chromosome mapping and organization of the human

THE JOURNAL OF BIOLOGICAL CHEMISTRY 8 1993 by The American Society for Biochemistry and Molecular Biology, Inc

Vol. 268, No. 6, Issue of February 25, pp. 4355-4361,1993 Printed in U.S.A.

Chromosome Mapping and Organization of the Human @-Galactoside ar2,6-Sialyltransferase Gene DIFFERENTIAL AND CELL-TYPE SPECIFIC USAGE OF UPSTREAM EXON SEQUENCES IN B-LYMPHOBLASTOID CELLS*

(Received for publication, April 10,1992)

XueCheng Wang, Annette Vertino, Roger L. Eddy$, Mary G. Byers$, Sheila N. Jani-Sait$, Thomas B. Shows$, and Joseph T. Y. LauO From the DeDartments of Molecular and Cellular Biolom and $Human Genetics, Roswell Park Cancer Institute, -”

Buffalo, Ne; York 14263

The human &galactoside a2,6-~ialyltransferase (EC 2.4.99.1) (SiaT-1) gene is localized to human chromosome 3 (q21-q28) by Southern analysis of somatic cell hybrids and by in situ hybridization of metaphase chromosomes. Comparative analysis between the human and the previously reported rat SiaT-1 genomic sequences demonstrates precise conservation of the in- tronlexon boundaries throughout the coding domains. Furthermore, there is extensive inter-species sequence similarity in some of the exons that contain information only for the 5’-leader regions. Human genomic sequences were also analyzed to reconcile reported differences in the 5”untranslated region in SiaT-1 mRNAs. In cultured cell lines of the B-lineage, Reh, Nalm-6, Jok-1, Ball-1, Daudi, and Louckes, the study demonstrates that three upstream exons, Exons(Y+Z) and Exon(X), are mutually exclusively utilized, result- ing in at least two distinct populations of SiaT- 1 mRNA being synthesized. None of these exons is present in the SiaT-1 mRNA isotype expressed in HepG2 human hepatoma cells. In all B-lymphoblastoid cell lines examined, the basal level SiaT- 1 mRNA is maintained by the expression of an isotype containing the Exons(Y+Z) sequence. The slightly smaller SiaT-1 mRNA, which contains the Exon(X) sequence but not Exons(Y+Z) sequence, is synthesized at a high level and found only in Jok-1, Daudi, and Louckes, the cell lines with mature B-cell phenotype. The study also provides further evidence that induced SiaT- 1 expression accompanies the appearance of CDw75, a putatively sialylated cell surface epitope and a marker of human mature B- lymphocytes. The SiaT-1 induction is the result of the appearance of a novel form of SiaT- 1 mRNA isotype.

There is much recent interest in role and regulation of the /%galactoside cu2,6-sialyltransferase (EC 2.4.99.1) (SiaT-1)’ in B-lymphocytes. The attention stems from CDw75, a human leukocyte cell-surface antigen expressed in mature and activated B-cells but not in B-cells at earlier stages of develop-

* This work was supported by Grants GM38193 (to J. T. Y. L.), HG00333 (to T. B. S.), and HD05196 (to T. B. S.) from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

8914; Fax: 716-845-8169. § T o whom correspondence should be addressed. Tel.: 716-845-

The abbreviations used are: SiaT-1, a2,6-sialyltransferase; kb, kilobase; bp, base pair; PCR, polymerase chain reaction; DAPI, diaminophenylindole.

ment nor in plasma cells. CDw75 is defined by monoclonal antibodies HH2 (Smeland et al., 1985), LN1 (Epstein et al., 1984), OKB-4 (Mittler et al., 1983), and EBU-141 (Dorken et al., 1989). CDw75, once erroneously believed to be a cell- surface variant of SiaT-1 (Stamenkovic et al., 1990), is a sialylated, cell-surface epitope requiring the catalytic media- tion of SiaT-l for its elaboration (Bast et d. , 1992; Munro et al., 1992; Stamenkovic et al., 1992; Guy and Andrew, 1991). A recent report contends that CDw75 is a ligand for CD22, a member of the immunoglobulin superfamily of integral mem- brane proteins (Wilson et al., 1991), and suggests a role for CDw75-CD22 in B cell-B cell interaction (Stamenkovic et al., 1991). A dramatic elevation in SiaT-1 mRNA levels parallels the appearance of CDw75 during the transition from the pre- B stage of development into mature, activated B-cells (Erik- stein et al., 1992; Stamenkovic et al., 1990, 1992).

SiaT-1 mediates the transfer of sialic acid onto exposed Galpl-4GlcNAc termini of N-linked oligosaccharides in serum and cell-surface glycoproteins. While little is known about SiaT-1 regulation in human lymphocytes, there has been substantial work on the SiaT-1 gene in rats, where sequences specifying the 403-amino acid catalytically active protein are divided among 5 exons that span over 45 kb of genomic DNA (Wang et al., 1990a; Svensson et al., 1990). A glucocorticoid responsive promoter region modulates SiaT-1 expression in liver (Wang et al., 1989). Furthermore, SiaT-1 mRNAs are heterogeneously sized in different rat tissues (O’Hanlon et al., 1989; Paulson et al., 1989). Transcript forms that retain only the 3”portion of SiaT-1 coding sequence have been characterized in rat kidney (O’Hanlon et al., 1989; Wen et al., 1992; O’Hanlon and Lau, 1992). At least one mRNA variant containing additional 5”untranslated sequences, but retaining the complete coding sequences, has been reported (Wen et al., 1992). From human tissues, since the initial isolation of a partial SiaT-1 cDNA from submaxillary gland (Lance et al., 1989), putatively full-length cDNAs have been isolated from placenta (Grundmann et al., 1990) and Daudi cells (Stamenkovic et al., 1990), a Burkitt lym- phoma B-lymphoblastoid cell line. Curiously, while the placenta and Daudi SiaT-1 cDNAs are in essential agreement over the protein coding sequence, the placenta cDNA predicts a completely different 5”untranslated sequence than the Daudi cDNA.

