ARCHIVES ITALIENNES DE BIOLOGIE

ARCHIVES ITALIENNES DE BIOLOGIEA JOURNAL OF NEUROSCIENCE,

PUBLISHED BY THE UNIVERSITY OF PISA,WITH THE SUPPORT OF THE UNIVERSITY OF PISA

The Archives Italiennes de Biologie have been edited in the past by the following Chief Editors:

C. Emery and A. Mosso Volume 1-7 (1882-1886) A. Mosso Volume 8-42 (1887-1904) A. Mosso and V. Aducco Volume 43-53 (1905-1910) V. Aducco Volume 54-94 (1910-1936) G. Moruzzi Volume 95-118 (1957-1980) O. Pompeiano Volume 119-140 (1981-2002) O. Pompeiano and P. Pietrini Volume 141-145 (2003-2007) P. Pietrini and B. Ghelarducci Volume 146 (2008)

Archives Italiennes de Biologie – a Journal of Neuroscience – was founded in 1882 and represents one of the oldest neuroscience journals in the world. Archives publishes original contributions in all the fields of neuroscience, including neurophysiology, experimental neuroanatomy and electron microscopy, neurobiology, neurochemistry, molecular biology, genetics, functional brain imaging and behavioral science.Archives Italiennes de Biologie also publishes monographic special issues that collect papers on a specific topic of interest in neuroscience as well as the proceedings of important scientific events.Archives Italiennes de Biologie is published in 4 issues per year and is indexed in the major collections of biomedical journals, including Medline, PubMed, Current Contents, Excerpta Medica. Since 2008, Archives Italiennes de Biologie is also published online. Submission of papers is now handled electronically through the website and published articles can be downloaded as PDF. For detailed instructions on how to submit manuscripts please visit the journal website at: http://www.architalbiol.org

Luigi Aloe (Roma)Evan Balaban (Montreal)Fabio Benfenati (Genova)Nicoletta Berardi (Firenze)

Mélanie Boly (Liege)Richard Boyle (Moffett Field)

Luigi Cervetto (Pisa)Enrico Cherubini (Trieste)

Patricia Churchland (San Diego)Fiorenzo Conti (Ancona)Maura Furey (Bethesda)

M. Felice Ghilardi (New York)James Haxby (Dartmouth)

Leszek Kubin (Philadelphia)Michela Matteoli (Milano)

Thomas Mergner (Freiburg)Paolo Nichelli (Modena)

Silvia Pellegrini (Pisa)Julian Paton (Bristol)

Emiliano Ricciardi (Pisa)Giuseppe Sartori (Padova)Jerry Siegel (Los Angeles)

Irene Tobler (Zürich)Giulio Tononi (Madison)

Johannes van der Steen (Rotterdam)Neil Watson (Burnaby)

CHIEF EDITORSPietro Pietrini and Brunello Ghelarducci (Pisa)

Published by Pisa University PressLungarno Pacinotti 43/44, 56126 Pisa, Italy

Printed in September 2014by Tipografia Monteserra S.n.c. - Vicopisano (PI)

Authorization of the Pisa Tribunal No. 20/1956Arch. Ital. Biol. ISSN: 0003-9829

EDITORIAL BOARD

MANAGING EDITORS

Emiliano Ricciardi and Paola d’Ascanio (Pisa)

EDITORIAL STAFF

Anna Gaglianese, Giusy Rota, Caterina Iofrida, Erika Melissari (Pisa)

I N S T R U C T I O N S T O A U T H O R S

Submission of a paper will be taken to imply that it has not previously been published, except in abstract form, and that it is not being considered for publication elsewhere. Decision as to publication of papers submitted to the Archives Italiennes de Biologie will be based on the opinion of the Editorial Board as to the significance and originality of the work.Manuscripts should be submitted electronically via the Archives Italiennes de Biologie page on http://www.architalbiol.org, hardcopy submission of manuscripts is no longer recommended. In cases of difficulty operating the electronic process, or in order to ship a hard-copy of your manuscript, please contact the Managing Editor for help and advice.

Authorship. To be identified as an author, the participant should have contributed to the conception and design of the project, drafted substantive portions of the paper or edited or revised same, and taken responsibility for the analysis and conclusions of the paper.

Preparation of manuscripts. Manuscripts should be prepared electronically using an appropriate Office Word compatible text-processing package, formatted for A4 or letter page size, double-spaced throughout, and using a 12 point font. Articles must be written in English. Text should flush left, and not be justified. Words should not be hyphenated. Pages should be numbered sequentially. In order to promote fairness of the review process, the Editors provide an anonymous review of all manuscripts. In order to achieve this, these separate pages will be removed from the manuscripts sent to referees. Authors will be sent reviewer comments. All reviewers will remain anonymous. Authors are informed about the Editors’ decision after the review process is completed.The main body of full-length paper should be divided into: Abstract, Introduction, Methods, Results, Discussion, References. Short com-munications should neither be divided in sections nor is the abstract required; the text should not be longer than 4 printed pages, including tables and illustrations. Review articles must have an initial Table of Contents and a final Summary. Book reviews, should not be longer than one printed page.

References. The style of citation should conform to that used in this issue.

Figures. Submit tables and illustrations as separate files, only as TIFF or EPS files, with a minimum resolution of 300dpi, 100% size repro-duction and with a format of to fit within a single or double column width of 80 mm and 160 mm, respectively, and a maximum page height of 220 mm. Authors can decide to only publish their figures in colour in the electronic publication (at no extra charge) whilst retaining greyscale reproduction in the printed version.

Supplementary files. The Archives Italiennes de Biologie now accept electronic supplementary material to support and enhance your scientific research. Supplementary files offer the author additional possibilities to publish supporting applications, movies, animation sequences, high-resolution images, background datasets, sound clips and more. Supplementary files supplied will be published online alongside the electronic version of your article. Authors should submit the material in electronic format together with the article and supply a concise and descriptive caption for each file.

Disclosure of competing interests and financial support is required at the time of submission. The authors are responsible for inform-ing the journal of any additional conflicts of interest or financial support that may arise prior to the date of publication of their paper.

Legal requirements. Papers describing animal experiments must be conducted in conformity with the legal requirements. Evidence should be provided by the Authors that they took adequate steps to ensure that animals did not suffer, and that the experiments were conducted in accordance with international standards on animal welfare as well as being compliant with local and national regula-tions. Studies are expected to be compliant with minimal standards as defined by the European Communities Council Directive 86/609/EEC and the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Papers describing experiments on human subjects should be conducted in conformity with the experimental protocol approved by the appropriate local ethics committee, and in compliance with national legislation and the Code of Ethics of the World Medical Association (Declaration of Helsinki), and must include a statement that written informed consent was obtained after the procedure(s) had been fully explained. If your submission does not contain written informed consent or Institutional Review Board approval, it will not be reviewed. Ethical and legal considerations require careful attention to the protection of a patient’s anonymity in any component (text, figures, tables, etc.) of the manuscript.

The Archives ltaliennes de Biologie are published in quarterly issues by the Department of Physiological Sciences, University of Pisa, Via S. Zeno 31, 1-56127, Pisa, Italy (Fax nr. +39-50-2213527; E-mail: [email protected]). The subscription price of Volumes, including postage, is € 140 (Italy) or € 180 (foreign countries). For payments, Subscribers may use Check made out to Pisa University Press S.r.l. or Bank Transfer to Banco Popolare, account no.: 000000002305, ABI 05034, CAB 14011 (Pisa University Press S.r.l. to be cited), IBAN: IT47W0503414011000000002305, SWIFT/BIC: BAPPIT21T51.Orders should be sent to: Paolo Vitali, Pisa University Press S.r.l., University of Pisa, Lungarno Pacinotti n. 43/44, I-56126, Pisa, Italy – Tel. + 39 50 2212055 – Fax + 39 50 2212945 – [email protected] for missing issues can only be considered if made immediately after receipt of the subsequent issue. Back volumes are also available.Subscribers’ data are treated in accordance with the provisions of the Legislative Decree, 30 June 2003, n. 196 – by means of computers operated by personnel, specifically responsible. These data are used by the Publisher to mail this publication. In accordance with Article 7 of the Legislative Decree no. 196/2003, subscribers can, at any time, view, change or delete their personal data or withdraw their use by writing to Pisa University Press, Lungarno Pacinotti 43, Pisa, Italy.

Cover image. The fine structure of the rabbit cerebellum as drawn by Camillo Golgi in his paper: “Recherches sur l’Histologie des Centre Nerveux”. IV. Circonvolutions cérébelleuses, Tav. V, Tome IV, Archives Italiennes de Biologie, Deuxième Annèe 1883.

T Y P E S O F P A P E R S

1. Research articles, reporting results of original research.2. Short communications, reporting on research in progress.3. Review articles, discussing current developments of interest to neuroscientists.4. Book reviews, discussing publications of general interest to neuroscientists.The journal will also consider articles on historical aspects of neuroscience, methodological papers, and editorials.

Introduction

Occlusal muscles dysfunctions may lead to Temporo-Mandibular Disorders (TMD, Cooper et al., 1991) characterized by pain, enhanced sympathetic activ-ity and increased daytime cortisol levels (Korszun et al., 2002; Light et al., 2009), likely depending on nociceptive trigeminal inputs (Sato and Schmidt, 1973; Bartsch et al., 2000). Moreover, a disregula-tion of the sympathetic activity in TMD has been recently proposed on the basis of pupillometric findings (Monaco et al., 2012). Finally, recent case reports showed that asymmetric sensorimotor tri-geminal signals are associated with asymmetries in

the activity of autonomic centres controlling verte-bral arteries haemodynamics and the pupil size (De Cicco, 2012a). In this instance, correction of the occlusal unbalance, which reversibly modified the left-right asymmetry in the masseter electromyo-graphic (EMG) activity, reduced pupils size asym-metries. The latter findings suggest that trigeminal centres exert a tonic control on autonomic structures and, thus, asymmetric trigeminal activity may create an unbalance in the activity of autonomic centres.In order to go deeper into the relation between tri-geminal and autonomic activity, we have studied whether 1) pain free TMD patients showing an asymmetry in the EMG activity of left and right

Sensorimotor trigeminal unbalancemodulates pupil size

V. DE CICCO1, E. CATALDO2, M. BARRESI3, V. PARISI4, D. MANZONI1

1 Department of Translational Research, University of Pisa, Italy; 2 Department of Physics, University of Pisa, Italy; 3 Department of Drug Sciences, University of Catania, Italy;

4 G.B. Bietti Foundation, IRCCS, Roma, Italy

A B S T R A C T

We studied whether the patients affected by Temporo-Mandibular Disorder (TMD) showing asymmetric electro-myographic (EMG) activity of masticatory muscles display asymmetries also in the pupils size in order to detect a possible tonic trigeminal control on autonomic centres. In 30 pain free TMD patients, we found a highly significant, positive correlation between left-right differences in EMG and pupils size. The asymmetry in the pupils size was likely induced by the asymmetric sensorimotor signals arising from the oro-facial region, as the pupils asymmetry decreased significantly after application of a cusp bite. Moreover, cusp bite wearing bilaterally increased the mydriasis induced by performing haptic tasks. Finally, unbalancing the occlusion by a precontact increased the diameter of the ipsilateral pupil and abolished the mydriasis induced by haptic tasks. In conclusion, trigeminal sensorimotor signals may exert a tonic control on autonomic structures regulating the pupils size at rest and dur-ing haptic tasks. Since task-associated mydriasis is correlated with task performance and is strictly proportional to the phasic release of noradrenaline at cerebral cortical level, present findings suggest that unbalanced trigeminal activity influences brain processes not directly related to the orofacial region.

Key wordsElevator muscles • Occlusion • Proprioceptive trigeminal signals • Autonomic control •

Pupil size • Locus Coeruleus

Corresponding Author: Prof. Diego Manzoni, Department of Translational Research and New Technologies in Medicine and Surgery, University of Pisa, via San Zeno 31, I-56127 Pisa, Italy - Tel.: +39 50 2213466 - Fax: +39 50 2213527 - Email: [email protected]

Archives Italiennes de Biologie, 152: 1-12, 2014.

2 V. DE CICCO ET AL.

masseter muscles displayed also asymmetries in the pupil size, and whether 2) a reduction of the EMG asymmetry by application of a cusp bite affected also pupil size asymmetry, thus indicating that sen-sorimotor trigeminal signals tonically modulate the activity of the autonomic structures controlling the pupil size. Finally, we assessed whether EMG asym-metries influenced the mydriasis associated with a sensorimotor task. In fact, the pupil size correlates with the changes in the neural activity occurring during task-associated “arousal” (Bradshaw, 1967; Bradley et al., 2008) and “mental effort” (Hess and Polt, 1964), as well as with task performance (Rajkoski et al., 1993).

Methods

SubjectsThe study protocol was in line with the declara-tion of Helsinki and was approved by the ethical committee of the San Domenico Clinics, (Rome, Italy). All participants signed an informed consent. Experiments were performed in 30 patients (age 25-45 years, 10 males and 20 females) affected by TMD (Dworkin et al., 1992), showing an asymmet-ric activity of masseters during clenching, and not exhibiting tooth loss and pain symptoms of any ori-gin. Participants under medication or reporting neu-rological, psychiatric, metabolic, endocrine symp-toms, or orthopaedic problems were not included in the study.

Preliminary evaluation and cusp bite manufacturingPatients were studied at least 2 hours after the lat-est caffeine containing beverages and cigarettes smoking. In a preliminary session, evaluation of the EMG activity of masseter muscles (and mandibu-lar kinematics) was performed during swallowing and clenching. Only subjects showing an asym-metry in EMG activity higher than 15% (quanti-fied as the ratio between the left-right difference and the left-right mean) (see Fig. 1A) were further investigated. Then, a fifteen-minutes transcutaneous electrical nerve stimulation (TENS) of trigeminal motor branches (Noaham and Kumbang, 2008), which activates muscles by direct stimulation of motor axons (Gomez and Christensen, 1991), was

performed (Fig. 1B). Stimulation was administered through four couples (cathode/anode) of electrodes (1600 mm2 of surface) applied at the level of incisura sigmoidea and of the submental region of both sides. Biphasic (cathodal/anodal) current pulses (0.54 msec duration, 21-25 mA intensity) were delivered by two I.A.C.E.R. stimulators (Martellago, Venice, Italy) leading to repeated contractions of masseters and mandible depressor muscles. The intensity of the left and right stimulation was adjusted in order to obtain a symmetric muscle activation (evaluated by EMG recording). Low frequency stimulation (0.618 Hz) was utilized for elevator muscles and higher frequency (40 Hz) for depressor muscles. In this way, alternated contraction and relaxation were observed in masseters, while mandible depressor muscles were tonically contracted, giving rise to small amplitude mandibular movements (1 mm). Following TENS, the mandibular resting posture was lowered and a dental impression was obtained in the new relative position of the arches by placing a self-hardening material between them. This dental impression was used to manufacture a cusp bite (Dao et al., 1994) modeled on the inferior dental arch. Cusp bite placement reduced the myoelectric unbalance, which decreased to less than 15% in all patients (Fig. 1C).

Experimental procedureThe experimental procedure is illustrated in Fig 2. At time 0 (t0), patients were studied at first in the habitual occlusal condition (without cusp bite, Bite OFF). The following measurements were made:1. EMG activity of masseters during clenching

effort;2. pupils size at rest, with the dental arches touching

each other;3. pupils size during performance of an haptic, sen-

sorimotor task (TanGram), with the dental arches touching each other.

All measurements were repeated during cusp bite wearing (Bite ON).Subjects were tested again after 90 days (t 90) of continuous cusp bite wearing (bite was taken off only during meal and teeth cleaning) in Bite ON condition. The following variables were studied:1. EMG activity of masseters in resting state;2. pupils size in resting state;3. pupils size during TanGram performance.

SENSORIMOTOR TRIGEMINAL UNBALANCE MODULATES PUPIL SIZE 3

Pupil size (2-3), but not EMG activity (1), was recorded again in Bite OFF condition and Bite ON condition.At this point a new test was administered to evaluate the effects possibly elicited by the induction of an occlusal deficit on the pupils size. For this purpose, after a further baseline evaluation of the pupils size (Bite OFF) at rest and during haptic task, zircon crystals were applied to the vestibular surface of the inferior canine tooth, on the side of the highest EMG activity (where the pupil was larger, see Fig. 4), in order to produce a pre-contact between the tooth

face and the palatal side of the upper canine. In this condition, pupil size measurements were performed both at rest and during haptic task.

Haptic taskAt the beginning of the experimental session, patients were instructed to perform a haptic task, which was practiced only once. The haptic task used in the present study was based on Tan Gram, consisting of a puzzle of triangular, square and par-allelogram-shaped forms. A piece of the puzzle (the parallelogram) was removed by the experimenter

Fig. 1. - Clinical evaluations of asymmetries in masseter activity and their correction by cusp bite. A) Patient in habit-ual occlusion (1). During clenching a left-right asymmetry in masseter EMG activity was observed (2). Calibration is indicated by vertical white lines and related numbers. B) Soon after the evaluation displayed in A the patient was submitted to transcutaneous electrical nerve stimulation (TENS, see text for further details), thus achieving a new mandibular resting posture. The picture illustrates the positioning of stimulating electrodes. A cusp bite appropriate for maintaining arches in this new position was prepared. C) One week following the initial session the patient was wearing the cusp bite prepared as indicated in B. The positioning of the cusp bite between the arches is shown in 1. Note that during clenching the masseter EMG activity was now more symmetric (2). The numbers on EMG traces represent the amplitude of the calibration bar (white vertical line).


and placed in the right hand of the subject, who had to reposition it in its original place without looking at his/her hand. Patients performed the task with their head placed in the pupillometer and the pupil size was monitored as soon as they began to explore the puzzle surface.

Data acquisitionPupil size measurements (mm) were performed in the same day time in all subjects and in standard condition of artificial lighting by using a corneal topographer-pupillographer (MOD i02, with chin support, CSO, Florence, Italy) made up of a stan-dard illuminator (halogen lamp, white light), a camera sensor CCD1/3”, with a 56 mm working distance. The operator monitored the iris image by the camera, which had an acquisition time of 33 msec. Measurements were performed for both eyes in photopic conditions (40 lux) and values were dis-played online on the computer screen.

The EMG activity of masseter muscles was recorded by Duo-trode surface Ag/AgCl electrodes (interelec-trode distance 19,5 mm, Myo Tronics, Seattle, WA, USA). Electrodes were placed on the masseters belly, along an axis joining the orbit corner to the man-dibular gonion, two cm far from the latter. The lead axis was parallel to the longitudinal axis of muscle fibres. Data were acquired at the sampling rate of 720 Hz by using an integrated system for EMG activ-ity and mandibular movement recording (K6-I; Myo Tronics). EMG signals were acquired with a lower cutoff frequency of 15 Hz, filtered with a notch (50 Hz), full-wave rectified and displayed on the instru-ments monitor. The instrument provided the mean value of the rectified EMG bursts produced during clenching. Recording was allowed by the instrument software only when the resistance of the two record-ing leads was comparable, which allows to minimize possible bias in the asymmetry evaluation due to the different size of the EMG signal of the two sides.

Fig. 2. - Flow chart of the experimental procedure. A. First experimental section (time 0). B. Second experimental section (time 90 days). The Black arrows mark the temporal order of the different tests performed. See text for further explanation.


Statistical analysis (SPSS.13)Analysis was performed by repeated measures ANOVA. For the size of each pupil and absolute left-right size differences a 2 time (t0, t90) x 2 condition (bite on, bite off) x 2 task (resting, haptic task) experimental design was used, while a single factor design (condition or time) was used for EMG activity values and absolute left-right differences recorded during clenching. In addition a 2 condi-tion (bite off, zircon) x 2 task (resting, haptic task) design was performed a t90 for pupils size. Finally, a 2 time (t0, t90) x 2 condition (bite on, bite off) analysis was performed for the increase in pupil size induced by the haptic task. The Greenhause-Geisser H correction was used when requested. Correlations between variables were assessed by linear regression analysis. Significance was set at p < 0.05

Results

A preliminary analysis excluded significant gender effects.

Effects of orthotic correction on pupil size and EMG activity at time 0During the first experimental session (t0), in habit-ual occlusion (bite OFF), all patients showed clear asymmetries of EMG activities (absolute left-right difference (mean ± SD = 50.9 ± 17.2, PV) and pupils size (0.326 ± 0.214, SD, mm). The distribu-tion histograms of the corresponding left (L)-right (R) differences was bimodal, with positive (L > R) and negative (R > L) values (Fig. 3). As shown in Fig. 4, the pupil size asymmetry was highly cor-related with the corresponding asymmetries in the

Fig. 3. - Changes in the distributions of left-right differences in EMG activity and pupil size induced by bite correc-tion. Distribution of the differences between left and right pupil size (in mm) and masseter EMG activity (in PV) have been shown in the upper and lower row, respectively. Data obtained during normal occlusion (Bite OFF) and following bite correction (Bite ON) have been shown in A-B and C-D, respectively.


EMG activity (r = 0.75, p < 0.0001; Y = 0.0066 X -0.005). In all instances, the side showing the larg-est EMG activity showed also the larger pupil size. The difference between the larger and smaller pupil size was highly significant (EMG activity: F(1,28) = 254.6, p < 0.0001; pupil size: F(1,28) = 67.46, p < 0.0001). Wearing the cusp bite reduced significantly the asymmetry in EMG activity (from 50.9 ± 17.2 to 14.1 ± 8.9 PV, F(1,28) = 125.8, p < 0.0001); as a consequence, the corresponding histogram became unimodal (Fig. 3C-D). The statistical analy-sis performed on the asymmetry in pupils size (and on the corresponding values) is shown in Table I. Decomposition of the significant time x condition x task interaction indicated that cusp bite reduced sig-nificantly the pupils asymmetry at rest (from 0.326 ± 0.214 to 0.11 ± 0.10 mm, p < 0.0001). The asym-metries in EMG activity and pupils diameters were still significantly correlated (r = 0.62, p < 0.0005), although greatly reduced. The correlation between the two variables was similar to that observed before bite wearing (Y = 0.006 X -0.007).As shown in Table II, the reduction in the asym-metry of masseter EMG was due to a significant increase in the activity of the less active masseter

muscle, whereas the reduction of the pupil asymme-try was due to a significant reduction in the diameter of both pupils, although larger in the largest one.

