Gene, 122 (1992) 297-304 0 1992 Elsevier Science Publishers

GENE

B.V. All rights reserved.

297

037%1119/92/$05.00

06806

Membrane ,U poly(A) transcription terminator

signal and 3’ flanking sequences for immunoglobulin-encoding genes

function

(Recombinant

heavy chain; p gene; polyadenylation;

termination)

DNA;

immunoglobulin

Nicolas J. Fasel”, Marga Rousseauxa, Nicole Dkglon”, Claude Bron a and Randolph Wall b,c

Hermann

mouse;

transcription

L. Govanb,

as a

Ronald Lawb*,

oInstitute of Biochemistry, Universityof Lausanne, 1066 Epalinges, Switzerland; b Department of Microbiology and Immunology, UCLA School of Medicine and ‘Molecular Biology Institute, Universityof California Los Angeles, Los Angeles, CA 90024, USA. Tel. (213)825-1244 Received

by J.K.C.

Knowles:

13 February

1992; Revised/Accepted:

16 February/l7

July 1992; Received

at publishers:

20 August

1992

SUMMARY

Developmentally regulated mechanisms involving alternative RNA splicing and/or polyadenylation, as well as transcription termination, are implicated in controlling the levels of secreted p (cl,), membrane p (pm) and 6 immunoglobulin (1g) heavy chain mRNAs during B cell differentiation (p gene encodes the p heavy chain). Using expression vectors constructed with genomic DNA segments composed of the ~1, polyadenylation signal region, we analyzed poly(A) site utilization and termination of transcription in stably transfected myeloma cells and in murine fibroblast L cells. We found that the gene segment containing the p, poly(A) signals, along with 536 bp of downstream flanking sequence, acted as a transcription terminator in both myeloma cells and L cell fibroblasts. Neither a 141-bp DNA fragment (which directed efficient polyadenylation at the p, site), nor the 536-bp flanking nucleotide sequence alone, were sufficient to obtain a similar regulation. This shows that the pm poly(A) region plays a central role in controlling developmentally regulated transcription termination by blocking downstream 6 gene expression. Because this gene segment exhibited the same RNA processing and termination activities in fibroblasts, it appears that these processes are not tissue-specific.

INTRODUCTION

Strict control of B lymphocyte for normal humoral responses.

development is essential During differentiation,

Correspondence to: Dr. N.J. Fasel, Institute Lausanne,

Ch. des Boveresses

Tel. (41-21)653 * Present School,

Epalinges,

University

of

Switzerland.

30 66; Fax (41-21)653-62-72.

address: Los Angeles,

Abbreviations:

of Biochemistry,

155, CH-1066

Norris

Cancer

Research

Hospital,

USC

Medical

CA 90033, USA. Tel. (213)224-65-49.

bp, base pair(s);

cat, choramphenicol

acetyl-transferase-

encoding gene; hph, gene encoding Hy phosphotransferase (Hy resistance); Hy, hygromycin; Ig, immunoglobulin; kb, kilobase or 1000 bp; p, gene encoding the p heavy chain; pm, membrane heavy chain; p,, secreted heavy chain; NP-40, nonidet P-40 detergent; nt, nucleotide(s); SDS, sodium dodecyl sulfate; SSC, 0.15 M NaCl/O.OlS M Na,.citrate pH 7.6; SV40, simian virus 40.

pre-B cells successively rearrange their Ig heavy chain and light chain genes to synthesize membrane-bound IgM on immature B cells. At the mature B cell stage, a new class of Ig, IgD, is present along with the IgM at the cell surface. Upon activation, B cells differentiate to plasma cells which turn off IgD synthesis and actively secrete pentameric IgM. The pm, ps and 6, mRNAs which code for the membrane and secreted heavy chains of IgM and membrane heavy chains of IgD, respectively, are products of a single large complex transcription unit (see Fig. 1). Control of the production of the various RNAs at different stages in B cell development appears to be achieved through a complex combination of transcriptional and post-transcriptional regulatory mechanisms (Wall and Kuehl, 1983; Rogers and Wall, 1984; Gough, 1987; Brown and Morrison, 1989). In terminally differentiated IgM-secreting cells, it appears that post-transcriptional processing mechanisms contribute

298

pLsP(A) I

0 Ml

iu/sgene

p,,, mRNA

H

M2

V

H

M3

V

M4

ml

P,,, P(A) 0 m2 a

II

V

b

IA to delta gene k

6 mRNA

Fig. 1. Schematic

representation

of the w-6 transcriptional

by lines. The two specific polyadenylation in murine and human restriction

are represented

by circles corresponding

The /1 exons are represented

unit and the respective

spond to sequences

by the letters a and b. The DNA segments

position

of their specific polyadenylation

specific to ps or p,,, mRNA.

