Propagating and banking genetically diverse human sapovirus strains using a human duodenal cell line: investigating antigenic differences between strains

ABSTRACT

There are four genogroups and 18 genotypes of human sapoviruses (HuSaVs) responsible for acute gastroenteritis. To comprehend their antigenic and virological differences, it is crucial to obtain viral stocks of the different strains. Previously, we utilized the human duodenum-derived cell line HuTu80, and glycocholate, a conjugated bile acid, to replicate and propagate GI.1, GI.2, and GII.3 HuSaVs (H. Takagi et al., Proc Natl Acad Sci U S A 117:32078–32085, 2020, https://10.1073/pnas.2007310117). First, we investigated the impact of HuTu80 passage number on HuSaV propagation. Second, we demonstrated that taurocholate improved the initial replication success rate and viral RNA levels in fecal specimens relative to glycocholate. By propagating 15 HuSaV genotypes (GI.1–7, GII.1–5, −8, and GV.1–2) and accomplishing preparation of viral stocks containing 1.0 × 10⁹ to 3.4 × 10¹¹ viral genomic copies/mL, we found that all strains required bile acids for replication, with GII.4 showing strict requirements for taurocholate. The deduced VP1 sequences of the viruses during the scale-up of serial passaged virus cultures were either identical or differed by only two amino acids from the original sequences in feces. In addition, we purified virions from nine strains of different genotypes and used them as immunogens for antiserum production. Enzyme-linked immunosorbent assays (ELISAs) using rabbit and guinea pig antisera for each of the 15 strains of different genotypes revealed distinct antigenicity among the propagating viruses across genogroups and differences between genotypes. Acquisition of biobanked viral resources and determination of key culture conditions will be valuable to gain insights into the common mechanisms of HuSaV infection.

IMPORTANCE

The control of human sapovirus, which causes acute gastroenteritis in individuals of all ages, is challenging because of its association with outbreaks similar to those caused by human norovirus. The establishment of conditions for efficient viral propagation of various viral strains is essential for understanding the infection mechanism and identifying potential control methods. In this study, two critical factors for human sapovirus propagation in a conventional human duodenal cell line were identified, and 15 strains of different genotypes that differed at the genetic and antigenic levels were isolated and used to prepare virus stocks. The preparation of virus stocks has not been successful for noroviruses, which belong to the same family as sapoviruses. Securing virus stocks of multiple human sapovirus strains represents a significant advance toward establishing a reliable experimental system that does not depend on limited virus-positive fecal material.

INTRODUCTION

Human sapoviruses (HuSaVs), similar to human noroviruses, are transmitted orally and cause acute gastroenteritis in people of all ages. Human-to-human contact and food-borne and water-borne HuSaV outbreaks have also been reported (1–6). The detection of HuSaVs has recently been facilitated through the development of improved primer sets and the utilization of next-generation sequencing techniques (1, 3, 7–14).

HuSaVs have a single-stranded, non-segmented RNA genome approximately 7,500 bases in length and possess two open-reading frames (ORFs). ORF1 (approximately 2,270–2,300 amino acids [aa] in length) comprises nonstructural proteins (NS1–7) and major capsid protein (VP1), whereas ORF2 encodes a minor capsid protein, VP2, of approximately 165 aa. The 5’- and 3-untranslated regions are less than 14 and 140 nucleotides in length, respectively (15). The surface structure of the HuSaV virion consists of 180 VP1 molecules. VP1 can be separated into an N-terminal arm (NTA), shell (S) domain, and protruding (P) domain. The P domain is further divided into two subdomains, P1 and P2, of which P2 forms the outermost virion structure (16, 17). HuSaVs are classified into four genogroups (GI, GII, GIV, and GV) and 18 genotypes (GI.1–7, GII.1–8, GIV.1, GV.1, and GV.2) based on VP1 sequences (9). GI.1 and GI.2 are the two most frequent HuSaV genotypes in public databases (18). A new genotype (GII.NA) candidate has also been reported for GII (19).

Propagation of HuSaVs has been challenging since their discovery by electron microscopy in the mid-1970s (1, 20–22). The HuSaV replication/propagation was not reported until 2020 (23). Using the human duodenum-derived HuTu80 cell line and including glycocholate (GlyCA), a conjugated bile acid physiologically abundant in the human duodenum (24, 25), in the culture medium resulted in the growth of HuSaV GI.1 and GII.3 to ~6 log₁₀-fold maximum from the input viral load, and serial viral passaging at an increased culture scale became possible (23). These HuSaV GI.1 and GII.3 strains were subsequently analyzed for their protein composition (VP1 and VP2) using purified virions (26), and the concentrated GI.1 strain was used to evaluate virus removal and inactivation in water treatment processes (27). We also reported the production of viruses by reverse genetics of the HuSaV GII.3 strain directly transfected into HuTu80 cells (26).

In 2023, Euller-Nicolas et al. reported HuSaV GI.1 and GI.2 replication using human small intestinal (duodenum, jejunum, and ileum) enteroids as well as HuTu80. Viral RNA levels increased approximately 3 log₁₀-fold (1115-fold maximum) and 2 log₁₀-fold in enteroids and HuTu80, respectively, when cultured in media supplemented with glycochenodeoxycholate (GCDCA) or GlyCA (28). Matsumoto et al. detected an increase in viral RNA in HuSaV GI.1, GI.2, GII.1, GII.3, and GIV.1 to ~2 log₁₀-fold (876-fold maximum) in iPS-derived human intestinal epithelial cells with bile-supplemented medium (29).

HuTu80 cells at passage number ≤80 were used in a previous study (23). In our preliminary experiment, increased VP1 protein levels in the cell culture supernatants were not detected using HuTu80 cells at passage number 40 when GI.1 and GII.3 HuSaV stock virus was employed as the inoculum (30).

In this study, we describe the successful propagation and serial passaging of multiple HuSaV strains of different genotypes using optimized growth conditions including bile acid use. Antigenic differences among HuSaV genotypes were previously evaluated using recombinant VP1, virus-like particles, or clinical specimens only for a limited number of genotypes (31–37). This is the first study to report minimal changes in sequences during serial passage with scaled-up propagation for 15 strains of different genotypes and the antigenic differences between them.

