A Comparative Study upon the Therapeutic Indices of Some Natural and Synthetic Anti-inflammatory Agents

Headnote

Abstract

Objective(s)

The aim of the present study was to compare the therapeutic indices of several agents used in treatment of inflammatory conditions which included: Vitamin E, hydro-alcoholic extracts of Glycyrrhiza glabra, Matricaria aurea, dexamethasone, piroxicam and diclofenac using Wehi-164 fibrosarcoma cells.

Materials and Methods

Cytotoxicity evaluation was based on vital dye exclusion assay. Matrix-metalloproteinases inhibition (MMPI) was assessed by gelatinase zymography method. The collected data were used to estimate the IC^sub 50^ (50% MMPI concentration), LC^sub 50^ (50% cytotoxicity concentration) and the therapeutic index (LC^sub 50^/IC^sub 50^).

Results

Among the natural anti-inflammatory agents used, M. aurea was the least toxic and the most effective inhibitor of MMP. Vitamin E not only increased MMP activity, but also was the most toxic of all the agents tested. Next in terms of toxicity to vitamin E was G. glabra. Diclofenac was more toxic than both piroxicam and dexamethasone.

Conclusion

Findings from this study suggest that medicinal plants reputed to have anti-inflammatory properties are not equally effective and safe. In order to assess the implications of these findings, further in vitro and in vivo studies are needed.

Keywords: Antinflammatory agents, Cytotoxicity, Gelatinase, Matrix-metalloproteinases, Therapuetic Index, Zymography

Introduction

Matrix-metalloproteinases (MMPs) are a family of multi-domain, zinc- and calciumdependent enzymes produced by a wide range of stromal and inflammatory cells and are capable of degrading a variety of components of the extra cellular matrix. The role of MMPs has been well established in a variety of inflammatory diseases such as arthritis (1) cardiovascular (2), metastasis of cancer (3) and in angiogenesis (4). Furthermore, MMPs (mainly membrane-type MMPs) have an important physiological role in embryo implantation, bone remodeling and organogenesis (5). MMPs can be constitutive (homeostatic MMPs), which are switched on in most cells under steady-state conditions, or inducible (pro-inflammatory MMPs), which have more complex promoter region and regulated by various agonists like in invasion of cancer cells. These findings highlight the complex nature of MMPs.

The biological activity involving MMPs are always dependent on maintaining a balance between proteinases and natural tissue inhibitors of matrix metalloproteinases (TIMPs), such as serum α^sub 2^-macroglobulin (6), monoclonal antibodies, and small molecules (both synthetic and natural products). Modulation of MMP activity, therefore, seems to be a useful therapeutic option in the treatment of both chronic and acute inflammatory conditions in which excess MMP activity has been implicated. However, despite their high potencies, MMPIs act nonspecifically which leads to side effects (7). The search for finding more selective and potent agents for clinical application in various inflammatory conditions is still continuing.

On the other hand, different classes of antiinflammatory drugs have been employed for treatment of a variety of inflammatory diseases which includes steroidal and non-steroidal as well as drugs obtained from medicinal plants (8). Because of their efficacy in the treatment of various inflammatory conditions, both steroidal and non-steroidal anti-inflammatory drugs (NSAIDs) are of the most frequently used classes of medicines in the world and account for more than 5% of all prescribed medications (9). In addition, the use of these agents is associated with various side effects such as gastrointestinal complications, and with prolonged use of corticosteroids, reduction in immune responses and dysfunction of HPA axis. Furthermore, there are few and limited controlled in vitro or clinical trials on the effects of herbal products. Therefore, their safety and efficacy have been always questioned by the approving authorities. However, anecdotal evidences suggest that herbs can be useful in the treatment of a variety of inflammatory conditions such as periodontal diseases (10) and in experimentally-induced granulation pouch (11). Among the ancient medicinal herbs that are still used for various inflammatory conditions are Glycyrrhiza glabra (12) and Matricaria aurea. While G. glabra has been suggested to be useful in the treatment of cancer (12), and in protection of liver following experimentally induced carbon-tetrachloride toxicity (13). M. aurea has been shown to have strong antioxidant properties (14) and traditionally used in case of urinary retention, urinary tract infections and kidney and testicle pain (15). Vitamin E, another potentially useful antioxidant agent, was found to contribute to the prevention of atherogenesis and restenosis arising from vascular injury associated with MMP-1 activation via its antioxidant activity (16).

