Cell Biology and Toxicology, Vol. 8, No. 2, 1992

139

EXPRESSION OF HEAT SHOCK-REGULATED HUMAN GROWTH HORMONE GENES CONTAINING OR LACKING INTRONS BY NIH-3T3 AND WISH CELL LINES SAMI ALOUANI*, PHILIPPE L'HOTE*, JEAN-BAPTISTE MARQ*, LOUIS-MARIE HOUDEBINE**, FRI~DI~RIC MONTANDON +, MARTINE C H E S S E B E U F - P A D I E U +, AND MICHEL DREANO* *Department of Genetic Engineering IntraCell S.A., Geneva-Carouge, Switzerland **Unit~ de Diff~renciation Cellulaire, INRA, Jouy-en-Josas, France +Laboratoire de Biochimie M~dicale, Facult~ de M~decine, Dijon, France

A plasmid containing the complete genomic DNA of the human growth hormone (ghGH) comprising four introns and driven by the human promoter of the human gene of the 70 kDa heat shock protein (hsp70) has been used to transfect mouse NIH-3T3 and human Wish cells. Selected cell lines were characterized for stable hGH secretion. Similarly in the same NIH-3T3 cells, the stable expression of the same plasmid construct, but containing the complementary DNA of the hGH gene (chGH), was compared in terms of the effect of introns on heterologous protein synthesis. Genomic hGH recombined cells synthetized, in a heat regulated fashion, matured hsp70/hGH hybrid mRNA able to drive the secretion of a 22 kDa polypeptide. Like the natural hGH, this polypeptide expressed the functional hormonal activity of prolactin on casein secretion by mammary cells. The time course of hGH secretion was prolonged in ghGH transcripts, while that of mRNA degradation appeared delayed, especially in Wish cells, as compared to chGH expression. In the human Wish cells the decay of endogenous hsp mRNA has been compared to that of recombinant hsp mRNA, demonstrating that this human hsp70/hGH hybrid mRNA was present in the cytoplasm during a longer period than the human endogenous hsp70 mRNA. In conclusion, similar levels of expression and resulting gene products were expressed from the chGH or the ghGH gene in an inducible manner. INTRODUCTION The interactions of introns in the course of gene expression in mammalian cells remain a controversial subject. Indeed, early studies using a series of recombinant Simian virus 40/mouse 1. Address all correspondence to: Martine Chessebeuf-Padieu, Laboratoire de Biochimie M6dicale, Facult~ de M~decine, 7, bd. Jean D'Arc, F-21033, Dijon, France. 2. Key words: hGH cDNA gene, hGH genomic gene, heat shock promoter, recombinant cell, mRNA expression, hGH secretion. 3. Abbreviations: chGH, hGH cDNA gene; ghGH, hGH, genomic gene; hGI-I, human growth hormone; hsp70, 70 kDa heat shock protein gene. Cell Biology and Copyright © 1992

Toxicology, Vol. 8, No. 2, pp. Princeton Scientific Publishing ISSN: 0742-2091

139-156

Co.,

Inc.

140

Alouani et al.

[$-globin hybrid gene clearly showed that splice sites played a role in the stability of recombinant transcripts and thus in mRNA accumulation (Hamer and Leder, 1979; Hamer et al., 1979; Gruss et al., 1979; Gruss and Khoury, 1980). It has also been observed that differential splices from identical primary Iranscripts resulted in the synthesis of protein variants with different properties (Breitbart et al., 1985). Furthermore, some introns contain regulatory sequences, for example, a glucocorticoid responsive element located in the first intron of genomic sequence of the human growth hormone (ghGH) (Moore et al., 1985; Slater et al., 1985) or a tissue-specific cellular enhancer within an intron of the immunoglobulin heavy chain gene (Banerji et al., 1983; Gillies et al., 1983; Picard and Schaffner, 1984). One pivotal function of introns is their capability to be alternatively spliced, leading from one single nucleotide sequence to the synthesis of proteins which differ in both cellular compartmentalization and function, as illustrated by transmembrane or soluble luteinizing hormone receptors (Tsai-Morris et al., 1990). In contrast, a series of cellular genes such as bean phaseolin (Gross et al., 1987), yeast actin (Ng et al., 1985), and chicken thymidine kinase (Chee et al., 1986) function without introns. On the other hand upon exposure to slress, an elevation of temperature, cells, from bacteria to man, respond by a sudden activation of a small set of genes directing the synthesis of heat shock proteins, accompanied by the suppression of much protein synthesis, active prior to the stress (Ashburner and Bonnet, 1979; Schlesinger et al., 1982). However, most hsp genes in all organisms, including eucaryotic cells, are free of introns, although there are few exceptions, such as the Drosophila melanogaster hsp83 gene (Yost and Lindquist, 1986), the ubiquitin gene in chicken embryo fibroblasts (Bond and Schlesinger, 1986), the human hsp27 gene (Hickey et al., 1986) and the Caenorhabditis elegans small hsp genes (Kay et al., 1987) contain introns. Intervening sequences are also present in Drosophila hsp cognate genes that encode proteins homologous to Drosophila HSP70, but which are not heat inducible (Ingolia and Craig, 1982). Hsp promoters have been used for the expression of foreign recombinant genes (Nover, 1987) including complex genes such as human B hepatitis virus surface antigen (HBsAg) (Dreano et al., 1987), flounder anti-freeze proteins (Rancourt et al., 1987), tissue-plasminogen activator (Sanzo et al., 1988), or chick neuronal nicotinic acetylcholine receptor (Ballivet et al., 1988)o In addition, hsp70 control elements have also been used to express genes containing intervening sequences such as the alcohol deshydrogenase gene (Bonner et al., 1984), a c-myc gene (Wurm et al., 1986), or an immunoglobulin heavy chain gene (Cattaneo and Neuberger, 1987). It was therefore of interest to observe the expression of the human growth hormone gene, hGH, by comparing hGH that contained (ghGH) with hGH that lacked introns (chGH) under the control of the promoter of the human gene of the 70 kDa heat shock protein (hsp70) using cloned recombinant mouse and human cell lines. This paper compares: 1) the expression of the hGH genomic gene (ghGH) driven by the human hsp70 promoter in human Wish and mouse NIH-3T3 cells to that of the hGH cDNA gene (chGH) as a similar hsp70 hybrid gene in the same mouse cell line, 2) the time course synthesis of the two hybrid hsp70/hGH mRNAs, 3) endogenous and recombinant hsp mRNAs in the human cell line, and 4) the hGH protein secreted by both cell lines.

