OUP user menu

Contamination of diverse nifH and nifH-like DNA into commercial PCR primers

Masahiro Goto, Shotaro Ando, Yusuke Hachisuka, Tadakatsu Yoneyama
DOI: http://dx.doi.org/10.1016/j.femsle.2005.03.042 33-38 First published online: 1 May 2005


False-positive results due to DNA contamination in PCR reagents have become a big problem in the amplification of small amounts of DNA. Recently, it was revealed that PCR reagents were contaminated with the nifH (dinitrogenase reductase) gene. We found that the PCR primers supplied by some manufacturers contained nifH gene and nifH-like DNA. This contamination resulted in false-positive results when searching for nifH genes in environmental samples. The sequences of the contaminating DNA appeared to be widely varied in the phylogenetic analysis of nifH. For this reason, great care should be taken when analyzing trace amounts of nucleotides.

  • Contamination
  • nifH
  • PCR
  • Primer

1 Introduction

PCR is a convenient method for detecting small amounts of target DNA. However, false-positive results due to its high sensitivity have become a big issue, especially in diagnostic laboratories [[[]. This problem is often caused by contamination with DNA from various sources including PCR reagents [[]. Many attempts have been made to eliminate such contamination, but the procedures are complicated and their effects often insufficient [[[3].

PCR has been used extensively to detect nitrogen-fixing bacteria in DNA extracted from ecosystems such as the marine environment [[4], termite guts [[5], soil and litter [[6], and plant tissues [[7,[8]. Reverse transcription (RT)-PCR has also been used to detect the expression of the nifH gene in mRNA extracted from those ecosystems [[9[2]. However, Zehr et al. [[3] reported that PCR reagents such as Taq DNA polymerase, buffers, dNTPs, and primers supplied by manufacturers were routinely contaminated with nifH gene sequences. They tried to remove contaminants by restriction enzyme digestion and ultrafiltration, but complete elimination was not possible.

We have investigated endophytic diazotrophs by amplifying nifH DNA extracted from plant tissues, and in the meantime we encountered a serious contamination problem. Consequently, we examined the source of contaminants.

2 Materials and methods

2.1 Preparation of PCR mixture

To avoid cross-contamination and contamination from aerosol, PCR reaction mixtures were always prepared on a clean bench with dedicated micropipettors and MultiGuard Barrier Tips (Sorenson BioScience, Salt Lake City, UT, USA). Sample tubes were always kept closed except when needed.

2.2 PCR amplification and electrophoresis

We used the following DNA polymerases: TaKaRa Taq (TaKaRa BIO, Shiga, Japan), TaKaRa Ex Taq (TaKaRa BIO), AmpliTaq Gold DNA Polymerase, LD (Applied Biosystems, Foster City, CA, USA), and Taq DNA Polymerase (Roche Diagnostics, Basel, Switzerland). Only PCR buffers, MgCl2 and dNTPs which were attached to each DNA polymerase were used, and the compositions of the PCR reaction mixtures (total 50 μl) were as described in each manufacturer's protocol. RNase-free water supplied with the RNeasy plant mini kit (QIAGEN, Hilden, Germany) was used to dilute the reaction mixtures. Primers were custom-ordered from Invitrogen Japan (Tokyo, Japan), TaKaRa BIO, and PROLIGO Japan (Kyoto, Japan). Primers ordered from Invitrogen Japan and TaKaRa BIO were received in dry form and were dissolved in the same RNase-free water used for the PCR reaction mixture. Primers ordered from PROLIGO Japan were received as liquids dissolved in TE buffer. Primers and conditions of the first and nested second round PCR were as described [[8], but the number of cycles and the annealing temperature of the first round PCR were changed to 30 cycles and 58 °C. The nested second round PCR was performed with primers from the same manufacturer and the same DNA polymerase used in the first round PCR. Ten microlitres of each nested PCR product was electrophoresed on a 2% (w/v) agarose gel and stained with ethidium bromide.

2.3 Reverse transcription

Reverse transcription was conducted using the TaKaRa RNA PCR Kit (AMV) Ver. 3.0 (TaKaRa Bio). The 10 μl of reaction mixture contained 1 × PCR buffer, 5 mM MgCl2, 1 mM dNTP mixture, 2.5 μM random 9 mers, 10 U of RNase inhibitor, and 2.5 U of reverse transcriptase. The reverse transcription condition was 10 min at 30 °C, 25 min at 42 °C, and 5 min at 99 °C. We used 2 μl of reverse transcription product as the first round PCR template DNA, as described above.

2.4 Cloning and sequencing

Among the PCR tubes that resulted in false-positive responses, two RT-nested PCR products using Invitrogen primers (conI1, I2) and seven nested PCR products using several manufacturer's primers and DNA polymerases (conI3, I4, I5, I6, conT1, T2, T3) were selected. They were cloned into Escherichia coli cells using the pST-Blue1 AccepTor Vector Kit (Novagen, Madison, WI, USA) following the protocol of the manufacturer. Colony PCR was performed with randomly selected colonies using T7 and U-19 primers, which were complementary to the vector. The products were purified with the Montage PCR Filter Unit (MILLIPORE, Billerica, MA, USA) and sequenced with the 310 Genetic Analyzer (Applied Biosystems) following the manufacturer's protocol.

