Search

Integrating the health-care enterprise (IHE) is an international initiative to promote the use of standards to achieve interoperability among health information technology systems. The Pathology and Laboratory Medicine domain within IHE has brought together subject matter experts, electronic health record vendors, and digital imaging vendors, to initiate development of a series of digital pathology interoperability guidelines, called "integration profiles" within IHE. This effort begins with documentation of common use cases, followed by identification of available data and technology standards best utilized to achieve those use cases. An integration profile that describes the information flow and technology interactions is then published for trial use. Real world testing occurs in "connectathon" events, in which multiple vendors attempt to connect their products following the interoperability guidance parameters set forth in the profile. This paper describes the overarching set of integration profiles, one of which has been published, to support key digital pathology use cases.

Background: The use of digital whole slide imaging for human epidermal growth factor receptor 2 (HER2) immunohistochemistry (IHC) could create improvements in workflow and performance, allowing for central archiving of specimens, distributed and remote interpretation, and the potential for additional computerized automation. Procedures: The accuracy, precision, and reproducibility of manual digital interpretation for HER2 IHC were determined by comparison to manual glass slide interpretation. Inter- and intra-pathologist reproducibility and precision between the glass slide and digital interpretations of HER2 IHC were determined in 5 studies using DAKO HercepTest-stained breast cancer slides with the Philips Digital Pathology System. In 2 inter-method studies, 3 pathologists interpreted glass and digital slides in sequence or in random order with a minimum of 7 days as a washout period. These studies also measured inter-observer reproducibility and precision. Another two studies measured intra-pathologist reproducibility on cases read 10 times by glass and digital methods. One additional study evaluated the effects of adding IHC control slides with each run, using 1 pathologist interpreting glass and digital slides randomized from the sets above along with appropriate controls for each slide in the set. Results: The overall results show that there is no statistical difference between the variance of performance when comparing glass and digital HER2 interpretations; and there were no effects noted when control tissues were evaluated in conjunction with the test slides. Conclusions: The results show that there is an equivalence of result when interpreting HER2 IHC slides in breast cancer by either glass slides or digital images. Digital interpretation can therefore be safely and effectively used for this purpose.

The categorization of intraductal proliferative lesions of the breast based on routine light microscopic examination of histopathologic sections is in many cases challenging, even for experienced pathologists. The development of computational tools to aid pathologists in the characterization of these lesions would have great diagnostic and clinical value. As a first step to address this issue, we evaluated the ability of computational image analysis to accurately classify DCIS and UDH and to stratify nuclear grade within DCIS. Using 116 breast biopsies diagnosed as DCIS or UDH from the Massachusetts General Hospital (MGH), we developed a computational method to extract 392 features corresponding to the mean and standard deviation in nuclear size and shape, intensity, and texture across 8 color channels. We used L1-regularized logistic regression to build classification models to discriminate DCIS from UDH. The top-performing model contained 22 active features and achieved an AUC of 0.95 in cross-validation on the MGH data-set. We applied this model to an external validation set of 51 breast biopsies diagnosed as DCIS or UDH from the Beth Israel Deaconess Medical Center, and the model achieved an AUC of 0.86. The top-performing model contained active features from all color-spaces and from the three classes of features (morphology, intensity, and texture), suggesting the value of each for prediction. We built models to stratify grade within DCIS and obtained strong performance for stratifying low nuclear grade vs. high nuclear grade DCIS (AUC = 0.98 in cross-validation) with only moderate performance for discriminating low nuclear grade vs. intermediate nuclear grade and intermediate nuclear grade vs. high nuclear grade DCIS (AUC = 0.83 and 0.69, respectively). These data show that computational pathology models can robustly discriminate benign from malignant intraductal proliferative lesions of the breast and may aid pathologists in the diagnosis and classification of these lesions.

