Histochem Cell Biol DOI 10.1007/s00418-016-1413-z

ORIGINAL PAPER

Characterization of the enhancer element of the Danio rerio minor globin gene locus Anastasia V. Nefedochkina1,2 · Natalia V. Petrova1 · Elena S. Ioudinkova1 · Anastasia P. Kovina1,2 · Olga V. Iarovaia1 · Sergey V. Razin1,2 

Accepted: 19 January 2016 © Springer-Verlag Berlin Heidelberg 2016

Abstract In Danio rerio, the alpha- and beta-globin genes are present in two clusters: a major cluster located on chromosome 3 and a minor cluster located on chromosome 12. In contrast to the segregated alpha- and beta-globin gene domains of warm-blooded animals, in Danio rerio, each cluster contains both alpha- and beta-globin genes. Expression of globin genes present in the major cluster is controlled by an erythroid-specific enhancer similar to the major regulatory element of mammalian and avian alphaglobin gene domains. The enhancer controlling expression of the globin genes present in the minor locus has not been identified yet. Based on the distribution of epigenetic marks, we have selected two genomic regions that might harbor an enhancer of the minor locus. Using transient transfection of constructs with a reporter gene, we have demonstrated that a ~500-bp DNA fragment located ~1.7 Kb upstream of the αe4 gene possesses an erythroidspecific enhancer active with respect to promoters present in both the major and the minor globin gene loci of Danio rerio. The identified enhancer element harbors clustered binding sites for GATA-1, NF-E2, and EKLF similar to the enhancer of the major globin locus on chromosome 3. Both enhancers appear to have emerged as a result of independent evolution of a duplicated regulatory element present in an ancestral single alpha-/beta-globin locus that existed before teleost-specific genome duplication.

* Sergey V. Razin [email protected] 1

Institute of Gene Biology, Russian Academy of Sciences, Vavilov Street 34/5, Moscow, Russia 119334

2

Molecular Biology Department, Biological Faculty, Lomonosov Moscow State University, Moscow, Russia 119992



Keywords  Zebrafish · Globin genes · Enhancer · Promoter · Chromatin

Introduction The globin gene domains constitute a favorable model that has been used for decades to study mechanisms of eukaryotic gene expression (Dillon and Sabbatini 2000; Higgs et al. 2006; Razin et al. 2003; Trimborn et al. 1999). In warmblooded vertebrate animals, alpha- and beta-globin genes are segregated into two domains located on different chromosomes that are regulated distinctly (Recillas-Targa and Razin 2001). The domain of the beta-globin genes demonstrates a lineage-specific change in DNAseI sensitivity and replication timing (Forrester et al. 1990). Transcriptional status of this domain is controlled by a block of enhancers known as the locus control region, LCR (Grosveld et al. 1987), which is flanked by insulator and has been shown to confer a position-independent and erythroid-specific expression pattern to linked genes in ectopic chromosomal positions (Grosveld et al. 1987; Li et al. 2002; Razin et al. 2003). The domain of alpha-globin genes is located in a chromosomal region that also harbors housekeeping genes, and this region resides in a DNAseI-sensitive open chromatin configuration in both erythroid and non-erythroid cells (Craddock et al. 1995). Expression of the alpha-globin genes is controlled by a potent enhancer known as the major regulatory element, MRE (Chen et al. 1997; Higgs et al. 1990; Jarman et al. 1991). The MRE is not flanked by an insulator and possesses only a limited ability to protect transgenes from position effects (Sharpe et al. 1992). In cold-blooded animals, including amphibians and fish, the alpha- and beta-globin genes are not segregated. Thus, in Danio rerio both alpha- and beta-globin genes are organized in two mixed clusters present on chromosomes 3

13



and 12 (Brownlie et al. 2003; Ganis et al. 2012). On either chromosome, the fused domain contains alternating alphaand beta-globin genes. With regard to their expression timing during development, both domains are split into individual stage-specific subdomains: embryonic and embryonic-larval in the minor locus and embryonic-larval and adult in the major locus (Brownlie et al. 2003; Ganis et al. 2012). In the both loci, these stage-specific subdomains are separated by an approximately 9 Kb spacers. In the course of individual development, the repertoire of expressed globin genes changes at least twice according to two developmental switches: from the embryonic to the embryonic-larval stage and from the embryonic-larval stage to the adult stage. Comparison of regulatory systems controlling the expression of globin genes in Danio rerio with those operating in warm-blooded animals may shed light on the evolution of these systems and help to understand the reasons for different regulation algorithms used in alpha- and beta-globin genes of birds and mammals. Little is known about the presence and distribution of regulatory elements in the Danio rerio joint alpha-/betaglobin gene domains. An erythroid-specific enhancer (a homolog of a homeothermic MRE occurring within a housekeeping gene NPRL3 neighboring the major globin gene locus on chromosome 3) is the only regulatory element mapped (Ganis et al. 2012). No regulatory element has been characterized for the minor globin gene locus (chromosome 12). Here, we report identification of an erythroid-specific enhancer element that is located immediately upstream of the cluster of alpha-/beta-globin genes on chromosome 12, is active with respect to promoters of alpha-globin genes, and shares some structural features with the MRE of the major globin gene locus.

Materials and methods Cell culture The avian erythroblastosis virus-transformed chicken erythroblast cell line HD3 [clone A6 of the line LSCC] (Beug et al. 1982a, b) was grown in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 2 % chicken serum and 8 % fetal bovine serum (FBS) at 37 °C under 5 % CO2. Chicken embryonic fibroblasts (CEFs) were isolated from 9-day-old chicken embryos according to the standard protocol and grown in DMEM supplemented with 8 % FBS and 2 % chicken serum.

