Roles of alternative splicing in infectious diseases: from hosts, pathogens to their interactions

Abstract

Alternative splicing (AS) is an evolutionarily conserved mechanism that removes introns and ligates exons to generate mature messenger RNAs (mRNAs), extremely improving the richness of transcriptome and proteome. Both mammal hosts and pathogens require AS to maintain their life activities, and inherent physiological heterogeneity between mammals and pathogens makes them adopt different ways to perform AS. Mammals and fungi conduct a two-step transesterification reaction by spliceosomes to splice each individual mRNA (named cis-splicing). Parasites also use spliceosomes to splice, but this splicing can occur among different mRNAs (named trans-splicing). Bacteria and viruses directly hijack the host's splicing machinery to accomplish this process. Infection-related changes are reflected in the spliceosome behaviors and the characteristics of various splicing regulators (abundance, modification, distribution, movement speed, and conformation), which further radiate to alterations in the global splicing profiles. Genes with splicing changes are enriched in immune-, growth-, or metabolism-related pathways, highlighting approaches through which hosts crosstalk with pathogens. Based on these infection-specific regulators or AS events, several targeted agents have been developed to fight against pathogens. Here, we summarized recent findings in the field of infection-related splicing, including splicing mechanisms of pathogens and hosts, splicing regulation and aberrant AS events, as well as emerging targeted drugs. We aimed to systemically decode host–pathogen interactions from a perspective of splicing. We further discussed the current strategies of drug development, detection methods, analysis algorithms, and database construction, facilitating the annotation of infection-related splicing and the integration of AS with disease phenotype.

Introduction

The inconsistency between adenovirus messenger RNA (mRNA) and its DNA transcription template has allowed researchers to understand that the genetic information transmission from DNA to RNA needs to undergo the removal of invalid information and the splicing of valid information. And such processes are called alternative splicing (AS) that enriches the repertoire of both transcriptome and proteome. It is estimated that 95% of human genes express more than one transcript isoform. Abnormal AS has been proven to be related to the onset and development of infectious diseases.

Infectious diseases are one kind of the most formidable and ubiquitous threats to public health worldwide. In the 21st century, human immunodeficiency virus (HIV) infection, malaria, and tuberculosis (TB) remain the three deadliest infectious diseases, while the outbreak of emerging infections such as corona virus disease 2019, hits the global public health system hard and causes a devastating influence on lives. Decoding the pathogen–host interactions is critical to control the spread of infectious diseases. Remodeling at the transcriptomic level is considered the main way that pathogens and hosts fight each other, and AS has been confirmed as a fast and convenient approach to realize such remodeling. The importance of many AS events in pathogen–host interactions has been documented.

In this review, we focused on AS alternations from the perspective of pathogens, the hosts, and their interactions. We first compared the similarities and differences between pathogen and host splicing mechanisms. Then, we collated all reported regulators and discussed the potential regulatory mechanisms in infectious diseases. We also summarized aberrant AS events in infectious diseases and illustrated their biological effects. Finally, targeted drugs were also described. This work aimed to decode the host–pathogen interactions from the angle of AS, and guide the development of targeted therapies to a certain extent.

Occurrence of AS

AS can be divided into different types including exon skipping, intron retention, alternative 3′ splice site (3′SS), alternative 5′SS, and mutually exclusive exon, as well as alternative polyadenylation and exitron [Figure 1]. Regardless of splicing type, the essence of AS is a two-step S_N2-type transesterification reaction.^[1] First, the 2′ hydroxyl group of the branch point (BP) adenine in the midstream and downstream sequence of the intron attacks its 5′SS, to generate a lariat structure. Then, the exposed 3′-OH at the 3′ end of the upstream exon attacks the 3′SS of the intron, then the upstream and downstream exons are ligated, and the lariat intron is released [Figure 2A].

Figure 1: Different types of alternative splicing. (A) Five major types of AS. (B) Schematic diagram of exitron splicing. (C) Schematic diagram of alternative polyadenylation splicing. (D) Four types of alternative polyadenylation splicing. APA: Alternative polyadenylation; AS: Alternative splicing; CDS: Coding sequence; mRNAs: Messenger RNAs; UTR: Untranslated region.

Figure 2: Splicing reactions of cis-splicing and trans-splicing. (A) The transesterification reactions in cis-splicing. (B) Schematic diagram of genic trans-splicing. Genic trans-splicing can occur among pre-mRNAs from different genes (top right) or the same genes (middle right and bottom right). (C) Schematic diagram of SL trans-splicing. SL exons with caps from SL RNAs are ligated to different structural genes, generating mature mRNA with the same 5′ end. BPS: Branch point site; mRNAs: Messenger RNAs; SL: Spliced leader; SS: Splice site.

In eukaryotes, the above reactions are catalyzed by a spliceosome. The spliceosomes of most fungi are similar to those of mammals, while the spliceosomes of parasites behave quite differently from those of mammals. Viruses and bacteria realize their splicing in a host-dependent manner, that is, utilizing the splicing machinery of their hosts. In the following part, we focus on the spliceosome behaviors in infectious diseases, to offer more mechanistic insights.

