Development of an Automated System for Ex Vivo Measuring the

Neuro Muscular Junction Functionality

Simona Pisu

, Emanuele Rizzuto

, Antonio Musarò

and Zaccaria Del Prete

DAHFMO-Unit of Histology and Medical Embryology, Sapienza University of Rome, Rome, Italy

Department of Mechanical and Aerospace Engineering, Sapienza University of Rome, Rome, Italy

1 RESEARCH PROBLEM

The present work is part of a research project, which

aims to characterize the nerve-muscle interaction in

Amyotrophic Lateral Sclerosis (ALS), a fatal

neuromuscular disease associated with motor neuron

degeneration, muscle atrophy and paralysis. The loss

of connection between muscle and nerve is a crucial

biological mechanism severely impaired in ALS. In

this context, to investigate the alterations in the

communication between muscle fibers and motor

neurons, we started studying the neuro muscular

junction (NMJ) from a functional point of view. The

capability of measuring the NMJ functionality can

therefore give essential information on its physio-

pathological conditions. Therefore, this study may

be useful to discriminate damages of the different

motor unit components in neuromuscular diseases

and, in long-term, to design more appropriate

therapeutic approaches.

In particular, the result that we expect to achieve

is to better clarify which element of the motor unit is

initially affected by this disease. In fact, the recently

proposed “dying-back” hypothesis supports the idea

that the damages begin in the NMJ and then spread

towards the motor neuron body, in opposition to the

more traditional idea according to which the muscle

is only a secondary target of the disease.

2 OUTLINE OF OBJECTIVES

In the described context, the aim of my PhD project

is to characterize the functionality of the

communication between muscle and nerve in

pathological mouse models, by developing new

automated experimental methodologies.

The measurement of NMJ functionality is

obtained by comparing muscle contractile response

elicited by nerve stimulation (indirect), with the

response of the same specimen to membrane

stimulation (direct). Since this latter stimulation

bypasses the neuronal signalling, any difference

between the two contractile responses may be

related to NMJ alterations. To date, I started

working with Soleus muscle-nerve specimens of

healthy Wild Type mice, to develop an experimental

system for studying NMJ functionality of ex vivo

muscle-nerve preparations. After that I’m

approaching the study of a ALS mouse model. In the

future years, I will approach the realization of new

experimental systems for testing NMJ functionality

in isotonic conditions and investigate directly in vivo

the muscle behavior. In all the systems We aim to

develop we will pay special attention to the accuracy

and the repeatability of the experimental procedures.

These metrological qualities will counteract the

negative effects due to the high variability that

usually arises when working with biological tissues.

3 STATE OF THE ART

To better define the alterations in the coupling

between motor neuron conduction and muscle

contraction we started evaluating the physio-

pathologic properties of both skeletal muscle and

NMJ, by stimulating the muscle directly and

indirectly. In particular, we studied standard muscle

contractile properties, as previously described by us

and by other research groups (Brooks SV

and

Faulkner JA, 1988; Del Prete et al, 2008). On the

other hand, although the NMJ functionality

technique has been extensively used on rats (Aldrich

et al., 1986; Van Lunteren and Moyer, 2004), only a

few works have been attempted on pathological

mouse models: Personius et al. measured the

diaphragm NMJ functionality of adult dystrophic

mdx mice (Personius and Sawyer, 2006), while Lee

et al. (Lee et al., 2011) and Ling et al. (Ling et al.,

2009) measured the NMJ properties respectively in

Pisu S., Rizzuto E., Musarò A. and Del Prete Z..

Development of an Automated System for Ex Vivo Measuring the Neuro Muscular Junction Functionality.

 2015 SCITEPRESS (Science and Technology Publications, Lda.)

Soleus and EDL of an animal model of spinal

muscle atrophy at a few days after birth. This

literature has been taken as a starting point, but these

studies are based on animal models of different ages

and sizes or different muscles, if compared with

ours. For this reason it was not possible to employ

the stimulation parameters as proposed.

The pathological model we decided to study at

first is the SOD1

G93A

mice (Gurney et al., 1994),

one of the most studied animal model of ALS

(Turner and Talbot, 2008). However, its skeletal

muscles contractile properties have been poorly

investigated and the NMJ functionality at all.

