7/28/2019 J. Clin. Microbiol.-1997-Lffler-3311-2

1/3

1997, 35(12):3311.J. Clin. Microbiol.

J Lffler, H Hebart, U Schumacher, H Reitze and H Einseleblood.of DNA of fungal pathogens from cultures andComparison of different methods for extraction

http://jcm.asm.org/content/35/12/3311Updated information and services can be found at:

These include:

CONTENT ALERTS

morecite this article),Receive: RSS Feeds, eTOCs, free email alerts (when new articles

http://journals.asm.org/site/misc/reprints.xhtmlInformation about commercial reprint orders:http://journals.asm.org/site/subscriptions/To subscribe to to another ASM Journal go to:

http://jcm.asm.org/cgi/alertshttp://jcm.asm.org/cgi/alertshttp://jcm.asm.org/cgi/alerts

7/28/2019 J. Clin. Microbiol.-1997-Lffler-3311-2

2/3

JOURNAL OF CLINICAL MICROBIOLOGY,0095-1137/97/$04.000

Dec. 1997, p. 33113312 Vol. 35, No. 12

Copyright 1997, American Society for Microbiology

Comparison of Different Methods for Extraction of DNA ofFungal Pathogens from Cultures and Blood

J. LOFFLER,1* H. HEBART,1 U. SCHUMACHER,2 H. REITZE,1 AND H. EINSELE1

Abt. Innere Medizin II, Medizinische Klinik und Poliklinik,1

and Abt. Medizinische Mikrobiologie,Hygieneinstitut,2 Eberhard-Karls-Universitat, 72076 Tubingen, Germany

Received 9 June 1997/Returned for modification 8 July 1997/Accepted 5 September 1997

Five commercially available extraction kits and an in-house DNA extraction method for the release of DNAfrom Candida albicans and Aspergillus niger cells were assessed for sensitivity, purity, duration, and cost. Allcommercially available kits helped to shorten the duration of DNA extraction. However, the sensitivity variedfrom 1 to 1,000 fungal cells/ml and costs varied from $0.10 to 2.30. The QIAmp Tissue kit was the commerciallyavailable assay that yielded the same sensitivity and purity of fungal DNA release as the in-house protocol but

was the most expensive method. In comparing these two extraction protocols, a 99% concordance of PCRresults for 125 blood samples analyzed could be demonstrated.

Invasive fungal infection is an increasing cause of morbidityand mortality in the immunocompromised patient. As culture-based detection methods show low sensitivity and poor speci-ficity, PCR-based assays have been recently developed to am-plify DNA of fungal pathogens (1, 7, 8).

In order to perform a sensitive, specific, and reliable PCR-based diagnostic test, availability of pure DNA lacking PCRinhibitors as well as a rapid and easy-to-perform DNA extrac-tion protocol is essential.

Protocols for extraction of DNA of fungal cells either arevery time-consuming (3) or show poor release of fungal DNA(10) compared to methods of extraction of DNA of humancells or viruses. Other protocols require additional lysis stepslike mechanical disruption or sonification or awkward toxicchemicals such as phenol-chloroform (4) or guanidine thiocy-anate (9).

Zymolyase, a -1,3-glucan laminaripentaohydrolase, may be

used for converting cells to spheroplasts. The spectrum of thisenzyme includes Candida, Torulopsis, and Saccharomycesstrains (6). To release DNA of filamentous fungi such as As-

pergillus niger, further lysis steps are necessary. Thus, mechan-ical destruction with glass beads (5) and freeze-thaw steps withliquid nitrogen or a heat alkali treatment have been success-fully applied.

Because of daily clinical practice, we had to speed up ourconventional DNA extraction method (2). Thus, we compareddifferent DNA extraction kits to determine their sensitivity,reliability, cost per sample, and duration for DNA release ofthe most common pathogenic yeast, Candida albicans, and A.

niger, a mold with a very complex cell wall. The method givingthe best results in the tests was adapted for the extraction ofDNA from clinical blood samples.

Prior to DNA extraction, Aspergillus isolates were precul-tured on Sabouraud agar for 72 h at 30C and Candida isolateswere precultured for 48 h at 30C. Plates were washed with 10ml of saline (0.9%) to obtain Aspergillus conidiae. Candidacolonies were inoculated in sterile saline. Thereafter, fungalsuspensions were adjusted photometrically (A530; McFarlandstandard of 0.5) to a concentration of 1 106 to 5 106

cells/ml. Defined numbers of serially diluted fungal cells (106

to 100) were spiked into uninfected EDTA (anticoagulant)-treated blood samples to test the sensitivity of the assays.

