, and they are assumed to exert various beneficial biological effects. However, there is little data on the potential biological interactions of these three phenolic acids which commonly occur together and are linked metabolically. In the present study, liver toxicity of the three phenolic acids was assessed on the three compounds singly and in various binary and one ternary combinations. A series of in vitro endpoints relevant to liver toxicity were evaluated in both a human (HepG2/C3A) and rat (MH1C1) hepatocyte cell line. The Combination Index (Cl) was calculated for each endpoint from both the concentration responses of the single compounds and the responses of the various binary and ternary mixtures. Both synergistic and antagonistic interactions were observed for some endpoints and some combinations of test agents. Interactions were most prevalent in measures of oxidative stress and cytochrome P450 activities in both cell types. There was only a 53% concordance between the rat and human cells which may be suggestive of species differences. The data suggest an approach for better characterizing the beneficial or adverse effects of complex botanical products through evaluation of interactions between individual phytochemical components.

Published by Elsevier GmbH.

Introduction

Drugs that produce overtly similar effects individually will sometimes show exaggerated or diminished effects when used concurrently (Tallarida 2001). Similar interactions can take place between drugs and dietary chemicals (Nahrstedt and Butterweck 2010). Botanicals used as both phytomedicines and as dietary supplements are complex mixtures of phytochemicals. However, safety and efficacy studies on these products often focus on single chemical components tested individually. Effects observed in studies on individual components often do not mimic those observed with the intact plant (Liu 2004; De Kok et al. 2008). Therefore, a rapid and quantitative assessment is necessary to identify potential synergistic or antagonistic effects from simple additive effects.

The Combination Index (Cl). also called the Interaction Index, is widely used to assess both beneficial and adverse interactions between pharmaceuticals (Zhao et at. 2004). The Cl has also been used to assess the interactive toxicities of environmental chemicals (McDermott et at. 2008). Khafif et at. (1998) used the Cl to quantitate chemopreventive synergism between botanical phenolics in cell cycle blockade using human oral epithelial cells. Saw et al. (2011) used the Cl to demonstrate synergism between phytochemical indoles and isothiocyanates in the induction of the antioxidant response in HepG2-C8 cells. The aforementioned studies on food-related chemicals focused on identifying beneficial interactions between phytochemicals. However, there appear to be no literature reports where the Cl has been used to assess potential harmful interactive effects between natural compounds occurring together in botanical products that are consumed as foods or as dietary supplements.

The liver plays a central role in the disposition and detoxication of xenobiotic and natural chemicals including drugs, dietary supplements, food additives, and food contaminants. This role of liver makes it a primary target of toxicity following oral exposure to a wide variety of chemicals. Therefore, identification of potential synergistic interactions that can alter liver function and, possibly, lead to liver toxicity is an important public health concern.

Rosmarinic acid (RA), caffeic acid (CA) and ferulic acid (FA) are common components of various foods such as thyme, rosemary, and basil, and of botanical dietary supplements either as the single compound or as a component of a botanical extract. There is evidence for the metabolic conversion of RA to CA and FA in vivo in both rats (Nakazawa and Ohsawa 1998) and humans (Baba et al. 2005). The reported in vivo biological activities of all three phenolic acids are similar and include antioxidant effects (Petersen and Simmonds 2003; Maurya and Devasagayam 2010; Srinivasan et al. 2005).

An in vivo study in rats (Nahrstedt and Butterweck 2010) evaluated pharmacological interactive effects (synergism) between individual phytochemicals isolated from St. John's Wort (Hypericurn perforatum L.). However, animal studies are time consuming and resource intensive, especially when looking for potential interactions following exposures to multiple agents. Therefore, validated in vitro cell culture systems are desirable for studying pharmacological interactions such as synergism or antagonism. Due to the limited lifespan, high cost, and significant batch-to-batch variability of fresh hepatocytes, liver hepatoma cell lines are commonly used to evaluate hepatotoxicity in vitro because of their availability and stable phenotype. HepG2/C3A cells, a clonal derivative of HepG2, have many desirable properties including strong contact inhibition of growth, high albumin production, low production of alpha-fetoprotein, measurable basal and inducible monooxygenase activities, and the ability to grow in glucose deficient medium (Kelly 1994). MH1C1 rat hepatoma cells show many similarities to cultured rat hepatocytes in the expression of biotransformation activities (Donato et al. 1994), albumin synthesis (Richardson et al. 1969), and growth in low glucose medium (Schamhart et al. 1979).

The current study was designed to investigate the ability of rapid, cell-based assays to identify potential synergistic and antagonistic interactions between phytochemicals that could alter the liver toxicity profiles of the individual compounds. Three phenolic acids, rosmarinic acid, caffeic acid and ferulic acid were selected as model phytochemicals. Studies were conducted in both human (HepG2/C3A) and rat (MH1C1) hepatocyte cell lines to assess species differences. Seven endpoints were evaluated which cover a variety of biological mechanisms relevant to hepatotoxicity including oxidative stress, mitochondrial membrane permeability, cellular neutral and polar lipid accumulation, CYP1A, 2B, and 3A activities, and cytolethality (Flynn and Ferguson 2008). Biological activity was evaluated for the compounds individually and in several binary and one ternary combination. Significant interactions, antagonistic or synergistic, were identified using the Combination Index along with appropriate confidence intervals.

Materials and methods

Materials

The human and rat hepatoma cell lines HepG2/C3A cells (CRL-10741) and MH1C1 cells (CCL-144) were purchased from American Type Culture Collection (ATCC) (Manassas, VA). Cell culture medium and medium supplements were obtained from Life Technologies/Gibco (Grand Island, NY). Fetal bovine serum was purchased from Hyclone (Logan, UT). Rosmarinic acid was purchased from INDOFINE (Hillsborough, NJ). Caffeic acid. ferulic acid, DMSO, and assay substrates (dihydrodichlo-rofluorescein, rhodamine 123, nile red. 7-ethoxyresorufin, 7-benzyloxyresorufin, resorufin, H33258, salicylamide, dicumarol) were purchased from Sigma (St. Louis, MO).

Cell culture

HepG2/C3A and MH1C1 cells were both cultured in Dulbecco's minimal essential medium (DMEM) with low glucose, pyruvate, Glutamax[TM] and pyridoxine. This basal medium was supplemented with MEM non-essential amino acids (1% final concentration), HEPES (10 mM final concentration) and fetal bovine serum (10% final concentration). No antibiotics were added to the medium. Cells were plated onto the inner 60 wells of 96-well tissue culture plates at a density of 6 x [10.sup.4] cells/[cm.sup.2] for HepG2/C3A cells and 12 x [10.sup.4] cells/[cm.sup.2] for MH1C1 cells. Cells were propagated and used up to a maximum of 5 passages at which point propagation began anew from a fresh vial of cryopreserved stock. Plates were incubated at 37[degrees]C in 5% C[O.sub.2]. Under these conditions, cells reached confluence in 5-6 days. Cells were treated with test compounds for 72 h beginning on culture day 8.

