Toxicology and Applied Pharmacology 223 (2007) 125 132

www.elsevier.com/locate/ytaap

Mixtures in the real world: The importance of plant self-defense toxicants,

mycotoxins, and the human diet
Joel L. Mattsson
14155 Pepin Place, Carmel, IN 46032, USA
Received 17 February 2006; revised 28 November 2006; accepted 23 December 2006
Available online 3 January 2007

Abstract
A perusal of research presented at the Annual Society of Toxicology Meetings, or in nearly any toxicology journal, will show that the
overwhelming emphasis of toxicology research is on synthetic chemistries. Because of substantial potency and exposure to natural chemicals, the
overwhelming focus on synthetic chemistries cannot lead to a realistic understanding of chemical risk to the general population. Natural chemicals,
simply because of their abundance and potency, may be as likely to be a public health concern and to be involved in chemical interactions (natural:
natural, natural:pharmaceutical; or natural:synthetic) as are environmental levels of synthetic chemicals. All plants have a mix of natural selfdefense chemistries and mycotoxins that, when tested in a manner comparable to synthetic pesticides, cause the entire spectrum of toxic effects. As
a further complication, plants also escalate much of their self-defense chemistry when attacked by insects and fungi, and damaged crops often have
higher mycotoxins levels. Effective crop protection will typically reduce the plant's levels of self-defense toxicants and mycotoxins, but may add
residues of synthetic pesticides or add some other risk variable. In addition, cooking may also alter the food chemistry (e.g., acrylamide). The
mixtures toxicologist needs to address the real world mixture of natural and synthetic chemicals. Public policy on crop-food safety cannot be
sensibly guided without these data and large voids in our understanding of risks from real-world mixtures cannot be in the public interest.
2007 Elsevier Inc. All rights reserved.
Keywords: Plant self-defense natural toxicant; Mycotoxins; Food; Diet; Pesticides; Mixtures; Risk-assesment

Introduction
People are exposed to natural toxicants, pharmaceuticals,
and synthetic chemicals. The purpose of this overview is to
convince toxicologists of the necessity to fully consider natural
toxicants when evaluating risk of the public to environmental
toxicants. The historical lack of systematic scientific and
regulatory consideration of natural toxicants as an integral
part of the human experience must, by its absence, lead to an
incomplete and possibly distorted understanding of overall
chemical risk to human health. The concepts presented here are
simple extensions of arguments presented by Bruce Ames and
Lois Gold (Ames, 1983; Ames and Gold, 1997; Ames, 2003).

Topic presented at the conference, Charting the Future: Building the

Scientific Foundation for Mixtures Risk Assessment. Atlanta, GA, Feb. 1617,
2005. The author was employed in 2005 by Dow AgroSciences LLC,
Indianapolis, IN 46268. The opinions expressed in this paper are solely those
of the author.
E-mail address: jmattsson@indy.rr.com.
0041-008X/$ - see front matter 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.taap.2006.12.024

As expressed in a biomonitoring paper by Paustenbach and

Galbraith (2006, p. 253): In fact, dietary exposure to these
naturally occurring toxicants can greatly exceed any type of
environmental exposures to these chemicals and the risks of
dietary intake can also be appreciable. Rigorous analyses must
consider that naturally occurring chemicals in the human diet
may cause adverse health effects similar to those present due to
industrial processes.
Obviously, humans are exposed to complex and highly
variable mixtures of natural and synthetic chemistries which
create opportunities for simple and complex interactions. The
term interaction as used here is very general and simply denotes
that, in one way or another, chemical A and chemical B (and
perhaps C, D, and E) alter the biological effect of one another.
Although there is no thorough or systematic approach to
study and regulation of natural toxicants, there are thousands of
papers and many textbooks about the natural toxicants of plants
and plant foods. Most of the natural toxicant literature deals
with acute or sub-chronic exposures. Very few of these studies
are chronic or address mixtures or interactions. A small

126

J.L. Mattsson / Toxicology and Applied Pharmacology 223 (2007) 125132

sampling of textbooks, by no means complete, would include

Toxicants Occurring Naturally in Foods (NAS, 1973), Toxicants of Plant Origin (Cheeke, 1989), Toxic Substances in Crop
Plants (D'Mello et al., 1991), Handbook of Plant and Fungal
Toxicants (D'Mello, 1997), and Foodborne Disease Handbook,
Volume 3 (Hui et al., 2001). Even a casual reading of these
textbooks will demonstrate that eating plants will inescapably
expose us to a myriad of complex and potent chemistries,
sometimes at surprisingly high levels.
The 624-page report on toxicants of foods by the National
Academy of Science (NAS, 1973) was prompted by several
concerns:
Perhaps the main one is the hope that it may contribute to a
more informed, realistic, and sensible attitude on the part of
the public toward the food supply (p. 2).
In any case, it is clear that the real challenge that we face is
the question of the long-term chronic toxicity, or lifetime
effects, of the known and yet unknown natural chemical
components of our foods. Such effects that might result
from ordinary patterns of consumption are of the greatest
potential importance, since they would be expected to affect
the largest number of people. Though the problem of
carcinogenesis has been emphasized above, similar attention
should be focused upon reproductive functions, mutagenesis, cardiovascularrenal diseases, mental disorders, and
other chronic ills of mankind of which the causes are
unknown (p. 580).
Twenty-eight years later, similar concerns about natural
toxicants of foods were expressed by Beier and Nigg (2001):
We are exposed to a plethora of natural food chemicals as
mixtures. For those that have been tested, the testing
procedures were limited in scope, design, and dose range
(p. 133).
There are no guidelines or regulations regarding naturally
occurring toxicants in food. Do we know what we have
placed in the marketplace? We do not. We have always
experimented with foods on ourselves. A basis for chemical
change(s) in our food supply should be established now. We
have the tools and scientific talent to obtain the knowledge
for that base (p. 136).
For risk assessments of environmental mixtures of chemicals
to be scientifically rational, our exposure to natural chemicals
found in foods cannot be ignored. Emphasis should be given to
those natural and synthetic chemicals that are most potent and to
which exposure is greatest. Good science is possible if both
natural and synthetic chemicals are evaluated by the same set of
toxicologic principles. There should be no double standard as
exists at present (Mattsson, 1996).
Interactions with natural chemicals
Most of the recognized interactions between natural and
synthetic chemistries are between natural food chemicals and

