Expression of Exogenous Wt-P53 Does Not Affect Normal Hematopoiesis: Implications for Bone Marrow Purging

R Scardigli1,2, G Bossi1, G Blandino1,3, M Crescenzi1, S Soddu1 and A Sacchi1 1Molecular Oncogenesis Laboratory, Regina Elena Cancer Institute, CRS, Via delle Messi d’Oro 156, 00158 Rome, Italy

Some gene therapy approaches for cancer treatment undertook a series of experiments to assess whether trans- attempt to transduce onco-suppressor genes into tumor duction of wt-p53 into normal hematopoietic cells is harm- cells. A central problem of this strategy is the targeting of ful. Two different wt-p53-recombinant retroviruses were tumor cells to avoid damage to normal ones. It has been used to infect primary, murine BM cells. Expression of noticed that transduction of wt-p53 into a large number of exogenous wt-p53 in these cells did not affect in vitro col- cancer cells induces tumor suppression. In contrast, some ony formation, and did not induce any observable effects observations suggest that introduction of exogenous wt- on morphology and differentiation. In contrast, the same p53 into nontransformed cells does not impair proliferation. viruses suppressed the tumor phenotype of v-src-transfor- If normal bone marrow (BM) cells are not affected by med 32D cells. These results might open the way to gene wt-p53 transduction, BM purging from p53-responding therapy approaches to leukemias with the p53 gene with- leukemic cells might be achieved in vitro by delivering out the need to target speciﬁcally and uniquely the tumor the wild-type onco-suppressor to all marrow cells. We cells, sparing the normal ones.

Keywords: onco-suppressor; retroviral vector; leukemia; gene therapy; tumor targeting

Introduction mutations.7–9 Various effects ranging from growth arrest7,10 to apoptosis11,12 or differentiation13–15 have been One of the critical steps in tumor gene therapy with described, making the wt-p53 gene a promising tool for tumor-suppressing agents (eg toxins, prodrugs, onco- tumor gene therapy. Interestingly, it was originally suppressor genes) is the transduction of the appropriate observed that introduction of exogenous wt-p53 into non- target cells. Tissue-specific methods for gene transfer or transformed, wt-p53 expressing cells, such as normal rat expression are being developed. Viral vectors are embryo fibroblasts,16,17 and VACO 235 adenoma cells7 directed to target specific cell types through chimeric coat causes no observable phenotype, suggesting that the vari- proteins containing the viral env and eukaryotic ligands, ous effects induced by wt-p53 expression might occur 1 such as EPO, or by bi-specific molecular adaptors that only in tumor cells. 2 recognize both env protein and cellular receptors. It was observed that low-level UV irradiation can acti- Another strategy is based on inserting tissue-specific pro- vate sequence-specific transcription by wt-p53 without an moters into retroviral long terminal repeats (LTRs). These increase in p53 protein levels. This activation can also be modifications drive exogenous gene transcription only in induced by antibodies specific to the C-terminal negative 3 particular tissues. An alternative approach to these stra- regulatory domain of p53 or by small peptides derived tegies might be offered by transduction of genes whose from this domain. It has thus been suggested that post- expression is detrimental only for tumor cells while not translational modifications of this negative regulatory harmful for normal ones. Some observations suggest that domain in vivo are rate-limiting steps for wt-p53 acti- the p53 onco-suppressor might belong to this type of vation.18,19 Recently, we found that expression of tem- gene. perature-sensitive p53 in its wild-type configuration Genetic modifications of the p53 onco-suppressor gene affects neither morphology nor proliferation of nontrans- are the most frequent alterations found in human can- formed, murine 32D myeloid progenitors maintained in 4–6 cers. In the past few years, a large number of experi- their normal culture conditions. In contrast, an acceler- ments have been performed to study the suppressor ated apoptotic cell death was observed after IL-3 with- activities of exogenous wt-p53 expression in tumor cells drawal,20 suggesting that this event might trigger the that have lost their normal alleles or bear p53 post-translational modifications that activate wt-p53, which would otherwise remain inactive in nonstressing conditions, in these nontransformed cells. However, Correspondence: S Soddu or A Sacchi when we transformed 32D cells with different oncogenes, Present addresses: 2Institut de Genetique et de Biologie Moleculaire et Cellulaire, Universite´ Louis Pasteur, Strasbourg, France; and 3Depart- wt-p53 expression was then able to induce tumor-sup- ment of Molecular Cell Biology, The Weizmann Institute, Rehovot, Israel pressing effects without the need for stressing stimuli.21 Received 4 February 1997; accepted 11 July 1997 These results suggest either that transformed cells might Exogenous wt-p53 does not affect normal hemopoiesis R Scardigli et al 1372 directly activate wt-p53 in the absence of stressing conditions, or that they are no longer able to maintain wt- p53 in an inactive form. We sought to take advantage of these findings for a new strategy of leukemia gene therapy. We proposed to purge the BM from leukemia by the functional targeting of tumor cells, while avoiding damage to normal ones, by delivering in vitro wt-p53 to all of the marrow cells. To begin to evaluate the feasibility of this strategy, we set up a series of experiments to verify whether the exogenous wt-p53 gene can be transduced into normal cells without untoward effects. Primary, murine BM cells were infected with two different recombinant retroviruses carrying the human wt-p53 cDNA and the puromycin (pBabe-p53) or the neomycin (pLp53SN) selectable markers. No differences in proliferation, differentiation, and colony formation in the presence of different cytokines were found between the infected and the control cells, indicating that normal murine BM cells can tolerate transduction of an exogenous wt-p53 gene. In contrast, the same wt-p53 expressing retroviruses were able to infect and induce a suppressing effect on 32Dv-src-transformed cells, suggesting the possibility of using a single agent to transduce both normal and tumor cells and obtain detrimental effects only in the latter.

