gene of Dictyostelium discoideum - NCBI

7 downloads 130015 Views 1MB Size Report
auto-digestion (reviewed by North, 1982). The cysteine ... et al., 1985) genes encodeproteins with a high degree of ... precise degree of enrichment has been published. ..... added to repair the deletion to the boundaries of the GC-rich region.
The EMBO Journal vol.6 no. 1 pp. 195 -200, 1987

Identification of a DNA sequence element required for efficient expression of a developmentally regulated and cAMP-inducible gene of Dictyostelium discoideum

Catherine J.Pears and Jeffrey G.Williams Imperial Cancer Research Fund, Clare Hall Laboratories, Blanche Lane, South Mimms, Herts EN6 3LD, UK Communicated by D.M.Glover

The cysteine proteinase 2 gene of Dictyostelium encodes a developmentally regulated sulphydryl proteinase which is first expressed late during cellular aggregation. The mRNA is slightly enriched in pre-stalk over pre-spore cells but we show here that it is expressed at a somewhat higher level in mature spore cells than in stalk cells. The mRNA is induced to accumulate precociously in response to exogenous cAMP and we show that a fragment of DNA containing 921 nucleotides upstream of the major start site of transcription directs regulated expression of the gene. Approximately 200 nucleotides upstream of the cap site there are two, adjacent, homologous G-rich regions of 9 nucleotides in length. We have constructed small internal deletions and point mutations which affect the distal element. We find a major reduction in regulated transcription indicating this element to be inportant in mediating cAMP induction of gene expression. Key words: cAMP-inducible genes/control elements/cysteine proteinase 2lDictyostelium discoideum

Introduction Upon starvation, individual amoebae of the slime mould Dictyostelium discoideum aggregate in response to pulses of cAMP to form a mound of cells in which differentiation occurs to yield stalk and spore cells. During differentiation, a large fraction of cellular protein is degraded and there are major changes in proteolytic activity which are presumed to be responsible for this auto-digestion (reviewed by North, 1982). The cysteine proteinase 1 (Williams et al., 1985) and cysteine proteinase 2 (Pears et al., 1985) genes encode proteins with a high degree of homology to plant and animal sulphydryl proteinases. Aside from the regions of the two proteins likely to be important in catalysis, the genes are highly diverged in sequence and they differ in both the number and position of introns (Pears et al., 1985). The two genes are, however, perfectly co-regulated during development. Both mRNA sequences are absent during growth and early aggregation but are abundantly expressed late during cellular aggregation. In addition to its role in chemotaxis, the rise in cAMP levels during aggregation appears to act to switch the pattern of gene expression - repressing accumulation of several markers of early aggregation (Yeh et al., 1978; Williams et al., 1980) and inducing the accumulation of later products (Klein, 1975; Gerisch et al., 1975; Town and Gross, 1978; Williams et al., 1980; Mehdy et al., 1983; Mullens et al., 1984; Haribabu et al., 1986). This has been shown by adding cAMP to cells early during aggregation, prior to the normal rise in cAMP levels, and measuring the effect on enzyme activity or specific mRNA concentration. The precise mechanism of action of cAMP is not known but there © IRL Press Limited, Oxford, England

is evidence that the activation of several genes is mediated by the cell-surface cAMP receptor (Oyama and Blumberg, 1986; Haribabu and Dottin, 1986). The cysteine proteinase 1 gene is inducible by cAMP (Williams et al., 1980) and we show here that the cysteine proteinase 2 gene is also cAMP inducible. The migrating slug is composed of an anterior portion contianing cells which differentiate to form stalk cells and a posterior portion containing spore-cell precursors. Pre-spore and pre-stalk cells can be purified by micro-dissection or by density gradient centrifugation. We have estimated the cysteine proteinase 1 mRNA sequence to be -2-fold, and the cysteine proteinase 2 mRNA to be 3-fold, enriched in pre-stalk over pre-spore cells (Pears et al., 1985). The cysteine proteinase 2 gene is present in a single copy in the genome (Pears et al., 1985) and displays an identical restriction map and pattern of temporal regulation to a gene encoding a protein named the pre-stalk-cathepsin, (Datta et al., 1986). The pre-stalk-cathepsin is described as being prestalk specific, (Datta et al., 1986) although no estimate of the precise degree of enrichment has been published. Here we have re-investigated the degree of pre-stalk enrichment of the cysteine proteinase 2 mRNA using a different method of gradient purification from that which we used previously and by micro-dissection. We have also analysed the expression of the gene in culminating stalk cells and in mature spores. In a previous study (Datta et al., 1986) the 5' portion of the pre-stalk-cathepsin gene has been used in a fusion with the Escherichia coli ,B-glucuronidase gene to show that a 2.5-kb segment of DNA upstream of the cap site is sufficient to allow regulated expression. We show here, using a 'mini-gene' from which an internal segment of the gene has been deleted, that a fragment containing 921 nucleotides of the upstream region is sufficient for regulated expression. We go on to show that mutations within a G-rich element located 200 nucleotides upstream of the cap site produce a drastic reduction in regulated expression of the gene. -

