The adult tooth pulp has a rich protective sensory innervation from the trigeminal ganglion (TG). The developmental establishment of this exclusively nociceptive innervation is not fully understood today. However, an increasing amount of data indicates that the tooth pulp cells produce molecular factors, which attract relevant nerve fibres from adjacent major nerve trunks. Presently, a wide variety of agents, which influence the growth of nerve fibres in various contexts, are known. Molecular factors, which might influence the innervation process of the developing tooth, are the focus of this thesis.

Based on gene-expression studies it has been suggested that nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF) and glial cell line-derived neurotrophic factor (GDNF) may be involved in the formation of a pulpal innervation. However, these studies did not examine how pulpal tissue/pulpal cells and TGs/TG neurones actually interact in vitro. In the first of the four studies included in this thesis we show that pulpal explants from neonatal rat pups elicit neurite growth from eo-cultured TG explants. It was also fotmd that application of exogenous NGF or GDNF, but not BDNF, to single TGs in vitro gives a rich neurite outgrowth. However, treatment of TG/pulpal co-cultures with antibodies against these neurotrophic factors does not fully block the neuritogenic effect of the pulpal explant. This seems to suggest that other, as yet tm.known factors are also involved. The second paper supplements the first one by illustrating the pulpal/TG interaction at the level of single cells. The results show that single TG neurones emit relatively linear neurites up to a single pulpal cell or a group of such cells. Upon reaching the pulpal cell/cells the neurite changes behaviour and forms a network of thin beaded branches in direct relation to the cell/cells. In co-cultures of TG neurones and a mixture of pulpal cells and 3T3 fibroblasts the neurones form neuritic networks exclusively in relation to the former. Finally, RT/PCR analysis revealed that both pulpal tissue and cultured pulpal cells express the genes for all known mammalian neurotrophins and GDNF-family members. The third paper compares the effects of cultured pulpal cells, and transfected 3T3 fibroblasts overexpressing selected neurotrophic factors on single TG neurones in terms of survival and neurite growth. It was found that the different 3T3 fibroblasts affect the neurones in different ways, and that none of them fully mimic the influences of wild-type pulpal cells. Also, the different 3T3 fibroblasts induce dissimilar neurite branching patterns. The pattern associated with NGF-overexpressing 3T3 cells is most similar to the one seen in co-cultures with TG neurones and pulpal cell.

Onset of nerve growth to the tooth pulp occurs comparatively late in development, although tooth pulp cells produce neurite growth stimulating molecular factors already during embryonic development, and in spite of the fact that nerve fibres from neurones expressing the appropriate receptor genes are located in the vicinity. These observations suggested the hypothesis that axon growth to prospective tooth pulps is transiently actively blocked during embryonic and early neonatal development. But, when amelogenesis and dentinogenesis have commenced nerve fibres enter the tooth pulp. This indicates that the hypothetical blockade decreases after birth, allowing an innervation of the tooth pulp to become established. In order to test this hypothesis, tooth-related tissue and TG explants from mice at various developmental stages were eo-cultured. It was found that mesenchymal tissue collected from early developing jaw or tooth primordia repel TG neurites, and that corresponding tissue from teeth in a more advanced stage of development attract TG neurites. One group of molecules with neurite growth-inhibiting properties are the semaphorins. Hence, we examined the presence of mRNAs for selected semaphorins in mesenchymal tissue collected from the jaw or from tooth pulps at different developmental stages. The results show that some of the selected semaphorins are expressed in a temporal pattern that matches the observed neuriteinhibiting effects seen in vitro and that conforms with the idea that there is an early pulpalderived blockage of nerve fibre growth in situ. Also, it was found that nerve fibres in the vicinity of the early developing tooth are immunoreactive for a relevant semaphorin receptor, neuropilin-1. After birth, when amelogenesis and dentinogenesis have commenced, nerve fibres in the tooth pulp are immunonegative for that receptor.

To conclude, our results show that pulpal tissue/pulpal cells can influence nerve fibres from TG explants/TG neurones in vitro. Also, we provide evidence that neurotrophins, GDNF-family members and semaphorins may contribute to this influence.