Frequency and patterns of first- and second-line drug resistance-conferring mutations in Mycobacterium tuberculosis isolated from pulmonary tuberculosis patients in a cross-sectional study in Tigray Region, Ethiopia

(1)

Frequency and patterns of ﬁ rst- and second-line drug

resistance-conferring mutations in Mycobacterium tuberculosis isolated from pulmonary tuberculosis patients in a cross-sectional study in Tigray Region, Ethiopia

Letemichael Negash Welekidan

^a,b,c,

*, Eystein Skjerve

^b

, Tsehaye Asmelash Dejene

^c

, Mengistu Welday Gebremichael

^d

, Ola Brynildsrud

^a,e

, Tone Tønjum

^f,g

,

Solomon Abebe Yimer

^g,h

aDepartmentofParaClinicalSciences,NorwegianUniversityofLifeSciences,P.O.Box369,0102Oslo,Norway

bDepartmentofProductionAnimalMedicine,NorwegianUniversityofLifeSciences,P.O.Box369,0102Oslo,Norway

cDepartmentofMedicalMicrobiologyandImmunology,DivisionofBiomedicalSciences,CollegeofHealthSciences,MekelleUniversity,P.O.Box1871, Mekelle,Ethiopia

dDepartmentofMidwifery,CollegeofHealthSciences,MekelleUniversity,P.O.Box1871,Mekelle,Ethiopia

eDepartmentofBacteriologyandImmunology,NorwegianInstituteofPublicHealth,P.O.Box222,0213Oslo,Norway

fDepartmentofMicrobiology,UnitforGenomeDynamics,UniversityofOslo,P.O.Box1072,0316Oslo,Norway

gDepartmentofMicrobiology,UnitforGenomeDynamics,OsloUniversityHospital,P.O.Box4950,0424Oslo,Norway

hCoalitionforEpidemicPreparednessInnovations,Oslo,Norway

ARTICLE INFO

Articlehistory:

Received9May2020

Receivedinrevisedform23October2020 Accepted17November2020

Availableonline3December2020

Keywords:

Rifampicin Isoniazid Fluoroquinolones Mutation

Mycobacteriumtuberculosis Ethiopia

ABSTRACT

Objectives:Tuberculosis (TB) is a preventable and treatable infectious disease, but the continuing emergenceandspreadofmultidrug-resistantTBisthreateningglobalTBcontrolefforts.Thisstudyaimed to describe the frequency and patterns of drug resistance-conferringmutations of Mycobacterium tuberculosis(MTB)isolatesdetectedfrompulmonaryTBpatientsinTigrayRegion,Ethiopia.

Methods:Across-sectionalstudydesignwasemployedtocollectsputumsamplesfrompulmonaryTB patientsbetweenJuly2018toAugust2019.CultureandidentiﬁcationtestsweredoneatTigrayHealth ResearchInstitute(THRI).Mutationsconferringrifampicin(RIF),isoniazid(INH)andﬂuoroquinolone (FQ) resistancewere determined in 227 MTB isolates using GenoType MTBDRplus and GenoType MTBDRsl.

Results:MutationsconferringresistancetoRIF,INHandFQsweredetectedin40/227(17.6%),41/227 (18.1%)and2/38 (5.3%)MTBisolates,respectively.ThemajorityofmutationsforRIF,INHand FQs occurredatcodonsrpoBS531L(70%),katGS315T(78%)andgyrAD94Y/N(100%),respectively.Thisstudy revealedasigniﬁcantnumberofunknownmutationsintherpoB,katGandinhAgenes.

Conclusion:HighratesofmutationsconferringresistancetoRIF,INHandFQswereobservedinthisstudy.

Alargenumberofisolatesshowedunknownmutations,whichrequirefurtherDNAsequencinganalysis.

Periodicdrugresistancesurveillanceandscaling-upofdrugresistancetestingfacilitiesareimperativeto preventthetransmissionofdrug-resistantTBinthecommunity.

licenses/by/4.0/).

1.Introduction

Tuberculosis (TB) is a preventable and treatable infectious disease. However, the continuing emergence and spread of multidrug-resistantTB(MDR-TB)isthreateningglobalTBcontrol efforts.Amongthe10millionincidentTBcasesreportedin2018, some484000patientshadrifampicin-resistantTB(RR-TB),andof these78%wereMDR-TB,deﬁnedasresistancetoatleastisoniazid (INH)andrifampicin(RIF),thetwomostpotentﬁrst-lineanti-TB

*Correspondingauthor.Presentaddress:DepartmentofMedicalMicrobiology and Immunology,DivisionofBiomedicalSciences, CollegeofHealthSciences, MekelleUniversity,P.O.Box1871,Mekelle,Ethiopia.

E-mailaddresses:[email protected], [email protected](L.N.Welekidan).

http://dx.doi.org/10.1016/j.jgar.2020.11.017

ContentslistsavailableatScienceDirect

Journal of Global Antimicrobial Resistance

j o u r n a l h o m ep a g e: w w w . e l s e v i e r . c o m / l o c at e / j g a r

(2)

drugs[1].AmongthereportedMDR-TBcasesin2018,6.2%were estimatedtobeextensivelydrug-resistantTB(XDR-TB)[1].Hence, in2019theWorldHealthOrganization(WHO)reportedthatthe worldwasnotontracktoreachthe2020milestonesoftheEndTB Strategy.

Ethiopiaisoneofthe14highTB,TB/HIVandMDR-TBburden countrieswithincidentTB,MDR/RR-TBcasesandoverallTBdeaths of165000,1600and26200,respectively,intheyear2018[1].

Themostimportantfactorsfortheemergenceand spreadof MDR-TBaremismanagementofTBtreatment,inappropriateuseof antimicrobialdrugsortheuseofineffectiveformulationsofdrugs suchastheuseofasingledrugs,poorqualitymedicinesorbad storageconditions,andprematuretreatmentinterruption.Hence, bacteriacanthenbetransmittedtoothersusceptibleindividuals.

Accumulationof point mutationsin coding regions fordrug targetsand/ordrug-convertingenzymesisamajormechanismfor acquiring resistance in Mycobacterium tuberculosis (MTB) [2].

