J Formos Med Assoc 2002 • Vol 101 • No 3

C.C. Chu, C.C. Huang, H.C. Kuo, et al

214

(J Formos Med Assoc2002;101:214–8)

Key words:X-linked recessive bulbospinalneuronopathyKennedy-Alter-Sung diseasesensory neuropathymorphometric analysis

Department of Neurology, Chang Gung Memorial Hospital and Medical College, Taipei, and 2Department of Neurologyand Neuroscience, Changhua Christian Hospital, Changhua.Received: 2 October 2001. Revised: 30 October 2001. Accepted: 8 January 2002.Reprint requests and correspondence to: Dr. Chin-Chang Huang, Department of Neurology, Chang Gung Memorial Hospital,199 Tung Hwa North Road, Taipei, Taiwan.

SENSORY NEUROPATHY IN X-LINKED

RECESSIVE BULBOSPINAL NEURONOPATHY

Chun-Che Chu, Chin-Chang Huang, Hung-Chou Kuo, Chin-San Liu,1 andChing-Shan Tsai1

X-linked recessive bulbospinal neuronopathy (X-BSN),also known as Kennedy-Alter-Sung disease, is an adult-onset spinal and bulbar amyotrophy characterized byslowly progressive muscle weakness and atrophy, fas-ciculations in bulbar and proximal limb muscles, andhyporeflexia or areflexia [1–4]. Endocrine featuressuch as gynecomastia, testicular atrophy, abnormallipid metabolism, and essential tremor may occasion-ally occur [5–7]. Neurophysiologic studies often dem-onstrate subclinical involvement of sensory nerves withdiminished or absent sensory nerve action potentials(SNAPs) and denervation changes, suggesting theinvolvement of sensory neurons in addition to bulbarand spinal motor neurons. The results of sural nervebiopsy are consistent with axonal atrophy anddegeneration, with secondary demyelination [8–10].An increased number of CAG repeats within the firstexon of the androgen receptor (AR) gene have been

Abstract: X-linked recessive bulbospinal neuronopathy (X-BSN) is an adult-onsetspinal and bulbar amyotrophy. Neurophysiologic studies demonstrate subclinicalinvolvement of sensory nerves with diminished or absent sensory nerve actionpotientials and denervation changes, indicating the involvement of sensoryneurons. We report the clinical features, findings of electrophysiologic study, andresults of morphometric analysis of sural nerve pathology in a patient with X-BSN.Molecular genetic studies were also performed in the patient and his threedaughters. Electrophysiologic studies revealed decreased amplitude sensory nerveaction potentials and the presence of high amplitude motor unit potentials in allmuscles tested. Sural nerve biopsy demonstrated axonal degeneration with apredominant loss of large myelinated fibers. Molecular genetic studies confirmedelongation of the CAG triplet repeats in exon 1 of the androgen receptor gene.Sequence analysis of the androgen receptor gene revealed that the number of CAGtriplet repeats was 45 in the patient and was 45 to 48 in the mutant allele but only19 to 30 in the normal allele in his three daughters. These findings suggest thatboth motor and sensory neurons are involved in X-BSN. Sural nerve biopsy andmolecular genetic analysis are helpful in differentiation between X-BSN and othermotor neuron diseases.

identified as the putative cause of X-BSN [11].Therefore, sural nerve biopsy and molecular geneticanalysis have become important in distinguishing X-BSN from other forms of motor neuron diseases, andeven in early detection of heterozygous female carriers[12, 13]. This report describes morphologic changes inthe sural nerve and the results of molecular geneticstudies in a family with X-BSN and correlates thesefindings with clinical manifestations and the results ofelectrophysiologic study.

