PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of cercorLink to Publisher's site
 
Cereb Cortex. 2010 June; 20(6): 1497–1505.
Published online 2009 October 27. doi:  10.1093/cercor/bhp211
PMCID: PMC2871376

Leading Process Branch Instability in Lis1+/− Nonradially Migrating Interneurons

Abstract

Mammalian forebrain development requires extensive migration, yet the mechanisms through which migrating neurons sense and respond to guidance cues are not well understood. Similar to the axon growth cone, the leading process and branches of neurons may guide migration, but the cytoskeletal events that regulate branching are unknown. We have previously shown that loss of microtubule-associated protein Lis1 reduces branching during migration compared with wild-type neurons. Using time-lapse imaging of Lis1+/− and Lis1+/+ cells migrating from medial ganglionic eminence explant cultures, we show that the branching defect is not due to a failure to initiate branches but a defect in the stabilization of new branches. The leading processes of Lis1+/− neurons have reduced expression of stabilized, acetylated microtubules compared with Lis1+/+ neurons. To determine whether Lis1 modulates branch stability through its role as the noncatalytic β regulatory subunit of platelet-activating factor (PAF) acetylhydrolase 1b, exogenous PAF was applied to wild-type cells. Excess PAF added to wild-type neurons phenocopies the branch instability observed in Lis1+/− neurons, and a PAF antagonist rescues leading process branching in Lis1+/− neurons. These data highlight a role for Lis1, acting through the PAF pathway, in leading process branching and microtubule stabilization.

Keywords: branching, leading process, microtubules, migration, PAFAH1b

Introduction

Cell migration is an essential component of neural development, yet its biologic basis is incompletely understood. Neuronal migration has been divided into 2 phases: leading process extension and nucleokinesis. Nucleokinesis requires cytoplasmic dynein, microtubules, microtubule-associated proteins, and cell polarity genes (reviewed in Tsai and Gleeson 2005). The role of several genes, including the human lissencephaly gene LIS1, in coupling the nucleus and centrosome is well established and is mediated through its interaction with cytoplasmic dynein and Ndel1 (Feng et al. 2000; Shu et al. 2004; Tanaka et al. 2004; Tsai et al. 2007).

In contrast, relatively little is known about the cytoskeletal events necessary for leading process navigation and branching. We have hypothesized that the leading process of migrating neurons has analogous functions to an axon growth cone, sensing cues and guiding migrating neurons (Nasrallah et al. 2006). Axon guidance is dependent on the actin cytoskeleton for protrusion as well as microtubules for stabilization of branches (Luo 2002; Dent and Gertler 2003). Migrating neurons may use common mechanisms. Expression of an enhanced green flourescent protein (eGFP)–actin construct in migrating cortical neurons reveals dynamic formation and retraction of microspikes and lamellapodia-like structures at the leading process (Schmid et al. 2004). Moreover, the microtubule-associated proteins Lis1 and doublecortin (Dcx) localize to the leading edge of migrating fibroblasts and neurons (Dujardin et al. 2003; Schaar 2004; Kholmanskikh et al. 2006). These data suggest that, in addition to a role in nucleokinesis, microtubule stabilization potentially plays a role in leading process dynamics. Support for this comes from directly observing migrating interneurons that show complex and dynamic leading process morphologies (Bellion et al. 2005; Nasrallah et al. 2006). In contrast, Lis1+/ interneurons have fewer branches during migration (Nasrallah et al. 2006), whereas Dcx-mutant cells reveal a more highly branched leading process and reduced stability of branches (Kappeler et al. 2006). Whether these effects on branching are mediated through the stabilization of microtubule is not known. An alternative possibility is that Lis1, functioning through its role as the noncatalytic regulatory subunit of the platelet-activating factor acetylhydrolase 1b (PAFAHIb), the enzyme that degrades platelet-activating factor (PAF) (Hattori et al. 1994), modulates leading process growth and branching.

To determine the cellular pathway(s) in which Lis1 functions to affect leading process branching, we performed a series of experiments using the Lis1+/ mouse and pharmacologic modulators of the PAF pathway. Time-lapse imaging of Lis1+/ cells migrating from medial ganglionic eminence (MGE) explant cultures indicated that the branching defect was not due to a failure to initiate branches but a defect in the stabilization of new branches. The addition of exogenous PAF to wild-type cells phenocopies the branch instability observed in Lis1+/ cells, and a PAF inhibitor rescues branching by leading process splitting in Lis1+/ cells. Finally, we find fewer stabilized (acetylated) microtubules in new leading processes in the absence of Lis1 signaling, whereas the addition of a PAF inhibitor restored the percentage of acetylated microtubules in Lis1+/ leading processes to wild-type levels. These data indicate that the leading process branching defect, but not nucleokinesis, results from a disruption of PAF signaling.

Materials and Methods

Mouse Strains and Genotyping

Lis1 mutant mice were obtained from Dr A. Wynshaw-Boris (University of California San Diego, La Jolla, CA) and maintained on a Black Swiss background, as previously described (Hirotsune et al. 1998; Nasrallah et al. 2006). Timed pregnant mice were considered embryonic day 0.5 on the morning a vaginal plug was identified. All embryos were morphologically staged (Theiler 1989) and genotyped by polymerase chain reaction as previously described (Hirotsune et al. 1998). Embryos for explant cultures (see below) were obtained from Dlx5/6 enhancer cre IRES GFP (abbreviated Dlx5/6 cre IRES GFP) transgenic mice (kindly provided by Dr K. Campbell, Children's Hospital Medical Center, Cincinnati, OH) maintained on a CD1 background. In addition, Lis1+/ mice were crossed with Dlx5/6 cre IRES GFP transgenic mice to generate a line of Dlx5/6 cre IRES GFP; Lis1+/ mice. The institution's animal care and use committee approved all animal breeding, handling, and experimental procedures.

