i

AXONOPATHY IN PERIPHERAL MYELIN PROTEIN 22 INSUFFICIENCY

A RESEARCH PAPER

SUBMITTED TO THE GRADUATE SCHOOL

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS

FOR THE DEGREE

MASTERS OF SCIENCE

BY

ATIQ ZAMANI

DR. DERRON BISHOP – ADVISOR

BALL STATE UNIVERSITY

MUNCIE, INDIANA

JULY 2010

ii

Abstract

Title: Axonopathy in Peripheral Myelin Protein 22 Insufficiency.

Student: Atiq Zamani

Degree: Master of Sciences

College: Sciences and Humanities

Date: July 2010

Pages: 33

The role that various myelin membrane proteins play during development and

disease processes is not well understood. To better understand their role in vivo we have

crossed transgenic mice possessing a single truncated pmp22 gene with mice expressing

yellow fluorescent protein in the cytoplasm of their neurons. The resulting double

transgenic mice were examined by a combination of confocal microscopy, transmission

electron microscopy, and immunohistochemistry to determine if pmp22 insufficiency

alters the structural integrity of myelin, glial cells, axons, or the subcellular milieu of

these various components. Axons from mice with pmp22 insufficiency developed

sprouts and debris localized to nodes with no signs of degeneration of a Wallerian type.

Ultrastructurally, the nodes accumulated tubovesicular structures as well as disrupted

cytoskeleton that did not appear to alter axon transport. Together, these results suggest

that pmp22 insufficiency leads to a non-lethal axonopathy that is restricted to nodes.

iii

Acknowledgements

I would like to thank my committee members for allowing me to work with them.

I would like to thank Dr. Derron Bishop for being an excellent thesis advisor. His

knowledge, passion, desire and dedication to research and sciences has helped me

through my thesis project. I have come to learn a great deal in only a short period of time

and I am ever grateful for his advice. I am honored to know and have shared this project

with him and I can say that I am a better researcher today from what he has shown me.

I thank Dr. Susan McDowell for her never ending dedication as a teacher and

mentor. I am grateful for her guidance throughout my career at Ball State University. I

very much appreciate all that she has done for me.

I thank Dr. Heather Bruns for being open and accepting to participate in my thesis

committee. Though I never had the honor of having her as a teacher, I have come to

admire her personality and interaction with students.

I thank Sharmon Knecht for not only being an amiable and wonderful friend, but

for her assistance and support in my research project. It is my pleasure to have known

and worked with her.

I thank my wonderful loving family for all their support and encouragement

throughout my life. I am here today because of the sacrifices they had to make in order

to provide for me. I am forever grateful and thankful.

iv

Table of Contents

Introduction Glial-Axonal Relationship 0-2 Peripheral Myelin Protein 22 – Expression Levels and Neuropathies 2-4 Demyelination and Axon Loss 4-6 Ultrastructural Changes during PMP22-Associated Neuropathies 6-7 Summary 7-7

Materials and Methods Animals 8-8

Confocal Microscopy and Analysis 8-9 Electron Microscopy and Analysis 9-11

Results 12-13

Discussion 14-17

References 18-20

v

LIST OF ACRONYMS/ABBREVIATIONS:

AChR – acetylcholine receptor ALS – amyotrophic lateral sclerosis CIDP – chronic inflammatory demyelinating polyneuropathy CMAP - compound muscle action potential CMT - Charcot-Marie-Tooth disease CMT1A - Charcot-Marie-Tooth disease type 1A CNS – central nervous system DRG – dorsal root ganglion ER - endoplasmic reticulum GBS – Guillain Barre’s syndrome HNPP – hereditary neuropathy with liability to pressure palsies MAG – myelin associated glycoprotein MBP – myelin basic protein MPZ – myelin protein zero MS – multiple sclerosis NCS – nerve conduction study NF – neurofilament NFH - heavy molecular weight neurofilament NFHP – phosphorylated heavy molecular weight neurofilament NFL – low molecular weight neurofilament NFM - middle molecular weight neurofilament NFMP – phosphorylated middle molecular weight neurofilament NMJ – neuromuscular junction PMP22 – peripheral myelin protein-22 PNS – peripheral nervous system PLP - proteolipid protein

vi

List of Figures

Figure 1. Nodal defects in pmp22 +/- axons. Page 21.

Figure 2. A statistically significant difference (p<0.001*) in nodal defects and sprouts of pmp22 +/- and controls. Page 22.

Figure 3. No statistically significant difference (p=0.440) in nodes/m of pmp22 +/- and controls. Page 23.

Figure 4. Demyelination in the pmp22 +/- axon. Page 24.

Figure 5. Subcellular defects in the nodes of the pmp22 +/- axon. Page 25.

Figure 6. Synaptic vesicle distribution in the pmp22 +/- and control axons. Page 26.

Figure 7. No statistically significant difference (p=0.620) in synaptic vesicle density in nerve terminals of pmp22 +/- and controls. Page 27.

Introduction

Glial-Axonal Relationship

Our ability to move, see, hear, smell, taste, and even our cognitive abilities rely

upon the coordinated long distance electric signaling between neurons and their targets.

Although neurons can vary in structure, they generally consist of a cell body (or soma),

dendrites and a single axon. The cell body and dendrites function by receiving

information from other neurons and pass an outgoing signal to the axon. This outgoing

signal is unique to the axon and consists of a rapidly repeating oscillation of the

membrane voltage (potential), called an action potential, which is due to the influx and

efflux of ions thorough selectively permeable membrane channels.

