Susceptibility to natural killer cells and down regulation of MHC class I expression in adenovirus 12 transformed cells are regulated by different E1A domains

EL SE VIER

Virus Research 45 (1996) 123-134

Virus Research

Susceptibility to natural killer cells and down regulation of MHC class I expression in adenovirus 12 transformed cells are regulated by different E1A domains Isabelle Huvent,

Chantal

Cousin,

Jean Claude

A l e x i K i s s 1, C l a u d i n e

Bernard,

D'Halluin*

I N S E R M U124, Institut de Recherche sur le Cancer de Lille, I place de Verdun, 59045 Lille cedex, France

Received 30 April 1996; accepted 3 July 1996

Abstract

All human adenoviruses transform rodent cells in vitro, but only cells transformed by serotypes belonging to subgroups A (Adl2) and B (Ad3) are tumorigenic for immunocompetent animals. In these cells, the expression of MHC-class I antigens is repressed and might allow them to escape from recognition by cytotoxic -f lymphocytes (CTL) and to develop in tumor. Furthermore, these cell lines appear resistant to lysis by natural killer (NK) cells. To determine the E1A domain(s) responsible for these properties several cell lines were created by transforming baby rat kidney (BRK) cells with a set of plasmids expressing different Ad2/Adl2 hybrid E1A gene products. The MHC class I gene expression was inhibited in cells expressing the Adl2 13S mRNA product and in cells transformed with Ad2/Adl2 hybrid E1A gene product harboring the C-terminal part of the conserved region (CR) 3 of Adl2. Susceptibility of these transformed cell lines to NK cells was determined by cytolytic assays. The results obtained suggest that two Adl2 E1A domains are required to induce resistance of the cell lines to NK cells. K e y w o r d s : Adenovirus; E1A; E1B; MHC-class I; Natural killer cells; Serotypes; Transformation

1. Introduction

H u m a n adenoviruses were first shown to be * Corresponding author. Tel: + 33 20 529700; fax: + 33 20 527083. Present address: Department of Microbiology, Semmelweis University of Medicine, Nagyvarad T6r 4, Budapest 1089, Hungary.

oncogenic in hamsters (Trentin et al., 1962). Following this discovery, several adenovirus serotypes were categorized into nononcogenic (Ad2, Ad5) and highly oncogenic (Adl2, AdlS, Ad31) groups by their capacity to induce tumors in rodents (Huebner et al., 1965). In spite of the differences in oncogenicity of different Ad serotypes in vivo, all of these viruses have the

0168- 1702/96/$l 5.00 © 1996 Elsevier Science B.V. All rights reserved P H SO168-1 702(96)01367-6

124

I. Huvent et al. / Virus Research 45 (1996) 123-134

capacity to transform normal rodent cells into immortalized cells. In general, the tumor inducing capacity of these transformed cells reflects the oncogenic phenotype of the transforming virus. Cells transformed by nononcogenic Ad2 and Ad5 are either non-tumorigenic or tumorigenic only when injected into immunodeficient animals, while cells transformed by highly oncogenic Adl2 are tumorigenic in immunocompetent as well as immunodeficient animals (reviewed in Bernards and van der Eb, 1984; Boulanger and Blair, 1991; Branton et al., 1985; D'Halluin and Lecl6re, 1992). The dependence of tumorigenicity on the immune status of the recipient suggests that surveillance by the immune system in the injected rodent plays a determining role in the tumorigenicity of adenovirus-transformed cells. The cellular immune compartment is involved in the recognition and elimination of tumor cells. Two components (specific and nonspecific) within the cellular immune compartment, act against potentially neoplastic cells. The specific component is mediated by the capacity of immune systemstimulated cytotoxic T lymphocytes (CTL) to recognize and destroy cells that express foreign cell-surface proteins associated with class I major histocompatibility complex (MHC) proteins. The nonspecific component is mediated by the capacity of natural killer (NK) cells and macrophages to eliminate neoplastic cells. The tumorigenicity of human adenoviruses in rodents depends mainly on the ability of different Ad serotypes to induce in transformed cells a resistance to the destruction caused by one or both of these components of the cellular immune system. The Adl2 E1A gene inhibits the expression of class I MHC proteins on the surface of transformed cells, thus removing a critical signal for recognition by CTL (Bernards et al., 1983; Jochemsen et al., 1984; Schrier et al., 1983; Vaessen et al., 1986). It has been shown that repression of class I MHC gene expression occurs mainly at the level of initiation of transcription in Adl2-transformed cells (Ackrill and Blair, 1988; Friedman and Ricciardi, 1988). Adl2-mediated transformation of cells derived from transgenic mice carrying the growth hormone gene under the control of the class I H-2K b regulatory elements

shows that sequences between - 2 0 1 5 and + 12 are sufficient to mediate down regulation of the transgene (Meijer et al., 1989). Furthermore, the early region 1A of nononcogenic Ad2/Ad5 can induce high levels of susceptibility to lysis by NK cells, while Adl2 E1A gene either induces lower levels of susceptibility or/and induces resistance to N K lysis (Cook and Lewis, 1984; Cook et al., 1986; Gallimore and Paraskeva, 1979; Kenyon and Raska, 1986; Mak et al., 1979; Sawada et al., 1985). The susceptibility of Ad2 transformed cells to N K celt lysis is directly correlated with the absence of in vivo tumorigenicity (Sheil et al., 1984). The susceptibility of Ad2/Ad5 transformed cells to NK cell lysis require a normal level of the E1A protein (Cook et al., 1987, 1993). When using recombined E1A region between Ad2 and Adl2, van der Eb and his colleagues demonstrated that the second exon is not important for the tumorigenicity of Ads from subgroup A (Jochemsen et al., 1984). Likewise, we constructed plasmids expressing Ad2/Adl2 hybrid E1A region and transformed primary cells. Then, measuring the levels of the MHC class I antigens at the transformed cells surface, we have tried to establish the structure-function relationships of Adl2 E1A proteins. The C-terminal domain of the Adl2 E1A CR3 is involved in the negative transcriptional regulation of the MHC class I gene. Furthermore, we analyzed the susceptibility of the transformed cells to NK cells. Only the cell lines transformed with the hybrid containing the second exon of the Adl2 E1A protein and a domain extending between amino acids 105 and 156 including an alanine rich sub-domain, located between CR2 and CR3 in adenovirus 12 E1A show a reduced susceptibility to N K cells.

