ABOVE: Cre-loxP recombination is lauded as a revolutionary tool for scientists to have precise control over gene manipulation. Modifired from © istock.com, DrAfter123

When Nat Sternberg, a molecular biologist at the Frederick Cancer Research Center, heard about bacteriophage P1, he was intrigued. Having previously worked on the head proteins of lambda (λ) phage, Sternberg had an appetite for studying phage-host relationships. P1 was largely unexplored, but Sternberg was up for the challenge.

Like λ phage, P1 infected Escherichia coli and entered the lysogenic cycle. In this stage, the bacteriophage genome integrated itself into the host cell’s genome to achieve replication without destroying the host cell. During lysogeny, the λ phage genome incorporated into the host DNA, but the P1 phage displayed a unique feature: It existed as an independent plasmid. P1 seemed to be a suitable model for studying plasmids, so Sternberg set off to construct a P1 library in a λ vector to study its genes. 

He focused on studying site-specific recombination of P1. P1 phage particles were cyclic, and researchers expected recombination to produce a circular map. However, Sternberg noted an unexpected characteristic of P1’s recombination map. It was linear.

I remember the day we saw the first autoradiograms coming off the machine. We were just overjoyed that this worked so well with high efficiency.

 —Jamey Marth, Sanford Burnham Prebys

Sternberg pondered about the underlying mechanism; he suspected that there must be a genetic hotspot for recombination located at the terminal ends of the linear genome to facilitate plasmid cyclization.1 

The genesis of Cre-loxP recombination

Consistent with his hypothesis, Sternberg’s experiments revealed a small fragment of P1 DNA at the end of the genetic map that was likely responsible for site-specific recombinase. He named this locus of crossover in P1, or a loxP recognition site. On further investigation, Sternberg performed deletion mutagenesis studies to identify another necessary component for recombination events, a P1 gene product he named Cre (an anagram for recombination).2 Cre recombined target sequences at two loxP recognition sites. 

He referred to the DNA sequences flanked by loxP sites as “floxed.” The products of Cre-mediated recombination depend on the orientation of the loxP sites. Two loxP sites oriented in the same direction will excise DNA as a circular loop, while loxP sites aligned in opposite directions will invert the DNA sequence between them. This Cre-loxP sandwich in the P1 bacteriophage laid the foundation for precise genetic manipulation.

When Sternberg moved to DuPont in 1984, a member of his lab who followed him to DuPont, Brian Sauer, continued the project. The biochemistry of the system was simple, but Sauer hoped that Cre’s prokaryotic origins would not hinder its ability to effectively work in eukaryotic cells. On testing loxP sites in yeast chromosomes, Sauer was thrilled to see that Cre recombinase readily recognized loxP sites and actively transported into the eukaryotic nucleus.3 He later showed that Cre-loxP functioned efficiently in mammalian cell lines.4   

His next question was whether this system could be implemented to target genes in the germline to obtain strains of genetically modified animals. 

The development of Cre-loxP mouse models

In the late 1980s, Nobel laureates Mario Capecchi at the University of Utah, Oliver Smithies at the University Wisconsin in Madison, and Sir Martin Evans at the University of Cambridge, pioneered methods for gene targeting in mouse embryo-derived stem (ES) cells and homologous recombination as a mechanism for manipulating genes in the mouse genome.5-7 This work demonstrated the feasibility of making specific mutations in ES cells and obtaining genetically modified knockout mice. This series of breakthrough studies catapulted transgenic mice into the spotlight and inspired researchers to follow suit.

Klaus Rajewsky, an immunologist from the University of Cologne, took a keen interest in the development of the immune system, especially that of B cells. He closely followed the work on ES cells and applied it to develop one of the earliest knockout models of the immune system by rendering mice deficient of B cells.8

However, the technology had its limitations. Rajewsky wanted to study the function of DNA polymerase β (polβ), but traditional knockouts were not a viable option. Mice needed the gene early in development, and eliminating this gene was fatal for them. 

In addition to targeting a gene of interest, geneticists needed to incorporate a selectable marker gene to find correctly recombined cells. Rajewsky used the resistance gene neomycin in his experiments, but he encountered an unexpected roadblock. “Another problem that arose in these early knockout experiments was the selection marker gene neomycin, which was still in the locus and disturbed the phenotypes,” recalled Rajewsky. “I, along with many others, thought about how this could be corrected. Then, we came across Cre recombinase systems.”9 When he read about this system, Rajewsky eagerly sought to introduce this technique into the gene targeting technologies he had already established.