Treatment with phenolic acids

In the concentration dependence study, the three phenolic acids were dissolved in DMSO and then diluted in culture medium to a series of concentrations (125, 250, 500, 1000 [micro]M) with 0.5% DMSO final concentration. Vehicle only (DMSO) was used as the control. Each agent was tested in three concurrent plates with six replicates for 72 h. In the interaction study, the three phenolic acids were tested in different binary combinations (1:3, 1:1, 3:1) and the single ternary combination (1:1:1) with the total final phenolic concentration of 500 [micro]]M to demonstrate potential interactive effects.

Endpoint assays

Endpoint assays were conducted essentially as described in Liu et al. (2011).

Statistical analysis

The Combination Index (CI) (Zhao et al. 2004) between two drugs A and B is:

CI = ([C.subA,X]/[IC.sub.X,A]) + ([C.sub.B,X]/[IC.sub.X,B])

where [C.sub.A,X] and [C.sub.B,X] are the concentrations of drug A and drug B used in combination to achieve X% drug effect. [IC.sub.X,A] and [IC.sub.X,B] are the concentrations for single agents to achieve the same effect. A CI of less than, equal to, and more than 1 indicates synergy, additivity, and antagonism, respectively.

[IC.sub.X,A] and [IC.sub.X,B] were determined by fitting the commonly-used log-logistic concentration response curve (Seefeldt et al. 1995), a variant of the Hill equation, to the single agent data:

y = C + D - C/1 + exp[b(log(x) - log([I.sub.50]))] (1)

where x is the compound concentration, C is the mean response at very high doses. D is the mean response at control, b is the slope (positive if responses decreases with dose or negative if response increases with concentration) and [I.sub.50] is the concentration that causes 50% of the response. D was set to 100 as all responses were standardized to percent of control. The parameters were estimated via nonlinear modeling using PROC NLIN in SAS (SAS[R] 9.2 2002-2008, SAS Institute Inc., Cary, NC) where the initial estimates of C, b and [I.sub.50] were determined via visual inspection and based on historical data.

Using these fitted curves, [IC.sub.X,A] and [IC.sub.X,B] were estimated by solving Eq. (1) for the concentration [X.sub.A] of compound A and the concentration [X.sub.B] of compound B which produced individually the same effect y as the combination [C.sub.A, X] and [C.sub.B,X] produced:

log([^.X].sub.A]) = log [I.sub.50A] + 1/[[^b].sub.A] + 1/[[^b].sub.A][log([[^.D].sub.A] - y) - log(y - [[^C.sub.A])]

and

log([[^.X].sub.B]) = log [I.sub.50B] + 1/[[^b].sub.B][log([[^.D].sub.B] - y) - log(y - [[^C].sub.B])

Then simply, [IC.sub.X,A] = exp(log([[^.X].sub.A]) and [IC.sub.X,B] = exp(log([^.X].sub.B]))

Bootstrap 90% confidence intervals were computed around the Cl using 1000 bootstrap samples, each containing 80% of the data. Since several of the bootstrap distributions were right skewed, the bias-corrected percentile method (Millard and Neerchal 2000) for computing the bootstrap confidence intervals was employed. A Ci that did not include 1.0 within the upper and lower confidence limits was considered indicative of a significant interaction.

Results

Concentration dependence study on individual compounds in HepG2 cells

Results are presented in Table 1. CA and FA gave a significant concentration-related decrease in oxidative stress while RA gave a biphasic close response by inhibiting oxidative stress at lower concentrations and increasing it at the highest concentration tested (1000 [micro]M). Both RA and CA caused a significant decrease in rho-damine 123 retention at all concentrations tested while FA had no significant effect. None of the compounds had a significant effect on neutral lipid content. RA significantly increased polar lipid content at the highest concentration tested, but CA and FA had no significant effect on polar lipid content. RA, CA and FA all induced CYP1A activity at the higher concentrations used, but maximum induction at 1000 [micro]M was nearly 10-fold greater following treatment with RA than with CA or FA. Both RA and CA significantly induced CYP2B/3A activity at the 1000 [micro]M concentration level while FA caused a significant, concentration-related decrease in CYP2B/3A activity. Cell viability, as determined by total cellular DNA content, was reduced significantly only by RA and only at the highest concentration tested (1000 [micro],M). Total cellular DNA content was increased by all concentrations of CA and FA, significantly at 125 and 1000 [micro]M of each.
Table 1 Concentration dependent activity of RA, CA and FA
in HepG2/C3A cells.

      Conc.    Oxidative    Rhodamine   Neutral     Polar       CYP1A
    ([mu]M)       stress          123    lipids    lipids    activity
                            retention

RA      125   98.8 [+ or   92.2 [+ or  110.8 [+  121.7 [+    120.0 [+
                  -] 0.7     -] 3.2 *     or -]  or -] 30  or -] 40.4
                                           20.6         3

        250   96.0 [+ or   90.7 [+ or   99.4 [+  108.4 [+    114.7 [+
                -] 0.8 *     -] 4.3 *     or -]     or -]       or -]
                                           30.2      37.6        50.1

        500   93.6 [+ or   51.1 [+ or  111.9 [+  124.7 [+   1785.2 [+
                -] 1.4 *     -] 1.1 *     or -]     or -]       or -]
                                           19.6      16.4     155.8 *

       1000  112.8 [+ or   42.0 [+ or  127.5 [+  147.9 [+   5761.6 [+
                -] 0.9 *     -] 0.5 *     or -]     or -]       or -]
                                           24.3    28.9 *    1487.5 *

CA      125  100.7 [+ or   92.5 [+ or  108.6 [+  112.2 [+    103.7 [+
                  -] 2.9     -] 3.6 *     or -]     or -]       or -]
                                           21.9      18.5        16.8

        250   98.6 [+ or   91.5 [+ or  115.4 [+  116.4 [+    139.6 [+
                  -] 1.2      -]3.2 *     or -]     or -]       or -]
                                           18.6      19.0        33.3

        500   97.1 [+ or   90.0 [+ or  120.0 [+  125.1 [+    278.4 [+
                -] 1.4 *     -] 3.3 *     or -]     or -]  or -] 37.1
                                           11.1      19.3           *

       1000   93.9 [+ or   46.0 [+ or   84.9 [+   87.5 [+    631.8 [+
                -] 1.1 *     -] 1.0 *     or -]     or -]       or -]
                                           20.5      19.8       47.7*

FA      125   98.3 [+ or  100.5 [+ or  100.4 [+  103.1 [+    126.6 [+
                  -] 1.2       -] 2.9     or -]        or       or -]
                                           23.2    -]17.5       130.9

        250   95.0 [+ or  101.6 [+ or  105.1 [+   96.1 [+    146.7 [+
                -] 1.2 *       -] 3.1     or -]     or -]       or -]
                                           18.2      10.0        54.8

        500    * 91.9 [+  102.2 [+ or  105.4 [+   90.6 [+    129.2 [+
               or -] 1.0       -] 5.2     or -]     or -]       or -]
                       *                   24.1      18.9        63.6

       1000   87.9 [+ or   97.3 [+ or  113.0 [+   87.2 [+    387.2 [+
               -]  1.7 *       -] 3.4     or -]     or -]  or -] 50.0
                                           11.7      10.4           *

