Octamer transcription factors bind to two different sequence ... - NCBI

The EMBO Journal vol.8 no.7 pp.2001 -2008, 1989

Octamer transcription factors bind to two different sequence motifs of the immunoglobulin heavy chain promoter

Iris Kemler, Edgar Schreiber, Michael M.Muller, Patrick Matthias and Walter Schaffner Institut fur Molekularbiologie II der Universitit Zurich, Honggerberg, 8093 Zurich, Switzerland Communicated by W.Schaffner

All promoters of immunoglobulin heavy chain genes contain three conserved sequence motifs: a heptamer motif CTCATGA, an octamer motif ATGCAAAT, and a TATA box. We show that, despite their different sequences, both the heptamer and the octamer motif are bound by the same octamer transcription factors (Oct factors, also referred to as OTFs), namely the lymphoidspecific proteins Oct-2A and Oct-2B, as well as the ubiquitous protein Oct-i. Even though binding to the octamer motif is stronger, a single heptamer motif can bind Oct proteins and mediate transcriptional activity in lymphoid cells. Furthermore, factor binding to the octamer motif facilitates binding to the nearby heptamer motif. We propose that the heptamer element plays a role early in B-cell differentiation to ensure that the heavy chain promoters are transcriptionally activated before the light chain promoters, which do not contain the heptamer motif. Key words: tissue-specific transcription/transcription factors/heavy chain promoter/octamer sequence

Introduction Immunoglobulin genes are expressed exclusively in cells of the B-lymphoid lineage. This tissue specificity is largely a property of the immunoglobulin enhancers and promoters (Falkner and Zachau, 1984; Grosschedl and Baltimore, 1985; Mason et al., 1985; Picard and Schaffner, 1985) (Banerji et al., 1983; Gillies et al., 1983; Neuberger, 1983), both of which are lymphoid specific. The octamer motif ATGCAAAT (also referred to as the decanucleotide sequence ATGCAAATNA) is a conserved element in the promoters of Ig heavy (IgH) and light chain (IgL) genes and in the IgH enhancer (Falkner and Zachau, 1984; Parslow et al., 1984). specific transcription (Dreyfus et al., 1987; Gerster et al., 1987; Wirth et al., 1987; Muller et al., 1988). However, the octamer motif has also been identified as an important control element in the promoters of non-tissue-specific genes such as the histone H2B (Sive et al., 1986) and small nuclear RNA (snRNA) genes (Mattaj et al., 1985; Ares et al., 1987). At least three nuclear proteins recognize the octamer sequence: Oct-I (also referred to as NF-Al, OBPIOO, OTF-1), a ubiquitous factor with an apparent mol. wt of 100 ©IRL Press

kd (Davidson et al., 1986; Singh et al., 1986; Sive and Roeder, 1986; Bohmann et al., 1987; Fletcher et al., 1987; Sturm et al., 1987); Oct-2A (NF-A2, OTF-2) of 60 kd (Landolfi et al., 1986; Staudt et al., 1986; Gerster et al., 1987; Scheidereit et al., 1987; Muller et al., 1988); and Oct-2B of 75 kd (Schreiber et al., 1988). The latter two proteins are only found in lymphoid cells. The IgH genes contain, as a second conserved promoter element, a heptamer sequence with the consensus CTCATGA. It is located 2-22 bp upstream of the octamer sequence (Eaton and Calame, 1987; Siu et al., 1987) and is required for full IgH promoter activity in lymphoid cells (Ballard and Bothwell, 1986; Eaton and Calame, 1987). Binding of proteins to the heptamer sequence has been demonstrated (Landolfi et al., 1987; Pruijn et al., 1987). In this paper we demonstrate that, despite their different sequences, both the heptamer and the octamer elements are bound by the same Oct proteins. Even though binding is stronger to the octamer motif, the heptamer motif alone can bind Oct protein and mediate transcriptional activity in lymphoid cells. Furthermore, protein binding to the octamer element facilitates binding to the heptamer element. We speculate that the heptamer element plays a role early in B-cell differentiation to ensure that the heavy chain promoters are transcriptionally activated before the light chain promoters, which do not contain the heptamer motif.

Results Proteins binding to the octamer and heptamer elements of IgH promoters To determine the nature of the proteins binding to the heptamer and octamer elements of the IgH promoter, we performed electrophoretic mobility shift (bandshift) assays.

