Inhibition of aspergillus flavus growth and aflatoxin b1 production in stored maize grains exposed to volatile compounds of trichoderma harzianum rifai.

Luisa Elena Mejía Agüero, Rafael Alvarado, Amaury Martínez and Blas Dorta

Luisa Elena Mejía Agüero. M.Sc. in Food Science and Technology, Universidad Central de Venezuela (UCV). Graduate Student, UCV, Venezuela.

Rafael Alvarado. Technician, Escuela Técnica Industrial Luis Caballero Mejías, Venezuela. Research Assistant, UCV, Venezuela.

Amaury Martínez. M.Sc. in Food Science and Technology, UCV, Venezuela. Professor, UCV, Venezuela.

Blas Dorta. Doctor in Biochemistry, Universidad Nacional de La Plata, Argentina. Professor, UCV, Venezuela. Address: Laboratorio de Procesos Fermentativos, Instituto de Biología Experimental, Facultad de Ciencias, UCV. Apartado 47114, Caracas 1050A, Venezuela. e-mail: bdorta@cantv.net.

SUMMARY

Inhibition of aflatoxin B1 production and A. flavus biomass accumulation in stored maize with volatile compounds (VCs) produced by T. harzianum was achieved under laboratory conditions. Gamma-irradiated (3Kgrey) maize grains inoculated with A. flavus were exposed to VCs produced by T. harzianum cultured for 15 days in Malt Extract Agar (MEA) or Defatted Corn Germ - Rice Husk (DCGRH) media. A. flavus biomass accumulation and aflatoxin B1 production decreased by 31.7% and 51.87%, respectively, when T. harzianum was cultured in MEA. No inhibitory effect was seen in DCGRH, at the same laboratory scale conditions. However, when VCs production on DCGRH medium was scaled 24´, the proliferation of A. flavus was significantly inhibited (90%). These results confirm the usefulness of T. harzianum as a biocontrol agent for toxigenic fungi in stored grains and its possible role in mycotoxin reduction.

Inhibición del crecimiento de aspergillus flavus y producción de aflatoxina b1 en granos de maíz almacenados bajo la acción de compuestos volátiles de trichoderma harzianum rifai.

Se demostró que el mecanismo basado en la acción de los compuestos volátiles (CVs) generados por Trichoderma harzianum, resultó factible en el control de crecimiento en biomasa de Aspergillus flavus y la producción de aflatoxina B1 en granos de maíz almacenados a escala de laboratorio. Los granos de maíz tratados con radiación gamma (3Kgrey) e inoculados posteriormente con A. flavus, fueron expuestos a los CVs producidos y acumulados por T. harzianum luego de 15 días de cultivo en los medios Agar Extracto de Malta (AEM) y Germen Desgrasado de Maíz - Cáscara de Arroz (GDMCA). Se observó una reducción de 31,7% en la biomasa de A. flavus acumulada en maíz y una reducción de 51,78% en la producción de aflatoxina B1 cuando se empleó el medio AEM para el cultivo de T. harzianum. No se observó inhibición cuando se empleó el medio GDMCA a la misma escala de laboratorio. Sin embargo, cuando la producción de CVs a partir del medio GDMCA fue escalada en 24´, la proliferación de A. flavus se redujo en un 90%. Los resultados obtenidos ponen de manifiesto el potencial de T. harzianum como agente biocontrolador de hongos micotoxigénicos en granos almacenados, y sus posibles implicaciones en la reducción de micotoxinas.

Inibição do crescimento de aspergillus flavus e produção de aflatoxina b1 em grãos de milho armazenados na presença de compostos voláteis de trichoderma harzianum rifai.

RESUMO

Demonstrou-se que o mecanismo baseado na acção dos compostos voláteis (CVs) gerados por Trichoderma harzianum, resultou viável no controlo do crescimento da biomassa de Aspergillus flavus e na produção de aflatoxina B1 nos grãos de milho armazenados a escala de laboratório. Os grãos de milho tratados com radiação gamma (3kgrey) e inoculados posteriormente com A. flavus foram expostos aos CVs produzidos e acumulados por T. harzianum logo após 15 dias de cultivo nos meios Agar Extracto de Malta (AEM) e Gérmen Desengordurado de Milho - Casca de Arroz (GDMCA). Observou-se uma redução de 31,7% na biomassa de A. flavus acumulada no milho, e uma redução de 51,78% na produção de aflatoxina B1 quando se aplicou ao meio AEM para o cultivo de Trichoderma harzianum. Não se observou inibição quando se aplicou ao meio GDMCA à mesma escala de laboratório. Por outro lado, quando a produção de CVs a partir do meio GDMCA foi a uma escala de 24×, a proliferação de A. flavus reduziu-se a 90%. Os resultados obtidos salientam e põem em evidência o potencial de T. harzianum como agente biocontrolador de fungos micotoxigenicos nos grãos armazenados, e as suas possíveis implicações na redução de micotoxinas.

