2024 年 47 巻 5 号 p. 905-911
Viruses require host cells to replicate and proliferate, which indicates that viruses hijack the cellular machinery. Human immunodeficiency virus type 1 (HIV-1) primarily infects CD4-positive T cells, and efficiently uses cellular proteins to replicate. Cells already have proteins that inhibit the replication of the foreign HIV-1, but their function is suppressed by viral proteins. Intriguingly, HIV-1 infection also changes the cellular metabolism to aerobic glycolysis. This phenomenon has been interpreted as a cellular response to maintain homeostasis during viral infection, yet HIV-1 efficiently replicates even in this environment. In this review, we discuss the regulatory role of glycolytic enzymes in viral replication and the impact of aerobic glycolysis on viral infection by introducing various host proteins involved in viral replication. Furthermore, we would like to propose a “glyceraldehyde-3-phosphate dehydrogenase-induced shock (G-shock) and kill strategy” that maximizes the antiviral effect of the glycolytic enzyme glyceraldehyde 3-phosphate dehydrogenase (GAPDH) to eliminate latently HIV-1-infected cells.
Viruses lack autonomous replication mechanisms and rely on the cellular environment within the host organism for replication. This necessitates their utilization of host cell enzymes, transcription and translation machinery, and other cellular components to facilitate the replication of their genetic material and the generation of progeny viral particles. On the other hand, the cell has a system for eliminating viruses and a struggle between the virus and the host is constantly taking place within the cell. Intriguingly, it has been reported that the metabolic state of cells changes in response to viral infection.1) Therefore, to gain a robust knowledge of viral replication, it is necessary to elucidate the regulation mechanism of viral replication considering not only the proteins involved in replication but also the metabolic state within the host cell.
Human immunodeficiency virus type 1 (HIV-1), causes AIDS, which is one of the three major infectious diseases, along with tuberculosis and malaria in the world. This virus was recognized and isolated when men began to develop progressive, unexplained immunodeficiency.2–4) It probably spread sporadically from nonhuman primates to humans throughout the 1900s.5) Starting with the discovery of azidothymidine (3′-azido-3′-deoxythymidine, zidovudine, or AZT) in 1987, a wide variety of anti-HIV drugs are now available for clinical use.6) However, further antiviral drug development is required owing to the emergence of drug-resistant viruses. In addition, HIV-1 is integrated into the human genome, making it difficult to eliminate the virus from the body. Currently, two treatment goals are proposed: sterilizing cure, which eliminates all viral-infected cells, or functional cure, which allows a lifelong control of the virus without treatment, even though complete eradication of the virus will not be achieved.7) Unfortunately, none of these have yet been achieved. Therefore, there is a need to properly characterize the adaptive capacity of HIV-1 from multiple perspectives and establish strategies to achieve cure. In this review, we summarize the functions of host proteins that contribute to viral infection and discuss our findings on the function of glycolytic enzymes in regulating HIV-1 replication and the importance of cellular metabolic changes associated with HIV-1 infection. We also discuss the importance of focusing on cellular metabolism in strategies to eliminate latently infected cells that prevent HIV-1 cure.
HIV-1 has positive-strand RNA as its genetic information and replicates mainly in CD4-positive T cells through the function of host proteins. HIV-1 penetrates into target cells through interactions of the envelope proteins on its viral membrane with the CD4 and either CC-chemokine receptor 5 (CCR5) or CXC-chemokine receptor 4 (CXCR4) on the cellular membrane.8–12) Subsequently, the reverse transcription process, which is the conversion of viral genome RNA to DNA, and the uncoating process, which is the release of the viral genome, proceed. During the reverse transcription, host-derived tRNALys3 is used as a primer.13–16) After membrane fusion, the core that encapsulates the viral genome is transported to the nucleus with the assistance of cleavage and polyadenylation specific factor 6 (CPSF6).17) During the uncoating, the cis-trans isomerization activity of the host protein peptidyl-prolyl isomerase Pin1 is used to induce conformational changes in the viral core, resulting in core disassembly.18) One of the hallmarks of HIV-1 replication and one of the causes of the viral reservoirs in vivo is integration of the viral genome into the host genome. At this step, lens epithelium–derived growth factor p75 splice variant (LEDGF/p75) is used as a linker that tethers the viral genome to the host genome.19,20) The transcription of the HIV-1 genome is initiated from 5′-LTR and involves several host-derived transcription factors, including nuclear factor-kappaB, and proceeds through the recruitment of host-cell positive transcription elongation factor b (P-TEFb) in addition to the viral transcriptional activator.21–23) The viral proteins translated by the host mechanism, together with the viral genomic RNA, assemble at the plasma membrane, and budding proceeds. At these steps, it is important that N-myristoyltransferase (NMT) modifies viral proteins with myristic acid and that charged multivesicular body protein (CHMP) 2A and CHMP4B, which form endosomal sorting complexes required for transport (ESCRT), detach from the plasma membrane.24–31) The essential proteins required for HIV-1 replication are described in Fig. 1 along with an overview of the viral replication mechanism. In order to effectively block HIV-1 replication, it is useful to inhibit the functions of the above mentioned proteins, but because of the abnormalities in host homeostasis as well as the virus, there are no approved antiviral drugs except for maraviroc, a CCR5 inhibitor. This fact indicates that the virus is hijacking host proteins critical to the host cells.