It’s just an illustration of how powerful genetics is. The more you can exquisitely control the process that you're working with, the more insight you're going to get into the question that you're asking.

 —Andrew McMahon, University of Southern California

Across the world, Jamey Marth, a molecular and cellular biologist at the Biomedical Research Center, recognized the method’s potential for modeling gene function. He wanted to study genes that regulated protein glycosylation in animals for closer recapitulation of the human system. 

Marth recalled early discussions with his team members revolving around Sauer’s previous work with recombinases that cleaved DNA in a very conservative and targeted manner in yeast and mammalian cells. Despite its success in mammalian cell lines, Marth was worried that it would not translate well in an animal model. It was possible that chromatin rearrangement during development might shut down the prokaryotic activity of Cre. However, he was undeterred and designed a Cre-expressing vector with two objectives: obtain Cre expression, and make the mutation cell specific. The results of his experiments surprised him.

He and his team demonstrated that Cre-loxP recombination efficiently deleted DNA sequences in specific developing T cells of transgenic animals in 1992.10 “I remember the day we saw the first autoradiograms coming off the machine,” recalled Marth. “We were just overjoyed that this worked so well with high efficiency.”

Around the same time, Marth received a call from Rajewsky. Rajewsky wanted to collaborate after reading Marth’s paper on the successful T cell cre transgenic line. Although Rajewsky initially wanted to target the polβ gene in B cells, Marth’s T cell transgenic line was an attractive alternative.

By crossing Marth’s T cell specific cre transgene with Rajewsky’s conditional polβ allele, Marth and Rajewsky developed mice that lacked DNA polymerase β in T cells.11 “When we did this work, this kind of technology opened the way to do lots of things beyond conditional targeting,” said Rajewsky.  

Cre-loxP recombination allows scientists to excise, insert, or invert specific DNA segments with unprecedented accuracy. This works through two key components: a Cre recombinase and loxP sequence recognition sites. The Cre enzyme identifies pairs of loxP sites (arrowheads), which flank (flox) the DNA, and catalyzes reciprocal DNA recombination between the two sites to excise a small piece of DNA.

     Infographic showing the breeding schematic to generate Cre-loxP tissue-specific knockout mice.
modified from © shutterstock.com, Gaspar Gomes Costa; © ISTOCK.COM, seamartini; designed by erin lemieux

To generate a tissue specific knockout mouse, researchers breed a mouse bearing a Cre transgene under a tissue- or cell-type specific or inducible promoter (A) with a homozygous floxed mouse (B).

The offspring are heterozygous for the floxed target gene (C) and breed with the homozygous floxed mouse (B).

The resulting experimental mouse is hemizygous for Cre and homozygous for loxP (D). This is the necessary genotype required to conditionally knock out the target gene in the specific tissue.

See full infographic: WEB | PDF

An inducible Cre-loxP system

The success of Rajewsky’s and Marth’s conditional gene targeting led to an explosion of different Cre transgenic lines with expression profiles in various tissues. Researchers developed additional levels of control over gene expression as a natural extension of this founding technology. Mainly, researchers developed temporal control to circumvent previous challenges that arose from global gene deletion or early developmental Cre recombinase activity. The incorporation of drug- or interferon-responsive promoters allowed researchers to control expression of Cre recombinase. Researchers used tetracycline, type I interferon, or tamoxifen to induce promoter activation12-14

Andrew McMahon, a developmental biologist from the University of Southern California, and his postdoctoral researcher Paul Danielian, who is now a biomedical editor at General Dynamics Information Technology, refined Cre recombinase control in vivo with mice using tamoxifen. “With a founding technology like Cre-loxP, a lot of resources got built around that,” recalled McMahon.

Danielian previously studied steroid hormone regulation, and his ideas guided their subsequent experiments. The duo fused the ligand binding domain of the estrogen receptor to Cre to generate a conditional form of it knowing that it would be sequestered in the cytoplasm in the absence of a ligand. Adding a ligand activated this fusion protein and translocated it to the nucleus to activate Cre activity and induce recombination. This was the first time anybody had conditionally removed gene activity in the developing fetus of the mammalian system.15 Since then, tamoxifen-inducible Cre-loxP has been one of the most widely used inducible systems. 

“It’s just an illustration of how powerful genetics is. The more you can exquisitely control the process that you’re working with, the more insight you’re going to get into the question that you’re asking,” said McMahon.

