RESEARCH ARTICLE

A genome-wide CRISPR screen reconciles the role of N-linkedglycosylation in galectin-3 transport to the cell surfaceSarah E. Stewart1,*, Sam A. Menzies2,*, Stephanie J. Popa1, Natalia Savinykh3, Anna Petrunkina Harrison3,Paul J. Lehner2 and Kevin Moreau1,‡

ABSTRACTGalectins are a family of lectin binding proteins expressed bothintracellularly and extracellularly. Galectin-3 (Gal-3, also known asLGALS3) is expressed at the cell surface; however, Gal-3 lacks asignal sequence, and the mechanism of Gal-3 transport to the cellsurface remains poorly understood. Here, using a genome-wideCRISPR/Cas9 forward genetic screen for regulators of Gal-3 cellsurface localization, we identified genes encoding glycoproteins,enzymes involved in N-linked glycosylation, regulators of ER-Golgitrafficking and proteins involved in immunity. The results of thisscreening approach led us to address the controversial role ofN-linked glycosylation in the transport of Gal-3 to the cell surface. Wefind that N-linked glycoprotein maturation is not required for Gal-3transport from the cytosol to the extracellular space, but is importantfor cell surface binding. Additionally, secreted Gal-3 is predominantlyfree and not packaged into extracellular vesicles. These data supporta secretion pathway independent of N-linked glycoproteins andextracellular vesicles.

KEY WORDS: Galectin, Glycosylation, Unconventional secretion

INTRODUCTIONGalectins are an evolutionarily conserved family of β-galactose-binding proteins. There are 15 members, all of which contain acarbohydrate-recognition domain (CRD). The family members canbe divided into three categories: (1) prototypic single CRD galectinsthat can form homodimers; (2) galectins that contain two tandemrepeat CRDs, and (3) galectin-3, a chimeric protein containing asingle CRD and a disordered N-terminal region that facilitatesoligomerization (Elola et al., 2015).Galectins belong to the leaderless class of proteins (defined by

the absence of signal peptide and transmembrane domains) thatfunction both in the cytoplasm and outside the cell. Their functionin the cytoplasm include roles in cell growth, apoptosis, the cellcycle and cellular immunity (Boyle and Randow, 2013; Liu andRabinovich, 2005; Nabi et al., 2015; Rabinovich et al., 2002). Whengalectins are outside the cell, they are known to be retained to theextracellular leaflet of the plasma membrane, typically through their

binding to N-linked glycans and core O-linked glycans onglycosylated proteins and lipids. From there, they modulate manycellular processes, including endocytosis, migration and adhesion(Elola et al., 2015; Lakshminarayan et al., 2014; Mazurek et al.,2012; Xin et al., 2015). In addition, galectins are found in the serum,where they regulate the activity of immune cells (Rabinovich et al.,2002).

Interestingly, galectin-3 (Gal-3, also known as LGALS3) isdetected at high levels in cardiac patients, and used as a marker forcardiovascular disease and heart failure (Jagodzinski et al., 2015;Medvedeva et al., 2016). Similarly, elevated levels of galectin-1(Gal-1, also known as LGALS1) are associated with poor prognosisin many cancers, including melanoma, lung, bladder and headcancers (Thijssen et al., 2015).

The mechanism of galectin secretion remains controversial. Asmentioned above, galectins lack a signal peptide and do not enterthe classical ER/Golgi secretory pathway (Hughes, 1999; Nickel,2003) and their secretion is not blocked by drugs that inhibit theclassical secretory pathway, such as brefeldin A and monensin(Lindstedt et al., 1993; Sato et al., 1993).

Currently, three major mechanisms have been proposed toexplain the unconventional secretion of leaderless proteins:(1) direct translocation across the plasma membrane, eitherthrough a transporter or by auto-transportation as in the case ofFGF2; (2) engulfment into extracellular vesicles (exosomes andmicrovesicles), and (3) capture into a membrane-boundcompartment, such as a secretory autophagosome, late endosomeor CUPS (Hughes, 1999; Nickel and Rabouille, 2009; Nickel andSeedorf, 2008).

Evidence is lacking for a mechanism involving directtranslocation of galectins across the membrane. In particular, atransporter has yet to be identified and auto-transportation by poreformation is also lacking (Hughes, 1999; Nickel and Rabouille,2009; Nickel and Seedorf, 2008; Rabouille, 2017). Galectinsecretion via microvesicles or exosomes, collectively termedextracellular vesicles (EVs), has been proposed (Cooper andBarondes, 1990; Mehul and Hughes, 1997; Sato et al., 1993;Seelenmeyer et al., 2008). Indeed, Gal-3 and Gal-1 are recruited tothe cytoplasmic leaflet of the plasma membrane, where they aresecreted in microvesicles generated by plasma membrane budding(Cooper and Barondes, 1990; Mehul and Hughes, 1997). However,contrasting reports show that Gal-1 secretion is not reduced whenplasma membrane blebbing is inhibited (Seelenmeyer et al., 2008).Furthermore, secretion in EVs does not explain how galectins aresubsequently delivered to the cell surface, although the EVs maybe disrupted in the extracellular space to release Gal-3 (Mehuland Hughes, 1997). It has also been proposed that Gal-1 isdirectly transported across the plasma membrane while coupledto glycoproteins or lipids on the inner leaflet of the membrane.Gal-1 secretion requires a functional CRD for cell surfaceReceived 19 May 2017; Accepted 17 July 2017

1University of Cambridge Metabolic Research Laboratories, Wellcome Trust-Medical Research Council Institute of Metabolic Science, University of Cambridge,Cambridge CB2 0QQ, UK. 2Department of Medicine, Cambridge Institute forMedical Research, Cambridge Biomedical Campus, Hills Road, Cambridge, CB20XY, UK. 3NIHR Cambridge BRC Cell Phenotyping Hub, Department of Medicine,University of Cambridge, Cambridge CB2 0QQ, UK.*These authors contributed equally to this work

‡Author for correspondence (km510@cam.ac.uk)

K.M., 0000-0002-3688-3998

3234

© 2017. Published by The Company of Biologists Ltd | Journal of Cell Science (2017) 130, 3234-3247 doi:10.1242/jcs.206425

Journal

ofCe

llScience

localization, and binding to glycoproteins, proteins orglycolipids may recruit galectins to the inner leaflet of theplasma membrane and mediate transport across the membrane tothe cell surface (Seelenmeyer et al., 2005). However, Chinesehamster ovary (CHO) cells lacking the ability to glycosylateglycoproteins efficiently secrete Gal-1 (Cho and Cummings,1995), suggesting that glycans do not play a role in secretion.Therefore, not only the mechanism of galectin secretion from thecell remains elusive, but also the role of glycosylation in thesecretion process.What is better established, however, is that moieties of N-linked

glycoproteins and lipids that are exposed to the extracellular leafletof the plasma membrane are important for restricting galectins tothe surface of cells after their secretion, and prevent them fromdiffusing in the extracellular space. For instance, exogenouspurified Gal-1, Gal-3 and Gal-8 (also known as LGALS) onlybind to the cell surface when N-linked glycosylation pathways areintact and N-linked glycans expressed at the cell surface (Patnaiket al., 2006).To identify key regulators of Gal-3 cell surface localization (the

sum of both its secretion and its retention) and clarify the role ofglycosylation, we performed a genome-wide CRISPR/Cas9forward genetic screen. The most significantly enriched genesidentified in the screen encodes glycoproteins, enzymes involved inN-linked glycan maturation and proteins regulating ER-Golgitrafficking.We focused on the role of two genes identified in the CRISPR

screen that encode proteins essential for N-linked glycosylation.When N-linked glycosylation was disrupted, the level of Gal-3on the cell surface decreased. This was not caused by disruptionof Gal-3 secretion from the cytosol to the extracellular space, asfree Gal-3 was detected in the medium. This demonstrates thatN-linked glycosylation is not required for secretion of Gal-3, butis essential for cell surface binding. These data support a model inwhich N-linked glycosylation is not required for secretion of Gal-3 to the extracellular space. Furthermore, we tested the role ofEVs in Gal-3 secretion and we conclude that they are notinvolved.

