
==== Front
Proc Natl Acad Sci U S A
Proc Natl Acad Sci U S A
PNAS
Proceedings of the National Academy of Sciences of the United States of America
0027-8424
1091-6490
National Academy of Sciences

38285943
202314309
10.1073/pnas.2314309121
videoVideoresearch-articleResearch Articlecell-bioCell Biology409
Biological Sciences
Cell Biology
A novel cysteine-rich adaptor protein is required for mucin packaging and secretory granule stability in vivo
Zhang Liping a https://orcid.org/0000-0001-8637-4975

Muirhead Kayla J. a b
Syed Zulfeqhar A. a c https://orcid.org/0000-0002-5825-9782

Dimitriadis Emilios K. d https://orcid.org/0000-0002-2089-5361

Ten Hagen Kelly G. Kelly.Tenhagen@nih.gov
a 1 https://orcid.org/0000-0002-2057-8120

aDevelopmental Glycobiology Section, National Institute of Dental and Craniofacial Research, NIH, Bethesda, MD 20892-4370
bAmbry Genetics, Aliso Viejo, CA 92656
cElectron Microscopy Core Facility, National Heart, Lung and Blood Institute, NIH, Bethesda, MD 20892
dTrans-NIH Resource on Biomedical Engineering and Physical Science, National Institute of Biomedical Imaging and Bioengineering, NIH, Bethesda, MD 20892
1To whom correspondence may be addressed. Email: Kelly.Tenhagen@nih.gov.
Edited by Carolyn Bertozzi, Stanford University, Stanford, CA; received August 18, 2023; accepted January 2, 2024

29 1 2024
6 2 2024
29 7 2024
121 6 e231430912118 8 2023
02 1 2024
Copyright © 2024 the Author(s). Published by PNAS.
2024
https://creativecommons.org/licenses/by-nc-nd/4.0/ This article is distributed under Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND).

Significance

Mucins are large proteins that protect epithelial surfaces throughout the body. Changes in mucin production and secretion are associated with colon cancer, inflammatory bowel disease, and cystic fibrosis. Here, we found that proper packaging of these large proteins prior to secretion is dependent on a small cysteine-rich adaptor protein. We show that in the absence of this protein, mucins are not properly compacted, resulting in secretory granules that are bloated and more fragile. Understanding the factors that are required for proper mucin biosynthesis, packaging, and secretion may provide insight into novel treatment strategies for diseases of the respiratory and digestive tracts.

Mucins are large, highly glycosylated extracellular matrix proteins that line and protect epithelia of the respiratory, digestive, and urogenital tracts. Previous work has shown that mucins form large, interconnected polymeric networks that mediate their biological functions once secreted. However, how these large matrix molecules are compacted and packaged into much smaller secretory granules within cells prior to secretion is largely unknown. Here, we demonstrate that a small cysteine-rich adaptor protein is essential for proper packaging of a secretory mucin in vivo. This adaptor acts via cysteine bonding between itself and the cysteine-rich domain of the mucin. Loss of this adaptor protein disrupts mucin packaging in secretory granules, alters the mobile fraction within granules, and results in granules that are larger, more circular, and more fragile. Understanding the factors and mechanisms by which mucins and other highly glycosylated matrix proteins are properly packaged and secreted may provide insight into diseases characterized by aberrant mucin secretion.

mucin
secretion
salivary glands
glycosylation
secretory granules
HHS | NIH | National Institute of Dental and Craniofacial Research (NIDCR) 100000072 Z01-DE-000713 Liping ZhangKayla J. MuirheadZulfeqhar A. SyedEmilios K DimitriadisKelly G Ten Hagen HHS | NIH | National Institute of Dental and Craniofacial Research (NIDCR) 100000072 ZIC DE000750-01 Liping ZhangKayla J. MuirheadZulfeqhar A. SyedEmilios K DimitriadisKelly G Ten Hagen
==== Body
pmcRegulated secretion is a highly ordered process where proteins destined to be secreted into the extracellular space are packaged into membranous secretory vesicles within cells and await a signal that then initiates vesicle fusion with the plasma membrane, followed by secretion of vesicle contents. While some cells secrete smaller proteins and peptides, others are responsible for secreting large, heavily modified, and multimerized extracellular matrix (ECM) proteins (1–6). The details of how these large proteins are targeted to and packaged within small secretory granules remain largely unknown.

Mucins are heavily O-glycosylated ECM proteins that form protective barriers, called mucous layers, along the epithelium in digestive, respiratory, and reproductive systems to protect tissues from damage and infection and maintain homeostasis (4, 7–11). Abnormal mucin secretion and mucous layer formation are associated with certain human diseases, such as colon cancer, cystic fibrosis, and inflammatory bowel disease (7, 12, 13). While the gel-forming mucins are divergent in sequence, all have highly O-glycosylated, repetitive domains rich in proline, threonine, and serine (PTS domains) along with cysteine-rich regions that are responsible for dimer and multimer formation (8). Upon secretion, these multimerized mucins are proposed to expand in a regulated fashion, aided by hydration of the O-linked glycans, to form net-like structures that line internal epithelia (4, 5, 14–17). The size and complexity of these mucin matrices lead to many questions regarding how cells organize and compact these macromolecules into the secretory granules before secretion occurs.

