Regulation of vimentin gene expression in the ocular lens

DEVELOPMENTAL

BIOLOGY

k-39,56-64

Regulation

(1990)

of Vimentin

Gene Expression

CHRISTINA M. SAX,* FRANCIS X.FARRELL,P “Laboratory

of Molecular TDepartment

in the Ocular Lens

ZENDRA E. ZEHNER,~

ANDJORAM

and Developmental Biology, National Eye Institute, National Institutes of Health, of Biochemisky and Molecular Biophysics and The Massey Cancer Cents, Medical Virginia Commcmwealth University, Richntond, Virginia 23298 Accepted

December

PIATIGORSKY* Bethesda, Maryland College of Virginia/

20892; and

20, 1989

Vimentin expression in the lens is striking due to the reported mesenchymal preference of vimentin and the epithelial origin of the lens. The amount of chicken vimentin mRNA levels determined by Northern blot analysis increased 3-fold from 7 to 14 days of embryonic lens development and then decreased lo-fold at 16 days of development, suggesting that post-transcriptional processes may contribute to the level of cytoplasmic vimentin mRNA during lens development. To analyze the mechanisms governing vimentin gene expression in the lens at the level of transcription, a series of chicken vimentin 5’-flanking region deletions were fused to the bacterial CAT gene and transfected into fibroblasts and lens cultures derived from three species. The -160 to +l sequence conferred equal promoter activity in cultured chicken lens epithelial cells and fibroblasts. The -321 to -160 sequences increased promoter activity in all cultures, but more strongly in fibroblasts than in lens cells. Sequence elements in the region -608 to -321 repressed promoter activity in lens cells and fibroblasts. Promoter activity was partially restored in fibroblasts but not in lens cells by -767 to -608 sequences. Vimentin gene expression in the lens thus appears to be controlled by multiple positiveand negative-acting elements in its 5’-flanking sequence. 0 1990 Academic Press, Inc.

This departure from the typical pattern of vimentin expression is also of potential significance with respect to the selection and regulation of enzyme/crystallins (Piatigorsky and Wistow, 1989). The crystallins comprise approximately 90% of the soluble protein of the lens. Both vimentin and enzyme/crystallins are found in many tissues, yet they are expressed in unexpectedly high concentration in the lens. Therefore, an understanding of vimentin gene regulation in the lens may provide insight into the regulation of the genes for the non-lens-specific crystallins (cuB-crystallins and enzyme/crystallins). Vimentin is encoded by a single copy gene in several species (Zehner and Paterson, 1983a; Quax et al, 1983; Lilienbaum et aZ., 1986). The proximal GC boxes of the chicken gene are specifically bound by Spl-enriched nuclear extracts (Sax et al, 1988). Both an upstream activating element which functions in fibroblasts (Sax et al., 1988) and those sequences which at least in part direct down-regulation during myogenesis (Sax et al, 1989) are localized in the region -321 to -160 of the chicken gene. These observations and those of others (Rittling and Baserga, 1987; Pieper et al, 1987) suggest that expression of the vimentin gene in several species is controlled by multiple c&acting sequence elements. While vimentin protein has been demonstrated in adult lenses (Bagchi et ah, 1985; Maisel and Ellis, 1984; Ramaekers et aZ., 1982), vimentin gene expression during embryonic lens development has not been analyzed

INTRODUCTION

Fiber cell formation in the ocular lens of vertebrates is characterized by a movement of equatorial epithelial cells toward the center of the growing lens, followed by cell elongation and loss of organelles in the terminal stages of differentiation (see Piatigorsky, 1981, for review). Cytoskeletal proteins are of special interest with respect to lens cell differentiation for several reasons. The lens cytoskeleton, composed of actin, tubulin, and vimentin (Bagchi et aZ., 1985; Maisel and Ellis, 1984; Ramaekers et ah, 1982), has been implicated in cell shape changes, such as the marked elongation characterizing fiber cell differentiation. The present study concerns the intermediate filament protein (IFP)’ vimentin. The IFP multigene family is composed of five members: the cytokeratins present in epithelial cells, desmin in muscle cells, glial fibrillary acidic protein in glial cells, the neurofilaments in neurons, and vimentin in cells of mesenchymal origin (see Lazarides, 1980, for review). Despite its mesenchymal connection, vimentin is one of the main cytoskeletal components synthesized in the lens (Bloemendal, 1981), making the lens an atypical epithelial tissue expressing vimentin. ’ Abbreviations used: @-gal, P-galactosidase; CAT, chloramphenicol acetyltransferase; DME, Dulbecco’s modified Eagle’s medium; IFP, intermediate filament protein; ZCLCZ, bacterial P-galactosidase gene; LTR, long terminal repeat; MSV, murine sarcoma virus; RSV, Rous sarcoma virus. 0012-1606/90 Copyright All rights

$3.00

0 1990 by Academic Press, Inc. of reproduction in any form reserved.

56

SAX

ET AL.

