Regulation of Nerve Growth Mediated by Inositol 1,4,5-Trisphosphate Receptors in Growth Cones

See allHide authors and affiliations

Science  27 Nov 1998:
Vol. 282, Issue 5394, pp. 1705-1708
DOI: 10.1126/science.282.5394.1705


The inositol 1,4,5-trisphosphate (IP3) receptor (IP3R) acts as a Ca2+ release channel on internal Ca2+ stores. Type 1 IP3R (IP3R1) is enriched in growth cones of neurons in chick dorsal root ganglia. Depletion of internal Ca2+ stores and inhibition of IP3 signaling with drugs inhibited neurite extension. Microinjection of heparin, a competitive IP3R blocker, induced neurite retraction. Acute localized loss of function of IP3R1 in the growth cone induced by chromophore-assisted laser inactivation resulted in growth arrest and neurite retraction. IP3-induced Ca2+ release in growth cones appears to have a crucial role in control of nerve growth.

In developing and regenerating neurons, the growth cone located at a distal tip of a neurite is thought to be the site governing nerve growth and axon guidance (1). Ca2+ is implicated in various signaling pathways in nerve growth and growth cone behavior (2,3). An optimal range of the concentration of intracellular Ca2+ ([Ca2+]i) in growth cones is required for proper nerve growth (2). Transient increase of [Ca2+]i in the growth cone is a key step in the signaling cascade for laminin-mediated growth cone navigation (4), cell adhesion–mediated neurite extension (5), growth cone migration (6), and axon turning (7). This transient increase of [Ca2+]i is thought to be caused by Ca2+ influx across the plasma membrane through Ca2+ channels such as the L-type voltage-dependent Ca2+ channel (8). IP3 triggers release of Ca2+ from internal Ca2+ stores into the cytosol through the IP3R. Whether IP3signaling and IP3-induced Ca2+ release (IICR) are involved in nerve growth has not been determined.

The IP3R1, a major neuronal member of the IP3R family, is highly expressed in the Purkinje neurons of the mouse cerebellum (9). IP3R1 was detected in chick dorsal root ganglia (DRG) and cerebellum as determined by protein immunoblotting (10) (Fig. 1A). IP3R1 was primarily detected by immunocytochemistry (11) in the central domain of the growth cone and along the neurite of chick DRG neurons (Fig. 1B).

Figure 1

(A) Expression of IP3R1 in chick DRG neurons. Immunoblots of IP3R1 with 4C11 mAb loaded with microsomes from ED11 chick DRG (lane 1), those from ED11 chick cerebellum (lane 2), those from postnatal day (PD) 20 mouse cerebellum (positive control) (lane 3), and those from PD20 IP3R1 knockout mouse cerebellum (negative control) (lane 4) (29). Immunodetection in all samples probed with nonspecific rat IgG was nil. (B) Distribution of IP3R1 in chick DRG neurons cultured on laminin. Phase-contrast images in top panels correspond to fluorescence images in bottom panels. Indirect immunofluorescence with 4C11 mAb is shown as anti-IP3R1. An experiment with nonspecific rat IgG is shown as a control. Other control experiments that omitted primary antibodies and probed with secondary antibodies gave completely negative results. Scale bar, 5 μm.

We characterized the effects of thapsigargin (TG) [which blocks Ca2+-adenosine 5′-triphosphatase (Ca2+-ATPase) on the endoplasmic reticulum and eventually depletes intracellular Ca2+ stores (12)] on neurite extension of chick DRG neurons cultured on laminin (13). Application of TG (10 μM) for 2 hours significantly inhibited neurite extension (P < 0.01) (Fig. 2A). Wash-out of TG resulted in recovery of the extension rate to that of untreated neurites (Fig. 2B). To distinguish effects of TG on neurite extension from those on neuritogenesis (which means initiation of neurite formation), we delayed drug treatments until 2 hours after the cells had been plated. This delayed treatment shifted the neurite length distribution and significantly inhibited the median neurite length 2 hours after addition of TG (Fig. 2C). Therefore, blockage of the TG-sensitive Ca2+ uptake in internal stores appears to inhibit neurite extension.