Time, bite and task effects on the pupils size and asymmetryDecomposition of the significant time x bite x task interaction illustrated in Table I indicated that a both t0 and t90 cusp bite wearing greatly decreased the asym-metry in pupils size at rest, due to significant reduc-tions in the diameters of both pupils (see Table II). On the other hand, on both occasions, the haptic task significantly increased the pupils size independently of bite conditions. During the haptic task, at variance with what observed at rest, bite correction increased the size of both pupils. As a consequence, cusp bite wearing significantly amplified the mydriasis observed during the haptic task, as indicated by decomposition of the significant time x condition effect observed for this parameter variable (larger pupil: F(1,28) = 8.63, p < 0.007, smaller pupil: F(1,28) = 33.1, p < 0.0001), with post-hoc comparison indicating significant dif-ferences between Bite OFF and Bite ON conditions for both pupils at t0 (largest pupil, p < 0.0001; smaller pupil, p < 0.0001) and t90 (largest pupil, p < 0.0001; smaller pupil, p < 0.0001) (Fig. 5, Table II). The larger task-associated mydriasis observed in Bite ON with respect to Bite OFF condition is illustrated in Fig. 6 for a representative patient.After the 90 days period elapsing from the first experi-mental session, in Bite ON condition both pupils showed a significantly smaller diameter at rest with respect to the first experimental session and a larger diameter during the haptic task. In Bite OFF condi-tion the only significant difference concerned the larger pupil, which showed a smaller diameter at rest with respect to the first session. As a consequence of the latter change, in Bite OFF the pupils asymmetry observed at rest was smaller in the second experimen-tal session with respect to the first one (t0: 0.33 ± 0.21; t90: 0.21 ± 0.20 mm, p < 0.0001). In this condition, the left-right differences in the pupil size measured at t90 was significantly correlated with that observed at t 0 (r = 0.952, Y = 0.738 X -0.023, p < 0.0005).

Effect of malocclusion on pupil sizeThe study of the effect of zircon crystal application showed a significant condition (Bite OFF/zircon) x task interaction (F(1,28) = 75,09, p < 0.0001) whose

Fig. 4. - Correlation between left-right side differences in pupil size and EMG activity. Scatter plot of left-right dif-ferences in pupil size and masseter EMG activity. Black squares and open circles represent subjects in normal occlusion (Bite OFF) and following bite correction (Bite ON), respectively. Positive and negative values on the abscissa and ordinate indicate predominance of the left and right side values, respectively. The regression line plot-ted on the graph is relative to data obtained in the Bite OFF condition and it is virtually indistinguishable from that obtained for Bite ON data. Its equation has been plotted in the rectangular box together with the value of the coef-ficient of correlation (r) and statistical significance.


decomposition indicated that the zircon-induced malocclusion increased significantly the pupil size in the resting condition (from 4.29 ± 0.79 to 5.06 ± 0.76, p < 0.0001), but not during the haptic task performance (Bite OFF: 5.05 ± 0.79; zircon: 5.14 ± 0.77, NS). As a consequence, malocclusion abol-ished the pupil dilation elicited by the task owing to the placement of a zircon crystal, which made the pupil size at rest as large as that observed during task performance with normal occlusion.

Discussion

Changes in basal pupil size: functional considerationsThe results of the present study indicate that the presence of an asymmetric EMG activity of masti-catory muscles during clenching is highly predictive of an asymmetry of the same sign in the pupils size. EMG asymmetries during clenching did not arise from asymmetries in the electrodes resistance and

Table I. - Summary of significant results.

Variable Smaller pupileffect (df = 28)

Larger pupileffect (df = 28)

Pupils asymmetryeffect (df = 28)

time F = 5.33* F = 42.2*** F = 8.14**

task F = 236.2*** F = 211.5*** F = 7.10*

bite ns ns F = 45.09***

time x task F = 33.84*** F = 122.8*** F = 5.83*

bite x task F = 33.8*** F = 88.8*** F = 4.11, p = 0.052

time x bite x task F = 33.1 F = 8.6** F = 29.77***

time 0

F(1,28) = 65.4***, 69.3*** F = 4.19*

Bite OFF

r < haptic, t = 8.9*** r < haptic, t = 0.06*** ns

Bite ON

r < haptic, t = 14.3*** r < haptic, t = 13.76*** r < haptic, t = 2.52*

Bite ON vs. OFF

resting, OFF > ON, t = 2.7* resting, OFF > ON, t = 4.99*** resting, OFF > ON, t = 7.15***

haptic, OFF < ON, t = 6.3*** haptic, OFF < ON, t = 3.87** haptic, OFF > ON, t = 5.33***

time 90 days

F(1,28) = 129.2*** F = 105.5*** F = 13.98**

Bite OFF

r < haptic, t = 8.1*** r < haptic, t = 9.84*** r < haptic, t = 4.01***

Bite ON

r < haptic, t = 19.16*** r haptic, t = 17.4*** r < haptic, ns

Bite ON vs. OFF

resting, OFF > ON, t = 4.47*** resting, OFF > ON, t = 5.82*** resting, OFF > ON, t = 3.95***

haptic, OFF < ON, t = 7.29*** haptic, OFF < ON, t = 4.20*** haptic, OFF > ON, t = 6.07***

time 0 vs. time 90 days

Bite ON

resting, t0 > t90, t = 5.256*** resting, t0 > t90, t = 9.86*** resting, ns

haptic, t0 < t90, t = 4.98*** haptic, t0 < t90, t = 5.11*** haptic, ns

Bite OFF

resting, ns resting, t0 > t90, t = 7.78*** resting, t0 > t90, t = 7.05***

haptic, ns haptic, ns haptic, ns

*** = p < 0.0001; ** = p < 0.01; * = p < 0.05


placement. They were strongly reduced by cusp bite placement and reappeared as soon as the cusp bite was removed. The asymmetric EMG activ-ity is likely due to an asymmetry in proprioceptive signals deriving from muscle spindles/periodontal receptors and/or in the efference copies of trigemi-nal motor signals. Present findings indicate that this sensorimotor unbalance exerts a tonic influence on sympathetic activity related to the control of the

pupils size. In fact, the reduction/abolition of the muscle asymmetry (and, as a consequence of the asymmetry in sensory and/or motor trigeminal sig-nals) by bite correction is immediately coupled to a drastic reduction in the asymmetry observed in pupil size. The remarkable stability of these effects was documented by the fact that they could be observed also 90 days after the first session. We may conclude that the development of a left-right asymmetry in the trigeminal sensory and/or motor signals induces a corresponding unbalance in the activity of left and right autonomic structures involved in the control of pupil size.Our findings can be accounted for by the fact that trigeminal afferents may affect the dilatator pupillae muscle by controlling the activity of preganglionic neurons located within the superior cervical gangli-on (Bartsch et al., 2000). It is known that trigeminal afferents are the origin of pathways running through well-known autonomic structures, such as the nucle-us tractus solitarii, the ventrolateral medulla, the A5 area, the ventrolateral part of the parabrachial nucleus and the Kolliker-Fuse nucleus (Panneton et al., 2000). Moreover, the peritrigeminal area sur-rounding the trigeminal motor nucleus is connected to the parvicellular reticular formation (Bourque and Kolta, 2001; Notsu et al., 2008), a structure mediat-ing autonomic reflexes (Esser et al., 1998). In addi-tion, preganglionic parasympathetic neurons located within the Edinger-Westphal nucleus, which induces miosis, receive afferents from the reticular forma-tion and vestibular nuclei (Breen et al., 1983), which

Table II. - Mean values of recorded variables.

Time Time 0 Time 90 days

Condition Bite OFF Bite ON Bite OFF Bite ON

Task Rest Haptic Haptic-Rest

Rest Haptic Haptic-Rest



larger pupil 4.42 ± 0.82 5.04 ± 0.78 3.96 ± 0.69 5.33 ± 0.71 4.29 ± 0.79 5.05 ± 0.77 3.77 ± 0.62 5.38 ± 0.70

smaller pupil 4.09 ± 0.72 4.70 ± 0.68 3.85 ± 0.63 5.16 ± 0.64 4.08 ± 0.70 4.68 ± 0.72 3.70 ± 0.58 5.24 ± 0.63

larger pupilmydriasis

0.62 ± 0.42 1.37 ± 0.54 0.60 ± 0.39 1.61 ± 0.50

smaller pupil mydriasis

0.60 ± 0.36 1.31 ± 0.48 0.76 ± 0.43 1.54 ± 0.42

EMG activity(hypertonic side)

162.3 ± 45.8 150.7 ± 31.2 137.4 ± 25.4

EMG activity(hypotonic side)

111.4 ± 40.8 140.6 ± 28.5 146.9 ± 25.8

The table reports the mean ± standard deviation values obtained for pupil size, EMG activity and task-related mydriasis, at time 0 and 90 days, during normal occlusion (Bite OFF) and while subjects weared a cusp bite that reduced the difference in EMG activity between left and right masseter (Bite ON).

Fig. 5. - Average increase in pupil size elicited by the haptic task in Bite ON and OFF conditions. The height of the columns represents the mean of the differences in pupil size observed between sensorimotor task and resting condition for all the subjects tested (n = 29). The error bars correspond to the standard deviation of the corresponding values. Data have been evalu-ated separately for the larger and smaller pupil, both in normal occlusion (Bite OFF) as well as following bite correction (Bite ON).


are known to receive trigeminal signals (Shammah-Lagnado et al., 1992; Diagne et al., 2006). Finally, trigeminal input may reach the Locus Coeruleus (LC), which is regarded as “an autonomic ganglion” (Van Bockstaele and Aston-Jones, 1995) and is acti-vated in parallel with the autonomic nervous system by several sensory stimulations (Elam et al., 1986; Bradley et al., 2008), probably through the paragi-gantocellularis (PGi) nucleus (Elam et al., 1986). In this respect, LC neurons may respond to trigeminal stimulation, as shown by transcutaneous electrical stimulation of the hamster’s pinna (Zhang and Guan, 2007); moreover, LC seems to receive afferents from the trigeminal mesencephalic nucleus, or at least from a brain region included between the LC and the latter structure (Cedarbaum and Aghajanian, 1978).Noradrenergic LC neurons inhibit the preganglionic parasympathetic neurons of the Edinger-Westphal nucleus (Szabadi and Bradshaw, 1996). This inhibi-tion is necessary to increase the pupil size, as the tonic activity of the iris constrictor would prevent pupil enlargement by dilatator pupillae (Wilhelm et al., 2001). This is probably the reason why, in

monkeys, modifications in the pupil size show an impressive covariation with the changes in the dis-charge of LC noradrenergic neurons (Rajkoski et al., 1993; Sterpenich et al., 2006). Compelling evidence indicate that the same occurs in humans (Gilzenrat et al., 2010: Murphy et al., 2014).Thus, the unbalance in the pupils size induced by asymmetric trigeminal sensory and/or motor sig-nals may develop in parallel to an unbalance in the activity of LC neurons. This hypothesis is consistent with the fact that occlusal disharmony increased the release of noradrenaline in the hypothalamic para-ventricular nucleus and that such an increase was abolished by the lesion of the ascending noradrener-gic bundle arising from LC (Yoshihara and Yawaka, 2011). It can be proposed that bite correction, which reduces the asymmetry in trigeminal signals, also reduces the asymmetries in LC neurons activity.

Changes in the pupil’s diameter modulation by the haptic taskCusp bite and zircon crystal placement did not change only the basal pupil size, but also the mydria-

Fig. 6. - Effect of wearing the cusp bite on the mydriasis associated to the haptic task. In a representative subject, photographs of a pupil have been taken in resting state (basal, upper row) and sensorimotor task (task, lower row) in both Bite OFF (A) and Bite ON (B) conditions.


sis induced with the haptic task. Mydriasis depends on the participants’ involvement in the performed task (Rajkoski et al., 1993) and is strictly propor-tional to the parallel task-related phasic release of noradrenaline at cerebral cortical level (Gabay et al., 2011). This release originates from the activation of LC, which modulates cortical arousal, (Carter et al., 2010; Samuels and Szabadi, 2008) and sensorimotor excitability (Matsutani et al., 2000).It is known that phasic LC activity improves task performance, while high tonic activity is detrimen-tal (Rajkoski et al., 1993; Gilzenrat et al., 2010). Given the relation observed between pupil size and LC discharge (Rajkoski et al., 1993; Sterpenich et al., 2006, Murphy et al., 2014), it can be proposed that bite correction, which decreases pupil size dur-ing rest and increases the task associated mydriasis, improves task performance. In contrast, the zyr-con-induced malocclusion, which induces opposite effects, should deteriorate task performance.On the basis of the present findings we propose that rebalancing the activity of trigeminal afferents decreases the tonic, but enhances the phasic release of norepinephrine at brain level during a sensorimo-tor task, improving the task performance. On the other hand, unbalancing the occlusion by zircon crystals increased basal pupil size but almost can-celed the mydriasis associated with a sensorimotor task, probably being detrimental to integrative neu-ral processes underlying the task.

Present results suggest that an asymmetry in trigeminal input would affect sensorimotor performanceA case report concerning a patient affected by a neurodegenerative diseases (De Cicco, 2012b) is in line with these observations, which are also consistent with the fact that asymmetric sensory stimulations attenuate the deficits following asym-metric brain lesions. In fact, in humans, lesions of the right posterior parietal lobe induce neglect of the contralateral part of body and space (Vallar et al., 1990; Karnath et al., 1993; Rorsman et al., 1999). The observation that, in animals, a second sym-metric lesion on the opposite side greatly decreases these symptoms in spite of doubling the extension of brain damage (Lomber and Payne, 1996) indicates that the symptoms of the former lesion depend upon the unbalance created in hemispheric brain activ-

ity. In humans, these symptoms are greatly attenu-ated when the tonic activity arising from vestibular receptors or muscle proprioceptors is enhanced on the ipsilateral or contralateral side, respectively (Vallar et al., 1990; Karnath et al., 1993; Rorsman et al., 1999). This indicates that asymmetric sensory stimulation promotes the brain activity rebalancing.In conclusion, our findings indicate the existence of a tonic control exerted by sensorimotor trigeminal signals on autonomic structures modulating pupil size and disclose the possibility that these sig-nals may also modulate sensorimotor functions not involving the orofacial region.

AcknowledgementsWe are grateful to Mrs. C. Pucci for typewriting of the manuscript and to Mr. P. Orsini and F. Montanari for valuable technical assistance.

References

Bartsch T., Jänig W., Häbler H.J. Reflex patterns in preganglionic sympathetic neurons projecting to the superior cervical ganglion in the rat. Auton. Neurosci., 83: 66-74, 2000.

Bourque M.J. and Kolta A. Properties and inter-connections of trigeminal interneurons of the lateral pontine reticular formation in the rat. J. Neurophysiol., 86: 2583-2596, 2001.

Bradley M.M., Miccoli L., Escrig M.A., Lang P.J. The pupil as a misure of emotional arousal and autonomic activation. Psychophysiology, 45: 602-607, 2008.

Bradshaw J. Pupil size as a measure of arousal dur-ing information processing. Nature, 216: 515-516, 1967.

Breen L.A., Burde R.M., Loewy A.D. Brainstem connections to the Edinger-Westphal nucleus of the cat: a retrograde tracer study. Brain Res., 261: 303-306, 1983.

Carter M., Yizhar O., Chikahisa S., Nguyen H., Adamantidis A., Nishino S., Deisseroth K., De Lecea L. Tuning arousal with optogenetic modula-tion of Locus Coeruleus neurons. Nat. Neurosci., 13: 1526-1533, 2010.

Cedarbaum J.M. and Aghajanian G.K. Afferent pro-jections to the rat locus coeruleus as determined by a retrograde tracing technique. J. Comp. Neurol., 178: 1-16, 1978.


Cooper B.C., Cooper D.L., Lucente F.E. Electromyography of masticatory muscles in cranio-mandibular disorders. Laryngoscope, 101: 150-157, 1991.

Dao T.T., Lavigne G.J., Charbonneau A., Feine J.S., Lund J.P. The efficacy of oral splints in the treat-ment of myofascial pain of the jaw muscles: a controlled clinical trial. Pain, 56: 85-94, 1994.

De Cicco V. Blood flow variations of cerebro-afferent vessels and of pupillary basal diameters induced by stomatognathic trigeminal proprioception: a case report. J. Med. Case Rep., 6: 275-282, 2012a.

De Cicco V. Central syntropic effects elicited by trigeminal proprioceptive equilibrium in subjects affected by Alzheimer Disease: a case report. J. Med. Case Rep., 6: 161-169, 2012b.

Diagne M., Valla J., Delfini C., Buisseret-Delmas C., Diague P.B. Trigeminovestibular and trigemino-spinal pathways in rats: Retrograde tracing com-pared with glutamic acid decarboxylase and gluta-mate immunohistochemistry. J. Comp. Neurol., 496: 759-772, 2006.

Dworkin S.F., Fricton J., Hollender L., Huggins K., LeResche L., Lund J., Mohi N., Ohrbach R., Palla S., Sommers E.E., Stohler C., Von Korff M., Widmer C.G. Research diagnostica criteria for temporomandibolar disorders: review, crite-ria, examinations and specifications critique. J. Craniomandib. Disord., 6: 301-355, 1992.

Elam M., Svensson T.H., Thoren P. Locus coeruleus neurons and sympathetic nerves: activation by cutaneous sensory afferents. Brain Res., 366: 254-261, 1986.

Esser M.J., Pronych S.P., Allen G.V. Trigeminal-reticular connections: Possible pathways for noci-ception-induced cardiovascular reflex responses in the rat. J. Comp. Neurol., 391: 526-544, 1998.

Gabay S., Pertzov Y., Henik A. Orienting of atten-tion, pupil size and the norepinephrine system. Atten. Percept. Psichophys., 73: 123-129, 2011.

Gilzenrat M.S., Nieuwenhuis S., Jepma M., Cohen J.D. Pupil diameter tracks changes in control state predicted by the adaptive gain theory of locus coeruleus function. Cogn. Affect. Behav. Neurosci., 10: 252-269, 2010.

Gomez C.E. and Christensen L.V. Stimulus-response latencies of two instruments delivering trans-cutaneous electrical neuromuscular stimulation (TENS). J. Oral Rehabil., 18: 87-94, 1991.

Hess E.H. and Polt J.M. Pupil size in relation to mental activity during simple problem-solving. Science, 143: 1190-1192, 1964.

Karnath H.O., Christ K., Hartje W. Decrease of con-tralateral neglect by neck muscle vibration and spatial orientation of trunk midline. Brain, 116: 383-396, 1993.

Korszun A., Young E.A., Singer K., Carlson N.E., Brown M.B., Crofford L. Basal circadian cortisol secretion in women with Temporo-mandibular dis-orders. J. Dental. Res., 81: 279-283, 2002.

Light K., Bragdon E., Grewen K., Brownley K., Girdler S., Maixner W. Adrenergic dysregulation and pain with and without acute beta-blockade in women with fibromyalgia and temporomandibular disorder. J. Pain, 10: 542-252, 2009.

Lomber S.G. and Payne B.R. Removal of two halves restores the whole: reversal of visual hemineglect during bilateral cortical or collicular inactivation in the cat. Vis. Neurosci., 13: 1143-1156, 1996.

Matsutani K., Tsuruoka M., Shinya A., Furuya R., Kawawa T. Stimulation of the Locus Coeruleus suppresses trigeminal sensorimotor function in the rat. Brain Res. Bull., 53: 827-832, 2000.

Monaco A., Cattaneo R., Mesin L., Ciarrocchi I., Sgolastra F., Pietropaoli D. Dysregulation of the autonomous nervous system in patients with tem-poromandibular disorder: a pupillometric study. Plos One, 7: e45424, 2012.

Murphy P.R., O’Connell R.G., O’Sullivan M., Robertson I.H., Balsters J.H. Pupil diameter cova-ries with BOLD activity in human locus coeruleus. Hum. Brain Mapp., [Epub ahead of print], 2014.

Nnoaham K.E., Kumbang J. Transcutaneous electri-cal nerve stimulation (TENS) for chronic pain. Cochrane Database Syst. Rev., Issue 3, 2008.

Notsu K., Tumori T., Yokota S., Semine J., Yasui Y. Posterior lateral hypothalamic axon terminal are in contact with trigeminal premotor neurons in the parvicellular reticular formation of the rat medulla oblongata. Brain Res., 1244: 71-81, 2008.

Rajkoski J., Kubiak P., Aston-Jones G. Correlations between locus coeruleus (LC) neural activity, pupil diameter and behaviour in monkey support a role of LC in attention. Pro. Soc. Neurosci. Abs., 19: 974, 1993.

Rorsman I., Magnusson M., Johansson B. Reduction of visuo-spatial neglect with vestibular galvanic stimulation. Scand. J. Rehabil. Med., 31: 117-124, 1999.

Samuels E.R. and Szabadi E. Functional neuroanato-my of the noradrenergic locus coeruleus: its roles in the regulation of arousal and autonomic function part I: principles of functional organisation. Curr. Neuropharmacol., 6: 235-253, 2008.


Sato A. and Schmidt R.F. Somatosympathetic reflex-es: afferent fibers, central pathways, discharge characteristics. Physiol. Rev., 53: 916-947, 1973.

Shammah-Lagnado S.J., Costa M.S., Ricardo J.A. Afferent connections of the parvocellular reticular formation: a horseradish peroxidase study in the rat. Neuroscience, 50: 403-425, 1992.

Sterpenich V., D’Argembeau A., Desseilles M., Balteau E., Albouy G., Vandewalle G., Degueldre C., Luxen A., Collette F., Maquet P. The Locus Coeruleus is involved in the successful retrieval of emotional memories in humans. J. Neurosci., 26: 7416-7423, 2006.

Szabadi E. and Bradshaw C. Autonomic pharmacol-ogy of D2-adrenoceptors. J. Physicopharmacol., 10 (Suppl 3): s6-18, 1996.

Vallar G., Sterzi R., Bottini G., Cappa S., Rusconi M. Temporary remission of left hemianesthesia after

vestibular stimulation. A sensory neglect phenom-enon. Cortex, 26: 123-131, 1990.

Van Bockstaele E.J. and Aston-Jones G. Integration in the ventral medulla and coordination of sym-pathetic, pain and arousal functions. Clin. Exp. Hypertens., 17: 153-165, 1995.

Yoshihara T. and Yawaka Y. Lesion of the ven-tral ascending noradrenergic bundles decrease the stress response to occlusal disharmony in rats. Neurosci. Lett., 503: 43-47, 2011.