RESULTS

Human sapoviruses propagate well in high-passage number HuTu80 cells

First, we investigated the impact of the number of HuTu80 cell passages on HuSaV growth. For this, we used cells at passage 40 or 141 in cultures supplemented with GlyCA for propagating HuSaV GI.1 (AK20 strain) and GII.3 (AK11 strain) P1 virus (23), as assessed by RT-qPCR and antigen (Ag)-ELISA.

In HuTu80 cells at passage 40, the amount of HuSaV GI.1 viral RNA was fivefold and ninefold higher at 5- and 10-days post-infection (dpi), respectively, than at 1 dpi, whereas RNA of HuSaV GII.3 at 5 and 10 dpi was 1.5 × 10⁷ and 2.2 × 10⁹ copies/mL, 130- and a 20,000-fold, respectively, compared with that at 1 dpi. In contrast, both GI.1 and GII.3 viral RNA levels increased to 3.4 × 10¹⁰ and 4.4 × 10¹⁰ (170,000-fold and 220,000-fold increase) and 8.3 × 10⁹ and 5.2 × 10¹⁰ copies/mL (17,000-fold and 110,000 -fold increase) at 5 and 10 dpi, respectively, in HuTu80 cells at passage 141 (Fig. 1A).

Fig 1.

HuSaV propagates more efficiently in HuTu80 cells from cultures with high passage number. P1 HuSaV stocks of GI.1-AK20 or GII.3-AK11 (2 × 10⁵ copies of viral RNA per well) were independently inoculated onto HuTu80 cells at passage 40 (P40) or 141 (P141). The culture medium was supplemented with sodium glycocholate (GlyCA). (A) HuSaV RNA levels and (B) HuSaV VP1 protein levels at 1, 5, or 10 days post-inoculation in the same cell culture supernatant are shown. Data represent the mean of two infection experiments. Error bars indicate standard deviation. The number above the bars indicates the mean fold-change in viral genome copy numbers between the two time points. The dotted line indicates the lowest quantification limit.

Next, both GI.1 and GII.3 were found to exhibit an increase in VP1 levels at 5 and 10 dpi compared with 1 dpi in the antigen (Ag) ELISA using the same supernatant from HuTu80 cells at passage 141, with an optical density (OD) value of 0.9 to 1.0 (Fig. 1B). However, VP1 protein was barely detectable (OD value ≤0.1) in the same cell supernatant at passage 40 (Fig. 1B).

These results showed that the number of HuTu80 cell passages influenced the propagation of HuSaV for the GI.1 and GII.3 strains, particularly when determined by Ag ELISA.

Taurocholate is more effective than glycocholate for the replication of human sapoviruses in feces

To obtain more diverse HuSaV strains, viral replication in high-passage number HuTu80 cells was examined using 16 genotypes (GI.1–7, GII.1–5, GII.7, GIV.1, GV.1, and GV.2) in human feces (Table 1). New GI.1-, GI.2-, and GII.3-positive feces were used in the present study, different from the sources used in the previous study (23). For 11 genotypic HuSaV strains, adding GlyCA increased the viral RNA levels from 1.1- to 4.8 log₁₀-fold. When we used taurocholate (TauCA) as an alternative conjugated bile acid, which is abundant in the duodenum, 2.4- to 6.4 log₁₀-fold viral RNA increases of 14 (GI.1–7, GII.1–5, and GV.1–2) strains of different genotypes were detected. Thus, the success rate of viral propagation from the feces was greater for TauCA than GlyCA, with the final viral RNA levels for the former ranging from 5.7 × 10⁸ to 2.3 × 10¹¹ copies/mL, whereas for the latter, these ranged from 3.8 × 10⁶ to 1.4 × 10¹¹ copies/mL (Table 2).

TABLE 1.

List of fecal suspensions tested for infection on HuTu80 cells

Sapovirus genotype	Strain	Viral RNA titer (copies/mL of fecal filtrate)	Fecal specimens collected (year and month)	Patient age (year [yr], month [mo])	Reference
GI.1	FS40	2.8 × 107	2016 Apr	1 yr 8 mo	This study
GI.2	FS25	6.4 × 106	2017 Apr	1 yr	This study
GI.3	D1736-A (D1736)	6.7 × 109	2008 Jun	40 yr	(38, 39)
GI.4	Chiba000496	1.1 × 1010	2000 Mar	5 yr	(39, 40)
GI.5	D1729-A (D1729)	2.1 × 1010	2008 Jun	31 yr	(38)
GI.6	Chiba000764	2.6 × 1010	2000 May	1 yr	(39, 40)
GI.7	D1714-B (D1714)	4.0 × 108	2008 Jun	58 yr	(38, 39)
GII.1	Kumamoto129	2.2 × 109	2010 Jun	8 yr 4 mo	(41)
GII.2	Kumamoto130	1.3 × 107	2010 Jul	0 yr 8 mo	(41)
GII.3	FS130	1.0 × 107	2017 May	2 yr 5 mo	This study
GII.4	D1739-A (D1739)	2.4 × 108	2008 Jun	27 yr	(38)
GII.5	Kashiwa1	3.3 × 108	2010 Apr	72 yr	(5)
GII.7	20072248 (Kumamoto82)	4.0 × 107	2008 Feb	2 yr 4 mo	(39, 41)
GII.8	AK764	2.9 × 107	2017 Nov	7 yr 8 mo	This study
GIV.1	Nagano10-2	1.2 × 1010	2007 Jul	16 yr	(42)
GV.1	D3302	1.1 × 1010	2018 Nov	Junior high school student	This study
GV.2	NGY1	4.1 × 1010	2012 Apr	Adult	(4)

TABLE 2.