Explant Culture

Explants

Embryonic day 14.5 (E14.5) brains from Dlx5/6 cre IRES GFP; CD1, Dlx5/6 cre IRES GFP; Lis1+/, or Dlx5/6 cre IRES GFP; Lis1+/+ animals were dissected in Hanks saline. The brains were embedded in 4% low–melting point agarose for coronal sectioning on a vibrating microtome (250 μm; LeicaVT1000S vibrotome; Leica Microsystems, Nussloch, Germany). Using an ocular micrometer, MGE and neocortical explants (200 × 200 μm) were dissected from forebrain slices, transferred onto cortical substrate cells (see below), and surrounded with a 1:1 mixture of growth factor reduced matrigel (Collaborative, Bedford, MA) and acidic collagen. MGE and cortical explants were placed within 200–300 μm of each other. Explants were incubated in 1:1 Dulbecco's modified Eagle's medium (DMEM):F12 (Invitrogen Corporation, Carlsbad, CA) with 10% fetal bovine serum (FBS, Invitrogen Corporation), 1 mM penicillin/streptomycin (Invitrogen Corporation), and 6.5 mg/mL glucose for 2–3 h and then transferred to differentiation media, containing 1:1 DMEM:F12 plus N2 supplement (1:100, Invitrogen Corporation), B27 (1:50, Invitrogen Corporation), 1 mM penicillin/streptomycin (Invitrogen Corporation), and 6.5 mg/mL glucose.

Cortical Substrate

Dissociated cortical substrate cells were prepared as previously described (Bellion et al. 2005), with the following modifications. Embryonic brains (E14.5) from CD1 mice were dissected, and meninges were removed in HANKS’ saline. The cortices were mechanically triturated by pipeting in 1:1 DMEM:F12 with 10% FBS, 1 mM penicillin/streptomycin, and 6.5 mg/mL glucose. Dissociated cells were plated at 2000 cells/mm2 on poly-D-lysine (10 μg/ml; Sigma, St Louis, MO), laminin-coated (10 μg/ml; BD Biosciences, San Jose, CA)-coated Delta T4 dishes (Bioptechs, Butler, PA) and cultured in differentiation media. After 2 days in culture, the cortical cultures were washed once with sterile phosphate-buffered saline (PBS) and frozen at −80 °C.

Immunostaining Explant Cultures

Cultures were fixed for 20 min in methanol (−20 °C) followed by a 15-min digestion in collagenase (1:100, Sigma). For double immunostaining of tyrosinated and acetylated tubulins, the explant cultures were blocked in 10% normal goat serum, 0.4% Triton X, and 0.05% NaN3 for 1 h at room temperature. The explants were then incubated in rat α–tyrosinated tubulin (1:750, overnight at 4 °C, Sigma) and reacted with either goat α–rat IgG Alexa633 or goat α−rat IgG Alexa488 (1:200, Molecular Probes/Invitrogen Corporation, Carlsbad, CA) for 1 h at room temperature. After washing in PBS, the explants were incubated in mouse α-acetylated tubulin (1:500, Sigma) and incubated with biotinylated goat α–mouse IgG2b (1:500; 1 h at room temperature; Jackson Immunoresearch, West Grove, PA). Finally, the explants were incubated in fluorescein isothiocyanate or Cy3 conjugated to streptavidin (1:500, Jackson Immunoresearch). For double immunostaining of neuronal class III β tubulin (TUJ1) and acetylated tubulin, the explant cultures were blocked in 10% normal goat serum for 1 h at room temperature and incubated in 0.4% Triton X (15 min at room temperature). Primary antibodies rabbit α-TUJ1 (1:400, Covance, Princeton, NJ) and rat α-acetylated tubulin (1:500, Sigma) were incubated simultaneously with explant cultures overnight at 4 °C. The explants were incubated with both goat α-rabbit IgG Alexa488 (1:200, Molecular Probes/Invitrogen Corporation) and biotinylated goat α-mouse IgG2b (1:500, Jackson Immunoresearch) for 1 h at room temperature. The biotinylated secondary antibody was reacted with Streptavidin-Cy3 (1:500, Jackson Immunoresearch) for 15 min at room temperature. Nuclei were counterstained with 4¢,6-diamidino-2-phenylindole (1:1000, Molecular Probes/Invitrogen Corporation). Images were obtained using a Leica DMR microscope equipped with epifluorescence and a Hamamatsu Orca digital camera.

Time-Lapse Video Microscopy

Twelve to sixteen hours after explanting MGE and neocortical explants onto substrate cells, Dlx5/6 cre IRES GFP–positive MGE cells migrating toward cortical explants were analyzed by time-lapse video microscopy. Culture dishes were placed in a heated chamber (Okolab; Warner Instruments, Hamden, CT) with a circulating mixture of CO2 (0.04 nL/min) and air (0.80 nL/min) on a Nikon TE-300 microscope equipped with a motorized stage (Prior, Cambridge, UK). Using ImagePro software (Media Cybernetics, Media, PA), fields of cells were live imaged over 2.5 h using a 20× objective; a phase image was taken every 5 min, whereas a fluorescent image was taken at the first and sixteenth time point for each field.

Pharmacology

Methyl carbamyl platelet–activating factor (mc-PAF; Cayman Chemicals, Ann Arbor, MI), a nonhydrolyzable PAF analog, was added to Dlx5/6 cre IRES GFP; CD1 and Dlx5/6 cre IRES GFP; Lis1 explant cultures at a final concentration of 1 μM, 30 min prior to imaging. PAF antagonist, BN52021 (Biomol, Plymouth Meeting, PA) was added to Dlx5/6 cre IRES GFP; Lis1 explant cultures at a final concentration of 50 μM, 30 min prior to imaging.

Data Analysis

All image analyses were performed manually by using ImagePro software. Dlx5/6 cre IRES GFP–positive MGE cells migrating toward cortical explants that also had leading process and/or branches in focus were analyzed. Analyses were performed blinded to genotype and treatment condition. Statistical differences between groups were calculated using Student's t-test where groups were defined as the mean from each slice analyzed.