Previous studies have demonstrated that distinct mucins undergo regulated restructuring to form compacted structures within secretory granules prior to secretion (18). Using Drosophila salivary glands (SGs), which synthesize and secrete multiple highly O-glycosylated mucins in response to developmentally regulated hormone pulses (19–24), we demonstrated that the mucins Sgs1 and Sgs3 form disc-like structures and bundled filaments, respectively, within secretory granules (18). This compaction process was dependent on genes controlling pH and calcium (18). Likewise, studies using portions of mammalian mucins have demonstrated that in the presence of calcium and an acidic pH, unique compacted structures can be formed in vitro, highlighting the importance of calcium and pH in mucin restructuring (5, 16, 17, 25). Other studies have suggested that other factors, such as small peptides might also be involved in mucin binding, multimer formation, or packaging (26).

Here, we identify a small cysteine-rich peptide that mediates mucin restructuring in vivo. Loss of this adaptor causes disruption of mucin compaction, resulting in secretory granules that have an altered mobile fraction, are larger and rounder in appearance, and are more fragile. This study highlights the biological importance of proper mucin compaction and identifies a unique factor essential for this process.

Results

Loss of a Small Peptide Results in Disrupted Mucin Restructuring and Bloated Secretory Granules.

Previous work from our group demonstrated that the highly glycosylated mucins, Sgs1 and Sgs3, undergo regulated restructuring to form compacted structures during secretory granule maturation (18). Sgs1 forms electron-lucent disc-like structures that decorate the outside of an electron-dense area and Sgs3 folds into parallel fibers that are arranged into bundles (18). The Drosophila salivary gland expresses a number of Sgs genes (Sgs1, Sgs3, Sgs4, Sgs5, Sgs5bis, Sgs7, Sgs8, and Eig71Ee), whose regulation is hormonally controlled during development (27). These genes encode highly O-glycosylated mucins rich in serine, threonine, and proline, as well as proteins much smaller in size whose function is currently unknown. The highly O-glycosylated mucins are produced in a regulated fashion to serve as the secreted “glue” that allows the larvae to adhere to a safe surface prior to undergoing metamorphosis, and thus are extremely important in the fly lifecycle. In an effort to determine additional factors required for mucin packaging/compaction within secretory granules, RNAi screens were performed to identify genes that affect secretory granule morphology. We visualized secretory granule formation and morphology with flies expressing the Sgs3-GFP (N-terminal portion of Sgs3 including regulatory regions linked to GFP) transgene, which was used previously to document factors essential for this process (18–22, 24, 28–30). Interestingly, RNAi knockdown of Sgs7 (using Sgs7 inverted repeat lines crossed with the c135-Gal4 driver line, which expresses in the larval SG) resulted in a dramatic alteration in secretory granule morphology (Fig. 1A). Measurements of secretory granule circularity and area revealed that loss of Sgs7 resulted in granules that were larger and more circular than WT (Fig. 1 A and B and SI Appendix, Fig. S1A). qPCR of all Sgs genes verified that Sgs7RNAiresulted in a dramatic and specific reduction in Sgs7 gene expression (Fig. 1C). Additionally, Coomassie staining (Fig. 1D) and immunoblotting using an Sgs7-specific antibody (Fig. 1E and SI Appendix, Fig. S1B) revealed the specific loss of a small protein of ~13 kDa in Sgs7RNAi SGs. This bloated secretory granule phenotype was also seen using transposon insertion mutations in Sgs7 (Sgs7Bl#24539/BL57993), which were examined in a background that expressed GFP in the cytoplasm to visualize secretory granule morphology indirectly (Fig. 1F). Sgs7Bl#24539/Bl#57993, which disrupts Sgs7 gene (SI Appendix, Fig. S1C) and protein expression (SI Appendix, Fig. S1E), also resulted in very circular, bloated secretory granules, further supporting a role for Sgs7 in secretory granule morphology and maturation (Fig. 1F and SI Appendix, Fig. S1F). Finally, we rescued the bloated secretory granule phenotype by expressing exogenous Sgs7 in the SGs of the Sgs7Bl#24539/Bl#57993 mutant (Sgs7Bl#24539/Bl#57993; c135>UAS-Sgs7-V5) (Fig. 1G and SI Appendix, Fig. S1F). Taken together, these data indicate that loss of Sgs7 results in a bloated secretory granule phenotype.

Fig. 1. Loss of Sgs7 results in bloated secretory granules and disrupted structure of Sgs3. (A) Compared with WT SGs, RNAi to Sgs7 (Sgs7RNAi) resulted in very round, bloated secretory granules in secretory cells. Green is Sgs3-GFP packaged in secretory granules. (B) The secretory granule circularities of WT and Sgs7RNAi were quantified as described previously (24). 200 granules of each genotype were analyzed. ****P < 0.0001. (C) Expression of each Sgs gene was quantitated in WT and Sgs7RNAi SGs by qPCR. Among six Sgs genes, only Sgs7 expression was dramatically decreased in Sgs7RNAi SGs, while others were unchanged. (D) Coomassie blue staining of proteins from WT and Sgs7RNAi SGs showed a specific decrease in a protein with the predicted size of Sgs7. (E) Western blot with Sgs7 antibody further confirmed Sgs7 protein level was dramatically decreased in Sgs7RNAi SGs. (F) Transposon mutations of Sgs7 (Sgs7Bl#24539/Bl#57993) also resulted in bloated secretory granules, similar to Sgs7RNAi. Green is GFP expressed in the cytoplasm of secretory cells. (G) Expression of Sgs7 in Sgs7 mutant SGs (Sgs7Bl#24539/Bl#57993; c135>UAS-Sgs7-V5) rescued the abnormal secretory granule phenotype. (H) TEM on secretory granules of WT and Sgs7RNAi. In WT secretory granules, Sgs3 is present as highly ordered bundled filaments (red arrowhead). In Sgs7RNAi, no bundled filaments were seen (black arrowhead) but other structures seen in WT were present at the periphery. (I) Western blot probed with Sgs3 antibody demonstrated that Sgs3 protein was present at similar levels in WT and Sgs7RNAi SGs, (black arrowheads). (Scale bars are 20 µm for (A), (F), and (G) and 600 nm for (H).)