285N

+l

...

v

k

]

I-,2

[

~.

Nq/K)

P2

T~V

P2

O.SV

1-13

IrA

pSV2I_)IIFR

P2

P2

DIIFR [O.SV I

gDNA of hGH 12.1 kb)

~P2

T.SV

T.SV

II3

NEK)

P2

pSP65

P2

1-13

P2

pSP65

FIGURE 1. Analysis of DNA and RNA recombinant in 3Cgl0 and Wg6 cell lines. FIG. 1A° Schematic representation of plasmids carrying hsp70/hGH hybrid genes (represented by a bidirectional arrow) encoded either by the 0.7 kb cDNA (pl7hGHdhfr and p17hGHneo, 1st and 2nd rows) or the 2.1 kb genomic DNA (p17ghGHneo~ 3rd row) containing four introns~ A, B, C, and D. Symbols correspond to the following sequences: HS: human hsp70 promoter; T.SV: Simian Virus 40 terminator; O.SV: origin of replication of the SV40; DHFR: dihydrofolate reductase gene; NEO: neomycin-resistance gene; pSV2DHFR and pSP65: expression vectors in which the transcription units were inserted; BI: BamHI; H3: HindIII; P2: PvuII; RI: EcoRI; SI: SalIo Protected fragments are included under the plasmid pl7hGHneo. The direction of transcription of hybrid genes is shown by arrows°

p 17ghGHneo [ I-IS

128N ~--Sl

+I

\

•

hGlI

H3.SI,BI

1 ! aN ~ t I3

1~5

I 1"1

/

BI

II3,SI,BI lr2 P2BI

I13.Sl t)2

rI)NA (0.7kb)

+ 1

pl7hGIlneo

RI

p 17hGl ldhfr I[I I IS

A

t'-O

t,o

Oo

142

AIouani et al.

FIG. lB. 20 ~tg DNA from Wg6 or 3Cgl0 cells and 1 and 0.1 ng of p17ghGHneo were digested with BamHI and, after migration on 1% agarose gel, transferred on nitrocellulose membrane (Schleicher and Schtiell, BA85). Hybridization was carried out using a nick translated 0.7 kb BamHI probe from pl7hGHneo corresponding to the complete hGH cDNA gene. Positions of DNA fragment markers are indicated by an arrow. FIG. 1C. RNA from heat-treated 3Cgl0 or Wg6 cells was analyzed by S1 mapping. The probe was the 5' end labeled (40000 cpm) 760 bp EcoRI-PvuII fragment from pl7hGHneo (Dreano et al., 1988). It comprised 480 bp of 5' flanking sequences and 113 bp of the RNA leader from the human hsp70 gene, 29 bp of the pSP65 polylinker which was used to insert hGH gene sequences, and 143 bp of the hGH coding sequences. S1 protected fragments were analyzed on 0.4 mm thick, 8% acrylamide-7M urea gels, using HaelII restriction digest of pBR322 and HindlII restriction digest of bacteriophage lambda DNA as size markers.

Cell Biology and Toxicology, Vol. 8, No. 2, 1992

143

MATERIALS AND METHODS

Cell lines: Transfection, Culture, Selection and Identification. Line 6:NIH-3T3 cells were co-transfected by the CaC12 method of Graham and van der Eb (1973) with two plasmids, pl7hGHdhfr and pEJ (Dreano et al., 1986). The pl7hGHdhfr (Fig. 1A, 1st row) contains chGH, the cDNA gene, driven by the human hsp70 promoter hsp70/chGtl) and ended by the SV40 terminator (T.SV) followed by the mouse dihydrofolate reductase (DHFR) gene in the other orientation and under the control of the SV40 promoter (O.SV) (Dreano et al., 1986). The pEJ is a plasmid carrying the gly12~val activated allele of the human cellular Ha-ras oncogene (Ha-rasEJ) (Tabin et al., 1982). One of the isolated foci of Ha-ras transformed cells gave rise to a selected cell line, denoted Line 6, which was found to secrete up to 3 ~g hGH per 106 cells in a 15 hr post-incubation period at 37*C following a heat shock at 42°C during 90 min (Dreano et al., 1986; Dreano et al., 1988a). It was cultured in DMEM (Gibco-BRL, F-95051 Cergy-Pontoise) supplemented with 10% fetal bovine serum (FBS) (Gibco-BRL) at 37°C in an incubator supplied with a humidified atmosphere of air/CO2 (95:5, v/v). Line 3Cg10:NIH-3T3 cells were transfected as above but with the plasmid pl7ghGHneo only. It comprises (Figure 1A, 3rd row) a BamHI 2.1 kb fragment from the plasmid pMThGH (Palmiter et al., 1983) containing the complete ghGH gene (a gift from Palmiter to researcher L.M.H.) under the control of human hsp70 promoter and ended by the SV40 terminator (T.SV) followed by a BamHI 6.0 kb fragment of pl7hGHneo (Figure 1A, 2nd row) (Dreano et al., 1988a) containing the neomycin resistance transposon of Escherichia coli (NEO) in the other orientation and under the control of SV40 promoter (O.SV). Therefore, the resulting plasmid carries the hsp70/ghGH hybrid gene, instead of the hsp70/chGH gene in the case of pl7hGHdhfr and pl7hGHneo. Cells were grown in the same medium as Line 6 cells. A cell line, denoted 3Cgl0, was isolated from spontaneously transformed cell foci which were further cloned from cells resistant to 400 ~tg/ml geneticin (Sigma, St. Louis, MO 63178). Line Wg6: Wish cells, transformed cells from human amnion tissue (ATCC CCL25), were similarly transfected as above with the same plasmid pl7ghGHneo alone. A cell line, denoted Wg6, grown as Line 6, was selected from clones resistant to 1500 ~tg/ml geneticin. Detection of the Prolactin Activity of Recombinant hGH. Rabbit mammary explants were cultured in the presence or the absence of either prolactin (0.04 I.tM) or hGH released in the culture medium of heat treated or control cell lines. In all cases, 0.83 laM insulin and 1.4 ~tM cortisol were added to amplify the lactogenic response. After one day of treatment, the presence of 13-casein was measured in tissue homogenates by RIA (Jahn et al., 1987). Blot Analysis, Immunoprecipitation, and hGH Assay. Classical cloning and Southern blotting techniques were carried out using procedures as described in Perbal (1988). For the western blot analysis of hGH, media of hGH synthesizing

