Single-chain antibody-fragment M6P-1 possesses a mannose 6-phosphate monosaccharide-specific binding pocket that distinguishes N-glycan phosphorylation in a branch-specific manner

Article abstract of DOI:10.1093/glycob/cwv093

The acquisition of mannose 6-phosphate (Man6P) on N-linked glycans of lysosomal enzymes is a structural requirement for their transport from the Golgi apparatus to lysosomes mediated by the mannose 6-phosphate receptors, 300 kDa cation-independent mannose 6-phosphate receptor (MPR300) and 46 kDa cation-dependent mannose 6-phosphate receptor (MPR46). Here we report that the single-chain variable domain (scFv) M6P-1 is a unique antibody fragment with specificity for Man6P monosaccharide that, through an array-screening approach against a number of phosphorylated N-glycans, is shown to bind mono- and diphosphorylated Man₆ and Man₇ glycans that contain terminal αMan6P(1 → 2)αMan(1 → 3)αMan. In contrast to MPR300, scFv M6P-1 does not bind phosphodiesters, monophosphorylated Man₈ or mono- or diphosphorylated Man₉ structures. Single crystal X-ray diffraction analysis to 2.7 ? resolution of Fv M6P-1 in complex with Man6P reveals that specificity and affinity is achieved via multiple hydrogen bonds to the mannose ring and two salt bridges to the phosphate moiety. In common with both MPRs, loss of binding was observed for scFv M6P-1 at pH values below the second pK_a of Man6P (pK_a = 6.1). The structures of Fv M6P-1 and the MPRs suggest that the change of the ionization state of Man6P is the main driving force for the loss of binding at acidic lysosomal pH (e.g. lysosome pH ～ 4.6), which provides justification for the evolution of a lysosomal enzyme transport pathway based on Man6P recognition.

Full text of DOI:10.1093/glycob/cwv093

Glycobiology, 2016, vol. 26, no. 2, 181–192
doi: 10.1093/glycob/cwv093
Advance Access Publication Date: 26 October 2015
Original Article
Glycan Recognition
Single-chain antibody-fragment M6P-1 possesses
a mannose 6-phosphate monosaccharide-specific
binding pocket that distinguishes N-glycan
phosphorylation in a branch-specific manner†
Ryan J Blackler2,‡, Dylan W Evans2,‡, David F Smith3,
Richard D Cummings3, Cory L Brooks2, Thomas Braulke4, Xinyu Liu5,
Stephen V Evans1,2, and Sven Müller-Loennies1,6
2Department of Biochemistry and Microbiology, University of Victoria, Victoria, BC Canada V8P 3P6, 3Department of
Biochemistry, National Center for Functional Glycomics, Emory University School of Medicine, Atlanta, GA 30322,
4
USA, Department of Biochemistry, Children’s Hospital, University Medical Center Hamburg-Eppendorf, D-20246
5
Hamburg, Germany, Department of Chemistry, University of Pittsburgh, 219 Parkman Avenue, 507-CSC, Pittsburgh,
PA 15260, USA, and 6Research Center Borstel, Leibniz-Center for Medicine and Biosciences, Parkallee 22, D-23845
Borstel, Germany
1To whom correspondence should be addressed: Tel: +49-4537-188-4700; Fax: +49-4537-188-4190; e-mail: sml@fz-borstel.de
(S.M.-L.); Tel: +1-250-472-4548; Fax: +1-250-721-8855; e-mail: svevans@uvic.ca (S.V.E.)
†The atomic coordinates and structure factors (pdb code 4RZC) have been deposited in the Protein Data Bank, Research
Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/).
‡These authors contributed equally to this work.
Received 13 May 2015; Revised 15 October 2015; Accepted 18 October 2015
Abstract
The acquisition of mannose 6-phosphate (Man6P) on N-linked glycans of lysosomal enzymes is a
structural requirement for their transport from the Golgi apparatus to lysosomes mediated by the
mannose 6-phosphate receptors, 300 kDa cation-independent mannose 6-phosphate receptor
(MPR300) and 46 kDa cation-dependent mannose 6-phosphate receptor (MPR46). Here we report
that the single-chain variable domain (scFv) M6P-1 is a unique antibody fragment with specificity
for Man6P monosaccharide that, through an array-screening approach against a number of phos-
phorylated N-glycans, is shown to bind mono- and diphosphorylated Man6 and Man7 glycans that
contain terminal αMan6P(1 → 2)αMan(1 → 3)αMan. In contrast to MPR300, scFv M6P-1 does not bind
phosphodiesters, monophosphorylated Man8 or mono- or diphosphorylated Man9 structures. Single
crystal X-ray diffraction analysis to 2.7 Å resolution of Fv M6P-1 in complex with Man6P reveals that
specificity and affinity is achieved via multiple hydrogen bonds to the mannose ring and two salt
bridges to the phosphate moiety. In common with both MPRs, loss of binding was observed for
scFv M6P-1 at pH values below the second pKa of Man6P (pKa = 6.1). The structures of Fv M6P-1
and the MPRs suggest that the change of the ionization state of Man6P is the main driving force
for the loss of binding at acidic lysosomal pH (e.g. lysosome pH ∼ 4.6), which provides justification
for the evolution of a lysosomal enzyme transport pathway based on Man6P recognition.
Key words: crystal structure, glycan array, lysosome, N-glycosylation, scFv
© The Author 2015. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
181

