Dynamics of peptidergic secretory granule transport are regulated by neuronal stimulation

Abstract

Background: Peptidergic neurons store and secrete the contents of large dense core vesicles (LDCVs) from axon terminals and from dendrites. Secretion of peptides requires a highly regulated exocytotic mechanism, plus coordinated synthesis and transport of LDCVs to their sites of release. Although these trafficking events are critical to function, little is known regarding the dynamic behavior of LDCVs and the mechanisms by which their transport is regulated. Sensory neurons also package opiate receptors in peptide-containing LDCVs, which is thought to be important in pain sensation. Since peptide granules cannot be refilled locally after their contents are secreted, it is particularly important to understand how neurons support regulated release of peptides.

Results: A vector encoding soluble peptidylglycine alpha-hydroxylating monooxygenase fused to green fluorescent protein was constructed to address these questions in cultured primary peptidergic neurons of the trigeminal ganglion using time lapse confocal microscopy. The time course of release differs with secretagogue; the secretory response to depolarization with K+ is rapid and terminates within 15 minutes, while phorbol ester stimulation of secretion is maintained over a longer period. The data demonstrate fundamental differences between LDCV dynamics in axons and growth cones under basal conditions.

Conclusions: Under basal conditions, LDCVs move faster away from the soma than toward the soma, but fewer LDCVs travel anterograde than retrograde. Stimulation decreased average anterograde velocity and increases granule pausing. Data from antibody uptake, quantification of enzyme secretion and appearance of pHluorin fluorescence demonstrate distributed release of peptides all along the axon, not just at terminals.

Figure 1

Trigeminal ganglion neurons are peptidergic. Trigeminal ganglia from P3-P5 rat pups were dissociated and maintained in culture for 2 days. Cells were fixed and processed for immunocytochemistry using antibodies to CGRP (green), substance P (red), and βIII tubulin (blue).

Figure 2

Expression of PHM-GFP allows quantification of LDCV dynamics and secretion. (A) Dissociated trigeminal ganglion neurons were transfected with vector encoding PHM-GFP and grown in culture for 2 days. Cells were fixed and processed for immunocytochemistry; PHM-GFP fluorescence was visualized in puncta throughout the cell soma and processes (green) along with CGRP (red). (B) A growth cone (labeled) and the preceding region of its axon are shown with immunofluorescent staining of CGRP (green) and βIII tubulin (blue); filamentous actin was visualized with a fluorescent phalloidin conjugate (red). F-actin enriched stress fibers in a non-neuronal cell are seen at the bottom of the image. (C) Duplicate cultures were analyzed at room temperature (23°C) and at 30°C; 30 minute medium collections were made under basal conditions and during stimulation with 1 μM PMA. Secretion was quantified by measuring units of PHM activity (pmol product/h) in basal and stimulated media and in cell extracts. (D) Cultures were stimulated for 5, 15, or 30 minutes at 37°C with 1 μM PMA or 50 mM KCl. Secretion was quantified by assaying units of PHM activity in basal and stimulated media; stimulated secretion was calculated by subtracting basal from total secretion (gray line); separate experiments with cultures of different densities were performed for PMA and KCl.

Figure 3

LDCV dynamics in growth cones and axons are fundamentally different under basal conditions. For analysis under basal conditions, neurons were kept in phenol-free L15 medium at or above 30°C. LDCVs containing PHM-GFP were tracked in growth cones and distal portions of axons. Examples of X-Y LDCV trajectories are shown; one showed rapid, directed motion (#219), and the other moved in a slow, dithering manner (#235). Velocity data for identified LDCV trajectories were pooled from 6 cells. The distribution of binned data is shown in the histogram (B), with the cumulative plot (C) emphasizing the differences between the two regions. LDCV trajectories were also analyzed to determine the distribution of linearity (D). Linearity was calculated by comparing how close the actual path was to a straight line representing the distance between the first and last point tracked. Cumulative direction was analyzed based on the mean angle of the trajectory (E) and assigned a score of 1 if anterograde or 0 if retrograde. N = 2760 LDCVs tracked.

Figure 4

An intact actin cytoskeleton is required for efficient LDCV transport. Trigeminal ganglion neurons were treated with 5 μM cytochalasin B for 1 hour and then fixed and processed for immunocytochemistry (A). Filamentous actin was visualized by staining with a TRITC-phalloidin conjugate (red) and microtubules were stained with an antibody to βIII tubulin (green). Inset: phalloidin staining is shown at higher magnification to illustrate puncta that appeared in cytochalasin-treated neurons. Neurons expressing PHM-GFP were pretreated with cytochalasin B in the same manner before live imaging; cumulative velocity (B) and cumulative linearity (C) were plotted as in Figure 3.

Figure 5

Secretagogues increase linearity but not cumulative velocity in axons. LDCVs containing PHM-GFP were imaged in the presence of PMA or KCl; the effects of secretagogue on dynamics were compared to behaviors observed under basal conditions. LDCV velocity and linearity were assessed in axons; data for each treatment group were pooled from 6 cells each. The distribution of cumulative velocity (A) and linearity (B) is shown for LDCVs under basal conditions and during stimulation with PMA and KCl. Both PMA and KCl had a significant effect (p < 0.01) on linearity, but not on velocity.

Figure 6

Secretagogues differentially affect instantaneous velocity and direction. Instantaneous velocity and direction were calculated for tracked LDCVs in neurons kept under basal conditions or exposed to PMA or KCl for 5 min. LDCVs moving in the retrograde direction were assigned a negative sign, and those moving in the anterograde direction received a positive sign (A); the distribution of binned data is shown for representative cells from each treatment group. The fraction of LDCVs stopped or moving in a given direction (B) is plotted as an average; the relationship between direction and velocity was also analyzed (C). One-way ANOVA with Bonferroni post-test (SPSS package): *, p < 0.02; **, p < 0.01.

Figure 7

PMA increases LDCV pausing time in axons. The effect of secretagogue on the percentage of time that LDCVs in axons spent paused was calculated; to be categorized as 0% paused, a LDCV had to be in continuous motion. The distribution of binned data is shown.

Figure 8

LDCV secretion occurs throughout neurons. Neurons were transfected with vector encoding PAM-GFP. After 2 days in culture, cells were stimulated with PMA in the presence of PAM ectodomain antibody for 15 min; internalized antibody was subsequently visualized in fixed, permeabilized cells with a Cy3 conjugated secondary antibody. Internalized PAM antibody was detected throughout the cell soma (top), along processes (bottom) and at growth cones (labeled). In the absence of secretagogue, almost no antibody uptake was observed (not shown).

Figure 9

Fusion events are detected by increases in PHM-pHluorin fluorescence. Neurons were co-transfected with vectors encoding PHM-pHluorin and DsRed. All images in this figure are from a single microscope field of one neuron; four other neurons yielded similar results. After 2 days in culture, growth cones of neurons expressing DsRed were identified, and image frames were acquired under basal conditions (A, top left). A horizontally oriented line was placed through the center of the axon and growth cone. Scans were acquired continuously along the indicated line during PMA stimulation; two examples are shown (C, left and middle). Distance in the x dimension is represented on the x-axis, with the cell soma out of the picture to the left and the growth cone to the right. Time is represented on the y-axis; the total time for each image displayed is 6 sec. At the conclusion of the experiment, the total LDCV population was visualized by adding 5 mM NH₄Cl to the bath solution; a frame (B, right) and line scan (C, right) are shown.