Severalstudieshavereportedthatcodon531oftherpoBgeneand codon 315 of the katG gene are found to have the highest mutational frequency for RIF resistance and INH resistance, respectively[3–6].However,otherstudiesconductedinEthiopia reported that INH resistancewas completely due tokatG gene mutationswithoutanymutationsintheinhAgene[7,8].Thereisa reported association between drug resistance mutations and speciﬁc lineages of MTB [8,9] and an association of speciﬁc mutation patterns with high patient mortality [5]. Hence, assessingthedrugresistanceburdenandcorrespondingmutation patternsinMTBisolatesindifferentcountriesandwithinregions of a country can provide better knowledge of the underlying mechanisms of drug resistance-conferring mutations. Besides servingasabasisforimprovingthenationalTBcontroleffortto addressdrug-resistantTB, suchstudies canalsoprovidecrucial informationtoselectthebesttherapeuticoptionsandtodevelop novel drugsthat canovercomeexisting resistancemechanisms.

Given thesigniﬁcance ofunderstandinganti-TB drugresistance mechanismsforTBcontrol,thereisalackofadequatestudieson thiscrucialresearchareainmanyhighTBburdencountries.

Afew studiesin Ethiopia haveassessedthedrugresistance burden and corresponding mutation patterns of MTB isolates detectedinvariousregionsofthecountry[3,8,10].However,no study to date has assessed the frequency of anti-TB drug resistance-conferringmutationsintheTigrayRegionofEthiopia.

Thisstudyaimedtodescribethefrequencyandpatternsofﬁrst- and second-line drug resistance-conferring mutations of MTB isolated from pulmonary TB patients in the Tigray Region, Ethiopia.

2.Materialsandmethods 2.1.Studysetting

ThisstudywasconductedinselectedhealthfacilitiesinTigray Region,Ethiopia, includingthefollowing six hospitals:Alamata Hospital (Southern Zone); Mekelle Hospital (Mekelle Special Zone); AdigratHospital (Eastern Zone);Adwa Hospital (Central Zone);Shire/SuhulHospital(North-WesternZone);andHumera/

Kahsay Abera Hospital (Western Zone). Tigray Region has an estimated total population of 5.13 million [11]. The region is administrativelydividedintosevenzones(includingonespecial zone, Mekelle). The geographical location of the hospitals and zonesofthestudyareaaredepictedinFig.1.

2.2.Studydesignandpopulation

Ahospital-basedcross-sectional studydesign wasemployed.

Data werecollected fromJuly 2018to August2019. The study populationcomprised allpresumptivepulmonaryTBcaseswho visited selectedhealth facilities in theregionduring the study period.

2.3.Samplesizeandsamplingprocedure

The sample size was determined by taking the required minimumnumberofMDR-TBpatientstobeenrolledinthestudy,

Fig.1.Mapofthestudyarea.

(3)

which wasestimatedtobe30–40cases.Thiswas toestablisha sample with a sufficient number of MDR-TB isolates to allow sufficientpowertodetectdifferenttypesofMDR-TBisolates.Based onanexpectedlevelof10%MDR-TBamongTBpatients[3],we aimedatrecruitingatleast300TBpatientsforthestudy.Without adjusting forclustering,theprecisionoftheestimatewouldbe 6.7–14.2%(95%confidenceinterval,relativeprecisionof3.3%).

HospitalsfromsixadministrativezonesofTigrayRegionwere selected based on availability of the GeneXpert¹ facility. A consecutive sampling technique was employed to recruit the studysampleinallofthehospitalsuntiltherequiredsamplesize wasobtained.

2.4.Inclusionandexclusioncriteria

PulmonaryTBpatientswhowerenotreceivingtreatment,aged 15 years with a GeneXpert¹-positive result who provided writteninformed consentwereincludedinthestudy. Critically ill patients fromwhom sociodemographic information, clinical dataandsputumsamplescouldnotbeobtained,patientsaged<15 yearsandthosewithextrapulmonaryTBwereexcludedfromthe study.

2.5.Sociodemographicdatacollection

Apre-testedstructuredquestionnairewasemployedtocollect sociodemographicdata,includingage,sex,residenceandhistoryof previousTBtreatment.Adequatetrainingindatacollectionwas providedtodatacollectorsbeforecommencementofthestudy.

2.6.Sputumsamplecollectionandtransportation

A 5–10mLsputumsamplewascollectedfromeligiblestudy participantsusingacodedandsterile50-mLFalcontubeaccording totherecommendationoftheEthiopianFederalMinistryofHealth [12].Sputumsampleswereproperlypackedandkeptat4Cfor transportationin anice bag toTigray HealthResearch Institute (THRI) following international standards of the WHO for the transport ofbiological substances(categoryB, UN-3373).Speci- mens arrivedwithin4–5daysofcollectionandwereprocessed within7daysfromthetimeofﬁrstcollection.

2.7.Sputumdecontaminationprocedure

Sputum samples were digested and decontaminated using freshlypreparedN-acetyl-L-cysteine(NALC)andsodiumhydroxide with a ﬁnal sodium hydroxide concentration of 1%. An equal volumeofNALCandsodiumhydroxidesolutionwasaddedtothe specimen and was incubated for 15 min [13]. Following centrifugationat3000gat4Cfor15min,thesedimentwas re-suspendedin2mLofsterilephosphate-bufferedsaline(PBS) (pH6.8).Finally,analiquotof0.5mLofsedimentwasinoculated intoaMycobacteriumGrowthIndicatorTube(MGIT)960andwas loadedontoaBACTECMGIT960instrumentandtwotofourdrops ofthesedimentwereinoculatedintoLowenstein–Jensen(LJ)egg medium.Aloopfulofthesedimentwasusedfordirectmicroscopic examinationusingthestandardZiehl-Neelsenstainingtechnique atTHRI[14].

2.8.Sputumculture

The decontaminated supernatant decanted sputum samples werecultured onLJ medium and in BACTEC MGIT 960culture medium (Becton Dickinson Microbiology Systems, Sparks, MD, USA) following standard operational procedures. Tubesfor the solidculturewereincubatedat37Cinaslantpositiontoensure anevendistributionoftheinoculumfor1weekandthereafterat 37Cinairforanother7weeksandwerecheckedonceaweekfor mycobacterialgrowth.Cultureswereconsiderednegativewhenno growthwasseenafter8weeksofincubationforsolidcultureand6 weeks(42days)for liquidculture.Growthofmycobacteriawas conﬁrmed by its typical colony morphology, acid-fast bacilli staining,Capiliaantigen test andinoculation ontoa blood agar platetoruleoutcontamination.