Human sapovirus viral loads in initial isolation and scale-up serial passages

Sapovirus genotype	Strain	Viral RNA (copies/mL) in the cell culture supernatant
		P0 (1 mL)		P1 (7 mL)	P2 (360 mL)
		GlyCA (log10 fold change level)	TauCA (log10 fold change level)	TauCA	TauCA
GI.1	FS40	1.4 × 1011 (3.8)	1.5 × 1011 (4.1)	4.5 × 1010	N.T.a
GI.2	FS25	9.9 × 109 (3.2)	7.5 × 109 (3.2)	2.5 × 1011	1.2 × 1010
GI.3	D1736	1.2 × 1011 (4.8)	2.3 × 1011 (5.3)	1.0 × 109	4.9 × 109
GI.4	Chiba000496	4.5 × 106 (1.5)	4.2 × 109 (4.5)	7.8 × 109	1.0 × 1010
GI.5	D1729	1.3 × 1011 (4.7)	1.7 × 1011 (4.8)	2.2 × 109	N.T.
GI.6	Chiba000764	1.4 × 1011 (4.4)	1.5 × 1011 (4.7)	3.4 × 1011	N.T.
GI.7	D1714	4.0 × 104 (0.0)	6.4 × 1010 (6.4)	2.6 × 1010	1.3 × 1010
GII.1	Kumamoto129	3.8 × 106 (1.2)	4.2 × 109 (3.2)	5.0 × 109	1.3 × 1010
GII.2	Kumamoto130	2.8 × 107 (1.1)	5.7 × 108 (2.4)	6.4 × 1010	N.T.
GII.3	FS130	2.4 × 1010 (3.2)	5.7 × 1010 (4.0)	1.0 × 1011	N.T.
GII.4	D1739	9.3 × 104 (0.0)	1.2 × 109 (4.1)	1.3 × 1010	5.0 × 109
GII.5	Kashiwa1	3.6 × 104 (0.0)	3.3 × 1010 (5.3)	1.7 × 1010	3.1 × 109
GII.7	Kumamoto82	4.5 × 104 (0.0)	4.6 × 104 (0.1)	N.T.	N.T.
GII.8	AK764	N.T.	9.1 × 107 (2.8)	6.0 × 109	1.4 × 109
GIV.1	Nagano10-2	6.7 × 105 (0.0)	7.3 × 105 (0.0)	N.T.	N.T.
GV.1	D3302	3.0 × 109 (2.4)	2.7 × 1010 (3.5)	1.3 × 1010	N.T.
GV.2	NGY1	8.6 × 108 (1.7)	6.8 × 1010 (4.1)	2.8 × 109	3.2 × 109

^a N.T., not tested.

GII.8-positive fecal specimens were subsequently identified during these trials. Viral growth was analyzed only with TauCA, resulting in an increase in viral RNA levels by 2.8 log₁₀, with the final viral RNA level reaching 9.1 × 10⁷ copies/mL (Table 2).

Therefore, a total of 15 HuSaV genotypes (GI.1–7, GII.1–5, GII.8, and GV.1–2) were propagated from virus-positive feces in HuTu80 cells; supplementation with TauCA resulted in a higher success rate than supplementation with GlyCA, as well as higher viral RNA levels.

Serial passage for viral propagation and scaling up

To obtain significant amounts of the viral stock of the 15 HuSaV strains, P0 viruses (grown from feces in 12-well plates) were inoculated onto new confluent high-passage number HuTu80 cells using TauCA-supplemented medium (~7 mL), and P1 viruses with final viral RNA levels ranging from 1.0 × 10⁹ to 3.4 × 10¹¹ viral genomic copies/mL were obtained (Table 2).

For nine of these genotypes (GI.2, GI.3, GI.4, GI.7, GII.1, GII.4, GII.5, GII.8, and GV.2), P1 viruses (grown in T75 flasks) were further inoculated onto confluent HuTu80 cells, and large volumes (approximately 360 mL) of P2 viruses (1.4 × 10⁹–1.3 × 10¹⁰ viral genomic copies/mL) (Table 2) were obtained. Virions were purified by ultracentrifugation, and particles with a diameter of approximately 40 nm having a characteristic sapovirus surface morphology were observed by transmission electron microscopy (Fig. 2). These preparations were then used as immunogens for antiserum production.

Fig 2.

Transmission electron micrographs of the purified human sapoviruses. Ultracentrifuge-purified P2 HuSaV particles of GI.2, GI.3, GI.4, GI.7, GII.1, GII.4, GII.5, GII.8, and GV.2. The scale bar indicates 50 nm.

Thus, all 15 genotypes of the HuSaV strains tested were capable of being serially passaged in HuTu80 cells at the scale in culture, and virus particles were visually characterized for nine of them.

Minimal or no viral amino acid changes occur during serial passaging

We investigated whether viral genome sequence changes occurred during scaled-up viral passaging. Analysis of the protein-encoding sequences revealed mutations in all but the NS5 and VP2 proteins. The hotspot mutation site was observed in VP1 (Table 3). Compared to the GI and GV strains, GII strains had fewer predicted amino acid changes, and four genotypes (GII.1, GII.2, GII.5, and GII.8) showed no sequence changes at all compared to the original fecal samples. The GV.2 strain showed the largest sequence change, which involved six amino acid residues (four in the NSs and two in VP1), three in the passage 1 (P1) virus, and three more in the passage 2 (P2) virus. Strain GI.5 was predicted to have VP1 shortened by one residue by replacing the C-terminal amino acid with a stop codon (Table 3).

TABLE 3.

Nucleotide and deduced amino acid changes in human sapovirus genomes during scale-up passages

