`
`CSL EXHIBIT 1055
`
`Page 1 of 59
`
`CSL V. Shire
`
`CSL EXHIBIT 1055
`CSL v. Shire
`
`Page 1 of 59
`
`

`

`
`
`Ellular and Molecular Llfl.‘ Science
`
`.w.
`
`
`
` formerly Experientia
`
`Scope
`The multidisciplinaryjournal CflrfLS, Cellular andMolecular Life Sciences (formerly Experientia) publishes
`articles, mini-reviews, reviews and mule-author reviews covering the latest aspects of biological and biomedi
`research. The journal will consider contributions focusing on molecular and cellular aspects of biomedicine,
`biology, immunology, molecular genetics, neuroscience, biochemistry, phannacology, and physiology related
`pharmacology.
`
`'
`
`Editor-in-chief
`
`Editorial Board
`
`Advisory Board
`
`
`
`C.C.F. Blake, Oxford
`D. Arigoni, Ziirich
`E. Carafoli, Zfirich
`TA. Biclrle, Basel
`KJA. Davies, Los Angeles, CA J. Engel, Basel
`G. Dirheimer, Strasbourg
`U. FeigeJhousand Oaks,
`T. Imoto, Fukuoka
`J.-M. Frere, Liege
`H. Jfimvall, Stockholm
`M. Go, Nagoya
`E. Kubll, Zfirich
`T. Kirchbausen, Boston, MA
`H.-G. Rammensee, Tiibingen
`K. Kuchler, Wien
`R. Timpl, Marflnsried
`F. Mayor Jr., Madrid
`W. Wabli, Lausanne
`R.I. Norman, Leicester
`M. Parnham, Zagreb
`P. Potier, Gif-sur-Yvette
`I. Slaninova, Praha
`P. Venetianer, Szeged
`
`Professor Dr. P. Jolles
`Laboratoire de Chimie
`des Substances Naturalles
`URA CNRS No. 401
`Museum National d’Histoire Naturelle
`63, rue Buffon
`F-75005 Paris I France
`
`Editorial Office
`
`CMLS
`Birkhfiuser Verlag
`P.O. Box 133
`CH—4010 Basel I Switzerland
`
`Tel. +41 61 205 0? 68
`Fax +41 61 205 07 95
`e—mail: CMLSedit@birkhauser.ch
`
`Visit our Web site
`http:/r‘mbirkhausench
`
`Cover
`
`llhistration of a low-mergy confermation ofN-aoetyllactosamine (lacNAc), a common terminal section of N—glycans. It
`me.g.bebmmdbymmflimlwfim(galmfins),safingmdmldngmmt(hnmymowdedbym. C.-W.vos1derLieth.
`Heidelberg, Germany).
`G. Router and H.-J. Gabius, p. 368, this issue.
`
`—————_—_____________________
`
`Page 2 of 59
`
`Page 2 of 59
`
`

`

`
`
`456
`
`Phorbol ester exposure activates an ARI-mediated
`increase in ERGO-l messenger RNA expression
`in human ovarian tumor cells
`Q. Li, L. Zhang, B. Tsang, K. Gardner,
`F. Bostick-Bruton and E. Reed
`(USA)
`
`467
`
`Protein kinase-dependent overexpression of the
`nuclear protein pirin in c-J UN and RAS
`transformed fibroblasts
`
`A.-C. Bergman, A. A. Aiaiya, W. Wendter,
`B. Binetruy, M. Shoshan, K. Sakagucht',
`T. Bergman, U. Kronenwett, G. Auer.
`E. Appetta, H. Jiirnuatt and S. Linder
`(Sweden, Germany, France and USA)
`
`Repetitive use of a phosphate-binding module in
`DNA polymerase ,6. Oct-l POU domain and
`phage repressors
`K. Yum, M. Shianyu. K. Kawatani and M. Gt?
`(Japan)
`
`Alloxan acts as a promidant only under reducing
`conditions: influence of melatonin
`H.-J. Brb‘mme, H. Ebeit, D. Pesehke and
`E. Pesehke
`
`(Germany)
`
`Repolarization abnormalities in cardiomyocytes of
`dogs with chronic heart failure: role of sustain-
`ed inward current
`A. I. Undrovinas, V. A. Maitsev and
`H. N. Sabban
`(USA)
`
`The involvement of the renin-angiotensin system in
`the regulation of cell proliferation in the rat
`endometrium
`M. Pawlikowski, G. Maren-Murat! and
`5'. Alaska
`
`472
`
`487
`
`494
`
`506
`
`A common
`
`ancestor
`
`for
`
`a
`
`subunit
`
`in
`
`the
`
`(Poiand)
`
`mitochondrial proton-translocating NADH:
`ubiquinone oxidoreductase (complex I) and
`short-chain dehydrogenasesireductases
`M. E. Baker, W. N. Grundy and C. P. Eikan
`(USA)
`
`511
`
`Announcement
`
`The Instructions for Authors appear on the inside back
`cover of this issue
`
`Page 3 of 59
`
`-:_
`
`- SSINO. 3, 1999
`
`.March 1999
`
`Immunotolerant functions of HLA-G
`E. D. Caroseiia, J. Dausset and N. Rouas-Freiss
`(France)
`
`Antisense RNA gene therapy for studying and
`modulating biological processes
`B. Weiss. G. Dauidkova and L.-W. Zhou
`(USA)
`
`Structural
`adhesion
`
`view of cadherin-mediated cell-cell
`
`J. R. Aiattia. H. Kurokawa and M. ikura
`(Canada)
`
`Eukaryotic glycosylation: whim of nature or
`multipurpose tool?
`G. Renter and H.-J. Genius
`(Germany)
`
`Peptidyl-prolyl cis-trans isomerases, a superfamily
`of ubiquitous folding catalysts
`.S'. "F. Garnet and M. A. Marahiet
`(Germany)
`
`Roles of the DNA mismatch repair and nucleotide
`excision repair proteins during meiosis
`D. T. Kirkpatrick
`USA
`(
`)
`
`' arch Articles
`
`Page 3 of 59
`
`

