DNase I hypersensitivity patterns of the serglycin proteoglycan gene in resting and phorbol 12-myristate 13-acetate-stimulated human erythroleukemia (HEL), CHRF 288-11, and HL-60 cells compared with neutrophils and human umbilical vein endothelial cells.

We mapped the DNase I-hypersensitive sites (DHSS) of the serglycin gene in resting and phorbol 12-myristate 13-acetate (PMA)-stimulated human erythroleukemia (HEL) and CHRF 288-11 cells, which have megakaryocytic characteristics, and HL-60 promyelocytic leukemia cells. We compared these DHSS with those of normal primary neutrophils and human umbilical vein endothelial cells. Several DHSS appear to be involved in regulating the level of endogenous expression and in the PMA response of hematopoietic cell lines. A DHSS unique to resting HL-60 cells and induced in CHRF 288-11 by PMA may explain the high degree of endogenous expression in HL-60 relative to HEL and CHRF (Schick, B. P., Petrushina, I., Brodbeck, K. C., and Castronuevo, P. (2001) J. Biol. Chem. 276, 24726-24735). A total of 4 DHSS in intron 1 and 6 in intron 2 are associated with the PMA response in a cell-specific manner. A DHSS in the 5'-flanking region and another in intron 1 lie in areas that have high homology with the orthologous murine serglycin locus and are rich in potential transcription factor binding sites. One DHSS in intron 1 and one in intron 2 are located within Alu repeats. Two DHSS found in DNA of normal primary neutrophils were different from those of the cell lines. One DHSS in exon 2 unique to neutrophils correlated with a previously unrecognized alternative splicing that removes exon 2. Human umbilical vein endothelial cells had a DHSS in intron 1 that was common with the cell lines. The different patterns of DHSS exhibited by the cells studied suggest that cell- and differentiation-specific alterations in chromatin structure may control serglycin gene expression.

tachment region consisting of multiple serine/glycine repeats. The GAG chains are clustered closely within this region, conferring a unique structure on this proteoglycan. The human serglycin proteoglycan carries chondroitin sulfate chains in most hematopoietic cells (1)(2)(3)(4), whereas mucosal mast cell serglycin carries chondroitin sulfate or heparin (5). The length and number of chondroitin sulfate GAG chains vary in a cellspecific fashion among these different cells. The serglycin proteoglycan is synthesized by normal hematopoietic cells, including megakaryocytes (6,7), granulocytes (eosinophils, basophils, and neutrophils), macrophages, lymphocytes, and mast cells (3, 4, 8 -10). It is also synthesized by a large number of hematopoietic tumor cells (11)(12)(13)(14), human umbilical vein endothelial cells (HUVEC) (15), and uterine decidua in mid-stage pregnant mice (16). Serglycin is likely to be important in packaging of proteins in secretory granules and in granule assembly and secretion. It is thought to be important also for modulation of the function of secreted granule proteins (17)(18)(19)(20). Serglycin can adhere to matrix proteins such as fibronectin and collagen (14,21,22) and can influence cell/matrix interactions (21). The intact serglycin proteoglycan could be involved in important modulatory roles in processes such as hematopoiesis, inflammation, and coagulation (3,8). The importance of the proteoglycan is shown by studies in which loss of correct heparin glycosaminoglycan structure causes marked defects in mast cell granules (23,24) and serglycin with abnormally short chondroitin sulfate chains is associated with platelet ␣ granule and bleeding defects in Wistar-Furth hereditary macrothrombocytopenic rats (22).
Two studies describe regulation of the serglycin gene by promoter analysis using deletion constructs. Avraham et al. (25) identify negative and positive elements in the murine serglycin promoter. A negative regulatory element was found between residues Ϫ250 and Ϫ190 and a positive regulatory element was identified between residues Ϫ118 and Ϫ81. The positive regulatory element was shown to function as an enhancer. Schick et al. (26) compared regulation of the human serglycin gene in three hematopoietic tumor cells, which include human erythroleukemia (HEL), CHRF 288-11, which have megakaryocytic characteristics, and promyelocytic HL-60 cells. By use of deletion constructs and site-directed mutagenesis two elements in the 5Ј-flanking region of the serglycin gene, (Ϫ80)ets and (Ϫ70)CRE sites, were found to be critical for gene expression. An increased expression of the serglycin mRNA in HEL and CHRF cells, but a markedly decreased expression in HL-60 cells, were observed after phorbol 12myristate 13-acetate (PMA) treatment. The effects of PMA on the promoter constructs were found to correlate with changes in mRNA expression.
