K
K-562
The human erythroleukemia cell line. It is highly sensitive to human and non-human primate NK-mediated
lysis.
Cytotoxicity Assays
Keyhole Limpet Hemocyanin (KLH)
Keyhole-Limpet Hemocyanin, a T-dependent protein
antigen derived from the hemocyanin of the mollusk
M. crenulata. One of the primarily T-dependent antigens used in ELISA studies.
Plaque Versus ELISA Assays. Evaluation of Humoral Immune Responses to T-Dependent Antigens
Immunoassays
3
3
3
Kanechlor
Polychlorinated Biphenyls (PCBs) and the Immune
System
Klebsiella, Infection and Immunity
3
Helen V Ratajczak
The major epidermal cells, which undergo a program
of terminal differentiation to the production of the stratum corneum. They act as signal transducers, converting nonspecific exogenous stimuli into the production
of cytokines, adhesion molecules and other inflammatory mediators.
Skin, Contribution to Immunity
Delayed-Type Hypersensitivity
3
3
Keratins
Keratins belong to the superfamily of intermediate filament proteins. They are the most abundant proteins in
epithelial cells and are known to be responsible for the
formation of cytoskeletal filaments by copolymerization. More than 20 type I keratins and about 15 type II
keratins are described. At least one of the type I and
one of the type II keratins are expressed by every
epithelial cell. For post-mitotic cornifying cells in the
epidermis, for example, a characteristic coexpression
of K1 and K10 keratin is observed.
Three-Dimensional Human Skin/Epidermal Models
and Organotypic Human and Murine Skin Explant
Systems
Boehringer Ingelheim Pharmaceuticals
900 Ridgebury Road
Ridgefield, CT 06877
USA
Synonyms
Encapsulatus
Definition
Klebsiella, also called Encapsulatus, is a genus of bacteria, of the tribe Escherichieae and family Enterobacteriaceae, containing several species causing infections primarily of the respiratory tract in man and
some of the lower animals (1).
Klebsiella is among the enteric bacilli included in the
coliform group, characterized as fermentative Gramnegative rods that inhabit the intestinal tract and nasopharynx of man and other animals without causing
disease. However, when the organisms get outside
these sites they cause serious disease. The Center for
Disease Control (CDC) in Atlanta lists the percentage
of endemic hospital infections caused by Klebsiella at
8% and of epidemic outbreaks at 3% of all pathogens.
Klebsiella pneumoniae (Friedländer's bacillus) has
been considered a significant respiratory pathogen
since 1882. Klebsiella is a Gram-negative bacterium
related to Enterobacter (formerly Aerobacter) and Serratia organisms which cause serious pulmonary and
urinary tract infections in hospitalized patients and
3
Keratinocytes
3
366
Klebsiella, Infection and Immunity
people with underlying diseases such as alcoholism,
diabetes, and chronic lung disease. Klebsiella is second to Escherichia coli as cause of Gram-negative
bacteremia but is usually more virulent than Escherichia coli in urinary tract infections. It is important to
differentiate among the bacteria because they have
wide differences in antibiotic susceptibility and pathogenicity. Biochemical tests are used to differentiate the
organisms (2).
Characteristics
The Gram-negative, non-spore-forming rods included
in the family Enterobacteriaceae are relatively small
(2–3 by 0.4–0.6 microns). The rods occur singly or in
pairs and are non-motile, lacking flagella. The Enterobacteriaceae grow readily on ordinary media under
aerobic or anaerobic conditions. They utilize glucose
fermentatively with the formation of acid or of acid
and gas, reduce nitrates to nitrites, and give a negative
oxidase reaction. Lactose fermentation, recognized by
the formation of colored colonies on solid media containing lactose and an appropriate indicator (e.g. neutral red) delineates the coliform organisms including
Klebsiella.
The genus Klebsiella is ubiquitous in nature: In the
environment it is in surface water, sewage, soil, and on
plants. In animals it is on mucosal surfaces of mammals such as humans, horses, or swine, which they
colonize. In humans the nasopharynx and the intestinal
tract are the most common habitant sites.
The genus Klebsiella consists of five species:
K. pneumoniae (subspecies pneumoniae, ozaenae,
and rhinoscleromatis), K. oxytoca, K. terrigena,
K. planticola, K. ornithinolytica. The most medically
important is K. pneumoniae followed, to a much lesser
degree, by K. oxytoca. K. terrigena and K. planticola
were originally considered to have no clinical significance and to be restricted to water, plants, and soil.