KEYWORDS / Aflatoxin B1 / Asperguillus flavus / Biocontrol / Trichoderma harzianum / Volatile Compounds /

The widespread occurrence of mycotoxins in cereal grains for human consumption and animal feed, as reflected in many reports (Mora and Lacey, 1997; Abdullah et al., 1998; Dalcero et al., 1998; Shöllenberger et al., 1999; Dutta and Das, 2001) indicates an increased proliferation of filamentous toxigenic fungi on these foodstuffs.

Maize grains are frequently contaminated by the mycotoxigenic fungi Aspergillus flavus, known to produce aflatoxin B1, one of the most potent carcinogenic and citotoxic compounds for man and some animal species (Foster et al., 1983; WHO/IARC, 1993; Lewis et al., 1999; Palanee et al., 2000).

The majority of fungal control methods for maize are based on the application of wide-spectrum chemical agents. These present, however, multiple deficiencies such as animal and human toxicity, environmental contamination, induced resistance and low effectiveness on seeds colonized by fungi in which mycotoxins have already been produced (OMS, 1992; PAHO/WHO, 1993; Calistru et al., 1997b; Gisi et al., 2002). This situation has necessitated the implementation of alternative control strategies at the levels of production and storage, especially when dealing with fungi with low humidity requirements. Among these new strategies is biological control, based on the antagonic action of one microorganism over another by attacking and contaminating cultures. So far, the most promising group for mycopathogen control belongs to the Genus Trichoderma, due to a variety of factors (Elad et al., 1983; Lorito et al., 1993; Calistru et al., 1997a, b; Yedidia et al., 1999, 2000; Howell et al., 2000; Cooney et al., 2001; Okigbo and Ikediugwu, 2001; Dal Bello et al., 2002; Humphris et al., 2002) including a reduction in micotoxin production when the mycopathogen grows together with Trichoderma species (Cooney et al., 2001). Among the control mechanisms associated with Trichoderma spp., there is an antagonic action with volatile compounds as shown by the fungistatic effects on wood-rotting fungi (Bruce et al., 1996, 2000; Calistru et al., 1997a; Wheatley et al., 1997; Humphris et al., 2002; Wheatley, 2002) and their efficacy in controlling apple ring rot (Kexiang et al., 2002). However, available information on the possible action of the volatile compounds of Trichoderma spp. on mycotoxigenic fungus is scarce and limited to a few studies (Calistru et al., 1997a).

Given that large-scale storage systems for maize operate as packed bed columns, and can be force-ventilated, the application of volatile compounds inhibiting mycotoxigenic fungi is seen as a possible control strategy. Thus, the central objective of this paper was to determine if the action of volatile compounds produced by Trichoderma harzianum is effective in preventing the proliferation of Aspergillus flavus and the subsequent production of aflatoxin B1 in stored maize.

Trichoderma harzianum (LPFIBE-5) was originally obtained from a corn plantation in Portuguesa State, Venezuela, and the micotoxigenic Aspergillus flavus (LPFIBE-20) was isolated from local maize seeds. Both fungal isolates were supplied by the Centro Venezolano de Colecciones de Microorganismos (CVCM), Venezuela. Fungi were maintained in periodical subcultures in slants of Malt Extract Agar (MEA; Hi Media Laboratories, Mumbai, India) supplemented with 2.35g·l^-1 glycerol.

To reduce the contaminant microbial load, commercial yellow corn was placed in covered aluminum containers (14.7´9.7´6cm) and irradiated at 3Kgrey in the Gamma Radiation Sterilization Plant at the Instituto Venezolano de Investigaciones Científicas (IVIC). Irradiated maize was kept at 4°C until used.

