OUP user menu

Comparative in silico analysis of PCR primers suited for diagnostics and cloning of ammonia monooxygenase genes from ammonia-oxidizing bacteria

Pilar Junier, Ok-Sun Kim, Verónica Molina, Petra Limburg, Thomas Junier, Johannes F. Imhoff, Karl-Paul Witzel
DOI: http://dx.doi.org/10.1111/j.1574-6941.2007.00437.x 141-152 First published online: 1 April 2008


Over recent years, several PCR primers have been described to amplify genes encoding the structural subunits of ammonia monooxygenase (AMO) from ammonia-oxidizing bacteria (AOB). Most of them target amoA, while amoB and amoC have been neglected so far. This study compared the nucleotide sequence of 33 primers that have been used to amplify different regions of the amoCAB operon with alignments of all available sequences in public databases. The advantages and disadvantages of these primers are discussed based on the original description and the spectrum of matching sequences obtained. Additionally, new primers to amplify the almost complete amoCAB operon of AOB belonging to Betaproteobacteria (betaproteobacterial AOB), a primer pair for DGGE analysis of amoA and specific primers for gammaproteobacterial AOB, are also described. The specificity of these new primers was also evaluated using the databases of the sequences created during this study.

  • Ammonia-oxidizing bacteria
  • PCR primers
  • amoC
  • amoA
  • amoB


Ammonia-oxidizing bacteria (AOB) are chemolithoautotrophic Gram-negative proteobacteria that fix CO2 with the reducing power obtained from ammonia oxidation (Prosser, 1989). They belong to two monophyletic lineages: Nitrosomonas spp. (including Nitrosococcus mobilis) and Nitrosospira spp. (including Nitrosolobus and Nitrosovibrio) form a closely related clade within the beta phylum (betaproteobacterial AOB) of proteobacteria, whereas Nitrosococcus oceani is affiliated to the gamma phylum (gammaproteobacterial AOB) of proteobacteria (Head et al., 1993; Purkhold et al., 2000; Purkhold et al., 2003).

Characterization of the species composition and diversity of AOB communities in nature has been hampered for a long time by difficulties in the isolation and culture of these microorganisms. Analysis of AOB communities has become accomplishable by applying culture-independent molecular approaches, which are based on the amplification of 16S rRNA genes by PCR (Bothe et al., 2000; Kowalchuk & Stephen, 2001) or the detection of 16S rRNA by FISH (Wagner et al.., 1993, 1995; Mobarry et al.., 1996). 16S rRNA genes are good phylogenetic markers, but are not necessarily related to the physiology of the target organisms (Kowalchuk & Stephen, 2001; Calvo & Garcia-Gil, 2004). Therefore, functional markers such as the genes encoding for key enzymes involved in ammonia-oxidation provide an alternative in ecological studies (Rotthauwe et al.., 1997). Diversity studies of AOB based on the sequence analysis of one of these genes, amoA, have shown a high resolution in separating closely related strains (Rotthauwe et al.., 1997; Alzerreca et al.., 1999; Aakra et al.., 2001; Norton et al.., 2002).

Ammonia monooxygenase (AMO) is a membrane-bound multiple subunit enzyme responsible for the conversion of ammonia to hydroxylamine (Hyman & Arp, 1992). The structural subunits of AMO in AOB are encoded by the genes amoC, amoA and amoB, which are organized in one operon (Norton et al.., 2002). The physical organization of the operon seems to be conserved in all AOB; multiple copies have been reported for betaproteobacterial AOB (Norton et al.., 2002), whereas so far it seems that in gammaproteobacterial AOB occurs as a single copy (Alzerreca et al.., 1999).