As many pathogenic fungi (e.g., A. niger) are resistant to lysisenzymes, we established a combined heat-alkali enzymaticDNA extraction protocol. After the hypotonic lysis of theerythrocytes with erythrocyte lysis buffer (10 mM Tris [pH 7.6],5 mM MgCl2, 10 mM NaCl) and enzymatic lysis of the leuko-cytes with leukocyte lysis buffer (10 mM Tris [pH 7.6], 10 mMEDTA, 50 mM NaCl, 0.2% sodium dodecyl sulfate, 200 g ofproteinase K [Boehringer, Mannheim, Germany] per ml), thesamples were incubated with 50 mM NaOH at 95C for 10 min.Samples were neutralized with 1 M Tris, pH 7.0. In order toproduce spheroplasts, pellets were incubated with 500 l ofZymolyase solution (300 g of Zymolyase [ICN, Costa Mesa,Calif.] per ml, 50 mM Tris [pH 7.5], 10 mM EDTA, 28 mM-mercaptoethanol [Sigma, Deissenhofen, Germany]) for 45

min at 37C. After an additional centrifugation step, the dif-ferent DNA extraction kits (QIAmp Tissue [Qiagen, Los An-geles, Calif.], GeneReleaser [BioVentures, Murfeesboro,Tenn.], Puregene D 6000 [Gentra, Minneapolis, Minn.], Dyna-beads DNA DIRECT [Dynal, Oslo, Norway] and DNAzol[Molecular Research Center, Cincinnati, Ohio]) were usedaccording to the manufacturers manual. Samples extracted bythe in-house method were treated with 1 M Tris-EDTA10%sodium dodecyl sulfate at 65C for 45 min for plast lysis. Then,samples were incubated with 5 M potassium acetate for 40 minat 20C for protein precipitation. After an additional centrif-ugation step (1,000 g, 30 min), DNA precipitation was car-ried out with cold isopropanol. DNA was purified with 70%ethanol, air dried for 1 h, and resuspended in 40 l of ddH2O.

After DNA extraction, spectral photometry was performed to

measure DNA content and purity. PCR was carried out withuniversal fungal primers binding within the 18S subunit rRNAmulticopy gene (5-ATTGGAGGGCAAGTCTGGTG, 5-CCGATCCCTAGTCGGCATAG) as published previously (2).

Amplification reactions were performed with standard buffers.PCR products were transferred onto nylon membranes (Am-ersham, Braunschweig, Germany) by Southern slot blot tech-nique. A species-specific probe for C. albicans and a genus-specific probe detecting most of the pathogenic Aspergillus spp.(30 pM each) were 5 labeled with digoxigenin (Roth,Karlsruhe, Germany). After hybridization, PCR products weredetected by antidigoxigenin-alkaline phosphatase (150 mU/ml;

* Corresponding author. Mailing address: Medizinische Klinik, Abt.II, Labor Dr. med. H. Einsele, Otfried-Mueller-Str. 10, 72076 Tu-bingen, Germany. Phone: 49 7071 2987355. Fax: 49 7071 293179.

3311

7/28/2019 J. Clin. Microbiol.-1997-Lffler-3311-2

3/3

Boehringer) and visualized by nitroblue tetrazolium and brom-chlorindoylphosphate (Boehringer).

Extractions and PCR were repeated five times. To check thereliability of the assays and the presence of inhibitors, a 129-bpfragment from the human DR- gene was coamplified (2).

The use of commercial kits did shorten the extraction pro-cedure, but quality and quantity of released DNA varied

widely. Table 1 compares the sensitivity, duration, and cost persample of the five different kits and the in-house protocol.

QIAmp Tissue and GeneReleaser proved to be as sensitiveas our in-house protocol (Fig. 1). All other kits lacked sensi-tivity for DNA compared to our in-house method. Kits result-ing in a DNA release of 2 to 3 log units below our standardmethod are not sensitive enough for reliable detection of lowamounts of fungal DNA in patients with febrile neutropenia.

We performed our test with C. albicans cells, as this is themost common pathogenic yeast in neutropenic patients. Fur-thermore, extraction of DNA of the filamentous fungus A.

nigerwas performed. Previous tests showed that DNA releasefrom A. niger, especially, is very difficult to obtain.

Smears in the agarose gel occurred when GeneReleaser orDNAzol was used, indicating that the samples contained im-pure DNA. This could potentially lead to a lower sensitivity in

the PCR assay or to false-negative samples because of inhibi-tion of DNA polymerization. In contrast, QIAmp Tissue

yielded good sensitivity and pure DNA. By using internal con-

trols to check the presence of PCR inhibitors, no false-negativeresults were obtained.

