Pigment granule translocation in red ovarian chromatophores from the palaemonid shrimp Macrobrachium olfersi (Weigmann, 1836): Functional roles for the cytoskeleton and its molecular motors

Abstract

The binding of red pigment concentrating hormone (RPCH) to membrane receptors in crustacean chromato- phores triggers Ca2+/cGMP signaling cascades that activate cytoskeletal motors, driving pigment granule translocation. We investigate the distributions of microfilaments and microtubules and their associated molecu- lar motors, myosin and dynein, by confocal and transmission electron microscopy, evaluating a functional role for the cytoskeleton in pigment translocation using inhibitors of polymer turnover and motor activity in vitro. Microtubules occupy the chromatophore cell extensions whether the pigment granules are aggregated or dispersed. The inhibition of microtubule turnover by taxol induces pigment aggregation and inhibits re- dispersion. Phalloidin-FITC actin labeling, together with tannic acid fixation and ultrastructural analysis, reveals that microfilaments form networks associated with the pigment granules. Actin polymerization induced by jasplaquinolide strongly inhibits RPCH-induced aggregation, causes spontaneous pigment dispersion, and inhibits pigment re-dispersion. Inhibition of actin polymerization by latrunculin-A completely impedes pigment aggregation and re-dispersion. Confocal immunocytochemistry shows that non-muscle myosin II (NMMII) co- localizes mainly with pigment granules while blebbistatin inhibition of NMMII strongly reduces the RPCH re- sponse, also inducing spontaneous pigment dispersion. Myosin II and dynein also co-localize with the pigment granules. Inhibition of dynein ATPase by erythro-9-(2-hydroxy-3-nonyl) adenine induces aggregation, inhibits RPCH-triggered aggregation, and inhibits re-dispersion. Granule aggregation and dispersion depend mainly on microfilament integrity although microtubules may be involved. Both cytoskeletal polymers are functional only when subunit turnover is active. Myosin and dynein may be the molecular motors that drive pigment aggregation. These mechanisms of granule translocation in crustacean chromatophores share various features with those of vertebrate pigment cells.

1. Introduction

Animal coloration has long provoked scientific curiosity, owing not only to esthetic appeal, but as an aspect of biological diversity that determines an individual’s performance and ultimately, species’ adapta- tion (Forsman et al., 2008). Color changes play important roles in mate signaling, warning to potential predators, thermoregulation, protection against UV radiation, and in camouflage. Among the Crustacea, camou- flage strategies like aposematism, mimetism and disruption (Stevens and Merilaita, 2009) depend on a strong, light-mediated capability to imitate substrate chromatic characteristics mediated by pigment movements within pigment bearing cells or chromatophores (reviewed in McNamara and Milograna, in press).
Like most crustacean chromatophores (Bauer, 2004), those of palaemonid shrimps are highly asymmetrical and consist of a spherical cell body containing a single nucleus and the main pigment granule mass, and one or two cell extensions that ramify into fine dendrites through which granules migrate (McNamara, 1981). These single cells form assemblages of 10–15 structurally and functionally linked effectors known as chromatosomes (McNamara and Taylor, 1987). Ultrastructurally, palaemonid chromatophores are characterized by various types of pigment granules, microtubules, well-developed, re- gionally differentiated smooth endoplasmic reticulum, occasional semi-spherical mitochondria, rough endoplasmic reticulum cisternae, polyribosomes and infrequent Golgi bodies (McNamara, 1981).