Wilhelm B., Giedke H., Lüdtke H., Bitter E., Hofmann A., Wilheim H. Daytime variations in central nervous system activation measured by a pupillographic sleepness test. J. Sleep Res., 10: 1-7, 2001.

Zhang J. and Guan Z. Pathways involved in soma-tosensory electrical modulation of dorsal cochlear nucleus activity. Brain Res., 1184: 121-131, 2007.

Introduction

Carpal tunnel syndrome (CTS) is one of the most common compressive neuropathies (Aroori and Spence, 2008). It is more common in women than in men, with a ratio of 3:1, most frequently between 40 and 60 years of age (Aroori and Spence, 2008). Recent findings have pointed out that CTS has the

potential to significantly limit the performance of activities of daily living (Massy-Westropp et al., 2000). In the work of Ansari et al. (2009) significant correlation between pain and paresthesia with chang-es in distal sensory latency was observed, while non significant correlation was noticed between positive Tinel’s sign with changes in distal sensory and/or distal motor latencies.

Electrodiagnostic evaluation of patients with carpal tunnel syndrome regarding the presence

of subjective and physical findingsD. RADOVIC1, M. LAZOVIC1,2, D. NIKOLIC3, I. PETRONIC2,3,

N. RADOSAVLJEVIC1, M. HRKOVIC1

1 Institute for Rehabilitation, Belgrade, Serbia; 2 Faculty of Medicine, University of Belgrade, Belgrade, Serbia; 3 Physical Medicine and Rehabilitation Department,

University Children’s Hospital, Belgrade, Serbia

A B S T R A C T

The aim of our study was to evaluate the changes of median nerve conduction velocities by electrodiagnostic pro-cedure in carpal tunnel syndrome (CTS) patients with and without subjective and physical findings. We have evalu-ated 116 patients that were diagnosed with CTS. Subjective findings: weakness, numbness and night pain were analyzed. Further physical findings were evaluated: Tinel’s sign, muscles hypotrophy and weakness according to muscle manual test (MMT). Duration of complaints was evaluated as well. Electroneurographic findings included: estimation of median nerve motor terminal latency (mMTL) and median nerve sensory terminal latency (mSTL), sensory velocity (mSV) and motor velocity (mMV). The patients who experienced night pain (p = 0.015) and those with muscles weakness on MMT (p = 0.016) had complained for significantly longer period. Statistically significant increase for mMTL values was noticed for patients with Tinel’s sign (p < 0.045), present muscles hypotrophy (p < 0.001) and weakness on MMT (p < 0.001). There is significant decrease for mMV in the group with present Tinel’s sign (p = 0.048), muscle hypotrophy (p = 0.003) and weakness on MMT (p = 0.002), and for mSV in the group with present muscle hypotrophy (p = 0.008) and group with weakness on MMT (p = 0.019). Multivariate logistic regressional analysis shown that only for hypotrophy, mMTL variable presents significant independent contributor (p = 0.009). For the diagnosis confirmation and treatment planning along with elecroneurography it is necessary to evaluate patients with CTS clinically, since different clinical manifestations are correlating in different degree with electroneurographic findings.

Key wordsCarpal tunnel syndrome • Electroneurography • Clinical findings

Corresponding Author: Diana Radovic, MD, Sokobanjska 17, 11000 Belgrade, Serbia - Tel.: +381638133345 - Email: [email protected]


14 D. RADOVIC ET AL.

Quick and reliable diagnosis of CTS can be made on the basis of characteristic history, physical exami-nation and electro-diagnostic findings (Lee et al., 2013). Early diagnosis and timely treatment induc-tion are important factors that could alter the course of CTS. However, as it was stated in the study of Gomes et al. (2006), that CTS can be present in patients without apparent symptoms as well in patients with negative neurophysiological studies. The CTS, discovered late and untreated, leads to the permanent damage of the median nerve (Mackinnon et al., 2000).We hypothesized that changes on median nerve due to the compression could be reflected on the pres-ence of subjective and/or physical findings. Clark et al., (2011) have indicated that sensory symptoms appear on index finger in 94% of evaluated patients with CTS. Therefore, the aim of our study was to evaluate the changes of median nerve conduction velocities by electrodiagnostic procedure in CTS patients with and without present subjective and physical findings.

Material and methods

Study groupWe have evaluated 116 patients 85 (73.3%) of which were females and 31 (26.7%) males, age of 55.04 ± 12.13 years that were referred to the Institute for Rehabilitation for diagnosis and treatment of CTS from February 2012 to February 2013. Prior to the inclusion in the study, the eligible participants were informed about the study protocol and they gave their informed consents. The study was approved by the Institutional Review Board and was carrying out according to the principles of good clinical practice. Patients with injuries and/or surgical interventions on upper extremities were excluded from the study.

Study parametersFurther parameters were analyzed: subjective find-ings including presence or absence of weakness, numbness and night pain. The presence or absence of Tinel’s sign, muscles hypotrophy and weak-ness according to muscle manual test (MMT) were evaluated on physical examination. Duration of complaints was evaluated for each of the above mentioned parameters.

Electroneurographic findings included: estimation of median nerve motor terminal latency (mMTL), median nerve sensory terminal latency (mSTL), median nerve sensory velocity (mSV) and median nerve motor velocity (mMV). The technique includ-ed the use of stimulative percutaneous electrodes and findings were obtained by surface electrodes. In order to achieve optimal results, patients were told to be relaxed, while they were placed in supine posi-tion with the room temperature between 22oC and 24oC (Matanovic et al., 2013). Previous reports indi-cated that temperature of extremities could influence nerve conduction velocities, thus the upper extremi-ties were warmed up to the temperature between 32oC and 36oC (Rutkove, 2001) before the examina-tion. mMTL was registered above abductor pollicis brevis muscle by stimulation on wrist joint at the distance of 7 cm (Cherian and Kuruvilla, 2006). Even though there are no exact tests for diagnostic accuracy to confirm CTS, we followed the clini-cal practice guidelines proposed by the American Academy of Orthopaedic Surgeons (AAOS, 2007). For the confirmation of CTS, we used mMTL values > 4.2 ms (Cherian and Kuruvilla, 2006). Sensory velocity was measured by stimulative electrode that was placed on the wrist joint and detection elec-trode on the point finger. For the confirmation of CTS, we used values for sensory fibers velocities of median nerve < 50 m/s, sensory latencies > 3.7 ms and amplitudes < 20 uV (DeLisa, 1983, Glowacki et al., 1996). Although motor velocities in majority of patients with CTS are within physiological ranges, we used velocities < 50 m/s as pathological values. To exclude polyneuropathy, we performed analysis of sensory and motoric conduction velocities of both median and ulnar nerves, and for radiculopathy exclusion for roots C8/Th1, we performed electro-myography testing by needle electromyography on abductor digiti minimi muscle.

Statistical analysisWe used whole numbers for the frequencies of evaluated parameters within subjective findings and those obtained on physical examination. Mean val-ues with standard deviation (SD) was used for inter-pretation of complaints duration, mMTL, mSTL, median sensory and motor velocities. Student’s t-test was used for assessing presence of statistical significance between evaluated parameters with

CARPAL TUNNEL SYNDROME AND ELECTRODIAGNOSTICS 15

normal distribution. In cases where there was not normal distribution of evaluated parameters for sta-tistical evaluation we used Mann-Whitney U test. For estimation of evaluated factors influence on electroneurographic findings we performed univari-ate logistic regression analysis. Multivariate logistic regression analysis was done for those parameters that were proven with significant importance on univariate analysis testing. Statistical significance was set at p < 0.05.

Results

Significantly longer complaints were present in patients who had subjective experience of night pain (p = 0.015) and those with muscles weakness on MMT (p = 0.016) (Table I). It was noticed that for subjective complaints there were no significant differences in mMTL values between groups of patients with and without evaluated parameters (p > 0.05), while statistically significant increase for mMTL values was noticed for patients with Tinel’s sign (p < 0.045), present muscles hypotrophy (p < 0.001) and weakness on MMT (p < 0.001) (Table I). There were no significant differences between evaluated values regarding presence or absence of both subjective and physical findings for mSTL (p > 0.05), thus univariate logistic regression was not performed for mSTL (Table I).

Concerning sensory and motor velocity values for parameters within subjective complaints there were non-significant changes between groups with and without present parameters (p > 0.05) (Table II). Significant decrease in motor velocities regarding evaluated parameters on physical examination was noticed for the group of patients with all positive pathological findings (for mMV: group with present Tinel’s sign (p = 0.048), muscle hypotrophy (p = 0.003) and weakness on MMT (p = 0.002)), while for mSV only in the group with present muscle hypotrophy (p = 0.008) and the group with weakness on MMT (p = 0.019) (Table II).Univariate logistic regressional analysis of the data for subjective findings revealed that mMV and mSV variables are significantly associated with the pres-ence of numbness, while duration of complaints variable significantly correlates with the night pain symptom (Table III). For the group of physical find-ings, it was shown that mMTL, mMV and mSV variables are significantly associated with the pres-ence of hypotrophy and decreased strength evalu-ated by MMT (weakness on MMT) (Table III). It was noticed as well that the duration of complaints variable significant correlates with weakness on MMT (Table III).Concerning the fact that majority of patients expe-rienced the numbness, the multivariate analysis was not performed for this parameter. On multivariate logistic regressional analysis it was shown that only

Table I. - Frequencies, duration and motor terminal latency values of evaluated parameters.

N Duration of complaints p value* mMTL

MV ± SD p value* mSTLMV ± SD p value*

Subjective findings

Numbness No 6 10.33 ± 12.88 0.449 6.92 ± 3.72 0.562 3.69 ± 0.54 0.159

Yes 110 10.55 ± 8.58 5.86 ± 1.73 4.00 ± 0.98

Weakness No 90 10.32 ± 9.21 0.230 5.85 ± 1.96 0.142 3.99 ± 0.95 0.344

Yes 26 11.27 ± 7.14 6.12 ± 1.49 4.02 ± 1.10

Night pain No 48 8.50 ± 7.67 0.015 6.19 ± 1.99 0.131 4.18 ± 1.23 0.847

Yes 68 11.97 ± 9.26 5.71 ± 1.76 3.87 ± 0.75

Physical findings

Tinels sign - 28 9.86 ± 9.38 0.284 5.64 ± 1.76 0.045 3.73 ± 0.39 0.625

+ 56 11.75 ± 9.08 6.45 ± 2.05 4.23 ± 1.24

Hypotrophy No 93 9.82 ± 7.47 0.257 5.52 ± 1.29 < 0.001 3.94 ± 1.02 0.873

Yes 22 13.95 ± 12.60 7.63 ± 2.81 4.24 ± 0.74

Weakness (MMT) No 73 8.88 ± 6.87 0.016 5.43 ± 1.31 < 0.001 3.89 ± 0.95 0.765

Yes 42 13.62 ± 10.80 6.78 ± 2.35 4.18 ± 1.01

mMTL = median nerve motor terminal latency; mSTL = median nerve sensory terminal latency; MMT = muscle manual test; * Mann Whitney U test


for hypotrophy parameter, mMTL variable presents significant independent contributor aside mMV and mSV (Table IV).There is a significant correlation between dura-tion of complaints and terminal latency, where longer duration was associated with increasement of mMTL (p = 0.013) (Table V). Significant cor-relation was found between duration of complaints and motor velocity changes, where longer duration was associated with reduction in motor velocity (p

= 0.016) (Table V). Non-significant correlation was found between duration of complaints and sensory velocity changes (p = 0.380) (Table V).

Discussion

Even though in the study of Macdonell et al., (1990) it was noticed that abnormalities of sensory veloci-ties when median nerve was stimulated were signifi-

Table II. - Sensory and motor velocity values of evaluated parameters.

N mSVMV ± SD p value mMV

MV ± SD p value

Subjective findings

Numbness No 6 50.85 ± 3.91 p = 0.259* 58.38 ± 6.75 p = 0.062*

Yes 110 39.41 ± 8.25 52.91 ± 5.87

Weakness No 90 39.69 ± 8.16 p = 0.519* 53.24 ± 5.93 p = 0.763**

Yes 26 40.83 ± 9.54 52.83 ± 6.28

Night pain No 48 38.94 ± 9.43 p = 0.385** 52.87 ± 7.20 p = 0.689**

Yes 68 40.57 ± 7.79 53.35 ± 5.01

Physical findings

Tinels sign - 28 42.29 ± 6.68 p = 0.110** 54.72 ± 4.44 p = 0.048**

+ 56 38.36 ± 9.98 51.95 ± 6.58

Hypotrophy No 93 40.97 ± 8.40 p = 0.008** 53.96 ± 5.41 p = 0.003**

Yes 22 34.25 ± 6.73 49.56 ± 7.35

Weakness (MMT) No 73 41.36 ± 8.55 p = 0.019** 54.48 ± 5.57 p = 0.002**

Yes 42 36.72 ± 7.56 50.81 ± 6.11

mSV = median nerve sensory velocity; mMV = median nerve motor velocity; MMT = muscle manual test; *Mann Whitney U test; ** Students t test

Table III. - Univariate logistic regression analysis of association between subjective and physical findings with duration of complaints and findings on electroneurography.

Subjective findings

Numbness Weakness Night pain

OR (95% CI) p OR (95% CI) p OR (95% CI) p

Duration of complaints 1.003 (0.911-1.104) 0.954 1.012 (0.965-1.062) 0.627 1.054 (1.002-1.108) 0.041

mMTL 0.806 (0.583-1.114) 0.192 1.074 (0.861-1.340) 0.528 0.869 (0.710-1.065) 0.177

mMV 0.853 (0.731-0.994) 0.041 0.989 (0.918-1.064) 0.761 1.014 (0.952-1.079) 0.668

mSV 0.790 (0.657-0.950) 0.012 1.017 (0.955-1.082) 0.603 1.024 (0.972-1.078) 0.382

Physical findings

Tinels sign Hypotrophy Weakness on MMT



mMTL 1.307 (0.961-1.778) 0.088 1.759 (1.312-2.359) 0.001 1.560 (1.195-2.036) 0.001

mMV 0.913 (0.833-1.001) 0.052 0.875 (0.790-0.968) 0.010 0.884 (0.812-0.962) 0.004

mSV 0.948 (0.888-1.013) 0.115 0.912 (0.849-0.981) 0.013 0.935 (0.882-0.992) 0.025mMTL = median nerve motor terminal latency; mSV = median nerve sensory velocity; mMV = median nerve motor velocity; MMT = muscle manual test


cantly lower on index finger, Wee and Abernathy (2003) suggested that neither index nor middle fin-ger evaluation might provide significant differences in sensory conduction abnormalities when median nerve was tested in patients with CTS. Different frequencies of positive findings regarding prolonged sensory latency in patients with mild, moderate and severe types of CTS were described in the study of Banach and Szczudlik (2004). For mild CTS forms significantly frequent were positive findings (pro-longed sensitive latency) on thumb finger, while for those with moderate and severe CTS, they found non significant difference between fingers (Banach and Szczudlik, 2004). However, it was underlined that the gold standard for the diagnosis of CTS considers to be electromyographical and electroneurographi-cal testing due to its high sensitivity (49-84%) and specificity (95-99%) (Robinson, 2007). In the event of nerve compression, pathological action potentials indicate denervation process and different degrees of reduction route at maximum voluntary contraction. Action potentials depend on the number of damaged and the remaining axons (Robinson, 2007).Our results demonstrated that the most frequent sub-jective finding was numbness (94.8%), followed by the night pain (58.6%). These findings are consistent with previous reports, where non-painful sensory disturbances in CTS patients were detected more

frequently than pain (Clark et al., 2011)Previously it was pointed out that electrodiagnostic studies do not correlate closely with the CTS sever-ity, particularly between mild to moderate degrees (Martin et al., 2005).Our findings are consistent to the certain degree with such observations particularly in terms of findings related to subjective parameters (numbness, weak-ness and night pain), where we found non-significant differences in mMTL, mSTL, mSV and mMV mean values between participants with and those without evaluated subjective parameters. When univariate logistic regression model was performed, we have pointed out that there is a significant correlation between mSV and mMV and numbness. Therefore, even though we found non-significant prolongation in mSV and mMV, in patients with CTS who expe-rience numbness, electrodiagnostic testing for mSV and mMV should be performed. These findings are consistent with previous reports, where it was stated that numbness and nocturnal pain are strong indi-cators for CTS (Gupta and Benstead, 1997). As a strong predictor of changes in mSV, particularly for median nerve, numbness was evaluated previously in the study of Ntani et al. (2013).Further, we have noticed that the duration of com-plaints is significantly associated with the presence of the night pain. Considering subjective parameters,

Table IV. - Multivariate logistic regression analysis of association between subjective and physical findings with duration of complaints and findings on electroneurography.

Subjective findings Physical findings

Night pain Hypotrophy Weakness on MMT



mMTL 1.576 (0.957-2.595) 0.074 2.370 (1.242-4.522) 0.009 1.612 (0.974-2.669) 0.063

mMV 0.972 (0.867-1.090) 0.626 0.952 (0.816-1.110) 0.528 0.930 (0.828-1.045) 0.225

mSV 1.055 (0.975-1.142) 0.180 0.954 (0.876-1.039) 0.277 0.974 (0.908-1.045) 0.460

mMTL = median nerve motor terminal latency; mSV = median nerve sensory velocity; mMV = median nerve motor velocity; MMT = muscle manual test

Table V. - Correlation between evaluated parameters and complaints duration.

Parameters N r p value

mMTL/Duration 116 0.229 0.013

mMV/Duration 114 -0.224 0.016

mSV/Duration 86 -0.096 0.380

mMTL = median nerve motor terminal latency; mSV = median nerve sensory velocity; mMV = median nerve motor velocity; r = Correlation coefficient


our findings could lead to the hypothesis that aside the pressure (per se) on sensory fibers of the median nerve in CTS, other factors might play to the certain degree some kind of a role in pain mechanism. Previous reports underlined the possibility of central sensitiza-tion mechanisms, somato-sensory impairments and even motor system modulation in pathogenesis of pain in patients with CTS (Fernandez-de-las-Penas et al., 2009; Tamburin et al., 2008; Tucker et al., 2007; Zanette et al., 2010). However, it should be kept in mind that central sensitization is a dynamic process which is under the influence of multiple factors, among them peripheral nocioceptive inputs, stress-ing out the role of peripheral and central sensitization mechanisms in CTS (de-la-Llave-Rincon et al., 2012).Given the facts above, considering clinical implica-tions for referring the patients with CTS and subjec-tive parameters that were evaluated in this study, there is a weak probability of acquiring information on nerve conduction studies regarding the prolon-gation of velocities in the event where subjective parameters (weakness and night pain) are both posi-tive or negative. Thus, it could be considered that electroneurography is of little diagnostic value for these subjective parameters in CTS patients.Even though we found that there are significant changes in mMTL and mMV in CTS patients with and without Tinel’s sign, further univariate analysis revealed that none of these variables (mMTL and mMV) are significantly associated with Tinel’s sign in patients with CTS. These findings give impres-sion for possible assumption that Tinel’s sign could not be considered as specific test for the evaluation of pathophysiological changes on affected median nerve in patients with CTS. Previous reports have stressed out that Tinel’s sign could be considered as more sensitive and specific test for the diagnosis of tenosynovitis of the hand rather than for CTS (El Miedany et al., 2008; Ibrahim et al., 2012).From the results of our study hypotrophia and weakness on MMT are shown to correlate more significantly with changes in mMTL, and motor and sensory conduction velocities of median nerve. When multivariate logistic regression analysis was preformed, significant independent factor for pro-longation in mMTL was shown only to be the muscle hypotrophy. These observations point out that muscle hypotrophy could be considered as more sensitive parameter then Tinel’s sign and weakness

on MMT in changes on electroneurographic evalua-tions in patients with CTS.We have demonstrated as well that as the duration of complaints is longer the more significant changes will be noticed for the mMTL and mMV, giving impression that motor fibers of median nerve are more sensitive to the influence of pathological pro-cess leading to the CTS.In the study of Chang et al., (2006), it was stated that motor conduction velocities evaluation is equal or even more sensitive than sensory conduction velocities investigation in the wrist to palm area for the patients with CTS. Our results correlate to the certain degree with such findings, where changes in motor velocities significantly correlated with more symptoms that were investigated then sensory con-duction velocities of median nerve for the patients with CTS.There are a few limitations in the study, particu-larly referring to the smaller number of eligible par-ticipants, thus further studies are warrant on larger population. Also, further studies could consider evaluation of changes regarding above mentioned parameters in patients considering dominant hand. Another limitation refers to the fact that using data from sensory nerve conduction parameters of differ-ent fingers could add more valuable results of the CTS evaluation to this study.In the conclusion, despite numerous statements regarding significance of electrodiagnostics evalua-tion in patients with CTS, our findings stressed out that for the diagnosis confirmation and treatment planning along with elecroneurography it is neces-sary to evaluate patients with CTS clinically. It is also important to underline that individual approach in the diagnosis of CTS is of great importance since an adequate treatment implementation will ultimate-ly bring to the best possible treatment outcome and better quality of life for these patients.

References

Ansari N.N., Adelmanesh F., Naghdi S., Mousavi S. The relationship between symptoms, clinical tests and nerve conduction study findings in carpal tun-nel syndrome. Electromyogr. Clin. Neurophysiol., 49: 53-57, 2009.

Aroori S. and Spence R.A. Carpal tunnel syndrome. Ulster Med. J., 77: 6-17, 2008.


Banach M. and Szczudlik A. The value of sensory latency measured on the four median innervated fingers for the diagnosis of the carpal tunnel syn-drome. Neurol. Neurochir. Pol., 38: 15-20, 2004.

Chang M.H., Liu L.H., Lee Y.C., Wei S.J., Chiang H.L., Hsieh P.F. Comparison of sensitivity of transcarpal median motor conduction velocity and conventional conduction techniques in elec-trodiagnosis of carpal tunnel syndrome. Clin. Neurophysiol., 117: 984-991, 2006.

Cherian A. and Kuruvilla A. Electrodiagnostic approach to carpal tunnel syndrome. Ann. Indian Acad. Neurol., 9: 177-182, 2006.

Clark D., Amirfeyz R., Leslie I., Bannister G. Often atypical? The distribution of sensory disturbance in carpal tunnel syndrome. Ann. R. Coll. Surg. Eng., 93: 470-473, 2011.

de-la-Llave-Rincón A.I., Puentedura E.J., Fernández-de-las-Peñas C. New advances in the mechanisms and etiology of carpal tunnel syndrome. Discov. Med., 13: 343-348, 2012.