182
R J Blackler et al.
phosphomannan linked to bovine serum albumin (BSA) was investi-
Introduction
gated by ELISA at pH 5.0, 6.0, 6.4 and 7.4 (Figure 2), as Man6P
High mannose-type Asn-linked oligosaccharides (N-glycans) terminat-
ing in mannose 6-phosphate (Man6P) function as recognition markers
for the transport of lysosomal hydrolases from the Golgi to lysosomes
(Luzio et al. 2007; Braulke and Bonifacino 2009). Man6P is recognized
by the 300 kDa cation-independent mannose 6-phosphate receptor
(MPR300) and 46 kDa cation-dependent mannose 6-phosphate recep-
tor (MPR46), which are P-type lectins (Castonguay et al. 2011) that me-
diate this transport. The MPR300 at the plasma membrane also
internalizes extracellular Man6P-containing proteins.
2
has a pKa ¼ 6:1 (Hassid and Ballou 1957; Gracy and Noltmann
1968). PMP is a mixture of glycans consisting of α1 → 3-linked Man
residues containing Man6P at the non-reducing terminus and a single
α1 → 2-linked Man at the reducing end (Parolis et al. 1998; Ferro et al.
2002). The optimum binding was observed at pH 7.4 and 6.4, and it
was about 2-fold reduced at pH 6.0. The affinity dramatically de-
creased with reduction in pH by three orders of magnitude and was es-
sentially lost at pH 5.0. In an ELISA inhibition assay D-mannose
6-sulfate at 100 mM was unable to inhibit the binding of scFv
M6P-1 to PMP-BSA.
Inherited defects in the phosphotransferase complex (UDP-N-
acetylglucosamine-1-phosphotransferase), the key enzyme in the for-
mation of Man6P, cause reduced levels or the complete absence of
Man6P residues, leading to the secretion of the lysosomal
hydrolases and, as a consequence, dysfunctional lysosomes. The asso-
ciated lysosomal storage disorders (LSDs) manifest as mucolipidosis
type II (MLII also known as I-cell disease) with severe skeletal ab-
normalities, hepato-splenomegaly, progressive psychomotor retard-
ation and early death, or MLIII (pseudo-Hurler polydystrophy,
MLIII) with clinically milder symptoms (Braulke et al. 2013). No
treatment is available for MLII or MLIII, whereas some LSDs
caused by mutations in genes encoding single lysosomal enzymes
are treatable by recombinant enzyme replacement therapy based
in most cases on Man6P-dependent uptake mechanisms for proper
lysosomal delivery in cells and tissues (Ortolano et al. 2014).
Thus, there is considerable interest in the structural determin-
ation and quantification of N-glycans containing Man6P, as well
as their positions in lysosomal glycoproteins and the structural
characterization of their specific interaction with MPRs. Several
studies of the Man6P-binding domains of MPR46 and MPR300
in complex with different ligands have revealed structural insights
into the binding of Man6P and provided explanations for the ob-
served pH optimum of binding and distinct trafficking functions
in the cellular environment (Dahms et al. 2008; Olson et al. 2008,
2015). Recently, we have isolated a rabbit antibody single-chain
variable fragment (scFv) designated M6P-1 by phage display
(Figure 1); M6P-1 is specific for Man6P residues and displays low
micromolar affinity similar to both MPRs (Müller-Loennies et al.
2010). The scFv M6P-1 was shown to bind to glycoproteins con-
taining Man6P, allowing the direct identification and quantification
of Man6P residues on proteins (Schröder et al. 2010; Madhavarao
et al. 2014) and the indirect determination of GlcNAc-1-
phosphotransferase or Man6P phosphatase activities (Pohl, Tiede,
et al. 2010; Makrypidi et al. 2012). Further, these studies have de-
monstrated that scFv M6P-1 has the potential to be a powerful add-
itional tool for the fast and economical diagnosis of MLII and MLIII
by simple western blotting (Müller-Loennies et al. 2010; Pohl,
Encarnacão, et al. 2010).
Specificity of scFv M6P-1 towards high mannose-type
N-glycans
The specificity of scFv M6P-1 towards high mannose-type N-glycans
was determined by interrogating a phosphorylated glycan array (Song
et al. 2009; Castonguay et al. 2012) with a biotinylated scFv M6P-1
(Figure 3). As positive controls, we measured binding of a soluble form
of the MPR300, which was purified from fetal bovine serum by
phosphomannan affinity chromatography as described previously
(Tong and Kornfeld 1989), and binding by concanavalin A (Con A,
a mannose-specific lectin). The glycans were printed on the array
and reactivity with Con A at 0.5 µg/mL indicated that all glycans on
the array were accessible to binding and present in sufficient amounts.
The differential binding observed by Con A is either due to differences
in the structures that impact Con A binding or slight differences in the
concentration of printed glycan. At higher Con A concentrations
(10 µg/mL) binding of the glycans is similar (Castonguay et al.
To better understand the specificity of scFv M6P-1 towards phos-
phorylated high mannose-type N-glycans and to explore its
anti-carbohydrate specificity in comparison with MPR P-type lectins,
we have determined its reactivity with a large number of high
mannose-type N-glycans and used single crystal X-ray diffraction to
determine the structure of the Fv in complex with Man6P.
Fig. 1. Primary structure of the variable domains of the light (VL) and heavy
chain (VH) scFv M6P-1 obtained by phage display from an immunized rabbit.
The sequence determination has been published (Müller-Loennies et al.
2010) and residues were numbered using a world wide web interface (http://
www.bioinf.org.uk/abs/) applying the Kabat convention (Martin 1996). Amino
acids involved in interactions with the ligand in the crystal structure are shown
with increased font size. This figure is available in black and white in print and in
color at Glycobiology online.
Results
Dependence of scFv M6P-1 binding on pH
The binding of scFv M6P-1 to an immobilized neoglycoconjugate
of pentamannose phosphate (PMP) from Hansenula holstii

Man6P specificity and structure of scFv M6P-1
183
The assumption was made that the space group was incorrect and
the data were examined for evidence of twinning. The data were re-
indexed into a primitive orthorhombic Bravais lattice with unit cell
axes a = 60.6 Å, b = 128.05 Å and c = 127.30 Å. Analysis of data
by PHENIX.xtriage (Adams et al. 2010) revealed a significant off-
origin peak at coordinates 0.000, 0.473, 0.473, with a height
28.44% of the origin. Twin law tests revealed pseudo-merohedral
twinning with the operator –h, l, k and an estimated twinned fraction
of 0.419 (Britton analyses), 0.437 (H-test) or 0.435 (maximum like-
lihood method). The structure was then solved by molecular replace-
ment in space group P212121 with four Fv molecules in the AU.
Multiple rounds of model building and refinement yielded a final
model with Rwork = 0.228 and Rfree = 0.269 and a final refined twin
fraction of 0.480.
Structure of Fv M6P-1 in complex with Man6P
Fig. 2. Binding of scFv M6P-1 to immobilized PMP-BSA at different pH in ELISA.
ScFv M6P-1 (Mw 28,228 Da) dissolved in the indicated buffers was titrated
starting from 5 µg/mL (177 nM) on PMP-BSA neoglycoconjugate immobilized
on polystyrene ELISA plates (85 ng/cup ≈ 24 pmol of PMP). The data were fitted
by non-linear regression to a logistic function using Origin v. 7.0 SR4
(OriginLab Corp., Northampton). Displayed are the duplicate data points and
the fitted curve. This figure is available in black and white in print and in
color at Glycobiology online.
The data collection and refinement statistics for Fv M6P-1 with
Man6P are given in Table I. The four crystallographically independ-
ent Fv molecules each contain a heavy and light chain, labeled H/L,
A/B, C/D and E/F. Excellent electron density was observed for the
Man6P monosaccharide in each of the four Fv molecules in the
AU (Figure 4A) and for the polypeptide chain generally, although
each light chain has regions of weak or interrupted electron density
around residues 7–10, 14–16 and 93–95B, opposite the combining
site (for numbering of residues see Figure 1). Residues were num-
bered using a world wide web interface (http://www.bioinf.org.uk/
abs/) (Martin 1996) applying the Kabat convention, with heavy
and light chain residues designated with an H or L, respectively,
and with amino acid insertions designated with capital letters to
keep conserved residues properly aligned. Superposition of the four
Fv in the AU showed that the complementarity determining regions
(CDR) of all four molecules in the AU adopt the same conformations
(Figure 4B), with the exception that in two of the four light chains,
SerL94 (chain L) or AsnL95A (chain B) of CDR L3 is not modeled
due to ambiguous electron density. Each of the four Fv in the AU
displays electron density consistent with a buffer-derived sulfate
ion linked by hydrogen bonds and a salt bridge to GlnL42 and
ArgL43, well outside the combining site. Two of the four Fv display
electron density for an additional sulfate ion coordinated to
ArgL95C in the combining site (chains D and F), which also main-
tain their salt bridge interactions to Man6P. The two Fv with this sul-
fate ion present have somewhat better electron density for residues
L93–L95B than the two without.
Fv M6P-1 makes four hydrogen bond contacts to the mannose
ring of Man6P through CDR H1 and H3 (Figure 5A). The side
chain of ThrH33 forms a hydrogen bond with the Man6P ring oxy-
gen, the backbone carbonyl of AspH95 forms a hydrogen bond to
OH-2, and the side chain of AspH95 and the backbone amide of
AsnH98 form hydrogen bonds to OH-3. There is an additional
hydrogen bond seen between ThrH33 main chain N and OH-2,
where the distance is observed to lie between 2.99 and 3.45 Å de-
pending on the AU. Fv M6P-1 binds to the phosphate group on
Man6P through CDR L1 and CDR L3 where both ArgL32 and
ArgL95C form salt bridges. There are small variations in the hydro-
gen bond distances between the four molecules in the AU due to
modest differences in refined orientations of protein side chains
and the ligand.
2012). The binding profile of the MPR300 revealed binding to all
phosphorylated glycans, including phosphodiesters and are therefore
in agreement with previous results (Song et al. 2009; Castonguay et al.
2012). Taken together, these control studies demonstrate the validity
of the phosphorylated glycan array.
The glycan array analysis (Figure 3) using the biotinylated scFv
M6P-1 revealed that it bound specifically mono- and diphosphorylated
Man6 (PM6 and 2(P)M6, respectively) as well as diphosphorylated
Man7 glycans 2(P)M7(1) and 2(P)M7(3) which all contained αMan6P
(1 → 2)αMan(1 → 3)αMan, i.e. Man6P in the A-arm. The strongest sig-
nal was obtained for PM6. Glycans lacking phosphate and those con-
taining αGlcNAc-1-P-6-αMan phosphodiesters were not bound by the
scFv. High mannose-type N-glycans that contained Man6P in the
C-arm, i.e. αMan6P(1 → 2)αMan(1 → 6)αMan in structures PM8(1),
and PM9 were also not bound by scFv M6P-1. Remarkably, this was
also the case for 2(P)M9 despite the presence of Man6P on the A-arm.
Crystal structure analysis: ambiguity in space group
assignment and pseudo-merohedral twinning of Fv
M6P-1 crystals
The Fv M6P-1 was generated by cleavage of the scFv linker, and sub-
sequently crystallized and soaked with Man6P. Diffraction data were
collected to 2.7 Å resolution, and initially indexed, integrated and
scaled in a primitive tetragonal Bravais lattice. Systematic absences in-
dicated space group P41212 or P43212. Cell content analysis predicted
two molecules in the asymmetric unit (AU); however, molecular re-
placement (using the coordinates of the variable domain dimer of
the only rabbit antigen binding fragment of an antibody (Fab) then de-
posited in the Protein Data Bank (Berman et al. 2000) (http://www.
pdb.org), 4HBC, as a search model initially located only one molecule
of the asymmetric unit. An extensively refined partial solution used as
a model eventually yielded a full solution in P41212, but refinement
stalled with R factors >40% and electron density that did not match
entire lengths of the polypeptide chain.
The soaked ligand Man6P is capable of mutarotation, but in each
case the electron density of the Man6P molecule is consistent with the
α-anomeric form. Superposition of β-mannose in the combining site