Molecular mechanism of AS in mammals

Spliceosome is a massive ribonucleoprotein (RNP) complex, including nearly 100 distinct proteins and five small nuclear RNPs (snRNPs). These snRNPs are composed of corresponding uridine-rich small nuclear RNA (snRNA) and some specific proteins. All snRNAs except U6 snRNA are first exported to the cytoplasm and bound to a series of heptameric Sm proteins to complete assembly, and then the generated snRNPs are transported to the nucleus to perform their respective roles.^[2] U6 snRNA directly combines with the Lsm complex in the nucleus to form U6 snRNP,^[2] a process that does not involve changes in subcellular localization. Spliceosomes are further classified into two main types according to their composition: U2- and U12-type. U2-type spliceosome splices most introns by recognizing the sequence of “CURACU”^[3] while U12-type spliceosome is only responsible for the removal of <0.5% introns by recognizing the 5′SS and 3′SS (AU-AC or GU-AG).^[4]

Each splicing step performed by the spliceosome is accompanied by repeated dissociation and assembly. In each splicing reaction, the spliceosome undergoes the sequential state transformation from early complex (E), pre-spliceosome (A), pre-B complex, pre-catalytic spliceosome (B), activated spliceosome (B^act), catalytically activated spliceosome (B^∗), step I complex (C), step II activated complex (C^∗), post-catalytic complex (P) to intron lariat spliceosome (ILS).^[5] First, U1 snRNP recognizes the 5′SS through the complementary base pairing, and then E complex is generated. Non-snRNP factors such as splicing factor 1 (SF1) and U2 auxiliary factor bind to the BP and 3′SS, respectively. U2 snRNP replaces SF1 through base complementary pairing, and aggregates at the BP, yielding A complex. Next, U4, U6, and U5 snRNPs come together to form a triple snRNPs that interact with A complex and thus urge the formation of pre-B complex. Subsequently, U1 snRNP dissociates, exposed 5′SS transfers to U6 snRNP, and U4 snRNP is unwound. With such conformational transitions, B, B^act, and B^∗ complexes are formed in turn. The first transesterification reaction occurs, and C and C∗ complexes are generated. Then, exons are ligated by U5 snRNA and a P complex generates. As the mature mRNA is released, the ILS complex is formed. After the complex dissociation, released snRNPs proceed to the next round [Figure 3].

Figure 3: Schematic diagram of splicing processes conducted by spliceosome in mammals. Complex A: Pre-spliceosome complex; Complex B^∗: Catalytically activated spliceosome complex; Complex B: Pre-catalytic spliceosome complex; Complex B^act: Activated spliceosome complex; Complex C^∗: Step II activated complex; Complex C: Step I complex; Complex E: Early complex; Complex P: Post-catalytic complex; ILS: Intron lariat spliceosome; mRNAs: Messenger RNAs; snRNP: Small nuclear ribonucleoprotein; SS: Splice site.

Molecular mechanism of AS in pathogens

High conservation between fungal and host spliceosomes, and limited evidence of bacterial splicing mechanism lead us to focus on the molecular splicing mechanism of parasites and viruses.

Parasites

Parasites can perform splicing by ligating exons from two separate pre-mRNAs (known as trans-splicing). In essence, trans-splicing is also a two-step transesterification reaction; however, it does not generate the lariat in the first step of cis-splicing but rather a Y structure intermediate.^[6] There are two kinds of trans-splicing: genic trans-splicing and spliced leader (SL) trans-splicing [Figures 2B and 2C]. The former ligates exons from different pre-mRNAs of the same gene, from transcripts of different genes or intergenic regions, or from transcript products of different chromosomes.^[7] Genic trans-splicing is observed in Drosophila, Caenorhabditis elegans, and mosquitoes.^[7] The later connects the SL exon of SL RNA to various pre-mRNAs, and then generates mature mRNAs with the same common sequence on the 5′ end.^[8] In contrast to snRNAs, SL RNAs dedicate SL exons to transcripts, triggering the consumption of themselves, while snRNAs are relatively constant and circulate in splicing rounds. Such splicing is reported in Euglenozoa, Dinoflagellates, and Ctenophores.^[7]

Viruses

Human immunodeficiency virus-1 (HIV-1) has three types of transcripts: the un-spliced (US), partially spliced (PS), and multiply spliced (MS) transcripts. After the invasion, HIV-1 RNA is reverse-transcribed and integrated into the host genome. Through utilizing host RNA polymerase II (RNAPII), integrated HIV-1 provirus generates a single full-length RNA,^[9] and initially, only MS transcripts are produced to encode the regulatory proteins. Then, PS and MS transcripts increase and produce the structural and accessory proteins. To improve the splicing efficiency, HIV-1 occupies multiple alternative 5′SS and 3′SS to disturb the normal recognition of spliceosome and thus cleaves at GT and AG dinucleotides to produce viral spliced mRNAs.^[10] HIV-1 also employs its own cis-acting elements and host trans-acting elements to regulate splicing, thereby modulating viral protein ratio and infectivity.^[11]

Human papillomavirus (HPV) has an approximately 8 kb DNA sequence that includes an early transcription region (E), a late region, and a long control region. In HPV-16, transcription initiates from an early promoter located at P97, and polycistronic mRNAs that encode E6 and E7 E1, E2, E8 E4, and E5 are formed.^[12] The E6E7 region is the most frequently spliced region and more than four potential isoforms have been found. Splice sites in the E6 open reading frames lead to the generation of a full-length E6 and several truncated E6^∗ isoforms.^[13] The splicing within E6 also regulates the generation of E7 mRNA due to the close position between E6 and E7.^[14,15] The splicing in the E1E2 region is also discovered.^[16] It is a pity that rare evidence reveals the molecular mechanism by which HPV takes advantage of the host splicing system. It is established that HPV-16 can drive host trans-acting factors to regulate its splicing.^[17]

Regulation of AS

Splicing regulators are shown in Figure 4 and relevant evidence is presented in Table 1.

