Analytical Review of Integument – Research project update June 2021
Why are we here?
We believe that every person should have the right to live life in the body that they genuinely want, even if it deviates from the standard human form. This is what freedom of form is about, and the Freedom of Form Foundation conducts research and advocacy to move the cause forward. To that end, we want to take a moment to be fully transparent with the state of the science and to start a serious discussion about what we can and cannot feasibly do in the coming years.
This analytical review of integument is about skin-level modifications like fur, scales, and feathers. Since much research already exists on transgenic delivery mechanisms, we are instead answering the question, what is it we want to deliver? Our job is to identify viable targets we can use.
For the sake of transparency, we are balancing some technical language here while staying as accessible as we can – but please feel free to reach out with questions!
Fig. 1: Summary of identified morphogenetic pathways and homologous structures. (1) Epidermis; (2) Follicle Outer Sheath / Stratum Basale; (3) Follicle Inner Sheath / Stratum Intermedium; (4) Hair Shaft; (5) Bulge Epidermal Stem Cells; (5a) Germinal Zone Epidermal Stem Cells; (6) Matrix; (7) Dermal Papilla; (8) Dermis; (9) Stratum Corneum; (10) Ramogenic Zone; (11) Feather Barb Ridge; (12) Feather Follicle Sheath; (13) Feather Sheath; (14) Pulp
How much is already possible for gene editing?
Gene therapy (i.e. gene editing in a clinical context) has already been used in humans, perhaps most notably in the FDA-approved treatment Luxturna (voretigine neparvovec) that cures a specific type of blindness. Combined with other approved gene therapies, as well as a large range of experimental gene therapies currently in clinical trials, adding single genes of up to 10 kilobases in size into patients is now well-established.
There are already very promising technologies for expanding this to larger genes or even multiple genes at once, ranging from human artificial chromosomes (HACs) to multi-stage approaches (Kazuki et al., 2021; Damme et al., 2009). Gene editing of stem cells in vitro before implantation is another exciting approach. Within this review, we will therefore not focus on gene-editing technologies, but rather focus on the genetic or mechanistic changes that are required. To use an analogy, let’s say you’re moving from one city to another. It’s important to first decide on a new place and what you’re bringing with you, before considering whether to get there by truck or several car trips! That first step of where to go and what to bring is the purpose of our current research.
Fur and hair across species
It’s thought that humans have lost our fur in part because of the deactivation of a single gene, KRT41P, that encoded a form of hair keratin. While KRT41 is functional in our closest furred cousins, chimpanzees, our version has become a pseudogene – the ghost of a once functional gene (Winter et al., 2001). We’ve found that the chimpanzee KRT41 open reading frame is 97.4% identical to human KRT41P, and using multiple sequence alignment in COBALT, we note that a single point mutation is able to restore the coding sequence (4141T>C in ENSG00000225438). If KRT41P lncRNA transcripts exist in human hair follicles, the best approach would be CRISPR to correct the original copy. If they do not, then an alternate strategy would be to introduce a new copy of the corrected gene by conventional integrating viral vectors (Milone & O’Doherty, 2018). Whether KRT41 expression is necessary and/or sufficient for fur growth, as suggested by its discoverers, remains to be seen and must be experimentally confirmed. Other technologies may be developed which could supersede these suggested methods.
Chimpanzee fur is not the end goal, though, because it’s rough and lacks the soft undercoat typical of most furred animals (Estes, 1999). To circumvent this, we must modulate hair follicle size and integrate much finer undercoat-like secondary hair follicles. Follicle size is primarily regulated by Wnt signalling (Lei et al., 2015; Zhu et al., 2013). While we could target existing follicles, there are not enough in human skin for full fur coverage (Otberg et al., 2004; Mangelsdorf et al., 2013). Since trichogenic (hair-producing) stem cells can already be cultured, we are pivoting towards using stem cells modified in vitro prior to transplantation (Lee et al., 2020).
Beyond cell transplantation, it may also be possible to derive new hair follicle stem cells from existing cells within skin. Unlike humans, most furred animals can regrow hair follicles even after their respective stem cells are destroyed, via a pathway called wound-induced hair follicle neogenesis (WIHN) (Harn et al., 2021). Since WIHN and its constituent signalling pathways (particularly Wnt and Shh signalling) are conserved across many species, and an inefficient form of it exists in humans, re-enabling WIHN in the absence of wounding is probably possible and a major arm of our work (Wang et al., 2015; Bhoopalam et al., 2020).
