Biochemical and Biophysical Research Communications 380 (2009) 710–714

Contents lists available at ScienceDirect

Biochemical and Biophysical Research Communications

journal homepage: www.elsevier.com/locate/ybbrc

Biotransformation of the neonicotinoid insecticides imidacloprid and

thiamethoxam by Pseudomonas sp. 1G
Gunjan Pandey *, Susan J. Dorrian, Robyn J. Russell, John G. Oakeshott
CSIRO Entomology, GPO Box 1700, Canberra, ACT 2601, Australia

a r t i c l e i n f o a b s t r a c t

Article history: We report the isolation of a Pseudomonas sp. which is able to transform imidacloprid and thiamethoxam
Received 23 January 2009 under microaerophilic conditions in the presence of an alternate carbon source. This bacterium, Pseudo-
Available online 29 January 2009 monas sp. 1G, was isolated from soil with a history of repeated exposure to imidacloprid. Both insecti-
cides were transformed to nitrosoguanidine (@NANO), desnitro (@NH), and urea (@O) metabolites and
a transformation pathway is proposed. This is the first conclusive report of bacterial transformation of
Keywords: the ‘magic nitro’ group which is responsible for the insect selectivity of neonicotinoid insecticides.
Neonicotinoids
Crown Copyright Ó 2009 Published by Elsevier Inc. All rights reserved.
Imidacloprid
Thiamethoxam
Degradation
Transformation
Pseudomonas
Bacteria
Magic nitro
Metabolites

Neonicotinoids are a relatively new class of synthetic organ-
ic insecticide, which are now widely used to control piercing
and sucking insect pests (aphids, leafhoppers, whiteflies etc.)
around the world. Annual sales of neonicotinoid insecticides
were around $US 1.56 billion in 2006, representing nearly
17% of the global insecticide market [1]. Neonicotinoids act
on the insect central nervous system by posing as an agonist
at the nicotinic acetylcholine receptor (nAChR), blocking the
nicotinergic neuronal pathways and causing accumulation of
the neurotransmitter acetylcholine [2–7]. Although vertebrates
also have nAChRs and nicotinergic neuronal pathways, the
neonicotinoids show a high level of insect selectivity, because
they have a much higher affinity for insects as compared to
vertebrate nAChRs. This selectivity of neonicotinoids is attrib-
uted to their so-called ‘magic nitro’ (nitroimine, @NANO2)
group [4,8].
Imidacloprid, N-[1-[(6-Chloro-3-pyridyl)methyl]-4,5-dihydroi-
midazol-2-yl]nitramide, was the first neonicotinoid to be widely
used commercially and remains by far the biggest product in the
class. Imidacloprid is a first generation chloropyridylmethyl
neonicotinoid but second-generation chlorothiazolylmethyl
neonicotinoids such as thiamethoxam, (EZ)-3-(2-chloro-1,3-
thiazol-5-ylmethyl)-5-methyl-1,3,5-oxadiazinan-4-ylidene(nitro)
* Corresponding author. Fax: +61 02 6246 4173.
E-mail address: Gunjan.Pandey@csiro.au (G. Pandey).
amine, are also now widely used for foliar, soil and seed treat-
ment [1,8].
Metabolism of imidacloprid and thiamethoxam has been
extensively studied in mammalian systems [4,8,9]. Although
other parts of the molecules can also be attacked, the ‘magic
nitro’ moiety has been a particular focus for these studies be-
cause slight alterations to it can completely reverse the selec-
tive toxicity of neonicotiniods for insects over vertebrates. For
instance, the selectivity ratio for insects over vertebrates
changes from 565 to 0.005 when this ‘magic nitro’ (@NANO2)
is metabolized to guanidine/desnitro (@NH2) derivatives [4,8].
Several human cytochrome P450 enzymes are known to reduce
imidacloprid to the nitroso (@NANO) derivative in an oxygen-
sensitive manner [10] and another oxygen-sensitive ‘neonicoti-
noid nitro-reductase’ from rabbit liver cytosol was recently
identified by Dick et al. as an aldehyde oxidase that readily
converts imidacloprid to nitrosoguanidine (@NANO) and ami-
noguanidine (NANH2) derivatives [11]. The latter is then acted
upon by unknown liver enzymes to produce other nitroimines
such as the corresponding guanidine (@NH) and urea (@O)
derivatives [12]. A recent study of thiamethoxam metabolism
has shown that its ‘magic nitro’ is also transformed to produce
nitrosoguanidine, aminoguanidine, guanidine and urea deriva-
tives in mice [9].
The fate of imidacloprid and thiamethoxam in the environ-
ment is not well understood but metabolites toxic to verte-