Sapovirus genotype	Strain	Accession number	Sequenced samples	Nonstructural proteins (NS)	VP1	VP2	Summary of amino acid changes
GI.1	FS40	LC788817a	P1 only				Not comparable
GI.2	FS25	LC788818a	P1 and P2	NS2-K195E (AAA→GAA) P2			NS2 in P2
GI.3	D1736	AB522396	original, P1, and P2	NS2-T96A (ACA→GCA) P1 NS2-A209G (GCA→GGA) P1 NS7-C1555b (TGC→TGT) P1	P2 domain-D2111H (GAT→CAT) P2		NS2 in P1 and VP1 in P2
GI.4	Chiba000496	AJ606693	Original, P1, and P2		P2 domain-S2088P (TCA→CCA) P1 T2163b (ACC→ACT) P1 P1 domain-R2277S (AGA→AGT) P1		VP1. P2 was identical to P1
GI.5	D1729	AB522393	Original and P1	NS4-I792T (ATC→ACC) P1 NS4-L903b (CTC→CTT) P1 NS7-P1417b (CCC→CCT) P1	P1 domain-Q2286 stop (CAA-TAA) P1		NS4 and VP1. P2 was identical to P1
GI.6	Chiba000764	AJ606694	Original and P1	NS7-Y1276b (TAT→TAC) P1 NS7-V1407I (GTC→ATC) P1	P1 domain-R2280S (AGA→AGT) P1		NS7 and VP1 in P1
GI.7	D1714	AB522390	Original, P1, and P2	NS2-K168E (AAG→GAG) P1			NS2. P2 was identical to P1
GII.1	Kumamoto129	AB689845	Original, P1, and P2				Not changed
GII.2	Kumamoto130	AB689846	Original and P1				Not changed
GII.3	FS130	LC788819a	P1 only				Not comparable
GII.4	D1739	AB522399	Original, P1, and P2		Shell domain-S1923P (TCG→CCG) P2		VP1 in P2
GII.5	Kashiwa1	LC190463	Original, P1, and P2	NS1-P60b (CCG→CCT) P1	F1923b (TTT→TTC) P2		Not changed
GII.8	AK764	LC787637a	Original, P1, and P2				Not changed
GV.1	D3302	LC787639	Original and P1	NS4-I800T (ATA→ACA) P1			NS4 in P1
GV.2	NGY1	AB775659	Original, P1, and P2	NS2-G76D (GGT→GAT) P1 NS3-T558A (ACA→GCA) P2 NS4-R677C (CGT→TGT) P1 NS6-D1115N (GAC→AAC) P2	P1 domain-E2231D (GAG→GAT) P1 P2 domain-V2099A (GTT→GCT) P2		NS2,4 and VP1 in P1 and further in NS3, NS6, and VP1 in P2

^a Sequence from P1 virus.

^b Synonymous nucleotide changes.

Our findings indicate that HuSaVs cultured using HuTu80 cells exhibit minimal or no amino acid changes during serial passage at the scale. No common shared mutations were observed.

Conjugated bile acids are required for propagation of all isolated human sapoviruses

Regarding the bile acid requirement for HuSaV growth, we confirmed that viruses could not be propagated at P1 under culture conditions without bile acids for any of the 15 genotypes (we tested bile acid conditions only in initial infection trials with virus-positive feces, see Table 2). The GI.7 and GII.5 strains initially grew (as judged by RT-qPCR) only with TauCA when examined directly from feces (Table 2), but viral growth was confirmed with both GlyCA and TauCA when HuSaV P1 stock was used. However, the P1 stock of GII.4 grew only in TauCA-supplemented medium, similar to the growth from feces (Table 2; Fig. 3). In the absence of bile acids, the signals at the VP1 level were low (OD values ≤ 0.1).

Fig 3.

Glycocholate and taurocholate requirements for human sapovirus propagation. Fifteen genotypes of human sapovirus P1 stocks (2 × 10⁶ copies of viral RNA per well) newly obtained in this study were inoculated onto HuTu80 cells with GlyCA, sodium taurocholate (TauCA), or without bile acids. HuSaV VP1 levels at 10 days post-inoculation were detected by Ag ELISA. Data represent the mean of two infection experiments, and error bars represent standard deviations.

These results indicate that the addition of conjugated bile acids is required for all 15 HuSaV genotypes grown in high-passage number HuTu80 cells and that the bile acid requirements for viral propagation are stricter for the GII.4 strain than for the other strains evaluated.

The immunogenicity of different human sapovirus genotypes differs

HuSaV genotypes were classified based on their VP1 sequences. Differences in immunogenicity among these HuSaV strains propagated in high-passage number HuTu80 cells were examined using antigen (Ag-)ELISAs. The VP1 amino acid lengths tested for the P1 viruses were 556–571 aa. Shared sequences among the genogroups were 45.1%–52.1%, and 76.8%–87.9% for GI, 70.6-85.8% for GII, and 88.2% for GV within the same genogroup (Fig. 4A).

Fig 4.

VP1 amino acid sequence identity and antigenic diversity among 15 human sapoviruses propagated in HuTu80 cells. Percent pairwise sequence identity of the VP1 protein among the strains of P1 stock are represented by a gray gradient in (A). The summary of Ag ELISA OD values for the strains with diluted P1 stock is shown in (B). The OD values for the strains with >0.2 are represented by a gray gradient.

As shown in Fig. 4B, comparing the HuSaV P1 virus stocks showed that the OD values of Ag ELISA of the 15 genotypes were 0.0–1.3. The OD signal was 0.0–0.1 for different genogroup strains, whereas for different genotypes within the same genogroup, it ranged from 0.0–1.0. Ag ELISA results for GI.6, GII.2, GII.3, and GII.5 indicated no obvious cross-reactivity to other genotypes in the same genogroup (max OD was 0.1). However, cross-reactivity with various signal intensities for the tested GI genotypes other than GI.6 was observed, especially in GI.2 (OD 0.3–0.9), GI.5 (OD 0.2–0.5), and GI.7 (OD 0.2–0.8) Ag ELISAs. Similarly, GII.4 Ag ELISA showed cross-reactivity with all the other tested GII genotypes (OD 0.2–0.9). Cross-reactivity reflected by OD signals (0.3 or 0.9) was also detected between the two GV strains of different genotypes using the GV Ag ELISA. Two-way or one-way cross-reactivity signals observed in GI, GII, and GV Ag ELISAs within the same genogroup did not simply correlate with VP1 aa identity and viral RNA copy number (Fig. 4A and B). However, in combination with antisera against other genotypes, none of the strains showed the same OD values as in the Ag ELISA against the respective genotypes (Fig. 4B).

Combining these results, we concluded that the immunogenicity of the propagated HuSaV strains varied among the different genotypes.

