Lysosomal exocytosis and resealing of damaged plasma membrane are essential for cellular homeostasis and tumor invasion. However, very little is known of the molecular machinery that regulates these physiological processes. Moreover, no mutations in any of the known regulators of lysosomal exocytosis in primary tumors of patients have been characterized. Here we demonstrate that RNF167-a, a lysosomal-associated ubiquitin ligase, negatively regulates lysosomal exocytosis by inducing perinuclear clustering of lysosomes. Importantly, we also characterized a set of novel natural mutations in RNF167-a, which are commonly found in diverse tumor types. We found that RNF167-a-K97N mutant, unlike the wild type, localizes in the cytoplasm and does not promote perinuclear lysosomal clustering. Furthermore, cells expressing RNF167-a-K97N exhibit dispersed lysosomes, increased exocytosis and enhanced plasma membrane repair. Interestingly, these functional features of RNF167-a-K97N were shared with a naturally occurring short version of RNF167 (isoform RNF167-b). In brief, the results presented here reveal a novel role of RNF167-a, as well as its natural variants RNF167-a-K97N and RNF167-b, as an upstream regulator of lysosomal exocytosis and plasma membrane resealing.
Lysosomes are specialized organelles with an inherent function in recycling and degrading the cargo delivered to them through endocytosis, phagocytosis and autophagy. In recent years, lysosomes have emerged as dynamic structures regulating physiological and pathological cellular processes such as antigen presentation, destruction of intracellular pathogens, plasma membrane repair, tumor invasion and metastasis, apoptotic cell death, metabolic signaling and gene regulation (Pu et al., 2016). These diverse functions of lysosomal vesicles are primarily facilitated by their positioning in the cells and bidirectional movement between the plasma membrane and perinuclear region (Bonifacino and Neefjes, 2017). The lysosomal movement towards the plus-end of microtubules (anterograde) and minus-end (retrograde) is mediated by kinesins and dynein motor proteins (Harada et al., 1998; Hollenbeck and Swanson, 1990), respectively. These motors generally do not bind lysosomes directly, instead the interaction is mediated by small GTPases like Rab7 (herein referring to the RAB7A isoform), ARL8B and their effector proteins and lipids (Bonifacino and Neefjes, 2017).
Under basal conditions, most lysosomes are concentrated in the center, known as the perinuclear cloud, and a few in the cell periphery (Jongsma et al., 2016). The distribution and motility of lysosomes are subject to change depending on various conditions. For instance, cytosolic acidification promotes anterograde movement whereas alkalinization favors retrograde transport of lysosomes (Heuser, 1989; Parton et al., 1991). In general, perinuclear accumulation of lysosomes occurs during starvation (Korolchuk et al., 2011), aggresome formation (Zaarur et al., 2014), drug-induced apoptosis (Yu et al., 2016), expression of pathogenic mutant forms of huntingtin (Erie et al., 2015) or leucine-rich repeat kinase 2 (LRRK2) (Dodson et al., 2012) and in lysosomal storage diseases (Li et al., 2016; Uusi-Rauva et al., 2012). Several protein complexes have been shown to regulate the intracellular positioning and movement of lysosomes. For example, the anterograde movement of lysosomes is mediated by a multisubunit complex, BLOC-1 related complex (BORC), which recruits ARL8B and its effector molecule SKIP (also known as PLEKHM2) (Pu et al., 2015; Rosa-Ferreira and Munro, 2011), whereas retrograde movement is mediated by Rab7 along with its effector molecule RILP (Jordens et al., 2001). Lysosomal movement and positioning play a decisive role in maintaining cellular homeostasis, dysregulation of which is a common event in various pathological conditions such as cancer and neurodegenerative diseases (Pu et al., 2016).
Extracellular acidification, a tumor microenvironment stimulus, promotes redistribution of lysosomes towards the cell periphery and this anterograde movement of lysosomes facilitates cancer growth, invasion and metastasis (Pu et al., 2016). During invasion through the dense extracellular matrix of neighboring tissue spaces or in metastasis, tumor cells are prone to plasma membrane damage (Jaiswal et al., 2014). Tumor cells cope with these kinds of stresses by upregulating plasma membrane resealing (PMR) via lysosomal exocytosis. The damaged plasma membrane triggers an elevation in intracellular Ca2+ levels, resulting in anterograde movement of lysosomes. This peripheral pool of lysosomes docked to the cell surface fuses with the plasma membrane and contributes to resealing of the damaged plasma membrane (Medina et al., 2011). It has been shown that S100A11, an annexin-binding protein that is overexpressed in various tumors, enhances PMR in metastatic cells and thereby protects them from plasma membrane stresses (Jaiswal et al., 2014). Importantly, inhibition of lysosomal exocytosis could reverse invasiveness and chemoresistance in aggressive sarcoma cells (Machado et al., 2015). Inhibitors of sodium-proton exchangers (NHEs) have been shown to be very effective in preventing cell-surface directed lysosome trafficking, thereby decreasing lysosomal exocytosis and cell invasion (Steffan et al., 2009), underlining the importance of lysosomal exocytosis in tumor progression.
Recent studies on RNF26, an endoplasmic reticulum (ER)-associated ubiquitin ligase demonstrated the role of ubiquitin ligases in late endosomal (LE) positioning (Jongsma et al., 2016). It has been reported that ER localization and RING domain activity of the RNF26 protein are essential for the perinuclear clustering of LE compartments. Another membrane-associated ubiquitin ligase is RNF167. Although some reports suggest RNF167 association with early endosomes (van Dijk et al., 2014), we and others have found that it mostly localizes to lysosomes (Lussier et al., 2012), with RNF167 overexpression inducing perinuclear clustering of lysosomes upon treatment with acetate Ringer's solution (Deshar et al., 2016). This ubiquitin ligase consists of an N-terminal signal peptide (SP), a protease-associated (PA) domain, a transmembrane (TM) domain followed by a really interesting new gene (RING) domain (Lussier et al., 2012; Nakamura, 2011; Yamazaki et al., 2013). It has been shown that the PA domain is essential for endosomal localization (van Dijk et al., 2014), whereas the RING domain confers ubiquitin ligase activity to the protein (Lussier et al., 2012; van Dijk et al., 2014). Information from the NCBI database suggests the existence of six different isoforms of RNF167 (Fig. S1A): RNF167-a (350 amino acids), RNF167-b (315 amino acids), RNF167-c (374 amino acids), RNF167-d (349 amino acids), RNF167-e (339 amino acids) and RNF167-f (308 amino acids), with one of the isoforms (RNF167-b) differing in the first 35 amino acids, which includes a putative SP, with respect to RNF167-a. At present, experimental evidence lacks information about the cellular localization and function of the RNF167-b isoform. In this study, we provide evidence to establish that RNF167-a is a negative regulator of lysosomal exocytosis. In contrast, RNF167-b and a naturally occurring tumor-associated mutant enhance lysosomal exocytosis and PMR. In summary, we report here for the first time how tumor-associated mutations could facilitate lysosomal exocytosis and might be responsible for plasma membrane repair in tumor cells.
