Effects of temperature and medium composition on inhibitory activities of gossypol-related compounds against aflatoxigenic fungi



Jay E. Mellon, U. S. Department of Agriculture, Agricultural Research Service, Southern Regional Research Center, 1100 Robert E. Lee Blvd., New Orleans, LA 70124, USA. E-mail: Jay.Mellon@ars.usda.gov



To investigate the effects of temperature and medium composition on growth/aflatoxin inhibitory activities of terpenoids gossypol, gossypolone and apogossypolone against Aspergillus flavus and A. parasiticus.

Methods and Results

The compounds were tested at a concentration of 100 μg ml−1 in a Czapek Dox (Czapek) agar medium at 25, 31 and 37°C. Increased incubation temperature marginally increased growth inhibition caused by these compounds, but reduced the aflatoxin inhibition effected by gossypol. Gossypolone and apogossypolone retained good aflatoxin inhibitory activity against A. flavus and A. parasiticus at higher incubation temperatures. However, increased temperature also significantly reduced aflatoxin production in control cultures. The effects of the terpenoids on fungal growth and aflatoxin production against the same fungi were also determined in Czapek, Czapek with a protein/amino acid addendum and yeast extract sucrose (YES) media. Growth of these fungi in the protein-supplemented Czapek medium or in the YES medium greatly reduced the growth inhibition effects of the terpenoids. Apogossypolone displayed strong anti-aflatoxigenic activity in the Czapek medium, but this activity was significantly reduced in the protein-amended Czapek and YES media. Gossypol, which displayed little to no aflatoxin inhibitory activity in the Czapek medium, did yield significant anti-aflatoxigenic activity in the YES medium.


Incubation temperature and media composition are important parameters involved in the regulation of aflatoxin production in A. flavus and A. parasiticus. These parameters also affect the potency of growth and aflatoxin inhibitory activities of these gossypol-related compounds against aflatoxigenic fungi.

Significance and Impact of the Study

Studies utilizing gossypol-related compounds as inhibitory agents of biological activities should be interpreted with caution due to compound interaction with multiple components of the test system, especially serum proteins.


Aspergillus flavus is a ubiquitous saprophytic fungus commonly found in tropical and subtropical climes (Klich 2002). This organism is also an opportunistic pathogen of a number of oilseed crops (e.g. cotton, maize, peanuts, tree nuts) and has agronomic significance due to its production of the potent carcinogenic mycotoxin aflatoxin B1 (CAST 2003). Aspergillus parasiticus is a closely related fungus capable of producing the B series, as well as the G series of aflatoxins.

Gossypol is an optically active disesquiterpene (C30) produced by the cotton plant (Gossypium hirsutum). The compound is principally located in lysigenous glands found throughout aerial tissues, including cottonseed, and along the external surfaces of nonglandular root tissue. Cottonseed kernels contain, on average, about 1·3% gossypol by weight (NCVT 2001). The compound is considered an antinutrient component of cottonseed and cottonseed meal (Bernardi and Goldblatt 1980), which limits its use as an animal feed and practicably precludes its use as a human protein source. The compound is known to exhibit a wide range of bioactivity that includes anticancer, antimicrobial and antiviral effects (Wang et al. 2009). Gossypolone is a related compound that is formed by oxidation of gossypol with ferric chloride (Haas and Shirley 1965). Although less actively studied, the compound has been reported to exhibit some anticancer effects; in general, this activity is reduced in comparison with gossypol (Gilbert et al. 1995; Blackstaffe et al. 1997; Dao et al. 2000). Apogossypolone is a related derivative that is formed by conversion of gossypol to apogossypol followed by oxidation to apogossypolone (Adams and Butterbaugh 1938). This compound has recently been reported to have stronger activity against some cancer cells and is of particular interest because it binds to and interferes with Bcl proteins that are associated with disrupting apoptosis mechanisms in mammalian cancer cells (Arnold et al. 2008; Zhan et al. 2009; Niu et al. 2012).

Gossypol contributes to plant defences through insecticidal activity (Bottger et al. 1964) and may be involved in other plant defence functions such as fungal inhibition. The (−)-enantiomer of gossypol is four times more active than the (+)-enantiomer in inhibition of conidia germination, mycelial growth and conidiophore development in A. flavus (Mellon et al. 2003). More recently, racemic gossypol, optically active gossypol, and a number of related gossypol derivatives, were found to inhibit in vitro growth of A. flavus (Mellon et al. 2011). In addition, gossypol, gossypolone and apogossypolone demonstrate significant growth inhibitory activity against a diverse collection of economically important filamentous fungi (Mellon et al. 2012).