It illustrates the processes from viral entry into cells to the production of new viruses. The essential proteins are enclosed in a square. Myr- represents the myristoyl group modification by NMT. Reverse transcription, integration, and maturation are catalyzed by viral enzymes, namely, reverse transcriptase, integrase, and protease, respectively.
Because viruses are foreign to the host, cells are equipped with antiviral proteins. In particular, proteins found specifically as “restriction factors” significantly reduce HIV-1 infectivity, but is inhibited by a virally encoded protein.32) One of the most well-known restriction factors is apolipoprotein B mRNA-editing enzyme catalytic subunit-like 3G (APOBEC3G).33) APOBEC3G deaminates cytidine residues in nascent negative single-stranded cDNA during reverse transcription, resulting in guanosine-to-adenosine hypermutation of the viral positive strand sequence.34–36) Namely, APOBEC3G exhibits antiviral activity through the gene editing of the virus and impairing its genetic integrity. On the other hand, HIV-1 encodes Vif, which counteracts APOBEC3G through APOBEC3G degradation by the ubiquitin-proteasome system, which is formed by Vif directly binding to APOBEC3G and recruiting it to a ubiquitin ligase complex.33,37,38) Colomer-Lluch et al. summarized the classical and well-documented SAM domain and HD domain-containing protein 1 (SAMHD1), tetherin, and tripartite motif protein 5α (TRIM5α) with APOBEC3G, as well as the more recently characterized myxovirus resistance 2 (MX-2), serine incorporator 3 (SERINC3), SERINC5, interferon-induced transmembrane proteins (IFITMs), schlafen 11, membrane-associated RING-CH 2 (MARCH2), and MARCH8.39) These findings indicate that HIV-1 can efficiently replicate while eliminating unfavorable restriction factors.
Since HIV-1 particles are approximately 100 nm in diameter, intraparticle space is limited, and incorporated proteins are likely to be important factors for replication. Interestingly, several groups, including ours, have shown that the glycolytic enzymes glyceraldehyde 3-phosphate dehydrogenase (GAPDH), alpha-enolase (ENO1), and pyruvate kinase muscle type 2 (PKM2) are incorporated into HIV-1 particles, even though they do not have an energy-producing system.40–44) These glycolytic enzymes are called “moonlighting proteins” because they exhibit diverse functions in addition to glycolysis, such as mRNA binding.45,46) Several studies showed that glycolytic enzymes are involved in the replication of various viruses such as the hepatitis B virus, Sendai virus, dengue virus, and influenza virus.47–53) Therefore, it was of interest to us whether HIV-1 replication is controlled by glycolytic enzymes.