However, recent reports do describe them as occurring
in human clinical specimens. Klebsiella pneumoniae is
the most important human pathogen of the Klebsiella group.
3
Pathogenicity Factors
K. pneumoniae produces multiple adhesins, which
help the microorganisms to adhere to host cells, a critical step in the infectious process. Some of the adhesins are fimbrial (pili) and others are non-fimbrial,
each with distinct receptor specificity. In
K. pneumoniae there are five adhesin types of which
two types (1 and 3) of pili predominate and play a role
in mediating adhesion to various epithelial cells.
Type 1 pili agglutinate guinea pig erythrocytes. The
agglutination is inhibited by mannose. This mannose
sensitive type 1 fimbriae is common in many members
of enterobacteria and plays a role in mediating adhe-
sion to the upper respiratory tract. The type 3 fimbriae
is characterized by its ability to agglutinate tannintreated erythrocytes and is designated mannose-resistant, Klebsiella-like (MR/K-HA) fimbriae. This type
of pili is made by many enteric genera and is capable
of binding to various human cells such as endothelial
cells and epithelial cells of the respiratory and urinary
tracts. Recent studies have shown K. pneumoniae can
internalize into epithelial cells. Three new types of
K. pneumonia adhesins have been reported:
* the R-plasmid-encoded non-fimbrial CF29K adhesin shown to mediate adherence to human intestinal
cell lines
* a new capsule-like extracellular adhesin that seems
to confer an aggregative pattern of adhesion to intestinal cell lines
* another fimbria-like adhesin designated KPF-28,
suggested to mediate adhesion to and colonization
of the human gut.
All members of the species produce complex acidic
polysaccharide capsules and large, moist, often very
mucoid colonies. The capsules are antiphagocytic and
are responsible for the organism’s invasive properties.
The capsules determine the pathogenicity of the bacteria and structurally form the basis for classification
into 77 capsular serotypes. The serotypes differ in their
pathogenicity and epidemiological relevance, with serotypes K1 and K2 considered especially likely to be
virulent. (3)
Lipopolysaccharides (LPS) are used to divide
K. pneumoniae into eight different serotypes (LPS,
O-antigen). The O1 serotype is the most common Oantigen found among clinical isolates. LPS is a major
factor in the ability of the bacterium to resist the host
serum bactericidal activity. LPS is able to activate
complement, with the deposition of C3b onto the
LPS molecule. However, the location of the C3b prevents the formation of the lytic membrane attack complex (see below) (3).
Two types of siderophores (high-affinity, low-molecular-weight iron chelators) are secreted by
K. pneumoniae. The siderophores compete effectively
for iron bound to host proteins. The siderophores provide the bacteria with iron, taking it from intracellular
(hemoglobin, ferritin, hemosiderin, myoglobin) and
extracellular (lactoferrin and transferrin) proteins.
Aerobactin, one of the siderophores, is considered to
have a virulence enhancing effect (3).
Preclinical Relevance
K. pneumoniae is found in the respiratory tract and
feces of 5 to 10% healthy subjects and is frequently
present as a secondary invader in the lungs of patients
with chronic pulmonary disease. It causes about 3% of
all acute bacterial pneumonias (2).
Klebsiella, Infection and Immunity
The most important predisposing factors to infection
from Klebsiella are granulocytopenia and qualitative
phagocyte defects, cellular immune dysfunction, humoral immune dysfunction, and splenectomy (3).
Relevance to Humans
Infection
Klebsiella is a pathogen causing severe pyogenic community-acquired pneumonia, which mainly affects immune-compromised people and has a high fatality rate
if untreated. Klebsiella is among the top forms of
pathogens causing infection in neonatal intensive
care units and is the second most common causative
agent of Gram-negative neonatal bacteremia.
Pneumonia caused by Klebsiella pneumoniae is
characterized by the production of thick gelatinous
sputum and a high bacterial population density in the
edema zones of the active lesions. The destructive action of the unphagocytized organism on the pulmonary
tissue interferes with antimicrobial therapy and often
results in chronic lung abscesses requiring surgical resection (2).
Klebsiella is an opportunistic pathogen which can give
rise to severe infections such as septicemia, pneumonia, urinary tract infections, and soft tissue infections.