Pigment aggregation is characterized by centripetal granule translocation into the chromatophore cell body from the cell extensions,
triggered by the release of red pigment concentrating hormone (RPCH) into the hemolymph via the sinus gland. RPCH is a small peptide neurosecretion or chromatophorotropin originating mainly in the eye- stalk X-organ (Fingerman, 1985). Pigment dispersion, often mediated by pigment dispersing hormone (PDH), consists of centrifugal granule translocation, and the spread of granules into the cell extensions (Rao, 1985). The aggregation mechanism is triggered by RPCH binding to a putative 7-span G protein-linked membrane receptor (see Hammers et al., 2011; Anders, 2012; Milograna et al., unpublished data) whose ac- tivation increases intracellular second messengers like Ca2+ and cGMP that activate interdependent signaling cascades (Nery et al., 1998; Ribeiro and McNamara, 2009; Milograna et al., 2010). Ultimately, these result in interactions between actin microfilaments and myosin motors (McNamara and Ribeiro, 1999; Boyle and McNamara, 2006), and also between microtubule arrays and other motor molecules like kinesin and dynein (McNamara and Boyle, 2009), all collectively known as cytoskeletal, mechano-chemical, force-transducing proteins. However, the mechanisms of pigment granule translocation and the specific proteins that regulate such molecular motors in crustacean chromatophores remain obscure.

Various molecular motors may underpin granule translocation in crustacean pigmentary effectors since different pigments can migrate simultaneously in opposing directions in polychromatic chromato- phores (McNamara, 1981). Further, ovarian and nerve cord red chromatophores contain two granule types with different surface char- acteristics: large (≈450 nm diameter) membrane limited granules; and small (≈140 nm diameter) carotenoid granules that lack a membrane (McNamara and Sesso, 1982, 1983). Pigment aggregation exhibits a biphasic velocity profile, displaying rapid and slow components puta- tively dependent on the activity of different motors (McNamara and Ribeiro, 1999). Actin-associated myosin II may drive pigment aggrega- tion, while pigment dispersion may be powered by microtubule- associated kinesin (Boyle and McNamara, 2006).

Studies on vertebrate pigment cells show that kinesin I and II partic- ipate in mammalian (Hara et al., 2000) and amphibian melanosome transport (Gross et al., 2002). An actin–myosin system functions in granule transport in rat (Wu et al., 1998), fish (Rodionov et al., 2003) and amphibian (Rogers et al., 1999) melanophores. Non-muscle myosin II isoforms play a role in cell cleavage, spindle formation, contractile activity of lamellipodia (Betapudi, 2010), and in the tractional forces un- derlying cell migration and motility (Lo et al., 2004). This myosin is the one of the motors responsible for vesicle transport (DePina et al., 2007), and is often associated with organelle membranes (DeGiorgis et al., 2002). Non-muscle myosin II seems a likely candidate for pigment granule transport in chromatophores, both crustacean and vertebrate, and should be better investigated.

Granule translocation in many vertebrate pigment cells involves complex molecular interactions, such as mutual modulation among dif- ferent motor classes (Even-Ram et al., 2007; Ali et al., 2008) that com- pete to move cargos towards their respective cytoskeletal polymers. Cargo-bearing motors may alternate from one cytoskeletal element to another during transport, and motors traditionally associated with microtubular transport, like kinesin for example, may participate in transport along actin microfilaments (Even-Ram et al., 2007). Microtubule- and microfilament-based motor proteins and pigment granules are functionally associated in fish and amphibian melano- phores (Gross et al., 2002).

While such cellular machinery has been examined in vertebrate chromatophores, little is known of the mechanisms that underlie gran- ule translocation in crustacean pigment cells. Here, we investigate the cytoskeletal polymers and respective molecular motors that drive pig- ment movements in the red ovarian chromatophores of a freshwater shrimp, Macrobrachium olfersi. We examine the locations and roles of actin microfilaments and microtubules, of non-muscle myosin II and other myosins, and of dynein in granule translocation. Understanding these mechanisms is fundamental to comprehending the functioning of pigment cells and color change mechanisms in invertebrates in gen- eral and their homologies with vertebrate pigment cells.