-

Results 77te cysteine proteinase 2 mRNA is expressed in both pathways of cellular differentiation and is inducible by cAMP The cysteine proteinase 2 gene is expressed at a high level from late in aggregation until culmination and in fractionated slugs the mRNA is enriched in pre-stalk over pre-spore cells (Pears et al., 1985). We have now re-investigated the degree of enrichment using a different Percoll gradient fractionation procedure (Ratner and Borth, 1983), from that which we used previously. We have also manually separated pre-stalk and pre-spore cells by microdissection of slugs. The pre-stalk and pre-spore fractions prepared by these two different methods, display an 10% crosscontamination (see legend to Figure IA). Both techniques of cell separation show the cysteine proteinase 2 mRNA to be only 1.5to 2-fold enriched in pre-stalk over pre-spore cells. The apparent degree of enrichment is actually somewhat lower than our previous estimate (Pears et al., 1985). A number of developmen-

195

C.J.Pears and J.G.Williams .t

A

1

_

W4

n.m

B

A

o ; 4 5

.L

4.

W

5_

a_1

4.5 7.5 hrs hrs 0-

+-.

*

*

TCGATGTTAA AGAAAATCTT TTTTTTTTTT TTTCACAGAT TTGGAAAAAT

-870

TAATTTACAT GTTCAAAGGG TTCTAAAAAA AAAAAAAAAA AAAAAAAAAA

-820

AATTATGAAG ATTCCACACA ATAATGATAT CATATTGTTG TGGGTAATGG

-770

ATTGAATCTT TTAAAATTCC GGAAATTTAA AATGGATATC TTAAAAAATA

-720

TAATTTTACA AAAATTTATT TTTTTTCTTT ATTTTTTATC TTTATTTTTT

-670

TTTTAAAAAA TTATTATCAC TCCATTGTAA TAAAAATACA TTTTGGGAAT

-620

TTAAAAAAAT AAAAAAAATA AAATTAATAA ATAAAAAAAA AAAAAAGAAA

-570

GGGTTGTTTT TTCATATTTT AATAGTCAAT AATTGTTTTC TTTGTTTTTA

-520

TAATTAATGT GCCATTTTTA TTCATTTATT TTTTTTTTAT TTTTTTTTTA

-470

ATTTATTTTT TTCTTTTTAT CATTCCTAAA AAAAAAAAAA AGCTATTAAT

-420

TAACCACAAA AAAAAAAAAA CATTTTTTTT TTTTTTTTTT TTTTTTTTTT

-370

TTAATGATAC AATTATTAAA AAACGTCATA AAATAATATG AAAAATCAAA

-320

AGAAATGTTT GATGATCCAA TAGAATTTAT TAAAATAATC ACCAATTGTT

-270

CGGGTGTGTT TTTTTTATTC TTTAAATTTT TTCTATTTAT GAAAACACAG l

-220

AAGTTAGGGG TGGGTTTTAT ATAGGTTTTT TTTATTTTTT TTTTATTATC

-170

ATAATTTATT ATCATTTTTA TATAAATTTT ATTTATTTTA ATTTTAATTA

-120

TTATTTTTTT TTTAAAAATA ATTATTTTTT TTTATTTTTT TTTTTTATAT

-70

TTAAAAAATT TTGTTTAAAA AAAAAAAAGG GTATAAATCT AATAAATTTA

-20

+1

Cysteine proteinas

TTCTTTTTTT TTTTTTTTTT CAAAATCACA ATTTTGTTTT ATAATTTTAT

+28

TTGTATAATA AATTTTATAT AAAAAAAAAC AAAAAAAAAA AAAAAAAAAA

+78

TTAATTATTT TAAAATG MET

+95

B WILD TYPE -AACACAGCGGGTGTGTTAAGTTAGGGGTGGGTTTTATAi i

IL30

A

x

2

Fig. 1. (A) Analysis of the relative abundance of the cysteine proteinase 2 mRNA in differentiated cells. Pre-stalk and pre-spore cells were separated either by Percoll density gradient centrifugation (Ratner and Borth, 1983) or by slug micro-dissection. Stalk and spore cells were purified as described in Materials and methods. The purity of the fraction was determined by staining with an antiserum specific to spore cells. The pre-spore fraction was estimated to be 10% contaminated with pre-stalk cells and the pre-stalk fraction to be 10% contaminated with pre-spore cells. A similar, independent, estimate of the degree of purity was obtained by analysing the RNA samples with a cDNA clone derived from the pre-spore-specific D19 mRNA (Barklis and Lodish, 1983) and with a probe for the highly prestalk-enriched pDd63 mRNA (K.A.Jermyn, R.R.Kay, M.Berks and J.G.Williams, submitted for publication). Total cellular RNA was isolated from the fractions and analysed by Northern transfer using the cysteine proteinase 2 cDNA clone pDd8 as a probe (Pears et al., 1985). Lane 1: anterior (pre-stalk) portion of micro-dissected slugs. Lane 2: posterior (prespore) portion of micro-dissected slugs. Lane 3: upper (pre-stalk) fraction from Percoll gradient. Lane 4: lower (pre-spore) fraction from Percoll gradient. Lane 5: stalk cells. Lane 6: spore cells. Densitometric scans of the autoradiogram indicated the following degrees of enrichment: microdissected slugs, pre-stalk:pre-spore ratio = 1.7:1; gradient fractionated slugs, pre-stalk:pre-spore ratio = 1.5:1; purified stalk and spore cells, stalk:spore ratio = 1:2. (B) Induction of the cysteine proteinase 2 mRNA by cAMP. Exponentially growing Ax2 cells were harvested, washed, resuspended at 107 cells/mi in 17 mM potassium phosphate buffer (pH 6.5) and shaken at 120 r.p.m. at 22°C. After 90 min of development, cAMP was added to 1 mM and cells were harvested 3 and 6 h later. Total cellular RNA was analysed as described in Figure IA.