Despite these findings, no previous study has attempted to correlate the MMPI activities of these synthetic and natural anti-inflammatory agents with their cytotoxicity and estimation of their therapeutic indices. The aims of the present study were twofold: Firstly, to compare IC^sub 50^ values of dexamethasone, piroxicam and diclofenac as well as three natural agents reported to have antiinflammatory properties: vitamin E, hydroalcoholic extract of G. glabra and M. aurea. Previous studies have shown a close relationship between MMP2 and MMP9 inhibition with anti-inflammatory properties of some of these agents used in this study (17). Secondly, the cytotoxicity produced by the above agents was compared by estimation of their LC^sub 50^ values. The aim of such comparative studies was to provide scientifically-based evidence of the therapeutic safety margins of agents which may be administered in the treatment of an inflammatory disease. Prior to becoming a drug for therapeutic use, a medicinal agent, synthetic or of herbal origin, vigorous in vitro and in vivo studies are needed to determine its safety and efficacy.

Materials and Method

Cell culture

The fibrosarcoma (Wehi-164, National Cell Bank of Iran, Pasteur Institute, Tehran) cell line was seeded at initial density of 20,000 cells/well in 96-well tissue culture plate. Cells were maintained in RPMI-1640 medium supplemented with 5% fetal calf serum, penicillin (100 units/ml) and streptomycin (100 µg/ml) under 5% CO2, 37 °C and saturated humidity (18).

Preparation of hydro-alcoholic extract of G. glabra and M. aurea (purchased from commercial market). The plant materials were identified by an expert botanist. The roots of G. glabra were cut and grinded into powder. Oven-dried M. aurea flowers were directly grinded. For each stage of extraction, 100 g of the powdered plant product was moistened for 2 hr with 70% ethanol solution and allowed to swell before being placed in one of a series of percolation chambers. The materials were repeatedly rinsed with 70% alcoholic solvent until the solvent became clear which required 24 to 48 hr of percolation. The collected percolation products were vacuum-dried and the extracts were dissolved and diluted in PBS to the desired concentrations. Percentages obtained for both extracts were approximately 7% (19).

Dose-response analysis

Sterile solutions of triplicate of two fold dilutions in PBS were prepared from pure substances of disodium dexamethasone sulphate (purchased from Daru Pahksh Pharmaceutical Co. Iran), piroxicam (purchased from Poorsina Pharmaceutical Co. Iran), diclofenac Na (purchased from Daru Pahksh Pharmaceutical Co, Iran), at concentrations of 10-200 µg/ml, vitamin E (dl-a-tocophirol acetate) (1 to 50 µg/ml) (purchased from Osveh Pharmaceutical Co, Iran), and G. glabra and M. aurea extracts (8 to 8000 µg/ml), were transferred to cultured cells. Non-treated cells, exposed to PBS alone, were used as control. Cells were cultured for 24 hr and then were subjected to colorimetric assay.

Vital dye exclusion cytotoxicity assay

Cell cultures were washed three times with ice-cold PBS, followed by fixation in a 5% formaldehyde solution. Fixed cells were washed three times and stained with 1% crystal-violet. Stained cells were washed, lysed and solubilized with 33.3% acetic acid solution. The density of developed purple color was read at 580 nm by colorimetric method (Tecan, Sunrise, Remote Touch Screen, Austria). The extent of color density was considered as the degree of cytotoxicity relative to background reading values for the control untreated cell culture (0% cytotoxicity). Furthermore, reference background readings for the plant extracts were subtracted from the relevant plant readings. Data from linear regression analysis of concentration-response curves were used in order to estimate the concentration that produced 50% cytotoxicity (LC^sub 50^).

Zymography

This technique was used to determine gelatinase A (MMP-2) and gelatinase B (MMP-9) activity in conditioned media according to modified Heussen and Dowdle method (18). Briefly, aliquots of conditioned media were subjected to electrophoresis in (2 mg/ml) gelatin-containing polyacrylamide gels in the presence of sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS- PAGE) under non-reducing conditions. The gels underwent electrophoresis for 3 hr at a constant voltage of 80 V. After electrophoresis, gels were washed and gently shaken in three consecutive washings in 2.5% Triton X-100 solution to remove SDS. Gel slabs were then incubated at 37 °C for 24 hr in 0.1 M Tris-HCl gelatinase-activation buffer (pH= 7.4) containing 10 mM CaCl2 and subsequently stained with 0.5% Coomassie Blue. After intensive destaining, proteolysis areas appeared as clear bands against a blue background. Quantitative evaluation of both surface and intensity of lysis bands were relatively compared to non-treated control wells and expressed as % MMPI relative to control gelatinase activity. Values obtained from standard references for gelatinase A and B and IC^sub 50^ values were determined from the linear regression analysis of concentrationresponse curves.

