Sequence-specific interaction of a conformational domain of p53 with DNA.

Mutations within a conserved "conformational" domain of the p53 protein have frequently been observed in a wide variety of human cancers. A hybrid protein containing the wild-type conformational domain of p53 fused to protein A bound to calf thymus DNA and a specific p53 DNA-binding motif. Hybrid proteins containing mutations in p53 bound to DNA less efficiently than wild-type hybrid protein. In addition, competition experiments showed that mutated p53 DNA-binding motif failed to interact with p53 hybrid proteins. The DNA-binding activity of wild-type p53 hybrid protein was inhibited by the metal chelator 1,10-phenanthroline. These results demonstrate that DNA-binding activity resides in the conformational domain of p53, providing a structural model for disruption of DNA binding by mutation. Furthermore, metal ions may regulate binding of p53 to DNA by modulating its conformation.

Epidermal growth factor receptor protein-tyrosine kinase activity in human cell lines established from squamous carcinomas of the head and neck.

Two cell lines established from tumors of the head and neck area at different clinical stages were found to differ in the expression and in the tyrosine kinase activity of the epidermal growth factor (EGF) receptor. Cell line 183A was derived from an early-stage tumor and cell line 1483 was derived from a tumor that had metastasized to lymph nodes. The 1483 cells displayed a higher plating efficiency and clonogenicity in soft agar, suggesting a more tumorigenic phenotype over the 183A cells. Analyses of EGF receptor levels by using R1 anti-EGF receptor serum indicated that the 1483 cells expressed 5-fold more receptor than did the 183A cells. EGF receptors isolated from each cell line were active for kinase activity in an immune complex kinase assay, using monoclonal R1 anti-EGF receptor antibody. The autophosphorylation activity of both receptors was stimulated by addition of EGF to isolated membrane preparations and intact cells, although the EGF receptor of the 1483 cells was much less responsive to EGF than the receptor from 183A cells. In addition, the 1483 receptor consistently incorporated about twice as much phosphate as did the 183A receptor in an immune complex kinase assay. These data suggest that the basal tyrosine kinase activity of the EGF receptor from 1483 cells may be more active than the EGF receptor kinase from 183 cells.

Analysis of P210bcr-abl tyrosine protein kinase activity in various subtypes of Philadelphia chromosome-positive cells from chronic myelogenous leukemia patients.

An altered c-abl gene product (P210bcr-abl) possessing associated tyrosine protein kinase activity was recently been reported in several blast chronic myelogenous leukemia (CML) cell lines. We have examined different morphological types of leukocytes directly obtained from patients at the blast crisis stage of CML for expression of P210bcr-abl tyrosine protein kinase activity. Phosphorylation of P210bcr-abl in an immune complex kinase assay using an anti-v-abl peptide serum was observed in blast cells from four Philadelphia chromosome (Ph1)-positive CML patients in blast crisis. P210bcr-abl protein kinase activity was detected regardless of whether the blast cells were of myeloid, lymphoid, or undifferentiated morphology. P210bcr-abl protein kinase activity was not detected in immune complexes either from leukocytes of four Ph1-negative CML patients in blast crisis, of five acute myelogenous leukemia patients, or in the promyelocytic cell line HL-60. Mature myeloid cells are associated with an inhibitory factor for not only P210bcr-abl protein kinase activity, but also protein kinases in general. Therefore, analyses of Ph1-positive benign phase CML myeloid cells, the majority of which are well differentiated, could not be successfully performed. The inhibition of P210bcr-abl protein kinase activity is not a specific property of mature cells from CML patients since granulocytes from a normal volunteer also demonstrated a similar effect. However, extracts of Ph1-positive cultured B-lymphocytes from a patient in benign phase demonstrated active P210bcr-abl protein indicating that the P210bcr-abl protein is expressed in an enzymatically active form in the earlier phases of CML. In addition to the previously reported P210 and P190 abl-related proteins, a novel Mr 53,000 protein was found to undergo phosphorylation at serine and tyrosine in immune complex kinase assays of two blast crisis CML cell lines (K562 and EM2) and in samples from blast crisis patients in which P210bcr-abl was detected. Peptide mapping by the Cleveland technique suggested that Mr 53,000 protein is unrelated to P210bcr-abl. Immune complex kinase assays of K562 cells with an anti-src serum (GD-11) yielded active c-src kinase and a Mr 50,000 phosphorylated protein, both of which were resistant to alkaline hydrolysis. Peptide mapping suggested that Mr 53,000 protein is related to Mr 50,000 protein which is precipitated with P210bcr-abl as an Mr 300,000 protein complex.

