Characterization of the O-Glycosylation Sites in the ChorionicGonadotropin b Subunit in Vivo Using Site-directed Mutagenesisand Gene Transfer*

(Received for publication, April 9, 1996, and in revised form, May 31, 1996)

Tadashi Sugahara, Mary R. Pixley, Fuad Fares‡, and Irving Boime§

From the Department of Molecular Biology and Pharmacology, Washington University School of Medicine,St. Louis, Missouri 63110

Human chorionic gonadotropin (CG) is a member ofa family of glycoprotein hormones which are het-erodimers containing two nonidentical subunits: a com-mon a and a hormone-specific b subunit. One of thedistinguishing features of the CGb subunit is the pres-ence of four serine acceptors clustered within the last 25amino acids. We previously demonstrated that this car-boxyl-terminal region is important for maintaining itsbiologic half-life, and when the sequence was geneti-cally fused to either the common a or follitropin b sub-units, O-glycosylation was observed. Because this car-boxyl-terminal sequence is located at the end of thesubunit, we considered this region a convenient in vivomodel for studying O-linked glycosylation in domainscontainingmultiple serine recognition sites. A CGb genewas engineered in which the N-linked sites were inacti-vated to eliminate background from those carbohydrategroups. Using this construct, we made a series of trun-cation and amino acid substitutions of acceptor serines,and these mutants were transfected into Chinese ham-ster ovary cells. O-Glycosylation was determined by[3H]glucosamine incorporation and glycanase sensitiv-ity of the products on SDS-polyacrylamide gels. We showthat the O-linked sites comprise independent repetitiveregions in which each acceptor serine has a recognitionsignal bounded by the next carboxy acceptor serinewithin four to five amino acids. It is also apparent thatrecognition of one site is not dependent on the glycosy-lation of another acceptor. Amino acid mutations in theacceptor regions demonstrated the importance of pro-line as a necessary feature for O-linked recognition inthe CGb sequence.

The placental hormone, human chorionic gonadotropin(hCG),1 and the pituitary hormone lutropin are members of theglycoprotein hormone family that also include follitropin andthyrotropin. They are heterodimers formed by a noncovalentassociation of a common a subunit and a hormone-specific bsubunit which determines receptor specificity (1). A unique

feature of the hCGb subunit is the carboxyl-terminal peptide(CTP) which bears four serine-linked oligosaccharides (2–4).One potential function of the CTP is to prolong plasma half-lifeof hCG (5–8) since its deletion from the hCGb subunit bysite-directed mutagenesis reduces the in vivo potency 3-foldcompared to wild type hCG (6). Previously, we demonstratedthat when the CTP is fused to either the carboxyl terminus ofthe follicle stimulating hormone b subunit (7) or to the commona subunit (8), it was O-glycosylated. This suggested that all ofthe information needed for O-glycosylation is contained withinthe CTP sequence.TheO-linked oligosaccharides clustered in the CGb carboxyl-

terminal region are composed of sialylated galactose/N-acetyl-galactosamine structures (4, 9). Attachment of N-acetylgalac-tosamine initiates the process and, although there iscontroversy regarding the precise point in the secretory path-way where this occurs (10; see references therein), recent ex-periments using immunoelectron microscopy identified UDP-GalNAc:polypeptide N-acetylgalactosaminyltransferase in thecis-Golgi compartment (10). The addition of galactose and sialicacid presumably occurs in the trans-Golgi. While the minimumsequence requirement for N-glycosylation of asparagine resi-dues is Asn-Xaa-(Ser/Thr) (where Xaa is any amino acid exceptproline), it is not known how the Ser/Thr sites are selected forO-glycosylation. Several investigators have examined thisprocess using in vitro systems comprised of synthetic peptidescontaining the glycosylation sites and purified UDP-GalNAc:polypeptide N-acetylgalactosaminyltransferase to define thefunctional acceptor unit. These studies showed that both typeand position of amino acids near the glycosylated serine orthreonine are important for glycosylation (11–19). Althoughthere is no clear consensus sequence analogous to the N-glyco-sylation sites, there are some preferences in the flanking aminoacids such as proline, serine, threonine, and alanine.Few studies have been performed in vivo (cell-culture) to

examine O-linked recognition (8, 20, 21). Such investigationsare important since there are multiple steps (and compart-ments) involved in the maturation of O-linked oligosaccha-rides. Based on our previous data, we considered the CGbsubunit a convenient model for studying O-glycosylation incells because the domain which contains the O-linked carbohy-drate is located at the carboxyl terminus, and because thepresence or absence of this region has no dramatic effect on theoverall folding of the subunit (6, 7). Thus, mutations within theCTP sequence should create minimal perturbations in the pro-tein. Here we address whether or not glycosylation of the fouracceptor serines in the CTP is governed by a single determi-nant or if multiple repetitive recognition units are involved.The amino acids flanking the acceptor sites were also examinedin an attempt to characterize the recognition sequence. Forthese studies, a construct was engineered in which the N-

