1. Cell Biology

Functional hierarchy among different Rab27 effectors involved in secretory granule exocytosis

  1. Kunli Zhao
  2. Kohichi Matsunaga
  3. Kouichi Mizuno
  4. Hao Wang
  5. Katsuhide Okunishi
  6. Tetsuro Izumi  Is a corresponding author
  1. Laboratory of Molecular Endocrinology and Metabolism, Department of Molecular Medicine, Institute for Molecular and Cellular Regulation, Gunma University, Japan
https://doi.org/10.7554/eLife.82821
Share this article
Cite this article
  1. Kunli Zhao
  2. Kohichi Matsunaga
  3. Kouichi Mizuno
  4. Hao Wang
  5. Katsuhide Okunishi
  6. Tetsuro Izumi
(2023)
Functional hierarchy among different Rab27 effectors involved in secretory granule exocytosis
eLife 12:e82821.
https://doi.org/10.7554/eLife.82821
  2. Download BibTeX
  3. Download .RIS

Abstract

The Rab27 effectors are known to play versatile roles in regulated exocytosis. In pancreatic beta cells, exophilin-8 anchors granules in the peripheral actin cortex, whereas granuphilin and melanophilin mediate granule fusion with and without stable docking to the plasma membrane, respectively. However, it is unknown whether these coexisting effectors function in parallel or in sequence to support the whole insulin secretory process. Here, we investigate their functional relationships by comparing the exocytic phenotypes in mouse beta cells simultaneously lacking two effectors with those lacking just one of them. Analyses of prefusion profiles by total internal reflection fluorescence microscopy suggest that melanophilin exclusively functions downstream of exophilin-8 to mobilize granules for fusion from the actin network to the plasma membrane after stimulation. The two effectors are physically linked via the exocyst complex. Downregulation of the exocyst component affects granule exocytosis only in the presence of exophilin-8. The exocyst and exophilin-8 also promote fusion of granules residing beneath the plasma membrane prior to stimulation, although they differentially act on freely diffusible granules and those stably docked to the plasma membrane by granuphilin, respectively. This is the first study to diagram the multiple intracellular pathways of granule exocytosis and the functional hierarchy among different Rab27 effectors within the same cell.

Editor's evaluation

This study examines how Rab27 and its different effectors regulate insulin secretory granule exocytosis. Using single and double knockouts and rescue experiments, the work presents convincing data to characterize the relative hierarchy of the Rab27 effectors, their action with the exocyst, and their roles in distinct types of exocytosis. Overall, this is a valuable study that sheds new light on the regulation of distinct forms of secretory granule exocytosis.

https://doi.org/10.7554/eLife.82821.sa0

Introduction

Synaptic vesicles must be docked and primed on the plasma membrane prior to stimulation for successful neuronal transmission to occur within 1 ms after electrical stimulation (Südhof, 2013). In contrast, secretory granules in exocrine and endocrine cells initiate exocytosis at a timepoint that is 1000-fold later than this, even when the Ca2+ concentration is abruptly upregulated by caged-Ca2+ compounds (Kasai, 1999). Their physiological release of bioactive substances likewise occurs much later. For example, digestive enzymes and insulin are secreted from pancreas cells continuously over a range of minutes to hours during food intake and subsequent hyperglycemia. To support such slow and consecutive exocytosis, the exocytic pathway can be neither single nor linear, and the requisite steps for recruiting granules from the cell interior to the cell limits are therefore critical. In fact, total internal reflection fluorescence (TIRF) microscopy monitoring of prefusion behavior in living cells has revealed that insulin granules residing beneath the plasma membrane prior to stimulation, and those recruited from the cell interior after stimulation, fuse in parallel, with some variability in their ratios, during physiological glucose-stimulated insulin secretion (GSIS) (Kasai et al., 2008; Ohara-Imaizumi et al., 2004; Shibasaki et al., 2007). In many secretory cells, a greater number of granules are clustered in the actin cortex at the cell periphery and/or along the plasma membrane compared with other cytoplasmic areas. It has traditionally been considered that granules docked to the plasma membrane form a readily releasable pool, whereas those accumulated within the actin cortex form a reserve pool. However, this may be an oversimplification, considering that there is a gating system to prevent spontaneous or unlimited vesicle fusion in regulated exocytosis. In fact, multiple Rab27 effectors that are involved in intracellular granule trafficking show complex and differential effects on exocytosis (Izumi, 2021). For example, granuphilin (also known as exophilin-2 and Slp4) mediates stable granule docking to the plasma membrane but simultaneously prevents their spontaneous fusion by interacting with a fusion-incompetent, closed form of syntaxins (Gomi et al., 2005; Torii et al., 2002). Another effector, exophilin-8 (also known as MyRIP and Slac2-c), anchors secretory granules within the actin cortex (Bierings et al., 2012; Desnos et al., 2003; Fan et al., 2017; Huet et al., 2012; Mizuno et al., 2011; Nightingale et al., 2009), which is considered to have dual roles in accumulating granules at the cell periphery and in preventing their access to the plasma membrane. Although another effector, melanophilin (also known as exophilin-3 and Slac2-a), similarly captures melanosomes within the peripheral actin network in skin melanocytes (Hammer and Sellers, 2012), it mediates stimulus-induced granule mobilization and immediate fusion to the plasma membrane in pancreatic beta cells (Wang et al., 2020). It is unknown, however, just how the different Rab27 effectors coexisting in the same cell function in a coordinated manner to support the entire exocytic processes. Neither is it known whether multiple secretory pathways and/or rate-limiting steps exist in the final fusion stages of regulated granule exocytosis. To answer these questions, we must first seek to determine whether each effector functions in sequence or in parallel. In this study, we compared the exocytic profiles in beta cells lacking the two effectors with those in cells deficient in each single effector and in wild-type (WT) cells. This yielded valuable insights into the functional hierarchy and relationship among the exocytic steps in which individual effectors are involved. We also found that the exocyst, which universally functions in constitutive exocytosis (Wu and Guo, 2015), plays roles in the physical and functional connections between different Rab27 effectors.

Results

Melanophilin exclusively functions downstream of exophilin-8 to mediate the exocytosis of granules recruited from the actin cortex to the plasma membrane after stimulation

In monolayer mouse pancreatic beta cells, insulin granules were unevenly distributed with accumulation in the actin cortex (Figure 1—figure supplement 1A). Because the Rab27 effectors, melanophilin and exophilin-8, are known to show affinities to the actin motors, myosin-Va and/or -VIIa (Hammer and Sellers, 2012, Izumi, 2021), these effectors are presumed to function on granules within this peripheral actin network. In fact, they showed a similar uneven distribution and were colocalized especially at the cell periphery (Figure 1—figure supplement 1B). To examine the functional relationship between these effectors, we generated melanophilin/exophilin-8 double-knockout (ME8DKO) mice by crossing exophilin-8-knockout (Exo8KO) mice (Fan et al., 2017) with melanophilin-knockout (MlphKO) mice (see ‘Materials and methods’). The doubly deficient beta cells displayed even distribution of insulin granules in the cytoplasm without accumulation at the cell periphery, in contrast to WT and MlphKO cells, but in a manner similar to that in Exo8KO cells (Figure 1A). Expression of exophilin-8, but not that of melanophilin, in ME8DKO cells at the endogenous level in WT cells redistributed them to the cell periphery (Figure 1B). This is consistent with the previous findings obtained from cells deficient in each single effector, which revealed that exophilin-8 is essential for granule accumulation within the actin cortex, whereas melanophilin is not (Fan et al., 2017; Wang et al., 2020).

Figure 1 with 1 supplement see all
Download asset Open asset
Exophilin-8, but not melanophilin, accumulates insulin granules in the action cortex.

(A) WT, MlphKO, Exo8KO, and ME8DKO beta cells were immunostained with anti-insulin antibody. A peripheral accumulation of insulin immunosignals was quantified under confocal microscopy: clustering of insulin immunosignals at least in one corner (upper) or along the plasma membrane (lower) were counted as positive. More than 100 cells were inspected in each of three independent experiments. Note that Exo8KO and ME8DKO cells do not show the peripheral accumulation of insulin granules in contrast to WT and MlphKO cells. (B) ME8DKO cells were infected by adenovirus expressing either HA-exophilin-8 or HA-melanophilin at the endogenous protein levels found in WT cells. LacZ was expressed in WT and ME8DKO cells as controls. The cell extracts were immunoblotted with the indicated antibodies (left). The ME8DKO cells expressing LacZ (upper), HA-exophilin-8 (middle), and HA-melanophilin (lower) were immunostained with anti-insulin and anti-HA antibodies (center), and a peripheral accumulation of insulin was quantified as in (A) (right). Insets represent higher magnification photomicrographs of a cell within the region outlined by frames. Note that expression of HA-exophilin-8, but not HA-melanophilin, rescues the peripheral granule accumulation in ME8DKO cells. Bars, 10 μm. ###p<0.001 by one-way ANOVA.

Figure 1—source data 1

Uncropped blot images of Figure 1B.

https://cdn.elifesciences.org/articles/82821/elife-82821-fig1-data1-v2.zip

ME8DKO mice show glucose intolerance without changes in body weight or insulin sensitivity, as found in each singly deficient mice (Figure 2—figure supplement 1A). The cells deficient in melanophilin and/or exophilin-8 all showed decreases in GSIS compared with WT cells in both islet batch and perifusion assays, although the decreases in ME8DKO cells tended to be larger than those in MlphKO cells (Figure 2A and B, Figure 2—figure supplement 1B and C). We next monitored insulin granule exocytosis directly by TIRF microscopy in living cells expressing insulin fused with enhanced green fluorescent protein (EGFP). The numbers of visible granules under TIRF microscopy were not different among the four genotypes of cells (Figure 2C). Lack of the change in Exo8KO cells suggests that granules residing beneath the plasma membrane are not necessarily derived or supplied from granules anchored within the actin cortex. The amount of insulin released in the medium correlated well with a reduction in the total number of fusion events detected as a flash followed by diffusion of insulin-EGFP fluorescence during glucose stimulation in each of the mutant cells (Figure 2D, Figure 2—figure supplement 1D). We previously categorized fused insulin granules into three types depending on their prefusion behaviors (Kasai et al., 2008): those having been visible prior to stimulation, ‘residents’; those becoming visible during stimulation, ‘visitors’; and those invisible until fusion, ‘passengers.’ MlphKO cells with the genetic background of C57BL/6N mice showed a specific decrease in the passenger-type exocytosis, particularly in a later phase, compared with WT cells (Figure 2D and E), which is consistent with previous findings in parental leaden cells with the genetic background of C57BR/cdJ mice (Wang et al., 2020). In contrast, both Exo8KO and ME8DKO cells with the same C57BL/6N genetic background displayed decreases in both the resident and passenger types of exocytosis, while changes in the markedly less frequent visitor type were difficult to compare among the cells. Because there are no significant differences in exocytic phenotypes between Exo8KO and ME8DKO cells, melanophilin is thought to function downstream of exophilin-8. Considering the previously identified role of each effector in beta cells (Fan et al., 2017; Wang et al., 2020), the passenger-type exocytosis mediated by melanophilin via interactions with myosin-Va and syntaxin-4 appears to be derived from granules anchored in the actin cortex by exophilin-8.

Figure 2 with 1 supplement see all
Download asset Open asset
Insulin secretory defects of beta cells doubly deficient in melanophilin and exophilin-8 are indistinguishable from those singly deficient in exophilin-8.

(A) Islets isolated from WT, MlphKO, Exo8KO, and ME8DKO mice at 12–18 weeks of age were preincubated in 2.8 mmol/L low glucose (LG)-containing KRB buffer at 37℃ for 1 hr. They were then incubated in new LG buffer for 1 hr followed by 25 mmol/L high glucose (HG) buffer for 1 hr. The ratios of insulin secreted in the media to that left in the cell lysates (Figure 2—figure supplement 1B) are shown (left; n = 9 from three mice each). (B) Islets from age-matched mice (19- to 25-week-old) were perfused with 16.7 mmol/L glucose buffer for 30 min, and the ratios of insulin secreted in the media to that left in the cell lysates are plotted as in Figure 2—figure supplement 1C. The area under the curve (AUC) is shown (n = 9 from three mice each). (C–E) A monolayer of islet cells from the above four kinds of mice were infected with adenovirus encoding insulin-EGFP and were observed by total internal reflection fluorescence (TIRF) microscopy (C, left). The numbers of visible granules were manually counted for WT (n = 10 cells from three mice), MlphKO (n = 9 from three mice), Exo8KO (n = 5 cells from three mice), and ME8DKO (n = 8 cells from three mice) cells (C, right). Insulin granule fusion events in response to 25 mmol/L glucose for 20 min were counted as in Figure 2—figure supplement 1D for WT (n = 21 cells from five mice), MlphKO (n = 10 from three mice), Exo8KO (n = 12 cells from three mice), and ME8DKO (n = 12 cells from three mice) cells. The observed fusion events were categorized into three types: residents, visitors, and passengers (D). Summary of the three modes of fusion events (E) in the first phase (left, 0–5 min) and the second phase (right, 5–20 min). Note that the decrease in insulin exocytosis in ME8DKO cells is greater than that in MlphKO cells, specifically due to the decrease in the resident-type exocytosis. Bar, 10 μm. #p<0.05, ##p<0.01, ###p<0.001 by one-way ANOVA.

