Porcine Ear Fibroblasts Research Articles

Transcriptome heterogeneity of porcine ear fibroblast and its potential influence on embryo development in nuclear transplantation.

There is heterogeneity among donor cells of the same source. Many studies have shown that donor cell affects the efficiency of somatic cell nuclear transfer (SCNT). However, the potential influence of donor cell heterogeneity on the efficiency of nuclear transplantation were rarely analyzed at the single-cell level. In this study, single-cell transcriptome sequencing was performed on 52 porcine ear fibroblasts randomly selected from the same source to compare their gene expression patterns. The results showed that 48 cells had similar gene expression patterns, whereas 4 cells (D11_1, D12_1, DW61_2, DW99_2) had significantly different gene expression patterns from those of other cells. There were no two cells with identical gene expression patterns. The gene expression patterns of D11_1, D12_1, DW61_2 and DW99_2 were analyzed, using the 48 cells with similar gene expression patterns as controls. Firstly, we used the R language statistics to select the differentially expressed genes in the 4 single cells, and identified the top 50 most significant differentially expressed genes. Then GO enrichment analysis and KEGG pathway analysis were performed on the differentially expressed genes. Enrichment analysis revealed that the main molecular functions of the differentially expressed genes included energy metabolism, protein metabolism and cell response to stimulation. The main pathways from KEGG enrichment were related to cell cycle, cell metabolism, and DNA replication. Finally, based on the above results and in consideration with the SCNT research progress, we discussed the potential effects of differential gene expression patterns of the 4 single cells on the embryonic development efficiency of nuclear transplantation. This study revealed transcriptional heterogeneity of porcine ear tissue fibroblasts and provided an effective method to analyze elite donor cells, thereby providing new ideas on improving the cloning efficiency of SCNT.

Generation of Marker-Free pbd-2 Knock-in Pigs Using the CRISPR/Cas9 and Cre/loxP Systems.

Porcine β-defensin 2 (PBD-2), expressed by different tissues of pigs, is a multifunctional cationic peptide with antimicrobial, immunomodulatory and growth-promoting abilities. As the latest generation of genome-editing tool, CRISPR/Cas9 system makes it possible to enhance the expression of PBD-2 in pigs by site-specific knock-in of pbd-2 gene into the pig genome. In this study, we aimed to generate marker-free pbd-2 knock-in pigs using the CRISPR/Cas9 and Cre/loxP systems. Two copies of pbd-2 gene linked by a T2A sequence were inserted into the porcine Rosa26 locus through CRISPR/Cas9-mediated homology-directed repair. The floxed selectable marker gene neoR, used for G418 screening of positive cell clones, was removed by cell-penetrating Cre recombinase with a recombination efficiency of 48.3%. Cloned piglets were produced via somatic cell nuclear transfer and correct insertion of pbd-2 genes was confirmed by PCR and Southern blot. Immunohistochemistry and immunofluorescence analyses indicated that expression levels of PBD-2 in different tissues of transgenic (TG) piglets were significantly higher than those of their wild-type (WT) littermates. Bactericidal assays demonstrated that there was a significant increase in the antimicrobial properties of the cell culture supernatants of porcine ear fibroblasts from the TG pigs in comparison to those from the WT pigs. Altogether, our study improved the protein expression level of PBD-2 in pigs by site-specific integration of pbd-2 into the pig genome, which not only provided an effective pig model to study the anti-infection mechanisms of PBD-2 but also a promising genetic material for the breeding of disease-resistant pigs.

Aging adult porcine fibroblasts can support nuclear transfer and transcription factor-mediated reprogramming.

Somatic cell nuclear transfer (SCNT) and induced pluripotent stem cells (iPSCs) technology are two classical reprogramming methods. Donor cell types can affect the reprogramming results in the above two methods. We here used porcine embryonic fibroblasts (PEFs) and adult porcine ear skin fibroblasts (APEFs) and adipose-derived stem cells (ADSCs) as donor cells for SCNT and source cells for iPSCs to study their invitro developmental capability and colony-formation efficiency, respectively. For SCNT, fusion and cleavage rate has no significant difference among PEFs, ADSCs and APEFs. The rate and total cell number of blastocysts in the APEF group were significant lower than that in PEFs and ADSCs. For transcription factor-mediated reprogramming, the reprogramming efficiency of ADSCs were significantly higher than PEFs and APEFs and there is no significant difference between PEFs and APEFs. Furthermore, PEFs, APEFs and ADSCs can be used to generate iPSCs. Fianlly, somatic cloned pigs could still be successfully generated from APEFs, suggesting terminally differentiated aging adult somatic cells could be reprogrammed into a totipotent state. Considering the easy availability of animal tissue and the costs of establishing cell lines, aging porcine ear fibroblasts can support nuclear transfer-mediated and transcription factor-based reprogramming.

