As

NNa

b

c

d

e

f

a

ARRAA

KNHGN

1

dn

eertcpll

dT

h0

Clinical Neurology and Neurosurgery 121 (2014) 19–22

Contents lists available at ScienceDirect

Clinical Neurology and Neurosurgery

jo ur nal home p age: www.elsev ier .com/ locate /c l ineuro

ssociation of HLA-DR/DQ polymorphisms with Guillain–Barréyndrome in Tunisian patients

ajiba Fekih-Mrissaa,b,∗, Meriem Mrada,c, Anis Riahid,e, Aicha Sayeha,c, Jamel Zaoualid,e,asreddine Gritli a,f, Ridha Mrissad,e

Hôpital Militaire Principal d’Instruction de Tunis, Service d’Hématologie, Laboratoire de Biologie Moléculaire, 1008 Montfleury, Tunis, TunisieAcadémie Militaire Fondouk Jédid, 8012 Nabeul, TunisieUniversité de Tunis El Manar, Faculté des Sciences de Tunis, 2092, El Manar, TunisieHôpital Militaire Principal d’Instruction de Tunis, Service de Neurologie, 1008 Montfleury, Tunis, TunisieUniversité de Tunis El Manar, Faculté de Médecine de Tunis, 1007 Tunis, TunisieUniversité de Monastir, Faculté de Pharmacie, 5000 Monastir, Tunisie

r t i c l e i n f o

rticle history:eceived 16 June 2013eceived in revised form 20 January 2014ccepted 10 March 2014vailable online 19 March 2014

eywords:euroimmunologyLAuillain–Barré syndrome (GBS)

a b s t r a c t

Human leukocyte antigen (HLA) alleles have been implicated in many autoimmune diseases. The aim ofthis study is to assess whether HLA-DR/DQ alleles confer susceptibility to Guillain–Barré syndrome (GBS)in a Tunisian population.Methods: The HLA-DR/DQ genotyping was performed using polymerase chain reaction sequence-specificprimers (PCR-SSP) in 38 patients with GBS and 100 healthy Tunisian control subjects.Results: GBS in Tunisian patients was found to be associated with the following alleles with theserelative patient versus control frequencies (pc denotes Bonferroni corrected probability values):DRB1*13 (23.68% vs. 9.0%; pc = 0.01), followed by DRB1*14 (22.36% vs.5.5%; pc < 10−3). Two haplotypes,DRB1*14/DQB1*05 and DRB1*13/DQB1*03, were found to be associated with susceptibility to GBS. How-

europathy ever DRB1*07/DQB1*02 and DRB1*03/DQB1*02 haplotypes were more frequently observed in controlsthan in patients (11.5% vs.7.9%; pc = 0.007 and 23% vs. 5.26%; pc < 10−3 respectively). These haplotypesseem to confer protection against the disease.Conclusion: Our data demonstrated a new GBS predisposition associated with HLA-DRB1*14 and DRB1*13.

edisp

Theses alleles could be pr

. Introduction

Guillain–Barré syndrome (GBS) is an acute demyelinatingisease of the peripheral nerves characterized by progressive weak-ess that usually follows a viral illness or immunization [1,2].

The exact pathogenetic mechanisms remain uncertain. How-ver, several features suggest an autoimmune component in thetiology of this disease. Both humoral and cell-mediated immuneesponses have been suspected of playing a role in disease suscep-ibility [3]. The hypothesis of an autoimmune/immune-mediatedause is supported by nerve biopsy studies that confirmed the

resence of lymphocytes and macrophages within the cellu-

ar inflammatory peripheral nerve infiltrate [4–6]. The humaneukocyte antigen (HLA) complex is crucial for immunity and

∗ Corresponding author at: Laboratoire de Biologie Moléculaire, Service’Hématologie, Hôpital Militaire Principal d’Instruction de Tunis, 1008 Montfleury,unis, Tunisie. Tel.: +216 22510488; fax: +216 70762084.

E-mail address: fnajiba@yahoo.fr (N. Fekih-Mrissa).

ttp://dx.doi.org/10.1016/j.clineuro.2014.03.014303-8467/© 2014 Elsevier B.V. All rights reserved.

osing genetic factors for GBS in the Tunisian population.© 2014 Elsevier B.V. All rights reserved.

its polymorphisms are specifically associated with autoimmuneinflammatory diseases [7].

Several genetic studies show that the major histocompatibil-ity complex (MHC) genes, especially HLA antigens represent thestrongest risk for developing several diseases with autoimmunefeatures [8,9]. Because GBS was thought to have an immune-mediated, possibly autoimmune component, the HLA moleculescarried by patients with this disorder have been evaluated in sev-eral studies. HLA and GBD association studies have given a varietyof associations with conflicting results [10–14]

The aim of the present study was to study the distribution ofHLA class II alleles in GBS Tunisian patients.

