INTRODUCTION
Viruses continue to be a global threat to human, animal, and plant health. Although significant advances have been achieved in understanding host-virus interactions, there are still many questions unanswered.
Caenorhabditis elegans is an animal model with many genes and biological processes that are evolutionarily conserved in higher eukaryotes. Thus, many fundamental discoveries in this model organism have been extrapolated into humans, such as apoptosis, RNA interference (RNAi), and microRNAs (
1–3). Likewise, in recent years,
C. elegans has become a model to study virus-host interactions following the discovery of Orsay virus, the first known natural virus of this organism (
4,
5). This
in vivo infection system has enabled the identification of novel host factors required for viral infection (
6–10). Further characterization of host-virus interactions by employing this genetically tractable model provides the opportunity to elucidate underlying mechanisms utilized by viruses to proliferate.
Orsay virus is a nonenveloped, single-stranded, positive-sense RNA virus related to viruses in the family
Nodaviridae (
4). The genome of Orsay virus is composed of two RNA segments, the first of which encodes an RNA-dependent RNA polymerase (RdRp) in the RNA1 segment (~3.4 kb). The RNA2 segment (~2.5 kb) encodes the viral capsid and a capsid-delta fusion protein that is generated by a ribosomal frameshifting mechanism (
11). The Orsay capsid has a T=3 icosahedral symmetry with 60 trimeric surface spikes (
12). In addition, a plasmid-based genetic reverse system was developed by generating transgenic animals harboring the Orsay virus cDNAs (
13). Orsay virus infects primarily intestinal cells, which leads to morphological changes of the intestine including fusion of intestinal cells, induction of vesicles, and disappearance of nuclei (
4,
14). Little is known about the host factors required for Orsay virus infection in
C. elegans. A few genes,
sid-3,
viro-2,
nck-1,
drl-1, and
hipr-1, essential for Orsay virus infection that act on early, prereplication stages of the virus life cycle have been identified (
6,
9,
10).
Viruses depend on host cells to produce viral proteins, replicate their genome, and assemble infectious particles to complete their viral life cycle. The building blocks and energy required for viruses are provided by the host cell. RNA viruses exploit the membranes and intracellular lipids of the host during infection (
15–18). One example is rotavirus, for which it was shown that colocalization of the lipid droplets (LDs) with the replication center and drugs that interfere with LD formation inhibited the viral RNA replication and production of viral progeny (
19). In addition, it has been shown that LDs are important in replication for multiple viruses like hepatitis C virus (HCV) (
20), dengue virus (
21), picornaviruses (
22,
23), noroviruses (
24), SARS-CoV-2 (
25–27), and DNA viruses like Marek’s disease virus (
28). Understanding the mechanisms employed by viruses to disrupt and exploit lipid metabolism may provide means to develop countermeasures against viruses.
The sterol-regulatory-element-binding protein (
srebp) is a transcription factor that belongs to the basic helix-loop-helix leucine zipper family and is important for the homeostasis of lipids in the cell. In mammals, there are two
srebp genes,
srebp1 and
srebp2.
srebp1 is mainly involved in the expression of fatty acid biosynthesis genes whereas
srebp2 is involved in cholesterol biosynthesis (
29). During the maturation of these proteins, the newly synthesized
srebp is located in the endoplasmic reticulum (ER) membrane as a precursor. When specific cellular lipids are low, the protein is transported to the Golgi complex and released by proteases. Then, the mature
srebp is translocated into the nucleus, where it induces the transcription of more than 30 lipogenic genes (
30). As
srebp has a vital role in cellular lipid metabolism, many viruses have subverted this transcription factor according to their needs. For example, it has been described that HCV (
31–34), coxsackievirus B3 (CVB3) (
35,
36), and human cytomegalovirus (
37,
38) promote the accumulation of lipids through induction of the
srebp pathway. Likewise, it has been reported that multiple viruses require lipids for efficient infection: SARS-CoV-2 and Ebola virus require cholesterol for viral entry (
39–41), dengue virus requires triglycerides as an energy source through β-oxidation (
42), and enteroviruses and flaviviruses require phosphatidylinositol-4-phosphate (PI4P) in the replication center for viral RNA replication (
43).
