RIN Family Proteins (RIN1, RIN2, and RIN3)
The RIN1 and RIN2 genes were first identified in a selection for expressed human cDNAs that suppress phenotypes associated with oncogenic RAS mutations in a yeast model system (Colicelli et al. 1991). Subsequent analysis demonstrated that RIN1 binds specifically to activated (GTP-bound) human HRAS (Han and Colicelli 1995; Han et al. 1997), suggesting that RIN1 is a downstream effector of RAS family proteins. RIN2 and RIN3 also bind to activated HRAS (Rodriguez-Viciana et al. 2004). All three RIN proteins contain a carboxy terminal RAS association (RA) domain that mediates this interaction (Colicelli (2004)).
It was subsequently reported that RIN1 encodes a VPS9-related guanine nucleotide exchange factor (GEF) function specific for the RAB5 family of early endocytosis GTPase (Tall et al. 2001). RIN2 and RIN3 also encode VPS9-type GEF domains upstream of the RA domain. This suggests that RIN family proteins are RAS effectors that promote RAB5 function, connecting RAS activation to receptor endocytosis.
Another defining feature of RIN proteins is their amino terminal SRC homology 2 (SH2) domain. The RIN1 SH2 domain mediates binding to activated receptor tyrosine kinases (Barbieri et al. 2003), which likely facilitates endocytosis and downregulation of these receptors. The RIN2 and RIN3 SH2 domains show some species variation (discussed below).
RIN1 also binds and activates ABL1 and ABL2 tyrosine kinases (Cao et al. 2008; Hu et al. 2005). This capacity requires an amino terminal fragment of RIN1, which may not be functionally conserved in RIN2 and RIN3.
RIN Family: Evolution and General Properties
Specific properties of each are described in the following sections.
RIN1, the defining member of the RIN protein family, is a RAS effector that couples cell signaling with receptor trafficking and cytoskeletal remodeling. It is localized to the cytoplasm and the plasma membrane. Binding of RIN1 to RAS is GTP-dependent and mediated by the carboxy-terminal region of RIN1 (Han et al. 1997). RIN1 overexpression suppresses fibroblast transformation by HRASG12V, likely by competing with RAS effectors that promote mitosis and suppress apoptosis (Wang et al. 2002). RIN1 localizes in cytoplasmic and membrane compartments, and this partition is regulated through a 14-3-3 interaction. Phosphorylation of RIN1-Ser351 by PKD enhances 14-3-3 binding, and mutation of this residue shifts RIN1 localization to the plasma membrane (Wang et al. 2002). This result suggests that 14-3-3 binding reduces access to RAS proteins, which are membrane tethered.
RIN1 encodes a VPS9 subfamily GEF domain that mediates endosome fusion and receptor endocytosis through activation of RAB5 family proteins (Tall et al. 2001). RIN1 preferential associates with and activates the RAB5A isoform (Chen et al. 2009). The SH2 domain of RIN1 binds to activated (tyrosine phosphorylated) EGFR to promote receptor downregulation (Barbieri et al. 2003). Through a proline-rich domain, RIN1 may recruit STAM2, a component of the ESCRT (endosomal sorting complex required for transport) machinery, and facilitate trafficking of ubiquitinated EGFR to lysosomes (Kong et al. 2007). RIN1 also promotes internalization of the TGF-β receptor through RAB5 activation. In this case, however, the result is increased signaling through SMAD2/3 and the transcription repressor SNAI1 (Hu et al. 2008). The contribution of RIN1 to growth factor receptor internalization and signaling may account for the observed silencing of RIN1 in breast tumor cells (Milstein et al. 2007) and enhanced expression in non-small cell lung adenocarcinoma cells (Tomshine et al. 2009).
Other downstream effectors of RIN1 are the ABL family non-receptor tyrosine kinases. The ABL SH3 domain mediates binding to a proline-rich motif in RIN1, leading to ABL-mediated phosphorylation of RIN1-Y36 and subsequent association with the ABL SH2 domain (Han et al. 1997; Afar et al. 1997). Binding of RIN1 to ABL relieves autoinhibition and stimulates ABL tyrosine kinase activity (Cao et al. 2008; Hu et al. 2005). As a positive regulator of ABL activity, RIN1 regulates actin remodeling. Mammary epithelial cells from Rin1−/− mice show extensive peripheral-actin networks, enhanced attachment to fibronectin, and increased cell motility (Hu et al. 2005). RIN1 potentiates the catalytic and transforming activity of the BCR-ABL1 fusion oncogene, and RIN1 over-expression increases BCR-ABL1 mediated leukemogenesis in a mouse model system (Afar et al. 1997). Importantly, the ABL1T315I mutant resistant to therapeutic kinase inhibitors remains responsive to positive regulation by RIN1 (Cao et al. 2008), and Rin1−/− bone marrow cells were refractory to transformation by BCR-ABL1T315I (Thai et al. 2011). These results suggest that the RIN1::ABL1 interaction may be a drugable vulnerability of oncogenic ABL fusion proteins.