ContextAssessment of accuracy and feasibility of whole slide imaging (WSI) for interinstitutional consultation in surgical pathology.ObjectivesTo train technical and pathologist staff in WSI technology, establish and evaluate a WSI workflow using training cases and second-opinion consultations, and assess diagnostic accuracy.DesignFirst, WSI training and evaluation using selected subspecialty service cases were performed and compared with the clinical glass slide (GS) diagnosis. Second, WSI and GS diagnoses of consecutive, second-opinion consultation cases were compared. Discrepancies underwent adjudication to determine a reference diagnosis. Participant observations on WSI initiation to practice were gathered.ResultsThere were 130 cases evaluated, with 123 correlations (94.6%) and 6 minor (4.6%) and 1 major (0.8%) discrepancies. The 74 consultation cases interpreted had 52 correlations (70.3%), and 18 minor (24.3%) and 4 major (5.4%) discrepancies. The WSI and GS adjusted major discrepancy rates in second-opinion consultations were 2.7% (2 of 74) and 4.1% (3 of 74), respectively. Statistical analysis showed that WSI was not inferior to GS interpretation. Pathologists agreed the software was easy to use and the images were adequate, but more time was spent rendering WSI interpretations.ConclusionsA significant learning curve was observed in the transition from the training set to clinical consultation cases associated both with WSI interpretation and adjustments to the digital analogs of routine GS workflow. Results from second-opinion consultations indicated that WSI interpretation was as accurate as GS interpretation among properly trained and experienced users. Overall, WSI-based practice appears feasible for second-opinion consultations.

Phosphorylation and subsequent activation of p70 S6 kinase (S6K) are events that are highly conserved in the cellular response to mitogens. The neuropeptide bombesin, which is a potent mitogen for Swiss 3T3 cells, stimulated a time- and dose-dependent activation of p70S6K as determined by gel mobility shift and immune complex kinase assays. This effect was inhibited by the immunosuppressant rapamycin at 10 nM, which also completely abolished bombesin-stimulated DNA synthesis at identical concentrations. In striking contrast, the combination of bombesin and insulin in synergy stimulated maximum DNA synthesis in these cells despite persistent inhibition of p70S6K by rapamycin throughout G1. These results indicate that activation of p70S6K is not required for the transition of quiescent cells to the S phase of the cell cycle. The inhibitory effects of rapamycin on bombesin-stimulated cell cycle progression did not involve accumulation of the cyclin-dependent kinase inhibitor p27kip1 but a striking inhibition of the expression of cyclin D1. This effect was circumvented by bombesin with insulin, suggesting that a rapamycin-insensitive pathway stimulated by this combination leads to cyclin D1 expression. Thus, these findings, dissociating the mitogenic effects of bombesin in synergy with insulin from activation of p70S6K, support the hypothesis that this kinase is a component of one of the parallel pathways that can lead to DNA synthesis rather than an obligatory point of convergence in mitogenic signaling.

The baculovirus expression system has been successfully used to overproduce a number of different protein products. In this report we describe the construction of a recombinant baculovirus containing the adenovirus E1A 13s cDNA sequence. Infection of insect cells with this virus results in the production of phosphorylated E1A protein. The phosphorylation pattern appears to be similar to the complex pattern associated with E1A protein synthesis in mammalian cells. Purified baculovirus generated E1A protein activated transcription of specific poIIII promoters both in microinjected Xenopus laevis oocytes and in HeLa cell in vitro transcription extracts. The protein also stimulates in vitro transcription of the poIIII transcribed VA1 gene.

The ATF-2 transcription factor can mediate adenovirus E1A-inducible transcriptional activation. Deletion analysis has indicated that the N-terminal region of ATF-2 is essential for this response. Furthermore, the N-terminus can activate transcription in the absence of E1A when fused to a heterologous DNA binding domain. However, in the intact protein this activation domain is masked. In this report we show that residues in the N-terminus required for activation are also required for mediating E1A stimulation. In particular two threonine residues at positions 69 and 71 are essential. These residues are phosphorylated in vivo and can be efficiently phosphorylated in vitro by the JNK/SAPK subgroup of the MAPK family. ATF-2 can bind to a UV-inducible kinase through a region in the N-terminus that is distinct from the sites of phosphorylation; this binding region is both necessary for phosphorylation by JNK/SAPK in vitro and for transcriptional activation in vivo. The activity of the N-terminus is stimulated by UV irradiation which stimulates the signalling pathway leading to JNK/SAPK. Finally, although ATF-2 binds to the E1A protein, the N-terminal activation domain is not required for this interaction. The results show that ATF-2, like other members of the ATF/CREB family of DNA binding proteins is regulated by specific signalling pathways.

