Search

Mitogen-activated protein kinases (MAPK) become activated during the meiotic maturation of oocytes from many species; however, their molecular targets remain unknown. This led us to characterize the activation of the ribosomal subunit S6 kinase of Mr 82 X 10(3) - 92 X 10(3) (p90rsk; a major substrate of MAPK in somatic cells) in maturing mouse oocytes and during the first cell cycle of the mouse embryo. We assessed the phosphorylation state of p90rsk by examining the electrophoretic mobility shifts on immunoblots and measured the kinase activity of immunoprecipitated p90rsk on a S6-derived peptide. Germinal vesicle stage (GV) oocytes contained a doublet of Mr 82 x 10(3) and 84 x 10(3) with a low S6 peptide kinase activity (12% of the maximum level found in metaphase II oocytes). A band of Mr 86 x 10(3) was first observed 30 minutes after GV breakdown (GVBD) and became prominent within 2 to 3 hours. MAPK was not phosphorylated 1 hour after GVBD, when the p90rsk-specific S6 kinase activity reached 37 % of the M II level. 2 hours after GVBD, MAPK became phosphorylated and p90rsk kinase activity reached 86% of the maximum level. The p90rsk band of Mr 88 x 10(3), present in mature M II oocytes when S6 peptide kinase activity is maximum, appeared when MAPK phosphorylation was nearly complete (2.5 hours after GVBD). In activated eggs, the dephosphorylation of p90rsk to Mr 86 X 10(3) starts about 1 hour after the onset of pronuclei formation and continues very slowly until the beginning of mitosis, when the doublet of Mr 82 X 10(3) and 84 X 10(3) reappears. A role for a M-phase activated kinase (like p34cdc2) in p90rsk activation was suggested by the reappearance of the Mr 86 X 10(3) band during first mitosis and in 1-cell embryos arrested in M phase by nocodazole. The requirement of MAPK for the full activation of p90rsk during meiosis was demonstrated by the absence of the fully active Mr 88 X 10(3) band in maturing c-mos -/- oocytes, where MAPK is not activated. The inhibition of kinase activity in activated eggs by 6-DMAP after second polar body extrusion provided evidence that both MAPK- and p90rsk-specific phosphatases are activated at approximately the same time prior to pronuclei formation.

Mos is normally expressed during oocyte meiotic maturation in vertebrates. However, apart from its cytostatic factor (CSF) activity, its precise role during mouse meiosis is still unknown. First, we analyzed its role as a MAP kinase kinase kinase. Mos is synthesized concomitantly with the activation of MAP kinase in mouse oocytes. Moreover, MAP kinase is not activated during meiosis in oocytes from mos -/- mice. This result implies that Mos is necessary for MAP kinase activation in mouse oocytes. Raf-1, another MAP kinase kinase kinase, is already present in immature oocytes, but does not seem to be active when MAP kinase is activated. Moreover, the absence of MAP kinase activation in mos -/- oocytes demonstrates that Raf-1 cannot compensate for the lack of Mos. These results suggest that Raf-1 is not involved in MAP kinase activation. Second, we analyzed the organization of the microtubules and chromosomes in oocytes from mos -/- mice. We observed that during the transition between two meiotic M-phases, the microtubules and chromosomes evolve towards an interphase-like state in mos -/- oocytes, while in the control mos +/- oocytes they remain in an M-phase configuration, as in the wild type. Moreover, after spontaneous activation, the majority of mos -/- oocytes are arrested for at least 10 hours in a third meiotic M-phase where they exhibit monopolar half-spindles. These observations present the first evidence, in intact oocytes, of a role for the Mos/.../MAP kinase cascade in the control of microtubule and chromatin organization during meiosis.

