.' ‘ .
.I— -~.,... .

' -. ..
<

 

LIBRARY
Michigan State
University

 

 

 

This is to certify that the

thesis entitled

Loca] Contro] of the Axona] Cytoskeieton

presented by

Peter w. Baas

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Phys10100V

  

Major professor

Date £6 3’ /€§77

0-7639

439

"u
3.)

47

.43‘4

 

MSU

LIBRARIES
m

 

 

 

 

RETURNINQ_MAI§E£BL§:
Piace in book drop to
remove this checkout from
your record. flfl§§ wiii
be charged if book is
returned after the date
stamped beiow.

 

 

 

 

 

 

LOCAL CONTROL OF THE AXONAL CYTOSKELETON

By

Peter William Baas

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Physiology

1987

 

ABSTRACT
LOCAL CONTROL OF THE AXONAL CYTOSKELETON
BY

Peter William Baas

The proximal segment of amputated neurites routinely regrows while
the distal segment only regrows if first it collapses to a "bead" of
axoplasm. Heidemann et al have proposed that stable MT fragments may
serve as nucleating elements for the local control of MT dynamics in the
axon (J. Cell Biol. 99:1289-1295). To test this proposal, regrowth of
"beads" was examined after recovery from treatments (cold or 0.1 ug/ml
nocodazole) that depolymerized the HTS in the bead to fragments too
short to "unfurl" during regrowth. Regrowth ensued immediately after
rewarming or rinsing out the nocodazole at the same rate as unarrested
regrowth. Beads treated with 1.0 ug/ml nocodazole in 11 of 13 cases
were devoid of MT fragments, failed to regrow, and never reassembled MTs
even hours after rinsing. These data indicate that regrowth requires MT
nucleation sites suggesting that tubulin in the bead and perhaps in the
neurite has a limited capacity for self-nucleation.

The polarity orientation of MTs in intact neurites was uniform with
[+] ends facing the growth cone. In uncollapsed amputated neurites, [+1
ends faced the cut in regrowing proximal segments, while [-1 ends faced
the cut in non~regrowing distal segments. Regrowth of collapsed distal
segments was accompanied by a MT polarity reversal that re—established

the characteristic relationship between the new growth cone and [+1 ends

 

 

of MTs. These results suggest that [+] ends of MTs are a requirement
for growth cone formation, and that axonal MTs are organized by an
interaction between MT assembly and growth cone motility.

Intact neurites treated with various MT poisons lost the cytoplasmic
demarcation between the ribosome rich soma and the ribosome deficient
neurite. Treated neurites showed many ribosomes in the first 100 um and
a decreasing gradient with increasing distances from the soma.
Ribosomes appeared in proximal but not distal segments of neurites
amputated prior to nocodazole treatment. These results suggest that MTs
in the neurite prevent the movement of ribosomes into the ax0plasm, and
thus may play an early role in the cytoplasmic compartmentation of

neurons .

 

 

dedicated to my father

11

 

ACKNOWLEDGMENTS

I would like to thank Dr. Steve Heidemann for guidance throughout my
graduate studies. I would also like to thank Stu Pankratz for expert
advice, and Leigh White and Gary Sinclair for technical assistance in
some of the work presented here. I would also like to thank all my
friends for support and encouragement.

iii

TABLE OF CONTENTS

List of Tables

List of Figures

List of Abbreviations

Introduction

Chapter 1.

Chapter 2.

Chapter 3.

Microtubule Reassembly from Nucleating
Fragments during the Regrowth of
Amputated Neurites.

Introduction

Materials and Methods

Results

Discussion

Microtubule Polarity Reversal Accom-
panies Regrowth of Amputated
Neurites.

Introduction

Materials and Methods

Results

Discussion

Role of Microtubules in the Cytoplas-
mic Compartmentation of Neurons.
Introduction

Materials and Methods

iv

viii

xi

25

29
31
34
118

52

54

 

Results

Discussion

List of References

56

67

69

 

LIST OF TABLES

Table 1: Polarity Orientation of MTs in Intact

Neurites 35
Table 2: Polarity Orientation of MTs in Amputated

Uncollapsed Neurites NO
Table 3: Polarity Orientation of MTs in Regrown

Neurites as

LIST OF FIGURES

Figure 1: Phase-contrast micrographs of a sequence of
collapse and regrowth of neurites after amputation. 12

Figure 2: Transmission electron micrograph of a
distal segment of an amputated neurite collapsed
to a bead and fixed before regrowth. 1”

Figure 3: (a) Transmission electron micrograph of

a cold treated bead. (b) Serial reconstuction of

9 consecutive sections through the cold treated

bead. 15

Figure u: Transmission electron microgaphs of beads
treated with (a) 0.1 ug/ml and (b) 1.0 ug/ml
nocodazole. 16

Figure 5: Transmission electron micrographs of (a)

a neurite lysed in GEP buffer supplemented with

2 M glycerol, (b) a neurite bundle lysed in GEP

alone, and (c) a cold treated bead lysed in GE?

alone. 18

Figure 6: Transmission electron micrographs of an

amputated distal segment that collapsed to a bead then

regrew neurites. (a) Region of the bead, (b) region

of the regrown neurite. 20

Figure 7: Transmission electron micrographs of

(a) an intact neurite treated with 20 ug/ml

nocodazole, rinsed with PBS, and returned to

undrugged medium for 15 minutes before fixa~

tion. 22

Figure 8: Phase-contrast micrograph of a bead
treated with 1.0 ug/ml nocodazole for

15-20 minutes, rinsed twice with PBS, then
treated with fresh medium containing M.6 X

vii

 

10‘8 M taxol.

Figure 9: Transmission electron micrograph of
a cross section through a neurite mass grown
from a cultured ganglion chunk treated under
the MT polarity revealing "hook" procedure.

Figure 10: Phase-contrast micrograph of a
firmly adherent neurite whose proximal and
distal segments did not collapse 20 minutes
after amputation.

Figure 11: Transmission electron micrograph
of a cross section through the distal seg-
ment of an amputated neurite that failed to
collapse 15 minutes after amputation sub-
sequently treated under MT-polarity re-
vealing "hook" conditions.

Figure 12: Phase-contrast micrograph of an
amputated neurite whose distal segment col-
lapsed and then regrew neurites.

Figure 13: Phase—contrast micrographs of an
amputated neurite whose distal segment (a) in-
completely collapsed to a bead that (b) regrew
a neurite from the remaining stub of the
original neurite.

Figure 1“: Transmission electron micrographs
of a cross section through a regrown neurite
treated under MT-polarity revealing "hook"
conditions.

Figure 15: Transmsission electron micrograph
of an embryonic chick sensory neuron in cul—
ture showing the clear demarcation between the
cytoplasm of the soma and axoplasm of the
neurites.

Figure 16: Transmission electron micrograph
of a neuron treated for 15-20 minutes with 1.0
ug/ml nocodazole showing the loss of the
demarcation between compartments.

viii

 

24

36

38

41

”3

44

47

57

59

Figure 17: Transmission electron micrograph
of (a) neurite, and (b) soma that were treated
with nocodazole, fixed, then treated for 20
hours with 0.5 mg/ml ribonuclease.

Figure 18: Transmission electron micrographs
of regions of different neurites treated for
15-20 minutes with 1.0 ug/ml nocodazole at
various distances from the soma. Regions ca.
(a) 50 um, (b) 150 um, (c) 200 um, and (d)
500 um from the soma.

Figure 19: Transmission electron micrographs
of (a) the proximal retraction bulb, and (b)
the collapsed distal segment of a neurite ampu-
tated 50-150 um from the soma prior to noco—
dazole treatment.

Figure 20: Transmission electron micrographs

of (a) a cold treared neurite, and (b) a neurite
treated with 2 ug/ml cytochalasin D for 15-20
minutes both at ca. 15 um from the soma.

ix

 

61

63

611

66

 

MT

IF

PBS

GEP Buffer

Bead

CW

CCW

distal segment

proximal segment

LIST OF ABBREVIATIONS

Microtubule

Intermediate Filament

Phosphate Buffered Saline

0.1 M PIPES, pH 6.9. with KOH, 1 mM EGTA, 2 mM GTP

Collapsed Distal Segment of Amputated Neurite

Clockwise

Counterclockwise

amputated neurite fragment freed from cell body

amputated neurite fragment attached to cell body

 

 

INTRODUCTION

The cytoskeleton is a set of structural proteins that are critical
for the development and maintenance of cell shape and cytoplasmic
organization in eucaryotic cells (1, 28). The cytoskeleton consists
principally of Inicrotubules (MTs), intermediate filaments (IFs), and
actin filaments, but also involves many regulatory and linking elements
(1, 28). The factors that regulate, organize, and integrate the
cytoskeleton in eucaryotic cells are poorly understood. An especially
useful model for studying cytoskeletal functions underlying cellular
morphology has been the neuron. The neuron shows a remarkable
anisotropic geometry, and is arguably the cell type whose function is
most intimately related to its shape. Neurons consist of a rounded cell
body that extends axons over distances many times the diameter of the
cell body (77). The cytoplasm of the axon is differentiated from that
of the cell body by the absence of protein synthetic machinery (57) and
the presence of a dense array of highly ordered cytoskeletal elements
(15, 16, SH, 63, 75, 77, 96). Actin filaments form a cortical network
below the plasma membrane of the axon, and extend in growing axons into
a motile leading edge termed the growth cone (16, 52, 53, 96). MTs and
IFs are present in parallel array within the shaft of the axon, but
terminate at the base of the growth cone (16, 53, 96). MTs are dynamic
structures generally regarded to provide plasticity to the axon during

axonal outgrowth, while IFs are thought to provide stability as the axon

 

 

matures (5“). There is widespread agreement that axonal growth depends
on both MT-based and actin-based cytoskeletal networks (”9, 53, 97). In
this regard, the axon is a system uniquely suited for studying the
integration of different elements of the cytoskeleton.

MTs in the axon are uniform in their polarity orientation (19, 32,
39, N1), and yet do not extend from a centriole or any observable MT
organizing structure (6“, 81, 98). McIntosh has argued that traditional
MT organizing centers function as if they were localized regions of MT
fragments available for elongation (67). Morris and Lasek (70) and
Brady et al (12) have Speculated that the spatial organization of axonal
MTs may be locally regulated by nucleation sites woven throughout the
the axon. Heidemann et al (91) and others (9, 12, 70) have proposed
that especially stable regions of otherwise labile MTs may serve as the
proposed nuclei. It is known however that the concentration of tubulin
in the axon is well above the concentration that permits gg.lg9!9
initiation of MT‘ assembly 13 ggtgg (71). If Heidemann's scheme is
correct, how is the local assembly of MTs in the axon limited to the
elongation of existing MTs?

The genesis of MT organization in the axon is an issue not addressed
by the local nucleation hypothesis. Lasek and his colleagues have
argued that MTs are assembled in the cell body and then transported in
orderly fashion down the axon as intact polymers where they are
ultimately degraded in the axon terminal (5". 55). This proposal has
recently been challenged, however, by evidence that the growth cone is a
major site of MT assembly in growing axons. Bamburg et al reported that
axonal elongation is 100 times more sensitive to MT poisons applied at

the growth cone than applied at the cell body or axonal shaft (5).

 

 

 

 

Bamburg's scheme is consistent with the local nucleation hypothesis in
that MT assembly in the axon would occur via the elongation of existing
MTs at the base of the growth cone. Joshi et al have suggested that the
elongation of MTs at the growth cone is regulated by the motility of the
growth cone releasing compressive force on the HTS (“9). Such an
intricately regulated system would likely demand that the HTS be
uniformly oriented in the axon in that the two ends ()f a MT‘ have
different assembly kinetics and equilibria (2, 6, 7, 25, 65, 66). Might
the genesis of MT organization in the axon be the result of an
integration between MT assembly and the actin-based motility' of the
growth cone?

