J Physiol Volume 585, Number 3, 711-719, December 15, 2007 DOI: 10.1113/jphysiol.2007.138776
Multiphoton imaging of chick retinal development in relation to gap junctional communication
David L. Becker1,
Kevin F. Webb1,
Christopher Thrasivoulou1,
Chih-Chi Lin1,
Roxana Nadershahi1,
Niki Tsakiri1 and
Jeremy E. Cook1
1 Department of Anatomy and Developmental Biology, University College London, Gower Street, London WC1E 6BT, UK
 |
Abstract
|
|---|
Neural progenitor cells in the developing retina extend processes that stretch from the basal vitread surface to the apical ventricular surface. During the cell cycle, the nucleus undergoes interkinetic nuclear migration (INM), moving in a vitread direction during G1, passing through S-phase at its peak and then, on entering G2, returning towards the ventricular surface where it enters M-phase and divides. We have previously shown that individual saltatory movements of the nucleus correlate with transient changes in cytosolic calcium concentration within these progenitor cells and that these events spread to neighbouring progenitors through connexin43 (Cx43) gap junction channels, thereby coordinating the migration of coupled clusters of cells. Disrupting coupling with pharmacological agents, Cx43-specific antisense oligodeoxynucleotides (asODNs) or dominant negative Cx43 (dnCx43) inhibits the sharing of calcium events, reducing the number that each cell experiences and significantly slowing INM. We have developed protocols for imaging migrating progenitor cells by confocal microscopy over relatively short periods, and by multiphoton microscopy over more extended periods that include complete cell cycles. We find that perturbing gap junctional communication not only slows the INM of progenitor cells but also apparently prevents them from changing direction at critical phases of the cell cycle. It also disrupts the migration of young neurons to their appropriate layers after terminal division and leads to their ectopic differentiation. The ability to perform extended time-lapse imaging over 3D volumes in living retina using multiphoton microscopy should now allow fundamental mechanisms governing development of the retinal neuroepithelium to be probed in detail.
(Received 16 June 2007;
accepted after revision 9 October 2007;
first published online 11 October 2007)
Corresponding author D. Becker: Department of Anatomy and Developmental Biology, University College London, Gower Street, London WC1E 6BT, UK. Email: d.becker{at}ucl.ac.uk
|
Introduction
|
|---|
In order to create a fully functional eye in just 21 days, the chick embryo needs to generate about 200 million neurons for each retina within about 10 days, have them migrate to the correct layers and then wire up the neuronal circuits appropriately. As in all parts of the developing central nervous system, co-ordination of this complex process requires communication between neural progenitor cells, and one of the key routes of communication is through gap junction channels. Neural progenitor cells in the developing chick retina are coupled together by channels constructed from connexin43 (Cx43) (Becker et al. 2002). This coupling facilitates the sharing of transient changes in cytosolic calcium levels (calcium events) as waves that pass between progenitor cells as they proliferate (Pearson et al. 2004, 2005). The nucleus of each neural progenitor cell usually undergoes a saltatory movement each time there is a calcium wave, moving in a vitread direction during G1, and then towards the ventricular zone in G2 where mitosis will take place (Fig. 1
). If the daughter cells of this mitosis remain as progenitors they will go through the cycle of interkinetic nuclear migration (INM) again. If either of them exits the cell cycle and differentiates, it will migrate towards the appropriate layer of the retina. In each case, its initial direction of movement will be vitread.