      CYP2B/3A    Total
      activity      DNA

RA  95.4 [+ or  94.8 [+
        -]15.0    or -]
                   18.7

      135.3 [+  92.0 [+
         or -]    or -]
          22.1     16.7

      136.1 [+  96.5 [+
         or -]    or -]
          10.4     13.4

      209.0 [+  83.2 [+
    or -] 49.6    or -]
             *   14.6 *

CA  95.0 [+ or    112.8
       -] 10.6    [+ or
                -] 18.5
                      *

      107.4 [+    109.2
         or -]    [+ or
          24.6       -]
                   20.4

      117.4 [+    109.4
         or -]    [+ or
          20.8       -]
                   18.9

      229.9 [+    127.3
    or -] 25.1    [+ or
             *  -] 20.5
                      *

FA  85.8 [+ or    107.9
       -] 10.5    [+ or
                 -] 9.9
                      *

    71.0 [+ or    106.8
      -] 6.4 *    [+ or
                     -]
                   11.3

    65.4 [+ or    105.4
      -] 9.4 *    [+ or
                     -]
                   11.3

    63.8 [+ or    114.5
     -] 15.6 *    [+ or
                -] 12.4
                      *

Results are expressed as mean % of control [+ or -] SD, RA, CA
and FA represent rosmarinic acid, caffeic acid and ferulic acid,
respectively.

* Significantly different from control (p < 0.05
for Dunnett' test).


Concentration dependence study on individual compounds in MH1C1 cells

Results are presented in Table 2. CA and FA gave a significant concentration-related decrease in oxidative stress while RA concentrations of 250 [mu]M and above significantly increased oxidative stress. Both RA and CA caused a significant concentration-related decrease in rhodamine 123 retention while FA had no significant effect. Both RA and CA caused a significant concentration-related decrease in both cellular neutral and polar lipid content while FA had no effect on either parameter. Both RA and CA caused a concentration-related increase in CYP1A activity while FA caused a significant concentration-related decrease in CYP1A activity. Both RA and CA significantly induced CYP2B/3A activity at higher concentration levels while FA had no significant effect on CYP2B/3A activity. Total cellular DNA content was reduced significantly by RA in a concentration-dependent manner. The response of total DNA content to CA was biphasic with a significant increase at 250 [mu]M and a significant decrease at 1000 [mu]M. FA had no significant effect on total cellular DNA content.
Table 2

Concentration dependent activity of RA, CA and FA in MH1C1 cells.

      Conc.   Oxidative   Rhodamine   Neutral    Polar      CYP1A
    ([mu]M)      stress         123    lipids   lipids   activity
                          retention

RA      125  98.5 [+ or  92.0 [+ or   89.7 [+  91.5 [+   101.7 [+
                 -] 0.8    -] 2.5 *     or -]    or -]      or -]
                                          7.9    4.3 *       11.1

        250    111.3 [+  64.7 [+ or   94.9 [+  86.6 [+   175.6 [+
              or -] 1.2    -] 0.8 *     or -]    or -]      or -]
                      *                   9.0    6.0 *     14.0 *
     11.2 *      -] 7.3

        500    117.0 [+   563 [+ or   78.8 [+  79.9 [+   294.7 [+
              or -] 1.9    -] 0.9 *     or -]    or -]   or -] 12
                      *                 7.1 *    4.3 *          *

       1000    112.8 [+  50.0 [+ or   67.1 [+  71.5 [+   426.0 [+
              or -] 1.0    -] 0.4 *     or -]    or -]      or -]
                      *                14.4 *    2.2 *     24.9 *

CA      125  99.7 [+ or  95.7 [+ or   84.9 [+    101.7    83.2 [+
                  -] 13      -] 3.0     or -]    [+ or      or -]
                                          9.4       -]        4.4
                                                  16.5

        250   993 [+ or  93.3 [+ or   89.0 [+  96.6 [+    80.8 [+
                 -] 0.7    -] 2.5 *     or -]    or -]      or -]
                                         12.8      213        7.1

        500  94.7 [+ or  81.4 [+ or   18.6 [+  35.7 [+   226.8 [+
               -] 1.0 *    -] 1.8 *     or -]    or -]      or -]
                                        5.5 *   10.0 *    113.3 *

       1000  84.4 [+ or  50.9 [+ or   14.0 [+  29.2 [+   509.8 [+
               -] 0.4 *    -] 0.6 *     or -]    or -]      or -]
                                        3.0 *    8.2 *     47.5 *

FA      125  98.0 [+ or    105.7 [+  109.9 [+  1123 [+  833 [+ or
                 -] 2.0       or -]     or -]    or -]   -] 8.0 *
                               21.8      20.2      203

        250   953 [+ or  93.7 [+ or  104.6 [+    119.0    81.1 [+
               -] 1.2 *      -] 3.1     or -]    [+ or  or -] 113
                                         22.4       -]          *
                                                  26.4

        500  91.1 [+ or  95.2 [+ or  113.4 [+    123.5    70.1 [+
               -] 1.8 *       -] 23     or -]    [+ or  or -] 8.2
                                         22.0       -]          *
                                                  17.1

       1000  83.8 [+ or  90.6 [+ or  124.1 [+    117.5    67.7 [+
               -] 1.8 *      -] 2.5     or -]    [+ or  or -] 5.7
                                         31.8   -] 183          *

    CYP2B/3A   Total
    activity     DNA

RA  102.7 [+   100.6
       or -]   [+ or
        17.6      -]
                 9.3

     1813 [+    90.2

       or -]   [+ or
               7.3 *

    342.5 [+    83.4
       or -]   [+ or
      28.4 *  -] 5.4
                   *

     5203 [+  773 [+
       or -]   or -]
      54.7 *   5.8 *

CA  106.6 [+   102.4
       or -]   [+ or
         6.2      -]
                12.5

    100.4 [+   106.8
       or -]   [+ or
         7.4      -]
                11.5
                   *

    543.1 [+   100.1
       or -]   [+ or
      28.0 *      -]
                11.0

    886.9 [+    89.1
       or -]   [+ or
      65.9 *  -] 8.1
                   *

FA  105.8 [+   100.5
       or -]   [+ or
        16.0      -]
                 9.9

    113.0 [+   102.9
       or -]   [+ or
         9.5      -]
                20.9

    103.6 [+   101.0
       or -]   [+ or
        13.7      -]
                10.5

     93.2 [+   100.0
       or -]   [+ or
        14.2      -]
                 9.5

Results are expressed as mean % of control [+ or -] SD. RA, CA and
FA represent rosmarinic acid, caffeic acid and ferulic acid,
respectively

* significantly different from control (P < 0.05 for Dunnett's test).


