Nerfin-1 is a nuclear regulator of axon guidance required for a subset of early pathfinding events in the developing Drosophila CNS. Nerfin-1 belongs to a highly conserved subfamily of Zn-finger proteins with cognates identified in nematodes and man. The neural precursor gene prospero is essential for nerfin-1 expression. Unlike nerfin-1 mRNA, which is expressed in many neural precursor cells, the encoded Nerfin-1 protein is only detected in the nuclei of neuronal precursors that will divide just once and then transiently in their nascent neurons. Although nerfin-1 null embryos have no discernible alterations in neural lineage development or in neuronal or glial identities, CNS pioneering neurons require nerfin-1 function for early axon guidance decisions. Furthermore, nerfin-1 is required for the proper development of commissural and connective axon fascicles. Nerfin-1 is essential for the proper expression of robo2, wnt5, derailed, G-oα47A, Lar, and futsch, genes whose encoded proteins participate in these early navigational events (Kuzin, 2005).

Initially discovered in a screen for neural precursor genes, nerfin-1 encodes a member of a subfamily of Zn-finger proteins with cognates identified in all metazoans (Brody, 2002a; Stivers, 2000). Nerfin-1 and other family members contain a highly conserved set of tandem Zn-fingers termed the EIN domain [named after the first three identified members: nematode egg laying-46 mutant (egl-46) (Desai, 1989; Desai, 1988); human insulinoma associated-1, IA-1 (Goto, 1992); and Drosophila nerfin-1 (Stivers, 2000). Loss of egl-46 function causes multiple defects in C. elegans nervous system development including abnormalities in cell migration, axonal outgrowth, and neurotransmitter production (Wu, 2001; Yu, 2003). The mammalian IA-1, shown to be a sequence-specific DNA binding protein (Breslin, 2002), is expressed during CNS development and in neuroendocrine tumors (Breslin, 2003; Kuzin, 2005 and references therein).

Although nerfin-1^null mutant embryos have no discernible alteration in either CNS or PNS lineage development, loss of nerfin-1 triggers axonal patterning defects throughout the CNS, but not in the PNS. These studies demonstrate that Nerfin-1 is required for the proper expression of a subset of factors that participate in early axon guidance decisions essential to both commissural and longitudinal connective axon fascicle development. Analysis of nerfin-1 expression reveals that Nerfin-1 protein is detected in only a subset of neural precursor cells that express its encoding message. Nerfin-1 accumulates in the nuclei of only those precursor cells that divide just once producing neurons and is then transiently expressed in their nascent offspring. Prolonged expression of Nerfin-1 in neurons interferes with later stages of CNS axon fasciculation and patterning. Finally, nerfin-1 is shown to be downstream of pros in the regulatory network(s) controlling axon guidance determinants. Epistasis experiments also demonstrate that the regulatory networks controlling nerfin-1 expression and those regulating the expression of lola or fru are separate from one another (Kuzin, 2005).

Although nerfin-1 mRNA expression is first detected in all early delaminating NBs, its encoded protein is detected only in NBs, specifically the MPs, which divide just once to produce interneurons. The transient expression of nerfin-1 mRNA in NBs, during the early phases of lineage development but not during intermediate or late stages, suggests that its NB expression is subject to temporal regulation (reviewed by Brody, 2002b). Following this initial phase of expression, nerfin-1 mRNA is detected in most if not all GMCs and nascent neurons and, again, Nerfin-1 protein is detected only transiently in a subset of these cells. Nerfin-1 protein is also transiently expressed in nascent PNS neurons. The transient expression of Nerfin-1 in neurons during the initial phases of axon development is consistent with the idea that it may be required for specific aspects of early axon guidance, particularly in CNS interneurons. In addition, the absence of detectable Nerfin-1 protein in many mRNA expressing cells suggests that its message may be translationally blocked and/or that the protein is rapidly degraded in these cells. A recent genome-wide screen for genes whose transcripts contain potential binding sites for the translational blocking micro-RNAs (miRNAs) has revealed that the 3′ UTR of the nerfin-1 message (1600 bases long) contains multiple predicted docking sites for nine different miRNAs (Enright, 2003). Protein instability motifs (PEST sites) within Nerfin-1 may also play a role in rapidly clearing Nerfin-1 from cells (Kuzin, 2005).

