The transcription factor ZEB1 (deltaEF1) promotes tumour cell dedifferentiation by repressing master regulators of epithelial polarity.

Epithelial to mesenchymal transition (EMT) is implicated in the progression of primary tumours towards metastasis and is likely caused by a pathological activation of transcription factors regulating EMT in embryonic development. To analyse EMT-causing pathways in tumourigenesis, we identified transcriptional targets of the E-cadherin repressor ZEB1 in invasive human cancer cells. We show that ZEB1 repressed multiple key determinants of epithelial differentiation and cell-cell adhesion, including the cell polarity genes Crumbs3, HUGL2 and Pals1-associated tight junction protein. ZEB1 associated with their endogenous promoters in vivo, and strongly repressed promotor activities in reporter assays. ZEB1 downregulation in undifferentiated cancer cells by RNA interference was sufficient to upregulate expression of these cell polarity genes on the RNA and protein level, to re-establish epithelial features and to impair cell motility in vitro. In human colorectal cancer, ZEB1 expression was limited to the tumour-host interface and was accompanied by loss of intercellular adhesion and tumour cell invasion. In invasive ductal and lobular breast cancer, upregulation of ZEB1 was stringently coupled to cancer cell dedifferentiation. Our data show that ZEB1 represents a key player in pathologic EMTs associated with tumour progression.

Anti-A2/RA33 autoantibodies are directed to the RNA binding region of the A2 protein of the heterogeneous nuclear ribonucleoprotein complex. Differential epitope recognition in rheumatoid arthritis, systemic lupus erythematosus, and mixed connective tissue disease.

The recently described anti-A2/RA33 autoantibodies occur in 20-40% of patients with RA, SLE, and mixed connective tissue disease (MCTD). They are directed to the A2 protein of the heterogeneous nuclear ribonucleoprotein complex (hnRNP-A2), an abundant nuclear protein associated with the spliceosome. The NH2-terminal half of the antigen contains two conserved RNA binding domains whereas its COOH-terminal part is extremely glycine-rich. The aim of this study was to characterize the autoepitopes of hnRNP-A2 and to investigate the effects of anti-A2/RA33 autoantibodies on possible functions of the antigen. Using bacterially expressed fragments, two major discontinuous epitopes were identified. One containing the complete second RNA binding domain was recognized by the majority of patients with RA and SLE but not by patients with MCTD. The second epitope contained sequences of both RNA binding domains and was preferentially targeted by patients with MCTD. When the RNA binding properties of the antigen were investigated, oligoribonucleotides containing the sequence motif r(UUAG) were found to bind to a site closely adjacent or overlapping with the epitope targeted by autoantibodies from patients with RA and SLE. Moreover, anti-A2/RA33 autoantibodies from patients with RA or SLE, but not from patients with MCTD, inhibited binding of RNA. Thus, anti-A2/RA33 autoantibodies recognize conformation-dependent epitopes located in a functionally important region of the antigen. Furthermore, the specific recognition of an epitope by MCTD patients may be used as another argument in favor of considering MCTD a distinct connective tissue disease.

Stromal-derived IGF2 promotes colon cancer progression via paracrine and autocrine mechanisms.

The insulin-like growth factor (IGF)2/IGF1 receptor (IGF1R) signaling axis has an important role in intestinal carcinogenesis and overexpression of IGF2 is an accepted risk factor for colorectal cancer (CRC) development. Genetic amplifications and loss of imprinting contribute to the upregulation of IGF2, but insufficiently explain the extent of IGF2 expression in a subset of patients. Here, we show that IGF2 was specifically induced in the tumor stroma of CRC and identified cancer-associated fibroblasts (CAFs) as the major source. Further, we provide functional evidence that stromal IGF2, via the paracrine IGF1R/insulin receptor axis, activated pro-survival AKT signaling in CRC cell lines. In addition to its effects on malignant cells, autocrine IGF2/IGF1R signaling in CAFs induced myofibroblast differentiation in terms of alpha-smooth muscle actin expression and contractility in floating collagen gels. This was further augmented in concert with transforming growth factor-ß (TGFß) signaling suggesting a cooperative mechanism. However, we demonstrated that IGF2 neither induced TGFß/smooth muscle actin/mothers against decapentaplegic (SMAD) signaling nor synergized with TGFß to hyperactivate this pathway in two dimensional and three dimensional cultures. IGF2-mediated physical matrix remodeling by CAFs, but not changes in extracellular matrix-modifying proteases or other secreted factors acting in a paracrine manner on/in cancer cells, facilitated subsequent tumor cell invasion in organotypic co-cultures. Consistently, colon cancer cells co-inoculated with CAFs expressing endogenous IGF2 in mouse xenograft models exhibited elevated invasiveness and dissemination capacity, as well as increased local tumor regrowth after primary tumor resection compared with conditions with IGF2-deficient CAFs. In line, expression of IGF2 correlated with elevated relapse rates and poor survival in CRC patients. In agreement with our results, high-level coexpression of IGF2 and TGFß was predicting adverse outcome with higher accuracy than increased expression of the individual genes alone. Taken together, we demonstrate that stroma-induced IGF2 promotes colon cancer progression in a paracrine and autocrine manner and propose IGF2 as potential target for tumor stroma cotargeting strategies.