2. Materials and methods 2.1. Plasmid constructions

Plasmids p2E1A, pI2E1A, pl2E1B have been previously described (Cousin et al., 1991; Leite et al., 1986, 1988). The p12 12S and p12 13S cDNA were kindly provided by Dr. Perricaudet (Paris).

1. Huvent et al. / Virus Research 45 (1996) 123 134

Ad2/Adl2 hybrid E1A plasmids were constructed by creating new restriction sites in derivates of the pSelect vector by site-directed mutagenesis as previously described (Lecl6re et al., 1993), and then by exchanging corresponding sequences (Fig. 3).

125

washing, the cells were analyzed for surface fluorescence in an EPICS (Coulter Electronics, FL). Assays were repeated on different lines cloned out for several transformation experiments. 2.4. Cytotoxicity assays

2.2. Cells and transfection procedures

Fischer 344 rats of the RT-11 haplotype were purchased from Iffa Credo (France). Primary cultures of baby rat kidney (BRK) cells were prepared from 6-days-old baby rat kidneys. Cells were grown in monolayer culture in Dulbecco modified Eagle medium (DMEM) with 10% fetal bovine serum (FBS). Transformation assays were performed by liposome transfection of plasmid DNA. BRK cells at approximately 70% confluence in culture bottles were co-transfected with 20 /~g plasmid expressing E1A + 20 /~g of pl2E1B. After 12 h, the medium was removed and then changed twice a week for 50 days. Cell lines were established from individual foci. Subcultures of primary BRK cells and established cell lines obtained from transformation foci were plated. At approximately 70-80% confluence they were fixed and stained with crystal violet. For NK cells susceptibility tests, target cells were labeled with 10 /~Ci of methyl [3H]thymidine (70 Ci/mmol; Amersham, UK) by Petri dishes for 24 h and washed twice in growth medium prior addition of NK cells. Total RNA were extracted from cells using guanidium/cesium chloride method and separated by electrophoresis on agarose gel.

Effectors cells were prepared essentially as previously described (Reynolds et al., 1981; Timonen and Saksela, 1980). In brief, spleen cells were obtained by mincing the tissue and passing it through a 50-gauge mesh screen in medium RPMI 1640 supplemented with 10% heat inactivated fetal calf serum (FCS). Mononuclear cells were isolated by Ficoll-Isopaque gradient centrifugation, washed twice in PBS, and suspended in RPMI supplemented with 10% heat inactivated FCS. Mononuclear cells were incubated for 2 h in culture flasks to remove adherent cells, and nonadherent cells were then counted and used in cytotoxicity assays. Cytotoxicity experiments were performed in 98 wells culture plates, each combination in triplicate in 0.2 ml. Effector:target ratio were 1:1, 10:1 and 25:1 or 50:1 and exposure time 16 h. To stop the cytotoxicity assay, 0.2 ml of PBS was added to each well, plates were centrifuged and supernatants were transferred; cells were lysed by addition of 0.2 ml NaOH 0.3 M, 30 min at room temperature. Macromolecular DNA was precipitated with TCA, filtrated, and filters were air dried and then counted. The cell supernatants were also counted to confirm the results obtained with cell pellets.

2.3. Fluorescence-activated cell sorter (FACS) analysis

3. Results

FACS analysis was performed to determine the cell surface density of MHC class I antigens. Cells were washed in phosphate buffered saline (PBS) and incubated with monoclonal antibody in excess. The monoclonal antibody used against the determinant RT-IA of rat class I MHC antigens was purchased from Biosys (France). After three washes, the cells were stained with the fluorescein isothiocyanate-conjugated IgG fraction of goat anti-mouse IgG (Institut Pasteur, Paris). After

3.1. Ad12 E I A large protein is responsible jbr the inhibition of M H C class I expression

All adenoviruses transform primary BRK cells, but the efficiency is serotype (subgroup) dependent. We showed that the E1A promoter is not involved in this phenomenon and that the Adl2 CR1 domain acted in decreasing the transformation frequency (Lecl6re et al., 1993). The BRK cells transformed with plasmids expressing 12S- or

126

L Huvent et al. / Virus Research 45 (1996) 123-134

13S-mRNA of Ad2 or Adl2 were analyzed for the expression of MHC class I antigen by flow cytofluorimetry (Fig. 1). The cell surface fluorescence of Adl2-transformed cells was reduced compared with that of Ad2-transformed cells and the product of Adl2 13S mRNA was sufficient to reduce the expression of MHC class I antigen. In cells transformed with Adl2 12S cDNA, the level of class I antigen appeared similar to that of Ad2 transformed cells (Fig. 1 and Table 1). The Ad2 and Adl2 13S mRNA products differ by at least two features: (i) the size of the 13S mRNA specific product (46 aa for Ad2, 30 aa for Adl2) resulting of the splice donor site position; and (ii) the presence of poly-alanine domain between the CR2 and CR3 in Adl2 E1A large protein (Fig. 2). 3.2. MHC class I gene expression in transformed cell lines with hybrid E1A gene

