A technical roadmap towards genetically engineered human embryos.

Last Updated: 26 October 2025

The Possible Approaches

At a high level, there are two possible approaches to creating genetically engineered human embryos:

1. Asexual reproduction (Inner Cell Mass Transplantation, ICMT)

2. Sexual reproduction (In-Vitro Gametogenesis, IVG)

Both these approaches rely on a set of foundation technologies which enable the human germline to be re-created from non-embryonic (somatic) cells and genetically manipulated in the laboratory.

---
title: Genetically Engineered Human Embryos
---
flowchart TB
  a1(eggs)-->|SCNT|a2[ESCs]
  b1((somatic cells))-->|OSKM|b2[iPSCs]
  b1-->a1
  a2-->IVG
  a2-->d1
  b2-->IVG
  b2-->d1
  subgraph engineering
    a2 & b2
  end
  subgraph IVG
    c1{sperm} & c2{eggs}
  end
  d1{ICM}-->|ICMT|d2{TE}
  d2-->embryo
  IVG-->|IVF|embryo
  subgraph embryo
    e(blastocyst)
  end

The Technological Foundations

The technology to genetically engineer human embryos builds on technical foundations laid out in the results of two Nobel-Prize winning discoveries:

1. Sir John Gurdon & Shinya Yamanaka - The Nobel Prize in Physiology or Medicine, 2012

“for the discovery that mature cells can be reprogrammed to become pluripotent”

2. Jennifer Doudna & Emmanuelle Charpentier - The Nobel Prize in Chemistry, 2020

“for the development of a method for genome editing”

The first set of discoveries makes it possible to use non-reproductive cells from individual’s body (e.g. skin cells) to generate pluripotent stem cells which behave like those of the early human embryo. Prior to these discoveries, it was not possible to access pluripotent stem cells without directly sampling tissue from early human embryos.

The second discovery provides a method to edit DNA within the genomes of cells. Prior to this discovery and its descendant technologies, editing the genome was possible but not programmable (constrained to specific sequences of DNA).

Sir John Gurdon, SCNT & ESCs

One of the earliest demonstrations of creating an embryo within the laboratory using non-embryonic tissue as a starting point involved the cloning of a frog (the Xenopus tadpole) based on genetic material from the nucleus of its intestinal cells.

The process, called somatic cell nuclear transfer (SCNT), involves the following steps:

1. Enucleating a Xenopus egg (removing its native genetic material)

2. Re-programming the egg by transplanting a foreign, intact nucleus from a somatic cell

Cytoplasmic factors carried over with the nucleus from the somatic cell result in re-programming of the egg. The re-programmed, diploid egg is functionally equivalent to the single-cell stage of the embryo (zygote). Electrical stimulation is used to trigger mitosis in the zygote and form an embryo.

Pluripotent cells from the ICM

Pluripotent cells (embryonic stem cells, ESCs) can be taken from the inner cell mass (ICM) of the 5-6 day old embryo (blastocyst). The ICM forms the foetus and yolk sac, while the rest of the embryo forms the trophectoderm (TE) which gives rise to the placenta:

---
title: Layers of the Blastocyst
---
flowchart TB
  subgraph TE
  a1[_trophoblast_]-->a2{**placenta**}
  end
  subgraph ICM
  b1[_hypoblast_]-->b2(**yolk sac**)
  c1[_epiblast_]-->c2(**foetus**)
  end

Takahashi, Yamanaka & iPSCs

In the publication Induction of Pluripotent Stem Cells from Adult Human Fibroblasts by Defined Factors, Takahashi, Yamanaka and colleagues were able to convert dermal fibroblasts from the skin cells of human adults back into cells which behave like ESCs.

Since they are not derived directly from the embryo, these cells are called induced pluripotent stem cells (iPSCs). iPSCs enable ESC-like cells to be derived entirely from non-germline cells, overcoming the requirement for egg cells in SCNT.

Microscopic comparison of natural embryonic stem cells (ESCs) and iPSCS.

The Yamanaka factors

The four factors which can be used to re-program adult somatic cells into iPSCs are:

Oct4, Sox2, Klf4, c-Myc

(OSKM)

Collectively, these factors are able to access and alter the activity of specialised genes in heterochromatin and reset the epigenetic landscape to restore the stem cell phenotype.

