Phenotypic Mutation 'feeble' (pdf version)
Mutation Type critical splice donor site (2 bp from exon)
Coordinate127,608,770 bp (GRCm38)
Base Change A ⇒ T (forward strand)
Gene Slc15a4
Gene Name solute carrier family 15, member 4
Synonym(s) PHT1, C130069N12Rik, PTR4
Chromosomal Location 127,595,664-127,632,897 bp (-)
MGI Phenotype Mice homozygous for an ENU-induced mutation display abrogation of both Toll-like receptor (TLR)-induced type I IFN and proinflammatory cytokine production by plasmacytoid dendritic cells. Conventional dendritic cells respond normally to TLR ligands.
Accession Number

NCBI RefSeq: NM_133895; MGI: 2140796

Mapped Yes 
Amino Acid Change
Institutional SourceBeutler Lab
Ref Sequences
Ensembl: ENSMUSP00000031367 (fasta)
Gene Model not available
SMART Domains

low complexity region 10 29 N/A INTRINSIC
Pfam:MFS_1 33 505 2.1e-9 PFAM
Pfam:PTR2 96 497 9e-94 PFAM
transmembrane domain 533 555 N/A INTRINSIC
Phenotypic Category decrease in response to injected CpG DNA, TLR signaling defect: type I IFN production by pDC
Penetrance 100% 
Alleles Listed at MGI

All alleles(4) : Gene trapped(3) Chemically induced(1)

Lab Alleles
AlleleSourceChrCoordTypePredicted EffectPPH Score
IGL01475:Slc15a4 APN 5 127601960 missense probably benign 0.41
IGL01543:Slc15a4 APN 5 127603766 missense probably benign 0.09
IGL01805:Slc15a4 APN 5 127608836 missense probably benign 0.36
IGL02958:Slc15a4 APN 5 127604665 missense possibly damaging 0.89
IGL03367:Slc15a4 APN 5 127601941 missense probably damaging 1.00
R0018:Slc15a4 UTSW 5 127602010 missense probably damaging 1.00
R0018:Slc15a4 UTSW 5 127602010 missense probably damaging 1.00
R0688:Slc15a4 UTSW 5 127608759 splice donor site probably benign
R0838:Slc15a4 UTSW 5 127617003 missense probably benign 0.42
R1544:Slc15a4 UTSW 5 127603768 missense probably benign 0.01
R1662:Slc15a4 UTSW 5 127608979 missense possibly damaging 0.79
R2504:Slc15a4 UTSW 5 127617239 missense possibly damaging 0.92
R2843:Slc15a4 UTSW 5 127604536 critical splice donor site probably null
R2845:Slc15a4 UTSW 5 127604536 critical splice donor site probably null
R2846:Slc15a4 UTSW 5 127604536 critical splice donor site probably null
R2900:Slc15a4 UTSW 5 127604536 critical splice donor site probably null
R2971:Slc15a4 UTSW 5 127604536 critical splice donor site probably null
R3018:Slc15a4 UTSW 5 127604536 critical splice donor site probably null
R3020:Slc15a4 UTSW 5 127604536 critical splice donor site probably null
R3021:Slc15a4 UTSW 5 127604536 critical splice donor site probably null
R3027:Slc15a4 UTSW 5 127604536 critical splice donor site probably null
R3053:Slc15a4 UTSW 5 127596682 missense possibly damaging 0.83
R3155:Slc15a4 UTSW 5 127604536 critical splice donor site probably null
R3432:Slc15a4 UTSW 5 127604536 critical splice donor site probably null
R4356:Slc15a4 UTSW 5 127604536 unclassified probably null
R4357:Slc15a4 UTSW 5 127604536 unclassified probably null
R4359:Slc15a4 UTSW 5 127604536 unclassified probably null
R4448:Slc15a4 UTSW 5 127604536 unclassified probably null
R4449:Slc15a4 UTSW 5 127604536 unclassified probably null
R4450:Slc15a4 UTSW 5 127604536 unclassified probably null
R4514:Slc15a4 UTSW 5 127604536 unclassified probably null
R4544:Slc15a4 UTSW 5 127604536 unclassified probably null
R4546:Slc15a4 UTSW 5 127604536 unclassified probably null
R4952:Slc15a4 UTSW 5 127603837 missense probably damaging 1.00
R5278:Slc15a4 UTSW 5 127616969 missense probably damaging 0.99
Mode of Inheritance Autosomal Recessive
Local Stock Live Mice, Sperm, gDNA
MMRRC Submission 034296-JAX
Last Updated 05/13/2016 3:09 PM by Anne Murray
Record Created unknown
Record Posted 01/11/2013
Other Mutations in This Stock Stock #: 0152 Run Code:
Validation Efficiency: 65/78

GeneSubstitutionChr/LocMutationPredicted EffectZygosity
Adck1 A to T 12: 88,431,151 Q185L probably benign Het
Fscn3 A to G 6: 28,429,967 probably benign Homo
Per1 T to G 11: 69,104,022 probably benign Het
Pkhd1 A to C 1: 20,522,894 I1665S possibly damaging Het
Tnrc6a C to A 7: 123,180,654 P1303T probably damaging Het
Usp47 T to C 7: 112,056,577 Y154H probably damaging Het
Zfp952 T to A 17: 33,003,221 C225S possibly damaging Het
Phenotypic Description

Figure 1. Identification of mutants with in vivo defects in pDC type I IFN production.  Mice were injected with the TLR9 ligand CpG and serum type I IFN levels were measure six hours later. The screen was validated by testing mice with mutations in genes known to be essential for the type I IFN response to CpG (A). Screening of ENU-mutagenized G3 mice revealed 5 new pedigrees with defects in the CpG-induced type I IFN response. These alleles include Tlr9 (CpG6), Irf7 (inept), and Ifr8 (Gemini).

