Identification of a coordinate regulator of interleukins 4, 13, and 5 by cross-species sequence comparisons

Science

Volumn 288, Issue 5463, 2000, Pages 136-140

  Loots, G G ^a,b   Locksley, R M ^c   Blankespoor, C M ^a   Wang, Z E ^c   Miller, W ^d   Rubin, E M ^a   Frazer, K A ^a  

^a LAWRENCE BERKELEY NATIONAL LABORATORY   (United States)

^b UNIVERSITY OF CALIFORNIA   (United States)

^c UNIVERSITY OF CALIFORNIA   (United States)

^d PENNSYLVANIA STATE UNIVERSITY   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

INTERLEUKIN 13; INTERLEUKIN 4; INTERLEUKIN 5;

ANIMAL CELL; ARTICLE; BASE MISPAIRING; CYTOKINE RELEASE; HUMAN; HUMAN CELL; IMMUNOREGULATION; NONHUMAN; PRIORITY JOURNAL; REGULATOR GENE; REGULATORY MECHANISM; SEQUENCE HOMOLOGY; SPECIES DIFFERENCE;

ANIMALS; BASE SEQUENCE; CHROMOSOMES, HUMAN, PAIR 5; CONSERVED SEQUENCE; DNA-BINDING PROTEINS; FUNGAL PROTEINS; GENE EXPRESSION REGULATION; HUMANS; INTERLEUKIN-13; INTERLEUKIN-4; INTERLEUKIN-5; KINESIN; MICE; MICE, TRANSGENIC; PHYSICAL CHROMOSOME MAPPING; REGULATORY SEQUENCES, NUCLEIC ACID; SACCHAROMYCES CEREVISIAE PROTEINS; SPECIES SPECIFICITY; TH1 CELLS; TH2 CELLS; TRANSGENES;

ANIMALIA; MAMMALIA; MUS MUSCULUS; RODENTIA;

EID: 0034616398     PISSN: 00368075     EISSN: None     Source Type: Journal    
DOI: 10.1126/science.288.5463.136     Document Type: Article

References (34)