2. Materials and methods

2.1. Subjects

The study population comprised 38 unrelated patients with GBS(30 male, 8 female) with a mean age ± SD of 39.44 ± 21.65 years.All patients fulfilled standard diagnostic criteria [15]. The criteria

20 N. Fekih-Mrissa et al. / Clinical Neurology and Neurosurgery 121 (2014) 19–22

Table 1Characteristics of GBS patients.

Characteristics Number of patients(n = 38)

Percentage (%)

SexFemale 8 21.05Male 30 78.99

Age at initial evaluation 39.44 ± 21.65 –

rlmwrao(dai

danp

2

oM

nLdwDa

2

iEr(tfbtarctP(

3

wys

Table 2Allele frequency in GBS patients and controls (a) HLA-DRB1 and (b) HLA-DQB1.

HLA % GBS patients(n = 38)

% Controls (n = 100) P-value

HLA-DRB1 typeDRB1*01 5.26 7.5 .51DRB1*03 5.26 21.5 .001a

DRB1*04 5.26 12.5 .07DRB1*07 10.52 15.5 <10(−3a

DRB1*11 13.15 13.5 .94DRB1*13 23.68 9.0 .002a

DRB1*14 22.36 5.5 <10(−3a

DRB1*15 10.52 12 .73

(b) HLA-DQB1 typeDQB1*02 21.0 31.0 .1DQB1*03 44.7 30.0 .1a

DQB1*05 7.8 16.0 .08DQB1*06 26.3 15.5 .19a

4. Discussion

In this study, we report an association of two haplo-types, DRB1*14/DQB1*05 and DRB1*13/DQB1*03, that have not

Table 3The frequencies of HLA DR/DQ haplotypes in GBS patients.

HLA DR/DQ allele DRB1*/DQB1 % GBS patients(n = 38)

% Controls(n = 100)

P-value

DRB1*01/DQB1*05 5.26 7.5 .97DRB1*03/DQB1*02 5.26 23 .007a

DRB1*04/DQB1*03 5.26 12.0 .09DRB1*07/DQB1*02 7.9 11.5 <10(−3a

DRB1*11/DQB1*03 10.52 13.0 .57DRB1*13/DQB1*03 14.75 4.0 .02DRB1*14/DQB1*05 13.5 1.5 <10(−3a

mean(range)

(14–81)

equired evidence of progressive motor weakness of more than oneimb, areflexia, and specific progression (symptoms and signs of

otor weakness that develop rapidly but cease to progress by foureeks into the illness), and the exclusion of other causes for neu-

opathy. Electrophysiological studies and cerebral spinal fluid (CSF)nalysis were investigated in all patients to support the diagnosisf GBS. All patients were treated with intravenous immunoglobulinIVIg) with a dosage of 0.4 g/kg body weight/day for five consecutiveays. All but one patient had clear signs of demyelinating neurop-thy. One patient died in the acute stage of the disease while in thentensive care department. All other patients recovered fully.

The controls consisted of 100 healthy matched individuals withiverse Tunisian origin similar to the patients, none of whom hadny history of GBS, family antecedent with GBS, nor any peripheraleuropathy. All participants gave informed consent before partici-ation in the study.

.2. HLA-typing by DNA amplification

Genomic DNA was extracted from peripheral blood samplesf patients and healthy individuals using the QIAamp®DNA Bloodini Kit (QIAGEN GmbH, Hilden, Germany).Low-resolution HLA typing was performed by PCR-SSP tech-

iques according to Micro SSP DNA Typing Trays DRB/DQB (Oneambda Inc. Canoga Park, CA, USA). Amplified DNA fragments wereetected by agarose gel electrophoresis (2.5% agarose gel), stainedith ethidium bromide, and UV transillumination. One LambdaNA/LMT software version 3.98 was used to detect specific DRB1nd DQB1 alleles.

.3. Statistical analysis

Allele frequencies were estimated by using the direct count-ng method. Haplotype frequencies were estimated using theM algorithm and deviations from Hardy-Weinberg equilib-ium were both performed using the Arlequin v.3.1 softwarehttp://cmpg.unibe.ch/software/arlequin3). All analyses related tohe case-control study were performed using the Statistical Packageor the Social Sciences v.16 (IBM, Armonk, NY, USA). Differencesetween cases and controls were evaluated by using the chi-squareest or Fisher’s test for qualitative variables. The odds ratios (OR)nd 95% confidence intervals (CI) were also calculated. The Bonfer-oni method was used to adjust for type I errors due to multipleomparisons. Corrected probability values (pc) were calculated forhe number of comparisons made, i.e., the number of alleles tested.robability values pc < 0.05 were considered statistically significantTable 1).

