Exogenous basement-membrane-like matrix stimulates adrenergic development in avian neural crest cultures

The development of quail trunk neural crest cultures was dramatically altered when the cultures were overlaid with a gel of reconstituted basement membrane (RBM) components derived from the Engelbreth-Holm-Swarm sarcoma. In the presence of the RBM gel overlay, the number of catecholamine-positive (CA+) cells that developed was increased 50-fold, while the final number of melanocytes and total cells was only half that seen in the control cultures. The presence of the RBM gel overlay did not alter the time of onset of differentiation of the CA+ cells or melanocytes. The stimulation of CA+ cell number was not observed with type IV collagen substrates, laminin substrates or type I collagen gel overlays with or without added laminin. The stimulation of CA+ cell development was dependent on initial plating density. The number of CA+ cells that developed in the presence of the RBM gel was proportional to the initial plating density at 80-320 cells mm-2, whereas no CA+ cells were observed below 20 cells mm-2 and only a few CA+ cells were detected at 40 cells mm-2. There was, however, extensive cell division and differentiation of melanocytes and unpigmented cells at the lower initial plating densities. When the RBM gel was used as a substrate, rather than as an overlay, a striking rearrangement of cells into interconnected strands was observed. After several days in culture, melanocytes, CA+ cells and unpigmented cells were present in these strands. These results indicate that molecules associated with a reconstituted basement-membrane-like matrix are a potent stimulatory influence on adrenergic development and also act to inhibit the production of other cell types in neural crest cultures.     

Substratum effects on cell dispersal, morphology, and differentiation in cultures of avian neural crest cells

Adhesive extracellular matrix (ECM) molecules appear to play roles in the migration of neural crest cells, and may also provide cues for differentiation of these cells into a variety of phenotypes. We are studying the influences of specific ECM components on crest differentiation at the levels of both individual cells and cell populations. We report here that the glycoproteins fibronectin and laminin differentially affect melanogenesis in cultures of avian neural crest-derived cells. Clusters of neural crest cells were allowed to form on explanted neural tubes for 24 and 48 hr, and then subcultured on uncoated glass coverslips or coverslips coated with fibronectin or laminin. The morphology of cells varied on the three substrata, as did patterns of cell dispersal. Crest cells dispersed most rapidly and extensively on fibronectin. In contrast, cells on laminin dispersed initially, but then assumed a stellate morphology and rapidly formed small aggregates. Cell dispersal was minimal on glass substrata, resulting in a uniformly dense distribution. These patterns of dispersal were similar in subcultures of both 24- and 48-hr clusters, although dispersal of cells from older clusters was less extensive. The rate and extent of melanogenesis correlated with patterns of cell dispersal. Cell from 24-hr clusters underwent melanogenesis significantly more slowly on fibronectin than on the other two substrata. Pigment cells began to differentiate by 2 days of subculture in the cell aggregates on laminin and in the dense centers of cultures on untreated glass. By 5 days, there was significantly more melanogenesis in cultures on laminin and glass than on fibronectin substrata. Melanogenesis in cultures of 48-hr clusters was more rapid and extensive on control (glass) substrata than on fibronectin or laminin, correlating with reduced cell dispersal. We conclude that fibronectin and laminin, which are found along neural crest migratory pathways in vivo, can affect melanogenesis in vitro by regulating patterns of cell dispersal.     

Effects of mesodermal tissues on avian neural crest cell migration

We have used microsurgical techniques to investigate the effects of embryonic mesodermal tissues on the pattern of chick neural crest cell migration in the trunk. Segmental plate or lateral plate mesenchyme was transplanted into regions encountered by neural crest cells. We found that neural crest cells are able to migrate through lateral plate mesenchyme but not through segmental plate tissue until this tissue differentiates into a sclerotome. After this stage, segmental migration is controlled by the subdivision of the sclerotome into a rostral and a caudal half; when the rostrocaudal orientation of the sclerotomes is reversed by rotating the segmental plate 180 degrees about its rostrocaudal axis, neural crest cells migrate through the portion of the sclerotome that was originally rostral.     

