Comparison of ZMAC and MARC-145 Cell Lines for Improving Porcine Reproductive and Respiratory Syndrome Virus Isolation from Clinical Samples

The MARC-145 cell line is commonly used to isolate porcine reproductive and respiratory syndrome virus (PRRSV) for diagnostics, research, and vaccine production, but it yields frustratingly low success rates of virus isolation (VI). The ZMAC cell line, derived from porcine alveolar macrophages, has become available, but its utilization for PRRSV VI from clinical samples has not been evaluated.

ABSTRACT

The MARC-145 cell line is commonly used to isolate porcine reproductive and respiratory syndrome virus (PRRSV) for diagnostics, research, and vaccine production, but it yields frustratingly low success rates of virus isolation (VI). The ZMAC cell line, derived from porcine alveolar macrophages, has become available, but its utilization for PRRSV VI from clinical samples has not been evaluated. This study compared PRRSV VI results in ZMAC and MARC-145 cells from 375 clinical samples (including 104 lung, 140 serum, 90 oral fluid, and 41 processing fluid samples). The PRRSV VI success rate was very low in oral fluids and processing fluids regardless of whether ZMAC cells or MARC-145 cells were used. Success rates of PRRSV VI from lung and serum samples were significantly higher in ZMAC than in MARC-145 cells. Lung and serum samples with threshold cycle (C_T) values of <30 had better VI success. PRRSV-2 in genetic lineages 1 and 8 was isolated more successfully in ZMAC cells than in MARC-145 cells, whereas PRRSV-2 in genetic lineage 5 was isolated in the two cell lines with similar success rates. For samples with positive VI in both ZMAC and MARC-145 cells, 14 of 23 PRRSV-2 isolates had similar titers in the two cell lines. A total of 51 of 95 (53.7%) ZMAC-obtained PRRSV-2 or PRRSV-1 isolates grew in MARC-145 cells, and all 46 (100%) MARC-145-obtained isolates grew in ZMAC cells. In summary, ZMAC cells allow better isolation of a wide range of PRRSV field strains; however, not all of the ZMAC-obtained PRRSV isolates grow in MARC-145 cells. This report provides important guidelines to improve isolation of PRRSV from clinical samples for further characterization and/or for producing autogenous vaccines.

INTRODUCTION

Porcine reproductive and respiratory syndrome (PRRS) is a swine disease characterized by reproductive failure and late-term abortion in sows, the birth of weak and stillborn pigs, and respiratory problems in piglets and growing pigs (1). PRRS causes tremendous economic loss to the swine industry worldwide. In a 2013 analysis, the economic impact of PRRS on the U.S. swine industry was estimated at $664 million annually, or $1.8 million a day (2). The causative agent, porcine reproductive and respiratory syndrome virus (PRRSV), is an enveloped and single-stranded positive-sense RNA virus. PRRSV genome RNA is about 15 kb in length and includes a 5′ untranslated region (UTR), 11 known open reading frames (ORFs), and a 3′ UTR (3, 4). PRRSV used to be classified into two distinct genotypes, including PRRSV-1 (type 1; European strain) and PRRSV-2 (type 2; North American strain). However, PRRSV-1 and PRRSV-2 have recently been regarded as two separate species (5). According to the International Committee on Taxonomy of Viruses release in 2018, PRRSV-1 (species name Betaarterivirus suid 1) and PRRSV-2 (species name Betaarterivirus suid 2) belong to the subgenera Eurpobartevirus and Ampobartevirus, respectively, which are classified in the genus Betaarterivirus in the subfamily Variarterivirinae of the family Arteriviridae in the suborder Arnidovirineae within the order Nidovirales.

Open reading frame 5 (ORF5), encoding the major viral envelope protein GP5, is most variable and is commonly used for sequencing to study the molecular epidemiology and/or genetic relatedness of PRRSV (4, 6). Between PRRSV-1 and PRRSV-2, ORF5 nucleotide sequences differ by ∼44%; within PRRSV-1, ORF5 nucleotide sequences differ by ∼30%; and within PRRSV-2, ORF5 nucleotide sequences differ by approximately 21% (1). Phylogenetic analyses of ORF5 sequences resulted in PRRSV-1 classification into four subtypes (7, 8) whereas PRRSV-2 was classified into 9 lineages and 37 sublineages (9), reflecting high genetic diversity in each genotype.

Various assays can be used for PRRSV detection and diagnosis, e.g., PCR, sequencing, virus isolation (VI), gross and microscopic lesion examination, immunohistochemistry, antibody testing, and so on (1, 10). PRRSV VI may have lower sensitivity than PCR and antibody assays (11), but PRRSV VI is still frequently needed for several applications. For example, (i) PRRSV VI is conducted to determine presence of infectious virus in samples. (ii) The commercial PRRSV vaccines formulated on the basis of a single virus strain do not always provide effective protection to swineherds due to high genetic and antigenic diversity among PRRSV field strains. In such cases, swine veterinarians often request isolation of PRRSV from clinical samples for producing farm-specific autogenous vaccines. (iii) PRRSV VI is needed to obtain isolates of new PRRSV strains for potential commercial vaccine development and/or update. (iv) PRRSV VI is required to obtain virus isolates for further characterizations of some PRRSV strains, including pathogenesis study, virus-host cell interaction study, etc. (v) Development and validation of some diagnostic assays (e.g., virus strain-specific indirect fluorescent antibody assay and virus strain-specific virus neutralization assay for antibody testing) rely on PRRSV cell culture isolates.

PRRSV shows highly restricted cell tropism in specifically differentiated and activated monocyte/macrophage lineages (12). Porcine alveolar macrophages (PAMs) derived from swine lungs were first applied to isolate PRRSV-1 prototype strain Lelystad in the Netherlands in 1991 (13). The use of PAMs to isolate PRRSV-2 strains was also previously reported (14) although PRRSV-2 prototype strain VR-2332 was first isolated in continuous cell line CL2621 in the United States in 1992 (15, 16). Nevertheless, CL2621 was a proprietary cell line of Boehringer Ingelheim Animal Health Inc. (16) and it was not widely available to most laboratories (17). Another continuous cell line, MARC-145, a subclone derived from monkey kidney cell line MA-104, was found to be permissive to PRRSV replication in 1993 (17). The St-Jude porcine lung (SJPL) cell line, which was in fact found to be of monkey instead of porcine origin based on karyotype and genetic analyses (18), was reported to be permissive to PRRSV infection (19). However, cytopathic effects were found to develop more slowly in SJPL cells than in MARC-145 cells and the efficiency of isolation of PRRSV from clinical samples in SJPL cells was not as high as in MARC-145 cells (19). Various groups have also reported that genetically modified cell lines expressing CD163 proteins are permissive to PRRSV infection (20, 21), but these cell lines have not been widely used for PRRSV isolation from clinical samples. Nowadays, MARC-145 is still the most commonly used cell line for PRRSV isolation, propagation, and titration in most laboratories. However, not all PRRSV strains grow in simian cell lines and some studies have indicated that PAMs are superior to simian cell lines for primary isolation of PRRSV field strains (22, 23). Nonetheless, virus isolation and propagation in PAMs face some disadvantages, including the short life span of primary cells that need to be prepared periodically, quality variation between batches of primary cells, heterogeneity of cell population, and high risk of contamination from other pathogens. The procedures of PAM preparation are laborious and technically difficult. In recent years, a ZMAC cell line derived from the lung lavage fluid samples from porcine fetuses was developed and ZMAC cells uniformly express typical macrophage markers (24). ZMAC cells were applied to propagate the modified live Prime Pac PRRSV strain with effective protection equivalent to that propagated in MARC-145 cells (24). However, the ZMAC cell line has not been thoroughly evaluated for primary PRRSV isolation from various clinical samples. In addition to being affected by the choice of cells used for VI, the success of isolation of PRRSV could also be affected by virus concentration in the samples as reflected by PRRSV PCR threshold cycle (C_T) values, by specimen types, and by the characteristics of the PRRSV strains as reflected by different phylogenetic lineages. The impacts of these factors on PRRSV VI have not yet been investigated in detail.

