Organization and evolution of the flavin-containing monooxygenase genes of human and mouse:...

Organization and evolution of the flavin-containingmonooxygenase genes of human and mouse: identificationof novel gene and pseudogene clustersDiana Hernandeza, Azara Janmohameda, Pritpal Chandana, Ian R. Phillipsb andElizabeth A. Shepharda

Objectives To date, six flavin-containing monooxygenase

(FMO) genes have been identified in humans, FMOs 1, 2, 3,

4 and 6, which are located within a cluster on chromosome

1, and FMO5, which is located outside the cluster. The

objectives were to review and update current knowledge of

the structure and expression profiles of these genes and

of their mouse counterparts and to determine, via a

bioinformatics approach, whether other FMO genes are

present in the human and mouse genomes.

Results and conclusions We have identified, for the first

time, a mouse Fmo6 gene. In addition, we describe a novel

human FMO gene cluster on chromosome 1, located 4 Mb

telomeric of the original cluster. The novel cluster contains

five genes, all of which exhibit characteristics of

pseudogenes. We propose the names FMO 7P, 8P, 9P, 10P

and 11P for these genes. We also describe a novel mouse

gene cluster, located approximately 3.5 Mb distal of the

original gene cluster on Chromosome 1. The novel mouse

cluster contains three genes, all of which contain full-

length open-reading frames and possess no obvious

features characteristic of pseudogenes. One of the genes

is apparently a functional orthologue of human FMO9P. We

propose the names Fmo9, 12 and 13 for the novel mouse

genes. Orthologues of these genes are also present in rat.

Sequence comparisons and phylogenetic analyses

indicate that the novel human and mouse gene clusters

arose, not from duplications of the known gene cluster, but

via a series of independent gene duplication events. The

mammalian FMO gene family is thus more complex than

previously realised. Pharmacogenetics 14:117–130 &

2004 Lippincott Williams & Wilkins

Pharmacogenetics 2004, 14:117–130

Keywords: FMO, flavin-containing monooxygenase, gene family,pseudogene, chromosome 1, evolution, human, mouse

aDepartment of Biochemistry and Molecular Biology, University College London,London, UK and bSchool of Biological Sciences, Queen Mary, University ofLondon, UK.

This study was supported by the Wellcome Trust, grant number 053590; P.C. isa recipient of a PhD studentship from the MRC, UK.

Correspondence to Elizabeth Shephard, Department of Biochemistry andMolecular Biology, University College London, Gower Street, London WC1E6BT, UK.Tel: + 44 20 76792321; fax: +44 20 76797193; e-mail: [email protected] Ian Phillips, School of Biological Sciences, Queen Mary, University of London,Mile End Road, London E1 4NS, U.K.Tel: +44 20 78826338; Fax +44 20 89830973; email: [email protected]

Received 20 August 2003Accepted 2 November 2003

IntroductionThe flavin-containing monooxygenases (FMOs) are,

after the cytochromes P450 (CYPs), the second largest

group of enzymes involved in the Phase I metabolism

of drugs and other xenobiotics [1–3]. FMOs catalyse

the NADPH-dependent oxidation of a wide range of

structurally diverse compounds that contain, as the site

of oxidation, a soft nucleophilic heteroatom, usually

nitrogen, phosphorus or sulphur [1]. Substrates for

FMO include drugs, such as tricyclic antidepressants,

antihistamines and inhibitors of monoamine oxidase A

and B [4]; dietary-derived compounds (e.g., trimethyla-

mine) [5]; and neurotoxins (e.g. N-methyl-1,2,3,6-tetra-

hydropyridine, MPTP) [6]. Cysteamine is an

endogenous substrate for both pig liver FMO1 [7] and

yeast FMO [8] and it has been suggested that the

oxidation of this compound to the disulphide may

provide a significant source of disulphide to maintain

the cellular thiol:disulphide balance [9]. Although the

FMOs exhibit activity towards a wide substrate range

their specificity is determined by the size of their

substrate cleft, which dictates entry or exclusion of a

particular chemical [2,10,11].

In humans, mutations in FMO3 cause the inherited

disorder trimethylaminuria [12]. Affected individuals

are unable to metabolize dietary-derived trimethyla-

mine (TMA) into TMA N-oxide, and consequently

excrete relatively large amounts of the malodorous

free amine in their urine, sweat and breath, which

imparts an unpleasant body odour reminiscent of

rotting fish. The disorder has thus been colloquially

named fish-odour syndrome. Mutations of this gene

are described in the trimethylaminuria database at

http://human-fmo3.biochem.ucl.ac.uk/Human_FMO3/

[13]. For recent reviews on the clinical and bio-

chemical aspects of trimethylaminuria see [14] and

[15].

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Original article 117

0960-314X & 2004 Lippincott Williams & Wilkins DOI: 10.1097/01.fpc.0000054152.92680.ea

To date, six FMO genes, named FMOs 1, 2, 3, 4, 5 and

6, have been identified in humans [16] (http://www.

sanger.ac.uk/HGP/Chr1/). In this paper we review and

update our current knowledge of these genes and of

their counterparts in the mouse, which is increasingly

being used as a model organism to investigate the role

of specific gene products in health and disease. We

demonstrate that the six genes have been conserved in

the two species during evolution. We also describe

novel FMO gene clusters identified on chromosome 1

of human and mouse.

MethodsRNA isolation and northern blot hybridization

Human FMO mRNA expression in endocrine tissues

was analysed using the Human Endocrine System

MTN Blot (BD CLONTECH UK, Basingstoke, UK).

Hybridization probes were cDNAs for FMO1 [17],

FMO2 [18], FMO3 [19], FMO4 [20] and FMO5 [21].

Total RNA was isolated from liver, lung, kidney, brain,

heart, testis and ovary of 129Sv and C57BL/6 adult (8–

9 weeks postpartum) or juvenile (5 weeks postpartum)

mice through the use of the ULTRASPEC RNA

isolation system (Biotecx, Houston, TX, USA). RNA

samples (30 �g) were denatured in formaldehyde and

electrophoresed through a 1% (w/v) agarose gel [22].

RNA was transferred to an optimized nylon membrane

(BDH, Lutterworth, UK). The hybridization probe

used to detect mouse FMO1 mRNA was a PCR-

amplified product derived from the 39UTR of the gene.

Hybridization conditions were as previously described

[17]. Expression data for human and mouse FMO

mRNAs were also obtained from the internet sites at

http://bioinformatics.weizmann.ac.il/cards/; http://www.

ncbi.nlm.nih.gov/UniGene and from the BLASTN

facility at http://www.ncbi.nih.gov/

Isolation of the mouse Fmo gene cluster

A mouse genomic DNA PAC library (RPC121) [23] was

obtained from the UK Human Genome Mapping

Project Resource Centre (HGMP) (Hinxton, UK) as

gridded filters. Each filter was screened, under moder-

ate stringency conditions (hybridization at 508C, final

wash at 508C in 0.1 3 standard saline citrate (SSC),

0.1% sodium dodecyl sulphate (SDS)), with a mixture

of human FMO1, 2, 3, 4 and 5 cDNAs [17–21]. Sixteen

positive PAC clones were identified and subsequently

obtained from the HGMP. DNA was isolated from all

16 clones and used to prepare five replica, dot-blot

filters. These were hybridized, under high stringency

conditions (hybridization at 658C, final wash at 658C in

0.1 3 SSC, 0.1% SDS), to individual known mouse

FMO cDNAs. Mouse I.M.A.G.E. cDNA clones were

obtained from the HGMP. FMO1: I.M.A.G.E: 1921107,

Accession No. AI316256; FMO2: I.M.A.G.E: 1432164,

Accession No. AA986385; FMO3: I.M.A.G.E: 1891165,



Accession No. AI322352. Plasmid DNAs were isolated

and their inserts sequenced (MWG Biotech, Germany).

The I.M.A.G.E. clones for FMOs 1, 2, 3 and 5 each

contain the entire coding sequence for the correspond-

ing protein. However, I.M.A.G.E: 692387 contains only

sequences derived from exons 1–6 of FMO4, plus 465

bp of intron 6.

PAC clones that gave a positive signal when hybridized

with individual mouse cDNAs were characterized by

restriction endonuclease digestion and Southern blot

analyses.

Identification and analysis of novel FMO genes of human,

mouse and rat

Novel FMO sequences were identified by a BLASTN

search of the human (http://www.ensembl.org/Homo_

sapiens/) and mouse (http://www.ensembl.org/Mus_

musculus/) genomes with cDNA sequences encoding

FMO5 of human (Accession No. NM_001461) and

mouse (Accession No. AI322352). The genomic region

containing the mouse Fmo6 gene was identified using

the human FMO6 putative cDNA as the search query.

Contig sequences identified by positive BLAST hits

were analysed using the ClustalW and Pustell matrix

programs of the MacVector 6.5.3 analysis package

(Oxford Molecular Ltd., Oxford, UK). Exonic se-

quences were located by comparison to the sequences

encoding cDNAs of human FMO5, Accession No.

NM_001461 [24]; mouse FMO5, Accession No.

U90535 [25]; human FMO3, Accession No. NM006894

[19]; and mouse FMO3, Accession No. U87147 [26].