pharmaceuticals. It is now recognized that furocoumarins in

grapefruit juice can sufficiently inhibit intestinal CYP3A4 to
cause clinically-relevant, reduced first-pass metabolism of
numerous pharmaceuticals. Bergamottin may be the most
important furocoumarin for this effect in grapefruit juice, and
concentrations of bergamottin are about 12 to 25 uM (He et al.,
1998).
Murray (2006) reported on other potential dietary interactions with P-450 enzymes. Some common dietary constituents
such as fats and carbohydrates can alter P-450 activity, and
chemicals in teas and cruciferous vegetables may also inhibit
human CYP enzymes. Thus, food constituents modulate CYP
expression and function by a range of mechanisms, with the
potential for both deleterious and beneficial outcomes (Murray,
2006).
Hu et al. (2005) have reviewed interactions between herbal
medicines and prescribed drugs, and many have caused serious
clinical problems. For example, hypericum (St. John's wort)
taken at the same time as oral contraceptives has been
associated with breakthrough bleeding and unplanned pregnancies. In spite of the risk of interactions, the authors state the
underlying mechanisms for the altered drug effects and/or
concentrations by concomitant herbal medicines are yet to be
determined.
An editorial by Kaneko and Ishigatsubo (2005) warns that
isoniazid, an effective and commonly used treatment for
tuberculosis, can interact with fish, cheeses, or wines that
have high histamine or tyramine contents. Isoniazid (INH) can
interfere with the metabolism of ingested histamine and
tyramine, and the potential for frightening and dangerous
interactions between INH and certain foods is presently little
known.
Alcohol is consumed world-wide, and the U.S. National
Institutes of Alcohol Abuse and Alcoholism (NIAAA, 1995)
have stated many medications can interact with alcohol,
leading to increased risk of illness, injury, or death. For
example, it is estimated that alcoholmedication interactions
may be a factor in at least 25 percent of all emergency room
admissions. An unknown number of less serious interactions
may go unrecognized or unrecorded. Natural contaminants of
foods are also a concern. There are more than 350 types of
mycotoxins that might be found in food or feeds (Doerr, 2003),
and concerns have been expressed about the likelihood of
interactions among mycotoxins (D'Mello et al., 1997; Speijers
and Speijers, 2004). Mycotoxin issues are discussed more fully
below.
The following text will illustrate that many natural chemistries of foods have substantial potency and sometimes a
surprisingly high content in foods, thereby increasing their
likelihood as participants in interactions of public health
significance.
Exposure and potency
Toxicology is based on the principle that risk is a function of
both exposure and potency. The body simply does not care
who or what made a chemical or why it was made. The only

J.L. Mattsson / Toxicology and Applied Pharmacology 223 (2007) 125132

relevant concerns are potency and dose. Since there is a great

deal of public discussion as well as regulatory and research
emphasis on human exposure to pesticide residues in foods,
these synthetic chemical residues will be used as a reference
point for natural toxicants.
Synthetic pesticide residues in food
There is abundant information on pesticide residues in foods.
The amount of allowable pesticide residue in foods is regulated.
First, a wide variety of required toxicology studies are
conducted, a relevant health effect is selected, and noobserved-adverse effects levels (NOAELs in mg/kg body
weight) are determined. An allowable one-day intake limit is
called an acute reference dose (ARfD, in mg/kg body weight),
and is typically 1/100 of an appropriate acute NOAEL.
Similarly, an allowable limit for lifetime exposure is called an
allowable daily intake (ADI, in mg/kg body weight per day),
and is typically 1/100 of the chronic or lifetime NOAEL.
Pesticide levels in food are regulated and monitored. The
maximum permissible residue levels, in ppm or mg/kg food, are
called tolerances by the USEPA and maximum residue levels
(MRL) in Europe. Tolerances are established for each pesticide
and for each crop by integration of toxicology data (ARfD and
ADI), good agricultural practices, and expected human
consumption.
Residues of synthetic pesticides in foods (including imports)
in the United States in 2003 were below the limit of detection in
about 50% of sampled foods (USDA, 2005). When detectable,
residue levels were typically a small fraction of one ppm. For
those fruits, vegetables and grains that had a pesticide detection
rate of 10% or higher, the mean amount of pesticide residue at
the 90th percentile was 0.13 ppm, with a 90th percentile median
of 0.02 ppm. At the 90th percentile, the mean levels of pesticide
residues were on average only 2% of the tolerance, with a
median of 1% (derived from Appendix K, USDA, 2005).
About 24% of foods contained residue of more than one
pesticide. Residues exceeded regulatory limits (tolerances) in
about 0.3% of samples (USDA, 2005). While exceeding
tolerance is a regulatory concern, the U.K. Pesticides Residue
Committee (PRC, 2004, p. 24) states MRLs [tolerances] are set
at levels that would not result in intakes high enough to cause
health risks. In most cases, residues at the MRL would result in
intakes considerably below both the ARfD and the ADI. So,
even if a residue is above the MRL, this does not automatically
result in an intake above the ARfD or the ADI. To say this
simply: Wide safety margins are designed into the tolerances.
Analytical results were similar in Great Britain (PRC, 2004).
About 70% of the samples tested had no detectable levels of
pesticides, and about 1% of residues were above the regulatory
limit.
Natural toxicants of foods
Unlike synthetic pesticides that are vigorously tested and
regulated to minimize human exposure to worrisome types of
toxicity and to worrisome levels in our foods, the natural

