Embryonic Schwann cell development: the biology of Schwann cell precursors and early Schwann cells

Transcription

1 J. Anat. (1997) 191, pp , with 2 figures Printed in the United Kingdom 501 Review Embryonic Schwann cell development: the biology of Schwann cell precursors and early Schwann cells K. R. JESSEN AND R. MIRSKY Department of Anatomy and Developmental Biology, University College London, UK (Accepted 6 May 1997) ABSTRACT The cellular events leading to the generation of Schwann cells from the neural crest have recently been clarified and it is now possible to outline a relatively simple model of the Schwann cell lineage in the rat and mouse. Neural crest cells have to undergo 3 main developmental transitions to become mature Schwann cells. These are the formation of Schwann cell precursors from crest cells, the formation of immature Schwann cells from precursors and, lastly, the postnatal and reversible generation of non-myelin- and myelin-forming Schwann cells. Axonal signals involving neuregulins are important regulators of these events, in particular of the survival, proliferation and differentiation of Schwann cell precursors. Key words: Neural crest; neuregulin; myelination. INTRODUCTION In retrospect, it is surprising how recently Schwann cell biologists started to pay significant attention to questions relating to the developmental origin of Schwann cells. Apart from classical studies indicating that in birds most Schwann cells emerge from the neural crest (Le Douarin & Smith, 1988), little attention was paid to defining early stages of Schwann cell development or to finding the factors that regulate specification and progression in this lineage, at a time when these questions were being studied in many other systems, including other lineages of the nervous system. Some of the first observations on early Schwann cells in mammals were carried out on embryonic rat nerves. They showed that nascent nerves contained glial cells that had flattened sheet-like processes that enveloped a large number of axons communally (Peters & Muir, 1959; Ziskind-Conhaim, 1988). In the human, similar cells were described in nerves of 12- wk-old embryos. Since cells with comparable structure and relationship with axons were found in both rat and human nerves right up to, and overlapping with, the time at which some Schwann cells start to assemble myelin sheaths, these morphological examinations failed to provide any pointers to the existence of distinctive developmental stages in glial development of early nerves, prior to myelination (Peters & Muir, 1959; Webster et al. 1973). Perhaps this is part of the reason why workers were disinclined to address the questions of when, how and from what kind of progenitor the Schwann cells in the nerves of newborn rodents arose. THE IDENTIFICATION OF SCHWANN CELL PRECURSORS IN RAT AND MOUSE NERVES In development of the rat embryo, nerves first extend into the hind limb in significant numbers at E13 14 (Reynolds et al. 1991). The first indication that the cells in these nerves differed significantly from Schwann cells in nerves of newborn animals (rats are born at E21 23) came from the observation that cells from E14 nerves underwent programmed cell death within 20 h when they were dissociated from the nerves and plated at moderate high density in vitro, while Schwann cells survived under identical con- Correspondence to Prof. K. R. Jessen, Department of Anatomy and Developmental Biology, University College London, Gower Street, London WC1E 6BT, UK. k.jessen ucl.ac.uk

2 502 K. R. Jessen and R. Mirsky Table 1. Some of the main differences between precurors and Schwann cells in the rat sciatic nerve* E14 Schwann cell precursors Die by apoptosis when removed from axons and plated in vitro The glial protein S100 is very low or absent from cytoplasm Do not divide in response to FGFs Are flattened with extensive cell-cell contacts in vitro Do not have autocrine survival loops Express very low levels of the transcription factor Krox-20 Motility * For references, see text. Schwann cells from newborn nerves Excellent survival under same conditions Express cytoplasmic S100 Divide in response to FGFs Are bi- or tripolar in vitro Have autocrine survival loops Express high levels of the transcription factor Krox-20 Motility ditions (Jessen et al. 1994). When we carried out a more extensive comparison of these cells it emerged that E14 cells and newborn Schwann cells differed with respect to a large number of cellular properties (Table 1). Our conclusion was that the developing rat sciatic nerve harboured at least 2 distinct types of cell during early development, one at E14 that we refer to as the Schwann cell precursor, and