Gene 275 (2001) 267 277
www.elsevier.com/locate/gene
Structure, chromosomal localization, and expression of the gene for mouse
ecto-mono(ADP-ribosyl)transferase ART5
Gustavo Glowackia, Rickmer Brarena, Marina Cetkovic-Cvrljeb, Edward H. Leiterb,
Friedrich Haaga, Friedrich Koch-Noltea,*
a
Institute for Immunology, University Hospital, Martinistrasse 52, 20246 Hamburg, Germany
b
The Jackson Laboratory, Bar Harbor, ME, USA
Received 7 May 2001; received in revised form 29 June 2001; accepted 11 July 2001
Received by A. Sippel
Abstract
Mono(ADP-ribosyl)transferases regulate the function of target proteins by attaching ADP-ribose to specific amino acid residues in their
target proteins. The purpose of this study was to determine the structure, chromosomal localization, and expression profile of the gene for
mouse ecto-ADP-ribosyltransferase ART5. Southern blot analyses indicate that Art5 is a single copy gene which maps to mouse chromosome
7 at offset 49.6 cM in close proximity to the Art1, Art2a and Art2b genes. Northern blot and RT-PCR analyses demonstrate prominent
expression of Art5 in testis, and lower levels in cardiac and skeletal muscle. Sequence analyses reveal that the Art5 gene encompasses six
exons spanning 8 kb of genomic DNA. The 50 end of the Art5 gene overlaps with that of the Art1 gene. A single long exon encodes the
predicted ART5 catalytic domain. Separate exons encode the N-terminal leader peptide and a hydrophilic C-terminal extension. Sequencing
of RT-PCR products and ESTs identified six splice variants. The deduced amino acid sequence of ART5 shows 87% sequence identity to its
orthologue from the human, and 37 and 32% identity to its murine paralogues ART1 and ART2. Unlike ART1 and ART2, ART5 lacks a
glycosylphosphatidylinositol-anchor signal sequence and is predicted to be a secretory enzyme. This prediction was confirmed by transfect-
ing an Art5 cDNA expression construct into Sf9 insect cells. The secreted epitope-tagged ART5 protein resembled rat ART2 in exhibiting
potent NAD-glycohydrolase activity. This study provides important experimental tools to further elucidate the function of ART5. q 2001
Published by Elsevier Science B.V.
Keywords: ADP-ribosylation; NAD-glycohydrolase; Gene family; Gene structure; Splice variants; Toxin homologue
1. Introduction of the target protein. ADP-ribosyltransferase activity of
bacterial toxins often is involved in the pathogenesis of
Mono(ADP-ribosyl)transferases (mADPRTs) are an disease (Moss and Vaughan, 1990; Aktories, 1991). For
important class of enzymes with known regulatory functions example, Diphtheria toxin ADP-ribosylates a diphthamide
as bacterial toxins and metabolic regulators (Moss and residue in elongation factor 2, thereby shutting off host cell
Vaughan, 1990; Aktories, 1991; Ludden, 1994). These protein synthesis (Honjo et al., 1968). Cholera and Pertussis
enzymes mediate the post-translational modification of toxins interfere with signal transduction by ADP-ribosylat-
specific target proteins by transferring the ADP-ribose ing the alpha-subunit of heterotrimeric G proteins at specific
moiety from NAD1 to specific amino acid residues in arginine or cysteine residues leading to uncoupling of
their target proteins. This usually inactivates the function surface receptors from their downstream effector molecules,
thereby affecting adenylate cyclase activity and ion flux
(Moss and Vaughan, 1990). Other toxins ADP-ribosylate
arginine residues in actin (C2, iota, SpvB, VIP2) and ras
Abbreviations: Gapd, glyceraldehyde 3 phosphate dehydrogenase; GPI,
(exoS) or asparagine residues in rho (C3) (Rappuoli and
glycosylphosphatidylinositol; mADPRT, mono(ADP-ribosyl)transferase;
Montecucco, 1997; Han et al., 1999; Otto et al., 2000). In
NAD1, nicotinamide adenine dinucleotide; PBS, phosphate-buffered
saline; PCR, polymerase chain reaction; RACE, rapid amplification of photosynthetic bacteria, a mADPRT (DRAT) regulates
cDNA ends; RT, reverse transcription; SDS-PAGE, sodium dodecyl sulfate
nitrogen fixation by ADP-ribosylating an arginine residue
polyacrylamide gel electrophoresis; UTR, untranslated region
of dinitrogenase reductase (Ludden, 1994).
* Corresponding author. Tel.: [phone]; fax: 149-40-
Mounting evidence indicates that ADP-ribosyltrans-
428034243.
ferases play important regulatory roles also in higher
E-mail address: [e-mail] (F. Koch-Nolte).
0378-1119/01/$ - see front matter q 2001 Published by Elsevier Science B.V.
