- Email: [email protected]
trans-Golgi and trans-Golgi network-localized Yip domain family proteins, which play a role in the Golgi reassembly and glycan synthesis
Experimental Cell Research 353 (2017) 100–108
Contents lists available at ScienceDirect
Experimental Cell Research journal homepage: www.elsevier.com/locate/yexcr
YIPF1, YIPF2, and YIPF6 are medial-/trans-Golgi and trans-Golgi network-localized Yip domain family proteins, which play a role in the Golgi reassembly and glycan synthesis
MARK
Jeerawat Soonthornsita,b, Noriko Sakaic, Yurika Sasakid, Ryota Watanabed, Shiho Osakod,e, ⁎ Nobuhiro Nakamuraa,c,d,e, a
Division of Engineering, Graduate School, Kyoto Sangyo University, Motoyama, Kamigamo, Kita, Kyoto 603-8555, Japan Department of Pre-clinic and Applied Animal Science, Faculty of Veterinary Science, Mahidol University, 999 Phutthamonthon Sai 4 Road Salaya, Phutthamonthon, Nakhon Pathom 73170 Thailand c Graduate School of Natural Science and Technology and School of Pharmacy, Kanazawa University, Kakuma, Kanazawa 920-1192, Japan d Department of Molecular Biosciences, Faculty of Life Sciences, Kyoto Sangyo University, Motoyama, Kamigamo, Kita, Kyoto 603-8555, Japan e Division of Life Sciences, Graduate School, Kyoto Sangyo University, Motoyama, Kamigamo, Kita, Kyoto 603-8555, Japan b
A R T I C L E I N F O
A BS T RAC T
Keywords: Golgi apparatus Vesicular transport Transmembrane protein Glycosylation
In this study, we attempted to explore the function of three uncharacterized mammalian homologs of yeast Yip domain family proteins—YIPF6, a homolog of Yip1p, and YIPF1 and YIPF2, which are homologs of Yif1p. Immunofluorescence staining revealed that YIPF1, YIPF2, and YIPF6 mainly localize in the medial-/transGolgi and also partially in the trans-Golgi network (TGN). On treatment with brefeldin A (BFA), the homologs co-migrated partly with medial-/trans-Golgi markers and also with a TGN marker in earlier time point, but finally redistributed within cytoplasmic punctate structures that were distinct from medial-/trans-Golgi and the TGN markers. YIPF6 formed a stable complex separately with YIPF1 and YIPF2, and knockdown of YIPF6 reduced YIPF1 and YIPF2 levels. These results suggest that YIPF6 forms complexes with YIPF1 and YIPF2 for their stable expression and localization within the Golgi apparatus. Knockdown experiments showed that YIPF1 and YIPF2, by contrast, are not necessary for the expression and localization of YIPF6. The structure of the Golgi apparatus and its disassembly after BFA treatment were not significantly affected by the knockdown of YIPF1, YIPF2, or YIPF6. However, reassembly of the Golgi apparatus after the removal of BFA was markedly delayed by the knockdown of YIPF1 and YIPF2, but not by that of YIPF6. These results strongly suggest that free YIPF6 after disassociating with YIPF1 and YIPF2 interferes with the reassembly of the Golgi apparatus. Knockdown of YIPF1 and YIPF2, but not that of YIPF6, also reduced intracellular glycans in HT-29 cells. Thus, we confirmed that YIPF1, YIPF2, and YIPF6 play a significant role in supporting normal glycan synthesis.
1. Introduction Yip domain family (YIPF) proteins are mammalian homologs of yeast Yip1p and Yif1p, which interact with Ypt family small GTPases, including Ypt1p and Ypt31p [1]. Yip1p and Yif1p form a complex with each other and are found in the ER, Golgi apparatus, and COPII vesicles. They were hypothesized to play a role in vesicle budding from the ER and/or fusion with the Golgi apparatus under regulation by Ypt proteins [1–4]. This was supported by the findings that Yip1p interacts with ER to Golgi SNAREs (Bos1p and Sec22p) [4] and YIP1 genetically interacts with genes coding an Rab-GDI (GDI1), COPII components
(SEC12, SEC13, and SEC23), a COPI component (SEC21) and a Golgi tethering protein (USO1) genes [3,5]. Both Yip1p and Yif1p are predicted to have five transmembrane segments; an N-terminal segment, which is exposed to the cytoplasm; and a short C-terminal segment, which is exposed to the lumen of the secretory pathway [6,7]. Yip1p and Yif1p show significant similarity at the transmembrane segments (Yip domain); two other non-essential proteins, Yip4p and Yip5p, also share the Yip domain [8]. Homologs of Yip1p and Yif1p are found in most eukaryotes, including fungi, protozoa, plants, and animals, indicating their essential role in cell functioning. However, the precise mode of their function remains
Abbreviations: YIPF, Yip domain family; ERGIC, ER-Golgi intermediate compartment; TGN, trans-Golgi network; BFA, brefeldin A; GnT-I, N-acetylglucosaminyltransferase-I; RFP, red fluorescent protein; GalT-I, β4-galactosyltransferase-I; PAS staining, Periodic acid-Schiff staining ⁎ Corresponding author at: Department of Molecular Biosciences, Faculty of Life Sciences, Kyoto Sangyo University, Motoyama, Kamigamo, Kita, Kyoto 603-8555, Japan. E-mail address: [email protected] (N. Nakamura). http://dx.doi.org/10.1016/j.yexcr.2017.03.011 Received 28 December 2016; Received in revised form 6 March 2017; Accepted 7 March 2017 Available online 09 March 2017 0014-4827/ © 2017 Elsevier Inc. All rights reserved.
Experimental Cell Research 353 (2017) 100–108
J. Soonthornsit et al.
2.4. Pull-down analysis
obscure [9]. Nine YIPF members are found in human cells [6,10]. Phylogenetic analyses have previously shown that YIPF4, YIPF5/YIP1A, YIPF6, and YIPF7/YIP1B are Yip1p homologs, while YIPF1, YIPF2, YIPF3, YIF1A, and YIF1B are Yif1p homologs (Fig. S1) [6]. While characterizing the human homologs, we previously found that each human Yip1p homolog (YIPF5 and YIPF4) works with a specific Yif1p homolog partner (YIF1A and YIPF3, respectively) [9,10]. Cellular distribution of each pair is different—YIPF5 and YIF1A mainly localize within the ER-Golgi intermediate compartment (ERGIC), while YIPF4 and YIPF3 mainly localize within the cis-Golgi. It is possible that each pair of YIPF proteins functions independently in their residing compartment by associating with a different set of proteins, in addition to functioning together in these compartments. Knockdown of Yip1p homologs (YIPF4 or YIPF5) causes marked reduction in the levels of their partner Yif1p homologs (YIPF3 or YIF1A), suggesting that Yif1p homologs are unstable in the absence of Yip1p homologs. The knockdown of YIPF4, YIPF3, YIPF5, and YIF1A causes significant fragmentation of the Golgi apparatus, highlighting their role in maintaining the Golgi structure. Here, we report that the Yip1p homolog YIPF6 forms complexes with the Yif1p homologs YIPF1 and YIPF2, and they localize in the medial-/trans-Golgi and TGN. Gene knockdown experiments showed that YIPF6, free of YIPF1 and YIPF2, interferes with the reorganization of the Golgi apparatus and glycan synthesis.
Pull-down analysis was performed as described previously [10]. Briefly, HeLa cells were incubated in a buffer containing digitonin and centrifuged to obtain a soluble extract. A part of the extract was kept as a sample of the extract, and the remaining extract was separated into two equal parts. Affinity-purified antibody for a YIPF protein was added to one part of the extract and affinity-purified IgG was added to another part. Protein A Sepharose CL-4B (GE Healthcare Bio- Science AB, Uppsala, Sweden) was used to pull down antibodies and bound materials. 2.5. SDS-PAGE, western blotting, and densitometry SDS-PAGE and western blotting were performed as described previously [10]. Chemiluminescence images were taken using LAS1000 and LAS4000mini (Fuji Photo Film. Inc., Tokyo, Japan). Densitometry was performed using ImageJ (National Institute of Health, Bethesda, USA). 2.6. siRNA transfection siRNA transfection was performed as described previously [10]. Briefly, HeLa cells or HT-29 cells were transfected with Stealth RNAi™ siRNAs (YIPF1: HSS122872, YIPF2: HSS149042, and YIPF6: equal amount mixture of HSS138898, HS138899, and HS138900, Invitrogen, Thermo Fisher Scientific Inc., Waltham, MA USA) by using Lipofectoamine 2000 and incubated for three days with a daily change in the medium before analyses. Silencer® Select Negative Control No. 1 siRNA (Thermo Fisher Scientific Inc.) was used as control.
2. Materials and methods 2.1. Cell culture, drug treatment, immunofluorescence staining, cell extraction, and immunoprecipitation These experiments were performed as described previously [10]. For examining the Golgi reassembly, HeLa cells were treated with a growth medium (Dulbecco's modified Eagles medium containing 10% fetal bovine serum) containing 0.25 µg/mL BFA for one hour to four hours. Then, the BFA containing medium was removed and the cells were washed with the growth medium for three times, and were incubated in the growth medium for one hour. HT-29 cells were originally purchased from ATCC and were kindly donated by Profs. A. Kurosaka and K. Sato (Kyoto Sangyo University) and cultured with the growth medium.
2.7. Immunofluorescent staining This was performed as described previously [14]. 2.8. Periodic acid-Schiff (PAS) staining 2×105 HT-29 cells were seeded on to coverslips in petri dishes 3.5 cm in diameter and were transfected with siRNA as described above. The cells were then stained using the PAS Kit (Sigma-Aldrich Co. LLC., St. Louis, MO, USA) according to the manufacturer's instructions, but with slight modifications. Briefly, the cells were fixed with a formalin–ethanol fixative solution (4% formaldehyde in 95% ethanol) and gently washed three times with tap water. The cells were then treated with the PAS solution for 5 min and washed five times with distilled water. Finally, the cells were treated with the Schiff reagent for 15 min and washed 10 times with tap water. The coverslips, containing the stained cells, were mounted in glycerol and observed under light microscope. Counter-staining with hematoxylin was omitted for the following image analysis. Micrographs were taken using a bright-field microscope with identical illumination and exposure settings for all samples. For each sample, 6–8 images were used for quantification with ImageJ. Two areas of cell clusters were selected from an image by the free-hand selection mode and two cell-free areas were selected from the same image. Mean intensity of an area was divided by pixel number of the area to estimate the area's cell density. The average density of two cell areas was subtracted by that of the two cell-free areas to obtain a specific staining density of an image. Finally, an average specific staining density of 6–8 images was calculated for each sample. For each experiment, the average specific densities of YIPF knocked-down cells were divided by that of negative control transfected cells to ascertain relative staining density of the cells. Statistical analysis was performed using StatMate V (ATMS Co., Ltd, Tokyo, Japan).