another at birth, about a week later, the specified Schwann cells (Jessen et al. 1994; Dong et al. 1995; Gavrilovic et al. 1995; Mirsky & Jessen, 1996). More recently, we have found that 2 d earlier, at E12, mouse nerves contain a Schwann cell precursor that is similar, albeit not identical, to the rat E14 precursor (unpublished). IN THE RAT SCIATIC NERVE SCHWANN CELLS ARE GENERATED FROM PRECURSORS AT E15 E17 To find out when Schwann cell precursors give rise to Schwann cells, we determined the phenotype of the cells in the sciatic nerve on successive days between E14 and birth (Jessen et al. 1994). When we examined the ability of the cells to survive for 20 h in routine culture media in the absence of neurons we found that cells from E15 nerves died like E14 cells, while most cells from E17 nerves survived like newborn cells. Similarly, nearly all E15 cells expressed very low or undetectable cytoplasmic levels of the glial protein S100 (see Table 1) while E17 cells, like newborn cells, contained readily detectable S100. The mitogenic response to basic fibroblast growth factor (FGF-2) also appeared between E15 and E17. A shift in the expression of other transcription factors (our unpublished observations) and in in vitro morphology and cell-cell relationships in culture also occurs with a similar time course, as does significant expression of the zinc finger protein Krox-20, a transcription factor of key importance in myelination (Topilko et al. 1994). The simplest view of these findings is that the nerve essentially contains precursors at E14 and E15 and Schwann cells from E17 onwards. The time between E15 and E17 is particularly important since during this period precursor cells progress to generate Schwann cells. This change in phenotype extends to many diverse and apparently unrelated cell features, indicating a major difference between the 2 developmental stages, and belying their morphological similarity mentioned above. Nevertheless, the transition occurs relatively abruptly, so that only during E16 does the sciatic nerve contain significant numbers of both cell types. As discussed more fully below, the generation of Schwann cells from precursors has now also been examined under well-defined conditions in culture. These studies remove any doubt that the change in the phenotype of whole cell populations described above represents conversions of individual precursor cells to Schwann cells. THE DEVELOPMENTAL ORIGIN OF SCHWANN CELLS: THE NEURAL CREST AND P EXPRESSION We know less about the formation of Schwann cell precursors from neural crest cells than we do about the generation of Schwann cells from precursors. Drawing a distinction between Schwann cell precursors and migrating neural crest cells is made more complicated by the fact that crest cells in the rat or mouse are almost certainly not a homogenous population, but likely to be a mixture of multipotential cells and cells at various stages of entry to the different crest-derived lineages. In addition to entering the Schwann cell lineage, crest cells fated to become Schwann cells must find outgrowing neurites, associate with them, and so become incorporated into embryonic nerves. The initial encounter between crest cells and axons is likely to take place in the anterior parts of the somites that form along both sides of the neural tube. As neural crest cells spread from their site of origin at the dorsal margin of the tube, many of them are directed through the anterior part of each somite, while the posterior part is avoided. Around this time, the first peripheral axons emerge from the ventrolateral part of the neural tube. They also select the anterior part of each somite in preference to the posterior part. Therefore, both axons and crest cells

3 Schwann cell development 503 are channeled to the same place at the same time (Rickman et al. 1985; Loring & Erickson 1987). Tracing the Schwann cell lineage one step closer towards its developmental origin would entail identifying those crest cells that have entered the Schwann cell lineage but have not yet joined outgrowing axons to become Schwann cell precursors within the newly formed nerves. Surprisingly, it is possible that this can be done by examining expression of the gene for the myelin protein P.P is the most abundant protein of Schwann cell myelin. Classically, it was considered a myelin restricted protein, made by myelin-forming cells as a part of a specific response to myelination-inducing axonal signals (Bray et al. 1981; Lemke, 1988). This picture has now to be modified, since it is now clear that low, but clearly detectable, basal levels of P are expressed not only in essentially all immature Schwann cells in the