Identification and differential subcellular localization of RNF167 isoforms
RNF167, a ubiquitin ligase of 350 amino acids is an endosomal/lysosomal associated protein with a C-terminal RING domain and N-terminal SP followed by a PA domain and a TM domain. However, in the NCBI database, we noted the presence of multiple RNF167 variants of different lengths (Fig. S1A), suggesting the existence of multiple RNF167 isoforms. Interestingly, a few of the transcripts, particularly the RNF167-b form, lack the region coding for the first 35 amino acids, raising the possibility that this particular form may differ from the full-length RNF167 (RNF167-a) in its localization to the membrane vesicles and lysosomal-associated function (Fig. 1A). To test such a possibility, first we assessed the expression of RNF167-b. We isolated total RNA from multiple human cell lines and performed nested PCR using primers to amplify amplicons of 334 bp and 138 bp for RNF167-a and RNF167-b isoforms (Fig. 1B). A strong band corresponding to 334 bp, expected from the expression of RNF167-a, was observed in all the cells that were tested. However, some of the cell lines also resulted in a 138 bp band, representing the expression of RNF167-b. To confirm that the 138 bp fragment was indeed amplified from RNF167-b, the DNA bands from colon cancer COLO 205 cell lines were extracted, sequenced and confirmed to contain sequences corresponding to RNF167-b (Fig. S1B). These results established the expression of a RNF167 variant without a predicted signal sequence.
To investigate the significance of putative SP, for subcellular localization of RNF167 and its function, we first generated a C-terminal EYFP-tagged RNF167-b (RNF167Δ1-35-EYFP) and RNF167-a and evaluated their expression pattern in HeLa cells. Interestingly, unlike RNF167-a, which showed punctate intracellular staining, the RNF167-b variant exhibited diffuse cytosolic staining indicative of a requirement for SP for localization to intracellular compartments. HeLa cells transfected with RNF167-EYFP isoforms were immunostained with GFP and antibodies specific for lysosomal membrane-associated protein 1 (LAMP1), a lysosomal marker. The results showed distinct colocalization of RNF167-a and lysosomes, whereas there was minimal association of RNF167-b with lysosomes (Fig. 1C). Quantification of colocalization and associated Pearson's coefficient values (Fig. 1D) revealed the requirement for SP for RNF167 localization to lysosomes. Interestingly, the SP of RNF167, (1-35)-EYFP, exhibited diffuse cytosolic staining. The results indicate that SP is essential but not sufficient for lysosomal localization of the protein (Fig. S1C).
Because expression of RNF167-a is reported to result in perinuclear clustering of lysosomes upon treatment with acetate Ringer's solution (Deshar et al., 2016), we examined the effect of acetate Ringer's solution on lysosomal clustering in cells expressing RNF167-b. HeLa cells transfected with plasmids carrying cDNA for EYFP or for C-terminal EYFP fusion of RNF167-a or RNF167-b were treated with acetate Ringer's solution (pH 6.9) and stained with anti-LAMP1 and anti-GFP antibodies. As expected, cells expressing EYFP showed dispersed lysosomes upon treatment. We noted that, consistent with the inability of RNF167-b to localize to lysosomal compartments, cells expressing RNF167-b showed substantially more dispersed lysosomes than cells expressing RNF167-a (Fig. 1E). To validate these findings independently, we assessed the distribution of lysosomes in cells by calculating the perinuclear index (PNI), a widely accepted method for quantifying lysosomal positioning (Marwaha et al., 2017). The PNI of RNF167-b was 23.24 compared with 93.57 for RNF167-a, revealing a regulatory role of RNF167-b variant in lysosomal positioning (Fig. 1F).
RNF167-a expression induces perinuclear clustering of lysosomes, but not mitochondria or early endosomes
To determine the specificity of RNF167-a mediated lysosomal positioning, we assessed the positioning of various intracellular compartments such as endosomes and mitochondria using the organelle-specific markers EEA1 (early endosomes), LAMP1 (lysosomes) and TOM20 (mitochondria) in cells expressing EYFP or RNF167-a-EYFP in the presence and absence of acetate Ringer's solution (pH 6.9). Treatment of the cells with acetate Ringer's solution is reported to affect mainly lysosomes, not other organelles (Heuser, 1989). As expected, we found that LAMP1-positive vesicles accumulated in the perinuclear region of RNF167-expressing cells, but the effect on distribution of other organelles was minimal (Fig. 2A; Figs S2,S3). Organelle distribution was quantified by calculating the PNI, as shown in Fig. 2B. Collectively, these findings suggest that RNF167-a status might be a major modifier of lysosomal positioning, which is important for physiological processes such as repair of the damaged plasma membrane and cellular invasion (Pu et al., 2016).
RNF167-a mediated lysosomal positioning is dependent on its E3 activity and localization to lysosomes
To establish that the localization of RNF167-a to the lysosomal compartments is essential for lysosomal clustering, we investigated the structure–function relationship of the PA domain in RNF167. To this end, we examined the lysosomal positioning in cells expressing RNF167ΔPA (Fig. 2C), a domain reported to be essential for the endosomal localization of RNF167 (van Dijk et al., 2014). As was the case with RNF167-b (Fig. 1C–E), RNF167ΔPA variant also failed to localize to the lysosomal compartments (Fig. 2D) and induce perinuclear clustering. Moreover, an E3 ligase-inactive mutant (RNF167-C233S) with an intact PA domain also failed to cluster lysosomes, suggesting the importance of ubiquitin ligase activity for lysosomal clustering. Quantification of the clustering, as measured by the PNI, is shown in Fig. 2E.
Knockdown of RNF167 with siRNA affects lysosomal positioning
To demonstrate further the role of RNF167, we next suppressed the levels of endogenous RNF167 in HeLa cells using selective siRNAs designed to target two independent regions of 3′UTR in RNF167. Transfection of specific siRNAs efficiently downregulated the expression of endogenous RNF167 mRNA compared with control siRNA (Fig. S4A). Lysosomal positioning in these cells with and without acetate Ringer's solution (pH 6.9) treatment was examined by immunostaining using anti-LAMP1 antibody and assessed in terms of the PNI (Fig. S4B,C). In contrast to untreated cells, a substantial difference in the lysosomal pattern (as reflected by the PNI) was observed in cells treated with acetate Ringer's solution (pH 6.9) (Fig. S4B,C). In cells transfected with control siRNA, the majority of lysosomes accumulated in the perinuclear region. Interestingly, the depletion of endogenous RNF167 resulted in the dispersal of LAMP1-positive compartments, with a significant shift of lysosomes from the perinuclear region to the periphery of the cells (Fig. 3A), indicative of anterograde transport. As expected, the PNI of specific siRNA-transfected cells was significantly less than for control siRN-treated cells (Fig. 3B). To corroborate the results further, we assessed lysosomal distribution by line profiling as an established method of choice (Willett et al., 2017). In cells treated with control siRNA, lysosomes were dispersed throughout the cells with a significant population at the perinuclear region (within 5 µm distance of the nuclear surface). On the other hand, RNF167 depletion resulted in increased accumulation of lysosomes at the cell periphery, which is defined as more than 15 µm from the nuclear surface (Fig. 3C). We further confirmed the specificity of the RNF167-mediated effect on lysosomes by performing experiments wherein HeLa cells were transfected with control siRNA or specific siRNA and reconstituted with EYFP, RNF167-a-EYFP or RNF167-b-EYFP. As expected, RNF167 siRNA transfection showed lysosomal dispersal in cells reconstituted with empty vector (Fig. 3D). A reversal in RNF167 depletion-induced lysosomal dispersion was observed in those cells cotransfected with RNF167-a-EYFP, but not RNF167-b-EYFP. Quantification of the clustering as measured by the PNI is shown in Fig. 3E. It is evident that those cells reconstituted with RNF167-a exhibit a higher PNI compared with RNF167-b, indicating that RNF167-a, not RNF167-b, selectively clusters lysosomes. Collectively, these findings indicate the importance of isoforms of RNF167 in lysosomal positioning upon extracellular acidification.