* This work was supported by National Institutes of Health ContractN01-HD92922 and a grant from the Monsanto Company. The costs ofpublication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked “advertisement”in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.‡ Current address: Dept. of Biochemistry and Biochemical Research

Unit, Kupat Holim-Lady Carmel Hospital, 7 Michal St., Haifa 34362,Israel.§ To whom correspondence should be addressed: Box 8103, Washing-

ton University School of Medicine, 660 South Euclid Ave., St. Louis, MO63110. Tel.: 314-362-2556; Fax: 314-362-2704.

1 The abbreviations used are: hCG, human chorionic gonadotropin;CTP, carboxyl-terminal peptide; PCR, polymerase chain reaction;HPLC, high performance liquid chromatography.

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 271, No. 34, Issue of August 23, pp. 20797–20804, 1996© 1996 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

20797

by guest on February 18, 2018http://w

ww

.jbc.org/D

ownloaded from

linked sites were inactivated to eliminate any background fromthose carbohydrate groups. This vector was used as a substratefor mutagenesis, and the resulting mutants were transfectedinto Chinese hamster ovary cells. Here we show that the fourO-linked sites are independent repeat units and that O-glyco-sylation of one site is not dependent on the recognition ofanother O-glycosylated acceptor.

EXPERIMENTAL PROCEDURES

Restriction enzymes used to prepare vectors were purchased fromPromega or Boehringer Mannheim. The enzymes used for PCR wereobtained from Stratagene (Pfu DNA polymerase). The DNA vector,M13mp19 was obtained from New England Biolabs. Oligonucleotideswere prepared by the Washington University Protein and Nucleic AcidChemistry Laboratory (St. Louis, MO). [35S]Cysteine (.1,000 Ci/mmol)was purchased from ICN Biochemicals (Irvine, CA). Cell culture mediaand reagents were prepared by the Washington University Center forBasic Cancer Research. Dialyzed calf serum, G418, fetal calf serum, andimmunoprecipitin were obtained from Life Technologies, Inc. Polyclonalantiserum directed against hCGb subunit was prepared in ourlaboratory.Mutagenesis and Vector Construction—Site-directed mutagenesis by

overlap PCR extension was described by Ho et al. (22). SalI-SalI frag-ments containing exon III of CGb gene (540 base pairs) were insertedinto M13mp19, and the RF DNA was isolated to create a convenientcassette DNA for PCR mutagenesis. Two complementary oligonucleo-tides (27-mers) were synthesized for each mutagenesis. Together withthese oligonucleotides, M13 reverse and forward universal primerswere used for the PCR. A series of truncated or amino acid substitutedmutants in the CTP region were constructed (Table I).These mutants were shuttled into the eukaryotic expression vector,

pM2HA (23, 24) containing the CGb exons I and II, resulting in a vectorcontaining the entire CGb gene. To perform the analyses without in-terference from the N-linked carbohydrates on O-linked carbohydrateanalysis, we used a mutant CGb gene (DAsn (1 1 2)) which lacksAsn-X-(Thr/Ser) consensus sequences at the Asn13 and Asn30 acceptorsites (24). The entire exon III cassette of all the mutants was sequencedto ensure that the mutations were correct and that there were nomisincorporations during the PCR.Transfection and Cell Culture—These constructs were transfected

into CHO cells as described previously (25), and stable transfectantswere selected by growing in culture medium containing 0.25 mg/mlneomycin analogue G418. Expression of the mutants was detected by

immunoprecipitation of metabolically labeled cells (see below). All sta-bly transfected CHO cell lines were maintained in medium A (Ham’sF12 medium supplemented with penicillin (100 units/ml), streptomycin(100 mg/ml), and glutamine (2 mM)) containing 5% (v/v) FCS and 0.125mg/ml G418 in a humidified 5% CO2 incubator.Metabolic Labeling and Protein Analysis—Cells were plated into