2.2. Plasmids and transfection A plasmid encoding N-acetylglucosaminyltransferase-I (GnT-I) tagged with red fluorescent protein (GnT-I-RFP) was produced by conjugating a cDNA encoding the monomeric red fluorescent protein [11] to the 3′-end of the GnT-I cDNA encoding GnT-I [12], and was inserted into pcDNA3 (Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA). Another plasmid encoding β4-galactosyltransferase-I (GalT-I) tagged with RFP (GalT-I-RFP) was purchased from Evrogen (pTagRFP-Golgi; Moscow, Russia). 2.3. Antibodies Rabbit polyclonal antibodies for YIPF1, YIPF2, and YIPF6 were produced and affinity-purified as described previously by using recombinant proteins [10]. Rabbit anti-human GM130 (GST-human GM130) was kindly donated by Drs. Sohda (Niigata University) and Misumi (Fukuoka University) [13]. The following antibodies were also purchased—mouse monoclonal anti-GM130 antibody (BD Biosciences, San Diego, CA, USA), CY3-conjugated anti-rabbit IgG, CY5-conjugated anti-rabbit IgG, CY5-conjugated anti-mouse IgG (Jackson ImmunoResearch Labs. Inc., West Grove, PA, USA), Alexa488-conjugated anti-mouse IgG (Molecular Probes, Thermo Fisher Scientific), and HRP-conjugated anti-rabbit IgG (Jackson ImmunoResearch). 101
Experimental Cell Research 353 (2017) 100–108
J. Soonthornsit et al.
Control YIPF1
YIPF6
YIPF2
B YIPF1
A
200 116 97.4 66
YIPF1
GnT-I
GM130
YIPF2
45 * 31
YIPF2
GM130
YIPF6
21.5 YIPF6
GM130
C
YIPF1 ++ - ooo o o 255 RRVALATIVTIVLLHMLLSVGCLAYFFDAPEMDHLPTTTATPNQTV AAAKSS
BFA 15min YIPF1
YIPF2
GnT-I
+ + oo o o o 255 TRLVATVLLSVVVLLHALLAMGCKLYFFQSLPPENVAPPPQITSLPSNIALSPTLPQSLAPS YIPF6 ++ 211 RRALAVYPVFLFYFVISWMILTFTPQ
YIPF2
Fig. 1. Characterization of anti-YIPF1, YIPF2, and YIPF6 antibodies. HeLa cells were analyzed by western blotting using anti-YIPF1, anti-YIPF2, and anti-YIPF6 antibodies. Molecular-size markers (kDa) are indicated on the left. Arrowheads on the right indicate the specific bands corresponding to each protein. An asterisk on the right of the YIPF2 line indicates a non-specific band. (B) HeLa cells were analyzed by double immunofluorescent staining using anti-YIPF1, anti-YIPF2, and anti-YIPF6 antibodies (CY3) together with an anti-GM130 antibody (Alexa488). Images taken using an epifluorescence microscope are shown. Scale bar=10 µm. (C) The predicted fifth transmembrane and cytoplasmic segments of YIPF1, YIPF2, and YIPF6. +, positively charged residue; −, negatively charged residue; o, predicted O-glycosylation sites; double underlines, predicted transmembrane segments. Position numbers of the first shown residue are indicated on the left.
YIPF6
3. Results
BFA 30min
3.1. Characterization of YIPF1, YIPF2, and YIPF6
YIPF1
Affinity-purified rabbit polyclonal antibodies were produced using recombinant YIPF1, YIPF2, and YIPF6 proteins. When HeLa cell lysate was analyzed using these antibodies with western blotting (Fig. 1A), a clear single band was detected for YIPF1 and YIPF6 antibodies and two bands were detectable for the YIPF2 antibody. A higher-mobility band detectable for YIPF2 turned out to be a non-specific reaction product, as a lower-mobility band was reduced by the knockdown of YIPF2 (Fig. 4 and S4). When these antibodies were used for immunofluorescent staining on HeLa cells, all antibodies stained juxtanuclear ribbon-like structures, closely localizing with GM130, a cis-Golgi marker protein (Fig. 1B). Specific staining of these structures was supported by the fact that the knockdown of YIPF1, YIPF2, and YIPF6 caused reduced staining (Fig. 5). Therefore, we concluded that these affinity-purified antibodies had adequate quality for further analyses. Molecular size of YIPF6 was predicted to be 26 kDa based on its amino acid sequence, which was identical to the molecular size of YIPF6 estimated using western blotting. Therefore, we deduced that YIPF6 did not undergo a major post-translational modification. YIPF6 has been previously predicted to have a short cytoplasmic segment with no glycosylation sites (Fig. 1C) [15]. Molecular sizes of YIPF1 and YIPF2 were predicted to be 35 and 34 kDa, respectively, based on their amino acid sequence, and their molecular sizes estimated using western blotting were 40 and 43 kDa, respectively. It was thus evident that YIPF1 and YIPF2 were post-translationally modified. This was
GnT-I
YIPF2
YIPF6
Fig. 2. Localization of YIPF1, YIPF2, and YIPF6. HeLa cells transiently expressing GnT-I-RFP; untreated (left gallery) and treated with BFA for 15 min (middle gallery) and 30 min (right gallery). The cells were then fixed and processed for immunofluorescence staining using anti-YIPF1 (upper panels), anti-YIPF2 (middle panels), and anti-YIPF6 (lower panels) antibodies (Alexa488), and were observed with fluorescence of RFP. Images were taken with a confocal microscope, and representative sliced images are shown. Scale bar=5 µm.
102
Experimental Cell Research 353 (2017) 100–108
J. Soonthornsit et al.
Pull down
likely attributable to O-glycosylation, as was the case with YIPF3 [9], considering that YIPF1 and YIPF2 are predicted to have five putative O-glycosylation sites in their cytoplasmic segments (Fig. 1C). No other glycosylation sites were detectable for YIPF1, YIPF2, and YIPF6.
E F1 C
E F2 C
E F6 C
Detection
YIPF1
3.2. YIPF1, YIPF2, and YIPF6 were localized in the medial-/transGolgi and TGN, but showed distinct dynamics from these compartment markers Fine localization of YIPF1, YIPF2, and YIPF6 was further analyzed using double staining with YIPF proteins and Golgi compartment markers—medial-Golgi, N-acetylglucosaminyltransferase-I tagged with red fluorescent protein (GnT-I-RFP); trans-Golgi, β4-galactosyltransferase-I tagged with RFP (GalT-I-RFP); TGN, TGN46 (Fig. 2, S2 and S3). Close inspection with confocal microscopy showed that YIPF1, YIPF2, and YIPF6 mainly localized within medial-/trans-Golgi and TGN, while less so within cis-Golgi, as deduced from the marker GM130 (Fig. 2 and S3; control). When HeLa cells were treated with brefeldin A (BFA), the bulk of medial-/trans-Golgi proteins relocated in the ER as evidenced by the localization in the nuclear membrane, whereas most of cis-Golgi proteins were retained within cytoplasmic punctate structures called “Golgi remnants,” which are distinct from the ER (Fig. S2, BFA 30 min) [9,16]. A TGN marker, TGN46, was also distributed distinctly from the cis-, medial- and trans-Golgi markers, consistent with the previous finding that TGN proteins are relocated in the endosomal compartment [17]. During BFA treatment, YIPF1, YIPF2, and YIPF6 showed prominent redistribution patterns that were different form the medial- and trans-Golgi marker proteins (Fig. 2 and S3; BFA), despite the former's close localization with the latter in non-treated cells. YIPF1, YIPF2, and YIPF6 were frequently observed along tubular intermediate structures stained with the medial- (Fig. 2; BFA; open arrows), trans-Golgi and also TGN markers after BFA treatment (Fig. S3; BFA; open arrows). After 30 min of BFA treatment, YIPF1, YIPF2, and YIPF6 mostly relocalized to the cytoplasmic punctate structures, with occasional colocalization with the medial-/trans-Golgi and TGN markers (arrows and open arrows). Strikingly, YIPF1, YIPF2, and YIPF6 was not found in the nuclear membrane and cytoplasmic reticular structures, which indicate the ER localization, in contrast with medial-/trans-Golgi markers (Fig. 2 and S3, open arrowheads). Therefore, it is likely that YIPF1, YIPF2, and YIPF6 were not retrieved back to the ER but mainly retained within the cytoplasmic structures similar to but distinct from Golgi remnants after BFA treatment.
YIPF2 YIPF6 YIPF4 YIPF5
Fig. 3. Complex formation of YIPF1 and YIPF2 with YIPF6. HeLa cell extract was pulled down with anti-YIPF1 (F1), anti-YIPF2 (F2), anti-YIPF6 (F6), and control IgG (C), as indicated on top of the panels. The precipitated material (F1, F2, F6, and C) were analyzed using western blotting with anti-YIPF1, anti-YIPF2, anti-YIPF6, anti-YIPF4, and anti-YIPF5 antibodies, as indicated in the left panel. A part (1/3.25) of the cell extract (E) was also analyzed as an input control.
3.4. Effects of YIPF1, YIPF2, and YIPF6 knockdown
F1+2 KD
F6 KD
F2 KD
Control
A
F1 KD
To analyze the function of YIPF1, YIPF2, and YIPF6, knockdown experiments were performed using pre-designed StealthRNAi siRNA (Thermo Fisher Scientific Inc.). Three available siRNAs for each YIPF proteins were used individually or in a mixture to knockdown the target proteins in HeLa cells. The result of western blotting showed that the siRNAs specifically and efficiently reduced the target proteins in similar levels after three days of transfection in all the conditions (Fig. 4 and S4 and unpublished observation), although slight cytotoxic effects, e.g. cell growth delay, cell detachment and cell rounding, were observed in
YIPF1 YIPF2
* YIPF6
B
3.3. YIPF6 forms a distinct complex with either YIPF1 or YIPF2
Relative intensity (%)
120
To analyze whether YIPF1 and YIPF2 formed a complex with YIPF6, pull-down assays were performed. HeLa cells were extracted by the digitonin-containing buffer, and YIPF1, YIPF2, and YIPF6 were pulled down from the extract using affinity-purified antibodies for each protein. The pulled-down proteins were then analyzed using western blotting with anti-YIPF antibodies. As seen in Fig. 3, each YIPF protein was efficiently pulled down by the corresponding antibody (F1, F2, and F6), but not by the control IgG (C). YIPF4 and YIPF5, which are Yip1p homologs localizing within cis-Golgi and ERGIC, respectively, were not pulled down by any of the antibodies, which confirmed that the pulldown assays were specific. The anti-YIPF6 antibody precipitated YIPF1 and YIPF2, while anti-YIPF1 and anti-YIPF2 antibodies precipitated the YIPF6 protein. This indicated that most YIPF1 and YIPF2 formed a complex with YIPF6. By contrast, YIPF2 was poorly pulled down by the anti-YIPF1 antibody, while YIPF1 was poorly pulled down by antiYIPF2 antibody. This result suggested that most YIPF1 or YIPF2 formed distinct complexes with YIPF6, and a ternary complex of YIPF1–YIPF2–YIPF6 was present in small quantity.
YIPF1
100
YIPF2
80
YIPF6
60 40 20 0
F1 KD
F2 KD F1+2 KD F6 KD
Fig. 4. Protein levels of YIPF1, YIPF2, and YIPF6 after knockdown. (A) HeLa cells were transfected with siRNAs targeting YIPF1 (F1 KD), YIPF2 (F2 KD), both YIPF1 and YIPF2 (F1+F2 KD), YIPF6 (F6 KD), and control double-stranded RNA (Control). The cells were then lysed and analyzed by western blotting using anti-YIPF1, anti-YIPF2, and anti-YIPF6 antibodies. Representative result of three independent replicates is shown. (B) YIPF protein levels were quantified using densitometry, and the relative amount of each protein compared with the control was shown. The average of three independent replicates is shown with s.e.m. (error bars).