Rin1 is most strongly expressed in mature forebrain neurons, with moderate expression in hematopoietic and epithelial cells (Dzudzor et al. 2010). This restricted expression of Rin1 is mediated in part by SNAI1. The Rin1 promoter sequence also contains a consensus recognition site for the transcription repressor REST. Deletion of this region surprisingly led to a reduction in reporter gene expression, suggesting that elements in this region enhance expression and may be positively regulating Rin1 expression in neuronal cells (Dzudzor et al. 2010).
Rin1−/−mice are viable and fertile and show no gross morphological abnormalities. However, they have elevated amygdala LTP (long-term potentiation) and enhanced fear conditioning, suggesting that Rin1 normally acts as a negative regulator of synaptic plasticity in this region (Dhaka et al. 2003). In addition, Rin1−/− mice are deficient in conditioned fear extinction and latent inhibition (Bliss et al. 2010). Rin1−/− mice have normal hippocampal-dependent learning, as well as normal motor learning, anxiety, and exploratory behavior, suggesting that Rin1−/− mice may be a useful model for studying neuropsychiatric conditions such as PTSD (post-traumatic stress disorder).
The RIN2 gene is widely expressed in mouse, based on analysis of mRNA levels ((Kajiho et al. 2003), BioGPS.gnf.org). The SH2, GEF, and RA domains first characterized in RIN1 are well conserved in RIN2, which has demonstrable guanine nucleotide exchange activity on RAB5 (Saito et al. 2002) and RAS interaction properties (Rodriguez-Viciana et al. 2004). However, RIN2 gene products in primates (human and chimpanzee) differ from their orthologs in other vertebrates (cow, dog, mouse, opossum, chicken, frog, and fish) in two notable aspects. First, the amino termini of primate RIN2 proteins extend about 50 residues beyond other vertebrate RIN2 gene products. Second, an arginine residue critical for the phosphotyrosine binding function of SH2 domains (mouse Rin2: FLVR122) is instead a histidine in primate RIN2 (human RIN2: FLVH171).
RIN3 was first described as a novel RAB5 guanine nucleotide exchange factor, isolated from a human leukocyte cDNA library based on a yeast two-hybrid screen for RAB5BQ79L interacting proteins (Kajiho et al. 2003).
RIN3 contains the GEF, RA, and SH2 domains conserved throughout the RIN gene family. It demonstrates guanine nucleotide exchange activity for RAB5 (Kajiho et al. 2003) and associates with activated RAS (Rodriguez-Viciana et al. 2004). The SH2 domain sequence of RIN3, like that of RIN2, shows a curious divergence at the FLVR motif. The sequences of most vertebrates, including human, encode the arginine residue critical for phosphotyrosine binding. But several species (pig, cow, cat, rat, and mouse) have a cysteine substitution at this key position, implying that RIN3 in these organisms may function somewhat differently.
RIN3 also associates with BIN1 (a.k.a. amphyphisin II), a membrane-bending protein involved in endocytosis (Kajiho et al. 2003). RIN3 can translocate to RAB5 positive endosomes and deletion of the RA domain caused RIN3 to be constitutively located to endocytic vesicles, suggesting that the RA domain has an autoinhibitory effect on RIN3’s endosomal localization (Yoshikawa et al. 2008).
In mice, RIN3 expression is highly enriched in mast cells with lower expression levels in other hematopoietic tissues including lymph node, bone, and T cells (http://symatlas.gnf.org). Protein levels in established human cell lines also indicate enrichment in mast cells (CJ and JC 2011).
RIN1-3 proteins connect RAS signal transduction with RAB5 activation and receptor endocytosis. RIN1 has a special role in ABL tyrosine kinase regulation, with a likely contribution to cytoskeleton remodeling. A mouse knockout model suggests RIN1 involvement in stimulation-induced signal transduction in multiple cell types. Mouse models of RIN2 and RIN3 deficiencies have not yet been described.
There are several outstanding questions to be answered regarding the biochemistry of RIN proteins. Particularly curious is a species-specific divergence in the SH2 domain arginine residue required for phosphotyrosine binding. Does this imply an alternate function for these SH2 domains? Further study is also needed to identify the full range of protein partners, and to determine the location and consequence of these interactions.
Evidence for the involvement of RIN proteins in human pathologies is just beginning to emerge. RIN1 appears to collaborate in tumorigenesis at multiple levels in ways that are complex and cell type specific. In addition, the fear learning and extinction phenotypes of Rin1−/− mice suggest that reduced RIN1 function in forebrain neurons could contribute to post-traumatic stress disorder. The correlation of RIN2 deficiency with connective tissue disorders provides another clear indication that RIN proteins play diverse and essential roles in human physiology.