A cellular protein activates the sequence-specific DNA-binding of p53 by interacting with the central conserved region.

Mutational inactivation of the p53 gene product is one of the most common genetic aberations so far identified in human cancers. The p53 protein suppresses the transformed phenotype by transactivation or repression of genes involved in cell growth control. Missense mutations in the p53 protein coding sequence observed in human cancers are clustered within a central conserved (conformational) domain spanning amino acid residues 100-300 of a total of 393. Using the conformational domain of p53 fused with protein A, we have shown that the p53 conformational domain possesses Zn+2-dependent, sequence-specific DNA-binding activity. In addition to binding DNA, this domain interacts with at least five cellular proteins ranging in sizes from 30K to 90K M(r) and with the SV40 large T antigen viral oncoprotein. We investigated these cellular proteins for their modulatory effects on the sequence-specific DNA binding activity of full-length wild-type p53. A mixture of p53 conformational domain-binding proteins in bulk enhanced the DNA-binding activity of p53 greater than two-fold. Selective elution of the p53-binding proteins from the p53 hybrid protein by using a sequential step-wise NaCl gradient implicated one protein of 35K M(r) as contributing to a greater than four-fold activation of p53 DNA-binding activity. A p53 conformational domain protein containing a tumor-derived mutation at amino acid 175 failed to associate with the 35K M(r) protein. We propose that proteins interacting with the conformational domain of wild type p53 regulate the DNA-binding activity of p53, thus providing a biochemical basis for the alterations in its function induced by point mutations.

Binding of cellular proteins to a conformational domain of tumor suppressor protein p53.

The genes regulated by p53, as well as the factors modulating its function, need to be identified before the mechanism of action of p53 in control of cell growth can be adequately understood. Binding of the SV40 large T-antigen protein to an evolutionally conserved (conformational) domain of p53 inhibits p53's DNA-binding and transcription activation activities. Cellular proteins might also bind to this same region of p53 to regulate its function. A hybrid protein composed of protein A fused to the conformational domain (amino acids 115-295) of p53 was expressed in Escherichia coli and used as an affinity probe for binding proteins in detergent lysates of non-small cell lung carcinoma (NSCLC) cells. The wild-type p53 hybrid protein associated with several major proteins of molecular weights 45 K, 56 K, and 70 K, as well as other minor species ranging in molecular weight from 30 K to 90 K. These proteins bound specifically to the p53 sequence of the hybrid protein. Protein A did not associate with these proteins and the two p53 hybrid proteins containing missense mutations at codons 273 and 175 exhibited a 40-80% weaker association. In addition, T antigen competed with the cellular proteins for binding to the conformational domain. The conditions of cell growth had a profound effect on the expression of the p53 binding proteins. Considerably more p53 binding proteins were expressed in actively growing cells than in cultures maintained under conditions for slow growth. Quantitative differences in expression of p53-binding proteins were observed among different NSCLC cell lines. The expression of p53-binding proteins was not restricted to NSCLC cell lines; detergent extracts of an osteosarcoma cell line yielded similar p53-binding proteins.

Altered expression of the p50 subunit of the NF-kappa B transcription factor complex in non-small cell lung carcinoma.

Recent studies have implicated a role for the p50 subunit of the NF-kappa B transcription factor complex in tumorigenesis. Therefore, we investigated the expression of the p50 subunit of the NF-kappa B transcription factor complex in paired normal and non-small cell lung carcinoma (NSCLC) tissues. Here we show that nine of 11 (81.8%) fresh NSCLC tissues expressed from two- to 20-fold higher levels of the p50 subunit than normal lung tissue. Thirteen NSCLC cell lines also exhibited high levels of p50. Alterations in the normal NF-kappa B/rel pathway of regulation may play a role in the genesis of NSCLC.

A novel NF-kappa B p65 spliced transcript lacking exons 6 and 7 in a non-small cell lung carcinoma cell line.