ml, m2, membrane

mini-exons;

p,p(A),

Abbreviations

S, M, Ml41

and dHi are defined by the positions

the three different mRNAs

site; pL,p(A), p, polyadenylation

predominantly to the increased expression of ps mRNA (Petersen and Perry, 1986; Tsurushita and Korn, 1987; Petersen et al., 1991; Watakabe et al., 1992), whereas transcription termination downstream from the pu, polyadenylation site but upstream from the ,u, site occurs only in certain myeloma cell lines (Kelley and Perry, 1986; Guise et al., 1988). Complex modes of regulation including transcription termination and post-transcriptional RNA processing choices appear to control 6 mRNA production. The IgM-secreting cells do not express detectable 6 heavy chain mRNA. Several studies now indicate that developmentally regulated transcription termination at or closely following the pm poly(A) site is the major mechanism used to shut down 6 mRNA expression in IgM-secreting cells (Mather et al., 1984; Yuan and Tucker, 1984; Fasel et al., 1986; Kelley and Perry, 1986; Galli et al., 1987; Law et al., 1987; Tish et al., 1991). We and others previously reported that transcription across the ~+6 gene in IgM-secreting cells terminated in myeloma cells over the sequences closely following the pm poly(A) site (Yuan and Tucker, 1984; Fasel et al., 1986; Kelley and Perry, 1986; Law et al., 1987; Tish et al., 1991). Our characterization of the sequences surrounding the pm poly(A) site revealed several features which suggested that this region could function as a terminator (Law et al., 1987). In the study reported here, we have used expression vectors containing minimal p gene segments required for polyadenylation to directly determine the intrinsic activity of the pm poly(A) region in controlling selective poly(A) addition and transcription termination. We

of specific

(pL,, pm and a) of this

sites (A). The open boxes at the end of the respective

are as follows: A, poly(A) tail; I, transcription

p. polyadenylation

by open boxes and the introns

to p. poly(A) and pL, poly(A). The two DNA elements conserved

sites: P, PstI, HIII, Hind111 and X, XbaI. In the lower part of the figure are represented

transcription domains;

genomes

unit and specific DNA segments.

sites are represented

transcripts

corre-

start point; Ml, M2, M3, M4, p constant

site; VDJ, variable

domain.

directly show that a 677-bp gene segment containing the pcl, poly(A) signals and 536 bp of 3’ flanking sequence effectively blocked transcription in myeloma cells and in fibroblast cells. DNA segments containing the ,u, poly(A) signals as well as downstream sequences are required for transcription termination. These findings indicate also that the processes in transcription termination which operate in terminally differentiated myeloma cells also occur in nonlymphoid cells.

RESULTS

AND DISCUSSION

(a) Expression vector constructs containing p gene M and S segments We prepared a series of expression vector constructs containing different 1g p gene segments (Fig. 1) encompassing the p”s(i.e., S) or pm (i.e., M or M141) polyadenylation signals. The S segment contained part of the Cp4 exon (including the splice donor site for pL, exon splicing) along with the 11, poly(A) site plus 206 bp of downstream sequence. The M segment contained ~~2 exon (lacking the splice acceptor site for Cp4 splicing) with the pL, poly(A) site plus 536 bp of downstream sequence. The Ml41 segment of the M fragment contained the pm poly(A) site flanked by 93 bp of upstream and 48 bp of downstream sequence. The dHi DNA fragment corresponds to an M DNA where the Ml41 segment has been removed. The expression vectors consisted of the early region SV40 promoter linked to the complete coding sequence of the hph

A

p-M

gene followed by ps and/or /Jo polyadenylation

s

sv 40 tl : 1.6kb A

t2:

2.5kb

t3:

2.4kb

p-M / S IHi

“A

B

a

b

c

c

have deleted their ~1genes or into mouse L fibroblast cells. Stably transfected cells were selected in Hy. Identical results were obtained with individual Hy-resistant cell clones or pools of Hy-resistant cells.

abc

(b) The M segment blocks polyadenylation

Fig. 2. Analysis

by Northern

cells transfected

with p-M/S

Structure

blot of cytoplasmic and p-M/SdHi

of the DNA constructs

(Madison,

RNA extracted

generated

from

(Panel A)

DNA constructs.

in Sp65 vector from Promega

WI). We used the SV40 early region promoter

comprised

be-

tween the PvuII site (nt 275) and the Hind111 site (nt 5173) (Reddy et al., 1978) and the Hy-resistance pLG88

(Gritz

gene as a BamHI

and Davies,

16-kb BarnHI-EcoRI clone in 1, Charon

restriction

fragment

4A (Blattner

(R. Deans, unpublished

fragment

1983). The p segments

results).

