ISG15 and HES4 Subtelomeric Interferon-Notch Axis (Chr1p36.33)

HES4 (Hairy and Enhancer of Split 4, located at NC_000001.11:998,962–1,000,172 on the negative strand, 4 exons, 221 amino acids, ~23.5 kDa) and ISG15 (Interferon-Stimulated Gene 15, located immediately telomere-distal at NC_000001.11:1,013,497–1,014,540 on the positive strand, 2 exons, 165-residue precursor cleaved to a 157-residue mature 15-kDa polypeptide) form a compact, divergently oriented 13.3-kb head-to-head subtelomere gene pair within the terminal megabase of human chromosome 1p36.33, immediately upstream of the neuromuscular synapse organizer AGRN (agrin, 1,020,120–1,056,118). Together, this GC-rich (>66% GC) subtelomeric neighborhood integrates two fundamental vertebrate signaling cascades: Notch developmental fate determination (HES4) and Type I/III interferon (IFN-α/β/λ) innate antiviral immunity (ISG15).

Despite their physical proximity within the same Chr1p36.33 subtelomeric chromatin domain, HES4 and ISG15 reveal contrasting evolutionary and clinical trajectories across placental mammals. Whereas HES4 encodes a canonical Notch-responsive basic helix-loop-helix–Orange (bHLH-O) transcriptional repressor that is conserved across humans, non-human primates, carnivores, and teleost fishes—yet was evolutionarily deleted from the genome of the laboratory house mouse (Mus musculus) and brown rat (Rattus norvegicus)—ISG15 encodes the archetype tandem ubiquitin-like (Ubl) modifier whose intracellular conjugation (ISGylation), extracellular cytokine signaling (LFA-1 → IFN-γ), and non-covalent stabilization of USP18 expose divergent species-specific phenotypes between mouse viral susceptibility and human Mendelian Type I interferonopathies and Mendelian Susceptibility to Mycobacterial Disease (MSMD).

Divergent Head-to-Head Promoter Topology and Non-B DNA Motifs

Within the GRCh38/T2T-CHM13v2.0 assembly of human Chr1p36.33 (1.00–1.06 Mb), HES4 (transcribed centromere-to-telomere on the minus strand) and ISG15 (transcribed telomere-to-centromere on the plus strand) are separated by a 13.3-kb intergenic regulatory interval rich in unmethylated CpG islands, G-quadruplex (G4) clusters, and alternating purine–pyrimidine Z-DNA-forming repeats ((TG)n/(CA)n and (CG)n).

This open euchromatic subtelomeric chromatin state permits rapid, independent signal-induced transcription at each promoter:

• The HES4 Promoter: Contains paired high-affinity CSL (RBPJ / CBF1) binding motifs (5′-YGTGRGAA-3′, arranged in a head-to-head Su(H) Paired Site [SPS]) that recruit the cleaved intracellular domain of Notch (NICD1–NICD4) together with the coactivator MAML1 (Mastermind-like 1) and histone acetyltransferase p300.
• The ISG15 Promoter: Contains two tandem Interferon-Stimulated Response Elements (ISREs, 5′-GAAANNGAAA-3′) bound by the canonical Type I/III interferon transcription factor complex ISGF3 (STAT1–STAT2–IRF9) as well as IRF3/IRF7 and PU.1—making ISG15 one of the fastest and most abundantly induced transcripts (increasing 100-fold to 1,000-fold within 2 to 6 hours) following viral infection or cytosolic nucleic acid sensing (RIG-I/MDA5/cGAS–STING).

Mode 1: Intracellular Co-Translational ISGylation (UBE1L–UBE2L6–HERC5)

Structurally, human ISG15 (15 kDa) is the oldest vertebrate member of the ubiquitin-like modifier family and consists of two tandem ubiquitin-like β-grasp domains (an N-terminal domain residues 1–78 and a C-terminal domain residues 79–157, sharing ~30% sequence identity with ubiquitin) connected by a flexible 6-residue hinge and terminating in the canonical diglycine conjugation motif Leu-Arg-Leu-Arg-Gly-Gly (LRLRGG) (exposed after cleavage of an 8-residue C-terminal pro-peptide by USP18 or UCHL3).

Crucially, every enzyme of the ISGylation cascade is itself transcriptionally co-induced by Type I interferons:

1. E1 Activating Enzyme (UBA7 / UBE1L): Uses Mg²⁺–ATP to adenylate the C-terminal Gly157 of ISG15 and form a high-energy E1–ISG15 thioester bond.
2. E2 Conjugating Enzyme (UBE2L6 / UbcH8): Accepts ISG15 via transthioesterification.
3. E3 Ligases (HERC5 in humans; HERC6 in mice; plus TRIM25/EFP and ARIH1): Unlike ubiquitin E3 ligases, which recognize specific degron motifs on mature folded substrates, human HERC5 physically associates with the 60S ribosomal subunit (polysomes). As a result, HERC5 co-translationally conjugates ISG15 onto lysine residues of newly synthesized nascent polypeptides during translation. In a virus-infected cell where the majority of actively translating mRNAs are viral capsids and polymerases, co-translational ISGylation sterically blocks the oligomeric assembly of influenza A (NS1/NP), SARS-CoV-2, HIV-1 (Gag), and papillomavirus capsids—explaining why coronaviruses (PLpro papain-like protease) and nairoviruses (OTU proteases) evolved dedicated viral de-ISGylating enzymes to strip ISG15 from viral proteins.