Schick et al. (26) showed the presence of DNase I-hypersensitive sites (DHSS) in the serglycin gene in resting hematopoietic tumor cells. DHSS, in their most general form, represent regions of less compact chromatin that is more susceptible to nuclease digestion and gaps in the nucleosome array of the chromatin fiber that reveal the location of regulatory motifs (27). In many instances regulatory regions have been localized to such open chromatin domains. For example, in the upstream region of the ␤-globin gene locus, there are multiple DHSS that comprise a region essential for high level expression of the various ␤-globin genes (28 -30). In the present report, we have extended our studies of the serglycin gene to mapping the serglycin DHSS in resting and PMA-stimulated HEL, CHRF 288-11, and HL-60 cells and have compared these DHSS with those of normal primary blood neutrophils and HUVEC. We also correlated the pattern of hypersensitivity with the previous studies on the endogenous levels of serglycin expression in both resting and PMA-stimulated hematopoietic tumor cells (11,12,14,26,31). Our data suggest that cell-and differentiation-specific alterations in chromatin structure may potentially control serglycin expression.

EXPERIMENTAL PROCEDURES
Cell Lines-HEL cells (32) and HL-60 promyelocytic leukemia cells (33) were obtained from the American Type Culture Collection and were cultured in RPMI 1640 with 10% fetal bovine serum, 2 mM glutamine. CHRF 288-11 cells (donated by Dr. Michael Lieberman, University of Cincinnati) (34) were cultured in Fischer's medium with 20% horse serum. All media contained 100 units/ml penicillin and 100 g/ml streptomycin. All culture media and supplements were from Invitrogen except for fetal bovine serum from Hyclone (Logan, UT). The cells were maintained at 37°C in the presence of 5% CO 2 and 95% humidity.
PMA Treatment of Hematopoietic Tumor Cells-HEL, CHRF 288-11, and HL-60 cells were resuspended at 2 ϫ 10 5 per ml. PMA was added at 0.162 M final concentration. The cells were incubated for 72 h, harvested by centrifugation, and washed with RPMI with no additives.
Isolation of Neutrophils from Human Peripheral Blood-Neutrophils were isolated from human peripheral blood according to standard procedures. Briefly, 480 ml of blood was drawn into 0.1 volume of 3.8% sodium citrate anticoagulant. Blood was centrifuged at 514 ϫ g for 15 min in a Beckman tabletop centrifuge with swinging bucket rotor to separate plasma, leukocytes, and erythrocytes. The leukocyte-rich buffy coat was collected and diluted with Krebs-Ringer glucose solution (120.0 mM NaCl, 4.5 mM KCl, 1.2 mM MgSO 4 , 1.0 mM CaCl 2 , 1.7 mM KH 2 PO 4 , 8.3 mM Na 2 HPO 4 , 10.0 mM glucose, pH 7.4) and layered over Ficoll-Paque Plus. After centrifugation at 2056 ϫ g for 30 min, the leukocyte suspension was separated into mononuclear cell and granulocyte populations. The granulocyte population typically contains Ͼ90% neutrophils. Contaminating erythrocytes were removed from the granulocyte suspension by two rounds of hypotonic lysis using ice-cold distilled water and 0.2% NaCl. After aspiration of the hemoglobincontaining supernatant, the granulocytes were suspended and washed in the Krebs-Ringer glucose solution, and a viable cell count was made by trypan blue exclusion. After centrifugation at 276 ϫ g for 8 min, the pelleted granulocytes were lysed in preparation for nuclear isolation.