`

`[SSN 1420-632X
`
`Surface Airmail lifted (SAL): For subscribers in Japan, India, Australia, and
`New Zealand, price lists indicating special carriage charges are available.
`Airmail delivery to all other countries is available upon request.
`Delivery (North America): Alljoumals are airspeeded to subscribers in the
`USA and Canada.
`Changes of Address: Please report any change of address at least 4 weeks
`before it is to take effect, giving your current as well as your future address.
`Claims: Claims for issues not received should be made promptly - at the very
`latest immediately upon receipt of the issue succeeding the missing number.
`Cancellations: No cancellations are accepted for the calendar year once sub-
`scription has been entered No refunds will be made. Cancellations must be
`received by September 30. 1999.
`Back Volumes of this journal are available. Prices are quoted at the current
`subscription rates per volume.
`
`Correspondence concerning advertisements. general publishing matters and
`procedures should he addressed to:
`Birkhauser Vet-lag, PO. Box :33. Cit-40m Basel. Switzerland
`Fax: +4] 61 205 01" 91. e-mail: cmls@birkhauset:ch
`
`30 reprints of each paper will be mplied free of charge. Prices for additional
`reprints can be obtained through the Publisher.
`
`Page 4 of 59
`
`Subscription Information
`The journal CHIS. Cellular and Molecular are Sciences formerly Experientia
`is published monthly by Birkht'ruser Verlag, in 1 volume per year.
`Annual subscription‘
`sFr. P488.-
`Single issue‘
`aFr.
`l48.~
`“ Suggested retail prices, postage and handling not included.
`
`Subscription orders should be addressed to:
`Birkhiuser Veriag, PO. Box [33. CPI-4010 Basel. Switzerland
`Fart: + 4] 61 205 O? 91 , e—mail: subscriptions@birkhauserzch
`or through your bookseller or subscription agency.
`
`USA Postmortem Send address changes to: CMLS. Cellular and Molecular
`Life Sciencesformerly Experientia. Bitkhauser Boston, 333 Meadowlands
`Parkway. Secaucus, NJ 01094. USA.
`Periodicals postage paid at New York, N.Y. and additional mailing offices. The
`North American agent for Bimuser Vet-lag is Birkhauser Boston, l'lS Filth
`Avenue. New York. NY 10010, USA.
`
`Copyright Statement
`Submission of the manuscript implies that (submitted) manuscripts have
`I
`been published, nor will be simultaneously submitted or published el :-
`By submitting a manuscript. the authors agree that the copyright for their
`is transferred to the Publisher if and when the article is accqsted for --
`.
`‘.
`The copyright covers the exclusive right to rupruduce and distribute the -
`by means of printing. reprinting or any other method deemed appropriate i
`Publisher including electronic means. photographic reproductions, | ' m -
`any other reproductions of similar nature, as well as translations.
`
`'
`
`:._
`
`Unless otherwise specified Birkhauser Verlag AG holds all rights. All =- '
`published in thisjournal are protected by copyright; they may not be dupl '-
`ated in hard copy or machine readable form or reproduced phoi ;.
`..-. --
`and they may not be redistributed. transmitted, or translated or stored on
`microfilm. in electronic data bases, on local servers, video disks etc.
`- n- .
`L
`prior written permission from Birkhauser Wiring AG. Connnetcial use or
`redistribution can only be allowed by specific licence agreement. Requests 1.;
`such permissions should be addressed to:
`Birklu'iuser Vet-lag. Journals Permissions Department, P.O. Box 133,
`(CH-40m Basel. Switzerland. e-mail: pennissions@birkhauser.ch
`
`'
`
`..
`Photocopying: Single photocopies of single articles may be made for 4
`use as allowed by national copyright laws. Permission of the Publisheratll
`ment of a fee is required for all other photocopying, including multiple or
`matic copying. copying for advertising or promotional purposes, resale. m -'
`forms of document delivery.
`In the USA, users may clear permissions and make payment through the
`Copyright Clearance Center Inc.. 222 Rosewood Drive. Denver's. MAIN-
`USA. In Germany, users may clear permissions and make payment titroud'r
`Verwertungsgesellschafl WORT, Munchen. In other countries where a load
`copyright clearance centre exists. please contact it for information on - »-
`permissions and payments.
`
`-'
`
`Disclaimer: While the advice and information in this journal is believed to
`true and accurate at the date of its going to press. neither the authors, the
`tors. nor the Publisher can accept any legal respomibility for any errors or
`omissions that may be made. The Publisher makes no warranty. express or
`implied, with respect to the material contained herein
`
`The use of general descriptive names, trade names. trademarks, etc. in this
`publication, even if not specifically identified, does not imply that these ‘-
`are not protected by the relevant laws and regulations The Publisher and
`editor can give no guarantee for the information on drug dosage and :.n .
`tion contained in this journal. The respective user must check its accuracy _
`consulting other sources of reference in each individual case.
`
`_
`
`.
`
`Abstractedfladesed in:
`Biochemistry Abstracts. Biological Abstracts. CAB. Intemau'onal, CAB
`access, Chemical Abstracts. Current Contents. EMBASEEWB .-
`Elsevier BIOBASElCtu-rent Awareness in Biological Sciences,
`Index Mdicquedline. Infon'nation Ventmes. Nutrition Abstracts 8c '
`Review of Plant Pathology, Springer Journals Preview Service.
`
`Printed on acid-free paper produced from chlorine-free pulp
`0 i999 Birkhfiuser Verlag, Basel - Boston - Berlin
`
`Page 4 of 59
`
`