PSCs as personal stem cells

Pluripotent stem cells (PSCs) - either ESCs from SCNT or iPSCs - are valuable for two reasons:

1. They can theoretically generate all the tissues and organs which make up a foetus.

2. They can be grown in laboratory conditions and cultured (doubled) for hundreds of generations.

In the long term, PSCs can be used to create a personal stem cell banks, held in reserve for:

1. Regenerative medicine: Cells are grown to replace losses due to damage or disease during life.

   • retinal pigment epithelium cells in age-related macular degeneration

   • pancreatic islet beta cells in type 2 diabetes

   • striatal dopaminergic neurons in Parkinson’s disease

   • motor neurons in spinal cord injury

2. Personalised medicine: Lab-grown models of an individual’s organs (organoids) based on their genetic material can be used to test different interventions and compounds for a disease before selecting the best one to administer.

Charpentier, Doudna & CRISPR

Based on a peculiar pattern of repetitive sequences found in the genomes of bacteria (called clustered regularly interspaced short palindromic repeats), it was discovered that bacteria have a defence system against viruses (bacteriophages) which consists of a programmable cursor that can move across the genome and create a break at a target DNA sequence.

The bacterial enzyme which performs this function (Cas9) can be re-programmed to target any short DNA sequence in the genome of other organisms, including humans.

The evolution of gene editors

More advanced versions of this enzyme developed by David R. Liu’s team (base editing, prime editing) can perform chemistry and templated synthesis of new sequences directly into the breaks created by Cas9 in the genome, by adding other enzyme modules.

Cas9, base editor and prime editor (with their respective DNA/RNA complexes).

These new editors extend Cas9’s cursor function to enable insertions, substitutions and deletions:

1. Base editors enable a change in the chemistry of a single base (using a deaminase, e.g. TadA)

2. Prime editors enable the templated synthesis of new sequences (using reverse transcription, RT).

---
title: The DNA Editors
---
flowchart
  A[_Cas9_]-->|single-base chemistry|B[**base editor**]
  B-->|sequence synthesis|C[**prime editor**]

Gene editors can correct mutations in DNA

CRISPR-based DNA editors have been used to correct mutations in DNA which are responsible for inherited diseases due to single-mutation (monogenic diseases), including:

1. relatively common disorders (e.g. sickle cell anaemia)

2. rare, isolated life-threatening diseases (e.g. CPS1 deficiency, a hereditary liver disease which causes an inability to break down ammonia from protein catabolism)

Iterative Engineering of PSCs

In contrast to the original fibroblasts (or any other committed cell type from the donor, which will have senescence limits on the number of divisions they can undergo), PSCs are effectively immortal and can be repeatedly cultured through a long-running series of experiments.

Leveraging this replicative immortality, it is possible to build an iterative workflow in which PSCs are cultured, edited and selected over many generations:

Iterative workflow for the genetic engineering of iPSCs.

After deriving PSCs through SCNT (ESCs) or re-programming (iPSCs), each cycle of engineering the PSCs would involve the following steps:

1. Performing one round of editing

2. Resetting the edited PSCs to the naive state

3. Separating the edited PSCs into individual cells

4. Replicating each cell into a mitotic clone

5. Screening a fraction of each clone

6. Selecting the clone with the correct edit

Resetting edited PSCs to a naive pluripotent state

PSCs from species other than mice tend to exist in a less potent developmental state called the ‘primed’ pluripotent state.

Primed PSCs cannot be separated into individual cells in the laboratory, since they will be unlikely to survive and replicate. This property (clonogenicity) is required for the screening and selection of successfully edited cells.

To remedy this, a partial Yamanaka cocktail consisting of Klf4 and a mutated form of Sox2 can be used to reset the PSCs to the naive pluripotent state between successive rounds of editing.

Screening and selecting PSCs

Depending on the efficiency of the editing strategy, a variable proportion of cells will be successfully edited within a group of PSCs. After resetting to the naive pluripotent state, PSCs can be separated into single cells and then replicated to form clones, and a fraction of each clone can be sampled to determine the clone with the correct edit.