Our initial investigations regarding the effects of a series of gossypol derivatives revealed that both gossypolone and apogossypolone (Fig. 1) were more effective than gossypol as growth inhibitors against an A. flavus strain (Mellon et al. 2011). Also, preliminary experiments suggested some variability regarding growth and aflatoxin production inhibition activities associated with these terpenoids when used to treat aflatoxigenic fungi at different incubation temperatures and medium conditions. Thus, an investigation was initiated to further define the effects of these treatment parameters on fungal growth and aflatoxin production in terpenoid-treated A. flavus/A. parasiticus cultures.

Figure 1.

Production of gossypolone and apogossypolone from gossypol.

Materials and methods

Fungal strains

Isolates were selected from the Southern Regional Research Center (SRRC) Permanent Culture Collection on the basis of pathogenicity and toxigenicity. Aspergillus flavus isolate 1000-F was obtained from infected cottonseed. Aspergillus flavus isolate AF13 (ATCC 96044) was derived from a field soil sample (Arizona, USA). Aspergillus parasiticus isolate 143-A (ATCC 201461) was obtained from an infected peanut sample (Uganda). Fungi were screened for purity and toxin production prior to the start of the experimental work. Fungal inocula were constructed from mature 5-day-old cultures in sterile inoculation medium (0·0005% Triton X-100; 0·2% agar); concentrations of 106 spores ml−1 were determined with a haemocytometer.

Terpenoid preparation/purification

Gossypol–acetic acid (1 : 1) was isolated from cottonseed soapstock as described previously (Dowd and Pelitire 2001). Pure gossypol was obtained by dissolving the acetic acid solvate in diethyl ether and washing the ether phase with equal volumes of water three times. The ether phase was then evaporated under vacuum, and the gossypol product was stored under vacuum for several days to remove residual ether. Proton NMR spectroscopy indicated that the product contained only trace levels of both ether and acetic acid. Gossypolone was prepared from gossypol acetic acid by mild oxidation in the presence of ferric chloride (Haas and Shirley 1965) (Fig. 1). Apogossypolone was prepared by the basic procedure of Adams and Butterbaugh (1938) and Zhan et al. (2009), first eliminating the formyl groups in concentrated base to form apogossypol and then oxidizing the inner benzyl rings to yield apogossypolone (Fig. 1). Product yields were comparable to prior reports, and the products were essentially pure by HPLC (Mellon et al. 2011). The products had the expected NMR, UV–Vis and mass spectrometric properties.

Fungal incubations

Standard Czapek Dox (Czapek) medium (Thom and Raper 1945) with the addition of 10 mg l−1 ZnSO4·7 H2O, 5 mg l−1 CuSO4·5 H2O and 2% (w/w) agar was utilized as the fungal growth medium for the temperature experiment. The medium was adjusted to pH 6·0 before heat sterilization, followed by equilibration to 60°C. Terpenoids were dissolved in acetone as a carrier solvent and were dispersed in the culture medium to yield a test concentration of 100 μg ml−1. Control media were also treated with an equivalent amount of acetone (i.e. 5% v/v). The media were introduced into sterile, disposable Petri plates (9 cm, 25 ml per plate) and were allowed to solidify. Plates were placed in a dark fume hood for 24 h to allow acetone to dissipate. Following acetone evaporation, plates were stored at 5°C in the dark until fungal inoculation. Six replicates of each treatment were single-point inoculated (2 μl) in the centre of the plate and incubated in the dark at 25, 31 or 37°C for up to 14 days.

The medium study utilized the standard Czapek medium (described above), a protein/amino acid-amended Czapek medium and a yeast extract sucrose (YES) medium. The amended Czapek medium contained 4 mg ml−1 bovine serum albumin (BSA), 0·48 mg ml−1 lysine and 4·42 mg ml−1 alanine to simulate protein/free amino acid concentrations found in the YES medium. The BSA was added as a stock solution (0·05 g ml−1) by sterile filtration (0·22 μm filter; Millipore Corp., Billerica, MA, USA). This medium was prepared as given above and equilibrated to 60°C. Standard YES medium (20 g l−1 agar; 150 g l−1 sucrose) was prepared using similar procedures to those given above. The terpenoids were introduced to the media in concentrations of 100 μg ml−1 as described in the temperature experiment. Six replicates of each treatment were prepared. All fungal cultures in the media study were incubated in the dark at 25°C for up to 17 days.

