- Email: [email protected]
Identification of a functional initiator sequence in the human MDR1 promoter
138
Biochirnica et Biophysica Acta, I 172 (1993) 138-146 (t~ 1993 Elsevier Science Publishers B.V. All rights reserved 0167-4781/93/$06.(1(I
BBAEXP 92474
Identification of a functional initiator sequence in the human MDR1 promoter Marion van Groenigen, Linda J. Valentijn and Frank Baas D~Tmrtment ~,(Neurology, Aeademic Medical Center, Amsterdam (Netherlands') and Di~ision of Molecular Biolo~9~', The Netherlands Cancer Institute, Amsterdam (Netherlands)
(Received 25 September 1992)
Key words: Initiator; MDR1; Promoter; P-glycoprotein; Transcriptional regulation The sequence reqmrements for proper transcriptional initiation of the downstream human multidrug resistance MDR1 (P1) promoter were determined using a transient expression system in HeLa cells. The MDR1 promoter has no TATA box and the transcription start site has a strong homology with the initiator (Inr) sequence identified in the murine terminal deoxynucleotidyltransferase ( T d T ) gene. A deletion analysis showed that sequences from - 6 to + 11 relative to the PI transcription start site were sufficient for proper transcriptional initiation, whereas deletion of sequences downstream of + 11 resulted in a strong reduction of properly initiated transcripts. In this transient assay system, both the MDR1 and TdT initiator require in Hela cells the presence of an upstream activating sequence such as the SV40 enhancer. This is in contrast to the transcription in in vitro systems, in which the initiator sequence is able to direct transcription in the absense of an enhancer. Analysis of mutations in the initiator sequence from - 8 to + 10 showed that the A and T nucleotides at position + 1 and +3, respectively, were essential, whereas other substitutions in this region had little effect on promoter activity.
Introduction
Mammalian cells can acquire resistance to a wide variety of non-related drugs when they are exposed to c h e m o t h e r a p e u t i c agents. This p h e n o m e n o n is called multidrug resistance ( M D R ) (for reviews see Refs. 1, 2 and 3). The molecular basis of one form of M D R , P-glycoprotein mediated M D R , is known. P-glycoproteins are e n e r g y - d e p e n d e n t effiux pumps, e n c o d e d by a small gene family and share a strong homology with bacterial transporter proteins. Transfection experiments have shown that the h u m a n M D R 1 P-glycoprotein gene can confer multidrug resistance [4,5]. Overexpression of P-glycoproteins in M D R cell lines can be due to transcriptional activation or gene amplification. In man, transcriptional regulation of the M D R 1 gene seems to play an important role. In the majority of h u m a n M D R cell lines, overexpression of the M D R 1 gene can not be explained by gene amplification [6,7] and we have shown previously that in the M D R derivatives of the h u m a n squamous lung cancer cell line SW-1573, the M D R 1 gene is transcriptionally
Correspondence to: F. Baas, Department of Neurology, K-214, Academic Medical Center, Meibergdreef 9, 1105 AZ Amsterdam, Netherlands.
activated [8]. In addition, induction of M D R 1 P-glycoprotein m R N A levels by heat shock and sodium arsenite is actinomycin D sensitive, suggesting transcriptional regulation [9]. Therefore, studies of the transcriptional regulation of the h u m a n M D R 1 P-glycoprotein gene are important to obtain insight into the mechanisms underlying P-glycoprotein mediated M D R . Previous studies have identified two transcriptional start sites of the M D R 1 gene in h u m a n M D R cell lines [10]. T h e downstream promoter, P1, which is used in normal tissues and cell lines has been cloned [10]. The upstream promoter, P2, has not been cloned yet. It is located 5' of the transcription initiation site of P1, and is active in M D R cell lines and tissues expressing vast amounts of M D R 1 m R N A . In all cases tested thus far, the majority of the transcripts are derived from P1 [10]. The downstream P1 promoter, which lacks a C A A T or T A T A box has been analyzed in transient expression systems [10-12]. In this system p r o m o t o r activity was only detected when an exogenously added e n h a n c e r was present in the transfected D N A constructs. A tissue-specific enhancer was identified located 10 kb upstream of the P1 promoter, however, this e n h a n c e r can only activate M D R 1 - C A T constructs in a limited set of adrenal and kidney derived cell lines, suggesting that additional sequences must be involved in regulation of M D R I
139 transcription in other tissues with high levels of MDR1 expression, like liver and colon [11]. To analyze the sequence requirements for basal promoter function, Cornwell [13] has used an in vitro transcription system. In this system, P1 promoter activity was not enhancer dependent, and deletion of sequences downstream of the initiation site strongly reduced basal transcription levels of the MDRI promoter [12]. To characterize the c/s-acting sequences involved in basal transcriptional regulation of the MDR1 promoter in more detail we have used a HeLa transient expression system. We show that only 17 nucleotides spanning the transcription initiation site ( - 6 - + 11) are sufficient for transcriptional initiation. This sequence shares a strong homology with the TdT initiator (lnr) sequence [14,15]. lnr elements are located at the transcription initiation site and can direct transcription from a RNA polymerase II promoter in the absence of a T A T A box. Using single and double base mutations of the initiation site we show that only two nucleotides of the Inr sequence are essential for initiator function, the A at position +1 and T at +3. Alteration of these nucleotides reduces transcription at least 20-fold, whereas single or double base mutations of other nucleotides of the initiator have only small effects on promoter strength. Materials and Methods
Plasmid constructions A 6 kb EcoRI fragment containing the MDR1 promoter region was isolated from a cosmid clone (hmdrl.1, F.B. and D. Haber unpublished results). This cosmid clone was isolated from a human genomic cosmid library by screening with the pmdrP2 probe [10]. Using the polymerase chain reaction (PCR) a NcoI site was introduced at the 3' end of exon 1. This site was used to fuse the MDR1 promoter to the human/3-globin gene at the A T G translation initiation codon in exon 1. For nomenclature of the clones the following abbreviations were used: m = M D R promoter;/3 =/3-globin gene; e = SV40 enhancer. Restriction sites were used to make the first series of 5' deletions (Pst I: -428, XbaI: - 198, XhoI: - 135, SacI: + 1, see Fig. 1). Exonuclease Bal31 was used for the subsequent 5' deletions. The 3' deletion A + 1 1 - + 139 was made by replacing the XhoI ( - 1 3 5 ) - N c o i (+ 139) fragment containing the MDR1 promoter with a synthetic DNA fragment containing the - 9 - + 11 MDR1 sequence (top strand: 5' T C G A C C T G A G C T C A T T C G A G T A G C 3'). The initiator constructs T d T / I n r and M D R 1 / I n r were made by replacing the PstI ( - 4 2 8 ) NcoI ( + 1 3 9 ) fragment of construct em/3 with a PCR fragment containing the TdT or MDR1 initiator and the first exon of MDR1 (Figs. 1 and 2). The PCR
m~
~
X m~e
om~
~
~-~ ~
/
.
.
~
~
~
r r r ~
~ ~
-52
~
-33 -15 -14 -13 ol
~
Sp6-m/~
.
N
~ -198 -135 -71 -61
.
N
.....
? , ~" ~ I, = i" ~
~/~r~///~-I
sp6 ao11-o139
_
_ _ ~ =/~\ oll
sp6-~o11-o139 . . . . .
~
~
'
o139
~ ;1~1 oll
o139
sp6
Fig. 1. MDR]-t-globin fusion genes and RNase protection constructs. MDR1 sequences are indicated with a solid bold line and exon 1 by an open box, t-globin sequences by a thin line and dashed boxes (exons), the SV40-enhancer is stippled, and vector sequences are indicated by an interrupted line. The MDR1 promoter is indicated by a rightward pointing arrow, the sp6 RNA polymerase promoter is indicated with the leftward arrow marked sp6. The top three lines show the structure of the M D R l - p r o m o t e r constructs with and without enhancer (X: XbaI at - 1 9 8 , P: Pstl at - 4 2 8 , N: Ncol at + 139). In mile and e m i the gene is cloned in two different orientations with respect to the SV40 enhancer. The 5' deletion constructs ( - 1 9 8 to + 1) contained a 40 nt spacer polylinker between the SV40 promoter and the MDR1 sequences. All these constructs were analyzed with an in vitro synthesized antisense riboprobe from sp6m t . This probe contains the MDR1 promoter region fused to t-globin cDNA for exons 1 and 2, and was linearized with X h o | at - 135 (see Materials and Methods). For the 3' deletion construct A + 1 1 - + 139 a matching riboprobe was made for RNase protection (sp6-A + 11 - + 139). All constructs were verified by dideoxy sequencing.
fragments were made using primers starting at - 6 flanked by a Pst I site and a primer containing the NcoI site at the MDR1//3-globin fusion. Initiator mutants were made by replacing the Xhol ( - 1 3 5 ) - N c o i ( + 139) MDR1 promoter fragment by P e R fragments made using degenerated oligonucleotides containing the initiator sequence and a primer overlapping the NcoI site at the MDR1//3-globin junction. The initiator primers contained an additional SalI site, which was used to ligate into the XhoI site at - 1 3 5 in the - 1 9 8 (XbaI) deletion mutant (see Fig. 2D). The entire P e R generated fragments and junctions as well as the deletion mutants and junctions were checked by sequencing using a primer at exon 1 of the /3-globin gene and cloned into the ~rvx derivatives ~-SVHN or ~-SVHNE [16]. The latter contains two
140 copies of the SV40 72-basepair repeat adjacent to a 40 nt polylinker (see Ref. 16 for a detailed map of the ~rSVHN and 7rSVHNE vectors).