This study aimed to compare the rates of success of PRRSV VI from clinical samples in ZMAC and MARC-145 cells. The correlation of PRRSV concentration, genetic lineage, and specimen types to VI success was analyzed. The infectious titers of virus isolates obtained in ZMAC and MARC-145 cells were compared. In addition, we investigated whether PRRSV isolates obtained in ZMAC cells would grow in MARC-145 cells and vice versa.

MATERIALS AND METHODS

Clinical samples.

At the Iowa State University Veterinary Diagnostic Laboratory (ISU VDL), when PRRSV PCR is requested, clinical samples are tested using a commercial PRRSV reverse transcription real-time PCR (RT-rtPCR), i.e., VetMAX PRRSV NA&EU reagent (Thermo Fisher Scientific), on the day that samples are received. Subsequently, the remaining samples are stored at −80°C in freezers on the same day. Threshold cycle (C_T) values of <37 were considered to represent positive results, and C_T values of ≥37 were considered negative. In this study, 375 clinical samples that were representative of different specimen types and C_T ranges were randomly selected from diagnostic cases submitted to the ISU VDL during 2017 to 2020 and used for virus isolation attempts in ZMAC and MARC-145 cells. Within 1 to 2 weeks after PRRSV PCR testing, each batch of samples were retrieved from −80°C freezers for PRRSV VI attempts in two cell lines on the same day. These included 305 PRRSV-2 PCR-positive clinical samples with various C_T ranges (109 serum samples with C_T values of 13.2 to 36.2, 96 lung samples with C_T values of 13 to 30.2, 59 oral fluid samples with C_T values of 22.0 to 34.0, and 41 processing fluid samples with C_T values of 15.6 to 30.9) and 70 PRRSV-1 PCR-positive clinical samples with various C_T ranges (31 serum samples with C_T values of 18.4 to 36.8, 8 lung samples with C_T values of 21.6 to 35.4, and 31 oral fluid samples with C_T values of 28.0 to 36.8).

Cells.

The MARC-145 cell line is a clone of the African monkey kidney MA-104 cell line (17). MARC-145 cells were cultured in regular RPMI 1640 medium supplemented with (final concentrations) 10% fetal bovine serum, 2 mM l-glutamine, 0.05 mg/ml gentamicin, 100 unit/ml penicillin, 100 µg/ml streptomycin, and 0.25 µg/ml amphotericin. The ZMAC cell line (ATCC PTA-8764) was derived from the lung lavage fluid samples from porcine fetuses (24) and was provided by Federico Zuckermann, University of Illinois, Urbana, IL, to Aptimmune Biologics, from which we obtained this cell line. ZMAC cells were cultured in suspension in the RPMI 1640 medium with l-glutamine and 25 mM HEPES (Corning, Oneonta, NY) supplemented with 1× MEM nonessential amino acids (Corning), 4 mM sodium pyruvate (Corning), 2 mM l-glutamine (Corning), 0.81% glucose (Corning), 10% fetal bovine serum (Sigma-Aldrich, St. Louis, MO), 0.01 µg/ml mouse macrophage colony-stimulating factor (mouse M-CSF; Shenandoah Biotechnology, Inc., Warwick, PA), 0.05 mg/ml gentamicin, 100 unit/ml penicillin, 100 µg/ml streptomycin, and 0.25 µg/ml amphotericin (all antibiotics from Invitrogen/Thermo Fisher Scientific, Carlsbad, CA). MARC-145 and ZMAC cells were maintained in an incubator at 37°C with 5% CO₂. In order to minimize potential variations of cells, MARC-145 cells from passage 41 (P41) to P55 and ZMAC cells from P3 to P12 were used in this study.

Virus isolation.

For lung samples, 10% lung homogenates were first prepared. Subsequently serum, lung homogenate, oral fluid, and processing fluid samples were filtered through 0.45-µm-pore-size membranes before inoculation into cells. Virus isolation was conducted in 24-well plates containing MARC-145 or ZMAC cells. For each batch of samples, inoculation into the two cell lines was conducted on the same day to avoid additional freeze-thaw of samples. For inoculation, 200 µl of samples was inoculated per well. After incubation for 1 h in a 37°C incubator with 5% CO₂, the inoculum was removed and 1 ml of the culture medium was added per well. The inoculated MARC-145 cells were examined daily for up to 5 days for development of cytopathic effects (CPE), and the inoculated ZMAC cells were observed for 3 days. Then, cell culture supernatants were harvested and the cell plates were fixed with 80% cold acetone for 10 min at room temperature followed by immunofluorescence staining. If the interpretation following immunofluorescence staining was unclear after two passages in cell cultures (e.g., if the immunofluorescence staining had some backgrounds and the result could not be clearly called positive or negative), the supernatants were tested by PRRSV RT-rtPCR to confirm VI results. If the first passage (P0) was VI negative, the P0 supernatants were inoculated into the respective cell lines for the second passage. The virus isolation result was considered negative if the second passage was still negative.

Immunofluorescence staining.

Both direct immunofluorescence staining and indirect immunofluorescence staining were initially conducted in two cell lines inoculated with PRRSV samples. For direct immunofluorescence staining, the acetone-fixed cells were incubated with pooled PRRSV nucleocapsid (N) protein-specific monoclonal antibodies SDOW17-F and SR30-F conjugated to fluorescein isothiocyanate (FITC) (Rural Technologies, Inc., Brookings, SD) for 1 h at 37°C. The antibody conjugates were decanted, and the cell plates were washed with phosphate-buffered saline (PBS) (1×; pH 7.4) 3 times for 5 min per time. Plates were read using an Olympus IX71 fluorescence microscope (Olympus America Inc., Center Valley, PA). For indirect immunofluorescence staining, the acetone-fixed cells were incubated with pooled PRRSV nucleocapsid protein-specific unconjugated monoclonal antibodies SDOW17-A and SR30-A (Rural Technologies, Inc.) for 45 min at 37°C. The antibodies were decanted, and the cell plates were washed three times (5 min each time) with PBS. Then, the cell plates were incubated with a goat anti-mouse secondary antibody conjugated with FITC (VWR International, Brooklyn, NY) for 30 min at 37°C. Three additional 5-min PBS washes were conducted, after which the plates were read using a fluorescence microscope as described above. After optimization of immunofluorescence staining, the inoculated MARC-145 cells were subjected only to direct immunofluorescence staining and the inoculated ZMAC cells were subjected only to indirect immunofluorescence staining.