The DNA binding specificities of CREB1 and CREB2 homodimers and the CREB2/cJun heterodimer were analyzed with a CASTing technique. All but one of the selected sequences varied from the consensus CRE (TGACGTCA) by three nucleotides or less. The profile of variations selected and the binding affinity for these sequences were unique for each CREB complex. The affinities were not effected by the palindromic nature of the sequences, but were strongly effected by flanking sequences. The strength of DNA binding in vitro correlated with the degree of transactivation observed in JEG-3 cells transfected with reporter plasmids harboring CRE variants, when hybrid CREB proteins fused to the VP16 activation domain were expressed. When native CREB proteins were expressed, the correlation was attenuated by the nature of the variant sequence. A CRE variant (TGACATCA) found in several natural promoters, exhibited the lowest basal transcription rate of the variants and a lower level of induction than expected when compared with the in vitro binding data. These results indicate that transactivation of DNA sequence elements is strongly effected by the strength of transcription factor binding, and that individual sequences can attenuate the level of induction.

The promoter motif CGTCA binds multiple cellular factors that mediate a variety of inducible events, including positive responses to raised cellular levels of cAMP and to the Adenovirus E1a protein. To date, at least ten mammalian cDNA clones have been isolated that encode distinct proteins capable of binding to this motif. However, in most cases the precise stimuli that may regulate these different factors have yet to be determined. We have previously shown that the abundant Hela protein ATF-43 forms a complex in vivo with the cyclic AMP response element binding protein (CREB). In this report we definitively show that ATF-43 is the product of the two published cDNA clones, ATF1 and TREB 36. We confirm that ATF1 efficiently heterodimerises with CREB and demonstrate that even though ATF1 and CREB homodimers, as well as the ATF1/CREB heterodimer efficiently bind to the CGTCA motif, the resulting DNA-protein complexes have significantly different stabilities. A region outside the DNA binding domain of ATF1 contributes to the instability of its interaction with DNA. We further show that despite ATF1's homology to CREB, it responds poorly to activation by protein kinase A. In light of our finding that in Hela cells the majority of CREB protein is heterodimerised with ATF1, we speculate on the functional significance of such heterodimers.

Various mutants of adenovirus E1A were assayed for their ability to complement the growth defect at the nonpermissive temperature for the cell line tsa14 which was isolated by immortalizing rat embryo fibroblasts with the thermolabile large T antigen of tsA58. This cell line grows indefinitely at the permissive temperature but undergoes rapid growth arrest upon shift up to the nonpermissive temperature. Since this growth arrest can be overcome by introduction of wild-type simian virus 40 large T antigen, human papillomavirus 16 E7, and adenovirus E1A, the tsa14 cells provided an excellent system for defining regions of E1A necessary for complementation of the growth defect. We demonstrate that conserved region 1 (CR1) is the region of E1A required for complementation. While CR2 of E1A has been shown to be required for the immortalization of primary cells and is also necessary for the binding of the 105-kDa retinoblastoma protein, mutations within this region did not abrogate complementation of the growth defect. However, since both CR1 and CR2 have previously been shown to be absolutely required for immortalization of primary cells by adenovirus E1A, this evidence suggests that the tsa14 system assays for the maintenance of proliferation and that this requires CR1.

