Phenotypic Mutation 'supermodel' (pdf version)
Mutation Type missense
Coordinate47,016,337 bp (GRCm38)
Base Change A ⇒ C (forward strand)
Gene Samd4
Gene Name sterile alpha motif domain containing 4
Synonym(s) 4933436G17Rik, 1700024G08Rik, Smaug, 1700111L17Rik
Chromosomal Location 46,882,854-47,105,815 bp (+)
MGI Phenotype FUNCTION: [Summary is not available for the mouse gene. This summary is for the human ortholog.] Sterile alpha motifs (SAMs) in proteins such as SAMD4A are part of an RNA-binding domain that functions as a posttranscriptional regulator by binding to an RNA sequence motif known as the Smaug recognition element, which was named after the Drosophila Smaug protein (Baez and Boccaccio, 2005 [PubMed 16221671]).[supplied by OMIM, Mar 2008]
PHENOTYPE: Mice homozygous for an ENU-induced allele exhibit leaness, myopathy and altered glucose metabolism. Mice homozygous for a spontaneous mutation exhibit kyphosis, abnormal gait, and decreased cortical bone thickness. [provided by MGI curators]
Accession Number

Ncbi Refseq: NM_001037221.2; NM_028966.3; NM_001163433.1; MGI:1921730

Mapped Yes 
Amino Acid Change Histidine changed to Proline
Institutional SourceBeutler Lab
Gene Model not available
SMART Domains Protein: ENSMUSP00000022386
Gene: ENSMUSG00000021838
AA Change: H86P

low complexity region 82 95 N/A INTRINSIC
low complexity region 292 305 N/A INTRINSIC
SAM 320 383 1.4e-7 SMART
low complexity region 445 463 N/A INTRINSIC
Predicted Effect probably damaging

PolyPhen 2 Score 1.000 (Sensitivity: 0.00; Specificity: 1.00)
(Using ENSMUST00000022386)
SMART Domains Protein: ENSMUSP00000098237
Gene: ENSMUSG00000021838
AA Change: H86P

low complexity region 82 95 N/A INTRINSIC
SAM 232 295 2.75e-6 SMART
low complexity region 357 375 N/A INTRINSIC
Predicted Effect probably damaging

PolyPhen 2 Score 1.000 (Sensitivity: 0.00; Specificity: 1.00)
(Using ENSMUST00000100672)
SMART Domains Protein: ENSMUSP00000122833
Gene: ENSMUSG00000021838

low complexity region 191 204 N/A INTRINSIC
SAM 219 282 1.4e-7 SMART
low complexity region 344 362 N/A INTRINSIC
Predicted Effect probably benign
SMART Domains Protein: ENSMUSP00000114621
Gene: ENSMUSG00000021838
AA Change: H86P

low complexity region 82 95 N/A INTRINSIC
SAM 232 295 2.75e-6 SMART
low complexity region 357 375 N/A INTRINSIC
Predicted Effect probably damaging

PolyPhen 2 Score 1.000 (Sensitivity: 0.00; Specificity: 1.00)
(Using ENSMUST00000137543)
Predicted Effect probably benign
Phenotypic Category
Phenotypequestion? Literature verified References
adipose tissue
skeleton phenotype
Alleles Listed at MGI

All alleles(2) : Targeted(1) Gene trapped(1)

Lab Alleles
AlleleSourceChrCoordTypePredicted EffectPPH Score
IGL00557:Samd4 APN 14 47052898 missense probably damaging 1.00
IGL01413:Samd4 APN 14 47016792 missense probably benign 0.01
B6584:Samd4 UTSW 14 47016337 missense probably damaging 1.00
G4846:Samd4 UTSW 14 47016319 missense probably damaging 1.00
R0096:Samd4 UTSW 14 47064297 missense possibly damaging 0.88
R0122:Samd4 UTSW 14 47016560 missense probably benign 0.44
R0562:Samd4 UTSW 14 47077509 missense probably damaging 1.00
R1247:Samd4 UTSW 14 47064328 critical splice donor site probably benign
R1247:Samd4 UTSW 14 47087758 small insertion probably benign
R1771:Samd4 UTSW 14 47089075 missense probably damaging 1.00
R1902:Samd4 UTSW 14 47074128 missense probably damaging 1.00
R1903:Samd4 UTSW 14 47074128 missense probably damaging 1.00
R2346:Samd4 UTSW 14 46884842 missense probably damaging 1.00
R4155:Samd4 UTSW 14 47052946 missense possibly damaging 0.74
R4498:Samd4 UTSW 14 47096109 missense probably damaging 1.00
R4510:Samd4 UTSW 14 47077585 missense probably benign 0.05
R4511:Samd4 UTSW 14 47077585 missense probably benign 0.05
R4658:Samd4 UTSW 14 47064246 missense probably damaging 1.00
R4871:Samd4 UTSW 14 47066463 missense probably damaging 1.00
R4991:Samd4 UTSW 14 47074010 missense probably damaging 0.97
R5432:Samd4 UTSW 14 47074062 missense probably benign 0.09
R5687:Samd4 UTSW 14 47016565 missense probably benign
R6035:Samd4 UTSW 14 47087872 missense probably damaging 1.00
R6035:Samd4 UTSW 14 47087872 missense probably damaging 1.00
R6254:Samd4 UTSW 14 47016631 missense probably damaging 1.00
R6366:Samd4 UTSW 14 47074150 critical splice donor site probably null
R6376:Samd4 UTSW 14 47052962 missense probably damaging 1.00
R6944:Samd4 UTSW 14 47016635 missense possibly damaging 0.94
X0018:Samd4 UTSW 14 47016696 missense possibly damaging 0.94
X0022:Samd4 UTSW 14 47074017 missense probably benign 0.45
Z0001:Samd4 UTSW 14 47016337 missense probably damaging 1.00
Mode of Inheritance Autosomal Recessive
Local Stock Live Mice, Sperm, gDNA
MMRRC Submission 037155-MU
Last Updated 2018-03-06 10:24 PM by Diantha La Vine
Record Created unknown
Record Posted 2014-06-18
Other Mutations in This Stock Stock #: B6584 Run Code: SLD00083
Coding Region Coverage: 1x: 85.5% 3x: 70.0%
Validation Efficiency: 133/150