Cystic fibrosis is a disease characterized by abnormalities in the epithelia of the lungs, intestine, salivary and sweat glands, liver, and reproductive systems, often as a result of inadequate hydration of their secretions. The primary defect in cystic fibrosis is the altered activity of a cAMP-activated Cl- channel, the cystic fibrosis transmembrane conductance regulator (CFTR) channel. However, it is not clear how a defect in the CFTR Cl- channel function leads to the observed pathological changes. Although much is known about the structural properties and regulation of the CFTR, little is known of its relationship to cellular functions other than the cAMP-dependent Cl- secretion. Here we report that cell volume regulation after hypotonic challenge is also defective in intestinal crypt epithelial cells isolated from CFTR -/- mutant mice. Moreover, the impairment of the regulatory volume decrease in CFTR -/- crypts appears to be related to the inability of a K+ conductance to provide a pathway for the exit of this cation during the volume adjustments. This provides evidence that the lack of CFTR protein may have additional consequences for the cellular function other than the abnormal cAMP-mediated Cl- secretion.

Three recent reports describe the first in vivo attempts at fetal gene therapy. The results underline the need for more intensive studies of the scientific and ethical implications of this new and perhaps more preventive approach to gene therapy.

The HOX11 homeobox gene was identified via the translocation t(10;14) in T cell leukaemia. To determine the function of this gene in mice, null mutations were made using homologous recombination in ES cells to incorporate lacZ into the hox11 transcription unit. Production of beta-galactosidase from the recombinant hox11 allele in +/- mutants allowed identification of sites of hox11 expression which included the developing spleen. Newborn hox11 -/- mice exhibit asplenia. Spleen formation commences normally at E11.5 in hox11 -/- mutant embryos but the spleen anlage undergoes rapid and complete resorption between E12.5 and E13.5. Dying spleen cells exhibit molecular features of apoptosis, suggesting that programmed cell death is initiated at this stage of organ development in the absence of hox11 protein. Thus hox11 is not required to initiate spleen development but is essential for the survival of splenic precursors during organogenesis. This function for hox11 suggests that enhanced cell survival may result from the t(10;14) which activates HOX11 in T cell leukaemias, further strengthening the association between oncogene-induced cell survival and tumorigenesis.

We have generated mice carrying the most common mutation in cystic fibrosis (CF), delta F508, within the cystic fibrosis (Cftr) gene. Mutant animals show pathological and electrophysiological changes consistent with a CF phenotype. delta F508-/- mice die from peritonitis and show deficiencies in cAMP-activated electrogenic Cl- transport. These mice produce delta F508 transcripts and show the temperature-dependent trafficking defect first described for the human delta F508 CFTR protein. A functional CFTR Cl- channel not demonstrated by null CF mice or present at 37 degrees C was detected following incubation of epithelial cells at 27 degrees C. Thus, these mice are an accurate delta F508 model and will be valuable for testing drugs aimed at overcoming the delta F508 trafficking defect.

Calcium-activated anion secretion is elevated in the pancreatic ductal epithelium of transgenic cf/cf mice which lack the cystic fibrosis transmembrane conductance regulator (CFTR). To elucidate whether this effect is due to increased activity of calcium-activated chloride channels, we have studied the relationship between CFTR and calcium-activated chloride currents in pancreatic duct cells isolated from Cambridge cf/cf mice. CFTR chloride currents activated by cAMP were detected in 59% (29/49) of wild-type cells and in 50% (20/40) of heterozygous cells. However, we could not detect any CFTR currents in the homozygous cf/cf cells (0/25). The maximum CFTR current density measured at a membrane potential of 60 mV was 23.5 +/- 2.8 pA/pF (n = 29) in wild-type cells, and about half that value, i.e. 12.4 +/- 1.6 pA/pF (n = 20) in heterozygotes (P = 0.004). Calcium-activated chloride currents were detected in 73% (24/33) of wild-type, 75% (21/28) of heterozygous and in 58% (7/12) of homozygous cf/cf cells. There was no significant difference between the steady-state calcium-activated current densities in the three genotypic groups; the current measured at 60 mV being 527 +/- 162 pA/pF (n = 24) from wild-type, 316 +/- 35 pA/pF (n = 21) from heterozygote and 419 +/- 83 pA/pF (n = 7) from homozygous cells. Our data suggest that lack of CFTR does not enhance the calcium-activated chloride conductance in murine pancreatic duct cells.

