E XP ER I ME NTAL C E LL RE S E ARCH
32 0 (2 014 ) 219 – 23 2
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/yexcr
Research Article
The ubiquitin ligase RNF126 regulates the retrograde sorting of the cation-independent mannose 6-phosphate receptor Christopher J. Smitha, C. Jane McGladea,b,n a
Department of Medical Biophysics, University of Toronto, Canada The Arthur and Sonia Labatt Brain Tumour Research Centre, The Hospital for Sick Children, Peter Gilgan Centre for Research and Learning, 686 Bay Street 17-9706, Toronto, Ontario, Canada M5G 0A4 b
article information
abstract
Article Chronology:
The ubiquitin proteasome system is central to the regulation of a number of intracellular sorting
Received 23 May 2013
pathways in mammalian cells including quality control at the endoplasmic reticulum and the
Received in revised form
internalization and endosomal sorting of cell surface receptors. Here we describe that RNF126, an
13 November 2013
E3 ubiquitin ligase, is involved in the sorting of the cation-independent mannose 6-phosphate
Accepted 15 November 2013
receptor (CI-MPR). In cells transiently depleted of RNF126, the CI-MPR is dispersed into Rab4
Available online 23 November 2013
positive endosomes and the efficiency of retrograde sorting is delayed. Furthermore, the stable
Keywords: RNF126 E3 ubiquitin ligase CI-MPR Retrograde sorting
knockdown of RNF126 leads to the lysosomal degradation of CI-MPR and missorting of cathepsin D. RNF126 specifically regulates the sorting of the CI-MPR as other cargo that follow the retrograde sorting route including the cholera toxin, furin and TGN38 are unaffected in the absence of RNF126. Lastly we show that the RING finger domain of RNF126 is required to rescue the decrease in CI-MPR levels, suggesting that the ubiquitin ligase activity of RNF126 is required for CI-MPR sorting. Together, our data indicate that the ubiquitin ligase RNF126 has a role in the retrograde sorting of the CI-MPR & 2013 Elsevier Inc. All rights reserved.
Introduction A subset of cargo within the endocytic system undergoes retrograde transport from the endosome to the TGN. Examples include TGN38/46, a protein of unknown function, the endoprotease furin, as well as the mannose-6-phosphate receptors. In addition, pathogens such as cholera toxin and shiga toxin hijack the
retrograde transport route in order to gain entry into the cell. Several distinct retrograde transport routes from both the early and late endosome have been identified. The different retrograde pathways can be characterized by adaptors that package cargo at the donor membrane, the use of clathrin or non-clathrin coated vesicles, as well as the SNAREs and tethers employed for fusion at the Golgi [1].
Abbreviations: CI-MPR, cation-independent mannose 6-phosphate receptor; EGFR, epidermal growth factor receptor; GGA, Golgilocalized ϒ ear containing, ARF binding protein; HEK, human embryonic kidney; PACS, phosphofurin acidic cluster protein-1; siRNA, small interfering RNA; TGN, trans-Golgi network; TfR, transferrin receptor n Corresponding author at: The Arthur and Sonia Labatt Brain Tumor Research Centre, The Hospital for Sick Children, Peter Gilgan Centre for Research and Learning, 686 Bay Street 17-9706, Toronto, Ontario, Canada M5G 0A4. Fax: þ1 416 813 8456. E-mail address: [email protected] (C.J. McGlade).
0014-4827/$ - see front matter & 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.yexcr.2013.11.013
220
E XP E RI ME N TA L CE L L R ES E ARC H
The cation-independent mannose 6-phosphate receptor (CI-MPR) and the cation-dependent mannose 6-phosphate receptor (CD-MPR) deliver lysosomal hydrolases to the lysosome. MPRs associate with newly synthesized lysosomal hydrolases modified with a Man-6-P moiety at the transGolgi network (TGN) and are subsequently sorted towards an endosomal compartment. Once at the endosome, MPRs release their cargo and are returned back to the Golgi for successive rounds of transport [2]. The CI-MPR has a complex trafficking itinerary and can be found at the Golgi, plasma membrane, early and late endosomes. The cytoplasmic tail of MPRs contains multiple sorting motifs that are recognized by a number of cargo selection proteins recruited to MPR containing compartments [3]. Sorting at the Golgi is dependent on the AP-1 and the Golgi-localized ϒ ear containing, ARF binding protein (GGA) which package cargo into carriers destined for the endosome [4–6]. Within the endocytic system, the MPRs associate with a number of adaptor proteins that regulate sorting of the CI-MPR back to the Golgi. At the early endosome, the clathrin adaptor proteins AP-1 [7], Phosphofurin Acidic Cluster Sorting Protein 1 (PACS-1) [8–10] and epsinR [11] have been shown to regulate CI-MPR sorting. In addition, the retromer complex has been shown to package the CI-MPR into transport carriers destined for the TGN [12–15]. From the late endosome, Rab9 and its effector TIP47 regulate the retrograde transport of CI-MPR [16,17]. The individual depletion of any of the retrograde machineries disrupt CI-MPR sorting suggesting that all are required for successful transport. Ubiquitin is a 76aa protein that can be covalently attached to substrate lysine residues. Protein ubiquitination involves the sequential action of an E1 ubiquitin activating enzyme, an E2 ubiquitin conjugating enzyme as well as an E3 ubiquitin ligase. The conjugation of ubiquitin to transmembrane cargo regulates the internalization and sorting of a number of cell surface receptors. In addition, ubiquitination of the endosomal sorting machinery has also been shown to regulate receptor downregulation. Until very recently, a role for the ubiquitin system in the retrograde sorting route had not been described [18]. Here we report that the ubiquitin ligase RNF126, regulates the retrograde sorting of CI-MPR. Depletion of RNF126 causes the CI-MPR to be retained in endosomes and an increase in the lysosomal degradation of the receptor. In addition, we show that the efficient sorting of the CI-MPR is dependent on the ubiquitin ligase activity of RNF126. Our studies provide evidence of a role for ubiquitin in the regulation of CI-MPR sorting.
32 0 (2 014) 219 – 23 2
Materials and methods cDNA constructs CD8-CIMPR was a gift from Matthew Seaman. CD25-TGN38 and CD25-Furin were gifts from Sergio Grinstein. Full length murine RNF126 and RNF126 C231A were cloned into HA-pcDNA 3.1.