Figure 2

Effects of inhibitors on neurite extension. (A) Effect of TG treatment. (a) Kinetics of neurite extension under TG (filled circles) or vehicle (open circles) treatments. TG was added at the time of plating (arrow). Data indicate mean neurite length and standard error of the mean (SEM). *, significantly lower than vehicle at each time point; P< 0.01 by Student's unpaired t test. (b) The distribution of DRG neurons with different neurite lengths under TG treatment (4 hours) is plotted as a cumulative histogram. Mean neurite length and SEM are shown in the inset table. Statistical analysis was by one-way analysis of variance (ANOVA), with all pairwise multiple comparison test (Tukey test). TG treatment is significantly different from control (P < 0.05). (B) Wash-out of TG. (a) TG (filled circles) was added at the time of plating (arrow). DRG neurons were treated with TG for 2 hours and washed with culture medium (arrowhead). *, significantly lower than vehicle at each time point; P < 0.01 by Student's unpaired t test. (b) The averages of growth rate (neurite extension length/2 hours) before and after wash-out. (C) Effect of delayed TG treatment. (a) DRG neurons were treated with TG 2 hours after plating (arrow). (b) TG treatment for 2 hours is significantly different from control;P < 0.05 by one-way ANOVA. (D) Effect of heparin and lithium. (a) Inhibition of neurite extension rate by heparin. Heparin was loaded by trituration, and the longest neurite of a cell loaded with heparin was measured at 2- and 3-hour culture time points, in the same cell. Data shown are average median growth rate (neurite extension length/hour) and SEM. *, significantly lower than control reagents; P < 0.01 by Student's unpairedt test. (b) LiCl treatment (filled circles) for 2 hours is significantly different from control (open circles);P < 0.05 by one-way ANOVA. (E) Microinjection of heparin or a control analog, de-N-sulfated heparin. (a) A growth cone is observed before the microinjection. Viability of the growth cone was verified during this period. (b) The neurite begins to retract and bend after injection (15 min after injection). (c) Heparin injection results in neurite retraction with bending, but not filopodial retraction (30 min after injection). (d, e, and f) Microinjection of a control analog did not affect growth cone behavior. Scale bar, 10 μm.

We also examined the effect of lithium ion and heparin, a competitive IP3R blocker, on neurite extension (13,14). Lithium is assumed to block the recycling of IP3 into inositol by inhibiting the hydrolysis of intermediate inositol phosphates, and this compound eventually results in loss of the IP3 signal (15). Exposure of cells to LiCl (50 mM) for 2 hours significantly inhibited neurite extension (Fig. 2D). In DRG neurons loaded with heparin by trituration (14), the neurite extension rate was significantly inhibited (Fig. 2D). The heparin analog, de-N-sulfated heparin, had no effect. TG, lithium, and heparin had no effect on growth cone morphology. We microinjected heparin or de-N-sulfated heparin into the cell body of DRG neurons and observed growth cone behavior by time-lapse video microscopy (16). Microinjection of heparin but not the control analog resulted in growth arrest of neurites and subsequent neurite retraction with bending (Fig. 2E, Table 1). Motility of filopodia was unaffected by heparin (Table 1). Because TG, lithium, and heparin had inhibitory effects on neurite extension, IP3 signaling and IP3R functions are implicated in neurite extension.

Table 1

Heparin microinjection and micro-CALI of IP3R1. Comparison of growth cone behavior with control experiments on phenotypes observed after heparin microinjection and micro-CALI of IP3R1 is shown. Growth arrest is defined as neurite extension after manipulation, a more than 90% reduction of the neurite extension rate before manipulation, and a transient neurite retraction within 5 min after manipulation. Neurite retraction is defined as neurite shown a decrease of its length by continuous retraction after manipulation. No phenotype was observed before microinjection of heparin and laser irradiation for CALI experiments, in all cases tested (n = 85 in total). Comparison of neurite extension rate after CALI of IP3R1 (from +5 min to +10 min shown in Fig. 3, C, D, and E) with that before CALI (from −5 min to +0 min) is shown. Data shown are the averages of neurite extension rate and SEM (micrometers per 5 min).

View this table:

To examine functions of IP3R in regions of the growth cone, we used the chromophore-assisted laser inactivation (CALI) technique, a method of protein ablation that has high spatial and temporal specificity (17–21). CALI uses laser irradiation to direct spatially restricted photogenerated hydroxyl free radical damage to targeted proteins through chromophore (malachite green, MG)-conjugated antibodies (18), thereby allowing one to inactivate protein function preferentially without affecting other cellular functions (17, 19–21). We used microscale CALI (micro-CALI) to test whether IP3R1 in the growth cone would potentiate nerve growth regulation. Effects of CALI of IP3R1 on IICR activity in vitro were measured with a luminescence spectrometer and the mouse cerebellum microsome fraction, in which the IP3R1 is specifically concentrated (9, 22). The 4C11 monoclonal antibody (mAb) itself blocks neither IP3 binding nor Ca2+release from microsomes (23). CALI of IP3R1 with chromophore-labeled 4C11 mAb caused about a 50% reduction in IICR activity. This inactivation depended on both the presence of chromophore-labeled 4C11 mAb and laser irradiation (Fig. 3A). CALI with chromophore-labeled nonspecific immunoglobulin G (IgG) did not affect IICR activity. CALI of IP3R1 had no inhibitory effect on Ca2+uptake activity by Ca2+-ATPase. Thus, CALI of IP3R1 causes a potent reduction of IICR activity in living cells.