1. Q. Li, S. Harju, K. R. Peterson, Trends Genet. 15, 403 (1999).
2. R. C. Hardison, J. Oeltjen, W. Miller, Genome Res. 7, 959 (1997); W. Miller, PipMaker (additional program information is available at http://globin.cse.psu.edu/ pipmaker).
3. R. C. Hardison, J. Oeltjen, W. Miller, Genome Res. 7, 959 (1997); W. Miller, PipMaker (additional program information is available at http://globin.cse.psu.edu/ pipmaker).
4. A. Abbas, K. M. Murphy, A. Sher, Nature 383, 787 (1996).
5. D. J. Symula et al., Nature Genet. 23, 241 (1999).
6. Human 5q31 sequence (clones; H14, H13, H23, H11, H16, H21, H18, H15, H24, H17, H26, H25, H81, H22, H20, and H82) was obtained (for a list of human clones submitted to GenBank, see http://www-hgc. lbl.gov/human-p1s.html) and assembled into nine contigs of the following sizes (given in the directionality of centromere to telomere): 380.0, 13.8, 155.3, 106.9, 155.5, 5.4, 8,9, 32.3, and 90.0 kb. On the basis of the physical map of the region, the eight gaps have estimated sizes of 12, 4, 5, 3, 16, 2, 5, and 5 kb, respectively. Seven mouse chromosome 11 bacterial artificial chromosomes (BACs) were isolated from
7. California Institute of Technology library CitbCJZ and sequenced in either draft or finished format: 24OC4 (202 kb), 3.5× draft; 111181 (151 kb), AC005742 finished; 219O10 (204 kb), 5.0× draft; 32119 (131 kb), 4.0× draft; 33O24 (150 kb), 3.5× draft; 54K15 (145 kb), 5.0X draft; and 327O23 (185 kb), AF248716 4.0X draft.
8. K. A. Frazer et al., Genome Res. 7, 495 (1997).
9. Human repetitive elements were masked with Repeat-Masker (A. F. A. Smit and P. Green; additional program information is available at http://www.genome. washington.edu/uwgc/analysistools/repeatmask.htm). Using gapped BLAST searches (default parameters), we compared masked 5q31 sequences with sequences in the National Center for Biotechnology Information (NCBI) databases, GenPept (release 113) and UniGene (30 September 1999), and analyzed the sequences for potential coding regions with GenScan [C. Burge and S. Kartin, J. mol. Biol. 268, 78 (1997)]. The following 14 genes in the interval were previously known and localized on the basis of exact GenPept database matches (NCBI numbers are given): Ubiquinone-binding protein (D50369), GDF-9 (NM_005260), KIF3A (AF041853), IL-4 (A00076), 1L-13 (U10307), RAD5O (U63139), 1L-5 (NM_000879), IRF1 (P10914), OCTN2 (NM_003060), OCTN1 (NM_003059), P4-hydmxylase alpha (II) (NM_004199), GM-CSF (P04141), 1L-3 (P08700), and long chain fatty acyl CoA synthetase 2/KIAA0837 (AF099740/AB020644). Four genes were identified on the basis of GenScan predictions, exact expressed sequence tag (EST) matches, and their similarities to known proteins, and they are referred to as a homolog of the gene to which they had the highest BLASTX score: pMU_72-homolog (NM_005935), APX-homolog (Q01613), Septin2-homolog (D86957), and Cyclin I-homolog (NM_006835). Of the five genes referred to by their UniGene number, one was identified solely on the basis of an exact UniGene match (Hs.70932), and four were identified by exact UniGene matches in conjunction with GenScan predictions and mouse and human conserved sequences (Hs.11637, Hs.77114, Hs.13308. and Hs.591082). Two database matches were determined to be pseudogenes: the Lim-domain match (X93510) was exact but only to the first 98 out of 1130 nucleotides in the mRNA, and the P4-hydroxylase alpha (II) pseudogene had neither an open reading frame nor an exact matching EST.