In
C. elegans the ortholog of human
srebp1 is
sbp-1, and there is no ortholog of
srebp2 (
44).
sbp-1 is involved in the regulation of lipogenic enzymes grouped in three main branches that generate the lipid precursors for the cell: the stearoyl-coenzyme A (CoA) desaturases
fat-6/fat-7, the stearoyl-coenzyme A desaturase
fat-5, and the monomethyl branched-chain fatty acids
elo-5/elo-6 (
45). RNA interference knockdown of
sbp-1 or mutations in the
sbp-1 gene lead to animals with low fat stores, high saturated fatty acid content, and reduced expression of lipogenic genes (
44,
46–48). Interestingly, mutation of
sbp-1 in
C. elegans also leads to ~2-fold accumulation of zinc (
49). A genetic suppressor screen identified that mutation in
sur-7, which encodes a member of the cation diffusion facilitator family, reduces the accumulation of zinc in
sbp-1 mutants and also restores lipid levels (
49), suggesting there is a link between zinc and lipid homeostasis. Only limited prior studies have investigated linkages between zinc and lipids. For example, zinc induces lipophagy in primary hepatocytes of yellow catfish (
50), zinc levels are reduced in patients with alcoholic fatty liver disease (
51), zinc supplementation reduces total cholesterol and triglycerides as found in a meta-analysis of 24 studies on humans, and zinc reverses alcoholic steatosis in mice (
52). Related to virus infection, antiviral effects of zinc supplementation against multiple viruses, including herpesviruses (
53), picornaviruses (
54,
55), influenza virus (
56), coronavirus (
57), HCV (
58–60), and HIV (
61,
62), have been described. While different mechanisms have been proposed to explain the putative antiviral effect of zinc, including inhibition of viral protein cleavage and inhibition of viral polymerase activity (
54,
55,
57), experimental data for these models are lacking, and it is not clear how zinc impacts virus infection.
Here, we show in vivo the impact of Orsay virus on lipid homeostasis as well as its dependence on lipid levels. We found that multiple transcription factors involved in lipid homeostasis play roles during Orsay virus infection, including the master regulator srebp-1/sbp-1. The reduced Orsay virus RNA levels in sbp-1 mutant animals could be rescued biochemically by supplementation with specific lipids, genetically by introduction of the suppressor gene sur-7, and by pharmacological treatment with a chelator of zinc, TPEN [N,N,N′,N′-tetrakis(2-pyridylmethyl)ethane-1,2-diamine]. Finally, direct treatment of wild-type (WT) animals with zinc reduced Orsay virus RNA levels. In these studies, we were able to observe correlated changes in lipid abundance and Orsay virus RNA levels, demonstrating a connection between zinc, lipids, and virus infection.
DISCUSSION
The genetic tractability of
C. elegans and its conservation of many pathways with mammals make it an excellent reductionist model to explore host-virus interactions. Use of Orsay virus enables the study of host-virus interactions
in vivo in a natural multicellular host. Here, we explored the role of lipids in Orsay virus infection. We demonstrate that Orsay virus impacts lipid homeostasis by reducing the amount of lipids ~60% by 48 hpi in animals, suggesting that Orsay virus alters lipid metabolism during viral infection. Interestingly, in human cells, dengue virus similarly reduces lipid abundance by 60% at 24 to 48 hpi (
42).
The dependence of Orsay virus on lipids is also supported by assessing the impact of genetic approaches to depleting lipids. Animals with mutations of multiple transcription factors as well as RNAi knockdown of those transcription factors had reduced Orsay virus RNA levels. Although by RNAi we observed phenotypes for only
sbp-1 and
mdt-15, defined mutants of additional transcription factors also led to lower Orsay virus RNA levels, most likely due to incomplete knockdown by RNAi. The observed dependence of Orsay virus in
C. elegans on lipids parallels the observation that depletion of lipids reduces infection by many pathogenic mammalian viruses, including SARS-CoV-2 (
76), hepatitis C virus (
77), Zika virus (
78), poliovirus (PV) (
79), and encephalomyocarditis virus (EMCV) (
80). Thus, Orsay virus infection of
C. elegans may serve as a robust model of these interactions.
As
sbp-1 is involved in the expression of many lipogenic genes, we sought to determine which lipid biosynthetic genes and corresponding lipid molecules are required during infection. We found genetically that of the three branches regulated by
sbp-1, the
fat-6/fat-7 and
elo-5/elo-6 branches affected Orsay virus infection but the
fat-5 pathway did not.
sbp-1(
ep79) mutants accumulate fatty acids like palmitic and palmitoleic acid, which belong to the
fat-5 branch, but have reduced levels of oleic acid, linoleic acid, and C15iso/C17iso, which correspond to the
fat-6/fat-7 and
elo-5/elo-6 branches, respectively (
48,
67). Feeding the
sbp-1(
ep79) mutants biochemically with multiple lipids from the
fat-6/fat-7 branch, including α-linoleic acid and γ-linoleic acid and dihomo-γ-linoleic acid, rescued viral RNA levels, but stearic acid, arachidonic acid, C15iso, or C17iso did not. The inability of stearic acid to rescue is predicted since stearic acid is the precursor of the
fat-6/fat-7 enzymes and the
sbp-1(
ep79) mutant accumulates stearic acid. The observation that multiple lipids downstream of
fat-6/fat-7 rescued Orsay virus RNA levels confirms the importance of this branch of lipid metabolism for Orsay virus. We did not observe rescue of the
sbp-1(
ep79) mutant by supplementation of C15iso or C17iso even though mutants with mutations in
elo-5 and
elo-6, the genes that synthesize these two lipids, respectively, have lower levels of Orsay virus infection. One possible explanation is that in the
sbp-1(
ep79) mutant there is a reported 50% reduction in the C15iso/C17iso lipid levels whereas in
elo-5 and
elo-6 mutants, which have a deletion of 379 bp and 184 bp of the first exon, respectively, the reduction is likely to be greater; therefore, the levels of these two lipids may not be limiting factors in the
sbp-1 background (
67).
