S

Sa

SMAa

b

c

d

e

a

ARRA

KCMGDPT

1

ivpG

cirac

1h

Journal of Clinical Virology 54 (2012) 359– 363

Contents lists available at SciVerse ScienceDirect

Journal of Clinical Virology

j ourna l ho mepage: www.elsev ier .com/ locate / j cv

hort communication

ensitive and rapid detection of ganciclovir resistance by PCR based MALDI-TOFnalysis

amuel Zürchera, Catherine Moosera, Alexander Urs Lüthia, Kathrin Mühlemanna,aria Teresa Barbania, Paul Mohacsib, Christian Garzonia,c, Meri Gorgievski-Hrisohoa,∗,

ndré Schallerd,f, Lukas Flatza,e,f

Institute for Infectious Diseases, University of Bern, 3010 Bern, SwitzerlandDepartment of Cardiology, University Hospital Bern (Inselspital), 3010 Bern, SwitzerlandDepartment of Infectious Diseases, University Hospital Bern (Inselspital), 3010 Bern, SwitzerlandDivision of Human Genetics, Department of Pediatrics, University Hospital Bern (Inselspital), 3010 Bern, SwitzerlandDepartment of Dermatology, University Hospital-CHUV, 1011 Lausanne, Switzerland

r t i c l e i n f o

rticle history:eceived 27 January 2012eceived in revised form 22 April 2012ccepted 27 April 2012

eywords:MVALDI-TOFanciclovirrug resistanceEXransplantation

a b s t r a c t

Background: Cytomegalovirus (CMV) infection is associated with significant morbidity and mortality intransplant recipients. Resistance against ganciclovir is increasingly observed. According to current guide-lines, direct drug resistance testing is not always performed due to high costs and work effort, even whenresistance is suspected.Objectives: To develop a more sensitive, easy applicable and cost-effective assay as proof of concept fordirect drug resistance testing in CMV surveillance of post-transplant patients.Study design: Five consecutive plasma samples from a heart transplant patient with a primary CMV infec-tion were analyzed by quantitative real-time polymerase chain reaction (rtPCR) as a surrogate marker fortherapy failure, and by direct drug resistance detection assays such as Sanger sequencing and the novelprimer extension (PEX) reaction matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF)based method.

Results: This report demonstrates that PEX reaction followed by MALDI-TOF analysis detects the A594Vmutation, encoding ganciclovir resistance, ten days earlier compared to Sanger sequencing and morethan 30 days prior to an increase in viral load.Conclusion: The greatly increased sensitivity and rapid turnaround-time combined with easy handling andmoderate costs indicate that this procedure could make a major contribution to improve transplantationoutcomes.

. Background

CMV infection is still a major cause of morbidity and mortal-ty in transplant recipients.1 Ganciclovir (GCV) or its oral form,

alganciclovir, is the medication of choice for treatment and pro-hylaxis of CMV infection after transplantation.2 Resistance againstCV may develop through a single point mutation in the viral

Abbreviations: A594V, alanine-594-valine; b.i.d., twice a day; CMV,ytomegalovirus; GCV, ganciclovir; MALDI-TOF, matrix-assisted laser desorptiononization time-of-flight; PEX, primer extension; rtPCR, real-time polymerase chaineaction; Tris–EDTA, Tris(hydroxymethyl)aminomethane–ethylenediaminetetra-cetic acid; UL97, phosphotransferase of cytomegalovirus; UL54, polymerase ofytomegalovirus; VL, viral load.∗ Corresponding author. Tel.: +41 31 632 35 62; fax: +41 31 632 49 66.

E-mail address: meri.gorgievski@ifik.unibe.ch (M. Gorgievski-Hrisoho).f These authors contributed equally to this work.

386-6532/$ – see front matter © 2012 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.jcv.2012.04.019

© 2012 Elsevier B.V. All rights reserved.

phosphotransferase (UL97) or polymerase (UL54).1,3 Drug resis-tance can be suspected if clinical disease occurs and CMV viral load(VL) does not decline under prolonged therapy. Subsequently, achange of medication (decreased immunosuppressive therapy withhigh dose GCV or additional foscarnet) or genotypic drug resistancetesting should be performed according to current guidelines.4,5

Sanger sequencing is still the most common technique for geno-typic drug resistance testing. Phenotyping methods are rarely usedin diagnostics due to a lengthy turnaround time and high laborrequirements.6,7

2. Objective

The aim of this study was to develop a more sensitive, easilyapplicable and economic method for early detection of the drugresistance encoding mutation A594V as a proof of concept for CMVsurveillance in a post-transplant patient.