Autocrine WNT2 signaling in fibroblasts promotes colorectal cancer progression.

The canonical WNT signaling pathway is crucial for intestinal stem cell renewal and aberrant WNT signaling is an early event in colorectal cancer (CRC) development. Here, we show for the first time that WNT2 is one of the most significantly induced genes in CRC stroma as compared to normal stroma. The impact of stromal WNT2 on carcinoma formation or progression was not addressed so far. Canonical WNT/ß-catenin signaling was assessed using a 7TGP-reporter construct. Furthermore, effects of WNT2 on fibroblast migration and invasion were determined using siRNA-mediated gene silencing. Tumor cell invasion was studied using organotypic raft cultures and in vivo significance was assessed via a xenograft mouse model. We identified cancer-associated fibroblasts (CAFs) as the main source of WNT2. CAF-derived WNT2 activated canonical signaling in adenomatous polyposis coli/ß-catenin wild-type colon cancer cells in a paracrine fashion, whereas no hyperactivation was detectable in cell lines harboring mutations in the adenomatous polyposis coli/ß-catenin pathway. Furthermore, WNT2 activated autocrine canonical WNT signaling in primary fibroblasts, which was associated with a pro-migratory and pro-invasive phenotype. We identified FZD8 as the putative WNT2 receptor in CAFs. Three-dimensional organotypic co-culture assays revealed that WNT2-mediated fibroblast motility and extracellular matrix remodeling enhanced cancer cell invasion of cell lines even harboring mutations in the adenomatous polyposis coli/ß-catenin pathway. Thus, suggesting a tumor-promoting influence on a broad range of CRC. In line, WNT2 also promotes tumor growth, invasion and metastasis in vivo. Moreover, high WNT2 expression is associated with poor prognosis in human CRC. The identification of the pro-malignant function of stromal derived WNT2 in CRC classifies WNT2 and its receptor as promising stromal targets to confine cancer progression in combination with conventional or targeted therapies.

Increased expression of extracellular proteins as a hallmark of human endothelial cell in vitro senescence.

A convenient way to study processes of aging in distinct human tissues consists of a molecular analysis of cells from the tissue in question, that were explanted and grown in vitro until they reach senescence. Using human umbilical vein endothelial cells (HUVEC), we have established an in vitro senescence model for human endothelial cells. A major hallmark of HUVEC in vitro senescence is the increased frequency of apoptotic cell death, which occurs as a determining feature of HUVEC senescence. Senescent endothelial cells are also found in vivo in atherosclerotic lesions, suggesting that the presence of such cells may contribute to the development of vascular pathology. To elucidate mechanisms underlying endothelial cell senescence and age-associated apoptosis, gene expression analyses were carried out. In these experiments, we observed the up-regulation of genes coding for extracellular proteins in senescent HUVEC. In particular, a significant upregulation of interleukin-8, VEGI, and the IGF-binding proteins 3 and 5 was observed. Upregulation of these genes was confirmed by both RT-PCR and Western blot. In the case of interleukin-8, a roughly 50-fold upregulation of the protein was also found in cellular supernatants. The extracellular proteins encoded by these genes are well known for their ability to modulate the apoptotic response of human cells, and in the case of interleukin-8, clear links to the establishment of atherosclerotic lesions have been defined. The results described here support a new model, where changes in the secretome of human endothelial cells contribute to vascular aging and vascular pathology.

Identification of a novel gene, CDCP1, overexpressed in human colorectal cancer.