Although no cross-regulation was detected between nerfin-1 and lola or fruitless, pros is required for full nerfin-1 expression. In addition to its role in axon guidance, pros has been shown to be required to bring precursor cells out of the proliferative state as a prelude to neuronal proliferative quiescence. Absence of pros results in an additional division of the GMC that results in increased numbers of cells in the CNS. In nerfin-1^null embryos, no evidence was found to indicate that loss of nerfin-1 affects precursor or neuron cell numbers nor does it trigger an increase in cell death, suggesting that nerfin-1 carries out a restricted repertoire of the pros functions, specifically those involved with axon guidance (Kuzin, 2005).

Analysis of nerfin-1^null embryos indicates that Nerfin-1 is needed by many, but not all, CNS neurons during the early phases of axon patterning. For example, while the development and patterning of ventral cord longitudinal connective fascicles is disrupted and both anterior and posterior commissure development is compromised, no significant disruptions are detected in motoneuron nerve tracks that exit the CNS, nor are significant disruptions detected in the PNS axon patterning (Kuzin, 2005).

It is likely that many of the subsequent defects in axon guidance, related to crossing segmental boundaries, defasciculation, and/or commissural development may have their origins in defects in the initial guidance decisions made by pioneering axons. Disruption in the initial pathfinding events can trigger subsequent misguidance of 'follower neurons' that rely on guidance cues laid down by pioneers. The pCC axon misguidance in nerfin-1 mutants could explain, in part, why many other axons fail to extend across the segmental border in nerfin-1mutants. Two aspects of pCC axon guidance are altered in nerfin-1 mutants: (1) the pCC interneuron fails to extend its axon in the proper anterior orientation and, (2) many of the misguided axons cross the midline in adjacent posterior segments. During normal development, the pCC ipsilateral axon pioneers the innermost longitudinal fascicle, the medial fascicle, and other axons project along the tract established by the pCC. Although defects in the initial anterior/posterior direction of pCC axon projection have not been reported in other mutant backgrounds, abnormal crossing of the midline has been observed in eve mutants (Fujioka, 2003) and also observed in robo mutant embryos; axons that normally pioneer ipsilateral projections project anteriorly and then cross the midline, and contralaterally projecting axons re-cross the midline multiple times. At this time, it is not known whether the follower axons also are defective in navigation, or whether their misguidance is solely due to initial axon pathfinding mistakes made by the pioneering neurons. Given the large numbers of neurons that express Nerfin-1, a scenario that includes both possibilities is favored (Kuzin, 2005).

Analysis of the expression dynamics of known axon guidance genes in nerfin-1^null embryos has revealed that nerfin-1 function is required for the proper expression of a subset of factors involved in at least two signaling events essential for early CNS axon patterning. Although Netrin/Frazzled signaling appears not to be affected by loss of nerfin-1, nerfin-1 function is required for the proper expression of specific genes involved in the decision to cross the midline and in the choice of commissures. One of the components of the Slit/Robo system, Robo2, requires Nerfin-1, either directly or indirectly, for full expression. Interestingly, altered expression of the other factors involved in the Slit/Robo pathway was not detected, specifically slit, comm, comm3, robo, and robo3 appear unaffected in nerfin-1^null embryos. Robo2 expression is restricted to axons that extend in the outer longitudinal pathway (the lateral fascicle), farthest from the midline and thus farthest from the source of Slit. The robo2 loss-of-function phenotype does not resemble that of the nerfin-1^null, suggesting that only a part of the nerfin-1 loss-of-function phenotype can be derived from its regulation of robo2 (Kuzin, 2005).

nerfin-1 function is also required for the proper expression of Drl and Wnt5, two factors involved in the choice between entering the AC or PC. Both derailed and Wnt5 mutants display abnormal projections of AC axons. Although not as severe as in nerfin-1^null embryos, the loss of Wnt5 also triggers breaks in longitudinal connectives, specifically the intermediate and lateral fascicles (Kuzin, 2005).