184
R J Blackler et al.
Fig. 3. Binding of scFv M6P-1 to glycans immobilized in an array on glass slides. The glycan microarray used in this analysis contained the schematically depicted
glycans that were printed in replicates of four at a concentration of 100 µM with the exception of glycans 3, 7, 11, 12, 15, 18 and 21, which were only available at lower
concentrations (<100 µM). Below a structure code is given where G = GlcNAc (blue squares), P = phosphate (light yellow circles) and M = Man (green circles).

Man6P specificity and structure of scFv M6P-1
185
Table I. Data collection and refinement statistics for liganded Fv
M6P-1
and natural lysosomal glycoproteins with antibodies. Yet, the crystal
structure analysis revealed that the lack of specificity for the glycosidic
linkage can be explained by a small pocket-type combining site that
specifically recognizes only the terminal carbohydrate residue. This
is in contrast to almost all other known antibodies against carbohy-
drates where, despite the dominance of a single residue providing
most of the binding energy, multiple residues of the carbohydrate anti-
gen are in contact with the antibody. From early studies, it was esti-
mated that the smallest antibody combining site can accommodate
two to three monosaccharide residues (Arakatsu et al. 1966; Bundle
et al. 1994). Thus, scFv M6P-1 is one of very few antibodies reported
to exhibit high affinity (Kd = 28 µM) (Müller-Loennies et al. 2010) for
a single monosaccharide residue.
Two different mechanisms for the antibody recognition of mono-
saccharides have been observed. First, there exist a number of crystal
structures of antibodies specific for various chlamydial oligosacchar-
ide antigens in complex with a 3-deoxy-α-D-manno-oct-2-ulosonic
acid (Kdo) monosaccharide (Brooks et al. 2008; Brooks, Blackler,
et al. 2010; Brooks, Müller-Loennies et al. 2010; Blackler et al.
2012), which have affinities for the monosaccharide in the range of
3–90 μM. In these cases, monosaccharide binding is achieved through
a pocket of conserved sequence containing specific hydrogen bonds
and a strong salt bridge to a sugar carboxylate.
Fv M6P-1 with Man6P
Resolution (Å)
Space group
a (Å)
30.1–2.72(2.79–2.72)a
P212121
60.6
b (Å)
128.1
c (Å)
127.3
No. of molecules per asymmetric unit
Rbsym
I/σ(I)
Completeness (%)
Redundancy
Unique reflections
Refinement
Rwork (%)
4
0.105 (0.392)
7.6 (2.4)
81.2 (86.3)
2.5 (2.5)
20805
22.6
Rfree (%)
26.7
No. of waters
r.m.s bonds (Å)
r.m.s. angles (°)
Mean B (Å2)
Wilson B (Å2)
Ramachandran
Favored
153
0.0090
1.489
30.4
33.0
94%
5%
Second, and in contrast, antibody S-20-4 has been determined to
achieve an affinity of 39 μM for a terminal monosaccharide of the
Vibrio cholerae Ogawa O-specific polysaccharide [oligomer of α1 →
2-linked 4-amino-4,6-dideoxy-D-mannose (perosamine) N-acylated
with 3-deoxy-L-glycero-tetronic acid] through a significant hydropho-
bic effect (Wang et al. 1998; Villeneuve et al. 2000).
Allowed
Outliers
1%
aValues in parentheses represent the highest resolution shell.
bRsym = ΣhklΣi|Ii(hkl) − 〈I(hkl)〉|/ΣhklΣiIi(hkl), where Ii(hkl) is the intensity of
each ith redundant observation of reflection hkl, and 〈I(hkl)〉 is the average
intensity of a given reflection hkl.
ScFv M6P-1 achieves specificity for mannose
6-phosphate through targeted hydrogen bonds and
charged-residue interactions
reveals steric clashes of OH-1 with the backbone of residues
TyrH32 and ThrH33.
Fv M6P-1 is a unique anti-carbohydrate antibody in its origin from
rabbit. Rabbit antibodies generally have longer and more diverse CDR
L3s compared with humans and mice (Lavinder et al. 2014), and Fv
M6P-1 follows this trend with a 12-residue CDR L3 [determined with
IMGT/V-QUEST (Brochet et al. 2008)]. This does not however appear
to be a critical factor in Man6P recognition as CDR L3 is backward
leaning (Figure 4B) and makes only one direct interaction from
ArgL95C to Man6P (Figure 5A). Fv M6P-1 displays no otherwise un-
usual features compared with other anti-carbohydrate antibodies
(Haji-Ghassemi et al. 2015).
ScFv M6P-1 achieves its affinity and specificity through several hydro-
gen bonds and important salt bridge interactions to the small carbo-
hydrate antigen. ScFv M6P-1 was previously determined not to bind
mannose, glucose or glucose 6-phosphate (Glc6P) (Müller-Loennies
et al. 2010), and binding to PMP-BSA was not inhibited by mannose
6-sulfate (this study). The crystal structure presents a clear rationale
for specific recognition of Man6P by scFv M6P-1, but not mannose,
as the scFv makes two salt bridges to the phosphate group from
ArgL32 and ArgL95C (Figure 5A). In this sense, M6P-1 is similar to
Kdo-binding antibodies that also utilize a salt bridge interaction
(Brooks et al. 2008; Brooks, Blackler et al. 2010; Brooks, Müller-
Loennies et al. 2010; Blackler et al. 2012).
The lack of measurable affinity for Glc6P can be traced to the
hydrogen bonds formed by the axial O2 of Man6P to AspH95 O
and ThrH33 N (depending on which molecule in the AU) that could
not be formed by the equatorial O2 on Glc. The lack of binding to
Man6S (with an ionization state of −1) indicates that strong
Discussion
ScFv M6P-1 recognizes a single carbohydrate residue
The Man6P-specific scFv M6P-1 was obtained by immunization with
PMP, implying a structural similarity between αMan6P(1 → 2)αMan
and αMan6P(1 → 3)αMan leading to cross-reaction of the yeast PMP
Glucose and Galactose in controls are depicted as blue and yellow circles, respectively. Numbers indicate the number of residues and isomers are distinguished by
numbers in brackets. N-Glycan residues are designated as schematically depicted in the lower panel following the nomenclature as published (Castonguay et al.
2012). The array was interrogated with biotinylated scFv M6P-1 and Con A and affinity purified fetal bovine serum MPR300. Bound scFv M6P-1 and Con A were
quantified after incubation with Cy5-labeled streptavidin and MPR300 was assayed using a rabbit polyclonal antibody and a Cy5-labeled anti-rabbit IgG as
described (Castonguay et al. 2012). Con A and MPR300 served as positive controls to validate the array. Relative fluorescence units were determined as the
average of four replicates, and the glycan numbers correspond to the glycans above. Biotinylated Con A at 0.5 µg/mL (left) was detected with Cy5 Streptavidin
(5 µg/mL); biotinylated scFv M6P-1 at 20 µg/mL (middle) detected with Cy5 Streptavidin (5 µg/mL); MPR300 at 5 µg/mL (right) detected with a rabbit polyclonal
antibody (1:250 dilution) and Cy5-labeled goat, anti-rabbit IgG at 5 µg/mL. Slides were scanned for Cy5 in a ProScaArray scanner (PerkinElmer) and average
fluorescence units were determined using the corresponding ProScanArray software.