We report the identification of a novel human tumor associated gene, CDCP1 (Cub Domain Containing Protein), which was identified using representational difference analysis and cDNA chip technology. The gene consists of eight exons, the upstream region of which neither contains a TATA- nor a CCAAT-box. However, a CpG island is located around the transcription start, which is found in approximately 60% of known genes. The CDCP1 gene was mapped to chromosome 3p21-p23 by fluorescence in situ hybridization. For expression profiling real time quantitative RT--PCR was performed using cell lines and laser capture microdissected colon cancer biopsies. CDCP1 mRNA is approximately 6 kb and highly overexpressed in human colon cancer and lung cancer. CDCP1 represents a putative transmembrane protein, containing three CUB domains in the extracellular part most likely involved in cell adhesion or interacting with the extracellular matrix.

Human interferon omega 1: isolation of the gene, expression in Chinese hamster ovary cells and characterization of the recombinant protein.

A gene encoding human interferon omega-1 (IFN-omega 1) was isolated from a cosmid library, sequenced and expressed in Chinese hamster ovary (CHO) cells under the control of an SV40-derived promoter/enhancer sequence. Culture supernatants of stably transfected cell clones contained biologically active IFN-omega 1 at concentrations up to 10 micrograms/l. Amplification of the expression vector containing a dhfr gene under methotrexate selection pressure resulted in yields up to 200 micrograms/l. Production of IFN-omega 1 was further enhanced 2- to 3-fold by propagation of the cells in the presence of n-butyrate. IFN-omega 1 was purified from culture supernatants by monoclonal antibody affinity chromatography. The resulting protein was at least 95% pure as determined by reverse-phase HPLC and size-exclusion HPLC. Sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS-PAGE) showed two bands of about the same intensity with apparent molecular masses of 24.5 and 22.5 kDa. Upon treatment with peptide:N-glycosidase F, both bands were shifted to lower molecular masses (20.5 and 18.5 kDa), indicating that CHO cell-derived IFN-omega 1 is glycosylated; Asn-78 was identified as the glycosylation site. Analysis of the carbohydrate moiety using glycosidases and lectins revealed the presence of biantennary complex oligosaccharides containing neuraminic acid. Amino acid sequencing showed that only about 40% of the molecules have the expected N-terminus, whereas the others carry two additional amino acids derived from the signal sequence. C-terminal amino acid sequencing using carboxypeptidase P demonstrated that the smaller form of the protein lacks nine amino acids. Disulfide bridges were shown to connect Cys residues 1 and 99 as well as 29 and 139, respectively, as in IFN-alpha. The specific antiviral activity of recombinant, glycosylated human IFN-omega 1 on human cells was 2.6 x 10(8) IU/mg, not significantly different from that of the authentic, human leukocyte-derived protein.

IGFBP7, a novel tumor stroma marker, with growth-promoting effects in colon cancer through a paracrine tumor-stroma interaction.

The activated tumor stroma participates in many processes that control tumorigenesis, including tumor cell growth, invasion and metastasis. Cancer-associated fibroblasts (CAFs) represent the major cellular component of the stroma and are the main source for connective tissue components of the extracellular matrix and various classes of proteolytic enzymes. The signaling pathways involved in the interactions between tumor and stromal cells and the molecular characteristics that distinguish normal 'resting' fibroblasts from cancer-associated or '-activated' fibroblasts remain poorly defined. Recent studies emphasized the prognostic and therapeutic significance of CAF-related molecular signatures and a number of those genes have been shown to serve as putative therapeutic targets. We have used immuno-laser capture microdissection and whole-genome Affymetrix GeneChip analysis to obtain transcriptional signatures from the activated tumor stroma of colon carcinomas that were compared with normal resting colonic fibroblasts. Several members of the Wnt-signaling pathway and gene sets related to hypoxia, epithelial-to-mesenchymal transition (EMT) and transforming growth factor-ß (TGFß) pathway activation were induced in CAFs. The putative TGFß-target IGFBP7 was identified as a tumor stroma marker of epithelial cancers and as a tumor antigen in mesenchyme-derived sarcomas. We show here that in contrast to its tumor-suppressor function in epithelial cells, IGFPB7 can promote anchorage-independent growth in malignant mesenchymal cells and in epithelial cells with an EMT phenotype when IGFBP7 is expressed by the tumor cells themselves and can induce colony formation in colon cancer cells co-cultured with IGFBP7-expressing CAFs by a paracrine tumor-stroma interaction.

Spectroscopic characterization of rhinoviral protease 2A: Zn is essential for the structural integrity.