Alignment of exonic sequences was optimized by

visual inspection and the putative cDNA sequences

assembled. In the case of pseudogenes exon 9 se-

quences were terminated at a position equivalent to

that of the stop codon of FMO5 of human and mouse.

The rat genome (http://www.ensembl.org/Rattus_nor-

vegicus/) was searched by BLASTN with assembled

cDNA sequences corresponding to the novel mouse

genes.

Splice sites were analysed and scored for their corre-

spondence to a consensus sequence using the Splice

Predictor Program [27] (available at http://www.fruitfly.

org/seq_tools/splice.html). This program scores sequen-

ces on a scale of 0 to 1.0, with scores closest to 1.0

being closest to the consensus.

Alignment of multiple protein or DNA sequences was

done with ClustalW (MacVector 6.5.3). Alignments

were edited manually to ensure the correct placement

of conserved protein motifs, and of exon boundaries of

pseudogenes that lack particular exons. The file was

converted for phylogeny analysis using the READSEQ

program developed by D.G. Gilbert (Indiana Univer-


118 Pharmacogenetics 2004, Vol 14 No 2

sity, Bloomingdale, IN, USA) (available at http://

bimas.dcrt.nih.gov/molbio/readseq/). The PAUP*4.0

beta 10 software package [28] was used for phyloge-

netic analyses and drawing of trees. The DNA tree was

constructed using a heuristic search and maximum

likelihood with settings corresponding to the HKY85

model.

Results and discussionThe FMO gene family of human and mouse – update

In this section we review and update our knowledge,

gained from experimental evidence and genome se-

quencing projects, of the structure and expression

profiles of FMO 1, 2, 3, 4, 5 and 6 genes. In mouse the

orthologous genes are denoted Fmo1, 2, 3, 4, 5 and 6.The identification of a gene and putative cDNA for

mouse FMO6 is described below.

cDNAs have been isolated for human FMOs 1 [17], 2

[18], 3 [19,29], 4 [20] and 5 [24]. FMO6 was identified

from the genome sequencing project (see below).

When first identified, the cDNA for human FMO4 was

given the name FMO2 [20]. Subsequently, FMOs were

re-named to reflect the chronological order in which the

proteins were identified [30].

Mouse FMO1, 3 and 5 cDNAs were cloned by conven-

tional means [25,26,31], FMO2 was identified from an

I.M.A.G.E. clone [32] and we have isolated a cDNA for

FMO4 (Accession No. AF461145) [33].

FMO gene structure

Experimental evidence and information derived from

genome sequencing projects shows that the FMO1, 2, 3,4 and 5 genes contain eight coding exons (numbered 2

to 9), the size and boundaries of which are highly

conserved, and at least one 59 non-coding exon (num-

bered 1) (Table 1) [34].

In contrast to other FMO mRNAs, FMO4 mRNAs

contain sequences derived from 10, not 9, exons, the

first two of which are entirely non-coding (Table 1). To

maintain the correlation with exons of other mamma-

lian FMOs, we have named the most upstream of the

FMO4 exons, exon 0. In the FMO4 gene of humans,

exons 0 and 1 are separated by an intron of 1772 bp.

Sequences derived from all 10 exons of FMO4 are

spliced to form the 4.3-kb mRNA present in human

liver [20]. However, the transcript of FMO4 is subject

to differential tissue-specific processing: in the pancreas

the predominant mRNA is 2.4 kb, whereas in other

tissues examined the major species is 4.3 kb (Fig. 1). It

is not known, however, whether the FMO4 transcript

undergoes tissue type-specific alternative splicing in

human brain, as has been observed in rat brain [35].

Analysis of mouse FMO4 cDNA clones shows that

alternative promoters are used in the transcription of

the Fmo4 gene. In a cDNA isolated from the amnion of

a 17-day-pregnant female (RIKEN clone, Accession

No. BY738979), sequences derived from exon 1 are

spliced to sequences from an upstream exon, which we

have named exon 0a. Exon 1 and exon 0a are separated

by an intron of 926 bp. A second cDNA, isolated from

mouse kidney (I.M.A.G.E. clone 4234388; Accession

No. BF782473), does not contain sequences from exon

0a, but instead has 59 sequence encoded by an exon,

which we have named 0b, which lies upstream of exon

0a. The two alternative exons (0a and 0b) are separated

by an intron of 42 bp.

The human FMO1 gene contains two non-coding exons


Table 1 FMO genes of human and mouse

Gene Non-coding exons Coding exonsAmino acid

residuesMajor expressionsite in adult Other sites of expression

FMO1 0, 1 2–9 532 Kidney Gut, pancreas, adrenal cortex and medulla, thymus, thyroid, testis,placenta, lung, hypothalamus, breast, brain tumours, fetal kidney, liver.

Fmo1 1 2–9 532 Kidney Embryonic and mammary tissues, brain, lung, heart, ovaries, testis, fetalkidney, liver.

FMO2 1 2–9 471a or 535b Lung Skeletal muscle, kidney, prostate gland, blood vessels, tumours.Fmo2 1 2–9 535 Kidney Liver, lung, mammary tissue, tumours.FMO3 1 2–9 532 Liver Lung, kidney, adrenal medulla and cortex, pancreas, thyroid, gut, brain.Fmo3 1 2–9 534 Liver (female) Lung, retina, blood vessels, fetal liver, male liver (up to 5 weeks of age).FMO4 0, 1 2–9 558 Kidney Liver, lung, spleen, testis, gut, adrenal medulla and cortex, pancreas,

thyroid, thymus, placenta, brain, tumours (parathyroid, melanoma, lung,adenocarcinoma).

Fmo4 0b, 0a, 1 2–9 561 Kidney Liver, colon, egg, amnion.FMO5 1 2–9 533 Liver Kidney, lung, gut, mammary gland, adipose tissue, spleen, pituitary,

placenta, brain, testis, pancreas.Fmo5 1 2–9 534 Liver Kidney, gut, lung, adrenal and pituitary glands, thymus, skin, eyeball,

uterus, ovary, bone marrow, bladder, eggs, epididymis, blastocysts,embryonic stem cells, fertilized eggs, brain, spinal cord.

aFMO2*2A allele; bFMO2*1 allele.

FMO genes and pseudogenes of human and mouse Hernandez et al. 119

(0 and 1), which are used in a tissue-specific manner to

create mRNAs with different leader sequences (Fig. 2).

Examination of available human FMO1 cDNA clones

shows that in the liver exon 0 is spliced to exon 2

(Accession No. NM 002021), whereas in the small

intestine exon 1 is spliced to exon 2 (Accession No.

AK097039). Exon 0 is separated from exon 1 by an

intron (numbered 0) of 9120 bp, and exon 2 is

separated from exon 1 by a 346-bp intron (numbered

1). A leader sequence of a third cDNA (Accession No.

BC047129), isolated from a library made from pooled

adult colon, kidney and stomach RNAs, is derived from

intron 1 sequence. A similar situation is seen in the

mouse, where all three published cDNAs isolated from

kidney have leader sequences derived from intron 1

(Accession Nos. B5784152; CB954312; CB599568) (Fig.

2). The presence of the intronic leader sequences may

be the result of incomplete splicing of transcripts

derived from promoters associated with exons 0 or 1.

Alternatively, they may be derived from a more down-

stream promoter situated within intron 1.

The Fmo1 gene of mouse, unlike that of humans, has

only one non-coding exon (numbered 1) (Fig. 2). All

cDNAs isolated to date, from mouse liver and lung,

contain sequences derived from exon 1. Comparison of

mouse Fmo1 to the rabbit FMO1 gene, which, like

human FMO1, has exons 0 and 1 [36], reveals that in

the mouse gene, exon 1 is composed of sequences

corresponding to exon 0, intron 0 and exon 1 of the

rabbit gene (Fig. 3). In mouse, the site corresponding

to the acceptor splice site of intron 1 of the rabbit gene

is mutated to GG (Fig. 3). Although this would

preclude splicing in mouse of ‘exon 0’ to ‘exon 1’

sequences, it would not prevent splicing of ‘exon 0’ to

exon 2, as is seen in human (Fig. 2) and rabbit [36].

Expression profiles of human and mouse FMOs 1, 2, 3, 4

and 5

To gain a better understanding of the tissue-specific

patterns of expression of FMO genes of human and

mouse we have used northern blot analysis (Figs 1 and

4) and expressed sequence tag (EST) information

obtained from databases (see Methods). A summary of

the sites of expression of the FMO genes is given in

Table 1.

In human and mouse the FMO5 gene is expressed in

many adult and fetal tissues and in the embryo, and is

the most ubiquitously expressed member of the FMOgene family. FMO5, however, does not catalyse the

oxidation of any of the diverse range of xenobiotic

chemicals that are substrates for other members of the

FMO family, except for short-chain amines [24]. This,

together with the widespread distribution of FMO5,

suggests that it may be more involved in endogenous

metabolism than in the detoxification of xenobiotics.

The expression in human and mouse of FMO4 is also

widespread. However, with the exception of the human

pancreas, the expression of this gene is low (Fig. 1).