Interactive effects of various compound combinations in HepG2 cells

Data are presented in Table 3. Corresponding Combination Indices with upper and lower confidence limits are presented in Table 4. Twenty-six significant interactions were observed in the various binary combinations of phenolic acids. Nineteen of these interactions were synergistic, and the remaining seven were antagonistic. Oxidative stress had the greatest number of interactions with 7 of the 9 binary combinations demonstrating a significant interaction. Three of these seven interactions were synergistic, and four were antagonistic. CYP1A activity had 5 significant interactions with 3 of these being synergistic and 2 antagonistic. For the remaining endpoints, all observed significant interactions for rhodamine 123 retention (three), for CYP2B/3A activity (four), and for neutral lipids (three) were synergistic. Of the four significant interactions for total DNA, three were synergistic and one was antagonistic. For the ternary combination of phenolic acids, synergistic interactions were observed for rhodamine 123 retention and CYP1A activity, and antagonistic interactions were observed for oxidative stress and total DNA.
Table 4

Calculated Combination Index and confidence limits in HepG2/C3A cells.

Mixture    RA ([mu]M)       0      0      0    125    250     375    375
           CA ([mu]M)     125    250    375      0      0       0    125
           FA ([mu]M)     375    250    125    375    250     125      0

CYP1A      Full sample   0.40  11.18  16.22   0.92      >    0.60   0.60
activity   CI                                        1000
           Median        0.39    5.4   5.53   0.79   26.2    0.75   0.69
           boot-strap
           CI
           Lower         0.32   0.03   0.48   0.73   80.2    0.46   0.01
           confidence
           limit
           Upper         0.42      >      >      >      >    0.76   0.75
           confidence           1000   I000   1000   1000
           limit

CYP2B/3A   Full sample   0.19   0.54   0.78   0.09   0.31    1.55   1.95
activity   CI
           Median        0.10   0.57   0.81   0.08   0.51    2.73  52.56
           boot-strap
           CI
           Lower         0.13   0.06   0.09   0.04   0.05    0.28   0.79
           confidence
           limit
           Upper         1.46   0.93   0.99   0.32   0.68  162.59  23.51
           confidence
           limit

Rhodamine  Full sample   0.57   0.76   0.88   1.01   0.88    1.01   0.89
123        CI
retention  Median        0.55   0.75   0.85   1.05   0.87    0.99   0.89
           boot-strap
           CI
           Lower         0.43   0.60   0.75   0.73   0.77    0.93   0.78
           confidence
           limit
           Upper        12.06  23.71  32.85  52.49   1.05    1.09   0.98
           confidence
           limit

Neutral    Full sample   0.82   0.44   0.20   0.46   0.45    0.43   0.26
lipids     CI
           Median        1.32   1.47   0.55   1.00   1.19    1.20   0.32
           boot-strap
           CI
           Lower         0.13   0.01      0   0.03      0    0.01   0.12
           confidence
           limit
           Upper         2.16   2.03   0.44   2.91   0.98    1.12   0.77
           confidence
           limit

Oxidative  Full sample   1.45   1.67   1.21   0.62   2.06  > 1000   0.18
stress     CI
           Median        1.42   1.65   1.21   0.60   1.93  > 1000   0.18
           boot-strap
           CI
           Lower         1.08   1.24   0.94   0.43   1.42  > 1000   0.10
           confidence
           limit
           Upper         2.11   2.56   1.74   0.96   3.49  > 1000   0.25
           confidence
           limit

Polar      Full sample   3.03   2.75   0.84   0.79   0.63    1.13   0.61
lipids     CI
           Median        4.31   2.72   1.00   0.82   1.13    2.32   0.69
           boot-strap
           CI
           Lower         0.83   0.41   0.17   0.01      0    0.17   0.08
           confidence
           limit
           Upper            >      >   7.72   3.00  42.11   94.44   1.19
           confidence    I000   1000
           limit

Total DNA  Full sample      >      0   0.41   0.13   0.21  > 1000      >
           CI            1000                                       1000
           Median           >   0.25   0.78    0.2   0.29  > 1000      >
           boot-strap    1000                                       1000
           CI
           Lower         0.60      0      0      0      0    1.10   0.74
           confidence
           limit
           Upper            >      0   1.09   0.59   0.38  > 1000      >
           confidence    1000                                       1000
           limit

Mixture    RA ([mu]M)      250    125  166.6
           CA ([mu]M)      250    375  166.6
           FA ([mu]M)        0      0  166.6

CYP1A      Full sample  > I000   1.45   0.48
activity   CI
           Median        53.29   1.16   0.55
           boot-strap
           CI
           Lower         96.38   0.92   0.31
           confidence
           limit
           Upper        > 1000      >   0.56
           confidence            1000
           limit

CYP2B/3A   Full sample    1.97   1.99   1.45
activity   CI
           Median         5.00  48.14  48.32
           boot-strap
           CI
           Lower          0.43   0.57   0.66
           confidence
           limit
           Upper        114.62   2.46   1.51
           confidence
           limit

Rhodamine  Full sample    0.81   0.72   0.78
123        CI
retention  Median         0.81   0.72   0.77
           boot-strap
           CI
           Lower          0.73   0.66   0.69
           confidence
           limit
           Upper          0.88   0.78   0.84
           confidence
           limit

Neutral    Full sample    0.14   1.72   1.08
lipids     CI
           Median         0.46   1.77   2.08
           boot-strap
           CI
           Lower          0.01   0.17   0.07
           confidence
           limit
           Upper          1.65      >   1.64
           confidence            1000
           limit

Oxidative  Full sample    0.47   0.77      >
stress     CI                           1000
           Median         0.44   0.78      >
           boot-strap                   1000
           CI
           Lower          0.28   0.55      >
           confidence                   1000
           limit
           Upper          0.64   1.83      >
           confidence                   1000
           limit

Polar      Full sample    0.12   0.59   2.03
lipids     CI
           Median         0.12   0.70   3.44
           boot-strap
           CI
           Lower          0.08   0.13   0.08
           confidence
           limit
           Upper          1.37   2.04      >
           confidence                   1000
           limit

Total DNA  Full sample    0.31   1.85      >
           CI                           1000
           Median         0.39   1.73      >
           boot-strap                   1000
           CI
           Lower             0   0.01   5.74
           confidence
           limit
           Upper          6.10      >      >
           confidence            1000   1000
           limit


Interactive effects of various compound combinations in MH1C1 cells

Data are presented in Table 5. Corresponding Combination Indices with upper and lower confidence limits are presented in Table 6. Thirty-seven significant interactions were observed in the MH1C1 cells in the various binary combinations of phenolic acids. There were 8 significant interactions for oxidative stress, 7 each for CYP1A and CYP2B/3A activities, 6 for neutral lipids, 4 for rhodamine 123 retention, 2 for polar lipids, and 3 for total DNA. For the ternary combination of phenolic acids, synergistic interactions were observed for rhodamine 123 retention, CYP1A activity and CYP 2B/3A activity, and an antagonistic interaction was observed for oxidative stress.

Comparisons between cell lines

Comparing the individual compounds (Tables 1 and 2), the effects on oxidative stress and rhodamine 123 retention were comparable between the human and rat cells in both direction and magnitude of response. However, there were some highly significant effects of RA and CA on both neutral and polar lipid content in the rat cells but not in the human cells. Effects of RA and CA on CYP1A activity were similar in direction but markedly different in magnitude between the two cell lines. FA significantly induced CYP1A activity in the human cells but significantly inhibited CYP1A activity in the rat cells. Effects of RA and CA on CYP2B/3A activity were similar in direction but different in magnitude between the two cell lines. FA significantly inhibited CYP2C/3A activity in the HepG2 cells but had no significant effect on this activity in the MH1C1 cells. The effects of RA on total DNA were comparable between the two cell lines. CA increased the total DNA in HepG2 cells, but CA had a biphasic response in MH1C1 cells by significantly increasing total DNA at 250 [mu]M and significantly decreasing total DNA at 1000 [mu]M. FA increased the total DNA in HepG2 cells but had no significant effect on total DNA in MH1C1 cells.