Differentiation of avian neural crest cells in vitro: Absence of a developmental bias toward melanogenesis

Summary Neural crest cells from quail embryos grown in standard culture dishes differentiate almost entirely into melanocytes within 4 or 5 days when chick embryo extract (CEE) or occasional lots of fetal calf serum (FCS) are included in the medium. Gel fractionation showed that the pigment inducing factor(s) present in these media is of high molecular weight (> 400 K daltons). In the absence of CEE, the neural tube can also stimulate melanocyte differentiation. Culture medium supplemented by selected lots of FCS permits crest cell proliferation but little overt differentiation after up to 2 weeks in culture if the neural tube is removed within 18 h of explantation in vitro. Subsequent addition of CEE to such cultures promotes complete melanocyte differentiation. Crest cells from White leghorn chick embryos also differentiate into melanocytes in the presence of CEE, but do not survive well in its absence. Melanocyte differentiation of crest cells from both quail and chick embryos can by suppressed by culturing under a dialysis membrane, even in the presence of the neural tube and CEE, but neuronal differentiation appears greatly enhanced.     

Pathways of avian neural crest cell migration in the developing gut

The NC-1 and E/C8 monoclonal antibodies recognize a similar population of neural crest cells as they migrate from vagal levels of the neural tube and colonize the branchial arch region of 2- to 3-day-old chicken embryos. Some of these immunoreactive cells then appear to enter the gut mesenchyme on the third day of incubation just caudal to the third branchial cleft. After entering the gut, these cells migrate in a rostral-caudal direction, using primarily the superficial splanchnic mesodermal epithelium of the gut as a substratum. The antigen-positive cells remain preferentially associated with the splanchnopleure. Few antigenic cells enter the mesenchyme surrounding the endoderm at anterior levels whereas they are found throughout the mesenchyme when nearing the umbilicus. At postumbilical levels, immunoreactive cells are distributed on both sides of the differentiating muscle layer but not within it. Although fibronectin immunoreactivity can be found throughout the wall of the gut, there is no apparent relationship between the distribution of fibronectin and the location of the immunoreactive cells. These results suggest that a mechanism more complex than a mere interaction with fibronectin may account for migration of crest-derived cells in the gut.     

Control of neural crest cell dispersion in the trunk of the avian embryo

Many hypotheses have been advanced to explain the orientation and directional migration of neural crest cells. These include positive and negative chemotaxis, haptotaxis, galvanotaxis, and contact inhibition. To test directly the factors that may control the directional dispersion of the neural crest, I have employed a variety of grafting techniques in living embryos. In addition, time-lapse video microscopy has been used to study neural crest cells in tissue culture. Trunk neural crest cells normally disperse from their origin at the dorsal neural tube along two extracellular pathways. One pathway extends laterally between the ectoderm and somites. When either pigmented neural crest cells or neural crest cells isolated from 24-hr cultures are grafted into the space lateral to the somites, they migrate: (1) medially toward the neural tube in the space between the ectoderm and somites and (2) ventrally along intersomitic blood vessels. Once the grafted cells contact the posterior cardinal vein and dorsal aorta they migrate along both blood vessels for several somite lengths in the anterior-posterior axis. Neural crest cells grafted lateral to the somites do not immediately move laterally into the somatic mesoderm of the body wall or the limb. Dispersion of neural crest cells into the mesoderm occurs only after blood vessels and nerves have first invaded, which the grafted cells then follow. The other neural crest pathway extends ventrally alongside the neural tube in the intersomitic space. When neural crest cells were grafted to a ventral position, between the notochord and dorsal aorta, in this intersomitic pathway at the axial level of the last somite, the grafted cells migrate rapidly within 2 hr in two directions: (1) dorsally, in the intersomitic space, until the grafted cells contact the ventrally moving stream of the host neural crest and (2) laterally, along the dorsal aorta and endoderm. All of the above experiments indicate that neither a preestablished chemotactic nor adhesive (haptotactic) gradient exists in the embryo since the grafted neural crest cells will move in the reverse direction along these pathways toward the dorsal neural tube. For the same reason, these experiments also show that dispersal of the neural crest is not directed passively by other environmental controls, since the cells can clearly move counter to their usual pathway and against such putative passive mechanisms.(ABSTRACT TRUNCATED AT 400 WORDS)     

Molecular mechanisms of avian neural crest cell migration on fibronectin and laminin

