Báo cáo Y học: Cloning and characterization of the mammalian-specific nicolin 1 gene (NICN1) encoding a nuclear 24 kDa protein doc

Naim2and Tosso Leeb1 1 Institute of Animal Breeding and Genetics and2Institute of Physiological Chemistry, School of Veterinary Medicine Hannover; 3 Institute of Molecular Biology, Medic

Cloning and characterization of the mammalian-specific nicolin 1

Bianca Backofen1,*, Ralf Jacob2,*, Katrin Serth3, Achim Gossler3, Hassan Y Naim2and Tosso Leeb1 1

Institute of Animal Breeding and Genetics and2Institute of Physiological Chemistry, School of Veterinary Medicine Hannover;

3

Institute of Molecular Biology, Medical School Hannover, Hannover, Germany

We have identified a novel mammalian gene, termed nicolin 1

gene (NICN1), that is present in human, dog and mouse,

whereas it is absent from the available genome sequences of

nonmammalian organisms The NICN1 gene consists of six

exons and spans about 6 kb of genomic DNA It encodes a

213 amino acid protein that does not belong to any known

protein family Experiments using green fluorescent protein

(GFP)-tagged nicolin 1 fusion proteins indicate that nicolin 1

is a nuclear protein Northern analysis and semiquantitative

RT-PCR demonstrated that the 2.5 kb NICN1 mRNA is expressed in a tissue-specific manner The highest NICN1 expression levels are found in brain, testis, liver, and kidney

On the other hand the NICN1 expression is weak in spleen, leukocytes, small intestine and colon The NICN1 gene is also expressed during development

Keywords: nicolin, comparative genomics, human, mouse, dog

The initial completion of the human genome sequence

revealed that the function of only about 15 000 of the

estimated 35 000 genes in the human genome is known [1]

Systematic approaches have been undertaken to identify

expressed sequences and full length cDNAs in several model

organisms, which generated a wealth of sequence

informa-tion on previously unidentified genes [2] However, at

present a most immediate task is the functional

character-ization of all the newly identified genes In contrast to

genomic and cDNA sequencing, no universal

high-through-put parallel approach is currently available to functionally

characterize mammalian genes It is therefore still necessary

to investigate single genes experimentally in order to learn

more about their functions

We have previously identified a novel gene on dog

chromosome 20q15.1-q15.2 by aligning the dog genomic

sequence to expressed sequence tag (EST) sequences from

human and mouse [3,4] In order to further characterize this

gene, for which we propose the name NICN1, we have now

analyzed the human and murine orthologs and provide

initial data on the expression profiles and cellular

localiza-tion of the protein product

M A T E R I A L S A N D M E T H O D S

General methods Standard molecular biological techniques were performed

as described elsewhere [5] DNA sequencing using dye primer sequencing chemistry was performed with the thermosequenase kit (Amersham Biosciences, Freiburg, Germany) and a LICOR 4200 L automated sequencer

Isolation of cDNA and genomic clones of the NICN1 gene

The isolation of the canine genomic BAC contig containing the canine NICN1 gene has been described previously [3] BLAST searches against the human and murine EST databases with the canine NICN1 cDNA sequence derived from DDBJ/EMBL/GenBank accession AJ012166 identi-fied a human (IMAGE: 433564) and a murine (IMAGE: 1349406) cDNA clone that harbored the complete open reading frame of the NICN1 gene These clones were obtained through the Resource Center/Primary Database

of the German Human Genome Project (http:// www.rzpd.de) The inserts of these two clones were com-pletely sequenced and submitted to the EMBL nucleotide database under accession AJ299740 and AJ299741 For the isolation of a murine genomic clone with the Nicn1 gene the murine RPCI-21 PAC library made from a female 129S6/ SvEvTac mouse [6] was initially screened with PCR primers Nicn1V (5¢-TTCGGGCTAGTCACACTTC-3¢) and Nicn1R (5¢-CTAGATGGGCACAGAAAGC-3¢), which were derived from the murine Nicn1 cDNA sequence The PCR was performed at an annealing temperature of 58C and produced a product of 181 bp on mouse genomic DNA Three independent Nicn1 PAC clones were identi-fied DNA from the PAC clone RPCI-21 469L13 was isolated using the Qiagen plasmid maxi kit (Qiagen, Hilden, Germany) To determine the genomic sequence of the murine Nicn1 gene a primer walking strategy was used until