144

Alouani et al.

cells, regardless of whether they were heat-treated, were concentrated ten times by lyophilisation. Samples were analyzed by electrophoresis on 15% polyacrylamide SDS gels (Laemmli, 1970) and western blotting was carried out using rabbit anti-hGH antibodies (Dako, CH-8045 Zurich) followed by a detection using gold-protein A complexes adapted from the method of Danscher and Norgaard, 1983). Cells were labeled by incubation of either heat-treated or control cells for 20 hr at 37*C in the presence of 200 [tCi of (35S)-L-met (NEN-du Pont de Nemours, F-75334 Paris). Supematants, clarified by a 3000 g centrifugation, were incubated with antibodies and protein-A sepharose previously incubated with an unlabeled supernatant of heat-treated NIH-3T3 cells. Precipitated proteins were analyzed by electrophoresis on 15% polyacrylamide SDS gels. Concentrations of hGH were measured by ELISA (Sensibeads, Terumo Medical Corporation, B-3030 Leuven).

$1 Protection Assay The transcripts of hsp/chGH and hsp/ghGH hybrid genes were analyzed by S 1 mapping using total cellular RNA from around 106 cells. Nuclease S 1 mapping was performed according to Dierks et al. (1981) and Weaver and Weissman (1979). RESULTS

Characterization of pl7ghGHneo Recombinant Cell Lines The selected cell lines, 3Cgl0 and Wg6, which resulted from the transfection of NIH-3T3 and Wish cells, respectively, with the genomic hGH plasmid, pl7ghGHneo, were expanded and hGH secretion was assessed. To verify that the ghGH gene was integrated into the cell genome, Southern analysis was carried out using a probe corresponding to the chGH gene (Figure 1A). Cellular DNA digested with BamHI (Figure 1B) showed that 3Cgl0 and Wg6 cell lines contained a fragment which co-migrated with the 2.1 kb BamHI restriction sequence of the ghGH gene. It was estimated that cells contained approximately one copy of the hGH gene sequence per genome. Analysis of RNA was performed by an S 1 mapping technique using a specific probe, the 0.7 kb EcoRI-PvulI fragment from pl7hGHneo (R1-P2, Figure 1A, 2nd row) carrying part of the chGH gene. Figure 1C shows that a fragment of a 285 base length was protected by the total RNA of Wg6 and 3Cgl0 cells, indicating that recombinant mRNA was correctly matured at least at the level of the first intron splice. In addition, these results demonstrate that the initiation of transcription in both cases was unique and corresponded to the start site of the human hsp70. Characterization of hGH Produced by Recombinant Cell Lines The hGH production by Line 6 cells has been previously described (Dreano et al., 1986 and 1988a). It reached up to 3 Ixg of hGH per 106 cells over a 15 hr period, following a heat shock (HS) of 90 min at 43"C. The production of hGH by recombinant cell lines resulting from transfection with the chGH or ghGH gene was compared. Spent media of the three cell lines, Line 6, 3Cgl0, and Wg6, were concentrated and analyzed by western blotting (Figure 2A). The hGH that was present in the medium of the 3Cgl0 (lane 3) and Wg6 (lane 5) cells

Cell Biology and Toxicology, Vol. 8, No. 2, 1992

145

submitted to HS reacted with rabbit and anti-hGH antibodies and co-migrated by electrophoresis with the same immune complexes which were obtained in the spent medium of Line 6 ceils after heat treatment (lane 1). This confirmed that the hGH secreted by ghGH transfected cells was correctly matured.

A

-.

1

2

3

4

5

6

FIGURE 2. Characterization of hGH secreted by recombinant 3Cgl0 or Wg6 cell fines. Cells, around 3.106 in a 90 mm diameter Petri dish containing 5 ml medium, were induced by a heat shock (HS) or not (Co) and post-incubated overnight at 37°C. FIG. 2A. Growth hormone secreted by recombinant cells either submitted to heat induction (HS) in lanes 1 (Line 6), 3 (3Cgl0), and 5 (Wg6) or not (Co) in lanes 2 (Line 6), 4 (3Cgl0), and 6 (Wg6) and compared to heat shocked NIH3T3 control cells in lane (-) was analyzed by western blot. Supernatants were concentrated by lyophilization to obtain around 5 I.tg hGH/ml in heat-induced cells and 20 I.tl were analyzed on a 15% polyacrylamide SDS. Human growth hormone binds to prolactin receptors and acts as a potent lactogenic hormone. Figure 2B shows that the culture of mammary explants supplemented with insulin (I) and cortisol (C) and incubated either without prolacfin (I+C) or with spent culture medium of each untransfected cells (NIH-3T3 or Wish) submitted to HS or not (Co) and of control recombinant cells (Co) from Line 6, Wg6, or 3Cgl0 produced little or no [3-casein. However, the same mammary explant cultures incubated with the spent medium of recombinant cells from Line 6, 3Cg 10, and Wg6, previously submitted to HS, produced amounts of [3-casein that

146

AIouani et al.

were similar to the amount produced in the control mammary culture treated with prolactin (I+C+P). The [3-casein production showed the same response plateau which was reached for 0.1 I.tg/ml prolactin (Jahn et al., 1987).