Results

Expression of exogenous wt-p53 is not harmful to 32D myeloid progenitors The murine 32D cell line is constituted by nontumori- genic, myeloid progenitors,22,23 which are dependent on IL-3 for survival and proliferation, and differentiate into granulocytes upon granulocyte colony-stimulating factor (G-CSF) stimulation.24 Because of these characteristics and the low predisposition to develop IL-3-independent Figure 1 32D parental and tsp53-transfectants maintained in the pres- subclones,25 these cells are considered an adequate model ence of IL-3 or induced to differentiate by G-CSF, at the indicated tempera- tures, were cytocentrifuged and stained with May–Grunwald Giemsa to for in vitro hemopoiesis. We found that expression of the evaluate morphological features. Each cell population cultured with IL-3 temperature-sensitive p53Val135 (ts-p53) mutant, that is constituted by undifferentiated blasts, whereas those incubated with G- behaves like wt-p53 at 32°C but not at 37°C, affects CSF are mostly constituted (80% of the alive cells) of differentiated poly- neither morphology nor proliferation of 32D cells main- morphonuclear cells. tained at a permissive temperature in their normal culture conditions.20 To verify whether exogenous wt-p53 murine leukemia virus and carry the selectable markers expression is also compatible with differentiation, we for puromycin or neomycin resistance, respectively. The incubated the ts-p53-expressing 32D-tsp53#6 and 32D- murine ⍀E and GP+E packaging cells were transfected tsp53#10 cells21 at the permissive temperature in the pres- with the different constructs to produce ecotropic retro- ence of G-CSF, to assess their granulocyte maturation, viruses, as described in Materials and methods. compared with parental 32D cells. Granulocytic differen- To confirm that the human p53 protein produced by tiation was morphologically evaluated on cytocentrifuge our packaging cells has wild-type configuration, we preparations fixed and stained with May–Grunwald immunoprecipitated the exogenous p53 with mAbs Giemsa. No difference in differentiation was observed PAb1801 and PAb1620. The first mAb recognizes human between tsp53-expressing and control cells after 12 days p53 in both wild-type and mutant configurations while of incubation at 32°C in the presence of G-CSF (Figure the second recognizes only wild-type configuration. As 1). Thus, exogenous wt-p53 expression interferes neither shown in Figure 2, both packaging cell lines – ⍀E/Babe- with morphology and proliferation, nor with differen- p53 and GP+E/Lp53 – express an exogenous p53 protein tiation of 32D cells. with wild-type immunoreactivity. Mouse primary BM cells were infected by 24 h coculti- Infection by retroviral vectors carrying wt-p53 does not vation with mitomycin-treated packaging cells in the affect the colony forming efficiency of primary BM cells presence of IL-3. The colony forming efficiency of these The results obtained with ts-p53 protein in 32D cells infected BM cells was tested by cultivation in methyl- prompted us to evaluate the effects of exogenous wt-p53 cellulose, in the presence of IL-3, GM-CSF or EPO, and transduction in primary BM cells. We inserted the human puromycin or G418 (see Materials and methods for wt-p53 cDNA into two different retroviral vectors – details). As shown in Table 1, cells infected with the wt- pBabe-puro and pLXSN – which derive from Moloney p53 recombinant retroviruses form colonies with the Exogenous wt-p53 does not affect normal hemopoiesis R Scardigli et al 1373

Figure 2 Biochemical analysis of exogenous p53 proteins in packaging cells. Cell lysates deriving from the indicated cells, and containing equal amounts of proteins were immunoprecipitated with the following Abs: 1, protein G as negative control; 2, PAb1801 (recognizing human wt- and mutant-p53); 3, PAb1620 (recognizing human wt-p53). Immunocomplexes were separated on a denaturing 10% polyacrylamide gel. Western blot analysis was performed with Ab7 polyclonal p53 anti-serum. The position of p53 protein is indicated. Human NB4 cells that express a mutant p53 protein were used as control for the immunoprecipitation. C2C12tsp53 cells that constitutively express the tsp53 protein were used as positive control for Western blotting.