The expression of FMO3 is more restricted and none of


FMO1

FMO3

FMO4

FMO5

2.8 kb

2.8 kb

4.3 kb

2.4 kb

4.3 kb

3 kb

1 2 3 4 5 6 7 8

Fig. 1

Northern blot of human RNA. Autoradiograms of human poly Aþ RNAisolated from the pancreas (1), adrenal medulla (2), thyroid (3), adrenalcortex (4), testis (5), thymus (6), small intestine (7) and stomach (8).The northern blot was hybridized separately to cDNA probes for humanFMO1, FMO3, FMO4 and FMO5. Arrows indicate the size of thedetected mRNA species.

HUMAN

MOUSE

0 1 2

1 2

9.1 kb

fetal liver

small intestine

6.7 kb

liverlung

Fig. 2

Origin of alternative leader sequences in human and mouse FMO1mRNAs. The number above each black box indicates the relevant exon.The alternative use is shown of exons 0 and 1 in human fetal liver andsmall intestine, respectively. The white boxes adjacent to exon 2 showthe region of intron that has been found in the leader sequence ofmouse kidney cDNA clones and in a cDNA clone isolated from pooledhuman kidney, stomach and colon. The size of the most 59 intron of theFMO1 genes in human and mouse is indicated.


the ESTs for this protein is derived from fetal or

embryonic tissues. In both humans and mouse FMO3expression is triggered at birth [37,38]. In adult male

mouse, however, the expression of the gene is subse-

quently switched off in the liver [38,39]. In contrast,

the pattern of Fmo3 expression in female mouse liver

during development is similar to that observed in

humans.

Many ESTs have been isolated for human FMO2. The

source of these ESTs indicates that the FMO2 gene is

expressed in a relatively small number of tissues.

Genotyping shows that Caucasians and Asians are

homozygous for an allele named FMO2*2A, which

contains a non-sense mutation at codon 472 [18].

Twenty-six percent of African-Americans, however,

have at least one copy of the FMO2*1 allele, which

encodes a full-length protein of 535 amino acid resi-

dues, designated FMO2.1 [40], that is catalytically

active [18]. Recently S-oxygenation activity was

demonstrated in lung microsomes isolated from a

heterozygous individual (FMO2*1/FMO2*2A) [41]. In-

dividuals possessing the FMO2*1 allele would be

expected to differ from those who are homozygous for

the nonfunctional FMO2*2A allele in their ability to

metabolize drugs and other xenobiotics that are sub-

strates for FMO2. The expression profile of FMO2indicates that such differences would be manifest not

only in the lung, the major site of FMO2 expression,

but also in tissues such as skeletal muscle, kidney and

prostate gland.


mouse AACGCTTCTTATCTCTTAGACCAGGCTCATCCTGTGTGCAGGGAACTCCCAAAC------ rabbit aacattccttatctcttagactgggttcatccggtgtgctgtgagctcccagaccaacag

mouse -ACTGGCTCAAA---ACAATTTTGATTGTTACTAGCTCTGGGATCCTAATTGTGTCAGGG rabbit cactgcctcacagtgaccattttGATCTCTGCTAGCTCCAGAATCCTAACTGTGGCAGGG

mouse CCACTGGAATGAGCAAATTACAGCTACTTAGAGTCAGTAACCCATAAAAATCTCCGATTC rabbit TTCTAAGGCACAAGGAATCACAAGTACACACAgtaagtaatcagatatactctccaattc

mouse CTCTGGGTGAAGAAGGTGGAGCTGAGTTTTTCCTGTTTTTTCTTATTGTACATCAGAGAG rabbit ccttggacgcaga-ggtggatggggacttttcctgttctgt--tattgtgcttcagagag

mouse --GAGTCGTCTCAACAGAACACAGCTCCCTCCACGAGTGACCCTGGGGGAATTTTTCGAC rabbit cagaatcacctaaccagatcacaactccctctgtccatgaccctggaagAACTGTAAAGC

mouse CATCCCTTGCTCCAACGTAAGGAACAGAACTTGAGACCTTGTCACAGGAACATAAAGTCA rabbit CACTCCCTGCTTCGAAGTCCAGGA-GGAGCTCCCGTACCTGTAACGGGACAGTAAAGGCA

mouse GATTG---CTAAACTTCTGCATTTACTGgt intron 1 > rabbit GATATTATCGAGGCTTTTGTGTACGTCGgt intron 1 >

Fig. 3

The non-coding exons of the mouse and rabbit FMO1 genes. The underlined lower case letters indicate the promoter region, the non-underlinedlower case letters intron 0, and the shaded lower case letters intron 0 donor and acceptor splice sites of the rabbit FMO1 gene. The mutated mouseacceptor splice site is denoted by the bold lettered GG.

Liver

Lung

Kidney

Brain

1 2 3 4 5 6

3.3 kb

3.3 kb

3.3 kb

3.3 kb

3.3 kb

4 kb

5 2 4 6 5 6

Testis Ovaries Heart

Fig. 4

Northern blot of mouse RNA. Autoradiograms of total RNA isolatedfrom various tissues of adult male 129Sv (1), adult female 129Sv (2),adult male C57BL/6 (3), adult female C57BL/6 (4), juvenile male129Sv (5), juvenile female 129Sv (6). Total RNA (30 �g) waselectrophoresed and hybridized with a cDNA probe for mouse FMO1.Arrows indicate the size of the mRNA species detected.


Unlike the situation in humans and some other mam-

mals, such as rabbit [42], FMO2 does not represent a

major isoform in mouse lung [33]. Mouse FMO2 was

first cloned from a female kidney cDNA library [32]

and most of the reported ESTs that match it exactly

are from kidney libraries, whereas only a few are from

libraries prepared from lung.

In both human and mouse (Fig. 4), FMO1 is a

predominant isoform in adult kidney [19] [43]. FMO1expression, although high in the liver of adult mouse

and other mammals, is switched off at birth in this

tissue in humans [16,37]. FMO1 activity has also been

detected in microsomal membranes isolated from hu-

man intestine [43].

Many substrates for FMO3 are also substrates for

FMO1. However, although the two enzymes exhibit

similar activities towards some compounds, such as

benzydamine [44] and tazarotene, a retinoid used for

the treatment of acne and psoriasis [45], they can differ

markedly in their ability to catalyse the oxygenation of

others, for example, trimethylamine [46] and methio-

nine [47]. The difference in the hepatic expression of

FMO3 between humans and other mammals thus has

implications for the extrapolation of drug metabolic

data from experimental animals to humans.

The FMO6 genes of human and mouse

Human FMO6

The Sanger Centre, as part of their chromosome 1

sequencing project, identified and mapped clone

PAC127D3 (Accession No.ALO21026), which con-

tained FMO2, FMO3 and a third, previously unknown,

FMO gene (http://www.sanger.ac.uk/HGP/Chr1/). This

gene, which is more similar to FMO3 than to FMOs 1,2, 4 and 5, was named FMO6 [48]. The gene has 9

exons, the exon–intron boundaries of which, with the

exception of the donor site for exon 9, are conserved

with those of FMO3. In FMO6 the acceptor site for

exon 9 is mutated from AG to AT. When the expres-

sion of FMO6 mRNA was examined in liver and

kidney, by reverse transcriptase (RT)–PCR, nine tran-

scripts, all shorter than that predicted, were observed

[49]. These transcripts were produced by the skipping

of exons and/or the use of alternative splice donor or

acceptor sites in introns 3, 4, 6 and 8, and none is likely

to encode a functional FMO protein.

Our analysis of available human FMO6 cDNAs indi-

cates that the production of aberrant mRNAs occurs

also in an endometrial adenocarcinoma (Accession nos.

AI978664 and BX282694) and in human lung epithelial

cells (Accession nos. CB853631, CA944858, BU684071),

through the use of alternative acceptor sites within

exon 9. In BU684071 the acceptor site, relative to the

first nucleotide of the translational start codon of

correctly spliced FMO3 mRNA, is at nucleotide 1287,

whereas in the other four ESTs an AG located at

position 1380 is used. The Splice Site Predictor Pro-

gram predicts a third splice site in exon 9, at position

1345, with a score of 0.70. However, no cDNAs using

this splice site have been isolated. As FMO6 exhibits

the properties of a pseudogene we suggest re-naming it

as FMO6P.

Mouse Fmo6

The mouse Fmo6 gene was identified by a BLASTN

search of the mouse genome with the human FMO6

putative cDNA sequence. An alignment to a region on

Chromosome 1, between the Fmo2 and Fmo3 genes,

was identified. The presence of Fmo6 was confirmed

experimentally by analysing the clone PAC 544G8,

which was known to contain Fmo2 and Fmo3, by PCR

with primers derived from the predicted exon 7

sequence, which are specific for Fmo6 and do not

amplify Fmo3 sequences (data not shown). By aligning

the selected genomic sequence to both human FMO6

and mouse FMO3 cDNAs, and comparison to the

intron/exon structure of known Fmo genes, we derived

a putative processed transcript, of 1599 bp, that con-

tained an open-reading frame of 533 codons (Fig. 5).