In order to perform the PCR technique routinely, low costsper assay are very important. We established our PCR assay

with a volume of 3 ml of blood, as this comparatively largeamount of blood might help to increase the sensitivity of theassay because of the higher yield of fungal DNA. Therefore, wetook 3 ml as the uniform level for all calculations. Thus, higher

costs than those calculated by the manufacturers occurred,especially with QIAmp Tissue; we calculated costs 23 timeshigher than those for a conventional method.

The kit resulting in highest sensitivity and purity wasadapted in order to test its value in clinical routine. A total of95 EDTA (anticoagulant)-treated blood samples from a se-lected group of nine neutropenic patients with proven or pre-sumed invasive fungal infection were tested. Moreover, 30blood samples from 10 healthy control persons were taken asnegative controls.

Blood samples were cultured by bedside inoculation intoBACTEC fungal medium and were tested daily for microbialgrowth by infrared detection (BACTEC 860; Becton Dickin-son). All blood samples were culture negative.

DNA extraction was performed by our in-house protocol

and with the QIAmp spin columns in parallel. All 30 bloodsamples from the 10 negative controls showed a negative PCRresult in both assays, irrespective of the DNA extraction pro-tocol applied.

Of the 95 samples from neutropenic patients, 52 were pos-itive and 43 were negative by the in-house method, and 51 werepositive and 44 were negative when QIAmp Tissue was used.Thus, a 99% correspondence between the two tests could bedocumented.

We thank T. Rogers, Hammersmith Hospital, London, United King-dom, for critical review of the manuscript.

REFERENCES

1. Buchman, T. G., M. Rossier, and W. G. Merz. 1990. Detection of surgicalpathogens by in vitro DNA amplification. Part I: rapid identification ofCandida albicans by in vitro amplification of a fungus-specific gene. Surgery

108:338347.2. Einsele, H., H. Hebart, G. Roller, J. Loffler, I. Rothenhofer, C. Mulle r, R.

Bowden, J.-A. van Burik, D. Engelhard, L. Kanz, and U. Schumacher. 1997.Detection and identification of fungal pathogens in blood by using molecularprobes. J. Clin. Microbiol. 35:13531360.

3. Gabal, M. 1992. Development of a chromosomal DNA probe for the labo-ratory diagnosis of aspergillosis. Mycopathologia 106:121129.

4. Haynes, K., T. Westerneng, J. Fell, and W. Moens. 1996. Rapid detectionand identification of pathogenic fungi by polymerase chain reaction ampli-fication of large subunit ribosomal DNA. J. Med. Vet. Mycol. 33:319325.

5. Hopfer, R., P. Walden, S. Setterquist, and W. Highsmith. 1993. Detectionand differentiation of fungi in clinical specimen using polymerase chainreaction (PCR) amplification and restriction enzyme analysis. J. Med. Vet.Mycol. 31:6575.

6. Kitamura, K., T. Kaneko, and Y. Yamamoto. 1971. Lysis of viable yeast cellsby enzymes of Arthrobacter luteus. Arch. Biochem. Biophys. 145:402404.

7. Makimura, K., S. Y. Murayama, and H. Yamaguchi. 1994. Detection of awide range of medically important fungi by the polymerase chain reaction.J. Med. Microbiol. 40:358364.

8. Melchers, W. J. G., P. E. Verweij, P. van den Hurk, A. van Belkum, B. E. DePauw, J. A. A. Hoogkamp-Korstanje, and J. F. G. M. Meis. 1994. Generalprimer-mediated PCR for detection ofAspergillus species. J. Clin. Microbiol.32:17101717.

9. Sandhu, G., B. Kline, L. Stockman, and G. Roberts. 1995. Molecular probesfor diagnosis of fungal infections. J. Clin. Microbiol. 33:29132919.

10. Tang, C., D. Holden, A. Aufauvre-Brown, and J. Cohen. 1993. The detectionof Aspergillus spp. by the polymerase chain reaction and its evaluation inbronchoalveolar lavage fluid. Am. Rev. Respir. Dis. 148:13131317.

TABLE 1. Comparison of different extraction methods

Method Sensitivitya Durationb Cost/samplec

In-house method 110 8 h 0.10QIAmp Tissue 110 4 h 2.30GeneReleaser 10 3 h, 15 min 1.80Puregene D 6000 100 5 h 0.80Dynabeads 100 4 h 1.10

DNAzol 1,000 4 h 1.00a Number of fungal cells detected per milliliter of blood after Southern blot

technique with species- or genus-specific digoxigenin-labeled oligonucleotide.b Total average duration of preparation of 12 blood samples for DNA extrac-

tion.c Cost in U.S. dollars calculated for one sample after Zymolyase treatment.

FIG. 1. Sensitivity testing for C. albicans by hybridization with species-spe-cific probe. DNA extraction was performed by six different procedures. Ampuwa,double-distilled H2O (negative control).

3312 NOTES J. CLIN. MICROBIOL.