tally regulated transcripts are known to be specifically degraded after cellular disaggregation (Chung et al., 1981; Mangiarotti et al., 1981) and we believe our previous slightly higher estimate of a 3-fold enrichment may be due to selective mRNA turnover during the much longer centrifugation used in the gradient 196

AIO

XBA

vZzzox>ZA>

v/////"ZZZ

AACACAGCGGGTG TCTAGAAGTTAGGGGTGGGTTTTATA V777777777

B&.HI link*r

-

CGGATCCG

Fig. 2. (A) The nucleotide sequence upstream of the coding region of the cysteine proteinase 2 gene. The sequence shown extends to a TaqI site at nucleotide -921. Nucleotide +1 is the start site of transcription determined by primer extension as described in the text. The cap site is located 93 nucleotides upstream of the initiation codon. The G-rich element at nucleoside -235 is underlined. (B) The nucleotide sequence of the G-rich element. The sequence adjacent to nucleotide -235 is shown with direct, (H) and inverted 0-4) repeats indicated. An XbaI site was created at nucleotide -225 by oligonucleotide-directed mutagenesis. The two internal deletion mutants A30 and /IO were created by exonucleolytic digestion from this site. The deletion mutants were re-circularized via a BamHI linker.

purification method of Tsang and Bradbury (1981). All three methods of separation, which have now been used, clearly demonstrate that this mRNA is present at a high relative concentration in pre-spore cells. We have also extended our analysis of the expression of this gene to terminally differentiated cells. Stalk cells and mature spores were purified and the relative content of the cysteine proteinase 2 mRNA determined by Northern transfer (Figure IA). The mRNA is expressed at an -2-fold higher level in spore cells. We have previously shown the cysteine proteinase 1 gene to be prematurely expressed in response to exogenous cAMP. The cysteine proteinase 1 and 2 mRNA sequences are temporally co-

Dictyostelium discoideum cysteine proteinase 2 gene Cysteine proteinase 2

4.5 7.5 hrs hrs O-

+-

+

ind III

5tin

* 0

Cysteine proteinase 2

Actin 0

a Kpn I

AmpR Fig. 3. Restriction map of the transformation vector containing the cysteine proteinase 2 mini-gene. The Fl origin, donated by the pEMBL vector (Dente et al., 1983) present in this construct, allows for the production of single-stranded DNA for sequence analysis and facilitates site-directed mutagenesis (A.Early and J.G.Williams, in preparation). The restriction fragment from the cysteine proteinase 2 gene present extends from nucleotide -921 and the 3' terminus lies within the 3' non-coding region of the cysteine proteinase 2 gene. A fragment derived from the 3' proximal region of the cysteine proteinase 1 gene was inserted downstream of the mini-gene to provide transcription termination and polyadenylation signals. This fragment extends into the coding region of a convergently transcribed gene of unknown function (A.Early and J.G.Williams, in preparation) and it therefore also provides transcriptional termination and polyadenylation signals for transcripts from the neomycin resistance gene. [The B1O vector of Nellen et al. (1984) does not contain a Dictyostelium trmination and polyadenylation signal downstream of the actin 6/neomycin resistance fusion gene which confers G418 resistance.]

regulated during development and it seemed likely, therefore, that the latter gene would also be inducible by cAMP. The data presented in Figure lB confirm this prediction. Addition of 1 mM cAMP at 1.5 h of development accelerates the first detectable appearance of the cysteine proteinase 2 mRNA sequence by 3-4 h. Characterization of the 5' proximal region of the cysteine proteinase 2 gene We have previously presented the nucleotide sequence of the entire coding region of the cysteine proteinase 2 gene and we have now determined 921 nucleotides of DNA sequence upstream of the initiation codon (Figure 2A). The major start site of transcription (nucleotide + 1) has been determined by primer extension to be 93 nucleotides upstream of the initiation codon (data not shown). There is a potential TATAA sequence element at nucleotide -32 and the cap site is immediately preceded by a run of 17 T residues. This is a common feature of Dictyostelium genes (Kimmel and Firtel, 1982). The entire upstream region is also typical in being predominantly composed of A or T residues, with an overall GC content of < 5 %. There is, however, a short region downstream of position -235 which is relatively G-rich. It contains two direct repeats of a 9-nucleotide element, with only a single nucleotide difference between the homologues. The first of the two elements forms part of a region of weak dyad symmetry (Figure 2B). Construction of a marked derivative of the cysteine proteinase 2 gene and analysis of its expression after transformation into Dictyostelium The construct shown in Figure 3, contains a 'mini-gene' derivative of the cysteine proteinase 2 gene inserted into a Dic-