Transcripts of the gene encoding the p65 subunit of the NF-kappa B/Rel transcription factor complex have been reported to undergo alternative splicing to generate one derivative lacking codons for amino acids (aa) 222 to 231 (p65 delta 1) and another that lacks codons for aa 13 to 25 (p65 delta 2) of the conserved Rel homology domain [Narayaran et al., Science 256 (1992) 317-320; Lyle et al., Gene 138 (1994) 265-266]. We have identified a third splicing event in a non-small-cell lung carcinoma cell line that potentially generates a novel p65 mRNA derivative lacking codons for aa 187 to 293 (p65 delta 3) of the Rel homology domain.

Electrophysiologic studies of cutaneous nerves of the forelimb of the cat.

The cutaneous innervation of the forelimb was investigated in 20 barbiturate-anesthetized cats by using electrophysiological techniques. The cutaneous area (CA) innervated by each cutaneous nerve was delineated in at least six cats by brushing the hair in the CA with a small watercolor brush while recording from the nerve. Mapping of adjacent CA revealed larger overlap zones (OZ) than were noted in the dog. Remarkable findings were that the brachiocephalic nerve arose from the axillary nerve and the CA comparable to that supplied by the cutaneous branch of the brachiocephalic nerve in the dog was supplied by a cutaneous branch of the suprascapular nerve. The CA supplied by the communicating branch from the musculocutaneous to the median nerve was similar in both species except that the communicating branch arose proximal to any other branches of the musculocutaneous nerve in the cat, whereas it was a terminal branch in the dog. The superficial branch of the radial nerve gave off cutaneous brachial branches in the cat proximal to the lateral cutaneous antebrachial nerve. The CA of the palmar branches of the ulnar nerve did not completely overlap the CA of the palmar branches of the median nerve as occurred in the dog; thus an autonomous zone (AZ) for the CA of the palmar branches of the median nerve is present in the cat, whereas no AZ existed for the CA of this nerve in the dog.

SV40 T-antigen as a dual oncogene: structure and function of the plasma membrane-associated population.

SV40 T-antigen (T-ag) is localized in both the nucleus (nT-ag) and plasma membrane (pmT-ag) of cells and provides multiple functions necessary for cell transformation. The pmT-ag population is structurally very similar to the nT-ag. Transport to the cell surface is by an unknown mechanism that does not involve the secretory pathway. The disposition of T-ag in the membrane exposes both the amino and the carboxyl terminus on the exterior of the cell. Nuclear-transport-defective mutants of T-ag can transform established cells in culture, but not primary cells, suggesting that non-nuclear forms of T-ag may mediate some transformation-related process(es). A non-cytolytic protein extraction technique utilizing 1-butanol solubilized from SV40-transformed cells a multimeric complex composed of pmT-ag and at least five cellular proteins ranging in size from 35,000 (35K) to 60K M. Both amino- and carboxylterminal T-ag-specific monoclonal antibodies co-precipitated T-ag and the 35-60K Mr proteins, but antibodies against the internal portion of T-ag precipitated only uncomplexed T-ag. The growth state of the cells markedly influenced the expression of the T-ag-containing surface complexes; more complexes were recovered from actively dividing cells than from confluent cell cultures, and suspension cells yielded more complexes than cells on a substratum. The complex exhibited a highly dynamic association with the cell membrane, as demonstrated by pulse-chase analysis. The characteristics of growth-dependent expression and rapid turnover rate suggest a functional role for the membrane complex. The identities of the cellular proteins in the complex with pmT-ag are unknown, although one member (56K) is recognized by p53-specific monoclonal antibodies.

Effect of sleep restriction on orthostatic cardiovascular control in humans.

We hypothesized that sleep restriction (4 consecutive nights, 4 h sleep/night) attenuates orthostatic tolerance. The effect of sleep restriction on cardiovascular responses to simulated orthostasis, arterial baroreflex gain, and heart rate variability was evaluated in 10 healthy volunteers. Arterial baroreflex gain was determined from heart rate responses to nitroprusside-phenylephrine injections, and orthostatic tolerance was tested via lower body negative pressure (LBNP). A Finapres device measured finger arterial pressure. No difference in baroreflex function, heart rate variability, or LBNP tolerance was observed with sleep restriction (P > 0.3). Systolic pressure was greater at -60 mmHg LBNP after sleep restriction than before sleep restriction (110 +/- 6 and 124 +/- 3 mmHg before and after sleep restriction, respectively, P = 0.038), whereas heart rate decreased (108 +/- 8 and 99 +/- 8 beats/min before and after sleep restriction, respectively, P = 0.028). These data demonstrate that sleep restriction produces subtle changes in cardiovascular responses to simulated orthostasis, but these changes do not compromise orthostatic tolerance.