Isolation of Nuclei and DNase I Digestion-Nuclei were isolated from the cells as described (36), except that IGEPAL CA-630 (Sigma) was used as detergent instead of Nonidet P-40. Approximately 3 ϫ 10 8 cells were harvested and washed in ice-cold phosphate-buffered saline. The PMA-treated cells were first washed with RPMI 1640 with no additives followed by washing with ice-cold phosphate-buffered saline. The adherent cells in the PMA-treated cultures were scraped gently from the flask with a rubber policeman. The cells were resuspended in 50 ml of ice-cold lysis buffer (40 mM potassium chloride, 10 mM sodium chloride, 10 mM magnesium chloride, 10 mM Tris-HCl, pH 7.5, 0.5% IGEPAL CA-630), kept on ice for 10 min, and centrifuged at 400 ϫ g for 10 min at 4°C. The cells were resuspended, and lysis buffer treatment was repeated. Subsequently, the nuclei were resuspended in lysis buffer, and the nuclear titer was adjusted to ϳ2 ϫ 10 7 nuclei/ml. Aliquots of 0.5 ml were distributed in tubes, and to each sample an appropriate amount of DNase I buffer (1 mM magnesium chloride, 5 mM calcium chloride) was added to a total volume of 40 l. The nuclei were incubated at 4°C with increasing amounts of DNase I (Roche Applied Science) up to 100 units for 4 min. The reactions were stopped by adding 60 l of 0.5 M EDTA, and the DNA was isolated by proteinase K-SDS digestion and potassium acetate precipitation followed by RNase A digestion, phenol-chloroform extraction, and ethanol precipitation. The final purified DNA pellet was dissolved in Tris-EDTA (pH 7.5).
Southern Blot Analysis-DNAs (5 or 10 g/sample) were digested with HindIII or KpnI, and the DNA fragments were separated in 1% agarose gels by electrophoresis and transferred to nylon membranes (Hybond Nϩ, Amersham Biosciences) using the Genie electrophoretic blotter (Idea Scientific, Minneapolis, MN) at 12 V for 1 h. The membranes were hybridized at 65°C overnight in a solution containing 50 mM Tris-HCl, pH 7.5, 1 M NaCl, 1% SDS, 10% dextran sulfate (Fisher), 0.05 mg/ml salmon sperm DNA, and the probes described below were radiolabeled by the random priming method (RadPrime DNA Labeling System by Invitrogen). The membranes were washed once with 2ϫ saline sodium citrate (SSC) at room temperature for 5 min followed by 0.5ϫ SSC at 65°C for 3-5 min. Two probes, H106 and H124, 5Ј and 3Ј to the HindIII restriction site in intron 1 of serglycin gene, and one probe, K108, 3Ј to the KpnI restriction site in exon 2, were generated (see Fig. 1). The H106 (106 bp) probe was described previously (26). This probe was used to map DHSS from the 5Ј end of the serglycin gene to the ϩ3321 HindIII site. H124 (124 bp) probe was generated using as forward primer nucleotides 3340 -3361 (5Ј-GTGGAATTGATAGAC-CAGGAGG-3Ј) and as reverse primer nucleotides 3463-3444 (5Ј-TG-CAGGCATGTTAGAGAAAC-3Ј) and mapped the gene from the 3Ј side of the ϩ3321 HindIII site. K108 (108 bp) probe, just 3Ј to the Kpn site at ϩ9012, was generated using as forward primer nucleotides 9012-9033 (5Ј-GCTGCAATCCAGACAGTAATTC-3Ј) and as reverse primer nucleotides 9119 -9101 (5Ј-AAAGGTCAGTCCTCAGACG-3Ј). These probes were based on the numbering sequence of Humphries et al. (39), and their sequences were confirmed by automated sequencing using the above forward and reverse primers.
