A three-component model of the spinal nerve ramification: Bringing together the human gross anatomy and modern Embryology.

Due to its long history, the study of human gross anatomy has not adequately incorporated modern embryological findings; consequently, the current understanding has often been incompatible with recent discoveries from molecular studies. Notably, the traditional epaxial and hypaxial muscle distinction, and their corresponding innervation by the dorsal and ventral rami of the spinal nerve, do not correspond to the primaxial and abaxial muscle distinction, defined by the mesodermal lineages of target tissues. To resolve the disagreement between adult anatomy and embryology, we here propose a novel hypothetical model of spinal nerve ramification. Our model is based on the previously unknown developmental process of the intercostal nerves. Observations of these nerves in the mouse embryos revealed that the intercostal nerves initially had superficial and deep ventral branches, which is contrary to the general perception of a single ventral branch. The initial dual innervation pattern later changes into an adult-like single branch pattern following the retraction of the superficial branch. The modified intercostal nerves consist of the canonical ventral branches and novel branches that run on the muscular surface of the thorax, which sprout from the lateral cutaneous branches. We formulated the embryonic branching pattern into the hypothetical ramification model of the human spinal nerve so that the branching pattern is compatible with the developmental context of the target muscles. In our model, every spinal nerve consists of three components: (1) segmental branches that innervate the primaxial muscles, including the dorsal rami, and short branches and long superficial anterior branches from the ventral rami; (2) plexus-forming intramural branches, the serial homolog of the canonical intercostal nerves, which innervate the abaxial portion of the body wall; and (3) plexus-forming extramural branches, the series of novel branches located outside of the body wall, which innervate the girdle and limb muscles. The selective elaboration or deletion of each component successfully explains the reasoning for the standard morphology and variability of the spinal nerve. Therefore, our model brings a novel understanding of spinal nerve development and valuable information for basic and clinical sciences regarding the diverse branching patterns of the spinal nerve.

Motor neurons with limb-innervating character in the cervical spinal cord are sculpted by apoptosis based on the Hox code in chick embryo.

In the developing chick embryo, a certain population of motor neurons (MNs) in the non-limb-innervating cervical spinal cord undergoes apoptosis between embryonic days 4 and 5. However, the characteristics of these apoptotic MNs remain undefined. Here, by examining the spatiotemporal profiles of apoptosis and MN subtype marker expression in normal or apoptosis-inhibited chick embryos, we found that this apoptotic population is distinguishable by Foxp1 expression. When apoptosis was inhibited, the Foxp1+ MNs survived and showed characteristics of lateral motor column (LMC) neurons, which are of a limb-innervating subtype, suggesting that cervical Foxp1+ MNs are the rostral continuation of the LMC. Knockdown and misexpression of Foxp1 did not affect apoptosis progression, but revealed the role of Foxp1 in conferring LMC identity on the cervical MNs. Furthermore, ectopic expression of Hox genes that are normally expressed in the brachial region prevented apoptosis, and directed Foxp1+ MNs to LMC neurons at the cervical level. These results indicate that apoptosis in the cervical spinal cord plays a role in sculpting Foxp1+ MNs committed to LMC neurons, depending on the Hox expression pattern.

Elucidation of target muscle and detailed development of dorsal motor neurons in chick embryo spinal cord.

The avian cervical spinal cord includes motoneurons (MNs) that send their axons through the dorsal roots. They have been called dorsal motoneurons (dMNs) and assumed to correspond to MNs of the accessory nerve that innervate the cucullaris muscle (SAN-MNs). However, their target muscles have not been elucidated to date. The present study sought to determine the targets and the specific combination of transcription factors expressed by dMNs and SAN-MNs and to describe the detailed development of dMNs. Experiments with tracing techniques confirmed that axons of dMNs innervated the cucullaris muscle. Retrogradely labeled dMNs were distributed in the ventral horn of C3 and more caudal segments. In most cases, some dMNs were also observed in the C2 segment. It was also demonstrated that SAN-MNs existed in the ventral horn of the C1-2 segments and the adjacent caudal hindbrain. Both SAN-MNs and dMNs expressed Isl1 but did not express Isl2, MNR2, or Lhx3. Rather, these MNs expressed Phox2b, a marker for branchial motoneurons (brMNs), although the intensity of expression was weaker. Dorsal MNs and SAN-MNs were derived from the Nkx2.2-positive precursor domain and migrated dorsally. Dorsal MNs remain in the ventral domain of the neural tube, unlike brMNs in the brainstem. These results indicate that dMNs and SAN-MNs belong to a common MN population innervating the cucullaris muscle and also suggest that they are similar to brMNs of the brainstem, although there are differences in Phox2b expression and in the final location of each population. J. Comp. Neurol. 521: 2987-3002, 2013. © 2013 Wiley Periodicals, Inc.

