Identification of host factor responsible for cell line permissiveness to African swine fever virus (ASFV)

by NextGenRnD®
8 min reading time

Endosome, mimivirus, replication factories and african swine fever virus (ASFV)

From a virologist’s point of view, the African swine fever virus (ASFV) has remarkable properties. In particular: (i) it replicates in the cytoplasm of the host cell as the opposite have not been convincingly demonstrated, whereas almost all DNA viruses replicate within the nucleus; (ii) it brings the viral enzymes required for transcription and replication within its core into the host cell like negative-strand RNA viruses or like retroviruses; (iii) it encodes polyproteins and replicates in close association with the endoplasmic reticulum (ER) membranes similarly to positive-strand RNA viruses, e.g., hepatitis C virus (HCV), poliovirus, etc.; (iv) it modifies the ER membranes similarly to HCV to change their curvature and to make them suitable for covering the core shell with an inner lipid membrane; (v) it is an arbovirus that can be transmitted by ticks like tick-borne encephalitis virus or Omsk hemorrhagic fever virus. This frenzy list goes on and on. At the end of the day, however, for swine, ASFV is very efficient and very deadly.

The need for the host factor, which would render a widely (or not so widely) used immortalized cell line permissive for efficient virulent genotype II (like Armenia/07) ASFV infection is not unique. Everyone in the ASFV field, especially biotech and pharma companies, wants this host factor. If such a host factor indeed exists, then it can be identified and below NextGenRnD will specify the experiments required to achieve this goal.

From a cell biologist’s point of view, ASFV is a very simple virus. It infects the host cell and then ends up in late endosome1. Upon release from the late endosome the core shell enclosing the DNA nucleoid enters the cytoplasm.

From a molecular biologist’s point of view, the genome structure and replication strategy of ASFV share many aspects with other large double-stranded DNA (dsDNA) viruses, especially with poxviruses. This similarity, however, is not exclusively enjoyed by poxviruses. Other large dsDNA viruses of Iridoviridae, Phycodnaviridae, and Mimiviridae families share many things in common with ASFV too2. These large dsDNA viruses are called nucleocytoplasmic large DNA viruses or NCLDVs for short.

The interesting thing about NCLDVs is that they are most likely not NCLDVs. Instead they might be purely cytoplasmic large DNA viruses or CLDVs for short. Why? A remarkable study was published in 2010, which demonstrated that the giant Mimivirus replicates in the cytoplasm, essentially like during vaccinia infection3. At the end of their paper, authors rightfully question whether NCLDVs are NCLDVs, or they are actually just CLDVs. But this is not the point why the Mimivirus has just been mentioned.

The Mimivirus study, mentioned above, demonstrated that each viral core delivered to the host cell cytoplasm forms a replication factory3. Another important observation evident from this study is that the “core-released” and actively replicating viral DNA stays associated with the core. Thus, the Mimivirus core serves as the crowding agent for the viral DNA. Due to the fact that the viral core shell is spherical and the newly generated viral DNA is bent around it.

There are several immortalized cells, like WSL cells (or sometimes called WSL-R4, WSL-Bu5, etc.), isolated from the wild boar. The WSL-R cells are remarkable cells, because they can be infected with any ASFV isolate6. The WSL-R, however, differ from the swine monocytes/macrophages as these cells does not develop a clear cytopathic effect, whereas swine macrophages do die in the course of ASFV infection.

One of the problems in ASFV research is very simple: the virulent strains do infect primary swine macrophages (or porcine alveolar macrophages, PAMs), whereas they do not infect the corresponding swine, monkey, or dog immortalized cells. Thus, the WSL-R cell line solves this problem. Both PAMs and WSL-R cells can be infected with the virulent genotype II Armenia/07 ASFV virus. The corresponding infection experiments were recently performed. In particular, when PAMs were infected with Armenia/07 the replication factory-characteristic, diffuse, and intense co-localization of synthesized viral DNA signal (stained with DAPI) and viral late protein (stained with anti-p72 dye-conjugated antibody) was observed7. However, when WSL-R cells were infected with Armenia/07 the punctate p72 signals were present and in a much lower number7, whereas the DNA signal was similar to the one observed in PAM infection. The same trend in the p72 staining pattern was evident for non-virulent attenuated ASFV. The authors concluded that the infection of WSL-R cells is not as productive as that of PAMs due to some block at a particular step of infection.

The p72 is the major protein of ASFV capsid, which encloses the inner membrane surrounding the viral core shell containing the nucleoid DNA. Importantly, the viral core assembly depends on the p72 expression8, and the p72 expression cannot take place if the viral DNA is not released into the cytosol (hence the name viral late protein). Thus, the fact that p72 is stained in a punctate pattern in WSL-R cells upon both virulent Armenia/07 and non-virulent naturally attenuated ASFV (NHV/P68) infection can be explained very easily. In particular, the release of ASFV DNA-containing core from the late endosome of infected WSL-R cells is non-efficient, or defective.