. Results

The disease was observed more often in men than womanith a ratio of 3.75. The mean age of disease onset was 39.44

ears. A comparison of the frequencies of tested HLA alleles intudy subjects is given in Table 2. The allelic frequencies were in

GBS, Guillain–Barré syndrome; P-value, probability value; significant P-value is inbold, P < 0.05.

a Corrected P-value.

Hardy–Weinberg equilibrium in all the samples. Probability val-ues (P values) were adjusted by applying the Bonferroni correction(pc). Among DRB1 alleles, HLA-DRB1*13 was increased significantlyin patients compared to controls (23.68% vs. 9.0%; pc = 0.01), fol-lowed by DRB1*14 (22.36% vs.5.5%; pc < 10−3). These two alleleswere associated with an increased risk for GBS. HLA-DRB1*03, how-ever, revealed a marked decrease in frequency (5.26% vs. 21.5%;pc = 0.001) in patients compared to controls, followed by DRB1*07(10.52% vs. 15.5%; pc < 10−3). These alleles seem to confer protec-tion against GBS. None of the DQB1 alleles tested in this studyconstituted a statistically significant risk for GBS.

Only haplotypes exhibiting significant linkage disequilibrium(LD) parameters between alleles were considered for study. Haplo-types with frequencies greater than 5% in either patients or controlswere compared between both groups in Table 3. A pronounced fre-quency of the DRB1*14/DQB1*05 haplotype among patients wasobserved which indicates a strong predisposing effect toward dis-ease (13.5% vs. 1.5%; pc < 10−3). This haplotype was followed insignificance of disease susceptibility by DRB1*13/DQB1*03 (14.75%vs. 4.0%; pc = 0.02).

Among the protective DR-DQ haplotypes, DRB1*07/DQB1*02and DRB1*03/DQB1*02 were significantly more prevalent amongcontrols in comparison with GBS patients (11.5% vs. 7.9%; pc = 0.007and 23% vs. 5.26%; pc < 10−3 respectively). These haplotypes seemto confer protection against GBS.

DRB1*15/DQB1*06 10.52 12.5 .82

GBS, Guillain–Barré syndrome; P-value, probability value; significant P-value is inbold, P < 0.05.

a Corrected P-value.

logy a

plTmftcod

iaacipw

oi[Dhwi(

wjaaaD

tadpDt

hDospatDtchdts

ls

T(rttww

[

[

[

[

[

[

[

[

[

[stee J, et al. HLA class II alleles are not a general susceptibility factor inGuillain–Barré syndrome. Neurology 2005;64:44–9.

N. Fekih-Mrissa et al. / Clinical Neuro

reviously been reported with GBS susceptibility. This researchinking HLA alleles with susceptibility to GBS is the first of its kind inunisia. Despite our small patient series, this research agrees withany studies reporting GBS to be more common in males than

emales [16]. For several autoimmune diseases, it has been shownhat sex-related factors appear to influence the risk associated witharriage of different HLA molecules [9,17]. Given the small numberf patients, it was not possible to investigate HLA associations withisease according to gender.

A limited numbers of studies, with inconsistent results, havenvestigated possible associations between HLA class II antigensnd GBS susceptibility [10,18]. The majority of studies did not findny association (or weak association) between GBS and carriage oflass II HLA molecules [11,21]. Another study was more nuanced ints conclusions and found that although there is no HLA allele thatrovides general susceptibility to GBS, HLA-DRB1*01 is associatedith increased disease severity [19].

A study in Mexico found that GBS was associated with carriagef HLA-DR3 [20]. Another study has also suggested a trend towardncreased frequency of carriage of HLA-DR3 among GBS patients18]. In the current study, we report associations of DRB1*14 andRB1*13 with GBS. Only one study of northern Chinese patientsave previously reported the association of HLA-DRB1*13 alleleith the two forms of GBS (acute inflammatory demyelinat-

ng polyneuropathy (AIDP) and acute motor axonal neuropathyAMAN)) [22].

In a study from England [14], the HLA-DQB1*03 associationas identified among GBS patients with a preceding Campylobacter

ejuni infection; however, no HLA association was found among GBSnd control subjects. In our case, this allele was initially found to bessociated with disease susceptibility, but this result was excludedfter applying the Bonferroni correction. However, the haplotypeRB1*13/DQB1*03 was found to be associated with GBS.

