SPI1 / SPIB / SPIC (PU.1)
SPI1 (PU.1), SPIB and SPIC are the class III branch of the ETS family. They read the same GGA(A/T) core as other ETS proteins but demand a purine-rich 5' flank, giving the hematopoietic PU-box AAAGAGGAAGTG - the one ETS subfamily that a scan can usually tell apart from the rest. SPI1 is required for macrophage, neutrophil, dendritic cell and B-cell development; SPIB acts in B cells and plasmacytoid dendritic cells; SPIC specifies red pulp macrophages. PU.1 also nucleates the ETS-IRF composite element (EICE), GGAAGTGAAAC, by recruiting IRF4 or IRF8 to DNA it has already bound. Reduced PU.1 dosage causes acute myeloid leukemia in mice, and SPI1 fusions occur in high-risk pediatric T-cell acute lymphoblastic leukemia.
The factors
Three human genes - SPI1 (PU.1, Spi-1), SPIB (Spi-B) and SPIC (Spi-C) - form the SPI subfamily. Each carries one ETS domain of about 85 residues with the family's winged helix-turn-helix fold and binds DNA as a monomer 8602247; outside the domain they have acidic and glutamine-rich activation regions and, in PU.1, a PEST region whose Ser148 must be phosphorylated for cooperative assembly with IRF partners 7797077. Systematic profiling of all human and mouse ETS proteins puts these three in their own class (class III), the most sequence-distinct group in the family 20517297,21548782; a tyrosine that sets specificity at the position immediately 3' of the core in other ETS proteins is replaced by asparagine here 20517297. Expression is hematopoietic and essentially non-overlapping with the rest of the family: SPI1 in myeloid cells, macrophages, dendritic cells, microglia and B cells; SPIB in B cells and plasmacytoid dendritic cells; SPIC narrowly in splenic red pulp macrophages, with none detected in monocytes, dendritic cells or the other tissue macrophages examined 19037245, and later also in a heme-induced, VCAM1-positive bone marrow macrophage subset 24630724. The first crystal structure of any ETS domain on DNA was PU.1's, which established the loop-helix-loop mode of DNA contact for the whole family 8602247.
The motif
The site is usually written AAAGAGGAAGTG, and the models grouped here agree on AAAGAGGAAGT. The invariant part is the same GGA(A/T) core all ETS proteins read, contacted by two invariant arginines in recognition helix H3 that hydrogen-bond the two guanines, with a wing lysine 5' of the core and a turn lysine on the opposite strand 3' of it 8602247. What sets this family apart is the 5' flank. Class I ETS proteins prefer ACC immediately 5' of GGA 20517297; the SPI proteins prefer purines - AGA, with a marked preference for G two bases 5' of the core - and tolerate a run of As further out 2180582,20517297,21548782. Because that difference spans three or four positions rather than one, class III motifs are the only ETS subfamily that a sequence scan separates from the rest with any reliability 20517297. Within the family, SPI1, SPIB and SPIC are not separable: their models are near-identical, and the Vierstra archetype that represents them is named for two of the three factors at once (AC0642:SPIB/SPI). Note that the published PU.1 site contains no CpG, so unlike the class I CCGGAAGT it is insensitive to CpG methylation. The composite EICE, GGAAGTGAAAC, is a PU.1 site fused to an IRF half-site 7797077,12453417 and is captured by only one model in this family - the CIS-BP entry M07944, whose consensus AAAGAGGAAGTGAAA runs from the PU.1 site into the IRF half-site.