10. Human repetitive elements were masked with Repeat-Masker (A. F. A. Smit and P. Green; additional program information is available at http://www.genome. washington.edu/uwgc/analysistools/repeatmask.htm). Using gapped BLAST searches (default parameters), we compared masked 5q31 sequences with sequences in the National Center for Biotechnology Information (NCBI) databases, GenPept (release 113) and UniGene (30 September 1999), and analyzed the sequences for potential coding regions with GenScan [C. Burge and S. Kartin, J. mol. Biol. 268, 78 (1997)]. The following 14 genes in the interval were previously known and localized on the basis of exact GenPept database matches (NCBI numbers are given): Ubiquinone-binding protein (D50369), GDF-9 (NM_005260), KIF3A (AF041853), IL-4 (A00076), 1L-13 (U10307), RAD5O (U63139), 1L-5 (NM_000879), IRF1 (P10914), OCTN2 (NM_003060), OCTN1 (NM_003059), P4-hydmxylase alpha (II) (NM_004199), GM-CSF (P04141), 1L-3 (P08700), and long chain fatty acyl CoA synthetase 2/KIAA0837 (AF099740/AB020644). Four genes were identified on the basis of GenScan predictions, exact expressed sequence tag (EST) matches, and their similarities to known proteins, and they are referred to as a homolog of the gene to which they had the highest BLASTX score: pMU_72-homolog (NM_005935), APX-homolog (Q01613), Septin2-homolog (D86957), and Cyclin I-homolog (NM_006835). Of the five genes referred to by their UniGene number, one was identified solely on the basis of an exact UniGene match (Hs.70932), and four were identified by exact UniGene matches in conjunction with GenScan predictions and mouse and human conserved sequences (Hs.11637, Hs.77114, Hs.13308. and Hs.591082). Two database matches were determined to be pseudogenes: the Lim-domain match (X93510) was exact but only to the first 98 out of 1130 nucleotides in the mRNA, and the P4-hydroxylase alpha (II) pseudogene had neither an open reading frame nor an exact matching EST.
11. Alignments between the human and mouse genomic sequences were computed with a dynamic programming method, scoring each nucleotide match as 1, each mismatch as -1, and each gap of length k as -6 to 0.2k. The percent identity plot displays the human positions and the percent identity of each segment of the alignment between successive gaps that has a length of at least 40 bp and at least 60% identity.
12. Sequences present in mature mRNAs were computationally identified either by exact GenPept matches or exact UniGene matches combined with partial GenPept matches and/or GenScan predictions of probability scores ≥0.15. Database matches include the 5′ and 3′ untranslated regions of mRNAs; therefore, some of the 155 conserved coding sequences, as defined in this study, are not translated.
13. To prevent the inclusion of RNA and RNA pseudogenes in this set of conserved noncoding sequences, we masked all tRNA and most of the known small nuclear RNA genes in the human 5q31 sequence with RepeatMasker. However, because computational screens are biased against RNA genes, it is possible that a fraction of these 90 conserved noncoding elements may actually be transcribed.
14. C. S. Osborne, M. A. Vadas, P. N. Cockerill. J. Immunol. 155, 226 (1995).
15. Genomic DNA was purchased from Clontech Laboratories (Palo Alto, CA) (catalog numbers dog 6950-1, rabbit 6960-1, rat 6750-1, mouse 6650-1, human 6550-1, porcine 6651-1, bovine 6850-1, and chicken 6852-1) Drosophila melanogaster and Fugu rubripes genomic DNA were isolated with standard methods. Polymerase chain reaction (PCR) amplifications were performed as follows: 100 ng of genomic DNA from each species was mixed with 200 μM of each deoxyribonucleoside triphosphate, 1 μM of each oligonucleotide primer (5 μM for degenerate primers), 5 μl of 10× PCR buffer (PerkinElmer). and five units of AmpliTaq DNA Polymerase (PerkinElmer) in a 50-μl volume. The samples were amplified under standard PCR reaction conditions in an automated thermal cycler (PerkinElmer 9700) for a total of 35 cycles with the following primer pairs: CNS-1 forward, 5′-TGATTTCTCGGCAGCCAGGGAGGGCC-3′; CNS-1 reverse, 5′-GGTGCCTGCGTCACCTCTGACCACAC-3′; CNS-2 forward, 5′-CCTCTCAGCATTTATCTTGGGC-3′; CNS-2 reverse, 5′-AGAGCCATAANNGTGTTTGGG-3′; CNS-3 forward, 5′-CNAGTNGNTCAGGGCNNGATGCCCAGG-3′; CNS-3 reverse, 5′-AAGGGNGTCTGNTCNTNCTGGAGCCTGCC-3′; CNS-4 forward, 5′-GCATGAAGNATTGNTGGCCC-3′; CNS-4 reverse, 5′-CTCTCTGGCNCTGGAACACC3′; CNS-5 forward, 5′-ACNGTTTTTNGTGTGCAGCACT-3′: CNS-5 reverse, 5′-ATTCTTTNAAAACCCCATATC-3′; CNS-6 forward, 5′-TAGNANAGTGAGGATGTCTG-3′; CNS-6 reverse, 5′-AAACCCCAGCNCTGGGCAAACAG-3′; CNS-8 forward, 5′-AAGTAAACNCTGNAAAANNTG-3′; CNS-8 reverse, 5′-CNCNNAAGTATACTTTGGAATCC-3′; CNS-9 forward, 5′-TNACTCNCAGTGACTGATNTTTG-3′; CNS-9 reverse, 5′-ATCNCCTCCNNGTNTCTTTGCAAC-3′; CNS-10 forward, 5′-CANGATGACTCAGCCAGCACAAG-3′; CNS-10 reverse, 5′-CCTNNTCTAGGAAATGGGCT-TGC-3′; CNS-11 forward, 5′-GGCAAANTGTCA