Exophilin-8 and melanophilin form a complex via the exocyst

To investigate the molecular basis by which exophilin-8 and melanophilin sequentially promote the passenger-type exocytosis, we first examined the expression level of each effector in cells lacking the other effector. Although the expression of exophilin-8 was not affected in MlphKO cells, that of melanophilin was decreased to 59.8 ± 9.0% (n = 4) in Exo8KO cells compared with WT cells (Figure 3A). However, this decrease should not affect GSIS in Exo8KO cells because a similar level of expression (65.4 ± 4.7%) in heterozygous MlphKO cells did not reduce it (Figure 3—figure supplement 1). We further found that the exophilin-8 and melanophilin expressed in the pancreatic beta cell line MIN6 form a complex (Figure 3B). These findings suggest that the protein stability of melanophilin partially depends on its interaction with exophilin-8, which is consistent with the model that melanophilin functions downstream of exophilin-8. However, they do not interact when expressed in HEK293A cells (Figure 3—figure supplement 2A), suggesting that they do not interact directly. The two effectors have been shown to interact with different proteins, except for Rab27a, in beta cells (Fan et al., 2017; Wang et al., 2020). The exophilin-8 mutant that loses its binding activity to RIM-BP2, and the melanophilin mutants that lose their binding activity to Rab27a, myosin-Va, or actin, all preserved their binding activity to the other WT effector in MIN6 cells (Figure 3—figure supplement 2B). To identify unknown intermediate proteins, we individually expressed Myc-TEV-FLAG (MEF)-tagged exophilin-8 and melanophilin in MIN6 cells, and we analyzed the proteins in each of the anti-FLAG immunoprecipitates using a liquid chromatography-tandem mass spectrometry (LC-MS/MS) (Figure 3—figure supplement 3). As a result, we identified the exocyst complex components, SEC8, SEC10, and EXO70, in both immunoprecipitates, which is consistent with the previous finding that exophilin-8 interacts with SEC6 and SEC8 in INS-1 832/13 cells (Goehring et al., 2007). We confirmed the endogenous interactions among melanophilin, exophilin-8, SEC6, and SEC10 in MIN6 cells (Figure 3C). The exocyst forms an evolutionarily conserved heterooctameric protein complex (Wu and Guo, 2015). To identify the components responsible for the interaction with two Rab27 effectors, we performed a visible immunoprecipitation (VIP) assay, which permits examination of large numbers of protein combinations and complicated one-to-many or many-to-many protein interactions simultaneously (Katoh et al., 2015). As a result, we found that exophilin-8 and melanophilin immediately, if not directly, interact with SEC8 and EXO70, respectively (Figure 3D). Although the exocyst components were found in both immunoprecipitates of melanophilin and exophilin-8, other interacting proteins such as RIM-BP2, RIM2, myosin-VIIa, and myosin-Va were identified in only one of the immunoprecipitates (Figure 3E), suggesting that the majority of each effector form a distinct complex in cells. In addition, although we previously reported that exophilin-8 primarily interacts with myosin-Vlla displaying a molecular mass of ~170 kDa in INS-1 832/13 rat cells (Fan et al., 2017), it interacted with mysosin-VIIa with an authentic mass of ~260 kDa in MIN6 mouse cells, which may reflect a species difference between these beta cell lines.

Figure 3 with 3 supplements see all
Download asset Open asset
Exophilin-8 and melanophilin interact through the exocyst in beta cells.

(A) Islet extracts (40 μg) from WT, MlphKO, Exo8KO, and ME8DKO mice were immunoblotted with antibodies toward the indicated proteins (left). Protein expression levels were quantified by densitometric analyses from 3 to 5 independent preparations (right). (B) MIN6 cells were infected with adenoviruses encoding control LacZ, mCherry-exophilin-8, and/or FLAG-melanophilin. After 2 days, the cell extracts underwent immunoprecipitation with mCherry nanobody. The immunoprecipitates (IP), as well as the 1% extracts (Input), were immunoblotted with anti-FLAG and anti-red fluorescent protein (RFP) antibodies. (C) MIN6 cell extracts were immunoprecipitated with rabbit anti-melanophilin antibody or control IgG, and the immunoprecipitates were immunoblotted with antibodies toward the indicated proteins. (D) HEK293A cells cultured in 10 cm dishes were transfected with mCherry-fused, exophilin-8 (left) or melanophilin (right) with the indicated EGFP-fused exocyst components. After 48 hr, the cell lysates were subjected to immunoprecipitation with mCherry-nanobody-bound glutathione-Sepharose beads. EGFP and mCherry fluorescence on the precipitated beads was observed by confocal microscopy. (E) MIN6 cells expressing FLAG-tagged, melanophilin or exophlin-8 were immunoprecipitated with anti-FLAG antibody, and the immunoprecipitates were immunoblotted with antibodies toward the indicated proteins. Note that the exocyst complex components exist in both immunoprecipitates, whereas RIM-BP2, RIM2, myosin-VIIa, and myosin-Va largely exist only in one of them. Bars, 100 μm. ###p<0.001 by one-way ANOVA.

Figure 3—source data 1

Uncropped blot images of Figure 3A, B, C and E.

https://cdn.elifesciences.org/articles/82821/elife-82821-fig3-data1-v2.zip

The exocyst functions only in the presence of exophilin-8

In mammalian cells, the exocyst complex is assembled after separate formation of the subcomplex 1 (SEC3, SEC5, SEC6, SEC8) and the subcomplex 2 (SEC10, SEC15, EXO70, EXO84) (Ahmed et al., 2018). Because exophilin-8 and melanophilin immediately bind SEC8 and EXO70, respectively (Figure 3D), it is conceivable that the holo-exocyst links the two effectors. Previous VIP assays detected interactions between SEC8 or SEC3 in the subcomplex 1 and SEC10 in the subcomplex 2 (Katoh et al., 2015). Therefore, silencing of SEC10 is expected to efficiently disrupt the formation of the holo-exocyst. However, SEC10 knockdown by specific siRNAs did not affect the granule localization of exophilin-8 or melanophilin (Figure 4—figure supplement 1). Furthermore, there was no effect on the peripheral accumulation of insulin granules in contrast to the case in exophilin-8-deficient cells (Figure 1A). However, SEC10 knockdown significantly decreased the colocalization ratio of melanophilin, but not that of exophilin-8, with SEC6, another exocyst component that is colocalized with insulin granules in MIN6 cells (Tsuboi et al., 2005; Figure 4A). Considering that the two effectors specifically interact with the component of different subcomplexes (Figure 3D), it is reasonable that silencing of SEC10 that disrupts the subcomplex 2 dissociates melanophilin from exophilin-8 associated with the subcomplex 1. To obtain further evidence that the two effectors associate through the exocyst, we downregulated SEC8 that is thought to interact with exophilin-8 (Figure 4—figure supplement 1) and found that SEC8 knockdown markedly disrupted the complex formation (Figure 4B). These findings indicate that the exocyst physically connects the two effectors on the same granule.

Figure 4 with 1 supplement see all
Download asset Open asset
Knockdown of the exocyst component disrupts the interaction between melanophilin and exophilin-8.

(A) HA-exophilin-8 and HA-melanophilin were expressed in Exo8KO and MlphKO monolayer beta cells, respectively, at the endogenous levels in WT cells as described in Figure 1B. They were then transfected with control siRNA or siRNA against SEC10 #11 or #12, as shown in Figure 4—figure supplement 1A. After fixation, the cells were coimmunostained with anti-HA and anti-SEC6 antibodies and were observed by confocal microscopy (left). Fluorescent intensity profiles along the indicated line of SEC6 and either HA-exophilin-8 or HA-melanophilin are shown (center). Colocalization was quantified by Pearson’s correlation coefficient (right, n = 18–26 cells from two mice each). Note that SEC10 knockdown (KD) induces dissociation of melanophilin, but not exophilin-8, from SEC6 (black arrowheads). (B) MIN6 cells were transfected with control siRNA or siRNA against SEC8 #12, as shown in Figure 4—figure supplement 1A, and the cell extracts were subjected to immunoprecipitation with control IgG or anti-melanophilin antibody. The immunoprecipitates (IP), as well as the 1% extracts (Input), were immunoblotted with anti-exophilin-8 and anti-melanophilin antibodies. Bars, 10 μm. ###p<0.001 by one-way ANOVA.

Figure 4—source data 1

Uncropped blot images of Figure 4B.

https://cdn.elifesciences.org/articles/82821/elife-82821-fig4-data1-v2.zip

SEC10 knockdown markedly decreased GSIS in WT cells, but induced no further decrease of GSIS in Exo8KO cells, which was already decreased by half compared with WT cells (Figure 5A, Figure 5—figure supplement 1). These findings suggest that the exocyst functions only in the presence of exophilin-8. As shown in Figure 2C, TIRF microscopy of Exo8KO cells expressing insulin-EGFP revealed no decrease in the number of visible granules (Figure 5B). Similarly, SEC10 knockdown does not affect the number of visible granules in either WT or Exo8KO cells. Consistent with insulin secretion assays (Figure 5A), it induces a marked decrease in glucose-stimulated fusion events in WT cells, but no additional decrease in Exo8KO cells (Figure 5C). Categorization by prefusion behavior revealed decreases in both resident and passenger types of exocytosis in WT cells, which phenocopied Exo8KO cells without SEC10 knockdown (Figure 2D). Again, SEC10 knockdown had no effects on either type of exocytosis in Exo8KO cells. The holo-exocyst appears to function with exophilin-8 in the same pathway because SEC10 knockdown is expected to preserve the interaction between exophilin-8 and the subcomplex 1, as shown in Figure 4A. Taken together, it seems that exophilin-8 and the exocyst mediate the passenger-type exocytosis when they interact with melanophilin on the same granule.

Figure 5 with 1 supplement see all
Download asset Open asset
The exocyst affects insulin granule exocytosis only in the presence of exophilin-8.

(A–C) WT or Exo8KO mouse islet cells were twice transfected with control siRNA or siRNA against SEC10 #11 or #12, as shown in Figure 4—figure supplement 1A, and were plated in a 24-well plate (A) or glass base dish (B, C). (A) The transfected monolayer cells were incubated for 1 hr in KRB buffer containing 2.8 mmol/L glucose and were then stimulated for 1 hr in the same buffer or in buffer containing 25 mmol/L glucose. Insulin levels secreted in the media and left in the cell lysates were measured (Figure 5—figure supplement 1), and their ratios are shown (n = 3 from three mice each). (B) The control or SEC10 knockdown (KD) cells (control siRNA-treated WT cells, n = 10 from three mice; SEC10 siRNA#11 or #12-treated WT cells, n = 11 from three mice; control siRNA-treated Exo8KO cells, n = 12 from three mice; SEC10 siRNA#11 or #12-treated Exo8KO cells, n = 12 from three mice) were infected with adenovirus encoding insulin-EGFP and were observed by total internal reflection fluorescence (TIRF) microscopy (left). The numbers of visible granules were manually counted (right). (C) The transfected cells (control siRNA-treated WT cells, n = 22 from three mice; SEC10 siRNA#11-treated WT cells, n = 8 from three mice; SEC10 siRNA#12-treated WT cells, n = 10 from three mice; control siRNA-treated Exo8KO cells, n = 15 from three mice; SEC10 siRNA#11-treated Exo8KO cells, n = 9 from three mice; SEC10 siRNA#12-treated Exo8KO cells, n = 7 from three mice) were infected with adenovirus encoding insulin-EGFP, and fusion events in response to 25 mmol/L glucose for 20 min were counted and categorized under TIRF microscopy as described in Figure 2D. Bar, 10 μm. #p<0.05, ##p<0.01, ###p<0.001 by one-way ANOVA versus control siRNA-treated WT cells.

Exophilin-8 promotes the exocytosis of granules residing beneath the plasma membrane only in the presence of granuphilin

As shown in Figure 2D, exophilin-8 deficiency also affects the resident-type exocytosis. Another Rab27 effector, granuphilin, is thought to be deeply involved in this type of exocytosis because beta cells lacking granuphilin display very few granules directly attached to the plasma membrane under electron microscopy (Gomi et al., 2005). Despite this docking defect, these cells display a marked increase in granule exocytosis, possibly because granuphilin interacts with and stabilizes syntaxins in a fusion-incompetent, closed form (Torii et al., 2002). To explore the functional relationship between the two effectors in this type of exocytosis, we generated granuphilin/exophilin-8 double-knockout (GE8DKO) mice. The GE8DKO cells also exhibited an increase in GSIS compared with Exo8KO cells (Figure 6A, Figure 6—figure supplement 1A), which indicates that a significant number of granules fuse efficiently without exophilin-8 and granuphilin, thus without prior capture in the actin cortex and stable docking to the plasma membrane. TIRF microscopy of these cells expressing insulin-EGFP revealed that, although the number of visible granules was markedly decreased in granuphilin-knockout (GrphKO) cells compared with WT cells, as expected, it was not further decreased in GE8DKO cells compared with GrphKO cells (Figure 6B). The number of fusion events during glucose stimulation under TIRF microscopy was correlated with the amount of insulin released in the medium in each mutant cell (Figure 6C). All the types of exocytosis were increased in GrphKO cells, suggesting that the absence of granuphilin facilitates granule easier access to the fusion-competent machinery on the plasma membrane. Simultaneous absence of exophilin-8 induced strikingly differential effects on each type of exocytosis in GE8DKO cells: the increases in the passenger and visitor types were reduced to the level found in Exo8KO cells, whereas the increase in the resident type was completely unaffected. The former finding is consistent with the view that the passenger and visitor types of exocytosis are derived from granules captured in the actin cortex by exophilin-8. However, the latter finding indicates that, at least in the absence of granuphilin, exophilin-8 is dispensable for the exocytosis of granules already residing beneath the plasma membrane prior to stimulation.

Figure 6 with 1 supplement see all
Download asset Open asset
Exophilin-8 deficiency strongly inhibits the resident-type exocytosis from granuphilin-positive, stably docked granules.