issued

are derived

of a BALB/c

and Tucker,

from from a

murine genomic

1984) subcloned

into pUC19

The murine S segment has a size of 610

bp from the PstI site present

in the C&4 exon to the Hind111 site down-

stream

site (Blattner

from the ps poly(A)

segment

(65 1 bp) corresponds

from nt 2255-2906

and Tucker,

to the XbaI-Hind111

in the reported

sequence

1984). The M

restriction

(Richards

fragment

et al., 1983). dHi

is a deletion mutant Ml41 was removed.

of M where the HincII fragment corresponding to Detailed maps are available upon request. Mouse

Sp2/0

line and L fibroblasts

B lymphoma

modified Eagle’s medium supplemented BRL),

1 mM L-glutamine

were transfected

Systems

protocol

et al., 1984). Stable lines were obtained

electronic as described

L cells were transfected

cipitation

and selection

presence

per ml. Sp2/0 cells cell processing previously

by selection in presence

Hy/ml (Calbiochem). technique,

in Dulbecco’s

with 10% fetal calf serum (Gibco,

and 10 pg of gentamycin

using a D.E.P.

trifuge or the DEAE-dextran

were grown

of 200 pg Hy/ml. (Panel B) Northern

clones

cen-

(Deans of 900 pg

by the Ca.phosphate

of Hy-resistant

pre-

was done in

phoresis on a 0.8% agarose gel in 10 mM Na.phosphate pH 6.5 (MacMaster and Carmichael, 1977). The samples were transferred to Genescreen Plus filters (DuPont-NEN) with a vacugene apparatus (Pharmacia/ LKB). Antisense RNA probes were generated by cloning DNA fragments S, M, Ml41 and hph into the vectors pGem-1 and pGem-2 (Promega, Madison, WI) and Sp6 or T7 RNA polymerase reactions were carried out according to Promega protocol. The filters were hybridized for 48 h at 55°C in a solution containing 50% formamide/ x SSC/ mM Na.phosphate

at a downstream

site In these experiments, we analyzed the effect of placing the p, poly(A) site in front of the p., poly(A) site. The two DNA constructs analyzed, p-M/S and p-M/SdHi, are shown in Fig. 2A. The dHi deletion in p-M/SdHi removes a 141-bp segment containing all the signals required for /Jo polyadenylation. Cytoplasmic RNA was prepared from transfected and untransfected Sp2/0 and analyzed in Northern blots (Fig. 2B). A single mRNA of 1.6-kb (tl) corresponding to an RNA polyadenylated at the pm site was detected in p-M/S transfected Sp2/0 cells (Fig. 2B, lane b). No RNA containing secreted p sequence was detected when the transfected cell RNA was hybridized with a ps probe (data not shown). A single mRNA of 2.4-kb (t3) was seen in p-M/SdHi transfected Sp2/0 myeloma cells (Fig. 2B, lane a). From the size of this polyadenylated RNA, we can conclude that the ps poly(A) site in this construct is functional. Therefore, the absence of the 2.4-kb transcript (t3) in p-M/S-transfected cells is not due to a defect in pL,poly(A) signal or to the lack of any truns-acting factor specific for ps. No hph-specific RNA was detected in nontransfected control myeloma cells (Fig. 2B, lane c). We also analyzed the expression of the p-M/S construct in transfected L cells. Cytoplasmic RNA from transfected and untransfected L cells was analyzed in Northern blots (Fig. 2C). No hph-specific RNA was detected in untrans-

blot analysis of transfected

Sp2/0 cells. Cytoplasmic RNA was isolated according to the NP40 lysis method. RNA (25 pg) was treated with glyoxal and subjected to electro-

2 x Denhardt’s/SO

signals. The

/Z./Z gene, isolated from the pLG88 vector (Gritz and Davies, 1983), confers Hy resistance in Sacchavomyces cerevisae (Gritz and Davies, 1983) and higher eukaryotes (Bloechlinger and Diggelmann, 1984). The structures of the . relevant construct(s) are included in the respective figures. After linearization using a unique restriction site present in the plasmid sequence, expression vector constructs were stably introduced into Sp2/0 plasmocytoma cells which

pH 6.5/0.2x

SDS/l

mM EDTA/

250 pg/ml denatured herring sperm DNA 500 pg/ml yeast RNA. filters were washed 5 x 15 min in 0.1 x SSC/O.l% SDS at 65°C

The and

exposed at -70°C on a X-AR Kodak film with an intensifier screen (Kronex). The sizes of the detected transcripts were measured by comparison to the migration

ofthe endogenous

28s and 18s ribosomal

RNAs.

The samples were loaded as follows: Sp2/0 cells expressing p-M/SdHi (lane a); Sp2/0 cells transfected by p-M/S (lane b); nontransfected Sp2/0 cells (lane c). (Panel C) Comparison of the expression of p-M/S in Sp2jO and L cells. The samples were loaded as follows: nontransfected L cells (lane a); Sp2/0 cells transfected by p-M/S (lane b); L cells transfected by p-M/S

(lane c). Abbreviations

and symbols

in the figure are as follows:

SV40, SV40 DNA promoter early region; hph, Hy-resistance gene; M, M segment; S, S segment; circle, polyadenylation site. The size of the different mRNAs (tl,t2 and t3) is given in kb.

