Phenotypic Mutation 'woodrat' (pdf version)
Mutation Type missense
Coordinate119,529,030 bp (GRCm38)
Base Change T ⇒ C (forward strand)
Gene Mbtps1
Gene Name membrane-bound transcription factor peptidase, site 1
Synonym(s) subtilisin/kexin isozyme-1, SKI-1, site-1 protease, S1P, 0610038M03Rik
Chromosomal Location 119,508,156-119,558,735 bp (-)
MGI Phenotype FUNCTION: [Summary is not available for the mouse gene. This summary is for the human ortholog.] This gene encodes a member of the subtilisin-like proprotein convertase family, which includes proteases that process protein and peptide precursors trafficking through regulated or constitutive branches of the secretory pathway. The encoded protein undergoes an initial autocatalytic processing event in the ER to generate a heterodimer which exits the ER and sorts to the cis/medial-Golgi where a second autocatalytic event takes place and the catalytic activity is acquired. It encodes a type 1 membrane bound protease which is ubiquitously expressed and regulates cholesterol or lipid homeostasis via cleavage of substrates at non-basic residues. Mutations in this gene may be associated with lysosomal dysfunction. [provided by RefSeq, Feb 2014]
PHENOTYPE: Mice homozygous for a gene trap allele die prior to implantation. Mice homozygous for an ENU-induced allele exhibit hypopigmentation, reduced female fertility, altered lipid homeostasis, and increased susceptibility to induced colitis. [provided by MGI curators]
Accession Number

NCBI RefSeq: NM_019709; MGI: 1927235

Amino Acid Change Tyrosine changed to Cysteine
Institutional SourceBeutler Lab
Gene Model not available
AlphaFold Q9WTZ2
SMART Domains Protein: ENSMUSP00000080117
Gene: ENSMUSG00000031835
AA Change: Y496C

signal peptide 1 22 N/A INTRINSIC
Pfam:Peptidase_S8 209 464 1.5e-43 PFAM
transmembrane domain 1000 1022 N/A INTRINSIC
Predicted Effect probably damaging

PolyPhen 2 Score 1.000 (Sensitivity: 0.00; Specificity: 1.00)
(Using ENSMUST00000081381)
SMART Domains Protein: ENSMUSP00000095965
Gene: ENSMUSG00000031835
AA Change: Y496C

signal peptide 1 22 N/A INTRINSIC
Pfam:Peptidase_S8 213 473 3.7e-45 PFAM
transmembrane domain 1000 1022 N/A INTRINSIC
Predicted Effect probably damaging

