Large genomic humanisations – The final frontier?

Mouse models that contain human genes are invaluable tools for preclinical research, however the targeted introduction of large human genomic sequences remains technically challenging, and few published lines exist.  The traditional approach of randomly inserting human genomic sequence is not without shortcomings.  We review gene targeting and genome engineering techniques that have been applied to generate large (>15kb), targeted human genomic Knock-ins.  These include OzBIG combined with goGermline^tm by Ozgene that has been successfully applied to generation of such lines.

Humanise the code

Genomically humanised mice are valuable tools for pre-clinical research that can be used to model human disease and evaluate the efficacy and efficiency of therapeutic interventions (Devoy et al. 2012; Zhu et al. 2019).  Antibodies or drugs may not be cross reactive with both a murine and human target. Antisense oligonucleotide and CRISPR-based gene therapy approaches may require the presence of a human genomic target.  This necessitates the humanisation of the therapeutic target for functional and validatory studies.  For example, application of a therapeutic agent to a mouse containing a mutant humanised gene giving rise to a disease state may be used to investigate validity of a therapeutic approach when compared to the phenotype of a mouse containing the wild-type humanised gene.

Consequently, the generation of humanised mice via the introduction of human genes into the murine genome has long been a goal of molecular geneticists.  Ideally, the human gene would be correctly regulated and expressed at a physiologically appropriate level with a correct tissue-expression profile to model as closely as possible the human biological state.  It may be appropriate to preserve splicing and the generation of alternate isoforms.  The humanized gene product may need to function correctly in the correct context of the murine metabolome.  The introduction of the gene within the genome should not introduce any adverse, deleterious or confounding phenotypes.  The murine copy of the gene may need to be removed or deactivated if its expression interferes with analysis, necessitating that the human gene compensate or complement for the loss of the murine gene.  Meeting these objectives presents a significant challenge both conceptually for allele design and technically for mouse model generation.

Mice can be humanised to varying degrees ranging from the introduction of a single point mutation into a murine gene to the introduction of mega bases of human genomic DNA.  Careful consideration of experimental objectives is required to determine what extent of humanisation is required.  While not without challenge, humanisations such as point mutations, the humanisation of a single exon, or the introduction of human cDNA within a murine gene can be made using standard gene targeting or gene editing approaches, but the introduction of large human genomic regions (15 to 100s of kilobases) in defined copy number at a defined locus remains technically demanding.

The transgenic legacy

Transgenic humanised mice, with insertions of human genomic sequence at random loci and in random copy number within the murine genome, has been generated for decades.  The construction of large BAC, YAC and PAC clone libraries, such as the RPCI-11 library (Osoegawa et al. 2001) constructed for the human genome project, where each clone within the library contains hundreds of kilobases of human genomic sequence, combined with pronuclear injection or electroporation technologies for the introduction of such clones into murine zygotes or ES cells, has facilitated the generation of transgenic lines since 1993 (Jakobovits et al. 1993; Schedl et al. 1993; Strauss et al. 1993).  Whilst introduction of a human genomic clone into the genome containing sufficient genomic encoded cis regulatory sequence enables copy number dependant, detectable levels of ectopic gene expression and has permitted the generation of a large number of useful transgenic lines (Van Keuren et al. 2009), there are significant drawbacks associated with this approach.  As both the number of copies of the transgene and the transgene insertion site(s) are random, transgenic founder lines need to be screened to identify a line with a biologically appropriate level of transgene expression (Chandler et al. 2007).  Gene dosage effects caused by changes in gene copy number as well as the genomic position of the transgene (chromosomal position effects) can effect transgene expression levels (Giraldo and Montoliu, n.d.).  Gene dosage effects are known to be an important driver in many disease states and thus obtaining biologically relevant expression levels is often of importance.  The random location(s) of transgene insertion(s) can also problematic.  It is technically demanding to ascertain the location of transgene insertion in any given transgenic line, and in most instances it remains unknown, although recently targeted locus amplification (Goodwin et al. 2019) and whole genome nanopore sequencing are being employed to determine transgene insertion sites (Suzuki et al. 2020).  It is therefore often unknown what gene, essential or otherwise, promoter, enhancer, chromatin remodelling site, or other regulatory element may have been disrupted by transgene insertion.  Furthermore, the BAC, YAC and PAC constructs that can be used to develop such lines may contain hundreds of thousands of bases of genomic DNA and may introduce passenger genes, in addition to the gene desired, into the mouse genome (Lusis, Yu, and Wang 2007).  If it is desirable, or necessary, to remove the endogenous murine copy of the gene of interest a knockout line must be either developed or obtained so that the humanised transgenic allele can be placed within the context of the knockout.