Correspondence toT Leeb, Institute of Animal Breeding and Genetics,

School of Veterinary Medicine Hannover, Bu¨nteweg 17p,

30559 Hannover, Germany.

Fax: + 49 511 9538582, Tel: + 49 511 9538874,

E-mail: Tosso.Leeb@tiho-hannover.de

Abbreviations: GFP, green fluorescent protein; NICN1, nicolin 1 gene;

NLS, nuclear localization signal.

*Note: These authors have contributed equally to the work of this

manuscript.

Note: The sequence data described in this paper have been submitted

to the EMBL nucleotide database under accession numbers

AJ012166, AJ299740, AJ299741, and AJ422131.

(Received 4 July 2002, revised 30 August 2002,

accepted 6 September 2002)

both strands in the region of interest were completely

sequenced Sequence data were analyzed withSEQUENCHER

4.0.5 (GeneCodes, Ann Arbor, MI, USA) and deposited

with the EMBL nucleotide database under accession

AJ437692 Further analyses were performed with the online

tools of the European Bioinformatics Institute (http://

www.ebi.ac.uk/), BLAST database searches in the

Gen-Bank database of the National Center for Biotechnology

Information NCBI (http://www.ncbi.nlm.nih.gov/) and the

RepeatMasker searching tool for repetitive elements (Smit,

A.F.A and Green, P., http://repeatmasker.genome.wash

ington.edu/)

Northern blot analysis

Northern blot analyses were performed using membranes

with 2 lg poly(A) RNA per lane from Clontech (BD

Biosciences Clontech, Heidelberg, Germany) according to

the manufacturer’s protocols With the human RNA filter

the 1194 bp EcoRI/NotI insert from the IMAGE 433564

clone was used as NICN1 probe With the mouse filter an

829-bp EcoRI fragment derived from the IMAGE 1349406

clone was used as Nicn1 probe For control experiments, the

human b-actin probes provided by Clontech were used The

probes were labeled with32P using the RadPrime labeling

kit (Invitrogen, Karlsruhe, Germany) After hybridization,

the membranes were exposed with intensifying screens

to Kodak BiomaxMS films (Amersham Biosciences,

Freiburg, Germany) at)80 C

RT-PCR analysis

Semiquantitative RT-PCR experiments were performed

using commercially available multiple tissue panels

(MTCTM) panels of first-strand cDNAs (BD Biosciences

Clontech) The concentration of the cDNAs in the used

panels is normalized to four different housekeeping genes

The PCR analyses of the panels was carried out according

to the manufacturer’s instruction For the human and

murine NICN1 RT-PCR, a fragment of 194 bp was

amplified using the primers NICN1_V362 (5¢-CAT

CACCACTGTGGCTGTC-3¢) and NICN1_R555

(5¢-CTCTGTCAGTGCCCACATC-3¢) and an annealing

temperature of 60C This experiment was performed two

times independently In control experiments

glyceraldehyde-3-phosphate dehydrogenase (GADPH) primers supplied

with the human and murine MTCTMpanels were used to

amplify an800 bp fragment from the housekeeping gene

GADPHusing an annealing temperature of 68C Aliquots

were taken from the PCR reactions during thermal cycling

and loaded on agarose gels for analysis

In situ hybridization on whole-mount mouse embryos

A digoxigenin-labeled antisense RNA probe for NICN 1

was generated using a 971-bp XhoI/SacI subfragment from

the murine IMAGE 1349406 clone and T7-RNA

poly-merase (Roche, Mannheim Germany) A sense RNA probe

using T3-RNA Polymerase (Roche, Mannheim, Germany)