`

`
` CMLS Cellular and Molecular Life Scion '
`
`CMLS, Cell. Mol. Life Sci. 55 (1999) 368-422
`1420-682XI99f030368-55 $ 1.50 + 0.20.“)
`© Birkhauser Verlag, Basel. 1999
`
`Review
`
`
`rather common type of posttranslational
`
`Eukaryotic glycosylation: whim of nature or multipurpose
`tool?!"
`
`G. Renter? and l-l.-J. Cabins“
`
`lnstitut fiir Physiologische Chemic, Tierarztliche Fakultéit, Ludwig-Maximilians-Universitat, Veterinarstr. 13,
`D-80539 Miinchen (Germany), Fax +49 89 21802508, e-mail: gabius@tiph.vetmed.uni-muenchen.de
`
`Received 2 October 1998; received after revision 1? November 1998: accepted 1? November 1998
`
`Abstract. Protein and lipid glycosylation is a ubiquitous
`phenomenon. The task of cataloguing the great struc-
`tural variety of the glycan part has demanded consider-
`able efforts over decades. This patient endeavor was
`imperative to discern the inherent rules of glycosylation
`which cannot affirm assumptions on a purely coinciden-
`tal nature of this type of protein and lipid modification.
`These results together with theoretical considerations
`uncover a salient property of oligosaccharides. In com-
`parison with amino acids and nucleotides, monosaccha—
`rides excel in their potential to serve as units of hardware
`for storing biological information. Thus, the view that
`glycan chains exclusively affect physicochemical proper—
`ties of the conjugates is indubitably flawed. This original
`concept has been decisively jolted by the discovery of
`endogenous receptors (lectins) for distinct glycan epi-
`topes which are as characteristic as a fingerprint or a
`
`signature for a certain protein (class) or cell type. Recent
`evidence documents that these binding proteins are even'
`endowed with the capacity to select distinct low-energy
`conformers of the often rather flexible oligosaccharidcs,
`granting entry to a new level of regulation of ligand
`affinity by shifting conformer equilibria. The assessment
`of the details of this recognition by X-ray crystallogra-
`phy, nuclear magnetic resonance spectroscopy. mi-
`crocalorimetry and custom-made derivatives is supposed
`to justify a guarded optimism in satisfying the need for
`innovative drug design in antiadhesion therapy.
`for
`example against viral or bacterial
`infections and un-
`
`wanted inflammation. This review presents a survey of
`the structural aspects of glycosylation and of evidence to
`poignantly endorse the notion that carrier~attached gly-
`can chains can partake in biological information transfer
`at the level of cell compartments. cells and organs.
`
`Key words. Cell adhesion; drug design; glycan metabolism: glycolipids; glycoproteins; glycosaminoglycans; lectins;
`recognition.
`
`Introduction
`
`Proteins are ubiquitous compounds in all living organ-
`isms with many well-established biological functions.
`They play important roles, for example as biocatalysts,
`transport and targeting devices, in establishing cell-type
`
`
`* Dedicated to Prof. Dr. F. Cramer on the occasion of his 75th
`birthday.
`** Corresponding author.
`1‘ Present address: Haferkarnp 28, D-24145 Kiel (Germany).
`
`Page 5 of 59
`
`specificity and in various aspects of cell sociology. Post-
`translational modification of proteins adds a further
`level of regulation to their activity and can thus enlarge
`their capacity to take part in regulatory events. The
`reversibility of the introduction of small
`inorganic
`groups into amino acid side chains such as phosphate
`allows proteins to be used as molecular switches. Evi-
`
`the capacity of the peptide backbone is not
`dently,
`restricted to serving as acceptor for a phosphate group-
`Another
`
`Page 5 of 59
`
`

