De-Extinction in Detail

De-extinction projects for animals consist of the same basic components: In Silico (computational) research of genomes, In Vitro (outside an organism) cellular engineering, In Vivo (in an organism) growth of engineered cells to a whole organism, Ex Situ (out of nature) captive management, and In Situ (in nature) wild management. For the passenger pigeon these parts are related such that we have divided the project into three phases. 


  Below you can tab through these phases for an in depth look at their individual components.

Phase 1.1In Silico Genome Research: The Band-tailed Pigeon Genome


DNA from fresh band-tailed pigeon tissue was obtained from a blood sample from “Sally”, a female band-tailed pigeon raised by professional breeder Sal Alvarez. DNA sequencers cannot sequence long pieces of DNA, so the extracted DNA molecules were sheared to different sized fragments in a controlled manner such that when they are sequenced they can be reassembled computationally by matching overlapping codes in the reads – creating long sequences of the genetic code. Short DNA fragments are not able to reassemble repetitive elements in the genome – for these hard to assemble regions longer sequence read information is needed. Through special preparation of DNA long fragment information can be retrieved even by sequencing short strands of DNA – by combining the short and long read information the original DNA sequence is pieced together – producing the reference assembly.

Bullet_BThe reference assembly of the band-tailed pigeon is ~1.1 billion base pairs of genetic code. In order to know what parts of the code are used to regulate cell functions and create traits the RNA from the organism needs to be sequenced. DNA controls traits by being transcribed (written) into RNA, which is then used by the cell to make proteins (protein synthesis) and control how little or how much protein is made (regulation). All these elements in the genome that are processed through an intermediate RNA stage are known as the transcriptome. Birds genomes contain ~17,000 encoded elements interspersed on different chromosomes. Different tissues in the body express different subsets of these 17,000 genes – therefore to discover and map them all we need to sequence RNA from multiple tissues. Once the RNA is sequenced the resulting reads can be mapped to the DNA reference assembly to identify and map the positions of genes and regulating elements.

The Final Product

Our reference genome has been assembled from DNA sequences from Sally, processed from 2012-2014. Gaps in the sequence were bridged using long fragment information made from DNA of a cell culture developed from one of Sally’s offspring in 2013. The RNA being used to map genes in the genome has been obtained from another of Sally’s progeny obtained in 2015.

Phase 1.2 – In Silico Genome Research: The Passenger Pigeon Genome

It is not possible to assemble the genome of the passenger pigeon in the same way that we can assemble overlapping fragments of the band-tailed pigeon for the following reasons:

  • The DNA has fragmented in an uncontrolled fashion, creating a mix of fragment sizes – genome assemblers must use fragment lengths of specific size ranges.
  • The DNA fragments from ancient specimens do not overlap perfectly due to degradation of the DNA code, causing mismatches in base pairs that were originally the same – these are “false mutations”.
  • The DNA is fragmented so short that long read is virtually impossible to obtain.

We overcome these issues by using the band-tailed pigeon genome to map the passenger pigeon DNA. We can then see where the fragments overlap despite mismatching base pairs. By sequencing enough fragments (to obtain what genomicists refer to as depth coverage) we can statistically rule out which mutations are false and what DNA code is real – this means we sequence enough DNA to have many DNA fragments that align to the same region of the genome. In the image above the passenger pigeon reads range from 1X coverage (a single read mapped to the reference base pair) to 4X (4 overlapping reads per reference base pair).

The Goal

When the mapping is done the computer will call the correct base pairs at each position and we take the consensus code from all the reads only at places with high enough depth coverage to be reliably accurate. This is not a perfect genome sequence of the passenger pigeon because it is mapped to the band-tailed pigeon, which may have different chromosome arrangements – but the genetic code is accurate. Birds, unlike mammals and other animals, have not developed major rearrangements of chromosomes through their evolution, so we believe that the passenger pigeons original genome is extremely close not only in code, but also in structure to the band-tailed pigeon genome.