Recently, protease 2A of human rhinovirus 2 (HRV2 2A) was shown to require a zinc ion for the formation of an active enzyme although zinc is not involved mechanistically. The data presented clearly show that the zinc ion bound to a picornaviral-specific motif represents an essential component of the native structure, probably representing a new Zn-binding motif. This structure, containing mostly beta-strand elements as shown by CD spectroscopy, changes drastically upon removal of zinc. The zinc-depleted form does represent an intermediate with mostly unchanged secondary structure, but not a fully denatured random coil as obtained by guanidinium hydrochloride. This is indicated by the blue-shifted fluorescence spectra and by CD. The native protein exhibited a cooperative phase transition at 53 degrees C. In contrast, the zinc-depleted form did not show any transition at all, again demonstrating the stabilizing role of the zinc ion. A structural intermediate was observed during thermal and pH denaturation that may represent a molten globule, as suggested by its ANS binding.

The molecular biology of human rhinoviruses.

A brief review of recent advances in the understanding of human rhinovirus molecular biology is presented. The importance of recent findings on the elucidation of serotypic diversity and their implications for the viral host-cell receptor site are emphasized. An introduction to the genome structure and to the pathway of gene expression in rhinoviruses leads on to a discussion of the crystal structure of human rhinovirus 14 (HRV14) and the antibody-inducing regions on the surface of the capsid. Evidence from these experiments indicates that four sites on HRV14 are responsible for inducing neutralizing antibodies. Amino-acid sequence comparisons reveal that these sites are different in other serotypes, strengthening the view that the sequences of these regions are fundamental in determining and defining rhinovirus serotypes. The crystal structure of HRV14 points to a depression in the viral capsid as being the site of binding to the host-cell receptor; however, residues involved in binding cannot yet be identified.

The structures of picornaviral proteinases.

Picornaviruses are a family of positive-strand RNA viruses the members of which include poliovirus, hepatitis A virus, rhinovirus, foot-and-mouth disease virus and encephalomyocarditis virus. The genetic information contained in the single-stranded, positive sense RNA genome is expressed as a single protein of around 2000 amino acids. This primary product of protein synthesis, designated the polyprotein, is subsequently cleaved into the mature viral proteins by proteinases present within it. The properties of the three defined proteolytic activities present in the picornaviruses are reviewed and the three-dimensional structures of the hepatitis A 3C proteinase and the leader proteinase of foot-and-mouth disease virus as well as a model of the structure of the HRV2 2A proteinase are compared with those of chymotrypsin, papain and streptomyces griseus A proteinase, respectively.

Anti-RA33 autoantibodies may recognize epitopes in the N-terminal region of hnRNP-A2 (RA33).

The nuclear autoantigen RA33, which is identical to the A2 protein of the heterogeneous nuclear ribonucleoprotein (hnRNP-A2), is a nucleic acid binding protein of 341 amino acids. The N-terminal part contains two RNA binding domains whereas the C-terminal part consists of a long glycine-rich region starting around amino acid 192. Autoantibodies to hnRNP-A2/RA33 can be detected in 20-40% of sera from RA, SLE and MCTD patients. So far, it has not been known which regions of A2/RA33 are recognized by these autoantibodies. To address this issue, tryptic fragments of natural A2/RA33 were investigated by immunoblotting using 14 sera from anti-RA33 positive patients with RA (n = 5), SLE (n = 5) and MCTD (n = 4). Most sera reacted with a 22 kD fragment comprising the N-terminal part of the protein. However, a smaller 18 kD fragment was recognized only by 3 RA and 3 MCTD sera whereas two of five SLE sera were reactive with two larger fragments of 26 and 29 kD. In order to further characterize the epitope(s) C-terminal deletion mutants of recombinant A2/RA33 were investigated by immunoblotting employing 27 sera from anti-RA33 positive patients with RA (n = 10), SLE (n = 8), and MCTD (n = 9). All sera recognized a fragment terminating at amino acid 212 which contained the complete N-terminal region as well as 20 amino acids of the glycine-rich section. Thus, these data indicate that the N-terminal part of A2/RA33 contains epitopes for antiA2/RA33 autoantibodies.(ABSTRACT TRUNCATED AT 250 WORDS)

Cleavage specificity of purified recombinant hepatitis A virus 3C proteinase on natural substrates.

