J Cutan Pathol 2009: 36: 296–301doi: 10.1111/j.1600-0560.2008.01024.xBlackwell Munksgaard. Printed in Singapore

Copyright # Blackwell Munksgaard 2008

Journal of

Cutaneous Pathology

Dermal sheet preparations in theevaluation of dermal innervation inParkinson’s disease and multiplesystem atrophy

Background: Evaluation of dermal nerve fibers in conventionalvertical sections is difficult because of the small number of fibersavailable for examination. In this study, we evaluated dermal sheetmounts for fibers in which the majority of fibers can be visualized.Methods: We compared the dermal small fiber density in sixParkinson’s disease (PD) and six multiple system atrophy (MSA)patients using dermal sheet preparations (DSP). DSP are based onepidermal-dermal separations and immunostaining of the entiredermis by the nerve growth factor receptor p75 antibody that stainsboth autonomic and sensory fibers.Results: The small fiber density was reduced in PD compared withMSA (p , 0.0001), suggesting the presence of small fiber neuropathyin PD.Conclusions: DSP offer a unique method of evaluation of dermalnerve fibers. This method can be used to evaluate small nerve fibers inmany neurological disorders such as MSA and PD.

Novak P, Marya NB, Whren K, Bhawan J. Dermal sheet preparationsin the evaluation of dermal innervation in Parkinson’s disease andmultiple system atrophy.J Cutan Pathol 2009; 36: 296–301. # Blackwell Munksgaard 2008.

Peter Novak1, Neil B. Marya2,Kara Whren2 and Jag Bhawan2

1Autonomic Laboratory, Department ofNeurology and2Dermatopathology Section, Department ofDermatology, Boston University School ofMedicine, Boston, MA, USA

Jag Bhawan, MD, Dermatopathology Section,Department of Dermatology, Boston UniversitySchool of Medicine, 609 Albany Street, J-309,Boston, MA 02118, USATel: (617) 638 5570Fax: (617) 638 5575e-mail: jbhawan@bu.edu

Accepted for publication February 26, 2008

Parkinsonism and autonomic impairment includingorthostatic hypotension, sphincter dysfunction, con-stipation and alteration in sweating is commonlyobserved in both Parkinson’s disease (PD) andmultiple system atrophy (MSA).1–3 Although thedysautonomia tends to be more severe in MSA, thedifferentiation of PD complicated by autonomicdysfunction from MSA can be a clinical challenge.3

Up to 10% of the patients in the PD brain bank in facthad MSA in one study.4

The lesion underlying the dysautonomia differs inMSA compared with PD. MSA is associated with

degeneration of the brainstem autonomic centers andintermediolateral nucleus of the spine.1,5 Dysautono-mia in PD is considered because of both central andperipheral origin as Lewy bodies are found in thehypothalamus, brainstem, intermediolateral nucleusof the spine and sympathetic ganglia.6 Functionalimaging studies showed cardiac sympathetic dener-vation in PD but not in MSA.7 Nevertheless, formaltesting of autonomic functions failed to distinguishMSA from PD based on pattern or severity ofdysautonomia.3 Evidence of functional postgangli-onic cholinergic impairment assessed by axon reflextesting was found in both patient groups.Reliable methods that would aid in the differenti-

ation of PD from MSA are critically needed in theCurrent address: Peter Novak, Department of Neurology, Univer-

sity of Massachusetts Medical School, Worcester, MA, USA.

296

clinic and research. Skin biopsies enable directvisualization of the dermal innervations that includeboth autonomic and sensory fibers. A recent semi-quantitative study found reduced dermal sympatheticnerve density in PD compared with controls thatsuggests the presence of peripheral autonomicneuropathy in PD.8 In this report, we present a uniquemethod of evaluation of the dermal nerve density byusing the dermal sheet preparations (DSP). DSP is theentire dermis adjusted to 1 mm in depth where manydermal nerve fibers can be visualized, evaluated andquantified (Fig. 1). This is in contrast to the con-ventional vertical section (Fig. 2) where only a fewfibers can be seen. Our DSP are based on a methoddescribed earlier by Tschachler et al.9 We havemodified the DSP by digesting collagen by collage-nase (see Methods) and then immunostaining withNGFr p75. The collagenase digestion allows pene-tration of the antibody deep into the dermis. TheNGFr p75 predominantly stains Schwann andperineuronal cells, supporting both sensory andautonomic fibers throughout the dermis.10,11

We processed skin biopsies in patients withMSA orPDcomplicated by severe autonomic dysfunction.Wetested the hypothesis that the dermal innervation isreduced in PD compared with MSA.