This derived cDNA has been assigned the Accession

No. BK001544. The amino acid sequence of the

putative polypeptide encoded by the open-reading

frame is 84% identical to that derived for human

FMO6, 76 and 72% identical to mouse and human

FMO3, respectively, but less than 60% identical to

FMOs 1, 2, 4 and 5 of mouse or human. The novel

mouse gene is therefore orthologous to human FMO6,and has thus been designated Fmo6. All introns of

Fmo6 begin and end, respectively, with GT and AG,

and there are no obvious features that would preclude

the production of a functional polypeptide.

To investigate whether the transcript of mouse Fmo6,like that of its human orthologue, may be subject to

incorrect splicing, we analysed available cDNA clones.

The Splice Site Predictor Program predicts two po-

tential acceptor sites for exon 4, one at the expected

exon-intron boundary (score 0.97) and the other 71 bp

upstream of this (score 0.98). Interestingly, both sites

are used. An I.M.A.G.E. clone (Accession No.

BQ715367) is spliced correctly between exons 3 and 4,

whereas a second clone (Accession No. BQ892177),

isolated from an olfactory epithelium library, was found

to contain an insertion of 71 nucleotides between

sequences derived from exons 3 and 4. The remaining

exons (5, 6 and 7) reported for the two I.M.A.G.E.

clones were spliced at the expected exon-intron bound-

aries. High splice site scores were obtained for the

predicted donor sites of exons 2 and 9 (between 0.99

and 1.00). The donor and acceptor site scores of exon 8

were both low, but predicted at positions identical with



those of other FMOs. Further experimental analysis is

required to establish whether the Fmo6 gene of mouse

should be classified as a gene or a pseudogene.

The FMO1, 2, 3, 4 and 6 genes are located on chromosome

1 of human and mouse

FMOs 1, 2, 3 and 4 were mapped experimentally to the

long arm of human chromosome 1, in the region 1q23–

25 (Fig. 6) [17,20,50,51], and, as described above,

FMO6 was identified by the Sanger Centre, as part of

their chromosome 1 sequencing project. The order of

the genes was determined experimentally by Southern

blot hybridization of independent YAC clones (data not

shown) and is the same as that determined by the

human genome sequencing project. The five genes

span a region of about 220 000 bp (Fig. 6). The most


FAD FMO3 MKKK-VAIIGAGVSGLAAIRSCLEEGLEPTCFERSDDVGGLWKFSDHIEEGRASIYQSVFTNSSKEMMCFPDFPYPDDFP FMO6 MGKK-VAIVGAGVSGLAAIRCCLEEGLDPICFERSIDVGGLWKFSSHAEEGRASIYQSVFTNSSKEMMCFPDFPYPDDFP FMO9 MVKKQIAVIGAGISGLGAIKCCLDEDLEPTCFERNDDIGGLWKFQKNASEKMPSIYRSVTINTSKEMMCFSDFPIPDHFP FMO12 MGVKRIAVIGAGVSGLGAIKCCLEEGLEPTCFEKKSDIGGLWKYEEIPKSGNLGIYKSLTCNTSKEMTAFSDYPIPDHYP FMO13 MEVKQIAIIGAGVSGLGAIKSCLEEGLEPTCFEKSNDIGGLWRYKETPENGRPGIYKSLTCNTSKEMTTFSDYPIPDHYP FMO5 MAKKRIAVIGAGASGLTCIKCCLEEGLEPVCFERSGDIGGLWRFQEAPEEGRASIYQSVVINTSKEMMCFSDYPIPDHYP

FMO3 NFMHHSKLQEYITSFAKEKNLLKYIQFETPVTSINKCPNFSTTGKWEVTTEKHGKKETAVFDATMICSGHHIFPHVPKDS FMO6 NYMHHSKLQEYITSFAQKKGLLRYIQFETLVSSIKKCSSFLTTGQWVVVTEKEGKQESVLFDAVMICSGHHVYPNMPTDS FMO9 NYMHNSKLMDYFRMYAKRFSLLDYIRFKTTVRSVRKRPDFHIHGQWDVVVETDGKQESLVFDGVLVCSGHHTDPHLPLKS FMO12 NYMHHSKMMEYLRMYARHFGLMKHIQFQTNVCNIKKRPDFSSSGQWDVVVETEEMQKTYIFDGIMVCSGHYTEKYFPLQD FMO13 NYMHHSKMMEYLRMYARHFGLMKHIQFQTRVCVVRKRPDFSSSGQWDVVVEADGKQKNYIFDGVMVCSGHYTEKYLPLQD FMO5 NYMHNSQVLEYFRMYAKEFDLLKYIQFKTTVCSVKKQPDFSTSGQWQVVTECEGKQQVDVFDGVLVCTGHHTDAHLPLES

NADPH FMO3 FPGLNRFKGKCFHSRDYKEPGIWKGKRVLVIGLGNSGCDIAAELSHVAQKVTISSRSGSWVMSRVWDDGYPWDMVVLTRF FMO6 FPGLEHFRGKCLHSRDYKGPGAFQGKKVLVIGLGNSASDIAVELSRLATQVIISTRSGSWIMSRVWNDGYPWDMVYVTRF FMO9 FPGIEKFEGCYFHSREYKSPEDYVGKRIIVVGIGNSGVDIAVELGRVAKQVFLSTRRGSWILHRVWNNGYPMDSSFFTRF FMO12 FEGISKFQGSYLHTWEYKHPDNFVGKRVAVIGLGNSGADVAGEISRVADQVFLSTRQGAWIWNRVWDHGEPMDTTVFTRY FMO13 FAGISKFQGSCLHSWEYKHPDSFVGKRVVVIGIGNSGADVANEISCVTEQVFLSTRRGTWIWNRVWDNGDPLDIALFTRY FMO5 FPGIEKFKGKYFHSRDYKNPVEFTGKRVIVIGIGNSGGDLAVEISHTAKQVFLSTRRGAWILNRVGKHGYPIDLLLSSRI

FMO3 QTFLKNNLPTAISDWWYTRQMNARFKHENYGLVPLNRTLRKEPVFNDELPARILCGMVTIKPNVKEFTETSAVFEDGTMF FMO6 TSFLRNILPSFVSDWLYIKKMNTWFKHENYGLMPLNGPLRKEPVFNDELPSRILCGMVTIKPIVTKFTETSAVFEDGTVF FMO9 HSFLQKILTTEAVNKYLEKTLNSRFNHAHYGLQPQHRPLSQHPTISDDLPNHIISGKVQVKPNVKEFTGTDVHFDDGTVE FMO12 NRAVQKICPRYIINRQMEKKLNGRFNHANYGLLPTHRILEQRTVLSDDLPNRIIIGKVKIKPNVKEFTSTSAIFEDG-TK FMO13 NRTVKSFYPTFLINRWTENKLNLRFNHANYGLQAKHRFLSHQSIFSDDLPNRIISGRVLVKTNVREFTSTSAIFEDG-SE FMO5 MYYLSRICGPSLKNNYMEKQMNQRFDHEMFGLKPKHRALSQHPTVNDDLPNRIIAGLVKVKGNVKEFTETAAIFEDGSRE

FMO3 EAIDCVIFATGYGYAYPFLDDSIIKSRNNEVTLYKGVFPPQLEKPTMAVIGLVQSLGATIPITDLQARWAAQVIKGTCTLL FMO6 EAIDCVIFATGYGYAYPFLDDSIIKSRNNEVTLYKGIFPPQLEKPTMAVIGLVQSLGAAIPTADLQARWAAKVFTSTCVLL FMO9 ENIDVVIFATGYSISFPFLGDLIA-VTDNEVSLYKLMFPPDLEKPTLAVIGLIQPLGIILPIAELQSRWAVRVFKGLSKLL FMO12 ENIDVVIFATGYKLSFPFLSDDSG-VLDNQYSMFKYVFPPELEKPTLAFIGILQPAGAILPTSELQSRWVVHVFKGIKKLL FMO13 EIVDVVVFATGYTLSFPFLDDSSE-ILDSKHTMFKFVFPPQLEKPTLAFIGILQPIGATIPTSELQSRWVTRVFAGLQKLL FMO5 DGIDVVIFATGYSFAFPFLEDSVK-VVKNKVSLYKKVFPPNLEKPTLAIIGLIQPLGAIMPISELQGRWATQVFKGLKKL

FMO3 PSVNDMMDDIDEKMGEKFKWYG---NSTTIQTDYIVYMDELASFIGAKPNLLWLFLKDPRLAVEVFFGPCSPYQFRLVGP FMO6 PTTNEMMDDIDEKMGKKLKWFG---QSHTLQTDYITYMDELSSFIGAKPNIPWLFLTDPQLALEVYFGPCSPYQFRLMGP FMO9 PSVKAMKADMDQR-KKAMEKRYVKTARHTIQVDHIEYMDEIASLAGVKPNLLLLFLSDPTLAMEVFFGPCTPYQYRLQGP FMO12 PSRRAMIADINRKNHQIMAKGSKKILQDHRRVTFVDYMDEIASEIGVKPNLLSLLLWDTKLAKEVFCGPCTSYQYRLQGP FMO13 PSQSNMMADINRK-KRKMEKEFVKSPRDVRRVPYIDYMDEIASEIGVKPNLLSFFLWDTKLAKEIFWGPCTPYQYRLQGP FMO5 PSQSEMMAEINKA-REEMAKRYVDSQRHTIQGDYIDTMEEIADLVGVRPNIQPLVFTDPRLALRLLLGPCTPVQYRLQGP