Perinuclear lysosomal clustering by RNF167-a is dynein dependent
In mammalian cells, long-range transport of organelles is mediated by microtubule-based motor proteins, including plus end-directed kinesins and minus end-directed dynein complexes (Rosa-Ferreira and Munro, 2011). Given that RNF167-a promotes the accumulation of lysosomes towards the perinuclear region, we hypothesized that RNF167-a-mediated juxtanuclear positioning of lysosomes is dynein dependent. To test this hypothesis, we investigated the lysosomal positioning in HeLa cells transfected with RNF167-a with or without p50-Dynamitin, overexpression of which is known to disrupt the dynein-dynactin complex (Choy et al., 2018). As postulated, we found that in cells expressing RNF167-a the peripheral pool of LAMP1-positive vesicles was significantly increased with cotransfection of Dynamitin-GFP compared with GFP alone (Fig. 4A). These observations suggest that RNF167-a regulates lysosomal positioning via a dynein-dependent retrograde transport mechanism. To validate the role of the dynein motor in this process further, experiments were done in the presence of the dynein inhibitor Ciliobrevin D (Firestone et al., 2012; Sainath and Gallo, 2015). HeLa cells transfected with RNF167-a were treated with Ciliobrevin D for 2 h before assaying for the lysosomal clustering. We noticed that RNF167-a could effectively cluster lysosomes in the perinuclear region in control cells, whereas treatment with Ciliobrevin D abrogated the lysosomal clustering (Fig. 4B), with an increased number of LAMP1-positive compartments in the cell periphery. These results indicate that perinuclear clustering driven by RNF167-a is specific for lysosomes and is dynein dependent.
RNF167-a regulates lysosomal exocytosis
To understand the physiological significance of RNF167-a, we focused on the regulation of lysosomal exocytosis in cells expressing RNF167-a. Eukaryotic cells reseal the damage on their plasma membrane by a process known as lysosomal exocytosis. Such plasma membrane damage is frequently observed in cells after exposure to mechanical stress or during metastasis (Castro-Gomes et al., 2016; Jaiswal et al., 2014). Damage to the plasma membrane leads to increased intracellular Ca2+ ions, lysosomal sensing of elevated intracellular Ca2+ and movement of lysosomes towards the plasma membrane in an anterograde fashion to help to repair the damage (Reddy et al., 2001).
We investigated the involvement of RNF167-a in lysosomal exocytosis caused by treatment with ionomycin, a known enhancer of intracellular Ca2+ levels (Encarnação et al., 2016; Rodríguez et al., 1997) for 10 min after 2 h of treatment with acetate Ringer's solution (pH 6.9). Exocytosis was measured by staining for LAMP1 on the plasma membrane using luminal epitope-specific mouse anti-LAMP1 antibody. As expected, the cell surface expression of LAMP1 in control HeLa cells was significantly increased upon ionomycin treatment (1.33% versus 86%). In contrast, cells stably expressing RNF167-a showed a substantial reduction in the surface expression of LAMP1 compared with control cells under the same conditions (3.33% versus 86% in control cells; Fig. 5A,B, upper panel). Three-dimensional images of the surface expression of LAMP1 are shown in Fig. S5A. We also evaluated the effect of RNF167-a in the presence or absence of extracellular Ca2+ using streptolysin-O (SLO), a bacterial toxin that induces damage in the plasma membrane. Consistent with results obtained from ionomycin treatment, HeLa cells stably expressing empty vector showed 72.2% LAMP1 staining at the plasma membrane in the presence of SLO+Ca2+, whereas RNF167-a positive cells showed only 19.5% under the same conditions (Fig. 5A,B, lower panel).
Conversely, cells in which RNF167 was depleted exhibited an increased cell surface expression of LAMP1 (Fig. 5C,D), indicative of exocytosis. To assess whether the effect of siRNA was specific and not off-target, we also looked at the surface staining of LAMP1 in siRNA-expressing HeLa cells transfected with EYFP, RNF167-a-EYFP or RNF167-b-EYFP. As expected, ionomycin treatment significantly increased the surface expression of LAMP1 in all the combinations used (Fig. 5E; Fig. S5B). However, cells reconstituted with RNF167-a-EYFP had less staining on the PM compared with cells expressing EYFP and RNF167-b-EYFP, as shown by quantification (Fig. 5E). Expression levels of constructs used for rescue experiments are shown in Fig. 5F and Fig. S5C. Together, these findings suggest that RNF167-a influences lysosomal exocytosis by regulating the retrograde movement of lysosomes.
Naturally occurring mutations in RNF167-a perturb its function
Given the importance of RNF167-a for perinuclear clustering of lysosomes and lysosomal exocytosis, we next investigated the effect of a set of naturally occurring mutations identified in RNF167-a from diverse cancer-types (http://cancer.sanger.ac.uk). We assessed the ability of tumor-associated (TA) mutants of RNF167-a (K97N, R200Q, C268R, R277L and L329P) to regulate lysosomal clustering and exocytosis. Results presented in Fig. 6A indicate that the effects of R200Q, C268R, R277L and L329P mutant versions of RNF167-a on lysosomal clustering were minimal. However, RNF167-K97N with a missense mutation in the PA domain exhibited distinct cytosolic expression and also prevented the perinuclear clustering of lysosomes. This suggests that one of the naturally occurring mutations in human tumors affects lysosomal clustering (Fig. 6B). Because lysosomal distribution is known to affect tumor progression and metastasis, we investigated the effect of these mutations on lysosomal exocytosis. Consistent with the results shown in Fig. 6A,B, only K97N but not other mutations or the wild type showed a substantial increase in surface expression of LAMP1 (Fig. 6C,D; Fig. S6B). Because RNF167-K97N is unable to induce perinuclear clustering, a greater number of lysosomes are available for docking and fuse with the plasma membrane, as evidenced by the increased surface expression of LAMP1. This indicates that the unhindered anterograde movement of lysosomes is an added advantage for tumor cells, mainly because these lysosomes can promote tumor progression by enhanced lysosomal exocytosis. Given that lysosomal clustering affects PMR, as shown with Rab3a depletion (Encarnação et al., 2016), we sought to determine the effect of RNF167-a and its variants on PMR.
