ArticlePDF Available

Effect of equilibrium relative humidity on ochratoxin A production by Aspergillus carbonarius in raw coffee beans

Braz. J. Food Technol., v.7, n.2, p.111-114, juL./dez., 2004 111 Recebido / Received: 11/09/2003. Aprovado / Approved: 12/04/2004.
Coffee; Ochratoxin A; Aspergillus carbonarius;
Equilibrium relative humidity / Café; Ocratoxina A;
Aspergillus carbonarius; Umidade relativa de equilíbrio.
Ochratoxin A (OTA) production by Aspergillus carbonarius was studied in raw
irradiated coffee beans under four conditions of equilibrium relative humidity (ERH): 80%,
87%, 95% and 100% at 25°C. Ochratoxin A production was quantified at 46 and 67 days.
There was no OTA production at 80% ERH, but an increase in ERH was observed of 87%,
91% and 94%, producing 0.88, 2.59 and 9.62µg/kg, respectively, after 67 days.
A produção de ocratoxina A (AO) por Aspergillus carbonarius foi estudada em café
irradiado beneficiado em quatro condições de umidade relativa de equilíbrio (URE): 80%,
87%, 95% e 100% a 25°C. Após 46 e 67 dias, procedeu-se à quantificação de ocratoxina A.
Não houve produção de AO a 80% de URE , porém observou-se um aumento de URE de
87%, 91% e 94%, correspondente a 0,88, 2,59 e 9,62µg/kg , após 67 dias de incubação.
Effect of Equilibrium Relative Humidity on
Ochratoxin A Production by Aspergillus
carbonarius in Raw Coffee Beans
Efeito da Umidade Relativa de Equilíbrio na
Produção de Ocratoxina A por Aspergillus
carbonarius em Café Cru Beneficiado
Hilary Castle MENEZES
Faculdade de Engenharia de Alimentos (FEA) - UNICAMP
Campinas-SP, Brazil
Marta Hiromi TANIWAKI
Instituto de Tecnologia de Alimentos (ITAL)
Av. Brasil, 2880, P. O. Box 139
CEP: 13070-178 Campinas-SP, Brazil
Beatriz Thie IAMANAKA
Instituto de Tecnologia de Alimentos (ITAL)
Av. Brasil, 2880, P. O. Box 139
CEP: 13070-178 Campinas-SP, Brazil
Braz. J. Food Technol., v.7, n.2, p.111-114, juL./dez., 2004 112
et al.
Effect of Equilibrium Relative Humidity
on Ochratoxin A Production by
Aspergillus carbonarius in Raw Coffee
Ochratoxin A (OTA) is produced in nature by three main
species, Aspergillus ochraceus, Aspergillus carbonarius and
Penicillium verrucosum. Recently, OTA producing A. ochraceus,
A. niger and A. carbonarius have been reported in coffee by
several authors (NAKAJIMA et al., 1997; TÉREN et al., 1997;
JOOSTEN et al., 2001; URBANO et al., 2001; TANIWAKI et al.
2003). Ochratoxin A is a mycotoxin which exhibits nephrotoxic,
immunosuppresive, teratogenic and carcinogenic properties
(CREPPY, 2002; ABARCA, 2001; VARGA et al., 2000). In the last
few years, coffee importing companies and researchers have
taken a greater interest in the micotoxicological quality of
raw coffee and it has been the object of regulation by coffee
importing countries. Based on the carcinogenicity of OTA,
some countries have introduced national limits on green
coffee: Italy, 8µg/kg; Greece, 20µg/kg; France, 5µg/kg and
Finland, 10µg/kg (ROMANI et al. 2000; CREPPY, 2002;
Until recently, A. carbonarius had not been considered
relevant, due to it being confused with A. niger. However, it
appears to be found in grapes and their derivatives (HEENAN
et al., 1998), coffee (JOOSTEN et al., 2001; TANIWAKI et al.,
2003) and cocoa (MATISSEK; RATERS, 2000).
The study of OTA production by A. carbonarius has
been mostly conducted in culture medium (HEENAN et al.,
1998) and in coffee cherries (JOOSTEN et al., 2001), but not
in raw coffee beans.