(A) Islets isolated from WT, Exo8KO, GrphKO, and GE8DKO mice at 12–17 weeks of ages were preincubated in low glucose (LG) KRB buffer for 1 hr. They were then incubated in another LG buffer for 30 min followed by high glucose (HG) buffer for 30 min. The ratios of insulin secreted in the media of that left in the cell lysates (Figure 6—figure supplement 1A) are shown as in Figure 2A (n = 12 from four mice each). (B) A monolayer of islet cells from the above four kinds of mice were infected with adenovirus encoding insulin-EGFP and were observed by total internal reflection fluorescence (TIRF) microscopy (left). The numbers of visible granules were manually counted for WT (n = 15 cells from four mice), Exo8KO (n = 11 cells from three mice), GrphKO (n = 14 cells from four mice), and GE8DKO (n = 17 cells from four mice) cells (right). (C) Fusion events in response to 25 mmol/L glucose for 20 min were counted and categorized as described in Figure 2D (WT: n = 14 cells from three mice; Exo8KO: n = 15 cells from three mice; GrphKO: n = 18 cells from three mice; GE8DKO: n = 20 cells from three mice). Note that the increases in the passenger and visitor types in GrphKO cells were eliminated by the simultaneous absence of exophilin-8 in GE8DKO cells, whereas the increase in the resident type was completely unaffected at all. (D, E) A monolayer of GrphKO (n = 13) and GE8DKO (n = 12) islet cells from three mice each were infected by adenoviruses encoding insulin-EGFP and KuO-granuphilin to mimic WT and Exo8KO cells, respectively, as described in Figure 6—figure supplement 1B. Insulin granule fusion events in response to 25 mmol/L glucose for 20 min were counted and categorized under TIRF microscopy as in Figure 6C, except that the granuphilin-positive and -negative granules were distinguished (D). There were no granuphilin-positive granules showing either passenger or visitor-type exocytosis. The fusion probability of granuphilin-positive and -negative granules is shown as the percentage of those granules displaying the resident-type exocytosis (E). Bar, 10 μm. #p<0.05, ###p<0.001 by one-way ANOVA. *p<0.05, *** p<0.001 by Student’s t test.

However, the above findings were unexpexted because exophilin-8 deficiency markedly affects the resident-type exocytosis in the presence of granuphilin (Figure 2D). We previously showed in WT cells that the resident-type exocytosis is heterogeneously derived from granuphilin-positive, immobile granules and granuphilin-negative, mobile granules (Mizuno et al., 2016). To directly assess the influence of exophilin-8 deficiency on granuphilin-mediated, docked granule exocytosis, we expressed Kusabira Orange-1 (KuO)-fused granuphilin in GrphKO and GE8DKO cells (Figure 6—figure supplement 1B) because exogenous granuphilin expressed in the presence of endogenous granuphilin abnormally accumulates insulin granules beneath the plasma membrane and severely impairs their exocytosis (Mizuno et al., 2016; Torii et al., 2004; Torii et al., 2002). Under TIRF microscopy, the numbers of granuphilin-positive, -negative, and total granules were not significantly different between the two cell types. We confirmed similar numbers of these visible granules in WT and Exo8KO cells by immunostaining endogenous granuphilin and insulin (Figure 6—figure supplement 1C), suggesting that the expression levels of KuO-granuphilin in GrphKO and GE8DKO cells are properly adjusted to mimic WT and Exo8KO cells, respectively. Consistently, the number of total fusion events as well as the level of passenger-type exocytosis in GE8DKO cells expressing KuO-granuphilin (mimic Exo8KO cells) were markedly decreased compared with that in GrphKO cells expressing KuO-granuphilin (mimic WT cells; Figure 6D). Furthermore, although the levels of resident-type exocytosis from both granuphilin-positive and -negative granules were significantly inhibited in mimic Exo8KO cells compared with mimic WT cells, the suppression was more complete toward granuphilin-positive granules. The fusion probability of granuphilin-positive granules, in which the number of fusion events is normalized by the number of visible granules, was also more strongly decreased (Figure 6E). These findings indicate that exophilin-8 primarily promotes the exocytosis of granules molecularly and stably tethered to the plasma membrane by granuphilin.

The exocyst promotes the exocytosis of granules residing beneath the plasma membrane in the absence of granuphilin

Because the exocyst is also involved in the resident-type exocytosis (Figure 5C), we next investigated its functional relationship with granuphilin. We first examined the effects of SEC10 knockdown in GrphKO cells expressing insulin-EGFP under TIRF microscopy. Although it did not further decrease the number of visible granules in GrphKO cells (Figure 7A), as was the case in GE8DKO cells (Figure 6B), it markedly decreased all the types of exocytosis, including the resident-type exocytosis (Figure 7B), in contrast to the case in GE8DKO cells (Figure 6C). We then investigated the effect of SEC10 knockdown in GrphKO expressing both insulin-EGFP and KuO-granuphilin. Again, SEC10 knockdown did not change the numbers of granuphilin-positive, -negative, and total granules (Figure 7—figure supplement 1). However, these cells, including those having received control siRNA, did not respond to glucose stimulation well, possibly because two times of siRNA transfection and two kinds of adenovirus infection are too much intervention for vulnerable primary beta cells. Nevertheless, these cells did respond to KCl-induced depolarization. Although SEC10 knockdown significantly reduced the numbers of fusion from both granuphilin-positive and -negative granules (Figure 7C), it specifically decreased the fusion probability of granuphilin-negative granules (Figure 7D). Therefore, the exocyst deficiency primarily affects the resident-type exocytosis from granuphilin-negative granules, in contrast to exophilin-8 deficiency (Figure 6E).

Figure 7 with 1 supplement see all
Download asset Open asset
Exocyst deficiency strongly inhibits the resident-type exocytosis from granuphilin-negative, untethered granules.

(A) GrphKO beta cells were transfected with control siRNA (n = 10 from three mice) or siRNA against SEC10 #11 or #12 (n = 19 from three mice) were infected with adenovirus encoding insulin-EGFP as described in Figure 5B. They were observed by total internal reflection fluorescence (TIRF) microscopy (left), and numbers of visible granules were manually counted (right). (B) The control (n = 10 from three mice) or SEC10 knockdown (KD) cells (n = 19 from three mice) were infected with adenovirus encoding insulin-EGFP, and fusion events in response to 25 mmol/L glucose for 20 min were counted and categorized under TIRF microscopy as described in Figure 6C. Note that in contrast to exophilin-8 knockout (Figure 6C), SEC10 knockdown suppresses the resident-type exocytosis in GrphKO cells. (C, D) GrphKO islet cells from three mice were transfected with control (n = 14) or SEC10 siRNAs (n = 15) twice, and were infected with adenoviruses encoding insulin-EGFP and KuO-granuphilin on the next day. Because these cells that had undergone two times siRNA transfection and two kinds of adenovirus infection failed to respond to 25 mmol/L glucose stimulation well, they were depolarized by 60 mmol/L potassium. Fusion events during 6 min were counted and categorized (C) as in Figure 6D. The fusion probability of granuphilin-positive and -negative granules is shown (D) as in Figure 6E. Note that SEC10 knockdown selectively suppresses it from granuphilin-negative granules. Bar, 10 μm. *p<0.05, **p<0.01, ***p<0.001 by Student t test.

Discussion

In this study, using distinct prefusion behaviors observed by TIRF microscopy as markers of differential exocytic routes, we investigated the functional hierarchy among different Rab27 effectors in pancreatic beta cells lacking one or two of these effectors. We first present evidence that exophilin-8 functions upstream of melanophilin to drive the passenger-type exocytosis. This type of exocytosis appears to be derived from granules within the actin network because exophilin-8 is essential for granule accumulation in the actin cortex and both effectors function via interactions with myosin-VIIa and myosin-Va, respectively, in beta cells (Fan et al., 2017; Wang et al., 2020). We also show that these two effectors are linked by the exocyst protein complex. Although this evolutionarily conserved protein complex is known to function in constitutive exocytosis, it has not been established whether it plays a universal role in regulated exocytosis. For example, Drosophila with mutation in SEC5 displays a defect in neurite outgrowth but not in neurotransmitter secretion (Murthy et al., 2003). In yeast, the exocyst interacts with Sec4, a member of the Rab protein on secretory vesicles, via exocyst component Sec15 (Guo et al., 1999). Further, Sec4 and Sec15 directly bind Myo2, the yeast myosin-V, for secretory vesicle transport (Jin et al., 2011). The exocyst components also interact with the SNARE fusion machinery: Sec6 with Snc2 (VAMP/synaptobrevin family in mammals) (Shen et al., 2013) and Sec9 (SNAP-25 family in mammals) (Dubuke et al., 2015), and Sec3 with Sso2 (syntaxin family in mammals) (Yue et al., 2017). Although proteins corresponding to Rab27 effectors are absent in yeast, they appear to be involved in similar interactions in mammalian cells. In fact, melanophilin interacts with Rab27a, myosin-Va, and syntaxin-4 in beta cells (Wang et al., 2020). In mammalian cells, the exocyst complex is assembled after the formation of two separate subcomplexes (Ahmed et al., 2018). The holo-exocyst appears to connect exophilin-8 and melanophilin because both subcomplex components commonly exist in the immunoprecipitates of two effectors in beta cells. We further show that exophilin-8 and melanophilin associate via different subcomplex components, SEC8 and EXO70, respectively, and that disruption of one subcomplex results in dissociation between the two effectors. To our knowledge, this is the first example showing that different effectors toward the same Rab form a complex in cells, which corroborates the previous suggestion that multiple effectors on the same granules may smooth the transition between consecutive intermediate exocytic processes (Izumi, 2021).

We then show that exophilin-8 knockout and SEC10 knockdown exhibit very similar defects in granule exocytosis in WT cells, indicating that they function together. Furthermore, neither affects the number of granules visualized by TIRF microscopy. The sole difference is that, although exophilin-8 deficiency erases the granule accumulation in the actin cortex, exocyst deficiency does not, which suggests that the exocyst functions downstream of exophilin-8. Consistent with this view, we could not find any effects of SEC10 knockdown in Exo8KO cells. In the pheochromocytoma cell line, PC12, the nascent granules generated are transported in a microtubule-dependent manner to the cell periphery within a few seconds (Rudolf et al., 2001). In skin melanocytes, melanosomes are dispersed throughout the cytoplasm using the myosin-Va motor along dynamic actin tracks assembled by the SPIRE actin nucleator (Alzahofi et al., 2020). Thus, even without prior capture in the actin cortex by exophilin-8, insulin granules may also be eventually transported close to the plasma membrane using these routes.

We next show that, although exophilin-8 deficiency inhibits the resident-type exocytosis in WT cells, it does not affect it despite its increase in GrphKO cells. However, in GrphKO cells expressing KuO-granuphilin, exophilin-8 deficiency almost completely inhibits the exocytosis of granuphilin-positive granules. Granuphilin-mediated, stably docked granules are thought to require priming machinery for fusion, such as Munc13, that converts a granuphilin-associated, closed form of syntaxin to a fusion-competent, open form (Mizuno and Izumi, 2022). Exophilin-8 can contribute to this process because it is associated with RIM-BP2, RIM, and Munc13 (Fan et al., 2017), which are known to have such a priming role in synaptic vesicle exocytosis (Brockmann et al., 2019; Brockmann et al., 2020). These priming factors could nonspecifically convert even granuphilin-free syntaxins into the open form because exophilin-8 deficiency also significantly inhibits the resident-type exocytosis from granuphilin-negative granules. In fact, the exophilin-8 mutant that loses binding activity to RIM-BP2 has no effect on the decreased insulin secretion in exophilin-8-deficient cells (Fan et al., 2017).

In contrast to exophilin-8 deficiency, SEC10 deficiency suppresses the resident-type exocytosis even in the absence of granuphilin. Furthermore, in GrphKO cells expressing KuO-granuphilin, it specifically decreases the fusion probability from granuphilin-negative granules showing this type of exocytosis. We previously showed that granuphilin-negative granules are more mobile and fusogenic compared with granuphilin-positive granules under TIRF microscopy (Mizuno et al., 2016). However, these diffusible granules must somehow be directionally mobilized to the plasma membrane for fusion. The exocyst likely functions in this process, considering that it generally tethers secretory vesicles to the plasma membrane prior to membrane fusion in a broad range of cells (Wu and Guo, 2015). This view also explains why SEC10 deficiency does not affect the fusion probability of granuphilin-positive granules that have already stably docked to the plasma membrane. However, this tethering step should occur within 100–200 nm distance from the plasma membrane because SEC10 knockdown has no effects on the numbers of granules visualized by TIRF microscopy whether granuphilin is present or not.