Cells expressing tumor-associated mutant RNF167-a-K97N can efficiently repair the plasma membrane
We examined the effect of lysosomal exocytosis mediated by RNF167-a-K97N on the ability of cells to reseal their plasma membrane. HeLa cells stably expressing the control vector, RNF167-a or RNF167-a-K97N were treated with acetate Ringer's solution (pH 6.9) for 2 h followed by 200 ng/ml Streptolysin-O (SLO) for 5 min. SLO, a bacterial toxin, induces pore formation in the plasma membrane and is commonly used to study membrane resealing (Corrotte et al., 2015). Briefly, cells were allowed to reseal the damage by incubating them in Hanks' balanced salt solution (HBSS) with or without Ca2+ at 37°C for 10 min. The efficiency of resealing was measured by propidium iodide (PI) staining as PI is impermeable to cells with an intact plasma membrane. As established earlier (Appelqvist et al., 2013), the percentage of PI staining in control cells upon Ca2+ treatment was much less than in those without Ca2+, confirming the requirement for Ca2+-dependent lysosomal exocytosis in maintaining membrane integrity. Consistent with lower LAMP1 surface expression (Fig. 5A,B), cells expressing RNF167-a were unable to repair the membrane resulting in a substantial increase in PI-positive cells, even in the presence of Ca2+, thus emphasizing the importance of lysosomal exocytosis in maintaining plasma membrane integrity and its regulation by RNF167-a (Fig. 7A,B). Importantly, cells expressing RNF167-a-K97N, a localization defective mutant with more dispersed lysosomes, were able to reseal the damage more effectively than cells expressing wild-type RNF167-a, leading to lower PI staining (Fig. 7C,D). This enhanced resealing was found only in cells expressing RNF167-a-K97N but not in cells expressing the other tested tumor-associated RNF167-a mutants. These ineffective mutants showed lysosomal clustering in the perinuclear region (Fig. 6A). This suggests that lysosomal exocytosis represents an important way to confer a survival advantage to cancer cells, in addition to other processes. The expression levels of RNF167 variants used is shown in Fig. S6C. Consistent with the above results, increased PMR in cells expressing RNF167-a-K97N as compared with RNF167-a was also confirmed by FACS analysis (Fig. 7E). It is evident from the PI fluorescence intensity that cells expressing empty vector or RNF167-a-K97N could effectively reseal SLO-induced damage in the presence of Ca2+ whereas RNF167-a was unable to do so, as shown in Fig. 7F.
Differential regulation of lysosomal exocytosis by isoform b
As shown in Fig. 1C,D, RNF167-b isoform is unable to associate with lysosomes or to induce perinuclear clustering of lysosomes. To determine whether this naturally occurring shorter version affects lysosomal exocytosis, we compared the cell surface expression of LAMP1 in HeLa cells stably expressing the RNF167-a or RNF167-b isoforms. Cells were treated with ionomycin at 37°C for 10 min after 2 h of incubation with acetate Ringer's solution (pH 6.9). Cells were immediately transferred to ice and stained with the luminal epitope-specific mouse anti-LAMP1 antibody. As expected, we noticed significant surface expression of LAMP1 in cells stably expressing RNF167-b isoform as compared with cells expressing RNF167-a isoform (Fig. 8A,B). Having found increased surface expression of LAMP1 in RNF167-b isoform-expressing cells, we measured the PMR capacity by PI staining as described above. As expected, cells expressing the RNF167-a isoform had a strong impairment in PMR. By contrast, RNF167-b isoform-expressing cells displayed efficient resealing, as assessed by a lower percentage of PI-positive cells in the presence of extracellular Ca2+ (Fig. 8C,D). In the presence of Ca2+, the RNF167-b isoform had about ninefold less PI-positive cells compared with cells expressing RNF167-a (9.66% PI-positive cells in the b-isoform and 89% in a). These findings indicate that cells stably expressing the RNF167-b isoform exhibit increased PMR through Ca2+-dependent lysosomal exocytosis upon extracellular acidification.
To further assess the effect of RNF167 expression on lysosomal exocytosis, we also measured the enzymatic activity of β-hexosaminidase, a lysosomal enzyme, in the culture supernatant. HeLa cells stably expressing empty vector, RNF167-a, RNF167-b or RNF167-a-K97N were treated with acetate Ringer's solution (pH 6.9) for 2 h followed by 5 µM ionomycin treatment for 10 min. Ionomycin treatment resulted in an expected release of β-hexosaminidase in empty vector cells. Consistent with results shown in Fig. 5A and Fig. S5A, a significant decrease in β-hexosaminidase activity was observed in cells expressing RNF167-a compared with empty vector. RNF167-a-K97N and RNF167-b showed no significant differences in activity (Fig. 8E), demonstrating the negative regulation of lysosomal exocytosis by RNF167-a.
We also assessed the lysosomal exocytosis in HeLa cells transiently transfected with EYFP-tagged RNF167-a and RNF167-b. As expected, cells expressing RNF167-b-EYFP exhibited increased surface expression of LAMP1 upon ionomycin treatment (Fig. S6D). This was in line with our finding that cells expressing RNF167-b-EYFP have more dispersed lysosomes upon extracellular acidification. Collectively, our results suggest that the two isoforms of RNF167 could affect tumor progression differently by promoting opposing functions in the regulation of lysosomal positioning, exocytosis and PMR.
Lysosomes undergo long-range intracellular transport along microtubules with the help of motor proteins (kinesins and dyneins) and short-range transport via actin-based myosin motors (Bonifacino and Neefjes, 2017). In this study, we show that RNF167-a, a lysosome-associated ubiquitin ligase, regulates the intracellular movement of lysosomes. We found that overexpression of RNF167-a promotes dynein-mediated retrograde transport of lysosomes, whereas depletion of endogenous RNF167 results in their anterograde movement. Our findings are further supported by the recent observation that ARL8B, one of the key regulators of lysosomal anterograde movement, is a substrate for RNF167 (Deshar et al., 2016). We demonstrated that the PA domain of RNF167-a is required for its localization to the lysosomal compartments and is essential for the perinuclear clustering of lysosomes. We further observed that RNF167-a localizes to Rab7-positive endosomes in addition to lysosomes (data not shown), raising the possibility that RNF167-a can also affect conventional Rab7-RILP mediated retrograde movement of lysosomes along the microtubules. It would be interesting to elucidate the role of RNF167 in recruiting dynein motors to lysosomes.