Cell fines and transfections HeLa, KB-3-1, KB 8-5 were grown in D M E M medium supplemented with 10% fetal calf serum, 2 mM glutamine, penicillin (50 units/ml) and streptomycin (50 ~ g / m l ) . SW1573 cell lines were grown in Ham's-F10, supplemented as above [8]. Cells were maintained in humidified a i r / 5 % C()2 at 37°C. Transfections were done using the calcium phosphate precipitation technique [17,18]. 10 p~g CsCl-purified plasmid DNA was used for transfection. In all experiments 2.5 /~g of a /3-globin gene with 128 nt of upstream sequences in ~-SVHNE [16] was included as a control for transfection efficiency. In the case of the 3' deletion clone A + 1 1 - + 139 an a-globin gene was used as transfection control. Cytoplasmic RNA was isolated 48-56 h after transfection using the NP-40 lysis method [18]. RNase protection RNase protection was carried out according to Zinn et al. [19]. In the case of the initiator mutants, with single base mismatches at the transcription start site only RNase T1 was used (concentration 300 U / m l ) . Hybridizations were done at 45°C for 16 h and RNase digestions at 30°C for 30 rain. Constructs used were: sp6-m/3 (Fig. 1), a 337 XbaI-NcoI MDR1 promoter fragment fused to a/3-globin cDNA from Nco 1 at exon 1 to BamHI at exon 2 cloned in SP72. For mutants 882, 891 and A + 1 1 - + 139 matching antisense constructs were made by replacing the MDR1 sequence from sp6-m/3 with the corresponding MDR1 sequences from the expression constructs (Fig. 2). Protected fragments for properly initiated transcripts of the MDR1 promoter are 448 nt for all constructs except the 3' deletion. In this case properly initiated transcripts will give a protected fragment of 320 nt. Transcripts from the /3-globin reference plasmid can also be detected with the MDR1/~-globin probe used for RNase protection, resulting in a protected fragment of 309 nt. For the TdT and MDR1 initiator constructs, matching antisense probes with the MDR1 and TdT sequences were used resulting in a protected fragment of 138 nt (Fig. 2). All transfection and RNase protection experiments were performed at least twice yielding similar results. Results
Identification of basal transcription elements of the human MDR1 promoter To determine the sequence requirements for MDR1 promoter function, we used a transient transfection system in which expression of the MDR1 promoter
could easily be measured. Promoter activity was assayed by RNase protection, in order to discriminate between properly initiated transcription and transcripts derived from cryptic promoters. The human MDR1 promoter was linked to the human /3-globin gene to distinguish endogenous MDR1 m R N A from transcripts derived from the transfected construct (Fig. 1 and Materials and Methods). The fusion gene was cloned in two plasmid vectors, pSP72 and ~-vx, and tested in the cell lines HeLa, SW-1573, KB 3-1 and the MDR resistant derivatives SW-1573/1R500 and KB 8-5 expressing the endogenous MDR1 gene at high levels [8,10]. In the absence of the SV-40 enhancer, the transfected MDR1/~9globin fusion gene with up to 5 kb of upstream MDR1 promoter sequences was transcriptionally inactive in all cell lines tested (data not shown). The combination of the ~-vx vector and HeLa cells showed the lowest level of background transcription and was used for the subsequent analysis of the MDR1 promoter. Using the first 428 nt of 5' MDR1 promoter sequences, high levels of transcription of the MDR1 promoter was dependent on the presence of an SV40 enhancer in a position dependent way. High levels of properly initiated transcripts were observed only when the SV40 enhancer was located 5' of the initiation site (compare constructs m/3, m/~e and era/3, Figs. 1, 3). A ......
~
~ N
B
~ ~ , ~
~ , o , ~ ~ , ~
N
D - -
-
~
--~ -198
~ -135
- ~ /
N
Fig. 2. Inr-mutants. MDR! sequences are indicated with a solid bold line and exon 1 by an open box, ~-globin sequences by a thin line and dashed boxes (exons), the SV40-enhancer is stippled, and vector sequences are indicated by an interrupted line. The position of the MDR1 and TdT lnr oligonucleotides are indicated by a small open box and a rightward pointing arrow at the transcription start site (see also legends Fig. 1). Constructs with the MDR1 and TdT lnr sequence ( - 6 - + 1 1 ) were made with the SV40 enhancer (A) or without (C). These constructs contain the entire first exon of MDR! ( + 1 1 - + 139) and have no upstream MDR1 sequences and were analyzed using an anti-sense probe starting at the NcoI (N) site (B). The construct used for the lnr single base mutational analysis is shown in D (see also Table liD. In this case a spacer (MDRI: - 198- - 135) was inserted between the SV40 enhancer and the Int. E: antisense probe used for RNase protection of the lnr mutants. This probe was derived from sp6m/3 and only detects spliced transcripts (Fig. 1).
141 5' and 3' deletions were analyzed to map basal promoter elements (Figs. 1, 4). Exonuclease Bal31 was used to prepare a 5' deletion series. All sequences 5' of - 1 3 can be deleted without a large effect on transcriptional activity (Fig. 4). MDRl-promoter activity ranged between 200% and 85% of the activity of the em/3 construct when normalized for the /3-globin transfection control (see Table I), whereas deletion of sequences up to + 1 abolished transcription completely. Deletion of the MDR1 sequences 3' of + 11 (clone A + 11- + 139) resulted in an strong decrease of properly initiated transcripts, while additional upstream starts were activated (Fig. 4).
Mutational analysis of the initiator sequence The deletion analysis shows that sequences from - 1 3 - + 11 surrounding the MDR1 initiation site are o ×
<
Z
~
•
o
~
517
I I lllll ~
fl
~
Inr
--
~-olobln
--
396--
298
--
220
--
Promotoractiui~ ~5'deletionconstructs 5' deletion
Activity a ( % )
-
100
198 - 71
100
- 61
100 60 60 200 160
-52 -33 - 15 -
14 13
85
+ 1
< 5
a Activity was determined by densitometry of autoradiographs using a LKB laser densitometer model 2202. Promoter activity of construct - 1 9 8 was set at 100%. In all cases the signal of the co-transfected /3-globin gene was used to correct for transfection efficiency and R N A recovery. All mutants were analyzed in at least two experiments yielding similar results. Activities are rounded at 10% values. T h e lower limit for detection is 5% of wild type promoter activity.
_
.
I I
TABLEI
sufficient for enhancer dependent transcription. This sequence is homologous to the initiator sequence, which has been shown to be sufficient for proper transcriptional initiation in in vitro and in vivo systems [14,15,20]. A comparison of the sequence of the MDR1 promoter with the sequence of the promoters of the terminal deoxynucleotidyl transferase promoter (TdT), the adenovirus major late promoter (MLP) and the promoter of the Drosophila gypsy retrotransposon is shown in Table II. All these promoters have been shown to contain functional initiator sequences [14,15,20]. To test whether the MDR1 promoter contains a functional initiator, constructs were made with either the MDR1 or TdTInr sequences ( - 6 - + 11) at the P1 promoter of the MDR1/~-globin gene (Fig. 2). The initiator was tested for transcriptional activity with (Fig. 2A) or without (Fig. 2C) the SV40 enhancer. In the presence of an SV40 enhancer, transcripts starting at the proper initiation site were detected (Fig. 5). The protected fragments for MDR1 and TdT Inr map at the same position. We will refer to this position as + 1. This position coincides with the major downstream P1 start [10]. In both MDR1 and TdT Inr constructs, the T A B L E II
Fig. 3. MDR1 promoter activity is enhancer dependent. RNase protection assay of 10 /zg cytoplasmic R N A from transfected HeLa cells, using a MDRl-Cl-globin probe (sp6m/3 linearized with XhoI, Fig. 1). The position of the protected fragments for transcription from the full length probe (fl), MDR1 Inr (Inr) and fl-globin transfection control (fl-globin) are indicated. Molecular weight markers are indicated at the left. sp6-Xho: input probe, tRNA: control hybridization with 10 ~ g tRNA. Lanes m/3, m/3e and e/3m show the signals obtained after transfection of H e L a cells with the corresponding constructs (see Fig. 1). Duplicate lanes show the analysis of two independent transfection experiments.
The MDR1 transcription start site has homology with promoters containing a functional initiator element -13
MDR1 TdT MLP Gypsy Consensus N = A , G, C or T. Y=CorT.