Scale development in reptiles
Scales share the same basic developmental machinery as hair and feathers – in fact, mammals have a reptilian common ancestor (Di-Poï & Milinkovitch, 2016). Since the basic machinery and underlying genes are the same, understanding the subtle differences in these pathways and learning from scale-like skin conditions may be enough to recreate scales in humans, and that is what we have begun to do here.
Embryonic scale development in reptiles begins with a flat bi-layered epidermis, which then develops a wavy surface with papillae comparable to those of hair or feather buds. These papillae form the asymmetric scale structures during epidermal stratification and keratinization (Alibardi 1998). Shifts in β-catenin expression (the end product of canonical Wnt signalling) and the appearance of NCAM occur while symmetric buds form asymmetric scales. During bud formation, β-catenin expression is limited to the outer epidermis, and NCAM is not yet present. β-catenin and NCAM are both expressed on the anterior end of the developing asymmetric scale at the boundary between the epidermis and the dermis. The NCAM signaler is likely responsible for most of the cell proliferation at this stage, while canonical Wnt signalling regulates β-catenin in the subpopulation of cells that express the latter highly (Wu et al. 2013).
While the connection between scale development and Wnt signalling is known, the effect of Wnt pathway perturbations on scale development is under-researched, in contrast to hair. Most recent studies have focused on the genetics of reptilian scale regeneration or evolutionary comparison with feathers, but the basic pathways for scale development are not well understood (Wu et al. 2014; Lai et al. 2018; Wu et al. 2018). Still, much can be learned from those studies – one in particular identified 180 specific genes that are thought to be exclusive to scale formation, including 24 transcription factors (Wu et al. 2018). We are thoroughly analyzing the underlying data as a starting point for additional research, which will be required to solidify enough understanding to recapitulate scales using modified human cells.
Feather production in avians
Feather development mechanisms are remarkably similar to those of scales, again coinciding with their evolutionary linkage, although the precise common ancestor is debated (Di-Poï & Milinkovitch, 2016; Yang et al., 2018 and 2020; Unwin & Martill, 2020). Subtle differences to these shared mechanisms cause feathers to grow much further outwards, form barbs, and vary in shape, colour, and pattern.
Since chickens grow both scales on their legs and feathers elsewhere on their skin, the general mechanisms between the two are driven by a joined genetic circuit. Recent research has revealed several genes that convert scales into feathers and vice versa within developing chickens, with the transcription factor SOX18 causing the most severe phenotypic change (Wu et al., 2017; Lai et al., 2018). This mechanism shows a deep link between scales and feathers that might be targetable.
On the other hand, the regulatory mechanisms that promote hair growth in mice actually inhibit feather production by feather-bearing skin cells in chickens, even with as few as 1 hair-producing mouse cell per 24 feather-producing chicken cells (Moscona and Moscona, 1965). This results in a barrier region between a feather-growing region and hair follicles. Since this was discovered in 1965 and has not been followed up, the mechanism for this inhibition is not known. We suspect that it occurs through secreted inhibitors of essential morphogenetic pathways, particularly Wnt signalling, such as via secreted Frizzled receptor-related proteins (SFRPs). However, as this is hypothetical, further experiments are necessary, starting with verification that hair-specific secreted inhibitors can block feather development. Depending on the extent of the issue, this may pose an obstacle to feather growth in human hair-bearing skin, and might first require permanent hair removal if this is in fact true.
One study identified an expansive list of genes involved in feather development, containing 193 gene names (Lowe et al., 2014). Many of these are shared with hair development in mammals (32 shared genes), and very interestingly, 3 have been shown to be positively selected during feather evolution (NOTCH1, COL3A1, and PCDH17). Notch signalling is a common and essential developmental pathway leading to downstream transcriptional regulation. The Notch1 protein is conserved between humans and birds, but differs in composition substantially (80.39% identical in chickens). Whether avian Notch1 differs substantially in function, or whether gene regulatory elements are the mechanism for positive selection, is an open question.
