Hippocampus Formation: an Intriguing Collaboration Henk Roelink
Total Page:16
File Type:pdf, Size:1020Kb
bb10g06.qxd 04/03/2000 01:10 Page R279 Dispatch R279 Hippocampus formation: An intriguing collaboration Henk Roelink Recent genetic studies have shown that the signalling temporal lobes are formed. In an adult rodent, the hip- factor Wnt3a is required for formation of the pocampus is consequently still found close to the dorsal hippocampus; the developmental consequences of Wnt midline where it is initially formed (Figure 1). signalling in the hippocampus are mediated by multiple HMG-box transcription factors, with LEF-1 being It is clear that Wnt family members are required for the required just for formation of the dentate gyrus. development of many embryonic structures, functions mediated by their effects on fundamental processes such Address: Department of Biological Structure and Center for Developmental Biology, University of Washington, Box 357420, as cell proliferation, differentiation, survival or mainte- Seattle, Washington 98117-7420, USA. nance. Obtaining a clear picture of how each Wnt acts is E-mail: [email protected] complicated by the relatively large size of the family, which has at least 18 members in amniotes [3]. Analyses of Current Biology 2000, 10:R279–R281 mice lacking one or several Wnt genes have revealed some 0960-9822/00/$ – see front matter remarkable phenotypes, often characterized by failure of © 2000 Elsevier Science Ltd. All rights reserved. induction of very specific anatomical structures. Wnt3A is no exception, and although several papers have been pub- Inductive interactions are fundamental to the formation of lished already on the more obvious defects of Wnt3a all brain structures. and signalling by molecules of the knockout mice, a recent paper from McMahon and col- Wingless/Wnt family is known to play a role in many of leagues [1] has reported a very interesting defect in the them. Recent studies of knockout mice defective for embryonic forebrain — the loss of hippocampal structures Wnt3a [1] or Lef1 [2] — which encodes a transcription that derive from the most anterior neural site of Wnt3a factor known to be a downstream target of Wnt signalling expression, located in the area where the hippocampus — have shown that Wnt signalling is also critical in the and corpus callosum will form. development of the hippocampus. The genetic results View metadata, citationhave and provided similar papers insight at core.ac.uk into the complexity of the process, Galceran et al. [2] have reported interesting observations brought to you by CORE indicating as they do that more than one Wnt is likely to complementary to those of Lee et al. [1]. They have found provided by Elsevier - Publisher Connector be involved, with multiple downstream transcription that loss of a critical element of a Wnt response pathway, factors responsible for development of different parts of the transcription factor known as lymphoid enhancer- the hippocampus binding factor 1 (LEF1), causes a relatively mild hippocam- pal phenotype, whereas expression of what is believed to be The hippocampus is a part of the brain familiar from many dominant-negative form of LEF1 results in failure of all studies of learning and memory — it is the classic site of the hippocampal structures to develop. LEF1 is a member of form of neuronal plasticity known as long-term potentiation, TCF/LEF family, transcription factors with characteristic and damage to the hippocampus in humans is associated HMG domains. TCF/LEFs are known to be downstream with severe loss of memory. The hippocampus arises from targets of the signalling pathway activated when a Wnt an area of proliferating cells at the dorsal edge of the telen- binds to its receptor. Activation of this pathway leads to sta- cephalon. This area is lined medially by the forming choroid bilisation and consequent accumulation of cytoplasmic β plexus. Lateral to the presumptive hippocampus, the neo- catenin, which translocated to the nucleus. cortex develops from the telencephalon. But the neocortex and hippocampus are not contiguous — they are separated In the nucleus, β catenin interacts with a TCF/LEF, via a by a transitional cortex known as the subiculum. At more specific motif at the amino terminus of the latter transcrip- anterior positions, this dorsal midline position is occupied by tion factor, thereby generating a transcriptionally active the corpus callosum, the major site of the axon bundles that complex. In the absence of a Wnt signal, a phosphorylation mediate communication between the two hemispheres. reaction catalysed by glycogen synthase kinase 3β (GSK3β) leads to instability and consequent downregulation of cyto- The early development of the hippocampus is very plasmic β catenin via the ubiquitin–proteasome pathway. similar in rodents — which will be discussed below — In the absence of nuclear β catenin, TCF/LEFs recruit the and in humans. Later in development, derivatives of the corepressor protein Groucho to the target gene