Methods

Subjects

Skin biopsies from six patients with PD and sixpatients with MSA were analyzed (Table 1). Allpatients were seen for evaluation of autonomicdysfunction in an Autonomic Laboratory at BostonMedical Center (Boston, MA, USA). PD patients(Table 1) had at least two of the three cardinal signs ofthe disease12 including resting tremor, bradykinesiaand rigidity. All PD patients also had orthostatic

hypotension. MSA patients had a combination ofautonomic, cerebellar and parkinsonian syndromeconsistent with a diagnosis of probable MSA accord-ing to the Consensus Criteria.1 All MSA subjects hada history of autonomic failure, urinary dysfunction orerectile dysfunction as defined in Gilman et al.1 Noneof the patients had a history of polyneuropathy or skinabnormalities on their legs. The Institutional ReviewBoard of the Boston University Medical Centerapproved the study.

Processing of skin biopsies

Skin biopsies from each patient were taken from twosites, 10 cm above the lateral knee (proximal site) and10 cm above the lateral malleolus (distal site). Afterthe skin was cleaned with isopropyl alcohol, the full-dermis thickness biopsy was performed with a 3-mmdiameter punch under local anesthesia using 1%lidocaine. Biopsies were placed intoMichel’s media to

Fig. 1. Image analysis. The digital photomicrograph of immunostained dermal sheet (A) was subjected to the background homogenization and to

the phase-symmetry processing (B) that enhances the nerve fibers. The nerve fibers were then detected using the threshold technique (320).

Fig. 2. Vertical section of skin stained with NGFr showing few nerve

fibers in the papillary dermis (320).

DSP in evaluating dermal innervation in PD and MSA

297

provide preservation and stored at 4�C until furtherprocessing.

Dermal sheet preparations

The dermis was separated from the epidermis byfloating the biopsy sample in 1 M NaCl solution for72 h followed by gently pulling the epidermis byforceps. This is similar to the salt-split skin techniqueused for immunofluorescence in the evaluation ofimmunobullous disorders. However, here we usedermis and discard the epidermis. Then, the depth ofthe dermal sheet was adjusted to about 1 mm. Thedermal sheetswere placed for 6 h at room temperaturein collagenase solution (1.9 mg Sigma Aldrich, St.Louis, MO, C0130) dissolved in 1 ml TESCA buffer[0.01 g of N-Tris(hydroxymethyl)methyl-2-aminoe-thanesulfonic acid in 3.6 ml of 0.1 M CaCl2, contain-ing additional 50 ml of 0.1 M CaCl2]. TESCA buffersolution was brought up to 1 ml using distilled water,and the entire collagenase solution was adjusted to pH7.4 using 1 M NaOH. After incubation with collage-nase, the dermal sheets were washed with PBS for5 min followed by fixation in acetone for 20 min.

The dermal sheets were then placed in peroxidaseblock (K4005; Dakocytomation, Carpinteria, CA,USA) for 10 min followed by two washes in tris-buffered saline (TBS) of 5 min each.The dermal sheetswere incubated overnight at 4�C inmonoclonalmouseanti NGFr (M3507; Dakocytomation, Carpinteria,CA, USA) at a dilution of 1 : 100. Dermal sheets werewashed in TBS twice for 10 min each followed byincubation in horseradish peroxidase-labeled second-ary mouse antibody (K4005; Dakocytomation) for 2 hat room temperature. The dermal sheets were thenwashed in TBS twice for 10 min each followed byincubation in 3-amino, 9-ethyl-carbazole substratechromogen (K4005; Dakocytomation) for 10 min.Finally, the sheets were washed in TBS for 10 min

and mounted on glass slides using aqueous mountingmedia for viewing through a light microscope.