103
Experimental Cell Research 353 (2017) 100–108
J. Soonthornsit et al.
extended ribbon-like structure in about the half of the cells ( > 40%) after two hours (W/O 2 h; F1 KD, F2 KD, and F1+2 KD). After four hours, the Golgi apparatus was reassembled into a compact ribbon-like structure in most of the cells (W/O 4 h; F1 KD, F2 KD, and F1+2 KD; ~80%). These results indicated that the Golgi reassembly was severely delayed in YIPF1 or YIPF2 knockdown cells. Surprisingly, the Golgi apparatus was reassembled almost normally in YIPF6 knocked-down cells at all the time points after the removal of BFA (F6 KD). Because YIPF6 knockdown caused a simultaneous reduction in YIPF1 and YIPF2 levels (Fig. 4 and S4), it is likely that the reduction in YIPF1 and YIPF2 per se did not inhibit the Golgi reassembly. Instead, the presence of free YIPF6, which lost the partner proteins YIPF1 and/or YIPF2, was likely the cause of the hampered Golgi reassembly. Whether such a hampered reassembly affects the function of the Golgi apparatus was further explored. Homozygous truncation mutation in YIPF6 has been previously shown to cause a defective formation and secretion of large secretory granules containing mucin from Paneth and goblet cells [18]. Therefore, we looked for model cell lines that can simulate intestinal granule secretion. Among the several cell lines tested, HT-29, a human colon carcinoma cell line [19], was suitable for our purpose because it accumulates glycans at high levels even under a growing culture condition in non-differentiated state. Western blotting analysis showed that the knockdown was successful for all YIPF proteins; however, the knockdown efficiency was slightly lower for HT-29 cells compared with that of HeLa cells (Fig. 7). Similar to HeLa cells, YIPF6 knockdown reduced YIPF1 and YIPF2 levels, while those of YIPF6 remained unchanged following YIPF1 and YIPF2 knockdown in HT-29 cells (Fig. 7A and B). The cells were then analyzed with PAS staining to check the quantity of glycans accumulated in the cells, and the staining was quantified using image analysis. As seen in Fig. 7C, YIPF1, YIPF2, and YIPF1–YIPF2 double knocked-down cells showed fainter staining, while that across YIPF6 knocked-down cells was more or less similar to the control cells. Quantitative analysis showed that the staining intensity of YIPF2 and YIPF1–YIPF2 double knockeddown cells was lower compared to the control cells; the difference was statistically significant (P < 0.01) (Fig. 7D). The staining intensity also tended to be lower for YIPF1 knocked-down cells, but was not statistically significant. The staining intensity of YIPF6 knocked-down cells was not different from that of the control cells. These results suggest that the knockdown of YIPF2—and possibly of YIPF1, but not of YIPF6—reduced the amount of intracellular glycans in HT-29 cells.
some conditions. These were strongly suggested to be off-target effects because they were not common for all the three siRNA or a mixture, which knocked down the target proteins in similar levels. Therefore, one of the three siRNAs for YIPF1 and YIPF2 while a mixture of the three siRNAs for YIPF6, which showed least off-target effects, were used for further analyses. Reduction in YIPF6 levels caused a severe reduction in those of YIPF1 and YIPF2 (Fig. 4, F6KD). This effect was not an off-target effect because all the three siRNAs reduced YIPF1 and YIPF2 to a similar level (Fig. S4). However, YIPF6 levels were not reduced due to the knockdown of YIPF1 (F1KD), but were slightly reduced due to that of YIPF2 (F2KD) along with a simultaneous knockdown of YIPF1 and YIPF2 (F1+2KD). In addition, YIPF1 levels remained unaffected by the knockdown of YIPF2 (F2KD) and YIPF2 levels reduced slightly after the knockdown of YIPF1 (F1KD). These results suggested that the presence of YIPF6, a Yip1p homolog, was essential for the stable expression of YIPF1 and YIPF2, Yif1p homologs, while YIPF6 expression was independent of YIPF1 and YIPF2. In addition, the expression of YIPF1 and YIPF2 was mostly independent of each other. These results were consistent with the previous finding that the stable expression of Yif1p homolog (YIPF3 and YIF1A) was dependent on the expression of the partner Yip1p homolog (YIPF4 and YIPF5, respectively) while the expression of Yip1p homolog was independent from the expression of the partner Yif1p homolog [9,10]. Immunofluorescence staining was performed to analyze whether the knockdown of YIPF1, YIPF2, or YIPF6 affected the morphology of the Golgi apparatus. siRNA-transfected cells were double stained with each of the YIPF and Golgi marker proteins (Fig. 5). Overall, staining of the YIPF proteins was consistent with the results of western blotting (Fig. 4). Fluorescence of YIPF proteins was markedly reduced owing to the knockdown of corresponding YIPF proteins. The fluorescence of YIPF1 was not significantly reduced by the knockdown of YIPF2 and vice-versa. Knockdown of YIPF6 caused a strong reduction in the fluorescence of YIPF1 and YIPF2, while knockdown of YIPF1 and YIPF2, individually or simultaneously, did not significantly affect YIPF6 staining, which agrees with the western blotting result. Therefore, we concluded that the localization of YIPF6 in the Golgi apparatus is not affected by the knockdown of YIPF1 or YIPF2. No considerable morphological changes were observed for cis-Golgi (GM130) (Fig. 5), medial-/trans-Golgi (giantin), and TGN (TGN46) (Fig. S6) following the knockdown of the YIPF proteins. These results suggested that the Golgi structure is resistant to the knockdown of YIPF1, YIPF2 or YIPF6 under normal cell culture conditions. Knockdown of YIPF1, YIPF2 or YIPF6 did not affect staining intensity and distribution of YIPF4 or YIPF5 (Fig. S5) further supporting that YIPF1, YIPF2 and YIPF6 were specifically knocked down. To explore the function of YIPF proteins, we analyzed whether the knockdown of YIPF proteins has any effect on the Golgi apparatus by using treatments that reversibly perturb the structure of the Golgi apparatus. We tried several drugs and conditions and found that knockdown of YIPF proteins affects the recovery from BFA induced Golgi disassembly. HeLa cells were treated with BFA for one hour and then washed and incubated with a growth medium up to four hours (Fig. 6A, B and S6). The Golgi apparatus was reassembled into a continuous ribbon-like structure in the majority ( > 70%) of the cells transfected with control siRNA one hour after the removal of BFA (W/ O 1 h, Cont). The number of the cells showing reassembled Golgi increased ( > 80%) after two hours (W/O 2 h), and it was finally restored to the level before BFA treatment (~90%) after four hours (compare N/T and W/O 4 h). In contrast, the Golgi apparatus became a bigger and brighter punctate structures spread in the cytoplasm but was not reassembled into a ribbon-like structure in most of the cells ( > 95%), which were either individually or simultaneously knocked-down YIPF1 or YIPF2, one hour after the removal of BFA (W/O 1 h; F1 KD, F2 KD, and F1+2 KD). The Golgi apparatus was then reassembled into a slightly
4. Discussion 4.1. Localization of YIPF1, YIPF2, and YIPF6 Immunofluorescent staining results strongly suggested that YIPF1, YIPF2, and YIPF6 mainly localized within medial-/trans-Golgi compartments and partly within the TGN. These results are consistent with the observation that transiently expressed YFP-tagged YIPF1, YIPF2, and YIPF6 are enriched in trans-Golgi compartments [7]. While most of the medial-/trans-Golgi proteins relocated to the ER following BFA treatment, YIPF1, YIPF2, and YIPF6 remained mainly within the cytoplasmic punctate structures. These structures were largely independent of the structures containing cis-Golgi and TGN markers. cisGolgi proteins including syntaxin5, GM130, YIPF3 and YIPF4 have been reported to relocalize to the structures called “Golgi remnants,” which overlap with ERGIC [9,16], while TGN proteins relocalize to structures overlapping with endosomes [17]. Thus, YIPF1, YIPF2, and YIPF6 mainly relocalized to the compartments independent of the cisGolgi/ERGIC and TGN/endosomes. It is unlikely that YIPF1, YIPF2, and YIPF6 were transported back to and re-exited the ER during recovery from BFA treatment to reach the compartments separate from the cis-Golgi/ERGIC or TGN/endosomes, considering that YIPF1, YIPF2, and YIPF6 did not show a significant overlap with ERGIC53 104
Experimental Cell Research 353 (2017) 100–108
J. Soonthornsit et al.
Control
YIPF1 KD
YIPF2 KD
YIPF1+2 KD
YIPF6 KD
YIPF1
GM130
YIPF2
GM130
YIPF6
GM130
Fig. 5. Distribution of YIPF1, YIPF2, and YIPF6 after knockdown. HeLa cells were transfected with siRNA as described in Fig. 4 and processed for double immunofluorescence staining using anti-YIPF1, anti-YIPF2, and anti-YIPF6 antibodies (CY3) together with anti-GM130 antibody (Alexa488). Images taken by epi-fluorescence microscope are shown. Scale bar=20 µm.
stable expression of Yif1p homologs. Immunoprecipitation experiments suggested that YIPF1 and YIPF2 formed a distinct complex with YIPF6 and that a ternary complex existed in a small quantity (Fig. 3). Preliminary experiments suggested that YIPF6 formed a hetero-oligomer (4–16 mer) with either YIPF1 or YIPF2, which has been previously observed for YIPF5–YIF1A and YIPF4–YIPF3 [9,10]. Although it is probable that YIPF1–YIPF6 and YIPF2–YIPF6 form independent complexes despite having YIPF6 as a common subunit, it remains to be seen whether these two complexes have a distinct function.
(unpublished observation) or GM130 (Fig. S3). Therefore, it is likely that majority of YIPF1, YIPF2, and YIPF6 resisted the retrograde flow and remained within the cytoplasmic structures that are similar to but not the same with the Golgi remnants. Detailed difference in dynamics between YIPF1, YIPF2, YIPF6 and other medial-/trans-Golgi proteins has to be clarified in future. 4.2. YIPF1 and YIPF2 stabilize by forming a complex with YIPF6 As is the case for other YIPF members, YIPF6, a Yip1 homolog, formed a complex with YIPF1 and YIPF2, Yif1p homologs (Fig. 3), and the expression of YIPF6 was essential for the stable expression of YIPF1 and YIPF2 (Figs. 4 and 7A and B). YIPF1, YIPF2, and YIPF6 did not interact with YIPF4 or YIPF5 (Fig. 3), suggesting that each Yip1p homolog forms a complex with a predesignated Yif1p homolog— YIPF5/YIP1A with YIF1A, YIPF4 with YIPF3, and YIPF6 with YIPF1 and YIPF2. For all these pairs, the expression of Yip1p homologs is necessary for the stable expression of Yif1p homologs. This supports the hypothesis that the expression of Yip1p homologs is vital for the
4.3. Absence of YIPF1 or YIPF2 inhibits the reassembly of the Golgi apparatus and reduces intracellular glycans Although knockdown of YIPF1 or YIPF2 induced no significant effect on the Golgi morphology under normal culture condition, it inhibited the Golgi reassembly after the removal of BFA (Fig. 6 and S6). This suggested that a minimal function of the cells pertaining to the Golgi assembly was retained under reduced YIPF1 or YIPF2 level. It is 105
Experimental Cell Research 353 (2017) 100–108
J. Soonthornsit et al.
A N/T
Cont
F1 KD
F2 KD
F1+2 KD
F6 KD
BFA
W/O 1h
W/O 2h
W/O 4h
B
100% 80% Cont F1 KD F2 KD F1+2 KD F6 KD
60% 40% 20% 0%
Cont
BFA
W/O 1h
W/O 2h
W/O 4h
Fig. 6. Effect of YIPF1, YIPF2 and YIPF6 knockdown on Golgi reassembly after BFA treatment. (A) HeLa cells were transfected with siRNA as described in Fig. 4 (N/T). The cells were then treated with BFA for 1 h (BFA) and control cells were further washed and incubated with normal medium for 1 h, 2 h or 4 h (W/O 1 h, W/O 2 h and W/O 4 h). The cells were fixed and processed for triple immunofluorescence staining using anti-GM130 (CY3), anti-giantin (CY5), and anti-TGN46 (Alexa488) antibodies. Representative images taken by an epi-fluorescence microscope are shown; however, only GM130 staining has been shown for simplicity and space limitation. Full sets of images are shown in Fig. S6. Scale bar=20 µm. (B) Relative numbers of the cells showing a ribbon-like structure by GM130 staining in (A) were quantified. Examples of the categorized cells are shown in Fig. S6. The average of three independent replicates is shown with s.e.m. (error bars).
