ESCDL-1, a new cell line derived from chicken embryonic stem cells, supports efficient replication of Mardiviruses

Abstract

Marek's disease virus is the etiological agent of a major lymphoproliferative disorder in poultry and the prototype of the Mardivirus genus. Primary avian somatic cells are currently used for virus replication and vaccine production, but they are largely refractory to any genetic modification compatible with the preservation of intact viral susceptibility. We explored the concept of induction of viral replication permissiveness in an established pluripotent chicken embryonic stem cell-line (cES) in order to derive a new fully susceptible cell-line. Chicken ES cells were not permissive for Mardivirus infection, but as soon as differentiation was triggered, replication of Marek's disease virus was detected. From a panel of cyto-differentiating agents, hexamethylene bis (acetamide) (HMBA) was found to be the most efficient regarding the induction of permissiveness. These initial findings prompted us to analyse the effect of HMBA on gene expression, to derive a new mesenchymal cell line, the so-called ESCDL-1, and monitor its susceptibility for Mardivirus replication. All Mardiviruses tested so far replicated equally well on primary embryonic skin cells and on ESCDL-1, and the latter showed no variation related to its passage number in its permissiveness for virus infection. Viral morphogenesis studies confirmed efficient multiplication with, as in other in vitro models, no extra-cellular virus production. We could show that ESCDL-1 can be transfected to express a transgene and subsequently cloned without any loss in permissiveness. Consequently, ESCDL-1 was genetically modified to complement viral gene deletions thus yielding stable trans-complementing cell lines. We herein claim that derivation of stable differentiated cell-lines from cES cell lines might be an alternative solution to the cultivation of primary cells for virology studies.

Fig 1. Differentiation increases permissiveness of cES to GaHV-2.

(A & B): cES cells were plated and infected 6 days post plating in reduced serum conditions with sorted CESC infected with vBAC20GFPVP22. Cells were fixed after an incubation of 6 days at 37°C. (A) cell monolayers were maintained in WE medium containing 1% FBS and 1.5% CS. (B) DMSO (64 μM) was added from day 4 after plating and until the end of the culture (scale bar represents 200 μm). (C) cES cells (passage 36) were exposed to HMBA and infected with sorted vBAC20GFPVP22-infected cells. Expression of VP22 was detected by the GFP signal and ICP4 by staining with monoclonal antibody E21 (red); cell nuclei were stained by Hoechst 33342. At late stages of infection, ICP4 is detected both in the cytoplasm and nucleus in VP22 expressing cells. At early stages of infection, when VP22 is barely detectable in the cells surrounding the highly infected cell, ICP4 staining is predominantly nuclear (arrow heads) indicating spread of virus from the originally infected cell to the neighbouring cells (scale bar represents 20 μm). (D) Induction of differentiation by HMBA increases susceptibility of cES cells to GaHV-2 infection. Comparison of the plaque counts at 4 days pi either on cES cells or on primary CESC exposed to differentiating drugs (2 independent experiments sampling 10 replicates for each condition with cES and 4 replicates with CESC). (E) Comparison of plaque sizes on either cES exposed to HMBA differentiation or CESC. For both cell types, HMBA was added in the maintenance medium after the infection with sorted vBAC20EGFPVP22-infected cells. Plaques appeared larger in cES differentiated cells. Plaque sizes from 80 plaques per experiment are shown as boxplots and whiskers (Tukey) (in B, P<0.001; Mann Whitney test).

Fig 2. The HMBA treatment modifies the transcriptomic profile of both CESCs and cES cells.

(A) The Venn diagram illustrates the genes that are differentially expressed between CESC and cES cells treated with or without HMBA and uninfected or infected with vBAC20UL49GFP inoculums. (B) A PCA analysis was performed on the common genes identified from this study and those provided by the datasets GSE61221 for the CEF-1, GSE38168 for the CEF-2 and GSE47191 for the DF1 cells.

Fig 3. Characterization of ESCDL-1.