Cell culture and transfection HeLa and 293 T cells were maintained in DMEM supplemented with 10% FBS. Stable cell lines were generated by transfecting HeLa cells with a non-silencing shRNA or with an shRNA with a sense strand targeting RNF126: 5′-CACTCAAACCCTATGGACT-3′ or with 5′-TGTATGTAGGAAGAGCTTA-3′ which targets Rabring7 (Open Biosystems). 24 h post transfection, cells were treated with 1 μg/ml puromycin and single colonies were selected. For cDNA expression, cells were seeded on 10 cm2 dishes and transfected using lipofectamine-2000 (Invitrogen). Cells were fixed 24 h after transfection, except in rescue experiments where cells were fixed 48 h after post-transfection. For siRNA treatment, cells were seeded on a 6 well dish and transfected with 40 pmol siRNA targeting RNF126: 5′-CCGGATTATATCTGTCCAAGA-3′ (Qiagen) or an All-Stars negative control siRNA (Qiagen) using lipofectamine2000 (Invitrogen). Experiments were performed 48 h post transfection. When cDNA and siRNA were co-transfected, cells were treated with 40 pmol siRNA 48 h before re-transfecting with both cDNA and RNA (1ug of DNA:30 pmol RNA) for an addition 24 h. Proteins were collected from media using trichloroacetic acid (TCA). Cells were grown in DMEM without serum for 24 h and the media was collected. TCA was added to a final concentration of 25% (v/v). Samples were incubated for 30 min at 4 1C and centrifuged for 5 min at 14000RPM. The pellet was washed 2 in cold acetone and re-suspended in sample buffer.
Antibodies and reagents The following commercial antibodies were used: Cathepsin-D (Millipore, Rabbit) CI-MPR (Thermo Scientific, mouse); GFP (Millipore, chicken); GGA3 (BD Biosciences, mouse); Giantin (Covance, rabbit); GM130 (BD Biosciences, mouse); HA (Novus Biologicals, Rabbit); TnfR (Millipore, rabbit); β-tubulin (Sigma, mouse); Rab7 (Cell Signaling, rabbit); Rab9 (Cell Signaling, rabbit). The following reagents were used: A555-CTxB (Molecular Probes), Leupeptin
Fig. 1 – Depletion of RNF126 causes the lysosomal degradation of the CI-MPR. (A) Depletion of RNF126 reduces the levels of the CIMPR. Non-silencing, RNF126 or Rabring7 knockdown lines were fixed and stained with anti-CI-MPR or anti-transferrin. Scale bar, 10 μm. The graph to the right shows the average volumetric fluorescence intensity above threshold measured on individual cells (n¼ 20) of 3 replicate experiments. Results are expressed as mean7SEM. (B) Non-silencing, RNF126 or Rabring7 knockdown lines were lysed in RIPA buffer, run on an SDS-PAGE gel under non-reducing conditions and analyzed with an anti-CIMPR antibody. A separate SDS-PAGE gel was run under reducing conditions and analyzed with anti β-tubulin. The graph to the right shows the mean CI-MPR immunoreactivity normalized to the non-silencing lines of 3 independent experiments. Error bars represent standard error of the mean. (C) The reduction in CI-MPR levels in the RNF126 knockdown lines can be rescued with leupeptin. Non-silencing and RNF126 knockdown lines were incubated with 4 mM leupeptin or a vehicle control for 24 h, lysed in RIPA, and analyzed as described in part B, n¼ 3. (D) The depletion of RNF126 leads to the secretion of immature cathepsin-D. Non-silencing and RNF126 knockdown lines were washed with PBS and incubated for 24 h in serum free media. The protein content from the media was precipitated using trichloroacetic acid. Lysate and media samples were analyzed with an antibody against cathepsin-D. Precursor (p), intermediate (i) and mature (m) forms are indicated. The lysate samples were run on separate SDS-PAGE gels and analyzed with an antibody against RNF126 or tubulin.
EX PE R IM EN TA L C ELL RE S EA RC H
(Sigma). RNF126 and Rabring7 antibodies were produced by immunizing rabbits with peptides corresponding to the carboxyterminal regions of each protein, RNF126: CSPSNENATSNS, Rabring7: CNRFSNDSQLHDRWTF. The following secondary antibodies were
3 20 (2014) 219 –2 32
221
used: HRP-linked anti-mouse IgG (GE Healthcare, sheep), Cy3conjugated donkey anti-mouse IgG, A488-conjugated donkey antirabbit IgG and Cy5-conjugated donkey anti-mouse IgG (Jackson ImmunoResearch).
222
E XP E RI ME N TA L CE L L R ES E ARC H
Immunocytochemistry Cells were fixed in a 2% PFA 30 mM sucrose solution for 30 min at room temperature and subsequently incubated in 100 mM glycine for 10 min. Cells were permeabilized with 0.1% TX-100 and blocked in a 3% normal donkey serum (Jackson ImmunoResearch). Coverslips were inverted onto primary antibody and incubated for 30 min at 37 1C in a humidified chamber, washed 3 10 min in 0.05% TX-100 and inverted onto secondary antibody for an addition 30 min at 37 1C. Coverslips were washed for an addition 3 10 min in PBS with 0.05% TX-100 before mounting in fluorescent mounting medium (Dako). Images were acquired using a Zeiss Axiovert 200 inverted fluorescent microscope using a 60 /1.35NA objective equipped with a back-thinned EM charged-coupled device camera (Hamamatsu, C9100-13) and a spinning disk confocal scan head. Equipment was driven by Volocity (PerkinElmer) acquisition software. All images were acquired at room temperature. The images collected through the focal planes of the Z-axis were subsequently collapsed into one image using Volocity for final presentation. Linear adjustments for contrast and brightness were performed using the Volocity software.
32 0 (2 014) 219 – 23 2
Therefore, this measurement protocol creates 4 non-overlapping ROIs on each cell found within a field of view and also compensates for variation in volume between cells. The sum fluorescence intensity within each ring is shown. Graphpad was used for statistical analysis. Unpaired students t-tests are reported.