glycans were also reduced in YIPF1 and YIPF2 knocked-down HT-29 cells (Fig. 7). This was attributable to either reduced glycan synthesis or enhanced glycan secretion. Enhanced glycan secretion is less likely to be the underlying cause because our preliminary experiment showed that the transport of VSV-G, a protein widely used as a secretorypathway marker, was not much affected by YIFP knockdown. The Golgi apparatus is the main compartment wherein O-linked glycan chains are synthesized. Therefore, it likely that glycan synthesis was reduced in the YIPF1 or YIPF2 knocked-down cells. It was probable that the presence of free YIPF6 caused incomplete or delayed recycling of Golgi enzymes inducing some defects in the glycan synthesis machinery while the structure of the Golgi apparatus was maintained somehow. A peculiar observation of our present study is that YIPF6 knockdown did not affect our analysis, considering that the deletion of the
probable that this function was strained with the massive need of reassembly after BFA removal. This hypothesis is supported by the observation that the Golgi apparatus was reassembled gradually, taking longer incubation times (~4 h) after BFA removal (Fig. 6 and S6). The Golgi localization of YIPF6 remained largely unaffected by the reduction in YIPF1 and YIPF2 levels (Fig. 5). This indicates that YIPF6 exists within the Golgi free of partner proteins. Because Yip1p/Yif1p has been proposed to be involved in vesicle fusion in the Golgi apparatus, it is probable that the YIPF1/YIPF2–YIPF6 complexes play a role in membrane fusion within the Golgi apparatus. Free YIPF6 may inhibit the fusion of the Golgi apparatus and associated transport vesicles during its reassembly after BFA removal, thereby leading to an incomplete reassembly in YIPF1 and YIPF2 knocked-down cells. Corresponding to the hampered Golgi reassembly, intracellular 106
Experimental Cell Research 353 (2017) 100–108
YIPF2 YIPF6
Y I P F1 + 2 K D
Y I P F6 K D
Y I P F2 K D
Relative intensity (%)
YIPF1
YIPF1 K D
C o nt
J. Soonthornsit et al.
200 YIPF1 150 YIPF2 100 YIPF6 50 0 F2KD
Relative density (%)
F1KD
F1+2KD
F6KD
120 100 80 60 40 20 0
F1
F2 F1+2 F6
Fig. 7. Effect of YIPF1, YIPF2 and YIPF6 knockdown on Glycan synthesis. HT-29 cells were transfected with siRNAs and analyzed by western blotting as described in Fig. 4. (B) YIPF protein levels were quantified and the average of three independent replicates is shown with s.e.m. (error bars). (C) HT-29 cells were transfected with siRNAs and then fixed and processed for PAS staining. Representative images taken with a bright-field microscope are shown. Scale bar=100 µm. (D) Relative density of PAS staining for each sample. The average of three independent replicates is shown with s.e.m. (error bars). Asterisks indicate the statistical significance compared with the control ascertained using ANOVA (P < 0.01).
(#24112525) from MEXT of Japan, Kanazawa University (2008, #2023202) and Kyoto Sangyo University (Individual research expenses: 2010-2016) (for NN).
whole transmembrane segment of YIPF6 has been reported to cause swelling of the ER and deprivation of secretory granules in Paneth and goblet cells in Klein-Zschocher mutant mice [18]. It is possible that the expression of N-terminal cytoplasmic fragment of YIPF6, as opposed to that of null YIPF6, is responsible for the mutant phenotype. However, we have not been able to observe any significant effect on the Golgi apparatus by the expression of a N-terminal cytoplasmic fragment of YIPF6 in our experimental conditions (unpublished observations). In order to clarify the precise function of YIPF6 together with YIPF1 and YIPF2, more in vitro and in vivo analyses are necessary.
Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at doi:10.1016/j.yexcr.2017.03.011. References [1] X. Yang, H.T. Matern, D. Gallwitz, Specific binding to a novel and essential Golgi membrane protein (Yip1p) functionally links the transport GTPases Ypt1p and Ypt31p, EMBO J. 17 (1998) 4954–4963. [2] H. Matern, X. Yang, E. Andrulis, R. Sternglanz, H.-H. Trepte, D. Gallwitz, A novel Golgi membrane protein is part of a GTPase-binding protein complex involved in vesicle targeting, EMBO J. 19 (2000) 4485–4492. http://dx.doi.org/10.1093/ emboj/19.17.4485. [3] M. Heidtman, C.Z. Chen, R.N. Collins, C. Barlowe, A role for Yip1p in COPII vesicle biogenesis, J. Cell Biol. 163 (2003) 57–69. http://dx.doi.org/10.1083/ jcb.200306118. [4] J. Barrowman, W. Wang, Y. Zhang, S. Ferro-Novick, The Yip1p.Yif1p complex is required for the fusion competence of endoplasmic reticulum-derived vesicles, J. Biol. Chem. 278 (2003) 19878–19884. http://dx.doi.org/10.1074/
Acknowledgments We would like to thank Drs. Yoshio Misumi (Fukuoka Univ.) and Miwa Sohda (Niigata Univ.) for the anti-human GM130 antibody, Prof. Ken-ichi Sato (KSU) for donating HT-29 cells, Prof. Akira Kurosaka (KSU) for helpful comments and suggestions, Mr. Yu Kori, Mr. Ryoichi Murakami (KSU) for their contributions, and Prof. Jez Simpson (University College Dublin) for collaboration related to this work. This work was supported by Grants-in-Aid for Scientific Research (#25440092), and for Scientific Research on Priority Areas 107
Experimental Cell Research 353 (2017) 100–108
J. Soonthornsit et al.
1211–1228. [13] M. Sohda, Y. Misumi, A. Yano, N. Takami, Y. Ikehara, Phosphorylation of the vesicle docking protein p115 regulates its association with the Golgi membrane, J. Biol. Chem. 273 (1998) 5385–5388. [14] J. Soonthornsit, Y. Yamaguchi, D. Tamura, R. Ishida, Y. Nakakoji, S. Osako, et al., Low cytoplasmic pH reduces ER-Golgi trafficking and induces disassembly of the Golgi apparatus, Exp. Cell Res. 328 (2014) 325–339. http://dx.doi.org/10.1016/ j.yexcr.2014.09.009. [15] C. Steentoft, S.Y. Vakhrushev, H.J. Joshi, Y. Kong, M.B. Vester-Christensen, K.T.B.G. Schjoldager, et al., Precision mapping of the human O-GalNAc glycoproteome through SimpleCell technology, EMBO J. 32 (2013) 1478–1488. http://dx.doi.org/ 10.1038/emboj.2013.79. [16] N. Nakamura, C. Rabouille, R. Watson, T. Nilsson, N. Hui, P. Slusarewicz, et al., Characterization of a cis-Golgi matrix protein, GM130, J. Cell Biol. 131 (1995) 1715–1726. [17] J. Lippincott-Schwartz, L. Yuan, C. Tipper, M. Amherdt, L. Orci, R.D. Klausner, Brefeldin A's effects on endosomes, lysosomes, and the TGN suggest a general mechanism for regulating organelle structure and membrane traffic, Cell 67 (1991) 601–616. http://dx.doi.org/10.1016/0092-8674(91)90534-6. [18] K. Brandl, W. Tomisato, X. Li, C. Neppl, E. Pirie, W. Falk, et al., Yip1 domain family, member 6 (Yipf6) mutation induces spontaneous intestinal inflammation in mice, Proc. Natl. Acad. Sci. USA 109 (2012) 12650–12655. http://dx.doi.org/ 10.1073/pnas.1210366109. [19] G. Huet, I. Kim, C. de Bolos, J.M. Lo-Guidice, O. Moreau, B. Hemon, et al., Characterization of mucins and proteoglycans synthesized by a mucin-secreting HT-29 cell subpopulation, J. Cell Sci. 108 (Pt 3) (1995) 1275–1285.
