Co-Cultivation Assays for Detecting Infectious Human-Tropic Porcine Endogenous Retrovirus (PERV)

1. Introduction

Porcine endogenous retroviruses (PERVs) are gammaretroviruses integrated in the genome of pigs, PERV-A and PERV-B are found in all pigs, PERV-C in many, but not all pigs [1]. PERV-A and PERV-B have been found to infect human cells in culture (human-tropic viruses), especially tumor and immortal cells [2, 3]. There are only a few reports on infection of primary cells [4-6]. PERV-C does not infect human cells, but only pig cells (ecotropic virus). However, recombinations between PERV-A and PERV-C have been detected in living pigs. In this case PERV-C acquired the receptor binding site for human PERV receptors and consequently, can infect human cells. PERV-A/C replicates at higher titers on human cells compared with the parental PERV-A [7-10]. Since these viruses are integrated into the genome, they cannot be eliminated as other pig viruses can. Therefore, they pose a special risk for xenotransplantation using pig cells, tissues, or organs. To detect PERVs and to analyze their expression, several methods have been developed [11-13]. However, the main question, which pigs release infectious human-tropic viruses able to infect human cells is difficult to answer. Neither the number of integrated proviruses (some of them are defect), nor the expression at the level of RNA or protein determine the release of virus particles [14]. This question can only be answered through infection assays, which are challenging to establish and, as will be shown below, have extremely limited value. Here, three commonly used assays and their limitations will be analyzed.

2. Co-cultivation with Pig Cells After Gamma-Irradiation

The first type of assay involves incubating human cells with gamma-irradiated pig cells (Figure 1A). Gamma or X irradiation, i.e., exposure to ionizing radiation with gamma rays, inhibits the proliferation of pig cells [15-17] and allows the isolation of living human cells after co-culture. The effects of gamma irradiation on cell viability depend on the time of exposure and culture conditions. Unfortunately, it may lead to an incomplete elimination of porcine cells and consequently to false-positive infection results. In addition, the gamma irradiation may have a negative effect on virus release (see next chapter).

This method was used when gamma irradiated (100 Gy) porcine kidney PK15 cells, which produce PERV, were incubated with several human cell lines and it was shown for the first time that PERV are able to infect human cells [2]. In addition to the human embryonic kidney 293 cells, MRC-5 fetal diploid fibroblasts, the rhabdomyosarcoma cell line RD, a B-cell line and two T-lymphocytic cell lines became productively infected. This method was also used when for the first time a human-tropic PERV released from pig lymphocytes irradiated with 2,000 rads from a 137Cs source was shown to infected human cells [7]. This virus was later identified as a recombinant PERV-A/C [8]. Furthermore, Garkavenko et al. [18] co-cultivated gamma irradiated (2000 rads from a 137Cs source) pig peripheral blood mononuclear cells (PBMCs) or PK15 cells with human 293 cells. The PBMCs were from a designated pathogen-free (DPF) herd of New Zealand pigs, Auckland Island pigs, among them were also PERV-C-positive pigs. No transmission of human-tropic PERV was observed in contrast to PK15 cells.

3. Effects of Gamma Irradiation on Donor Pig Cells

Ionizing radiation induces cellular damage as indicated by DNA damage, and reduction of DNA damage repair and cell survival. Other indicators of cellular damage may also include cell cycle alterations, glycolysis effects and cell apoptosis [15]. The DNA damage induced by irradiation could result in direct and indirect biological effects associated with reactive oxygen species (ROS). Increased ROS, DNA double strand breaks (DSB), cellular apoptosis, and inhibition of cell proliferation were observed [15, 16]. Gamma radiation at a dose as low as 5 Gy can cause DNA strand breaks and chromosome aberrations. Irradiated PBMCs retain cellular functions, such as cytokine secretion, cytotoxicity, and co-stimulation, but are unable to proliferate [17]. Irradiated PBMCs and pig embryonic kidney cells also retain the ability to release PERV [7, 18]. Due to the negative effect on the DNA and subsequently on gene and virus expression as well as the potential impossibility to eliminate all pig cells, this scenario is significantly different from the situation in xenotransplantation.