In summary, our findings are the first to determine the functional hierarchy among different Rab27 effectors expressed in the same cell, although the exact mechanism remains unknown how these different steps are regulated in a coordinated manner. Some granules trapped in the actin cortex by the action of exophilin-8 stimulus-dependently fuse either immediately by the action of melanophilin (the passenger-type exocytosis) or after staying beneath the plasma membrane for a while (the visitor-type exocytosis) (Figure 8A). The exocyst may help tether granules to the plasma membrane in these types of exocytosis. Other granules somehow transported to the cell periphery without capture in the actin cortex are stably tethered to the plasma membrane by the action of granuphilin or remain untethered beneath the plasma membrane (Figure 8B). The former granules stimulus-dependently fuse after the granuphilin-associated, closed form of syntaxins are converted into the fusion-competent, open form, possibly by the action of exophilin-8-associated priming factors, such as RIM-BP2. The latter granules fuse stochastically after directed to the plasma membrane by the action of the exocyst. It should be noted, however, that granule exocytosis is not abrogated in beta cells lacking any of the single or double Rab27 effectors. This may suggest the existence of other secretory pathways. For example, other Rab27 effectors, such as exophilin-7 (also known as JFC1 or Slp1), can also promote some types of exocytosis (Wang et al., 2013). However, given that granules can arrive close to the plasma membrane and fuse efficiently without prior capture in the actin cortex by exophilin-8 and/or stable docking to the plasma membrane by granuphilin, Rab27 effectors are thought to be evolved to play regulatory roles that prevent spontaneous or inappropriate fusion in regulated secretory pathways, as suggested previously (Izumi, 2021). In any case, it is now evident that there are multiple redundant paths and rate-limiting processes toward the final fusion step in granule exocytosis, which evolved to ensure the robust performance of the mechanisms governing the secretion of vital molecules, such as insulin.

Figure 8
Download asset Open asset
A schematic model for the functional relationship among different Rab27 effectors and the exocyst in insulin granule exocytosis.

At least, four different types of insulin granule exocytosis are discriminated by total internal reflection fluorescence (TIRF) microscopic analyses: the passenger type and the visitor type derived from granules anchored in the actin cortex (A), and the resident type from granules docked to the plasma membrane or from granules that remain untethered beneath the plasma membrane (B). Exophilin-8, melanophilin, granuphilin, and the exocyst differentially regulate each type of exocytosis. See details in ‘Discussion.’

Materials and methods

Key resources table
Reagent type (species) or resourceDesignationSource or referenceIdentifiersAdditional information
Gene (Mus musculus)MyRIP
(Exophilin-8)		GenBankGene ID: 245049
Gene (M. musculus)Mlph
(Melanophilin)		GenBankGene ID: 171531
Gene (M. musculus)Sytl4
(Granuphilin)		GenBankGene ID: 27359
Cell line (Homo sapiens)HEK293AInvitrogenCat#: R70507;
RRID:
CVCL_6910		The cell line has been authenticated and tested negative for mycoplasma
Cell line (M. musculus)InsulinomaMiyazaki et al., 1990RRID:CVCL_0431The cell line has been authenticated and tested negative for mycoplasma
Transfected construct (M. musculus)On-Target plus non-targeting pool siRNAHorizon Discovery LtdCat#: D-001810-10-05Silencer Select
Transfected construct (M. musculus)siRNA to Exoc5 (SEC10)Horizon Discovery LtdCat#: 105504On-Target plus Set of 4 siRNA (J-047583-11 and -12)
Transfected construct (M. musculus)siRNA to Exoc4 (SEC8)Horizon Discovery LtdCat#: 20336On-Target plus Set of 4 siRNA (J-051541-12-0050)
Biological sample (M. musculus)Primary pancreatic beta cellsThis paperFreshly isolated islets from male mouse pancreas
AntibodyAnti-Rab27a/b (rabbit polyclonal)IBLCat#: 18975; RRID:AB_494635WB (1:2000)
AntibodyAnti-Rab27a (mouse monoclonal)BD BiosciencesCat#: 558532; RRID:
AB_647327		IF (1:100)
AntibodyAnti-Exophilin-8 (goat polyclonal)AbcamCat#: ab10149; RRID:AB_296882IF (1:100)
AntibodyAnti-Melanophilin (goat polyclonal)Everest BiotechCat#: EB05444; RRID:AB_2146092WB (1:2000)
AntibodyAnti-Melanophilin (rabbit polyclonal)ProteintechCat#: 10338-1-AP;
RRID:
AB_2146104		WB (1:2000)
AntibodyAnti-Myosin-VIIa (rabbit polyclonal)AbcamCat#: ab3481;
RRID:
AB_303841		WB (1:2000)
AntibodyAnti-Myosin-Va (rabbit polyclonal)Cell Signaling TechnologyCat#: 3402;
RRID:
AB_2148475		WB (1:2000)
AntibodyAnti-RIM-BP2 (rabbit polyclonal)ProteintechCat#: 15716-1-AP;
RRID:AB_2878173		WB (1:2000)
AntibodyAnti-RIM1/2 (rabbit polyclonal)Synaptic SystemsCat#: 140213;
RRID:
AB_2832237		WB (1:2000)
AntibodyAnti-SEC6 (mouse monoclonal)Assay DesignsCat#: ADI-VAM-SV021;
RRID:
AB_10618264		IF (1:100), WB (1:2000)
AntibodyAnti-SEC8 (rabbit polyclonal)ProteintechCat#: 11913-1-AP;
RRID:
AB_2101565		WB (1:2000)
AntibodyAnti-SEC10 (rabbit polyclonal)ProteintechCat#: 17593-1-AP;
RRID:
AB_2101582		WB (1:2000)
AntibodyAnti-EXO70 (rabbit polyclonal)ProteintechCat#: 12014-1-AP;
RRID:
AB_2101698		WB (1:2000)
AntibodyAnti-FLAG (rabbit polyclonal)Sigma-AldrichCat#: F7425;
RRID:
AB_439687		WB (1:2000)
AntibodyAnti-HA (rabbit polyclonal)MBLCat#: 561;
RRID:
AB_591839		WB (1:2000)
AntibodyAnti-HA (rat monoclonal)RocheCat#: 11867423001;
RRID:
AB_390918		IF (1:100)
AntibodyAnti-β-actin (mouse monoclonal)Sigma-AldrichCat#: A5316;
RRID:
AB_476743		WB (1:10,000)
AntibodyAnti-red fluorescent protein (RFP) (rabbit polyclonal)MBLCat#: PM005;
RRID:AB_591279		WB (1:10,000)
Commercial assay or kitAlphaLISA insulin kitPerkinElmerCat#: AL350HV/C/FInsulin detection
Commercial assay or kitInsulin high range kitCisbioCat#:
62IN1PEG		Insulin detection
Commercial assay or kitInsulin ultra sensitive kitCisbioCat#:
62IN2PEH		Insulin detection
Recombinant DNA reagentpEGFP-C3 (plasmid)AddgeneCat#: 53755-53762;
RRID:
Addgene_53755–53762		GFP version of SEC3, SEC5, SEC6, SEC8, SEC10, SEC15, EXO70, and EXO84
Software, algorithmImageQuant TL softwareCytivaRRID:SCR_014246Quantify immunoblotting
Software, algorithmNIS Element ViewerNikonRRID:SCR_014329Quantify colocalization and analyze fusion events
Software, algorithmProtein pilot softwareSCIEXRRID:SCR_018681Mass spectrometry
OtherRhodamine-conjugated phalloidinThermo Fisher ScientificCat#: R415IF (1:100)
OtherAnti-FLAG M2
Affinity Gel		Sigma-AldrichCat#: A2220IP
OtherProtein G
Sepharose 4
Fast Flow		GE Healthcare BiosciencesCat#: GE17-0618-01IP
OtherLipofectamine 3000InvitrogenCat#: L3000001Transfection
OtherLipofectamine RNAiMaxInvitrogenCat#: 13778075Transfection

Mice and pancreatic islet cell preparation

Request a detailed protocol

Animal experiments were performed according to the rules and regulations of the Animal Care and Experimental Committees of Gunma University (permit number: 22-010; Maebashi, Japan). Only male mice and their tissues and cells were phenotypically characterized in this study. Leaden (C57J/L) mice with nonfunctional mutation of the gene encoding melanophilin (Mlph) (Matesic et al., 2001) were purchased from The Jackson Laboratory (Strain #:000668, RRID:IMSR_JAX:000668), and were backcrossed with C57BL/6N mice 10 times to generate MlphKO mice. Exo8KO mice in the genetic background of C57BL/6N mice were described previously (Fan et al., 2017). ME8DKO mice were obtained by mating Exo8KO mice with the MlphKO mice described above. GrphKO mice in the genetic background of C3H/He mice were described previously (Gomi et al., 2005). The male Exo8KO mice were mated with the female GrphKO mice. Because the granuphilin and exophilin-8 genes are on the mouse X and 9 chromosomes, respectively, the resultant F1 generation is either male (Grph-/Y, Exo8+/-) or female (Grph+/-, Exo8+/-). By crossing these F1 mice, GE8DKO mice, as well as the WT, GrphKO, and Exo8KO mice, were generated in the F2 generation and used for experiments. Although the resultant F2 mice have a mixture of C57BL/6N and C3H/He genomes, we expected that significant phenotypic changes due to the loss of Rab27 effectors may be preserved despite any influences due to randomly distributed differences in the genome. In fact, we found similar differences in exocytic profiles between WT and Exo8KO cells both in the C57BL/6N background (Figure 2) and in the mixture of the C57BL/6N and C3H/He backgrounds (Figure 6). Furthermore, GrphKO cells in the mixture of the C57BL/6N and C3H/He backgrounds showed changes in granule localization and exocytosis (Figure 6), consistent with the reported phenotypes of GrphKO cells in the C3H/He background (Gomi et al., 2005). Pancreatic islet isolation and dissociation into monolayer cells and insulin secretion assays were performed as described previously (Gomi et al., 2005; Wang et al., 2020). Briefly, islets were isolated from cervically dislocated mice by pancreatic duct injection of collagenase solution, and size-matched five islets were cultured overnight in a 24-well plate. Monolayer islet cells were prepared by incubation with trypsin-EDTA solution and were cultured for further 2 days. Insulin released from isolated islets or monolayer cells was measured by an AlphaLISA insulin kit (PerkinElmer) or insulin high range and ultra-sensitive kits (Cisbio).

DNA manipulation

Request a detailed protocol

Mouse cDNAs of granuphilin (Wang et al., 1999), exophilin-8 (Mizuno et al., 2011), and melanophilin (Wang et al., 2020) were cloned previously. Human cDNAs of SEC3, SEC5, SEC6, SEC8, SEC10, SEC15, EXO70, and EXO84 in the pEGFP-C3 vector were gifts from Dr. Channing J. Der (Addgene plasmid # 53755-53762; http://n2t.net/addgene:53755-53762; RRID:Addgene_53755-53762; Martin et al., 2014). Adenoviruses encoding insulin-EGFP and KuO-granuphilin were described previously (Kasai et al., 2008; Mizuno et al., 2016). Hemagglutinin (HA)-, FLAG-, MEF-, One-STrEP-FLAG (OSF), and mCherry-tagged exophilin-8 and melanophilin were made previously (Fan et al., 2017; Wang et al., 2020). To express exogenous protein, HEK293A cells were transfected with the plasmids using Lipofectamine 3000 reagent (Invitrogen), whereas MIN6 and primary pancreatic beta cells were infected with adenoviruses.

Cell lines, antibodies, and immunoprocedures

Request a detailed protocol

MIN6 cells (RRID:CVCL_0431) were originally provided by Dr. Jun-ichi Miyazaki (Osaka University; Miyazaki et al., 1990). HEK293A cells (RRID:CVCL_6910) were purchased from Invitrogen (Cat# R70507). Guinea pig anti-insulin serum was a gift from H. Kobayashi (Gunma University). Rabbit polyclonal anti-exophilin-8 (αExo8N) and anti-granuphilin (αGrp-N) antibodies are described previously (Fan et al., 2017; Yi et al., 2002). mCherry nanobody was a gift from Drs. Y. Katoh and K. Nakayama (Kyoto University) (Katoh et al., 2016). The sources of commercially available antibodies and their concentrations used are listed in Key Resources Table. Cells were lysed by lysis buffer consisting of 50 mmol/L Tris-HCl, pH 7.5, 150 mmol/L NaCl, 10% (w/v) glycerol, 100 mmol/L NaF, 10 mmol/L ethylene glycol tetraacetic acid, 1 mmol/L Na3VO4, 1% Triton X-100, 5 μmol/L ZnCl2, 1 mmol/L phenylmethylsulfonyl fluoride, and cOmplete Protease Inhibitor Cocktail (Roche). Immunoblotting and immunoprecipitation were performed as described previously (Matsunaga et al., 2017; Wang et al., 2020). VIP assay using the nanobody was performed as described previously (Katoh et al., 2015). Briefly, HEK293A cells on 10 cm dish were transfected with pEGFP-Sec and either pmCherry-Melanophilin or pmCherry-Exophilin-8 by Lipofectamine 3000. After 48 hr, cells were lysed by 1 mL of lysis buffer, and the cell lysate was centrifuged at 10,000 × g for 10 min, and the supernatant was subjected to immunoprecipitation with a 5 μL gel volume of mCherry nanobody-bound glutathione Sepharose. The beads were washed three times with lysis buffer and transferred to 35 mm glass base dishes (Glass φ12, IWAKI). Green and red fluorescence of beads was observed by confocal laser scanning microscopy. Acquisition of images was performed under fixed conditions. For immunofluorescence procedures, monolayer primary beta cells seeded at 3–5 × 104 cells on a poly-L-lysine-coated glass base dish were cultured overnight in RPMI-1640 medium (11.1 mmol/L glucose) supplemented with 10% fetal calf serum and were fixed by 4% paraformaldehyde for 30 min at room temperature. The cells were washed by phosphate buffered saline (PBS) and permeabilized by PBS containing 0.1% Triton X-100 and 50 mmol/L NH4Cl for 30 min. After blocking with 1% bovine serum albumin in PBS for 30 min, the cells were immunostained and observed under confocal microscopy with a ×100 oil immersion objective lens (1.49 NA). Quantification of immunoblot signals was performed using ImageQuant TL software (Cytiva). Quantification of colocalization between two proteins was performed using NIS Element Viewer software (Nikon).