A x 2 +

minvene

Fig. 4. Analysis of the expression of the cysteine proteinase 2 mini-gene after transformation into Dictyostelium. Total RNA was extracted from Ax2 cells and from a pooled population of cells, transformed with the cysteine proteinase 2 minigene, which were subjected to development as described in the legend to Figure lB. The minigene transcript, AKpn, (780 bases) was resolved from the transcripts of the endogenous gene (1335 bases) on a 1.5% formaldehyde agarose gel, and analysed by Northern transfer as described in Materials and methods.

tyostelium transformation vector based upon the BlO vector of Nellen et al. (1984). A KpnI fragment of 920 nucleotides was removed from the cysteine proteinase 2 gene to create a minigene with an in-frame fusion. Axenically grown cells were transformed with this construct, selected clonally with G418 and then pooled to yield populations derived from several hundred resistant colonies (Nellen and Firtel, 1985). These were grown in suspension in axenic medium in the presence of increasing concentrations of G418 to a final level of 10 ,ug/ml. This results in gene amplification (Nellen and Firtel, 1985) and we estimate the final copy number of the construct to be 40-50 by Southern transfer (data not shown). Restriction mapping revealed no detectable rearrangements of the transformed DNA and indicated that the copies were present as head-to-tail tandem arrays in high mol. wt DNA. The resistant population was subjected to development in suspension and at 1.5 h cAMP was added to 1 mM. Total cellular RNA was isolated at 4.5 and 7.5 h of development and analysed by Northern transfer. The mini-gene produces a discrete transcript of the expected size which is perfectly co-regulated with, and present at approximately equal concentrations to, transcripts from the endogenous gene (Figure 4). Thus 921 nucleotides of 5' proximal sequence of the cysteine proteinase 2 gene directs regulated expression during normal development and is also sufficient to confer cAMP inducibility. Mutagenesis of the upstream region of the gene and analysis by transformation In order to delineate the control elements necessary for cysteine proteinase 2 gene expression, we prepared an 'external' deletion series in which DNA between the TaqI site at -921 and -

197

C.J.Pears and J.G.Williams

4.5 7.5 hrs hrs

4.5 75 hrs hrs

O--t-+

e

9

A 30

.::

o-+--. +=

w

e

of residual expression seen with A30 and A 10 is not caused by a small percentage of cells aberrantly expressing the gene at high levels.

4.5 ITh hrs hrs

9

C.ysteine proteinase 2 A KpnrI

Xba I

a IO

Fig. 5. Analysis of the expression of the derivative of the mini-gene with mutations in the G-rich elements. Development of pooled populations of transformed D. discoideum Ax2 cells was performed as in the legend to Figure 4, and the resulting Northern transfers probed with 5' and 3' proximal KpnI fragments from the cysteine proteinase 2 cDNA clone pDd8 (Pears et al., 1985).

the cap site was progressively removed. Correctly regulated expression was observed using a fragment extending only 327 nucleotides upstream of the cap site (data not shown). However we observed a gradual reduction in the absolute level of transcription as the extent of the deletion was increased. We believe this is most likely to be due to inhibition of downstream elements by vector sequences juxtaposed as a result of the deletion. In order to circumvent the problem, we decided to prepare 'internal' deletions in the G-rich elements located around nucleotide -230. A comparison with a co-regulated gene (see Discussion) strongly suggested that these elements might play a role in regulating gene expression. The region is devoid of useful restriction sites and, in order to manipulate the DNA sequence in this area, it was necessary to create a unique restriction site. An XbaI site was created at nucleotide -225 by site-directed mutagenesis (Figure 2B), the DNA was restricted at this site and small internal deletions were introduced. After re-circularization via a BamHI linker, the nucleotide sequence of several constructs, containing deletions, was determined and two were chosen for further analysis (Figure 2B). The A30 mutant contains a 30-bp deletion, which removes both of the direct repeats, and AMO contains a 10-bp deletion which removes only the upstream element. [Both constructs retain the extreme 5' portion of the element of dyad symmetry (Figure 2B)]. These deletion mutants and the XbaI mutant were transformed into Dictyostelium and analysed as described above. The copy number was estimated by Southern transfer to be 40-50 copies/cell for the wild type and XbaI constructs and 10-20 for A30 and AlO, (data not shown). Analysis of the expression of the mutated constructs (Figure 5) showed that, when normalized to the level of endogenous cysteine proteinase 2 and to the relative copy number, the largest deletion, A30, reduced the expression of the mini-gene by - 50-fold. The smaller deletion, AI0, also reduced expression by a similar degree. The three point mutations introduced to create the XbaI site result in a reduction of regulated expression by 20-fold. These results were confirmed by repeating the analysis using at least three individual clones of transformed cells for each of the mutant constructs. The levels of expression obtained were comparable to those found for the pooled populations (data not shown). Hence the low level -

