Dire Wolves Return—But Are They Truly the Originals?

Around 13,000 years ago, dire wolves (Aenocyon dirus) roamed the grasslands of North America as some of the largest and most powerful predators of their time. Today, their fossilized remains not only offer glimpses into a lost past but also illuminate possibilities for the future. Now, Dallas-based biotechnology company Colossal Biosciences is making headlines with a groundbreaking project that brings these extinct predators back to life through genetic engineering—pushing the boundaries of evolutionary biology, ethics, and biotechnology.

Although dire wolves bear a striking resemblance to gray wolves, large-scale paleogenomic analyses conducted in 2021 revealed that this similarity is misleading. Genetic data show that dire wolves diverged from other wolf-like canids around six million years ago and are more closely related to black-backed jackals and South American canids. Their evolutionary history highlights the complexity of hybridization and speciation processes. The genetic analyses were based on remarkably well-preserved remains: a 13,000-year-old tooth from Ohio and a 72,000-year-old skull from Idaho.

DNA extracted from these specimens was mapped not to classical reference genomes, but to a newly constructed iterative dire wolf genome (aenDir). This approach helped eliminate reference bias, which is a common obstacle in low-quality ancient DNA analysis, and enabled a more accurate and detailed reconstruction of the species’ phylogenetic history. The results revealed that dire wolves carry genetic signatures from both wolf-like ancestors and South American lineages.

Colossal’s approach differs significantly from classical cloning. Instead of using somatic cells, scientists extracted endothelial progenitor cells (EPCs) from gray wolf blood and edited 14 gene regions with 20 mutations to produce dire wolf-specific traits. These genetic modifications restored key features such as white fur, a broader jaw, stronger limbs, and the characteristic howl. Among the altered genes, NCAPG and LCORL were strongly linked to the large body size observed in dire wolves.

A total of 45 embryos were created and implanted into two domestic hound breeds. As a result, two male pups, Romulus and Remus, were born in October 2024, followed by a female, Khaleesi, in January 2025. Delivered via cesarean section, the pups were bottle-fed immediately after birth. Within just a few weeks, they began howling and displaying hunting behaviors—evidence that the genetic reprogramming had successfully reinstated not just the physical traits but also the behavioral instincts of the species.

Can Ancient Behavioral Codes Be Revived?

Today, Romulus, Remus, and Khaleesi live in a 2,000-acre secured wildlife reserve in the United States. This facility allows scientists to monitor their physical health, behavior, and social development. However, dire wolves are naturally pack animals with complex social structures, and a group of three individuals cannot replicate the dynamics of a natural pack. This raises the question of whether genetically resurrected individuals can truly reflect the ethology of their original species—not just in form, but in social function.

Bringing dire wolves back to life is not solely a triumph of genetic engineering. It is also the product of advanced paleogenomic techniques, ancient DNA analysis, and phylogenetic modeling—forming a multilayered scientific achievement.

Ancient DNA Extraction and Library Preparation

Researchers extracted DNA from two different fossil specimens: a 13,000-year-old tooth (DireSP) from Sheriden Pit, Ohio, and a 72,000-year-old skull (DireGB) from Gigantobison Bay, Idaho. Researchers surface-sterilized the samples using 0.5% bleach and isolated DNA with buffers optimized for short, degraded molecules typical of ancient DNA. Single-stranded DNA (ssDNA) libraries were then constructed to preserve the integrity of the highly fragmented ancient genetic material.

Sequencing, Mapping, and Reference Bias

Using the NovaSeq 6000 platform, researchers obtained 13.2 billion paired-end reads from DireSP and 78 billion from DireGB. Initial mappings to the gray wolf genome (canLor) were hindered by reference bias—where sequences preferentially align to closely related species, leading to inaccurate results.

To overcome this, the team constructed an iterative dire wolf-specific paleogenome (aenDir). This process began with mapping to the gray wolf genome, extracting dire wolf-specific sequences, and building a new reference in cycles. The aenDir genome increased mapping coverage by 2.8× for DireGB and 1.5× for DireSP while reducing reference bias by 70–80%.

Phylogenetic and Ancestral Tree Analyses

Phylogenetic analyses placed dire wolves as a distinct clade that diverged earlier than black-backed jackals and gray wolves. However, only 44% of the genomic loci supported this topology, while others suggested alternative evolutionary relationships. This discrepancy likely results from incomplete lineage sorting and past hybridization events—common in rapidly diversifying species.

The most plausible model shows that approximately 61% of the dire wolf genome shares ancestry with wolf-like species, while 39% traces back to South American canids. This hybrid origin helps explain inconsistencies in phylogenetic trees and reflects a complex evolutionary past shaped by admixture.

Genes Under Positive Selection

Eighty genes were identified as being under positive selection: 49 episodic and 31 continuous. Many of these are associated with body growth and morphology:

• NCAPG: Strongly linked to large body size; also associated with growth in horses, goats, and cattle.
• LCORL: Negatively correlated with growth and potentially co-regulated with NCAPG.

Genes related to male reproductive biology also showed signs of selection:

• ROS1, TP73, AKAP3, PDGFA, NUTM1: Involved in spermatogenesis and testicular development.

Genetic Divergence and Morphological Adaptation

Genetic comparison revealed only 0.09% sequence divergence between dire wolves and gray wolves—a surprisingly low figure possibly distorted by reference genome bias. Despite overlapping habitats, there was no evidence of gene flow between them, indicating that dire wolves were reproductively and behaviorally isolated from their close relatives.

Methodological Challenges and Limitations

Working with ancient DNA presents unique technical challenges:

• DNA damage: High rates of 5’ C>T deamination.
• Low sequencing depth: Only 6% of DireSP reads achieved >5× coverage.
• Structural variant detection: The iterative mapping method is optimized for small-scale mutations, not large genomic rearrangements.

Nevertheless, the creation of the aenDir reference genome significantly improved mapping quality and serves as a valuable foundation for future comparative genomic research.

Romulus, Remus, and Khaleesi are more than just laboratory products; they represent a pivotal convergence of evolutionary history and technological possibility. These individuals are the first real evidence that extinct species can be brought back not just genetically, but behaviorally and ecologically.

Yet with this power comes great responsibility. As we acquire the ability to reshape nature, we must also recognize the ethical, ecological, and scientific boundaries of such interventions.

Science has given voice to the long-silent fossils of the past by translating their secrets into genetic code. Now, the question before us is: how will we write the next chapter?

We got inspired by those articles to create this content

Colossal, Time, Time, Biorxiv