was used as a negative control in the experiment Mouse

embryos at days 8.5, 9.5 and 10.5 were dissected in NaCl/Pi

and fixed overnight at 4C in 4% paraformaldehyde After

washing once with NaCl/P the embryos were dehydrated

through a methanol series and stored in 100% methanol at )20 C Whole-mount in situ hybridization was performed following a standard procedure [7] with minor modifications using an InsituPro #10.000 (INTAVIS Bioanalytical Instruments AG, Bergisch Gladbach, Germany) In brief, the hybridization was carried out at 65C After extensive washes, alkaline phosphatase-conjugated anti-digoxigenin antibody Fab fragment (Roche, Mannheim, Germany) was diluted 1 : 5000 and applied The coloration was performed

in BM Purple substrate (Roche, Mannheim, Germany) for several hours at 37C

Expression constructs and transient transfection

of COS-1 cells The entire protein coding sequence of the human NICN1 cDNA was PCR amplified with Pfu turbo DNA poly-merase (Stratagene, Heidelberg, Germany) and cloned into the BamHI/XhoI sites of the pEGFP-C1 or pEGFP-N1 vector (BD Biosciences Clontech) The resulting expression constructs were termed pEGFPNICN1 [green fluorescent protein (GFP) is fused to the N-terminus of nicolin 1] or pNICN1EGFP (GFP is fused to the C-terminus of nicolin 1) The protein coding regions of these two constructs were verified by DNA sequencing

COS-1 cells were transiently transfected with DNA by using DEAE-dextran essentially as described previously [8] The plasmids were diluted in 1.5 mL of serum-free DMEM (Gibco Life Technologies, Eggenstein, Germany), contain-ing 500 lg of DEAE-dextran and incubated at room temperature for 30 min The cells were rinsed twice with NaCl/Pi and overlaid with the DNA–complexsolution followed by incubation at 37C for 1.5 h in a CO2 incubator Subsequently, the solution was removed and

10 mL of DMEM containing 600 lg chloroquine was added for 3 h The cells were used 48–60 h after transfection for immunoprecipitation Cells for confocal analysis were treated in the same manner except that no chloroquine was added

Biosynthetic labeling of transfected cells and immunoprecipitation of cell extracts COS-1 cells were cultured in DMEM supplemented with 10% fetal calf serum, 50 UÆmL)1penicillin and 50 mgÆmL)1 streptomycin (denoted complete medium) They were transfected without DNA (mock) or with 2 lg of the appropriate recombinant DNA using DEAE-dextran as described above Transiently transfected COS-1 cells were labeled 48–60 h post-transfection for 2 h with 80 lCi of [35S]methionine in methionine-free DMEM containing 2% fetal calf serum, 50 UÆmL penicillin and 50 mgÆmL)1 streptomycin (denoted met-free medium) Cells were solu-bilized with 1 mLÆdish)1of cold lysis buffer essentially as described before [8] The lysates were incubated with the mouse anti-GFP antibody B34 (Babco) and precipitated with protein A-Sepharose Following immunoprecipitation, the protein A-Sepharose beads were washed three times with washing buffer A (0.5% Triton X-100, 0.05% sodium deoxycholate in NaCl/Pi) and three times with washing buffer B (500 mMNaCl, 10 mMEDTA, 0.5% Triton X-100

in 125 mMTris/HCl pH 8.0) prior to analysis of the samples

by SDS/PAGE and phosphorimaging

Confocal fluorescence microscopy

Confocal images of living cells were acquired 2 days after

transfection on a Leica TCS SP2 microscope with an· 63

water planapochromat lens (Leica Microsystems)

GFP-images of transfected COS-1 cells on coverslips were

obtained with the 488 nm excitation line of an argon laser

and an emission range between 505 and 525 nm for

fluorescence detection as described previously [9]