Klebsiella species have been implicated in chronic inflammatory upper respiratory tract infections:
K. ozenae in ozena, a progressive fetid atrophy of
the nasal mucosa; and K. rhinoscleromatis in rhinoscleroma, a destructive granuloma of the nose and
pharynx. The main targets of Klebsiella are hospitalized immune-compromised hosts, particularly those
with serious granulocytopenia, with severe underlying
diseases. Klebsiella is the causative agent for 5%–7%
of all hospital-acquired infections and is among the
most important nosocomial pathogens. Klebsiella is
among the eight most important pathogens in hospitals, second only to E.coli as the most common cause
of Gram-negative sepsis. Klebsiella infections are observed in almost any body site, although infections of
the urinary and respiratory tracts predominate. Klebsiella infections are associated with reactive arthritis in
some individuals and may cause cutaneous infection (2).
3
Treatment
Kanamycin, gentamycin, the polymyxins, chloramphenicol, cephalothin, and streptomycin are commonly used in treatment. In urinary tract infections
nalidixic acid and nitrofurantoin are effective.
Although some Klebsiella strains are resistant, most
strains of Klebsiella are susceptible to cephalothin,
distinguishing this bacterium from Enterobacter and
Serratia, which produce a cephalosporinase (2).
Strains of Klebsiella have emerged which are antibiotic-resistant including strains which produce ex-
367
tended spectrum β-lactamase and a new Klebsiella
species (K. planticola and K. terrigens, respectively).
Immunity
Symptomatic K. pneumoniae infections exhibit a severe inflammatory process. Innate immune mechanisms include phagocytosis by polymorphonuclear leukocytes, deprivation of the bacteria of iron, and activation of complement. Two pathways of complement
activation have been described:
* in the classical pathway the so-called natural Klebsiella-specific antibodies react with the Klebsiella
to activate complement
* in the alternative pathway, activation of complement is achieved by antigens on the bacterial surface via the properdin system.
Both pathways lead to the activation of C3 and the
formation of C3b on the bacterial surface, mediating
phagocytosis and helping form the terminal C5b–C9
complex which lyses the bacteria. The alternative
pathway of complement activation is considered the
major innate immunity against K. pneumoniae.
Several of the pathogenicity factors described more
fully above help the bacteria avoid these innate immune mechanisms. Although the lipopolysaccharide
(LPS) present in the capsule which surrounds the bacterium activates complement, and C3b is deposited,
the location of the C3b is on the longest O-polysaccharide side, far away from the bacterial cell membrane. Therefore the formation of the lytic membrane
attack complex (C5b–C9) is prevented and bacterial
cell death does not take place. Also the bacteria secrete
low-molecular-weight iron chelators, called siderophores, that compete effectively for iron bound to
host proteins. Perhaps the most important protective
mechanism of the bacteria is their capsules which are
composed of complex acidic polysaccharides. The
capsule protects the bacteria from phagocytosis and
inhibits the activation of or uptake of complement
components.
Other non-specific innate immunity mechanisms
which protect against Klebsiella infection include nitric oxide and the T helper 1-type cytokines: tumor
necrosis factor-α (TNFα), interferon-γ, macrophage
inflammatory protein-2, lipopolysaccharide-binding
protein (LPB), CD-14, interleukin-12, γ-interferon,
and nitric oxide. In contrast, Th2-driven immune responses (e.g., IL-4 and IL-10) appear to be detrimental
to the host (4).