GAPDH, ENO1, and PKM2 inhibit viral reverse transcription. Previous studies revealed that reverse transcription is initiated within budding viral particles and template RNA, reverse transcriptase, primers, and deoxyribonucleotide triphosphates (dNTPs) must be incorporated into the viral particles for accurate enzymatic reactions. In HIV-1 reverse transcription, viral genomic RNA is the template, primers and dNTPs are supplemented by host-derived materials and virus-specific reverse transcriptase catalyzes the reactions. It is noted that non-aminoacylated tRNALys3 is used as the primer, and priming proceeds within the virus particles.13) It has been reported that increased levels of packaged tRNALys3 increase the infectivity of the virus.54) The packaging of tRNALys3 into viral particles requires the viral proteins Pr55gag and Lysyl-tRNA synthetase (LysRS) and the interaction of the tRNALys3-LysRS complex with Pr55gag.55–57) In this situation, the aminoacylation of tRNALys3 is inhibited by HIV-1 Vpr by directly binding to LysRS.58) Our first report demonstrating the involvement of GAPDH in reverse transcription showed that the amount of GAPDH expression levels in virus-producing cells was correlated with the amount of GAPDH incorporated into virus particles, and that the amount of GAPDH incorporated into virus particles was inversely correlated with the amount of tRNALys3 in virus particles.41) Furthermore, co-immunoprecipitation and yeast two-hybrid-based interaction analysis showed that GAPDH interacts directly with Pr55gag as a tetramer at two sites and competitively inhibits the interaction between Pr55gag and the tRNALys3-LysRS complex.41,59) These findings are also supported by the observation that increased expression levels of LysRS in virus-producing cells decrease the amount of GAPDH incorporated into the viral particles. On the other hand, ENO1 and PKM2 are also incorporated into the viral particles and prevent viral reverse transcription.43,44) The finding that PKM2 is a key host factor that adversely affects HIV-1 infectivity by blocking virion recruitment of tRNALys3 was featured on the cover of Biol. Pharm. Bull., 2018;41(4).: 612–618. Furthermore, albeit an ENO1-specific finding, ENO1 in HIV-1 target cells contributes to the inhibition of viral integration.60) Although a more detailed analysis of the antiviral activity of ENO1 and PKM2 is required, the findings suggest that each of these glycolytic enzymes exerts its antiviral activity independently. Taken together, these studies indicate that glycolytic enzymes involved in the backbone of energy metabolism collaboratively contribute to the replication of various viruses, including HIV-1.
Glycolytic enzymes such as GAPDH, ENO1, and PKM2 exhibit antiviral activity, but there is no protein encoded in the viral genome to counter them. Therefore, these cannot be called “restriction factors.” However, we hypothesized that the shift in the host cell metabolism associated with infection may be the means for HIV-1 to counter the anti-HIV-1 effects of these glycolytic enzymes. As reviewed by Thaker et al., it has long been known that viral infection alters the host cell metabolism, a phenomenon called metabolic reprogramming, and the mechanisms and effects of metabolic reprogramming have been actively investigated.1) Aerobic glycolysis, first observed in cancer cells, allows for a rapid ATP supply to the cells.61) It has been shown that cells infected with HIV-1 also show enhanced aerobic glycolytic metabolism, resulting in an enhanced viral production.62,63) In HIV-1-infected cells, the expressions of glucose transporter 1, which is one of the main glucose transporter in activated T cells, and hexokinase (HK), which is one of the rate-limiting enzymes in glycolysis, are upregulated and facilitate the utilization of glycolysis.64,65) Chang et al. demonstrated that the inhibition of aerobic glycolysis in T-cells using galactose instead of glucose in the cell culture medium induced a moonlighting function of GAPDH as an mRNA-binding protein.66) Interestingly, our study examining the function of glycolytic enzymes in HIV-1 replication under similar conditions showed that the incorporation of GAPDH, ENO1, and PKM2 into the viral particles is enhanced in cultures grown on galactose medium, which contributes to the inhibition of reverse transcription.63) The function of GAPDH based on the cellular metabolic state that we currently hypothesize is shown in Fig. 2. These findings suggest that providing an environment for GAPDH, ENO1, and PKM2 to exert their moonlight function can inhibit the efficient replication of HIV-1 by inhibiting HIV-1 reverse transcription.
(A) In a state where cells utilize aerobic glycolysis, GAPDH is engaged in the glycolytic flux, which prevents it from interacting with viral proteins and being incorporated into viral particles. Under this situation, the tRNALys3-LysRS complex is readily incorporated into viral particles. As a result, the viral particle acquires tRNALys3, a primer for the reverse transcription. (B) In a state where cells do not utilize aerobic glycolysis, the glycolytic flux diminishes, causing GAPDH to perform functions other than glycolysis. As a result, instead of the tRNALys3-LysRS complex, GAPDH interacts with viral proteins and is incorporated into virus particles, thereby preventing the incorporation of the tRNALys3-LysRS complex.