Figure 2. The feeble mutation causes defective pDC function. Injection of feeble homozygous mice with CpG-A (TLR9 stimulation) or resiquimod (TLR7 stimulation) does not induce serum type I IFN (A, left), and highly enriched pDC cultures generated from homozygous feeble bone marrow lac type I IFN responses to the indicated TLR stimuli (A, right). In vitro-derived feeble pDCs generate reduced levels of secreted proinflammatory cytokines in response to the indicated stimuli (B). Normal pDC numbers are present in the feeble spleen, and normal percentages of pDCs develop from feeble bone marrow in vitro (C). Figure is reproduced from (1).

Figure 3. The feeble mutation affects type I IFN mRNA expression and protein production. Lack of intracellular type I IFN in response to CpG stimulation in feeble pDCs even in the presence of exogenous recombinant IFN (rIFN-β). Populations were first gated on live cells expressing B220.  FSC, forward scatter (A). Lack of IFN-α gene induction in CpG-stimulated feeble pDCs using two degenerate primer sets. Inept pDCs with a mutation in Irf7 were used as controls (B). Uptake of Alexa488 conjugated CpG is normal in feeble pDCs (C). Except for "Media", all conditions included addition of CpG. Incubations were performed at 37°C except for "Ice". CpG on ice reflects background extracellular fluorescence occurring without active uptake. Treatment with DNAse was to remove cell surface-bound DNA. Trypan Blue was to quench extracellular fluorescence. Right hand graph depicts a merge of wild type and feeble data. Figure is reproduced from (1).

Figure 4. The feeble mutation does not affect cDC function. In vitro-derived feeble cDCs produce normal to near-normal TNF-α levels in response to the TLR2/6 ligand peptidoglycan (PGN), the TLR4 ligand LPS, resiquimod, and CpG (A). Feeble cDCs produce normal IFN levels in response to LPS stimulation (B) and transfected double-stranded DNA (C). TicamLps2/Lps2 cDCs, which are defective in TLR3 and TLR4 signaling, were used as negative controls in (B). Figure is reproduced from (1).

Figure 5. Reduced autoimmunity in Slc15a4 mutant (i.e., feeble) C57BL/6-Faslpr mice. Mutant and WT mice were analyzed for pDC development and function at 4 mo of age, disease manifestations at 8 mo of age, and survival. (A) Frequency of pDCs in spleen. Cells were assessed by flow cytometry using anti-CD11c and anti-PDCA-1 antibodies. (B) In vivo and in vitro type I IFN production in response to TLR stimulation. Serum IFN-α levels in CpG-challenged mice, and IFN-α production by BM-differentiated pDCs stimulated with CpG or R848, were determined by ELISA. (C) In vitro B cell proliferation in response to TLR9 engagement. Purified B cells from mutant and WT mice were stimulated in vitro with CpG. B cells from Tlr9-/- C57BL/6 mice were used as negative controls. (D) Anti-chromatin autoantibodies. (E) Anti-nuclear autoantibodies (ANA). (F) Survival. Figure reproduced from (2).

Figure 6. T-independent and T-dependent responses in the Slc15a4 mutant (i.e., feeble) C57BL/6-Faslpr mice. Mutant and WT mice were immunized with NP-ficoll or NP-CGG, and titers of low- and high-affinity anti-NP antibodies were determined by ELISA using high-substituted (NP-23) and low-substituted (NP-4) antigen-conjugates, respectively. (A) T-independent response to NP-ficoll and day 5 post-immunization. (B) T-dependent response to NP-CGG at day 28 post-immunization. Figure reproduced from (2).

Figure 7. Cellular changes in Slc15a4 mutant (i.e., feeble) C57BL/6-Faslpr mice. Spleen and LNs of mutant and WT mice were examined at 8 mo of age. (A) Spleen and LN organ weights. (B) pDC and cDC numbers in spleen. Flow cytometry was used to identify pDCs (PDCA-1+CD11clow) and cDCs (CD11c+) cells. (C) T cell numbers in spleen. CD4+, CD8+, and CD4-CD8- (double-negative, DN) T cells were identified by flow cytometry after gating on TCRβ+ cells. (D) B cell numbers in spleen. Spleen cells were examined by flow cytometry. Figure reproduced from (2).

The feeble mutation was discovered in a screen of ethylnitrosourea (ENU)-mutagenized G3 C57BL/6J mice looking for reduced type I interferon (IFN) responses to CpG DNA challenge in vivo (1) (Figure 1).  This screen is designed to identify mutations that specifically affect plasmacytoid dendritic cell (pDC) development and function.  pDCs are the primary type I IFN producing cell type in response to Toll-like receptor 9 (TLR9), which senses CpG DNA.  Homozygous feeble mice also fail to produce type I IFN following in vivo challenge with the TLR7 ligand, the single-stranded RNA mimic resiquimod (Figure 2A, left), while highly enriched pDC cultures generated from homozygous feeble bone marrow using FLT3 ligand (see the record for warmflash) lack type I IFN responses to the indicated TLR stimuli including viruses (Figure 2A, right). Cultured pDCs also displayed impaired proinflammatory cytokine responses including reduced production of tumor necrosis factor (TNF)-α, interleukin (IL)-6 and IL-12p40 (Figure 2B) (1). Despite this defect in pDC function, feeble mice display normal frequencies of splenic pDCs, and normal percentages of pDCs develop from feeble bone marrow in vitro (Figure 2C). The defect in type I IFN production by feeble pDCs is not due to an inability to secrete type I IFN because Ifna transcripts andintracellular type I IFN protein could not be detected in in vitro-generated feeble pDCs, even when exogenous type I IFN was added to circumvent the requirement for the IFN feed-forward loop [(1); Figure 3A,B].  Uptake of CpG by feeble pDCs is normal, suggesting that the feeble defect occurs after TLR ligands areinternalized (Figure 3C).  