Comparison of Human and Murine Serglycin Gene Loci-We obtained the sequences of the human and murine serglycin loci using the Celera human and murine data bases and determined areas of high homology (Ͼ75%) using the Visual tools for alignment program (37,38). We compared mouse Celera SERGLYCIN Gene File mCG11329 on chromosome 10 scaffold number GAx6k02T2QGNO 19000001-19500000 to human Celera SERGLYCIN Gene file hCG96142 on chromosome 10q22 scaffold number GAx5L2HTUSATD 68306K-68322K. These sequence were aligned with GenBank TM clone M90058 and with the same sequence as found in Humphries et al. (39).
Preparation of Neutrophil Granules-After isolation from blood, the neutrophils were washed in Krebs-Ringer glucose solution (120 mM NaCl, 4.5 mM KCl, 1.2 mM MgSO 4 , 1.0 mM CaCl 2 , 1.7 mM KH 2 PO 4 , 8.3 mM Na 2 HPO 4, 10.0 mM glucose, pH 7.3). Viable cell counts were made by hemacytometer using trypan blue exclusion. Cells were resuspended to a concentration of ϳ3 ϫ 10 7 cells/ml in 1ϫ disruption buffer (100.0 mM KCl, 3.0 mM NaCl, 3.5 mM MgCl 2 , 10.0 mM PIPES, 1.0 mM AT-P(Na) 2 , 0.5 mM phenylmethylsulfonyl fluoride, pH 7.3). Cells were subjected to subcellular fractionation using a "cell cracker," a stainless steel homogenizer (H & Y Enterprises, Redwood City, CA). In it the cells were gently lysed by shearing when the suspension was forced by syringe to pass by a tungsten carbide ball-bearing through a narrow gauge channel in the homogenizer. The lysate was collected and centrifuged in a Beckman tabletop centrifuge with swinging bucket rotor for 10 min at 1100 rpm to pellet cell debris and nuclei. Then, to achieve subcellular fractionation, the post-nuclear supernatant was collected and layered over a discontinuous Percoll gradient consisting of three layers of different densities (1.050, 1.090, and 1.120 g/ml) according to the method of Borregaard (8,28,29). The preparation was spun in a Beckman J-21 centrifuge with a JA 21 fixed angle rotor at 20,000 rpm (37,000 ϫ g) for 20 -30 min at 4°C. The subcellular fractionation yielded four bands designated ␣, ␤1, ␤2, and ␥, corresponding to the neutrophil granule populations 1°(azurophil), 2°(specific), 3°(gelatinase), and secretory vesicles, respectively. The bands were collected by Pasteur pipette. To remove Percoll and to pellet the band proteins, the samples were then ultracentrifuged in a Beckman tabletop ultracentrifuge with a fixed angle rotor at 50,000 rpm for 1 h at 4°C. The protein pellets were collected and frozen at Ϫ80°C until further analysis.
Neutrophil mRNA Analysis-RNA was isolated using Trizol (Life Sciences), and reverse transcription-PCR was carried out as described previously (12). The primers flanked the entire expected cDNA. The sequences were 5Ј-atgatgcagaagctactcaaa (forward) and 5Ј-gtgtcaaggtgggaaaatcc (reverse). The bands were sequenced by automated sequencing in the institutional facility at the Kimmel Cancer Center.
Neutrophil Proteoglycan Analysis-Proteoglycans were solubilized from the granule fractions in 8 M urea, 0.2% Triton X-100, 0.1 M NaCl, 50 mM Tris HCl, pH 8.0, and isolated on Q-Sepharose (Amersham Biosciences) using a step gradient of 0.1-1.0 M NaCl in 50 mM Tris HCl. Proteoglycans from HL-60 cells and culture medium were isolated by this protocol. The proteoglycans were digested with chondroitinase ABC (Associates of Cape Cod), and the core proteins were analyzed by SDS-PAGE and Western blotting as described previously (15). Nitrocellulose blots were incubated for 1 h in 5% dry milk-blocking buffer to prevent nonspecific protein interactions. Blots were then incubated with chicken ␣-serglycin polyclonal antibody at 1:100 for 1-16 h followed by rinsing in wash buffer to remove unbound antibody. Subsequently, blots were incubated with alkaline-phosphatase-conjugated ␣-chicken IgG monoclonal antibody (Sigma) at 1:1000 for 1 h followed by rinsing in wash buffer. Finally, ␣-serglycin immunoreactivity was detected by incubation of the blots with the alkaline phosphatase substrate and developer, 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium (Kirkegard and Perry Laboratories).