Vimentin

yet. We demonstrate here that chicken vimentin mRNA levels increase through 14 days of embryonic development and then decrease at 16 days of development. In addition, we used a series of vimentin promoter/CAT gene fusions to provide evidence for both positive and negative control elements in the 5’-flanking sequence of the chicken vimentin gene operating in both transfected lens cells from several species and chicken fibroblasts. MATERIALS

Northern

AND

METHODS

Analysis

Total cytoplasmic RNA was isolated from the whole lenses of 7-, 9-, 14-, and 16-day chicken embryos as described by Chirgwin et al. (1979). Five micrograms of each sample was size-fractionated on a 6.6% formaldehyde, 1% agarose gel and transferred onto a nitrocellulose filter. Duplicate panels were probed with 1 X lo6 cpm/ml of either of the following nick-translated 32Plabeled fragments: pE8, an 850-bp chicken vimentin cDNA (Zehner and Paterson, 1983a), or pEW1’7, a chicken 61-crystallin cDNA subclone (subclone of pEW7) (Wawrousek et al., 1986). Filters were hybridized in 50% formamide, 3X SSPE, 5X Denhardt’s solution, 0.1% SDS, 100 pg/ml denatured Escherichia coli DNA, and 10 mg/ml poly(A) at 42°C. Following hybridization filters were washed twice at room temperature in 2X SSC, 0.1% SDS for a total of 30 min and twice at room temperature in 0.1X SSC, 0.1% SDS for a total of 30 min. Densitometric Scanning Autoradiograms

of Northern

Blot

Northern blot autoradiogram signal densities were quantitated using a DU-8 UV-Visible Spectrophotometer (Beckman) at 600 nm, with peak areas being computed by an internal microprocessor. The autoradiograms in Fig. 1 were scanned by densitometry and plots of the peak area for each time point are shown. Each of these autoradiogram exposures was choosen because the densitometry scans reflected visual changes in signal intensity, i.e., a shorter exposure of the vimentin Northern blot yields signals which could not be detected by the densitometer, while longer exposures of the 6crystallin Northern blot yielded signals which saturated the X-ray film (not shown). Primary

Cultures

Primary embryonic lens cells were prepared by modifying the procedure of Borras et al. (1988). Briefly, lenses were isolated from 14-day chicken embryos, disrupted once with forceps, incubated in 0.025% trypsin at 37°C for 90 set, drawn up and down twice in a

Expression

in the Lens

57

pipet, incubated an additional 90 set in trypsin, and seeded onto collagen-coated 60-mm dishes (six lenses per dish) in DME, 10% fetal calf serum, and 0.1 mg/ml gentamicin. Patches of lens epithelial cells were grown at 37°C in 10% CO2 for 48 hr, washed extensively, and then used for transfection. Cultures enriched for primary embryonic fibroblasts were established from the breast muscle of 14-day chicken embryos (Chepelinsky et al, 1985). Breast muscle was minced, trypsinized at 37°C for 30 min, filtered through sterile lens-paper, and plated onto non-collagen-coated plates in DME, 10% fetal calf serum, and 0.1 mg/ml gentamicin. The cultures were washed extensively after 24 hr and maintained up to 72 hr before plating for transfection. While these cultures may be a mixed population of fibroblasts and myoblasts, no fusion of individual cells or myotube formation was observed. Lens Cell Lines The rabbit lens epithelial cell line N/N1003A was originally established from the lenses of newborn rabbits by Reddan et al. (1986). The mouse lens epithelial cell line NKR-11 was established from the lenses of 4- to 7-week-old Nakano strain mice by Russell et al. (1977). The mouse lens epithelial cell line aTN4-1, clonally derived from aTN4 cells, was obtained from P. Russell. aTN4 cells (Yamada et al., 1989) were established from the lens of a transgenic mouse (Mahon et ah, 1987) in which the transgene was composed of the SV40 T antigen driven ‘by a lens-specific cYA-crystallin promoter. All three lines were maintained on non-collagen-coated plates in DME, 10% fetal calf serum, and 0.1 mg/ml gentamicin at 37°C and 5% COe. Plasmids Fragments derived from the 5’-flanking region of the chicken vimentin gene were cloned into the multicloning site of p8CAT (Dente et al., 1983; Sax et aL, 1989). The construction of pcV-321 and pcV-160 has previously been described (Sax et al., 1988). The 3’ end of pcV-767 was generated by RsaI digestion at +87 and Bal31 nuclease digestion to remove the initiator ATG of the chicken vimentin gene, while the 5’ end was generated by Hind111 digestion at -767. The 5’ ends of pcV-608 and pcV-568 were generated by BstNI and PstI digestion, respectively, of pcV-767. Control chloramphenicol acetyltransferase (CAT) plasmids paA366,-CAT (Chepelinsky et a& 1985), pRSV-CAT (Gorman et al, 1982a), pSVO-CAT, and pSV2-CAT (Gorman et d, 1982b) have been previously described. Internal control plasmids consisted of either the RSV LTR (pTB1) (Borris et aL, 1988) or the MSV LTR (pMSV-pgal) (Bouvagnet et al, 1987) linked to the bacterial lacx gene.

58

DEVELOPMENTAL

BIOLOGY

VOLUME

139,199O RESULTS

Tranqfection Primary chicken embryonic fibroblasts, rabbit lens cells (N/N1003A), and mouse lens cells ((uTN~-1, NKR-11) were plated at a density of 3 X lo5 tells/60-mm dish 1 day before transfection. All cells were transfected via the calcium phosphate-DNA precipitation method of Graham and Van der Eb (1973). For each dish, 10 pg of a given vimentin promoter-CAT plasmid and 1 pg of an internal control plasmid (pTB1 for primary chicken cultures, pMSV$gal for mammalian cell lines) were employed. Cultures were harvested 48 hr post-transfection and the cells disrupted by repeated freezing and thawing in 0.25 M Tris, pH 7.8. CAT and /3-Galactosidase Assays CAT activity was assayed in primary chicken cultures via the TLC-method described by Gorman et al. (1982b) and in mammalian cell lines via the two-phase fluor diffusion method described by Neumann et aZ. (1987). @Galactosidase activity was assayed in all cells as described by Nielsen et al. (1983).