To begin to understand how the loss of Sgs7 is resulting in bloated secretory granules, we next performed transmission electron microscopy (TEM) on SGs from WT and Sgs7RNAi larvae. Interestingly, Sgs7RNAi resulted in the complete and specific loss of the highly ordered Sgs3 bundled filaments (Fig. 1H and SI Appendix, Fig. S1D). While the Sgs3 filaments were lost, the Sgs1 discs were still present and associated with the electron-dense regions as is seen in WT (Fig. 1H). Interestingly, however, the electron-dense/Sgs1 regions appeared on the periphery of the secretory granule rather than being present throughout the granule (Fig. 1H), suggesting that the unfolded, expanded Sgs3 may be pushing these other structures to the periphery. The loss of the Sgs3 filaments was not due to the loss of Sgs3 itself, as the Sgs3 protein was still detected by western blots at equivalent levels in WT and Sgs7RNAi SGs (Fig. 1). These results suggest that the loss of Sgs7 disrupts the regulated restructuring of Sgs3, leading to expansion of Sgs3, which causes rounder and larger secretory granules in SGs.

Cysteine-Rich Domains Are Required for Sgs7 and Sgs3 Interactions.

Sgs7 encodes a small unglycosylated protein (~8 kDa predicted molecular mass) that is co-expressed with the highly glycosylated Sgs3 (~32 kDa predicted molecular mass as an unglycosylated monomer) during the third instar larval period in response to ecdysone pulses (31–34). To understand how Sgs7 may be affecting the folded structure of Sgs3, we examined where each protein resides within WT SGs. We created a Sgs7-v5 transgenic line under the control of the UAS promoter in the background of the Sgs3-GFP line and expressed Sgs7 using a Gal4 driver that drives expression in the SG (c135>UAS-Sgs7-V5; Sgs3-GFP). As shown in Fig. 2A, Sgs7 localized to secretory granules with Sgs3. Additionally, western blots of SG lysates as well as the secreted material from SGs (glue plug) revealed that Sgs7 was also present in the secreted fraction (Fig. 2B). These results suggest that Sgs7 is packaged and secreted from the same secretory granules as Sgs3.

Fig. 2. Sgs7 and Sgs3 colocalize in secretory granules. (A) Staining of SGs with the anti-V5 antibody showed that Sgs7 (red) co-localizes with truncated Sgs3-GFP transgene (green) in secretory granules. (Scale bar, 20 µm.) (B) Western blot probed with the Sgs7 antibody showed that Sgs7 protein is present in secretory cells before secretion (SG lysates) and also within the secreted material (glue plug). (C) Sgs7 (green) expressed in Drosophila S2R+ cells localized to the Golgi apparatus, as detected by the cis-Golgi marker GM130 (red) and the trans-Golgi marker Golgin245 staining (magenta). DNA staining is shown in blue. (D) Full-length Sgs3 (green) expressed in S2R+ cells formed secretory granule-like structures and did not localize to Golgi apparatus (GM130 or Golgin245, magenta). (E) When co-expressed with Sgs3 (green), Sgs7 (red) no longer localizes to the Golgi (Golgin245, magenta) but co-localized with Sgs3 in vesicle-like structures. (Scale bar, 5 µm.)

To determine what governs the colocalization of Sgs7 and Sgs3, we next expressed Sgs3 and Sgs7 in Drosophila S2R+ cells (Fig. 2 C–E). Sgs7 expressed in these cells localized to regions of the Golgi apparatus, as demonstrated by colocalization with the Golgi markers GM130 and Golgin245 (Fig. 2C). When Sgs3 was expressed in these cells, it localized to round structures that were distinct from the Golgi, as there was no colocalization with GM130 or Golgin245 (Fig. 2D). Interestingly, when Sgs3 and Sgs7 were expressed in the same cells, Sgs7 colocalized with Sgs3 in the secretory granule-like structures (Fig. 2E). These results suggest that Sgs3 and Sgs7 interact and that Sgs3 can relocalize Sgs7 from the Golgi to secretory granule-like structures.