127

toxicants of food plants are not similarly constrained. There are

numerous opportunities for toxicity and for interactions from
natural chemicals of foods simply because of their great
numbers and sometimes high levels.
The following references are but a tiny fraction of those in
the literature, but will illustrate the diversity and typical
concentrations of natural toxicants in foods. In 1983, Bruce
Ames wrote a review of the substantial amounts of natural
carcinogens in foods (Ames, 1983). Ames stated that plants in
nature synthesize toxic chemicals in large amounts, apparently
as a primary defense against the hordes of bacterial, fungal, and
insect and other animal predators (p. 1256). Ames' paper
discusses many potential carcinogens that occur in the human
diet at trace to very high levels. For example, pyrrolizidine
alkaloids are carcinogenic, mutagenic, and teratogenic and are
present in thousands of plant species (often at > 1% by weight),
some of which are ingested by humans, particularly in herbs and
herbal teas and occasionally in honey (p. 1257).
Pyrrolizidine alkaloids (PA) are a recognized cause of human
intoxication, liver cirrhosis, veno-occlusive disease and liver
cancer, particularly in Jamaica, India, and parts of Africa
(Coulombe, 2001). In a referenced but non-peer reviewed
correspondence, Daughton (2005) points out that regulatory
authorities have identified PAs as a major human health threat,
especially for fetuses and infants. In particular, Daughton
(2005) identifies PAs as a missing but plausible candidate in the
etiology of some specific cancer clusters because PAs can
exhibit aperiodic cycles of high expression. Honey was cited
as an example, since PA levels can vary two or more orders of
magnitude within the same foraging location and by time of
year.
Because of potentially high levels of PAs, the Food
Standards Australia New Zealand (FSANZ) released a consumer advisory on 9 February 2004 to limit consumption of
honey from particular plants. Some batches of honey can attain
levels of PAs of 1 mg/kg, and levels of PAs found in various
grain commodities in Australia have ranged from < 50 to
>6000 g/kg (ANZFA, 2001). The provisional tolerable daily
intake (PTDI) for PAs was established by ANZFA (2001) at
1 g/kg body weight per day, with a caveat that further
characterisation of the potential human health risk from
exposure to PAs in food is not possible because there is
currently inadequate dietary exposure information.
The U.S. National Research Council conducted an extensive
review of food carcinogens in 1996. While concluding that
human risk of cancer from natural and synthetic carcinogens in
our foods is low, the U. S. National Research Council (NRC,
1996, Appendix A) provides data on natural chemical
concentrations of some possible or probable rodent carcinogens
found in food: allyl isothiocyanate in horseradish (2000 ppm)
and black mustard seed (10,000 ppm); benzaldehyde (found in
over 40 foods) in white bread (510 ppm); capsaicin in hot red
peppers (up to 10,000 ppm of dried fruit); estragol in basil and
oregano (approx. 100 ppm), in tarragon (approx. 10,000 ppm),
in anise and star anise (approx. 50,000 ppm); furfurol in cocoa
and coffee (55 to 255 ppm), in wine (trace to 10 ppm), in many
fruits (trace to 1 ppm); safrole in cocoa, nutmeg, mace, black

128

J.L. Mattsson / Toxicology and Applied Pharmacology 223 (2007) 125132

pepper (approx. 2000 ppm); benz[a]anthracine in coconut oil

(0.5 to 13.7 ng/g), in broiled fish (0.6 to 2.9 ng/g), in smoked
fish (0.2 to 189 ng/g); benzo[a]pyrene in margarine (0.9 to
36 ng/g), in broiled meat (0.17 to 50 ng/g), in smoked fish (1 to
78 ng/g), ham (up to 14.6 ng/g), spinach (7.4 ng/g), tea (0.3 to
15.8 ng/g). This listing demonstrates that many natural
carcinogens or putative carcinogens are present in our foods
at levels that are far greater than synthetic pesticides.
Not all natural toxicants of plant foods are made by the plant
or are external contaminants. There has recently been a great
deal attention paid to levels of acrylamide in some fried and
oven-baked foods (Dybing and Sanner, 2003; CFSAN, 2005).
Acrylamide is a natural product of high cooking temperatures in
certain carbohydrate-rich foods, and acrylamide is mutagenic,
carcinogenic, and neurotoxic. Concentrations from CFSAN,
2005 data were highest in ground coffee (27609 ppb), potato
chips (5051970 ppb), cereals (25534 ppb), crackers (39
1540 ppb), cookies (34955 ppb), and other foods. Mean
dietary intake of acrylamide in Norwegian adults was about 0.5
ug/kg/day. Intake at the 97.5 percentile was about 3 higher for
adults and 45 higher in adolescents (Dybing and Sanner,
2003). When evaluated by linear extrapolation methods
(Dybing and Sanner, 2003), carcinogenic risks from dietary
acrylamide were very high (90th percentile intake lifetime
cancer risk for males was 13 cases per 10,000). Recently, the
FAO/WHO Expert Committee on Food Additives (FAO/WHO,
2005) used Bench Mark Dose methods and calculated margins
of exposure for cancer (MOE = NOAEL/Exposure). For average
exposure, the MOE was 300, and 75 for those with high
exposure. The committee expressed concern that these MOE's
were low for genotoxic carcinogens.
Perhaps the most studied of the plant self-defense toxicants
are the various glycoalkaloids (GA) of potatoes. GA are also
abundant in green peppers, red peppers, and green tomatoes.
Most self-defense chemistry of plants is located at the interface
with the outside world. It is logical, therefore, that concentrations of potato GA are highest in the skins (300 to 600 ppm),
and lowest in the flesh (12 to 50 ppm). Whole tubers contain
about 10 to 180 ppm unless damaged. When stimulated by light
(greened potatoes) or damaged physically or by infection, levels
of GA can soar (bitter tubers, 250 to 800 ppm) (MAFF, 1996;
Percival and Dixon, 1997).
Potato GAs cause acute gastrointestinal toxicity in humans at
about 2 mg/kg body weight, and cause serious neurological
toxicity and possibly death when exposure exceeds 3 mg/kg
body weight (Morris and Lee, 1984; Mensinga et al., 2005). As
noted by Hopkins (1995), this is a worringly steep dose
response curve. If manufactured as a synthetic pesticide, the
acute potency of GA would result in a WHO classification Ia,
extremely hazardous (death <5 mg/kg oral dose) or Ib, highly
hazardous (death < 50 mg/kg oral dose) (WHO, 2002).
Hopkins (1995) wrote that it seems very clear, on the basis
of the established principles of evaluating chemical toxic
potential, that the potato glycoalkaloids are deserving of the
most detailed experimental attention and that, thus far, they
have been subject of only limited investigation (p. 325). It was
pointed out by Hopkins (1995) that humans appear to be