PII: S0378-1119(01)00608-4
268 G. Glowacki et al. / Gene 275 (2001) 267 277
animals (Haag and Koch-Nolte, 1997). In the mouse, for with primers N11 and N41, purified and radiolabeled to high
example, cell surface mADPRTs have been implicated in specific activity (.108 cpm/mg). Hybridization and wash-
regulating myogenesis, long-term potentiation in hippocam- ing were performed essentially as described previously
pal neurons, and the activity of cytotoxic T cells (Zolk- (Koch-Nolte et al., 1995; Braren et al., 1998).
iewska and Moss, 1993; Schuman et al., 1994; Wang et
2.3. Isolation and sequencing of genomic and cDNA clones
al., 1996). The first mammalian mADPRT (ART1) was
cloned from rabbit skeletal muscle (Zolkiewska et al.,
A 129/SvJ mouse genomic P1 DNA library (Genome
1992). Related genes, designated ART2 ART7, have been
Systems, St. Louis, MO) was screened by PCR with primers
cloned from mouse, human, and chicken (Koch-Nolte and
N00 and N31. Purified P1 DNAs were subjected to restric-
Haag, 1997). We have previously cloned and characterized
tion mapping and suitable restriction fragments were
the mouse Art1 gene and mapped it to chromosome 7 at
subcloned into pBluescript (Stratagene) for sequence
offset 49.6 cM (Koch-Nolte et al., 1996a; Braren et al.,
analyses. Mouse heart and testis  marathon cDNA
1998). We now report the cloning, characterization, and
(BALB/c) were purchased from Clontech and subjected to
expression of the mouse Art5 gene and show that it maps
50 and 30 RACE reactions according to the manufacturer s
to the same region on mouse chromosome 7.
protocol with nested primers N32 and N33 for 50 RACE and
In the case of Art2, defects in gene structure and/or
N02 and N03 for 30 RACE. For defining the 50 cap site of
expression have been found to coincide with susceptibility
Art5 cDNA, testis CapSite cDNAe (Eurogentec) was
for autoimmune disease in certain rat and mouse laboratory
amplified with primers N32 and N07 according to the manu-
strains (Greiner et al., 1987; Koch-Nolte et al., 1995).
facturer s instructions. PCR amplification products were
Considering the presumptive regulatory role of mADPRTs,
cloned into the pCR2.1 vector (Invitrogen). All sequences
it is conceivable that defects in other gene family members
were obtained by dideoxy sequence analysis with appropri-
may also be of clinical relevance. The description of the
ate vector- and Art5-specific primers using the big dye-
mouse Art5 gene presented here provides the basis for
terminator sequencing kits (Applied Biosystems).
applying established transgene and knock-out technologies
Sequences obtained from PCR products were confirmed
to further analyze the function of the ART5 protein and to
by sequencing clones obtained from two separate PCR reac-
determine the functional significance of Art5 gene defects.
tions. The nucleotide sequences described here have been
deposited in the EMBL database (Accession numbers:
2. Materials and methods
AJ295722, Y08028, and Y16835).
2.1. PCR primers and PCR reactions
2.4. Amino acid sequence alignment and structure
prediction analyses
Primers derived from the sequence of the mouse Art5 gene
Multiple sequence alignments, sequence distance and
were as follows: N02, ACT CTC TGG AGT TAT GAT CAG
phylogeny calculations were performed with the DNAstar
ACC TG; N03, GCT GCA GCT CTC CAG AGC TGG ACC;
and Sequencher programs on a Macintosh computer. Signal
N00, AGG ATG ATT CTG GAG GAT CTG CTG ATG;
peptide cleavage sites were predicted with the Signal P
N33, GCG TGC AAG CTG AGG CAG CTG AG; N32,
program (www.cbs.dtu.dk/services/SignalP/).
CCA GGC CTC TTG TGC TGC TTC CCA; N99, CTG
CTT CCT GCA GCC GTT CAA AGC CC; N06, AGA
2.5. RT-PCR analyses
CAG ATT TGG CGA CTT AAC TAG C; N05, CTG ATC
TCA GGC CAG GAC TAG GC; N11, GAT ACC TTT GAT
PCR analyses were carried out on cDNAs from different
GAT GCC TAT GTG GGC TG; N41, GAC AAC GCC TGG
mouse tissues normalized with respect to the transcript
ATA GGA GCC CCA AAA CA. PCR reactions were carried
levels for six house-keeping enzymes (MTC panel, Clon-
out in 20 ml reaction volumes on purified plasmid or P1 DNA
tech). cDNA fragments were amplified with primer pairs
(0.1 10 ng) or cDNA (Clontech) (10 50 ng) in 1 Ł PCR
N00 and N99, N06 and N99, and N05 and N99. PCR reac-
buffer (Perkin Elmer) with 2.5 units of Amplitaq-Gold poly-
tions were performed in 20 ml reaction volumes with the
merase (Perkin Elmer) for 25 cycles (20 s at 958C, 20 s at
Advantage polymerase mix (Clontech), 0.2 mg template
558C, and 60 s at 728C). Samples were incubated for 8 min at
cDNA, and 100 ng of primers derived from two separate
948C prior to the first PCR cycle.
exons. The polymerase mix included the TaqStart antibody
(Clontech) for automatic hot start PCR. Cycling conditions
2.2. Southern and Northern blot analyses
were 948C for 20 s, 708C for 30 s, and 728C for 180 s for
Genomic DNAs were prepared as described (Koch-Nolte cycles 1 3; the annealing temperature was reduced to 658C
et al., 1995) or purchased from The Jackson Laboratory (Bar in cycles 4 6 and to 608C thereafter. Aliquots (6.5 ml) were
Harbor, ME), restriction digested and subjected to Southern removed after the 26th, 31st, and 36th cycles and analyzed
blot analysis. Northern blots were purchased from Clontech. by electrophoresis on 1% agarose gels and ethidium
A 545 bp fragment of mouse Art5 was generated by PCR bromide staining.