jbc.M302406200. [5] C.Z. Chen, M. Calero, C.J. DeRegis, M. Heidtman, C. Barlowe, R.N. Collins, Genetic analysis of yeast Yip1p function reveals a requirement for Golgi-localized rab proteins and rab-Guanine nucleotide dissociation inhibitor, Genetics 168 (2004) 1827–1841. http://dx.doi.org/10.1534/genetics.104.032888. [6] A. Shakoori, G. Fujii, S.-I. Yoshimura, M. Kitamura, K. Nakayama, T. Ito, et al., Identification of a five-pass transmembrane protein family localizing in the Golgi apparatus and the ER, Biochem. Biophys. Res. Commun. 312 (2003) 850–857. http://dx.doi.org/10.1016/j.bbrc.2003.10.197. [7] T. Kranjc, E. Dempsey, G. Cagney, N. Nakamura, D.C. Shields, J.C. Simpson, Functional characterisation of the YIPF protein family in mammalian cells, Histochem. Cell Biol. (2016). [8] M. Calero, N.J. Winand, R.N. Collins, Identification of the novel proteins Yip4p and Yip5p as Rab GTPase interacting factors, FEBS Lett. 515 (2002) 89–98. [9] K. Tanimoto, K. Suzuki, E. Jokitalo, N. Sakai, T. Sakaguchi, D. Tamura, et al., Characterization of YIPF3 and YIPF4, cis-Golgi localizing Yip domain family proteins, Cell Struct. Funct. 36 (2011) 171–185. [10] Y. Yoshida, K. Suzuki, A. Yamamoto, N. Sakai, M. Bando, K. Tanimoto, et al., YIPF5 and YIF1A recycle between the ER and the Golgi apparatus and are involved in the maintenance of the Golgi structure, Exp. Cell Res. 314 (2008) 3427–3443. http:// dx.doi.org/10.1016/j.yexcr.2008.07.023. [11] R.E. Campbell, O. Tour, A.E. Palmer, P.A. Steinbach, G.S. Baird, D.A. Zacharias, et al., A monomeric red fluorescent protein, Proc. Natl. Acad. Sci. USA 99 (2002) 7877–7882. http://dx.doi.org/10.1073/pnas.082243699. [12] D.T. Shima, K. Halder, R. Pepperkok, R. Watson, G. Warren, Partitioning of the Golgi apparatus during mitosis in living HeLa cells, J. Cell Biol. 137 (1997)
108
Contents lists available at ScienceDirect
Experimental Cell Research journal homepage: www.elsevier.com/locate/yexcr
YIPF1, YIPF2, and YIPF6 are medial-/trans-Golgi and trans-Golgi network-localized Yip domain family proteins, which play a role in the Golgi reassembly and glycan synthesis
MARK
Jeerawat Soonthornsita,b, Noriko Sakaic, Yurika Sasakid, Ryota Watanabed, Shiho Osakod,e, ⁎ Nobuhiro Nakamuraa,c,d,e, a
Division of Engineering, Graduate School, Kyoto Sangyo University, Motoyama, Kamigamo, Kita, Kyoto 603-8555, Japan Department of Pre-clinic and Applied Animal Science, Faculty of Veterinary Science, Mahidol University, 999 Phutthamonthon Sai 4 Road Salaya, Phutthamonthon, Nakhon Pathom 73170 Thailand c Graduate School of Natural Science and Technology and School of Pharmacy, Kanazawa University, Kakuma, Kanazawa 920-1192, Japan d Department of Molecular Biosciences, Faculty of Life Sciences, Kyoto Sangyo University, Motoyama, Kamigamo, Kita, Kyoto 603-8555, Japan e Division of Life Sciences, Graduate School, Kyoto Sangyo University, Motoyama, Kamigamo, Kita, Kyoto 603-8555, Japan b
A R T I C L E I N F O
A BS T RAC T
Keywords: Golgi apparatus Vesicular transport Transmembrane protein Glycosylation
In this study, we attempted to explore the function of three uncharacterized mammalian homologs of yeast Yip domain family proteins—YIPF6, a homolog of Yip1p, and YIPF1 and YIPF2, which are homologs of Yif1p. Immunofluorescence staining revealed that YIPF1, YIPF2, and YIPF6 mainly localize in the medial-/transGolgi and also partially in the trans-Golgi network (TGN). On treatment with brefeldin A (BFA), the homologs co-migrated partly with medial-/trans-Golgi markers and also with a TGN marker in earlier time point, but finally redistributed within cytoplasmic punctate structures that were distinct from medial-/trans-Golgi and the TGN markers. YIPF6 formed a stable complex separately with YIPF1 and YIPF2, and knockdown of YIPF6 reduced YIPF1 and YIPF2 levels. These results suggest that YIPF6 forms complexes with YIPF1 and YIPF2 for their stable expression and localization within the Golgi apparatus. Knockdown experiments showed that YIPF1 and YIPF2, by contrast, are not necessary for the expression and localization of YIPF6. The structure of the Golgi apparatus and its disassembly after BFA treatment were not significantly affected by the knockdown of YIPF1, YIPF2, or YIPF6. However, reassembly of the Golgi apparatus after the removal of BFA was markedly delayed by the knockdown of YIPF1 and YIPF2, but not by that of YIPF6. These results strongly suggest that free YIPF6 after disassociating with YIPF1 and YIPF2 interferes with the reassembly of the Golgi apparatus. Knockdown of YIPF1 and YIPF2, but not that of YIPF6, also reduced intracellular glycans in HT-29 cells. Thus, we confirmed that YIPF1, YIPF2, and YIPF6 play a significant role in supporting normal glycan synthesis.
1. Introduction Yip domain family (YIPF) proteins are mammalian homologs of yeast Yip1p and Yif1p, which interact with Ypt family small GTPases, including Ypt1p and Ypt31p [1]. Yip1p and Yif1p form a complex with each other and are found in the ER, Golgi apparatus, and COPII vesicles. They were hypothesized to play a role in vesicle budding from the ER and/or fusion with the Golgi apparatus under regulation by Ypt proteins [1–4]. This was supported by the findings that Yip1p interacts with ER to Golgi SNAREs (Bos1p and Sec22p) [4] and YIP1 genetically interacts with genes coding an Rab-GDI (GDI1), COPII components
(SEC12, SEC13, and SEC23), a COPI component (SEC21) and a Golgi tethering protein (USO1) genes [3,5]. Both Yip1p and Yif1p are predicted to have five transmembrane segments; an N-terminal segment, which is exposed to the cytoplasm; and a short C-terminal segment, which is exposed to the lumen of the secretory pathway [6,7]. Yip1p and Yif1p show significant similarity at the transmembrane segments (Yip domain); two other non-essential proteins, Yip4p and Yip5p, also share the Yip domain [8]. Homologs of Yip1p and Yif1p are found in most eukaryotes, including fungi, protozoa, plants, and animals, indicating their essential role in cell functioning. However, the precise mode of their function remains
Abbreviations: YIPF, Yip domain family; ERGIC, ER-Golgi intermediate compartment; TGN, trans-Golgi network; BFA, brefeldin A; GnT-I, N-acetylglucosaminyltransferase-I; RFP, red fluorescent protein; GalT-I, β4-galactosyltransferase-I; PAS staining, Periodic acid-Schiff staining ⁎ Corresponding author at: Department of Molecular Biosciences, Faculty of Life Sciences, Kyoto Sangyo University, Motoyama, Kamigamo, Kita, Kyoto 603-8555, Japan. E-mail address: [email protected] (N. Nakamura). http://dx.doi.org/10.1016/j.yexcr.2017.03.011 Received 28 December 2016; Received in revised form 6 March 2017; Accepted 7 March 2017 Available online 09 March 2017 0014-4827/ © 2017 Elsevier Inc. All rights reserved.
Experimental Cell Research 353 (2017) 100–108
J. Soonthornsit et al.
2.4. Pull-down analysis
obscure [9]. Nine YIPF members are found in human cells [6,10]. Phylogenetic analyses have previously shown that YIPF4, YIPF5/YIP1A, YIPF6, and YIPF7/YIP1B are Yip1p homologs, while YIPF1, YIPF2, YIPF3, YIF1A, and YIF1B are Yif1p homologs (Fig. S1) [6]. While characterizing the human homologs, we previously found that each human Yip1p homolog (YIPF5 and YIPF4) works with a specific Yif1p homolog partner (YIF1A and YIPF3, respectively) [9,10]. Cellular distribution of each pair is different—YIPF5 and YIF1A mainly localize within the ER-Golgi intermediate compartment (ERGIC), while YIPF4 and YIPF3 mainly localize within the cis-Golgi. It is possible that each pair of YIPF proteins functions independently in their residing compartment by associating with a different set of proteins, in addition to functioning together in these compartments. Knockdown of Yip1p homologs (YIPF4 or YIPF5) causes marked reduction in the levels of their partner Yif1p homologs (YIPF3 or YIF1A), suggesting that Yif1p homologs are unstable in the absence of Yip1p homologs. The knockdown of YIPF4, YIPF3, YIPF5, and YIF1A causes significant fragmentation of the Golgi apparatus, highlighting their role in maintaining the Golgi structure. Here, we report that the Yip1p homolog YIPF6 forms complexes with the Yif1p homologs YIPF1 and YIPF2, and they localize in the medial-/trans-Golgi and TGN. Gene knockdown experiments showed that YIPF6, free of YIPF1 and YIPF2, interferes with the reorganization of the Golgi apparatus and glycan synthesis.
Pull-down analysis was performed as described previously [10]. Briefly, HeLa cells were incubated in a buffer containing digitonin and centrifuged to obtain a soluble extract. A part of the extract was kept as a sample of the extract, and the remaining extract was separated into two equal parts. Affinity-purified antibody for a YIPF protein was added to one part of the extract and affinity-purified IgG was added to another part. Protein A Sepharose CL-4B (GE Healthcare Bio- Science AB, Uppsala, Sweden) was used to pull down antibodies and bound materials. 2.5. SDS-PAGE, western blotting, and densitometry SDS-PAGE and western blotting were performed as described previously [10]. Chemiluminescence images were taken using LAS1000 and LAS4000mini (Fuji Photo Film. Inc., Tokyo, Japan). Densitometry was performed using ImageJ (National Institute of Health, Bethesda, USA). 2.6. siRNA transfection siRNA transfection was performed as described previously [10]. Briefly, HeLa cells or HT-29 cells were transfected with Stealth RNAi™ siRNAs (YIPF1: HSS122872, YIPF2: HSS149042, and YIPF6: equal amount mixture of HSS138898, HS138899, and HS138900, Invitrogen, Thermo Fisher Scientific Inc., Waltham, MA USA) by using Lipofectoamine 2000 and incubated for three days with a daily change in the medium before analyses. Silencer® Select Negative Control No. 1 siRNA (Thermo Fisher Scientific Inc.) was used as control.
2. Materials and methods 2.1. Cell culture, drug treatment, immunofluorescence staining, cell extraction, and immunoprecipitation These experiments were performed as described previously [10]. For examining the Golgi reassembly, HeLa cells were treated with a growth medium (Dulbecco's modified Eagles medium containing 10% fetal bovine serum) containing 0.25 µg/mL BFA for one hour to four hours. Then, the BFA containing medium was removed and the cells were washed with the growth medium for three times, and were incubated in the growth medium for one hour. HT-29 cells were originally purchased from ATCC and were kindly donated by Profs. A. Kurosaka and K. Sato (Kyoto Sangyo University) and cultured with the growth medium.
2.7. Immunofluorescent staining This was performed as described previously [14]. 2.8. Periodic acid-Schiff (PAS) staining 2×105 HT-29 cells were seeded on to coverslips in petri dishes 3.5 cm in diameter and were transfected with siRNA as described above. The cells were then stained using the PAS Kit (Sigma-Aldrich Co. LLC., St. Louis, MO, USA) according to the manufacturer's instructions, but with slight modifications. Briefly, the cells were fixed with a formalin–ethanol fixative solution (4% formaldehyde in 95% ethanol) and gently washed three times with tap water. The cells were then treated with the PAS solution for 5 min and washed five times with distilled water. Finally, the cells were treated with the Schiff reagent for 15 min and washed 10 times with tap water. The coverslips, containing the stained cells, were mounted in glycerol and observed under light microscope. Counter-staining with hematoxylin was omitted for the following image analysis. Micrographs were taken using a bright-field microscope with identical illumination and exposure settings for all samples. For each sample, 6–8 images were used for quantification with ImageJ. Two areas of cell clusters were selected from an image by the free-hand selection mode and two cell-free areas were selected from the same image. Mean intensity of an area was divided by pixel number of the area to estimate the area's cell density. The average density of two cell areas was subtracted by that of the two cell-free areas to obtain a specific staining density of an image. Finally, an average specific staining density of 6–8 images was calculated for each sample. For each experiment, the average specific densities of YIPF knocked-down cells were divided by that of negative control transfected cells to ascertain relative staining density of the cells. Statistical analysis was performed using StatMate V (ATMS Co., Ltd, Tokyo, Japan).