Nucleic acid extraction.

Viral nucleic acids were extracted from clinical samples or cell culture lysates using a MagMAX Pathogen RNA/DNA kit (Thermo Fisher Scientific, Waltham, MA) and a Kingfisher Flex instrument (Thermo Fisher Scientific) following the instructions of the manufacturer. Nucleic acids were eluted into 90 µl of elution buffer.

PRRSV reverse transcription real-time PCR.

A commercial PRRSV RT-rtPCR reagent, VetMAX PRRSV NA&EU (Thermo Fisher Scientific), was used to test samples for the presence or absence of PRRSV RNA following the manufacturer’s instructions with some modifications. PRRSV RT-rtPCR was set up in a 20-μl reaction mixture containing the following: 6.50 μl of TaqMan Fast 1-Step master mix (Thermo Fisher Scientific), 0.80 µl of AmpliTaq 360 DNA polymerase (Thermo Fisher Scientific) (5 U/µl), 2.0 μl of 10× PRRSV Primer and Probe Mix V2 (Thermo Fisher Scientific), 2.7 μl nuclease-free water, and 8 μl nucleic acid extract. Amplification reactions were performed on an ABI 7500 Fast instrument (Thermo Fisher Scientific) under the following conditions: 1 cycle of 50°C for 5 min, 1 cycle of 95°C for 20 s, and 40 cycles of 95°C for 3 s and 60°C for 30 s. This commercial PRRSV RT-rtPCR can simultaneously detect and differentiate PRRSV-1 and PRRSV-2. A threshold cycle (C_T) value of <37 was considered positive and a C_T value of ≥37 was considered negative for either PRRSV-1 or PRRSV-2. The analytical sensitivity of this commercial PRRSV RT-rtPCR was estimated to be 5 to 10 genomic copies per reaction.

PRRSV ORF5 sequencing by the Sanger method and phylogenetic analyses.

PRRSV-2 ORF5 sequencing was performed on the clinical samples using the Sanger method following previously described procedures (25). The 234 PRRSV-2 ORF5 sequences determined for the clinical samples in this study together with 841 reference sequences representing the nine lineages of PRRSV-2 (9, 26) were aligned using the progressive method (FFT-NS-2) in MAFFT v7.407 (27). The four independent phylogenetic trees were constructed in IQ-TREE v1.6.12 (28) using maximum likelihood with a stochastic algorithm and systematical nucleotide substitution with 10 categories of a FreeRate heterogeneity model (SYM+R10) with 1,000 bootstrap replicates. The nucleotide substitution model was identified using ModelFinder (29) in IQ-TREE, which computed the log-likelihood scores of parsimony trees for 286 nucleotide substitution models with the Akaike information criterion (AIC), the corrected Akaike information criterion (AICc), and the Bayesian information criterion (BIC). To identify the best tree, tree topology tests were performed using resampling estimated log likelihood (RELL) approximation with 10,000 replicates (30), a weighted Kishino-Hasegawa (KH) test (31), a weighted Shimodaira-Hasegawa test (32), and an approximately unbiased (AU) test (33). The phylogenetic tree was annotated using MEGA6.0 (34). Nucleotide identities were determined by the use of the MegAlign 15 program in DNASTAR Lasergene 15 software.

Virus titration.

Twenty-three PRRSV-2 PCR-positive samples that were VI positive in both MARC-145 and ZMAC cell lines were selected for virus titer comparisons. Basically, 23 virus isolates obtained in MARC-145 cells were titrated in MARC-145 cells and 23 virus isolates obtained in ZMAC cells were titrated in ZMAC cells. Briefly, each virus isolate was serially diluted 10-fold and inoculated into MARC-145 cells or ZMAC cells grown in 96-well plates (100 µl per well), with five replicate wells per dilution. The plates were incubated at 37°C with 5% CO₂ for 3 to 5 days. Viral CPE was recorded daily. The cell culture supernatants were harvested, and the cell plates were fixed with 80% cold acetone for 10 min at room temperature. Immunofluorescence staining on the fixed cells was conducted as described above. The virus titers were calculated according to the method described by Reed and Muench (35) and expressed as 50% tissue culture infective doses (TCID₅₀) per milliliter.

Virus growth cross-checking.

Eighty-two PRRSV-2 isolates obtained in ZMAC cells were inoculated into MARC-145 cells to evaluate their growth. Similarly, 45 PRRSV-2 isolates obtained in MARC-145 cells were inoculated into ZMAC cells to evaluate their growth. Thirteen PRRSV-1 isolates obtained in ZMAC cells were inoculated into MARC-145 cells and 1 PRRSV-1 isolate obtained in MARC-145 cells was inoculated into ZMAC cells to evaluate their growth. When PRRSV-1 and PRRSV-2 isolates were both counted, 95 isolates obtained in ZMAC cells were inoculated into MARC-145 cells and 46 isolates obtained in MARC-145 cells were inoculated into ZMAC cells for evaluating their growth.

Statistical analyses.

McNemar’s test was performed to investigate the effects of cell lines or lineage classification on PRRSV VI outcomes. The effects of C_T values and specimen types on PRRSV VI were evaluated by penalized maximum likelihood estimates. Fisher’s exact test was performed to examine the results of virus growth cross-checking.

RESULTS

Cytopathic effects and optimization of immunofluorescence staining of ZMAC and MARC-145 cells inoculated with PRRSV.

In mock-inoculated MARC-145 and mock-inoculated ZMAC cells, no viral CPE was observed. In PRRSV-inoculated MARC-145 cells, viral CPE generally showed up during 2 to 5 days postinoculation (dpi) compared to 1 to 2 days in PRRSV-inoculated ZMAC cells. Immunofluorescence staining was conducted to confirm PRRSV VI in MARC-145 and ZMAC cells. Mock-infected and PRRSV-infected MARC-145 cells can be readily distinguished by either direct immunofluorescence staining (Fig. 1A and B) or indirect immunofluorescence staining (Fig. 1C and D). However, the mock-infected ZMAC cells exhibited quite strong background staining under the conditions of the direct immunofluorescence procedure (Fig. 1E) that made it difficult to clearly distinguish those cells from the PRRSV-infected ZMAC cells (Fig. 1F). In contrast, when indirect immunofluorescence staining was conducted, the mock-infected and PRRSV-infected ZMAC cells were readily distinguishable (Fig. 1G and H). Therefore, for PRRSV VI confirmation, direct immunofluorescence staining was conducted on the inoculated MARC-145 cells and indirect immunofluorescence staining on the inoculated ZMAC cells.

FIG 1.