In contrast, feeble peritoneal macrophages and conventional dendritic cells (cDCs) generated in vitro from GM-CSF-treated bone marrow cells produce near normal amounts of TNF-α following stimulation with all TLR stimuli, including TLR7 and TLR9 ligands ((1); Figure 4A; TLR Signaling Screen).  Furthermore, TRIF-dependent signaling through TLR4 is intact in cDCs from feeble mice (Figure 4B).  Finally, feeble cDCs and macrophages transfected with double-stranded DNA produce IFN, demonstrating that DNA sensing through DNA cytosolic sensors is intact in these mice (Figure 4C; Double-stranded DNA Macrophage Screen).  These macrophages also secrete type I IFN normally in response to viral infection (Ex Vivo Macrophage Screen for Control of Viral Infection). B cell proliferative responses to CpG were also normal and no defect was found in cross presentation of antigens. NK cell activity is normal in the in vivonatural killer (NK) cell and CD8+ cytotoxic T lymphocyte (CTL) cytotoxicity screen, and mice show normal susceptibility to mouse cytomegalovirus (MCMV Susceptibility and Resistance Screen) and Rift Valley Fever Virus (In Vivo RVFV Susceptibility Screen). Feeble mice contain normal numbers of immune cell types. 


Collectively, these observations demonstrate that the feeble mutation selectively impairs pDC activity while the function of cDCs and other immune cell types remains relatively normal.


A congenic C57BL/6-Faslpr mouse carrying the feeble mutation was recently studied to determine the role of pDCs in the development of lupus (2).  In this congenic strain, pDCs were present at normal frequencies in the spleen, but did not produce type I IFNs in response to TLR7 or TLR9 ligands [Figure 5; (2)].  B cells from this strain had normal in vitro proliferation TLR9 ligands and efficient T-dependent and T-independent humoral responses [Figure 6; (2)]. In addition the levels of cDCs, T cells (CD4+ and CD8+), and B cells (CD21-CD23-) were reduced [Figure 7; (2)].  Although the B cell responses in these mice were efficient, the strain had reduced disease manifestations and extended survival (2). Furthermore, they propose that the deficiency in TLR signaling and proinflammatory cytokine production in the pDCs is due to inefficient signaling by endosomal TLRs  (2). This study proposed that pDCs, via type I IFN hyperproduction and a subsequent abnormal innate immune response, have a significant role in the pathogenesis of systemic autoimmunity (2).

Nature of Mutation

Figure 8. Mapping and identification of the feeble mutant. LOD score mapping the feeble mutation to chromosome 5 (A) and sequence showing the T to A transversion in the Slc15a4 gene (B). Aberrant Slc15a4 transcripts caused by the feeble splicing mutation. Please see the text for details. The aberrant 73 amino acids encoded by the first transcript are shown in red. Black signifies amino acid coding regions (C). Figure is adapted from (1).

The feeble phenotype was mapped by crossing homozygous mutant mice to C3H/HeN mice and backcrossing the F1 progeny to the mutant stock: based on 46 meioses, the mutation mapped to chromosome 5 with a peak LOD score of 4.58 (Figure 8A), and on 112 meioses, localized to a 7.2 Mb critical region between markers located 125.310 and 132.581 Mb from the centromere.  This region contained 55 genes, of which all coding and splice junction bases (10 nucleotides at each end of the intron) were sequenced at the genomic DNA level, using a combination of semiautomated capillary sequencing and whole genome sequencing.  A T→A transversion in the donor splice site in intron 2 of Slc15a4 was identified (position 8623 in the Genbank genomic region NC_000071 for linear genomic DNA sequence of Slc15a4) (Figure 8B).

            <--exon 2 intron 2-->
The mutated nucleotide is indicated in red; the splice donor sequence is shown in blue. The mutation results in at least two aberrant transcripts (Figure 8C). The first aberrant transcript lacks 260 bp derived from exons 2 and 3 leading to a frame shift with 73 aberrant amino acids prior to chain termination. The second transcript lacks all nucleotides specified by exon 2 and the same nucleotides specified by exon 3 that are missing from the first aberrant transcript, resulting in the loss of 348 bp and 116 amino acids.
Protein Prediction

Figure 9. Topography and domain structure of the SLC15A4/PHT1 transporter. Shown are the 12 transmembrane-spanning domains (TM). The transport of proton plus histidine or peptide in the cytoplasm is indicated by the arrow. The feeble mutation occurs in the donor splice site of intron 2 and results in truncation of the normal sequence in TM5 (small red "x") or causes deletion of TMs 5 and 6 (large red "X").