PolyPhen 2 Score 1.000 (Sensitivity: 0.00; Specificity: 1.00)
(Using ENSMUST00000098362)
Meta Mutation Damage Score Not available question?
Is this an essential gene? Essential (E-score: 1.000) question?
Phenotypic Category Autosomal Recessive
Candidate Explorer Status CE: no linkage results
Single pedigree
Linkage Analysis Data
Penetrance 100% 
Alleles Listed at MGI
All alleles(30) : Targeted, knock-out(1) Targeted, other(1) Gene trapped(27) Chemically induced(1)
Lab Alleles
AlleleSourceChrCoordTypePredicted EffectPPH Score
Muskrat UTSW 8 119538137 missense probably damaging 1.00
packrat UTSW 8 119528961 missense probably damaging 1.00
R0194:Mbtps1 UTSW 8 119535369 missense probably damaging 1.00
R0270:Mbtps1 UTSW 8 119538117 splice site probably benign
R0485:Mbtps1 UTSW 8 119522601 splice site probably benign
R1269:Mbtps1 UTSW 8 119520277 missense probably damaging 1.00
R1351:Mbtps1 UTSW 8 119518162 missense possibly damaging 0.95
R1536:Mbtps1 UTSW 8 119546125 missense probably benign 0.01
R1542:Mbtps1 UTSW 8 119546247 splice site probably null
R1543:Mbtps1 UTSW 8 119542069 splice site probably benign
R1580:Mbtps1 UTSW 8 119538900 missense possibly damaging 0.79
R1587:Mbtps1 UTSW 8 119518219 missense probably damaging 0.96
R1715:Mbtps1 UTSW 8 119542730 missense probably benign 0.40
R1845:Mbtps1 UTSW 8 119522493 missense probably benign 0.13
R2147:Mbtps1 UTSW 8 119538859 missense probably benign 0.01
R2157:Mbtps1 UTSW 8 119542727 missense probably benign 0.01
R2416:Mbtps1 UTSW 8 119538917 missense probably damaging 1.00
R2910:Mbtps1 UTSW 8 119546037 missense possibly damaging 0.82
R2911:Mbtps1 UTSW 8 119546037 missense possibly damaging 0.82
R3079:Mbtps1 UTSW 8 119531205 missense probably benign 0.40
R3079:Mbtps1 UTSW 8 119538863 missense probably damaging 1.00
R3080:Mbtps1 UTSW 8 119531205 missense probably benign 0.40
R3080:Mbtps1 UTSW 8 119538863 missense probably damaging 1.00
R4116:Mbtps1 UTSW 8 119541652 missense probably benign 0.00
R4296:Mbtps1 UTSW 8 119522499 missense possibly damaging 0.95
R4602:Mbtps1 UTSW 8 119535347 missense probably damaging 1.00
R4603:Mbtps1 UTSW 8 119535347 missense probably damaging 1.00
R4610:Mbtps1 UTSW 8 119535347 missense probably damaging 1.00
R4611:Mbtps1 UTSW 8 119535347 missense probably damaging 1.00
R4729:Mbtps1 UTSW 8 119525420 missense probably damaging 1.00
R4868:Mbtps1 UTSW 8 119508928 missense probably benign 0.01
R4893:Mbtps1 UTSW 8 119518193 missense probably damaging 1.00
R4999:Mbtps1 UTSW 8 119533348 missense probably damaging 1.00
R6056:Mbtps1 UTSW 8 119515602 missense probably benign
R6062:Mbtps1 UTSW 8 119531091 missense possibly damaging 0.94
R6237:Mbtps1 UTSW 8 119528961 missense probably damaging 1.00
R6617:Mbtps1 UTSW 8 119538137 missense probably damaging 1.00
R7215:Mbtps1 UTSW 8 119524568 missense possibly damaging 0.82
R7275:Mbtps1 UTSW 8 119542750 missense probably benign
R7794:Mbtps1 UTSW 8 119538884 missense probably damaging 1.00
R8029:Mbtps1 UTSW 8 119547805 start gained probably benign
R8104:Mbtps1 UTSW 8 119529055 missense possibly damaging 0.85
R8205:Mbtps1 UTSW 8 119520338 missense probably damaging 1.00
R8351:Mbtps1 UTSW 8 119546184 missense probably benign 0.01
R8487:Mbtps1 UTSW 8 119541674 missense probably damaging 1.00
R8753:Mbtps1 UTSW 8 119508862 missense possibly damaging 0.94
R9155:Mbtps1 UTSW 8 119508954 missense probably benign 0.06
R9168:Mbtps1 UTSW 8 119521863 missense probably benign 0.01
R9172:Mbtps1 UTSW 8 119533369 missense probably damaging 1.00
RF019:Mbtps1 UTSW 8 119525550 missense probably damaging 1.00
X0017:Mbtps1 UTSW 8 119531124 missense probably damaging 1.00
X0027:Mbtps1 UTSW 8 119522547 missense probably damaging 1.00
Mode of Inheritance Autosomal Recessive
Local Stock Live Mice, Embryos, gDNA
MMRRC Submission 012829-UCD
Last Updated 2022-04-12 3:40 PM by External Program
Record Created unknown
Record Posted 2009-01-22
Phenotypic Description
Figure 1.  The woodrat coat color phenotype at four months of age.

The woodrat (wrt) phenotype was identified as a visible recessive variant among G3 mice homozygous for mutations induced by ENU (1). Homozygous woodrat mice exhibit coat hypopigmentation beginning on postnatal day 8, which progresses with time so that by 4 weeks of age wrt/wrt coats are composed of both white and black hairs, varying in pigment content along the length of each hair shaft. Adult wrt/wrt mice have a more homogeneous coat composed of hairs with a white base and a grey/brown tip (Figure 1). Studies confirmed that the wrt mutation led to the observed progressive hypopigmentation (2).


Compound heterozygosity for a null mutation (a gene trap allele) and the wrt allele (Mbtps1wrt/trap) results in embryonic lethality, while 53% and 54% of the expected numbers of Mbtps1+/trap or Mbtps1wrt/wrt mice, respectively, survive to term (2). Because a single copy of the gene (Mbtps1+/trap) is associated with diminished survival to term and causes no visible phenotype, while Mbtps1wrt/wrt is associated with about the same impairment of survival to term but does cause a visible phenotype, it is inferred that the wrt allele encodes a protein with ≤50% of wild type activity. In addition, it appears that the survival threshold likely falls between 25% and 50% of wild type enzyme activity levels.  It was determined that the wrt allele causes maternal-zygotic effect lethality (i.e. homozygous females can only give birth to offspring carrying at least one wild-type copy of the Mbtps1 locus) (2).  Further examination determined that death and resorption of maternal-zygotic mutant embryos (homozygote embryos derived from homozygote mothers) occurs before the eighth day of gestation (2).