198

Discussion The cysteine proteinase 2 gene is first expressed late during cellular aggregation prior to overt cellular differentiation. The early appearance of pre-stalk-enriched sequences, such as the cysteine proteinase 2 mRNA, has been held to show that, contrary to the morphological and biochemical evidence (Takeuchi et al., 1977; Oohata, 1983; Schaap, 1983; Krefft et al., 1984), prestalk differentiation precedes pre-spore differentiation (Mehdy et al., 1983; Mehdy and Firtel, 1985). It is obviously therefore of great importance to understand the precise pattern of expression of this class of gene. We have confirmed in this paper our previous data which showed only a low degree of enrichment of this mRNA sequence in pre-stalk over pre-spore cells. There is no published estimate of the degree of enrichment of the prestalk-cathepsin of Datta et al. (1986) and therefore it is difficult to assess the apparent inconsistency in results. However, the strongest piece of data suggesting a high degree of pre-stalk enrichment was immunocytochemical localization of the protein (Datta et al., 1986) and it may be that the protein is more highly localized than the mRNA because of selective translation or protein stability. The marginal degree of enrichment which we observe in prestalk cells appears unlikely to be of major functional significance because the mRNA is expressed at comparable levels in stalk and spore cells. If the intracellular concentration of the mRNA is any guide to protein levels, then the presence of cysteine proteinase 2 mRNA suggests a continued need for this enzyme during the terminal stages of stalk cell differentiation. Its presence in mature spore cells is of some interest since it has previously been shown that cysteine proteinase levels are extremely low in mature spores, with a rapid increase during spore germination (North and Cotter, 1984). The cysteine proteinase 2 mRNA is therefore presumably present in an inactive form in the mature spore and possibly plays a role in re-structuring the cell during spore germination. It appears unlikely to play a role in bacterial digestion after emergence from the spore because we are totally unable to detect the mRNA in vegetative amoebae (Pears et al., 1985). The cysteine proteinase 1 and 2 genes are both expressed coordinately with the major rise in cAMP levels during cellular aggregation. We have previously shown that cAMP, added early during development, induces precocious accumulation of the cysteine proteinase 1 mRNA and we have now shown that the cysteine proteinase 2 gene is also cAMP inducible. Expression of the cysteine proteinase 1 gene is known to be regulated at the transcriptional level (Presse et al., 1986). While there is as yet no direct evidence for the cysteine proteinase 2 gene, it is perfectly co-regulated with the cysteine proteinase 1 gene and hence seems also likely to be controlled at this level. We have investigated the DNA sequence elements involved in transcriptional regulation by manipulation of the DNA upstream of the gene and transformation into Dictyostelium. The region upstream of the cysteine proteinase 2 gene contains an extremely high proportion of A and T residues, as is typical of non-coding regions of Dictyostelium genomic DNA (Kimmel and Firtel, 1982). While there is a precedent for a region of DNA as simple as a poly(dT) tract acting to maintain an elevated level of transcription (Struhl, 1985), it seemed logical

Dictyostelium discoideum cysteine proteinase 2 gene

to assume that the unusually G-rich region at nucleotide -230 might be involved in regulation. Suggestive evidence for this derived from a comparison with the cysteine proteinase 1 gene which also contains two copies of a direct repeat of a very Grich element located - 200 nucleotides upstream of the cap site (D.Driscoll and J.G.Williams, in preparation). We obtained more direct evidence for a role of the G-rich repeats by constructing an external deletion series. However, we observed a gradual reduction in the level of expression as separation of vector sequences from thecap site was reduced. Inhibition by vector sequences seems especially likely to be a problem in Dictyostelium because of the unusual nature of the genome. It might also explain why the level of expression we obtain with the undeleted construct was much lower than expected given the high copy number present in the transformants. We cannot, however, rule out alternative possibilities such as destabilization of the RNA because of the internal deletion used to create the mini-gene or the presence of a generalized enhancer of transcription upstream

of nucleotide -921. The internal deletion, which removes both copies of the Grich element, drastically reduces the level of regulated expression. However, we obtain an equivalent reduction by deleting only the distal element. This suggests either that the downstream element is unnecessary or that both elements must be intact for efficient transcription. The existence of two G-rich elements upstream of the cysteine proteinase 1 gene, present as a direct repeat (D.Driscoll and J.G.Williams, in preparation), suggest that the latter explanation may be correct. There are many precedents for the presence of multiple control elements upstream of eukaryotic genes (Scheidereit et al., 1983; Dierks et al., 1983; Carter et al., 1984; McKnight et al., 1984; Dudler and Travers, 1984; Renkawitz et al., 1984; Stuart et al., 1984; Goodbourn et al., 1985) and in the case of the metallothionein gene there is direct evidence to show that more than one copy of the control element is needed to obtain inducible gene transcription (Searle et al., 1985). However, proof that multiple elements are required in this case will necessitate more extensive manipulations, such as the specific deletion of the downstream element. The point mutations introduced to create the XbaI site result in a major decrease in regulated transcription. This result confirms the importance of the upstream element by ruling out any possible artefactual inhibition due to the DNA rearrangement involved in creating the analogous deletion mutant (A10). The upstream element forms part of a region of weak dyad symmetry (Figure 2B) and the point mutations reduce the degree of symmetry. Clearly more point substitutions are necessary before any firm conclusions can be drawn but dyad symmetry is a common feature of both prokaryotic (reviewed by Pabo and Sauer, 1984) and eukaryotic control elements (Pelham and Bienz, 1982; Karin et al., 1984; Hope and Struhl, 1986; Sen and Baltimore, 1986; Treisman, 1986; Montminy et al., 1986). The element which we have identified is necessary for efficient expression of a cAMP-inducible gene and it is therefore of interest to compare it with elements suggested to play an analogous role in higher eukaryotes. A canonical palindromic element (TGACGTCA) is present upstream of many cAMP inducible genes (Montminy et al., 1986) and a synthetic oligonucleotide containing a version of this consensus sequence acts to confer cAMP inducibility (Comb et al., 1986). There is no apparent homology between this sequence and the G-rich region of the cysteine proteinase 2 gene. However, in contrast to several mammalian mRNA-inducible genes (Boney et al., 1983; Montminy et al., 1986), there is no direct evidence that cAMP acts