Comparing the sensitivity in detecting compound interactions (Tables 3 and 4), 59% of the Combination indices were statistically significant for the rat cell data, whereas the human cell data resulted in statistically significant combination indices only 43% of the time. The rat and human cell data were not always in agreement in the direction of the interaction (antagonism or synergism). The full sample combination indices agreed between cell lines on the direction of the interaction only 53% of the time. That is, the human cell and rat cell CI's were either both above 1 or both less than 1 only 53% of the time. The median bootstrap CI's agreed slightly more at 57% of the time.

Discussion

Our data show that various binary combinations and one ternary combination of the three phenolic acids, CA, RA and FA, at a total phenolic concentration of 500 [mu]M can have interactive effects on several measures of liver function that are significantly different from the effects predicted by simple additivity. That non-additive interactions were seen most often with oxidative stress (8 of 10 phenolic acid combinations in the HepG2/C3A cells, 9 of 10 phenolic acid combinations in the MH1C1 cells) supports in vivo data which suggest that antioxidant benefits of phenolic mixtures found in whole botanical products are superior to those observed with single phenolic compounds (Liu 2004: De Kok et al. 2008: Saw et al. 2011).The next most prevalent interactions were seen with CYP1A activity (6 of 10 phenolic acid combinations in the HepG2/C3A cells, and 8 of 10 phenolic acid combinations in the MH1C1 cells). Our findings support an earlier study in HepG2 cells that showed that a combination of tea catechins inhibits TCDD-induced CYP1A activity more effectively than the individual catechins do (Williams et al. 2003) and in vivo data which suggest that the anti-carcinogenic activity of many foods and dietary supplements is attributable to their modulation of cytochrome P450 activities that can generate carcinogens (De Kok et al. 2008: Nahrstedt and Butterweck 2010).

Construction of isobolograms as described by Berenbaum (1989) is an alternate, but closely related to the CI. means of quantifying interactions between phytopharmaceuticals (Wagner and Ulrich-Merzenich 2009). However, we feel that the CI offers some computational advantages over isobolograms, especially in the determination of confidence intervals. In addition, our statistical approach was distinctive in that we opted to use nonparametric bootstrapping to estimate the confidence intervals around the Cl as the shape of the Cl parent distribution appeared skewed. The use of Eq. (1) to fit the dose response curves merits further exploration. We used Eq. (1) since it is the accepted gold standard for close response data. However, we found that such dose--response curves may not be appropriate for several endpoints whose trends were not faithful to a logistic pattern.

Daily intake of phenolic acids, from both dietary and supplement sources, could be several grams per day (Clifford 1999; Ovaskainen et al. 2008). Plasma concentrations of intact phenolics in humans following oral exposure are low, typically on the order of 1 [mu]M (Scalbert and Williamson 2000). However, the total plasma antioxidant capacity following ingestion of phenolic-rich foods indicates much higher plasma levels of total phenolic compounds, on the order of 50 p.M. This observation suggests that phenolics are metabolized quickly, primarily by methylation or conjugation with sulfate or glucuronic acid, but continue to circulate in the plasma as the metabolites. Although the in vivo conversion of RA to CA has been attributed to esterases in gut flora, our preliminary data (not shown) suggests that HepG2/C3A cells possess both the esterase activity to convert RA to CA and the catechol-O-methyltransferase activity to convert CA to FA. The methylation and conjugation (sulfation and glucuronidation) of hyclroxycinnamic acids, including CA and FA, by HepG2 cells has been reported previously (Mateos et al. 2006). The contribution of these metabolites to beneficial or harmful effects of phenolics is unknown. It should be noted that interpretation of the relevance of phenolic levels used in the current study (500 [mu]M) based on blood concentrations measured following in vivo exposure should be cautious given that peak liver concentrations of compounds can be much higher than steady state blood levels because of first-pass effects.

Cytochrome P450 induction and inhibition are important factors in many areas of pharmacology. pharmacokinetics and toxicology, usually leading to unfavorable clinical events such as drug-drug or food-drug interactions. CYP1A1 is poorly expressed in human liver, but it is highly inducible. CYP1A2 is expressed principally in the liver and is also inducible. CYP1A2 metabolizes more than 100 clinical drugs (e.g., clozapine, tacrine, tizanidine, and theophylline), a number of procarcinogens (e.g., benzo[a]pyrene and aromatic amines), and several important endogenous compounds (e.g., steroids) (Zhou et al. 2010). Therefore, induction or inhibition of CYP1A may provide a partial explanation for some clinical drug/herbal interactions. In the current study, RA and CA induced CYP1A activities concentration-dependently in both HepG2/C3A and MH1C1 cells in vitro. However, co-treatment with RA and CA caused significant interactions in both cell lines (Tables 3 and 5), especially for HepG2/C3A cells where a marked synergism was noted with a mixture of 25% CA and 75% RA with a total phenolic acid concentration of 500 [mu]M. This synergism could lead to interactions and possible toxicity when CYP1A substrate drugs are co-administered with food or dietary supplements containing both RA and CA.
Table 3

Interactions of FA. RA and CA in HepG2/C3A cells.

  CA       RA       FA      Oxidative   Rhodamine
([mu]M)  ([mu]M)  ([mu]M)    stress         123
                                        retention

    125      375        0   94.6 [+ or  51.4 [+ or
                            -] 1.5 (s)  -] 2.8 (s)

    250      250        0   94.8 [+ or  53.0 [+ or
                            -] 0.7 (s)  -] 2.8 (s)

    375      125        0   96.4 [+ or  55.1 [+ or
                                -] 1.6  -] 3.3 (s)

    125        0      375   94.7 [+ or  88.8 [+ or
                            -] 1.1 (a)      -] 7.0

    250        0      250   96.4 [+ or  91.5 [+ or
                            -] 1.1 (a)      -] 4.7

    375        0      125   96.1 [+ or  90.6 [+ or
                                -] 1.3      -] 4.1

      0      I25      375   97.8 [+ or  94.8 [+ or
                            -] 3.7 (s)      -] 4.8

      0      250      250  113.3 [+ or  71.4 [+ or
                            -] 1.1 (a)      -] 3.2

      0      375      125  117.6 [+ or  62.6 [+ or
                            -] 1.0 (a)      -] 2.3

    167      167      167  106.8 [+ or  65.9 [+ or
                            -] 2.0 (a)  -] 3.5 (s)

  Neutral    Polar        CYP1A       CYP2B/3      Total
  lipids     lipids     activity     activity       DNA

  106.7 [+  122.0 [+    3135.4 [+  173.9 [+ or   102.9 [+
 or -] 6.5     or -]  or -] 859.8      -] 14.6      or -]
       (s)       5.0          (s)                     8.3