RESULTSA genome-wide screen identifies genes required for Gal-3cell surface localizationOwing to the limited knowledge about Gal-3 trafficking from thecytosol to the cell surface and its regulation, we set out to identifygenes required for cell surface localization of Gal-3. HeLa cellsin steady state suspension (sHeLa) express Gal-3 on their surface(Fig. S1A) and there is a small proportion detectable in the medium(Fig. S1B). To be found on the outer leaflet of the cell, Gal-3 mustbe secreted from the cytosol through an unconventional proteintrafficking pathway. Therefore, we performed a genome-wideCRISPR/Cas9 forward genetic screen in sHeLa and enriched forcells with decreased cell surface Gal-3 (Fig. 1A). To ensure optimalscreening parameters, sHeLa stably expressing Cas9 nuclease(sHeLa-Cas9) were analysed for Gal-3 surface expression by flowcytometry. Gal-3 surface expression was largely homogenous;however, the small population of Gal-3-negative cells were removedin a pre-clear cell sort to optimize screening parameters. Theresulting population (∼1×108 cells) was then transduced with theGeCKO v2 single-guide RNA (sgRNA) library, containing 123,411guide RNAs targeting 19,050 genes, at a multiplicity of infection of∼0.3 (Shalem et al., 2014; Timms et al., 2016). Untransduced cellswere removed through puromycin selection, and rare cells that had

reduced cell surface Gal-3 were enriched by two rounds offluorescence-activated cell sorting (FACS) (Fig. 1A,B). ThesgRNA abundance of the enriched population was quantified bydeep sequencing and compared to the control unsorted population(Fig. 1C) (König et al., 2007; Timms et al., 2016). Strikingly themost significantly enriched genes identified in this screen coded forGolgi enzymes involved in N-linked glycosylation or proteinsregulating ER-Golgi transport (Fig. 1C,D). These include solutecarrier family 35 member A2 (SLC35A2), mannosyl (alpha-1,3-)-glycoprotein beta-1,2-N-acetylglucosaminyltransferase (MGAT1),mannosidase alpha class 1A member 2 (MAN1A2) and componentof oligomeric Golgi complex 1 (COG1) (Fig. 1C,D; Table S1). Tofurther analyse the function of the genes identified in this screen, weapplied bioinformatic pathway analysis to the 200 most enrichedgenes (Huang et al., 2009a,b). This analysis showed that, of thegenes with known function, many genes coded for glycoproteinssuch as integrin β3 (ITGB3), laminin subunit beta 2 (LAMB2)and basigin (CD147, also known as BSG), or proteins involved inthe transport of glycoproteins within the Golgi and to the cellsurface including ADP-ribosylation factor 1 (ARF1) and ADP-ribosylation factor-like protein 3 (ARL3), and proteins with rolesin immunity, including NLR family pyrin domain-containing 2(NLRP2) and tripartite motif-containing proteins 5 and 34(TRIM5, TRIM34) (Table 1). Interestingly, several proteinsidentified in this screen are known Gal-3 interactors either on thecell surface, such as the integrins, laminins and CD147, or in thecytosol for the TRIMs (Chauhan et al., 2016; Priglinger et al.,2013).

No core ER proteins or enzymes required for N-linkedglycosylation upstream of the Golgi were identified in the screen.Furthermore, not all subunits of the COG family were identified.sgRNAs targeting Gal-3 itself were also not enriched in the screen.Analysis of the control unsorted population shows that the screenwas not saturating, and that 7% of the sgRNAs in the library werenot present and ∼20% showed a coverage of <200 cells/sgRNA(data not shown). Five sgRNAs targeting Gal-3 were efficientlyrepresented, yet these cells were not enriched during the screen. Thismay indicate that these guides were not effective at targeting Gal-3,or that Gal-3 deletion is lethal or decreased cell growth. Anotherexplanation for the lack of Gal-3 sgRNA enrichment is that Gal-3secreted by other surrounding cells is able to bind to the surface ofGal-3 null cells, masking the effect of the knockout in our FACSassay. This scenario could also affect other knockout cells, wherethe sgRNA targets key regulators of Gal-3 secretion butglycosylated binding partners on the cell surface are unaffected.However, some of the hits identified in the screen, such as TRIM34,TRIM5, ARHGAP30 andARHGAP9, are not known to regulate theglycosylation pathway. Therefore, it remains unclear why Gal-3 wasnot enriched in this screen.

Additionally, we did not identify genes previously linked tounconventional secretion, such as those coding for the LC3 (alsoknown as MAP1LC3) proteins, GABARAP, GRASP55 (alsoknown as GORASP2) and components of endosomal sortingcomplexes required for transport (ESCRT) (Table S1) (Nickel andRabouille, 2009; Rabouille, 2017). To further confirm thatautophagy, the GRASP55 and the ESCRT pathways do notregulate cell surface localization of Gal-3, we used LC3 andGABARAP knockout cells (Fig. S2A), GRASP55 esiRNA (Fig.S2B) and a Vps4 (also known as VPS4A) dominant negative mutant(Fig. S2C). In all cases the cell surface expression of Gal-3 remainedunaffected, further verifying the absence of these genes in theCRISPR/Cas9 forward genetic screen.

3235

RESEARCH ARTICLE Journal of Cell Science (2017) 130, 3234-3247 doi:10.1242/jcs.206425

Journal

ofCe

llScience

N-linked glycosylation is required for cell surface binding butnot galectin secretionBecause it remains controversial whether glycosylation is requiredfor galectin trafficking to the cell surface, we aimed to determine

whether defective N-linked glycosylation decreases Gal-3trafficking to the cell surface, or N-linked glycoproteins aresimply required for Gal-3 cell surface binding. Tunicamycinblocks the transfer of N-acetylglucosamine-1-phosphate from

Fig. 1. ACRISPR/Cas9-mediated genetic screen identifies genes required for cell surface localization of Gal-3. (A) Schematic of the CRISPR/Cas9 screenin sHeLa to identify genes required for Gal-3 cell surface localization. Cells were transduced with a lentiviral sgRNA library (sgRNA transduction indicated bycolours in the nucleus) and cells that were successfully transduced were selected with puromycin. After selection, the population was split into two; one half wassorted by FACS to enrich for cells that have less Gal-3 on the surface (Gal-3 is represented by small orange shapes on the cell surface) and the other was notsorted to represent the entire library. After two rounds of enrichment, the DNA from both the enriched population and the unsorted library was harvested, andenriched sgRNAs were identified by sequencing. Targeted genes were then plotted according to their relative enrichment. (B) CRISPR-mediated mutagenesiswas performed on sHeLa using the GeCKO v2 sgRNA library, and rare cells with decreased surface Gal-3 expression were selected by two sequential roundsof FACS. Cell surface Gal-3 was measured on live cells using an anti-Gal-3 antibody conjugated to Alexa Fluor 647. (C) Plot illustrating the hits from theCRISPR screen. The RSA algorithm was used to identify the significantly enriched genes targeted in the selected cells. Themost significantly enriched genes arelabelled. (D) Schematic of the N-linked glycosylation pathway within the Golgi. Genes identified to be important for Gal-3 surface localization by the CRISPRscreen are highlighted in red, and those chosen for further study (MGAT1 and SLC35A2) are shown in bold.

3236

RESEARCH ARTICLE Journal of Cell Science (2017) 130, 3234-3247 doi:10.1242/jcs.206425

Journal

ofCe

llScience

UDP-N-acetylglucosamine to dolichol phosphate, the first step inN-linked glycosylation (Fig. 2A) and was used to inhibitglycosylation in sHeLa. Cells were treated with increasingconcentrations of tunicamycin to inhibit N-linked glycosylation,and the level of cell surface Gal-3 was assessed by flow cytometry.Cell surface Gal-3 decreased as the concentration of tunicamycinincreased (Fig. 2B). Propidium iodide was used to measure cellviability and cells remained viable at all tunicamycin concentrations(Fig. 2B, right). In agreement with the results of the CRISPR screen,this shows that a reduction in N-linked glycosylation (and thuscomplex glycans at the cell surface) decreases cell surface Gal-3.To investigate whether N-linked glycosylation is required for

transport of Gal-3 from the cytosol to the extracellular space, thesupernatant of sHeLa cells treated with tunicamycin was analysedby western blotting. In this assay, if N-linked glycosylation is indeedrequired for Gal-3 transport there should be a reduction in the levelof Gal-3 in the supernatant compared to untreated cells. Conversely,if N-linked glycosylation is only required for cell surface bindingand not for secretion, there should be an increase in free Gal-3measured in the supernatant. Western blotting showed aconcentration-dependent increase in Gal-3 in the supernatant aftertunicamycin treatment (Fig. 2C). A similar trend was observed forGal-1 (Fig. 2C, left). The effectiveness of tunicamycin treatmentwas confirmed by assessing the relative expression of the ERresident protein BiP (GRP78, also known as HSPA5), whereincreased expression indicates ER stress (Fig. 2C). Actin andannexin A2 were used as negative controls, with low levelsdetectable in the supernatant upon tunicamycin treatment (Fig. 2C).These results support the suggestion that Gal-3 and Gal-1 requireN-linked glycans to bind to the cell surface (Patnaik et al., 2006).

These data also suggest that secretion of galectins from the cytosolto the extracellular space is independent of N-linked glycosylation.

N-linked glycan maturation mediated by MGAT1 andSLC35A2 is required for Gal-3 cell surface binding but notsecretionThe use of tunicamycin to block N-linked glycosylation providesproof of principle, but there may be confounding factors due to off-target effects. Therefore, to validate the findings of the CRISPRscreen and investigate the role of N-linked glycan maturation, wetargetedMGAT1 and SLC35A2; two genes that were highly enrichedin the screen and are known to be specifically required for N-linkedglycosylation (Fig. 1C,D). In the cis-Golgi, MGAT1 adds N-acetylglucosamine to the sugar backbone of glycoproteins,initiating complex N-linked glycosylation (Fig. 1D). SLC35A2acts later in the trans-Golgi, transporting UDP-galactose into thetrans-Golgi network for addition onto glycoproteins (Fig. 1D).

We generated MGAT1 and SLC35A2 CRISPR knockout celllines using guide RNAs from an independent CRISPR/Cas9 library(Wang et al., 2015). This provides an additional control for off-target effects as the guide RNAs differed from those used in ouroriginal CRISPR screen. Single cell cloning, using FACS, wascarried out to obtain knockout clones for MGAT1 and SLC35A2(Fig. S3). To evaluate the presence of CRISPR-induced mutationsin the MGAT1 or SLC35A2 genes, the targeted region of thegene was amplified and sequenced. Alignments and tracking ofindels by decomposition (TIDE) analysis confirmed that MGAT1and SCL35A2 contain CRISPR-induced insertions and deletions(Fig. S3) (Brinkman et al., 2014). MGAT1 and SLC35A2 clonescontained a combination of homozygous and compoundheterozygous deletions likely to disrupt gene function (Fig. S3).As a positive control, additional MGAT1 and SLC35A2 clones thatexpressed cell surface Gal-3 to a similar level as untargeted cells(Gal-3 positive) were isolated; these contained no insertions ordeletions in the targeted region (Fig. S3).