Cre-loxP and CRISPR

The Cre-loxP system remains the gold standard method for conditional gene regulation in mice, but it can be costly and time consuming. So, over the last few years, researchers wondered whether to complement this method with another powerful gene editing technique. They found their opportunity with the discovery of clustered regularly interspaced short palindromic repeats (CRISPR) and CRISPR associated nuclease, Cas9, which together enable highly specific DNA alterations at precise locations within the genome.16

The Cre-loxP-CRISPR combination sparked researchers’ interest in investigating its potential applications. CRISPR could introduce specific mutations or genetic variations and the Cre-loxP recombination system could precisely excise or integrate the genetic elements. 

For Stefan Hans, a developmental geneticist at the Dresden University of Technology, this two-step approach offered the best of both technologies for his studies on neuron regeneration in zebrafish.17 The speed of transgenic zebrafish generation and the ability to precisely visualize the target cells were key advantages.

“While you can be pretty sure that the cells are mutant, you also need to know where your cells are to understand how they behave. So, this feature is an important aspect because the putative mutant cells are labeled, and we can easily identify and take out these cells for analysis,” said Hans. Although the combination of these techniques was demonstrated in mammalian cells a few years prior, Hans was the first to use it in zebrafish research in 2021.18

The Cre-loxP system left a lasting influence on conditional gene editing, and modern advances have taken gene function and development to new heights. “Techniques come and go with new technology, It’s just like night and day. So, when you have a technique that’s lasted 30 years with no replacement technology, I think that’s kind of remarkable,” said Marth. 


  1. Sternberg N. Demonstration and Analysis of P1 Site-specific Recombination Using λ-P1 Hybrid Phages Constructed In Vitro. Cold Spring Harb Symp Quant Biol. 1979;43:1143-1146.
  2. Sternberg N, Hamilton D. Bacteriophage P1 site-specific recombination: I. Recombination between loxP sites. J Mol Biol. 1981;150(4):467-486.
  3. Saur B. Functional expression of the cre-lox site-specific recombination system in the yeast Saccharomyces cerevisiae. Mol Cell Bio. 1987;7(6):2087-2096.
  4. Saur B, Henderson N. Site-specific DNA recombination in mammalian cells by the Cre recombinase of bacteriophage P1. Proc Natl Acad Sci U S A. 1988;85(14):5166-5170.
  5. Evans MJ, Kaufman MH. Establishment in culture of pluripotent cells from mouse embryos. Nature. 1981;292:154-156.
  6. Doetschman T, et al. Targetted correction of a mutant HPRT gene in mouse embryonic stem cells. Nature. 1987;330:576-578.
  7. Thomas KR, Capecchi MR. Site-directed mutagenesis by gene targeting in mouse embryo-derived stem cells. Cell. 1987;51:503-512.
  8. Kitamura D, et al. A B cell-deficient mouse by targeted disruption of the membrane exon of the immunoglobulin μ chain gene. Nature. 1991;350:423-426.
  9. Gu H, et al. Independent control of immunoglobulin switch recombination at individual switch regions evidenced through Cre-loxP-mediated gene targeting. Cell. 1993;73(6):1155-1164.
  10. Orban PC, et al. Tissue- and site-specific DNA recombination in transgenic mice. Proc Natl Acad Sci. 1992;89(15):6861-6865.
  11. Gu H, et al. Deletion of a DNA Polymerase β Gene Segment in T Cells Using Cell Type-Specific Gene Targeting. Science. 1994;265(5168):103-106.
  12. St-Onge L, et al. Temporal Control of the Cre Recombinase in Transgenic Mice by a Tetracycline Responsive Promoter. Nucleic Acids Research. 1996;24(19):3875-3877.
  13. Kühn R, et al. Inducible gene targeting in mice. Science.1995;269(5229):1427-1429.
  14. Metzger D, et al. Conditional site-specific recombination in mammalian cells using a ligand-dependent chimeric Cre recombinase. Proc Natl Acad Sci U S A, 1995;92(15):6991-6995.
  15. Danielian PS, et al. Modification of gene activity in mouse embryos in utero by a tamoxifen-inducible form of Cre recombinase. Cell. 1998;8(24):1323-1326.
  16. Jinek M, et al. A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity. Science. 2012;337(6096):816-821.
  17. Hans S, et al. Cre-Controlled CRISPR mutagenesis provides fast and easy conditional gene inactivation in zebrafish. Nat Commun. 2021;12:1125.