Case Report

Clinical and family historyA 50-year-old man developed progressive limb weakness inall limbs over a 5-year period. He found that he could not lift

J Formos Med Assoc 2002 • Vol 101 • No 3 215

Sural Nerve Pathology in Bulbospinal Neuronopathy

dumbbells and dance as usual beginning 5 years prior to thishospital visit in 1998. He had become thinner in the last 3years and noted progressive body weight loss, generalizedweakness, and muscle wasting, especially in the proximallimbs. He had difficulty in raising his arms and rising fromsquatting. Neurologic examinations disclosed mild muscleweakness and wasting in all limbs with full muscle strength indistal limbs and grade 4/5 in proximal limbs according to theScale of the Medical Research Council of Great Britain.Tendon reflexes were all absent. Sensory modalities includ-ing pin-pricks, temperature, touch, vibration, and positionsensations were intact. Neither Romberg’s sign nor sensoryataxia was noted. Fasciculations were noted in the perioralarea, tongue, and extremities as well as mild postural handtremor. There was neither obvious muscle cramping onexercise, dysarthria, nor dysphagia. Gynecomastia and tes-ticular atrophy were also noted. Review of family historydisclosed similar symptoms in his uncle, but no symptomswere found in his three daughters.

Biochemical dataBiochemical studies including serum concentrations ofcholesterol, triglyceride, glucose, and glycohemoglobin, pro-tein electrophoresis, and lipoprotein electrophoresis were allwithin normal ranges. Serum creatinine phosphokinase waselevated to 845 U/L (normal, 15–130 U/L) with 98.4% MMform and 1.6% MB form. Serum concentrations of prolactin,luteinizing hormone, follicular stimulating hormone, 17-estradiol, progesterone, testosterone, cortisol, triiodothy-ronine, tetraiodothyronine, thyroid stimulating hormone, andgrowth hormone were normal.

Electro-neurophysiologic findingsNerve conduction studies demonstrated normal distallatencies, amplitudes of compound motor action potentials,and nerve conduction velocities in median, ulnar, peroneal,and tibial nerves, but decreased SNAP amplitudes in allsensory nerves tested [sural nerve, 7.5 µV (normal, > 10 µV);median nerve, 7.0 µV (normal, > 10 µV); ulnar nerve, 4.8 µV(normal, > 8 µV)]. Electromyographic studies showed in-creased amplitudes in motor unit potentials in the left flexordigitorum indicis, biceps, anterior tibialis, and vastus medialis.There were neither fibrillations nor positive waves. Soma-tosensory evoked potential study revealed absence of wavesin bilateral L2 responses and absence of P40 following righttibial nerve stimulation. These data indicated a peripheralsensory conduction defect in both legs and normal peripheraland central conduction from both hands.

Sural nerve biopsySural nerve biopsy specimens were obtained from the leftankle in July 1998. Approximately 1.5 cm of the nervefascicles was fixed in a 3% glutaraldehyde solution anddivided into two segments, one for epoxy sections and theother for teased fiber preparation. Semi-thin sections of 0.6µm were stained with toluidine blue and examined under alight microscope. The fiber sizes of each myelinated nervefiber in these fascicles were measured in a morphometricanalyzer (Leica Q500MC image processing and analysis

system; Median Cybernetics, Silver Spring, MD, USA). Ultra-thin sections were stained with uranyl acetate and lead citrateand studied using a Jeol JEM-200EXII electron microscope.About 100 nerve fibers were teased under a dissecting micro-scope and observed under a light microscope.

Sural nerve biopsy specimens consisted of five fasciculi.On light microscopic examination, transverse sections of thesural nerve revealed a loss of myelinated nerve fiber (MF),affecting predominantly the large MF population (Fig. 1).Some fibers had thin myelin suggesting remyelination. A fewclusters of regenerating fibers were scattered around MFs.Neither interstitial changes nor thickened perineurium wasseen.

Electron microscopic examination confirmed axonal de-generation with abnormal accumulation of neurofilamentsand microtubules (Fig. 2). Remyelination features character-ized by a relatively thin myelin sheath to the diameter of theaxons were observed. Disruption of the myelin sheath withcytoplasmic debris was also observed. Unmyelinated fibersappeared relatively normal.

Morphometric study of the sural nerve showed a moder-ate reduction in nerve fiber density (2,150–4,623/mm2; control,6,000–10,000/mm2). A histogram of MFs showed a unimodaldistribution with a left shift and a decreased amplitude of thepeak of large MFs (Fig. 3). Preferential involvement of largeMFs was also demonstrated by comparison of the density ofsmall MFs (diameter < 7 µm) with the density of large MFs(diameter > 7 µm). Only 21.5% of MFs in the patient werelarge, while 43% of MFs in a normal sample taken from a 45-year-old man were large MFs. Teased fiber examinationshowed few fibers with segmental demyelination orremyelination.