. +11
GAA
GCCTGAGCTC
ATTCGAGTAG
C
CAT
CAGAGCCCTC
ATTCTGGAGA
C
'CGC
GTTCGTCCTC
ACTCTCTTCC
G
TTG
TCTTCTACTC
AGTTCAAATC
T
CTC
ANTY
142 activity of the Inr sequence was enhancer dependent (data not shown). We have introduced single and double base mutations in the MDR1 initiator, to identify the precise sequence requirements for lnr function (Table lII). All constructs were analyzed in the em/3-198 plasmid in which sequences from - 1 9 8 - - 1 3 5 were fused to - 9 (see Fig. 2 and Table III). The sequences ( - 1 9 8 - 135), which were not essential for enhancer-dependent transcription (see Fig. 4) were included to act as spacer between the SV40 enhancer and the transcription initiation site in order to discriminate between properly initiated transcripts and transcripts derived from cryptic promoters. We have modified the RNase protection procedure to map transcription initiation in constructs with mutations around the initiation site. Under standard condi-
< Z
¢0
J I 517
tions single base mismatches are partially cleaved and therefore it is impossible to discriminate between properly initiated and background transcription in case of mismatches at the initiation site. When RNase T1 is used instead of RNase A, it is possible to distinguish between proper and non-specific transcription initiation, even when a probe with several nucleotide mismatches is used (Fig. 6 and legend). All Inr mutants were analyzed using RNase T1 and anti-sense RNA probe derived from construct 891 (see Table III). A representative example of the analysis of the mutants is shown in Fig. 7. Even though two prominent additional bands are present, the use of RNase protection allows discrimination of the properly initiated transcripts from these major bands. The largest band maps at the position of full-length probe and probably reflects transcripts derived from the plasmid vector. The origin of
JI
II
J
J
-inr
396
--
minr ~,-globin 298
--
221
--
154
--
145
--
~
~~
-~a-Olobln
--
Fig. 4. Deletion analysis of the MDR1 promoter. 5' and 3' deletion constructs were transfected to HeLa cells and 10/~g cytoplasmic RNA was analyzed by RNase protection. 5' deletions (lanes - 198- + 1) were analyzed with a riboprobe derived from sp6-m/3 linearized with Xhol and the 3' deletion (lane A + 1 1 - + 139) was analyzed with a riboprobe derived from sp6-a + 1 1 - + 139 (see Fig. 1). All deletion constructs are shown except - 135 which was analyzed in a separate experiment and showed results similar to - 198. The position of the protected fragments for transcription from the MDR1 Inr (inr), /3-globin transfection control (/3-globin) and molecular weight markers are indicated, tRNA: control hybridization with 10/~g tRNA. The intensity of the protected fragment corresponding to properly initiated transcripts (inr) from the 5' deletions was normalized to the fl-globin band by densitometry (see Table I). The 3' internal deletion was assayed with a different anti-sense probe (sp6-d + 11- + 139), completely homologous to the construct used for transfection (Fig. 1). Because in this case properly initiated transcripts of the 3' deletion are about 300 nt, the/3-globin gene could not be used as an internal control. In this case a plasmid expressing the a-globin gene was used as transfection control. All experiments were performed at least two times yielding similar results.
143 < < z
E
I
I
I
E
~
reduce transcription initiation at least 20-fold, whereas other mutations affected transcriptional initiation at most 3-fold. The bases at - 4 and + 5 - + 11 have not been tested individually, but are not conserved between the Inr sequences of TdT, MLP and gypsy (see Table II). In conclusion, these results suggest that the A at +1 and the T at + 3 in the Inr region from - 6 - + 11 are essential for transcription initiation.
E
I
El e
m I ~
--852
Discussion
--fl
The analysis of the MDR1 promoter showed the presence of a functional initiator sequence, which is sufficient for enhancer-dependent transcription in HeLa cells. Sequences between - 4 2 8 and - 6 do not seem to be essential for basal promoter activity (see Figs. 4 and 5). Sequences 3' of + 11, however, largely affect the amount of properly initiated transcripts, suggesting the presence of a regulatory element between +11 and +139. At least one regulatory element (MDRI: + 4 1 - + 53) has been identified in this region [13]. This sequence has a strong homology with the SV40 downstream element ( + 2 1 - + 33). Whether this
|
--
~
~
+~.
--160 -- 153 - - Inr
W
~
--128 TABLE III
~
~1~
lO9
Mutational analysis of the MDRI initiator Clone
Sequence
Activity a
-8
~
D
g
m
Fig. 5. The MDR1 and TdT Inr are functionally equivalent. RNase protection assay of HeLa cells transfected with MDR1 and TdT Inr constructs. The position of the full length probe (fl), transcripts derived from the lnr (inr) and molecular weight markers are indicated. In order to map initiation, matching anti-sense probes (sp6mdr, sp6-TdT, see Fig. 2B) were used. The probes lack the/3-globin sequences of the fusion gene. These probes could be used since HeLa cells do not express endogenous MDR1 mRNA. No signals were obtained with mock transfected cells and cells transfected with constructs without the SV40 enhancer (data not shown). The lanes sp6-mdr and sp6-TdT show the probes for the MDR and TdT Inr constructs. Lanes mdr-tRNA and TdT-tRNA show the signals after control hybridizations of 10 /zg tRNA with the MDR and TdT probes, respectively.
the second band from the top is unknown. This additional band might reflect activation of a cryptic promoter in the spacer sequences used (MDR: - 1 9 8 -135). The single point mutations at + 1 and + 3
938 (wt) 881 882 8811 887 888 894 891 896 90 91 911 ~ 913
92 931 933 94 consensus
+1
+10
C T G A G C T C A T T C G A G T A G C A X C T
T
C
A
T
T
A
C
T A T G T
c
A A
T A G C
100 90 100 90 60 70 60 80 120 50 80 100 40
T A
c T
< 5 50 <5 30
N N N A N T Y
a Activity was determined by densitometry of autoradiographs using a LKB laser densitometer model 2202. Promoter activity of the wild type construct 938 was set at 100%. In all cases the signal of the co-transfected'/3-globin gene was used to correct for transfection efficiency and RNA recovery. All mutants were analyzed in at least two experiments yielding similar results. Activities are rounded at 10% values. Due to background signals, the lower limit for detection is 5% of wild type promoter activity. b This clone contains an additional mutation at + 12 (G --, T).
144 is the only regulatory element 3' of + 11 remains to be determined. C o m p u t e r - a i d e d sequence analysis revealed the presence of consensus binding sites 3' of the lnr sequence for: IE1.2 (nt 62-67, Ref. 21); NF-E1.5 (nt 81-86, Ref. 22); H4TF-1 (nt 94-99, Ref. 23) present in the Transcription Factor database (release 4) of the G C G D N A analysis program. Regulatory elements 3' of the start site of a polymerase II (pol II) p r o m o t e r lacking a T A T A box have been identified in several other genes like the murine dhfr gene [24], the Drosophila ultrabithorax gene [25], Antennapedia gene [26], engrailed gene [27] and gypsy retrotransposon [20]. Gypsy contains a functional Inr element, whereas the others show sequence homology with the lnr element [20]. The similarity in p r o m o t e r structure of the MDR1 p r o m o t e r with these promoters suggests that they are regulated by a specific set of transcription factors. T h e analysis of the MDR1 p r o m o t e r showed that activation of Inr mediated transcription by the SV40 e n h a n c e r was position dependent. This effect could be due to the fact that in the era/3 derived constructs the SV40-enhancer acts as an upstream activating sequence and not as an enhancer. However, competition of a cryptic p r o m o t e r in the ~rvx vector with the initiator of the fusion gene for the SV40-enhancer in the m/3e construct could play a role. Competition of cryptic promoters for SV40 e n h a n c e r is not a general feature of the ~-vx vector, since the/3-globin p r o m o t e r analyzed in ~-vx is responsive to the SV40 e n h a n c e r irrespective of its position [28]. However, the/3-globin p r o m o t e r contains a T A T A box and competition has been detected in o t h e r systems with promoters lacking a T A T A box [29]. Therefore, we have tested whether the inclusion of a T A T A box affected the MDR1 p r o m o t e r activity and its response to the SV40 enhancer. T h e inclusion of a T A T A box located 30 nt upstream of the MDR1 initiator did not increase promoter activity in the constructs m/~, em/3 and m/3e (data not shown). These experiments suggest that the position d e p e n d e n t effect of the SV40 e n h a n c e r on MDR1 p r o m o t e r activity cannot be explained by simple competition of a T A T A less p r o m o t e r with a cryptic p r o m o t e r in the ~-vx vector. To analyze the sequence requirements of a functional initiator in more detail we have m u t a t e d all base positions between - 6 and + 4, with exception of G 4, which has been shown to be not essential for Inr function [15]. T h e data presented in this paper suggest that only two nucleotides, A +~ and T +3 are essential for initiator function, whereas single base mutations of other nucleotides of the MDR1 initiator affect transcription at most 3-fold (Table IlI). Based on a sequence comparison of Inr containing p r o m o t e r s , the d e g e n e r a t e d consensus s e q u e n c e Y A + ] Y T C Y Y Y was p r o p o s e d by Roy et al. [30]. O u r
probe RNase
517--
M A
N A
N T1
~ii!ii
Inr 396
~
298--
~-globin
221 - -
Fig. 6. RNase protection using probes with mismatches at the initiation site. Autoradiogram of an RNase protection experiment using RNase A or RNase TI in combination with a probe containing several mismatches. RNA from HeLa cells transfected with lnr mutant 891 (Table II1) and analyzed with anti-sense probe complementary to 891 (M for matching) and 882 (N for not matching). The positions of the expected fragments and molecular weight markers are indicated. In the case of a matching probe (M) and RNase A several bands are visible in the region of 400-500 nt. One band at the Inr position and several longer bands due to background transcription. When a probe with three bases mismatch in the transcription start region is used (N) all fragments due to background transcription are cleaved at the transcription start site and co-migrate with properly initiated transcripts. This can be circumvented by using RNase T1 in stead of RNase A (lane N, TI). For this test we have used a probe and RNA combination with 3 mismatches in the initiation region which can be cleaved by RNase TI. All other probe Inr mutant combinations used in this study have less mismatches. This experiment shows that under the conditions used, RNase T1 does not cleave small mismatches in RNA-RNA hybrids. We have also tested other probe combinations yielding similar results (data not shown).