DNase I Hypersensitivity Patterns of the Serglycin Gene in
Resting and PMA-treated HEL, CHRF 288-11, and HL-60 Cells-The DHSS sites of the control and PMA-treated cell lines were determined by analysis of Southern blots with three different probes: H106, H124, and K108. The results of these analyses are summarized in Fig. 1. We will describe the results in terms of the DHSS found with each of these probes.
Sites Identified in DNA of Resting and PMA-treated Hematopoietic Cells with Probe H106 -We reported recently (26) that mapping of DHSS of serglycin in three hematopoietic tumor cells, HEL, CHRF, and HL-60, with the H106 probe disclosed two hypersensitive sites that were common to all three cell lines. We now designate these sites as DHSS I and DHSS II, as shown in Figs. 1 and 2, a and b. DHSS I is at Ϫ276 bp (numbering is from Ref. 39) in the gene and appears as a 3.6-kb fragment on the Southern blot, and DHSS II is at ϩ825 bp in the gene and is seen as a 2.5-kb fragment on the Southern blot. DHSS I and DHSS II appeared as very faint bands in HEL compared with CHRF and HL-60 cells (Fig. 2, a and b). We also described previously the DHSS III at ϩ2075 bp in intron 1, identified as a 1.25-kb fragment on the Southern blot (26), which is unique to HL-60 among the resting cells. We now show that DHSS I and DHSS II remain after PMA treatment of all three cell lines. No new DHSS were identified in HEL DNA with the H106 probe. However, PMA treatment of CHRF cells induced DHSS III (Fig. 2d), and PMA treatment of HL-60 cells induced two more hypersensitive sites in intron 1, DHSS IV at ϩ3090 bp (0.235 kb on the blot) and DHSS V at ϩ3200 bp (0.125 kb on the blot) ( Figs. 1 and 2e).
The expected parent band of 5.2 kb (closed arrowhead) and an extra band of 23.1 kb (designated as a), not a DHSS, were observed in both the resting and PMA-stimulated hematopoietic tumor cells (Fig. 2, a-e). A band of 8.1 kb (designated as b),

FIG. 5. Analysis of serglycin DNase I-hypersensitive sites in neutrophils.
Nuclei from resting neutrophils were isolated and digested with increasing concentrations of DNase I, and DNA was isolated and purified (see "Experimental Procedures"). DNA was digested with HindIII or KpnI, electrophoresed on 1% agarose gels, and blotted. The HindIII blots were hybridized with the 32 P-labeled H106 and then H124 probes, and the KpnI digest was hybridized with the K108 probe. not a DHSS, was seen only in resting CHRF cells (Fig. 2, a and  b). The origin of this band is not known.
Sites Identified with Probe H124 in Resting and PMA-treated Hematopoietic Cells-DHSS VI at ϩ5325 bp appeared as a 2.0-kb fragment (Fig. 3, b and e) in DNA from undifferentiated and PMA-stimulated HL-60 cells and PMA-treated CHRF cells (Fig. 3d) but was not seen in resting CHRF cells or in HEL cells with or without PMA treatment. The site was visualized at much lower DNase concentration in HL-60 cells (5 units/ml) compared with CHRF (18 units/ml) (Fig. 3, a and b) and, thus, is stronger in the HL-60 cells. A new site, DHSS VII, in intron 1 at ϩ5675 bp, was identified in HEL cells after PMA treatment. The expected 11-kb parent band was observed in both the resting and PMA-stimulated hematopoietic tumor cells (Fig. 3, a-e).