Several different approaches are being taken to provide protection against Klebsiella. Ribosomal immunotherapy combine ribosomes from different bacterial
strains and provides increased innate as well as specific immunity. Vaccines include lysates and proteoglycans. Newer approaches include genetic inactiva-
K
KLH ELISA
1. Asimov I, Bassett DL, Beamer PR et al. (eds) (1966)
Stedman’s Medical Dictionary, 21st ed. Williams &
Wilkins Co., Baltimore
2. Sonnenwirth AC (1973) The enteric bacilli and similar
Gram-negative bacteria. In: Davis BD, Dulbecco R, Eisen
HN, Ginsberg HS, Wood WB, McCarty M (eds) Microbiology including immunology and molecular genetics,
2nd ed. Harper & Row, New York, pp 769–771
3. Sahly H, Podschun R, Ullmann U (2000) Klebsiella
infections in the immunocompromised host. Adv Exp
Med Biol 479:237–249
4. Tsai WC, Stroeter RM, Zisman DA et al. (1997) Nitric
oxide is required for effective innate immunity against
Klebsiella pneumoniae. Infect Immun 65:1870–1875
5. Szostak MP, Hensel A, Eko FO et al. (1996) Bacterial
ghosts: non-living candidate vaccines. J Biotech 44:161–
170
6. Ratajczak HV, Thomas PT, House RV et al. (1995) Local
versus systemic immunotoxicity of isobutyl nitrite
following subchronic inhalation exposure of female
B6C3F1 mice. Fund Appl Toxicol 27:177–184
KLH ELISA
Plaque Versus ELISA Assays. Evaluation of Humoral Immune Responses to T-Dependent Antigens
3
Knock-In
A gene targeting approach in which the knockout construct used to remove a mouse gene carries a functional gene (usually not of mouse origin).
Knockout, Genetic
Transgenic Animals
Jeanine L. Bussiere . Brad Bolon
Amgen Inc.
One Amgen Center Drive
Thousand Oaks, CA 91320-1799
USA
Synonyms
Genetically engineered mouse, gene-targeted mouse,
knockout mouse, KO mouse, null mutant mouse, −/−
mouse, targeted mutant mouse, tm mouse
Definition
Gene-targeted or “knockout” animals have been created to specifically lack an endogenous gene using
molecular and cellular genetic engineering techniques
(1). Homologous recombination is employed to replace the endogenous gene in an embryonic stem
(ES) cell with engineered DNA. The DNA insertion
may be a nonsense sequence that merely interrupts the
endogenous gene, or it may contain a functional gene
that encodes a different protein (a knock-in).
3
References
Knockout, Genetic
3
Klebsiella is not a focused concern of the regulatory
environment.
An animal in which genetic code for a specific protein
has been removed.
Animal Models of Immunodeficiency
Transgenic Animals
Knockout, Genetic
3
Regulatory Environment
Knockout Animal
3
tion of bacteria, e.g. the formation of bacterial ghosts
created by expression of a cloned PhiX174 gene E
which results in lysis of the bacteria. The latter has
been shown to induce specific humoral and cellular
immune responses and to confer protective immunity (5).
For immunotoxicity studies, a host resistance model is
used in mice to monitor pulmonary host defense mechanisms (6).
3
368
Characteristics
Conventional Knockout Technology
Targeting protocols employ homologous recombination between identical flanking sequences of nucleotides on:
* a targeting construct bearing an engineered DNA
sequence
* the endogenous gene on a chromosome within ES
cells.
Individual ES cells are cultured and then challenged
with cytotoxic agents to remove those in which targeting was inaccurate. This selection is possible because
one or more chemical-sensitive elements are located
adjacent to the engineered sequence on the targeting
construct; these elements promote death only in those
ES cells in which the correct recombination event has
not occurred. Surviving ES cells are cultured and injected into blastocysts, where they are incorporated at
random into all the tissues of the developing embryo.
3
3
Knockout, Genetic
3
3
Conditional Knockout Technology
Gene targeting can be limited to specific life stages
and/or tissues by using a targeting construct that
bears a site-specific recombinase (1). The prototype
for this paradigm is the Cre/loxP system, in which
the bacterial enzyme Cre excises any DNA located
between two loxP sites (a short nucleotide sequence
that does not occur in vertebrate DNA). This procedure requires the creation of two lines of genetically
altered mice: a gene-targeted line in which the engineered sequence contains a functioning gene flanked
by two loxP sites; and a transgenic line incorporating
the Cre gene. The two parent lines of mice are normal,
but crossing them results in progeny in which the
loxP-flanked gene has been excised in cells that express the transgenic Cre protein. Gene inactivation is
limited to a single tissue by placing the Cre transgene
under the control of a tissue-specific gene promoter.
Another form of conditional knockout mice utilizes
chemically mediated inhibition of a particular gene
product at the relevant stage of life. Conditional
knockouts are especially useful for studying developmentally essential genes in adult mice where global
knockout during gestation would result in embryonic
lethality. In addition, this approach is more similar to
the clinical situation where inhibition of a gene product occurs after birth (generally in adults) and avoids
any adaptations or compensations that may occur in
the animals by knock-ing out the gene product during
embryonic development.
