As part of the International Knockout Mouse Consortium (IKMC), the Wellcome Trust Sanger Institute (WTSI) is generating 160 knockout mouse lines per year. WTSI also reports broad-based primary phenotyping of each mouse line as part of the International Mouse Phenotyping Consortium (IMPC). Analysis of the first 250 knockout lines from the primary phenotyping pipeline has been published recently (White et al, Cell 2013). One third of the targeted genes are chosen for likely involvement in human diseases (e.g. as shown in GWAS studies), one third are chosen from suggestions of participating scientists and the remainder are randomly selected from the IKMC embryonic stem cell pool. Knockout lines that are embryonically lethal are phenotyped as heterozygotes, whilst the rest are phenotyped as homozygotes.
Most of the knockouts lines are generated by a knock-in first approach (Skarnes et al, Nature 2011). This technique combines the advantages of a reporter-tagged allele with a conditional knockout approach (see figure).
Schematic of the ‘knockout-first’ conditional allele (Skarnes et al, Nature 2011, p. 338)
The knock-in first approach relies on the identification of a critical exon common to all transcript variants that, when deleted, creates a frame-shift mutation. The critical exon is defined by a computer-assisted procedure and flanked by loxP sites. Upstream, an FRT-flanked lacZ-loxP-neomycin cassette is inserted. The lacZ-loxP-neomycin cassette contains a splice acceptor and also codes for a poly-A tail. In its full-length form (tm1a), the construct thus generates a null allele by creating a shortened transcript that lacks the critical exon. A lacZ reporter knockout (tm1b) can be generated by exposing mice or cells to Cre recombinase. Alternatively, exposure to Flp recombinase will create a wild type allele (tm1c), which can then be used to generate conditional knockout strains by crossing the mice to a tissue-specific Cre-expressing strain.
Historically, approximately 25% of knockout lines at Sanger have been phenotyped as tm1b alleles and 75% as tm1a alleles. An internal study showed that 60% of the tm1a lines were complete knockouts, 16% were hypomorphs and 24% had residual gene activity. As a consequence, from 2015, all lines are being phenotyped in their tm1b form. More detailed information on the constructs used can be found in the targeting strategy section of the IMPC website.
How to order specific knockout mouse or ES cell lines from the IMPC
OBCD does not distribute materials, but mice and ES cells can be requested via the IMPC for a small charge from repositories in Europe or the USA. Please search for your gene of interest on the IMPC website to find out which products are available.
To uphold the principles of the 3R’s, skeletal samples that would otherwise not be used in the broad-based primary phenotype screen are analysed by the OBCD consortium. To maximise the use of samples, tissues from 16 week-old female mice (n=4) are analysed in a rapid-throughput bone phenotyping platform (Bassett et al, PLoS Genet 2012) whilst samples from 16 week-old males (n=6) are analysed in a rapid-throughput joint phenotyping platform (Waung et al, Osteoarthr Cartil 2014). Importantly, this allocation reflects the use of gold standard mouse provocation models for detailed investigation of bone loss (ovariectomy in females) and osteoarthritis (surgical destabilisation of the medial meniscus in males).
To identify new genetic determinants of bone mass and strength, and cartilage integrity, limbs from the knockout strains are analysed for a number of structural and biomechanical variables. The techniques being used are:
Bone phenotyping platform
- Digital x-ray microradiography – femur length and bone mineral content
- Micro CT – trabecular bone (BV/TV, Tb.N, Tb.Th) and cortical bone (Ct.Th, BMD)
- Biomechanical testing – three-point bend testing of femur (yield, maximum and fracture loads, stiffness, toughness) and vertebra compression testing (yield and maximum loads, stiffness)
- Statistical analysis
Joint phenotyping platform
- Joint surface imaging
- Digital x-ray microradiography – bone mineral content
- Micro CT – articular cartilage volume and thickness and subchondral bone (BV/TV, BMD)
- Statistical shape modelling – whole femur
All IKMC mice are generated in the same C57BL/6 genetic background, allowing reference data to be generated for each phenotype parameter from 50–70 sex-matched 16 week-old wild-type mice. Outlier phenotypes are called for each parameter if the mean value from n=4 (female) or n=6 (male) mice lies more than 2.0 standard deviations away from the mean of the reference range. Complex statistical approaches, together with principal component analysis, are used to ensure that significant abnormal phenotypes resulting from simultaneous but smaller variances in multiple parameters are also identified.