In order to analyze the Adl2 E1A domain responsible for the inhibition of MHC class I expression we constructed hybrid E1A regions between Adl2 and Ad2 by exchanging domains after creating restriction sites by directed mutagenesis as shown in Fig. 3. The resulting plasmids were transfected into primary BRK cells and approximately ten transformed cell lines

E ¢-

0

6

30

6

30

6

30

Log FI Fig. 1. Cell surface expression of class I antigens by FACS analysis. Patterns from log fluorescence intensity versus the cell number showed the antigen levels on surface of the cells transformed by p2E1A (A), p2(13S) (B), p2(12S) (C), pl2E1A (D), p12(13S) (E), p12(12S) (F). Panels D and E exhibit a fluorescence intensity inferior to 6.0 for more than 60% of cells; more than 81% of cells from the other patterns possess a log F1. I. superior to 6.0.

were established from cell individual loci arising from transformation assays. Some differences were observed in the morphologies of transformed cell lines. Cells transformed by p2C12 + Elb and p 2 S 1 2 + E l b were somewhat flat, spindle shaped (not shown) and resembled the Adl2El-transformed cells, while Ad2El-transformed cells grew to high densities and possessed a very rounded shape. Therefore, Adl2 E1A gene products entailed less morphological changes than Ad2 E1A region, and in this, we agree with Shiroki et al. (1979). Recombinants p2C12 and p2S12 that retained the C-terminal of Adl2 protein, perhaps failed to induce an epithelial cell growth factor in transformed BRK cells, as described for Ad2 (Quinlan et al., 1988). Total R N A were prepared from cells and Fig. 4 shows the expression levels of E1A gene in adenovirus transformed cells determined by Northern blot analysis. Northern blot hybridization with E1A probe demonstrated that the levels of 13S mRNA expression weakly changed between the different cell lines tested, a lower level was observed in p2C12 and p12C2 cell lines and a higher level in p2C12S2 cell lines. Three different cell lines transformed by each hybrid, before the 20th passage, were characterized by the rate of MHC class I antigens on cell surface by flow cytofluorimetry (Fig. 5 and Table 1). A lowered level of MHC class I antigens was found on the surface of cells transformed by the Ad2/Ad 12 recombinants harboring the Adl2 CR3 region (p2S12, p2C12) (Fig. 5(E,F)). Moreover, cells transformed with p12S2, p2C12S2, in fact hybrids possessing Ad2 CR3 region, were also tested for the expression of the class I antigens on cell surface; they expressed high levels of class I antigens as Ad2 El-transformed cells. The major difference between Ad2 and Adl2 E1A region, in regulation of MHC class I expression, appeared located in the C-terminal part of the CR3. The E1A domain extending between amino acids 105 and 156 including an alanine rich sub-domain, located between CR2 and CR3 in adenovirus 12 E1A was not responsible for the inhibition of MHC class I expression.

L Huvent et al. / Virus Research 45 (1996) 123 134

127

Table 1 MHC class I expression and resistance to NK cell lysis of adenovirus transformed cells Plasmidsa

Cell lines

MHC class I expressionb

NK cell resistance':

p2E1A

250-a 250-b 251-a 251-c 252-d 252-f 1201-b 1201-c 1221-a 1221-g 1222-b 1222-c 1229-a 1229-b 1229-d 1228-a 1228-b 1228-c t232-b 1232-c 1232-d 1231-a 1231-b 1231-e 1230-a 1230-b

31.6 ( _+15.5) 36.5 (_+ 13.5) 44.9 ( _+32.0) 39.4 ( + 27.3) 11.7 ( + 7.5) 10.8 (-+6.9) 1.9 ( -+ 1.1) 1.7 (+ 1.5) 2.0 ( + 1.2) 1.9 ( _+0.9) 10.4 ( -+ 6.5) 11.2 ( -+ 7.2) 7.5 (_+4.3) 6.8 ( + 4.1) ND 8.0 ( _+3.9) 7.2 ( _+3.5) 8.5 ( + 4.3) 41.2 ( _+22.5) 38.4 ( -+ 23.5) ND 36.1 (+ 14.2) 37.5 ( _+19.7) 35.8 (_+ 13.5) 26.9 ( -+ 13.2) 24.3 ( + 12.5)

24.3 ( + 3.5) 22.9 (_+4.1) 21.7 ( _ 5.4) 26.3 ( _+3.1) 18.2 ( -+4.3) 20.7 (-+5.1) 72.6 ( _+3.8) 79.8 (_+4.5) 82.8 ( + 3.0) 74.8 ( + 3.9) 69.6 ( + 4.5) 76.2 ( + 4.3) 56.9 (+ 3.6) 52.7 ( _+4.2) 59.1 ( _+5.4) 24.0 ( _+5.2) ND 19.7 ( _+6.2) 15.3 ( -+ 3.8) 17.8 ( -+ 4.5) 16.4 ( + 3.1) 13.5 (_+4.7) ND 12.9 (_+2.9) 23.7 ( -+ 5.5) 24.3 ( + 3.4)

p2E1A-13S p2El A-12S pt2ElA pI2E1A-13S pl2E1A-12S p2C12 p2S12 p12C2 p12S2 p2C12S2

aPlasmids used in cell transformation. bMean fluorescence _+S.D. ~Percent of resistant target cells determined for ratio effector/target of 25/1.