Expression analysis of green fluorescent protein in tissues and organs in α-1,3 galactosyltransferase knockout pigs

The pig is an ideal source to provide organs because its organ size and physiology are similar to humans. However, an acute rejection will ensue after pig-to-human xenotransplantation. The α-1,3 galactosyltransferase gene knockout (GTKO) pigs were generated in recent years, and could solve the problem of hyperacute rejection. But due to lack of reporting genes, the rejection status of cells and organs post pig-to-human xenotransplantation cannot be visualized. In this study, we introduced the enhanced green fluorescent protein (EGFP) gene driven by the CAG promoter into GTKO porcine ear fibroblasts. Then we produced transgenic pigs expressing the EGFP gene by nuclear transfer technology. Expression levels of EGFP in different tissues and organs of the cloned pig were investigated by Nightsea DFP-1 Fluorescent Protein Flashlight, fluorescence microscope and quantitative PCR assays. The results showed that the protein and transcript of EGFP were expressed in all tissues and organs of the GTKO pig, but the expression was weak in the liver and central nervous system. In conclusion, we have successfully produced the transgenic GTKO pigs expressing EGFP in all tested tissues and organs, which builds up a good basis to track transplanted cells or tissues.

Cell-free extract from porcine induced pluripotent stem cells can affect porcine somatic cell nuclear reprogramming.

Pretreatment of somatic cells with undifferentiated cell extracts, such as embryonic stem cells and mammalian oocytes, is an attractive alternative method for reprogramming control. The properties of induced pluripotent stem cells (iPSCs) are similar to those of embryonic stem cells; however, no studies have reported somatic cell nuclear reprogramming using iPSC extracts. Therefore, this study aimed to evaluate the effects of porcine iPSC extracts treatment on porcine ear fibroblasts and early development of porcine cloned embryos produced from porcine ear skin fibroblasts pretreated with the porcine iPSC extracts. The ChariotTM reagent system was used to deliver the iPSC extracts into cultured porcine ear skin fibroblasts. The iPSC extracts-treated cells (iPSC-treated cells) were cultured for 3 days and used for analyzing histone modification and somatic cell nuclear transfer. Compared to the results for nontreated cells, the trimethylation status of histone H3 lysine residue 9 (H3K9) in the iPSC-treated cells significantly decreased. The expression of Jmjd2b, the H3K9 trimethylation-specific demethylase gene, significantly increased in the iPSC-treated cells; conversely, the expression of the proapoptotic genes, Bax and p53, significantly decreased. When the iPSC-treated cells were transferred into enucleated porcine oocytes, no differences were observed in blastocyst development and total cell number in blastocysts compared with the results for control cells. However, H3K9 trimethylation of pronuclear-stage-cloned embryos significantly decreased in the iPSC-treated cells. Additionally, Bax and p53 gene expression in the blastocysts was significantly lower in iPSC-treated cells than in control cells. To our knowledge, this study is the first to show that an extracts of porcine iPSCs can affect histone modification and gene expression in porcine ear skin fibroblasts and cloned embryos.

High efficient differentiation of functional hepatocytes from porcine induced pluripotent stem cells.

Hepatocyte transplantation is considered to be a promising therapy for patients with liver diseases. Induced pluripotent stem cells (iPSCs) provide an unlimited source for the generation of functional hepatocytes. In this study, we generated iPSCs from porcine ear fibroblasts (PEFs) by overexpressing Sox2, Klf4, Oct4, and c-Myc (SKOM), and developed a novel strategy for the efficient differentiation of hepatocyte-like cells from porcine iPSCs by following the processes of early liver development. The differentiated cells displayed the phenotypes of hepatocytes, exhibited classic hepatocyte-associated bio-functions, such as LDL uptake, glycogen storage and urea secretion, as well as possessed the metabolic activities of cytochrome P-450 (CYP) 3A and 2C. Furthermore, we compared the hepatocyte differentiation efficacy of our protocol with another published method, and the results demonstrated that our differentiation strategy could significantly improve the generation of morphological and functional hepatocyte-like cells from porcine iPSCs. In conclusion, this study establishes an efficient method for in vitro generation of functional hepatocytes from porcine iPSCs, which could represent a promising cell source for preclinical testing of cell-based therapeutics for liver failure and for pharmacological applications.