View larger version (131K):
[in this window]
[in a new window]
|
Figure 1. Confocal microscope image of a section through an E5 chick retina
Nuclei are stained red with propidium iodide and gap junction hemichannels are stained green. The phases of the cell cycle and the direction of interkinetic nuclear migration (INM) in each phase are marked (M, G1, S and G2) on the proliferating neuroepithelial cells. VZ marks the ventricular zone surface adjoining the retinal pigment epithelium (RPE). Scale bar, 25 µm.
|
|
We have previously shown, using conventional laser scanning confocal microscopy to image such cells for short periods, that if Cx43 protein levels are reduced with Cx43-specific asODNs or if gap junctional communication is impaired by application of pharmacological blocking agents or transfection with a dominant negative form of Cx43 then the rate of movement of the neural progenitor cells is significantly slowed. As INM is intimately linked to the phases of the cell cycle, cell proliferation is impeded, and this can result in the formation of abnormally small eyes (Becker & Mobbs, 1999; Pearson et al. 2005). A similar positive correlation between gap junctional coupling and speed of cell migration has been reported for a variety of embryonic cells such as neural crest (Huang et al. 1998; Bannerman et al. 2000). It also appears that calcium transients passing through the gap junctions drive the movement of both developing neurons (Komuro & Rakic, 1996, 1998; Komuro et al. 2001) and neural progenitor cells (Weissman et al. 2004; Pearson et al. 2002, 2004, 2005).
In the present study, we have begun to examine the long-term consequences of reduced gap junctional communication on cell-cycle exit and the migration of young postmitotic neurons to the appropriate layer of the retina. We have used multiphoton microscopy to image the dynamic events of proliferation, migration and differentiation and their relationships to gap junctional communication between neural progenitor cells of the chick retina over periods of 14–18 h during which they may pass through two full cell cycles. Imaging of retinal development by conventional laser scanning confocal microscopy is fairly straightforward for short periods such as 2–3 h but long-term imaging requires several modifications to the technique. The use of multiphoton excitation is important for reducing phototoxicity, but it generates its own problems in respect of energy absorption. Tissue substrates, culture media and methods of labelling cells all need to be chosen with these problems in mind. Temperature control, the gassing of media, prevention of medium evaporation, the loss of signal with depth and the handling and analysis of very large data sets are additional problems that need to be overcome for successful long-term imaging of neuronal development.
|
Methods
|
|---|
Tissue preparation
Fertilized eggs from White Leghorn or Rhode Island Red chickens were incubated at 38°C. On embryonic day 5 (E5) the embryos were removed and decapitated, and the retinae were dissected free in Krebs solution (Pearson et al. 2005). Retinae were mounted on white Millipore (Watford, UK) filters and labelled ballistically with either DiI, Oregon Green or pIRES-eGFP, using a Helios (Bio-Rad, Hemel Hempstead, UK) gene gun (Gan et al. 2000; Pearson et al. 2005). Retinae were cultured in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal calf serum but without Phenol Red, at 36°C, with 5% CO2. On some occasions, whole-mount retinae were loaded with calcium indicators such as Fluo4-AM or Rhod-2-AM (Molecular Probes/Invitrogen, Paisley, UK) for 30 min at 36°C prior to imaging. Immunostaining for connexins and or neuronal markers was carried out on paraformaldehyde fixed whole-mounted or cryosectioned retinae as described by Becker et al. (1995). Either propidium iodide or bis-benzimide (Sigma, Poole, UK) was often used as a nuclear counterstain.
Imaging
Multiphoton imaging of retinae was performed on a Leica SP2MP microscope (Leica, Milton Keynes, UK) with a Spectra-Physics (Didcot, Oxon) Millennia 8 W pump laser driving a Tsunami Ti:Sapphire laser at 840 nm, set to deliver 10–70 mW of average power to the specimen according to the brightness of the fluorescence signal. Detection of the signal was via the non-de-scanned detectors, where green and red signals were separated out. The microscope utilized a Gibraltar fixed stage and was housed inside a Ludin chamber which maintained the temperature at 37°C with a delivery of 5% CO2 to the specimen. The specimen was kept stable at the bottom of a 55 mm culture dish with a platinum harp, and viewed using a x40 (0.8 NA) or x63 (0.9 NA) ceramic dipping objective (Fig. 2
). Once the specimen was visualized, ultra-pure paraffin oil (BDH Laboratory Supplies, Poole, UK) was added to the top of the preparation to prevent evaporation of the culture medium. Images from 4D data sets were viewed using either the Imaris (Bitplane, Zurich, Switzerland) or Volocity (Improvision, Coventry, UK) software packages. A key consideration and limitation of analysis is the handling of very large data sets. Typically, imaging through a 100 µm thick retina in 1 µm steps with a 512 x 512 pixel resolution in a single spectral channel every 10 min (as required to see fine detail and rapid growth) generates around 150 MB of data per hour. Thus, during a single experiment of 14–18 h, several gigabytes (GB) of voxel data are generated, which even the fastest of current PCs (with 16 GB RAM and quadruple 64-bit 3 GHz processors) finds difficult to handle as a single data set. Segmentation of data sets to follow small clusters of cells is required in order to render cells and track them in 4D.