`

`Review Article
`
`369
`
`tures. analysis, conformation and biosynthesis, and out-
`line the current concept of the biological role of these
`fascinating molecules.
`
`The sugar code
`
`The term ‘complex carbohydrates’ should not solely be
`thought to connote a technically demanding task of
`cataloguing structures. Principally, oligomers of any
`class of biomolecules can establish a coding system for
`information storage, resembling words in a language. In
`this sense, structural complexity entails the intriguing
`capacity to serve as versatile hardware for biological
`information transfer. The capacities for this duty are
`not equal for the biochemical coding systems. Amino
`acids and nucleotides are linked nearly exclusively in
`
`H0
`
`OH
`0
`
`OH OH
`
`HO
`
`OH
`0 OH
`
`OH
`
`OH
`
`[3 - lurenoee
`
`open chain
`(Fischer projection]
`
`/
`
`\ H
`
`c: - pyranose
`
`[5 - pyranose
`
`HO
`
`0
`
`HO
`
`[10
`
`0mo"H0
`
`Reeves confonnatlon
`4'01
`
`Figure l. Mutarotation of D-glucose with the different representa-
`tions of D-glucose as Fischer projection. Haworth projection and
`Reeves ‘C. chair conformation.
`
`
`
`. Cell. Mol. Life Sci. Vol. 55. 1999
`
`3.: ification is protein glycosylation. A potential indi-
`"s of a biochemical role of the carrier-immobilized
`units is furnished by the observation that a sec-
`abundant class of biomolecules shares the property
`
`_phatide phosphatidylinositol with its lipid anchor
`and hydrophilic inositol head group is likewiSe
`' ed as a means of glycan attachment. Glyco-
`teins and glycolipids are referred to as glycoconju-
`-
`and members of
`the group of
`complex
`
`_ parts are directed into the outer environment and
`constitute the glycocalyx. While the glycan chains
`picls rarely extend beyond a dozen monosaccharide
`and therefore remain in close vicinity to the cell
`' ace,
`those of glycoproteins can protrude further
`the cell membrane due to either long glycan chains
`linkage to an amino acid at a remote position within
`peptide chain.
`the early years of research on these compounds,
`-_ hasis had to be laid on the development of anaiyti-
`' ools to allow structure determination as a prerequi-
`
`for any taking stock of functional aspects. Now a
`of sophisticated and further continuously improved
`ods is available, consequently making it easier to
`the initially puzzling complexity of saccharide se-
`_.ces on record [145]. Only after having pursued
`- ctural analysis. which revealed a wide range of
`.3? can structures. could attention be turned to the enzy—
`tic routes of biosynthesis and to the delineation of a
`tive physiological meaning, these lines of reasoning
`for example, in reviews by Roseman [7] and
`interburn and Phelbs [8]. This historical development
`-. the different phases of research on complex carbohy-
`tes is' nicely described by a series of authoritative
`_ ews [9— 18]. Currently. the enormous strides taken in
`- area of sugar-mediated recognition suggest
`that
`T
`’3; can constituents of proteins and lipids establish an
`'-"habetical system, adding to the well-accepted coding
`iii ortance of nucleic acids and peptide backbones [19,
`, Citing from the article of Winterburn and Phelbs
`.-;.'-'1 ‘carbohydrates are ideal for generating compact
`"ts with explicit
`informational properties since the
`utations on linkages are larger
`than can be
`hieved by amino acids. and. uniquely in biological
`.= lymers, branching is possible”. Therefore, ‘the signifi»
`fence of the glycosyl residues is to impart a discrete
`ugnitional role on the protein’ (or any other carrier
`.ckbone). This notion has meanwhile continuously
`-_"~r enough support to keep it
`in the eye of a
`adership from different areas of life sciences [1, 21—
`6]. In order to familiarize the reader with this concept.
`' s will summarize the major features of glycan struc—
`
`Page 6 of 59
`
`Page 6 of 59
`
`