CRISPR-induced deletions led to a loss of target proteinexpression in both MGAT1 clones and SLC35A2 clones,assessed by western blotting (Fig. 3A). MGAT1 and SLC35A2protein levels in the Gal-3-positive clones are similar to those in thewild type (Fig. 3A). MGAT1 and SLC35A2 are both essential forN-linked glycosylation, so defective glycosylation would beexpected on all N-linked glycoproteins. To assess this, lysosomal-associated membrane protein-2 (LAMP2) glycoforms wereanalysed by western blotting. MGAT1- and SLC35A2-deficientclones expressed a lower molecular weight form of LAMP2compared to wild-type and Gal-3-positive sHeLa (Fig. 3A). Thisindicates that there are fewer mature N-linked glycans added toLAMP2 when MGAT1 or SLC35A2 is absent.

To confirm that the loss of MGAT1 and SLC35A2 leads to adecrease in the expression of Gal-3 on the cell surface, as identifiedin the initial CRISPR screen, we assessed cell surface Gal-3 usingflow cytometry. Gal-3-positive clones expressing MGAT1 andSLC35A2 were indeed positive for Gal-3 at a comparable level towild-type sHeLa (Fig. 3B). Likewise, MGAT1- and SLC35A2-deficient clones obtained from the Gal-3-negative populationshowed a marked reduction in the expression of Gal-3 (Fig. 3B).Therefore, the Gal-3-negative phenotype seen by flow cytometrycan be attributed to CRISPR-mediated knockout of MGAT1 andSLC35A2, further validating the original CRISPR screen.

To assess whether loss of MGAT1 or SLC35A2 impacts thetransport of Gal-3 from the cytosol to the extracellular space, weanalysed Gal-3 secretion by western blotting. Our results show that

Table 1. Pathway analysis of the 200 most significantly enriched genesidentified in the genome-wide CRISPR screen for Gal-3 cell surfacelocalization

Category Count

GlycoproteinsADAM10, ABCC3, CD302, CD33, FRAS1, NDST2, NHLRC3,APLP2, ARTN, CTNND2, CLCA1, CHRM4, C2, DSG3, FAP,FGF17, GABRA1, GIF, HS6ST2, HIST1H2BD, ITGB3, IL17D,LAMB2, LUM, MAN1A2, MPZL3, NPR1, OLFML2A, OR2S2,OR51B6, OR51G2, PHYHIP, PRL, PCDHB6, SLC36A4,SLC39A9, SLC6A7, SUSD3, TOR1a, TMED10

40

Golgi complex/Golgi membraneADAM10, ARF1, ARL3, NDST2, COL4A3BP, COG1, COG3,COG5, MAN1A2, MGAT1, PACS1, SAR1A, SLC35A2, TMED10,TMEM165, UNC50

16

Cell junctionARF1, KIAA1462, CTNND2, CHRM4, DSG3, DLG2, FAP,GABRA1, ITGB3, TOR1A

10

Protein transportARF1, ARL3, COG1, COG3, COG5, NXT2, SAR1A, TMED10,UNC50

9

Cell adhesionCD33, KIAA1462, CTNND2, DSG3, FAP, ITGB3, LAMB2, MPZL3,PCDHB6

9

ImmunityNLRP2, TIRAP, C2, DCSTAMP, LRMP, MAP3K5, TRIM5

7

Protein phosphorylationADAM10, COL4A3BP, DGUOK, MAP3K5, NPR1, OOEP, STK38

7

GTPase activationDEPDC5, ELMOD2, RAP1GDS1, ARHGAP30, ARHGAP9,TBC1D22A

6

Congenital disorders of glycosylationCOG1, COG5, SLC35A2, TMEM165

4

3237

RESEARCH ARTICLE Journal of Cell Science (2017) 130, 3234-3247 doi:10.1242/jcs.206425

Journal

ofCe

llScience

in both MGAT1- and SLC35A2-deficient cells, Gal-3 is readilydetected in the extracellular medium (Fig. 3C). Furthermore, inMGAT1- and SLC35A2-deficient cells there was an increase in therelative amount of Gal-3 in the supernatant compared to the wild-

type sHeLa control (Fig. 3C). This was not seen in the Gal-3-positive wild-type clones, which remained similar to wild-typesHeLa (Fig. 3C). Gal-1 also showed a similar trend in SLC35A2knockout cells (Fig. S4). Therefore, a lack of N-linked glycosylation

Fig. 2. Tunicamycin decreases cell surface Gal-3 while increasing the level of Gal-3 in the medium. (A) Schematic of tunicamycin inhibition of N-linkedglycosylation. Tunicamycin blocks the transfer of N-acetylglucosamine-1-phosphate from UPD-N-acetylglucosamine to dolichol phosphate, the first step in N-linked glycosylation. (B) Tunicamycin reduces cell surface localization of Gal-3. sHeLa were treated with increasing concentrations of tunicamycin diluted inserum-free medium for 24 h. Cell surface Gal-3 was measured on live cells using an anti-Gal-3 antibody conjugated to Alexa Fluor 647; cell viability was alsoassessed by flow cytometry using propidium iodide. Unstained cells are shown in grey. Quantification is shown on the right. Data are mean±s.e.m. from biologicalreplicates (n=3); *P<0.05 (two-sample Student’s t-test comparing cells with treated with the indicated tunicamycin concentrations to untreated cells). (C)Tunicamycin increases the levels of Gal-3 in the culture supernatant. Western blot analysis of cell lysates and supernatants of sHeLa treated with increasingconcentrations of tunicamycin (24 h at 37°C in serum-free medium). Note that the tunicamycin treatment was efficient, as shown by the increased level of BiP anddecreased level of CD29. Quantification is shown on the right. Data are mean±s.e.m. from biological replicates (n=3); *P<0.05 (two-sample Student’s t-testcomparing cells treated with the indicated tunicamycin concentrations to untreated cells).

3238

RESEARCH ARTICLE Journal of Cell Science (2017) 130, 3234-3247 doi:10.1242/jcs.206425

Journal

ofCe

llScience

Fig. 3. MGAT1 and SLC35A2 knockout abrogates Gal-3 cell surface binding but not secretion. (A) Western blot analysis of MGAT1- and SLC35A2-deficientsHeLa. Cell lysates were assessed for either MGAT1 or SLC35A2 protein levels after CRISPR/Cas9 targeting and single cell cloning based on Gal-3 surfaceexpression. LAMP2wasalso assessed to analyse defects in glycosylation, and actinwasusedas a loading control. (B) Cell surface localization ofGal-3 is decreasedin MGAT1- and SLC35A2-deficient sHeLa measured by flow cytometry. Cell surface Gal-3 was measured on live cells using an anti-Gal-3 antibody conjugated toAlexaFluor 647.Grey, noantibody; black line, untransfected; pink dotted line, sgMGAT1-positive clone; blue, sgMGAT1-negative clone 1; green, sgMGAT1-negativeclone 2. The same respective colours are used for sgSLC35A2 in the lower panels. (C) Gal-3 is secreted from MGAT1- and SLC35A2-deficient sHeLa. Wild type,positive control and negative clones forMGAT1 (left) andSLC35A2 (right) cellswere incubated in serum-freemedium for 24 h, and the cells andmediumassessedbywestern blotting. Gal-3 was assessed in the lysate and medium (supernatant); actin was used as a loading control and control for cell lysis. Exposure times areindicated to allow relative comparisons between blots to illustrate the large increase inGal-3 in the supernatant compared to actin. Quantification of MGAT1 (left) andSLC35A2 (right) is shown in the bottompanels. Data aremean±s.e.m. frombiological replicates (n=3); *P<0.05 (two-sampleStudent’s t-test comparing each cell linetowild-type cells). Note that the same actin loading control blots are shown in the top panels of A andC, and also in the top panel of Fig. 5A, as the blots were strippedand reprobed with the indicated antibodies. Similarly, the bottom panel of A and the top panel of Fig. 5B are the same blots.

3239

RESEARCH ARTICLE Journal of Cell Science (2017) 130, 3234-3247 doi:10.1242/jcs.206425

Journal

ofCe

llScience

due to MGAT1 or SLC35A2 deficiency leads to reduced galectinbinding to the cell surface and an increase in galectin in thesupernatant. This is indicative of a binding defect and not areduction in secretion.