markedly more susceptible than rats and mice, and that it would
be important to identify an appropriate species for hazard
assessment studies. Potato GA are also structurally related to
jervine, a recognized teratogen in sheep (Gaffield and Keeler,
1996). GA and their metabolites had a teratogenic potential of
0% to 50% of jervine in hamsters treated by gavage (Gaffield
and Keeler, 1996), and dietary solanidine has caused abortion of
fetuses in mice (Friedman et al., 2003).
The recent work of Mensinga et al. (2005) with human
volunteers lends gravity to the concerns of Hopkins (1995).
Mensinga et al. (2005) reported nausea and vomiting in one
person given 1.25 mg GA/kg body weight, and that the
clearance of GA from serum of human subjects takes more than
24 h (t1/2 range for alpha chaconine 2784 h; alpha-solanine 5
42 h). These clearance times indicate that daily consumption of
potato products may cause accumulation of glycoalkaloids,
which may consequentially lead to adverse health effects.
In Great Britain, it was estimated that average, long-term
GA intake was about 2.4 mg/person/day, with the 97.5
percentile person ingesting 7.8 mg/person/day (MAFF, 1996).
If a body weight of 70 kg is assumed, the daily dose of GA
would approximate 0.03 to 0.1 mg/kg/day. A probabilistic oneday estimate of GA intake for children 1 to 6 years of age was
0 mg/kg body weight at the 50th percentile, 0.2 mg/kg at the
80th percentile, and 0.7 mg/kg at the 95th percentile (Table 3 of
Mattsson, 2000). Exposure would be less for people eating
only peeled potatoes. Cooking does not reduce GA exposure,
and damaged or greened potatoes can substantially increase GA
exposure. Although reports of potato poisoning in humans are
rare, it appears that safety margins are often small. It remains
unexplored if there are cofactors that would alter the risk for
GA overdose. GA may aggravate inflammatory bowel disease
(Patel et al., 2002), which raises questions of altered absorption
of dietary GAs in certain subpopulations.
Furocoumarins also appear to be plant self-defense chemicals. Furocoumarins are found in parsnips, celery, parsley, citrus
fruits, peas, beans, peanuts, and other crops. Levels of
furocoumarins can vary widely depending on the type of
plant and degree of activation of furocoumarin chemistries
(MAFF, 1996). Typical levels of furocoumarins in parsnips are
20 ppm, organic parsnips 48 ppm, celery 0.9 ppm, organic
celery 10 ppm, parsley 38 ppm, Seville oranges 12 ppm.
Damage to plants significantly elevated furocoumarin levels,
and cooking moderately reduced furocoumarin levels (MAFF,
1996).
The furocoumarins are both mutagenic and carcinogenic.
Furocoumarins are photoactivated, and the most commonly
reported adverse effect is photodermatitis from either direct
contact with the crop or from oral ingestion. Celery photodermatitis is the best studied. The threshold for phototoxicity
is estimated at 18 ppm fresh weight for acute exposures, and
79 ppm for repeated exposures (Diawara and Trumble,
1997). The long-term furocoumarin consumption by the
average adult was estimated to be 0.24 to 1.28 mg/person/
day (roughly 0.003 to 0.018 mg/kg/day); while on any single
occasion could be as much as 10 higher. Children, on
average, might ingest higher levels (MAFF, 1996).

J.L. Mattsson / Toxicology and Applied Pharmacology 223 (2007) 125132

The National Academy of Sciences (NAS, 1973) stated

concerns about the health consequences of long-term dietary
habits. A large, comprehensive study was conducted by White
et al. (2000a) based upon expectations that a long-term
traditional Japanese lifestyle, culture and diet would be
protective of brain function, and acculturation to a Western
lifestyle and diet would be a risk factor. The subjects were
Japanese that had immigrated to Hawaii, and the data led
rapidly to a focus on tofu and brain function and brain atrophy.
The results were nicely summarized in an editorial by Grodstein
et al. (2000): [it was found] contrary to their own and others
expectations, that men who consumed greater amounts of tofu
during midlife appeared to score worse on cognitive tests, to
have lower brain weight and to demonstrate ventricular
enlargement on MRI . Post hoc, the results were substantially
replicated in the men's wives who presumably had a similar
diet. A criticism (Guo et al., 2000) and further analyses were
conducted (White et al., 2000b), with no change in conclusions.
As White et al. (2000a) stated: In this study population, 20% to
25% of the burden of cognitive impairment appears attributable
to midlife tofu consumptionan effect size of enormous public
health importance (pp. 252253). Although there are many
caveats about the meaning of these data, they do point to a need
for careful consideration of long-term dietary factors and human
health, and avoidance of assumptions about the health aspects
of foods in the absence of adequate data.
A change in consumption patterns of a traditional food can
lead to health problems when safety margins are small. A
vegetable (Sauropus androgynus) that is common in Malaysia
became a health food for weight loss and other benefits in
Taiwan (Ger et al., 1997; Hsiue et al., 1998). For those seeking
a health benefit, consumption increased from a typical onemeal per week (150 g) to about 130 g/day, and instead of
cooking, the vegetable was often prepared as a juice although
cooking was still common. Patients were mostly women, and
moderate to severe pulmonary disease (bronchiolitis obliterans)
occurred in a dose-related fashion, and a small number of
deaths occurred. The consumption period for 49 patients with
obstructive lung disease ranged from 12 to 150 days (Table 3 of
Hsiue et al., 1998), and the etiologic agent in the vegetable is
unknown.
Plants are commonly invaded by fungi, and many fungi make
toxins of public health concern. For example, the prevalence of
neural tube defects (NTD) doubled among Mexican-American
women along the TexasMexico border in 19901991 (Missmer et al., 2006). In 1989, there was an outbreak of equine
leukoencephalomalacia that was particularly severe in Texas,
and it was suggested that the equine outbreak and the human
NTD outbreak were related to high fumonisin levels in corn. The
epidemiological evaluation of the NTD data supported the
possible etiological role of corn fumonisins (Missmer et al.,
2006). A review by Marasas et al. (2004) indicated fumonisins
are able to inhibit embryonic sphingolipid metabolism and this
appears to interfere with folate utilization to produce embryotoxicity and neural tube defects (p. 713). Marasas et al.
concluded that diets low in folate but with elevated levels of
fumonisins might be of particular concern.