Uptake experiments For antibody uptake, Cells plated on coverslips were washed with PBS pre-chilled to 4 1C, transferred to a new 24 well dish and incubated with DMEM containing 0.2% BSA for 30 min at 4 1C. Next, coverslips were flipped onto a drop of primary antibody diluted in 0.2% BSA, incubated for 30 min at 4 1C and then washed PBS. Coverslips were transferred to a new 24 well dish and incubated in DMEM with 0.2% BSA pre-warmed to 37 1C for the indicated period of time, washed with PBS, and fixed in PFA. For A555-CTxB uptake, HeLa cells plated on coverslips were inverted onto a drop of DMEM containing 0.2% BSA and 4ug/ml A555-CTxB and incubated for 30 min at 4 1C, washed in PBS, and returned to 37 1C for the indicated period of time before fixation.
Immunoblotting Image analysis To measure the mean fluorescence intensity, the level of background fluorescence was determined by measuring 5 regions of interest (ROIs) devoid of any cells. Individual cells were selected and the Cy3 intensity per unit volume was determined using Volocity. For colocalization analysis, images were first deconvolved using the fast restorative function in Volocity. A point spread function was applied to each channel with the following parameters: Numerical aperture¼ 1.2, medium refractive index¼ 1.33 and emission wavelengths of 561 nm (Cy3) or 520 nm (GFP). A threshold value for each image was determined by measuring 5 ROIs devoid of any cells. Individual cells were selected and the Pearson's correlation above threshold was determined using the colocalization function in Volocity. Graphpad (prism) was used for statistical analysis. Unpaired students t-tests are reported. To measure the extent of dispersion, a measurement protocol was designed in Volocity that measured the total fluorescence intensity in eroded ROIs emanating from the outside of the cell inwards. Using the ‘find objects’ function in Volocity, individual cells within a field of view were selected. To find the entire volume of the cell, the ‘find objects’ function detected the fluorescence intensity above an automatically determined threshold. Objects with a volumeo100 μm3 were excluded and the noise from the detected objects was removed using a fine filter. The second ROI was determined by finding the volume of the entire cell using the same parameters but was subsequently eroded in volume by 2 iterations. Similarly, the third ROI was eroded in volume by 5 iterations (objects with a volume ofo10 μ m3 were excluded) and the fourth ROI eroded by 9 iterations. Next, the total fluorescence intensity contained within the identified ROIs were subtracted from each other to create nonoverlapping rings of intensity. ROI 2 was subtracted from ROI 1 to create an outer ring. ROI 3 was subtracted from ROI 2 to create a middle outer ring. ROI 4 was subtracted ROI 3 to create a middle inner ring while the inner ring corresponds to the fourth ROI.
Cells were lysed in RIPA buffer (150 mM NaCl, 20 mM Tris (Ph 7.5), 1% NP-40, 1% sodium deoxycholate, 0.1% SDS) and boiled in SDSLaemmli sample buffer. When immunoblotting for the CI-MPR, samples were prepared in non-reducing sample buffer. Proteins were separated by SDS-PAGE and transferred to PVDF membrane (Pall Corp).
Results RNF126 regulates the endosomal sorting of the CI-MPR The E3 ubiquitin ligases RNF126 and Rabring7 have been indicated in the endocytic sorting of the EGFR and other cell surface receptors [19–21]. To investigate additional roles for these E3 ubiquitin ligases in membrane traffic, we examined the steady state distribution of endosomal cargo that follows different trafficking itineraries to that of EGFR HeLa cell lines that stably express either a non-silencing shRNA, an shRNA that targets RNF126, or an shRNA targeting Rabring7 were generated [21]. The distribution of the transferrin receptor was comparable between non-silencing and both knockdown suggesting that the endocytosis and recycling of this receptor is unaffected (Fig. 1A). In addition, the distribution and intensity of CI-MPR staining was comparable between non-silencing and Rabring7 knockdown lines. However, in cell lines stably depleted of RNF126 we noticed a reduction in the total levels of the CI-MPR by both immunofluorescence and western blot (Fig. 1A and B). To test whether depletion of RNF126 was causing the CI-MPR to be degraded in the lysosome, we treated non-silencing and RNF126 knockdown lines with leupeptin, an inhibitor of lysosomal hydrolazes, or a vehicle control. Treatment of the RNF126 knockdown lines with leupeptin partially restored the levels of the CI-MPR to those observed in the non-silencing lines, suggesting that CI-MPR is degraded in the lysosome in cell lines depleted of RNF126 (Fig. 1C).
EX PE R IM EN TA L C ELL RE S EA RC H
170 IB: Tubulin
34
IB: RNF126
2.5 2.0 1.5 1.0 0.5 0.0 Control si
55
the treatment with an siRNA directed against RNF126 increased the total CI-MPR levels by western blot (Fig. 2A). We next immunostained control and RNF126 depleted cells for the CIMPR. In cells treated with RNF126 siRNA, the CI-MPR appeared in peripherally dispersed vesicles compared to the tightly packed perinuclear staining pattern observed in the control siRNA treated cells (Fig. 2B). To quantify the degree of dispersion in the staining pattern, we devised a protocol that measured the fluorescence in 4 non-overlapping regions of interest mapped over the volume of the cell. Using this protocol, we found that total CI-MPR staining intensity in the control siRNA treated cells decreased from the inner most region to the cell periphery. In cells depleted of RNF126, CI-MPR staining intensity remained relatively constant in all regions of interest, confirming that the distribution of the CI-MPR had shifted towards the cell periphery (Fig. 2C).