B

"•0.8. ~0,6 O

~ 0.4 :~i!l

:iii

~ 0.2 d~

!'.~i

O

~00

I:

i.t.C l+C-t-P

Co

HS

NIH-3T3

ng hGltperml:

0

0

line 6

3Cgl0

Co It$ Wg6

2 240

46 252

17 272

Co

HS

' Co

HS

Co

H$

Wish

0 0

FIG. 2B. Culture media were collected and aliquots were added to the culture medium of mammary gland explants from a pregnant rabbit. One day latter, 13-casein concentration was measured in cytoplasmic extracts of mammary cells by RIA (n = 3) (Jahn et al., 1987): I, C, and P correspond to insulin (0.83 IxM), cortisol (1.4 I.tM), and prolactin (0.04 I.tM), respectively. The two first bars correspond to control mammary explants treated with prolactin (I+C÷P) or not (I+C). Values in the bottom of the Figure indicate the concentration of hGH in ng per ml of culture medium in each cell line (Co and HS), NIH-3T3, Line 6, 3Cgl0, Wg6, and Wish, respectively.

Time Course Secretion of hGH by 3CglO and Wg6 Cells after Heat Induction The optimal condition for the secretion of hGH by ghGH recombinant cell lines, 3Cgl0, and Wg6, as established over a 96 hr period during which a heat treatment at 43°C was applied for various lengths of time at the beginning, followed by post-incubation at 37 ° for the remaining time. Figure 3A shows that: 1) hGH secretion depended on the length of heat treatment, 2) maximum levels of secretion were obtained after a heat treatment of 1 hr° Figure 3A also shows that longer heat shock periods tended to significantly decrease hGH synthesis in 3Cg 10 (NI/-I-3T3) while it remained more stable in Wg6 (Wish) cells, and that amounts of hGH, 3 to 4 l.tg/106 cells/24 hr found in the spent medium of heat-treated cells were comparable to those

Cell Biology and Toxicology, Vol. 8, No. 2, 1992

147

(A) 4000

2000.

..~

f2t,

0

1

2

3

4

Heat shock (hours)

03) 6000

=

5000,

4OOO ~o o t.~i

30OO

c~

(.9 2OOO

"4" 3Cgl0

1000

0

• 0

I

20

I

40

I

60

i

80

100

lengths of post incubation at 37~C (hours) F I G U R E 3. Secretion of hGH by 3Cgl0 or Wg6 cells after different length of heat induction. Around 106 cells per 50 mm diameter Petfi dish from either 3Cgl0 or Wg6 cell lines were induced as follows: FIG. 3A. heat treatment at 43°C for the indicated times and post-incubation overnight at 37°C; FIG. 3B. heat treatment of 43°C for 1 hour and post-incubation at 37°C for the indicated length of time.

FIGURE 4. S1 analysis of mRNA in recombinant 3Cg10 or Wg6 cell lines after different lengths of time following a heat induction. Messenger RNAs from ceils, around 106 cells per 50 mm diameter Petri dish heat treated for 2 hr at 43°C, were extracted after different lengths of time of post-incubation at 37°C, as indicated in the Figure, and were analyzed by $1 mapping procedure using probes labeled at both extremities. FIG. 4A. Shows mRNA from Line 6 cells expressing the chGH gene using a EcoRI-HindlII 600 bp probe from plasmid pl7Jo (Dreano et al., 1986) and containing around 480 bp of 5' non-transcripted sequence from the hsp70 gene and the 113 bp of RNA leader (Figure 1A). FIG. 4B. Shows mRNA from 3Cgl0 ceils expressing the ghGH gene using a pl7Jo EcoRI-SalI 620 bp probe containing around 480 bp of 5' non-transcripted sequence from the hsp70 gene and the 113 bp of RNA leader and 15 bp of the pSP65 polylinker (Figure 1A). (C) Show mRNA from Wg6 cells expressing the ghGH gene using the same probe as in (B). Protected fragments are indicated by an arrowhead, and reference marker fragments, 564 and 125 bases, corresponding to HindlII restriction sequences of bacteriophage lambda DNA, by arrows.

Oo

Cell Biology and Toxicology, VoL 8, No. 2, 1992

149

F I G U R E 5. Double S1 analysis of endogenous and recombinant hsp mRNA from Wg6 recombinant ceils. RNAs from heat-treated Wg6 cells were prepared 0, 1, 3, 7, 10, and 24 hours after the end of heat induction. Double S 1 analysis of mRNA was carried out using a mixture of a specific probe for the hsp hybrid gene hsp70/hGH corresponding to a EcoRI-SalI 620 bp fragment from the plasmid pl7hGHneo and of a specific probe for the endogenous hsp70 gene corresponding to a BglII 1.7 kb fragment from the plasmid p17 (Voellmy et al., 1985)o Protected fragments are indicated by an arrowhead and reference marker fragments are indicated by arrows.

150

AIouani et al.

obtained with Line 6 cells (Dreano et al., 1986; Dreano et al., 1988a). The time course of hGH secretion by 3Cgl0 and Wg6 cells following a heat shock at 43°C during 1 hr is shown in Figure 3B. Secretion of hGH during the 96 hr period after the onset of heat treatment showed a secretion pattern under either a saturated (3Cgl0) or an increasing rate (Wg6) mode. In similar conditions, Line 6 cells, chGH gene, secreted hGH for only 24 hr (Dreano et al., 1986).