Since the publication of the first amoA sequence of Nitrosomonas europaea (McTavish et al.., 1993), the number of partial and full-length sequences available in public databases has increased significantly. Several PCR primers to amplify amoA have been published (Holmes et al.., 1995; Sinigalliano et al.., 1995; Rotthauwe et al.., 1997; Juretschko et al.., 1998; Nold et al.., 2000; Purkhold et al.., 2000; Hoshino et al.., 2001; Nicolaisen & Ramsing, 2002; Norton et al.., 2002; Okano et al.., 2004). The analysis of AMO-encoding genes has been extended to amoC and amoB (Purkhold et al.., 2000; Norton et al.., 2002; Calvo & Garcia-Gil, 2004), and more recently functional genes homologous to those in AOB have been described in Archaea (Konneke et al.., 2005; Treusch et al.., 2005). Some of these primers were designed when only a few sequences were available. Considering the new sequence information accumulated in recent years, including the complete genomes of Nitrosomonas europaea (Chain et al.., 2003), Nitrosococcus oceani (Klotz et al.., 2006) and Nitrosospira multiformis, sequence analysis can contribute to estimate the advantages and failures of the available primers, and to assist the development of new strategies to study the structure of AOB communities. In this study all available amoCAB sequences from recognized AOB species, and whenever possible the sequences from uncultured clones, were used to characterize the published PCR primers and to propose new primers for the amplification of the amoCAB operon.

Materials and methods

Sequences and alignments

for in silico analyses, the nucleotide sequences of amo genes were downloaded from GenBank using entrez (http://www.ncbi.nlm.nih.gov/). Protein sequences were retrieved from Swissprot using expasy (http://www.expasy.org). The analyzed sequences were: (a) 16 sequences of amoC from both beta- and gammaproteobacterial AOB, (b) eight sequences of the related subunit of the particulate methane monooxygenase (pmoC), (c) one amoC sequence from the recently described ammonia-oxidizing archaeon Candidatus Nitrosopumilus maritimus (Konneke et al.., 2005), (d) 32 amoB sequences from beta- and gammaproteobacterial AOB; (e) seven pmoB sequences from methane-oxidizing bacteria (MOB), (f) two amoB sequences from crenarchaeota and (g) 2669 sequences of amoA and the related α subunit of the particulate methane monooxygenase (pmoA) from cultured and uncultured AOB. Although amoA sequences from crenarchaeota were considered, they differed widely and were excluded from the analysis.

The amoC, pmoC, amoB, pmoB, amoA and pmoA sequences were integrated in arb (Ludwig et al.., 2004). A database of complete and partial sequences of amoA from recognized AOB species was also prepared in arb. Before the analysis, the sequences were verified manually and those including STOP codons or erroneous starting points were omitted. To simplify the presentation of the results, sequence similarity is shown only for amoA sequences from cultured AOB (11 different phylogenetic clusters) and 10 pmoA sequences. The complete databases are available at http://cegg.unige.ch/ammoniaoxigenase. Sequences were aligned using clustalw included in arb.


To simplify the comparison between primers that had been designed in different studies, this study proposes a standardized designation system according to the name of the targeted gene, followed by information on the position and orientation of the primers. An example of such designation is as follows: amoA31f in which ‘amoA’ indicates the gene targeted, ‘31’ the position in the alignment and ‘f’ the direction of the primers (forward). Additional letters at the end of the designation indicate modifications such as shorter versions (s), wobble positions (IUPAC code), probe for FISH (p) or primer specific for gammaproteobacterial AOB (Gam). The new designation is always given in parenthesis after the original designation of the primer [e.g. AMO-F (amoA21f)].

Analysis of the primers was carried out using the software oligo 6.0 (Table 1). The position of each primer was determined after alignment of all the sequences in arb. Specificity was evaluated using blast (http://www.ncbi.nlm.nih.gov/BLAST/) for short, nearly exact matches and also match probe in arb. Because the different Tm values in the presence of several mismatches calculated by oligo (Table 1) do not take into account the position of the mismatch, additional analyses were carried out with match probe in arb. The match probe subroutine of arb calculates two different parameters for specificity: number of mismatches and weight of the mismatches. The last parameter depends on the number, position and kind of mismatches. A maximum number of five mismatches was allowed in the analysis. New primers were designed by visual inspection of the multiple alignments or using the software genefisher (http://bibiserv.techfak.uni-bielefeld.de/genefisher/). The newly designed primers were also analyzed with oligo v.6.0 (Table 1).