G. Glowacki et al. / Gene 275 (2001) 267 277 269
2.6. Expression of recombinant ART5 protein and enzyme fic polymorphism with those of other loci previously
assays mapped in the BSS interspecific cross clearly showed link-
age of Art5 to markers on medial chromosome 7 (Fig. 2),
Mouse ART1 and ART5 were produced as His6x FLAG-
just distal to where Art2a and Art2b had been previously
tagged recombinant proteins in Sf9 insect cells as described
mapped using the same backcross panel (Prochazka et al.,
previously for ART2.1 and ART2.2 (Koch-Nolte et al.,
1991; Koch-Nolte et al., 1996a). Art5 showed the identical
1996b). The coding region of mouse Art5 was amplified by
strain distribution pattern as Art1 (0/94 recombinants vs. 1/
PCR with primers BF6, ACT CGT TCT AGA ATG ATT
94 recombinants with the Art2a/Art2b tandem pair) (Koch-
CTG GAG GAT CTG CTG ATG and BR6, AAG AGG
Nolte et al., 1996a). The latter show no recombinants with
AGA TCT GGG TCC AGC TCT GGA GAG CTG; that of
the guanyl cyclase 2d (Gucyd2d) locus currently listed at
mouse Art1 was amplified with BF9, GTC ACC TCT AGA
offset 48 cM and Art5 shows no recombinants with the
ATG AAG ATT CCT GCT ATG ATG TCT and BR9L, GCA
cholecystokinin b receptor locus (Cckbr) at offset 49.6
GGA GGA TCC AAT GGA GCC TGG GGC TGA GCT AC.
cM (map positions were retrieved in June 1999 from the
Amplification products were digested with XbaI and BglII
Mouse Genome Database, MGD, at http://www.informa-
and cloned into a derivative of pVL1393 expression vector
tics.jax.org).
(Pharmingen) containing a C-terminal chimeric His6x- and
FLAG-tag. Sf9 and Hi5 cells were transfected with purified
3.2. Structure of the Art5 gene
plasmid (3 mg/3 Ł 106 cells), BaculoGold DNA (0.5 mg)
(Pharmingen), and Cellfectin (GIBCO/BRL). Cell superna-
We confirmed the suspicion that Art5 and Art1 are in
tants were harvested 4 days after transfection and used for
two amplification rounds of infection. Sepharose-immobi-
lized FLAG-tag specific monoclonal antibody M2 was
purchased from Sigma. Soluble or immobilized proteins
were incubated in 50 ml 20 mM Tris (pH 8.0), 1 mM ADP-
32
ribose, 5 mM DTT, 0.5 mCi P-labeled NAD1 (Amersham)
and 10 mM NAD1 (Sigma) for 60 min at 378C. Where indi-
cated, reactions also contained 2 mM agmatine (Sigma).
Proteins were analyzed by SDS-PAGE and Western blot
analyses; supernatants were analyzed by thin layer chroma-
tography as described previously (Haag et al., 1995; Koch-
Nolte et al., 1996b; Braren et al., 1998).
3. Results and discussion
3.1. Southern blot analyses show that Art5 is a single copy
gene in close linkage with the Art1 and Art2 genes on mouse
chromosome 7
Considering that a single Art1 gene and two tandemly
linked, functional Art2 genes have been observed in the
mouse (Prochazka et al., 1991; Braren et al., 1998), it was
of interest to determine whether the Art5 gene might also
have been duplicated. To this end we performed Southern
blot analyses of restriction digested genomic mouse and rat
DNAs with an Art5-specific probe (Fig. 1). A single band
was obtained with all of the enzymes tested except for those
Fig. 1. Southern blot analyses of the Art5 gene. (A) Genomic DNAs were
with a cognate site in the probe used (e.g. SacI: Fig. 1, lanes
digested with SacI (lanes 1 and 6), HindIII (lanes 2 and 7), BamHI (lanes 3
1 and 6). These results indicate that the Art5 gene is a single
and 8), EcoRI (lanes 4 and 9) and TaqI (lanes 5 and 10) and subjected to
copy gene in the mouse and in the rat.
Southern blot analysis with radiolabeled Art5-specific cDNA probe
With TaqI we observed a restriction fragment length N11xN41 (the localization of this probe in exon 4 is illustrated in Fig. 3).
DNAs were from a C57BL/6J mouse (lanes 1 5) and from the rat cell line
polymorphism in genomic DNAs from C57BL/6J and
C58NT (lanes 6 10). (B) Genomic DNAs were digested with HindIII (lanes
Mus spretus (SPRET/Ei) (1.7 vs. 2 kb). We used this
1, 4, and 7), PstI (lanes 2, 5, and 8), and BamHI (lanes 3, 6, and 9) and
RFLP to map the gene in the C57BL/6J Ł SPRET/Ei
subjected to Southern blot analysis as in (A). DNAs were from a farmhouse
(BSS) interspecific backcross panel (Rowe et al., 1994).
mouse caught in Northern Germany (lanes 1 3), from a C57BL/6J mouse
Comparison of the haplotype distribution of the Art5-speci- (lanes 4 6), and from inbred Mus spretus (lanes 7 9).
270 G. Glowacki et al. / Gene 275 (2001) 267 277
close physical proximity by molecular cloning and sequen- The results showed that the two genes are arranged in a head
cing of the Art5 gene. Primers N00 and N31 were derived to head fashion. The distance between the two polyadenyla-
from the mouse Art5 cDNA sequence (Okazaki et al., 1996; tion sites is approximately 15 kb. The finding that the P1
Haag and Koch-Nolte, 1997) and were used to isolate three genomic DNA clones containing the Art1 Art5 gene cluster
clones containing the Art5 gene from a 129/SvJ mouse do not overlap with P1 clones containing the Art2 gene pair
genomic P1 library by PCR screening. Given the close link- indicates that these respective Art gene pairs are separated
age of Art5 with Art1 and the two Art2 genes, we also by at least 40 kb.
