- Email: [email protected]
Selective stabilization of microtubules within the proximal region of developing axonal neurites
Brain Research Bulletin, Vol. 48, No. 3, pp. 255–261, 1999 Copyright © 1999 Elsevier Science Inc. Printed in the USA. All rights reserved 0361-9230/99/$–see front matter
PII S0361-9230(98)00019-7
Selective stabilization of microtubules within the proximal region of developing axonal neurites Thomas B. Shea* Center for Cellular Neurobiology and Neurodegeneration Research, Department of Biological Sciences, University of Massachusetts Lowell, Lowell, MA, USA [Received 30 June 1997; Revised 12 December 1997; Accepted 19 January 1998] ABSTRACT: This study examined the distribution of labile and stable microtubules (MTs) during axonal neurite elaboration in NB2a/d1 cells using immunocytochemical markers of unmodified (tyrosinated; Tyr), modified (detyrosinated [Glu] and acetylated [Acet]) and total tubulin. Prominent total and Tyr tubulin immunoreactivity was relatively evenly distributed throughout axonal neurites. By contrast, Acet or Glu immunoreactivity was relatively concentrated within the proximal region of the neurite. Ultrastructural analyses demonstrated an array of longitudinal MTs that apparently span the entire neurite length. The observed differential localization of modified tubulin subunits in axonal neurites of these cells may therefore derive from selective stabilization of proximal regions of full-length axonal MTs. This was substantiated by the observation of Acet immunoreactivity on 30 –50% of MTs within the most proximal axonal region, along with a proximal-distal decline to ≤5% of Acetimmunoreactive MTs, in immunoelectron microscopy (immunoEM) analyses. Microinjected biotinylated subunits were initially detected in assembled form within soma and proximal neurites, indicative of ongoing tubulin subunit incorporation into MTs within, and/or MT translocation into, proximal neurites. Because acetylation and detyrosination are functions of MT age, their concentration in this region despite deposition and/or transport of biotinylated tubulin suggests that a subset of axonal MTs undergoes subunit turnover and/or translocation at rates vastly slower than that of the majority of axonal MTs. Selective stabilization of the proximal region of a subset of axonal MTs may serve to construct a relatively stationary scaffold against which other axonal elements could translocate to more distal axonal regions for continued axonal outgrowth. © 1999 Elsevier Science Inc.
stability, these modifications do not themselves confer stability [16]. The methods by which neurons generate MT populations differing in stability, as well as overall methodologies by which tubulin undergoes axonal transport (i.e., as monomer or polymer or both), remain the subject of controversy (e.g., [7,15,21,23–25,32]; for reviews, see [2,9,14,17,18,22]). By interacting with each other and with other cytoskeletal elements, MTs provide structural support for the growing axon (e.g., [26] and references therein). In addition, stable MTs may also act as a scaffold against which relatively labile MTs may translocate during continued axonal elongation [24]. If this is indeed the case, one would expect to find a population of selectively stabilized MTs within at least the proximal portion of axonal neurites, which may or may not extend along most or all of the axonal length. In the present study, therefore, the distribution of modified and unmodified MTs during axonal neurite outgrowth was examined. NB2a/d1 neuroblastoma cells were utilized in this study, as they have provided a useful and unique model system for examining, via an array of biochemical, ultrastructural and morphological methodologies, the events associated with the transition of a mitotic neuroblast into a postmitotic, polarized neuronal cell. These events include the sequential deposition of unmodified MTs and neurofilaments, followed by modification of these cytoskeletal elements (by Acet/Glu and extensive phosphorylation, respectively) in a manner that spatially and temporally resembles that described for neurons in situ and in culture (e.g., [27–29,31] and references therein). It is demonstrated herein that selective stabilization of a subset of axonal MTs accompanies axonal elaboration, and it is suggested that this subset may play a structural role as a “scaffold” to facilitate translocation of additional cytoskeletal elements during continued axonal elongation.
KEY WORDS: Microtubules, Acetylation, Detyrosination, Cytoskeleton, Axon, Neuronal differentiation.
MATERIALS AND METHODS
INTRODUCTION
Cells and Culture Conditions
Axonal outgrowth is accompanied by substantial deposition of microtubules (MTs) [10,13]. Microtubule populations of varying stability exist within axonal neurites [6,33]. Stabilized MTs are denoted by the posttranslational tubulin modifications of acetylation (Acet) and detyrosination (Glu) [4,8], which accumulate as a function of MT age [3,11,16]. Notably, while the degree of MT Acet and Glu provides a reliable index of MT age and relative
NB2a/d1 cells [27,29,31] were plated in Dulbecco’s Modified Eagle Medium containing 10% horse serum and 25 mg/ml gentamycin (Gibco, Grand Island, NY, USA) in either 10-cm2 plates or in multiwell chamber slides (Lab-Tek, Naperville, MD, USA) in a humidified atmosphere of 95% air and 5% CO2. Outgrowth of neurites demonstrating multiple characteristics of axons was in-
* Address for correspondence: Thomas B. Shea, Center for Cellular Neurobiology and Neurodegeneration Research, Department of Biological Sciences, University of Massachusetts Lowell, One University Avenue, Lowell, MA 01854, USA. Fax: 508-934-3044; E-mail: [email protected]
255
256
SHEA
duced by 7 continuous days of treatment with 1 mM dibutyryl adenosine 39,59-cyclic monophosphate (dbcAMP; Sigma Chemical Co., St. Louis, MO, USA) beginning 24 h after plating as described [27–29,31]. Medium was routinely changed every 3 days, with the inclusion of dbcAMP where appropriate. In some experiments, cultures were treated with 1026 M colchicine or 10 mM taxol for up to 2 h.
rescence intensity was approximately equivalent along the entire neurite length, or “Distal . Proximal” or “Distal , Proximal” when one half exhibited more intense immunostaining than the other. Neurites were scored in this manner rather than by the use of a multisegmented “mask” approach [5], because this simplified approach facilitated quantitation of relative MT isoform distribution between proximal and distal neurite halves.
Fixation and Extraction
Electron Microscopy
Unassembled tubulin was extracted from cells under conditions that stabilize existing MTs according to the procedure of Brown et al. [7] with minor modifications. Cells were rinsed with phosphate buffered saline (PBS) then extracted for 5 min in 60 mM PIPES (pH 6.9) containing 10 mM EGTA, 2 mM MgCl2, 1% saponin and 10 mM taxol. As described previously [7], substitution of 1 mM CaCl2 for EGTA and elimination of taxol depleted virtually all tubulin staining due to MT depolymerization (not shown). In some experiments, cultures were fixed with 4% paraformaldehyde for 15 min at room temperature without extraction.
To monitor MT length along axonal neurites, additional cultures were fixed for electron microscopic analyses according to Baas and Black [4] and subjected to serial sectioning parallel to the culture surface. Serial sections presenting full-length or near fulllength axonal profiles were utilized in further analyses. For immunoelectron microscopy (immuno-EM), cells were fixed according to Baas and Black [4], reacted with 6-11B-1 as described above, followed by goat anti-mouse antibody conjugated to colloidal gold [30]. Immuno-EM localization and quantification of Acet MTs was carried out on collages of prints at 10,000 –20,0003 spanning the perikaryon and at least the first 20 mM of the axon from three separate cells. These micrographs were scored at 1-mM intervals perpendicular to the axon, and the percentage of labeled MTs in each 1-mM segment was then calculated by determining the Number of Labeled MTs/Number of Total MTs 3 100. Additional micrographs including the final 15 mM of axonal neurites adjacent to growth cones, and from random areas within central axonal neurite regions, were similarly examined and quantified. As an index of antibody specificity, it was noted that neurofilaments were never decorated by colloidal gold in these analyses, whereas neurofilaments were readily labeled in such analyses using a neurofilament-directed primary antibody in previous analyses [30].
Immunofluorescent Microscopy Cells were processed for immunofluorescence according to Brown et al. [7]. Cells were fixed by immersion in 220°C methanol for 6 min, followed by rehydration in PBS and blocking for 10 min with 4% normal goat serum in PBS. Cells were simultaneously double labeled by incubation for 45 min at 37°C with a mixture of a rat monoclonal antibody (Y/L12, 1:100 dilution; Jackson Immunochemicals) directed against tyrosinated (Tyr) a-tubulin and mouse monoclonal antibodies directed against either all forms of a-tubulin (1:1000 dilution in PBS; Sigma) or Acet a-tubulin (6-11B-1; 1:20 dilution; Sigma), then incubated for an additional 45 min at 37°C with a mixture of fluorescein-conjugated goat anti-rat antibody and Texas red-conjugated goat anti-mouse antibody; each of these secondary antibodies had been preadsorbed by the manufacturer against the heterologous rodent IgG to eliminate potential cross-reactivity (Jackson Immunochemicals). Additional cultures were singly labeled with the above antibodies and with rabbit polyclonal antiserum (“Glu”; which reacts specifically with Glu a-tubulin) [11] followed by goat anti-rabbit secondary antibody. Cells were mounted in a 1:1 mixture of PBS and glycerol and stored in the dark at 4°C. Elimination of primary antibody resulted in no reaction above background.