2. Material and methods

2.1. Ethical procedures

Collection of animals, their maintenance in the laboratory and use were authorized by the relevant Brazilian federal agency, the Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais Renováveis (IBAMA/DIREN) under permits #070/2004 and #18559-1 to JCM.Female freshwater shrimp, Macrobrachium olfersi (Weigmann 1836), possessing immature ovaries with red chromatophores visible in the fibrous ovarian capsule, were collected by sieving the marginal vegetation of the Paúba River (23° 47′ 51.30″ S; 45° 32′ 33.61″ W), São Paulo State, Brazil. This species is neither endangered nor protected. The shrimps were transported to the laboratory in 30-L carboys contain- ing aerated water from the collection site where they were maintained at room temperature (23 °C) in 80-L tanks furnished with aerated Paúba River or local spring water. The shrimps were fed diced beef, carrot and beetroot on alternate days.

2.2. Immunocytochemistry: epifluorescence and confocal microscopy

For preparations requiring maximum red pigment dispersion within the chromatophores, the shrimps were held on a black background for 2 h before dissection. The intact ovaries were dissected and fixed/permeabilized in sodium biphosphate buffer (PBS; 100 mmol L−1 NaH2PO4 and 100 mmol L−1 Na2HPO4, pH 7.4), containing 4% p-formaldehyde (Sigma-Aldrich, USA) and 0.3% Triton X-100 (Sigma-Aldrich, USA) for 15 min at room temperature (23 °C). The ovarian preparations were rinsed in 100 mmol L−1 glycine in PBS for 5 min, then incubated at room temperature for 1 h in PBS containing 0.1% Triton X-100, 2% bovine serum albumin and 5% goat serum (Sigma-Aldrich, USA) to block non-specific antibody binding.

The whole ovaries were then incubated with their respective prima- ry antibodies in PBS [mouse monoclonal anti-β-tubulin, diluted 1:100 (Sigma-Aldrich, USA); rabbit polyclonal anti-skeletal and smooth muscle myosin, 1:50 (Sigma-Aldrich, USA); polyclonal anti-non- muscle myosin II, 1:100 (a gift from Prof. Roy Larson, FMRP); and mono- clonal anti-dynein heavy chains, 1:100 (Sigma-Aldrich, USA)] for 48 h at 4 °C, rinsed in PBS and incubated for 3 h with their respective Alexa Fluor 488-conjugated secondary antibodies (goat anti-rabbit IgG or goat anti-mouse IgG, diluted 1:400, Life Technologies, USA, gifts from Prof. Roy Larson, FMRP) at room temperature. They were then re- fixed in 4% p-formaldehyde in PBS for 15 min when necessary. The whole ovaries were carefully bisected in PBS with ophthalmic scissors, and the fine ovarian capsules containing the red chromatophores were separated from the oocytes, mounted on glass slides with Vectashield Mounting Medium and covered with 0.25 mm Knítel Glaser cover slips. Slides were stored in the dark at 4 °C.

Negative control preparations lacking primary antibodies were prepared using ovarian capsules containing chromatosomes. These cap- sules were incubated with either goat anti-mouse Alexa Fluor 488- or goat anti-rabbit Alexa Fluor 488-conjugated secondary antibodies. Such preparations showed no significant labeling (Supplementary Fig. 1). These red pigments absorb light at wavelengths between 400 and 600 nm, with peak absorbance at 500 nm and peak emission at 568 nm (Boyle, 2005).

Phalloidin-fluorescein isothiocyanate (phalloidin-FITC) (495/520 nm, Sigma-Aldrich, USA) (Wehland et al., 1977) was used to label actin micro- filaments in the chromatophores. After fixing and permeabilizing as above, whole ovaries were also incubated for 3 h in 300 nmol L−1 phalloidin-FITC in PBS at room temperature. Their ovarian capsules were then dissected and mounted as described above.Fluorescence images were acquired using a Leica DFC 300FX camera coupled to a Leica DM 5000B microscope, or Leica TCS SP2 or SP5 laser scanning confocal microscopes.

2.3. Transmission electron microscopy

To reveal the cytoskeleton during pigment dispersion, intact ovaries containing chromatophores in the limiting ovarian capsule were firstly incubated in 10−8 mol L−1 RPCH for 15 min at room temperature (23 °C) to fully aggregate the pigments. These were then dispersed by rinsing for 30 min in physiological saline (PS, 364.0 ± 1.9 mOsm/kg H2O, pH 7.4, containing [in mmol L−1] Na+ 180, K+ 5, Ca2+ 5.5 and Mg2+ 1 as chlorides [≈198 mmol L−1], Na2HCO3 2.5, HEPES 5 and glucose 2) (McNamara et al., 1990).