Another study of GBS patients, who when grouped accordingo subtype AIDP or AMAN, found that certain DQB1 moleculesre associated with AIDP but not AMAN [23]. McCombe found aecreased frequency for carriage of DR5 in the GBS group as com-ared to controls [21]. In the current study, HLA-DRB1*03 andRB1*07 alleles were more frequently found in controls compared

o the patient group.The frequency of many HLA DR/DQ polymorphisms was

igh among our Tunisian subjects; however, the frequency ofRB1*14/DQB1*05 and DRB1*13/DQB1*03 were particularly low inur Tunisian cohort of controls (1.5% and 4.0% respectively). Othermaller studies conducted in Tunisia confirm our result. For exam-le, Ayed et al. found a frequency of 2.5% of DRB1*14/DQB1*05nd 4.5% for DRB1*13/DQB1*03 in the general Tunisia popula-ion [24]. Similarly, Hajjej et al. found frequencies of 0.97% forRB1*14/DQB1*05 and 5.82% for DRB1*13/DQB1*03 [25]. Given

hat these haplotypes were found at significantly greater frequen-ies in our Tunisian patients, it is a strong indication that theseaplotypes are risk factors for GBS. Certainly GBS is a fairly rareisorder and, due to the rather small number of cases included inhis study, the observed associations should be confirmed by largertudies.

The diverse nature of allelic associations with GBS among popu-ations can perhaps be explained by the varying predominance ofome alleles among different populations and ethnicities.

Among demyelinating diseases, the study of HLA associations inunisian patients has been investigated only in multiple sclerosisMS) and chronic inflammatory demyelinating polyradiculoneu-opathy (CIDP) [8,26]. The alleles associated with susceptibility

o GBS are different from those of MS. DRB1*13 was foundo be associated to CIDP and GBS; however, the haplotypesere different: DRB1*13/DQB1*06 was associated with CIDPhile DRB1*13/DQB1*03 was associated with GBS. Perhaps these

[

[

nd Neurosurgery 121 (2014) 19–22 21

differences are a consequence of the diverse pathogeneses of GBScompared with that of other HLA II-related autoimmune diseases.For example, while MS patients may experience a polyphasicoccurrence of neurologic events (i.e., relapsing-remitting MS), GBSpatients have a monophasic course of disease.

Given that GBS is a fairly rare disorder and due to the rather smallnumber of cases included in this study, the observed associationsshould be confirmed by a larger study. Further, it would be benefi-cial to study HLA genotyping of GBS patients by subtype, AIDP andAMAN, to resolve any distinction of HLA associations between sub-types. These results could suggest different mechanisms of diseasedevelopment.

To our knowledge, this is the first study performed to analyzethe association of HLA DRB1/DQB1 alleles on GBS susceptibility inTunisia. The validity of our results awaits confirmation with largerscale studies.

Acknowledgment

We would like to thank Dr. Christian Winchell for his precioushelp in correcting this manuscript.

References

[1] Ho TW, McKhann GM, Griffin JW. Human autoimmune neuropathies. Annu RevNeurosci 1998;21:187–226.

[2] Hughes RA, Hadden RD, Gregson NA, Smith KJ. Pathogenesis of Guillain–Barrésyndrome. J Neuroimmunol 1999;100:74–97.

[3] Arnason BG, Soliven B. Acute inflammatory demyelinating neuropathy. In: DyckPJ, Thomas PK, editors. Peripheral neuropathy. 3rd ed. Philadelphia: WB Saun-ders; 1993. p. 1437–97.

[4] Pollard JD, McCombe PA, Baverstock J, Gatenby PA, McLeod JG. Class II antigenexpression and T lymphocyte subsets in chronic inflammatory demyelinatingpolyneuropathy. J Neuroimmunol 1986;13:123–34.

[5] Winer J, Hughes S, Cooper J, Ben-Smith A, Savage C. Gamma delta T cells infil-trating sensory nerve biopsies from patients with inflammatory neuropathy. JNeurol 2002;249:616–21.

[6] Illés Z, Shimamura M, Newcombe J, Oka N, Yamamura T. Accumulation ofValpha7.2–Jalpha33 invariant T cells in human autoimmune inflammatorylesions in the nervous system. Int Immunol 2004;16:223–30.

[7] Horton R, Wilming L, Rand V, Lovering RC, Bruford EA, Khodiyar VK, et al. Genemap of the extended human MHC. Nat Rev Genet 2004;5:889–99.