Confocal Microscopy Additional cultures processed for Acet or Tyr immunoreactivity were subjected to digital confocal microscopy under epifluorescent optics using a video camera mounted on a Zeiss inverted microscope; the camera was interfaced with a Z-axis drive stage controller operated by Oncor Imaging software (Gaithersburg, MD, USA) in a Macintosh Quadra-950 computer [34]. Central sections obtained in a Z-scale through-focus series that transected the entire axonal length were subjected to densitometric analyses as described above.
Microdensitometric Analyses
Nomenclature
To quantify the relative distribution along axonal neurites of MTs containing or lacking various a-tubulin isoforms, between 50 and 100 randomly selected cells in multiple microscopic fields from duplicate cultures reacted with each antibody were examined under ultraviolet and phase-contrast optics. Images were captured via a Dage cooled CCD camera operated by a Scion LG-3 frame grabber in a Macintosh PowerPC 7100AV controlled by NIH Image (1.57). The length of each neurite shaft was determined using the software’s “neurite labeling macro” using the phasecontrast image, and the neurite was divided into halves. The relative intensity of tubulin immunofluorescence was determined by outlining each half of the axonal neurite shaft with the program’s freehand selection tool and measuring the integrated fluorescent intensity. The on-screen outline of the shaft was shifted to an adjacent area devoid of cells or debris, and the “background” intensity was measured and subtracted from each axonal neurite segment to generate a corrected fluorescence intensity. For quantitative analyses of tubulin isoform distribution, neurites were scored as “Distal 5 Proximal” where the corrected tubulin fluo-
According to the convention of previous studies [4], references are made herein to “Acet” (acetylated), “Tyr” (tyrosinated)” or “Glu” (detyrosinated) MTs solely to simplify the description of results, and they only denote the presence of MTs exhibiting prominent immunoreactivity towards one or more of the specific antibodies utilized in this study; this terminology is not meant to imply that the total or even the majority of a-tubulin subunits in individual MTs, or regions of MTs, have undergone such a posttranslational modification. Indeed, it has been shown that up to 40% of the individual MTs in axons of sympathetic neurons contain both stable (modified) and labile (unmodified) domains [4] and that most cellular domains contain variously modified MTs [1] (see also Fig. 6 herein). Moreover, lack of Acet immunoreactivity in our analyses cannot be interpreted to indicate the complete absence of this modification, but rather only denotes relative levels of acetylation; this consideration is particularly important in interpretation of immuno-EM analyses, in which thin sections are viewed rather than the entire axon as in immunofluorescent analyses. Similarly, references to “modified” and “unmodified” MTs
STABILIZATION OF AXONAL MICROTUBULES
FIG. 1. Acet immunoreactivity is concentrated within the proximal half of axonal neurites. Panels present individual cells doubly labeled for immunofluorescent analyses of total and Acet immunoreactivity (upper two panels) or total and Tyr immunoreactivity (lower two panels) as indicated. Note the more even distribution of Tyr and total microtubule immunofluorescence along the entire neurite length versus the relative concentration of Acet immunofluorescence within the proximal neurite half.
are intended only to indicate the presence or absence of Glu and/or Acet a-tubulin subunits and do not address the presence or absence of any other undisclosed modifications. RESULTS The distribution of total and modified tubulin was examined in NB2a/d1 cells that had been treated with dbcAMP for 7 days. Total tubulin immunoreactivity was prominent throughout the perikaryon and axonal neurite shaft and was also present within the growth cone at an apparently reduced intensity relative to the rest of the cell (Fig. 1). Tyrosine immunoreactivity presented a profile similar to that of total tubulin immunoreactivity, indicating the presence of Tyr MTs throughout the perikaryal and axonal MT array (Fig. 1). Acet MTs were also detected within perikarya and axonal neurites but were more concentrated within the proximal than the distal region of axonal neurites (Fig. 1). Confocal analyses of central regions of axonal neurites provided a further indication of selective localization of Acet immunoreactivity within the proximal region of axonal neurites (Fig. 2). Detyrosinated MTs presented a distribution similar to that of Acet, but not Tyr, MTs (Fig. 3). Densitometric analyses confirmed the above visual interpretations. Total MTs were relatively evenly distributed within the proximal and distal halves of the majority of axonal neurites, suggesting the presence of similar numbers of MTs along the length of most neurites (Figs. 1, 2, and 4). Although this distribution was mirrored by that of Tyr MTs, Acet and Glu MTs were instead selectively localized within the proximal neurite half (Figs. 1, 2, and 4). This substantial concentration of Acet and Glu MTs within the proximal portion of the neurite was not reflected within the total MT distribution, because no corresponding increase in total MTs was observed within the proximal versus distal neurite half (Fig. 4). This disparity suggested that (1) either the stable MT population(s) represented a subpopulation of total MTs and therefore did not influence the distribution of total tubulin immunoreactivity or (2) that these posttranslational modifications were regionally concentrated within the proximal portion of MTs that spanned the full length of axonal neurites. To resolve this issue, we carried out ultrastructural examinations of axonal MTs. Immuno-EM confirmed that Acet immunoreactivity was indeed
257
FIG. 2. Confocal microscopic analysis of Acet immunoreactivity. Panels present cells doubly labeled for total and Acet immunofluorescence as indicated and subjected to confocal analysis as described in Materials and Methods. Upper panels present gray-scale images of the central aspect of the axonal length; lower panels present the same images with intensified contrast generated by removal of gray-scale and retention of only white pixels. The small arrow denotes the growth cone region, whereas the larger arrowhead denotes a portion of the perikaryon. Note that confocal analyses further denote the apparent relative concentration of Acet immunoreactivity within the proximal neurite half.
localized to MTs (Fig. 5A). A longitudinally oriented array of parallel MTs was observed along the length of axonal neurites (Fig. 5B). Because most axonal neurites did not lie precisely within the plane of section along their entire length, we sectioned the entire length of selected axonal neurites and assembled serial section reconstructions. Following examination of serial sections of the entire length of three such neurites, only a single definitive MT end was located within the axonal shaft itself of one out of all three neurites (not shown), whereas all other MT profiles terminating within a given section appeared within the subsequent section (e.g., Figs. 5C–E). These analyses indicate that differentiated NB2a/d1 cells elaborate MTs that span the entire axonal length. Immuno-EM analyses demonstrated that approximately 30 –50% of MTs within the most proximal axonal region were Acet and that this percentage declined dramatically within the first 20 mM of the axonal shaft (Fig. 5F). Low levels (e.g., 0 –5%) of
FIG. 3. Detyrosinated immunoreactivity is concentrated with proximal neurites. Cells were singly labeled for total, Tyr, Acet or Glu immunoreactivity. The arrowheads denote the position of the growth cone for each cell. Note that total and Tyr immunofluorescence are more evenly distributed along the neurite length than are Acet and Glu immunofluorescence.
258
FIG. 4. Relative distribution of modified, unmodified and total tubulin immunoreactivity in developing axonal neurites. Between 50 and 100 randomly selected cells in multiple microscopic fields from duplicate cultures reacted with each antibody were examined under phase-contrast and immunofluorescent optics. Neurites were bisected into equal proximal and distal halves according to their respective phase-contrast images using the “Neurite Length Macro” of NIH Image (see Materials and Methods). Neurites were then scored according to their relative intensity of immunostaining as “Distal . Proximal,” “Distal 5 Proximal” or “Distal , Proximal” as determined under immunofluorescent optics. Note that at all measured times during outgrowth, the distribution of Tyr immunoreactivity roughly parallels that of total immunoreactivity. By contrast, Glu and Acet immunoreactivity are either evenly distributed along the neurite length or, unlike Tyr immunoreactivity, are concentrated within the proximal half of the neurite.