The controversy of whether MTs are assembled principally in the cell
body or at the growth cone is one of a number of issues that question
the relative contributions of the cell body and local (in the axon) cues
for the regulation of the axonal cytoskeleton. Solomon has argued that
the pattern of outgrowth and branching of axons is dictated by
"endogenous determinants" from the cell body (85), whereas others have
argued that the growth cone is more reSponsive to environmental cues
once it crawls away from the cell body (13, 1M, 22, M7, 59). Shaw and
Bray reported that the collapsed distal segments of amputated neurites,
although physically separated from the cell body, are able to regrow
neurites entirely similar to normal neurites on the light microscopic
level (82). Do the MTs within the regrown neurites enjoy the same
degree of spatial organization as do MTs in normal neurites? If so,
what does this say about the relative contributions of the cell body and
local influences to MT organization in axons?

Lasek and Brady have argued that the organization of cytoplasm in

 

 

the neuron is not a Special case, but rather is an exaggeration of
themes common across eucaryotic cells (57). Protein synthetic machinery
tends to be localized to the perikaryon, while the differentiated
cytoplasm specific to the cell type tends to be found at the periphery
of the cell. In the case of the neuron, the long distances traversed by
axons demand the existence of a complex system for the transport down
the axon of substances synthesized in the perikaryon. Although axonal
transport has been extensively studied (33, 36, “3, SM, 55), to my
knowledge no experimental evidence exists for the mechanism by which the
cytoplasm in neurons is compartmented into "translational" and
"expressional" types of cytoplasm. Lasek has proposed that ribosomes
are localized to the perikarya of mature neurons by attachments to
cytoskeletal elements within the perikaryon (5”). Lasek and Katz have
pointed out that the "granular" cytoskeleton of the perikaryon is devoid
of MTs (56). Might the exclusion of ribosomes from axoplasm during
neuronal development be attributable in part to the cytoskeleton,
specifically the MTs, within the axon rather than (or in addition to)
the cytoskeleton of the perikaryon?

The research presented in the next three chapters represents an
attempt to address the questions put forth in this introduction.
Chapter 1 exploits the collapse/regrowth model to test the validity of
the local nucleation hypothesis. Chapter 2 similarly exploits the
collapse/regrowth model to explore the proposed relationship between MT
organization in the axon and the actin-based motility of the growth
cone. Chapter 3 examines the role of MTs within the axon in the

exclusion of ribosomes from axoplasm during early neuronal deveIOpment.

 

 

CHAPTER 1. Microtubule Reassembly from Nucleating Elements
During the Regrowth of Amputated Neurites

INTRODUCTION

Microtubules (MTs) are dynamic structures that are critical for the
growth and maintenance of axons (16, 53, 95, 96). Axonal MTs enjoy a
high degree of spatial organization (15, 39, H1, 75), yet are apparently
not under the influence of a traditional MT organizing center (6A, 81,
98). Indeed, a growing body of evidence suggests that local and
environmental cues play an important role in the control of the axonal
cytoskeleton (13, 22, N7, 59). Lasek and his colleagues (12, 70) have
speculated that cytoskeletal nucleation sites may be woven throughout
the axon for the local maintenance of cytoskeletal morphology.
Heidemann et al have reported that MTs of cat sympathetic nerves are
oriented in polar fashion (39), and are able to recapitulate their polar
organization during recovery from cold or drug induced depolymerization
(41). Heidemann et al proposed that MT fragments that resist
depolymerization may play a role in MT organization by acting as
nucleating "seeds" for MT elongation (N1). Morris and Lasek (70) and
Brady et a1 (12) have similarly suggested that stable polymers in the
cytoskeleton may play an important role in cytoskeletal organization.
Biochemical evidence (HS) suggesting that stable MTs exist 33 133:9 as

short, disassembly-resistant regions of longer, otherwise labile MTs has

 

 

 

been confirmed by observations lg 1112 (50, 79). Stable MTs have been
shown to be biochemically distinct from the labile polymer (9, 12, ”2,
HS, ”8, 92). Nucleated assembly from stable fragments seems attractive
in that elongation from an existing MT is faster and more energetically
favorable than 95 £219 initiation (2. 73, 7”). A functional role for
stable fragments as nucleation seeds would lend substance to
speculations on individual nucleating elements for each MT (84, 90).
Stable fragments may also provide the physical "cap" against
treadmilling proposed by Kirschner (51), and may be of particular
importance in MT organization if axonal MTs share the kind of "dynamic
instability" recently reported of MTs in 31339 (69).

Shaw and Bray reported that distal segments of amputated neurites
from cultured chick sensory neurons collapse to shortened segments or
"beads" of axoplasm that subsequently regrow neurites entirely similar
to ordinary neurites on the light microscopic level (82). Wessells et
al (93) demonstrated the same phenomenon in ciliary ganglion neurons and
provided evidence for the importance of growth cone motility in
cytoskeletal organization during regrowth. George and Lasek have
suggested that the collapse of amputated neurites may involve the same
kinds of regulatory mechanisms that coordinate the cytoskeleton during
slow transport (35). Although completely divorced from the cell body,
the amputated neurites clearly contain all the informational cues
required for cytoskeletal reorganization underlying collapse and
regrowth. In an effort to understand this cytoskeletal reorganization we

have examined the disassembly and reassembly of MTs during the collapse

and regrowth of amputated neurites.

 

 

 

 

 

 

MATERIALS 532 METHODS

Cell Culture

Embryonic chick sensory neurons were cultured rising a procedure
slightly modified from that of Shaw and Bray (82). Dorsal root ganglia
were dissected from the lumbosacral regions of 12 day embryos, and
placed in L-15+ medium (L-15 purchased from Gibco Laboratories, Grand
Island, NY supplemented with 0.6% glucose, 2 mM L-glutamine, 100 U/ml
penicillin, and 100 mg/ml streptomycin). The ganglia were rinsed twice,
then treated with 0.25% trypsin for 25 minutes at 37oc. The trypsin was

removed, L-15+

with 10% fetal calf serum (Hazleton Dutchland, Inc.,
Denver, PA) was added, and the ganglia were triturated with a pipette
into a single cell dispersion. The cells were rinsed twice, and plated
in L-15+ with 10% fetal calf serun, 0.6% methyl cellulose (Methocel ARM,
Dow Chemical Company, Midland, MI), and 100 ug/ml nerve growth factor
from mouse saliva (17). The cells were plated into 35 mm Corning tissue
culture dishes at a density sufficiently low so as to render

non-neuronal contamination inconsequential. Cultures were kept in a

humid, 37°C incubator for 18-30 hours prior to experimentation.

Amputation Experiments

Cultures were covered with a thin layer of mineral oil to prevent
evaporation and maintain pH, and placed on a microscope stage warmed to
37°C with a Sage air curtain incubator. Temperatures of warming

cultures were monitored by submerging the feedback thermister of the air

 

 

 

 

 

 

curtain incubator directly into the culture medium, and taking the
cooling mode as an indicator that the set temperature had been reached.
Control experiments using larger volunes of fluid showed the air curtain
incubator to be accurate within 1°C. Amputations were performed with a
Leitz micromanipulator using glass needles made on a microelectrode
puller. The needles were often broken to a diameter of about 10
microns, and neurites were cut from directly above to minimize lateral
displacement. In an effort to maximize the number of amputated neurites
that collapsed completely to beads of axoplasm, i.e. with no visible
processes remaining after collapse, unbranched neurites with modest
growth cones were selected for amputation and cut approximately 100 um
from the growth cone. Regrowth of collapsed neurites was arrested in
three sets of experiments using conditions known to depolymerize MTs.
Cultures were cold treated by incubating the dishes in a -5°C bath for 2
hours, or treated with either 0.1 ug/ml or 1.0 ug/ml nocodazole (Aldrich
Chemical Company, Milwaukee, WI) for 15-20 minutes at 37°C. After these
treatments, regrowth properties were examined by rewarming the cold
treated beads on the microscope stage, or rinsing the nocodazole-treated
beads twice with PBS then returning them to undrugged medium.

In another set of experiments, cold treated beads were permeabilized
under conditions slightly modified from a procedure reported by Cande
(23) to depolymerize spindle MTs in PtK1 cells. In these experiments
cultures were gently lysed after cold treatment for 4-5 minutes at 37°C
in a buffer containing 0.1 M PIPES adjusted to pH 6.9 with KOH, 1 mM
EGTA, and 2.5 mM GTP (GEP buffer), with 0.08-0.11 Brij 58. Normal
neurites were also lysed in this buffer with or without added 2M

81Ycerol. The latter conditions have been found to stabilize assembled

 

 

MTs against dissassembly by dilution while allowing diffusion of free
tubulin in a variety of cells (8, 86).

Regrowth properties of collapsed neurites treated with 1.0 ug/ml
nocodazole were examined 15 and 30 minutes after rinsing by exposing
them to either 2.3 x 10"l3 M or ”.6 x 10-8 M taxol. Taxol, a drug
reported to promote MT assembly (80), was a gift from the Deve10pmental

Therapeutics Program of the National Cancer Institute.

Electron Microscopy

Cultures were fixed for electron microscopy using 1 of 2 methods.
Neurites and untreated collapsed neurites were fixed by replacing the

medium with a solution containing 0.1 M cacodylate, 1 mM MgCl and 2%

2!
glutaraldehyde for 20-30 minutes. Because collapsed neurites treated
under MT depolymerizing or lysis conditions tended to lift from the dish
surface during exchanges of fluid, these cultures were fixed by the
addition of an equal amount of the same medium or buffer containing Ni
glutaraldehyde. The cultures were then rinsed twice in 0.1 M cacodylate
with 5% sucrose, treated with 0.15‘ tannic acid for 5 minutes, rinsed

twice, postfixed in 1% 050 for 5 minutes, dehydrated in ethanol series,

u
and embedded in Polybed 812 (Polysciences, Inc., Warrington, PA). Thin
sections cut parallel to the substratum were stained with uranyl acetate

and lead citrate, and observed with a Phillips 300 transmission electron

microscope.

 

 

 

Serial Reconstruction pf Beads

In order to determine whether the MTs left in collapsed neurites
after cold treatment were in the form of fragments, a modification of
the method of Nicklas (72) was used for serial reconstruction. Outer
boarders and MTs together with many membranous vesicles within the
collapsed neurite were traced from the micrographs onto transparent
plastic sheets. The tracings were aligned first by the membranous
registration markers, then adjusted to maximize the matches of MT ends
in consecutive sections by moving the tracings within a range of two MT
diameters. MT length and orientation were then depicted in the form of

a composite drawing.

 

 

 

am

Collapse pf Amputated Neurites

Our studies confirmed the observations of Shaw and Bray (82).
Distal segments of amputated neurites collapsed to beads of axoplasm
(hereafter "beads") which then regrew neurites almost immediately.
Because we were interested in examining collapse and regrowth
separately, we selected for amputation unbranched neurites with modest
growth cones (Figure 1a), and amputated distal segments approximately
100 microns in length. We found such neurite segments most likely to
completely collapse to beads before beginning regrowth. Figure 1 shows
a typical sequence of amputation, collapse, and regrowth. Fifty three
amputations were performed in a set of experiments to monitor regrowth.
In 16 cases, regrowth began prior to the complete collapse of’ the
neurite. In 37 cases, however, the amputated neurite collapsed to a
flattened sphere or "bead" of axoplasm with no visible processes prior
to the onset of regrowth. Initially, the growth cone flaired and began
to "melt" almost immediately upon severing the neurite. Then, as
reported by Shaw and Bray (82), the distal segment coiled (Figure 1b) as
it collapsed to a bead (Figures 1c and d). Electron micrographs of
distal segments at various times after amputation show varying degrees

of‘ MT depolymerization roughly correlated to the degree of' coiling

suffered by the neurite.

 

 

Figure 1: Phase-contrast micrographs of a sequence of collapse and
regrowth of neurites after uputation. (a) Neurite before amputation.
(b) 3 minutes after amputation, growth cone is flaired and neurite is
coiled and retracted. (c) 12 minutes after unputation, collapse to a
bead is almost complete. (at) 13 minutes after amputation, collapse is
coaplete. and minor regrowth has begun. (e) 16 minutes after amputation,
mltipolar regrowth: are apparent. (f) 18 minutes after amputation,
regrowths have lengthened. Bar, 20 um.