We have examined the molecular interactions of avian neural crest cells with fibronectin and laminin in vitro during their initial migration from the neural tube. A 105-kDa proteolytic fragment of fibronectin encompassing the defined cell-binding domain (65 kDa) promoted migration of neural crest cells to the same extent as the intact molecule. Neural crest cell migration on both intact fibronectin and the 105-kDa fragment was reversibly inhibited by RGD-containing peptides. The 11.5-kDa fragment containing the RGDS cell attachment site was also able to support migration, whereas a 50-kDa fragment corresponding to the adjacent N-terminal portion of the defined cell-binding domain was unfavorable for neural crest cell movement. In addition to the putative "cell-binding domain," neural crest cells were able to migrate on a 31-kDa fragment corresponding to the C-terminal heparin-binding (II) region of fibronectin, and were inhibited in their migration by exogenous heparin, but not by RGDS peptides. Heparin potentiated the inhibitory effect of RGDS peptides on intact fibronectin, but not on the 105-kDa fragment. On substrates of purified laminin, the extent of avian neural crest cell migration was maximal at relatively low substrate concentrations and was reduced at higher concentrations. The efficiency of laminin as a migratory substrate was enhanced when the glycoprotein occurred complexed with nidogen. Moreover, coupling of the laminin-nidogen complex to collagen type IV or the low density heparan sulfate proteoglycan further increased cell dispersion, whereas isolated nidogen or the proteoglycan alone were unable to stimulate migration and collagen type IV was a significantly less efficient migratory substrate than laminin-nidogen. Neural crest cell migration on laminin-nidogen was not affected by RGDS nor by YIGSR-containing peptides, but was reduced by 35% after addition of heparin. The predominant motility-promoting activity of laminin was localized to the E8 domain, possessing heparin-binding activity distinct from that of the N-terminal E3 domain. Migration on the E8 fragment was reduced by greater than 70% after addition of heparin. The E1' fragment supported a minimal degree of migration that was RGD-sensitive and heparin-insensitive, whereas the primary heparin-binding E3 fragment and the cell-adhesive P1 fragment were entirely nonpermissive for cell movement.(ABSTRACT TRUNCATED AT 400 WORDS)     

Dynamic alterations in gene expression after Wnt-mediated induction of avian neural crest

The Wnt signaling pathway is important in the formation of neural crest cells in many vertebrates, but the downstream targets of neural crest induction by Wnt are largely unknown. Here, we examined quantitative changes in gene expression regulated by Wnt-mediated neural crest induction using quantitative PCR (QPCR). Induction was recapitulated in vitro by adding soluble Wnt to intermediate neural plate tissue cultured in collagen, and induced versus control tissue were assayed using gene-specific primers at times corresponding to premigratory (18 and 24 h) or early (36 h) stages of crest migration. The results show that Wnt signaling up-regulates in a distinct temporal pattern the expression of several genes normally expressed in the dorsal neural tube (slug, Pax3, Msx1, FoxD3, cadherin 6B) at "premigratory" stages. While slug is maintained in early migrating crest cells, Pax3, FoxD3, Msx1 and cadherin 6B all are down-regulated by the start of migration. These results differ from the temporal profile of these genes in response to the addition of recombinant BMP4, where gene expression seems to be maintained. Interestingly, expression of rhoB is unchanged or even decreased in response to Wnt-mediated induction at all times examined, though it is up-regulated by BMP signals. The temporal QPCR profiles in our culture paradigm approximate in vivo expression patterns of these genes before neural crest migration, and are consistent with Wnt being an initial neural crest inducer with additional signals like BMP and other factors maintaining expression of these genes in vivo. Our results are the first to quantitatively describe changes in gene expression in response to a Wnt or BMP signal during transformation of a neural tube cell into a migratory neural crest cell.     

Extracellular matrix from normal but not Steel mutant mice enhances melanogenesis in cultured mouse neural crest cells

The Steel mutation is a non-cell-autonomous defect in mice that affects the development of several stem cell populations, including germ cells, hematopoietic cells, and neural crest-derived pigment cells. To characterize the environmental lesion caused by the Steel mutation, we have compared the ability of normal and mutant extracellular matrix material to support the differentiation of normal mouse neural crest cells in vitro. Extracellular matrix deposited by cultured skin cells isolated from normal fetuses enhanced melanogenesis by crest cells over that observed on plastic substrata. In contrast, matrix material produced by Steel-Dickie (Sld) fetal skin cells failed to enhance melanogenesis. Adrenergic differentiation by neural crest-derived cells was promoted equally by both normal and mutant extracellular matrix compared to control substrata. We conclude that the environmental defect in mutant embryos selectively affects a melanogenic subpopulation of neural crest cells and resides, at least in part, in the extracellular matrix.     