As a first step to understanding the expression and regulation of SiaT-1 in human B-lymphocytes, human SiaT-1 genomic sequences were isolated and comparatively analyzed against known rat genomic information. We report the close inter-species similarity between the rat and human SiaT-1 genes that includes not only sequence similarity but also

4355

4356 Human Sialyltransferase Gene

precise conservation of intron/exon boundaries. Multiple exons encoding 5“untranslated sequences are differentially utilized to generate the placenta and Daudi forms of mRNA. In lymphoblastoid cells exhibiting B-cell phenotypes at different stages of maturation, CDw75 expression is accompanied by the induction of SiaT-1 mRNA. Furthermore, while the placental SiaT-1 form is expressed at low levels in all cell lines examined, elevated SiaT-1 mRNA levels in CDw75 positive cells result from the expression of the Daudi form. Finally, we document that the SiaT-1 gene (SiaT-1) resides in 3q27- q28 of human chromosome 3.

EXPERIMENTAL PROCEDURES

Cell Culture-The human lymphoblastoid lines, Jok-1, Daudi, Reh, and MT-2 were purchased from Dr. James Pauley (Roswell Park Cancer Institute, Buffalo, NY). The lymphoblastoid lines, Ball-1, Louckes, and Nalm-6 were gifts from Drs. Sahai Srivastava, John Yates, and Ben Seon (Roswell Park Cancer Institute, Buffalo, NY), respectively. These cells were maintained in suspension in RPMI 1640 supplemented with 10% fetal bovine serum, 100 units/ml peni- cillin, and 100 pg/ml streptomycin. HepG2 (a gift from Dr. Heinz Baumann, Roswell Park) was maintained in minimal essential me- dium (GIBCO) supplemented with 10% fetal bovine serum.

Isolation and Characterization of Genomic Sequences-A library of Sau3A partial digest of human placental DNA cloned into EMBL-3 SP6/T7 (Clontech Lab, Palo Alto, CA) was screened using probes derived from human and rat SiaT-1 cDNA clones (see “Nucleic Acid Probes” below). Insert DNA fragments were initially characterized by restriction digestion and Southern blot analysis. Human genomic DNA fragments that hybridized to SiaT-1 cDNA probes were sub- cloned into plasmid vectors. Complete sequence determination was performed for all genomic exon regions to determine intron interrup- tions and that the genomic sequences are in complete agreement with known SiaT-1 cDNA information. Double-stranded sequencing was performed using Sequenase (modified T7 polymerase from U. S. Biochemicals) by the dideoxynucleotide chain termination method.

Analysis of SiaT-1 mRNA Expression-RNA isolation from cells and Northern analysis was performed as described previously (Wang et al., 1989). First strand cDNAs were synthesized from 5 pg of poly(A)+ mRNA (TimeSaver cDNA Synthesis Kit, Pharmacia LKB Biotechnology Inc.). Parallel cDNA synthesis reactions were performed using either the synthetic oligonucleotides HST-p3 (5’- CTCTGGTTTGGCCTTGG-3’) or HST-p4 (5”AACTTGATGCC- TGGTCC-3’) as primers for reverse transcription. Both these oligonucleotides are complementary to the coding region of human SiaT- 1 within Exon(I1). For each PCR reaction, 1% of the resultant cDNA was used. PCR primers used were HST-p5 (5’- CTGCTTCTGGCTAATC-3’), HSTp6 (5”AAAGGGAGCCGA- TACCGACC-3’), HST-pl5 (5”GGGGAGAGCCAGTTGCGC-3’), and HST-pl9 (5’-GCATCCCTTCTCCCATAC-3’). Each reaction, in a total of 40 pl, is comprised of 0.2 mM of each dinucleotide triphos- phate, 1.5 units of Taq polymerase, 3.0 pl of cDNA, and appropriate concentrations of primers as given in the legend of Fig. 6. The relative location of each of these primer sequences is diagrammed in Fig. 2 A .

Human SiaT-I cDNA Clones and Nucleic Acid Probes-The probes for SiaT-1 gene isolation and for Northern hybridization of SiaT-1 mRNA were generated from a number of different human SiaT-1 cDNA clones. pHST1, which spans two-thirds of the SiaT-1 coding domain (Lance et al., 1989) and pHST2, which covers the 3”untranslated region’ were from a human submaxillary gland cDNA library. pHL6O-1, a SiaT-1 clone containing (Y+Z+I+II) form, was from HL-60 mRNA (data not shown). pJOK-8 is a cDNA derived from Jok-1 mRNA and contains SiaT-1 sequence with the configuration of (X+I+II) (data not shown). Both pHL6O-1 and pJOK-8 were derived from cDNA synthesized using HST-p3 or HST-p4 as primers for reverse transcription. For the gene mapping studies, the 2.0-kb insert from pHST2 was used as probe. Exon(X)- and Exons(Y+Z)- specific probes were derived from PCR-amplified fragments from the appropriate cloned cDNA and primer pairs HL-p6/HL-p7 and HL- p20/HL-p19, respectively (see Fig. 5, legend, and Fig. 4, C-E). The probe for hybridization of human terminal transferase mRNA is a 1.1-kb EcoRI/EcoRI fragment of pT711 (a generous gift from Dr. Fred Bollum, Uniformed Services University of the Health Sciences, Bethesda, MD). The pT711 fragment covers the 3’-half of the ter-

* K. H. Lau and J. T. Y. Lau, unpublished data.

minal transferase coding region and 200 bp of the 3”untranslated sequence.

Immunofluorescence Microscopy-The cells were plated on coverslips, washed 3 times in phosphate-buffered saline (pH 7.4), fixed in acetone for 5 min at room temperature, and incubated in blocking solution (0.1% bovine serum albumin, 0.3% normal goat serum) for 1 h at room temperature. The slides were then washed 3 times in phosphate-buffered saline (10 min per wash), incubated for 1 h at room temperature in a humidified chamber with HH2 (anti-CDw75) (a gift from Dr. Steinar Funderud, Norwegian Radium Hospital, Norway) diluted in blocking solution. Slides were then washed 3 times in phosphate-buffered saline, incubated for 30 min at room temperature with rhodamine-conjugated goat anti-mouse IgM (Kirkegaard & Perry Lab, Inc., Gaithersburg, MD) diluted in blocking solution. Following the incubation, the slides were washed 3 times with phosphate-buffered saline, once with distilled water, and mounted on microscope slides using antifed solution containing 1.5 pg/ml diaminophenylindole (DAPI), 1.5 pg/ml p-phenylenediamine, and 2% 1,4- diazabicyclo-[2,2,2]octane in glycerol. Slides were analyzed on a Ni- kon Optiphot microscope equipped for epifluorescence microscopy. G-2A and UV-2A filter combinations were used for the detection of rhodamine and DAPI, respectively.