Considering a single axon can be a meter in length, two mechanisms have evolved

to ensure that action potentials are conducted quickly: increasing axonal diameter and

myelin. While increasing the interior diameter of the axon improves conduction velocity

by decreasing the longitudinal resistance, it is not efficient in the mammalian nervous

system because of the large number of neurons. The combination of a large number of

large diameter neurons would necessitate an extremely large nervous system. An

alternative is the development of myelin. Myelin is a lipid rich multilamellar membrane

of axons (Schweigreiter, Roots et al. 2006). In myelinated axons, action potentials are

restricted to non-myelinated segments, known as nodes of Ranvier (Poliak and Peles

2

2003). Current influx at a node can rapidly spread in the myelinated internode segment,

due to the reduced membrane capacitance, resulting in action potential conduction

velocities as fast as 150 meters/sec. The function of myelin is known to be identical in

central and peripheral nervous systems; however, there are differences in the cell biology

of myelination where oligodendrocytes surround as many as fifty axons in the CNS while

Schwann cells associate with only one axon in the PNS (Stevens and Fields 2002).

The glial-axonal interaction is not only essential for speeding action potential

conduction velocity, but also for the survival of the axons (Martini 2001). If the myelin

is disrupted, axon degeneration can often ensue. For instance, demyelination is the

hallmark of many neurological diseases including multiple sclerosis (MS), Guillain-

Barre’s syndrome (GBS), and various peripheral neuropathies. However, how

demyelination leads to axonal degeneration is unknown and there is currently no

treatment to hinder this pathological process. One way to better understand this

relationship is to understand the functional role of the proteins that hold the compact

layers of myelin together.

Peripheral Myelin Protein 22 – Expression Levels and Neuropathies

Although there are many different proteins that hold layers of myelin together,

one of the more abundant is peripheral myelin protein 22 (PMP22). The pmp22 gene

encodes a 22kD tetraspan membrane glycoprotein that resides on the surface of

myelinating Schwann cells localized to compact myelin (Suter and Forscher 1998).

Newly-synthesized PMP22 in the myelinating Schwann cells rarely makes it to the

membrane since >90% of it is degraded in the endoplasmic reticulum (ER) (Pareek,

3

Notterpek et al. 1997). This quality control mechanism appears to be functionally

essential, since certain mutations in PMP22 may cause aberrant trafficking and

accumulation of proteins within intracellular organelles, which has been proposed to be a

pathogenic mechanism of certain neuropathies (Colby, Nicholson et al. 2000). In

addition to simple mutations in PMP22, its expression levels also appear to be critical to

axon function since alterations of pmp22 expression results in clinically distinct

disorders: hereditary neuropathy with liability to pressure palsies (HNPP) and Charcot-

Marie-Tooth disease type 1A (CMT1A). HNPP and CMT1A are among the most

frequently inherited neurological diseases and are caused by deletions or duplication of

chromosome 17p11.2-12 containing pmp22 gene (Lupski and Garcia 1992; Chance and

Pleasure 1993).

HNPP results from insufficient levels of PMP22 and leads to symptoms that

include transient, asymmetric, multifocal sensory motor deficits (Li, Zhang et al. 2004).

These symptoms are usually triggered by physical activities involving compression,

repetitive movement, or stretching suggesting that axons with insufficient amounts of

PMP22 are vulnerable to mechanical forces. In human patients, electrophysiology

demonstrates that compound action potential conduction velocity is indeed slowed at

sites subject to mild mechanical compression (Li and Li 2002).

Charcot-Marie-Tooth disease (CMT) is a group of inherited neuropathies with a

prevalence of one in 2500 people (Shy, Garbern et al. 2002). CMT1A is the most

common form of CMT, affecting one-half of all CMT cases (Ionasescu, Ionasescu et al.

1993; Wise, Garcia et al. 1993). CMT1A is an autosomal dominant neuropathy caused

by a 1.4 Mb duplication of chromosome 17p11.2-12 (Raeymaekers, Timmerman et al.

4

1991). CMT1A patients develop a slowly progressive, symmetrical, demyelinating

neuropathy with secondary axonal loss (Krajewski, Turansky et al. 1999).

Demyelination and Axon Loss

Demyelination is a pathological process that takes place in many neurological

disorders, such as MS, GBS, and chronic inflammatory demyelinating polyneuropathy

(CIDP). Although demyelination leads to a slowed conduction velocity, it often does not

correlate with clinical symptoms. Axon loss, on the other hand, does appear to correlate

with a number of different demyelinating diseases (Krajewski, Lewis et al. 2000;

Bjartmar and Trapp 2003). The cellular and molecular mechanisms responsible for

demyelination mediated axon loss remains unclear. Further, whether or not the axon

degeneration in these instances occurs through previously described mechanisms, such as

Wallerian degeneration, is also unknown.

One way that demyelination could mediate axon loss is through direct

connections to the underlying axonal cell membrane. Myelinated axons are composed of

four discrete compartments: the node of Ranvier, the paranode, the juxtaparanode, and

the internode. Each of these compartments contains a unique, non-overlapping set of

protein constituents (Scherer 2002). The paranodal region, comprised of loops of

Schwann cell membrane that interact with axonal membrane adjacent to the node of

Ranvier, contain Schwann cell proteins myelin associated glycoprotein (MAG),

Connexin 32 (Cx32), and axonal proteins Caspr and Contactin. These proteins participate

in Schwann cell-axonal interaction and define the nodal region. The juxtaparanodal

region, the portion of the Schwann cell and the interacting axonal membrane adjacent to

5

the paranode, contains voltage gated potassium channels that are expressed on the

axolemma. The internodes, the remaining segment flanked by two nodes of Ranvier,

contain the myelin structural proteins, myelin protein zero (MPZ), PMP22, and myelin

basic protein (MBP), which participate in forming the tightly-compacted myelin sheath.