Migration speed of the nucleus was calculated from the sequential displacement of the front edge of the nucleus. We adjusted for drift of the field by selecting a stationary point as reference. A cell was defined as being stopped if its incremental speed was less than 0.2 μm/min; this number was empirically determined to be the speed at which we could no longer distinguish directed movement versus random motion in the culture and/or the resolution of identifying the same point in a cell in consecutive frames. Length of the leading process and the number of branches were directly measured at each time point (5 min apart). For branch analysis, the formation and lifetime of each new leading process was plotted over time. The method of branch formation was classified as interstitial or by leading process splitting (Fig. 1A,B), as previously described for axon growth cones (Luo 2002). New branches present for ≥15 min were defined as stable. An existing multibranched leading process was distinguished from a newly formed leading process by comparing successive frames in each movie; a leading process split that persisted for ≥15 min was then defined as a new branch, generating a cell with a multibranched leading process. See Figure 1 for further description of interstitial branching (Fig. 1A) and leading process splitting (Fig. 1B).

Figure 1.
Diagram representing interstitial branching (A) and leading process splitting (B) drawn from cells observed in this study (A, Fig. 3; B, Supplementary Material WT.mov). (A) An interstitial branching is characterized by a leading process that initially ...

Results

We previously showed that Lis1+/− interneurons migrating in embryonic slice culture and in the brain have a reduced speed of migration, an elongated leading process, and reduced number of branches compared with wild-type cells (Nasrallah et al. 2006). To further characterize the branching defect in Lis1+/− interneurons, we used an explant coculture system, modified from Bellion et al. (2005). Cortex and MGE explants from Dlx5/6 cre IRES GFP; Lis1+/− and Dlx5/6 cre IRES GFP; Lis1+/+ E14.5 forebrain were dissected and arranged 200–400 μm apart from each other on a coverslip coated with the surface proteins from cerebral cortical cells. Because the Dlx5/6 enhancer drives green flourescent protein (GFP) expression in interneurons (Stuhmer et al. 2002), we limited our analysis specifically to GFP+ interneurons migrating from the MGE. Time-lapse video microscopy was used to analyze the migration of the MGE-derived interneurons toward the cortical explants.

Similar to our data from slice culture experiments, Dlx5/6 cre IRES GFP; Lis1+/− cells migrated from explant cultures with a reduced speed of nuclear migration (0.73 ± 0.23 μm/min; n = 27 cells derived from 6 cortical slices from 3 animals and from 3 litters) when compared with wild-type littermate cells (0.85 ± 0.26 μm/min; n = 68 cells derived from 16 cortical slices from 8 animals from 3 litters; P < 0.05; Fig. 2A). Slowed nuclear migration results in elongation of cell morphology and a longer leading process in Dlx5/6 cre IRES GFP; Lis1+/− cells (Fig. 2B). The average length of the leading process was 40.4 ± 4.6 μm in Dlx5/6 cre IRES GFP; Lis1+/+ cells compared with 59.5 ± 7.5 μm in Dlx5/6 cre IRES GFP; Lis1+/− cells (P < 0.001). In contrast to our previous studies, however, we observed a similar average number of leading processes per cell in wild-type (2.44 ± 0.73 leading processes) and mutant cells (2.19 ± 0.66 leading processes; P > 0.05) migrating from explant cultures (Fig. 2C; also see Supplementary Materials for movies of Dlx5/6 cre IRES GFP; Lis1+/+ and Dlx5/6 cre IRES GFP; Lis1+/− cells.

Figure 2.
Comparison of nuclear migration and cell morphology in Dlx5/6 cre IRES GFP; Lis1+/− migrating interneurons and mc-PAF–treated(1 μM) Dlx5/6 cre IRES GFP wild-type cells. Speed of nuclear migration is reduced in Dlx5/6 cre IRES GFP ...

Quantification of the number and lifetime of new branches per cell per 100 min shows that Dlx5/6 cre IRES GFP; Lis1+/− cells form a similar number of new branches as Dlx5/6 cre IRES GFP; Lis1+/+ cells (Fig. 2A). Though all wild-type and mutant cells analyzed formed new branches (Fig. 3A, white bars), there is a significant reduction in the formation of new stable branches in mutant cells (Fig. 3A, black bars). Whereas 64.6 ± 10% (n = 24 cells) of new branches are stable (lifetime ≥15 min) in Dlx5/6 cre IRES GFP; Lis1+/+ cells, only 36.8 ± 10% (n = 13 cells) of new branches in mutant cells last 15 min or more (P < 0.05; Fig. 3A).

Figure 3.
Reduced branch stability and leading process splitting in Dlx5/6 cre IRES GFP; Lis1+/− and mc-PAF–treated cells. Both Dlx5/6 cre IRES GFP; Lis1+/− and mc-PAF–treated cells make similar numbers of new branches compared with ...

We observed 2 types of branching in migrating interneurons (Fig. 4A–D; see also Fig. 1A,B and movies provided in the Supplementary Material). Interneurons can form interstitial branches, by extending a new branch from the main trunk of the leading process (Fig. 4A–D, arrows), analogous to interstitial branching during axon guidance (Luo 2002). Alternatively, branches can form by splitting of the leading process to form 2 sister branches, as observed by others (Fig. 4D arrowhead; Kappeler et al. 2006). Although Dlx5/6 cre IRES GFP; Lis1+/− cells could form branches by either method (Fig. 3B), they have a reduced rate of branching by leading process splitting (3.6 ± 1.5 splits/100 min; n = 27 cells) compared with Dlx5/6 cre IRES GFP; Lis1+/+ cells (5.5 ± 1.9 splits/100 min; n = 68; P < 0.001; Fig. 3C).

Figure 4.
Time-lapse sequence of leading process splitting and interstitial branch formation in migrating interneurons. Branch formation was observed at 5-min intervals (panels A–D). Extension of a new branch from the main trunk of the leading process or ...