Figure 4: Regulator mechanism of alternative splicing. A: Adenine; ESE: Exon splicing enhancer; ESS: Exon splicing silencer; G: Guanine; hnRNP: Heterogeneous nuclear ribonucleoproteins; ISE: Intron splicing enhancer; ISS: Intron splicing silencer; RNA Pol II: RNA polymerase II; m⁶A: N6 position of adenosine; mRNA: Messenger RNA; snRNA: Small nuclear RNA; snRNP: Small nuclear ribonucleoprotein; SRSF: Serine- and arginine-rich splicing factor; U: Uracil.

Table 1 - Infectious disease-related splicing regulators.

Regulatory factors	Pathogen	Pathogen component	Host target	Action pattern∗	Biological effect∗	Reference
Spliceosome
SnRNP	HSV-1	ICP27	SnRNPs	Changing cellular localization	Inhibiting host splicing	[19]
SnRNA	IAV	NS1	U6	Inhibiting the formation of U6-U2 and U6-U4 complexes	Inhibiting host splicing	[72]
SnRNP-related proteins	HIV-1	Vpr	SAP145	Inhibiting the formation of SAP145-SAP49 complex	Inhibiting the spliceosome assembly and host splicing	[73]
	MRV	μ2	EFTUD2, PRPF8, and SnRNP200	Inhibiting expression	Altering the splicing of specific genes	[18]
Trans-acting factors
SRSFs and related proteins	HBV	NA	SRSF2	Inhibiting expression	Promoting the splicing of viral pre-genomic RNA	[24]
					Promoting the generation of viral HpZ/P′
	HIV-1	NA	SC35 and 9G8	Promoting expression	Promoting splicing at HIV-1 A3 site	[25]
		NA	SRSF1	Promoting expression	Promoting the splicing at viral A1 and A2 site	[25]
		Tat protein	SRSF2	Promoting phosphorylation	Promoting TAU-exon10 skipping	[74]
	HSV-1	ICP27	SRSF2	Changing cellular localization	Inhibiting host pre-mRNA splicing	[19]
	MTB	NA	SRSF2 and SRSF3	Inhibiting expression	Promoting IL-4-exon2 skipping	[26]
					Inhibiting TLR4-exon2 or 3 skipping
	Reovirus	μ2 protein	SRSF2	Changing cellular localization	Changing host splicing	[27]
					Promoting viral replication
hnRNPs	EBV	EBER1	AUF1 and hnRNP D	Inhibiting the binding of AUF1 to AU-rich elements	Changing host splicing	[75]
	IAV	NS1	hnRNP K	Changing cellular localization	Changing host splicing;	[33]
					Promoting viral replication
Other RBPs	DV	NS5	RBM10	Promoting degradation	Increasing SAT1-exon4 inclusion	[36]
					Promoting viral replication
Co-transcriptional regulation	HCMV	NA	RNAPII	Promoting phosphorylation	Changing the splicing of viral RNAs	[38]
				Changing cellular localization
Epigenetic modifications
DNA level	Escherichia coli	NA	Host genes	Decreasing the methylation level	Promoting LGR4-exon5 skipping	[39]
RNA level	Adenovirus	NA	m6A writing enzymes	Changing cellular localization	Increasing m6A level of viral mRNA	[76]
					Increasing the splicing efficiency within adenovirus late transcriptional units
	DV Zika virus WNV HCV	NA	CIRBP	Decreasing the m6A level	Decreasing the intron retention of CIRBP	[40]
Protein level†	HPV-31	E7	SETD2	Increasing the H3K36me3 level	Regulating splice site selection and the generation of viral L1 RNAs	[41]
RNA structures	HIV-1	NA	NA	Promoting the generation of the SL2	Promoting the exon 6D inclusion of viral mRNA	[42]
Phase separation	EBV	EBNA2	SRSF1 and SRSF7	Forming phase-separated droplets	Promoting MPPE1-exon11 skipping	[45]

^∗Researches that do not provide detailed action patterns or biological effects are not listed.