129

Animals are more often afflicted with mycotoxin overexposure, and Doerr (2003) nicely summarizes the dilemma:
There are, perhaps, half a dozen or more mold genera with
which we are usually most concerned (although there are many,
many more that can adversely affect animals). Fusarium,
Aspergillus, Alternaria, Penicillium, Stachybotrys, and Helminthosporium are representative of this group. For each genus
there are numerous toxigenic species. Many of those species
have the capacity to produce more than one kind of mycotoxin,
which, in part, accounts for the more than 350 recognized
toxins. And, each, in turn, may infest those several feed
components mentioned above and leave behind one or more
toxic residues. What begins as our concept of a single mold and/
or single mycotoxin is seen by the animal as a multitude of toxic
challenges. D'Mello et al. (1997) state: Routinely, animal
toxicity problems occur in which the quantity of the individual
mycotoxins found in the suspected feed(s) does not explain the
observed syndromes. Currently, the combined effects of
mycotoxins on animal and human health have aroused concern
because synergistic activities present a unique set of problems in
defining both toxicity and food safety guidelines (p. 294).
Concerns about interactions among mycotoxins have also been
expressed by Speijers and Speijers (2004). These concerns for
possible interactions among mycotoxins, and between fumonisins and folate, demonstrate the potential interactive complexity
of this class of chemistry. I have not yet identified papers
dealing with possible interactions between mycotoxins and
other potential toxicants of foods, or between mycotoxins and
synthetic chemicals, but these remain relevant questions.
Hammond et al. (2004) reported on fumonisin levels in U.S.
corn from 107 locations. Over 3 years, total fumonisin levels in
60% of the samples were above 2 ppm, which is the U.S. FDA
guidance level for human food. Levels ranged from not detectable
(<0.5 ppm) to about 25 ppm. Aflatoxin levels were most often not
detectable at 20 ppb, but a couple of samples were 1500 ppb or
higher. Mycotoxin levels typically were lower in corn-borerprotected Bt hybrids. That levels of fumonisins in food can
sometimes be high is illustrated by the U.K. Food Standards
Agency (2003) report of the market withdrawal of maize meal that
contained fumonisins at about 4.7 and 20 ppm. As with other
natural toxicants, it appears that mycotoxins are commonly in our
foods at levels appreciably higher than synthetic pesticides with
an implication that potential risk also would be greater.
Mycotoxins can be found in unexpected items. Ochratoxin A, a
carcinogenic mycotoxin, has been measured in red and white wines
(Shepard et al., 2003). Ochratoxin levels were about 0.2 ug/L, with
a high level of 0.4 ug/L. Mycotoxins are also found in grain dust,
raising questions about inhalation exposure (Nordby et al., 2004).
Mean levels reported were 31 ppb for deoxynivalinol, 62 ppb for
T-2, and 130 ppb for HT-2. Maximum levels were 10 to 20 higher.
Discussion of perceptions of natural toxicants
Voluntary versus involuntary exposure
Because people voluntarily eat food, some people argue
that they also voluntarily eat the natural pesticides and other

130

J.L. Mattsson / Toxicology and Applied Pharmacology 223 (2007) 125132

toxins of foods. It takes only a moment to recognize that this

is not true. Hardly anyone in the general public knows of the
great variety and quantity of natural toxicants of foods. There
is no systematic study of these natural toxicants (Beier and
Nigg, 2001), and certainly, there is no systematic education of
the public about these toxicants. In contrast, synthetic
pesticides are extensively studied, regulated, and monitored,
and the public is constantly exposed to realistic and to
alarmist information about synthetic pesticides in their food.
Given that organic produce is now commonly available, and
it is widely advertised that these products do not have
pesticides, it could be argued that consumption of synthetic
pesticides is far more voluntary than consumption of natural
toxicants of foods.
Exposure to natural toxicants is under human control
Clearly, plants must have natural-pesticide defenses and
plants will commonly be contaminated by fungal toxins.
Consequently, some people argue that this is just the way it
is. This perspective is, or course, incorrect. The levels of
natural toxicants of plants are very much under human control
when this control is exercised. The profiles of natural toxicants
vary between different kinds of plants, and levels vary among
different cultivars of a variety of plant (e.g., glycoalkaloids and
potato variety; MAFF, 1996, p.13). Humans make the
selections, and these selections determine exposure. In addition,
levels of many self-defense chemistries will increase with insect
and fungal stress, and humans determine how this crop stress is
to be managed (Mattsson, 2000). Consequently, all of these
natural-toxicant features are to a greater or lesser degree under
human control.
Many think that using synthetic pesticides will simply add to
the burden already imposed by natural pesticides. This
perspective is partly true and partly false. Plant self-defense
systems are both constituitive (at relatively fixed levels), or can
be induced in response to crop damage (Hammerschmidt and
Schultz, 1996; Karban and Baldwin, 1997). Pesticide residues
will add to a plant's constituitive defenses, but if the pesticides
are effective in reducing crop damage, they will reduce the selfdefense chemistries that are induced by damage. In addition,
protected crops are less vulnerable to fungal invasion and
mycotoxin contamination.
The biological complexity of plants leads to an important
question: Is it safer to eat an insect and fungal damaged plant
that has higher levels of stress-induced self-defense toxicants
and higher levels of fungal toxins, or to eat a plant that was
protected with pesticides? Similarly, is it safer to eat a plant that
was selected for natural, high levels of insect and fungal
resistance to avoid synthetic pesticide residues, or to eat a plant
with lower levels of natural defense that is augmented (when
necessary) by synthetic pesticides. At this time, there is little
specific data to guide these decisions. These trade-offs are riskversus-risk paradigms that are virtually unexplored but should
comfortably fit within the expertise of the mixtures toxicologist
(see Hicks et al., 2000, for an example with mycotoxins and
fungicides).