 

 

 

 

Ultrastructure pf Collapsed Neurites with and without MT

Depolymerization

We examined the ultrastructure of 10 amputated, untreated neurites
collapsed to beads. All ten clearly showed many long MTs in addition to
multiple cross-sections and apparent fragments (Figure 2). Electron
micrographs of u beads that were cold treated for 2 hours in a -5°C bath
(Figure 3a) show the marked absence of any long MTs similar to those
found in the untreated beads. In order to confirm that the HTS left in
the cold treated bead were indeed short fragments, we reconstructed 9
consecutive serial sections, approximately 751, of one typical bead.
The reconstruction (Figure 3b) shows many cross sections of MTs and 158
longitudinal fragments. In this particular bead all MTs were found to
be shorter than 0.5 microns in length while over 801 were shorter than
0.2 microns in length. Treatment of collapsed neurites with 0.1 ug/ml
nocodazole for 15—20 minutes likewise resulted in substantial MT
depolymerization. We examined u such beads and found short fragments as
well as some longer MTs in the peripheral regions (Figure Na). However,
treatment with a higher dose of 1.0 ug/ml nocodazole appeared to
completely depolymerize MTs. We examined sections from 3 such beads and
could find no unambiguous MTs (Figure ub).

We wished to determine if the MT fragments that persisted in the
cold treated bead were merely "mass action fragments" that could not
depolymerize in the high tubulin concentration of the cold treated bead,
or if they were in some way distinct from the majority of MT mass that
had depolymerized. We reasoned that detergent lysis would release

soluble tubulin dimer allowing depolymerization of any "mass action

 

 

 

Transmission electron micrograph of a distal segment of an

Figure 2

Many

amputated neurite collapsed to a bead and fixed before regrowth.

MTs of varying lengths are apparent. Bar, 0.5 um.

 

 

 

 

 

 

Figure 3: (a) Transmission electron micrograph of a cold treated bead.
(b) Serial reconstruction of 9 consecutive sections through the cold
treated bead. Arrows mark MT fragment longitudinal sections. Recon-
structions shows MT longitudinal sections as line segments and MT cross
sections as dots. Bar, 0.5 um.

 

 

 

 

Figure 11: Transmissicn electron micrcgraphs of beads treated with (a)
0.1 ug/ml and (b) 1.0 ug/lnl nocodazole. HT fragments remain in the
lower dcsed bead (arr-curs), while depolymerization is complete in the
higher dosed bead. Arrow in (b) marks anbiguous vesicular element. Bar.
0.5 um.

 

 

 

 

 

 

l7

fragments." The efficacy of this treatment in depolymerizing MTs in
neurites was examined by comparing the MT array left in intact neurites
lysed in a MT reassembly buffer ('GEP" see Methods section) with the
array left in neurites lysed in the same buffer supplemented with 2 M
glycerol. The latter conditions have been found to stabilize assembled
MTs against dissassembly by dilution while allowing diffusion of free
tubulin in a variety of cells (8, 86). Light microscope observations of
neurites lysed in GEP without glycerol showed a transient beading within
1 minute of adding lysis buffer, an effect symptomatic of MT
depolymerization in neurites (26, H3, 50). Neurites lysed in GEP with
glycerol showed no such beading. Electron micrographs of neurites lysed
in GEP with glycerol show a dense array of long MTs (Figure 53), while
electron micrographs of neurites lysed in GE? alone show substantial MT
depolymerization (Figure 5b). Electron micrographs of‘ cold treated
beads lysed in GEP without glycerol (Figure Sc) show MT fragments
similar to those MT fragments in unlysed, cold treated beads. That is,

lysis of cold treated beads caused no major additional MT disassembly.

Regrowth from Collapsed Neurites

Regrowth of neurites from beads began either within the first 30
seconds after collapse or often before collapse was complete. We
followed regrowth in some cases for “5-60 minutes after collapse. In
these cases the total length of the regrown neurites approximated that
Of the original distal segment, although the bead never entirely

disappeared. The fastest and most robust regrowth typically occurred

 

 

 

 

 

Figure 5: Transmission electron micrographs of (a) a neurite lysed in
GEP buffer supplemented with 2 M glycerol. (b) a neurite bundle lysed in
GEP buffer alone. and (c) a cold treated bead lysed in GEP alone. Lysis
in GEP alone permitted substantial MT disassembly by dilution in normal
neurites but apparently permitted no further depolymerization in the
cold treated beads. Bar, 0.5 ml.

 

 

0V8

pri

neu
OOH

ori

 

 

 

19

over the first 15 minutes. Out of 53 amputations, 16 began regrowth
prior to the complete collapse of the neurite. In these cases, the
remnants of the growth cone or neurite elongated to form the regrown
neurite(s). In the 37 cases in which the amputated neurite collapsed
completely there was no relationship between the number and/or
orientation of regrown neurites with the orginal neurite/growth cone.
In all cases regrowth ensued at rates of 3.5-".5 microns/minute per
regrown neurite averaged per bead. These rates were at least double
those reported by Shaw and Bray (82), perhaps due to our use of plastic
instead of glass substrata. Electron micrographs of regrown beads show
paraxial MT arrays in the regrown neurites (Figures 6a and b) entirely
similar to those in ordinary neurites, and fewer MTs in the bead than
before regrowth (Figure 6a).

After rewarming, regrowth from cold treated beads was very similar
to regrowth from untreated beads. Cold arrested regrowth resumed within
30 seconds of the cultures reaching a narrow temperature range within
1-20C of 37°C. The rates of regrowth from cold treated beads were in
the same range of 3.5-“.5 microns/minute as regrowth from untreated
beads.

The regrowth of nocodazole treated beads was more difficult to study
in that the rinsing required to remove the nocodazole often caused
detachment of the bead from the dish surface. Both cold and nocodazole
treated beads were less adherent to the substratum than untreated beads.
However, cold treated beads did not require a medium change and thus
could be regrown with minimal disturbance. Only 6 of NO beads remained
attached after rinsing of the lower dose (0.1 ug/ml) of nocodazole and

only 5 of 50 beads survived rinsing of the higher dose (1.0 ug/ml). All

 

 

 

 

 

.5... m6 Jam .3750: £8.30.“ oz» ho coamot 2:

.802 on» he came: 23 50.3.2.0: 30.30... :23 omen m 0» 839300 cos»
03.50: possesses co co anaemuosoue c9500: Saunas—acute no menu:

 

 

 

 

6 of
secox
micrt
noco
from

The

reg:
Eigl
ret
Ele

med

SN

MT:

21

6 of the beads treated with the lower dose began regrowth within 30
seconds of their return to undrugged medium at a rate of 3.5-u.5
microns/minute. However, the beads treated with the higher dose of
nocozole showed very poor regrowth. Only 1 of the S beads recovering
from the higher dose regrew and this was only after a 20 minute delay.
The other u never regrew even after u hours of observation.

We wished to determine whether MT assembly was occurring without
regrowth in beads rinsed after treatment with 1.0 ug/ml nocodazole.
Eight beads were treated with 1.0 ug/ml nocodazole, rinsed twice, and
returned to undrugged medium. Seven of the 8 failed to regrow.
Electron micrographs of 2 such beads fixed after 15 minutes in undrugged
medium and 3 such beads fixed after 30 minutes in undrugged medium show
no MTs (not shown). In 2 of the 30 minute cases we were able to obtain
sections covering approximately 901 of the bead and all were devoid of
MTs. The single bead that regrew did so after a 15 minute delay, its
regrown neurite contained MTs as expected (not shown) and the bead was
on the same plate as a non-regrowing bead found to be devoid of MTs. In
order to determine whether the apparent inability of such beads to
initiate MT assembly was due to the incomplete removal of nocodazole by
our rinsing method, we examined the recovery of normal neurites treated
with a 20 fold higher dose (20 ug/ml) of nocodazole. These unamputated
neurites retained a very few MT fragments (Figure 7a) and reassembled a
dense array of HTS after 15 minutes in undrussed medium (Figure 7b)’

In order to determine whether beads treated with 1.0 ug/ml

nocodazole were inviable and/or whether MT reassembly was no longer

All u beads treated with

Possible, we treated such beads with taxol.
8

1-0 UE/ml nocodazole for 15 minutes, rinsed and treated with ”.6 x 10

 

'1

 

 

Figure 7: Transmission electron micrographs of (a) an intact neurite
treated with 20 ug/ml nocodazole, rinsed with PBS, and returned to
undrugged mediua for 15 minutes before fixation. An apparently normal MT
array has been recovered. Bar, 0.5 um.

 

 

 

 

M tax

Such

unarrl

noood

but 5

23

M taxol rapidly regrew after a delay of about 3 minutes (Figure 8).
Such regrown neurites contained an array of MTs similar to those of
unarrested regrown beads (not shown). Six of 6 beads treated with
nocodazole, rinsed and treated with 2.3 x 10-5 M taxol failed to regrow

but showed substantial MT reassembly in the beads.

 

Figure 8: Phase-contrast micrograph of a bead treated with 1.0 ug/ml
nocodazole for 15-20 minutes, rinsed twice with PBS, then treated with
fresh medium containing LG x 10' M taxol. The head began robust
regrowth 3 minutes after adding the taxol. Arrow marks bead. Bar, 20 um.

 

 

 

 

gs

dis

56E

 

 

W

We report here an investigation of the role of MT assembly and
disassembly in the regrowth of' neurites from the collapsed, distal
segments of amputated neurites of chick sensory neurons. Fully
collapsed beads contained many long MTs (Figure 2). One possibility was
that these MTs might simply "unfurl" into the axon during regrowth in a
manner roughly analogous to Lasek's proposal for slow axonal transport
of MTs in the form of assembled polymers (55). However, beads treated
at —5°C for 2 hours (Figure 3) or with 0.1 ug/ml nocodazole for 15-20
minutes (Figure Ma) contained only short fragments of‘ MTs, and yet
regrew with no delay after reaching 37°C or rinsing out the nocodazole
in the absence of HTS long enough to "unfurl". The rates of regrowth in
all 3 cases were indistinguishable and in the range of 3.5-4.5
microns/minute. The rapidity of regrowth and lack of a delay prior to
the onset of regrowth argue in favor of nucleated reassembly of MTs and
against the the possibility of g3 flgvg initiation (2, H6, 73, 74).
Indeed, beads treated with a higher dose of 1.0 ug/ml nocodazole were
devoid of MTs (Figure ub), and failed to regrow in 11 out of 13 cases.
Although the poor adhesion of the treated beads to the substrata might
also disable regrowth by not allowing a growth cone to form, beads
treated with cold or with the lower dose of nocodazole were equally
precariously attached yet regrew without delay. Taxol, not known to
increase cell adhesion, enabled beads treated with 1.0 ug/ml nocodazole
to regrow. Incomplete removal of nocodazole by our rinsing method also
seems an unlikely factor in that a very high dose of 20 ug/ml was

rapidly reversible in normal neurites (Figure 7). Further, neurites

25

 

 

 

treai

3600

Nocc

vart‘

8101

tha

 

26

treated with the lower dose of nocodazole began regrowtrx within 30
seconds of rinsing. The rapidity of this reversal would not be expected
if nocodazole was sequestered by some peculiarity of the bead.
Nocodazole poisoning has been found to be immediately reversible in a
variety of cell types (27, 83). We conclude that rapid regrowth of
axonal fragments is dependent upon the presence of MTs or MT fragments
that serve as nucleating elements for MT elongation.

In the absence of nucleating elements, the expectation on the basis
of MT assembly in vitrg would be a delay in regrowth while initial
nucleation occurred, after which assembly would occur rapidly (2, 46,
73, 7”). In 2 out of 13 cases, beads treated with the higher dose of
nocodazole apparently sufficient to completely depolymerize MTs indeed
responded in this fashion, while the majority, 11 of 13, did not. One
possibility is that in the 2 cases of regrowth the 1.0 ug/ml nocodazole
was not sufficient to completely depolymerize bead MTs. More likely,
given the 15-20 minute delay before regrowth, is that in these cases the
bead tubulin managed self-nucleated MT assembly after complete
depolymerization. However, the failure of 11/13 to regrow even after
extended periods suggests a limited capacity of bead tubulin for
self-nucleation. This suggestion is supported by the observation that
beads that failed to regrow after 15 or 30 minutes in undrugged medium
were still devoid of MTs. Treatment of beads rinsed of 1.0 ug/ml
nocodazole with taxol, a drug that promotes MT assembly (80), was
sufficient to stimulate bead tubulin into initiating assembly g2 £239,
In H out of 4 cases, beads treated with “.6 x 10-8 M taxol after being
rinsed of 1.0 ug/ml nocodazole regrew after a 3 minute delay (Figure 8).