Histochem Cell Biol

The fish were kept under standard conditions in aquariums 20 L in volume at a temperature of 28 °C and illuminance corresponding to a 12 h day and 12 h night. All zebrafish experiments and procedures were performed as approved by the Ethics Committee of the Institute of Gene Biology. Gene expression analysis Approximately 10 zebrafishes were collected at different time intervals after fertilization (12–74 dpf) and homogenized in TRIzol Reagent (Invitrogen, Life Technologies, Carlsbad, CA, USA). Adult fish were snap-frozen in liquid nitrogen and were also homogenized in TRIzol. Total RNA was extracted using the manufacturer’s instructions and was treated with DNase I (Fermentas UAB, Vilnius, Lithuania) to remove residual DNA. RNA (2 µg) was reverse transcribed in a total volume of 20 µl for 1 h at 42 °C using 0.2 µg random hexamer primers and 200 U reverse transcriptase (Fermentas UAB) in the presence of 20 U of ribonuclease inhibitor (Fermentas UAB). The cDNAs obtained were analyzed by TaqMan™ real-time polymerase chain reaction (real-time PCR) using the CFX96 real-time PCR detection system (Bio-Rad, Hercules, CA, USA). For analysis of the expression of the mature mRNA of globin genes, the amplicon was designed within an exon. For analysis of the non-spliced transcripts of globin genes, amplicons were designed in such a way that they spanned intron–exon junctions. The PCR primers and TaqMan™ probes are presented in Table 1. The levels of the luciferase genes (firefly and Renilla) transcription in the transiently transfected HD3 cells were estimated using a semi-quantitative RT-PCR. The constructs with the firefly luciferase reporter gene expressed under the control of βe2Δ promoter with or without RE1Δ regulatory element were transfected in HD3 cells. The firefly luciferase basic vector was used as a control. The vector RL-CMV harboring Renilla luciferase gene expressed under the control of CMV promoter was co-transfected as the transfection efficiency control. One day after the transfection, the cells were collected and homogenized in Trizol Reagent. Total RNA was extracted and treated with DNase I. RNA (2 µg) was reverse transcribed as described above. The reaction without the reverse transcriptase was used as a control (RT(−) control). The cDNAs obtained were analyzed using a semi-quantitative PCR. After electrophoretic separation in 1 % agarose gel, the amplification products were visualized by ethidium bromide staining. The PCR primers (‘Luc’ for the firefly luciferase and ‘R-Luc’ for the Renilla luciferase) are presented in the Table 1.

Zebrafish maintenance Chromatin immunoprecipitation (ChIP) Wild-type zebrafish (Danio rerio; AB type) were staged, raised, and maintained under standard laboratory conditions as described (Kimmel et al. 1995; Westerfield 2000).

13

Blood was extracted from 18 dpf zebrafish. Fish were euthanized by an overdose of tricaine (ethyl 3-aminobenzoate

Histochem Cell Biol Table 1  Primers used for expression analysis

αe4 exon direct

5′ CGAGGGTGCGAAGAACATAA 3′

αe4 exon reverse

5′ GTTTGTAGGATGTATTCGCAGTTG 3′

βe3 exon direct

5′ GCTCTTGATATTATCCATGTTGTTG 3′

βe3 exon reverse

5′ CACAGAGGCAATCATGGCTAA 3′

βe2 exon direct

5′ CTCGGCGTAGGTGTTCTTGAT 3′

βe2 exon reverse

5′ ATCCCTGGACTCAGAGATACTTTG 3′

αe5 exon direct

5′ GTGGACTCGGTCTGGCTGTT 3′

αe5 exon reverse

5′ GCGTGCAATTCACTGAGGTT 3′

αe4 junction direct

5′ GCACCACTTAGCAGGTTCCA 3′

αe4 junction reverse

5′ AACAAAGAATACGTAAGTGTACTGGAC 3′

βe3 junction direct

5′ CGTTATAGTTTCTGTATGTACAAATATCC 3′

βe3 junction reverse

5′ TCGAGACCTTGACAAGGTAATCTT 3′

βe2 junction direct

5′ CAATCACGATTGTCAGGCAGT 3′

βe2 junction reverse

5′ TCTGAGACTCACCTGTTCCGTT 3′

αe5 junction direct

5′ CAACGGCCTGCTGAACCT 3′

αe5 junction reverse

5′ CTGCAACTGTTGGACTCAATTAG 3′

αe4 exon taqman

5′(FAM)TGGCACCAC(BHQ1)TTAGCAGGTTCCATG-P 3′

βe3 exon taqman

5′(FAM)AGCGCACGG(BHQ1)TGTTGTGGTCCT-P 3′

βe2 exon taqman

5′(FAM)TGCGACCT(BHQ1)TTGGGTTGTTAATGATG-P 3′

αe5 exon taqman

5′(FAM)ACGACCT(BHQ1)TTTCAACGGCCTGCTG-P 3′

αe4 junction taqman

5′(FAM)TGTAGGA(BHQ1)TGTATTCGCAGTTGGTAGGC-P 3′

βe3 junction taqman

5′(FAM)CAAACCGAA(dT-BHQ1)AATTAAACGTGTTGACGAAGT-P 3′

βe2 junction taqman

5′(FAM)CGGCCAGCAGC(BHQ1)TGAAACAGAGG-P 3′

αe5 junction taqman

5′(FAM)AGCTGAGAG(BHQ1)TCGACCCTGCTAACTTCA-P 3′

slc4a exon direct

5′ CGCTACAAGCATTACCTTAGTGAC 3′

slc4a exon reverse

5′ AGTCCTCCAAAGGTGATCGC 3′

slc4a exon taqman

5′(FAM)CAGCGAAGTAGATGAAGATGACAGCAGA(BHQ1) 3′

Luc dir

5′ GCTATGAAACGATATGGGCTGAAT 3′

Luc rev

5′ AACCGTGATGGAATGGAACAACA 3′

R-luc dir

5′ TTGTGCCACATATTGAGCCAGTAGC 3′

R-luc rev

5′ TAATGTTGGACGACGAACTTCACCTT 3′

methanesulfonate, 100 mg/L). The blood was collected into red cell isolation buffer (0.14 M NaCl, 20 mM HEPES pH 8.0, 5 % fetal bovine serum, and 100 U heparin/mL). For each ChIP, a minimum of 2 × 107 cells were used. For NChIP analysis of histone modifications, the cells were pelleted briefly and then resuspended in a lysis buffer solution containing 0.32 M sucrose, 100 mM NaCl, 50 mM KCl, 10 mM Tris–HCl (pH, 8.0), 0.5 mM CuSO4, and 1 % Triton X-100. This, and all other solutions, contained 5 mM sodium butyrate (to inhibit histone deacetylase activity) and a protease inhibitor cocktail (Sigma-Aldrich, St. Louis, MO, USA). After 15 min of incubation on ice, the cells were pelleted and then resuspended in the same buffer without CuSO4. The nuclei were pelleted, washed twice with digestion buffer (50 mM Tris–HCl [pH 8.0], 3 mM MgCl2, 1 mM CaCl2, 0.3 M sucrose, and 100 mM NaCl), and resuspended in 0.5 ml of digestion buffer. Micrococcal nuclease (Fermentas UAB) was added to the samples at a concentration of 60 U/μl, and the samples were incubated