Statistical analysis

All experiments were performed in triplicates. The graphic data were presented as percentage of mathematical means relative to control baseline values before addition of agents tested ± SEM. In evaluation of differences between recorded data the P-value< 0.05 was considered significant using ANOVA followed by Dunnets' post-hoc test. Prism software was employed to produce a linear regression analysis from the concentration-response curves and determination of both LC^sub 50^ and IC^sub 50^ by extrapulation of the mathematical module: Y=bX + a, where the coefficients a (intercept) and b (slope) and X is the corresponding concentration at which Y (LC^sub 50^ or IC^sub 50^) occurred.

Results

Comparison of cytotoxic effects (LC^sub 50^ values)

Figure 1 and Table 1 show the concentrationpercentage relationships of viable Wehi-164 cells curves relative to untreated control group and the corresponding LC^sub 50^ values of diclofenac, piroxicam, and dexamethasone, respectively. LC^sub 50^ values for vitamin E, diclofenac, dexamethasone and piroxicam were 25, 82.3, 104, 131 µg/ml, respectively. Corresponding LC^sub 50^ values for G. glabra and M. aurea hydro-alcoholic extracts were 465 and 1305 µg/ml, respectively (Table 1 and Figure 1). LC^sub 50^ values of the plant extracts against Wehi-164 fibrosarcoma, as a sensitive cell line (17), were significantly lower relative to vitamin E, dexamethasone, piroxicam, and diclofenac (Table 1).

Comparison of inhibition of MMP (MMPI) activities

The IC^sub 50^ value of hydro-alcoholic extract of G. glabra was 2.59 fold higher than that of M. aurea (9288 versus 3588 µg/ml) (Figure 2 and Table 1), while the corresponding values for diclofenac, dexamethasone and piroxicam were 83.6, 136.3, 143.8 µg/ml respectively (Table 1). The percentage of MMPI activity of diclofenac, over all concentration range used (10-200 µg/ml) was significantly greater than those recorded for piroxicam and dexamethasone (Figure 1). Vitamin E produced an increase in the MMP activities at the concentration range of 1 to 10 µg/ml. This increase in MMP activity was found to be maximal at 10 µg/ml reaching 8 fold higher than baseline value recorded for untreated control cells. At concentrations above 10 µg/ml, vitamin E produced a gradual reduction from the maximum reaching 400% of the baseline value at 50 µg/ml (Figure 1).

Comparison of the therapeutic indices (LC^sub 50^ / IC^sub 50^ ratios)

The ratios of the cytotoxicity (LC^sub 50^) to antiinflammatory effectiveness (IC^sub 50^) of the studied agents were used as measure for the therapeutic index. The results showed that vitamin E was the most toxic of all agents used; not only it increased the MMP activity but also it had the lowest LC^sub 50^ value at 25 µg/ml. On the other hand, G. glabra was found to be the second in line to vitamin E in terms of its LC^sub 50^/IC^sub 50^ ratio, having a therapeutic index of 0.05, while diclofenac was the safest with a therapeutic index of 0.98 (Table 1).

Discussion

In this study, using Wahi-164 fibrosarcoma cell line, zymography method of analysis and vital dye cytotoxicity exclusion assay, the therapeutic indices (LC^sub 50^/IC^sub 50^ ratios) of two NSAIDs namely diclofenac, piroxicam, and a steroidal agent, dexamethasone, were measured and compared with those produced by vitamin E and hydro-alcoholic extracts of G. glabra and M. aurea. The results showed that the descending order of IC^sub 50^ values (MMPI efficacy) was in the following sequence: diclofenac > dexamethasone > piroxicam > M. aurea > G. glabra. Contrary to these agents, vitamin E (10 µg/ml) produced an eight fold increase in MMP activity. Furthermore, the descending order of LC^sub 50^ values (cytotoxic effect) was: vitamin E > diclofenac > dexamethasone > piroxicam > G. glabra> M. aurea. However, when comparative therapeutic indices of these agents were calculated, diclofenac was found to be the most efficacious and the least toxic among the tested compounds. The descending order of these ratios was: diclofenac > piroxicam > dexamethasone > M. aurea >G. glabra > vitamin E.