Image processing

The immunostained samples were photographed intodigital form by a light microscope (Olympus AH-2;Olympus JapanCo., Ltd,Tokyo, Japan) using a digitalcamera (Olympus DP-70; Olympus Japan Co., Ltd,Tokyo, Japan) with resolution of 11.65 pixels per mm.The image processing software was written inMatlab� 6.1 (The MathWorks, Inc., Natick, MA,USA). The color images were converted to gray-scaleimages (Fig. 1B). Then, the background of the imageswas homogenized with the morphological processingwhereby images were divided into circles withdiameters of 15 pixels and the average densities overthat area were equalized. The resulting homogenizedimages were normalized within the range of 0–1.Then, the image noise was reduced by a two-dimensional median filter having a length of 5 pixels.The detection of fibers was performed using thephase-symmetry method (PSM), which is a robustmethod for detection of linear structures such aslines.13,14 The detected lines served as a template fordetection of nerve fibers in the original images. In ourapplication, the PSM considers all linear structuresthat have symmetrical phase as being nerve fibers.The final detection was achieved by using a thresholdtechnique that was in the range of 0.1–0.3. The fiberdensity was expressed in percent as ratio of area ofdetected fibers to the whole image.

Statistical analysis

The overall difference of fiber density betweendiagnoses (PD vs. MSA) and biopsy sites (distal vs.proximal) was analyzed by 2 3 2 (diagnoses 3 sites)

Table 1. Clinical characteristics of patients�

No. Age (years) Gender DiseaseDuration ofdisease (years) H&Y Clinical characteristics Medication

1 58 M PD 2 2.5 Pasym (L . R), OH and syncopal episodes Rop, Pyr, C/L and Pro2 69 M PD 2 2.5 Pasym (L . R), dizziness and syncope C/L and Pra3 80 M PD 5 2.5 Pasym (L . R), OH and balance problems C/L, Pra and Pro4 72 M PD 15 3 Pasym (L . R), OH, gait freezing, MF and D C/L5 71 M PD 10 2.5 Pasym (L . R), OH, dysphonia, MF and D C/L, Ama and Rop6 73 M PD 2 2.5 Pasym (R . L) and Esw C/L7 70 M MSA 9 – Pasym (R . L), OH and incontinence C/L, Pyr, Pro, Ama and Flu8 80 M MSA 7 – Pasym (R . L), OH and incontinence C/L, Flu and Pro9 59 M MSA 6 – Pasym (L . R) and OH C/L, Flu, Pro and Pyr10 76 M MSA 2 – Pasym (L . R), OH and respiratory failure C/L and Pyr11 66 M MSA 2 – Pasym (R . L), OH and depression C/L, Pro and Flu12 68 M MSA 5 – Pasym (L . R), OH and depression C/L, Pro, Pyr, Flu and Yoh

Ama, amantadine; C/L, carbidopa/levodopa; D, dyskinesia; Esw, excessive sweating; Flu, fludrocortisone; H&Y, Hoehn and Yahr stage; L, left side; MF,motor fluctuations; MSA, multiple system atrophy; PD, Parkinson’s disease; Pasym, symmetric parkinsonism; Pra, pramipexole; Pro, ProAmatine; Pyr,pyridostigmine; R, right side; Rop, ropinirole; Yoh, yohimbine.�Clinical characteristics show predominant symptoms.

Novak et al.

298

MANOVA (multivariable analysis of variance). Ageand duration of disease were included as covariates.JMP 5.1 (SAS, Cary, NC, USA) software was used forstatistical analyses.

Results

Figure 3 shows a typical example of dermal stainingin PD andMSA patients. A rich subepidermal plexuswith numerous fibers and Schwann cells is seen in theMSA patient (Fig. 3A). The morphology of fibersvaried;many fibers were linear, while others exhibitedfragmentation, swelling, thinning and beading. Thefiber density and number of Schwann cells arereduced in the PD patient (Fig. 3B). In the PDpatient, even visible fibers stain with less immunore-activity (IR) than that in the MSA patient, and thestaining is more fragmented. Nerve fiber densities foreach subject are presented in Table 2.MANOVA showed higher average fiber density in

the MSA group (expressed as means and respectivestandard deviations, proximal site 9.16 � 1.28%and distal site 8.55 � 2.01%) than in the PD group(proximal site 4.76 � 0.82% and distal site 3.94 �1.48%, p , 0.0001). There were significant differ-ences in both proximal sites (p , 0.0001) and distalsites (p , 0.0007) between the MSA group andthe PD group. Comparing the fiber density withindiagnoses, there was no significant difference betweenthe proximal and distal sites both in the PD group(p , 0.26) and in the MSA group (p , 0.51). Theeffects of age and duration of disease were notsignificant.