TIRF microscopy

Request a detailed protocol

TIRF microscopy was performed as described previously (Wang et al., 2020). Briefly, monolayer islet cells on glass base dish were infected with adenovirus encoding insulin-EGFP. Two days thereafter, the cells were preincubated for 30 min in 2.8 mmol/L low glucose (LG)-containing Krebs-Ringer bicarbonate (KRB) buffer at 37℃. They were then incubated in 25 mmol/L high glucose (HG)-containing buffer for 20 min or 60 mmol/L high potassium buffer for 6 min. TIRF microscopy was performed on a ×100 oil immersion objective lens (1.49 NA). The penetration depth of the evanescent field was 100 nm. Images were acquired at 103 ms intervals. Fusion events with a flash followed by diffusion of EGFP signals were manually selected and assigned to one of three types: residents, which were visible for more than 10 s prior to fusion; visitors, which became visible within 10 s of fusion; and passengers, which were not visible prior to fusion. In case of coinfection with adenovirus encoding KuO-granuphilin, sequential multi-color TIRF microscopy was performed as described previously (Mizuno et al., 2016). Briefly, EGFP was excited using a 488 nm solid-state laser, whereas KuO was excited using a 561 nm laser. Excitation illumination was synchronously delivered from an acousto-optic tunable filter-controlled laser launch. A dual-band filter set (LF488/561A; Semrock) was applied on a light path.

Mass spectrometry

Request a detailed protocol

MIN6 cells (2 × 107 cells in ten 15 cm dishes) were infected with adenovirus encoding FLAG, FLAG-melanophilin, or MEF-exophilin-8. The anti-FLAG immunoprecipitates were subjected to gel electrophoresis and visualized by Oriole fluorescent gel staining (Bio-Rad). Specific bands were excised and digested in gels with trypsin, and the resulting peptide mixtures were analyzed by a LC-MS/MS system, as described previously (Matsunaga et al., 2017). All MS/MS spectra were separated against the Mus musculus (mouse) proteome data set (UP000000589) at the Uniplot using Protein pilot software (SCIEX).

Silencing of the exocyst component in mouse pancreatic islet cells

Request a detailed protocol

The On-Target plus Set of four siRNA against mouse Exoc5 (J-047583-11 and -12; Cat# 105504) and mouse Exoc4 (J-051541-12-0050; Cat# 20336), and the On-Target plus non-targeting pool siRNA were purchased from Horizon Discovery Ltd. Mouse pancreatic islet cells suspended in 1 × 105 cells/280 μL were transfected with 100 nmol/L siRNA using Lipofectamine RNAiMAX reagent (Invitrogen). After plating on a 24-well plate or glass base dish for 48 hr, the cells were transfected with the same siRNA for the second time. After another 48 hr, the cells were subjected to immunoblotting analyses, immunofluorescent staining, insulin secretion assays, or TIRF microscopy after infection with adenovirus encoding insulin-EGFP.

Statistical analysis

Request a detailed protocol

All quantitative data were assessed as the mean ± SEM. p-Values were calculated using Student’s t test or a one-way ANOVA with a Tukey multiple-comparison using GraphPad Prism software.

Data and resource availability

Request a detailed protocol

All data generated or analyzed during this study are included in the manuscript and supporting files. All noncommercially available resources generated and/or analyzed during this study are available from the corresponding author upon reasonable request.

Data availability

All data generated or analyzed during this study are included in the manuscript and supporting files.

References

Decision letter

  1. Jonathan S Bogan
    Reviewing Editor; Yale School of Medicine, United States
  2. Anna Akhmanova
    Senior Editor; Utrecht University, Netherlands
  3. Jing Hughes
    Reviewer; Washington University, United States
  4. Alistair Hume
    Reviewer; University of Nottingham, School of Life Sciences, United Kingdom

Our editorial process produces two outputs: (i) public reviews designed to be posted alongside the preprint for the benefit of readers; (ii) feedback on the manuscript for the authors, including requests for revisions, shown below. We also include an acceptance summary that explains what the editors found interesting or important about the work.

Decision letter after peer review:

Thank you for submitting your article "Functional hierarchy among different Rab27 effectors involved in secretory granule exocytosis" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Anna Akhmanova as the Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Jing Hughes (Reviewer #2); Alistair Hume (Reviewer #3).

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential revisions:

1) Metabolic phenotyping of the DKO mouse needs to be performed, ideally including dynamic measurements of insulin secretion (Reviewer 2)

2) Additional data are require to strengthen the connection between the exocyst and the Rab27 effectors by additional loss of function experiments combined with live cell imaging of the components. It would also be important to clarify how melanophilin expression is affected by Exo8 KO (Reviewer 1).

3) The manuscript should use more precise language, noting the limitations of the interpretation and providing alternate explanations, particularly for the possibility that another pathway(s) for insulin secretion exists in these preparations (Reviewer 3).

Reviewer #1 (Recommendations for the authors):

1) The authors show both biochemically and by granule imaging that GSIS is reduced from β cells of exophilin-8 and melanophilin KO mice. However, information about granule density at the plasma membrane and insulin content in the cells is lacking. It is therefore difficult to know whether the secretory defect can also be explained by a relative insulin deficiency.

2) A problem with the study is that the authors follow the dynamic process of insulin secretion by live cell TIRF microscopy of glucose-stimulated cells, but then look at the localization of the Rab27 effectors by immunofluorescence staining of fixed cells, where the culture conditions at the time-point of fixation are unknown (e.g. glucose concentration). This makes it difficult to directly compare the results obtained by these two methods. The paper would be much strengthened from live cell imaging of Exophilin-8, melanophilin, granuphilin (as is done elegantly in Figure 6) and components of the exocyst complex. Such experiments will allow the authors to draw much more accurate conclusions about the identity of different granule subpopulation and how they relate to the different Rab27 effectors. This becomes particularly important for the exocyst components, where the authors data indicate that these would be present on a majority of granules that undergo exocytosis.

3) The authors show that melanophilin levels are reduced by 50% in exophilin-8 KO β cells, and suggest based on this that Exo8 may be required for stabilization of Melanophilin. This however, complicates the interpretation of the data in this model. How can the authors know that part of the exophilin-8 KO phenotype is not due to the simultaneous reduction in melanophilin expression? Although the authors show that the KO phenotype can be rescued by re-expression of exo-8, they don't show what happens with melanophilin expression under these conditions. It would be essential to show this, and also to test to what extent melanophilin re-expression affect the phenotype of the Exo-8 KO cells.

4) In figure 1 – supplement 1 the authors show a strikingly high degree of colocalization between exo-8 and Mlph. However, based on functional data and on the model shown in Figure 7, these two proteins would only be expected to colocalize on a small subset of granules that immediately fuse with the plasma membrane. What is the reason for this discrepancy?

5) The authors demonstrate, through biochemical assays, interactions between the exocyst complex and both Exo-8 and Mlph. Based on results from IP-experiments, the authors conclude that Exo-8 interacts with subcomplex I via Sec8 and that Mlph interacts with subcomplex 2 via Exo70. These are interesting results that corroborates earlier studies indicating a link between insulin granule mobilization/secretion and the exocyst complex.

However, it is very difficult to envisage how this interaction plays into the regulation of granule mobilization and exocytosis. In particular, it is unclear how the interaction with the exocyst complex affects the localization of the Rab27 effectors and what its role is in Grph-dependent exocytosis. The evidence for the Rab27 effectors associating with the exocyst complex would be strengthened by showing the loss of effectors from Rab27 positive granules following knockdown of Sec8 (for Exo-8) and Exo70 (for Mlph) and showing that the interaction between Exo-8 and Mlph is lost following KD by IP. It would also be relevant to determine to what extent Grph-dependent exocytosis also require the Exocyst complex (in the model it does not, but this is not experimentally tested). To show colocalization with Sec6 as is done now (Figure 3A,B) is not appropriate since it is unclear what granule pool is labelled by this protein and whether its localization is affected by Sec10 KD. In fact, it is unclear to me why the authors chose to reduce the expression of Sec10 and not Sec8/Exo70, which might have allowed the authors to more specifically dissect the Exo-8/Mlph pathways.

6) The components of the exocyst complex can have functions independent of the exocyst. Since the authors only reduce the expression of Sec10, it may be difficult to draw conclusions about the entire complex. More extensive loss of function studies would be required or at a minimum this possibility should be discussed.

7) The authors state that exophilin-8 and melanophilin were re-expressed at the endogenous level (Figure 1B), which is also what the Western blot shows. However, re-expression of exophilin-8 results in a much larger proportion of cells with peripheral granule accumulation (compare Figure 1A to Figure 1B; 45% and 65%). A likely explanation for this is that not all cells are transduced with the adenovirus. It would therefore be better to not say at the endogenous level but instead just talk about re-expression.

8) Could the authors explain the rational for the choice of glucose concentrations when stimulating the β cells. 2.8 mM is sub-physiological and 25 mM is supra-physiological. More reasonable would have been to increase the glucose concentration from 5 mM to 10 mM.

9) It is shown that exo-8 is not involved in Grph-mediated exocytosis of docked granules but still somehow involved in exocytosis of docked granules. How can Exo-8 deficiency primarily influence the release of granules that are already docked at the plasma membrane? Based on the proposed mechanism, Exo-8 would be expected to primarily influence exocytosis of granules that are mobilized from the cytosol along actin filaments.

10) It would be appropriate to also include Grph KO and WT cells in the experiments presented in Figure 6, so that one can compare the degree of rescue obtained.

11) How do the authors propose that actin is involved? Is there an actin-independent (granuphilin) and actin-dependent (melanophilin) form of exocytosis in β cells?

12) Quantify the degree of colocalization between Sec6 and insulin in Figure 3 —figure supplement 1.

13) Cropped images in figure 1A and 1B are not correctly cropped.

14) The small decrease in Pearson correlation coefficient reported in Figure 3C should not be artificially inflated by y-axis cropping.

15) The authors several times talk about "immediate" binding. Could they please clarify what they mean by this. Is it direct binding?

16) The authors state that GSIS from the GE8DKO islets exhibit an intermediate level between GrphKO and Exo8KO, but this is not what is shown in Figure 5A (there are no differences at all).

17) In Figure 5C, Exo-8 KO is without effect on insulin secretion visualized by granule imaging (total and resident granules). This is different from other experiments and complicates interpretations also from the GE8DKO cells. Can the authors please comment on the reason for this.

18) How can the authors know that they are recording exocytic events from β cells and not from other islet cell types?

Reviewer #2 (Recommendations for the authors):

Suggestions for additional experiments or clarifications:

– While Exo8 and melanophilin loss-of-function mouse models have been previously characterized (Fan et al. 2017, Wang et al. 2020), the metabolic phenotype of double knockout ME8DKO mice is not presented here. It would be informative to include some standard in vivo metabolic testing such as IPGTT, body morphometrics, fasting glucose and diabetes incidence on chow or challenge diets, to see if these mice phenocopy Exo8KO and what does the loss of this pathway mean for diabetes propensity.

– Dynamic insulin secretion as the authors have previously performed (again, in Fan et al. 2017 and Wang et al. 2020) would complement the static secretion data in Figure 1C by showing whether the secretory defect lies in 1st or 2nd phase, and how this might correspond to the release of different insulin pools – resident, passenger, or visitor.

– The observation that melanophilin-only knockouts (MlphKO) have a milder phenotype than Exo8KO or ME8DKO suggests that there may be effectors other than melanophilin downstream of Exo8. Now that we have double KO data, what are the authors' conclusion regarding role of melanophilin, is it redundant and therefore expendable for Exo8 or Rab27 function? Under what physiologic conditions might this melanophilin-mediated, passenger-type exocytosis pathway be important?

Reviewer #3 (Recommendations for the authors):

In this manuscript Zhao et al. investigated how multiple Rab27 effectors work to regulate insulin secretion by murine pancreatic b-cells. They do this by comparing the phenotypes of b-cells/islets lacking effectors doubly or singly. Their main findings/contributions are that

1. Mlph works downstream of Myrip/exophilin-8 to mobilise granules for fusion from the actin network to the plasma membrane after stimulation.

2. Mlph and exophilin-8 interact via the exocyst

3. Down-regulation of exocyst affects exocytosis in cells expressing exophilin-8

4. Exophilin-8 promotes fusion of granules docked by granuphilin at the membrane

5. Exophilin-8 not required for Grph related granule docking at the plasma membrane

A model for how the three effectors coordinate ISG secretion. According to this model there are 2 insulin secretion pathways in b-cells; (a) where Exo8 acts upstream of Mlph and with actin/Myosin Va/VIIa, exocyst and syntaxin 4 to move dock granules in actin and promote exocytosis, and (b) where Exo8 works in an antagonistic manner with Grph promoting secretion of granules docked at the membrane by Grph.

In support of 1 they show that:

insulin granules are docked at the cortex (a prelude to fusion) in WT and MlphKO, but not Exo8KO and double ME8DKO b-cells (Figure 1).

In ME8DKO cells they show that expression of Exo8 but not Mlph restores cortical distribution of granules (Figure 1).

Using TIRFM analysis of insulin secretion in pancreatic b-cells that fusion of non-resident granules (passenger and visitor types) is reduced in MlphKO cells compared with controls in response to stimulation (Figure 1).

In support of 2 they show that:

over-expressed/tagged and endogenous Exo8 and mlph can be co-precipitated in MIN6 cells (Figure 2).

exocyst components Sec6 and Sec10 co-IP with Mlph and Exo8 (Figure 2).

Depletion of sec10 reduces Mlph association with sec6 positive insulin granules (Figure 3).

However, it would be important to show this directly e.g., by showing the results of Mlph/Exo8 IP in exocyst depleted cells to confirm that Mlph/Exo8 interaction is via the exocyst.