DISCUSSION

In this study, we propagated 15 HuSaV genotypes using HuTu80 cells and acquired sufficient stocks of these viruses to enable further work. Two factors contributed to the success of this endeavor, the first being the high passage number of the HuTu80 cells, as shown in Fig. 1. The results were reproduced in another laboratory using the HuSaV GI.1-AK20 stock. Viral RNA levels increased when using HuTu80 cells after >80 passages but not at <40 passages (Dr. Nobutaka Shirasaki, the corresponding author of [ref (27)], personal communication). By Ag using ELISA, we were unable to detect substantial VP1 protein signals in the cell culture supernatants from GI.1 and GII.3 HuSaV using HuTu80 at passage 40. In contrast, significant viral RNA levels of GII.3 were detected at passages 40 and 141 (2.2 × 10⁹ and 5.2 × 10¹⁰ viral genome copies/mL, respectively) (Fig. 1). We confirmed the infectivity of the progeny viruses by infecting new high-passage-numbered HuTu80 cells with 10 dpi cell culture supernatants from GII.3-infected P40 and P141 cells. We found similarly increased VP1 protein levels (OD value of 1.1) at 10 dpi when 2 × 10⁷ copies per well of P40 and 2 × 10⁵ copies per well of P141, respectively, were used (data not shown). These results indicate that an increase in viral nucleic acids does not always reflect levels of infectious progeny virus produced, and our study data suggest that higher passage cells produce more infectious progeny virus; however, the mechanism of virion assembly in HuSaV remains unknown. Nonetheless, this is an important practical finding when preparing progeny virus stocks as a resource.

The second key factor for the efficient propagation of the 15 HuSaV genotypes was the introduction of a new conjugated bile acid, TauCA into the cultures. TauCA was selected because of its similar structural feature and physiological abundance (i.e., GlyCA and TauCA, which are formed by the addition of glycine and taurine to cholic acid, respectively, are produced at a 3:1 ratio in humans (24)), and its reduced toxicity to HuTu80 cells. Therefore, we could add 2 mM TauCA at twice the concentration of GlyCA. This resulted in a higher success rate of viral replication/propagation from virus-positive feces and higher viral RNA levels in the P0 virus stock compared to the use of GlyCA.

We further confirmed that passage number 113 and 226 HuTu80 cells supported the propagation of HuSaV GI.1-AK20 and GII.3-AK11 strains at the protein level supplemented with GlyCA or TauCA, but not in passage number 47 HuTu80 cells (Fig. S1A). The preliminary Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis among mock-infected controls with these three different passage number HuTu80 cells supplemented with GlyCA showed that six pathways, namely, neuroactive ligand–receptor interaction, cell adhesion molecules, pathway in cancer, PI3-Akt signaling pathway, ECM–receptor interaction, and proteoglycans in cancer, categorized as Environmental Information Processing and Human Diseases, by KEGG, were all enriched among the three different passage number cell combinations (Fig. S1B). These results indicated that higher-passage number HuTu80 cells had a distinct gene expression profile compared with the low-passage number cells and resulted in a more suitable phenotype for HuSaV infection/replication/growth. We confirmed that our high passage number of cells (passage 226) were actually HuTu80 cells using the cell line authentication service by ATCC (data not shown). In the future, it will be interesting to identify the critical cellular pathways and factors involved in HuSaV infection and growth.

The failure of GII.7 and GIV.1 strain propagation (Table 2) may have simply been due to the fecal specimens themselves, the conditions for virus infection, or growth conditions of GII.7 and GIV.1, which differ from those of other genotypes (e.g., different susceptible cells or different bile acids essential for propagation). Additional tests on other GIV.1-positive fecal specimens from the same outbreak (42) did not show increased HuSaV viral RNA levels in HuTu80 cells (data not shown). Euller-Nicolas et al. also reported the failure of GIV.1 to replicate in an intestinal enteroid system (28). In contrast, Matsumoto et al. reported a low but detectable increase of RNA levels (~13.9 fold) in GIV.1-positive stool specimens using iPS-derived intestinal epithelial cells (29).

As feces positive for the GII.6 strain were not available, we performed in vitro synthesized capped RNA transfection-based reverse genetics for HuSaV GII.6 SaKaeo-15 (GenBank accession number AY646855) as described (26), with GlyCA or TauCA supplementation. However, the virus could not be propagated (as detected by RT-qPCR) even after repeated blind passages of the transfected cell culture supernatant (data not shown).

In this study, the addition of bile acids (GlyCA or TauCA) was shown to be essential for HuSaV propagation, not only directly from virus-positive feces but also for P1 virus. HuSaV strain GII.4 replicated only with the addition of TauCA in both feces (P0) and P1 (propagated from P0) samples (Table 2; Fig. 3). Such different bile acid requirements are interesting from a physiological perspective for future studies on the mechanism of infection because these conjugated bile acids are abundant in bile secreted in the small intestine, especially in the upstream duodenum, in the human intestinal tract (24, 25).However, in contrast to our study results and those reported by Euller-Nicolas et al. (28), the results obtained by Matsumoto et al. indicated that addition of bile is not always necessary for HuSaV replication in iPS-derived intestinal epithelial cells (29).

The results of this study carry some public health implications. The HuSaV virus-positive fecal suspensions used in this study were collected from donors of various ages (Table 1), and it was demonstrated that the virus remained infectious for long periods (>20 years) when kept frozen. To the best of our knowledge, the present report also describes the first instance of infective HuSaV in an individual without acute gastroenteritis; the GII.5 strain was isolated from the feces of an elderly (72 year old) asymptomatic food preparation worker during a suspected foodborne outbreak (Table 1) (5). The HuSaV GI.7- and GII.4- positive feces used in this study were co-positive for norovirus GII.6 (1.3 × 10⁸ copies/g stool) and norovirus GI.1 (1.2 × 10⁷ copies/g stool), respectively, both of which were from patients during a gastroenteritis outbreak caused by consuming undercooked clams (38, 43), but norovirus was not detected by next-generation sequencing (NGS) analysis of any tested HuSaV stock (data not shown). This suggested that norovirus does not replicate under the culture conditions used. This finding is consistent with a previous report that norovirus does not propagate in HuTu80 cells (44).