4. Separation of Pig and Human Cells by Porous Membranes

The second assay to study PERV release and infection of human cell is based on co-cultivation of pig and human cells using a double chamber system (Figure 1B). The human cells are located in the lower chamber, the pig cells in the upper one, separated by a porous membrane. This membrane with a pore size of 0.4 µm allows exchange of medium, small molecules and viruses. In this scenario all cells are alive during the whole co-cultivation. However, this set up does not take into account the cell-to cell transmission of viruses, which is a common mode of retroviral infection (see next chapter).

5. Cell-To-Cell Transmission of Retroviruses

Retroviruses can infect target cells either by cell-free transmission or by direct cell-to-cell spread. Using cell-to-cell spread retroviruses may circumvent humoral immunity in vivo. The cell-to-cell spread of the human immunodeficiency virus type 1 (HIV-1) and the human T-cell lymphotropic virus type 1 (HTLV-1) are well studied (for review see [19]). Cell-to-cell spread takes place at specialized contact-induced virological synapses between infected and uninfected cells. In the case of HIV-1, the large extracellular loop of CD63 and the transmembrane envelope protein gp41 of HIV-1 were essential for the establishment of viral synapses [20]. Gamma-retroviruses are not well studied in this context. The murine leukemia virus (MuLV), a close relative to PERV, utilizes virus-induced filopodia for efficient dissemination [21]. These filopodia originate from non-infected cells and interact, through their tips, with infected cells. The viral envelope glycoprotein (Env) on an infected cell interacts with the receptor molecules on a target cell and generates a stable bridge. Viruses then move along the outer surface of the filopodial bridge toward the target cell. Virus spreading was shown to be sensitive to reagents affecting the actin/myosin machinery [21]. The assembly of MuLV is directed towards sites of cell-cell contact [22]. There are two mechanisms, surface-based transmission and de novo assembly at the site of contact [23]. B cells infected with Friend MuLV (F-MuLV) have been shown to form virological synapses within the lymph node of living mice as shown by intravital staining [24, 25]. In vivo virological synapses showed one striking difference when compared with synapses in vitro: whereas in vitro synapsis formation for F-MuLV is entirely driven by Env–receptor interactions, in vivo F-MuLV-infected primary B cells selectively interact with CD4⁺ T cells and less with CD8⁺ T cells, despite the fact that both cell types express similar mCAT-1 receptor levels and are equally susceptible to cell-free F-MuLV. Therefore, other factors in addition to the interaction of Env and receptor must contribute to the selectivity of F-MuLV spreading in vivo [25]. To summarize, cell-to-cell spread is an important mode of retrovirus transmission. PERV is likely to utilize cell-to-cell transmission to infect target cells; however, the underlying mechanism remains to be determined.