32P-end-labelled oligonucleotides (Figure 1A), containing either the octamer sequence of the human histone H2B promoter (H2B-octa), or the octamer and heptamer elements of the murine 104-2 IgH promoter (hep+oct+), were incubated with nuclear extracts from non-lymphoid (HeLa) or lymphoid (BJA-B) cells and electrophoresed through a native polyacrylamide gel. Figure lB shows the bandshift experiments performed with HeLa nuclear extract. Fragment H2B-octa (lane 2) gave rise to one major retarded band (complex H-1), resulting from binding of one Oct-I protein (Fletcher et al., 1987). In contrast, with the oligonucleotide containing both the heptamer and octamer element (hep+oct+), two major shifted bands, H-1 and H-2, were detected (lane 3). The faster migrating protein-DNA complex H-I has the same mobility as the complex formed with the H2B-octa fragment (compare lanes 2 and 3), and can also be attributed to binding of one Oct-I molecule. Complex H-2 seems to require a heptamer sequence in addition to an intact octamer sequence, since it is detected only with the hep+oct+ probe and not with any of the other

2001

I.Kemler et al. Ai

heptarer

'- t'4- '1 A

hl A

ocUIITier

'

.-F

-f

H2B-octa

.

.-

'it

AA --': *-'

*-'

.o;.:::uA . CT' -T .: hep ocF P }-s i ,L-' A

hep oct~

1-

I.;--

--

;"-

,';!

.;

\"

>,

i-\

s

_p

_*

_ -

--X ;

hep oct

-

-k:

,

hep+oct

xT

.; .C v r-.

;

...

-

hepta3ner

-

-

;ostarner

m.. e.eO. t.

-

C)KA

.-

h-1I L,-

..7T.-. 6> ~K- K;S.X.

A fa

I)

-N'.

i t-

; I ;K _ _ , (.C _K _ _ _

A

4., ,-.-

_

L.

",

j.,: ".

B

LI'-

:'

4 E..' :-

soo

am --

--

s

*

0

a

+D

. _,

Fig. 1. The octamer and heptamer elements of the IgH promoter bind nuclear proteins from HeLa and BJA-B extract. (A) Regulatory motifs of a typical VH promoter and synthetic oligonucleotides used in this study. H2B-octa corresponds to the octamer sequence from the human histone H2B promoter (Sive et al., 1986). Hep+oct+ contains the heptamer and octamer elements of the murine 104-2 IgH promoter (Ballard and Bothwell, 1986). Hep-oct+, hep+oct-, hep-oct- contain mutations in either the heptamer or the octamer element, or in both. The heptamer and the octamer elements are boxed, mutated nucleotides are indicated in black. (B, C) Native 4% polyacrylamide gel of oligonucleotides incubated with HeLa (B, lanes 2-6) or BJA-B (C, lanes 1-5) nuclear extract. F indicates free fragment. Specific DNA-protein complexes are indicated as H-1 and H-2 for HeLa cell extract and B-I to B-6 for BJA-B cell extract. The bands B-1, B-2 and B-3 represent binding of Oct-2A, Oct-2B and Oct-I respectively. The faster migrating bands are non-specific and probably result from proteins binding to single-stranded oligonucleotides. All bandshift reactions shown in panels B and C were run on the same gel.

probes used, where either or both of the two elements were mutated (lanes 4-6). This suggests that complex H-2 results from the simultaneous binding of two proteins, presumably to the heptamer and the octamer elements of oligonucleotide hep+oct+ (see below), whereas complex H-1 results from binding of only one protein molecule. Interestingly, oligonucleotide hep+oct-, which has a mutated octamer sequence, gave rise to a retarded band exactly comigrating with H-1 (lane 5). This suggests that the same protein that otherwise binds to the octamer motif may bind to the heptamer sequence. In this case, however, the heptamerspecific band was weaker than that resulting from bona fide octamer binding (compare lanes 5 and 4). As expected, the 2002

Fig. 2. Octamer transcription factors bind to the heptamer element. (A) Competition experiments with HeLa nuclear extract, using a 500-fold molar excess of competitor DNA. As octamer-specific competitor (S) oligonucleotide fD (containing the octamer element of the IgH enhancer) and as unspecific competitor (U) oligonucleotide fd (containing a mutated octamer element), were used (for details see Materials and methods). (B) Bandshift experiment with gel-purified Oct-2A. DNA-protein complexes are indicated as B-1 to B-3. (C) Bandshift experiment with in vitro produced Oct-2 protein. Lane 1, oligonucleotide hep+oct+ incubated with rabbit reticulocyte lysate without synthetic RNA (negative control); lanes 2-6, rabbit reticulocyte lysate programmed with in vitro generated Oct-2 rnRNA, incubated with the oligonucleotides indicated above each lane; lane 7, H2B-octa oligonucleotide incubated with BJA-B nuclear extract. The bands B-1, B-2 and B-3 represent binding of Oct-2A, Oct-2B and Oct-I respectively; F, free fragment. (D) Bandshift experiment with oligonucleotide hep+oct+ (lane 2) and oligonucleotide VI: 5'-CGAGCCTAGAGCTAATGATATAGCAGAAAGACATGCAAATTAGGCCG-3' (lane 3) and 10 Ag HeLa nuclear extract. Lane 1, no extract.