Time Course of mRNA Synthesis After Heat Induction RNA levels as steady state concentrations were measured after different periods of postincubation at 37*C following a 2-hour heat shock induction at 43*C. Line 6 cells were analyzed using a 0.6 kb EcoRI-HindlII probe from the pl7Jo plasmid (Dreano et al., 1986) containing a 5' non-transcribed sequence and 113 bp of the RNA leader (Figure 1A)o In order to avoid hybridization with endogenous hsp70 sequences in the human Wish cells, a second probe was designed, a 620 bp EcoRI-SalI fragment from pl7Jo containing the same 113 bp sequence as above plus 15 bp of the pSP65 polylinker used to insert the the hsp70/hGH hybrid gene. This probe hybridized specifically with the hybrid gene sequence and not with the endogenous hsp70 sequence. In Line 6 cells, the 113 base long fragment (Figure 1A) was protected by mRNA and the optimum level of hybrid mRNA was detected 1 hr after the end of heat shock and started to decrease 9 hr after (Figure 4A). In 3Cgl0 (Figure 4B) and Wg6 (Figure 4C) cells, the 128 base long fragment (Figure 1A) was also protected and the optimum level of hybrid mRNA was obtained 7 hr after the end of heat induction. In these ghGH recombinant cells (Figures 4B and 4C) hybrid mRNA remained present during 24 hr and even 48 hr (results not shown) after the end of heat induction, peculiarly, in the case of Wg6 cells (Figure 4C). Further Analysis of Endogenous and Foreign mRNAs Taking into account that Line 6 cells had been co-tranfected with the chGH gene and the HarasEJ cellular oncogen and because this oncogen can play a role on the transcription of hsp (Simon et al., 1987), while 3Cgl0 and Wg6 cells were transfected with the ghGH plasmid only, we compared the kinetics of endogenous and foreign hsp mRNA within Wg6 cells using a mixture of two probes: 2) a 1.7 kb BglII fragment from p17 (Voellmy et al., 1985) which contains around 1.2 kb of the 5' non-transcribed sequence, the 118 bp of the complete RNA leader, as well as 362 bp of the coding sequence of the human hsp70 gene and 2) the 620 bp EcoRI-SalI fragment of pl7Jo, as described above, were used in double S1 mapping experiment with the total RNA of Wg6 cells retrieved at different times after induction. Figure 5 shows that two fragments, the larger was 480 base long and the smaller 128 base long were protected. They correspond to the expected size of protected fragments by endogenous hsp70 mRNA and recombinant hsp70 mRNA respectively. The maximum level of expression of endogenous hsp70 occurred immediately after the end of the heat shock and decreased sharply 1 hr after. Hsp70/hGH hybrid mRNA continued to accumulate and was still detectble 24 hr after the end of heat shock, confirming that the hybrid hsp70 mRNA was more stable than the endogenous hsp mRNA.

Cell Biology and Toxicology, Vol. 8, No. 2, 1992

151

DISCUSSION The selection and the characterization of Line 6, a NIH-3T3 cell line transfected with the plasmid pl7hGHdhfr which expresses the chGH gene both in vitro (Dreano et al., 1986; Dreano et al., 1988a) and in vivo (Dreano et al., 1988b), has been previously described. This chGH gene which encodes the cDNA copy is under the control of the promoter of the human hsp70 gene transfected with the plasmid pl7hGHdhfr. The transfection of Wish and NIH-3T3 cells with the plasmid pl7ghGHneo in which the ghGH gene, under the same hsp70 promoter, is the cellular gene containing four introns, has also permitted the isolation of cell lines expressing a large amount of hGH upon heat induction. The basal level of expression without heat induction remained very low; and hGH production, controlled by heat induction, reached secretion ratios corresponding to heat induced over non-induced hGH release ranging from 100- to 1000-fold. Levels of production in selected clones were comparable, in the order of 3-4 lag per 106 cells after a single heat shock induction. They were similar to that of Line 6 cells expressing the intronless hGH gene of the hsp/chGH hybrid gene. Polypeptides synthesized by cells transfected with hsp/chGH hybrid genes, where hGH was encoded from either chGH or ghGH, were both efficiently secreted. It is interesting to note that some stress proteins are involved in this process, particularly the BIP/GRP78 protein which has at least 60% homology with the HSP70 protein, and binds transiently to a variety of nascent wild type exocytic proteins (Gething and Sambrook, 1989). Furthermore, both types of recombinant cells secreted hGH with an apparent Mr equal to 22000 Da. This polypeptide elicited a lactogenic hormonal activity, the induction of 13-casein synthesis by mammary gland explants, when these explants were incubated with the spent medium of each heat-treated recombinant cells. Therefore, the two types of recombinant gene directed the synthesis of a polypeptide which was correctly matured and secreted into a functional hGH.

Role of lntrons or Intervening Sequences The consequence of either the absence or the presence of introns for the production of recombinant genes is more and more frequently documented. For example, in transiently transfected cultured cells, a SV40/rabbit-fl-globin cDNA gene (Buchman and Berg, 1988) as well as the mouse ribosomal protein cDNA gene rpL32 (Chung and Perry, 1989) were found to be inactive, while in both cases the addition of intervening sequences significantly increased (up to 400-fold) RNA production. Cis-regulatory elements were also related within the largest intron of the adenosine deaminase gene in transgenic mice (Aronow et al., 1989). In contrast, the ino tronless gamma-globin gene was expressed at 24% of the level of an intron-containing gene in eryhtoid cells using viral infection (Rixon et al., 1990). In contrast to our work on chGH or ghGH recombinant cells, Pasleau et al. (1987) found that eucaryotic cells transiently transfected with the bovine growth hormone (bGH) gene lacking introns, i.e., cbGH, expressed 2 to 7 times less bGH than cells transfected with the genomic gene, gbGH. Deng et al. (1989) similarly observed a low-level expression of the intronless mouse thymidylate synthase (ts) gene (cts) that had the normal 5' and 3' flanking regions as compared to the genomic ts (gts) during their transient expression in ts- hamster V79 cells. However, inclusion of introns 5 and

152

Alouani et al.

6 only in cts provoked an 8- to 9- fold increase of mRNA and protein synthesis. In addition, introns 5 and/or 6 could almost be deleted except a 60 bases sequence next to each exon without losing their stimulatory effect. However, all of the above in vitro studies only dealt with the transient expression of the gene vector, replicative or not, because investigations on stable recombinant cells were not reported.