Discussion

The main finding of our study is that the small fiberdensity as reflected by the p75 IR was significantlylower in the PD than in the MSA patients. The p75

stains predominantly Schwann cells, perineuriumandin less extend axons.11 Then, reduced p75 IR in PDgroup predominantly reflects loss of Schwann cells.The p75 expression through Schwann cells plays anactive role in the maintenance of a normal innerva-tion. The regulation and function of p75 is complexand incompletely understood. Increased expression ofp75 was described in regenerating motor nerves,15,16

experimental neuropathies17 and glial cells,15 whilep75 can be downregulated in injured sensoryneurons.15 Human spinal cord lesion, e.g. centrallesion, results in marked decrease of the p75 IR inboth neural and non-neural tissue.18 In our study, weused p75 predominantly as a neuronal marker. Aprevious study using simultaneous costaining of p75with axonal marker protein gene product (PGP) 9.5showed no significant difference in nerve fiber densitybetween both staining in humans,19 suggesting thatp75 is a reasonable pan-neuronal marker. A distinctadvantage of p75 over PGP 9.5 is higher contrast ofstained fibers.19 However, because p75 stains pre-dominantly Schwann cells, the loss of p75 IR is

Fig. 3. Dermal nerves in 82-year-old MSA (A) and 73-year-old PD (B) patients. Both skin samples were taken from calf. Note reduced density

of nerve fibers in the PD patient. Arrows point to Schwann cells (320). MSA, multiple system atrophy; PD, Parkinson’s disease.

Table 2. Results�

No.

Small fiber density (%)

Proximal Distal

1 5.54 3.742 5.10 4.223 4.7 2.254 5.77 6.185 4.79 3.656 3.55 2.697 7.8 6.338 8.14 7.239 9.16 8.6410 9.56 7.5111 8.64 7.5312 11.72 11.49

�The density of small fibers is expressed in percent for proximal anddistal sites. The subject number corresponds to the same subject as inTable 1.

DSP in evaluating dermal innervation in PD and MSA

299

because of either axonal neuropathy with thesecondary Schwann cell degeneration or primarydegeneration of Schwann cells. It can be argued that ifPD is a primary peripheral Schwann cell disorder,then it can also be expected to have Schwann celldegeneration in large fibers. This would translate inlarge fiber neuropathies that are not observed inPD.20

This line of reasoning suggests that the Schwann cellloss observed in our PD group is secondary to axonaldamage and not vice versa. Nevertheless, onlysimultaneous staining of p75 with an axonal markercan clarify the status of underlying axons.

Whether degeneration in PD preferentially affectsautonomic or sensory fibers cannot be assessed inthis study because NGFr p75 stains both types ofnerves. Several lines of evidence point to a selectiveloss of autonomic fibers with sparing of sensory fibersin PD. First, although many patients with PDcomplain of subjective sensory symptoms includingnumbness and burning limb sensation,20 the objec-tive electrophysiological and imaging evidencepoints to a deficit in sensory processing at the basalganglia level21 and large fiber sensory neuropathy isnot a part of PD as evidenced by normal nerveconduction studies.20 Second, in a recent semi-quantitative study, Dabby et al. found decreasedautonomic innervations of the blood vessels, sweatglands and erector pili muscles in PD patients.8

Nevertheless, it would be advantageous to modifythe DSP technique with a specific staining forautonomic fibers, for example to visualize theadrenergic fibers using tyrosine hydroxylase stain-ing, to provide direct quantitative evidence ofpostganglionic autonomic degeneration in PD.

Our study suggests that skin biopsy using DSP canbe helpful in differentiation between the PDandMSApatients, the task that at present constitutes a consider-able clinical challenge because of overlap of dysauto-nomia and parkinsonism in both disorders.3However,we cannot rule out possibility that the results arebecause of inadequate sample size or patient selectionbias. As such, the results need to be reproduced ina larger study.

Conclusions

Our study shows that dermal skin biopsy can behelpful in the evaluation of small dermal nerve fibersby using DSP. This method provides an opportunityto study dermal nerve fibers not only in PD andMSA patients but also in other neurological disordersas well. Furthermore, additional immunostainingwith different markers may allow further delineationbetween autonomic and nonautonomic nerve fibers.This method has a clear advantage over the con-ventional vertical cut-stained sections.

Acknowledgment

The authors thank P. Kovesi for providing software to calculate the

phase symmetry.

References

1. Gilman S, Low PA, Quinn N, et al. Consensus statement on the

diagnosis of multiple system atrophy. J Neurol Sci 1999; 163: 94.