In 6 other cases a higher dose of 2.3 x 10-5 M taxol failed to stimulate

 

 

 

regrowth.
Letournea
from chi<
higher d<
assembly
MT assem
of the
experime
is char:
Lasek (
subunit
given t
factors
evidenc
is m
MT dis
was mu
in im
trans;
degm
01“ MT
0011a
into
term“
30001

“0001

27

regrowth, but did cause substantial MT reassembly in the beads.
Letourneau and Ressler had previously reported that neurite outgrowth
from chick sensory neurons was limited to low levels of taxol and that
higher doses inhibit neurite outgrowth (60). The ready regrowth when MT
assembly is stimulated by taxol suggests that it was indeed the lack of
MT assembly that inhibited regrowth previously, rather than inviability
of the bead or loss of competent tubulin. It is unclear from our
experiments whether the limited capacity of tubulin to initiate assembly
is characteristic of neuronal tubulin or of the bead only. Morris and
Lasek (71) have found in squid axoplasm that the amount of tubulin in
subunit form is higher than what would be predicted by in 11319 studies
given the total tubulin concentration. They concluded that there are
factors in the cell that inhibit the polymerization of tubulin. Our
evidence suggests that such factors may function in part by inhibiting
93 3919 initiation. On the other hand, we found that the bead had some
MT disassembly properties not shared by the intact neurite. Nocodazole
was much more effective at completely depolymerizing MTs in beads than
in intact neurites (Figures Nb and 7a). Lasek's model for the slow
transport of cytoskeletal elements maintains that MTs are selectively
degraded in axon terminals (5”). Burton has recently reported evidence
of MT disassembly at the distal ends of intact frog axons (18). During
collapse of the neurite, its terminal growth cone becomes incorporated
into the bead. If a degradation factor similar to that in synaptic
terminals exists iii the growth cones of‘ advancing neurites, it might
account for both the increased sensitivity of bead MTs to higher
nocodazole concentrations and their limited ability to self-nucleate.

If neuronal tubulin has only a limited capacity for self—nucleation,

 

the i
cytosl
wide
12, u
regio
fragh
fragn
bead1
reas
depo
lysi
uou]
Figl
com
neu

eup

28

the importance of MT nucleating elements would be paramount for
cytoskeletal reorganizations occurring during axonal growth. Abundant
evidence suggests that certain neuronal MTs are especially stable (9,
12, “2, “5, “8, 92), and that these MTs exist in the form of short
regions of otherwise labile MTs (“5, 50, 79). We wondered whether the
fragments left in the cold treated bead were merely "mass action
fragments" maintained because of a high tubulin concentration in the
bead, or if they were intrinsically stable regions of longer MTs. We
reasoned that if the fragments were "mass action fragments" they would
depolymerize if the free tubulin were allowed to escape from the bead by
lysing. If, however, the fragments were intrinsically stable, they
would not depolymerize upon subunit dilution. Indeed, as shown in
Figure 5. MT fragments persisted in cold treated beads lysed under
conditions that caused substantial MT depolymerization in normal
neurites. All available evidence indicates that the lysis buffer we
employed causes MT disassembly by dilution since it supports MT assembly
13 31319 while allowing soluble tubulin to diffuse from lysed cells (8,
23. 86). We interpret this as indirect evidence that the fragments in
the cold treated beads were different from the majority of the MT mass.
Job et al (MS), Brady et a1 (12), and others have provided evidence that
cold stable MTs differ in either MAP or tubulin composition. The
fragments left in the cold treated beads might similarly vary from the

solubule MT fraction for these reasons.

 

sun

of th
and 1
polar
asset
cont
orie
tem
cont

uni:

CHAPTER 2. Microtubule Polarity Reversal Accompanies Regrowth
of Amputated Neurites

INTRODUCTION

Microtubules (MTs) are intrinsically polar structures. The two ends
of the MT differ in the concentration of subunits needed for assembly
and their rates of elongation (2, 6, 7, 25, 65, 66). The role of MT
polarity in regulating cellular processes that are dependent upon MT
assembly, such as axonal growth (15, 75), is largely unexplored. Axons
contain dense paraxial arrays of MTs that are uniform in their polarity
orientation; the [+1 or "fast growing" ends of MTs are found in the axon
terminal (19, 32, 39, #1). The cellular and molecular factors
contributing to this uniform polarity orientation are lanknown. The
uniform polarity orientation of MTs extending from centrosomes or other
MT organizing centers indicates that such structures regulate MT
assembly in a polar fashion (30, 31, 37). However, unlike most
interphase MT arrays, axonal MTs do not extend from a centrosome or any
observable organizing structure (64, 81, 98). Although some studies
have postulated that MTs are first assembled in the cell body and then
transported down the axon as intact polymers (10, 55), others have
reported persuasive evidence that the growth cone is the major site of
MT assembly supporting axonal growth (5). Consequently, there is a

controversy whether the growth cone or cell body is the principle site

29

 

 

of MT or
impu
system i
neurons
reportet
segment
disasse
neurite
polar h
is the
cell 1
neurit
segmen
distal
growth
contra

regrow

30

of MT organization in axonal growth.

Amputated neurites of cultured chick sensory neurons are a useful
system in which to investigate the principle site of MT organization in
neurons and the role of MT polarity in axonal growth. We previously
reported that collapse and subsequent regrowth of the amputated neurite
segment freed from the cell body (the distal segment) involves MT
disassembly and reassembly (3). Here we wished to determine whether
neurites regrown from the collapsed distal segment contain a uniformly
polar MT array despite the absence of the cell body. Also of interest
is the difference in the regrowth of the neurite segment attached to the
cell body (hereafter proximal segment) and the distal segment in
neurites that remain extended after amputation. Extended proximal
segments routinely form new growth cones at their cut ends. However,
distal segments that remain extended after amputation do not form new
growth cones or show signs of regrowth at their cut ends (1”). This
contrasts with the routine formation of new growth cones and robust
regrowth by distal segments that first collapse to "beads" of axoplasm
(3, 82). Do differences in the ability of neurite segments to regrow

correlate with differences in MT polarity?

 

mm

C_e1_l

dete

were

intc

MATERIALS AER METHODS
Cell Culture

Embryonic chick sensory neurons were cultured by two methods. For
determination of MT polarity in intact neurites, dorsal root ganglia
were dissected from the lumbosacral regions of 12 day embryos and cut
into two or three pieces with forceps. The chunks were cultured (using
a modification of the method of Vlodavsky et a1, Ref. 91) on polylysine
treated tissue culture dishes in L-15+ medium (L-15 purchased from Gibco
Laboratories, Grand Island, NY supplemented with 0.6% glucose, 2 mM
L-glutamine, 100 U/ml penicillin, and 100 mg/ml streptomycin) containing
10% fetal calf serum, 20 ug/ml 7s NGF, and 5 ug/ml laminin
(Polysciences, Inc., Warrington, PA). The chunks were allowed to grow
neurites for “8 hours prior to experimentation. For amputation
experiments, dissociated cells were cultured as previously described (3)

on untreated or polylysine treated tissue culture plastic. Dissociated

cells were allowed to grow neurites for 12-30 hours prior to

experimentation.

Amputation Expriments

Amputation experiments were performed on a microscope stage

prewarmed to 37°C with a Sage air curtain incubator (as previously

described in Ref. 3). Unbranched neurites with modest growth cones were

selected and amputated ca. 100 um from the growth cone in an effort to

31

 

 

promot

[a
0

rear

the

subs

perr

del

uer

00F

32

promote complete collapse to "beads" of axoplasm.

MT Polarity Determination

The polarity orientation of MTs in the neurites of cultured neurons
was determined by the "hook method" previously described (19, 32, 39,
40, “1). The ganglia chunk preparation was used to increase the number
of neurites sectioned at one time. The polarity of MTs in amputated and
regrown neurites was determined using modifications designed to decrease
the tendancy of the neurite segments to loosen and wash away from the
substratum during hook treatment. The modifications reported here
permit the routine polarity determination of MTs in even these very
delicate poorly adhered cell processes. After amputations, cultures
were rinsed briefly in PBS, then exposed to a MT assembly mixture
containing 1.2 mg/ml beef brain MT protein (twice cycled in the presence
of glycerol) in 0.5 M PIPES, pH 6.9, 2.51 DMSO, 0.05 mM EDTA, 0.5 mM

EGTA' 0-5 mM M801 0.5 mM GTP, and 0.06-0.081 Brij 58 for 2 minutes at

2'
o
37°C. The mixture was then removed and replaced for 30 minutes at 37 C

with an identical mixture not containing the Brij 58 detergent- An

equal quantity of “1 glutaraldehyde in 0.1 M cacodylate was then added

to stabilize lysed neurites from being rinsed away. After 2-3 minutes

cultures were rinsed twice in 2% glutaraldehyde in 0-1 M cacodylate in

order to remove excess polymerized tubulin. Neurites were then aIIOWEd

to fix for 30 minutes after which cultures were again gently rinsed (5

times) in 0.1 M cacodylate with 5% sucrose. This procedure reSUIted in

ca. 75% of the lysed neurites remaining attached to the substratum.

 

After
acid, 1
030 1
Polybe
circle
market
Thin
chtai
asym
elec
flip
(in
and
his

SE(

33

After fixation, cultures were treated for 5 minutes in 0.15% tannic
acid, rinsed twice in 0.1 M cacodylate with 51 sucrose, postfixed in 1%
030“ for 5 minutes, dehydrated in ethanol series, and embedded in
Polybed 812 (Polysciences, Inc.). Samples were then relocated and
circled with a diamond marker objective. Orientation of the neurite was
marked by a series of circles colinear with the length of the neurite.
Thin sections were cut perpendicular to the substratum in order to
obtain cross-sections of the neurites. Blockfaces were cut in
asymmetric shapes so that samples could be quickly relocated on the
electron microsc0pe, and to make clear whether any sections had been
flipped. Sections were picked up on 0.75% formvar coated 75 mesh grids
(Ernest F. Fullam, Inc., Schenectady, NY), stained with uranyl acetate
and lead citrate, and observed with a Phillips 300 transmission electron
microscope. Care was taken to relocate the regrown neurite on thin

sections with great certainty (even though the rinsing procedure "as

quite effective in removing exogenous polymerized tubulin) in that much

0? the membrane was extracted in many cases. Hooks were analyzed as

previously described (”0).

 

mill

Ma;

from

end:

she
rev
cul

Hm

RESULTS

Heidemann and others have reported that in axons of nerve fibers
from intact animals MTs are oriented with their [+1 or "fast growing"
ends pointed away from the cell body toward the axon terminal (19, 32,
39, “1). In order to determine whether neurites of cultured neurons
share this uniform polarity orientation, we performed MT-polarity
revealing "hook" experiments on bundles of neurites grown out from
cultured chunks of embryonic chick dorsal root ganglia. In this method
"hooks" of MT protofilaments are formed on the sides of MTs. Hooks with
a clockwise (CW) orientation indicate that the [+3 end of the MT is
pointed toward the observer, while hooks with a counterclockwise (CCW)
orientation indicate that the [-] end of the MT is pointed toward the
observer. The results of 3 separate experiments are recorded in Table
1. Of those MTs that were unambiguously hooked, 97% had clockwise hooks
as seen from the growth cones looking toward the ganglia (Figure 9).
That is, like MT in axons from intact animals, neurite MTs were found to

have their [+1 ends in the distal terminal of the neuronal process.