The SV40 promoter is expressed well in the fission yeast S. pombe, and it initiates transcription at the same site as in mammalian cells. The majority of the enhancer sequences, however, do not contribute to this activity. DNAase I footprint analysis of the promoter revealed the presence of an AP-1-like factor in S. pombe cells that protects a region of the promoter almost identical to that protected by human AP-1. The specificity of binding of the yeast and mammalian AP-1 proteins was found to be similar. We have found two AP-1-like binding activities in budding yeast cells, one of which appears quite distinct from the binding activity of the product of the budding yeast GCN4 gene. We also demonstrate that in fission yeast the AP-1 binding site can act as an upstream activating sequence. The DNA-protein complexes containing the mammalian AP-1 and fission yeast AP-1-like factors are sensitive to phosphatase treatment, indicating that they may be phosphorylated.

We have investigated the E1A-inducible E3 promoter of adenovirus type 5 with respect to its ability to bind specific nuclear proteins. Four distinct nucleoprotein-binding sites were detected, located between positions-7 to -33, -44 to -68, -81 to -103, and -154 to -183, relative to the E3 cap site. These sites contain sequences previously shown to be functionally important for efficient E3 transcription. No major qualitative or quantitative differences were found in the binding pattern between nucleoprotein extracts prepared from uninfected or adenovirus-infected HeLa cells. Competition experiments suggest that the factors binding to the -154 to -183 and -81 to -103 sites are the previously identified nucleoproteins, NF1 and AP1, respectively. The factor binding to the -44 to -68 site, which we term ATF, also interacts with other E1A-inducible promoters and is very similar and probably identical to the factor that binds to the cAMP-responsive element of somatostatin. We have purified this factor, which is a protein of 43 kD in size.

A product of the adenovirus E1A gene is a positive regulator of early viral gene expression. In this report we show that E1A regulates at the transcriptional level and that sequences located 5' to the early viral regions contain sites which confer regulation by the E1A gene product. We constructed chimeric genes in which the sequences at the 5' end of the E2A, E3, and E4 regions were fused to the structural sequences of either the herpes simplex virus thymidine kinase gene, the bacterial gene encoding the enzyme neomycin phosphotransferase, or the chloramphenicol acetyltransferase gene. In all cases, expression of the chimeric genes was induced by a product of the E1A region. It was also found that the insertion of a fragment from the left-hand end of the adenovirus type 5 genome into a plasmid harboring the thymidine kinase gene resulted in elevated frequencies of transformation of TK- cells to TK+. The elevated transformation frequencies were only detected when the insert and tk gene were covalently joined. This effect occurred even when the insert was several kilobase upstream from, and regardless of its orientation to, the transcriptional initiation site of the tk gene. We propose that this region of the adenovirus type 5 genome harbors a cis-acting enhancer of transcription.

The ability of injected heterologous mRNAs to compete with endogenous mRNAs in Xenopus oocytes was assayed. In confirmation of previous reports, globin mRNA translation in oocytes results in a concomitant decrease in endogenous protein synthesis. In contrast, injection of adenovirus 5 mRNA into the oocyte results in a stimulation of endogenous protein synthesis. The stimulation is dose dependent and does not require nuclear transcription in the oocyte. Preliminary mapping data suggest that the stimulatory RNA is a product of one of the viral immediate early genes.

The human ATF and AP1 transcription factors bind to highly related DNA sequences. Their consensus binding sites differ by a single nucleotide, but this single change is crucial in determining factor binding specificity. We have previously identified an AP1 (yAP1) binding activity in yeast. In this report we identify a yeast ATF (yATF) binding activity whose specificity can be distinguished from that of yAP1 by the same crucial nucleotide that distinguishes binding of human ATF and AP1. The ATF binding site can act as an efficient upstream activating sequence in vivo, suggesting that yATF is a transcriptional activator. The yATF DNA-binding complex is phosphorylated and the binding activity of partially purified yATF can be enhanced in vitro by the addition of protein kinase A, indicating that the phosphorylation state of yATF may be important in determining its ability to bind DNA.