3.3. Two domains o f A d l 2 E I A proteins are responsible f o r resistance to N K cells

The B R K cells transformed with plasmids expressing 12S- or 1 3 S - m R N A o f A d 2 or Ad12 were analyzed for the susceptibility to N K cells. Cell lines transformed by adenovirus 12 genomic E 1 A or each 12S or 13S c D N A appeared to be resistant to N K cells, in contrast adenovirus 2 transformed cells were susceptible to N K cells (Fig. 6 and Table 1). P r i m a r y B R K cells were resistant to N K cell lysis (not shown). In order to analyze the adenovirus 2 E 1 A domain(s) responsible for the susceptibility to N K cells, the same obtained cell lines were characterized for the susceptibility to N K cells purified f r o m rat spleen as described in Section 2. The transformed cells by hybrid aden-

oviruses were susceptible to lysis by N K cells as those transformed by adenovirus type 2 (Fig. 7 and Table 1); only the cell lines transformed with the hybrid p2C12 in cooperation with adenovirus 12 E1B region show a susceptibility intermediate between that o f adenovirus 2- and 12-transformed cells. This cell line expressed a n o r m a l level o f E 1 A m R N A . This hybrid (p2C12) E 1 A gene p r o d u c t contains the N-terminus, CR1 and the first part o f C R 2 d o m a i n o f adenovirus 2 E 1 A and the C-terminal d o m a i n o f CR2, the entire C R 3 and the second exon o f adenovirus 12 E1A, the unique alanine rich d o m a i n o f adenovirus 12 E 1 A is also present. But this d o m a i n present also in the p2C12S2 transformed cells was not sufficient to provide resistance to N K cell and causes lysis (Fig. 7 and Table 1). These results suggest

128

I. Huvent et al./ Virus Research 45 (1996) 123-134

Adl2 Ad2

KpnI MR-TEMTPLVLSYQEADDILEHLVDNF-FNEVPSDDDLYVPSLYELYDLD - MRHIICHGGVITEEMAASLLDQLIEEVLADNLPPPSHFEPPTLHELYDLD V

Adl2 Ad2

- VESAGEDNNEQAVNEFFPESLILAASEGLFL-PEPPV ........ - V-TAPEDPNEEAVSQIFPDSVMLAVQEGIDLLTFPPAPGSPEPPHLSRQP

Adl2 Ad2

-

48 50

-

CRI

I

-

LSPVC

89 99

CR2

E ..... PIGGECMPQLHPEDMDLLCYEMGFPCSDSEDEQDENGMAHVSAS EQPEQRALGPVSMPNLVPEVIDLTCHEAGFPPSDDEDEEGE .........

134 140

ClaI CR3

Adl2 Ad2

- AAAAAADREREEFQLDHPELPGHNCKSCEHHRNSTGNTDLMCSLCYLRAY ............ EFVLDYVEHPGHGCRSCHYHRRNTGDPDIMCSLCYMRTC SmaI

Adl2 Ad2

- NMFIYSPVSDNEPEPNSTL ............................... - GMFVYSPVSEPEPEPEPEPEPARPTRRPKLVPAILRRPTSPVSRECNSST

203 229

Adl2 Ad2

.... DGDERPSPPKLGSAVPEGVIKPVPQRVTGRRRCAVESILDLIQEEER - DSCDSGPSNTPPEIHPVVPLCPIKPVAVRVGGRRQ-AVECIEDLLNESGQ

250 278

Adl2 Ad2

- EQTVPVDLSVKRPRCN ..... PLDLSCKRPR-P

184 179

$

266 289

I

GAG

GTA

CCC

ATC ATG

E

V

P

M

D

P

G

P CCT CCT

I ATC

D GAT

P CCC

G GGG

L CTT CTG L

P CCA GTA V KpnI

GAT GAT

CCC CCC

GGG GGA

ClaI

Adl2

Ad2

SmaI

Fig. 2. Comparison of the large E1A gene product sequences of Ad serotypes 2 and 12. Protein sequences are aligned. The new restriction sites KpnI, ClaI and SmaI and the changes that occurred are shown. The conserved regions are indicated.

that at least two adenovirus 12 EIA domains were required to confer lysis resistance to NK cells: one of them extended between residues 105 and 156, including the alanine rich domain and the second was contained in the second exon of Adl2 E1A.

4. Discussion The Adl2 transformed cells are more oncogenic than the Ad2/Ad5 transformed cells (Gallimore and Paraskeva, 1979; Mak et al., 1979) and this difference results essentially from the decrease of MHC class I antigens in cells transformed by the highly oncogenic Adl2 and the resistance to NK cells lysis. Although, each 12S and 13S mRNA product of Adl2 transform primary BRK cells with the same efficiency (Lecl6re et al., 1993), the

level of MHC-class I antigen was reduced only in transformed BRK cells expressing the large E1A protein of Adl2 (Fig. 1 and Table 1). This result agrees with the down regulation of a transgene under the control of H-2K b regulatory elements in transgenic mice (Meijer et al., 1989). When using Ad2/Adl2 E1A hybrids, we could have separated the regions required for down regulation of MHC-class I antigen from those involved in susceptibility to NK cells. The C-terminal domain of the CR3 was responsible for down regulation of MHC class I gene expression (Fig. 5 and Table 1). Liu and Green (1990, 1994) described several zones within the E1A conserved region 3 indicating that the C-terminal end of E1A CR3 binds the transcription factor ATF-2. E1A associated with these factors would then act as a co-activator in the trans-activation process

1. Huvent et al./ Virus Research 45 (1996) 123-134

(Dynlacht et al., 1991). Four acidic residues located between 140 and 149 function as an activating region (Lillie and Green, 1989), but in Adl2 E1A protein, this domain is divided by an alanine rich sequence. An alanine rich domain has been shown responsible for inhibitor effect of the Drosophila even-skipped protein (Han and Manley, 1993). This neutral amino-acid sequence of 20 residues did not seem to act in the down regulation of MHC class I expression, since in our experiments hybrids p2C12 and p2S 12 showed the same weak levels of MHC gene expression. Webster and Ricciardi (1991) proposed a model in which the trans-activating domain of E1A binds to different cellular protein targets through the zinc finger region (aa 147-177) and the carboxyl region (aa 183-188). Their substitution mutants in the finger region, similar to the Adl2 corresponding region, did not abolish the C4 zinc finger structure required for trans-activation. Thus, to ensure that the binding of the ATF-2 with the carboxyl region of Adl2 CR3 domain is responsible for down regulation, additional studies will be necessary including the construction of an Ad2/Adl2 E1A hybrid possessing the aminoacid residues 180-188 of Adl2 region. 46

121,136142

77

:i

~1

i

CRI

p2s12

CR2

..........