7

9

14

Northern

Days

9

7

16

11

of Embryonic

13

16

Development

of Embryonic

Lenses

We first analyzed the pattern of vimentin gene expression during embryonic chicken lens development. Total RNA was isolated from the lenses of 7-, 9-, 14-, and 16-day embryonic chickens and subjected to Northern blot analysis using a chicken vimentin cDNA probe (Zehner and Paterson, 1983a). Vimentin mRNA levels increased steadily from 7 to 14 days of chicken embryonic lens development (Fig. lA), exhibiting an overall 3-fold increase in abundance (as revealed by densitometry scanning). Vimentin mRNA levels then decreased by 8- to lo-fold at 16 days of embryonic development. The same result was obtained in two separate experiments. We believe this decrease was not due to degradation of the 16-day sample since the 28 S and 18 S rRNA bands of all time points exhibited relatively equal intensity and integrity upon ethidium bromide staining (not shown). At all ages examined the characteristic vimentin mRNA doublet was present. These forms arise from the utilization of multiple polyadenylation sites

28s

7

Blot Analysis

9

16

14

-

17 Days

of Embryonic

Development

FIG. 1. Northern blot analysis of vimentin (A) and al-crystallin (B) mRNA levels. Total cytoplasmic RNA was isolated from the whole lenses of 7-, 9-, 14-, and Is-day chicken embryos, size-fractionated on a 6.6% formaldehyde, 1% agarose gel, and transferred to nitrocellulose. Duplicate panels were probed with 1 X lo6 cpm/ml of nick-translated mP-labeled fragments: ES, a chicken vimentin cDNA (A); EW17, a chicken bl-crystallin cDNA (B). The migration positions of the 28 S and 18 S rRNA species are denoted. These autoradiograms were scanned using a DU-8 spectrophotometer to quantitate the signal densities. A plot of the peak area for each band is shown below the respective blot. For vimentin densitometry scans we have shown the individual intensities of the vimentin doublet (0 for 2030- and 2052-nt species, 0 for the 1804-nt species).

SAX

ET AL.

Vimentin

(Zehner and Paterson, 1983b). The bands of this doublet were present at relatively equal intensity, in agreement with studies on l-week-old chicken lenses (Capetanaki et al, 1983). The ij-crystallin genes are the first crystallin genes expressed during chicken embryonic lens development and are preferentially expressed in the lens (see Wistow and Piatigorsky, 1988, for review). The 61-crystallin cDNA used as a probe here cannot distinguish between 61- and b2-crystallin mRNAs, but 98-99% of d-crystallin mRNA present in the developing chicken lens is derived from the al-crystallin gene (Parker et aL, 1988). In contrast to vimentin, 6-crystallin mRNA levels continued to increase during embryonic lens development and exhibited no decrease at 16 days of development (Fig. lB), in agreement with previous studies (Milstone et al., 1976). Densitometry scanning indicated that a-crystallin mRNA increased twofold (per unit of total lens RNA) from 7 to 16 days of embryonic development. The b-crystallin autoradiogram (Fig. 1B) was exposed onetenth as long as the vimentin autoradiogram (Fig. 1A). Thus, vimentin mRNA is much less abundant than b-crystallin mRNA during this period, as expected. Regulatory Elements Utilized in Primary Fibroblasts and Lens Cultures

Chicken

As a first step in understanding the mechanisms which account for changes in vimentin mRNA levels we have attempted to define those sequences which regulate vimentin gene expression. We have fused a series of DNA fragments flanking the 5’ end of the chicken vimentin gene to the bacterial CAT gene (Fig. 2) and analyzed the expression of the resulting plasmids in primary embryonic chicken lens epithelial cells and fibroblast cultures. Fibroblast cultures were established from the breast muscle of 14-day embryonic chickens

Expresim

59

in the Lens

and thus may be a mixed population of fibroblasts and myoblasts. However, no fusion of individual cells or myotube formation was observed when cells were maintained for up to 6 days in culture. The results of three separate transfections, normalized for transfection efficiency by consideration of the internal control pTB1 (see Methods), are summarized in Table 1. The specific activity (pmoles of [14C]chloramphenicol acetylated per unit of ,&galactosidase activity) represents the level of CAT activity arising from each plasmid controlled for differing transfection efficiencies across cultures by standardization to the amount of P-galactosidase activity. The results showed that pcV-160 contains sequences which confer similar promoter strengths in both cultured chicken lens cells and fibroblasts. Sequences contained within the -321 to -160 region (pcV-321) activated transcription an additional threefold in lens and sevenfold in fibroblasts. The region -568 to -321 (pcV-568) reduced promoter activity from pcV-321 levels to near background levels in lens cells (1.6% of pcV-321 CAT activity). The addition of upstream sequences in pcV-608 and pcV-767 did not significantly alter the level of pcV-568 expression in lens cells. Therefore, sequences contained within -568 to -321 are sufficient to suppress promoter activity in chicken lens cells. While pcV-568 exhibited a similar degree of repression in chicken fibroblasts (2.7% of pcV-321 activity) as in lens cultures, promoter activity was further reduced by sequences contained within -608 and -568 (pcV-608). Therefore, -608 to -568 sequences behave differently with cell type in that this region is capable of further repressing promoter activity in fibroblasts but not in lens cells. In contrast to lens cell experiments, the addition of sequences within the region -767 to -608 (pcV-767) partially restored transcriptional activity in fibroblasts to levels that were 26% of those ob-