Distribution of fibronectin in the early phase of avian cephalic neural crest cell migration

Immunofluorescence and immunoperoxidase labeling for fibronectin was used to study the early events of cephalic neural crest cell migration in avian embryos. Prior to crest cell appearance, fibronectin was associated with the basement membranes of all tissues. The loose mesenchymal cells were also surrounded by this glycoprotein. The crest cell individualization phase included a transient rounding up and a rapid increase in cell number in a very limited space. Whereas the neural tube basement membrane was not formed dorsally at the site of emergence of crest cells, it was partially fused laterally with the ectoderm basement membrane apparently preventing immediate crest cell emigration. Further increase in cell number occurred concomitantly with their penetration between the two developing basement membranes of the neural tube and the ectoderm. The localization of migrating crest cells is apparently greatly influenced by local interactions between the ectoderm and the neural tube, whose morphogenesis differs considerably at each axial level: at the mesencephalic-rhombencephalic levels, crest cells rapidly reached a cell-free space that was mostly devoid of fibronectin. Further migration occurred laterally in that space while pioneer crest cells became surrounded by fibronectin in their environment. Crest cells progressed as a confluent multicellular layer with an apparent velocity of 70 μm/hr. At the prosencephalic and median rhombencephalic levels, crest cells accumulated between the fibronectin-rich basement membranes of the ectoderm and the neural tube. Pioneer crest cells were arrested at the site of attachment of the ectoderm and the neural tube basement membranes (i.e., optic vesicles and otic placodes). Crest cells resumed their migration when more space became available during the constriction of the optic vesicles and the invagination of the otic placodes.     

Pathways and mechanisms of avian trunk neural crest cell migration and localization

Crest cells individualized at the dorsal border of the neural tube, while they became surrounded by a fibronectin-rich matrix. Crest cells initiated their migration between the basement membranes of the neural tube and the ectoderm. In the vagal region, crest cells migrated in a fibronectin-rich environment between the ectoderm and the dermomyotome, very rapidly reaching the apex of the pharynx. In the trunk region, crest cells opposite the bulk of the somite accumulated at the junction between the somite, the neural tube, and the ectoderm; they resumed their migration at the onset of the dissociation of the somite into dermomyotome and sclerotome. Migration occurred more ventrally along the neural tube; nevertheless, the formation of the rapidly expanding sclerotome prevented crest cells from reaching the paranotochordal region. Thereafter, crest cells accumulated between the neural tube, the dermomyotome, and the sclerotome, where ultimately they formed the dorsal root ganglia. In contrast, cells opposite the intersomitic space did not encounter these obstacles and utilized a narrow pathway formed between the basement membranes of the two adjacent somites. This pathway allowed crest cells to reach the most ventral regions of the embryo very rapidly; they accumulated along the aorta to form the aortic plexuses, the adrenal medulla, and the sympathetic ganglia. The basic features of the migration pathways are (1) a strict delimitation by the fibronectin-rich basement membranes of the surrounding tissues, (2) a formation of space concomitant with the migration of crest cells, (3) a transient existence: continued migration is correlated with the presence of fibronectin, whereas cessation is correlated with its focal disappearance. The crest cells are characterized by their inability to traverse basement membranes and penetrate within tissues. We propose that the combination of active proliferation, unique motility properties, and the presence of narrow pathways are the major mechanisms ensuring correct directionality. Morphologically defined transient routes of migration along with developmentally regulated changes in the extracellular matrix and in the adhesive properties of crest cells are most probably involved in their stabilization in defined territories and their aggregation into ganglia.     

The avian neural crest as a model system for the study of cell lineages

Under the influence of environmental factors, the neural crest gives rise to numerous cell types and is therefore, by definition, a pluripotential structure. However, it was not clear until recently to what extent each individual neural crest cell possessed multiple capacities for differentiation. As a result of in vivo and in vitro approaches aimed at solving this problem, it has become apparent that the neural crest is made up of cells in different states of determination and that some lineages are segregated very early. In particular, analysis of clones obtained from single cells grown in culture has shown that, although many individual neural crest cells are pluripotential to varying degrees, others are apparently committed to give rise to only one derivative. The role of the embryonic microenvironment in the emergence of phenotypic diversity is probably complex, certain factors acting to promote the survival of selected subpopulations of fully determined progenitors, while others may direct partly committed precursors towards a specific developmental fate.     