GeneSubstitutionChr/LocMutationPredicted EffectZygosity
1700102H20Rik C to T 17: 3,559,578 Homo
2810474O19Rik C to T 6: 149,329,346 H1297Y probably damaging Het
Acadl T to A 1: 66,848,473 probably benign Het
Astn2 C to T 4: 65,992,387 V403M probably damaging Het
Clcc1 C to T 3: 108,672,913 T302I probably damaging Homo
Hormad1 T to A 3: 95,570,696 probably benign Homo
Rnf213 C to T 11: 119,426,069 T1007I probably damaging Het
Rrh T to C 3: 129,811,742 N239D probably damaging Homo
Samd4 A to C 14: 47,016,337 H86P probably damaging Homo
Slc27a2 T to C 2: 126,561,642 L195P possibly damaging Het
Srek1ip1 T to C 13: 104,817,374 probably benign Het
Tars2 T to C 3: 95,742,150 probably null Homo
Zfp37 A to T 4: 62,191,378 V521E probably damaging Het
Phenotypic Description
Figure 1. The supermodel (spmd) phenotype. (A) Leanness and kyphosis in a male spmd homozygote at 8 weeks of age. An age-matched male WT C57BL/6J mouse is shown at the top of the left panel. (B) Kaplan-Meier survival curve (n = 11 WT mice, 11 spmd mice). P = 0.018, Log-rank test. (C) Body weight of mice maintained on regular chow or HFD was monitored weekly for 20 weeks beginning at 4 weeks of age. Kaplan-Meier curves showing deaths of spmd homozygotes during the period of monitoring (right panel). Mouse numbers at start of experiment: regular chow, n = 6 male and 7 female WT mice, n = 10 male and 6 female spmd mice; HFD, n = 13 male and 11 female WT mice, n = 8 male and 5 female spmd mice. (D) Total fat and muscle volume of 8 week old male homozygous spmd (n = 3) and WT littermates (n = 3) determined by CT analysis. (E) Representative photographs of eWAT and iWAT from 9 week old male mice. (F) Weights of eWAT and iWAT normalized to body weight in 16 week old mice (n = 4 male and 4 female WT mice, 4 male and 5 female homozygous spmd mice). Data in C, D, and F represent means ± SD. P values were determined by Student’s t test unless indicated. Figure obtained from (1).
Figure 2. Reduced fat and muscle volume in spmd mice. CT analysis of 8 week old male homozygous spmd (n = 3) and WT littermates (n = 3). (A) Fat and muscle volume relative to total body volume. (B) Visceral fat volume and subcutaneous fat volume relative to total fat volume. (C) Bone mineral density and bone mineral content. (D) Weight of iBAT normalized to body weight in 16 week old mice (n = 4 male and 4 female WT mice, 4 male and 5 female homozygous spmd mice) and representative photographs of iBAT from 9 week old male and female mice. Data represent means ± SD. P values were determined by Student’s t test. Figure obtained from (1).
Figure 3. Adipocyte defects and myopathy in spmd mice. (A) Sections from eWAT of 8-10 week old WT and homozygous spmd mice were examined by bright field (left panels) and fluorescence microscopy (right panels). (B) Hind limb muscle sections from 8-10 week old WT and homozygous spmd mice. H&E staining of gastrocnemii (upper panels). Green arrowhead indicates myofibers with irregular shapes. Metachromatic ATPase staining (middle panels), and myosin (Type I, slow) staining (lower panels). Filled black triangles indicate soleus (Sol.); open triangles indicate gastrocnemius (Gast.). Scale bars, 60 μm. Figure obtained from (1).
Figure 4. Increased glycogen content of spmd muscle. Sections from gastrocnemius muscles of 12 week old WT and homozygous spmd mice were stained with H&E (upper panels) and Periodic Acid-Schiff (PAS) stains (lower panels). Green arrowheads indicate muscle fibers with centralized nuclei. Scale bars, 60 μm. Figure obtained from (1).
Figure 5. Altered glucose metabolism in spmd mice. (A) Blood glucose was measured in mice fasted for 5 hr. Each data point represents one mouse, and the bar indicates the mean of all values; error bars represent SEM. (B) Serum insulin was measured in male WT (n = 6) and homozygous spmd mice (n = 6) fasted for 5 hr. (C) Intraperitoneal glucose tolerance test. Blood glucose was measured at the indicated times after i.p. glucose injection in male WT (n = 10) and homozygous spmd mice (n = 10). For 0, 15, 30, 60, 90, and 120 min, respectively, P = 0.026, 0.198, 0.0009, 0.0015, 0.0017, 0.0020. (D) Insulin tolerance test. Blood glucose was measured at the indicated times after i.p. insulin injection in male WT (n = 7) and 15 homozygous spmd mice (n = 9). For 20, 30, and 50 min, respectively, P = 0.067, 0.009, 0.010. (E) Hyperglycemic clamp. Plasma insulin was measured at the indicated times during glucose infusion in male WT (n = 8) and homozygous spmd mice (n = 5). For 0, 5, 10, 15, 30, 60, 90, and 120 min, respectively, P = 0.050, 0.040, 0.100, 0.063, 0.059, 0.027, 0.045, 0.040. Data in B, C, D, and E represent means ± SEM. P values were determined by Student’s t test. Figure obtained from (1).
Figure 6. Elevated energy expenditure and reduced mobility of spmd mice. (A-C) Metabolic profiling of WT (n = 6) and homozygous spmd mice (n = 6) over a 72-hr period. (A) Oxygen consumption and carbon dioxide production. (B) Ambulatory motion and rearing/jumping motions as measured by cage beam breaks. (C) Food intake per day normalized to body weight. (D and E) Expression of mitochondrial uncoupling genes and brown fat markers in eWAT (D) and skeletal muscle (E) of WT (n = 5) and homozygous spmd mice (n = 5). Expression levels were normalized to 18S rRNA expression. Data represent means ± SD. P values were determined by Student’s t test. Figure obtained from (1).
Figure 7. Partial rescue of spmd phenotype by transgenic expression of wild type Samd4. (A) Construction of the modified BAC containing Samd4 isoform 2 cDNA. Organization of Samd4 genomic DNA encompassed by the original BAC, RP23-178L4 (top). A cDNA construct encoding GFP-tagged Samd4 isoform 2 (bottom) was used to replace exon sequences downstream from the start codon in the BAC. Positions of primers used for genotyping are indicated. (B) PCR genotyping of transgenic mice using the primers indicated in A. (C) Protein expression of Samd4-GFP in transgenic mice. Lysates of heart tissue from a WT mouse, Samd4spmd/spmd mouse, and Samd4spmd/spmd; Tg(Samd4) mouse were subject to immunoprecipitation using anti-GFP-coupled beads. Immunoprecipitates were immunoblotted using GFP antibody. (D) A Samd4spmd/spmd mouse (left panel) and a Samd4spmd/spmd; Tg(Samd4) mouse (right panel) at 8 weeks of age. (E) Body weights of 8 week old WT, Samd4spmd/spmd, and Samd4spmd/spmd; Tg(Samd4) mice. Each data point represents one mouse, and the bar indicates the mean of all values; error bars represent SEM. (F and G) CT analysis of 10 week old male Samd4spmd/spmd (n = 3), Samd4spmd/spmd; Tg(Samd4) (n = 5), and WT mice (n = 3). (F) Fat and muscle volume relative to total body volume. (G) Visceral fat volume and subcutaneous fat volume relative to total fat volume. Data represent means ± SD. P values were determined by Student’s t test. Figure obtained from (1).