Slc15a4 encodes a 574 amino acid, twelve-membrane spanning protein that is a member of the solute carrier family (SLC) 15 in mammals (Figure 9).  The solute-carrier gene (SLC) superfamily encodes intrinsic membrane transporters comprising 55 gene families and 362 putatively functional protein-coding genes. The gene products include passive transporters, symporters and antiporters that transport a wide variety of substrates including amino acids and oligopeptides, glucose and other sugars, inorganic cations and anions, bile salts, carboxylate and other organic anions, acetyl coenzyme A, essential metals, biogenic amines, neurotransmitters, vitamins, fatty acids and lipids, nucleosides, ammonium, choline, thyroid hormone and urea (3).


Members of the mammalian SLC15 family have been shown to be proton-dependent oligopeptide transporters and are included in the proton-coupled oligopeptide transporter (POT) superfamily or the peptide transporter (PTR) family, which spans all species.  Some members of the POT family exhibit limited sequence similarity to protein members of the major facilitator superfamily.  The SLC15 family is composed of four members: two proteins of the peptide transporter (PEPT) series (SLC15A1/PEPT1 and SLC15A2/PEPT2) that transport exclusively di- and tripeptides and two proteins of the peptide/histidine transporter (PHT) series (SLC15A3/PHT2 and SLC15A4/PHT1) that are able to transport histidine as well as selected peptides (4).  Mouse PHT1 has 85% identity to its human homologue and 49% identity with PHT2, but displays only weak homology with the PEPT proteins (roughly 30% similarity) (5-7).  Other than PHT2, PHT1 is most similar to the NTR1/NTR5 peptide/histidine transporters found in Arabidopsis thaliana with roughly 30% identity and 50% similarity (5)


All members of the POT superfamily are predicted to contain 12 transmembrane domains (TMD), with the N- and C-termini facing the cytosol (Figure 9).  The TMDs of mouse PHT1 are predicted to occur at amino acids 42-62, 69-89, 98-118, 161-181, 197-217, 225-245, 319-339, 364-384, 406-426, 460-480, 491-511, and 536-556 (Uniprot).  Unlike PEPT1 and PEPT2, there is no large extracellular loop between TMD9 and 10.  Despite significant differences in sequence homology amongst this family, small protein stretches can be highly conserved including three motifs characteristic of POT/PTR transporters.  The first motif (EXCERFXYYG) starts at the outward-facing end of TMD1 and stretches into the first loop.  The second motif (GXXXADXXXGKXXTI) begins in the C-terminal half of TMD2 and ends in the N-terminal half of TMD3.  The third motif (FYXXINXG) occurs in TMD5 (4).  Variations of these motifs are found in PHT1.  Point mutations introduced into these regions in PEPT1 and PEPT2 cause loss of transporter function, indicating that these domains are necessary for the formation of the transporter pore and substrate-binding (8-12)


Crystalstructures of the oligopeptide transporters do not exist and much of the transport mechanism of these proteins (and 3-D structure) is inferred from computer modeling and mutational analysis [reviewed by (13;14)].  Computer modeling of PEPT1 suggests that peptide transport occurs through a central pore.  TMD1-4 and 7-9 are predicted to determine or contribute to substrate binding characteristics and affinity in this central pore (8;15;16), and most of the residues shown to be essential for function in PEPT1 and PEPT2 are facing this pore and are presumably needed for peptide binding and proton binding.  Some of these residues are conserved in PHT1 and include Y79 in TMD2, Y106 in TMD3, Y202 and N206 in TMD5, F326, L329 and W333 in TMD7, D386 in TMD8, and E466 in TMD10.  The residues in TMD7 are proposed to be structurally important, but many of the other residues are thought to be involved in substrate affinity and rate of transport (11;13;17;18).  An arginine residue inTMD7 of PEPT1 is also important for the coupling of proton and peptide transport (19).  This residue is conservatively changed to a lysine (K321) in PHT1.  None of the histidines that likely play a role in the pH dependency and proton binding of these transporter proteins are conserved in PHT1.  This includes a histidine (H57 in hPEPT1) that is predicted to be the residue protonated as the first step in the transport cycle and is essential for PEPT1 transporter function (9;16;18).  Like PHT1, proton-dependent peptide transporters in bacteria also lack this residue, but still display similar peptide transport characteristics (20).  Alignment of PHT-like sequences showed three conservative histidine residues (H349, 446 and 522) in the extracellular loops between TMD7 and TMD8, TMD9 and TMD10, and TMD11 and TMD12, but the importance of these residues to PHT1 function has not been determined (21).  These residues are not conserved in the PEPT transporters.


PHT1 is predicted to be N-glycosylated at amino acids 136, 143, 222, 437 in the hydrophilic regions predicted to face the extracellular matrix.  The protein also contains 12 predicted protein kinase A (PKA) and protein kinase C (PKC) phosphorylation sites, most of which are located in the predicted cytosolic portion of the protein (5;21).  Some of these sites are confirmed to be phosphorylated in mouse PHT1.  In response to IFNγ in macrophages, PHT1 is phosphorylated on serines 281 and 299 (22), and PHT1 peptides phosphorylated on S292 have been recovered from mouse liver (23).  PHT1 also contains an acidic di-leucine motif [(D/E)xxxL(L/I)] at amino acids 4-9 that is known to be required for binding to AP-3 (see the record for bullet gray), an adaptor protein complex that transports protein from the late Golgi to the vacuole (21).


The feeble mutation results in abnormal splicing of the Slc15a4 gene resulting in the transcription of two aberrant products.  One of these transcripts would result in a protein that is truncated within TMD5, while the second transcript would produce a protein lacking TMDs 5 and 6 (Figure 8).  It is likely that neither of these transcripts produce functional or partially functional proteins.