To increase our understanding of rat embryos in culture and to attempt the isolation of blastocyst-derived cell lines, we examined the initial growth behaviour of rat blastocysts from four strains of rat on four different feeder cell layers. The feeders used were a continuous cell line of murine embryonic fibroblasts (STO), primary mouse (MEF) or primary rat (REF) embryonic fibroblasts, and a continuous cell line of rat uterine epithelial cells (RUCs). A medium that gave optimum plating efficiencies for murine ES cells was used in the rat embryo culture. Each culture system allowed hatching and attachment of the blastocysts, that is, the behaviour was similar on each feeder and each strain for the first 2 days in culture. Subsequently, there was a rapid differentiation of the Inner Cell Mass (ICM) cells on fibroblastic feeder cell layers (STO > MEF > REF), and this was generally complete after 3-6 days in primary culture. On RUCs, the ICM was found to increase in size without differentiation up to and including day 4 and in some cases longer. Embryo-derived cells were obtained by disaggregating and passaging ICMs on REF and RUC feeders. Rounded, refractile, and epithelial-like cells were isolated on REF and colonies of ES-like cells on the RUCs. The ES-like cells were positive for expression of alkaline phosphatase and stage-specific embryonic-antigen 1. This is an important first step towards the derivation and culture of pluripotent ES cells from the rat.

During evolution, up to 10% of the mammalian genome may have arisen by rare retroposition events. This process involves reverse transcription of RNA intermediates that originate from retroviral and retroviral-like sequences, highly and middle repetitive DNA elements, and processed pseudogenes. The mechanism, and contemporary nature, for retrotransposition of the viral family and long interspersed elements has been well studied; however, it has proven difficult to demonstrate that the process by which pseudogenes retropose is continuing. In this report a mutation in the murine hypoxanthine-guanosine phosphoribosyl transferase (hprt) gene, which was previously isolated following retroviral infection of ES cells, is shown to result from a de novo retroposition of an alpha-tubulin pseudogene. Repair of this insertion by homologous recombination restores the activity of the hprt locus, thus confirming the site of mutation. This retroposon bears all the hallmarks of a naturally processed pseudogene [intron loss, presence of a poly(A) tail, and target site duplication] while the retroposition event took place at a known time in well-defined conditions, during retroviral infection of ES cells. The study of this mutation demonstrates that under appropriate conditions pseudogenes of protein-coding genes can still retropose in the mammalian genome. The coincidence of this mutagenic event with retroviral infection suggests that in this situation the reverse transcriptase may have had a retroviral origin, which would implicate a retroviral role in facilitating pseudogene formation.

1. The delta F508 mutation of the cystic fibrosis (CF) gene is of high frequency in man (1 in 25) and in homozygotes causes cystic fibrosis. It is suggested that cystic fibrosis heterozygotes withstand secretory diarrhoea better than normal individuals and so are genetically advantaged. This hypothesis has been examined by measuring electrogenic chloride secretion in gut epithelia of normal and heterozygous CF mice. 2. Chloride secretory responses of normal and heterozygous colonic epithelia to forskolin, vasoactive intestinal polypeptide (VIP), isoprenaline, cholera toxin, heat-stable enterotoxin (STa), guanylin, carbachol and lysylbradykinin were examined. No significant differences in responses of tissues of the two genotypes were found. 3. Responses of normal and heterozygous ileal epithelia to forskolin and glucose were investigated. Heterozygous tissues responded as well as normal tissues. 4. Frusemide (furosemide) caused virtually identical inhibition of the chloride secretory responses to forskolin in colonic epithelia of both genotypes. 5. No evidence to support the genetic advantage hypothesis in ileal or colonic epithelia of the null CF mouse has been found, at least for acute responses. If the hypothesis is true then either (a) other non-cystic fibrosis transmembrane conductance regulator (non-CFTR) transport processes are involved, (b) prolonged exposure to secretagogues is required, or (c) delta F508 CFTR is responsible for the protective effect.

Cystic fibrosis is a common severe autosomal recessive genetic disease which is caused by dysfunction of an epithelial cell surface cAMP activated Cl- channel. The effects of this dysfunction are pleoitropic but the human morbidity results from the effects in the respiratory epithelium. Gene therapy is an attractive possible treatment, the gene required is well characterised and only low-level expression is required. The cellular target is accessible and the clinical effects of treatment should be readily assayable. This chapter reviews current proposals for suitable gene delivery mechanisms and vectors and discusses the clinical trials, the results from the first of which are now becoming available.