2.9.Drugsusceptibilitytestingforﬁrst-andsecond-lineanti-TBdrugs usingthelineprobeassay(LPA)

All culture-positive isolates were subjected to GenoType MTBDRplus,andisolatesthatwereresistanttobothINHandRIF (MDR-TB) were tested again by GenoType MTBDRsl genotypic method following the manufacturer’s instructions (GenoType¹ MTBC;HainLifescience,Nehren,Germany)[15].

Fig.2. Zonaldistributionofthe227tuberculosis(TB)casesintheTigrayRegion,NorthernEthiopia,July2018toAugust2019.Note:numbersareinpercent.

(4)

2.10.Qualityassuranceandqualitycontrol

All laboratory analyses were carried out following standard operatingprocedures.BoththesolidcultureandLPAprocedures werecheckedandvalidated.TheMTBreferencestrainH37Rvwas used as a quality control organism throughout the LPA test.

Moreover,bothstartandendcontrolswereusedduringeachbatch of specimenprocessingand DNAextraction, anda no-template controlwasalsousedforLPAreagents.

2.11.Dataentryandstatisticalanalysis

DatawereenteredusingEpiData3.1dataentrysoftware.After cleaningandvalidation,dataweretransferredintoStataSE15/SE forWindows(StataCorpLP,CollegeStation,TX,USA)forstatistical analysis. Descriptive statistics were computed and frequencies werepresentedintables.

3.Results

Atotalof227MTBculture-positiveparticipantswereincluded inthestudy,ofwhich144(63.4%)weremaleand83(36.6%)were female,withameanstandarddeviationageof3413.8years (range15–85years).Themajorityofthestudyparticipants(127/

227;55.9%)wereurbandwellers.Thezonaldistributionindicated thatTBcaseswerehighestintheEasternZone(27.3%),followedby MekelleSpecialZone(21.6%)(Fig.2).

3.1.Frequencyofdrugresistance

Ofthe227isolatesinthisstudy,40(17.6%)wereRIF-resistant,of which21(52.5%),16(40.0%)and3(7.5%)werefromnew,relapse andtreatmentfailurecases,respectively.Amongthe41isolates (18.1%)resistanttoINH,23(56.1%),15(36.6%)and3(7.3%)were detectedfromnew,relapseandfailurecases,respectively.There weretwoisolates(0.9%)monoresistanttoRIFandthreeisolates (1.3%)monoresistanttoINH.

The overall proportionof MDR-TBwas 38 (16.7%),with the majorityfromnewcases(20/38;52.6%)followedbyrelapsecases (15/38;39.5%)andfailurecases(3/38;7.9%).Amongthe38MDR- TBisolatestestedforresistancetosecond-lineanti-TBdrugs,only2 (5.3%)FQ-resistantisolatesweredetected,ofwhich1wasfroma relapse case and 1 was from a treatmentfailure case (Fig. 3;

Table1).

3.2.Frequencyofmutationsforresistancetoﬁrst-andsecond-line anti-TBdrugs

Resistance-conferring mutations were observed in 40/227 (17.6%),41/227(18.1%)and 2/38(5.3%)isolatesfor RIF,INHand FQs,respectively(Tables1and2).

Among the 40 RIF-resistant isolates, the majority (70%) of mutationsoccurredatcodonS531LoftherpoBgene,followedby mutationsatcodonsH526Y(10%),H526D(7.5%)andD516V(5%).A total of 20% of RIF-resistant isolates were associated with the absenceof wild-type(WT) probeswithoutthedevelopmentof Fig.3.Flowchartoftheprocedureusedtorecruitallparticipantsinthestudy,TigrayRegion,Ethiopia,July2018toAugust2019.MDR-TB,multidrug-resistanttuberculosis;

pre-XDR-TB,pre-extensivelydrug-resistanttuberculosis.

Table1

Patientcharacteristicsandfrequencyofresistancetoﬁrst-andsecond-lineantituberculardrugsinrelationtopatienttuberculosiscategories,TigrayRegion,Ethiopia,July 2018toAugust2019.

Variable RIF-resistant(n=40)[n(%)] INH-resistant(n=41)[n(%)] MDR(n=38)[n(%)] FQ-resistant(n=2)[n(%)]

New Relapse Failure New Relapse Failure New Relapse Failure New Relapse Failure

Sex

Male 9(22.5) 13(32.5) 2(5.0) 11(26.8) 12(29.3) 2(4.9) 8(21.1) 12(31.6) 2(5.3) 0(0) 1(50.0) 1(50.0) Female 12(30.0) 3(7.5) 1(2.5) 12(29.3) 3(7.3) 1(2.4) 12(31.6) 3(7.9) 1(2.6) 0(0) 0(0) 0(0) Age(years)

15–24 7(17.5) 2(5.0) 1(2.5) 6(14.6) 2(4.9) 1(2.4) 6(15.8) 2(5.3) 1(2.6) 0(0) 0(0) 0(0) 25–34 6(15.0) 7(17.5) 1(2.5) 7(17.1) 7(17.1) 1(2.4) 6(15.8) 7(18.4) 1(2.6) 0(0) 1(50.0) 0(0) 35–44 6(15.0) 4(10.0) 1(2.5) 7(17.1) 3(7.3) 1(2.4) 6(15.8) 3(7.9) 1(2.6) 0(0) 0(0) 1(50.0)

45–54 1(2.5) 1(2.5) 0(0) 1(2.4) 1(2.4) 0(0) 1(2.6) 1(2.6) 0(0) 0(0) 0(0) 0(0)

55 1(2.5) 2(5.0) 0(0) 2(4.9) 2(4.9) 0(0) 1(2.6) 2(5.3) 0(0) 0(0) 0(0) 0(0)

Residence

Urban 14(35.0) 9(22.5) 3(7.5) 15(36.6) 8(19.5) 3(7.3) 13(34.2) 8(21.1) 3(7.9) 0(0) 1(50.0) 1(50.0)

Rural 7(17.5) 7(17.5) 0(0) 8(19.5) 7(17.1) 0(0) 7(18.4) 7(18.4) 0(0) 0(0) 0(0) 0(0)

Total 21(52.5) 16(40.0) 3(7.5) 23(56.1) 15(36.6) 3(7.3) 20(52.6) 15(39.5) 3(7.9) 0(0) 1(50.0) 1(50.0) RIF,rifampicin;INH,isoniazid;MDR,multidrug-resistant;FQ,ﬂuoroquinolone.