DeLisa J. Nerve Conduction Velocity Techniques. Raven Press, New York, 1983 (p. 33-39).

El Miedany Y., Ashour S., Youssef S., Mehanna A., Meky F.A. Clinical diagnosis of carpal tunnel syn-drome: old tests-new concepts. Joint Bone Spine, 75: 451-457, 2008.

Fernández-de-las-Peñas C., De-la-Llave-Rincón A.I., Fernández-Carnero J., Cuadrado M., Arendt-Nielsen L., Pareja J. Bilateral widespread mechan-ical pain sensitivity in carpal tunnel syndrome: Evidence of central processing in unilateral neu-ropathy. Brain, 132: 1472-1479, 2009.

Gomes I., Becker J., Ehlers J., Nora D.B. Prediction of the neurophysiological diagnosis of carpal tun-nel syndrome from the demographic and clinical data. Clin. Neurophysiol., 7: 964-971, 2006.

Glowacki K.A., Breen C.J., Sachar K., Weiss A.P. Electrodiagnostic testing and carpal tunnel release outcome. J. Hand Surg. Am., 21: 117-21, 1996.

Gupta S.K. and Benstead T.J. Symptoms experienced by patients with carpal tunnel syndrome. Can. J. Neurol. Sci., 24: 338-342, 1997.

Ibrahim I., Khan W.S., Goddard N., Smitham P. Carpal tunnel syndrome: a review of the recent literature. Open Orthop. J., 6: 69-76, 2012.

Lee H.J., Kwon H.K., Kim D.H., Pyun S.B. Nerve conduction studies of median motor nerve and median sensory branches according to the severity of carpal tunnel syndrome. Ann. Rehabil. Med., 37: 254-62, 2013.

Macdonell R.A., Schwartz M.S., Swash M. Carpal tunnel syndrome: which finger should be tested? An analysis of sensory conduction in digital branches of the median nerve. Muscle Nerve, 13: 601-606, 1990.

Mackinnon S.E., Novak C.B., Landau W.M. Clinical diagnosis of carpal tunnel syndrome. JAMA, 284: 1924-1926, 2000.

Martin B.I., Levenson L.M., Hollingworth W., Kliot M., Heagerty P.J., Turner J.A., Jarvik J.G. Randomized clinical trial of surgery ver-sus conservative therapy for carpal tunnel syn-drome [ISRCTN84286481]. BMC Musculoskelet. Disord., 6: 2, 2005.

Massy-Westropp N., Grimmer K., Bain G. A sys-tematic review of the clinical diagnostic tests for carpal tunnel syndrome. J. Hand Surg., 25: 120-127, 2000.

Matanovic D., Popovic S., Parapid B., Petronic I., Nikolic D. Neurophysiological evaluation in newly diagnosed diabetes mellitus type 1. Cent. Eur. J. Med., 8: 503-508, 2013.

Ntani G., Palmer K.T., Linaker C., Harris E.C., Van der Star R., Cooper C., Coggon D. Symptoms, signs and nerve conduction velocities in patients with suspected carpal tunnel syndrome. BMC Musculoskelet. Disord., 14: 242, 2013.

Robinson L.R. Electrodiagnosis of Carpal Tunnel Syndrome. Phys. Med. Rehabil. Clin. N. Am., 18: 733-746, 2007.

Rutkove S.B. Effects of temperature on neuromuscular electrophysiology. Muscle Nerve, 24: 867-882, 2001.

Tamburin S., Cacciatori C., Marani S., Zanette G. Pain and motor function in carpal tunnel syn-drome: A clinical, neurophysiological and psycho-physical study. J. Neurol., 255: 1636-1643, 2008.

Tucker A.T., White P.D., Kosek E., Pearson R.M., Henderson M., Coldrick A.R., Cooke E.D., Kidd B.L. Comparison of vibration perception thresholds in individuals with diffuse upper limb pain and carpal tunnel syndrome. Pain, 127: 263-269, 2007.

Zanette G., Cacciatori C., Tamburin S. Central sen-sitization in carpal tunnel syndrome with extrater-ritorial spread of sensory symptoms. Pain, 148: 227-236, 2010.

Wee A.S., Abernathy S.D. Carpal tunnel syndrome: comparison of the median sensory nerve conduc-tion findings from the index and middle fingers. Electromyogr. Clin. Neurophysiol., 43: 251-253, 2003.

www.aaos.org/research/guidelines/CTS_guideline.pdf

Introduction

Imitation is a social behavior that contributes to the transmission of intentions and to creating mental and social links between humans; it can be considered as copying something seen, naturally or artificially elic-ited by others’ acting (Mühlau et al., 2005). When individuals are asked to imitate the movements of someone standing in front of them, they can act in at least two ways. In anatomical imitation the imitator activates exactly the same effectors (hence the same

nervous mechanisms) as the model, i.e., the muscles that raise the right arm to mimic the model’s lifted right arm. In the second mode, defined as specular or mirror mode, the imitator activates the effectors sharing an external spatial reference with those acti-vated by the model, i.e., raising the left arm to mimic the model’s lifted right arm (Koski et al., 2003; Franz et al., 2007; Press et al., 2009).In the past, the theme of imitation has been addressed in relation to several aspects: some investigators focused on imitative performances with respect to

Anatomical or mirror mode imitation?A behavioral approach

C. PIERPAOLI1, L. FERRANTE2, T. MANZONI1†, M. FABRI1

1 Dipartimento di Medicina Sperimentale e Clinica, Facoltà di Medicina e Chirurgia, Università Politecnica delle Marche, Ancona, Italy; 2 Dipartimento di Scienze Biomediche e Sanità Pubblica,

Facoltà di Medicina e Chirurgia, Università Politecnica delle Marche, Ancona, Italy

A B S T R A C T

Imitation can occur in at least two forms: one, which can be defined as anatomical, is based primarily on the men-tal construct of the body schema and allows recognition of correspondences between own body anatomy and that of other individuals. The other form, defined as specular or mirror mode, is most probably based on the allocation of some form of attention to the same region of the environmental space both by model and imitator, and to the objects it contains.This study investigated the behavioral strategy of imitation in normal subjects, to assess whether they carried out task instructions using an anatomical or a mirror perspective.Twenty seven adults were asked to imitate intransitive meaningful and meaningless gestures shown by a model in video clips. Instructions about how to perform them were provided before each trial. Trials were free (intended to produce mirror imitation) or driven (intended to produce anatomical imitation); further driven trials were administered to verify participants’ knowledge of bodily laterality and were used as control. Performances were interpreted as anatomical or mirror imitation, according to the observation of anatomical or spatial reference frames between stimulus and imitator.The results revealed that in spontaneous imitation the mirror mode was more frequent (61% of responses), in line with previous studies. The novel finding was the prevalence (93% of responses) of anatomical imitation in tasks involving detailed driven instructions.

Key wordsImitation • Mirror mode • Anatomical mode • Mental Rotation

Corresponding Author: Dr. Chiara Pierpaoli, Dipartimento di Medicina Sperimentale e Clinica, Sezione di Neuroscienze e Biologia cellulare, Università Politecnica delle Marche, via Tronto 10/A, 60020 Ancona, Italy - Tel.: +39 071 220 6193/6050 - Fax: +39 071 220 6052 - Email: [email protected]


ANATOMICAL OR MIRROR MODE IMITATION? A BEHAVIORAL APPROACH 21

stimulus type (hand/head vs. finger configurations; Lausberg and Cruz, 2004; novel vs. well known actions; Rumiati and Tessari, 2002; transitive vs. intransitive gestures; Carmo and Rumiati, 2009; biological vs. non biological movements; Crescentini et al., 2011; object-oriented vs. non object-oriented movements; Wohlschläger and Bekkering, 2002). Other researchers concentrated on stimulus position with respect to the observer (e.g., degrees of rotation at 0°, 180°; Press et al., 2009; Krause and Kobow, 2013). Further studies explicitly instructed partici-pants on how to execute imitative performances (e.g., anatomical vs. mirror perspective-taking; Sambrook, 1998; Bekkering et al., 2000; Anquetil and Jeannerod, 2007). Still other researchers focused on a combina-tion of these factors (Stürmer et al., 2000; Heyes and Ray, 2004; Kitada et al., 2010). The aspect shared by most of these investigations is the evaluation of reac-tion times and levels of response accuracy.At variance with those studies the present investiga-tion was undertaken to establish in which conditions a subject uses anatomical vs. mirror perspective-taking.Although various approaches have been proposed to explain the mechanisms whereby subjects adopt an anatomical or a mirror imitative response, their results are not concordant (Schofield, 1976; Gleissner et al., 2000; Iacoboni et al., 2001).A tendency to imitate has been documented already in the immediate postnatal period (Meltzoff and Decety, 2003; Meltzoff, 2007), and conflicting theories and notions have been devised to account for the way imitation is accomplished through perspective-taking.Wapner and Cirillo (1968) investigated imitative com-petence in children aged 8 to 18 years, focusing on the ‘transposition’ of left-right relations, a feature of spa-tial development. They showed a human model (the experimenter himself) enacting hand/ear movements with the right or left hand (e.g., grasping the right ear with the left hand) or actions directed to an external object (e.g., touching a colored object), and instructed 240 young subjects to “Do just like I do. Watch me. I want you to do exactly as I do”. The authors noted that 8 year olds were likely to use a mirror imitation strategy, and that increasing age involved its gradual replacement by the anatomical mode.Press et al. (2009) replicated this study in their Experiment 1, and showed that when participants were asked to “do what the model does” at six dif-

ferent viewing angles (0°, 60°, 120°, 180°, 240° and 300°) they performed more often anatomically matching acts. In Experiment 2 participants were divided into two groups and instructed to imitate using an anatomically and not anatomically matching mode, respectively; the results disclosed performance differences for some viewing angles, performances being more accurate when subjects used mirror imi-tation from viewing angles of 180° and 240°.Bekkering and co-workers (2000) investigated the way children imitate gestures modelled by an exper-imenter placed in front of the observer. They dem-onstrated that when children (aged 3 to 5 years) were asked to imitate as if standing in front of a mirror (“Try to imitate me as if you were my mirror. You do what I do”), their performances were generally cor-rect, and especially so when the action was directed to a target (ear) on the same side of the body (right hand to right ear), thus documenting unquestionable mirroring competences.Avikainen et al. (2003) tested the mirroring ability of adults, and found that subjects made significantly fewer errors when using the mirror mode (“imitate as if looking at a mirror”) than when they were asked to use the opposite hand (anatomical correspondence).Franz et al. (2007) investigated the mirror and non-mirror mode of imitation assuming that the natural tendency of humans is to mirror movements, and that anatomical performances would replace the mirror mode in presence of certain stimulus information (i.e., in Experiment 1, participants were wearing wedding bands on a finger of the left hand). When the mental alignment between the self and the stimulus does not require mental spatial transformation, performance is faster and in mirror mode; when rotation of one’s body representation is required for alignment with respect to the stimulus a spatial transformation is necessary, resulting in an anatomical performance. In their pilot tests, Franz and colleagues found that when the instructions asked to imitate the model in non-mirror mode reaction times were longer than when they asked to imitate in mirror mode.The results of these investigations, some document-ing a predominance of the anatomical mode and others a predominant use of the mirror mode of imitation, suggested to us that the original imitation mechanism is according to a mirror perspective; however, situations requiring a mental operation (i.e., involving a specific comparison between the

22 C. PIERPAOLI ET AL.

self and the stimulus) result in the adoption of the anatomical perspective.The aim of the present study was to explore the relationships between instructions and imitative per-spective-taking based on the assumption that adult subjects, confronted with a visuomotor stimulus, i.e. a model in 3rd person perspective performing intransi-tive gestures with both arms, i) will use a mirror mode strategy when acting spontaneously; and ii) will adopt the anatomical mode when instructed to use the same or the opposite limb compared with the model.To test the first assumption we asked adult par-ticipants to imitate a set of gestures by instructing them to “imitate what you see” and “imitate using the limb of your choice” (free imitation tasks). To test the second assumption participants were asked to “imitate using the same limb as the model” and “imitate using the limb opposite to the model’s” (driven tasks).The tendency to imitate a model in 3rd person perspec-tive using a mirror strategy is probably explained by the need to create a social space in which model and imitator share a similar spatial topography (Iacoboni et al., 2001). The decision to mirror acts executed by others is likely related to cortical patterns of activity, consistent with the existence of the mirror neuron system, and reflecting a close connection between the mirror strategy and the specific neuron system that matches observed and executed actions. Investigations of cortical activation during anatomi-cally and not anatomically matching gestures have highlighted different functional MRI patterns of frontoparietal activation for mirror and anatomical imitation (Koski et al., 2003), generating the hypoth-esis of a close link between mirror imitation and the mirror neuron system. In addition, the facts that the anatomical perspective is more often associated with executive errors (Ishikura and Inomata, 1995), and that mirror mode imitation involves shorter reaction times than the anatomical mode (Koski et al., 2003; Franz et al., 2007) suggest that the two mechanisms might be underpinned by distinct neural processes.Therefore, if in free tasks subjects seem to follow the spontaneous mechanism of imitation, the imita-tive strategy adopted in driven tasks can probably be explained with the ability of adult humans to put themselves in another’s perspective, a mental opera-tion that may allow a greater understanding of the other’s gestures.

In free trials subjects were likely to imitate gestures using their dominant limb. In general, people tend to use their preferred hand for more demanding tasks, and usually codify visuomotor stimuli (i.e., in laterality judgments) favoring the left or right side according to their handedness (left side for the left-handed, right side for the right-handed). It has been reported that right-handed subjects exposed to a decision-making task involving right-left judgments about presented stimuli (e.g., hands depicted in dif-ferent manners) recognized the right hands faster than the left (Gentilucci et al., 1998); similar find-ings were reported by Ionta and Blanke (2007), who described faster responses in spatial choice reaction tasks where visual stimuli lay in the visual field cor-responding to the dominant hand. The hypothesis has thus been advanced that individuals imagine actions differently according to their handedness (Casasanto, 2009), and that mental representations of actions and body-related movements may be dif-ferent in relation to the preferred body part used in everyday life acts by right- and left-handed subjects (Willems et al., 2010).In our free trials, one possibility was therefore that participants would use their preferred limb (i.e., the right limb, since all were right-handed according to the Oldfield test). However, this was immediately disproved by the finding that most responses were in the mirror mode, and very few involved the limb that they employed most often in everyday activi-ties. In driven trials, where the possibility to use the preferred limb was excluded by task instructions, the main finding was that most participants adopted the anatomical imitation strategy.

Methods

ParticipantsTwenty seven subjects (14 males, age range 21-53 years, mean 33, SD = 7.39) gave their informed con-sent to participate in the study, approved by the local Ethics Committee. All were right-handed according to the Edinburgh Inventory (Oldfield, 1971) and all had normal or corrected to normal visual acuity.

StimuliStimuli consisted of intransitive upper limb gestures enacted by a model in video clips that were shown


in sequential order. Both meaningful (n = 84) and meaningless (n = 84) gestures were included, to focus on imitative perspective-taking and avoid per-formance biases related to the stimulus. The mean-ingless gestures were modifications of the meaning-

ful gestures; they involved the same body parts but differed in the relationship between hand/arm/trunk and the target of the movement. Both sets of move-ments included body-related (Fig. 1A1-3 and B1-3) and body-unrelated (Fig. 1A4-6 and B4-6) gestures.

Fig. 1. - Gestures executed by the model in the video clips. A1-6: meaningful gestures; B1-6 meaningless gestures; A1-3 and B1-3 body-related gestures; A 4-6 and B 4-6 body-unrelated gestures.


Each video clip depicted a model (in 3rd person perspective; 180° viewing angle) who started from a rest position and performed one gesture. Clips were filmed with a digital camera (Sony Cyber Shot, DSC, W210 Steady Shot, 12.1 Mega Pixels, Optical zoom 4x) keeping the size, frame (film shot) and light of the stimulus (gesturing model) constant across clips. When each clip began the model was standing still with the arms by the sides, facing the imitator. There were 6 gestures: (1) military salute (open hand held diagonally close to the forehead), (2) silence (closed hand with index finger upright, close to the lips), (3) crazy (index finger sticking out of a closed hand repeatedly hitting the forehead), (4) bye-bye (open hand, waving 45° to one side and the other, repeatedly), (5) more or less (hand open in front of the body; moving repeatedly 45° to the left and to the right), and (6) stop (open hand, away from and in front of the body; see Carmo and Rumiati, 2009, Appendix A). Each gesture was per-formed with either limb in separate clips. The clips were edited using E-PRIME software (Psychology Software Tools Inc., Pittsburgh, PA), and projected to the center of the visual field on a PC screen (TFT LCD 15,4” WXGA Acer CrystalBrite, resolution 1280x800). The distance between the screen and the subject’s eyes was adjusted to 57 cm (1 cm on the screen represented 1° of visual field).

Design and procedureThe experimental protocol lasted 40 minutes. Before beginning, each participant was instructed about the aim and features of the study. They were asked to sit in front of a computer monitor; the experimenter, sit-ting on their left, explained that the study consisted of two sets of trials, one showing meaningful and the other meaningless gestures, and that at the beginning of each trial they would be instructed about the way to perform the required task. As expected, all partici-

pants demonstrated an understanding of the content of meaningful gestures and failed to understand the meaningless gestures. All subjects were shown an identical list of 168 video clips.A fixation point lasting 2 s was followed by a 250 ms alert sound and then by the clip showing the model against a white background; this was followed by a variable period of blank screen allowing for gesture imitation by the subject. The experimenter merely annotated the limb (right or left) used by the par-ticipant in response to each right and left gesture executed by the model. As soon as the gesture had been executed the experimenter prompted the next video clip.Each set consisted of 7 trials, named with the prefix “S” followed by numbers from 0 to 6. In both mean-ingful and meaningless sets, trials were presented in sequential order from S0 to S6 starting with S0, continuing with S1, S2, S3, S4, S5, and finishing with S6 (Table I).S0 (“imitate what you see”; in Italian: “imita ciò che vedi”) asked subjects to imitate what they saw in the video clips; S1 and S6 (“imitate using the limb of your choice”; in Italian: “imita utilizzando un arto a tuo piacimento”) asked them to make the gesture using any limb they wished; in S3 (“imitate using the right limb”; in Italian: “imita usando l’arto destro”) and S5 (“imitate using the left limb”; in Italian: “imita usando l’arto sinistro”) they were asked to imitate the gestures using their right and left limb, respectively. In S2 (“imitate using the same limb as the model”; in Italian: “imita usando lo stesso arto usato dalla modella”) and S4 (“imitate using the limb opposite to the model’s”; in Italian: “imita usando l’arto opposto a quello usato dalla modella”) participants were asked to use the same and the opposite limb as the model, respectively.Trials and gestures were presented to all subjects in the same order, since order was not expected to

Table I. - Description of the contents of trials and instructions.

Blocks of trials Type of instruction

Free trials S0 Free instructions “Imitate what you see”

S1 and S6 “Imitate using the limb of your choice”

Driven trials S2 (same) Driven instructions “Imitate using the same limb as the model”

S4 (opposite) “Imitate using the limb opposite to the model’s”

Driven control trials S3 (right limb) Driven control instructions “Imitate using the right limb”

S5 (left limb) “Imitate using the left limb”


influence performances and to result in a carryover effect. The sequence of the video clips is reported in Table II. Reaction times were not measured nor was movement accuracy evaluated.S0 and S1 were free trials whose main difference regarded participant’ interpretation of the degree of freedom they allowed. The wording “imitate what you see” and “imitate using the limb of your choice” aimed at leaving the amplest freedom. Administration of two trials with a very similar instruction content in close succession was directed at helping participants become familiar with the study aims and content and at obtaining an imita-tion as natural as possible (reinforcement through repetition).– S2 and S4 were driven tasks.– S6 was another free task. The reason for adding a

further free trial was to verify whether the driven trials S2 and S4 would influence the spontaneous way participants imitate.

– S3 and S5 were driven tasks testing right/left dis-crimination competence and were used as control trials.

No further explanations were provided after the instructions given before each trial.

Statistical methodsThe response variable of interest, Y, was binary and concerned whether the subject’s gesture was performed in the mirror (Y = 1) or the anatomical (Y = 0) mode.Some covariates were analyzed. A dichotomized scoring method was used to operationalize them. The factors considered as possible explanatory vari-ables in the statistical analysis and their levels were therefore:– Gmean: the model’s gesture was meaningful

(Gmean = 1) or meaningless (Gmean = 0);– Gbody: the gesture was body-related (Gbody =

1) or body-unrelated (Gbody = 0);– Limb: the model moved the right upper limb

(Limb = 1) or the left upper limb (Limb = 0);

– Imit: the trial was characterized by driven (Imit = 1) or free instruction (Imit = 0).

A logistic linear model was used to group binary data with a random effect associated with the sub-jects participating in the study, who were a random sample from a normal population.To summarize the extent to which the sample data are fitted by a given (current) statistical model we used deviance, which can be regarded as a measure of the lack of fit between model and data. In general, the larger the deviance the poorer the fit. Statistical analysis of data was carried out with R statistical program (R Core Team, 2013; see also Collett, 2003) and the Microsoft Excel® program.

Results

The dataset consisted of 4536 potential observa-tions: 2 sets of trials; 7 trials to a set; 12 clips/trial; 27 participants). Driven trials S3 and S5 were used as control trials and were therefore excluded from the statistical analysis. Their subtraction (1296 observations) left 3240 observations. Null responses (n = 449) were also excluded from the analysis: they consisted of failure to make the required gesture (e.g., because of loss of concentration) or of perfor-mance correction by participants. The final dataset comprised 2791 observations.Statistical analysis of the final dataset was performed by calculating the statistical frequency of each factor (Gmean, Gbody, Limb, Imit and Y; Table III).As regards the covariate Gmean, out of a total of 2791 valid observations, 1198 related to meaning-less gestures (Gmean = 0; 43%), and 1593 to mean-ingful gestures (Gmean = 1; 57%). For the covariate Gbody there were 2791 valid observations, of which 1394 regarded body-related gestures (Gbody = 0; 49.9%) and 1397 body-unrelated gestures (Gbody = 1; 50.1%). With regard to the covariate Limb, there were 2791 valid observations, 1394 relating to ges-tures executed by the model with her left upper limb

Table II. - Fixed sequence of the gestures per trial.

