|Home | About | Journals | Submit | Contact Us | Français|
Hearing loss is a costly and growing problem for the elderly population worldwide with millions of people being affected. There are currently two prosthetic devices available to minimize problems associated with the two forms of hearing loss: hearing aids that amplify sound to overcome middle ear based conductive hearing loss and cochlear implants that restore some hearing after neurosensory hearing loss. The current presentation provides information on the treatment of neurosensory hearing loss. Although the cochlear implant solution for neurosensory hearing loss is technologically advanced; it still provides only moderate hearing capacity in neurosensory deaf individuals. Inducible stem cells and molecular therapies are appealing alternatives to the cochlear implant and may provide more than a new form of treatment as they hold the promise for a cure. To this end, current insights into inducible stem cells that may provide cells for seeding the cochlea with the hope of new hair cell formation are being reviewed. Alternatively, similar to induction of stem cells, cells of the flat epithelium that remains after hair cell loss could be induced to proliferate and differentiate into hair cells. In either of these strategies, hair cell specific genes known to be essential for hair cell differentiation or maintenance such as ATOH1, POU4F3, GFI1, and miRNA-183 will be utilized with the hope of completely restoring hearing to all patients with hearing loss.
In the United States, over three million hearing disabilities are registered. The aging population of developed countries will see a significant rise in this problem with an expected frequency of approximately 50% of people suffering from some degree of hearing loss by age 65. Thus, many retirees may find themselves cut off from the social communications they enjoyed throughout their lives. Hearing loss can be attributed to two distinct problems: conductive hearing loss in which the function of the middle ear is partially or completely lost and neurosensory hearing loss. Neurosensory hearing loss is an irreversible loss of hair cells in the cochlea, followed over time (10 years of more) by loss of sensory neurons. Hair cell loss can be attributed to environmental factors (such as loud noises), ototoxic medications (cysplatin, aminoglycosides) and genetic predispositions. Age related hearing loss may represent an accumulation of environmental and genetic factors compounded by genetic predisposition, cumulative effects of ototoxic treatments, environmental insults that include infection, and increasing mutational load with aging in mitochondrial DNA.
Some vertebrates such as bony fish and birds have the capacity to regenerate damaged or lost hair cells through proliferation of supporting cells followed by transdifferentiation of postmitotic supporting cells into hair cells.1 Because humans (and other mammals) lack this regenerative capacity in the ear, thousands of people worldwide currently rely on cochlear implants (CI) as the only treatment for neurosensory hearing loss.2-4 Although CI represent a technological breakthrough, they have inherent limitations that prevent them from completely restoring hearing loss to the level of natural hearing. Even the best implant electrodes that rely on residual functions of the apical, low frequency end of the cochlea,5, 6 still cannot match the innate ability of the fully functional organ of Corti. The advancement of molecular and cellular therapy using pluripotent stem cells and continued expansion of our knowledge of the molecular basis of embryonic hair cell formation, sets the stage for multiple attempts worldwide to develop new solutions for hearing loss that have the goal of restoring the lost hair cells and thus natural hearing. Attempts at seeding the ear with pluripotent stem cells, with the intent to form new hair cells are on the horizon and have been successfully completed in animal models for sensory neurons.7 Additionally, although they are still in the developmental phase, current studies are working toward ways to genetically up-regulate transcription factors in the remaining epithelium that replaces the organ of Corti after hair cell loss to regenerate hair cells and their associated supporting cells. In parallel, molecular and cellular therapeutic progress is working to retain sensory neurons after hair cell loss and thus ensure long term viability of CI or, should hair cell regeneration be achieved, connect regenerated hair cells with the brain.2, 3 This review will describe the current state of these attempts and highlight necessary steps that must be taken for the successful reversal of hearing loss.
There are many forms of hearing disorders: the two most prominent are conductive and neurosensory losses. Patients with conductive loss are viable candidates for hearing aids. Hearing aids amplify the sound waves from the environment causing the tympanic membrane to become more responsive, which allows the sound energy to reach the inner ear at audible levels. Necessarily, hearing aids rely on a healthy inner ear with functional mechanoelectric transducers: the hair cells. The other form of hearing loss is neurosensory, most commonly caused by a loss of hair cells. In patients suffering from this form of loss, amplification of sound waves by hearing aids will not enhance sound perception as the sensory organ of Corti is critically damaged through the loss of hair cells. Currently, CI are the only treatment available to these patients. CI take the sound waves directly from the environment to the spiral ganglion, bypassing much of the inner ear. Only patients with complete or nearly complete deafness are considered candidates for CI.