Mode 2: Extracellular Cytokine Signaling via Leukocyte LFA-1 (CD11a/CD18)

In addition to intracellular protein conjugation, unconjugated ("free") monomeric and dimeric ISG15 is secreted via non-classical exocytosis/secretory lysosomes by monocytes, neutrophils, lymphocytes, and epithelial cells. Extracellular ISG15 functions as a true immunomodulatory cytokine: it binds the αL I-domain of the leukocyte integrin LFA-1 (CD11a/CD18, encoded by ITGAL/ITGB2) on Natural Killer (NK) cells and T lymphocytes. Engagement of LFA-1 activates Src-family kinases (Lck/Fyn), Syk synergizing with Interleukin-12 (IL-12) to drive potent secretion of Interferon-γ (IFN-γ) and IL-10—essential for macrophage activation and clearance of intracellular Mycobacterium tuberculosis and attenuated BCG vaccine strains.

Mode 3: Non-Covalent Stabilization of USP18 and Prevention of Human Type I Interferonopathy

In 2012–2015, Jean-Laurent Casanova, Dusan Bogunovic, and colleagues discovered a striking evolutionary divergence between human and mouse ISG15 by studying patients with biallelic loss-of-function mutations in ISG15:

• Paradoxical Viral Resistance vs. Mycobacterial Susceptibility: Unlike Isg15^−/− mice (which are hypersusceptible to viral infections), human ISG15-deficient patients do not suffer from severe viral infections; instead, they present with Mendelian Susceptibility to Mycobacterial Disease (MSMD) (due to loss of extracellular ISG15–LFA-1-driven IFN-γ production) alongside basal ganglia calcifications and systemic Type I interferonopathy resembling Aicardi–Goutières syndrome (AGS).
• Molecular Basis (ISG15–USP18 Non-Covalent Shielding): The ISG15-specific isopeptidase USP18 (UBP43) performs two independent roles: (1) it catalytically cleaves ISG15 from conjugated proteins, and (2) it binds the cytoplasmic Box1/Box2 region of the Type I IFN receptor subunit IFNAR2, displacing JAK1 to shut off STAT1/STAT2 phosphorylation (the master negative-feedback brake on Type I IFN signaling). In humans (but not in mice), intracellular free ISG15 binds non-covalently to USP18, shielding USP18 from SKP2-mediated ubiquitination and proteasomal degradation. Consequently, human cells lacking ISG15 degrade their USP18 protein, leaving IFNAR1/IFNAR2 unbraked—which drives hyper-amplified constitutive ISG expression (conferring broad viral resistance at the cost of neuroinflammatory cerebral calcification).

Structural Mechanism of bHLH-Orange and WRPW Corepressor Recruitment

HES4 is a 221-residue member of the mammalian Drosophila hairy and Enhancer of split (HES1–HES7) family of transcriptional repressors. Like HES1, HES4 contains three signature structural modules:

1. N-Terminal Basic Helix-Loop-Loop (bHLH) DNA-Binding & Dimerization Domain (residues 31–88): Contains a characteristic conserved proline residue in the basic region that bends the DNA-recognition α-helix, directing homo- or heterodimeric HES4 to bind N-box (5′-CACNAG-3′) and class C E-box (5′-CACGCG-3′) sites in target promoters.
2. Central Orange Domain (residues 96–139): A two-helix bundle that dictates partner specificity for heterodimerization with other HES or HEY factors (HES1–HES4 dimers).
3. C-Terminal WRPW Tetrapeptide Motif (Trp-Arg-Pro-Trp, residues 218–221): Directly binds the C-terminal WD40 β-propeller domain of TLE1–TLE4 (Transducin-Like Enhancer of Split / mammalian Groucho) corepressors, recruiting Histone Deacetylases (HDAC1/HDAC2) and Sirtuins (SIRT1) to deacetylate histones and condense target chromatin.

Why Mouse Models Miss Human HES4 Biology in T-Cell and Neural Crest Specification

Comparative phylogenetics and synteny analysis across vertebrate genomes demonstrate that HES4 originated early in gnathostome evolution (present in zebrafish Danio rerio, Xenopus, chicken, dog, cow, and primates), but was deleted via subtelomeric chromosomal rearrangement in the ancestor of Muroidea rodents (Mus musculus and Rattus norvegicus).