Electrogenic ion transport in the isolated colonic epithelium from normal and transgenic mice with cystic fibrosis (CF mice) has been investigated under short-circuit current (Isc) conditions. Normal tissues showed chloride secretion in response to carbachol or forskolin, which was sensitive to the Na-K-2Cl cotransport inhibitor, frusemide. Responses to both agents were maintained for at least 12 h in vitro, but the responses to carbachol changed in format throughout this period. By contrast CF colons failed to show the normal secretory responses to carbachol and forskolin, most preparations showing a decrease in Isc that was immediately reversed by frusemide. In CF colons addition of Ba2+ ions or tetraethylammonium (TEA+) to the apical bathing solution antagonised the reduction in Isc caused by the secretagogues. It is concluded that the reduction in Isc in CF colons is due to electrogenic K+ secretion and this was confirmed by flux studies using rubidium-86. In normal colons exposed to TEA+ the responses to forskolin were greater, but not significantly so, presumably because the minor K(+)-secretory responses are dominated by major chloride-secretory responses. Again rubidium-86 fluxes showed an increase of K+ secretion in normal colons receiving forskolin. Since the amiloride-sensitive current was not different in CF and normal colons there was no evidence that the CF mice were stressed in a way that increased mineralocorticoid levels and hence K+ secretion. Knowledge of the phenotype of the colonic epithelium of the CF mouse sets the baseline from which attempts at gene therapy for the gut must be judged.

The effect of lysylbradykinin on electrogenic chloride secretion in the epithelium of the mouse colon has been investigated. The peptide was active only when applied to the basolateral surface and its effects were inhibited by the B2 receptor antagonist, Hoe 140, also applied to the same surface. The chloride channel blocker, niflumic acid, also inhibited the response to kinin when added apically. Cyclo-oxygenase inhibition with piroxicam attenuated the responses to kinin, indicating involvement of prostaglandins in the responses. It is concluded that lysylbradykinin increases chloride secretion by acting via B2 receptors and, as with other tissues, brings about secretion through the agency of multiple messengers. In colonic epithelia from cystic fibrosis (CF) mice lysylbradykinin was without effect, suggesting that the final effector process involves apically located cystic fibrosis transmembrane conductance regulator (CFTR) chloride channels.

The LIM domain protein rbtn2 is associated with T cell acute leukemias. We demonstrate that rbtn2 is a nuclear protein expressed in the erythroid lineage in vivo, and using homologous recombination, we show that it is essential for erythroid development in mice. The homozygous rbtn2 null mutation leads to failure of yolk sac erythropoiesis and embryonic lethality around E10.5. Moreover, in vitro differentiation of yolk sac tissue from homozygous mutant mice and sequentially targeted double-mutant ES cells demonstrates a block to erythroid development. This shows a pivotal role for a LIM domain protein in lineage specification during mammalian development and suggests that RBTN2 and GATA-1 are critical at similar stages of erythroid differentiation.

The c-mos proto-oncogene encodes a 37-39K cytoplasmic serine/threonine kinase implicated in the meiotic maturation events during murine spermatogenesis and oogenesis. In Xenopus, ectopic expression of pp39mos can promote both the meiotic maturation of oocytes and also arrest the cleavage of blastomeres. To elucidate the role of pp39mos we have generated homozygous mutant mice by gene targeting in embryonic stem cells. These mice are viable and mutant males are fertile, demonstrating that pp39mos is not essential for spermatogenesis. In contrast, mutant females, have a reduced fertility because of the failure of mature eggs to arrest during meiosis. c-mos-/- oocytes undergo germinal vesicle breakdown and extrusion of both polar bodies followed in some cases by progression into cleavage. Mutant females also develop ovarian cysts. These results demonstrate that a major role for pp39mos is to prevent the spontaneous parthenogenetic activation of unfertilized eggs.

A variety of cystic fibrosis gene therapy approaches based on viral (adenovirus, retrovirus, and adeno-associated virus) and non-viral (liposomes and receptor-mediated endocytosis) routes are currently being assessed for safety and efficacy. Of these, the trials involving liposomal and adenoviral vectors are the most advanced, as both have been shown to correct the cystic fibrosis Cl- conductance defect in vivo.