Previous studies reported a deficit in the generation

of maximum force in hind limbs muscles of

SOD1

G93A

mice, when compared to wild type

littermates (Hegedus et al., 2008; Derave et al.,

2003). Some authors detected a reduced number of

motor neurons associated with disruption of the

neuromuscular junction (Ngo et al., 2012; Fischer et

al, 2003).

Temporal analysis of axon and NMJ

degeneration in transgenic mice indicate that

motorneuron pathology begins distally from the

synaptic area, and then proceeds towards soma in a

retrograde dying back manner

(Luc Dupuis and Jean-

Philippe Loeffler, 2009).

4 METHODOLOGY

To the proposed aims, I have developed a system to

measure, ex vivo, the muscle contractile response

due to indirect stimulation trough the nerve and the

muscle contractile response due to direct membrane

stimulation. The proposed system is fully automated,

through the use of a custom-made software, that

allows to precisely control the experiment, making

all the tests repeatable and accurate.

4.1 Experimental Setup

The muscle to be tested is excised with its intact

innervation and vertically mounted in an oxygenated

(95% O

and 5% CO

) and temperature controlled

chamber (30°C), containing a bicarbonate-buffered

solution at pH 7.4. One end of the muscle is linked

to a fixed clamp and the other end is connected to

the lever-arm of an Aurora Scientific Instruments

Inc. (ASI) 300B actuator/transducer. The apparatus

allows to stimulate the muscle both directly and

indirectly. For direct stimulation, two platinum

electrodes are located 2 mm from each side of the

muscle and the electrical stimuli are current pulses

of 300 mA generated by an ASI 701C stimulator.

For indirect stimulation, the nerve is sucked into a

suction electrode (A-M Systems Inc.) and

supramaximal pulses were delivered to the nerve by

an ASI 701B stimulator.

4.2 Protocol

We have developed a protocol that allows to study in

Soleus muscles several parameters of the isometric

muscular contraction and of the NMJ functionality,

proposed in the literature, in a single test. Indeed,

Personius et al. applied the fatigue protocols at two

frequencies on different specimens and van Lunteren

et al. studied the intratetanic fatigue, using separate

repetitive stimulations, delivered or directly on the

muscle or through the nerve.

The protocol is composed of four parts. Initially,

the muscle is stimulated with 4 single pulses to

measure twitch force (F

) and kinetic parameters,

namely time to peak (TTP), half relaxation time

(1/2RT) and maximum value of force time

derivative (dF/dt), for both membrane and nerve

stimulations, as shown in figure 1.

Figure 1: Twitch force (Ftw), time to peak (TTP), half

relaxation time (1/2RT) and maximum value of force time

derivative (dF/dt) can be measured with single pulse

stimulation.

After that, 8 shuffled stimulations in the range

between 20 Hz, which is the lowest summation

frequency, and 80 Hz, the Soleus tetanic frequency,

are used to measure the force-frequency curves for

both muscle and nerve stimulations. Finally, the

NMJ functionality parameters, namely

Neurotransmission Failure (NF) and Intratetanic

Fatigue (IF), are measured with two isometric

fatigue paradigms. These paradigms are made up of

14 stimulations via the nerve and 1 stimulation on

the membrane, to induce a specific stress in the

NMJ. The first paradigm is based on pulse trains

delivered at a physiological firing frequency of

35 Hz while the second is composed of pulse trains

delivered at the tetanic frequency. We employed

DevelopmentofanAutomatedSystemforExVivoMeasuringtheNeuroMuscularJunctionFunctionality

0.8 s pulse trains, with a rest period of 1.2 s between

each of them, to stimulate the soleus, differently

from Personius et al. that used 0.33s pulse train to

stimulate the diaphragm membrane and the phrenic

nerve, with a rest period of 0.67s after each train.

This differences are due to the different fiber

composition of the two muscles. This sequence is

repeated 20 times, for a total paradigm of 10 min.

The Neurotransmission Failure parameter

compares the force decrease due to the nerve

conduction to that due to muscle contraction during

the paradigm.

where F is the force decrease after nerve stimulation

and MF is the force decrease after membrane

stimulation .

Intratetanic Fatigue represents the force drop

which can occur within the same tetanic contraction,

in case of stimulation via the nerve.

where F

is the force at the last pulse and F

MAX

the maximum value of force within the single train

of pulses (see figure 2).

Figure 2: Intratetanic Fatigue is the force decrease within

the same pulse train. From van Lunteren et al., 2004.

