Position-Dependent Variegation of Globin Transgene Expression
Total Page:16
File Type:pdf, Size:1020Kb
Proc. Natl. Acad. Sci. USA Vol. 92, pp. 5371-5375, June 1995 Genetics Position-dependent variegation of globin transgene expression in mice GRAHAM ROBERTSON*, DAVID GARRICK*, WENLIAN WU*, MARGOT KEARNS*, DAVID MARTINt, AND EMMA WHITELAW*f *Department of Biochemistry, University of Sydney, New South Wales 2006, Australia; and tDepartment of Pediatrics, University of Washington School of Medicine and Fred Hutchinson Cancer Research Center, 1124 Columbia Street, Seattle, WA 98104 Communicated by Stuart H. Orkin, The Children's Hospital, Boston, MA, February 9, 1995 (received for review December 19, 1994) ABSTRACT Expression of genes in eukaryotes has com- with development (15, 16) and is not copy number dependent monly been analyzed in a whole tissue, and levels ofexpression (14-17). have been interpreted as the result of equivalent rates of We have generated transgenic mouse lines using globin transcription in every cell. We have produced transgenic promoter-aHS-40 constructs linked to the Escherichia coli mouse lines that express 18-galactosidase under the control of lacZ gene rather than a globin structural gene. Histochemical globin promoters linked to the major tissue-specific regula- staining for 13-galactosidase (13-gal) activity provides a simple tory element of the a-globin locus, which permits the analysis cell-by-cell assay for transgene expression in individual eryth- of transgene expression in individual red blood cells. We find rocytes. We find that the transgene is expressed in a hetero- that expression of the transgene within all mouse lines is cellular pattern. Furthermore, the number of 13-gal-expressing heterocellular. Individual cells either do not express the cells varies greatly between different transgenic lines derived transgene at all or express it at a level characteristic of that from the same construct but is consistent within a line. Our line. The number of 18-galactosidase-expressing cells varies analysis suggests that variation- in total transgene activity greatly between different lines of transgenic mice at any between lines is mainly due to differences in the number of defined stage of development, but within a transgenic line, cells committed to transgene expression rather than different individual mice have strikingly similar numbers ofexpressing levels of promoter activity. It also implicates position effects cells. This suggests that the degree of heterocellular expres- from the site of transgene integration in determining the sion is determined by the site of integration, as is seen in degree of heterocellular expression. position-effect variegation. MATERIALS AND METHODS The possibility that transgenes may be expressed in a mosaic Transgene Constructs. The globin promoter-lacZ-aHS-40 pattern is rarely considered. Traditionally, expression of globin constructs were similar to that used previously (18), except that transgenes is analysed by lysis and extraction of mRNA from the globin promoters were synthesized by PCR amplification. erythroid cells, and measurements are taken to be represen- The 127-bp C promoter was from +6 to - 127, and the C/ y tative of expression levels in all cells. Expression levels deter- hybrid promoter had the same sequence but with bases -83 to mined for different transgenes, or at different developmental -112 replaced by bases -134 to -193 of the 'y promoter, stages, are thus interpreted as the result of an equivalent rate thereby exchanging the regions containing the GATA and of transcription in every cell. However, such differences could CACC motifs of the two promoters; the sequences were also arise from changes in the proportion of cells expressing confirmed by the Sanger method. Prior to microinjection, the gene, which would not be evident when cells were exam- DNA fragments were excised from vector sequences with Not ined as a group. It is commonly assumed that within a given I and Kpn I and purified by agarose gel electrophoresis. tissue, each cell transcribes a transgene, but this has been Transgenic Mouse Production. Transgenic mice were gen- shown not to be true in several studies (1-8). Inactivation of erated by microinjection into the pronuclei of fertilized eggs a gene in varying numbers of cells upon integration near from the outbred P.O. mouse strain. Transgenic progeny were inactivating chromatin has been described in Drosophila (9) identified, and copy number was determined by standard and yeast (10) and termed position-effect variegation. The Southern blot analysis of tail DNA. Hemizygous lines were behavior of transgenes may resemble this phenomenon. established by mating transgenic founders to P.O. mice. The individual globin genes are poorly expressed in trans- 18-gal Activity in Whole Cells. Twelve and one-half-day genic mice when linked only to their promoters. However, embryos were