12-well dishes (300,000–350,000 cells/well) in 1 ml of medium A sup-plemented with 5% FCS 1 day before labeling. For continuous labelingexperiments, cells were washed twice with medium B (cysteine-freemedium A supplemented with 5% dialyzed calf serum) and labeled for6 h in 1 ml of this medium B containing 20 mCi/ml [35S]cysteine. Mediaand cell lysates were prepared and treated as described (26). All immu-noprecipitates were resolved on 15% SDS-polyacrylamide gels by themethod of Laemmli (27). Gels were soaked for 10 min in 1 M sodiumsalicylate, dried, and autoradiographed. Cells were also metabolicallylabeled with [3H]glucosamine, and labeled GalNAc was determined byHPLC on acid hydrolyzed protein using an Aminex Bio-Rad columnaccording to Green et al. (28).

RESULTS

Truncated CTPMutants—The hCGb subunit has two N- andfour O-linked oligosaccharides, and two forms of the subunitare expressed in CHO cells; this is apparently caused by thepresence of one or two N-linked oligosaccharides (24, 29). Com-pared to the lysate form, one of the distinguishing features ofthe secreted protein is a mobility shift of about 10 kDa which isprimarily due to the addition of O-linked chains immediatelyprior to secretion. This difference in electrophoretic mobility isthe basis for the assay of O-linked acceptor efficiency of themutants. Because the heterogeneity created by the N-linkedcarbohydrates could complicate the analysis (see below), weconstructed a mutant (CGbDAsn (1 1 2)), which lacks theN-linked consensus sequences (24). We first made a series oftruncated mutants containing stop codons at various sites inthe CTP to: (a) calibrate the system for relating protein migra-tion and the number of sites O-glycosylated (b) assess if theserine acceptor sites are independent from one another (seeTable I). Three variants containing a stop codon at eitherresidue 122 (T22), 128 (T28), and 133 (T33) were constructed(Table I, Fig. 1). If a variant is not O-glycosylated, the electro-phoretic mobility of lysate and media should be the same. The

TABLE ITruncated or amino acid substituted mutants in CTP region

Biosynthesis of O-Linked Oligosaccharides20798

by guest on February 18, 2018http://w

ww

.jbc.org/D

ownloaded from

Asn mutant containing the complete CTP unit undergoes thelysate/medium change as discussed above (Fig. 1, lanes 3 and4). There is a progressive decrease in mobility as the amount ofdeleted sequence is increased; however, it is evident that theloss of a significant portion of the CTP sequence as in the T33mutant does not eliminate all of the acceptor activity in theremaining serine sites (lanes 5 and 6). The secreted form of theT22 mutant exhibits no apparent molecular weight change(lanes 9 and 10) even though it contains the normal O-glycosy-lation site at Ser121, whereas the T28 mutant which containsan additional six amino acids including the Ser127 acceptorexhibits a lysate/medium modification (lanes 7 and 8); thus, atleast one site appears O-glycosylated in T28 (see below). Themutant T33 has three potential serine O-glycosylation sitesand manifests a molecular weight change greater than thatseen in mutant T28. The increase in the apparent mass corre-sponds to an additional O-linked carbohydrate unit. Theseresults suggest that each acceptor site in the CTP comprises anindependent recognition unit.To address whether absence of the N-linked carbohydrate

affected recognition, the corresponding CTP mutants contain-ing the two intact Asn-acceptor sites expressed in CHO cellswere treated with endoglycosidase F (Fig. 1B). Because thisenzyme specifically releases the N-linked oligosaccharides, notO-linked carbohydrates, the digested proteins should be equiv-alent to the (DAsn (1 1 2)) mutants. In the absence of Endo F,both the lysate and media proteins of the wild type and the CTPmutants display multiple forms. The lysate/medium profiles ofthe treated samples with Endo F are identical to that seen withDAsn (1 1 2) variants (Fig. 1A). As expected, the digestionproduct of the intra- and extracellular forms of the T22 mutantwere identical, indicating no modification occurred (data notshown). These data imply that the lack of the N-linked oligo-saccharides do not affect acceptor recognition in these mutants.Characterization of the CTP Modification—The above data