RNF126 si
IB: CI-MPR
fold change in CI-MPR immunoreactivity
RNF126 si
Control si
The CI-MPR serves to sort lysosomal hydrolases from the Golgi to the endocytic system. We next analyzed whether the depletion of RNF126 resulted in the missorting of the lysosomal aspartyl protease, cathepsin D. Cathepsin D is synthesized in a precursor form that subsequently matures into an intermediate and mature form upon reaching the acidic milieu of the endosomal and lysosomal compartments. Under normal conditions, cathepsin D is efficiently sorted towards the endocytic system and little is secreted from the cell [22]. We analyzed both intracellular and extracellular cathepsin D by immunoblot. Compared to control, the depletion of RNF126 caused an increase in the amount of secreted immature cathepsin D, suggesting that CI-MPR cargo are inefficiently sorted in these cells (Fig. 1D). To further assess the short-term effects of RNF126 knockdown on the CI-MPR, we used siRNA to deplete RNF126 from HeLa cells. In contrast to observations in the stable knockdown lines,
RNF126 si sum fluorescence intensity (10 )
4.0 P=0.006
CI-MPR
9
Control si
223
3 20 (2014) 219 –2 32
Control si RNF126 si
3.0
2.0
1.0
Outer ring
middle outer ring
middle inner ring
Inner ring
0.0
Fig. 2 – Transient knockdown of RNF126 leads to the dispersal of the CI-MPR into peripheral structures. (A) Transient depletion of RNF126 increases the total levels of the CI-MPR. HeLa cells were treated with control or an siRNA directed against RNF126. Cells were lysed 48 h post transfection, run on an SDS-PAGE gel under non-reducing conditions and analyzed with an anti-CI-MPR antibody. A separate gel was run under reducing conditions and analyzed by immunoblotting with antibodies against β-tubulin and RNF126. The graph to the right shows the mean CI-MPR immunoreactivity relative to control between 3 independent experiments. Bars represent standard error of the mean. (B) The depletion of RNF126 leads to the dispersal of the CI-MPR. HeLa cells treated with a control or siRNA directed against RNF126 were fixed 48 h post-transfection and stained with an anti-CI-MPR antibody. Scale bar, 10 μm. (C) CI-MPR staining in the cell periphery is increased in HeLa cells treated with RNF126 siRNA. Total fluorescence intensity in 4 regions of interest mapped over the volume of the cell (see materials and methods). 15 fields of view were quantified for each condition between 3 replicate experiments. Results are expressed as mean7SEM.
32 0 (2 014) 219 – 23 2
3.0
Control si RNF126 si
2.0
1.0
Outer ring
Inner ring
0.0 middle outer ring
RNF126 si
Giantin
Control si
sum fluorescence intensity (109)
E XP E RI ME N TA L CE L L R ES E ARC H
middle inner ring
224
Fig. 3 – The Golgi complex is mildly disrupted in RNF126 depleted cells. (A) The depletion of RNF126 leads to the dispersal of Giantin. HeLa cells treated with an control siRNA or an siRNA directed against RNF126 were fixed and stained for Giantin. Scale bar, 10μm. Images were analyzed as in Fig. 2C. Six fields of view were examined for each condition between 3 replicate experiments. Results are expressed as mean7SEM.
Disruptions of CI-MPR retrograde sorting machinery are sometimes accompanied by Golgi fragmentation. We therefore examined the Golgi in cells depleted of RNF126. Cells treated with control or RNF126 siRNA were stained with antibodies directed against Giantin. In cells depleted of RNF126, the Giantin staining appeared slightly more dispersed but remained on one side of the nucleus (Fig. 3). This phenotype is similar to that observed upon depletion of mVps26, SNX1, GCC88 or Arl5b [12,23–25] but is not as fragmented as in cells depleted of the Golgin GCC185 [26]. This suggests that the depletion of RNF126 has only a mild effect on Golgi structure.
CFP-Rab11 into cells treated with an siRNA directed against RNF126 and measured the colocalization between the CI-MPR and each Rab protein (Fig. 5A). In control siRNA treated cells, the degree of overlap between the CI-MPR and Rab4 or Rab5 were roughly equivalent, and greater than the colocalization observed between the CI-MPR and Rab11. In cells treated with the RNF126 siRNA, the overlap between the CI-MPR and Rab4 increased, while the colocalization between the CI-MPR with either Rab5 or Rab11 remained comparable to control (Fig. 5B). This suggests that the depletion of RNF126 causes the CI-MPR to accumulate in a Rab4 positive endocytic recycling compartment.
The depletion of RNF126 causes the redistribution of the CI-MPR into an endocytic recycling compartment
RNF126 regulates the retrograde sorting of the CI-MPR
We reasoned that the dispersed CI-MPR structures may be endosomal compartments. To determine the identity of the CIMPR containing structures, we co-stained control and RNF126 depleted cells for the CI-MPR and a number of endocytic markers. Control and RNF126 depleted cells were co-stained with antibodies against the CI-MPR and the transferrin receptor, a marker of the endocytic recycling compartment, or the late endosomal marker Rab7. We also co-stained with Rab9, an important regulator of CI-MPR retrograde sorting [16,27,28] (Fig. 4A). In cells depleted of RNF126, there was a significant increase in the degree of overlap between the CI-MPR and the transferrin receptor, suggesting that a higher proportion of CI-MPR is found in the endocytic recycling compartment under these conditions. Depleting RNF126 from cells did not alter the degree of overlap between the CI-MPR and Rab7 or Rab9 (Fig. 4C) suggesting that RNF126 does not play a role in Rab9 mediated CI-MPR sorting [29]. After endocytosis, the transferrin receptor is sorted through Rab4, Rab5 and Rab11 domains. These Rab proteins can be found on the same endosome, but do not significantly intermix over time [30]. To further define the location of the CI-MPR in cells depleted of RNF126, we transfected either CFP-Rab4, CFP-Rab5 or
Steady state distribution of the CI-MPR is achieved through balancing the amount of receptor delivered to the endosome versus the quantity of receptor removed by retrograde transport. To examine whether depletion of RNF126 affected the retrograde sorting route, we performed antibody uptake experiments. HeLa cells were treated with control or RNF126 siRNA and transfected with a CD8-CIMPR chimera [12]. Cells were chilled, incubated with anti-CD8, and either immediately fixed or warmed to 37 1C to allow the internalization of the antibody bound cargo for either 8 or 24 min before fixation (Fig. 6A). In cells treated with the control siRNA, the majority of the CD8-CIMPR had been sorted towards the center of the cell at the 24min time point. However, in cells depleted of RNF126, the sorting of CD8-CIMPR towards the inner region of the cell was delayed. Significantly less receptor had reached the central most region of the cell compared to control (Fig. 6B). Furthermore, the degree of overlap between the CI-MPR and the TfnR was significantly increased in cells treated with RNF126 siRNA, compared to control siRNA, at the 24 min time point (Fig. 6C). Taken together, these results suggest that the CI-MPR is retained in a transferrin receptor positive compartment in cells depleted of RNF126, which is consistent with a role for RNF126 in regulating the retrograde sorting of the CI-MPR.