When nerve fibers are demyelinated, a number of these proteins become redistributed.

This redistribution may contribute to axonal degeneration. For example, conspicuous

reorganization of molecular architecture has been found in Cx32 and pmp22 knockout

mice, the two chronic demyelination animal models (Neuberg, Sancho et al. 1999).

A second possible mechanism that could explain demyelination mediated axon

loss is through cytoskeletal changes in the axon. Myelination affects the phosphorylation

of axonal neurofilaments (NF). NFs are the principal element of the axonal cytoskeleton

and are composed of three subunits, referred to as low (NFL), medium (NFM), and heavy

(NFH) molecular weight NF proteins (Grant and Pant 2000). They all share a similar

protein structure that consists of an N-terminal head domain, a α-helical central domain,

and a C-terminal tail domain. The C-terminal tail domain forms sidearms that protrude

from the NF backbone and are thought to regulate both the rate of axonal transport and

the caliber of axons. The best evidence for this comes from experiments in Trembler

mice that have a chronic demyelinating neuropathy. These mice have decreased NF

phosphorylation, increased NF density, along with decreased axonal transport (de Waegh,

Lee et al. 1992). Transplantation of a segment of a Trembler nerve into a normal nerve

also produces similar changes in axons that have regenerated through the nerve graft but

not in the surrounding nerve, demonstrating that this effect is induced by contact with the

6

abnormal Schwann cells (Sahenk, Chen et al. 1999). Together, these results argue that

glial contact can alter the underlying axonal cytoskeleton.

Finally, abnormalities with the Schwann cell itself may lead to axon degeneration

independent of demyelination. Such uncoupling of axonal loss from segmental

demyelination has been previously observed in several disorders. Human patients with

proteolipid protein 1 (PLP1) deletion develop axonal loss without demyelination as well

as PLP1 knockout mice (Griffiths, Klugmann et al. 1998; Garbern, Yool et al. 2002).

Transgenic mice in which the glial-specific protein, 2’,3’-cyclic nucleotide

phosphodiesterase (CNP1) was deleted also develop axonal degeneration without

demyelination (Lappe-Siefke, Goebbels et al. 2003). Mice with deficient MAG develop

a late-onset axonal degeneration with well preserved myelin (Hanemann, Gabreels-Festen

et al. 2001).

Ultrastructural changes during PMP22-Associated Neuropathies

Perhaps not surprising is the notion that alternations in PMP22 levels can lead to

disruption of myelin structure. For instance, focal hypermyelination in HNPP was first

described by Behse and colleagues (Behse, Buchthal et al. 1972). Madrid and Bradley

later described tomaculated myelin in various neuropathies due to its sausage shape

(Bradley, Madrid et al. 1975). This pathological phenomenon was subsequently found in

many other neuropathies, including anti-myelin associated glycoprotein (MAG)

paraproteinemic neuropathy, CMT1B (due to P0 gene mutations), CMT4B (due to

myotubularin-related 2 gene mutation), along with several animal models of peripheral

nerve disorders (Sander, Ouvrier et al. 2000). In general, tomacula are distributed mainly

7

in the paranodal regions and occasionally in the internodal regions as a result of

redundant myelin infoldings and outfoldings. Myelin in the tomacula appears to have

reduced stability and tend to degenerate over time (Adlkofer, Naef et al. 1997).

Whether or not tomacula confer toxicity to the underlying axon remains

unanswered. Axons encased by tomacula frequently appear constricted. Sometimes, this

constriction can be so severe that axons can become difficult to discern at the

ultrastructural level (Bradley, Madrid et al. 1975). Studies in MAG knockout mice

suggest that redundant myelin folding in tomacula is a result of axonal shrinkage based

on the observation of overall reduction of axonal diameters in mag-/- mice (Yin,

Crawford et al. 1998). However, tomacula were subsequently found in the early life of

these mice (1-month-old) when axonal diameters had not yet been reduced (Cai, Sutton-

Smith et al. 2002). Together, it is unclear how axons lose caliber and whether or not

dysmyelination is its cause or consequence.

Summary

The role that various myelin proteins play in regulating the glial-axonal

relationship remains unclear. Therefore, the primary focus of this work is to examine

axons from mice with insufficient levels of PMP22 with the highest spatial resolution

possible to determine if this reduction results in a structurally distinct axonopathy.

Materials and Methods

Animals. All animals were provided by Dr. Jun Li from Vanderbilt University School of

Medicine. All animal handling was approved by the Vanderbilt University School of

Medicine IACUC. Briefly, pmp22 was inactivated through homologous recombination

(Adlkofer, Martini et al. 1995). Pmp22+/- mice were crossed with yfp+/+ mice that

expressed yellow fluorescent protein in the cytoplasm of their neurons under the

influence of the thy-1 promoter. After deep anesthesia, double transgenic mice were

transcardially perfused with 4.0% paraformaldehyde. Both left and right gluteus

maximus, soleus, omoyhoid, and sternomastoid muscles were carefully dissected from

the animals. One of the paired muscles was postfixed in 4.0% paraformaldehyde while

its corresponding contralateral limb was postfixed in 2.5% paraformaldehyde with 4%

glutaraldehyde. The former muscles were used for confocal microscopy while the latter

muscles were used for electron microscopy.