Molecular genetic analysisGenomic DNA was extracted from blood cells of the patientand his three daughters. Polymerase chain reaction (PCR) wasperformed as previously described [13] in a 100-µL reactionmixture containing 500 ng genomic DNA, 30 pmol of theforward primer 5'-TCCAGAATCTGTTCCAGAGCGTGC-3'

Fig. 1. Transverse semi-thin section of sural nerve biopsy showing amoderately reduced density of myelinated fibers and axonaldegeneration. (Toluidine blue stain, x 200 before reduction)

J Formos Med Assoc 2002 • Vol 101 • No 3

C.C. Chu, C.C. Huang, H.C. Kuo, et al

216

and the reverse primer 5'-GCTGTGAAGGTTGCTGTT-CCTCAT-3', 1.5 units of FastStart Taq DNA polymerase(Roche Diagnostics, Mannheim, Germany), 20 nmol of eachdeoxyribonucleotide triphosphate (dNTP), 2 µL of 10x PCRbuffer (pH 8.3), and 4 µL 5x GC-rich solution. The reactionmixture was subjected to PCR in a PTC-0200 DNA EnginePCR machine (MJ Research, Inc., Boston, MA, USA) consist-ing of denaturation at 94°C for 1 minute, annealing at 64°Cfor 30 seconds, and primer extension at 72°C for 2 minutes for36 cycles. Half of the amplified DNA from each sample wasanalyzed by electrophoresis at 100 v for 2.5 hours on a 4%SeaKem LE agarose gel with a 100-bp DNA ladder as a sizemarker. The remaining PCR product was purified by ethanolprecipitation to eliminate primers and dNTPs and was thenligated into a pGEM-t vector and the recombinant DNA wastransfected into E. coli DH 5α cells. White colonies wereisolated and nucleotide sequencing was then performed oneach of the positive clones using the SequiThermTM cycle

Fig. 2. Electron microscopic examination showing axonal degenera-tion with abnormal accumulation of neurofilaments and neurotubules.(x 8,000 before reduction)

sequencing kit (Epicentre Technologies Co., Madison, WI,USA) with the sequencing primer 5'-TCCAGAATCTGTTCC-AGAGCGTGC-3'. More than three clones were sequenced toobserve the genetic instability of exon 1 of the AR gene foreach of the individuals examined.

PCR-amplified DNA fragments encompassing the CAGtriplet repeat in the first exon of the AR gene were obtainedfrom the proband and other members of his family (Fig. 4). Anelongated 360-bp PCR product was amplified from exon 1 ofthe AR gene of the X-BSN patient and three carriers, but a 290-bp DNA fragment was obtained from the normal allele in ahealthy unrelated male as a normal control. The number ofCAG repeats in exon 1 of the AR gene determined by DNAsequencing was 45 in the patient with X-BSN. His threedaughters had 45/29, 45/30 and 48/19 CAG triplet repeats inthe two alleles in exon 1 of the AR gene. However, in theunrelated healthy male control, only 22 CAG triplet repeatswere found in exon 1 of the AR gene (Fig. 4).

Discussion

This study describes the clinical manifestations, resultsof electrophysiologic study, and pathologic changes inthe sural nerve in a patient with X-BSN confirmed bymolecular analysis. Although he did not have sensorysymptoms, sensory neuropathy was documented from abnormal nerve conduction velocity studies withdecreased amplitude SNAPs in sensory nerves.Furthermore, sural nerve biopsy revealed axonal de-generation with loss of large MFs. These data indicatethat subclinical involvement of sensory nerves canoccur in patients with X-BSN.

X-BSN may be confused with progressive bulbarpalsy variants of motor neuron disease. Many of thesecases at first appear to be sporadic, with proximal

Fig. 3. Histogram ofmyelinated nerve fi-ber diameter : A)unimodal distribu-tion with a decreasedamplitude of the peakin large myelinatedfibers. B) Normal con-trol showing a bimo-dal distribution inmyelinated fibers.