  105.1 [+  111.0 [+  174.3 [+ or  177.1 [+ or   104.3 [+
     or -]     or -]      -] 42.6       -] 7.5  or -] 7.5
      12.8      15.1          (a)                     (s)

  109.3 [+  119.7 [+  191.1 [+ or  186.1 [+ or   103.2 [+
     or -]     or -]      -] 38.1      -] 14.1      or -]
      21.4      22.8                                  9.0

  115.4 [+  124.1 [+  531.7 [+ or   75.5 [+ or   104.0 [+
     or -]     or -]     -] 143.9      -] 16.9         or
      10.1       3.6          (s)                  -]11.0

  115.2 [+  124.8 [+  491.2 [+ or   90.5 [+ or   105.7 [+
 or -] 5.6     or -]     -] 238.8      -] 14.7      or -]
                 4.4                       (s)        9.3

  112.6 [+  123.0 [+  447.0 [+ or   94.5 [+ or   105.3 [+
 or -] 5.4     or -]     -] 210.6   -] 7.6 (s)      or -]
       (s)       8.9                                  9.8

  128.3 [+  122.1 [+  137.0 [+ or   78.9 [+ or   102.8 [+
     or -]     or -]      -] 23.2   -] 8.0 (s)  or -] 9.5
      15.5      10.7                                  (s)

  126.8 [+  130.0 [+  120.2 [+ or  146.3 [+ or    99.7 [+
or -] 20.7     or -]      -] 31.1      -] 23.3  or -] 8.3
       (s)      24.9          (a)          (s)        (s)

 1 19.6 [+  131.7 [+  758.3 [+ or  164.8 [+ or    94.9 [+
     or -]     or -]     -] 702.5      -] 24.8  or -] 6.3
      16.8      22.7          (s)                     (a)

  132.4 [+  136.0 [+  163.7 [+ or  160.1 [+ or    99.6 [+
     or -]     or -]      -] 13.4      -] 23.3      or -]
      48.7      51.5          (s)                11.4 (a)

Results are expressed as mean % of control [+ or -] SD. RA. CA and
FA represent rosmarinic acid, caffeic acid and ferulic acid,
respectively, s, synergistic interaction: a. antagonistic
interaction as determined by the Combination Index (CI).
Table 5

Interactions of FA. RA and CA in MH1C1 cells.

 CA       RA       FA      Oxidative   Rhodamine
([mu]M)  ([mu]M)  ([mu]M)    stress         123
                                        retention

    125      375        0  127.8 [+ or  57.2 [+ or
                            -] 3.0 (a)  -] 0.8 (s)

    250      250        0  114.6 [+ or  59.3 [+ or
                            -] 5.8 (s)  -] 1.3 (s)

    375      125        0   96.3 [+ or  60,9 [+ or
                            -] 1.8 (s)  -] 2.3 (s)

    125        0      375   90.2 [+ or  89.7 [+ or
                            -] 1.3 (s)      -] 3.6

    250        0      250   91.6 [+ or  89.3 [+ or
                                -] 1.4      -] 4.5

    375        0      125   83.2 [+ or  70.1 [+ or
                            -] 1.6 (s)  -] 4.6 (s)

      0      125      375  102.1 [+ or  83.1 [+ or
                            -] 3.5 (s)      -] 2.2

      0      250      250  114.9 [+ or  74.6 [+ or
                            -] 1.2 (a)      -] 1.4

      0      375      125  118.8 [+ or  67.5 [+ or
                            -] 2.0 (a)      -] 0.8

    167      167      167  106.4 [+ or  71.0 [+ or
                            -] 1.7 (a)  -] 1.0 (s)

 Neutral     Polar     CYP1A      CYP2B/3A    Total
  lipids    lipids    activity    activity     DNA

  38.6 [+   28,9 [+    188.3 [+    594.7 [+  95.9 [+
or -] 8.2     or -]       or -]  or -] 78.7    or -]
      (s)   6.1 (s)        16.5         (s)      7.1

  39.7 [+   37.3 [+    201.8 [+    574.2 [+  95.1 [+
or -] 9,4     or -]  or -] 20.5       or -]    or -]
      (s)      10.7         (s)        62.2      7.6

  44.9 [+   33.0 [+    199.2 [+    722.6 [+  98.7 [+
or -] 8.3     or -]       or -]  or -] 95.1    or -]
      (s)      13.2        15.9         (s)      6.5
                                                 (s)

  72.9 [+   74.4 [+  71.8 [+ or    101.1 [+    105.8
    or -]     or -]      -] 6.1       or -]    [+ or
     15.1      12.8         (a)        14.1   -] 9.4
                                                 (s)

  73.9 [+   79.2 [+  70.4 [+ or    540.2 [+    104.5
    or -]     or -]      -] 4.0  or -] 84.5    [+ or
     10.0      11.5         (a)         (s)   -] 7.8

  90.2 [+   86.3 [+  69.5 [+ or    615.0 [+    101.1
or -] 8.1     or -]      -] 9.1  or -] 87.8    [+ or
      (a)      12.8         (a)         (s)   -] 9.8
                                                 (s)

  107.9[+  102.7 [+  85.1 [+ or    124.7 [+  96.1 [+
    or -]     or -]      -] 7.7  or -] 28.9    or -]
 11.8 (a)   6.7 (a)         (a)         (a)      7.2

  98.0 [+  100.0 [+    111.6 [+    237.5 [+  91.3 [+
    or -]     or -]   or -] 9.1  or -] 39.0    or -]
 15.5 (a)       8.5         (a)         (a)      4.1

  77.0 [+   86.6 [+    149.1 [+    384.8 [+  91.0 [+
    or -]     or -]  or -] 10.8   or -]61.0    or -]
     13.5       7.1         (s)         (s)      4.5

  52.1 [+   67.9 [+    176.8 [+    496.3 [+  98.2 [+
    or -]     or -]   or -] 8.8  or -] 30.2    or -]
      8.0       4.5         (s)         (s)      8.9

Results are expressed as mean % of control [+ or -] SD. RA. CA
and FA represent rosmannic acid, caffeic acid and ferulic acid,
respectively, s. synergistic interaction; a. antagonistic
interaction as determined by the Combination Index (CI).
Table 6 Calculated Combination Index and confidence limits
in MH1C1 cells.