34

WIM—

Sam

I....
ass

. i
|,"F,‘r

 

 

Table l; Polarity Orientation of MTs in Intact Neurites

 

ngplg Neurites Counted 9!. ggfl' 19!
1 50 152 3 98
2 100 296 13 96
3 65 2112 8 97
Total 215 690 2a 97

 

a
as seen by an observer standing at the growth cone.

 

35

 

 

 

 

 

Figure 9: Transmission electron micrograph of a cross section through a
neurite mass grown from a cultured ganglion chunk treated under the MT
polarity revealing "hook" procedure. Hooks are predominately CH (arrows)
indicating MTs are oriented with their [*1 ends pointing away from the
cell body toward the growth cone. Bar, 0.8 mm.

0

  
 
  

. .9.| ’3 r
of ; l
00 1:53;“ r

. '0‘ '_
9' g} i“. .‘ \;.'

 

 

 

 

 

 

=5
(v
-e

c011

his!

seg

8N)

di:

th:

 

 

Regrowth and MT Polarity Orientation ip Amputated Uncollapsed Neurite

Segments

Following amputation, neurite segments either remain extended or
collapse depending on how firmly they adhere to the substratum. On the
highly adhesive substrate of polylysine treated tissue culture plastic,
in over 15 cases we observed little or no collapse of either neurite
segment within the first 15-20 minutes after amputation. While new
growth cones routinely formed at the cut ends of the segments proximal
to the cell body, growth cones never formed at the cut ends of the
distal segments severed from the cell body. This finding is similar to
that of Bray et al on neurites amputated under the liighly adhesive
condition of serum free medium (1H).

All further amputation experiments were performed on the moderately
adhesive surface of untreated tissue culture plastic. Approximately 9
out of 10 distal segments of neurites amputated on this substratum
collapsed, while roughly 1 out of 10 remained extended. This allowed us
to examine regrowth and MT polarity in both collapsed and extended
neurite segments under the same substrate conditions, which are known to
affect growth cone formation (1%, 59, 93). We observed 7 amputated
neurites whose proximal and distal segments remained extended after
amputation. For these uncollapsed amputated segments, again the cut
ends of the proximal segments routinely formed new advancing growth
cones usually within 15-20 minutes, while the cut ends of the distal
segments never formed new growth cones (Figure 10). Some cut ends of
Proximal and distal segments retracted a few microns and formed rounded

varicosities of cytoplasm (retraction bulbs) on their cut ends. These

37

 

 

 

 

 

 

 

Figure 10: Phase-contrast micrograph of a firmly adherent neurite whose
proximal and distal segments did not collapse 20 minutes after
amputation. The cut end of the segment proximal to the cell body has
formed a new growth cone (thin arrow), while the cut end of the segment
distal to the cell body remains crushed in appearance. Thick arrow marks
scar (Tom nputation with glass needle. Bar, 20 um.

 

 

 

 

 

retr

segr

forr

ret‘

rep

dis

usi

for

per

1191

 

 

39

retraction bulbs formed new growth cones on cut ends of proximal
segments, while retraction bulbs on cut ends of distal segments never
formed a growth cone. In a very few (less than 5%) cases a distal
retraction bulb formed one or two wispy filopodia. We previously
reported that retraction bulbs are sites of MT disassembly (3), and MT
disassembly is known to stimulate formation of filopodia (14, 50).

We examined the polarity orientation of MTs in both the proximal and
distal segments of uncollapsed neurites 15 minutes after amputation
using the very gentle conditions for MT-polarity revealing hook
formation described in Materials and Methods. These conditions
permitted the determination of MT polarity in even these delicate
neurite segments. Although the degree of hooking was low (generally
less than 30%), the results were very consistant in all samples we
analyzed. We used the original growth cone of the neurite prior to
amputation as the reference point for both proximal and distal segments.
In all M proximal and all M distal segments we analyzed, over 901 of the
unambiguously hooked MTs were clockwise (Table 2, Figure 11). That is,
in both segments [+1 ends were pointed away from the cell body toward
the original growth cone. Plus ends of MTs faced the cut end of the
regrowing proximal segments, while [—] ends of MTs faced the cut end of

the non-regrowing distal segments.

 

 

 

 

 

 

Table 2; Polarity Orientation of MTs i Amputated

Uncollapsed Neurites

e i Q

 

Sample Proximal/Distal pg 993 1 CW
1 Proximal 26 2 93
1 Distal 38 2 95
2 Proximal 51 1 98
2 Distal 32 3 91
3 Proximal 38 2 95
3 Distal 38 1 97
i
“ Proximal 11 O 100
u Distal 18 0 100

 

6
,side of the cut with respect to the cell body.
as seen by an observer standing at the original growth cone of p
the neurite. j
l

40

 

 

Figure 11: Transmission electron micrograph of a cross section through
the distal segment of an amputated neurite that failed to collapse 15
minutes after amputation subsequently treated under MT-polarity
revealing "hook" conditions. Hooks are predominately CW (arrows)
indicating [#1 ends of MTs pointing toward the growth cone. ie. {-3 ends
of MTs pointed toward the cut. The polarity of MTs was similarly [+1
ends toward the original growth cone in proximal segments, ie. [+1 ends
pointed toward the cut. Bar, 0.8 um.

 

 

Polarity Orientation of MTs in Neurites Regrown form Collapsed Distal

_.——-—-

Segments

As previously reported (3, 82, 93), distal segments of neurites that
underwent significant collapse ‘to beads of axoplasm after amputation
routinely regrew neurites (Figure 12). We found a close correlation
between the degree of collapse and the ability' to regrow. Distal
segments that retracted less than approx. 60% of their initial amputated
length did not regrow. Distal segments that did collapse and regrow
formed at least two new advancing growth cones in all cases. In 62
collapse/regrowth events we monitored, 18 "beads" began to regrow before
completely retracting all remnants of the original neurite. In all
these cases, the remaining short stub of neurite always gave rise to a
regrowing neurite (Figure 13). In the no cases of regrowth from a bead
with no visible processes, only 7 regrew a neurite colinear with the
original. In 23 of the nu cases the bead regrew tripolar or tetrapolar
neurites (Figure 1f). In 1N other cases, the bead regrew bipolar
neurites but at a significant angle to the original neurite (Figure 12).
Except in the cases in which stubs of the original neurite remained, we
could find no correlation between the pattern of regrowth and any aspect
of the original neurite, growth cone, or neuron.

MT polarity orientation in neurites regrown from beads of axoplasm
was determined by the same method used for uncollapsed, amputated
neurites. We used the bead of axoplasm as the point of reference for MT
polarity because the bead occupied the position of the original growth
cone on the unamputated neurite. Table 3 shows that 85% to 100% of the

hooked MTs were counterclockwise (as seen from the bead, see Figure 14)

42

 

 

 

 

 

 

Figure 12: Phase-contrast micrograph of an amputated neurite whose
distal segment collapsed and then regrew neurites. The regrown

neuritu regrew bipolar at a sharp angle to the orientation of the
original neurite. The cut end of the proximal segment formed a
retraction bulb, then a new growth cone (thin arrow). Thick arrow marks
the sear n-om mutation with glass needle. Bar, 20 In.

 

 

 

 

 

 

 

 

 

 

Figure 13: Phase-contrast micrographs of an amputated neurite whose
distal segment (a) incompletely collapsed to a bead that (b) regrew a
neurite from the remaining stub of the original neurite. The proximal
cut end formed a new advancing growth cone. Bar, 20 u.

 

 

 

 

 

 

 

 

 

 

45

regardless of the direction or pattern of regrowth in all 9 regrown

neurites we were able to examine. That is, [+] ends of MTs were now in

the newly formed growth cone and [-] MTs ends were located in the

position of the original growth cone.

 

 

 

 

 

Table 3; Polarity Orientation pf MTs lg Regrown Neurites

Sample Pattern
1 bipolar
2 bipolar
3 bipolar
“ multipolar
5 bipolar
6 bipolar
7 multipolar
8 bipolar
9 bipolar

 

Collapse

complete

complete

complete

complete

complete

complete

complete

incomplete

incomplete

29

100

21

H6

10

I‘m.

1

O
O
3:

100

100

100

85

100

100

100

 

a
as seen by an observer standing at the bead (original growth cone).

46

 

 

 

Figure 1“: Transmission electron micrographs of a cross section through
a regrown neurite treated under MT-polarity revealing "hook" conditions.
Hooks are predominately CCW (arrow) indicating [+1 ends of MTs pointed
away from the bead. The polarity of MT: was [+1 ends away form the bead
regardless of the pattern of regrowth. Bar, 0.8 um.

 

 

 

 

 

same
axon:
ends

from
neur

cut

dist

regr
sin:
nel
cen

0011]

an;
and
Nei
0u1

re

 

 

mission

We report here that MTs in the neurites of cultured neurons have the
same uniform polarity orientation as previously reported for MTs in the
axons of mature neurons (19, 32, 39, A1). The [+1 or "fast growing"
ends of MTs are found in the growth cone or axon terminal. As expected
from this MT polarity orientation in intact neurites, MTs in amputated
neurites that did not collapse were oriented with [+1 ends facing the
cut end of the proximal segment, and [-1 ends facing the cut end of the
distal segment. Our data show that MTs reorganized and reversed their
polarity during the regrowth of neurites from collapsed distal segments
of amputated neurites. Prior to amputation or collapse [+1 MT ends face
an observer standing at the growth cone, while after collapse and
regrowth the stationary observer sees [-1 ends of MTs. This finding is
similar to that of McNiven et al who showed that arms of fish
melanophores severed from the cell body form a new pigment migration
center from which MTs are organized with a uniform polarity reversed
compared to the original orientation (68). In our experiments [+1 ends
of MTs were closely correlated with the formation of new growth cones in
amputated neurites both at the proximal segment of uncollapsed neurites
and at the growth cones newly formed from collapsed distal segments.
Neither neurite elongation nor growth cone formation was observed at the
cut ends of the distal segments that remained extended. Distal segments
regrew neurites only if they initially collapsed to a signficant degree.
Since MT disassembly accompanies neurite collapse (3), this finding
suggests that MT reorganization is a requirement for growth cone

formation in the distal segment. After reorganization [+1 MT ends faced

 

 

 

 

the i
depen
asses
consi
100

appl
been
and

form
him

his

 

49

the growth cones. Although axonal growth has long been thought to
depend upon MT elongation (15, 5”. 75), our results suggest that
assembly at the [+1 ends is required for growth. This interpretation is
consistent with the finding of Bamburg et al that axonal extension is
100 times more sensitive to colcemid applied to growth cones than
applied elsewhere (5). The formation of growth cones has previously
been shown to be stimulated after experimentally induced MT disassembly
and reassembly (1H). Wessells et al speculated that growth cone
formation requires cytoskeletal ends (93). Our data support this
hypothesis and suggest that growth cone formation requires [+1 ends of
MTs in the immediate vicinity of the axonal surface/cortex.'

We find particularly interesting the ability of beads of axoplasm
completely separated from the cell body to reorganize MTs so that
regrown neurites share the same uniform MT polarity orientation
characteristic of intact neurites and axons. We found that the number
and spatial orientation of regrown neurites was independent of any
observable quality of‘ the original neurite. This suggests that the
potential for forming neurites is widespread around the margin of the
axoplasmic bead and that neurite formation is occurring g5 ppyp. The
collapsed distal segment can apparently reorganize a uniformly polar MT
array and support neurite growth without contributions from the cell
body such as centrioles (81, 87), endogenous determinants (85). or slow
axonal transport (33, 55). At the very least, our experiments show that
local, ie. in the neurite, processes are sufficient for the task of
organizing MT assembly during neurite growth. Based on the
ultrastructural similarity of regrown and normal neurites (3) and the

similarity of regrowth to normal growth, it is our view that the same

 

 

 

mean
and
asse
stru
prev
neur
outg
pror
con
fon
tha
elo
end
cor
rd
cm

th

 

50

mechaniwn of MT organization obtains in both normal neurite outgrowth
and the regrowth studied here. We suggest that organization of MT
assembly at the growth cone is crucial for axonal growth. No organizing
structure other than MTs or MT fragments were observed in this or
previous ultrastructural study (3) of axoplasmic beads and their regrown
neurites. We speculate that MT organization accompanying neurite
outgrowth is not the result of‘ a MT organizing structure but ()f a
process, the interaction of MT assembly with the motility of the cell
cortex. Neurite outgrowth begins with motile activity such as filopodia
formation and ruffling by the cortex of the cell/bead (93). We suggest
that in order for the typical growth cone structure to form and neurite
elongation to begin, a region of the motile margin must "recruit" [+1
ends of MTs. Joshi et al have proposed that MT elongation at the growth
cone is then regulated by the actin—based advance of the growth cone
releasing compressive force on the MTs, thus lowering their critical
concentration for assembly (U9). We speculate that [+1 ends of MTs at
the growth cone are an important element in this scheme in that they
would have a competitive assembly advantage over [—1 ends at any given
tubulin concentration and degree of compression.