show that the lysate/media modifications are confined betweenthe Ser121 and carboxyl terminus 145. To show that thechanges in the mobility are due to O-linked glycosylation, sam-ples of media from the truncated mutants (T22, T28, and T33)were treated with O-glycanase (Genzyme) (Fig. 2). Becausesialic acid substitution on Gal-GalNAc blocks glycanase hydrol-ysis, samples of T33 and T28 were first treated with neuramin-idase which reduces the size of the truncated and wild typeforms (panel B, lanes 3, 7, and 11). Subsequent addition ofO-glycanase further decreases the mass to the correspondinglysate form (lanes 4, 8, and 12). As expected, the enzymes hadno effect on the migration of the unmodified T22 mutant (panelB, lanes 13–16).Further evidence for the presence of O-linked carbohydrate

in these truncated mutants is the incorporation of the carbo-hydrate precursor [3H]glucosamine (Fig. 3). Cells were labeledovernight in media containing 50 mCi/ml of [3H]glucosamine.Because the N-linked sites are mutated, incorporated label willreflect only the presence of O-linked carbohydrate. The se-creted forms of DAsn, T33, T28, but not T22, incorporate 3H(lanes 3–6), and they co-migrate with the corresponding 35S-labeled proteins (lane 2). Moreover, when the O-linked oligo-saccharides were acid-hydrolyzed from the DAsn (1 1 2) mu-tant and analyzed by HPLC (28), two major peakscorresponding to GalNAc and GluNAc in a 15:1 ratio wereobserved. These data are in agreement with the predominanceof GalNAc, not GlcNAc in the native CGb O-linked carbohy-drates. Taken together, the data show that the changes inmobility of the secreted mutants are due to the presence ofO-linked carbohydrate.Point Mutations in CTP and Characterizations of a Single

Acceptor Site—While data with the T28 mutant appeared todefine a single serine recognition site, we could not exclude thatthe three tandem serine residues at 118–120 and the terminalSer127 were also alternative acceptors. To address this issue, a

FIG. 1. Expression of truncated CGb mutants from transfected CHO cells. CHO cells were incubated for 6 h in the presence of 25 mCi/ml[35S]cysteine (A). Lysate (L) and media (M) fractions were immunoprecipitated with CGb-specific antiserum and were resolved on 15% polyacryl-amide gels. Autoradiography was overnight. bWT, CGb wild type (lanes 1 and 2); DAsn, the CGb mutant without the two N-linked sites atasparagines 13 and 30 (lanes 3 and 4). The mutants T22 (lanes 9 and 10), T28 (lanes 7 and 8), and T33 (lanes 5 and 6) terminate with amino acids121, 127, and 132, respectively (see also Table I). B shows the effect of removing the N-linked oligosaccharides from the truncated mutants byendoglycosidase F digestion (Endo F). The T28 and T33 variants containing the two N-linked glycosylation sites were treated with Endo F asdescribed (22).

Biosynthesis of O-Linked Oligosaccharides 20799

by guest on February 18, 2018http://w

ww

.jbc.org/D

ownloaded from

modified T28 mutant was constructed in which serines 118–120 were converted to alanine (S3 1 T28) (Fig. 4A, lanes 7 and8). Despite this change, the lysate/medium shift characteristicof T28 (lanes 11 and 12) is still observed. Further, when avariant was constructed in which only the acceptor serine 121was mutated to alanine (DSer1 1 T28) (Fig. 4B, lanes 5 and 6),

the acceptor activity was abolished. These data show that inthe T28 mutant only Ser121 is the acceptor and that the termi-nal Ser127 is not glycosylated. It is also evident that the entirecomplement of serine acceptor sites are glycosylated in the S3variant of CGb DAsn(1 1 2), i.e. containing the complete CTPsequence (Fig. 4A; lanes 5 and 6). Thus, serine residues 118–

FIG. 2. Sensitivity of the CGb trun-cated mutants to neuraminidase andO-glycanase. The CGb truncated mu-tants T33, T28, and T22 (A) and DAsn (11 2) secreted by transfected CHO cellswere immunoprecipitated and treatedwith either neuraminidase alone for 18 hor with neuraminidase for 1 h and a sub-sequent incubation with O-glycanase for17 h (B).

FIG. 3. [3H]Glucosamine incorpora-tion into CGb mutants. CHO cells ex-pressing DAsn, T33, T28, and T22 wereincubated overnight with 50 mCi/ml[3H]glucosamine (lanes 3–6). Secretedprotein was immunoprecipitated and re-solved on 15% polyacrylamide gel (lanes3–6). [35S]Cysteine-labeled DAsn (lanes 1and 2) and T22 (lanes 7 and 8) were alsoincluded. The gel was exposed overnightat 270 °C.