- 767 k - 608 I - 56% I

I

pcv-608

1

~A-568

+

pcv-321

+

WV-160

-321 -160

FIG. 2. Chicken vimentin 5’-flanking region and promoter-CAT plasmids. A schematic representation of the chicken vimentin 5’-flanking region indicates the position of a CCAAT box (CAT), five GC boxes (GCl-G&5), regions similar to functional SV40 enhancer elements (B, C), and two regions which share sequence similarity with the hamster vimentin flanking region (16 bp, 18 bp). +l denotes the transcription initiation start site. Shown here are those five fragments fused to the bacterial CAT gene in the expression vector p8CAT. The resulting plasmids are referred to as peV-767, pcV-608, pcV-568, pcV-321, and pcV-160, respectively.

60

DEVELOPMENTAL BIOLOGY TABLE 1 CAT ACTIVITY IN PRIMARY CHICKEN LENS AND FIBROBLAST CULTURES Primary lens cells

Primary fibroblasts”

Specific activityb

Fold”

pcV-767 pcV-608 pcV-568 pcV-321 pcV-160 p8CAT

0.5 0.5 0.4 25 8.1 0.2

+ 0.2 -I- 0.2 + 0.2 _+5.0 + 3.2 + 0.1

2.5 2.5 2.0 125 41 1.0

paA366,-CAT pRSV-CAT pSVO-CAT

1.7 * 0.9 83 f 0.3 0.1 k 0.0

17 830 1.0

Plasmid

Specific activity 229 2.8 24 881 126 3.1

Fold”

t- 36 k 0.4 f 3.7 _t 171 + 13 + 0.5

73 0.9 7.7 284 40 1.0

3.1 + 0.5 2740 + 535 2.0 f 0.4

1.5 1370 1.0

a Primary fibroblasts were prepared from the breast muscle of 14day chicken embryos. These cultures may be a mixed population of fibroblasts and myoblasts, although no fusion of individual cells or myotube formation was observed when cultures were maintained up to 6 days in culture. b Specific activity = pmoles of [‘4C]chloramphenicol acetylated per unit of fl-galactosidase activity. Transfections were performed three times for each primary culture type using two different preparations of plasmid DNA, and CAT activity assayed by the TLC method of Gorman et al. (1982b). The mean CAT activity and standard deviation are shown for each. ’ Fold induction of CAT activity for a promoter-CAT plasmid relative to the promoterless vector. p8CAT is a promoterless vector for pcV-767, pcV-608, pcV-568, pcV-321, and pcV-160, while pSVO-CAT is a promoterless vector for pSV2-CAT and pczA366,-CAT.

VOLUME 139,199O

served for pcV-321. Sequences in the region -767 to -608 therefore behave differently in the regulation of vimentin gene expression in lens cells and fibroblasts. We have also transfected a series of control plasmids to ensure that these primary cultures represent the in situ lens and fibroblast cell types. pRSV-CAT, which is typically expressed in a wide variety of cells (Gorman et al., 1982a), is amply expressed in both cultures. paA366,, a murine cuA-crystallin promoter-CAT gene fusion, has been shown to exhibit lens-specific expression (Chepelinsky et aZ., 1985; Overbeek et al, 1985). Here, paA366,-CAT directed CAT activity l’i’-fold over pSVO-CAT levels in our primary chicken lens cultures, but only 1.5-fold over background in the primary chicken embryonic fibroblasts. We conclude that these primary cultures do in fact represent tissue-specific expression systems. Regu1atcn-p Elements Utilized in Mouse and Rabbit Lens Cell Lines To assess the conservation of regulatory elements across species, we have also transfected the chicken vimentin promoter-CAT plasmids into mouse (NKR-11 and cuTN4-1) and rabbit (N/N1003A) lens epithelial cell lines. The results of multiple transfections of each cell line are shown in Table 2. Chicken vimentin 5’-flanking sequences contained within pcV-160 acted as a promoter in all three cell lines. CAT activity was 6-fold over the promoterless vector p8CAT in transfected NKR-11 cells, 3.3-fold in aTN4-1 cells, and l.&fold in N/N1003A cells (see below for discussion of this low

TABLE 2 CAT ACTIVITY IN MOUSE AND RABBIT LENS CELL LINES Mouse NKR-11 Plasmid pcV-767 pcV-608 pcV-568 pcV-321 pcV-160 p8CAT pSV2-CAT polA366,CAT pSVO-CAT

Specific activity” 407 252 1457 3030 1777 295

f 3~ + -+ f +

Mouse aTN4-1 Foldb

Specific activity”

Rabbit N/N1003A Foldb

120 83 567 554 844 36

1.4 0.8 4.9 10 6.0 1.0

97 + 90 I!z 128 * 432 f 285 f 87?

10 20 3 11 35 24

1.1 1.0 1.5 5.0 3.3 1.0

16900 + 1042 510 + 267 287 k 188

58 1.8 1.0

692 + 278 312 + 13 94 + 10

7.4 3.3 1.0

Specific activity” 135 * 127 + 172 + 380 k 24Ok 137 +

Foldb

12 12 17 89 71 27

0.9 0.9 1.3 2.8 1.8 1.0

2245 + 936 367 f 38 120 + 25

18 3.0 1.0

a Specific activity = fmole of [aH]acetylchloramphenicol produced per minute per unit @-galactosidase activity. NKR-11 and N/N1003A cells were transfected a total of four times, while olTN4-1 cells were transfected twice. Two different preparations of plasmid DNA were used. CAT activity was assayed by the two-phase fluor diffusion method of Neumann et al. (1987). The mean CAT activity and standard deviation are shown for each. b Fold induction of CAT activity for a promoter-CAT plasmid relative to the promoterless vector. p8CAT is a promotorless vector for pcV-767, pcV-608, pcV-568, pcV-321, and pcV-160, while pSVO-CAT is a promoterless vector for pSV2-CAT and p(uA366.-CAT.