Inocula of both A. flavus and T. harzianum were prepared from sporulated cultures developed at ambient temperature during 15 days on MEA slants. Spores were collected by adding sterile 0.1% Tween 80 in distilled water and shaking by hand. Spore counts were made using a Neubauer haemocytometer.

Volatile compounds (VC) production of T. harzianum was carried out at two different scales, employing roux bottles (small scale) and aluminum trays (large scale), with work surfaces of 216 and 2591cm², respectively. Roux bottles were loaded with 330ml of MEA medium or, alternatively, 154g of a solid medium formulated with defatted corn germ - rice husk (DCGRH), a modification of the BH medium described by Dorta et. al. (1990). In the latter medium, the rice bran was substituted by defatted corn germ, a byproduct supplied by Promesa, Aragua State, Venezuela. Both media (MEA and DCGRH) were sterilized at 120°C during 20min.

In the large scale experiment, DCGRH medium (400g) was poured uniformly over the entire surface of the sterilized aluminum trays. In this case, the medium was prepared in Erlenmeyer flasks, sterilized and inoculated with T. harzianum spores before being transferred to the trays. Each tray was covered with a polyethylene bag (previously sterilized by gamma radiation at 2.5Mrad). The whole operation was performed in a laminar flow cabinet.

MEA cultures were started by adding 10ml of spore suspension to a concentration of 10⁶ spores/ml, while DCGRH cultures were inoculated with 10⁶ spores/g of solids. Incubations were done at 25±1°C for 15 days in the presence of continuous artificial light.

Antagonism of T. harzianum vs A. flavus

Roux bottles containing T. harzianum grown on MEA and DCGRH media for two weeks, and their controls (media without inocula), were connected to glass columns (2.5cm diameter and 30cm long) packed with 62g of maize inoculated with Aspergillus flavus at 10³ spores/g. The roux bottle-column system remained connected allowing gaseous exchange during 15 days (Figure 1a), after which columns were dismantled to determine the accumulated fungal biomass and the aflatoxin B1 levels present.

Aluminum trays containing 15 day developed cultures of T. harzianum, and their respective controls, were connected through a tubing system to 4l plastic containers packed with 2kg of maize inoculated with Aspergillus flavus at the range of 10³ spores/g (Figure 1b). Volatile compounds of T. harzianum were transferred to the plastic containers using aquarium peristaltic pumps which operated 1min daily, to guarantee a complete change of the void volume of the corn packed beds.

The total A. flavus biomass accumulated in each column was determined indirectly using measurements of N-acetylglucosamine, following the method of Ride and Drisdale (1972). Columns containing contaminated and treated maize were sterilized at 120ºC during 20min and oven dried at 60ºC to constant weight. Immediately afterwards, the total content of each column was powdered with an intermediate mill (Thomas Wiley, Swedesboro, NJ, USA) until a particle size of 60 mesh was obtained, and dried again at 60ºC to constant weight. Subsamples of known weights 50-80mg were taken from each treatment, and processed following the protocol described by Ride and Drisdale (1972).

For the determination of G_m A. flavus was cultivated in MEA plates. Mycelia were harvested by centrifugation, washed twice in distilled water, and oven dried at 60°C until constant weight. A known weight sample was processed using the Ride and Drisdale (1972) protocol for N-acetylglucosamine determination.

Aflatoxin determination was performed using Aflatest immunoaffinity columns (P³10ng/g; Vicam, Watertown, MA, USA). For each treatment 50g samples were taken and processed according to the procedure described by the manufacturer.

Variance analyses and means comparison tests were applied to the results with the statistical package 2001 Statistica for Windows (StatSoft, Inc., Tulsa, AZ, USA).

Effect of VCs from T. harzianum on A. flavus proliferation

The effectiveness of malt extract as a culture medium for some species of Trichoderma, in antagonic processes mediated by volatile compounds, has been established in earlier studies (Wheatley et al., 1997). Although in those studies reference is made to volatile compound action on a specific group of wood-rotting fungi (Neolentinus lepideus, Postia placenta, Gloeophyllum trabeum and Coriolus versicolor, among others), such action could be extended to other fungi of phytosanitary importance, such as A. flavus. In this study, the effect of VCs produced by T. harzianum in MEA was tested on the proliferation of A. flavus inoculated on previously irradiated maize grains. The corresponding analysis of variance revealed significant differences (p<0.05) between the values of A. flavus biomass accumulated on irradiated maize and those of non-exposed samples (Table I).