Rat Slc15a4 was cloned from brain and retina.  Northern analysis indicated abundant mRNA expressed in the brain and eye with low levels found in the lung and spleen. No mRNA was found in the pancreas, kidney, intestine, liver, heart and skeletal muscle. In situ hybridization in the brain found strong expression of Slc15a4 mRNA in the hippocampus, cerebellum, and pontine nucleus.  Weak to moderate expression was detected in other regions of the brain (5).  RT-PCR and Southern blot analyses demonstrated rat Slc15a4 mRNA expression throughout the rat gastrointestinal tract, the placenta, prostate, liver and thymus (24)


By contrast, RT-PCR analysis of human SLC15A4 revealed expression mainlyin skeletal muscle, followed by kidney, heart, and liver, with relatively little expression in colon and brain.  Northern analyses with EST clones suggested primary expression in the skeletal muscle and spleen (7)SLC15A4 mRNA is also present in retinal pigmented epithelium (RPE) (25).  Human PHT1 protein was detectedat the villous epithelium in intestinal sections (21).


Figure 10. Slc15a4 mRNA is expressed in pDCs but not cDCs. Figure from reference (1).

According to SymAtlas, human SLC15A4 is highly expressed in dendritic cells (DCs), and moderately expressed in natural killer (NK) cells.  Low levels of expression are also found in other immune cell types. SLC14A4 mRNA levels are increased in the inflamed colons of patients with intestinal bowel disease (26).


Mouse Slc15a4 cDNA is reported to be detected from embryonic day 1 and 16, inner ear, liver, retina and testis (MGI).  In situ hybridization of the mouse brain found Slc15a4 expressed in most regions, with highest expression in the thalamus and cerebellum (Allen Brain Atlas).  The GEO database shows expression in cDCs and pDCs from mouse spleen (27), and in hematopoietic progenitor cells (28).  Further analysis of Slc15a4 expression in pDCs and cDCs found high levels of expression specifically in pDCs [(1); Figure 10]. Large-scale proteomic screens have identified phosphorylated PHT1 peptides in IFNγ-treated macrophages and mouse liver (22;23).


By similarity, PHT1 is predicted to be localized to the plasma membrane, and human PHT1 is apparently detected at the plasma membrane of villous epithelium (21).  Speculation concerning the intracellular location of PHT1 has been based on the fact that both PHT1 and PHT2 proteins contain di-leucine motifs that are reported to localize proteins to endosomes, lysosomes, and vacuoles via AP-3.  Both rat PHT1 and PHT2 localize to lysosomes when expressed in cell culture systems (6), and in vitro studies characterizing PHT1 and PHT2 peptide transporter functions suggests that these proteins may function intracellularly (6;21).  SLC14A4 protein is expressed in early endosomes in the human embryonic kidney epithelial cell line HEK293T (26).



Nutrient transport across the plasma membrane of cells is an essential process that requires the presence of a large number of membrane proteins with highly specialized functions. The transport pathways mediating nutrient uptake in bacteria, yeast, and plants are mainly energized by a transmembrane electrochemical proton gradient, but nutrient transport in mammalian cells is mostly driven either by the electrochemical sodium (Na+) or purely substrate gradients. However, some mammalian transporter systems, including oligopeptide transport, utilize proton gradients to drive the transport of substrates. Proton-dependent transmembrane transport of short oligopeptides occurs in all living organisms, and provides an efficient means of taking up large amounts of amino acids in peptide form. In all species from bacteria to vertebrates, oligopeptide transporters belonging to the POT/PTR family specifically transport di- and tripeptides, although ATP-dependent oligopeptide transporters in bacteria, plants and yeast are able to transport peptides greater than three amino acids in length. The absence of these transporters in higher eukaryotes may be due to the ability of these longer peptides to activate the immune system (4).  Some POT/PTR transporters, particularly PEPT1 and PEPT2, are able to transport a broad spectrum of drugs and antibiotics, and play important roles in drug handling and disposition in vivo.  Studies of PEPT1 and PEPT2 suggest that PTR transporters can essentially transport all di and tripeptides, although peptides consisting of L-amino acids are preferred (14)


The first step in peptide transport by POT/PTR proteins is thought to be proton binding by a histidine residue (see Protein Prediction), followed by transfer of the proton to the C-terminal carboxyl group of the peptide substrate, which is then translocated across the membrane.  The proton then dissociates from the peptide in the cytoplasm (19).  Although this model may apply to all PTR transporters, the substrate:H+ stoichiometry is variable.  For instance, the high affinity PEPT2 catalyzes uptake of two and three protons with neutral and anionic dipeptides, respectively, while the low affinity PEPT1 catalyzes uptake of one H+ per neutral peptide (14).  Transport properties of these transporters depend on several factors including protonation state of the substrate, and intracellular and extracellular pH (29).      