for 10 min at 28 °C. The reactions were stopped by the addition of EDTA to a final concentration of 10 mM. The solution was pelleted, and the supernatant containing the solubilized chromatin was used for further analysis. Immunoprecipitation reactions were performed overnight at 4 °C in 1 ml of solution containing 50 mM NaCl, 20 mM Tris– HCl (pH 8.0), 5 mM EDTA, 20 mM sodium butyrate, protease inhibitor cocktail, 100 μL of the supernatant fraction, and 5 μg of rabbit polyclonal antibodies against modified histones (Active Motif, Carlsbad, CA, USA). For XChIP analysis of GATA1 deposition, the blood was extracted from 20 dpf zebrafish. Fish are euthanized in an overdose of tricaine (ethyl 3-aminobenzoate methanesulfonate, 100 mg/L). The blood was collected into red cell isolation buffer (0,14 M NaCl, 20 mM HEPES pH 8.0, 5 % fetal bovine serum and 100 U heparin/mL). For each ChIP, a minimum of 1 × 107 cells were used. The cells were fixed with 1 % v/v formaldehyde in the red cell isolation buffer at room temperature for 10 min. The formaldehyde

13



Histochem Cell Biol

Table 2  Primers used for NChIP analysis

Table 3  Primers used for XChIP analysis

a direct

5′ AATGCTATCGCACTTGAGGTC 3′

a direct

5′ CCACAGCTTGAGTCACTGATAGG 3′

a reverse

5′ CAGGTCCATTCTTAGAGATTATGTG 3′

a reverse

5′ TGCAGCTATGATGAGCATGTACT 3′

b direct

5′ CGAGGGTGCGAAGAACATAA 3′

b direct

5′ CGAGGGTGCGAAGAACATAA 3′

b reverse

5′ GTTTGTAGGATGTATTCGCAGTTG 3′

b reverse

5′ GTTTGTAGGATGTATTCGCAGTTG 3′

c direct

5′ CACCAACACACTTGCATTGAAA 3′

c direct

5′ CACCAACACACTTGCATTGAAA 3′

c reverse

5′ ATGCGGTGGTGTCCAAATC 3′

c reverse

5′ ATGCGGTGGTGTCCAAATC 3′

Neg direct

5′ TGACTCTAGGCCTTAATCTACCAT 3′

Neg direct

5′ ACTTGTTGTTAATGTGAAACCACG 3′

Neg reverse

5′ TACAAACCATTTAACAGCCATTC 3′

Neg reverse

5′ CATCCGCTTATGATGCTTTTG 3′

MRE direct

5′ CTCAATTATGCTGAGCACTGTGTA 3′

MRE direct

MRE reverse

5′ CCCTCCATGACTCTGCACTTT 3′

a taqman

5′ (FAM)ACCCTTGAGGACGAT(BHQ1) ATTACAATCATACGC-P 3′

b taqman

5′(FAM)TGGCACCAC(BHQ1)TTAGCAGGTTCCATG-P 3′

c taqman

5′(FAM)CTGGACAGT(BHQ1)TTGGCACCCTGTAGGAA-P 3′

Neg taqman

5′(FAM)CTGTGGCCAGAT(BHQ1)TGACCACTGAAGG-P 3′

5′ CTCAATTATGCTGAGCACTGTGTA 3′ MRE reverse 5′ CCCTCCATGACTCTGCACTTT 3′ a taqman 5′ (FAM)TGCTGGCACT(BHQ1)GGCTCTGATAAACC-P 3′ b taqman 5′(FAM)TGGCACCAC(BHQ1)TTAGCAGGTTCCATG-P 3′ c taqman 5′(FAM)CTGGACAGT(BHQ1)TTGGCACCCTGTAGGAA-P 3′ Neg taqman 5′(FAM)AGGACT(BHQ1)GAGATGGCAAAGCAAGAACG-P 3′

MRE taqman 5′(FAM)CGCATGT(BHQ1)CAGTTGTTGCATTTTTAGG-P 3′

MRE taqman 5′(FAM)CGCATGT(BHQ1)CAGTTGTTGCATTTTTAGG-P 3′

was inactivated by adding glycine to the final concentration 0.125 M. The fixed cells were pelleted at 1100 g for 5 min at 4 °C, washed with red cell isolation buffer without heparin and containing 1 mM PMSF and 1 × protease inhibitor cocktail (ThermoScientific), pelleted again, and resuspended in 300 μl of lysis buffer (50 mM Tris–HCl pH 8.0, 1 % SDS, 10 mM EDTA, 1 mM PMSF, and 1× protease inhibitor cocktail (ThermoScientific)). The lysate was incubated for 15 min on ice and sonicated with a Bioruptor (Diagenode) (high power, 20 cycles for 30 s each with 30 s intervals). The cell debris was removed using a microcentrifuge (5 min, 13 000 rpm, 4 °C), and the supernatant was diluted tenfold with the immunoprecipitation buffer (16.7 mM Tris–HCl, pH 8.0, 167 mM NaCl, 2 mM EDTA, 1,1 % Triton X-100, 0.01 % SDS, 1 mM PMSF, 1x protease inhibitor cocktail (ThermoScientific)). The cell lysates were incubated with 5 μg of anti-Gata-1a (IN), Z-Fish™ (AnaSpec) antibodies overnight at 4 °C with rotation. Immune complexes were collected using magnetic beads coated with immobilized protein G (Invitrogen Life Technologies, Carlsbad, CA, USA). The DNA samples eluted from the beads were analyzed by TaqMan real-time PCR. The sequences of the primers and the TaqMan probes are listed in Tables 2 and 3.