Immunofluorescence staining of PRRSV in ZMAC and MARC-145 cells. Examples of direct immunofluorescence staining of mock- and PRRSV-infected MARC-145 and ZMAC cells, using FITC-conjugated PRRSV monoclonal antibodies, are depicted in panels A, B, E, and F. Examples of indirect immunofluorescence staining of mock- and PRRSV-infected MARC-145 and ZMAC cells, using unconjugated PRRSV monoclonal antibodies and FITC-conjugated goat anti-mouse secondary antibody, are depicted in panels C, D, G, and H.

Comparison of PRRSV-1 VI from clinical samples in ZMAC and MARC-145 cells.

Seventy PRRSV-1 PCR-positive clinical samples (31 serum, 8 lung, and 31 oral fluid samples) with various C_T values were used for VI attempts. None of the 31 PRRSV-1 oral fluid samples was VI positive in either ZMAC or MARC-145 cells (Table 1). Among the PRRSV-1 serum and lung samples, 25.8% (8/31) and 0% (0/31) of serum samples, 62.5% (5/8) and 12.5% (1/8) of lung samples, and 33.3% (13/39) and 2.6% (1/39) of serum plus lung samples were VI positive in ZMAC and MARC-145 cells, respectively (Table 1). The results of 2-by-2 table analyses of PRRSV-1 VI outcomes in ZMAC and MARC-145 cells are summarized in Table 2. The only PRRSV-1 sample that was VI positive in MARC-145 cells was also VI positive in ZMAC cells. Among the 13 PRRSV-1 samples that were VI positive in ZMAC cells, 1 was VI positive and 12 were VI negative in MARC-145 cells (Table 2). The PRRSV-1 VI success rates were significantly higher in ZMAC cells than in MARC-145 cells for serum (P = 0.0047), lung (P = 0.0455), and serum plus lung (P = 0.0005).

TABLE 1.

Outcomes of PRRSV isolation from clinical samples in ZMAC and MARC-145 cells

Virus type	Specimen	No. of samples	% of samples (no. of positive samples/total no. of samples)
			ZMAC VI+	MARC-145 VI+
PRRSV-2	Serum	109	47.7 (52/109)	9.2 (10/109)
	Lung	96	68.8 (66/96)	45.8 (44/96)
	Oral fluid	59	0 (0/59)	0 (0/59)
	Processing fluid	41	9.8 (4/41)	7.3 (3/41)
	All	305
PRRSV-1	Serum	31	25.8 (8/31)	0 (0/31)
	Lung	8	62.5 (5/8)	12.5 (1/8)
	Oral fluid	31	0 (0/31)	0 (0/31)
	All	70

TABLE 2.

Comparison of PRRSV isolation results in MARC-145 and ZMAC cells in clinical samples

Virus type	Specimen(s)	Cell line(s)	No. of samples
			MARC-145 VI+	MARC-145 VI−	Total
PRRSV-2	Serum	ZMAC VI+	10	42	52
		ZMAC VI−	0	57	57
		Both	10	99	109
	Lung	ZMAC VI+	37	29	66
		ZMAC VI−	7	23	30
		Both	44	52	96
	Serum and lung	ZMAC VI+	47	71	118
		ZMAC VI−	7	80	87
		Both	54	151	205
	Processing fluid	ZMAC VI+	3	1	4
		ZMAC VI−	0	37	37
		Both	3	38	41
PRRSV-1	Serum	ZMAC VI+	0	8	8
		ZMAC VI−	0	23	23
		Both	0	31	31
	Lung	ZMAC VI+	1	4	5
		ZMAC VI−	0	3	3
		Both	1	7	8
	Serum and lung	ZMAC VI+	1	12	13
		ZMAC VI−	0	26	26
		Both	1	38	39

The success rates of isolation of PRRSV-1 from serum and lung samples with different C_T values are summarized in Table 3. However, due to small number of PRRSV-1 samples, impacts of PRRSV-1 C_T values on the VI outcomes were not statistically analyzed.

TABLE 3.

Distribution of PRRSV PCR C_T values and virus isolation outcomes in ZMAC and MARC-145 cells based on specimen types^a

Virus type	CT value	Serum value(s)				Lung				Serum and lung
		Total no. of samples	CT mean ± SD	No. (%) of ZMAC VI+specimens	No. (%) ofMARC-145 VI+ specimens	Total no. of samples	CT mean ± SD	No. (%) of ZMAC VI+specimens	No. (%) ofMARC-145 VI+ specimens	Total no. of samples	CT mean ± SD	No. (%) of ZMAC VI+specimens	No. (%) of MARC-145 VI+ specimens
PRRSV-2	<20	22	16.90 ± 1.98	19 (86.4)	5 (22.7)	31	17.07 ± 1.67	29 (93.5)	21 (67.7)	53	17.00 ± 1.79	48 (90.6)	26 (49.1)
	20–25	32	22.47 ± 1.38	23 (71.9)	5 (15.6)	41	22.44 ± 1.39	29 (70.7)	19 (46.3)	73	22.45 ± 1.38	52 (71.2)	24 (32.9)
	25–30	31	27.72 ± 1.52	9 (29.0)	0	23	27.17 ± 1.40	8 (34.8)	4 (17.4)	54	27.49 ± 1.48	17 (31.5)	4 (7.4)
	30–33	17	31.41 ± 0.75	1 (5.9)	0	1	30.20	0	0	18	31.34 ± 0.78	1 (5.6)	0
	33–37	7	34.23 ± 1.05	0	0	0	NA	0	0	7	34.23 ± 1.05	0	0
	All	109	24.99 ± 5.61	52 (47.7)	10 (9.2)	96	21.92 ± 4.16	66 (68.8)	44 (45.8)	205	23.55 ± 5.21	118 (57.6)	54 (26.3)
PRRSV-1	<20	1	18.40	0	0	0	NA	0	0	1	18.40	0	0
	20–25	6	22.90 ± 1.00	3 (50)	0	5	23.04 ± 0.93	5 (100.0)	1 (25.0)	11	22.96 ± 0.92	8 (72.7)	1 (9.1)
	25–30	9	27.43 ± 1.07	4 (44.4)	0	1	27.50	0	0	10	27.44 ± 1.01	4 (40.0)	0
	30–33	11	31.18 ± 0.96	1 (9.1)	0	0	NA	0	0	11	31.18 ± 0.96	1 (9.1)	0
	33–37	4	35.08 ± 1.65	0	0	2	34.85 ± 0.78	0	0	6	35.00 ± 1.33	0	0
	All	31	28.58 ± 4.41	8 (25.8)	0	8	26.55 ± 5.40	5 (62.5)	1 (12.5)	39	28.16 ± 4.63	13 (33.3)	1 (2.6)

^aPRRSV-1 VI success rates were significantly higher in ZMAC than in MARC-145 cells for serum samples (P = 0.0047), lung samples (P = 0.0455), and serum samples plus lung samples (P = 0.0005). PRRSV-2 VI success rates were significantly higher in ZMAC than in MARC-145 cells for serum samples (P < 0.0001), lung samples (P = 0.0002), and serum samples plus lung samples (P < 0.0001). PRRSV-2 VI success rates were significantly higher in lung samples than in serum samples in either MARC-145 cells (P < 0.0001) or ZMAC cells (P = 0.0028). In the category of C_T values of <20, the mean C_T values were not significantly different between serum and lung samples (P = 0.75). In the category of C_T values of 20 to 25, the mean C_T values were not significantly different between serum and lung samples (P = 0.92). In the category of C_T values of 25 to 30, the mean C_T values were not significantly different between serum and lung samples (P = 0.17). C_T values significantly affected the success rate data for PRRSV-2 VI, with lower VI success rates in samples having higher C_T values in MARC-145 cells (for serum samples, P = 0.0027; for lung samples, P = 0.0003; for serum samples plus lung samples, P < 0.0001) and in ZMAC cells (serum, P < 0.0001; lung, P < 0.0001; serum plus lung, P < 0.0001). NA, not applicable.