Fungal growth determination

Fungal colony diameters were measured on a daily basis. Measurements were continued until control colonies filled their plates, at which time the entire treatment series for the fungal strain was terminated. Each colony diameter consisted of an average of two measurements taken at 90° to each other. Colony areas were calculated from diameter measurements; inhibition was taken as the ratio of the treatment and control areas expressed as a percentage.

An analysis of variance was conducted on the colony areas with SAS (version 9.3, Statistical Analysis System, Cary, NC, USA) Proc anova. Least significant difference testing of the treatments (α = 0·05) was conducted for each set of fungi and growth condition.

Aflatoxin analysis

Following termination of the experiment, three plates chosen at random from each treatment were extracted for toxin determination. The entire contents from each plate were macerated, placed in 75 ml of 65% acetone (v/v, aqueous) and shaken in an orbital shaker (25°C, 125 rev min−1) for at least 1 h. Solids were removed by filtration through qualitative filter paper (Whatman #4, Whatman Ltd., Maidstone, England). Methylene chloride, 25 ml, was added to the aqueous acetone filtrate. The biphasic mixture was shaken, the phases were allowed to separate, and the organic (lower) phase was collected. Water was removed with anhydrous sodium sulfate, and the solvent was evaporated at ambient temperature. Each sample was resuspended in 5 ml of methylene chloride and transferred to a 1-dram vial. Contents were again allowed to dry by evaporation and resolvated in a small aliquot of acetone (volume depending on toxin level). Four microlitres of each sample was spotted on silica gel G thin-layer plates (J T Baker, Phillipsburg, NJ, USA), which were developed in diethyl ether/methanol/water (96 : 3 : 1) mobile phase. Aflatoxins B1 and G1 were quantified directly on thin-layer plates by fluorescence densitometry (Shimadzu 9301PC, Shimadzu Corp., Columbia, MD, USA), comparing Rf values to aflatoxin standards (Sigma Chemical Co., St. Louis, MO, USA). Aflatoxin data were transformed to normalize area differences between control and terpenoid-treated plates. Aflatoxin analysis methodology has been previously described (Mellon et al. 2012).


Incubation of the A. flavus/A. parasiticus isolates (AF13, 1000-F, 143-A) at temperatures above 25°C resulted in the expected increased control growth rates. However, no further increase in growth rates was observed by raising the incubation temperature above 31°C (Table 1). In most cases, growth inhibition for the tested terpenoids was only marginally improved with increased incubation temperature (Table 1). Growth inhibition of A. parasiticus by apogossypolone was an exception to this trend, increasing from about 40% inhibition at 25°C to 71% at 37°C (Table 1). Complete growth inhibition of A. parasiticus 143A by gossypolone on the Czapek medium was observed, as found in a previous study (Mellon et al. 2012).

Table 1. Aflatoxigenic fungal growth in the presence of gossypol derivatives at different incubation temperaturesa
Fungal isolateTemperature, °CGrowth Periodb, daysMean colony area at the end of growth periodc, cm2LSD
  1. a

    Grown on Czapek Dox medium. All compounds tested at a 100 μg ml−1 concentration. Values expressed as area mean ± standard deviation (n = 6).

  2. b

    Day prior to control colony filling plate.

  3. c

    Means within a row followed by different superscripted letters are significantly different according to least significant difference (LSD) test at α = 0·05.

  4. d

    No growth observed.