R E S U L T S A N D D I S C U S S I O N

Analysis of the NICN1 cDNA sequence

During the detailed analysis of a 162-kb dog genomic DNA

sequence [3,4], we observed the presence of a previously

undescribed gene that showed highly significant matches to

several ESTs in the databases The novel gene was termed

nicolin 1 (NICN1) A BLAST search against the

EST-database with the NICN1 cDNA resulted in 85 exclusively

mammalian hits with E-values < 10)10 The database

searches clearly verified the existence of NICN1 transcripts

in different mammalian species such as human, mouse, rat,

cattle and pig, whereas homologous sequences are absent

from nonmammalian species including the completely

sequenced model organisms We obtained a human and a

murine NICN1 cDNA clone from the IMAGE collection

and determined the complete sequences of these cDNAs

The comparative analysis of the human, canine, and murine

NICN1 cDNA revealed a highly conserved open reading

frame in all three species In all three species, the encoded

NICN1 protein consists of 213 amino acids and has a

calculated molecular weight of 24 kDa (Fig 1) The amino

acid sequences of the NICN1 proteins are 94% identical

between human and dog, and 89% identical between

human and mouse The NICN1 protein does not belong to

any known protein family It does not possess a signal

sequence at the N-terminal end for translocation into the

endoplasmic reticulum and it does not contain a

hydro-phobic transmembrane region, which could function as a

membrane-anchoring domain

Genomic organization of the NICN1 gene The genomic DNA sequences of the canine and murine NICN1genes were determined from genomic BAC or PAC clones and submitted to the EMBL nucleotide database (accessions AJ012166, AJ422131) Partial genomic sequences of the human NICN1 gene were available from the human genome draft sequence (contig accession NT_022439) A comparison between the genomic sequences and the previously obtained cDNA sequences revealed a conserved genomic organization of sixclosely spaced exons for the canine [3,4] and murine NICN1 gene Southern blotting confirmed that the murine Nicn1 gene is a single copy gene (data not shown), whereas in the human genome

a processed pseudogene of the NICN1 gene, termed NICN2P, is located at HSA Xp11.22–11.3 (EMBL acces-sion AL591503) According to the current human and murine genome sequences the human NICN1 gene is located at HSA 3p21 while the murine ortholog Nicn1 is located at MMU 9F

The location of the exon/intron boundaries in the NICN1gene with respect to the protein coding sequence is completely conserved between human, dog and mouse (Table 1) In dog, all the introns have the canonical GT-AG dinucleotides at the splice junctions, whereas in mouse, the third intron belongs to the rare class of GC-AG introns that are also recognized by U1/U2 containing spliceosomes [10] In human, dog and mouse the NICN1 gene is immediately followed by the AMT gene for aminomethyltransferase, an enzyme of glycine meta-bolism [4,11] The promoters of the canine and murine NICN1genes lack a TATA box-motif In both species, the sequences in the promoter regions fulfill the criteria of CpG islands (CpG obs/exp > 0.6 and GC > 50%) although the GC content is just barely above 50% In the mouse the transcription start site could be tentatively assigned as a full length murine cDNA sequence is available through the RIKEN mouse cDNA project [2; EMBL accession AK013602]

Expression of theNICN1 gene

To investigate the expression profile of the NICN1 gene in different adult tissues, we performed Northern blot analyses with human and murine RNAs (Fig 2) In human and mouse, specific signals were obtained at a size of approxi-mately 2.5 and 2.3 kb, respectively The size of the murine transcript is in good agreement with our murine genomic data, which allow the calculation of a cDNA size of 2133 nt without the poly(A) tail Expression levels of the NICN1 gene were measured from the signal intensities on the autoradiograms To validate the expression analysis we also performed semiquantitative RT-PCR analysis on normal-ized RNAs from different tissues (Fig 3) The expression data from these two experiments are summarized in Fig 4 Although there is some variation between the expression level seen in the Northern analyses and in the RT-PCRs, the two experiments clearly indicate a tissue-specific expression pattern of the NICN1 gene Tissues with strong NICN1 expression are brain, testis, liver and kidney Intermediate expression was observed in heart, skeletal muscle, placenta, pancreas and lung, whereas in peripheral blood leukocytes, spleen, thymus, small intestine and colon the observed

Fig 1 Alignment of the nicolin 1 amino acid sequences of human,

mouse, and dog Dashes represent identical amino acids The positions

of the exon boundaries with respect to the protein sequence are

indi-cated by arrows.