Transgenic Alternatives to Knockout Technology
Gene targeting techniques are quite time consuming,
with production of knockout progeny often requiring a
year or more. Other genetic engineering strategies
have been developed to reduce or delete gene function
rapidly and simply without resorting to actual removal
of the endogenous gene (1). Three such techniques
require production of transgenic mice, or animals in
which DNA has been added rather than removed.
* First, the transgenic protein may act as a
dominant negative inhibitor that overrides the activity of
the endogenous protein without interacting with it.
*
*
Second, the transgenic protein may bind and inactivate the endogenous protein.
Third, a transgenic protein with cytotoxic activity
may be used to kill specific populations of cells that
normally harbor the endogenous gene.
Physiological Impact of Genetic Knockout
The physiological properties of the targeted gene and
the replacement sequence will determine what the
functional significance of the engineered null mutation
will be to the knockout mouse. All targeting procedures disrupt the normal coding sequence of an endogenous gene, thereby preventing expression and function of its protein product (i.e. knockout). Insertion of
an engineered sequence that contains a functional gene
with constitutive activity (knock-in) may replace the
role of the deleted endogenous gene, or yield a different functional or structural alteration. Both the genetic
and phenotypic composition of knockout mice must be
defined before useful information may be gathered
regarding the mechanism, efficacy, and potential toxicity of the deleted gene product. The impact of null
mutations and knock-in replacement genes is examined either in vivo or in likely target tissues in vitro.
The consequences of genetic modification typically
are investigated in young adult mice using various
combinations of conventional anatomic, biochemical,
clinical, and molecular methods. Morphologic assessment is considered the “gold standard” for phenotypic
analysis. Specific endpoints that might be noted at the
gross or microscopic level include alterations in the
size, shape, color or location of organs, or the presence
of aberrant elements (e.g. extra organs, tumors).
Knockout and transgenic mice often are structurally
normal even if functional abnormalities are apparent,
while many engineered mice appear to lack both structural and functional defects. However, subtle phenotypes (functional and/or structural changes resulting
from the genetic engineering event) sometimes may
be unmasked using pharmacological challenges or
other physiological stressors (2,3).
3
Progeny are born as chimeras (with tissues containing both normal and gene-targeted cells), grown to
adulthood, and then bred to determine whether or
not the targeted cells are present in the gonads and
contributing to gamete production. Chimeric animals
in which germline transmission occurs are used as the
parental generation (founders) to breed homozygous
knockout animals. This knockout technology currently
is suitable only for certain strains of mice, as this species is the only one in which reliable ES cell lines have
been produced.
369
Preclinical Relevance
Knockout (and transgenic) mice are rapidly gaining
acceptance as routine tools for mechanistic research
and offer considerable promise for generating specific
models of toxicological importance. Knockout mice
have been used to assess drug specificity, to investigate mechanisms of toxicity, and to screen for mutagenic and carcinogenic activities of xenobiotics. Similarly, the impact of novel therapeutic candidates can be
estimated in knockout mice; generation of viable and
fertile animals with null mutations for a potential target protein implies that pharmacological inhibition of
the molecule in vivo will elicit no major adverse effects. Furthermore, the apparent lack of an in vivo
K
3
370
Knockout, Genetic
phenotype could be used in conjunction with substantial evidence of in vitro efficacy to support the selection of a likely no observable adverse effect level
(NOAEL) for use in preclinical pharmacology and
toxicology studies. Particular emphasis in future pharmacology and toxicology studies will be directed toward conditional knockout mice (to evaluate the impact of chemically-mediated inhibition of a particular
gene product at the relevant stage of life) and “humanized” knock-in animals (in which the endogenous
mouse gene is replaced with the homologous human
gene to examine its role in disease or drug metabolism). Humanized mice are of particular importance as
these animals can be employed to evaluate the efficacy
and toxicity of human proteins that are not pharmacologically active in normal rodents or that induce a
neutralizing antibody response that limits long-term
exposure. It is important to remember that a “humanized” mouse is still a mouse, and that any phenotype,
or absence thereof, in mice bearing a human gene
knock-in models—but is not strictly analogous to—
normal human biology (4). One particular criticism
is that “humanized” mice manufacture one or a few
human proteins of interest, but other proteins that interact with the human molecules are still of mouse
origin. The physiological effect of human-mouse protein interactions may differ slightly—or substantially
—from that of the normal human-human association.