These drawbacks necessitate the development of techniques that permit the targeted replacement of endogenous murine genomic sequences with their human genomic orthologue.  Such approaches allow the precise insertion of a single copy of a human gene at a define location within the genome.  Despite decades of mouse model generation there remain few examples of genomically targeted mice, particularly in the tens of kilobases to hundreds of kilobases range (Zhu et al. 2019).  It is of interest therefore to review techniques that have been applied for the generation of genomically humanised mice in the 15-100s of kb range.

Targeted approaches

An initial approach utilises recombineered bacterial artificial chromosomes (BACs) to target human genomic sequence into the genome of murine embryonic stem (ES) cells by homologous recombination.  The approach, termed VelociGene®, was initially used to generate large genomic deletions (Valenzuela et al. 2003), and subsequently applied to the generation of large genomic humanisations.  A targeting construct (BACvec) encoding the desired human genomic region to be knocked-in flanked with murine genomic sequence homologous to the targeted locus is introduced into murine ES cells and integrates into the murine genome at the targeted locus by homologous recombination (HR) at low frequency.  The BACvec contains a site-specific recombinase recognition sequence flanked drug selection cassette to permit selection of transfected clones that are screened for correct integration of the human gene by loss of allele qPCR.

This approach has been successfully applied in the generation of a number of humanised lines.  A little over 10kb the murine interleukin 33 gene (Il33) was replaced with approximately 15kb of human genomic sequence in a single targeting event to generate a model of severe inflammatory airway disease and investigate the effect of an IL-33 neutralising antibody (Allinne et al. 2019).  The colony stimulating factor 1 gene (Csf1) was humanised for the study of human myelopoiesis in vivo in a Rag2 knockout background (Rathinam et al. 2011).  Approximately 16.5kb of the murine Csf1 gene was replaced with the orthologous 17.5kb, humanising most of the coding region of the gene and 3’ UTR, leaving the gene under the regulation of murine promoter elements.  A mouse model susceptible to MERS-CoV infection to enable in vivo testing of antiviral antibodies raised against the virus was generated by the introduction of 82 kb human DPP4 genomic DNA from exons 2–26, including the 3′UTR, in place of 79-kb murine Dpp4 (Pascal et al. 2015).  A mouse model of DITRA was generated by the humanisation of the interleukin 36 genes (Il1f6, Il1f8, Il1f9) and their receptor, Il1rl2 (Hovhannisyan et al. 2020).  The humanisation of the interleukin 36 gene cluster involved the introduction of 88.9kb of human genomic DNA.

Significantly more complex models have also been generated such as the humanisation of the murine IgH immunoglobin heavy chain and IgK immunoglobin light chain loci (Macdonald et al., n.d.) and T cell receptor genes (Moore et al. 2021).  Contrary to the models described previously, these mice were largely generated by the stepwise insertion of large regions of human genomic DNA, up to 210kb per targeting event, into a locus that had previously been deleted by cre-mediated recombination.  In these approaches humanisation is achieved not by direct replacement of murine genomic sequence with corresponding human sequence in a single targeting event, but rather the sequential insertion of human genomic sequence at a locus that no longer contains the murine orthologous sequence.  In the instance of the humanised immunoglobin model 940kb of human sequence was introduced at the IgH locus in a process that required nine ES cell targeting steps and 480kb was introduced at the Igk locus in eight ES cell targeting steps.  The two lines were then crossed to generate the final fully humanised line.  Notably, targeting frequencies for were low for the steps that required insertion of large human genomic sequences, between 0.1 and 0.5%.

Targeting of human genomic sequence into the murine genomic locus has also been achieved via a multi-step process where deletion of the targeted locus is first achieved by homologous recombination with a BACvec, rather than cre-mediated recombination, followed by subsequent targeting events for the insertion of genomic DNA.  This approach was followed in the generation. of humanised C3 via the introduction of 53.5kb of human genomic (Devalaraja-Narashimha et al. 2021) and humanised Fcg receptors lines via the introduction 69kb followed by a further 109kb of human genomic DNA (Gillis et al. 2017).