188

r CR3

'~

s c

~

"!11~

129

E I A 13S

actin 1

2

3

4

5

6

7

Fig. 4. Expression of EIA gene in adenovirus transformed cells. Total RNA was extracted from cells using the guanidium/cesium chloride method, separated by electrophoresis on agarose gel and transferred to a nylon membrane. Its hybridization to nick translated 32p-labeled E1A gene was examined. The probe was a mixture of the Ad2 and Adl2 EIA gene at an equivalent specific radio-labeling. Similar concentrations of RNA (30 pg) were used, as judged by hybridization to a actin gene (lower panel). HEK-293, as control, cell and BRK transformed cell lines were indicated.

During the time course of these experiments, similar approaches were used by Williams and Graham and theirs collaborators; they show that two regions within the first exon of Adl2 E1A influence tumorigenicity (Jelinek and Graham,

1 5 o ~

C

K

K .,=,~

.

.

.

.

.

.

.

c

i5° : 150

-7

p~2C2 s p12S2

![, ; ~

~

', ~ ~ ; ~

50

. . . . . . . .

c

s

p2C12S2

Fig. 3. Ad2/Adl2 hybrid E1A plasmids. Plasmids were constructed by creating new restriction sites by site-directed mutagenesis: Clal at nucleotide (nt) 814 and Smal at nt 964 in Adl2 E1A and Kpnl at nt 655 in Ad2 E1A, and then by exchanging homologous sequences. Upper line: the conserved regions of the Ad2 E1A gene. Numbers indicate AA location of the Ad2 E1A protein. Thick lines, Adl2; thin lines, Ad2.

6

30

6

30

6

30

Log FI

Fig. 5. Cell surface expression of class 1 antigens in cells transformed with hybrid E1A region. Patterns from log fluorescence intensity versus the cell number showed the antigen levels on surface of the cells transformed by p2(13S) (A), p12(13S) (D), p2C12S2 (B), p12C2 (C), p2S12 (E), p2C12 (F). Panels D and F exhibit a fluorescence intensity inferior to 6.0 for more than 60% of cells.

130

I. Huvent et al. / Virus Research 45 (1996) 123 134

100

80 v

60 c

40

._~

o~

~ 2o i

1/1

!

5/1

!

25/1

effectors : target ratio Fig. 6. Specificlysis of Ad2- and Adl2-E1A transformedcells by NK cells. The established cell lines were labeled and incubated with NK cells for 16 h and average of percentage of lysis cells from three independent experiments was estimated by counting the TCA precipitable radiolabeled fraction of the cell pellet. Cells transformed by p2E1A (A), p2(13S) (+), p2(12S) ([]), pl2E1A (*), p12(13S) (O), p12(12S) (0). 1992; Jelinek et al., 1994; Telling and Williams, 1994). These two first exon domains are involved in down regulation of M H C class I gene (Pereira et al., 1995). Another factor influence the tumorigenicity of adenovirus E1 transformants may be the ability of their E1A products to encode CTL epitopes, in Hooded Lister rats only Ad5 E1A proteins encode CTL epitopes but not Adl2 E1A (Pereira et al., 1995). Furthermore, the tumorigenicity of hamster and mouse cells transformed by adenovirus type 2 and 5 is not influenced by the level of class I antigens expressed on the cells, but a result of the susceptibility to N K cells (Haddada et al., 1986). Cells transformed by hybrids encoding amino terminal Adl2 residues up to amino acid 156 were susceptible to N K cells lysis. Furthermore, cells transformed by hybrid E1A gene encoding the specific Adl2 domain present between CR2 and CR3 were also susceptible to N K cells lysis. These results suggest that two Adl2 E1A domains were required and acted in concert to confer resistance to N K cells lysis, the second exon and the domain

extending between amino acids 105 and 156, including a domain rich in Alanine, present in both Adl2 E I A proteins (12S and 13S m R N A products), between the CR2 and the CR3 (Fig. 1). Sequence analysis show that this sequence is absent from the known E1A sequences of other human adenoviruses. However, the amino acids between CR2 and CR3 of Adl2 bear a close similarity to the analogous region of the highly oncogenic simian adenovirus type 7 (Kimelman et al., 1985) (Fig. 8), suggesting that this region, which is unique to these two unrelated oncogenic adenoviruses, may play a role in their oncogenicity. To investigate this possibility, cells were transformed with hybrids harboring Adl2 E1A sequences extending between CR2 and CR3, introduced into Ad2 E1A gene. Transformed cells were susceptible to N K cells lysis at the same level as the Ad2 transformed cells. Furthermore this domain is not involved in the down regulation of M H C class I expression; the Adl2 12S m R N A product does not inhibit the M H C class I expression (Fig. 1 and Meijer et al., 1989) and the transformed cell lines (three independent clones) by the p2C12S2 expressed a high level of class I antigen (Fig. 4). In conclusion, two Adl2 E1A domains were required to confer resistance to N K cells lysis, the second exon and the domain rich in alanine, present in Adl2 E1A large protein, between the CR2 and the CR3 (Fig. 2). Recently, it has been reported that the E1A second exon of adenovirus type 5 is required for induction of target cell susceptibility to lysis by natural killer cells (Krantz et al., 1996). The results obtained with adenovirus transformed cells appear in conflict to the general purpose concerning the relation between the susceptibility to N K cells and expression of M H C class I expression. In general, a high level of class I antigen, in particular the haplotype H-2D d in mice, protects cells from N K cells lysis (for review see Moretta et al., 1994; Raulet, 1994). Corroboration of this effect came from analysis of the destruction of cells from f l 2 m - mice by N K cells. The first studies showed that transplant of fetal liver or bone marrow cells from p 2 m - mice were rejected by irradiated normal mice (Bix et al., 1991). The lysis of f l 2 m - cells may be restricted

131

I. Huvent et al. / Virus Research 45 (1996) 123-134

100

A

.....'"if!:!'"---:...