Figure 2.  Increased susceptibility to DSS-induced colitis in woodrat mice.  (A) Percent initial weight of C57BL/6J and woodrat (Mbtps1wrt/wrt) mice on the indicated days after administration of 2% DSS.  C57BL/6J mice receiving 0% DSS served as controls.  n = 9-20 mice per group. *p<0.05, **p<0.005, two tailed Mann-Whitney test for C57BL/6J on 2% DSS versus woodrat.  (B) Representative colons and cecums from untreated C57BL/6J mice (top), C57BL/6J mice treated for 9 days with 2% DSS (center), and woodrat mice treated for 9 days with 2% DSS (bottom).  Figure reproduced from reference (1).
Figure 3. Mbtps1wrt (wrt) are resistant to LCMV Cl13 infection despite immunodeficiency. (A) Mbtps1wrt are resistant to persistent LCMV Cl13 infection but the kinetics of viremia are not altered during infection with the acute LCMV Armstrong (ARM) parent strain. Mean and SEM are shown from 6–12 mice/group at the indicated times post infection. (B) GP33 ex vivo peptide stimulation and (C) viral peptide specific T cell enumeration with tetramers from splenocytes 5dpi. (D) Mbtps1wrt mount a CTL immune response to LCMV Cl13, 9 dpi, when WT CTL exhibit exhaustion. GP33 and (E) NP396 ex vivo peptide stimulation from splenocytes. Mean and SEM are shown from 6 mice/group. *=P-value<0.05. **=P-value<0.005. Obtained from (3).
Figure 4.  Mbtps1wrt are susceptible to establishment of LCMV Cl13 infection in serum, spleen, liver and kidney, but have an early sustained defect in the bone marrow.

A) Mbtps1wrt are susceptible to LCMV Cl13 replication in serum, spleen, liver and kidney. Mean and SEM are shown from 4–6 mice/group at the indicated times post infection. B) In the bone marrow, monocytes and DC have defective establishment of LCMV Cl13 in Mbtps1wrt. Mean and SEM are shown from 5 mice/group, 5 dpi. *=P-value<0.05. **=P-value<0.005. Obtained from (3).

Figure 5. Adoptive immunotherapy with Mbtps1wrt bmDC eliminates viral persistence

B) wrt bmDC have higher levels of MHC molecules (class I and class II) and costimulatory molecules (CD80 and CD86). Flow cytometric analysis of bmDC from WT and wrt mice are shown with Mean and SEM for both % of cells in highly activated gate as well as gMFI for MHC and costimulatory molecules from one of 2 similar independent experiments (N=4). *=P-value<0.05. Obtained from (3).

When challenged with 2% DSS in drinking water (DSS-induced Colitis Screen), wrt homozygotes develop severe and sometimes fatal colitis, as measured by weight loss relative to wild type controls (Figure 2A).  DSS-induced weight loss is similar in wrt/wrt and wild type mice when they are also treated with antibiotics, indicating that the enhanced susceptibility of wrt mice to DSS-induced colitis depends on commensal flora. In wrt homozygotes, nine days of 2% DSS treatment causes bleeding in the colon, and a shortening and thickening of the colon indicative of edema and inflammation, which were not observed in wild type mice (Figure 2B). Histological analysis demonstrates leukocyte recruitment and crypt loss in the colons of DSS-treated wrt/wrt animals. In addition, wrt/wrt mice have increased levels of interleukin (IL)-1β and IL-6 in epithelium and lamina propria both before and after seven days of DSS administration, as measured in the supernatants of distal colonic explants after 24 hours in culture. The gut epithelial barrier was found to be intact in naïve wrt/wrt mice. 


The unfolded protein response (UPR) was investigated in wrt homozygous and heterozygous mice by determining the protein levels of glucose-regulated protein (GRP)78 (also known as BiP and Hspa5) and GRP94. Colons of wild type mice, wrt heterozygotes, and wrt homozygotes showed similar levels of GRP78 and GRP94 before DSS treatment, but wrt homozygotes had significantly lower levels of both proteins after three days of DSS administration. Interestingly, in wrt heterozygotes, GRP78 and GRP94 protein levels were intermediate between wild type and homozygous mutant levels. GRP78 and GRP94 mRNA expression levels are reduced in colonic epithelial cells from wrt mice even before DSS treatment. Using chimeras made by reciprocal bone marrow transplantation between wild type and wrt mice, both DSS-induced colitis and reduced colon levels of GRP78 and GRP94 proteins were found to be conferred by non-hematopoietic wrt mutant cells. Induction of the UPR by intraperitoneal injection of tunicamycin is markedly more toxic to wrt mice, with 10/11 dying within 30 hours of treatment, compared to wild type mice, where only 1/11 died during the same time. Tunicamycin treatment causes colitis in wrt, but not wild type mice.