Morphogenesis of sclerotome and neural crest in avian embryos

Summary The distribution of sclerotome and neural crest cells of avian embryos was studied by light and electron microscopy. Sclerotome cells radiated from the somites towards the notochord, to occupy the perichordal space. Neural crest cells, at least initially, also entered cell-free spaces. At the cranial somitic levels they moved chiefly dorsal to the somites, favouring the rostral part of each somite. These cells did not approach the perichordal space. More caudally (i.e. trunk levels), neural crest cells initially moved ventrally between the somites and neural tube. Adjacent to the caudal half of each somite, these cells penetrated no further than the myosclerotomal border, but opposite the rostral somite half, they were found next to the sclerotome almost as far ventrally as the notochord. However, they did not appear to enter the perichordal space, in contrast to sclerotome cells.When tested in vitro, sclerotome cells migrated towards notochords co-cultured on fibronectin-rich extracellular material, and on collagen gels. In contrast, neural crest cells avoided co-cultured notochords. This avoidance was abolished by inclusion of testicular hyaluronidase and chondroitinase ABC in the culture medium, but not by hyaluronidase from Streptomyces hyalurolyticus. The results suggest that sclerotome and neural crest mesenchyme cells have a different distribution with respect to the notochord, and that differential responses to notochordal extracellular material, possibly chondroitin sulphate proteoglycan, may be responsible for this.     

Hensen's node regulates avian neural crest differentiation in vitro

Different anteroposterior (AP) regions of the neural crest normally produce different cell types, both in vivo and in vitro. AP differences in neural crest cell fates appear to be specified in part by mechanisms that act prior to neural crest cell migration. We, therefore, examined the possibility that the fates of neural crest cells, like those of neural tube cells, can be regulated by interactions with Hensen's node. Using a transfilter co-culture system, we found that young (stage 3+ to 4) Hensen's node up-regulates the expression of two cranial-specific phenotypes (fibronectin and smooth muscle actin immunoreactivities) in mass cultures of trunk neural crest cells, and down-regulates the expression of a trunk-specific phenotype (melanin synthesis). The changes in phenotype produced by exposure to young Hensen's node were not accompanied by changes in the proliferation of either fibronectin immunoreactive cells or melanocytes. The capacity of Hensen's node to elicit changes in trunk neural crest cell phenotype decreased as the developmental age of the node increased and was lost by stage 6. In addition, old Hensen's node did not stimulate the expression of trunk-specific phenotypes in cranial neural crest cells, suggesting that cranial- and trunk-specific phenotypes are induced by different mechanisms. © 1996 John Wiley & Sons, Inc.     

Ethanol-responsive gene expression in neural cell cultures.

We have studied the molecular mechanisms underlying neuronal adaptation to chronic ethanol exposure. NG108-15 neuroblastoma cells were used to perform a detailed analysis of ethanol-induced changes in neuronal gene expression. High resolution, quantitative two-dimensional (2-D) gel electrophoresis of in vitro translation products showed both dose-dependent increases and decreases in specific mRNA abundance following treatment with ethanol at concentrations seen in actively drinking alcoholics (50-200 mM). Dose response curves for representative members of the increasing or decreasing response groups had very similar profiles, suggesting that similar mechanisms may regulate members of a response group. Some mRNAs that increased with ethanol treatment appeared identical to species induced by heat shock while other mRNAs were only induced by ethanol. We conclude that chronic ethanol exposure can produce specific coordinate changes in expression of neuronal mRNAs, including some members of the stress protein response. However, the overall pattern of ethanol-responsive gene expression is distinct from the classical heat shock subgroup of stress proteins response. Changes in gene expression and specifically, mechanisms regulating a subset of stress protein expression, could be an important aspect of neuronal adaptation to chronic ethanol seen in alcoholics.     