View this table:

Primers analyzed in this study

Tm ( °C) oligo with different number of mmLoop Tm ( °C)References
GeneNew designationOriginal designationSequence 5′–3′PositionLength (bp)Deg.01234
amoAamoA21fAMO-FAGA AAT CCT GAA AGC GGC21–3818162.255.548.942.235.5NSinigalliano et al. (1995)
amoA34fGCG GCR AAA ATG CCG CCG GAA GCG34–5724186.481.476.471.466.4105Molina et al. (2007)
amoA40fAMO-F2AAG ATG CCG CCG GAA GC40–5617168.761.654.647.540.4NJuretschko et al. (1998)
amoA49fGAG GAA GCT GCT AAA GTC49–6618153.646.940.233.626.9NThis study
amoA60r304RTAY CGC TTC CGG CGG CAT TTT CGC CGC34–6027275.870.164.458.753.067.0Norton et al. (2002)
amoA121famoA-3FACC TAC CAC ATG CAC TT121–13717151.044.036.929.922.8NWebster et al. (2002)
amoA151fA189GGN GAC TGG GAC TTC TGG151–16818459.052.445.739.032.4NHolmes et al. (1995)
amoA154f301FGAC TGG GAC TTC TGG CTG GAC TGG AA154–17926167.962.256.550.845.1NNorton et al. (2002)
amoA154fsGAC TGG GAC TTC TGG154–16815146.338.330.322.314.3NThis study
amoA187famoA-1FFCAA TGG TGG CCG GTT GT187–20317164.457.350. et al. (2001)
amoA310famoA-3FGGT GAG TGG GYT AAC MG310–32617451.144.036.929.922.8NPurkhold et al. (2000)
amoA332famoA-1FGGG GTT TCT ACT GGT GGT332–34918158.351.645.038.331.6NRotthauwe et al. (1997)
amoA332fHYamoA1F modGGG GHT TYT ACT GGT GGT332–34918658.852.145.438.832.1NStephen et al. (1999)
amoA337pA337TTC TAC TGG TGG TCR CAC TAC CCC ATC AAC T337–36731256.050.244.538.833.1NOkano et al. (2004)
amoA359rCamoA-4RGGG TAG TGC GAC CAC CAG TA340–35920165. et al. (2002)
amoA627rCGT ACC TTT TTC AAC CAT CC608–62720162. study
amoA665rAMO-R2GCT GCA ATA ACT GTG GTA648–66518153.446.740.133.426.7NJuretschko et al. (1998)
amoA680rA682 modAAV GCV GAG AAG AAW GC664–680171851.544.437.430.323.3NNold et al. (2000)
amoA681rA682GAA SGC NGA GAA GAA SGC664–681181654.447.841.134.527.8NHolmes et al. (1995)
amoA686rAMO-RGAT ACG AAC GCA GAG AAG669–68618154.948.341.634.928.3NSinigalliano et al. (1995)
amoA820rAmoA-2R′CCT CKG SAA AGC CTT CTT C802–82019456.149.843.537.230.93.0Okano et al. (2004)
amoA822ramoA-2RCCC CTC KGS AAA GCC TTC TTC802–82221465.059.253.547.842.13.0Rotthauwe et al. (1997)
amoA822rTCamoA-2R-TCCCC CTC TGC AAA GCC TTC TTC802–82221170.264.558.747.341.63.0Nicolaisen & Ramsing (2002)
amoA822rTGamoA-2R-TGCCC CTC TGG AAA GCC TTC TTC802–82221169.263.557.852.146.311.0Okano et al. (2004)
amoA828r302RTTT GAT CCC CTC TGG AAA GCC TTC TTC802–82827170.264.458.753.047.330.0Norton et al. (2002)
amoBamoB44ramoB-4RGCT AGC CAC TTT CTG G29–4416151.944.436.929.421.941.0Purkhold et al. (2000)
amoB160famoBMfTGG TAY GAC ATK AWA TGG160–17718847.040.333.627.020.3NCalvo & Garcia-Gil (2004)
amoB506r308RTCC CAG CTK CCG GTR ATG TTC ATC C482–50625468.863.157.451.645.9NNorton et al. (2002)
amoB660ramoBMrRCG SGG CAR GAA CAT SGG643–660181662.856.149.542.836.1NCalvo & Garcia-Gil (2004)
amoB1179rCCA AAR CGR CTT TCC GG1164–117917461.053.946.939.832.7NThis study
amoB1179rGamGCA AAG CGG CTG TCT GG1164–117917164.857.850.743.736.6NThis study
amoCamoC58fCTA YGA CAT GTC RCT GTG G58–7219451.545.138.832.526.2NThis study
amoC763f305FGTG GTT TGG AAC RGI CAR AGC AAA763–786211661.856.150.444.739.0NNorton et al. (2002)
  • New primer designations consider: target gene (amo followed by A, B or C), position in the alignment and orientation (forward, f; reverse, r. Modifications of the original primer sequence are shown in IUPAC code after the letter indicating the orientation of the primer. Other designations: p=probe for FISH; s=shorter version; Gam=specific for gamma-AOB. For amoA the positions were defined according to the sequence of Nitrosomonas europaea (L08050). For amoB and amoC the positions were defined according to the sequences of the Nitrosomonas europaea genome (BX321859). Melting temperature was calculated by nearest neighbor method. Deg.=number of different sequences due to wobble positions. mm=number of mismatched positions. N=no loops detected.