rat, irrespective of whether they will subsequently myelinate or not, but also in Schwann cell precursors (Lee et al. 1997). The striking increase in P synthesis that is induced by axons selectively in those cells that form myelin is therefore a strong up-regulation of pre-existing basal levels rather than novel gene expression. Significantly, we found that P is also present in a subpopulation of migrating neural crest cells in rodents, in agreement with findings in the chick (Bhattacharyya et al. 1991; Lee et al. 1997). There are no published reports of P expression in cells other than Schwann cells and in view of this tight lineage-specificity of this gene in late development it is likely, although not proven, that P expression in crest cells marks those cells that have just entered the Schwann cell lineage. It will be interesting to examine the relationship between P expression in crest cells and expression of markers of entry to other crest lineages. If P expression emerges as a specific marker of the glial lineage, it follows that many crest cells have selected a glial fate even at the early stages of crest migration. While basal P expression might provide the continuity that enables us to trace Schwann cell development as far back as the migrating crest, differences are also evident between crest cells and precursors. As already indicated, one of these is the relationship of these cells to surrounding tissues, since, unlike the crest, one of the hallmarks of precursors is to be intimately associated with axons, forming with them the compact structure early embryonic nerves. Furthermore, neuregulins do not appear to promote crest cell survival although they act as potent survival factors for Schwann cell precursors (see below) and precursors, unlike crest cells, express the protein GAP-43 (Jessen et al. 1994). USING NEUREGULIN, AXONS REGULATE PRECURSOR SURVIVAL AND DIFFERENTIATION The experiments discussed so far build up to a simple picture of the Schwann cell lineage (Fig 1). It appears that the generation of the non-myelin and myelinforming cells in mature nerves from neural crest cells, involves the formation of 2 main intermediates, the Schwann cell precursor, typically found in rat nerves at E14 and 15 (mouse E12 and 13) and the immature Schwann cell, present from E17 (mouse E15) to around birth. At that time, the cells start to differentiate, first along the myelin pathway, with the mature non-myelin-forming cells appearing later. The lineage therefore involves 3 main transition points, i.e. the transition of crest cells to precursors, of precursors to immature Schwann cells and lastly the final, and largely reversible, formation of the 2 mature Schwann cell types (Stewart et al. 1995; Mirsky & Jessen 1996). It appears that a single signal, neuregulin, is of paramount importance in regulating the embryonic phase of these events. Neuregulins (also known as NDF, heregulin, ARIA, GGF or SMDF) are generated from a single gene, exist both in membranebound and soluble isoforms and regulate growth and differentiation of several cell types in vitro (Ben- Baruch & Yarden, 1994; Gassman & Lemke, 1997). As mentioned above, E14 rat Schwann cell precursors undergo programmed cell death when they are removed from embryonic nerves and placed in neuron-free cultures. Dorsal root ganglion (DRG) neurons secrete protein(s) that prevent this death, and isolated axonal membranes also possess activity that supports precursor survival (Jessen et al. 1994; Ratner, Jessen & Mirsky, unpublished). It is likely, therefore, that the survival of Schwann cell precursors in vivo is regulated by signals from the axons with which they associate. These signals also drive, or permit, the next step in the development of the lineage, since the precursors mature to generate Fig. 1. The Schwann cell lineage. Regulation by axon-associated neuregulin (NRG) is indicated.

4 504 K. R. Jessen and R. Mirsky Schwann cells in vitro in the presence of the neuronderived proteins. There is now strong evidence that a key component of this signal is neuregulin. Beta forms of neuregulin, but not alpha forms, support precursor survival in vitro for several days and, during this period, the cells alter their phenotype from that of precursors to that of Schwann cells. The time course of this Schwann cell generation in vitro is similar to that with which the glial population of developing nerves switches from the precursor phenotype to the Schwann cell phenotype in vivo. Most importantly, the action of the signal from DRG neurons that supports precursor survival and differentiation is blocked by a soluble hybrid protein containing the extracellular domain of the ErbB-4 receptor, a high