Also, direct observation of maize packed in columns exposed to VCs from T. harzianum cultured in DCGRH, suggests a differential effect associated with their concentration. An axial gradient in the proliferation of A. flavus was observed, being lower in the immediate surroundings of the point of entry of the volatiles and higher at the distal end.

An experimental system for increasing the VC concentration using connected trays in parallel was designed on the basis of the area scale destined for T. harzianum culture (Figure 1b).

With this system, the cultured surface of T. harzianum increased 24´ and the weight of treated maize increased 60´. The VCs produced by T. harzianum were transferred to the corn by pumping fermentation gases in sufficient quantities to produce a complete daily turnover of the void volume of the packed corn bed. Under these conditions, A. flavus proliferation was reduced by 90% after 15 days (Table I). Clearly, the antagonic activity of the VCs reduces A. flavus proliferation in stored maize, and the production of VCs may also be scaled up for implementation at greater volumes. However, direct application of fermentation gases of T. harzianum to stored maize causes an increase in the moisture content, characteristic of solid-state fermentation systems. Thus, their use must be accompanied by the use of dehydration devices to ensure the low-humidity conditions necessary for maize storage systems.

Effect of VCs of T. harzianum on aflatoxin B1 levels

VCs produced by T. harzianum cultured 15 days in MEA significantly reduced (p<0.05) levels of specific aflatoxin (ng aflatoxin/mg of A. flavus biomass) in inoculated maize (10³ spores/ml) as compared to non-exposed maize (Table I).

Although the inhibition of aflatoxin B1 production by volatiles generated from cotton leaves has been reported (Zeringue and McCormick, 1990; Zeringue et al., 1996; Greene et al., 1999), this study represents the first report of the inhibiting effect produced by the VCs of T. harzianum on A. flavus proliferation in stored maize.

Trichoderma spp. volatiles identified as inhibitors of wood rotting fungi (Bruce et al., 1996; Wheatley et al., 1997) and those identified in cotton leaves (Zeringue et al., 1990; Zeringue et al., 1996) have shown similarities with hexanal, octanal and 1-heptanal compounds. These findings suggest that volatile aldehydes produced by Trichoderma spp. should be responsible for the inhibiting effect on the micotoxigenic fungus A. flavus.

Further work is required to characterize the chemical nature of the volatile compounds reported in this study. It will help establish relationships with volatile agents previously identified by other authors, and raise the possibility for synthesizing such compounds and investigating its use as control agents on A. flavus growing in stored grains.

1. Abdullah N, Nawawi A, Othman I (1998) Survey of fungal counts and natural occurrence of aflatoxins in Malaysian starch-based foods. Mycopathologia 143: 53-58.        [ Links ]

2. Bruce A, Kundzewicz A, Wheatley R (1996) Influence of culture age on the volatile organic compounds produced by Trichoderma aureoviride and associated inhibitory effects on selected wood decay fungi. Mat. Organismen. 30: 79-94.        [ Links ]

3. Bruce A, Wheatley R, Humphris S, Hackett C, Florence M (2000) Production of volatile organic compounds by Trichoderma in media containing different amino acids and their effects on selected wood decay fungi. Holzforschung 54: 481-486.        [ Links ]

4. Calistru C, McLean M, Berjak P (1997a) In vitro studies on the potential for biological control of Aspergillus flavus and Fusarium moniliforme by Trichoderma species. Mycopathologia 137: 115-124.        [ Links ]

5. Calistru C, McLean M, Berjak P (1997b) In vitro studies on the potential for biological control of Aspergillus flavus and Fusarium moniliforme by Trichoderma species. Mycopathologia 139: 115-121.        [ Links ]

6. Cooney JM, Lauren DR, di Menna ME (2001) Impact of competitive fungi on Trichothecene production by Fusarium graminareum. J. Agric. Food Chem. 49: 522-526.        [ Links ]

7. Dal Bello GM, Mónaco CI, Simón MR (2002) Biological control of seedling blight of wheat caused by Fusarium graminareum with beneficial rhizosphere microorganisms. World J. Microbiol. Biotechnol. 18: 627-636.        [ Links ]

8. Dalcero A, Magnoli C, Luna M, Ancasi G, Reynoso MM, Chiacchiera S, Miazzo R, Palacio G (1998) Mycroflora and naturally occurring mycotoxins in poultry feeds in Argentina. Mycopathologia 141: 137-43.        [ Links ]