Confocal Microscopy and Analysis. Muscles fixed in 4.0% paraformaldehyde were

placed in 35mm petri dishes (Corning Glass Works, Coming, NY) containing Sylgard

Sillicon elastomer (Dow Coming Corporation, Midland, MI) are pinned with fine

minutien pins (Fine Science Tools, Foster City, CA) for stabilization during confocal

9

imaging and photoconverstion. First the tissue are cleaned of any debris and then washed

with 3 changes of 0.1 M phosphate buffer saline (PBS). Acetylcholine receptors were

labeled with -bungarotoxin conjugated to Alexa-594 for one hour and then washed with

3 changes of 0.1 M PBS buffer. Muscles were transferred to a slide place in glycerol

with 1.0% DABCO and sealed with #1.5 coverslips. YFP was imaged using a 488nm

laser line and detected with a 505-530nm band pass filter. Alexa 543nm laser line and

detected with a 560nm long pass filter. All images were collected with a 40x, 1.2NA

objective lens and sampled at Nyquist limits. Image stacks were projected to 2D using a

maximum pixel projection algorithm.

Preterminal axon length was determined by measuring the distance between the

last heminode to as far as the axon could be traced proximally. The proximal

measurements were always stopped at a node, so as to get a more accurate measurement

of internode distance (see below). We avoided axons that traveled mostly in the z-

dimension to get closer to Euclidean measurements of distance. Nodal sprouts were

counted when a branch emanated from a node that did not innervate acetylcholine

receptors. Small spheres of fluorescent debris that were not connected to an axon were

counted. Nodal sprouts and debris were pooled and counted as nodal defects. Internode

distance was determined by dividing preterminal distance by the number of nodes. Nodal

defects and internode distance were compared between the experimental and control

animals for statistical significance using a student’s t-test.

Electron microscopy and Analysis. Muscles destined for electron microscopy were first

analyzed by confocal microscopy. After confocal microscopy, areas of interest were

10

targeted for serial section transmission electron microscopy by placing fiducial markers

in adjacent muscle fibers. Crystals of 1, 1-dioctadecyl-3,3,3,3-

tetramethylindodicarbocyanine-5,5-disulfonic acid (Dil, Invitrogen, Carlsband, CA) were

applied to surface of the selected muscle of the axon or NMJ of interest using

iontophoresis. Briefly, a micropipette puller (Kopf, Tujunga, California) was used to pull

a micropipette (World Precision Instrument, Sarasota, FL) to a resistance of 5-10 MΩ. A

1.0% solution of Dil in 100% dicholormethylene (Sigma-Aldrich, St. Louis, MO) was

loaded into the micropipette. The pipette was next threaded with a silver chloride

electrode to an SD9 stimulator (Grass Instruments, Quincy, MA). The microelectrode

was inserted into the muscle near the axon and DiI was unloaded with an electrical pulses

of 10v. To render the Dil crystal electron dense, DiI was illuminated near its excitation

peak with a 40x water-immersion lens (Ziess) in the presence of 3, 3-Diaminobenzidine

(DAB, 5.0mg/ml, Sigma-Aldrich).

After photoconversion, the tissue of interest was then stained with 1.0% osmium

tetroxide (Electron Microscopy Sciences, Hatfield, PA) reduced in 1.5% potassium

ferrocyanide ( Sigma-Aldrich) for thirty minutes. Tissue was then washed with 4

changes of 0.1 M sodium cacodylate buffer (Electron Microscopy Sciences) and

dehydrated in an ascending ethanol series(25, 50, 75 and 100% ethanol) (AAPER

Alcohol and Chemical Company, Shelbyville, KY). After dehydration steps, the tissue

block was place in a final 2 changes of propylene oxide (Electron Microscopy Sciences)

and infiltrated with Araldite 502/Embed812 resin and polymerized at 60 degrees C for 48

hours (Electron Microscopy Sciences). Next after 48 hours the polymerized tissue block

was trimmed to appropriate size into a trapezoid shape and thick sectioned (0.5-1.0 m)

11

with a glass knife on a Leica Ultracut R microtome (Leica, Wetzlar, Germnay). Thick

sections were carefully examined individually with an Axiovert 35M light microscope

(Zeiss) to locate the fiducial markers. Once the markers were located the block was then

trimmed down into a smaller trapezoid where only the area the tissue and area of interest

was. The tissue was then sectioned (60-80nm) into hundreds of serial sections containing

the axon or NMJ of interest. The sections were cut with a diamond knife (Diatome) and a

Leica UC6 ultramicrotome (Leica). Thin sections were then picked up on 1.2% pioloform

(Ted Pella, Inc. Redding, CA, in choloroform, Sigma-Aldrich) coated slot grids (Electron

Microscopy Sciences). The grids were counterstained with 2.0% aqueous uranyl acetate

and Reynold’s lead citrate. Sections were viewed at accelerating voltage of 75kV on a

Hitachi H-600 TEM. Electron micrographs were captured at 5000X and scanned at a

resolution of 1200dpi with an Epson Perfection V700 Photo Scanner.

Serial electron micrographs were imported into the Reconstruct software

developed by Kristen Harris and John Fiala (http://synapse-web.org/tools/index.stm).