16. CAATGTTC-3′; CNS-11 reverse, 5′-CTGTCANAGCCACACAGAAG3′; CNS-12 forward, 5′-TCCACATTTTCTTNCCTTTG-3′; CNS-12 reverse, 5′-GTNTCNCTGCCCTTTGATG-3′; CNS-13 forward, 5′-GGNTGAGATNCTGGAGGCTC-3′; CNS-13 reverse, 5′-GAGCAGGTCTGACNNGGGTG-3′; CNS-14 forward, 5′-TTGGCAATTCCCCTGAAA C-3′; CNS-14 reverse, 5′-AAGCTKAGYTCTGGCAGG-3′; CNS-15 forward, 5′-AAGNNTGTT GCTANGGTCACTGTG-3′; CNS-15 reverse, 5′-GCAGNTGTGGTTTTGAGANGTTCA T-3′; CNS-16 forward, 5′-CTCCCACATCCTTGGGAGGG-3′; and CNS-16 reverse, 5′-CCAGNAGCCAGGGACACACC-3′. Amplified PCR products were analyzed by gel electrophoresis on 3% Nusieve GTG agarose gels (BioWhittaker Molecular Applications, Rockland, ME), extracted with QIAquick Gel Extraction Kits (QIA-GEN, Valencia, CA), and sequenced with Big Dye chemistry (PE Applied Biosystems, Foster City, CA).
17. 4 (pH 7.2), 1 mM EDTA. and 1% SDS]. Probes were generated by PCR amplification of human genomic DNA using primer pairs of each CNS element and then gel purified (12). A CNS element was defined as single copy if none of the five lanes containing DNA (one of each restriction digest) had more than one or two bands (assumed to be a polymorphism),
18. Gridded high-density dog and baboon libraries (BACPAC Resources, Roswell Park Cancer Institute, Buffalo, NY) were screened by hybridization with CNS-1 probes generated by PCR amplification of dog and human genomic DNA (12), respectively. Content mapping of 9 dog BACs and 14 baboon BACs for the presence of KIF3A, CNS-2, IL-4, CNS-1, IL-13, and RAD5O by PCR defined the location of CNS-1 to the IL-4 through IL-13 region.
19. CN5-1 sequences (human, mouse, dog, rat, cow, and rabbit) were searched (the Transcription Factor Database is available at http://transfac.gbf-braunschweig. de/TRANSFAC/index.html) for consensus binding sites of four proteins [NF-AT (nuclear factor of activated T cells), c-maf, GATA-3, and STAT6 (signal transducer and activator of transcription 6)] known to regulate IL-4 transcription. No conserved consensus binding sites were found.
20. N. Takemoto et al., Int. Immunol. 10, 1981 (1998); S. Agarwal and A. Rao, Immunity 9, 765 (1998).
21. N. Takemoto et al., Int. Immunol. 10, 1981 (1998); S. Agarwal and A. Rao, Immunity 9, 765 (1998).
22. YAC A94G6 (450 kb) (6) was retrofitted with plys2neo vector (gift from K. Peterson) as described [ B. C. Lewis, N. P. Shah, B. S. Braun, C. T. Denny, Genet. Anal. Tech. Appl. 9, 86 (1992)]. The yeast shuttling vector, pRS406.CNS-1.LoxP, was constructed as follows: A 2.4-kb Sac I fragment containing human CNS-1 was cloned into pR5406 (Stratagene) to generate pRS406.CNS-1. Oligonucleotides, LoxP-PmI I (forward 5′-GTGTAACT TCGTATAGCATACATTATACGAAGTTATCAC-3′; reverse 5′-GTGATAACTTCGTATAATGTATGCTATACGAAGTTACAC) and LoxP-Sph I (forward 5′-CTAACTTCGTATAGCATACATTATACGAAGTTATGCATG-3′; reverse 5′-CATAACTTCGTATAATGTATGCTATACGAAGTTAGCATG-3′), containing LoxP sequences with sticky ends were synthetically synthesized, annealed in vitro, and subcloned into PmI I and Sph I sites of the pRS406.CNS-1 vector, creating pRS406.CNS-1.LoxP This vector was linearized at the PfIM I site, and the pop-in/pop-out method [K. Duff, A. McGuigan, C. Huxley, F. Schultz, J. Hardy, Gene Ther. 1,1 (1993)] was used to modify the retrofitted A94G6 YAC. YAC DNA was isolated at a final concentration of ∼1 ng/ml and microinjected into fertilized FVB mouse eggs using standard proce-dures as previously described (6).