2.2. Plasmids and transfection A plasmid encoding N-acetylglucosaminyltransferase-I (GnT-I) tagged with red fluorescent protein (GnT-I-RFP) was produced by conjugating a cDNA encoding the monomeric red fluorescent protein [11] to the 3′-end of the GnT-I cDNA encoding GnT-I [12], and was inserted into pcDNA3 (Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA). Another plasmid encoding β4-galactosyltransferase-I (GalT-I) tagged with RFP (GalT-I-RFP) was purchased from Evrogen (pTagRFP-Golgi; Moscow, Russia). 2.3. Antibodies Rabbit polyclonal antibodies for YIPF1, YIPF2, and YIPF6 were produced and affinity-purified as described previously by using recombinant proteins [10]. Rabbit anti-human GM130 (GST-human GM130) was kindly donated by Drs. Sohda (Niigata University) and Misumi (Fukuoka University) [13]. The following antibodies were also purchased—mouse monoclonal anti-GM130 antibody (BD Biosciences, San Diego, CA, USA), CY3-conjugated anti-rabbit IgG, CY5-conjugated anti-rabbit IgG, CY5-conjugated anti-mouse IgG (Jackson ImmunoResearch Labs. Inc., West Grove, PA, USA), Alexa488-conjugated anti-mouse IgG (Molecular Probes, Thermo Fisher Scientific), and HRP-conjugated anti-rabbit IgG (Jackson ImmunoResearch). 101
Experimental Cell Research 353 (2017) 100–108
J. Soonthornsit et al.
Control YIPF1
YIPF6
YIPF2
B YIPF1
A
200 116 97.4 66
YIPF1
GnT-I
GM130
YIPF2
45 * 31
YIPF2
GM130
YIPF6
21.5 YIPF6
GM130
C
YIPF1 ++ - ooo o o 255 RRVALATIVTIVLLHMLLSVGCLAYFFDAPEMDHLPTTTATPNQTV AAAKSS
BFA 15min YIPF1
YIPF2
GnT-I
+ + oo o o o 255 TRLVATVLLSVVVLLHALLAMGCKLYFFQSLPPENVAPPPQITSLPSNIALSPTLPQSLAPS YIPF6 ++ 211 RRALAVYPVFLFYFVISWMILTFTPQ
YIPF2
Fig. 1. Characterization of anti-YIPF1, YIPF2, and YIPF6 antibodies. HeLa cells were analyzed by western blotting using anti-YIPF1, anti-YIPF2, and anti-YIPF6 antibodies. Molecular-size markers (kDa) are indicated on the left. Arrowheads on the right indicate the specific bands corresponding to each protein. An asterisk on the right of the YIPF2 line indicates a non-specific band. (B) HeLa cells were analyzed by double immunofluorescent staining using anti-YIPF1, anti-YIPF2, and anti-YIPF6 antibodies (CY3) together with an anti-GM130 antibody (Alexa488). Images taken using an epifluorescence microscope are shown. Scale bar=10 µm. (C) The predicted fifth transmembrane and cytoplasmic segments of YIPF1, YIPF2, and YIPF6. +, positively charged residue; −, negatively charged residue; o, predicted O-glycosylation sites; double underlines, predicted transmembrane segments. Position numbers of the first shown residue are indicated on the left.
YIPF6
3. Results
BFA 30min
3.1. Characterization of YIPF1, YIPF2, and YIPF6
YIPF1
Affinity-purified rabbit polyclonal antibodies were produced using recombinant YIPF1, YIPF2, and YIPF6 proteins. When HeLa cell lysate was analyzed using these antibodies with western blotting (Fig. 1A), a clear single band was detected for YIPF1 and YIPF6 antibodies and two bands were detectable for the YIPF2 antibody. A higher-mobility band detectable for YIPF2 turned out to be a non-specific reaction product, as a lower-mobility band was reduced by the knockdown of YIPF2 (Fig. 4 and S4). When these antibodies were used for immunofluorescent staining on HeLa cells, all antibodies stained juxtanuclear ribbon-like structures, closely localizing with GM130, a cis-Golgi marker protein (Fig. 1B). Specific staining of these structures was supported by the fact that the knockdown of YIPF1, YIPF2, and YIPF6 caused reduced staining (Fig. 5). Therefore, we concluded that these affinity-purified antibodies had adequate quality for further analyses. Molecular size of YIPF6 was predicted to be 26 kDa based on its amino acid sequence, which was identical to the molecular size of YIPF6 estimated using western blotting. Therefore, we deduced that YIPF6 did not undergo a major post-translational modification. YIPF6 has been previously predicted to have a short cytoplasmic segment with no glycosylation sites (Fig. 1C) [15]. Molecular sizes of YIPF1 and YIPF2 were predicted to be 35 and 34 kDa, respectively, based on their amino acid sequence, and their molecular sizes estimated using western blotting were 40 and 43 kDa, respectively. It was thus evident that YIPF1 and YIPF2 were post-translationally modified. This was
GnT-I
YIPF2
YIPF6
Fig. 2. Localization of YIPF1, YIPF2, and YIPF6. HeLa cells transiently expressing GnT-I-RFP; untreated (left gallery) and treated with BFA for 15 min (middle gallery) and 30 min (right gallery). The cells were then fixed and processed for immunofluorescence staining using anti-YIPF1 (upper panels), anti-YIPF2 (middle panels), and anti-YIPF6 (lower panels) antibodies (Alexa488), and were observed with fluorescence of RFP. Images were taken with a confocal microscope, and representative sliced images are shown. Scale bar=5 µm.
102
Experimental Cell Research 353 (2017) 100–108
J. Soonthornsit et al.
Pull down
likely attributable to O-glycosylation, as was the case with YIPF3 [9], considering that YIPF1 and YIPF2 are predicted to have five putative O-glycosylation sites in their cytoplasmic segments (Fig. 1C). No other glycosylation sites were detectable for YIPF1, YIPF2, and YIPF6.
E F1 C
E F2 C
E F6 C
Detection
YIPF1
3.2. YIPF1, YIPF2, and YIPF6 were localized in the medial-/transGolgi and TGN, but showed distinct dynamics from these compartment markers Fine localization of YIPF1, YIPF2, and YIPF6 was further analyzed using double staining with YIPF proteins and Golgi compartment markers—medial-Golgi, N-acetylglucosaminyltransferase-I tagged with red fluorescent protein (GnT-I-RFP); trans-Golgi, β4-galactosyltransferase-I tagged with RFP (GalT-I-RFP); TGN, TGN46 (Fig. 2, S2 and S3). Close inspection with confocal microscopy showed that YIPF1, YIPF2, and YIPF6 mainly localized within medial-/trans-Golgi and TGN, while less so within cis-Golgi, as deduced from the marker GM130 (Fig. 2 and S3; control). When HeLa cells were treated with brefeldin A (BFA), the bulk of medial-/trans-Golgi proteins relocated in the ER as evidenced by the localization in the nuclear membrane, whereas most of cis-Golgi proteins were retained within cytoplasmic punctate structures called “Golgi remnants,” which are distinct from the ER (Fig. S2, BFA 30 min) [9,16]. A TGN marker, TGN46, was also distributed distinctly from the cis-, medial- and trans-Golgi markers, consistent with the previous finding that TGN proteins are relocated in the endosomal compartment [17]. During BFA treatment, YIPF1, YIPF2, and YIPF6 showed prominent redistribution patterns that were different form the medial- and trans-Golgi marker proteins (Fig. 2 and S3; BFA), despite the former's close localization with the latter in non-treated cells. YIPF1, YIPF2, and YIPF6 were frequently observed along tubular intermediate structures stained with the medial- (Fig. 2; BFA; open arrows), trans-Golgi and also TGN markers after BFA treatment (Fig. S3; BFA; open arrows). After 30 min of BFA treatment, YIPF1, YIPF2, and YIPF6 mostly relocalized to the cytoplasmic punctate structures, with occasional colocalization with the medial-/trans-Golgi and TGN markers (arrows and open arrows). Strikingly, YIPF1, YIPF2, and YIPF6 was not found in the nuclear membrane and cytoplasmic reticular structures, which indicate the ER localization, in contrast with medial-/trans-Golgi markers (Fig. 2 and S3, open arrowheads). Therefore, it is likely that YIPF1, YIPF2, and YIPF6 were not retrieved back to the ER but mainly retained within the cytoplasmic structures similar to but distinct from Golgi remnants after BFA treatment.
YIPF2 YIPF6 YIPF4 YIPF5
Fig. 3. Complex formation of YIPF1 and YIPF2 with YIPF6. HeLa cell extract was pulled down with anti-YIPF1 (F1), anti-YIPF2 (F2), anti-YIPF6 (F6), and control IgG (C), as indicated on top of the panels. The precipitated material (F1, F2, F6, and C) were analyzed using western blotting with anti-YIPF1, anti-YIPF2, anti-YIPF6, anti-YIPF4, and anti-YIPF5 antibodies, as indicated in the left panel. A part (1/3.25) of the cell extract (E) was also analyzed as an input control.
3.4. Effects of YIPF1, YIPF2, and YIPF6 knockdown
F1+2 KD
F6 KD
F2 KD
Control
A
F1 KD
To analyze the function of YIPF1, YIPF2, and YIPF6, knockdown experiments were performed using pre-designed StealthRNAi siRNA (Thermo Fisher Scientific Inc.). Three available siRNAs for each YIPF proteins were used individually or in a mixture to knockdown the target proteins in HeLa cells. The result of western blotting showed that the siRNAs specifically and efficiently reduced the target proteins in similar levels after three days of transfection in all the conditions (Fig. 4 and S4 and unpublished observation), although slight cytotoxic effects, e.g. cell growth delay, cell detachment and cell rounding, were observed in
YIPF1 YIPF2
* YIPF6
B
3.3. YIPF6 forms a distinct complex with either YIPF1 or YIPF2
Relative intensity (%)
120
To analyze whether YIPF1 and YIPF2 formed a complex with YIPF6, pull-down assays were performed. HeLa cells were extracted by the digitonin-containing buffer, and YIPF1, YIPF2, and YIPF6 were pulled down from the extract using affinity-purified antibodies for each protein. The pulled-down proteins were then analyzed using western blotting with anti-YIPF antibodies. As seen in Fig. 3, each YIPF protein was efficiently pulled down by the corresponding antibody (F1, F2, and F6), but not by the control IgG (C). YIPF4 and YIPF5, which are Yip1p homologs localizing within cis-Golgi and ERGIC, respectively, were not pulled down by any of the antibodies, which confirmed that the pulldown assays were specific. The anti-YIPF6 antibody precipitated YIPF1 and YIPF2, while anti-YIPF1 and anti-YIPF2 antibodies precipitated the YIPF6 protein. This indicated that most YIPF1 and YIPF2 formed a complex with YIPF6. By contrast, YIPF2 was poorly pulled down by the anti-YIPF1 antibody, while YIPF1 was poorly pulled down by antiYIPF2 antibody. This result suggested that most YIPF1 or YIPF2 formed distinct complexes with YIPF6, and a ternary complex of YIPF1–YIPF2–YIPF6 was present in small quantity.
YIPF1
100
YIPF2
80
YIPF6
60 40 20 0
F1 KD
F2 KD F1+2 KD F6 KD
Fig. 4. Protein levels of YIPF1, YIPF2, and YIPF6 after knockdown. (A) HeLa cells were transfected with siRNAs targeting YIPF1 (F1 KD), YIPF2 (F2 KD), both YIPF1 and YIPF2 (F1+F2 KD), YIPF6 (F6 KD), and control double-stranded RNA (Control). The cells were then lysed and analyzed by western blotting using anti-YIPF1, anti-YIPF2, and anti-YIPF6 antibodies. Representative result of three independent replicates is shown. (B) YIPF protein levels were quantified using densitometry, and the relative amount of each protein compared with the control was shown. The average of three independent replicates is shown with s.e.m. (error bars).