Pharmacologically NSAIDs are clinically proved effective anti-inflammatory drugs and are thought to act through inhibition of cyclooxygenase enzyme. Other studies have linked their anti-inflammatory action with their MMPI effects (17). However, other mechanisms explaining their cytotoxicity have been proposed: While Saunders et al (21), postulated that NSAIDs cause depletion of intracellular polyamines - naturally occurring growth factors found in all cells - thus leading to decrease in growth and death in cancer cells while others suggested a direct increase in permeabilization of cell membrane to calcium ions (22).

Findings from our study showed a relative close correlation between MMPI activities and the cytotoxic effects of the tested NSAIDs. However, is there any direct relationship between cytotoxicity and MMPI activity? This seems to be unlikely for three reasons: Firstly, MMPs are known to be secretory products released under both physiological and pathological conditions, while cellular death is orchestrated by intracellular processes leading to apoptosis or necrosis. Secondly, following the disappointing outcomes of MMPIs as modulators of cellular growth in treatment of cancer (23), these agents have been advocated, with different degrees of success, in the treatment of various acute and chronic inflammatory conditions (24). Thirdly, as shown from our results, vitamin E was found to cause an increase in MMP activity, yet it also produced cytotoxicity. These observations suggest that inhibition of MMP activity is not related to the cytotoxicity and these two processes are orchestrated independently by different mechanisms.

The results from estimated LC^sub 50^ and IC^sub 50^ values indicate that there are differences between the two hydro-alcoholic extracts from G. glabra roots and M. aurea flowers. G. glabra was three fold more cytotoxic (LC^sub 50^ 465 versus 1450) and 2.6 fold less potent as an anti-inflammatory natural product (IC^sub 50^ of 9288 versus 3584) than M. aurea. Consequently, M. aurea is shown to be a safer and a more effective anti-inflammatory agent of natural origin than G. glabra but less effective than conventionally used NSAIDs and dexamethasone. Prudently, further in vivo and in vitro studies on the anti-inflammatory properties of the active constitutes of this these medicinal plant are needed. On the other hand, the findings from examining extract from G. glabara comes in agreement with a recent report in which the aqueous extract of G. glabra was found to have significant, in vivo and in vitro anti-proliferative and antiangiogenesis effects on Ehrlich ascites tumor cells (12). Based on these findings it seems that the extract from this medicinal plant has a potential role as a supplemental source in cancer therapy.

Despite extensive researches on vitamin E, its exact role in prevention or treatment of a given inflammatory disease and on the mechanism by which it produces the proposed effects is not clearly defined. While it has been regarded merely as an antioxidant (25), others dispute this as the main mechanism of action and argue that it has no in vivo protective effect against oxidative damage or prevention of diseases which have at their basis an oxidative insult (26). Furthermore, in our study, in contrast to other tested agents, findings from vitamin E experiments showed that it increased MMP activity. This finding comes in agreement with those reported by Chakraborti et al where pretreatment of the microsomes with vitamin E (1 mM) preserved the increase in MMP-2 activity, Ca2 +ATPase activity and also ATP-dependent Ca2 + uptake in the microsomes following exposure to peroxynitrite (27).

The other unexpected aspect of vitamin E was its cytotoxic property. It was found to have the lowest LC^sub 50^ and accordingly was the most toxic of all the tested agents. Antiproliferative effects of vitamin E were previously reported for neuroblastoma cells (28) and different formulations of vitamin E produced varying degrees of apoptosis (29, 30). The exact mechanism that mediate vitamin E cytotoxic effects can not be clarified from present experiments and further works need to be undertaken. However, it is possible that vitamin E may have an indirect effect via interaction with the pro-antioxidants in the culture media. These effects were not tested in our experiments and need further assessments.

MTT, LDH leakage, and dye exclusion assays are the most commonly used cytotoxicity assays in in vitro toxicological studies on different cell lines. However, previous studies have shown that these tests produced variable results, which are attributed to the differences in the sensitivities of the cell lines employed as well as to the agents used (31). The toxicity results from our studies may not be conclusive. In order to increase the reliability of these results and to distinguish between cytotoxicity and organelle toxicity, measurement of other inflammatory markers such as NFkB transcriptional activity, TNFa or IL-1, and different cells lines such as Jurkatt and other lymphoid cell lines using MTT and other standard tests need to be carried out. Despite these limitations, the results of this study showed that the used agents clearly have different inhibitory potencies on MMP2 and MMP9 isoenzymes produced by Wehi 164 fibrosarcoma cells which are known for their high yield of MMP secretory activity (32).