Establishment and Characterization of Novel Porcine Induced Pluripotent Stem Cells Expressing hrGFP

Induced pluripotent stem (iPS) cells have been established in various animal species since 2006. The pig is a potentially useful model in human regenerative medicine, and the characters of porcine embryonic stem (pES) cells were much similar with human embryonic stem (hES) cells. In present study, the traceable humanized recombinant green fluorescent protein expressing porcine induced pluripotent stem (piPS/hrGFP+) cells were generated from porcine ear fibroblasts (pEF) by introducing four human transcription factors (Sox2, Oct4, Klf4, and c-Myc) constructed in lentivirus vectors. The piPS/hrGFP+ cells continuously and steadily expressed hrGFP signal continuously and steadily for more than 90 passages over 20 months despite repeated subcultures. They also retained the typical defined characteristics including continuous proliferation with undifferentiated status, expression of ES pluripotent pluripotency markers (Oct4, AP, SSEA-3, SSEA-4, TRA-1-60, and TRA-1-81), and maintenance of a normal karyotype (36+XY). Three embryonic germ layers were also successfully revealed from in vitro differentiation by EB formation. Various histological analysis and immunohistochemical staining of the teratomas revealed various tissues derived from three embryonic germ layers, including neural tissues, keratin-containing epidermal tissues, skeletal muscle, smooth muscle, cartilage, adipose tissues, and glandular structures. These results support that piPS/hrGFP+ cells can be generated from pEF by direct reprogramming, and these traceable piPS/hrGFP+ cells would be beneficial for future application on cell transplantation and tissue regeneration.

Generation of Leukemia Inhibitory Factor-Dependent Induced Pluripotent Stem Cells from the Massachusetts General Hospital Miniature Pig

The generation and application of porcine induced pluripotent stem cells (iPSCs) may enable the testing for safety and efficacy of therapy in the field of human regenerative medicine. Here, the generation of iPSCs from the Massachusetts General Hospital miniature pig (MGH minipig) established for organ transplantation studies is reported. Fibroblasts were isolated from the skin of the ear of a 10-day-old MGH minipig and transduced with a cocktail of six human factors: POU5F1, NANOG, SOX2, C-MYC, KLF4, and LIN28. Two distinct types of iPSCs were generated that were positive for alkaline phosphatase activity, as well as the classical pluripotency markers: Oct4, Nanog, Sox2, and the surface marker Ssea-1. Only one of two porcine iPSC lines differentiated into three germ layers both in vitro and in vivo. Western blot analysis showed that the porcine iPSCs were dependent on LIF or BMP-4 to sustain self-renewal and pluripotency. In conclusion, the results showed that human pluripotent factors could reprogram porcine ear fibroblasts into the pluripotent state. These cells may provide a useful source of cells that could be used for the treatment of degenerative and genetic diseases and agricultural research and application.