Targeted approaches have not been limited to VelociGene®.  Recombinase mediated cassette exchange (RMCE) which utilises a recombinase rather than HR to direct recombination of a donor cassette with the desired genomic locus, was adapted as recombinase-mediated genomic replacement (RMGR) for the insertion of large human genomic sequences (Wallace et al. 2007).  Heterologous lox sites are introduced into the genome flanking the targeted locus by double gene targeting using HR to create a ‘landing pad’.  A donor BAC is also modified so that the same heterologous lox sites flank the human genomic region to be introduced into the genome.  Cells are transfected with both the modified BAC and a cre expression vector so that cre recombinase mediates the exchange of murine and human genomic sequences.  Being recombinase driven, the efficiency of genomic exchange is often high, however the approach is initially time-consuming with the requirement to create a landing pad within the murine genome by HR prior to performing the genomic exchange.  The approach was used to replace 85kb of the murine a globin locus with the117kb orthologous human genomic locus (Wallace et al. 2007).  The murine murine IgH immunoglobin heavy chain and IgK immunoglobin light chain loci have again also been humanised following a similar approach with sequential integration of large BAC clones of up to 313kb and efficiencies of 30-94% (Lee et al. 2014).

OzBIG combined with goGermline^TM for the generation of targeted genomically humanised mice

At the time of writing, the OzBIG approach has been successfully applied in targeting large Knock-ins in 28 ES cell lines for the generation of genetically modified mice. These include models for the study of rare genetic disorders, child cancer and the introduction of human genes with no murine orthologue such as APOL1 for the study of kidney disease (Partick, M et al. 2022). This represents a significant contribution in a field where few large genomically humanised lines exist. The largest successful targeting to date introduces a little under 120kb of human genomic sequence in a single targeting event and we continue to investigate approaches for the introduction of larger sequences. Humanisations of up to 240kb are feasible by double gene targeting. The OzBIG approach recombineers human BAC clones to generate constructs that are targeted directly into the ES cell genome by HR. Targeted ES cell clones are injected into goGermlineÔ blastocysts (Koentgen et al. 2016) to ensure efficient generation of exclusively ES cell derived mice carrying the large genomic insertion (Fig. 1)