2.5 Phylogenetic analysis

The sequences of clones were submitted to the DDBJ nucleotide sequence database under Accession Nos. AB198366 to AB198391. The homology of clones to the recorded nifH sequences was searched based on the DDBJ DNA sequence database with the BLAST program [[4]. A phylogenetic tree was constructed from DNA sequences by the DNASIS Pro Ver.2.6 (Hitachi Software Engineering, Tokyo, Japan) with the neighbor-joining method [[5], and bootstrap analyses of 100 repeated samplings were performed.

3 Results and discussion

Our laboratory had encountered the problem of contamination before the report of Zehr et al. [[3]. First, we renewed all PCR reagents, water, disposables and instruments, but false-positive results were found on negative control lanes in electrophoresis gels. Secondly, we prepared reaction mixtures in other rooms using new equipment, but the negative control still showed PCR amplification, probably caused by contamination. Thirdly, we suspected manufacturing contamination of Taq DNA polymerase and tried various polymerases from different manufacturers, but all polymerases we used amplified contaminated DNA. However, an interesting result was obtained when we used primers made by certain manufacturers. Using primers obtained from Invitrogen, all negative control lanes showed a band on the agarose gel (Fig. 1(a)). When we used primers obtained from TaKaRa, one of six negative control lanes showed the band (Fig. 1 (b)), and when we used primers from PROLIGO, no negative control lanes showed the band (Fig. 1 (c)). Positive controls on the test of PROLIGO's primers were better amplified than those from other manufacturers (Fig. 1). All PCR reagents and other variables except the primers were the same during these tests, including the lot numbers of reagents.

Figure 1

Test of primers from three manufacturers. (a) Primers from Invitrogen Japan. (b) Primers from TaKaRa BIO. (c) Primers from PROLIGO Japan. No template: without template DNA; positive control: 10 ng DNA extracted from Bradyrhizobium japonicum USDA 110 was added.

Additional test repetitions confirmed that primers from Invitrogen showed false bands in all of the negative control lanes, while primers from TaKaRa showed, on average, false bands in about one quarter of the negative control lanes, and primers from PROLIGO showed no bands in the negative control lanes in electrophoresis gels (Table 1). All tests were performed with TaKaRa Ex Taq, which had shown the most favorable amplification efficiency during our observations. These results indicated that primers from Invitrogen and TaKaRa were seriously or slightly contaminated with DNA, which had complementary sequences with four primers designed from the NifH amino acid sequence [[8]. We do note that the contamination level may come and go over different lots. Primers used for the tests were all from the same lots. We searched other lots of Invitrogen and PROLIGO primers; however, we did not find any different result.

View this table:
Table 1

Frequency of false-positive results with primers from different manufacturers

Manufacturer of primersNumber of testsFalse-positive results
TaKaRa BIO185
  • One aliquot dispensed from a master reaction premix counted as one test. Results were summed for all repetitions of PCR amplification with TaKaRa Ex Taq.

Two RT-nested PCR products and seven nested PCR products using primers from Invitrogen and TaKaRa with several DNA polymerases without template RNA/DNA were cloned and sequenced. The clones are described in Table 2 based on their base sequences. Clones obtained from the same PCR product were all grouped into the same group. A phylogenetic tree (Fig. 2) was constructed with the sequences of the representative clones shown in Table 2, the nifH sequences of well-known diazotrophic bacteria, and contaminant sequences reported by Zehr et al. [[3].

View this table:
Table 2

Clones obtained from RT-nested PCR and nested PCR products

TypeClones which are similar to each otherRepresentative clonesManufacturer of primersDNA polymerase
Group IconI1-1, 2, 3, 4, conI2-1, 2, 3, 4conI1-1InvitrogenTaKaRa Taq
Group IIconI3-1, 2, conI6-1, 2conI3-1InvitrogenTaKaRa Taq (conI3- 1, 2), Roche Taq (conI6-1, 2)
Group IIIconI4-1, 2, 3, 4, conI5-1conI4-1InvitrogenTaKaRa Taq (conI4-1, 2, 3, 4), AmpliTaq Gold, LD (conI5-1)
Group IVconT1-1, 2, 3, 4conT1-1TaKaRa BIOTaKaRa Taq
Group VconT2-1, 2, 3conT2-1TaKaRa BIOTaKaRa Ex Taq
Group VIconT3-1, 2conT3-1TaKaRa BIORoche Taq
  • Clones that exceeded 99% base sequence similarity with each other were divided into groups and a representative clone was chosen from each group. The conI1 and conI2 clones were from RT-nested PCR products without the addition of template RNA. The other clones were from nested PCR products without template DNA.

Figure 2

Phylogenetic relationships among the sequences of representative clones listed in Table 2 and the nifH sequences from the DNA database. DNA database accession numbers are shown in parentheses. Rhodobacter capsulatus bchL and Plectonema boryanum frxC were used as the outgroup. The scle bar denotes 0.1 substitutions per site. Bootstrap values (percentages) for neighbor joining from 100 repeated samplings are shown for each node, with values less than 50% being omitted.