probed these P1 clones with specific probes for each of Comparison of cDNA and genomic DNA sequences
the other genes. The results revealed that each of the three showed that the Art5 gene is composed of six exons
P1 clones does indeed contain Art5 and Art1 but not Art2a (Figs. 3B and 4). Fig. 3B shows a comparison of the
or Art2b. Three overlapping restriction fragments contain- Art5 exon/intron structure with those of the mouse Art1,
ing Art5 and Art1 were subcloned and sequenced (Fig. 3A). rat ART2b, and chicken ART7 genes (Davis and Shall,
1995; Haag et al., 1996; Braren et al., 1998). A common
feature of these genes is the presence of a long exon encod-
ing the predicted catalytic domain. The genes differ in the
50 and 30 regions. While the ATG initiation codon of
chicken ART7 is contained in the first exon, the 50 UTRs
of the mammalian ART genes are split into at least three
exons. Similarly, the C-terminal ends of the mammalian
ARTs are encoded by distinct exons whereas the C-term-
inal end of chicken ART7 is encoded by the same exon as
the catalytic domain (Fig. 3B). We note further that Art1
and Art5 both contain a small 30 exon (encoding seven and
ten amino acids, respectively) which is lacking in rat ART2
and chicken ART7 (Fig. 3B).
The finding that the Art1 and Art5 genes overlap at their
50 ends (Fig. 3A) suggests that their expression may be
regulated by a common promoter and/or regulatory
element(s). Indeed, Art1 and Art5 genes are coexpressed
in some tissues, e.g. heart and skeletal muscle (see Fig. 7).
Furthermore, the differently sized Art5 transcripts in testis
(1.6 kb) vs. skeletal muscle (1.35 kb) could reflect use of
alternative promoters, as has been described for rat ART2
(Kuhlenb�umer et al., 1997). The results of our 50 RACE
and RT-PCR analyses (see Fig. 8) support this: exon 1 of
Art5 is preferentially used in skeletal muscle, while Art5
transcripts in testis evidently derive from a transcription
start site at the beginning of a long version of exon 2.
Note that the testis-specific transcription start site of Art5
lies outside of the Art1 gene, while the muscle-specific tran-
scription start site overlaps with the 50 end of the Art1 gene
(Fig. 3A). Further studies will be required to define the
common and gene-specific regulatory elements in the
Fig. 2. Chromosomal mapping of the Art5 gene. Map figure (top) from The
Art1/Art5 gene pair.
Jackson BSS backcross panel showing part of chromosome 7 with loci
linked to Art5. The map is depicted with the centromere toward the top.
A 3 cM scale bar is shown to the right of the figure. Loci mapping to the
3.3. Features of the predicted Art5 gene product
same position are listed in alphabetical order. Raw data from The Jackson
Laboratory were obtained from http://www.jax.org/resourdes/documents/
Fig. 4 shows the nucleotide and deduced amino acid
cmdata. The known positions of human orthologues are depicted on the
sequences of the Art5 coding region and the primers used
left (top). Haplotype analyses are shown on the bottom. Each column
for PCR and sequence analyses. The sequence shown devi-
represents the chromosome identified in the backcross progeny that was
inherited from the (C57BL/6J Ł SPRET/Ei) F1 parent. Solid boxes indicate ates from the cDNA sequence published by Okazaki et al.
the presence of a C57BL/6J allele; open boxes indicate the presence of a
(1996) by two frame shift mutations and a dinucleotide
SPRET/Ei allele. Loci are listed in order with the most proximal at the top.
mutation. We obtained the same sequence as the 129/SvJ
The number of offspring inheriting each type of chromosome is listed at the
mouse genomic sequence shown in all available EST
bottom of each column. The percent recombination between adjacent loci is
sequences and in cDNA from BALB/c mouse. Most of
given to the right of the figure, with the standard error (SE) for each percent
recombination. the coding sequence for the predicted secretory protein
G. Glowacki et al. / Gene 275 (2001) 267 277 271
Fig. 3. Structure of the Art5 gene. (A) Map of the closely linked Art5 (right) and Art1 (left) genes. Restriction enzyme sites are marked by vertical bars (S, SacI;
E, EcoRI). Exons are boxed and coding regions are shaded. Fragments obtained by subcloning P1 DNA into plasmid vectors are shown on top with their
corresponding fragment length indicated in kilobases (kb). The PCR fragment obtained with primers N11 and N41 that was used as a probe for the Southern and
Northern blot analyses shown in Figs. 1 and 7 is indicated below exon 4 of Art5. Note that Art1 contains an exon (exon 1*) in addition to those reported by
Braren et al. (1998) (database Accession number: X95825). This exon is contained within the cDNA sequence obtained by Okazaki et al. (1996) (database
Accession number: U31510, residues 193 245) from a lymphoma cell line, but was contained neither in 50 RACE nor in PCR products obtained from skeletal
muscle cDNA (G.G. and R.B., unpublished data). (B) Schematic diagram of the exon/intron structure of the mouse Art5 gene compared to those of the mouse
Art1, rat ART2b, and chicken ART7 genes. Exons are boxed, coding regions are shaded and introns are shown as triangles. The exact lengths of exons and the
approximate lengths of introns are given in basepairs.
is contained within a single 730 bp long exon (exon 4 in contain two potential N-glycosylation sites, albeit at differ-
Fig. 3). Three separate exons encode the 50 UTR (exons ent positions: N102 and N197 in mART5; N65 and N248 in
1 3), the presumptive N-terminal leader peptide is encoded mART1. Both proteins also carry the characteristic R-S-
by the last 60 nt of exon 3, and the 30 UTR is encoded by EXE active site motif of arginine-specific mADPRTs
a single exon (exon 6). The predicted cleavage site for (R174, S184, and E220XE222 in ART5) (Domenighini
the signal peptide (A23/V24) is encoded at the 50 end of and Rappuoli, 1996; Koch-Nolte et al., 1996b).