SHEA turnover or preferential subunit Acet within proximal versus distal neurite regions. Following treatment with the MT-stabilizing drug, taxol, the entire MT length was Acet (Fig. 7). Acetylated MTs furthermore accumulated within the growth cone, a region from which they were typically absent prior to taxol treatment (e.g., Fig. 1). These data indicate that there is no inherent inability for Acet MTs to accumulate along the entire neurite length and therefore indicate that selective Acet of MTs occurs within the proximal neurite region. Regional tubulin incorporation and/or MT translocation was further investigated by microinjection of biotinylated tubulin. Biotinylated tubulin associated with MTs within minutes after microinjection (Fig. 8). Consistent with the presence of labile (Tyr) MTs within the soma and proximal neurites, biotinylated subunits were initially detected within MTs within these regions. These analyses were unable to distinguish, and were not intended to distinguish, whether the appearance of MT-associated biotinylated subunits within proximal neurites was a function of localized assembly, transport of newly assembled perikaryal MTs into axonal neurites or both mechanisms. Nevertheless, these analyses indicate ongoing subunit incorporation within, and/or MT translocation into, proximal neurites. Because Acet is a function of MT age [3,11,16], one interpretation of the persistence of a concentration of Acet MTs within proximal neurites, despite localized subunit exchange and/or transport, is that the Acet MTs represent a subset of axonal MTs which either do not undergo subunit turnover or translocation or do so at rates vastly slower than those of the overall axonal MT population. Longer time points were not conclusive, because biotinylated tubulin rapidly dispersed throughout, and associated with MTs throughout, the entire neurite (not shown; e.g., see [20]). DISCUSSION
Acet-reactive MTs observed in immuno-EM analyses of segments 16 –20 were also detected within the final 15 mM of the axonal shaft up to the growth cone (Fig. 5F), as well as within random sections from “central” axonal regions (i.e., sections in which neither the perikaryon nor growth cone could be detected within a 15 to 20-mM radius; not shown). These data indicate that the proximal region of a subpopulation of the total axonal MTs is selectively acetylated. Acetylation and Glu are considered as markers for stabilized MTs [4]. To determine whether or not selective accumulation of Acet and Glu immunoreactivity within proximal neurites corresponded with selective MT stabilization in this region, the relative stability of MTs within axonal regions was probed by treatment with colchicine. Colchicine treatment substantially depleted MTs from both perikarya and neurites (Fig. 6). Axonal MTs survived colchicine treatment longer than perikaryal MTs, suggesting that they were preferentially stabilized as compared to perikaryal MTs. Continued colchicine treatment depleted distal axonal MTs and reduced proximal axonal MTs. The last surviving MTs were a subset of the MTs located within the proximal neurite; these are designated a “subset” because an overall reduction in all modified and unmodified MTs accompanied extended colchicine treatment (Fig. 6). The only Tyr immunoreactivity surviving colchicine treatment for 120 min precisely colocalized with Acet immunoreactivity, indicating that these relatively colchicine-resistant MTs (or regions of MTs) contained both Tyr and Acet subunits. These findings indicate that Acet immunoreactivity indeed denotes selectively stabilized MTs (or regions of MTs). Next, a different approach was used to address whether the selective stabilization of MTs/regions of MTs within the proximal half of axonal neurites was a consequence of differential subunit
The selective accumulation within NB2a/d1 axonal neurites of Acet and Glu MTs has been previously documented [27]. In the present study, differential distribution of these modifications between proximal and distal axonal neurite regions was demonstrated. These findings are consistent with the previous demonstration of selective stabilization of MTs within proximal axonal regions [3]. In contrast to selective concentration of Acet and Glu MTs within proximal neurites, most neurites contained similar amounts of total and Tyr MTs along their entire length. A minority of neurites displayed somewhat more Tyr and total MTs in the proximal half as did those displaying more Tyr and total MTs in the distal half, consistent with the demonstration that the addition of tubulin subunits to axonal MTs occurs at both the perikaryal and growth cone end of neurites [7]. These findings collectively demonstrate the existence of multiple MT populations within the developing axonal neurites of dbcAMP-treated NB2a/d1 cells. Information regarding the relative stability of axonal and perikaryal MT populations was provided by colchicine treatment. Microtubules displaying Acet immunoreactivity persisted the longest during colchicine treatment, indicating that MTs possessing this modification were selectively stabilized. Moreover, proximal neuritic Acet MT immunoreactivity persisted beyond that of perikaryal Acet MT immunoreactivity, despite the presence of significantly greater initial Acet immunoreactivity within many neuronal perikarya, suggesting that Acet MTs (or regions of MTs) within proximal axonal neurites are relatively more stable than those within perikarya. Although the vast majority of Tyr immunoreactivity was rapidly depleted by colchicine, the ultimate colocalization of Tyr and Acet immunoreactivity within the proximal neurite suggested that stable MTs contained both Acet and Tyr subunits (e.g., see also [1,4]). The notion that differentiating
STABILIZATION OF AXONAL MICROTUBULES
259
FIG. 6. Colchicine treatment differentially affects perikaryal and axonal, and Acet versus Tyr, immunoreactivity. Images present cultures following 0 –120 min in the presence of colchicine. At each time point, the identical field is presented as visualized under fluorescein and rhodamine optics to depict Tyr or Acet MTs, respectively. Colchicine treatment resulted in rapid, substantial depletion of MTs from both perikarya and neurites. Axonal MTs survived colchicine treatment longer than perikaryal MTs (e.g., compare 60-min colchicine treatment to 0 min), suggesting that they were preferentially stabilized as compared to perikaryal MTs (arrow at 60 min denotes most proximal end of axonal neurite). Continued colchicine treatment (e.g., 120 min) depleted distal axonal MTs and reduced proximal axonal MTs. Note that the residual Tyr immunoreactivity remaining after 120 min of colchicine treatment corresponds precisely with Acet immunoreactivity within proximal neurites (arrows).
NB2a/d1 cells selectively regulate MT stability was confirmed by the rapid accumulation of MTs along the entire neurite length following taxol-induced MT stabilization. Our analyses did not address the site(s) of MT nucleation. However, the length of axonal neurites (approximately 100 mM)
FIG. 5. Ultrastructural analyses of tubulin immunoreactivity and microtubule (MT) distribution in axonal neurites. Immunoelectron microscopy with 6-11B-1 confirmed that Acet immunoreactivity under these conditions is associated with MTs (panel A); in this representative micrograph, 15/15 colloidal gold particles are on or adjacent to MTs. A densely packed longitudinally oriented array of parallel MTs was observed along the length of axonal neurites (panel B). Additional panels (C–E) depict representative serial reconstructions of portions of one axonal neurite. Only a single definitive MT end was located within any of the axonal shafts examined (not shown; see text). All other MT profiles apparently terminating within a given section appeared within an adjacent section; arrows depict this in panels C–E, which present reconstructions from three adjacent sets of longitudinal sections. These analyses indicate that, under these conditions of differentiation, NB2a/d1 cells elaborate MTs that span the entire axonal length. The accompanying graph (panel F) presents quantification of collodial-gold-labeled MTs within axonal neurites. Micrographs from three separate cells subjected to immunoelectron microscopy with 6-11B-1 containing longitudinal axonal profiles were assembled in a collage that contained a portion of the perikaryon and the initial 20 mM of proximal axon, and scored at 1-mM intervals. Similar quantitation was carried out on collages that spanned the final 15 mM of the axonal neurite and contained at least a portion of the growth cone. The percentage of Acet MTs in each 1-mM segment was then calculated by determining the Number of Labeled MTs/Number of Total MTs 3 100. The graph presents the mean 6 standard deviation.
FIG. 7. Taxol-induced microtubule (MT) stabilization fosters accumulation of Acet immunoreactivity. Panels present distal portions of neurites from taxol-treated and untreated cells doubly labeled for total and Acet immunoreactivity as indicated. Only the distal portions of neurites are presented. Total tubulin immunoreactivity was prominent along the distal shaft, whereas Acet immunoreactivity was barely detectable. Mild total tubulin immunoreactivity, but no Acet immunoreactivity, was also present within the growth cone (arrows). Following taxol treatment for 2 h, however, Acet immunoreactivity was prominent within the distal neurite, and both Acet and total tubulin immunoreactivity were also prominent within the growth cone.
260
FIG. 8. Biotinylated subunits were initially detected within microtubules (MTs) within soma and proximal neurites. Cells were injected with biotinylated tubulin (Bt-tub). Ten minutes after injection, cells were fixed either with paraformaldehyde without extraction (which reveals both assembled and unassembled subunits) or with methanol following saponin extraction (which, as described previously, depletes unassembled subunits but retains subunits that have assembled into MTs). Comparison of unextracted and extracted cells indicates that substantial biotinylated subunits were depleted at this early time point. In addition to biotinylated immunoreactivity within perikarya, immunoreactivity within unextracted and extracted neurites was concentrated within the proximal half; arrows in phase image denote the entire neurite length.