To identify potential regions of interaction between Sgs7 and Sgs3, we next performed deletions of select portions of Sgs3 (Fig. 3A). Deletion of the T-rich domain only (Sgs3∆T) or the repetitive, highly-O-glycosylated PTTTK domain only (Sgs3∆P) had no effect on colocalization of Sgs3 and Sgs7 (Fig. 3 A and B). Interestingly, Sgs3 lacking only the cysteine-rich C-terminal domain (Sgs3∆C) failed to colocalize with Sgs7 and also failed to form secretory granule-like structures (Fig. 3 A and B). Sgs3∆C resulted in Sgs3 signal being scattered throughout the cytoplasm in small puncta. However, constructs lacking both the highly-O-glycosylated PTTTK region and the cysteine-rich C-terminal region of Sgs3 (Sgs3∆PC) were able to form secretory granule-like structures but still lacked colocalization with Sgs7, suggesting that the C-terminal region is required for secretory granule formation only when the highly O-glycosylated region is present (Fig. 3 A and B). Taken together, these results suggest that Sgs3 and Sgs7 colocalize via the cysteine-rich C-terminal region of Sgs3. Additionally, the absence of the C-terminal region in the presence of the highly-O-glycosylated PTTTK region results in Sgs3 being dispersed throughout the cytoplasm rather than in secretory granule-like structures.

Fig. 3. Sgs3 and Sgs7 interact through cysteine bonding. (A) Diagram of WT Sgs3 and different domain deletions used for co-expression with Sgs7. N, N-terminal domain; T-rich, threonine-rich domain; PTTTK, repetitive mucin domain; C, cysteine-rich C-terminal domain. (B) The deletions of the T-rich domain (Sgs3∆T, green) or the PTTTK domain (Sgs3∆P, green) co-localized with Sgs7 (red). Deletions that removed the C-terminal region (Sgs3∆PC or Sgs3∆C, green) resulted in loss of colocalization with Sgs7 (red). Additionally, Sgs3∆C disrupted the formation of large secretory granule-like structures. (C) Alignment of Sgs7 and Sgs3 revealed considerable homology within the cysteine-rich regions of both proteins. Cysteines are highlighted. Mutating the cysteines within the cysteine-rich domain of Sgs3 (Sgs3M, green) (D) or Sgs7 (Sgs7M, red) (E) resulted in the loss of colocalization of these proteins. Additionally, Sgs3M was unable to form granule-like structures. (F) Western blots demonstrate that Sgs3 (Sgs3-FLAG) was co-precipitated with WT Sgs7 (Sgs7-AV5) but not with the cysteine mutated Sgs7 (Sgs7M-AV5). (G) Mutated Sgs7 (Sgs7M) was unable to rescue the bloated secretory granules (Sgs7Bl#24539/Bl#57993; c135>UAS-Sgs7M-V5). Green is the GFP expressed in the cytoplasm of secretory cells. (H) AlphaFold2 was used to model complex formation between Sgs3 and Sgs7. Formation of tetramers by two Sgs7 molecules and two Sgs3 molecules. Red is intra-disulfide-bond and magenta is inter-disulfide-bond. White dotted box shows a magnified view of the intra- and inter-disulfide-bonds formed in the tetramer. There is one intra-disulfide-bond in each Sgs7 (C31-C38) and Sgs3 (C264-C271), two inter-disulfide-bonds between two Sgs7 molecules (two C41-C65 bonds) and Sgs3 (two C274-C298 bonds), and four inter-disulfide-bonds between two Sgs7 and two Sgs3 molecules (two Sgs7C67-Sgs3C281 bonds and two Sgs7C48-Sgs3C300 bonds). See also Movie S1. (Scale bars are 5 µm for (B), (D), and (E) and 20 µm for (G).)

Sgs7 consists of a cysteine-rich region that shares considerable homology with the C-terminal cysteine-rich domain of Sgs3 (Fig. 3C), suggesting that these two regions may physically interact. We next set out to determine whether the colocalization of Sgs3 and Sgs7 was driven by their cysteine-rich domains. We mutated each cysteine (C) in this region of Sgs3 to alanine (A) and again looked for colocalization with Sgs7 in Drosophila S2R+ cells. As shown in Fig. 3D, mutating the eight cysteines within the C-terminal cysteine-rich domain of Sgs3 (Sgs3M) abrogated colocalization with Sgs7. This also resulted in the disruption of secretory granule formation of Sgs3, suggesting that the cysteines within the C-terminal cysteine-rich domain of Sgs3 are not only responsible for interacting with Sgs7 but also responsible for secretory granule formation when the highly O-glycosylated regions are present. We then performed the reciprocal experiment, by mutating the cysteines within Sgs7 to create the construct Sgs7M. As shown in Fig. 3E, mutating the cysteines within Sgs7 abrogated its colocalization with Sgs3, suggesting that the cysteines within Sgs7 are responsible for interacting with Sgs3. Additionally, to study the interaction of Sgs7 and Sgs3, we performed immunoprecipitation using V5-tagged Sgs7 or Sgs7M (containing the cysteine mutations) to see whether Sgs3 can be co-immunoprecipitated. As shown in Fig. 3F, Sgs3 was co-immunoprecipitated with wild-type Sgs7 but not Sgs7M. We next examined whether Sgs7M could rescue the bloated secretory granule phenotype of Sgs7Bl#24539/Bl#57993. As shown in Fig. 3G and quantitated in SI Appendix, Fig. S1F, Sgs7M was not able to rescue the bloated secretory granule phenotype of Sgs7Bl#24539/Bl#57993.