(A) ESCDL-1 cells (passage 52) display a mesenchyme cell morphology with numerous actin stress fibres (Alexa Fluor^® 594 phalloidin). Nuclei appear in blue due to Hoechst 33342 staining, mitochondria in green by staining with monoclonal antibody 4C7 and Alexa Fluor^® 488 anti-mouse IgG. (B) ESCDL-1 express vimentin as a major intermediate filament protein. Proteins were extracted from chicken keratinocyte line K8 (KcES) and ESCDL-1 and western blots were probed with anti-actin JLA-20, anti-vimentin AMF17b, or anti-cytokeratin type I or II antibodies. The apparent molecular masses of actin (45 kDa) and vimentin (55 kDa) are similar to those described in the publications describing the monoclonal antibodies. ESCDL-1 cells do not express type I or II cytokeratins, which are detected in KcES extracts. Molecular weight markers (PageRuler^™ Plus prestained protein ladder—Thermo Scientific) are on the left side of each blot.

Fig 4. Replication of vBAC20UL17mRFP on ESCDL-1.

(A) vBAC20UL17mRFP disseminates as efficiently as its parental virus vBAC20 on ESCDL-1. Viral plaques were counterstained with a mixture of monoclonal antibodies as described and areas were determined from a total of 180 plaques per virus in 2 experiments (Tukey box and whisker plots– Mann Whitney test—p = 0.015). (B) Kinetics of viral replication on ESCDL-1 at P41. ESCDL-1 cells were infected at an m.o.i of 0.02 and cell-associated virus was quantified from day 1 to day 4 on ESCDL-1 at intermediate or high passages, or on primary CESC (error bars represent 2 standard deviations). (C) Detection of viral proteins in cell extracts. Infected (lanes 1 & 2) or non-infected (lane 3) cell extracts were subjected to PAGE and blotting. Total cell-extracts were loaded in lane 1 and 3, while NP-40 insoluble cell extract was loaded in lane 2. Vimentin detection illustrates the relative amount of protein loaded. Protein pUL17-mRFP was detected by using a monoclonal anti-RFP; the upper part of the gel shows the detection of the fusion protein at the predicted apparent molecular mass (~110 kDa) while the lower part of the gel shows the presence of mRFP possibly cleaved from the fusion protein. (D) Comparison of vBAC20UL17mRFP production at passages 9 and 10 on low (P37-P47, white bars), intermediate (P57-P69, vertical bars) and high (P97-P104, black bars). The limited decline of virus production on intermediate / high passage cells at P9 was not confirmed at P10 (error bars represent 2 standard deviations).

Fig 5. TEM analysis of vBAC20 morphogenesis in ESCDL-1.

(A) Overview of an infected cell with intranuclear A (black triangle) and B (black arrowheads) capsids and intracytoplasmic C capsids (white ellipses). The white arrow points to an image of capsid tegumentation in the cytoplasm. (B) Intranuclear accumulation of small particles (SP– 30 to 35 nm in diameter) arranged in a pseudo-crystalline structure in the vicinity of A capsids (black triangle). (C) Accumulation of primary enveloped virions in distended cisternae of the nuclear membrane (black arrows point to enveloped C capsids). (D) C capsid undergoing secondary envelopment: electron dense material, possibly of tegument origin, surrounding the capsid is surrounded by a membrane in close vicinity to the Golgi (bar represents 0.2 μm). (E) Multiple intracytoplasmic C capsids (white ellipses) close to an enveloped cytoplasmic particle (white arrowhead).

Fig 6. ESCDL-1 stably expressing the YGFP Venus support GaHV-2 replication.

(A) Venus expression by ESCDL-1-Venus at passage 8 post selection initiation (scale bar indicates 100 μm). Venus YGFP is equally distributed in the nucleus and cytoplasm. (B) Comparison of plaque areas of vBAC20UL17mRFP between the parental cell line and the ESCDL-1-Venus/Clone7. The plaque areas of virus produced on ESCDL-1-Venus/Clone7 (green Tuckey box and whiskers plots) and on ESCDL-1 (shaded Tuckey box and whiskers plot) were measured for the parental and the Venus expressing cell-lines at matching passages (2 independent experiments). Differences existing between the cells or viruses were non significant (Mann-Whitney test with P values over 0.5). (C) vBACRB1BUL17mRFP plaques on ESCDL-1-Venus/Clone7: note the strong up-regulation of Venus expression in infected cells (scale bar indicates 200 μm).