`

`
`
`laut in German language. If now multiple saccharide
`chains, each containing a set of possible structural
`variations, are attached to one protein backbone as is-
`the case for many glycoproteins, one can grasp the
`immediate relevance of these calculations and findings
`for the theoretical content of information stored in the
`glycoconjugate.
`Posttranslational modification by glycosylaticn can.
`even lead to the creation of discrete subsets of glycc~
`proteins (glycoforms. according to Rademacher et a].
`[39]). They share an identical polypeptide sequence but
`differ in the glycan composition. As will exemplarin
`be outlined in the section on glycans and disease.
`monitoring the relative proportions of serum transfer-
`rin glycoforms has diagnostic value. It is indicative of
`clinically defined inherited syndromes or of impair-
`ment of liver function by distinct causes. In line with
`the given reasoning,
`it
`is tempting to speculate that
`this potential is not wasted as a costly whim of nature
`but
`rather establishes a multipurpose tool
`that
`is
`taken advantage of in diverse circumstances. Before
`the functional aspect can be discussed. it is essential to
`provide an insight into the structural organization of
`protein and lipid glycosylaticn.
`
`in which the two anomeric centers are linked to -
`other. Sequence permutation doubles this number,
`this calculation. the potential for alteration of the ' 5.;
`size (pyranoscifuranose, see fig. 1) has not even beefi,
`incorporated. Using two amino acids or nucleoti -...;
`only two different structures can be generated.
`If We allow the formation of any linear or bran -.-..__tj
`trisaccharide from a pool of 20 different monosacclurg
`rides, we end up with 9 >< 10" possibilities. whe
`only 8000 tripeptides can be formed from the
`amino acids [19]. For a hexamer this gap widens ttf»
`more than seven orders of magnitude. This enormous.
`coding capacity of oligosaccharides
`renders
`them-
`highly suitable hardware for
`information storage.
`However,
`it
`is essential
`to add that
`the theoretical
`number of combinations has by far not been un-
`earthed in nature, probably because the equally high-
`number of corresponding glycosyltransferases is not-
`available in the organism [5. 27—29]. A compensatory
`diversification is found instead by addition of ‘small
`substituents’ such as phosphate, acetate or sulfate [19,.
`26. 30m38]. As will be illustrated later
`in seventh
`
`places, single or multiple modifications with such sub-
`stituents can even occur on one monosaccharide.
`residue, affecting its character as code letter. For ex;
`ample,
`the 4-sulfation of N-acetyl-D-galactosamine
`(GalNAc) can be likened to the formation of an um—
`
`370
`
`G. Renter and H.—J. Gabius
`
`linear chains via peptide or phosphodiester bonds.
`Only the primary sequence of the monomeric building
`units in the chains allows for variability in these two
`cases.
`This
`situation
`changes
`drastically when
`cligosaccharides are considered. It
`is pertinent to re-
`call aspects of the basic biochemistry of sugars to ap-
`preciate these extraordinary properties of this type of
`oligomer, which can reach an unprecedented level of
`storage capacity.
`
`of
`teaches mutarotation
`biochemistry
`Textbook
`monosaccharides such as D-glucose primarily from the
`perspective of a measurable physical parameter {opti-
`cal rotation). Since linear forms of. for example. any
`hexose are normally present only in minute amounts,
`hemiacetals or -ketals are invariably found. The mu-
`tarotation of a hexose yields four different ring struc-
`tures for each monomeric unit. as depicted in figure 1
`for D—glucose (Gle). Viewed from the perspective of
`evaluation of hardware abilities, this result is of rele-
`vance. Similarly far~reaching is
`the fact
`that each
`monosaccharide presents a variety of hydroxyl groups.
`Formation of a bond between two monosaccharide
`units by condensation can engage diverse possibilities,
`a daunting task for the organic chemist. When the
`anomeric center of a monosaccharide participates in
`the glycosidic bond, each anomer can theoretically re-
`act with one of the five, chemically nearly equivalent
`hydroxyl groups of another hexopyranose unit. Figure
`2 shows that permutations on the basis of only these
`
`
`
`Figure 2. Possible linkage points for a disaccharide consisting of
`Ga] and Ole. The different structures that can be formed result
`from variations in (i) anomeric configuration (a or 5}. (ii) ring
`size. that is furnnose or pyranose, with only pyranose depicted in
`the figure {see fig.
`I).
`(iii)
`linkage points as indicated by the
`arrows. with the dotted line for the formation of nonreducing
`disaccharides involving both anomeric centers, and {iv} the se-
`quence of the two monomers. Excluding linkage via the two
`anomeric centers. each C! hydroxyl group can be conjugated to
`any of the four available hydroxyl groups at C2, C3. C4 and C6
`in at- or fi-configuration. yielding 8 isomers for each of the two
`anomeric centers.
`
`Page 7 of 59
`
`Page 7 of 59
`
`