It is well known that the duration of incubation, the concentrations of the test agents as well as the type of employed dye are among the important detrimental experimental conditions that need to be considered carefully when performing cytotoxicity tests (31, 33). In order to limit the effects of these confounding factors, we allocated the 24 hr as the fixed time point and any cellular damage that has occurred at the IC^sub 50^ or EC50 concentrations of the tested agents was assumed to be similar. Previous studies using 24 hr incubation period was found reliable for measurement of IC^sub 50^ values (33) and, in our preliminary experiments, this time point was found to be suitable and gave reproducible results. These findings are merely a small step in proposing a simple method for evaluating the safety and efficacy of various anti-inflammatory agents both from synthetic and natural sources.

What is the relevance of these findings to clinical practice? Inhibition of a selected MMP may have a potential beneficial role in the treatment of inflammatory-based conditions. To achieve such a target, ideally we need to have a drug which selectively, safely and effectively inhibits a specifically implicated MMP. In this in vitro study we used the therapeutic index as an index for assessment of safety profile and showed that, among the tested agents, the hydro-alcoholic extract of M. aurea was a safer inhibitor of MMP-2 and MMP-9 than vitamin E and G. glabra. Prudently, more detailed studies on the active constituent(s) of this medicinal plant need to be undertaken in order to utilize it in clinical practice. On the other hand, herbal treatments are becoming increasingly popular, are often used in complementary medicine for treatment of various ailments and conditions, and are often associated with an intuitive feeling that "naturalness" is a guarantee of harmlessness. The findings from this study have shown that a product from natural origin does not simply means that it is a safe agent. Using these simple in vitro tests, we have shown that vitamin E and G. glabara are potent cytotoxic agents and highlight the importance of vigorous preclinical investigations on the possible adverse effects associated with herbal or natural agents. Such studies may pave the way for better understanding of the potential role of the herbal agents as therapeutic tools in modern health systems.

Acknowledgment

The authors wish to express their thanks for the financial support of Deputy for Research Affairs of Ahvaz Jundishapur University of Medical Education, Ahvaz, Iran.

References

References

1. DiMartino MJ, High W, Galloway WA, Crimmin MJ. Preclinical antiartherric activity of matrix metalloproteinase inhibitors. Ann N Y Acad Sci 1994; 732:413-413.

2. Janssens S, Lijinen HR. What has been learned about the cardiovascular effects of matrix metalloproteinases from moue models? Cardiovasc Res 2006; 69:585-594.

3. Egelblad M, Werb Z. New functions for matrix metalloproteinases in cancer progression. Nat Rev Cancer 2002; 2:161-174.

4. Johnson C, Sung HJ, Lessner SM, Fini ME, Galis ZS. Matrix metalloproteinase-9 is required for adequate angiogenic revascularization of ischemic tissues: potential role in capillary branching. Circ Res 2004; 94:262- 268.

5. Barrett AJ, Rawlings ND, Woessnar JF. Handbook of Proteolytic Enzymes. Academic Press; 2004.

6. Woessner JF Nagase H. I nhibition of the MMPs. Textbook of Matrix Metalloproteinases and TIMPs. Oxford University Press; 2000.

7. Chirco R, Liu XW, Jung KK, Kim HR. Novel functions of TIMPs in cell signaling. Cancer Metastasis Rev 2006; 25:99-113.

8. Whittaker M, Floyd CD, Brown P, Gearing AJ. Design and therapeutic application of matrix metalloproteinase inhibitors. Chem Rev 1999; 99:2735-2776.

9. Smalley WE, Ray WA, Daugherty JR, Griffin MR. Nonsteroidal anti-inflammatory drugs and the incidence of hospitalizations for peptic ulcer disease in elderly persons. Am J Epidemiol 1995; 141:539-545.

10. Tipton DA, Lyle B, Babich H, Dabbous MKh. In vitro cytotoxicity and anti-inflammatory effects of myrrh oil on human gingival fibroblasts and epithelial cells. In Vitro Toxicol 2003; 17:301-310

11. Kimura I, Yoshikawa M, Kobayashi S, Sugihara Y, Suzuki M, Oominami H, et al. New triterpenes, myrrhanol A and myrrhanone A, from guggulgum resin and their potent anti-inflammatory effect on adjuvant-induced air pouch granuloma of mice. Bioorg Med Chem Lett 2001; 11:985-989.

12. Sheela ML, Ramakrishna MK, Salimath BP. Angiogenic and proliferative effects of the cytokine VEGF in Ehrlich ascites tumor cells is inhibited by Glycyrrhiza glabra. Int Immunopharmacol 2006; 6:494-498.