Because standard mouse immunology and neurobiology models lack Hes4, human-specific single-cell RNA-seq and ATAC-seq studies were required to uncover why hominids retained HES4:

1. Early Human Thymocyte and T-Lineage Commitment: When human CD34+ hematopoietic progenitor cells enter the thymus and engage DLL4 (Delta-like ligand 4)–Notch1, HES4 and HES1 are the two earliest Notch1 target genes induced. However, HES4 and HES1 play non-redundant roles: HES4 specifically represses alternative myeloid (CEBPA, SPI1/PU.1) and natural killer cell fates at the earliest double-negative (CD34+ CD1a−) stage while permitting B-lineage potential prior to full Notch1 commitment, whereas HES1 enforces later CD34+ CD1a+ T-cell specification.
2. Neural Crest, Osteogenesis, and 1p36 Deletion Syndrome: In human cranial neural crest and mesenchymal stem cells, HES4 interacts with TWIST1 and RUNX2 to balance osteogenic differentiation; in high-grade osteosarcoma, constitutive Notch activation of HES4 blocks terminal osteoblast differentiation and correlates with pulmonary metastasis. Moreover, terminal deletions of Chr1p36.33 (Monosomy 1p36 syndrome, the most common terminal deletion syndrome in humans, ~1 in 5,000 births) hemizygously delete HES4, ISG15, and AGRN, contributing to craniofacial dysmorphism, congenital heart defects, and developmental delay.

Why do ISG15-null mice die rapidly from viral infections whereas ISG15-null humans are unusually resistant to viruses?

The opposite viral phenotypes of ISG15 deficiency in mice versus humans result from a single evolutionary divergence in ISG15–USP18 binding affinity. In mice, USP18 (the negative-feedback inhibitor of the IFNAR2 receptor) is intrinsically stable and does not require Isg15 for protection from proteasomal degradation; therefore, deleting Isg15 in mice eliminates antiviral ISGylation without increasing baseline interferon signaling, rendering Isg15^−/− mice vulnerable to influenza, herpes, and sindbis viruses. In humans, intracellular free ISG15 is required to stabilize USP18 against SKP2-mediated degradation. Consequently, ISG15-deficient human cells lose USP18, experience sustained hyper-activation of the entire STAT1/STAT2 interferon transcriptome (OAS, MX1, PKR, IFIT), and resist broad-spectrum viral replication.

Why is ISG15 one of the highest-ranked diagnostic biomarkers distinguishing bacterial from viral febrile illness and systemic lupus erythematosus (SLE)?

Because the ISG15 promoter on Chr1p36.33 resides in constitutively poised subtelomeric euchromatin with tandem high-affinity ISRE sites, ISG15 mRNA and intracellular/extracellular ISG15 protein rise within 2 hours of Type I (IFN-α/β) or Type III (IFN-λ) interferon exposure—reaching 100-fold to 1,000-fold induction in whole blood during acute viral infections (SARS-CoV-2, influenza, RSV) and active SLE or dermatomyositis flares ("IFN signature"), while remaining near baseline during pure extracellular bacterial infections (which primarily induce IL-6, IL-1β, and CRP rather than Type I IFN).

Why did the house mouse (Mus musculus) lose HES4 while retaining HES1?

In vertebrate subtelomere evolution, terminal chromosome regions undergo elevated rates of double-strand break repair, non-allelic homologous recombination, and Robertsonian translocation. During the karyotypic reshuffling that generated the acrocentric rodent chromosomes of the Muroidea lineage, the ancestral syntenic block corresponding to human Chr1p36.33 underwent a breakpoint between Hes4 and Isg15/Agrn, deleting Hes4 while transferring Isg15 and Agrn to mouse chromosome 4 (Chr4 E2). Because murine Hes1 (on mouse Chr16) compensated for core Notch oscillatory repression during mouse somitogenesis and neurogenesis, the loss of Hes4 was tolerated in murids.

1. Durfee, L. A., Lyon, N., Seo, K., & Huibregtse, J. M. (2010). "The ISG15 conjugation system broadly targets newly synthesized proteins: implications for the antiviral function of ISG15." Molecular Cell, 38(5), 722–732. https://doi.org/10.1016/j.molcel.2010.05.002
2. Bogunovic, D., Byun, M., Durfee, L. A., et al. (2012). "Mycobacterial disease and impaired IFN-γ immunity in humans with inherited ISG15 deficiency." Science, 337(6102), 1684–1688. https://doi.org/10.1126/science.1224026
3. Zhang, X., Bogunovic, D., Payelle-Brogard, B., et al. (2015). "Human intracellular ISG15 prevents interferon-α/β over-amplification and auto-inflammation." Nature, 517(7532), 89–93. https://doi.org/10.1038/nature13801
4. De Obaldia, M. E., Bell, J. J., Wang, X., et al. (2013). "T cell development requires constraint of the myeloid regulator C/EBP-α by the Hes1 and Hes4 target genes of Notch." Nature Immunology, 14(12), 1277–1284. https://doi.org/10.1038/ni.2760