0006-291X/$ - see front matter Crown Copyright Ó 2009 Published by Elsevier Inc. All rights reserved.
doi:10.1016/j.bbrc.2009.01.156
 G. Pandey et al. / Biochemical and Biophysical Research Communications 380 (2009) 710–714 711

brates have been reported from soil [4,6]. In this paper, we re- Results and discussion
port transformation of imidacloprid and thiamethoxam at the
‘magic nitro’ pharmacophore by a newly isolated soil Attempts to set up cultures from the soil sample in which imi-
bacterium. dacloprid or thiomethoxam was provided as the sole source of car-
bon and nitrogen did not show any degradation under aerobic or
microaerophilic conditions. However degradation of about 70% of
Materials and methods an initial concentration of 50 ppm imidacloprid was observed
within 14 days after five rounds of subculturing under microaero-
Chemicals. Dichloromethane, imidacloprid, thiomethoxam, for- philic conditions when filter-sterile 10 mM glucose was added as a
mic acid, and D-glucose were obtained from Sigma–Aldrich (St. carbon source (data not shown). When different dilutions of this
Louis, MO). The solvents were obtained from Merck (Whitehouse enrichment were plated onto nutrient agar plates, 20 morphologi-
Station, NJ). cally distinct bacterial colonies appeared after 2 days of incubation
Soil samples, bacterial strains and culture conditions. Soil samples at 28 °C. These were picked and checked individually for imidaclo-
were collected from a golf course at Kamba, ACT with seven years’ prid degradative activity in microaerophilic liquid cultures as
history of imidacloprid exposure. above. Three of these strains, 1G, 1W, and GP2, were found to de-
Imidacloprid and thiamethoxam stock solutions were pre- grade 70% imidacloprid from an initial concentration of 50 ppm
pared in dichloromethane and used at final concentrations of in 14 days. All three showed 99% 16S rRNA gene sequence identity
50 ppm (196 lM) in minimal media as previously described with the genus Pseudomonas. These strains were also tested for
[13]. The dichloromethane was evaporated by shaking the media thiamethoxam degradation under similar conditions of microaero-
well before inoculating 1% (w/v) of the soil sample. Cultures philic growth in the presence of 10 mM glucose. They also de-
were then set up under a variety of aerobic, anaerobic and graded about 70% of 50 ppm thiamethoxam within 14 days
microaerophilic conditions at 28 °C. Microaerophilic conditions under these conditions (data not shown). Since the three strains
were maintained in anaerobic jars using Campylobacter gas gen- metabolized both imidacloprid and thiamethoxam at similar rates
erating kits (Oxoid Limited, Hampshire, UK). and initial data showed that they produced similar metabolites
Three strains capable of degrading the two neonicotinoids were (data not shown), only one strain, 1G, was selected for further
isolated on plates after five subcultures as above. These strains studies.
were identified by 16s rRNA gene sequencing (Genbank Accession The HPLC analysis which showed 70% depletion of the imi-