We previously purified and characterized functionally the Escherichia coli-expressed product of the human subgroup C adenovirus E1A 13S mRNA (B. Ferguson, N. Jones, J. Richter, and M. Rosenberg, Science 224:1343-1346, 1984; B. Krippl, B. Ferguson, M. Rosenberg, and H. Westphal, Proc. Natl. Acad. Sci. USA 81:6988-6992, 1984). We have now expressed in E. coli and purified the protein product encoded by the human subgroup C adenovirus E1A 12S mRNA and have compared the functional properties of this protein with those of the E1A 13S mRNA product. Using microinjection techniques to introduce these proteins into mammalian cells, we found that the E1A 12S mRNA product, like the 13S mRNA product, localized rapidly to the cell nucleus and induced adenovirus gene expression. Although both E1A gene products localized to the nucleus and stimulated adenovirus gene transcription, these proteins did not directly bind to DNA under conditions in which a known DNA-binding protein, the human c-myc gene product, bound DNA efficiently. Thus, the E1A and myc gene products, which have been related both structurally and functionally, exhibit distinctly different biochemical properties.

The regulation of adenovirus type 5 gene expression by the E1a gene product was examined in microinjected Xenopus laevis oocytes. Chimeric genes were constructed which included the promoter region of early adenovirus type 5 gene 3 and the structural sequence which codes for the bacterial enzyme chloramphenicol-3-O-acetyltransferase (CAT). A plasmid containing this chimeric gene as well as plasmids containing the E1a gene were coinjected into oocyte nuclei. The presence of the E1a gene was shown to increase CAT activity by up to 8.5-fold over basal levels. Synthesis of the functional product from the E1a gene requires the removal of intron sequences by RNA splicing. The E1a gene and a derivative that precisely lacks the intron were equally effective in increasing CAT activity, suggesting that splicing of the primary E1a transcript is efficiently accomplished in the oocyte nucleus. This was confirmed by directly examining the E1a mRNAs by the S1 mapping procedure. A protein extract from adenovirus type 5-infected HeLa cells enriched for the E1a protein may supplant the E1a plasmid in enhancing CAT activity. Synthesis of the CAT enzyme after gene injection is invariant in oocytes from the same frog, but oocytes from different frogs show a high degree of variability in their ability to synthesize the CAT enzyme. Microinjected X. laevis oocytes appear to be an extremely useful system to study the effects of protein elements on transcription.

The human type C adenovirus E1a 13S messenger RNA encodes a gene product, that positively regulates the transcription of viral genes and certain cellular genes and is involved in the transformation of primary mammalian cells. The E1a gene product was expressed at high levels in Escherichia coli. In a Xenopus oocyte microinjection assay, the purified Escherichia coli-produced protein activated the E1a-responsive adenovirus E3 promoter and functioned as efficiently as the E1a gene itself.

The purified Escherichia coli-expressed human subgroup C adenovirus E1A 13S mRNA product induces expression from the adenovirus type 5 E3 promoter when injected into Xenopus oocytes. In the present communication, the E. coli-expressed E1A 13S and 12S mRNA products are shown to undergo a posttranslational modification in microinjected Xenopus oocytes, which causes a 2- to 4-kDa increase in apparent molecular size, identical to that occurring in HeLa cells expressing the E1A gene. The E. coli-expressed E1A proteins are similarly modified in vitro in a soluble rabbit reticulocyte lysate. The modified form of the E1A proteins preferentially localizes to the oocyte nucleus following cytoplasmic microinjection. The use of various deleted forms of E1A protein synthesized in E. coli shows that a first exon-encoded domain of E1A, residing between amino acid residues 23 and 120, is sufficient for the posttranslational modification and nuclear localization of E1A and also for the trans-activation of the E3 promoter by E1A in Xenopus oocytes. These results suggest that the posttranslational modification of E1A protein may be important for its function.

To identify sequences within the adenovirus-5 E3 promoter necessary for E1A trans-activation, a series of promoter deletion mutants were constructed and analysed. A region between positions -82 and -105 was shown to be critical both for E1A induced expression as well as uninduced expression. The importance of this region was confirmed by constructing hybrid promoters consisting of E3 and Herpes simplex virus thymidine kinase sequences. The E1A insensitive tk promoter could be converted to an E1A sensitive promoter by replacing sequences upstream of position -79 with the corresponding region of the E3 promoter. This critical region of the E3 promoter contains a sequence 5' AGATGACTA3' which is also present in important upstream regions of the E2A and E4 promoters.