Lipids could be required for one or more steps of the Orsay virus life cycle, such as viral entry, proper release and trafficking of infectious particles into the cytoplasm, replication of viral RNA, viral assembly, or egress. To better define the stage at which
sbp-1 acts, we specifically assessed viral replication by employing an
in vivo replicon system where Orsay virus replication is initiated from an inducible integrated transgene (
6). We found that virus RNA levels were reduced in the
sbp-1(
ep79) mutant, indicating that one or more lipids regulated by
sbp-1 are needed for virus RNA replication. While a number of studies have defined the importance of
srebp-1, the human ortholog of
sbp-1, for viral infection, none of the studies have implicated a specific stage of the viral life cycle that is dependent on
srebp1 (
32,
35,
37,
81). Thus, our findings demonstrate that the step of viral RNA replication is affected in the
sbp-1(
ep79) mutant, although it is possible that additional other steps of the viral life cycle may be also affected.
These findings demonstrate that specific lipids regulated by
sbp-1 are important for Orsay virus infection. The exact mechanism by which these lipids are necessary for an efficient virus replication is still unknown, but several possibilities center on roles involving the viral replication center. It has been shown that oleic acid plays a role in the viral replication step of hepatitis C virus (HCV), as its depletion disrupts the integrity of membranous HCV replication centers and renders HCV RNA susceptible to nuclease-mediated degradation (
82). Another possibility is that the lipids could be the precursors for components of the replication center that help to recruit the RdRp. Such is the case for coxsackievirus B3 (CVB3) and poliovirus (PV) RNA polymerases that show a high affinity for PI4P lipids (
43) or the p92 RNA polymerase from tomato bushy stunt virus (TBSV), which recognizes phosphatidylethanolamine (PE) to form a complex associated with the membrane of peroxisomes (
83). Alternatively, or in addition, lipids can directly modulate viral enzymatic activities. For example, the autocatalytic cleavage of the PV 3CDpol proteins, which are the precursors of the polymerase (3Dpol) and protease 3C, is attenuated when bound to PI4P lipids (
84). Likewise, it has been shown that PE stimulates the enzymatic activity of TBSV viral polymerase and enhances its association with viral RNA (
83), whereas binding to phosphatidylglycerol lipids inhibits its activity (
85). The ability of some lipids, but not others, to rescue Orsay virus infection in
C. elegans provides an opportunity to further dissect the precise biochemical interactions required for virus infection. More studies of the localization of the critical lipids, as well as their downstream products, are necessary to understand how Orsay virus affects and is dependent upon lipid metabolism
in vivo.
One strength of the
C. elegans system is its genetic tractability, which enables detailed dissection of pathways by genetic approaches including suppressor screens. A previous suppressor screen of the
sbp-1(
ep79) mutant lipid defect found that mutation of
sur-7, which encodes a transporter of zinc, restores lipid homeostasis (
49). We found that mutation of
sur-7 rescued the lipid levels as well as the Orsay virus infection in the
sbp-1 background. These results suggested that zinc might be playing a role in the homeostasis of lipids and thus affecting the virus infection. In concordance with this hypothesis, zinc chelation in the
sbp-1 background rescued both lipid levels and Orsay virus infection. It is only due to the unbiased nature of the forward genetic suppressor screening that this linkage between zinc, lipid homeostasis, and virus infection was hypothesized. Interestingly, other papers have shown a correlation between zinc and lipids (
50,
86), and antiviral effects of zinc supplementation against multiple mammalian viruses have been described (
53–55,
57,
61,
87,
88). Although some mechanisms have been proposed to explain the role of zinc during virus infection, like the inhibition of viral protein cleavage and processing as well as inhibition of the viral polymerase activity, the experimental data supporting these mechanistic models are lacking (
54,
55,
57). Interestingly, the importance of lipids for many of these zinc-sensitive viruses including coronaviruses (
26,
89), picornaviruses (
43,
90), and hepatitis E virus (
91), has been reported. Our collective data support a novel hypothetical mechanistic model wherein zinc reduces virus infection via the depletion of lipids and may have broad implications for zinc-sensitive viruses (
Fig. 7). In addition, this reaffirms the great potential of model organism studies to elucidate novel mechanisms and dissect pathways that are broadly important across host species.