With respect to the immune system, the physiological
functions and pathways for many genes important to
normal immune function have been investigated using
knockout (and transgenic) mice (5). Again, humanized
mice are of particular importance in modeling the
human immunologic response, as they have several
advantages over conducting immunopharmacology
and immunotoxicity studies in nonhuman primates
(the only alternative if the human protein is not active
in rodents). Rodent studies are simple, relatively inexpensive, and can include enough experimental subjects to achieve suitable statistical power. More importantly, immunotoxicity assays are well characterized in the mouse, in contrast to the nonhuman primate.
However, three caveats must be kept in mind when
using genetically engineered mice for immunotoxicity
assessment:
* the emphasis on morphologic assessment as the
usual standard for phenotypic analysis means that
the immune function of most genetically engineered
mice is poorly characterized
* conclusions reached using a standard knockout
mouse (in which the gene is missing throughout
gestation and postnatal life) may not accurately reflect disease or pharmacological interventions in
which genetic function is nullified only during
adulthood (the most likely clinical scenario)
*
most critically, the background strain on which the
null mutation is carried (mice with different genetic
backgrounds) respond very differently to immune
stimuli.
The standard background of knockout mice is a mixture of C57BL/6 (the predominant component, derived
from the blastocyst) and S129 (the major source of ES
cells for gene targeting). Further, not all S129 ES cell
lines are comparable, and knockout mice often are not
bred to achieve genetic homogeneity on a suitable genetic background. Therefore, it is critical to first assess
the immunopharmacologic response using standard
immune assays in the background strain (C57BL/6)
prior to testing knockout mice back-crossed to increase the C57BL/6 gene fraction, or to back-cross
the null mutation onto mice with a genetic background
relevant to immunotoxicity assessment (i.e. B6C3F1).
Relevance to Humans
The data generated from knockout and transgenic mice
have been used to model immune responses based on
the similarity between vertebrate immune systems.
Such comparisons are made even more relevant by
using “humanized” mice. Generally, the aim of such
studies has been to discern molecular pathways and
physiological functions, or to examine the efficacy of
immunopharmacologic manipulations. Toxicity bioassays routinely are performed in knockout animals for
some purposes (e.g. mutagenicity and carcinogenicity), but the relevance of such mice with respect to
conventional immunotoxicity testing remains to be
proven. With the increasing number of protein therapeutics on the market, these data become even more
important to demonstrate that the knockout mice are a
viable alternative to immunotoxicity testing in nonhuman primates, and are relevant to the findings seen in
humans.
Regulatory Environment
Preclinical efficacy and safety studies, especially
chronic studies, are notoriously difficult to perform
when the candidate therapeutic agent is a human protein. Due to these difficulties, regulatory agencies are
concerned with alternative means of assessing risk.
These alternatives may include testing in nonhuman
primates, homologous proteins in the appropriate animal species, or in assessing knockout or transgenic
mice. It should be understood that these are all surrogates to testing the clinical candidate in humans, and
that each of these options has its own set of caveats.
However, “humanized” knockout and transgenic mice
should provide a reasonable alternative, especially for
immunotoxicity protocols in which the mouse response is well characterized.
Kupffer Cells
KO Mouse
Knockout, Genetic
Kupffer Cells
Specialised, macrophage-like cells in the liver. Kupffer
cells phagocytise foreign particles, bacteria and old
blood cells.
Fish Immune System
3
1. Bolon B, Galbreath E, Sargent L, Weiss J (2000) Genetic
engineering and molecular technology. In: Krinke G (ed)
The Laboratory Rat. Academic Press, London, pp 603–
634
2. Bolon B, Galbreath EJ (2002) Use of genetically
engineered mice in drug discovery and development:
Wielding Occam's razor to prune the product portfolio.
Int J Toxicol 21:55–64
3. Doetschman T (1999) Interpretation of phenotype in
genetically engineered mice. Lab Anim Sci 49:137–143
4. Liggitt HD, Reddington GM (1992) Transgenic animals
in the evaluation of compound efficacy and toxicity: will
they be as useful as they are novel? Xenobiotica
22:1043–1054
5. Ryffel B (1997) Impact of knockout mice in toxicology.
Crit Rev Toxicol 27:135–154
3
References
371
K