4.3 Software

The experimental setup is controlled by a Windows

PC and a custom-made software in LabView 2011.

The interface between the computer and the

equipment is managed by the National Instruments

data acquisition board NI PCI Express 6353. This

choice allows to have a high flexibility in the

management of the experiments. The software is

used to control synchronously and automatically the

pulse stimulators and the actuator/transducer. In

particular, all the stimulation parameters are

manageable by means of simple text files provided

to the software as input.

During the experiment the software

automatically changes the acquisition frequency

setting it to 20 kHz in case of stimulation with single

pulses, and to 1 kHz during the remaining

stimulations. This choice is due to the fact that the

temporal parameters measured during a single pulse

stimulation are of the order of tens of milliseconds,

and an accuracy of 50s is therefore needed to point

out any significant difference. During stimulations at

frequencies higher than the single pulse, no temporal

parameters are calculated, therefore is sufficient to

acquire the data at 1 kHz to correctly sample the

signals of interest avoiding an overloading of the

system.

The software so designed and realized is

extremely flexible and functional. In fact, it is able

to perform all the main tests to define the

mechanical characteristics of muscle tissue and NMJ

functionality. This versatility has been extremely

useful when carrying out the preliminary tests aimed

at determining the optimal stimulation parameters,

such as the durations of the pulses, the stimulation

frequencies and the waiting times. On the other

hand, during the trial it is possible to perform

different protocols in a simple way.

5 EXPECTED OUTCOME

The proposed protocol allows the measurement of

the NMJ functionality in Soleus muscle of any

mouse model, by testing isometric contraction. We

expect to be able to characterize the NMJ

functionality of other muscle types by the

application of this protocol with only minor

modifications.

5.1 Future Plans

To perform a functional characterization of the

muscle-nerve preparations that better represents the

muscle and NMJ in vivo behavior, I am confident of

being able to develop a similar methodology to

study the muscle isotonic contraction. To do this, a

tool which allows the suction electrode for nerve

stimulation to follow the muscle shortening will be

designed and realized. I am also going to develop a

in vivo/in situ methodology that can allow to study

different muscles or muscular groups to give

information on the nerve conduction in addition to

NMJ and muscle functionality. Based on this latter

technique, the long-term goal is to study a possible

improvement of muscle functionality following

stimulation training protocols.

BIOSTEC2015-DoctoralConsortium

6 STAGE OF THE RESEARCH

6.1 Setting Stimulation Parameters

To determine the optimal pulse parameters to

stimulate the sciatic nerve I started performing

preliminary tests on Soleus muscle-nerve

preparations of four months-old Wild Type mice. At

first, I checked the current intensity necessary to

elicit the nerve stimulation: this was found to be

between 5 mA and 10 mA. The lower limit is the

first current intensity that causes the maximum

twitch contraction force. Increasing the current sent

through the suction electrode beyond 10 mA it

generates a current field so intense as to induce also

the direct stimulation of the muscle membrane

through the solution. Event that should absolutely be

avoided.

Once set the optimal current value, I looked for

the relation between the width of single pulses and

the twitch specific force. To this purpose 10 Wild

Type Soleus specimens were stimulated with single

pulses of widths between 0.2 ms and 2 ms, the range

most used in the literature. There was no significant

variation of specific force varying the pulses length.

However, in correspondence of 1.4 ms duration the

forces showed the highest value, as shown in figure

Figure 3: Twitch specific forces developed stimulating via

the nerve with pulses of width between 0.2 and 2 ms.

The choice has been also supported by tests of

tetanic stimulation in which trains of pulses of 1.4

ms caused a developed force on average higher than

the other widths.

Once we have determined the two parameters of

the stimulation pulse, we moved to verify the

validity of the method. In a healthy specimen is

expected that direct and indirect stimulations bring

the muscle to contract in the same manner and

develop the same forces. In fact, as shown in figure

4, the single pulse tests showed the same twitch

force and kinetic contraction parameters in both

cases of stimulation: the direct and the indirect one.

Figure 4: Twitch force and df/dt of WT Soleus stimulated

directly and indirectly.

The force-frequency relations are also

comparable in both cases of stimulation (see figure

5).

Figure 5: Force-frequency curves from WT Soleus directly

and indirectly stimulated.