fragments containing potential regulatory elements were obtained using PCR amplification of zebrafish genomic DNA with the pairs of primers presented in Table 4. The fragments were cloned into the pGL3 vector (Promega Corporation, Madison, WI, USA) in the following order: the test fragment containing the predicted promotor was cloned upstream from the luciferase gene, and the test fragment containing the potential enhancer was cloned downstream of the luciferase gene. Fragments containing the potential enhancer were cloned in two possible orientations: direct genomic and reverse genomic. Transfection of the luciferase constructs [1 µg of various test constructs +50 ng of pRLCMV (Promega Corp.)] into HD3 and CEF was performed using TurboFect® (Fermentas UAB). The assays for luciferase activity were performed using the Dual-Luciferase® Reporter Assay System (Promega Corp.), according to the manufacturer’s instructions. The luciferase activity was normalized to Renilla luciferase activity.

Transfection and luciferase activity assays DNA digestion by restriction enzymes, gel electrophoresis isolation of DNA fragments, and cloning were performed according to standard protocols (Maniatis et al. 1982). The

13

Results Transcription of globin genes located in the minor locus The minor locus of Danio rerio globin genes was reported to harbor two pairs of α/β globin genes (αe4–βe3 and βe2– αe5). Transcription of βe3 was observed only at the embryonic stage of development; βe2 and αe5 are active at the larval stage, and αe4 appears to be permanently inactive (Ganis et al. 2012). To verify this transcription pattern, we

Histochem Cell Biol Table 4  Primers used for the preparation of the genetic constructs

βa1 direct

5′ TAATCTCGAGGATGTCCTTAGGTGCTTGC 3′

Xho I

βa1 reverse

5′ TAATCTCGAGTTTCTCAAGTGAGGTTTCTTCAG 3′

Xho I

MRE direct

5′ TAATGGATCCTATATGTCTCTGCAGCAAATTCAA 3′

Bam HI

MRE reverse

5′ TAATGGATCCTTTGAGTCGATGCGTTGT 3′

Bam HI

RE1 direct

5′ TAATGGATCCCAATAGGTTAGTCAGGTA 3′

Bam HI

RE1 reverse

5′ TAATGGATCCTTCTAATGTACAAGTTCT 3′

Bam HI

RE1Δ direct

5′ TAATGGATCCCGATATTACAATCATACG 3′

Bam HI

RE1Δ reverse

5′ TAATGGATCCTATCGTCTGCTTTATCTA 3′

Bam HI

RE2 direct

5′ TAATGGATCCATCATTAATTCAGCACTG 3′

Bam HI

RE2 reverse

5′ ATTAGGATCCCTACACAACACAGCAATC 3′

Bam HI

RE(1 + 2) direct

5′ TAATGGATCCACAGCACATTAATTAGGA 3′

Bam HI

RE(1 + 2) reverse

5′ ATTAGGATCCGCTAACGGATATACATTT 3′

Bam HI

βe3 direct

5′ TAATGCTAGCGGACAGAGCTTTAAATAG 3′

Nhe I

βe3 reverse

5′ ATTAGCTAGCATCATGCAGTCAGAACAG 3′

Nhe I

βe2/αe5 direct

5′ ATATGCTAGCCTGTCAAGAACAAGAATC 3′

Nhe I

βe2/αe5 reverse

5′ TAATGCTAGCCAGACTGGGCACCTACAT 3′

Nhe I

isolated RNA from a pool of 10 Danio rerio collected at various time intervals after fertilization (12–74 days postfertilization [dpf]) as well as from adult fish. The relative transcription levels of the four globin genes present in the minor locus were then determined using RT-PCR. The results (Fig. 1) matched those reported by Ganis team with the exception that low-level transcription of αe4 (approximately 2 % of the level of αe5 transcription) was detected. In contrast to the transcripts of αe5, the transcripts of αe4 were also detected in adult fish. We should also mention that test-amplicons located fully within a predicted exon of this gene produced a much higher signal than amplicons covering exon/intron junctions. Thus, the αe4 transcript appears to be spliced. The αe5 and βe2 genes were transcribed at approximately the same high level from 18 to 26 dpf (Fig. 1). In contrast, transcription of βe3 was below detection level at 12 dpf and subsequently. This result agrees with the previous observation that this gene is expressed only at the very early embryonic stage (up to 3 dpf) (Ganis et al. 2012). Identification of an erythroid‑specific enhancer located upstream to the minor globin gene locus of Danio rerio Eukaryotic enhancers, even those controlling the same genes in different organisms, usually do not possess prominent sequence identity. However, enhancers can be recognized due to the presence of clustered binding sites for transcription factors and certain chromatin signatures. The enhancers controlling erythroid genes usually harbor binding sites for erythroid-specific transcription factors, including GATA-1/2, NF-E2, and EKLF (Flint et al. 2001), and co-localize with erythroid-specific DNase I-hypersensitive sites (Flint et al. 2001). In a previous study, Ganis et al. (Ganis et al. 2012)

suggested that a regulatory element of the minor globin locus of Danio rerio may be located upstream to the gene cluster where a strong GATA1 deposition site co-localizes with a DNase I-hypersensitive site (Fig. 2). The whole αe4 gene is also located within a DNase I-hypersensitive region, in contrast to the βe2, βe3, and αe5 genes. Thus, there is a possibility that a regulatory element is located also within this gene (Ganis et al. 2012). To acquire more information about the possible location of enhancer(s) within these two areas of interest, we have studied the association of these elements with various chromatin marks, namely H3K4me1 and H3K4me3, because the concomitant presence of these marks is typical for active enhancers (Pekowska et al. 2011) and H3K27Ac because in humans this modification also constitutes a part of an active enhancer chromatin signature (RadaIglesias et al. 2011). In addition, the association of the same elements with GATA1 transcription factor in 18 dpf zebrafish red blood cells has been studied as the previously published analysis (Ganis et al. 2012) was performed on the red blood cells of adult zebrafish. Based on the distribution of epigenetic marks, we have focused attention on the two areas that could harbor enhancers (Fig. 2). The first one is located at a distance of 1.7 Kb upstream from the αe4 gene. This area harbors an erythroid-specific DNaseI-hypersensitive site and GATA1 deposition site and was previously considered to harbor a potential enhancer (Ganis et al. 2012). This area possesses high levels of H3K4me3 and H3K27Ac but only a moderate level of H3K4me1. The second area roughly co-localizes with the αe4 gene and is marked by the simultaneous presence of DNase I-hypersensitive site(s), H3K27Ac, H3K4me3, and H3K4me1 (at a moderate level). To check the selected regions for the presence of erythroid-specific enhancers, we performed transient

13



Histochem Cell Biol

A

αe4

βe3

βe2

1 kb

αe5

B

Rel.un.