1. Guanylin, a 15 amino acid endogenous gut peptide, increased the short circuit current (SCC) in the epithelium of the mouse colon, but only when applied to the apical and not the basolateral surface. 2. By use of selective blockers of epithelial ion transport and modification of the bathing solution, it was concluded that guanylin increased electrogenic chloride secretion but also had a minor effect on electrogenic sodium absorption. In addition there were small residual currents which remained unresolved. 3. The threshold concentration of guanylin causing a SCC increase was less than 50 nM, but at concentrations 40 times greater no indication of a maximally effective concentration was found. 4. Two guanylin isomers with the same amino acid sequence but with the disulphide bridges joined in an alternate fashion showed no activity. Thus only guanylin with the greatest structural homology to heat stable enterotoxin (STa) showed biological activity. 5. The action of guanylin was virtually eliminated in colonic epithelia from transgenic cystic fibrosis (CF) mice. As these animals lack the chloride channel coded by the CF gene sequence, it is likely that the final effector process in murine colonic epithelia involves the CFTR (cystic fibrosis transmembrane conductance regulator) chloride channel. 6. Opportunistic infections of the gut generating STa lead to diarrhoeal conditions via an action of the toxin on apical guanylin receptors. Thus, as discussed, the CF heterozygote may have a genetic advantage in this circumstance.

The cellular volume of crypts isolated from 2- to 3-week-old mouse small intestine has been measured to assess the capacity of the epithelial cells to respond to secretagogues. Vasoactive intestinal polypeptide (VIP) or carbachol, respectively cAMP- and calcium-mediated secretagogues, produced a reduction crypt volume attributed to KCl loss through channels activated by the agonists. Consistent with the participation of separate chloride channels, 4,4'-diisothiocyanatostilbene-2,2'-disulphonic acid (DIDS) blocked the carbachol- but not the VIP-induced volume decrease, whilst glibenclamide abolished the VIP effect without affecting the carbachol-induced volume decrease. Animals homozygous for a disrupted cftr gene, introduced by gene targeting, were also used as the source for crypt isolation. In these CFTR (-/-) crypts. VIP failed to elicit any reduction in cellular volume, while the response to carbachol was indistinguishable from that seen in crypts from age-matched control animals. These results are consistent with murine CFTR being a cAMP-activated chloride channel inhibited by glibenclamide and resistant to DIDS. A separate chloride conductance activated by calcium mobilization in small-intestinal crypts appears to be independent of CFTR.

We have used gene targeting in embryonic stem cells to introduce an HPRT mini-gene into the coding sequence of the murine cystic fibrosis gene (cftr). This insertion introduces a termination codon in frame with the cftr coding sequence to terminate prematurely the CFTR protein within the first nucleotide binding domain. Animals homozygous for the cftr disruption fail to thrive and display a range of symptoms including meconium ileus, distal intestinal obstructions, gastrointestinal mucus accumulation and blockage of pancreatic ducts. The animals also show lacrimal gland pathology. Tracheal and caecal transepithelial current measurements demonstrate the lack of a cAMP activatable Cl- channel. These animals will prove useful for the evaluation of new therapeutic drugs and gene therapy strategies.

Cystic fibrosis (CF) is a lethal inherited disorder affecting about 1 in 2,000 Caucasians. The major cause of morbidity is permanent lung damage resulting from ion transport abnormalities in airway epithelia that lead to mucus accumulation and bacterial colonization. CF is caused by mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene that encodes a cyclic-AMP-regulated chloride channel. Cyclic-AMP-regulated chloride conductances are altered in airway epithelia from CF patients, suggesting that the functional expression of CFTR in the airways of CF patients may be a strategy for treatment. Transgenic mice with a disrupted cftr gene are appropriate for testing gene therapy protocols. Here we report the use of liposomes to deliver a CFTR expression plasmid to epithelia of the airway and to alveoli deep in the lung, leading to the correction of the ion conductance defects found in the trachea of transgenic (cf/cf) mice. These studies illustrate the feasibility of gene therapy for the pulmonary aspects of CF in humans.