Identification of cellular host factor(s) rendering cells permissive to ASFV

In the previous section, we deduced that the release of ASFV DNA-containing core from the late endosome of infected WSL-R cells is non-efficient, or defective. At the same time, the release of ASFV DNA-containing core from the endosomes of PAMs is perfectly efficient. Thus, there should be some host factor(s) within the late endosome that allow(s) the efficient release of DNA-containing ASFV core from the endosome. Thus, this host factor can be identified through comparing the protein subsets (i.e., proteomes) of endosomes isolated from ASFV-infected PAMs and WSL-R cells.

Specialized fractionation protocols can be easily used to isolate the pure late endosome fraction. The subsets of cell proteins present in late endosomes’ fractions of PAMs and WSL-Rs infected with ASFV should then be analyzed by high-resolution two-dimensional gel electrophoresis (2-DE). Subsequently, the proteins should be stained with Coomassie Blue, the protein “spots” of interest excised, and digested with trypsin. Next, tryptic cleavage fragments should be eluted from polyacrylamide matrix using TFA/H2O and TFA/acetonitrile, separated using RP-HPLC and sequenced. This methodology allows sequencing picomole amounts of protein.

High-level experimental approach

The ASFV, mentioned in previous section, can be either genotype II (Armenia/07) or genotype I (NHV/P68) or any other strain of ASFV. Let us assume that we are using the Armenia/07 strain to infect PAM and WSL-R cells. First, we passage the PAM and WSL-R cells, harvest 5-10x106 cells of each and isolate the corresponding late endosome fractions, which we are going to name PAM-LE and WSL-R-LE. The PAM-LE and WSL-R-LE is our negative control samples. Second, we grow ~2.5-5x106 PAM and WSL-R cells and infect them with Armenia/07, then wait for 48 hours, harvest and isolate the corresponding late endosome fractions (Armenia/07-PAM-LE and Armenia/07-WSL-R-LE respectively). Before the isolation of late endosome fractions we count the cells and make sure we start with the equal quantities, then after we have isolated LEs, we quantify the protein content of the LE-fractions to be sure we load the equal amounts of our samples. Third, we run the 2-DE and stain the resulting gels with Coomassie Blue (or silver-stain). Fourth, we compare PAM-LE vs. Armenia/07-PAM-LE and WSL-R-LE vs. Armenia/07-WSL-R-LE to understand how ASFV infection influences the expression of host cell proteins and which viral protein spots are appearing in Armenia/07-infected late endosome samples. Fifth, we compare PAM-LE vs. WSL-R-LE and Armenia/07-PAM-LE vs. Armenia/07-WSL-R-LE to identify protein spots not present, over-represented, or under-represented. These spots are the candidate host factors responsible for the permissiveness of PAMs to Armenia/07. Sixth, we “micro-sequence” these spots to identify the corresponding proteins. Seventh, we generate the transient expression vectors encoding identified codon-optimized proteins (with intron within their coding DNA sequence) and transfect these into the cell line of interest (e.g., WSL-R, PK15, ST, Vero, MDCK, or EFN-R[see below]). Alternatively, we might want to create (either using CRISPR or any other approach) the cell lines expressing identified proteins either constitutively or when we want them to (inducible expression). Subsequently, we perform the Armenia/07 infection and after 48 hours we analyze the following during multiple passages: (i) cytopathic effect development; (ii) replication factories staining pattern (DAPI & anti-p72); (iii) western blot of p72; (iv) virus yield analysis; (v) virus genome integrity and absence of deletions or recombination events; and (vi) virulence status. This way, we will be able to validate that one or several of the identified host factors is the one responsible for Armenia/07 infection permissiveness.

The spectrometric typing, the origin of WSL-R and EFN-R cells

In 2010, there was an interesting publication from the Friedrich-Loeffler-Institut (Germany), which has a Collection of Cell Lines in Veterinary Medicine (CCLV), which contains WSL-R cell line. In particular, the Karger et al. figured out that they could phenotype the cell lines by simply “flying” their proteomes using MALDI-TOF4. This is essential, when one would like to rapidly establish if there is a potential contamination of the cell culture. The “interesting” aspect of the publication was the phylogenetic tree they reconstructed after flying the proteomes of all the cell lines they had. In particular, the clustering was mainly driven by taxonomy and allowed the determination of unknown species. In other words, the swine cell lines were within one taxa. Interestingly, the WSL-R and one other cell line, EFN-R, were within one clade, whereas swine PK 15 and ST cell lines belonged to other clades4. Thus, EFN-R is an additional porcine cell line (however, not immortalized, at least originally) that might be used as a host for ASFV genotype II infection. Both WSL-R and EFN-R cell lines are available from Friedrich-Loeffler-Institut (Germany).

The EFN-R cell line can be relevant for ASFV infection as the assembly and envelopment of herpesvirus-like vesicles was studied in detail in this cell line9. This is relevant because ~100-nm viral vesicle formation at the nuclear membrane, its envelopment and egress into cytosol was observed. The two last processes are relevant for ASFV as well. Thus, EFN-R cell line can be used in concert with PAMs and WSL-R cells, as specified in the previous section.

Contact NextGenRnD for details.

References