SAX

ET AL.

Vimentin Expression

induction value). Sequences contained between -321 and -160 (pcV-321) further activated transcription in all three transfected lines (lo-fold over p8CAT in NKR-11, 5-fold in aTN4-1, and 2.8-fold in N/Nl003A). Addition of -608 to -321 sequences to the promoter repressed transcriptional activity in all three cell lines to levels comparable with promoterless p8CAT levels. For the most part, an element(s) within -568 to -321 (pcV-568) is responsible for this substantial repression of activity except in the case of the mouse cell line NKR-11. Here, complete repression required inclusion of -608 to -568 (pcV-608) sequences. Sequences located in the region -767 to -608 did not significantly restore CAT activity in any of the lens cell lines. Therefore, the vimentin promoter-CAT plasmids behave qualitatively the same in the transfected mammalian lens cells as they did in the homologous chicken lens cells. While the induction of each vimentin promoter-CAT construction over p8CAT in N/Nl003A cells was low, the same pattern of CAT expression was obtained in four separate transfections using two different preparations of plasmid DNAs. Therefore, we believe the trend reflects regulatory events occurring within these lens cells, and the low levels of CAT activity may be due to species divergence of the transcriptional apparatus. We have also transfected paA366,-CAT into these cell lines as a control and for comparison of vimentin and crystallin transcriptional regulation (Table 2). pnA366,-CAT was expressed in both the N/Nl003A and the aTN4-1 cell lines, consistent with previous observations (Reddan et ak, 1986; Nakamura, personal communication). paA366,-CAT appears slightly less active in NKR-11 cells, as might be expected based on the low level of cYA-crystallin protein detected in these cells (Russell et aZ., 1977, and personal communication). pSVZCAT, which is expressed in a wide variety of cells (Gorman et ah, 1982a), was amply expressed in the three lens cell lines. DISCUSSION

Our data suggest that the unexpectedly precipitous decline in vimentin mRNA between 14 and 16 days of embryonic development accounts for the reduction in vimentin protein in differentiating lens cells during development (Ellis et al, 1984). Since vimentin may act as an anchor for cell nuclei (Georgatos and Bloebel, 1987), reduced need for vimentin may be due to the loss of fiber cell nuclei during development (Modak and Perdue, 1970). It is noteworthy that vimentin expression ceases in mammalian erythrocytes when the cell nuclei are extruded (Dellagi et al, 1983), while vimentin persists in the nucleated avian erythrocyte (Woodcock, 1980). Although we cannot be certain at this point

in the Lens

61

whether the decrease in vimentin mRNA levels is due to transcriptional or post-transcriptional events, the dramatic decrease raises the possibility that post-transcriptional processes are involved. Furthermore, preliminary nuclear run-on studies suggest that transcription initiation of the chicken vimentin gene is not significantly different at 14 and 16 days of embryonic development (Sax, unpublished). Indeed, it is possible that post-transcriptional events, such as mRNA stabilization, may play a significant role in recruitment of gene for crystallins in the lens (see Piatigorsky and Wistow, 1989). Although the quantitative aspects are not identical, our data show that the 5’-flanking sequence of the chicken vimentin gene has positive and negative regulatory functions in the lens as in mouse L cells (Sax et aL, 1988) and muscle cells (Sax et a& 1989). In the lens these include positive modulation between -321 and +1 and putative negative modulation between -608 and -321. An interesting difference between lens and fibroblasts exists in the region -767 to -608, which can partially restore promoter activity in chicken fibroblasts but not in lens cells. Moreover, the putative negative-acting sequences, characterized by Farrell et al,2 behave somewhat differently in the different cell types. The -568 to -321 stretch suppressed activity in all the cells examined here. The -608 to -568 sequence provided further repression in chicken fibroblasts and the mouse NKR-11 cell line, in contrast to its behavior in chicken primary lens epithelial cells, mouse aTN4-1, and rabbit N/Nl003A lens cell lines where these sequences did not further suppress transcription. This difference among lens cultures may be due to subtle variations in the transcriptional apparatus, perhaps due in part to changes in culture during the generation of cell lines. Possibly, the factor(s) which represses promoter activity in NKR-11 cells requires the presence of sequences in the -608 to -568 region for efficient binding and repression. In this regard it is interesting to note that a 23-bp sequence is present twice in the chicken vimentin 5’-flanking region: once at -605 to -584 and again at -483 to -461 (17 of 23 bp similarity). Duplication of this element in pcV-608 may yield greater silencing than found in pcV-568. Alternatively, the differentiation state of NKR-11 cells may be important in this regard. In contrast to crTN4-1 and N/Nl003A cells, the NKR-11 cell line expresses very little a-crystallin but rather primarily y-crystallins (Russell et al, 1977), a feature of the more differentiated lens fiber cells (for review see Piatigorsky, 1981). The chicken and hamster vimentin gene 5’-flanking *F. X. Farrell, C. M. Sax, and Z. E. Zehner. A negative element controls vimentin gene expression. Manuscript in preparation.