Although the Lis1+/− nuclear migration defect has been linked to interactions of Lis1with dynein and Ndel1 (Shu et al. 2004), how Lis1 functions at the leading process is poorly understood. In addition to its role in the Lis1/dynein/Ndel1 complex, Lis1 is also the β subunit of PAFAHIb (Hattori et al. 1994), and previous studies have shown PAF induces cytoskeletal changes and growth cone collapse in axons (Clark et al. 1995; McNeil et al. 1999). The reduction of Lis1 levels in mutant cells alters PAFAHIb degradation of PAF and therefore alters PAF signaling (Prescott et al. 2000).

We hypothesized that reduced branching at the tip of the leading process and branch instability in Lis1+/− interneurons are caused by increased PAF signaling. To test this hypothesis, we added either control media or media containing 1 μM mc-PAF, a nonhydrolyzable PAF analog (Bix and Clark 1998), to Dlx5/6 cre IRES GFP; CD1 or Dlx5/6 cre IRES GFP; Lis1+/+ cells 30 min prior to imaging. Acute treatment with mc-PAF caused a small but significant reduction (P < 0.05) in the speed of nuclear migration (0.93 ± 0.15 μm/min for control cells, n = 89 cells; 0.87 ± 0.19 μm/min for mc-PAF–treated cells, n = 94 cells; both control and mc-PAF–treated cells were derived from 16 cortical slices from 8 animals and from 3 litters; Fig. 2D). Despite the small nuclear migration defect, the average length of the leading process was not different between control (44.5 ± 5.3 μm) and mc-PAF–treated cells (45.7 ± 4.9 μm/min; Fig. 2E). Similar to Dlx5/6 cre IRES GFP; Lis1+/− cells migrating from explant cultures, control (2.22 ± 0.30 leading processes) and mc-PAF–treated cells (2.00 ± 0.38 leading processes; P = 0.096) have a similar average number of leading processes per cell (Fig. 2F). Although mc-PAF–treated cells form a similar number of new branches as control cells (Fig. 3A, compare gray bars of DlxGFP control and DlxGFP mc-PAF), there is a significant reduction in new stable branches (Fig. 3A, cross-hatched bars), similar to Dlx5/6 cre IRES GFP; Lis1+/− interneurons. In addition, mc-PAF–treated cells have a reduced frequency of branching by leading process splitting (4.1 ± 0.76 splits/100 min; n = 94 cells) compared with untreated cells (5.4 ± 0.91 splits/100 min; n = 89; P < 0.001; Fig. 3D).

If modulation of PAF signaling is the underlying defect resulting in the observed leading process defect in Lis1+/− cells, then blocking PAF signaling in Lis1+/− cells would be expected to rescue the branching phenotype. We added either dimethyl sulfoxide (DMSO) or a PAF antagonist, BN52021 (also called gingkolide B, 50 μM) in DMSO, to Dlx5/6 cre IRES GFP; Lis1+/+ or Dlx5/6 cre IRES GFP; Lis1+/− explant cultures 30 min prior to imaging. BN52021 is a noncompetitive PAF inhibitor shown to bind at cerebral cortical synaptosomal membranes and other intracellular binding sites (Marcheselli et al. 1990). BN52021 also antagonizes PAF by accelerating the kinetics of PAFAH1b activity (Bonin et al. 2004). Acute treatment with BN52021 rescued the nuclear migration defect in Dlx5/6 cre IRES GFP; Lis1+/− cells (Fig. 5A). Dlx5/6 cre IRES GFP; Lis1+/− cells treated with BN52021 migrated faster (0.91 ± 0.18 μm/min; n = 18 cells; control and BN52021-treated Lis1+/− cells were derived from 14 cortical slices from 7 animals from 3 litters) than Dlx5/6 cre IRES GFP; Lis1+/− cells that received DMSO (0.69 ± 0.11 μm/min; n = 8 cells; P < 0.01). Furthermore, the nuclear migration speed of BN52021-treated mutant cells was comparable to that of Dlx5/6 cre IRES GFP; Lis1+/+ cells treated with BN52021 (0.91 ± 0.14 μm/min; n = 18 cells; control and BN52021-treated Lis1+/+ cells were derived from 14 cortical slices from 7 animals from 3 litters; P = 0.97). Unexpectedly, however, the rescued speed of nuclear migration in Dlx5/6 cre IRES GFP; Lis1+/− BN52021-treated cells did not rescue the elongation of the leading process (BN52021-treated mutant cells 51.4 ± 12.1 μm; DMSO-treated mutant cells 57.9 ± 16.1 μm; Fig. 5B). BN52021 treatment also increased the average number of leading processes and frequency of leading process splitting in Dlx5/6 cre IRES GFP; Lis1+/− cells (Fig. 5C,D). The average number of leading processes increased from 1.34 ± 0.11 leading processes in DMSO-treated mutant cells to 2.19 ± 0.75 leading processes in BN52021-treated mutant cells (P < 0.01). The frequency of branching by leading process splitting in Dlx5/6 cre IRES GFP; Lis1+/− cells (5.4 ± 1.4 splits/100 min) was restored to wild-type levels with BN52021 (5.6 ± 1.1 splits/100 min; P = 0.67).

Figure 5.
A PAF antagonist, BN52021, rescues Dlx5/6 cre IRES GFP; Lis1+/− defects. (A) BN52021 rescues nuclear migration speed in Dlx5/6 cre IRES GFP; Lis1+/− cells. DMSO-treated Dlx5/6 cre IRES GFP; Lis1+/− cells have a significantly reduced ...