Sites Identified with Probe K108 in Resting and PMA-treated Hematopoietic Cells-No DHSS were detected with K108 in resting HEL, CHRF, or HL-60 cells or in PMA-treated HEL and CHRF cells (not shown). The data for PMA-treated HL-60 cells are shown in Fig. 4. Six bands representing the hypersensitive sites, DHSS X at ϩ9,451 bp, DHSS XI at ϩ9,871 bp, DHSS XII at ϩ10,001 bp, DHSS XIII at ϩ10,251 bp, DHSS XIV at ϩ11,031 bp, and DHSS XV at ϩ13,651 bp were detected in intron 2. These sites are shown at 0.45, 0.87, 1.0, 1.25, 2.03, and 4.65 kb, respectively, on the Southern blot.
A band of 7.7 kb, not a DHSS, was observed in the blots probed with K108 from all three PMA-stimulated hematopoietic tumor cells, whereas a 6.1-kb band, also not a DHSS, was seen in resting HEL, CHRF, and HL-60 cells (not shown).
DNase I Hypersensitivity Pattern of the Serglycin Gene in Neutrophils and HUVEC-We extended our analysis of the chromatin structure of the serglycin gene to resting and normal primary blood neutrophils and HUVEC. No DHSS were detected with the H106 probe in either cell (Figs. 1 and 5a) (data not shown for HUVEC). Analysis of the neutrophil DNA with the H124 probe revealed DHSS VII at 5675 bp in intron 1 (band at 2.35 kb). When neutrophil DNA was digested with KpnI and probed with K108 ( Figs. 1 and 5c) DHSS VIII was detected at 9076 bp, shown at 0.075 kb on the Southern blot, and was mapped to exon 2. Another site, DHSS IX, was mapped to intron 2 at 9177 bp. These sites are unique to neutrophils.
The expected parent band for the H106 probe was observed (designated as b in Fig. 5a), but not the 11 kb for the H124 probe. Other bands appeared on the Southern blot at 23.1 kb from H106, H124, and K108 (designated as a in Fig. 5, a-c) and 12.5 kb from K108 (designated as c in Fig. 5c).
The results for HUVEC are shown in Fig. 6. Using the H124 probe, DHSS VI, which was found in some hematopoietic tumor cells but not neutrophils, was identified. The expected 11-kb parent band for the H124 probe was observed. A band at 8 kb was present but did not seem to be a DHSS. No hypersensitive sites were observed when HUVEC DNA was probed with H106, but a band (other than DHSS) at 23.1 kb was found (data not shown) as in the neutrophils. No DHSS were detected with the K108 probe but a fragment other than DHSS was seen at 9.4 kb on the Southern blot (not shown).
Relationship of DHSS in Exon 2 to Alternative Splicing of Serglycin in Neutrophils-Analysis of the core protein of serglycin from isolated neutrophil granules revealed the presence of a 21-kDa core, which is considerably smaller than the 27-28-kDa core that is generally found in hematopoietic cells. A minor band of this size was also detected in serglycin from HL-60 cells (Fig. 7). To determine the origin of this band, serglycin mRNA was subjected to reverse transcription-PCR. The expected band of 509 bp was seen on agarose gel electrophoresis. In addition, there was a faint band at 361 bp (not shown). The 361-bp band was reamplified with the same primers and sequenced. The sequencing showed that the reduced size of this band was due to complete absence of the 2nd exon. The partial sequence around this alternative splice site was tcatcctggttctggaatcctcagttcaaGAaagacgagaatccagg, where the capital letters show the junction. The loss of exon 2 (49 amino acids) is consistent with the difference in M r between the neutrophil granule bands, which we have identified, and the expected 27-28-kDa band. We hypothesize that the DHSS in exon 2, which is unique to neutrophils, may be responsible for this alternative splicing.  (12,13,26) and CHRF (14,26) and in K562 cells (31), all of which have megakaryocytic characteristics. In contrast, serglycin mRNA expression is down-regulated in myeloid cell lines such as HL-60 in response to PMA (26,31,40,41). The mechanisms for the differences in basal activity between the cell lines and the opposite responses to PMA in HL-60 compared with cell lines with megakaryocytic potential are not understood. We propose that differences in chromatin organization among resting cells and changes induced by PMA, as demonstrated by the presence of DHSS, may explain some of these observations. Several of the DHSS appear to be involved in regulating the level of endogenous serglycin expression and in the PMA response in the hematopoietic cell lines. DHSS III, which is unique to HL-60 cells in resting cells, may explain the high degree of serglycin expression in resting HL-60 cells relative to HEL and CHRF. DHSS VI may be related to the endogenous high expression in HL-60 cells and the previously reported increased serglycin expression in PMA-treated CHRF cells (14,26). DHSS VII may be related to the increased serglycin expression in HEL cells. DHSS IV, V, and X-XV appear only in PMA-treated HL-60 and may have a combined effect on the suppression of serglycin expression by PMA in these cells. It is interesting to compare the DHSS of normal versus tumor cells used in these studies. The DHSS VI at ϩ5325 bp seen in HUVEC DNA was seen in resting and PMA-treated HL-60 cells and PMA-treated CHRF cells, suggesting that a common transcriptional mechanism may be involved, but this DHSS was not seen in neutrophil DNA. The absence of this band from neutrophil DNA may be due to the maturity of the blood cells compared with HL-60-like bone marrow progenitors.