View larger version (37K):
[in this window]
[in a new window]
|
Figure 2. Electroporation and microscope set-up
A, a manipulator is used to bring two gold-plated plates into close apposition, sandwiching the retina on its Millipore filter (B) and making a circuit by contacting the 1 µl of plasmid with 1% Fast Green placed on the retinal surface. A series of 5 pulses of 50 V for 5 ms per pulse is delivered to the retina and Fast Green can be seen to enter the tissue. C, diagram of the imaging set-up showing the dipping objective focused on the retina mounted on a white Millipore filter, held in place by a platinum harp. Phenol Red free DMEM covers the specimen and a layer of paraffin oil prevents evaporation of the medium whilst allowing its gassing.
|
|
Manipulation of gap junctional communication
Gap junctional communication was perturbed in several ways. Carbenoxolone (CBX) (100 µM), Flufenamic acid (FFA; 100 µM) and 18-
-glycyrrhetinic acid (18--GA; 50 µM) were purchased from Sigma. Inhibitors of gap junctional communication are notoriously ‘dirty’ compounds, which often produce other undesired effects in cells; for example, CBX is known to inhibit Ca2+ channels at micromolar concentrations (Vessey et al. 2004; Cadetti & Thoreson, 2006). To guard against these non-specific effects, use of the two different pharmacological agents was complemented by more specific molecular methods targeted to reduce connexin expression. Cx43-specific antisense oligodeoxynucleotides (asODNs) were delivered in 30% PluronicTM gel (Sigma) as described in detail by Becker et al. (1999). This typically produces a 95% knockdown of Cx43 protein within 2 h. Transfection of the retinae with pIRES vectors expressing eGFP alone or eGFP with wild-type (wt) or dominant negative (dn) Cx43 was achieved either ballistically, by coating 1 µm gold particles with vectors and delivering them into the ventricular surface of the tissue with the Helios gene gun, or by electroporation (Becker et al. 2001 and Fig. 2). Expression of the vectors and maturation of the eGFP required about 5–6 h, and the transfected cells were imaged either overnight or on the next day.
|
Results
|
|---|
Technical considerations
Using multiphoton excitation at 840 nm, we achieved good excitation of all of our dyes. However, we still suffered from signal drop-off as we focused deeper than 50 µm into the retina. This could be reliably overcome by increasing the gain of the photomultiplier tube (PMT) at progressively greater depths to keep the detected signal constant, automatically repeating this at each time point using the depth attenuation software embedded in our Leica system.
In order to overcome problems of relative movement of microscope and tissue associated with temperature changes, we found it necessary to heat the Ludin chamber to 37°C for at least 1 h (preferably 2 h) prior to imaging and to allow the retinal preparation to acclimatize for at least 30 min before time-lapse imaging began. Weighting the Millipore filter down with a ring of platinum wire prevented drift of the specimen over time (Fig. 2). Applying high-grade paraffin oil to the surface of the culture medium before imaging began prevented evaporation in the heated environment over long periods. This not only kept the objective immersed in the medium but also prevented changes in the concentration of its constituents, which could have resulted in toxic conditions.