intracellular second messenger to activate gene expression in Dictyostelium. Indeed there is evidence to suggest that, for several genes, it may not be involved (Oyama and Blumberg, 1986). A large number of mRNA sequences are inducible by cAMP in Dictyostelium and some variant of the sequence we have defined would be expected to be common to all these genes. The Dll gene, which encodes a cysteine-rich protein of unknown function has an extremely G-rich tract at -200 nucleotides upstream of the start site of transcription (Barklis et al., 1985) but there is as yet no direct evidence that it is involved in gene activation. The effect of the two deletion mutations examined here is very large (> 50-fold) but there is a residual level of gene expression and these transcripts are both developmentally regulated and cAMP inducible. One possible explanation for this observation is that neither deletion is sufficiently extensive to entirely remove the regulatory element. This seems very unlikely, since the remaining sequence retains only a single C residue from within the most distal part of the element of dyad symmetry. A second possibility is that the G-rich element we have deleted is necessary for efficient transcription but that developmental regulation is conferred by an element located elsewhere within the region. We cannot at present rule out this possibility. However, the explanation which we favour is that the region at nucleotide -230 is responsible for developmental regulation but upon deletion of this element, a second element acts to confer developmentally regulated expression at a low level. Again, precedents from other inducible genes suggest that multiple control elements are likely to be present. There are several G-rich regions (at -780 and -850) which may fulfil this role and we are currently investigating this possibility using double mutations and fusions with the promoters of other regulated genes. as the

Materials and methods Cell culture and development Axenically grown cells of the strain Ax2 (obtained from J.Ashworth) were developed on agar as previously described (Pears et al., 1985). Development of cells in suspension was performed as in Williams et al. (1979). Total nucleic acid was isolated by phenol extraction (Jacobson, 1976). The separation of prestalk and pre-spore cells from the strain V12M2 (obtained from G.Gerisch) was carried out by micro-dissection of slugs or by the Percoll density gradient centrifugation method of Ratner and Borth (1983). Using culminates of the strain Ax2, stallc cells were isolated as in Morrisey et al. (1984) and spores were isolated as described by Devine,K.M. et al. (1982). Transformation of Ax2 cells was carried out by a minor modification of the procedure of Nellen et al. (1984). The conditions of glycerol shock were reduced to 15% glycerol for 2 min as we found the method employed by Nellen et al. (1984), for strain Ax3, to be toxic for the Ax2 strain. Pooled populations of transformants were selected for high copy number by growth in 10 jg/ml G418 (Nellen and Firtel, 1985).

Analysis of RNA Primer extension was performed using a synthetic oligonucleotide of 26 residues in length which was 5'-end labelled using T4 polynucleotide kinase. Hybridization was performed with 20 itg of total RNA as described by Devine,J.M. et al. (1982), except that the hybridization temperature was 47°C. The oligonucleotide was complementary to the mRNA from nucleotide +51 to +77, relative to the A of the ATG initiation codon. The products were analysed on a 7.5% denaturing acrylamide gel in parallel with dideoxy sequencing reactions (Sanger et al., 1977) directed by a single-stranded template of a subclone of the gene primed with the same oligonucleotide. Northern transfer with 20 itg of total cellular RNA was performed as described previously (Williams et al., 1985) using as a probe a mixture of 5' and 3' proximal KpnI fragments from the cysteine proteinase 2 cDNA clone, pDd8 (Pears et al., 1985). The results were analysed where using an LKB ultrascan XL laser scanning densitometer. In those situations there was a very large difference in strength between signals, (such as when analysing the deletion mutants in Figure 5) dilutions of the RNA samples were analysed in parallel in order to ensure linearity of response.