1 2 3 4 5 6 7 8 9 10 11 12

Gesture Military salute

Stop Military salute

Bye bye

Crazy More or less

Silence Crazy More or less

Bye bye

Silence Stop

Side of the model’s limb Left Right Right Left Right Left Right Left Right Right Left Left


(Limb = 0; 59.9%), and 1397 relating to right-arm gestures (Limb = 1; 50,1%). As regards the covari-ate Imit, out of 2791 valid observations 1665 were relative to free tasks (Imit = 0; 60%) and 1126 to driven tasks (Imit = 1; 40%). Finally, as regards the variable of interest Y, out of 2791 valid observations (S0 + S1 + S2 + S4 + S6), 1698 involved anatomical imitation (Y = 0; 61%) and 1093 involved mirror mode imitation (Y = 1; 39%).Since the aim of the study was to investigate the link between imitative perspective-taking (Y) and trial instructions (Imit), assuming that mirror and anatomical imitation depend on the way subjects codify the experimenter’s instruction, the frequency of Imit×Y was compared before applying the binary logistic regression. The interesting finding was that even though the number of anatomical performances was higher than that of mirror performances (Y = 0 = 61%; Y = 1 = 39%), this seemed to depend on the

Imit covariate. In fact, the results of Imit×Y (Fig. 2) showed that of the 1665 free instructions (Imit = 0), 651 resulted in anatomical imitation (Y = 0; 39%) and 1014 in mirror imitation (Y = 1; 61%). Of the 1126 driven instructions (Imit = 1), 1047 were executed in the anatomical mode (Y = 0; 93%), and 79 in the mirror mode (Y = 1; 7%). Of the 1698 ana-tomical performances (Y = 0), 651 were responses to free instructions (Imit = 0; 38%) and 1047 to driven tasks (Imit = 1; 62%). Finally, 1014 of the 1093 mirror performances (Y = 1) were responses to free instructions (Imit = 0; 93%) and 79 to driven tasks (Imit = 1; 7%).After frequency calculation binary logistic regres-sion was applied. The deviance and the correspond-ing number of degrees of freedom for some of the possible logistic regression models fitted to the data were evaluated. In defining the probability of mirror mode gestures S = P(Y = 1), 7 equations were calcu-

Table III. - Frequencies of the covariates and the variable of interest.

Value = 0 Value = 1 Tot.

Gmean 1198 1593 2791

Gbody 1394 1397 2791

Limb 1394 1397 2791

Imit 1665 1126 2791

Y 1698 1093 2791

Fig. 2. - Bar histogram showing participant performances in free and driven imitation trials, expressed as number of observations. In free trials (bars on the left) subjects showed a preference for mirror mode imitation (light bars; 1014/1665 total observations, 61%). In driven trials (bars on the right), participants showed a marked preference for the anatomical strategy (dark bars; 1047/1126 total observations; 93%).


lated to describe the statistical relationship between all covariates and the response variable.Imit (deviance = 2481) and Limb (deviance = 3482) were the variables with the highest predictive power, since they involved the lowest deviance compared with fitting a single constant term (deviance = 3538). Gbody (constant + Gbody; deviance = 3538) and Gmean (constant + Gmean; deviance = 3544) were the single terms influencing Y values least.It should be noted that the deviance for the single-variable model is considerably larger than that for models involving more than one term (i.e., consider-ing Imit, Limb and Gmean and interactions between Imit and the other two variables; deviance = 2336); however, the aim of the study was to find the signifi-cant factors associated with the probability of Y = 1 and Y = 0, not to evaluate the contribution of all fac-tors to Y. Since addition of Gbody to the model did not improve the deviance, Gbody was not included.Finally, writing K = b0 + b1 × Imit + b2 × Limb + b3 × Gmean + b4 × Imit × Limb + b5 × Imit × Gmean for the linear component of the model derived from the fixed effects, the logistic model became

where = P(Y = 1) is the probability that the subject’s gesture is executed in a mirror perspective, V is the standard deviation of the random effect, and Z is the standard normal random variable.In determining the variables that are significantly related to the response probability regardless of the lack of fit, we then chose model number 7 (Table IV) because it involved the greatest predictive power (deviance = 2378) of the models that included only significant variables.No practice or fatigue effects affected performances, since performance quality (correctness) was not measured; in addition each trial had a short dura-tion and a brief rest was envisaged between S0 to S6. There was no order effect, because subjects did not know what the next task would involve until the experimenter instructed them at the beginning of each trial; moreover the experimental design was of the observational type (i.e., a questionnaire-like study), where the order of the questions has no power to shape responses nor to create a bias.S3 and S5 demonstrated that all participants dis-

tinguished between right and left limb. These were control trials and were excluded from the statistical analysis.

Discussion

The present study was designed to gain insights into the imitative modes adopted by healthy adult subjects left free to imitate or required to use the same or the opposite limb with respect to a model. The results indicate that the mirror mode was used more often in performances involving free imitation (61%) and that the anatomical mode was the pre-dominant strategy (93%) when instructions required using the same or the opposite limb, suggesting that instruction content strongly affected the mode of imitation used by subjects. These findings partially agree with previous studies, albeit obtained in dif-ferent conditions.To learn whether gestures were imitated in the mir-ror or the anatomical mode (Y variable), we ana-lyzed factors related to the stimulus (Limb, Gmean and Gbody) and a measure related to the task (Imit). The hypothesis was that since Imit was the sole covariate distinguishing free from driven trials it could be the one exerting the strongest influence on Y (imitative perspective-taking; 0 anatomical, 1 mirror). Data analysis confirmed the hypothesis by showing that the imitation mode was based on the instructions describing how tasks were to be executed (Imit). As shown in Table IV, Imit had the strongest associative link with Y (Std. Error = 0.294; p < 0.001). To assess which covariate, if any, influenced Y, a within-subjects repeated measures design was adopted: Imit (free instructions vs. driv-en instructions) × Limb (model’s right upper limb vs. model’s left upper limb) × Gmean (meaningful vs. meaningless gesture) × Gbody (body-related vs. body-unrelated).The instructions (Imit) used in the present work are not wholly comparable to those used in previous studies of imitation. Earlier investigations frequently employed instructions such as “do what I do” or “do what the experimenter does” (see Press et al., 2009). Those used in our free trials were slightly different, since participants where asked to “imitate what you see” and “imitate using the limb of your choice”. These instructions aimed at disconnecting


the subject’s decision about the limb to be used from the stimulus; the task to be performed was the sole instruction content. In other surveys driven trials involved explicit instructions by the demonstrator (the experimenter or a model in videos) to execute anatomical performances (see Press et al., 2009; Sudo et al., 2012), or “to imitate movements of the left side of the model’s body with the movements of the left side of their own bodies, and vice versa for movements of the right side of the model’s body” (see Heyes and Ray, 2004). In our study only the concept of same and opposite were specified to guide performances, without further explanation of their meaning. Use of these terms was directed at differentiating the putative spontaneous behav-ior characterizing imitation in free trials from the strategy adopted to follow instructions that required making a sort of comparison between the self and the stimulus. In driven tasks S2 and S4 the instruc-tions directed the subject’s attention to the model’s limb by asking them to relate performances to some characteristic of the model using the same or the opposite body parts, thus prompting subjects to process the concepts of same and opposite. The situ-ation could involve one of two types of reasoning based on spatial or anatomical frames of reference. When subjects chose the anatomical reference frame they first needed to decide whether the limb used by the model was the right or the left, then they had to rotate their own body image to fit the anatomical ref-erence of the model’s; finally they had to select the same (or opposite) anatomical effector with respect the model’s.In defining a performance as mirror or anatomical a link was assumed between the criteria of spatial stimulus response compatibility theory and instruc-tion content (see Table V). In general, we interpreted as mirror the responses involving spatial match and

as anatomical those lacking this feature. According to the stimulus response mapping process, a com-patible condition includes spatial correspondence between the observer’s right side and the right side of the stimulus (and vice versa). A response con-forming to spatial stimulus response compatibility was defined as mirror. Lack of spatial match led the response to be considered as anatomical.Therefore in driven trial S2 responses where the concept of same was interpreted as “in the same spa-tial line” were defined as mirror and those that were inconsistent with this interpretation were considered as anatomical. In driven trial S4 pairs “model’s right limb vs. participant’s right limb” and “model’s left limb vs. participant’s left limb” were defined as mir-ror, and pairs “model’s right limb vs. participant’s left limb” and “model’s left limb vs. participant’s right limb” as anatomical (see Table V). In the lat-ter case the opposite limb was the one that in spatial terms lay in the diagonal trajectory between the model’s and the imitator’s limb.In driven tasks the effect based on the correspon-dence between the spatial location of the stimulus and the effector, which in the responses given in free trials was manifest, was replaced by a different criterion, i.e., the anatomical identity between imita-tor’s and model’s body parts, a sort of ability to put oneself in another’s perspective; a mental operation that may allow a greater understanding of the other’s gestures. The reason why spatial compatibility was not produced in driven trials may be that driven instructions depend on the so-called own body trans-formation (May and Wendt, 2012), a mechanism of visuospatial transformation through which subjects mentally assume the position of the stimulus to reproduce the performances observed. This phe-nomenon is clearly evident when an observer makes left-right judgments.

Table IV. - Parameter estimates and standard errors of the estimates for model number 7.

Variable Estimate Std. Error z value P

Fixed effects

Intercept, b0 1.275 0.249 5.12 < 0.001

Imit, b1 -2.948 0.294 -10.03 < 0.001

Limb, b2 -1.154 0.116 -9.94 < 0.001

Imit×Limb, b3 1.123 0.274 4.1 < 0.001

Random effects

Std. Dev., V 1.204


Subjects watching the model and asked to use the same or the opposite limb (S2 and S4) were required to assess two parameters: 1. which limb was used by the model; and 2. which of their own limbs corre-sponded with the model’s same or opposite, in terms of the effector to be recruited. These questions relate to handedness and laterality judgments.In the present study participants adopted a right-left judgment to decide whether the limb used by the model was the right or the left even though no explicit laterality judgment was required.After establishing whether the body part used by the model was the right or the left, the next mechanism consisted of putting oneself in the other person’s shoes (Jeannerod, 2007; Parsons, 1994, 1995) to achieve a body parts match. It may be hypothesized that such a mechanism is mental rotation (see Shepard and Metzler, 1971).

Final conclusions

Considering that: i) the shift from mirror to ana-tomical imitation occurs at a certain stage in human development (Wapner and Cirillo, 1968); ii) the shift from mirror to anatomical mode is temporally in line with the development of the highest cogni-tive functions, such as mental spatial representation (Wapner and Cirillo, 1968); iii) the shift from mirror to anatomical imitation allows subjects to create and manipulate information relative to the body image, comparing the self with the environment, repre-sented by the stimulus to be imitated (Wapner and Cirillo, 1968); iv) there exist two ways to execute imitative performances: a primary and more spon-taneous mode involving a spatial match between the self and the stimulus (mirror mode) and another

involving more complex cognitive processes (from spatial stimulus response compatibility effect to mental rotation), it may be hypothesized that the shift occurs when the subject is required to process the concept of same and opposite (body parts) in relation to some characteristics of the stimulus, i.e., when the subject is required to perform a cognitive task. According to Wapner and Cirillo the shift is related to the stage of human development; the pres-ent findings indicate that in adult subjects the adop-tion of the mirror or the anatomical strategy depends on the type of task.Sudo et al. (2012) investigated egocentric mental transformations of the self in mirror and anatomical tasks. They tested subjects imitating hand actions presented at different spatial orientations using mir-ror and anatomical perspectives. They evaluated the correctness and the response latency of the strategy used, and discovered that faster performances were in mirror mode in the facing situation (180°), and in anatomical mode in the side by side situation (0°).In a previous study by Jackson et al. (2006), visuo-motor stimuli executed with hands and feet were presented in 1st and 3rd person perspective to partici-pants, who were asked to observe and imitate them; the time taken to imitate movements was shorter in 1st person perspective. Moreover functional imag-ing data demonstrated strong sensorimotor region activation in 1st person perspective. These results demonstrated that the compatibility between the movement to be executed and the model signifi-cantly influences motor performance.Since mirror imitation is faster than the anatomical mode and does not require spatial transformation, the hypothesis may be advanced that sheer cognitive economy leads to spontaneous imitation via the mir-ror mode, whereas the anatomical imitation strategy

Table V. - Description of the relationship between the model and the imitator, in defining mirror and anatomical performances.

Model vs. ImitatorFree instructions Driven instructions

S0, S1 and S6 S2 S4

Model’s right limb vs. Imitator’s right limb

Non spatial compatibility (anatomical mode)

Non spatial compatibility(anatomical mode)

Spatial compatibility(mirror mode)

Model’s left limb vs. Imitator’s left limb


Non spatial compatibility(anatomical mode)


Model’s right limb vs. Imitator’s left limb




Model’s left limb vs. Imitator’s right limb





is adopted when performance involves a link to spe-cific characteristics of the stimulus.In conclusion, the present findings highlight a sig-nificant difference in the way healthy adults imitate observed movements; the mode of imitation adopted depends on task and situation. The two modes prob-ably obey distinct mental processes that may be underpinned by specific neural circuits or cortical areas whose maturation depends on the stage of brain development.

AcknowledgementsThis work was supported by Ministero Istruzione, Università e Ricerca, PRIN 2009. We are grateful to all the volunteers who participated in the study and to Word Designs (www.silviamodena.com) for language editing.

References

Anquetil T. and Jeannerod M. Simulated actions in the first and in the third person perspectives share common representations. Brain Res., 1130: 125-129, 2007.

Avikainen S., Wohlschläger A., Liuhanen S., Hänninen R., Hari R. Impaired mirror-image imi-tation in Asperger and high-functioning autistic subjects. Curr. Biol., 13: 339-341, 2003.

Bekkering H., Wohlschläger A., Gattis M. Imitation of gestures in children is goal-directed. Q. J. Exp. Psychol. A., 53: 153-164, 2000.

Carmo J.C. and Rumiati R.I. Imitation of transitive and intransitive actions in healthy individuals. Brain Cogn., 69: 460-464, 2009.

Casasanto D. Embodiment of abstract concepts: good and bad in right- and left-handers. J. Exp. Psychol. Gen., 138: 351-367, 2009.

Collett D. Modelling Binary Data, Capman Hall, London, CRC, 2003.

Crescentini C., Mengotti P., Grecucci A., Rumiati R.I. The effect of observed biological and non bio-logical movements on action imitation: an fMRI study. Brain Res., 1420: 80-92, 2011.

Development Core Team A language and environ-ment for statistical computing. Foundation for Statistical Computing, Vienna, Austria, 2008.

Franz E. A., Ford S., Werner S. Brain and cogni-tive processes of imitation in bimanual situations: Making inferences about mirror neuron systems. Brain Res., 1145: 138-149, 2007.

Gentilucci M., Daprati E., Gangitano M. Implicit visu-al analysis in handedness recognition. Conscious Cogn., 7: 478-493, 1998.

Gleissner B., Meltzoff A.N., Bekkering H. Children’s coding of human action: cognitive factors influ-encing imitation in 3-year-olds. Dev. Sci., 3: 405-414, 2000.

Heyes C. and Ray E. Spatial S-R compatibility effects in an intentional imitation task. Psychon. Bull. Rev., 11: 703-708, 2004.

Iacoboni M., Koski L.M., Brass M., Bekkering H., Woods R.P., Dubeau M.C., Mazziotta J.C., Rizzolatti G. Reafferent copies of imitated actions in the right superior temporal cortex. Proc. Natl. Acad. Sci. USA., 98: 13995-13999, 2001.

Ionta S. and Blanke O. Differential influence of hands posture on mental rotation of hands and feet in left and right handers. Exp. Brain Res., 195: 207-217, 2009.

Ishikura T. and Inomata K. Effects of angle of model-demonstration on learning of motor skill. Percept. Mot. Skills, 80: 651-658, 1995.

Jackson P.L., Meltzoff A.N., Decety J. Neural cir-cuits involved in imitation and perspective-taking. Neuroimage, 31: 429-439, 2006.

Jeannerod M. Being oneself. J. Physiol. Paris, 101: 161-168, 2007.

Kitada R., Dijkerman H.C., Soo G., Lederman S.J. Representing human hands haptically or visually from first-person versus third-person perspectives. Perception, 39: 236-254, 2010.

Koski L., Iacoboni M., Dubeau M.C., Woods R.P., Mazziotta J.C. Modulation of cortical activity dur-ing different imitative behaviors. J Neurophysiol, 89: 460-471, 2003.

Krause D. and Kobow S. Effects of model orientation on the visuomotor imitation of arm movements: The role of mental rotation. Hum. Mov. Sci., 32: 314-327, 2013.

Lausberg H. and Cruz R.F. Hemispheric specialisa-tion for imitation of hand-head positions and fin-ger configurations: a controlled study in patients with complete callosotomy. Neuropsychologia, 42: 320-334, 2004.

May M. and Wendt M. Separating mental transfor-mations and spatial compatibility effects in the own body transformation task. Cogn. Process., 13: 257-260, 2012.

Meltzoff A.N. ‘Like me’: a foundation for social cog-nition. Dev. Sci., 10: 126-134, 2007.


Meltzoff A.N. and Decety J. What imitation tells us about social cognition: a rapprochement between developmental psychology and cognitive neurosci-ence. Phil. Trans. R. Soc. Lond. B., 358: 491-500, 2003.

Mühlau M., Hermsdörfer J., Goldenberg G., Wohlschläger A.M., Castrop F., Stahl R., Röttinger M., Erhard P., Haslinger B., Ceballos-Baumann A.O., Conrad B., Boecker H., Left inferior parietal dominance in gesture imitation: an fMRI study. Neuropsychologia, 43: 1086-1098, 2005.

Oldfield R.C. The assessment and analysis of handed-ness: the Edinburgh Inventory. Neuropsychologia, 9: 97-113, 1971.

Parsons L.M. Temporal and kinematic properties of motor behavior reflected in mentally simulated action. J. Exp. Psychol. Hum. Percept. Perform., 20: 709-730, 1994.

Parsons L.M., Fox PT., Downs J.H., Glass T., Hirsch T.B., Martin C.C., Jerabek P.A., Lancaster J.L. Use of implicit motor imagery for visual shape discrimination as revealed by PET. Nature, 375: 54-58, 1995.

Press C., Ray E., Heyes C. Imitation of lateralized body movements: doing it the hard way. Laterality, 14: 515-527, 2009.

Rumiati R.I. and Tessari A. Imitation of novel and well-known actions: the role of short-term memo-ry. Exp. Brain Res., 142: 425-433, 2002.

Sambrook T.D. Does visual perspective matter in imitation? Perception, 27: 1461-1473, 1998.

Shepard R.N. and Metzler J. Mental rotation of three-dimensional objects. Science, 171: 701-703, 1971.

Shofield W.N. Do children find movements which cross the body midline difficult? Q. J. Exp. Psychol., 28: 571-582, 1976.

Stürmer B., Aschersleben G., Prinz W. Correspondence effects with manual gestures and postures: a study of imitation. J. Exp. Psychol. Hum. Percept. Perform., 26: 1746-1759, 2000.

Sudo T., Heari T., Mogi K. Egocentric mental trans-formation of self: effects of spatial relationship in mirror-image and anatomic imitations. Exp. Brain Res., 221: 27-32, 2012.

Wapner S. and Cirillo L. Imitation of a model’s hand movements: age changes in transposition of left-right relations. Child Dev., 39: 887-894, 1968.

Willems R.M., Toni I., Hagoort P., Casasanto D. Body specific motor imagery of hand actions: neu-ral evidence from right- and left-handers. Psychol. Sci., 21: 67-74, 2010.

Wohlschläger A. and Bekkering H. Is human imi-tation based on a mirror-neuron system? Some behavioural evidence. Exp. Brain Res., 143: 335-341, 2002.

Introduction

In healthy individuals, sleep is beneficial and restor-ative. On the contrary, in patients with respiratory or heart disease, sleep can precipitate myocardial ischemia, arrhythmias and even death (Verrier and Mittleman, 2005; Verrier and Harper, 2010). Heart diseases cause 30% of the deaths in the world (WHO, 2009). Results from a wealth of studies indicate that cardiac events occur predominantly in

the early morning hours (Muller et al., 1985; Muller et al., 1987a; Muller et al., 1987b; Pepine, 1991). This may be linked to the prevalent occurrence of rapid eye movement (REM) sleep at the end of the sleep period, prior to the transition to wakefulness. Because REM sleep is associated with large fluctua-tions in blood pressure, heart rate (HR) and sympa-thetic activity (Somers et al., 1993; Vanoli et al., 1995; Vaughn et al., 1995), autonomic surges aris-ing from the last REM sleep episode could stimulate

Statistical, spectral and non-linear analysisof the heart rate variability during

wakefulness and sleepV. BRANDO1, S. CASTRO-ZABALLA2, A. FALCONI2,

P. TORTEROLO2, E.R. MIGLIARO1

1 Laboratorio de Fisiología Cardiovascular; 2 Laboratorio de Neurobiología del Sueño, Departamento de Fisiología, Facultad de Medicina, Universidad de la República, Montevideo, Uruguay

A B S T R A C T

As a first step in a program designed to study the central control of the heart rate variability (HRV) during sleep, we conducted polysomnographic and electrocardiogram recordings on chronically-prepared cats during semi-restricted conditions.We found that the tachogram, i.e. the pattern of heart beat intervals (RR intervals) was deeply modified on passing from alert wakefulness through quiet wakefulness (QW) to sleep. While the tachogram showed a rhythmical pattern coupled with respiratory activity during non-REM sleep (NREM), it turned chaotic during REM sleep.Statistical analyses of the RR intervals showed that the mean duration increased during sleep. HRV measured by the standard deviation of normal RR intervals (SDNN) and by the square root of the mean squared difference of successive intervals (rMSSD) were larger during REM and NREM sleep than during QW. SD-1 (a marker of short-term variability) and SD-2 (a marker of long-term variability) measured by means of Poincaré plots increased during both REM and NREM sleep compared to QW. Furthermore, in the spectral analysis of RR intervals, the band of high frequency (HF) was larger in NREM and REM sleep in comparison to QW, whereas the band of low frequency (LF) was larger only during REM sleep in comparison to QW. The LF/HF ratio was larger during QW compared either with REM or NREM sleep. Finally, sample entropy analysis used as a measure of complexity, was higher during NREM in comparison to REM sleep. In conclusion, HRV parameters, including complexity, are deeply modified across behavioral states.