Our data suggests that Lis1, functioning through the PAF pathway, plays a role in leading process growth by stabilizing new branches. How new branches form and grow is not well studied in migrating neurons. One possible mechanism is through the stabilization of invading microtubules, as observed in axon guidance (reviewed in Luo 2002). Assuming that leading process branching in migrating neurons forms by a similar mechanism, we would predict fewer stabilized microtubules in the leading process of Lis1+/− cells. Microtubules are tyrosinated when they first polymerize but later become detyrosinated and may be acetylated (Bulinski and Gundersen 1991; Bulinski 2007). We examined the distribution of tyrosinated (tyr) and acetylated (acetyl) tubules at the leading edge of migrating interneurons from Lis1+/− and Lis1+/+ cells (Dlx5/6 cre IRES GFP mice were not used to permit double labeling with antibodies to tyrosinated tubules and acetylated tubules). The majority (69% for Lis1+/+ and 72% for Lis1+/−) of leading processes colabeled for both tyr tubulin and acetyl tubulin (Fig. 6A,B). In contrast, 24% fewer leading processes from Lis1+/− migrating neurons were labeled with the stabilized (acetyl) form of tubulin alone (Fig. 6C–E) when compared with Lis1+/+ migrating neurons (n = 47 cells for Lis1+/+ and 32 cells for Lis1+/−, data from 2 different explant experiments). We next asked whether the PAF inhibitor BN52021-mediated rescue of leading process branching in Lis1+/− cells (see Fig. 5C,D) correlates with a greater percentage of microtubule acetylation in leading process branches. Lis1+/+ and Lis1+/− MGE explant cultures, treated with DMSO or BN52021 (50 μM), were immunolabeled for neuronal class III β tubulin (TUJ1) and acetylated tubules. TUJ1 immunolabeling allowed visualization of all leading processes. We then calculated the percentage of acetyl tubule immunopositive (immuno+) leading processes [(number of acetyl tubule immuno+ leading processes)/(number of all TUJ1 immuno+ leading processes)]. DMSO- and BN52021-treated Lis1+/+ cells had a comparable percentage of acetylated leading processes (87 ± 1.0%, n = 24 cells from 2 separate experiments, and 89 ± 3.0%, n = 29 cells from 3 separate experiments, respectively), whereas DMSO-treated Lis1+/− cells had only 65 ± 2.0% acetylated leading processes (n = 25 cells from 2 separate experiments; Fig. 6F). The addition of BN52021 to Lis1+/− cells increased the percentage of acetylated leading processes to wild-type levels (82 ± 6.0%, n = 33 cells from 3 separate experiments; Fig. 6F). Together, these data indicate that the leading process of migrating Lis1+/− cells can move and branch normally, but new branches are unstable due to the failure to stabilize invading microtubules. Furthermore, inhibition of PAF signaling in Lis1+/− cells rescues branching at the leading process, perhaps by increasing microtubule acetylation in newly formed branches.

Figure 6.
Distribution of tyrosinated (tyr) and acetylated (acetyl) tubules at the leading edge of migrating interneurons (A,B). Representative wild-type interneurons migrating in explant culture with leading process branches coimmunolabeled for tyr- and acetyl- ...

Discussion

Non-radial cell migration (NRCM) of interneurons from the ganglionic eminence to the neocortex requires interaction of the migrating cells with guidance cues and permissive substrates. Like the sensory function of an axon growth cone, it has been hypothesized that the leading process of migrating cells also senses and responds to environmental cues and directs migration. However, relatively little is known about leading process dynamics and how branching is regulated. Previous studies from our laboratory have shown Lis1+/− cells have a nuclear migration defect as well as a reduced number of leading processes (Nasrallah et al. 2006). Although many studies indicate a role for Lis1 in nuclear positioning (Morris et al. 1998; Lee et al. 2003; Shu et al. 2004; Tanaka et al. 2004), the branching defect raised the possibility of another cellular role for Lis1. However, whether Lis1 functions at the leading edge through microtubules and dynein/Ndel1, the PAF pathway, or an unknown pathway is unclear.

Using an explant culture system, we have shown that Dlx5/6 cre IRES GFP; Lis1+/− interneurons have reduced branch stability. Although Lis1+/− cells make a normal number of new branches, fewer of these new branches are stable compared with wild-type cells. In addition, wild-type and mutant cells form branches either by interstitial branching or division of the leading process (Fig. 1). However, the frequency of branching by leading process splitting is reduced in Lis1+/− cells. The branch instability phenotype and reduced branching at the tip of the leading process were similar to previous reports of PAF-induced growth cone collapse and retraction in cultured-hippocampal neurons (Clark et al. 1995; McNeil et al. 1999). To determine whether Lis1’s role in the PAF pathway may mediate the branching defects seen in Lis1+/− cells, we added mc-PAF, a nonhydrolyzable PAF analog, to Dlx5/6 cre IRES GFP; CD1 or Dlx5/6 cre IRES GFP; Lis1+/+ cells. Increased PAF signaling phenocopies the reduced branch stability and frequency of leading process splitting in wild-type cells. Moreover, addition of BN52021, a PAF inhibitor, to Dlx5/6 cre IRES GFP; Lis1+/+ or Dlx5/6 cre IRES GFP; Lis1+/− cultures rescued the frequency of leading process splitting in Lis1+/− cells to wild-type levels. The average number of leading processes was also increased in Lis1+/− cells treated with BN52021. Furthermore, we show that Lis1+/− leading process branch instability may be the result of decreased acetylation of microtubules at the leading edge and that addition of BN52021 to Lis1+/− cells restores the percentage of acetylated leading processes to wild-type levels. Though further experiments are needed, these results suggest that Lis1 functions to mediate leading process branching in migrating interneurons through PAF signaling, specifically affecting microtubule acetylation and stabilization of the leading process.

Comparison of Lis1+/− Migration Defects in Slice Culture and Explant Culture

Quantification of NRCM dynamics in Lis1+/− and wild-type littermate slice cultures revealed reduced speed of nuclear migration, elongation of the leading process, and reduced number of branches in Lis1+/− cells (Nasrallah et al. 2006). In order to directly observe branching events, we used an explant culture system modified from Bellion et al. (2005) in which cells migrate on the surface proteins of cortical cells. Quantification of NRCM dynamics of Dlx5/6 cre IRES GFP; Lis1+/+ and Dlx5/6 cre IRES GFP; Lis1+/− cells showed both a reduced speed of nuclear migration and elongation of the leading process in Lis1+/− cells, similar to results in slice culture. However, we did not observe a significant reduction in the average number of leading processes in mutant cells. This discrepancy likely reflects the difficulty in observing branches in many focal planes present in slice culture (Nasrallah et al. 2006). In addition, many of the branches observed in Dlx5/6 cre IRES GFP; Lis1+/− cells migrating from explants were short in length and short-lived and were most likely not accounted for in the slice culture experiments. The inclusion of these unstable branches may explain why we found no difference in the average number of leading processes in wild-type and Dlx5/6 cre IRES GFP; Lis1+/− cells migrating from explant cultures.