EX PE R IM EN TA L C ELL RE S EA RC H
CI-MPR
225
merge
RNF126 si
Control si
TfR
3 20 (2014) 219 –2 32
CI-MPR
merge
Rab9
CI-MPR
merge
RNF126 si
Control si
RNF126 si
Control si
Rab7
P
32 0 (2 014 ) 219 – 23 2
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/yexcr
Research Article
The ubiquitin ligase RNF126 regulates the retrograde sorting of the cation-independent mannose 6-phosphate receptor Christopher J. Smitha, C. Jane McGladea,b,n a
Department of Medical Biophysics, University of Toronto, Canada The Arthur and Sonia Labatt Brain Tumour Research Centre, The Hospital for Sick Children, Peter Gilgan Centre for Research and Learning, 686 Bay Street 17-9706, Toronto, Ontario, Canada M5G 0A4 b
article information
abstract
Article Chronology:
The ubiquitin proteasome system is central to the regulation of a number of intracellular sorting
Received 23 May 2013
pathways in mammalian cells including quality control at the endoplasmic reticulum and the
Received in revised form
internalization and endosomal sorting of cell surface receptors. Here we describe that RNF126, an
13 November 2013
E3 ubiquitin ligase, is involved in the sorting of the cation-independent mannose 6-phosphate
Accepted 15 November 2013
receptor (CI-MPR). In cells transiently depleted of RNF126, the CI-MPR is dispersed into Rab4
Available online 23 November 2013
positive endosomes and the efficiency of retrograde sorting is delayed. Furthermore, the stable
Keywords: RNF126 E3 ubiquitin ligase CI-MPR Retrograde sorting
knockdown of RNF126 leads to the lysosomal degradation of CI-MPR and missorting of cathepsin D. RNF126 specifically regulates the sorting of the CI-MPR as other cargo that follow the retrograde sorting route including the cholera toxin, furin and TGN38 are unaffected in the absence of RNF126. Lastly we show that the RING finger domain of RNF126 is required to rescue the decrease in CI-MPR levels, suggesting that the ubiquitin ligase activity of RNF126 is required for CI-MPR sorting. Together, our data indicate that the ubiquitin ligase RNF126 has a role in the retrograde sorting of the CI-MPR & 2013 Elsevier Inc. All rights reserved.
Introduction A subset of cargo within the endocytic system undergoes retrograde transport from the endosome to the TGN. Examples include TGN38/46, a protein of unknown function, the endoprotease furin, as well as the mannose-6-phosphate receptors. In addition, pathogens such as cholera toxin and shiga toxin hijack the
retrograde transport route in order to gain entry into the cell. Several distinct retrograde transport routes from both the early and late endosome have been identified. The different retrograde pathways can be characterized by adaptors that package cargo at the donor membrane, the use of clathrin or non-clathrin coated vesicles, as well as the SNAREs and tethers employed for fusion at the Golgi [1].
Abbreviations: CI-MPR, cation-independent mannose 6-phosphate receptor; EGFR, epidermal growth factor receptor; GGA, Golgilocalized ϒ ear containing, ARF binding protein; HEK, human embryonic kidney; PACS, phosphofurin acidic cluster protein-1; siRNA, small interfering RNA; TGN, trans-Golgi network; TfR, transferrin receptor n Corresponding author at: The Arthur and Sonia Labatt Brain Tumor Research Centre, The Hospital for Sick Children, Peter Gilgan Centre for Research and Learning, 686 Bay Street 17-9706, Toronto, Ontario, Canada M5G 0A4. Fax: þ1 416 813 8456. E-mail address: [email protected] (C.J. McGlade).
0014-4827/$ - see front matter & 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.yexcr.2013.11.013
220
E XP E RI ME N TA L CE L L R ES E ARC H
The cation-independent mannose 6-phosphate receptor (CI-MPR) and the cation-dependent mannose 6-phosphate receptor (CD-MPR) deliver lysosomal hydrolases to the lysosome. MPRs associate with newly synthesized lysosomal hydrolases modified with a Man-6-P moiety at the transGolgi network (TGN) and are subsequently sorted towards an endosomal compartment. Once at the endosome, MPRs release their cargo and are returned back to the Golgi for successive rounds of transport [2]. The CI-MPR has a complex trafficking itinerary and can be found at the Golgi, plasma membrane, early and late endosomes. The cytoplasmic tail of MPRs contains multiple sorting motifs that are recognized by a number of cargo selection proteins recruited to MPR containing compartments [3]. Sorting at the Golgi is dependent on the AP-1 and the Golgi-localized ϒ ear containing, ARF binding protein (GGA) which package cargo into carriers destined for the endosome [4–6]. Within the endocytic system, the MPRs associate with a number of adaptor proteins that regulate sorting of the CI-MPR back to the Golgi. At the early endosome, the clathrin adaptor proteins AP-1 [7], Phosphofurin Acidic Cluster Sorting Protein 1 (PACS-1) [8–10] and epsinR [11] have been shown to regulate CI-MPR sorting. In addition, the retromer complex has been shown to package the CI-MPR into transport carriers destined for the TGN [12–15]. From the late endosome, Rab9 and its effector TIP47 regulate the retrograde transport of CI-MPR [16,17]. The individual depletion of any of the retrograde machineries disrupt CI-MPR sorting suggesting that all are required for successful transport. Ubiquitin is a 76aa protein that can be covalently attached to substrate lysine residues. Protein ubiquitination involves the sequential action of an E1 ubiquitin activating enzyme, an E2 ubiquitin conjugating enzyme as well as an E3 ubiquitin ligase. The conjugation of ubiquitin to transmembrane cargo regulates the internalization and sorting of a number of cell surface receptors. In addition, ubiquitination of the endosomal sorting machinery has also been shown to regulate receptor downregulation. Until very recently, a role for the ubiquitin system in the retrograde sorting route had not been described [18]. Here we report that the ubiquitin ligase RNF126, regulates the retrograde sorting of CI-MPR. Depletion of RNF126 causes the CI-MPR to be retained in endosomes and an increase in the lysosomal degradation of the receptor. In addition, we show that the efficient sorting of the CI-MPR is dependent on the ubiquitin ligase activity of RNF126. Our studies provide evidence of a role for ubiquitin in the regulation of CI-MPR sorting.
32 0 (2 014) 219 – 23 2
Materials and methods cDNA constructs CD8-CIMPR was a gift from Matthew Seaman. CD25-TGN38 and CD25-Furin were gifts from Sergio Grinstein. Full length murine RNF126 and RNF126 C231A were cloned into HA-pcDNA 3.1.