Figure 3

CALI of IP3R1. (A) CALI of IP3R1 in IICR activity in vitro. Relative IICR activity was calculated to normalize by the following equation: (IICR)/(Ca2+ release by ionomycin treatment). Ratio to the control experiment without treatment was calculated and displayed. Data shown are the averages and SEM. *, significantly different from all control groups; P < 0.02 by Student's unpaired t test. (B) Trituration loading of chromophore (MG)–labeled 4C11 mAb was confirmed by immunocytochemistry with a secondary antibody (2 hours of culture after trituration). Unloaded neurons were not stained. Scale bar, 5 μm. (C) Micro-CALI of IP3R1 in growth cone. Sequential panels from top to bottom show the time course of growth cone behavior during the observation (every 5 min). (a) A growth cone is first observed for 5 min before laser irradiation (from −5 min to +0 min). (b) A region of growth cone is chosen for micro-CALI (white outline: laser spot) and laser irradiated for 5 min (from +0 to +5 min). (c) Theneurite stalls and starts retracting after laser irradiation (+5 min). Neurite bending also appears (arrow). (d and e) Neurite is retracting, whereas filopodia are not retracting (from +10 to +15 min). (D) Micro-CALI of IP3R1 in neurite shaft does not affect neurite extension. (E) Control micro-CALI with MG-nonspecific IgG (micro-CALI of IgG) in growth cone has no effect. Scale bars in (C), (D), and (E), 10 μm.

We loaded chromophore-labeled 4C11 mAb or nonspecific IgG into DRG neurons by trituration, monitored loading of antibodies into neurites and growth cones by observing coincidental loading of fluorescein isothiocyanate (FITC)–conjugated dextran (molecular size: 150 kD) or FITC-labeled nonspecific IgG, and confirmed loading by immunocytochemistry (Fig. 3B). The region of interest was laser irradiated for 5 min, and the growth cone was observed by time-lapse video microscopy before, during, and after laser irradiation (24). Micro-CALI of IP3R1 in growth cones resulted in growth arrest of neurites and subsequent neurite retraction, within 5 min after termination of the irradiation (Fig. 3C). Growth arrest of neurites or neurite retractions were observed in 19 of 26 (73.1%) or 14 of 26 (53.8%) IP3R1 micro-CALI experiments (cells), respectively (Table 1; P < 0.01 by χ2 test). In 7 of 26 cases, the neurite retraction was accompanied by neurite bending (Fig. 3C, panel c, and Table 1). Filopodial motility was not affected (Fig. 3C andTable 1). In contrast, parallel micro-CALI experiments with chromophore-labeled nonspecific IgG showed no perturbation of growth cone behavior or neurite extension (Fig. 3E and Table 1). Laser irradiation alone toward the growth cone had no effect on growth cone behavior (n = 11). Neurite retraction appears not to be a nonspecific trauma response to CALI effects, because micro-CALI of calcineurin (19), myosin-V (20), talin (21), or vinculin (21) in growth cones causes various different phenotypes in filopodia but not in neurites. In contrast to the results of micro-CALI in growth cones, micro-CALI of IP3R1 in regions of the neurite shaft did not affect growth cone behavior (Fig. 3D and Table 1). This suggests that IP3R1 in growth cones has an important role in neurite extension, whereas IP3R1 in neurite shafts is unlikely to function in neurite extension. The IP3R1 may need to be recruited to the growth cone to function with other signaling molecules.

Local loss of IP3R1 function within the growth cone results in growth arrest and neurite retraction, even though the neurons do undergo Ca2+ influxes across the plasma membrane. Although a change in [Ca2+]i within the growth cone in response to inactivation of IP3R1 was not detectable in calcium imaging experiments with a fura-2 indicator (25), our findings do suggest that IICR within the growth cone, which may cause a small transient change in [Ca2+]i, is a key regulatory factor governing neurite extension. Although the molecular events downstream of Ca2+ fluxes are not well understood in nerve growth, calmodulin (CaM) (26), an intracellular Ca2+ receptor protein; calcineurin (19), a CaM-dependent protein phosphatase; and Ca2+-CaM–dependent protein kinase II (27) regulate nerve growth. Double immunostaining of IP3R1 and microtubules revealed that the distribution of IP3R1 was associated with that of microtubules such as tubulin in the growth cone (28). Thus, IICR through IP3R1 in the growth cone and its downstream effectors might act locally to regulate microtubule assembly and promote neurite extension. This notion is also supported by our findings that IP3R1 appears not to be distributed in filopodia and that inactivation of this molecule does not affect filopodial motility. [Ca2+]imobilization by IICR could modulate Ca2+ influxes through Ca2+ channels. Therefore, IICR from internal stores and Ca2+ influx may act together to direct nerve growth.

  • * To whom correspondence should be addressed. E-mail: kohtaro{at}


Stay Connected to Science

Navigate This Article