During pigment dispersion, the ovaries with chromatophores in the ovarian capsules were fixed in a primary fixative (containing in mM, NaCl 100, KCl 5, CaCl2 5, MgCl2 1, sodium cacodylate 100, p-formaldehyde 200, glutaraldehyde 250 and 2% tannic acid, pH 7.4) for 1 h at room temperature, then rinsed twice in PS on ice. Secondary fixation was performed for 1 h at 4 °C in 2% osmium tetroxide in 200 mmol L−1 sodium cacodylate-buffered saline, then rinsed twice in 2% aqueous uranyl acetate for 1 min and incubated in uranyl acetate for 1 h (Goldman et al., 1979).

The ovaries were dehydrated on ice in an ethanol series, followed by propylene oxide at room temperature, then embedded in Araldite 502 resin (Pelco, Ted Pella Inc., Redding CA, USA) after infiltration overnight on a rotary mixer. Thin sections (80 nm) of the chromatophores within the ovarian capsule were cut on a diamond knife using a Sorval Porter-Blum MT2-B ultramicrotome, stained with aqueous 2% uranyl acetate and 2% lead citrate (Reynolds, 1963) and observed using a Philips EM 208 electron microscope at an accelerating voltage of 100 kV.

2.4. Physiological assays: cytoskeletal polymers and molecular motors

The ovarian chromatosome preparation employed here has been de- scribed in detail previously (McNamara and Ribeiro, 1999). Briefly, after holding the shrimp on a black background for 2 h to induce maximum pigment dispersion in the ovarian chromatosomes, the ovary was dis- sected, sectioned transversally and the anterior half transferred to an acrylic micro-perfusion chamber (154 μL volume) containing physio- logical saline (PS), which was then sealed with a coverslip and gravity perfused (0.7 mL/min, 25 mm H2O pressure). The preparations were observed at 200 × magnification employing transmitted and incident light using a Wild M10 stereoscopic microscope coupled to a Sony DXC-151A CCD video camera and Trinitron monitor (Sony Corporation, Tokyo, Japan).

The chromatosome preparations were perfused with the various in- dividual reagents for 30 to 50 min prior to adding RPCH. Their effect on pigment distribution and translocation velocity was evaluated by di- rectly quantifying pigment distribution in the cell extensions of chromatosomes measuring 180 to 220 μm in diameter, at 2-min inter- vals, against a calibrated ocular reticule. Each experiment was repeated 7–12 times, using a single chromatosome from each preparation. All experiments were performed at room temperature (≈23 °C).

The chromatophores were initially stabilized by 10-min perfusion with PS prior to reagent testing; PS was also used to wash out RPCH and reagents. In low calcium, EDTA-buffered saline (LCS) (383 ± 0.3 mOsm/kg H2O, N = 5), Ca2+ was replaced by 5.6 mmol L−1 choline chloride in 2 mmol L−1 ethylenediamine tetra-acetic acid (EDTA, Sigma, MO, USA) (residual Ca2+ 9.2 × 10−11 M). LCS was used to induce maxi- mum possible pigment dispersion when required, since low extracellular Ca2+ likely causes pigment dispersion by inhibiting RPCH-receptor bind- ing and the pigment aggregation ensuing from a putative Ca2+-induced Ca2+-release mechanism (McNamara and Ribeiro, 2000).

Chromatophore pigments were stimulated to aggregate using red pigment-concentrating hormone (RPCH, Peninsula Laboratories Inc., San
Carlos, CA, USA) perfused at a final pharmacological concentration of 5 or 10 nmol L−1 where indicated (Josefsson, 1983; Rao and Fingerman, 1983; Fingerman, 1985) or a calcium ionophore (A23187, 25 μmol L−1). The roles of microtubules, actin microfilaments and their respective
molecular motors in pigment translocation were evaluated using the following inhibitors and activators (as final concentrations in the per- fusate). All stock solutions were prepared in distilled water, 0.25–0.5% DMSO or 0.25% methanol. All salts and reagents used were purchased from Sigma (St. Louis, MO, USA) unless otherwise indicated.