23. YAC A94G6 (450 kb) (6) was retrofitted with plys2neo vector (gift from K. Peterson) as described [ B. C. Lewis, N. P. Shah, B. S. Braun, C. T. Denny, Genet. Anal. Tech. Appl. 9, 86 (1992)]. The yeast shuttling vector, pRS406.CNS-1.LoxP, was constructed as follows: A 2.4-kb Sac I fragment containing human CNS-1 was cloned into pR5406 (Stratagene) to generate pRS406.CNS-1. Oligonucleotides, LoxP-PmI I (forward 5′-GTGTAACT TCGTATAGCATACATTATACGAAGTTATCAC-3′; reverse 5′-GTGATAACTTCGTATAATGTATGCTATACGAAGTTACAC) and LoxP-Sph I (forward 5′-CTAACTTCGTATAGCATACATTATACGAAGTTATGCATG-3′; reverse 5′-CATAACTTCGTATAATGTATGCTATACGAAGTTAGCATG-3′), containing LoxP sequences with sticky ends were synthetically synthesized, annealed in vitro, and subcloned into PmI I and Sph I sites of the pRS406.CNS-1 vector, creating pRS406.CNS-1.LoxP This vector was linearized at the PfIM I site, and the pop-in/pop-out method [K. Duff, A. McGuigan, C. Huxley, F. Schultz, J. Hardy, Gene Ther. 1,1 (1993)] was used to modify the retrofitted A94G6 YAC. YAC DNA was isolated at a final concentration of ∼1 ng/ml and microinjected into fertilized FVB mouse eggs using standard proce-dures as previously described (6).
24. K. Wagner et a/., Nucleic Adds Res. 25, 4323 (1997).
25. G. G. Loots, unpublished observations.
26. For FISH, slides of lymphocytes isolated from 4-to 6-week-old Fl transgenics were prepared and hybridized with two human P1's (H23 and H24) (5) as described [E. D. Green et al., in Mapping Genomes, vol. 4 of Genome Analysis: A Laboratory Manual Series, B. Birren et al., Eds. (Cold Spring Harbor Press, Cold Spring Harbor, NY, 1999), pp. 303-413].
27. ni phenotype. Cells were activated using irradiated antigen-presenting cells with monoclonal antibodies against β T cell receptor and CD28 with IL-2; for Th1 conditions, recombinant murine IL-12 and antibody to IL-4 were used, and for Th2 conditions, recombinant murine IL-4 was used.
28. + T cells were analyzed with flow cytometry (positive cells were gated as compared to isotype antibody controls and are based on at least 10,000 flow cytometric events) for expression of the designated murine and human cytokines by intracellular cytokine detection after 4 hours of restimulation with phorbol myristate acetate and ionomycin (PMA/IONO) (4). Supernatants from activated T cells were collected (either 72 hours after the primary stimulation with antibodies and irradiated antigen presenting cells or 24 hours after restimulation of 7-day-old cultures with PMA/IONO) and analyzed with enzyme-linked immunosorbent assay (ELISA) for human IL-5 and murine IL-13.
29. Total RNA was isolated with RNA-STAT-60 (TELTEST"B"). Five micrograms of RNA was reverse-transcribed into cDNA (Superscript II, Gibco), and expression levels were measured with TaqMan Syber-Green quantitative PCR assay (PerkinElmer). The sequences of the TaqMan primer pairs used to quantify mRNA are as follows: human IL-4 forward, 5′-ACAGCCTCA
30. CAGAGCAGAAGACT-3′; human IL-4 reverse, 5′-GT-GTTCTTGGAGGCAGCAAAG-3′; human IL-5 forward, 5′-ATAAAAATCACCAACTGTGCACTGAA-3′; human IL-5 reverse, 5′-CAAGTTTTTGAATAGTCTTTCCACAGTAC-3′; human IL-13 forward, 5′'-CAGAAGCTCCGCTCTGCAAT-3′; human IL-13 reverse, 5′-ACACGTTGATCAGGGATTCCA-3′; human KIF3A forward, 5′-CGCAGTCTCGAGAGCGTCAA-3′; human KIF3A reverse, 5′-ACACCGGGTGCGCAGA-3′; human RAD50 forward, 5′-TGTTGGCTGGCAGGATCTTT-3′; human RAD50 reverse, 5′-CGTGAGACCCGCGAATCT.
31. 3′; mouse glyceraldehyde phosphate dehydrogenase (mGAPDH) forward, 5′-GGCAAATTCAACGCCACAGT-3′; and mGAPDH reverse, 5′-CCTCACCCCATTT-GATGTTAGTG-3′.
32. R. M. Locksley and Z. E. Wang, unpublished observations.
33. K. A. Frazer, unpublished observation.
34. We thank A. Nyugen, W. Dean, and K. Lewis for DNA sequencing; J. f. Cheng for isolating mouse chromosome 11 BACs; K. Frankel for assistance with sequence assembly; K. U. Wagner for providing the Cre recombinase transgenic mice; I. Plajzer-Frick for assistance with FISH; C. McArthur for flow cytometry; and K. Peterson and M. Dunaway for critical reading of the manuscript. Supported by the following grants: U.S. Department of Energy contract DEAC0376SF00098, NIH GM-5748202 (K.A.F.), Howard Hughes Medical Institute, AI30663 and NIH HL56385 (R.M.L), and National Library of Medicine LM05110 (W.M.).