Cell culture and transfection HeLa and 293 T cells were maintained in DMEM supplemented with 10% FBS. Stable cell lines were generated by transfecting HeLa cells with a non-silencing shRNA or with an shRNA with a sense strand targeting RNF126: 5′-CACTCAAACCCTATGGACT-3′ or with 5′-TGTATGTAGGAAGAGCTTA-3′ which targets Rabring7 (Open Biosystems). 24 h post transfection, cells were treated with 1 μg/ml puromycin and single colonies were selected. For cDNA expression, cells were seeded on 10 cm2 dishes and transfected using lipofectamine-2000 (Invitrogen). Cells were fixed 24 h after transfection, except in rescue experiments where cells were fixed 48 h after post-transfection. For siRNA treatment, cells were seeded on a 6 well dish and transfected with 40 pmol siRNA targeting RNF126: 5′-CCGGATTATATCTGTCCAAGA-3′ (Qiagen) or an All-Stars negative control siRNA (Qiagen) using lipofectamine2000 (Invitrogen). Experiments were performed 48 h post transfection. When cDNA and siRNA were co-transfected, cells were treated with 40 pmol siRNA 48 h before re-transfecting with both cDNA and RNA (1ug of DNA:30 pmol RNA) for an addition 24 h. Proteins were collected from media using trichloroacetic acid (TCA). Cells were grown in DMEM without serum for 24 h and the media was collected. TCA was added to a final concentration of 25% (v/v). Samples were incubated for 30 min at 4 1C and centrifuged for 5 min at 14000RPM. The pellet was washed 2 in cold acetone and re-suspended in sample buffer.
Antibodies and reagents The following commercial antibodies were used: Cathepsin-D (Millipore, Rabbit) CI-MPR (Thermo Scientific, mouse); GFP (Millipore, chicken); GGA3 (BD Biosciences, mouse); Giantin (Covance, rabbit); GM130 (BD Biosciences, mouse); HA (Novus Biologicals, Rabbit); TnfR (Millipore, rabbit); β-tubulin (Sigma, mouse); Rab7 (Cell Signaling, rabbit); Rab9 (Cell Signaling, rabbit). The following reagents were used: A555-CTxB (Molecular Probes), Leupeptin
Fig. 1 – Depletion of RNF126 causes the lysosomal degradation of the CI-MPR. (A) Depletion of RNF126 reduces the levels of the CIMPR. Non-silencing, RNF126 or Rabring7 knockdown lines were fixed and stained with anti-CI-MPR or anti-transferrin. Scale bar, 10 μm. The graph to the right shows the average volumetric fluorescence intensity above threshold measured on individual cells (n¼ 20) of 3 replicate experiments. Results are expressed as mean7SEM. (B) Non-silencing, RNF126 or Rabring7 knockdown lines were lysed in RIPA buffer, run on an SDS-PAGE gel under non-reducing conditions and analyzed with an anti-CIMPR antibody. A separate SDS-PAGE gel was run under reducing conditions and analyzed with anti β-tubulin. The graph to the right shows the mean CI-MPR immunoreactivity normalized to the non-silencing lines of 3 independent experiments. Error bars represent standard error of the mean. (C) The reduction in CI-MPR levels in the RNF126 knockdown lines can be rescued with leupeptin. Non-silencing and RNF126 knockdown lines were incubated with 4 mM leupeptin or a vehicle control for 24 h, lysed in RIPA, and analyzed as described in part B, n¼ 3. (D) The depletion of RNF126 leads to the secretion of immature cathepsin-D. Non-silencing and RNF126 knockdown lines were washed with PBS and incubated for 24 h in serum free media. The protein content from the media was precipitated using trichloroacetic acid. Lysate and media samples were analyzed with an antibody against cathepsin-D. Precursor (p), intermediate (i) and mature (m) forms are indicated. The lysate samples were run on separate SDS-PAGE gels and analyzed with an antibody against RNF126 or tubulin.
EX PE R IM EN TA L C ELL RE S EA RC H
(Sigma). RNF126 and Rabring7 antibodies were produced by immunizing rabbits with peptides corresponding to the carboxyterminal regions of each protein, RNF126: CSPSNENATSNS, Rabring7: CNRFSNDSQLHDRWTF. The following secondary antibodies were
3 20 (2014) 219 –2 32
221
used: HRP-linked anti-mouse IgG (GE Healthcare, sheep), Cy3conjugated donkey anti-mouse IgG, A488-conjugated donkey antirabbit IgG and Cy5-conjugated donkey anti-mouse IgG (Jackson ImmunoResearch).
222
E XP E RI ME N TA L CE L L R ES E ARC H
Immunocytochemistry Cells were fixed in a 2% PFA 30 mM sucrose solution for 30 min at room temperature and subsequently incubated in 100 mM glycine for 10 min. Cells were permeabilized with 0.1% TX-100 and blocked in a 3% normal donkey serum (Jackson ImmunoResearch). Coverslips were inverted onto primary antibody and incubated for 30 min at 37 1C in a humidified chamber, washed 3 10 min in 0.05% TX-100 and inverted onto secondary antibody for an addition 30 min at 37 1C. Coverslips were washed for an addition 3 10 min in PBS with 0.05% TX-100 before mounting in fluorescent mounting medium (Dako). Images were acquired using a Zeiss Axiovert 200 inverted fluorescent microscope using a 60 /1.35NA objective equipped with a back-thinned EM charged-coupled device camera (Hamamatsu, C9100-13) and a spinning disk confocal scan head. Equipment was driven by Volocity (PerkinElmer) acquisition software. All images were acquired at room temperature. The images collected through the focal planes of the Z-axis were subsequently collapsed into one image using Volocity for final presentation. Linear adjustments for contrast and brightness were performed using the Volocity software.
32 0 (2 014) 219 – 23 2
Therefore, this measurement protocol creates 4 non-overlapping ROIs on each cell found within a field of view and also compensates for variation in volume between cells. The sum fluorescence intensity within each ring is shown. Graphpad was used for statistical analysis. Unpaired students t-tests are reported.