exon 4. Fig. 6 shows a comparison of the deduced mouse ART5
Fig. 5 shows hydropathy profiles of ART5 and ART1. amino acid sequence with that of its proposed orthologue
Like ART1, ART5 also has a hydrophobic N-terminal signal from the human hART5 (our own unpublished observation),
peptide, but lacks the C-terminal hydrophobic GPI-anchor the mouse and human ART1 and the chicken ART6A and
signal. In addition to four cysteine residues which are ART7 paralogues. Note that mouse ART5 shows significant
conserved among the ART family, both proteins contain sequence identity to human ART5 even in the N- and C-
two extra cysteine residues near the C-terminus: C254 and terminal regions. ART5 is slightly more similar to ART1
C270 in ART5; C280 and C284 in ART1 (the latter pair is (37% identity) than to ART6A or ART6B (31 and 33%
also found in chicken ART7, see Fig. 6). Both proteins identity, respectively).
272 G. Glowacki et al. / Gene 275 (2001) 267 277
Fig. 4. Nucleotide and deduced amino acid sequences of Art5. Exon sequences are shown in upper case, and flanking intronic sequences are shown in lower
case. Four conserved cysteine residues are shown in circles, two non-conserved cysteine residues are marked by triangles, and the predicted catalytic glutamic
acid (E) residue and two potential N-linked glycosylation sites are boxed. The signal peptide cleavage site is marked by a double arrow, and the polyadenyla-
tion signal is boxed. Positions and orientation of PCR primers are marked by horizontal arrows above the nucleotide sequence. Cryptic splice sites in exons 2 4
are marked by slanted arrows (donor sites slanted toward the right, acceptor sites toward the left). Internal SacI and XhoI cleavage sites are marked by
downward arrows. Differences between the sequence and that published by Okazaki et al. (1996) (U60881) are marked as follows: a 1 bp deletion, *; a 1 bp
insertion, #; and a dinucleotide mutation, ^.
G. Glowacki et al. / Gene 275 (2001) 267 277 273
3.5. Art5 transcripts are subject to extensive alternatively
splicing
The pattern of differently sized Art5-specific bands
observed in our Northern and RT-PCR analyses (compare,
for example lanes 6 and 8 in panel b of Fig. 7A,B) suggested
that Art5 gene transcripts may be alternatively spliced. In
order to test this hypothesis, we cloned and sequenced PCR
amplification products obtained from different tissues with
forward primers derived from exons 1 (N05), 2 (N06), or 3
(N00) in combination with a reverse primer from exon 6
(N99). The results reveal a number of alternative splice
variants containing shortened versions of exons 2 4 (Fig.
8). These splice variants evidently derive from usage of
cryptic splice sites in these exons (two in exon 2, one in
exon 3, and three in exon 4) (these sites are marked by
arrows in Fig. 4). All splice variants contain the coding
region for the N-terminal signal peptide and are predicted
to encode truncated secretory polypeptides. Cap finder and
50 RACE analyses indicated that Art5 transcripts in testis
originate from a transcription start site at the 50 end of an
extended version of exon 2 (536 bp). In contrast, Art5 tran-
scripts in skeletal muscle apparently originate from a
Fig. 5. Hydropathy profiles of mouse ART5 and ART1. Four conserved
distinct transcription start site upstream of exon 1 (Fig. 8).
cysteine residues are marked by open circles and are connected by lines
The latter contains a much shorter version of exon 2 (111
indicating the predicted disulfide bonds. Two non-conserved cysteine resi-
bp) (Fig. 8).
dues are shown by shaded circles. N-linked glycosylation sites are marked
It is tempting to speculate that differences in the compo-
by forks. Proposed active site residues arginine (R), glutamic acid (E), and
serine (S) are also marked. Predicted N-terminal leader peptides and C- sition and sequence of the 50 UTR may permit regulation of
terminal GPI-anchor signal peptides are shaded.
enzyme levels at the translational level. Indeed, extensive
alternative splicing of the 50 UTR has been observed also for
rat ART2 (Kuhlenb�umer et al., 1997) and mouse Art3 genes
(our own unpublished observations). Intriguingly, we also
3.4. Expression of the Art5 gene is most prominent in testis
find Art1 and Art5 splice variants in regions coding for
catalytic important residues in the C-terminal half of the
In order to determine the expression profile of the
protein (Fig. 8) (Braren et al., 1998). The biological func-
Art5 gene, we performed Northern blot analyses using
tion, if any, of truncated ART1/ART5 proteins lacking
the exon 4-specific probe N11xN41. The results show a
mADPRT enzyme activity is unclear. It is conceivable,
prominent 1.6 kb Art5-specific band in testis (Fig. 7A,b,
for example, that these truncated proteins function as domi-
lane 8) and weaker 1.35 kb bands in heart and skeletal
nant negative mutants. Further analyses should address this
muscle (Fig. 7A,b, lanes 1 and 6) but not in any of the
question.
other tissues analyzed. For comparison, we performed a
similar analysis with a probe derived from the correspond-
ing exon of the mouse Art1 gene (Fig. 7A,a). Similar to
3.6. Recombinant ART5 is secreted by transfected insect
Art5, Art1 is prominently expressed in heart and skeletal
cells and displays potent NAD-glycohydrolase activity
muscle, but in contrast to Art5, Art1 is not expressed in
testis.