study suggest that the pyrimidines at positions ( - 1, + 5 , + 6 , + 7) are not essential. Also y + 2 and C +4 are not strictly required for Inr function, mutations of Y +2 to A and C +4 to T reduce transcription to 50 and 30% of wild type (mutants 931 and 94, Table III). W h e t h e r the double mutant, resembling the Drosophila gypsy lnr element is still functional remains to be tested.
145
~' E
517
¢~
~1
m
o
~
--
396 -ibbin 298
--
221
--
154
--
145
--
Fig. 7. Mutational analysis of the MDR1 initiator. RNase protection of 10/xg cytoplasmic RNA from HeLa cells transfected with lnr mutants (see Table III). All mutants were analyzed with a probe complementary to mutant 891. The positions of the protected fragments for the Inr and /3-globin transfection control are indicated. The fragments longer than Inr are due to background transcription. This figure is a composition of different exposures of the same experiment. For each sample at least two different exposures were scanned yielding similar results. Results of the densitometric analysis are shown in Table 1II. All mutants were analyzed at least two times yielding similar results.
T h e a p p a r e n t lack o f s e q u e n c e c o n s e r v a t i o n of t h e lnr e l e m e n t m i g h t be e x p l a i n e d by the p r e s e n c e of m u l t i p l e Inr b i n d i n g proteins. D i f f e r e n t p r o t e i n s could b i n d to specific v a r i a n t s o f the d e g e n e r a t e d Inr consensus s e q u e n c e . Two d i f f e r e n t Inr b i n d i n g p r o t e i n s YY1 a n d T F I I - I have a l r e a d y b e e n i d e n t i f i e d in H e L a ceils [30,31]. F o r o n e o f t h e s e p r o t e i n s , T F I I - I , strong c o o p e r a t i v e b i n d i n g with t h e helix-loop-helix p r o t e i n U S F 43 has b e e n d e m o n s t r a t e d [30]. T h e large effect of the 3' d e l e t i o n (A + 1 1 - + 139) on t r a n s c r i p t i o n initiation, is c o m p a t i b l e with such interactions. In conclusion, the M D R 1 P1 p r o m o t e r activity is d e p e n d e n t on the i n i t i a t o r s e q u e n c e l o c a t e d at the t r a n s c r i p t i o n initiation site a n d on d o w n s t r e a m l o c a t e d s e q u e n c e s ( + 1 1 - + 139). S e q u e n c e s 5' o f - 6 a r e not essential for e n h a n c e r - d e p e n d e n t t r a n s c r i p t i o n in H e L a cells. T h e i d e n t i f i c a t i o n o f t h e s e r e g u l a t o r y e l e m e n t s a n d t h e i r r e s p o n s e to e x o g e n o u s l y a d d e d e n h a n c e r s e q u e n c e s , p r o v i d e s the basis for f u r t h e r r e s e a r c h tow a r d s e l u c i d a t i n g the m e c h a n i s m of t r a n s c r i p t i o n a l activation o f the h u m a n M D R 1 gene.
Acknowledgements
W e t h a n k C. Lincke, J.J.M. Smit, Drs. P.A. Bolhuis, P. Borst, A . H . Schinkel, G. Z a m a n a n d J. Z o m e r d i j k for v a l u a b l e discussions a n d critical c o m m e n t s on this m a n u s c r i p t . This r e s e a r c h was f u n d e d by T h e N e t h e r lands O r g a n i z a t i o n for Scientific R e s e a r c h (F.B.). References
1 Endicott, J.A. and Ling, V. (1989) Annu. Rev. Biochem. 58, 137-171. 2 Gottesman, M.M. and Pastan, I. (1988) J. Biol. Chem. 263, 12163-12166. 3 Schinkel, A.H. and Borst, P. (1991) Semin. Cancer Biol. 2, 213-226. 4 Ueda, K., Cardarelli, C., Gottesman, M.M. and Pastan, I. (1987) Proc. Natl. Acad. Sci. USA 84, 3004-3008. 5 Lincke, C.R., Van der Bliek, A.M., Schuurhuis, G.J., Van der Velde-Koerts, T., Smit J.J.M. and Borst P. (1990) Cancer Res. 50, 1779-1785. 6 Van der Bliek, A.M., Baas, F,, Van der Velde-Koerts, T., Biedler, J.L., Meyers, M.B., Ozols, R.F., Hamilton, T.C., Joenje, H. and Borst P. (1988) Cancer Res. 48, 5927-5932.
146 7 Shen, D.W., Fojo, A.T., Chin, J.E., Roninson, I.B., Richert, N., Pastan, I. and Gottesman, M.M. (1986) Science 232, 643-645. 8 Baas, F., Jongsma, A.P., Broxterman, H.J., Arceci, R.J., Housman, D., Scheffer, G.L., Riethorst, A., Van Groenigen, M., Nieuwint, A.W. and Joenje, H. (1990) Cancer Res. 50, 5392-5398. 9 Chin, K.V., Tanaka, S., Darlington, G., Pastan, I. and Gottesman, M.M. (1989) J. Biol. Chem. 265, 221-226. 1(1 Ueda, K., Pastan, I. and Gottesman, M.M. (1987) J. Biol. Chem. 262, 17432-17436. 11 Kohno, K., Sato, S., Uchiumi, T., Takano H., Kato, S. and Kuwano, M. (1990) J. Biol. Chem. 265, 19690-19696. 12 Chin, K.V., Ueda, K., Pastan, I. and Gottesman, M.M. (1992) Science 255, 459-462. 13 Cornwell, M.M. (1990) Cell Growth Diff. 1,607-615. 14 Smale, S.T. and Baltimore, D. (1989) Cell 57, 103-113. 15 Smale, S.T., Schmidt, M.C., Berk, A.J. and Baltimore, D. (1990) Proc. Natl. Acad. Sci. USA 87, 4509-4513. 16 Treisman, R.H. and Maniatis, T. (1985) Nature 315, 72-75. 17 Graham, F.L. and Van der Eb, A.J. (1973) Virology 52, 456-462. 18 Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2rid edition, Cold Spring Harbor University Press, Cold Spring Harbor. 19 Zinn, K., Dimaio, D. and Maniatis, T. (1983) Cell 34, 865-879.
20 Jarrel, K.A. and Meselson, M. (19911 Proc. Natl. Acad. Sci. USA 88, 102-104. 21 Ghazal, P., Lubon, H., Flechstein, B. and Henninghausen. L. (1987) Proc. Natl. Acad. Sci. USA 84, 3658-3662. 22 Wall, L., De Boer, E. and Grosveld, F. (1988) Genes Dev. 2, 1089-1100. 23 Dailey, L., Hanly, S.M., Roeder, R.G. and tteintz, N. (1986) Proc. Natl. Acad. Sci. USA 83, 7241-7245. 24 Means, A.L. and Farnham, P.J. (1990) Mol. Cell. Biol. 10, 653661. 25 Biggin, M.D. and Tjian, R. (1988) Cell 53, 699 711. 26 Perkins, K.K., Dailey, G.M. and Tjian, R. (1988) Genes Dev. 2, 1615-1626. 27 Soeller, W.C., Poole. S.J. and Kornberg, T. (1988) Genes Dev. 2, 68-81. 28 Green, M.R., Treisman, R.H. and Maniatis, T. (1983) Cell 35, 135-148. 29 Wieland, S., Schatt, M.D. and Rusconi, S. (1990) Nucleic Acids Res. 18, 5113-5118. 30 Roy, A.L., Meisterernst, M., Pognonec, P. and Roeder, R.G. (1991) Nature 354, 245-248. 31 Seta, E., Shi, Y. and Shenk, T. (1991) Nature 354, 241-244.