Lymphocytic choromeningitis virus (LCMV) is a pathogenic arenavirus, and it depends on S1P to cleave its envelope glycoprotein.  Without S1P, the virus cannot replicate (3). Following infection with a persistant variant of LCMV Armstrong strain, LCMV clone 13 (LCMV Cl13), woodrat animals were found to resist persistent infection. The LCMV Cl13 strain exhibit extended persistence in wild-type animals, LCMV Cl13 was not detected in the woodrat animals at 3 weeks post-infection (3).   Examination of T cell number and activity as well as antibody responses in the woodrat animals determined that the animals had reduced immune responses compared to wild-type animals (Figure 3).  To understand the protective effect of the mutation, the animals were supplemented with a 5% cholesterol diet and subsequently exposed to LCMV Cl13.  Although cholesterol has been shown to be essential for the viral life cycle and host immune responses (4-6), cholesterol repletion did not revert woodrat resistance. Examination of interferon (IFN)alpha levels did not find a significant increase in the woodrat animals, indicating that IFN production was not leading to woodrat resistance.  Examination of bone marrow cells found that woodrat dendritic cells (DCs) and their monocyte precursors were resistant to infection by LCMV Cl13; the splenic DC reaction to LCMV Cl13 was equivalent to wild-type (Figure 4). The differential phenotype between the bone marrow and splenic DCs indicates that there might be differences in the expression of Mbtps1.  Indeed, examination of Mbtps1 expression found that there are decreased levels in the bone marrow; target genes of Mbtps1 were also decreased in the woodrat bone marrow.  Examination of LCMV Cl13 viral replication in woodrat bone marrow-derived macrophages and fibroblasts showed that there is no difference in viral growth kinetics when compared to wild-type (3).   In contrast, viral replication was almost completely null in the bone marrow-derived DCs.  Adoptive transfer of woodrat bone marrow-derived DCs to wild-type hosts conferred resistance of persistent viral infection (3) (Figure 5B).

Nature of Mutation
The wrt mutation was mapped to Chromosome 8, and corresponds to an A to G transition at position 2041 of the Mbtps1 transcript, in exon 13 of 24 total exons.
491  -L--T--E--C--P--Y--M--W--P--Y--C-
The mutated nucleotide is indicated in red lettering, and results in a tyrosine to cysteine change at residue 496 of the Site 1 protease (S1P).
Illustration of Mutations in
Gene & Protein
Protein Prediction
Figure 6.  Domain structure of S1P. (A) The signal peptide, prosegment, catalytic domain, growth factor/cytokine receptor family motif (GFCR), region encoded by exon 21, and transmembrane domain (TM) are indicated.  The positions of the catalytic triad residues (D218, H249, S414), and N338 of the oxyanion hole are shown above the diagram.  Amino acid sequences surrounding cleavage sites (downward arrows) for generation of S1P-B, S1P-C, and the shed form of S1P are shown.  Note that in addition to the major cleavage site after K137, cleavage after F133 has also been shown to occur, although to a lesser extent (). The woodrat mutation results in a tyrosine to cysteine change at residue 496 of the Site 1 protease. This image is interactive. Other mutations found in S1P are noted in red. Click on each mutation for more information. (B) Amino acid numbers and molecular masses are shown for each processed form of S1P.

Mbtps1 encodes the Site 1 protease (S1P), also known as subtilisin-kexin-isoenzyme 1 (SKI-1), which was simultaneously discovered by two groups as the enzyme responsible for the Site 1 cleavage of sterol regulatory element binding proteins (SREBPs), and as a proprotein convertase with specificity for nonbasic amino acids (7;8). The family of proprotein convertases (PCs) comprise nine mammalian serine proteases that convert secretory precursor proteins into active forms by proteolysis at selected sites [reviewed in (9)].  Seven of the nine known PCs are specific for basic amino acids and cleave at single or paired basic residues. In contrast, only two known PCs, S1P and proprotein convertase subtilisin kexin 9 (PCSK9), are specific for nonbasic amino acids. S1P recognizes the motif (R/K)-X-(L/I/V)-Z↓, where X is any amino acid, and Z is any amino acid except Pro, Cys, Glu, Asp or Val (7;10-12). Mouse S1P is 98% identical to that found in rat, and 96% identical to human S1P, with 100% identity among the three catalytic domains (7).


S1P is a type I transmembrane protein with its N-terminal proteolytic domain projecting into the ER lumen. It contains 1052 amino acids, comprising a signal peptide (amino acids 1-17), a prosegment (amino acids 18-186), and a catalytic protease domain (amino acids 218-414) (Figure 6) (7).  Asp, His and Ser residues, known as the catalytic triad, make up the active site of S1P, as is characteristic of all serine proteases; they are located at positions 218, 249, and 414, respectively (7). Asn338 of S1P forms part of an oxyanion hole, another characteristic of serine proteases, which serves to stabilize the transition state and thereby lower the activation energy of the peptide cleavage reaction. Following the catalytic domain, where all seven basic amino acid-specific PCs contain a P-domain, S1P has an extended C-terminal region with unknown function (amino acids 415-1052).  This region contains a conserved growth factor/cytokine receptor family motif (amino acids 849-861), followed by a putative 24-amino acid hydrophobic transmembrane segment, and a highly basic C-terminal tail that may regulate cellular localization (7;13).  S1P may be alternatively spliced into forms lacking amino acids 430-483 or 858-901, regions found within the extended C-terminal tail (7). Notably, amino acids 858-901, encoded by exon 21, encompass part of the growth factor/cytokine receptor family motif. Deletion of exon 21 results in abrogation of prosegment processing, retention in the endoplasmic reticulum (ER), and abrogation of protease activity towards a substrate, suggesting that the motif may be important for the proper folding of the protein that allows its processing (14). S1P functions most efficiently near neutral pH and at 2-3 mM Ca2+ (11;15).