Micrographs were aligned to each other and then manually traced to provide us with a 3D

model of our area of interest. Section thickness was determined using the cylindrical

diameters method (Fiala and Harris 2001). Nerve terminal volume was determined from

the manual traces and section thickness. Synaptic vesicles were counted and normalized

using the Ambercrombie method for cell counting. A student’s t-test was used to

determine if synaptic vesicle density was statistically significant between the

experimental and control animals.

Results

We first crossed yfp16 +/+ mice with pmp22+/- mice to create double transgenic

mice that express fluorescent protein in the cytoplasm of their neurons along with a

deficiency in pmp22 in their peripheral nerves. We then used confocal microscopy to

examine peripheral nerves and neuromuscular junctions in these mice and fluorescent

controls. Confocal microscopy revealed no evidence of denervated or partially

innervated muscle fibers suggesting pmp22 insufficiency is not immediately lethal to

axons. However, we noticed in pmp22+/- mice that peripheral axons had a number of

small branches that did not innervate muscle fibers (Figure 1). These sprouts emanated

from nodes of Ranvier where myelin does not cover the axon. In other cases, we did not

notice sprouts, but rather found accumulations of fluorescent debris near the nodes

(Figure 1).

We compared the total number of nodal defects (i.e., nodal sprouts and debris) in

pmp22+/- and control mice. We found a statistically significant difference (p<0.001*) in

nodal defects between that pmp22+/- axons (0.021 +/- 0.014 nodal defects/m) compared

to controls (0.006 +/- 0.008nodal defects/m) (see Figure 2). To determine if this result

was simply due to a difference in the number of nodes per micron in pmp22+/- axons

compared to control, we determined the average internode distance. There was no

statistically significant difference in internode distance between pmp22+/- axons (36.0 +/-

13

17.1 m) compared to controls (38.8 +/- 16.1 m) (see Figure 3). Together, these results

suggest pmp22+/- axons may have a vulnerability restricted to their nodes.

To better understand what pathology may be underlying the nodal defect, we took

advantage of the far greater resolution of the transmission electron microscope to

examine the nodal and paranodal regions in pmp22+/- axons. In the paranodal region, we

noticed layers of compact myelin that surrounded the axon had begun to separate (Figure

4). This demyelination occurred on both the outer and inner layers suggesting the entire

compact myelin had been compromised. At the node, we noticed an accumulation of

tubulovesicular structures that we were unable to find in controls (Figure 5). Further, the

cytoskeleton in regions of demyelination had lost its orderly parallel structure axial to the

membrane. Since the cytoskeleton is essential for both anterograde and retrograde

transport of organelles in axons, such a localized disruption could interfere with organelle

delivery to axon terminals.

We decided to count the number of synaptic vesicles in serial electron

micrographs of nerve terminals to see if there are fewer vesicles in potentially axon

transport impaired pmp22+/-mice (Figure 6). We examined eight nerve terminals from

three pmp22+/- and four nerve terminals from two control mice. In some nerve terminals,

we found evidence of reduced vesicle density in pmp22+/-. However, this difference was

not statistically significant different between the pmp22+/- compared to controls ( =

522.3 verses 617.3 vesicles/m3, respectively) (Figure 7).

Discussion

Our results demonstrate that pmp22+/- axons exhibit nodal sprouting and debris at

one month of age. This difference was not simply due to an increased number of nodes

since the average internode distance was not significantly different between pmp22+/-

axons and controls. We were interested in determining what consequences, if any, these

defects might have upon the axon and nerve terminals. To this end, we took advantage of

transmission electron microscopy to examine the subcellular milieu of pmp22+/- axons.

At the paranodal regions of these axons we found evidence of

dysmyelination/demyelination in compact myelin layers that surround the axon.

We also noticed that the cytoskeleton seemed to have lost its orderly parallel

structure. The function and survival of neurons are dependent upon axonal transport,

where newly synthesized proteins and organelles from the cell body located in the spinal

cord are transported long distances along the axon. Cytoskeleton disruption, which has

been known to impair axon transport in disease models such as the fALS model of

amyotrophic lateral sclerosis (Dutta and Trapp 2006), can lead to a decrease in trophic

support as well as decreased organelle density in synaptic terminals. Despite disruption

of the cytoskeleton, we were unable to find a statistically significant difference in vesicle

density between pmp22+/- axons compared to controls.

15

Our data suggests that even though there are nodal defects coupled with

demyelination and cytoskeleton rearrangements, these changes do not lead to a Wallerian

type degeneration resulting in denervation or partially innervated muscle fibers. In such a

case, we would have expected that the membrane be severed at some point causing the

distal axon to fragment, become engulfed by Schwann cells, and finally disappear

entirely. That we were unable to find evidence of this Wallerian type of degeneration

underscores the idea that pmp22 insufficiency does not appear to lead to a denervating

axonopathy.

Instead, pmp22+/- insufficiency appears to result in an abnormality restricted to the

nodes. In a normal myelinated axon, voltage gated sodium channels are highly

concentrated at the nodes (Kazarinova-Noyes and Shrager 2002), resulting in ionic flux

and action potential generation from node to node. The efficiency of action potential

propagation, therefore, is dependent on the both myelin and nodal integrity. Any changes

or damages in the myelin or node can alter the delicate balance of sodium channel density

and/or local metabolism. For instance, in demyelinating diseases such as MS, myelin

damage often results in the upregulation of sodium channels (Stys, Waxman et al. 1992).