The small fraction of MTs that displayed polarity revealing hooks in
amputated neurites is a regretable but probably unavoidable shortcoming
of these experiments. Conditions for forming hooks on MTs always
involve changing the fluid surrounding the cells and at least one fluid
must have detergent (40). These requirements were nearly incompatible
with the study of amputated neurites because of their precarious
attachment to the culture substratum (3). Despite a great deal of trial

and error, we were unable to define conditions that produced a high

 

 

fractic
segmeni
with h
oursel
Nevert
findir

affect

 

 

Sl

fraction of MTs with hooks and preserved the attachment of the neurite
segments. It is possible that neuronal MTs are difficult to decorate
with hooks: in all previous studies of MT polarity in neuronal tissue by
ourselves and others the extent of "hooking" was low (19. 32, 39, “1).
Nevertheless, we would point to the consistency of our results and the
finding in cat sympathetic nerve that the degree of hooking does not

affect the conclusion with respect to MT polarity (H1).

 

 

 

CHAPTER 3. Role of Microtubules in the Cytoplasmic
Compartmentation of Neurons

INTRODUCTION

The exaggerated geometry of the neuron has made it a uniquely
suitable model in which to study the organization of cytoplasm (16, 54,
57, 75, 77, 89, 96). Neurons display a high degree of regional
specialization (20, 36, 38, 57, 76, 77). Lasek and Brady have argued
that the cytoplasm of neurons can be compartmentally divided into the
ribosome rich "translational" cytOplasm of the perikaryon and the
ribosome deficient "expressional" cytoplasm of the axon (57). This kind
of organization can also be noted in pancreatic endocrine cells (57),
certain protozoa (11), fibroblasts (34), and HeLa cells (24, 44, 58),
but is pronounced in neurons due to their extension of translational
cytoplasm over long distances in the form of axons. The restriction of
protein synthetic machinery to the perikaryon in neurons demands the
existence of a complex system of axonal transport to nourish the
axoplasm. Axonal transport has been extensively studied (36, 43, S4,
55), but to our knowledge there has been no eXperimental work examining
how ribosomes are localized to the perikaryon. Lasek (54) has proposed
that "free" ribosomes in neurons are restricted from entry into axons by
cytoskeletal attachments in the perikaryon. Most ribosomes in immature

and growing neurons however appear not to be immobilized around the

52

 

 

EIIII____________________________________——_i
53

nucleus, nor are they anchored to endoplasmic reticulum (16, 21, 26, 62,
63, 89, 96), and yet they are still restricted from freely diffusing
into the axoplasm.

The cytoplasm of the axon is further distinguished from the soma by
the presence of dense paraxial arrays of cytoskeletal elements critical
for axonal tranSport (15, 16, 36, 43, 54, 57). Lasek and Katz have
noted that the "agranular" cytoskeleton of the axon is unique from the
"granular" cytoskeleton of the perikaryon by the marked absence of
microtubules (MTs) in the latter (56). The restriction of ribosomes to
one compartment and MTs to the other suggests that the selective
exclusions may be coordinated. We previously reported that the
ultrastructure of PC12 neurites becomes very similar in appearance to
that of the soma after treatments that depolymerize MTs (49). We
recently reported preliminary evidence that the number of visible
ribosomes in the axon—like neurites of cultured embryonic chick sensory
neurons is substantially increased after MT depolymerization (4). These
results suggested that MTs in the neurite may play a role in the
restriction of ribosomes to the soma. We report here studies designed

to determine the role of' MTs in cytoplasmic compartmentation during

early neuronal development.

 

 

 

 

IF’!

 

 

MATERIALS AND METHODS

Cell Culture

Embryonic chick sensory neurons were cultured as previously

described (3).

Experimental Treatments

Cultures were treated with various drugs by replacing the medium
with fresh medium containing the drug. Cultures were treated for 15-20
minutes with 1.0 ug/ml nocodazole (Aldrich Chemical Company, Milwaukee,
WI), 10 ug/ml vinblastine sulfate, 100 ug/ml griseofulvin, or 2 ug/ml
cytochalasin D (Sigma Chemical Company, St. Louis, MO). Other cultures
were partially submerged for 2 hours in a -5°C circulating cold bath.
In some experiments, neurites were amputated at various distances from
the soma before drug treatment using our modification (3) of the method

of Shaw and Bray (82).

 

Electron Microscopy

Samples were prepared for electron microscopy as previously
described (3). In one set of experiments cultures were incubated after
fixation in 0.5 mg/ml ribonuclease (Sigma Chemical Company, St. Louis,

MO) in 0.13 M cacodylate for 20 hours prior to osmication (to confirm

54

 

 

 

55

dense particles as ribosomes as described in Ref. 88). As an estimate
of ribosomal density, ribosomes were counted and averaged in random 0.5
x 0.5 um regions of longitudinal thin sections ca. 50 nm thick. At

least 5 sections of each sample were examined in this fashion.

 

 

 

 

 

 

Met
ten
tvm
per
reg
dif
MTs

betr

org;
the

naur

 

 

RESULTS

Ultrastructure pf Control Neurons

 

Although the ultrastructure of embryonic chick sensory neurons has
been examined in some detail (26, 96), we wished to further examine the
ultrastructure of control neurons with particular emphasis on the
density of ribosomes in the soma, hillock, and neurite regions. Most
neurites were 1-1.5 um in width and over 200 um in length. Figure 15
shows a typical control neuron of these dimensions. The somata of these
control neurons contained dense populations of ribosomes most of which
appeared to be unattached to endOplasmic reticulum as has been
previously reported of ribosomes in the somata of growing neurons (21,
29, 61, 62, 63, 78, 89). The ribosomes appeared in densities of ca.
150-300 per square um in longitudinal thin sections (see Materials and
Methods section). Ribosomes were identified by their diameter and
tendency to cluster into rosettes. Unlike mature neurons (57) but
typical of embryonic neurons (62, 63, 89), ribosomes extended to the
periphery of the soma and sometimes into the hillock region. In the
region of the hillock the cytoplasm of the axon became markedly
different in appearance from that of the soma. Dense paraxial arrays of
MTs and neurofilaments filled the axoplasm. The sharp demarcation
between the soma and neurite was also characterized by what appeared to
be a limitation of the free communication of rnitochondria and other
organelles between compartments. Ribosomal density quickly dropped past
the hillock region. The density of ribosomes in the first 15-20 um of

neurites was less than than 15 ribosomes per square um. Neurites that

56

 

Figure 15:

Tran-lesion electron micrograph of am embryonic chick
aenaory neuron in culture showing the clear duration between the

cytopla- of the some and axoplaI- of the neurites. Ribosomes densely
populate the some but are . ,, bar, 0.5 In.

 

 

 

 

 

58

were particularly short and stout (greater than 1.5 um in width and less
than 200 um in length) were less consistant in their ultrastructural
characteristics. Some of these neurites contained lower densities of
MTs, more ribosomes, and less clear cytoplasm demarcations between
compartments. Due to these inconsistancies, we selected for our

compartmentation studies neurites with the more typical dimensions.

Ultrastructure 93 Nocodazole Treated Neurons

Neurites grown under the conditions described above (see Materials
and Methods section) were sufficiently adherent to the substratum that
treatment with 1.0 ug/ml nocodazole for 15-20 minutes caused minimal if
any retraction of the neurites. Longer exposure times generally caused
marked retraction, lifting of the neurites from the substratum, or
fragmentation of the "beads" along the length of the neurite that often
accompany MT depolymerization (26, 43, 50). Neurites treated for 15-20
minutes showed substantial MT depolymerization, but were also markedly
different from control neurites in other aspects of their
ultrastructure. Electron micrographs of the soma, hillock, and proximal
neurite show the apparent dissolution of the clear compartmental
demarcation observed in control neurites (Figure 16). Ribosomes and
other organelles appeared to be in free communication between the soma
and neurite in this region. Ribosomes populated the hillock region and
proximal neurite in densities typical of the soma (150-300 per sqare
um). Also as previously reported of PC12 neurites treated under MT

depolymerizing conditions (49, 50), many regions of the nocodazole

 

 

 

 

 

 

 

 

Figure 16: Transmission electron micrograph of a neuron treated for
15-20 minutes with 1.0 ug/ml noocdaxole showing loss of demarcation
between compartments. Bar, 0.5 um.

I ,
, . I .I,‘
. “5”“6‘6’.
\

“’3‘

“3.42173: "

 

 

 

 

 

I7—7

60

treated neurites contained a large number of small internal
non-mitochondrial vesicles. We have reported preliminary evidence that
the additional vesicles are attributable to an increase in membrane
internalization that accompanies MT depolymerization in neurites (4).
Ribosomes in neurites were identified by their diameter and
similarity in appearance to the electron dense particles found in
control somata. In order to confirm that the "nocodazole induced"
particles were indeed ribosomes, we used a method previously employed by
Steward (88) for the identification of ribosomes in dendrites. After
nocodazole treatment and fixation with glutaraldehyde, cultures were
treated for 20 hours with 0.5 mg/ml ribonuclease. Steward reported that
such treatment causes a partial digestion of ribosomes that is clearly
detectable under the electron microscope. We found the treatment to be
very effective in digesting the nocodazole-induced particles in proximal
regions of neurites (Figure 17a), and equally effective at digesting the

particles that were almost certainly ribosomes in the soma (Figure 17b).

5;; Ribosomes Invading the Neurite from the Soma during 5:

Depolymerization?

In order to examine whether the ribosomes that appeared in
nocodazole treated neurites were arising due to decompartmentation from
the soma, or from elements already comprising the neurite, we examined
the ultrastructure of treated neurites at various distances from the
soma. Six of 6 neurites examined in regions past the hillock but within

the first 100 um from the soma were found to contain ribosomes in

 

 

 

 

 

Figure 17: Transmission electron micrograph of (a) neurite, and (b)
some that were treated with nocodazole, fixed, then treated for 20
hours with 0.5 mg/ml ribonuclease. The ribonuclease dissolved the
"nocodazole induced" particles in the neurite as well as the ribosomes
in the soma. Bar, 0.5 um.

   
   

’- n .
‘ is"; :‘ ".1
3151‘ “3’37. #3”)-

03'
a

 

 

 

62

densities typical of the soma (150-300 per square um, Figure 18a). Four
of 4 neurites examined in regions 100-200 um from the soma contained
many fewer ribosomes than in the first 100 um, but generally still
higher than in control neurites (20-60 per sqare um, Figure 18b).
Distal regions of each length range contained fewer ribosomes than
proximal; Figure 18c shows a neurite ca. 200 um from the soma with only
sparse somewhat ambiguous ribosomes. In four of 4 neurites examined in
regions beyond 200 um, we found no unambiguous ribosomes (Figure 18d).
In order to further test whether the ribosomes were moving down the
neurite after MT depolymerization, we amputated neurites at distances
between 50-150 um from the soma prior to nocodazole treatment. In both
cases in which we were able to observe both proximal and distal
fragments under the electron microscope, we found a dense population of
ribosomes (ca. 300 per sqare um) in the retraction bulb and proximal
segment still attached to the soma (Figure 19a), but few or no ribosomes

in the collapsed distal segment (Figure 19b).