Somatic Cell Hybrid Analysis-A 2-kb cDNA fragment (pHST2) that corresponds to the 3’-untranslated region of human SiaT-1 mRNA (see “Nucleic Acid Probes”) was used to probe Southern blots for the somatic cell hybrid analysis and in situ hybridization to chromosomes. This 2-kb region, in its entirety, is located within Exon(V1) of the human SiaT-1 gene. The probe hybridizes to a single band on total human genomic blot, and does not recognize mouse genomic DNA (data not shown). Somatic cell hybrids and the method of analysis are described in detail previously (Shows et al., 1982,1984; Shows, 1983). Briefly, 36 hybrids derived from 17 unrelated human cell lines and 4 mouse cell lines, all previously characterized by karyotypic analysis and by mapped enzyme markers, were used. Southern blots of the DNA prepared from these hybrids were probed for the presence of the human SiaT-1 sequence.

In Situ Hybridization to Metaphase Chromosomes-Fluorescent in situ hybridization to metaphase chromosomes was performed as described by Cherif et al. (1989) and Fan et al. (1990). The DNA probe (see “Somatic Cell Hybrid Analysis”) was biotinylated, hybridized to the chromosome spreads, and detected by fluorescein-conjugated avidin (Vector Labs). Slides were evaluated with a Nikon fluorescence microscope. Twenty-two metaphases were examined, with Q (DAPI counterstaining) and R banding (propidium iodide counterstaining) used to confirm the identity of the chromosome.

RESULTS

Chromosomal Mapping of Human SiaT-1 Gene-Southern blot analysis of somatic cell hybrids using a probe that spe- cifically recognizes Exon(V1) of the human SiaT-1 gene was employed to determine chromosomal assignment (see “Exper- imental Procedures”). The result of this analysis is summa- rized in Table I, where the scoring was determined by the presence (+/) or absence (-/) of human SiaT-1 signal and compared to previously defined presence (/+) or absence (/ -) of the individual human chromosomes in each hybrid. For example, of the 18 hybrids that are known to contain human chromosome 3, all 18 hybridized positively (+/+ column) and none showed negative hybridization (-/+ column) for SiaT- 1. Fifteen other hybrids are known to be missing chromosome 3; none of these scored positively (+/- column), and all 15 were negative (-/- column) for the human SiaT-1 sequence. Thus, a 0% discordancy indicates a perfect segregation of the probe with a specific chromosome and that the SiaT-1 gene is localized on chromosome 3. Furthermore, hybrids with translocated but no intact chromosome present were not used for tabulation of discordancy. There were three additional hybrids that contain chromosome 3 translocations but no intact chromosome 3. These hybrids are TSL-2 (with 17ter- 1 7 ~ 1 3 + 3p21-3pter translocation), XTR-2 and XTR- 3BSAGB (both with 3pter-3q21 + Xq28-Xqter translocation). All three of these hybrids scored negative for SiaT-1 signal (data not shown), further localizing the SiaT-1 gene to the q21 to qter region of chromosome 3.

Human Sialyltransferase Gene 4357

~

Concordant No. Discordant No. Chromosome Of hybrids Of Discordancy

(+/+) (-/-) (+/-) (-/+)

1 7 2 9 3 18 4 11 5 13 6 10 7 12 8 9 9 3

10 14 11 8 12 13 13 11 14 11 15 12 16 11 17 13 18 11 19 5 20 10 21 12 22 6 x 10

15 9 10 9 15 0 14 6 10 5 13 8 12 5

7 9 14 13 6 4

10 7 9 5

10 7 9 7

11 5 15 7 6 5

10 7 14 13 6 8 4 6

11 12 10 5

%

1 31 7 46 0 0 4 29 8 36 5 36 6 31

11 56 2 47

11 43 7 44 9 39 8 42 9 44 7 34 3 28 9 42 8 42 4 47

12 56 14 56 7 53 6 35

Fluorescent in si tu hybridization to metaphase chromosomes was performed independently to the somatic cell hybrid analysis to map the sialyltransferase gene. Of the 22 metaphases examined, a fluorescent signal was detected at 3q27- 3q28 on both chromatids of a single chromosome 3 in 17 out of the 22 cells. One of the 17 cells had signals on both chromatids of both chromosome 3 at 3q27-3q28. This result is illustrated as an idiogram in Fig. 1.

Isolation and Characterization of Human SiaT-1 Genomic Sequences-Genomic clones from a human placental DNA library were isolated using probes generated from existing human and rat SiaT-1 cDNA sequences. Probes specific to Exon(I1) and Exon(V1) were used for a comparative Southern blot analysis between the cloned DNA and total human genomic DNA. Genomic and cloned DNA resulted in an identical set of signals that indicate the correct representation of the human genomic sialyltransferase sequence on the cloned DNA (data not shown). Coding sequences for the 405- amino acid human SiaT-1 protein (Lance et al., 1989; Grund- mann et al., 1990), as in rat SiaT-1 (Wang et al., 1990a), are divided among 5 exon regions (shaded regions in Exon(I1) to Exon(V1)) that span over 40 kb of human genomic DNA (Fig. 2A) .

Fig. 2B is a diagrammatic comparison of the intron/exon boundaries in the coding domains between human (H) and rat (R) SiaT-1 sequences that clearly demonstrates the precise conservation between the human and rat genes.

In both rat and human SiaT-1 genes, Exon(1) carries information for the 5”untranslated region. In addition to Exon(I), three additional human exons, Exons(X), (Y), and (Z), that encode 5”untranslated sequences were identified (see Fig. 1). The human placenta SiaT-1 cDNA isolated by Grundmann et al. (1990) was generated from mRNA containing Exon(Y) and Exon(Z) sequences 5’ to Exon(1). A cDNA clone, pHL6O- 1 , isolated in this laboratory from HL-60 cells, contains identical 5”untranslated leader sequence to the placenta cDNA (data not shown). The Daudi cDNA isolated by Sta- menkovic et al. (1990) resulted from an mRNA in which Exon(X) sequence was used instead of Exons(Y+Z) se-

23 22

14

1? I I .- I

FIG. 1. Idiogram of the distribution of signals on both chromatids of chromosome 3. In situ hybridization of human metaphase chromosomal spreads were performed as described under “Ex- perimental Procedures.” 77% of the cells examined (17/22) had signals on both chromatids of a single chromosome 3 a t 3q27-q28. Additionally, one cell had signals on both chromatids of chromosome 3, for a total of 18 double signals.

quences. A cDNA from Jok-1 cells, pJOK-8, is apparently analogous to this Daudi cDNA. pJOK-8, in addition, includes an additional 25-bp sequence 5‘ to the sequence in the Daudi clone (see Fig. 3E).