^†Some relevant evidence was listed in the part of “Trans-acting factors”.AU: Adenine and uracil; AUF1: AU-rich element binding factor 1; CIRBP: Cold-inducible RNA binding protein; DV: Dengue virus; EBER1: Epstein-Barr virus-encoded RNA 1; EBNA2: Epstein-Barr virus-encoded nuclear antigen 2; EBV: Epstein–Barr virus; EFTUD2: Elongation factor Tu GTP binding domain containing 2; HBV: Hepatitis B virus; HCMV: Human cytomegalovirus; HCV: Hepatitis C virus; HIV-1: Human immunodeficiency virus-1; HPV-31: Human papillomavirus 31; hnRNP D: Heterogeneous nuclear ribonucleoprotein D; hnRNPs: Heterogeneous nuclear ribonucleoproteins; HSV-1: Herpes simplex virus-1; IL-4: Interleukin 4; IAV: Influenza A virus; LGR4: Leucine rich repeat containing G protein-coupled receptor 4; MPPE1: Metallophosphoesterase 1; mRNA: Messenger RNA; m⁶A: N6 position of adenosine; MRV: Mammalian orthoreovirus; MTB: Mycobacterium tuberculosis; NA: Not available; NS: Nonstructural protein; Vpr: Viral protein R; PRPF8: Pre-mRNA processing factor 8; RBM10: RNA binding motif protein 10; RBPs: RNA-binding proteins; RNAPII: RNA polymerase II; RNP: Ribonucleoprotein; SAP145-SAP49: Splicing-associated protein 145-splicing-associated protein 49; SAT1: Spermidine/spermine N1-acetyltransferase 1; SETD2: Histone methyltransferase SET domain-containing 2; SL2: Stem loop 2; SnRNA: Small nuclear RNA; SnRNP: Small nuclear RNP; SRSF: Serine- and arginine-rich splicing factor; TAU: Translational andrology and urology; TLR4: Toll like receptor 4; WNV: West Nile virus.

Spliceosome

Pathogen-induced changes in the abundance and localization of both snRNAs and snRNP-related proteins can modulate AS by affecting the recognition of splice sites. In general, pathogens secrete some proteins to interact with host snRNP-related proteins to arrest the expression of these proteins. For instance, mammalian orthoreovirus (MRV) releases μ2 protein to interact with host U5 snRNP-related proteins and thus inhibit their expression.^[18] Such alternations further affect the splicing of interleukin 34 (IL34), thereby influencing the immune response of the host. In contrast to abundance alternations, pathogen-induced re-localization is rarely observed.^[19] While poliovirus has been reported to use its 2A(pro) to regulate the assembly of snRNAs and snRNP-related proteins, downstream biological effects have not been elucidated.^[20]

Cis-acting elements

Cis-acting elements are the short-conserved sequence elements within the pre-mRNA. These elements regulate splicing by promoting or inhibiting the recognition of adjacent splice sites and the binding of trans-acting factors. Both mutations in cis-acting elements and impaired cis-acting element motifs in adjacent introns or exons can generate a certain impact on splicing.^[21,22] It is a pity that we fail to collect relevant evidence on infectious diseases.

Trans-acting factors

Trans-acting factors are RNA-binding proteins (RBPs) that specifically bind the cis-acting elements to regulate splicing.

Serine- and arginine-rich splicing factors (SRSFs)

SRSFs family members have certain structural similarities that share one or two N-terminal RNA recognition motif (RRM) domains and an arginine/serine-rich (RS) domain in the C-terminal region. The activity of SRSFs is controlled by the phosphorylation of serine residues within their RS domain. In general, SRSFs bind their RRM domains to exon splicing enhancer (ESE) or intron splicing enhancer and utilize their RS domains to recruit splicing components, then activate splicing. Such an action pattern is position-dependent, that is, SRSFs can also inhibit splicing when they bind to an intron.^[23] The expression, modification, and cellular position of SRSFs all generate significant influence on splicing. Hepatitis B virus, HPV, and Mycobacterium tuberculosis (MTB) have been reported to manipulate host SRSF expression to regulate splicing and benefit themselves.^[24-26] While HPV-16 exerts its E2 protein to enhance the SRSF1 phosphorylation, and then facilitates its own splicing and replication.^[12,17] Unlike the aforementioned pathogens, MRV utilizes its μ2 protein to mediate the localization transition of SRSF2 from the nucleus to the cytoplasm, therefore changing host splicing and enhancing its own replication and pathogenicity.^[27] Some special regulatory mechanisms have also been discovered.^[28]

Heterogeneous nuclear ribonucleoproteins (hnRNPs)

hnRNPs family members typically harbor the RRM, K homology (KH), and glycine-rich domains. RRM and KH domains are responsible for binding to pre-mRNA sequence, while the glycine-rich domain interacts with other hnRNPs.^[29] Likewise, the post-translational modification of hnRNPs partly determines the activity of these proteins. Upon activation, hnRNPs are more inclined to inhibit splicing, whereas this family can also act as a splicing activator in some special situations.

Changing the hnRNPs’ expression is one of the most common strategies employed by pathogens to regulate AS. For example, HPV-16 increases the host hnRNP A1 expression to promote the binding of hnRNP A1 and viral late regulatory element, thereby facilitating viral splicing and improving the expression efficiency of viral late gene.^[30] The roles of hnRNPs’ modifications in some infectious diseases have also been reported.^[31,32] Changes in the subcellular location of hnRNPs also influence splicing. Thompson et al^[33] reported that influenza A viruses (IAVs) migrated host hnRNP H to nuclear speckles (NSs) and then changed the host splicing to influence viral replication.