Fig 7. Constitutive expression of pUL49 (VP22) in ESCDL-1 complements the deletion of UL49 in GaHV2 genome.

(A & B) Constitutive expression of pUL49 (VP22) in uncloned cell populations (A) and in ESCDL-1-UL49/clone 2 (B). VP22 staining by anti pUL49 Mabs and an Alexa Fluor^® 488 goat anti-mouse detects filamentous material between 2 strongly positive nuclei that appear to be still bound after cell division (white arrowhead). (C to H) Complementation of replication for BAC20ΔUL49 on ESCDL-1-UL49/clone 7 (C, E, G) and absence of viral dissemination in non-complementing ESCDL-1 (D, F, H). Viral replication was detected using a chicken hyper immune serum revealed by an Alexa Fluor^® 488 goat anti-chicken conjugate together with an anti-ICP4 Mab (C, D), a mixture of anti-gI and -gE Mabs (E, F), or a mixture of anti-pUL49 (VP22) Mabs (G, H) all revealed by an Alexa Fluor^® 594 goat anti-mouse conjugate. The restoration of pUL49 (VP22) expression is associated with the viral replication (G). Early-late (ICP4) and late (gE-gI) antigens are detected in isolated ESCDL-1 cells (D & F) and in panel H the arrow points to an isolated cell in which vBAC20ΔUL49 undergoes an aborted replication cycle as revealed by the polyclonal anti-MDV serum without detection of VP22. Scale bar represents 50μm.

Fig 8. Expression of pUL37 in ESCDL-1 complements the deletion of UL37 ORF in BACRB-1B.

Upper panel (1–4): Transfection of BACRB-1BΔ37 yields viral plaques on complementing cells: viral plaques were detected in ESCDL-1-UL37 by staining with Mabs B17 (anti-VP22), K11 (anti-gB) and E21 (anti-ICP4) and an Alexa Fluor^® 488 GAM conjugate (1). In ESCDL-1, BACRB-1BΔ37 did not yield a virus that could disseminate and only isolated positive cells could be seen (2). As a control, BACRB-1BUL17mRFP was transfected in either complementing or non-complementing parental cells, producing viral plaques on both (3 & 4). Scale bar = 200 μm. Middle Panel (5 to 8): vBACRB-1BΔ37 can be serially passaged in complementing cells and virus multiplication induces pUL37 expression in the ESCDL-1-UL37: BACRB-1BΔ37 (5,6) or BAC RB-1B (7,8) were transfected either in ESCDL-1-UL37 complementing cells or in ESCDL-1 and passaged once in the same cells. The development of viral infection by passage 2 of the vBACRB-1BΔ37 virus is seen in complementing cells (green fluorescence in 5) and coincides with the expression of pUL37 in infected cells (red fluorescence in 5); in non-complementing cells the same virus passage does not replicate (6). The parental virus (vBACRB-1B) transfected and passaged in the same conditions replicated equally well on ESCDL-1 and on ESCDL-1-UL37 (7 & 8). Scale bar = 50μm. Lower Panel (9 to 12): vBACRB-1BΔ37 may be passaged at least 3 times in complementing cells and does not revert to a replicating virus when plated on non-complementing cells. The 3rd passage of vBACRB-1BΔ37 yielded typical viral plaques in complementing cells (9) whereas the same virus did not form plaques in ESCDL-1 (10). Again vRB-1B at the same passage developed equally well in both cells (staining as in the upper panel, except for HOECHST 33342). Scale bar = 200 μm.

Fig 9. vBACRB-1BΔ37 and vBACRB-1B dissemination in complementing and parental ESCDL-1.

(A) Comparison of dissemination of vBACRB-1B on ESCDL-1 or ESCDL-1-UL37: plaque areas were significantly larger on ESCDL-1-UL37 (Tukey box and whiskers plot—P< 0.0001, Mann Whitney test). (B) Comparison of plaque areas of parental and Δ37 viruses at their 3^rd passage (on complementing cells only for the Δ37 virus). Both viruses were plated on ESCDL-1-UL37 and plaque areas were measured at day 4 after plating. Statistically significant differences are indicated with an asterisk (Tukey box and whiskers plot—P = 0.0004, Mann Whitney test).