Where it acts
PU.1 sites are found at myeloid and B-lineage promoters and, more heavily, at the enhancers that define those lineages; the classic example is the macrophage colony-stimulating factor receptor (CSF1R) promoter, where PU.1 directs macrophage-specific expression 8264604. PU.1 behaves as a lineage "pioneer" in the sense that its binding precedes and is required for much of the accessible chromatin landscape of myeloid cells 20513432, so its motif is enriched in macrophage and B-cell DNase and ATAC footprints far more than in other cell types. Two named composite elements dominate its cooperative biology. The EICE (ETS-IRF composite element, GGAAGTGAAAC) was defined at immunoglobulin light-chain enhancers, where DNA-bound and Ser148-phosphorylated PU.1 recruits IRF4 (originally Pip) or IRF8 7797077,12453417; cooperativity between the two DNA-binding domains has been measured at 20- to 40-fold 12453417. The crystal structure of the PU.1/IRF4/DNA complex shows the DNA bent into an S shape that juxtaposes the two domains across the central minor groove, and explains why IRF1 and IRF2 are anti-cooperative with PU.1 at the same element 12453417. PU.1 sites also cluster with CEBP, RUNX and GATA sites in myeloid enhancers.
Biology and disease
Spi1-null mice lack macrophages, neutrophils and B cells and have severely disrupted myeloid and lymphoid development 8079170,8896458, the most striking single-gene hematopoietic phenotype among the ETS family. Dosage matters as much as presence: reducing PU.1 expression to roughly a fifth of normal in mice causes acute myeloid leukemia 15146183, placing SPI1 among the transcription factors whose partial loss is leukemogenic. SPIC is required for red pulp macrophage development and splenic iron recycling 19037245. In human disease, recurrent SPI1 fusions (STMN1-SPI1, TCF7-SPI1) define a high-risk subgroup of pediatric T-cell acute lymphoblastic leukemia 28671687, and PU.1 activity is a recurring theme in myeloid leukemia biology more broadly.
Reading a hit
A purine-rich AAAGAGGAAGTG match is one of the few ETS hits that genuinely localizes the signal to a subfamily - but only to the subfamily. SPI1, SPIB and SPIC cannot be told apart by sequence, and the atlas's representative model for this family is itself named for two factors. The distinction from class I ETS is real but not absolute: degenerate members of this family, such as HOCOMOCO's short SPI1.H14CORE.1.S.B model whose consensus is only GGAAGT, will match generic ETS sites, and the GGAA core alone is expected millions of times per strand in a 3.1-Gb genome. Cell type carries much of the inference: outside myeloid, B-lineage or microglial contexts a PU-box match is far more likely to be a chance GGAA than a bound PU.1 site. Two caveats specific to this family: only one of its twelve models (CIS-BP M07944, AAAGAGGAAGTGAAA) reaches into the IRF half-site, so scanning with the rest will find the PU.1 half of an EICE without flagging the composite - look explicitly for GTGAAAC a few bases 3' of the core; and the JASPAR SPIB model (MA0081.3) is stored on the reverse strand as CACTTCCTCTT, the same site as AAGAGGAAGTG.
How this was worked out
PU.1 is cloned as a macrophage- and B-cell-specific activator that recognizes the purine-rich PU box, GAGGAA, through a carboxy-terminal DNA-binding domain related to the ets oncogene.
Still open then: How the domain read the PU box, and what set it apart from other ETS sites, was unknown.
Two Spi1 knockouts establish PU.1 as essential for blood development: one gives late-gestation lethality with a multilineage defect in lymphoid and myeloid progenitors; the second (1996) gives live-born pups lacking macrophages, neutrophils, B and T cells, in which T cells and a few neutrophils later appear while B cells and macrophages never do.
Still open then: The two lines disagreed on whether PU.1 acts at lineage commitment of a multipotential progenitor or only at later B-cell and macrophage differentiation.
Pip (IRF4) is cloned as the lymphoid partner that binds the immunoglobulin light-chain enhancer composite element only alongside DNA-bound PU.1, and complex formation requires phosphorylation of PU.1 Ser148.
The PU.1 ETS domain-DNA crystal structure, the first for the family, shows a winged helix-turn-helix contacting ten base pairs through a loop-helix-loop, with two recognition-helix arginines on the GGA core and conserved wing and turn lysines on either side.