Delayed onset of futsch expression was observed in nerfin-1^null embryos. In loss-of-function futsch mutant alleles, the development of the lateral-most longitudinal connectives is reduced. Some of the axon fasciculation defects seen in nerfin-1^null embryos could be explained in part by delayed futsch expression, especially those exhibited by pioneering axons that make up the lateral fascicle (Kuzin, 2005).

Reduced expression of G-oα47A and Lar were also observed. G-oα47A is required for the proper development of Fas2-positive connectives. However, the axon guidance phenotype of G-oα47A mutants is more severe than in nerfin-1^null embryos; in the G-oα47A mutant: AC and PC do not separate. The greater severity in the G-oα47A mutant could be explained by the fact that loss of nerfin-1 does not completely ablate G-oα47A expression. Lar has been found to regulate motoneuron axon guidance decisions outside of the CNS. The lack of motoneuron axon patterning defects in nerfin-1^null embryos could be explained by the fact that Lar expression, in mutants, was just reduced and not ablated (Kuzin, 2005).

Although some of the axon guidance defects in the nerfin-1^null mutants can be explained, in part, as a result of the altered expression of the above genes, certain aspects of the patterning defects are distinct from those described for these genes. For example, patterning defects in pCC axons have been observed due to loss of eve or robo, but the reversed anterior/posterior orientation of many of the pioneering pCC axons in nerfin-1^null embryos has not been seen in other mutant backgrounds. This suggests that Nerfin-1 may regulate additional, as yet uncharacterized pathfinding determinants. Therefore, it will be very interesting to identify additional genes regulated by Nerfin-1 since these are very likely to include new members of the axon guidance machinery (Kuzin, 2005).

GENE STRUCTURE

cDNA clone length - 3200

Bases in 5' UTR - 159

Exons - 1

Bases in 3' UTR - 1631

PROTEIN STRUCTURE

Amino Acids - 469

Structural Domains

Full-length sequence analysis of a novel cDNA detected in a screen for neural precursor genes (Brody, 2002a; Stivers, 2000) and an overlapping cDNA obtained from the Berkeley EST Project (LD18634.5) revealed that they correspond to a previously uncharacterized gene that codes for a 469 aa Zn-finger protein (Genbank accession nos. AF203690 and CG13906). Database homology searches for related Drosophila proteins have uncovered an additional uncharacterized gene (Genbank nos. AC009183; CG12809). The two predicted proteins show high homology in their Zn-finger motifs. Based on the observation that both are expressed only in the nervous system and that they encode related proteins, the genes were named nerfin-1(located at 61E1) and nerfin-2 (at 85F) -- 'nerfin' refers to nervous finger. Homology searches of other genomes have revealed that the nerfins belong to a highly conserved Zn-finger gene subfamily marked by the proteins' first two Zn-fingers, termed here the EIN domain (pronounced EE-in) after the first three identified proteins: the nematode Egg laying mutant, Egl-46 (Genbank no. U64853); the human Insulinoma associated protein, IA-1 (Goto, 1992); and the Drosophila Nerfin-1. Alignment of all known EIN domains shows that homology is highest in the predicted Zn-finger alpha helical DNA reading heads, suggesting that they may have similar, if not identical, DNA-binding specificity. Crystallographic analyses of other Zn-finger protein-DNA complexes have demonstrated that amino acids within each finger's alpha helix play key roles in conferring DNA-binding specificity. Comparison of the complete primary structures of the known EIN domain proteins reveals that they contain one to three additional Zn-fingers and also harbor other motifs found in developmentally important transcription factors (Stivers, 2000).

date revised: 28 December 2004

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.