At the DNA level, each clone can be screened whole genome sequencing (looking for on-target editing efficiency and the absence of off-target effects elsewhere in the genome).

At the epigenetic level, an important consideration during selection and screening is avoiding clones which suffer from loss of imprinting. This is a major known issue affecting PSCs whose causal origins are unclear (could be re-programming or in-vitro culture conditions) but mechanistically often involves the hyper-methylation of imprinted genes.

Screening each PSC clone for proper imprinting involves two additional layers of sequencing:

1. RNA sequencing to quantify levels of gene expression from each allele, and

2. Bi-sulfite sequencing to check DNA methylation status at the imprinting control regions (ICRs).

Asexual Reproduction (Cloning)

Engineered iPSCs can be used to create embryos in which the foetus is genetically derived from the individual whose cells were used for iPSC generation. This pathway involves cloning the iPSC donor.

Inner Cell Mass Transplantation

Since iPSCs can only give rise to the layers of the ICM, they need to be combined with a TE layer to re-constitute the whole embryo.

In mouse models, it is possible to do this using tetraploid complementation. In primates, a working approach is to create an IVF embryo and then remove the native ICM, replacing it instead with an exogenous ICM (trophoblast replacement, TR).

These techniques can be grouped under the umbrella term of inner cell mass transplantation (ICMT), in which ICM cells donated from an exogenous source (e.g. iPSCs) are injected into the blastocoele of an embryo which is either missing an ICM (TE-only) or has had its native ICM removed.

Based on successful experiments with TR in NHPs, iPSCs may need to be pre-aggregated (clumped) into an ICM structure prior to injection into the blastocoele in order to maximise the likelihood of successful engraftment with the TE.

The Chimeric Embryo

Once engrafted, the ICM from the iPSCs and TE should form a chimeric embryo.

Within the chimeric embryo, a foetus will develop which consists of genetic material from the iPSCs.

---
title: The Chimeric Embryo
---
flowchart LR
subgraph donors
  a1{somatic cells}
  b1{gametes}
end
a1-->|iPSCs|a2
b1-->|IVF - ICM|b2
subgraph blastocyst
  a2[ICM]
  b2[TE]
end
a2-->a3
b2-->b3
subgraph embryo
  a3(**foetus**, yolk sac)
  b3(**placenta**)
end

Loss of Genetic Diversity

Since cloning an individual is a form of asexual reproduction, this pathway creates a bottleneck in genetic diversity (which relies on sexual recombination under normal circumstances).

Since it bypasses a critical step in natural evolution, this pathway should only be applied in the individual setting and not on a population-wide scale.

Sexual Reproduction (IVG)

The alternative pathway to creating genetically engineered human embryos involves sexual reproduction using eggs and sperm derived from engineered PSCs.

In-Vitro Gametogenesis

In-vitro gametogenesis, or IVG, involves the differentiation of PSCs into the gametes.

This body of work is currently led by the teams of two Japanese scientists, Mitinori Saitou and Katsuhiko Hayashi. One of the central motivations for this work is to enable laboratory-derived sperm (for males > 60 YO) or eggs (for females > 35 YO) in order to extend reproductive longevity.

This is particularly significant as a technological advance for female fertility, since invasive retrieval of eggs and storage in vitrification is currently the only way to save the ability to have children as the quality of eggs decline with age.

The aim of IVG is to recreate the physiological process of gametogenesis in-vitro. The two components necessary to re-capitulate from nature are:

1. The precursor cells which give rise to the gametes (primordial germ cells, PGC)

2. The somatic cells which form the supportive gonadal niche

If the two components are independently derived from PSCs and then re-aggregated in-vitro, they will form gonadal organoids which support the full in-vitro development of functional gametes.