Table 1 Colony forming efﬁciency of p53 recombinant retrovirus-infected BM cells

Virus Selection drug No. of colonies No. of colonies No. of colonies in IL-3 in GM-CSF in EPO

12345a 12a 12a

Mock Puromycin 0 0 0 0 0 0 0 0 0 Mock None 40 80 30 40 118 105 17 29 22 Babe-p53 Puromycin 20 16 20 12 15 12 10 11 3 Babe-puro Puromycin 28 28 17 16 12 10 0 20 10 Mock G418 0 0 0 0 0 0 Mock None 15 20 93 35 24 22 Lp53SN G418 5 15 23 25 5 7 LXSN G418 2 10 14 20 1 3

aExperiment number. same efficiency as those infected with the viruses carry- p53-, control- and mock-infected cells (Figure 3a, EPO). ing only the selectable markers. However, the number of To confirm granulocytic–macrophage differentiation, the colonies obtained from infected cells was only one-quar- capacity to phagocytose and reduce NBT was also ter of the number obtained with mock-infected cells. To assessed on cells infected with either LXSN or Lp53SN test whether this reduction in colony formation was due and grown in the presence of GM-CSF (Figure 3b, upper to low infection efficiency or to infection-mediated tox- panel). Fifty per cent of the colonies analyzed had NBT- icity, BM cells were plated, after infection by LXSN or positive cells in both LXSN- and Lp53SN-infected popu- Lp53SN, in methylcellulose with IL-3 in the presence or lations. The percentage of NBT-positive cells was similar absence of the selection drug G418. In the absence of in each colony analyzed independent of the virus used G418 the number of colonies was similar between mock- for the infection (Figure 3b, lower panel). Thus, infection and retrovirus-infected cells indicating that the infection with wt-p53 recombinant retroviruses does not induce per se is not toxic (data not shown). phenotypic changes, and does not affect in vitro colony To evaluate whether the infected cells that form colo- formation and differentiation capacity of primary BM nies in methylcellulose differentiate normally, cytospin cells. preparations of single colonies were prepared and stained. May–Grunwald Giemsa staining was used to Detection of the exogenous wt-p53 gene in cell colonies assess cell morphology in the presence of IL-3 or GM- The recombinant retroviruses used for BM cell infection CSF. Typical BM monocytes with large cytoplasms carry the wt-p53 gene and the selection markers in the (Figure 3a, IL-3) and polymorphonuclear cells (Figure 3a, region comprised between the LTRs. However, to ascer- GM-CSF) were found in either uninfected and recombi- tain that our cells had integrated and expressed the nant retrovirus-infected cells. Benzidine reactivity was exogenous wt-p53 gene, the presence of exogenous p53 employed to verify erythroid differentiation in the pres- cDNA was verified on genomic DNA from single colo- ence of EPO. No differences were observed between wt- nies grown in methylcellulose, by PCR amplification of Exogenous wt-p53 does not affect normal hemopoiesis R Scardigli et al 1374 (Figure 4, lower panel). The p53 cDNA was amplified from 25 of 32 colonies of Lp53SN-infected BM cells tested. Only in these 25 positive colonies was it also possible to amplify the endogenous cyclin-B1 gene as control (data not shown), indicating that seven colonies were p53 negative for PCR technical problems. Interestingly, a similar result was obtained from the colonies (10 of 10) grown in the absence of G418 selection suggesting that the reduction of colony formation observed after infection and selection (Table 1) is due to inefficient recombinant retrovirus gene expression rather than infection. Further studies are in progress better to clarify this issue. Expression of exogenous wt-p53 protein was evaluated by indirect immunofluorescence on bulk-infected BM cells or on cell colonies grown in methylcellulose. In the first condition, the infected BM cells were selected for 2 days in puromycin- or G418-containing medium, separated from dead cells by density gradient centrifugation, and cytocentrifuged on to slides. For colony analysis, the colonies grown, after infection, in methylcellulose in the presence of IL-3, GM-CSF or EPO, and the selection drug, were recovered and cytocentrifuged. The cytospin preparations were fixed and reacted with anti-p53 mAb DO- 1, which preferentially recognizes p53 of human origin. Expression of exogenous p53 was found both in bulk- infected and single-colony cells (Figure 5A). Moreover, exogenous p53 expression was confirmed in bulk- infected BM cells by Western blot analysis (Figure 5B).