following 7 days of treatment with dbcAMP vastly exceeds that of the perikaryon, indicating, as previously suggested [14], that MTs must undergo elongation within the axonal neurite. The initial accumulation of biotinylated tubulin within proximal neurites following its injection into the cell body is consistent both with the possibility of subunit addition to the proximal end of axonal MTs and the observation in this study of labile MTs throughout the neurite. However, a concentration of stabilized MT markers on a subset of MTs within the proximal neurite was observed in this study. Because the accumulation of Acet and Glu subunits within MTs is time dependent [3,11,16], subunits may not be added to the proximal end of this particular Acet- and Glu-rich MT subset. Rather, they are consistent with the exclusive addition of subunits to the distal end of this particular subset and/or suggest that this MT subset undergoes exceptionally slow, if any, turnover and its proximal end is preserved. These data are furthermore inconsistent with models suggesting that tubulin subunits are added exclusively to MTs within the proximal neurite region and that the MT array translocates en bloc along the axonal neurite length; such a model would necessitate that the oldest MTs or MT regions (and therefore those containing the highest proportion of Acet and Glu immunoreactivity) would be detected within distal neurite regions. These data are inconsistent with the structural hypothesis of axonal transport as originally proposed, which held that MTs were transported as a highly cross-linked network [17]. However, they are fully consistent with the subsequent refinement of the structural hypothesis, which considers that individual MTs may translocate independently of one another [18,19], provided that allowances are made for some MTs/MT populations to translocate substantially more slowly than others; this latter possibility allows for the observed preferential accumulation of age-related tubulin modifications within the proximal neurite region. A recent study of transport of radiolabeled tubulin [12] supports this latter possibility. In these studies, newly synthesized tubulin was conveyed along sensory axons following radiolabeling of developing dorsal root ganglia in situ. However, some newly synthesized tubulin accumulated along developing axons substantially after the passage of the transported wave, suggesting that a portion of tubulin may contribute to the development of a relatively stationary MT population in growing axons. Selective stabilization of certain MTs could provide a nidus for MT elongation during axon out-
SHEA growth [20]. In addition, selective stabilization of even a subset of MTs within the proximal axonal region represents a potential mechanism to provide structural support for the oldest region of the developing neurite, yet it still permits relative plasticity within the distal region, which is associated with continued growthrelated remodeling. This possibility is consistent with the delay in accumulation of Acet and Glu MT immunoreactivity within dbcAMP-induced axonal neurites until 3 days after treatment [27]. Moreover, the existence of a stationary population of MTs has been postulated to serve as a scaffold against which moving MTs derive motive force [24]. It was further noted that any such putative stationary MT population must be relatively small, because it is not reflected in total MT distribution [24]. The finding that the accumulation of Acet and Glu immunoreactivity is not reflected in the distribution of total MT immunoreactivity in the present study concurs with this latter suggestion of Reinsch et al. [24]. Previous studies have demonstrated that NB2a/d1 axonal neurites attain a degree of stability derived from interactions of axonal MTs and neurofilaments and, furthermore, that these interactions are mediated in part by the MT-associated proteins MAP1B and tau [26]. It remains possible that those MTs exhibiting strong Acet and Glu immunoreactivity within proximal neurites represent those MTs that have undergone relatively long-lasting MT-MT or MT-neurofilament interactions and that the proximal domains of these MTs have accumulated the age-related modifications of Acet and Glu as a consequence of a resultant delay in turnover. In consideration of these and previous findings, it is suggested that selective stabilization of MTs within the proximal region of developing axons may provide a relatively stationary scaffold against which moving MTs and/or other membranous and cytoskeletal organelles translocate to more distal axonal regions for continued axonal outgrowth. ACKNOWLEDGEMENTS
The author thanks Peter Paskevich for assistance with electron microscopy, Drs. Steven Vincent and Francine Benes for assistance with confocal microscopy, Ms. Mary Lou Beermann for expert technical assistance, and Dr. Peter Baas for his generous gift of biotinylated tubulin and his continued interest and advice. This research was supported by the NIMH.
REFERENCES 1. Ahmad, F. J.; Pienkowski, T. P.; Baas, P. W. Regional differences in microtubule dynamics in the axon. J. Neurosci. 13:856 – 866; 1993. 2. Baas, P. W. Microtubules and axonal growth. Curr. Op. Cell Biol. 9:29 –36; 1997. 3. Baas, P. W.; Ahmad, F. J.; Pienkowski, T. P.; Brown, A.; Black, M. M. Sites of microtubule stabilization for the axon. J. Neurosci. 13:2177– 2185; 1993. 4. Baas, P. W.; Black, M. M. Individual microtubules in the axon consist of domains that differ in both composition and stability. J. Cell Biol. 111:495–509; 1990. 5. Black, M. M.; Slaughter, T.; Moshiach, S.; Obrocka, M.; Fischer, I. Tau is enriched on dynamic microtubules in the distal region of growing axons. J. Neurosci. 16:3601–3619; 1996. 6. Brady, S. T.; Tytell, M.; Lasek, R. J. Axonal transport and axonal tubulin: Biochemical evidence for cold-stability. J. Cell Biol. 98:847– 858; 1984. 7. Brown, A.; Slaugther, T.; Black, M. M. Newly-assembled microtubules are concentrated in the proximal and distal regions of growing axons. J. Cell Biol. 119:867– 882; 1992. 8. Cambray-Deakin, M. A.; Burgoyne, R. D. Posttranslational modifications of a-tubulin: Acetylated and detyrosinated forms in axons of rat cerebellum. J. Cell Biol. 104:1569 –1574; 1987.
STABILIZATION OF AXONAL MICROTUBULES 9. Cleveland, D. W.; Hoffman, P. H. Slow axonal transport models come full circle: Evidence that microtubule sliding mediates axon elongation and tubulin transport. Cell 67:453– 456; 1991. 10. Daniels, M. The role of microtubules in the growth and stabilization of nerve fibers. Ann. N. Y. Acad. Sci. 253:535–544; 1975. 11. Gundersen, G. G.; Khawaja, S.; Bulinski, J. Postpolymerization detyrosination of a-tubulin: A mechanism for subcellular differentiation of microtubules. J. Cell Biol. 105:251–264; 1987. 12. Hoffman, P. N. Pulse-labeled tubulin is retained along growing axons. Soc. Neurosci. Abstr. 22:1986; 1996. 13. Jacobs, J. R.; Stevens, J. K. Changes in the organization of the neuritic cytoskeleton during nerve growth factor-activated differentiation of PC12 cells: A serial electron microscopic study of the development and control of neurite shape. J. Cell Biol. 103:663– 673; 1986. 14. Joshi, H. C.; Baas, P. W. A new perspective on microtubules and axon growth. J. Cell Biol. 121:1191–1196; 1993. 15. Keith, C. H.; Blane, K. Site of tubulin polymerization in PC12 cells. J. Neurochem. 54:1258 –1268; 1990. 16. Khawaja, S.; Gundersen, G. G.; Bulinski, J. C. Enhanced stability of microtubules enriched in detyrosinated tubulin is not a direct function of detyrosination level. J. Cell Biol. 106:141–149; 1988. 17. Lasek, R. J. Translocation of the axonal cytoskeleton and axonal locomotion. Philos. Trans. R. Soc. Lond. B Biol. Sci. 299:313–327; 1982. 18. Lasek, R. J. Polymer sliding in axons. J. Cell Sci. 5(suppl. 5):161–179; 1986. 19. Lasek, R. J. Studying the intrinsic determinants of neuronal form and function. In: Lasek, R. J.; Black, M. M., eds. Intrinsic determinants of neuronal form and function. New York: Liss; 1988:1–58. 20. Li, Y.; Black, M. M. Microtubule assembly and turnover in growing axons. J. Neurosci. 16:531–544; 1996. 21. Miller, K. E.; Joshi, H. C. Tubulin transport in neurons. J. Cell Biol. 133:1355–1366; 1996. 22. Ochs, S. Axoplasmic transport and its relation to other nerve functions. New York: Wiley; 1982.
261 23. Okabe, S.; Hirokawa, N. Turnover of fluorescently labeled tubulin and actin in the axon. Nature (Lond.) 343:479 – 482; 1990. 24. Reinsch, S. S.; Mitchison, T. J.; Kirschner, M. W. Microtubule polymer assembly and transport during axonal elongation. J. Cell Biol. 115:365–379; 1991. 25. Sabry, J.; O’Connor, T. P.; Kirschner, M. W. Axonal transport of tubulin in T1 pioneer neurons in situ. Neuron 14:1247–1256; 1995. 26. Shea, T. B.; Beermann, M. L. Respective roles of neurofilaments, microtubules, MAP1B and tau in the outgrowth and stabilization of axonal neurites. Mol. Biol. Cell. 5:863– 875; 1994. 27. Shea, T. B.; Beermann, M. L.; Nixon, R. A. Post-translational modification of a-tubulin by acetylation and detyrosination in NB2a/d1 neuroblastoma cells. Dev. Brain Res. 51:195–204; 1990. 28. Shea, T. B.; Beermann, M. L.; Nixon, R. A.; Fischer, I. Microtubuleassociated protein tau is required for axonal neurite elaboration by neuroblastoma cells. J. Neurosci. Res. 32:363–374; 1992. 29. Shea, T. B.; Fischer, I.; Sapirstein, V. S. Effects of retinoic acid on growth and morphologic differentiation of mouse NB2a neuroblastoma cells in culture. Dev. Brain Res. 21:307–314; 1985. 30. Shea, T. B.; Paskevich, P. A.; Beermann, M. L. The protein phosphatase inhibitor okadaic acid increases axonal neurofilaments and neurite caliber, and decreases axonal microtubules in NB2a/d1 cells. J. Neurosci. Res. 35:507–521; 1993. 31. Shea, T. B.; Sihag, R. K.; Nixon, R. A. Neurofilament triplet proteins of NB2a/d1 neuroblastoma: Posttranslational modification and incorporation into the cytoskeleton during differentiation. Dev. Brain Res. 43:97–109; 1988. 32. Takeda, S.; Funakoshi, T.; Hirokawa, N. Tubulin dynamics in neuronal axons of living zebrafish embryos. Neuron 14:1257–1264; 1995. 33. Tashiro, T.; Kurokawa, M.; Komiya, Y. Two populations of axonally transported tubulin differentiated by their interactions with neurofilaments. J. Neurochem. 43:1220 –1225; 1984. 34. Vincent, S. L.; Khan, Y.; Benes, F. M. Cellular colocalization of dopamine D1 and D2 receptors in rat medial prefrontal cortex. Synapse 19:112–120; 1995.