No.

of

mye

linat

ed f

iber

s/m

m2

1,000

800

600

400

200

01 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Diameter of myelinated fibers (µm)

1,000

800

600

400

200

01 2 3 4 5 6 7 8 9 10 11 12 13 14 15

A B

J Formos Med Assoc 2002 • Vol 101 • No 3 217

Sural Nerve Pathology in Bulbospinal Neuronopathy

muscles affected first, followed by bulbar muscles [1–4]. Dysarthria and dysphagia are associated with atro-phy and weakness of facial, jaw, and tongue muscles.However, contraction fasciculations of the perioralmuscles, gynecomastia, and other endocrine abnor-malities are characteristic [6, 7, 11]. In addition,hyporeflexia or areflexia and essential tremor mayoccur. Laboratory test results are usually normal ex-cept for a modest elevation in creatine kinase [9, 14].A high level of creatine kinase (mean, 1078 ± 873 U/ L;range, 274_3457 U/L) in 26 of 28 patients with X-BSNhas been described [9]. Similar results may also befound in spinal muscular atrophies or other motorneuron diseases [15]. Correct diagnosis of this disor-der is important for both prognosis and genetic coun-seling [12, 13, 16, 17].

Our previous reports examined clinical, bio-chemical, and molecular studies of X-BSN patients [13,

16, 17]. However, electrophysiologic studies and nervepathology data have not been emphasized. Kennedy etal found abnormal SNAPs in three cases and mild lossof nerve fibers in the peripheral nerves at autopsy inone [1]. In 1982, Harding et al found most patients hadsmall or unrecordable SNAPs with absence of clinicalsensory impairments [2]. A detailed clinicopathologicstudy with sural nerve biopsy and morphologic obser-vations found that lower motor neurons were markedlydepleted through all spinal segments and brainstemmotor nuclei [4]. Primary sensory neurons were lessseverely affected. A quantitative study of primary sen-sory axons suggested that a distally accentuatedaxonopathy was the salient pathologic process [4].Sensory nerve involvement was found not only in X-BSN patients but also in carriers in a large family andsural nerve biopsies confirmed a decrease in MFs andthe presence of atrophic axons with demyelinating andremyelinating changes [8]. Some carriers may presentsigns of chronic dennervation and low-amplitude SNAPs[9]. Our study showed similar findings of sensory nerveinvolvement when compared with previous studies [4,5, 8, 9]. Therefore, the identification of sensory neur-opathy by electrophysiologic study and sural nervepathology is helpful in differentiation of this diseasefrom other motor neuron diseases. Finally, X-BSNshould be diagnosed by molecular genetic analysis.

The pathogenesis of sensory nerve involvement inX-BSN is still unclear. The highly selective involvement offibers in the fasciculus gracilis and predominant lower leginvolvement in clinical sensory nerve conduction velocitystudies suggest that the involvement in primary sensoryneurons is caudally dominant in terms of rostrocaudalsegmental distributions [4]. Recently, it has been sug-gested that this disease may share nerve degenerationmechanisms with X-linked Charcot-Marie-Tooth (CMTX)disease [8, 12], as the CMTX locus seems to be close to theAR gene [18–20]. However, the clinical and neuropatho-logic features of peripheral nerve involvement are quitedifferent in these two diseases. Further investigation offunction of the AR gene is warranted.

References

1. Kennedy WR, Alter M, Sung JH: Progressive proximalspinal and bulbar muscular atrophy of late onset: a sex-linked recessive trait. Neurology 1968;18:671–80.

2. Harding AE, Thomas PK, Baraitser M, et al: X-linkedrecessive bulbospinal neuronopathy: a report of tencases. J Neurol Neurosurg Psychiatry 1982;45:1012–9.

3. Amato AA, Prior TW, Barohn RJ, et al: Kennedy’s disease:a clinicopathologic correlation with mutations in theandrogen receptor gene. Neurology 1993;43:791–4.

Fig. 4. Pedigree of the family with X-linked bulbospinal neuronopathy(X-BSN) and molecular analysis of CAG repeat expansion in exon 1of the androgen receptor gene in the patient and some family members.Five sets of DNA fragments encompassing the CAG repeat region inexon 1 of the androgen receptor gene were amplified using thepolymerase chain reaction (PCR) and revealed an elongated 360-bpPCR product in one X-BSN patient ( ; I-1) and three carriers ( ;II-1, II-2 and II-3), while a 290-bp product was identified in a normalcontrol. = Patient; = carrier; M = 100-bp DNA ladder; N = normalmale control.