2. Kaufmann H. Multiple system atrophy. Curr Opin Neurol

1998; 11: 351.

3. Riley DE, Chelimsky TC. Autonomic nervous system testing

may not distinguish multiple system atrophy from Parkinson’s

disease. J Neurol Neurosurg Psychiatry 2003; 74: 56.

4. Colosimo C, Albanese A, Hughes AJ, de Bruin VMS, Lees AJ.

Some specific clinical features differentiate multiple system

atrophy (striatonigral variety) from Parkinson’s disease. Arch

Neurol 1995; 52: 294.

5. Daniel SE. The neuropathology and neurochemistry of

multiple system atrophy. In Bannister R, Mathias CJ, eds.

Autonomic failure. A textbook of clinical disorders of the

autonomic nervous system. Oxford: Oxford University Press,

1999; 321.

6. Wakabayashi K, Takahashi H. Neuropathology of autonomic

nervous system in Parkinson’s disease. 1997; 38 (Suppl. 2): 2.

7. Goldstein DS. Functional neuroimaging of sympathetic inner-

vation of the heart. Ann N Y Acad Sci 2004; 1018: 231.

8. Dabby R, Djadetti R, Shahmurov M, et al. Skin biopsy for

assessment of autonomic denervation in Parkinson’s disease.

J Neural Transm 2006; 113: 1169.

9. Tschachler E, Reinisch CM, Mayer C, Paiha K, Lassmann H,

Weninger W. Sheet preparations expose the dermal nerve

plexus of human skin and render the dermal nerve end

organ accessible to extensive analysis. J Invest Dermatol 2004;

122: 177.

10. Bothwell M. p75(NTR): a receptor after all. Science 1996;

272: 506.

11. Liang Y, Johansson O. Light and electron microscopic

demonstration of the p75 nerve growth factor receptor in

normal human cutaneous nerve fibers: new vistas. J Invest

Dermatol 1998; 111: 114.

12. Hughes AJ, Daniel SE, Kilford L, Lees AJ. Accuracy of clinical

diagnosis of idiopathic Parkinson’s disease: a clinico-pathological

study of 100 cases. J Neurol Neurosurg Psychiatry 1992; 55: 181.

13. Kovesi PD. MATLAB and octave functions for computer vision

and image processing. School of Computer Science & Software

Engineering. Perth, Australia: The University of Western

Australia, 2000, available at http://www.csse.uwa.edu.au/

�pk/research/matlabfns (accessed 12 November, 2006).

14. Kovesi PDi. Symmetry and asymmetry from local phase. In

the 10th Australian Joint Conference on Artificial Intelligence,

poster proceedings, pages 185–190, Perth, 1997.

15. Zhou XF, Rush RA, McLachlan M. Differential expression of

the p75 nerve growth factor receptor in glia and neurons of

the rat dorsal root ganglia after peripheral nerve transection.

J Neurosci 1996; 16: 2901.

16. Anton ES, Weskamp G, Reichardt LF, Matthew WD. Nerve

growth factor and its low-affinity receptor promote Schwann

cell migration. Proc Natl Acad Sci U S A 1994; 91: 2795.

17. Conti G, Baron PL, Scarpini E, et al. Low-affinity nerve growth

factor receptor expression in sciatic nerve during P2-peptide

induced experimental allergic neuritis. Neurosci Lett 1995;

199: 135.

Novak et al.

300

18. Lopez SM, Perez-Perez M, Marquez JM, Naves FJ,

Represa J, Vega JA. p75 and TrkA neurotrophin receptors

in human skin after spinal cord and peripheral nerve injury,

with special reference to sensory corpuscles. Anat Rec 1998;

251: 371.

19. Antunes SL, Liang Y, Neri JA, Haak-Frendscho M, Johansson

O. The expression of NGFr and PGP 9.5 in leprosy reactional

cutaneous lesions: an assessment of the nerve fiber status using

immunostaining. Arq Neuropsiquiatr 2003; 61: 346.

20. Koller WC. Sensory symptoms in Parkinson’s disease. Neurol-

ogy 1984; 34: 957.

21. Boecker H, Ceballos-Baumann A, Bartenstein P, et al. Sensory

processing in Parkinson’s and Huntington’s disease: investiga-

tions with 3D H(2)(15)O-PET. Brain 1999; 122: 1651.

DSP in evaluating dermal innervation in PD and MSA

301