Interaction of Promoters and Splicing Mechanism The heat shock which induces the hsp promoter profoundly modifies the formation of ribonucleoprotein particles (Hickey and Weber, 1982; Mayrand and Pederson, 1983), and provokes alterations in the transcriptional machinery (Lindquist, 1986). In cells submitted to a severe heat shock, precursors of hsp83 mRNA containing introns were found to accumulate; however, cells pretreated with a mild heat shock prior to a severe heat shock were able to produce mature hsp83 mRNA (Yost and Lindquist, 1986). These results suggest that inhibition of the splicing mechanism constitutes a supplementary step of regulation of non-heat shock mRNA and protein syntheses during stress, while heat shock proteins themselves play a protective role at the level of splicing. Recently, another relation between the heat shock phenomenon and the regulation of transcription was described using Trypanosoma cells (Muhich and Boothroyd, 1989). In these cells, mRNA synthesis involved a trans-splicing pathway which was sensitive to heat shock, with the exception of hsp70 mRNA. Splicing of transcribed mRNAs depends on mechanisms specific to host cells. For instance, maturation of eucaryotic genes did not occur in plant cells, and a nematode hsp gene, which contains introns, was not matured in mouse cells (Kay et al., 1987). Bornstein et al. (1988) have observed that the human collagen al(I)/hGH hybrid gene containing deletions and/or inversions of intronic sequence resulted in a markedly reduced hGH synthesis in chick tendon fibroblasts and not in 3T3 cells when transiently transfected by this plasmid. It was concluded that such intervening sequences exhibited critical interactive effects with the promoter of the collagen gene. In our hands, Xenopus oocytes injected with pl7hGHneo were able to secrete around 3 ng of hGH per oocyte, while, in similar experimental conditions, oocytes injected with pl7ghGHneo did not produce any hGH (data not shown). In transgenic mice, the genomic rat growth hormone gene (grGH) under the control of the metallothionein promoter was expressed more efficiently than the intronless (crGH) gene, upon Zn or Cd induction, although heterologous intervening sequences inserted between the same promoter and the crGH gene markedly improved its expression (Palmiter et al., 1991).

Expression of mRNA from hsp Hybrid Genes The most significant difference observed in vitro between both types of recombinant cells appeared in the time course of hybrid mRNA accumulation as well as of hGH secretion. In ghGH recombinant cells, this phenomenon was more significant with Wg6 cells than with 3Cgl0 cells as compared to chGH recombinant Line 6 cells. A certain number of mechanisms could explain this phenomenon: 1) mRNA precursor transcripts from ghGH require processing which in principle should be submitted to a linkage to spliceosomes leading to nuclear stabi-

Cell Biology and Toxicology, Vol. 8, No. 2, 1992

153

lization, etc., delaying the migration of mRNA from the nucleus to the translational machinery, 2) introns containing regulatory sequences, such as tissue-specific factors, transcriptional enhancers, etc., may lead to interactions with factors involved in the regulation of the hsp70 promoter, 3) cells stressed by the heat treatment maintained an "auto-induction," due to the presence of intracellular unfolded proteins which can re-induce the heat shock system (Ananthan et al., 1986), and 4) the half-life of mRNA transcribed from ghGH may be more stable than that from chGH. To examine the transcription kinetics, the cell line expression was investigated at the mRNA level. Hybrid hsp70/hGH mRNA transcribed from the ghGH gene was detectable 24 hr (Figures 4B and C) and even 48 hr (result not shown) after induction, while hybrid hsp70/mRNA transcribed from the chGH gene of Line 6 cells started to disappear after a 9hour period following the end of heat induction (Figure 4A). In order to examine if cells were not maintained in a slress state, the mRNA of Wg6 cells was analyzed using the two different probes described in the Result section. Endogenous hsp70 mRNA and exogenous hybrid hsp70/hGH mRNA produced by Wg6 cells were analyzed by S 1 mapping. Results indicate that hsp hybrid mRNA was present in cytoplasmic extracts for up to 24 hours while endogenous hsp rnRNA was barely detectable 3 hours after the end of heat treatment (Figure 5). This observation suggests that mRNA synthesis from the hybrid gene containing introns remained active during a much longer time than that of endogenous hsp mRNA, and/or that former mRNAs were more stable. We can speculate that recombinant hybrid mRNa did not migrate from the nucleus using the specific pathway of synthesis of endogenous hsp mRNA, but followed the preferential pathway of non-hsp mRNA in which RNA will be matured. In conclusion, although it was suggested that transcriptional activation is regulated by a complex set of interactions mediated by DNA-binding proteins between intronic and 5'-flanking promoter sequences, we showed that such interactions are not essential using hsp70 promoter. This promoter allowed, in a fair number of host ceils, a powerfully inducible transcriptional activation of genes that did and did not contain intervening sequences. ACKNOWLEDGMENTS We thank M. Garcia, J. Marquez, and H. Grabowski for their excellent technical work, D. Rungger for the critical reading of the manuscript, and P. Bromley for support. This work was supported by grants from Battelle Memorial Institute, IntraCel S.A., the Fonds National Suisse de la Recherche Scientifique, n°3.542-0.86, and the Biotechnology Action Programme of the European Community and the Ligue Bourguignone Contre le Cancer. REFERENCES

ANANTHAN, J., GOLDBERG, A.L., and VOELLMY, R. (1986). "Abnormal proteins serve as eukaryotic stress signals and trigger the activation of heat shock gene." Science 232:522-524.

154

Alouani et al.