Biosynthesis of O-Linked Oligosaccharides20800

by guest on February 18, 2018http://w

ww

.jbc.org/D

ownloaded from

120 on the amino side of Ser121 are not required for recognitionat Ser121 or the other acceptor sites. Moreover, a DSer1 mutantdoes not affect glycosylation of the other serines (data notshown) which implies O-linked acceptor activity of the down-stream sites is not dependent on proximal recognition of Ser121.It appears then that Ser120 defines the amino-terminal borderof the CTP.Having identified at least one recognition site in the T28

mutant, we created a series of stop and amino acid substitutionmutants between the T22 and T28 mutants to further definethe structural requirement for the acceptor activity of Ser121

(Fig. 5A). A mutant truncated at amino acid 123 (T24) was notglycosylated (Fig. 5B, lanes 11 and 12); the presence of theadditional Lys-Ala was not sufficient for recognition. However,if a variant included the adjacent Pro124 (T25), some glycosy-lation was detected, although it was incomplete as evident bythe presence of the unglycosylated form (lanes 9 and 10). Mu-tants which result in extending the protein from Pro125 toSer127 become more efficiently glycosylated; the intensity of theupper band increases at the expense of the lower band (lanes 4,6, 8, and 10). Adding the next serine acceptor site, i.e. Ser127

results in quantitative O-glycosylation of Ser121. Even thoughthe T28 mutant has the Ser127 glycosylation site, the apparentmolecular weight of this mutant is comparable to that observedfor the truncated mutants T25–T27. Thus, while Ser127 in theT28 mutant is not glycosylated, it is required for maximalacceptor activity of Ser121, which implies that Ser127 definesthe carboxy border of the Ser121 recognition sequence.Although acceptor activity of Ser121 is dependent on the

intervening sequence between 121 and 127, it is not clear if thenature of the amino acid, i.e. proline and/or the length of thesequence is critical. To address this question, the Pro in thepartial acceptor mutant T25 was changed to several differentamino acids (Fig. 5C). If length is the sole determinant, thenacceptor activity should remain unaffected by the amino acidchanges, whereas if the proline at 124 is crucial, changing this

residue should affect O-glycosylation. No detectable acceptoractivity was seen when the Pro was mutated to Arg (lanes 3 and4) or Asp (not shown). The glycosylation efficiency of variantscontaining either Leu, Ala, or Gln was 30–50% of that seen inT25, and extending the sequence from 122 to 127 with only Alareplacements did not increase the extent of glycosylation overthat seen with T25 Ala mutant (Fig. 5D). These data show thaton the carboxy side of the Ser121 acceptor site, the type of aminoacid is an essential feature for maximal glycosylation efficiencyand that increasing the peptide length is insufficient to over-come the lack of the Pro residues.We examined the sequence requirements for the second ac-

ceptor site at Ser127. The migration of T33 stop mutant corre-sponds to a CTP derivative containing two O-linked sites. Thisindicates that the sequence Ser127 to Ser132 encompasses anadditional recognition unit (Fig. 6). If the T33 mutant is trun-cated at Ser130 (T31), two bands are detected (lane 6); themobility shift of the lower band is the same as that of the T28stop mutant, and that of the upper band is comparable to theunmodified T33 mutant. No nonglycosylated form of either ofthese mutants is observed. Thus, Ser127 in the T31 mutant is aweak acceptor in the absence of the Pro131 and Ser132 residues.Moreover, Ser132 of the T33 mutant is not an acceptor becausethe mobility of the secreted T33 mutant is the same as that ofthe upper band of the T31 mutant devoid of Ser132. Althoughwe cannot exclude the possibility that Ser130 in the T31 mutantis an alternative acceptor, based on our data, its carboxyl-terminal location makes this unlikely. These results furthersupport the conclusion that the recognition signals for themultiple serine acceptors in the CTP is defined on the carboxyside of each acceptor by the adjacent serine glycosylation site.

DISCUSSION

Newly synthesized glycoproteins undergo numerous modifi-cations in the secretory pathway; for O-glycosylation, severalcompartments are involved. In many cases, the O-linked sites

FIG. 4. Identification of the glycosy-lation site in the mutant T28. A, ala-nine mutations in the CTP includingserines 118–120 (S3) or serine 120 (S20)of the full-length DAsn construct (lanes1–6) or in the T28 truncated form (lanes7–12). Lanes 1, 2 and 11, 12 show themigration of the nonmutated full-lengthCTP and T28 forms, respectively. In B,the serine at 121 in T28 was converted toalanine (DSer1; lanes 5 and 6). The migra-tion of the control DAsn (lanes 1 and 2)and T28 (lanes 3 and 4) are shown.