double mutant hep-oct- did not give rise to any octamerspecific complexes (lane 6). The pattern observed with BJA-B nuclear extract was more complex (Figure IC). The three retarded bands observed with the oligonucleotide H2B-octa (lane 1), designated B-1, B-2 and B-3, are attributed to binding of Oct-2A, Oct-2B and Oct-1, respectively (Gerster et al., 1987; Schreiber et al., 1988). Oct-2B was recently described by Schreiber et al. (1988) and shown, on the basis of limited proteolysis experiments, to be closely related to Oct-2A. A complex

Oct proteins bind to the IgH heptamer element

pattern was seen with the fragment hep+oct+ (lane 2). Bands B- I and B-2 represent binding of one protein molecule of Oct-2A or Oct-2B, whereas the upper three bands B4, B-5 and B-6 are assumed to result from the simultaneous binding of two factors to the same DNA fragment, because they cannot be detected after mutation of either the heptamer or the octamer motif (lanes 3 and 4). Complex B4 migrates at approximately the same position as complex B-3 (compare lanes 2 and 1) and probably consists of a mixture of DNA fragments bound by two Oct-2A molecules and DNA fragments bound by one Oct-I protein.

0ffm

-

_

-EJI

-rn-mm-s

'

.,

im-,

The heptamer element is bound by octamer transcription factors Protein-DNA complexes formed by an oligonucleotide containing only the heptamer element migrate in a bandshift gel at the same position as the complexes generated with the octamer sequence (Figure lB and C). Therefore it seemed plausible that the heptamer motif was also recognized by Oct proteins. To test this possibility, we performed binding competition experiments with HeLa extract (Figure 2A) and further binding studies with a gel-purified preparation of Oct-2A (Figure 2B) and in vitro produced protein from cloned Oct-2 cDNA (Figure 2C). In the competition experiment using HeLa nuclear extract (Figure 2A), complex H-1 formed either with the octamer (hep-oct+, lane 3) or with the heptamer element (hep+oct-, lane 6), and complexes H-1 and H-2 (lane 1), generated with oligonucleotide hep+oct+, were greatly reduced or eliminated by the presence of an excess of unlabelled fragment containing the octamer element of the IgH enhancer (lanes 2, 4 and 7), but not by the same amount of a mutated octamer sequence (lanes 5 and 8) or by the CCAAT element from the al-globin promoter (data not shown). We next separated nuclear extracts from BJA-B cells in a SDS-protein gel and renatured the eluted proteins for DNA binding, as described by Schreiber et al. (1988). The fraction containing Oct-2A was used for a bandshift experiment with the above-described oligonucleotides. Figure 2B shows that this specifically enriched Oct-2A preparation can bind to both the octamer containing oligonucleotide hep-oct+ (lane 3) and the heptamer containing oligonucleotide hep+oct- (lane 4), in both cases forming complex B-1. With the wild-type oligonucleotide hep+oct+, a slower migrating complex could be detected in addition to complex B-1, probably resulting from binding of two Oct-2A molecules to the same DNA fragment. To demonstrate conclusively that an Oct factor binds to the heptamer element, we used protein obtained from cloned Oct-2 cDNA (Muller et al., 1988). For this, run-off RNA was synthesized from the clone pGemC +, containing this cDNA, and subsequently translated in vitro in a rabbit reticulocyte lysate. The in vitro produced Oct-2 protein was used for a bandshift assay with the five oligonucleotides shown in Figure IA. This revealed that the in vitro synthesized Oct-2 protein bound to all DNA fragments except for the double mutant hep-oct- (Figure 2C, lanes 2-6). The retarded bands migrated at the same position as the Oct-2A from BJA-B nuclear extract (lane 7, complex B-1). The complex formed with oligonucleotide hep+octcould be competed out with an excess of unlabelled fragment containing an octamer motif, but not with the same amount

DI MN 8

:.

.:- ;-.' x El 7' ,I

wMa

Fig. 3. Protein binding to the heptamer element is facilitated by protein binding to the octamer element (A) Left panel: autoradiograph of a titration experiment using constant amounts of oligonucleotide hep+oct+ and increasing amounts of HeLa nuclear extract (as indicated above each lane). Protein-DNA complexes H-1 and H-2 are indicated by arrows. Right panel: Binding curves for complexes H-1 (open circles) and H-2 (solid circles). The radioactivity in each band, quantitated by scintillation counting, is plotted against the amount of HeLa nuclear extract used. (B) Titration experiment with oligonucleotide hep-oct+. (C) Titration experiment with oligonucleotide H2B-octa. In lanes 1-9 of (A-C) the following amounts of poly(dI/dC) were used: 3, 0.012, 0.025, 0.05, 0.1, 0.2, 0.4, 0.8, 1.6 zg. Furthermore, each binding reaction contained 5 jig of bovine serum albumin.

of a mutated fragment (data not shown). Control experiments performed with lysates primed with no synthetic RNA (lane 1) showed a weak, slower migrating band, which we attribute to an octamer binding activity present in the lysate itself, presumably Oct-I. Since binding of two molecules of Oct-2A results in a complex migrating at approximately the same position as the octamer binding activity of the lysate, it cannot be concluded from this experiment whether two in vitro

2003

I.Kemler et al.