Results and discussion

Sequence analysis of amoA primers

Sequence matching of the amoA primers was analyzed in the ARB database prepared in this study. The complete alignment extended over 829 nucleotide positions, which were numbered according to the sequence of Nitrosomonas europaea (L08050). The majority of the amoA sequences were found in the region between positions 340 and 802. Therefore, the comparison of primers annealing outside of this region was limited to only a few sequences from the following clusters: Nitrosospira cluster 3, Nitrosomonas europaea, Nitrosomonas oligotropha, Nitrosomonas cryotolerans, gammaproteobacterial AOB and Methylococcus capsulatus (Table 2).

Table 2

Comparison of the primer sequences with the ARB database

The primer pair AMO-F (amoA21f) and AMO-R (amoA686r) (Sinigalliano et al.., 1995), which had been derived from one sequence of Nitrosomonas europaea available at that time, proved to be highly specific for the Nitrosomonas europaea cluster (Table 2). In the GenBank search, the forward primer AMO-F (amoA21f) matched perfectly sequences from Nitrosomonas europaea. In contrast, AMO-F (amoA21f) has three to five mismatches with some sequences of Nitrosospira cluster 3, and more than five mismatches with Nitrosospira multiformis, two sequences from the Nitrosomonas oligotropha cluster, Nitrosomonas cryotolerans, MOB and gammaproteobacterial AOB. Additionally, the comparison with clonal sequences from uncultured organisms showed that this primer has five mismatches to another region of pmoA. The reverse primer AMO-R (amoA686r) matched perfectly with only three sequences of the Nitrosomonas europaea cluster, but possessed two to more than five mismatches with other sequences of this cluster. AMO-R (amoA686r) also has two to four mismatches with almost all sequences from cultured betaproteobacterial AOB, 1190 sequences from uncultured betaproteobacterial AOB and pmoA from Methylococcus capsulatus. This primer has more than five mismatches with all other MOB and amoA of gammaproteobacterial AOB. According to this study's sequence analysis (Table 2), the AMO-F (amoA21f) and AMO-R (amoA686r) pair may be suitable to amplify AOB closely related to Nitrosomonas europaea and to exclude other AOB groups under stringent PCR conditions. An experimental evaluation (Sinigalliano et al.., 1995) had shown that this primer pair can also amplify amoA from Nitrosomonas cryotolerans and Nitrosococcus oceani, but this conclusion is not supported by the in silico evaluation and can only be explained by the use of PCR conditions favoring low specificity.

The primer pair AMO-F2 (amoA40f) and AMO-R2 (amoA665r) (Juretschko et al.., 1998) was published to increase the sensitivity of amoA detection using a nested PCR approach from templates prepared with the primers AMO-F (amoA21f) and AMO-R (amoA686r) (Sinigalliano et al.., 1995), considered above. AMO-F2 (amoA40f) matches perfectly eight of the 14 sequences analyzed (both Nitrosospira and Nitrosomonas spp.) and have one mismatch with Nitrosospira sp. NpAV, two with Nitrosospira multiformis, three with Nitrosomonas cryotolerans, and more than five with MOB or gammaproteobacterial AOB. The sequence analysis suggests that AMO-F2 (amoA40f) may be suitable to target betaproteobacterial AOB in general. In contrast, the primer AMO-R2 (amoA665r) seems to match sequences from the Nitrosomonas europaea cluster (including environmental clones), better than other clusters, matching perfectly only three sequences from the Nitrosomonas europaea cluster (Table 2). AMO-R2 (amoA665r) has four high-weighted mismatches to sequences from the Nitrosospira lineage and one to four mismatches with different weight with other Nitrosomonas sequences. Because of the restricted spectrum of matches of AMO-R2 (amoA665r), the authors conclude that AMO-F2 (amoA40f) may be suitable as a general primer for amplifying betaproteobacterial AOB, but it should be combined with another reverse primer to accomplish this goal. Although the AMO-F2 (amoA40f) and AMO-R2 (amoA775r) pair was originally designed for nested amplification from products prepared with the primers AMO-F (amoA21f) and AMO-R (amoA686r) in order to increase PCR sensitivity, this approach seems to have limited applicability considering that the primer pair used in the first round of PCR (AMO-F and AMO-R) appear to be biased for amplification of Nitrosomonas europaea.