Interestingly, different mutations in RNF167-a have been identified in several patient tumor samples. Analysis of the COSMIC database (http://cancer.sanger.ac.uk/cosmic) shows that 67.16% of changes are a result of missense substitution, followed by 19.4% synonymous substitution, 10.45% frameshift deletion and 1.49% each of inframe deletion and nonsense substitutions. RNF167 is a ubiquitin ligase, and mutations in the RING domain can affect its ligase activity. Moreover, mutations in other domains such as PA or SP can affect not only the stability of the protein but also its localization and function. RNF167-a-V98G, a TA mutant identified in patients with lung cancer, harbors a missense mutation in the PA domain that was shown to abrogate localization of the protein to endosomal compartments (van Dijk et al., 2014). Because of the cross-talk between lysosomal positioning and PMR and its importance in tumor progression, the effect of RNF167-a and its TA mutations on lysosomal exocytosis and PMR processes have been investigated in this paper. Out of the five mutations tested (K97N, R200Q, C268R, R277L and L329P), one mutation in the PA domain (K97N) was unable to localize to lysosomes or induce perinuclear clustering. This is consistent with the observations made using RNF167ΔPA (Fig. 2D), indicating the importance of lysosomal localization for RNF167-a function. To our surprise, RNF167-C268R, which harbors a missense mutation in the RING domain, was still able to induce perinuclear clustering, unlike RNF167-C233S, a ligase-inactive mutant that has been reported to block perinuclear clustering of lysosomes (Deshar et al., 2016). To understand these differences, we assessed the lysosomal association of C268R and C233S, two distinct ligase-inactive mutants of RNF167 by measuring the Pearson's coefficient. It was evident from the Pearson's coefficient that C268R had a better lysosomal association than C233S (Fig. S6A). The differences in clustering may be attributed to differences in their intracellular localization or reasons yet to be found.
Because cytosolic variants of RNF167 (K97N or the b-isoform) are unable to cluster lysosomes, it is logical to think that RNF167-a exerts its function by directly targeting lysosome-associated proteins that regulate lysosomal motility. One of the lysosomal proteins reported in overexpression studies to be targeted by RNF167-a is ARL8B (Deshar et al., 2016), an ADP-ribosylation factor-like protein that promotes the anterograde movement of lysosomes. The fact that a E3 ligase-inactive mutant (RNF167-a-C268R) was not able to induce lysosomal positioning changes indicates that (1) degradation of ARL8B is not required for lysosomal mobility, (2) RNF167 E3 mutant can degrade ARL8B indirectly via a complex involving a to-be-identified E3, or (3) RNF167-a exerts its effect by involving other lysosomal positioning mechanisms such as calcium signaling, cholesterol accumulation (Johansson et al., 2007) or pH that are independent of E3 function. Future studies in this direction are necessary to unravel these processes.
Intracellular mobility of lysosomes plays an important role in tumor cell invasion and metastasis (Steffan et al., 2010). Cells are prone to plasma membrane damage during invasion of the extracellular matrix; tumor cells partly offset this damage by enhanced lysosomal exocytosis, which helps in plasma membrane repair (Jaiswal et al., 2014). Acidic extracellular pH, a specific feature of the tumor microenvironment, has been shown to stimulate the anterograde movement of lysosomes. Hence, it is conceivable that mutations in regulatory molecules of exocytosis favoring anterograde movement are beneficial to tumor progression. In fact, it is known that tumor cells change the expression level of some of the regulators of anterograde movement for their advantage. For example, skeletal muscle sarcoma cells enhance lysosomal exocytosis by downregulating the expression of NEU1, a sialidase and known negative regulator of lysosomal exocytosis. Deficiency of NEU1 leads to the accumulation of over-sialylated LAMP1, which actively participates in lysosomal exocytosis (Machado et al., 2015).
To our knowledge, no TA mutations in regulators of lysosomal exocytosis have been characterized. In this report, we investigate the effect of several TA mutations of RNF167-a on lysosomal exocytosis and PMR and demonstrate that one of the mutations, K97N, affected lysosomal exocytosis (measured by immunostaining using luminal epitope-specific anti-LAMP1 antibody) and PMR. The K97N mutant was reported in kidney tumor samples and the lysine residue at the 97th position is highly conserved. The reason for increased lysosomal exocytosis and PMR exhibited by K97N-expressing cells is because of the dispersed lysosomes compared with cells expressing wild-type RNF167-a. The effect on lysosomal exocytosis is unique for K97N but not for the other mutations (C268R, R277L and L32P) tested. Also, K97N but not other mutants showed defects in lysosomal localization. It is possible that these mutations of RNF167 can influence yet-to-be-discovered pathways and result in tumor progression without affecting lysosomal distribution. It would be interesting to look at lysosomal positioning and exocytosis in primary cells derived from tumor samples of patients with mutations in RNF167-a.
Interestingly, our findings also reveal an important role for the RNF167-b isoform, which lacks the SP domain. We demonstrated that SP is also required for lysosomal localization in addition to the PA domain, as RNF167-b showed cytosolic localization. We confirmed the expression of RNF167-b in COLO-205 cells by sequencing the amplified PCR products. Moreover, we found that RNF167-b, unlike RNF167-a, enhances lysosomal exocytosis and PMR. We provide evidence for functional differences between the two variants. Differential expression of isoforms with antagonistic functions has been associated with tumor progression. For example, Intersectin1 (ITSN1) is a highly conserved scaffold protein and exists in two isoforms referred to as the long isoform (ITSN1-L) and short isoform (ITSN1-S). ITSN1-S promotes glioma development whereas ITSN1-L inhibits glioma progression both in vivo and in vitro. These opposing functions of ITSN1-S and ITSN1-L are highly regulated by alternative splicing (Shao et al., 2019). Similarly, alternative splicing of CD99 leads to two isoforms, full-length (CD99wt) and a truncated short form (CD99sh) (Bernard et al., 1997; Hahn et al., 1997). CD99 is a transmembrane glycoprotein that plays a crucial role in cell adhesion, apoptosis, differentiation of T cells and thymocytes (Alberti et al., 2002; Bernard et al., 1997), etc. Studies from osteosarcoma and prostate cancer cells revealed that the full-length form inhibits anchorage-dependent growth, migration and metastasis whereas the shorter form promotes tumor progression (Scotlandi et al., 2007). In addition to this, alternative splicing of BCLX (also known as BCL2L1) produces two isoforms, BCLX (L) and BCLX (S). BCLX (L) promotes cell growth and proliferation and inhibits programmed cell death, whereas BCLX (S) promotes cell apoptosis (Boise et al., 1993; Schwerk and Schulze-Osthoff, 2005).
Hence, it is conceivable that isoforms of RNF167 could also play a major role in maintaining homeostasis at the organism level. In summary, our study identifies RNF167-a as an important regulator in Ca2+-dependent lysosomal exocytosis and PMR. We demonstrate that K97N, one of the tumor-associated mutants of RNF167-a, is unable to localize to the lysosomal compartments and that cells expressing this mutant efficiently repair the plasma membrane by enhanced lysosomal exocytosis. We also show that a localization defective variant of RNF167, isoform b, lacking amino acids 1-35, also fails to cluster lysosomes in the perinuclear region and shows a phenotype similar to that of RNF167-a-K97N. Our data reveal new clues for the regulation of lysosomal exocytosis and tumor progression.