The objective of this research was to study OTA
production by A. carbonarius in raw coffee beans submitted
to different conditions of equilibrium relative humidity (ERH)
of 80%, 87%, 95% and 100% at a temperature of 25°C, in
order to acquire data that might be useful in preventing
ochratoxin A production in coffee beans.
2.1 Coffee
Raw arabica coffee from the experimental farm at the
Campinas Agronomy Institute (IAC, Campinas, SP, Brazil) was
dried and husked. The beans were sterilized by Cobalt 60
irradiation at 10kGy at the Agricultural Centre for Nuclear
Energy (CENA/USP, Piracicaba, SP, Brazil).
2.2 Selection and preparation of the Aspergillus
carbonarius spore suspensions
Three ochratoxin A producing isolates of A. carbonarius
were isolated from coffee grown in Brazil. The strains used in
this test, ITAL 168, 169 and 170, were isolated by TANIWAKI
et al. (2003). They were inoculated into malt extract agar (MEA)
and incubated at 25°C for 5 days. The inoculum was prepared
by mixing the cultures followed by transference into a test
tube containing 40 ml phosphate buffer plus 0.1% tween 80
and glass beads. The suspension was agitated for 1 minute.
A spore concentration of 107 CFU/mL was obtained by the
dilution technique.
2.3 Inoculation of A. carbonarius spores into the
According to the methodology proposed by
PALACIOS-CABRERA et al. (2001), a 1mL aliquot of spore
suspension (107 CFU/mL) was inoculated into 10g of sterilized
soil with a particle size of 300µm. 4g of this mixture were
then added to 100g of coffee and mixed for 15min (Wagner
mixer). This procedure was repeated for a total weight of
1.5kg of contaminated coffee.
2.4 Determination of water activity in the raw coffee
The water activity (aw) was determined using the
Aqualab 3TE (Decagon-USA), equipment.
2.5 Calibration curve for moisture content
The initial moisture content of the raw coffee was
determined according to the vacuum oven method at 70ºC
for 24h. The Norm International ISO method 1447 (1978)
was also used, drying the coffee at 130ºC for 6h, keeping it in
desiccators at room temperature for 15h and then drying it
again at 130ºC for 4h. The range of coffee moisture content
used to make the calibration curve was from 9 to 20%.
2.6 Preparation of saturated salt solutions for the
determination of Equilibrium Relative Humidity
Saturated solutions of ammonium sulphate (NH4)2SO4),
sodium potassium tartrate (NaKC4H4O6) and lead nitrate
(Pb(NO3)2) were used to obtain ERH values of 80, 87 and
95%, respectively. The solutions were left in the desiccators
for 7 days to reach equilibrium. De-ionised water was placed
in the desiccators to obtain 100% of ERH.
2.7 Preparation of the isotherm trials at 25ºC with
4 different values of equilibrium relative
humidity (80, 87, 95% and 100%)
Plastic pots containing 25g of dried coffee beans were
inoculated with 1g of soil containing a spore suspension as
mentioned in item 2.3. Five pots were placed in each desiccator
(4 contaminated samples and one non-inoculated control),
and incubated at 25ºC. The ERH values tested were 80, 87,
95% and 100%, obtained using saturated solutions. The non-
inoculated pot was weighed at weekly intervals to determine
the increase in moisture and the time needed for the coffee
beans to reach equilibrium with the salt solution, based on
the change in bean weight. The beans reached the ERH of
Braz. J. Food Technol., v.7, n.2, p.111-114, juL./dez., 2004 113
et al.
Effect of Equilibrium Relative Humidity
on Ochratoxin A Production by
Aspergillus carbonarius in Raw Coffee
the salt solutions and OTA production was quantified after
46 days (equilibrium). All trials were carried out twice.
2.8 Extraction and quantification of ochratoxin A
in the raw coffee beans
Ochratoxin A was extracted based on the method of
PITTET et al. (1996). The coffee beans were extracted with a
solution of methanol: 3% sodium bicarbonate (50:50). The
extracts were filtered and diluted with phosphate buffered
saline and applied to an immunoaffinity column (Vicam,
Watertown, MA), immobilized with OTA specific monoclonal
antibody. After washing, the OTA was eluted with HPLC grade
methanol and quantified by reverse-phase HPLC using a
fluorescence detector. The mobile phase used was acetonitrile/
4mM sodium acetate with a 0.5% acetic acid solution (42:58).