199

C.J.Pears and J.G.Williams Insertion of the cysteine proteinase 2 gene into a transformation vector and sequence analysis A TaqI restriction fragment containing the entire coding region and 921 bp of upstream sequence from the cysteine proteinase 2 gene was purified from the genomic clone pDdG4 (Pears et al., 1985). This was inserted into the AccI site of the Dictyostelium transformation vector pBlO (A.Early and J.Williams, in preparation). A Sau3A restriction fragment containing the 3' termini of the cysteine proteinase 1 gene and of an unknown, convergently transcribed gene (A.Early, personal communication) was cloned downstream of the cysteine proteinase 2 gene. The sequence upstream of the cysteine proteinase 2 gene was determined from a deletion series of the above construct (Henikoff, 1984). The deletion mutants were analysed either by the chain termination method (Sanger et al., 1977) or by chemical cleavage (Maxam and Gilbert, 1980). In the latter case restriction fragments were labelled either with T4 polynucleotide kinase or the Klenow fragment of DNA polymerase I. Oligonucleotide-directed mutagenesis and internal deletion A synthetic oligonucleotide of 28 residues in length was synthesized, complementary to a region between nucleotide -218 and -236. It contained three mismatches which, when rendered double stranded by replication and repair in E. coli, created an XbaI restriction site at -225. Mutagenesis was performed by the twoprimer method (Norris et al., 1983) and transformed colonies were screened by hybridization to the oligonucleotide primer (Zoller and Smith, 1982) as described by Mason et al. (1985). Internal deletions were created by cleaving the mutated DNA at the XbaI site and incubating with T4 DNA polymerase in the absence of nucleotide triphosphates, where it acts as a 3' to 5' exonuclease (O'Farrell, 1981). After 1 min under these conditions an excess of dATP and dTTP was added to repair the deletion to the boundaries of the GC-rich region. The projecting 5' termini were removed with Neurospora endonuclease and repaired with the Klenow fragment of DNA polymerase 1. After re-circularization, via a BamHI linker (with the sequence CGGATCCG), the DNA was transformed into DH5 cells, a derivative of DH1 (Hanahan, 1985) and the nucleotide sequence was established by chemical cleavage (Maxam and Gilbert, 1980).

Acknowledgements We would like to thank Keith Jermyn for the pre-stalk and pre-spore cell separations, lain Goldsmith for providing the synthetic oligonucleotides and Niall Dillon and Donna Driscoll for critical reading of the manuscript.

References Barklis,E. and Lodish,H.F. (1983) Cell, 32, 1139-1148. Barklis,E., Pontius,B. and Lodish,H.F. (1985) Mol. Cell Biol., 5, 1473-1479. Boney,C., Fink,D., Schlichter,D., Carr,K. and Wicks,W.D. (1983) J. Biol. Chem., 258, 4911-4918. Carter,A.D., Felber,B.K., Walling,M.J., Jubier,M.F., Schmidt,C.J. and Hamer,D.H. (1984) Proc. Natl. Acad. Sci. USA, 81, 7392-7396. Chung,S., Landfear,S.M., Blumberg,D.D., Cohen,N.S. and Lodish,H.F. (1981) Cell, 24, 785-797. Comb,M., Birnberg,N.C., Seasholtz,A., Herbert,E. and Goodman,H.M. (1986) Nature, 323, 353-356. Datta,S., Gomer,R.H. and Firtel,R.A. (1986) Mol. Cell Biol., 6, 811-820. Dente,L., Cesareni,G. and Cortese,R. (1983) Nucleic Acids Res., 11, 1645-1655. Devine,J.M., Tsang,A.S. and Williams,J.G. (1982) Cell, 28, 793-800. Devine,K.M., Morrissey,J.H. and Loomis,W.F. (1982) Proc. Natl. Acad. Sci. USA, 79, 7361-7365. Dierks,P., Van Ooyen,A., Cochran,M.D., Robkin,K., Reiser,J. and Weissman,C. (1983) Cell, 32, 695-706. Dudler,R. and Travers,A.A. (1984) Cell, 38, 353-398. Gerisch,G., Fromm,H., Huesgen,A. and Wick,U. (1975) Nature, 255, 547-549. Goodbourn,S., Zinn,K. and Maniatis,T. (1985) Cell, 41, 509-520. Hanahan,D. (1985) Techniques for Transformation of E. coli, DNA Cloning Vol. 1: A Practical Approach. IRL Press, Ch. 6, pp. 109-135. Haribabu,B. and Dottin,R.P. (1986) Mol. Cell Biol., 6, 2402-2408. Haribabu,B., Rakjovic,A. and Dottin,R.P. (1986) Dev. Biol., 113, 436-442. Henikoff,S. (1984) Gene, 28, 351-359. Hope,I.A. and Struhl,K. (1986) Cell, 46, 885-894. Jacobson,A. (1976) In Last,J.A. (ed.), Methods in Molecular Biology 8. Marcel Dekker Inc., New York, pp. 161-209. Karin,M., Haslinger,A., Holtgreve,H., Richards,R.I., Krauter,P., Westphal,H.M. and Beato,M. (1984) Nature, 308, 513-519. Kimmel,A.R. and Firtel,R.A. (1982) In Loomis,W.F. (ed.), The Development of Dictyostelium discoideum. Academic Press, pp. 234-324. Klein,C. (1975) J. Biol. Chem., 250, 7134-7138. Kreffi,M., Voet,L., Gregg,J.H., Meirhofer,H. and Williams,K.L. (1984) EMBO J., 3, 201-206.