Absorption of light energy by dark objects or pigmented tissues was found to be a serious problem when using multiphoton rather than conventional laser excitation, and could result in extensive tissue destruction. It was overcome by minimizing the laser power, by using white rather than black Millipore filters to mount the retinae, and by making sure that all of the pigmented retinal epithelium was removed in the dissection. The presence of Phenol Red in the culture medium was also found to have a negative effect on the detection of the emitted fluorescence signal. Imaging of cells, with and without Phenol Red in the medium, showed that in its presence approximately 33% of the light was lost, resulting in a dimmed image with significantly reduced sensitivity (Fig. 3
).
View larger version (34K):
[in this window]
[in a new window]
|
Figure 3. Images of live chick retinal ganglion cells, ballistically labelled with the carbocyanine dye, DiI
The parameters for image acquisition were kept constant. Image B was acquired with Phenol Red in the medium, whereas image A was acquired in a medium without Phenol Red. The typical decrease in signal in the presence of Phenol Red is about 33%. Scale bar, 50 µm.
|
|
An additional problem associated with the use of multiphoton excitation was a tendency for tungsten particles, when used for ballistic labelling, to cause severe flare in the images (Fig. 4
). This appeared to be the result of saturation of the PMT. Flare was never seen when imaging with conventional laser excitation; nor was it seen when gold particles were used, so gold was preferred for ballistic labelling in this context, as well as for transfection.
View larger version (28K):
[in this window]
[in a new window]
|
Figure 4. Tungsten particles are unsuited to ballistic labelling of cells intended for multiphoton microscopy
A, multiphoton imaging of cells ballistically labelled with DiI-coated tungsten particles generated a flaring artifact in the images which appeared to result from saturation of the PMT. This may relate to absorption of energy by the dark tungsten particles. It was not seen when gold particles were used. B, intensity profile plots along the flare lines show saturation followed by gradual recovery of the PMT. Scale bar, 50 µm.
|
|
Effects of manipulating gap junctional communication on retinal development
In control retinae, labelled neuroprogenitor cells were seen moving towards the vitread surface during G1 phase, then towards the ventricular surface in G2 phase, where they underwent division (Fig. 5
). This process typically took about 8–9 h. However, when gap junctional communication was perturbed with pharmacological agents such as FFA and CBX, the rate of movement of all cells was severely reduced. Similar effects were seen when communication was reduced in cells transfected with dnCx43 or when Cx43 protein was reduced by application of asODNs.
View larger version (57K):
[in this window]
[in a new window]
|
Figure 5. Projected, rendered and time-lapse images of living progenitor cells
A, a projection of a Z series taken through the entire thickness of the retina, showing cells ballistically labelled with DiI. B, an image from one time point in a 4D data set which has been surface rendered and slightly rotated to show the profiles of labelled cells. C, a series of images of a single labelled cell over 390 min. The ventricular surface is uppermost and the nucleus of the cell can be seen to be migrating towards the vitread surface during G1 before returning towards the ventricular surface during G2. Scale bars: A and B, 50 µm, C, 25 µm.
|
|
Visualizing the migration of DiI-labelled cells over 16–18 h, in retinae in which communication had been blocked with FFA or CBX, revealed significant changes in migratory behaviour. At the start of imaging, cells were seen to migrate in both the vitread and sclerad (ventricular) directions, just as in control retinae but slower. However, in the absence of gap junctional communication these cells never turned around and headed back in the opposite direction. Instead, having reached either the vitread or the ventricular surface of the retina, they remained there. Immunostaining with neuronal markers such as TUJ1 (which stains neuronal class III β-tubulin but not the β-tubulin of glial cells) or for MAP1b (microtubule associated protein 1b, which is expressed in neuronal cell bodies and actively growing neurites) showed that, in control retinae at these stages, neurons were only present in the developing retinal ganglion cell layer, at the vitread surface. However, in retinae where communication had been blocked for 16–18 h, significant amounts of neuronal staining could be found in cells at both the vitread and ventricular surfaces (Fig. 6
).