expression was weak Independent evidence from separate

experiments provides confirmation for these rough

classifi-cations in many of the investigated tissues However, in liver

strong NICN1 expression was observed in both mouse

experiments and in the human RT-PCR analysis but not in

the human Northern analysis One possible explanation for

this apparent discrepancy could be that the NICN1

expression in liver is not constantly high but dependent on

the metabolic state of the organ As the RNA for the

Northern blot and the RT-PCR analysis had been derived

from two different anonymous human donors the

differ-ences in the observed NICN1 expression levels in these two

samples might reflect differences in the physiological liver

conditions

In addition to the analysis of adult tissues, the RT-PCR analysis provided evidence that the murine Nicn1 gene is also expressed during development (Fig 3) A more detailed analysis of the expression of the Nicn1 gene at days 8.5, 9.5, and 10.5 of murine development was performed by in situ hybridization of digoxigenin-labeled riboprobes to whole-mount mouse embryos These experiments showed a uniform staining of the mouse embryos (data not shown) indicating the presence of Nicn1 transcripts in all tissues of the developing embryos at the investigated developmental stages

Fig 2 Northern blot analysis of the human and murine NICN1 genes.

Membranes containing 2 lg poly(A) RNA per lane from different

tissues were hybridized with 32 P-labeled probes After the

hybridiza-tions with human or murine NICN1 probes, control experiments with

b-actin probes were performed to verify the amounts of target RNA

that had been present in different lanes Exposure times were 24 h for

the NICN1 hybridizations, 2 h for the human b-actin control

hybrid-ization, and 30 min for the b-actin control hybridization on the mouse

blot.

Fig 3 Semiquantitative RT-PCR analysis of human and murine NICN1 genes Normalized cDNAs from different tissues were used as templates

in semiquantitative RT-PCR experiments In the upper five panels, the amplification of a NICN1 fragment after 24, 26, 30, 34 and 38 cycles of PCR is shown The NICN1 amplification was performed with identical primers and conditions on the human and murine cDNA panel In the lower panel the amplification of a GADPH control fragment after 26 cycles of PCR is shown to verify that similar amounts of cDNA had been present Note that the used cDNA panels are normalized to four different housekeeping genes, therefore slight differences in the amount

of GADPH are expected M, size standard (1 kb ladder); neg, negative control without cDNA template.

Table 1 Exon-intron junctions of the murine Nicn1 gene Exon sequences are shown in uppercase letters and intron sequences in lowercase letters Untranslated parts of the exons are shown in italics The conserved dinucleotides at exon-intron junctions are shown in bold Note that intron 3 belongs to the rare class of GC-AG introns while all the other introns have the typical GT-AG splice junctions For the first exon the transcription start site is shown instead of a 3¢ splice site For the last exon the polyadenylation signal (underlined) and the most common polyadenylation site (bold underlined) are shown instead of a 5¢ splice site Numbers refer to the corresponding positions in the Nicn1 cDNA with +1 at the adenosine of the translation initiation codon.

Intron phase

Intron size

Cellular localization of nicolin 1

Expression constructs were prepared, in which the GFP was

fused to the N- or C-terminal ends of human nicolin 1

COS-1 cells were transiently transfected with these two

constructs termed pEGFPNICN1 and pNICN1EGFP To

assess the expression of the full length fusion proteins,

transfected COS-1 cells were metabolically labeled with

[35S]methionine and the cellular extracts were

immunopre-cipitated using an anti-GFP antibody As shown in Fig 5, a NICN1-GFP fusion protein with an apparent molecular weight of about 53 kDa was detected, which corresponds well to the expected size of 53 143 Da

We next investigated the potential localization of the NICN1-GFP fusion protein in transfected COS-1 cells using confocal laser microscopy A strong nuclear fluores-cence became visible 48 h after transfection of cells with either pEGFPNICN1 or with pNICN1EGFP but not with the pEC1 control vector (Fig 5) A strong GFP-staining of the fusion proteins was detected in the nuclei, whereas staining of the cytosol and the nucleoli of trans-fected cells was less intensive These data provide strong evidence for a nuclear localization of nicolin 1 regardless of the mode of fusion with the GFP (i.e N- or C-terminal) Apparently the fusion of GFP had no substantial influences

on the structural features or on putative signals that sort the

Fig 4 Quantification of NICN1 expression in different tissues The graphs were calculated from the experiments shown in Figs 2 and 3.