PII S0361-9230(98)00019-7
Selective stabilization of microtubules within the proximal region of developing axonal neurites Thomas B. Shea* Center for Cellular Neurobiology and Neurodegeneration Research, Department of Biological Sciences, University of Massachusetts Lowell, Lowell, MA, USA [Received 30 June 1997; Revised 12 December 1997; Accepted 19 January 1998] ABSTRACT: This study examined the distribution of labile and stable microtubules (MTs) during axonal neurite elaboration in NB2a/d1 cells using immunocytochemical markers of unmodified (tyrosinated; Tyr), modified (detyrosinated [Glu] and acetylated [Acet]) and total tubulin. Prominent total and Tyr tubulin immunoreactivity was relatively evenly distributed throughout axonal neurites. By contrast, Acet or Glu immunoreactivity was relatively concentrated within the proximal region of the neurite. Ultrastructural analyses demonstrated an array of longitudinal MTs that apparently span the entire neurite length. The observed differential localization of modified tubulin subunits in axonal neurites of these cells may therefore derive from selective stabilization of proximal regions of full-length axonal MTs. This was substantiated by the observation of Acet immunoreactivity on 30 –50% of MTs within the most proximal axonal region, along with a proximal-distal decline to ≤5% of Acetimmunoreactive MTs, in immunoelectron microscopy (immunoEM) analyses. Microinjected biotinylated subunits were initially detected in assembled form within soma and proximal neurites, indicative of ongoing tubulin subunit incorporation into MTs within, and/or MT translocation into, proximal neurites. Because acetylation and detyrosination are functions of MT age, their concentration in this region despite deposition and/or transport of biotinylated tubulin suggests that a subset of axonal MTs undergoes subunit turnover and/or translocation at rates vastly slower than that of the majority of axonal MTs. Selective stabilization of the proximal region of a subset of axonal MTs may serve to construct a relatively stationary scaffold against which other axonal elements could translocate to more distal axonal regions for continued axonal outgrowth. © 1999 Elsevier Science Inc.
stability, these modifications do not themselves confer stability [16]. The methods by which neurons generate MT populations differing in stability, as well as overall methodologies by which tubulin undergoes axonal transport (i.e., as monomer or polymer or both), remain the subject of controversy (e.g., [7,15,21,23–25,32]; for reviews, see [2,9,14,17,18,22]). By interacting with each other and with other cytoskeletal elements, MTs provide structural support for the growing axon (e.g., [26] and references therein). In addition, stable MTs may also act as a scaffold against which relatively labile MTs may translocate during continued axonal elongation [24]. If this is indeed the case, one would expect to find a population of selectively stabilized MTs within at least the proximal portion of axonal neurites, which may or may not extend along most or all of the axonal length. In the present study, therefore, the distribution of modified and unmodified MTs during axonal neurite outgrowth was examined. NB2a/d1 neuroblastoma cells were utilized in this study, as they have provided a useful and unique model system for examining, via an array of biochemical, ultrastructural and morphological methodologies, the events associated with the transition of a mitotic neuroblast into a postmitotic, polarized neuronal cell. These events include the sequential deposition of unmodified MTs and neurofilaments, followed by modification of these cytoskeletal elements (by Acet/Glu and extensive phosphorylation, respectively) in a manner that spatially and temporally resembles that described for neurons in situ and in culture (e.g., [27–29,31] and references therein). It is demonstrated herein that selective stabilization of a subset of axonal MTs accompanies axonal elaboration, and it is suggested that this subset may play a structural role as a “scaffold” to facilitate translocation of additional cytoskeletal elements during continued axonal elongation.
KEY WORDS: Microtubules, Acetylation, Detyrosination, Cytoskeleton, Axon, Neuronal differentiation.
MATERIALS AND METHODS
INTRODUCTION
Cells and Culture Conditions
Axonal outgrowth is accompanied by substantial deposition of microtubules (MTs) [10,13]. Microtubule populations of varying stability exist within axonal neurites [6,33]. Stabilized MTs are denoted by the posttranslational tubulin modifications of acetylation (Acet) and detyrosination (Glu) [4,8], which accumulate as a function of MT age [3,11,16]. Notably, while the degree of MT Acet and Glu provides a reliable index of MT age and relative
NB2a/d1 cells [27,29,31] were plated in Dulbecco’s Modified Eagle Medium containing 10% horse serum and 25 mg/ml gentamycin (Gibco, Grand Island, NY, USA) in either 10-cm2 plates or in multiwell chamber slides (Lab-Tek, Naperville, MD, USA) in a humidified atmosphere of 95% air and 5% CO2. Outgrowth of neurites demonstrating multiple characteristics of axons was in-
* Address for correspondence: Thomas B. Shea, Center for Cellular Neurobiology and Neurodegeneration Research, Department of Biological Sciences, University of Massachusetts Lowell, One University Avenue, Lowell, MA 01854, USA. Fax: 508-934-3044; E-mail: [email protected]
255
256
SHEA
duced by 7 continuous days of treatment with 1 mM dibutyryl adenosine 39,59-cyclic monophosphate (dbcAMP; Sigma Chemical Co., St. Louis, MO, USA) beginning 24 h after plating as described [27–29,31]. Medium was routinely changed every 3 days, with the inclusion of dbcAMP where appropriate. In some experiments, cultures were treated with 1026 M colchicine or 10 mM taxol for up to 2 h.
rescence intensity was approximately equivalent along the entire neurite length, or “Distal . Proximal” or “Distal , Proximal” when one half exhibited more intense immunostaining than the other. Neurites were scored in this manner rather than by the use of a multisegmented “mask” approach [5], because this simplified approach facilitated quantitation of relative MT isoform distribution between proximal and distal neurite halves.
Fixation and Extraction
Electron Microscopy
Unassembled tubulin was extracted from cells under conditions that stabilize existing MTs according to the procedure of Brown et al. [7] with minor modifications. Cells were rinsed with phosphate buffered saline (PBS) then extracted for 5 min in 60 mM PIPES (pH 6.9) containing 10 mM EGTA, 2 mM MgCl2, 1% saponin and 10 mM taxol. As described previously [7], substitution of 1 mM CaCl2 for EGTA and elimination of taxol depleted virtually all tubulin staining due to MT depolymerization (not shown). In some experiments, cultures were fixed with 4% paraformaldehyde for 15 min at room temperature without extraction.
To monitor MT length along axonal neurites, additional cultures were fixed for electron microscopic analyses according to Baas and Black [4] and subjected to serial sectioning parallel to the culture surface. Serial sections presenting full-length or near fulllength axonal profiles were utilized in further analyses. For immunoelectron microscopy (immuno-EM), cells were fixed according to Baas and Black [4], reacted with 6-11B-1 as described above, followed by goat anti-mouse antibody conjugated to colloidal gold [30]. Immuno-EM localization and quantification of Acet MTs was carried out on collages of prints at 10,000 –20,0003 spanning the perikaryon and at least the first 20 mM of the axon from three separate cells. These micrographs were scored at 1-mM intervals perpendicular to the axon, and the percentage of labeled MTs in each 1-mM segment was then calculated by determining the Number of Labeled MTs/Number of Total MTs 3 100. Additional micrographs including the final 15 mM of axonal neurites adjacent to growth cones, and from random areas within central axonal neurite regions, were similarly examined and quantified. As an index of antibody specificity, it was noted that neurofilaments were never decorated by colloidal gold in these analyses, whereas neurofilaments were readily labeled in such analyses using a neurofilament-directed primary antibody in previous analyses [30].
Immunofluorescent Microscopy Cells were processed for immunofluorescence according to Brown et al. [7]. Cells were fixed by immersion in 220°C methanol for 6 min, followed by rehydration in PBS and blocking for 10 min with 4% normal goat serum in PBS. Cells were simultaneously double labeled by incubation for 45 min at 37°C with a mixture of a rat monoclonal antibody (Y/L12, 1:100 dilution; Jackson Immunochemicals) directed against tyrosinated (Tyr) a-tubulin and mouse monoclonal antibodies directed against either all forms of a-tubulin (1:1000 dilution in PBS; Sigma) or Acet a-tubulin (6-11B-1; 1:20 dilution; Sigma), then incubated for an additional 45 min at 37°C with a mixture of fluorescein-conjugated goat anti-rat antibody and Texas red-conjugated goat anti-mouse antibody; each of these secondary antibodies had been preadsorbed by the manufacturer against the heterologous rodent IgG to eliminate potential cross-reactivity (Jackson Immunochemicals). Additional cultures were singly labeled with the above antibodies and with rabbit polyclonal antiserum (“Glu”; which reacts specifically with Glu a-tubulin) [11] followed by goat anti-rabbit secondary antibody. Cells were mounted in a 1:1 mixture of PBS and glycerol and stored in the dark at 4°C. Elimination of primary antibody resulted in no reaction above background.