Comparison of PRRSV-2 VI from clinical samples in ZMAC and MARC-145 cells.

Among the 305 PRRSV-2 PCR-positive samples, 47.7% (52/109) and 9.2% (10/109) of serum samples, 68.8% (66/96) and 45.8% (44/96) of lung samples, 0% (0/59) and 0% (0/59) of oral fluid samples, and 9.8% (4/41) and 7.3% (3/41) of processing fluid samples were VI positive in ZMAC and MARC-145 cells, respectively (Table 1). When the serum and lung samples were combined, the success rates of PRRSV-2 VI in ZMAC and MARC-145 cells were 57.6% (118/205) and 26.3% (54/205), respectively.

The PRRSV-2 VI results in ZMAC and MARC-145 cells are summarized in Table 2. Among the 41 processing fluid samples, three samples were PRRSV-2 VI positive in both ZMAC and MARC-145 cells whereas one sample was VI positive in ZMAC cells but VI negative in MARC-145 cells. Totals of 10 serum and 37 lung samples were positive in both ZMAC and MARC-145 cells. However, among the 52 serum and 66 lung samples that were VI positive in ZMAC cells, 42 serum and 29 lung samples were VI negative in MARC-145 cells. With the serum and lung samples combined, 47 serum and lung samples were VI positive in both ZMAC and MARC-145 cells whereas 71 serum and lung samples were VI positive in ZMAC cells but negative in MARC-145 cells.

Statistical analyses indicated that the PRRSV-2 VI success rates were significantly higher in ZMAC cells than in MARC-145 cells for serum (P < 0.0001), lung (P = 0.0002), and serum plus lung (P < 0.0001) (Table 2; see also Table 3). In addition, the PRRSV-2 VI success rates were significantly higher in lung samples than in serum samples in either MARC-145 cells (P < 0.0001) or ZMAC cells (P = 0.0028). PRRSV-2 VI success rates were not significantly different between the two cell lines for processing fluid samples.

Impact of C_T values on PRRSV-2 VI outcomes.

Among the 59 oral fluid samples with PRRSV-2 PCR C_T values of 22 to 34, none was VI positive in either the ZMAC or MARC-145 cell line. Among the 41 processing fluid samples with PRRSV-2 PCR C_T values of 15.6 to 30.9, 3 samples (C_T values of 15.6, 16.1, and 16.3) were VI positive in both cell lines and 1 sample (C_T 21.0) was VI positive in the ZMAC cells but negative in the MARC-145 cells. Due to absence of or a low number of VI positive samples, the impact of C_T values on VI outcomes was not analyzed for oral fluid and processing fluid samples. The success rates of isolating PRRSV-2 from serum and lung samples with different C_T ranges are summarized in Table 3. The rates of success of PRRSV-2 VI from serum and lung samples were 90.6% and 49.1% (C_T values of <20), 71.2% and 32.9% (C_T values of 20 to 25), 31.5% and 7.4% (C_T values of 25 to 30), and 5.6% and 0% (C_T values of 30 to 33) in ZMAC and MARC-145 cells, respectively. The trend is clear that the success rates of PRRSV-2 VI decreased in serum and lung samples with increasing C_T values. The PRRSV-2 VI success rate was very low for samples with C_T values of 30 to 33, and there was no VI success seen with samples with C_T values of >33 in either ZMAC or MARC-145 cells. Statistical analyses based on the penalized maximum likelihood estimates method indicated that C_T values significantly affected the success rates of PRRSV-2 VI in MARC-145 cells (for serum samples, P = 0.0027; for lung samples, P = 0.0003; for serum plus lung samples, P < 0.0001) and in ZMAC cells (for serum samples, P < 0.0001; for lung samples, P < 0.0001; for serum plus lung samples, P < 0.0001).

Impact of genetic lineages on PRRSV-2 VI outcomes.

A total of 305 PRRSV-2 clinical samples (109 serum, 96 lung, 59 oral fluid, and 41 processing fluid samples) were used for VI attempts in this study. Due to unavailability of some clinical samples (very limited volume) and failure of ORF5 sequencing on some clinical samples, ORF5 sequences were obtained from 234 PRRSV-2 samples, including 91 serum, 85 lung, 44 oral fluid, and 14 processing fluid samples (see Table S1 in the supplemental material). Phylogenetic analyses revealed that these 234 samples contained PRRSV-2 belonging to lineage 1 (106 samples, including 48 serum, 33 lung, 21 oral fluid, and 4 processing fluid samples), lineage 5 (70 samples, including 27 serum, 30 lung, 12 oral fluid, and 1 processing fluid samples), lineage 6 (1 processing fluid sample), lineage 8 (52 samples, including 16 serum, 17 lung, 11 oral fluid, and 8 processing fluid samples), and lineage 9 (5 lung samples), as described in detail in Table S1. As shown in Fig. S1 in the supplemental material, PRRSV-2 strains in the 234 clinical samples were distributed across different subclusters in each lineage, demonstrating that the selected clinical samples accurately represented the genetic diversity of PRRSV-2 in each lineage. PRRSV-2 VI was negative from all oral fluid samples. Four processing fluid samples were VI positive, but the sequences were not available. Therefore, only 176 serum and lung samples with lineage information were used to assess the impact of lineages on VI outcomes. PRRSV-2 VI outcomes for serum and lung samples classified into different lineages are summarized in Table 4. Only a small number (n = 5) of samples included PRRSV-2 lineage 9 and hence were excluded from the statistical analyses. For lineage 1 and lineage 8 samples (serum plus lung), the VI success rates were significantly higher in ZMAC cells than in MARC-145 cells (P < 0.0001 for lineage 1 and P = 0.0039 for lineage 8). In contrast, for the lineage 5 serum plus lung samples, the rates of VI success were not significantly different between ZMAC and MARC-145 cells (P = 0.3173). Currently, in the United States, there are five commercial PRRSV-2 modified live virus vaccines that are approved for use: Ingelvac PRRS MLV vaccine (lineage 5), Ingelvac PRRS ATP vaccine (lineage 8), Fostera PRRS vaccine (lineage 8), Prime Pac PRRS RR vaccine (lineage 7), and Prevacent PRRS vaccine (lineage 1). The ORF5 nucleotide identities of PRRSV-2 strains in 176 serum and lung samples were compared to those in five commercial PRRSV-2 vaccines, and the results are summarized in Table 5. For the 81 serum and lung samples with the lineage 1 PRRSV sequences, all had ORF5 sequences with <91% similarity to the lineage 1 Prevacent PRRS vaccine virus and were considered wild-type strains (Table 5). These lineage 1 wild-type PRRSV-2 strains appeared to grow better in ZMAC cells than in MARC-145 cells (Table 4). Of the 33 serum and lung samples containing lineage 8 PRRSVs, 2 had ORF5 sequences similar (99.8% to 100% nucleotide identity) to the lineage 8 Ingelvac PRRS ATP vaccine virus, 27 had ORF5 sequences similar (99% to 100% nucleotide identity) to the lineage 8 Fostera PRRS vaccine virus, and 4 had wild-type ORF5 sequences (Table 5). Among the 27 samples with Fostera PRRS vaccine-like sequences, 18 were VI positive in ZMAC cells and 7 were VI positive in MARC-145 cells, suggesting that it may be more efficient to isolate Fostera PRRS vaccine-like virus in ZMAC cells than in MARC-145 cells. For the 57 serum and lung samples with lineage 5 PRRSV sequences, 55/57 had ORF5 sequences similar (98.8% to 100% nucleotide identity) to the Ingelvac PRRS MLV vaccine virus (Table 5); VI outcomes in serum and lung samples containing lineage 5 virus were not significantly different in ZMAC and MARC-145 cells (Table 4). This suggests that the rates of success of isolation of the Ingelvac PRRS MLV vaccine-like virus are similar in ZMAC and MARC-145 cells.