Aspergillus flavus, AF13251251·1 ± 1·0A5·45 ± 0·28C5·09 ± 0·55C9·28 ± 0·99B0·94
31750·5 ± 1·2A2·90 ± 0·28D4·75 ± 0·87C6·17 ± 1·23B1·19
37752·0 ± 2·1A4·25 ± 0·30C4·43 ± 1·77C6·17 ± 0·65B1·70
A. flavus, 1000F251251·7 ± 0·7A5·76 ± 1·03C4·98 ± 0·32C8·41 ± 1·70B1·28
31750·5 ± 1·2A3·17 ± 0·68C4·20 ± 0·97C5·60 ± 1·15B1·24
37751·1 ± 1·0A5·00 ± 0·69C3·50 ± 1·05C6·38 ± 1·63B1·39
A. parasiticus, 143A251450·1 ± 1·2A19·6 ± 2·3C d 29·8 ± 2·9B2·79
311050·5 ± 2·7A11·1 ± 0·8C d 24·3 ± 1·5B2·30
371047·6 ± 2·6A11·6 ± 0·8B d 13·7 ± 2·5B2·61

Incubation temperature had a pronounced effect on the aflatoxin production of the fungi grown on Czapek medium. Increasing the temperature from 25 to 31°C decreased aflatoxin B1 production by several hundred fold in the A. flavus control cultures and by 20-fold in the A. parasiticus control cultures (Table 2). The same temperature change caused about a 200-fold reduction in aflatoxin G1 production by the A. parasiticus isolate. Increasing the incubation temperature to 37°C resulted in near cessation of aflatoxin production (both B and G series) for this isolate (data not shown). Apogossypolone exhibited strong aflatoxin inhibitory activity (99–100% inhibition) at all incubation temperatures. Increased temperatures raised the aflatoxin inhibition effects of gossypolone in the A. flavus isolates. Gossypol demonstrated some moderate anti-aflatoxin activity (B1) at 25°C in the A. parasiticus isolate, but that activity was not observed at higher temperatures (Table 2).

Table 2. Effect of gossypol-related compounds on aflatoxin production in Aspergillus flavus and A. parasiticus at different incubation temperaturesa
Fungal isolateTemperature, °CToxin typeControl plate toxin level (μg)% toxin inhibition with treatmentb
  1. a

    Data are reported as the average toxin per plate (n = 3). None of the control or treated cultures of A. parasiticus produced aflatoxins at 37°C. All cultures grown on standard Czapek media.

  2. b

    Toxin levels were adjusted to account for colony size differences as related in Table 1.

  3. c

    Gossypolone-treated cultures of A. parasiticus exhibited no growth.

A. flavus, AF1325B111·3 ± 1·804100
31B10·034 ± 0·0080100100
37B10·023 ± 0·0250100100
A. flavus, 1000-F25B118·9 ± 4·5035100
31B10·068 ± 0·0250100100
37B10·005 ± 0·0050100100
37G10·0020 ± 0·00050100100
A. parasiticus, 143A25B134·5 ± 6·256 c 99
31B11·70 ± 0·251 c 100
25G195·9 ± 15·40 c 99
31G10·429 ± 0·1140 c 100

Growth of these aflatoxigenic fungi on either the protein-amended Czapek or the YES media essentially eliminated growth inhibition exhibited by gossypol or gossypolone (Table 3). Although this same inhibitory activity was reduced in the apogossypolone-treated A. flavus cultures grown on the amended Czapek and YES media, some inhibition was still observed. Apogossypolone-treated A. parasiticus cultures did not follow this trend; they demonstrated increased growth inhibition levels in the richer media (Table 3).

Table 3. Aflatoxigenic fungal growth in the presence of gossypol derivatives in different culture mediaa
Fungal isolateMediumbGrowth periodc, daysMean colony area at the end of growth periodd, cm2LSD
  1. a

    Cultures grown at 25°C. Compounds tested at a 100 μg ml−1 concentration. Values expressed as mean ± standard deviation (n = 6).

  2. b

    Media designations: CZ = Czapek Dox (standard); CZ/protein = Czapek Dox amended with 4 mg ml−1 bovine serum albumin, 0·48 mg ml−1 lysine and 4·42 mg ml−1 alanine; YES = yeast extract sucrose (standard).

  3. c

    Day prior to control colony filling plate.

  4. d

    Means within a row followed by different superscripted letters are significantly different according to least significant difference (LSD) test at α = 0·05.