(5)

corresponding mutant bands and were reported as unknown mutations.TheunknownmutationsforRIFresistanceoccurredat rpoBWT3(5%),rpoBWT4(2.5%),rpoBWT6(2.5%),rpoBWT7(5%) andrpoBWT8(5%).

Ofthe41INH-resistantstrains,32isolates(78%)hadmutations atkatGgenecodonS315Tand5%ofisolateshadmutationsatinhA genecodonC15T.Atotalof19.5%ofINH-resistantisolatesshowed missing WT bandwithout the presence of mutation bands, of which17.1%wereobservedinthekatGgeneregionand2.4%inthe inhAgeneregion.TherewasnoisolatewithmutationsbothinkatG andinhAgeneregions.

Themostfrequentresistance-conferringmutationforFQswas observed at gyrA codon D94Y/N (100%). However, an equal frequencyofothermutationsatgyrAcodonD94G(50%)andgyrA codonA90V(50%)wereobservedinFQ-resistantisolates(Fig.4;

Table2) 4.Discussion

This is the ﬁrst report assessing the frequency of drug resistance-conferringmutationstoﬁrst-andsecond-lineanti-TB drugsinTigrayRegion,Ethiopia.

Inthepresentstudy,amongthe40(17.6%)RIF-resistantisolates, 70%oftherpoBgenemutationsoccurredatcodonS531L,followed by 10%atcodon H526Y,7.5% atcodonH526Dand 5%atcodon D516V.ThisﬁndingisinlinewithpreviousstudiesfromEthiopiaas well as the ﬁndings of other studies from countries including Vietnam,TaiwanandIndia.AtStPeter’sTBSpecializedHospitalin Addis Ababa(Ethiopia),thehighest prevalenceofmutationwas observedatcodonS531L(68.7%),followedbyH526Y(6.6%)and H526D(4%)[16].InVietnam,themostcommonpointmutations for the rpoB gene occurred at codon 531 (37.8%), followed by codons526(23%)and516(9.5%)[17].InTaiwan,mutationswere observedatcodons531(49.4%),526(20.4%)and516(8.6%)[18]of therpoBgene.Themostfrequentmutationsinthestudyconducted inIndiawereobservedatcodonsS531L(57.81%),H526Y/D(14.8%) andD516V(6.42%)[19].

ThefrequencyofmutationatcodonS531Linourstudyishigher thaninonestudyconductedatStPeter’sTBSpecializedHospital

(68.7%)[16] aswell asstudies inVietnam (37.8%) [17], Taiwan (49.4%)[18] andIndia(57.8%)[19].Conversely,thefrequencyof mutationatS531LoftherpoBgeneobservedinthecurrentstudyis lowerthan studiesconductedin anotherstudyat StPeter’sTB SpecializedHospital(81.3%)[20],inJigjiga,Ethiopia(80%)[3]and inSouthwestEthiopia(82%)[10].

Mutationsin therpoBgene resultingin aminoacidchanges withinthe81-bpcoreregionoftheRNApolymeraseβ-subunitare foundin96%ofRIF-resistantMTBstrains[17].Thisstudyrevealed an overall proportion of 92.5% of mutations conferring RIF resistance, which is comparable with the findings in a report fromTaiwan(91.1%)[18].Ourfindingislowerthanthestudyin Jigjiga,Ethiopia(100%)[3]and higherthanthefindingsfromSt Peter’sTBSpecializedHospital,AddisAbaba(82%) [16],Amhara Region(1.3%) [7],SouthwestEthiopia(85.3%) [10]and Vietnam (76%)[17].

In the current study, 20% unknown mutations conferring resistance to RIF were observed. Similarly, other studies reported that rates of unknown mutations conferring RIF resistancewere16%atStPeter’sTBSpecializedHospital,Addis Ababa,Ethiopia[16],14.7%inSouthwestEthiopia[10]and21%

in India [19]. Several studies have shown that mutationsassociatedwithRIFresistancecanbefoundoutside ofthe81-bphotspotregionoftherpoBgenesuchasatcodons 490, 535, 504, 541, 553 and 572 [21,22] which cannot be detected byLPA. Thisis supportedbya DNAsequencing study performed for RIF drug resistance in Vietnam where 20.3% of strains had novel mutations at codon 490 [17]. Generally, mutationsoccurringin therpoB hotspotregiondidnotexhibit thesamelevelofresistancetoRIF.Themostcommonmutations (65–86%)thatoccuratcodons531and526areassociatedwitha highlevelofresistancetoRIF.Mutationsatcodon516resultedin low-levelresistanceofMTBtoRIF.Raremutationsassociatedwith RIFresistancehavebeenfoundintheamino-terminalregionof rpoB [21,22]. Most reference laboratories that use molecular methods(LPA)onlyexaminethe81-bpregionof therpoBgene.

Hence,itisrecommendedtoscreenforamino-terminalmutations toidentifythepresenceofphenotypicRIFresistance-conferring mutations[23].

Table2

Frequencyofmutationsconferringresistancetoﬁrst-andsecond-lineantituberculardrugsinTigrayRegion,Ethiopia,July2018toAugust2019,determinedbylineprobe assay.