Confocal Microscopy Additional cultures processed for Acet or Tyr immunoreactivity were subjected to digital confocal microscopy under epifluorescent optics using a video camera mounted on a Zeiss inverted microscope; the camera was interfaced with a Z-axis drive stage controller operated by Oncor Imaging software (Gaithersburg, MD, USA) in a Macintosh Quadra-950 computer [34]. Central sections obtained in a Z-scale through-focus series that transected the entire axonal length were subjected to densitometric analyses as described above.
Microdensitometric Analyses
Nomenclature
To quantify the relative distribution along axonal neurites of MTs containing or lacking various a-tubulin isoforms, between 50 and 100 randomly selected cells in multiple microscopic fields from duplicate cultures reacted with each antibody were examined under ultraviolet and phase-contrast optics. Images were captured via a Dage cooled CCD camera operated by a Scion LG-3 frame grabber in a Macintosh PowerPC 7100AV controlled by NIH Image (1.57). The length of each neurite shaft was determined using the software’s “neurite labeling macro” using the phasecontrast image, and the neurite was divided into halves. The relative intensity of tubulin immunofluorescence was determined by outlining each half of the axonal neurite shaft with the program’s freehand selection tool and measuring the integrated fluorescent intensity. The on-screen outline of the shaft was shifted to an adjacent area devoid of cells or debris, and the “background” intensity was measured and subtracted from each axonal neurite segment to generate a corrected fluorescence intensity. For quantitative analyses of tubulin isoform distribution, neurites were scored as “Distal 5 Proximal” where the corrected tubulin fluo-
According to the convention of previous studies [4], references are made herein to “Acet” (acetylated), “Tyr” (tyrosinated)” or “Glu” (detyrosinated) MTs solely to simplify the description of results, and they only denote the presence of MTs exhibiting prominent immunoreactivity towards one or more of the specific antibodies utilized in this study; this terminology is not meant to imply that the total or even the majority of a-tubulin subunits in individual MTs, or regions of MTs, have undergone such a posttranslational modification. Indeed, it has been shown that up to 40% of the individual MTs in axons of sympathetic neurons contain both stable (modified) and labile (unmodified) domains [4] and that most cellular domains contain variously modified MTs [1] (see also Fig. 6 herein). Moreover, lack of Acet immunoreactivity in our analyses cannot be interpreted to indicate the complete absence of this modification, but rather only denotes relative levels of acetylation; this consideration is particularly important in interpretation of immuno-EM analyses, in which thin sections are viewed rather than the entire axon as in immunofluorescent analyses. Similarly, references to “modified” and “unmodified” MTs
STABILIZATION OF AXONAL MICROTUBULES
FIG. 1. Acet immunoreactivity is concentrated within the proximal half of axonal neurites. Panels present individual cells doubly labeled for immunofluorescent analyses of total and Acet immunoreactivity (upper two panels) or total and Tyr immunoreactivity (lower two panels) as indicated. Note the more even distribution of Tyr and total microtubule immunofluorescence along the entire neurite length versus the relative concentration of Acet immunofluorescence within the proximal neurite half.
are intended only to indicate the presence or absence of Glu and/or Acet a-tubulin subunits and do not address the presence or absence of any other undisclosed modifications. RESULTS The distribution of total and modified tubulin was examined in NB2a/d1 cells that had been treated with dbcAMP for 7 days. Total tubulin immunoreactivity was prominent throughout the perikaryon and axonal neurite shaft and was also present within the growth cone at an apparently reduced intensity relative to the rest of the cell (Fig. 1). Tyrosine immunoreactivity presented a profile similar to that of total tubulin immunoreactivity, indicating the presence of Tyr MTs throughout the perikaryal and axonal MT array (Fig. 1). Acet MTs were also detected within perikarya and axonal neurites but were more concentrated within the proximal than the distal region of axonal neurites (Fig. 1). Confocal analyses of central regions of axonal neurites provided a further indication of selective localization of Acet immunoreactivity within the proximal region of axonal neurites (Fig. 2). Detyrosinated MTs presented a distribution similar to that of Acet, but not Tyr, MTs (Fig. 3). Densitometric analyses confirmed the above visual interpretations. Total MTs were relatively evenly distributed within the proximal and distal halves of the majority of axonal neurites, suggesting the presence of similar numbers of MTs along the length of most neurites (Figs. 1, 2, and 4). Although this distribution was mirrored by that of Tyr MTs, Acet and Glu MTs were instead selectively localized within the proximal neurite half (Figs. 1, 2, and 4). This substantial concentration of Acet and Glu MTs within the proximal portion of the neurite was not reflected within the total MT distribution, because no corresponding increase in total MTs was observed within the proximal versus distal neurite half (Fig. 4). This disparity suggested that (1) either the stable MT population(s) represented a subpopulation of total MTs and therefore did not influence the distribution of total tubulin immunoreactivity or (2) that these posttranslational modifications were regionally concentrated within the proximal portion of MTs that spanned the full length of axonal neurites. To resolve this issue, we carried out ultrastructural examinations of axonal MTs. Immuno-EM confirmed that Acet immunoreactivity was indeed
257
FIG. 2. Confocal microscopic analysis of Acet immunoreactivity. Panels present cells doubly labeled for total and Acet immunofluorescence as indicated and subjected to confocal analysis as described in Materials and Methods. Upper panels present gray-scale images of the central aspect of the axonal length; lower panels present the same images with intensified contrast generated by removal of gray-scale and retention of only white pixels. The small arrow denotes the growth cone region, whereas the larger arrowhead denotes a portion of the perikaryon. Note that confocal analyses further denote the apparent relative concentration of Acet immunoreactivity within the proximal neurite half.
localized to MTs (Fig. 5A). A longitudinally oriented array of parallel MTs was observed along the length of axonal neurites (Fig. 5B). Because most axonal neurites did not lie precisely within the plane of section along their entire length, we sectioned the entire length of selected axonal neurites and assembled serial section reconstructions. Following examination of serial sections of the entire length of three such neurites, only a single definitive MT end was located within the axonal shaft itself of one out of all three neurites (not shown), whereas all other MT profiles terminating within a given section appeared within the subsequent section (e.g., Figs. 5C–E). These analyses indicate that differentiated NB2a/d1 cells elaborate MTs that span the entire axonal length. Immuno-EM analyses demonstrated that approximately 30 –50% of MTs within the most proximal axonal region were Acet and that this percentage declined dramatically within the first 20 mM of the axonal shaft (Fig. 5F). Low levels (e.g., 0 –5%) of
FIG. 3. Detyrosinated immunoreactivity is concentrated with proximal neurites. Cells were singly labeled for total, Tyr, Acet or Glu immunoreactivity. The arrowheads denote the position of the growth cone for each cell. Note that total and Tyr immunofluorescence are more evenly distributed along the neurite length than are Acet and Glu immunofluorescence.
258
FIG. 4. Relative distribution of modified, unmodified and total tubulin immunoreactivity in developing axonal neurites. Between 50 and 100 randomly selected cells in multiple microscopic fields from duplicate cultures reacted with each antibody were examined under phase-contrast and immunofluorescent optics. Neurites were bisected into equal proximal and distal halves according to their respective phase-contrast images using the “Neurite Length Macro” of NIH Image (see Materials and Methods). Neurites were then scored according to their relative intensity of immunostaining as “Distal . Proximal,” “Distal 5 Proximal” or “Distal , Proximal” as determined under immunofluorescent optics. Note that at all measured times during outgrowth, the distribution of Tyr immunoreactivity roughly parallels that of total immunoreactivity. By contrast, Glu and Acet immunoreactivity are either evenly distributed along the neurite length or, unlike Tyr immunoreactivity, are concentrated within the proximal half of the neurite.