In support of 3 they show that:

using ELISA assays that glucose dependent insulin secretion is reduced in wild-type, but not Exo8KO, cells when Sec10 is KD (Figure 4A).

In support of 4 they show that:

insulin secretion in double granuphilin (Grph)/Exo8 DKO b-cells is reduced compared with single Grph KO cells (Figure 5A). Related to this they show that granule number in the TIRF field is affected GrphKO and GE8DKO but not Exo8KO. This seems surprising given that in Figure 1 the authors show that the number of docked granules is increased in Exo8KO cells.

In support of 5

they use TIRFM to show that granules are docked at the membrane by Exo8KO but not GrphKO or GE8DKO (Figure 5B).

Overall, their data support the model shown in Figure 6.

This is an interesting study but leaves open some important questions regarding insulin secretion and the role of Rab27a and its effectors in this process.

The finding that GE8DKO islets release more insulin than wild-type (Figure 5A) is striking and suggests that another pathway(s) for insulin secretion exists in these preparations. Can the authors discuss the possible molecular mechanism of this pathway in their manuscript/include this in their model to present a more complete picture of the mechanisms of granule exocytosis?

Related to this while I understand that it may be out of the remit of this project to characterise Grph/Exo8 independent pathways from scratch it would nevertheless be highly relevant to this study on Rab27a effectors for the authors to investigate the possibility that Rab27 and Mlph are involved in this pathway. For example, by generating triple Mlph, Exo8 and Grph8 KO islets or depleting in Mlph in GE8DKO islets and examining the effect of this on exocytosis and how this relates to exocytosis in the absence of Rab27a.

Regarding the possible different pathways have the authors noticed any differences in the kinetic profile of fusion events as monitored by TIRFM this might provide interesting insight into the relationship and mechanism between the proposed different pathways.

https://doi.org/10.7554/eLife.82821.sa1

Author response

Essential revisions:

1) Metabolic phenotyping of the DKO mouse needs to be performed, ideally including dynamic measurements of insulin secretion (Reviewer 2)

We have performed the above experiments and presented the data in Figure 2B, 2E, and Figure 2—figure supplement 1.

2) Additional data are require to strengthen the connection between the exocyst and the Rab27 effectors by additional loss of function experiments combined with live cell imaging of the components. It would also be important to clarify how melanophilin expression is affected by Exo8 KO (Reviewer 1).

We have performed additional loss of function experiments combined with live cell imaging of the components and presented the data in Figure 7 and Figure 7—figure supplement 1. We found differential roles in the resident type exocytosis between exophilin-8 and the exocyst. Please see our responses to Reviewer 1’s comments 2) and 5). Furthermore, we clarify the effect of melanophilin expression in Exo8KO cells by performing experiments in heterozygous MlphKO cells (Figure 3—figure supplement 1). Please see our response to Reviewer 1’s comment 3).

3) The manuscript should use more precise language, noting the limitations of the interpretation and providing alternate explanations, particularly for the possibility that another pathway(s) for insulin secretion exists in these preparations (Reviewer 3).

We have asked a professional external editor in the US to correct English and reflected the changes in a revised manuscript. Furthermore, we have added description about the existence of another pathway in the last part of Discussion.

Reviewer #1 (Recommendations for the authors):

1) The authors show both biochemically and by granule imaging that GSIS is reduced from β cells of exophilin-8 and melanophilin KO mice. However, information about granule density at the plasma membrane and insulin content in the cells is lacking. It is therefore difficult to know whether the secretory defect can also be explained by a relative insulin deficiency.

We have added the data of insulin content in Figure 2—figure supplement 1B and presented the data of granule density in Figure 2C.

2) A problem with the study is that the authors follow the dynamic process of insulin secretion by live cell TIRF microscopy of glucose-stimulated cells, but then look at the localization of the Rab27 effectors by immunofluorescence staining of fixed cells, where the culture conditions at the time-point of fixation are unknown (e.g. glucose concentration). This makes it difficult to directly compare the results obtained by these two methods. The paper would be much strengthened from live cell imaging of Exophilin-8, melanophilin, granuphilin (as is done elegantly in Figure 6) and components of the exocyst complex. Such experiments will allow the authors to draw much more accurate conclusions about the identity of different granule subpopulation and how they relate to the different Rab27 effectors. This becomes particularly important for the exocyst components, where the authors data indicate that these would be present on a majority of granules that undergo exocytosis.

We have described in Materials and methods that monolayer primary β cells were cultured in RPMI-1640 medium (11.1 mmol/L glucose) before the fixation.

About the identity of different granule subpopulation related to the different Rab27 effectors, we would first like the reviewer to understand that, although each Rab27 effector regulates a distinct exocytic step, all Rab27 effectors previously examined in pancreatic β cells, including Noc2 that also interacts with Rab2a located in the perinuclear region (Matsunaga, J Cell Sci, 2017), are not differentially located on specific granules where they function, but are distributed on almost all insulin granules throughout the cytoplasm via interaction with multiple Rab27a molecules on single granules (see granuphilin for Yi et al., Mol Cell Biol; exophilin-8 for Fan et al., ELife, 2017; and melanophilin for Wang et al., Diabetes, 2020). For example, we previously showed by both TIRF and confocal microscopy that granuphilin and melanophilin, which should separately mediate the resident type and the passenger type, respectively, are colocalized on granules (Supplementary Figure 6 in Wang et al., Diabetes, 2020). Thus, each effector is not necessarily located on different granule subpopulation. As discussed previously (Izumi, Cell Struct Funct, 2021), the fate of each individual granule is not defined by the presence of the particular Rab27 effector per se but appears to be stochastically determined by its interactions with specific proteins and membrane lipids in a local milieu around the granule. As the reviewer noticed, the exocyst component, Sec6, is also distributed to almost all the granules (Figure 4).

For live cell imaging, it is very important to express exogenous fluorescent protein in a physiological condition. For example, overexpressed granuphilin changes granule distribution (Torii et al., J Biol Chem, 2004) and severely inhibits granule exocytosis (Torii et al., Mol Cell Biol, 2002). Furthermore, we could not perform live cell imaging previously in the case of melanophilin, because mCherry-melanophilin expressed in the presence of endogenous melanophilin was not located on granules, but aberrantly located along the actin microfilaments, and because its expression at the endogenous level in melanophilin-deficient cells display too weak fluorescence to monitor granule exocytosis (please see the details in Wang et al., Diabetes, 2020). Therefore, we always had to express fluorescent Rab27 effectors in the absence of endogenous proteins. This is thought to be particularly critical for the exocyst that generally functions as a heterooctameric protein complex, because exogenous expression of one fluorescent component unbalances the quantities among the components. In support of this view, we found marked effects of SEC10 knockdown on granule exocytosis despite the partial downregulation (Figure 4—figure supplement 1A). Therefore, decent works involving live cell imaging of the exocyst have been performed in knockin cells in both yeast and mammalian cells (see examples, Boyd et al., J Cell Biol 2004, Donovan et al. J Cell Biol, 2015; Ahmed et al., Nat Commun, 2018). Given that our data in Figures 3 and 4 suggest that the holo-exocyst connects only a part of exophilin-8 and melanophilin, we do not want to perform experiments overexpressing one component of the exocyst, which unlikely functions as a good marker of the holo-exocyst and likely yields uncertain and possibly artificial results. We would like to do live cell imaging of the exocyst in primary β cells, after we generate the knockin mouse for one of the components in the future study. We hope that the editors and the reviewers understand that the main purpose of this study is to identify the functional relationship among different Rab27 effectors.

We also would like to point out that exophilin-8 and the exocyst are anyway involved in both resident and passenger types of exocytosis, as shown in Figures 2D and 5C, and therefore, their live-cell imagimg would not identify different granule subpopulation mediating the specific type of exocytosis. It should also be noted that TIRF microscopy does not provide direct information in the cell interior. For example, we could not get information where granules mediating the passenger exocytosis, which appear only in one frame prior to fusion, have been originally located before stimulation.

However, in the case of granuphilin that molecularly tether granules to the plasma membrane, live cell imaging by TIRF microscopy generates useful information. Given that the evanescent field illuminates fluorescence within 100-200 nm of the coverslip and that the radius of insulin granules is 150-200 nm, granuphilin-positive granules should harbor granuphilin in their hemisphere that is closer to the plasma membrane, and so are likely physically tethered to the plasma membrane via interaction with the closed form of syntaxins. In fact, we previously found that granules visible under TIRF microscopy can be divided into two classes: granuphilin-positive, immobile granules and granuphilin-negative, mobile granules (Mizuno et al., Sci Rep, 2016). Therefore, we can discriminate granules physically tethered to the plasma membrane and those just locating nearby. Please also understand that granuphilin-negative granules under TIRF microscopy mean its absence in their hemisphere that is closer to the plasma membrane but may still harbor granuphilin in their opposite hemisphere. To investigate the role of the exocyst in the resident type exocytosis, we performed new experiments in response to the reviewer’s point 5 below.

3) The authors show that melanophilin levels are reduced by 50% in exophilin-8 KO β cells, and suggest based on this that Exo8 may be required for stabilization of Melanophilin. This however, complicates the interpretation of the data in this model. How can the authors know that part of the exophilin-8 KO phenotype is not due to the simultaneous reduction in melanophilin expression? Although the authors show that the KO phenotype can be rescued by re-expression of exo-8, they don't show what happens with melanophilin expression under these conditions. It would be essential to show this, and also to test to what extent melanophilin re-expression affect the phenotype of the Exo-8 KO cells.

Although melanophilin must be kept at the endogenous level as described above, it is not easy to exactly compensate the decreased amount of endogenous melanophilin in each Exo8KO cell. Therefore, we instead examined the heterozygous MlphKO cells that may express it at a half level of the endogenous melanophilin. Indeed, the level of endogenous melanophilin in the heterozygous cells was 65.4 ± 4.7% of that in WT cells, which was equivalent to the level in Exo8KO cells (59.8 ± 9.0%). The finding that these heterozygous cells did not show reduced insulin secretion (Figure 3—figure supplement 1) suggests that the decrease in melanophilin expression is unlikely responsible for the reduced insulin secretion in Exo8KO cells.

4) In figure 1 – supplement 1 the authors show a strikingly high degree of colocalization between exo-8 and Mlph. However, based on functional data and on the model shown in Figure 7, these two proteins would only be expected to colocalize on a small subset of granules that immediately fuse with the plasma membrane. What is the reason for this discrepancy?

As described above in response to the reviewer’s point 2, the localization of each Rab27 effector is not restricted to specific granules. The localization of Rab27 effectors is determined via interaction with multiple Rab27 molecules existing on granules. For example, we previously showed the melanophilin E14A mutant deficient in Rab27 binding are not colocalized with granules at all (Wang et al., Diabetes, 2020). Therefore, the two proteins are highly colocalized on each insulin granule, although a subset of these proteins are expected to interact with, as the reviewer points out.

5) The authors demonstrate, through biochemical assays, interactions between the exocyst complex and both Exo-8 and Mlph. Based on results from IP-experiments, the authors conclude that Exo-8 interacts with subcomplex I via Sec8 and that Mlph interacts with subcomplex 2 via Exo70. These are interesting results that corroborates earlier studies indicating a link between insulin granule mobilization/secretion and the exocyst complex.

However, it is very difficult to envisage how this interaction plays into the regulation of granule mobilization and exocytosis. In particular, it is unclear how the interaction with the exocyst complex affects the localization of the Rab27 effectors and what its role is in Grph-dependent exocytosis. The evidence for the Rab27 effectors associating with the exocyst complex would be strengthened by showing the loss of effectors from Rab27 positive granules following knockdown of Sec8 (for Exo-8) and Exo70 (for Mlph) and showing that the interaction between Exo-8 and Mlph is lost following KD by IP. It would also be relevant to determine to what extent Grph-dependent exocytosis also require the Exocyst complex (in the model it does not, but this is not experimentally tested). To show colocalization with Sec6 as is done now (Figure 3A,B) is not appropriate since it is unclear what granule pool is labelled by this protein and whether its localization is affected by Sec10 KD. In fact, it is unclear to me why the authors chose to reduce the expression of Sec10 and not Sec8/Exo70, which might have allowed the authors to more specifically dissect the Exo-8/Mlph pathways.

more specifically dissect the Exo-8/Mlph pathways.

In response to the reviewer’s question, we first examined the effects of SEC10 knockdown on the intracellular localization of exophilin-8 and melanophilin, and found that SEC10 knockdown does not affect their colocalization with insulin or Rab27a (Figure 4—figure supplement 1). It is not surprising, because Rab27 effectors is localized on granules via interaction with Rab27, as described above. Therefore, it is not expected that any effectors will be lost from Rab27 positive granules even if other exocyst component is downregulated.

The reason we select SEC10 to downregulate the exocyst is that this component plays a key role to form the holo-exocyst, because previous VIP assays indicate that the two subcomplexes are biochemically linked via the interactions between Sec10 in the subcomplex 2 and Sec8 and/or Sec3 in the subcomplex 1 (Katoh et al., J Cell Sci, 2015). If SEC10 knockdown disrupts the interaction between exophilin-8 and melanophilin, the holo-exocyst likely links the two effectors. Consistent with this view, SEC10 knockdown partially disrupted the colocalization with SEC6 of melanophilin, but not that of exophilin-8 (Figure 4A).

However, according to the reviewer’s suggestion, we also tried to knockdown SEC8 and EXO70. The available siRNAs downregulated SEC8 significantly (Figure 4—figure supplement 1A), but unfortunately not EXO70 (see Author response image 1). In fact, siRNAs against EXO70 was reported to be inefficient (Ahmed et al., Nat Commun, 2018). We thus investigated whether the interaction between exophilin-8 and melanophilin is lost following SEC8 KD by IP, and found that it markedly disrupted the complex formation (Figure 4B), which further supports the involvement of the exocyst in the interaction of the two effectors.