The flow rate was 1ml/min. The equipment used was a
Shimadzu LC-10VP system (Shimadzu Corporation, Japan),
with a fluorescence detector set at 330nm excitation and
470nm emission. The HPLC was fitted with an ODS Hypersil
(5µm, 25mm X 4.6mm) pre-column and SupelcosilTM LC-18
(5µm, 250mm X 4.6mm) column (Supelco, USA). The detection
limit of this method was 0.1µg/kg OTA.
Figure 1 presents the calibration curve to show the
correlation between the moisture contents determined by
the vacuum oven and ISO methods (the former standard
method for routine analysis of raw coffee beans).
The calibration curve was important to determine the
value of the initial moisture content according to the ISO
method, adopted as the reference to prepare a raw coffee
isotherm (GOUGH, 1975; BUCHELI et al., 1998). The ISO (1978)
method is specific for raw coffee moisture content. The results
obtained from the moisture content isotherm are presented
in Figure 2.
Coffee moisture content (dry weight basis)
y = 1.0492x + 4.5083
R2 = 0.9974
0 5 10 15 20
Vacuum oven method
ISO Method
FIGURE 1. Calibration curve of raw coffee moisture content
determined by two methodologies (ISO and vacuum oven).
y = 214.36x2 - 268.92x + 93.891
R2 = 0.9981
y = 218.65x2 - 272.57x + 99.425
R2 = 0.9974
0,65 0,7 0,75 0,8 0,85 0,9 0,95
Water activity (25ºC)
Moisture content (dry weight basis)
Vacuum oven
ISO method
FIGURE 2. Adsorption isotherm of raw coffee at 25°C.
TABLE 1. Production of ochratoxin A (µg/kg) by Aspergillus
carbonarius in raw coffee beans at 25°C, with ERH values of
80, 87, 95 and 100%.
Ochratoxin A (µg/kg)* Time
(days) 80% 87% 95% 100%
46 ND** 0.31 2.57 7.59
67 ND 0.88 2.59 9.62
* Average of two repetitions
** Not detected: limit of detection 0.1µg/kg
Both adsorption isotherms were constructed
based on the moisture gain at different water activities.
The initial moisture content was determined by two
methodologies, ISO and vacuum oven at 70ºC for 24h.
The criteria established that the coffee had reached
equilibrium when the weight of the sample showed no
further changes in mass gain to the third decimal place.
The tests at 80% and 87% satisfied the criteria, with
corresponding water activity values of 0.79 and 0.85
respectively. However, at 95% and 100% ERH, the coffee
did not reach equilibrium because the samples were
visibly mouldy. The water activities for 95% and 100%
ERH were 0.91 and 0.94, respectively, at the time when
the samples were analysed. PALACIOS-CABRERA;
TANIWAKI (2003) compared the method of the Brazilian
Agriculture Ministry to that of ISO, to determine the
moisture content of green coffee. In this study, the
difference between these methodologies was 1%.
Table 1 presents the production of OTA by A.
carbonarius in coffee after 46 and 67 days. There was a
gradual increase in ochratoxin A production with increase
in water activities after 46 and 67 days. OTA production by
A. carbonarius in raw coffee was very low in all the
experiments, as well as at a water activity of 0.94.
Braz. J. Food Technol., v.7, n.2, p.111-114, juL./dez., 2004 114
et al.
Effect of Equilibrium Relative Humidity
on Ochratoxin A Production by
Aspergillus carbonarius in Raw Coffee
JOOSTEN et al., (2001) did not detect ochratoxin A
production by A. carbonarius, at water activities of 0.85 and
0.91 in coffee cherries, but at 0.94 the production was 230µg/
kg after 14 days. In the present study, A. carbonarius produced
a very small amount of OTA at 0.85 and 0.91, which
corresponded to the ERH of 87% and 95%; at 0.94 (100% of
ERH), the maximum production found was 9.6µg/kg after 67
days. These differences may be due to the difference in the
substrate used JOOSTEN et al. (2001) used coffee cherries
with the pulp, which is rich in nutrients and free sugars that
have more carbon source availability for fungi to grow
(ROGERS et al., 1999; BUCHELI; TANIWAKI, 2002). In contrast,
in the present study, the coffee beans were husked, dried and
The production of ochratoxin A in coffee is complex
and many factors may be involved such as substrate, strain
variability, moisture content, temperature and time of contact
between fungi and substrate. However, OTA production in
coffee can be avoided using adequate agriculture, harvesting,
drying, storage and transport practice. The present work
shows that in coffee stored at water activity below 0.80, OTA
production by A. carbonarius can be severely inhibited.