Reference list

• Allinne, Jeanne, George Scott, Wei Keat Lim, Dylan Birchard, Jonas S. Erjefält, Caroline Sandén, Li-Hong Ben, et al. 2019. “IL-33 Blockade Affects Mediators of Persistence and Exacerbation in a Model of Chronic Airway Inflammation.” Journal of Allergy and Clinical Immunology 144 (6): 1624-1637.e10. https://doi.org/10.1016/j.jaci.2019.08.039.
• Brosh, Ran, Jon M. Laurent, Raquel Ordoñez, Emily Huang, Megan S. Hogan, Angela M. Hitchcock, Leslie A. Mitchell, et al. 2021. “A Versatile Platform for Locus-Scale Genome Rewriting and Verification.” Proceedings of the National Academy of Sciences 118 (10): e2023952118. https://doi.org/10.1073/pnas.2023952118.
• Chandler, Kelly J., Ronald L. Chandler, Eva M. Broeckelmann, Yue Hou, E. Michelle Southard-Smith, and Douglas P. Mortlock. 2007. “Relevance of BAC Transgene Copy Number in Mice: Transgene Copy Number Variation across Multiple Transgenic Lines and Correlations with Transgene Integrity and Expression.” Mammalian Genome 18 (10): 693–708. https://doi.org/10.1007/s00335-007-9056-y.
• Devalaraja-Narashimha, Kishor, Karoline Meagher, Yifan Luo, Cong Huang, Theodore Kaplan, Anantharaman Muthuswamy, Gabor Halasz, et al. 2021. “Humanized C3 Mouse: A Novel Accelerated Model of C3 Glomerulopathy.” Journal of the American Society of Nephrology 32 (1): 99–114. https://doi.org/10.1681/ASN.2020050698.
• Devoy, Anny, Rosie K. A. Bunton-Stasyshyn, Victor L. J. Tybulewicz, Andrew J. H. Smith, and Elizabeth M. C. Fisher. 2012. “Genomically Humanized Mice: Technologies and Promises.” Nature Reviews Genetics 13 (1): 14–20. https://doi.org/10.1038/nrg3116.
• Esvelt, Kevin M., Jacob C. Carlson, and David R. Liu. 2011. “A System for the Continuous Directed Evolution of Biomolecules.” Nature 472 (7344): 499–503. https://doi.org/10.1038/nature09929.
• Gillis, Caitlin M., Friederike Jönsson, David A. Mancardi, Naxin Tu, Héloïse Beutier, Nico Van Rooijen, Lynn E. Macdonald, Andrew J. Murphy, and Pierre Bruhns. 2017. “Mechanisms of Anaphylaxis in Human Low-Affinity IgG Receptor Locus Knock-in Mice.” Journal of Allergy and Clinical Immunology 139 (4): 1253-1265.e14. https://doi.org/10.1016/j.jaci.2016.06.058.
• Giraldo, Patricia, and Lluís Montoliu. n.d. “Size Matters: Use of YACs, BACs and PACs in Transgenic Animals.”
• Goodwin, Leslie O., Erik Splinter, Tiffany L. Davis, Rachel Urban, Hao He, Robert E. Braun, Elissa J. Chesler, et al. 2019. “Large-Scale Discovery of Mouse Transgenic Integration Sites Reveals Frequent Structural Variation and Insertional Mutagenesis.” Genome Research 29 (3): 494–505. https://doi.org/10.1101/gr.233866.117.
• Hovhannisyan, Zaruhi, Nengyin Liu, Sara Khalil-Aguero, Casandra Panea, Jeffrey VanValkenburgh, Ruoyu Zhang, Wei Keat Lim, et al. 2020. “Enhanced IL-36R Signaling Promotes Barrier Impairment and Inflammation in Skin and Intestine.” Science Immunology 5 (54): eaax1686. https://doi.org/10.1126/sciimmunol.aax1686.
• Jakobovits, A., A. L. Moore, L. L. Green, G. J. Vergara, C. E. Maynard-Currie, H. A. Austin, and S. Klapholz. 1993. “Germ-Line Transmission and Expression of a Human-Derived Yeast Artificial Chromosome.” Nature 362 (6417): 255–58. https://doi.org/10.1038/362255a0.
• Koentgen, Frank, Jiangwei Lin, Markella Katidou, Isabelle Chang, Mona Khan, Jacqui Watts, and Peter Mombaerts. 2016. “Exclusive Transmission of the Embryonic Stem Cell‐derived Genome through the Mouse Germline.” Genesis (New York, N.y. : 2000) 54 (6): 326–33. https://doi.org/10.1002/dvg.22938.
• Lee, E.-Chiang, Qi Liang, Hanif Ali, Luke Bayliss, Alastair Beasley, Tara Bloomfield-Gerdes, Laura Bonoli, et al. 2014. “Complete Humanization of the Mouse Immunoglobulin Loci Enables Efficient Therapeutic Antibody Discovery.” Nature Biotechnology 32 (4): 356–63. https://doi.org/10.1038/nbt.2825.
• Lusis, Aldons J., Janet Yu, and Susanna S. Wang. 2007. “The Problem of Passenger Genes in Transgenic Mice.” Arteriosclerosis, Thrombosis, and Vascular Biology 27 (10): 2100–2103. https://doi.org/10.1161/ATVBAHA.107.147918.
• Macdonald, Lynn E, Margaret Karow, Sean Stevens, Wojtek Auerbach, William T Poueymirou, Jason Yasenchak, David Frendewey, et al. n.d. “Precise and in Situ Genetic Humanization of 6 Mb of Mouse Immunoglobulin Genes.”
• Mitchell, Leslie A, Laura H McCulloch, Sudarshan Pinglay, Henri Berger, Nazario Bosco, Ran Brosh, Milica Bulajić, et al. 2021. “De Novo Assembly and Delivery to Mouse Cells of a 101 Kb Functional Human Gene.” Edited by L M Steinmetz. Genetics 218 (1): iyab038. https://doi.org/10.1093/genetics/iyab038.