obtained from P.O. females mated with hem- when also linked to sequences associated with four DNase I izygous transgenic males. After bleeding whole embryos with hypersensitive sites located 6-18 kb upstream of the 13-globin intact yolk sacs into PBS, erythrocytes were gently pelleted in structural genes, termed the 13-globin locus control region (13 a microcentrifuge and fixed, washed, and stained with 5-bro- LCR), levels of expression similar to the endogenous mouse mo-4-chloro-3-indolyl ,B-D-galactopyranoside (X-Gal) as de- globin genes are observed. All lines give high levels of expres- scribed (18) for at least 24 h at 37°C. Adult red blood cells sion, and these levels are copy number dependent, indicating obtained by tail bleeding hemizygous transgenic mice 3 months that each copy of the transgene is independently expressed (11, of age or older were similarly stained with X-Gal. 12). A regulatory element located 40 kb upstream of the P-gal Activity in Lysates. Transgene activity was determined 1 using lysates of 12.5-day postcoitum (dpc) blood. Twelve and a-globin locus (aHS-40) is similar to the LCR in that it gives one-half-day embryos obtained from P.O. females mated with high levels of erythroid-specific expression in cell culture (13) hemizygous transgenic males were bled into 200 and transgenic mice (14-16) regardless of integration site; how- individually ,ul ever, expression of an a-globin gene linked to alHS40 declines Abbreviations: 3 LCR, f-globin locus control region; aHS-40, regu- latory element located 40 kb upstream of the a-globin locus; 3-gal, The publication costs of this article were defrayed in part by page charge 13-galactosidase; X-Gal, 5-bromo-4-chloro-3-indolyl 13-D-galactopyra- payment. This article must therefore be hereby marked "advertisement" in noside; dpc, day(s) postcoitum. accordance with 18 U.S.C. §1734 solely to indicate this fact. iTo whom reprint requests should be addressed. 5371 Downloaded by guest on October 1, 2021 5372 Genetics: Robertson et al Proc. NatL Acad ScL USA 92 (1995) of PBS, and the blood cells from positive embryos (determined tween the numbers of expressing cells at the embryonic and by X-Gal staining of 20 ,u of this suspension) were pooled. adult stages, with the highest at 12.5 dpc (27/SA and 29/7B) After counting the pooled cells on a hemocytometer, the cells decreasing to some of the lowest as adults. This variability was were spun for 30 s in a microcentrifuge at 4°C and were not a unique feature of the C/^y hybrid promoter, as it was also resuspended in 100 ,1. of ice-cold 250 mM Tris (pH 7.4). The observed in the majority of mouse lines generated with dif- cells were lysed by four cycles of freeze-thawing, and the ferent globin promoters: a 127-bp C promoter (Fig. lb and cellular debris was removed by centrifugation at 4°C for 10 min Table 1), a 550-bp promoter, and a 570-bp a-globin promoter in a microcentrifuge. One microliter of the resulting lysate was (data not shown). Within a given transgenic line, however, there assayed for ,B-gal activity using the substrate o-nitrophenyl was strong consistency between individual mice (Table 1). ,B-D-galactopyranoside as described (19) with the absorbance A possible explanation for this variation is that the promoter of the samples determined at 413 nm. is in fact active in all blood cells but at different levels in each line, resulting in different starting concentrations of 13-gal RESULTS protein, and that the 13-gal activity decays to an undetectable level in some cells as they circulate. To ensure that the numbers High-level, erythroid-specific expression of the lacZ gene observed were not an underestimate, whole cells were stained under the control of the {-globin promoter and the aHS-40 has continuously for extended periods. A plateau in the number of been reported (18). In this study, transgenic lines generated stained cells is reached in both high and low percentage lines with a C/y hybrid promoter give similar patterns of expression. between 5 and 20 h (Fig. 2a), whereas 3-gal activity in lysates Analysis of erythroid cell lysates revealed 1-gal activity in all from the same cells is constant up to 30 h (Fig. 2b). We lines tested at 12.5 dpc (Table 1), but 13-gal was below conclude that the cells that do not stain after 24 h are truly detectable levels in lysates from adult mice for 9 of 11 lines negative. To ascertain whether erythrocytes might lose 1-gal (data not shown). This suggests that the embryonic/fetal activity with age, embryonic and adult cells were incubated in promoter in the majority of these lines is being developmen- cell culture medium. No significant decline in the number of tally regulated. In the previous study with the 550-bp C staining cells was observed over 5 days (data not shown). Since promoter alone, a group of 5 lines all showed clear develop- the vast majority of erythrocytes at 12.5 dpc were recently mental regulation (18).