Uptake experiments For antibody uptake, Cells plated on coverslips were washed with PBS pre-chilled to 4 1C, transferred to a new 24 well dish and incubated with DMEM containing 0.2% BSA for 30 min at 4 1C. Next, coverslips were flipped onto a drop of primary antibody diluted in 0.2% BSA, incubated for 30 min at 4 1C and then washed PBS. Coverslips were transferred to a new 24 well dish and incubated in DMEM with 0.2% BSA pre-warmed to 37 1C for the indicated period of time, washed with PBS, and fixed in PFA. For A555-CTxB uptake, HeLa cells plated on coverslips were inverted onto a drop of DMEM containing 0.2% BSA and 4ug/ml A555-CTxB and incubated for 30 min at 4 1C, washed in PBS, and returned to 37 1C for the indicated period of time before fixation.
Immunoblotting Image analysis To measure the mean fluorescence intensity, the level of background fluorescence was determined by measuring 5 regions of interest (ROIs) devoid of any cells. Individual cells were selected and the Cy3 intensity per unit volume was determined using Volocity. For colocalization analysis, images were first deconvolved using the fast restorative function in Volocity. A point spread function was applied to each channel with the following parameters: Numerical aperture¼ 1.2, medium refractive index¼ 1.33 and emission wavelengths of 561 nm (Cy3) or 520 nm (GFP). A threshold value for each image was determined by measuring 5 ROIs devoid of any cells. Individual cells were selected and the Pearson's correlation above threshold was determined using the colocalization function in Volocity. Graphpad (prism) was used for statistical analysis. Unpaired students t-tests are reported. To measure the extent of dispersion, a measurement protocol was designed in Volocity that measured the total fluorescence intensity in eroded ROIs emanating from the outside of the cell inwards. Using the ‘find objects’ function in Volocity, individual cells within a field of view were selected. To find the entire volume of the cell, the ‘find objects’ function detected the fluorescence intensity above an automatically determined threshold. Objects with a volumeo100 μm3 were excluded and the noise from the detected objects was removed using a fine filter. The second ROI was determined by finding the volume of the entire cell using the same parameters but was subsequently eroded in volume by 2 iterations. Similarly, the third ROI was eroded in volume by 5 iterations (objects with a volume ofo10 μ m3 were excluded) and the fourth ROI eroded by 9 iterations. Next, the total fluorescence intensity contained within the identified ROIs were subtracted from each other to create nonoverlapping rings of intensity. ROI 2 was subtracted from ROI 1 to create an outer ring. ROI 3 was subtracted from ROI 2 to create a middle outer ring. ROI 4 was subtracted ROI 3 to create a middle inner ring while the inner ring corresponds to the fourth ROI.
Cells were lysed in RIPA buffer (150 mM NaCl, 20 mM Tris (Ph 7.5), 1% NP-40, 1% sodium deoxycholate, 0.1% SDS) and boiled in SDSLaemmli sample buffer. When immunoblotting for the CI-MPR, samples were prepared in non-reducing sample buffer. Proteins were separated by SDS-PAGE and transferred to PVDF membrane (Pall Corp).
Results RNF126 regulates the endosomal sorting of the CI-MPR The E3 ubiquitin ligases RNF126 and Rabring7 have been indicated in the endocytic sorting of the EGFR and other cell surface receptors [19–21]. To investigate additional roles for these E3 ubiquitin ligases in membrane traffic, we examined the steady state distribution of endosomal cargo that follows different trafficking itineraries to that of EGFR HeLa cell lines that stably express either a non-silencing shRNA, an shRNA that targets RNF126, or an shRNA targeting Rabring7 were generated [21]. The distribution of the transferrin receptor was comparable between non-silencing and both knockdown suggesting that the endocytosis and recycling of this receptor is unaffected (Fig. 1A). In addition, the distribution and intensity of CI-MPR staining was comparable between non-silencing and Rabring7 knockdown lines. However, in cell lines stably depleted of RNF126 we noticed a reduction in the total levels of the CI-MPR by both immunofluorescence and western blot (Fig. 1A and B). To test whether depletion of RNF126 was causing the CI-MPR to be degraded in the lysosome, we treated non-silencing and RNF126 knockdown lines with leupeptin, an inhibitor of lysosomal hydrolazes, or a vehicle control. Treatment of the RNF126 knockdown lines with leupeptin partially restored the levels of the CI-MPR to those observed in the non-silencing lines, suggesting that CI-MPR is degraded in the lysosome in cell lines depleted of RNF126 (Fig. 1C).
EX PE R IM EN TA L C ELL RE S EA RC H
170 IB: Tubulin
34
IB: RNF126
2.5 2.0 1.5 1.0 0.5 0.0 Control si
55
the treatment with an siRNA directed against RNF126 increased the total CI-MPR levels by western blot (Fig. 2A). We next immunostained control and RNF126 depleted cells for the CIMPR. In cells treated with RNF126 siRNA, the CI-MPR appeared in peripherally dispersed vesicles compared to the tightly packed perinuclear staining pattern observed in the control siRNA treated cells (Fig. 2B). To quantify the degree of dispersion in the staining pattern, we devised a protocol that measured the fluorescence in 4 non-overlapping regions of interest mapped over the volume of the cell. Using this protocol, we found that total CI-MPR staining intensity in the control siRNA treated cells decreased from the inner most region to the cell periphery. In cells depleted of RNF126, CI-MPR staining intensity remained relatively constant in all regions of interest, confirming that the distribution of the CI-MPR had shifted towards the cell periphery (Fig. 2C).