Biochirnica et Biophysica Acta, I 172 (1993) 138-146 (t~ 1993 Elsevier Science Publishers B.V. All rights reserved 0167-4781/93/$06.(1(I
BBAEXP 92474
Identification of a functional initiator sequence in the human MDR1 promoter Marion van Groenigen, Linda J. Valentijn and Frank Baas D~Tmrtment ~,(Neurology, Aeademic Medical Center, Amsterdam (Netherlands') and Di~ision of Molecular Biolo~9~', The Netherlands Cancer Institute, Amsterdam (Netherlands)
(Received 25 September 1992)
Key words: Initiator; MDR1; Promoter; P-glycoprotein; Transcriptional regulation The sequence reqmrements for proper transcriptional initiation of the downstream human multidrug resistance MDR1 (P1) promoter were determined using a transient expression system in HeLa cells. The MDR1 promoter has no TATA box and the transcription start site has a strong homology with the initiator (Inr) sequence identified in the murine terminal deoxynucleotidyltransferase ( T d T ) gene. A deletion analysis showed that sequences from - 6 to + 11 relative to the PI transcription start site were sufficient for proper transcriptional initiation, whereas deletion of sequences downstream of + 11 resulted in a strong reduction of properly initiated transcripts. In this transient assay system, both the MDR1 and TdT initiator require in Hela cells the presence of an upstream activating sequence such as the SV40 enhancer. This is in contrast to the transcription in in vitro systems, in which the initiator sequence is able to direct transcription in the absense of an enhancer. Analysis of mutations in the initiator sequence from - 8 to + 10 showed that the A and T nucleotides at position + 1 and +3, respectively, were essential, whereas other substitutions in this region had little effect on promoter activity.
Introduction
Mammalian cells can acquire resistance to a wide variety of non-related drugs when they are exposed to c h e m o t h e r a p e u t i c agents. This p h e n o m e n o n is called multidrug resistance ( M D R ) (for reviews see Refs. 1, 2 and 3). The molecular basis of one form of M D R , P-glycoprotein mediated M D R , is known. P-glycoproteins are e n e r g y - d e p e n d e n t effiux pumps, e n c o d e d by a small gene family and share a strong homology with bacterial transporter proteins. Transfection experiments have shown that the h u m a n M D R 1 P-glycoprotein gene can confer multidrug resistance [4,5]. Overexpression of P-glycoproteins in M D R cell lines can be due to transcriptional activation or gene amplification. In man, transcriptional regulation of the M D R 1 gene seems to play an important role. In the majority of h u m a n M D R cell lines, overexpression of the M D R 1 gene can not be explained by gene amplification [6,7] and we have shown previously that in the M D R derivatives of the h u m a n squamous lung cancer cell line SW-1573, the M D R 1 gene is transcriptionally
Correspondence to: F. Baas, Department of Neurology, K-214, Academic Medical Center, Meibergdreef 9, 1105 AZ Amsterdam, Netherlands.
activated [8]. In addition, induction of M D R 1 P-glycoprotein m R N A levels by heat shock and sodium arsenite is actinomycin D sensitive, suggesting transcriptional regulation [9]. Therefore, studies of the transcriptional regulation of the h u m a n M D R 1 P-glycoprotein gene are important to obtain insight into the mechanisms underlying P-glycoprotein mediated M D R . Previous studies have identified two transcriptional start sites of the M D R 1 gene in h u m a n M D R cell lines [10]. T h e downstream promoter, P1, which is used in normal tissues and cell lines has been cloned [10]. The upstream promoter, P2, has not been cloned yet. It is located 5' of the transcription initiation site of P1, and is active in M D R cell lines and tissues expressing vast amounts of M D R 1 m R N A . In all cases tested thus far, the majority of the transcripts are derived from P1 [10]. The downstream P1 promoter, which lacks a C A A T or T A T A box has been analyzed in transient expression systems [10-12]. In this system p r o m o t o r activity was only detected when an exogenously added e n h a n c e r was present in the transfected D N A constructs. A tissue-specific enhancer was identified located 10 kb upstream of the P1 promoter, however, this e n h a n c e r can only activate M D R 1 - C A T constructs in a limited set of adrenal and kidney derived cell lines, suggesting that additional sequences must be involved in regulation of M D R I
139 transcription in other tissues with high levels of MDR1 expression, like liver and colon [11]. To analyze the sequence requirements for basal promoter function, Cornwell [13] has used an in vitro transcription system. In this system, P1 promoter activity was not enhancer dependent, and deletion of sequences downstream of the initiation site strongly reduced basal transcription levels of the MDRI promoter [12]. To characterize the c/s-acting sequences involved in basal transcriptional regulation of the MDR1 promoter in more detail we have used a HeLa transient expression system. We show that only 17 nucleotides spanning the transcription initiation site ( - 6 - + 11) are sufficient for transcriptional initiation. This sequence shares a strong homology with the TdT initiator (lnr) sequence [14,15]. lnr elements are located at the transcription initiation site and can direct transcription from a RNA polymerase II promoter in the absence of a T A T A box. Using single and double base mutations of the initiation site we show that only two nucleotides of the Inr sequence are essential for initiator function, the A at position +1 and T at +3. Alteration of these nucleotides reduces transcription at least 20-fold, whereas single or double base mutations of other nucleotides of the initiator have only small effects on promoter strength. Materials and Methods
Plasmid constructions A 6 kb EcoRI fragment containing the MDR1 promoter region was isolated from a cosmid clone (hmdrl.1, F.B. and D. Haber unpublished results). This cosmid clone was isolated from a human genomic cosmid library by screening with the pmdrP2 probe [10]. Using the polymerase chain reaction (PCR) a NcoI site was introduced at the 3' end of exon 1. This site was used to fuse the MDR1 promoter to the human/3-globin gene at the A T G translation initiation codon in exon 1. For nomenclature of the clones the following abbreviations were used: m = M D R promoter;/3 =/3-globin gene; e = SV40 enhancer. Restriction sites were used to make the first series of 5' deletions (Pst I: -428, XbaI: - 198, XhoI: - 135, SacI: + 1, see Fig. 1). Exonuclease Bal31 was used for the subsequent 5' deletions. The 3' deletion A + 1 1 - + 139 was made by replacing the XhoI ( - 1 3 5 ) - N c o i (+ 139) fragment containing the MDR1 promoter with a synthetic DNA fragment containing the - 9 - + 11 MDR1 sequence (top strand: 5' T C G A C C T G A G C T C A T T C G A G T A G C 3'). The initiator constructs T d T / I n r and M D R 1 / I n r were made by replacing the PstI ( - 4 2 8 ) NcoI ( + 1 3 9 ) fragment of construct em/3 with a PCR fragment containing the TdT or MDR1 initiator and the first exon of MDR1 (Figs. 1 and 2). The PCR
m~
~
X m~e
om~
~
~-~ ~
/
.
.
~
~
~
r r r ~
~ ~
-52
~
-33 -15 -14 -13 ol
~
Sp6-m/~
.
N
~ -198 -135 -71 -61
.
N
.....
? , ~" ~ I, = i" ~
~/~r~///~-I
sp6 ao11-o139
_
_ _ ~ =/~\ oll
sp6-~o11-o139 . . . . .
~
~
'
o139
~ ;1~1 oll
o139
sp6
Fig. 1. MDR]-t-globin fusion genes and RNase protection constructs. MDR1 sequences are indicated with a solid bold line and exon 1 by an open box, t-globin sequences by a thin line and dashed boxes (exons), the SV40-enhancer is stippled, and vector sequences are indicated by an interrupted line. The MDR1 promoter is indicated by a rightward pointing arrow, the sp6 RNA polymerase promoter is indicated with the leftward arrow marked sp6. The top three lines show the structure of the M D R l - p r o m o t e r constructs with and without enhancer (X: XbaI at - 1 9 8 , P: Pstl at - 4 2 8 , N: Ncol at + 139). In mile and e m i the gene is cloned in two different orientations with respect to the SV40 enhancer. The 5' deletion constructs ( - 1 9 8 to + 1) contained a 40 nt spacer polylinker between the SV40 promoter and the MDR1 sequences. All these constructs were analyzed with an in vitro synthesized antisense riboprobe from sp6m t . This probe contains the MDR1 promoter region fused to t-globin cDNA for exons 1 and 2, and was linearized with X h o | at - 135 (see Materials and Methods). For the 3' deletion construct A + 1 1 - + 139 a matching riboprobe was made for RNase protection (sp6-A + 11 - + 139). All constructs were verified by dideoxy sequencing.
fragments were made using primers starting at - 6 flanked by a Pst I site and a primer containing the NcoI site at the MDR1//3-globin fusion. Initiator mutants were made by replacing the Xhol ( - 1 3 5 ) - N c o i ( + 139) MDR1 promoter fragment by P e R fragments made using degenerated oligonucleotides containing the initiator sequence and a primer overlapping the NcoI site at the MDR1//3-globin junction. The initiator primers contained an additional SalI site, which was used to ligate into the XhoI site at - 1 3 5 in the - 1 9 8 (XbaI) deletion mutant (see Fig. 2D). The entire P e R generated fragments and junctions as well as the deletion mutants and junctions were checked by sequencing using a primer at exon 1 of the /3-globin gene and cloned into the ~rvx derivatives ~-SVHN or ~-SVHNE [16]. The latter contains two
140 copies of the SV40 72-basepair repeat adjacent to a 40 nt polylinker (see Ref. 16 for a detailed map of the ~rSVHN and 7rSVHNE vectors).