Tying it all together with patterning mechanisms
Even with the means to convert human hair into fur, scales, or feathers, we still need to ensure their correct distribution over the skin. While manual placement is possible, it would be far superior to allow some degree of self-organization using developmental patterning mechanisms. We must therefore understand both natural patterning and prospects for artificial patterning mechanisms.
Before fur, feathers, or scales grow on skin, a ‘blueprint’ pattern forms on the skin to instruct the skin where to produce instances of each structure. One way in which these blueprints are naturally formed is through so-called Turing patterning (Turing, 1952; Sick et al., 2006), which form regularly spaced spots of chemical signals. Strangely enough, most Turing patterns are semi-disordered, like the hair on your skin that is distributed semi-randomly. However, some are highly ordered and symmetrical, like those for the scales of a snake or the feathers of flighted birds. This symmetry is obtained by adding another signal to the existing mechanism.
In flighted birds, this occurs through a wave of ectodysplasin A (EDA), symmetrically diffusing from the dorsal midline outwards on the skin (Ho et al., 2019; Inaba et al., 2019). We expect that symmetric scales use a similar mechanism. Asymmetrically patterned animals use EDA as well, though, so symmetry-specific EDA must be studied further.
An important consideration for our purposes is that since most mechanisms discussed here are activated at the embryonic stage, the patterns that they produce tend to be very small – on the order of hundreds of micrometers (Ouyang et al., 1994; Ho et al., 2019). For comparison, the distance between adult scales or feathers would be on the order of centimeters. As a result, we need to develop Turing patterns with longer intervals. Interestingly, angelfish have stripes that are also generated by a Turing pattern activated in the adult organism, with the pattern size being on the order of centimeters (Kondo and Asai, 1995). Although there is little known about the exact mechanisms here, it does show potential patterning mechanisms can exist at the length scale we need.
One promising model of such large-scale patterning requires that distant cells can communicate through long cellular protrusions that skip over a lot of neighboring cells (Owen et al., 2020; Volkening & Sandstede, 2018). This increases the size of the patterns, since the molecules that make up the pattern are accelerated through this tube. Unfortunately, such a mechanism is also much harder to engineer, due to the much more complex cellular mechanics.
Another way to produce the patterns we need is by using fully synthetic Turing patterns. For example, patterns have been engineered using annealing nucleic acids as the reacting agents, which makes the pattern both programmable and highly orthogonal, having very little crosstalk with other mechanisms and thus allowing the construction of more complex patterns. Furthermore, a simple DNA-based system created in a gel was capable of producing patterns of a centimeter length scale, which is what we want, in just a few hours (Chen and Seelig, 2020). Note that the short pieces of DNA used here are only used as a diffusible signal, and are unrelated to the DNA that stores genes in your body. Such a mechanism opens up many possibilities of complex pattern formation, such as pattern reshaping (Chirieleison et al., 2013). However, whether this can be implemented in the extremely constrained extracellular matrix of skin is unknown, as is the effect of interfering extracellular nucleases (Baker et al., 1998), which might destroy the diffusing DNA molecules before patterns can be formed. Our work aims to address some of these considerations.
Where do we go from here?
We have begun and will continue to describe a complete set of gene targets and stem cell-based approaches for reproducing fur, scales, and feathers in consenting adult humans. At the conclusion of this project, we will identify the best candidate targets for further validation, and at this time we will plan further validation studies, such as via academic collaborations or contract research organizations.
Our work marks the first step towards achieving freedom of form via biomedical science, and we are excited for what the future may hold. While it will be difficult to achieve a one-size-fits-all approach for creating any particular total body layout, and transition is likely to be a multistep process with varying degrees of invasiveness for each individual alteration, it is our expert opinion that anthropomorphic animal forms are achievable at minimum using near-term biotechnology coupled with microsurgical approaches.
We know that the nature of our work is controversial and that there will be many rightful questions about our methods and approach. With that in mind, we invite anyone with questions to ask them on social media or reach out to us directly, and we will endeavor to answer them to the best of our ability.
PS: If there are STEM professionals out there who read this and are interested in working with us, please do not hesitate to reach out!
The Project Team
Analytical Review of Integument
Freedom of Form Foundation
501(c)(3) EIN 82-4415111