Alon Kahana, M.D., Ph.D.
Research ProjectsDr. Kahana in his lab


The Kahana Lab studies biology at the intersection of embryology, adult tissue regeneration, stem cells and cancer. As a practicing ophthalmologist and oculoplastic orbital surgeon, Dr. Kahana’s research is focused on disorders of the orbit and visual pathway, but with broad application to a variety of diseases and tissue types. The lab is particularly interested in projects that would lead to breakthroughs in regenerative medicine to cure blinding conditions and cancer.

Extraocular Muscle Regeneration

Binocular vision in humans requires coordinated movement of both eyes, controlled by 6 EOMs per eye. In children, development of binocular vision in general and stereopsis (binocular depth perception) in particular requires properly coordinated eye movement. Poorly coordinated eye movement, i.e., strabismus, can result in amblyopia (neurologic vision loss) and/or permanent loss of stereopsis. In adults, strabismus can result in diplopia (double vision), leading to loss of visual function, as well as in social stigma and isolation. Strabismic conditions are very common, with a pediatric prevalence of approximately 3%, with a total prevalence of 2-5%. While the majority of ophthalmic disorders have benefitted tremendously from advances in molecular genetics and cell biology, treatment of strabismic conditions still relies on occlusion, prisms and surgery, with high rates of failure and potential complications that can result in loss of visual function. The transformative goal of this project is to cure strabismus and preserve vision through regenerative approaches, with further applications to a multitude of skeletal muscle disorders.

Coordinated eye movement requires intact neurologic control and functional EOMs. Disease and fibrosis (e.g. thyroid eye disease, trauma) can cause irreversible muscle damage. In addition, cranial nerve disorders can secondarily lead to atrophy or contracture of EOMs. Unfortunately, a damaged or maldeveloped muscle will result in poor function, irrespective of neurologic input. While mammalian muscles have a very limited capacity to regenerate de novo, there is experimental evidence of limited EOM repair following injury in mammals . Hence, there is potential for therapeutic EOM repair, but we need a better mechanistic understanding of how to trigger and control it.

In order to address the limitations in the field, our laboratory has developed a novel model of EOM injury and regeneration that takes advantage of the genetic accessibility and regenerative capacity of adult zebrafish. Importantly, we have discovered that adult zebrafish EOMs regeneration begins with reprogramming of “post-mitotic” myocytes – multinucleated, syncytial – into dedifferentiated myoblasts capable of robust proliferation to restore the cell number required to regenerate lost tissue. The discovery of myocyte dedifferentiation has been controversial. However, vertebrate cell reprogramming and dedifferentiation is also well studied in adult zebrafish retina, bone, cartilage, heart, and liver, providing excellent models for developing a mechanistic understanding of the reprogramming process. Since myocyte differentiation and biology are extremely well-conserved across species, there is every reason to believe that studying dedifferentiation in zebrafish will enhance our understanding of the potential for dedifferentiation and regeneration in human muscle. Specifically, we hypothesize that mammalian myocytes have gained, through evolution, molecular repressor pathways to prevent myocyte dedifferentiation in order to reduce the risk of dysregulated reprogramming resulting in malignant transformation. It follows that a mechanistic understanding of dedifferentiation would identify molecular activators and repressors that could be utilized to develop novel therapies. The discovery of induced pluripotent stem (iPS) cell reprogramming through the activation of just 4 transcription factors (TFs) – MYC, KLF4, SOX2 and OCT4 – lends significant credibility to the notion that mammalian cell reprogramming is possible once the basic biology is understood.

Our investigations, now published, have revealed that adult zebrafish EOM myocytes reprogram into dedifferentiated myoblasts capable of reentering the cell cycle by 20 hours post injury (hpi). The reprogramming process involves both transcriptional and cytoplasmic events, including FGF-dependent induction of autophagy to disassemble the sarcomeres and the excess nuclei in the multinucleated starting cells. The lab has now turned its attention to understanding the transcriptional and epigenetic processes that underlie myocyte reprogramming.

Cancer Stem Cells

The paradigm that has dominated cancer research over the past 30 years is that cancer is a genetic disease driven by mutation and subsequent selection of malignant cell clones. This has resulted in a massive effort to characterize the mutational spectrum in each type of cancer and to apply this knowledge to specifically target these mutations in individual patient tumors – “targeted therapy.” Despite the potential promise of “personalized oncology treatment,” this approach has led to only modest benefits characterized by short-lived responses for the majority of cancer patients. At the same time, there is growing appreciation that epigenetic mechanisms play a key role in cancer development and treatment resistance. These epigenetic mechanisms in cancer mimic embryologic development and generate tumors that display a hierarchical organization, at the apex of which are tumor cells that demonstrate stem cell-like properties. These cancer stem cells (CSCs) mediate tumor metastasis and play a pivotal role in treatment resistance. The transformative goal of this project is to cure cancer by targeting CSCs.