186
R J Blackler et al.
Fig. 4. Crystal structure of scFv M6P-1 (A). Fo − Fc omit electron density maps contoured to 2σ for Man6P in each of the four Fv molecules in the AU. The refined
coordinates of the corresponding Man6P are superposed for clarity. (B) Superposition of M6P-1 binding sites of the four molecules in the asymmetric unit,
displaying α-carbon backbones of CDR loops and the Man6P molecule of each unit. Chains H/L in green, A/B white, C/D tan and E/F blue.
Fig. 5. Comparison of Man6P-binding sites of Fv M6P-1 and bovine MPR. Stereoviews of the binding sites of (A) Fv M6P-1 (light chain green, heavy chain tan), (B) the
CRD of MPR46 (pdb code 1M6P; Roberts et al. 1998) and (C) the CRD formed by the N-terminal domains 1–3 of MPR300 (pdb code 1SYO; Olson et al. 2004), showing
hydrogen bonds and salt bridges as dashed yellow spheres.

Man6P specificity and structure of scFv M6P-1
187
charged-residue interactions are imperative for binding, and supports
the ELISA data that showed that binding is lost at pH below the
second pKa of Man6P (pK = 6.1, Hassid and Ballou 1957; Gracy
and Noltmann 1968). The scFv M6P-1 was crystallized at a pH of
8.5, which would result in a charge of −2 for the phosphate moiety
thus allowing for strong salt bridges to be formed.
The observed lack of binding to phosphodiesters can be explained
by the loss of one negative charge on the phosphate because a charge
of −1 is insufficient for binding. In addition, the combining site of the
Fv is sterically restricted surrounding the phosphate (Figure 6A) and
could only accommodate a phosphodiester with the second sugar
directed upwards away from the binding site.
ScFv M6P-1 is specific for glycans containing terminal
αMan6P(1 → 2)αMan(1 → 3)αMan
In addition to its inability to bind non-phosphorylated or phospho-
diester glycans, scFv M6P-1 differentiates between phosphomonoester
glycans with specificity for those containing αMan6P(1 → 2)αMan
(1 → 3)αMan (Figure 3, structures 16–18). The lack of scFv M6P-1
binding to glycans PM8(1), PM9 and 2(P)M9 showed that it does
not cross-react with αMan6P(1 → 2)αMan(1 → 6)αMan and indi-
cated that glycans with two consecutive terminal α(1 → 2)-linkages
as in αMan6P(1 → 2)αMan(1 → 2)αMan(1 → 3)αMan are not
bound. Therefore, the observed cross-reaction between PMP and
N-glycans is due to the fact that the scFv is permissive for the change
from α(1 → 3) to α(1 → 2) only in the terminal linkage but not in the
penultimate position.
While Man6P monosaccharide is capable of mutarotation, each
of the Man6P molecules is observed clearly bound in the α-
configuration, which points the anomeric oxygen away from the bind-
ing pocket (Figure 6A). The CDR loops do not extend significantly
beyond the level of Man6P in this pocket nor do many side chains pro-
trude. With the exception of TrpH31, which could possibly stack
against additional mannose residues, there is little potential for steric
hindrance of glycans displaying a terminal αMan6P. However, when
αMan6P(1 → 3)αMan(1 → 3)αMan of PMP from the complex struc-
ture with MPR46 (PDB 1C39) is aligned with bound Man6P in the
M6P-1 structure, there are significant clashes with the second and
third mannose residues, with the third completely overlapping the pro-
tein. Yet, as M6P-1 was generated using PMP, alternate conforma-
tions of this ligand must exist.
Comparison of scFv M6P-1 and P-type lectin Man6P
recognition strategy
The MPR46 and MPR300 are well-studied P-type lectins that bind
high mannose-type glycans containing Man6P (reviewed in Dahms
et al. 2008) allowing a comparison of the antibody combining site
with the lectin carbohydrate recognition domains (CRDs). MPR46
forms stable homodimers with a single Man6P CRD per monomer
and binds almost exclusively phosphomonoesters (Distler et al.
1991; Olson et al. 2008; Song et al. 2009). In contrast, the
MPR300 contains four independent Man6P-binding sites and binds
strongly to almost all phosphomono- as well as phosphodiesters,
such as Man-P-GlcNAc (Song et al. 2009) (Figure 5). It contains
two high-affinity Man6P-binding sites and a low-affinity-binding
site that has been shown to bind Man-P-GlcNAc with higher affinity
than Man6P (Chavez et al. 2007; Olson et al. 2010).
Considering ligand flexibility, the shallow-binding site and the
dominance of the phosphate in the interaction, binding of Fv M6P-1
to αMan6P(1 → 2)αMan(1 → 2)αMan(1 → 3)αMan or αMan6P
(1 → 2)αMan(1 → 6)αMan appeared feasible, in principal. An alterna-
tive explanation for the lack of binding may therefore be a conform-
ational change induced in the antigen due to the presence of Man
(Figure 3, residue I) inducing steric crowding which renders not only
Man6P on the C-arm but also on the A-arm inaccessible due to back-
folding of the C-arm towards the core GlcNAc. Such conformational
rearrangements affecting the whole triantennary structure have been
predicted for higher N-glycans from molecular dynamics simulations
(Balaji et al. 1994; Qasba et al. 1994) and nuclear magnetic resonance
experiments (Vliegenthart et al. 1983; Kamiya et al. 2013). Neverthe-
less, the binding to PM8 and PM9 by MPR300 domain 9 (Bohnsack
et al. 2009) showed that the Man6P on the C-arm is in fact accessible
to protein binding. Therefore, the surface of M6P-1 adjacent the
Man6P pocket must clash with the PM9, 2(P)M9 and PM8(1) epi-
topes bound by MRP300.
ScFv M6P-1 will recognize several glycoproteins possessing
terminal Man6P residues, which facilitates the diagnosis of
MLII and MLIII by western blotting demonstrating the loss or sig-
nificantly lowered levels of several Man6P-containing proteins, re-
spectively, in extracts of cultured fibroblast of patients compared
with healthy individuals (Müller-Loennies et al. 2010; Pohl,
Encarnacão, et al. 2010). The lack of scFv M6P-1 binding to
higher Man8 or Man9 N-glycan structures that are phosphory-
lated on the C-arm will certainly lead to an underestimation of
the total Man6P content of glycoproteins based on western blot
analysis. However, structural analyses by mass spectrometry indi-
cate that the glycan structures containing <8 mannose residues are
by far the most abundant on lysosomal enzymes, e.g. arylsulfatase
A (Schröder et al. 2010).
The crystal structures of the bovine MPR46 in complex with pen-
tamannosyl phosphate (PMP, PDB code 1C39) (Olson et al. 1999) and
of the N-terminal domains 1–3 of MPR300 in complex with Man6P
(PDB code 1SZ0) (Olson et al. 2004) have revealed in both MPRs five
conserved interactions from Tyr, Gln, Arg, Tyr and Glu to the pyra-
nose ring (Figure 5). A total of 11 possible hydrogen bonds to the pyr-
anose ring lead to a high-affinity-binding site.
Surprisingly, the phosphate is bound by the high-affinity-binding
site in domains 1–3 of the MPR300 only by two hydrogen bonds
without any salt bridge interaction, yet it has been reported to ex-
hibit an affinity for Man6P similar to MPR46 (Tong et al. 1989;
Tong and Kornfeld 1989), which binds the phosphate through one
salt bridge (His105) and three hydrogen bonds (Asn103, Asn104)
and loses binding to Man6P upon His105 deprotonation (Olson
et al. 2008).
In contrast to the lectin MPR structures, the interaction between
the antibody Fv M6P-1 and Man6P is dominated by the phosphate
moiety, which forms mono- and bidentate salt bridges through two ar-
ginine side chains (ArgL32, ArgL95C). The combining site of Fv
M6P-1 displays only four hydrogen bonds to the pyranose ring of
Man6P. Two of these bonds are from the protein main chain atoms
in Fv M6P-1, whereas all hydrogen bonds in the MPRs to the pyranose
ring are from side chains.
The combining site of Fv M6P-1 is similar to the MPR46 binding
site with steric restriction about the phosphate moiety (Figure 6A and B),
consistent with its inability to accommodate phosphodiesters.
The MPR300-binding site formed by domains 1–3 is less restricted
than that of MPR46 (Figure 6C); however, a Man-P-GlcNAc
phosphodiester-binding site is also located in domain 5 (Bohnsack
et al. 2009). Determination of the solution NMR structure of domain
5 revealed a binding site less sterically crowded than MPR46 (Olson
et al. 2010).