3. Results

3.1. The chromatophore cytoskeleton and its associated molecular motors

The ovarian chromatosomes are embedded in a fine ovarian capsule consisting of fibroblasts and other cell types that exhibit abundant mi- crotubules and microfilaments. The red pigment granules contained within the chromatosomes are intensely autofluorescent at the rhoda- mine wavelength (550 nm), as seen by epifluorescence (Fig. 1A) and confocal microscopy (Fig. 1C, D).

The chromatophore cytoskeleton is rich in microtubules, particularly in the cell extensions, as clearly visualized by epifluorescence and con- focal fluorescence microscopy in chromatosomes with aggregated pig- ments labeled with anti-β tubulin antibodies (Fig. 1A, D). Narrow central cores of one or two microtubule bundles project from the edge of the aggregated pigment mass into the long cell extensions (Fig. 1A, D, F, G). These extensions are not visible in phase contrast fields (Figs. 3B, E, 5A), since they contain no identifying pigment granules. Minor microtubule bundles are also present within the pigment mass that forms each constituent chromatophore (Fig. 1D, G).

Actin microfilaments are distributed abundantly throughout the fibro- blasts that constitute the ovarian capsule in which the chromatosomes are immersed (Fig. 2B). Owing to microfilament density in the capsule tis- sue, it is difficult to visualize actin within the chromatophore cytoplasm (Fig. 2B, C) by confocal microscopy. However, microfilaments often per- fectly accompany the edges of pigment-filled extensions (Fig. 2D).

3.2. Physiological assays of cytoskeletal and molecular motor function

Ovarian chromatophores exhibit typical pigment aggregation– dispersion cycles of about 30-min duration each, reversibly induced by red pigment concentrating hormone, cyclic dibutyryl guanosine monophosphate or the calcium ionophore A23187 (Fig. 6A–H, see Boyle and McNamara, 2008; Ribeiro and McNamara, 2009).

4. Discussion

Our findings disclose a role for cytoskeletal actin and tubulin, and their associated molecular motors, during pigment granule movements in crustacean chromatophores. We demonstrate actin microfila- ments unequivocally for the first time and reveal that polymerized microtubules occupy the cell extensions during both granule aggrega- tion and dispersion; we reveal physiological roles for both these poly- mers in pigment movements. Myosin and dynein are intimately associated with the granules themselves, and we explore the functional role of these motors in granule translocation. We also reveal the partic- ipation of non-muscle myosin II in pigment translocation, a novel func- tion for this motor.

4.1. Cytoskeletal polymers and pigment granule translocation

4.1.1. Microtubules

Confocal immunofluorescence microscopy shows that the microtu- bule bundles in M. olfersi chromatophores extend fully from the perikar- yon into the cell extensions whether the pigments are dispersed or aggregated, suggesting sustained microtubule polymerization/tubulin turnover independently of the state of pigment distribution. A mi- crotubular cytoskeleton also projects into the chromatophore cell ex- tensions in the shrimps Palaemon serratus (Chassard-Bouchaud and Hubert, 1971), P. affinis (McNamara, 1980) and Palaemonetes vulgaris (Robison and Charton, 1973). A functional role for microtubules in pig- ment granule translocation in crustacean chromatophores, however, has been the object of a protracted and controversial debate. Pigment aggregation in epidermal chromatophores of P. affinis is unaffected by the microtubule disrupting agents colchicine and vinblastine (McNamara, 1980). Further, colchicine does not inhibit RPCH- or A23187-induced aggregation in M. olfersi ovarian chromatophores, but does cause partial aggregation in cells with dispersed pigments (McNamara and Ribeiro, 1999). Colchicine inhibits aggregation in Palaemonetes vulgaris ovarian chromatophores only at non- physiological concentrations (25 mM, Robison and Charton, 1973), like- ly a nonspecific effect, since lumicolchicine, a UV-derived colchicine an- alog that does not interact with microtubules, also inhibits RPCH- induced aggregation in M. potiuna chromatophores (Tuma et al., 1995). Like colchicine, the inhibition by taxol of tubulin turnover dem- onstrated here caused considerable pigment aggregation in M. olfersi chromatophores, showing that impeding addition of tubulin subunits to the microtubule positive poles can induce spontaneous aggregation. This suggests that microtubules and their associated motors may sus- tain pigment dispersion within the cell extensions. Unlike colchicine, taxol also decreases RPCH-triggered aggregation velocity (Table 1), supporting this notion.