300 fected L cells (Fig. 2C, lane a). A 1.6-kb mRNA (tl) resulting from polyadenylation at the II, site was detected in L cells transfected with the p-M/S construct (Fig. 2C, lane c). This species co-migrated with the 1.6-kb mRNA seen in p-M/S-transfected Sp2/0 myeloma cells (Fig. 2C, lane b). In some clones but not in all, the Sp2/0 cells also contained a smaller RNA species designated tX (cf. Fig. 5B, lane a and Fig. 5C, lane a) never observed in L cells (Fig. 5D, lane d). From the sizes and properties of the tX poly(A)-containing RNA species seen here and in some other blots, it is likely that these variant RNAs result from processing at a cryptic splice in the hph sequence. It is not

(c) The M segment terminates transcription Selective usage of pm poly(A) site in the p-M/S

clear why this site is utilized in Sp2jO cell clones but not in L cells. The presence of these aberrant tX species is independent of the construction used and therefore does not affect the interpretation of the results. The level of polyadenylated tl RNA was reproducibly higher in transfected Sp2/0 cells than in transfected L cells. We have not determined if this difference resulted from higher integrated copy number in the myeloma cells. As in transfected Sp2/0 myeloma cells, L cells transfected with p-M/S did not contain any RNA (t2) resulting from polyadenylation at the downstream ps site.

A

p-cat-4

II

Sp2 / Onuclei

B

F

D

Fig. 3. Analysis

vector

could result either from preferential usage of the first polyadenylation signal encountered in a complex transcription unit or from termination of transcription by the p, poly(A) signals and downstream sequences in the M segment. To determine if this result was due to transcription termination, we carried out run-on transcription assays on nuclei isolated from cells tranfected with a vector in which the M segment was inserted between hph and the bacterial cat gene (Fig. 3A). These gene segments could be used as probes to avoid ambiguities resulting from the presence of eukaryotic repetitive sequences in the p gene. Labelled RNA from transcription assays in nuclei isolated from either p-cat-4-transfected Sp2/0 cells or L cells was hybridized to a purified 330-bp DNA fragment from the hph gene (probe I) or to a purified 330-bp fragment from the cat gene (probe II; Fig. 3A). Hybridization to the hph segment (probe I) was 3 to 4-fold stronger than that to the cat segment (probe II) at two different DNA concentrations (Fig. 3B) as quantified by densitometry scanning. Direct comparison of the observed intensity of hybridization and polymerase II loading is possible since both probe fragments were the same

5pshPhI

-

5 pg cat II

-

of the effect of the M segment on transcriptional

lpghph

l~gcatI1

I

5 fig

cat

5pgmp8

-

1 pg

-

1 fig

cat

mp8

activity. Detailed maps are available upon request. (Panel A) Structure

of DNA p-cat-4

construct and position of the probes in the transcription unit. (Panel B) Nascent RNA chains were elongated in vitro in isolated nuclei from Sp2jO cells transfected with p-cat-4 and hybridized to 5 pg or 1 pg of specific restriction fragments (DNA probe I or probe II) fixed on a nylon filter. (Panel C) In vitro elongated RNA isolated from L cells transfected with p-cat-4 hybridized to 5 pg of probe I or probe II. (Panel D) Nascent RNA chains were elongated in vitro in isolated nuclei from Sp2/0 cells transfected with p-cat-4 construct or cat gene. (Panel E) As in (D) but hybridization to single-stranded DNAs,

and hybridized to 5 pg or 1 pg of single stranded DNA containing M13mp8 or to M13mp8 containing the cat gene.

the hph

301 length. Similarly, labelled RNA was hybridized to singlestranded M13mp8 DNAs containing the hph or cat gene