6. Use of Selection-Marker Resistant Human Cells

The third assay is based on co-culture of pig cells with human cells which are transfected with a selection marker (Figure 1C). This method is the easiest to handle and yields the purest population of human cells. The human cells are made resistant against a selection medium. Example of such selection markers are the neomycin phosphotransferase (neo) gene, which confers resistance to geneticin or G418, a toxic aminoglycoside, in eukaryotic cells [26]. Another example is hygromycin B, an aminocyclitol antibiotic produced by Streptomyces. Hygromycin B phosphotransferase isolated from Escherichia coli has been shown to permit direct selection for hygromycin B resistance following transfection of eukaryotic cell lines [26]. To study PERV infection of human cells, this method was first used by us in 2008 [27]. Human 293 cells were made neomycin-resistant (293neo+) and after co-cultivation with pig cells, selection medium was applied to eliminate pig cells. In these investigations, it was shown that islet cells from German landrace pigs did not produce infectious human-tropic PERV [27]. Screening for PERV infection was done using a PERV-specific PCR. To confirm that all pig cells were eliminated after cultivation with selection medium, a PCR detecting pig COII was performed. Later Kono et al. [28] used hygromycin-resistant 293 cells. After co-cultivation with PK15 cells, PERV proviruses were found in hygromycin-resistant 293 cells by PCR and virus particles were found by electron microscopy. The absence of the porcine alpha-1,3-galactosyltranferase gene indicated that all PK15 cells were removed. Following these studies, the authors performed an infectivity assessment of PERV studying the expression of PERV using high-throughput sequencing. Since not all integrated proviruses produce infectious human-tropic viruses due to some being defective, understanding the expression of full-length proviruses is crucial [28]. The ability of PERVs to infect human cells varies among pig species due to differences in the number of infectious proviruses [14, 29]. To summarize: in addition to its advantages over the other two methods—specifically, (i) avoiding treatments like gamma irradiation, which could impact PERV expression and replication in donor cells, (ii) enabling direct contact between human and pig cells, (iii) allowing both, detection of both free virus and cell-to-cell transmission, and (iv) allowing for complete removal of pig cells—something not fully achieved in the first assay —, this method also offers the added benefits of (v) not requiring expensive equipment and being less time-consuming.

7. What do the Assays Tell Us?

Regardless of the assay performed, the selection of donor pig cells and recipient human cells is crucial to the experimental outcome. The results remain inherently limited, and depend on the type of pig cells and human cells. A cell line like PK15 releases infectious human-tropic PERV, whereas normal pig lymphocytes do not [30]. However, after stimulation with a mitogen pig PBMCs may release infectious virus [31-33]. This depends on the pig breed; minipigs more often release infectious human-tropic virus, usually PERV-A/C [34]. Other pig primary cells also release PERV infectious for human cells [6], but PBMCs stimulated with a T cell mitogen and 12-O-tetradecanoyl-phorbol-13-acetate (TPA) were the best producers [30-35]. Concerning the target cells, human 293 cells are highly susceptible to PERV infection, likely due to the loss of intracellular restriction factors [36]. In contrast, primary human cells were not successfully infected with PERV released from pig cells, yet. Only a human-cell-adapted PERV, which had a high replication rate due to an increased number of transcription factor binding sites in its long terminal repeats (LTR) [10], was able to achieve infection primary human PBMCs [4].

Transplanting a vascularized pig organ involves the transfer of various cell types, including stem cells. Stem cells are known for their high expression of endogenous retroviruses, which decreases as they differentiate [37]. The transplant recipient also consists of diverse cell types, and it remains uncertain whether certain primary human cells exhibit susceptibility comparable to 293 cells. Of particular interest is evaluating whether human stem cells can be infected by PERV.

Therefore, these assays have limited utility in assessing the risk posed by PERV in transplants derived from specific pig breeds. The limitations of in vivo models—such as the transplantation of pig organs into non-human primates—are well documented, particularly due to the absence of fully functional PERV receptors in these models [38, 39]. Moreover, combining in vivo and in vitro assays does not enhance predictive accuracy. As a result, there is currently no definitive or reliable experimental method to assess the risk of PERV transmission; only long-term monitoring of actual xenotransplant recipients can ultimately provide conclusive answers.

8. Conclusions

The assay based on co-culture of pig cells with human cells which are transfected with a selection marker best simulates the conditions of in vivo xenotransplantation. This method is the easiest to handle and yields the purest population of human cells. However, like the other assays it only indicates whether a specific pig cell type releases PERVs and whether a specific human cell type is susceptible to infection. A negative infection result does not necessarily reflect the in vivo situation, where a transplanted organ consists of multiple pig cell types interacting with a diverse range of human cells within a living organism. The capability of this assay as well as the other assays to assess the risk posed by PERV in transplants derived from specific pig breeds is limited.