In rat liver, SiaT-1 transcription is initiated at Exon(1) (Wang et al., 1990a, 1990b). However, a larger SiaT-1 mRNA isoform has been identified in rat kidney that contains additional 5”untranslated sequences that are encoded by two additional upstream exons, Exon(-1) and Exon(0) (Wen et al., 1992). Sequence comparison of the exons that carry 5’- untranslated sequences between human and rat is shown in Fig. 3. From this comparison, it is clear that Exon(Z) is the human homolog of the rat Exon(0) (Fig. 3C). Sequence similarity between human Exon(Y) and rat Exon(-1), is not as dramatic (Fig. 30). The sum of this data suggests that the SiaT-1 mRNA expressed in HL-60 cells and also identified by Grundmann et al. (1990) in the placenta to be the human homolog of the rat kidney SiaT-1 form containing Exon(-l+O). Human Exon(X) sequence was derived from genomic analysis and a cDNA isolated from Jok-1 cells (see “Experimental Procedures”) and is shown in Fig. 3E. The Daudi cell-derived cDNA reported by Stamenkovic et al. (1990) contains 90 of the 115 bp shown in Fig. 3E. There is no known rat equivalent to Exon(X).


” 115 bp

R Ex II Ex 111 Ex IV Ex v Y

g t a a a l . . . t c l t t l l c ~ ~ TGGTACCAGAAT. g l a a a l . . . l c l g l l t c ~ a TGGTATCAGAAA.

Ex 111 Ex IV Ex V Ex VI

~ I ~ ~ ~ ~ . . . F I ~ C C ~ O C ~ O TTGGTTACCACA.

H: .“““..Jala~al.G.ccclclc~.a TATCATCATCATG

a ~ a a a ~ . . . c t c m c ~ a t s a TTAGTCACCACA. R: TCCGGCATGCTGG a l a s a l . . . o l a l a a c c s a TATCATCATCATG S G M L G

FIG. 2. Genomic organization of human SiaT-1. Panel A , diagrammatic representation of exon organization. Shaded regions denote protein coding sequences. Arrows andp3, p4, p5, p6, p15, andp19 (see “Experimental Procedures”) indicate position and polarity of synthetic oligonucleotide primers used in PCR analysis (see “Results”). The genomic spatial arrangement between Exon(X) and Exons(Y) and (Z) is undetermined. Each of these exon regions is isolated to a separate X genomic clone, suggesting vast intronic distances between these exons. Panel B , comparison of intron/exon boundaries within the coding region of human and rat SiaT-l genes. The sequence boundaries are displayed for Exons(II), (III), (IV), (V), and (VI) (boxed regions) and their respective introns (lower case letters). For each, nucleotide seauences of human ( H ) and rat ( R ) are Dresented. The single letter amino acid codes predicted by the exon domains are displayed above a n i below the respective’human and rat nucleotide sequence;

.~

SiaT-1 mRNA Expression in Cells of the B-lineage-In order t o expand on the notion that SiaT-1 expression is concomi- tan t with that of CDw75 during B-cell development, six B- lymphoblastoid lines, all well characterized for phenotype with respect to their developmental stages, were chosen for examination. The Reh line represents a precursor state of B- null, T-null stage of lymphocyte development (Rosenfeld et al., 1977). The Nalm-6 line is an often used line representing the pre-B stage of development; Jok-1, Ball-1, Daudi, and Louckes are lines exhibiting the mature B-cell phenotype (Minowada et al., 1981a, 1981b; Rossowski and Srivastava, 1983; Wang et al., 1988). Fig. 4 summarizes our assessment of CDw75 expression in these cells. CDw75 expression was tested using the anti-CDw75 antibody, HH2, and visualized with rhodamine-conjugated goat anti-mouse IgM (see “Experimen- tal Procedures”). Fig. 4 displays the immunofluorescence micrographs (panel ZI) in which the corresponding fields of cells are shown in light micrographs (panel I ) for comparison. As expected, Reh and Nalm-6 lines, both representing earlier states of B-cell development, are negative for CDw75. Jok-1, Daudi, and Louckes, cell lines exhibiting the mature B phenotype are positive for CDw75. It is curious that the Ball-1, a line that reportedly also exhibits the mature B phenotype, is not positive for CDw75.

In order to confirm that Reh and Nalm-6, but not the other lines, represent B-cells in the early stages of development, the lymphoblastoid lines were assessed for expression of terminal deoxynucleotidyltransferase, an enzyme limited to lymphocytes in early stages of development (Peterson et al., 1984; Sasaki et al., 1982). Shown in Fig. 5, panel A , is an RNA blot probed for the expression of terminal transferase mRNA. Terminal transferase transcripts are present in Reh (lane 3 ) , and at a lower level, in Nalm-6 (lane 4 ) , but absent in Jok-1 (lane 5 ) , Ball-1 (lane 6), Daudi (lane 7) , and Louckes (lane 8). As controls, terminal transferase is also not expressed in HepG2, a cell line of hepatic origin, and in MT-2, a cell line

of the T-lineage (Miyoshi et al., 1981; Koyanagi et al., 1984) (lanes 1 and 2, respectively). Fig. 5, panel B, shows a parallel RNA blot probed for SiaT-1 expression. As expected, SiaT-1 is an abundant transcript in HepG2. In contrast, MT-2 expresses little or no SiaT-1. For the B-lymphoblastoid cells, Reh and Nalm-6 cells that are CDw75 negative expressed low levels of SiaT-1 mRNA; Jok-1, Daudi-1, and Louckes, in parallel with the presence of CDw75 epitope, contains dramatically elevated SiaT-1 mRNA levels. Again, Ball-1, a line reportedly exhibiting the mature B phenotype but stained negatively for CDw75, expresses low SiaT-1 mRNA levels comparable to that found in Reh and Nalm-6. Furthermore, B-lineage cells express a t least two species of SiaT-1 mRNAs, discernible by a slight difference in mobility on RNA gels (Fig. 5B). Low SiaT-1 transcript levels in Reh, Nalm-6, and Ball-1 appears to be the result of expression of a slightly larger mRNA. Dramatically higher levels of SiaT-1 transcripts in Jok-1, Daudi, and Louckes appear to be due to the expression of a slightly smaller mRNA form. Hepatic SiaT-1 mRNA (HepG2) appears to migrate faster than either of the B-cell forms (Fig. 5B, lane 1 ) .

