Engineering Transcription Factors to Program Cell Fate Decisions

Thumbnail Image




Kabadi, Ami Meda


Gersbach, Charles A

Journal Title

Journal ISSN

Volume Title

Repository Usage Stats



Technologies for engineering new functions into proteins are advancing biological research, biotechnology, and medicine at an astounding rate. Building on fundamental research of natural protein structure and function, scientists are identifying new protein domains with previously undescribed properties and engineering new proteins with expanded functionalities. Such tools are enabling the precise study of fundamental aspects of cellular behavior and the development of a new class of gene therapies that manipulate the expression of endogenous genes. The applications of these gene regulation technologies include but are not limited to controlling cell fate decisions, reprogramming cell lineage commitment, monitoring cellular states, and stimulating expression of therapeutic factors.

While the field has come a long way in the past 20 years, there are still many limitations. Historically, gene therapy and gene replacement therapies have relied on over-expression of natural transcription factors that activate specific endogenous gene networks. However, natural transcription factors are often inadequate for generating efficient, fast, and homogenous cellular responses. Furthermore, most natural transcription factors have complex structures and functions that are difficult to improve or alter by rational design. This thesis presents three novel and widely applicable methods for engineering transcription factors for programming cell fate decisions in primary human cells. MyoD is the master transcription factor defining the myogenic lineage. Expression of MyoD in certain non-myogenic lineages induces a coordinated change in differentiation state. We use MyoD as a model for developing our protein engineering techniques because myogenesis is a well-studied pathway that is characterized by an easily detected change in phenotype from mono-nucleated to multinucleated cells. Furthermore, efficient generation of myocytes in vitro presents an attractive patient-specific method by which to treat muscle-wasting diseases such as muscular dystrophy.

We first demonstrate that we can improve the ability of MyoD to convert human dermal fibroblasts and human adipose-derived stem cells into myocyte-like cells. By fusing potent modular activation domains to the MyoD protein, we increased myogenic gene expression, myofiber formation, cell fusion, and global reprogramming of the myogenic gene network. The engineered MyoD transcription factor induced myogenisis in a little as ten days, a process that takes three or more weeks with the natural MyoD protein.

While increasing the potency of transcriptional activation is one mechanism by which to improve transcription factor function, there are many other possible routes such as increasing DNA-binding affinity, increasing protein stability, altering interactions with co-factors, or inducing post-translational modifications. Endogenous regulatory pathways are complex, and it is difficult to predict specific amino acid changes that will produce the desired outcome. Therefore, we designed and implemented a high-throughput directed evolution system in mammalian cells that allowed us to enrich for MyoD variants that are successful at inducing expression of the myogenic gene network. Directed evolution presents a well-established and currently unexplored approach for uncovering amino acid substitutions that improve the intrinsic properties of transcription factors themselves without any prior knowledge. After ten rounds of selection, we identified amino acid substitutions in MyoD that increase expression of a subset of myogenic gene markers in primary human cells.

Rather than guide cell fate decisions by expressing an exogenous factor, it may be beneficial to activate expression of the endogenous gene locus. In comparison to delivering the transcription factor cDNA, expression from the endogenous locus may induce chromatin remodeling and activation of positive feedback loops to stimulate autologous expression more quickly. Recent discoveries of the principles of protein-DNA interactions in various species and systems has guided the development of methods for engineering designer enzymes that can be targeted to any DNA target site. We make use of the RNA-guided Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)/Cas9 system to induce expression of the endogenous MyoD gene in human induced pluripotent stem cells (iPSCs). Through complementary base pairing, chimeric guide RNAs (gRNAs) direct a Cas9 transcriptional activator to a target DNA sequence, leading to endogenous gene expression. A current limitation of CRISPR/Cas9-based gene regulation is the potency of transcriptional activation and delivery of the CRISPR/Cas9 components. To address these limitations, we first developed a platform to express Cas9 and up to four gRNAs from a single lentiviral vector. We then optimized the gRNAs and Cas9 transcriptional activator to induce endogenous MyoD expression and differentiate iPSCs into myocyte-like cells.

In summary, the objective of this work is to develop protein engineering techniques to improve both natural and synthetic transcription factor function for programming cell fate decisions in primary human cells. While we focus on myogenesis, each method can be easily adapted to other transcription factors and gene networks. Engineered transcription factors that induce fast and efficient remodeling of gene networks have widespread applications in the fields of biotechnology and regenerative medicine. Continuing to develop these tools for modulating gene expression will lead to an expanded number of disease models and eventually the efficient generation of patient-specific cellular therapies.





Kabadi, Ami Meda (2015). Engineering Transcription Factors to Program Cell Fate Decisions. Dissertation, Duke University. Retrieved from


Dukes student scholarship is made available to the public using a Creative Commons Attribution / Non-commercial / No derivative (CC-BY-NC-ND) license.