200

Mangiarotti,G., Lefebvre,P. and Lodish,H.F. (1981) Dev. Biol., 89, 82-91. Mason,P.J., Jones,M.B., Elkington,J.A. and Williams,J.G. (1985) EMBO J., 4, 205-211. Maxam,A. and Gilbert,W. (1980) Methods Enzymol., 65, 499-560. McKnight,S.L., Kingsbury,R.C., Spence,A. and Smith,M. (1984) Cell, 37, 253-262. Mehdy,M.C. and Firtel,R.A. (1985) Mol. Cell Biol., 5, 705-713. Mehdy,M.C., Ratner,D. and Firtel,R.A. (1983) Cell, 32, 763-771. Montminy,M.R., Sevarino,K.A., Wagner,J.A., Mandel,G. and Goodman,R.H. (1986) Proc. Natl. Acad. Sci. USA, 83, 6682-6686. Morrissey,J.H., Devine,K.M. and Loomis,W.F. (1984) Dev. Biol., 103, 414-424. Mullens,I.A., Franke,J., Kapples,D.J. and Kessin,R.H. (1984) Eur. Int. Biochem., 142, 409-415. Nellen,W. and Firtel,R.A. (1985) Gene, 39, 155-163. Nellen,W., Silan,C. and Firtel,R.A. (1984) Mol. Cell Biol., 4, 2890-2898. Norris,K., Norris,F., Christianson,L. and Fiil,N. (1983) Nucleic Acids Res., 11 5103-5112. North,M.J. (1982) Microbiol. Rev., 46, 308-340. North,M.J. and Cotter,D.A. (1984) Exp. Mycol., 8, 47-54. O'Farrell,P. (1981) Bethesda Res. Lab. Focus, 3, 1. Oohata,A. (1983) J. Embryol. Exp. Morphol., 74, 311-319. Oyama,M. and Blumberg,D.D. (1986) Proc. Natl. Acad. Sci. USA, 83 4819-4823. Pabo,C. and Sauer,R. (1984) Annu. Rev. Biochem., 53, 293-321. Pears,C.J., Mahbubani,H. and Williams,J.G. (1985) Nucleic Acids Res., 13, 8853-8866. Pelham,H.R.B. and Bienz,M. (1982) EMBO J., 1, 1473-1477. Presse,F., Bogdenovsky-Sequeval,D., Mathieu,M. and Felenbok,B. (1986) Mol. Gen. Genet., 203, 333-340. Ratner,D. and Borth,W. (1983) Exp. Cell Res., 143, 1- 13. Renkawitz,R.G., Schutz,G., Von der Ahe,D. and Beato,M. (1984) Cell, 37, 503-510. Sanger,F., Nichlen,S. and Coulson,A.R. (1977) Proc. Natl. Acad. Sci. USA, 74, 5463-5467. Schaap,P. (1983) Roux's Arch. Dev. Biol., 192, 86-94. Scheidereit,C., Geisse,S., Westphal,H.M. and Beato,M. (1983) Nature, 304, 749-752. Searle,P.F., Stuart,G.W. and Palmiter,H.D. (1985) Mol. Cell. Biol., 5, 1480-1489. Sen,R. and Baltimore,D. (1986) Cell, 46, 705-716. Struhl,K. (1985) Proc. Natl. Acad. Sci. USA, 82, 8419-8423. Stuart,G.W., Searle,P.F., Chen,H.Y., Brinster,R.L. and Palmiter,R.D. (1984) Proc. Natl. Acad. Sci. USA, 81, 7318-7322. Takeuchi,I., Hayashi,M. and Tasaka,M. (1977) Cell differentiation and pattern formation in Dictyostelium, Development and Differentiation in the Cellular Slime Mould Dictyostelium discoideum. Elsevier/North Holland Biomedical Press, pp. 1-16. Town,C.D. and Gross,J.D. (1978) Dev. Biol., 63, 412-420. Treisman,R. (1986) Cell, 46, 567-574. Tsang,A.S. and Bradbury,J.M. (1981) Exp. Cell Res., 132, 433-441. Williams,J.G., Lloyd,M.M. and Devine,J.M. (1979) Cell, 17, 903-913. Williams,J.G., Tsang,A.S. and Mahbubani,H. (1980) Proc. Natl. Acad. Sci. USA, 77, 7171-7175. Williams,J.G., North,M.J. and Mahbubani,H. (1985) EMBO J., 4, 999-1006. Yeh,R.P., Chan,F.K. and Coukell,M.B. (1978) Dev. Biol., 66, 361-374. Zoller,M.J. and Smith,M. (1982) Nucleic Acids Res., 10, 6487-6500.

Received on 20 October 1986

Note added in proof The pre-stalk cathepsin gene (Datta et al., 1986) displays an identical nucleotide sequence to the cysteine proteinase 2 gene, both within and upstream of the coding region, and is therefore certainly the same gene [Datta,S. and Firtel,R. (1987) Mol. Cell. Biol., in press]. In this same paper, these authors also present evidence to show that the G-rich elements at nucleotide -210 are essential for efficient expression of the gene (R.Firtel, personal communication).