View larger version (76K):
[in this window]
[in a new window]
|
Figure 6. Effects of perturbing gap junctional communication on the laminar distribution of TUJ1-positive neurons
A–C, 90 deg rotations of confocal Z stacks taken through whole-mount retinae that have been stained for the neuronal marker TUJ1 (red) and for DNA in nuclei (blue) with the ventricular surface uppermost. In the control retina (A) TUJ1 positive cells are restricted to the ganglion cell layer at the vitread surface. However, in retinae where gap junctional communication has been perturbed with CBX (B) or Cx43-specific antisense (C), TUJ1 positive cells can be found at both the ventricular and the vitread surfaces, suggesting that when communication is perturbed, ectopic neuronal differentiation can occur. Scale bars, 50 µm.
|
|
Progenitor cells that had been transfected with eGFP, eGFP-wtCx43 or eGFP-dnCx43 16–18 h previously were also visualized, and monitoring their locations within the retina was straightforward as cell bodies could be easily seen to be in the VZ (with somata appearing to contact the ventricular surface), in the GCL (with somata lined up close to the vitread surface and appearing to contact the emerging optic axon layer) or in the intermediate zone (with bipolar processes traversing the retina; see Fig. 7B
). We found eGFP and eGFP-wtCx43 to be distributed rather evenly throughout the retina: in cells at the vitread surface, traversing the retina or at the ventricular surface. However, of the cells expressing eGFP-dnCx43, about 65% were located at the ventricular surface of the retina, where mitosis normally takes place (Fig. 7). Non-parametric statistical analysis of the data was performed, comparing all groups of data (Kruskal–Wallis ANOVA by ranks) and pairs of groups (post hoc Mann–Whitney U tests). The fraction of cells expressing eGFP-dnCx43 that was found to be in the VZ was significantly greater than that for either eGFP or eGFP-wtCx43 (P < 0.0001) but there was no significant difference between the locations of cells expressing eGFP and eGFP-wtCx43 (P < 0.13). In addition, there were significantly fewer eGFP-dnCx43-expressing cells in the GCL compared to those expressing eGFP (P < 0.0002) or eGFP-wtCx43 (P < 0.0008). Closer examination of the eGFP-dnCx43-expressing cells at the ventricular surface revealed that they had begun to grow neurites (Fig. 7C), implying that they had differentiated ectopically, having failed to migrate away from the ventricular surface as differentiating neurons should.
View larger version (31K):
[in this window]
[in a new window]
|
Figure 7. Effects of perturbing gap junctional communication on the laminar distribution of eGFP-expressing neurons and progenitors
A, a projection of a Z stack of images through the entire thickness of a retina that has been ballistically transfected with eGFP. B, rotation of the data set in A by 90 deg to show the laminar locations of eGFP-positive cells throughout the retina, including compact cell bodies at both the vitread and ventricular surfaces as well as bipolar cells that traverse the entire thickness of the retina. C, a high power, rendered image of a single, live eGFP-dnCx43-expressing cell at the ventricular surface. The cell is clearly starting to grow neurites at this ectopic position within the retina. D, percentage counts of eGFP-positive cells at three distinct retinal laminar locations: VZ (ventricular zone), INT (intermediate zone) and GCL (ganglion cell layer), as defined in the text. Perturbation of gap junctional communication, whether by transfection with dnCx43 or by the addition of CBX to the culture medium, greatly increased the proportion of eGFP-expressing cells at the ventricular surface. The effects of the two treatments also appeared to be additive, one possible reason being that CBX takes effect sooner than transfection, another that it also affects communication among the neighbours of the transfected cells. Scale bars: A and B, 50 µm, C, 10 µm.
|
|
Overnight incubation of retinae, transfected with either eGFP or eGFP-wtCx43, in media containing the gap junction blocker CBX had an effect similar to that seen in the non-communicating eGFP-dnCx43 cells. The majority of eGFP- and eGFP-wtCx43-expressing cells were found in the VZ, being present there in significantly (P < 0.0001) greater numbers than when these transfects were cultured in normal media (Fig. 7, Table 1
). Treament with CBX caused the percentage of eGFP- and eGFP-wtCx43-expressing cells in the VZ to be greater (about 80%) than when cells expressed eGFPdnCx43 but were not treated with CBX (about 65%).