RNF126 si
IB: CI-MPR
fold change in CI-MPR immunoreactivity
RNF126 si
Control si
The CI-MPR serves to sort lysosomal hydrolases from the Golgi to the endocytic system. We next analyzed whether the depletion of RNF126 resulted in the missorting of the lysosomal aspartyl protease, cathepsin D. Cathepsin D is synthesized in a precursor form that subsequently matures into an intermediate and mature form upon reaching the acidic milieu of the endosomal and lysosomal compartments. Under normal conditions, cathepsin D is efficiently sorted towards the endocytic system and little is secreted from the cell [22]. We analyzed both intracellular and extracellular cathepsin D by immunoblot. Compared to control, the depletion of RNF126 caused an increase in the amount of secreted immature cathepsin D, suggesting that CI-MPR cargo are inefficiently sorted in these cells (Fig. 1D). To further assess the short-term effects of RNF126 knockdown on the CI-MPR, we used siRNA to deplete RNF126 from HeLa cells. In contrast to observations in the stable knockdown lines,
RNF126 si sum fluorescence intensity (10 )
4.0 P=0.006
CI-MPR
9
Control si
223
3 20 (2014) 219 –2 32
Control si RNF126 si
3.0
2.0
1.0
Outer ring
middle outer ring
middle inner ring
Inner ring
0.0
Fig. 2 – Transient knockdown of RNF126 leads to the dispersal of the CI-MPR into peripheral structures. (A) Transient depletion of RNF126 increases the total levels of the CI-MPR. HeLa cells were treated with control or an siRNA directed against RNF126. Cells were lysed 48 h post transfection, run on an SDS-PAGE gel under non-reducing conditions and analyzed with an anti-CI-MPR antibody. A separate gel was run under reducing conditions and analyzed by immunoblotting with antibodies against β-tubulin and RNF126. The graph to the right shows the mean CI-MPR immunoreactivity relative to control between 3 independent experiments. Bars represent standard error of the mean. (B) The depletion of RNF126 leads to the dispersal of the CI-MPR. HeLa cells treated with a control or siRNA directed against RNF126 were fixed 48 h post-transfection and stained with an anti-CI-MPR antibody. Scale bar, 10 μm. (C) CI-MPR staining in the cell periphery is increased in HeLa cells treated with RNF126 siRNA. Total fluorescence intensity in 4 regions of interest mapped over the volume of the cell (see materials and methods). 15 fields of view were quantified for each condition between 3 replicate experiments. Results are expressed as mean7SEM.
32 0 (2 014) 219 – 23 2
3.0
Control si RNF126 si
2.0
1.0
Outer ring
Inner ring
0.0 middle outer ring
RNF126 si
Giantin
Control si
sum fluorescence intensity (109)
E XP E RI ME N TA L CE L L R ES E ARC H
middle inner ring
224
Fig. 3 – The Golgi complex is mildly disrupted in RNF126 depleted cells. (A) The depletion of RNF126 leads to the dispersal of Giantin. HeLa cells treated with an control siRNA or an siRNA directed against RNF126 were fixed and stained for Giantin. Scale bar, 10μm. Images were analyzed as in Fig. 2C. Six fields of view were examined for each condition between 3 replicate experiments. Results are expressed as mean7SEM.
Disruptions of CI-MPR retrograde sorting machinery are sometimes accompanied by Golgi fragmentation. We therefore examined the Golgi in cells depleted of RNF126. Cells treated with control or RNF126 siRNA were stained with antibodies directed against Giantin. In cells depleted of RNF126, the Giantin staining appeared slightly more dispersed but remained on one side of the nucleus (Fig. 3). This phenotype is similar to that observed upon depletion of mVps26, SNX1, GCC88 or Arl5b [12,23–25] but is not as fragmented as in cells depleted of the Golgin GCC185 [26]. This suggests that the depletion of RNF126 has only a mild effect on Golgi structure.
CFP-Rab11 into cells treated with an siRNA directed against RNF126 and measured the colocalization between the CI-MPR and each Rab protein (Fig. 5A). In control siRNA treated cells, the degree of overlap between the CI-MPR and Rab4 or Rab5 were roughly equivalent, and greater than the colocalization observed between the CI-MPR and Rab11. In cells treated with the RNF126 siRNA, the overlap between the CI-MPR and Rab4 increased, while the colocalization between the CI-MPR with either Rab5 or Rab11 remained comparable to control (Fig. 5B). This suggests that the depletion of RNF126 causes the CI-MPR to accumulate in a Rab4 positive endocytic recycling compartment.
The depletion of RNF126 causes the redistribution of the CI-MPR into an endocytic recycling compartment
RNF126 regulates the retrograde sorting of the CI-MPR
We reasoned that the dispersed CI-MPR structures may be endosomal compartments. To determine the identity of the CIMPR containing structures, we co-stained control and RNF126 depleted cells for the CI-MPR and a number of endocytic markers. Control and RNF126 depleted cells were co-stained with antibodies against the CI-MPR and the transferrin receptor, a marker of the endocytic recycling compartment, or the late endosomal marker Rab7. We also co-stained with Rab9, an important regulator of CI-MPR retrograde sorting [16,27,28] (Fig. 4A). In cells depleted of RNF126, there was a significant increase in the degree of overlap between the CI-MPR and the transferrin receptor, suggesting that a higher proportion of CI-MPR is found in the endocytic recycling compartment under these conditions. Depleting RNF126 from cells did not alter the degree of overlap between the CI-MPR and Rab7 or Rab9 (Fig. 4C) suggesting that RNF126 does not play a role in Rab9 mediated CI-MPR sorting [29]. After endocytosis, the transferrin receptor is sorted through Rab4, Rab5 and Rab11 domains. These Rab proteins can be found on the same endosome, but do not significantly intermix over time [30]. To further define the location of the CI-MPR in cells depleted of RNF126, we transfected either CFP-Rab4, CFP-Rab5 or
Steady state distribution of the CI-MPR is achieved through balancing the amount of receptor delivered to the endosome versus the quantity of receptor removed by retrograde transport. To examine whether depletion of RNF126 affected the retrograde sorting route, we performed antibody uptake experiments. HeLa cells were treated with control or RNF126 siRNA and transfected with a CD8-CIMPR chimera [12]. Cells were chilled, incubated with anti-CD8, and either immediately fixed or warmed to 37 1C to allow the internalization of the antibody bound cargo for either 8 or 24 min before fixation (Fig. 6A). In cells treated with the control siRNA, the majority of the CD8-CIMPR had been sorted towards the center of the cell at the 24min time point. However, in cells depleted of RNF126, the sorting of CD8-CIMPR towards the inner region of the cell was delayed. Significantly less receptor had reached the central most region of the cell compared to control (Fig. 6B). Furthermore, the degree of overlap between the CI-MPR and the TfnR was significantly increased in cells treated with RNF126 siRNA, compared to control siRNA, at the 24 min time point (Fig. 6C). Taken together, these results suggest that the CI-MPR is retained in a transferrin receptor positive compartment in cells depleted of RNF126, which is consistent with a role for RNF126 in regulating the retrograde sorting of the CI-MPR.
EX PE R IM EN TA L C ELL RE S EA RC H
CI-MPR
225
merge
RNF126 si
Control si
TfR
3 20 (2014) 219 –2 32
CI-MPR
merge
Rab9
CI-MPR
merge
RNF126 si
Control si
RNF126 si
Control si
Rab7
P