A

p-Ml41

ISiM Ml41 s I v///

hph

sv40

segments to exclude that the observed signals are due to wrong strand transcription. As shown in Fig. 3D, hybridization of labelled RNA to the hph gene probe is fivefold stronger than that to the cat gene probe as measured by densitometry scanning. Two different DNA concentrations (1 and 5 pg) were also used to ensure that hybridization was performed in DNA excess conditions. When M13mp8 single-stranded DNA is used as a background control, hybridization is fourfold weaker than that of the cat gene probe as shown in Fig. 3E, thus only 5% of the signal can be attributed to background hybridization. These experiments show for the first time that the M segment (composed of the p, polyadenylation signals and 543 bp of downstream flanking sequence) can block polymerase II progression along a transcription unit. Furthermore, this termination can occur out of the normal context in the ~+6 complex transcription unit. The transcription termination seen with the isolated M segment occurs over a relatively short stretch of DNA as it has been previously reported in certain myeloma cell lines (Kelley and Perry, 1986; Guise et al., 1988). Thus, the isolated M segment is able to reproduce a pattern of termination seen in terminally differentiated B lineage cells (Yuan and Tucker, 1984; Fasel et al., 1986; Kelley and Perry, 1986; Galli et al., 1987; Law et al., 1987; Guise et al., 1988; Tish et al., 1991). The transcription terminating activity of the isolated M segment is apparently not restricted to myeloma cells, since the same results were obtained in L cells where transcription of the downstream cat test gene segment was reduced by the same factor (Fig. 3C). (d) M segment sequences sufficient for membrane p polyadenylation do not block downstream gene expression We next carried out transfection experiments to see if a short DNA fragment (i.e., 141 bp long) containing all the signals for p”m polyadenylation could block expression of downstream sequences (Fig. 4A). Two findings suggested that this segment might terminate transcription. First, we showed that deletion of the Ml41 fragment in the vector, p-M/SdHi (Fig. 2A), allowed expression of polyadenylated RNA from a downstream gene segment (Fig. 2B). Second, the Ml41 DNA fragment has been previously found to contain an unusual nuclear factor binding site with the properties expected for eukaryotic termination signals which function in an orientation-specific and directional manner (Law et al., 1987). Sp2/0 myeloma cells were transfected with the p-M141/ S/M vector in which the 141-bp fragment containing the pL, poly(A) signals was inserted before an S segment followed by a complete M segment (Fig. 4A). Cytoplasmic RNA was analysed by Northern blot to determine which poly-

%a

M

6

d

tl : 1.55kb

A A

B

t2:

2.05kb

t3:

2.35kb

a

S probe S

probe Fig. 4. Analysis

by Northern

cells transfected

with p-M141/S/M.

segment as an H&II (Panel A) Structure

fragment

blot of cytoplasmic

RNA extracted

Ml41 fragment

from nt 2281-2422

of the DNA construct

is contained (Richards

from

in the M

et al., 1983).

introduced

into Sp2/0 (panel

B) or in L cells (panel C). The samples were separated

on a 0.8% agarose

gel in 10 mM Na.phosphate and the Northern M141-specific

pH 6.5 (MacMaster

blots were probed

riboprobe.

from Sp2/0 transfected

Samples

and Carmichael,

1977),

using either an hph-, an S- or an

are loaded as follows: (panel B) RNA

with p-M141/S/M

probe (lane a) or to an S (lane b) anti-sense

hybridized probe;

probed

to hph anti-sense

(panel C) RNA from

with Ml41

(lane a) or pL,

L cells transfected

with p-M141/S/M

(lane b) anti-sense

RNA probe. The size of the different mRNAs

(tl, t2

and t3) is given in kb.

adenylation sites were used in generating polyadenylated mRNA. The results in Fig. 4B (lane a) show the three hph-specific RNA species (tl, t2, t3) suggesting that all three poly(A) sites were used as confirmed when the blots are analysed with an hph or an M141-specific probe (data not shown). The tX RNA species was also observed. Only the t2 and t3 RNA species hybridized with an S probe thereby confirming the structures of these species (Fig. 4B, lane b). The expression of RNA species polyadenylated at the S and M segment poly(A) sites (located downstream from M141) clearly indicates that expression of downstream segments is not blocked by the Ml41 fragment. This shows that these minimal p, poly(A) signals are not sufficient to obtain the steady-state RNA pattern observed when a complete M segment is present. The results (Fig. 2) showed that deletion of the M 14 1 sequences from the complete M segment restored efficient utilization of the poly(A) site in the downstream S segment (i.e., p-M/SdHi). This finding indicates that the M segment sequences downstream from the pm poly(A) signals in Ml41 also are not capable of blocking transcription. Although we cannot formally exclude that the different steady-state RNA pattern is influenced by differences in RNA stability, we tend to

302 favour a model where polymerase when the complete M segment

II loading is affected only is present. Accordingly,

both the ,u~ poly(A) signals and the downstream flanking sequence in the M segment would be necessary elements for diminishing downstream DNA expression. The sequences in the Ml41 segment are clearly capable of directing highly efficient polyadenylation at the pm site. Polyadenylated tl RNA is the most abundant species seen in these experiments (Fig. 4A, lane a). The t3 RNA polyadenylated at the third poly(A) site (i.e., in M) was second in abundance. The t2 RNA polyadenylated at the ,u~poly(A) site in the S segment was the least abundant of the three RNA species. These results indicate that the different poly(A) sites in this vector are used with quite different efficiencies. The most distal poly(A) site (i.e., present in the M segment) is used with higher efficiency than the preceding poly(A) site (i.e., located in the S segment) in agreement with results shown in Fig. 5B. Both results indicate that the poly(A) signals and other sequences comprising the S segment do not block expression of a downstream DNA fragment. Expression of RNA species from the p-M141/S/M vector was also analysed in L cells. The same pattern of polyadenylated RNA species was seen in L cells and in Sp2/0 myeloma cells (Fig. 4C, lanes a and b). Polyadenylated