TABLE 4.

Virus isolation outcomes for different PRRSV-2 lineages in ZMAC and MARC-145 cells^a

Lineage	Serum values			Lung values			Serum and lung values
	Total no. of samples	No. (%) of ZMAC VI+ samples	No. (%) of MARC-145 VI+ samples	Total no. of samples	No. (%) of ZMAC VI+ samples	No. (%) of MARC-145 VI+ samples	Total no. of samples	No. (%) of ZMAC VI+ samples	No. (%) of MARC-145 VI+ samples
1	48	29 (60.4)	5 (10.4)	33	24 (72.7)	8 (24.2)	81	53 (65.4)	13 (16.0)
5	27	4 (14.8)	2 (7.4)	30	15 (50)	20 (66.7)	57	19 (33.3)	22 (38.6)
8	16	6 (37.5)	1 (6.3)	17	13 (76.5)	8 (47.1)	33	19 (57.6)	9 (27.3)
9	0	0	0	5	3 (60)	2 (40)	5	3 (60)	2 (40)
All	91	39 (42.9)	8 (8.8)	85	55 (64.7)	38 (44.7)	176	94 (53.4)	46 (26.1)

^aORF5 sequencing was successfully performed on 234 PRRSV-2 PCR-positive specimens (91 serum, 85 lung, 44 oral fluid, and 14 processing fluid samples) used for VI. Since all oral fluid samples were virus isolation negative and the four VI-positive processing fluid samples did not have sequences available, only 176 serum and lung samples were used to analyze the impact of lineages on the VI outcomes. For lineage 1 and lineage 8 samples (serum and lung), the VI success rates were significantly higher in ZMAC cells than in MARC-145 cells (P < 0.0001 for lineage 1 and P = 0.0039 for lineage 8). For the lineage 5 serum and lung samples, the VI success rates determined for the ZMAC and MARC-145 cells were not significantly different (P = 0.3173).

TABLE 5.

ORF5 nucleotide identities of 176 PRRSV-2 serum and lung samples compared to five commercial PRRSV-2 vaccines

Lineage	No. of clinical serum and lung samples	% ORF5 nucleotide identity of PRRSV-2 samples to commercial vaccine:
		Ingelvac MLV (L5)	Ingelvac ATP (L8)	Fostera (L8)	Prime pac (L7)	Prevacent (L1)
1	81	83.7–88.4	83.3–87.6	83.7–88.2	84.6–88.4	85.4–90.7
5	57	89.2–100a	84.7–90.7	85.9–91.7	86.6–92	83.4–87.1
8	33	83.4–91.9	85.2–100b	86.2–100c	84.7–93.2	82.2–86.7
9	5	86.6–88.6	87.1–89.4	90–92	88.1–90.5	84.1–86.1
All	176

^aAmong the 57 lineage 5 samples, 55 had ORF5 sequences similar to that of Ingelvac MLV vaccine (98.8% to 100% nucleotide identity).

^bAmong the 33 lineage 8 samples, 2 had ORF5 sequences similar to that of Ingelvac ATP vaccine (99.8% to 100% nucleotide identity).

^cAmong the 33 lineage 8 samples, 27 had ORF5 sequences similar to that of Fostera (99.0% to 100% nucleotide identity).

Comparison of virus titers of PRRSV-2 isolates obtained in ZMAC and MARC-145 cells.

In order to compare the infectious titers of viruses isolated in ZMAC and MARC-145 cells, 23 PRRSV-2 isolates obtained in ZMAC cells were titrated in ZMAC cells and 23 PRRSV-2 isolates obtained in MARC-145 cells were titrated in MARC-145 cells; these virus isolates were derived from the same set of 23 clinical samples. As shown in Fig. 2, most of the isolates (14/23) obtained in ZMAC and MARC-145 cells had similar infectious titers whereas 6 ZMAC isolates had higher titers than the MARC-145 isolates and 3 ZMAC isolates had lower titers than the MARC-145 isolates.

FIG 2.

Comparison of infectious titers (TCID₅₀/ml) of 23 PRRSV-2 isolates obtained in ZMAC and MARC-145 cells. The same sets of 23 clinical samples were VI positive in both ZMAC and MARC-145 cells. The isolates obtained in ZMAC cells were titrated in ZMAC cells, and the isolates obtained in MARC-145 cells were titrated in MARC-145 cells. Six isolates that had higher titers in ZMAC cells than in MARC-145 cells are boxed with a dashed line. Fourteen ZMAC and MARC-145 isolates that had similar titers are boxed with a solid line. The lineage (e.g., L1) and RFLP pattern (e.g., 1-7-4) of each isolate are also indicated.

Cross-checking of growth of PRRSV isolates in ZMAC and MARC-145 cells.

In order to determine whether isolates derived from ZMAC cells can grow in MARC-145 cells and vice versa, 82 PRRSV-2 isolates and 13 PRRSV-1 isolates obtained in ZMAC cells were inoculated into MARC-145 cells and 45 PRRSV-2 isolates and one PRRSV-1 isolate obtained in MARC-145 cells were inoculated into ZMAC cells to evaluate their growth. The data are summarized in Tables 6, 7, and 8 (see also Table S1).

TABLE 6.