CHOglycosylationmutants also efficiently secretegalectinsTo further assess the role of N-linked glycosylation in the transportof galectins to the cell surface and their secretion, several well-characterized CHO cell lines with glycosylation defects were usedto validate our data (Stanley, 1989). These include an MGAT1 loss-of-function mutant (Lec1), an SLC35A2 loss-of-function mutant(Lec8), and an MGAT1 and SLC35A2 loss-of-function doublemutant (Lec3.2.8.1) (Stanley, 1989). The aberrations in the N-linked glycans produced from each cell line are depicted in Fig. 4A.These cell lines were previously used to analyse galectin-glycan

binding specificity, demonstrating that N-linked glycans are themajor ligand for Gal-1, -3 and -8 binding at the cell surface (Patnaikand Stanley, 2006). These mutant CHO lines were used here tofurther assess Gal-3 and Gal-1 cell surface binding and secretion.Flow cytometry analysis of Gal-3 expression on the surface of CHOcells showed that MGAT1 (Lec1) and SLC35A2 (Lec8) single loss-of-function mutant lines, as well as the MGAT1/SLC35A2(Lec3.2.8.1) double-mutant line, all exhibited a decrease in thelevel of Gal-3 detectable on the cell surface compared to the wildtype (Pro5) (Fig. 4A). This phenotype was reversed in an MGAT1rescue cell line, confirming that the loss of Gal-3 on the surface iscaused by the loss-of-function mutation in the MGAT1 gene(Fig. 4B) (Chen and Stanley, 2003; Kumar and Stanley, 1989).Western blot analysis showed that Gal-3 was secreted by MGAT1(Lec1) and SLC35A2 (Lec8) loss-of-function cells as expected(Fig. 4C). Furthermore, the level of Gal-3 detectable in the mediumwas substantially higher than in the wild-type (Pro5) CHO andMGAT1 rescue cell lines (Fig. 4C). This was also evident when Gal-1 secretion was assessed (Fig. 4C). These data are consistent withresults obtained in sHeLa lines and further confirm that N-linkedglycosylation is not required for Gal-3 secretion.

Secreted Gal-3 is primarily free and not packaged into EVsThus far, we have shown that N-linked glycan maturation is notrequired for the transport of Gal-3 from the cytosol to theextracellular space and is a regulatory element that retainsgalectins at the cell surface. Next, we set out to investigatewhether secreted Gal-3 is free in the medium or packaged into EVs.There are conflicting data in the literature as to whether galectins aresecreted via EVs (Cooper and Barondes, 1990; Mehul and Hughes,1997; Sato et al., 1993; Seelenmeyer et al., 2008). To investigatethis, the medium from wild-type, MGAT1- or SCL35A2-deficientcells was collected and subjected to differential centrifugation.Briefly, cells were removed at 300 g, then the cell debris wasremoved at 3000 g and EVs pelleted at 100,000 g. The supernatantand EV pellets were separated after centrifugation at 100,000 g andeach assessed for Gal-3 by western blotting. The data show similarlevels of Gal-3 in the medium after removing EVs at 100,000 g,indicating that the majority of the secreted Gal-3 is free and notpackaged in vesicles (Fig. 5A,B). Gal-3 was detectable in the100,000 g EV pellet of all cell lines, although the levels weresomewhat variable, and there was a small increase in the amount ofboth actin and Gal-3 detected in the EV pellets from MGAT1-deficient clones (Fig. 5A,B). It is important to note that the EVpellets were 50× concentrated compared to the supernatant samples(Fig. 5A,B). To further assess the composition of the 100,000 gpellet, we analysed the tetraspanin CD63, which is known to be

enriched in exosomes (Escola et al., 1998). The 100,000 g pelletwas CD63 positive and therefore contained some exosomes(Fig. 5A,B). Owing to impaired glycosylation, CD63 runs as asmaller form in theMGAT1- and SLC35A2-deficient EVs (Fig. 5A,B).The lack of glycosylation on CD63 seems to affect antibodydetection, and the naked nonglycosylated form was detected betterthan the glycosylated form. Therefore, it is difficult to comment onthe relative levels of CD63 in the EV pellets of the MGAT1- andSLC35A2-deficient cells compared to the wild-type controls.However, we believe that the lack of MGAT1 or SLC35A2 doesnot affect the formation or level of EVs.

We also assessed whether Gal-3 secreted from CHO MGAT1(Lec1), SLC35A2 (Lec8) and MGAT1/SLC35A2 double(Lec3.2.8.1) mutant cell lines is also free and not packaged intoEVs. In agreement with the MGAT1- and SLC35A2-deficientsHeLa, the levels of Gal-3 secreted from the CHO MGAT1 (Lec1),SLC35A2 (Lec8) and MGAT1/SLC35A2 double (Lec3.2.8.1)mutant lines remained unchanged after a 100,000 g centrifugationstep (Fig. 5C). There was a small increase in the level of Gal-3 andactin detectable in the EV pellets of MGAT1 (Lec1), SLC35A2(Lec8) and MAGT1/SLC35A2 double (Lec3.2.8.1) mutantscompared to wild-type (Pro5) and rescue lines (Fig. 5C). Thismay also be reflected in the MGAT1-deficient cells but is not thecase for SLC35A2-deficient cells, in which the level was morevariable (Fig. 5A,B). Therefore, any difference in the level of EVssecreted is trivial and is unlikely to significantly contribute to thelevels of secreted Gal-3. Owing to differences in the species of thecells, we were unable to evaluate CD63 in the EV pellets of CHOcells. Taken together, these results show that Gal-3 associated withEVs is a small proportion of the total secreted Gal-3 and, therefore,cannot be the primary route for trafficking outside the cell.

N-linked glycoproteins are required for the recruitment ofintracellular Gal-3 to damaged lysosomal membranesGal-3 has important roles in regulating cell death and immunity, andis recruited to endolysosomes, lysosomes and phagosomes inresponse to induced organelle damage and damage caused bybacterial infection (Aits et al., 2015; Feeley et al., 2017; Maejimaet al., 2013; Paz et al., 2010). In addition, Gal-3 interacts withTRIM16 to coordinate autophagy to protect against cell damage andbacterial invasion (Chauhan et al., 2016). Recruitment to lysosomesor Shigella-disrupted phagosomes is dependent on Gal-3 binding toN-linked glycans, as shown using a Gal-3 CRD mutant and CHOMGAT1 mutant (Lec1) cells, respectively (Aits et al., 2015; Pazet al., 2010). Therefore, N-linked glycans are not only important forcell surface localization of Gal-3, but are also central for therecruitment of Gal-3 to damaged lysosomes. To further characterizeour MGAT1- and SLC35A2-deficient sHeLa lines, we assessed theability of Gal-3 to redistribute from the cytosol to the membrane ofleaky lysosomes (Maejima et al., 2013). To do so, we expressedGFP fused to Gal-3 in wild-type, MGAT1- and SLC35A2-deficientsHeLa lines. All cell lines were then treated with L-Leucyl-L-Leucine methyl ester (LLOMe) to induce lysosomal leakiness, andwe assessed the recruitment of GFP-Gal-3 to the site of damage byimmunofluorescence (Maejima et al., 2013). In wild-type cells,GFP-Gal-3 was efficiently redistributed from the cytosol to the siteof lysosomal damage, colocalizing with LAMP2-positive puncta(Fig. 6A). However, in MGAT1- and SLC35A2-deficient cells therecruitment of GFP-Gal-3 to LAMP2 positive damaged lysosomeswas reduced (Fig. 6A). We also assessed recruitment of LC3, asdamage to lysosomes should initiate autophagy to degrade thedysfunctional organelle (Maejima et al., 2013). As expected, in

3240

RESEARCH ARTICLE Journal of Cell Science (2017) 130, 3234-3247 doi:10.1242/jcs.206425

Journal

ofCe

llScience

Fig. 4. MGAT1 andSLC35A2mutation in CHOLec cells reducesGal-3 cell surface binding but does not affect secretion. (A) Gal-3 cell surface localization isdecreased in MGAT1 and SLC35A2 mutant CHO lines. Cell surface Gal-3 was measured on live MGAT1 (Lec1), SLC35A2 (Lec8) and double mutant (Lec3.2.1.8)CHO Lec cells compared towild-type (Pro5) cells by flow cytometry using an anti-Gal-3 antibody conjugated to Alexa Fluor 647. Grey, no antibody; brown, wild type;red, mutants. Predicted N-linked glycans for each cell line are shown on the histograms. Sugar symbols: purple triangle, fructose; green circle, mannose; orangecircle, galactose; blue square, N-acetylglucosamine; pink trapezoid, sialic acid. Quantification is shown on the right. Data are mean±s.e.m. from biological replicates(n=3); *P<0.05 [two-sampleStudent’s t-test comparing eachmutant CHO line towild-type (Pro5) cells]. (B)Gal-3 cell surface localization is rescued inMGAT1 rescueCHOLec cellsmeasuredby flow cytometry.Gal-3wasmeasuredas inA.Grey, noantibody; brown,wild type; red,mutants; purple, rescue.Quantification is shownonthe right. Data are mean±s.e.m. from biological replicates (n=3); *P<0.05 [two-sample Student’s t-test comparing the MGAT1 mutant and rescue lines to wild type(Pro5) cells]. (C)Gal-3 is secreted fromMGAT1andSLC35A2mutant CHOLec cells.Wild-type (Pro5),MGAT1 (Lec1),MGAT1 rescue, SLC35A2mutant (Lec8) andthe double mutant (Lec3.2.8.1) were incubated in EX-CELL 325 PF CHO for 48 h, and the cells and medium assessed by western blotting. Gal-3, Gal-1 and actinwere analysed in the cell lysatesandmedium (supernatant). Exposure timesare indicated forcomparison.Quantification is shownon the right. Data aremean±s.e.m.from biological replicates (n=3); *P<0.05 [two-sample Student’s t-test comparing each mutant CHO cell line to wild-type (Pro5) cells].