Still open then: The structure did not explain the purine-rich 5' flank that distinguishes the SPI proteins.
The PU.1/IRF4/DNA ternary structure shows the composite element bent into an S shape that juxtaposes the two DNA-binding domains across the central minor groove, accounting for cooperativity with IRF4 and IRF8 and anti-cooperativity with IRF1 and IRF2.
Hypomorphic alleles that lower PU.1 to about 20% of normal cause acute myeloid leukemia in mice, whereas complete or 50% loss does not, so graded reduction rather than absence of PU.1 is leukemogenic.
PU.1, together with small sets of lineage-determining partners (C/EBP in macrophages, E2A, EBF and Oct-2 in B cells), establishes most promoter-distal H3K4me1 regions; PU.1 binding initiates nucleosome remodeling followed by H3K4 monomethylation.
Still open then: Whether PU.1 can open closed sites alone, or only with its collaborating factors, was not resolved.
- 2017mediumTranscriptome and targeted capture sequencing, transduction of mouse stem/progenitor cells28671687
Recurrent STMN1-SPI1 and TCF7-SPI1 fusions mark 3.9% (7 of 181) of pediatric T-cell acute lymphoblastic leukemias, a distinct subgroup with uniformly poor survival; the fusion proteins keep transcriptional activity and block maturation of mouse progenitors.
Still unresolved
- The base immediately 5' of the GGA core is reported inconsistently: the PU box and the ChIP-derived models read GAGGAA, whereas systematic in vitro profiling characterizes class III by G at -2 and C at -1, and one HT-SELEX SPIB model here reads AAAAGCGGAAGT; if C is genuinely favored, the site can contain a CpG, and the methylation insensitivity of the classic PU.1 site would not extend to it.
- How much PU.1 alone selects myeloid and B-cell enhancers is unsettled: its binding initiates nucleosome remodeling and H3K4 monomethylation at many distal sites, yet those sites are established together with lineage partners such as C/EBP in macrophages and E2A in B cells, and whether PU.1 can open closed chromatin without them is not settled by the studies cited here.
This is one of the few ETS subfamilies a scan can separate from the others, because its purine-rich 5' flank differs over three or four positions rather than one. It cannot separate its own members: SPI1, SPIB and SPIC share one motif, and the atlas representative is a Vierstra archetype named for two of them. Expression usually decides - SPIC is confined to iron-recycling macrophages of the splenic red pulp and bone marrow, SPIB to B cells and plasmacytoid dendritic cells - so a PU-box hit outside a hematopoietic context deserves scepticism. The family is small (12 models) and includes short, degenerate entries that behave like a generic ETS site. Only one of the twelve models reaches into the IRF half-site of the EICE composite, so a scan will usually register the PU.1 half-site and miss the element that gives it specificity.