|
Discussion
|
|---|
We have shown that it is possible to image migrating neural progenitor cells in whole-mount embryonic chick retinae over extended periods that include complete cell cycles by making appropriate modifications to existing techniques. When gap junctional communication is perturbed in the short term, we have previously found that progenitor migration continues, although at a significantly reduced rate (Pearson et al. 2005). However, the extended imaging techniques that we describe here have revealed that impaired gap junctional communication results not only in retarded saltatory movement (INM) of progenitor cell nuclei as they traverse the ventricular zone but also in the failure of both progenitor cell nuclei and young postmitotic neurons to change direction at the limits of their travel, and thus their failure to leave either the vitread or ventricular side of the developing retina. Regardless of whether communication was blocked throughout the progenitor population (with drugs or Cx43-specific asODNs), or in a subset of individual cells (by transfection with dnCx43), labelled cells gradually accumulated in these two layers, and in many cases developed the morphological and immunocytochemical characteristics of immature ectopic neurons. The effects of gap junction blockers, such as CBX, were more pronounced than those of eGFP-dnCx43, producing a higher percentage of eGFP cells in the VZ (80% versus 65%). This is likely to reflect the fact that the gap junction blockers start to act immediately but the transfected eGFP-dnCx43 construct takes at least 6 h to become expressed and affect communication.
A similar case of ectopic neuronal differentiation at the ventricular surface of the retina was reported by Brittis & Silver (1994) after they cultured developing rat eyes with chondroitin sulphate proteoglycan applied to the ventricular surface. Brittis & Silver suggested that the abnormal presence of this glycosaminoglycan at the ventricular surface might lead to ectopic binding of a growth factor that promotes differentiation, but they also noted that it might act by elevating connexin expression, as it does in cultured hepatocytes (Spray et al. 1987). Similarly, the present observations could be interpreted as implying that impaired gap junctional communication in young neurons causes the precocious elaboration of neurites, which then secondarily restrict the subsequent migration of the cell body. However, the parallel effects of impaired communication on INM in cycling progenitor cells cannot be explained by precocious differentiation, so a primary effect of impaired communication on migration seems more likely, with ectopic differentiation as a secondary consequence. We are continuing to investigate this issue.
Clearly, appropriate levels of gap junctional communication between retinal neural progenitor cells are essential for their normal cell cycle progression, differentiation and migration to appropriate terminal sites. Other parts of the CNS, such as the spinal cord and cerebral cortex, undergo similar processes during development (reviewed by Hollyday, 2001; Marín & Rubenstein, 2003), so it is likely that gap junctional communication has a fundamental role in the proliferation, differentiation and migration of neurons throughout the CNS, co-ordinating the formation of its normal architecture.