Neurogenin2 expression together with NeuroM regulates GDNF family neurotrophic factor receptor α1 (GFRα1) expression in the embryonic spinal cord.

In many regions of the nervous system, the combinatorial action of transcriptional factors specifies the individual fate of neuronal subtypes. Contrary to this, we report that a single transcriptional factor controls a phenotype shared by different subtypes of neurons, namely the expression of a neurotrophic factor receptor in the spinal cord. Along the dorsoventral axis of the chick embryo spinal cord, the expression pattern of a specific receptor for glial cell line derived-neurotrophic factor (GDNF family of receptors α1: GFRα1) was related to that of two basic helix-loop-helix (bHLH) transcriptional factors (NeuroM and Neurogenin2: Ngn2). In ovo electroporation in the chick embryo revealed that the overexpression of NeuroM alone was sufficient to induce ectopic GFRα1 expression without overt neuronal differentiation, whereas the suppression of NeuroM activity resulted in the specific loss of GFRα1 expression, indicating that NeuroM may act as a differentiation factor for GFRα1 expression. Ngn2 overexpression was also sufficient to induce precocious GFRα1 expression. However, the forced expression of both obligate suppressor and activator forms of Ngn2 also induced aberrant GFRα1 expression. Thus, any deviation from an optimum level of Ngn2 expression resulted in aberrant GFRα1 expression. Consistent with this, manipulation of Ngn2 expression levels by other bHLH factors also resulted in ectopic GFRα1 expression. For example, the downregulation by Ascl1 and the upregulation by Ptf1a induced ectopic GFRα1 expression, irrespective of endogenous expression patterns of Ascl1 and Ptf1a (Ascl1/Ptf1) in the spinal cord. The suppression of Ascl1/Ptf1a activities abolished Ngn2 and GFRα1 expression, even in Ascl1/Ptf1a-negative regions. These data indicate the presence of a distinct regulatory sequence for a determinant of GFRα1 expression, in which Ascl1/Ptf1a may competitively intervene to stochastically modulate default Ngn2 expression levels. Thus, Ngn2 together with NeuroM serves as readout to regulate GFRα1 expression, which occurs in multiple subtypes of spinal neurons.

Expression pattern of LRR and Ig domain-containing protein (LRRIG protein) in the early mouse embryo.

The combination of leucine-rich repeat (LRR) and immunoglobulin-like (Ig) domains is found in the domain architecture of the Trk neurotrophin receptor protein. Recently dozens of such proteins simultaneously carrying LRR and Ig domains as the Trk receptors have been identified. Given the significant biological roles of Trk and such newly identified proteins, we have searched the public database for human proteins with LRR and Ig domains (collectively termed the leucine-rich repeat and Ig domain-containing protein, LRRIG protein, in this study), and have analyzed the mRNA expression pattern of mouse orthologs of obtained human LRRIG proteins at embryonic day 10. The list of the LRRIG proteins includes 36 human proteins: four LINGO, three NGL, five SALM, three NLRR, three Pal, two ISLR, three LRIG, two GPR, two Adlican, two Peroxidasin-like proteins, three Trk neurotrophin receptors, a yet unnamed protein AAI11068, and three AMIGO. Some molecules (LINGO2, LINGO4, NGL1, SALM1, SALM5, and TrkB) were expressed exclusively in neuronal tissues, whereas others (ISLR1, GPR124, and Adlican2) exhibited non-neuronal expression profiles. However, the majority of LRRIG protein family exhibited broad mRNA tissue-expression profiles.

Ficolin A and ficolin B are expressed in distinct ontogenic patterns and cell types in the mouse.

