\makeatletter
\@ifundefined{HCode}
{\documentclass[screen,CRCHIM,Unicode,biblatex,published]{cedram}
\addbibresource{crchim20250737.bib}
\newenvironment{noXML}{}{}
\def\tsup#1{$^{{#1}}$}
\def\tsub#1{$_{{#1}}$}
\def\ndash{\text{--}}
\usepackage[T1]{fontenc}
\def\thead{\noalign{\relax}\hline}
\def\tbody{\noalign{\relax}\hline}
\def\endthead{\noalign{\relax}\hline}
\def\tabnote#1{\vskip4pt\parbox{.92\linewidth}{#1}}
\RequirePackage{etoolbox}
\def\jobid{crchim20250737}
%\graphicspath{{/tmp/\jobid_figs/web/}}
\graphicspath{{./figures/}}
\newcounter{runlevel}
\let\MakeYrStrItalic\relax
\csdef{Seqnsplit}{\\}
\def\refinput#1{}
\def\back#1{}
\def\botline{\\\hline}
\def\dollar{\$}
\def\og{\guillemotleft}
\def\fg{\guillemotright}
\def\Lbreak{\break}
\def\mn{\phantom{$-$}}
\def\0{\phantom{0}}
\def\mathbi#1{\text{\textbf{\textit{#1}}}}
\usepackage{multirow} 
\def\nrow#1{\@tempcnta #1\relax%
\advance\@tempcnta by 1\relax%
\xdef\lenrow{\the\@tempcnta}}
\def\morerows#1#2{\nrow{#1}\multirow{\lenrow}{*}{#2}}
\usepackage{hyperref}
\makeatletter
\g@addto@macro{\UrlBreaks}{\UrlOrds}
\gappto{\UrlBreaks}{\UrlOrds}
\DOI{10.5802/crchim.425}
\datereceived{2025-09-12}
\daterevised{2025-09-23}
\dateaccepted{2025-09-25}
\ItHasTeXPublished
}
{\documentclass[crchim]{article}
\usepackage[T1]{fontenc}
\def\CDRdoi{10.5802/crchim.425}
\def\href#1#2{\url[#1]{#2}}
\let\refinput\input
\makeatletter
\def\CDRsupplementaryTwotypes#1#2{}
\let\splittabular\relax
}
\makeatother

\usepackage{upgreek}

\dateposted{2025-11-28}
\begin{document}

\begin{noXML}

\CDRsetmeta{articletype}{review}

\title{Lectins in the sweet strategy of lung pathogens: from structural
glycobiology to design of glycomimetic pathoblockers}

\alttitle{Lectines et glyco-strat\'{e}gie des pathog\`{e}nes
pulmonaires : de la glycobiologie structurale au concept de
glycomim\'{e}tiques pathobloqueurs}

\author{\firstname{Anne} \lastname{Imberty}\CDRorcid{0000-0001-6825-9527}}
\address{Univ. Grenoble Alpes, CNRS, CERMAV, 38000 Grenoble, France}
\email{anne.imberty@cermav.cnrs.fr}

\editornote{Article submitted by invitation}
\alteditornote{Article soumis sur invitation}

\thanks{CNRS, CBH-EUR-GS (grant no. ANR-17-EURE-0003), Glyco@Alps
(grant no. ANR-15-IDEX-02)}

\keywords{\kwd{Lectins}\kwd{Protein--carbohydrate
interactions}\kwd{Pathoblockers}\kwd{Glycomimetics}}

\altkeywords{\kwd{Lectines}\kwd{Interactions
prot\'{e}ine-sucre}\kwd{Pathobloqueur}\kwd{Glycomim\'{e}tiques}}

\begin{abstract}
Glycan epitopes exposed on the host cell surface serve as receptors for
a wide range of proteins expressed by microorganisms. Among these,
lectins that are produced by viruses, bacteria, and fungi in either
soluble or surface-attached forms, play key roles in infection
processes. The fine specificity that many of these glycan-binding
proteins display toward human glycan epitopes is the result of
long-term co-evolution. Structural insights into lectin architecture,
their binding sites, and the atomic details of their interactions with
glycans are crucial for designing antimicrobial agents capable of
competing with infectious processes and potentially supplementing
antibiotics in the current fight against antimicrobial resistance. This
review highlights how structural lectinology can serve as a starting
point for fruitful collaborations with synthetic carbohydrate chemists
in the design, synthesis, and characterization of glycomimetics acting
as pathoblockers. Various strategies have been developed to target
lectins, exemplified here by soluble lectins from the lung pathogens
\textit{Pseudomonas aeruginosa} and members of the \textit{Burkholderia
cepacia} complex. The affinity for monosaccharide ligands has been
enhanced through heterocyclic functionalization using classical
medicinal chemistry approaches, with the best results obtained by
mimicking natural glycan receptors. Alternatively, the oligomeric
nature of lectins allows the design of divalent or multivalent ligands,
which exhibit much stronger affinity due to the cluster effect.
Covalent inhibitors and non-carbohydrate glycomimetics also represent
promising leads for future strategies to combat antimicrobial
resistance.
\end{abstract}