We have successfully disrupted the cftr (cystic fibrosis transmembrane conductance regulator) gene at its endogenous locus in embryonic stem cells by gene targeting. We are using a double replacement strategy to introduce subtle mutations into exon 10. We report here the first step of creating a null mutation by insertion of a functional hprt (hypoxanthine phosphoribosyl transferase) mini-gene into exon 10 of the cftr gene. Targeted embryonic stem cell clones were identified by PCR screening and confirmed by Southern blot analysis. One of the cftr targeted clones has been injected into recipient blastocysts and shown to contribute to chimaeras. The targeted clones will now be used as the starting point for a second gene targeting step to remove the hprt gene in exon 10 with the concomitant introduction of the delta F508 mutation or other mutations.

We have examined the possibility that the 5'-UT region of the human cardiac actin (CH-actin) mRNA is responsible for regulating translation of this transcript during skeletal myogenesis. Genes were constructed which consisted of the murine leukaemia virus promoter driving the Escherichia coli LacZ coding region with and without the CH-actin 5'-UT region. These constructs were transfected into L6 myoblasts that were subsequently differentiated into myotubes. The presence of the CH-actin 5'-UT region appeared to have no effect on expression of the LacZ reporter gene. EGTA blocks myogenesis and inhibits translation of muscle-specific transcripts (Endo, T. and Nadal-Ginard, B. (1987) Cell 49, 515-526) but EGTA had no effect on expression of the chimaeric LacZ transcripts. Thus, if the CH-actin transcript is subject to translational regulation, it must be mediated by sequences other than those of the 5'-UT region.

Many viral oncogenes encode protein-tyrosine kinase activities. However, important in vivo substrates of these enzymes have yet to be identified. Recently, type I topoisomerases were shown to be in vitro substrates for two tyrosine kinases. Following tyrosine phosphorylation, topoisomerase I activity was reduced 10-fold (Tse-Dinh et al. Nature 312: 785-786, 1984). To determine whether topoisomerase I activity was modulated by tyrosine phosphorylation in vivo, we have measured topoisomerase I activity in nuclear lysates prepared from both normal fibroblasts and cells transformed by two different viral oncogenes (v-abl, v-src). Under a variety of experimental conditions, we have found no evidence to support the notion that type I topoisomerase activity is modulated by tyrosine phosphorylation in vivo.

Tyrosine residues of middle-T and tyrosine phosphorylation are thought to be important in the transformation of cultured rodent cells by polyomavirus. Of the potential tyrosine sites in the carboxyl-terminal half of middle-T, tyrosines 297, 315, and 322 have been studied previously, whereas tyrosine 250 has not. Two mutant plasmids, XD121 and pT250, encode polyomavirus middle-T species in which the tyrosine 250 residue is affected. XD121 is a deletion mutant in which the region encoding tyrosine 250, together with three adjacent amino acids, is deleted, whereas pT250 is a point mutant in which the tyrosine 250 codon has been converted to a phenylalanine codon. The plasmids were handicapped in transforming ability, as judged by focus formation on a monolayer of Rat-1 cells. Both demonstrated a reduction in the number of foci produced and a lag in the time of appearance of foci when compared with wild-type plasmid. The importance of residue 250 in this phenotype was indicated by the observation that plasmids containing multiple mutations proximal to the tyrosine 250 codon were wild type in their transforming ability. Furthermore, a revertant of pT250 (pT250-w.t.), which utilized the alternative tyrosine codon of TAC, was shown to regain full transforming activity. A combined-mutant plasmid, pTH, encodes a middle-T species in which both tyrosines 250 and 315 are converted to phenylalanine. This plasmid was totally defective in the transformation of rodent cells in a focus formation assay; however, it did impart a small measure of anchorage-independent growth when the encoded protein was expressed in NIH 3T3 cells. The in vitro kinase activity and pp60c-src association of the mutant middle-T antigens were examined. These assays demonstrated a reduction in phosphate acceptor activity for the middle-T species encoded by pT250 and pTH. Quantitative kinase assays showed that all of the tyrosine-mutant middle-T species, encoded by pAS131 (containing the tyrosine 315 codon-to-phenylalanine codon mutation), pT250, and pTH, were able to enhance pp60c-src kinase activity but only at levels which were intermediate and which reflected their transforming abilities relative to wild type.