Exons 30

ae5

0,12

be2

0,10

be3

0,08

be3

0,06

ae4

ae4

20

Rel.un.

40

0,02

C 0,6

ae4

0,4 0,3 0,2

0,06

be3

Introns

ae4

0,04 0,02

0,1 0,0

0,08

be3

Exons

12 16 18 20 22 24 26 44 60 74 adult

dpf

Rel.un.

Rel.un.

0,00

12 16 18 20 22 24 26 44 60 74 adult

dpf

0,5

be2

0,04

10 0

ae5

Introns

12 16 18 20 22 24 26 44 60 74 adult

dpf

0,00

12 16 18 20 22 24 26 44 60 74 adult

dpf

Fig. 1  Expression profiles of the globin genes of the minor locus throughout development. a A scheme demonstrating the genomic positions of the globin genes in the minor locus. b Relative levels of exonic and intronic sequences of the αe4, βe3, βe2, and αe5-globin genes in RNA isolated from zebrafish at different time intervals after fertilization as indicated below the x-axis. The relative units shown

on the y-axis (rel. un.) are the values normalized to the level of the slc4a1a exonic sequence. Different line colors correspond to the expression profile of the different globin genes from the minor locus. The error bars represent the standard error of the mean (SEM) for two independent experiments. c The same as b, but the scale on the y-axis is adjusted to show transcription of the αe4 gene

transfection of constructs with a reporter gene (firefly luciferase) under control of various α- and β-globin gene promoters. The selected genomic fragments were PCRamplified and inserted downstream to the reporter gene. The cultured erythroid cells of Danio rerio are currently unavailable. Thus, we transfected the prepared constructs into chicken erythroid HD3 cells (Beug et al. 1982a, b). To this end, it should be noted that heterologous cellular models are frequently used to study regulatory elements of Danio rerio genome (Silva and Conceição 2015; Zhu et al. 2012). In a preliminary experiment, we verified the ability of HD3 cells to support transcription of a reporter gene placed under control of the Danio rerio βa1 promoter and the previously characterized (Ganis et al. 2012) MRE enhancer from the Danio rerio major globin gene locus. The results presented in Fig. 3 demonstrate that the βa1 promoter is active in HD3 cells and, furthermore, that activity of this promoter in HD3 cells is stimulated by the

Danio rerio MRE enhancer. We concluded that heterologous chicken erythroid cells can be used to identify Danio rerio enhancers. Next, we tested the ability of genomic fragments harboring the selected regions 1 and 2 to stimulate transcription directed by the βa1 promoter. The schematics showing the constructs and the results of luciferase assays are presented in Fig. 3. It is evident that among all the genomic segments tested, only the segment harboring region 1 possessed a weak enhancer activity. Interestingly, this activity was suppressed when region 1 was cloned within a longer fragment also bearing region 2 (Fig. 3, construct βa1 + RE(1 + 2)). By contrast, a 500-bp fragment bearing the central part of the region 1 demonstrated increased enhancer activity (Fig. 3, construct βa1 + REΔ). Region 2 by itself did not possess any enhancer activity. In the next set of experiments, we cloned the promoters of globin genes present in the minor locus (αe5, βe2,

13

Histochem Cell Biol

αe4

βe3

αe4

RE1

1 kb

RE2

a

B

b

c

control βa1 βa1+MRE βa1+RE1 βa1+RE1 rev βa1+RE1∆ βa1+RE1∆ rev βa1+RE2 βa1+RE2 rev βa1+RE(1+2) βa1+RE(1+2) rev

GATA I Adult RBC

C

0.4

30

0.3

20

0.2

10

0.1

a b c

0.0

a b c

Neg MRE

GATA 1 Larval RBC

40

Neg MRE

a b c

H3K27Ac Larval RBC

0

Neg MRE

a b c

H3K4me1 Larval RBC 15 12 9 6 3 0

Neg MRE

% Input

60 50 40 30 20 10 0

RE1 RE1∆

1 kb

RE2

B

DNAse I Adult RBC

H3K4me3 Larval RBC

βe3

A

A

Fig. 2  Position prediction of the regulatory elements of the minor globin locus. A A scheme showing the position of the fragments (RE1 and RE2) containing potential regulatory elements for analysis. The black boxes (a–c) represent the regions analyzed in the ChIP experiments by the TaqMan real-time PCR. B The distribution of DNAseI-hypersensitive sites and GATA1 deposition sites in the area harboring potential regulatory elements. The arrows indicate DNase I-hypersensitive sites and strong GATA1 deposition sites according to Ganis et al. (2012). C The distribution of modified histones (H3K4me3, H3K4me1, and H3K27Ac) and GATA1 deposition sites within the area harboring the potential regulatory elements in 18 dpf zebrafish red blood cells. The y-axis indicates the bound fraction percentage of the input fraction. On the x-axis, the letters represent the test fragments as shown in scheme A. MRE indicates the test fragment located within the erythroid-specific enhancer of the major globin locus, the positive control. Het indicates the test fragment located within the heterochromatic region located on the long arm of chromosome 4 (Howe et al. 2013), the negative control

and βe3) and tested the ability of the identified enhancer element to stimulate transcription directed by these promoters. Genes αe5 and βe2 are located close to each other in a head-to-head orientation. One may suppose that the expression of these genes is controlled by a bidirectional promoter. We tested the promoter activity of the whole fragment separating βe2 and αe5 (1880 bp fragment, constructs βe2 and αe5 in Fig. 4b, c represent different orientations of this large fragment in respect to the reporter gene) and shorter fragments located upstream to αe5, βe2, and βe3 (see schematics in Fig. 4a). Although the long and the short tested fragments possessed similar promoter activity (Fig. 4b), only the activity of the minimal promoters of βe2 and αe5 was significantly stimulated by the enhancer under study (Fig. 4c). The activity of the βe3 minimal