In order to test the prediction that ART5 is a secretory
Using a more sensitive RT-PCR assay, we compared
enzyme we cloned the Art5 cDNA into a baculovirus
expression of Art5 and Art1 using gene-specific primers
expression construct so as to fuse His6x- and FLAG-epitope
derived from separate exons. The results confirm the North-
tags to the C-terminal amino acid. Transfection of this
ern blot analyses with respect to expression in testis, heart
construct into the Sf9 and Hi5 insect cell line led to the
and skeletal muscle (Fig. 7A, lanes 8, 1, and 6, respectively).
appearance of a 37 kDa recombinant protein in the super-
Moreover, lower transcript levels could also be detected for
natant of transfected cells, which could be precipitated with
both genes in lung, and for Art1 in liver and fetal tissues.
immobilized anti FLAG-antibody (Fig. 9A, lane 3).
Screening the EST database (dbEST, release June 2000)
Fig. 9 shows comparative enzyme assays of mouse
for related ESTs with the BLASTn program revealed the
ART1, ART2.2, and ART5 recombinant proteins immobi-
presence of six Art5-specific ESTs, all derived from
lized on matrix-bound M2-anti-FLAG monoclonal antibo-
embryonic tissues.
dies. These immunoprecipitates were incubated with
274 G. Glowacki et al. / Gene 275 (2001) 267 277
Fig. 6. Amino acid sequence alignment of mouse, human, and chicken ARTs. The deduced amino acid sequence of mouse ART5 (mART5) was aligned with
those of its proposed human orthologue (hART5) and of its paralogues from mouse and human (mART1 and hART1, respectively) and chicken (chART6A and
chART7, respectively). The hydrophobic N- and C-terminal signal sequences are shaded (the latter is present only in mouse and human ART1). Conserved and
proposed active site residues arginine (R), glutamic acid (E), and serine (S) are marked by shaded circles. Calculated percentage amino acid sequence identities
between these proteins are shown in the table at the bottom. Sequences were compiled from database Accession numbers X95825 and S74683 (ART1), Y08028
and Y16835 (ART5), and D31864 and X82397 (ART6a and ART7).
[32P]NAD1 in the absence (Fig. 9A C) or presence of the The immunoblot in Fig. 9A shows that recombinant
arginine analogue agmatine (Fig. 9D). Protein-ADP-ribosy- ART1 and ART5 differ only slightly in Mr (lanes 2 and
lation was analyzed by SDS-PAGE autoradiography (Fig. 3), while ART2.2 is approximately 5 kDa smaller (lane 1).
9B); NAD-glycohydrolysis and agmatin-ADP-ribosylation The autoradiogram of the same blot (Fig. 9B) reveals little if
were analyzed by thin layer chromatography (Fig. 9C,D). any labeling of recombinant protein, but strong labeling of
G. Glowacki et al. / Gene 275 (2001) 267 277 275
Fig. 7. Northern blot (A) and RT-PCR (B) analyses of mouse Art5 and mouse Art1 gene expression. (A) A mouse multiple tissue Northern blot containing 2 mg
polyA1 RNA per lane was hybridized first with radiolabeled Art5 exonic probe N11xN41 (see Fig. 4) (b), stripped and reprobed with an Art1-specific probe (a),
and finally with a probe specific for Gapd (c). RNAs were from the following tissues: lane 1, heart; lane 2, brain; lane 3, spleen; lane 4, lung; lane 5, liver; lane
6, skeletal muscle; lane 7, kidney; lane 8, testis. (B) A panel of mouse cDNAs was subjected to PCR amplification with Art5-specific primers N00 and N99 (a),
with Art1-specific primers M16 and M90 (b) and with Gapd-specific primers GAPDF and GAPDR (Clontech) (c). Reaction products were analyzed by agarose
gel electrophoresis and were visualized by staining with ethidium bromide. The visible bands correspond to the expected sizes of 1100 bp in the case of Art5,
1130 bp in the case of Art1, and 1000 bp in the case of Gapd. RNAs were from the following tissues: lane 1, heart; lane 2, brain; lane 3, spleen; lane 4, lung; lane
5, liver; lane 6, skeletal muscle; lane 7, kidney; lane 8, testis; lane 9, 7 day embryo; lane 10, 11 day embryo; lane 11, 15 day embryo; lane 12, 17 day embryo.
cDNAs in lanes 1 8 are from a BALB/c mouse, and cDNAs in lanes 9 12 are from a Swiss Webster mouse.
the M2 light chain in case of ART2.2 (lane 1) and ART1 ribosyl-agmatine (lanes 1 and 3), while ART5 does not
(lane 3), but not ART5 (lane 2) or the control precipitate (lane 2). Instead, ART5 efficiently catalyzes the hydrolysis
(lane 4). of [32P]NAD1 to [32P]ADP-ribose (and nicotinamide) (Fig.
The thin layer chromatogram (Fig. 9D) shows that 9C,D, lane 2). Control precipitates do not catalyze
ART2.2 and ART1 catalyze the formation of [32P]ADP- [32P]NAD1 (lane 4). Similar results have been obtained
Fig. 8. Schematic diagram of Art5 cDNAs obtained by RT-PCR, 50 RACE and Cap-finder analyses. The six exons of the major splice variant are boxed and the
coding region is shaded as in Fig. 3. Arrows mark the cryptic splice sites in exons 2 4. Tissues from which cDNAs were derived, primers used for PCR, and
lengths of PCR products are shown on the right. For comparison, the structures of the full-length cDNA published by Okazaki et al. (1996) (U60881) and of
EST W12489 (this study) are shown on top.