There is limited protection from co-evolution

A comforting idea is that natural toxicants and humans have
co-existed since the beginning of time, so humans will have
developed resistance to natural toxicants. While partly true, it
takes only a few moments to recognize that human tolerance to
natural toxicants of plants is limited. For example, humans are
at an evolutionary disadvantage to most plants. The human
reproductive cycle is long compared to most plants, giving most
plants an evolutionary advantage. Insects are in a far better
evolutionary situation than humans, and yet plants are quite
effective in keeping up with evolutionary resistance of insects
and fungi. If a plant's rate of evolving natural pesticides could
not keep up with the evolving defenses of insects and fungi, this
plant would cease to exist. It is an evolutionary war that does not
offer a simplistic answer for human safety.
In addition, plants have ecological and geographic distributions. If humans did not evolve in those locations, then there
was no opportunity for co-evolution. An example is the
solanaceous glycoalkaloids. Krasowski et al. (1997) speculate
that atypical genes for butyrylcholinesterase are especially high
in geographic areas where foods that contain high levels of
glycoalkaloids may have been consumed for thousands of years
(South and Central America, Europe, some areas of the Middle
East). These atypical genes occur at a very low rate in Africa
and Asia where there was less evolutionary pressure from
glycoalkaloids.
Of all plants, humans can eat only a few. Survival specialists
recommend you should eat only those plants or fruits that you
recognize as safe (U.S. Army Survival Manual FM 21-76). To
do otherwise is to put your life at risk. This simple observation
points to the main evolutionary capability of animals; the ability
to learn which foods are safe or are safe under certain conditions
(Pfister, 1999). Learning is far faster that co-evolution, but
requires an ability to make an association between ingesting a
food and a subsequent illness. Cognitive associations with food
that promptly make you sick are readily remembered and can
cause strong avoidance behaviors (even if the food itself was not
the cause of the illness). It would be very difficult to learn an
association to foods that cause toxicity only after chronic
ingestion or when illness occurs a long time after eating the
food.
Today, agriculturalists may select new plants for enhanced
pest resistance, or may use traditional methods to modify plants
for increased pest resistance. There is no systematic evaluation
of the safety of these traditionally-modified foods (Fenwick et
al., 1990), and there clearly is no evolutionary opportunity to
adjust to quantitative or qualitative changes in pest resistance
that are being made today.
Natural and synthetic toxicology are the same
Some people argue that the human body somehow
differentiates between natural and synthetic chemicals. There
are, of course, many natural and synthetic molecules with which
humans have minimal or no evolutionary experience. If this lack
of evolutionary experience meant humans had no detoxification

J.L. Mattsson / Toxicology and Applied Pharmacology 223 (2007) 125132

defense against new to humans molecules, it is hard to imagine

how the species has survived and has migrated world-wide.
Ames and Gold (1997) and Gold et al. (2002) have
addressed this natural versus synthetic issue as a common
misconception. A generic approach to detoxification is
probably essential to survival, as a need for a specific
detoxification mechanism for each and every molecule in
nature is not imaginable. Toxicologists would be hard pressed to
identify molecules, with the exception perhaps of the very
smallest, that are recognized and detoxified as whole
molecules. The generic nature of detoxification is the key to
modern toxicology testing of synthetic chemistries. Systematic
toxicology screening can provide a high degree of assurance
that, even if humans have never before been exposed to a
specific molecule, that this novel molecule can be reasonably
detoxified by those mechanisms that evolved against natural
toxicants.
Mechanisms of toxicity appear to be the same for natural and
synthetic carcinogens. Overall, the basic mechanism involved
in the entire process of carcinogenesisfrom exposure of the
organism to expression of tumorsare qualitatively similar, if
not identical, for synthetic and naturally occurring carcinogens
(NRC, 1996, p. 9.).
Summary and conclusions
Regulators and toxicologists are in an awkward ethical
position in regards to hazard and risk assessment of natural
toxicants. While we have the scientific tools, their application
for natural toxicants is conspicuous by its absence. As stated by
Speijers (1995), of most inherent plant toxins at best only
limited data are available, which makes it impossible to perform
an accurate safety evaluation. This limited knowledge of
inherent plant toxins permits the mystical claim of safety on
the basis of history of food use, and thus the development of
specific food safety regulation has been postponed. This
deficiency in safety evaluation of natural toxicants is greatly
compounded by an absence of any realistic attempt to
understand the role of natural toxicants as a significant
component of everyone's daily exposure to chemicals, natural
and synthetic.
There appear to be no toxicologically meaningful differences
between so-called natural and synthetic chemicals. In the body,
chemicals are chemicals, and risk of harm is driven by potency
and exposure. The natural:synthetic distinction may serve
political and funding objectives, but availability of funding
appears to have grossly distorted the balance of toxicological
research toward synthetic chemicals. Using synthetic pesticide
residues in foods as a point of reference, it is apparent that there
are far higher levels of natural toxicants in food than of synthetic
pesticide residues. Ames (1983) estimated that dietary
intake of natures pesticides is likely to be several grams per
dayprobably at least 10,000 times higher than the dietary
intake of man-made pesticides (p. 1258). Beier and Nigg
(2001) estimated that people might consume 60 ug per day of
synthetic pesticides and herbicides, while natural pesticide
consumption is about 4.8 to 6 million ug/day. These numbers

131

result in a natural pesticide potential of 80,000 to 100,000 times

that of synthetic chemicals in the diet (p. 132).
Although the exposure estimates of Ames (1983) and Beier
and Nigg (2001) contrast natural and synthetic pesticides, the
important point is not limited to pesticides. The point is that
people are exposed to large amounts of natural toxicants in their
daily lives, and the risk implications of these exposures are
largely ignored in contemporary toxicology. Essers et al. (1998)
recommended that in light of the small safety margins for
many inherent toxicants in plant food, there is a need for more
accurate risk assessments of these substances in food. There is
very little data on these substances (p. 170). Essers et al. make
several recommendations on how to address food safety, and
propose that priorities should driven by the severity of the health
risk involved, the degree of human exposure, and the feasibility
of conducting a useful study.
While in general agreement with the proposals of Essers et
al. (1998), I would propose that we need to take one step further
back. As stated earlier in the paper, humans are exposed to a
mixture of natural and synthetic toxicants at the same time.
Because of scarce resources, priorities should be determined
from our best estimates of risk to people from all chemical
sources. In this manner, there would be a realistic balance of
studies on both natural and synthetic chemistries driven by
sound scientific principles. Breaking down the artificial
concepts of natural and synthetic will foster opportunities to
discover interactions between natural:natural, natural:synthetic,
and synthetic:synthetic, again based upon sound scientific
principles. Only a realistic view of human exposure can lead to
realistic risk assessments and risk management.
References
Ames, B., 1983. Dietary carcinogens and Anticarcinogens. Science 221,
12561264.
Ames, B., 2003. An enthusiasm for metabolism. J. Biol. Chem. 278 (7),
43694380.
Ames, B., Gold, L.S., 1997. Environmental pollution, pesticides, and the
prevention of cancer: misconceptions. FASEB 11, 10411052.
ANZFA, 2001. Australia New Zealand Food Authority. Pyrrolizidine Alkaloids
in Food. Available at http://www.anzfa.gov.au.
Beier, R.C., Nigg, H.N., 2001. Toxicology of naturally occurring chemicals in
food. In: Hui, Y.H., Smith, R.A., Spoerke Jr., D.G. (Eds.), Foodborne
Disease Handbook, Second Edition, vol. 3. Plant Toxicants, Marcel Dekker,
New York, pp. 37185.
Cheeke, P.R., 1989. Toxicants of Plant Origin. CRC Press, Boca Raton.
CFSAN, 2005. Center for Food Safety and Applied Nutrition, U.S. Food and
Drug Administration. Exploratory Data on Acrylamide in Food (2002June
2005).
Coulombe Jr., R.A., 2001. Natural toxins and chemopreventives in plants. In:
Helferich, W., Winter, C.K. (Eds.), Food Toxicology. CRC Press, Boca
Raton, pp. 137161.
Daughton, C.G., 2005. Overlooked in Fallon? Environ. Health Perspect. 113 (4),
A224.
Diawara, M.M., Trumble, J.T., 1997. Linear furanocoumarins. In: D'Mello, J.P.
F. (Ed.), Handbook of Plant and Fungal Toxicants. CRC Press, Boca Raton,
pp. 175189.
D'Mello, J.P.F. (Ed.), 1997. Handbook of Plant and Fungal Toxicants. CRC
Press, Boca Raton.
D'Mello, J.P.F., Duffus, C.M., Duffus, J.H. (Eds.), 1991. Toxic Substances in
Crop Plants. The Royal Soc. Chem., Cambridge.
D'Mello, J.P.F., Porter, J.K., Macdonald, A.M.C., Placinta, C.M., 1997.