Other RBPs

The RNA-binding FOX (RBFOX), neuro-oncological ventral antigen (NOVA), and RNA-binding motif (RBM) families also belong to trans-acting factors. The effect of the RBFOX family on exon splicing depends on the binding position, while NOVA proteins mainly target the intron splicing.^[34,35] RBM family functions in a flexible way and their action pattern has not been fully summarized. A recent study reported that Dengue virus non-structural 5 protein increased the inclusion of spermidine/spermine N1-acetyltransferase 1-exon4 and thus favored viral replication by inducing the degradation of RBM10.^[36]

Co-transcriptional regulation

Transcription that occurs simultaneously with splicing (termed as co-transcription splicing) can influence splicing and thus play a regulatory role. Two potential models have been proposed to explain the co-transcriptional regulation.^[37] One is the recruitment model, referring to RNAPII using its own C-terminal domain to recruit trans-acting factors to pre-mRNA and then promote the recognition of splicing sites and splicing. Another one is the kinetic coupling model, indicating that the elongation rate of RNAPII alters the duration of spliceosome-splice site interactions and thus determines whether exons are skipped or retained. These two mechanisms are not mutually exclusive, and sometimes they can co-exist. Changes in RNAPII (expression, modification, etc.) can influence splicing significantly. For example, human cytomegalovirus infection leads to the increased phosphorylation level of RNAPII and changed RNAPII localization, which further regulates the splicing of viral RNA.^[38]

Epigenetic modifications

In infectious diseases, the regulatory roles of DNA or RNA methylation in splicing are relatively less reported. Ju et al^[39] found that Escherichia coli infection induced the methylation of leucine-rich-repeat-containing G-protein-coupled receptor 4-exon5, to affect the exon recognition and subsequent splicing. While Gokhale et al^[40] documented that the Flaviviridae family decreased the N6 position of adenosine (m⁶A) modification of host cold-inducible RNA binding protein (CIRBP) mRNA, to inhibit the expression of long CIRBP isoform and thus promote viral replication. In addition to the modification of trans-acting factors, the modification of histones also plays a regulatory role. For instance, HPV-31 applies its E7 protein to augment the expression of histone methyltransferase SET domain-containing 2 (SETD2) and the level of H3K36me3 that is enriched in HPV-31 genome, further recruiting trans-acting factors and preventing the splicing defects in late viral RNAs brought by SETD2 depletion.^[41]

RNA structures

RNA structures act by determining the exposure degree of RNA sequence, recruiting trans-acting proteins, or adjusting the movement speed of RNAPII. Jablonski et al^[42] documented that HIV-1 drove the generation of the stem loop 2 by inducing the T-to-C mutation and thus promoted the exon 6D inclusion, achieving its enhanced infectivity. In the host, the RNA structural transformation inhibits the hnRNP H activity and then reverses the inhibition effect of this trans-acting factor.^[42]

Phase separation

Phase separation refers to the phenomenon that single-phase molecular complexes separate into two phases (a dense phase and a dilute phase) that then stably co-exist, mediating the formation of membrane-less organelles (NSs, etc).^[43] Phase separation is emerging as a splicing regulator. At the molecular level, NSs contain abundant trans-acting factors and snRNPs, and such composition is strongly suggestive of the potential that NSs promote splicing by offering a separate space.^[44] From a more microscopic perspective, exons are preferentially sequestered into NSs through binding to SRSFs, whereas introns tend to be excluded due to their binding to hnRNPs.^[44] Space partition places the exon–intron boundaries at the NS interfaces, exposes splice sites located at these boundaries, and hence catalyzes the splicing reactions. EBV-encoded nuclear antigen 2 has been verified to harbor the potential to perform phase separation and aggregate SRSF1 and SRSF7 together into the same phase, inducing the cancer-associated splicing landscapes.^[45] The exploration of phase separation is in its infancy and plenty of areas deserve to be investigated.

Specific AS Events of Infectious Diseases

Decoding these events and their biological roles is beneficial to further understanding host–pathogen interactions. Based on available evidence, we preferentially describe AS events related to viral and bacterial infections. Relevant evidence is shown in Table 2.

Table 2 - Infectious disease-related splicing events.