`

`
`
`OH
`
`O
`
`NHAc
`
`NH
`
`[O ‘.
`\c/
`.'
`[
`
`KfNH
`HIV/0'20
`
`0%?
`
`NH
`
`x
`
`0
`
`Cr:-
`
`o
`
`
`NHAc
`
`\NH
`
`a
`
`b
`
`3. Linkage points for the covalent attachment of common
`d Owglycans to protein. The GlcNAcfi l-Asn-X-Ser sequon is
`eristic for most N-glycans. The carbohydrate chain is
`a at the C-4 of GlcNAc as indicated by the wavy line {a};
`.- of core fucosylation, see figure IS. The GalNAcxl-Ser
`re of mucin-type O-linked glycoproteins can be extended
`tion of further saccharides to C-3 and C-6, leading to the
`'_uction of branches (.5).
`
`w a attachment to proteins
`
`,glycoproteins. the glycan chains are nearly always
`"a
`= ntly linked to suitable functional groups of amino
`in side chains within a protein. The elucidation of the
`_=:"-" ures 'ol‘ the linkage region between carbohydrate
`_ protein was
`initiated with the description of
`NAcfil-Asn for N-glycans and GalNAcorl-Ser or
`for O—glycans (fig. 3) [1, 10. 25, 26, 40]. Ensuing
`ta rch considerably extended our knowledge in this
`
`re 4. Structure of the «4Xylfi l-Ser linkage of glycosarninogly—
`4":
`W s.
`
`Page 8 of 59
`
`a.
`
`l.Mol.Lil‘e Sci. VOLSS. 1999
`
`Review Article
`
`371
`
`3‘
`
`b
`
`6
`
`H30
`
`H0
`
`«mo
`
`0
`
`OH
`
`0H
`
`5
`
`HN/
`Mcéo
`L__
`
`H0
`
`H
`
`Ho
`
`H0
`
`H0
`
`HO
`
`0
`
`0H
`
`HN
`°\/kc4’o
`l
`
`OH
`
`.
`
`HN/
`Ody/0
`l
`
`OH
`
`NHAc
`
`0
`
`OH
`
`N
`
`o
`
`H\\c//
`
`[.
`
`NH
`
`d
`
`/C-."':.-o
`
`Figure 5. Less common Ser glycosylations are found in the (a)
`-3Fucal-Ser linkage and (b) Glcfi l-Ser linkage in EGF-Iike do-
`mains. Another possible mode of O-glycosylation as well as one of
`glucose attachment to protein are presented by the linkages (c)
`GlcNACfil-Ser and (d) Glcfll-Asn. Type (c) is predominantly
`found in intracellular proteins in the nucleus and cytoplasm.
`
`area [15. 28. 29, 41, 42]. The current status of delin-
`eation of the rules that govern the selection of suitable
`glycosylation sites in the peptide sequence has enabled
`explanation of the nonrandom nature of this glycan
`attachment. It has been recognized that for N-glycosidic
`linkage a peptide sequence of Asn-x-Seri'l‘hr (a sequon)
`is required, in which the variable amino acid X seems to
`be important for the efficiency of recognition of the
`resulting sequon [43—45]. Additionally, the amino acid
`following a sequon (the Y-position in Asn-X-Serfl'hr-
`
`Page 8 of 59
`
`