Gene FailingWTband Codonanalysed Developedmutationband Aminoacidchange RIF-resistant(n=40)[n(%)]

rpoB rpoBWT3 516 rpoBMUT1 Asparticacid→valine 1(2.5)

rpoBWT3 513–517 Notdetected Unknown 2(5.0)

rpoBWT4 516 rpoBMUT1 Asparticacid→valine 1(2.5)

rpoBWT8 531 rpoBMUT3 Serine→leucine 19(47.5)

rpoBWT7 526 rpoBMUT2A Histidine→tyrosine 3(7.5)

rpoBWT7 526 rpoBMUT2B Histidine→asparticacid 3(7.5)

Notdetected 531 rpoBMUT3 Serine→leucine 9(22.5)

Notdetected 526 rpoBMUT2A Histidine→tyrosine 1(2.5)

Gene FailingWTband Codonanalysed Developingmutationband Aminoacidchange INH-resistant(n=41)[n(%)]

katG katGWT 315 katGMUT1 Serine→threonine 31(75.6)

katGWT 315 Notdetected Unknown 7(17.1)

Notdetected 315 katGMUT1 Serine→threonine 1(2.4)

inhA inhAWT1 –15 inhAMUT1 Cysteine→threonine 2(4.9)

inhAWT2 –8 Notdetected Unknown 1(2.4)

Gene FailingWTband Codonanalysed Developingmutationband Aminoacidchange FQ-resistant(n=2)[n(%)]

gyrA gyrAWT3 94 gyrAMUT3B Asparticacid→tyrosine/asparticacid→asparagine 1(50.0) gyrA Notdetected 94 gyrAMUT3B Asparticacid→tyrosine/asparticacid→asparagine 1(50.0)

gyrA Notdetected 94 gyrAMUT3C Asparticacid→glycine 1(50.0)

gyrA Notdetected 90 gyrAMUT1 Alanine→valine 1(50.0)

WT,wild-type;MUT,mutant;RIF,rifampicin;INH,isoniazid;FQ,ﬂuoroquinolone.

(6)

Theobserveddifferencesinthefrequencyofmutationscouldbe linkedtoseveralfactors.First,itmayberelatedtotheassociation betweenanti-TB drugresistance and lineagediversitiesof MTB strains[8].Second,itmaybeduetotheassociationofthevarious MTBlineageswithgeographiclocationsandthesociodemographic backgroundofstudyparticipants[24].Astudyreportedthelink betweenrpoBgenesinglenucleotidepolymorphisms(SNPs)and RIFresistance,andthesewereMTBlineage-speciﬁc[25].Molecular analysisofMTBstrainsatStPeter’sTBSpecializedHospital,Addis Ababa,a nationalreferral hospital,showedthepresenceofhigh straindiversityofMTBacrossthecountry[20].

Mutations in the catalase-peroxidase gene (katG) and the enoyl-acylcarrierproteinreductasegene(inhA)havebeenfoundto accountmoreforINHresistance[17].Abateetal.reportedthatthe frequencyofmutationsatkatGgenecodonS315Trangedfrom50– 95%dependingonthegeographicdistribution[16,26],andthisisin linewiththecurrentstudy.Alowfrequencyofmutationsatcodon C15ToftheinhApromoter(4.9%)wasreportedinthisstudy.Our studyﬁndingissupportedbyotherresearcherswhoreportedthat the katG gene codon S315T was the most prevalent mutation responsible for INH resistance [27–29]. Besides, several study reportsfromEthiopiaindicatedthatINHresistancewasduesolely to katGmutations, and noinhA promoter regionmutation was observed[7,8,20].

ThefrequencyofmutationsforS315TofthekatGgeneandC15T ofinhAinthecurrentstudyishigherforS315T(64%)andlowerfor C15T (19%) compared with a global report [30]. Moreover,the frequencyoftheinhAgenemutationinourstudyislowerthanthat reportedfromSouthwestEthiopia(9.8%)[10]andIndia(24.75%) [19],buthigherthanthestudyreportfromStPeter’sTBSpecialized Hospital,AddisAbaba,Ethiopia(0.8%)[16].

AnoverallmutationproportiontoINHinthekatGregulatory gene occurredin 78%of isolates.Thisﬁnding islowerthanthe study results reported from St Peter’s TB Specialized Hospital, Addis Ababa, Ethiopia (93%) [16], Jigjiga, Ethiopia (83.3%) [3], SouthwestEthiopia(90.2%)[10]andChina(86.2%)[31].Thestudy resultishigherthantheﬁndingsreportedfromAmharaRegion, Ethiopia(3.5%)[7],India(62.6%)[19]and Vietnam(76.8%)[17].

VariationsinthefrequencyofmutationsatcodonS315TofkatG and in inhA in the differentstudies largely depend on the TB

prevalenceinthevariouscountriesacrosstheworldanddiffer- encesingeographiclocation[23,32].INHremainsoneofthemost potentfirst-lineanti-TBdrugs.Theeffectofthisessentialdrugis jeopardisedbytheemergenceofdrug-resistantMTBstrains.Thisis associated with clinically significant or high levels of INH resistanceowing tothe absence of significant loss of bacterial fitnessresultingfromfrequentmutation[8,33].MutationatkatG indicatesahighlevelofresistancetoINHsuchthattreatmentusing INHdoesnothaveaneffectontheMTBstrains,whilstmutationsin inhAindicatelow-levelresistancetoINHandthepossibilitytouse INH in thetreatment at a highdose. Mutationat codon S315T resultsinanenzymethatdoesnothavetheabilitytoactivateINH butmaintains50%ofitscatalase-peroxidaseactivity.Thus,the alteredcatalase-peroxidaseprovideshigh-levelresistancetoINH whileretainingalevelofoxidativeprotectionthatissufficientto enabletheorganismtomaintaindetoxifyingactivityagainsthost antibacterialradicals[23].

TheabsenceofanyisolatewithdoublemutationsinbothkatG and the inhA promoter regionis similar toprevious studiesin Ethiopia[3,10].However,thefindingisincontrasttoseveralother studyreports.DoublemutationsatbothkatGandinhAregulatory regionsinthesameisolatewerefoundinstudiesconductedatSt Peter’sTB SpecializedHospital,AddisAbaba, Ethiopia(2%) [16], South India (0.1%) [28] and Shanghai, China (1.1%) [31]. These findingssuggestthatgeneticstrainbackgroundmayinfluencethe level ofINH resistance conferred byparticular mutations:MTB lineage 2 (Beijing genotype) was associated with any drug resistanceandlineage1wasassociatedwithinhApromotercodon C15Tmutations[34].Previousstudies inEthiopiarevealedthat MTB lineages 1 and 2 were among the majority of lineages identified, and MTBisolates that demonstratedresistancetoat leastoneofthetesteddrugsbelongedtoMTBlineage2[20,35].

Some studies indicated that most mutations linked to FQ resistance occurred in the gyrA gene (codons 90–94) [27,36].

Likewise, the most frequent FQ resistance-conferring mutation wasobservedatgyrAcodonD94Y/N(100%)inthepresentstudy.