enhancers, human telencephalon expand considerably and fold to actively repressing their transcription [4]. give rise to the temporal lobe. The human hippocampus makes up the floor of the temporal lobe. In rodents, this Earlier analyses of Tcf and Wnt3a knockout mice indi- telencephalic expansion is much less dramatic, and no cated that a TCF/LEF is indeed required for Wnt3a bb10g06.qxd 04/03/2000 01:10 Page R280 R280 Current Biology Vol 10 No 7 Figure 1 Diagrams of the dorsal left cerebral hemisphere of (a) wild-type, The phenotypes of Wnt3a–/– and Lef1–LacZ/Lef1–LacZ mice (b) Lef1–/– and (c) Wnt3a–/– or Lef1–LacZ/Lef1–LacZ mouse (c) are similar and characterized by the absence of the embryos. In (a), the hippocampus, transitional cortex and neocortex hippocampus. Multilayered neocortical structures can be are indicated. In Lef1–/– mice (b), the typical ‘S’ shape of the observed at the position that would normally be occupied by hippocampus can still be seen, but the dentate gyrus is absent. the hippocampus. signal transduction. The abnormalities exhibited in the other hippocampal structures can be recognized. This paraxial mesoderm of Tcf1; Lef1 double mutant mice and phenotype is much less severe than the Wnt3a phenotype, of Wnt-3a single mutants are almost identical — addi- in which all hippocampal structures are missing, implying tional neural tissue is formed at the expense of paraxial that not all Wnt3a signaling in the hippocampus is mesoderm [5,6]. The domains of Lef1 and Tcf1 expres- mediated by LEF1. sion, although more widespread, often include the domains of Wnt3a expression. The sites where the Other TCF/LEF genes are expressed in this signaling domains of Wnt3a and Lef1 expression overlap include the field and it is likely that they are involved in Wnt3a area that gives rise to the hippocampus. The hippocampal signal transduction. The involvement of other phenotypes detected in Wnt3a [1] and Lef1 [2] mutant TCF/LEFs is further evidenced by the phenotype of a mice indicate that products of these genes act together in mutant in which a LacZ transgene is integrated in-frame certain aspects of hippocampus development. into the Lef1 gene; this lesion eliminates the HMG DNA- binding domain from the gene’s protein product, but The hippocampal phenotype of the Wnt3a mutant mice is leaves the β catenin-binding domain intact. Mice consistent with the gene’s domain of expression, which homozygous for this allele display a complete loss of hip- can be detected at early stages in the so-called cortical pocampal structures. It thus appears that the fusion hem — the longitudinal axis of which predicts the location protein has a dominant-negative action, though the mech- of hippocampus development. Although other Wnt genes anism of this is not yet known. More familiar dominant- are expressed in this domain, Wnt3a is the earliest to be negative forms of TCFs have been made by deletion of expressed in this structure, and so the most likely to be the β catenin-binding domain, resulting in non-productive involved in initial inductive processes. The Wnt3a mutant association with DNA [7]. mice appear to lack all hippocampal structures; in their place, the most medial cortical structure contains a broad An indication of how the Lef1–LacZ protein has domi- cortical plate, characteristic of neocortex, which abuts the nant-negative activity comes from the observation that the choroid plexus of the lateral ventricle (Figure 1). The level of Lef`1–LacZ protein in Lef`1–LacZ/Lef`1–LacZ choroid plexus of the lateral ventricles forms normally in homozygotes is much higher than in Lef`1–LacZ/+ het- both Wnt3a and Lef1 mutants. erozygotes, which are phenotypically normal. This differ- ence in expression level cannot be explained solely by the The phenotypes of mice carrying two different Lef1 one versus two Lef`1–LacZ alleles. The Lef1–LacZ protein mutant alleles indicate that multiple TCF/LEFs are shows dominant-negative activity only in those cells in involved in transduction of Wnt3a signals in the develop- which is highly expressed. As this protein is expected to ing hippocampus. One construct removes Lef1 activity be able to bind β catenin, but not DNA, its dominant-neg- entirely, and this was found to cause a phenotype in which ative property is most likely due to the ‘mopping up’ of granule cells of the dentate gyrus do not form, although cytoplasmic β catenin. This would prevent the activation bb10g06.qxd 04/03/2000 01:10 Page R281 Dispatch R281 of other TCFs, which would then associate with the tran- References scriptional repressor Groucho. Another interesting impli- 1. Lee SM, Tole S, Grove E, McMahon AP: A local Wnt-3a signal is required for development of the mammalian hippocampus. cation of the elevated Lef`1–LacZ expression level in Development 2000, 127:457-467. homozygotes is that Lef-1, or some other TCF, normally 2. Galceran J, Miyashita-Lin EM, Devaney E, Rubenstein JL, Grosschedl R: Hippocampus development and generation of dentate gyrus downregulates activity of the Lef1 promoter.