Questions on CDC report  "Starlink is not the cause of Allergies"


Masaharu Kawata

(Assistant Professor, School of Science, Nagoya University, Japan

translated by Katsumi Yamada from Japanese article


Starlink has been a worldwide controversy since last year for its potential

allergenicity as pointed out by Scientific Advisory Panel of EPA, on the protein introduced into Bt corn developed by Aventis CropScience which is not approved for food use in USA and other countries.


 Starlink was detected in the corn imported by Japan and Korea and concern for its safety has mounted during this one year over the world.


The Center of Disease Control (CDC) and Food and Drug Administration (FDA) of USA have announced March they will soon test the blood from the people who said they became ill after eating corn products last fall.  FDA

has received 51 of calls of allergenic response after eating corn product such as Tacos since its announcement of the Starlink allergenicity. In addition hundreds persons claimed food companies as taken illness by eating corn products.  The number was informed to FDA.


Their study whether or not the StarLink corn had caused allergies to the people had reported on June 11, 2001.  The report is on the blood

sample test if antibody (IgE) against Cry9C protein that the genetically modified corn produces is contained in the serum from those reported severe allergic reaction.


 The report concluded that the allergic reaction of those reported was not

caused by Cry9C protein.  However there are some problems on the test derived this conclusion that has various shortcomings as stated herein therefore not acceptable from scientific view points..




(1) Outline of the test

The tests were carried out on 51 people who claimed to have suffered

allergic reaction between July 1,2000 and November 30,2000.  23 out of 51 people were rejected as not satisfying the definition of CDC allergenicity  category.   Allergy reaction is defined if the claimer suffered,

1) anaphylactic shock(including dizziness and discomfort)  in one hour after taking corn product,  2) symptoms on skin and oral cavity and pharynges (hives, anthema, itch, swelling and burning pain in throat, etc.) within 12 hours of corn product intake, and  3) symptoms in digestive tract such as flatus, loose bowels, nausea, convulsion within 12 hours of corn product intake.


The number of people corresponds to these definition was 28, among those

subjected to interview or telephone questionnaire later were 24 who were aged between 5 to 74(average 36), 13 male and 11 female. Their residences are spread over 15 states and none from the same locality.  17 people offered blood serum and the report is based on the serum analyses of these 17 samples.  The  ELISA method was employed to detect Cry9C anti-body that is developed by FDA.


 Three test groups were prepared as (A) Blood serum from 17 claimers,  (B) Blood serum from people allergic to cat, plants, peanut etc. representing

hypersensitive to usual allergen,  (C) Blood serum of 21 people preserved before Starlink corn was marketed (before 1996) and is used as negative control against group A.  Positive control was prepared from goat serum by injecting purified Cry9C protein to it.


After 3 repetitive tests, CDC concluded that the reactivity of group A

against Cry9C allergen was identical to group B and lower than group C.

The serum with positive anti-body from goat reacted violently with Cry9C

protein which was claimed as an endorsement of sensitivity of the system .


Points in question:

(1) The data obtained from 3 repetitive tests is spreading widely and values are not reproducible. The average absorbancy of the reaction plate (at 450nm) in group C is 0.098, 0.078, and 0.171.  The average for Groups A  and B also shows fluctuations in each test.  This is indicative that the test method used is unstable therefore its authenticity is questionable.


(2)The average reactivity of the 21 serum samples from group C intended as

negative control was, oddly enough, much higher than any samples in groups A and B by over 20% .  This result implies that the blood serum from the individuals before StarLink corn was commercialized is more reactive to Cry9C than those claimed to have suffered allergy as eating the corn.   CDC, after apparent brain racking, came up with an excuse that the blood serum had been freeze preserved before 1996 over a long period of time that it could be different from that of fresh blood samples. Nevertheless, the group C average absorbance level was used as the criteria to determine whether the serum has significant reactivity with Cry9C or not at a level of 2.5 times its absorbance level.  Only those surpassed this level is determined as positive reaction with Cry9C protein.

This is indisputably a mistake in selecting the negative control which should have been fresh blood samples from present Americans who have minimal possibility of taking StarLink corn products.  Freeze preserved serum showing higher reactivity than test samples should never be used as negative control.


This kind of doubtful data may be unacceptable for pear-reviewed scientific paper that goes under scrutiny of editorial board by third party.




(3) Suspicious Cry9C allergen:

The Cry9C protein used as allergen in the test by CDC and FDA was not one purified from StarLink corn but from synthesized protein substitute derived from E.coli using plasmid recombinant containing Cry9C gene as described in the report by FDA.  The purified Cry9C protein sample was supplied by Aventis CropScience Com.


According to SAP Report No.2000-06 dated Dec.1.2000, the Cry9C protein

produced by corn is different in molecular size from those produced by

original Cry9C protein in the Bacillus thuringiensis and is resistant to digestion, and possibly be glycosylated.  This phenomena is commonly seen when genes from prokaryotes such as bacteria is expressed in eukaryotic cell.  And the sugar chain is well known to enhance allergenicity of a protein.


If the Cry9C protein used in the test is produced using E.coli, it ma be bacterial protein without sugar chain which may well be non-reactive to

the serum from group A individuals.  Also, it is highly reasonable that the goat serum used as positive control is highly reactive to the antigen, because it is sensitized against purified Cry9C protein from E.coli.

In addition the CDC and FDA reports do not test reactivity of Cry9C protein with other immuno-protein like IgG and IgA.


We have found many examples of this kind of data comparison that are incomparable and make look scientific, and is the same disguise tactics used in the application for approval of Roudup Ready Soybean by Monsanto in Japan.


The results by CDC and EPA may not assure the Cry9C protein in StarLink corn has no allergenicity.