In mammals, peptide uptake occurs mainly in the intestine and kidneys primarily through the action of PEPT1 and PEPT2, respectively.  Peptide uptake is indirectly Na+ dependent in these tissues based on the requirement of Na+/H+ exchanger activity to recover from the cellular acid loading caused by proton-dependent oligopeptide transport (30).  Due to its expression in the intestine, PEPT1 is important for the absorption of many drugs, while PEPT2 contributes to the clearance of these drugs through the kidney (4;14).  PEPT2 is also expressed at high levels in the brain.  Characterization of PEPT2-deficient mice suggests that PEPT2 is the primary oligopeptide transporter in the kidney (31;32), and is necessary for the transport of peptides from the cerebrospinal fluid to the brain (32;33).  Despite these defects, PEPT2-deficient mice remain healthy and fertile.  PEPT1-deficient mice do not exist, but disruption of the PEPT1 orthologue PEP-2 (OPT-2) in Caenorhabditis elegans resulted in severely retarded development, reduced numbers of progeny, and body size due to the reduced absorption of peptide nutrients in the intestine (34).  PEPT1 expression in the intestine is regulated by food intake and leptin levels (14;35).  PEPT1 and PEPT2 are expressed in additional tissues, but characterization of peptide transport in these areas remain uncharacterized.  Unlike PEPT1 and PEPT2, PHT1 and PHT2 are able to transport the single amino acid histidine, in addition to a narrower range of di- and tripeptides (5;21)


In humans, PEPT1 may have an influence on the development of inflammatory bowel disease (IBD; OMIM #266600) such as Crohn's disease (CD) and ulcerative colitis (UC), which are chronic relapsing illnesses of the gastrointestinal tract.  PEPT1 mediates intracellular uptake of bacterial products that can induce inflammation such as the neutrophil-chemotactic molecule N-formyl-methionyl-leucyl-phenylalanine (fMLP) (36).  A recent study has shown that PEPT1 is also responsible for the cytosolic uptake of the bacterial-derived muramyl dipeptide (MDP) in epithelial cells.  Binding of MDP to NOD2, a protein often mutated in CD, induces NF-κB activation with downstream production of the proinflammatory cytokines interleukin 8 (IL-8) and monocyte chemoattractant protein-1 (MCP-1) (37).  Moreover, PEPT1 expression is induced in the inflamed colons of some patients with IBD (38;39), possibly due to the stimulatory effect that proinflammatory molecules such as TNF-α and IFNγ have on transporter trafficking to the cell membrane (40;41). However, studies on SLC15A1 polymorphisms and connection to CD were inconclusive (42).  The anti-inflammatory tripeptide KPV (Lys-Pro-Val) is transported by PEPT1 in the intestine and is able to ameliorate intestinal inflammation in chemically-induced colitis in mice (43)


Naturally occurring variants of human PEPT2 have been shown to significantly alter substrate transport in vitro (44;45), and SLC15A2 polymorphisms are associated with increased vulnerability to the neurotoxic effects of low-level lead exposure due to the presence of higher levels of lead in the blood (46).  Thus, human SLC15A2 polymorphisms may have important consequences for drug metabolism and disposition.


DCs are immune cells, whose main function is to process antigen material and present it on the surface to other cells of the immune system, thus functioning as antigen-presenting cells (APCs).  They activate helper T cells, cytotoxic T cells, and B cells, by presenting them with antigens derived from the pathogen, along with non-antigen specific costimulatory signals.  pDCs are a rare subtype of circulating DCs found in the blood as well as in peripheral lymphoid organs.  They are distinguished from conventional DCs by being derived from lymphoid precursors rather than myeloid precursors, and by differences in cell-surface markers (47).  As components of the innate immune system, these cells express endosomally-localized TLR 7 and 9, which enable the detection of internalized viral and bacterial nucleic acids, such as ssRNA (via TLR7) or CpG DNA (via TLR9).  TLR7 and 9 signaling occurs via the adaptor protein myeloid differentiation (MyD) 88 (see pococurante and lackadaisical) resulting in the activation of the NF-κB pathway and production of proinflammatory cytokines such as TNF-α (see panr1), as well as the production of large amounts of type I IFN.  Type I IFNs are pleiotropic anti-viral proteins mediating a wide range of effects. Both TNF-α and type I IFN production by pDCs depends on components of the MyD88 signaling pathway including interleukin receptor associated kinase (IRAK)-1 and IRAK-4 (see otiose), interferon response factor (IRF) 5, TNF receptor–associated factor 6 (TRAF6), inhibitor of kappa-B kinase-α (IKKα), osteopontin, and IRF7 (see the record for inept)   [reviewed in (48;49)].  The multispanning ER membrane protein UNC93B (mutated in 3d), which is required for trafficking of TLRs 3, 7, and 9 to the endosomal compartment (50), is also necessary for pDCs to sense nucleic acids. pDCs from TLR9-deficient or TLR7-deficient mice (see records for CpG1 and rsq1) fail to produce type I IFN in response to various viral and bacterial pathogens [reviewed by (51)].  An acidic endosomal environment has been shown to be essential for appropriate TLR7 and TLR9 signaling (52-54).


Putative Mechanism

The phenotype of feeble mice demonstrates that PHT1/SLC15A4 is critical for the pDC response to TLR signaling.  The closely related transporter, PHT2, is strongly expressed in certain immune cell types with phagocytotic activity.  The putative localization of this protein to intracellular vesicles such as endosomes, suggests an important role in phagocytosis of pathogens.  During phagocytosis, antigenic organisms or dead and dying cells are digested into amino acids and small peptides.  PHT2 may contribute to the transport of these peptides across the phagosomal membrane into the cytosol (6).  PHT1 may have an analogous function to PHT2 in the endosomal compartment of pDCs by regulating endosomal TLR7 and TLR9 signaling either by transporting a critical component into or out of the endosome, or by maintaining the appropriate pH necessary for TLR activation.   