MATERIALS AND METHODS
Cell culture and stable cell preparation
HeLa and HEK-293T cells were cultured in Dulbecco's modified Eagle's medium (DMEM with 4.5 g/l glucose, L-glutamine, 3.7 g/l sodium bicarbonate and sodium pyruvate) (AL007A, HiMedia, Mumbai, Maharashtra, India) supplemented with 10% FBS (10270-106, Gibco-Thermo Fisher Scientific, Waltham, MA, USA) and 1% penicillin-streptomycin (A001A, HiMedia, Mumbai, Maharashtra, India). All cells were kept in an incubator at 37°C with 5% CO2 and tested routinely for contamination. Cell lines used in this study were all obtained from ATCC.
HeLa cells stably expressing wild-type RNF167 was generated as follows: HEK-293T cells were transfected with 1:1:2 ratio of pCAG-VSVG, pUMVC and pMYs-IRES-Puro RNF167 wild-type plasmids using Lipofectamine 3000 reagent (L3000-015, Invitrogen, Carlsbad, CA, USA) as per the manufacturer's instructions. At 4 days post-transfection, retrovirus containing supernatant was harvested and used to infect HeLa cells in the presence of 4 μg/ml polybrene. Stable cell lines were selected using 1 μg/ml puromycin selection. Other stable cell lines mentioned in this study were also prepared using a similar protocol.
Recombinant DNA constructs
The RNF167Δ1-35 construct was generated by cloning the coding sequence of RNF167 obtained by PCR amplification using primers FP-5′-GAAACTCGAGATGGACTTTGCAGACCTTCCAGC-3′ and RP-5′-AAAGGGCCCGGACCAGGATAACAGGGGAAG-3′, followed by in-frame cloning into XhoI-ApaI sites of pEYFP-N1 (Clontech, Palo Alto, CA, USA). Wild-type RNF167 and all the variants were cloned into pMYs-IRES-Puro (Cell Biolabs, San Diego, CA, USA) using the Infusion cloning kit (Takara Bio, Otsu, Japan) with primers FP-5′-GGTGGTACGGGAATTATGCACCCTGCAGCCTTC-3′ and RP-5′-ATTTACGTAGCGGCCTTAGACCAGGATAACAGGG-3′.
For RNF167Δ1-35, the forward primer was 5′-GGTGGTACGGGAATTATGGACTTTGCAGACCTTC-3′ with the same reverse primer as mentioned above. All primers used in this study were synthesized and bought from Sigma Aldrich (St Louis, MO, USA).
The following plasmids were obtained from Addgene: pCAG-VSVG (deposited by Arthur Nienhuis and Patrick Salmon, Addgene # 35616) and pUMVC (deposited by Bob Weinberg, plasmid # 8449) (Stewart et al., 2003). A mammalian expression plasmid encoding Dynamitin-GFP (Choy et al., 2018) was a gift from Dr Roberto Botelho (Ryerson University, Canada). Plasmids encoding C-terminal GFP fusion of RNF167 C268R, RNF167 R277L, RNF167 L329P and RNF167ΔPA (van Dijk et al., 2014) were generously provided by Dr Ruth H. Palmer (University of Gothenburg, Sweden).
Amino acid substitutions in RNF167 were made using site-directed mutagenesis by Prime Star GXL-DNA polymerase (R050A, Takara Bio, Otsu, Japan) according to the manufacturer's protocol with the following primers: RNF167 K97N, 5′-GCAACTTTGACCTCAATGTCCTAAATGCCCAG-3′ and 5′-CTGGGCATTTAGGACATTGAGGTCAAAGTTGC-3′; RNF167 R200Q, 5′-GTTGTATCCAGCACCAGAAACGGCTCCAGCG-3′ and 5′-CGCTGGAGCCGTTTCTGGTGCTGGATACAAC-5′; RNF167 C233S, 5′-GATGTCTGTGCCATTAGCCTGGATGAATATGAG-3′ and 5′-CTCATATTCATCCAGGCTAATGGCACAGACATC-3′.
Antibodies and reagents
The following antibodies were used in this study: rabbit anti-LAMP1 (1:500; ab24170, Abcam, Cambridge, MA, USA), mouse anti-GFP (1:1000; 632375, Takara Bio, Otsu, Japan), mouse anti-FLAG (1:1000; F3165, Sigma Aldrich, St Louis, MO, USA), mouse anti-LAMP1 (1:50; H4A3, Developmental Studies Hybridoma Bank, Iowa City, IA, USA), mouse anti-TOM20 (1:300; sc-17764, Santa Cruz Biotechnology, Santa Cruz, CA, USA), rabbit anti-GFP (1:1000; A-6455, Thermo Fisher Scientific, Rockford, IL, USA), rabbit anti-EEA1 (1:100; 2411S, Cell Signaling Technology, Beverly, MA, USA), mouse anti-RNF167 (1:1000; sc-515405, Santa Cruz Biotechnology), mouse anti-GAPDH (1:2000; 10-10011, Abgenex, Odisha, India), rabbit anti-mouse IgG (H+L) secondary antibody, HRP (1:1000; 61-6520, Thermo Fisher Scientific), Cy5-AffiniPure donkey anti-mouse IgG (715-175-150, Jackson Immuno Research Laboratories, West Grove, PA, USA), goat anti-rabbit IgG (H+L) Alexa Fluor 568 (A-11036), goat anti-mouse IgG (H+L) Alexa Fluor 488 (A-11001), goat anti-mouse IgG (H+L) Alexa Fluor 568 (A-11031). All the Alexa fluorophore-conjugated antibodies were used at a dilution of 1:1000 and purchased from Thermo Fisher Scientific.
Ionomycin (I3909), DMSO (D8418), Streptolysin-O (SAE0089), propidium iodide (P4170), DAPI (D8417), Polybrene (H9268) and puromycin (P8833) were from Sigma (St Louis, MO, USA), LysoTracker Red (L7528), and Prolong Gold antifade reagent (P36934) were from Thermo Fisher Scientific. Ciliobrevin D (250401) was from Calbiochem (San Diego, CA, USA) and KOD polymerase (71661) from Novagen (Madison, WI, USA). All chemicals, unless otherwise specified, were obtained from Sigma Aldrich.
HeLa cells were transfected with 1 μM each of control siRNA, siRNF167_1, siRNF167_2 separately using RNAimax (13778-075, Thermo Fisher Scientific) in reduced serum-containing medium (Opti-MEM, 5185-034, Gibco-Thermo Fisher Scientific, Waltham, MA, USA). The complex was removed after 8-10 h of transfection and cells were allowed to grow in complete medium. siRNA oligos were purchased from Eurogentec (Seraing, Belgium) along with control siRNA (SR-CL000-005). The sequences of siRNAs were as follows: RNF167_1, 5′-AAG CAG AGG GAC UGG GUC UUU-3′; RNF167_2, 5′-GGG ACU GGG UCU UCA CUU CUU-3′. Cells were either treated with acetate Ringer's solution (10 mM glucose, 70 mM sodium acetate, 5 mM KCl, 80 mM NaCl, 2 mM NaH2PO4, 2 mM CaCl2, 1 mM MgCl2 and 10 mM HEPES, final pH 6.9) for 2 h (Marwaha et al., 2017) or harvested after 24 h of transfection for immunostaining and quantitative real-time PCR (qRT-PCR), respectively.