FMO3 GKWSGARNAILTQWDRSLKPMKTRVVSKVQKSC--SHFYSRLLRLLAVPVLLIALFLVLI FMO6 GKWDGARNAILTQWKRTVKPTRTRAVGEAQR----PRHLYDLLRMLFFPVFFLAVLLTFY FMO9 GKWDGARRAILTQRERIIKPLKTRITSEKSRSAPGLFWIKMALFGLAFLVPSLTYFSYICQ FMO12 GKWDGARAAILTQRERMLKPLRTRVVKQSHLS--HLSWVKSACVVVFFFISVVVIMHITSH FMO13 GKWAGARAAILTQRDRILKPLRSRVLKNSETSSSSLFWVRCICAVIFPFVSVFAIIHAIYQ FMO5 GKWAGARKTILTTEDRVRKPLMTRVVERDSSGG-SLVTVRVLMLAVAFFAVILAYF

Fig. 5

Sequence comparison of mouse FMOs 3, 5, 6, 9, 12 and 13. The GXGXXG/A motifs, characteristic of FAD-pyrophosphate and NADPH-pyrophosphate-binding sites, and the conserved EGLEP and FATGY sequences are shown as shaded boxes.



FMO3 FMO6 FMO2 FMO1 FMO4

26.9 kb

20 kb

23.2 kb 25.6 kb 37.4 kb 27.6 kb

24 kb 40 kb 20 kb

�220 kb

q21.1 q24.2 q24.3

�22 Mb �4 Mb

FMO5

39 kb

Human Chromosome 1

FMO7p FMO8p FMO9p FMO10p FMO11p

�320 kb

? kb 14.5 kb 18.9 kb 17.1 kb 29.4 kb

Mouse Chromosome 1

Fmo13 Fmo12 Fmo9150 kb 30 kb

13 kb 13.2 kb 17 kb�210 kb

H2H1

�3.5 Mb

�200 kb

Fmo4 Fmo1 Fmo2 Fmo6 Fmo320.5 kb 10.5 kb 28.3 kb 6.9 kb

15.9 kb 36.7 kb 23.8 kb 20.5 kb 29.1 kb

Mouse Chromosome 3F2

Fmo5

24.3 kb

(a)

(b)

83 kb 81 kb 18 kb 81 kb

Fig. 6

Diagrammatic representation of the chromosomal location of human and mouse FMO genes. (a) The location of FMO5 and of the two gene clustersare shown in their cytogenetic bands on human chromosome 1. Each gene is represented by a filled square, with its size given below. Theapproximate sizes of each cluster and the distances between them (www.ensembl.org) are shown. (b) The location of Fmo5 on Chromosome 3 andof the two Fmo clusters on Chromosome 1 are shown in their cytogenetic bands. Each gene is represented by a filled square, with its size givenbelow. The size of the two clusters and the distance between them are also given. The arrows represent the direction in which the genes aretranscribed. The diagrams are not drawn to scale.


centromeric gene in the cluster is FMO3, followed by

FMO6, FMO2, FMO1 and FMO4. The genes are all

transcribed in the centromeric to telomeric direction.

To determine the chromosomal location of the mouse

genes, Fmos1, 2, 3, 4 and 6, we identified PAC clones,

from a mouse genomic library, that hybridized to mouse

Fmo cDNA sequences. The complement of Fmo genes

in each PAC was assessed by PCR, using primers based

on the mouse cDNA sequences. Three clones were

found to be positive for both Fmo1 and Fmo4, two were

positive for both Fmo2 and Fmo3, and two were posi-

tive for Fmo4 only (data not shown), indicating that, as

is the case in humans, in the mouse Fmo4 lies next to

Fmo1, and Fmo2 lies next to Fmo3. Analysis by

fluorescence in situ hybridization (FISH), using the

PAC clones as probes, located the Fmo 1, 2, 3 and 4genes in a cluster towards the distal end of mouse

Chromosome 1 (data not shown) in a region that is

syntenic with human chromosome 1 [52].

From comparative mapping data [52], it would appear

that the whole of the region of synteny between human

chromosome 1 and mouse Chromosome 1 is inverted in

relation to the telomere and centromere. Thus, in this

region, genes that are more centromeric in humans are

more distal in mice and this appears to be the case for

the FMO clusters of mouse and man. The order and

orientation of the five genes within the cluster, as

indicated by our analysis of YAC and PAC clones and

the order and orientation derived by the Sanger Centre,

is cent-FMO3-FMO6-FMO2-FMO1-FMO4-tel in humans,

whereas in mouse the genes are orientated cent-Fmo4-Fmo1-Fmo2-Fmo6-Fmo3-tel and are transcribed from

the telomere to the centromere (Fig. 6).

The FMO5 genes of human and mouse

Of the three independent PAC clones that were posi-

tive for Fmo5, none contained any other Fmo gene

(data not shown), indicating that in mouse, as is the

case in humans, this gene is separated from the main

Fmo cluster. FISH analysis mapped Fmo5 to mouse

Chromosome 3 (data not shown). In humans FMO5 is

located on the long arm of chromosome 1 in the region

1q21.1 [51,53], between the genes PRKAB2 and BCL9,a region syntenic to mouse Chromosome 3.

Identification of novel FMO genes in human and mouse

A second FMO gene cluster on human chromosome 1

Through an analysis of the human genome we have

identified a second FMO gene cluster on the long arm

of chromosome 1, located approximately 4.27 Mb closer

to the centromere than the cluster that contains FMOs1, 2, 3, 4 and 6. This second cluster contains five genes,

which span a region of about 330 000 bp. By comparison

with both FMO5 and FMO3 exonic sequences, potential

exons were identified in each of the five genes. All five

genes have the characteristics of pseudogenes (see

below). The potential DNA coding sequence of each

pseudogene is between 57 and 65% identical to the

genes for FMO1, FMO2, FMO3, FMO4, FMO5 and

FMO6 (Table 2). No pseudogene is therefore obviously

derived from any of the six known FMO genes. We

therefore suggest the names FMO7P, FMO8P, FMO9P,FMO10P and FMO11P for these novel sequences.

FMO7P is the most centromeric and FMO11P the most

telomeric of the genes in this cluster (Fig. 6A). The

evidence for the existence of these genes and the

features that characterize them as pseudogenes are

described below. Unless stated otherwise, the reference

numbers given to locate each of the genes on chromo-

some 1 (http://www.sanger.ac.uk/HGP/Chr1/) are as

follows: the first indicates the A of the ATG start codon

and designates the beginning of the potential protein-

coding sequence; the second defines the position, in

exon 9 of the pseudogene, that corresponds to that of

the last nucleotide of the stop codon in FMO5.

Human FMO7P This gene is the most centromeric in

the cluster. Only sequences corresponding to exons 3, 4,

5 and 7 could be assigned with confidence. The 59 end

of exon 3 and the 39 end of exon 7 correspond, respec-

tively, to nucleotides 162 094 790 and 162 101 690 of

chromosome 1. No sequence corresponding to exon 6

was identified, even after a rigorous analysis of the 2769

nucleotides that separate exons 5 and 7. Translation of

sequences with some similarity to exons 2, 8 and 9

indicated possible remnants of these three exons. How-

ever, as they could not be assigned with confidence,

these regions were not used in the calculation of nucleo-

tide identities or for phylogenetic analyses (Table 2 and

Fig. 7).

Human FMO8P Eight potential exons, that together

are 57 to 60% identical to the 8 coding exons of the

known human FMO genes (Table 2), were identified

for the second gene in the cluster, FMO8P. This gene

is located between nucleotides 162 185 557 and

162 200 081 of chromosome 1. In this region of the

chromosome, a novel transcript, ENST0000031389,

comprising four potential exonic sequences is predicted

by the Sanger Centre. The first two exons of this

transcript correspond to exons 4 and 5 of a FMO gene,

whereas the last two exons do not correspond to FMOsequences. The predicted transcript contains a 19 449-

bp intron between its third and fourth exons. Closer

analysis of this ‘intronic’ sequence revealed that it con-

tained sequences corresponding to exons 6, 7, 8 and 9 of

a FMO gene. Exon 3 of FMO8P is located within the

predicted transcript ENST0000031389 and is separated

from exon 4 by an intron of 455 bp. Exon 2, which is not

contained within ENST0000031389, is separated from

exon 3 by an intron of 3637 bp. The acceptor splice site

for exon 7 is mutated to CG and thus a transcript



encoding a functional FMO protein would not be pro-

duced. In addition, exons 6 and 7 contain in-frame stop

codons, exon 8 a frame-shift mutation and exon 9 a

premature stop codon.

Human FMO9P FMO9P lies between nucleotides

162 231 846 and 162 250 846. This region of the genome

is annotated with two overlapping transcripts,

NM_138784 and a novel predicted transcript ENST

00000322141 (http://www.ensembl.org/Homo_sapiens/).