WefoundanequalfrequencyofmutationsatgyrAcodonD94Gand codonA90V.Consistentwithourresults,otherstudiesshowedthat themostfrequentmutationsthatconfersFQ resistancewereat codon94followedbycodon90[36–38].Unlikeourstudyﬁndings, Fig.4.Correspondingmutationpatternsofrifampicin(RIF)-andisoniazid(INH)-resistantMycobacteriumtuberculosis(MTB)isolatesbyGenoTypeMTBDRplusassay.Lanes1 and2,mutationsconferringRIFresistanceatrpoBWT8wherethebandsareabsentandthecorrespondingrpoBMUT3haveappeared,andINHresistanceatkatGwherethe WTbandisabsentandthecorrespondingkatGMUT1appeared,whichindicatesmultidrug-resistantMTBisolates.However,lanes3–10aresusceptibleisolatesforRIFand INHowingtothepresenceofallWTbandsandtheabsenceofallMUTbands.P=H37Rv(positivecontrolstrain).N=waterasanegativecontrol.WT,wild-type;MUT,mutant.

(7)

fewotherstudiesreportedthatA90V(57%)[21]andD91A(47.6%) [27]werethemostfrequentmutationsconferringFQresistance.

Previous studies showed signiﬁcant evidence regarding the associationbetweengyrAmutationsandFQ-resistant,MDR,pre- XDRandXDRisolates,suggestingthatmutationsingyrAcanactas asurrogatediagnosticmarkerforFQ-resistantandMDR-TBanda possibleindicatorofpre-XDR-TBorXDR-TB[27,37].

NoneoftheisolateshadmutationsinthegyrBgeneconferring resistancetoFQs,concordantwithstudyreportsfromBotswana [39]andMorocco[27].Conversely,astudyreportedtheoccurrence ofmutationsinbothgyrAandgyrBgenes[37].Asystematicreview byAvalosetal.andresearchperformedinHebeiProvince,China, statedthattherewerevariabilitiesinthefrequencyofmutationsin thegyrAgenegeographically[37,40].

Weweresurprisedtoobserveunexpectedmutationpatternsin rpoB,katGandgyrAgenes,whereboththeWTandcorresponding mutantbandappeared.Thisindicatesthepresenceofsusceptible and resistant isolates, which were characterised as a hetero- resistantpopulationofMTBstrains.Ahighfrequencyofhetero- resistancewasobservedatgyrA(codonsD94Y/N,D94GandA90V), followedbyrpoB(codonsS531LandH526Y)andkatG(S315T).This ﬁndingisinlinewithotherstudiesaroundtheworld[8,19,36].

Thisfindinghighlightedthattherewereisolateswithidentical mutations,whichmightindicatetheacquisitionofdrugresistance that typically confers a reduction in fitness cost. This may consequently contribute to the spread of drug-resistant TB in thepopulation.Geneticdiversityanalysisrevealedthat28.6%of drug-resistant MTB strains were clustered, which showed the existenceofclonaltransmissionbetweenasmallnumberofstrains [37].Therefore,earlydiagnosisandtreatmentwithdefinitiveanti- TB drugs and strengthening of the DOTS (directly-observed treatment, short-course) programme are imperative to prevent thetransmissionofacquireddrug-resistantMTBinthecommuni- ty.

5.Conclusion

Themostcommonresistance-conferringmutationstoRIF,INH andFQsoccurredatrpoBcodon531(70%),katGcodon315(78%) and gyrA codon94 (100%),respectively. Thestudyrevealedthe presenceofasignificantnumberofunknownmutationsinrpoB, katGandinhAgenes,whichwereassociatedwithdrugresistanceto RIFandINH.Hence,thesefindingshighlighttheneedforapplying advanced moleculartechniques (whole-genome sequencing) to help identify all genetic mutations that are relevant to drug resistance throughout the genome. Additionally, our findings suggest the need for conducting periodic surveillance of drug resistance-conferring mutations, the early diagnosis and treat- mentofTB,andscaling-upofdrugsusceptibilitytestingfacilitiesto preventandcontrolthetransmissionofdrug-resistantTBinthe community.

Funding

This study was ﬁnancially supported by an institutional collaborationphaseIVbetweentheNorwegianUniversityofLife Sciences(NMBU),MekelleUniversity(MU)andHawassaUniversi- ty(HU)[CRPO/CHS/PhD//MUNMBU/028/2010].Thefundingsource didnothaveanyinvolvementinthestudydesign,datacollection, analysisand interpretationof data,writingofthereportorthe decisiontosubmitthearticleforpublication.

Competinginterests Nonedeclared.

Ethicalapproval

EthicalapprovalwasobtainedfromMekelleUniversity,College ofHealthSciencesEthicalReviewandResearchCommittee[ERC 1438/2018],theMinistryofScienceandHigherEducation,Ethiopia [SHE/S.M/14.4/708/19] andthe RegionalCommitteefor Medical ResearchEthicsinEasternNorway(REKØst)[2018/1118/REKsør- øst A]. Written informed consent was secured from all study participantsbeforecommencementofthestudy.

Acknowledgments

TheauthorsarethankfultotheNorwegianVeterinaryInstitute (NVI), Tigray Health Research Institute(THRI), EthiopianPublic Health Institute(EPHI) and Armauer HansenResearch Institute (AHRI) for providing laboratory materials and reagents. The authorswouldalsoliketothankthestudyparticipantsfortheir willingnesstoparticipateinthestudy.Theauthorsthankallofthe datacollectorsaswellastheTigrayRegionalHealthBureau(TRHB) and all administrators of selected hospitals for facilitating the study.

References

[1]WorldHealthOrganization(WHO).Globaltuberculosisreport2018.Geneva, Switzerland:WHO;2018..[Accessed9December2020]https://apps.who.int/

iris/handle/10665/274453.

[2]CollF,McNerneyR,PrestonMD,Guerra-AssunçãoJA,WarryA,Hill-Cawthorne G,etal.Rapiddeterminationofanti-tuberculosisdrugresistancefromwhole- genomesequences.Genome Med 2015;7:51,doi:http://dx.doi.org/10.1186/

s13073-015-0164-0.

[3]BrhaneM,KebedeA,PetrosY.Moleculardetectionofmultidrug-resistant tuberculosisamongsmear-positivepulmonarytuberculosispatientsinJigjiga town, Ethiopia. Infect Drug Resist 2017;10:75–83, doi:http://dx.doi.org/

10.2147/IDR.S127903.