Immunofluorescence and confocal microscopy
HeLa cells were transfected with RNF167 constructs as mentioned in the figure legends. Briefly, cells were transfected at 18-20 h of plating (∼60% confluency) using Lipofectamine 3000. Post transfection (24 h), cells were treated with acetate Ringer's solution (pH 6.9) for 2 h. Cells were fixed in 4% PFA in PHEM buffer (60 mM Pipes, 10 mM EGTA, 25 mM HEPES and 2 mM MgCl2, final pH 6.8). Immunostaining was performed as previously described (Marwaha et al., 2017). Briefly, cells were permeabilized in blocking solution (0.2% saponin and 5% FBS in PHEM buffer) for 30 min. Cells were washed in 1× PBS, followed by 1 h incubation with the primary antibody in staining buffer (0.2% saponin in PHEM buffer). These cells were further incubated with a secondary antibody in the staining buffer for 30 min and washed in 1× PBS to remove unbound antibodies. Nuclear staining was performed using 10 μg/ml DAPI for 1 min, then cells were washed in 1× PBS and mounted in ProLong Gold antifade reagent. All incubations were performed at room temperature unless otherwise stated.
To detect the luminal part of LAMP1 in the plasma membrane, cells seeded on glass coverslips were treated with acetate Ringer's solution (pH 6.9) for 2 h. These cells were treated with either DMSO or 5 μM ionomycin for 10 min and then immediately transferred onto ice. Immunostaining was performed as previously described (Jaiswal et al., 2014). Briefly, cells were incubated with mouse anti-LAMP1 (H4A3) for 30 min at 4°C, followed by fixation using 4% PFA in 1× PBS for 10 min and incubation with secondary antibody for 1 h. Next, cells were stained with 10 μg/ml DAPI and mounted in ProLong Gold antifade reagent.
To detect endogenous levels of TOM20, cells were subjected to permeabilization using 50 µg/ml digitonin in 1× PBS for 8 min followed by 1 h blocking with 3% BSA in 1× PBS. The primary and secondary antibodies were diluted in blocking buffer and incubated for 1 h each at room temperature. DAPI staining and mounting were performed as mentioned in the previous paragraph. For EEA1 staining, cells were permeabilized using 0.1% Triton X-100 in 1× PBS for 10 min followed by 1 h incubation in blocking buffer (0.1% Triton X-100 and 4% goat serum in 1× PBS). Both primary and secondary antibodies were diluted in 1×PBS containing 0.02% Triton X-100 and 1% goat serum and incubated overnight at 4°C and 1 h at room temperature, respectively. Cells were washed three times, stained for DAPI and mounted as mentioned previously. Single-plane confocal images were acquired using a confocal laser scanning microscope (Zeiss LSM 880, Germany) equipped with a 63× oil immersion objective.
For live-cell imaging, cells were transfected with the indicated constructs on glass-bottom dishes (MatTek Corporation, P35G-1.0-20-C). Post transfection (24 h), the imaging dish was loaded into a sealed live-cell imaging chamber (37°C and 5% CO2) for imaging. Single-plane confocal images were acquired using a confocal laser scanning microscope (Leica TCS SP5 II, Wetzlar, Germany) equipped with a 63×1.4 NA oil immersion objective. For quantification, images were imported into ImageJ (National Institutes of Health, Bethesda, MD, USA) software.
Quantification of lysosomal distribution and colocalization analysis
Lysosomal distribution was calculated in terms of the perinuclear index (PNI) (Marwaha et al., 2017). Briefly, average LAMP1 intensities were calculated for the whole cell (Itotal), within 5 µm distance of the nuclear surface (Iperinuclear) and more than 10 µm from the nuclear surface (Iperipheral). These intensities were normalized as I<5=(Iperinuclear/Itotal)−100 and I>10=(Iperipheral/Itotal)−100 and the PNI was calculated as the difference between I<5 and I>10 [PNI=(I<5−I>10)×100].
Line profiling was performed as described earlier (Willett et al., 2017). The average intensity of LAMP1-positive vesicles was calculated from three independent lines drawn from the nuclear surface. For all three lines, intensities were grouped into three categories (<5 µm, 5-15 µm and >15 µm) and the fractional intensities of each of these regions was calculated using the total intensity of that particular line. These fractional values were plotted against the distance from the nuclear surface.
Pearson's correlation coefficient was determined using the JACoP plugin of ImageJ (National Institutes of Health, Bethesda, MD, USA) software. The coefficient was calculated using single-plane images with threshold adjusted to 17-255 for LAMP1 and 20-255 for RNF167 (default setting 0-255).
Plasma membrane repair assays
For immunofluorescence, HeLa cells stably expressing empty vector, wild-type RNF167 and variants were first treated with acetate Ringer's solution (pH 6.9) for 2 h. Cells were washed three times with ice-cold HBSS, followed by 200 ng/ml SLO treatment for 5 min on ice. SLO-containing buffer was replaced with HBSS or HBSS containing 4 mM CaCl2 at 37°C to allow PMR and incubated further for 10 min at 37°C. Next, nuclear staining was performed using 50 μg/ml propidium iodide (PI) for 1 min followed by three washes in HBSS at 37°C. Cells were fixed in 4% PFA for 10 min. These cells were further washed, incubated with 10 μg/ml DAPI and mounted in ProLong Gold antifade reagent. A minimum of 100 cells from each field was used for calculating the percentage of PI-positive cells in each group.
For FACS analysis, HeLa cells stably expressing vector, wild-type RNF167 and other variants were treated as mentioned above. After resealing, cells were trypsinized immediately. Cell pellets were washed with flow cytometer buffer (1% FBS and 2 mM EDTA in PBS) and resuspended in 250 μl flow cytometer buffer. Before analysis on a BD accuriC6 system (Beckman Coulter, Fullerton, CA, USA), cells were also stained with 50 μg/ml PI. At least 10,000 cells were used for analysis. The intensity of PI-positive cells was represented as a histogram and the percentage of PMR calculated as described earlier (Encarnação et al., 2016).