The increasing demand for ATP exceeds the production capabilities of existing

mitochondria. Na+/K+ ATPase pumps increase their activity in an attempt to maintain

ionic gradients placing a further increase on the metabolic demands of oxidative

metabolism. As a result, the localized accumulations of intracellular Na+ ions result in

excess Ca2+ ions entering the cell through the Na+/Ca2+ exchanger resulting in a Ca2+-

mediated axon damage along the axons. Other genetic manipulations have similarly lead

to axon loss possibly through other cellular mechanisms. For instance, overexpressing

16

P0, a major structural myelin protein (Previtali, Quattrini et al. 2000), leads not only to

demyelination in the peripheral nervous system in mice, but also axon terminal loss

within six months (Bjartmar, Yin et al. 1999).

Observations of both MS and P0 overexpression lead to the intriguing possibility

that nearly any manipulation to myelin can result in axon loss. However pmp22

insufficiency appears to progress differently than both MS and overexpression of P0. In

MS, where autoimmune loss of myelin over the long term results in large scale

denervation of CNS tracts, sclerotic areas of former myelinated axons serve as the

hallmark of the disease. Considering we were unable to confirm axon loss in up to four-

month old pmp22+/- animals, it appears any compromise induced by pmp22+/-

insufficiency does not follow a similar course to MS or P0 overexpression. Our results

strengthen the notion that damage to myelin does not universally lead to axon loss.

One reason for differences in MS and pmp22+/- insufficiency may simply be

related to differences between the glial cell type and the relationship they share with

central and peripheral axons. Oligodendrocytes in the central nervous system myelinate

many axons compared to Schwann cells which provide surround only one axon. It is this

insulating and branching of oligodendrocytes to other axons that may confer its

debilitating properties. A second factor may be related to the immune response generated

in MS compared to pmp22+/- mice. We found no evidence of immune cell infiltration

into the peripheral nerves of pmp22+/- mice indicating an immune response was not

generated despite the cellular debris. This is in sharp contrast to MS where immune cell

infiltration is a hallmark of the disease. Perhaps the immune cells confer the ultimate

damage to axons that is not conferred in pmp22+/- insufficiency.

17

During the early phases of MS, a relapsing-remitting phase is characterized by

slowed and even stopped action potential conduction. Pmp22+/- mice have recently been

shown to suffer conduction block following short term mechanical occlusion (Bai, Zhang

et al. 2010). Although Na+ channel density did not change in these studies, conduction

block can induced significantly quicker in pmp22+/- animals compared to controls. This

effect could be due to biophysical properties of the axons since these animals also

showed contrictions in the paranodal regions of the sciatic nerve. Our studies could not

confirm these severe constrictions in the distal axons of leg muscles perhaps suggesting

the insufficiency primarily attacks the more proximal large caliber parts of the axon as

compared to the smaller more distal terminations. This is in contrast to other disease

models, such as fALS, that have been previously described as distal axonopathies that

tend to progress more proximally over time (Fischer, Culver et al. 2004).

Transgenic pmp22 knockout mice have been an accepted model of Charcot-

Marie-Tooth disease where dysmyelination/demyelination often leads to axonal changes.

Tomacula are well-described structures in these mice that we did not encounter in our

study. Although recent reports (Bai, Zhang et al. 2010), show these structures in

proximal portions of sciatic nerves, they are not nearly as abundant as in pmp22-/- mice.

The ultrastructure of tomacula in pmp22+/-in these more proximal regions was not

explored as in the detail of our study. An examination of more proximal portions of

sciatic nerves in both pmp22+/- and pmp22-/- mice might better elucidate the role this

protein plays in the glial axon relationship in the peripheral nervous system.

References

Adlkofer, K., R. Martini, et al. (1995). Hypermyelination and demyelinating peripheral neuropathy in Pmp22-deficient mice. Nat Genet 11(3): 274-80.

Adlkofer, K., R. Naef, et al. (1997). Analysis of compound heterozygous mice reveals that the Trembler mutation can behave as a gain-of-function allele. J Neurosci Res 49(6): 671-80.

Bai, Y., X. Zhang, et al. (2010). Conduction block in PMP22 deficiency. J Neurosci 30(2): 600-8.

Behse, F., F. Buchthal, et al. (1972). Hereditary neuropathy with liability to pressure palsies. Electrophysiological and histopathological aspects. Brain 95(4): 777-94.

Bernd, P. (2008). The role of neurotrophins during early development. Gene Expr 14(4): 241-50.

Bjartmar, C. and B. D. Trapp (2003). Axonal degeneration and progressive neurologic disability in multiple sclerosis. Neurotox Res 5(1-2): 157-64.

Bjartmar, C., X. Yin, et al. (1999). Axonal pathology in myelin disorders. J Neurocytol 28(4-5): 383-95.

Bradley, W. G., R. Madrid, et al. (1975). Recurrent brachial plexus neuropathy. Brain 98(3): 381-98.

Cai, Z., P. Sutton-Smith, et al. (2002). Tomacula in MAG-deficient mice. J Peripher Nerv Syst 7(3): 181-9.

Chance, P. F. and D. Pleasure (1993). Charcot-Marie-Tooth syndrome. Arch Neurol 50(11): 1180-4.

Colby, J., R. Nicholson, et al. (2000). PMP22 carrying the trembler or trembler-J mutation is intracellularly retained in myelinating Schwann cells. Neurobiol Dis 7(6 Pt B): 561-73.

de Waegh, S. M., V. M. Lee, et al. (1992). Local modulation of neurofilament phosphorylation, axonal caliber, and slow axonal transport by myelinating Schwann cells. Cell 68(3): 451-63.