5,6 3,4 0,4 0,8 0

4,8 6,6

2,3 2,3 50 100 150 200 250 normalized luciferase activity

8,3

300

Fig. 3  Chicken erythroid HD3 cells can be used to assess expression of Danio rerio erythroid-specific enhancers and promoters. a A scheme showing the position of the fragments (RE1, RE1Δ, and RE2) containing potential regulatory elements for analysis. b The results of transient transfection experiments. The names of the constructs are shown on the left side of the diagram. The diagram shows normalized luciferase activity. The activity observed for the pGL3 control vector (black columns) was arbitrarily considered as ‘100,’ and the other data were normalized accordingly. The numbers near the columns indicate the fold difference between the normalized luciferase activity of the enhancer element and the normalized luciferase activity of the promoter element. MRE represents the enhancer from the Danio rerio major globin gene locus, and βa1 is the globin gene promoter from the major locus. Error bars represent the SEM for three independent experiments

promoter was only slightly stimulated by this enhancer. This correlates well with the low transcriptional level of the βe3 gene (see Fig. 1). Based on the results presented in Figs. 3 and 4a, one may conclude that the enhancer present in the upstream area of the Danio rerio minor globin gene locus is active with respect to promoters of globin genes located in both the major and the minor alpha-/beta-globin gene loci. The activity of luciferase in extracts of cells transfected with constructs harboring this reporter gene is commonly considered to depend linearly on transcription level. Still, this is an indirect method of the transcriptionlevel estimation. To verify that the enhancer identified in the above-described experiments indeed stimulate the transcription of reporter gene, we have estimated by RT-PCR analysis the levels of the reporter genes transcript in HD3 cells transfected by pGL3 basic construct, the construct with βe2Δ promoter and the construct with βe2Δ promoter and the identified enhancer (RE1Δ regulatory element). The results obtained (Fig. 5) demonstrated that inclusion of this enhancer element in the constructs results in a significant increase of the reporter gene transcription level. To assay the tissue specificity of the identified enhancer, the transfection experiments were repeated with chicken primary fibroblasts. Although the αe2 and βe5 promoters

13



Histochem Cell Biol

A

1 kb

αe4

RE1 RE1∆

B

βe3 RE2

βe3∆

βe2/αe5 αe5∆

βe2∆

βe3

βe3∆ βe2

1,6

βe3∆

βe2∆ αe5

1,4

βe2

αe5∆ 0

2 4 6 8 10 12 normalized luciferase activity

αe5∆ αe5∆+RE1∆ erythroid

non-erythroid

βe3

C

without promotor βe3

D

αe5

βe2

0,95

6,4

prom+RE1∆ rev

1,7

1,9

βe2∆

βe2∆ βe2∆+RE1∆

3,1

1,1

prom prom+RE1∆

1,9

6,4

4,2

αe5

1,3 1,3

αe5∆

0 10 20 30 40 0 10 20 30 40 normalized luciferase activity

Fig. 4  Identification of an erythroid-specific enhancer located upstream of the zebrafish minor globin gene locus. a A scheme demonstrating the position of the cloned fragments used for testing the promoter and enhancer activities. b Testing the ability of the globin gene promoters of the minor locus to stimulate luc + transcription. The diagrams show normalized luciferase activity. The activity observed for the pGL3 control vector was arbitrarily considered to be ‘100,’ and the other data were normalized accordingly. The value of the normalized luciferase activity for the plasmid without the promoter (pGL3 basic vector) is presented for comparison. Error bars represent the SEM for three to six independent experiments. c Testing the ability of the regulatory element 1 (RE1) to stimulate transcription directed by the globin gene promoters of the minor locus. The

were active in these cells, their activity was not stimulated by the enhancer under study (Fig. 4d). Consequently, the identified enhancer appears to be erythroid-specific.

0

5

10

15

20

25

Using a transient transfection assay, we have demonstrated the presence of an erythroid-specific enhancer in a short (500 bp) DNA fragment located ~1.7 Kb upstream to the αe4 gene of the Danio rerio minor globin gene locus. The identified enhancer is erythroid-specific and is active with respect to promoters of globin genes located in both the major and the minor loci. Danio rerio is becoming a popular model system for molecular biology. Identification of a

13

30

35

40

normalized luciferase activity

diagrams show the normalized luciferase activity as described above. The numbers near the columns indicate the fold difference between the normalized activity of the reporter gene driven by promoter with enhancer or by promoter only. Error bars represent the SEM for three to six independent experiments. d Testing the ability of the regulatory element 1 (RE1) to enhance the promoter activity of the βe2 and αe5 promoters in erythroid (HD3) and non-erythroid (CEFs) chicken cells. The diagrams show the normalized luciferase activity as described above. The numbers near the columns indicate the fold difference between the normalized activity of the reporter gene driven by promoter with enhancer or by promoter only. Error bars represent the SEM for two independent experiments

22 cycles 24 cycles

26 cycles RT(-)

А B C А B C

А B C А B C

firefly luciferase

Discussion

3,1

2,4

0,72

RT(-) А B

C А B C

Renilla luciferase

Fig. 5  The semi-quantitative RT-PCR analysis of the luciferase mRNA in the transfected HD3 cells. Agarose gel electrophoresis of RT-PCR products representing firefly and Renilla luciferase mRNAs extracted from HD3 cells transfected with the following constructs: a pGL3 vector without any regulatory elements. b pGL3 vector with βe2Δ promoter. c pGL3 vector with βe2Δ promoter and RE1Δ regulatory element. The vector RL-CMV with Renilla luciferase gene controlled by CMV promoter was co-transfected as the transfection efficiency control. The PCR products were analyzed after the 22, 24, 26 cycles of the polymerase chain reaction. The RT(−) control represents the RT reaction without the reverse transcriptase