A

p-S i M

tl t2:

B

a

Fig. 5. Analysis

:

1.9 kb 2.2kb

b

by Northern

blot of the expression

of p-S/M in Sp2jO or

in L cells. (Panel A) DNA constructs and expected mRNAs (tl, t2) polyadenylated at specific sites in the p-S/M construct. The different segments used in these constructs have been described earlier. Abbreviations and symbols in the figure are as follows: hph, Hy-resistanceencoding gene; M, M segment; S, S segment; SV40, SV40 DNA promoter early region; circle, polyadenylation site. The size of the different mRNAs separated on a 0.8% agarose gel (tl, t2) is given in kb. (Panel B) Northern blot analysis of Sp2/0 cells (lane a) or L cells (lane b) transfected with p-S/M. The same amount (25 pg) of RNA was loaded in each lane.

RNA species (tl, t2, t3) were present at approximately the same ratio as in Sp2/0 myeloma cells. These findings certainly suggest that the regulatory mechanisms which direct poly(A) site selection and transcription termination in terminally differentiated B lineage cells are also operative in fibroblasts. (e) The S segment does not affect downstream gene expression We next analysed S and M RNA species produced in myeloma cells or L cells transfected with vector constructs in which the ps site was separated from the downstream pm site by 0.3 kb. Both Sp2/0 cells and L cells transfected with p-S/M contained two polyadenylated RNA species, tl (i.e., S) and t2 (i.e., M) which hybridized with the /@z-specific probe (Fig. 5B, lanes a and b, also Fig. 5D, lane a). In both cells, the t2 species was at least two to three times more abundant than the tl species indicating preferential usage of the p, over the ps polyadenylation signals with this construct. Preferential usage of the second polyadenylation site is observed providing good evidence that the RNA polymerase II does not stop in the S segment. (f) The p,,, poly(A) region contains two sequences (box A and box B) conserved between mouse and human The pm sequences and downstream flanking sequences comprising the M segment contain multiple features which could function in transcription termination. A protein factor binding site which includes an unusual palindromic sequence predicted to have orientation-dependent activity has been defined immediately upstream from the AATAAA pm poly(A) signal in the Ml41 segment (Law et al., 1987). The DNA directly downstream from the pm poly(A) site contains two noncoding sequences which are highly conserved (i.e., box A and box B, a and b, Fig. 6) (Fasel et al., 1986) in location and sequence (> 80% identical) between murine (Richards et al., 1983) and human (Milstein et al., 1984) p genes. Such high levels of sequence conservation in noncoding sequences is characteristic of important regulatory elements. The conserved sequence designated box A in Fig. 6 contains the sequence, AAACCTTTCCTGC, which is also found downstream from the Ig heavy chain gene a, poly(A) site. It will be interesting to test the termination activity of these conserved DNA elements in conjunction with the M 14 1 segment containing the pL, poly(A) signals. It also would be interesting to know if isolated elements from the Ml41 segment cannot function in termination without a functional poly(A) site as it was tested for SV40 (Connely and Manley, 1988). The discoveries described here provide the foundation for definitive studies to identify the functional elements in the ~1, segment and flanking region and to characterize their contributions to transcription termination.

303

ml

m2 (GG=W 3'UT II XHi

P

(=) 15

26

I Hi

0

HI11

Box A

Box B

----ACTTTATTGTGAAGGAATTTTGTTTTGTTTTTC~CCTTTCCTGC----//----CCTGATGGAAG:AAGGGAAGTAGGGCAG:AGAAAATTCCAGGCCT----

Mouse

I

I

4214

4318

I 465:

4697

--__G_---C_G__--G_-TT--___--C~___----__A_---___-----_,,____-______---_G_----G_C_~_A--__G__-GGG_________--___ I I I 3029 283; 2986 2789

Human

B Human Pm

AAAACTTTCCTGC

nouse pm

AAACCTTTCCTGC

Mouse

a,

GAACCTTTCCC

Fig. 6. Conserved

GC

A I 1939

I

1926 sequences

in the pL, polyadenylation

(Box A and Box B) conserved

between

tive elements (GGGAGA),,(AG),,.

murine

et al., 1980). The numbering

of the murine

(Box A) is located

the numbering 3’ untranslated present

sequence

sequences

sequence;

in the human

Hi, HincII

follows the designations

line indicates

of the II, poly(A) region containing

nt identity.

in the National

p, AATAAA Abbreviations

Biomedical

sequence.

region and in the & region. The numbering

Research

Foundation of the human

are as follows: ml, ml mini-exon;

of the murine s

sequence

by two repeti-

to nt 4274 (Early et al., 1980; Rogers

The numbering

site; HIII, Hind111 site; P, PstI site; X, XbaI site. (Panel B) Representation

and the murine k

the two sequences

the mini exons ml and m2 with their 3’UT followed

is re p resented by a circle and corresponds

25 nt 3’ of the postulated

in Milstein et al. (1984). The dashed

representation

Open boxes represent

The murine poly(A) site [pL, poly(A)]

TGCMD).