Exogenous wt-p53-expressing BM cells can form spleen colonies The experiments described above indicate that BM cells can tolerate exogenous wt-p53 gene transduction when cultured in vitro. To verify whether these cells can survive and proliferate in vivo, we evaluated whether they maintained the capacity to produce spleen colonies. Sub- lethally irradiated F1-hybrid mice were injected with 5 × 105 syngeneic cells that had been mock-infected or infected with Babe-p53 or Babe-puro recombinant retroviruses. Ten days after injection, spleen colonies were dis- sected away from surrounding tissues, cultured in suspension for 2 days in the presence of IL-3 and puromycin, separated from dead cells by density gradient, and plated in methylcellulose in the presence of IL-3 and puromycin. No drug-resistant colonies grew from the mice injected Figure 3 (a) Primary uninfected bone marrow cells (BM), bone marrow cells infected with Babe-puro (puro), or Babe-p53 (p53) were plated in with mock-infected BM cells while 25, 16, 31, 27 and 35 methylcellulose in the presence of the indicated cytokines. Cells from single colonies, or 30, 27 and 21 colonies grew from the spleens colonies grown with IL-3 or GM-CSF were cytocentrifuged and stained of the mice injected with Babe-p53- or Babe-puro-infected with May–Grunwald Giemsa to evaluate cell morphology. Cells from BM cells, respectively. This result indicates that wt-p53 single colonies grown with EPO were reacted with benzidine before cyto- recombinant retrovirus-infected BM cells maintain the centrifugation to detect hemoglobin synthesis. (b) Primary bone marrow capacity to proliferate in vivo and colonize mouse spleen. cells infected with LXSN or Lp53SN were plated in methylcellulose in the presence of GM-CSF. Granulocytic–macrophage differentiation was confirmed by NBT reactivity. Two positive cells from the indicated infec- Exogenous wt-p53 carried by recombinant retrovirus is tions, rich in black granules are shown (upper panel). The percentages of active in stressing and transforming conditions NBT-positive cells observed in different single colonies are also reported The purpose of our study was to verify the absence of (lower panel). any phenotypic changes induced by exogenous wt-p53 expression in normal BM cells. To determine whether the lack of effect in normal cells was due to the capacity of DNA sequences corresponding to p53 exons 7 and 8. The these cells to tolerate transduction of the exogenous wt- two primers used for PCR encompass a 288-bp fragment p53 gene, rather than to an alteration in the protein enco- of p53 cDNA. This fragment was amplified only in the ded by our recombinant retroviruses, we sought to infect colonies derived from BM cells infected with the retro- parental 32D and 32Dv-src-transformed cells to ascertain viruses carrying the wt-p53 gene (Figure 4, upper panel). exogenous wt-p53 activity. We had already shown that Southern blot hybridization was performed with a p53 exogenous wt-p53 expression in 32D cells accelerates probe to confirm the specificity of the amplified products apoptosis upon IL-3 withdrawal20 while in 32Dv-src- Exogenous wt-p53 does not affect normal hemopoiesis R Scardigli et al 1375

Figure 4 PCR analysis of transduced p53 cDNA. Primary uninfected bone marrow cells (BM), or bone marrow cells infected with Babe-puro (puro), LXSN, Babe-p53 (p53), or Lp53SN were plated in methylcellulose in the presence of the indicated cytokines. Upper panel: cells from single colonies were processed for ampliﬁcation of a human p53 sequence of 288 bp. Lower panel: Southern blot hybridization of the gel shown in the upper panel with a p53 probe. SV-p53 is a plasmid carrying the entire human p53 cDNA that was used as positive control.