Recently, several regions for primer design have been identified based on reverse translation of protein alignment in the amoCAB operon (Norton et al.., 2002). The primer 304R (amoA60r) is located near the 5′ end of amoA and allows, in combination with the primer 305F (amoC763f), the amplification of the intergenic region between amoC and amoA. This primer does not perfectly match any of the amoA sequences of cultured AOB (Table 2), having three to four mismatches of high weight in all cases examined. Additionally, the primer 304R (amoA60r) possesses a very stable loop structure (Table 1), which is not desirable for PCR. The experimental evaluation (Norton et al.., 2002) showed that 304R (amoA60r), in combination with 305F (amoC763f), amplified the variable intergenic region of Nitrosospira sp. NpAv, Nitrosospira briensis, Nitrosospira sp. 39–19, Nitrosospira tenuis, Nitrosospira multiformis, Nitrosomonas europaea, Nitrosomonas eutropha, Nitrosomonas sp. AL212, Nitrosomonas sp. JL21, Nitrosomonas sp. GH22 and Nitrosomonas cryotolerans. However, according to the sequence analysis, this should be only possible under low specificity of PCR (Table 1). This intergenic region can be relevant for community studies because the size of the products obtained from each species is different and the nucleotide sequence is highly variable. However, modification of the primers (for example shortening of primers or designing new primers) might be considered for application with environmental samples.

The primers 301F (amoA154f) and 302R (amoA828r) were designed as a primer pair to amplify a core region of 675 bp from amoA in 14 AOB (Norton et al.., 2002). The primer 301F (amoA154f) matches perfectly Nitrosospira briensis, Nitrosovibrio tenuis, Nitrosospira sp. 39-19 and Nitrosomonas cryoloterans, but has one to four mismatches with all other Nitrosospira and Nitrosomonas sequences, three to four mismatches with gammaproteobacterial AOB and two to three mismatches with pmoA (Table 2). The primer 302R (amoA828r) only targets amoA of gammaproteobacterial AOB, because its target region is deleted in the amoA of gammaproteobacterial AOB. Among betaproteobacterial AOB the primer 302R (amoA828r) matches perfectly only the sequences from Nitrosovibrio tenuis and Nitrosomonas europaea, but has one mismatch of low weight with Nitrosospira sp. Np 39–19, one to two mismatches of intermediate weight with Nitrosospira briensis, Nitrosospira multiformis, Nitrosospira sp. NpAV, Nitrosomonas sp. GH22 and Nitrosomonas sp. TK794, and three to four mismatches of high weight with Nitrosomonas sp. AL212, Nitrosomonas sp. JL21 and Nitrosomonas cryotolerans. Because of their length and base composition, both 301F (amoA154f) and 302R (amoA828r) have a very high Tm (Table 1), and therefore PCR conditions (for example salt and formamide concentration) have to be modified. Considering that the forward primer 301F (amoA154f) has potential to match simultaneously beta- and gammaproteobacterial AOB and MOB, the shorter version amoA154fs is suggested as a modification with lower Tm (Table 1) and significantly higher sequence similarity for all of the sequences (Table 2).