Cancer stem cells can arise from dysregulation of normal tissue stem cell pathways that regulate their self-renewal, or from reprogramming of more differentiated cells through activation of core transcription factors resembling those activated in induced pluripotent stem (IPS) cells. A unifying concept behind these observations is that cancer is a disease of “aberrant stemness,” which is achieved through a unique stem cell epigenetic state and maintained via a combination of intrinsic and extrinsic signals that also involve the tumor microenvironment. According to this model, rapid development of resistance to molecularly targeted therapies may result from the plasticity of the stem cell state with activation of alternate pathways to maintain the “aberrant stemness.” This suggests that rather than targeting specific mutations, which are but one pathway to “aberrant stemness,” we should develop strategies to target “stemness” itself, i.e. the shared downstream consequence of these mutations.

In order to elucidate the fundamental molecular mechanisms that drive and maintain “stemness” in CSCs, we propose to utilize a novel model of dedifferentiation in adult zebrafish that is robust and results in a large population of reprogrammed dedifferentiated progenitor cells.  Our preliminary data reveal high-level concordance between the molecular genetic and epigenetic pathways in dedifferentiated zebrafish cells and human CSCs. An advantage of this approach is that it permits the de-coupling of “stemness” from other secondary changes driven by oncogenesis and genetic instability in tumors.  Utilizing this concordance, we propose to identify shared candidate pathways and nuclear states that define the stem cell state in dedifferentiated cells. Using genome wide conformation capture (Hi-C), 3D-FISH, RNA sequencing (RNA-Seq), non-coding RNA analysis, and chromatin immunoprecipitation-sequencing (ChIP-Seq), we will probe the nuclear architecture alterations and chromatin remodeling that underlie reprogrammed stemness. Further experimental and computational comparisons with zebrafish embryonic neural crest and human embryonic stem cells will add fundamental insights into the differences between the embryonic versus reprogrammed stem cell states, to further narrow putative targets for therapeutic targeting. These putative targets would be validated utilizing in vitro human and transgenic mouse models of breast cancer stem cells. Once validated, the zebrafish dedifferentiation model would be utilized to screen through and develop drugs that target “stemness,” taking advantage of the ease of genetic manipulation and drug delivery as well as the physiologically relevant microenvironment of the zebrafish. At the conclusion of the full project, we would identify key molecular pathways and develop drugs that  directly target cancer “stemness”.

Thyroid Eye Disease

Thyroid disorders are among the most common endocrine disorders, affecting 20 million Americans, 30 million Europeans, and millions more throughout the developing world, particularly in areas of low iodine intake.[1-3] In areas with adequate iodine dietary intake, thyroid disorders are usually related to auto-immunity, ranging from Graves disease (hyperthyroid autoimmune disease) to Hashimoto’s (hypothyroid autoimmune disease), with associated thyroid autoantibodies.

With wide spread availability of thyroid hormone-containing medications – both synthetic and natural – the endocrinologic derangement can essentially be cured. In hypothyroid states, patients are supplemented with thyroid hormone, and in the hyperthyroid state, the inflamed thyroid gland can be ablated surgically (thyroidectomy) or via radioactive iodine (RAI) administration, followed by thyroid hormone supplementation to achieve a euthyroid endocrine status. If thyroid gland ablation is not desired, a euthyroid state can still be achieved in many cases by inhibiting the activity of thyroperoxidase using either methimazole or propylthiouracil. Either way, management of the endocrine disorder is usually straight forward and generally successful.

A key sequela of autoimmune thyroid conditions is the development of thyroid eye disease (TED), also known as Graves orbitopathy, thyroid associated ophthalmopathy, and other similar terms that pertain to the same disorder. Importantly, while the treatment of dysthyroid endocrine disorders is usually straight forward, the treatment of TED is anything but, and long after the endocrine disorder has been treated and stabilized, patients continue to wrestle with the after effects of TED, namely the disfigurement, discomfort, social isolation, diplopia, dry eyes and sometimes even loss of vision. The transformative goal of this project is to cure TED and prevent its occurrence in the first place.