SLC15A4/PHT1 was recently identified as a putative transporter for NOD1 ligands in early endosomes (26).  NOD1 is a NOD2-related protein that also detects fragments of bacterial peptidoglycan in the cytoplasm and activates the NF-κB pathway (55).  The mechanism underlying the internalization of these bacterial peptides into the cytosol appears to involve clathrin-dependent endocytosis and endosomal transport, a process that was pH-dependent and occurred prior to the acidification mediated by vacuolar-type (V) H+-ATPases.  As mentioned above, PHT localizes to early endosomes in HEK293T cells, and knock downs of SLC15A4 mRNA using siRNA constructs resulted in a significant decrease of NF-κB activation by NOD1 ligands in these cells.  These data suggest that PHT1 is needed to transport bacterial-derived peptides into the endosome prior to their internalization into the cytosol and recognition by NOD receptors (26).


As PHT1 possesses an acidic dileucine motif, it is possible that localization to the endosomal compartment may be dependent on AP-3.  Mutations in AP-3 subunits, as well as in proteins belonging to the biogenesis of lysosome-related organelle complexes (BLOC)-1, BLOC-2, and BLOC-3 complexes, cause Hermansky-Pudlak syndrome (HPS; OMIM #203300) in humans and HPS-like disease in mice (56-60), which is characterized by oculocutaneous albinism, impaired platelet aggregation, and in some cases, pulmonary fibrosis, neutropenia, and mild immunodeficiency.  These phenotypes are attributed to lysosome related organelle (LRO) defects in melanosomes, platelet dense granules, lamellar bodies of type II alveolar epithelial cells, and lytic granules of CTL and NK cells.  These cell-type specific organelles share various characteristics with lysosomes and endolysosomes, such as an acidic intralumenal pH (61).  A similar organelle within pDCs, also dependent on AP-3 and HPS protein function, may be required for TLR-dependent induction of both type I IFN and proinflammatory cytokines.


Figure 11. A membrane trafficking pathway is important for pDC function. Mice with defects in the Ap3b1 subunit of the AP-3 complex (pearl and bullet gray) and mutations in genes encoding BLOC-1 (salt and pepper) and BLOC-2 (toffee) components lack serum type I IFN in response to CpG injection compared with ENU-mutagenized G3 mice (A). Ap3b1pearl/pearl mice display normal numbers of splenic pDCs and generate normal percentages of bone marrow-derived pDCs (B). Highly enriched in vitro-derived Ap3b1pearl/pearl  pDCs display defective type I IFN responses (C), but bone marrow-derived cDCs appear to be normal (D). Unc93b13d/3d pDCs were used as controls in C. The ARF1 inhibitor, Brefeldin A, prevents type I IFN gene production in wild type pDCs (E). Figure from (1).

Figure 12. The HPS proteins form several protein complexes (AP-3, BLOC-1, BLOC-2, and BLOC3) that are involved in trafficking of proteins to lysosomal-related organelles (LROs) from the TGN and affect the synthesis of these organelles. Particular HPS complexes may affect only a subset of LROs (see text). It is though that BLOC-1 and AP-3 mediate early steps of vesicle trafficking from the early endosome, while BLOC-2 and BLOC-3 are likely involved at later stages. Studies have shown physical interactions between BLOC-1 and AP-3, and BLOC-1 and BLOC-2. For simplicity, only the major cargo proteins affected by each HPS complex are shown (ATP7A, tyrosinase; Tyrp1, LAMP1). The presence of ATP7A in maturing melanosomes allows the influx of copper and activates tyrosinase. BLOC-1 is known to bind to the vesicle fusion protein syntaxin 13, which is localized to the stage 1 melanosome-coated endosome. BLOC-2 and AP-3 may interact with clathrin. Melanosomal maturation is shown separately, along with the stages affected by mutations in each HPS complex. Mutations in the δ subunit of the AP-3 complex affect melanosomal maturation between stages III and IV, but, as mentioned above, the AP-3 complex likely mediates early vesicle trafficking.

In order to test this hypothesis, mice with mutations in the AP-3b1 subunit of AP-3 and mice with mutations in various members of the BLOC complexes, were tested for their ability to respond to in vivo injected CpG.  Ap3b1pearl/pearl (56) and Ap3b1bullet gray/bullet gray mice, the salt and pepper allele of Dtnbp1, which encodes dysbindin, a component of the BLOC-1 complex (62), and the toffee allele of Hps5, which encodes a component of the BLOC-2 complex (63), all failed to produce type I IFN after in vivo challenge with CpG DNA (Figure 11A).  Further analysis of Ap3b1pearl/pearlmice revealed that although splenic pDCs are present (Figure 11B) and can be generated in vitro from bone marrow, the cells cannot produce type I IFN and TNF-α in response to TLR9 stimulation (Figure 11C).  cDCs from Ap3b1pearl/pearlmice were functionally normal (Figure 11D).  Moreover, treatment of pDCs at the time of CpG stimulation with an inhibitor of the AP-3 activating GTP binding protein ARF1 results in inhibition of type I IFN gene induction by pDCs (Figure 11E).  The similarity of these phenotypes to those found in feeble animals suggests that these proteins may function in the same pathway as PHT1 in pDCs [Figure 12; (1)]. A similar study of AP-3 function in pDCs suggests that AP-3 is required for the type I IFN response by trafficking TLR9 to a specialized subcellular compartment (64).  


Mutations in the lysosomal trafficking regulator gene Lyst (see the record for souris) and the GTPase-encoding Rab27a (see the record for concrete) also cause albinism, platelet dysfunction, and immunological defects (65;66). However, mice with homozygous mutations in these genes (charlotte gray, ashen and concrete) responded normally to in vivo CpG challenge as did june gloom mice homozygous for a mutation in the pigmentation-specific gene Slc45a2 (1).