360 S. Zürcher et al. / Journal of Clinical Virology 54 (2012) 359– 363

Fig. 1. Primer extension (PEX) reaction is based on the same principle as Sanger sequencing. However, the extension reaction of the reverse PEX primer is terminatedafter just one (wild-type) or two nucleotides (mutant) by a ddNTP (dideoxynucleotide). Analysis of more than one mutation could be performed by multiplexing the PEXr e desiP es) cat

3

3

s(f(wclGca

3

Fpwep3C(paaPwa

3

mTT

concentration was best for improving sensitivity while maintainingspecificity11 (Fig. 1).

eaction and MALDI-TOF detection. For each mutation a specific PEX primer had to bEX primer + 1 nucleotide) and mutant product (reverse PEX primer + 2 nucleotidime-of-flight) analysis.

. Study design

.1. Patient

The study patient was a 55-year old female who had undergonetem cell transplantation in August 2008 and a heart transplantHTX) in September 2009 for amyloidosis. She was routinely testedor CMV load by quantitative real-time polymerase chain reactionrtPCR). Five consecutive plasma samples (days 0, 10, 20, 40 and 55)ere also analyzed genotypic while developing the GCV resistance

ausing mutation A594V in March 2010. Day 0 was defined as theast sample showing genotypic wild-type constitution for A594V.enotyping of UL97 was performed by Sanger sequencing andompared to the novel PEX/MALDI-TOF (primer extension/matrix-ssisted laser desorption ionization time-of-flight) based assay.

.2. PCR amplification

DNA was extracted using NucliSENS easyMAG (bioMérieux,rance). Amplification of CMV UL97 was performed as describedreviously8 with the following modifications. The original methodas based on sequencing of two overlapping PCR products. We

stablished a new PCR (supplemental data) combining the forwardrimer of the first amplicon (5′-TGGCCGACGCTATCAAATTT-′) with the reverse primer of the second amplicon (5′-GACACGAGGACATCTTG-3′) using HotStarTaq DNA PolymeraseQIAGEN, Switzerland). The PCR product (codons 439–645) encom-asses the so-called “canonical” mutations at codons 460, 520nd 592–603 which confer GCV resistance.9 PCR products werenalyzed by gel electrophoresis and purified by QIAquick PCRurification Kit (QIAGEN, Switzerland). The purified ampliconsere then subjected to Sanger sequencing and PEX/MALDI-TOF

nalysis.

.3. Sanger sequencing

Cycle sequencing reaction was performed using BigDye Ter-inator v3.1 kit (Life Technologies Europe BV, Switzerland).

he sequence was acquired on a genetic analyzer AB-3130 (Lifeechnologies Europe BV, Switzerland). Electropherograms were

gned. Mass differences between the reverse PEX primer, wild-type product (reversen be easily detected by MALDI-TOF (matrix-assisted laser desorption ionization

processed using SeqMan (DNASTAR, USA) and then ana-lyzed in silico by an open access interpretation tool athttp://www.informatik.uni-ulm.de/ni/staff/HKestler/hcmv.10

3.4. PEX reaction

The five plasma samples (days 0–55) previously analyzedby Sanger sequencing were retrospectively tested by the novelPEX/MALDI-TOF based assay. PEX reaction was performed withthe reverse PEX primer (5′-CTTGCCGTTCTCCAAC-3′), using TER-MIpol DNA Polymerase (Solis BioDyne, Estonia) (supplementaldata). Products were purified using Micro Bio-Spin 6 chromatog-raphy columns in Tris–EDTA (BioRad, USA). A high PEX primer

Fig. 2. Course of CMV load measured by quantitative real-time polymerase chainreaction (rtPCR). x-Axis: time span of CMV observation, y-axis: CMV load incopies/ml. Samples where genotyping was performed are indicated by *: (days 0,10, 20, 40 and 55).

S. Zürcher et al. / Journal of Clinical Virology 54 (2012) 359– 363 361

Fig. 3. Comparison of five consecutive plasma samples of a patient with a primary CMV (cytomegalovirus) infection while developing the mutation A594V (GCG to GTG)causing ganciclovir resistance by Sanger sequencing (column A), and by PEX/MALDI-TOF analysis (primer extension/matrix-assisted laser desorption ionization time-of-flight) (column B). Day 0 was defined as the last wild-type only constitution for A594V analyzed by Sanger sequencing and the novel PEX/MALDI-TOF based method. Thex-axis represents the molecular mass over charge while the y-axis shows the relative amount of detected ions. Expected molecular masses are: reverse PEX primer 4814.4 Da,wild-type product 5129.5 Da (reverse PEX primer + 1 nucleotide) and mutant product 5404.5 Da (reverse PEX primer + 2 nucleotides). Day 47 shows similar patterns as day40 (data not shown).