Infectious disease	Targets	Isoforms	Action pattern	Biological effect∗	References
Viral infections
EBV	MPPE1	Isoforms with and without exon11	Increasing MPPE1-exon11 skipping	Inducing EBV-related tumorigenesis	[45]
	STAT1	STAT1α and STAT1β	Increasing STAT1β expression	Promoting cell proliferation	[28,51]
HBV	NA	Viral spliced isoform SP1RNA	Increasing SP1RNA expression	Reducing the recruitment of inflammatory monocytes/macrophages	[77]
HIV-1/AIDS	CCNT1	Isoforms with and without exon7	Increasing CCNT1-exon7	Suppressing the transcriptional activation of HIV-1 and maintaining the latency of infected CD4+ T cells	[46,48,49]
	HLA-1	HLA-A11 and HLA-A11svE4	Increasing HLA-1-exon4 skipping	Inhibiting the activation of natural killer cells	[78]
	RUNX1	RUNX1a, RUNX1b, and RUNX1c	Inhibiting the expression of RUNX1b and RUNX1c	Promoting HIV-1 replication and latency	[50]
High-risk HPV	E6 and E7	E6, E6∗I, E6∗II, E6∗III, E6∗IV, E6∗V, E6∗VI, E6^E7, E6^E7∗I and E6^E7∗II	Increasing E6∗I expression	Inhibiting E6-mediated degradation of p53 and promoting the DNA damage	[79]
HSV-1	MxA	MxA and varMxA	Increasing the exon14–16 skipping of MxA	Promoting HSV-1 infection	[80]
IAV	TP53-i9	P53α, p53β, and p53γ	Increasing the expression of p53α and p53β	Promoting viral replication	[81]
Bacterial infections
TB	IL-4	IL-4 and IL-4 delta2	Increasing IL-4-exon2 skipping	Inhibiting the biological effects of IL-4	[82]
	IL-7	IL-7 and IL-7δ5	Increasing IL-7-exon5 skipping	Promoting STAT5 phosphorylation and prolonging T cells’ life	[56]
	IL-7R	sIL-7R and mIL-7R	Inhibiting the expression of sIL-7R and mIL-7R	Inhibiting the T cell sensitivity to IL-7	[57]
	IL-12Rβ	IL-12Rβ1 and IL-12Rβ1ΔTM	Increasing the IL-12RB-exon14 skipping	Promoting DC migration and MTB-specific T cell activation	[54]
	IL-32	IL-32 and IL-32γ	Increasing the IL-32γ expression	Enhancing the colocalization of MTB and lysosomes and reducing the MTB load	[83]
	RAB8B	Isoforms with and without terminal three exons	Increasing the skipping of terminal three exons	Disturbing the macrophage autophagy to promote the intracellular survival of MTB	[53]
Parasitic infections
Trypanosomiasis	HDAC7	Isoforms with and without exon15	Increasing HDAC7-exon15 skipping	Inhibiting host cell cycle pathways	[84]

^∗Researches that do not provide detailed biological effects are not listed. AIDS: Acquired immune deficiency syndrome; CCNT1: Cyclin T1; DC: Dendritic cell; EBV: Epstein–Barr virus; HBV: Hepatitis B virus; HDAC7: Histone deacetylase 7; HIV-1: Human immunodeficiency virus 1; HLA-A: Major histocompatibility complex, class I, A; HPV: Human papillomavirus; HSV-1: Herpes simplex virus 1; IAV: Influenza A virus; IL-4: Interleukin 4; IL-7: Interleukin 7; IL-7R: Interleukin 7 receptor; IL-12Rβ: Interleukin 12 receptor subunit beta; IL-32: Interleukin 32; mIL-7R: Membrane-bound interleukin-7 receptor; MPPE1: Metallophosphoesterase 1; MPP: Mycoplasma pneumoniae pneumonia; MTB: Mycobacterium tuberculosis; NA: Not available; RUNX1: RUNX family transcription factor 1; sIL-7R: Soluable interleukin-7 receptor; SP1RNA: Spliced 1 RNA; STAT: Signal transducer and activator of transcription; TB: Tuberculosis.

Viral infections

HIV-1 infection/AIDS

The genome integration and splicing hijack mechanism of HIV-1 inevitably perturbates host AS. Byun et al^[46] documented 494 changed AS events of 427 genes in HIV-infected CD4⁺ T cells and functional enrichment suggested that these 427 genes were closely related to HIV-1 life-cycle-related pathways. Cyclin T1 (CCNT1)-exon7 skipping is remarkably promoted after HIV-1 infection. CCNT1 is a component of the positive transcription elongation factor b (P-TEFb) complex, as well as a cofactor of the HIV-1 Tat protein.^[47] The exclusion of CCNT1-exon7 results in the loss of the key Cyclin_N motif and thus the key domain for the P-TEFb complex, ultimately suppressing the transcriptional activation of HIV-1 and maintaining the latency of infected CD4⁺ T cells.^[48,49]

RUNX family transcription factor 1(RUNX1) also undergoes differential splicing during the HIV-1 infection. RUNX1 is a transcriptional regulator gene and involved in the regulation of HIV-1 latent infection. Currently, three host RUNX1 isoforms have been identified: RUNX1b, RUNX1c, and RUNX1a. Both RUNX1b and RUNX1c work as the negative regulators of HIV-1 replication, whereas RUNX1a plays the opposite role.^[50] HIV-1 is able to attenuate the production of RUNX1b and RUNX1c and raises Tat protein level by upregulating long non-coding RNA uc002yug.2, eventually enhancing the activity of HIV-1 long terminal repeat and further promoting viral replication and latency.^[50]

Epstein–Barr virus (EBV) infection

EBV usually utilizes its BS-MLF1 (SM) protein to influence the selection of splice sites in signal transducer and activator of transcription 1 (STAT1) and then upregulates the ratio of STAT1β and STAT1α isoforms.^[28,51] STAT1α is the active form that plays an anti-proliferation or pro-apoptotic role, while STAT1β is the truncated form that exerts the opposite effect.^[52] By controlling the STAT1 splicing, EBV breaks the original balance to promote cell proliferation and subsequently induces tumorigenesis.

In EBV-infected individuals, Peng et al^[45] found that metallophosphoesterase 1 (MPPE1)-exon11 skipping is significantly promoted. The appearance of exon11 introduces a premature termination codon into transcripts, resulting in translation failure. EBV prefers to splice MPPE1-exon11 out to upregulate the production of functional MPPE1 protein and then induce EBV-related tumorigenesis.