13. Rajesh MG, Latha MS. Protective activity of Glycyrrhiza glabra Linn. on carbon tetrachloride-induced peroxidative damage. Indian J Pharmacol 2004; 36:284-287.

14. Al-Ismail KM, Aburjai T. Antioxidant activity of water and alcohol extracts of chamomile flowers, anise seeds and dill seeds. J Sci Food Agric 2004; 84:173-178.

15. Abu-Rabia A. Palestinian plant medicines for treating renal disorders: An inventory and brief history. Alter Complem Ther? 2005; 11:295-300.

16. Orbe J, Rodri'guez JA, Arias R, Belzunce M, Nespereira B, Pe'rez-Ilzarbe M, et al. Antioxidant vitamins increase the collagen content and reduce MMP-1 in a porcine model of atherosclerosis: implications for plaque stabilization. Atherosclerosis 2003; 167:45-53.

17. Mirshafiey A, Khodadadi A, Rehmy BH, Khorramizadeh MR, Eslami MR, Razavi A, et al. Sodium alginate as a novel therapeutic option in experimental colitis. Scand J Immunol 2005; 61:316-321.

18. Mosmann T. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J Immunol Methods 1983; 65:53-63.

19. Williamson EM, Okapako DT, Evans FJ. Pharmacological Methods in phytotherapy research, volume 1: Selection, preparation and pharmacological evaluation of plant material. John Willey and Sons; Chichester? 1989.p.15-23.

20. Heussen C, Dowdle EB. Electrophoretic analysis of plasminogen activator in polyacrylamide gels containing sodium dodecyl sulfate and copolymerized substrates. Ann Biochem 1980; 102:196-202.

21. Saunders FR, Hughes A, Heather M, Wallace HM. Does polyamine depletion influence NSAID-induced cytotoxicity in colorectal cancer? Abstracts Toxicol 2006; 226:12-77.

22. Tomisato W, Tanaka K, Katsu T, Kakuta H, Sasaki K, Tsutsumi S, et al. Membrane permeabilization by non steroidal anti-inflammatory drugs. Biochem Biophys Res Commun 2004; 323: 1032-1039.

23. Overall CM, Lopez-Otin C. Strategies for inhibition of cancer: Innovation for the post-trial era. Nat Rev Cancer 2002; 2: 657-672.

24. Parks WC, Wilson CL, Lopez-Boado YS. Matrix metalloproteinases as modulators of inflammation and innate immunity. Nat Rev Immunol 2004; 4:617-629

25. Traber M, Atkinson J. Vitamin E: antioxidant and nothing more. Free Radic Biol Med 2007; 43:4-15.

26. Azzi A. Molecular mechanism of a-tocopherol action. Free Radic Biol Med 2007; 43: 16-21.

27. Chakraborti S, Mandal A, Das S, Chakrabort Ti. Inhibition of Na+/Ca2+ exchanger by peroxynitrite in microsomes of pulmonary smooth muscle: role of matrix metalloproteinase-2. Biochim Biophys Acta 2004; 1671:70- 78.

28. Slack RS. The effect of vitamin A and E on neuroblastoma growth and differentiation. Thesis submitted in partial fulfilment of the requirements for the degree of Doctorate of Philosophy in Biochemistry. Ottawa, Canada, 1991.

29. Theriault A, Chao JT, Wang Q, Gapor A, Adeli K. Tocotrienol: a review of its therapeutic potential. Clin Biochem 1999; 32: 309-319.

30. Kline K, Yu W, Sanders BG.Vitamin E: Mechanisms of action as tumor cell growth inhibitors. J Nutr 2001; 131:161S-163S.

31. Fotakis G, Timbrell JA. In vitro cytotoxicity assays: Comparison of LDH, neutral red, MTT and protein assay in hepatoma cell lines following exposure to cadmium chloride. Toxicol Lett 2006; 160:171-177.

32. Shariftabrizi A, Nifli AP, Ansari M, Saadat F, Ebrahimkhani MR, Alizadeh N, et al. Matrix metalloproteinase 2 secretion in WEHI 164 fibrosarcoma cells is nitric oxide-related and modified by morphine. Eur J Pharmacol 2006; 530:33-39.

33. Romero D, Gomez-Zapata M, Luna A, Garcia-Fernandez AJ.Morphological characterisation of BGM (Buffalo Green Monkey) cell line exposed to low doses of cadmium chloride. Toxicol In Vitro 2003; 17:293-299.