SHEA turnover or preferential subunit Acet within proximal versus distal neurite regions. Following treatment with the MT-stabilizing drug, taxol, the entire MT length was Acet (Fig. 7). Acetylated MTs furthermore accumulated within the growth cone, a region from which they were typically absent prior to taxol treatment (e.g., Fig. 1). These data indicate that there is no inherent inability for Acet MTs to accumulate along the entire neurite length and therefore indicate that selective Acet of MTs occurs within the proximal neurite region. Regional tubulin incorporation and/or MT translocation was further investigated by microinjection of biotinylated tubulin. Biotinylated tubulin associated with MTs within minutes after microinjection (Fig. 8). Consistent with the presence of labile (Tyr) MTs within the soma and proximal neurites, biotinylated subunits were initially detected within MTs within these regions. These analyses were unable to distinguish, and were not intended to distinguish, whether the appearance of MT-associated biotinylated subunits within proximal neurites was a function of localized assembly, transport of newly assembled perikaryal MTs into axonal neurites or both mechanisms. Nevertheless, these analyses indicate ongoing subunit incorporation within, and/or MT translocation into, proximal neurites. Because Acet is a function of MT age [3,11,16], one interpretation of the persistence of a concentration of Acet MTs within proximal neurites, despite localized subunit exchange and/or transport, is that the Acet MTs represent a subset of axonal MTs which either do not undergo subunit turnover or translocation or do so at rates vastly slower than those of the overall axonal MT population. Longer time points were not conclusive, because biotinylated tubulin rapidly dispersed throughout, and associated with MTs throughout, the entire neurite (not shown; e.g., see [20]). DISCUSSION
Acet-reactive MTs observed in immuno-EM analyses of segments 16 –20 were also detected within the final 15 mM of the axonal shaft up to the growth cone (Fig. 5F), as well as within random sections from “central” axonal regions (i.e., sections in which neither the perikaryon nor growth cone could be detected within a 15 to 20-mM radius; not shown). These data indicate that the proximal region of a subpopulation of the total axonal MTs is selectively acetylated. Acetylation and Glu are considered as markers for stabilized MTs [4]. To determine whether or not selective accumulation of Acet and Glu immunoreactivity within proximal neurites corresponded with selective MT stabilization in this region, the relative stability of MTs within axonal regions was probed by treatment with colchicine. Colchicine treatment substantially depleted MTs from both perikarya and neurites (Fig. 6). Axonal MTs survived colchicine treatment longer than perikaryal MTs, suggesting that they were preferentially stabilized as compared to perikaryal MTs. Continued colchicine treatment depleted distal axonal MTs and reduced proximal axonal MTs. The last surviving MTs were a subset of the MTs located within the proximal neurite; these are designated a “subset” because an overall reduction in all modified and unmodified MTs accompanied extended colchicine treatment (Fig. 6). The only Tyr immunoreactivity surviving colchicine treatment for 120 min precisely colocalized with Acet immunoreactivity, indicating that these relatively colchicine-resistant MTs (or regions of MTs) contained both Tyr and Acet subunits. These findings indicate that Acet immunoreactivity indeed denotes selectively stabilized MTs (or regions of MTs). Next, a different approach was used to address whether the selective stabilization of MTs/regions of MTs within the proximal half of axonal neurites was a consequence of differential subunit
The selective accumulation within NB2a/d1 axonal neurites of Acet and Glu MTs has been previously documented [27]. In the present study, differential distribution of these modifications between proximal and distal axonal neurite regions was demonstrated. These findings are consistent with the previous demonstration of selective stabilization of MTs within proximal axonal regions [3]. In contrast to selective concentration of Acet and Glu MTs within proximal neurites, most neurites contained similar amounts of total and Tyr MTs along their entire length. A minority of neurites displayed somewhat more Tyr and total MTs in the proximal half as did those displaying more Tyr and total MTs in the distal half, consistent with the demonstration that the addition of tubulin subunits to axonal MTs occurs at both the perikaryal and growth cone end of neurites [7]. These findings collectively demonstrate the existence of multiple MT populations within the developing axonal neurites of dbcAMP-treated NB2a/d1 cells. Information regarding the relative stability of axonal and perikaryal MT populations was provided by colchicine treatment. Microtubules displaying Acet immunoreactivity persisted the longest during colchicine treatment, indicating that MTs possessing this modification were selectively stabilized. Moreover, proximal neuritic Acet MT immunoreactivity persisted beyond that of perikaryal Acet MT immunoreactivity, despite the presence of significantly greater initial Acet immunoreactivity within many neuronal perikarya, suggesting that Acet MTs (or regions of MTs) within proximal axonal neurites are relatively more stable than those within perikarya. Although the vast majority of Tyr immunoreactivity was rapidly depleted by colchicine, the ultimate colocalization of Tyr and Acet immunoreactivity within the proximal neurite suggested that stable MTs contained both Acet and Tyr subunits (e.g., see also [1,4]). The notion that differentiating
STABILIZATION OF AXONAL MICROTUBULES
259
FIG. 6. Colchicine treatment differentially affects perikaryal and axonal, and Acet versus Tyr, immunoreactivity. Images present cultures following 0 –120 min in the presence of colchicine. At each time point, the identical field is presented as visualized under fluorescein and rhodamine optics to depict Tyr or Acet MTs, respectively. Colchicine treatment resulted in rapid, substantial depletion of MTs from both perikarya and neurites. Axonal MTs survived colchicine treatment longer than perikaryal MTs (e.g., compare 60-min colchicine treatment to 0 min), suggesting that they were preferentially stabilized as compared to perikaryal MTs (arrow at 60 min denotes most proximal end of axonal neurite). Continued colchicine treatment (e.g., 120 min) depleted distal axonal MTs and reduced proximal axonal MTs. Note that the residual Tyr immunoreactivity remaining after 120 min of colchicine treatment corresponds precisely with Acet immunoreactivity within proximal neurites (arrows).
NB2a/d1 cells selectively regulate MT stability was confirmed by the rapid accumulation of MTs along the entire neurite length following taxol-induced MT stabilization. Our analyses did not address the site(s) of MT nucleation. However, the length of axonal neurites (approximately 100 mM)
FIG. 5. Ultrastructural analyses of tubulin immunoreactivity and microtubule (MT) distribution in axonal neurites. Immunoelectron microscopy with 6-11B-1 confirmed that Acet immunoreactivity under these conditions is associated with MTs (panel A); in this representative micrograph, 15/15 colloidal gold particles are on or adjacent to MTs. A densely packed longitudinally oriented array of parallel MTs was observed along the length of axonal neurites (panel B). Additional panels (C–E) depict representative serial reconstructions of portions of one axonal neurite. Only a single definitive MT end was located within any of the axonal shafts examined (not shown; see text). All other MT profiles apparently terminating within a given section appeared within an adjacent section; arrows depict this in panels C–E, which present reconstructions from three adjacent sets of longitudinal sections. These analyses indicate that, under these conditions of differentiation, NB2a/d1 cells elaborate MTs that span the entire axonal length. The accompanying graph (panel F) presents quantification of collodial-gold-labeled MTs within axonal neurites. Micrographs from three separate cells subjected to immunoelectron microscopy with 6-11B-1 containing longitudinal axonal profiles were assembled in a collage that contained a portion of the perikaryon and the initial 20 mM of proximal axon, and scored at 1-mM intervals. Similar quantitation was carried out on collages that spanned the final 15 mM of the axonal neurite and contained at least a portion of the growth cone. The percentage of Acet MTs in each 1-mM segment was then calculated by determining the Number of Labeled MTs/Number of Total MTs 3 100. The graph presents the mean 6 standard deviation.
FIG. 7. Taxol-induced microtubule (MT) stabilization fosters accumulation of Acet immunoreactivity. Panels present distal portions of neurites from taxol-treated and untreated cells doubly labeled for total and Acet immunoreactivity as indicated. Only the distal portions of neurites are presented. Total tubulin immunoreactivity was prominent along the distal shaft, whereas Acet immunoreactivity was barely detectable. Mild total tubulin immunoreactivity, but no Acet immunoreactivity, was also present within the growth cone (arrows). Following taxol treatment for 2 h, however, Acet immunoreactivity was prominent within the distal neurite, and both Acet and total tubulin immunoreactivity were also prominent within the growth cone.
260
FIG. 8. Biotinylated subunits were initially detected within microtubules (MTs) within soma and proximal neurites. Cells were injected with biotinylated tubulin (Bt-tub). Ten minutes after injection, cells were fixed either with paraformaldehyde without extraction (which reveals both assembled and unassembled subunits) or with methanol following saponin extraction (which, as described previously, depletes unassembled subunits but retains subunits that have assembled into MTs). Comparison of unextracted and extracted cells indicates that substantial biotinylated subunits were depleted at this early time point. In addition to biotinylated immunoreactivity within perikarya, immunoreactivity within unextracted and extracted neurites was concentrated within the proximal half; arrows in phase image denote the entire neurite length.