translated Oct-2A proteins can bind simultaneously to the oligonucleotide hep+oct+. The additional faster migrating complexes seen in lanes 2-6 might be ascribed to binding of shorter in vitro translation products (see also Miller et al., 1988). These experiments clearly show that octamer transcription factors are able to recognize the conserved heptamer element of IgH promoters, even in the absence of an additional octamer element.

a.. .. _

Protein binding to the octamer element facilitates protein binding to the heptamer element To analyze whether Oct protein binding to the heptamer motif was facilitated by Oct binding to the octamer motif, we performed a titration experiment in which a constant amount of the oligonucleotide hep+oct+ was incubated with increasing amounts of HeLa nuclear extract (Figure 3A). In this experiment the concentration of the competitor poly(dI/dC) was adjusted in proportion to the amount of extract, in order to avoid non-specific competition at low protein concentrations. Low extract concentrations generated predominantly complex H-1, whereas higher concentrations gave rise to the additional complex H-2. Quantitation of the protein-DNA complexes H-I and H-2 by scintillation counting revealed that complex H- I increased approximately linearly with increasing extract, whereas complex H-2 increased exponentially. However, this exponential increase of complex H-2 formation could merely be the result of complex stabilization due to high extract concentrations. To show that this is not the case, we performed the same titration experiments with the oligonucleotides hep-oct+ (Figure 3B) and H2B-octa (Figure 3C), which contain only one binding site. In both cases the amount of DNA-protein complex increased linearly with increasing extract concentration. Furthermore, the same result was also obtained with fragment hep+oct- (data not shown). Thus, these results show that Oct-I binding to the octamer motif facilitates binding to the heptamer motif. Methylation interference analysis reveals Oct-1 contact sites over the heptamer element To identify the sites of protein binding to the heptamer and octamer elements we performed methylation interference analysis (Siebenlist and Gilbert, 1980) with the oligonucleotides H2B-octa, hep+oct+, hep+oct- and hep-oct+. In all cases both strands were partially methylated with dimethyl sulphate and incubated with HeLa nuclear extract. After separation of the complexes from the unbound DNA in a preparative bandshift assay, the DNA was eluted from the gel and subjected to chemical cleavage (Materials and methods). The results of the methylation interference analysis are shown in Figure 4A (upper strand) and Figure 4B (lower strand), and are summarized in Figure 4C. In complex H-2, generated with the wild-type oligonucleotide hep+oct+ (compare with Figure IB, lane 3), protein was bound over both the octamer and the heptamer elements, whereas in complex H-1, formed with the same fragment, protein interactions were seen only with the octamer element. In the mutant oligonucleotide hep+oct-, the only site of interaction was the heptamer element and not the mutated octamer binding site. Similarly, in the case 2004

Fig. 4. Methylation interference analysis with HeLa nuclear extract reveals contact sites over the heptamer element. (A, B) Autoradiographs of the methylation interference pattern of the upper and lower strands of the indicated oligonucleotides. For comparison, the lower strand of oligonucleotide H2B-octa is shown together with the upper strand of the other templates. Lanes labelled H-1 and H-2 (lanes 2, 5, 6, 9, 12) correspond to the protein-bound probe and lanes labelled F (lanes 3, 7, 10, 13) correspond to the free probe. The nucleotides whose methylation completely or partially interferes with protein binding are indicated by solid or open triangles respectively. (C) Summary of the methylation interference patterns.

of the mutant heptamer site (hep-oct+), the only protein binding was seen with the intact octamer sequence.

Oct proteins bind to the IgH heptamer element

A

..,

-R,--

-

t;

-Z -

B ,.X

I

it

2 3 4 Fig. 5. The heptamer element stimulates transcription in lymphoid cells. (A) Plasmids used for the transfection experiments. Oligonucleotides containing the heptamer and the octamer elements of the 70ZH+ promoter (Nelson et al., 1983) and the respective mutants were introduced upstream of the f-globin TATA box in the OVEC vector (Westin et al., 1987). The heptamer and the octamer elements are boxed, the mutated nucleotides are indicated in black. OVEC-2, containing only the 3-globin TATA box, served as negative control. All constructs contained an SV40 enhancer downstream of the 3-globin gene (see also Materials and methods). (B) RNase mapping of cytoplasmic RNA from BJA-B cells transfected with the constructs shown in (A). Test indicates the position of the correctly initiated RNA and ref indicates the position of the reference signal from OVECREF. Lane M, size marker (HpaII-cleaved pBR322).