|
References
|
|---|
Bannerman P, Nichols W, Puhalla S, Oliver T, Berman M & Pleasure D (2000). Early migratory rat neural crest cells express functional gap junctions: evidence that neural crest cell survival requires gap junction function. J Neurosci Res 61, 605–615.[CrossRef][Medline]Becker DL, Bonness V, Catsicas M & Mobbs P (2002). Changing patterns of ganglion cell coupling during chick retinal development. J Neurobiol 52, 280–293.[CrossRef][Medline]
Becker DL, Ciantar D, Catsicas M, Pearson R & Mobbs P (2001). Use of pIRES vectors to express EGFP and connexin constructs in studies of the role of gap junctional communication in the early development of the chick retina and brain. Cell Commun Adhes 8, 355–359.[Medline]
Becker DL, Evans WH, Green CR & Warner A (1995). Functional analysis of amino acid sequences in connexin43 involved in intercellular communication through gap junctions. J Cell Sci 108, 1455–1467.[Abstract]
Becker DL, McGonnell I, Makarenkova H, Patel K, Tickle C, Lorimer J & Green CR (1999). Roles for 1 connexin in morphogenesis of chick embryos using a novel antisense approach. Dev Genet 24, 33–42.[CrossRef][Medline]
Becker DL & Mobbs P (1999). Connexin 1 and cell proliferation in the developing chick retina. Expl Neurol 156, 326–332.[CrossRef]
Brittis PA & Silver J (1994). Exogenous glycosaminoglycans induce complete inversion of retinal ganglion cell bodies and their axons within the retinal neuroepithelium. Proc Natl Acad Sci U S A 91, 7539–7542.[Abstract/Free Full Text]
Cadetti L & Thoreson WB (2006). Feedback effects of horizontal cell membrane potential on cone calcium currents studied with simultaneous recordings. J Neurophysiol 95, 1992–1995.[Abstract/Free Full Text]
Gan WB, Grutzendler J, Wong WT, Wong RO & Lichtman JW (2000). Multicolor ‘DiOlistic’ labeling of the nervous system using lipophilic dye combinations. Neuron 27, 219–225.[CrossRef][Medline]
Hollyday M (2001). Neurogenesis in the vertebrate neural tube. Int J Dev Neurosci 19, 161–173.[CrossRef][Medline]
Huang GY, Cooper ES, Waldo K, Kirby ML, Gilula NB & Lo CW (1998). Gap junction-mediated cell-cell communication modulates mouse neural crest migration. J Cell Biol 143, 1725–1734.[Abstract/Free Full Text]
Komuro H & Rakic P (1996). Intracellular Ca2+ fluctuations modulate the rate of neuronal migration. Neuron 17, 275–285.[CrossRef][Medline]
Komuro H & Rakic P (1998). Orchestration of neuronal migration by activity of ion channels, neurotransmitter receptors, and intracellular Ca2+ fluctuations. J Neurobiol 37, 110–130.[CrossRef][Medline]
Komuro H, Yacubova E, Yacubova E & Rakic P (2001). Mode and tempo of tangential cell migration in the cerebellar external granule layer. J Neurosci 21, 527–540.[Abstract/Free Full Text]
Marín O & Rubenstein JLR (2003). Cell migration in the forebrain. Annu Rev Neurosci 26, 441–483.[CrossRef][Medline]
Pearson RA, Catsicas M, Becker DL, Bayley P, Luneborg NL & Mobbs P (2004). Ca2+ signaling and gap junction coupling within and between pigment epithelium and neural retina in the developing chick. Eur J Neurosci 19, 2435–2445.[CrossRef][Medline]
Pearson R, Catsicas M, Becker DL & Mobbs P (2002). Purinergic and muscarinic modulation of the cell cycle and Ca2+ signalling in the chick retinal ventricular zone. J Neurosci 22, 7569–7579.[Abstract/Free Full Text]
Pearson RA, Luneborg NL, Becker DL & Mobbs P (2005). Gap junctions modulate interkinetic nuclear movement in retinal progenitor cells. J Neurosci 25, 10803–10814.[Abstract/Free Full Text]
Spray DC, Fujita M, Sáez JC, Choi H, Watanabe T, Hertzberg EL, Rosenberg LC & Reid LM (1987). Proteoglycans and glycosaminoglycans induce gap junction synthesis and function in primary liver cultures. J Cell Biol 105, 541–551.[Abstract/Free Full Text]
Vessey JP, Lalonde MR, Mizan HA, Welch NC, Kelly ME & Barnes S (2004). Carbenoxolone inhibition of voltage-gated Ca channels and synaptic transmission in the retina. J Neurophysiol 92, 1252–1256.[Abstract/Free Full Text]
Weissman TA, Riquelme PA, Ivic L, Flint AC & Kriegstein AR (2004). Calcium waves propagate through radial glial cells and modulate proliferation in the developing neocortex. Neuron 43, 647–661.[CrossRef][Medline]
|
Acknowledgements
|
|---|
We would like to thank BSc and MSc project students who have contributed to pilot projects on the retina over the years. This work is funded by the BBSRC.
Top
Abstract
Introduction