In the case of porcine sapovirus, the GIII.1 TC-Cowden strain, which can be propagated in cell cultures supplemented with bile acid (15, 45, 46), multiple conserved mutations in the nonstructural proteins NS7 (polymerase) and VP1 (15, 47) were noted. In contrast, the HuSaVs propagated in the present study exhibited no common mutations, suggesting that they were not adaptogenic mutations essential for viral growth. In addition, GII.1, GII.5, and GII.8 HuSaVs did not mutate at all from the original sequence throughout the comparable viral protein-coding region sequences even after P1 and/or P2 virus serial passaging with scale-up once and/or twice (Table 3). This further supports the utility of this approach. Amino acid changes in VP1 were most frequent in protruding domains 1 and 2 (both at P1 and P2), and less frequent in the shell domain (16, 17) (Table 3). The VP1 C-terminus amino acid residue (Q) in the GI.5 strain was truncated in the serially passaged strain despite being highly conserved among GI strains (15), suggesting that this residue is not essential for viral growth.

In the present study, we verified antigenic differences between 15 genotypes of HuSaV strains propagated in HuTu80 cells by Ag ELISA using antisera raised against the corresponding genotype virus-like particles (VLPs) or purified viruses. This enabled us to confirm that different genotypes have distinct immunogenic properties. Some of these differences among genogroups and genotypes of HuSaVs have been previously reported for GI.1, GI.5, GII.3, GIV.1, and GV.1 VLPs expressed in insect cells and GII.2 and GII.3 VLPs expressed in mammalian cells using Ag ELISA with rabbit and guinea pig antisera raised against these VLPs (32, 33, 48). We confirmed these findings from earlier research, although Ag-ELISA for GI.5 showed one-way cross-reactive signals for the HuSaV GI.1 P1 passage virus (OD 0.4, Fig. 4B).

HuSaV genotyping was based on the VP1 sequence (9, 49, 50). In this study, we demonstrated that current genotyping reflects different antigenicities using 15 cell culture-propagated strains of different genotypes. It was recently reported that a mouse antiserum prepared against the VLP of the HuSaV GII.1 strain failed to block infection and replication of the HuSaV GII.3 strain (29). Such infection inhibition experiments with panels of isolated HuSaVs as well as antisera or monoclonal antibodies raised against them will be useful for determining the neutralizing ability and relevant epitopes in the future.

This study allowed for the propagation of more strains of different genotypes (15 of the 18 HuSaV genotypes categorized as GI, GII, and GV) than in a previous study (23), thereby securing viral stocks and preparing specific antisera for each genotype. This accomplishment led to the creation of experimental viral infection systems for various HuSaV strains of different genotypes with diverse antigenic properties, without requiring human feces. The utilization of HuTu80 cells highly susceptible to multiple HuSaV strains will accelerate the identification of genotype-specific or common factors essential for viral infection and propagation. This should also help identify conditions for disinfecting viral contamination and assist in developing agents to prevent HuSaV infection and propagation, as well as assisting in the implementation of neutralization assays for future vaccine design purposes.

MATERIALS AND METHODS

Cells

HuTu80 cells (ATCC #HTB-40) were maintained and in T25 or T75 flasks with Iscove’s modified Dulbecco’s medium (IMDM) (Sigma, #I3390), supplemented with 2 mM glutamine (Gibco), 3% heat-inactivated fetal bovine serum (FBS) (Biosera, Kansas City, MO, USA), and antibiotics (100 U/mL penicillin and 100 µg/mL streptomycin (Gibco)) at 37°C in a humidified 5% CO₂ atmosphere. The cells were passaged once a week by using a splitting ratio of 1:10. HuTu80 cells at passage numbers 40, 47, or 113–226 were used in this study.

Bile acids

Sodium GlyCA or sodium TauCA, both purchased from Nacalai Tesque (Kyoto, Japan), were added as supplements to the HuSaV cultures. Consistent with previous studies (23), the optimal concentration of TauCA was determined to be the highest tested; no cytotoxicity was detected, as observed by microscopy.

Viruses and fecal specimens

Human sapovirus (HuSaV) P1 stocks of GI.1 AK20 and GII.3 AK11 (GenBank accession numbers LC715151 and LC715150, respectively) (23) and HuSaV-positive fecal suspensions positive for one of the 17 genotypes (GI.1–7, GII.1–5, and GII.7, GII.8, GIV.1, GV.1, and GV.2) (viral RNA levels of 6.4 × 10⁶ to 4.1 × 10¹⁰ /mL of stool suspension) (Table 1) were used for viral growth studies. These fecal specimens were collected from individuals 8 months to 72 years of age between the years 2000 and 2018 (4, 5, 38–42, 49) and stored in a freezer. Nine samples collected from patients <9 years of age were from sporadic infection cases, and the remaining eight samples were from cases related to outbreaks (Table 1). Fecal sample collection and pathogen screening were performed at regional public health institutes. Virus isolation from fecal specimens was performed at the National Institute of Infectious Diseases with the approval of the Institutional Ethics Committee. In the case of an outbreak due to a single genotype, the fecal suspension with the highest viral RNA amount was chosen for propagation.

Assessment of the effect of the number of HuTu80 cell passages

To assess the impact of host cell passage number on the propagation of HuSaVs, a confluent monolayer of HuTu80 cells (approximately 1 × 10⁶ cells) was utilized (at passage numbers 40 or 141 or another set at passage numbers 47, 113, or 226 grown in a 12-well plate). The culture medium was replaced with 0.5 mL of a medium supplemented with 1 mM GlyCA or 2 mM TauCA (virus growth medium) before inoculation. Subsequently, 20 µL of the diluted P1 viral stock of GI.1 AK20 or GII.3 AK11 (23) (approximately 2 × 10⁵ viral RNA copies per well) was added to each well. The plates were then incubated overnight at 37°C in 5% CO₂ and air. The cell monolayers were washed twice with Hanks' balanced salt solution without Mg²⁺ and Ca²⁺ (HBSS (-) ([FUJIFILM Wako Pure Chemical Corporation], after which the virus growth medium (1.0 mL) was added to each well, and the cells were incubated for 10 days. The supernatants of passage numbers 40 or 141, collected at 1, 5, and 10 dpi, were analyzed using RT-qPCR to detect HuSaV genomic RNA (7) and Ag ELISA to detect the HuSaV VP1 protein, as described previously (23). The supernatants of passage numbers 47, 113, and 226 without inoculated virus washing at 1, 5, and 10 dpi were analyzed using Ag ELISA to detect the HuSaV VP1 protein.