Mixture    RA ([mu]M)       0       0       0    125    250     375

           CA ([mu]M)     125     250     375      0      0       0

           FA ([mu]M)     375     250     125    375    250     125

CYP1A      Full sample  10335  207.82  313.33  >1000  >1000    0.76
activity   C1

           Median        4.58    5.85    7.41  >1000  >1000    0.76
           boot-strap
           C1

           Lower         5.09   43.46    8.54  >1000  >1000    0.66
           confidence
           limit

           Upper        >1000   >1000   >1000  >1000  >1000    0.86
           confidence
           limit

CYP2B/3A   Full sample   2.67    0.04    0.05  >1000  >1000    0.57
activity   C1

           Median boot   2.30    0.04    0.05  >1000  >1000    0.58
           strap C1

           Lower         0.33    0.03    0.03  >1000  >1000    0.49
           confidence
           limit

           Upper         8.12    0.04    0.59  >1000  >1000    0.70
           confidence
           limit

Rhodamine  Full sample   0.70    0.90    0.66   1.12   1.01    0.90
123        C1
retention

           Median        0.70    0.90    0.63   1.11   0.97    0.87
           boot-strap
           C1

           lower         0.48    0.71    0.62   0.84   0.88    0.84
           confidence
           limit

           Upper         1.01    1.22    0.79   1.76   1.92   >1000
           confidence
           limit

Neutral    Full sample   0.40    1.16   >1000  >1000  >1000    0.03
lipids     C1

           Median        0.33    0.98   >1000  >10OO  >1000    0.01
           boot-strap
           C1

           Lower         0.02    0.51    3.57   2.80   2.74       0
           confidence
           limit

           Upper         4.25    4.89   >1000  >1000  >1000   >1000
           confidence
           limit

Oxidative  Full sample   0.86    0.94    0.45   0.53   1.56   >1000
stress     C1

           Median        0.88    0.97    0.43   0.59   1.56   >1000
           boot-strap
           C1

           Lower         0.73    0.81    0.13   0.37   1.06   >1000
           confidence
           limit

           Upper         1.00    1.07    0.53   0.70   2.45   >1000
           confidence
           limit

Polar      Full sample   0.06    1.10    2.55  >1000  >1000    0.65
lipids     C1

           Median        0.04    1.10    2.56  >1000  >1000    0.20
           hoot-strap
           C1

           Lower            0       0    0.87   1.26   0.66       0
           confidence
           limit

           Upper        >1000   >1000   >1000  >I000  >1000   >1000
           confidence
           limit

Total DNA  Full sample   0.06    0.47    0.49   0.84   1.28    3.08
           C1

           Median        0.02    0.53    0.89   0.80   2.37  116.13
           boot-strap
           C1

           Lower            0       0    0.25   0.37   0.36    0.39
           confidence
           limit

           Upper         0.92    1.85    0.67   1.38   7.06  263.03
           confidence
           limit

Mixture    RA ([mu]M)     375    250    125  166.6

           CA ([mu]M)     125    250    375  166.6

           FA ([mu]M)       0      0      0  166.6

CYP1A      Full sample   0.95      0   0.51   0.59
activity   C1

           Median        0.98   0.04   0.67   0.63
           boot-strap
           C1

           Lower         0.81      0      0   0.47
           confidence
           limit

           Upper         1.02   0.04  >1000   0.63
           confidence
           limit

CYP2B/3A   Full sample   0.81  >1000   0.06   0.62
activity   C1

           Median boot   0.80  >1000   0.06   0.62
           strap C1

           Lower         0.73   0.05   0.05   0.58
           confidence
           limit

           Upper         0.90  >1000   0.06   0.67
           confidence
           limit

Rhodamine  Full sample   0.60   0.66   0.67   0.82
123        C1
retention

           Median        0.60   0.66   0.67   0.77
           boot-strap
           C1

           lower         0.55   0.60   0.62   0.76
           confidence
           limit

           Upper         0.67   0.72   0.74   0.93
           confidence
           limit

Neutral    Full sample      0   0.24   0.62   0.04
lipids     C1

           Median           0   0.22   0.70   0.02
           boot-strap
           C1

           Lower            0      0      0      0
           confidence
           limit

           Upper            0   0.89   0.90   1.57
           confidence
           limit

Oxidative  Full sample  >1000   0.49   0.46  >1000
stress     C1

           Median       >1000   0.49   0.47  >1000
           boot-strap
           C1

           Lower        >1000   0.40   0.31  >1000
           confidence
           limit

           Upper        >1000   0.60   0.63  >1000
           confidence
           limit

Polar      Full sample      0   0.67   0.45   0.03
lipids     C1

           Median           0   0.70   0.48   0.01
           hoot-strap
           C1

           Lower            0      0      0      0
           confidence
           limit

           Upper            0   1.14   1.04  >1000
           confidence
           limit

Total DNA  Full sample   0.85   0.75   0.44   2.72
           C1

           Median        0.51   0.61   0.46   3.87
           boot-strap
           C1

           Lower         0.52   0.51   0.17   0.08
           confidence
           limit

           Upper         1.73   4.77   0.78  >1000
           confidence
           limit


Differences were observed between the human and rat cell lines that could reflect species differences. For MH1C1 cells, RA and CA also exerted exaggerated effects on CYP1A activity, but to a much lesser extent. This phenomenon can be attributed to the different basal expression and inducibility of CYP1A in HepG2/C3A and MH1C1 cells. It is well known that CYP450 induction responses in the rat differ from those in the human due to sequence differences in the ligand domain of the nuclear receptor genes and CYP response elements (Lin 2006). Rodent cell lines have different CYP1A, CYP2B and CYP3A expression levels and inducibilities compared to human cell lines (Westerink et al. 2008). Therefore, it is important to include both cell lines in the early-phase screening for drug-drug interactions.

In summary, our data show that phenolic acids to which humans are routinely exposed in the diet or through dietary supplements, can interact synergistically or antagonistically on various aspects of normal liver metabolism in ways that could impact their health benefits or health hazards. In vitro assays using both human and rat hepatocytes along with data analysis using the Combination Index allow the objective determination of potential interactions between multiple food-related chemicals at multiple concentration combinations so that more relevant benefit and risk analyses can be conducted on chemically complex products.

Conflict of interest statement

The authors declare no conflicts of interest.

Acknowledgment

The authors thank Mr. Michael O'Donnell, Jr. for helpful discussions on statistics and experimental design.

0944-7113/S - see front matter. Published by Elsevier GmbH. Imp://dx.doi.org/10.1016/j.phymed.2012.12.013

Abbreviations: RA, rosmarinic acid; CA, caffeic acid; FA. ferulic acid; Cl. Combination Index; CYP1A, cytochromes P4501A1 and P4501A2; CYP21313A. cytochromes P45028 and P4503A.

* Corresponding author at: U.S. Food and Drug Administration, MOD-1 Laboratories, 8301 Muirkirk Road. Laurel, MD 20708, United States. Tel.: +1 301 210 6377: fax: +1 301 2104699.

E-mail address: thomas.flynniofda.hhs.gov (T.J. Flynn).

References

Baba, S., Osakabe, N., Natsume, M., Yasuda, A., Moto. Y., Hiyoshi, K., Takano, H., Yoshikawa. T., Terao. J., 2005. Absorption, metabolism, degradation and urinary excretion of rosmarinic acid after intake of Perilla frutescens extract in humans. European Journal of Nutrition 44. 1-9.

Berenbaum, M.C., 1989. What is synergy? Pharmacological Review 41.93-141. Clifford. M.N., 1999. Chlorogenic acids and other cinnamates--nature, occurrence and dietary burden. Journal of the Science of Food and Agriculture 79. 362-372.

De Kok, T.M., van Breda, S.C., Manson, M.M., 2008. Mechanisms of combined action of different chemopreventive dietary compounds: a review. European Journal of Nutrition 47 (Suppl. 2), 51-59.