Key wordsPoincaré plot • REM • Spectral analysis • Autonomic system

Corresponding Author: Dr. Pablo Torterolo, Facultad de Medicina, Universidad de la República, General Flores 2125, 11800 Montevideo-Uruguay - Tel.: +598 2924 34 14 ext. 3234 - Email: [email protected]


HRV DURING SLEEP 33

platelet aggregation, disrupt plaques, or precipitate coronary spasm or arrhythmia, which may become manifest only upon arousal and could be inappropri-ately attributed to events during wakefulness rather than to sleep (Verrier et al., 1996). Thus, in spite of its quiescence, sleep is not a protected state.As early as the 2nd century, Galen of Pergamon observed that sleep is accompanied by a decrease in HR (Furley and Wilkie, 1984). In 1923 it was reported that HR increases during dreaming, and an increase in HR was observed by Aserinsky and Kleitman in their pioneer paper about REM sleep (when most dreams occur) (Macwilliam, 1923; Aserinsky and Kleitman, 1953). Nowadays, there is consensus that, by modulating autonomic as well as endocrine function, both non-REM (NREM) and REM sleep impact upon the activity of the heart. In this regard, an accepted way to assess the autonomic influence on the cardiac rhythm is to study the HR variability (HRV). HRV values are derived from the analysis of the oscillations of heart beat intervals (RR) dura-tion recorded by means of the electrocardiogram (ECG). Statistical methods, spectral, and non-linear analysis have been widely applied to determine HRV (Akselrod et al., 1981; Task Force of the ESC and the NASPE, 1996; Brennan et al., 2001; Contreras et al., 2007). In fact, it has been demonstrated in humans that the power spectral density (PSD) of the high frequency band of the spectrum (HF > 0.15 Hz) reflects parasympathetic activity. On the other hand, the low frequency band (LF; 0.04-0.15 Hz) is considered either as a marker of sympathetic activity or a parameter dependent on both sympathetic and vagal influences. The LF/HF ratio is also widely used as a relative marker of sympathetic nervous activity or sympatho-vagal balance (Akselrod et al., 1981; Pomeranz et al., 1985; Malliani et al., 1994). Factors such as age (Migliaro et al., 2001), or diseases as diabetes, sepsis or narcolepsy are associated with changes in HRV (Pontet et al., 2003; Contreras et al., 2007; Fronczek et al., 2008).The central nervous system control of the HRV is not fully understood, principally during sleep. As a first step in a program designed to investigate the central control of HRV during sleep utilizing the cat as an animal model, in the present study we analyze HRV indexes (statistical, spectral as well as non-linear methods) during undisturbed sleep and wakefulness cycles.

Methods

Four adult cats were used in this study. The animals were determined to be in good health by veterinarians of the Department of Laboratory Medicine of our Institution. All experimental procedures were con-ducted in accordance with the National Animal Care law (#18611) and with the “Guide to the care and use of laboratory animals” (8th edition, National Academy Press, Washington D.C., 2010). Furthermore, the Institutional Animal Care Committee approved the experimental procedures (Protocol #071140-001227-09). Adequate measures were taken to minimize pain, discomfort or stress of the animals, and efforts were made in order to use the minimal number of animals necessary to produce reliable data.Surgical procedures were the same as those employed in our previous studies (Torterolo et al., 2009; Castro et al., 2013). Briefly, each animal was pre-medicated with Xylazine®, atropine and antibi-otics. Anesthesia, was induced with ketamine, and maintained with a gas mixture of isoflourane in oxy-gen. The head was positioned in a stereotaxic frame and stainless steel screw electrodes were placed in the frontal and parietal bones to record the electroen-cephalogram (EEG), as well as in the orbital portion of the frontal bone to record the electro-oculogram (EOG). Bipolar electrodes were implanted in both lateral geniculate nuclei in order to monitor ponto-geniculo-occipital (PGO) waves. A Winchester plug (connected to the electrodes) and a chronic head-restraining device were bonded to the skull with acrylic cement. After the animals recovered from surgery, they were adapted to the recording envi-ronment for a minimum of two months. This long adaptation period is important to avoid cardiovas-cular long-term modifications that are elicited after surgery in the cat (Sei et al., 1989).Experimental sessions were conducted from 11 AM to 3 PM, in a temperature-controlled environment (21-23°C). During these sessions, the head of the cat was held in stereotaxic position by a head-restraining device. The EEG, EOG, PGO, neck electromyogram (EMG; electrodes were placed acutely on the skin over neck muscles), ECG (electrodes were placed acutely on the skin over the pre-cordial region), and respiratory activity by means of a micro-effort piezo crystal infant sensor were recorded. These signals were stored in a digital format by a CED A/D con-

34 V. BRANDO ET AL.

verter® and the Spike-2® software. The animals did not exhibit any signs of stress or discomfort dur-ing these sessions, as evidenced by their quiescent behavior and the fact that they fell asleep a few minutes after the beginning of the recording session.

Sleep and HRV analysesThe states of quiet wakefulness (QW), NREM and REM sleep were determined on the basis of poly-somnographic records that were divided into 10-sec-ond epochs and analyzed according to standard criteria (Ursin and Sterman, 1981).Occasionally, as in our previous studies, alert wake-fulness (AW) was elicited by means of auditory stimulation (Castro et al., 2013).RR intervals were measured from the ECG recordings using Spike-2® software. The software detects each R wave and calculates the duration of RR intervals. The RR intervals detected were visually inspected for identifying and correcting false positives or artificial long intervals (lack of detection) and arrhythmias that could generate abnormal values of HRV.For qualitative analysis the corrected RR intervals were plotted against time (tachogram). The tacho-gram was converted and analyzed as a waveform with a sample rate of 16.4 Hz. Spectrogram, auto-correlation and cross-correlation (against respiratory activity) functions were performed by means of the Spike-2® software.For selected 500-s periods of wakefulness, NREM and REM sleep, the magnitude squared coherence was analyzed (Castro et al., 2013). Coherence analysis is a measure of similitude in activity in a given frequency. Magnitude squared coherence was measured using the Spike-2® script COHER 1S. By means of this algorithm we analyzed the coherence between two waveforms channels (tachogram and respiratory activity) that were recorded simultane-ously. This period of analysis was divided into 32 time-blocks for a sample rate of 16.4 Hz, with a bin size of 512 samples (and a resolution of 0.016 Hz. To normalize the data we applied the Fisher z’ transform to coherence values (Castro et al., 2013).Quantitative analysis of the HRV was performed as follow. The data matrix for HRV analysis consisted of 48 lists of 750 validated RR intervals (expressed in ms). Sixteen lists of RR intervals were obtained during stable and artifact-free QW, NREM and REM sleep episodes. Each animal contributed with

4 lists of intervals per behavioral state (12 in total).The data were analyzed using the Kubios software package for HRV analysis (Kubios HRV 2.1 soft-ware University of Kuopio, Finland, www.kubios.uef.fi); similar processing of the data has been per-formed in previous reports of our group (Migliaro and Contreras, 2001; Migliaro et al., 2001; Pontet et al., 2003; Contreras et al., 2007). The following indexes were analyzed:a) Statistical. Mean RR interval, standard devia-

tion of normal RR intervals (SDNN, called NN after removal of abnormal or non-sinusal inter-vals), and the square root of the mean squared difference of successive intervals (rMSSD). SDNN provides an overall examination of HRV, while rMSSD yields a more accurate analysis about short-term modifications of HRV (Von Neumann, 1941; Malik, 1995).

b) Spectral analysis. It was performed by means of an autoregressive model of order 16. The power of the LF and HF bands, as well as the LF/HF ratio was estimated. We considered that the time of analysis was not enough to evaluate the very low frequency band (VLF), hence it was not con-sidered (Task Force of the ESC and the NASPE, 1996). The normalized spectral HRV indices LFnu and HFnu were not calculated because reflect the same aspect of the autonomic balance as the LF/HF ratio (Burr, 2007). Considering that LF and HF bands definitions are usually set for human’s recordings, we performed a preliminary analysis of the HRV in the frequency domain in the cat in frequencies up to 1 Hz. We carefully chose the definition of HF limits according to the respiratory rate (12-30 per minute in the cat), and then we set the HF limits in 0.15-1 Hz. The val-ues for the LF band were set at 0.04-0.15 Hz; we would like to stress that slow oscillations within these frequencies are readily observed after bilat-eral vagotomy in the cat (Baust and Bohnert, 1969). A previous report in the cat utilized simi-lar LF and HF bands definitions (Abbott, 2005).

c) Poincaré plots. This method is a geometric procedure that allows a beat-to-beat analysis of the HRV. It is based on a scatter-plot of each RR interval as a function of the preceding one and yields information on instantaneous HRV in a graphic representation (Woo et al., 1992). Each plot displays the beat-to-beat variability

HRV DURING SLEEP 35

(variation in RRn+1 at a given value of RRn) that is reflected in the scatter of values on the y-axis for a given value of x, as well as the overall varia-tion, which is reflected in the absolute extent of dispersion of points on the axes. We analyzed this plot quantitatively, following the procedure described by Tulppo et al. (1996). The entire plot was fitted into an ellipse with a center in the mean RR value; the longitudinal (major) and the perpendicular transversal (minor) axes were traced in the ellipse. The standard deviation of the data along the longitudinal (major) axis characterizes long-term variability and is called SD-2. The standard deviation of the data in the direction of transverse (minor) axis is called SD-1, and reflects beat to beat variation of inter-vals or short term variability (Tulppo et al., 1996; Brennan et al., 2001; Lerma et al., 2003).

d) Sample Entropy. In order to evaluate the complex-ity (relative lack of predictability) of heart rhythm we computed the Sample Entropy (SampEn), with the algorithm developed by Richman and Moorman, (2000). Briefly, SampEn express the negative natural logarithm of the conditional probability that two sequences similar for m points remain similar at the next point. Lower entropy values indicate a more predictable (less complex) time series.

Statistical analysisValues are reported as the mean ± S.D. The statistical significance between the mean numbers of different variables compared during sleep and wakefulness were evaluated utilizing the one-way Kruskal-Wallis analysis of variance (ANOVA) and Bonferroni as post-hoc test. The Mann-Whitney test was also uti-lized in particular cases (see Results). Correlations were evaluated by means of the Spearman’s correla-tion method. All the statistical analyses were done using the Instat£ software. The null hypothesis was rejected with a p value < 0.05.

Results

Qualitative analysisA representative polysomnographic recording of a complete sleep cycle is shown in Fig. 1. The EEG, the EEG spectrogram, the tachogram and its spectrogram

as well as the EMG are shown. The raw recording of the EEG and its spectrogram show that high ampli-tude and low frequency waves are present during NREM sleep. During this sleep state regular smooth oscillations are readily observed in the tachogram. On passing to REM sleep the amplitude of the EEG waves decreased while the frequency increased; a marked decrease in EMG activity also go along with the EEG activation. In addition, the HR and the pat-tern of organization of the intervals deeply change; larger irregularities with pronounced increases and decreases in the HR characterized REM sleep.The spectrogram of the tachogram is exhibited for one sleep cycle in Fig. 1 and for a longer period of time (that included several behavioral states transi-tions) in Fig. 2A. During NREM sleep there is a narrow HF band (at 0.25-0.3 Hz in these examples, but can reach higher frequencies as it is shown in Fig. 7) that becomes disorganized on entering into REM sleep. However, this frequency band values are variable within one REM sleep episode and among different episodes; in fact, HF activity below and above this narrow band is usually present during REM sleep (see Fig. 7 for an example).As shown in Figs. 1 and 2A, LF band potency is enhanced during REM sleep. During NREM sleep LF is reduced. In Fig. 2A it is also shown that the process of arousal usually induces a large increase in HR followed by a less marked bradycardia; this pat-tern of heart activity is translated as LF band incre-ments in the spectrogram. Interestingly, very brief, barely discernible, arousal episodes can induce LF increments. In Fig. 2B it is exhibited that when the animal is alerted by auditory stimulation there are brief increments in HR that induce enhancements of LF band in the spectrogram.The relationship between RR intervals and respirato-ry activity (RA) is shown in Figs. 3 and 4. Note that a narrow and high amplitude HF band (at ≈ 0.3 Hz in these examples) is present both in the tachogram and in the RA during NREM sleep (Fig. 3). The narrow HF band values present in the tachogram and RA is lost on entering into REM sleep. Autocorrelations as well as cross-correlations and coherence functions between tachogram and RA are shown in Fig. 4A, B and C, respectively. These analyses demonstrate the tight coupling that occurs between the tacho-gram and RA during NREM sleep. This coupling decreases during both QW and REM sleep.

36 V. BRANDO ET AL.

Quantitative analysisAs shown above, wakefulness is very heteroge-neous. Then, we were very careful to include in the analysis episodes of QW without peaks of HR acceleration (as shown in Fig. 2B). Representative intervals chosen for the quantitative analysis are shown in Fig. 5.Statistical analysis of the data (Table I) showed that the mean RR interval were longer (HR decreased) dur-ing both sleep states in comparison to QW. The SDNN and rMSSD values were also higher during NREM and REM sleep in comparison to QW (Table I).In Fig. 6, examples of Poincaré plots of three repre-sentative cats are depicted. The visual observation of the cloud of points in the plot clearly shows that the beat-to-beat variability is smaller during QW than in NREM and REM sleep; this result was confirmed by the analysis of SD-1. The overall variability

assessed by SD-2 was also larger during REM and NREM sleep than during QW (Table I).SD-1 and rMSSD values, both signs of fast HRV changes, were highly correlated (R2 = 1) during QW, NREM and REM sleep. SD-2 and SDNN that describe the long-term variability showed a moder-ate correlation during QW (R2 = 0.69), the correla-tion increased during NREM sleep (R2 = 1) and was still high during REM sleep (R2 = 0.85).Representative PSD of the tachograms are illus-trated in Fig. 7. It is evident that LF and HF changed through behavioral states. HF power increased dur-ing NREM and REM sleep in comparison to QW (Table I). LF power increased during REM sleep in comparison to QW. LF during REM was larger in comparison with NREM sleep, although the differ-ence was marginally not significant in the post-hoc test (p = 0.05). On the other hand, LF/HF was higher

Fig. 1. - Polysomnographic recording and tachogram during a sleep cycle. The EEG, the tachogram, and their cor-responding spectograms are shown during NREM and REM sleep, as well as during an arousal that followed the REM sleep episode. bpm, beats per minute; EEG, electroencephalogram; EMG, electromyogram; HR, heart rate.

HRV DURING SLEEP 37

in QW than in NREM and REM sleep (Table I). The peak frequency of the LF and HF did not change throughout behavioral states (Table I).

Sample entropyNon-parametric ANOVA yield a marginal signifi-cance (p = 0.05) in SampEn values across behavioral states (Table I). However, SampEn during NREM becomes significantly higher than during REM sleep

when both states were compared with the Mann-Whitney test (p = 0.02).

Discussion

In the present report we evaluated the HRV of the cat during sleep and wakefulness. Visual inspec-tion of the tachograms and their spectrograms

Fig. 2. - HRV during several sleep and wakefulness transitions. A. The hypnogram, the tachogram and its spectro-gram are shown across several sleep cycles. B. The tachogram and its spectrogram are exhibited during quiet (QW) and alert wakefulness (AW). bpm, beats per minute; HR, heart rate.

38 V. BRANDO ET AL.

showed noticeable differences in QW, NREM and REM sleep. The different mathematical approaches employed confirmed that HRV is different across behavioral states.

Technical considerationsThe cat is considered the standard animal model for sleep research. This species spends almost fourteen hours per day sleeping, and its sleep episodes are consolidated, which allow some studies that are dif-ficult to perform in animals with short sleep cycles such as rodents (Ursin and Sterman, 1981; Zepelin et al., 2005). In addition, the cat exhibits weak circa-dian influences compared to other species (Sterman

et al., 1965; Mishina et al., 2006); therefore, this model also excels in differentiating sleep from circadian regulation in cardiovascular parameters, which may complicate the interpretation of results (Burgess et al., 1997).Cardiovascular control during sleep is often com-pared with that occurs during wakefulness (Silvani and Lenzi, 2005). However, wakefulness entails a wide repertoire of physiological responses (move-ments for exploration, feeding, etc.). In order to reduce physiological variability, we took advantage of recording cats under semi-restricted conditions. In such experimental setting, the unique difference across behavioral states is the state per se; in other

Fig. 3. - Tachogram and respiratory activity during a sleep cycle. The tachogram and its spectrogram, as well as the respiratory activity and its spectrogram are shown. bpm, beats per minute; HR, heart rate; RA, respiratory activity.

HRV DURING SLEEP 39

words, there are not changes in the body position or movements that would alter the cardiovascular physiology. Furthermore, it is important to con-sider that in the quantitative analysis of this study, wakefulness was not only quiet, but also relaxed (an example is shown in Fig. 5). Periods of alert wake-fulness, with peaks of HR (as shown in Fig. 2) were not included in the quantitative analysis.

HRV indexes and sleepStatistical methods

Our data showed that the HR during QW is higher than during NREM and REM sleep; there were not

significant differences between the last two states. Similar findings were obtained in cats (Baust and Bohnert, 1969), rats (del Bo et al., 1982; Yang et al., 2003), and also in humans (Vaughn et al., 1995); however in humans the common trend in the litera-ture is that during REM sleep HR is higher than in NREM sleep. Phasic periods of REM sleep that are marked by a high frequency of eyes movements, muscle twitches and, by burst of PGO waves (in cats), are accompanied by surges in HR (as well as in blood pressure) in several species (Parmeggiani, 2005); a predominance of number and/or intensity of phasic periods in humans could explain the HR acceleration during this sleep state.

Fig. 4. - Relationship between the tachogram and the respiratory activity. A. Autocorrelation function (ACF) of the tachogram during quiet wakefulness (QW), non REM (NREM) and REM sleep. B. Cross correlation function (CCF) between tachogram and respiratory activity. C. z´-coherence between the tachogram and respiratory activity.

40 V. BRANDO ET AL.

In comparison to QW, the SDNN and rMSSD increased during both NREM and REM sleep. Hence, statistical evaluation suggests that both short and long-term variability increased during sleep in comparison to QW.

Poincaré plots

HRV assessed by Poincaré plots resulted in an increase in the short-term variability (expressed by SD-1) and long-term variability (SD-2) during both NREM and REM sleep in comparison to QW.As we stated above, SD-1 and rMSSD are both of them are indexes of beat-to-beat or short-term vari-ability (Task Force of the ESC and the NASPE, 1996; Tulppo et al., 1996). In agreement with this concept, we found a perfect correlation between both values during wakefulness and sleep. Correlation between SD-2 and SDNN, both representing long-term (over-all) variability, have also been reported (Task Force of the ESC and the NASPE, 1996; Brennan et al., 2001). In fact, we found that SDNN and SD-2 were highly correlated during NREM and REM sleep, but the correlation decreased during QW.Raetz et al. (1991) analyzed the HRV in the cat during wakefulness and sleep with Poincaré plots (but without analyzing SD-1 and SD-2) (Raetz et al., 1991). The scatter-plots depicted in their paper have similarities with the plots presented here and, in accordance to our results, they showed a low beat-to-beat variability during wakefulness. However, in contrast to our study, utilizing the interval range as a measure of long term (overall) HRV they observed a large variability during wakefulness, as was previ-ously reported (Baust and Bohnert, 1969). We con-sider that this difference could be due to the “active” type of wakefulness analyzed in these studies (i.e., animals recorded under unrestrained conditions) that produces long trains of high HR which are common-ly followed by marked bradycardia; this sequence of events produces an increase in the overall RR range (Raetz et al., 1991). REM sleep also had a larger overall range of intervals in comparison to NREM sleep (Baust and Bohnert, 1969; Raetz et al., 1991). On the contrary, we observed that there was not difference in the long-term variability analyzed by means of SD-2 between REM vs. NREM sleep; the differences in the analytical tools may explain this discrepancy.

Spectral analysis

The frequency domain analysis is especially impor-tant because it is widely used, not only in preclinical, but also in clinical studies. To the best of our knowl-edge, this is the first report that performed heart rate spectral analysis during sleep in the cat.

Fig. 5. - RR intervals during quiet wakefulness and sleep. The regularity of the oscillations during NREM compared to REM sleep and quiet wakefulness are evident. Tachograms were constructed from 750 RR intervals during wakefulness and sleep (700 of these RR intervals are shown in the graphics). The horizontal line indicates the mean RR interval duration.

HRV DURING SLEEP 41

We found that, compared to QW, HF power increased during NREM and REM sleep. No significant differ-ences were found in HF between REM and NREM sleep. However, note that although the narrow HF frequency peak linked to respiratory activity during NREM sleep becomes disorganized during REM sleep, higher and lower frequencies than the respira-tory rate that were included as HF band, did increase during REM sleep. In comparison to wakefulness, an HF power increment during NREM sleep has been reported both in humans and rats (Berlad et al., 1993; Yang et al., 2003). Whereas it has been described that HF decreases during human REM sleep (Berlad et al., 1993), there are also reports in humans and rats showing that HF values do not decrease in the transi-tion from NREM to REM sleep (Ako et al., 2003; Yang et al., 2003), as in the present study.LF increased during REM sleep in comparison to QW, and there was a clear trend to increase in comparison to NREM sleep. Note that we used the minimal num-ber of animals necessary to produce reliable scientific data and a conservative non-parametric statistical test. We believe that with an increase in “n” the difference between REM and NREM sleep would become sig-nificant. In agreement with our results, relative large LF power values were usually observed during REM sleep in man (Berlad et al., 1993; Ako et al., 2003).The LF/HF ratio was larger during QW; this is also a common finding both in humans and rats (Jurysta et al., 2003; Yang et al., 2003).

Complexity

In tachograms as in others time series, the concept of variability could be expanded with the analysis of the complexity of the data distribution. This point is par-ticularly relevant in studies where heart rate is modi-fied by the fluctuation of the autonomic flow induced by sleep or respiration (Porta et al., 2001; Viola et al., 2011). Our data showed that the regular tachogram of NREM sleep yielded larger complexity (SampEn) values in comparison to REM sleep; interestingly, similar results were obtained by Vigo et al., (2010) in humans. In fact, Lewis and Short, (2007) have shown that SampEn of the RR intervals is larger during met-ronomic (regular) breathing compared to spontaneous breathing. Since sympathetic activation reduces while parasympathetic increases entropy of RR time series (Yeragani et al., 1993), the intense variation in the activity of the autonomic system during REM and NREM sleep may explain our results (see below).