importance of Ochratoxin A- producing Aspergillus spp. Journal
of food Protection, v.6, p.903-906, 2001.
BUCHELI, P.; MEYER, I.; VUATAZ, A.; VIANI, R. Industrial storage of
green robusta coffee under tropical conditions and its impact on
raw material quality and ochratoxin A content. J. Agric. Food.
Chem., v.46, p.4507-4511, 1998.
BUCHELI, P.; TANIWAKI, M.H. Review: Research on the origin, and on
the impact of post-harvest handling and manufacturing on the
presence of ochratoxin A in coffee. Food Additives and
Contaminants, v.19, p.655-665, 2002.
CREPPY, E. E. Update of survey, regulation and toxic effects of
mycotoxin in Europe. Toxicology Letters, v.127, p.19-28, 2002.
GOUGH, M.C. A simple technique for the determination of humidity
equilibria in particulate foods. Journal of. Stored. Products
Research, v.11, p.161-166, 1975.
HEENAN, C.N.; SHAW, K.J.; PITT, J.I. Ochratoxin A production by
Aspergillus carbonarius and A.niger isolates and detection using
coconut cream agar. Journal of Food Mycology, v.1, n.2,
p.67-72, 1998.
BUCHELI, P. Production of ochratoxin A by Aspergillus carbonarius
on coffee cherries. International Journal of Food
Microbiology, v.65, p.39-44, 2001.
MATISSEK, R.; RATERS, M. Ochratoxin A in cocoa and human health
aspects. 13th International Cocoa Research Conference, Kota
Kinabalu, Malaysia, 9-14 October, 2000.
aflatoxin B1 and ochratoxin A in commercial green coffee beans by
high-performance liquid chromatography linked with
immunoaffinity chromatography. Food and Agricultural
Immunology, v.9, p.77-83, 1997.
NORM INTERNATIONAL ISO 1447. Green coffee- Determination of
moisture content (routhine method), 1978.
teor de umidade do café cru beneficiado: Comparação entre
diferentes metodologias. Brazilian Journal of Storage, v.6,
p.25-29, 2003.
Optimisation of the inoculation of Aspergillus ochraceus in coffee
for isothermal studies simulating storage and Marine Transport
of raw coffee. 19 th ASIC Coffee Conference, Trieste, Italy,
p.14-18, May, 2001.
chromatographic determination of ochratoxin A in pure and
adulterated soluble coffee using na immunoaffinity column cleanup
procedure. Journal of Agricultural and Food Chemistry,
v.44, p.3564-3569, 1996.
the content of sugars, sugar alcohols, myo-inositol, carboxylic
acids and inorganic anions in developing grains from different
varieties of Robusta (Coffee canephora) and Arabica (C. arabica)
coffees. Plant Science, v.149, p.115-123, 1999.
DALLA ROSA, M. Screening on the Occurrence of Ochratoxin
A in green coffee beans of different origins and types, Journal
of Agricultural and Food Chemistry, v.48, p.3616-3619,
source of ochratoxin A in Brazilian coffee and its formation in
relation to processing methods. Intern. J. Food Microbiol.,
v.82, p.173-179, 2003.
TÉREN, J.; PALÁGYI, A.; VARGA, J. Isolation of ochratoxin producing
aspergilli from green coffee beans of different origin. Cereal
Research Communications, v.25, p.303-304, 1997.
Occurrence of ochratoxin A producing fungi in raw Brazilian coffee.
J. Food Prot., v.64, p.1226-1230, 2001.
VARGA, J.; RIGÓ, K.; TÉREN, J. Degradation of ochratoxin A by
Aspergillus species. International Journal of Food
Microbiology, v.59, p.1-7, 2000.