Expression of the human gene encoding the major heat shock protein, HSP70, was induced during cell growth by serum stimulation and after infection with adenovirus 5. In this study we showed that HSP70 gene expression could be induced by adenovirus 5 infection, even in the absence of exogenous serum factors. Whereas serum stimulation induced the expression of the endogenous HSP70 gene, it had no effect on early adenovirus promoters. However, expression of both the cellular HSP70 gene and the adenovirus E3 promoter were activated during adenovirus infection. By using a collection of reconstructed mutant viruses, we identified the 13S product of the E1A region as the specific transcriptional trans-activator of the HSP70 gene.

Type 5 adenovirus mutants that differentially express E1A 13S, 12S, or 9S mRNAs were constructed to study the role of their gene products in transformation. H5dl520 expresses the 243-amino-acid (AA) protein encoded in the 12S mRNA but not the 13S mRNA-encoded 289-AA protein. This mutant transformed both cloned rat embryo fibroblast (CREF) cells and baby rat kidney (BRK) cells at a frequency 40-100 times greater than did wild-type viruses. In addition, all of the foci produced were fibroblastic and grew very slowly at 32 degrees C. In contrast, H5dl21, which was mutated so that only the 54-AA protein encoded by the 9S mRNA was synthesized, did not transform either cell type. DNA transfection studies with plasmids from which these mutants were constructed demonstrated that the differences in transformation frequencies were not as marked. The plasmid pD1/D2, which directs the synthesis of the 54-AA protein only, was found to transform baby rat kidney cells at low frequency, provided the gene was linked to a fragment from the simian virus 40 genome containing the transcriptional enhancer element.

The adenovirus E1A gene controls initial stages of oncogenic transformation (1–5) and activates transcription of a variety of viral and host cell genes (6–11). We have a strong interest in the molecular details of these important regulatory functions. For this reason, some of us (12) have inserted the E1A coding sequence in a prokaryotic expression vector in order to generate the quantities of pure E1A proteins that are required to study their action. Here, we summarize results, to be published in detail elsewhere, which have allowed us to define more closely those sequences of the E1A gene that determine the intracellular location of E1A proteins as well as their functions during the virus life cycle.

Most genes in eukaryotic cells and their viruses contain non-coding intervening sequences (introns)1–5. These sequences are excised from each transcript during maturation of mRNA by a process known as RNA splicing. Although very little is known about the mechanism by which splicing occurs or its biological significance, several studies have suggested that the removal of introns is a necessary prerequisite to the accumulation of stable, cytoplasmic mRNA6–10. Here we describe a mutant of adenovirus 5 (Ad5) containing a deletion within early region Ela; the deletion prevents splicing of the normal Ela transcripts. The unspliced transcript is found in the cytoplasm of infected cells, showing that for this gene, splicing is not necessary for the biosynthesis of cytoplasmic message.

We efficiently expressed, in Escherichia coli, and purified the protein product encoded by the human adenovirus type 12 (Ad12) 13S mRNA. The functional properties of the E1A protein were analyzed by introducing the protein by microinjection or protoplast fusion into living mammalian cells. We showed that the E. coli-expressed E1A protein induces gene expression of the adenovirus type 5 (Ad5) E1A deletion mutant Ad5dl312. The purified E1A protein rapidly and quantitatively localized to the cell nucleus after microinjection into the cytoplasm. In addition, we raised high-titered monospecific antibodies to the purified Ad12 E1A protein. Using deleted forms of an adenovirus type 2 and Ad5 hybrid (Ad2/5) E1A protein, we showed that all of the epitopes conserved between Ad2/5 E1A and Ad12 E1A protein that are recognized by the Ad12 E1A-specific antiserum map to within the first exon-encoded amino-terminal half of the protein.