All sexually reproducing organisms are composed of two types of cells: the 'mortal' somatic cells, which form the body of the organism, and the 'immortal' germ cells, which produce the next generation. During development, precursor germ cells (better known as primordial germ cells, PGCs) are created in one part of the embryo, often far away from their final destination. They must then migrate to the somatic part of the future gonads, where they become mature germ cells — sperm or eggs. Genetic analyses of fruitflies, zebrafish and mice have led to the identification of several genes that are required for PGC migration1. Moreover, these studies have suggested that the somatic tissues that line the migratory path provide attractive, repulsive and survival cues that direct the germ-cell precursors1, 2. Such cues have so far proved elusive. But, writing respectively in Cell and on page 279 of this issue, Doitsidou et al.3 and Knaut et al.4 show that a pair of evolutionarily conserved molecules — the protein SDF-1 and its receptor CXCR4 — act to guide PGCs in zebrafish.
Cell migration is important for many biological processes, including embryonic development, immune responses, wound repair and the spread of tumour cells (metastasis). In many cases, proteins known as chemokines have been shown to trigger and guide cell migration. Initially identified as 'defence' proteins because of their ability to regulate immune-cell movement, chemokines have since also been found to be involved in events as diverse as blood and blood-vessel formation and embryo development5.
Broadly speaking, these proteins work by interacting with members of a superfamily of receptors found on the surface of motile cells. These are the G-protein-coupled receptors, so called because they interact with intracellular 'on/off' switches known as G proteins. The interaction between chemokine and receptor activates intracellular signal-transduction cascades, which induce cell movement, survival and gene expression in a manner that depends in part on the precise cell type and signal. For instance, during directed cell migration, the responding cells detect a small concentration gradient of the chemokine, such that the leading edge of the cell orients towards the signal's source. The activated receptor at the front edge stimulates a specific intracellular enzyme, leading to the recruitment of proteins that contain a structural region known as the pleckstrin-homology domain, involved in protein– protein interactions. This is thought to locally amplify the response to the signal6.
One chemokine–receptor pair consists of the chemokine SDF-1 and the receptor CXCR4. The receptor was first identified as being important for HIV-1 to enter cells, and the chemokine–receptor duo has also been implicated in the metastasis of breast-cancer cells7, 8. Moreover, genetic analysis of these two proteins in mice has shown that they are required for the development of blood cells (for example, in the migration of mature B cells to the bone marrow) and in the central nervous system (in the migration of granule cells within the layers of the cerebellar cortex, for instance)9, 10, 11, 12, 13, 14. Doitsidou et al.3 and Knaut et al.4 now show that SDF-1 and CXCR4 are also involved in PGC migration in zebrafish.
The two groups used different techniques to tackle the problem of how PGCs are guided to their destination in zebrafish embryos. Doitsidou et al.3 started by inactivating a large group of genes, one at a time, that they thought might be important for PGC migration. To do so, they used a technique known as 'morpholino antisense knockdown'. This led them to cxcr4b, one of two genes that encode CXCR4 proteins in zebrafish; when this gene was inactivated, PGCs failed to reach the somatic gonad and instead dispersed throughout the fish. Meanwhile, Knaut et al.4 made use of a large-scale genetic screen conducted by the Tübingen 2000 Screen Consortium, which randomly generated many thousands of mutant zebrafish. From these mutants Knaut et al. identified one, known as odysseus (ody), in which PGCs likewise failed to reach the gonad. The authors show that the mutation in ody animals occurs in the cxcr4b gene.
Both groups show that cxcr4b is expressed in PGCs, and Knaut et al. demonstrate that such expression is necessary for normal PGC migration. Moreover, it appears that when cxcr4b is inactivated or mutated, PGCs move, but they cannot read a chemotactic gradient3, 4. And although ody mutant PGCs can recruit a fluorescently labelled pleckstrin-homology domain to their leading edge, the localization of this domain is unstable and reflects the randomness of the migratory path4.