The aim of this study was to determine the effects of water activity (a(w)) (0.92-0.98), temperature (5-45 °C) and incubation time (5-60 days) on growth and ochratoxin A (OTA) production by Aspergillus niger and Aspergillus carbonarius on maize kernels using a simple method. Colony diameters of both strains at 0.92 a(w) were significantly lower than those at 0.96 and 0.98 a(w) levels. The optimum growth temperature range for A. niger was 25-40 °C and for A. carbonarius 20-35 °C. A. niger produced OTA from 15 to 40 °C, and the highest OTA level was recorded at 15 °C. The concentration of OTA produced at 0.92 a(w) was significantly lower than those at 0.96 and 0.98 a(w). A. carbonarius produced OTA from 15 to 35 °C and the maximum concentration was achieved at 15 °C, although not differing statistically from the concentration detected at 20 °C. At 0.98 a(w) the OTA concentration was significantly higher than at 0.96 and 0.92 a(w). Our results show that maize supports both growth and OTA production by A. niger and A. carbonarius. The studied strains were able to produce OTA in maize kernels from the fifth day of incubation over a wide range of temperatures and water availabilities. Although the limit of quantification of our method was higher than that required for the analysis of OTA in food commodities, it has proved to be a useful and rapid way to detect OTA production by fungi inoculated onto natural substrates, in a similar way as for pure culture. Both species could be a source of OTA in this cereal in temperate and tropical zones of the world.
Full-text available
Green Robusta coffee was stored in silos for 8 months under industrial conditions in Thailand, and subjected to air-conditioning, aeration, and nonaeration, and compared to bag storage under ambient conditions. Air-conditioning clearly reduced the relative humidity (RH) of the silo atmosphere and the moisture content (MC) and the water activity (aw) of green coffee. Overall storage behavior was better for coffee in the aerated silo (RH = 68%; MC = 13.0%; aw = 0.69) than for coffee stored in bags (RH = 81%; MC = 13.5%; aw = 0.72). Aeration provided an efficient means to reduce MC and aw during the rainy period. Glucose content was linked positively with the increase of a woody/rubbery note in coffee cup quality and increased by 50% between months 3 and 6 when green coffee MC and aw increased the most. Glucose is a potential green coffee quality marker. Under the tested storage conditions, neither the growth and presence of ochratoxin A (OTA)-producing fungi nor consistent OTA production was found. OTA contamination appeared to have occurred before storage. Keywords: Coffee; storage; silo; glucose; mold; ochratoxin A
Full-text available
Changes in concentration of mono- and oligosaccharides, sugar alcohols, myo-inositol, carboxylic acids and inorganic anions in coffee grains were analysed during grain development in three cultivars of Coffea arabica L (Arabica) and two of C. canephora L var. Robusta (Robusta) by high performance anion exchange chromatography coupled to pulsed electrochemical detection (HPAE-PED). The majority of the components analysed either decreased in concentration during the first half of the development period or accumulated steadily during the latter half of the period. The profiles are taken to indicate relationships between the perisperm, the principal tissue in the young grain and the endosperm during maturation. While most of the free sugar in the mature grain is accounted for by sucrose, fructose and glucose are both at higher concentrations in the perisperm. Considerable amounts of myo-inositol (3–4% dry weight (DW)) are found in young grains, while only the phosphorylated form phytic acid occurs in mature grains (0.3–0.6% DW). Quinic acid, which is present in very low amounts in mature endosperm, represents between 6 and 16% DW in young grains, this possibly being the major precursor pool for the high amounts of chlorogenic acids (5–10% DW) which are a characteristic of mature coffee grains. Of the other organic acids analysed, citric and malic acids are dominant in the mature grain, with higher concentrations in Arabica than Robusta. The results are discussed with respect to the potential implications for transport mechanisms in developing coffee grains and also the importance of the compounds analysed for industrial quality and flavour.