Dutta, R. and B. D. Trapp (2006). [Pathology and definition of multiple sclerosis]. Rev Prat 56(12): 1293-8.

Fiala, J. C. and K. M. Harris (2001). Cylindrical diameters method for calibrating section thickness in serial electron microscopy. J Microsc 202(Pt 3): 468-72.

Fischer, L. R., D. G. Culver, et al. (2004). Amyotrophic lateral sclerosis is a distal axonopathy: evidence in mice and man. Exp Neurol 185(2): 232-40.

Garbern, J. Y., D. A. Yool, et al. (2002). Patients lacking the major CNS myelin protein, proteolipid protein 1, develop length-dependent axonal degeneration in the absence of demyelination and inflammation. Brain 125(Pt 3): 551-61.

19

Grant, P. and H. C. Pant (2000). Neurofilament protein synthesis and phosphorylation. J Neurocytol 29(11-12): 843-72.

Griffiths, I., M. Klugmann, et al. (1998). Axonal swellings and degeneration in mice lacking the major proteolipid of myelin. Science 280(5369): 1610-3.

Hanemann, C. O., A. A. Gabreels-Festen, et al. (2001). Axon damage in CMT due to mutation in myelin protein P0. Neuromuscul Disord 11(8): 753-6.

Ionasescu, V. V., R. Ionasescu, et al. (1993). Charcot-Marie-Tooth neuropathy type 1A with both duplication and non-duplication. Hum Mol Genet 2(4): 405-10.

Kazarinova-Noyes, K. and P. Shrager (2002). Molecular constituents of the node of Ranvier. Mol Neurobiol 26(2-3): 167-82.

Krajewski, K., C. Turansky, et al. (1999). Correlation between weakness and axonal loss in patients with CMT1A. Ann N Y Acad Sci 883: 490-2.

Krajewski, K. M., R. A. Lewis, et al. (2000). Neurological dysfunction and axonal degeneration in Charcot-Marie-Tooth disease type 1A. Brain 123 ( Pt 7): 1516-27.

Lappe-Siefke, C., S. Goebbels, et al. (2003). Disruption of Cnp1 uncouples oligodendroglial functions in axonal support and myelination. Nat Genet 33(3): 366-74.

Li, A. H., J. M. Zhang, et al. (2004). Human acupuncture points mapped in rats are associated with excitable muscle/skin-nerve complexes with enriched nerve endings. Brain Res 1012(1-2): 154-9.

Li, S. and Y. Li (2002). [Neurology]. Zhonghua Yi Xue Za Zhi 82(24): 1669-71. Lupski, J. R. and C. A. Garcia (1992). Molecular genetics and neuropathology of

Charcot-Marie-Tooth disease type 1A. Brain Pathol 2(4): 337-49. Martini, R. (2001). The effect of myelinating Schwann cells on axons. Muscle Nerve

24(4): 456-66. Neuberg, D. H., S. Sancho, et al. (1999). Altered molecular architecture of peripheral

nerves in mice lacking the peripheral myelin protein 22 or connexin32. J Neurosci Res 58(5): 612-23.

Pareek, S., L. Notterpek, et al. (1997). Neurons promote the translocation of peripheral myelin protein 22 into myelin. J Neurosci 17(20): 7754-62.

Poliak, S. and E. Peles (2003). The local differentiation of myelinated axons at nodes of Ranvier. Nat Rev Neurosci 4(12): 968-80.

Previtali, S. C., A. Quattrini, et al. (2000). Epitope-tagged P(0) glycoprotein causes Charcot-Marie-Tooth-like neuropathy in transgenic mice. J Cell Biol 151(5): 1035-46.

Raeymaekers, P., V. Timmerman, et al. (1991). Duplication in chromosome 17p11.2 in Charcot-Marie-Tooth neuropathy type 1a (CMT 1a). The HMSN Collaborative Research Group. Neuromuscul Disord 1(2): 93-7.

Sahenk, Z., L. Chen, et al. (1999). Effects of PMP22 duplication and deletions on the axonal cytoskeleton. Ann Neurol 45(1): 16-24.

Sander, S., R. A. Ouvrier, et al. (2000). Clinical syndromes associated with tomacula or myelin swellings in sural nerve biopsies. J Neurol Neurosurg Psychiatry 68(4): 483-8.

Scherer, S. S. (2002). Myelination: some receptors required. J Cell Biol 156(1): 13-5.

20

Schweigreiter, R., B. I. Roots, et al. (2006). Understanding myelination through studying its evolution. Int Rev Neurobiol 73: 219-73.

Shy, M. E., J. Y. Garbern, et al. (2002). Hereditary motor and sensory neuropathies: a biological perspective. Lancet Neurol 1(2): 110-8.

Stevens, B. and R. D. Fields (2002). Regulation of the cell cycle in normal and pathological glia. Neuroscientist 8(2): 93-7.

Stys, P. K., S. G. Waxman, et al. (1992). Ionic mechanisms of anoxic injury in mammalian CNS white matter: role of Na+ channels and Na(+)-Ca2+ exchanger. J Neurosci 12(2): 430-9.

Suter, D. M. and P. Forscher (1998). An emerging link between cytoskeletal dynamics and cell adhesion molecules in growth cone guidance. Curr Opin Neurobiol 8(1): 106-16.

Wise, C. A., C. A. Garcia, et al. (1993). Molecular analyses of unrelated Charcot-Marie-Tooth (CMT) disease patients suggest a high frequency of the CMTIA duplication. Am J Hum Genet 53(4): 853-63.