Figure 5 Exogenous wt-p53 expression in BM cells. (A) Cytocentrifuge preparations were obtained with the following cells: BM cells infected with Babe-p53 (a, b, e and f), BM cells infected with Babe-puro (c, d, g and h). a, b, c and d are cells maintained as bulk population; e, f, g and h are individual colonies grown in methylcellulose. The cells were fixed and permeabilized, and exogenous human p53 was detected by indirect immunofluorescence with mAb DO-1 (a, c, e and g); b, d, f and h are the relative phase-contrast fields. (B) BM cells infected with LXSN or Lp53SN and maintained as a bulk population were lysed, run on 10% SDS-PAGE, blotted, and reacted with Ab7 polyclonal p53 anti-serum, as described in Figure 2. transformed cells it induces monocytic differentiation temperature,20 acceleration of apoptosis was found in the without affecting cell viability and proliferation rates.21 Lp53SN-infected 32D cells compared with LXSN-infected Thus, 32D and 32Dv-src cells were infected with LXSN control (Figure 6c). or Lp53SN, and selected in G418-containing media. To Monocytic differentiation of 32Dv-src cells was perfor- verify whether exogenous p53 protein was in wild-type med by the ␣-naphthyl esterase lympho–monocytic configuration, 32Dv-src-infected cells were lysed and p53 marker. Approximately 20% of exogenous wt-p53- protein was immunoprecipitated by mAbs PAb240, expressing cells were positive for ␣-naphthyl esterase which recognizes mutant p53, or PAb1620, which recog- (data not shown), as already reported for ts-p53-express- nizes wt-p53. As shown in Figure 6a, only PAb1620 was ing and pBabe-p53-infected 32Dv-src cells.21 Thus, p53 able to precipitate the exogenous p53 protein indicating protein encoded by our recombinant retroviruses is in that it is expressed in 32Dv-src cells and is in wild-type wild-type configuration and is able to induce biological configuration. To verify whether exogenous p53 protein effects in the appropriate conditions (ie growth factor was correctly localized in the nucleus, 32D-infected cells withdrawal in nontransformed cells, or expression in were analyzed by indirect immunofluorescence. As tumor cells). shown in Figure 6b, fluorescence was mostly present in the nuclei, as expected for wt-p53 protein. The apoptotic Discussion rate of 32D cells after IL-3 withdrawal was evaluated by the trypan blue exclusion test. As reported for ts-p53- We evaluated the effects induced by infection with wt- expressing 32D cells maintained at the permissive p53 recombinant retroviruses in nontransformed, Exogenous wt-p53 does not affect normal hemopoiesis R Scardigli et al 1376 Moreover, preliminary in vivo experiments indicate that wt-p53 retrovirus-infected BM cells injected i.v. in sublethally irradiated mice can produce puromycin-resistant spleen colonies. In contrast, infection of nontransformed 32D and transformed 32Dv-src cells by the same retroviruses induced acceleration of apoptosis upon IL-3 withdrawal and monocytic differentiation, respectively, as already observed after ts-p53 overexpression at the permissive temperature.20,21 Taken together, these results strongly indicate that transduction of wt-p53 is not harmful for normal cells in the absence of stressing stimuli, whereas it has a suppressor activity on tumor cells. It was previously observed that infection by wt-p53 recombinant adenovirus of a nontransformed fibroblast cell line or normal human bronchial epithelial cells in vitro, or normal lung cells in vivo, does not modify the in vitro cell proliferation rate, and is not toxic in vivo for the lung.26–29 Our data extend these observations in a cell system – the hematopoietic one – whose homeostasis is regulated by a precise balance of proliferation, apoptosis and differentiation. They might open the way to BM purging from leukemias by p53 gene delivery without the need to target specifically and uniquely the tumor cells. Indeed, on the basis of our data, it can be speculated that infection of both normal BM and leukemic cells should result in survival of the normal cells and suppression of tumorigenicity of the leukemic ones. This approach will probably require a higher efficiency of transduction/expression than the 25% we obtained in our drug-selected cells. We are currently trying different infection protocols in the presence of different cytokines. Moreover, in vivo BM reconstitution experiments are in progress to verify the complete safety of exogenous wt- p53 expression. The molecular mechanisms underlying the different responses to wt-p53 expression of normal versus tumor cells are unknown. It has been recently shown that in normal fibroblasts, upon induction of DNA damage, p53 is able to bind DNA even in the absence of increased protein levels,18,19 indicating that p53 is present in normal cells in an inactive form. We previously showed that wt- p53 overexpression induced no effects in nontransformed 32D cells when they were incubated in the presence of the appropriate growth factors. In contrast, an accelerated apoptotic rate was observed