Some of the DHSS lie in regions of the human serglycin gene, which are of interest because of their homology with the murine serglycin gene locus (Table I) and because they are in regions that are rich with transcription factor binding sites, which may be involved in regulation of serglycin expression. For example, DHSS I at Ϫ276 bp is within the 65% conserved region between mouse and human serglycin and is only 101 bases upstream from the proximal promoter region, which is 95% conserved (39,42,43). It is about 200 bp upstream from the critical ets and CRE sites, which were found to be the two major regulatory elements for constitutive expression of serglycin gene in HEL and CHRF cells (26). Similarly, DHSS II is found only 151 bp 5Ј to a region in intron 1 that has high homology to a similarly placed region of intron 1 in the murine gene locus and is the only intronic region with Ͼ75% homology. This homology was revealed by a search of the Celera human and murine data bases (Table I). This region is rich in PEA-1 and AP-1 potential transcription factor binding sites. DHSS III, which is unique to HL-60 cells and PMA-treated CHRF cells and is located in intron 1 at ϩ2075 bp (26), contains multiple potential E-box sites, motifs of the sequence CANNTG that bind basic helix-loophelix proteins, and may be related to the high basal serglycin expression in resting HL-60 relative to resting HEL or CHRF cells.
Several of the DHSS appear to be located well within or very close to Alu repeats. Alu repeats are members of the SINE (short interspersed element) family of human repetitive DNA sequences (44). These Alu elements were initially estimated at Ͼ500,000 copies in the haploid human genome (45). A recent detailed analysis of the draft sequence of the human genome (46) has shown that of more than one million copies Alu elements are the most abundant short interspersed elements. The typical Alu repeats are ϳ300 bp in length and are commonly found in introns, 3Ј-untranslated regions of genes, and intergenic genomic regions (47). Alu repeats are found only every 3-5 kb in the human genome (48). The human serglycin is unusual because it contains an average of Ͼ1 Alu element/kb (39); 19 Alu repetitive DNA sequences are interspersed in the two introns of the human serglycin gene, and two others are present in the 5Ј-flanking region. Several of the DHSS in intron 1 of the serglycin gene lie within or very close to Alu repeats. DHSS VII at ϩ5675 bp lies in an Alu region between nucleotides ϩ5479 and ϩ5756, and DHSS X lies in the Alu region from 9333 to 9622 bp. DHSS XI at 9871 bp and DHSS XII at ϩ10001 bp are between two closely apposed Alu regions from 9627-9759 bp and 10392-10519 bp. Our study is the first to implicate cell-specific and differentiation-specific Alu-related DHSS in the regulation of expression of a gene.