188
R J Blackler et al.
Fig. 6. Binding surfaces of Fv M6P-1 and MPRs. Top-down and side-view surface representations of Fv M6P-1, MPR46 (PDB code 1C39) and MPR300 domains 1–3
(PDB code 1SZ0). In each side-view, the model is clipped through the binding pocket. The dashed line in the top-down view indicates the location of the clipping
plane. (A) Fv Man6P in complex with Man6P. Displayed above the side-view is a model of array glycan #16 for size comparison. (B) MPR46 in complex with PMP,
showing only αMan6P(1 → 3)αMan(1 → 3)αMan as observed in that structure. Manganese ion is displayed as a mauve sphere. (C) MPR300 domains 1–3, sequentially
colored yellow, orange, and mauve, in complex with Man6P.

Man6P specificity and structure of scFv M6P-1
189
(Müller-Loennies et al. 2010). Fv was produced from M6P-1 scFv
by incubation with subtilisin at a ratio of 1:300 (w/w) to digest the
flexible linker so as to prevent oligomerization. The digest was per-
formed at room temperature in 20 mM HEPES, pH 7.2, with
150 mM NaCl for 3 h and quenched by the addition of 1 mM
PMSF (G-biosciences). The Fv was further purified by size-exclusion
chromatography (Phenomenex BioSep SEC-s3000 column) using
HEPES, pH 7.2, with 150 mM NaCl as the mobile phase.
pH-dependent binding of scFv M6P-1 and MPRs stems
from ionization of phosphate
The scFv M6P-1, MPR46 and MPR300 all have similar observed af-
finities for Man6P with KD values in the range of 7–28 µM (Tong
and Kornfeld 1989; Müller-Loennies et al. 2010). The common
theme is the presence of charged residues making ion-dipole
and/or salt bridge interactions. Neither MPR binds Man6P at pH
<5, which has been explained in MPR46 by the disruption of elec-
trostatic inter-subunit interactions and the protonation of E133,
one of the critical Man6P-binding residues (Waheed et al. 1990;
Waheed and von Figura 1990; Olson et al. 2008). Interestingly,
the strongest binding of scFv M6P-1 to PMP was seen at pH 6.4
and 7.4, with ∼2-fold reduced binding at pH 6.0 and no observed
binding at pH 5.0, yet the binding site possesses no residues with
pKa in this range. It is therefore likely that the loss of affinity at
pH 5.0 is mainly driven by a change of the protonation state of the
Preparation of mannose-6-phosphate
Man6P was prepared from benzyl 2,3,4-tri-O-benzyl α-D-mannoside
(Lu et al. 2005). In brief, benzyl 2,3,4-tri-O-benzyl β-D-mannoside
was phosphorylated with dibenzyl N,N-diethylphosphoramidite and
treated with mCPBA to give the phosphate triester. Hydrogenolysis
of the resulting benzyl 2,3,4-tri-O-benzyl-1-O-dibenzylphosphoryl
β-D-mannoside over Pd/C and purification using ion exchange resin
(Dowex® Na+ form) gave the disodium salt of Man6P.
2
phosphate moiety of Man6P ( pKa ¼ 6:1, Hassid and Ballou 1957;
Gracy and Noltmann 1968).
While the effect of phosphate ionization on the loss of Man6P
binding by MPR46 at acidic pH was not discussed for the release of
cargo into the endosomal compartment (Olson et al. 2008), it is likely
a critical factor in pH-dependent binding (Tong and Kornfeld 1989)
given the extensive charge coordination in the binding site (Figure 5B).
Although there are no salt bridges to Man6P observed in the structure
of MPR300 domains 1–3, there are potential long-range charge inter-
actions from Arg391, Lys358 and Lys125 (between 4.8 and 6.9 Å)
that may be weakened by Man6P protonation. The sequence align-
ment of MPR300 domains 3, 5 and 9 with MPR46 reveals histidine
and lysine residues of MPR300 domain 9 in a similar position to
His105 of MPR46 that forms a salt bridge to Man6P (Reddy et al.
2004). Therefore, Man6P protonation may also explain the mechan-
ism by which the MPR300 releases its cargo into the acidic lumen of
endosomes, for which structural changes of the protein have not been
observed, and provides an explanation why a system based on Man6P
recognition has evolved for shuttling glycoproteins to the lysosome.
Biotinylation and ELISA-binding assays
The scFv M6P-1 was biotinylated using sulfo-NHS-LC-biotin (Ther-
mo Fisher Scientific, Bonn, Germany). The reagent was dissolved to
10 mM in water and 0.05 mL added to 0.7 mL of purified scFv
(1.0 mg/mL) in PBS, pH 7.4. After 2 h on ice excess reagent was re-
moved by dialysis against PBS. The activity of the biotinylated scFv
was verified by ELISA against PMP-BSA using a secondary antibody
(ascites from mAb 9E10) against the c-myc tag as described (Müller-
Loennies et al. 2010) and an HRP-conjugate of streptavidin (Dianova,
Germany).
For the ELISA-binding assays at different pH, the scFv (1 mg/mL in
PBS, pH 7.4) was diluted to the starting concentration of 0.005 mg/
mL in citrate-phosphate buffer containing 150 mM NaCl with a pH
adjusted to 5.0, 6.0 or 6.4. The binding curves were determined as de-
scribed (Müller-Loennies et al. 2010) in these buffers and compared
with the binding curve of scFv in PBS, pH 7.4.
ELISA inhibition was performed as described (Müller-Loennies
et al. 2010) using mannose 6-sulfate (sodium salt, Dextra), Man6P
and Man at starting concentrations of 100 mM.
Conclusions
The Fv M6P-1 is shown to distinguish N-glycan phosphorylation in
a branch-specific manner, with specificity for A-branch terminal
αMan6P(1 → 2)αMan(1 → 3)αMan present on glycans PM6, 2(P)
M6, 2(P)M7(1) and 2(P)M7(3). This specificity and the predominance
of glycans with fewer than 8 Man residues on lysosomal enzymes sup-
port the potential utility of scFv M6P-1 in the diagnosis of MLII and
MLIII or in protein purification for enzyme replacement therapy. The
crystal structure of Fv M6P-1 in complex with Man6P reveals that the
basis for its ability to react with terminal αMan6P is a shallow
monosaccharide-binding pocket with multiple hydrogen bonds to
the pyranose ring and two salt bridges to the phosphate moiety. The
loss of Man6P binding by scFv M6P-1 <pH 6.0 is found to be due to
Man6P protonation and the associated weakening of salt bridge inter-
actions, and a similar effect occurring in the MPRs with Man6P
protonation rationalizes the evolution of the Man6P-based lysosomal
enzyme transport pathway.
Glycan microarray analysis
The phosphorylated glycan array was produced and the analyses were
performed as previously described, and the data were reported as his-
tograms of average Relative Fluorescent Units corresponding to glycan
structures identified by chart ID (Song et al. 2009; Castonguay et al.
2012). The carbohydrate structures are depicted in Figure 3. As nega-
tive controls lacto-N-neotetraose (#22), NA2 (#23), PBS (#24 and 25)
and biotin (#26) were included. The positive controls were carried out
with biotinylated concanavalin A (Vector Labs) at 0.5 µg/mL detected
with Cy5-labeled Streptavidin (Invitrogen) and a soluble form of the
MPR300 at 5 µg/mL detected with a rabbit anti-MPR300 polyclonal
antibody (B14.5 at 1:250 dilution) and Cy5-labeled anti-rabbit IgG
(Invitrogen) at 5 µg/mL. The MPR and B14.5 antibody were kindly
provided by Nancy Dahms (Medical College of Wisconsin, Milwau-
kee, WI). The biotinylated scFv M6P-1 was analyzed on the CFG gly-
can microarray and those data, which are available on the CFG
website (http://www.functionalglycomics.org), indicated that it
bound only mannose-6-phosphate among the 610 glycans printed
on v5.1 of the CFG array. The biotinylated scFv M6P-1 was applied
to the phosphorylated glycan array at 200, 20 and 2 µg/mL in buffer
containing EDTA and detected with Cy5-labeled Streptavidin.
Materials and methods
Preparation of Fv M6P-1
The generation of scFv M6P-1 by immunization and phage display
and its expression and purification have been described previously