Bacterial infection diseases

Tuberculosis

In macrophages, Kalam et al^[53] found that MTB increased the generation of the RAB8B truncated isoform by promoting the skipping of terminal three exons. The premature stop codons render the truncated isoform incapable of translation, consequently interfering with macrophage autophagy and allowing the survival of virulent MTB.^[53] While in dendritic cells, Robinson et al^[54] reported that MTB invasion observably promoted the interleukin 12 receptor subunit beta 1 (IL-12Rβ1)-exon14 skipping (the IL-12(p40)2 receptor) and increased the production of a short isoform, IL-12Rβ1ΔTM. In the full-length transcript, exon14 is responsible for encoding the transmembrane sequence, while its loss introduces an early stop codon in IL-12Rβ1ΔTM. IL-12Rβ1ΔTM can enhance IL-12Rβ1-dependent DC migration and MTB-specific T cell activation by increasing the affinity of IL-12Rβ1 and IL-12(p40)2 or facilitating IL-12(p40)2-dependent signaling.^[54,55]

Based on clinical cohorts, many genes have been found to be differentially spliced between TB patients and healthy controls (HCs). For instance, interleukin-7 (IL-7) isoforms without exon5 (named IL7-δ5), predominantly expressed in MTB-positive granuloma tissue, can drive the phosphorylation of STAT5 in T cells and prolong the life of T cells.^[56] Lundtoft et al^[57] discovered that MTB infection can impair T cell sensitivity to IL-7 by declining plasma soluble IL-7 receptor (IL-7R) (IL-7R without exon5 and exon6) and membrane-type IL-7R in T cells.

Mycoplasma pneumoniae pneumonia (MPP)

Granulysin (GNLY) is an antimicrobial protein located in the cytotoxic granules of natural killer (NK) and cytotoxic T cells. During the MPP development, GNLY-intron1 is preferential to be excluded to promote the generation of a shorter GNLY transcript.^[58] Intriguingly, the Ensembl database indicates that transcripts with intron1 (such as ENST00000470974.1 and ENST00000489980.5) fail to encode proteins. The above evidence demonstrates that the host may promote the generation of GNLY transcripts with translational potential to enhance their immune defense capacity.

It is also reported that the alternative 3′SS splicing of solute carrier family 11 member 1 (SLC11A1) increases in MPP children.^[58] SLC11A1 can transport divalent cations (Mn2⁺ and Fe2⁺), and inhibit pathogen growth by eliminating bivalent metals.^[59] Based on the above studies, we speculate that this event may be involved in MPP by regulating ion metabolism or protective enzyme production. More exploration is needed to verify this speculation.

Targeted Agents for AS in Infectious Diseases

Several inhibitors targeting aberrant splicing in viral, parasitic, and fungal infections are developed. In the HIV-1 infection, most agents act by targeting trans-acting factors. For example, indole derivatives (IDC16) suppresses the ESE-dependent splicing activity of SRSF1 and then inhibits the formation of key viral proteins, disrupting the generation of full-length HIV-1 pre-mRNA and the assembly of HIV-1 particles.^[60] ABX464 exerts its action by preventing the Rev-mediated export of US transcripts to the cytoplasm and thus inhibits HIV-1 replication.^[61] Reported anti-IAV agents mainly target specific splicing events.^[62,63] Besides, drugs against other viruses, parasites, and fungi infections are also documented.^[64-71] Of note, drugs targeting bacterial infection have not been reported. Similar splicing regulatory patterns in different infections lead us to hold the view that these developed drugs may have certain applicability for bacterial infections. Further research is needed. And the side and toxic effects of developed drugs still need to be further assessed. All detailed information is listed in Table 3.

Table 3 - Developed targeted agents for AS in infectious diseases.

Infection	Pathogens	Potential agent	Mechanism∗	Reference
Viral infection	HIV-1	SRPIN340	SRPIN340 inhibits the phosphorylation of SRSF4 and promotes its degradation by arresting SRPK1 and SRPK2.	[85]
		IDC16	IDC16 inhibits the ESE-dependent splicing activity of SRSF1 and then the formation of key viral proteins, thereby disrupting the generation of full-length HIV-1 pre-mRNA and the assembly of HIV-1 particles.	[60]
		1C8	1C8 affects the SRSF10-dependent splicing of HIV-1 transcripts by promoting the dephosphorylation of SRSF10 and increasing its interaction with hTra2β.	[86]
		ABX464	ABX464 prevents Rev-mediated export of unspliced HIV-1 transcripts to the cytoplasm and interacts with the Cap binding complex, thereby inhibiting viral replication.	[61]
		Chlorhexidine	Chlorhexidine decreases the levels of unspliced and single-spliced viral RNAs, and reduces the Rev accumulation.	[87]
		5342191†	5342191 decreases the levels of unspliced and single-spliced viral RNAs, reduces the accumulation of viral proteins by promoting MEK1/2-ERK1/2 signaling and thus inhibits viral replication.	[88]
		GPS491‡	GPS491 decreases the levels of unspliced and single-spliced viral RNAs, as well as the accumulation of three essential viral proteins.	[89]
	Influenza virus	PPMO	PPMO promotes TMPRSS2-exon5 skipping to increase the expression of inactive TMPRSS2 form and thus suppress viral infectivity.	[62]
		Clypearin Corilagin Pinosylvine	Drugs target CLK1 to disturb the splicing of M2 viral mRNA.	[90]
		Ro-3306	Ro-3306 inhibits host CDK1 to affect the splicing of M2 viral mRNA, leading to the restriction of viral replication. It also directly binds to PB2 to inhibit viral RNA replication.	[63]
	HCMV	NMS-873	NMS-873 inhibits the splicing switch from IE1 to IE2 and thus decreases IE2 expression by inhibiting VCP.	[65]
Parasitic infection	Trypanosoma cruzi	MAO	MAO prevents trans-splicing of pre-TcIP(3)R mRNA to inhibit the growth and infectivity of parasite.	[67]
	Malarial parasite	TCMDC-135051(1)	TCMDC-135051(1) kills malarial parasite by inhibiting PfCLK3.	[68]
	Toxoplasma gondii	Altiratinib	Altiratinib induces the intron retention in both T. gondii and P. falciparum by inhibiting PRP4K/CLK3.	[69]
	Plasmodium falciparum
Fungal infection	Candida albicans	Bleomycin	Bleomycin selectively inhibits the self-splicing of Group I intron, leading to the accumulation of pre-mRNA.	[70]
	Cryptococcus neoformans	CMN	CMN acts as a Prp8 intein splicing inhibitor and specifically inhibits the Prp8-Intein-containing fungi. CMN also reduces virulence factors, melanin production, and biofilm formation.	[71]