`

`3?2
`
`G. Renter and H.-.l. Gabius
`
`Y) has a notabIe impact on glycosylation efficiency,
`especially for Asn-X-Ser sequons, Pro, Glu and Trp
`causing the most negative effect [46. 47]. For human and
`bovine plasma proteins C. human von Willebrand fac-
`tor and human CD69 an unusual Asn-X—Cys sequon
`has been identified for N~glycan attachment [48, 49].
`The definition of the sequon allows potential glycosyla-
`tion sites in any protein to be deduced on the basis of
`amino acid sequence data. No equivalent peptide se-
`quence requirement has been described so far for classi-
`cal mucin-type O-glycosylation. The initiation of chain
`growth by several transferases with varying degrees of
`stringency of substrate recognition has at least suggested
`general rules and an algorithm of predictive value [50.
`51]. A substantial functional redundancy of these trans-
`ferases was inferred by mutagenic recombination of one
`gene with a deletion in the catalytic region, which did
`not
`impair O~glycan synthesis [52]. The lack of an
`0~glycosylation sequcn does not mean that the reaction
`of the uridine diphosphate (UDP)-GalNAc:polypeptide
`N-acetngaIactosaminyltransferases can take place with-
`out sequence requirements. The spatially accessible ac-
`ceptor
`amino
`acid
`appears
`to
`be
`surrounded
`preferentially by Ser, Thr, Pro. Ala and Gly residues
`[28, 29, 50, 53]. Serfl‘hr residues not only serve as
`O-glycan acceptors but also as adaptors for another
`class of glycans. The third type of complex carbohy-
`drates are the glycosaminoglycans using the glycan-
`protein linkage 4Xylfi l-Ser/Thr (fig. 4} [54]. Usually, a
`trisaccharide 4Gchfl l-3Cialfil-3Cialfil- is attached to
`the 4-position of Xyl before the typical building blocks
`of the individual glycosaminoglycan are added [55].
`In addition to allowing GalNAc and Xyl attachment,
`the hydroxyl groups of Serfl‘hr are sites for the incorpo-
`ration of other glycan modifications. Despite their com-
`paratively low frequency they should not be overlooked
`in this context. On the contrary,
`they illustrate the
`existence of cytoplasmic glycosylation pathways inde-
`pendent of these classical routes, each case an economic
`argument against a merely superfluous expression of the
`engaged glycosyltransferases. Epidermal growth factor-
`(EGF)-like domains of several glycoproteins including
`blood coagulation factors VII. IX and X11, plasminogen
`activators and protein Z are the targets of a sequon-de-
`pendent O-fuccsylation and O—glucosylation [56. 57].
`The enzymatic introduction of oc-L-fucose (a-L-Fuc);
`(fig. 5a) can initiate extension to GIcfll-3Fucrx l-SerfThr
`or to a sialylated N—acetyllactosamine sequence at the
`3-position [58, 59]. The sequon dependence of fucosyla-
`tion in the EGF-Iike domains of tissue-type plasmino-
`gen activator (t-PA) and urckinase was reinforced by
`the absence of this modification in the human chimeric
`plasminogen activator Kztu-PA lacking this sequence
`element [56, 60]. The consensussequence for the reac-
`tion of this fucosyltransferase is Cys-X-X-GIy-Gly—Serf
`
`Eukaryotic glycosylatioq
`
`Thr-Cys-, which can also be found in other types of
`proteins besides EGF module-bearing proteins [56, 591
`Once added, protein-bound L-l‘ucose has been intimated
`to be important for mediating binding of the fucosy.
`lated protein to human HepGZ hepatcma cells, because
`fucose-free EGF-like domains lack this capacity [6]].
`Since binding and degradation of IPA is not inhibited
`by 50 mM L-fucose in the case of human and rat
`hepatoma cells [62],
`this suggestion requires further
`experimental support. Another possibility concerns the
`
` 3.
`
`OH
`
`0
`
`OH
`
`H
`
`H
`
`/
`
`o-
`HN
`(Kg/O\/i\C/7
`é',
`l
`
`o
`
`Ho
`
`HO
`
`0H
`
`0
`
`H
`
`,
`
`/'
`
`HN
`
`b
`
`c
`
`Figure 6. C—C linkage of Man and Trp in human RNase {ri-
`bonuclease) U, (a). This is currently the only example known
`without N- or O-glycosidic linkage between saccharide and
`protein. Neither biological implications nor enzymatic ways for
`biosynthesis are known so far. Structure of the Manel-P-Scl‘
`linkage as an example of phosphoglycosylation (b). Instead of
`D-mannose, Fuc, GlcNAc and Xyl may also be found. MancI-Ser
`linkage as common glycosylation motif in yeast (:1.
`
`
`
`Page 9 of 59
`
`Page 9 of 59
`
`