Dlx5/6 cre IRES GFP; Lis1+/− Cells Have Reduced Branch Stability and Reduced Frequency of Branching by Leading Process Splitting

The leading process of migrating interneurons has cytoskeletal similarities with the extending axon growth cone (Schmid et al. 2004), raising the possibility that they have common regulatory pathways. Previously observed PAF-induced growth cone collapse and retraction in cultured neurons (Clark et al. 1995) are analogous to our data showing retraction of new leading process branches and a reduced frequency of leading process splitting in Dlx5/6 cre IRES GFP; Lis1+/− cells compared with wild-type cells.

Although Dlx5/6 cre IRES GFP; Lis1+/− and Dlx5/6 cre IRES GFP; Lis1+/+ cells make similar numbers of new branches, a significantly reduced percentage of new branches were stable in Dlx5/6 cre IRES GFP; Lis1+/− cells compared with wild-type littermate cells. These results indicate that Dlx5/6 cre IRES GFP; Lis1+/− cells can initiate new branches normally, but there is a defect in stabilizing or maintaining the branch. Studies in axons have shown that branching involves local destabilization of the actin and microtubule cytoskeletons, followed by extension of a filopodium, which is then stabilized by microtubule invasion (reviewed in Luo 2002). Lis1 has been shown to bind and stabilize microtubules (Sapir et al. 1997; Faulkner et al. 2000; Smith et al. 2000). Assuming that branches form by a similar mechanism in migrating neurons, microtubule instability in Lis1+/− cells could explain the instability of newly formed branches. Our data on the distribution of tyrosinated and acetylated microtubules at the leading edge of migrating interneurons from Lis1+/− and Lis1+/+ cells show that significantly fewer leading processes from Lis1+/− migrating neurons were labeled with the stabilized (acetyl-) form of microtubules alone. Together, these data suggest that Lis1 may be involved in the conversion of newly polymerized microtubules to stabilized microtubules.

PAF Analog Phenocopies Dlx5/6 cre IRES GFP; Lis1+/− Branching Defects in Wild-Type Cells and PAF Antagonist Rescues Branching at the Leading Process in Dlx5/6 cre IRES GFP; Lis1+/− Cells

Previous studies have demonstrated a role for PAF signaling in regulating neuronal migration but did not examine the effects of PAF on branching. Interestingly, both increased PAF signaling and loss of signaling through the mutation of the PAF receptor disrupt cerebellar granule cell migration (Bix and Clark 1998; Tokuoka et al. 2003). In addition, PAF receptor knockout animals (Pafr−/−) crossed to Lis1+/− animals show a further slowing of granule cell migration in vitro compared with Pafr+/−; Lis1+/− animals (Tokuoka et al. 2003). These studies suggest that an optimal level of PAF signaling is necessary for normal migration. Consistent with these data, we found application of mc-PAF to wild-type cells reduces the speed of nuclear migration, though this defect did not alter the length of the leading process. Similar to Lis1+/− cells, mc-PAF–treated cells have reduced branch stability and reduced branching by leading process splitting. Moreover, application of the PAF inhibitor BN52021 to Dlx5/6 cre IRES GFP; Lis1+/− cells rescues the frequency of branching to wild-type levels, increases the average number of leading processes, and restores the acetylation of microtubules in Lis1+/− leading processes to wild-type levels. Previous studies have shown that the effects of mc-PAF on granule cell migration can be reversed with simultaneous application of PAF antagonists, BN52021 or trans-2,5-bis-{3,4,5-trimethoxyphenyl}-1,3-dioxolane (Bix and Clark 1998). However, our results are the first example of rescue of Lis1+/− leading process defects with a PAF antagonist.

Alternate Cellular Pathways

Although our data suggest that Lis1 acts through the PAF pathway to regulate branch stability, we cannot rule out the possibility that other cellular processes also modulate the leading edge. Recent studies suggest the PAFAH1b α1 and α2 subunits interact with reelin signaling. Both α1 and α2 bind to the reelin receptor VLDLR (Zhang et al. 2007), and α2 binds Dab1 in a phosphorylation-independent manner (Zhang et al. 2009). In vivo data also support this interaction; whereas crossing Lis1+/− mice with reeler mice does not dramatically alter the severe hippocampal layering defect seen in reeler mice alone (Assadi et al. 2003), the addition of Pafah1b3 (encoding the α1 subunit) mutations in triple Lis1+/−; Reln−/−; Pafh1b3+/− mutants results in almost complete loss of hippocampal structures (Assadi et al. 2008). Furthermore, adding the PAF inhibitor BN52021 to PC12 cells shows that its proacetylhydrolase effect may be related to downregulation of α1 subunit expression, leading to a relative increase in α2 expression (Bonin et al. 2004). Therefore, although the rescue of leading process branching we observed with addition of BN52021 to Lis1+/ interneurons may reflect the ability of BN52021 to increase PAF hydrolysis, the relative increase in the ratio of α2/α1 subunit expression may also affect reelin or other signaling pathways. The N-terminal region of reelin regulates the complexity of cortical pyramidal neuron apical dendrites, lending further support that reelin may be important for other types of branching (Chameau et al. 2009).

Another possibility is that the Lis1/Ndel1/dynein complex may also regulate branching, and perhaps, increased PAF signaling shifts the equilibrium of Lis1 binding to PAFAH1b so that less Lis1 can participate in Ndel1/dynein complex. In this model, the movement of microtubules subunits loaded with dynein and Lis1 toward the plus end of formed microtubules and dissociation at the terminal end (Yamada et al. 2008) would be modulated by PAF signaling. Future studies will be necessary to test this hypothesis and determine whether titrating levels of Ndel1 or dynein can also affect branching.