The supermodel phenotype was initially identified in G3 mice derived from N-ethyl-N-nitrosourea (ENU)-mutagenized C57BL/6J stock (1). The supermodel mouse exhibited a markedly lean body and thoracic kyphosis by six weeks of age [Figure 1A; (1)]. Both male and female supermodel mice were infertile. In addition, the supermodel mice exhibited shortened life spans (i.e., lethality by three months) compared to wild-type mice (Figure 1B). Supermodel mice exhibited lower body weight, body mass index, and body length compared to wild-type mice (Figure 1C). When fed a high fat diet, the supermodel homozygotes increased their initial body weights by 1.3-fold (males) or 1.1-fold (females), compared to wild-type mice that increased their body weights by 1.5-fold (males) or 1.8-fold (females) over a 6-month period (Figure 1C). Circulating levels of cholesterol and high-density lipoprotein (HDL) were reduced whereas low-density lipoprotein (LDL) was elevated in supermodel mice compared to wild-type mice (1). Analysis by computed tomography (CT) determined that the weight difference between the supermodel and wild-type mice was due to reduced fat and muscle tissue in the supermodel mice (Figure 1D). The weights of epigonadal white adipose tissue (eWAT) and inguinal WAT (iWAT) from supermodel mice were significantly reduced relative to those in wild-type mice (Figure 1E & F). The reductions in fat and muscle volume in the supermodel mice were proportional to the overall reduction in whole body volume [Figure 2A; (1)]. In addition, the proportion of total fat volume corresponding to visceral fat was decreased, while the proportion of total fat volume corresponding to subcutaneous fat was increased, indicating that fat distribution in the supermodel mice was altered (Figure 2B). Bone mineral density or bone mineral content was not changed in the supermodel mice (Figure 2C). Histological examination of WAT from 8-10 week old supermodel mice determined that numerous white adipocytes exhibited heterogeneous morphology characterized by reduced cell size and abnormal fat droplet accumulation in the cytoplasm [Figure 3A; (1)]; the weight and morphology of interscapular brown adipose tissue (iBAT) was not significantly changed. The hind limb muscle from the supermodel mice exhibited scattered focal myopathy (i.e., focal myofibers with irregular shapes and heterogeneous sizes) in the gastrocnemius, soleus, estensor digitorum longus (EDL), and tibialis anterior (TA) muscles (Figure 3B). Several myofibers exhibited centralized nuclei, indicating constitutive myofiber degeneration and regeneration [Figure 4; (1)].


Homozygous supermodel mice at 10-12 weeks of age exhibited reduced fasting glucose and insulin levels compared to wild-type mice [Figure 5A & B; (1)]. Intraperitoneal glucose tolerance tests determined that the supermodel mice had delayed clearance of glucose from the blood (Figure 5C), elevated insulin sensitivity (Figure 5D), and impaired insulin release by pancreatic beta cells (Figure 5E). Metabolic cage studies of 10-12 week old homozygous supermodel mice determined that they had enhanced oxygen consumption and carbon dioxide production over a 72-hour period compared to wild-type mice [Figure 6A; (1)]. However, the supermodel mice were hypoactive (i.e., decreased ambulatory movement and exploratory rearing/jumping) compared to wild-type mice (Figure 6B); caloric intake was comparable between the supermodel and wild-type mice (Figure 6C). Expression of mitochondrial uncoupling genes Ucp1, Upc2, and Ucp3 as well as brown fat markers Ppara and Cidea were elevated in supermodel eWAT (Figure 6D) and skeletal muscle (Figure 6E) compared to those in wild-type mice.


A BAC was constructed that contained a Samd4-GFP cDNA construct encoding isoform 2 [Figure 7A; (1); see below for more details on the isoforms of Samd4]. Expression of the BAC in homozygous supermodel mice resulted in an increase in their body size and weight (Figure 7B-E). The mice that expressed the BAC exhibited similar fat volume:total body volume and muscle volume:total body volume ratios as wild-type mice (Figure 7F) and correction of visceral and subcutaneous fat distribution (Figure 7G).