Biosynthesis of O-Linked Oligosaccharides 20801

by guest on February 18, 2018http://w

ww

.jbc.org/D

ownloaded from

are clustered, and it is not clear if recognition of multiple sitesinvolves overlapping sequences or if the acceptors are eachdefined by discrete signals. Here we used the CTP sequence inthe CGb subunit as a model to address this question since itcontains four closely spaced O-linked sites at the carboxyl endof the protein. Truncated mutants terminating at each Seracceptor were constructed and expressed in CHO cells. The

data show that the CTP region is not a single O-glycosylationsignal, but it is a domain composed of multiple recognitionunits bounded by a Ser-to-Ser motif. A question that arises iswhether or not linkage of the initial GalNAc occurs proces-sively, i.e. in an ordered sequence or randomly. Our data favorthe latter because, when Ser121 was changed to Ala, glycosyla-tion of the remaining sites in the cluster was not impaired. In

FIG. 5. Modification of acceptor activity of Ser121. A describes a series of truncated mutants (see also Table I) where stop codons werecreated at the following amino acids: 127 (T27), 126 (T26), 125 (T25), 124 (T24), and 122 (T22). The autoradiograph in B shows the correspondinglysate (L) and medium (M) forms of these mutants. C shows that proline is necessary for maximal O-glycosylation of Ser121. The carboxyl-terminalproline in T25 (lanes 1 and 2) was converted to Arg (lanes 3 and 4), Ala (lanes 5 and 6), Leu (lanes 7 and 8), or Gln (lanes 9 and 10). D describesa series of alanine (Ala) substitutions on the T25 template. The construct containing an Ala substitution for Pro124 in T25 is designated (ala)1, i.e.

2121Ser-

122Lys-

123Ala-

124Ala

The mutants (ala)2 to (ala)4 refer to increasing chain length to 127 amino acids. The mutant T28 which contains (ala)4 terminates with Ala insteadof the acceptor Ser at 127.

Biosynthesis of O-Linked Oligosaccharides20802

by guest on February 18, 2018http://w

ww

.jbc.org/D

ownloaded from

addition, the independent behavior of the individual sites inthe truncated mutants appears to preclude a coupled mecha-nism for sequential O-linked addition.We attempted to characterize a recognition sequence re-

quired for Ser-linked glycosylation in the CTP. There have beennumerous investigations using in vitro systems containing oli-gopeptides and purified GalNAc transferase to study this prob-lem (13–19). Efforts to identify a consensus sequence usingsuch systems has been difficult because of the absence of mostsecondary and tertiary structure in the peptide substrates andthe presence of multiple GalNAc transferases (30, 31). It is ofinterest that a GalNAc transferase was recently characterizedusing a CGb-derived peptide containing the Ser121, Ser127, andSer132 acceptor sites (30). While our data do not assess if one ormore enzymes are responsible for the observations seen here,the homogeneous changes observed favor a single GalNActransferase. Because of the relative discrete changes observed,we could test certain predictions concerning the recognitionsequences based on the previously reported in vitro observa-tions. It was proposed that accessibility to the serine acceptorsite is a key factor for recognition by the GalNAc transferase(32). This could be a major consideration for mucins where thedensity of O-linked acceptor is high, but the CTP, with itsproline-induced b turns, is flexible, fully exposed to solvent,and thus to the glycosyltransferases. Moreover, accessibilitycannot be the exclusive determinant because in the mutanttruncated at residue 124, glycosylation at Ser121 required thePro124. Substitution of Ala, Gln, or Leu at residue 124 resultedin marked reduction of acceptor efficiency, and mutating Pro toArg or Asp eliminated acceptor activity. This result shows thata proline residue within at least three amino acids carboxy toSer121 is required for recognition. These results are consistentwith studies of the amino acid sequences surrounding theknown O-glycosylation sites (derived from the National Bio-medical Research Foundation Protein Data Base) (11, 12)which indicated the need for one or more proline residueswithin five amino acids carboxy to the acceptor Thr/Ser. More-over, the acceptor efficiency of a T27 mutant, in which theproline residues were all replaced with Ala (-Ser-Lys-Ala-Ala-Ala-Ala-) was comparable to the T25 mutant containing thePro124 3 Ala change. These data show that in addition to

