Decision Criteria for the Evaluation of Nontransgenic Animals 
From Transgenic Animal Research

Connie L. Bacon, D.V.M., Residue Evaluation and Planning
Division, U.S. Department of Agriculture, Food Safety and
Inspection Service, 300 12th Street, S.W.
Washington, DC  20250*

Tremendous advances have been made in livestock production in the last 
two decades.  Artificial insemination and embryo transfer are now 
routine procedures in many cattle operations.  Biotechnology promises 
even more advances in the production of livestock than artificial 
insemination or embryo transfer.  In the near future it may be possible 
to develop commercially important changes in livestock in one 
generation that would have previously taken many years to accomplish 
through traditional selective breeding practices.  Presumably the 
genetically engineered livestock of the future will provide producers 
with more efficient animals and consumers with products that were 
produced using fewer feed additives, are lower in fat and cholesterol 
and contain fewer pharmaceutical, biological and chemical residues.

In 1982, the first transgenic animals that expressed the injected 
foreign DNA were produced.  These mice, containing a rat growth hormone 
gene linked to a metallothionein promoter, are probably the best known 
transgenic animals produced to date.  Thousands of transgenic mice have 
been produced since that time, with a wide variety of transgenes and 
for many different purposes.  In 1985 the first attempts were made to 
develop transgenic livestock using the same methods employed in mice.  
The results were very disappointing and it is now recognized that the 
production of transgenic cattle, sheep, goats, and swine is far more 
difficult than the production of transgenic mice.

There appear to be four major reasons that transgenic livestock are so 
difficult to produce.  First, in murine eggs the pronuclei are readily 
visible, making microinjection a fairly simple procedure.  On the other 
hand, the cytoplasm of the larger mammals is very dense and special 
techniques must be employed in order to visualize the pronuclei.  These 
techniques often reduce the viability of the embryos.  Second, the 
recovery of suitable eggs for microinjection is more difficult.  Fewer 
eggs per donor female are obtainable in livestock.  Third, optimum in 
vitro culture conditions have yet to be established for pig, sheep, and 
cattle eggs through morula or blastocyst stages.  The fourth factor 
that makes transgenic livestock more difficult to produce than mice is 
the long generational interval.  The gestation of pigs, sheep, goats, 
and cattle range from 114 to 283 days with the onset of puberty ranging 
from six months to two years.  This can be contrasted to the mouse 
which has a gestation of 20 days and a 28-day onset to puberty.

The efficiency of producing transgenic mice has been reported to 
approach 20% of the injected ova.  The efficiency of producing 
transgenic livestock is less than 2%.  Due to this extreme inefficiency 
and the economics of housing, maintenance, and equipment required to 
produce transgenic livestock, it has been estimated that the average 
cost of producing a transgenic cow approaches $1 million.  Therefore, 
due to the large number of nontransgenic animals which result from 
transgenic animal experiments and the high cost of maintaining these 
animals, it is not surprising that researchers are interested in 
slaughtering the nontransgenic animals which are a product of these 
types of experiments.

The production of mosaic transgenic animals is often a complication for 
researchers working with transgenic mice.  Mosaics are animals which do 
not contain the transgene in all cells of the body.  This is thought to 
occur in mice fairly often because of the timing of fertilization and 
microinjection of the one-cell embryo.  It is hypothesized that the 
chromosomal DNA has already begun to replicate at the time of 
microinjection.  In the production of transgenic livestock, the 
frequency of occurrence of mosaic animals is very low when one-cell 
embryos are microinjected.  This is most likely due to the longer 
development time of the larger animals compared to mice.  The DNA is 
not replicating as quickly in the larger mammals in this early stage 
and first cleavage occurs much later than it does in mice.  

The mosaic mice which have been produced through microinjection 
techniques are uniformly mosaic in the somatic cells.  Therefore, all 
organ systems possess approximately the same percentage of cells 
containing the transgene.  

Mosaic transgenic animals can also be produced by using a viral vector 
as the means of introducing the transgene into an embryo which is 
usually at the 8-16 cell cleavage stage.  Mosaicism in this situation 
is the desired outcome of the researcher and is a technique often 
employed to study gene expression.

A variety of methods are used to test the animals which result from 
transgenic animal experiments for the presence of the transgene.  The 
two most commonly used methods are Southern Hybridization and the 
Polymerase Chain Reaction (PCR).  Both of these methods can be highly 
specific and highly sensitive under the proper conditions.  The 
sensitivity and specificity are based on the knowledge of the precise 
DNA sequence which was introduced into the embryo.