3 linical

3

Mmawwa

4

m(ttufTec2tticw

ufipddpd

5

atsrlm(

slbUirpa

tmtcwstt

62 S. Zürcher et al. / Journal of C

.5. MALDI-TOF analysis

Mass spectrometry analysis was performed on a MicromassALDI-TOF micro MX (Waters, USA) operated in positive linearode. A mixture of 0.3 M 3-hydropicolinic acid, 0.5 M picolinic

cid and 0.3 M ammonium fluoride with a volume ratio at 9:1:1as used as matrix.12 1 �L of purified PEX product and 1 �L matrixere mixed together, applied on a stainless steel target plate and

ir-dried at room temperature.

. Results

The course of the CMV load shows a 10-fold increase over threeonths from December 2009 (4.5 log10 copies/ml) to mid March

5.4 log10 copies/ml) followed by a severe increase of 100-foldo (7.4 log10 copies/ml) in only two weeks (Fig. 2). GCV resis-ance was suspected and the mutation A594V was confirmedsing Sanger sequencing. Therapy was subsequently changed tooscarnet, according to current international recommendations.5

he sequential Sanger sequencing electropherograms showed themergence of the mutation A594V (GCG to GTG) over time (Fig. 3,olumn A). On day 0 and day 10, no mutation was detectable. On day0, the first appearance of a minor subpopulation with the muta-ion A594V became evident as a small peak of thymidine (red) athe same position as the high peak of cytidine (blue) represent-ng the predominant wild-type virus. On day 40, the resistanceausing mutation (GTG) was the predominant form. On day 55, theild-type virus (GCG) had almost completely disappeared.

By analyzing the PEX/MALDI-TOF generated spectra (Fig. 3, col-mn B), the presence of the mutant population at day 10 as arst small peak representing the mutant PEX product (reverse PEXrimer extended by two nucleotides) was detectable in the repro-uction. On day 40, the mutant population was predominant. Onay 55, the wild-type population was still represented by a cleareak, whereas using Sanger sequencing, the wild-type was barelyetectable.

. Discussion

Early detection of evolving drug resistance is essential for timelynd adequate adaptation of therapy in the management of post-ransplant patients.5,6 VL monitoring is now routinely used as aurrogate marker to detect therapy failure. The method is fast,elatively inexpensive and well established in most diagnosticaboratories, but gives neither information about the underlying

echanism of drug resistance mutation nor of its level of resistanceIC50 ratio).

Direct drug resistance testing is usually performed using Sangerequencing.6 However, this method requires the presence of ateast 20–30% of the viral quasispecies in order to reliably and unam-iguously detect the mutation.1 In addition, Sanger sequencing ofL97 and UL54 is time consuming, labor- and cost-intensive requir-

ng the separate analysis of at least three products. Hence, it is notoutinely performed, even in cases where drug resistance is sus-ected. This results in a major delay in drug resistance detectionnd adequate adaptation of therapy.

Alternative genotyping methods are available, but have limita-ions and are not commonly in use: for example rtPCR for direct

utation analysis lacks the ability to detect all clinically impor-ant mutations in UL97 and UL54 in a single assay.13 Results of aommercial MALDI-TOF based sequencing method were published

ith comparable sensitivity, costs and turnaround time to Sanger

equencing.14 Current opinions on next generation sequencingechnologies are diverse. In a review from 2010, citing a subpopula-ion detection limit of <1%, this method is currently still considered

Virology 54 (2012) 359– 363

too expensive and time consuming.1 A recent study on pyrose-quencing of CMV UL97 claimed a detection limit of merely 6% withrapid data acquisition. However, the necessary work effort, specialtechnical skills and costs were not mentioned.15

A limitation of this novel PEX/MALDI-TOF based assay is thatonly known mutations can be detected. An advantage of this assayis its easy adaptation for including other important mutationsby adding specific PEX primers with no significant increase ofexpenses.

By applying the novel PEX/MALDI-TOF based method to a com-plex CMV clinical case with the known A594V mutation, we coulddemonstrate as a proof of concept that this assay has increased sen-sitivity while costs, work effort and turnaround time are similar tosequencing of just one amplicon.

While the mutation A594V was detected on day 10 by the novelPEX/MALDI-TOF based method, it was only on day 20 that Sangersequencing revealed the emergence of the mutation. QuantitativertPCR finally showed a relevant VL increase of one log on day 47.