Mental health services for medical students: perceptions of students, student affairs deans, and mental health providers.

PURPOSE: To examine (1) whether there is any consistency among medical schools in mental health services provided and (2) how these services are perceived by student affairs deans, mental health service providers, and the students themselves. METHOD: Questionnaires were sent in October 1991 to the student affairs dean (or director), the individual responsible for student mental health services, and a student representative in each of the 126 U.S. and Canadian medical schools. Data were sought regarding personnel, individuals served, location, hours, administration, funding, confidentiality, administrative referrals, and respondents' suggestions for improvement. Possible differences among the three groups of respondents were tested by chi-square. RESULTS: Responses were received from 75 student affairs deans, 53 mental health providers, and 30 students. There was much diversity among schools in services provided, especially in the areas of administration and funding. Although perceptions of the three respondent groups were often the same, they differed significantly in a number of areas. Suggestions for improvement of services involved funding, personnel, hours, confidentiality and privacy, specialty services, preventive and support programs, and visibility. The suggestion most frequently made by the students was for increased information and visibility. CONCLUSION: The differences among schools coupled with the differing perceptions within schools indicate a need for a comprehensive consideration of what kinds of mental health services are needed and how they can best be made accessible to a diverse body of students experiencing a variety of academic and personal challenges.

The in vitro penetration and distribution of T-2 toxin through human skin.

The penetration and distribution of T-2 toxin in excised human abdominal skin has been determined for a dose range of 1.0-2.6 micrograms/cm2 skin using an ethanol vehicle and a saline receptor solution. In all cases the overall percentage penetration of T-2 after 48 h was low, the greatest amounts of toxin being present in the stratum corneum with less in the epidermis and relatively little in the dermis. Vehicle: skin partition coefficients support this finding. Neither penetration nor distribution were changed by a rabbit serum receptor solution. Electron micrographs showed that at 1.8 micrograms/cm2 and above the contents of the intercellular space are leached out to leave the integument as a porous membrane. The distribution of T-2 within the skin after 48 h would suggest that for doses up to 2.6 micrograms/cm2 the irritative and inflammatory effects on the skin would be of more immediate concern than would systemic toxicity.

Skin effects of trichothecenes and their amelioration by decontamination.

The ability of trichothecenes, in particular T2 toxin (T2), to cause irritant effects on the skin has been investigated in laboratory rodents and rabbits. Quantitatively, T2 was found to be highly potent in this respect, causing irritant reactions on rat skin at contamination densities of less than 1 microgram.cm-2. The first appearance of skin effects was delayed for approximately 6 h after application, irrespective of the dose applied. Similarly, variation in solvent or injection into or beneath the skin failed to accelerate the onset of the irritant reaction. An aqueous soap solution was largely effective in removing low doses of T2 from the skin, but was ineffective in removing larger doses. However, washing the contaminated skin with polyethylene glycol 300 was very effective at removing even large doses of T2 from the skin. The macrocyclic trichothecene verrucarin A was of comparable irritancy to T2 on rat skin. Diacetoxyscirpenol and nivalenol were less potent.

Selective compartmentalization of different mdm2 proteins within the nucleus.