190
R J Blackler et al.
Glycan Interaction Core (Emory University School of Medicine, Atlanta, GA)
that assisted with analysis of samples on the array.
Crystallization of Fv M6P-1
Purified Fv was concentrated to 14 mg/mL using Amicon Ultra centri-
fugal filters. Crystallization screens were set up using an Art Robbins
Instruments Gryphon robot. Small needle-like crystals appeared in
Hampton Index screen (Hampton Research, Aliso Viejo, CA) condition
#6 (0.1 M Tris, pH 8.5, with 2 M ammonium sulfate) after 4 days.
Conflict of interest statement
S.M.-L. and T.B. hold a patent on the application of scFv M6P-1.
Ligand soaking, data collection, structure determination
and refinement
Abbreviations
AU, asymmetric unit; BSA, bovine serum albumin; CDR, complementarity de-
termining region; CRD, carbohydrate recognition domain; Fab, antigen binding
fragment of an antibody; Fv, variable fragment of an antibody (scFv-fragment
after enzymatic cleavage of the linker); Glc, glucose; Glc6P, glucose 6-phosphate;
GlcNAc-1-phosphotransferase, UDP-N-acetylglucosamine:lysosomal enzy-
me N-acetylglucosamine-1-phosphotransferase; Man, mannose; Man6P, man-
nose 6-phosphate; ML, mucolipidosis; MPR46, 46 kDa cation-dependent
mannose 6-phosphate receptor; MPR300, 300 kDa cation-independent man-
nose 6-phosphate receptor; PMP, pentamannose phosphate from Hansenula
holstii; scFv, single-chain fragment variable; VH, variable domain of the anti-
body heavy chain; VL, variable domain of the antibody light chain.
Crystals were soaked in mother liquor with 2 mM Man6P and flash
frozen to −160°C with an Oxford Cryostream 700 crystal cooler
(Oxford Cryosystems, Devens, MA) using CryoOil (MiTeGen, Ithaca,
NY) as cryoprotectant. Data were collected on beamline CMCF-ID at
the Canadian Light Source synchrotron (Saskatoon, SK) and pro-
cessed using HKL2000 (HKL Research, Charlottesville, VA). Data
quality were analyzed using PHENIX.xtriage (Adams et al. 2010).
The structure of liganded Fv M6P-1 was solved by molecular replace-
ment using PHASER (McCoy et al. 2007) with the variable domain of
the antibody heavy chain (VH) and variable domain of the antibody
light chain (VL) of a rabbit Fab (PDB code 4HBC) as a model. Manual
fitting of Fo − Fc and 2Fo − Fc electron density maps was carried out
with Coot (Emsley et al. 2010) and SetoRibbon (S. V. Evans, unpub-
lished). TLS and restrained amplitude-based twin refinement allowing
isotropic thermal motion was carried out with REFMAC5 as imple-
mented in CCP4 (Winn et al. 2011). The final refinement and model
statistics are given in Table I.
References
Adams PD, Afonine PV, Bunkoczi G, Chen VB, Davis IW, Echols N, Headd JJ,
Hung LW, Kapral GJ, Grosse-Kunstleve RW, et al. 2010. PHENIX: A com-
prehensive Python-based system for macromolecular structure solution.
Acta Crystallogr D Biol Crystallogr. 66:213–221.
Arakatsu Y, Ashwell G, Kabat EA. 1966. Immunochemical studies on
dextrans. V. Specificity and cross-reactivity with dextrans of the antibodies
formed in rabbits to isomaltonic and isomaltotrionic acids coupled to bo-
vine serum albumin. J Immunol. 97:858–866.
Visualization and graphics
Crystal structure figures were produced using SetoRibbon (S. V.
Evans, unpublished) and UCSF Chimera (Pettersen et al. 2004). Chi-
mera is developed by the Resource for Biocomputing, Visualization
and Informatics at the University of California, San Francisco (sup-
ported by NIGMS P41-GM103311).
Balaji PV, Qasba PK, Rao VS. 1994. Molecular dynamics simulations of high-
mannose oligosaccharides. Glycobiology. 4:497–515.
Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H,
Shindyalov IN, Bourne PE. 2000. The Protein Data Bank. Nucleic Acids
Res. 28:235–242.
Blackler RJ, Müller-Loennies S, Brade L, Kosma P, Brade H, Evans SV. 2012.
Antibody recognition of chlamydia LPS: Structural insights of inherited im-
mune responses. In: Kosma P, Müller-Loennies S, editors. Anticarbohydrate
Antibodies. Vienna: Springer. p. 75–120.
Bohnsack RN, Song X, Olson LJ, Kudo M, Gotschall RR, Canfield WM,
Cummings RD, Smith DF, Dahms NM. 2009. Cation-independent man-
nose 6-phosphate receptor: A composite of distinct phosphomannosyl bind-
ing sites. J Biol Chem. 284:35215–35226.
Sequence analysis
The sequence of Fv M6P-1 was analyzed with UCL AbCheck and the
structure was numbered according to the Kabat method (http://www.
bioinf.org.uk/abs/; Martin 1996).
Braulke T, Bonifacino JS. 2009. Sorting of lysosomal proteins. Biochim Biophys
Funding
Acta. 1793:605–614.
This work was supported in part by grants from the Natural Re-
sources and Engineering Research Council of Canada and the Canad-
ian Institute for Health Research to S.V.E . S.V.E. is a recipient of a
senior scholarship from the Michael Smith Foundation for Health
Research. R.J.B. is a recipient of an NSERC CGSD Award. X.L. thanks
support from NIH (P50CA121973). A Grant P41GM103694
(R.D.C.) to the National Center for Functional Glycomics also sup-
ported work.
Braulke T, Raas-Rothschild A, Kornfeld S. 2013. I-cell disease and pseudo-
Hurler polydystrophy: Disorders of lysosomal enzyme phosphorylation
and localization. In: Valle D, Vogelstein B, Kinzler KW, Antonarakis SE,
Ballabio A, Scriver CR, Sly WS, Bunz F, Gibson KM, Mitchell G, editors.
The Online Metabolic and Molecular Bases of Inherited Disease.
New York: McGraw-Hill.
Brochet X, Lefranc MP, Giudicelli V. 2008. IMGT/V-QUEST: The highly cus-
tomized and integrated system for IG and TR standardized V-J and V-D-J
sequence analysis. Nucleic Acids Res. 36:W503–W508.
Brooks CL, Blackler RJ, Sixta G, Kosma P, Müller-Loennies S, Brade L,
Hirama T, MacKenzie CR, Brade H, Evans SV. 2010. The role of CDR
H3 in antibody recognition of a synthetic analog of a lipopolysaccharide
antigen. Glycobiology. 20:138–147.
Brooks CL, Müller-Loennies S, Borisova SN, Brade L, Kosma P, Hirama T,
Mackenzie CR, Brade H, Evans SV. 2010. Antibodies raised against chla-
mydial lipopolysaccharide antigens reveal convergence in germline gene
usage and differential epitope recognition. Biochemistry. 49:570–581.
Brooks CL, Müller-Loennies S, Brade L, Kosma P, Hirama T, MacKenzie CR,
Brade H, Evans SV. 2008. Exploration of specificity in germline monoclonal
Acknowledgments
We thank Nadine Harmel, Christine Schneider and Ute Agge for their technical
assistance. We also thank Dr. C.R. MacKenzie from the Institute of Biological
Sciences at the National Research Council of Canada (Ottawa) for the kind gift
of mAb 9E10 ascites. The authors acknowledge The Consortium for Functional
Glycomics funded by the NIGMS GM62116 and GM98791 for services pro-
vided by the Glycan Array Synthesis Core (The Scripps Research Institute,
LaJolla, CA) that produced the mammalian glycan microarray and the Protein-

