Inner ear lesions in congenital cytomegalovirus infection of human fetuses.

Congenital cytomegalovirus (CMV) infection is the leading cause of non-hereditary congenital sensorineural hearing loss (SNHL). The natural course and the pathophysiology of inner ear lesions during human fetal CMV infection have not yet been reported. Inner ear lesions were investigated in six CMV-infected fetuses aged 19-35 postconceptional weeks and correlated with central nervous system (CNS) lesions. All the fetuses had high viral loads in the amniotic fluid and severe visceral and CNS lesions visible by ultrasound. Diffuse lesions consisting of both cytomegalic cells containing inclusion bodies and inflammation were found within all studied structures including the inner ear, brain, other organs, and placenta, suggesting hematogenous dissemination. Cochlear infection was consistently present and predominated in the stria vascularis (5/6), whereas the supporting cells in the organ of Corti were less often involved (2/6). Vestibular infection, found in 4/6 cases, was florid; the non-sensory epithelia, including the dark cells, were extensively infected. The endolymphatic sac was infected in 1 of 3 cases. The severity of inner ear infection was correlated with the CNS lesions, confirming the neurotropism of CMV. This study documenting infection of the structures involved in endolymph secretion and potassium homeostasis in fetuses with high amniotic fluid viral loads suggests that potassium dysregulation in the endolymphatic compartment of the inner ear may lead to secondary degeneration of the sensory structures. In addition, the occurrence of SNHL depends on the intensity and duration of the viral infection and inflammation.

Viral vector tropism for supporting cells in the developing murine cochlea.

Gene-based therapeutics are being developed as novel treatments for genetic hearing loss. One roadblock to effective gene therapy is the identification of vectors which will safely deliver therapeutics to targeted cells. The cellular heterogeneity that exists within the cochlea makes viral tropism a vital consideration for effective inner ear gene therapy. There are compelling reasons to identify a viral vector with tropism for organ of Corti supporting cells. Supporting cells are the primary expression site of connexin 26 gap junction proteins that are mutated in the most common form of congenital genetic deafness (DFNB1). Supporting cells are also primary targets for inducing hair cell regeneration. Since many genetic forms of deafness are congenital it is necessary to administer gene transfer-based therapeutics prior to the onset of significant hearing loss. We have used transuterine microinjection of the fetal murine otocyst to investigate viral tropism in the developing inner ear. For the first time we have characterized viral tropism for supporting cells following in utero delivery to their progenitors. We report the inner ear tropism and potential ototoxicity of three previously untested vectors: early-generation adenovirus (Ad5.CMV.GFP), advanced-generation adenovirus (Adf.11D) and bovine adeno-associated virus (BAAV.CMV.GFP). Adenovirus showed robust tropism for organ of Corti supporting cells throughout the cochlea but induced increased ABR thresholds indicating ototoxicity. BAAV also showed tropism for organ of Corti supporting cells, with preferential transduction toward the cochlear apex. Additionally, BAAV readily transduced spiral ganglion neurons. Importantly, the BAAV-injected ears exhibited normal hearing at 5 weeks of age when compared to non-injected ears. Our results support the use of BAAV for safe and efficient targeting of supporting cell progenitors in the developing murine inner ear.

Non-cell-autonomous planar cell polarity propagation in the auditory sensory epithelium of vertebrates.

Sensory epithelia of the inner ear require a coordinated alignment of hair cell stereociliary bundles as an essential element of mechanoreceptive function. Hair cell bundle alignment is mediated by core planar cell polarity (PCP) proteins, such as Vangl2, that localize asymmetrically to the circumference of the cell near its apical surface. During early phases of cell orientation in the chicken basilar papilla (BP), Vangl2 is present at supporting cell junctions that lie orthogonal to the polarity axis. Several days later, there is a striking shift in the Vangl2 pattern associated with hair cells that reorient towards the distal (apical) end of the organ. How the localization of PCP proteins transmits planar polarity information across the developing sensory epithelium remains unclear. To address this question, the normal asymmetric localization of Vangl2 was disrupted by overexpressing Vangl2 in clusters of cells. The BP was infected with replication-competent retrovirus encoding Vangl2 prior to hair cell differentiation. Virus-infected cells showed normal development of individual stereociliary bundles, indicating that asymmetry was established at the cellular level. Yet, bundles were misoriented in ears infected with Vangl2 virus but not Wnt5a virus. Notably, Vangl2 misexpression did not randomize bundle orientations but rather generated larger variations around a normal mean angle. Cell clusters with excess Vangl2 could induce non-autonomous polarity disruptions in wild-type neighboring cells. Furthermore, there appears to be a directional bias in the propagation of bundle misorientation that is towards the abneural edge of the epithelium. Finally, regional bundle reorientation was inhibited by Vangl2 overexpression. In conclusion, ectopic Vangl2 protein causes inaccurate local propagation of polarity information, and Vangl2 acts in a non-cell-autonomous fashion in the sensory system of vertebrates.

Induction of cytomegalovirus-infected labyrinthitis in newborn mice by lipopolysaccharide: a model for hearing loss in congenital CMV infection.

Congenital cytomegalovirus (CMV) infection is the most common infectious cause of sensorineural hearing loss in children. Here, we established an experimental model of hearing loss after systemic infection with murine CMV (MCMV) in newborn mice. Although almost no viral infection was observed in the inner ears and brains by intraperitoneal (i.p.) infection with MCMV in newborn mice, infection in these regions was induced in combination with intracerebral (i.c.) injection of bacterial lipopolysaccharide (LPS). The susceptibility of the inner ears was higher than that of the brains in terms of viral titer per unit weight. In the labyrinths, the viral infection was associated with the mesenchymal vessels and accompanied by inflammatory cells induced by LPS, causing hematogenous targets of infection in the labyrinths. Viral infection also spread in the perilymph regions such as the scala tympani and scala vestibuli, probably from infected brains via meningogenic and cochlear nerve routes. Viral infection was not observed in the scala media in the endolymph, including the Corti organ. However, viral infection was observed in the spiral limbus, including the stria vascularis. These results suggest that hearing loss caused by labyrinthitis after congenital CMV infection may be enhanced by inflammation caused by systemic bacterial infection in the neonatal period.

A glucocorticoid reduces adverse effects of adenovirus vectors in the cochlea.

Gene transfer using a recombinant adenovirus is a powerful tool for research and clinical applications, but its cytotoxicity and immune response limit its use, especially when repeated application of the vector is necessary. This study investigated the effects of dexamethasone (DEX)-induced immunosuppression on the outcome of adenovirus gene transfer in guinea pig inner ears. Animals received DEX for 29 days. Their inner ear was inoculated with 5 micro l of adenovirus vector twice, on days 5 and 26. Auditory brainstem response was measured on days 1, 8 and 29. The animals were sacrificed on day 29, and reporter gene expression was evaluated. In control animals that received no DEX, postinoculation threshold shifts and lesions in the organ of Corti were observed and reporter gene expression was absent. In contrast, DEX-treated ears were largely protected, and transduction of inner ear cells was readily apparent. These data demonstrate that immunosuppressive treatment can reduce the negative consequences of repeated adenovirus-mediated gene therapy.