DEVELOPMENTAL BIOLOGY

62

sequences have a very similar arrangement of promoter elements (Zehner et al, 1987). As expected, therefore, the hamster promoter-CAT plasmid containing 176 bp of upstream sequence exhibits high activity in lens cells (Pieper et a& 1987). Surprisingly, however, 599 bp of hamster promoter sequence did not reduce CAT activity in transfected hamster lens cells (Pieper et al, 1987). By contrast, we showed that plasmids containing the chicken vimentin gene sequence between -321 and -608 have lower CAT activity than that with only 321 bp of 5’4anking sequence when expressed in lens cells and fibroblasts. This difference may be due to cell type, to the absence of upstream negative regulation by the hamster vimentin gene, or to the requirement of more than 599 bp of 5’-flanking sequence of the hamster vimentin gene to repress promoter activity. The hamster sequence is not known further than 200 bp upstream for sequence comparisons. Species divergence has been observed between the regulatory elements of the chicken and the hamster vimentin genes expressed during erythropoiesis (Ngai et aZ., 1987). Although vimentin expression in vivo is typically restricted to cells of mes-

A

Element

VOLUME 139.1990

enchymal origin, most cells grown in culture express vimentin regardless of embryonic origin (Franke et al., 1979; Traub et aL, 1983; Virtanen et ah, 1981). In fact the hamster vimentin 5’-flanking region directs gene expression differently in transgenic mice and cells cultured from that mouse (Pieper et aL, 1989). While some of the regulatory elements we have identified here may be related to growth in culture, the expression differences observed between the primary chicken lens cells and fibroblast cultures as well as between the mammalian cell lines suggest that at least some of these regulatory elements are those that function in the intact ocular lens. We have compared the -767 to -568 region of the chicken vimentin gene to a putative negative-acting region of the chicken 81-crystallin gene. This region, like that of vimentin, functions in both lens and nonlens cells (Borrbs et ak, 1988). Three similar regions, denoted Nl, N2, and N3, are identified and presented in Fig. 3A. Interestingly, these three sequences are either adjacent or overlapping in the vimentin gene, and two are in close proximity in the al-crystallin gene (Fig. 3B). Ad-

Sequence

Nl

vim

N2

N3

GCACAGGA

CGGGAGGC

lllllllI

-466

IlIIlIII

to -451

dl

GCACAGGAACTGGAGGC

- 495 to - 479

vim

TGAGGGGGGCC

-450

to -440

-129

to -119

-465

to -455

-151

to -140

IIIIIIIIIII

dl

TGAGGGGGGCC

vim

CACAGGAC

GGG

llllIIIl dl

III

CACAGGACAGGG

23bp

B

r VIM---

dl ---p

I

I

I

I

I

I

-490

-480

-151

-140

-130

-120

I -110

FIG. 3. Comparison of vimentin and 61-crystallin negative sequences. (A) Comparison of the negative-acting 5’-flanking sequences of the chicken vimentin (-608 to -321) and al-crystallin (-603 to -120) genes indicates the presence of three similar sequence elements (Nl, N2, N3). (B) Schematic diagram indicating the position of Nl (I%!),N2 @mm), and N3 @) in the chicken vimentin (vim) and 61-crystallin (61) genes. Also shown is a 9-bp match with the murine major histocompatibility class I gene negative element (Q (Borrirs et CL&1988). The position of an 18-bp segment (-461 to -444) sharing 72% sequence identity between the chicken and hamster vimentin genes and a 23-bp sequence (-483 to -461) repeated in the chicken vimentin 5’-flanking region are denoted as such.

SAX ET AL.

Vimentin

ditionally present in this area is an 1%bp segment sharing 72% sequence similarity between the chicken and the hamster vimentin genes (Zehner et CCL,1987) and one of the 23-bp repeated elements (see above). The 1% and 23-bp elements exhibit subregions of similarity with each other. This arrangement of sequence elements is reminiscent of the modular motifs that comprise enhancers (Ondek et al., 1988). A modular arrangement of repressor motifs may allow for diversity and versatility of transcriptional regulation in different cell types and may account for the differing degrees of repression observed in the lens and fibroblast cultures employed here. Finally, our results suggest that the vimentin gene is not regulated in the same manner as the lens-specific crystallin genes, namely, al- and crA-crystallin. The vimentin and bl-crystallin genes exhibit different expression patterns during embryonic lens development, and pcV-321 is expressed in lens cells of all the species examined here while the cYA-crystallin plasmid paA366,-CAT is not expressed significantly in the NKR-11 cell line. It will be interesting to compare vimentin gene expression with that of aB-crystallin or the enzyme/crystallins, which are functional in many tissues but preferentially expressed in the lens (see Piatigorsky and Wistow, 1989). We are grateful to Dr. Paul Russell for providing the (r-TN4-1 cell line, Dr. Peggy Zelenka for assistance in nuclear run-on transcription analyses, Drs. Terete Borras and Diane Borst for careful reading of the manuscript, and Ms. Dawn Chicchirichi for assistance in preparing the text. Z.E.Z. is supported by Public Health Service Grant AM33310 and grants from the Muscular Dystrophy Association and the Jeffress Trust. REFERENCES BAGCHI, M., CAPORALE, R. S., WECHTER, R. S., and MAISEL, H. (1985). Vimentin synthesis by ocular lens cells. Exp. Eye Res. 40,385-392. BLOEMENDAL, H. (1981). “Molecular and Cellular Biology of the Eye Lens,” pp. 9’7-113. Wiley, New York. BORRKS,T., PETERSON, C. A., and PIATIGORSKY, J. (1988). Evidence for positive and negative regulation in the promoter of the chicken 61-crystallin gene. Dev. BioL 127,209-219. BOUVAGNET, P. F., STREHLER, E. E., WHITE, G. E., STREHLER-PAGE, M.-A., NADAL-GINARD, B., and MAHDAVI, V. (198’7). Multiple positive and negative 5’ regulatory elements control the cell-type-specific expression of the embryonic skeletal myosin heavy-chain gene. Mol. CelL BioL 7,4377-4389. CAPETANAKI, Y. G., NGAI, J., FLYTZANIS, C. N., and LAZARIDES, E. (1983). Tissue-specific expression of two mRNA species transcribed from a single vimentin gene. Cell 35,411-420. CHEPELINSKY, A. B., KING, C. R., ZELENKA, P. S., and PIATIGORSKY, J. (1985). Lens-specific expression of the chloramphenicol acetyltransferase gene promoted by 5’ flanking sequences of the murine aA-crystallin gene in explanted chicken lens epithelia. Proc. NutL Acad. Sci. USA 82,2334-2338. CHIRGWIN, J. M., PRZYBYLA, A. E., MACDONALD, R. J., and RUTTER, W. J. (1979). Isolation of biologically active ribonueleic acid from sources enriched in ribonuclease. Bioch.emi&y l&5294-5299.