In order to assess further the molecular basis of SiaT-1 expression during B-cell maturation, probes that are specific to Exon(X) or Exons(Y+Z) were used to hybridize parallel RNA blots. As clearly demonstrated in Fig. 5, panel C, Exon(X) sequence is found only in Jok-1, Daudi, and Louckes lines but not in Reh, Nalm-6, and Ball-1. In contrast, Ex- ons(Y+Z) sequences are found in all B-lineage cell lines examined (Fig. 5, panel D). In addition, it is noteworthy that while the hepatoma line, HepG2, expresses high SiaT-1 levels, the HepG2 mRNA contains neither Exon(X) nor Exons(Y+Z) sequences (Fig. 5, lane 1, panels B, C, and D).

Polymerase chain reaction was used to characterize further the SiaT-1 mRNAs expressed in B-lineage cells. cDNA synthesized from Jok-1 and Nalm-6 RNAs were used. To assess the context of Exon(X) sequence in SiaT-1 mRNA from Jok-

Human Sialyltransferase Gene

A 4359

FIG. 3. Direct sequence comparison of 5”untranslated exons of human and rat SiaT- 1 genes. The DNA sequences were aligned by the method of Myers and Miller (1988). Panel A illus- trates the 5’ portion of Exon(I1) including 49 bases downstream of the initiator codon (underlined). The sequences of noncoding Exons(I), (Z), and (Y) are shown and compared to their putative rat counterparts in panels B, C, and D, respectively. Panel E displays the nucleotide sequence (Exon(X) of which there is no known rat counterpart. Arrow indicates the beginning of the SiaT-1 cDNA sequence reported by Stamen- kovic et al. (1990). The additional sequence (upstream to the arrow) was present in the cDNA clone, pJOK-8, that was derived from Jok-1 cells (see “Ex- perimental Procedures”). The first nucleotide in each exon is designated as position 1. Nucleotide identity (*) between the respective human and rat sequences is shown for each exon domain.

B

C

D

H S l m T E x 2 1- GCTTGTTTTCCTGCTCAGAACAAAGTGACTTCCCTG4ACACATC-TTCAT

RS I a TExZ H ** xa.+xIx*x* x x* * *** +x x m

1- CCGTG-GTT~TGCTCTC;CAC-A-GTGGCTCTCCTGTCTCGACCARCAT x *Hal.*

H S I a T E x Z 50- ~~TEATTCACACCAACCTGAAGAAAAAGTTCAGCTGCTGCGTCCTGGTCT *X****-* ****** ***************- ** * ***I***

RS I T E x 2 48- T~TTCATACCAACTTGAAGAMAAGTTCAQCCTCTTCATCCTC~OTCT

HS I a T E x 1

R S I a T E x l

W S I a T E x l

R S i s T E x l

H S I m T E x l

R S l a T E x l

H S I m T E x Z

R S l s T E x O

HS I a TEXZ

R S l a T E x O

HSI a T E x Z

R S i e T E x O

H S I a T I E x Y

1-

10-

50-

60-

97-

110-

ATGAGTTTTGATCATCCTGAGAAAAATGGGCCTTGGCCTGCAGACCCAAT y** *~)+*************+**** ***-***** *)+He* * * ATWTTTTTGATCATCCTGAGAAAAATGAGCCTTGGCCTCCAGACCTAGT

- ~ A - A - C C T T C C C T C C C A T G ~ T A A T A A T T C C T A A T T C C T ~ ~ ~ T - G A

GAAGTAACCTCTTTCTCATGGAGAACAGTGCTGGCTCCTGAGGATCTGGA -* x *x *x xxxx*x ** c*xx** ***x+***-, x* x*

AGGCCTGCCOCCCCTGGGGGATTAOCCAGAA(XAC3 *****xx x**** .& xx,*X*x***CXx*+*H+

GGGCCTGCAGCCCCAG4GGGATTAGCCAGAAGCAG

1- GCAAGGCtrAtPGCCAGTTCGCAGAGCCCTGCAACCAGCAGTCCA~GAAGT *c+-, -*x*** ** *+ x** .& x+ x*** x x, +e+

1- GGGTAGAGCCAGCTGTGCCAAGCTCCACAGCCACAGCCAGTG~T~~~~A~

64- GGTGAATGTCATGGAGCCCAGCTGAAATGAAATGGACTGG-CCCCCTTGAGCCTG

SI- A A T C A G T G C C C C G G A A G T C A A C T G A A C T G G C C C C C T T * * ** * *** ** **+x**** ****+ ***cl~+*. ) l **

103-.TCCC-AAGCCCTGGTGCCAGGTGTCCATGCCCGTGCTGAG

101- CCCCAAATACCTGATACCCACTTTCCAGCCCGATGCTGAG t*t ** **** t ** * +*x* r** *+*****

1- GCCCCGGCGTTA-ACAAAG-GGAGCCGATACCGACCGGCGTGGGCCCGGAG

R S I a T l E x - 1 25-

H S i s T I E x Y 5 0-

RS I T 1 Ex- 1 73-

H S l a T l E x Y 99-

R S l a T I E x - 1 1 2 2 -

CCCACGTGTGAGGGGCCGCAGAACCTACACC-ACC-GCGTCTGTARGTG )x* x * x x x XI x* x +x* x.&* x*** x x x