276 G. Glowacki et al. / Gene 275 (2001) 267 277
Fig. 9. Enzyme assays with immunopurified recombinant mouse ART5. FLAG-tagged mART2.2 (lane 1), mART5 (lane 2), and mART1 (lane 3) were
immunoprecipitated from the supernatants of respective Sf9 cell transfectants with M2-Sepharose matrix. Control precipitations were performed with super-
natant of mock-transfected Sf9 cells (lane 4). Precipitates were washed and incubated with [32P]NAD1 for 60 min at 378C in the absence (A C) or in the
presence (D) of 2 mM agmatine. Proteins were precipitated and analyzed by SDS-PAGE (A,B), and supernatants were analyzed by TLC (C,D). (A)
Immunoblot developed with M2-monoclonal antibody (1:2500) followed by peroxidase conjugated sheep anti-mouse IgG and the enhanced chemilumines-
cence system. (B) Autoradiograph of the same blot after quenching of chemiluminescence. (C,D) Autoradiographs of thin layer chromatographs. Arrows mark
the positions of (a) FLAG-tagged ART proteins, (b) M2-light chain, (c) [32P]ADP-ribosylated agmatin, (d) [32P]NAD, and (e) [32P]ADP-ribose. Migration of
marker proteins (Seeblue, Novex) is indicated on the left.
also with recombinant mouse ART5 produced in Escheri- the mouse Art5 gene provides the basis for applying trans-
chia coli (Okazaki et al., 1996; Weng et al., 1999). These gene and knock-out technologies to further probe the func-
authors further observed that incubation of recombinant tion of this intriguing enzyme. The mouse will also provide
ART5 at high NAD concentrations (1 mM) resulted in a convenient model for testing whether recombinant ART5
auto-ADP-ribosylation of ART5 and a change of the or its inhibitors hold promise as tools for experimental inter-
NAD-glycohydrolase activity into agmatine-ADP-ribosyl- ventions.
transferase activity (Weng et al., 1999). With respect to its
strong basal NAD-glycohydrolase and little if any agma-
Acknowledgements
tine-ADP-ribosyltransferase activity, ART5 most closely
resembles recombinant rat ART2A (Haag et al., 1995).
The results described here in part represent the graduate
thesis of G.G. We thank Marion Nissen, Maren K�hl,
3.7. Conclusions
Roman Girisch, and Vivienne Welge for excellent technical
assistance. This work was supported by grant No310/3 from
Our results show that the Art5 gene maps to a cluster of
the Deutsche Forschungsgemeinschaft to F.K.N. and by
four Art genes on mouse chromosome 7, where it partially
stipends from the Boehringer Ingelheim Fonds, the
overlaps in a head to head fashion with the Art1 gene. Simi-
Deutscher Akademischer Austausch Dienst, and the
larities in the structures of the Art5 and other Art-encoding
Deutsche Gesellschaft f�r Immunologie to G.G. F.K.N.,
genes indicate that these genes were derived from a
F.H., and E.H.L. supervised the study. M.K. performed
common ancestor by gene duplication. Distinctive features
the experiments described in Figs. 1 and 7A, M.C.C. those
of the Art5 gene include extensive alternative splicing and
described in Fig. 2, G.G. and R.B. those described in Fig. 3,
the lack of a sequence encoding a C-terminal GPI-anchor
and G.G. those described in Figs. 4 9. G.G. and F.K.N.
signal peptide. In the latter respect, Art5 resembles chicken
wrote the paper.
ART6 and ART7 (Tsuchiya et al., 1994; Davis and Shall,
1995). However, the differences of these genes in their
exon/intron structures (Fig. 3B), the relatively low degree References
of amino acid sequence identity (Fig. 6), and the markedly
Aktories, K. (Ed.), 1991. ADP-Ribosylating Toxins (Current Topics in
different expression patterns  Art5 is expressed mainly in
Microbiology and Immunology). Springer Verlag, Berlin.
testis (Fig. 7), ART6 and ART7 in hematopoietic cells
Braren, R., Glowacki, G., Nissen, M., Haag, F., Koch-Nolte, F., 1998.
(Tsuchiya et al., 1994; Davis and Shall, 1995)  all indicate
Molecular characterization and expression of the gene for mouse
that Art5 most likely is a paralogue rather than the species
NAD 1 :arginine ecto-mono(ADP-ribosyl)transferase, Art1. Biochem.
orthologue of chicken ART6 and ART7. The availability of J. 336, 561 568.
G. Glowacki et al. / Gene 275 (2001) 267 277 277
Davis, T., Shall, S., 1995. Sequence of a chicken erythroblast mono(ADP- E., Thiele, H.G., Koch-Nolte, F., Haag, F., 1997. Expression of the RT6
ribosyl)transferase-encoding gene and its upstream region. Gene 164, mono(ADP-ribosyl)transferases is regulated by two promoter regions.
371 372. Adv. Exp. Med. Biol. 419, 265 270.
Domenighini, M., Rappuoli, R., 1996. Three conserved consensus Ludden, P.W., 1994. Reversible ADP-ribosylation as a mechanism of
sequences identify the NAD-binding site of ADP-ribosylating enzymes, enzyme regulation in procaryotes. Mol. Cell. Biochem. 138, 123 129.
expressed by eukaryotes, bacteria and T-even bacteriophages. Mol. Moss, J., Vaughan, M., 1990. ADP-Ribosylating Toxins and G Proteins:
Microbiol. 21, 667 674. Insights into Signal Transduction. American Society for Microbiology,
Greiner, D.L., Mordes, J.P., Handler, E.S., Angelillo, M., Nakamura, N., Washington, DC.
Rossini, A.A., 1987. Depletion of RT6.1 1 T lymphocytes induces Okazaki, I.J., Kim, H.J., Moss, J., 1996. Cloning and characterization of a
diabetes in resistant biobreeding/Worcester (BB/W) rats. J. Exp. Med. novel membrane-associated lymphocyte NAD:arginine ADP-ribosyl-
166, 461 475. transferase. J. Biol. Chem. 271, 22052 22057.