The fatigue paradigm is derived from Personius

et al., however the Soleus stimulation parameters

were modified accordingly to its fiber composition

and to the standard membrane stimulation

parameters. Since the Soleus muscle is composed by

a higher percentage of slow fibers than the

diaphragm, it needs a longer stimulation time to

develop the maximum force. This stimulation time is

estimated to be 0.8 s and 1 s is necessary to the

muscle relaxation. Therefore, the fatigue paradigms

were set as follows: one 0.8 s pulse train delivered

directly on the membrane, followed by fourteen

0.8 s pulse trains delivered through the nerve, with a

rest period of 1.2 s after each pulse train. By

repeating this series of stimulations for 20 times, as

for the diaphragm, each paradigm takes 10 minutes.

Some preliminary tests showed that a rest time of

15 min is necessary between the 35 Hz protocol and

the 80 Hz one to have repeatable results not affected

by muscle fatigue. The rest time that the specimens

needed to recover their physiological properties after

the first fatigue paradigm was considered adequate

when the force generated by the muscle at the

DevelopmentofanAutomatedSystemforExVivoMeasuringtheNeuroMuscularJunctionFunctionality

beginning of the second fatigue paradigm was at

least 90% of its maximum force.

The application of the fatigue protocol showed

that the force developed by nerve stimulation

decreases very rapidly while the muscle continues to

be able to produce a much greater force as result of

direct stimulation, as shown in figure 6.

Figure 6: Example of a fatigue paradigm.

The calculation of the Neurotransmission Failure

gave results comparable to those existing in the

literature for the WT muscles. Likewise was not

revealed Intratetanic Fatigue in case of direct

stimulation of the muscle. On the other hand, in case

of indirect stimulation the IF is similar to that shown

by van Lunteren.

6.2 Transgenic Model

Once set the method on the wild type model, to

focus our attention on a pathologic model we started

studying the Soleus of SOD1

G93A

mice at the end

stage of the disease. At this age, in fact, significant

NMJ damages are expected thus allowing to obtain a

further confirmation of the validity of the method

and to test its sensitivity. The first experimental

results showed that transgenic muscles presents a

significant slowdown of the kinetic parameters if

directly stimulated, and a further slowdown in case

of nerve stimulation.

Analysis of force-frequency curves revealed

significant differences between WT and transgenic

muscles. At the tetanic frequency a decrease of

about 20% was reported for SOD1

G93A

Soleus

specific force, in comparison to the controls. Once

again, a worsening in the TG Soleus response was

reported when stimulated through the nerve. A

significant decrease of specific forces was, in fact,

measured for all the tested frequencies. The

maximum specific force developed in case of nerve

stimulation appears halved if compared to the direct

stimulation. These results highlight the sensitivity of

the method to discriminate even small temporal

differences and to separate the components of

muscle contraction due to muscle inner damages

from the ones due to NMJ conduction defects.

The analysis of the forces measured during the

80 Hz fatigue paradigm pointed out a limitation of

the method when calculating intratetanic fatigue at

the end of the protocol. In fact, in literature this

parameter is measured during repeated nerve

stimulations of about 2.5 minutes. We tried to

calculate the IF during the entire fatigue protocol

and we observed that at the beginning the generated

forces are always of the order of tens of mN so it is

possible to reliably measure the force decrease

within the same pulse train. On the contrary, at the

end of the same paradigm, the muscles were

exhausted and the contraction forces developed in

case of nerve stimulation were near to zero,

especially in the transgenic model. In this situation,

the sensitivity of the method resulted heavily

reduced, and the IF values calculated from the

seventh minute on, were the forces are about 7 mN,

basically expresses noise. For this reason we are

evaluating to shorten the paradigm. See figure 7.

Figure 7: Maximum force developed by the same

specimen at the first nerve stimulation end at the last nerve

stimulation during the fatigue paradigm.

Because of this reason, I decided to compute NF

and IF only in the first seven minutes of stimulation

of the 80 Hz fatigue protocol. Results showed higher

IF values, starting from the first stimulations, in

transgenic muscles stimulated trough the nerve

compared to the WT model.

On the contrary the analysis of

Neurotransmission Failure did not show any

alteration in the transgenic model. A possible

explanation of this may be that at the end-stage of

the pathology the transgenic

skeletal muscles are

also severely compromised and can hide the NMJ

defects.

In conclusion, the proposed experimental

technique allows to determine the NMJ functionality

separately from the muscle contractile properties in

isolated muscle-nerve preparations of pathological

mouse models. Preliminary results obtained from the

SOD1

G93A

model are in accordance with the

literature, showing muscle contraction defects and

NMJ impairment.