Cell fines and transfections HeLa, KB-3-1, KB 8-5 were grown in D M E M medium supplemented with 10% fetal calf serum, 2 mM glutamine, penicillin (50 units/ml) and streptomycin (50 ~ g / m l ) . SW1573 cell lines were grown in Ham's-F10, supplemented as above [8]. Cells were maintained in humidified a i r / 5 % C()2 at 37°C. Transfections were done using the calcium phosphate precipitation technique [17,18]. 10 p~g CsCl-purified plasmid DNA was used for transfection. In all experiments 2.5 /~g of a /3-globin gene with 128 nt of upstream sequences in ~-SVHNE [16] was included as a control for transfection efficiency. In the case of the 3' deletion clone A + 1 1 - + 139 an a-globin gene was used as transfection control. Cytoplasmic RNA was isolated 48-56 h after transfection using the NP-40 lysis method [18]. RNase protection RNase protection was carried out according to Zinn et al. [19]. In the case of the initiator mutants, with single base mismatches at the transcription start site only RNase T1 was used (concentration 300 U / m l ) . Hybridizations were done at 45°C for 16 h and RNase digestions at 30°C for 30 rain. Constructs used were: sp6-m/3 (Fig. 1), a 337 XbaI-NcoI MDR1 promoter fragment fused to a/3-globin cDNA from Nco 1 at exon 1 to BamHI at exon 2 cloned in SP72. For mutants 882, 891 and A + 1 1 - + 139 matching antisense constructs were made by replacing the MDR1 sequence from sp6-m/3 with the corresponding MDR1 sequences from the expression constructs (Fig. 2). Protected fragments for properly initiated transcripts of the MDR1 promoter are 448 nt for all constructs except the 3' deletion. In this case properly initiated transcripts will give a protected fragment of 320 nt. Transcripts from the /3-globin reference plasmid can also be detected with the MDR1/~-globin probe used for RNase protection, resulting in a protected fragment of 309 nt. For the TdT and MDR1 initiator constructs, matching antisense probes with the MDR1 and TdT sequences were used resulting in a protected fragment of 138 nt (Fig. 2). All transfection and RNase protection experiments were performed at least twice yielding similar results. Results
Identification of basal transcription elements of the human MDR1 promoter To determine the sequence requirements for MDR1 promoter function, we used a transient transfection system in which expression of the MDR1 promoter
could easily be measured. Promoter activity was assayed by RNase protection, in order to discriminate between properly initiated transcription and transcripts derived from cryptic promoters. The human MDR1 promoter was linked to the human /3-globin gene to distinguish endogenous MDR1 m R N A from transcripts derived from the transfected construct (Fig. 1 and Materials and Methods). The fusion gene was cloned in two plasmid vectors, pSP72 and ~-vx, and tested in the cell lines HeLa, SW-1573, KB 3-1 and the MDR resistant derivatives SW-1573/1R500 and KB 8-5 expressing the endogenous MDR1 gene at high levels [8,10]. In the absence of the SV-40 enhancer, the transfected MDR1/~9globin fusion gene with up to 5 kb of upstream MDR1 promoter sequences was transcriptionally inactive in all cell lines tested (data not shown). The combination of the ~-vx vector and HeLa cells showed the lowest level of background transcription and was used for the subsequent analysis of the MDR1 promoter. Using the first 428 nt of 5' MDR1 promoter sequences, high levels of transcription of the MDR1 promoter was dependent on the presence of an SV40 enhancer in a position dependent way. High levels of properly initiated transcripts were observed only when the SV40 enhancer was located 5' of the initiation site (compare constructs m/3, m/~e and era/3, Figs. 1, 3). A ......
~
~ N
B
~ ~ , ~
~ , o , ~ ~ , ~
N
D - -
-
~
--~ -198
~ -135
- ~ /
N
Fig. 2. Inr-mutants. MDR! sequences are indicated with a solid bold line and exon 1 by an open box, ~-globin sequences by a thin line and dashed boxes (exons), the SV40-enhancer is stippled, and vector sequences are indicated by an interrupted line. The position of the MDR1 and TdT lnr oligonucleotides are indicated by a small open box and a rightward pointing arrow at the transcription start site (see also legends Fig. 1). Constructs with the MDR1 and TdT lnr sequence ( - 6 - + 1 1 ) were made with the SV40 enhancer (A) or without (C). These constructs contain the entire first exon of MDR! ( + 1 1 - + 139) and have no upstream MDR1 sequences and were analyzed using an anti-sense probe starting at the NcoI (N) site (B). The construct used for the lnr single base mutational analysis is shown in D (see also Table liD. In this case a spacer (MDRI: - 198- - 135) was inserted between the SV40 enhancer and the Int. E: antisense probe used for RNase protection of the lnr mutants. This probe was derived from sp6m/3 and only detects spliced transcripts (Fig. 1).
141 5' and 3' deletions were analyzed to map basal promoter elements (Figs. 1, 4). Exonuclease Bal31 was used to prepare a 5' deletion series. All sequences 5' of - 1 3 can be deleted without a large effect on transcriptional activity (Fig. 4). MDRl-promoter activity ranged between 200% and 85% of the activity of the em/3 construct when normalized for the /3-globin transfection control (see Table I), whereas deletion of sequences up to + 1 abolished transcription completely. Deletion of the MDR1 sequences 3' of + 11 (clone A + 11- + 139) resulted in an strong decrease of properly initiated transcripts, while additional upstream starts were activated (Fig. 4).
Mutational analysis of the initiator sequence The deletion analysis shows that sequences from - 1 3 - + 11 surrounding the MDR1 initiation site are o ×
<
Z
~
•
o
~
517
I I lllll ~
fl
~
Inr
--
~-olobln
--
396--
298
--
220
--
Promotoractiui~ ~5'deletionconstructs 5' deletion
Activity a ( % )
-
100
198 - 71
100
- 61
100 60 60 200 160
-52 -33 - 15 -
14 13
85
+ 1
< 5
a Activity was determined by densitometry of autoradiographs using a LKB laser densitometer model 2202. Promoter activity of construct - 1 9 8 was set at 100%. In all cases the signal of the co-transfected /3-globin gene was used to correct for transfection efficiency and R N A recovery. All mutants were analyzed in at least two experiments yielding similar results. Activities are rounded at 10% values. T h e lower limit for detection is 5% of wild type promoter activity.
_
.
I I
TABLEI
sufficient for enhancer dependent transcription. This sequence is homologous to the initiator sequence, which has been shown to be sufficient for proper transcriptional initiation in in vitro and in vivo systems [14,15,20]. A comparison of the sequence of the MDR1 promoter with the sequence of the promoters of the terminal deoxynucleotidyl transferase promoter (TdT), the adenovirus major late promoter (MLP) and the promoter of the Drosophila gypsy retrotransposon is shown in Table II. All these promoters have been shown to contain functional initiator sequences [14,15,20]. To test whether the MDR1 promoter contains a functional initiator, constructs were made with either the MDR1 or TdTInr sequences ( - 6 - + 11) at the P1 promoter of the MDR1/~-globin gene (Fig. 2). The initiator was tested for transcriptional activity with (Fig. 2A) or without (Fig. 2C) the SV40 enhancer. In the presence of an SV40 enhancer, transcripts starting at the proper initiation site were detected (Fig. 5). The protected fragments for MDR1 and TdT Inr map at the same position. We will refer to this position as + 1. This position coincides with the major downstream P1 start [10]. In both MDR1 and TdT Inr constructs, the T A B L E II
Fig. 3. MDR1 promoter activity is enhancer dependent. RNase protection assay of 10 /zg cytoplasmic R N A from transfected HeLa cells, using a MDRl-Cl-globin probe (sp6m/3 linearized with XhoI, Fig. 1). The position of the protected fragments for transcription from the full length probe (fl), MDR1 Inr (Inr) and fl-globin transfection control (fl-globin) are indicated. Molecular weight markers are indicated at the left. sp6-Xho: input probe, tRNA: control hybridization with 10 ~ g tRNA. Lanes m/3, m/3e and e/3m show the signals obtained after transfection of H e L a cells with the corresponding constructs (see Fig. 1). Duplicate lanes show the analysis of two independent transfection experiments.
The MDR1 transcription start site has homology with promoters containing a functional initiator element -13
MDR1 TdT MLP Gypsy Consensus N = A , G, C or T. Y=CorT.