affinity receptor for neuregulin that is not known to bind to any other growth factor (Dong et al. 1995). Neuregulin is also present at the appropriate time and place in rat and mouse embryos: strikingly high expression of neuregulin mrna is seen over the motor neurons of the ventral horn in the spinal cord and over DRG neurons, the major sources of axons in embryonic peripheral nerves (Marchionni et al. 1993; Orr-Urtreger et al. 1993; Meyer and Birchmeier, 1994; Ho et al. 1995). Taken together, these results strongly suggest that neuronally derived neuregulin is likely to act as a survival maturation factor for Schwann cell precursors. FGFs also support precursor survival, in combination with IGF-1 (Gavrilovic et al. 1995). However, FGFs plus IGF-1 only support survival in the short term and most cells die during the second day of exposure to these factors in vitro (Dong et al. 1995). The ability to study Schwann cell generation from precursors in vitro, in the presence of neuregulin, should make it easier to analyse how this key step in Schwann cell development is regulated. It is already clear that precursors generate Schwann cells in a process that is largely independent of proliferation: precursor differentiation proceeds equally in cultures in which cell division is fast or slow, and precursors are capable of converting to Schwann cells without dividing at all. Distinct signals are also being identified that, respectively, accelerate and delay Schwann cell generation (our unpublished observations). An interesting question that is still unresolved is whether action of neuregulin on Schwann cell precursors is simply to block death or whether this signal also promotes differentiation. Thus it is possible that by ensuring survival, neuregulin permits the operation of an intrinsic differentiation programme in the precursor cells that dictates their conversion to Schwann cells. Alternatively, neuregulin is required to drive this differentiation programme, in addition to its survival promoting function. HOW IS SCHWANN CELL SURVIVAL REGULATED? We have argued above that the survival of precursors depends acutely on axonal survival signals mediated by neuregulin. This arrangement is likely to have biological advantages since it would help to match glial numbers to axon numbers. It might also have a morphogenetic role by ensuring that precursors that strayed from axons to inappropriate locations away from the nerve would die. It would, however, be calamitous if the same arrangement applied to postnatal Schwann cells, since nerve damage and consequent loss of axonal contact would then threaten large scale Schwann cell death. It might therefore be expected that postnatal Schwann cells possess an alternative, axon independent, way of surviving. We have found that such a mechanism does indeed exist in the form of an autocrine survival loop (unpublished) (Fig. 2). Importantly, such survival loops do not exist in Schwann cell precursors. In our initial comparison of survival of precursors and Schwann cells (Jessen et al. 1994), the cells were grown at a density sufficient for Schwann cells to achieve a concentration of secreted survival factors that ensured survival. Precursors are unable to do this, irrespective of cell density, and therefore die in the absence of neuronal signals (Jessen et al. 1994). Schwann cells can be seen to die in a similar way in vitro but only if they are Fig. 2. Both axon-derived and autocrine signals regulate survival in the Schwann cell lineage. There is a shift from an exclusive reliance on autocrine signals to establishment of autocrine survival loops as precursors generate Schwann cells (for references, see text). The panel showing adult nerves is hypothetical since autocrine mechanisms in adult Schwann cells have not been examined. Ax, axons; Pre, Schwann cell precursors; Sch, Schwann cells.

5 Schwann cell development 505 plated at extremely low cell density. In agreement with this, transection of adult nerves does not appear to lead to Schwann cell death, while degeneration of embryonic axons leads to death of precursors, at least in the chick (Ciutat et al. 1996; Trachtenberg & Thompson, 1996). Schwann cells in newborn nerves may be in an intermediate stage, having developed the ability of axon independent, autocrine, survival support, while partially relying on axonal survival signals (Grinspan et al. 1996; Syroid et al. 1996; our unpublished observations). It will be of great interest to identify the autocrine Schwann cell growth factors. REFERENCES BEN-BARUCH N, YARDEN Y (1994) Neu differentiation factors: a family of alternatively spliced neuronal and mesenchymal factors. Proceedings of the Society for Experimental