3241

RESEARCH ARTICLE Journal of Cell Science (2017) 130, 3234-3247 doi:10.1242/jcs.206425

Journal

ofCe

llScience

Fig. 5. Secreted Gal-3 is predominantly soluble and not packaged in EVs. (A) Soluble Gal-3 is secreted fromMGAT1-deficient sHeLa.Wild type, positive controland negative clones forMGAT1-deficient cells were incubated in serum-freemedium for 24 h. The cellswere collected and lysed, whereas themediumwas subjectedto differential centrifugation at 300, 3000 and 100,000 g. A sample of the medium was collected after each centrifugation step. Gal-3 was assessed in the lysate, theentire 100,000 gEV pellet andmedium (supernatant). Actin was used as a loading control and control for cell lysis. Exposure times are indicated for comparison. The100,000 g EV pellets were also analysed by western blotting for levels of glycosylated and nonglycosylated CD63. (B) Soluble Gal-3 is secreted from SLC35A2-deficient sHeLacells.Wild type, positive control andnegative clones for SLC35A2-deficient cellswere incubated in serum-freemedium for 24 h. Sampleswere treatedasdescribed inA. (C)SecretedGal-3 fromMGAT1andSLC35A2mutantCHOLeccells is soluble.Wild type (Pro5),MGAT1 (Lec1),MGAT1 rescue, SLC35A2mutant(Lec8) and the doublemutant (Lec3.2.8.1) were incubated in EX-CELL 325 PFCHO for 48 h, and the cells andmedium collected. The cells, 100,000 gEVpellet andmedium were processed as in A. Gal-3 and actin were analysed by western blotting and exposure times are indicated for comparison. CD63 was not analysed in thisexperiment as it does not cross-react with hamster CD63.Note that the sameactin loading control blots are shown in the top panel of B and the bottompanel of Fig. 3Aas the blots were stripped and reprobed with the indicated antibodies. Also, the top panel of A and the top panels of Fig. 3A and C, are the same blots.

3242

RESEARCH ARTICLE Journal of Cell Science (2017) 130, 3234-3247 doi:10.1242/jcs.206425

Journal

ofCe

llScience

wild-type cells treated with LLOMe, GFP-Gal-3-positive punctawere also mRFP-LC3 positive (Fig. 6B). In the MGAT1- andSLC35A2-deficient cells, recruitment of GFP-Gal-3 to mRFP-LC3-positive damaged lysosomes was impaired (Fig. 6B). This furtherconfirms that N-linked glycan maturation is required for therecruitment of Gal-3 to damaged lysosomes and autophagosomes,essential for cellular homeostasis and defence.

DISCUSSIONCell surface expression of galectins is essential for cellularhomeostasis. Despite having important functions in theextracellular space, the mechanism of galectin secretion remainsunclear. Galectins do not enter the classical secretory pathway, asthey do not contain a signal peptide and their secretion is not affectedby drugs that block this pathway (Hughes, 1999). Therefore, theymust exit the cell though an unknown unconventional proteintrafficking pathway. Currently, there are limited data available toexplain the mechanisms of galectin trafficking from the cytosol to the

extracellular space, and current theories are controversial. Here, weapplied a genome-wide CRISPR screen using the GeCKO v2 libraryto identify regulators of Gal-3 cell surface localization. Followingmutagenesis and enriching for cells with reduced Gal-3 expression atthe cell surface, many genes coding for glycoproteins or proteinsrequired for N-linked glycan maturation were identified. While thisscreen returned many important regulators of Gal-3m it is apparentthat the screen was not saturating, as shown by the sequencing datafrom the control unsorted population. However, five sgRNAstargeting Gal-3 were efficiently represented in the control unsortedpopulation, yet these cells were not enriched during sorting. Oneexplanation for this is that there is free Gal-3 in the medium, secretedby surrounding cells, that binds to the surface of Gal-3-deficient cellsmasking their Gal-3-negative phenotype. This could also mask otherimportant hits where secretion of Gal-3 is impaired but glycosylationis normal. Unfortunately, wewere not able to assess the levels of Gal-3 in the medium and, therefore, do not know if this explains the lackof Gal-3 sgRNA enrichment. Moreover, there are hits identified in

Fig. 6. Recruitment of GFP-Gal3 to damaged lysosomes is reduced in MGAT1- and SLC35A2-deficient cells. (A) Wild-type (control), MGAT1-deficient(clone 1) or SLC35A2-deficient (clone 1) sHeLa transiently expressing GFP-Gal-3 for 24 h were treated with 1 mM LLOMe for 3 h. Cells were fixed with PFA,permeabilized with Triton X100 and subjected to immunocytochemistry using an anti-LAMP2 antibody, then processed for confocal microscopy. The intensity ofLAMP2 and Gal-3 signals measured using ImageJ in a minimum of 20 cells per condition is shown on the right. (B) Wild-type (control), MGAT1-deficient (cl1) orSLC35A2-deficient (cl1) sHeLa transiently expressing GFP-Gal-3 and mRFP-LC3 for 24 h were treated with 1 mM LLOMe for 3 h. Cells were fixed with methanoland processed for confocal microscopy. Colocalization (Pearson’s coefficient) between Gal-3 and LC3 is shown on the right. Data are mean±s.e.m. fromindividual cells (n>20); *P<0.05 (two-sample Student’s t-test). Scale bars: 10 μm.

3243

RESEARCH ARTICLE Journal of Cell Science (2017) 130, 3234-3247 doi:10.1242/jcs.206425

Journal

ofCe

llScience

this screen that are not known to regulate glycosylation (such asTRIM34, TRIM5, ARHGAP30 and ARHGAP9), which should alsobe masked in this scenario. Additionally, we did not detect any ERproteins required for glycosylation, upstream of the Golgi, which issomewhat surprising. Loss of these proteinsmay be lethal or decreasecell proliferation. It is also important to note that the mostsignificantly enriched genes identified by the screen may not bethose most important for mediating Gal-3 surface localization; it maysimply be that they survive well and are therefore enriched better thanothers.Although the screen was not saturating, the results obtained here

are consistent with the literature as Gal-3 is known to bind to N-linked glycans present on the cell surface (Patnaik et al., 2006). Thisis also consistent with the notion that Gal-3 requires N-linkedglycans to facilitate trafficking from the cytosol to the cell surface(Seelenmeyer et al., 2005). However, this is controversial, and it wasimportant to establish whether glycoproteins carrying N-linkedsugar moieties are required for transport of Gal-3 from the cytosol tothe extracellular space. Using tunicamycin and two differentMGAT1 and SLC35A2 mutant cell lines, we demonstrate thatGal-3 cell surface binding is dependent on the expression ofcomplex N-linked glycans; however, Gal-3 is efficiently secreted inthe absence of N-linked glycans. The secretion of both Gal-3 andGal-1 was unperturbed in the absence of N-linked glycosylation,clearly demonstrating that their secretion is independent of both theclassical secretory pathway and any pathway requiring complexglycoproteins and lipids for transport.The role of EVs in galectin secretion has been controversial, with

conflicting reports in the literature (Cooper and Barondes, 1990;Mehul and Hughes, 1997; Sato et al., 1993; Seelenmeyer et al.,2008). Here, we demonstrate that transport of Gal-3 from the cytosolto the extracellular space is not primarily mediated by EVs in sHeLaand CHO cell lines. Owing to the increased levels of Gal-3detectable in the medium, MGAT1- and SLC35A2-deficient cellsprovide an excellent system for assessing whether extracellular Gal-3 is packaged into EVs. Using differential centrifugation, we showthat the vast majority of Gal-3 detected in the medium is free andsoluble, indicating that Gal-3 is not packaged into EVs. These datasupport an EV-independent pathway for Gal-3 trafficking to the cellsurface and secretion into the extracellular space.It has previously been shown that Gal-3 is redistributed from the

cytosol to glycoproteins on the luminal membrane of damagedendolysosomes/phagosomes (Aits et al., 2015; Feeley et al., 2017;Maejima et al., 2013; Paz et al., 2010). Once associated with themembrane of the damaged organelle, Gal-3 stimulates autophagy toclear the threat (Maejima et al., 2013). Furthermore, it has beenshown that Gal-3 is recruited to Shigella-containing phagosomes inwild-type CHO cells but not in MGAT1 (Lec1) mutant CHO cells(Paz et al., 2010). Given these previous data, we tested the MGAT1-and SLC35A2-deficient sHeLa in this context. As expected, Gal-3recruitment to damaged lysosomes is impaired in the MGAT1- andSLC35A2-deficient cell lines. These data, shown by us and others,may explain why people with congenital disorders of glycosylation(CDG) suffer from recurrent infections (Albahri, 2015; Grünewaldet al., 2002; Monticelli et al., 2016). CDG are rare genetic disorderswhere glycosylation of multiple proteins is deficient or defectivedue to mutations in the glycosylation pathway; these mutations canoccur in COG1, MGAT1 and SLC35A2 genes among many others(Albahri, 2015; Grünewald et al., 2002). CDG cause a range oforgan malfunctions; in almost all cases the nervous system isaffected and symptoms include developmental disabilities, ataxiahypotonia, hyporeflexia and immunological defects (Albahri, 2015;

Grünewald et al., 2002; Monticelli et al., 2016). It is becomingincreasingly apparent that patients with immunological defects aremore likely to have mutations in resident ER and Golgi enzymes(Monticelli et al., 2016). Consistent with this, our data and previousdata from Paz and colleagues suggest that patients with certainforms of CDG could have a reduced ability to sense bacterial or viralentry in the cytosol due to a lack of galectin recruitment to the site ofinfection (Paz et al., 2010).