CGGGCGGCCGC-CACCGAGCGTGCTGAGCAACCGCAGCCTCCCCCtCCtn

TAGGC-GTTGCTTGCAGAGTGTffiTCCCTTAGATCTTGCTCACCQC3GTTA *t* x ** * * * x *** x t *** tw *

GAGTGCAGCGA

CAGATCAG-GA +* *** **

E + H S I a T I E x X 1 TAGGGCCGCACAACCAGGCAGGGCGTGGAAGCTCTGCATCCCTTCTCCCA

H S i a T I E x X 5 1 TACCTTGCTCTACACATCTCTTCATCTGTATCCTCTGCAGCATCATTGAT

H S l a T l E x X 1 0 1 GATAAACCAGTAAAT

1 cells, synthetic oligonucleotides HST-pl9 (complementary to Exon(X)) and HST-p5 (complementary to Exon(1)) were used as primers for the PCR analysis (see Fig. 2A and “Ex- perimental Procedures” for details). As seen in Fig. 6, lune I, PCR amplification of Jok-1 cDNA resulted in a 211-bp product, as expected if Exon(X) sequence is fused directly 5’ of Exon(1). No other PCR products are seen, suggesting the absence of Exon(X) isoforms containing additional sequences between Exon(X) and Exon(1). To examine the context of Exons(Y+Z) in SiaT-1 transcripts in Nalm-6 cells, oligonucleotides HST-pl5 (complementary to region in Exon(Z)) and HST-p5 (complementary to region in Exon(1)) were used. The resultant 269-bp product (Fig. 6, lane 2) is expected if and only if Exon(Z) sequences are joined directly to Exon(1). If HST-p6 (complementary to Exon(Y)) was used instead of HST-pl5 in PCR analysis of Nalm-6 cDNA, a 368-bp DNA is the resultant product. The 368-bp DNA is predicted if

Exon(Y) is fused immediately 5’ to Exon(Z) and Exon(1) (see Fig. 2.4 ).

DISCUSSION

Whatever the role and precise molecular structure of CDw75, it is clear that induced SiaT-1 expression accompanies the appearance of CDw75. Furthermore, SiaT-1 induction is the result of appearance of a novel SiaT-1 mRNA isotype. Consistent with the CDw75 being a marker for mature, active B-cells, three of the four B-lymphoblastoid cell lines with mature B phenotype displayed CDw75 and exhib- ited high SiaT-1 expression. The one exception, Ball-1, is CDw75 negative and also has a basal level of SiaT-1 expression. The two precursor B-cell lines, ‘Reh and Nalm-6, both scored negative for CDw75 and expressed basal levels of SiaT- 1. MT-2, a T-lineage cell line, expressed very low levels of SiaT-1 and does not display CDw75.


FIG. 4. Expression of CDw75 in human lymphoblastoid cells. Human lymphoblastoid cells, Reh, Nalm-6, Jok-1, Ball-1, Daudi, and Louckes were fixed on microscope coverslips, incubated with anti-CDw75 HH2 monoclonal antibody, and analyzed as described under "Experimental Procedures." In panel I , cells were visualized with DAPI, and a UV-2A filter was used. In panel ZI, a G-2A filter was used for epifluorescence visualization of rhodamine-conjugated secondary antibody.

1 2 3 4 5 6 7 8 " 7 .. .

A

B

C

FIG. 5. Northern blot analysis of HepG2, MT-2, Reh, and B-lineage lymphoblastoid cells, Nalm-6, Jok-1, Ball-1, Daudi, a n d Louckes. Panels A-D are duplicate RNA blots. In each panel, 5 pg of total RNA was isolated from HepG2, MT-2, Reh, Nalm-6, Jok- 1, Ball-1, Daudi, and Louckes (lanes 1-8 of each panel, respectively). The following probes were used (see also "Experimental Procedures"). Panel A, 1.1-kb fragment of human terminal transferase cDNA sequence; panel R, 1.3-kb human SiaT-1 cDNA fragment which spans Rxons(I1) to (VI); panel C, a 68-bp Exon(X) sequence of human SiaT- 1 gene; and panel D, 239-bp cDNA fragment of Exon(Y+Z) sequence of human SiaT-1 gene. The 68-bp Exon(X) fragment (nucleotide Nos. 36 to 103 in Fig. 4 E ) and the 236-bp Exon(Y+Z) fragment (nucleotide 15 in Fig. 4 0 to nucleotide 141 in Fig. 4C) are generated from PCR amplification of existing cDNA clones.

M r 1 2 3

bp * 368 + 269 4- 211

FIG. 6. PCR analysis of Nalm-6 and Jok-1 SiaT-1 mRNA. cDNA from Jok-1 (lane I ) or cDNA from Nalm-6 (lanes 2 and 3 ) was used as target DNA in the PCR analysis. The primers used were as follows: lane I , HST-pl9 (0.5 p ~ ) and HST-p5 (1.0 p ~ ) ; lane 2, HST-pl5 (0.5 p ~ ) and HST-p5 (1.0 p ~ ) ; and lane 3, HST-p6 (0.5 p ~ ) and HST-p5 (1.0 p ~ ) . Each cDNA was initially denatured at 93 "C for 5 min. Taq polymerase was then added, and the reactions were allowed to anneal for 20 min in the presence of HST-pl9 at 48 "C (lane I ) , HST-pl5 at 58 "C (lane 2 ) , or HST-p6 at 58 "C (lane 3 ) . This was followed by extension at 72 "C for 5 min. HSTg5 was then added to all reactions, and target sequences were amplified for 33-37 cycles at the following conditions: 93 "C, 60 s; 37 "C, 60 s; 72 'C, 115 s.

We have defined two upstream exons, Exon(Y) and Exon(Z), that, together with Exon(1) and a portion of Exon(I1) form the 5"untranslated leader of an isotype of human SiaT- 1 mRNA. Sequence similarity between Exons(Y+Z) and rat

Exons(-l+O) strongly suggests that this mRNA isotype is the human homolog of the isotype identified in rat; the Ex- ons(-l+O) containing rat isotype is believed to be the "con- stitutive" SiaT-1 form expressed in most tissues (Wen et al., 1992). Consistent with this notion is that human Exons(Y+Z) containing mRNAs are expressed in a number of cells and tissues. A cDNA representative of this isotype was initially isolated from placenta (Grundmann et al., 1990). We have also isolated a cDNA with the same isotype from HL-60, a cell line of myelocytic lineage. Furthermore, we showed that the basal SiaT-1 mRNA in B-lymphoblastoid cells are Ex- ons(Y+Z) containing transcripts (see Fig. 50).

Elevated SiaT-1 expression in Jok-1, Daudi, and Louckes cells results from induction of the Exon(X)-containing isotype (see Fig. 5C), which appears as a slightly faster migrating species than Exon(Y+Z) transcripts on RNA blots (Fig. 5B). As demonstrated by PCR analysis of Exons(Y+Z) mRNA from Nalm-6 cells and Exon(X) mRNA from Jok-1 cells (Fig. 6), usage of these exons in the generation of SiaT-1 mRNAs is mutually exclusive, a t least within the B- lymphoblastoid samples examined.