• Moore, Michael J., Maggie Zhong, Johanna Hansen, Hans Gartner, Craig Grant, Mei Huang, Faith M. Harris, et al. 2021. “Humanization of T Cell–Mediated Immunity in Mice.” Science Immunology 6 (66): eabj4026. https://doi.org/10.1126/sciimmunol.abj4026.
• Osoegawa, Kazutoyo, Aaron G. Mammoser, Chenyan Wu, Eirik Frengen, Changjiang Zeng, Joseph J. Catanese, and Pieter J. De Jong. 2001. “A Bacterial Artificial Chromosome Library for Sequencing the Complete Human Genome.” Genome Research 11 (3): 483–96. https://doi.org/10.1101/gr.169601.
• Pandey, Smriti, Xin D. Gao, Nicholas A. Krasnow, Amber McElroy, Y. Allen Tao, Jordyn E. Duby, Benjamin J. Steinbeck, et al. 2024. “Efficient Site-Specific Integration of Large Genes in Mammalian Cells via Continuously Evolved Recombinases and Prime Editing.” Nature Biomedical Engineering, June. https://doi.org/10.1038/s41551-024-01227-1.
• Partick, M, Brechun, K, Kranz, H, Reiss, M, Keller, M, Clark, M.M, Gauntlett, J.C, and Askew, G.R. 2022. “Introducing OzBIG: A Robust Process for Generating Mouse Models with 20 – 240kb Targeted Replacement.” In . Vol. Poster # 59. Helsinki.
• Pascal, Kristen E., Christopher M. Coleman, Alejandro O. Mujica, Vishal Kamat, Ashok Badithe, Jeanette Fairhurst, Charleen Hunt, et al. 2015. “Pre- and Postexposure Efficacy of Fully Human Antibodies against Spike Protein in a Novel Humanized Mouse Model of MERS-CoV Infection.” Proceedings of the National Academy of Sciences 112 (28): 8738–43. https://doi.org/10.1073/pnas.1510830112.
• Rathinam, Chozhavendan, William T. Poueymirou, Jose Rojas, Andrew J. Murphy, David M. Valenzuela, George D. Yancopoulos, Anthony Rongvaux, Elizabeth E. Eynon, Markus G. Manz, and Richard A. Flavell. 2011. “Efficient Differentiation and Function of Human Macrophages in Humanized CSF-1 Mice.” Blood 118 (11): 3119–28. https://doi.org/10.1182/blood-2010-12-326926.
• Roth, Timothy B., Benjamin M. Woolston, Gregory Stephanopoulos, and David R. Liu. 2019. “Phage-Assisted Evolution of Bacillus Methanolicus Methanol Dehydrogenase 2.” ACS Synthetic Biology 8 (4): 796–806. https://doi.org/10.1021/acssynbio.8b00481.
• Schedl, Andreas, Lluís Montoliu, Gavin Kelsey, and Günther Schütz. 1993. “A Yeast Artificial Chromosome Covering the Tyrosinase Gene Confers Copy Number-Dependent Expression in Transgenic Mice.” Nature 362 (6417): 258–61. https://doi.org/10.1038/362258a0.
• Strauss, William M., Jessica Dausman, Caroline Beard, Carol Johnson, Jeanne B. Lawrence, and Rudolf Jaenisch. 1993. “Germ Line Transmission of a Yeast Artificial Chromosome Spanning the Murine Α1(I) Collagen Locus.” Science 259 (5103): 1904–7. https://doi.org/10.1126/science.8096090.
• Suzuki, Osamu, Minako Koura, Kozue Uchio-Yamada, and Mitsuho Sasaki. 2020. “Analysis of the Transgene Insertion Pattern in a Transgenic Mouse Strain Using Long-Read Sequencing.” Experimental Animals 69 (3): 279–86. https://doi.org/10.1538/expanim.19-0118.
• Valenzuela, David M., Andrew J. Murphy, David Frendewey, Nicholas W. Gale, Aris N. Economides, Wojtek Auerbach, William T. Poueymirou, et al. 2003. “High-Throughput Engineering of the Mouse Genome Coupled with High-Resolution Expression Analysis.” Nature Biotechnology 21 (6): 652–59. https://doi.org/10.1038/nbt822.
• Van Keuren, Margaret L., Galina B. Gavrilina, Wanda E. Filipiak, Michael G. Zeidler, and Thomas L. Saunders. 2009. “Generating Transgenic Mice from Bacterial Artificial Chromosomes: Transgenesis Efficiency, Integration and Expression Outcomes.” Transgenic Research 18 (5): 769–85. https://doi.org/10.1007/s11248-009-9271-2.
• Wallace, Helen A.C., Fatima Marques-Kranc, Melville Richardson, Francisco Luna-Crespo, Jackie A. Sharpe, Jim Hughes, William G. Wood, Douglas R. Higgs, and Andrew J.H. Smith. 2007. “Manipulating the Mouse Genome to Engineer Precise Functional Syntenic Replacements with Human Sequence.” Cell 128 (1): 197–209. https://doi.org/10.1016/j.cell.2006.11.044.
• Zhang, Weimin, Ilona Golynker, Ran Brosh, Alvaro Fajardo, Yinan Zhu, Aleksandra M. Wudzinska, Raquel Ordoñez, et al. 2023. “Mouse Genome Rewriting and Tailoring of Three Important Disease Loci.” Nature 623 (7986): 423–31. https://doi.org/10.1038/s41586-023-06675-4.
• Zhu, Fei, Remya R. Nair, Elizabeth M. C. Fisher, and Thomas J. Cunningham. 2019. “Humanising the Mouse Genome Piece by Piece.” Nature Communications 10 (1): 1845. https://doi.org/10.1038/s41467-019-09716-7.