Human

region. (Panel A) Shematic

and human.

database

(reference

sequence

represents

m2, m2 mini-exon;

of the three homologous represents

the numbering

3’UT,

nt sequences in Sitia et al.

(1985)

(g) Conclusions In these studies we have addressed the role of the p, poly(A) site region in regulating selective transcription termination in Ig-secreting myeloma cells and in the L cell fibroblast line to determine the lymphoid specifity of this process. Our findings indicate that the pL, poly(A) region plays a central role in terminating ongoing transcription when linked to the 536-bp downstream flanking DNA segment. This is the first experimental demonstration of transcription termination activity mediated by this gene segment outside of its natural environment. The decreased polymerase II loading produced by this segment was almost comparable to that seen over these sequences (i.e., SO-90% reduction) of the endogenous ~+6 gene in myeloma or hybridoma cells (Yuan and Tucker, 1984; Fasel et al., 1986; Kelley and Perry, 1986). Interestingly, the M segment exhibited identical termination activity in transfected L cell fibroblasts, strongly suggesting that these mechanisms are not tissue-specific. We also demonstrated that the downstream M segment sequences or the elements in the vicinity of the pm segment (M141) are not sufficient

to reproduce the M segment effect. Identical results were seen in L cells transfected with this construct. These results strongly suggest that functional elements in both gene segments are required for the processes in polymerase II down-loading. Accordingly, this suggests that 6 gene transcription at earlier stages in B cell development is due to an anti-termination mechanism which allows at least partial read-through the pL, and downstream segment.

ACKNOWLEDGEMENTS

We are indebted to Dr. R.J. Deans for helpful discussions, Dr. H. Diggelmann for providing the hph gene and Z. Freiwald for help in preparing the manuscript. This work was supported by grants from the Swiss National Fund for Scientific Research to C.B. (Grant No. 3126320.89), from the Swiss Cancer League to N.F. (FOR.383.90.2) and by grants (Nos. CA12800 and GM40185) from the United States National Institutes of Health to R.W.

304 REFERENCES

regulation

of immunoglobulin

during B-lymphocyte Blattner, F.R. and Tucker, P.W.: The molecular bulin D. Nature 307 (1984) 417-422. Bloechlinger,

K. and Diggelmann,

as a selectable eucaryotic Brown,

marker

H.: Hygromycin

for DNA

transfer

B phosphotransferase

experiments

tory and membrane Clin. Immunol.

S.L.: Regulation

with higher

immunoglobulin

Immunopathol.

during lymphocyte

nal is required

of secre-

development.

for transcription

mRNA

termination

sig-

by RNA polymerase

II.

Genes Dev. 2 (1988) 440-452. Deans,

R.J., Denis, K., Taylor,

munoglobulin

A. and Wall, R.: Expression

of an im-

Proc. Natl. Acad.

Sci. USA 81 (1984) 1292-1296. Hood,

J., Davis,

M., &lame,

L.: Two mRNAs

K., Bond,

with different

M., Wall, R. and

3’ ends encode

bound and secreted forms of immunoglobulin

membrane

mu chain. Cell 20 (1980)

C., Govan,

H., Hermanson,

and Wall, R.: Control of p+6 gene expression opment.

In: Szentivary,

Antibodies,

Structure,

vention in Disease.

A., Maurer, Synthesis,

Plenum,

Galli, G., Guise, J., McDevitt, tive position termination

and strengths

G., Law, R.

in B lymphocyte

P.H. and Janicki,

Function

devel-

B.W (Eds.),

and Immunologic

Inter-

M., Tucker, of poly(A)

are critical to membrane

sion during B cell development.

P.W. and Nevins, J.: Rela-

and Perry,

message.

Trends Genet.