Role for Lis1 at the Leading Process

Our data clearly indicate a role for Lis1 at the leading process of migrating neurons, in addition to its role in nucleokinesis. A function for Lis1 at the leading process was previously postulated based on its expression in the leading edge of neurons and migrating fibroblasts (Sasaki et al. 2000; Aumais et al. 2001; Dujardin et al. 2003). In addition, Lis1+/− cerebellar neurons have reduced f-actin at the leading process and fewer filopodia (Kholmanskikh et al. 2003) and mislocalization of IQGAP1 at the leading edge (Kholmanskikh et al. 2006), suggesting that Lis1 plays a role in actin regulation. Furthermore, in vivo evidence in Drosophila shows that Lis1 functions in neural processes; mutant DLis1 mushroom bodies have proliferation defects and reduced dendritic branching. A similar phenotype was observed in Dhc64C (cytoplasmic dynein heavy chain) mutants, suggesting that at least some of roles of Lis1 in neural processes are mediated through its microtubule- and dynein-associated functions (Liu et al. 2000). Our data suggest that Lis1 and PAF signaling interact to regulate branching in the leading process of migrating neurons, and we speculate that Lis1 functions in the conversion of newly polymerized microtubules to a stabilized form of microtubules.

Supplementary Material

Supplementary material can be found at: http://www.cercor.oxfordjournals.org/

Funding

National Institutes of Health (NS45034 and HD026979).

Supplementary Material

[Supplementary Data]

Acknowledgments

We would like to thank members of the Golden laboratory for their discussion and support as well as critical reading of this manuscript.

Conflict of Interest: None declared.