Nature of Mutation

The supermodel mutation was mapped by bulk segregation analysis (BSA) of F2 mice (n = 43) generated from an intercross of F1 mice produced by intracytoplasmic injection of sperm from supermodel homozygotes into eggs from C57BL/10J mice (1). Linkage of the phenotype (LOD = 5.41) with a single nucleotide polymorphism (SNP) at base pair 55,054,219 on chromosome 14 was detected.  Whole genome SOLiD sequencing identified an A to C transversion at base pair 47,016,337 on chromosome 14 within Samd4.  The mutation corresponds to base pair 133,373 in the Genbank genomic region NC_000080 and base pair 994 in the cDNA sequence ENSMUST00000022386 within exon 3 of 13. The mutation also corresponds to base pair 883 in the cDNA sequence ENSMUST00000100672 within exon 3 of 12.


81      -S--L--L--L--T--H--L--P--L--L--K-


Genomic numbering corresponds to NC_000080. The mutated nucleotide is indicated in red lettering, and results in a histidine (H) to proline (P) substitution at amino acid 86 in the 711 amino acid Samd4 isoform 1 (ENSMUSP00000022386) and the 623 amino acid Samd4 isoform 2 (ENSMUSP00000098237; NP_083242).


Protein Prediction

Figure 8. The protein structure of mouse Samd4. Samd4 has two SSR domains (SSR1 and SSR2) as well as a SAM. The H86P supermodel mutation is marked by an asterisk and is within the SSR2 domain. Abbreviations: SSR, Smaug similarity regions; SAM, sterile alpha motif.

Figure 9. Crystal structure of the RNA binding domain of Drosophila Smg The RBD of Smg contains both a PHAT and SAM domain. Figure corresponds to PDB ID: 1OXJ. The helices are numbered according to (5). Abbreviations: PHAT, pseudo-HEAT repeat analogous topology; SAM, sterile alpha motif.

Samd4 contains two Smaug (Smg) similarity regions (SSR) at the N-terminus [SSR1 (amino acids 1-64) and SSR2 (amino acids 65-127); Figure 8], which are conserved between Drosophila Smg and vertebrate Samd4 (2). SSR domains have no defined function in Smg or Samd4 (3;4)


Samd4 binds to Smg recognition elements (SREs) on target RNAs through a sterile alpha motif (SAM) (amino acids 323-396, UniProt; amino acids 320-383, SMART) (2;3;5). The SRE forms a loop structure comprised of four to five bases (CNGG or CNGGN, where N is any base) (3;5;6).  Although a specific amino acid is not required at position 2 or 5, the separation mediated by the amino acid at position 2 is essential for RNA binding (3). Proteins that have SAM domains (e.g. Samd4, Samd4B, tumor protein 63 (Tp63), Tp73, Eph receptor B2 (EphB2), and EphA4 [see the record for frog]) are often highly conserved and are involved in regulating signaling and developmental processes, often by translational repression of their respective targets (Samd4), transcriptional activation (p63 and p73), or by functioning as receptor tyrosine kinases (the Eph receptors) (3-5;7-9). The Smg SAM domain also mediates the interaction of Smg with Cup, an eIF4E-binding protein (discussed below in “Background”) (3;5;10). The 70- to 100-amino acid SAM domains can either be in a monomeric arrangement or can form homo- or heterooligomers (3;5).  SAM domains can also associate with other SAM-containing proteins as well as with tyrosine phosphorylated SH2 domains (7).  In contrast to other SAM-containing proteins, Samd4 does not oligomerize (5).  


The crystal structure of the Drosophila Smg RNA binding domain (RBD) is comprised of a SAM domain and a C-terminal pseudo-HEAT repeat analogous topology (PHAT) domain that is essential for high-affinity RNA binding [Figure 9; PDB:1OXJ; (5); residues 596-763 of Smg]. The PHAT domain in Drosophila Smg may be necessary for an interaction with a corepressor or with other proteins during nos translational regulation (5;11). Mammalian homologs of Smg do not contain the PHAT domain, suggesting a species-specific role for the PHAT domain in Smg function (5). The RBD is comprised entirely of α- and 310-helices (5). This structure is unique in that most other RBDs contain substantial amounts of β-sheets (5;6;12); the RBDs in Pumilio (13), finO (14) and SRP19 (15) are predominantly α-helical. The SAM domain is comprised of one long α-helix (α5), three short α-helices (α1, α3, α4) and one 310-helix (h2) arranged in a globular bundle with a hydrophobic core (3;5).  The core residues are the most conserved among species; residues on the surface show more sequence variability (3). Basic residues on the surface of the SAM domain are essential for RNA binding (3;5)


Samd4 has two putative 14-3-3 mode I ([R/K]XX(pS/pT)XP) binding motifs [RSV(pS)LT, amino acids 251-265; KTR(pS)LP, amino acids 655-660] (15). Samd4 also contains a putative serine/threonine kinase Akt/PKB phosphorylation site (RXRXX[S/T]) at amino acids 163-168 (16). Samd4 has a putative tyrosine phosphorylation site at Tyr613 within the SAM domain. Phosphorylation at Tyr613 may modulate binding to mRNAs as well as binding of SH2 domain-containing proteins or low-molecular weight phosphotyrosine phosphatase (4;5;17).


The mouse Samd4 gene produces multiple protein-coding transcripts. The two most common transcripts in the mouse are ENSMUST00000022386 encoding the canonical 711-amino acid Samd4 protein, and ENSMUST00000100672 encoding the 623-amino acid protein (2).


The supermodel mutation (H86P) is within the SSR2 region of both isoforms of the Samd4 protein (1).