Phase 1In Silico Genome Research: Genome Comparisons

Bullet_ASome mutations matter to de-extinction while others do not. Consider this –  your parents and siblings all have unique mutations, but you are all human. Those unique mutations make you different from one another, but are not what make you different than a chimpanzee. The mutations that make you different from a chimpanzee will be shared by all humans. Therefore unique diversity is not evolutionarily significant for discovering the traits that make a passenger pigeon (though they will be important later for developing genetic diversity in a viable population). These shared mutations are called fixed differences.

By scanning the genomes of the passenger pigeon and band-tailed pigeon side by side we can identify the fixed differences. Not all of these differences affect traits – some mutations are silent or are not coded into proteins – these build up over time by a process called genetic drift, and are not acted on by evolutionary selection. The fixed mutations that change traits will form the blueprint of the passenger pigeon.

Bullet_BWhen the candidate fixed differences are identified we will synthesize these codes through a process known as DNA printing – making the genetic code into viable DNA molecules again. 

The Goal

The end result is a library, or bank, of digital and physical genetic codes. These DNA molecules will be used to edit the genomes of band-tailed pigeons.

Phase 2.1In Vitro Revival: Primordial Germ Cell isolation, Culturing, and CRISPR/Cas9 Genome Editing

Birds have not yet been cloned, and cloning may never be possible, but genome editing in birds has been achieved with various methods. The method we have begun establishing protocols for involves the use of primordial germ cells (PGCs) – a kind of stem cell that becomes sperm and eggs.

In a developing embryo PGC’s can be isolated from the germinal crescent, a region in the early stage embryo where the cells form, or from the gonads at a later stage in embryonic development. We are currently learning how to isolate PGCs using rock pigeons as practice surrogates.

Once isolated these cells can be grown in a liquid medium of nutrients, called culture. Cultured cells will live and grow in the lab, allowing us to make many gene changes without needing to hatch birds every time we edit the genome. The right culture conditions for different bird species require a lot of development to get right. This task is currently the next component that Revive & Restore will start.

Bullet_AOnce PGCs are in culture we can use CRISPR/Cas9 technology to edit the genome. We will design CRISPR/Cas9 complexes designed to target specific regions of the genome. Once injected inside a cell CRISPR/Cas9 will bind to the band-tailed pigeon genome  at these target sites and cut the DNA.

Bullet_BThe synthesized passenger pigeon DNA from Phase 1.3 will be integrated into the genome by the cell’s own DNA repair mechanism – homologous recombination. The ends of the passenger pigeon DNA code will match the DNA at the cut sites, allowing the cell to piece the two codes together.

Bullet_CAfter genome editing a process termed “allele replacement” is complete. The band-tailed pigeon DNA has been removed and overwritten with passenger pigeon DNA. This results in a PGC culture that is now slightly passenger pigeon – by repeating the process we will eventually create PGCs that harbor newly created passenger pigeon genomes that resemble a sort of hybrid DNA code between modern band-tails and extinct passenger pigeons.

The Goal

The goal is that the hybrid genome produces a bird that not only carries the genetic legacy of an extinct species, but looks and behaves like extinct passenger pigeons.

Phase 2.3 In Vivo Revival: Germ-line Chimera generation

Bullet_AIn order to turn the engineered PGCs into a living bird we must put the cells through a process called germ-line transfer. The PGCs will be injected into the bloodstream of a developing band-tailed pigeon. The cells will migrate and colonize the gonads (ovaries or testes) of the embryo. This process has been done in chickens.

Bullet_BThe embryo develops into a germline chimera – this is not a hybrid bird. Every tissue of the bird is band-tailed pigeon, but the gonads will host a population of sperm or eggs that is made up of native band-tailed pigeon cells and a portion of the engineered cells we injected.

Bullet_CWhen a male germ-line chimera is mated to a female germ-line chimera they will have three types of offspring – pure band-tailed pigeons, band-tail-passenger pigeon hybrids, and when an engineered sperm meets an engineered egg cell the result will be a fully formed de-extinct passenger pigeon. Band-tailed pigeon offspring can be used to create more chimeras or to release to the west coast wilderness to ensure the survival of the species. If hybrid birds are fertile they can be used to breed passenger pigeon traits through back-breeding. The new passenger pigeons (birds carrying a fully edited set of alleles from each parent) will form the first true generation of de-extinct passenger pigeons.