Stem Cell Precursors

In the stem cell stage of IVG, key signalling molecules (cytokines) which are responsible for body patterning in the early embryo (in particular, BMP) can be used to form both the germ cell precursors (PGC-like cells, PGCLCs) and supportive somatic cells:

---
title: IVG in Gonadal Organoids
---
flowchart LR

subgraph stem cells
  a1((PSCs))-->b1(testicular somatic cells)
  a1-->|BMP|b2((PGCLCs))
  a1-->b3(ovarian somatic cells)
end
subgraph rTestis
  b1-.->c1{spermatogonia}
  b2-.->c1
end
subgraph rOvaries
  b2-.->c2{oocytes}
  b3-.->c2
end

In-Vitro Oogenesis (IVO)

In-vitro oogenesis occurs in reconstituted ovaries (rOvaries) and involves three steps:

1. In-vitro differentiation (IVD, 3 weeks) - converting PGCLCs to oocytes in first meiosis (M1)

2. In-vitro growth (2 weeks) - growing the M1 oocytes

3. In-vitro maturation (IVM, 1-2 days) - triggering the M1 oocytes to enter second meiosis (M2)

During the IVD stage, co-culture with ovarian somatic cells (fetal ovarian somatic cell like cells, FOSLCs, which can be generated from the PSCs as well) results in the formation of in-vitro immature (primordial) follicles (PGCLCs → M1 oocytes, FOSLCs → granulosa cells).

During the in-vitro growth stage, granulosa cells in the follicles will proliferate and form projections to the oocytes, resulting in mature (antral) follicles.

Finally, in-vitro maturation (IVM) happens quickly and results in the formation of fertile haploid M2 oocytes. The three stages are driven by changes to the culture media (hormones, growth factors and oxygen levels mimicking the physiological changes in the ovaries during the natural ovulation cycle).

In mice, the M2 oocytes can be fertilised by sperm and yield live births in ~5% of cases. The offspring of these mice appear healthy and fertile (can make offspring of their own).

In-Vitro Spermatogenesis (IVS)

In similar fashion to IVO, in-vitro spermatogenesis involves the creation of a re-constituted testis (rTestis) by re-aggregating PGCLCs with embryonic testicular somatic cells. The resulting organoid will form seminiferous tubules in which intermediate cell types during the natural development of sperm (gonocytes, prospermatogonia) up to spermatogonia (diploid, pre-meiotic precursors to sperm) and in some cases immature (non-motile, tailless) round spermatids.

Since the spermatids are non-motile, physical injection into fertile eggs are required to produce offspring (intra-cytoplasmic sperm injection, ICSI).

Epigenetic Re-programming and Chromosomal Stability

The two most important issues with IVG-derived gametes based on the above studies in mice are:

1. Failure to properly complete the re-establishment stage of epigenetic re-programming, resulting in the incomplete silencing of genes when progressing from earlier to later stages of development

2. Chromosomal aneuploidy in a large proportion of derived gametes, due to impaired pairing of chromosomes at homologous regions during the prophase of meiosis I

Until these two factors are resolved, the gametes from IVG are akin to naturally-sourced gametes from aged individuals. For reproductive success, this is tantamount to undergoing an IVF procedure involving a high attrition rate of embryos (non-viable, with developmental defects).

XY→ XX Conversion in PSCs to Enable IVO in a Male

In the context of male-and-male reproduction, Hayashi’s team have shown that it is possible to convert the XY karyotype into an XX karyotype in PSCs before performing IVG.

The following steps were taken:

1. XY PSCs were cultured over many generations until some mitotic clones lost the Y chromosome

2. These rare XO clones were treated with reversine, which interferes with assembly of a cytoskeletal structure in cells (the spindle) responsible for proper segregation of duplicated chromosomes from mitosis into daughter cells

The second step of this process results in XX cells which can be developed through IVG into oocytes. These male-derived oocytes were fertilised by sperm from another male mouse and led to successful birth of male-parented offspring.

Partial Human IVG with Xenogeneic Somatic Cell Grafts

IVG has been partially successful in the male and female germline setting using a combination of:

• PGCLCs derived from human iPSCs

• Supportive somatic gonadal cells from mice

When re-aggregated, gonadal organoids were successfully created (called xenogeneic re-constituted testes and ovaries, xrTestes and xrOvaries respectively). The xrTestes were able to produce prospermatogonia, while the xrOvaries produced early oocytes.

Neither of these were tested for fertilisation.

X Chromosome Re-Activation in the Female Germline

PGC differentiation into PGCLCs in cells from human females occurs at a lower efficiency of compared to the equivalent process using cells from males.