Samd4 was expressed in the muscle, fat tissue, brain, kidney, heart, testis, and liver of the mouse (2;18). In the human, SAMD4A (designated as KIAA1053) was expressed in heart, skeletal muscle, spinal cord, whole adult and fetal brain, adult lung, liver, ovary, kidney, and testis; little to no expression was detected in pancreas, spleen, and fetal liver (19).


Samd4 protein was detected in mouse and rat brain extracts by Western blot analysis (2). Samd4 was enriched in neuronal postsynaptic densities in a membrane-free fraction of synaptoneurosomes that was enriched in cortical cytoskeleton and synapse-associated elements (e.g. polyribosomes and the marker protein PSD-95) (2). Analysis of human Samd4 expression after transfection in mammalian cells found that Samd4 was concentrated in 0.5 - 2.0 micrometer cytoplasmic granules distributed throughout the cytoplasm (2). After treatment with leptomycin B, a drug that inhibits the nuclear export receptor CRM1, Samd4 staining of the nucleus increased, indicating that Samd4 is shuttled between the nucleus and the cytoplasm (2).


In Drosophila, Smg RNA-binding activity occurs prior to the onset of zygotic transcription, indicating that Smg is synthesized from maternal mRNA (4).  Smg mRNA is detected at high levels in ovaries, but no Smg protein is detected, indicating that there is translational regulation of Smg or rapid degradation of the Smg protein in the ovaries (4). Tadros et al. determined that there are three smg isoforms that have poly(A) tails approximately 75 nucleotides in length in the mature, stage 14 oocyte; after egg activation, the length of the poly(A) tail is extended to 175 nucleotides thereby promoting translation (20). Pan-Gu, a kinase complex, is required for the translation of smg mRNA by relieving repression by translation repressors after egg activation (20). In the embryo, Smg protein expression was high during early embryogenesis, but Smg protein could not be detected later in embryonic development, indicating a need for SRE-dependent translational repression during early embryogenesis that was not needed at later stages (4;21). The distribution of smg mRNA and the Smg protein was uniform through the early embryo, but at later stages of embryonic development the protein was concentrated at the posterior pole of the embryo in the cytoplasm with foci of concentrated Smg levels (4). The areas of Smg-containing foci contained polar granules, large ribonucleoprotein structures that are involved in germ cell specification (4)


There are several RNA-binding protein families that regulate the routing, activation, or disposal of transcripts through interaction with cis-acting elements in the 3’ untranslated region (UTR) of mRNAs (2;10;22). Although several RNA-binding proteins have been identified, the molecular interactions that allow proteins bound to mRNA to influence the translation machinery are poorly understood (10). It is estimated that as many as 2-8% of human genes encode RNA-binding proteins (22). It is unknown which, and how many, of the proteins bind to small RNAs, ribosomal RNAs, or mRNAs (22). mRNA-binding proteins can be classified as either global, group-specific or type-specific (22). Global mRNA-binding proteins (e.g. poly(A)-binding protein) recognize a large group of mRNAs without recognizing a unique sequence (22). Group-specific mRNA-binding proteins bind to a sub-set of the mRNAs from the global group, possibly to mRNAs that are structurally and/or functionally related [e.g. those that contain AU-rich elements (AREs) or SREs] (22).  Type-specific RNA-binding proteins [e.g. iron response element (IRE)-binding protein (23) and histone mRNA stem-loop-binding protein (24)] recognize highly unique mRNA sequences (22). For example, type-specific mRNA-binding proteins can recognize sequences in mRNAs that encode proteins needed in large amounts within short time intervals during biological processes (e.g. during the cell cycle) (22;24). However, with the exception of the ARE, few sequence elements common to large groups of mRNAs are known (22).


The known functions of the Drosophila Samd4 homolog, Smg, as well as studies describing Samd4 in mammals are described in detail, below.


Figure 10. Smaug functions in Drosophila embryonic development through the repression of maternal transcripts. (A) In the bulk cytoplasm of the developing Drosophila embryo, Smaug binds to hairpin structures in the 3’ UTR of nos, resulting in the repression of nos translation and its subsequent degradation (see part C for a more detailed illustration of Smaug function). In the pole plasm, Oskar (Osk), a protein that can both activate the translation of nos mRNA as well as block deadenylation and translational repression, interacts with the RNA-binding domain of Smaug, activating nos translation. Nos translation in the posterior pole plasm subsequently facilitates abdominal segmentation and posterior development in the Drosophila embryo (not shown). See the text for more details. (B) The 5’ cap (7mG) is recognized by members of the eukaryotic initiation factor (eIF)4F complex (i.e., eIF4E, eIF4G, and eIF4A).  eIF4E recruits eIF4G to the transcript. eIF4G is a scaffolding protein that subsequently recruits eIF4A and eIF3; eIF3 recruits the 40S ribosomal subunit.  (C) Smaug represses nos translation in the bulk cytoplasm through recruitment of the eIF4E-binding protein, Cup, thereby blocking the eIF4E/eIF4G interaction. Smaug also recruits the CCR4-Not complex that regulates target mRNA translation through deadenylation (via the CCR4 and Caf1 subunits). The Not4 subunit of the CCR4-Not complex also functions as an E3 ligase that ubiquitinates target proteins including Not4, Not1, and transcription initiation factor IID (TFIID), leading to co-translational proteasomal degradation. Some figures are based on solved crystal structures: Smaug, PDB:1OXJ; 5’ cap eIF4E, PDB: 1EJ1; eIF4G, PDB:1UG3; eIF4E-Cup, PDB:4AXG; eIF4A, PDB:2G9N.