Nos. FJ648813, FJ648814, and FJ648815). dacloprid and thiamethoxam above also detected the concomi-
HPLC and LC–MS anlaysis. Imidacloprid and thiamethoxam tant appearance of one metabolite peak in each case (data not
depletion was measured quantitatively by HPLC with a photodiode shown). However, the extracted ion chromatogram obtained
array detector operating at 270 nm. An Agilent series LC system from MS analysis of the HPLC stream resolved three metabolite
controlled by Agilent TOD Software (version A.01.00) was used peaks in each case. This observation suggested transformation
(Agilent Technologies, Santa Clara CA). The mobile phase consisted of the ‘magic nitro’ group as the diode array detector was set
of acetonitrile:water (18:82 v/v) containing 0.1% formic acid (v/v). at a wavelength of 270 nm which is the characteristic absor-
The solvent mixture was pumped at 0.7 ml min 1. An AquaÒ C18, bance of this group. The three metabolites were then character-
5 lM particle size, 250  4.6 mm (Phenomenex, Torrance CA) col-
umn was used at 25C.
MS analysis was performed using an LC/MSD TOF mass spec-
trometer (Agilent Technologies) with an electrospray ionization
source. The mass spectrometer was connected to the HPLC
stream after the diode array detector. Scanning was done in po-
sitive ion mode. Nitrogen flow was maintained at 12 l min 1.
The capillary temperature and spray voltage were 350C and
3 kV, respectively. The fragmenter and skimmer were set at
120 and 60 V, respectively, for MS-TOF scans in the range of
50–300 m/z.
For metabolite analysis, 1% (v/v) aerobically grown seed cul-
ture of strain 1G in nutrient broth was inoculated in 50 ml of
minimal medium and incubated under non-shaking microaero-
philic conditions at 28 °C for 14 days. Culture (1 ml) was col-
lected, centrifuged at 12,000g and filtered with 0.22 lm filters
(Millipore, Billerica MA). The filtrate (10 ll) was used for the
HPLC analysis and 400 ll of filtrate was transferred into a
1.5 ml centrifuge tube for metabolite extraction. The filtrate
was extracted four times, each time with 5 min vortexing fol-
lowed by centrifugation for 3 min at 16,000g. Thereafter, the
upper, organic layer was transferred into a glass vial. The first
extraction involved 800 ll dichloromethane and then 10 ll of
20% NaOH was added, followed by two extraction cycles with
800 ll chloroform–acetonitrile (5:2). Finally, 20 ll of concen-
trated HCl was added to the separated aqueous fraction, fol-
lowed by a fourth extraction with chloroform–acetonitrile as
above. The combined extracts were dried by a stream of high
purity nitrogen and finally dissolved in 100 ll of acetonitrile–
water (20:80) and analysed with LC–MS. The experiment was
Figure 1. LC–MS extracted ion chromatograms showing metabolite peaks of: (A)
performed in triplicate with appropriate controls. imidacloprid and; (B) thiomethoxam.
712 G. Pandey et al. / Biochemical and Biophysical Research Communications 380 (2009) 710–714