in the same cells after IL-3 Figure 6 Wild-type structure and function of retroviral carried p53. (a) withdrawal.20 These results suggested that nontransfor- Cell lysates deriving from the cells indicated, and containing equal amounts of proteins were immunoprecipitated with the following Abs: 1, med 32D cells, as well as normal BM cells, can tolerate protein G as negative control; 2, PAb240 (recognizing mutant-p53); 3, exogenous wt-p53 as long as the action of the endogen- PAb1620 (recognizing human wt-p53). Total lysate from GP+E-Lp53SN ous p53 is not required (eg during deprivation of survival was loaded as positive control for Western blot analyses. SDS/PAGE and factors). Interestingly, when we transformed these 32D- Western blotting were performed as in Figure 2. Human HEL92.1.7 cells ts-p53 expressing cells with different oncogenes, they that express a mutant p53 protein were used for PAb240 immunoprecipit- became sensitive to wt-p53 action by directly responding ation control. (b) Cytospin preparations of a: 32D-Lp53SN or b: 32D- 21 LXSN were analyzed by indirect immunofluorescence with anti-p53 mAb with growth arrest or differentiation to its expression. as described in Figure 2. (c) Effect of exogenous p53 expression on viability It is possible that the apparatus necessary to activate of 32D cells upon IL-3 withdrawal. The cells, 4 × 104/ml, indicated in the endogenous p53 protein upon DNA damage can also legends were plated without IL-3, and viability was determined, every control exogenous wt-p53 expression in normal cells, 12 h, by trypan blue exclusion. The mean of three different experiments while tumor cells might lack this type of regulation. is reported. Another possible divergence between normal and transformed cells in exogenous wt-p53 regulation might be myeloid 32D progenitors and primary BM cells. We due to its degradation. Normal cells might be more active found that transduction of the wt-p53 gene does not affect in degrading exogenous p53 protein than their transfor- morphology, proliferation, and differentiation capacity of med counterparts. The two mechanisms are not mutually both 32D and primary BM cells, as long as they are main- exclusive. Understanding the molecular mechanisms that tained in the presence of growth or differentiation factor determine the different responses of normal and tumor supplies (IL-3, G-CSF, GM-CSF or EPO and serum). cells to wt-p53 expression will likely help in developing Exogenous wt-p53 does not affect normal hemopoiesis R Scardigli et al 1377 new gene therapy approaches based on the onco-sup- Polymerase chain reaction (PCR) pressor p53 gene. PCR amplification was performed on cell lysates using a PCR kit (Perkin Elmer Cetus, Norwalk, CT, USA), follow- Materials and methods ing the manufacturer’s instructions. Single colonies derived from methylcellulose cultures were isolated with ␮ m Cell cultures a micropipette, lysed in 50 l of lysis buffer (10 m Tris- m Parental 32Dcl.3 murine hematopoietic progenitors and HCl pH 8; 0.5 m EDTA; 0.0001% SDS; 0.01% Triton ␮ ° their derivatives 32D-tsp53#6 and 32D-tsp53#10, which X100; 100 g/ml proteinase K), incubated for 1 h at 55 C, ␮ express the temperature-sensitive p53Val135 mutant,21 and used in a 100 l PCR reaction after proteinase K heat were cultured in RPMI-1640 medium (BIO-Whittaker, inactivation. A fragment of 288 bp, corresponding to ′ Walkersville, MD, USA) supplemented with 10% heat- parts of exons 7 and 8, was amplified using a 5 primer ′ inactivated fetal bovine serum (FBS) (GIBCO BRL, Life complementary to exon 7 5 (TAGTGTGGTGGT ′ ′ Technology, Grand Island, NY, USA), 5% conditioned GCCCTATGAGCCG)3 , and a 3 primer complementary ′ ′ medium from the murine myelomonocytic cell line to exon 8 3 (CCTCGTGATTCGCTCGTGACGGGTT)5 . ␮ WEHI-3B as a source of crude IL-3 (IL-3/CM),30 and gen- Samples of 10 l from each PCR reaction were visualized tamicin (BIO-Whittaker). WEHI-3B and 32Dv-src cells21 on 1.5% agarose gel by ethidium-bromide staining. The were cultured in RPMI-1640 with 10% FBS. ⍀E31 and amplification specificity was controlled by blotting the GP+E32 ecotropic packaging cell lines were cultured in PCR products on to nylon membranes and hybridizing DMEM medium (BIO-Whittaker) supplemented with with the human p53 cDNA by standard procedures. 