Using AlphaFold2 for generating structural predictions of Sgs7 and Sgs3, we obtained best-fit predictions that implicate the formation of tetramers composed of two Sgs7 molecules and two Sgs3 molecules linked together through the cysteine-rich domains (Fig. 3H and Movie S1). Indeed, the size of Sgs7 on western blots suggests dimer/multimer formation that is dependent on cysteines (SI Appendix, Fig. S2 A and B). Taken together, our results indicate that Sgs3 and Sgs7 interface with one another through cysteine-mediated interactions and these interactions are required for the biological function of Sgs7-mediated restructuring/compaction of Sgs3.

Disruption in Mucin Compaction Results in Fragile Secretory Granules.

We noticed when performing experiments that Sgs7RNAi SGs often displayed ruptured secretory granules, suggesting that altered mucin compaction may influence secretory granule stability. We therefore performed experiments to quantitate differences in secretory granule biophysical properties and stability upon loss of Sgs7. To interrogate how the disruption in mucin restructuring may be affecting the internal dynamics of the secretory granules, we performed florescence recovery after photobleaching (FRAP) experiments to examine the internal mobility of granular contents. We used the Sgs3-GFP line, that expresses a truncated version of Sgs3 (under the control of its own promoter) fused to GFP (31, 32). This truncated Sgs3 lacks most of the glycosylated region and the cysteine-rich regions predicted to mediate intermolecular associations with Sgs7 and thus localizes throughout the entire secretory granule (Fig. 1A, WT). We photobleached the central regions of individual secretory granules containing the truncated Sgs3-GFP in both WT and Sgs7RNAi third instar larvae and measured the fluorescence intensity before and after photobleaching (Fig. 4A). The averaged normalized FRAP curve for 20 granules of WT or Sgs7RNAi were fit to single exponential equations (Fig. 4B), which allows one to estimate the florescence recovery half-time (t1/2) (Fig. 4C). While there was no significant difference in t1/2 between WT and Sgs7RNAi (Fig. 4C) (indicating no difference in fluorescence recovery time), there was a significant difference in the mobile fraction between WT and Sgs7RNAi (Fig. 4D). As shown in Fig. 4D, loss of Sgs7 results in a significant increase in the mobile fraction within secretory granules relative to WT, suggesting a less structured/more fluid content.

Fig. 4. Sgs7RNAi results in altered intragranular mobility and fragile secretory granules. (A) FRAP analysis of secretory granules of live SGs from WT or Sgs7RNAi larvae. A small circular region of interest (ROI) in a secretory granule containing Sgs3-GFP (shown in white) was defined, and the fluorescence intensity change in ROI before and after the bleach (Fpre-bleach and Fpost-bleach) was monitored over time. (B) Fluorescence intensity values of 20 secretory granules from WT or Sgs7RNAi were plotted and used to generate values for recovery half times (t1/2) (C) and mobile fraction (Mf) (D), which were calculated based on the equations described in Snapp et al. (35). (E) AFM on secretory granules from WT or Sgs7RNAi larvae to calculate elastic modulus of granule contents. The beginning and end time of incubation of WT (F) and Sgs7RNAi (G) SGs in a hypotonic solution. See also Movies S2 and S3. Note the dark regions, indicating ruptured secretory granules, in the end time point of Sgs7RNAi. (Scale bars are 5 µm for (A) and 50 µm for (F) and (G).)

We next performed atomic force microscopy (AFM) to quantify the elastic modulus of the contents of secretory granules isolated from WT and Sgs7RNAi SGs as described previously (33). Isolated secretory granules were compressed using a flat, tipless AFM cantilever and changes in secretory granule radii were measured at various forces (33). Consistent with the increased mobility seen for Sgs7RNAi, we observed a significant decrease in the elastic modulus (a measure of the elasticity of granule contents) of Sgs7RNAi granules relative to WT (Fig. 4E). Taken together, our results suggest that the unstructured Sgs3 mucin results in secretory granules with a more mobile, fluid content that is less elastic in nature.

Finally, to assess secretory granule stability under different osmotic conditions, we incubated WT and Sgs7RNAi SGs in a hypotonic solution and imaged granules over time (30 s). Interestingly, granules from Sgs7RNAi SGs began to rupture almost immediately (indicated by a GFP flash followed by a dark region where the secretory granule used to be) in the hypotonic solution whereas those from WT SGs were stable over the course of the experiment (Fig. 4 F and G and Movies S2 and S3). Taken together, our results identify a unique cysteine-rich adaptor protein required for the compaction of a specific mucin. Moreover, we demonstrate that proper mucin compaction is necessary to ensure secretory granule stability in vivo. Our results have implications for mucin packaging in human secretory tissues.

Discussion

Here, we identify a unique cysteine-rich adaptor protein that is essential for proper mucin restructuring and compaction, providing information regarding how large ECM proteins are compacted prior to secretion. Moreover, we demonstrate that disruption of this compaction resulted in secretory granules that were less stable and ruptured upon changes in the osmotic environment. Indeed, Sgs7RNAi SGs incubated in hypotonic conditions displayed secretory granules that ruptured almost immediately, whereas WT granules were resistant to osmotic changes over the course of the experiment. Additionally, FRAP and AFM experiments indicate that the biophysical characteristics of the secretory granule contents were dramatically altered in the absence of Sgs7, when Sgs3 was not properly compacted/restructured. The data are consistent with a model where the unstructured, expanded Sgs3 mucin results in granule contents being more fluid, with less structure. The expansion of Sgs3, with its repetitive PTS domains rich in O-linked GalNAc typical of mucins, may also explain the susceptibility to osmotic changes. It is well known that the glycosylated regions of mucins are readily hydrated leading to unique rheological and viscoelastic properties upon secretion. Our results suggest that mucin compaction not only allows these large multimeric proteins to fit within secretory granules but further protects against premature hydration or expansion prior to secretion. Taken together, our results demonstrate that a small adaptor protein is required for ordered mucin restructuring, which is necessary for secretory granule integrity and stability in vivo.