Hepatitis A virus (HAV) 3C proteinase expressed in Escherichia coli was purified to homogeneity, and its cleavage specificity towards various parts of the viral polyprotein was analyzed. Intermolecular cleavage of the P2-P3 domain of the HAV polyprotein gave rise to proteins 2A, 2B, 2C, 3ABC, and 3D, suggesting that in addition to the primary cleavage site, all secondary sites within P2 as well as the 3C/3D junction are cleaved by 3C. 3C-mediated processing of the P1-P2 precursor liberated 2A and 2BC, in addition to the structural proteins VP0, VP3, and VP1-2A and the respective intermediate products. A clear dependence on proteinase concentration was found for most cleavage sites, possibly reflecting the cleavage site preference of 3C. The most efficient cleavage occurred at the 2A/2B and 2C/3A junctions. The electrophoretic mobility of processing product 2B, as well as cleavage of the synthetic peptide KGLFSQ*AKISLFYT, suggests that the 2A/2B junction is located at amino acid position 836/837 of the HAV polyprotein. Furthermore, using suitable substrates we obtained evidence that sites VP3/VP1 and VP1/2A are alternatively processed by 3C, leading to either VP1-2A or to P1 and 2A. The results with regard to intermolecular cleavage by purified 3C were confirmed by the product pattern derived from cell-free expression and intramolecular processing of the entire polyprotein. We therefore propose that polyprotein processing of HAV relies on 3C as the single proteinase, possibly assisted by as-yet-undetermined viral or host cell factors and presumably controlled in a concentration-dependent fashion.

Cleavage specificity of human rhinovirus-2 2A protease for peptide substrates.

Substrate requirements of the human rhinovirus serotype-2 2A protease have been examined using synthetic peptides. A chromogenic peptide with a sequence of TRPIITTA-p-nitroanilide was found to be cleaved efficiently by the 2A protease with an apparent Km value of 95 microM, which allowed the protease activity to be monitored and measured continuously using a spectrophotometer. Competition cleavage assays reveal this peptide was cleaved over 10-fold more efficiently than the 16-mer peptide derived directly from its native processing site. On the basis of these data, we conclude that the P1' glycine residue is not absolutely needed for the 2A cleavage to occur and the essential residues required for the 2A activity would exist within the N-terminal side of the scissile bond.

Variability in gene expression patterns of Ewing tumor cell lines differing in EWS-FLI1 fusion type.

Type 1 and type 2 EWS-FLI1 fusion products result from variation in breakpoint locations arising from the t(11;22)(q24;q12) recurrent chromosomal translocation in Ewing's sarcoma family tumors (EFT). Previously, studies from our institution (updated in the present communication at a median follow-up of more than 6 years) and others suggested a prognostic difference for EFT patients with localized disease depending on the type of EWS-FLI1 fusion present in the tumor. It has been suggested that the observed clinical discrepancies result from different transactivation potentials of the various EWS-FLI1 fusion proteins. In an attempt to identify genes whose expression levels are differentially modulated by structurally different EWS-FLI1 transcription factors, we have used two related PCR-based subtractive approaches, cDNA representational difference analysis (cDNA-RDA) and linker-capture subtraction (LCS) to compare transcript representations in cDNA pools of type 1 versus type 2 EFT cell lines. About 800 clones obtained by the two approaches were analyzed by dot blot hybridization to cDNA pools. Eighty-six clones showing the highest variability in signal intensities on the dot blots were further hybridized to individual EFT cell line RNAs on Northern blots, and four of them were additionally studied by real-time quantitative PCR (RTQ-PCR). Although interindividual variations in gene expression patterns in the range of one- to several-fold were observed, no correlation to specific EWS-FLI1 fusion types could be identified. Among the genes differentially expressed in individual EFT cell lines are several previously implicated in tumor growth, invasion, and metastasis. Although our data may have revealed candidate genes whose composite expression pattern may be relevant for the biology of individual EFT, they do not support a role of distinct EWS-FLI1 fusion types for EFT prognosis based on different transactivation potentials.

Human rhinovirus serotype 2: in vitro synthesis of an infectious RNA.

A complete cDNA copy of human rhinovirus serotype 2 RNA was placed under the control of a T7 RNA polymerase promoter. An in vitro transcribed RNA containing two extra G residues at the 5' end gave rise to plaques on transfection into HeLa cells. The efficiency was approximately half that obtained with viral RNA. On the contrary, an in vitro synthesized RNA containing 16 additional nucleotides at the 5' end was not infectious. This ability to make an infectious in vitro transcribed RNA will be useful in studying the characteristics of viruses using the human rhinoviral minor group receptor.