132

J.L. Mattsson / Toxicology and Applied Pharmacology 223 (2007) 125132

Fusarium mycotoxins. In: D'Mello, J.P.F. (Ed.), Handbook of Plant and

Fungal Toxicants. CRC Press, Boca Raton, pp. 287301.
Doerr, J.A., 2003. Mycotoxin interactions in livestock. http://animal.cals.
arizona.edu/swnmc/papers/2003/doerr.pdf.
Dybing, E., Sanner, T., 2003. Risk assessment of acrylamide in foods. Toxicol.
Sci. 75, 715.
Essers, A.J.A., Alink, G.M., Speijers, G.J.A., Alexander, J., Bouwmeister, P-H.,
van den Brandt, P.A., Ciere, S., Gry, J., Herrman, J., Kuiper, H.A., Mortby,
E., Renwick, A.G., Shrimpton, D.H., Vainio, H., Vittozzi, L., Koeman, J.H.,
1998. Food plant toxicants and safety. Risk assessment and regulation of
inherent toxicants in plant foods. Environ. Toxicol. Pharmacol. 5, 155172.
FAO/WHO, 2005. Joint FAO/WHO Expert Committee on Food Additives.
Sixty-fourth meeting, Rome, Feb. 817.
Fenwick, G.R., Johnson, I.T., Hedley, C.L., 1990. Toxicity of disease-resistant
plant strains. Trends Food Sci. Technol. 2325 (July).
Friedman, M., Henika, P.R., Mackey, B.E., 2003. Effect of feeding solanidine,
solasodine and tomatidine to non-pregnant and pregnant mice. Food Chem.
Toxicol. 41, 6171.
Gaffield, W., Keeler, R.F., 1996. Induction of terata in hamsters by solanidane
alkaloids derived from Solanum tuberosum. Chem. Res. Toxicol. 9,
426433.
Ger, L.P, Chiang, A.A., Lai, R.S., Chen, S.M., Tseng, C.J., 1997. Association of
Sauropus androgynus and bronchiolitis obliterans syndrome: a hospitalbased case-control study. Am. J. Epidemiol. 145 (9), 842849.
Gold, L.S., Slone, T.H., Manley, N.B., Ames, B.N., 2002. Misconceptions
About the Causes of Cancer. Vancouver, Canada: Fraser Institute (2002).
http://www.fraserinstitute.ca/shared/readmore.asp?sNav=pb&id=D477
Grodstein, F., Mayeux, R., Stampfer, M.J., 2000. (Editorial). Tofu and cognitive
function: food for thought. J. Am. Coll. Nutr. 19 (2), 207209.
Guo, C., Wilkens, L.R., Maskarinec, G., Murphy, S., 2000. Letter to the Editor:
examining associations of brain aging with midlife tofu consumption. J. Am.
Coll. Nutr. 19 (4), 467468.
Hammerschmidt, R., Schultz, J.C., 1996. Multiple defenses and signals in plant
defense against pathogens and herbivores. In: Romeo, et al. (Ed.),
Phytochemical Diversity and Redundancy in Ecological Interactions.
Plenum Press, New York, pp. 121154.
Hammond, B.G., Campbell, K.W., Pilcher, C.D., Degooyer, T.A., Robinson, A.E.,
McMillen, B.L., Spangler, S.M., Riordan, S.G., Rice, L.G., Richard, J.L.,
2004. Lower fumonisin mycotoxin levels in the grain of Bt corn grown in the
United States in 20002002. J. Agric. Food Chem. 52 (5), 13901397.
He, K., Iyer, K.R., Hayes, R.N., Sinz, M.W., Woolf, T.F., Hollenberg, P.F., 1998.
Inactivation of cytochrome P450 by bergamottin, a component of grapefruit
juice. Chem. Res. Toxicol. 11, 252259.
Hicks, L.R., Brown, D.R., Storch, R.H., Bushway, R.J., 2000. Need to
determine the relative developmental risks of Fusarium mycotoxin
deoxynivalenol (DON) and Benomyl (BEN) in wheat. Hum. Ecol. Risk
Assess. 6 (2), 341354.
Hopkins, J., 1995. The glycoalkaloids: naturally of interest (but a hot potato).
Food Chem. Toxicol. 33 (4), 323328.
Hsiue, T.R., Guo, Y.L., Chen, K.W., Chen, C.W., Lee, C.H., Chang, H.Y., 1998.
Doseresponse relationship and irreversible obstructive ventilatory defect in
patients with consumption of Sauropus androgynus. Chest 113 (1), 7176.
Hu, Z., Yang, X., Ho, P.C., Chan, S.Y., Heng, P.W., Chan, E., Duan, W., Koh, H.
L., Zhou, S., 2005. Herbdrug interactions: a literature review. Drugs 65 (9),
12391282.
Hui, Y.H., Smith, R.A., Spoerke Jr., D.G., 2001. Foodborne Disease Handbook,
Second Edition. Plant Toxicants, vol. 3. Marcel Dekker, New York.
Kaneko, T., Ishigatsubo, Y., 2005. Isoniazid and food interactions: fish, cheese,
and wine. Intern. Med. 44 (11), 11201121.
Karban, R., Baldwin, I.T., 1997. Induced Responses to Herbivory. The
University of Chicago Press, Chicago.
Krasowski, M.D., McGehee, D.S., Moss, J.M., 1997. Natural inhibitors of
cholinesterases: implications for adverse drug reactions. Can. J. Anesth. 44
(5), 525534.
MAFF, 1996. Glycoalkaloids. Inherent Natural Toxicants in Food. Ministry of
Agriculture, Fisheries and Food. Food Surveillance Paper No. 51. The
Stationary Office, London.
Marasas, W.F., Riley, R.T., Hendricks, K.A., Stevens, V.L., Sadler, T.W.,