.......~ ....-.:. .....

............!!)£. .............. ,

o~

70

"'""1~ "..

\t ~t

50

.,.-... \'...

.~_

~. -o 20

I

I

I

I

i

I

111

511

25/1

1/1

5/1

25/1

effectors

• target ratio

Fig. 7. Specific lysis of Ad2/Adl2 hybrid EIA transformed cells by NK cells. The established cell lines were labeled and incubated with NK cells for 16 h and average of percentage of lysis cells from three independent experiments was estimated by counting the TCA precipitable radiolabelled fraction of the cell pellet. Cells transformed by p2C12 shown in intermediate level of specific lysis; about 50% for a ratio 25/1. Cells transformed by p2E1A ( + ); pl2EIA (e); p2S12 (11); p12S2 (A) (A) and p2C12 (*); p2C12S2 ([]); p12K2 ( , ) ; p2K12 (©); pl2E1A, (e) (B).

to hematopoietic cell types. N K cells failed to lyse cells of an embryonic fibroblast cell line established from f l 2 m - embryos (Zijlstra et al., 1992). The high level of class I antigen, in association with /?2 microglobulin, protect cells perhaps by masking a specific receptor recognized by N K cells. Possible explanations for results obtained with adenovirus transformed cells are: (i) the association of peptides from E1A proteins with M H C

class I antigen, a s H - 2 D b in adenovirus type 5 infected cells (Kast et al., 1989), reduces the accessibility to the specific receptor; (ii) E1A proteins modify expression level of specific receptor for N K cells on target cells, as the NKR-P1, a type II integral membrane protein (Giorda et al., 1990); and (iii) cell shape can act upon the accessibility of specific receptor(s). These possibilities are not mutually exclusive. A peptide presented by class I

Ad2 ~¢@~E~G~EDE ..... EGEE . . . . . . . . . . . . . . . . . . . . FVL-~E~RS~ Ad7 ..... ST~,QSIHTAVNSGVK~S . . . . D v ~ t < L - ~ ~ Ad ~ ~ ~ ~ D. . . . . . . . S QAI QN ~ S HGVQAVS . . . . E S m S - ~ ~ ~ ~ ~ Ad ~ ~ ~ ~ i ~ ~ ..... S~MA~VSAS~.~R~S~ ~ ~ ~ SA~ ~ L ~ ~ % ~ % ~ % ~ % ~ E GE H S QVE T E RKMAEAAAA~AAAAARRE QD D FR L- ~ H ~ : ~ ~ Ad 4 1 D ~ ~ ~S~AD EAE E RAE E E E TAVS N YVN IAE GAS .... Q LVL - ~ : i ~ @ : ~ ~ Ad 4 O ~ ' : ~ ~ ~ T DEAT E - -AE E EAAH P T YVN- - E N E N . . . . E LV L - ~ t ~ @ : ~ TAd E - - _ : l £ - - l : i - . i l ± - . .... EEVEAASGLAFETNEEV . . . . . . . . E G F V F P ~ ~ CR2

CR3

Fig. 8. Comparison of the large E1A proteins of human, simian and tupaia adenoviruses. Protein sequences were aligned and shown from CR2 until the first part of CR3. Alanine residues were in bold.

132

I. Huvent et al. / Virus Research 45 (1996) 123 134

MHC antigen and recognized by CTL is encoded by the Adl2/Ad5 second exon (Kast et al., 1989; Sawada et al., 1994; Urbanelli et al., 1989). Furthermore, specific receptors would be involved, for example in mice, the Ly49 antigen is expressed by a subpopulation of spleenic NK cells and, more importantly, in H-2 b strain, the Ly49 + and Ly49 subsets have been shown to differ in their specificity. Ly49 + NK cells do not lyse H-2 d and H-2 k tumor target cells, whereas Ly49- NK cells do (Karlhofer et al., 1992). Furthermore, the IFN increases class I MHC antigen expression without inducing resistance to natural cell killing (Routes, 1992, 1993).

Acknowledgements We are grateful to Dr M. Perricaudet for providing the plasmids p12 13S and p12 12S cDNA. I. Huvent is a recipient of fellowships from the Minist6re de la Recherche et de la Technologie. A. Kiss performed his work in the framework of UNESCO-ROSTE Research Grants Project in Virology of the European Network 'Man Against Virus' (Grant No. 222.227.1). This work was supported by the Institut National de la Sant6 et de la Recherche M6dicale, the Universit6 de Lille II and the ARC (Association pour la Recherche contre le Cancer, Grant No. 415-89).