The metabolic state of HIV-1-infected cells was also shown to affect HIV-1 latency. Latent HIV-1 infection has been interpreted to be caused by a replication-competent but transcriptionally repressed provirus, which presents a major obstacle to HIV-1 cure. HIV-1 latency has been shown to be triggered by several mechanisms that lead to the silencing of viral expression, including epigenetic DNA modification.67) Interestingly, we found that the inhibition of glycolysis using galactose inhibits virus production.63) Shytaj et al. showed that an enhanced glycolysis in early HIV-1 infection is required for active HIV-1 replication, and then a progressive downregulation of glycolysis in late HIV-1 infection is associated with the onset and maintenance of the HIV-1 latency period.68) Furthermore, Mutascio et al. have shown that CD8-positive T cells contribute to the metabolic reprogramming of CD4-positive T cells and arrest the glycolysis of CD4-positive T cells, resulting in a latent infection state.69) These findings suggest that in the development of therapeutic strategies targeting HIV-1 latency, increasing glycolytic activity would facilitate the elimination of latently infected cells by the host immune system.
It has been 40 years since the identification of HIV-1, but the strategies necessary to achieve a cure remain controversial. Two approaches have been proposed.70) In the context of functional cure, a method called “block and lock” has been proposed, aiming to maintain latently infected cells that are deeply silenced using latency-promoting agents. Furthermore, in the context of sterilizing cure, a method called “shock and kill” has also been proposed. In this approach, latently infected cells are forcefully reactivated using latency-reversing agents (LRAs), such as protein kinase C agonists and histone deacetylase inhibitors, and then eliminated by CD8-positive T cells. Since individual “shock and kill” strategies to date have ineffectively reduced reservoir size in clinical trials, combining several LRAs with different mechanisms of action may be required to address the heterogeneity of latent HIV-1 reservoirs. Strategies are also underway to find LRA combinations that do not reduce the ability of immune cells to eliminate activated HIV-1-infected cells and to find LRAs with improved delivery to lymph nodes.
In this review, we highlight the importance of focusing on glycolytic enzymes with anti-HIV-1 activity in infected cells and metabolic shifts in HIV-1-infected cells. Focusing on the energy-producing glycolytic enzymes, the rate-limiting enzymes in the glycolytic system are known to be HK, 6-phosphofructokinase 1, and PKM. Therefore, it is considered that the increase in the expression level of an enzyme such as GAPDH alone, which is not the rate-limiting enzyme in glycolysis, does not directly promote the glycolysis flux itself and increase the production of ATP. Interestingly, seven out of ten glycolytic enzymes, including GAPDH, ENO1, and PKM2, and seven out of eight citric acid cycle enzymes have been reported to have moonlighting functions.71) Yamaji et al. reported that glycolytic GAPDH activity was not elevated in vascular endothelial cells under hypoxic conditions, despite the increased expression level of GAPDH,72) suggesting that not all newly induced GAPDH molecules contribute to glycolysis. Intriguingly, Ceriani et al. found that when human peripheral blood mononuclear cells (PBMCs) and CD4+ T cells are treated with representative LRAs for 24 h, PEP005 increased the expression level of GAPDH, whereas a STING agonist decreased the expression level,73) suggesting that the combination of LRAs that can efficiently upregulate GAPDH expression raises the possibility that not only the activation of latently infected cells but also the infectivity of newly produced viruses can be inhibited by GAPDH. In addition, since the reservoir size in vivo has been successful reduced using romidepsin and immune checkpoint blockade (ICB) nivolumab targeting PD-1,74) there is also scope to examine the effects of cocktail therapy using combinations of LRAs selected to efficiently upregulate GAPDH expression in latently infected cells and ICB (Fig. 3).
A combination of LRAs that can efficiently upregulate GAPDH expression will not only activate latently infected cells, but also reduce the viral infectivity owing to the antiviral effect of GAPDH in virus-producing cells, leading to the elimination of infected cells via immune systems stimulated by immune checkpoint blockade (ICB).
The authors declare no conflict of interest.