103
Experimental Cell Research 353 (2017) 100–108
J. Soonthornsit et al.
extended ribbon-like structure in about the half of the cells ( > 40%) after two hours (W/O 2 h; F1 KD, F2 KD, and F1+2 KD). After four hours, the Golgi apparatus was reassembled into a compact ribbon-like structure in most of the cells (W/O 4 h; F1 KD, F2 KD, and F1+2 KD; ~80%). These results indicated that the Golgi reassembly was severely delayed in YIPF1 or YIPF2 knockdown cells. Surprisingly, the Golgi apparatus was reassembled almost normally in YIPF6 knocked-down cells at all the time points after the removal of BFA (F6 KD). Because YIPF6 knockdown caused a simultaneous reduction in YIPF1 and YIPF2 levels (Fig. 4 and S4), it is likely that the reduction in YIPF1 and YIPF2 per se did not inhibit the Golgi reassembly. Instead, the presence of free YIPF6, which lost the partner proteins YIPF1 and/or YIPF2, was likely the cause of the hampered Golgi reassembly. Whether such a hampered reassembly affects the function of the Golgi apparatus was further explored. Homozygous truncation mutation in YIPF6 has been previously shown to cause a defective formation and secretion of large secretory granules containing mucin from Paneth and goblet cells [18]. Therefore, we looked for model cell lines that can simulate intestinal granule secretion. Among the several cell lines tested, HT-29, a human colon carcinoma cell line [19], was suitable for our purpose because it accumulates glycans at high levels even under a growing culture condition in non-differentiated state. Western blotting analysis showed that the knockdown was successful for all YIPF proteins; however, the knockdown efficiency was slightly lower for HT-29 cells compared with that of HeLa cells (Fig. 7). Similar to HeLa cells, YIPF6 knockdown reduced YIPF1 and YIPF2 levels, while those of YIPF6 remained unchanged following YIPF1 and YIPF2 knockdown in HT-29 cells (Fig. 7A and B). The cells were then analyzed with PAS staining to check the quantity of glycans accumulated in the cells, and the staining was quantified using image analysis. As seen in Fig. 7C, YIPF1, YIPF2, and YIPF1–YIPF2 double knocked-down cells showed fainter staining, while that across YIPF6 knocked-down cells was more or less similar to the control cells. Quantitative analysis showed that the staining intensity of YIPF2 and YIPF1–YIPF2 double knockeddown cells was lower compared to the control cells; the difference was statistically significant (P < 0.01) (Fig. 7D). The staining intensity also tended to be lower for YIPF1 knocked-down cells, but was not statistically significant. The staining intensity of YIPF6 knocked-down cells was not different from that of the control cells. These results suggest that the knockdown of YIPF2—and possibly of YIPF1, but not of YIPF6—reduced the amount of intracellular glycans in HT-29 cells.
some conditions. These were strongly suggested to be off-target effects because they were not common for all the three siRNA or a mixture, which knocked down the target proteins in similar levels. Therefore, one of the three siRNAs for YIPF1 and YIPF2 while a mixture of the three siRNAs for YIPF6, which showed least off-target effects, were used for further analyses. Reduction in YIPF6 levels caused a severe reduction in those of YIPF1 and YIPF2 (Fig. 4, F6KD). This effect was not an off-target effect because all the three siRNAs reduced YIPF1 and YIPF2 to a similar level (Fig. S4). However, YIPF6 levels were not reduced due to the knockdown of YIPF1 (F1KD), but were slightly reduced due to that of YIPF2 (F2KD) along with a simultaneous knockdown of YIPF1 and YIPF2 (F1+2KD). In addition, YIPF1 levels remained unaffected by the knockdown of YIPF2 (F2KD) and YIPF2 levels reduced slightly after the knockdown of YIPF1 (F1KD). These results suggested that the presence of YIPF6, a Yip1p homolog, was essential for the stable expression of YIPF1 and YIPF2, Yif1p homologs, while YIPF6 expression was independent of YIPF1 and YIPF2. In addition, the expression of YIPF1 and YIPF2 was mostly independent of each other. These results were consistent with the previous finding that the stable expression of Yif1p homolog (YIPF3 and YIF1A) was dependent on the expression of the partner Yip1p homolog (YIPF4 and YIPF5, respectively) while the expression of Yip1p homolog was independent from the expression of the partner Yif1p homolog [9,10]. Immunofluorescence staining was performed to analyze whether the knockdown of YIPF1, YIPF2, or YIPF6 affected the morphology of the Golgi apparatus. siRNA-transfected cells were double stained with each of the YIPF and Golgi marker proteins (Fig. 5). Overall, staining of the YIPF proteins was consistent with the results of western blotting (Fig. 4). Fluorescence of YIPF proteins was markedly reduced owing to the knockdown of corresponding YIPF proteins. The fluorescence of YIPF1 was not significantly reduced by the knockdown of YIPF2 and vice-versa. Knockdown of YIPF6 caused a strong reduction in the fluorescence of YIPF1 and YIPF2, while knockdown of YIPF1 and YIPF2, individually or simultaneously, did not significantly affect YIPF6 staining, which agrees with the western blotting result. Therefore, we concluded that the localization of YIPF6 in the Golgi apparatus is not affected by the knockdown of YIPF1 or YIPF2. No considerable morphological changes were observed for cis-Golgi (GM130) (Fig. 5), medial-/trans-Golgi (giantin), and TGN (TGN46) (Fig. S6) following the knockdown of the YIPF proteins. These results suggested that the Golgi structure is resistant to the knockdown of YIPF1, YIPF2 or YIPF6 under normal cell culture conditions. Knockdown of YIPF1, YIPF2 or YIPF6 did not affect staining intensity and distribution of YIPF4 or YIPF5 (Fig. S5) further supporting that YIPF1, YIPF2 and YIPF6 were specifically knocked down. To explore the function of YIPF proteins, we analyzed whether the knockdown of YIPF proteins has any effect on the Golgi apparatus by using treatments that reversibly perturb the structure of the Golgi apparatus. We tried several drugs and conditions and found that knockdown of YIPF proteins affects the recovery from BFA induced Golgi disassembly. HeLa cells were treated with BFA for one hour and then washed and incubated with a growth medium up to four hours (Fig. 6A, B and S6). The Golgi apparatus was reassembled into a continuous ribbon-like structure in the majority ( > 70%) of the cells transfected with control siRNA one hour after the removal of BFA (W/ O 1 h, Cont). The number of the cells showing reassembled Golgi increased ( > 80%) after two hours (W/O 2 h), and it was finally restored to the level before BFA treatment (~90%) after four hours (compare N/T and W/O 4 h). In contrast, the Golgi apparatus became a bigger and brighter punctate structures spread in the cytoplasm but was not reassembled into a ribbon-like structure in most of the cells ( > 95%), which were either individually or simultaneously knocked-down YIPF1 or YIPF2, one hour after the removal of BFA (W/O 1 h; F1 KD, F2 KD, and F1+2 KD). The Golgi apparatus was then reassembled into a slightly
4. Discussion 4.1. Localization of YIPF1, YIPF2, and YIPF6 Immunofluorescent staining results strongly suggested that YIPF1, YIPF2, and YIPF6 mainly localized within medial-/trans-Golgi compartments and partly within the TGN. These results are consistent with the observation that transiently expressed YFP-tagged YIPF1, YIPF2, and YIPF6 are enriched in trans-Golgi compartments [7]. While most of the medial-/trans-Golgi proteins relocated to the ER following BFA treatment, YIPF1, YIPF2, and YIPF6 remained mainly within the cytoplasmic punctate structures. These structures were largely independent of the structures containing cis-Golgi and TGN markers. cisGolgi proteins including syntaxin5, GM130, YIPF3 and YIPF4 have been reported to relocalize to the structures called “Golgi remnants,” which overlap with ERGIC [9,16], while TGN proteins relocalize to structures overlapping with endosomes [17]. Thus, YIPF1, YIPF2, and YIPF6 mainly relocalized to the compartments independent of the cisGolgi/ERGIC and TGN/endosomes. It is unlikely that YIPF1, YIPF2, and YIPF6 were transported back to and re-exited the ER during recovery from BFA treatment to reach the compartments separate from the cis-Golgi/ERGIC or TGN/endosomes, considering that YIPF1, YIPF2, and YIPF6 did not show a significant overlap with ERGIC53 104
Experimental Cell Research 353 (2017) 100–108
J. Soonthornsit et al.
Control
YIPF1 KD
YIPF2 KD
YIPF1+2 KD
YIPF6 KD
YIPF1
GM130
YIPF2
GM130
YIPF6
GM130
Fig. 5. Distribution of YIPF1, YIPF2, and YIPF6 after knockdown. HeLa cells were transfected with siRNA as described in Fig. 4 and processed for double immunofluorescence staining using anti-YIPF1, anti-YIPF2, and anti-YIPF6 antibodies (CY3) together with anti-GM130 antibody (Alexa488). Images taken by epi-fluorescence microscope are shown. Scale bar=20 µm.
stable expression of Yif1p homologs. Immunoprecipitation experiments suggested that YIPF1 and YIPF2 formed a distinct complex with YIPF6 and that a ternary complex existed in a small quantity (Fig. 3). Preliminary experiments suggested that YIPF6 formed a hetero-oligomer (4–16 mer) with either YIPF1 or YIPF2, which has been previously observed for YIPF5–YIF1A and YIPF4–YIPF3 [9,10]. Although it is probable that YIPF1–YIPF6 and YIPF2–YIPF6 form independent complexes despite having YIPF6 as a common subunit, it remains to be seen whether these two complexes have a distinct function.
(unpublished observation) or GM130 (Fig. S3). Therefore, it is likely that majority of YIPF1, YIPF2, and YIPF6 resisted the retrograde flow and remained within the cytoplasmic structures that are similar to but not the same with the Golgi remnants. Detailed difference in dynamics between YIPF1, YIPF2, YIPF6 and other medial-/trans-Golgi proteins has to be clarified in future. 4.2. YIPF1 and YIPF2 stabilize by forming a complex with YIPF6 As is the case for other YIPF members, YIPF6, a Yip1 homolog, formed a complex with YIPF1 and YIPF2, Yif1p homologs (Fig. 3), and the expression of YIPF6 was essential for the stable expression of YIPF1 and YIPF2 (Figs. 4 and 7A and B). YIPF1, YIPF2, and YIPF6 did not interact with YIPF4 or YIPF5 (Fig. 3), suggesting that each Yip1p homolog forms a complex with a predesignated Yif1p homolog— YIPF5/YIP1A with YIF1A, YIPF4 with YIPF3, and YIPF6 with YIPF1 and YIPF2. For all these pairs, the expression of Yip1p homologs is necessary for the stable expression of Yif1p homologs. This supports the hypothesis that the expression of Yip1p homologs is vital for the
4.3. Absence of YIPF1 or YIPF2 inhibits the reassembly of the Golgi apparatus and reduces intracellular glycans Although knockdown of YIPF1 or YIPF2 induced no significant effect on the Golgi morphology under normal culture condition, it inhibited the Golgi reassembly after the removal of BFA (Fig. 6 and S6). This suggested that a minimal function of the cells pertaining to the Golgi assembly was retained under reduced YIPF1 or YIPF2 level. It is 105
Experimental Cell Research 353 (2017) 100–108
J. Soonthornsit et al.
A N/T
Cont
F1 KD
F2 KD
F1+2 KD
F6 KD
BFA
W/O 1h
W/O 2h
W/O 4h
B
100% 80% Cont F1 KD F2 KD F1+2 KD F6 KD
60% 40% 20% 0%
Cont
BFA
W/O 1h
W/O 2h
W/O 4h
Fig. 6. Effect of YIPF1, YIPF2 and YIPF6 knockdown on Golgi reassembly after BFA treatment. (A) HeLa cells were transfected with siRNA as described in Fig. 4 (N/T). The cells were then treated with BFA for 1 h (BFA) and control cells were further washed and incubated with normal medium for 1 h, 2 h or 4 h (W/O 1 h, W/O 2 h and W/O 4 h). The cells were fixed and processed for triple immunofluorescence staining using anti-GM130 (CY3), anti-giantin (CY5), and anti-TGN46 (Alexa488) antibodies. Representative images taken by an epi-fluorescence microscope are shown; however, only GM130 staining has been shown for simplicity and space limitation. Full sets of images are shown in Fig. S6. Scale bar=20 µm. (B) Relative numbers of the cells showing a ribbon-like structure by GM130 staining in (A) were quantified. Examples of the categorized cells are shown in Fig. S6. The average of three independent replicates is shown with s.e.m. (error bars).