In silico modeling using AlphaFold2 (https://colab.research.google.com/github/sokrypton/ColabFold/blob/main/AlphaFold2.ipynb) suggests that Sgs7 and Sgs3 form a tetrameric structure by virtue of inter- and intramolecular disulfide bonding. Sgs7 and Sgs3 C-terminal domains share similar sequences, including eight cysteines. While it is known that the cysteine-rich regions of mammalian mucins are important for the formation of disulfide-bonded dimers and multimers (5, 14–17, 25), little is known regarding the binding of other cysteine-rich proteins. Experimentally, we found that Sgs7 could rescue the compaction of Sgs3 only when the C-terminal cysteines of both proteins were present. Cell culture experiments further confirmed that the cysteines in both Sgs7 and Sgs3 are necessary for the interaction between these two proteins. In both salivary glands and cell culture, we found wild-type Sgs7 can form dimers and multimers, while the cysteine mutations just appear as monomers (SI Appendix, Fig. S2 A and B). These results suggest disulfide bond formation between Sgs7 molecules may help stabilize the interaction between Sgs7 and Sgs3. Similar to Sgs3 in SGs, we found that wild-type Sgs3 in cell culture can form granule-like structures that interact with Sgs7. However, upon loss of C-terminal cysteine-rich region of Sgs3 (Sgs3∆C), the structures disappeared, and proteins dispersed in cytoplasm, suggesting that the cysteine-rich region is necessary for proper Sgs3 structure formation. Interestingly, we found constructs that deleted both the heavily O-glycosylated repeated PTTTK region and cysteine-rich region (Sgs3∆PC) were able to form granule-like structures, suggesting that the cysteine-rich region is specifically required to mediate the organization of proteins containing the repeated, highly glycosylated PTTTK region. Our results suggest that small cysteine-rich proteins may be uniquely required to help package proteins with large repetitive domains. We look forward to versions of AlphaFold2 that will be able to include the dense glycosylation of these repetitive regions in modeling predictions in the future.

Interestingly, we found loss of Sgs7 disrupted the organized structures of one specific mucin, Sgs3, within secretory granules, while the restructuring of other mucins was unaffected. Sgs3 and Sgs7 form a tightly linked cluster on the 3L chromosomal arm, suggesting that there has been evolutionary pressure to ensure the co-expression of these two proteins (34, 36, 37). This further suggests that there may be other adaptors that are specific for different mucins. Indeed, other small, secreted glue proteins have been characterized from SGs, such as Sgs5, Sgs5bis, and Sgs8 (37, 38). In the future, it will be very interesting to investigate whether these small proteins are involved in the restructuring and compaction of mucins in secretory granules.

Previous work with mammalian mucins has also found evidence for non-mucin proteins upon treatment with reducing agents (26). While we did not find obvious sequence homologs to Sgs7, mammalian proteins with similar functional/structural regions exist. Indeed, many of the identified interacting mammalian proteins had properties similar to Sgs7, including a high density of cysteines and small size (26). In fact, there is evidence that higher-order mucin multimers and mucin rheological properties are influenced by binding with these non-mucin proteins (26). The trefoil factor family peptides (TFF1, TFF2, and TFF3) are cysteine-rich small proteins co-expressed and secreted with mucins (39). The TFF domain has six conserved cysteines, forming three intramolecular disulfide bonds and producing a three-clover-leaf structure, which results in structurally and functionally stable TFF peptides (39, 40). TFF gene knock-out mice showed altered structure of mucus layers and increased mucosal injury and inflammation, suggesting that TFF peptides facilitate mucus layer organization to play protective and mucosal healing roles in the digestive system (39, 40). Additionally, both in vivo and in vitro studies found that TFF peptides increase the viscosity and elasticity of mucus gels and enhance epithelium restitution rates (41, 42). Interestingly, one study in Xenopus laevis found that the ortholog of TFF2 is packaged with the MUC6 ortholog in the same secretory granules within the esophageal goblet cells (43). In the future, it will be informative to investigate whether TFFs are also involved in mucin organization within secretory granules as well as what role Sgs7 may be playing once secreted with Sgs3.

Finally, Sgs7 was not only present with Sgs3 in the SGs but was also secreted along with Sgs3. Sgs7 is predicted to be an antimicrobial peptide (AMP) by several online AMP prediction tools (antimicrobial peptide scanner, https://www.dveltri.com/ascan/v2/; CAMPR3, https://www.camp.bicnirrh.res.in/prediction.php; DBAASP, https://dbaasp.org/tools?page=property-calculation), suggesting that it may not only aid in mucin compaction prior to secretion but also serve a protective function once secreted. Indeed, evidence exists for the role of mucus/mucins in optimizing the activity of antimicrobial peptides (44). Whether Sgs3 and Sgs7 maintain interactions upon secretion and how they may function as a complex extracellularly will be a future area of investigation. Understanding the factors involved in mucin biosynthesis, packaging, secretion, and function will be key to designing intelligent approaches to restoring aberrant mucin synthesis as well as designing novel antimicrobial strategies.