TED is considered an inflammatory orbitopathy characterized by a multitude of ophthalmic signs and symptoms. These include eyelid retraction (most common), conjunctival chemosis, eyelid edema, ocular surface disease (unstable tear film and exposure keratopathy), exophthalmos, restrictive strabismus with diplopia, orbital compartment syndrome, compressive or stretch optic neuropathy, chronic pain, facial deformity, and vision loss that includes the possibility of blindness and loss of the eye from corneal perforation. Thyroid associated periorbitopathy refers to changes that occur to the periocular skin and subcutaneous tissues. And TED is commonly associated with a more systemic dermopathy, most commonly manifesting as pretibial myxedema.

TED is associated with a hyperthyroid state in 90% of cases but it can also occur in hypothyroid and euthyroid patients. Risk factors include smoking, fluctuating T3 levels, post-ablative hypothyroidism, high levels of thyroid stimulating immunoglobulin, radioactive iodine treatment (particularly in the absence of post-treatment steroids), and associated dermopathy (pretibial myxedema). The pathogenesis of TED is thought to involve antigenic mimicry, with autoantibodies against the TSH receptor and associated IGF receptor attacking orbital tissues. This would, in turn, lead to activation of orbital fibroblasts and release of proinflammatory cytokines, with a subsequent cycle of infiltration, fibrosis and transdifferentiation into myofibroblasts and adipocytes. However, antigenic mimicry fails to explain the reason for the selective involvement of orbital tissues and skin, since most tissues in the body express TSH and IGF1 receptors. Furthermore, efforts to trigger the orbitopathy immunologically with administration of autoantibodies have been only marginally successful.

To explain these discrepancies, we hypothesize that the orbit and skin share a unique physiologic trait that predisposes them to pro-fibrotic inflammation in response to a dysthyroid condition. We further hypothesize that TED is driven by a local trigger that is distinctly separate from the inflammatory process that propagates the disease, i.e. a “Trigger-Propagation” hypothesis. We are currently pursuing novel approaches to test our hypothesis and identify the triggers for de novo cytokine release in orbital tissues of patients with autoimmune thyroid dysfunction.


Figure 1: A zebrafish lateral rectus muscle undergoing dedifferentiation, with myofibers in green down and to the left, and dedifferetiated proliferating mesenchymal cells in beige at top right. DAPI marks nuclei (blue).


Figure 2: A craniectomy view of a regenerating lateral rectus muscle (left) and control (right). Proliferating nuclei marked in red. DAPI marks nuclei (blue).


Figure 3: A close-up of figure 2, showing the transition from an area of muscle to an area of dedifferentiated mesenchymal cells that underwent reprogramming to become proliferating myoblasts.



Movie 1: Adult zebrafish display a stereotypical optokinetic response (OKR)


Adult zebrafish display a stereotyped pattern of eye movements, known as an optokinetic reflex (OKR), in response to moving visual stimuli. This response is evoked by placing zebrafish on a stationary pedestal within a rotating optokinetic drum that is marked on the inner wall by alternating black and white stripes. Generation of this response requires (1) vision, and (2) extraocular muscle activity. In the movie, the first 10 seconds reveal left-sided eye pursuits followed by rapid right-sided saccades in response to counter-clockwise rotating stripes. At 11 seconds, drum rotation is changed to clockwise, and the fish responds appropriately by tracking the stripes to the right. This adaptive trait is found in most vertebrates and allows for moving images to be stabilized on the retina. The Kahana lab is using this behavioral assay to assess extraocular muscle function in our extraocular muscle regeneration research.


Movie 2: Rostral Neural Crest Migrates in Dorsal and Ventral Waves Around the Eye to Populate the Frontonasal Process and Periorbital Mesenchyme


Multiphoton time-lapse imaging of Tg(sox10::EGFP) from 12 to 30 hpf demonstrated that sox10 expressing cells migrate into the head via waves that migrate dorsal and ventral to the eye to populate the frontonasal process and periorbital mesenchyme.


Movie 3: Thyroid hormone signaling is required for initiation of ventral wave migration of rostral neural crest cells. (Multiphoton time-lapse microscopy)


Following morpholino knockdown of the thyroid hormone receptor Thraa (which would block thyroid hormone signaling), the ventral wave fails to migrate, while cells of the dorsal wave die following migration.


Movie 4: Exogenous retinoic acid rescues the phenotype of thyroid hormone receptor knockdown


Time-lapse imaging of Tg(sox10::EGFP) from 12 to 30 hpf of Thraa MO knockdowns treated with 1 nM RA (starting at 12 hpf) demonstrated rescue of migration of the ventral wave of the rostral neural crest.


Kahana research facilities
Last Modified: Friday, 27-Jan-2017 10:29:13 EST