The primer amoA-1FF (amoA187f) (Hoshino et al.., 2001) was originally designed to amplify Nitrosomonas europaea in combination with the primer amoA-2R (amoA822r), for in situ PCR. In the sequence analysis, amoA-1FF (amoA187f) fully matches Nitrosomonas europaea, Nitrosococcus mobilis and Nitrosospira sp. NpAV, and has one to two mismatches with the other Nitrosomonas and Nitrosospira sequences (Table 2). The low number of mismatches with some sequences from other clusters of both lineages (Nitrosovibrio tenuis, Nitrosospira sp. Np 39-19, Nitrosomonas sp. C-113a and also uncultured clones) suggests that the amoA-1FF (amoA187f) is probably not specific for Nitrosomonas europaea.

The primer combination amoA-3F (amoA310f) and amoB-4R (amoB44r) was designed to amplify part of amoA from the gammaproteobacterial AOB Nitrosococcus halophilus (Purkhold et al.., 2000). amoa-3F (amoA310f) matches perfectly only the sequence from this species, has one mismatch with the two other gammaproteobacterial AOB, two or three mismatches with MOB and three to four mismatches of high weight with Nitrosospira multiformis, Nitrosospira sp. Np 39-19 and the majority of Nitrosomonas sequences (Table 2). According to the analysis with oligo (Table 1), highly stringent conditions are needed for reliable results with amoA-3F (amoA310f).

The primer pair amoA-1F (amoA332f) and amoA-2R (amoA822r) (Rotthauwe et al.., 1997) is the most widely used to amplify amoA in environmental studies, despite the differences in the Tm between the primers (Table 1). amoA-1F (amoA332f) is located in a region conserved in all betaproteobacterial AOB; it matches perfectly or with one to two mismatched sequences from betaproteobacterial AOB, but it does not match sequences from gammaproteobacterial AOB. The primer amoA1F mod (amoA332fHY), which is a modified version including two wobble positions to increase sequence identity with cultured betaproteobacterial AOB (Stephen et al.., 1999), matched the same spectrum of sequences, but produced differences in the weight of the mismatches (Table 2) and Tm (Table 1). The primer amoA-2R (amoA822r) matched only sequences from betaproteobacterial AOB, but had lower specificity than 302R (amoA828r) (Norton et al.., 2002). Several variants of amoA-2R (amoA822r) have been proposed, including the primer amoA-2R′ (amoA820r) (Okano et al.., 2004), a shorter version with lower sequence similarity to the target region and additional unspecific matches in other regions of amoA from both cultured and uncultured species. Other variants of amoA-2R (amoA822r) have been proposed specifically for denaturing gradient gel electrophoresis (DGGE), in order to reduce the number of wobble positions that usually generate double bands in denaturing gels. These include amoA-2R-TC (amoA822rTC) (Nicolaisen & Ramsing, 2002) or amoA-2R-TG (amoA822rTG) (Okano et al.., 2004). These primers matched the same sequences as the original version but showed differences in the weight of the mismatches (Table 2) and higher Tm (Table 1).

In addition to the primers described to amplify amoA, the probe A337 (amoA337p) (Okano et al.., 2004) has been published for FISH. Although this probe has, in most of the cases, mismatches with sequences from cultures of all betaproteobacterial AOB clusters, it has fewer than five mismatches with all sequences from cultured betaproteobacterial AOB and all sequences from uncultured clones, suggesting that it is located in a region suitable for the design of a general primer for detection of betaproteobacterial AOB.

Sequence analysis of primers for simultaneous detection of amoA and pmoA

The common evolutionary origin of AMO and particulate methane monooxygenase (pMMO) (Holmes et al.., 1995) suggests the possibility of finding conserved regions for designing primers that amplify both genes. The primer pair A189 (amoA151f) and A682 (amoA681r) was used for this purpose (Holmes et al.., 1995). A189 (amoA151f) is located in the same conserved region as 301F (amoA154f) (Norton et al.., 2002) and amoA154fs. It has a perfect match with the majority of sequences from beta- and gammaproteobacterial AOB and MOB (Table 2). The reverse primer A682 (amoA681r) matches perfectly only sequences from Nitrosospira clusters 2 and 0. A further modification of this primer, A682 mod (amoA680r) (Nold et al.., 2000), was designed to increase the sensitivity for gammaproteobacterial AOB. However, as shown in Table 2, the matches with cultured AOB improved only slightly.