In each experiment the tissue with the highest expression level was set to 100 The height of the other bars indicates the relative amount of NICN1 mRNA in the other tissues (A) In the human Northern blot, NICN1 signal intensi-ties were normalized with respect to the b-actin signals (B) In the human RT-PCR analysis NICN1 signal intensities after 34 cycles were normalized with respect to the GADPH signals after 26 cycles (C) In the mouse Northern blot, Nicn1 signal intensities were normalized with respect to the b-actin signals (D) In the mouse RT-PCR analysis, Nicn1 signal intensities after 30 cycles were normalized with respect to the Gapdh signals after 26 cycles.

Fig 5 Cellular localization of the nicolin 1 protein (A) Confocal images of representative COS-1 cells transfected with pNICN1EGFP

or pEGFPNICN1 48 h after transfection The cells show a strong nuclear fluorescence indicating a translocation of the fusion proteins into the nucleus Nuclear translocation is seen with both N- and C-terminal NICN1-GFP fusion proteins In contrast, cells that were transfected with the control plasmid pEGFP-C1 encoding GFP without any fused sequences showed only cytoplasmic fluorescence (B) To confirm the correct expression of full-length fusion proteins an immunoprecipitation of COS-1 cell extracts was performed Transi-ently transfected COS-1 cells were labeled for 2 h with [ 35 S]methionine Following immunoprecipitation with mAb anti-GFP the samples were analyzed by SDS/PAGE on a 10% slab gel After electrophoresis the gels were fixed and analyzed on a phosphorimaging device The observed size of the precipitated material corresponds to the expected size of the fusion protein Similar results were obtained with the con-struct pNICN1EGFP.

nicolin 1 protein to the nucleus Interestingly, translation of

the cDNA sequence did not reveal a canonical basic nuclear

localization signal (NLS) that has been found for many

other nuclear proteins [12] This indicates that NICN1 is

translocated into the nucleus by a pathway different from

that utilizing the classical NLS signal [13] For instance,

nuclear targeting may occur through O-linked

N-acetylglu-cosaminyl residues, which have been demonstrated to direct

neoglycoproteins to the nucleus [14]

In conclusion, we have identified a novel mammalian

protein designated nicolin 1 The nicolin 1 genes (NICN1)

were characterized in human, mouse and dog The observed

tissue-specific expression of the NICN1 gene and the nuclear

localization of the nicolin 1 protein provide initial evidence

for future studies to investigate the function of this novel

protein

A C K N O W L E D G M E N T S

We thank H Klippert and S Neander for expert technical assistance.

We would also like to thank F Martins-Wess for help with the

RT-PCR experiments.

R E F E R E N C E S

1 International Human Genome Sequencing Consortium (2001)

Initial sequencing and analysis of the human genome Nature 409,

860–921.