The Goal

The overall goal of this phase is to produce a parent generation of chimeras and a first generation stock population of fully edited birds – the new passenger pigeons.

Phase 3.1Ex Situ Restoration: Ex Situ Captive Breeding

Bullet_AIn order to generate large numbers of new passenger pigeons that breed and behave naturally will require several different parent flocks of birds. The germ-line chimeras can be continually bred with controlled lighting to produce large numbers of offspring. As many as 50 offspring per year can be born from a single pair of birds. A flock of 5 breeding pairs, each carrying slightly different genetic lines of passenger pigeon, can produce nearly 200 offspring each year. Band-tailed pigeons can live over ten years and start breeding in their first year. To ensure health and welfare for the birds pairs will be taken out of cycle periodically. By rotating different sets of pairs in and out of cycle a flock of six breeding pairs (5 pairs breeding at a time) could produce almost 2,000 offspring in their lifetime. The pairs will be mixed and matched in order to generate genetic diversity between the different engineered strains.

Bullet_BThe eggs from germ-line chimeras will be transferred to surrogate parents for incubation and parenting. The goal is to produce a community of surrogate parents that breed in similar societies to passenger pigeons, so that our new passenger pigeons develop with the proper behavioral culture. Band-tailed pigeons nest in trees like passenger pigeons did, but do not nest in tight communities. Rock pigeons will nest in dense communities, but not on tree branches. Rock pigeons may be trained and raised to use nest platforms on tree branches, or band-tailed pigeons may be able to be conditioned to tolerate close proximity – it is likely a combination of surrogate parents will be used to foster the first generation of passenger pigeons. These captive bred birds will be housed in environments of natural trees and forest plant species, preventing domesticated traits from imprinting. Their food will not be supplied in dishes, but strewn about through undergrowth and foliage – stimulating natural foraging behavior.

Bullet_CThe first generation of new passenger pigeons once sexually mature will breed and raise their own offspring in natural cycles without the intervention of caretakers. The only human interactions with the birds will be health examinations with veterinary staff, otherwise monitoring will be designed to be remote. We want to ensure that birds bred in captivity are prepared for the wild.

The Goal

Eventually there will be no need for the chimeras and surrogate parents, once the captive population has reached a size that individuals can be removed for wild release without endangering the continuance of the flock we will have reached our goal of a viable captive population.

Phase 3.2In Situ Restoration: In Situ Wild Release

Bullet_AThe first step in wild release will be a “soft release” . Large flight aviaries encompassing entire trees set within forest locations will be constructed at sties within the former breeding range (spring locations) and roosting/foraging ranges (Autumn and Winter) of the passenger pigeon. These structures will allow flock managers to monitor how the birds cope with natural weather conditions. The enclosures can be designed to introduce other species to observe interactions – such as how the birds’ response to squirrels, rabbits, deer, etc.

Passenger pigeons were nomadic; to prevent the birds from acclimating to specific sites and becoming “migratory” the birds will be rotated randomly between multiple spring, summer, and winter sites. Moving the birds will be the one of the projects biggest challenges. Cranes and geese raised in captivity have been taught flight movements by airplanes. It may be possible to fly passenger pigeons using specialized drones – but a more natural solution may be achievable using homing pigeons as surrogate flocks to lead the passenger pigeons from one site to another. This strategy poses a big risk – the birds will be completely free temporarily. The birds may scatter and not follow the surrogate flock. All birds will be implanted with micro-GPS trackers to trace their movements – in this way we can locate and retrieve birds that wander, but more importantly we can observe if the birds are forming the tight social units that historic passenger pigeons did. In time it may be possible to stop using the homing pigeons to lead new batches of passenger pigeons.

Bullet_BOnce it appears that the pigeons bred and raised in soft release rotation are adapted to natural conditions and flock in the manner that is desired we can begin releasing test flocks to the wild. The first birds free in the wild will still be monitored by GPS for study. This population will likely be designated as an experimental population for research until it is self sustaining.

The Goal

Our goal is to release the first test flocks in 2032. The wild population will be stocked from soft release sites until it appears to be sustaining. This may be possible in less than twenty years with a target goal of 10,000 wild birds exhibiting self-sustained population growth.