Histochem Cell Biol

DHS minor locus

MRE major locus 1

100 NF-E2

EKLF

200 bp

GATA1

Fig. 6  Comparison of transcription factor binding site profiles in MRE and in the identified enhancer element. The figures indicate the binding motifs for erythroid transcription factors. NF-E2 binding sites are shown as blue circles, EKLF binding sites are shown as green ovals, and GATA boxes are shown as purple ellipses, respectively. Positions above and below the lines indicate the presence of elements on opposite DNA strands. The alignment was centered on the first NF-E2 binding site

new erythroid-specific enhancer in the genome of Danio rerio may provide new opportunities for further research involving transient transfections and/or creating transgenic zebrafish. Analysis of the DNA sequence of the identified enhancer element demonstrated the presence of clustered recognition sites for erythroid-specific transcription factors NF-E2, EKLF, and GATA1. In this respect, the identified enhancer is similar to the MRE located upstream to the major globin gene locus (Fig. 6). Analysis of distribution of binding sites for erythroid-specific transcription factors in the vicinity of the minor globin gene locus of several bony fishes shows that enhancer identified in the present study may be present in similar position in genomes of other fishes. Thus, in the platyfish (Xiphophorus maculatus) genome—two GATA-1 binding sites, one NF-E2 binding site, and two EKLF binding sites—are clustered in a 250bp genomic region located at a distance of 2.5 kb from the boundary of the minor globin gene locus. Similar cluster of binding sites for erythroid-specific transcription factors is located at a distance of 1.2 Kb from the boundary of the minor globin gene locus in the genome of fugu (Takifugu rubripes). The minor globin gene locus of bony fishes is likely to have appeared as a result of a full genome duplication that occurred 320–400 million years ago [teleost-specific genome duplication, TGD (Opazo et al. 2013)], which provided the genetic material for an extraordinary functional diversity of teleost hemoglobin isoforms, perhaps facilitating adaptation to marine and freshwater environments. Phylogenetic analysis and comparative genomics provided the evidence that a common ancestor of gars (order Lepisoteiformes) and teleosts had multiple alpha- and beta-globin genes, localized in one cluster (Opazo et al. 2013). The major and minor loci of alpha-/

beta-globin genes emerged as a result of TGD in the stem lineage of teleost fishes. These two loci further evolved via local duplications and deletions (Opazo et al. 2013; Yukuto and Mutsumi 2010). Nevertheless, the traces of TGD can be easily found in the Danio rerio genome. Thus, the genes shisa9 and mlk2 are present in the vicinity of both the minor and major globin gene clusters of Danio rerio. Orthologs of these genes can be found in the vicinity of alpha-globin gene clusters in human and chicken (Opazo et al. 2013). At the same time, in Lepisosteidae (gars) that were separated from the main branch of bony fishes before TGD, there is only one cluster of alpha-/ beta-globin genes possessing the same genomic neighbors (Opazo et al. 2013). Although there is no extended genomic synteny upstream of the minor and major alpha-/ beta-globin clusters of Danio rerio, genes rhbdf1a (major locus) and rhbdf1b (minor locus) are paralogs whose origin can be easily explained by TGD (Hardison 2008; Opazo et al. 2013). During subsequent evolution, a part of the upstream region of the minor globin locus that included the gene MPG and a portion of the NPRL3 gene might have been lost. The MRE element located in the intron of the NPRL3 gene possibly survived, but being located in a non-transcribed region between rhbdf1b and the globin gene cluster was partially transformed, giving rise to the enhancer element characterized in the present study. Taking into account this reasoning, it is not surprising that this enhancer is active with respect to the globin gene promoters located in both the minor and major loci. Interestingly, this enhancer significantly stimulates the activity of the βe2 and αe5 promoters but not that of the βe3 promoter. To this end, it may be of interest that in the chicken, the early embryonic genes possess all the necessary regulatory elements in the vicinity of the promoter; thus, that their expression does not depend on establishing communication with any remote enhancer (Ioudinkova et al. 2011; Ulianov et al. 2012). However, in chickens, switching of embryonic globin gene expression is realized via substitution of a population of cells expressing the embryonic globin genes by a population of cells expressing the adult genes. In zebrafish minor globin gene locus, the so-called maturational switching occurs. The change of expression profile occurs in the same circulating erythrocytes. Yet even in this case, the basic principle that promoters of embryonic genes are self-sufficient appears to be valid. Consequently, this mechanism is likely to be evolutionary ancient. An unexpected observation made in this study is that the αe4 gene previously reported to be inactive is, in fact, transcribed, albeit at a low level. Furthermore, the transcript of this gene is likely to be spliced. Elucidation of the functional significance (if any) of this transcript represents an interesting task for future research.

13

Acknowledgments  This work was supported by the Russian Science Foundation (project #14-24-00022). Author contributions  AVN, NVP, ESI, OVI, and SVR conceived and designed the experiments. AVN and APK performed the experiments. AVN, NVP, APK, OVI, and SVR analyzed the data. SVR wrote the paper. Compliance with ethical standards  Conflict of interest  The authors declare that they have no conflict of interest Ethical standards  All applicable international, national, and/ or institutional guidelines for the care and use of animals were followed. All procedures performed in studies involving zebrafish were in accordance with the ethical standards of the institution at which the studies were conducted.