Materials and Methods

Full details on the materials and methods are described in SI Appendix. The methods include fly strains and genetics; gene cloning and constructs used; antibody preparation; staining Drosophila tissues and cells; live SG imaging and FRAP; AFM; real-time PCR; western blotting; and immunoprecipitation.

Supplementary Material

Appendix 01 (PDF)

Click here for additional data file.

Movie S1. Three-dimensional movie of disulfide bonds formed between tetramers consisting of Sgs7 and Sgs3.

Movie S2. WT salivary gland expressing Sgs3-GFP was incubated in a hypotonic solution and granules were imaged over time (30 seconds).

Movie S3. Sgs7RNAi salivary gland expressing Sgs3-GFP was incubated in a hypotonic solution and granules were imaged over time (30 seconds).

We would like to thank our colleagues for many helpful discussions. We also thank the Bloomington Stock Center and the Vienna Drosophila RNAi Center for providing fly stocks, antibodies, and other reagents. This research was supported by the Intramural Research Program of the NIDCR at the NIH (Z01-DE-000713 to K.G.T.H.) and the NIDCR Imaging Core (ZIC DE000750-01).

Author contributions

L.Z., K.J.M., Z.A.S., and K.G.T.H. designed research; L.Z., K.J.M., Z.A.S., and E.K.D. performed research; L.Z. and K.J.M. contributed new reagents/analytic tools; L.Z., K.J.M., Z.A.S., E.K.D., and K.G.T.H. analyzed data; and L.Z. and K.G.T.H. wrote the paper.

Competing interests

The authors declare no competing interest.

Data, Materials, and Software Availability

All study data are included in the article and/or supporting information.