Cross-checking of growth of PRRSV-2 isolates in ZMAC and MARC-145 cells for 82 PRRSV-2 isolates obtained in ZMAC cells and reinoculated into MARC-145 cells^a

Initial VI result in ZMAC cells	Initial VI result in MARC-145 cells	Result for ZMAC isolates reinoculated into MARC-145 cells	No. (%) of isolates	No. of isolates of indicated lineageb
Pos	Pos	Pos	33 (40.2)	6L1 + 16L5 + 8L8 + 1L9 + 2 no seq
Pos	Neg	Pos	14 (17.1)	6L1 + 2L5 + 2L8 + 1L9 + 3 no seq
Pos	Pos	Neg	5 (6.1)	4L1 + 1 no seq
Pos	Neg	Neg	30 (36.6)	17L1 + 8L8 + 5 no seq
Subtotal			82 (100)	33L1 + 18L5 + 18L8 + 2L9 + 11 no seq

^aPos, positive; Neg, negative.

^bData are presented as the number of isolates of the indicated lineage (e.g., “6L1” represents six isolates of lineage 1) or as the number of isolates with no sequence identity (no seq).

TABLE 7.

Cross-checking of growth of PRRSV-2 isolates in ZMAC and MARC-145 cells for 45 PRRSV-2 isolates obtained in MARC-145 cells and reinoculated into ZMAC cells^a

Initial VI result in MARC-145 cells	Initial VI result in ZMAC cells	Result for MARC-145 isolates reinoculated into ZMAC cells	No. (%) of isolates	No. of isolates of indicated lineageb
Pos	Pos	Pos	38 (84.4)	10L1 + 16L5 + 8L8 + 1L9 + 3 no seq
Pos	Neg	Pos	7 (15.5)	6L5 + 1L8
Pos	Pos	Neg	0
Pos	Neg	Neg	0
Subtotal			45 (100)	10L1 + 22L5 + 9L8 + 1L9 + 3 no seq

^aPos, positive; Neg, negative.

^bData are presented as the number of isolates of the indicated lineage (e.g., “10L1” represents 10 isolates of lineage 1) or the number of isolates with no sequence identity (no seq).

TABLE 8.

Cross-checking of growth of PRRSV-1 isolates in ZMAC and MARC-145 cells for 13 PRRSV-1 isolates obtained in ZMAC cells and reinoculated into MARC-145 cells^a

Initial VI result in ZMAC cells	Initial VI result in MARC-145 cells	Result for ZMAC isolates reinoculated into MARC-145 cells	No. (%) of isolates
Pos	Pos	Pos	1 (7.7)
Pos	Neg	Pos	3 (23.1)
Pos	Pos	Neg	0
Pos	Neg	Neg	9 (69.2)
Subtotal			13 (100)

^aPos, positive; Neg, negative.

Among the 82 PRRSV-2 isolates obtained in ZMAC cells, 47 (57.3%) grew and 35 (42.7%) did not grow in MARC-145 cells. The 47 ZMAC isolates that grew in MARC-145 cells were obtained from 47 clinical samples, 33 of which were also VI positive in MARC-145 cells. The 35 ZMAC isolates that did not grow in MARC-145 cells were derived from 35 clinical samples, 5 of which were VI positive in MARC-145 cells and 30 of which were VI negative in MARC-145 cells. The genetic lineage information corresponding to the PRRSV-2 isolates in each category is provided in Tables 6, 7, and 8 as well. The 82 PRRSV-2 isolates obtained in ZMAC cells included 33 lineage 1, 18 lineage 5, 18 lineage 8, 2 lineage 9, and 11 with sequences unavailable. Among them, 12/33 lineage 1, 18/18 lineage 5, 10/18 lineage 8, and 2/2 lineage 9 isolates obtained in ZMAC cells grew in MARC-145 cells. Statistical analysis using Fisher’s exact test indicated that there was a significant difference (P < 0.0001) between the different genetic lineages of ZMAC-derived PRRSV-2 isolates regarding their ability to grow in MARC-145 cells.

All of the 45 PRRSV-2 isolates obtained in MARC-145 cells grew in ZMAC cells. Among them, 38 original samples were VI positive and 7 original samples were VI negative in ZMAC cells. The 45 PRRSV-2 isolates obtained in MARC-145 cells included 10 lineage 1, 22 lineage 5, 9 lineage 8, 1 lineage 9, and 3 with sequences unavailable.

When the 13 PRRSV-1 isolates obtained in ZMAC cells were inoculated into MARC-145 cells, 4 (30.8%) grew and 9 (69.2%) did not grow in MARC-145 cells. When one PRRSV-1 isolate obtained in MARC-145 cells was inoculated into ZMAC cells, it grew.

DISCUSSION

Currently, the MARC-145 cell line is most commonly used in veterinary diagnostic and research laboratories for PRRSV isolation, propagation, and titration despite the low rate of success of primary isolation from clinical samples. PAMs are reported to be more sensitive than MARC-145 cells for VI (23), but it has also been reported that PRRSV strains differ in their abilities to replicate in PAMs and MA-104 cells (22). Ideally, PRRSV VI should be attempted in both PAMs and MARC-145 cells to achieve better outcomes. However, PAMs are primary cells and some disadvantages (e.g., a requirement for periodical preparation, quality variations between batches of PAMs, etc.) have limited the widespread use of PAMs. An immortalized cell line, ZMAC, derived from PAMs, has recently become available (24) and could potentially improve PRRSV VI. We initially planned to compare PRRSV VI results in PAM, ZMAC, and MARC-145 cells. However, two batches of PAMs prepared by us did not have good qualities (low percentage of cell viability). In order to avoid obtaining biased results from these PAMs, we decided not to include PAMs for the comparisons in the present study. Thus, in the current study, comparison of PRRSV VI results from clinical samples was conducted in MARC-145 and ZMAC cells.

Conventionally, specimens collected from individual pigs for PRRSV testing include serum or blood specimens, blood swabs, tissue samples (e.g., lung, tonsil, and lymph nodes), bronchoalveolar lavage fluid samples, semen samples, and so on. In order to reduce the testing cost, pooling the individual samples prior to testing is a common strategy. However, in recent years, there has been growing use of population-based specimen types (e.g., oral fluid and processing fluid) for PRRSV surveillance testing (36, 37). In the current study, 375 clinical samples, including four specimen types (serum, lung, oral fluid, and processing fluid) were used for PRRSV VI comparison in MARC-145 and ZMAC cells. Among them, 305 samples were PRRSV-2 PCR positive and 70 samples were PRRSV-1 PCR positive. PRRSV-1 and PRRSV-2 VI was attempted from 90 oral fluid samples with C_T ranges of 22.0 to 36.8, but there was no VI success with any tested oral fluids in either ZMAC or MARC-145 cells, indicating that PRRSV VI from oral fluid is still a big challenge. It is possible that chemical and physical properties of oral fluid may negatively affect PRRSV VI, but the details of the mechanisms remain undetermined. Since 2017, processing fluid has been increasingly used for monitoring PRRSV status in breeding herds (37, 38). In the current study, 41 PRRSV-2 PCR-positive processing fluid samples were included for VI attempts. Only four processing fluid samples with very low C_T values (15.6, 16.1, 16.3, and 21.0) had positive VI results (see Table S1 in the supplemental material), suggesting that PRRSV can be isolated from processing fluid but that the overall success rate is low even when both MARC-145 and ZMAC cells are used. In this study, 140 serum samples (109 PRRSV-2 PCR positive and 31 PRRSV-1 PCR positive) and 104 lung samples (96 PRRSV-2 PCR positive and 8 PRRSV-1 PCR positive) were also included for VI evaluation. For clinical serum and/or lung samples, the success rates of PRRSV-2 and PRRSV-1 VI were significantly higher in ZMAC cells than in MARC-145 cells. In addition, the PRRSV-2 VI success rates were significantly higher in lung than in serum samples in either the MARC-145 or ZMAC cell line. Therefore, with respect to success rates, the preferred specimen types for PRRSV VI are lung > serum > processing fluid and oral fluid.