β-hexosaminidase assay was performed as described earlier (Rodríguez et al., 1997). In brief, HeLa cells stably expressing empty vector, RNF167-a, RNF167-a-K97N or RNF167-b were treated with acetate Ringer's solution (pH 6.9) for 2 h followed by DMSO or 5 μM ionomycin for 10 min at 37°C. Cells were immediately transferred to ice and supernatant containing released β-hexosaminidase was collected and centrifuged to remove cellular debris. In parallel, cells were lysed using 1% TritonX-100 in 1× PBS and the cell lysate clarified by centrifugation at 16,000 g for 10 min at 4°C. The supernatant containing released β-hexosaminidase and appropriately diluted cell lysates were incubated with 50 μl of 6 mM 4-methylumbelliferyl-N-acetyl-β-d-glucosaminide (M2133, Sigma) in citrate-phosphate buffer (40 mM sodium citrate, 88 mM Na2PO4, pH 4.5) separately and incubated for 15 min at 37°C. Reactions were stopped by adding a 50 μl stop solution (2 M Na2CO3 and 1.1 mM glycine, pH 10.2). The β-hexosaminidase activity was measured using the SpectraMax M5 multidetection reader (Molecular Devices, San Jose, CA, USA) at excitation/emission wavelengths of 365/450 nm, respectively. Total β-hexosaminidase activity is the sum of released activity in the supernatant and intracellular activity in the cell lysate. The percentage β-hexosaminidase released was calculated as β-hexosaminidase in the supernatant relative to the total activity.
RNA isolation and relative qRT-PCR
Total RNA was extracted using the RNeasy Mini Kit (74104, Qiagen, Valencia, CA, USA) following the manufacturer's recommendations. RNA (1 μg) was reverse transcribed using oligo-DT and the SuperScript III First-Strand Synthesis System (18080051, Thermo Fisher Scientific). Relative qRT-PCR was performed in a total volume of 25 µl with 12.5 µl Syber Green master mix (1725121, BioRad Laboratories, Hercules, CA, USA), 1 µl gene-specific primer mix, 5 µl cDNA and 6.5 µl sterile water. GAPDH was used to normalize the expression of RNF167 in all the samples used. The normalized values were expressed as relative to control siRNA (considering the mRNA level of RNF167 in control siRNA cells as 100%). Successful amplification resulted in an amplicon size of 178 bp in the coding region of RNF167 (from 542-719 base pairs of NM_015528.3). The following primers were used: RNF167, FP-5′-CTGTGGTTGTGGCCGCTGTGCTGTGG-3′ and RP-5′-CTGCAGGCATTGTCTGGGTGAGCCTCC-3′; GAPDH, FP-5′-ATGGGGAAGGTGAAGGTCGGAGTCAAC-3′ and RP-5′-GATCACAAGCTTCCCGTTCTCAGCCTTGAC-3′.
cDNAs prepared from different cell lines (ZR-75, COLO 205, SUM159, HCC1937, MIA PaCa, HeLa and MCF 10A) were subjected to two rounds of PCR (nested PCR) with the following primers: FP-5′-GTTTGAAGGTCTCGCGAGATCG-3′ and RP-5′-CCAGGATTTACGGGTCTTCCGAC-3′. The purified product from the first set of PCR was used as a template for the second step of PCR with the following primers: FP-5′-TCCCACCCAGCTCCACTAAACG-3′ and RP-5′-GCTGGAAGGTCTGCAAAGTCC-3′. Amplicons of expected sizes were purified and confirmed by sequencing.
Cells were washed with ice-cold 1× PBS and harvested at 850×g in ice-cold 1× PBS. Pellets were lysed in lysis buffer containing 50 mM Tris-Cl pH 7.5, 1 mM EDTA pH 8, 150 mM sodium chloride, 1% TritonX-100, 1 mM phenylmethylsulfonyl fluoride (PMSF, P7626, Sigma), protease inhibitor cocktail (P8340, Sigma) and phosphatase inhibitors 2 and 3 (P5726 and P044, Sigma) for 30 min. Samples were centrifuged at 13,500×g for 10 min and the supernatants heated at 95°C for 10 min in Laemmli buffer. Samples were transferred to the PVDF membrane (IPVH00010, Merck, Darmstadt, Germany) and immunoblotting performed using indicated antibodies. Membranes were visualized using chemiluminescent HRP substrate (WBKLS0500, Merck) and images acquired using the ChemiDoc XRS+ System (Bio-Rad Laboratories, Hercules, CA, USA). To detect the expression levels of RNF167-a, its tumor-associated variants and RNF167-b, cells were pretreated with proteasomal and lysosomal inhibitors (25 μM MG132 and 0.2 μM Bafilomycin) for 4 h before harvesting.
Multiple sequence alignment
Amino acid sequences of different isoforms of RNF167 were retrieved from the NCBI database with the following IDs: NP_056343.1, NP_001307294.1, NP_001357233.1, NP_001357236.1, NP_001357237.1, NP_001357242.1. These sequences were aligned using Kalign (Lassmann and Sonnhammer, 2005) with gap open penalty of 11, gap extension penalty of 0.85 and terminal gap penalty of 0.45. Pictorial representations of the isoforms are based on the multiple sequence alignment.
Data were collected from at least three independent experiments and are shown as mean±s.d. P-values were calculated using the two-tailed unpaired Student's t-test (for two parameters) or one-way ANOVA or two-way ANOVA (for comparing multiple groups) followed by Tukey's test, as mentioned in the figure legends, with significance levels set at α=0.05 (ns, not significant, *P<0.05, **P<0.01, ***P<0.001 and ****P<0.0001). All statistical analyses were performed in Graph Pad Prism 8 (Version 8.2.0).
The authors gratefully acknowledge the following people for their generous gift of reagents: Dr Roberto Botelho, Ryerson University, Canada (Dynamitin-GFP); Dr Ruth H. Palmer, University of Gothenburg, Sweden (C-terminal GFP-tagged RNF167-C268R, RNF167-R277L, RNF167-L329P and RNF167ΔPA). The authors also acknowledge technical help from Keerthi S. Nair and Fabi Rasheed.
Conceptualization: S.V.N., N.D.N., S.M.S.; Methodology: S.V.N., S.M.S.; Validation: S.V.N., N.D.N., S.M.S.; Formal analysis: S.V.N., N.D.N., R.P.N., S.M.S.; Investigation: S.V.N., N.D.N., R.P.N.; Resources: R.K., S.M.S.; Writing - original draft: S.V.N., S.M.S.; Writing - review & editing: S.V.N., R.K., S.M.S.; Visualization: S.V.N., S.M.S.; Supervision: S.M.S.; Project administration: S.M.S.; Funding acquisition: S.M.S.
S.V.N. and R.P.N. acknowledge financial support from the Department of Science and Technology, Ministry of Science and Technology, India INSPIRE programme for PhD and undergraduate fellowships, respectively. N.D.N. acknowledges financial support from the Indian Institute of Science Education and Research Thiruvananthapuram (IISER TVM). This work was partly supported by funding from the Department of Science and Technology Science and Engineering Research Board (DST-SERB; grant EMR/2016/008048), the Department of Biotechnology, Ministry of Science and Technology, India (grant BT/PR21325/BRB/10/1554/2016) and the Indian Institute of Science Education and Research Thiruvananthapuram (IISER TVM) awarded to S.M.S.
Peer review history
The peer review history is available online at https://jcs.biologists.org/lookup/doi/10.1242/jcs.239335.reviewer-comments.pdf
The authors declare no competing or financial interests.