In female PGCs, it was found that one of the X chromosomes (which is normally inactivated, by a long non-coding RNA XIST) can partially re-activate due to progressive repression of XIST in female PGCs during long periods of culture.

Sourcing Somatic Cells for PSCs

The Accumulation of Somatic Mutations with Ageing

A known problem with somatic cells derived from older individuals is the burden of somatic mutations accumulated in the genome over the lifetime (e.g. from radiation, chemical exposures, cellular senescence and associated failure of DNA repair and replication mechanisms).

These accumulated mutations can increase the risk of cancer formation. Cancer-causing (oncogenic) mutation profiles are also likely to confer a mitotic fitness advantage to carrier cells within the somatic cell population, which will result in a bias towards the genetic profiles of those cells when generating iPSCs.

If cells with a higher somatic mutation burden are used to derive PSCs, it would be necessary to add a large number of additional steps to screen and correct these mutations at the engineering stage.

Cord Blood Cells

The likely best candidate source of somatic cells for PSC generation from the time of birth are cord blood cells (haematopoeitic stem cells).

Haematopoetic stem cells (HSCs) are easily obtained from the umbilical cord and arise from the mesoderm (the middle germ layer in the embryo when it gastrulates, or organises into 3 layers). The mesoderm is the developmental origin of connective tissue and muscle in the developing embryo.

HSCs can be re-programmed to iPSCs through treatment with a cocktail of transcription factors (OSNL: Oct4, Sox2, Nanog, Lin28). HSCs are already routinely banked for allogeneic purposes (donated to other individuals with conditions such as aplastic anaemia or leukaemia who suffer from a depletion of the blood cells in the bone marrow). In the allogeneic setting, HSCs in cord blood banks have to be checked for immune compatibility (HLA-matched) which may limit their application, but in the autologous setting (use on self) there would be no need for such matching.

In light of this comparison, banking HSCs for personal germline preservation and regenerative medicine through autologous re-programming is far less contingent than banking for allogeneic transplantation. The former application (creating an iPSC store) should be the more compelling reason for parents to bank cord blood cells on behalf of their children.

The Future of Reproduction

Balancing the current state of technology with evolutionary considerations as well as the traditional concept of the genetic relationship between parents and their children, the ideal approach to genetic engineering in human embryos would be one which:

1. Preserves built-in mechanisms for genetic diversification (by relying on sexual reproduction)

2. Respects the contribution of both parents to the genetic constitution of the child

3. Utilises the most integral state of the genome from each parent as the starting material

4. Leverages stem cell banking infrastructure at the foundation of regenerative medicine

5. Extends the reproductive longevity of both parents (by relying on banked cells)

In this approach, two individuals with cord blood banked from birth (male-female or male-male; female-female reproduction has not yet been demonstrated) would undertake the following:

1. Re-programming each of their HSC stores into iPSCs

2. Optimisation of their respective iPSC genomes through genetic engineering (correcting any disease-causing mutations that will transmit to their embryos)

3. Differentiation of the engineered iPSCs into a sperm stock and egg stock (IVG)

4. IVF to produce an embryo

The future of reproduction, enhanced by genetic engineering, looks like:

---
title: Genetic Engineering in Sexual Reproduction
--- 
flowchart TB
  a1((cord blood cells))-->|OSNL|b1[iPSCs]
  b1-->|engineering|IVG
  subgraph IVG
    c1{sperm} & c2{eggs}
  end
  IVG-->|IVF|embryo
  subgraph embryo
    e(blastocyst)
  end

Appendix

Primed pluripotency in PSCs

One of the key challenges in working with human PSCs is maintaining their naive pluripotent state.

In contrast to mouse PSCs which benefit from a temporary suspension of developmental potential (diapause), human PSCs tend to stabilise in the primed pluripotent state:

Oct4 Regulates the Naive-Primed Switch

A key mechanism by which PSCs are driven towards the naive state or the primed state is the orchestration of gene networks by control regions of DNA (cis-regulatory elements, CREs) associated with the OCT4 gene.

Depending on which OCT4 CRE is activated, PSCs will have different behavioural biases:

1. The proximal enhancer (PE) orchestrates genetic programs for proliferation and differentiation (priming)

2. The distal enhancer (DE) orchestrates programs for PSC self-renewal (naivety)

The PE is activated by Oct4 homo-dimers, while the DE is activated by Oct4-Sox2 hetero-dimers.