Drosophila Smg

Silencing of maternal mRNAs and Nos

Early Drosophila embryonic development is controlled by proteins translated from mRNAs produced by the mother (25;26). To transfer developmental control to the zygotic genome, a subset of maternal mRNAs (e.g. Hsp83, nanos, string, Pgc, and cyclin) are degraded with a concomitant activation of the zygotic genome. Smg and Bicoid stability factor (BSF), the two major regulators of the maternal-zygote transition, are highly conserved (20;25-29). In mature oocytes, maternal transcripts are stable, but after egg activation, many of the maternal mRNAs are destabilized (20;28). In Drosophila, the degradation of maternal mRNAs is complete by interphase of nuclear division cycle 14, approximately two hours after fertilization (20;25). The 3’ UTRs of the maternal mRNAs are enriched in both microRNA (miRNA; small non-coding RNAs) target sites and SREs; unstable maternal mRNAs also contain PUF-domain protein binding sties and AREs for rapid deadenylation (20;30;31).


In Drosophila, mRNA turnover is generally initiated by Pan2-Pan3- and CCR4-Not-mediated deadenylation of the 3’ poly(A) tail [Figure 10;(26;32-35)]; the deadenylase poly(A)-ribonuclease (PARN) contributes to this process in mammals, but not in Drosophila (36;37).  Shortening of the poly(A) tail reduces the binding of Pab1, a poly(A) binding protein, to the tail resulting in a reduced rate of translation followed by decapping by Dcp1/Dcp2 and the 5’ to 3’ degradation by the exonuclease Xrn1 (38). In contrast, 3’ UTR-mediated translational control can also occur when the length of the poly(A) tail remains constant (such as in the case of nos, oskar (osk) mRNAs as well as the lipoxygenase mRNA from rabbit reticulocytes) (4;39). In the case of the aforementioned mRNAs, regulatory elements within the 3’ UTRs are responsible for translational repression [i.e. lipoxygenase and osk have multiple copies of a motif with no secondary structure (4;39;40), while nos has stem-loop SREs (4;21;41)].


Smg controls both transcript stability and translation within the first two hours of Drosophila embryogenesis [Figure 10; (10;20;26)]. One important Smg target is maternal transcripts encoding Nos, a key regulator of anterior-posterior pattern formation. At fertilization, nos mRNA is present throughout the bulk cytoplasm and the posterior pole plasm of the embryo. However, nos mRNA is translated only in the posterior pole plasm, while nos mRNA in the bulk cytoplasm is translationally repressed and degraded. Such an expression pattern generates a Nos protein gradient originating from the posterior pole plasm, which is necessary to promote proper anterior and posterior cell fates (21;41-44).  Nos protein, in addition to Pumilio (Pum) and Brain Tumor (Brat), represses the translation of Hunchback (Hb) to facilitate abdominal segmentation and the posterior development of the embryo (45;46). Deletion of the region of nos that binds Smg leads to accumulation of Nos in the bulk cytoplasm and head development defects (41). There are two SREs within stem-loop structures in the 3’ UTR of nos that are necessary and sufficient for translational repression by Smg as well as to direct nos for degradation (21;26;28;41;47)


Loss-of-function alleles were used to initially determine that Smg is necessary for the repression of nos mRNA in the bulk cytoplasm as well as for normal progression through the late nuclear division cycles in the Drosophila embryo (27)Smg mutant embryos displayed aberrant expression of Nos in the bulk cytoplasm resulting in lethal body-patterning defects (3;27;27). In addition, there were cell cycle defects during later nuclear divisions and an inability to develop to the midblastula transition (27).  In another study using a null allele of smg, nos mRNA was stabilized after two hours of development and the poly(A) tail of the nos mRNA was longer in the null embryos than in wild-type embryos (32)


Smg continues to be expressed in Drosophila embryos after the expression of nos mRNA declines, indicating a possible role for Smg after embryogenesis (4)


The CCR4-Not complex

It is estimated that in Drosophila, over 1,600 of 7,745 maternal mRNAs are destabilized after fertilization (20;28). Following egg activation in Drosophila, approximately two-thirds of the destabilized maternal mRNAs (712 of 1,069) are targeted for Smg-mediated suppression through recruitment of the CCR4-Not complex to the 3’ UTRs of the mRNAs (20;25;26;32). The CCR4-Not complex is an L-shaped complex comprised of several conserved subunits including CCR4, Not1-5, Caf40, Caf130, and Caf1; all of the subunits, except Not4, are associated with a stable complex (38;48-50). The CCR4-Not complex regulates gene expression by utilizing its two main enzymatic functions, ubiquitination (via the Not4 subunit) and deadenylation (via the CCR4 and Caf1 subunits) (Figure 10).  The CCR4, Caf1, and Not4 subunits of the complex are proposed to act independently of the complex as well as to bind and dock the complex onto targets (38)


In addition to its role in deadenylating maternal transcripts, the CCR4-Not complex plays a role in response to glucose depletion and stress by ubiquitinating target proteins and regulating several cellular processes including the DNA damage response, and transcription elongation by allowing stalled RNA pol II to resume transcription by changing the configuration of the polymerase relative to the template (35;38;51)


Interaction of Smg with the eIF4E-binding protein, Cup

The SRE-dependent repressor complex is comprised of TraI, Me31B, the CCR4-Not complex, Smg, and Cup (47). Smg represses nos translation in the bulk cytoplasm through recruitment of the eIF4E-binding protein, Cup, thereby blocking formation of the translation initiation complex eIF4F (10). Cup is distributed throughout the Drosophila embryo and is homologous to 4E-T, a human nucleocytoplasmic shuttling protein that transports eIF4E into the nucleus (10;52;53); it has been proposed that spatial regulation of nos may involve disrupt­­­ing Cup and/or Smg function specifically at the posterior (10). In addition to Smg-related functions, Cup is also involved in oocyte growth, maintenance of chromosome morphology, and establishment of egg chamber polarity (10;52). Cup may serve as an adaptor protein that is used by multiple translational repressors to interact with eIF4E (10).