In all cases where protein binding was seen over the octamer motif (H2B-octa, hep+oct+, hep-oct+), this

produced the characteristic methylation interference pattern found with extracts from several other cell lines (Staudt et al., 1986; Pruijn et al., 1987; Muller et al., 1988; Schreiber et al., 1988). In the oligonucleotides hep+oct+ and hep+oct-, protein binding to the heptamer element was abolished by methylation of both adenines and the guanine of the top strand, one adenine of the bottom strand and the adenine residue which lies between the heptamer and the octamer element in the bottom strand. It should be noted that the methylation interference patterns generated with the mutants hep+oct- and hep-oct+ are complementary subsets of the interferences observed with the wild-type fragment in complex H-2. With the oligonucleotide hep+oct+ only modifications of nucleotides in the octamer motif abolish formation of complex H-1. This suggests that the binding affinity of Oct-I is higher for the octamer than for the heptamer sequence. Consistent with this notion, we could show by competition experiments with HeLa nuclear extract that the affinity of Oct-I for the octamer element is -8-fold higher than for the heptamer element (data not shown). These results show that the complex formed with the

oligonucleotide hep+oct- is due to protein binding directly to the heptamer element. This was confirmed by a bandshift experiment performed with the heptamer and octamer elements of the VI IgH promoter (Eaton and Calame, 1987), where the two elements are separated by 14 nucleotides: 5 '-CGAGCCTAGAGCTAATGATATAGCAGAAAGACATGCAAATTAGGCCG-3'. As shown in Figure 2D, Oct-I can also bind to the octamer and heptamer motifs of the V1 promoter and generate the complex H-2 (lane 3), although this is weaker than the complex formed with the oligonucleotide hep+oct+ (lane 2). The heptamer element can stimulate transcription in lymphoid cells To understand its functional significance, we next asked whether the heptamer element alone was sufficient to stimulate transcription in lymphoid cells (BJA-B). For this, we constructed the plasmids shown in Figure 5A, where the heptamer and octamer elements or their respective mutants were inserted upstream of the rabbit f-globin TATA box. All plasmids carry the SV40 enhancer downstream of the globin gene and as an internal control plasmid OVEC-REF was cotransfected (Westin et al., 1987). RNase mappings were performed to measure the amount of globin mRNA

2005

I.Kemler et al.

transcribed from both the test and control genes. All bands were excised and quantitated by scintillation counting. A promoter containing only the TATA box (OVEC-2) produced very few transcripts in BJA-B cells (lane 1). Addition of either the histone octamer element (H2B-octa) or the heptamer and octamer elements as they are found in the promoter of the 70ZH+ heavy chain gene (70Zhep+oct+) gave a strong signal (lanes 2 and 3). Mutation of the heptamer motif (7OZhep-oct+) reduced the transcriptional activity to 70% as compared to wild-type (lane 4), whereas the mutation of the octamer element (70Zhep+oct-, lane 5) led to a more dramatic decrease in transcription efficiency, resulting in 10% of the original level. With the double mutant very few transcripts could be detected (lane 6). The low level of activity obtained with the heptamer element alone (70Zhep+oct-) still represented a 2-fold stimulation of transcription as compared to the double mutant 7OZhep-oct- (lanes 5 and 6). These results show that the heptamer element is required for full transcriptional activity in lymphoid cells and that this element alone can stimulate transcription, albeit weakly.

Discussion The same protein recognizes two different sequences Here we have demonstrated that, despite very different sequences, both the heptamer and the octamer motifs of IgH promoters bind the same Oct proteins, namely the lymphoidspecific factors Oct-2A and Oct-2B, as well as the ubiquitous protein Oct-1. For Oct-2A, we have shown that gel-purified protein as well as in vitro translated protein from a cloned Oct-2 cDNA bind to the heptamer sequence. It has been shown by others that octamer transcription factors can interact with degenerate octamer sites present in the SV40 enhancer (Sturm et al., 1987), in the promoters of homeotic genes of Drosophila melanogaster (Ko et al., 1988; Thali et al., 1988) and with the TAATGARAT (R=purine) motif of herpes virus immediate early promoters (Baumruker et al., 1988; Gerster and Roeder, 1988; O'Hare and Goding, 1988). However, none of these sequences is as dissimilar to the octamer element as the heptamer sequence. Specific recognition of divergent sequences by the same factor has been reported in several other cases in eukaryotes (Davidson et al., 1986; Johnson et al., 1987; Pfeifer et al., 1987; Hwung et al., 1988; Barberis et al., 1989). We entertain the possibility that the heptamer and the octamer motif, although displaying no sequence homology, create a similar binding site with the DNA in a particular conformation that may be further stabilized by flanking base pairs. Thus, if the consensus octamer sequence is degenerate (e.g. the heptamer element), Oct factor binding to this sequence might be stabilized by flanking nucleotides that possibly favour the correct DNA conformation for Oct binding. Consistent with this, recent experiments of Baumruker et al. (1988) showed that Oct-I binds to divergent DNA sequences, including the TAATGARAT motif, provided that the flanking nucleotides show some, as yet poorly understood, 'permissivity'. Not only one, but in fact three different proteins (Oct-1, Oct-2A and Oct-2B) are able to recognize the heptamer and octamer element. Moreover, Drosophila homeotic proteins (Ubx and Abd-B) can trans-activate from promoters