accessibility, the function of the Pro residue is critical, and thatchain length cannot be a sole determinant. Moreover, thatreplacement of the proline with a charged residue eliminatedacceptor activity is consistent with the recent transfectionstudies of Nehrke et al. (21) using a unique reporter constructcomprised of sequences encoding a single O-glycosylation siteand epitopes that facilitate isolation of the glycosylated protein.Data were presented that charged residues in the vicinity ofthe O-glycosylation site significantly depressed glycosylationefficiency when the reporter is expressed in COS-7 or MCF-7cells. Thus, depending on the position in the sequence, chargedamino acids may quench alternative Ser/Thr sites in the CTP,e.g. Ser118/120/130 and Thr140.Although the CTP is readily accessible to the glycosyltrans-

ferases, not all of the Ser/Thr residues between 118 and 145 areglycosylated, e.g. the adjacent serines 118–120 at the amino-terminal side of the Ser121 acceptor site. Moreover, neither ofthese residues are acceptors, even when Ser121 in T28 waseliminated. Although Ser/Thr residues adjacent to the normalacceptor sites may be less efficiently glycosylated in vitro (13,18, 19), the absence of acceptor function by Ser118–120 suggeststhat this region is the base of a b turn and is not as readilyexposed to the glycosyltransferase(s). Given the importance ofthe 26–110 disulfide bond in the CGb subunit for assemblywith the a subunit (33), Ser118–120 may represent a spacer forthe major structure and the acceptor region of Ser121 in theO-linked rich domain. Thus, the folding induced by this disul-fide bond could be a factor in defining the amino end of the CTPregion. This is consistent with recent mutagenesis studies oferythropoietin which show that the conformation of the proteinaffects O-linked acceptor efficiency (20). Our data suggest thatthe GalNAc transferase is very precise for a particular site.This implies that the transfer of carbohydrate to the nativeacceptor suppresses the recognition at other local sites, or thepresence of GalNAc at a Ser abolishes its function as a recog-nition sequence for an adjacent acceptor Ser.Transfected cells expressing CGb provide a feasible system

for studying structure-function of site-specific O-glycosylation.One major issue is whether the sequences required for theinitial step inO-glycosylation, i.e.GalNAc addition, also encodeinformation for the maturation of the growing oligosaccharide

FIG. 6. Carboxyl-terminal borderfor the Ser127 recognition signal. Theinactive Ser127 site in T28 is converted toan acceptor when a construct containingthree additional amino acids was created(T31; lanes 5 and 6). Maximal recognitionis observed in T33 which terminates withSer132 (lanes 3 and 4).

Biosynthesis of O-Linked Oligosaccharides 20803

by guest on February 18, 2018http://w

ww

.jbc.org/D

ownloaded from

chain. This model could address the influence of the flankingprotein sequences and the effect of cell type on the oligosaccha-ride structure by inserting the CTP sequence at defined regionsin a variety of proteins. Our cellular approach, together withmass spectroscopy of purified protein and oligosaccharides, willfurther address the specificity of recognition both for Thr andSer O-glycosylation as well as the enzymatic mechanismsinvolved.

Acknowledgments—We thank Dr. Jacques U. Baenziger and Dr.David Ben-Menahem for critical review of the manuscript and SusanCarnes for her assistance in its preparation. We are grateful to Drs.Susan Daniels-McQueen and Jacques U. Baenziger for their expertisein the GalNAc analysis.

REFERENCES

1. Pierce, J. G., and Parsons, T. F. (1981) Annu. Rev. Biochem. 50, 465–4952. Birken, S., and Canfield, R. E. (1977) J. Biol. Chem. 252, 5386–53923. Keutmann, H. T., and Williams, R. M. (1977) J. Biol. Chem. 252, 5393–53974. Kessler, M. J., Mise, T., Ghai, R. D., and Bahl, O. P. (1979) J. Biol. Chem. 254,

7909–79145. Kalyan, N. K., and Bahl, O. P. (1983) J. Biol. Chem. 258, 67–746. Matzuk, M. M., Hsueh, A. J. W., LaPolt, P., Tsafriri, A., Keene, J. L., and

Boime, I. (1990) Endocrinology 126, 376–3837. Fares, F. A., Suganuma, N., Nishimori, K., LaPolt, P. S., Hsueh, A. J. W., and

Boime, I. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 4304–43088. Furuhashi, M., Shikone, T., Fares, F., Sugahara, T., Hsueh, A. J. W., and

Boime, I. (1995) Mol. Endocrinol. 9, 54–639. Cole, L. A., Birken, S., and Perini, F. (1985) Biochem. Biophys. Res. Commun.