Ficolins are a group of proteins characterized by the presence of collagen-like and fibrinogen-like domains. Two of three human ficolins, L-ficolin and H-ficolin, are serum lectins that form complexes with mannose-binding lectin-associated serine proteases (MASPs) and play important roles in the lectin complement pathway. The other human ficolin, M-ficolin, is a non-serum-type ficolin that is expressed in monocytes. Little is known about the physiological roles of ficolins. In this study, we delineated the ontogeny and cell types that express ficolins in mice. RT-PCR analysis showed that the expression pattern of ficolin A expression was closely similar to that of Masps, suggesting that these molecules may function in coordination as components of the lectin complement pathway. The cell types that express ficolin A mRNA in both adult liver and spleen were identified as macrophages by in situ hybridization. Ficolin B exhibited a distinct ontogeny pattern that switched from embryonic liver to postnatal bone marrow and spleen. The cells that express ficolin B mRNA were identified as belonging to the myeloid cell lineage by magnetic sorting and by subsequent RT-PCR in bone marrow cells. Thus, the different spatial-temporal expression patterns of ficolins A and B suggest that these molecules play distinct roles in the prenatal and postnatal stages.

Differential expression of the GDNF family receptors RET and GFRalpha1, 2, and 4 in subsets of motoneurons: a relationship between motoneuron birthdate and receptor expression.

Previous studies have demonstrated the expression of specific members of the glial cell line-derived neurotrophic factor (GDNF) receptor family alpha (GFRalpha) in subsets of motoneurons (MNs) in the developing mouse spinal cord. We examined the expression pattern of GFRalpha and RET in the avian lumbar spinal cord during the period of programmed cell death (PCD) of MNs by using double labeling in situ hybridization and immunohistochemistry. In the lateral motor column (LMC) of the lumbar spinal cord, a laminar organization of GFRalpha expression was observed: GFRalpha1-positive MNs were located in the medial LMC; GFRalpha1-, 2-, and 4-positive MNs were situated in the lateral LMC; and GFRalpha4-positive MNs were located in the intermediate LMC. The species of GFRalpha receptor that was expressed in MNs was found to be related to their birthdates. The expression of subpopulation-specific transcriptional factors was also used to define MNs that express a specific pattern of GFRalpha. This analysis suggests that motor pools as defined by these transcriptional factors have unique expression patterns of GFRalpha receptor. Early limb bud ablation did not affect the expression of GFRalpha in the spinal cord, indicating that regulation of receptor expression is independent of target-derived signals. Finally, GDNF mRNA expression was found in the limb during the PCD period of MNs. In conclusion, these results indicate that time of withdrawal from the mitotic cycle may specify the expression pattern of GFRalpha in subsets of MNs and that GDNF may function as a target-derived neurotrophic factor for specific subpopulations of MNs.

Zic1 promotes the expansion of dorsal neural progenitors in spinal cord by inhibiting neuronal differentiation.

The role of Zic1 was investigated by altering its expression status in developing spinal cords. Zic genes encode zinc finger proteins homologous to Drosophila Odd-paired. In vertebrate neural development, they are generally expressed in the dorsal neural tube. Chick Zic1 was initially expressed evenly along the dorsoventral axis and its expression became increasingly restricted dorsally during the course of neurulation. The dorsal expression of Zic1 was regulated by Sonic hedgehog, BMP4, and BMP7, as revealed by their overexpressions in the spinal cord. When Zic1 was misexpressed on the ventral side of the chick spinal cord, neuronal differentiation was inhibited irrespective of the dorsoventral position. In addition, dorsoventral properties were not grossly affected as revealed by molecular markers. Concordantly, when Zic1 was overexpressed in the dorsal spinal cord in transgenic mice, we observed hypercellularity in the dorsal spinal cord. The transgene-expressing cells were increased in comparison to those of truncated mutant Zic1-bearing mice. Conversely, we observed a significant cell number reduction without loss of dorsal properties in the dorsal spinal cords of Zic1-deficient mice. Taken together, these findings suggest that Zic1 controls the expansion of neuronal precursors by inhibiting the progression of neuronal differentiation. Notch-mediated inhibition of neuronal differentiation is likely to act downstream of Zic genes since Notch1 is upregulated in Zic1-overexpressing spinal cords in both the mouse and the chick.