Initial viral propagation trials with conjugated bile acid supplementation

Confluent HuTu80 cell monolayers (passage number >120), grown in 12-well plates, were inoculated with HuSaV. Before inoculation, the culture medium was replaced with 0.5 mL of the virus growth medium (cell culture medium supplemented with GlyCA at 1 mM or TauCA at 2 mM). HuSaV-positive fecal suspensions diluted in minimum essential medium with Earle’s salts and 0.05% sodium bicarbonate supplemented with 0.5% lactalbumin, 0.2% bovine serum albumin, and antibiotics) were sterilized using a 0.22-µm centrifuge filter, and 20 µL (6 × 10⁴–8 × 10⁶ copies of viral RNA) was added to each well. The plates were then incubated overnight at 37°C with 5% CO₂. The cell monolayers were washed twice with HBSS (−). Finally, the virus growth medium (1.0 mL) was added to each well. These cultures were incubated for 7–10 days, and HuSaV RNA levels in the cell-free supernatants (P0: the initial passage of the virus in cell culture supernatants) were quantified using RT-qPCR (7).

Preparation of P1 virus stock

To prepare passage 1 (P1) HuSaV stock, the P0 virus stock was inoculated onto confluent monolayers of HuTu80 cells (passage number >130) grown in T75 flasks. Before inoculation, the culture medium was replaced with approximately 7 mL/flask virus growth medium containing 2 mM TauCA. An inoculum (~5 × 10⁶ copies of viral RNA/flask) was added and flasks incubated overnight. The cells were subsequently washed twice with HBSS (–), and the virus growth medium was added (~10 mL/flask). The supernatants were collected at 7–10 dpi, clarified by centrifugation at 10,000 × g for 30 minutes, and filtered through a 0.22-µm syringe filter.

Scaled-up further serial passages for virion purification

To purify virions for use as immunogens, P2 passaged viruses of nine HuSaV genotypes (GI.2, GI.3, GI.4, GI.7, GII.1, GII.4, GII.5, GII.8, and GV.2) were obtained by inoculating P1 virus stock onto confluent monolayers of HuTu80 cells (passage number >140) grown in 12–14 T225 flasks. Before inoculation, the culture medium was replaced with 26–30 mL virus growth medium containing 2 mM TauCA per T225 flask. Five hundred microliters of the inoculum (containing 5 × 10⁶ copies of viral RNA/flask) was added to each flask, and the cells were cultured for 7–11 days. HuSaV virions were purified from cell culture supernatants collected from the infected cells by centrifugation at 10,000 × g for 60 minutes. The supernatants were concentrated by sucrose-cushioned ultracentrifugation at 100,000 × g for 3 hours using a Beckman SW 32 Ti rotor. The resulting pellet was resuspended in phosphate-buffered saline (PBS) overnight at 4°C. For CsCl gradient centrifugation, 4.5 mL of the sample was mixed with 2.1 g of CsCl and centrifuged at 100,000 × g for 24 hours at 10°C using a Beckman SW55Ti rotor. The visible band corresponding to the virion was collected by using a syringe directly from the side of the centrifuge tube, diluted with PBS (–), centrifuged for 2 hours at 154,000 × g using a Beckman TLA55 rotor, and the pellet was resuspended in PBS (–). The purified viral particles were placed on a formvar and carbon-coated grid for 45 seconds, rinsed with distilled water, and stained with a 2% uranyl acetate solution. The specimens were observed using a transmission electron microscope (HT7700; Hitachi High Technologies, Tokyo, Japan) at 80 kV. The purified viruses were subjected to SDS-PAGE and Coomassie Brilliant Blue staining, as previously described (26).

Confirmation of the requirement for conjugated bile acid supplementation for viral propagation

To confirm the requirement for bile acids for HuSaV propagation, a confluent monolayer of HuTu80 cells (passage number >200) in 24-well plates was used. The culture medium was replaced with 0.5 mL of a medium supplemented with 1 mM GlyCA, 2 mM TauCA, or without bile acids before inoculation. Subsequently, 20 µL of each diluted P1 viral stock of the newly propagated 15 strains of different genotypes (approximately 2 × 10⁶ copies of viral RNA per well) listed in Table 2 and Fig. 4A and B was added. The cells were then incubated for 10 days without washing. The supernatants collected at 10 dpi were analyzed using Ag ELISA for each genotype to detect the HuSaV VP1 protein.

RNA extraction and RT-qPCR

Viral RNA was extracted from 100 µL of fecal suspensions or cell culture supernatants using High-Pure RNA Isolation Kits (Roche Diagnostics, Mannheim, Germany) following the manufacturer’s instructions. cDNA was synthesized as follows: 5 µL of RNA samples was mixed with a 5 µL mixture including 50 pmol random hexamer (Takara, Shiga, Japan), 1 mM dNTPs, 20 units of recombinant RNase inhibitor (Takara), and 30 units of ReverTra Ace (Toyobo, Osaka, Japan). Reactions were performed at 30°C for 10 minutes, 42°C for 30 minutes, and 95°C for 5 minutes. qPCR was performed using 2.5 µL of cDNA with mixed forward primers (HuSaV-F1:5′-GGCHCTYGCCACCTAYAAYG-3, HuSaV-F2:5′-GACCARGCHCTCGCYACCTAYGA-3′, and HuSaV-F3:5′-GCWRYKGCWTGYTAYAACAGC-3), reverse primer (HuSaV-R: 5′-CCYTCCATYTCAAACACTA-3′), and probe (HuSaV-TPa:, 5′-FAM‐CCNCCWATRWACCA‐MGB‐NFQ-3′), which can detect all the human genotypes (7), using the QuantiTect Probe PCR Master Mix (Qiagen, Hilden, Germany) under the following conditions: 95°C for 15 minutes followed by 45 cycles of a two-step PCR (95°C for 15 seconds and 60°C for 60 seconds), as described previously (7). The amplification data were collected and analyzed using 7500 Real-Time software v2.0.6 (Applied Biosystems). A tenfold serial dilution from 2.5 × 10⁸ to 2.5 × 10¹ copies per well of the plasmid containing GII.2 (AY237420) sapovirus target region sequences (51) was used to create a standard curve to calculate viral genomic RNA copy numbers.