Together, these data demonstrate that galectin cell surface bindingand secretion are two distinct events. This is consistent with previousstudies, which have shown that Gal-3 secretion is unaffected bydisruptions in the secretory pathway (Cho and Cummings, 1995;Lindstedt et al., 1993; Sato et al., 1993). Exactly which domains orsequences are essential for mediating galectin secretion are alsocontroversial. It has been shown that the flexible N-terminal domainon Gal-3 is important for secretion; however, this flexible N-terminaldomain is absent in other galectins (Menon and Hughes, 1999).Therefore, if there is a common unconventional secretory pathwayutilized by the galectin family, it would be somewhat surprising if thiswas located in the only domain that is not conserved across thegalectin family. By contrast, other studies have found that the CRD isessential for the effective secretion of Gal-1 (Seelenmeyer et al.,2005). However, in our analyses, the CRD mutant Gal-3 (R186S),which is unable to bind GlcNAc, did not show any defects in Gal-3secretion compared to the wild type (Fig. S5) (Salomonsson et al.,2010). It is possible that there are differences in the requirements forsecretion between galectin family members, but this would be verysurprising as the galectins are highly similar and common transportmechanism would be expected.

Regardless of the exact mechanism, it may be expected thatgalectins are not secreted via the conventional secretory pathway astheir ligand (complex carbohydrates) is a major component of thelumen of the ER and Golgi. If galectins had to move through the ERand Golgi they would come into contact with their ligand, bind, andpotentially interrupt the movement of other proteins through thesecretory pathway. Therefore, having a separate pathway fortrafficking galectins to the cell surface is an excellent way ofensuring that they only meet their ligands where required.

Finally, hundreds of genetic disorders that result fromdeficiencies in different glycosylation pathways have beendescribed, including several neurological diseases such as autism,epilepsy and CDG (Freeze et al., 2015). Additionally, cancer cellsare known to deeply alter the glycosylation pathway inducing hypo-or hyper-glycosylation (Pinho and Reis, 2015). As such, it would beinteresting to study whether the alterations in several signallingpathways described in these diseases are associated with adysregulation of cell surface galectins, given the important role ofgalectins in signal transduction and cell-to-cell interactions.

MATERIALS AND METHODSCell cultureSuspension HeLa cells were cultured in DMEM D6546 (Molecular Probes)plus 10% fetal bovine serum (FBS), 2 mM L-glutamine and 100 Uml−1

penicillin/streptomycin in 5% CO2 at 37°C. LC3 and GABARAP knockoutHeLa cells were cultured as described (Nguyen et al., 2016). Lec cells(CHO), obtained from Pamela Stanley (Albert Einstein College ofMedicine, New York City, USA), were cultured in MEM alpha,nucleosides (Molecular Probes, 22571038), plus 10% FBS and100 Uml−1 penicillin/streptomycin in 5% CO2 at 37°C.

Antibodies and reagentsAntibodies used were rat polyclonal anti-Gal-3 [Biolegend; 125408; 1/2000in western blotting (WB)], rat polyclonal anti-galectin-3 conjugated to

3244

RESEARCH ARTICLE Journal of Cell Science (2017) 130, 3234-3247 doi:10.1242/jcs.206425

Journal

ofCe

llScience

Alexa Fluor 647 (Biolegend; 125402; 1/100 in flow cytometry), rabbitpolyclonal anti-galectin-1 (a generous gift from Walter Nickel, HeidelbergUniversity, Germany; 1/500 in WB), mouse monoclonal anti-annexin A2(BD Biosciences; 610071; 1/1000 in WB), rabbit polyclonal anti-actin(Sigma-Aldrich; A2066; 1/2000 in WB), rabbit polyclonal anti-BiP(Abcam; ab21685: 1/1000 in WB), mouse monoclonal anti-LAMP2(Biolegend; 354302; 1/1000 in WB, 1/100 in immunofluorescence),rabbit polyclonal anti-SLC35A2 (Cambridge Bioscience; HPA036087;1/500 in WB), rabbit polyclonal anti-MGAT1 (Abcam; ab180578; 1/1000in WB), mouse monoclonal anti-human CD63 (Thermo Fisher Scientific;10628D; 1:500 in WB), rabbit polyclonal anti-GFP (Clontech; 632592;1/2000 in WB), rabbit polyclonal anti-LC3B (Novus Biologicals; NB100-2220; 1/2000 in WB), rabbit polyclonal anti-GABARAP (Abgent;AP1821a; 1/1000 in WB), monoclonal anti-CD29 (BD Biosciences;Clone 18/CD29; 1/2000 in WB) and rabbit polyclonal anti-GRASP55(Proteintech; 10598-1-AP; 1/1000 in WB).

Reagents included tunicamycin (New England Biolabs; 12819), L-Leucyl-L-leucine methyl ester (Sigma-Aldrich; L7393), propidium iodidesolution (Biolegend; 421301), QuickExtract DNA extraction solution(Epicenter; QE0905T), Herculase II fusion DNA polymerase (Agilent;600675). Oligonucleotides for MGAT1 and SLC35A2 CRISPR targetingand sequencing were obtained from Sigma-Aldrich (Table S2). MISSIONesiRNA against GRASP55 was also from Sigma-Aldrich (EHU056901).

PlasmidspSpCas9(BB)-2A-GFP (PX458) [Addgene plasmid #48138, deposited byFeng Zhang (Ran et al., 2013)], pEGFP-hGal3 [Addgene plasmid #73080,deposited by Tamotsu Yoshimori (Maejima et al., 2013)], pmRFP-LC3[Addgene plasmid #21075, deposited by Tamotsu Yoshimori (Maejimaet al., 2013)] and lentiCas9-Blast [Addgene plasmid #52962, deposited byFeng Zhang (Sanjana et al., 2014)] were used. Vps4 wild type and EQmutant were a gift from Colin Crump (Crump et al., 2007).

CRISPR screenThe Cas9 nuclease was stably expressed in suspension HeLa cells bylentiviral transduction (Sanjana et al., 2014). Then, ∼1×108 cells weretransduced with the GeCKO v2 sgRNA library [Addgene cat #1000000047,deposited by Feng Zhang (Shalem et al., 2014)] at a multiplicity of infectionof ∼0.2. Untransduced cells were removed from the library throughpuromycin selection (1 mg/ml) commencing 48 h after transduction. Rarecells that had lost cell surface Gal-3 were then enriched by sequential roundsof FACS, with the first sort taking place 7 days after transduction with thesgRNA library and the second sort 14 days later. Genomic DNA wasextracted (Puregene Core Kit A, Qiagen) from both the sorted cells and anunselected pool of mutagenized cells. sgRNA sequences were amplified bytwo rounds of PCR, with the second round primers containing the necessaryadaptors for Illumina sequencing (Table S2). Sequencing was carried outusing a 50 bp single-end read on an Illumina HiSeq2500 instrument, using acustom primer binding immediately upstream of the 20 bp variable segmentof the sgRNA. The 3′ ends of the resulting reads were trimmed off theconstant portion of the sgRNA, and then mapped to an index of all of thesgRNA sequences in the GeCKO v2 library using Bowtie 2. The resultingsgRNA count tables were then analysed using the RSA algorithm with thedefault settings (Rivest et al., 1978).

Bioinformatics pathway analysisThe first 200 hits identified in the CRISPR screen were loaded to theanalysis wizard of DAVID Bioinformatics Resources 6.8 to perform apathway analysis (Huang et al., 2009a,b). According to the algorithm, onlythose genes with known function are included in the pathway analysis andhence not all genes will appear in the tabulated results (Table 1).

CRISPR-mediated gene disruptionFor CRISPR/Cas9-mediated gene disruption, oligonucleotides (Sigma-Aldrich) (Table S2) for top and bottom strands of the sgRNAwere annealed,and then cloned into the Cas9 expression vector pSpCas9(BB)-2A-GFP(PX458) as previously described (Ran et al., 2013). Transfected cells weresorted for GFP fluorescence and clones were isolated by FACS based on a

loss of cell surface Gal-3. Gene disruption was verified by collectinggenomic DNA from clonal lines with QuickExtract DNA extraction solutionand amplifying the CRISPR/Cas9 targeted region with primers flanking≥200 base pairs either side of the expected cut site (Table S2). PCR productswere sequenced by Sanger sequencing. Insertions and deletions analysed bysequence alignment and TIDE (Brinkman et al., 2014). The TIDE R code,kindly provided by Prof. van Steensel (The Netherlands Cancer Institute,Amsterdam, The Netherlands), was used to analyse clones containingdeletions >50 bp.

Flow cytometryCells were washed once with serum-free medium, incubated at 4°C for30 min with an anti-Gal-3 antibody conjugated to Alexa Fluor 647, washedagain and analysed on a FACSCalibur (BD Biosciences) equipped withlasers providing 488 nm and 633 nm excitation sources. Alexa Fluor 647fluorescence was detected with an FL4 detector (661/16 BP). For sorting,cells were immunostained as above and FACS was carried on an Influx (BDBiosciences) or Aria-Fusions (BD Biosciences) cell sorter equipped withlasers providing 488 nm and 640 nm excitation sources. Alexa Fluor 647fluorescence was detected with a 670/30 BP detector on the Influx and AriaFusion cell sorters.

Immunofluorescence microscopyFor immunofluorescence microscopy, cells were cultured on coverslips,fixed with 4% paraformaldehyde in PBS for 5 min and permeabilized with0.1% Triton X100 in PBS for 5 min. Coverslips were incubated withprimary antibodies for 2 h, washed three times with PBS, and incubatedwith secondary antibodies for 30 min. Samples were mounted usingProLong Gold antifade reagent with DAPI (4,6-diamidino-2-phenylindole;Invitrogen) and observed using a Leica SP8 laser confocal microscope.