322 POTENTIAL OF PRIMARY PORCINE KIDNEY CELLS FOR THE GENERATION OF TRANSGENIC PIGS VIA SOMATIC CELL NUCLEAR TRANSFER

The generation of genetically tailored pig models for biomedical research using somatic cell nuclear transfer (SCNT) is an efficient and precise approach, whereas the outcome is crucially dependent on the source of nuclear donor cells. Especially for site-directed mutagenesis by homologous recombination, including the generation of single cell clones, the demands on the target cells are high. Different primary cells used for SCNT have been tested for their efficiency in SCNT experiments, but further characterisation of the specific cell types, their morphology, proliferation, lifespan, and stability of karyotype is mostly lacking. We have evaluated the potential of 2 primary porcine kidney cell lines (PKC) isolated from juvenile pigs by a simple collagenase digestion and culture in collagen-coated dishes as cell source for SCNT, including their morphology, proliferation capacity, transfection efficiency, and capacity to support full-term development of SCNT embryos after additive gene transfer or homologous recombination. Single cell clones generated by subcloning of PKC at passage 3 showed different morphologies, proliferation rates, and lifespan, indicating that PKC culture is a mixed population of different types of fibroblasts and/or other cells types. The PKC could be maintained in culture for up to 71 passages without signs of senescence and decreased proliferation, exhibiting a stable karyotype containing 74% normal chromosome numbers (2N = 38) determined from metaphase spreads. In contrast, porcine fetal fibroblasts (PFF) and porcine ear fibroblasts (PEF) could be not be passaged more than 20 times. The calculation of growth curves at passage 4 to 5 showed that PKC exhibited a higher proliferation rate with a population doubling time of 16.6 to 18.4 h compared with PFF (23.2. h) and PEF (32.9 h). Furthermore the determination of the developmental competence after SCNT using PKC at passage 4 in 3 independent experiments and in vitro cultivation for 7 days resulted in a higher blastocyst rate (21%) compared with that in PFF (9.1%) and PEF (4.3%). The comparison of different transfection methods (lipofection, nanofection, conventional electroporation, nucleofection), using an expression vector for green fluorescent protein (GFP), showed that the NucleofectorTM technology gave the best results with transfection efficiencies of 70 to 89%, high fluorescence intensity, low cytotoxicity, good cell proliferation, and almost no morphological signs of stress. So far, around 150 cloned piglets using 18 different gene constructs have been produced using stable transfected PKC after additive gene transfer and targeting of 3 different loci. These findings demonstrate that among the 3 tested types of donor cells, PKC, PFF, PEF, primary PKC have outstanding potential for the production of genetically modified pigs by SCNT. This work is supported by the DFG (FOR535, FOR793), the Bayerische Forschungsstiftung, and Mukoviszidose e.V.

327 PRODUCTION OF TRIPLE TRANSGENIC hHO-1/GGTA-1–/–/hCD55 TRANSGENIC PIGS USING SLEEPING BEAUTY TRANSPOSITION AND ZINC-FINGER NUCLEASES

Advances in xenotransplantation (pig-to-baboon or human transplantation) require multi-transgenic pigs with a homozygous knockout (KO) of the α1,3-galactosyltransferase gene (GGTA-1, encoding for Gal-epitopes) to control the hyperacute rejection response. To achieve prolonged survival of the porcine xenograft, transgenic expression of additional immune modulatory genes on the GGTA-1–/– background is considered the solution of choice. The acute vascular rejection (AVR) is primarily caused by endothelial cell activation and leads to rejection of GGTA-1–/– pig organs. Here, we set out to produce triple transgenic pigs with high expression of the human complement regulatory gene hCD55 (human decay acceleration factor, hDAF) and the anti-apoptotic and anti-inflammatory gene human heme oxygenase-1 (hHO-1), respectively. Porcine ear fibroblasts carrying a hHO-1 transgene (Petersen et al. 2011 Xenotransplantation 18, 355–368) were transfected with zinc-finger nucleases (ZFN) targeting the GGTA-1 gene, leading to a biallelic KO of the GGTA-1 gene in ~1% of the transfected cells (Hauschild et al. 2011 Proc. Natl. Acad. Sci. USA 108, 15 010). Somatic cell nuclear transfer (SCNT) was accomplished with cells from double transgenic cell line (hHO-1/GGTA-1–/–), and one pregnancy was terminated to obtain fetal cell cultures. Fluorescence-activated cell sorting (FACS) analysis revealed that all 11 fetuses (C6F1-F11) were lacking Gal-epitopes. One pure fetal cell culture (C6F5) was used for co-transfection with the newly designed SB-CAGGS-hDAF-Puror transposon, based on the Sleeping Beauty transposon plasmid pT2/HB, together with the SB transposase 100X plasmid (both kindly provided by Dr. Zoltan Ivics). After 14 days of selection with puromycin, cell colonies were picked and expanded to obtain cells for analysis and as a backup for SCNT. The PCR amplification with hDAF-specific primers revealed that ~50% of picked colonies had integrated the transgene. Two hDAF-positive colonies were pooled and used for SCNT. One pregnancy was sacrificed and 4 hHO-1/GGTA-1–/–/hCD55 fetuses were obtained for real-time PCR analysis and subsequent use in SCNT. Real-time PCR showed elevated hCD55 expression representing a 0.9-fold (0.7- to 1.0-fold) expression of the housekeeping gene glyceraldehyde 3-phosphate dehydrogenase (GAPDH), while FACS analysis confirmed absence of Gal-epitopes in all 4 fetuses. Binding of human-specific anti-CD55 antibodies showed high expression of hCD55 by FACS measurement. Triple transgenic fetuses C8F2 and C8F4 showed higher expression of hCD55 by real-time PCR and by FACS analysis compared with fetuses C8F1 and C8F3. Fetuses C8F3 and C8F4 were used for re-cloning experiments to produce live offspring, which will further be characterised and used in transplantation experiments. This work underlines the importance of new genomic technologies to further improve the efficiency of the generation of transgenic pigs suitable for xenotransplantation.