Man6P specificity and structure of scFv M6P-1
191
antibody recognition of a range of natural and synthetic epitopes. J Mol
Biol. 377:450–468.
Olson LJ, Castonguay AC, Lasanajak Y, Peterson FC, Cummings RD,
Smith DF, Dahms NM. 2015. Identification of
a fourth mannose
Bundle DR, Eichler E, Gidney MA, Meldal M, Ragauskas A, Sigurskjold BW,
Sinnott B, Watson DC, Yaguchi M, Young NM. 1994. Molecular recogni-
tion of a Salmonella trisaccharide epitope by monoclonal antibody Se155-4.
Biochemistry. 33:5172–5182.
Castonguay AC, Lasanajak Y, Song X, Olson LJ, Cummings RD, Smith DF,
Dahms NM. 2012. The glycan-binding properties of the cation-independent
mannose 6-phosphate receptor are evolutionary conserved in vertebrates.
Glycobiology. 22:983–996.
6-phosphate binding site in the cation-independent mannose 6-phosphate
receptor. Glycobiology. 25:591–606.
Olson LJ, Dahms NM, Kim JJ. 2004. The N-terminal carbohydrate recognition
site of the cation-independent mannose 6-phosphate receptor. J Biol Chem.
279:34000–34009.
Olson LJ, Hindsgaul O, Dahms NM, Kim JJ. 2008. Structural insights into
the mechanism of pH-dependent ligand binding and release by the cation-
dependent mannose 6-phosphate receptor.
J Biol Chem. 283:
Castonguay AC, Olson LJ, Dahms NM. 2011. Mannose 6-phosphate receptor
homology (MRH) domain-containing lectins in the secretory pathway.
Biochim Biophys Acta. 1810:815–826.
Chavez CA, Bohnsack RN, Kudo M, Gotschall RR, Canfield WM, Dahms NM.
2007. Domain 5 of the cation-independent mannose 6-phosphate receptor
preferentially binds phosphodiesters (mannose 6-phosphate N-acetylgluco-
samine ester). Biochemistry. 46:12604–12617.
10124–10134.
Olson LJ, Peterson FC, Castonguay A, Bohnsack RN, Kudo M, Gotschall RR,
Canfield WM, Volkman BF, Dahms NM. 2010. Structural basis for recog-
nition of phosphodiester-containing lysosomal enzymes by the cation-
independent mannose 6-phosphate receptor. Proc Natl Acad Sci USA.
107:12493–12498.
Olson LJ, Zhang J, Lee YC, Dahms NM, Kim JJ. 1999. Structural basis
Dahms NM, Olson LJ, Kim JJ. 2008. Strategies for carbohydrate recognition by
the mannose 6-phosphate receptors. Glycobiology. 18:664–678.
Distler JJ, Guo JF, Jourdian GW, Srivastava OP, Hindsgaul O. 1991. The
binding specificity of high and low molecular weight phosphomannosyl
receptors from bovine testes. Inhibition studies with chemically syn-
for recognition of phosphorylated high mannose oligosaccharides by the
cation-dependent mannose 6-phosphate receptor.
J Biol Chem. 274:
29889–29896.
Ortolano S, Vieitez I, Navarro C, Spuch C. 2014. Treatment of lysosomal stor-
age diseases: Recent patents and future strategies. Recent Pat Endocr Metab
Immune Drug Discov. 8:9–25.
thesized 6-O-phosphorylated oligomannosides.
J Biol Chem. 266:
21687–21692.
Parolis LA, Parolis H, Kenne L, Meldal M, Bock K. 1998. The extracellular
polysaccharide of Pichia (Hansenula) holstii NRRL Y-2448: The phos-
phorylated side chains. Carbohydr Res. 309:77–87.
Emsley P, Lohkamp B, Scott WG, Cowtan K. 2010. Features and development
of Coot. Acta Crystallogr D Biol Crystallogr. 66:486–501.
Ferro V, Li C, Fewings K, Palermo MC, Linhardt RJ, Toida T. 2002. Determin-
ation of the composition of the oligosaccharide phosphate fraction of Pichia
(Hansenula) holstii NRRL Y-2448 phosphomannan by capillary electro-
phoresis and HPLC. Carbohydr Res. 337:139–146.
Gracy RW, Noltmann EA. 1968. Studies on phosphomannose isomerase. 3. A
mechanism for catalysis and for the role of zinc in the enzymatic and the
nonenzymatic isomerization. J Biol Chem. 243:5410–5419.
Haji-Ghassemi O, Blackler RJ, Martin Young N, Evans SV. 2015. Antibody rec-
ognition of carbohydrate epitopes. Glycobiology. 25:920–952.
Hassid WZ, Ballou CE. 1957. Phosphate esters. In: Pigman WW, editor. The
Carbohydrates. Chemistry, Biochemistry, Physiology. New York: Academ-
ic Press. p. 172–187.
Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC,
Ferrin TE. 2004. UCSF Chimera – A visualization system for exploratory
research and analysis. J Comput Chem. 25:1605–1612.
Pohl S, Encarnacão M, Castrichini M, Müller-Loennies S, Muschol N,
Braulke T. 2010. Loss of N-acetylglucosamine-1-phosphotransferase
gamma subunit due to intronic mutation in GNPTG causes mucolipidosis
type III gamma: Implications for molecular and cellular diagnostics. Am J
Med Genet A. 152A:124–132.
Pohl S, Tiede S, Marschner K, Encarnacao M, Castrichini M, Kollmann K,
Muschol N, Ullrich K, Müller-Loennies S, Braulke T. 2010. Proteolytic
processing of the gamma-subunit is associated with the failure to form
GlcNAc-1-phosphotransferase complexes and mannose 6-phosphate residues
Kamiya Y, Yanagi K, Kitajima T, Yamaguchi T, Chiba Y, Kato K. 2013.
Application of metabolic 13C labeling in conjunction with high-field
nuclear magnetic resonance spectroscopy for comparative confor-
mational analysis of high mannose-type oligosaccharides. Biomolecules.
3:108–123.
Lavinder JJ, Hoi KH, Reddy ST, Wine Y, Georgiou G. 2014. Systematic char-
acterization and comparative analysis of the rabbit immunoglobulin reper-
toire. PLoS ONE. 9:e101322.
on lysosomal enzymes in human macrophages. J Biol Chem. 285:
23936–23944.
Qasba PK, Balaji PV, Rao VS. 1994. Molecular dynamics simulations of oligo-
saccharides and their conformation in the crystal structure of lectin-
carbohydrate complex: Importance of the torsion angle psi for the orienta-
tion of alpha 1,6-arm. Glycobiology. 4:805–815.
Reddy ST, Chai W, Childs RA, Page JD, Feizi T, Dahms NM. 2004. Identifica-
tion of a low affinity mannose 6-phosphate-binding site in domain 5 of
the cation-independent mannose 6-phosphate receptor. J Biol Chem.
279:38658–38667.
Lu W, Navidpour L, Taylor SD. 2005. An expedient synthesis of benzyl
2,3,4-tri-O-benzyl-beta-D-glucopyranoside and benzyl 2,3,4-tri-O-benzyl-
beta-D-mannopyranoside. Carbohydr Res. 340:1213–1217.
Luzio JP, Pryor PR, Bright NA. 2007. Lysosomes: Fusion and function. Nat Rev
Mol Cell Biol. 8:622–632.
Roberts DL, Weix DJ, Dahms NM, Kim JJ. 1998. Molecular basis of lysosomal
enzyme recognition: Three-dimensional structure of the cation-dependent
mannose 6-phosphate receptor. Cell. 93:639–648.
Madhavarao CN, Agarabi CD, Wong L, Müller-Loennies S, Braulke T,
Khan M, Anderson H, Johnson GR. 2014. Evaluation of butyrate-induced
production of a mannose-6-phosphorylated therapeutic enzyme using par-
allel bioreactors. Biotechnol Appl Biochem. 61:184–192.
Schröder S, Matthes F, Hyden P, Andersson C, Fogh J, Müller-Loennies S,
Braulke T, Gieselmann V, Matzner U. 2010. Site-specific analysis of
N-linked oligosaccharides of recombinant lysosomal arylsulfatase A pro-
duced in different cell lines. Glycobiology. 20:248–259.
Makrypidi G, Damme M, Müller-Loennies S, Trusch M, Schmidt B, Schluter H,
Heeren J, Lübke T, Saftig P, Braulke T. 2012. Mannose 6 dephosphoryla-
tion of lysosomal proteins mediated by acid phosphatases Acp2 and
Acp5. Mol Cell Biol. 32:774–782.
Song X, Lasanajak Y, Olson LJ, Boonen M, Dahms NM, Kornfeld S,
Cummings RD, Smith DF. 2009. Glycan microarray analysis of P-type lec-
tins reveals distinct phosphomannose glycan recognition. J Biol Chem.
284:35201–35214.
Martin AC. 1996. Accessing the Kabat antibody sequence database by com-
Tong PY, Gregory W, Kornfeld S. 1989. Ligand interactions of the cation-
independent mannose 6-phosphate receptor. The stoichiometry of mannose
6-phosphate binding. J Biol Chem. 264:7962–7969.
Tong PY, Kornfeld S. 1989. Ligand interactions of the cation-dependent man-
nose 6-phosphate receptor. Comparison with the cation-independent man-
nose 6-phosphate receptor. J Biol Chem. 264:7970–7975.
puter. Proteins. 25:130–133.
McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ.
2007. Phaser crystallographic software. J Appl Crystallogr. 40:658–674.
Müller-Loennies S, Galliciotti G, Kollmann K, Glatzel M, Braulke T. 2010.
A novel single-chain antibody fragment for detection of mannose 6-
phosphate-containing proteins: Application in mucolipidosis type II pa-
tients and mice. Am J Pathol. 177:240–247.
Villeneuve S, Souchon H, Riottot MM, Mazie JC, Lei P, Glaudemans CP,
Kovac P, Fournier JM, Alzari PM. 2000. Crystal structure of an