Overexpression of the mdm2 protooncogene protein, which can lead to the inactivation of normal p53, has been observed in some human cancers. The mdm2 gene is positively regulated by p53, providing for a feedback loop in the control of both p53 and mdm2 activity. The expression of the mdm2 and p53 proteins in different non-small cell lung carcinoma (NSCLC) cell types harboring wild-type or mutant p53, or lacking p53 altogether, were investigated to determine whether a correlation exists between the expression of these two proteins. The mdm2 protein was expressed at very low levels in all NSCLC lines examined, regardless of the p53 status. To determine whether mdm2 could be induced by p53 in NSCLC, NSCLC cells were transfected with a recombinant adenovirus expressing high levels of wild-type p53. The highest levels of exogenous wild-type p53 were observed in p53-null H358 and H1299 cells and in H226b cells expressing endogenous wild-type p53 were observed in p53-null H358 and H1299 cells and in H226b cells expressing endogenous wild-type p53. In these cells, wild-type p53 induced the expression of 90/92K M(r) mdm2 proteins, as well as several faster-migrating mdm2-related species exhibiting relative mobilities of 76/78K, 57/59K, 46K, 28K, and 12K. Northern analyses of H358 and H1299 cells transfected with wild-type p53 showed that these cells expressed three species of mdm2 mRNA of 5.5, 4.6-3.8, and 2.1 Kb in size. Subcellular fractionation revealed that the 90/92K M(r) mdm2 protein species was localized to both the crude plasma membrane/cytoplasmic and nuclear fractions, and that the smaller mdm2 proteins associated selectively with different nuclear substructures. The 76/78K, 57/59K, and 46K Mr(r) mdm2 proteins may be derived by differential splicing of the 5.5 Kb mRNA, and their differential compartmentalization within the nucleus suggests that each has a distinct function, potentially in the regulation of p53 and other gene products.

Two-dimensional gel analysis of p53-mediated changes in protein expression.

The p53 gene product suppresses both tumor cell growth and the cellular transformation process promoted by oncogenes. Although several genes are known to be positively regulated by p53, it is unclear how many gene expression events are involved in p53-mediated growth arrest and apoptosis. Two-dimensional gel electrophoresis was employed to investigate changes in whole-cell protein expression in p53-null H1299 cells that were transfected with a recombinant adenovirus expressing wild-type p53 at high efficiency. The two-dimensional gel analysis was restricted to proteins ranging in mass from 12,000 to 80,000 daltons. A total of 17 proteins were induced and one was repressed within 16 h of expression of exogenous wild-type p53 in H1299 cells. These results indicate that p53-mediated growth suppression involves a complex array of gene expression events.

Retinoblastoma protein in non-small cell lung carcinoma, cells arrested for growth by retinoic acid.

The growth of non-small cell lung carcinoma (NSCLC) cells was inhibited by retinoic acid after 16 h of treatment. However, the growth of all cell lines except one became refractory to retinoic acid after 48 h of exposure. The expression of the hyperphosphorylated retinoblastoma protein species (RB p110) was observed to be increased fivefold to tenfold in NSCLC cells within 16 h of exposure to retinoic acid. In the H460a and H226b cell lines, p110 showed some conversion to the underphosphorylated p105 form after 24 h of retinoic acid treatment. After 48 h, p105 became the predominant form, along with a 60K M(r) species. After 72 h, expression of all RB protein species became almost undetectable in H460a and H226b cells, concomitant with increases in cell growth rates. A different pattern of RB expression was observed in the H322j and H358 cell lines. In these cells, both p110 and p105 were induced within 8 h of retinoic acid treatment. Furthermore, the elevated levels and the phosphorylation state of RB in retinoic acid-treated H322j and H358 cells remained essentially unchanged for up to 72 h and only low amounts of the 60K M(r) species were detected. I believe that a post-translational mechanism is responsible for the retinoic acid-mediated induction of RB, since levels of RB mRNA remained relatively unchanged during the time course of retinoic acid exposure. I conclude that retinoic acid inhibited the growth of NSCLC cells by inducing high levels of RB and that increases in RB levels occurred as a result of either an increase in stability of the protein or by downregulation of an RB-specific protease. The inhibition of growth by retinoic acid must involve other molecular events, since the H596b cell line, which lacks RB protein, exhibited growth properties in the presence of retinoic acid similar to those of cells in which RB expression was induced.

Cyclin A1 is a p53-induced gene that mediates apoptosis, G2/M arrest, and mitotic catastrophe in renal, ovarian, and lung carcinoma cells.