In order to determine whether these ultrastructural alterations were
specifically attributable to MT depolymerization, and not to either a
specific side effect of nocodazole or to general insult, we examined the
ultrastructure of at least 2 neurites at each of the 3 distances
described above (under 100 um, 100-200 um, and beyond 200 um from the
soma) given other experimental treatments. Neurites treated with 10

ug/ml vinblastine, a MT poison with somewhat different properties than

 

 

 

Figure 18: Tran-lesion electron micrographs of regions of different
neurites treated for 15-20 minutes with 1.0 ug/ml nocodazole at various
distances from the some. legions on. (a) 50 um, (b) 150 um, (c) 200 um,
and (d) 500 um ”on the aoma. Ribosomes occupy proximal regions at ca.
the one density as in control somata, and more distal regions with
decreasing densities at increasing distances from the soma. Arrows mark
sparse somewhat ambiguous ribosomes at 200 ll. Increases in small
vesicles varied in different regions with no correlation to distance
from the some. Bar, 0.5 In.

              
    

      

~

*- ~w~rrz ”is 8'5? ’- .1 9 It‘s-'7 “ .
’ “fir'T-‘Wser' "‘73? “ W"-

   
   

\

 

 

 

 

 

 

 

Figure 19: Transission electron micrographs of (a) the proximal
retraction bulb, and (b) the collapsed distal scglent cf a neurite
amputated 50-150 m fro. the some prior to nocodazole treatment.
Ribosomes densely populate the proxinel segaent still attached to the
soma. but are absent or nearly so in the distal segment. Bar, 0.5 us.

w.k, ,"

‘.

 

 

 

 

65

nocodazole or colchicine (28), or cooled to -S°C for 2 hours displayed
the same ultrastructural changes as those treated with nocodazole
(Figure 208). Treatment with 100 ug/ml griseofulvin, another MT poison
with a different mechanism of action than nocodazole (28). was less
reliable in depolymerizing HTs. The ultrastructure of those neurites in
which griseofulvin was ineffective in depolymerizing MTs was entirely
similar to control neurites, while that of the neurites in which
griseofulvin was effective in depolymerizing MTs clearly showed the same
axoplasmic alterations as nocodazole treated neurites. Neurites treated
wifl1 2 ug/ml cytochalashi D, a potent actin disruptor (49), did not
display increases in ribosomes over control levels (Figure 20b).
Neurites suffering general insult by amputation alone also did not show

increases in ribosomes.

 

 

 

Figure 20: Transmission electron micrographs of (a) a cold treated
neurite. and (b) a neurite treated with 2 ug/ml cytochalasin D for 15-20
minutes both at ca. 15 um from the soma. The cold treated neurite shows
substantial 1T depolymerization and a density of ribosomes similar to

that of the soma. The cytochalasin treated neurite is nearly devoid of
ribosomes. Bar, 0.5 um.

 

 

 

 

DISCUSSION

Our evidence indicates that ribosomes are able to freely move from
the soma into the axon-like neurites of embryonic chick sensory neurons
after treatments that depolymerize the dense MT arrays normally present
in the neurites. We found densities of ribosomes similar to that of the
soma in the first 100 um of the neurites, and a decreasing gradient with
increasing distance from the soma (Figures 16 and 18). It is unlikely
that we were simply unmasking ribosomes already present in the neurite
in that ribosomes are not masked by MTs in dendrites (88) nor in initial
segments of immature axons (16, 21, 26, 62, 63, 89, 96). Distal
segments of neurites severed prior to nocodazole treatment dispayed no
increase in ribosomes, while proximal segments still attached to the
soma displayed increases typical of unsevered nocodazole treated
neurites (Figure 17). The increase in ribosomes was reproducible with
various MT poisons with different mechanisms of action, but. was not
reproducible by actin disruption with cytochalasin, nor by general
insult caused by amputation alone. We interpret these results as
suggesting that MT depolymerization specifically liberates ribosomes
from an influence that normally limits their entry into axons during
early neuronal development.

It is unlikely given the rapidity of the response that the movement
of ribosomes observed here can be attributable to the same mechanism
underlying the classic "chromatolytic" release of ribosomes from the
N1331 substance caused by axotomy (78). Furthermore, very few ribosomes
are bound to endoplasmic reticulum ("Nissl substance") even prior to MT

depolymerization in immature and growing neurons (21, 29, 61, 62, 63,

67

 

 

68

78. 89). Lasek (5“) has proposed that "free" ribosomes are actually
attached to cytoskeletal elements of the perikaryon in a fashion similar
to that observed in the perikarya of other cell types (24, 3“, an, 58,
9”). Ribosomes in embryonic neurons however appear not to be
immobilized around the nucleus as in mature neurons (57). Furthermore,
in direct contrast to our results, perikaryal attachments of ribosomes
to microtrabeculae in other cell types have been shown to be sensitive
to actin disruption but not MT depolymerization (RH, 58).

We would argue that the maintenance of cytoplasmic compartmentation
in embryonic neurons is primarily dependent upon the cytoskeleton within
the neurite rather than the soma. We suggest that the presence of a
dense array of HTS in neurites provides a physical barrier to the entry
and movement of ribosomes through the axoplasm. This suggestion is
substantiated by our finding that shorter stouter neurites with lower MT
densities less consistently contained sharp cytoplasmic demarcations
between compartments than did the more typical, thin neurites with
densely packed MTs. The finding that ribosomes extend to the periphery
of the soma in immature neurons but are restricted, even those not
membrane bound, to the perikarya of mature neurons suggests that the
proposed perikaryal attachments may be developmentally established. We
propose that the assembly of a dense array of MTs in axons during early
neuronal development provides an impediment to the movement of ribosomes
into the axoplasm. This may be an an initial measure toward cytoplasmic
compartmentation until perikaryal attachments to endOplasmic reticulum
and microtrabeculae can be secured. On the other hand, it may be the

principal mechanism by which cytoplasmic compartmentation is maintained

throughout the life of the neuron.

 

 

 

 

 

LI ST OF REFERENCES

 

 

 

1.

2.

3.

9.

10.

11.

 

LIST OF REFERENCES

Alberts, B., D. Bray, J. Lewis, M. Raff, K. Roberts, and J.D.
Watson. 1983. Molecular Biology of the Cell. Garland
Publishing, Inc., New York. 549-609.

Allen, C., and 0.0. Borisy. 1974. Structural polarity and
directional growth of microtubules of Chlamydomonas flagella. J.
Mol. Biol. 90: 381—402.

Baas, P.W., and 3.8. Heidemann. 1986. Microtubule reassembly from
nucleating fragments during the regrowth of amputated neurites.
J. Cell Biol. 103: 917-927. (Chapter 1 of dissertation.)

Baas, P.W., G.I. Sinclair, L.J. Lamp, and S.R. Heidemann. 1986.
Major alterations in axoplasm accompany microtubule
depolymerization in neurites. J. Cell Biol. 103: 79a. (abstact).

Bamburg, J.R., D. Bray, and K. Chapman. 1986. Assembly of
microtubules at the tip of growing axons. Nature 321: 788-790.

Bergen, L., and 6.6. Borisy. 1980. Head-to—tail polymerization of
microtubules in vitro: electron microscopic analysis of seeded
assembly. J. Cell Biol. 84: 141-150.

Binder, L., W. Dentler, and J.L. Rosenbaum. 1975. Assembly of
chick brain tubulin onto flagellar microtubules from
Chlamydomonas and sea urchin sperm. Proc. Natl. Acad. Sci. USA
72: 1122-1126.

Black, M.H., and L.A. Greene. 1982. Changes in the colchicine
susceptibility of microtubules associated with neurite outgrowth:
studies with nerve growth factor-responsive PC12 pheochromocytoma
cells. J. Cell Biol. 95: 379-386.

Black, M.M., J.M. Cochran, and J.T. Kurdyla. 1984. Solubility
properties of neuronal tubulin: evidence for labile and stable
microtubules. Brain Res. 295: 255-265.

Black, M.M., Keyser, P., and E. Sobel. 1986. Interval between the
synthesis and assembly of cytoskeletal proteins in cultured
neurons. J. Neurosci. 6: 1004-1012.

Bowser, S.S., S.M. McGee-Russell, and C.L. Rieder. 1984. Multiple
fission in Allogromia sp., strain NF (Formaminiferida): release,
disperal, and ultrastructure of offspring. J. Protozool. 31:

69

 

 

 

 

 

 

272-275.

12. Brady, S.T., M. Tytell, and R.J. Lasek. 1984. Axonal tubulin and
axonal microtubules: biochemical evidence for cold stability. J.
Cell Biol. 99: 1716-1724.

13. Bray. D. 1973. Branching patterns of individual sympathetic
neurons in culture. J. Cell Biol. 56: 702-712.

14. Bray, D., C. Thomas, and G. Shaw. 1978. Growth cone formation in
cultures of sensory neurons. Proc. Natl. Acad. Sci. 75:
5226-5229.

15. Bray, D., and D. Gilbert. 1981. Cytoskeletal elements in neurons.
Annu. Rev. Neurosci. 4: 505-523.

16. Bunge, M.B. 1973. Fine structure of nerve fibers and growth cones
of isolated sympathetic neurons in culture. J. Cell Biol. 56:
713-735.

17. Burton L.E., W.H. Wilson, and £.M. Shooter. 1978. Nerve growth
factor from mouse saliva. J. Biol. Chem. 7807-7812.

18. Burton, P.R. 1985. Microtubules of frog olfactory axons: their
length, number and the apparent fate of their distal ends. J.
Cell Biol. 101: 125a (abstract).

19. Burton, P.R., and J.L. Paige. 1981. Polarity of‘ axoplasmic
microtubules in the olfactory nerve of the frog. Proc. Natl.
Acad. Sci. USA 78: 3269-3273.

20. Caoeres, A., G.A. Banker, and L. Binder. 1986. Immunocytochemioal
localization of tubulin and microtubule-associated protein 2
during the development of hyppocampal neurons in culture. J.
Neurosci. 714-722.

21. Caley, D.W., and D.S. Maxwell. 1968. An electron microscopic
study of neurons during postnatal deve10pment of the rat cerebral
cortex. J. Comp. Neurol. 133: 17-28.

22. Campenot, R.B. 1977. Local control of neurite development by
nerve growth factor. Proc. Natl. Acad. Sci. USA 74: 4516-4519.

23. Cande, W.Z., J. Snyder, D. Smith, K. Summers, and J.R. McIntosh.
1974. A functional mitotic spindle prepared from mammalian cells
in culture. Proc. Natl. Acad. Sci. USA 71: 1559-1563.

24. Cervera, M., G. Dreyfuss, and S. Penman. 1981. Messenger RNA is
translated when associated with the cytoskeletal framework in
normal and VSV-infected HeLa cells. Cell 23: 113-120.

25. Cote, R.H., and 6.6. Borisy. 1981. Head-to—tail polymerization of
microtubules in vitro. J. Mol. Biol. 150: 577-602.

 

71

Fine structural changes in neurons associated
J.

26. Daniels, M.P. 1973.
with colchicine inhibition of nerve fiber formation in vitro.

Cell Biol. 58: 463—470.

27. DeBrabander, M., G. Geuens, R. Van de Veire, F. Thone, F. Aerts, L.
Desplanter, J. De Cree, and M. Borgers. 1977. The effects of R
17934 a new antimicrotubular substance, on the ultrastructure of

neoplastic cells in vivo. Eur. J. Cancer 13: 511-528.

28. Dustin, P. 1984. Microtubules (second edition). Springer-Verlag,
New York.

29. Eschner, J., and P. Glees. 1963. Free and membrane-bound
ribosomes in the maturing neurons of the chick and their possible

functional significance. Experientia 19: 301—303.

30. Euteneuer, U., and J.R. McIntosh. 1980. Polarity of midbody and
phragmoplast microtubules. J. Cell Biol. 87: 509-515.

Polarity of some motility

31. Euteneuer, U., and J.R. McIntosh. 1980.
related microtubules. Proc. Natl. Acad. Sci. USA 78: 372-376.

32. Filliatreau, 0., and L. DiGiamberdino. 1981. Microtubule polarity
in myelinated axons as studied after decoration with tubulin.

