Our results constitute the first report on the morphological characterization of the human MSO nucleus, component of the SOC known as an important site of convergence of auditorium information from both ears (Moore,
1987; Heffner and Masterton,
1990; Schofield,
2002), in perinatal life and on the pathological aspects of this peculiar structure in SIUDS/SIDS.
Studies on the pathophysiology of SIDS have focused on the brainstem of infants with important results pointing to a reflexogenic lethal mechanism (Rossi and Matturri,
1995). Reflexogenic disorders can in fact lead to inhibitory action on the sinoatrial node with possible cardiac arrest. In particular, a subset of SIDS from sudden loud noises (cardioauditory death) has been reported by Guntheroth (
1995). We can suppose in these cases the presence of alterations of brainstem nuclei involved in auditory functions, above all of the MSO, given the high sensitivity of its neurons even to very low-frequency sounds (Yin and Chan,
1990).
The most prominent observation emerged from this study was the loss of neurons and fibers in the MSO region in high percentages of suddenly dead victims.
Nevertheless alterations of the MSO are been reported also in autism, a neurological developmental disorder characterized by cognitive and communicative impairments and restricted repetitive behaviors (Lord et al.,
2000; Rice,
2009), in addition to some degree of hearing deficits (Rosenhall et al.,
1999; Kellerman et al.,
2005).
Kulesza and Mangunay (
2008) reported significant differences in the cytoarchitecture of the MSO between autistic individuals and controls, mainly represented by smaller round neurons with shorter dendrites, similar to the MSO neurons found by us in SIUDS and SIDS cases. They interpret the morphological alterations in autistic MSO as related to disfunctions of the auditory pathway. However, we suppose an involvement of the MSO in broader functions than those related to hearing.
Interesting, Rodier et al. (
1996) observed a near-complete absence of the superior olive and of the facial nucleus in the brainstem of a 21-year-old autistic woman. This neuropathologic pattern recalls the one present in almost half of the SIUDS victims. In fact, in our study, MSO dysgenesis was frequently associated with hypoplasia of the F/PFc, in addition to the hypodevelopment of the contiguous RTN.
So, we believe that these three adjacent pontine structures are involved in the same pattern of neuronal development. This idea is based on the works of Altman and Bayer (
1980) and Chisaka et al. (
1992), showing that the facial nucleus and the superior olive not only form at the same embryological stage, but also derive from the same rhombomeres, precisely the fourth and fifth rhombomers. Accordingly, Carpenter et al. (
1993) showed that a complete failure of formation of these rhombomers cause lack of the facial nucleus and superior olive in transgenic knockout mice.
Evident demonstrations don't exist in literature that the RTN arises from the same rhombomers. Different investigators refer, however, that neurons located around the facial motor nucleus, called “parafacial respiratory group” (pFRG), together with neurons of the confined RTN, serve as a single pre-inspiratory and/or expiratory-modulating complex in newborns (Onimaru and Homma,
2003; Janczewski and Feldman,
2006). Then it is plausible the involvement of the RTN in the ontogenetic process of the parafacial complex. Moreover, experimental studies on Phox2b expression in rats provide genetic evidence for a similar embryological origin of the two nuclei (Dubreuil et al.,
2008).
Since both the F/PFc and the RTN have an essential role in the respiratory rhythm-generating circuit and in maintenance of breathing, we think that also the MSO could have a share in the control of the ventilator activity, in addition to hearing. Our hypothesis is supported by genetic studies showing that the key player in the development of this rhombomeric region of the hindbrain is the proneural transcription factor
mouse atonal homolog 1 (
Math1). Mice lacking
Math1 lose multiple components of the arousal systems and die shortly after birth from an apparent inability to initiate respiration (Ben-Arie et al.,
2000; Wang et al.,
2005; Rose et al.,
2009).
So, also the alterations of the MSO observed by us in this study and those reported in autism may indicate a disruption of the excitatory/inhibitory eupneic inputs reaching this nucleus, with different degree of severity and pathological consequences, even leading to death in a vulnerable period of perinatal development.
The presence of the triad of abnormalities (MSO, F/PFc, and RTN hypoplasia) confined to victims of sudden fetal death leads to wonder whether breathing alterations can cause death during intrauterine life. It is known that the respiratory-related neuronal network is active before birth checking ventilatory-like rhythmic movements in the mammalian fetus (Barcroft and Barron,
1937; Boddy and Dawes,
1975). Therefore, it becomes vital only after birth. Thus, defects of this occasional respiratory activity in prenatal life would be not sufficient to justify the fetal death. One possibility is that these neuronal structures participate not only in breathing but, more extensively, are essential to the control of all the vital functions.
In autism the MSO disorders are frequently associated with additional anomalies of the brainstem and cerebellum. Increased cell packing density, cell body size decrease with limited dendritic arbor, reduced binding of GABA and nicotinic receptors in different brainstem nuclei and regions and decreased number of cells in cerebellar cortex layers have been reported (Bauman and Kemper,
1985,
2005; Gaffney et al.,
1987; Palmen et al.,
2004; Kulesza and Mangunay,
2008; Schmitz and Rezaie,
2008). In particular, a reduction in the number of Purkinje cells is considered one of the most consistent pathological feature of autism (Gaffney et al.,
1987; Arin et al.,
1991; Bauman and Kemper,
1996; Whitney et al.,
2008).
The high incidence of Purkinje cell loss in cerebellar cortex and in particular the high correlation with the MSO dysmorphology represents an additional point of convergence between sudden perinatal death and autistic disorders. In addition it enhances our assumption of a MSO involvement in breathing, given the fundamental role of the Purkinje cells in eupneic ventilation (Xu et al.,
2004).
As we did in our previous studies (Lavezzi et al.,
2005,
2007,
2010,
2012a,
b), we ascribe once again the neuropathological findings of this study to smoke absorption in fetal life. Prenatal chronic nicotine exposure primarily alters the development of fetal cholinergic and other neurotransmitter systems, so potentially affecting brainstem centers critical to cardiorespiratory and, more generally, all the autonomic functions control (Duncan et al.,
2009). The noxious role of multiple neurotoxins in cigarette smoking during and after pregnancy should be taken into account also in the etiology of autism for which currently underlying biological causes remain to be established.
The conclusion of this work is that considerable remodeling takes place in the MSO of sudden death victims, in addition to neurological alteration of brainstem and cerebellum previously highlighted by us (Matturri et al.,
2002; Lavezzi et al.,
2004,
2010,
2012a,
b; Matturri and Lavezzi,
2007; Lavezzi and Matturri,
2008a). This leads to attribute to the MSO more important functions than those well-known related to hearing. However, this hypothesis should be tested in a larger sample of SIUDS/SIDS for a more detailed understanding of the pathogenesis and neurobiology of these syndromes, and ultimately to more effective preventive interventions, given the relationship with prenatal smoke absorption here reported.