2006

containing an octamer element, apparently by binding to this sequence (Thali et al., 1988). Both the Oct proteins and the Drosophila homeoproteins have in common a 60 amino acid conserved homeodomain. This has become evident after the recent cloning of cDNAs for Oct-I and Oct-2 proteins by several groups (Clerc et al., 1988; Ko et al., 1988; Muller et al., 1988; Scheidereit et al., 1988; Sturm et al., 1988). Sequence analysis revealed that Oct-I and Oct-2 share a second region of sequence similarity, upstream of the homeodomain, called the POU-specific box (Herr et al., 1988). Both conserved domains are required for DNA binding of Oct-1, irrespective of whether a consensus octamer sequence or a TAATGARAT sequence are used (Sturm and Herr, 1988). Since both Oct-I and Oct-2 can bind to the heptamer sequence CTCATGA, and since the only conserved region between the two proteins is the POU/homeobox domain, we consider it likely that this domain is also responsible for the binding to the heptamer sequence. Does the heptamer element regulate IgH gene transcription? It is known that Ig heavy chain promoters are activated before Ig light chain promoters during B-cell differentiation (reviewed in Yancopoulos and Alt, 1986). Since the heptamer element is conserved in IgH but not in IgL promoters (Ballard and Bothwell, 1986; Eaton and Calame, 1987), this element might be involved in the timing of IgH expression. One possibility is that at an early stage of Bcell differentiation an additional factor, perhaps together with Oct, facilitates transcription from heavy chain promoters via the heptamer motif. Alternatively, co-operative binding of Oct factors to the octamer and heptamer elements might help to regulate IgH expression. By titration experiments we show that Oct binding to the octamer motif facilitates protein binding to the heptamer motif. However, mutation of the heptamer element does not have a drastic effect on transcriptional activity in the cell line analysed here, which represents a mature B-cell. In cells representing late stages of the B-cell differentiation, a single octamer element together with a TATA box is sufficient to constitute a good promoter (Dreyfus et al., 1987; Wirth et al., 1987; Muller et al., 1988). Co-operative binding to the heptamer and octamer elements might be advantageous in a situation where less octamer protein is present, as in pre-B-cells. Indeed, two different pre-B cell lines have been shown to contain low Oct-2 protein amounts, as compared to mature B-cells (Staudt et al., 1988). Therefore, the heptamer element might be crucial for the timing of IgH expression in the early stages of B-cell differentiation, by allowing co-operative binding of Oct proteins to the promoter.

Materials and methods Preparation of nuclear extracts, bandshift experiments and methylation interference analysis Nuclear extracts from HeLa cells and BJA-B cells were prepared as described by Schreiber et al. (1988). For bandshifts, 1-3 fmol of 32P-end-labelled oligonucleotides [sequences as in the 104-2 IgH promoter (Ballard and Bothwell, 1987)] were incubated

with 10 ntg HeLa or BJA-B nuclear extract and 4 ig poly(dI/dC) in 15 id of a buffer containing 4% Ficoll, 30 mM KCI, 1 mM EDTA, 1 mM DTT

Oct proteins bind to the and 0.25 mg/ml bovine serum albumin (BSA, Boehringer Mannheim). After 15 min incubation at room temperature the binding reaction was separated on a native 4 % polyacrylamide gel (20:1 crosslink) in 0.25 xTBE buffer at 10 V/cm for 2 h at room temperature. Gel-purified Oct-2 protein was obtained from BJA-B nuclear extracts as described by Schreiber et al. (1988). Five microlitres of the total renatured protein fraction of 1500 'l was used for each bandshift analysis, 100 ng of poly(dI/dC) was used as unspecific carrier. Competition experiments were done by mixing a 500-fold molar excess of competitor DNA to the binding reaction before adding the nuclear extract. As competitor DNAs, oligonucleotides fD and fd were used (Gerster et al., 1987), both derived from the murine IgH enhancer (corresponding to nucleotides 518-564). fD:

5'-CTGAGCAAAACACacCCgGGGTAATTTGCATTTCTAAAATAAGTcGA-3

fd:

S'-CTGAGCAAAACACacCCgGGGTAATgTtCAgTTCTAAAATAAGTcGA-3'

The Ephrussi element (f) is indicated in italics, the octamer element (D/d) is underlined; point mutations are indicated by lower-case letters. Methylation interference analyses were performed as described by Holler et al. (1988).

In vitro translation Run-off transcripts were prepared by transcribing I yig of XbaI-linearized pGemC+ (Muller et al., 1988) in a 30 j4l reaction containing 2.5 mM of all rNTPs, 30 mM Tris-HCI (pH 7.5), 15 mM MgCI2, 15 mM DTT, 40 U RNasin and 80 U T7 RNA polymerase (New England Biolabs) at 37°C for 40 min. Translation reactions were done as described by Rusconi and Yamamoto (1987) with 7 Itl RNA-DNA mixture and 20 i1 rabbit reticulocyte lysate (Promega) in a final volume of 30 1a. For bandshifts, 1 141 of the translation reaction was used in each binding reaction.