9. Dorta B, Bosch A, Arcas JA, Ertola RJ (1990) High level of sporulation of Metarhizium anisopliae in a medium containing byproducts. Appl. Microbiol. Biotechnol. 33: 712-715.        [ Links ]

10. Dutta TK, Das P (2001) Isolation of aflatoxigenic strain of Aspergillus and detection of aflatoxin B1 from feeds in India. Mycopathologia 151: 29-33.        [ Links ]

11. Elad Y, Chet I, Boyle P, Henis Y (1983) Parasitism of Trichoderma spp on Rhizoctonia solani and Sclerotium rolfsii - Scanning Electron Microscopy and Fluorescence Microscopy. Phytopathology 73: 85-88.        [ Links ]

12. Foster PL, Eisenstadt E, Miller JH (1983) Base substitution mutations induced by metabolically activated aflatoxin B1. Proc Natl. Acad. Sci. 80: 2695-2698.        [ Links ]

13. Gisi U, Sierotzki H, Cook A, McCaffery A (2002) Mechanisms influencing the evolution of resistance to Qo inhibitor fungicides. Pest Manag. Sci.. 58: 859-867.        [ Links ]

14. Greene-McDowelle DM, Ingber B, Wright MS, Zeringue HJ, Bhatnagar D, Cleaveland T (1999) The effects of selected cotton-leaf volatiles on growth, development and aflatoxin production of Aspergillus parasiticus. Toxicon 37: 883-893.        [ Links ]

15. Howell CR, Hanson LE, Stipanovic RD, Puckhaber LS (2000) Induction of Terpenoid synthesis in cotton roots and control of Rhizoctonia solani by seed treatment with Trichoderma virens. Phytopathology 90: 248-252.        [ Links ]

16. Humphris SN, Bruce A, Buultjens E, Wheatley RE (2002) The effects of volatile microbial secondary metabolites on protein synthesis in Serpula lacrymans. FEMS Microbiol. Lett.. 210: 215-219.        [ Links ]

17. Kexiang G, Xiaoguang L, Yonghong L, Tianbo Z, Shuliang W (2002) Potencial of Trichoderma harzianum and T. atroviride to control Botryosphaeria berengeriana f. sp. piricola, the cause of apple ring rot. J. Phytopathol. 150: 271-276.        [ Links ]

18. Lewis CW, Smith JE, Anderson JG, Freshney RI (1999) Increased cytotoxicity of food-borne mycotoxins toward human cell lines in vitro via enhanced cytochrome p450 expression using the MTT bioassay. Mycopathologia 148: 97-102.        [ Links ]

19. Lorito M, Harman GE, Hayes CK, Broadway RM, Tronsmo A, Woo SL, Di Pietro A (1993) Chitinolytic enzymes produced by Trichoderma harzianum: Antifungal Activity of purified endochitinase and chitobiosidase. Phytopathology 83: 302-307.        [ Links ]

20. Martínez A, Trucksess M, Park D (1987) Analysis of Venezuelan corn for aflatoxin and Aspergillus flavus or Aspergillus parasiticus contamination. In Llewellyn GC, O’Rear CE (Eds.) Biodeterioration Research 1. Plenum. New York, USA. pp:111-118.        [ Links ]

21. Mazzani C, Borges O, Luzón O, Barrientos V, Quijada P (1999) Incidencia de Aspergillus flavus, Fusarium moniliforme, aflatoxinas y fumonisinas en ensayos de híbridos de maíz en Venezuela. Fitopatol. Venez. 12: 9-13.        [ Links ]

22. Mazzani C, Borges O, Luzón O, Barrientos V, Quijada P (2000) Fusarium moniliforme, fumonisinas y Aspergillus flavus en granos de híbridos de maíz en el estado Guárico, Venezuela. Rev. Fac. Agron (LUZ) 17: 185-195.        [ Links ]

23. Mazzani C, Borges O, Luzón O, Barrientos V, Quijada P (2001) Occurrence of Fusarium moniliforme and fumonisins in kernels of Maize hybrids in Venezuela. Braz. J. Microbiol. 32: 345-349.        [ Links ]

24. Mora M, Lacey J (1997) Handling and aflatoxin contamination of white maize in Costa Rica. Mycopathologia 138: 77-89.        [ Links ]