\begin{altabstract}
Les \'{e}pitopes glycanes expos\'{e}s \`{a} la surface des cellules
h\^{o}tes servent de r\'{e}cepteurs \`{a} un large \'{e}ventail de
prot\'{e}ines exprim\'{e}es par les microorganismes. Parmi elles, les
lectines, produites par des virus, des bact\'{e}ries et des
champignons, jouent un r\^{o}le cl\'{e} dans les processus infectieux.
La sp\'{e}cificit\'{e} fine que pr\'{e}sente un bon nombre de ces
prot\'{e}ines envers les \'{e}pitopes glycanes humains r\'{e}sulte
d'une co\'{e}volution. Les donn\'{e}es structurelles sur l'architecture
des lectines, leurs sites de liaison et les d\'{e}tails atomiques de
leurs interactions avec les glycanes sont essentielles pour concevoir
des agents antimicrobiens capables d'interf\'{e}rer avec les processus
infectieux et de compl\'{e}ter les antibiotiques dans la lutte actuelle
contre la r\'{e}sistance aux antimicrobiens. Cette revue met en
\'{e}vidence comment la lectinologie structurale peut constituer un
point de d\'{e}part pour des collaborations fructueuses entre chimistes
des glucides et biologistes mol\'{e}culaires en vue de concevoir,
synth\'{e}tiser et caract\'{e}riser des glycomim\'{e}tiques agissant
comme pathobloquants. Diverses strat\'{e}gies d\'{e}velopp\'{e}es pour
cibler les lectines sont illustr\'{e}es sur les lectines solubles des
pathog\`{e}nes pulmonaires \textit{Pseudomonas aeruginosa} et
\textit{Burkholderia cepacia}. L'affinit\'{e} pour les ligands
monosaccharidiques a \'{e}t\'{e} renforc\'{e}e par une
fonctionnalisation h\'{e}t\'{e}rocyclique utilisant des approches
classiques de chimie m\'{e}dicinale, les meilleurs r\'{e}sultats
\'{e}tant obtenus en mimant les r\'{e}cepteurs glycanes naturels. Par
ailleurs, la nature oligom\'{e}rique des lectines permet la conception
de ligands divalents ou multivalents, qui pr\'{e}sentent une
affinit\'{e} beaucoup plus forte gr\^{a}ce \`{a} un effet
d'avidit\'{e}. Les inhibiteurs covalents et les glycomim\'{e}tiques non
glucidiques constituent \'{e}galement des pistes prometteuses pour de
futures strat\'{e}gies visant \`{a} lutter contre la r\'{e}sistance aux
antimicrobiens. 
\end{altabstract}

\maketitle

{\vspace*{12pt}}

\twocolumngrid

\end{noXML}

\section{Introduction}\label{sec1}

Lectins are proteins that specifically recognize and bind to
carbohydrates, including glycoconjugates present on the surface of all
living cells~\cite{1}. These glycans participate in a wide range of
biological processes, most of which are mediated through the
recognition of the so-called \textit{glycocode} by protein
readers~\cite{2}.  Indeed, the structural diversity of monosaccharides,
their linkages, and conformational states provide an optimal system for
signal transmission. The evolutionary expansion of glycan complexity,
driven by the diversification of biosynthetic enzymes, has been
paralleled by the emergence of a broad repertoire of glycan-binding
proteins (GBPs), including lectins, capable of mediating both
physiological and pathological functions~\cite{3}.

Traditionally, lectins are defined as proteins that bind specifically
to certain sugars and thereby cause agglutination of particular cell
types~\cite{4}. This definition excludes both enzymes and antibodies of
the adaptive immune system. Lectins can exist as \mbox{soluble} proteins or as
modules integrated into larger proteins, where their primary function
is to engage glycan targets. In pathogenic bacteria, for example,
adhesins act as virulence factors by mediating attachment to host
surfaces, often through pili or flagella that incorporate one or more
lectin domains~\cite{5}. Similarly, many bacterial toxins, such as
those responsible for cholera or tetanus, require glycan recognition
for binding to host membranes prior to cell entry or membrane
disruption~\cite{6}. Because lectins frequently present multiple
binding sites, achieved either through oligomerization or tandem
repeats, they are often called agglutinins. While the affinity of a
single binding site is typically low, this is compensated by
multivalency, resulting in strong avidity (the so-called Velcro effect)
when glycans are presented in multiple copies~\cite{7}.

Blocking lectin--glycan interactions can interfere with infection
processes and biofilm formation. As such, lectins are increasingly
recognized as promising targets for the development of
pathoblockers~\cite{8}. These compounds are designed to inhibit
virulence factors or other pathogenic mechanisms without affecting
bacterial viability. By targeting processes such as adhesion, invasion,
quorum sensing, and biofilm formation, pathoblockers disarm pathogens
rather than kill them. This approach exerts minimal selective pressure,
thereby reducing the risk of resistance development, while preserving
the beneficial commensal microbiota. Pathoblockers are now considered
as alternative, or more likely adjunctive, therapies alongside
antibiotics, and hold significant promise in the fight against
multidrug-resistant pathogens~\cite{9}.