Transfections and RNA analysis BJA-B cells were transfected by the DEAE -dextran procedure with 4 yg test plasmid and 1 yg of reference plasmid as described by Gerster et al. (1987). After incubation of the cells for -40 h, cytoplasmic RNA was extracted (Gerster et al. 1987). Forty micrograms of RNA was used for hybridization to a radioactive complementary strand RNA probe (spanning positions -37 to + 179 of the vector; Westin et al., 1987), generated by SP6 RNA polymerase. Hybridization was done at 37°C overnight. Hybridization products were digested with RNase A (6 itg/ml) and RNase TI (10 U/ml) at 37°C for 60 min and separated on a 6% polyacrylamide/7 M urea gel. Experiments were quantitated by scintillation counting of excised bands. Construction of plasmids The plasmid 70Zhep+oct+ was constructed by cloning the double-stranded

oligonucleotide.

5'-CGAGTATCCTGCTCATGAATATGCAAATCCTCTGG-3' 3'-TCGAGCTCATAGGACGAGTACTTATACGTTTAGGAGACCAGCT-5'

IgH heptamer element

References Ares,M., Chung,J.-S., Giglio,L.

and Weiner,A.M. (1987) Genes Dev., 1, 808-817. Ballard,D.W. and Bothwell,A. (1986) Proc. Natl. Acad. Sci. USA, 83,

9626-9630.

Banerji,J., Olson,L. and Schaffner,W. (1983) Cell, 33, 729-740. Barberis,A., Superti-Furga,G., Vitelli,L., Kemler,I. and Busslinger,M. (1989) Genes Dev., in press. Baumruker,T., Sturm,R. and Herr,W. (1988) Genes Dev., 2, 1400-1413. Bohmann,D., Keller,W., Dale,T., Scholer,H.R., Tebb,G. and Mattaj,I.A. (1987) Nature, 325, 268-272. Clerc,R.G., Corcoran,L.M., LeBowitz,J.H., Baltimore,D. and Sharp,P.A. (1988) Genes Dev., 2, 1570-1581.

Davidson,I., Fromental,C., Augereau,P., Wildeman,A., Zenke,M. and Chambon,P. (1986) Nature, 323, 544-548. Dreyfus,M., Doyen,N. and Rougeon,F. (1987) EMBO J., 6, 1685-1690. Eaton,S. and Calame,K. (1987) Proc. Natl. Acad. Sci. USA, 84, 7634-7638.

Ephrussi,A., Church,G.M., Tonegawa,S. and Gilbert,W. (1985) Science,

227, 134-140. Falkner,F.G. and Zachau,H.G. (1984) Nature, 310, 71-74. Fletcher,C., Heintz,N. and Roeder,R.G. (1987) Cell, 51, 773-781. Gerster,T., Matthias,P., Thali,M., Jiricny,J. and Schaffner,W. (1987) EMBO J., 6, 1323-1330. Gerster,T. and Roeder,R.G. (1988) Proc. Natl. Acad. Sci. USA, 85, 6347-6351. Gillies,S.D., Morrison,S.L., Oi,V.T. and Tonegawa,S. (1983) Cell, 33, 717-728.

Grosschedl,R. and Baltimore,D. (1985) Cell, 41, 885-897. Herr,W., Sturm,R.A., Clerc,R.G., Corcoran,L.M., Baltimore,D., Sharp,P.A., Ingraham,H.A., Rosenfeld,M.G., Finney,M., Ruvkun,G. and Horvitz,H.R. (1988) Genes Dev., 2, 1513-1516. Holler,M., Westin,G., Jiricny,J. and Schaffner,W. (1988) Genes Dev., 2, 1127-1135. Hwung,Y.-P., Crowe,D.T., Wang,L.-H., Tsai,S.Y. and Tsai,M.-J. (1988) Mol. Cell. Biol., 8, 2070-2077. Johnson,P.F., Landschulz,W.H., Graves,B.J. and McKnight,S.L. (1987) Genes Dev., 1, 133-146. Ko,H.-S., Fast,P., McBride,W. and Staudt,L.M. (1988) Cell, 55, 135-144. Landolfi,N.F., Capra,J.D. and Tucker,P.W. (1986) Nature, 323, 548-551. Landolfi,N.F., Capra,J.D. and Tucker,P.W. (1987) Proc. Natl. Acad. Sci. USA, 84, 3851-3855. Landolfi,N.F., Yin,X.M., Capra,J.D. and Tucker,P.W. (1988) Nucleic Acids Res., 16, 5503-5514. Mason,J.O., Williams,G.T. and Neuberger,M.S. (1985) Cell, 41, 479-487. Mattaj,I.W., Lienhard,S., Jiricny,J. and De Robertis,E.M. (1985) Nature, 316, 163-167. Muller,M.M., Ruppert,S., Schaffner,W. and Matthias,P. (1988) Nature, 336, 544-551. Nelson,K.J., Haimovich,J. and Perry,R.P. (1983) Mol. Cell.Biol., 3, 1317-1332.