2 Kawai, J., Shinagawa, A., Shibata, K., Yoshino, M., Itoh, M.,

Ishii, Y., Arakawa, T., Hara, A., Fukunishi, Y., Konno, H.,

Adachi, J., Fukuda, S., Aizawa, K., Izawa, M., Nishi, K.,

Kiyo-sawa, H., Kondo, S., Yamanaka, I., Saito, T., Okazaki, Y.,

Gojobori, T., Bono, H., Kasukawa, T., Saito, R., Kadota, K.,

Matsuda, H., Ashburner, M., Batalov, S., Casavant, T.,

Fleisch-mann, W., Gaasterland, T., Gissi, C., King, B., Kochiwa, H.,

Kuehl, P., Lewis, S., Matsuo, Y., Nikaido, I., Pesole, G.,

Quac-kenbush, J., Schriml, L.M., Staubli, F., Suzuki, R., Tomita, M.,

Wagner, L., Washio, T., Sakai, K., Okido, T., Furuno, M., Aono,

H., Baldarelli, R., Barsh, G., Blake, J., Boffelli, D., Bojunga, N.,

Carninci, P., de Bonaldo, M.F., Brownstein, M.J., Bult, C.,

Flet-cher, C., Fujita, M., Gariboldi, M., Gustincich, S., Hill, D.,

Hofmann, M., Hume, D.A., Kamiya, M., Lee, N.H., Lyons, P.,

Marchionni, L., Mashima, J., Mazzarelli, J., Mombaerts, P.,

Nordone, P., Ring, B., Ringwald, M., Rodriguez, I., Sakamoto,

N., Sasaki, H., Sato, K., Schonbach, C., Seya, T., Shibata, Y.,

Storch, K.F., Suzuki, H., Toyo-oka, K., Wang, K.H., Weitz, C.,

Whittaker, C., Wilming, L., Wynshaw-Boris, A., Yoshida, K.,

Hasegawa, Y., Kawaji, H., Kohtsuki, S & Hayashizaki, Y (2001) Functional annotation of a full-length mouse cDNA collection Nature 409, 685–690.

3 Leeb, T., Neumann, S., Deppe, A., Breen, M & Brenig, B (2000) Genomic organization of the dog dystroglycan gene DAG1 locus

on chromosome 20q15.1-q15.2 Genome Res 10, 295–301.

4 Leeb, T., Breen, M & Brenig, B (2000) Genomic structures and sequences of two closely linked genes (AMT, TCTA) on dog chromosome 20q15.1 fi q15.2 Cytogenet Cell Genet 89, 98–100.

5 Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seid-man, J.G., Smith, J.A & Struhl, K (1995) Current Protocols in Molecular Biology John Wiley & Sons Inc, New York.

6 Osoegawa, K., Tateno, M., Woon, P.Y., Frengen, E., Mammoser, A.G., Catanese, J.J., Hayashizaki, Y & de Jong, P.J (2000) Bacterial artificial chromosome libraries for mouse sequencing and functional analysis Genome Res 10, 116–128.

7 Wilkinson, D.G (1992) In Situ Hybridization: a Practical Approach (Wilkinson, D.G., ed.), pp 75–83 IRL, Oxford, UK.

8 Jacob, R., Alfalah, M., Gru¨nberg, J., Obendorf, M & Naim, H.Y (2000) Structural determinants required for apical sorting of an intestinal brush-border membrane protein J Biol Chem 275, 6566–6572.

9 Jacob, R., Weiner, J.R., Stadge, S & Naim, H.Y (2000) Addi-tional N-glycosylation and its impact on the folding of intestinal lactase-phlorizin hydrolase J Biol Chem 275, 10630–10637.

10 Burset, M., Seledtsov, I.A & Solovyev, V.V (2001) SpliceDB: database of canonical and non-canonical mammalian splice sites Nucleic Acids Res 29, 255–259.

11 Backofen, B & Leeb, T (2002) Genomic organization of the murine aminomethyltransferase gene (Amt) DNA Sequence 13, 179–183.

12 Kalderon, D., Roberts, B.L., Richardson, W.D & Smith, A.E (1984) A short amino acid sequence able to specify nuclear loca-tion Cell 39, 499–509.

13 Jans, D.A., Xiao, C.Y & Lam, M.H (2000) Nuclear targeting signal recognition: a key control point in nuclear transport? Bioessays 22, 532–544.

14 Duverger, E., Pellerin-Mendes, C., Mayer, R., Roche, A.C & Monsigny, M (1995) Nuclear import of glycoconjugates is distinct from the classical NLS pathway J Cell Sci 108, 1325–1332.

S U P P L E M E N T A R Y M A T E R I A L

The following material is available from http://www blackwell-science.com/products/journals/suppmat/EJB/ EJB3232/EJB3232sm.htm

Fig S1 Whole-mount in situ hybridization on mouse embryos with Nicn1 riboprobes