Interaction of Smg with Hsp83

Hsp83 was subsequently identified as another target of Smg during early Drosophila embryogenesis although Smg-binding sites have not been identified (20;28;54). Hsp83 associates with various cellular signaling proteins such as steroid hormone receptors, Src-like kinases, and the serine/threonine kinase Raf  (55). Association between Hsp83 and the steroid hormone receptors mediates ligand binding and activation of the receptors; the role of Hsp83 in Raf activation is not known (55).  Smg recruits CCR4-Not which deadenylates/destabilizes the hsp83 transcript in the bulk cytoplasm, leading to degradation of the transcript; hsp83 is protected from degradation in the posterior pole plasm (20;28;54). Smg-mediated regulation of hsp83 is independent of Cup and SRE motifs (2;26)


Interaction of Smg with Oskar

The RNA-binding domain of Smg also interacts with Oskar (Osk), a protein that can both activate the translation of nos mRNA as well as block deadenylation and translational repression; overexpression of Smg can block Osk-mediated activation of nos mRNA (32;41;42;56;57).  During early oogenesis, osk mRNA is distributed throughout the cytoplasm, but during the maturation process of the oocyte, osk is localized to the posterior pole of the oocyte; repression of osk translation is alleviated upon localization to the posterior pole (4;40;56;58). Once Osk is expressed it can then recruit nos mRNA to the posterior pole of the embryo (4;40;43;44;46;56). Zaessinger et al. showed that Osk does not affect Smg and CCR4-Not complex association, but instead Osk prevents Smg from binding the nos mRNA, resulting in a lack of nos mRNA deadenylation, and stabilization and translation of the mRNA (32).  Whether Osk directly binds to Smg to mediate this effect is not known (32).


Table 1 summarizes the functions of the Smg-interacting proteins.


Table 1.  Interactors with Drosophila Smg

Protein or Complex




Regulates abdominal segmentation and anterior-posterior pattern formation in Drosophila embryo; Smg is essential for nos localization


CCR4-Not complex

Regulates Smg-mediated degradation of maternal mRNAs by deadenylation



Prevents the binding of eIF4G to eIF4E; mediates interaction between Smg and eIF4E


Hsp83 (i.e. Hsp90)

Associates with various cellular signaling proteins; mediates ligand binding and activation of the steroid hormone receptors



Represses translation of several mRNAs including hunchback in Drosophila development



Prevents deadenylation of nos mRNA; activates translation of nos mRNA


Pan-Gu kinase complex

Required for smg translation



Mammalian Samd4

When overexpressed in BHK cells, human Samd4 repressed translation of SRE-containing reporter mRNAs without affecting their stability (2).  Samd4 overexpression also induced the formation of cytoplasmic granules containing polyadenylated mRNA and the RNA binding proteins Staufen-1 and PABP. When the cells were fractionated by sucrose gradient centrifugation, Samd4 was found in cytoplasmic 20S particles lacking polysomes (2).  The mechanism of translation repression by mammalian Samd4 has not been reported and the targets of Samd4 are currently unknown (3).  Statistical analysis has predicted that the RNA consensus sequence required for Smg binding is expected to occur once in every 10,000-100,000 bases, indicating that Smg homologs could potentially regulate a large number of mRNAs (3).  In Drosophila, ~18% of cDNA sequences contain at least one putative SRE (3).  It has been proposed that in addition to the SAM domain, another domain (e.g. the SSR domains) or RNA binding activity may be involved in target specificity (3).  In addition, a target sequence may need to have multiple SREs to bind Samd4 (3).


There are two Smg homologous genes present in the mammalian genome, Samd4 and Samd4b; Samd4 is located on mouse chromosome 14, while Samd4b is on mouse chromosome 7 (2-4).  Northern blot and RT-PCR analyses of human tissues revealed that Samd4b is expressed in both embryonic and adult tissues with the highest expression in adult testis and lowest in adult liver (7). Samd4B inhibits AP-1, p53, and p21-mediated transcriptional activities (7).

Putative Mechanism
Figure 11. Samd4 is required for mTORC1 signaling. (A) Lysates of muscle or eWAT from 12-15 week old WT and homozygous spmd mice were immunoblotted using the indicated antibodies. (B and C) Flag-tagged wild type Samd4 or Samd4-H86P were transiently overexpressed in 293T cells or stably expressed in C2C12 myoblasts. Myc-tagged 14-3-3ζ protein was co-overexpressed in 293T cells. Flag or Myc immunoprecipitates were immunoblotted using the indicated antibodies. In B, lanes 1 and 4 contain 10% of the IPs from C2C12 cells; lanes 1 and 2 contain 100% of the IPs from 293T cells. In C, lanes 2 and 3 contain, respectively, 50% and 100% of the IPs. (D) In vitro kinase assay. 5 μg of purified recombinant proteins visualized by Coomassie blue staining (left panel). Recombinant proteins were incubated for 30 min with ATP and HA-tagged activated Akt immunoprecipitated from 293T cells. Reaction mixtures were immunoblotted with phospho-Akt substrate and HA antibodies (right panels). Figure obtained from (1).