Sequence analysis of amoC and amoB primers

Both amoB and amoC are likely to be good alternatives as functional markers for molecular studies on AOB because, they code for essential parts of the multi-subunit AMO enzyme, which may be involved in the active site by extrapolation with the homologous pmoC and pmoB (Lieberman & Rosenzweig, 2005; Balasubramanian & Rosenzweig, 2007), and have a suitable size for phylogenetic inferences (amoC has around 800 bp and amoB is the longest of the three genes with more than 1200 bp). However, compared with amoA, amoB and amoC have been neglected despite the their potential for additional sequence information.

Consequently, only a few primers have been described to amplify these genes. Primer 305F (amoC763f) (Norton et al.., 2002) was designed to be used in combination with 304R (amoA60r) to generate a PCR product encompassing the 3′ end of amoC, the intergenic region with amoA and the 5′ part of amoA (see Fig. 1). Alignment with amoC sequences showed that primer 305F (amoC763f) does not match perfectly any of the sequences analyzed (Fig. 2) and possesses a significant difference in Tm (Table 1) with 304R (amoA60r). The primer 305F (amoC763f) has between one and six mismatches with betaproteobacterial AOB and more than 10 mismatches with gammaproteobacterial AOB and MOB. The two copies of amoC that are not located in the amoCAB operon of betaproteobacterial AOB had more mismatches at different positions with the primers (Fig. 2), suggesting that new primers can be designed to target these singleton copies specifically.


Schematic diagram of the amoCAB operon in beta- (a) and gamma-AOB (b). IS, intergenic regions. The position and orientation of the different primers are shown by arrows. For primer designation see Table 1.


Alignment using clustalw of amoC primers with all sequences available. Matches with the primer sequences are indicated by dots. Matches in wobble positions are shown as shaded. The asterisk denotes the amoC copies of Nitrosomonas europaea and Nitrosospira multiformis not belonging to the amoCAB operon.

The primer amoB-4R (amoB44r) (Purkhold et al.., 2000), which was designed to amplify amoAB from Nitrosococcus halophilus in combination with the primer amoA-3F (amoA 310f), does not match perfectly any sequence analyzed (Fig. 3). This region is not highly conserved either in gamma- or in betaproteobacterial AOB.


Alignment using clustalw of amoB primers with all sequences available. Matches with the primer sequences are indicated by dots. Matches in wobble positions are shown as shaded. Dashes represent gaps in the alignment.

The primer pair amoBMf (amoB160f) and amoBMr (amoB660r) (Calvo & Garcia-Gil, 2004) has been published recently in order to use amoB as an alternative molecular marker for AOB. Both primers target regions relatively conserved in beta- and some gammaproteobacterial AOB (Fig. 3), but so far they have not been used extensively in environmental samples. The annealing temperature suggested for this primer pair (Calvo & Garcia-Gil, 2004) is significantly higher than the calculated values (Table 1).

The primer 308R (amoB506r) (Norton et al.., 2002) was proposed to be combined with 305F (amoC763f) as an alternative to obtain the full length of the amoA gene and its flanking regions. In the alignment with amoB sequences (Fig. 3), this primer had 10–11 mismatches with sequences from gammaproteobacterial AOB and is therefore probably suitable only for betaproteobacterial AOB.

Very recently, the amoB sequences from two Archaea have been deposited in GenBank (Konneke et al.., 2005; Treusch et al.., 2005). These partial sequences were too short for sequence comparison with the majority of primers analyzed here. The primers amoBMf (amoB160f) and 308R (amoB506r) presented more than 12 mismatches and are not expected to target these sequences.

Description of new primers for amplification of the amoCAB operon

To examine the possibility of amplifying the almost complete amoCAB operon, sequence conservation was inspected in the few sequences available for the flanking genes amoC and amoB. The primers amoC58f and amoB1179r (Table 1) were designed to amplify the largest segment possible of the operon, which includes the three genes and the intergenic regions. The size of the PCR product is variable due to differences in the length of the genes and especially of the intergenic regions, but should be around 2900 bp. Matching of the primer amoC58f with the amoC sequences available in GenBank is shown in Fig. 2. A blast search retrieved only sequences from betaproteobacterial AOB and did not have any unspecific match. This primer matched perfectly the sequences from betaproteobacterial AOB, except for the amoC copies of Nitrosomonas europaea and Nitrosospira multiformis that are not located in an operon. These extra copies of amoC are expected to be excluded from the amplification because of the difference in the sequence but also the use of the reverse primer amoB1179r, which is located at the end of the amoB gene. The primer amoB1179r matches in a highly conserved region of amoB from betaproteobacterial AOB and Nitrosococcus halophilus (Fig. 3). In a blast search, it matched all amoB from betaproteobacterial AOB. In a modification of this primer (amoB1179rGam) the specificity is shifted to target only gammaproteobacterial AOB.