Author response image 1
Download asset Open asset

In response to the reviewer’s question “what the role of the exocyst is in Grph-dependent exocytosis”, we first performed TIRF experiments in GrphKO cells. SEC10 knockdown did not further decrease the number of visible granules in GrphKO cells (Figure 7A), as found in GE8DKO cells (Figure 6B). However, it markedly decreased the resident type exocytosis (Figure 7B), in contrast to the case in GE8DKO cells (Figure 6C). To directly assess the influence of exocyst deficiency on granuphilin-mediated, docked granule exocytosis, we then expressed Kuo-Grph in GrphKO cells, as performed in Figure 6D,E. As shown in Figure 7C,D, SEC10 knockdown specifically decreased the fusion probability of granuphilin-negative granules (Figure 7D), in contrast that exophilin-8 deficiency markedly decreased that of granuphilin-positive granules (Figure 6E). As described in Discussion, these data suggest that the exocyst promote the exocytosis of granuphilin-negative, mobile granules by tethering them to the plasma membrane, as found in other constitutive secretory pathways. In contrast, exophilin-8 appears to promote fusion of granuphilin-mediated docked (tethered) granules by converting the associated, closed form of syntaxins into the open form.

Although we could not measure glucose-induced insulin secretion in the above experiments, we hope that the editors and the reviewers understand the vulnerable nature of primary islet cells after isolation. To our knowledge, there is no reports to do such multiple interventions, two times transfections and two kinds of adenovirus infections, for live cell imaging of primary β cells. In general, it is difficult to monitor two kinds of fluorescent proteins in addition to insulin-EGFP to examine their function in granule exocytosis in living β cells, without generating the knockin mouse that expresses multiple fluorescent protein instead of endogenous proteins.

6) The components of the exocyst complex can have functions independent of the exocyst. Since the authors only reduce the expression of Sec10, it may be difficult to draw conclusions about the entire complex. More extensive loss of function studies would be required or at a minimum this possibility should be discussed.

As described in response to the reviewer’s comment 5), we chose SEC10 knockdown to examine the involvement of the holo-exocyst, and performed additional experiments using SEC8 siRNA (Figure 4B). To our knowledge, although the exocyst has been investigated more than quarter century, the literature, including live cell imaging of the exocyst (Ahmad et al., Nat Commun, 2018), indicates that it functions as the holo-complex in cells.

7) The authors state that exophilin-8 and melanophilin were re-expressed at the endogenous level (Figure 1B), which is also what the Western blot shows. However, re-expression of exophilin-8 results in a much larger proportion of cells with peripheral granule accumulation (compare Figure 1A to Figure 1B; 45% and 65%). A likely explanation for this is that not all cells are transduced with the adenovirus. It would therefore be better to not say at the endogenous level but instead just talk about re-expression.

It is inevitable that adenovirus infection is somewhat inhomogeneous. We just meant the endogenous level in average.

8) Could the authors explain the rational for the choice of glucose concentrations when stimulating the β cells. 2.8 mM is sub-physiological and 25 mM is supra-physiological. More reasonable would have been to increase the glucose concentration from 5 mM to 10 mM.

We simply thought that the stronger stimulation makes the phenotypic differences more evident in the mutant cells. The glucose stimulation from 5 or 10 mM may be weak to detect significant differences.

9) It is shown that exo-8 is not involved in Grph-mediated exocytosis of docked granules but still somehow involved in exocytosis of docked granules. How can Exo-8 deficiency primarily influence the release of granules that are already docked at the plasma membrane? Based on the proposed mechanism, Exo-8 would be expected to primarily influence exocytosis of granules that are mobilized from the cytosol along actin filaments.

No, we did state that exophilin-8 is involved in Grph-mediated exocytosis, as shown in Figure 6. We also proposed that Exo8 promotes the priming reaction to remove a Grph-mediated exocytic clamp by associated proteins such as RIM-BP2, RIM, and Munc13. To support this view, we show that the absence of exophilin-8 inhibits the fusion of granuphilin-positive, docked granules more strongly compared with granuphilin-negative, undocked granules (Figure 6D,E). At present, we have no evidence that Exo8 primarily influences exocytosis of granules that are mobilized from the cytosol along actin filaments, because, under TIRF microscopy, the number of visible granules before stimulation is not changed by exophilin-8 deficiency (Figures 2C, 6B).

10) It would be appropriate to also include Grph KO and WT cells in the experiments presented in Figure 6, so that one can compare the degree of rescue obtained.

These experiments are performed to discriminate the effect of exophilin-8 deficiency on granuphilin-positive, docked granules from that on granuphilin-negative, untethered granules. We adjusted the expression level of granuphilin at endogenous level, as shown in Figure 6—figure supplement 1. Furthermore, the actual numbers of fusion events in WT and Exo8KO cells (Figure 6B) are equivalent to those found in mimic WT and mimic Exo8KO cells (Figure 6D).

11) How do the authors propose that actin is involved? Is there an actin-independent (granuphilin) and actin-dependent (melanophilin) form of exocytosis in β cells?

We should discriminate the role of actin tracks from that of the actin cortex in vesicle transport. For example, granules may reach the plasma membrane by generating actin tracks by another Rab27 effector, SPIRE, as shown in skin melanocytes (Alzahofi et al., Nat Commun, 2020). Based on the data in Figure 2, we suggest that melanophilin-mediated passenger type exocytosis is derived from granules captured within the actin cortex by Exophilin-8. Consistently, the increase in the passenger type exocytosis in the absence of granuphilin is markedly reduced to the level found in Exo8KO cells (Figure 6C). However, as shown in Figures 2C and 6B, the absence of exophilin-8 does not affect the number of granules visible under TIRF microscopy, which suggests that granules residing close to the membrane are not necessarily derived from granules anchored in the actin cortex.

12) Quantify the degree of colocalization between Sec6 and insulin in Figure 3 —figure supplement 1.

We have removed this data, because the reviewer mentioned in the comment 5) that Sec6 is not appropriate since it is unclear what granule pool is labelled by this protein and whether its localization is affected by Sec10 KD. We instead examined the effects of SEC10 knockdown on the granule localization of exophilin-8 and melanophilin, and quantified the degree of colocalization (Figure 4—figure supplement 1).

13) Cropped images in figure 1A and 1B are not correctly cropped.

Although the images were correctly cropped in the original file, they appear to be improperly cropped during the conversion process to the PDF file. We have corrected them.

14) The small decrease in Pearson correlation coefficient reported in Figure 3C should not be artificially inflated by y-axis cropping.

We have corrected the y-axis in Figure 4A (original Figure 3C).

15) The authors several times talk about "immediate" binding. Could they please clarify what they mean by this. Is it direct binding?

In general, coimmunoprecipitation experiments do not prove direct binding. However, visible IP assays show that only one of the exocyst components can interacts with exophilin-8 or melanophilin, and therefore we used the word “immediate”.

16) The authors state that GSIS from the GE8DKO islets exhibit an intermediate level between GrphKO and Exo8KO, but this is not what is shown in Figure 5A (there are no differences at all).

We corrected the sentence as below:

GE8DKO cells also exhibited an increase in GSIS compared with Exo8KO cells.

17) In Figure 5C, Exo-8 KO is without effect on insulin secretion visualized by granule imaging (total and resident granules). This is different from other experiments and complicates interpretations also from the GE8DKO cells. Can the authors please comment on the reason for this.

Compared with the experiments of Figure 2D that were performed in mice with the same B6 genetic background, the experiments in Figure 6C (original Figure 5C) were performed in mice with a mixture of B6 and C3H genetic backgrounds (see Materials and methods), which may have increased variance of the data and decreased the significance in differences. However, the mean values and differences between WT and Exo8KO cells are similar between these studies.

18) How can the authors know that they are recording exocytic events from β cells and not from other islet cell types?

Although we and others performing similar TIRF microscopic experiments did not discriminate β cells from other cells beforehand, we previously showed that most (90%) of the monolayer cells from pancreatic islets are β cells. See the examples of WT and melanophilin-deficient cells (Wang et al., Diabetes 2020). Furthermore, we did not include the data of cells that did not respond at all to glucose stimulation, which may be non-β cells or become inert during preparation and/or incubation.

Reviewer #2 (Recommendations for the authors):

Suggestions for additional experiments or clarifications:

– While Exo8 and melanophilin loss-of-function mouse models have been previously characterized (Fan et al. 2017, Wang et al. 2020), the metabolic phenotype of double knockout ME8DKO mice is not presented here. It would be informative to include some standard in vivo metabolic testing such as IPGTT, body morphometrics, fasting glucose and diabetes incidence on chow or challenge diets, to see if these mice phenocopy Exo8KO and what does the loss of this pathway mean for diabetes propensity.

We compared the metabolic phenotypes, such as body weight, blood glucose concentrations during IPGTT and those during ITT including fasting glucose concentrations among WT, MlphKO, Exo8KO, and ME8DKO mice. Each mouse strain displayed correlated insulin secretory capacities in batch (Figure 2A), islet perifusion (Figure 2B; Figure 2—figure supplement 1C), and TIRF microscopic assays (Figure 2D,E; Figure 2—figure supplement 1D). However, it should be noted that, because these effectors are widely and differentially expressed in other endocrine cells, their deficiency may differently affect secretion of other hormones influencing glucose tolerance, such as incretin and glucagon.

– Dynamic insulin secretion as the authors have previously performed (again, in Fan et al. 2017 and Wang et al. 2020) would complement the static secretion data in Figure 1C by showing whether the secretory defect lies in 1st or 2nd phase, and how this might correspond to the release of different insulin pools – resident, passenger, or visitor.

We have performed islet perifusion experiments (Figure 2—figure supplement 1C) and compared the time course of insulin secretion with that of granule exocytosis under TIRF microscopy (Figure 2—figure supplement 1D). The passenger type exocytosis is particularly decreased in a later 2nd phase in mutant cells (Figure 2E). However, the resident type exocytosis also tended to decrease in a later phase (Figure 2E).

– The observation that melanophilin-only knockouts (MlphKO) have a milder phenotype than Exo8KO or ME8DKO suggests that there may be effectors other than melanophilin downstream of Exo8. Now that we have double KO data, what are the authors' conclusion regarding role of melanophilin, is it redundant and therefore expendable for Exo8 or Rab27 function? Under what physiologic conditions might this melanophilin-mediated, passenger-type exocytosis pathway be important?

As shown in Figures 2A and 2D, MlphKO cells have a milder phenotype compared with Exo8KO or ME8DKO cells, because of the lack of decrease in the resident type exocytosis. However, melanophilin deficiency induces similar decreases in the passenger type exocytosis whether exophilin-8 is present (MlphKO cells) or not (ME8DKO cells), which indicates that melanophilin does not merely play a redundant role, but a unique role in this type of exocytosis. In support of this view, both effectors function through different protein complexes, RIM-BP2-myosin-VIIa and myosin-Va-syntaxin-4, respectively (Fan et al., 2017 and Wang et al., 2020).

Reviewer #3 (Recommendations for the authors):

In this manuscript Zhao et al. investigated how multiple Rab27 effectors work to regulate insulin secretion by murine pancreatic b-cells. They do this by comparing the phenotypes of b-cells/islets lacking effectors doubly or singly. Their main findings/contributions are that

1. Mlph works downstream of Myrip/exophilin-8 to mobilise granules for fusion from the actin network to the plasma membrane after stimulation.

2. Mlph and exophilin-8 interact via the exocyst

3. Down-regulation of exocyst affects exocytosis in cells expressing exophilin-8

4. Exophilin-8 promotes fusion of granules docked by granuphilin at the membrane

5. Exophilin-8 not required for Grph related granule docking at the plasma membrane

A model for how the three effectors coordinate ISG secretion. According to this model there are 2 insulin secretion pathways in b-cells; (a) where Exo8 acts upstream of Mlph and with actin/Myosin Va/VIIa, exocyst and syntaxin 4 to move dock granules in actin and promote exocytosis, and (b) where Exo8 works in an antagonistic manner with Grph promoting secretion of granules docked at the membrane by Grph.

In support of 1 they show that:

insulin granules are docked at the cortex (a prelude to fusion) in WT and MlphKO, but not Exo8KO and double ME8DKO b-cells (Figure 1).

In ME8DKO cells they show that expression of Exo8 but not Mlph restores cortical distribution of granules (Figure 1).

Using TIRFM analysis of insulin secretion in pancreatic b-cells that fusion of non-resident granules (passenger and visitor types) is reduced in MlphKO cells compared with controls in response to stimulation (Figure 1).

In support of 2 they show that:

over-expressed/tagged and endogenous Exo8 and mlph can be co-precipitated in MIN6 cells (Figure 2).

exocyst components Sec6 and Sec10 co-IP with Mlph and Exo8 (Figure 2).

Depletion of sec10 reduces Mlph association with sec6 positive insulin granules (Figure 3).

However, it would be important to show this directly e.g., by showing the results of Mlph/Exo8 IP in exocyst depleted cells to confirm that Mlph/Exo8 interaction is via the exocyst.

In support of 3 they show that:

using ELISA assays that glucose dependent insulin secretion is reduced in wild-type, but not Exo8KO, cells when Sec10 is KD (Figure 4A).