We were the first to identify cyclin A1 as a p53-induced gene by cDNA expression profiling of p53-sensitive and -resistant tumor cells [Maxwell S. A. and Davis G. E. (2000) Proc. Natl. Acad. Sci. USA 97, 13009-13014]. We show here that cyclin A1 can induce G2 cell cycle arrest, polyploidy, apoptosis, and mitotic catastrophe in H1299 non-small cell lung, TOV-21G ovarian, or 786-0 renal carcinoma cells. More cdk1 protein and kinase activities were observed in cyclin A1-induced cells than in GFP control-induced cells. Thus, cyclin A1 might mediate apoptosis and mitotic catastrophe through an unscheduled or inappropriate activation of cdk1. Two primary renal cell carcinomas expressing mutated p53 exhibited reduced or absent expression of cyclin A1 relative to the corresponding normal tissue. Moreover, renal carcinoma-derived mutant p53s were deficient in inducing cyclin A1 expression in p53-null cells. Cyclin A1 but not cyclin A2 was upregulated in etoposide-treated tumor cells undergoing p53-dependent apoptosis and mitotic catastrophe. Forced upregulation of cyclin A2 did not induce apoptosis. The data implicate cyclin A1 as a downstream player in p53-dependent apoptosis and G2 arrest.

Posttranslational regulation of p53 tumor suppressor protein function.

Alteration of the p53 gene by deletion and mutation is the most common denominator yet identified among human cancers (Hollstein et al., 1991; Caron de Fromental and Soussi, 1992). The involvement of the p53 gene in such a broad scope of human cancers warrants further investigation into its mechanism of action in regulating cell growth. The wild-type p53 protein restricts cell growth in the G1 phase of the cell cycle by regulating the transcription of genes and possibly, by influencing DNA replication. Elucidating the cell growth restriction of p53 will require identification and characterization of the genes whose expression is regulated by p53 and the proteins that interact with p53 to regulate its DNA-binding and transactivation functions. A model for the regulation of p53 biochemical function is proposed that extends further and builds on the conformational hypothesis for regulation of p53 function hypothesized by Milner (1991) and Ullrich et al. (1992a). The conformational hypothesis of regulation of p53 function states that the conformation of p53 determines whether it expresses growth-suppressing or growth-promoting biological activity. Mutations observed in human cancer lock p53 in a growth-promoting conformation. We expand the conformational hypothesis in a regulatory model that includes binding proteins, kinases/phosphatases, redox modifier proteins, and homo/hetero-oligomerization, which modulate the tertiary structure of the protein. Different conformational modes of p53 interact differently with initiation complexes at gene promoters and at origins of DNA replication. Each form of p53, depending on its interaction with proteins and gene transcription-initiation complexes, will mediate distinct biological effects on cells ranging from growth suppression to growth promotion. Furthermore, depending on its conformational state, p53 can repress or activate other transcription factors thus indirectly affecting gene regulation. We propose that each cell and tissue type expresses unique quantities and types of p53-binding proteins and modifying enzymes that regulate the interaction of p53 with promoters of genes necessary for control of growth of a specific cell or tissue. It is anticipated that defects in the expression of p53 regulatory proteins are involved in a portion of those tumors expressing normal p53. Defects in the p53 biochemical pathway may thus be even more prevalent in human cancers than is now realized.

Transient stabilization of p53 in non-small cell lung carcinoma cultures arrested for growth by retinoic acid.

Proliferation of five non-small cell lung carcinoma (NSCLC) cultures was inhibited after 16 h exposure to retinoic acid. We investigated whether expression of the p53 protein correlated with the growth pattern of NSCLC lines observed in the presence of retinoic acid Levels of wild-type p53 protein underwent fivefold increases in lines H460a and H226b after 16 to 48 h treatment with 5 microM retinoic acid but then decreased to undetectable amounts in these cell lines after 72 h retinoic acid treatment. Levels of p53 transcripts remained unchanged during the time of increases in protein expression in retinoic acid-treated H460a cells, suggesting that a post-translational mechanism was involved in the increased expression of the protein. Pulse-chase analysis demonstrated that wild-type p53 was significantly more stabile in H460a cells treated with retinoic acid, exhibiting a half-life greater than 6 h, in contrast to 3 h for the protein in untreated control cells. The retinoic acid-mediated effect was specific for wild-type p53, since expression of mutant p53 in the H596b and H322j cell lines remained relatively unchanged even after 72 h exposure to retinoic acid. We conclude that retinoic acid induces stabilization of wild-type p53 in NSCLC cells by a post-translational mechanism. Furthermore, increases in expression of p53 were not responsible for the retinoic acid-induced transient inhibition of growth of NSCLC cells, since the growth of H358 p53-null cells also was inhibited by retinoic acid.