Aspergillus flavus, AF13CZ1353·0 ± 1·0A5·95 ± 0·48C6·54 ± 1·52C8·97 ± 2·80B2·03
CZ/protein1052·2 ± 1·1A52·8 ± 1·2A50·9 ± 1·1A34·8 ± 4·2B2·77
YES637·9 ± 0·4C43·7 ± 1·9A40·5 ± 0·5B12·3 ± 0·4D1·20
A. flavus, 1000FCZ1454·3 ± 1·5A7·93 ± 1·48C6·22 ± 1·33D9·68 ± 0·87B1·60
CZ/protein1052·1 ± 1·4A52·4 ± 1·3A48·4 ± 0·7B33·7 ± 3·8C2·61
YES635·7 ± 1·1C41·8 ± 2·1A38·0 ± 1·0B12·0 ± 0·1D1·54
A. parasiticus, 143ACZ1444·2 ± 2·0A24·9 ± 4·0C1·43 ± 3·5D30·4 ± 2·0B3·63
CZ/protein1150·9 ± 1·0A47·8 ± 2·4B48·6 ± 1·3B28·4 ± 1·4C1·97
YES747·5 ± 1·4C49·2 ± 1·5B50·9 ± 1·3A18·6 ± 0·6D1·51

In the 25°C medium experiment, gossypol and gossypolone demonstrated little aflatoxin inhibition on the Czapek medium (Table 4). Apogossypolone, however, yielded strong anti-aflatoxin activity (95–97% inhibition) with all of the tested fungal isolates, as was observed at 25°C in the temperature experiment (Table 2). Gossypol and gossypolone yielded similar low aflatoxin inhibition on the amended Czapek medium. Growth of the A. flavus isolates on the amended Czapek medium completely eliminated the anti-aflatoxin activity demonstrated by apogossypolone (Table 4). This same terpenoid did show moderate inhibitory activity (c. 60%) in the A. parasiticus isolate on the amended Czapek medium. Gossypolone demonstrated little to no anti-aflatoxin activity with these isolates grown on the YES medium (Table 4), and apogossypolone also yielded a reduced capacity for aflatoxin inhibition in this medium. Surprisingly, gossypol, which showed little to no activity in the Czapek or amended Czapek media, did provide good anti-aflatoxin activity in the YES medium (Table 4).

Table 4. Effect of gossypol-related compounds on aflatoxin production in Aspergillus flavus and A. parasiticus grown in different culture mediaa
Fungal isolateMediumbToxin typeControl plate toxin level (μg)% toxin inhibition with treatmentc
  1. a

    Data are reported as the average toxin per plate (n = 3). All cultures were grown at 25°C. Compounds tested at a 100 μg ml−1 concentration.

  2. b

    Media designations: CZ = Czapek Dox (standard); CZ/protein = Czapek Dox amended with 4 mg ml−1 bovine serum albumin, 0·48 mg ml−1 lysine and 4·42 mg ml−1 alanine; YES = yeast extract sucrose (standard).

  3. c

    Toxin levels were adjusted to account for colony size differences as related in Table 3.

  4. d

    Gossypolone-treated Czapek Dox cultures of A. parasiticus exhibited negligible growth.

A. flavus, A-13CZB17·1 ± 0·460097
CZ/proteinB190·7 ± 5·51800
YESB1135 ± 149602
A. flavus, 1000-FCZB15·06 ± 0·950095
CZ/proteinB159·5 ± 2·32600
YESB1339 ± 1175060
A. parasiticus, 143ACZB132·4 ± 3·018 d 95
CZ/proteinB166·3 ± 2·3112959
YESB1231 ± 752016
CZG183·2 ± 8·30 d 95
CZ/proteinG1266 ± 7264564
YESG1448 ± 2078851


The dramatic reduction in aflatoxin production with increased incubation temperature observed with the control cultures is consistent with previous studies (Schindler et al. 1967). This temperature effect has been attributed to reduced expression of functional aflatoxin biosynthetic enzymes or a nonfunctional AFLR transcription factor (OBrian et al. 2007). Prevalence of this negative environmental effect on aflatoxin production begs the question as to why aflatoxin contamination is a problem in field conditions at all, given that ambient temperatures in some endemic problem areas (e.g. cotton, southwest United States; maize, southeast United States) can be well above 37°C. However, field temperatures are not constant, but instead cycle through highs and lows. These low temperatures (about 25°C) would place aflatoxigenic fungi well within a toxin production range, and once produced, the toxin is temperature stable (Raters and Matissek 2008).

Effects of increased temperature on growth inhibition expressed by the gossypol-related compounds on these fungi appeared to be marginally increased. Elevated temperature also increased the potency of aflatoxin inhibitory activity shown by gossypolone and apogossypolone. In the case of gossypol-treated A. parasiticus cultures, the moderate anti-aflatoxin activity exhibited at 25°C was not observed at 31°C.