Expression

in the Lens

DELLAGI, K., VAINCHENKER, W., VINCI, G., PAULIN, J. C. (1983). Alteration of vimentin intermediate sion during differentiation of human hemopoietic Biol Organ J. 2,1509-1514. DENTE, L., CESARENI, G., and CORTESE, R. (1983). family of single stranded plasmids. Nucleic

63 D., and BROUET, filament exprescells. Eur. Mol. pEMBL: A new Acids Res. 11,

1645-1655.

ELLIS, M., ALOUSI, S., LAWNICZAK, J., MAISEL, H., and WELSH, M. (1984). Studies on lens vimentin. Exp. Eye Res. 38,195-202. FRANKE, W. W., SCHMID, E., WINTER, S., OSBORN,M., and WEBER, K. (1979). Widespread occurrence of intermediate-sized filaments of the vimentin-type in cultured cells from diverse vertebrates. Exp. Cell Res. 123,25-46. GEORGATOS, S. D., and BLOEBEL, G. (1987). Two distinct attachment sites for vimentin along the plasma membrane and the nuclear envelope in avian erythrocytes: A basis for vectorial assembly of intermediate filaments. J CeU BioL 10.5,105-115. GORMAN, C. M., MERLINO, G. T., WILLINGHAM, M. C., PASTAN, I., and HOWARD, B. H. (1982a). The Rous sarcoma virus long terminal repeat is a strong promoter when introduced into a variety of eukaryotic cells by DNA-mediated transfection. Proc NatL Acad. Sci. USA 79,6777-6781. GORMAN, C. M., MOFFAT, L. F., and HOWARD, B. H. (1982b). Recombinant genomes which express chloramphenicol acetyltransferase in mammalian cells. Mol. CelL BioL 2,1014-1051. GRAHAM, F., and VAN DER EB, A. (1973). A new technique for the assay of infectivity of human adenovirus 5 DNA. Virology 52, 456-467.

LAZARIDES, E. (1980). Intermediate filaments as mechanical integrators of cellular space. Nature (London) 283,249-256. LILIENBAUM, A., LEGAGNEUX, V., PORTIER, M.-M., DELLAGI, K., and PAULIN, D. (1986). Vimentin gene: Expression in human lymphocytes and in Burkitt’s lymphoma cells. Eur. Mol. BioL 0rgun. J. 5, 2809-2814. MAHON, K. A., CHEPELINSKY, A. B., KHILLAN, J. S., OVERBEEK, P. A., PIATIGORSKY, J., and WESTPHAL, H. (1987). Oncogenesis of the lens in transgenic mice. Science 235,1622-1628. MAISEL, H., and ELLIS, M. (1984). Cytoskeleton proteins of the aging human lens. Curr. Eye Res. 3,369-381. MILSTONE, L. M., ZELENKA, P., and PIATIGORSKY, J. (1976). d-Crystallin mRNA in chick lens cells: mRNA accumulates during differential stimulation of 6-crystallin synthesis in cultured cells. Dev. BioL 48,197-204.

MODAK, S. P., and PERDUE, S. W. (1970). Terminal lens cell differentiation. I. Histological and microspectrophotometric analysis of nuclear degeneration. Exp. Cell Res. 59,43-56. NEUMANN, J. R., MORENCY, C. A., and RUSSIAN, K. 0. (1987). A novel rapid assay for chloramphenicol acetyltransferase gene expression. BioTechniqws 5,444-447. NGAI, J., BOND, C., WOLD, B. J., and LAZARIDES, E. (1987). Expression of transfected vimentin genes in differentiating murine erythroleukemia cells reveals divergent &s-acting regulation of avian and mammalian vimentin sequences. MoL Cell. BioL 7,3955-3970. NIELSEN, D. A., CHOU, J., MACKRELL, A. J., CASADABAN, M. J., and STEINER, D. F. (1983). Expression of a preproinsulin-fi-galactosidase gene fusion in mammalian cells. Proc. NatL Acad Sci. USA 80, 5198-5202. ONDEK, B., GLOSS, L., and HERR, W. (1988). The SV40 enhancer contains two distinct levels of organization. Nature (London) 333, 40-45.