The prosegments of many peptidases, including PCs, act both as intramolecular chaperones and inhibitors of the associated protease. Enzyme activation typically requires release of the prosegment in an organelle-specific manner. Likewise, S1P is a zymogen activated by autocatalytic cleavage of its prosegment in the ER. First, a signal peptidase removes the signal peptide by cleavage at the sequence LVVLLC17↓GKKHLG resulting in S1P-A (amino acids 18-1052), which is inactive (11;12). Then, sequential autocatalytic cleavages at R134SLK137↓YA and R183RLL186↓RA occur to generate the active S1P-B (amino acids 138-1052) and S1P-C (amino acids 187-1052), respectively, plus the major ~24 kDa and 14 kDa prosegment cleavage fragments (11;12). These fragments are further processed into 10- and 8-kDa products. Mutational analysis demonstrates that Arg130 and Arg134 are critical for autocatalytic processing of the prosegment, as well as for the subsequent efficient exit of S1P from the ER (14). Once cleaved, the 14 kDa prosegment fragment reportedly remains tightly bound to S1P, unlike the prosegments of other PCs, which are released by acidic pH and high Ca2+ concentrations found in the trans-Golgi network (16). In addition, only the full length 24-kDa prosegment, but none of the smaller fragments, is inhibitory for S1P protease activity towards a synthetic peptide substrate (11). Based on these findings, the 14 kDa prosegment fragment has been hypothesized to serve a chaperone-like function.


S1P exists as both cell surface transmembrane and extracellularly shed protein forms (7;8). Shedding occurs through cleavage at the sequence KHQKLL953↓SIDL, a non-canonical S1P recognition sequence in which the two leucine residues adjacent to the cleavage site are critical (14). Evidence suggests that S1P catalyzes its own cleavage and shedding, but whether proteolysis occurs in cis or in trans is unknown.  Using cell surface biotinylation followed by immunoprecipitation, S1P has been localized at the cell membrane in transiently transfected HEK293 cells. In addition, soluble S1P accumulates in the cell media, which may be blocked by treatment with brefeldin-A, suggesting that S1P is cleaved and shed from the plasma membrane (14). However, it remains possible that S1P is secreted from cells, as S1P mutants lacking the transmembrane and cytosolic domains fail to be cleaved at the L953 shedding site but are still detected in the cell media (14;15).


The wrt mutation predicts the amino acid substitution Y496C, located within the extended C-terminal tail. Protein expression and catalytic activity of the mutant protein have not been tested. Based on the wrt phenotype, the mutant protein is hypothesized to possess ≤50% of wild type protein activity, but the nature of its dysfunction is unknown (see Phenotypic Description).


Northern blot analysis of mRNA from adult rat demonstrates that Mbtps1 is expressed in most tissues, including brain (striatum hippocampus, hypothalamus), liver, thymus, spleen, kidney, and heart (7). It is also highly expressed in adrenal, anterior pituitary, and thyroid glands. A similar widespread expression pattern is detected in two-day-old rat by in situ hybridization, with additional expression in skin, muscles, cardiac muscles, bones, teeth, and intestine (7).


Full length and processed forms of the membrane-anchored S1P localizes to the endoplasmic reticulum (ER) and Golgi membranes (7;11;12;17). The short C-terminal sequence after the transmembrane domain is oriented in the cytosol, with the bulk of the protein oriented in the lumen (Figure 3). S1P has been found associated with other cytoplasmic compartments that may represent lysosomes and/or endosomes (7). Treatment of cells with brefeldin-A does not alter the production of the prosegment fragments, and S1P-B is endo H-sensitive, indicating that S1P prosegment cleavage occurs in the ER (11;12). S1P-C is endo H-resistant, indicating a localization in the Golgi.  An active, soluble form of S1P generated by cleavage after Leu953 can be shed into the culture medium of cells expressing the complete protein (11;12).  The 24 kDa prosegement fragment may also be shed into the culture medium (12).


PCs play a central role in processing a wide range of protein precursors, including hormones, growth factors, transcription factors, bacterial toxins and viral glycoproteins. Consequently, they have been implicated in multiple disease states, participating in the pathogenesis of dyslipidemias and atherosclerosis (13), Alzheimer’s disease (18), tumorigenesis (19), and several bacterial and viral infections (20-23).