References Ackrill, A.M. and Blair, G.E, (1988) Regulation of major histocompatibility class I gene expression at the level of the transcription in highly oncogenic adenovirus-transformed rat cells. Oncogene 3, 483-487. Bernards, R. and van der Eb, A.J. (1984) Adenovirus: transformation and oncogenicity. Biochim. Biophys. Acta 783, 187- 204. Bernards, R., Schrier, P.I., Houweling, A., Bos, J.L. and van der Eb, A.J. (1983) Tumorigenicity of cells transformed by adenovirus type 12 by evasion of T cell immunity. Nature (London) 305, 776 779. Bix, M., Liao, N.-S., Zijlstra, M., Loring, J., Jaenisch, R. and Raulet, D. (1991) Rejection of class I MHC deficient haemopoietic cells by irradiated MHC-matched mice. Nature (London) 349, 329-331. Boulanger, P.A. and Blair, G.E. (1991) Expression and inter-

actions of human adenovirus oncoproteins. Biochem. J. 275, 281-299. Branton, P.E., Bailey, S.T. and Graham, F.L. (1985) Transformation by human adenoviruses. Biochim. Biophys. Acta 780, 67 94. Cook, J.L. and Lewis, A.M. Jr. (1984) Differential NK cell and macrophage killing of hamster cells infected with nononcogenic or oncogenic adenovirus. Science 224, 612615. Cook, J.L., Walker, T.A., Lewis, A.M. Jr., Ruley, H.E., Graham, F.L. and Pilder, S.H. (1986) Expression of the adenovirus EIA oncogene during cell transformation is sufficient to induce susceptibility to lysis by host inflammatory cells. Proc. Natl. Acad. Sci. USA 83, 6965-6969. Cook, J.L., May, D.L., Lewis, A.M. Jr. and Walker, T.A. (1987) Adenovirus EIA gene induction of susceptibility to lysis by natural killer cells and activated macrophages in infected rodent cells. J. Virol. 61, 3510 3520. Cook, J.L., Wilson, B.A., Wolf, L.A. and Walker, T.A. (1993) E1A oncogene expression level in sarcoma cells: an independent determinant of cytolytic susceptibility and tumor rejection. Oncogene 8, 625-635. Cousin, C., Winter, N., Gomes, S. and D'Halluin, J.C. (t991) Cellular transformation by enteric adenoviruses. Virology 181,277 287. D'Halluin, J.C. and Lecl6re, V. (1992) The adenovirus E1A gene: immortalizating nuclear oncogene prototype. Bull. Inst. Pasteur 90, 45 65. Dynlacht, B.D., Hoey, T. and Tjian, R. (1991) Isolation of co-activators associated with the TATA-binding protein that mediate transcriptional activation. Cell 66, 563-576. Friedman, D.J. and Ricciardi, R.P. (1988) Adenovirus type 12 E1A gene represses accumulation of MHC class I mRNAs at the level of transcription. Virology 165, 303 305. Gallimore, P.H. and Paraskeva, C. (1979) A study to determine the reasons for differences in the tumorigenicity of rat cell lines transformed by adenovirus 2 and adenovirus 12. Cold Spring Harbor Symp. Quant. Biol. 44, 703-713. Giorda, R., Rudert, W.A., Vavassori, C., Chammbers, W.H., Hiserodt, J.C. and Trucco, M. (1990) NKR-PI, a signal transduction molecule on natural killer cells. Science 249, 1298-1301. Haddada, H., Lewis, A.M. Jr., Sogn, J.A., Colignan, J.E., Cook, J.L., Walker, T.A. and Levine, A.S. (1986) Tumorigenicity of hamster and mouse cells transformed by adenovirus types 2 and 5 is not influenced by the level of class l major histocompatibility antigens expressed on the cells. Proc. Natl. Acad. Sci. USA 83, 9684-9688. Han, K. and Manley, J.L. (1993) Transcriptional repression by the Drosophila even-skipped protein: definition of a minimal repression domain. Genes Dev. 7, 491 503. Huebner, R.J., Casey, M.J., Chanock, R.M. and Schell, K. (1965) Tumors induced in hamsters by a strain of adenovirus type 3: sharing of tumor antigens and 'neoantigens' with those produced by adenovirus type 7 tumors. Proc. Natl. Acad. Sci. USA 54, 381-388. Jelinek, T. and Graham, F.L. (1992) Recombinant human adenoviruses containing hybrid adenovirus type 5 (Ad5)/

L Huvent et al. / Virus Research 45 (1996) 123 134

Adl2 E1A genes: characterization of hybrid E1A proteins and analysis of transforming activity and host range. J. Virol. 66, 4117 4125. Jelinek, T., Pereira, D.S. and Graham, F.L. (1994) Tumorigenicity of adenovirus-transformed rodent cells is influenced by at least two regions of adenovirus type 12 early region IA. J. Virol. 68, 888 896. Jochemsen, A.G., Bos, J.L. and van der Eb, A.J. (1984) The first exon of region E1A genes of adenoviruses 5 and 12 encodes separate functional protein domain. EMBO J. 3, 2923-2927. Karlhofer, F.M., Ribaudo, R.K. and Yokoyama, W.M. (1992) MHC class I alloantigen specificity of Ly-49 ÷ ll-2-activated natural killer cells. Nature (London) 358, 66-68. Kast, W.M., Offringa, R., Peters, P.J., Voordouw, A.C., Meloen, R.H.. van der Eb, A.J. and Melief, C.J. (1989) Eradication of adenovirus El-induced tumors by ElA-specific cytotoxic T lymphocytes. Cell 59, 603-614. Kenyon, D.J. and Raska, K. Jr. (1986) Region EIA of highly oncogenic adenovirus 12 in transformed cells protects against NK but not LAK cytolysis. Virology 155, 644654. Kimelman, D., Miller, J.S., Porter, D. and Roberts, B.E. (1985) Ela regions of the human adenoviruses and of the highly oncogenic simian adenovirus 7 are closely related. J. Virol. 53, 399 409. Krantz, C.K., Routes, B.A., Quinlan, M.P. and Cook, J.L. (1996) E1A second exon requirements for induction of target cell susceptibility to lysis by natural killer cells: Implications for the mechanism of action. Virology 217, 23 32. Leclere, V., Huvent, I., Verwaerde, P., Cousin, C. and D'Halluin, J.C. (1993) Comparison between E1A gene from oncogenic and non-oncogenic adenoviruses in cellular transformation. Arch. Virol. 132, 343 357. Leite, J.P.G., Collart, J.F. and D'Halluin, J.C. (1988) The regulation of E1A: the role of both EIA mRNA products of the subgroup B adenoviruses on the early promoters of the subgroup C adenoviruses. Gene 69, 111 120. Leite, J.P.G., Niel, C. and D'Halluin, J.C. (1986) Expression of the chloramphenicol acetyl transferase gene in human cells under the control of early adenovirus subgroup C promoters: effect of E1A gene products from other subgroups on gene expression. Gene 41, 207 215. Lillie, J.W. and Green, M.R. (1989) Transcription activation by the adenovirus EIA protein. Nature 338, 39-44. Liu. F. and Green, M.R. (1990) A specific member of the ATF transcription factor family can mediate transcription activation by the adenovirus Ela protein. Cell 61, 1217-1224. Liu, F. and Green, M.R. (1994) Promoter targeting by adenovirus Ela through interaction with different cellular DNA-binding domains. Nature 368, 520-525. Mak, S., Mak, 1., Smiley, 1.R. and Graham, F.L. (1979) Tumorigenicity and viral expression in rat cells transformed by Adl2 virions or by EcoRl C fragments of Adl2 DNA. Virology 98, 456 460.