BIOSTEC2015-DoctoralConsortium

REFERENCES

Aldrich, T. K., Shander, A., Chaudhry, I., and Nagashima,

H. (1986). Fatigue of isolated rat diaphragm: role of

impaired neuromuscular transmission. Journal of Ap-

plied Physiology, 61(3):1077–1083.

Brooks, S. V. and Faulkner, J. A. (1988). Contractile prop-

erties of skeletal muscles from young, adult and aged

mice. Journal of Physiology, 404:71–82.

Del Prete, Z., Musaro`, A., and Rizzuto, E. (2008).

Measur- ing mechanical properties, including isotonic

fatigue, of fast and slow MLC/mIgf-1 transgenic

skeletal mus- cle. Annals of biomedical engineering.

Derave, W., Van Den Bosch, L., Lemmens, G., Eijnde, B.

O., Robberecht, W., and Hespel, P. (2003). Skele- tal

muscle properties in a transgenic mouse model for

amyotrophic lateral sclerosis: effects of creatine treat-

ment. Neurobiology of disease, 13(3):264–272.

Dupuis, L., Gonzalez de Aguilar, J. L., Echaniz-Laguna,

A., Eschbach, J., Rene, F., Oudart, H., Halter, B.,

Huze, C., Schaeffer, L., Bouillaud, F., and Loeffler, J.

P. (2009). Muscle mitochondrial uncoupling

dismantles neuromuscular junction and triggers distal

degenera- tion of motor neurons. PloS one,

4(4):e5390.

Fischer, L. R., Culver, D. G., Tennant, P., Davis, A. A.,

Wang, M., Castellano-Sanchez, A., Khan, J., Polak,

M. A., and Glass, J. D. (2003). Amyotrophic lateral

sclerosis is a distal axonopathy: evidence in mice and

man. Experimental Neurology, 185(2):232–240.

Gurney, M. E., Pu, H., Chiu, A. Y., Canto, M. C. D., Pol-

chow, C. Y., Alexander, D. D., Caliendo, J., Hentati,

A., Kwon, Y. W., and Deng, H. X. (1994). Mo- tor

neuron degeneration in mice that express a hu- man

Cu,Zn superoxide dismutase mutation. Science,

264(5166):1772–1775.

Hegedus, J., Putman, C. T., Tyreman, N., and Gordon, T.

(2008). Preferential motor unit loss in the SOD1 G93A

transgenic mouse model of amyotrophic lateral

sclerosis. The Journal of physiology, 586(14):3337–

3351.

Lee, Y. I., Mikesh, M., Smith, I., Rimer, M., and

Thompson, W. (2011). Muscles in a mouse model of

spinal mus- cular atrophy show profound defects in

neuromuscu- lar development even in the absence of

failure in neu- romuscular transmission or loss of

motor neurons. De- velopmental biology, 356(2):432–

444.

Ling, K. K. Y., Lin, M.-Y., Zingg, B., Feng, Z., and Ko,

C.-P. (2009). Synaptic defects in the spinal and neu-

romuscular circuitry in a mouse model of spinal mus-

cular atrophy. PloS one, 5(11):e15457–e15457.

Ngo, S. T., Baumann, F., Ridall, P. G., Pettitt, A. N.,

Henderson, R. D., Bellingham, M. C., and Mc-

Combe, P. A. (2012). The relationship between

Bayesian motor unit number estimation and histo-

logical measurements of motor neurons in wild-type

and SOD1(G93A) mice. Clinical Neurophysiology,

123(10):2080–2091.

Personius, K. E. and Sawyer, R. P. (2006). Variability and

failure of neurotransmission in the diaphragm of mdx

mice. Neuromuscular disorders : NMD, 16(3):168–

177.

Turner, B. J. and Talbot, K. (2008). Transgenics, toxicity

and therapeutics in rodent models of mutant SOD1-

mediated familial ALS. Progress in neurobiology,

85(1):94–134.

Van Lunteren, E., Moyer, M., and Kaminski, H. J. (2004).

Adverse effects of myasthenia gravis on rat phrenic di-

aphragm contractile performance. Journal of Applied

Physiology, 97(3):895–901.

DevelopmentofanAutomatedSystemforExVivoMeasuringtheNeuroMuscularJunctionFunctionality