In 1984 the first CI were approved for commercial distribution and by 2005, CI had successfully restored at least partial hearing to over 100 000 people.8 Theoretically, the concept of a CI is simple. An external microphone receives sound waves from the environment, changes the mechanical energy into a digital signal through a speech processor, and converts that signal to an electrical impulse which travels through a wire placed strategically in the cochlea. Along the insulated wire, focal active points function as electrodes that stimulate nerve endings which trigger activation of the spiral ganglia and prompt a cascade of impulses ending in the brain for auditory interpretation.9 Because of the crude “robot-like” sound produced by the CI, the patient needs to train continuously with speech therapists to improve word recognition in a process called mapping. With successful surgery and therapy, patients will progress from near deafness to an acceptable level of effective communicative ability.10
While this may seem simple, we are limited by both the physical principles of the peripheral auditory system and knowledge of the appropriate biological and mechanical manipulative steps required for restoration of hearing. How to replicate approximately 15000 hair cells is still unknown, but recent trends have made improvements in sound quality and speech discrimination and recognition by changing the number and placement of electrodes in the cochlea, using hybrid implants, testing bilateral application of CI and altering the number of channels along with the rate of stimulation of the electrodes.
By manipulating the number and placement of electrodes along the cochlea, patients may improve pitch discrimination. A 2007 study showed that “turning off” (essentially reducing the number) of electrodes deeper than 560 degrees into the cochlea improved consonant and vowel recognition. There were two reasons for this improvement: 1) a reduction in interaction between overlapping apical electrodes (electrode interaction causes a degree of interference) and 2) an improvement of the alignment of the analysis filters to the estimated overall pitch.11 This combined technique or the insertion of a short-cochlear or hybrid implant may be ideal for older patients who have retained low but not high frequency hearing.12 These hybrid implants improve word and sentence recognition for profound high frequency hearing loss,10, 13-15 the most frequent of neurosensory hearing loss. In these patients, the implanted ear will receive electronic stimulation for the first 10-20 mm and acoustic residual hearing for the remainder of the cochlea. In the non-implanted ear, the patient will only hear the residual acoustic stimulation, perhaps with amplification of a hearing aid.12, 13, 15 Also, low frequency acoustic hearing improves pitch recognition which has positive implications in melody recognition.16
Research is also underway to improve sound localization and speech recognition in poor signal to background noise environments, such as “competitive talking”. Using bilateral application of CI has shown promising results thus far.17 Increasing the number of channels (from 8 to 16) and rate of electrode stimulation (813 pulses per second [pps] to 5 100 pps) has also allowed patients to discriminate signal from background allowing better word recognition.18
While many of these new technologies show promise and have allowed patients to improve their hearing and quality of life, there is still a great debate as to pros and cons of CI. Benefits have already been mentioned, i.e. restoration of hearing to a degree and the ability to resume/restore viable communication, but it is important to note that CI do not “cure” the patient of hearing loss or restore “normal” hearing. CI patients have deteriorating inner ears. Hair cells are dying, and with hair cell death, important neurotrophic factors are not being supplied to sensory neurons. In time, these neurons will die and not even CI will enable sound perception. This area has shown promising developments as more has been discovered about delivering neurotrophic factors, independent of hair cells, in CI recipients to prolong nerve and thus, hearing, longevity.3, 19
Research is making strides to enhance sound quality and discrimination between signal and background noises 17 but this technology is still inferior to human natural functions. Furthermore, CI restrict laboratory tests such as magnetic resonance imaging and can be inconvenient (battery replacement, removal prior to water exposure, mechanical pieces). Cochlear implantation also has a negative societal stigma. The 2000 documentary “Sound and Fury”, highlighted the chasm among the deaf community. The deaf community has long rejected technology that allows a deaf person to hear.8 To this day, the National Association of the Deaf does not widely accept CI as a primary treatment option; a stance that alienates those in the deaf community who choose to undergo CI surgery.8
While implants and general knowledge of the ear have improved, CI is still not as good as human natural hearing and certainly will never provide a cure. Technology is quickly progressing so that someday we may be able to satisfactorily approximate the natural hearing.20 Meanwhile, molecular biology is racing ahead in genetic manipulation of existing cells as well as embryonic and adult stem cell technologies. With our growing knowledge base in functional genomics, it would be possible to over-express regulatory genes causing the transdifferentiation of supporting cells into hair cells.21 The drawback to this approach, however, is that it would deplete our supply of supporting cells, causing structural disorganization to the organ of Corti.21 Although stem cell technology is, in many ways, still in its infancy, in the long run it may be possible to differentiate inducible pluripotent stem cells (iPS cells), through a well designed series of intermediates, guided by numerous transcriptional and diffusible factors, to replace the lost hair cells. Although many major hurdles still exist, this treatment has the potential to become the gold standard for restoration of hearing loss.