. +11
GAA
GCCTGAGCTC
ATTCGAGTAG
C
CAT
CAGAGCCCTC
ATTCTGGAGA
C
'CGC
GTTCGTCCTC
ACTCTCTTCC
G
TTG
TCTTCTACTC
AGTTCAAATC
T
CTC
ANTY
142 activity of the Inr sequence was enhancer dependent (data not shown). We have introduced single and double base mutations in the MDR1 initiator, to identify the precise sequence requirements for lnr function (Table lII). All constructs were analyzed in the em/3-198 plasmid in which sequences from - 1 9 8 - - 1 3 5 were fused to - 9 (see Fig. 2 and Table III). The sequences ( - 1 9 8 - 135), which were not essential for enhancer-dependent transcription (see Fig. 4) were included to act as spacer between the SV40 enhancer and the transcription initiation site in order to discriminate between properly initiated transcripts and transcripts derived from cryptic promoters. We have modified the RNase protection procedure to map transcription initiation in constructs with mutations around the initiation site. Under standard condi-
< Z
¢0
J I 517
tions single base mismatches are partially cleaved and therefore it is impossible to discriminate between properly initiated and background transcription in case of mismatches at the initiation site. When RNase T1 is used instead of RNase A, it is possible to distinguish between proper and non-specific transcription initiation, even when a probe with several nucleotide mismatches is used (Fig. 6 and legend). All Inr mutants were analyzed using RNase T1 and anti-sense RNA probe derived from construct 891 (see Table III). A representative example of the analysis of the mutants is shown in Fig. 7. Even though two prominent additional bands are present, the use of RNase protection allows discrimination of the properly initiated transcripts from these major bands. The largest band maps at the position of full-length probe and probably reflects transcripts derived from the plasmid vector. The origin of
JI
II
J
J
-inr
396
--
minr ~,-globin 298
--
221
--
154
--
145
--
~
~~
-~a-Olobln
--
Fig. 4. Deletion analysis of the MDR1 promoter. 5' and 3' deletion constructs were transfected to HeLa cells and 10/~g cytoplasmic RNA was analyzed by RNase protection. 5' deletions (lanes - 198- + 1) were analyzed with a riboprobe derived from sp6-m/3 linearized with Xhol and the 3' deletion (lane A + 1 1 - + 139) was analyzed with a riboprobe derived from sp6-a + 1 1 - + 139 (see Fig. 1). All deletion constructs are shown except - 135 which was analyzed in a separate experiment and showed results similar to - 198. The position of the protected fragments for transcription from the MDR1 Inr (inr), /3-globin transfection control (/3-globin) and molecular weight markers are indicated, tRNA: control hybridization with 10/~g tRNA. The intensity of the protected fragment corresponding to properly initiated transcripts (inr) from the 5' deletions was normalized to the fl-globin band by densitometry (see Table I). The 3' internal deletion was assayed with a different anti-sense probe (sp6-d + 11- + 139), completely homologous to the construct used for transfection (Fig. 1). Because in this case properly initiated transcripts of the 3' deletion are about 300 nt, the/3-globin gene could not be used as an internal control. In this case a plasmid expressing the a-globin gene was used as transfection control. All experiments were performed at least two times yielding similar results.
143 < < z
E
I
I
I
E
~
reduce transcription initiation at least 20-fold, whereas other mutations affected transcriptional initiation at most 3-fold. The bases at - 4 and + 5 - + 11 have not been tested individually, but are not conserved between the Inr sequences of TdT, MLP and gypsy (see Table II). In conclusion, these results suggest that the A at +1 and the T at + 3 in the Inr region from - 6 - + 11 are essential for transcription initiation.
E
I
El e
m I ~
--852
Discussion
--fl
The analysis of the MDR1 promoter showed the presence of a functional initiator sequence, which is sufficient for enhancer-dependent transcription in HeLa cells. Sequences between - 4 2 8 and - 6 do not seem to be essential for basal promoter activity (see Figs. 4 and 5). Sequences 3' of + 11, however, largely affect the amount of properly initiated transcripts, suggesting the presence of a regulatory element between +11 and +139. At least one regulatory element (MDRI: + 4 1 - + 53) has been identified in this region [13]. This sequence has a strong homology with the SV40 downstream element ( + 2 1 - + 33). Whether this
|
--
~
~
+~.
--160 -- 153 - - Inr
W
~
--128 TABLE III
~
~1~
lO9
Mutational analysis of the MDRI initiator Clone
Sequence
Activity a
-8
~
D
g
m
Fig. 5. The MDR1 and TdT Inr are functionally equivalent. RNase protection assay of HeLa cells transfected with MDR1 and TdT Inr constructs. The position of the full length probe (fl), transcripts derived from the lnr (inr) and molecular weight markers are indicated. In order to map initiation, matching anti-sense probes (sp6mdr, sp6-TdT, see Fig. 2B) were used. The probes lack the/3-globin sequences of the fusion gene. These probes could be used since HeLa cells do not express endogenous MDR1 mRNA. No signals were obtained with mock transfected cells and cells transfected with constructs without the SV40 enhancer (data not shown). The lanes sp6-mdr and sp6-TdT show the probes for the MDR and TdT Inr constructs. Lanes mdr-tRNA and TdT-tRNA show the signals after control hybridizations of 10 /zg tRNA with the MDR and TdT probes, respectively.
the second band from the top is unknown. This additional band might reflect activation of a cryptic promoter in the spacer sequences used (MDR: - 1 9 8 -135). The single point mutations at + 1 and + 3
938 (wt) 881 882 8811 887 888 894 891 896 90 91 911 ~ 913
92 931 933 94 consensus
+1
+10
C T G A G C T C A T T C G A G T A G C A X C T
T
C
A
T
T
A
C
T A T G T
c
A A
T A G C
100 90 100 90 60 70 60 80 120 50 80 100 40
T A
c T
< 5 50 <5 30
N N N A N T Y
a Activity was determined by densitometry of autoradiographs using a LKB laser densitometer model 2202. Promoter activity of the wild type construct 938 was set at 100%. In all cases the signal of the co-transfected'/3-globin gene was used to correct for transfection efficiency and RNA recovery. All mutants were analyzed in at least two experiments yielding similar results. Activities are rounded at 10% values. Due to background signals, the lower limit for detection is 5% of wild type promoter activity. b This clone contains an additional mutation at + 12 (G --, T).
144 is the only regulatory element 3' of + 11 remains to be determined. C o m p u t e r - a i d e d sequence analysis revealed the presence of consensus binding sites 3' of the lnr sequence for: IE1.2 (nt 62-67, Ref. 21); NF-E1.5 (nt 81-86, Ref. 22); H4TF-1 (nt 94-99, Ref. 23) present in the Transcription Factor database (release 4) of the G C G D N A analysis program. Regulatory elements 3' of the start site of a polymerase II (pol II) p r o m o t e r lacking a T A T A box have been identified in several other genes like the murine dhfr gene [24], the Drosophila ultrabithorax gene [25], Antennapedia gene [26], engrailed gene [27] and gypsy retrotransposon [20]. Gypsy contains a functional Inr element, whereas the others show sequence homology with the lnr element [20]. The similarity in p r o m o t e r structure of the MDR1 p r o m o t e r with these promoters suggests that they are regulated by a specific set of transcription factors. T h e analysis of the MDR1 p r o m o t e r showed that activation of Inr mediated transcription by the SV40 e n h a n c e r was position dependent. This effect could be due to the fact that in the era/3 derived constructs the SV40-enhancer acts as an upstream activating sequence and not as an enhancer. However, competition of a cryptic p r o m o t e r in the ~rvx vector with the initiator of the fusion gene for the SV40-enhancer in the m/3e construct could play a role. Competition of cryptic promoters for SV40 e n h a n c e r is not a general feature of the ~-vx vector, since the/3-globin p r o m o t e r analyzed in ~-vx is responsive to the SV40 e n h a n c e r irrespective of its position [28]. However, the/3-globin p r o m o t e r contains a T A T A box and competition has been detected in o t h e r systems with promoters lacking a T A T A box [29]. Therefore, we have tested whether the inclusion of a T A T A box affected the MDR1 p r o m o t e r activity and its response to the SV40 enhancer. T h e inclusion of a T A T A box located 30 nt upstream of the MDR1 initiator did not increase promoter activity in the constructs m/~, em/3 and m/3e (data not shown). These experiments suggest that the position d e p e n d e n t effect of the SV40 e n h a n c e r on MDR1 p r o m o t e r activity cannot be explained by simple competition of a T A T A less p r o m o t e r with a cryptic p r o m o t e r in the ~-vx vector. To analyze the sequence requirements of a functional initiator in more detail we have m u t a t e d all base positions between - 6 and + 4, with exception of G 4, which has been shown to be not essential for Inr function [15]. T h e data presented in this paper suggest that only two nucleotides, A +~ and T +3 are essential for initiator function, whereas single base mutations of other nucleotides of the MDR1 initiator affect transcription at most 3-fold (Table IlI). Based on a sequence comparison of Inr containing p r o m o t e r s , the d e g e n e r a t e d consensus s e q u e n c e Y A + ] Y T C Y Y Y was p r o p o s e d by Roy et al. [30]. O u r
probe RNase
517--
M A
N A
N T1
~ii!ii
Inr 396
~
298--
~-globin
221 - -
Fig. 6. RNase protection using probes with mismatches at the initiation site. Autoradiogram of an RNase protection experiment using RNase A or RNase TI in combination with a probe containing several mismatches. RNA from HeLa cells transfected with lnr mutant 891 (Table II1) and analyzed with anti-sense probe complementary to 891 (M for matching) and 882 (N for not matching). The positions of the expected fragments and molecular weight markers are indicated. In the case of a matching probe (M) and RNase A several bands are visible in the region of 400-500 nt. One band at the Inr position and several longer bands due to background transcription. When a probe with three bases mismatch in the transcription start region is used (N) all fragments due to background transcription are cleaved at the transcription start site and co-migrate with properly initiated transcripts. This can be circumvented by using RNase T1 in stead of RNase A (lane N, TI). For this test we have used a probe and RNA combination with 3 mismatches in the initiation region which can be cleaved by RNase TI. All other probe Inr mutant combinations used in this study have less mismatches. This experiment shows that under the conditions used, RNase T1 does not cleave small mismatches in RNA-RNA hybrids. We have also tested other probe combinations yielding similar results (data not shown).