188. Guise, J.W., Lim, P.K., Yuan, D. and Tucker, P.W.: Alternative and membrane

by transcriptional

forms of immunoglobulin termination

opment.

in stable

expres-

p chain is

mRNA

production

and post-transcritional

con-

during lymphocyte

devel-

Natl. Acad.

brane polyadenylation signal, possible role in transcription tion. Proc. Natl. Acad. Sci. USA 84 (1987) 8160-8164. Haimovich,

J. and Perry,

R.P.:

terminaMode

B., Dhar, R., Subramanian,

P.K., Celma, M.L. and Weissman,

Sci. USA 83

of

N., &in,

B.S.,

S.M.: Thegenome

J.E., Gilliam, A.C., Shen, A., Tucker, P.W. and Blattner,

Unusual

sequences

region. Nature

in the murine

immunoglobulin

F.R.:

p-6 heavy chain

306 (1983) 483-487.

J., Early, P., Carter,

C., &lame,

K., Bond,

M., Hood,

L. and

Wall, R.: Two RNAs with different 3’ ends encode membrane-bound mu chain. Cell 20 (1980) 303-

312. J. and Wall, R.: Immunoglobulin

lymphocyte

differentiation.

RNA

Adv. Immunol.

A., Kikutani,

rearrangements

in B

35 (1984) 39-59.

H., Hammerling,

U. and Stavnezer,

J.:

The regulation of membrane-bound and secreted CIchain biosynthesis during the differentiation of the B cell lymphoma 1.29. J. Immunol. 135 (1985) 2859-2864. N. and Hozumi,

N.: Parameters

lation of delta heavy chain gene expression.

that govern the regt-

Mol. Cell. Biol. 10 (1991)

5340-5348. N. and Korn,

J.L.:

Effects of intron

of mouse 1~heavy chain mRNA.

length on differential

Mol. Cell. Biol. 7 (1987)

2602-2605. Wall, R. and Kuehl, M.: Biosynthesis Annu. Rev. Immunol.

Law, R., Kuwabara, M., Briskin, M., Fasel, N., Hermanson, G., Sigman, S. and Wall, R.: Protein binding site at the immunoglobulin m mem-

K.J.,

Proc.

of simian virus 40. Science 200 (1978) 494-498. Richards,

processing

Nucleic Acids Res. 14 (1986) 5431-5447.

E.L., Nelson,

intron.

of pm and ~LS

sites and is dependent

2321. Reddy, V.B., Thimmappaya,

Tsurushita,

plasmacytoma

140 (1988) 3988-3998.

Kelley, D.E. and Perry, R.P.: Transcriptional trol of immunoglobulin

production

by an increase in cleavage-polyadenylation efficiency but no measurable change in splicing efficiency. Mol. Cell. Biol. 11 (1991) 2324-

Tish, R., Kondo,

J. Immunol.

Regulated

Petersen, M.L., Gimmi, E.R. and Perry, R.P.: The developmentally regulated shift from membrane to secreted jr mRNA is accompanied

Sitia, R., Rubartelli,

in Escherichiu coli and Sacchavomyces cerevisae. Gene 25 (1983) 179-

transfectants.

R.P.:

on the length of the ps-pm

Rogers,

Genes Dev. 1 (1987) 471-481.

3 (1987) 238-239.

Mather,

M.L.

and secreted forms of immunoglobulin

sites as well as transcription versus secreted p chain expres-

Gritz, L. and Davies, J.: Plasmid encoded hygromycin B resistance: the sequence of hygromycin B phosphotransferase gene and its expression

sion of secreted

E.V. and Rabbits, J.M.: The sequence ofhuman

mRNA requires linkage of the poly(A) addition

Rogers,

New York, 1986, pp. 35-46.

Gough, N.: Putting a stop to an immunoglobulin

regulated

Sci. USA 74 (1977)

p-6 intron reveals possible vestigial switch segments.

Pan, J., Ghosh,

313-319. Fasel, N., Briskin, M., Carter,

Proc. Natl. Acad.

(1986) 8883-8887.

heavy chain gene into lymphocytes.

Early, P., Rogers,

immunoglobulin Petersen,

polyadenylation

varies

Nucleic Acids Res. 12 (1984) 6523-6533.

50 (1989) 155-170.

S. and Manley, J.L.: A functional

glyoxal and acridine-orange. 4X35-4838. Milstein, C.P., Deverson,

of the production

expression

Cell 36 (1984) 329-338.

MacMaster, G. and Carmichael, G.G.: Analysis of single and double strand nucleic acids on polyacrylamide and agarose gels by using

cells. Mol. Cell. Biol. 4 (1984) 2929-2931.

S.L. and Morrison,

Connely,

biology of immunoglo-

mu- and delta-chain

maturation.

Watakabe, ternative

A., Sakamoto, exon sequence

and regulation

of immunoglobulins.

1 (1983) 393-422. H. and Shimura,

Y.: Repositioning

of mouse IgM pre-mRNA

activates

of an alsplicing

of the preceding intron. Gene Express. 3 (1992) 175-184. Yuan, D. and Tucker, P.: Transcriptional regulation of p-6 heavy chain in normal murine B lymphocytes.

J. Exp. Med. 160 (1984) 564-576.