[8] Messadi A, Fekih-Mrissa N, Ouerhani S, Zaweli J, Louatti I, Layouni S, et al. HLAclass II alleles and multiple sclerosis in Tunisian patients. Clin Neurol Neurosurg2010;112:849–52.

[9] Fekih-Mrissa N, Klai S, Zaouali J, Gritli N, Mrissa R. Association of HLA-DR/DQpolymorphism with myasthenia gravis in Tunisian patients. Clin Neurol Neu-rosurg 2013;115:32–6.

10] Winer JB, Briggs D, Welsh K, Hughes RA. HLA antigens in the Guillain–Barrésyndrome. J Neuroimmunol 1988;18:13–6.

11] Hillert J, Osterman PO, Olerup O. No association with HLA-DR, -DQ or -DP allelesin Guillain–Barré syndrome. J Neuroimmunol 1991;31:67–72.

12] Kaslow RA, Sullivan-Bolyai JZ, Hafkin B, Schonberger LB, Kraus L, Moore MJ,et al. HLA antigens in Guillain–Barré syndrome. Neurology 1984;34:240–2.

13] Latovitzki N, Suciu-Foca N, Penn AS, Olarte MR, Chutorian AM. HLA typing andGuillain–Barré syndrome. Neurology 1979;29:743–5.

14] Rees JH, Vaughan RW, Kondeatis E, Hughes RA. HLA-class II alleles inGuillain–Barré syndrome and Miller Fisher syndrome and their associationwith preceding Campylobacter jejuni infection. J Neuroimmunol 1995;62:53–7.

15] Asbury AK. Diagnostic considerations in Guillain–Barré syndrome. Ann Neurol1981;9:1–5.

16] Govoni V, Granieri E. Epidemiology of the Guillain–Barré syndrome. Curr OpinNeurol 2001;14:605–13.

17] Celius EG, Harbo HF, Egeland T, Vartdal F, Vandvik B, Spurkiand A. Sex andage at diagnosis are correlated with the HLA-DR2, DQ6 haplotype in multiplesclerosis. J Neurol Sci 2000;178:132–5.

18] Stewart GJ, Pollard JD, McLeod JG, Wolnizer CM. HLA antigens in theLandry–Guillain–Barré syndrome and chronic relapsing polyneuritis. Ann Neu-rol 1978;4:285–9.

19] Geleijns K, Schreuder GM, Jacobs BC, Sintnicolaas K, van Koningsveld R, Meul-

20] Gorodezky C, Varela B, Castro-Escobar LE, Chávez-Negrete A, Escobar-GutiérrezA, Martínez-Mata J. HLA-DR antigens in Mexican patients with Guillain–Barrésyndrome. J Neuroimmunol 1983;4:1–7.

21] McCombe PA, Csurhes PA, Greer JM. Studies of HLA associations in maleand female patients with Guillain–Barré syndrome (GBS) and chronic

2 logy a

[

[

[

[

Eur J Med Genet 2006;49:43–56.

2 N. Fekih-Mrissa et al. / Clinical Neuro

inflammatory demyelinating polyradiculoneuropathy (CIDP). J Neuroimmunol2006;180:172–7.

22] Monos DS, Papaioakim M, Ho TW, Li CY, McKhann GM. Differential distri-bution of HLA alleles in two forms of Guillain–Barré syndrome. J Infect Dis1997;176:S180–2.

23] Magira EE, Papaioakim M, Nachamkin I, Asbury AK, Li CY, Ho TW, et al. Dif-

ferential distribution of HLA-DQ beta/DR beta epitopes in the two forms ofGuillain–Barré syndrome, acute motor axonal neuropathy and acute inflam-matory demyelinating polyneuropathy (AIDP): identification of DQ betaepitopes associated with susceptibility to and protection from AIDP. J Immunol2003;170:3074–80.

[

nd Neurosurgery 121 (2014) 19–22

24] Ayed K, Ayed-Jendoubi S, Sfar I, Labonne MP, Gebuhrer L. HLA class-I and HLAclass-II phenotypic, gene and haplotypic frequencies in Tunisians by usingmolecular typing data. Tissue Antigens 2004;64:520–32.

25] Hajjej A, Hmida S, Kaabi H, Dridi A, El Gaaied A, Boukef K. HLA genes in SouthernTunisians (Ghannouch area) and their relationship with other Mediterraneans.

26] Mrad M, Fekih-Mrissa N, Mansour M, Seyah A, Riahi A, Gritli N, Mrissa R.Association of HLA-DR/DQ polymorphism with chronic inflammatory demyeli-nating polyradiculoneuropathy (CIDP) in Tunisian patients. Transfus Apher Sci2013;49:623–6.