Embryonic stem (ES) cells come from the inner cell mass (ICM) of the pre-implantation blastocyst. Human ES cells are pluripotent (not totipotent) and maintain the ability to differentiate into any cell type derived from the three germ layers: ectoderm (skin, nerve), mesoderm (muscle, bone, and blood), and endoderm (liver, pancreas, lungs, and gastrointestinal tract). In addition, all ES cells have the capability of self-renewal/regeneration. Most of the studies done to date are on mouse ES. More recently, however, studies are being done on human ES cells with the hope of developing treatments with clinical applications.
All embryonic stem cells express a subset of genes shown to be critical for the maintenance of pluripotency. These genes include Oct4, NANOG, and Sox2.15 Absence or deregulation of these genes causes the cell to lose its pluripotent capacity and differentiate. Oct4 (alias, POU5F1) is a transcription factor, and part of the POU (bipartite DNA-binding domain) family. Oct4 is essential for pluripotency of the cells in the early embryo and Oct4-deficient mouse embryos develop to the blastocyst stage, but the ICM is not pluripotent and instead is destined to develop into trophoectoderm.23, 24 Furthermore, a less than two-fold increase in Oct4 expression leads to differentiation into primitive endoderm and mesoderm.25 In both cases, pluripotency is lost.
Nanog is a homeodomain protein. A decrease in Nanog expression leads to differentiation of the ICM into parietal endoderm-like cells; whereas an increase in Nanog leads to self-renewal/propagation with an increase in Inhibitor of Differentiation Protein (ID) production. Increased p53 (tumor protein 53, TP53) causes a decrease of Nanog and leads to differentiation.26
Sox2 is a member of the sex determining region Y-related High Mobility Group (HMG) box family that encodes transcription factors with a single HMG DNA-binding domain. Sox2 is expressed in pluripotent and multipotent cells. In Sox2 null mice the ICM does not form and, as in the Oct4 mutants, the cells that survive develop into trophoectoderm.27 Sox2 and related SOX genes are essential for neuronal stem cells as demonstrated by conditional mutant mouse lines,28 and are important factors upstream of hair cell formation in the ear.22, 29 These conditional mutations have a wild-type phenotype under certain (permissive) environmental conditions and a mutant phenotype under other (restrictive) conditions.
Human ES cells are being studied with the hope that they may someday replace solid organ transplantation and treat chronic diseases such as Parkinson’s. Differentiation protocols for human ES cells exist for many cell types including cardiomyocytes, hepatocytes, hematopoietic cells, neurons and endothelial cells. Despite recent progress, several obstacles remain including the risk of tumor development and immune rejection of the implanted cells by the host.
One source of immunogenicity from cultured ES cells comes from the use of animal products such as fetal calf serum or mouse fibroblast cells during the development process. This not only introduces the risk of transmission of infection, but also carries the possibility of the human ES cells taking up animal antigens into the cell line. Techniques such as the use of human skin fibroblasts as feeder cells help to eliminate some of this risk. The other source of immunogenicity from ES cells arises from the rejection of the ES cells by the host. This occurs when the ABO and HLA type from the ES cells does not match that of the host. Although it is possible to reduce the mismatch between ES cells and the host, generating a perfect match for the host is not possible. Therefore, the use of ES cells for transplantation into the human host will require the use of immunosuppressive therapy. These therapies not only cause reduced ability to fight opportunistic infections and suppress tumors, in addition the side effects of such agents include kidney failure, osteoporosis, diabetes and hypertension.30
One solution to this immunogenicity issue could be cloning; a process in which the nucleus of an oocyte is replaced with that of a somatic cell from the donor. The cells are then harvested at the blastocyst stage providing stem cells that are ABO and HLA matched with the exception of mitochondrial antigens from the oocyte donor. This process has been accomplished with animal cells, but not performed using human cells. This technique is laden with ethical issues as the product, if it were not destroyed at the blastocyst stage, could develop into a cloned person.30
Although embryonic stem cells have received much attention, due to their capacity to differentiate into virtually any cell type, there are sources of stem cells in the adult as well. Such adult stem cells have lost the ability to differentiate into all cell types, but retain the capacity to differentiate into a subset of cell types. Importantly, these cells also retain the ability to continue to divide, although at a slower rate than ES cells, to maintain their own pool and to give rise to daughter cells that replenish lost tissue. Although adult stem cells likely exist in the inner ear,31 they are difficult, if not impossible, to retrieve without the destruction of the very organ that they should restore and therefore are not likely to be useful in human therapies. A few other examples of organ systems with adult stem cells include the immune system, the hematopoietic system, the eye, skin and hair follicles, finger nails, and even one neurosensory system: the olfactory system.32
Here a few known sources of easily accessible adult stem cells will be highlighted. Focusing on these skin derived cells will provide significant cellular material for each person to be used without rejection to replace lost hair cells. It is also anticipated that transformation of ectodermal adult stem cells into embryonic ectoderm derived hair cells could prove to be easier than starting from adult bone marrow.
Limbal epithelial stem cells (LESC) are a population of ectodermally-derived stem cells that reside at the junction of the cornea and the conjunctiva of the eye. These cells are responsible for replenishing the corneal epithelium throughout life. Adult stem cells in the cornea are smaller than nearby epithelial cells. An increase in size is associated with increased differentiation and a loss of growth capacity. Cultured LESCs are one of the few adult stem cell populations that have been successfully used for clinical treatment.33
The hair follicle is another source of pluripotent, neural crest-derived stem cells. These stem cells have been isolated from the whisker follicles of mice and shown to continue to divide and develop into neurons, smooth muscle cells, Schwann cells and melanocytes. Sieber-Blum et al.34 were also able to direct the differentiation of these cells into other neural crest-derived cell types by culturing the cells in the presence of various regulatory proteins (neuregulin1 [NGR1] for Schwann cells and bone morphogenetic protein [BMP2] for chondrocytes).34 Furthermore, hair follicles drive the replacement of hairs in a cycling fashion. To do so, hair follicles contain stem cells.35 These cells have already been shown to express GATA3,36 a gene critical for early inner ear development and hearing,37, 38 and could be transformed into neurosensory precursors in the presence of other critical ear specific genes such as NEUROG1 and FOXG1 22 in tissue culture.
Olfactory epithelium also has profound stem cell potential as it not only retains the capacity to regenerate olfactory receptor neurons, but is also surgically accessible. This specialized tissue also produces several of the same genes that are expressed in the developing inner ear including NEUROG1, NEUROD1 and FOXG1.32, 39 Understanding the molecular basis of proliferation, regulation and differentiation of this continuingly replenishing sensory system could provide great insights for the molecular blocks that apparently affect similar replacement of sensory hair cells in the mammalian ear.
Recently, a new technique for generating stem cells was described in which retroviral transduction was used to generate ectopic expression of various transcription factors. Oct4, Sox2, c-MYC and Kruppel-like factor-4 (KLF4) were determined to be necessary to “reprogram” fibroblast cells into pluripotent stem cells (also called induced pluripotent stem [iPS] cells) that have many of the same features as embryonic and adult stem cells, including self renewal and differentiation into various cell types. Further, when injected under the skin of mice, some but not all of these cells formed tumors that contained cell types from all three germ layers.40 The same transformation of adult cells was achieved in parallel using retroviral vectors to transduce Oct4, Sox2, Nanog, and Lin28 in the mouse 41 to generate iPS cells. IPS cells may have less transformation potential than ES cells but since iPS cells are reprogrammed adult cells that are directly derived from a given patient; the major problem of immunohistochemical incompatibility is essentially overcome.
The enormous potential of induced hematopoietic progenitors from iPS cells derived from skin fibroblasts was demonstrated by the cure of a humanized form of sickle cell anemia in a mouse.42 This proof of principle experiment coupled with the successful induction of pluripotent stem cells from fibroblasts provides hope in using iPS cells for cellular gene therapy. However, due to the retroviral transfection needed to randomly insert “stem cell” genes to form iPS cells, there is still a strong risk of tumor formation, making them currently unsuitable for human trials.43 Next generation approaches are necessary that allow stable insertion of those genes into the skin derived cells without any risk for tumor transformation. One option that is being explored is the use of small molecules that can slip through the cell membrane and into the nucleus to turn on specific genes.43 Another approach that just begins to be explored is the use of bacteriophage derived integrases.44 These enzymes recognize approximately 48 pseudo-integration sites in the human genome and would thus allow stable insertion of ‘stem cell’ transcription factors without tumor induction.44 Combined with proper genetically engineered driving systems that allow activation of those integrated gene “on command”,45 use of these synthetic gene regulators or circuits could possibly result in regulated gene expression with tumor formation.
In order to eventually manipulate iPS cells to generate cochlear hair cells, it should be firstly understood how the cell fate of the sensory cells of the inner ear is initially designated and the molecular mechanisms by which these cells then differentiate into functional hair cells. Only once these steps are understood it would be possible to recapitulate this process in pluripotent stem cells using a minimal and essential set of core transcription factors.
The inner ear develops from a patch of ectodermal cells called the otic placode.46, 47 This patch of cells eventually gives rise to all the cells of the inner ear, producing the three-dimensional labyrinth structure as well as the neurosensory epithelia required for normal hearing and balance. Inherent in this process is the differentiation of these specified ectodermal cells into sensory neurons, hair cells, supporting cells, and non-sensory epithelial cells.
Induction of the otic placode from the embryonic ectoderm requires the generation of molecular gradients of diffusible molecules produced by the surrounding tissues. Such factors that have been characterized for mammalian ear development include Sonic hedgehog from the floor plate and notochord, fibroblast growth factors from the mesoderm and neuroectoderm, wingless related nuclear transcription factor from the hindbrain, and BMPs from the ectoderm. Dimerized BMP receptors phosphorylate SMADs, which then enter the nucleus to regulate over 500 genes. The diffusible factors mentioned above commonly complex in their intracellular signal pathway with SMADs and enhance the transformation from ectoderm to the otic placode. Acting together, these molecules induce the formation of the otic placode, committing these cells to develop into the inner ear. The first steps of transformation of ectoderm to otic ectoderm are not yet fully characterized molecularly. However, certain critical events are now specified on a molecular basis 22, 48, 49 that lead to the upregulation of a number of transcription factors such as EYA1, FOXI1, GATA3, and FOXG1 (to name but a few).
Upon transformation to the otic placode, NEUROG1 is upregulated within a subpopulation of cells as an early indication of neurosensory commitment of otic placode cells.50, 51 NEUROG1 commits these cells to a neurosensory fate. Although the expression of NEUROG1 is necessary for neurosensory cell fate determination, it is not sufficient. Other genes involved in this specification include but are not limited to GATA3, PAX2/8, FOXG1, FOXI1, ISLET1, Sox2, LMX1A, EYA1 and SIX1. Loss of function analysis using general or conditional null mutations of several of these genes has revealed their essential role in these processes. However, neither the full complement of necessary transcription factors nor their hierarchy or interactive network is fully understood, limiting our abilities to pick the most relevant genes for the intended transformation of iPS cells into otic placode-like stem cells.
For example, no neurons will form in the absence of NEUROG1.50 In contrast, these neuronal precursors, with NEUROG1 present, will continue their specialization into a neuron. Alternatively, precursors which express Sox2 followed by Atoh1 will differentiate into hair cells.29 Recent studies have shown that by embryonic day 10.5, the neurosensory precursors consist of three distinct populations: neuronal precursors that form neurons, sensory precursors that form hair cells and precursors that can switch their fates and differentiate into either neurons or hair cells.22, 52 It is this latter group that has a great therapeutic interest. If it is possible to transform iPS cells into otic placode neurosensory stem cells, it would also be possible to create neurosensory cells that could give rise to both neurons and hair cells. This concept, initially proposed based on the apparent loss of hair cells in NEUROG1 null mutants 50, 53 was recently been demonstrated in cell lineage analysis using tamoxifen inducible NEUROG1-CreERT.52 Next it will be discussed whether it is possible that co-expression of NEUROG1 with Sox2, GATA3 and Oct4 suffices to transform skin derived iPS cells into cells that respond to the ear environment to differentiate as hair cells.
With the current understanding of the molecular basis of inner ear development in general, and hair cell formation in particular, what we have learned could be applied to the generation of iPS cells with the hope of developing a permanent solution to hearing loss. Because the technique for inducing pluripotency in adult cells is demanding, the percentage of cells that are actually transformed into iPS cells is quite low.54 Maybe, one explanation for this is that adult stem cells are present in small numbers in the tissue used for the transformation; and that it is only in these adult stem cells, that are not terminally differentiated and are already poised for regeneration, that the added transcription factors are able to cause ‘dedifferentiation’ to produce ES-like cells. A previous group has already used ES cells for seeding the inner ear of chicken embryos 55 and thus provided proof of principle for cell based therapy. The following section outlines a proposal for the mouse which avoids the use of ES cells by generating iPS lines (Figure 1).
The transcription factor, GATA3, has been demonstrated in adult stem cells of several tissues including the epidermis and hair follicle,56 kidney,57 and hematopoietic cells.56 Furthermore, a Tg-GATA3(GFP) transgenic mouse has been developed.57 Using the GFP tagged protein, it is therefore possible to use flow cytometry to separate GATA3-GFP positive adult stem cells in the skin from fully differentiated skin cells. Transformation of the GATA3-positive cells with Sox2 and Oct4 should yield a high percentage of ES-like colonies, while the other set of cells should yield few or none. If successful, such an approach will not only yield a higher percentage of iPS cells but would also predispose those cells toward otic transformation through the co-expression of Sox2, GATA3, and Oct4. Only few additional factors are needed to fully unfold the capacity of these cells to behave like otocyst neurosensory stem cells. The most likely factors to be tested are EYA1, NEUROG1, FGF10 and ISLET1 (Figure 2).
It is possible that those cells, once seeded into the ear, can pick up proper cues to differentiate into hair cells. Thus, it may suffice to transfer these cells to the ear that has lost hair cells. However, it is possible that the adult ear will not provide enough contextual information to induce such differentiation since embryonic transcription regulators necessary for this transformation are no longer active. If so, it will be necessary to drive the otocyst-transformed iPS cells to differentiate into hair cells. In this second transfection four genes may be inserted: Atoh1, Pou4f3, Gfi1, and miRNA-183.58 The rational for choosing these genes is as follows:
Atoh1 (formerly Math1) is a basic helix-loop-helix gene that is closely related to the Drosophila proneuralgene, atonal. This highly-conserved gene is expressed in all sensory hair cells of the inner ear and is critical for hair cell generation as fully differentiated hair cells are not found in Atoh1 null mice.59 Atoh1 does not appear to be required for the establishment, or cell fate specification, of hair cells,60 but does play a critical role in the maturation or differentiation of sensory hair cells from within the established epithelium.61, 62
Pou4f3 (formerly, Brn3c) is another gene that is critical for hair cell development and, most importantly, maintenance.63 In the inner ear, Pou4f3 is expressed specifically in auditory and vestibular hair cells. Further, mice with a mutation of the Pou4f3 gene are deaf and have impaired balance due to a complete loss of auditory hair cells with subsequent loss of sensory neurons.64, 65 Data suggest that, while Pou4f3 is not required for cell fate specification, it is required for hair cell differentiation and survival.66
Growth factor independence 1 (Gfi1) is a possible downstream target of Pou4f3;66 and its expression in the ear is confined to sensory hair cells and neurons. Gfi1 mutant mice display behavioral defects that are consistent with inner ear anomalies and do not respond to noise. Although Gfi1-deficient mice initially develop inner ear hair cells, these hair cells are disorganized. The outer hair cells of the cochlea are improperly innervated and express neuronal markers that are not normally expressed in these cells. Furthermore, Gfi1 mutant mice lose all cochlear hair cells around the time of birth.67 Thus, Gfi1 is also required for inner ear hair cell differentiation and survival. It is still unclear whether Gfi1 is regulated by Pou4f3; however, the insect ortholog of the Gfi1-null, senseless, has the capacity to form sensory cells even in the absence of atonal, the insect ortholog of Atoh1.68, 69
Finally, miRNA-183 is one of the many microRNAs that act post-transcriptionally to regulate gene expression. miRNA-183 is expressed specifically in the hair cells of the inner ear and is thought to play a role in silencing genes that support non-hair cell fates.70 Additionally, the miRNA-183 gene is very highly conserved evolutionarily, demonstrating strong expression in the inner ear of the lamprey and hag fish as well as in putative sensory organs of the acorn worm.71 The restricted expression along with its high conservation suggests that miRNA-183 is a critical player in inner ear hair cell specification. Indeed, our unpublished data strongly support the notion that specific miRNA expression is necessary for other transcription factors to achieve the desired cell transformation. Presumably, miRNA-183 eliminates transcripts that might otherwise persist through cellular transformations and/or hones the repertoire of transcripts up-regulated by hair cell specific transcription factors to more effectively transform cells into hair cells (Figure 2).
In analogy to the effect of multiple gene expression to transform adult cells into iPS cells,41 the co-expression of four or more factors could drive such iPS cells to differentiate in the ear as hair cells. To achieve full transformation of these iPS cells into hair cells, it may be necessary to inject iPS cells treated with transient expression of Sox2, GATA3, Oct4 and NEUROG1 (and/or EYA1, FOXG1, FGF) into the cochlea and activate the hair cell specific genes using regulatable expression systems such as tamoxifen or less detrimental systems such as bacterial promoter fragments driven by antibiotics.72
Seeding of the cochlea with pluripotent stem cells has at least one major problem; that is, how to ensure that the cells take up residence specifically in the auditory sensory epithelium. A solution to this problem is, instead of inserting stem cells, to induce the remaining cells of the auditory epithelium to proliferate and differentiate into functional hair cells. To realize this goal, it is first important to understand the nature of the remaining cells in the auditory epithelium. After an ototoxic insult, hair cells die fairly rapidly, leaving differentiated supporting cells. Some attempts have been made, in animal models, to cause the transdifferentiation of these cells into sensory hair cells. This process, however, leads to a decrease in the number of supporting cells available and is only feasible for a short time following ototoxic insult.21 Over time, or with more severe injury, the supporting cells disappear, leaving behind a layer of flat epithelium. This layer will be the recipient of developing gene therapy.
Specifically, research strives to identify a gene that continues to be expressed in these cells despite the loss of hair cells and perhaps even the degeneration of supporting cells. Currently, the neurotrophin 3 (Ntf3) gene shows promise. Ntf3 is expressed primarily in supporting cells and is one of the neurotrophic factors required for sensory neuron survival in the cochlea.73 Moreover, expression of the neurotrophin Bdnf in undifferentiated cells of the organ of Corti is already known for Atoh1 null mice 60 and the microarray analysis of Atoh1 null mice also show some low level of expression of both Bdnf and Ntf3. Indeed, close examination of several genes show a topological expression in the sensory epithelium of E18.5 Atoh1 null mice (Figure 3), consistent with our previously published data.60 Of importance, Ntf3 expression continues in these cells into adult stages 3 and thus may be expressed in the cochlea after hair cell loss. Indeed, it is likely that this neurotrophin in not yet dedifferentiated or lost supporting cells is supporting the sensory neurons found to survive after hair cell loss. This idea is in line with the rapid loss of 85% of sensory neurons if this neurotrophin is lost in embryos or adults.3, 73
To analyze this question, the authors proposed and have developed the Pou4f3 Dreidel mutant mouse as an ideal model system for the cochlea in which hair cells are lost and only the supporting cells/flat epithelium remains (Figure 4). In the inner ear, the Pou4f3 protein is found only in sensory hair cells; and mice carrying a targeted deletion of the Pou4f3 gene show complete loss of auditory hair cells during the late embryonic and early postnatal period.65 Using in situ hybridization technique, it is feasible to look for Ntf3 expression in the remaining cells of the organ of Corti. If this gene is indeed still expressed in the remaining cells and continues to be expressed over time, it is an ideal candidate. The preliminary data using Ntf3 in situ hybridization show some very low levels of expression. This is currently being analyzed further using quantitative polymerase chain reaction and microarray of microdissected cochleae and vestibular sensory epithelia.