In addition to the improved sensitivity of the novel PEX/MALDI-TOF assay, its outstanding applicability results in lower costs andhence the possibility to perform the test more frequently. Thisincreases the chance of timely detection of drug resistance withrapid and adequate adaptation of the therapy. Therefore the novelPEX/MALDI-TOF assay has the potential to be implemented in rou-tine surveillance following transplantation, with possible improvedpatient outcomes. Prospective clinical studies are still needed tofurther validate this novel method’s positive impact on clinicaloutcomes.

Funding

No funding.

Competing interests

No conflict of interest.

Ethical approval

Ethical approval was not required.

Acknowledgments

We thank Avril Lüthi, Louise Vilén Zürcher, Mary McKay Vilénand John Looney for thoughtful review of the manuscript.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.jcv.2012.04.019.

References

1. Lurain NS, Chou S. Antiviral drug resistance of human cytomegalovirus. ClinMicrobiol Rev 2010;23:689–712.

2. Humar A, Snydman D. Cytomegalovirus in solid organ transplant recipients. AmJ Transplant 2009;9:S78–86.

3. Chou S. Cytomegalovirus UL97 mutations in the era of ganciclovir and maribavir.Rev Med Virol 2008;18:233–46.

4. Boivin G, Goyette N, Rollag H, Jardine AG, Pescovitz MD, Asberg A, et al.Cytomegalovirus resistance in solid organ transplant recipients treated withintravenous ganciclovir or oral valganciclovir. Antivir Ther 2009;14:697–704.

5. Kotton CN, Kumar D, Caliendo AM, Asberg A, Chou S, Snydman DR, et al. Inter-national consensus guidelines on the management of cytomegalovirus in solid

organ transplantation. Transplantation 2010;89:779–95.

6. Drew WL. Cytomegalovirus resistance testing: pitfalls and problems for theclinician. Clin Infect Dis 2009;50:733–6.

7. Hamprecht K, Eckle T, Prix L, Faul C, Einsele H, Jahn G. Ganciclovir-resistantcytomegalovirus disease after allogeneic stem cell transplantation: pitfalls of

linical

1

1

1

1

1

S. Zürcher et al. / Journal of C

phenotypic diagnosis by in vitro selection of an UL97 mutant strain. J Infect Dis2003;187:139–43.

8. Castor J, Cook L, Corey L, Jerome KR. Rapid detection directly from patientserum samples of human cytomegalovirus UL97 mutations conferring ganci-clovir resistance. J Clin Microbiol 2007;45:2681–3.

9. Chou S, Waldemer RH, Senters AE, Michels KS, Kemble GW, Miner RC, et al.Cytomegalovirus UL97 phosphotransferase mutations that affect susceptibilityto ganciclovir. J Infect Dis 2002;185:162–9.

0. Chevillotte M, von Einem J, Meier BM, Lin FM, Kestler HA, Mertens T. A newtool linking human cytomegalovirus drug resistance mutations to resistance

phenotypes. Antiviral Res 2010;85:318–27.

1. Wise CA, Paris M, Morar B, Wang W, Kalaydjieva L, Bittles AH. A standardprotocol for single nucleotide primer extension in the human genome usingmatrix-assisted laser desorption/ionization time-of-flight mass spectrometry.Rapid Commun Mass Spectrom 2003;17:1195–202.

1

Virology 54 (2012) 359– 363 363

2. Schaller A, von Kanel T, Gehr B, Sanz J, Gallati S. Rapid and reliable genotypingof polymorphic loci modifying correct splicing of CFTR pre-mRNA using massspectrometry. Eur J Hum Genet 2007;15:53–61.

3. Gohring K, Mikeler E, Jahn G, Hamprecht K. Rapid simultaneous detec-tion by real-time PCR of cytomegalovirus UL97 mutations in codons460 and 520 conferring ganciclovir resistance. J Clin Microbiol 2006;44:4541–4.

4. Posthuma CC, van der Beek MT, van der Blij-de Brouwer CS, van derHeiden PL, Marijt EW, Spaan WJ, et al. Mass spectrometry-based com-parative sequencing to detect ganciclovir resistance in the UL97 gene of

human cytomegalovirus. J Clin Virol: Off Publ Pan Am Soc Clin Virol 2011;51:25–30.

5. Schindele B, Apelt L, Hofmann J, Nitsche A, Michel D, Voigt S, et al. Improveddetection of mutated human cytomegalovirus UL97 by pyrosequencing. Antimi-crob Agents Chemother 2010;54:5234–41.