Figure 12. Mammalian Samd4 is required for mTORC1 signaling.  mTORC1 signaling is stimulated by numerous upstream signals including growth factors and insulin, which bind and activate their receptors, leading to the recruitment and activation of PI3K. Phosphorylation of PIP2 by PI3K generates PIP3, which recruits Akt to the inner leaflet of the cell membrane, where Akt is phosphorylated and activated by PDK1. Active Akt promotes mTORC1 signaling by inhibitory phosphorylation of PRAS40 and the TSC1/TSC2 complex, both negative regulators of mTORC1. Phosphorylation of PRAS40 reduces its binding to Raptor, releasing mTORC1 from inhibition by PRAS40. The 14-3-3 proteins bind to and inhibit the functions of phosphorylated TSC2 and PRAS40, thereby promoting mTORC1 signaling. In the mTORC1 complex, mTOR interacts with Raptor, PRAS40, Deptor, and mLST8 to phosphorylate target proteins, including S6 kinase 1 (S6K1) and eIF-4E-binding protein 1 (4E-BP1), in a rapamycin-sensitive manner. mTORC1-associated signaling mediates several downstream effects in the cell including suppression of autophagy, activation of transcription leading to mitochondrial metabolism, increased protein synthesis, proliferation, and growth factor production. Samd4 is necessary for mTORC1 signaling, as shown by reduced phosphorylation of 4E-BP1 and S6 in supermodel homozygotes. Wild type but not mutant Samd4 is phosphorylated by Akt in vitro, and binds to 14-3-3 proteins, suggesting that Samd4 promotes mTORC1 signaling in conjunction with these proteins. Abbreviations: PI3K, phosphoinositide 3 kinase; PDK1; phosphoinositide dependent kinase 1; Raptor, regulatory-associated protein of mTOR; PRAS40, proline rich Akt substrate 40 kDa; mLST8, mammalian lethal with Sec13 protein 8; Deptor, DEP domain-containing mTOR interacting protein; TSC1, tuberous sclerosis 1; IRS1, insulin receptor substrate 1; PIP2, phosphatidylinositol-3,4-bisphosphate; PIP3, phosphatidylinositol-3,4,5-triphosphate.

Figure 13. Samd4 may promote mTORC1 signaling that promotes glucose metabolism in muscle and adipose tissue. The mTORC1 complex regulates numerous cellular processes (left). mTORC1 positively regulates glucose metabolism, and we hypothesize that Samd4 is an upstream regulator of mTORC1 in this process (right). Both 14-3-3 and Akt are positive regulators of mTORC1 through inhibiting PRAS40 and TSC2. Samd4 interacts with 14-3-3 proteins in a muscle cell line and is phosphorylated by Akt in vitro.

Supermodel mice phenocopied muscle- and adipose-specific Raptor mutants (60;61). Raptor is scaffold protein that regulates substrate binding in the mTORC1 complex, a complex that regulates cell growth and metabolism and affects signaling pathways that regulate nutrient uptake, protein and lipid synthesis, energy expenditure, and autophagy (62). mTORC1 target eIF4E binding protein 1 (4E-BP1) and S6 (a substrate of the mTORC1 target S6 kinase 1 (S6K1)) were hypophosphorylated in the muscle and adipose tissues from homozygous supermodel mice (Figure 11A). C2C12 myoblast cells that stably expressed wild-type Flag-Samd4 or Flag-Samd4supermodel  were used to examine Samd4-interacting proteins. Immunoprecipitation identified 220 and 212 putative Samd4 interactors in the Samd4 and Samd4supermodel lysates, respectively. None of the known mTORC1 complex components were found in either of the Samd4 immunoprecipitates (1). However, six out of seven mammalian 14-3-3 protein isoforms interacted with wild-type Samd4; three of the isoforms were completely absent, while three were reduced in the Samd4supermodel  immunoprecipitate (1). An interaction between Samd4 and 14-3-3 proteins was confirmed by immunoprecipitation using Flag-Samd4-expressing C2C12 cells as well as 293T cells overexpressing Flag-Samd4 and Myc-14-3-3ζ (Figure 11B). However, Flag-Samd4supermodel  failed to pull down 14-3-3 proteins (Figure 11B). Consistent with these data, wild-type Samd4 but not Samd4supermodel was phosphorylated on serine(s) within a 14-3-3 binding motif (Figure 11C). As the Akt/PKB phosphorylation consensus sequence is similar to the 14-3-3 binding motif, Akt-mediated phosphorylation of Samd4 was examined by in vitro kinase assay. When incubated with ATP, Akt phosphorylated wild-type Samd4 but failed to phosphorylate Samd4supermodel (Figure 11D). Because the supermodel mutation lies outside putative Akt phosphorylation sites present in Samd4, it was proposed that the supermodel mutation may disrupt the conformation of Samd4, subsequently preventing its phosphorylation (1). Taken together with the supermodel phenotype, these results indicate that Samd4 functions, putatively through 14-3-3 proteins, in metabolic regulation in conjunction with mTORC1 (Figure 12).

Primers Primers cannot be located by automatic search.

Supermodel genotyping is performed by amplifying the region containing the mutation using PCR followed by sequencing of the amplified region to detect the nucleotide change.  The following primers were used for PCR amplification and sequencing:


Primers for PCR amplification




Primers for sequencing

Supermodel_seq (F): 5’- CACAGTTGAGAACATTTAGCCAG -3’

Supermodel_seq (R): same as PCR (R) primer, above.


PCR program

1) 94°C        2:00

2) 94°C        0:30

3) 57°C        0:30

4) 72°C        1:00

5) repeat steps (2-4) 29x

6) 72°C        7:00

7) 4°C           


The following sequence (from Genbank genomic region: NC_000080.5 of the linear genomic sequence of Samd4) is amplified:


133141                                      tgaccg atcccatcca cagttgagaa

133201 catttagcca gggttaagaa gtcgacgctc taagccacag aaatttagag tgcggtgtct
133261 cgtgtgatcg tgggtgtatt tgctcatgtg tcctgtgtac tttccccttc aggaatcatt
133321 aaccaatggc aacaggaatc caaggataaa gtgatttccc ttctgctaac tcacctgcct
133381 ttgctgaagc caggaaacct cgacgcgaaa gcagagtata tgaaactgct gcccaagatc
133441 ctggcacact ctatcgaaca caaccagcac attgaggaga gcaggcagct gctgtcctat
133501 gctttgatcc acccagccac ttccttggaa gaccgcagcg cactagccat gtggctgaat
133561 cacttggagg accgcacatc caccagcttt ggtagccaga accgaggccg ctcggactct
133621 gtggattatg ga


PCR primer binding sites are underlined. The forward sequencing primer binding sites are highlighted; overlapping sequence with the PCR primer is italicized.  The mutated A is highlighted in red.

Science Writers Anne Murray
Illustrators Victoria Webster
AuthorsSungyong Won, Elizabeth Hanley, Bruce Beutler