The autonomic system as a determinant of the HRV during sleep

The interactions between local reflex and central cardiovascular regulatory mechanisms are complex, and many of the control mechanisms involved are intrinsically non-linear (Silvani and Lenzi, 2005). The neural control of the cardiovascular system

Table I. - Statistical, spectral, Poincaré indexes of HRV and sample entropy during sleep and wakefulness.

QW NREM REM P < 0.05

Statistical Indexes

Mean RR (ms) 302.7 ± 23.4 344.0 ± 26.3 349.9 ± 30.5 QW vs. NREM, QW vs. REM

SDNN (ms) 14.9 ± 4.3 27.1 ± 12.2 30.1 ± 11.5 QW vs. NREM, QW vs. REM

rMSSD (ms) 7.9 ± 3.5 18.6 ± 13.8 19.3 ± 14.2 QW vs. NREM, QW vs. REM

Spectral Indexes

LF (ms2/Hz) 134.0 ± 91.4 217.0 ± 103.8 425.3 ± 368.3 QW vs. REM

HF (ms2/Hz) 64.11± 70.0 553.3 ± 800.9 486.9 ± 623.1 QW vs. NREM, QW vs. REM

LF/HF 3.5 ± 2.1 1.33 ± 1.3 1.89 ± 1.6 QW vs. NREM, QW vs. REM

LF Peak (Hz) 0.05 ± 0.01 0.05 ± 0.03 0.04 ± 0.03 NS

HF Peak (Hz) 0.3 ± 0.1 0.27 ± 0.1 0.27 ± 0.1 NS

Poincaré Indexes

SD1 (ms) 5.6 ± 2.4 13.1 ± 9.8 13.4 ± 10.2 QW vs. NREM, QW vs. REM

SD2 (ms) 24.7 ± 9.5 37.2 ± 14.5 43.5 ± 14.5 QW vs. NREM, QW vs. REM

Sample Entropy 1.12 ± 0.27 1.15 ± 0.25 0.96 ± 0.18 P = 0.05

QW, quiet wakefulness; NREM, non-REM sleep; REM, REM sleep; NS, not significant.

42 V. BRANDO ET AL.

is accomplished by the autonomic outflow to the heart and vessels. The autonomic outflow includes both the reflex contribution of peripheral factors (baroreceptors, chemoreceptors, etc.) and central commands that change as a function of behavioral states, emotion, exercise, etc. The summation of these influences on the autonomic cardiovascular outflow makes it difficult to evaluate the effect of sleep on a single factor, such as the HRV (Silvani and Lenzi, 2005).A wealth of data indicate that the basic autonomic feature of NREM sleep is the functional prevalence of parasympathetic influences associated with quies-cence of sympathetic activity (Parmeggiani, 2005). NREM sleep is characterized by a down-regulation of cardiovascular activity; however the magnitude of this down-regulation depends on the species. In the cat, there is a decrease in blood pressure and heart rate, while the stroke volume is unchanged (Parmeggiani, 2005).

The basic autonomic feature of REM sleep is the great variability in sympathetic activity (phasic increments in a background of a tonic decrease in sympathetic tone) associated with phasic changes in tonic parasympathetic drive (Baust and Bohnert, 1969; Parmeggiani, 2005). In contrast to wakeful-ness and NREM sleep, this variability is attributed to a feedback-independent, “open loop” system that regulates REM sleep (Parmeggiani, 2005).In the present study, during NREM sleep we observed a decrease in HR, a strong regularity of the variations of the RR intervals, as well as an increase in the complexity of the RR series. Furthermore, in comparison to QW, during both REM and NREM sleep there was a larger short and long-term variability of the intervals. What is the role of vagal activity in this variability? The type of HRV observed in periods of high respiratory sinus arrhythmia as in NREM sleep, has been attributed to vagal influences (Pomeranz et al., 1985). It is

Fig. 6. - Poincaré plots during quiet wakefulness and sleep. The distribution or “clouds” of points are shown for 3 representative cats (C1 to C3) during quiet wakefulness, NREM and REM sleep.

HRV DURING SLEEP 43

known that decelerations are mostly accompanied by prior accelerations in HR and increases in blood pressure, and involve baroreceptor activation (Baust and Bohnert, 1969; Dickerson et al., 1993a). In dogs, these transient pauses of the HR are present during sleep in the transition from NREM to REM sleep, and also during phasic periods of REM sleep (Dickerson et al., 1993a). Primary vagally-mediated

decelerations in HR (not preceded by increases in HR or arterial blood pressure), are present mainly during tonic REM sleep in the cat (Verrier et al., 1998). In humans, Guilleminault et al. (1984) observed periods of sinus arrest during REM sleep in apparently healthy young adults; some of these individuals experienced periods of asystole of up to 9 s during REM sleep (mainly during phasic REM sleep), and the administration of muscarinic block-ers reduced the duration of the nocturnal asystole (Guilleminault et al., 1984). The authors concluded that the nocturnal asystole was the result of exagger-ated vagal tone. Hence, these data suggest a strong parasympathetic activation not only during NREM sleep, but also during REM sleep, and is consistent with our results of higher HF and lower LF/HF ratio during REM sleep in comparison to QW.We also found that LF was higher in REM sleep than in QW; in addition, a clear trend to higher val-ues of LF was observed in REM sleep when com-pared with NREM sleep. It is likely that the phasic increase in sympathetic activity during REM sleep is responsible for the increase in LF, as was suggested in humans (Ako et al., 2003). In fact, phasic periods of REM sleep seem to be related to sympathetic-induced increases in HR; in dogs, 90% of the HR surges are concentrated during phasic REM sleep (Dickerson et al., 1993b). Therefore, phasic incre-ments in the activity of both autonomic branches during REM sleep could explain the short and long-term HRV found during this state.LF value was low during QW. This result is due to the fact that in our preparation the animals were quiet and relaxed, and this index deeply depends on the level of arousal (Task Force of the ESC and the NASPE, 1996; Bonnet and Arand, 1997), as was shown in Fig. 2.

Conclusions and future directions

In the present report, we conducted a meticulous analysis of the HRV during sleep and wakefulness in the cat. HRV increases during sleep, but the charac-teristics of this variability differ between REM and NREM sleep.Although positron emission tomography (PET) stud-ies in humans suggest a role of the right amygdaloid complex in the HRV during REM sleep (Desseilles et al., 2006), the neural mechanisms that modulate

Fig. 7. - Power spectrum histograms during quiet wake-fulness and sleep. These representative histograms show large differences between quiet wakefulness and sleep. PSD, power spectral density.

44 V. BRANDO ET AL.

HRV during sleep are not known in detail. We consider that the cat is an excellent model for deter-mining the central mechanisms that are responsible for the control of HRV during sleep. Here, we have completed the first step of our long-term plan by set-ting the foundation of the HRV during sleep in this animal model.

AcknowledgementsThis study was supported by the “Programa de Desarrollo de Ciencias Básicas, PEDECIBA” and by the ANII-FCE-2-2011-1-7030 grant from Uruguay. We are grateful to Drs. Jack Yamuy and Paola Contreras for their critical comments of the manuscript.

References

Abbott J.A. Heart rate and heart rate variability of healthy cats in home and hospital environments. J. Feline Med. Surg., 7: 195-202, 2005.

Ako M., Kawara T., Uchida S., Miyazaki S., Nishihara K., Mukai J., Hirao K., Ako J., Okubo Y. Correlation between electroencephalography and heart rate variability during sleep. Psychiatry Clin. Neurosci., 57: 59-65, 2003.

Akselrod S., Gordon D., Ubel F.A., Shannon D.C., Berger A.C., Cohen R.J. Power spectrum analysis of heart rate fluctuation: a quantitative probe of beat-to-beat cardiovascular control. Science, 213: 220-222, 1981.

Aserinsky E. and Kleitman N. Regularly occurring periods of eye motility, and concomitant phenom-ena, during sleep. Science, 118: 273-274, 1953.

Baust W. and Bohnert B. The regulation of heart rate during sleep. Exp. Brain Res., 7: 169-180, 1969.

Berlad I.I., Shlitner A., Ben-Haim S., Lavie P. Power spectrum analysis and heart rate variability in Stage 4 and REM sleep: evidence for state-specific changes in autonomic dominance. J. Sleep Res., 2: 88-90, 1993.

Bonnet M.H. and Arand D.L. Heart rate variability: sleep stage, time of night, and arousal influences. Electroencephalogr. Clin. Neurophysiol., 102: 390-396, 1997.

Brennan M., Palaniswami M., Kamen P. Do existing measures of Poincare plot geometry reflect nonlin-ear features of heart rate variability? IEEE Trans Biomed. Eng., 48: 1342-1347, 2001.

Burgess H.J., Trinder J., Kim Y., Luke D. Sleep and cir-cadian influences on cardiac autonomic nervous sys-tem activity. Am. J. Physiol., 273: H1761-1768, 1997.

Burr R.L. Interpretation of normalized spectral heart rate variability indices in sleep research: a critical review. Sleep, 30: 913-919, 2007.

Castro S., Falconi A., Chase M.H., Torterolo P. Coherent neocortical 40-Hz oscillations are not present during REM sleep. Eur. J. Neurosci., 37: 1330-1339, 2013.

Contreras P., Canetti R., Migliaro E.R. Correlations between frequency-domain HRV indices and lagged Poincare plot width in healthy and diabetic subjects. Physiol. Meas., 28: 85-94, 2007.

del Bo A., Ledoux J.E., Tucker L.W., Harshkfield G.A., Reis D.J. Arterial pressure and heart rate changes during natural sleep in rat. Physiol. Behav., 28: 425-429, 1982.

Desseilles M., Vu T.D., Laureys S., Peigneux P., Degueldre C., Phillips C., Maquet P. A prominent role for amygdaloid complexes in the Variability in Heart Rate (VHR) during Rapid Eye Movement (REM) sleep relative to wakefulness. Neuroimage, 32: 1008-1015, 2006.

Dickerson L.W., Huang A.H., Nearing B.D., Verrier R.L. Primary coronary vasodilation associated with pauses in heart rhythm during sleep. Am. J. Physiol., 264: R186-196, 1993a.

Dickerson L.W., Huang A.H., Thurnher M.M., Nearing B.D., Verrier R.L. Relationship between coronary hemodynamic changes and the phasic events of rapid eye movement sleep. Sleep, 16: 550-557, 1993b.

Fronczek R., Overeem S., Reijntjes R., Lammers G.J., van Dijk J.G., Pijl H. Increased heart rate variability but normal resting metabolic rate in hypocretin/orexin-deficient human narcolepsy. J. Clin. Sleep Med., 4: 248-254, 2008.

Furley D. and Wilkie J.S. Galen on respiration and the arteries. 1984. New Jersey: Princeton University press.

Guilleminault C., Pool P., Motta J., Gillis A.M. Sinus arrest during REM sleep in young adults. N. Engl. J. Med., 311: 1006-1010, 1984.

Jurysta F., van de Borne P., Migeotte P.F., Dumont M., Lanquart J.P., Degaute J.P., Linkowski P. A study of the dynamic interactions between sleep EEG and heart rate variability in healthy young men. Clin. Neurophysiol., 114: 2146-2155, 2003.

Lerma C., Infante O., Perez-Grovas H., Jose M.V. Poincare plot indexes of heart rate variability cap-

HRV DURING SLEEP 45

ture dynamic adaptations after haemodialysis in chronic renal failure patients. Clin. Physiol. Funct. Imaging, 23: 72-80, 2003.

Lewis M.J. and Short A.L. Sample entropy of electro-cardiographic RR and QT time-series data during rest and exercise. Physiol. Meas., 28: 731-744, 2007.

Macwilliam J.A. Some applications of physiology to medicine: III. Blood pressure and heart rate action in sleep and dreams: Their relation to haemor-rhages, angina, and sudden death. Br. Med. J., 2: 1196-1200, 1923.

Malik M. Geometrical Methods for Heart Rate Variability Assessment. 47-61. In: Malik M and Camm J. Armonk, N. F. P. C. I. (Ed.), Heart Rate Variability, 1995.

Malliani A., Lombardi F., Pagani M. Power spectrum analysis of heart rate variability: a tool to explore neural regulatory mechanisms. Br. Heart J., 71: 1-2, 1994.

Migliaro E.R. and Contreras P. Short-Term Analysis of Heart Rate Variability in Diabetic Patients. Latin American Archives of Cardiovascular Sciencies, 31-36, 2001.

Migliaro E.R., Contreras P., Bech S., Etxagibel A., Castro M., Ricca R., Vicente K. Relative influence of age, resting heart rate and sedentary life style in short-term analysis of heart rate variability. Braz. J. Med. Biol. Res., 34: 493-500, 2001.

Mishina M., Watanabe N., Watanabe T. Diurnal variations of blood pressure in cats. J. Vet. Med. Sci., 68: 243-248, 2006.

Muller J.E., Ludmer P.L., Willich S.N., Tofler G.H., Aylmer G., Klangos I., Stone P.H. Circadian variation in the frequency of sudden cardiac death. Circulation, 75: 131-138, 1987a.

Muller J.E., Stone P.H., Turi Z.G., Rutherford J.D., Czeisler C.A., Parker C., Poole W.K., Passamani E., Roberts R., Robertson T., et al. Circadian varia-tion in the frequency of onset of acute myocardial infarction. N. Engl. J. Med., 313: 1315-1322, 1985.

Muller J.E., Tofler G.H., Willich S.N., Stone P.H. Circadian variation of cardiovascular disease and sympathetic activity. J Cardiovasc Pharmacol, 10 Suppl 2: S104-109; discussion S110-101, 1987b.

Parmeggiani P.L. Physiologic regulation in sleep. 185-191. In: Kryger M.H. et al. (Eds.), Principles and practices of sleep medicine. Philadelphia, Elsevier-Saunders, 2005.

Pepine C.J. Circadian variations in myocardial isch-emia. Implications for management. Jama, 265: 386-390, 1991.

Pomeranz B., Macaulay R.J., Caudill M.A., Kutz I., Adam D., Gordon D., Kilborn K.M., Barger A.C., Shannon D.C., Cohen R.J., et al. Assessment of autonomic function in humans by heart rate spec-tral analysis. Am. J. Physiol., 248: H151-153, 1985.

Pontet J., Contreras P., Curbelo A., Medina J., Noveri S., Bentancourt S., Migliaro E.R. Heart rate vari-ability as early marker of multiple organ dysfunc-tion syndrome in septic patients. J. Crit. Care, 18: 156-163, 2003.

Porta A., Guzzetti S., Montano N., Furlan R., Pagani M., Malliani A., Cerutti S. Entropy, entropy rate, and pattern classification as tools to typify com-plexity in short heart period variability series. IEEE Trans. Biomed. Eng., 48: 1282-1291, 2001.

Raetz S.L., Richard C.A., Garfinkel A., Harper R.M. Dynamic characteristics of cardiac R-R intervals during sleep and waking states. Sleep, 14: 526-533, 1991.

Richman J.S., Moorman J.R. Physiological time-series analysis using approximate entropy and sample entropy. Am. J. Physiol. Heart Circ. Physiol., 278: H2039-2049, 2000.

Sei H., Morita Y., Morita H., Hosomi H. Long-term profiles of sleep-related hemodynamic changes in the postoperative chronic cat. Physiol. Behav., 46: 499-502, 1989.

Silvani A. and Lenzi P. Reflex cardiovascular con-trol in sleep. In: Parmeggiani, P.L., Velluti, R.A. (Eds.), The physiologic nature of sleep. London, Imperial College Press, 2005.

Somers V.K., Dyken M.E., Mark A.L., Abboud F.M. Sympathetic-nerve activity during sleep in normal subjects. N. Engl. J. Med., 328: 303-307, 1993.

Sterman M.B., Knauss T., Lehmann D., Clemente C.D. Circadian sleep and waking patterns in the laboratory cat. Electroencephalogr. Clin. Neurophysiol., 19: 509-517, 1965.

Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology Heart rate variability: standards of measurement, physiological interpretation and clinical use. Circulation, 93: 1043-1065, 1996.

Torterolo P., Benedetto L., Lagos P., Sampogna S., Chase M.H. State-dependent pattern of Fos protein expression in regionally-specific sites within the preoptic area of the cat. Brain Res., 1267: 44-56, 2009.

Tulppo M.P., Makikallio T.H., Takala T.E., Seppanen T., Huikuri H.V. Quantitative beat-to-beat analysis of heart rate dynamics during exercise. Am. J. Physiol., 271: H244-252, 1996.

46 V. BRANDO ET AL.

Ursin R. and Sterman M. Manual for standard-ized scoring of sleep and waking states in adult cats. 1981. Los Angeles: BIS/BRI, University of California.

Vanoli E., Adamson P.B., Ba L., Pinna G.D., Lazzara R., Orr W.C. Heart rate variability during spe-cific sleep stages. A comparison of healthy sub-jects with patients after myocardial infarction. Circulation, 91: 1918-1922, 1995.

Vaughn B.V., Quint S.R., Messenheimer J.A., Robertson K.R. Heart period variability in sleep. Electroencephalogr. Clin. Neurophysiol., 94: 155-162, 1995.

Verrier R.L. and Harper R.M. Cardiovascular physi-ology: central and autonomic regulation. In: Kryger M.H. et al. (Eds.), Principles and practices of sleep medicine. Philadelphia, 2010.

Verrier R.L., Lau T.R., Wallooppillai U., Quattrochi J., Nearing B.D., Moreno R., Hobson J.A. Primary vagally mediated decelerations in heart rate dur-ing tonic rapid eye movement sleep in cats. Am. J. Physiol., 274: R1136-1141, 1998.

Verrier R.L. and Mittleman M.A. Sleep-related car-diac risk. 1161-1170. In: Kryger M.H. et al. (Eds.), Principles and practices of sleep medicine. Philadelphia, Elsevier-Saunders, 2005.

Verrier R.L., Muller J.E., Hobson J.A. Sleep, dreams, and sudden death: the case for sleep as an auto-nomic stress test for the heart. Cardiovasc. Res., 31: 181-211, 1996.

Vigo D.E., Dominguez J., Guinjoan S.M., Scaramal M., Ruffa E., Solerno J., Siri L.N., Cardinali D.P. Nonlinear analysis of heart rate variability within independent frequency components during the sleep-wake cycle. Auton. Neurosci., 154: 84-88, 2010.

Viola A.U., Tobaldini E., Chellappa S.L., Casali K.R., Porta A., Montano N. Short-term complexity of cardiac autonomic control during sleep: REM as a potential risk factor for cardiovascular system in aging. PLoS One, 6: e19002, 2011.

Von Neumann J.R. The mean square successive dif-ference. Ann. Math. Stat., 12: 153-162, 1941.

WHO Global health risk: mortality and burden of disease attributable to selected major risks. 2009. Geneva: WHO Press.

Woo M.A., Stevenson W.G., Moser D.K., Trelease R.B., Harper R.M. Patterns of beat-to-beat heart rate variability in advanced heart failure. Am. Heart J., 123: 704-710, 1992.

Yang C.C., Shaw F.Z., Lai C.J., Lai C.W., Kuo T.B. Relationship between electroencephalogram slow-wave magnitude and heart rate variability during sleep in rats. Neurosci. Lett., 336: 21-24, 2003.

Yeragani V.K., Srinivasan K., Vempati S., Pohl R., Balon R. Fractal dimension of heart rate time series: an effective measure of autonomic function. J. Appl. Physiol. (1985), 75: 2429-2438, 1993.

Zepelin H., Siegel J.M., Tobler I. Mamalian Sleep. 91-100. In: Kryger M.H. et al. (Eds.), Principles and practices of sleep medicine. Philadelphia, Elsevier-Saunders, 2005.

Introduction

Autophagy is a process responsible for the degra-dation of intracellular material and is evolution-arily conserved among all eukaryotes. In autophagy, cytoplasmic components are engulfed by double-membrane-bound structures (autophagosomes) and delivered to lysosomes/vacuoles for degradation (Mizushima et al., 2008). Recent studies report that autophagy has a great variety of physiological and pathophysiological roles as occurs in Alzheimer disease and amyotrophic lateral sclerosis (Hara et al., 2006; Kabuta et al., 2006; Ravikumar et al., 2010; Nixon, 2013), neuronal death and cerebral ischemia (Shi et al., 2012). In addition, intracellular autophagic processes, including macroautophagy,

microautophagy and chaperone-mediated autophagy (CMA) are important for their contribution to pep-tide generation and/or sorting and presentation by Major Histocompatibility Complex (MHC) class II molecules to cluster of differentiation 4 (CD4) T lymphocytes (Strawbridge and Blum, 2007; Crotzer and Blum, 2008). Autophagy is also responsible for the turnover of aged intracellular proteins and organ-elles, participates in the elimination of microorgan-isms, neoplastic cells and is also an effective anti-aging mechanism (Lleo et al., 2007). Autophagy, a process of “self-eating”, is often activated in response to starvation and stress (Levine and Klionsky, 2004). The molecular details of autophagy have been well characterized in Saccharomyces cerevisiae by stud-ies in proteins encoded by a group of genes that

Sheep primary astrocytes under starvation conditions express higher amount of LC3 II

autophagy marker than neuronsE. MURA, G. LEPORE, M. ZEDDA, S. GIUA, V. FARINA

Department of Veterinary Medicine, University of Sassari, Italy

A B S T R A C T

Autophagy is a general term for the degradation of cytoplasmic components within lysosomes. Recent studies have clearly demonstrated that autophagy has a greater variety of physiological and pathophysiological roles than expected, such as starvation adaptation, intracellular protein and organelle clearance, development, anti-aging, elimination of microorganisms, cell death, tumor suppression and antigen presentation. MAP-LC3 is one of the most common markers to evaluate autophagic processes.In our study, the autophagic activity in neurons and astrocytes from sheep brain under starving conditions was evaluated. In order to detect LC3 immunoreactivity, confocal analysis by double immunofluorescence was per-formed together with the cell type markers: GFAP to identify astrocytes, E-III tubulin to identify neurons. The results show that astrocytes are characterized by LC3-positive areas, which increase in a time-dependent manner. In contrast, LC3 immunoreactivity was very weak in neurons. Therefore, it can be assumed that astrocytes show a higher capability than neurons to cope with stress and exhibit a stronger autophagic response.