NM_138784 is the transcript of the Ensembl gene

ENSG00000143151 and comprises six exons. Sequence

comparisons reveal that four of these correspond to

exons 2, 3, 4 and 5 of a FMO gene, whereas the other

two do not correspond to FMO sequences. The novel

predicted transcript ENST00000322141 contains exons

7, 8 and part of 9 of a FMO gene. Further examination of

the genomic region between these two transcripts re-

vealed a FMO exon 6 sequence, located between the

predicted sixth exon of NM_138784 and the predicted

first exon of ENST00000322141. Exon 6 of FMO9P is

flanked by both mutated acceptor (AT) and donor (AT)

splice sites, which accounts for the lack of detection of

exon 6 by the automated transcript assembly programs

used by the genome consortium. In addition, exon 2 of

FMO9P has a 1-bp deletion, which causes a shift in the

reading frame and a premature stop signal at codon 7.

The gene is thus unable to produce a functional FMO

protein.

Transcript NM_138784 is related to I.M.A.G.E. clone

3997249, which was isolated from the bladder carci-


Table

2DNAse

quence

identity

betw

eenFMOsofhumanandmouse

FMO1

FMO2

FMO3

FMO4

FMO5

FMO6P

FMO7P

FMO8P

FMO9P

FM010P

FMO11P

FMO1

FMO2

FMO3

FMO4

FMO5

FMO6

FMO9

FMO12

FMO1

–FMO2

65

–FMO3

62

63

–FMO4

61

63

60

–FMO5

60

63

61

61

–FMO6P

63

63

74

60

60

–FMO7P

59

61

57

57

65

56

–FMO8P

57

60

59

59

60

57

59

–FMO9P

59

61

60

62

65

59

63

65

–FMO10P

57

59

58

57

62

57

61

71

65

–FMO11P

60

59

59

58

64

59

55

78

67

69

–FMO1

85

62

62

59

58

62

57

56

57

56

56

–FMO2

62

83

62

61

62

62

57

59

62

58

59

61

–FMO3

63

64

83

60

61

75

56

58

60

57

60

63

64

–FMO4

60

63

60

83

60

59

55

57

60

56

57

59

61

60

–FMO5

60

63

61

60

84

60

64

59

65

61

61

58

61

60

58

–FMO6

62

61

74

60

59

84

54

57

59

56

58

61

61

76

59

60

–FMO9

59

61

61

61

65

60

62

64

83

64

67

57

62

61

59

64

60

–FMO12

58

59

59

57

62

58

58

69

64

75

73

57

58

59

57

63

58

65

–FMO13

59

60

58

57

62

60

59

69

65

77

74

59

60

59

57

63

58

66

80

Sha

ded

area

s¼

hum

an;u

nsha

ded

area

s¼

mo

use.

Per

cent

iden

titie

sw

ere

bas

edo

nth

ep

rote

in-c

od

ing

reg

ions

ofa

llkn

ow

nex

ons

.The

sew

ere

3to

5an

d7

forFMO7P

;2to

7an

d9

forFMO10P

;2,3

and

6to

9fo

rFMO11P

;an

d2

to9

for

allo

ther

gen

es.

FMO4Fmo4

FMO5

Fmo5

Fmo9

FMO9P

FMO10P

Fmo12

Fmo13FMO8P

FMO11p

FMO7PFmo1

FMO1

Fmo6

FMO6P

FMO3

Fmo3

Fmo2FMO2

Fig. 7

Phylogenetic tree of the FMO genes of human and mouse.


noma clone library NIH_MCG_53. Because of the

frame-shift in exon 2 of FMO9P the predicted start

codon of the I.M.A.G.E. clone lies downstream, within

exon 3 of FMO9P, and is out of frame. Thus no section

of the amino acid sequence predicted from this cDNA

corresponds to a FMO protein.

Human FMO10P The FMO10P gene lies between

nucleotides 162 269 228 and 162 302 376. Analysis of this

genomic sequence revealed seven regions, which to-

gether have 57 to 62% nucleotide sequence identities to

exons 2 to 7 and exon 9 of known FMOs (Table 2). Exon

8 could not be assigned with certainty and is not

included in the phylogenetic analyses described below.

An in-frame stop signal at codon 152 precludes the

formation of functional protein. A number of other

mutations are present in exons 7 and 9.

Human FMO11P FMO11P, the most telomeric gene in

the cluster, is located between nucleotides 162 383 413

and 162 413 114 of chromosome 1. Exons 2 to 9 were

identified. However, exon 4 is corrupted by the insertion

of a short, 38-bp repeat sequence that is present many

times in the human genome, and therefore this exon

was not included in further analyses described below.

Exon 5 too, is poorly conserved, although the last 17

codons can be aligned with exon-5 sequences of other

FMO genes. This exon also was excluded from phyloge-

netic analyses. The gene contains a number of other

mutations. For example, exon 2 has a 1-bp deletion

resulting in a change in the reading frame, from codon

36, and the formation of an in-frame stop codon in exon

3.

A second mouse Fmo gene cluster on Chromosome 1

We have identified three novel Fmo genes within

the mouse genome (http://www.ensembl.org/Mus_

musculus/). The three genes span about 210 000 bp and

form a cluster on mouse Chromosome 1 in band H2.

They lie about 3.5 Mb, closer to the telomere than the

known Fmo cluster and are transcribed from the

telomeric to centromeric direction.

As described below, all three of the genes encode full-

length open-reading frames (Fig. 5) and possess no

obvious features that would categorize them as pseudo-

genes. Thus, until experimental evidence proves other-

wise, they should be regarded as functional and,

consequently, we propose that their names should not

be appended with a ‘p’ for pseudogene. Exonic regions

of the most distal gene of the cluster (Fig. 6B) are 83%

identical at the nucleotide level to exonic sequences of

the human pseudogene FMO9P. This is the same

degree of identity exhibited by orthologous FMOs of

human and mouse (Table 2). The mouse gene is

apparently a functional orthologue of human FMO9Pand we therefore propose that it be designated Fmo9.

The exonic regions of the other two genes in the

cluster are 80% identical to each other at the nucleotide

level. The human genes to which they have greatest

similarity are the pseudogenes FMO10P and 11P. The

exonic regions of the most proximal and middle genes

in the cluster, respectively, are 77 and 75% identical to

FMO10P and 74 and 73% identical to FMO11P. It is

not clear from this degree of sequence identity whether

either of the mouse genes is orthologous to one of the

human pseudogenes, and, if so, which would be the

orthologous pair(s). Therefore we propose that they be

named Fmo12, for the middle gene of the cluster, and

Fmo13, for the most proximal gene (Fig. 6B). Each of

the three genes is described below. We give the

location of the genes by chromosome nucleotide num-

bers. The first number indicates the A of the ATG start

codon and designates the beginning of the potential

protein-coding sequence; the second defines the posi-

tion of the last nucleotide of the stop codon (http://

www.ensembl.org/Mus_musculus/).

Mouse Fmo9 The Fmo9 gene lies between nucleotides

167 349 105 and 167 332 118 and is designated as tran-

script ENSMUST00000027843. A cDNA clone (Acces-

sion No. NM_172844), isolated from a 0-day neonate

head library, corresponds exactly to the eight exonic

sequences of Fmo9 and contains an open-reading frame

of 539 codons (Fig. 5). The deduced amino acid se-

quence of the predicted protein (Fig. 5) is about 65%

identical to that of FMO5. This demonstrates that Fmo9can produce a correctly spliced mRNA and is therefore

not a pseudogene in the mouse.

Three other transcripts have been predicted for this

region of the genome (http://www.ensembl.org/Mus_

musculus/genome). Two of these are partial transcripts,

ENS MUST00000027843 and ENSMUST0000002784,

which overlap with NM_172844, and lack, respectively,

exons 3 and 8 and exons 6 and 8. The third, EN-

SMUST00000027842, comprises 10 exons. Analysis of

this sequence showed that the predicted transcript is a

chimera of exons 2 to 7 of Fmo9 and exons 8 and 9 of

Fmo12, with a predicted intron of 189019 bp between

exons 7 and 8 (see below). None of these transcripts is

likely to exist.

Mouse Fmo12 Fmo12, the middle gene of the cluster,

lies between nucleotides 167 312 150 and 167 298 987

and is located within the 59 region of the predicted

189 019-bp intronic sequence of transcript EN-

SMUST00000027842 (see above). The nucleotide se-

quences of the 8 coding exons of Fmo12 have overall

identities of 57 to 63% to the coding exons of Fmos 1, 2,3, 4, 5 and 6 (Table 2). There are no obvious aberrant

splice donor or acceptor sites. The gene is predicted to

encode a polypeptide of 537 amino acid residues (Fig. 5)

that is 46 to 58% identical to FMOs 1, 2, 3, 4, 5, 6 and 9



(data not shown). The derived cDNA has been assigned

the Accession No. BK001546.