References Beug H, Doederlein G, Freudenstrein C, Graf T (1982a) Erythroblast cell lines transformed by a temperature sensitive mutant of avian erythroblastosis virus. A model system to study erythroid differentiation in vitro. J Cell Physiol 1:195–207. doi:10.1002/ jcp.1041130427 Beug H, Palmieri S, Freudenstein C, Zentgraf H, Graf T (1982b) Hormone-dependent terminal differentiation in vitro of chicken erythroleukemia cells transformed by its mutants of avian erythroblastosis virus. Cell 28:907–919. doi:10.1016/0092-8674(82)90070-8 Brownlie A et al (2003) Characterization of embryonic globin genes of the zebrafish. Dev Biol 255:48–61. doi:10.1016/ S0012-1606(02)00041-6 Chen H, Lowrey CH, Stamatoyannopoulos G (1997) Analysis of enhancer function of the HS-40 core sequence of the human alpha-globin cluster. Nucleic Acids Res 25:2917–2922. doi:10.1093/nar/25.14.2917 Craddock CF, Vyas P, Sharpe JA, Ayyub H, Wood WG, Higgs DR (1995) Contrasting effects of alpha and beta globin regulatory elements on chromatin structure may be related to their different chromosomal environments. EMBO J 14:1718–1726 Dillon N, Sabbatini P (2000) Functional gene expression domains: defining the functional units of eukaryotic gene regulation. BioEssays 22:657–665. doi:10.1002/15211878(200007)22:73.0.CO;2-2 Flint J et al (2001) Comparative genome analysis delimits a chromosomal domain and identifies key regulatory elements in the alpha globin cluster. Hum Mol Genet 10:371–382. doi:10.1093/ hmg/10.4.371 Forrester WC, Epner E, Driscoll MC, Enver T, Brice M, Papayannopoulou T, Groudine M (1990) A deletion of the human b-globin locus activation region causes a major alteration in chromatin structure and replication across the entire b-globin locus. Gene Dev 4:1637–1649. doi:10.1101/gad.4.10.1637 Ganis JJ et al (2012) Zebrafish globin switching occurs in two developmental stages and is controlled by the LCR. Dev Biol 366:185–194. doi:10.1016/j.ydbio.2012.03.021 Grosveld F, van Assandelt GB, Greaves DR, Kollias B (1987) Position-independent, high-level expression of the human b-globin gene in transgenic mice. Cell 51:975–985. doi:10.1016/0092-8674(87)90584-8

13

Histochem Cell Biol Hardison RC (2008) Globin genes on the move. J Biol 7:35. doi:10.1186/jbiol92 Higgs DR, Wood WG, Jarman AP, Sharpe J, Lida J, Pretorius I-M, Ayyub H (1990) A major positive regulatory region located far upstream of the human a-globin gene locus. Gene Dev 4:1588– 1601. doi:10.1101/gad.4.9.1588 Higgs DR et al (2006) How transcriptional and epigenetic programmes are played out on an individual mammalian gene cluster during lineage commitment and differentiation. Biochem Soc Symp 73:11–22. doi:10.1042/bss0730011 Howe K et al (2013) The zebrafish reference genome sequence and its relationship to the human genome. Nature 496:498–503. doi:10.1038/nature12111 Ioudinkova ES, Ulianov SV, Bunina D, Iarovaia OV, Gavrilov AA, Razin SV (2011) The inactivation of the pi gene in chicken erythroblasts of adult lineage is not mediated by packaging of the embryonic part of the alpha-globin gene domain into a repressive heterochromatin-like structure. Epigenetics 6:1481–1488. doi:10.4161/epi.6.12.18215 Jarman AP, Wood WG, Sharpe JA, Gourdon G, Ayyub H, Higgs DR (1991) Characterization of the major regulatory element upstream of the human a-globin gene cluster. Mol Cell Biol 11:4679–4689. doi:10.1128/MCB.11.9.4679 Kimmel CB, Ballard WW, Kimmel SR, Ullmann B, Schilling TF (1995) Stages of embryonic development of the zebrafish. Dev Dyn 203:253–310. doi:10.1002/aja.1002030302 Li Q, Peterson KR, Fang X, Stamatoyannopoulos G (2002) Locus control regions. Blood 100:3077–3086. doi:10.1182/ blood-2002-04-1104 Maniatis T, Fritsch EF, Sambrook J (1982) Molecular cloning: a laboratory manual. Harbor Lab. Press, Cold Spring Opazo JC, Butts GT, Nery MF, Storz JF, Hoffmann FG (2013) Wholegenome duplication and the functional diversification of teleost fish hemoglobins. Mol Biol Evol 30:140–153. doi:10.1093/ molbev/mss212 Pekowska A et al (2011) H3K4 tri-methylation provides an epigenetic signature of active enhancers. EMBO J 30:4198–4210. doi:10.1038/emboj.2011.295 Rada-Iglesias A, Bajpai R, Swigut T, Brugmann SA, Flynn RA, Wysocka J (2011) A unique chromatin signature uncovers early developmental enhancers in humans. Nature 470:279–283. doi:10.1038/nature09692 Razin SV, Farrell CM, Recillas-Targa F (2003) Genomic domains and regulatory elements operating at the domain level. Int Rev Cytol 226:63–125. doi:10.1016/S0074-7696(03)01002-7 Recillas-Targa F, Razin SV (2001) Chromatin domains and regulation of gene expression: familiar and enigmatic clusters of chicken globin genes. Crit Rev Eukaryot Gene Expr 11:227–242. doi:10.1615/CritRevEukarGeneExpr.v11.i1-3.110 Sharpe JA, Chan-Thomas PS, Lida J, Ayyub H, Wood WG, Higgs DR (1992) Analysis of the human alpha globin upstream regulatory element (HS-40) in transgenic mice. EMBO J 11:4565–4572 Silva I, Conceição N (2015) Cloning, characterization and analysis of the 5′ regulatory region of zebrafish xpd gene. Comput Biochem Physiol B Biochem Mol Biol 185:47–53. doi:10.1016/j. cbpb.2015.04.003 Trimborn T, Gribnau J, Grosveld F, Fraser P (1999) Mechanisms of developmental control of transcription in the murine alpha and beta-globin loci. Genes Dev 13:112–124. doi:10.1101/ gad.13.1.112 Ulianov SV, Gavrilov AA, Razin SV (2012) Spatial organization of the chicken beta-globin gene domain in erythroid cells of embryonic and adult lineages. Epigenetics Chromatin 5:16. doi:10.1186/1756-8935-5-16

Histochem Cell Biol Westerfield M (2000) The Zebrafish book: a guide for the laboratory use of zebrafish (Danio rerio), 4th edn. Eugene, Oregon Yukuto S, Mutsumi N (2010) Teleost fish with specific genome duplication as unique models of vertebrate evolution. Environ Biol Fish 88:169–188. doi:10.1007/s10641-010-9628-7

Zhu K, Wang H, Gul Y, Zhao Y, Wang W, Liu S, Wang M (2012) Expression characterization and the promoter activity analysis of zebrafish hdac4. Fish Physiol Biochem 38:585–593. doi:10.1007/s10695-011-9540-x

13