10% FBS. Indirect immunofluorescence × 4 Retroviral vectors and packaging cells Approximately 4 10 cells were spun on to a glass slide The construction of pBabe-puro and pLXSN retroviral with a cytocentrifuge. Cytospin preparations were air- vectors has been described.31,33 Briefly, pBabe-puro and dried, fixed with 2% formaldehyde for 10 min at room pLXSN are defective Moloney murine leukemia virus- temperature (RT), permeabilized with 0.1% Triton X100 for 1 min, blocked for 30 min at RT in PBS, containing based vectors containing, respectively, the bacterial puro ° or neo gene downstream of the SV40 early promoter. The 3% FBS and incubated for 1 h at 37 C with mAb DO–1 (Ab-6 from Oncogene Science, Uniondale, NY, USA), pBabe-p53 and pLp53SN vectors were obtained by 34 inserting the human wt-p53 cDNA, derived from which reacts preferentially with human p53 protein. After washing, cytospins were incubated for 45 min at pCMVp53,7 into the unique BamHI site of pBabe-puro 37°C with biotin-conjugated goat anti-mouse IgG (Pierce and pLXSN vectors. Europe, Oud Beijerland, The Netherlands), and sub- Virus-producing cell lines were generated by stable sequently incubated for 45 min at 37°C with fluorescein- transfection of ⍀E and GP+E ecotropic packaging cells conjugated streptavidin (Becton Dickinson, Cedex FR). with the retroviral vectors. Vector DNAs were trans- ⍀ + fected into E and GP E cells by calcium-phosphate pre- Morphological and differentiation analyses cipitation. Selection of stably transfected mixed popu- × 4 ␮ Approximately 4 10 cells cultured in suspension, or lations was performed with 2 g/ml of puromycin or 1.5 cells derived from single colonies grown in methylcellu- mg/ml of G418 for 2 weeks. Retroviral titrations were lose, were spun on to glass slides with a cytocentrifuge. 31 performed as described. For morphological analysis, cytospin preparations were fixed and stained with May–Grunwald Giemsa (Sigma, Retroviral infections St Louis, MO, USA) and observed with a light micro- Primary BM cells were obtained from the long bones of scope. To evaluate granulocytic–macrophage differen- dead C57Bl/6 mice. Approximately 1 × 106 BM cells were ␮ tiation of bone marrow cells the ability to phagocyte and incubated for 1 h with 8 g/ml polybrene and then reduce nitroblue tetrazolium (NBT) was measured. Single infected in IL-3-containing medium by in vitro coculti- colonies were resuspended in 100 ␮l of RPMI and then × 6 vation with 1 10 mitomycin-treated, retrovirus produc- reacted with 100 ␮l of NBT solution (Sigma), as ing packaging cells. After 16 h of cocultivation, BM cells described.15 NBT positivity was determined on May– were removed from the packaging monolayer and Grunwald Giemsa-stained cytospins observed with a replated in fresh IL-3-containing medium. After a further light microscope. For evaluation of erythroid differen- 24 h, BM cells were assayed for in vitro colony formation. tiation, the cells were reacted with benzidine as 35 Colony formation described, before cytocentrifugation. Positive blue stained cells were observed with a light microscope. Infected BM cells (1 × 105) were suspended in 2 ml of methylcellulose culture mixture (Stemcell Technologies, Spleen colony formation c/o Terry Fox Laboratory, Vancouver, Canada), in the Primary BM cells were obtained from DBA-2 mice pre- ␮ presence of IL-3, GM-CSF or EPO, and 2 g/ml puromy- treated with 5-FU (150 mg/kg i.v. 48 h before harvesting cin or 1.5 mg/ml G418, and plated into two 35-mm Petri BM cells) and infected as described above. Infected cells dishes. Optimal cytokine concentrations were determined were selected for 48 h in the presence of puromycin. Live by dose–response colony formation assays: 10% of IL- cells were separated from dead ones by density gradient, 3/CM; 10 U/ml of recombinant m-EPO (Boehringer- and injected i.v. (5 × 105 per mouse) in sublethally X- Mannheim Italia, Milan, Italy), plus 5% IL-3/CM; 1.63 irradiated (6 Gy) DBA-2 mice. U/ml of recombinant m-GM-CSF (Genetics Institute, Cambridge, MA, USA). After 12 days of incubation, BM Immunoprecipitation and Western blotting colonies were counted under an inverted microscope and Approximately 1 × 106 ⍀E and GP+E ecotropic packaging analyzed for morphology and exogenous p53 expression. cell lines were incubated for 30 min on ice in lysis buffer Exogenous wt-p53 does not affect normal hemopoiesis R Scardigli et al 1378 (20 mm Tris pH 7.8; 50 mm NaCl; 5 mm EDTA; 0.5% Na- 13 Shaulsky G, Goldfinger N, Peled A, Rotter V. Involvement of deoxycholate; 0.2% SDS; 1% Nonidet P40; 1 mm PMSF; wild-type p53 in pre-B-cell differentiation in vitro. Proc Natl Acad 16.5 ␮g/ml aprotinin). Lysates containing equivalent Sci USA 1991; 88: 8982–8986. amounts of proteins were subjected to immunoprecipit- 14 Feinstein E, Gale RP, Reed J, Canaani E. Expression of the normal p53 gene induces differentiation of K562 cells. Oncogene ation by incubation for 90 min at 4°C with anti-p53 mono- 1992; 7: 1853–1857. clonal antibodies PAb1801, PAb240 or PAb1620 15 Soddu S et al. Wild-type p53 gene expression induces granulo- (Oncogene Science, Uniondale, NY, USA). Immuno-com- cytic differentiation of HL-60 cells. Blood 1994; 83: 2230–2237. plexes were precipitated with ImmunopurePlus protein 16 Finlay CA, Hinds PW, Levine AJ. The p53 proto-oncogene can G (Pierce Europe), washed twice with lysis buffer, boiled act as a suppressor of transformation. Cell 1989; 57: 1083–1093. for 5 min in sample buffer, analyzed on 10% SDS/PAGE, 17 Eliyahu D et al. Wild-type p53 can inhibit oncogene-mediated and blotted on to nitrocellulose membranes (Bio-Rad, focus formation. Proc Natl Acad Sci USA 1989; 86: 8763–8767. Hercules, CA, USA). Filters were probed by rocking for 2 18 Hupp TR, Meek DW, Midgley CA, Lane DP. Regulation of the h at RT with p53 polyclonal anti-serum PAb-7 (Oncogene specific DNA binding function of p53. Cell 1992; 71: 875–886. Science) diluted in TBST (20 mm Tris-HCl pH 7.8, 150 19 Hupp TR, Sparks A, Lane DP. Small peptides activate the latent m sequence-specific DNA binding function of p53. Cell 1995; 83: m NaCl, 0.02% Tween 20). Immunoreactivity was 237–245. determined using the ECL-reaction kit (Amersham, 20 Blandino G et al. Wild-type p53 modulates apoptosis of normal, Arlington Heights, IL, USA), following the manufac- IL-3 deprived, hematopoietic cells. Oncogene 1995; 10: 731–737. turer’s instructions. 