study suggest that the pyrimidines at positions ( - 1, + 5 , + 6 , + 7) are not essential. Also y + 2 and C +4 are not strictly required for Inr function, mutations of Y +2 to A and C +4 to T reduce transcription to 50 and 30% of wild type (mutants 931 and 94, Table III). W h e t h e r the double mutant, resembling the Drosophila gypsy lnr element is still functional remains to be tested.
145
~' E
517
¢~
~1
m
o
~
--
396 -ibbin 298
--
221
--
154
--
145
--
Fig. 7. Mutational analysis of the MDR1 initiator. RNase protection of 10/xg cytoplasmic RNA from HeLa cells transfected with lnr mutants (see Table III). All mutants were analyzed with a probe complementary to mutant 891. The positions of the protected fragments for the Inr and /3-globin transfection control are indicated. The fragments longer than Inr are due to background transcription. This figure is a composition of different exposures of the same experiment. For each sample at least two different exposures were scanned yielding similar results. Results of the densitometric analysis are shown in Table 1II. All mutants were analyzed at least two times yielding similar results.
T h e a p p a r e n t lack o f s e q u e n c e c o n s e r v a t i o n of t h e lnr e l e m e n t m i g h t be e x p l a i n e d by the p r e s e n c e of m u l t i p l e Inr b i n d i n g proteins. D i f f e r e n t p r o t e i n s could b i n d to specific v a r i a n t s o f the d e g e n e r a t e d Inr consensus s e q u e n c e . Two d i f f e r e n t Inr b i n d i n g p r o t e i n s YY1 a n d T F I I - I have a l r e a d y b e e n i d e n t i f i e d in H e L a ceils [30,31]. F o r o n e o f t h e s e p r o t e i n s , T F I I - I , strong c o o p e r a t i v e b i n d i n g with t h e helix-loop-helix p r o t e i n U S F 43 has b e e n d e m o n s t r a t e d [30]. T h e large effect of the 3' d e l e t i o n (A + 1 1 - + 139) on t r a n s c r i p t i o n initiation, is c o m p a t i b l e with such interactions. In conclusion, the M D R 1 P1 p r o m o t e r activity is d e p e n d e n t on the i n i t i a t o r s e q u e n c e l o c a t e d at the t r a n s c r i p t i o n initiation site a n d on d o w n s t r e a m l o c a t e d s e q u e n c e s ( + 1 1 - + 139). S e q u e n c e s 5' o f - 6 a r e not essential for e n h a n c e r - d e p e n d e n t t r a n s c r i p t i o n in H e L a cells. T h e i d e n t i f i c a t i o n o f t h e s e r e g u l a t o r y e l e m e n t s a n d t h e i r r e s p o n s e to e x o g e n o u s l y a d d e d e n h a n c e r s e q u e n c e s , p r o v i d e s the basis for f u r t h e r r e s e a r c h tow a r d s e l u c i d a t i n g the m e c h a n i s m of t r a n s c r i p t i o n a l activation o f the h u m a n M D R 1 gene.
Acknowledgements
W e t h a n k C. Lincke, J.J.M. Smit, Drs. P.A. Bolhuis, P. Borst, A . H . Schinkel, G. Z a m a n a n d J. Z o m e r d i j k for v a l u a b l e discussions a n d critical c o m m e n t s on this m a n u s c r i p t . This r e s e a r c h was f u n d e d by T h e N e t h e r lands O r g a n i z a t i o n for Scientific R e s e a r c h (F.B.). References
1 Endicott, J.A. and Ling, V. (1989) Annu. Rev. Biochem. 58, 137-171. 2 Gottesman, M.M. and Pastan, I. (1988) J. Biol. Chem. 263, 12163-12166. 3 Schinkel, A.H. and Borst, P. (1991) Semin. Cancer Biol. 2, 213-226. 4 Ueda, K., Cardarelli, C., Gottesman, M.M. and Pastan, I. (1987) Proc. Natl. Acad. Sci. USA 84, 3004-3008. 5 Lincke, C.R., Van der Bliek, A.M., Schuurhuis, G.J., Van der Velde-Koerts, T., Smit J.J.M. and Borst P. (1990) Cancer Res. 50, 1779-1785. 6 Van der Bliek, A.M., Baas, F,, Van der Velde-Koerts, T., Biedler, J.L., Meyers, M.B., Ozols, R.F., Hamilton, T.C., Joenje, H. and Borst P. (1988) Cancer Res. 48, 5927-5932.
146 7 Shen, D.W., Fojo, A.T., Chin, J.E., Roninson, I.B., Richert, N., Pastan, I. and Gottesman, M.M. (1986) Science 232, 643-645. 8 Baas, F., Jongsma, A.P., Broxterman, H.J., Arceci, R.J., Housman, D., Scheffer, G.L., Riethorst, A., Van Groenigen, M., Nieuwint, A.W. and Joenje, H. (1990) Cancer Res. 50, 5392-5398. 9 Chin, K.V., Tanaka, S., Darlington, G., Pastan, I. and Gottesman, M.M. (1989) J. Biol. Chem. 265, 221-226. 1(1 Ueda, K., Pastan, I. and Gottesman, M.M. (1987) J. Biol. Chem. 262, 17432-17436. 11 Kohno, K., Sato, S., Uchiumi, T., Takano H., Kato, S. and Kuwano, M. (1990) J. Biol. Chem. 265, 19690-19696. 12 Chin, K.V., Ueda, K., Pastan, I. and Gottesman, M.M. (1992) Science 255, 459-462. 13 Cornwell, M.M. (1990) Cell Growth Diff. 1,607-615. 14 Smale, S.T. and Baltimore, D. (1989) Cell 57, 103-113. 15 Smale, S.T., Schmidt, M.C., Berk, A.J. and Baltimore, D. (1990) Proc. Natl. Acad. Sci. USA 87, 4509-4513. 16 Treisman, R.H. and Maniatis, T. (1985) Nature 315, 72-75. 17 Graham, F.L. and Van der Eb, A.J. (1973) Virology 52, 456-462. 18 Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2rid edition, Cold Spring Harbor University Press, Cold Spring Harbor. 19 Zinn, K., Dimaio, D. and Maniatis, T. (1983) Cell 34, 865-879.
20 Jarrel, K.A. and Meselson, M. (19911 Proc. Natl. Acad. Sci. USA 88, 102-104. 21 Ghazal, P., Lubon, H., Flechstein, B. and Henninghausen. L. (1987) Proc. Natl. Acad. Sci. USA 84, 3658-3662. 22 Wall, L., De Boer, E. and Grosveld, F. (1988) Genes Dev. 2, 1089-1100. 23 Dailey, L., Hanly, S.M., Roeder, R.G. and tteintz, N. (1986) Proc. Natl. Acad. Sci. USA 83, 7241-7245. 24 Means, A.L. and Farnham, P.J. (1990) Mol. Cell. Biol. 10, 653661. 25 Biggin, M.D. and Tjian, R. (1988) Cell 53, 699 711. 26 Perkins, K.K., Dailey, G.M. and Tjian, R. (1988) Genes Dev. 2, 1615-1626. 27 Soeller, W.C., Poole. S.J. and Kornberg, T. (1988) Genes Dev. 2, 68-81. 28 Green, M.R., Treisman, R.H. and Maniatis, T. (1983) Cell 35, 135-148. 29 Wieland, S., Schatt, M.D. and Rusconi, S. (1990) Nucleic Acids Res. 18, 5113-5118. 30 Roy, A.L., Meisterernst, M., Pognonec, P. and Roeder, R.G. (1991) Nature 354, 245-248. 31 Seta, E., Shi, Y. and Shenk, T. (1991) Nature 354, 241-244.