Gelineau-van Waes, J., Missmer, S.A., Cabrera, J., Torres, O., Gelderblom,
W.C., Allegood, J., Martinez, C., Maddox, J., Miller, J.D., Starr, L., Sullards,
M.C., Roman, A.V., Voss, K.A., Wang, E., Merrill Jr., A.H., 2004.
Fumonisins disrupt sphingolipid metabolism, folate transport, and neural
tube development in embryo culture and in vivo: a potential risk factor for
human neural tube defects among populations consuming fumonisincontaminated maze. J. Nutr. 134, 711716.
Mattsson, J.L., 1996. Lets end the double standard for natural versus
manufactured chemicals. J. Occup. Health 38, 9496.
Mattsson, J.L., 2000. Do pesticides reduce our total exposure to food borne
toxicants? Neurotoxicology 21 (12), 195202.
Mensinga, T.T., Sips, A.J.A.M., Rompelberg, C.J.M., van Twillert, K.,
Muelenbelt, J., van den Top, H.J., van Egmond, H.P., 2005. Potato
glycoalkaloids and adverse effects in humans: an ascending dose study.
Regul. Toxicol. Pharmacol. 41 (1), 6672.
Missmer, S.A., Suarez, L., Felkner, M., Wang, E., Merrill Jr., A.H., Rothman,
K.J., Hendricks, K.A., 2006. Exposure to fumonisins and the occurrence of
neural tube defects along the Texas-Mexico border. Environ. Health
Perspect. 114 (2), 237241.
Morris, S.C., Lee, T.H., 1984. The toxicity and teratogenicity of Solanaceae
glycoalkaloids, particularly those of the potato (Solanum tuberosum): a
review. Food Technol. Aust. 36 (3), 118124.
Murray, M., 2006. Altered CYP expression and function in response to dietary
factors: potential roles in disease pathogenesis. Curr. Drug Metab. 7 (1),
6781.
NAS, 1973. Toxicants Occurring Naturally in Foods, 2nd ed. National Academy
of Sciences, Washington, DC.
NIAAA, 1995. National Institute on Alcohol Abuse and Alcoholism No. 27 PH
355. http://pubs.niaaa.nih.gov/publications/aa27.htm.
Nordby, K-C., Halstensen, A.S., Elen, O., Clasen, P-E., Langseth, W.,
Kristensen, P., Eduard, W., 2004. Trichothecene mycotoxins and their
determinants in settled dust related to grain production. Ann. Agric. Environ.
Med. 11, 7583.
NRC, 1996. Carcinogens and Anticarcinogens in the Human Diet. National
Research Council. National Academy Press, Washington, DC.
Paustenbach, D., Galbraith, D., 2006. Biomonitoring: is body burden relevant to
public health? Regul. Toxicol. Pharmacol. 44 (3), 249261.
Patel, B., Schutte, R., Sporns, P., Doyle, J., Jewel, L., Fedorak, R.N., 2002.
Potato glycoalkaloids adversely affect intestinal permeability and aggravate
inflammatory bowel disease. Inflamm. Bowel Dis. 8 (5), 340346.
Percival, G.C., Dixon, G.R., 1997. Glycoalkaloids. In: D'Mello, J.P.F. (Ed.),
Handbook of Plant and Fungal Toxicants. CRC Press, Boca Raton, pp. 1935.
Pfister, J.A., 1999. Behavioral strategies for coping with poisonous plants. In:
Launchbaugh, K.L., Sanders, K.D., Mosley, J.C. (Eds.), Grazing Behavior in
Livestock and Wildlife. Idaho Forest, Wildlife and Range Exp. Sta. Bull.
#70, pp. 4559.
PRC, 2004. Annual Report of the Pesticides Residue Committee, http://www.
pesticides.gov.uk/uploadedfiles/Web_Assets/PRC/PRCannualreport2004.pdf.
Shepard, G.S., Fabiani, A., Stockenstrom, S., Mshicileli, N., Sewram, V., 2003.
Quantitation of ochratoxin A in South African wines. J. Agric. Food Chem.
51 (4), 11021106.
Speijers, G.J.A., 1995. Toxicological data needed for safety evaluation and
regulation on inherent plant toxins. Nat. Toxins 3, 222226.
Speijers, G.J., Speijers, M.H., 2004. Combined toxic effects of mycotoxins.
Toxicol. Lett. 153 (1), 9198.
U.K. Food Standards Agency, 2003. http://www.food.gov.uk/news/newsarchive/
maize.
USDA, 2005. Pesticide Data Program. Annual Summary Calendar Year 2003.
United States Department of Agriculture. http://www.ams.usda.gov/science/
pdp/Summary2003.pdf.
White, L.R., Petrovitch, H., Ross, G.W., Masaki, K., Hardman, J., Nelson, J.,
Davis, D., Markesbery, W., 2000a. Brain aging and midlife tofu
consumption. J. Am. Coll. Nutr. 19 (2), 242255.
White, L.R., Petrovitch, H., Ross, G.W., Masaki, K., Hardman, J., Nelson, J.,
Davis, D., Markesbery, W., 2000b. Response to Letter to the Editor. J. Am.
Coll. Nutr. 19 (4), 467468.
WHO, 2002. The WHO Recommended Classification of Pesticides by Hazard
and Guidelines to Classification 20002002. WHO/PCS/01.5.