what-when-how
In Depth Tutorials and Information
These two constructs allowed for testing of the regula-
tory sequences within the normal COL1A1 sequence as
well as a means to quantitatively assay the expression of
the minigenes relative to the endogenous gene at both
the RNA and protein levels. 25 The shortened human pro-
αI(I) collagen chains associate with normal murine pro-
α1(I) collagen and pro-α2(I) collagen chains via disulfide
bonds, resulting in “procollagen suicide,” with degrada-
tion of both normal and shortened pro-α chains, caus-
ing a depletion of normal type I collagen and altering
extracellular matrix organization. 26 The minigene has
varying levels of expression in transgenic mice, leading
to a range of phenotypes. Mice expressing high levels of
the minigene are phenotypically similar to lethal vari-
ants of human OI, while mice expressing moderate lev-
els of transgene have fractures, reduced bone mineral
and collagen content as well as altered (brittle) matrix. 25
Mice expressing high levels of the minigene displayed
perinatal lethality as high as 90%. 26 A unique aspect of
this model is the partial downregulation of the mutant
COL1A1 using antisense oligonucleotides. 26 When mice
carrying the antisense oligonucleotide were bred to mice
carrying the minigene, the lethality rate fell to ~27%, 26
presumably due to an increase in normal type I collagen
production. Weaknesses of this model include the low
yield of transgenic mice following injection (about 10%
of injected embryos survive to sexual maturity) 26 and the
variable copy number (1-50) in transgenic animals mak-
ing genotype to phenotype correlations difficult.25 25 The
unusual hybrid nature of the human and mouse proteins
also make direct translation to human OI difficult.
type II and was the first model to have a glycine substitu-
tion in the triple helical domain. Due to the variation in
mutant construct copy number between fetuses within a
litter the interpretation of genotype-to-phenotype corre-
lations is difficult. Perinatal lethality is also a limitation.
BrtlII and BrtlIV Mice ( Col1a1 )
BrtlII and BrtlIV mice are knockin mouse models
derived using a cre-lox knockin strategy. 29 The construct
was designed to contain the G349C Col1a1 mutation, a
glycine-to-cysteine substitution at position 349 in the
triple helical domain of the α1(I) collagen chain (present
in a moderate type IV OI patient 30 ), a L359M substitution
(leucine-to-methionine substitution to allow evaluation
of the transgene protein expression and incorporation
into matrix), and a GUC ribozyme cut site for poten-
tial therapeutic studies. 29 This original construct was
designed to generate a null allele by blocking expression
of the mutant allele and thus result in a mild OI type I
phenotype. However, the placement of the floxed cas-
sette resulted in alternative splicing of the mutant allele
giving rise to a lethal phenotype. This knockin model,
referred to as the BrtlII mouse, models human OI type
II, 29 reproducing the skeletal outcome described in OI
patients with similar molecular rearrangements. 31
To derive the BrtlIV mouse, the BrtlII chimera (F1) was
mated to a cre-recombinase mouse to remove the floxed
cassette. BrtlIV /+ mice, heterozygous for the G349C sub-
stitution in the α1(I) chain, represent a knockin model for
OI type IV, 29 even though they demonstrate phenotypic
variability ranging from approximately 30% perinatal
lethality to long-term survival with moderately severe
OI. 29 BrtlIV mice exhibit autosomal dominant genetic
transmission, and model a glycine substitution mutation
with the physiologic levels and tissue distribution of OI
collagen mutations.
BrtlIV /+ mice which die a few hours after birth had
similar sizes and weights as normal littermates. 29 They
had flared ribcages, multiple rib and long bone fractures,
abnormal vertebral bodies and decreased calvarial min-
eralization. BrtlIV /+ mice which survive the perinatal
period exhibit a growth deficiency; they were ~50% of
the wild-type size until 6 weeks of age, then increased in
size to ~80% of wild-type by adulthood. 29 Throughout
development BrtlIV /+ mice exhibit disorganized trabecu-
lae, thin, abnormally mineralized cortices, reduced cross-
sectional area, cortical thickness, bending moment of
inertia and bone mineral density (BMD), and relatively
high bone turnover rates. 32,33 The upregulated bone turn-
over is believed to be due to unsynchronized osteoblasts
and osteoclasts leading to the generation of more bone-
resorbing osteoclasts. 32 Post-pubertal mice have decreased
calvarial mineralization, rib and long bone deformities
and fractures, osteoporotic vertebral bodies and features
G859C Mouse ( Col1a1 )
The G859C mice were generated using α1(I) colla-
gen DNA constructs (designed to produce a glycine-to-
cysteine substitution at position 859 in the triple helical
domain of α1(I) collagen chain) by site-directed mutagen-
esis. These constructs were microinjected into fertilized
eggs, developed in utero and resulted in embryos that
were examined within 1 hour of birth. 27 Of nine pups
that died perinatally, seven carried between 0.25 and six
copies of the transgene. In contrast, none of the surviving
pups carried the transgene. 27 Radiographs showed that
bones of transgenic pups were pliable, poorly mineral-
ized and developmentally retarded with short, wavy ribs
and short, broad long bones. 27 Additionally, increased
bone resorption was also indicated by large cavities
within the bone. 28 Phenotypic severity in this model is
proportional to mutant gene expression; mice with the
most severe phenotype express the highest levels of the
mutant gene. 27 The substitution of a glycine with a cys-
teine leads to production of disulfide-bonded αI(I):αI(I)
dimers. 28 Though this model is not based on a known
patient genotype, it is phenotypically similar to lethal OI
Search WWH ::




Custom Search