ARONOW, B., LATTIER, D., SILBIGER, R., DUSING, M., HUTTON, J., JONES, G., STOCK, J., McNEISH, J., POTTER, S., WITTE, D., and WIGINTON, D. (1989). "Evidence for a complex regulatory array in the first intron of the human adenosine deaminase gene." Genes & Develop. 3:1384-1400. ASHBURNER, M. and BONNER, J.J. (1979). "The induction of gene activity in Drosophila by heat shock." Cell 17:241-254. BALLIVET, M., NEF, P., COUTURIER, S., RUNGGER, D., BADER, C.R., BERTRAND, D., and COOPER, E. (1988). "Electrophysiology of a chick neuronal nicotinic acetylcholine receptor expressed in Xenopus oocytes after cDNA injection." Neuron 1:847-852. BANERJI, J., OLSON, L., and SCHAFFNER, W. (1983). "A lymphocyte-specific cellular enhancer is located downstream of the joining region in immunoglobulin heavy chain genes." Cell 33:729-740. BOND, U. and SCHLESINGER, M.J. (i986). "The chicken ubiquitin gene contains a heat shock gene promoter and expresses an unstable mRNA in heat-shocked ceils." Mol. Cell. Biol. 6:4602-4610. BONNER, J.J., PARKS, C., PARKER-THORNBURG, J., MORTIN, M.A., and PELHAM, H.R.B. (1984). "The use of promoter fusions in Drosophila genetics: isolation of mutations affecting the heat shock response." Cell 37:979-991. BORNSTEIN, P., McKAY, J., LISKA, D.J., APONE, S., and DEVARAYALU, S. (1988). "Interactions between the promoter and first intron are involved in transcriptional control of o~1(I) collagen gene expression." Mol. Cell. Biol. 8:4851-4857. BREITBART, R.E., NGUYEN, H.T., MEDFORD, R.M., DESTREE, A.T., MAHDAVI, V., and NADAL-GINARD, B. (1985). "Intricate combinatorial patterns of exon splicing generate multiple regulated troponin T isoforms from a single gene." Cell 41:67-82. BUCHMAN, A.R. and BERG, P. (1988). "Comparison of intro-dependent and intron-independent gene expression." Mol. Cell. Biol. 8: 4395-4405. CATTANEO, A. and NEUBERGER, M.S. (1987). "Polymeric immunoglobulin M is secreted by transfectants of non-lymphoid cells in the absence of immunoglobulin J chain." EMBO 6:2753-2758. CHEE, P.P., KLASSY, R.C., and SLIGHTOM, J.L. (1986). "Expression of a bean storage protein phaseolin "minigene" in foreign plant tissues." Gene 41:47-57. CHUNG, S. and PERRY, R.P. (1989). "Importance of introns for expression of mouse ribosomal protein gene rpL32." Mol. Cell. Biol. 9:2075-2082. DANSCHER, G. and NORGAARD, J.O.R. (1983). "Light microscopic visualization of colloidal gold on resin-embedded tissue." J. Histochem. Cytochem. 31:1394-1398. DENG. T., LI, Y., and JOHNSON, F. (1989). "Thymidylate synthase gene expression is stimulated by some (but not all) introns." Nuc. Acids Res. 17:645-658. DIERKS, P., VAN OOYEN, A., MANTEI, N., and WEISSMANN, C. (1981). "DNA sequences preceding the rabbit [3-globin gene are required for mouse L cells RNA formation in mouse L cells of ~-globin RNA with the correct 5' terminus." Proc. Natl. Acad. Sci. 78: 1411-1415. DREANO, M., FOUILLET, X., BROCHOT, J., VALLET, J.M., MICHEL, M.L., RUNGGER, D., and BROMLEY, P. (1986). "High-level, heat-regulated synthesis of proteins in eukaryotic cells." Gene 49:1-8. DREANO, M., BROCHOT, J., MYERS, A., CHENG-MEYER, C., RUNGGER, D., VOELLMY, R., and BROMLEY, P. (1987). "Heat-regulated expression of the hepatitis B virus surface antigen in the Human Wish cell line." Virus Res. 8:43-59. DREANO, M., FISCHBACH, M., MONTANDON, F., SALINA, C., PADIEU, P., and BROMLEY, P. (1988a). "Production of secretable proteins using the passage in vivo as tumours of cells carrying heat-inducible constructs." Biotechnology 6:953-958. DREANO, M., MARQ, J.B., and BROMLEY, P. (1988b). "Antibody formation against heat-induced gene products expressed in animals." Biotechnology 6:1340-1343. GETHING, M.J. and SAMBROOK, J. (1989). "Protein folding and intracellular transport: studies of influenza virus haemagglutinin." Biochem. Soc. Symp. 55:155-166.