Physical influences on neural crest cell migration in avian embryos: contact guidance and spatial restriction

Several ideas on how neural crest (NC) cell migration in bird embryos might be dependent on the physical qualities of the internal embryonic environment were studied. Contact guidance has been suggested to direct NC cells ventrally in the trunk, but this has been subject to doubt (see Newgreen and Erickson, 1986, Int. Rev. Cytol. 103, 118-119). On reexamination, in situ extracellular matrix (ECM) and cell processes on the medial face of the somites were found appropriately oriented for this function. In addition, tissue culture models of oriented ECM could induce orientation of NC cells which mimicked that observed in the embryo. It is concluded that in this situation, oriented structures contribute to directed migration of NC cells in vivo, but the mechanism of contact guidance (i.e., steric or adhesive guidance) could not be ascertained. Contact guidance, in the form of steric guidance, has also been suggested as limiting ventrad NC cell movement at the midbrain level due to an insurmountable ridge on the side of the midbrain. The presence of this ridge was confirmed but it is unlikely to be responsible for prevention of ventrad migration, because, although it subsides very rapidly, the cells still refuse to move ventrad, and because models of this ridge in vitro proved to be no obstacle to NC cells. NC cell migration is also described as being limited by gross space between other organs or tissues. In vitro, NC cells could penetrate Nucleopore filters with pore diameters of 0.86 micron or greater. Observation of cell-free spaces in embryos showed that these were almost all much larger than the minimum pore size established experimentally. It is therefore concluded that in general the dimensions of gross tissue spaces probably do not set important limits for NC cell migration, but that the dimensions of transiently distensible microspaces between ECM fibrils may be a critical physical parameter.     

Transforming growth factor-beta alters differentiation in cultures of avian neural crest-derived cells: effects on cell morphology, proliferation, fibronectin expression, and melanogenesis.

Neural crest cell differentiation is responsive to a variety of extrinsic signals that include extracellular matrix (ECM) molecules and growth factors. Transforming growth factor-beta (TGF-beta) has diverse, cell type-specific effects, many of which involve regulation of synthesis of ECM molecules and their cell surface receptors. We are studying both separate and potentially interrelated influences of ECM and growth factors on crest differentiation and report here that TGF-beta alters several aspects of crest cell behavior in vitro. Clusters of quail neural crest cells were cultured in the presence and absence of 400 pM TGF-beta 1 and examined at 1, 3, and 5 days. When examined at 5 days, there was a dramatic decrease in the number of melanocytes in treated cultures, regardless of the onset or duration of TGF-beta treatment. With continuous TGF-beta treatment, or with treatment only during crest cluster formation on explanted neural tubes, many cells increased in area, becoming extremely flat. These changes were evident beginning on Day 3. While quantitative analyses of video images documented the size increase, several aspects of motility were relatively unchanged. Synthesis of fibronectin (FN) by approximately 11% of cells on Day 3 and 31% of cells on Day 5 was demonstrated by immunocytochemistry and was associated with a sixfold increase in FN mRNA by Day 5. Experiments which correlated FN immunoreactivity with incorporation of bromodeoxyuridine suggested that the population of large, flat, FN-positive cells did not proliferate selectively and that there was a slower rate of proliferation in TGF-beta-treated cultures than in untreated cultures. The large FN-immunoreactive cells resemble cells derived from cephalic neural crest and raise interesting questions concerning potential roles for TGF-beta in regulating crest differentiation in vivo.     

PTEN基因对早期胚胎神经嵴细胞迁移的作用研究

目的 初步探讨PTEN基因在早期神经嵴细胞迁移中的作用.方法 首先胚胎整体的原位杂交和免疫荧光方法检测鸡胚胎内源性的PTEN基因及蛋白水平的表达情况;其次,利用鸡胚胎体内半侧神经管转染的方法,使神经管一侧PTEN基因过表达,对侧神经管为正常对照侧;最后,通过Pax7的整体胚胎免疫荧光表达观察PTEN基因对其标记的部分神经嵴细胞迁移的影响.结果 内源性PTEN基因在mRNA和蛋白水平表达显示,其在早期胚胎HH4期的神经板即开始明显的表达;通过半侧过表达PTEN基因后观察到过表达PTEN基因侧的头部神经嵴细胞迁移与对照侧相比明显受到抑制,但对躯干部的影响并不明显.结论 PTEN基因可能抑制早期胚胎头部神经嵴细胞的迁移.