Figure 2. LC–MS TOF spectrum of the pseudo molecular ions (m/z + H+) of imidacloprid (D) and its presumptive desnitro/guanidine (A), nitrosoguanidine (B) and urea (C)
metabolites.

ised by protonated molecular mass ion [M+H+] analysis in LC–
MS (ES+) studies (Fig. 1). In the presence of imidacloprid, three
metabolites (retention times 14.01, 19.82, and 22.76 min) corre-
sponding to elemental compositions of C9H11ClN4 + H+,
C9H10ClN5O + H+, and C9H10ClN3O + H+ appeared (Fig. 2). Simi-
larly, in the presence of thiamethoxam, three metabolites
(retention times 14.05, 19.65, and 22.72 min) corresponding to
elemental compositions of C8H11ClN4OS + H+, C8H10ClN5O2S + H+
and C8H10ClN3O2S + H+ appeared (Fig. 3). These empirical for-
mulas can best be interpreted as the desnitro, nitrosoguanidine,
and urea metabolites of imidacloprid and thiamethoxam,
respectively. The observed mass spectra of all six metabolites

Figure 3. LC–MS TOF spectrum of the pseudo molecular ions (m/z + H+) of thiomethoxam (D) and its presumptive desnitro/guanidine (A), nitrosoguanidine (B) and urea (C)
metabolites.
 G. Pandey et al. / Biochemical and Biophysical Research Communications 380 (2009) 710–714 713

Cl
Cl O
N S
N N N N N CH3

N N
NO2 NO2

Imidacloprid Thiamethoxam

N N N
NO2 NO NH2

Magic nitro Nitrosoguanidine Aminoguanidine

(nitroimine) metabolite metabolite

NH

Desnitro / guanidine
metabolite

Urea
metabolite

Figure 4. Proposed metabolic route for ‘magic nitro’ group transformation of imidacloprid and thiomethoxam in Pseudomonas sp.1G.

match precisely with the in silico generated mass spectra of
their respective elemental compositions.
In mammalian systems the ‘magic nitro’ group of imidacloprid
is postulated to be reduced to nitrosoguanidine and aminoguani-
dine and then cleaved to the guanidine and urea derivatives
[4,8,9]. A similar pathway has been reported for thiamethoxam
in mice, rats, goats and hens [9]. By contrast there has been no re-
port to date describing any imidacloprid or thiamethoxam miner-
alizing bacterial isolate. In fact there have been no previous reports
of any thiamethoxam metabolism by bacteria. In the only reported
bacterial activity against the ‘magic nitro’ group of the neonicoti-
noids, Leifsonia sp. strain PC-21 was found to co-metabolically
transform imidacloprid in tryptic soy broth supplemented with
D-glucose and succinate [14 and see below]. Similar co-metabolic
transformation has been observed in this study with D-glucose in
minimal medium. In an evolutionary context, it appears that the
complete set of genes required for mineralisation of neonicotinoids
has not been recruited within one bacterium since the commercial
release of the compounds in the early 1990s.
Recently Dai et al. have reported the isolation of Stenotropho-
monas maltophilia CGMCC 1.1788 which is capable of two-step
transformation of imidacloprid to an olefin-imidacloprid via ethyl-
ene hydroxylation and dehydrogenation [15]. While this study did
not implicate transformation of the ‘magic nitro’ group another re-
cent finding has done so. Specifically guanidine and urea metabo-
lites were recovered from metabolism of imidacloprid (37–58% of
25 ppm over 4 weeks) by the Leifsonia sp. strain PC-21 [14], albeit
this latter study did not detect the nitrosoguanidine metabolite of
imidacloprid as observed in the present study. The Leifsonia study
utilised complex tryptic soy broth media supplemented with 1%
succinate or D-glucose, so it is difficult to conclude much about
the pathway involved.
The oxygen-sensitive aldehyde oxidase from rabbit liver cytosol
(see Introduction) degrades imidacloprid to an aminoguanidine
metabolite in a two-step reduction via a nitrosoguanidine interme-
diate [11] but we could not detect an aminoguanidine metabolite in
this study. Nevertheless the oxygen sensitivity of the aldehyde oxi-
dase is interesting given that strain 1G requires microaerophilic
conditions to transform the ‘magic nitro’ group of neonicotinoids.
It could be that a similar aldehyde oxidase operates in our strain
to reduce the ‘magic nitro’ to nitrosoguanidine under microaero-
philic conditions. We suggest that the ‘magic nitro’ group of imida-
cloprid and thiamethoxam is transformed to produce
nitrosoguanidine metabolites. The parent molecules and/or the
nitrosoguinidine would be further converted to a non-toxic urea
metabolite via a more toxic desnitro/guanidine intermediate
(Fig. 4). Such a scheme for ‘magic nitro’ metabolism in mammalian
systems indeed has been proposed previously in the literature
[4,8,9]. Cloning of the gene-enzyme system(s) responsible for the
degradation we have described in Pseudomonas strain 1G is cur-
rently underway so we can further characterize these reactions.
Acknowledgments
This work was supported in part by Orica Australia and Horti-
culture Australia Ltd. The authors gratefully thank Dr. Rinku Pan-
dey for critically reading the manuscript and providing valuable
inputs.
714 G. Pandey et al. / Biochemical and Biophysical Research Communications 380 (2009) 710–714