`

`
`-:-'._.-: --Cell. Mol. Life Sci. Vol. 55. 1999
`
`Review Article
`
`373
`
`R
`(I?
`o
`"Ca A xHN
`ii
`NH
`('3 wO/$?O/
`°
`
`.
`
`m1—2Manu1-6Manu146lefla14mmm1
`a
`1
`Babiesdutmsht
`2
`lGalen
`
`{Gm-9L6
`0
`o
`l
`ii
`"
`TfiTNO/ \(cmuCHa
`0x. Acne.an
`‘1'le
`fi
`c
`
`ules are embedded in the cell membrane via the fatty acids (often
`of this Family of molec
`7. GPI anchor structure. All members
`JI'acylglyool. Starting with D-glucosamine.
`of the phospholipid part which can comprise ceramide. mono- or diacylgtycerol or alkyl
`hot! to this membrane-embedded anchor
`wsaccharide that varies structurally among the different glycoproteins is covalently attac
`each central Denannose unit is a major source of structural
`phosphoethanolamine} at
`rporation of substituents (saccharides,
`al-Z-linked mannose residue of the core glycan.
`.
`ethanolamine is introduced to the
`hydrophobic C-terminus. which generates the
`idation follows cleavage of the peptide backbone 10412 residues away from the
`Asn. Gly. Ala. Ser or Cys), yielding its glypiation. Potential
`terminus. A protein is attached by its w-site (preferentially Asp.
`f a modification at the inositol residue will hamper Pl-PLC
`-_r sites for (G)PI-PLC and GPl-PLD are indicated. Introduction 0
`(G)Pl-PLCID: (glycosyl) phosphatidylinositol phospholipases CID.
`
`
`
`'ve influence of fucosylation at ThrlS on the
`h factor activity of the EGF-like domain in hu-
`urokinase-type plasminogen activator [63. 64]. As
`ted above. the EGF module for example in t-PA
`_ not only harbor the 0-fuccsylation. It also carries
`'j' O-linked glucose unit (fig. Sb) in the consensus
`'1 ence Cys-X-Ser-X~Pro-Cys of the EGF domain [56.
`.1". The glucose moiety may further be substituted at
`'
`' -position by Xylrxl-3Xylot1-, as found in blood
`tion factors VII and IX, protein 2, t-PA and
`s- thrombospondin [56. 65].
`ntrast to the rather restricted appearance of these
`modifications [56], a fi-linked GIcNAc (fig. 5c) can
`Mached to an extended array of proteins [6640}.
`far, no elongation of this type of modification has
`reported. It is confined to a single monosaccharide
`to the protein, and is present on a steadily
`_ g list of intracellular proteins. for example cy-
`eletal constituents such as erythrocyte band 4.]
`-' . neurofilaments, cytokeratins 8. 13 and 18. tran-
`.
`-;:-: factor Spl or c-myc, estrogen receptor and
`-[70, 71]. A reciprocal relationship to phosphoryla-
`is assumed for the reversible formation of Glc-
`_ fil-SerfThr, as such glycosylation sites may also be
`ts of protein kinase(s) action. Functional implica-
`5 also include a role in transcriptional initiation by
`A polymerase I! and in cytoskeletal assembly and
`etion [TO—72]. GlcNAcylation is also possible
`hydroxyproline in slime molds. where a GalfFuc-
`taining pentasaccharide appears to be the nonim—
`ogenic
`product of
`a
`hitherto
`unrecognized
`plasmic O-glycosylation pathway [73]. The lack of
`unogenicity is reason to suspect occurrence of simi-
`structures in mammals. Hydroxyl groups of amino
`a side chains can further accommodate conjugation
`'- th DI-Gal (Ser in plant glycoproteins such as the
`
`potato lectin}. with e-Glc (Tyr in glycogenin, the primer
`for glycogen bicsynthesis),
`fi-Gal
`(Hyl
`in collagen)
`which can be extended to a disaccharide by al,2-linked
`Glc (surprisingly. the presence of this otherwise uncom-
`mon disaccharide O-linked to Thr is also known from
`the surface layer glycoprotein of halobacteria [74, 75])
`and cat-Man (Serg'Thr in audystroglycan which can be
`extended to the Siaa2-3Galfi1-461cNAcfil-ZMan se-
`quence and in yeast glycoproteins) [17, 29. 41. 16-78].
`The O-mannosylation in mammals has originally been
`described for derivatives of a GlcNAcfi 1-3Man core
`structure in the chondroitin sulfate proteoglycan of rat
`brain [79. 80]. A consensus sequence for yeast and
`mammalian O-mannosylation has not yet been delin-
`eated, if it is possible at all. Yeast cells employ dolichol—
`phosphate-fi-Man as donor for the first transfer to the
`protein acceptor in the endoplasmic reticulum by the
`doliehol-phosphate-Manzprotein mannosyltransferase
`[77, 31].
`In this context, an even more exceptional analogy to
`members of a different urkingdom can be drawn for a
`
`
`
`Figure B. Substitution pattern at neuraminic acid. The whole
`group of possible deriva