Donato, M.T., Bassi, A.M., Gomez-Lechon, Penco, S., Herrero, E., Adamo, D., Castell, J.V., Ferro, M., 1994. Evaluation of the xenobiotic biotransformation capability of six rodent hepatoma cell lines in comparison with rat hepatocytes. In Vitro Cellular and Developmental Biology 30A. 574-580.

Flynn, T.J., Ferguson. M.S., 2008. Multiendpoint mechanistic profiling of hepatotoxicants in HepG2/C3A human hepatoma cells and novel statistical approaches for development of a prediction model for acute hepatotoxicity. Toxicology In Vitro 22,1618-1631.

Kelly, J.H., 1994. Permanent human hepatocyte cell line and its use in a liver assist device (LAD). US Patent No. 5290684.

Khalif, A., Schantz, S.P., Chou, T.-C., Edelstein, D., Sacks, P.G., 1998. Quantitation of chemopreventive synergism between (-)-epigallocatechin-3-gallate and curcumin in normal. premalignant and malignant human oral epithelial cells. Carcinogenesis 19,419-424.

Lin, J.H., 2006. CYP induction-mediated drug interactions: in vitro assessment and clinical implications. Pharmaceutical Research 23. 1089-1116.

Liu, R.H., 2004. Potential synergy of phytochemicals in cancer prevention: mechanisms of action. Journal of Nutrition 134. 3479S-3485S.

Liu, Y., Flynn, T.J., Ferguson, M.S., Hoagland. EM., Yu, L., 2011. Effects of dietary phenolics and botanical extracts on hepatotoxicity-related endpoints in human and rat hepatoma cells and statistical models for prediction of hepatotoxicity. Food and Chemical Toxicology 49, 1820-1827.

Mateos, R., Goya, L, Bravo, L. 2006. Uptake and metabolism of hydroxycinnamic acids (chlorogenic, caffeic, and ferulic acids) by HepG2 cells as a model of the human liver. journal of Agricultural and Food Chemistry 54, 8724-8732.

Maurya, D.K., Devasagayam, T.P., 2010. Antioxidant and prooxidant nature of hydroxycinnamic acid derivatives ferulic and caffeic acids. Food and Chemical Toxicology 48, 3369-3373.

McDermott, C., Allshire, A., van Pelt, F., Heffron, J.J.A., 2008. In vitro exposure of Jurkat T-cells to industrially important organic solvents in binary combination: interaction analysis. Toxicological Sciences 101, 263-274.

Millard, S.M., Neerchal, N.K., 2000. Environmental Statistics Using S-PLUS. Chapman Hall/CRC Press, Boca Raton, pp. 261-263.

Nahrstedt, A., Butterweck, V., 2010. Lessons learned from herbal medicinal products: the example of St. John's wort. journal of Natural Products 73, 1015-1021. Nakazawa, T., Ohsawa, K., 1998. Metabolism of rosmarinic acid in rats. journal of Natural Products 61, 993-996.

Ovaskainen, M.-L., Torronen, R., Koponen, J.M., Sinkko, H., Hellstrom, J., Reinivuo, H., Mattila, P., 2008. Dietary intake and major food sources of polyphenols in Finnish adults. Journal of Nutrition 138, 562-566.

Petersen, M., Simmonds, MS., 2003. Rosmarinic acid. Phytochemistry 62, 121-125.

Richardson, U.I., Tashjian Jr., Levine, L., 1969. Establishment of a clonal strain of hepatoma cells which secrete albumin, journal of Cell Biology 40, 236-247.

Saw, C.L-L. Cintron. M., Wu, T.-Y., Guo, Y., Huang, Y., Jeong, W.-S., Knog, A.-N.T., 2011. Pharmacodynamics of dietary phytochemical indoles I3C and DIM: induction of Nrf2-mediated phase II drug metabolizing and antioxidant genes and synergism with isothiocyanates. Biopharmaceutics and Drug Disposition 32, 289-300.

Scalbert, A., Williamson, G., 2000. Dietary intake and bioavailability of polyphenols. Journal of Nutrition 130, 2073S-2085S.

Schamhart, D.H., van de Poll, K.W., van Wijk, R., 1979. Comparative studies of glucose metabolism in HTC, RLC, MH1C1, and Reuber H35 rat hepatoma cells. Cancer Research 39, 1051-1055.

Seefeldt, S.S., Jensen, J.E., Fuerst, E.P., 1995. Log-logistic analysis of herbicide dose-response relationships. Weed Technology 9, 218-227.

Srinivasan, M., Rukkumani, R., Ram Sudheer, A., Menon, V.P., 2005. Ferulic acid, a natural protector against carbon tetrachloride-induced toxicity. Fundamental and Clinical Pharmacology 19, 491-496.

Tallarida, R.J., 2001. Drug synergism: its detection and applications. Journal of Pharmacology and Experimental Therapeutics 298, 865-872.

Wagner, H., Ulrich-Merzenich, G., 2009. Synergy research: approaching a new generation of phytopharmaceuticals. Phytomedicine 16, 97-110.

Westerink, W.M., Stevenson. J.C., Schoonen, W.G., 2008. Pharmacologic profiling of human and rat cytochrome P450 1A1 and 1A2 induction and competition. Archives of Toxicology 82, 909-921.

Williams, S.N., Pickwell, G.V., Quattrochi, LC., 2003. A combination of tea (Camellia senensis) catechins is required for optimal inhibition of induced CYP1A expression by green tea extract. Journal of Agricultural and Food Chemistry 51, 6627-6634.

Zhao, L, Wientjes. G., Au, J.L.S., 2004. Evaluation of combination chemotherapy: integration of nonlinear regression, curve shift, isobologram, and combination index analyses. Clinical Cancer Research 10, 7994-8004.

Zhou, S.F., Wang, B., Yang, L.P., Liu, J.P., 2010. Structure, function, regulation and polymorphism and the clinical significance of human cytochrome P450 1A2. Drug Metabolism Reviews 42. 268-354.

(1.) Current address: Allen County Kentucky School, Scottsville, KY 42164. United States.

Yitong Liu (a), (c), Thomas J. Flynn (a), (*), Martine S. Ferguson (b), Erica M. Hoagland (a), (1)

(a.) Division of Toxicology. Center for Food Safety and Applied Nutrition, U.S. Food and Drug Administration, Laurel, MD 20708, United States

(b.) Division of Public Health and Biostatistics, Center for Food Safety and Applied Nutrition, U.S. Food and Drug Administration, College Park, MD 20740, United States

(c.) Oak Ridge Institute for Science and Education. Oak Ridge, TN 37831, United States
COPYRIGHT 2013 Urban & Fischer Verlag
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2013 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Liu, Yitong; Flynn, Thomas J.; Ferguson, Martine S.; Hoagland, Erica M.
Publication:Phytomedicine: International Journal of Phytotherapy & Phytopharmacology
Article Type:Report
Geographic Code:1USA
Date:May 15, 2013
Words:8348
Previous Article:Hesperidin attenuates cisplatin-induced acute renal injury by decreasing oxidative stress, inflammation and DNA damage.
Next Article:ESCOP on-line: new monographs available!
Topics:

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters