Twenty years have passed since the discovery of hair cell regeneration in birds (Corwin & Cotanche, 1988; Ryals & Rubel, 1988). The initial excitement caused by this discovery has been followed by steady progress in understanding the fundamental mechanisms that recently culminated in research evidence of hair cell regeneration in both the auditory and vestibular portions of the mammalian inner ear (Kawamoto et al., 2003; Izumikawa et al., 2005; Staecker et al., 2007).
Clinical audiologists are faced with the responsibility of translating these basic science findings into potential patient application. They raise important questions: When will hair cell regeneration be a reality for my patients? What will be the mea-sures of candidacy? What will the impact of hair cell regeneration be in my patients who use or are candidates for hearing aids or other amplification devices? Will hearing aids or cochlear implants continue to be needed in the face of hair cell regeneration?
Unfortunately, specific answers to these questions are not yet possible. However, the more audiologists know about current state of hair cell regeneration research, the more they can help patients understand how that research can influence their future care. The following are some of the latest findings and techniques involved in the study of hair cell regeneration—and their clinical implications.
Building a Hair Cell
Two different mechanisms come into play when hair cells are replaced after injury in birds and other non-mammals: cell division (mitosis) and transdifferentiation. These two mechanisms are illustrated in Figure 1 [PDF]. In the case of mitosis (Figure 1a [PDF]) the supporting cells regenerate by cell division and differentiation. This process involves two steps: 1) the cell divides, duplicating its DNA, and becomes two undifferentiated cells, and 2) at least one of these cells is stimulated to differentiate into a hair cell.
This two-step process is seen during early embryonic development of the inner ear. In adult birds, supporting cells in the inner ear retain this capacity, but in mammals this capacity is under strict genetic control and does not automatically occur after injury.
The exact stimulus for this re-entry into the cell cycle has been the subject of many studies on hair cell regeneration over the last 20 years. Understanding and manipulating cell cycle regulation and control is one of three primary approaches scientists use to initiate hair cell regeneration in mammals. These studies have been greatly aided by investigations of the basic molecular and genetic mechanisms involved in early development. Scientists reason that understanding how these early developmental signals work will be very useful in understanding how we might recapitulate development after injury.
Transdifferentiation is a mechanism in which supporting cells convert to hair cells without going through cell division. Figure 1b [PDF] shows migration of a supporting cell to the basement membrane followed by differentiation into a hair cell at the luminal surface.
One advantage of this mechanism is that it only involves only one step. Cells need not replicate DNA, grow, and form duplicate cells; they need only to convert from one cell type to another. Although the transdifferentiation of supporting cells may seem to be a simpler way of replacing dead hair cells, there is a downside. If supporting cells transdifferentiate to hair cells, then what would replace the supporting cell? Would structural alterations in the organ of Corti caused by the loss of supporting cells significantly alter recovery of function?
Understanding the genetic determinants of cell transdifferentiation is critical in using this mechanism for future hair cell regeneration therapies in mammals or humans. Two specific techniques have employed transdifferentiation of either resident or non-resident cells to replace lost hair cells in the mammalian ear: gene therapy and stem cells.
Cell Cycle Controls
Several lines of evidence show that cells in the mammalian cochlea can be induced to divide and form new hair cells (White et al., 2006, and review by Breuskin et al., 2008; Stone & Cotanche, 2007). So why don't they spontaneously do this when hair cells are injured?
Many molecular and genetic controls are involved in moving a cell through the cell cycle to form a new cell (for a more thorough discussion of cell cycle and hair cell regeneration see Ryals, Matsui, & Cotanche, 2007). Some cell cycle factors are growth factors, mitogenes, cyclins, cyclin-dependent kinases, and tumor suppressor genes. Each has been the topic of research related to its application to hair cell regeneration. For example, several studies have analyzed the potential role of growth factors on initiating re-entry into the cell cycle and promoting cell division. Such studies show that even when growth factors are present, supporting cells of the mammalian cochlea do not spontaneously re-enter the cell cycle.
This finding reinforces the idea that some factor is present in the mammalian cochlea that prohibits re-entry even when stimulated. One such genetic factor has been identified: p27kip1 (Lowenheim et al., 1999; Chen & Segil, 1999). This gene is in a family of genes called "tumor suppressors." As the name implies, these genes are important regulators of cell division, and their presence usually means that cells are strongly inhibited from continuing division. Investigators found that the absence of the p27kip1 gene in the inner ear resulted in an overproduction of hair cells in the mouse cochlea. This abundance of hair cells results in significant hearing loss. Too many hair cells seem to be just as bad as too few.
Other genes in this family, such as p19/ink4d and Rb1, are also important in regulating cell division and hair cell production. Investigators have shown that the presence of these genes can result in proliferation of hair cells in the adult mammalian cochlea, but that these cells die shortly after they are produced. So if scientists are to stimulate cells in the mammalian cochlea to re-enter the cell cycle, they need to understand not only how to manipulate the inhibitory genetic controls that regulate the number of hair cells produced, but also how to keep these cells alive and functional.
After cell division, the second critical step in building a new hair cell is to provide the appropriate signals that will instruct it to become a hair cell. Two excellent review articles discuss mechanisms that guide hair cell fate and organization of the cochlea (Bryant et al., 2002; Kelley, 2002). These mechanisms involve transcription factors that bind to specific parts of DNA and control the transfer of genetic information from DNA to RNA. Simply put, transcription factors provide the "code" that tells the cell what to do. Atoh1 (previously termed Math1) is a transcription factor that signals undifferentiated cells to become hair cells. In animals missing Atoh1, hair cells do not develop (Bermingham et al., 1999). Atoh1 is present in the developing cochlea but is "turned off" in adulthood.
Any future therapy attempting to replace hair cells in the cochlea will need to be aware that Atoh1 must be present or "turned on" for the cells to become hair cells. Sox2 is another transcription factor involved in hair cell fate. The identification of these factors has helped us come closer to replacing hair cells in the mammalian cochlea. The Atoh1 transcription factor has already been used, in a gene therapy approach, to replace hair cells in the cochleae of guinea pigs deafened with ototoxic drugs.
Whether a cell becomes a hair cell or a supporting cell is also influenced by local signals where neighboring cells prevent the overproduction of a particular cell type. Typically this signaling pathway is turned on when adjacent cells die. Understanding this pathway will not only help us ensure that cells produced after hair cell death become hair cells, but will also help control the appropriate number and type of cells produced. This cell-cell pathway involves a receptor protein (Notch) that signals the transcription factors Hes1and Hes5. Activation of this pathwayresults in the correct number of sensory and nonsensory cells (Kelley, 2002; Zine & de Ribaupierre, 2002).
Several investigators have now shown that this pathway plays a role in hair cell replacement in birds, zebrafish, and potentially in mammals (for reviews, see Stone & Cotanche, 2007; Breuskin et al., 2008). Further understanding of the controls regulating the initiation and cessation of this cell-cell signaling will be crucial to future therapeutic approaches for hair cell replacement in the mammalian cochlea. As previous studies have shown, we need the right type and also the appropriate number of cells in the cochlea in order to restore hearing function.
Gene therapy is an experimental treatment that involves introducing genetic material into a person's cells to fight disease (see the Human Genome Project Web site). Briefly, a gene is delivered to a cell using a carrier known as a "vector." The most common types of vectors used in gene therapy are viruses. To use this approach in the inner ear, investigators first created hair cell loss by using an ototoxic drug and then inserted the Atoh1 gene into remaining cells of the inner ear through a viral vector. These investigations showed that new hair cells can be created by stimulating remaining supporting cells to transdifferentiate into hair cells in both the mammalian cochlea and vestibular system (Kawamoto et al., 2003; Izumikawa et al., 2005; Staecker et al., 2007). These authors have shown evidence that the newly produced hair cells provide some return of functional hearing and balance. More recently investigators have been able to use a similar gene transfer approach in utero to produce new hair cells (Gubbles et al., 2008, see Figure 2 [PDF]). This approach is a major step toward the possibility of gene therapy for congenital deafness.
These important findings are providing the impetus for continuing study and replication in many laboratories across the country. Although gene therapy has been shown to be a remarkably promising technique, there are several limitations that must be more fully explored before this technique can be considered in humans (see sidebar).
Stem Cell Therapy
Stem cells have received considerable attention over the last several years because of their potential to repair or regenerate damaged cells and tissues in the human body. Because these cells have the potential to become any cell type, investigators have proposed their use to replace inner ear hair cells lost to injury, disease, or aging. Embryonic stem cells maintain the ability to become any cell type; other non-embryonic stem cells maintain the ability to become varied cell types along a more restricted line (i.e., neural stem cells can become not only neurons, but also neural supporting cells like glia).
Early studies showed that adult utricular cells retained stem cell capabilities and, when transplanted to a developing chick otocyst, formed hair cells in the inner ear (Li & Heller, 2003). More recent studies, summarized by Martinez-Mondedero and Edge (2007), have shown that cells in both the early developing cochlea and more mature vestibular system retain stem cell properties and have the potential for use in replacing hair cells and neurons in the adult inner ear.
Current research strategies are directed toward stem cell transplants to restore or replace degenerating neurons or hair cells. Many studies focus on the use of stem cells to restore auditory nerve fibers or spiral ganglion cells; success in this area might have a much shorter translation to clinical application through the use of cochlear implant technology. Stem cells are excellent candidatesfor this therapy, as they have the potential to increase thenumber of auditory neurons available for electrical excitationvia a cochlear implant.
The success of any therapy to replace hair cells or neurons will require the following critical steps:
Conversion of embryonic stem cells to neurectodermprogenitors (in the case of inner ear progenitors this step may not be necessary)
Directed differentiation of progenitorsinto auditory neurons or hair cells
Delivery of cells into their targetsite within the delicate cochlea
Successful integrationof transplanted cells with existing neural populations
Several stem cell types have been delivered into the mammaliancochlea for the replacement of auditory neurons, including bonemarrow stem cells, neural stem cells, inner ear stem cells, and embryonic stem cells. Investigators have found evidence that these cell types can integrate into the target site (although they don't always do this), express neuronal markers, and grow neuronal processes. Some reports show that these cells can survive for up to 13 weeks (Coleman et al., 2007). Although such findings illustrate thepromise of this therapy, many questions remain unanswered and further investigation in animal models will be necessary before this technique can be used in humans (see sidebar).
We are much closer to human hair cell regeneration therapy than 20 years ago when the first research was published. We now know that it is possible to restore inner ear hair cells in mature mammals. That is a huge step forward, but not big enough to begin regenerating hair cells in humans.
As these therapies are developed, clinical audiologists will continue to play a critical role. Candidacy for hair cell regeneration therapy must be evaluated in terms of the underlying genetic, anatomical, and biochemical mechanisms that contribute to the etiology of the hearing loss. Therefore, accurate and specific diagnostic assessments must be developed for multiple etiologies underlying sensorineural hearing loss. Audiologists will be called upon to provide those diagnostic measures.
Some audiologists have wondered about the impact of future regenerative therapies on patients who have been treated with amplification devices or cochlear implants. Will hair cell regeneration eliminate the need for hearing aids and cochlear implants? The short answer is that is unlikely. If the potential for safely initiating hair cell regeneration in humans does become a possibility, the first attempts will probably result in incomplete replacement of all structures necessary for normal hearing.
Even incomplete replacement, however, could be very good news for patients with sensorineural hearing loss and for audiologists. Patients with difficult-to-fit hearing losses (such as those with a very severe-to-profound loss, a corner audiogram, or a severe-to-profound sharply sloping high-frequency hearing loss) could benefit greatly from the restoration of just a few hair cells. Adding a few hair cells connected to the appropriate neural fibers might increase sensitivity to the point at which amplification can be useful. Increasing the number of auditory neurons that can be stimulated by a cochlear implant could potentially improve CI performance and lower stimulation requirements.
It seems hard to imagine that improved medical treatment of sensorineural hearing loss through therapies such as hair cell regeneration can be anything but positive for audiologists and the patients they treat. Further research into the basic cellular, molecular, and genetic mechanisms underlying hair cell regeneration—with the goal of translating research into clinical practice—holds promise for all patients with what was once thought to be permanent
sensorineural hearing loss.