These findings suggest that the constant turnover of tubulin subunits may play a role in pigment aggregation, and particularly in maintaining granule dispersion. However, given that microtubules remain polymer- ized in the pigment-aggregated state, and that depolymerization does not affect aggregation, microtubule depolymerization clearly does not underpin pigment aggregation.

4.1.2. Actin microfilaments

We demonstrate unequivocally abundant actin microfilaments in crustacean chromatophores for the first time by transmission electron microscopy using tannic acid-fixed preparations from M. olfersi. It is challenging to confirm actin microfilaments within the chromatophore cytoplasm using FITC-conjugated phalloidin, owing to interference from actin filaments in the ovarian capsule fibroblasts. However, actin micro- filaments clearly define pigment granule trails, suggesting a functional role in translocation. Our physiological assays corroborate these structural findings, and emphasize clearly a crucial role for actin microfilaments in pigment translocation. Firstly, jasplaquinolide, a microfilament-polymerizing agent, significantly inhibits RPCH-induced aggregation and completely inhibits pigment dispersion (Table 1). Appar- ently, stable actin polymers in the cytoplasm impede granule movements; in contrast, actin filament depolymerization or active turn- over seems to accompany pigment translocation. Secondly, latrunculin- A, an inhibitor of microfilament polymerization completely inhibits RPCH-triggered pigment aggregation (Table 1). Directly increasing intra- cellular [Ca2+] by using a Ca2+ ionophore to bypass signaling events asso- ciated with the cell membrane only provokes minor pigment aggregation. Pigment dispersion is also completely inhibited (Table 1), showing that granule translocation is highly dependent on actin polymerization. These findings suggest strongly that the mechanisms of both pigment ag- gregation and dispersion rely on actin monomer–microfilament turnover. A role for actin in crustacean pigment translocation has long been suggested from physiological findings. Cytochalasin B, an alkaloid that prevents actin polymerization, inhibits RPCH–induced pigment aggre- gation in a variety of caridean shrimp chromatophores (Robison and Charton, 1973; Tuma et al., 1995; McNamara and Ribeiro, 1999). It also induces spontaneous pigment aggregation, the natural unstimulated state, in crab chromatophores (Lambert and Crowe, 1976).

4.2. Molecular motors and pigment granule translocation

4.2.1. Myosin

Confocal immunofluorescence microscopy reveals clearly non-muscle myosin II in the ovarian chromatophores of M. olfersi, mainly associated with pigment granules but also with the cytoskeleton, since the antibody labels some areas lacking pigments. In epidermal and nerve cord chro- matophores of P. affinis (McNamara, 1980), P. northropi and M. olfersi (McNamara and Sesso, 1982), and in several fish chromatophores (Taylor, 1992), carotenoid pigment granules show structural continuity with large networks of smooth endoplasmic reticulum cisternae. The pig- ment mass forms a single structure that exhibits elastic properties (McNamara and Ribeiro, 1999) and behaves like an intracellular spring matrix moved by molecular motors (Boyle and McNamara, 2008). Fur- ther, non-muscle myosin II lacks the traditional cargo-binding tail sites (Rosenfeld et al., 2003) and appears to cause contraction of actin net- works in various cell types (Sandquist and Swenson, 2006).