Site-directed mutagenesis was used to change Lys-128 of the simian virus 40 large-T nuclear location signal to Met, Ile, Arg, Gln, Asn, Leu, or His. Except for the large-T antigen of the Arg mutation, which was present in cytoplasmic and nuclear compartments, the resultant proteins were unable to enter the nucleus. By contrast, mutations at other sites within the signal were generally less severe in their effect. In some cases (Lys-128 to Gln, Asn, and His), the apparently cytoplasmic variants were able to support limited plasmid DNA replication, suggesting that low levels of large-T antigen undetectable by immunofluorescence were present in the nucleus. Such mutants did not support viral DNA replication. We conclude that there is a strong requirement for a basic residue at position 128 in the large-T nuclear location signal, with Lys the preferred residue.

Protein-tyrosine kinases, either in the form of growth-factor receptors or as the polypeptide products of oncogenes, appear to be important in the regulation of cell growth and transformation. A major question, however, is how their substrates mediate changes in gene expression (reviewed in refs 6-8). Because binding of proteins to specific DNA sequences represents the most direct mechanism for regulating transcription, we have investigated the possibility that some DNA-binding proteins may be substrates of protein-tyrosine kinases. Here, we present evidence for nuclear phosphotyrosyl-proteins in murine fibroblasts transformed by the v-abl protein-tyrosine kinase. Furthermore, we have found that these proteins are not significantly phosphorylated in normal NIH 3T3 cells. Finally, using affinity competition chromatography with bacterial and mouse DNA, we have demonstrated that some of these proteins preferentially bind to mouse DNA. The identification of phosphotyrosyl-proteins with selective DNA-binding properties suggests a possible mechanism through which protein-tyrosine kinases may effect changes in gene transcription.

Aberrations in nuclear proto-oncogene organisation and/or gene expression have been implicated in cell transformation mediated by the v-abl gene. For example, it has been suggested that amplification of the c-myc proto-oncogene is a co-operative event in v-abl induced fibroblast transformation. We have investigated amplification of the c-myc, p53 and c-fos nuclear proto-oncogenes in several Abelson murine leukaemia virus (A-MuLV) transformed fibroblast lines. None of these proto-oncogenes were detectably rearranged or amplified in v-abl transformed Swiss 3T3 lines. In contrast, NIH3T3 fibroblasts transformed by the v-abl gene consistently showed a 4 to 16-fold amplification of the c-myc gene. These data show that c-myc gene amplification is not an obligatory event associated with A-MuLV transformation, but may be restricted to cell lines derived from NIH3T3. c-myc gene amplification also did not correlate with a reduced latency period for tumour induction in nude mice. In addition, c-myc amplification was not selected during tumourigenesis, indicating that this event is not required for A-MuLV transformed Swiss 3T3 cells to display a full tumourigenic phenotype.

A short sequence of predominantly basic amino acids Pro-Pro-Lys-Lys-Lys-Arg-Lys-Val from SV40 Large T is responsible for the normal nuclear location of the protein. Alteration of Lys-128 to each of six different residues other than Arg renders Large T cytoplasmic, whereas single amino acid changes in the surrounding region impair but do not prevent nuclear accumulation. When transposed to the amino terminus of cytoplasmic Large T species, or Escherichia coli beta-galactosidase or of chicken muscle pyruvate kinase, the sequence around Lys-128 of Large T is able to direct the recipient protein to the nucleus. This demonstrates that these amino acids can be sufficient for nuclear location and can act as a nuclear location signal. A computer search of over 2500 proteins reveals that some other nuclear proteins (for example, BK virus Large T, SV40 VP2 and adenovirus 72kDa DNA binding protein) contain very similar basic tracts, but so too do some presumed non-nuclear proteins (for example, poliovirus VP3). We suggest that the related sequence acts as the nuclear location signal in the other nuclear proteins but that the sequence does not function in all cases, perhaps because it is not accessible. A similar, but shorter or less basic sequence, was detected in a number of other nuclear proteins, for example, polyoma virus Large T, SV40 VP1 and several histones. However, such sequences were also found in many other proteins. Perhaps the shorter basic sequences can also act as nuclear location signals, but to be functional they need to be exposed (for example, at the amino terminus of the protein as in SV40 VP1) or to be present in multiple copies.