The amount of viral load in samples can be reflected by C_T values of real-time PCR. This study further examined the effect of C_T values on PRRSV-2 VI success in ZMAC and MARC-145 cells. Obviously, the PRRSV-2 VI success rate in ZMAC and MARC-145 cells tended to decrease in lung and serum samples with higher C_T values (corresponding to lower viral loads). PRRSV-2 was poorly isolated from samples with C_T values of >30, and no PRRSV VI success was observed with samples with C_T values of >33 in either the ZMAC or MARC-145 cell line. Based on the data from this study, for cost-effectiveness, it is recommended to conduct PRRSV VI from clinical samples with C_T values of <30.

Restriction fragment length polymorphism (RFLP) analysis of ORF5 has been used to describe the genetic diversity of PRRSV-2 throughout North America since the late 1990s (39) despite its apparent drawbacks (40–42). In 2010, a classification system based on the phylogenetic relatedness of the ORF5 sequences was proposed, and global PRRSV-2 strains were classified into 9 lineages and 37 sublineages (9). In this study, 234 PRRSV-2 ORF5 sequences from the clinical samples (serum, lung, oral fluid, and processing fluid) were classified into four major lineages (lineages 1, 5, 8, and 9) and the effect of genetic lineages on PRRSV-2 VI was investigated for the first time in detail using 176 ORF5 sequences from serum and lung samples (Table 4). Lung and serum samples containing PRRSV-2 in lineages 1 and 8 had significantly higher VI success rates in ZMAC than in MARC-145 cells. PRRSV-2 VI outcomes in serum and lung samples containing lineage 5 virus were not significantly different in ZMAC and MARC-145 cells.

In a previous study, virus titers in porcine endometrial (PEE) cells, MARC-145 cells, and PAMs inoculated with a PRRSV VR-2332 isolate were compared (43). The isolate grown in MARC-145 cells had the lowest virus titer compared to those in PEE cells and PAMs, and the isolate obtained in PAMs had higher titer than that obtained in PEE cells (43). However, in the current study, 70% of field isolates had similar infectious titers in ZMAC and MARC-145 cells and 25% of field isolates obtained in ZMAC cells had higher virus titers than those obtained in MARC-145 cells. Hence, ZMAC cells can be regarded as potential host cells for PRRS vaccine production, yielding virus titers similar to those yielded by the MARC-145 cells that are currently used in most situations. In the present study, we demonstrated that all PRRSV isolates obtained in MARC-145 cells grew in ZMAC cells. Therefore, there would not be a concern about adapting MARC-145 virus isolates to grow in ZMAC cells if ZMAC cells are adopted for vaccine production. However, the laboratories of some autogenous vaccine production companies still use MARC-145 cells for PRRSV propagation and vaccine production. On the basis of the results from the current study, 42.7% of PRRSV-2 isolates and 69.2% of PRRSV-1 isolates obtained in ZMAC cells did not grow in MARC-145 cells. Accordingly, autogenous vaccine production companies may sometimes fail to propagate the virus in MARC-145 cells if PRRSV isolates obtained in ZMAC cells in diagnostic laboratories are forwarded to them. Maybe introduction of the use of ZMAC cells in the laboratories of some PRRSV autogenous vaccine production companies should be considered.

Overall, a better PRRSV VI outcome can be achieved in ZMAC cells than in MARC-145 cells. The details of the mechanisms remain to be elucidated. It is suspected that the mechanisms are related to virus genetic diversity and the interaction between viral proteins and host cell receptors. A number of molecules have been described as potential cellular receptors for PRRSV, but only two of them have been studied in detail: sialoadhesin (Sn) and CD163 (44). Sialoadhesin is a macrophage-restricted lectin, and sialoadhesin on PAMs binds to sialic acids on the PRRSV virion, responsible for viral attachment and internalization (45, 46). However, Sn alone does not induce viral uncoating, and another molecule, CD163, is required for viral uncoating and productive infection in PAMs (47, 48). It was found that the level of PRRSV infection was reduced up to 70% when PAMs were incubated at 37°C with either Sn- or CD163-specific antibodies, but no PRRSV infection was observed after PAMs were incubated with both Sn- and CD163-specific antibodies (47). In addition, PRRSV production increased 10 to 100 times in nonpermissive cells coexpressing Sn and CD163 compared to cells expressing only CD163 (47). Thus, Sn and CD163 were found to be important for effective PRRSV infection in PAMs. Unlike PAMs, MARC-145 cells express only CD163 and not Sn (21, 44), which may explain the relatively inefficient PRRSV infection seen in MARC-145 cells compared to PAMs. In addition, monkey CD163 on MARC-145 cells has some amino acid differences from swine CD163 on PAMs (21) and it is unclear whether these CD163 differences also affect the efficiency of PRRSV replication in MARC-145 compared to PAM cells. Expression of Sn and CD163 on ZMAC cells has not yet been explored, and it remains to be deciphered in future studies.

In conclusion, under the conditions of this study, we demonstrated the following. (i) It is a challenge to isolate PRRSV from oral fluid and processing fluid samples regardless of whether ZMAC cells or MARC-145 cells are used. The preferred specimen types for PRRSV VI are lung > serum > processing fluid and oral fluid. (ii) For clinical serum and/or lung samples, the rates of success of PRRSV VI were significantly higher in ZMAC cells than in MARC-145 cells. When PRRSV VI results are negative in MARC-145 cells, PRRSV VI attempts in ZMAC cells are recommended. (iii) For cost-effectiveness, it is recommended to conduct PRRSV VI from clinical samples with C_T values of <30. (iv) For samples containing lineage 1 and lineage 8 PRRSV-2, VI success rates were significantly higher in ZMAC than in MARC-145 cells. For samples containing lineage 5 Ingelvac PRRS MLV, the VI success rates were similar in two cell lines. If the PRRSV-2 genetic lineage information is unknown in a clinical sample, the VI attempts will result in a better overall outcome in ZMAC cells than in MARC-145 cells. (v) Most of the PRRSV-2 isolates obtained in ZMAC cells and MARC-145 cells had similar infectious titers. (vi) Not all of the PRRSV-2 isolates obtained in ZMAC cells grew in MARC-145 cells, while all of the PRRSV-2 isolates obtained in MARC-145 cells grew in ZMAC cells. These data can directly provide some guidance to swine veterinarians, producers, and researchers to better serve their needs of obtaining PRRSV isolates for further characterization or for producing autogenous vaccines.