Cell Biology and Toxicology, Vol. 8, Noo 2, 1992

155

GILLIES, S.D., MORRISON, S.L., OI, V.T. and TONEGAWA, S. (1983). "A tissue-specific transcription enhancer element is located in the major intron of a rearranged immunoglobin heavy chain gene." Cell 33:717-728. GRAHAM, F.L. and VAN DER EB, A.J. (1973). "A new technique for the assay of infectivity of human adenovirus 5 DNA." Virology 52:456-467. GROSS, M.K., KAINZ, M.S., and MERRILL, G.F. (1987). "Introns are inconsequential to efficient formation of cellular thymidine kinase mRNA in mouse L cells." Mol. Cell. Biol. 7:45764581. GRUSS, P., LAI, C.J., DHAR, R., and KHOURY, G. (1979). "Splicing as a requirement for biogenesis of functional 16S mRNA of Simian Virus 40. Proc. Natl. Acad. Sci. 76:4317-4321. GRUSS, P. and KHOURY, G. (1980). "Rescue of a splicing defective mutant by insertion of an heterologous intron." Nature 286:634-637. HAMER, D.H. and LEDER, P. (1979). "Splicing and the formation of stable RNA." Cell 18:12991302. HAMER, D.H., SMITH, K.D., BOYER, S.H., and LEDER, P. (1979). "SV40 recombinants carrying rabbit 13-globin gene coding sequences." Cell 17:725-735. HICKEY, E.D. and WEBER, L.A. (1982). "Modulation of heat-shock polypeptide synthesis in HeLa cells during hyperthermia and recovery." Biochemistry 21:1513-1521. HICKEY, E., BRANDON, S.E., POTTER, R. , STEIN, G., STEIN, J., and WEBER, L.A. (1986). "Sequence and organization of genes encoding the human 27 kDa heat shock protein." Nucl. Acids Res. 14:4127-4145. INGOLIA, T.D. and CRAIG, E.A. (1982). "Drosophila gene related to the major heat shock-induced gene is transcribed at normal temperatures and not induced by heat shock." Proc, Natl. Acad. Sei. 79:525-529. JAHN, G., DUSANTER-FOURT, I., KELLY, P.A., HOUDEBINE, L.M., and DJIANE, J. "Measurement by radioimmunoassay of casein content in Rabbit mammary gland during pregnancy and after prolactin stimulation in organ culture." Proc. Soc. Exp. Biol. Med. 184:19-23. KAY, R.J., RUSSNAK, R.H., JONES, D., MATHIAS, C., and CANDIDO, E.P.M. (1987). "Expression of intron-containing C. elegans heat shock genes in mouse cells demonstrates divergence of 3' splice site recognition sequences between nematodes and vertebrates, and an inhibitory effect of heat shock on mammalian splicing apparatus." Nucl. Acids. Res. 15:37233741. LAEMMLI, U.K. (1970). "Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 277, 680-685. LINDQUIST, S. (1986). "The heat-shock response." Annu. Rev. Biochem. 55:1151-1191. MAYRAND, S. and PEDERSON, T. (1983). "Heat shock alters nuclear ribonucleoprotein assembly in Drosophila cells." Mol. Cell. Biol. 3:161-171. MOORE, D.D., MARKS, A.R., BUCKLEY, D.I., KAPLER, G., PAYVAR, F., and GOODMAN, H.M. (1985). "The first intron of the human growth hormone gene contains a binding site for glucocorticoid receptor." Proc. Natl. Acad. Sci. 82:699-702. MUHICH, M.L. and BOOTHROYD, J.C. (1989). "Synthesis of trypanosome hsp70 mRNA is resistant to disruption of trans-splicing by heat shock." J. Biol. Chem. 264:7107-7110. NG, R., DOMDEY, H., LARSON, G., ROSSI, J.J., and ABELSON, J. (1985). "A test for intron function in the yeast actin gene." Nature 314:183-184. NOVER, L. (1987). "Expression of heat shock genes in homologous and heterologous systems." Enzyme Microb. Technol. 9:130-144. PALMITER, R.D., NORSTEDT, G., GELINAS, R.E., HAMMER, R.E., and BRINSTER, R.L. (1983). "Metallothionein-human GH fusion genes stimulate growth of mice." Science 222:809-814. PALMITER, R.D., SANDGREN, E.P., AVARBOCK, M.R., ALLEN, D.D., and BRINSTER, R.L. (1991). "Heterologous introns can enhance expression of transgenes in mice." Proc. Natl. Acad. Sci. 88:478-482. PASLEAU, F., LEUNG, F., and KOPCHICK, J.J. (1987). "A comparison of bovine growth hormone expression directed by bGH genomic or intronless DNA in transiently transfected eukaryotic cells." Gene 57:47-52. PERBAL, B. (1988). A practical guide to molecular cloning. A Wiley-Interscience Publication.

156

Alouani et al.

PICARD, D. and SCHAFFNER, W.A. (1984). "A lymphocyte-specific enhancer in the mouse immunoglobin k gene." Nature 307: 80-82. RANCOURT, D.E., WALKER, V.K., and DAVIES, P.L. (1987). "Flounder antifreeze protein synthesis under heat shock control in transgenic Drosophila melanogastero" Molo Cello Biolo 7:21882195. RIXON, M.W., HARRIS, E.A.S., and GELINAS, R.E. (1990). "Expression of the human gammaglobin gene after retroviral transfer to transformed erythroid cells. Biochemistry 29:43934400. SANZO, K., BRIGTWELL, B., and WARREN, T.G. (1988). "Temperature regulated expression of tissue plasminogen activator in a heterologous cell line. Proc. Miami Biotechnol. Winter Symp. 8:58. SCHLESINGER, M., ASHBURNER, M., and TISSIERIES, A. (1982). Heat Shock from Bacteria to Man. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. SIMON, M.C., KITCHENER, K., HAO, H.T., HICKEY, E., WEBER, L., VOELLMY, R., HEINTZ, N., and NEVINS, J.R. (1987). "Selective induction of human heat shock gene transcription by the adenovirus EIA gene products, including the 12S EIA product." Mol. Cell. Biol. 7:2884-2890. SLATER, E.P., RABENAU, O., KARIN, M., BAXTER, J.D., and BEATO, M. (1985). "Glucocorticoid receptor binding and activation of a heterologous promoter by dexamethasone by the first intron of the human growth hormone gene." Mol. Cell. Biol. 5:2984-2992. TABIN, C.J., BRADLEY, S.M., BARGMANN, C.I., WEINBERG, R.A., PAPAGEORGE, A.G., SCOLNICK, E.M., DAHR, R., LOWRY, D.R., and CHANG, E.H. (1982). "Mechanism of activation of a human oncogene." Nature 300:143-149. TSAI-MORRIS, C.H., BUCZKO, E., WANG, W., and DUFAU, M.L. (1990). "Intronic nature of the rat luteinizing hormone receptor gene defines a soluble receptor subspecies with hormone binding activity." J. Biol. Chem. 265:19385-19388° VOELLMY, R., AHMED, A., SCHILLER, P., BROMLEY, P., and RUNGGER, D. (1985). "Isolation and functional analysis of a human 70,000-dalton heat shock protein gene segment." Proc. Natl. Acad. Sci. 82:4949-4953. WEAVER, R.F. and WEISSMAN~ C. (1979). "Mapping of RNA by a modification of the Berk-Sharp procedure: the 5' termini of 15S ~-globin mRNA precursor and mature 10S ~-globin mRNA have identical map coordinates." Nucl. Acids Res. 7:1175-1192. WURM, F.M., GWINN, K.A., and KINGSTON, R.E. (1986). "Inducible overproduction of the mouse c-myc protein in mammalian cells." Proc. Natl. Acad. Sci. 83:5414-5418. YOST~ H.J. and LINDQUIST, S. (1986). "RNA splicing is interrupted by heat shock and is rescued by heat shock protein synthesis." Cell 45:185-193.

received: 8/15/91 accepted: 2/14/92