133

Meijer, 1., Jochemsen, A.G., de Wit, C.M., Bos, J.L., Morello, D. and van der Eb, A.J. (1989) Adenovirus type 12 EIA down regulates expression of a transgene under control of a major histocompatibility complex class I promoter: evidence for transcriptional control. J. Virol. 63, 4039-4042. Moretta, L., Ciccone, E., Mingari, M.C., Biassoni, R. and Moretta, A. (1994) Human natural killer cells: origin, clonality, specificity, and receptors. Adv. Immun. 55, 341 380. Pereira, D.S., Rosenthal, K.L. and Graham, F.L. (1995) Identification of adenovirus EIA regions which affect MHC class I expression and susceptibility to cytotoxic T lymphocytes. Virology 211, 268 277. Quinlan, M.P., Whyte, P. and Grodzicker, T. (1988) Growth factor induction by the adenovirus type 5 Ela 12S protein is required for immortalization of primary epithelial cells. Mol. Cell. Biol. 8, 3191--3203. Raulet, D.H. (1994) MHC class l-deficient mice. Adv. Immunol. 55, 381 -421. Reynolds, C.W., Timonen, T. and Herberman, R.B. (1981) Natural Killer (NK) cell activity in the rat. 1. Isolation and characterization of the effector cells. J. lmmunol. 127, 282 289. Routes, J.M. (1992) IFN increases class I MHC antigen expression on adenovirus-infected human cells without inducing resistance to natural killer cell killing. J. Immunol. 149, 2372 2377. Routes, J.M. (1993) Adenovirus E1A inhibits IFN-induced resistance to cytolysis by natural killer cells. J. Immunol. 150, 4315 4322. Sawada, Y., Fohring, B., Shenk, T.E. and Rfiska, K. Jr. (1985) Tumorigenicity of adenovirus-transformed cells: region E1A of adenovirus 12 confers resistance to natural killer cells. Virology 147, 413.-421. Sawada, Y., Raskova, J., Fujinaga, K. and Raska, K. Jr. (1994) Identification of functional domains of adenovirus tumor-specific transplantation antigen in types 5 and 12 by viable viruses carrying chimeric EIA genes. Int. J. Cancer 57, 598 603. Schrier, P.I., Bernards, R.T., Vaessen, R.T.M.J.. Houweling. A. and van der Eb, A.J. (1983) Expression of class I major histocompatibility antigens switched off by highly oncogenic adenovirus 12 in transformed rat cells. Nature (London) 305, 771 775. Sheil, J.M., Gallimore, P.H., Zimmer, S.G. ~md Sopori, M.L. (1984) Susceptibility of Adenovirus 2-transformed rat cell lines to natural killer (NK) cells: direct correlation between NK resistance and in vivo tumorigenesis. J. Immunol. 132, 1578 1582. Shiroki, K., Shimojo, H., Sawada, Y., Uemizu, Y. and Fujinaga, K. (1979) Incomplete transformation of rat cells by a small fragment of Adl2 DNA. Virology 95, 127-136. Telling, G.C. and Williams, J. (1994) Constructing chimeric typel2/type5 adenovirus EIA gene and using them to identify an oncogenic determinant of adenovirus type 12. J. Virol. 68, 877 887.

134

L Huvent et al. / Virus Research 45 (1996) 123-134

Timonen, T. and Saksela, E. (1980) Isolation of human NK cells by density gradient centrifugation. J. Immunol. Methods 36, 285-291. Trentin, J.J., Yabe, Y. and Taylor, G. (1962) The quest for human cancer viruses. Science 137, 835-838. Urbanelli, D., Sawada, Y., Raskova, J., Jones, N.C., Shenk, T. and Raska, K. (1989) C-terminal domain of the adenovirus EIA oncogene product is required for induction of cytotoxic T lymphocytes and tumor-specific transplantation immunity. Virology 173, 607-614. Vaessen, R.T.M.J., Houweling, A., Israel, A., Kourilsky, P.

and van der Eb, A.J. (1986) Adenovirus E1A-mediated regulation of class I MHC expression. EMBO J. 5, 335341. Webster, L.C. and Ricciardi, R.P. (1991) trans-dominant mutants of E1A provide genetic evidence that the zinc finger of the trans-activating domain binds a transcription factor. Mol. Cell. Biol. 11, 4287 4296. Zijlstra, M., Auchincloss, H.J., Loring, J.M., Chase, C.M., Russell, P.S. and Jaeniseh, R. (1992) Skin graft rejection by /]2-microglobulin-deficient mice. J. Exp. Med. 175, 885893.