192
R J Blackler et al.
anti-carbohydrate antibody directed against Vibrio cholerae O1 in complex
Waheed A, von Figura K. 1990. Rapid equilibrium between monomeric, dimer-
ic and tetrameric forms of the 46-kDa mannose 6-phosphate receptor at 37
degrees C. Possible relation to the function of the receptor. Eur J Biochem.
193:47–54.
Wang J, Villeneuve S, Zhang J, Lei P, Miller CE, Lafaye P, Nato F, Szu SC,
Karpas A, Bystricky S, et al. 1998. On the antigenic determinants of the li-
popolysaccharides of Vibrio cholerae O:1, serotypes Ogawa and Inaba. J
Biol Chem. 273:2777–2783.
with antigen: Molecular basis for serotype specificity. Proc Natl Acad Sci
USA. 97:8433–8438.
Vliegenthart JFG, Dorland L, Halbeek Hv. 1983. High-resolution, 1H-nuclear
magnetic resonance spectroscopy as a tool in the structural analysis of
carbohydrates related to glycoproteins. In: Tipson RS, Derek H, editors.
Advances in Carbohydrate Chemistry and Biochemistry. Academic Press.
p. 209–374.
Waheed A, Hille A, Junghans U, von Figura K. 1990. Quaternary structure of
the Mr 46,000 mannose 6-phosphate specific receptor: Effect of ligand, pH,
and receptor concentration on the equilibrium between dimeric and tetra-
meric receptor forms. Biochemistry. 29:2449–2455.
Winn MD, Ballard CC, Cowtan KD, Dodson EJ, Emsley P, Evans PR,
Keegan RM, Krissinel EB, Leslie AG, McCoy A, et al. 2011. Overview of
the CCP4 suite and current developments. Acta Crystallogr D Biol
Crystallogr. 67:235–242.