following 7 days of treatment with dbcAMP vastly exceeds that of the perikaryon, indicating, as previously suggested [14], that MTs must undergo elongation within the axonal neurite. The initial accumulation of biotinylated tubulin within proximal neurites following its injection into the cell body is consistent both with the possibility of subunit addition to the proximal end of axonal MTs and the observation in this study of labile MTs throughout the neurite. However, a concentration of stabilized MT markers on a subset of MTs within the proximal neurite was observed in this study. Because the accumulation of Acet and Glu subunits within MTs is time dependent [3,11,16], subunits may not be added to the proximal end of this particular Acet- and Glu-rich MT subset. Rather, they are consistent with the exclusive addition of subunits to the distal end of this particular subset and/or suggest that this MT subset undergoes exceptionally slow, if any, turnover and its proximal end is preserved. These data are furthermore inconsistent with models suggesting that tubulin subunits are added exclusively to MTs within the proximal neurite region and that the MT array translocates en bloc along the axonal neurite length; such a model would necessitate that the oldest MTs or MT regions (and therefore those containing the highest proportion of Acet and Glu immunoreactivity) would be detected within distal neurite regions. These data are inconsistent with the structural hypothesis of axonal transport as originally proposed, which held that MTs were transported as a highly cross-linked network [17]. However, they are fully consistent with the subsequent refinement of the structural hypothesis, which considers that individual MTs may translocate independently of one another [18,19], provided that allowances are made for some MTs/MT populations to translocate substantially more slowly than others; this latter possibility allows for the observed preferential accumulation of age-related tubulin modifications within the proximal neurite region. A recent study of transport of radiolabeled tubulin [12] supports this latter possibility. In these studies, newly synthesized tubulin was conveyed along sensory axons following radiolabeling of developing dorsal root ganglia in situ. However, some newly synthesized tubulin accumulated along developing axons substantially after the passage of the transported wave, suggesting that a portion of tubulin may contribute to the development of a relatively stationary MT population in growing axons. Selective stabilization of certain MTs could provide a nidus for MT elongation during axon out-
SHEA growth [20]. In addition, selective stabilization of even a subset of MTs within the proximal axonal region represents a potential mechanism to provide structural support for the oldest region of the developing neurite, yet it still permits relative plasticity within the distal region, which is associated with continued growthrelated remodeling. This possibility is consistent with the delay in accumulation of Acet and Glu MT immunoreactivity within dbcAMP-induced axonal neurites until 3 days after treatment [27]. Moreover, the existence of a stationary population of MTs has been postulated to serve as a scaffold against which moving MTs derive motive force [24]. It was further noted that any such putative stationary MT population must be relatively small, because it is not reflected in total MT distribution [24]. The finding that the accumulation of Acet and Glu immunoreactivity is not reflected in the distribution of total MT immunoreactivity in the present study concurs with this latter suggestion of Reinsch et al. [24]. Previous studies have demonstrated that NB2a/d1 axonal neurites attain a degree of stability derived from interactions of axonal MTs and neurofilaments and, furthermore, that these interactions are mediated in part by the MT-associated proteins MAP1B and tau [26]. It remains possible that those MTs exhibiting strong Acet and Glu immunoreactivity within proximal neurites represent those MTs that have undergone relatively long-lasting MT-MT or MT-neurofilament interactions and that the proximal domains of these MTs have accumulated the age-related modifications of Acet and Glu as a consequence of a resultant delay in turnover. In consideration of these and previous findings, it is suggested that selective stabilization of MTs within the proximal region of developing axons may provide a relatively stationary scaffold against which moving MTs and/or other membranous and cytoskeletal organelles translocate to more distal axonal regions for continued axonal outgrowth. ACKNOWLEDGEMENTS
The author thanks Peter Paskevich for assistance with electron microscopy, Drs. Steven Vincent and Francine Benes for assistance with confocal microscopy, Ms. Mary Lou Beermann for expert technical assistance, and Dr. Peter Baas for his generous gift of biotinylated tubulin and his continued interest and advice. This research was supported by the NIMH.
REFERENCES 1. Ahmad, F. J.; Pienkowski, T. P.; Baas, P. W. Regional differences in microtubule dynamics in the axon. J. Neurosci. 13:856 – 866; 1993. 2. Baas, P. W. Microtubules and axonal growth. Curr. Op. Cell Biol. 9:29 –36; 1997. 3. Baas, P. W.; Ahmad, F. J.; Pienkowski, T. P.; Brown, A.; Black, M. M. Sites of microtubule stabilization for the axon. J. Neurosci. 13:2177– 2185; 1993. 4. Baas, P. W.; Black, M. M. Individual microtubules in the axon consist of domains that differ in both composition and stability. J. Cell Biol. 111:495–509; 1990. 5. Black, M. M.; Slaughter, T.; Moshiach, S.; Obrocka, M.; Fischer, I. Tau is enriched on dynamic microtubules in the distal region of growing axons. J. Neurosci. 16:3601–3619; 1996. 6. Brady, S. T.; Tytell, M.; Lasek, R. J. Axonal transport and axonal tubulin: Biochemical evidence for cold-stability. J. Cell Biol. 98:847– 858; 1984. 7. Brown, A.; Slaugther, T.; Black, M. M. Newly-assembled microtubules are concentrated in the proximal and distal regions of growing axons. J. Cell Biol. 119:867– 882; 1992. 8. Cambray-Deakin, M. A.; Burgoyne, R. D. Posttranslational modifications of a-tubulin: Acetylated and detyrosinated forms in axons of rat cerebellum. J. Cell Biol. 104:1569 –1574; 1987.
STABILIZATION OF AXONAL MICROTUBULES 9. Cleveland, D. W.; Hoffman, P. H. Slow axonal transport models come full circle: Evidence that microtubule sliding mediates axon elongation and tubulin transport. Cell 67:453– 456; 1991. 10. Daniels, M. The role of microtubules in the growth and stabilization of nerve fibers. Ann. N. Y. Acad. Sci. 253:535–544; 1975. 11. Gundersen, G. G.; Khawaja, S.; Bulinski, J. Postpolymerization detyrosination of a-tubulin: A mechanism for subcellular differentiation of microtubules. J. Cell Biol. 105:251–264; 1987. 12. Hoffman, P. N. Pulse-labeled tubulin is retained along growing axons. Soc. Neurosci. Abstr. 22:1986; 1996. 13. Jacobs, J. R.; Stevens, J. K. Changes in the organization of the neuritic cytoskeleton during nerve growth factor-activated differentiation of PC12 cells: A serial electron microscopic study of the development and control of neurite shape. J. Cell Biol. 103:663– 673; 1986. 14. Joshi, H. C.; Baas, P. W. A new perspective on microtubules and axon growth. J. Cell Biol. 121:1191–1196; 1993. 15. Keith, C. H.; Blane, K. Site of tubulin polymerization in PC12 cells. J. Neurochem. 54:1258 –1268; 1990. 16. Khawaja, S.; Gundersen, G. G.; Bulinski, J. C. Enhanced stability of microtubules enriched in detyrosinated tubulin is not a direct function of detyrosination level. J. Cell Biol. 106:141–149; 1988. 17. Lasek, R. J. Translocation of the axonal cytoskeleton and axonal locomotion. Philos. Trans. R. Soc. Lond. B Biol. Sci. 299:313–327; 1982. 18. Lasek, R. J. Polymer sliding in axons. J. Cell Sci. 5(suppl. 5):161–179; 1986. 19. Lasek, R. J. Studying the intrinsic determinants of neuronal form and function. In: Lasek, R. J.; Black, M. M., eds. Intrinsic determinants of neuronal form and function. New York: Liss; 1988:1–58. 20. Li, Y.; Black, M. M. Microtubule assembly and turnover in growing axons. J. Neurosci. 16:531–544; 1996. 21. Miller, K. E.; Joshi, H. C. Tubulin transport in neurons. J. Cell Biol. 133:1355–1366; 1996. 22. Ochs, S. Axoplasmic transport and its relation to other nerve functions. New York: Wiley; 1982.
261 23. Okabe, S.; Hirokawa, N. Turnover of fluorescently labeled tubulin and actin in the axon. Nature (Lond.) 343:479 – 482; 1990. 24. Reinsch, S. S.; Mitchison, T. J.; Kirschner, M. W. Microtubule polymer assembly and transport during axonal elongation. J. Cell Biol. 115:365–379; 1991. 25. Sabry, J.; O’Connor, T. P.; Kirschner, M. W. Axonal transport of tubulin in T1 pioneer neurons in situ. Neuron 14:1247–1256; 1995. 26. Shea, T. B.; Beermann, M. L. Respective roles of neurofilaments, microtubules, MAP1B and tau in the outgrowth and stabilization of axonal neurites. Mol. Biol. Cell. 5:863– 875; 1994. 27. Shea, T. B.; Beermann, M. L.; Nixon, R. A. Post-translational modification of a-tubulin by acetylation and detyrosination in NB2a/d1 neuroblastoma cells. Dev. Brain Res. 51:195–204; 1990. 28. Shea, T. B.; Beermann, M. L.; Nixon, R. A.; Fischer, I. Microtubuleassociated protein tau is required for axonal neurite elaboration by neuroblastoma cells. J. Neurosci. Res. 32:363–374; 1992. 29. Shea, T. B.; Fischer, I.; Sapirstein, V. S. Effects of retinoic acid on growth and morphologic differentiation of mouse NB2a neuroblastoma cells in culture. Dev. Brain Res. 21:307–314; 1985. 30. Shea, T. B.; Paskevich, P. A.; Beermann, M. L. The protein phosphatase inhibitor okadaic acid increases axonal neurofilaments and neurite caliber, and decreases axonal microtubules in NB2a/d1 cells. J. Neurosci. Res. 35:507–521; 1993. 31. Shea, T. B.; Sihag, R. K.; Nixon, R. A. Neurofilament triplet proteins of NB2a/d1 neuroblastoma: Posttranslational modification and incorporation into the cytoskeleton during differentiation. Dev. Brain Res. 43:97–109; 1988. 32. Takeda, S.; Funakoshi, T.; Hirokawa, N. Tubulin dynamics in neuronal axons of living zebrafish embryos. Neuron 14:1257–1264; 1995. 33. Tashiro, T.; Kurokawa, M.; Komiya, Y. Two populations of axonally transported tubulin differentiated by their interactions with neurofilaments. J. Neurochem. 43:1220 –1225; 1984. 34. Vincent, S. L.; Khan, Y.; Benes, F. M. Cellular colocalization of dopamine D1 and D2 receptors in rat medial prefrontal cortex. Synapse 19:112–120; 1995.