A novel transcript, ENSMUST00000046662, predicted

from the region in which Fmo12 lies, is derived from 6

exons. Five of these correspond to exons 2, 3, 4, 5 and

7 of the Fmo12 gene, whereas the sixth is 53 bp and is

not related to a Fmo sequence. The Splice Site

Predictor Program gives donor site scores for exons 2, 3,

4, 5, 6, 7 and 8 as 0.96, 0.99, 0.75, 0.65, 0.90, 0.98 and

1.00, respectively. Relatively high scores, of 0.85 and

0.79, respectively, are predicted for sites 179 bp down-

stream of exon 4 and 43 bp downstream of exon 8. As

no ESTs have been reported for Fmo12 it is not known

whether these predicted alternative splice sites are

used. The splice donor sites for exons 3, 4, 5, 6, 8 and 9

have scores of 0.92, 0.95, 0.99, 0.83, 0.95 and 0.60,

respectively. The program did not predict a donor site

for exon 7, although a consensus AG is present at the

correct position in the gene. A site with a high score of

0.91 was predicted 35 bp upstream of the beginning of

exon 7. However, the expected splice junctions for this

exon are identified in the predicted transcript EN-

SMUST00000046662.

Mouse Fmo13 Fmo13 lies at the proximal end of the

gene cluster between nucleotides 167 160 135 and

167 144 915 (Fig. 6B). Three transcripts have been pre-

dicted from the region in which this gene is located:

ENSMUST00000027842 contains exons 8 and 9 of a

Fmo gene; ENSMUST00000027844 has 4 exons, three

correspond to exons 6, 7 and 8 of a Fmo gene, whereas

the fourth does not correspond to a Fmo sequence;

ENSMUST00000061918 has 4 exons, corresponding to

Fmo exons 3, 4, an extended 5 and 7. Close analysis of

the genomic region from which these transcripts are

predicted revealed a Fmo exon 2, 1370 bp upstream of

exon 3. All donor splice sites begin with GT and all

acceptor sites end with AG. The Splice Site Predictor

Program gives a donor splice site score for exon 2 of

0.98. It is possible, therefore, that a mRNA derived from

exons 2 to 9 of Fmo13 could be produced. The mRNA

would encode a polypeptide of 539 amino acid residues

that is 47 to 58% identical to FMOs 1, 2, 3, 4, 5, 6 and 9,

and 73% identical to FMO12. The derived cDNA has

been assigned Accession No. BK001545. No EST clones

derived from this gene have been reported.

Evolution of FMO genes

The nucleotide sequences of members of the human

pseudogene cluster (FMOs 7P, 8P, 9P, 10P and 11P) aremore similar to each other than to members of the

known gene cluster (FMOs 1, 2, 3, 4 and 6) (Table 2).

This indicates that, although both clusters contain five

genes, the pseudogene cluster did not arise via a

complete duplication of the gene cluster. Instead, it

appears to have arisen via a series of independent gene

duplication events. This also seems to be the case for

members of the novel mouse gene cluster (Fmos 9, 12and 13) (Table 2). The novel human and mouse genes

are more similar to FMO5 than to FMOs 1, 2, 3, 4 and 6(Table 2).

Phylogenetic analyses of human and mouse FMOs,

based on nucleotide (Fig. 7) and amino acid sequences

(data not shown), suggest that an ancestral gene gave

rise, via a series of duplications, to five genes, FMO1,FMO2, the precursor of FMOs 3 and 6, FMO4, and

another gene. Soon after, this last gene gave rise to

FMO5, FMO7P, and a third gene, which subsequently

duplicated to yield FMO9 and the precursor of the

remaining FMO genes and pseudogenes. A series of

duplications of the latter gene gave rise to FMOs 8P,10P, 11P, 12 and 13.

From the calculated rate of evolution of FMOs [16] andtheir amino acid sequence identities, we can estimate

the time at which some of the gene duplications took

place. The duplications that gave rise to FMOs 1, 2, theprecursor of 3 and 6, 4, and the precursor of the

remaining FMO genes are estimated to have occurred

some 275 million years ago (mya). FMOs 5 and 9diverged about 210 to 240 mya, whereas FMOs 3 and 6,and 12 and 13, arose some 140 mya. Thus, the most

recent common ancestor of human and mouse, and

indeed of all placental mammals (about 80 mya), is

predicted to have had a cluster containing genes for

FMOs 1, 2, 3, 4 and 6, a separate gene for FMO5, and

a second cluster, containing genes for FMOs 9, 12 and

13. It is possible that the second cluster contained

additional genes, subsequently lost in the mouse line-

age, or that genes arose specifically in the human

lineage and then became pseudogenes.

Human FMO9P and mouse Fmo9 are clearly ortholo-

gues (Table 2, Fig. 7). Although it is not possible to

assign orthologues of Fmos 12 and 13 unambiguously,

FMO10P and FMO11P are, respectively, approximately

76 and 74% identical to Fmos 12 and 13 (Table 2).

Given that 10P and 11P are pseudogenes, and thus

would have accumulated mutations more rapidly since

losing their function, it is possible that they may

represent human orthologues of mouse Fmos 12 and 13.

Analysis of the rat genome reveals that it contains a

cluster of orthologues of mouse Fmos 9, 12 and 13 (data

not shown). Two of these, rat FMOs 9 and 12, containopen-reading frames that encode polypeptides that are

92 and 89% identical to their mouse orthologues.

FMO13, however, contains a premature stop codon and

is apparently a pseudogene in rat and thus should be

designated FMO13P.

The mammalian FMO gene family is thus more com-



plex than previously realized. In addition to the main

gene cluster of FMOs 1, 2, 3, 4 and 6, and the separate

FMO5, which appear to be present in all mammals

investigated to date, a second cluster, containing genes

encoding three novel FMOs (9, 12 and 13), is present

in rodents, and a pseudogene cluster containing five

members (FMOs 7P, 8P, 9P, 10P and 11P), is present inhumans.

References1 Ziegler DM. Recent studies on the structure and function of multi-

substrate flavin-containing monooxygenases. Ann Rev Pharmacol Toxicol1993; 33:179–199.

2 Rettie AE, Meier GP, Sadeque AJ. Prochiral sulfides as in vitro probes formultiple forms of the flavin-containing monooxygenase. Chem Biol Inter-act 1995; 96:3–15.

3 Cashman JR. Stereoselectivity in S- and N-oxygenation by the mammalianflavin-containing and cytochrome P-450 monooxygenases. Drug Met Rev1998; 30:675–707.

4 Cashman JR. Human flavin-containing monooxygenase: substrate specifi-city and role in drug metabolism. Curr Drug Metab 2000; 1:181–191.

5 Marks R, Greaves MW, Prottey C, Hartop PJ. Trimethylaminuria: the useof choline as an aid to diagnosis. Br J Dermatol 1977; 96:399–402.

6 Cashman JR, Ziegler DM. Contribution of N-oxygenation to the metabo-lism of MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine) by variousliver preparations. Mol Pharmacol 1986; 29:163–167.

7 Duffel MW, Logan DJ, Ziegler DM. Cysteamine and cystamine. MethodsEnzymol 1987; 143:149–154.

8 Suh JK, Poulsen LL, Ziegler DM, Robertus JD. Redox regulation of yeastflavin-containing monooxygenase. Arch Biochem Biophys 2000;381:317–321.

9 Ziegler DM, Duffel MW, Poulsen LL. Studies on the nature and regulationof the cellular thio:disulphide potential. Ciba Found Symp 1979; 2:191–204.

10 Guo WX, Ziegler DM. Estimation of flavin-containing monooxygenaseactivities in crude tissue preparations by thiourea-dependent oxidation ofthiocholine. Anal Biochem 1991; 198:143–148.

11 Kim YM, Ziegler DM. Size limits of thiocarbamides accepted as sub-strates by human flavin-containing monooxygenase 1. Drug Metab Dispos2000; 28:1003–1006.

12 Dolphin CT, Janmohamed A, Smith RL, Shephard EA, Phillips IR.Missense mutation in flavin-containing mono-oxygenase 3 gene, FMO3,underlies fish-odour syndrome. Nat Genet 1997; 17:491–494.

13 Hernandez D, Addou S, Lee D, Orengo C, Shephard EA, Phillips IR.Trimethylaminuria and a human FMO3 mutation database. Hum Mutat2003; 22:201–213.

14 Mitchell SC, Smith RL. Trimethylaminuria: the fish malodor syndrome.Drug Metab Dispos 2001; 29:517–521.

15 Cashman JR, Camp K, Fakharzadeh SS, Fennessey PV, Hines RN,Mamer OA, et al. Biochemical and clinical aspects of the human flavin-containing monooxygenase form 3 (FMO3) related to trimethylaminuria.Curr Drug Metab 2003; 4:151–170.

16 Phillips IR, Dolphin CT, Clair P, Hadley MR, Hutt AJ. McCombie RR,et al. The molecular biology of the flavin-containing monooxygenases ofman. Chem Biol Interact 1995; 96:17–32.

17 Dolphin C, Shephard EA, Povey S, Palmer CN, Ziegler DM, Ayesh R,et al. Cloning, primary sequence, and chromosomal mapping of a humanflavin-containing monooxygenase (FMO1). J Biol Chem, 1991;266:12379–12385.