glycans were also reduced in YIPF1 and YIPF2 knocked-down HT-29 cells (Fig. 7). This was attributable to either reduced glycan synthesis or enhanced glycan secretion. Enhanced glycan secretion is less likely to be the underlying cause because our preliminary experiment showed that the transport of VSV-G, a protein widely used as a secretorypathway marker, was not much affected by YIFP knockdown. The Golgi apparatus is the main compartment wherein O-linked glycan chains are synthesized. Therefore, it likely that glycan synthesis was reduced in the YIPF1 or YIPF2 knocked-down cells. It was probable that the presence of free YIPF6 caused incomplete or delayed recycling of Golgi enzymes inducing some defects in the glycan synthesis machinery while the structure of the Golgi apparatus was maintained somehow. A peculiar observation of our present study is that YIPF6 knockdown did not affect our analysis, considering that the deletion of the
probable that this function was strained with the massive need of reassembly after BFA removal. This hypothesis is supported by the observation that the Golgi apparatus was reassembled gradually, taking longer incubation times (~4 h) after BFA removal (Fig. 6 and S6). The Golgi localization of YIPF6 remained largely unaffected by the reduction in YIPF1 and YIPF2 levels (Fig. 5). This indicates that YIPF6 exists within the Golgi free of partner proteins. Because Yip1p/Yif1p has been proposed to be involved in vesicle fusion in the Golgi apparatus, it is probable that the YIPF1/YIPF2–YIPF6 complexes play a role in membrane fusion within the Golgi apparatus. Free YIPF6 may inhibit the fusion of the Golgi apparatus and associated transport vesicles during its reassembly after BFA removal, thereby leading to an incomplete reassembly in YIPF1 and YIPF2 knocked-down cells. Corresponding to the hampered Golgi reassembly, intracellular 106
Experimental Cell Research 353 (2017) 100–108
YIPF2 YIPF6
Y I P F1 + 2 K D
Y I P F6 K D
Y I P F2 K D
Relative intensity (%)
YIPF1
YIPF1 K D
C o nt
J. Soonthornsit et al.
200 YIPF1 150 YIPF2 100 YIPF6 50 0 F2KD
Relative density (%)
F1KD
F1+2KD
F6KD
120 100 80 60 40 20 0
F1
F2 F1+2 F6
Fig. 7. Effect of YIPF1, YIPF2 and YIPF6 knockdown on Glycan synthesis. HT-29 cells were transfected with siRNAs and analyzed by western blotting as described in Fig. 4. (B) YIPF protein levels were quantified and the average of three independent replicates is shown with s.e.m. (error bars). (C) HT-29 cells were transfected with siRNAs and then fixed and processed for PAS staining. Representative images taken with a bright-field microscope are shown. Scale bar=100 µm. (D) Relative density of PAS staining for each sample. The average of three independent replicates is shown with s.e.m. (error bars). Asterisks indicate the statistical significance compared with the control ascertained using ANOVA (P < 0.01).
(#24112525) from MEXT of Japan, Kanazawa University (2008, #2023202) and Kyoto Sangyo University (Individual research expenses: 2010-2016) (for NN).
whole transmembrane segment of YIPF6 has been reported to cause swelling of the ER and deprivation of secretory granules in Paneth and goblet cells in Klein-Zschocher mutant mice [18]. It is possible that the expression of N-terminal cytoplasmic fragment of YIPF6, as opposed to that of null YIPF6, is responsible for the mutant phenotype. However, we have not been able to observe any significant effect on the Golgi apparatus by the expression of a N-terminal cytoplasmic fragment of YIPF6 in our experimental conditions (unpublished observations). In order to clarify the precise function of YIPF6 together with YIPF1 and YIPF2, more in vitro and in vivo analyses are necessary.
Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at doi:10.1016/j.yexcr.2017.03.011. References [1] X. Yang, H.T. Matern, D. Gallwitz, Specific binding to a novel and essential Golgi membrane protein (Yip1p) functionally links the transport GTPases Ypt1p and Ypt31p, EMBO J. 17 (1998) 4954–4963. [2] H. Matern, X. Yang, E. Andrulis, R. Sternglanz, H.-H. Trepte, D. Gallwitz, A novel Golgi membrane protein is part of a GTPase-binding protein complex involved in vesicle targeting, EMBO J. 19 (2000) 4485–4492. http://dx.doi.org/10.1093/ emboj/19.17.4485. [3] M. Heidtman, C.Z. Chen, R.N. Collins, C. Barlowe, A role for Yip1p in COPII vesicle biogenesis, J. Cell Biol. 163 (2003) 57–69. http://dx.doi.org/10.1083/ jcb.200306118. [4] J. Barrowman, W. Wang, Y. Zhang, S. Ferro-Novick, The Yip1p.Yif1p complex is required for the fusion competence of endoplasmic reticulum-derived vesicles, J. Biol. Chem. 278 (2003) 19878–19884. http://dx.doi.org/10.1074/
Acknowledgments We would like to thank Drs. Yoshio Misumi (Fukuoka Univ.) and Miwa Sohda (Niigata Univ.) for the anti-human GM130 antibody, Prof. Ken-ichi Sato (KSU) for donating HT-29 cells, Prof. Akira Kurosaka (KSU) for helpful comments and suggestions, Mr. Yu Kori, Mr. Ryoichi Murakami (KSU) for their contributions, and Prof. Jez Simpson (University College Dublin) for collaboration related to this work. This work was supported by Grants-in-Aid for Scientific Research (#25440092), and for Scientific Research on Priority Areas 107
Experimental Cell Research 353 (2017) 100–108
J. Soonthornsit et al.
1211–1228. [13] M. Sohda, Y. Misumi, A. Yano, N. Takami, Y. Ikehara, Phosphorylation of the vesicle docking protein p115 regulates its association with the Golgi membrane, J. Biol. Chem. 273 (1998) 5385–5388. [14] J. Soonthornsit, Y. Yamaguchi, D. Tamura, R. Ishida, Y. Nakakoji, S. Osako, et al., Low cytoplasmic pH reduces ER-Golgi trafficking and induces disassembly of the Golgi apparatus, Exp. Cell Res. 328 (2014) 325–339. http://dx.doi.org/10.1016/ j.yexcr.2014.09.009. [15] C. Steentoft, S.Y. Vakhrushev, H.J. Joshi, Y. Kong, M.B. Vester-Christensen, K.T.B.G. Schjoldager, et al., Precision mapping of the human O-GalNAc glycoproteome through SimpleCell technology, EMBO J. 32 (2013) 1478–1488. http://dx.doi.org/ 10.1038/emboj.2013.79. [16] N. Nakamura, C. Rabouille, R. Watson, T. Nilsson, N. Hui, P. Slusarewicz, et al., Characterization of a cis-Golgi matrix protein, GM130, J. Cell Biol. 131 (1995) 1715–1726. [17] J. Lippincott-Schwartz, L. Yuan, C. Tipper, M. Amherdt, L. Orci, R.D. Klausner, Brefeldin A's effects on endosomes, lysosomes, and the TGN suggest a general mechanism for regulating organelle structure and membrane traffic, Cell 67 (1991) 601–616. http://dx.doi.org/10.1016/0092-8674(91)90534-6. [18] K. Brandl, W. Tomisato, X. Li, C. Neppl, E. Pirie, W. Falk, et al., Yip1 domain family, member 6 (Yipf6) mutation induces spontaneous intestinal inflammation in mice, Proc. Natl. Acad. Sci. USA 109 (2012) 12650–12655. http://dx.doi.org/ 10.1073/pnas.1210366109. [19] G. Huet, I. Kim, C. de Bolos, J.M. Lo-Guidice, O. Moreau, B. Hemon, et al., Characterization of mucins and proteoglycans synthesized by a mucin-secreting HT-29 cell subpopulation, J. Cell Sci. 108 (Pt 3) (1995) 1275–1285.
jbc.M302406200. [5] C.Z. Chen, M. Calero, C.J. DeRegis, M. Heidtman, C. Barlowe, R.N. Collins, Genetic analysis of yeast Yip1p function reveals a requirement for Golgi-localized rab proteins and rab-Guanine nucleotide dissociation inhibitor, Genetics 168 (2004) 1827–1841. http://dx.doi.org/10.1534/genetics.104.032888. [6] A. Shakoori, G. Fujii, S.-I. Yoshimura, M. Kitamura, K. Nakayama, T. Ito, et al., Identification of a five-pass transmembrane protein family localizing in the Golgi apparatus and the ER, Biochem. Biophys. Res. Commun. 312 (2003) 850–857. http://dx.doi.org/10.1016/j.bbrc.2003.10.197. [7] T. Kranjc, E. Dempsey, G. Cagney, N. Nakamura, D.C. Shields, J.C. Simpson, Functional characterisation of the YIPF protein family in mammalian cells, Histochem. Cell Biol. (2016). [8] M. Calero, N.J. Winand, R.N. Collins, Identification of the novel proteins Yip4p and Yip5p as Rab GTPase interacting factors, FEBS Lett. 515 (2002) 89–98. [9] K. Tanimoto, K. Suzuki, E. Jokitalo, N. Sakai, T. Sakaguchi, D. Tamura, et al., Characterization of YIPF3 and YIPF4, cis-Golgi localizing Yip domain family proteins, Cell Struct. Funct. 36 (2011) 171–185. [10] Y. Yoshida, K. Suzuki, A. Yamamoto, N. Sakai, M. Bando, K. Tanimoto, et al., YIPF5 and YIF1A recycle between the ER and the Golgi apparatus and are involved in the maintenance of the Golgi structure, Exp. Cell Res. 314 (2008) 3427–3443. http:// dx.doi.org/10.1016/j.yexcr.2008.07.023. [11] R.E. Campbell, O. Tour, A.E. Palmer, P.A. Steinbach, G.S. Baird, D.A. Zacharias, et al., A monomeric red fluorescent protein, Proc. Natl. Acad. Sci. USA 99 (2002) 7877–7882. http://dx.doi.org/10.1073/pnas.082243699. [12] D.T. Shima, K. Halder, R. Pepperkok, R. Watson, G. Warren, Partitioning of the Golgi apparatus during mitosis in living HeLa cells, J. Cell Biol. 137 (1997)
108