References [8] M. Tomizawa, J.E. Casida, Neonicotinoid insecticide toxicology: mechanisms of

selective action, Annu. Rev. Pharmacol. Toxicol. 45 (2005) 247–268.
[9] K.A. Ford, J.E. Casida, Unique and common metabolites of thiamethoxam,
[1] P. Jeschke, R. Nauen, Neonicotinoids—from zero to hero in insecticide
clothianidin, and dinotefuran in mice, Chem. Res. Toxicol. 19 (2006) 1549–
chemistry, Pest. Manag. Sci. 64 (2008) 1084–1098.
1556.
[2] I. Yamamoto, J.E. Casida, Neonicotinoid Insecticides and the Nicotinic
[10] D.A. Schulz-Jander, J.E. Casida, Imidacloprid insecticide metabolism: human
Acetylcholine Receptor, Springer, Verlag, Tokyo, Japan, 1999.
cytochrome P450 isozymes differ in selectivity for imidazolidine oxidation
[3] D. Wollweber, K. Tietjen, Chloronicotinyl insecticides: a success of the new
versus nitroimine reduction, Toxicol. Lett. 132 (2002) 65–70.
chemistry, in: I. Yamamoto, J.E. Casida (Eds.), Neonicotinoid Insecticides and
[11] R.A. Dick, D.B. Kanne, J.E. Casida, Identification of aldehyde oxidase as the
the Nicotinic Acetylcholine Receptor, Springer, Verlag, Tokyo, Japan, 1999.
neonicotinoid nitroreductase, Chem. Res. Toxicol. 18 (2005) 317–323.
[4] M. Tomizawa, J.E. Casida, Selective toxicity of neonicotinoids attributable to
[12] D.A. Schulz-Jander, W.M. Leimkuehler, J.E. Casida, Neonicotinoid insecticides:
specificity of insect and mammalian nicotinic receptors, Annu. Rev. Entomol.
reduction and cleavage of imidacloprid nitroimine substituent by liver
48 (2003) 339–364.
microsomal and cytosolic enzymes, Chem.Res. Toxicol. 15 (2002) 1158–1165.
[5] M. Tomizawa, J.E. Casida, Structure and diversity of insect nicotinic
[13] D. Prakash, A. Chauhan, R.K. Jain, Plasmid-encoded degradation of p-
acetylcholine receptors, Pest. Manag. Sci. 57 (2001) 914–922.
nitrophenol by Pseudomonas cepacia, Biochem. Biophys. Res. Commun. 224
[6] M. Tomizawa, D.L. Lee, J.E. Casida, Neonicotinoid insecticides: molecular
(1996) 375–381.
features conferring selectivity for insect versus mammalian nicotinic
[14] J.C. Anhalt, T.B. Moorman, W.C. Koskinen, Biodegradation of imidacloprid by
receptors, J. Agric. Food Chem. 48 (2000) 6016–6024.
an isolated soil microorganism, J. Environ. Sci. Health B 42 (2007) 509–514.
[7] M.Y. Liu, J. Lanford, J.E. Casida, Relevance of [3H]imidacloprid binding site in
[15] Y.J. Dai, S. Yuan, F. Ge, T. Chen, S.C. Xu, J.P. Ni, Microbial hydroxylation of
house fly head acetylcholine receptor to insecticidal activity of 2-
imidacloprid for the synthesis of highly insecticidal olefin imidacloprid, Appl.
nitromethylene- and 2-nitroimino-imi-dazolidines, Pestic. Biochem. Physiol.
Microbiol. Biotechnol. 71 (2006) 927–934.
46 (1993) 200–206.