Supporting Information

This article is a PNAS Direct Submission.
==== Refs
1 I. Raote, V. Malhotra, Protein transport by vesicles and tunnels. J. Cell Biol. 218 , 737–739 (2019).30718263
2 K. M. Valentijn, J. E. Sadler, J. A. Valentijn, J. Voorberg, J. Eikenboom, Functional architecture of Weibel-Palade bodies. Blood 117 , 5033–5043 (2011).21266719
3 A. Ermund, S. Trillo-Muyo, G. C. Hansson, Assembly, release, and transport of airway mucins in pigs and humans. Ann. Am. Thorac. Soc. 15 , S159–S163 (2018).30431338
4 M. E. Johansson , Composition and functional role of the mucus layers in the intestine. Cell Mol. Life Sci. 68 , 3635–3641 (2011).21947475
5 D. Ambort , Calcium and pH-dependent packing and release of the gel-forming MUC2 mucin. Proc. Natl. Acad. Sci. U.S.A. 109 , 5645–5650 (2012).22451922
6 M. E. Johansson, G. C. Hansson, Mucus and the goblet cell. Dig. Dis. 31 , 305–309 (2013).24246979
7 G. C. Hansson, Mucus and mucins in diseases of the intestinal and respiratory tracts. J. Intern. Med. 285 , 479–490 (2019).30963635
8 G. C. Hansson, Mucins and the microbiome. Annu. Rev. Biochem. 89 , 769–793 (2020).32243763
9 T. Pelaseyed, G. C. Hansson, Membrane mucins of the intestine at a glance. J. Cell Sci. 133 , jcs240929 (2020).32169835
10 M. E. Johansson, G. C. Hansson, Immunological aspects of intestinal mucus and mucins. Nat. Rev. Immunol. 16 , 639–649 (2016).27498766
11 Y. Yao , Mucus sialylation determines intestinal host-commensal homeostasis. Cell 185 , 1172–1188.e1128 (2022).35303419
12 L. Arike, G. C. Hansson, The densely O-glycosylated MUC2 mucin protects the intestine and provides food for the commensal bacteria. J. Mol. Biol. 428 , 3221–3229 (2016).26880333
13 M. E. Johansson, H. Sjovall, G. C. Hansson, The gastrointestinal mucus system in health and disease. Nat. Rev. Gastroenterol. Hepatol. 10 , 352–361 (2013).23478383
14 M. Kesimer, A. M. Makhov, J. D. Griffith, P. Verdugo, J. K. Sheehan, Unpacking a gel-forming mucin: A view of MUC5B organization after granular release. Am. J. Physiol. Lung Cell Mol. Physiol. 298 , L15–22 (2010).19783639
15 H. E. Nilsson , Intestinal MUC2 mucin supramolecular topology by packing and release resting on D3 domain assembly. J. Mol. Biol. 426 , 2567–2579 (2014).24816392
16 C. Ridley , Assembly of the respiratory mucin MUC5B: A new model for a gel-forming mucin. J. Biol. Chem. 289 , 16409–16420 (2014).24778189
17 S. Trillo-Muyo , Granule-stored MUC5B mucins are packed by the non-covalent formation of N-terminal head-to-head tetramers. J. Biol. Chem. 293 , 5746–5754 (2018).29440393
18 Z. A. Syed, L. Zhang, D. T. Tran, C. K. E. Bleck, K. G. Ten Hagen, Regulated restructuring of mucins during secretory granule maturation in vivo. Proc. Natl. Acad. Sci. U.S.A. 119 , e2209750119 (2022).36252017
19 M. Ashburner, G. Richards, Sequential gene activation by ecdysone in polytene chromosomes of Drosophila melanogaster. III. Consequences of ecdysone withdrawal. Dev. Biol. 54 , 241–255 (1976).825403
20 G. Korge, Larval saliva in Drosophila melanogaster: Production, composition, and relationship to chromosome puffs. Dev. Biol. 58 , 339–355 (1977).407116
21 D. T. Tran, K. G. Ten Hagen, Real-time insights into regulated exocytosis. J. Cell Sci. 130 , 1355–1363 (2017).28302911
22 I. F. Zhimulev, N. N. Kolesnikov, Synthesis and secretion of mucoprotein glue in the salivary gland of Drosophila melanogaster. Wilehm Roux Arch. Dev. Biol. 178 , 15–28 (1975).28305064
23 D. T. Tran, A. Masedunskas, R. Weigert, K. G. Ten Hagen, Arp2/3-mediated F-actin formation controls regulated exocytosis in vivo. Nat. Commun. 6 , 10098 (2015).26639106
24 S. Ji , A molecular switch orchestrates enzyme specificity and secretory granule morphology. Nat. Commun. 9 , 3508 (2018).30158631
25 G. Javitt , Assembly mechanism of mucin and von Willebrand factor polymers. Cell 183 , 717–729.e716 (2020).33031746
26 O. W. Meldrum , Mucin gel assembly is controlled by a collective action of non-mucin proteins, disulfide bridges, Ca(2+)-mediated links, and hydrogen bonding. Sci. Rep. 8 , 5802 (2018).29643478
27 M. Monier, V. Courtier-Orgogozo, Drosophila glue: A promising model for bioadhesion. Insects 13 , 734 (2022).36005360
28 S. K. Beckendorf, F. C. Kafatos, Differentiation in the salivary glands of Drosophila melanogaster: Characterization of the glue proteins and their developmental appearance. Cell 9 , 365–373 (1976).825230
29 M. A. Mary Boyd, The hormonal control of salivary gland secretion in Drosophila melanogaster: Studies in vitro. J. Insect Physiol. 23 , 517–523 (1977).
30 D. T. Tran, K. G. Ten Hagen, Mucin-type O-glycosylation during development. J. Biol. Chem. 288 , 6921–6929 (2013).23329828
31 A. Biyasheva, T. V. Do, Y. Lu, M. Vaskova, A. J. Andres, Glue secretion in the Drosophila salivary gland: A model for steroid-regulated exocytosis. Dev. Biol. 231 , 234–251 (2001).11180965
32 J. Burgess , AP-1 and clathrin are essential for secretory granule biogenesis in Drosophila. Mol. Biol. Cell 22 , 2094–2105 (2011).21490149
33 V. Parpura, J. M. Fernandez, Atomic force microscopy study of the secretory granule lumen. Biophys J. 71 , 2356–2366 (1996).8913576
34 T. E. Crowley, M. W. Bond, E. M. Meyerowitz, The structural genes for three Drosophila glue proteins reside at a single polytene chromosome puff locus. Mol. Cell Biol. 3 , 623–634 (1983).6406838
35 E. L. Snapp, N. Altan, J. Lippincott-Schwartz, Measuring protein mobility by photobleaching GFP chimeras in living cells. Curr. Protoc. Cell Biol. Chapter 21, Unit 21, 19 , 21.1.1–21.1.24 (2003).
36 M. D. Garfinkel, R. E. Pruitt, E. M. Meyerowitz, DNA sequences, gene regulation and modular protein evolution in the Drosophila 68C glue gene cluster. J. Mol. Biol. 168 , 765–789 (1983).6411930
37 F. Borne, R. J. Kulathinal, V. Courtier-Orgogozo, Glue genes are subjected to diverse selective forces during Drosophila development. Genome Biol. Evol. 13 , evab248 (2021).34788814
38 E. M. Shore, G. M. Guild, Larval salivary gland secretion proteins in Drosophila structural analysis of the Sgs-5 gene. J. Mol. Biol. 190 , 149–158 (1986).3098981
39 W. Hoffmann, Trefoil factor family (TFF) peptides and their diverse molecular functions in mucus barrier protection and more: Changing the paradigm. Int. J. Mol. Sci. 21 , 4535 (2020).32630599
40 E. Aihara, K. A. Engevik, M. H. Montrose, Trefoil factor peptides and gastrointestinal function. Annu. Rev. Physiol. 79 , 357–380 (2017).27992733
41 M. A. Jarva , Trefoil factors share a lectin activity that defines their role in mucus. Nat. Commun. 11 , 2265 (2020).32404934
42 L. Thim, F. Madsen, S. S. Poulsen, Effect of trefoil factors on the viscoelastic properties of mucus gels. Eur. J. Clin. Invest. 32 , 519–527 (2002).12153553
43 H. Schwarz, W. Hoffmann, Subcellular localization of the TFF peptides xP1 and xP4 in the Xenopus laevis gastric/esophageal mucosa: Different secretion modes reflecting diverse protective functions. Int. J. Mol. Sci. 21 , 761 (2020).31979419
44 J. F. Pierre , Peptide YY: A Paneth cell antimicrobial peptide that maintains Candida gut commensalism. Science 381 , 502–508 (2023).37535745