Reduction in growth inhibition by the treatment compounds in the amended Czapek and YES media suggests that these polyphenolic compounds might affect some aspect of nitrogen assimilation, as providing a medium with a variety of nitrogen sources appears to block the inhibition. The near complete loss of growth inhibition exhibited by gossypol and gossypolone on the rich media suggests that other factors might also contribute to the effects. This might be explained by the formation of Schiff base-type complexes between the aldehyde moieties of the terpenoids (Fig. 1) and amine groups associated with proteins or free lysine. For gossypol, this reaction occurs freely with both proteins (King et al. 1958; Conkerton and Frampton 1959) and amine-containing compounds (Kenar 2006). Such derivatization could allow for some of the inhibitors to be covalently bound and sequestered either inside or outside the cell, which would lower the effective concentration of these compounds, thereby reducing their activity. A similar reduction in the anticancer activity of (−)-gossypol has been reported when the growth media contains either serum (Blackstaffe et al. 1997) or BSA (Huang et al. 2006). Apogossypolone would not be subject to this process, because it does not possess free aldehyde groups (Fig. 1), which might account for the less severe reduction in growth inhibitory activity observed for this compound (Table 3).

While this seems to account for the general data trends, other factors might also be contributing to the overall effects. As gossypol has been reported to inhibit hexose transport (Pérez et al. 2009), some inhibition of carbon assimilation might also be present. Higher initial sugar concentrations were part of the YES media, which might contribute to reduced growth inhibition. As no inhibition (possibly even a little enhanced growth) was observed in the gossypol- or gossypolone-treated cultures of this medium (Table 3), this might be a secondary contributor to the observed results. In addition, it is also possible that the BSA present in the amended Czapek medium might physically bind the compounds, as the biological function of this serum protein is to transport lipophilic substances. This effect would also potentially reduce their effective cellular concentrations. However, unlike Schiff base formation, this mechanism could also affect apogossypolone levels, which might account for the reduced inhibition seen with this compound in the amended Czapek medium compared with the standard Czapek or YES media.

For the media experiment, standard Czapek medium was amended with BSA and the amino acids lysine and alanine to simulate in a defined medium the protein and free amino acid levels found in the YES medium. Growth of these aflatoxigenic fungal isolates (controls) on either the amended Czapek or YES media resulted in a significant increase in aflatoxin production (both B and G classes). This result is not surprising, because it is known that A. flavus grown on a medium containing a readily accessible carbon source (free sugar) and a proteinaceous nitrogen source (oilseed storage proteins) increases aflatoxin B1 production by 4- to 10-fold (Mellon and Cotty 1998).

The mechanism of aflatoxin inhibitory activity expressed by apogossypolone is currently unknown. There is good evidence to support the model that later stages of aflatoxin biosynthesis occur in vesicles termed ‘aflatoxisomes’ (Chanda et al. 2009). Further, the export of aflatoxin outside of the mycelium appears to occur through exocytosis by fusion of these vesicles with the fungal plasmalemma membrane (Chanda et al. 2010). Possibly, apogossypolone exerts its anti-aflatoxigenic effects either by interfering with the formation of aflatoxisomes or by interfering with transport of aflatoxin substrates through vesicle membranes. The severe reduction in apogossypolone anti-aflatoxigenic activity by inclusion in the amended Czapek or YES media cannot be explained by concentration reduction effects due to Schiff base formation between aldehyde moieties and free amines. In addition, expression of significant aflatoxin inhibitory activity by gossypol in the YES medium, with little to no activity expressed in the Czapek or amended Czapek media, suggests this result occurs by means of one or more mechanisms distinct from that of apogossypolone.

In this study, the potency of the gossypol-related compounds as fungal growth and aflatoxin inhibitors was somewhat affected by incubation temperature and greatly affected by the metabolic environment to which they are subjected. The differences in activity observed among these closely related compounds suggest that the activities are likely effected through a multitude of mechanisms. Inconsistent literature regarding the activity and toxicity of these compounds is not uncommon. The results support a possible explanation to this variable activity in that the effects are derived as the result of a several mechanisms, and consequentially, the net effects tend to be sensitive to the exact conditions and protocols used for testing.