21 Soddu S et al. Wt-p53 induces diverse effects in 32D cells expressing different oncogenes. Mol Cell Biol 1996; 16: 487–495. 22 Greenberger JS et al. Demonstration of permanent factor-depen- Acknowledgements dent multipotential (erythroid-neutrophil-basophil) hematopo- We wish to thank Dr B Calabretta for helpful suggestions ietic progenitor cell lines. Proc Natl Acad Sci USA 1983; 80: and discussion at the beginning of this project. We are 2931–2935. grateful to Drs H Land, F Belardelli and C Peschle for 23 Migliaccio G et al. Selection of lineage-restricted cell lines immortalized at different stages of hematopoietic differentiation generous gifts of cells, plasmids and viruses. This work from the murine cell line 32D. J Cell Biol 1989; 109: 833–841. was partially supported by Associazione Italiana per la 24 Valtieri M et al. Cytokine-dependent granulocytic differen- Ricerca sul Cancro, Progetto finalizzato Italia-USA, and tiation. Regulation of proliferative and differentiative responses Ministero Sanita`. RS was the recipient of an AIRC fellow- in a murine progenitor cell line. J Immunol 1987; 138: 3829–3835. ship. 25 Askew DS, Ashmun RA, Simmons BC, Cleveland JL. Constitut- ive c-myc expression in an IL-3-dependent myeloid cell line sup- presses cell cycle arrest and accelerates apoptosis. Oncogene References 1991; 6: 1915–1922. 1 Kasahara N, Dozy AM, Kan YW. Tissue-specific targeting of 26 Clayman GL et al. In vivo molecular therapy with p53 retroviral vectors through ligand-receptor interactions. Science adenovirus for microscopic residual head and neck squamous 1994; 266: 1373–1376. carcinoma. Cancer Res 1995; 55: 1–6. 2 Etienne-Julan M et al. Cell targeting by murine recombinant 27 Wills KN et al. Development and characterization of recombi- retroviruses. Bone Marrow Transplant 1992; 9 (Suppl. 1): 139–142. nant adenoviruses encoding human p53 for gene therapy of can- 3 Ferrari G et al. A retroviral vector containing a muscle-specific cer. Hum Gene Ther 1994; 5: 1079–1088. enhancer drives gene expression only in differentiated muscle 28 Zhang W-W et al. Safety evaluation of Ad5CMV-p53 in vitro and fibers. Hum Gene Ther 1995; 6: 733–742. in vivo. Hum Gene Ther 1995; 6: 155–164. 4 Baker SJ et al. Chromosome 17 deletions and p53 gene mutations 29 Lesoon-Wood LA et al. Systemic gene therapy with p53 reduces in colorectal carcinomas. Science 1989; 244: 217–221. growth and metastases of a malignant human breast cancer in 5 Levine AJ, Momand J, Finlay CA. The p53 tumor suppressor nude mice. Hum Gene Ther 1995; 6: 395–405. gene. Nature 1991; 351: 453–456. 30 Lee JC, Hapel AJ, Ihle JN. Constitutive production of a unique 6 Hollstein M, Sidransky D, Vogelstein B, Harris CC. p53 lymphokine (IL-3) by the Wehi-3 cell line. J Immunol 1982; 128: Mutations in human cancer. Science 1991; 253: 49–53. 2393–2398. 7 Baker SJ et al. Suppression of human colorectal carcinoma cell 31 Morgenstern JP, Land H. Advanced mammalian gene transfer: growth by wild-type p53. Science 1990; 249: 912–915. high titre retroviral vectors with multiple drug selection mark- 8 Chen P, Chen Y, Bookstein R, Lee WH. Genetic mechanisms of ers and a complementary helper-free packaging cell line. Nucleic tumor suppression by the human p53 gene. Science 1990; 250: Acids Res 1990; 18: 3587–3596. 1576–1580. 32 Markowitz D, Goff S, Bank A. A safe packaging line for gene 9 Ozturk M, Ponchel F, Pulsleux A. p53 As a potential target in transfer: separating viral genes on two different plasmids. J Virol cancer therapy. Bone Marrow Transplant 1992; 9 (Suppl. 1): 1988; 62: 1120–1124. 164–170. 33 Miller AD, Rosman GJ. Improved retroviral vectors for gene 10 El-Deiry WS et al. WAF-1, a potential mediator of p53 tumor- transfer and expression. Biotechniques 1989; 7: 980–990. suppression. Cell 1993; 75: 817–825. 34 Gannon JV, Greaves R, Iggo R, Lane DP. Activating mutations 11 Yonish-Rouach E et al. Wild-type p53 induces apoptosis of in p53 produce a common conformational effect. A monoclonal myeloid leukaemic cells that is inhibited by interleukin-6. Nature antibody specific for the mutant form. EMBO J 1990; 9: 1595– 1991; 352: 345–347. 1602. 12 Ryan JJ, Danish R, Gottlieb CA, Clarke MF. Cell cycle analysis 35 Orkin SH, Harosi FI, Leder P. Differentiation in erythro- of p53 induced cell death in murine erythroleukemia cells. Mol leukemic cells and their somatic hybrids. Proc Natl Acad Sci USA Cell Biol 1993; 13: 711–719. 1975; 72: 98–102.