Key wordsAutophagy • LC3 II • Astrocytes • Neurons • Sheep

Corresponding Author: Prof. Vittorio Farina, Department of Veterinary Medicine, University of Sassari, via Vienna 2, 07100 Sassari, Italy - Tel.: +39 079 229460 - Email: [email protected]


48 E. MURA ET AL.

regulate autophagy, which are genetically conserved in eukaryotes, up to man (Kihara et al., 2001). Autophagy is regulated by several kinases, particu-larly serine/threonine protein kinases such as mam-malian Target Of Rapamycin (mTOR), adenosine monophosphate (AMP)-activated protein kinase, protein kinase B (Akt), mitogen-activated protein kinase (ERK, p38 and JNK) and protein kinase C that are often deregulated in cancer and are important therapeutic targets (Sridharan et al., 2011). Finally, the microtubule-associated protein light chain 3 (MAP-LC3) is a major constituent of the autophago-some, a double membrane structure that sequesters the target organelle/protein and then fuses with endolysosomes where the contents – and LC3 – are degraded. During autophagy, the cytoplasmic form of MAP LC3 (LC3 I) is processed and recruited to the autophagosomes, where LC3 II is generated by site-specific proteolysis and lipidation near to the C-terminus. The hallmark of autophagic activation is thus the formation of cellular autophagosomes containing LC3 II, while autophagic activity is mea-sured biochemically as the amount of LC3 II that accumulates in the absence or presence of lysosomal activity. LC3 is synthesized as pro-LC3, which is immediately processed by Atg4 into the cytosolic form, LC3 I. During autophagosome formation, LC3 I covalently links to phosphatidylethanolamine (PE) and is incorporated into autophagosome membranes where it recruits cargo. This event is mediated by the autophagocytosis-associated protein 3 and 7 (Atg3 and Atg7). It was recently shown that WIPI2 gene, which in humans encodes WD repeat domain phosphoinositide-interacting protein 2, the human orthologue of Atg18, appears to be required for LC3 lipidation and is involved in positively regulating this process (Iwata et al., 2005; Behrends et al., 2010; Polson et al., 2010; Riley et al., 2010).In this work, the protocols have been followed necessary to obtain primary neural cells from sheep brain (Mura et al., 2012). Aim of this study was to evaluate LC3 immunoreactivity in primary neural cell cultures. Since autophagy may differ in neurons and non-neuronal cells (Yue et al., 2009; Di Malta et al., 2012a, b), in this investigation the possible differences in LC3 II immunoreactivity between astrocytes and neurons in primary cultures from fetal sheep brain have been examined under starving-induced stress conditions.

Methods

Tissue collectionFetuses were available at a local slaughterhouse, when sheep were slaughtered casually pregnant, so fetuses underwent immediate death. Among them, five 40-day fetuses were collected and quickly transported to lab hood on ice. Samples came from animals slaughtered for human consumption and treated in accordance with European laws (86/609/EEC) following the procedures concerning animal welfare during slaughter process. The gestational age has been chosen as suitable to obtain viable neu-rons and astrocytes from sheep brain (Richard et al., 1998). Under sterile conditions, the cranial cavity was opened, the brain visualized and the frontal cor-tex isolated. The whole process of tissue sampling was carried out in two hours after slaughtering. Fetal age was established measuring crown-rump length using a standard reference table (McGeady et al., 2006). Samples were minced into small fragments and stored in liquid nitrogen.

Primary cultures from sheep brainA papain dissociation system kit (Worthington Biochemical Co., Lakewood, NJ, USA) was used to dissociate cells following what previously reported (Mura et al., 2012). Cells were then suspended in basal medium consisting of a 1:1 mixture of Dulbecco’s modified Eagle’s Medium (DMEM) and Ham’s F-12, supplemented with penicillin (30 mg/l), streptomycin (50 mg/l), sodium bicarbonate (2.4 g/l), insulin (10 Pg/ml), transferrin (10 Pg/ml), sodium selenite (10-8 M) and 10% fetal calf serum (FCS, Sigma, St. Louis, MO, USA). Cells were seeded in 24-well plates, previously coated with poly-L-lysine at a density of 5 x 105/ml medium and incubated at 37°C. A minimum of 200 cells per field were count-ed under confocal microscope at 40X magnification and percentages were determined from the obtained mean values of cells immunopositive to anti-LC3 II antibody. Cells were classified as unstained when they were completely not immunoreactive.

In vitro starvation models and rapamycin treatmentPrimary cultures at around 80% confluence (7 days from seeding) underwent one of the following pro-tocols:

DIFFERENCES IN AUTOPHAGY BETWEEN NEURONS AND ASTROCYTES IN VITRO 49

1. basal medium for 72 h;2. medium containing the autophagy inductor rapa-

mycin (Sigma) 50 PM, following the protocol of Tsvetkov et al. (2010) as positive control for 72 h;

3. serum-deprived medium for 24, 48 and 72 h.

ImmunofluorescenceIn order to detect LC3 immunoreactivity in neurons and astrocytes, double immunofluorescence was performed using LC3 II antibody mixed together with cell type markers. In particular, the procedure was as follows: the primary cultures after treatment were fixed in 4% paraformaldehyde at room temper-ature for 20 min. After fixation, cells were washed with phosphate buffer, then permeabilized with 0.1 % Triton X100 for 10 min, treated with 5% bovine serum albumin (BSA, Sigma) buffer for 30 min to block endogenous nonspecific sites and incubated overnight with one of the following monoclonal primary antibodies against: glial fibrillary acidic protein (GFAP), diluted 1:500 to identify astrocytes, and E-III tubulin, diluted 1:200 to identify neurons. In addition, a polyclonal anti LC3 II antibody, dilut-ed 1:100 was used. Cells were then incubated with a mixture of two secondary antibodies anti-rabbit rhodamine (TRITC), diluted 1:100, and anti-mouse fluorescein (FITC), diluted 1:100. All antibod-ies were from Sigma. All samples underwent blue nuclear counterstaining for fluorescence microscopy with Hoechst no. 33342 reagent. Finally, labeled cultures were mounted with FlourSaveTM Reagent (Calbiochem, San Diego, CA, USA). Cultures were observed and photographed using a Leica TCS con-focal microscope. Negative controls were performed by substituting primary antibodies with bovine serum albumin in PBS. Under these conditions no immunostaining was observed.

Results

Characterization of primary culturesPrimary cell cultures obtained from sheep brain were viable and contained a heterogeneous cell population. Two cell populations were mainly pres-ent: i) E-III tubulin immunoreactive cells that were neurons, mostly bipolar-shaped displaying small and round cell bodies with thin extensions (Fig. 1), ii) GFAP-immunoreactive cells showing the typical

morphology of astrocytes with an irregular body shape and multiple processes (Fig. 2). Because of the marked discrepancy in the percentage of the two cell types in favor of astrocytes, a bar graph has been believed unnecessary.

Fig. 1. - Cell characterization. Immunocytochemical detection of the neuron-specific marker E-III tubulin in primary cultures. Nuclei are counterstained with Hoechst blue flurescent dye no. 33342. Neurons are mostly bipolar-shaped and display small and round cell bodies with thin extensions. Confocal microscopy. Bar = 50 Pm.

Fig. 2. - Cell characterization. Immunocytochemical detection of the glial-specific marker GFAP in primary cultures. Astrocytes show an irregular body shape and multiple processes. Confocal microscopy. Bar = 50 Pm.

50 E. MURA ET AL.

Detection of LC3 II in basal and rapamycin conditionsIn order to verify LC3 II expression in basal con-ditions, cells were cultured in medium containing 10% FCS. Immunofluorescence assay did not show LC3 II positive cells. As showed in Fig. 3, neither showed neurons nor astrocytes positive areas to LC3 II in basal conditions. In contrast, treatment with 50 PM rapamycin induced LC3 II immunopositive areas in both neurons and astrocytes (Fig. 4).

Detection of LC3 II in starving conditionsThe cultures grew in serum deprivation conditions for 24, 48 and 72 h. At 24 h, astrocytes started showing LC3 II immunopositive areas, that became more extended at 48 and 72 h (Fig. 5). Indeed, under serum-starving conditions, astrocytes are the cell type showing the highest percentage of LC3 II immunopositive cells. Indeed, the bar graph in Fig. 6 indicates that astrocytes represent 30, 60 and 90% of LC3 II immunopositive cells after 24, 48 and 72 h starving condition respectively. Those cells are indeed characterized by exten-sive LC3 II-immunoreactive cytoplasmic vacuol-ization involving the perinuclear region as well as many cytoplasmic processes. The morphological

features of astrocytes in toto were retained almost unchanged. Neurons displayed a very weak LC3 II immunopositivity after 24 h (no more than 2% posi-tive cells), that was still weak after 48 h treatment (no more than 7% positive cells). Moreover, at 72 h the almost complete disappearance of neuronal extensions could be observed together with the absence of LC3 II immunopositivity (Figs. 6, 7).

Discussion

In the present work a protocol was followed to obtain sheep primary cultures and an immunocytochemical assay was performed to evaluate LC3 II expression in neurons and astrocytes from fetal sheep brain. Viable neural cell cultures were obtained success-fully from 40-day old fetuses. The LC3 II expres-sion detected during starvation was compared to that found in basal conditions and after autophagy induction by 50 PM rapamycin treatment (positive control). This resulted in an extent of cytoplasmic vacuolization that was greater than in basal condi-tions both in neurons and astrocytes. Rapamycin (mTOR inhibitor), which mimics cellular starvation by blocking signals required for cell growth and

Fig. 3. - LC3 II determination in basal conditions. Neurons (A) and astrocytes (B) do not show immunopositive areas. Confocal microscopy. Bar = 50 Pm.


Fig. 4. - LC3 II determination after a 72 h-exposure to 50 PM rapamycin. The arrows (A, B) indicate LC3 II small red-stained immunopositive areas mostly located in the perinuclear region. The separate channels, FITC for E-III tubulin (C) and GFAP (D), as well as TRITC for LC3 II (E, F) are also shown. Confocal microscopy. Bar = 50 Pm (A, C, E), 80 Pm (B, D, F).

52 E. MURA ET AL.

proliferation, has become one of the most widely-used autophagy inducers (Jung et al., 2010). The induction of autophagy was assessed by detecting an increase in the autophagosomal membrane form of LC3 II (Zeng et al., 2007; Barilli et al., 2008; Stanfel et al., 2009; Tsvetkov et al., 2010). The increased

protein expression of LC3 by rapamycin we detected was evaluated successfully by immunocytochemical assay.Confocal analysis after a time-range of 24, 48 and 72 h of serum deprivation showed abundant autophago-somes in astrocytes, infrequently observed in neu-

Fig. 5. - LC3 II determination in astrocytes at 24 h, 48 h and 72 h starvation. Cells are characterized by LC3-immunoreactive cytoplasmic areas (stained in red) in the perinuclear region as well as in many cytoplasmic proc-esses that increase in a time dependent manner. The separate channels, FITC (D, E, F) for GFAP and TRITC (G, H, I) for LC3 II are also shown. Confocal microscopy. Bar = 50 Pm (A, C), 80 Pm (B).


rons. Despite high LC3 expression was induced in neurons by 50 PM rapamycin, in starving conditions the LC3 autophagy marker is expressed in astro-cytes more markedly than in neurons. The major capability in activating autophagic mechanisms of astrocytes is appreciable not only qualitatively, but also quantitatively. Indeed, 30, 60 and 90% LC3 II-immunopositive astrocytes were found after 24, 48 and 72 h starving condition respectively. A key observation in our study is that astrocytes seem to be more responsive to stress than neurons as exhibit a stronger autophagic ability to counteract these stresses. This seems to be confirmed by the mor-phological alterations observed in neurons and not in astrocytes after 72 h starvation when co-cultured. Indeed, vacuoles were apparent in the cytoplasm of astrocytes in all three starvation times. This could explain why astrocytes are essential for neuronal survival and repair (Giffardi and Swanson, 2005).Several studies report that astrocytes appear to be more resistant than neurons to most stress conditions both in vitro and in vivo. Astrocytes may be more able than neurons to cope with increased intracellu-lar free Zn2+ because they possess a greater capacity for buffering ([Zn2+]i) excess (Dineley et al., 2000) and are generally capable to counteract stresses

more effectively than neurons due to their high antioxidant reserve (Barker et al., 1996) The more marked increase in expression levels of antioxi-dant enzymes may render astrocytes more suitable to counteract oxidative stress than neuronal cells (Schmuck et al., 2002). Finally, in vivo observations show that astrocytes are less vulnerable to ischemia than neurons (Gürer et al., 2009).The role of glial cells, particularly astrocytes, in the pathophysiology of cerebral ischemia has become an important subject of research (Qin et al., 2010). Brain energetics, water content and ion homeostasis, inflammation, production of trophic factors, vas-cular regulation, neurogenesis and vasculogenesis, among others, are all under the influence of glial cells. As a consequence, glial cells, especially astro-cytes, have been considered as promising targets for novel therapeutic approaches in brain protection (Nedergaard and Dimagl, 2005).In this paper, the experimental model is based on sheep fetuses available at a local slaughterhouse, when sheep were slaughtered casually pregnant. For this reason, we emphasize that a recent (2011) Joint Statement of the Society for Neuroscience (USA), the Federation of European Neuroscience Societies, and the Japan Neuroscience Society advocates the

Fig. 6. - Percentages of LC3 II-positive cells under starving conditions at 24, 48 and 72 h. The number of immunoposi-tive astrocytes is markedly higher than neurons and it increases in a time-dependent manner.

54 E. MURA ET AL.

use of laboratory animals, but also stresses the necessity to replace and reduce their number. The use of primary cell cultures derived from the sheep, a species commonly used in animal production rep-resents a step forward towards this humane goal.

Funding sourceWork supported by a grant from the Regione Autonoma della Sardegna in the Master and Back program.

AcknowledgementsThe Authors wish to thank dr. Antonello Floris for his technical support.

References

Barilli R., Visigalli R., Sala R., Gazzola G.C., Parolari A., Tremoli E., Bonomini S., Simon A., Closs E.I., Dall’Asta V., Bussolati O. In human endothelial

Fig. 7. - LC3 II determination in neurons at 24 h, 48 h and 72 h starvation. Cells show a very weak LC3 immunopo-sitivity at all times and a strong reduction of cytoplasmic extensions at 72 h. The separate channels, FITC (D, E, F) for E-III tubulin and TRITC (G, H, I) for LC3 II are also shown. Confocal microscopy. Bar = 30 Pm (A), 50 Pm (B, C).


cells rapamycin causes mTORC2 inhibition and impairs cell viability and function. Cardiovasc. Res., 78: 563-571, 2008.

Barker J.E., Bolaños J.P., Land J.M., Clark J.B., Heales S.J. Glutathione protects astrocytes from peroxynitrite-mediated mitochondrial damage: implications for neuronal/astrocytic trafficking and neurodegeneration. Dev. Neurosci., 18: 391-396, 1996.

Behrends C., Sowa M.E., Gygi S.P., Harper J.W. Network organization of the human autophagy system. Nature, 466: 68-76, 2010.

Crotzer V.L. and Blum J.S. Cytosol to lysosome transport of intracellular antigens during immune surveillance. Traffic, 9: 10-16, 2008.

Di Malta C., Fryer J.D., Settembre C., Ballabio A. Astrocyte dysfunction triggers neurodegeneration in a lysosomal storage disorder. Proc. Natl. Acad. Sci. USA, 109: E2334-2342 (Epub 2012 Jul 23), 2012a.

Di Malta C., Fryer J.D., Settembre C., Ballabio A. Autophagy in astrocytes: A novel culprit in lyso-somal storage disorders. Autophagy, 8: 1871-1872, 2012b.

Dineley K.E., Scanlon J.M., Kress G.J., Stout A.K., Reynolds I.J. Astrocytes are more resistant than neurons to the cytotoxic effects of increased [Zn2+]

i. Neurobiol. Dis., 7: 310-320, 2000.

Giffardi R.G. and Swanson R.A. Ischemia-induced programmed cell death in astrocytes. Glia, 50: 299-306, 2005.

Gürer G.., Ozdemir Y., Erdemli E., Can A., Dalkara T. Astrocytes are more resistant to focal cerebral ischemia than neurons and die by a delayed necro-sis. Brain Pathol., 19: 630-641, 2009.

Hara T., Nakamura K., Matsui M., Yamamoto A., Nakahara Y., Suzuki-Migishima R., Yokoyama M., Mishima K., Saito I., Okano H., Mizushima N. Suppression of basal autophagy in neural cells causes neurodegenerative disease in mice. Nature, 441: 885-889, 2006.

Iwata A., Riley B.E., Johnston J.A., Kopito R.R. HDAC6 and microtubules are required for autoph-agic degradation of aggregated huntingtin. J. Biol. Chem., 280: 40282-40292, 2005.

Jung C.H., Ro S.H., Cao J., Otto N.M., Kim D.H. mTOR regulation of autophagy. FEBS Lett., 584: 1287-1295, 2010.

Kabuta T., Suzuki Y., Wada K. Degradation of amyotrophic lateral sclerosis-linked mutant Cu, Zn-superoxide dismutase proteins by macroau-

tophagy and the proteasome. J. Biol. Chem., 281: 30524-30533, 2006.

Kihara A., Noda T., Ishihara N., Ohsumi Y. Two distinct Vps34 phosphatidylinositol 3-kinase com-plexes function in autophagy and carboxypepti-dase Y sorting in Saccharomyces cerevisiae. J. Cell Biol., 152: 519-530, 2001.

Levine B. and Klionsky D.J. Development by self-digestion: Molecular mechanisms and biological functions of autophagy. Dev. Cell, 6: 463-477, 2004.

Lleo A., Invernizzi P., Selmi C., Coppel R.L., Alpini G., Podda M., Mackay I.R., Gershwin M.E. Autophagy: highlighting a novel player in the autoimmunity scenario. J. Autoimmun., 29: 61-68, 2007.

McGeady T.A., Quinn P.J., FitzPatrick E.S., Ryan M.T. Veterinary Embryology. Oxford. Blackwell Publishing, 2006.

Mizushima N., Levine B., Cuervo A.M., Klionsky D.J. Autophagy fights disease through cellular selfdigestion. Nature, 451: 1069-1075, 2008.

Mura E., Suman M., Montelli S., Peruffo A., Cozzi B., Farina V. Characterization of an established endothelial cell line from primary cultures of fetal sheep hypothalamus. Res. Vet. Sci., 94: 388-393, 2012.

Nedergaard M. and Dirnagl U. Role of glial cells in cerebral ischemia. Glia, 50: 281-286, 2005.

Nixon R.A. The role of autophagy in neurodegenera-tive disease. Nat. Med., 19: 983-997, 2013.

Polson H.E., de Lartigue J., Rigden D.J., Reedijk M., Urbé S., Clague M.J., Tooze S.A. Mammalian Atg18 (WIPI2) localizes to omegasome-anchored phagophores and positively regulates LC3 lipida-tion. Autophagy, 6: 506-522, 2010.

Qin A.P., Liu C.F., Qin Y.Y., Hong L.Z., Xu M., Yang L., Liu J., Qin Z.H., Zhang H.L. Autophagy was activated in injured astrocytes and mildly decreased cell survival following glucose and oxygen deprivation and focal cerebral ischemia. Autophagy, 6: 738-753, 2010.

Ravikumar B., Sarkar S., Davies J.E., Futter M., Garcia-Arencibia M., Green-Thompson Z.W., Jimenez-Sanchez M., Korolchuk V.I., Lichtenberg M., Luo S., Massey D.C., Menzies F.M., Moreau K., Narayanan U., Renna M., Siddiqi F.H., Underwood B.R., Winslow A.R., Rubinsztein D.C. Regulation of mammalian autophagy in physiology and pathophysiology. Physiol Rev., 90: 1383-1435, 2010.

56 E. MURA ET AL.

Richard O., Duittoz A.H., Hevor T.K. Early, middle, and late stages of neural cells from ovine embryo in primary cultures. Neurosci. Res., 31: 61-68, 1998.

Riley B.E., Kaiser S.E., Shaler T.A., Ng A.C., Hara T., Hipp M.S, Lage K., Xavier R.J., Ryu K.Y., Taguchi K., Yamamoto M., Tanaka K., Mizushima N., Komatsu M., Kopito R.R. Ubiquitin accumu-lation in autophagy-deficient mice is dependent on the Nrf2-mediated stress response pathway: a potential role for protein aggregation in autopha-gic substrate selection. J. Cell Biol., 191: 537-552, 2010.

Schmuck G, Röhrdanz E, Tran-Thi Q.H, Kahl R, Schlüter G. Oxidative stress in rat cortical neu-rons and astrocytes induced by paraquat in vitro. Neurotox. Res., 4: 1-13, 2002.

Shi R., Weng J., Zhao L., Li X.M., Gao T.M., Kong J. Excessive autophagy contributes to neuron death in cerebral ischemia. CNS Neurosci. Ther., 18: 250-260, 2012.

Sridharan S., Jain K., Basu A. Regulation of autopha-gy by kinases. Cancers, 3: 2630-2654, 2011.

Stanfel M.N., Shamieh L.S., Kaeberlein M., Kennedy B.K. The TOR pathway comes of age. Biochim. Biophys. Acta, 1790: 1067-1074, 2009.

Strawbridge A.B. and Blum J.S. Autophagy in MHC class II antigen processing. Curr. Opin. Immunol., 19: 87-92, 2007.

Tsvetkov A.S., Miller J., Arrasate M., Wong J.S., Pleiss M.A., Finkbeiner S.A. Small-molecule scaf-fold induces autophagy in primary neurons and protects against toxicity in a Huntington disease model. Proc. Natl. Acad. Sci. USA., 107: 16982-16987, 2010.

Yue Z., Friedman L., Komatsu M., Tanaka K. The cellular pathways of neuronal autophagy and their implication in neurodegenerative diseases. Biochim. Biophys. Acta, 1793: 1496-1507, 2009.

Zeng Z., Sarbassov d.D., Samudio I.J., Yee K.W., Munsell M.F., Ellen Jackson C., Giles F.J., Sabatini D.M., Andreeff M., Konopleva M. Rapamycin derivatives reduce mTORC2 signaling and inhibit AKT activation in AML. Blood, 109: 3509-3512, 2007.

Documents

ARCHIVES ITALIENNES DE BIOLOGIE