^∗Researches that do not provide detailed mechanisms are not listed.

^†This agent refers to compound N-[4-chloro-3-(trifluoromethyl) phenyl]-7-nitro-2,1,3-benzoxadiazol-4-amine.

^‡This agent refers to thiazole-5-carboxamide derivative. AS: Alternative splicing; CDK1: Cyclin dependent kinase 1; CLK1: CDC Like Kinase 1; CLK3: CDC Like Kinase 3; CMN: Calcimycin; ERK: Extracellular signal-regulated kinase; ESE: Exon splicing enhancer; HCMV: Human cytomegalovirus; HIV-1: Human immunodeficiency virus-1; IDC16: Indole derivative; IE: Immediate early gene; MAO: Morpholino antisense oligo; MEK: Mitogen-activated protein/extracellular signal-regulated kinase (ERK) kinase; mRNA: Messenger RNA; PPMO: Peptide-conjugated phosphorodiamidate morpholino oligomer; PRP4K: Pre-mRNA processing factor 4 kinase; SRSF: Serine- and arginine-rich splicing factor; SRPIN340: A selective SRPK inhibitor; SRPK1: Serine- and arginine-rich splicing factor protein kinase 1; SRPK2: Serine- and arginine-rich splicing factor protein kinase 2; TcIP(3)R: Inositol 1,4,5-trisphosphate receptor of Trypanosoma cruzi; TMPRSS2: Transmembrane serine protease 2; VCP: Valosin containing protein.

Conclusions and Future Perspectives

In summary, fungi and parasites rely on their own spliceosomes to shape their own transcription profiles, while viruses and bacteria do this by hijacking the host's splicing system. Their hijacking is realized by multiple pathways (affecting trans-acting factors, epigenetics, etc.). Splicing regulation is a complicated process, involving many factors, and the synergistic or antagonistic interactions among these factors. Changes in regulators trigger a series of gene splicing changes and thus influence the immune, growth, or metabolism-related processes. To date, most attention has been given virus-related splicing. In bacterial or parasitic infections, existing data so far are obtained by observing phenotype changes, and the underlying molecular mechanisms behind these phenotypes remain unknown.

Current strategies for developing targeted agents include targeting AS events and specific regulators. In infectious diseases, targets used in the former strategy are usually pathogen-derived molecules. Compared with the former, the latter strategy can generate wider biological effects. While one side effect that can be expected is to interfere with normal biological processes in which the target regulator is involved. In addition to developing targeted drugs, we should also pay attention to the potential of these events as biomarkers. The diagnostic/predictive ability, performance stability, and applicable populations of these AS biomarkers need to be further comprehensively evaluated.

Detection technologies and analytic methods are important for AS-related researches. High-throughput sequencing is the most common technology to detect AS; however, short read length restricts its detection accuracy. Emerging third-generation sequencing has longer read length, and is more suitable for the AS detection. Strikingly, advances in single-cell RNA sequencing (scRNA-seq) facilitate the AS detection from the bulk level to a finer level. Smart-seq2 allows researchers to observe AS in isolation from cellular heterogeneity. However, Smart-seq2 captures fewer cells than other scRNA-seq methods, which restricts the number of observable objects. Algorithms for processing Smart-seq2 data (such as expedition) are also limited. For other scRNA-seq platforms, some AS analytic methods (scAPA, scDAPA, scMAPA, etc.) have been developed. However, these algorithms only allow us to identify the alternative polyadenylation splicing.

To this end, we emphasize to construct some infection-related AS public databases that comprehensively summarize and annotate detailed AS events based on large private or public data. If possible, clinical variations can be incorporated. Based on these comprehensive databases, scholars can conclude the splicing rules of infectious diseases, thus facilitating the discovery of broad-spectrum anti-infective targets or promising biomarkers.

Funding

This work was funded by the National Natural Science Foundation of China (No. 82272416).

Conflicts of interest

None.