SDF-1 is the major chemokine that binds to CXCR4b, so is it, too, necessary for PGC migration? The new papers suggest that it is: it seems that germ-cell precursors are guided towards somatic cells that are producing high levels of SDF-1. The migratory path of zebrafish PGCs is complex15, but each step appears to follow the dynamic expression pattern of SDF-1 (refs 3, 4; Fig. 1a–d). Moreover, alterations in the pattern of SDF-1 expression, resulting from mutation of embryonic tissue-patterning genes, correlate with defects in PGC migration3. And reducing SDF-1 expression produces the same effects as mutating cxcr4b. Finally, incorrect expression of SDF-1 at a new location is sufficient to attract PGCs, and this response requires CXCR4b (Fig. 1e–g)3, 4.
Figure 1: Guiding primordial germ cells (PGCs) in zebrafish.
a–d,Normal migration of PGCs (red), incorporating the results of Doitsidou et al.3 and Knaut et al.4. a, Four clusters of PGCs originate at random positions along edges of the embryo and migrate towards the dorsal midline. b, At the beginning of somitogenesis (segment formation), sdf-1 RNA (green) is expressed strongly near the first segment, and PGCs move towards this location. c, During somitogenesis the four PGC clusters move towards higher levels of sdf-1. (Note, however, the high levels of sdf-1 in the hindbrain — large horizontal stripes — where PGCs do not go.) d, At the end of embryo development PGCs associate with sdf-1-expressing cells at the position of the future gonad (green vertical bar). e–g, Studies of SDF-1 and its receptor CXCR4b (refs 3, 4). e, After SDF-1 is depleted (blue–grey vertical bars) or in ody mutants (which lack functional CXCR4b; not shown), germ cells scatter throughout the embryo. f, Subsequent misexpression of SDF-1 attracts PGCs to a new position. g, ody mutant PGCs (not shown) and CXCR4b-depleted PGCs (purple) do not migrate towards misexpressed SDF-1.
High resolution image and legend (63K)These studies raise many interesting questions. For instance, are SDF-1 and CXCR4b the only chemokine–receptor pair that guides PGC migration in zebrafish? Possibly not: although PGCs travel to many tissues in which SDF-1 is strongly expressed, they ignore other tissues that express high levels of SDF-1 (Fig. 1c). So, additional signals, modifications of SDF-1, or interactions of SDF-1 with the proteins of the extracellular matrix in which cells are embedded may alter the chemokine's activity as a PGC attractant, and could be needed to keep these cells on course. Such secondary signals may not be sufficient, however, to attract or repel PGCs, as these cells do not appear to have a favoured destination when the levels of CXCR4b and SDF-1 are decreased.
Also, how does SDF-1 work? Do somatic cells release SDF-1, producing a long-range gradient that attracts PGCs from far away? Or are PGCs instead simply 'caught' when they move close to somatic cells that are expressing SDF-1 on their surface? Which signalling molecules within PGCs mediate the migratory response? And finally, how do germ cells 'know' when they have arrived at their destination? Zebrafish embryos provide an excellent system in which to address these questions.
It will also be interesting to learn whether this chemokine–receptor pair is involved in PGC migration in other organisms. Mice lacking either CXCR4 or SDF-1 have been analysed, but details of PGC migration have not yet been reported. Interestingly, these mice die as embryos with defects in several tissues9, 10, 11, 12, 13, 14. By contrast, in zebrafish, the ody mutants survive, and the SDF-1–CXCR4b pair seems to be required only during the migration of PGCs and cells known as lateral-line cells3, 4, 16. Different G-protein-coupled chemokine receptors, such as CXCR4a, may carry out other functions of CXCR4 in zebrafish. In fruitflies, no clear relatives of SDF-1 and CXCR4 have been identified, but G-protein-coupled receptors do seem to be important for PGC migration17. Indeed, in all organisms studied so far PGCs are seen crawling along, rather like other cells that use signals from G-protein-coupled receptors to trigger movement up chemical gradients6. So such signalling may be a major determinant, and the common denominator, of migrating PGCs throughout the animal kingdom.