The sequences of group V clones were very different from known nifH sequences, and their amino acid sequences had less than 60% similarity with those of the recorded NifH. These clones may not be the nifH gene, but it is true that PCR reagents were contaminated with these DNA fragments. Sequences of clones obtained in this study were not matched with those of Zehr et al. [[3]; conT3-1 was the closest with 81–90% similarity. Sequences divided into six groups were highly diverse among known nifH genes (Fig. 2), suggesting that there were diverse contaminants in commercial PCR reagents.

Clones of groups II and III, whose sequences were almost the same in each group, were detected from two PCR reaction mixtures with DNA polymerases from different manufactures (Table 2). This implies that contaminants did not originate from DNA polymerases, attached buffers, or dNTPs, and that primers contaminated at the manufacturing step were the most likely candidate for the main source of nifH and nifH-like DNA contamination. However, this does not mean that primers are the only source of contamination. Considering that primers were contaminated with nifH, the same or similar kind of contamination in DNA polymerases, attached buffers, dNTPs, or RT-PCR kit components is highly possible, although we have so far been unable to find such contamination in these materials.

According to the result of a Blast search [[4], the sequences obtained in this study displayed high similarity with some sequences in the DDBJ DNA database. In particular, some partial nifH gene sequences of clones obtained from mine spoils (Accession No. AJ716424 and others) were almost the same as the sequences of group IV clones within the error range of PCR amplification. These clones might be derived from contamination. Needless to say, it is possible that the same bacteria were in both mine spoils and PCR reagents. It is difficult to ascertain whether they existed in the environment.

We also want to point out that the sequences of group II clones were nearly the same as the nifH gene sequences of Delftia tsuruhatensis strain HR4 (Accession No. AY544164), Klebsiella variicola strain F2R9 (Accession No. AY367392), and Klebsiella pneumoniae 342 (Accession No. AY242355). The sequences of group IV clones were also the same as the partial nifH gene sequences of Paenibacillus sp. g2 (Accession No. AY373368) and K. oxytoca clone CC1103A1 (Accession No. AY221827). Of course, it is possible that these bacteria existed in PCR reagents, or that some bacteria contaminating PCR reagents might possess the nifH gene of these bacteria, but it cannot be excluded that these sequences in the database might be adversely affected because of contamination. It is rather strange that three species (D. tsuruhatensis, K. variicola, and K. pneumoniae) and two species (Paenibacillus sp. and K. oxytoca) had the same nifH sequences, even if we take into consideration that nifH genes are sometimes transferred laterally [[6] and that some nitrogen-fixing bacteria possess multiple nifH and nifH-like genes [[5]. Interestingly, Zehr et al. [[3] detected the contamination of the 16S rRNA gene of Delftia sp. in PCR reagents.

It is well known that commercial DNA polymerases are contaminated E. coli and Thermus aquaticus DNA during their manufacture [[7]. As far as we know, these bacteria do not possess the nifH gene. However, PCR reagents, especially custom-ordered primers, are often contaminated with the nifH and nifH-like DNA. As Zehr et al. [[3] mentioned, bacterial contamination of ultrapure water might be the reason of the apparent contamination of primers. Bradyrhizobium sp., Burkholderia sp., Pseudomonas stutzeri, P. saccharophilia, and P. fluorescens were isolated from ultrapure water [[8,[9]. The results of PCR showed that Bradyrhizobium sp. and P. saccharophilia had the nifH gene [[9]. The others might also have a nifH gene [[0[2]. Kulakov et al. [[9] were able to isolate bacteria including nitrogen fixers from all ultrapure water systems used in their study. Ultrapure water might be contaminated with bacteria ubiquitously. Those bacteria would form biofilms and survive the oligotrophic environment [[9]. If the bacteria residing in ultrapure water induce the contamination of primers, its levels would be influenced by the synthesis time, even if they were made by the same manufacturer. Thus, primers should be tested at every purchase, even if they were made by a manufacturer that had previously produced non-contaminated primers.

Care should also be taken regarding the contamination of other PCR reagents. As described above, negative control lanes in an electrophoresis gel do not always show DNA bands, even though the aliquots of reaction mixture were dispensed from the same master premix (Fig. 1 (b)). We also recommend that researchers test primers with as many controls as possible. As Zehr et al. [[3] indicated, it is difficult to eliminate DNA contamination completely. We also tried DNase treatments and UV exposure of PCR reagents, but we could not decontaminate the mixtures completely without decreasing their amplification efficiency (data not shown). PCR is an excellent method for detecting extremely small amounts of DNA; however, we must take great care to determine whether false-positive results are being generated because of its overtly high sensitivity.


We thank Dr. S. Otsuka, Laboratory of Soil Cosmology, the University of Tokyo, for reading this manuscript and making helpful comments. We also thank Ms. R. Kugimiya and Dr. M. Hidaka, Laboratory of Molecular and Cellular Breeding, the University of Tokyo, for their useful advice.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
View Abstract