[4]TracevskaT,JansoneI,BrokaL,MargaO,BaumanisV.MutationsintherpoB andkatGgenesleadingtodrugresistanceinMycobacteriumtuberculosisin Latvia. J Clin Microbiol 2002;40:3789–92, doi:http://dx.doi.org/10.1128/

JCM.40.10.3789-3792.2002.

[5]VidyarajCK,ChitraA,SmitaS,MuthurajM,GovindarajanS,UsharaniB,etal.

Prevalenceofrifampicin-resistantMycobacteriumtuberculosisamonghuman- immunodeﬁciency-virus-seropositivepatientsandtheirtreatmentoutcomes.

J Epidemiol Glob Health 2017;7:289–94, doi:http://dx.doi.org/10.1016/j.

jegh.2017.09.002.

[6]SharmaBK,BhandariS,MaharjanB,ShresthaB,RajBanjarM.Rapiddetection of rifampicin and isoniazid resistant Mycobacterium tuberculosis using Genotype MTBDRplus assay in Nepal. Int Sch Res Notices 2014;2014:648294,doi:http://dx.doi.org/10.1155/2014/648294.

[7]BiadglegneF,TessemaB,RodloffAC,SackU.Magnitudeofgenemutations conferringdrugresistanceinMycobacteriumtuberculosisisolatesfromlymph nodeaspiratesinEthiopia.IntJMedSci2013;10:1589–94,doi:http://dx.doi.

org/10.7150/ijms.6806.

[8]AlelignA,ZewudeA,MohammedT,TolosaS,AmeniG,PetrosB.Molecular detectionofMycobacteriumtuberculosissensitivitytorifampicinandisoniazid inSouthGondarZone,NorthwestEthiopia.BMCInfectDis2019;19:343.

[9]GagneuxS,BurgosMV,DeRiemerK,EncisoA,MuñozS,HopewellPC,etal.

Impact of bacterial genetics on the transmission of isoniazid-resistant Mycobacteriumtuberculosis.PLoS Pathog2006;2:e61,doi:http://dx.doi.org/

10.1371/journal.ppat.0020061.

[10]TadesseM,AragawD, DimahB,EfaF,AbdellaK,KebedeW,etal. Drug resistance-conferringmutationsinMycobacteriumtuberculosisfrompulmo- nary tuberculosis patients in Southwest Ethiopia. Int J Mycobacteriol 2016;5:185–91,doi:http://dx.doi.org/10.1016/j.ijmyco.2016.02.009.

[11]FederalDemocraticRepublicofEthiopiaCentralStatisticalAgency.Population projectionofEthiopiaforallregionsatweredalevelfrom2014–2017.Addis Ababa,Ethiopia:CentralStatisticalAgency;2017.

[12]MinistryofHealthofEthiopia.Nationalcomprehensivetuberculosis,leprosy andTB/HIVtrainingmanualforhealthcareworkers.Participants’manual.

AddisAbaba,Ethiopia:MinistryofHealth;2016.

[13]WorldHealthOrganization(WHO).Mycobacteriologylaboratorymanual.1st ed.Geneva,Switzerland:WHO;2014.

[14]KentPT,KubicaGP.Publichealthmycobacteriology:aguideforthelevelIII laboratory.Atlanta,GA:USDepartmentofHealthandHumanServices,Centers forDiseaseControlandPrevention;1985.

[15]HainLifescience.GenoTypeMTBDRplusv.2.0. Moleculargeneticassayfor identiﬁcationoftheM.tuberculosiscomplexanditsresistancetorifampicin and isoniazid from clinical specimens and cultivated samples. Nehren, Germany:HainLifescience;2015.

(8)

[16]AbateD,TedlaY,MeressaD,AmeniG.Isoniazidandrifampicinresistance mutationsandtheireffectonsecond-lineanti-tuberculosistreatment,Addis Ababa,Ethiopia.IntJTubercLungDis2014;18:946–51.

[17]MinhNN,VanBacN,SonNT,LienVTK,HaCH,CuongNH,etal.Molecular characteristicsofrifampin-andisoniazid-resistantMycobacteriumtuberculosis strainsisolatedinVietnam.JClinMicrobiol2012;50:598–601,doi:http://dx.

doi.org/10.1128/JCM.05171-11.

[18]JouR,ChenHY,ChiangCY,YuMC,SuIJ.Geneticdiversityofmultidrug- resistantMycobacteriumtuberculosisisolatesandidentiﬁcationof11novel rpoBallelesinTaiwan.JClinMicrobiol2005;43:1390–4,doi:http://dx.doi.org/

10.1128/JCM.43.3.1390-1394.2005.

[19]Polu GP,MohammadShaik J, KantaKota NM,KarumanchiD, AllamUS.

AnalysisofdrugresistancemutationsinpulmonaryMycobacteriumtubercu- losisisolatesintheSoutherncoastalregionofAndhraPradesh,India.BrazJ InfectDis2019;23:281–90,doi:http://dx.doi.org/10.1016/j.bjid.2019.07.002.

[20]DamenaD,TolosaS,HailemariamM,ZewudeA,ChimusaR,MihretA,etal.

GeneticdiversityanddrugsusceptibilityproﬁlesofMycobacteriumtuberculo- sisobtainedfromSaintPeter’sTBSpecializedHospital,Ethiopia.PLoSOne 2019;14:e0218545,doi:http://dx.doi.org/10.1371/journal.pone.0218545.

[21]CavusogluC, HilmiogluS,Guneri S,Bilgic A,CengizC,SuleyhaH,et al.

CharacterizationofrpoBmutationsinrifampin-resistantclinicalisolatesof MycobacteriumtuberculosisfromTurkeybyDNAsequencingandlineprobe assay. J Clin Microbiol 2002;40:4435–8, doi:http://dx.doi.org/10.1128/

JCM.40.12.4435.

[22]AktasE,DurmazR,YangD,YangZ.Molecularcharacterizationofisoniazidand rifampin resistance of Mycobacterium tuberculosis clinical isolates from Malatya, Turkey. Microb DrugResist 2005;11:94–9, doi:http://dx.doi.org/

10.1089/mdr.2005.11.94.