Since its discovery in 1998, S1P has been shown to cleave diverse substrates. A group of these substrate proteins are activated from a latent state through regulated intramembrane proteolysis (RIP), a process by which transmembrane proteins are cleaved within a membrane-spanning helix to release cytosolic domains that enter the nucleus (or other cellular compartments) to regulate transcription (or other functions) (24). RIP typically requires an initial cleavage in the extracytoplasmic domain of the substrate before the intramembrane proteolytic event. S1P and the Site 2 protease (S2P) function sequentially to carry out RIP for an expanding list of proteins, with S1P performing the initial extracytoplasmic cleavage (typically in the Golgi), and S2P mediating intramembrane cleavage. RIP substrates for S1P and S2P include the SREBPs (8), and CREB/ATF (cAMP response element binding/activating transcription factor) family proteins including ATF6, CREBH, CREB4, OASIS, and Luman, all of which are transcription factors of the basic leucine zipper family (25-29) (see below). Another major group of S1P substrates consists of the surface glycoproteins of viruses, which includes those of the Arenaviruses lymphocytic choriomeningitis virus (LCMV) and Lassa fever virus, and the Bunyavirus Crimean-Congo hemorrhagic fever virus (20;23;30). S1P has been shown to cleave pro-brain derived neurotrophic factor (BDNF) in vitro (7).

Figure 7.  The SREBP pathway and the ATF6-mediated ER stress pathway. (Left) When sufficient sterols are present in the cell, via interactions between the C-terminal regulatory domain (Reg) on SREBPs and WD motifs on SCAP, the SREBPs form a complex with SCAP and remain in the ER.  Retntion in the ER depends on interaction between the sterol-sensing domain (SSD) of SCAP and either Insig-1 or -2.  The SCAP-Insig-1 interaction is strongly enhanced by cholesterol, while the SCAP-Insig-2 interaction absolutely requires cholesterol.  In the absense of sterols, SCAP and Insig do not interact, and SCAP is freed to transport SREBPs to the Golgi via COPII vesicles.  There, sequential cleavage of SREPBs by S1P and S2P (red lines), releases the basic helix-loop-helix-leucine-zipper transcription factor for translocation to the nucleus and activation of target genes.  Genes encoding acetyl CoA synthetase (Acss2), fatty acid synthetase (Fasn), and the LDL receptor (Ldlr) are among those activated by SREBPs. (Right) Under normal conditions, ATF6 is held in th ER by interactions with the chaperone GRP78.  ER stress, caused by the accumulation of unfolded proteins in the ER, results in release of GRP78, exposing a Golgi localization sequence on ATF6 that targets it to the Golgi in COPII vesicles.  There sequential cleavages by S1P and S2P frees the basic leucine zipper and transactivation domain (TA) of ATF6 to translocate to the nucleus, where it activates genes including CHOP, GRP78, and XBP1.

RIP was first found to occur during the activation of SREBPs, now the paradigm for this type of processing (Figure 7). SREBPs are transcriptional activators of genes responsible for the biosynthesis and uptake of cholesterol and unsaturated fatty acids. SREBPs are bound to the ER membrane and nuclear envelope in a hairpin orientation, with the N and C termini projecting into the cytosol [reviewed in (31)]. The N-terminal segment is a basic helix-loop-helix-leucine zipper transcription factor; the C-terminal segment forms a complex with the membrane-anchored regulatory protein SREBP-cleavage activating protein (SCAP). Between the cytosolic N and C termini is a segment of ~80 amino acids consisting of two membrane-spanning helices separated by a ~30 amino acid ER luminal loop. S1P initiates the processing of SREBP-1 and SREBP-2 by cleaving this loop at the sequence RSVL↓S (32).  S2P subsequently cleaves the N-terminal segment within its transmembrane region, releasing it for entry into the nucleus (33;34). When sterols accumulate in cells, SREBPs remain membrane-anchored in a complex with the sterol-sensing SCAP protein within the ER. Retention in the ER requires either Insig-1 or Insig-2, which bind to the sterol-sensing domain of SCAP (35;36). Thus, cleavage of SREBPs by S1P is blocked, and gene transcription declines. When cellular demand for sterols increases, Insig proteins release SCAP, which then transport SREBPs to the Golgi apparatus via coatamer protein complex II (COPII) vesicles, where the SREBPs encounter active S1P and are cleaved to an active state (37-41).