In rat, a SiaT-1 transcript isotype that is transcriptionally initiated a t Exon(1) and devoid of additional 5' sequences is expressed to high levels in liver (Wang et al., 1990a, 1990b; Weinstein et al., 1987). This is consistent with our current observation that HepG2, a human hepatoma cell line, expresses high levels of a SiaT-1 isotype that is faster migrating than either the Exon(X) or the Exons(Y+Z) forms (see Fig. 5B, lane I). Moreover, the HepG2 transcript did not hybridize to probes for Exons(X) or (Y+Z) sequences (see Fig. 5, C and D, respectively).

Taken together, the data strongly suggests that at least two and, likely, three physically distinct promoter regions are operative in the expression of SiaT-1. Yet another promoter region is responsible for a class of transcripts and proteins in rat kidney that retains less than 50% of the coding sequence and do not have SiaT-1 catalytic capabilities (Wang et al., 1990b; O'Hanlon and Lau, 1992). Of the promoters that can transcribe the complete coding sequence, one resides immediately 5' of Exon(1) (Wang et al., 1990b) and is responsible for the glucocorticoid-responsive SiaT-1 expression in hepatic tissues (Wang et al., 1990a). Since Exons(X), (Y), and (Z) reside 5' of Exon(I), at least one and likely two upstream promoter regions must exist. The collective data strongly suggest that Exon(X) and Exons(Y+Z) type mRNAs have distinctly different 5'-regions, and hence, different transcrip- tional initiation points and promoter regions. First, differen-

Human Sialyltransferase Gene 4361

tial Northern hybridization (Fig. 5 ) is highly suggestive that sequences for Exon(X) and Exons(Y+Z) do not reside on the same transcripts. Second, independent cDNA isolates from a number of different cell types contain either Exon(X) or Exon(Y+Z), but never together. Finally, PCR analysis (Fig. 6) of SiaT-1 mRNAs in Nalm-6 and Jok-1 cells confirm the mutual exclusiveness in the utilization of these exon regions for the synthesis of two SiaT-1 mRNA forms. The other possibility is that a single common upstream promoter mediates the transcription of both Exon(X) and Exons(Y+Z) transcripts. However, if only one upstream promoter is involved, a hypothetical shared region 5' of Exons(X) and (Y+Z) is expected. To date, there is no evidence for a shared 5' -region.

At least 100 distinct glycosyltransferases, including 10 or more sialyltransferases, are believed to be required to elabo- rate the complete complement of mammalian glycosidic link- ages (reviewed in Paulson and Colley, (1989), Sadler et al. (1982), and Beyer et al. (1981)). Chromosomal assignments have been reported for only a few of the glycosyltransferases. By in situ hybridization and somatic cell hybrid analysis, we have assigned the SiaT-1 gene to q27-q28 of human chromosome 3. Among other glycosyltransferases with known chromosomal localizations, the (31,4-galactosyltransferase resides at 9q13-p21 (Duncan et al., 1986); N-acetylglucosaminyltrans- ferase I resides on chromosome 5 (Hull et al., 1991); two of the closely related family of al,3-fucosyltransferases reside on chromosome 19 while a third resides on chromosome 11 (Weston et al., 1992). Enzymatic analysis of human-murine cell hybrids was also suggestive that another sialyltransferase, the a2,3-sialyltransferase involved in the modification of the 0-linked Gal@l,3GalNAca-R, resides on chromosome 11 (de Heij et al., 1988). Our assignment of SiaT-1 to chromosome 3 is consistent with the current notion that sequences encoding the glycosyltransferases are dispersed in the human genome. I t remains to be determined whether the rest of the sialyltransferase genes are likewise dispersed.

Genes of higher eukaryotes have often been classified into two broad categories. Expression of tissue-specific genes such a s albumin or ovalbumin are restricted to a limited number of cell types, while housekeeping genes are expressed in all cell types. SiaT-1, as well as many other glycosyltransferases, do not fit well into either category. Most tissues express SiaT- 1, but the level of expression varies dramatically (O'Hanlon et al., 1989; Paulson et al., 1989). Tissue differences in SiaT- 1 expression, at least in part, is the result of regulation of a single gene by multiple and distinct promoter regions. Tissue differences in expression is achieved in a different manner by another glycosyltransferase, the cul,3-fucosyltransferase, an enzyme with an equally broad range of tissue specificity. In contrast to SiaT-1 in which a single gene is regulated by multiple promoter regions, multiple fucosyltransferase genes encoding catalytically similar enzymes are differentially expressed in different tissues (Weston et al., 1992; Kumar et a[., 1991). It remains to be determined if similar mechanisms regulate the tissue distribution of the other glycosyltransferases.

REFERENCES Bast, B. J. E. G., Zhou, L.-J., Freeman, G. J., Colley, K. J., Ernst, T. d . , Munro,

J. M., and Tedder, T. F. (1992) J. Cell Biol. 116 , 423-435

Bever. T. A,. Sadler. J. E.. Rearick. J. I.. Paulson. J. C.. and Hill. R. L. (1981) Adu. Enzymol. 62, 23-i75

I _ . , . .

Cherif, D., Bernard, O., and Berger, R. (1989) Hum. Genet. 81,358-362 de Heij, H. T., Tetteroo, P. A. T., Geurts van Kessel, A, H. M., Schoenmaker,

E., Visser, F. J., and Van den Eijnden, D. H. (1988) Cancer Res. 4 8 , 1489- 1493

Do-rken, B., Moller, P., Pezzutto, R., Schwartz-Albiez, R., and Moldenhauer, G. (1989) in Leukocyte Typing IV. White Cell Differentiation Antigem (Knapp, W., Dorken, B., Gilks, W. R., Rleber, E. P., Schmidt, R. E., Stein, H., and von dem Borne, A. E. G. K., eds) pp. 109-110, Oxford University Press,

Duncan, A. M. V., McCorquodale, M. M., Morgan, C., Rutherford, T. J., Appert, Oxford

H. E., and McCorquodale, D. J. (1986) Biochem. Bwphys. Res. Commun. 141 , 1185-1188

Epstein, A. L., Marder, R. J., Winter, J. N., and Fox, R. I. (1984) J. Immunol. 133, 1028-1036

Erikstein, B. K., Funderud, S., Beiske, K., Aas-Eng, A,, De Lange Davies, C., Blomhoff, H. K., and Smeland, E. B. (1992) Eur. J. Immunol. 22,1149-1155