Biol. Cell 42: 69-72.
and J. Sjostrand. 1975. Slow axonal
blockade by interruption of contact
Brain Res. 86: 67-73.

The spatial distribution

33. Frizell, M., W.G. McLean,
transport of proteins;
between cell body and axon.

3“. Fulton, A., K. Wan, and S. Penman. 1980.
of polyribosomes in 3T3 cells and the associated assembly of
proteins into skeletal framework. Cell 20: 849-857.

Contraction of isolated

35. George, 5.8., and R.J. Lasek. 1983.
neural processes: a model for studying cytoskeletal translocation

J. Cell Biol. 97: 267a (abstact).
1980. Intracellular transport in

in neurons.

Grafstein, 8., and D.S. Forman.

36.
Physiological Rev. 60:

neurons.
Dynein—microtubule

37. Haimo, L.T., and B.R. Telzer. 1982.
interactions: ATP sensitive dynein binding and the structural
polarity of mitotic microtubules. Cold Spring Harbor Symp.

Quant. Biol. 46: 207-218.
38. Hammerschlag, R. 1983/84.
they are going?...Speculations on the role of molecular address

How do neuronal proteins know where

markers. Devel. Neurosci.
39. Heidemann, B.R., J.M. Landers, and M.A. Hamborg. 1981. Polarity

orientation of axonal microtubules. J. Cell Biol. 91: 661—665.
90. Heidemann, S.R., and (L Euteneuer. 1982. Microtubule polarity

 

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

S3.

72

determination based on conditions for tubulin assembly in vitro.
Methods Cell Biol. 24: 207-216.

Heidemann, S.R., M.A. Hamborg, S.J. Thomas, B. Song, S. Lindley,
and D. Chu. 1984. Spatial organization of axonal microtubules.
J. Cell Biol. 99: 1289-1295.

Hesketh, J.E. 1984. Differences in the polypeptide composition
and enzyme activity between cold-stable and cold-labile
microtubules and study of microtubule alkaline phosphatase
activity. FEBS Lett. 169: 313-318.

Rorie, H., T. Takenaka, and M. Kaiho. 1983. Effects of disruption
of microtubules on translocation of particles and morphology in
tissue cultured neurites. Brain Res. 288: 85-93.

Howe, J.G., and J.W.B. Hershey. 1984. Translocational initiation
factor and ribosome association with the cytoskeletal framework
fraction from HeLa cells. Cell 37: 85-93.

Job, D., C.T. Rauch, B.R. Fischer, and R.L. Margolis. 1982.
Recycling of cold stable microtubules: evidence that cold
stability is due to substoichiometric polymer blocks.
Biochemistry 21: 509-515.

Johnson, K.A., and 6.6. Borisy. 1977. Kinetic analysis of
microtubule self-assembly in vitro. J. Mol. Biol. 117: 1-32.

Johnston, R.N., and N.K. Wessells. 1980. Regulation of the
elongation of nerve fibers. Curr. Top. Dev. Biol. 16: 165-206.

Jones, D.H., E.G. Gray, and J. Barron. 1980. Cold stable
microtubules in brain studied in fractions and slices. J.
Neurocytol. 9: 493-504.

Joshi, H.C., D. Chu, R.E. Buxbaum, and S.R. Heidemann. 1985.
Tension and compression in the cytoskeleton of PC12 neurites. J.
Cell Biol. 101: 697-705.

Joshi, H.C., P. Baas, D.T. Chu, and S.R. Heidemann. 1986. The
cytoskeleton of neurites after microtubule depolymerization.

Kirschner, M.W. 1980. Implications of treadmilling for the
stability and polarity of actin and tubulin polymers in vivo. J.
Cell Biol. 86: 330-334.

Kuczmarski, B.R., and J.L. Rosenbaum. 1979. Studies of the
organization and localization of actin and myosin in neurons. J.
Cell Biol. 80: 356-371.

Landis, S.C. 1983. Neuronal growth cones. Annu. Rev. Physiol.
45: 567—580.

 

 

 

 

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

 

73

Lasek, R.J. 1981. The dynamic ordering of the neuronal
cytoskeleton. Neurosci. Res. Program Bull. 19: 7-32.

Lasek, R.J. 1982. Translocation of the neuronal cytoskeleton and
axonal locomotion. Philos. Trans. R. Soc. Lond. B Biol. Sci.
299: 313-327.

Lasek, R.J., and M.J. Katz. 1979. The axon and its origins. In:
Nerve and Muscle Research. (Proceedings of the IVth international
Congress of Neuromuscular Diseases, Montreal, Sept. 14-21, 1978.)
Amsterdam: Elsevier Science Publishers. 232-240.

Lasek, R.J., and S.T. Brady. 1981. The axon: a prototype for
studying expressional cytoplasm. Cold Spring Harb. Symp. Quant.
Biol. 46: 113-124.

Lenk, R., L. Ransom, Y. Kaufmann, and S. Penmann. 1977. A
cytoskeletal structure with associated polyribosomes obtained
from HeLa cells. Cell 10: 67-78.

Letourneau, P.C. 1975. Cell—to—substratum adhesion and guidance
of axonal elongation. Dev. Biol. 44: 92-101.

Letourneau, P.C., and A.H. Ressler. 1984. Inhibition of neurite
initiation and growth by taxol. J. Cell Biol. 98: 1355-1362.

Lindholm, 0.8., and J.A. Khawaja. 1983. Distribution and protein
synthetic activities of neuronal free and membrane-bound
ribosomes during postnatal development of rat cerebral cortex.
Neurosci. 9: 645-651.

Lyser, K.M. 1964. Early differentiation of motor neuroblasts in
the chick embryo as studied by electron microscopy. I. General
Aspects. Dev. Biol. 10: 433-466.

Lyser, K.M. 1968. Early differentiation of motor neuroblasts in
the chick embryo as studied by electron microscopy. II.
Microtubules and neurofilaments. Devel. Biol. 17: 117-142.

Lyser, K.M. 1968. An electron microscopic study of centrioles in
differentiating neuroblasts. J. Embryol. Exp. Morphol. 20:
343-354.

Margolis, R.L., and L. Wilson. 1978. Opposite end assembly and
disassembly of microtubules at steady state in vitro. Cell 13:
1-8.

Margolis, R.L., and 1" Wilson. 1981. Microtubule treadmills-
possible molecular machinery. Nature 293: 705-711.

McIntosh, J.R. 1983. In: Spatial Organization of Eukaryotic
Cells. J.R. McIntosh, Ed. Alan R. Liss, New York.

McNiven, M.A., M. Wang, and K.R. Porter. 1984. Microtubule

 

 

 

 

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

74

polarity and the direction of pigment transport reverse
simultaneously in surgically severed melanophore arms. Cell 37:
753-765.

Mitchison, T., and M. Kirschner. 1984. Dynamic instability of
microtubule growth. Nature (Lond.) 312: 237-242.

Morris, J.R., and R.J. Lasek. 1982. Stable polymers of the axonal

cytoskeleton: the axoplasmic ghost. J. Cell Biol. 92: 192-198.

Morris, J.R., and R.J. Lasek. 1984. Monomer-polymer equilibria in

the axon: direct measurement of tubulin and actin as polymer and
monomer in axoplasm. J. Cell Biol. 98: 2064-2076.

Nicklas, R.B., D.F. Kubai, and T.S. Hays. 1982. Spindle
microtubules and their mechanical associations in anaphase. J.
Cell Biol. 95: 91-104.

Olmsted, J.B., J.M. Marcum, K.A. Johnson, C. Allen, and G.G.
Borisy. 1974. Microtubule assembly: some possible regulatory
mechanisms. J. Supramol. Struct. 2: 429-450.

Oosawa, F., and S. Asakura. 1975. The Thermodynamics of the
Polymerization of Protein. Academic Press, Inc. New York. 41-47.

Pachter, J.S., R.K.H. Liem, and M.L. Shelanski. 1984. The
neuronal cytoskeleton. Adv. Cell. Neurobiol. 5: 113-142.

Peng, I., L.I. Binder, and M.M. Black. 1986. Biochemical and
immunological analyses of cytochemical domains of neurons. J.
Cell Biol. 102: 252-262.

Peters, A., S.L. Palay, and PL de Webster. 1979. The Fine
Structure of the Nervous System: The Neurons and Supporting
Cells. Saunders, Philadelphia.

Price, D.L, ans Porter, K.R. 1972. The response of ventral horn
neurons to axonal transection. J. Cell Biol. 53: 24-37.

Sahenk, 2., and S. Brady. 1983. Morphological evidence for stable
regions in axonal microtubules. J. Cell Biol. 97: 210a
(abstract).

Schiff, P.B., J. Fant, and 8.8. Horwitz. 1979. Promotion of
microtubule assembly in vitro by taxol. Nature (Lond.) 277:
665-667.

Sharp, G.A., K. Weber, and M. Osborn. 1982. Centriole number and
process formation in established neuroblastoma cells and primary
dorsal root ganglion neurones. Eur. J. Cell Biol. 29: 97-103.

Shaw, 0., and D. Bray. 1977. Movement and extension of isolated
growth cones. Exp. Cell Res. 104: 55-62.

 

 

 

75

 

83. Snyder, J.A., and B.T. Hamilton. 1982. Loss of mitotic
centrosomal microtubule initiation capacity at the
metaphase—anaphase transition. Eur. J. Cell Biol. 27: 191-199.

84. Solomon, F. 1980. Organizing microtubules in cytoplasm. Cell 22:
331-332.

85. Solomon, F. 1981. Specification of cell morphology by endogenous
determinants. J. Cell Biol. 90: 547-553.

86. Solomon, F., M. Magendantz, and A. Salzman. 1979. Identification
with microtubules of one of the microtubule-associated proteins.

Cell 18: 431-438.

87. Spiegelman, B.M., M.A. Lopata, and M.W. Kirschner. 1979. Multiple
sites for the initiation of‘ microtubule assembly in mammalian
cells. Cell 16: 239-252.

88. Steward, O. 1983. Polyribosomes at the base of dendritic spines
of central nervous system neurons: their possible role in synapse
construction and modification. Cold Spring Harb. Symp. Quant.
Biol. 48: 745-759.

89. Tennyson, V.M. 1965. Electron microscopic study of the developing
neuroblast of the dorsal root gangion of the rabbit embryo. J.
Comp. Neurol. 124: 267-318.

90. Tucker, J.B. 1979. Spatial organization of microtubules. In
Microtubules. K. Roberts and J.S. Hyams, editors. Academic
Press, Ltd. London. 315-357.

91. Vlodavsky, I., A. Levi, I. Lax, Z. Fuks, and J. Schlessinger.

1982. Induction of cell attachment and morphological
differentiation in a pheochromocytoma cell line and embryonal
sensory cells by the extracellular matrix. Dev. Biol. 93:
285-300.

92. Webb, 8.0., and L. Wilson. 1980. Cold stable microtubules from
brain. Biochemistry 19: 1991-2001.

93. Wessells, N.K., S.R. Johnson, and R.P. Nuttall. 1978. Axon
initiation and growth cone regeneration in cultured motor
neurons. Exp. Cell. Res. 117: 335-345.

94. Wolosewick, J.J., and K.R. Porter. 1979. Microtrabecular lattice
of the cytoplasmic ground substance. Artifact or reality. J.
Cell Biol. 82: 114-139.

95. Yamada, K.M., B.S. Spooner, and N.K. Wessells. 1970. Axon growth:
role of microfilaments and microtubules. Proc. Natl. Acad. Sci.
USA 66: 1206-1212.

96. Yamada, K.M., B.S. Spooner, and N.K. Wessells. 1971. Ultra—
structure and function of growth cones and axons of cultured

 

76

nerve cells. J. Cell Biol. 49: 614-635.

97. Yamada, K.M., and N.K. Wessells. 1971. Axon elongaton: effect of
nerve growth factor on microtubule proteins. Exp. Cell Res. 66:

98. Zenker, W., and E. Hohberg. 1973. An alpha-nerve fiber: number of
neurotubules in the stem fiber and in the terminal branches. J.
Neurocytol. 2: 143-148.

 

still-