18 Dolphin CT, Beckett DJ, Janmohamed A, Cullingford TE, Smith RL,Shephard EA, et al. The flavin-containing monooxygenase 2 gene (FMO2)of humans, but not of other primates, encodes a truncated, nonfunctionalprotein. J Biol Chem 1998; 273:30599–30607.

19 Dolphin CT, Cullingford TE, Shephard EA, Smith RL, Phillips IR.Differential developmental and tissue-specific regulation of expression ofthe genes encoding three members of the flavin-containing monooxygen-ase family of man, FMO1, FMO3 and FM04. Eur J Biochem 1996;235:683–693.

20 Dolphin CT, Shephard EA, Povey S, Smith RL, Phillips IR. Cloning,primary sequence and chromosomal localization of human FMO2, a newmember of the flavin-containing mono-oxygenase family. Biochem J 1992;287:261–267.

21 Janmohamed A, Dolphin CT, Phillips IR, Shephard EA. Quantification andcellular localization of expression in human skin of genes encoding flavin-

containing monooxygenases and cytochromes P450. Biochem Pharma-col 2001; 62:777–786.

22 Fourney RM, Miyakoshi J, Day RS, Paterson MC. Northern blotting:efficient RNA staining and transfer. Focus (BRL) 1988; 10:5–7.

23 Osoegawa K, Tateno M, Woon PY, Frengen E, Mammoser AG, CataneseJJ, et al. Bacterial artificial chromosome libraries for mouse sequencingand functional analysis. Genome Res 2000; 10: 116–128.

24 Overby LH, Buckpitt AR, Lawton MP, Atta-Asafo-Adjei E, Schulze J,Philpot RM. Characterization of flavin-containing monooxygenase 5(FMO5) cloned from human and guinea pig: evidence that the uniquecatalytic properties of FMO5 are not confined to the rabbit ortholog. ArchBiochem Biophys 1995; 317:275–284.

25 Cherrington NJ, Falls JG, Rose RL, Clements KM, Philpot RM, Levi PE,et al. Molecular cloning, sequence, and expression of mouse flavin-containing monooxygenases 1 and 5 (FMO1 and FMO5). J Biochem MolToxicol 1998; 12:205–212.

26 Falls JG, Cherrington NJ, Clements KM, Philpot RM, Levi PE, Rose RL,et al. Molecular cloning, sequencing, and expression in Escherichia coli ofmouse flavin-containing monooxygenase 3 (FMO3): comparison with thehuman isoform. Arch Biochem Biophys 1997; 347:9–18.

27 Reese MG, Eeckman FH, Kulp D, Haussler D. Improved splice sitedetection in Genie. J Comput Biol 1997; 4:311–323.

28 Swofford, DL, PAUP*: Phylogenetic Analysis Using Parsimony (and OtherMethods) 4.0 Beta. 2002, Sunderland, MA: Sinauer Associates, Inc.

29 Lomri N, Gu Q, Cashman JR. Molecular cloning of the flavin-containingmonooxygenase (form II) cDNA from adult human liver. Proc Natl AcadSci U S A 1992; 89:1685–1689.

30 Lawton MP, Cashman JR, Cresteil T, Dolphin CT, Elfarra AA, Hines RN,et al. A nomenclature for the mammalian flavin-containing monooxygenasegene family based on amino acid sequence identities. Arch BiochemBiophys 1994; 308:254–257.

31 Itoh K, Nakamura K, Kimura T, Itoh S, Kamataki T. Molecular cloning ofmouse liver flavin containing monooxygenase (FMO1) cDNA and char-acterization of the expression product: metabolism of the neurotoxin,1,2,3,4-tetrahydroisoquinoline (TIQ). J Toxicol Sci 1997; 22:45–56.

32 Karoly ED, Rose RL. Sequencing, expression, and characterization ofcDNA expressed flavin-containing monooxygenase 2 from mouse.J Biochem Mol Toxicol 2001; 15:300–308.

33 Janmohamed A, Hernandez D, Chandan P, Phillips IR, Shephard EA. Celltype-specific expression of the flavin-containing mono-oxygenase family ofmouse. submitted.

34 Dolphin CT, Riley JH, Smith RL, Shephard EA, Phillips IR. Structuralorganization of the human flavin-containing monooxygenase 3 gene(FMO3), the favored candidate for fish-odor syndrome, determineddirectly from genomic DNA. Genomics 1997; 46:260–267.

35 Lattard V, Longin-Sauvageon C, Benoit E. Cloning, sequencing andtissue distribution of rat flavin-containing monooxygenase 4: two differentforms are produced by tissue-specific alternative splicing. Mol Pharmacol2003; 63:253–261.

36 Luo Z, Hines RN. Identification of multiple rabbit flavin-containing mono-oxygenase form 1 (FMO1) gene promoters and observation of tissue-specific DNase I hypersensitive sites. Arch Biochem Biophys 1996;336:251–260.

37 Koukouritaki SB, Simpson P, Yeung CK, Rettie AE, Hines RN. Humanhepatic flavin-containing monooxygenases 1 (FMO1) and 3 (FMO3)developmental expression. Pediatr Res 2002; 51:236–243.

38 Falls JG, Blake BL, Cao Y, Levi PE, Hodgson E. Gender differences inhepatic expression of flavin-containing monooxygenase isoforms (FMO1,FMO3, and FMO5) in mice. J Biochem Toxicol 1995; 10:171–177.

39 Falls JG, Ryu DY, Cao Y, Levi PE, Hodgson E. Regulation of mouse liverflavin-containing monooxygenases 1 and 3 by sex steroids. Arch BiochemBiophys 1997; 342:212–223.

40 Whetstine JR, Yueh M, McCarver DG, Williams DE, Park C, Kang JH,et al. Ethnic differences in human flavin-containing monooxygenase 2(FMO2) polymorphisms: detection of expressed protein in African-Amer-icans. Toxicol Appl Pharmacol 2000; 168:216–224.

41 Krueger SK, Martin SR, Yueh MF, Pereira CB, Williams DE. Identificationof active flavin-containing monooxygenase isoform 2 in human lung andcharacterization of expressed protein. Drug Metab Dispos 2002; 30:34–41.

42 Williams DE, Ziegler DM, Nordin DJ, Hale SE, Masters BS. Rabbit lungflavin-containing monooxygenase is immunochemically and catalyticallydistinct from the liver enzyme. Biochem Biophys Res Commun 1984;125:116–122.

43 Yeung CK, Lang DH, Thummel KE, Rettie AE. Immunoquantitation ofFMO1 in human liver, kidney, and intestine. Drug Metab Dispos 2000;28:1107–1111.

44 Lang DH, Rettie AE. In vitro evaluation of potential in vivo probes for



human flavin-containing monooxygenase (FMO): metabolism of benzyda-mine and caffeine by FMO and P450 isoforms. Br J Clin Pharmacol2000; 50:311–314.

45 Attar M, Dong D, Ling KH, Tang-Liu DD. Cytochrome P450 2C8 andFlavin-containing monooxygenases are involved in the metabolism oftazarotenic acid in humans. Drug Metab Dispos 2003; 31:476–481.

46 Lang DH, Yeung CK, Peter RM, Ibarra C, Gasser R, Itagaki K, et al.Isoform specificity of trimethylamine N-oxygenation by human flavin-containing monooxygenase (FMO) and P450 enzymes: selective catalysisby FMO3. Biochem Pharmacol 1998; 56:1005–1012.

47 Ripp SL, Itagaki K, Philpot RM, Elfarra AA. Methionine S-oxidation inhuman and rabbit liver microsomes: evidence for a high-affinity methionineS-oxidase activity that is distinct from flavin-containing monooxygenase 3.Arch Biochem Biophys 1999; 367:322–332.

48 Phillips IR, Dolphin CT, McCombie RR, Beckett DJ, Janmohamed A,Smith RL, et al. Molecular genetics of the flavin-containing monooxy-genases. In First International Workshop on Trimethylaminuria 1999;Bethesda, MD, USA.

49 Hines RN, Hopp KA, Franco J, Saeian K, Begun FP. Alternative proces-sing of the human FMO6 gene renders transcripts incapable of encodinga functional flavin-containing monooxygenase. Mol Pharmacol 2002;62:320–325.

50 Shephard EA, Dolphin CT, Fox MF, Povey S, Smith R, Phillips IR.Localization of genes encoding three distinct flavin-containing mono-oxygenases to human chromosome 1q. Genomics 1993; 16:85–89.

51 McCombie RR, Dolphin CT, Povey S, Phillips IR, Shephard EA. Localiza-tion of human flavin-containing monooxygenase genes FMO2 and FMO5to chromosome 1q. Genomics 1996; 34:426–429.

52 Carver EA, Stubbs L. Zooming in on the human-mouse comparative map:genome conservation re-examined on a high resolution scale. GenomeRes 1997; 7:1123–1137.

53 Gelb BD, Zhang J, Cotter PD, Gershin IF, Desnick RJ. Physical mappingof the human connnexin 40 (GJA5), flavin-containing monooxygenase 5,and natriuretic peptide receptor genes on 1q21. Genomics 1997;39:409–411.



Organization and evolution of the flavin-containing monooxygenase genes of human and mouse:...

Documents

Transcript of Organization and evolution of the flavin-containing monooxygenase genes of human and mouse:...