Viral genome sequencing and analysis

The genomic sequence of HuSaV from nucleic acids extracted from fecal specimens or cell culture supernatants was determined using NGS. Library preparation and Illumina MiSeq sequencing were conducted following previously established protocols (52). Briefly, individual specimens were used for library construction employing a NEBNext Ultra II RNA Library Prep Kit for Illumina (New England Biolabs, Ipswich, MA, USA), incorporating barcoded adapters to generate a 200-bp fragment library. The resulting libraries were purified using AMPure XP magnetic beads (Beckman Coulter, Brea, CA, USA) and were assessed for quality on the 4150 TapeStation System (Agilent Technologies) with D1000 ScreenTape. Nucleotide sequencing was performed on an Illumina MiSeq sequencer (Illumina, San Francisco, CA, USA) with a MiSeq Reagent Kit v2 (Illumina) to generate 151-bp paired-end reads. Data analysis was carried out with CLC Genomics Workbench v22 (QIAGEN). The complete or nearly complete nucleotide sequence of HuSaV strains was obtained by de novo assembly and mapping reads to the reference in CLC Genomics Workbench.

Antisera

Rabbit or guinea pig antisera against GI.1 (AY237422), GI.5 (AB253740), GII.2 (AY237420), GII.3 (AY603425), and GV.1 (AY646856) VLPs were previously reported (32, 33), Antisera against GI.6 (AB455803) VLPs (36) were generated anew. These VLPs had 98.6%–99.8% aa identity (1–8 aa differences) with the VP1 aa of the corresponding genotype propagated in this study (data not shown). The VP1 aa identity of each strain was calculated using Genetyx Mac version 21.2.2 with MAFFT alignment, and then “% identity Matrix” with the parameter of “Gaps are takes into account” was used. In addition, for GI.2, GI.3, GI.4, GI.7, GII.1, GII.4, GII.5. GII.8, and GV.2, purified virions (10^9-10 viral RNA copies/rabbit or guinea pig) were used as immunogens for antiserum preparation by at least three subcutaneous immunizations every 2 weeks. Complete Freund’s adjuvant was used for the initial immunization and incomplete Freund’s adjuvant for subsequent boosting.

Antigen ELISA for the detection of HuSaV capsid VP1 protein

Detection of VP1 by Ag ELISA for each genotype was performed as follows: 96-well microtiter plates (Immulon 2HB Flat Bottom Plates, Code #3455, Thermo) were treated with rabbit hyperimmune antiserum produced against human SaV VLPs or virions at 1:1,000 (GI.2,–3, -4,–5, and −7; GII.1,–4, -5, and –8; GV.1; and GV.2), 1:2,000 (GI.1, GI.6, and GII.2), or 1:5,000 (GII.3) dilutions in 0.05 M carbonate buffer (pH 9.6), left overnight at 4°C, and then blocked with PBS (–) containing 0.25% casein for at least 2 hours at room temperature or overnight at 4°C. The plates were then washed thrice with PBS (–) containing 0.1% Tween 20 (PBS-T), and 50 µL of the tested samples was added to each well. After incubating the cell culture supernatants for 1 hour at room temperature, the plates were washed with PBS-T, and 50 µL of guinea pig hyperimmune antiserum raised against human SaV VLPs or virions was added for 1 hour at room temperature. The dilution for each genotype was the same as for rabbit antisera. The plates were washed, and 50 µL of 1:4,000 HRP-conjugated goat anti-guinea pig IgG (ROCKLAND Code #606–103-129) was added. The plates were incubated for 1 hour at room temperature and then washed with PBS-T before adding 50 µL/well of 1 mM of the substrate 3, 3', 5, 5'-tetramethylbenzidine and 0.01% H₂O₂ in citrate buffer (pH 3.5) for 30 minutes at room temperature. The reaction was stopped with addition of 50 µL/well of 1N H₂SO₄, and the absorbance was then measured at 450 nm, with 750 nm as the reference, by using a microplate reader.

To assess viral replication, 10 µL of the supernatant from cultures at 1 or 10 dpi was used to assess the effect of the number of cell passages, whereas 50 µL from cultures at 10 dpi was used to assess the bile acid requirement for propagation of the 15 genotypes of HuSaVs. For the antigenic reactivity test, 50 µL of diluted P1 stock in the virus growth medium (1.3 × 10⁷–4.3×10⁹ copies of viral RNA per 50 µL) (Fig. 4B), which showed sufficient OD values (1.0–1.3) by the corresponding genotype-targeting Ag ELISA, was used.

Statistical analysis

All statistical analyses were performed on GraphPad Prism 9.5.1. Error bars in the graphs denote the standard deviations.

Cellular gene expression profiling

Total RNA was extracted from confluent monolayers of HuTu80 cells (passage numbers 47, 113, or 226) individually cultured in a T25 flask supplemented with 1 mM GlyCA overnight, lysed in 1.8 mL of TRIzol, and frozen. Total RNA extraction, mRNA-library preparation using the Illumina TruSeq Stranded mRNA Sample Prep Kit (Illumina), NGS sequencing by an Illumina NovaSeq (Illumina), data processing including gene enrichment and functional annotation analysis, and pathway analysis for significant gene list based on gProfiler (https://biit.cs.ut.ee/gprofiler/orth), and KEGG pathway (http://www.genome.jp/kegg/pathway.html), were all outsourced and performed at Macrogen Incorporated (Tokyo, Japan).

DATA AVAILABILITY

The sequences of all human sapovirus strains described in this paper have been deposited in public databases GenBank/DDBJ/EMBL under the accession no. AB522390, AB522393, AB522396, AB522399, AB689845, AB689846, AB775659, AJ606693, AJ606694, LC190463, LC715150, LC715151, LC787637, LC787639, LC788817, LC788818, and LC788819.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/jvi.00639-24.

Fig. S1. jvi.00639-24-s0001.pdf.

Differences of HuSaV propagation and cellular gene pathway expression pattern among different passage numbered HuTu80 cells.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.