Fan, Y.-S., Davis, L. M., and Shows, T. B. (1990) Proc. Natl. Acad. Sci. U. S. A. 87,6223-6227

Grundmann, U., Nerlich, C., Rein, T., and Zettlmeissl, G. (1990) Nucleic Acids Res. 18,667

Guy, K., and Andrew, J. M. (1991) Immunology 74 , 206-214 Hull, E., Sarkar, M., Spruijt, M. P. N., Hoppener, J. W. M., Dunn, R., and

Koyanagi, Y., Hinuma, Y., Schneider, J., Chosa, T., Hunsmann, G., Kobayashi, Schachter, H. (1991) Biochem. Biophys. Res. Commun. 176,608-615

N., Hatanaka, M., and Yamamoto, N. (1984) Med. Microbiol. Immunol. 173 , 197-1 ~n

Kumar, R., Potvin, B., Muller, W. A,, and Stanley, P. (1991) J. Biol. Chem. L e . IT"

266.21777-21783 Lance,", P., Lau,~K. M., and Lau, J. T. Y. (1989) Biochem. Biophys. Res.

Commun. 164,225-232 Minowada, J., Koshiba, H., Sagawa, K., Kubonishi, I., Lok, M. S., Tatsumi, E.,

Han, T., Srivastava, B. I. S., and Ohnuma, T. (1981a) J. Cancer Res. Clin.

Minowada, J., Koshiba, H., Sagawa, K., Kubonishi, I., Lok, S., Tatsumi, E., Oncol. 101,91-100

Han, T., Srivastava, B. I. S., and Ohnuma, T. (1981b) Cancer Res. Clin. Oncol. 101,91-100

Mittler, R. S., Talle, M. A,, Carpenter, K., Rao, P. E., and Goldstein, G. (1983)

Miyoshi, I., Kubonishi, I., Yoshimoto, S., Akagi, T., Ohtsuki, Y., Shiraishi, Y.,

Munro, S., Bast, B. J. E. G., Colley, K. J., and Tedder, T. F. (1992) Cell 6 8 ,

J. Immunol. 1 3 1 , 1754-1761

Nagata, K., and Hinuma, Y. (1981) Nature 294 , 770-771

1 nnl Myers, E. W., and Miller, W. (1988) Comput. Appl. Biosci. 4 , 11-17 O'Hanlon, T. P., and Lau, J. T. Y. (1992) Glycobiology 2 , 257-266 O'Hanlon, T. P., Lau, K. M., Wang, X., and Lau, J. T. Y. (1989) J. Biol. Chem.

""v

264. 17389-17394 Paulson,J. C., andColley, K. J. (1989) J. Biol. Chem. 264 , 17615-17618 Paulson, J. C., Weinstein, J., and Schauer, A. (1989) J. Biol. Chem. 264,10931-

Peterson, R. C., Cheung, L. C., Mattaliano, R. J., Chang, L. M. S., and Bollum,

Rosenfeld, C., Goutner, A,, Choquet, C., Venuat, A. M., Kayibanda, B., Pico, J.

Rossowski, W., and Srivastava, B. I. S. (1983) Br. J. Cancer 4 7 , 158-160 Sadler, J. E., Beyer, T. A., Oppenheimer, C. L.. Paulson, J. C., Prieels, J. P.,

Rearick, J. I., and Hill, R. L. (1982) Methods Enzymol. 83,458-514 Sasaki, R., Takaku, F., Lin, Y.-L., and Bollum, F. J. (1982) Leukemia Res. 6 ,

197-302 Shows, T. B., Sakaguchi, A. Y., and Naylor, S. L. (1982) Adu. Hum. Genet. 12 ,

341-452 Shows, T. B. (1983) in Current Topics in Biological and Medical Research

339, Alan R. Liss, New York (Rattazzi, M. C., Scandalios, J. G., and Whitt, G. S., eds) Vol. 10, pp. 323-

Shows, T. B., Eddy, R. L., Haley, L. L., Byer, M., Henry, M., Fujita, T., Matsui,

Smeland, E., Funderud, S., Ruud, E., Blomhoff, H. K., and Godal, T. (1985) H., and Taniguchi, T. (1984) Somat. Cell Mol. Gen. 10,315-318

Stamenkovic, I., Asheim, H. C., Deggerdal, A,, Blomhoff, H., Smeland, R. B., Scand. J . Immunol. 21,205-214

Stamenkovic, I., Sgroi, D., Aruffo, A., Sy, M.-S., and Anderson, T. (1991) Cell and Funderud, S. (1990) J. Exp. Med. 172 , 641-643

Stamenkovic, I., Sgroi, D., and Aruffo, A. (1992) Cell 6 8 , 1003-1004 6 6 , 1133-1144

Svensson, E. C., Soreghan, B., and Paulson, J. C. (1990) J. Biol. Chem. 265 ,

10934

F. J. (1984) Proc. Natl. Acad. Sci. U. S. A. 81,4363-4367

L., and Greaves, M. F. (1977) Nature 267,841-843

2 0 8 ~ ~ n ~ m Wang, D., Liebowitz, D., Wang, F., Gregory, C., Rickinson, A., Larson, R.,

Wang, X., O'Hanlon, T. P., and Lau, J. T. Y. (1989) J. Bid. Chem. 2 6 4 , 1854-

-"" ""_ Springer, T., and Kieff, E. (1988) J. Virol. 6 2 , 4173-4184

1 R5Q Wang, X., O'Hanlon, T. P., Young, R. F., and Lau, d . T. Y. (1990a) Glycobiology

Wang, X., Smith, T. J . , and Lau, J. T. Y. (1990b) J. Biol. Chem. 2 6 5 , 17849-

""

1 , 25-31

17853 Weinstein, J., Lee, E. U., McEntee, K., Lai, P. H., and Paulson, J. C. (1987) J.

Wen, D. X., Svensson, E. C., and Paulson, J. C. (1992) J . Biol. Chem. 267 ,

Weston, B. W., Nair, R. P., Larsen, R. D., and Lowe, J. B. (1992) J. Bid. Chem.

Wilson, G. L., Fox, C. H., Fauci, A. S., and Kehrl, J . (1991) J. Exp. Med. 173 ,

Biol. Chem. 262, 17735-17743

2512-2518

267,4152-4160

137-146

chromosome mapping and organization of the human

Documents