The aim of this work was to compare methods to determine the moisture content in wet weight (w.w.) of raw coffee beans. Six methodologies were evaluated in the range of 9 to 17 percent: A) oven with forced ventilation (ISO method); B) oven with forced ventilation (Brazilian Agriculture Ministry method); C) Vacuum oven at 98°C for 24 h; D) Vacuum oven at 70°C for 24 h; E) Capacitance method utilizing Gehaka G 600 equipment; F) calorimetric method, utilizing the infrared "AND" (130°C for 15 min). The ISO 1447 is the official method for raw coffee, and it was used as a standard for comparison with other methods.
Aspergillus strains were isolated from green coffee beans of different origin, and their ochratoxin producing ability was examined. The most prevalent species found in most samples was A. niger; some isolates of this species produced ochratoxins A and B. A. ochraceus strains were also isolated, some of which produced ochratoxins. The other Aspergillus species found in green coffee beans include A. flavus, A. fumigatus, A. tamarii, A. oryzae and A. parasiticus, none of which produced ochratoxins. Our results indicate that A. ochraceus is not the only Aspergillus species responsible for ochratoxin contamination of green coffee beans.
The natural occurrence of aflatoxin B1 (AFB1) and ochratoxin A (OTA) has been surveyed in 47 samples of commercial green coffee beans. Samples were collected in Nagoya City, Japan, during 1988–93. Using a combination of immunoaffinity chromatography and high‐performance liquid chromatography analysis, AFB1 and OTA were quantified with detection limits of 2 ng kg and 0.1 μg kg respectively. Positive rates and levels of AFB1and OTA were 32% and 2.0–32.9 ng kg, and 30% and 0.1–17.4 μg kg respectively. The samples contaminated with AFB1 or OTA were mainly imported from African and Asian countries, and several samples were found to be positive for these two mycotoxins for the first time in Japan, although their levels were low.
An apparatus is described which incorporates a pump, a hygrometer and a sample of food. It can be used to determine the equilibrium relative humidity of the air between the food particles in three minutes. It has been used to determine the moisture sorption isotherms for maize, tea, cocoa, parchment coffee and green coffee at 28°C.
An immunoaffinity column/HPLC procedure was developed to quantify low levels of ochratoxin A (OTA) in green coffee beans, roasted coffee beans, and soluble (instant) coffee with greater than 80% recoveries. The method was used to survey 116 soluble coffee samples from various countries and different manufacturers and showed contamination levels ranging from “not detectable” to 15.9 μg/kg. The highest levels of OTA were detected among soluble coffees that had been adulterated with coffee husks and/or coffee parchments (mean contamination level:  5.9 μg/kg). By comparison, OTA concentrations in pure soluble coffee samples were significantly lower, with a mean contamination level of 1.1 μg/kg. Although higher than normal figures have been found in some products purchased in East European countries, the results of this survey indicate that pure soluble coffee is not a major source of OTA in the diet, with estimated intakes being well within safety limits. Keywords: Ochratoxin A; soluble coffee; analysis; HPLC; adulteration
Mycotoxin contamination of agricultural products is a serious health hazard throughout the world. Besides attempts to eliminate mycotoxins from contaminated substrates by physical and chemical methods, the ability of microbes to degrade mycotoxins is now being widely examined. In this study, several Aspergillus species were examined for their ability to degrade ochratoxin A. A. fumigatus and black Aspergillus strains were found to detoxify ochratoxin A in culture media. The kinetics of ochratoxin A detoxification by an atoxigenic A. niger strain was examined by thin layer chromatography, high-performance liquid chromatography and an immunochemical technique. A. niger CBS 120.49 was found to effectively eliminate ochratoxin A from both liquid and solid media, and the degradation product, ochratoxin alpha, was also decomposed.
Since to our knowledge no data are available in the literature regarding the influence of green coffee type and origin on ochratoxin A (OTA) content, determinations were carried out in order to assess the level of OTA contamination in green coffee samples of different provenience. A total of 162 samples of green coffee beans from various countries (84 from Africa, 60 from America, and 18 from Asia) were analyzed for OTA. Both the amount and the variability of OTA levels were tested as a function of green coffee provenience. The results showed that 106 of the overall samples were positive for OTA, with concentration ranging from 0 to 48 microg/kg (ppb). In particular, it was possible to verify that African samples were more contaminated with respect to samples of other origin in terms of frequency and level of OTA; the highest concentrations observed were 18 and 48 microg/kg in two samples from The Congo.