Potential of primary kidney cells for somatic cell nuclear transfer mediated transgenesis in pig

BackgroundSomatic cell nuclear transfer (SCNT) is currently the most efficient and precise method to generate genetically tailored pig models for biomedical research. However, the efficiency of this approach is crucially dependent on the source of nuclear donor cells. In this study, we evaluate the potential of primary porcine kidney cells (PKCs) as cell source for SCNT, including their proliferation capacity, transfection efficiency, and capacity to support full term development of SCNT embryos after additive gene transfer or homologous recombination.ResultsPKCs could be maintained in culture with stable karyotype for up to 71 passages, whereas porcine fetal fibroblasts (PFFs) and porcine ear fibroblasts (PEFs) could be hardly passaged more than 20 times. Compared with PFFs and PEFs, PKCs exhibited a higher proliferation rate and resulted in a 2-fold higher blastocyst rate after SCNT and in vitro cultivation. Among the four transfection methods tested with a GFP expression plasmid, best results were obtained with the NucleofectorTM technology, resulting in transfection efficiencies of 70% to 89% with high fluorescence intensity, low cytotoxicity, good cell proliferation, and almost no morphological signs of cell stress. Usage of genetically modified PKCs in SCNT resulted in approximately 150 piglets carrying at least one of 18 different transgenes. Several of those pigs originated from PKCs that underwent homologous recombination and antibiotic selection before SCNT.ConclusionThe high proliferation capacity of PKCs facilitates the introduction of precise and complex genetic modifications in vitro. PKCs are thus a valuable cell source for the generation of porcine biomedical models by SCNT.

236 GENE TARGETING USING ZINC FINGER NUCLEASES (ZFN) IN THE PORCINE FIBROBLAST

Gene targeting is a genetic technique that utilises homologous recombination between an engineered exogenous DNA fragment and the endogenous genome of an animal. In domestic animals, gene targeting has provided an important tool for producing knockout pigs for the α1,3-galactosyltransferase gene (GGTA1) to use in xenotransplantation. The frequency of homologous recombination is a critical parameter for the success of gene targeting. The efficiency of homologous recombination in somatic cells is lower than that in mouse embryonic stem cells. The application of gene targeting to somatic cells has been limited by its low efficiency. Recently, knockout rat and mouse were generated by introducing nonhomologus end joining (NHE)-mediated deletion or insertion at the target site using zinc-finger nucleases (ZFN). Therefore, the development of effective knockout and knock-in techniques in domestic animals is very important in biomedical research. In this study, we investigated homologous recombination events at the cytidine monophospho-N-acetylneuraminic acid hydroxylase (CMAH) gene locus using ZFN in porcine primary fibroblast. The CMAH-targeted ZFN plasmid and mRNA were purchased from Sigma-Aldrich (St Louis, MO, USA). Porcine ear fibroblasts cells were obtained from a 10-day-old male Chicago miniature pigs. The fibroblasts were cultured in DMEM containing 15% fetal bovine serum, 1 × nonessential amino acids, 1 × sodium pyruvate, 10–4 M β-mercaptoethanol, 100 unit mL–1 penicillin and 100 μg mL–1 streptomycin. The cells were trypsinized and resuspended at a concentration of 1.25 × 107 cells mL–1 in F10 nutrient mixture. Four hundred microliters of the cell suspension was electroporated in a 4-mm cuvette with 4 pulses of 1 ms duration using 400V capacitive discharges using the CMAH neo targeting vector and ZFN plasmid or RNA. The CMAH neo targeting vector consists of the neomycin resistance gene (neo) as a positive selectable marker gene, 789-bp 5′ arm and 763-bp 3′ arm from exon 8 of CMAH gene. After selection of G-418, PCR analysis was performed using 64 colonies transfected with ZFN plasmid and 48 colonies transfected with ZFN RNA. As a result, 19 positive colonies were identified in colonies transfected with ZFN plasmid and 15 colonies were identified in colonies transfected with ZFN RNA. The targeting efficiency was 29.7 and 31.6% in the colonies transfected with ZFN plasmid and ZFN RNA, respectively. To our knowledge, this study provides the first evidence that the efficiency of gene targeting using ZFN was higher than that of conventional gene targeting in the porcine fibroblast. These cell lines may be used in production of CMAH knockouts for xenotransplantation.