The application of amoA for phylogenetic inference is partially limited due to short length and high conservation of the fragment analyzed (Purkhold et al.., 2003). Therefore, one of the main challenges for applying this gene as a functional molecular marker is the search of primers that allow the amplification of a longer amoA fragment. Different conserved positions were detected in the amoA alignment. The primer amoA34f was designed to target positions close to the 5′ region of the gene that can be used in combination with primers for the 3′ region of the gene such as amoA-2R (Rotthauwe et al.., 1997) or 302R (Norton et al.., 2002) to amplify almost the whole of amoA. This primer retrieved sequences from all beta AOB included in this study (Table 2), and has been already used to characterize AOB communities in marine environments (Molina et al.., 2007). The wider spectrum of betaproteobacterial AOB recognized by the primer amoA34f, compared with the primer amoAF (Sinigalliano et al.., 1995), makes amoA34f a better option for PCR in environments not dominated by Nitrosomonas-like AOB.

The primers amoA121f and amoA359rC were designed to amplify an internal fragment from betaproteobacterial AOB suitable for DGGE. The primer amoA121f matches perfectly all Nitrosospira spp. and some Nitrosomonas spp. and with one to four mismatches Nitrosomonas eutropha, Nitrosomonas sp. GH22, Nitrosomonas sp. TK794, Nitrosomonas sp. AL212, Nitrosomonas sp. JL21. It has more than five mismatches with sequences from gammaproteobacterial AOB (Table 2). The reverse primer amoA359rC, having nine bases overlap with amoA-1F (amoA332f) (Rotthauwe et al.., 1997), matches perfectly the sequences from all Nitrosospira clusters and displays high similarity with the Nitrosomonas clusters. A former version of the primer combination amoA121fgc-amoA359r, which was originally designated amoA-3F/amoA-4R, designed in the laboratory was previously used by other research groups to analyze the impact of soil management on the diversity of AOB in soil (Webster et al.., 2002). The primer amoA359rC reported in this manuscript is an improved variant of the original primer designated amoA4-R, which was used in the DGGE without wobble positions to avoid artifacts. Besides the use of this primer combination for DGGE, the size of the expected PCR product makes it also potentially useful for quantification of AOB by real-time PCR.

Although the number of amoA sequences from gammaproteobacterial AOB is very limited (only two complete sequences), the primer pair amoA49f and amoA627r was designed to tentatively amplify a fragment of 559 bp exclusively from gammaproteobacterial AOB. These primers, when checked in GenBank by blast, matched only the sequences used for primer design. Similarly, the evaluation in arb showed that the forward primer matches only the two gammaproteobacterial AOB while the reverse primer has three mismatches with Nitrosococcus halophilus but no mismatches with Nitrosococcus oceani. Between one and three mismatches were recorded with some MOB and more than five with all betaproteobacterial AOB (Table 2). These primers have a similar melting temperature (Table 1), desirable for specific amplification.


The re-examination of specific primers to amplify the amoCAB operon carried out in this study by sequence analyses indicates possible strength and weakness of primers to study community composition of AOB in environmental samples. The use of new primers targeting new regions in the complete operon can contribute to the information on the evolution and function of the amoCAB operon in AOB. Additionally, nested amplification offers the possibility of increasing PCR sensitivity for AOB detection in environmental samples.

Authors' contribution

P.J. and O.-S.K. contributed equally to this paper.


This research was partially supported by G.I.F. (German–Israel Foundation) grant no. I-711-83.8/2001 and BSF (Binational Science Foundation) grant no. 2002-206. The authors wish to thank the Max-Planck Society and the G.I.F. for financial support of P. J. and O.-S.K. during this study.


  • Editor: Michael Wagner


View Abstract