Alu repeats have been implicated in the transcriptional regulatory regions of a number of genes. It is interesting to compare the serglycin Alu-related DHSS to those that have been implicated in regulation of other genes. These include a negative regulatory domain in the promoter region of human erythropoietin receptor gene extending from nucleotides Ϫ1050 to Ϫ450 (49), a transcriptional silencer in the third intron of the Wilms' tumor gene WT1 (50), two hypersensitive sites distal to the promoter region of keratin 18 gene in transgenic mice (51), and two adjacent cis-acting regulatory elements in the 5Ј sequence of the ␥ chain gene, both of which are part of an Alu repeat (52). The ␥ chain element between nucleotides Ϫ445 and Ϫ336 was a positive element in both basophils and T cells, whereas the other between nucleotides Ϫ365 and Ϫ264 acted as a negative element in basophils and a positive one in T cells. Several studies have TABLE I Conserved regions between human and mouse serglycin gene A search from the Celera human and murine database revealed several conserved regions between human and mouse serglycin using the criteria of a 75% identity over 100 bp. These conserved regions are shown with the corresponding DNA regions found in the clone reported by Humphries et al. (39). The start of the sequence of the clone is at Ϫ1844 bp based on the numbering of Humphries et al. (39) and 1 based on the numbering of GenBank™. This corresponds to 8127 bp (human) in the Celera database. TATC is a reverse GATA site exactly at Ϫ1748 in the human gene. shown physiological relevance of Alu regions. Recombination events between Alu elements resulted in the deletion of a portion of the human factor VIII gene in a patient with hemophilia A (53), and the duplication of seven exons in low density lipoprotein receptor gene was found in a patient with familial hypercholesterolemia (54). Genetic disorders can result from different types of mutations that arise after the insertion of an Alu repeat (47). Alu insertions account for ϳ0.1% of all human genetic disorders, such as hemophilia, neurofibromatosis, and breast cancer (55). The physiologic importance of the potential involvement of the DHSS associated with Alu regions in serglycin gene regulation remains to be determined. The DHSS found in DNA of normal neutrophils were different from those from the HL-60 cell line, which would be representative of immature myeloid cells. Striking was the mapping of DHSS VIII to exon 2 in the neutrophils and DHSS IX to a proximal region of intron 2. We identified an mRNA in neutrophils that codes for a serglycin core protein with complete deletion of exon 2. We hypothesize that the DHSS VIII in this region is involved in establishing the configuration needed for this alternative splicing. This mRNA is consistent with the serglycin core protein from isolated neutrophil granules, which migrates at about 21 kDa after chondroitinase digestion. The major core protein in HL-60 cells (as with all other known human serglycins) is 26 -28 kDa, but a minor component at 21 kDa is also present. This mRNA was a minor component of the neutrophil reverse transcription-PCR products but may be more highly translated into protein in the mature cells than the full-length mRNA, because its product is the only one that we could detect on Western blotting. Alternatively, the proteoglycan with the full-length core protein may have been secreted during the preparation of the granules or present in a compartment that we did not analyze. Stellrecht et al. (9) report a differential level of serglycin proteoglycan gene expression in the granulocytes as they matured from promyelocytes to segmented neutrophils. In situ hybridization to normal bone marrow and peripheral blood leukocytes demonstrated that the gene was expressed in the promyelocytes at an ϳ2-fold greater level than in the segmented neutrophils, and the expression decreased as the granulocytes matured. It is possible that the differences we observed in the DHSS between HL-60 cells and neutrophils reflect stage-specific differences in regulation.
The significance of the exon 2 deletion for serglycin biology is uncertain but may be explained by our previous finding that the N-terminal portion of the sequence encoded by exon 2 contains a potential heparin binding site, and dimerization of the peptide sequence YPTQRARYQWVRCNP via the cysteine residue results in significant binding affinity to low molecular weight heparin (56). This peptide comprises the N-terminal region of the mature protein in cells that do not delete exon 2, and thus, the deletion would generate a different N-terminal sequence that could influence neutrophil-specific interactions of the serglycin core protein with other proteins. These interactions may also be influenced by the apparent low sulfation of neutrophil granule serglycin that we have reported here. Alternatively, if all or a portion of the protein region coded by exon 2 is normally removed from the mature proteoglycan by proteolysis and maintains independent function as either a monomer or dimer either within or after secreted from other cells, this function would be absent in neutrophils.