ImmunoblottingAll samples were resolved by 12% SDS-PAGE and transferred topolyvinylidene difluoride membranes for blotting. Membranes wereblocked with 5% (w/v) skim milk powder in PBS containing 0.1% Tween20 (PBST) for 30 min at room temperature.Membraneswere then probedwithan appropriate dilutionof primaryantibodyovernight at 4°C.Membraneswerewashed three times in PBST before incubation in diluted secondary antibodyfor 1 h at room temperature.Membraneswerewashed as before and developedwith Amersham ECLWestern Blotting Detection Reagent RPN2106 for thedetection of proteins in the cell lysates, or Cyanagen Westar XLS100 for thedetection of proteins in the secreted fractions, using a Bio-Rad Chemi DocXRS system. Membranes were stripped with Restore Plus (Thermo FisherScientific, 46430) according to the manufacturer’s instructions.

Secretion assayTo measure the secretion of galectins, cells were washed with serum-freemedium and incubated for 24 h for sHeLa or 48 h for CHO (Lec). For sHeLa,the serum-free medium was DMEM plus 2 mM L-Glutamine. For CHO(Lec), the serum-free medium was EX-CELL 325 PF CHO (Sigma-Aldrich,C985Z18). Cell supernatants were then collected, centrifuged at 300 g toremove potential remaining cells, either filtered at 0.22 µm or centrifuged at3000 g to remove cell debris, mixed with sample buffer [50 mMTris-HCl pH6.8, 2% SDS (w/v), 0.1% Bromophenol Blue, 10% glycerol and 100 mMDTT] and boiled at 100°C for 5 min. Cell pellets were lysed in lysis buffer(20 mM Tris-HCl pH 6.8, 137 mM NaCl, 1 mM EDTA, 1% Triton X-100and 10% glycerol) at 4°C for 10 min, and insoluble material was removed bycentrifugation at 10,000 g for 10 min at 4°C. Sample bufferwas added and thesamples boiled (as above). Cell lysates and cell supernatants were thensubjected to SDS-PAGE. Densitometry was performed in ImageJ and thedifferences in the levels of secreted Gal-3 were calculated in each cell line bydividing the amount of Gal-3 in the supernatant by the amount of Gal-3 in thelysate. These values were then used to calculate the fold change relative to thecontrol cells.

Removal of extracellular vesiclesCells were processed as described for the secretion assay, except after the3000 g centrifugation step the medium was collected and centrifuged at

3245

RESEARCH ARTICLE Journal of Cell Science (2017) 130, 3234-3247 doi:10.1242/jcs.206425

Journal

ofCe

llScience

100,000 g for 60 min at 4°C. After each centrifugation step a sample of themediumwas collected for western blotting. The extracellular vesicle pellet wasresuspended in a small volume of nonreducing sample buffer [50 mM Tris-HCl pH 6.8, 2% SDS (w/v), 0.1% Bromophenol Blue, and 10% glycerol].Half of the EV pellet samplewas taken andDTT added to a final concentrationof 100 mM. All samples were boiled and resolved by SDS-PAGE. The entireconcentrated EV pellet sample was loaded on two gels (reduced andnonreduced) for each sample due to the small scale of the assay. Therefore, theEV pellet was 50× more concentrated than the equivalent supernatant.

Statistical analysisSignificance levels for comparisons between groups were determined withtwo-sample Student’s t-test.

AcknowledgementsWe thank Walter Nickel for Gal-1 antibody, Michael D’Angelo for help andexpertise with CRISPR/Cas9 INDEL analysis, Bas van Steensel for providing theTIDE R code, Matt Castle for R code expertise, Nuno Rocha for help with theCRISPR cloning method, Pamela Stanley for Lec CHO cell lines, MichaelLazarou for the LC3 and GABARAP knockout cells, and Colin Crump for the Vps4vectors.

Competing interestsThe authors declare no competing or financial interests.

Author contributionsConceptualization: S.E.S., S.A.M., P.J.L., K.M.; Methodology: S.E.S., S.A.M., S.J.P., N.S., A.P.H., P.J.L., K.M.; Validation: S.E.S., S.J.P., P.J.L., K.M.; Formalanalysis: S.E.S., S.A.M., S.J.P.; Investigation: S.E.S., S.A.M., K.M.; Data curation:K.M.; Writing - original draft: S.E.S., K.M.; Writing - review & editing: S.E.S., S.A.M.,S.J.P., N.S., A.P.H., P.J.L., K.M.; Supervision: K.M.; Project administration: K.M.;Funding acquisition: A.P.H., P.J.L., K.M.

FundingThis work was supported by the Wellcome Trust [100574/Z/12/Z; 101835/Z/13/Z toP.J.L.; Wellcome Trust PhD studentship to S.A.M. and S.J.P.], Medical ResearchCouncil [MRC_MC_UU_12012/5 to S.E.S. and K.M.] and National Institute forHealth Research [to N.S. and A.P.H.]. Deposited in PMC for release after 6 months.

Supplementary informationSupplementary information available online athttp://jcs.biologists.org/lookup/doi/10.1242/jcs.206425.supplemental

ReferencesAits, S., Kricker, J., Liu, B., Ellegaard, A.-M., Hamalisto, S., Tvingsholm, S.,Corcelle-Termeau, E., Høgh, S., Farkas, T., Holm Jonassen, A. et al. (2015).Sensitive detection of lysosomal membrane permeabilization by lysosomalgalectin puncta assay. Autophagy 11, 1408-1424.

Albahri, Z. (2015). Congenital disorders of glycosylation: a review.Am. J. Pediatr. 1,6-28.

Boyle, K. B. and Randow, F. (2013). The role of ‘eat-me’ signals andautophagy cargo receptors in innate immunity. Curr. Opin. Microbiol. 16,339-348.

Brinkman, E. K., Chen, T., Amendola, M. and van Steensel, B. (2014). Easyquantitative assessment of genome editing by sequence trace decomposition.Nucleic Acids Res. 42, e168.

Chauhan, S., Kumar, S., Jain, A., Ponpuak, M., Mudd, M. H., Kimura, T., Choi,S. W., Peters, R., Mandell, M., Bruun, J.-A. et al. (2016). TRIMs and galectinsglobally cooperate and TRIM16 and galectin-3 co-direct autophagy inendomembrane damage homeostasis. Dev. Cell 39, 13-27.

Chen, W. and Stanley, P. (2003). Five Lec1 CHO cell mutants have distinct Mgat1gene mutations that encode truncated N-acetylglucosaminyltransferase I.Glycobiology 13, 43-50.

Cho, M. and Cummings, R. D. (1995). Galectin-1, a beta-galactoside-binding lectinin Chinese hamster ovary cells. II. Localization and biosynthesis. J. Biol. Chem.270, 5207-5212.

Cooper, D. N. and Barondes, S. H. (1990). Evidence for export of a muscle lectinfrom cytosol to extracellular matrix and for a novel secretory mechanism. J. CellBiol. 110, 1681-1691.

Crump, C. M., Yates, C. and Minson, T. (2007). Herpes simplex virus type 1cytoplasmic envelopment requires functional Vps4. J. Virol. 81, 7380-7387.

Elola, M. T., Blidner, A. G., Ferragut, F., Bracalente, C. and Rabinovich, G. A.(2015). Assembly, organization and regulation of cell-surface receptors by lectin-glycan complexes. Biochem. J. 469, 1-16.

Escola, J.-M., Kleijmeer, M. J., Stoorvogel, W., Griffith, J. M., Yoshie, O. andGeuze, H. J. (1998). Selective enrichment of tetraspan proteins on the internalvesicles of multivesicular endosomes and on exosomes secreted by human B-lymphocytes. J. Biol. Chem. 273, 20121-20127.

Feeley, E. M., Pilla-Moffett, D. M., Zwack, E. E., Piro, A. S., Finethy, R., Kolb,J. P., Martinez, J., Brodsky, I. E. and Coers, J. (2017). Galectin-3 directsantimicrobial guanylate binding proteins to vacuoles furnished with bacterialsecretion systems. Proc. Natl. Acad. Sci. USA 114, E1698-E1706.

Freeze, H. H., Eklund, E. A., Ng, B. G. and Patterson, M. C. (2015).Neurological aspects of human glycosylation disorders. Annu. Rev.Neurosci. 38, 105-125.

Grunewald, S., Matthijs, G. and Jaeken, J. (2002). Congenital disorders ofglycosylation: a review. Pediatr. Res. 52, 618-624.

Huang, W., Sherman, B. T. and Lempicki, R. A. (2009a). Bioinformaticsenrichment tools: paths toward the comprehensive functional analysis of largegene lists. Nucleic Acids Res. 37, 1-13.

Huang, W., Sherman, B. T. and Lempicki, R. A. (2009b). Systematic andintegrative analysis of large gene lists using DAVID bioinformatics resources. Nat.Protoc. 4, 44-57.

Hughes, R. C. (1999). Secretion of the galectin family of mammalian carbohydrate-binding proteins. Biochim. Biophys. Acta 1473, 172-185.

Jagodzinski, A., Havulinna, A. S., Appelbaum, S., Zeller, T., Jousilahti, P.,Skytte-Johanssen, S., Hughes, M. F., Blankenberg, S. and Salomaa, V.(2015). Predictive value of galectin-3 for incident cardiovascular disease andheart failure in the population-based FINRISK 1997 cohort. Int. J. Cardiol. 192,33-39.