Once the Ntf3 or another gene is characterized, a genetically engineered DNA vector or vectors containing the genes described above for pluripotency and cell division (i.e., GATA3, Sox2, NANOG, and Oct4) and for hair cell differentiation (ATOH1, POU4F3, GFI1, and miRNA-183) can be generated using the promoter fragments from the Ntf3 gene to drive their expression. In this way, only cells in which Ntf3 is upregulated, namely the cells of the flat epithelium, will the genes to induce cell division and differentiation be expressed. Similarly, promoter fragments of such genes still expressed in sensory epithelia that have lost all hair cells can be used to initiate hair cell differentiation only in iPS cells seeded into the ear that happen to land on the sensory epithelia. Finding such genes and isolating their promoters to allow selective gene activation only in sensory epithelia is crucial for either approach.
One concern for the use of iPS in the treatment of human diseases is the length of time required for the development of the individual iPS lines. For acute illnesses such as spinal cord injury or myocardial infarction, it would not be practical to begin generation of the cell line following the injury.43 In contrast, adult-onset hearing loss is frequently a slowly progressing process. With early identification, possibly aided by screening for known and suspected genes that lead to hearing loss, iPS lines could be initiated well before hearing loss progressed to a state requiring treatment. Patients with more rapid hearing loss, such as that induced by ototoxic agents, would still likely benefit from iPS because, as discussed above, the sensory neurons survive long after the hair cells have died; and there is hope that even after the neurons die, new hair cells might be able to induce the growth of neurites to the organ of Corti through the production of neurotrophins.7, 22, 31
While the above outlined process will likely be able to generate hair cells in a flat organ of Corti, there are still some obstacles that must be addressed. Initially, it is critical not only to find a way to turn on proliferation, but also a means to turn it off. Without adequate cell cycle control, any of the techniques discussed above have the risk of forming tumors. Furthermore, because the cochlea is an enclosed space surrounded by bone, a tumor of the inner ear would likely destroy this delicate organ very rapidly as well as the adjacent facial nerve as in Schwannoma of the VIIIth nerve.
Another critical issue for restoration of hearing is the rebuilding of the appropriate cytoarchitecture required for organization and polarization of the hair cells within the organ of Corti. In a properly functioning cochlea, the hair cells are organized in a three and one radial configuration. Additionally, each individual hair cell has a direction, or polarity, defined by the kinocilia and stereocilia pattern on the apical surface. In order for the hair cells to be stimulated by sound waves in the basilar membrane, they must all align perpendicular to the long axis of the cochlea. Although it is known that genes such as Foxg1 are involved in determining the number of rows of hair cells and the polarity of hair cells in the organ of Corti,74 we do not yet have a firm grasp of how these features are defined in the normal cochlea.24 Clearly, organization of the cochlear hair cells will be, at least in part, governed by their surroundings. It has yet to be seen how much influence the infrastructure (such as the spiral artery) and gene expression from non-sensory cells adjacent to the organ of Corti will have on the developing hair cells, should the attempts to regenerate them outlined above be successful.
Although the use iPS cells may provide relief from the ethical concerns of using embryonic stem cells, the major ramifications of gene therapy must not be forgotten. Beyond somatic therapy, genetically reprogrammed iPS cells are also suitable for germline alteration when injected into blastocysts, thus providing the potential of correcting inherited diseases in future generations. With this technology, it is possible that DNA from genetically altered stem cells could be passed on to future generations.75 It is, therefore, also possible that gene therapy may progress into genetic design, where one will be able to pass on desired traits and refrain from passing on less desirable ones. This may reduce the genetic burden on future generations but it will also limit the variability of the gene pool in the population. Moving forward and manipulating germlines without absolute knowledge of the underlying transcription factors, would mean playing the “Sorcerer’s apprentice” and running the risk of permanently altering human species.
Fortunately, for the treatment of hearing loss, the ear is a closed system, surrounded by bone and therefore fairly self-contained. Using iPS cells to regenerate hair cells and restore hearing will not interfere with the patient’s germline. Thus, any gene therapy implemented will remain in that patient and will not permanently modify the genome or affect future generations.
Parts of this investigation were conducted in a facility constructed with support from Research Facilities Improvement Program Grant from the National Center for Research Resources, National Institutes of Health. We acknowledge the use of the confocal microscope facility of the NCCB, supported by EPSCoR EPS-0346476 (CFD 47.076)
Fundings. This work was supported by grants from the National Institute of Health USA [RO1 DC 005590 (BF); DC005009 (KB)].