The amount of genomic DNA needed to generate a detectable hybridization 
signal using the Southern Hybridization method depends on a number of 
factors.  These include the proportion of the genome that is 
complementary to the probe, the size of the probe and its specific 
activity, and the amount of genomic DNA transferred to the filter.  
Under optimum conditions, this method is capable of detecting 0.1 pg of 
DNA complementary to the probe with an autoradiographic exposure of 
several days.  Therefore, a sequence 1000 bp in length that occurs only 
once in the genome (1 part in 3 million in mammals) can be detected on 
an overnight exposure, if 10 ug of genomic DNA is transferred to the 
filter and hybridized to a probe several hundred nucleotides in length.  
The use of PCR to amplify the gene sequence of interest allows the 
detection by Southern Hybridization of a single copy gene to be 
detected in the presence of a  1013-fold excess of irrelevant DNA.

These two methods, Southern Hybridization and PCR, can then be used to 
test for the presence of the transgene in animals.  Southern 
Hybridization and PCR are also capable of detecting mosaic animals.  
These techniques are able to detect one copy of the transgene in 10% or 
fewer of the animal's cells.  The most extreme mosaic reported to date 
is a mouse with the transgene present in 15% of its cells.  The failure 
to detect the transgene by these methods supports the conclusion that 
the foreign DNA failed to incorporate into the genome of the animal.

The criteria that may be used to determine if the animals are 
nontransgenic are: 1) failure to detect the presence of the transgene 
by Southern Hybridization, the Polymerase Chain Reaction, or other 
appropriate scientific methods, 2) absence of a measurable gene 
product, 3) absence of transgene- associated traits, and 4) a healthy 
appearance.

FSIS has concluded that these nontransgenic animals (determined by some 
or all of the above methods) can be slaughtered safely under Title 9, 
Code of Federal Regulations (CFR) Sections 309.17, Livestock Used In 
Research.  Animals exposed to a viral vector require prior approval by 
the Animal and Plant Health Inspection Service (APHIS).  Under 9 CFR 
309.17, the researcher must submit an application for slaughter to the 
Residue Evaluation and Planning Division (REPD), Science and 
Technology, Food Safety and Inspection Service. This application must 
contain data demonstrating the methods employed to differentiate 
transgenic animals from nontransgenic animals.  The application must 
also include such information as the number, age, sex, and identifying 
marks, such as tatoos or eartags, of the animals proposed for 
slaughter, and pharmaceuticals, biologics, or chemicals the animals 
were administered and the last date of administration, and the proposed 
date and establishment of slaughter.  Upon receipt of all necessary 
information, REPD will review and evaluate the application.  Provided 
that all criteria outlined in 9 CFR 309.17 are met, the animals 
described in the application will be approved for slaughter.  These 
animals are identified upon presentation to the USDA 
Inspector-In-Charge at the slaughter plant as having been involved in 
research.  They are maintained as a separate lot throughout the 
slaughter procedure.  If the animals appear normal on antemortem and 
postmortem inspection they are passed for human consumption.

In addition to the research currently being conducted with transgenic 
livestock, there is also active research aimed at producing transgenic 
poultry.  The production of transgenic poultry poses some unique 
problems for researchers, but the food safety of these birds would be 
assessed in a similar manner as transgenic livestock.  If the birds are 
shown to be nontransgenic, then they may be eligible for slaughter 
under Title 9, CFR 381.75, Poultry Used in Research.  The requirements 
of this regulation mirror those of 9 CFR 309.17.

FSIS feels that it is appropriate to slaughter these nontransgenic 
animals under 9 CFR 309.17 as livestock used in research or 9 CFR 
381.75 as poultry used in research.  On June 22, 1990, FSIS presented 
the proposed decision criteria for the slaughter of nontransgenic 
animals from transgenic animal research to USDA's Agricultural 
Biotechnology Research Advisory Committee (ABRAC).  ABRAC unanimously 
voted to endorse the process by which FSIS will evaluate and present 
for slaughter such animals produced in the course of transgenic animal 
research.

Animals which are shown to contain the transgene by Southern 
Hybridization and/or the Polymerase Chain Reaction must be evaluated 
separately.  FSIS is currently in the process of drafting a statement 
of the process which will be used to evaluate the food safety of meat, 
poultry, and meat and poultry products derived from products of 
biotechnology.  FSIS plans to present this statement to ABRAC.

*Questions concerning the contents of this paper should be directed to 
Dr. Pat Basu, Director - Technology Transfer and Coordination Staff, 
USDA, FSIS, Rm 4911 South Bldg, Washington, D.C. 20250 or phone (202) 
720-8623.

September 1990 Slightly revised December 1991