Haag, F., Koch-Nolte, F., 1997. ADP-Ribosylation in Animal Tissues: Otto, H., Tezcan-Merdol, D., Girisch, R., Haag, F., Rhen, M., Koch-Nolte,
Structure, Function and Biology of Mono(ADP-Ribosyl)Transferases F., 2000. The spvB gene-product of the salmonella enterica virulence
and Related Enzymes. Plenum Press, New York. plasmid is a mono(ADP-ribosyl)transferase. Mol. Microbiol. 37, 1106
Haag, F., Andresen, V., Karsten, S., Koch-Nolte, F., Thiele, H., 1995. Both 1115.
allelic forms of the rat T cell differentiation marker RT6 display nico- Prochazka, M., Gaskins, H.R., Leiter, E.H., Koch-Nolte, F., Haag, F.,
tinamide adenine dinucleotide (NAD)-glycohydrolase activity, yet only Thiele, H.G., 1991. Chromosomal localization, DNA polymorphism,
RT6.2 is capable of automodification upon incubation with NAD. Eur. and expression of Rt-6, the mouse homologue of rat T-lymphocyte
J. Immunol. 25, 2355 2361. differentiation marker RT6. Immunogenetics 33, 152 156.
Haag, F., Kuhlenb�umer, G., Koch-Nolte, F., Wingender, E., Thiele, H.G., Rappuoli, R., Montecucco, C., 1997. Guidebook to Protein Toxins and
1996. Structure of the gene encoding the rat T cell ecto-ADP-ribosyl- Their Use in Cell Biology. Oxford University Press, Oxford.
transferase RT6. J. Immunol. 157, 2022 2030. Rowe, L., Nadeau, J., Turner, R., Frankel, W., Letts, V., Eppig, J., Ko, M.,
Han, S., Craig, J.A., Putnam, C.D., Carozzi, N.B., Tainer, J.A., 1999. Thurston, S., Birkenmeier, E., 1994. Maps from two interspecific back-
Evolution and mechanism from structures of an ADP-ribosylating cross DNA panels as a community mapping resource. Mammal.
toxin and NAD complex. Nat. Struct. Biol. 6, 932 936. Genome 5, 253 274.
Honjo, T., Nishizuka, Y., Hayaishi, O., Kato, I., 1968. Diphtheria toxin- Schuman, E.M., Meffert, M.K., Schulman, H., Madison, D.V., 1994. An
dependent adenosine diphosphate ribosylation of aminoacyl transferase ADP-ribosyltransferase as a potential target for nitric oxide action in
II and inhibition of protein synthesis. J. Biol. Chem. 243, 3553 3555. hippocampal long-term potentiation. Proc. Natl. Acad. Sci. USA 91,
Koch-Nolte, F., Haag, F., 1997. Mono(ADP-ribosyl)transferases and 11958 11962.
related enzymes in animal tissues. Emerging gene families. Adv. Exp. Tsuchiya, M., Hara, N., Yamada, K., Osago, H., Shimoyama, M., 1994.
Med. Biol. 419, 1 13. Cloning and expression of cDNA for arginine-specific ADP-ribosyl-
Koch-Nolte, F., Klein, J., Hollmann, C., K�hl, M., Haag, F., Gaskins, H.R., transferase from chicken bone marrow cells. J. Biol. Chem. 269,
Leiter, E.H., Thiele, H.G., 1995. Defects in the structure and expression 27451 27457.
of the genes for the T cell marker Rt6 in NZW and (NZB x NZW)F1 Wang, J., Nemoto, E., Dennert, G., 1996. Regulation of CTL by ecto-
mice. Int. Immunol. 7, 883 890. nictinamide adenine dinucleotide (NAD) involves ADP-ribosylation
Koch-Nolte, F., K�hl, M., Haag, F., Cetkovic-Cvrlje, M., Leiter, E.H., of a p56lck-associated protein. J. Immunol. 156, 2819 2827.
Thiele, H.G., 1996a. Assignment of the human and mouse genes for Weng, B., Thompson, W.C., Kim, H.J., Levine, R.L., Moss, J., 1999.
muscle ecto mono (ADPribosyl)transferase to a conserved linkage Modification of the ADP-ribosyltransferase and NAD glycohydrolase
group on human chromosome 11p15 and mouse chromosome 7. Geno- activities of a mammalian transferase (ADP-ribosyltransferase 5) by
mics 36, 215 216. auto-ADP-ribosylation. J. Biol. Chem. 274, 31797 31803.
Koch-Nolte, F., Petersen, D., Balasubramanian, S., Haag, F., Kahlke, D., Zolkiewska, A., Moss, J., 1993. Integrin alpha 7 as substrate for a glyco-
Willer, T., Kastelein, R., Bazan, F., Thiele, H.G., 1996b. Mouse T cell sylphosphatidylinositol-anchored ADP-ribosyltransferase on the
membrane proteins Rt6-1 and Rt6-2 are arginine/protein mono(ADPri- surface of skeletal muscle cells. J. Biol. Chem. 268, 25273 25276.
bosyl)transferases and share secondary structure motifs with ADP-ribo- Zolkiewska, A., Nightingale, M.S., Moss, J., 1992. Molecular characteriza-
sylating bacterial toxins. J. Biol. Chem. 271, 7686 7693. tion of NAD:arginine ADP-ribosyltransferase from rabbit skeletal
Kuhlenb�umer, G., Rothenburg, S., Matthes, M., Hollmann, C., Wingender, muscle. Proc. Natl. Acad. Sci. USA 89, 11352 11356.