Since some non-muscle myosin II may not be directly associated with the pigment granules themselves, pigment movement might occur in part through endoplasmic reticulum contractility (McNamara, 1980; Taylor, 1992) and not involve the direct binding of this motor to a granule. Alternatively, non-muscle myosin II may express a tail segment homologous to cargo receptors, which might enable cargo binding (DePina et al., 2007).

The inhibition of non-muscle myosin II by blebbistatin strongly inhibits pigment aggregation and reduces translocation velocity (Table 1), inducing spontaneous dispersion even in the presence of RPCH, revealing this particular myosin class to be present and functionally involved in pigment translocation. Inhibition of non- muscle myosin II by blebbistatin also partially inhibits melanosome aggregation in fish retinal pigment epithelium (Barsoum and King- Smith, 2007).

Further, non-muscle myosin II may not be the only aggregation motor involved, since inhibition by blebbistatin is only partial. Other motors may play a part, especially in rapid phase aggregation, which is unaffected when non-muscle myosin II is inhibited. Our confocal mi- croscopy analysis of labeling with an anti-skeletal and smooth muscle myosin antibody reveals a pan-myosin located in areas occupied by pig- ment granules. Using the same antibody, Boyle and McNamara (2006) demonstrated a stable association between myosin and red pigment granules in M. olfersi chromatophores and showed functional inhibition of RPCH-triggered pigment aggregation in vitro. Inhibition of myosin ATPase by butanedione monoxime (McNamara and Ribeiro, 1999) also suggests a role for myosins in driving pigment aggregation.

Interestingly, protein kinase G (PKG) may activate molecular motors or their associated protein regulators during pigment aggregation in
M. olfersi ovarian chromatophores since PKG inhibition produces an aggregation profile very similar to non-muscle myosin II inhibition (see Milograna et al., 2012, Fig. 4).

4.2.2. Dynein

Dynein is also associated intimately with the pigment granules as re- vealed here by confocal immunofluorescence microscopy in intact
M. olfersi ovarian chromatophores. Western blotting for dynein in M. olfersi chromatophores has been inconclusive: a monoclonal, anti- dynein intermediate chain antibody revealed a single, well-defined band at 150 kDa (Ribeiro and McNamara, 2000), heavier than tradition- al 74-kDa dynein intermediate chains (Hirokawa, 1998; Rogers et al., 1999). EHNA, an inhibitor of dynein ATPase induces 30% pigment aggregation in our preparations, probably attributable to increasing cGMP titers, since EHNA also inhibits PDE2, a cGMP-stimulated phos- phodiesterase (Podzuweit et al., 1995). EHNA also inhibits RPCH- induced pigment aggregation by 20%, significantly reducing fast phase aggregation (Table 1), which suggests dynein to be one of the aggrega- tion motors other than myosins. Dynein inhibition also abolishes completely pigment re-dispersion, an effect likely related to the elevat- ed cytosolic cGMP titers maintained while PDE2 is inhibited. Curiously, sodium metavanadate, also a dynein-ATPase inhibitor, does not affect RPCH-triggered aggregation (McNamara and Ribeiro, 1999), which is different from its physiological effect in fish chromatophores (Beckerle and Porter, 1982; Rodionov and Borisy, 1997; Oshima, 2001).

This investigation reveals that pigment granule translocation depends on both microfilament and microtubule polymers associated with subunit turnover, and on different molecular motors. Dynein may sustain microtubule-based translocation during initial rapid phase aggregation. Subsequent slow phase aggregation likely depends on non-muscle myosin II bound to the granules, and translocation along actin microfilaments. Alternatively, non-muscle myosin II may act via the actin-mediated contraction of the pigment mass as a whole, producing a spring-like effect. Distinct molecular motors appear to be associated with the cytoskeleton, smooth endoplasmic reticulum and pigment granules, providing rapid granule translocation in response to neuropeptide signaling. Such mechanisms in crustacean chromatophores share various features with those of vertebrate pigment cells.