Discussion
In the present study we explored the appearance of Annexin A7 during mouse development at the mRNA and protein level and focused on the central nervous system during embryogenesis. Northern blot and immunohistochemistry analysis show the expression of Annexin A7 in all embryonic tissues from day E5 on, the earliest day studied. Differentiating cells, like the neural ectoderm, which is the origin of the cells belonging to the nervous system, show Annexin A7 immunoreactivity mainly in the cytosol. Endodermal as well as mesodermal cells exhibit a similar subcellular localization of the protein.
In the developing brain we noted a striking change in the subcellular distribution of Annexin A7. Cells in the stratum germinativum of the neopallial cortex, which surrounds the lateral ventricle, show at E13 and E15 a staining for Annexin A7 mainly in the cytosol. But this staining largely disappears and at the following day E16 we detect Annexin A7 in the nucleus of cells in the intermediate and marginal zone of the neopallium. The different expression patterns of Annexin A7-positive cells from the ventricular germinative zone to the marginal zone of the later neocortex observed in this study are similar to the developmental patterns of Tenascin-C-positive astroglial precursors following the guidance of the radial glial cells [18]. Moreover, the patterns resemble that of migrating and differentiated neurons described by Berry et al. [13]. Neurons that are generated prenatally in the proliferative ventricular layer of the neopallial cortex subsequently migrate through the intermediate zone to form the different cortical cell layers in a declining inside-out gradient of cell maturation. These observations suggested that the subcellular localization of Annexin A7 depends on developmental stage and cell type.
In the adult brain we generally observed a nuclear localization for Annexin A7 in neurons whereas astrocytes exhibited both a cytosolic as well as a nuclear staining. However, pyramidal neurons of the isocortex and Purkinje-cells of the cerebellum exhibited a cytosolic stain of intermediate intensity including their neurites and dendrites. We have previously reported the expression of Annexin A7 in human temporal neocortex and hippocampus obtained from neurosurgery for therapy-refractory epilepsy and found the two Annexin A7 isoforms restricted to the cytoplasm and nuclei of astrocytes, whereas neurons lacked any signal [19]. The hippocampal area showed typical signs of Ammon's horn sclerosis, but the adjacent temporal neocortex tissue did not show any histopathological alterations. This data is in discrepancy with our actual observation in the murine brain. To determine if this is due to different methods in immunofluorescence staining or indeed different expression patterns in mouse and human, we repeated the immunofluorescence of human brain. This time we investigated sections from human parietal cortex of autopsy brain. The astroglial expression of Annexin A7 could be confirmed, although not all astrocytes exhibited the nuclear staining. Pyramidal neurons however indicated a distinct staining of Annexin A7 most prominent along their dendrites. The different results obtained for the neurons may be explained by the fact that we previously used frozen tissue sections or that they were derived from patients suffering from temporal lobe epilepsia. On the other hand the actual human brain tissue used was from aged patients and did not correspond to the age of the mice included in this study. In cultured cells after cell damage or apoptosis (unpublished observations) or in cells treated with a Ca2+-ionophore Annexin A7 translocated from the cytoplasm to cellular membranes [19]. We therefore favor the hypothesis that Annexin A7 in the sensitive neurons of the human autopsy brain may have similarly translocated to the cellular membrane. This property of translocation and membrane binding is common to all annexins and commercially available kits for apoptosis detection employ recombinant AnnexinA5. The presence of nuclear Annexin A7 in murine brain was confirmed by controls using sections from the AnxA7-/- mouse and by a biochemical extraction of the protein from the nucleus. Both Annexin A7 isoforms described in brain tissue seemingly are expressed by neurons and astrocytes, which was shown using total cell extracts of cultured neuo-2a, PC-12, and C6 cells. In addition to their expression in brain, both Annexin A7 isoforms have only been described in heart muscle and red blood cells [5-8].
Although the inactivation of the annexin A7 gene did not interfere with the viability and development of knock out mice [10], their generation allowed us to address the role of Annexin A7 in specific cell types [8,9]. Indeed, when we analyzed primary astrocytes from an AnxA7-/- mouse for Ca2+-dependent signaling processes, we found that they exhibited a significantly increased velocity of mechanically induced astrocytic Ca2+-waves as compared to wild type [9]. This led us to propose, that Annexin A7 can act as a Ca2+-buffer and is involved in Ca2+-homeostasis. In neurons Ca2+ ions play major roles in various physiological and pathophysiological processes [20-25]. One can speculate about an involvement of Annexin A7 in the regulation of these Ca2+-dependent processes, propositions that need further investigation. However, such roles are confirmed for heart function by studies of Schrickel et al. [submitted], who described an involvement of Annexin A7 in the maintenance of a regular cardiac electrophysiology and Ca2+-homeostasis.
An altered subcellular location during embryogenesis was also reported for Annexin A11. The developing gray matter of the rat embryonic spinal cord exhibited primarily nuclear localization of Annexin A11, while immunoreactivity was lost from the nuclei in the adult spinal cord [26]. In contrast, our studies show a relocation of Annexin A7 from the cytosol to the nucleus in cells of the embryonic neuronal tissue. The Annexin A7 distribution is determined by a variety of factors. An important one appears to be Ca2+, which promotes binding of Annexin A7 to membranes and also allows aggregation of annexins [27]. Binding partners of Annexin A7 such as sorcin might represent additional factors [28]. Although the members of the annexin family are generally found at the plasma membrane, in the cytoplasm or in association with the cytoskeleton, Annexins A1, A2, A4, A5 and A11 have been described to be localized at least partially in the nucleus [26,29,30]. Studies with human foreskin fibroblasts demonstrated that Annexin A1, A4 and A5 are all present in the nucleus at higher concentration than in the cytosol [31]. Raising intracellular Ca2+ led to relocation of these annexins to the nuclear membrane. An important role for annexins in mediating the Ca2+-signal within the nuclei of the fibroblasts was proposed. These results mirror studies with stably transfected C6 cells, in which high intracellular Ca2+-concentrations induced a marked redistribution of Annexin A7 from its localization in the nucleoplasm to the nuclear membrane [19].
None of the annexins contains a typical nuclear localization signal and their mechanism of nuclear import remains to be elucidated. For Annexin A11 it was shown that nuclear localization is mediated by its N-terminal region, which also contains a binding side for the S100 protein calcyclin [29]. More recently Tomas and Moss [32] showed, that Annexin A11 and S100A6 assemble at the nuclear envelope during nuclear breakdown. Their role in this process is not known. In general it seems that the nuclear localization of the annexins is actively regulated. For example the nucleocytoplasmic compartmentalization of Annexin A2 is controlled by sequestration of the AnxA2/p11 complex modulated by phosphorylation and by a nuclear export signal found in the AnxA2 3–12 region [33]. One function of Annexin A2 in the nucleus, that appears not to involve binding of p11, has been suggested by its purification as part of a primer recognition protein complex that enhances DNA polymeraseα-activity in vitro [34]. The annexins may participate in a nuclear response to initial cell stimulation or to a Ca2+-transient, presumably by regulating DNA replication. For Annexin A7 such a pathway is very speculative at the moment. Future studies are however directed by these findings and will concentrate on the identification of the nuclear localization signal of Annexin A7 as well as on the role of Annexin A7 in the nuclear compartment.