More recent studies have identified S1P and S2P as the proteases mediating RIP for several ER stress-responsive CREB/ATF transcription factors. ATF6 is an ER-resident transcription factor activated in response to ER stress, which is perceived by cells as an excess of unfolded or misfolded proteins in the ER and triggers the unfolded protein response (UPR). The UPR facilitates the folding, processing, export, and degradation of misfolded proteins in the ER.  ATF6 is a type II transmembrane protein with a basic leucine zipper domain in the cytosol and a stress-sensing domain in the ER lumen (42). Under normal conditions, ATF6 is retained in the ER through interaction with the chaperone GRP78 (43). The accumulation of unfolded or misfolded proteins in the ER results in release of ATF6 and its translocation to the Golgi in COPII vesicles, where it is cleaved by S1P and S2P similarly to SREBPs (28;44;45). The cleaved cytosolic domain traffics to the nucleus, where it binds to sequences known as ER stress responsive elements (ERSE), and activates transcription of UPR target genes including the chaperones GRP78, GRP94, CCAAT/enhancer-binding protein homologous protein (CHOP), and X-box binding protein 1 (XBP1) (Figure 4) (46). In a similar manner, CREBH, CREB4, Luman, and OASIS are activated in response to ER stress by S1P- and S2P-mediated RIP (25;27;29;47).


The physiological function of S1P has been investigated in mice with a targeted deletion of S1P. Homozygous null mutations of S1P result in embryonic lethality before day 4 due to abnormal epiblast formation and subsequent embryo implantation (48;49). However, an inducible liver-specific knockout of S1P is viable, and was used to study the requirement for S1P in the processing of SREBPs in the liver (48). The Mbtps1 gene was flanked with loxP sites, and mice homozygous for the floxed allele and heterozygous for a transgene encoding the Cre recombinase under control of the interferon (IFN)-inducible liver-specific MX1 promoter were generated. Intraperitoneal injection of poly I:C was used to induce IFN, activating Cre recombination and inactivation of S1P in livers. These animals exhibited reduced levels of nuclear SREBPs, and 74% and 64% reductions of cholesterol and fatty acid biosynthesis, respectively, in hepatocytes. Low density lipoprotein (LDL) receptor mRNA and LDL receptor-mediated clearance of plasma LDL declined by 50%. However, plasma cholesterol levels fell, suggesting that LDL production was reduced. The merely partial effects of conditional deletion of S1P on SREBP activation and lipoprotein biosynthesis were unexpected, as S1P function was presumed to be nonredundant. The data suggest that another protease may substitute for S1P in the liver. Alternatively, the 77-90% efficiency of Cre-mediated recombination may not be sufficient to abolish S1P function.

Putative Mechanism

Inflammatory bowel disease (IBD) is a chronically recurring inflammatory disorder of the intestine, with causative contributions from genetic, environmental and immunological factors (OMIM #266600). The clinical appearance of IBD is heterogeneous, and can include diarrhea, abdominal pain, rectal bleeding, fever, weight loss, and signs of malnutrition (50). Several recent reports provide evidence for a causal role for ER stress in IBD. For example, a mutation of the Muc2 gene (see record for Schlendrian), which encodes the highly glycosylated mucin-2 protein that constitutes the majority of the intestinal mucus barrier, leads to ER stress and spontaneous colitis in mice (51). In these animals, ER stress was attributed to the aberrant oligomerization and accumulation of mucin-2. Mice lacking IRE1β, an ER stress sensor expressed in intestinal epithelial cells, are more susceptible to DSS-induced colitis relative to wild type mice (52).  These mice exhibit increased levels of the chaperone GRP78 in the colonic mucosa, indicative of ER stress. Mice lacking XBP1, a UPR response protein activated by IRE1, in intestinal epithelial cells display a loss of secretory Paneth cells and goblet cells in the intestinal epithelia, as well as spontaneous inflammation in the ileum (53).  Elevated levels of GRP78 and CHOP, and IRE1 hyperactivation in the small intestine were also detected in these mice.  In addition, single nucleotide polymorphisms (SNPs) in the XBP1 locus have been correlated with IBD in humans (53).


The wrt phenotype provides the first evidence that the ATF6-mediated UPR, initiated by S1P, is required for prevention of IBD and maintenance of intestinal homeostasis. Other ER stress-induced S1P substrates, including CREBH, CREB4, Luman, and OASIS, may also play a role in the prevention of IBD. In contrast to XBP1 and IRE1 mutant mice, which have higher basal levels of ER stress response proteins such as GRP78 and CHOP relative to wild type mice, wrt mice have normal basal protein levels of both GRP78 and GRP94 which fail to be upregulated in response to ER stress. Mouse embryonic fibroblasts from ATF6α knockout mice similarly display normal basal levels of ER chaperones including GRP78 and GRP94, but fail to induce them upon ER stress (54). These data support the finding that ATF6 is required for full IRE1α-dependent induction of UPR genes, but IRE1α is not required for ATF6-dependent transcriptional activation of target genes (55). In the absence of IRE1 or XBP1, compensatory activation of the ATF6-mediated UPR may result in elevated levels of GRP78 and CHOP.