References
- Kodandapani R, Pio F, Ni CZ, et al. A new pattern for helix-turn-helix recognition revealed by the PU.1 ETS-domain-DNA complex. Nature 1996;380:456-60. PMID 8602247
- Wei GH, Badis G, Berger MF, et al. Genome-wide analysis of ETS-family DNA-binding in vitro and in vivo. EMBO J 2010;29:2147-60. PMID 20517297
- Hollenhorst PC, McIntosh LP, Graves BJ. Genomic and biochemical insights into the specificity of ETS transcription factors. Annu Rev Biochem 2011;80:437-71. PMID 21548782
- Zhang DE, Hetherington CJ, Chen HM, et al. The macrophage transcription factor PU.1 directs tissue-specific expression of the macrophage colony-stimulating factor receptor. Mol Cell Biol 1994;14:373-81. PMID 8264604
- Scott EW, Simon MC, Anastasi J, et al. Requirement of transcription factor PU.1 in the development of multiple hematopoietic lineages. Science 1994;265:1573-7. PMID 8079170
- McKercher SR, Torbett BE, Anderson KL, et al. Targeted disruption of the PU.1 gene results in multiple hematopoietic abnormalities. EMBO J 1996;15:5647-58. PMID 8896458
- Eisenbeis CF, Singh H, Storb U. Pip, a novel IRF family member, is a lymphoid-specific, PU.1-dependent transcriptional activator. Genes Dev 1995;9:1377-87. PMID 7797077
- Escalante CR, Brass AL, Pongubala JM, et al. Crystal structure of PU.1/IRF-4/DNA ternary complex. Mol Cell 2002;10:1097-105. PMID 12453417
- Kohyama M, Ise W, Edelson BT, et al. Role for Spi-C in the development of red pulp macrophages and splenic iron homeostasis. Nature 2009;457:318-21. PMID 19037245
- Rosenbauer F, Wagner K, Kutok JL, et al. Acute myeloid leukemia induced by graded reduction of a lineage-specific transcription factor, PU.1. Nat Genet 2004;36:624-30. PMID 15146183
- Seki M, Kimura S, Isobe T, et al. Recurrent SPI1 (PU.1) fusions in high-risk pediatric T cell acute lymphoblastic leukemia. Nat Genet 2017;49:1274-1281. PMID 28671687
- Klemsz MJ, McKercher SR, Celada A, et al. The macrophage and B cell-specific transcription factor PU.1 is related to the ets oncogene. Cell 1990;61:113-24. PMID 2180582
- Heinz S, Benner C, Spann N, et al. Simple combinations of lineage-determining transcription factors prime cis-regulatory elements required for macrophage and B cell identities. Mol Cell 2010;38:576-89. PMID 20513432
- Haldar M, Kohyama M, So AY, et al. Heme-mediated SPI-C induction promotes monocyte differentiation into iron-recycling macrophages. Cell 2014;156:1223-1234. PMID 24630724
12 motifs in this family — JASPAR 3 · HOCOMOCO 5 · CIS-BP 3 · Vierstra 1
| motif | database | consensus | bits |
|---|---|---|---|
| SPIB.H14CORE.2.SM.BSMB | HOCOMOCO | gAAtGAGGAAGTa | 19.4 |
| SPIB.H14CORE.1.S.CSC | HOCOMOCO | AAAAGCGGAAGT | 18.9 |
| M07944_2.00 | CIS-BP | AAAGAGGAAGTGAaA | 18.7 |
| SPIB.H14CORE.0.P.BPB | HOCOMOCO | AAAGAGGAAGTGa | 17.2 |
| SPI1.H14CORE.0.P.BPB | HOCOMOCO | AAAGAGGAAGTG | 16.1 |
| Spi1 (MA0080.7) | JASPAR | AAAGAGGAAGTG | 14.5 |
| SPIB (MA0081.3) | JASPAR | tCACTTCCTCTT | 14.3 |
| SPIC (MA0687.2) | JASPAR | AAAGaGGAAGTA | 13.2 |
| M03005_2.00 | CIS-BP | AAAGaGGAAGTA | 13.2 |
| AC0642:SPIB/SPI:Ets | Vierstra | AAAGaGGAAGT | 12 |
| SPI1.H14CORE.1.S.BSB | HOCOMOCO | GGAAGT | 9.8 |
| M02678_2.00 | CIS-BP | AgaGGAA | 9.1 |
HOCOMOCO tags read as assay + quality grade: P ChIP-seq, S HT-SELEX, M methyl-HT-SELEX, G GHT-SELEX, I SMiLE-seq, B PBM; grade A means two or more assay types agreed. A P-only matrix is a record of where the protein was crosslinked, which need not be a sequence it binds itself.
The factors, elsewhere
These entries are written from the primary literature and every citation is checked against PubMed, but this one can still be out of date, too confident, or simply wrong — especially where a motif’s published site and its database matrices disagree. Corrections and additions are welcome, and they are the fastest way to make this page better for everyone.