I

II1 2 3

M N I-1 II-1 II-2 II-3

1

J Formos Med Assoc 2002 • Vol 101 • No 3

C.C. Chu, C.C. Huang, H.C. Kuo, et al

218

4. Sobue G, Hashizume Y, Mukai E, et al: X-linked recessivebulbospinal neuronopathy. A clinicopathological study.Brain 1989;112:209–32.

5. Warner CI, Servidei S, Lange DJ, et al: X-linked spinalmuscular atrophy (Kennedy’s syndrome): a kindred withhypobetalipoproteinemia. Arch Neurol 1990;47:1117–20.

6. Arbizu T, Santamaria J, Gomez JM, et al: A family withadult spinal and bulbar muscular atrophy, X-linked in-heritance and associated testicular failure. J Neurol Sci1983;59:371–82.

7. Yamamoto R, Shimada T, Takenaka M, et al: Two cases ofprogressive bulbospinal muscular atrophy of sex-linkedrecessive trait (Kennedy-Alter-Sung) with type IVhyperlipoproteinemia and fatty liver. Clin Neurol 1984;24:411–46.

8. Guidetti D, Vescovini E, Motti L, et al: X-linked bulbarand spinal muscular atrophy, or Kennedy disease: clinical,neurophysiological, neuropsychological and molecularstudy of a large family. J Neurol Sci 1996;135:140–8.

9. Mariotti C, Castellotti B, Pareyson D, et al: Phenotypicmanifestations associated with CAG-repeat expansion inthe androgen receptor gene in male patients and het-erozygous females: a clinical and molecular study of 30families. Neuromusc Disord 2000;10:391–7.

10. Wilde J, Moss T, Thrush D: X-linked bulbo-spinalneuronopathy: a family study of three patients. J NeurolNeurosurg Psychiatry 1987;50:279–84.

11. La Spada AR, Wilson EM, Lubahn DB, et al: Androgenreceptor gene mutations in X-linked spinal and bulbarmuscular atrophy. Nature 1991;352:77–9.

12. Fischbeck KH, ar-Rushdi N, Pericak-Vance M, et al:X-linked neuropathy: gene localization with DNA probes.Ann Neurol 1986;20:527–32.

13. Huang CC, Chen RS, Chu NS, et al: X-linked recessivebulbospinal neuronopathy: clinical and molecular studiesin a Taiwanese family. J Formos Med Assoc 1998;97:354–9.

14. Holezu Y, Yanai S, Nagai N, et al: A case of Kennedy-Alter-Sung (KAS) syndrome presenting as hypersexuality andelevated serum CK: usefulness of genetic analysis. RinshoShinkeigaku Clin Neurol 1996;36:471–4.

15. Rudnik-Schoneborn S, Luteenrath S, Borkowska J, et al:Analysis of creatine kinase activity in 504 patients withproximal spinal muscular atrophy types I–III from thepoint of view of progression and severity. Eur Neurol 1998;39:154–62.

16. Liu CS, Chang YC, Chen DF, et al: Type IV hyperlipopro-teinemia and moderate instability of CAG triplet expan-sion in the androgen-receptor gene. Lipid, sex hormoneand molecular study in a Chinese family with Kennedy-Alter-Sung disease. Acta Neurol Scand 1995;92:398–404.

17. Chen RS, Huang CC, Chu NS, et al: Random X chromo-some methylation patterns in the carriers with clinicalphenotypic expressions of X-linked recessive bulbospinalneuronopathy. Acta Neurol Scand 1999;100:249–53.

18. Ionasescu VV, Burns TL, Searby C, et al: X-linkeddominant Charcot-Marie-Tooth neuropathy with 15cases in a family genetic linkage study. Muscle Nerve1988;11;1154–6.

19. Le Guen E, Raise N, Guggenheim M, et al: Linkageanalysis between dominant X-linked Charcot-Marie-Toothdisease, and 15 Xq11-Xq21 microsatellites in a new largefamily: three new markers are closely linked to the gene.Neuromusc Disord 1996;4:463–9.

20. Bergoffen J, Scherer K, Wang S, et al: Connexin muta-tions in X-linked Charcot-Marie-Tooth disease. Science1993;262:2039–42.