The precise mechanism by which ER stress causes colitis remains under investigation, but evidence suggests it involves the induction of both inflammatory and apoptotic signaling. While the goal of the UPR is to adapt to a changing environment and restore normal ER function, when adaptation fails and homeostasis is not restored, ER-mediated pathways alert the cell (and possibly neighboring cells) by activating NF-κB, which induces a variety of genes involved in host defense. In the case of excessive and prolonged ER stress, apoptotic signaling pathways are activated in order to rid the body of the dysfunctional cells [reviewed in (56)]. Thus, while UPR signaling initially combats apoptosis by attempting to repair damage, it may later promote it if homeostasis is not reestablished. ER stress is associated with a variety of diseases, including neurodegeneration, ischemia/reperfusion injury, heart disease, and diabetes (56).


Activation of PERK, IRE1α, and ATF6α by ER stress has been shown to initiate apoptotic signaling (56). However, ATF6α-deficient cells exhibit increased apoptosis upon ER stress (54). It has been proposed that this increased apoptosis may be due to a compromised ability of ATF6α-deficient cells to adapt to ER stress by downregulating UPR signaling and thereby prevent eventual apoptosis (52). The prolonged ER stress-induced elevation of CHOP, which has been shown to induce apoptosis (57) and spliced XBP1 levels in cells lacking ATF6α supports this hypothesis.  Likewise, a failure in adaptation to ER stress in the colons of wrt mice may, through aberrant prolonged activation of ATF6-independent UPR signaling, result in activation of apoptotic signaling pathways. Death of intestinal epithelial cells may cause disruption of the gut epithelial barrier and permit lumenal antigens to penetrate into subepithelial tissues leading to inflammation.


ER stress in the intestine and colon may also directly cause inflammation, without a disruption of the intestinal barrier. Kaser et al. report that even minor deficiencies in XBP1 expression within intestinal epithelial cells lead to spontaneous enteritis, while leaving the intestinal barrier largely intact (53). In addition, XBP1 deficiency in a small intestinal cell line promotes a proinflammatory state in which TNFα and flagellin activate elevated JNK/SAPK signaling in the absence of microbes and cytokines (53). Such signaling may in turn contribute to the increased apoptosis of Paneth, goblet and intestinal epithelial cells observed in XBP1 mutant mice. A deficiency of ATF6 signaling may directly promote a proinflammatory state in woodrat colons, a hypothesis supported by the finding that woodrat mice display increased levels of IL-1β and IL-6 in epithelium and lamina propria even before DSS administration, suggestive of a low level of constitutive inflammation.

Primers Primers cannot be located by automatic search.
Woodrat genotyping is performed by amplifying the region containing the mutation using PCR, followed by sequencing of the amplified region to detect the single nucleotide transition. The same primers are used for PCR amplification and for sequencing. The woodrat mutation destroys an Afl III restriction enzyme site in the Mbtps1 genomic sequence, and genotyping may also be performed by PCR amplification of the region containing the mutation followed by Afl III restriction enzyme digestion.
PCR program
1) 94°C             2:00
2) 94°C             0:30
3) 60°C             0:30
4) 72°C             0:30
5) repeat steps (2-4) 34X
6) 72°C             3:00
7) 4°C              ∞
The following sequence of 300 nucleotides (from Genbank genomic region NC_000074 for linear genomic sequence of Mbtps1) is amplified:
29541                       acaccctcag tgccagccag gacaccctac aagcaggcct
29581 ttagtgtgca gccccactgt gctacctgtg cccggcactc tctgatgtgg tctacagtaa
29641 cactctcttc ttgtttcagc ctgagtccta gctacatcga cctgactgag tgtccctaca
29701 tgtggcccta ctgctcccag cctatctact atggaggaat gccaacaatc gttaatgtca
29761 ccatcctcaa tggcatgggc gtcacaggaa gaattgtgga taaggtgaaa cttctctgtg
29821 tgacgtggcc cagactttgt
Primer binding sites are underlined; the Afl III site destroyed by the woodrat mutation is highlighted in gray; the mutated A is indicated in red lettering.
Restriction Digest
Instead of sequencing, digest PCR reactions with Afl III. Run on 2% agarose gel with heterozygous and C57BL/6J controls.
Products: woodrat allele- 300 bp; wild type allele- 158 bp, 142 bp.
Science Writers Eva Marie Y. Moresco, Anne Murray
Illustrators Diantha La Vine
AuthorsKatharina Brandl, Sophie Rutschmann, Karine Crozat, Michael Berger, Xin Du, Bruce Beutler
Edit History
2010-12-23 2:53 PM (current)
2010-12-16 2:36 PM
2010-12-16 2:36 PM
2010-11-11 3:39 PM
2010-11-11 3:11 PM
2010-03-12 9:00 AM
2010-03-09 3:03 PM
2010-02-03 3:02 PM