25. Okigbo RN, Ikediugwu FEO (2001) Biological control of tuber surface mycroflora of yams (Dioscorea rotundata). Trop Sci. 41: 85-89.        [ Links ]

26. OMS (1992) Consecuencias sanitarias del empleo de plaguicidas en la agricultura. Organización Mundial de la Salud. Geneve, Switzerland. 128 pp.        [ Links ]

27. PAHO/WHO (1993) Pesticides and health in the Americas. Division of Health and Environment. Environmental, Series Nº12. Washington, DC, USA. 109 pp.        [ Links ]

28. Palanee T, Dutton MF, Chuturgoon A (2000) Cytotoxicity of aflatoxin B1 and its chemically synthesized epoxide derivative on the A459 human epithelioid lung cell line. Mycopathologia 151: 155-159.        [ Links ]

29. Raybaudi R, Martínez A (2000) Incidencia e identificación de la micobiota en granos de maíz (Zea mays L) y comparación de medios de cultivo para la determinación de mohos toxigénicos. Fitopatol. Venez. 13: 15-18.        [ Links ]

30. Ride JP, Drisdale RB (1972) A rapid method for the chemical estimation of filamentous fungi in plant tissue. Physiol. Plant Pathol. 2: 7-15.        [ Links ]

31. Shöllenberger M, Suchy S, Terry H, Drochner W, Martin H (1999) A survey of Fusarium toxins in cereal-based foods marketed in an area of southwest Germany. Mycopathologia 147: 49-57.        [ Links ]

32. Wheatley RE (2002) The consequences of volatile organic compound mediated bacterial and fungal interactions. Ant. Leeuw. 81: 357-364.        [ Links ]

33. Wheatley R, Hackett C, Bruce A, Kundzewicz A (1997) Effect of substrate composition on production of volatile organic compounds from Trichoderma spp: Inhibitory to wood decay fungi. Int. Biodeter. Biodegrad. 39: 199-205.        [ Links ]

34. WHO/IARC (1993) Monographs on the evaluation of carcinogenic risk to humans. Vol 56. International Agency for Research on Cancer (IARC). Lyon, France. 245 pp.        [ Links ]

35. Yedidia I, Benhamou N, Chet I (1999) Induction of defense responses in cucumber plants (Cucumis sativus L.) by the biocontrol agent Trichoderma harzianum. Appl. Env. Microbiol. 65: 1061-1070.        [ Links ]

36. Yedidia I, Benhamou N, Kapulnik Y, Chet I (2000) Induction and accumulation of PR protein activity during early stages of root colonization by the mycoparasite Trichoderma harzianum strain T-203. Plant Physiol. Biochem. 38: 863-873.        [ Links ]

37. Zeringue HJ, McCormick SP (1990) Aflatoxin production in cultures of Aspergillus flavus incubated in atmospheres containing selected cotton leaf-derived volatiles. Toxicon 28: 445-448.        [ Links ]

38. Zeringue HJ, Brown RL, Neucere JN, Cleveland TE (1996) Relationships between C₆-C₁₂ alkanal and alkenal volatile contents and resistance of maize genotypes to Aspergillus flavus and aflatoxins production. J. Agric. Food Chem. 44: 403-407.        [ Links ] ]]> 1 1998 143 53-58 2 1996 30 79-94 3 2000 54 481-486 4 1997 a 137 115-124 5 1997 b 139 115-121 6 2001 49 522-526 7 2002 18 627-636 8 1998 141 137-43 9 1990 33 712-715 10 2001 151 29-33 11 1983 73 85-88 12 1983 80 2695-2698 13 2002 58 859-867 14 1999 37 883-893 15 2000 90 248-252 16 2002 210 215-219 17 2002 150 271-276 18 1999 148 97-102 19 1993 83 302-307 20 1987 111-118 21 1999 12 9-13 22 2000 17 185-195 23 2001 32 345-349 24 1997 138 77-89 25 2001 41 85-89 26 OMS 1992 128 27 PAHO/WHO 1993 109 28 2000 151 155-159 29 2000 13 15-18 30 1972 2 7-15 31 1999 147 49-57 32 2002 81 357-364 33 1997 39 199-205 34 WHO/IARC 1993 245 35 1999 65 1061-1070 36 2000 38 863-873 37 1990 28 445-448 38 1996 44 403-407