[23]SomoskoviA,ParsonsLM,SalﬁngerM.Themolecularbasisofresistanceto isoniazid,rifampin,andpyrazinamideinMycobacteriumtuberculosis.Respir Res2001;2:164–8,doi:http://dx.doi.org/10.1186/rr54.

[24]MaharjanB,NakajimaC,IsodaN,ThapaJ,PoudelA,ShahY,etal.Genetic diversityanddistributiondynamics ofmultidrug-resistantMycobacterium tuberculosisisolatesinNepal.SciRep2018;8:16634, doi:http://dx.doi.org/

10.1038/s41598-018-34306-w.

[25]ZawMT,EmranNA,LinZ.Mutationsinsiderifampicin-resistancedetermining regionofrpoBgeneassociatedwithrifampicin-resistanceinMycobacterium tuberculosis. J Infect Public Health 2018;11:605–10, doi:http://dx.doi.org/

10.1016/j.jiph.2018.04.005.

[26]BostanabadS,TitovL,BahrmandA,NojoumiS.Detectionofmutationin isoniazid-resistant Mycobacterium tuberculosis isolates from tuberculosis patientsinBelarus. IndianJMed Microbiol2008;26:143–7,doi:http://dx.

doi.org/10.4103/0255-0857.40528.

[27]OudghiriA,KarimiH,ChetiouiF,ZakhamF,BourkadiJE,ElmessaoudiMD,etal.

Molecularcharacterizationofmutationsassociatedwithresistancetosecond- linetuberculosisdrugamongmultidrug-resistanttuberculosispatientsfrom highprevalencetuberculosiscityinMorocco.BMCInfectDis2018;18:98,doi:

http://dx.doi.org/10.1186/s12879-018-3009-9.

[28]Muthaiah M, Shivekar SS, Cuppusamy Kapalamurthy VR, Alagappan C, SakkaravarthyA,BrammacharyU.Prevalenceofmutationsingenesassociated withrifampicinandisoniazidresistanceinMycobacteriumtuberculosisclinical

isolates.JClinTubercOtherMycobactDis2017;8:19–25,doi:http://dx.doi.org/

10.1016/j.jctube.2017.06.001.

[29]IsakovaJ,SovkhozovaN,VinnikovD,GoncharovaZ,TalaibekovaE,Aldasheva N,etal. MutationsofrpoB,katG,inhA andahp genesin rifampicinand isoniazid-resistant Mycobacterium tuberculosis in Kyrgyz Republic. BMC Microbiol2018;18:22,doi:http://dx.doi.org/10.1186/s12866-018-1168-x.

[30]SeifertM,CatanzaroD,CatanzaroA,RodwellTC.Geneticmutationsassociated withisoniazidresistanceinMycobacteriumtuberculosis:asystematicreview.

PLoS One 2015;10:e0119628, doi:http://dx.doi.org/10.1371/journal.

pone.0119628.

[31]CaoX,ZhanQ,GuoY,YangJ,LiuY,WanB,etal.Genotypiccharacterizationof katG,inhA,andahpCinisoniazid-resistantMycobacteriumtuberculosisclinical isolatesinShanghai,China.JundishapurJMicrobiol2020,doi:http://dx.doi.

org/10.5812/jjm.95713.

[32]AndoH,KondoY,SuetakeT,ToyotaE,KatoS,MoriT,etal.IdentiﬁcationofkatG mutationsassociatedwithhigh-levelisoniazidresistanceinMycobacterium tuberculosis. AntimicrobAgents Chemother2010;54:1793–9,doi:http://dx.

doi.org/10.1128/AAC.01691-09.

[33]PymAS,Saint-JoanisB,ColeST.EffectofkatGmutationsonthevirulenceof Mycobacteriumtuberculosisandtheimplicationfortransmissioninhumans.

InfectImmun2002;70:4955–60,doi:http://dx.doi.org/10.1128/IAI.70.9.4955- 4960.2002.

[34]FennerL,EggerM,BodmerT,AltpeterE,ZwahlenM,JatonK,etal.Effectof mutation and genetic background on drug resistance in Mycobacterium tuberculosis.AntimicrobAgentsChemother2012;56:3047–53,doi:http://dx.

doi.org/10.1128/AAC.06460-11.

[35]NuruA,MamoG,WorkuA,AdmasuA,MedhinG,PieperR,etal.Genetic diversityofMycobacteriumtuberculosiscomplexisolatedfromtuberculosis patientsinBahirDarCityanditssurroundings,NorthwestEthiopia.Biomed ResInt2015;2015:174732,doi:http://dx.doi.org/10.1155/2015/174732.

[36]GardeeY,DreyerAW,KoornhofHJ,OmarSV,SilvaP,BhyatZ.Evaluationofthe GenoTypeMTBDRslversion2.0assayforsecond-linedrugresistancedetection of Mycobacterium tuberculosis isolates in South Africa. J Clin Microbiol 2017;55:791–800,doi:http://dx.doi.org/10.1128/JCM.01865-16.

[37] LiQ,GaoH,ZhangZ,TianY,LiuT,WangY,etal.Mutationandtransmission proﬁlesofsecond-linedrugresistanceinclinicalisolatesofdrug-resistant Mycobacterium tuberculosis from Hebei Province, China. Front Microbiol 2019;10:1838,doi:http://dx.doi.org/10.3389/fmicb.2019.01838.

[38]LiJ,GaoX,LuoT,WuJ,SunG,LiuQ,etal.AssociationofgyrA/Bmutationsand resistancelevelstoﬂuoroquinolonesinclinicalisolatesofMycobacterium tuberculosis.EmergMicrobesInfect2014;3:e19,doi:http://dx.doi.org/10.1038/

emi.2014.21.

[39]MogashoaT,MelamuP,DerendingerB,LeySD,StreicherEM,IketlengT,etal.

DetectionofsecondlinedrugresistanceamongdrugresistantMycobacterium tuberculosisisolatesinBotswana.Pathogens2019;8:208,doi:http://dx.doi.org/

10.3390/pathogens8040208.

[40]AvalosE,CatanzaroD,CatanzaroA,GaniatsT,BrodineS,AlcarazJ,etal.

FrequencyandgeographicdistributionofgyrAandgyrBmutationsassociated with ﬂuoroquinolone resistance in clinical Mycobacterium tuberculosis isolates:a systematic review.PLoS One 2015;10:e0120470,doi:http://dx.

doi.org/10.1371/journal.pone.0120470.