References

  • Assadi AH, Zhang G, Beffert U, McNeil RS, Renfro AL, Niu S, Quattrocchi CC, Antalffy BA, Sheldon M, Armstrong DD, et al. Interaction of reelin signaling and Lis1 in brain development. Nat Genet. 2003;35:270–276. [PubMed]
  • Assadi AH, Zhang G, McNeil R, Clark GD, D'Arcangelo G. Pafah1b2 mutations suppress the development of hydocephalus in compound Pafah1b1; Reln and Pafah1b1; Dab1 mutant mice. Neurosci Lett. 2008;439:100–105. [PMC free article] [PubMed]
  • Aumais JP, Tunstead JR, McNeil RS, Schaar BT, McConnell SK, Lin SH, Clark GD, Yu-Lee LY. NudC associates with Lis1 and the dynein motor at the leading pole of neurons. J Neurosci. 2001;21:RC187. [PubMed]
  • Bellion A, Baudoin JP, Alvarez C, Bornens M, Metin C. Nucleokinesis in tangentially migrating neurons comprises two alternating phases: forward migration of the Golgi/centrosome associated with centrosome splitting and myosin contraction at the rear. J Neurosci. 2005;25:5691–5699. [PubMed]
  • Bix GJ, Clark GD. Platelet-activating factor receptor stimulation disrupts neuronal migration in vitro. J Neurosci. 1998;18:307–318. [PubMed]
  • Bonin F, Ryan SD, Migahed L, Mo F, Lallier J, Franks DJ, Arai H, Bennett SA. Anti-apoptotic actions of the platelet-activating factor acetylhydrolase I alpha2 catalytic subunit. J Biol Chem. 2004;279:52425–52436. [PubMed]
  • Bulinski JC. Microtubule modification: acetylation speeds anterograde traffic flow. Curr Biol. 2007;17:R18–R20. [PubMed]
  • Bulinski JC, Gundersen GG. Stabilization of post-translational modification of microtubules during cellular morphogenesis. Bioessays. 1991;13:285–293. [PubMed]
  • Chameau P, Inta D, Vitalis T, Monyer H, Wadman WJ, van Hooft JA. The N-terminal region of reelin regulates postnatal dendritic maturation of cortical pyramidal neurons. Proc Natl Acad Sci USA. 2009;106:7227–7232. [PubMed]
  • Clark GD, McNeil RS, Bix GJ, Swann JW. Platelet-activating factor produces neuronal growth cone collapse. Neuroreport. 1995;6:2569–2575. [PubMed]
  • Dent EW, Gertler FB. Cytoskeletal dynamics and transport in growth cone motility and axon guidance. Neuron. 2003;40:209–227. [PubMed]
  • Dujardin DL, Barnhart LE, Stehman SA, Gomes ER, Gundersen GG, Vallee RB. A role for cytoplasmic dynein and LIS1 in directed cell movement. J Cell Biol. 2003;163:1205–1211. [PMC free article] [PubMed]
  • Faulkner NE, Dujardin DL, Tai CY, Vaughan KT, O'Connell CB, Wang Y, Vallee RB. A role for the lissencephaly gene LIS1 in mitosis and cytoplasmic dynein function. Nat Cell Biol. 2000;2:784–791. [PubMed]
  • Feng Y, Olson EC, Stukenberg PT, Flanagan LA, Kirschner MW, Walsh CA. LIS1 regulates CNS lamination by interacting with mNudE, a central component of the centrosome. Neuron. 2000;28:665–679. [PubMed]
  • Hattori M, Adachi H, Tsujimoto M, Arai H, Inoue K. Miller-Dieker lissencephaly gene encodes a subunit of brain platelet-activating factor acetylhydrolase. Nature. 1994;370:216–218. [PubMed]
  • Hirotsune S, Fleck MW, Gambello MJ, Bix GJ, Chen A, Clark GD, Ledbetter DH, McBain CJ, Wynshaw-Boris A. Graded reduction of Pafah1b1 (Lis1) activity results in neuronal migration defects and early embryonic lethality. Nat Genet. 1998;19:333–339. [PubMed]
  • Kappeler C, Saillour Y, Baudoin JP, Tuy FP, Alvarez C, Houbron C, Gaspar P, Hamard G, Chelly J, Metin C, et al. Branching and nucleokinesis defects in migrating interneurons derived from doublecortin knockout mice. Hum Mol Genet. 2006;15:1387–1400. [PubMed]
  • Kholmanskikh SS, Dobrin JS, Wynshaw-Boris A, Letourneau PC, Ross ME. Disregulated RhoGTPases and actin cytoskeleton contribute to the migration defect in Lis1-deficient neurons. J Neurosci. 2003;23:8673–8681. [PubMed]
  • Kholmanskikh SS, Koeller HB, Wynshaw-Boris A, Gomez T, Letourneau PC, Ross ME. Calcium-dependent interaction of Lis1 with IQGAP1 and Cdc42 promotes neuronal motility. Nat Neurosci. 2006;9:50–57. [PubMed]
  • Lee WL, Oberle JR, Cooper JA. The role of the lissencephaly protein Pac1 during nuclear migration in budding yeast. J Cell Biol. 2003;160:355–364. [PMC free article] [PubMed]
  • Liu Z, Steward R, Luo L. Drosophila Lis1 is required for neuroblast proliferation, dendritic elaboration and axonal transport. Nat Cell Biol. 2000;2:776–783. [PubMed]
  • Luo L. Actin cytoskeleton regulation in neuronal morphogenesis and structural plasticity. Annu Rev Cell Dev Biol. 2002;18:601–635. [PubMed]
  • Marcheselli VL, Rossowska MJ, Domingo M-T, Braquet P, Bazan NG. Distinct platelet-activating factor binding sites in intracellular membranes of rat cerebral cortex. J Biol Chem. 1990;265:9140–9145. [PubMed]
  • McNeil RS, Swann JW, Brinkley BR, Clark GD. Neuronal cytoskeletal alterations evoked by a platelet-activating factor (PAF) analogue. Cell Motil Cytoskeleton. 1999;43:99–113. [PubMed]
  • Morris NR, Efimov VP, Xiang X. Nuclear migration, nucleokinesis and lissencephaly. Trends Cell Biol. 1998;8:467–470. [PubMed]
  • Nasrallah IM, McManus MF, Pancoast MM, Wynshaw-Boris A, Golden JA. Analysis of non-radial interneuron migration dynamics and its disruption in Lis1+/- mice. J Comp Neurol. 2006;496:847–858. [PubMed]
  • Prescott SM, Zimmerman GA, Stafforini DM, McIntyre TM. Platelet-activating factor and related lipid mediators. Annu Rev Biochem. 2000;69:419–445. [PubMed]
  • Sapir T, Elbaum M, Reiner O. Reduction of microtubule catastrophe events by LIS1, platelet-activating factor acetylhydrolase subunit. EMBO J. 1997;16:6977–6984. [PubMed]
  • Sasaki S, Shionoya A, Ishida M, Gambello MJ, Yingling J, Wynshaw-Boris A, Hirotsune S. A LIS1/NUDEL/cytoplasmic dynein heavy chain complex in the developing and adult nervous system. Neuron. 2000;28:681–696. [PubMed]
  • Schaar BT. Cortical development deNUDEd. Neuron. 2004;44:213–214. [PubMed]
  • Schmid RS, Shelton S, Stanco A, Yokota Y, Kreidberg JA, Anton ES. alpha3beta1 integrin modulates neuronal migration and placement during early stages of cerebral cortical development. Development. 2004;131:6023–6031. [PubMed]
  • Shu T, Ayala R, Nguyen MD, Xie Z, Gleeson JG, Tsai LH. Ndel1 operates in a common pathway with LIS1 and cytoplasmic dynein to regulate cortical neuronal positioning. Neuron. 2004;44:263–277. [PubMed]
  • Smith DS, Niethammer M, Ayala R, Zhou Y, Gambello MJ, Wynshaw-Boris A, Tsai LH. Regulation of cytoplasmic dynein behaviour and microtubule organization by mammalian Lis1. Nat Cell Biol. 2000;2:767–775. [PubMed]
  • Stuhmer T, Puelles L, Ekker M, Rubenstein JL. Expression from a Dlx gene enhancer marks adult mouse cortical GABAergic neurons. Cereb Cortex. 2002;12:75–85. [PubMed]
  • Tanaka T, Serneo FF, Higgins C, Gambello MJ, Wynshaw-Boris A, Gleeson JG. Lis1 and doublecortin function with dynein to mediate coupling of the nucleus to the centrosome in neuronal migration. J Cell Biol. 2004;165:709–721. [PMC free article] [PubMed]
  • Theiler K. The house mouse: atlas of embryonic development. New York: Springer-Verlag; 1989. p. 178.
  • Tokuoka SM, Ishii S, Kawamura N, Satoh M, Shimada A, Sasaki S, Hirotsune S, Wynshaw-Boris A, Shimizu T. Involvement of platelet-activating factor and LIS1 in neuronal migration. Eur J Neurosci. 2003;18:563–570. [PubMed]
  • Tsai JW, Bremner KH, Vallee RB. Dual subcellular roles for LIS1 and dynein in radial neuronal migration in live brain tissue. Nat Neurosci. 2007;10:970–979. [PubMed]
  • Tsai LH, Gleeson JG. Nucleokinesis in neuronal migration. Neuron. 2005;46:383–388. [PubMed]
  • Yamada M, Toba S, Yoshida Y, Haratani K, Mori D, Yano Y, Mimori-Kiyosue Y, Nakamura T, Itoh K, Fushiki S, et al. LIS1 and NDEL1 coordinate the plus-end-directed transport of cytoplasmic dynein. EMBO J. 2008;27:2471–2483. [PubMed]
  • Zhang G, Assadi AH, McNeil RS, Beffert U, Wynshaw-Boris A, Herz J, Clark GD, D'Arcangelo G. The Pafah1b complex interacts with the reelin receptor VLDLR. PLoS ONE. 2007;2:e252. [PMC free article] [PubMed]
  • Zhang G, Assadi AH, Roceri M, Clark GD, D'Arcangelo G. Differential interaction of the Pafah1b alpha subunits with the reelin transducer Dab1. Brain Res. 2009;1267:1–8. [PMC free article] [PubMed]

Articles from Cerebral Cortex (New York, NY) are provided here courtesy of Oxford University Press