Biology and Medicine 206, BHATTACHARYYA A, FRANK E, RATNER N, BRACKENBURY R (1991) P is an early marker of the Schwann cell lineage in chickens. Neuron 7, BRAY GM, RASMINSKY M, AGUAYO AJ (1981) Interactions between axons and the sheath cells. Annual Review of Neuroscience 4, CIUTAT D, CALDERO J, OPPENHEIM RW, ESQUERDA JE (1996) Schwann cell apoptosis during normal development and after axonal degeneration induced by neurotoxins in the chick embryo. Journal of Neuroscience 16, DONG Z, BRENNAN A, LIU N, YARDEN Y, LEFKOWITZ G, MIRSKY R et al. (1995) NDF is a neuron-glia signal and regulates survival, proliferation, and maturation of rat Schwann cell precursors. Neuron 15, GASSMAN M, LEMKE G (1997) Neuregulins and neuregulin receptors in neural development. Current Opinion in Neurobiology 7, GAVRILOVIC J, BRENNAN A, MIRSKY R, JESSEN KR (1995) Fibroblast growth factors and insulin growth factors combine to promote survival of rat Schwann cell precursors without induction of DNA synthesis. European Journal of Neuroscience 7, GRINSPAN JB, MARCHIONNI MA, REEVES M, COULALOGLOU M, SCHERER SS (1996) Axonal interactions regulate Schwann cell apoptosis in developing peripheral nerve: neuregulin receptors and the role of neuregulins. Journal of Neuroscience 16, HO W-H, ARMANINI MP, NUIJENS A, PHILLIPS HS, OSHEROFF PL (1995) Sensory and motor neuron-derived factor. A novel neuregulin variant highly expressed in sensory and motor neurons. Journal of Biological Chemistry 270, JESSEN KR, BRENNAN A, MORGAN L, MIRSKY R, KENT A, HASHIMOTO Y et al. (1994) The Schwann cell precursor and its fate: a study of cell death and differentiation during gliogenesis in rat embryonic nerves. Neuron 12, LE DOUARIN NM, SMITH J (1988) Development of the peripheral nervous system from the neural crest. Annual Review of Cell Biology 4, LEE M-J, BRENNAN A, BLANCHARD A, ZOIDL G, DONG Z, TABERNERO A et al. (1997) P is constitutively expressed in the rat neural crest and embryonic nerves and is negatively and positively regulated by axons to generate non-myelin-forming and myelinforming Schwann cells respectively. Molecular and Cellular Neuroscience, in press. LEMKE G (1988) Unwrapping the genes of myelin. Neuron 1, LORING JF, ERICKSON CA (1987) Neural crest pathways in the trunk of the chick embryo. Developmental Biology 121, MARCHIONNI MA, GOODEARL ADJ, CHEN MS, BERMINGHAM- MCDONOGH O, KIRK C, HENDRICKS M et al. (1993) Glial growth factors are alternatively spliced erbb2 ligands expressed in the nervous system. Nature 362, MEYER D, BIRCHMEIER C (1994) Distinct isoforms of neuregulin are expressed in mesenchymal and neuronal cells during mouse development. Proceedings of the National Academy of Sciences of the USA 91, MIRSKY R, JESSEN KR (1996) Schwann cell development, differentiation and myelination. Current Opinion in Neurobiology 6, ORR-URTREGER A, TRAKHTENBROT L, BEN-LEVY R, WEN D, RECHAVI G, LONAI P et al. (1993) Neural expression and chromosomal mapping of Neu differentiation factor to 8p12-p21. Proceedings of the National Academy of Sciences of the USA 90, PETERS A, MUIR AR (1959) The relationship between axons and Schwann cells during development of peripheral nerves in the rat. Quarterly Journal of Experimental Physiology 64, REYNOLDS ML, FITZGERALD M, BENOWITZ LI (1991) GAP-43 expression in developing cutaneous and muscle nerves in the rat hindlimb. Neuroscience 41, RICKMAN M, FAWCETT JW, KEYNES RJ (1985) The migration of neural crest cells and the growth of motor axons through the rostral half of the chick somite. Journal of Embryology and Experimental Morphology 90, STEWART HJS, MIRSKY R, JESSEN KR (1995) The Schwann cell lineage: embryonic and early postnatal development. In Glial Cell Development, Basic Principles and Clinical Relevance (ed. Jessen KR, Richardson WD), pp Oxford: Bios Scientific Publishers Ltd. SYROID DE, MAYCOX PR, BURROLA PG, LIU N, WEN D, LEE K-F et al. (1996) Cell death in the Schwann cell lineage and its regulation by neuregulin. Proceedings of the National Academy of Sciences of the USA 93, TOPILKO P, SCHNEIDER-MANOURY S, LEVI G, BARON-VAN EVERCOOREN A, CHENNOUFI ABY, SEITANIDOU T et al. (1994) Krox-20 controls myelination in the peripheral nervous system. Nature 371, TRACHTENBERG JT, THOMPSON WJ (1996) Schwann cell apoptosis at developing neuromuscular junctions is regulated by glial growth factor. Nature 379, WEBSTER H DE F, MARTIN JR, O CONNELL MF (1973) The relationships between interphase Schwann cells and axons before myelination: a quantitative electron microscopic study. Developmental Biology 32, ZISKIND-CONHAIM L (1988) Physiological and morphological changes in developing peripheral nerves of rat embryos. Developmental Brain Research 42,