Konig, R., Chiang, C.-Y., Tu, B. P., Yan, S. F., DeJesus, P. D., Romero, A.,Bergauer, T., Orth, A., Krueger, U., Zhou, Y. et al. (2007). A probability-basedapproach for the analysis of large-scale RNAi screens. Nat. Methods 4, 847-849.

Kumar, R. and Stanley, P. (1989). Transfection of a human gene that corrects theLec1 glycosylation defect: evidence for transfer of the structural gene for N-acetylglucosaminyltransferase I. Mol. Cell. Biol. 9, 5713-5717.

Lakshminarayan, R., Wunder, C., Becken, U., Howes, M. T., Benzing, C.,Arumugam, S., Sales, S., Ariotti, N., Chambon, V., Lamaze, C. et al. (2014).Galectin-3 drives glycosphingolipid-dependent biogenesis of clathrin-independent carriers. Nat. Cell Biol. 16, 595-606.

Lindstedt, R., Apodaca, G., Barondes, S. H., Mostov, K. E. and Leffler, H.(1993). Apical secretion of a cytosolic protein by Madin-Darby canine kidneycells. Evidence for polarized release of an endogenous lectin by a nonclassicalsecretory pathway. J. Biol. Chem. 268, 11750-11757.

Liu, F.-T. and Rabinovich, G. A. (2005). Galectins as modulators of tumourprogression. Nat. Rev. Cancer 5, 29-41.

Maejima, I., Takahashi, A., Omori, H., Kimura, T., Takabatake, Y., Saitoh, T.,Yamamoto, A., Hamasaki, M., Noda, T., Isaka, Y. et al. (2013). Autophagysequesters damaged lysosomes to control lysosomal biogenesis and kidneyinjury. EMBO J. 32, 2336-2347.

Mazurek, N., Byrd, J. C., Sun, Y., Hafley, M., Ramirez, K., Burks, J. andBresalier, R. S. (2012). Cell-surface galectin-3 confers resistance to TRAIL byimpeding trafficking of death receptors in metastatic colon adenocarcinoma cells.Cell Death Differ. 19, 523-533.

Medvedeva, E. A., Berezin, I. I., Surkova, E. A., Yaranov, D. M. and Shchukin,Y. V. (2016). Galectin-3 in patients with chronic heart failure: association withoxidative stress, inflammation, renal dysfunction and prognosis. MinervaCardioangiol. 64, 595-602.

Mehul, B. and Hughes, R. C. (1997). Plasma membrane targetting, vesicularbudding and release of galectin 3 from the cytoplasm of mammalian cells duringsecretion. J. Cell Sci. 110, 1169-1178.

Menon, R. P. and Hughes, R. C. (1999). Determinants in the N-terminal domainsof galectin-3 for secretion by a novel pathway circumventing the endoplasmicreticulum-Golgi complex. Eur. J. Biochem. 264, 569-576.

Monticelli, M., Ferro, T., Jaeken, J., Dos Reis Ferreira, V. and Videira, P. A.(2016). Immunological aspects of congenital disorders of glycosylation (CDG): areview. J. Inherit. Metab. Dis. 39, 765-780.

Nabi, I. R., Shankar, J. and Dennis, J. W. (2015). The galectin lattice at a glance.J. Cell Sci. 128, 2213-2219.

Nguyen, T. N., Padman, B. S., Usher, J., Oorschot, V., Ramm, G. and Lazarou,M. (2016). Atg8 family LC3/GABARAP proteins are crucial for autophagosome-lysosome fusion but not autophagosome formation during PINK1/Parkinmitophagy and starvation. J. Cell Biol. 215, 857-874.

Nickel, W. (2003). The mystery of nonclassical protein secretion. A current viewon cargo proteins and potential export routes. Eur. J. Biochem. 270,2109-2119.

Nickel, W. and Rabouille, C. (2009). Mechanisms of regulated unconventionalprotein secretion. Nat. Rev. Mol. Cell Biol. 10, 148-155.

Nickel, W. and Seedorf, M. (2008). Unconventional mechanisms of proteintransport to the cell surface of eukaryotic cells. Annu. Rev. Cell Dev. Biol. 24,287-308.

Patnaik, S. K. and Stanley, P. (2006). Lectin-resistant CHO glycosylation mutants.Methods Enzymol. 416, 159-182.

3246

RESEARCH ARTICLE Journal of Cell Science (2017) 130, 3234-3247 doi:10.1242/jcs.206425

Journal

ofCe

llScience

Patnaik, S. K., Potvin, B., Carlsson, S., Sturm, D., Leffler, H. and Stanley, P.(2006). Complex N-glycans are the major ligands for galectin-1, -3, and -8 onChinese hamster ovary cells. Glycobiology 16, 305-317.

Paz, I., Sachse, M., Dupont, N., Mounier, J., Cederfur, C., Enninga, J.,Leffler, H., Poirier, F., Prevost, M.-C., Lafont, F. et al. (2010). Galectin-3,a marker for vacuole lysis by invasive pathogens. Cell. Microbiol. 12,530-544.

Pinho, S. S. and Reis, C. A. (2015). Glycosylation in cancer: mechanisms andclinical implications. Nat. Rev. Cancer 15, 540-555.

Priglinger, C. S., Szober, C. M., Priglinger, S. G., Merl, J., Euler, K. N., Kernt, M.,Gondi, G., Behler, J., Geerlof, A., Kampik, A. et al. (2013). Galectin-3 inducesclustering of CD147 and integrin-beta1 transmembrane glycoprotein receptors onthe RPE cell surface. PLoS ONE 8, e70011.

Rabinovich, G. A., Rubinstein, N. and Fainboim, L. (2002). Unlocking the secretsof galectins: a challenge at the frontier of glyco-immunology. J. Leukoc. Biol. 71,741-752.

Rabouille, C. (2017). Pathways of unconventional protein secretion. Trends CellBiol. 27, 230-240.

Ran, F. A., Hsu, P. D., Wright, J., Agarwala, V., Scott, D. A. and Zhang, F.(2013). Genome engineering using the CRISPR-Cas9 system. Nat. Protoc. 8,2281-2308.

Rivest, R. L., Shamir, A. and Adleman, L. (1978). A method for obtaining digitalsignatures and public-key cryptosystems. Commun. ACM 21, 120-126.

Salomonsson, E., Carlsson, M. C., Osla, V., Hendus-Altenburger, R., Kahl-Knutson, B., Oberg, C. T., Sundin, A., Nilsson, R., Nordberg-Karlsson, E.,Nilsson, U. J. et al. (2010). Mutational tuning of galectin-3 specificity andbiological function. J. Biol. Chem. 285, 35079-35091.

Sanjana, N. E., Shalem, O. and Zhang, F. (2014). Improved vectors and genome-wide libraries for CRISPR screening. Nat. Methods 11, 783-784.

Sato, S., Burdett, I. and Hughes, R. C. (1993). Secretion of the baby hamsterkidney 30-kDa galactose-binding lectin from polarized and nonpolarized cells: a

pathway independent of the endoplasmic reticulum-Golgi complex.Exp. Cell Res.207, 8-18.

Seelenmeyer, C., Wegehingel, S., Tews, I., Kunzler, M., Aebi, M. and Nickel,W. (2005). Cell surface counter receptors are essential componentsof the unconventional export machinery of galectin-1. J. Cell Biol. 171,373-381.

Seelenmeyer, C., Stegmayer, C. and Nickel, W. (2008). Unconventional secretionof fibroblast growth factor 2 and galectin-1 does not require shedding of plasmamembrane-derived vesicles. FEBS Lett. 582, 1362-1368.

Shalem, O., Sanjana, N. E., Hartenian, E., Shi, X., Scott, D. A., Mikkelsen,T. S., Heckl, D., Ebert, B. L., Root, D. E., Doench, J. G. et al. (2014).Genome-scale CRISPR-Cas9 knockout screening in human cells. Science343, 84-87.

Stanley, P. (1989). Chinese hamster ovary cell mutants with multiple glycosylationdefects for production of glycoproteins with minimal carbohydrate heterogeneity.Mol. Cell. Biol. 9, 377-383.

Thijssen, V. L., Heusschen, R., Caers, J. and Griffioen, A. W. (2015). Galectinexpression in cancer diagnosis and prognosis: A systematic review. Biochim.Biophys. Acta 1855, 235-247.

Timms, R. T., Menzies, S. A., Tchasovnikarova, I. A., Christensen, L. C.,Williamson, J. C., Antrobus, R., Dougan, G., Ellgaard, L. and Lehner, P. J.(2016). Genetic dissection of mammalian ERAD through comparative haploid andCRISPR forward genetic screens. Nat. Commun. 7, 11786.

Wang, T., Birsoy, K., Hughes, N.W., Krupczak, K. M., Post, Y., Wei, J. J., Lander,E. S. and Sabatini, D. M. (2015). Identification and characterization of essentialgenes in the human genome. Science 350, 1096-1101.

Xin, M., Dong, X.-W. and Guo, X.-L. (2015). Role of the interaction betweengalectin-3 and cell adhesion molecules in cancer metastasis. Biomed.Pharmacother. 69, 179-185.

3247

RESEARCH ARTICLE Journal of Cell Science (2017) 130, 3234-3247 doi:10.1242/jcs.206425

Journal

ofCe

llScience