Intermolecular cleavage of hepatitis A virus (HAV) precursor protein P1-P2 by recombinant HAV proteinase 3C.

Active proteinase 3C of hepatitis A virus (HAV) was expressed in bacteria either as a mature enzyme or as a protein fused to the entire polymerase 3D or to a part of it, and their identities were shown by immunoblot analysis. Intermolecular cleavage activity was demonstrated by incubating in vitro-translated and radiolabeled HAV precursor protein P1-P2 with extracts of bacteria transformed with plasmids containing recombinant HAV 3C. Identification of cleavage products P1, VP1, and VPO-VP3 by immunoprecipitation clearly demonstrates that HAV 3C can cleave between P1 and P2 as well as within P1 and thus shows an activity profile similar to that of cardiovirus 3C.

Using proteinase trapping to detect revertants of inactive rhinoviral 2A proteinase mutants.

The 2A proteinase of human rhinovirus 2 cleaves itself off the growing polyprotein at its own N terminus during translation; this property was used to develop an in vivo screening system with the lacZ gene fragment of M13mp18. The fusion of an active 2A proteinase to the C-terminus of the alpha-fragment did not affect alpha-complementation, as the proteinase cleaved itself off the alpha-fragment. However, an inactive 2A proteinase remained fused to the alpha-fragment hindering alpha-complementation. Random mutations were then introduced into the 2A gene site by PCR amplification. Mutants defective in alpha-complementation (thus containing an inactive 2A proteinase) were obtained at an efficiency of 5%, mutants showing reduced 2A activity at an efficiency of 1%. Mutants showing reduced or no 2A activity were then subjected to PCR mutagenesis. Three mutants reactivating an inactive 2A proteinase were examined and the compensatory changes determined.

Mutational analyses support a model for the HRV2 2A proteinase.

The proteinase 2A of human rhinovirus 2 is a cysteine proteinase which contains a tightly bound Zn ion thought to be required for structural integrity. A three-dimensional model for human rhinovirus type 2 proteinase 2A (HRV2 2A) was established using sequence alignments with small trypsin-like Ser-proteinases and, for certain regions, elastase. The model was tested by expressing selected proteinase 2A mutants in bacteria and examining the effect on both intramolecular ("cis") and intermolecular ("trans") activities. The HRV2 proteinase 2A is proposed to have a two domain structure, with the catalytic site and substrate binding region on one face of the molecule and a Zn-binding motif on the opposite face. Residues Gly 123, Gly 124, Thr 121, and Cys 101 are proposed to be involved in the architecture of the substrate binding pocket and to provide the correct environment for the catalytic triad of His 18, Asp 35, and Cys 106. Residues Tyr 85 and Tyr 86 are thought to participate in substrate recognition. The presence of an extensive C-terminal helix, in which Asp 132, Arg 134, Phe 130, and Phe 136 play important roles, explains why mutations in this region are generally detrimental to proteinase activity. The proposed Zn-binding motif comprises Cys 52, Cys 54, Cys 112, and His 114. Exchange of these residues inactivates the enzyme. Furthermore, as measured by atom emission spectroscopy, Zn was absent from purified preparations of proteinase 2A in which His 114 had been replaced by Asn. The absence of disulphide bridges was confirmed by subjecting highly purified HRV2 proteinase 2A to one- and two-step alkylation procedures.

2A proteinase of human rhinovirus cleaves cytokeratin 8 in infected HeLa cells.

Rhino- and enteroviruses encode two proteinases, 2A and 3C, which are responsible for the processing of the viral polyprotein and for cleavage of several cellular proteins. To identify further targets of the 2A proteinase of human rhinovirus serotype 2 (HRV2), an in vitro cleavage assay followed by two-dimensional electrophoresis was employed. Cytokeratin 8, a member of the intermediate filament group of proteins, was found to be proteolytically cleaved in vitro by the 2A proteinase of HRV2 and of coxsackievirus B4 and in vivo during HRV2 infection of HeLa cells. The cleavage results in removal of 14 amino acids from the N-terminal head domain of cytokeratin 8. However, other intermediate filament proteins (cytokeratins 7 and 18 and vimentin) were not cleaved in the course of the HRV2 infection. Compared with the processing of the eucaryotic translation initiation factors 4GI and 4GII, cleavage of cytokeratin 8 occurs late in the infection cycle at the time of the onset of the cytopathic effect.