---
title: Regulation of iPSC Outcomes via Oct4
---
flowchart LR
%% OCT4 CREs
subgraph OCT4
  a1
  b1[DE]
end
%% Protein dimers
c{_Oct4-Oct4_}---a1[PE]
a1-->a2(_proliferation_)
%% iPSC phenotypes
d{_Oct4-Sox2_}---b1
b1-->b2(_self-renewal_)

Super-Sox Resets Naive Pluripotency in PSCs

One way in which PSCs can be driven towards the naive pluripotent state is through treatment of the cells with a mutant form of Sox2.

A team led by Sergiy Velychko showed that this mutant (called “super-Sox”, or Sox2-17) increases the ability of Sox2 to bind to Oct4, forming more hetero-dimers which promote the naive state.

Oct4-Sox2 binding the OCT4 distal enhancer (OCT4-DE), adapted from the Velychko team's publication.

The super-Sox mutant can be applied in two ways:

1. Replacing Sox2 with super-Sox when re-programming somatic cells into iPSCs

2. Treating primed PSCs with super-Sox and Klf4 (S*K) to reset then to the naive pluripotent state

The second application of super-Sox can be used to reset pluripotency during long-running experiments which require PSCs to be maintained in the naive state.

Parental Imprinting

An important non-Mendelian mode of inheritance in humans involves the parental imprinting of a number of genes (at least 80). For certain genes such as IGF2 (insulin growth factor 2, which regulates foetal growth and development), only one of the two copies (alleles) of the gene inherited from each parent is active, since sperm and eggs will regulate the allele differently using epigenetic marks. In the case of IGF2:

1. Local DNA methylation activates IGF2 in sperm

2. The absence of local DNA methylation in eggs results in inactivity of IGF2

There are evolutionary reasons for the imprinting mechanism, related to the control of gene dosage. For example, IGF2 is thought to be imprinted in order to prevent the growth of the foetus from exceeding the supportive limits of the placenta.

The epigenetic marks of imprinted genes are wiped and re-established in every generation, at the transition stage between the reproductive stem cells (PGC, primordial germ cells) and gametes (sperm/eggs). The two key stages in this process are:

1. Erasure of epigenetic marks from the current generation within the PGCs

2. Establishment of new epigenetic marks depending on the sex of the gametes

Tetraploid complementation

Tetraploid complementation involves injecting iPSCs into modified embryos without an ICM.

During this procedure, a 2-cell IVF embryo is fused using electrical pulses (electro-fusion, which disrupts the membrane barrier) to make a single cell with 2 x 2N chromosomes.

The resulting tetraploid cells form an embryo consisting only of the TE layer (missing a native ICM). To complete the embryo, iPSCs can be injected into the cavity (blastocoele) to form a new ICM.

Index

germline

The cells which convey the inherited genetic material of a species. They include the gametes (reproductive cells - sperm and eggs) and the early cells of the embryo.

somatic

Of or pertaining to cells in the body which do not constitute the germline of an organism.

pluripotent

Capable of forming all types of cells except for cells in extra-embryonic tissue (e.g. the placenta).

yolk sac

Provides early nutrition and gas exchange to the early embryo, before the placenta takes over, and is the first site of haematopoesis.

heterochromatin

Tightly compacted chromosomal DNA, which is a typical characteristic of differentiated cells.

mitotic clone

To be distinguished from an organism-level clone, a mitotic clone is a group of cells arising from the division of a single cell in culture.

tetraploid cells

From a developmental biology standpoint, polyploidy causes any cells biased towards forming the ICM to undergo apoptosis, whereas in nature TE cells can benefit from polyploidy to support the high endocrine output and tissue-invasive implantation behaviour of the developing placenta.

meiosis

The replication, crossing-over (homologous recombination) and segregation of chromosomes into haploid gametes (each containing only 1 copy of each chromosome). This process introduces genetic diversity into offspring at the level of the individual germline.

• meiosis 1 involves the replication and homologous recombination of chromosomes

• meiosis 2 involves the segregation of chromosomes into halves (chromatids)