containing the heptamer and the octamer element of the 70ZH + heavy chain promoter (Nelson et al., 1983), into the Sacl and Sall sites of the OVEC vector (Westin et al., 1987) in front of the fl-globin TATA box. Subsequently, upstream ,B-globin sequences were removed by cloning the 3400 bp SacI -Hindml fragment of this plasmid into the SacI-HindlII sites of pUC 18. Mutant oligonucleotides containing point mutations in either the heptamer or the octamer element, or both, and the oligonucleotide H2Bocta were subcloned the same way to create plasmids 70Zhep+oct-, 70Zhep-oct+, 7OZhep-oct- and H2B-octa. All constructs contained an SV40 enhancer downstream of the 3-globin gene. For the transfection analysis we used the heptamer and the octamer elements of the 70ZH + promoter, which is known to be functional (Nelson et al., 1983). No differences were observed in Oct binding to these sequences compared to the 104-2 promoter, either in bandshift experiments or in a methylation interference analysis. OVEC-2 is a derivative of OVEC, lacking upstream j3-globin sequences. As a reference gene the plasmid OVEC-REF (Westin et al., 1987) was used.

Acknowledgements We would like to thank Stefan Tanno for synthesizing the oligonucleotides used for our studies, Dr Sandro Rusconi for help with in vitro translations and Fritz Ochsenbein for graphical work. We also thank Drs Deborah Maguire, Daniel Schumperli and Roberto Cattaneo for critical reading of the manuscript and valuable comments. This work was supported by the Kanton Zurich and by the Swiss National Science Foundation.

Neuberger,M.S. (1983) EMBO J., 2, 1373 -1378. O'Hare,P. and Goding,C.R. (1988) Cell, 52, 435-445.

Parslow,T.G., Blair,D.L., Murphy,W.J. and Granner,D.K. (1984) Proc. Natl. Acad. Sci. USA, 81, 2650-2654. Pfeifer,K., Prezant,T. and Guarente,L. (1987) Cell, 49, 19-27. Picard,D. and Schaffner,W. (1985) EMBO J., 4, 2831-2838. Pruijn,G.J.M., van Driel,W., van Miltenburg,R.T. and van der Vliet,P.C. (1987) EMBO J., 6, 3771-3778. Rusconi,S. and Yamamoto,K.R. (1987) EMBO J., 6, 1309-1315. Scheidereit,C., Heguy,A. and Roeder,R.G. (1987) Cell, 51, 783-793. Scheidereit,C., Cromlish,J.A., Gerster,T., Kawakami,K., Balmaceda,C.G., Currie,A. and Roeder,R.G. (1988) Nature, 336, 551-557. Schreiber,E., Matthias,P., Muller,M.M. and Schaffner,W. (1988) EMBO J., 7, 4221-4229. Siebenlist,U. and Gilbert,W. (1980) Proc. Natl. Acad. Sci. USA, 77, 122-126.

Singh,H., Sen,R., Baltimore,D. and Sharp,P.A. (1986) Nature, 319, 154-158. Siu,G., Springer,E.A., Huang,H.V., Hood,L.E. and Crews,S.T. (1987) J. Immunol., 138, 4466-4471. Sive,H.L. and Roeder,R.G. (1986) Proc. Natl. Acad. Sci. USA, 83,

6382 -6386.

Sive,H.L., Heintz,N. and Roeder,R.G. (1986) Mol. Cell. Biol., 3329-3340.

6.

Staudt,L.M., Singh,H., Sen,R., Wirth,T., Sharp,P.A. and Baltimore,D. (1986) Nature, 323, 640-643. 2007

I.Kemler et al.

Staudt,L.M., Clerc,R.G., Singh,H., LeBowitz,J.H., Sharp,P.A. and Baltimore,D. (1988) Science, 241, 577-580. Sturm,R.A. and Herr,W. (1988) Nature, 336, 601-604. Sturm,R., Baumruker,T., Franza,B.R.,Jr and Herr,W. (1987) Genes Dev., 1, 1147-1160. Sturm,R.A., Das,G. and Herr,W. (1988) Genes Dev., 2, 1582-1599. Thali,M., Muller,M.M., DeLorenzi,M., Matthias,P. and Bienz,M. (1988) Nature, 336, 598-601. Westin,G., Gester,T., Muller,M.M., Schaffner,G. and Schaffner,W. (1987) Nucleic Acids Res., 15, 6787-6798. Wirth,T., Staudt,L. and Baltimore,D. (1987) Nature, 329, 174-178. Yancopoulos,G.D. and Alt,F.W. (1986) Annu. Rev. Immunol., 4, 339-368. Received January 11, 1989; revised March 28, 1989

Note added in proof Independently, Poellinger and Roeder (1989) have also shown the binding of Oct factors to the heptamer element of IgH promoters (Mol. Cell. Biol., 9, 747-756; Nature, 337, 573-576).

2008