Yin, X., T. O. Crawford, et al. (1998). Myelin-associated glycoprotein is a myelin signal that modulates the caliber of myelinated axons. J Neurosci 18(6): 1953-62.

21

Fig

ure

1.

Nod

al d

efec

ts in

pm

p22

+/-

axon

s. P

anel

(A

) sh

ows

a co

nfoc

al p

roje

ctio

n fr

om a

pm

p22

+/-

mus

cle

whe

re a

xons

app

ear

gree

n, a

nd a

cety

lcho

line

rec

epto

rs a

ppea

r re

d. H

ighe

r m

agni

fica

tion

im

ages

indi

cate

d by

the

boxe

d re

gion

s ap

pear

in

pane

ls C

, D a

nd E

. P

anel

(B

) sh

ows

a co

nfoc

al

proj

ecti

on f

rom

a c

ontr

ol a

nim

al. P

anel

s C

, D a

nd E

sho

w e

vide

nce

of n

odal

spr

outs

and

deb

ris

(arr

ows)

. S

cale

bar

= 2

0.0 m

in p

anel

s A

and

B, a

nd 5

.0

m in

pan

el C

, D a

nd E

.

22

pmp22(+/-)pmp22 (+/+)

0.05

0.04

0.03

0.02

0.01

0.00

Noda

l Def

ects

/um

Figure 2. A statistically significant difference (p<0.001*) in nodal defects and sprouts of pmp22 +/- and controls. Nodal defects and sprouts in pmp22 +/- and controls were

observed using confocal microscopy. A mean value of 0.006 +/- 0.008 SD defects/m,

and 0.021 +/- 0.014 SD defects/m was observed for pmp22 +/- and controls respectively. Data was analyzed using a Student’s t-test.

*

23

pmp22 (+/-)pmp22 (+/+)

80

70

60

50

40

30

20

10

Inte

rnod

e Le

ngth

(um

)

Figure 3. No statistically significant difference (p=0.440) in nodes/m of pmp22 +/- and controls. Nodal distance per micron in pmp22 +/- and controls was measured using

confocal microscopy. A mean value of 36.0 +/- 17.1 SDm, and 38.8 +/- 16.1 SD m was measured for pmp22 +/- and controls respectively. Data was analyzed using a Student’s t-test.

24

Fig

ure

4.

Dem

yeli

nat

ion

in t

he

pmp2

2 +

/- ax

on.

The

top

left

pan

el s

how

s a

conf

ocal

pro

ject

ion

from

a p

mp2

2 +

/- ax

on (

gree

n) th

at

form

s a

syna

pse

over

ace

tylc

holi

ne r

ecep

tors

(re

d).

A s

urfa

ce r

ende

ring

fro

m s

eria

l sec

tion

s fr

om th

e bo

xed

regi

on is

exp

ande

d be

low

. A

sin

gle

elec

tron

mic

rogr

aph

has

been

inse

rted

into

the

rend

erin

g at

its

exac

t loc

atio

n. T

he e

lect

ron

mic

rogr

aph

is e

xpan

ded

in th

e ri

ght p

anel

. A

rrow

s sh

ow r

egio

ns o

f de

mye

lina

tion

. S

cale

bar

= 1

.0

m.

25

Fig

ure

5.

Su

bce

llu

lar

def

ects

in t

he

nod

es o

f th

e pm

p22

+/-

axon

. T

he to

p le

ft p

anel

sho

ws

a co

nfoc

al p

roje

ctio

n fr

om a

pm

p22

+/-

axon

(gr

een)

that

for

ms

a sy

naps

e ov

er a

cety

lcho

line

rec

epto

rs (

red)

. A

sur

face

ren

deri

ng f

rom

ser

ial s

ecti

ons

from

th

e bo

xed

regi

on is

exp

ande

d be

low

. A

sin

gle

elec

tron

mic

rogr

aph

has

been

inse

rted

into

the

rend

erin

g at

its

exac

t loc

atio

n. T

he

elec

tron

mic

rogr

aph

is e

xpan

ded

in th

e ri

ght p

anel

and

sho

ws

tubu

love

sicu

lar

stru

ctur

es w

ithi

n th

e un

mye

lina

ted

node

. T

he

cyto

skel

eton

thro

ugh

this

reg

ion

is d

isru

pted

as

wel

l. S

cale

bar

= 1

.0

m.

26

Fig

ure

6.

Syn

apti

c ve

sicl

e d

istr

ibut

ion

in t

he

pmp2

2 +

/- an

d c

ontr

ol a

xon

s. P

anel

(A

) sh

ows

an e

lect

ron

mic

rogr

aph

from

a p

mp2

2 +

/- ax

on th

roug

h a

sing

le n

erve

term

inal

. P

anel

(B

) sh

ows

an e

lect

ron

mic

rogr

aph

from

a c

ontr

ol ax

on th

roug

h a

nerv

e te

rmin

al.

Sca

le b

ar =

1.0

m

.

27

pmp22 (+/-)pmp22 (+/+)

1100

1000

900

800

700

600

500

400

300

200

Vesi

cles

/um

3

Figure 7. No statistically significant difference (p=0.620) in synaptic vesicle density in nerve terminals of pmp22 +/- and controls. Vesicle density of nerve terminals in pmp22+/- and controls were counted from serial section transmission electron micrographs. A mean value of 522 +/- 228 SDvesicles/m3, and 617 +/- 430 SD

vesicles/m3 was counted for pmp22+/-and controls respectively. Data was analyzed using a Student’s t-test.