In support of 4 they show that:

insulin secretion in double granuphilin (Grph)/Exo8 DKO b-cells is reduced compared with single Grph KO cells (Figure 5A). Related to this they show that granule number in the TIRF field is affected GrphKO and GE8DKO but not Exo8KO. This seems surprising given that in Figure 1 the authors show that the number of docked granules is increased in Exo8KO cells.

In support of 5

they use TIRFM to show that granules are docked at the membrane by Exo8KO but not GrphKO or GE8DKO (Figure 5B).

Overall, their data support the model shown in Figure 6.

This is an interesting study but leaves open some important questions regarding insulin secretion and the role of Rab27a and its effectors in this process.

The finding that GE8DKO islets release more insulin than wild-type (Figure 5A) is striking and suggests that another pathway(s) for insulin secretion exists in these preparations. Can the authors discuss the possible molecular mechanism of this pathway in their manuscript/include this in their model to present a more complete picture of the mechanisms of granule exocytosis?

It is possible that other Rab27 effectors compensate for the absence of Granuphilin and Exophilin-8. For example, exophilin-7 is involved in undocked granule exocytosis in the absence of granuphilin, although exophilin-7-knockout β cells display only modestly reduced exocytosis (Wang et al., Mol Biol Cell, 2013). However, we would like to suggest the possibility that secretory vesicles can fuse efficiently once they are directed to the plasma membrane using machineries such as cytoskeletons and/or the exocyst that are present in almost all cells. For example, in PC12 cells, SNAREs in native plasma membranes are constitutively active even if not engaged in fusion events (Lang et al., J Cell Biol 2002;158:751–760). Therefore, regulated exocytic pathways are thought to be evolved by acquiring additional gating machinery to prevent spontaneous or inappropriate fusion. We would also like to suggest that granule clustering sites in the actin cortex or along the plasma membrane reflect the presence of such gating steps (in contrast, no vesicles accumulate near the target membrane in constitutive exocytic pathways). We would like to propose that Rab27 effectors play such regulatory roles, rather than essential roles, in exocytosis, as discussed previously (Izumi, Cell Struct Funct, 2021). If so, the finding that GE8DKO cells release more insulin than WT cells is not so surprising. I briefly describe our thought in Discussion as below.

In summary, our findings are the first to determine the functional hierarchy among different Rab27 effectors expressed in the same cell, although the exact mechanism remains unknown how these different steps are regulated in a coordinated manner.

It should be noted, however, that granule exocytosis is not abrogated in β cells lacking any of the single or double Rab27 effectors. This may suggest the existence of other secretory pathways. For example, other Rab27 effectors, such as exophilin-7 (also known as JFC1 or Slp1), can also promote some types of exocytosis (Wang et al., 2013). However, given that granules can arrive close to the plasma membrane and fuse efficiently without prior capture in the actin cortex by exophilin-8 and/or stable docking to the plasma membrane by granuphilin, Rab27 effectors are thought to be evolved to play regulatory roles that prevent spontaneous or inappropriate fusion in regulated secretory pathways, as suggested previously (Izumi, 2021).

Related to this while I understand that it may be out of the remit of this project to characterise Grph/Exo8 independent pathways from scratch it would nevertheless be highly relevant to this study on Rab27a effectors for the authors to investigate the possibility that Rab27 and Mlph are involved in this pathway. For example, by generating triple Mlph, Exo8 and Grph8 KO islets or depleting in Mlph in GE8DKO islets and examining the effect of this on exocytosis and how this relates to exocytosis in the absence of Rab27a.

If melanophilin has function in the absence of exophilin-8, ME8DKO cells should show additional defects compared with Exo8KO cells, which we could not detect (Figure 2). Furthermore, considering that melanophilin deficiency only affects the passenger type exocytosis (Figure 2D), we think it unlikely that additional melanophilin deficiency suppresses the increase in the resident type exocytosis in GrphKO or GE8DKO cells (Figure 6). Furthermore, as described above, constitutive-like exocytic pathways or other effectors such as exophilin-7 may function in the absence of granuphilin and exophilin-8.

Regarding the possible different pathways have the authors noticed any differences in the kinetic profile of fusion events as monitored by TIRFM this might provide interesting insight into the relationship and mechanism between the proposed different pathways.

Although we have not examined differences in kinetic profiles of fusion events between WT and mutant cells, we previously showed no obvious differences in time courses or amplitudes of fluorescence intense changes during fusion reaction on average, among resident, visitor, and passenger types of exocytosis in WT cells (Kasai et al., Traffic 2008; Wang et al., Diabetes 2020).

https://doi.org/10.7554/eLife.82821.sa2

Article and author information

Author details

  1. Kunli Zhao

    Laboratory of Molecular Endocrinology and Metabolism, Department of Molecular Medicine, Institute for Molecular and Cellular Regulation, Gunma University, Maebashi, Japan
    Contribution
    Data curation, Formal analysis, Validation, Investigation
    Competing interests
    No competing interests declared

  2. Kohichi Matsunaga

    Laboratory of Molecular Endocrinology and Metabolism, Department of Molecular Medicine, Institute for Molecular and Cellular Regulation, Gunma University, Maebashi, Japan
    Contribution
    Data curation, Formal analysis, Supervision, Validation, Investigation, Methodology
    Competing interests
    No competing interests declared

  3. Kouichi Mizuno

    Laboratory of Molecular Endocrinology and Metabolism, Department of Molecular Medicine, Institute for Molecular and Cellular Regulation, Gunma University, Maebashi, Japan
    Contribution
    Supervision, Funding acquisition, Validation, Methodology
    Competing interests
    No competing interests declared

  4. Hao Wang

    Laboratory of Molecular Endocrinology and Metabolism, Department of Molecular Medicine, Institute for Molecular and Cellular Regulation, Gunma University, Maebashi, Japan
    Contribution
    Supervision, Funding acquisition, Methodology
    Competing interests
    No competing interests declared

  5. Katsuhide Okunishi

    Laboratory of Molecular Endocrinology and Metabolism, Department of Molecular Medicine, Institute for Molecular and Cellular Regulation, Gunma University, Maebashi, Japan
    Contribution
    Resources
    Competing interests
    No competing interests declared

  6. Tetsuro Izumi

    Laboratory of Molecular Endocrinology and Metabolism, Department of Molecular Medicine, Institute for Molecular and Cellular Regulation, Gunma University, Maebashi, Japan
    Contribution
    Conceptualization, Data curation, Supervision, Funding acquisition, Validation, Visualization, Writing - original draft, Project administration, Writing - review and editing
    For correspondence
    tizumi@gunma-u.ac.jp
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-0974-7384

Funding

Japan Society for the Promotion of Science (JP19H03449)

  • Tetsuro Izumi

Japan Society for the Promotion of Science (JP20K06535)

  • Kouichi Mizuno

Japan Society for the Promotion of Science (JP20K15742)

  • Hao Wang

Eli Lilly Japan

  • Tetsuro Izumi

Teijin Pharma

  • Tetsuro Izumi

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

The authors thank T Nara, E Kobayashi, and T Ushigome for colony maintenance of mice, and S Shigoka for preparing the manuscript. This work was supported by Japan Society for the Promotion of Science KAKENHI grants JP19H03449 to TI, JP20K06535 to KMi, and JP20K15742 to HW. It was also supported by grants from Eli Lilly Japan Donation and Teijin Pharma Donation to TI. This work was the result of using research equipment shared in the MEXT Project for promoting public utilization of advanced research infrastructure (program for supporting the introduction of the new sharing system) grant number JPMXS0420600122.

Ethics

Animal experiments were performed according to the rules and regulations of the Animal Care and Experimental Committees of Gunma University (permit number: 22-010; Maebashi, Japan).

Senior Editor

  1. Anna Akhmanova, Utrecht University, Netherlands

Reviewing Editor

  1. Jonathan S Bogan, Yale School of Medicine, United States

Reviewers

  1. Jing Hughes, Washington University, United States
  2. Alistair Hume, University of Nottingham, School of Life Sciences, United Kingdom

Version history

  1. Received: August 18, 2022
  2. Preprint posted: August 23, 2022 (view preprint)
  3. Accepted: February 21, 2023
  4. Accepted Manuscript published: February 21, 2023 (version 1)
  5. Version of Record published: March 6, 2023 (version 2)

Copyright

© 2023, Zhao et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 903
    Page views
  • 155
    Downloads
  • 1
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Kunli Zhao
  2. Kohichi Matsunaga
  3. Kouichi Mizuno
  4. Hao Wang
  5. Katsuhide Okunishi
  6. Tetsuro Izumi
(2023)
Functional hierarchy among different Rab27 effectors involved in secretory granule exocytosis
eLife 12:e82821.
https://doi.org/10.7554/eLife.82821

Categories and tags

Research organism

Further reading

    1. Cell Biology
    2. Medicine

    Glucagon-like peptide-1 receptor activation stimulates PKA-mediated phosphorylation of Raptor and this contributes to the weight loss effect of liraglutide

    Thao DV Le, Dianxin Liu ... Julio E Ayala
    Research Article Updated

    The canonical target of the glucagon-like peptide-1 receptor (GLP-1R), Protein Kinase A (PKA), has been shown to stimulate mechanistic Target of Rapamycin Complex 1 (mTORC1) by phosphorylating the mTOR-regulating protein Raptor at Ser791 following β-adrenergic stimulation. The objective of these studies is to test whether GLP-1R agonists similarly stimulate mTORC1 via PKA phosphorylation of Raptor at Ser791 and whether this contributes to the weight loss effect of the therapeutic GLP-1R agonist liraglutide. We measured phosphorylation of the mTORC1 signaling target ribosomal protein S6 in Chinese Hamster Ovary cells expressing GLP-1R (CHO-Glp1r) treated with liraglutide in combination with PKA inhibitors. We also assessed liraglutide-mediated phosphorylation of the PKA substrate RRXS*/T* motif in CHO-Glp1r cells expressing Myc-tagged wild-type (WT) Raptor or a PKA-resistant (Ser791Ala) Raptor mutant. Finally, we measured the body weight response to liraglutide in WT mice and mice with a targeted knock-in of PKA-resistant Ser791Ala Raptor. Liraglutide increased phosphorylation of S6 and the PKA motif in WT Raptor in a PKA-dependent manner but failed to stimulate phosphorylation of the PKA motif in Ser791Ala Raptor in CHO-Glp1r cells. Lean Ser791Ala Raptor knock-in mice were resistant to liraglutide-induced weight loss but not setmelanotide-induced (melanocortin-4 receptor-dependent) weight loss. Diet-induced obese Ser791Ala Raptor knock-in mice were not resistant to liraglutide-induced weight loss; however, there was weight-dependent variation such that there was a tendency for obese Ser791Ala Raptor knock-in mice of lower relative body weight to be resistant to liraglutide-induced weight loss compared to weight-matched controls. Together, these findings suggest that PKA-mediated phosphorylation of Raptor at Ser791 contributes to liraglutide-induced weight loss.

    1. Cell Biology
    2. Developmental Biology

    Cylicins are a structural component of the sperm calyx being indispensable for male fertility in mice and human

    Simon Schneider, Andjela Kovacevic ... Hubert Schorle
    Research Article

    Cylicins are testis-specific proteins, which are exclusively expressed during spermiogenesis. In mice and humans, two Cylicins, the gonosomal X-linked Cylicin 1 (Cylc1/CYLC1) and the autosomal Cylicin 2 (Cylc2/CYLC2) genes, have been identified. Cylicins are cytoskeletal proteins with an overall positive charge due to lysine-rich repeats. While Cylicins have been localized in the acrosomal region of round spermatids, they resemble a major component of the calyx within the perinuclear theca at the posterior part of mature sperm nuclei. However, the role of Cylicins during spermiogenesis has not yet been investigated. Here, we applied CRISPR/Cas9-mediated gene editing in zygotes to establish Cylc1- and Cylc2-deficient mouse lines as a model to study the function of these proteins. Cylc1 deficiency resulted in male subfertility, whereas Cylc2-/-, Cylc1-/yCylc2+/-, and Cylc1-/yCylc2-/- males were infertile. Phenotypical characterization revealed that loss of Cylicins prevents proper calyx assembly during spermiogenesis. This results in decreased epididymal sperm counts, impaired shedding of excess cytoplasm, and severe structural malformations, ultimately resulting in impaired sperm motility. Furthermore, exome sequencing identified an infertile man with a hemizygous variant in CYLC1 and a heterozygous variant in CYLC2, displaying morphological abnormalities of the sperm including the absence of the acrosome. Thus, our study highlights the relevance and importance of Cylicins for spermiogenic remodeling and male fertility in human and mouse, and provides the basis for further studies on unraveling the complex molecular interactions between perinuclear theca proteins required during spermiogenesis.

    1. Cell Biology
    2. Structural Biology and Molecular Biophysics

    Baited reconstruction with 2D template matching for high-resolution structure determination in vitro and in vivo without template bias

    Bronwyn A Lucas, Benjamin A Himes, Nikolaus Grigorieff
    Research Advance

    Previously we showed that 2D template matching (2DTM) can be used to localize macromolecular complexes in images recorded by cryogenic electron microscopy (cryo-EM) with high precision, even in the presence of noise and cellular background (Lucas et al., 2021; Lucas et al., 2022). Here, we show that once localized, these particles may be averaged together to generate high-resolution 3D reconstructions. However, regions included in the template may suffer from template bias, leading to inflated resolution estimates and making the interpretation of high-resolution features unreliable. We evaluate conditions that minimize template bias while retaining the benefits of high-precision localization, and we show that molecular features not present in the template can be reconstructed at high resolution from targets found by 2DTM, extending prior work at low-resolution. Moreover, we present a quantitative metric for template bias to aid the interpretation of 3D reconstructions calculated with particles localized using high-resolution templates and fine angular sampling.