Gene targeting with zinc finger nucleases to produce cloned eGFP knockout pigs

We report using zinc finger nucleases (ZFNs) and somatic cell nuclear transfer (SCNT) to generate pigs with a knockout (KO) mutation of an enhanced green fluorescent protein (eGFP) transgene. ZFNs are synthetic modular proteins composed of a FokI endonuclease domain linked to a sequence-specific zinc finger DNA-binding domain. Pairs of ZFNs bind to the target region, allowing FokI dimerization and subsequent DNA cleavage. Mutation of an eGFP transgene by using ZFNs has been demonstrated in rats (Geurts et al., 2009); however, there are no reports of ZFN-mutated livestock offspring. Using ZFNs to increase the efficiency of gene modification may advance the production of agriculturally and clinically relevant animal models, particularly in species where modification is difficult. All animal procedures followed an approved IACUC protocol. Adult porcine ear fibroblasts hemizygous for the eGFP transgene (Whitworth et al., 2009), from a grandson of the founder animal, were cultured to 75% confluency, trypsinized, and cotransfected (Ross et al., 2010) with a pair of ZFN plasmids that bind to opposing strands at the eGFP target site (Sigma–Aldrich CompoZr®) and a red fluorescent CAG-tomato plasmid as a transient selectable fluorophore reporting transfection efficiency. Transfected fibroblasts were cultured for 96 h and selected for red fluorescence by automated cell sorting (FACS; Fig. 1A,B) and seeded into 96-well plates (100 cells/well) based on CAG-tomato expression (2% of total cells sorted). A second round of FACS using negative selection of eGFP fluorescence enriched for eGFP KO cells (~5% of sorted cells). PCR of genomic DNA from fibroblasts in wells containing predominantly non-green cells was used to amplify a fragment bracketing the ZFN targeting site. The PCR product was cloned into E. coli, and 16 colonies showed mutations by Sanger sequencing. These mutations at the ZFN cleavage site included a 6-bp deletion, a 333-bp deletion, and a 222-bp deletion replaced with a 113-bp inversion of the deleted sequence (Supplemental Information http://animalsciences.missouri.edu/faculty/prather/). Fibroblast colonies determined to carry ZFN-induced mutations were used as donor cells for SCNT, and embryo transfer (n = 3; Whitworth et al., 2009). One Day-12 embryo collection recovered 7 of 9 embryos that did not fluoresce. Quality sequence was obtained from four embryos, and confirmed that the two embryos that fluoresced had an intact eGFP and two embryos that did not fluoresce were mutated. One of seven piglets recovered via Cesarean section at term for the other two pregnancies fluoresced (Fig. 1C). DNA sequencing confirmed that this piglet had unaltered sequence at the predicted ZFN cut site, while the non-fluorescing piglets had various deletions and insertions. While a direct comparison to a 12% mutation rate in the rat made by microinjection (Geurts et al., 2009) cannot be made because of our FACS preselection and SCNT methods, six out of seven founders being mutated is consistent with their results. This study establishes ZFN-based gene modification in a large animal model. The methods used are anticipated to be useful in selective knockout of endogenous swine genes of agricultural and biomedical importance without introducing any transgenic sequence into the genome. Figure 1 A: Control porcine ear fibroblasts; eGFP and Hoechst DNA stain fluorescence. B: Porcine ear fibroblasts treated with eGFP-specific ZFNs; eGFP and Hoechst DNA stain fluorescence. GFP fluorescence is absent in a high proportion of the treated fibroblasts ...