In humans, SLC15A4 was found to be a candidate susceptibility locus for systemic lupus erythematosus (SLE; OMIM #152700) in genome-wide association studies (67).  Patients with this systemic autoimmune disease often display an “interferon signature” with upregulation of IFN target genes (68).  Thus, PHT1 may be an attractive target for pharmacotherapy aimed at preventing aberrant type I IFN production as occurs in diseases such as SLE.


Primers Primers cannot be located by automatic search.
Feeble genotyping is performed by amplifying the region containing the mutation using PCR, followed by sequencing of the amplified region to detect the single nucleotide change.
Primers for PCR amplification
PCR program
1) 94°C             2:00
2) 94°C             0:30
3) 56°C             0:30
4) 72°C             1:00
5) repeat steps (2-4) 29X
6) 72°C             7:00
7) 4°C               ∞
Primers for sequencing
Feeble_seq(R):  5’- 5’- CCTCACCCATGCTGGAAGTG -3’
The following sequence of 872 nucleotides (from Genbank genomic region NC_000071 for linear DNA sequence of Slc15a4) is amplified: 
23554                                     tgacttt gagcactagg acttgctggc
23581 atttggagag gcatctagct gctggaggga ttgagtgtga gatggctttt gtcaggagac
23641 tcgctgaatc tgtaaacagc actttgacac tgtccttctt gagatcctgt gtatctgctt
23701 tcaatttgta gatgatcagg tgctaggtag gtatggtaac ggcatatagc tgctgtgggt
23761 ggttgggagg tcattacata tgaaaacact acagcttgtc atgtttaaag agtaaacgga
23821 aggtgtagaa gatcacattt tcttatgcat ttctcttctc tcttacttta ggttaaagat
23881 cgaggtccag aagccactcg gagatttttc aattggtttt actggagcat taatttggga
23941 gcaatcctgt cattaggagg tattgcctat attcagcaga atgtgagctt tttcacaggc
24001 tacctgattc ccacagtctg tgtggccatt gctttcctgg tcttcctctg tggccagagt
24061 gtcttcatca ccaagcctcc tgacggcagt gccttcactg acatgttcag aattctgacc
24121 tacagttgct gctcccagag aggagggcag cggagaagtg ggtgagtgac acaactcctt
24181 gctccccgca gagaaagtcc cagctgaggg ctgggctgtg tggtttgtgt ctggtcttct
24241 agccggcttc attcaggaag tgtgtggtat aaacccctcc attaggatag cagagccctg
24301 agttcaatac tctgcccaac tccttcactt ccagcatggg tgagggtgcc ttgctgctgg
24361 gtagcagatt ttgaaggtag ggtctgcaga agatatcatg gagacattgc gtgctttaga
24421 cacgg
PCR primer binding sites are underlined; sequencing primer binding sites are highlighted in gray; the mutated T is shown in red text.
  67. Han, J. W., Zheng, H. F., Cui, Y., Sun, L. D., Ye, D. Q., Hu, Z., Xu, J. H., Cai, Z. M., Huang, W., Zhao, G. P., Xie, H. F., Fang, H., Lu, Q. J., Xu, J. H., Li, X. P., Pan, Y. F., Deng, D. Q., Zeng, F. Q., Ye, Z. Z., Zhang, X. Y., Wang, Q. W., Hao, F., Ma, L., Zuo, X. B., Zhou, F. S., Du, W. H., Cheng, Y. L., Yang, J. Q., Shen, S. K., Li, J., Sheng, Y. J., Zuo, X. X., Zhu, W. F., Gao, F., Zhang, P. L., Guo, Q., Li, B., Gao, M., Xiao, F. L., Quan, C., Zhang, C., Zhang, Z., Zhu, K. J., Li, Y., Hu, D. Y., Lu, W. S., Huang, J. L., Liu, S. X., Li, H., Ren, Y. Q., Wang, Z. X., Yang, C. J., Wang, P. G., Zhou, W. M., Lv, Y. M., Zhang, A. P., Zhang, S. Q., Lin, D., Li, Y., Low, H. Q., Shen, M., Zhai, Z. F., Wang, Y., Zhang, F. Y., Yang, S., Liu, J. J., and Zhang, X. J. (2009) Genome-Wide Association Study in a Chinese Han Population Identifies Nine New Susceptibility Loci for Systemic Lupus Erythematosus. Nat Genet. 41, 1234-1237.
Science Writers Nora G. Smart
Illustrators Diantha La Vine
AuthorsAmanda L. Blasius, Carrie N. Arnold, Philippe Georgel, Sophie Rutschmann, Yu Xia, Pei Lin, Charles Ross, Xiaohong Li, Nora G. Smart, and Bruce Beutler
Edit History
06/15/2011 3:24 PM (current)
02/01/2011 3:01 PM
01/07/2011 7:44 AM
12/08/2010 3:34 PM
12/01/2010 9:31 AM
11/12/2010 2:41 PM
11/12/2010 11:05 AM
11/11/2010 1:41 PM
11/11/2010 1:09 PM
11/11/2010 1:04 PM
11/11/2010 12:27 PM
11/11/2010 11:49 AM
11/11/2010 11:47 AM
11/11/2010 9:37 AM
11/11/2010 9:35 AM
11/10/2010 4:14 PM
11/10/2010 4:12 PM
11/10/2010 4:09 PM
11/10/2010 3:59 PM
11/10/2010 3:48 PM
10/29/2010 3:48 PM
10/29/2010 11:13 AM
10/29/2010 11:06 AM