126, 333–33910. Roth, J., Wang, Y., Eckhardt, A., and Hill, R. L. (1994) Proc. Natl. Acad. Sci.

U. S. A. 91, 8935–893911. Wilson, I. B. H., Gavel, Y., and Heijne, G. V. (1991) Biochem. J. 275, 529–53412. O’Connell, B. C., Tabak, L., and Ramasubbu, N. (1991) Biochem. Biophys. Res.

Commun. 180, 1024–103013. Elhammer, Å. P., Poorman, R. A., Brown, E., Maggiora, L. L., Hoogerheide, J.

G., and Kezdy, F. J. (1993) J. Biol. Chem. 268, 10029–1003814. Young, J. D., Tsuchiya, D., Sandlin, D. E., and Holroyde, M. J. (1979) Bio-

chemistry 18, 4444–444815. Briand, J. P., Andrews, S. P., Jr., Cahill, E., Conway, N. A., and Young, J. D.

(1981) J. Biol. Chem. 256, 12205–1220716. Hughes, R. C., Bradbury, A. F., and Smyth, D. G. (1988) Carbohydr. Res. 178,

259–26917. Wang, Y., Abernethy, J. L., Eckhardt, A. E., and Hill, R. L. (1992) J. Biol.

Chem. 267, 12709–1271618. O’Connell, B. C., Hagen, F. K., and Tabak, L. A. (1992) J. Biol. Chem. 267,

25010–2501819. Wang, Y., Agrwal, N, Eckhardt, A. E., Stevens, R. D., and Hill, R. L. (1993)

J. Biol. Chem. 268, 22979–2298320. Elliott, S., Bartley, T., Delorme, E., Derby, P., Hunt, R., Lorenzini, T., Parker,

V., Rohde, M., and Stoney, K. (1994) Biochemistry 33, 11237–1124521. Nehrke, K., Hagen, K. F., and Tabak, L. A. (1996) J. Biol. Chem. 271,

7061–706522. Ho, S. N., Hunt, H. D., Horton, R. M., Pullen, J. K., and Pease, L. R. (1989)

Gene (Amst.) 77, 51–5923. Sachais, B. S., Snider, R. M., Lowe, J. A., III, and Krause, J. E. (1993) J. Biol.

Chem. 268, 2319–232324. Matzuk, M., and Boime, I. (1988) J. Biol. Chem. 263, 17106–1711125. Matzuk, M. M., Kornmeier, C. M., Whitfield, G. K., Kourides, I. A., and Boime,

I. (1988) Mol. Endocrinol. 2, 95–10026. Corless, C. L., Matzuk, M. M., Ramabhadran, T. V., Krichevsky, A., and

Boime, I. (1987) J. Cell Biol. 104, 1173–118127. Laemmli, U. K. (1970) Nature 227, 680–68528. Green, E. D., van Halbeek, H., Boime, I., and Baenziger, J. U. (1985) J. Biol.

Chem. 260, 15623–1563029. Matzuk, M. M., Krieger, M., Corless, C. L., and Boime, I. (1987) Proc. Natl.

Acad. Sci. U. S. A. 84, 6354–635830. Sørensen, T., White, T., Wandall, H. H., Kristensen, A. K., Roepstorff, P., and

Clausen, H. (1995) J. Biol. Chem. 270, 24166–2417331. White, T., Bennett, E. P., Takio, K., Sørensen, T., Bonding, N., and Clausen, H.

(1995) J. Biol. Chem. 270, 24156–2416532. Hill, H. D., Jr., Schwyzer, M., Steinman, H. M., and Hill, R. L. (1977) J. Biol.

Chem. 252, 3799–380433. Suganuma, N., Matzuk, M., and Boime, I. (1989) J. Biol. Chem. 264,

19302–19307

Biosynthesis of O-Linked Oligosaccharides20804

by guest on February 18, 2018http://w

ww

.jbc.org/D

ownloaded from

Tadashi Sugahara, Mary R. Pixley, Fuad Fares and Irving Boime Using Site-directed Mutagenesis and Gene Transferin VivoSubunit

β-Glycosylation Sites in the Chorionic Gonadotropin OCharacterization of the

doi: 10.1074/jbc.271.34.207971996, 271:20797-20804.J. Biol. Chem. 

  http://www.jbc.org/content/271/34/20797Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/271/34/20797.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on February 18, 2018http://w

ww

.jbc.org/D

ownloaded from