OVERBEEK, P. A., CHEPELINSKY, A. B., KHILLAN, J. S., PIATIGORSKY, J., and WESTPHAL, H. (1985). Lens-specific expression and developmental regulation of the bacterial chloramphenicol acetyltransferase gene driven by the murine cYA-crystallin promoter in transgenic mice. Proc. Natl. Acad Sci. USA 82,7815-7819.

64

DEVELOPMENTAL BIOLOGY

PARKER, D. S., WAWROUSEK, E. F., and PIATIGORSKY, J. (1988). Expression of the &crystallin genes in the embryonic chicken lens. Dev. Biol 126,375-381. PIATIGORSKY, J. (1981). Lens differentiation in vertebrates. A review of cellular and molecular features. D$wentiation 19,134-153. PIATIGORSKY, J., and WISTOW, G. J. (1989). Enzyme/crystallins: Gene sharing as an evolutionary strategy. Cell 57,197-199. PIEPER, F. R., SCHAART, G., KRIMPENFORT, P. J., HENDERIK, J. B., MOSHAGE, H. J., VAN DE KEMP, A., RAMAEKERS, F. C., BERNS, A., and BLOEMENDAL, H. (1989). Transgenic expression of the musclespecific intermediate filament protein desmin in nonmuscle cells. J. Cell Biol 108,1009-1024. PIEPER, F. R., SLOBBE, R. L., RAMAEKERS, C. S., CUYPERS,H. T., and BLOEMENDAL, H. (1987). Upstream regions of the hamster desmin and vimentin genes regulate expression during in vitro myogenesis. Eur. Mol. Biol Orgun. J. 6,3611-3618. QUAX, W., EGBERTS, W. V., HENDRIKS, W., QUAX-JEUKEN, Y., and BLOEMENDAL, H. (1983). The structure of the vimentin gene. Cell 35, 215-223. RAEMAEKERS, F. C. S., POELS, L. G., JAP, P. H. K., and BLOEMENDAL, H. (1982). Simultaneous demonstration of microfilaments and intermediate-sized filaments in the lens by double immunofluorescence. &p. Eye Res. 35,363-369. REDDAN, J. R., CHEPELINSKY, A. B., DZIEDZIC, D. C., PIATIGORSKY, J., and GOLDENBERG, E. M. (1986). Retention of lens specificity in long-term cultures of diploid rabbit lens epithelial cells. wrentiation 33,168-174. RITTLING, S. R., and BASERGA, R. (1987). Functional analysis and growth regulation of the human vimentin promoter. Mol. Cell Biol. 7,3908-3915. RUSSELL, P., FUKUI, H. N., TSUNEMATSU, Y., HUANG, F. L., and KINOSHITA, J. H. (1977). Tissue culture of lens epithelial cells from normal and Nakano mice. Invest. Ophthdmol. Visual Sci. 16,243-246. SAX, C. M., FARRELL, F. X., TOBIAN, J. A., and ZEHNER, Z. E. (1988).

VOLUME 139.1990

Multiple elements are required for expression of an intermediate filament gene. Nucleic Acids Res. 16, 8057-8076. SAX, C. M., FARRELL, F. X., and ZEHNER, Z. E. (1989). Down-regulation of vimentin gene expression during myogenesis is controlled by a 5’-flanking sequence. Gene 78,235-242. TRAUB, U. E., NELSON, W. J., and TRAUB, P. (1983). Polyacrylamide gel electrophoretic screening of mammalian cells cultured in vitro for the presence of the intermediate filament protein vimentin. J. Cell Sci. 62.129-147. VIRTANEN, I., LEHTO, V.-P., LEHTONEN, E., VARTIO, T., STENMAN, S., KURKI, P., WAGER, O., SMALL, J. V., DAHL, D., and BADLEY, R. A. (1981). Expression of intermediate filaments in cultured cells. J. Cell Sci. 50.45-63. WAWROUSEK, E. F., NICKERSON, J. M., and PIATIGORSKY, J. (1986). Two &crystallin polypeptides are derived from a cloned al-crystallin cDNA. FEBS Lett. 205,235~240. WISTOW, G. J., and PIATIGORSKY, J. (1988). Lens crystallins: The evolution and expression of proteins for a highly specialized tissue. Anna Rev. Biochem. 57,479-504. WOODCOCK, C. L. F. (1980). Nucleus-associated intermediate filaments from chicken erythrocytes. J. Cell Biol. 85,881-889. YAMADA, T., NAKAMURA, T., WESTPHAL, H., RUSSELL, P. (1989). Synthesis of cu-crystallin by a cell line derived from the lens of a transgenie animal. Cum: Eye Res., in press. ZEHNER, Z. E., LI, Y., ROE, B. A., PATERSON, B. M., and SAX, C. M. (1987). The chicken vimentin gene. Nucleotide sequence, regulatory elements, and comparison to the hamster gene. J. Biol. Chem 262, 8112-8120. ZEHNER, Z. E., and PATERSON, B. M. (1983a). Characterization of the chicken vimentin gene: Single copy gene producing multiple mRNAs. Proc. N&l. Acau! Sci. USA 80.911-915. ZEHNER, Z. E., and PATERSON, B. M. (1983b). Vimentin gene expression during myogenesis: Two functional transcripts from a single copy gene. Nucleic Acids Res. 11, 8317-8332.