\section{Structural information on lectins from pathogens involved in
airway infections}\label{sec2} 

The airway epithelium is covered by a dense glycocalyx composed of
\textit{O}-glycosylated mucins, \textit{N}-glycoproteins, and
glycolipids. The composition and frequency of glycan epitopes in the
respiratory tract vary according to tissue type, with notable
differences in the sialylated glycans expressed in the upper versus
lower airways~\cite{10}. Glycan signatures are also altered in certain
pathologies: for instance, cystic fibrosis is associated with
abnormally thick mucus and modification of the relative abundance of
sialic acid and fucose residues~\cite{11}. From the nasal epithelium to
the alveoli, the respiratory tract constitutes the primary entry point
for numerous airborne pathogens, which can lead to life-threatening
diseases such as tuberculosis or Middle East Respiratory Syndrome.
Many of these pathogens are hospital-associated and exhibit multidrug
resistance, including \textit{Pseudomonas aeruginosa}~\cite{12} and
members of the \textit{Burkholderia cepacia} complex (BCC)~\cite{13}.
Emerging fungal infections further exacerbate this burden, as they are
often characterized by increasing antifungal resistance and delayed
diagnosis.

While mucins play a protective role by trapping pathogens, many
microorganisms have evolved mechanisms to exploit host glycans in order
to colonize and invade airway tissues. Lectins, in particular, play a
central role in these processes, mediating specific recognition of
epithelial glycans and thereby facilitating infection or biofilm
formation. A summary of lectins implicated in airway infections,
together with their specificities and structural data, is provided in
Table~\ref{tab1}.

%tab1
\begin{table*}
\caption{\label{tab1}Lectins from pathogens responsible for airway
infections with available structural data}
\tabcolsep=3pt
\begin{tabular}{ccccc}
\thead
Species & Lectins & \parbox[t]{3pc}{\centering Structural class} 
& \parbox[t]{5pc}{\centering Monosaccharide specificity} & 
\parbox[t]{3pc}{\centering PDB code}\vspace*{2pt} \\
\endthead
\multicolumn{5}{c}{Bacteria} \\ 
\textit{Bordetella pertussis} & PTX & AB\tsub{5} toxin & NeuAc & 1PTO
\\ 

\textit{Burkholderia ambifaria} & BambL & ${\upbeta}$-propeller & Fuc &
3ZW0 \\ 

\textit{Burkholderia cenocepacia} & \parbox[t]{8pc}{\centering Bc2L-A,
Bc2L-C-Cter Bc2L-C-Nter} & \parbox[t]{6pc}{\centering 2-calcium lectin
TNF${\upalpha}$-like} & \parbox[t]{3pc}{\centering Man\Lbreak Fuc} &
\parbox[t]{3pc}{\centering 2VNV\Lbreak 3WQ4}\vspace*{2pt} \\ 

\textit{Mycoplasma pneumoniae} & P40/P90 adhesin &
\parbox[t]{6pc}{\centering Mycoplasma adhesin}\vspace*{2pt} & NeuAc & 6TLZ \\ 

\textit{Pseudomonas aeruginosa} & \parbox[t]{6pc}{\centering Pyocin
LecA/PA-IL LecB/PA-IIL} & \parbox[t]{7pc}{\centering Monocot-lectin
like 1-calcium lectin 2-calcium lectin} & \parbox[t]{3pc}{\centering
Man\Lbreak ${\upalpha}$Gal\Lbreak Fuc} & 
\parbox[t]{3pc}{\centering 4LEA\Lbreak 1OKO\Lbreak 1GZT}\vspace*{2pt} \\ 

\textit{Yersinia pestis} & PsaA & Bacterial adhesin & Gal & 2HB0
\vspace*{5pt} \\ 

\multicolumn{5}{c}{Fungi} \\ 
\textit{Aspergillus fumigatus} & AFL1/FleA & ${\upbeta}$-propeller &
Fuc & 4AGI \\ 

\textit{Scedosporium apiospermum} & SapL1 & ${\upbeta}$-propeller & Fuc
& 6TRV \vspace*{5pt} \\ 

\multicolumn{5}{c}{Viruses} \\ 

\parbox[t]{10pc}{\centering Human coronavirus (HCoV) MERS-CoV
SARS-CoV-2} & Spike glycoprotein & \parbox[t]{6pc}{\centering
Coronavirus spike protein} & \parbox[t]{6pc}{\centering 9Ac-NeuAc\Lbreak NeuAc\Lbreak
NeuAc} & \parbox[t]{3pc}{\centering 8OPM\Lbreak 6Q04\Lbreak 7QUR}\vspace*{2pt} \\ 

Influenza virus & Hemagglutinin & Influenza hemag & NeuAc & 3HTQ
\botline
\end{tabular}
\tabnote{Fuc: fucose; Man: mannose, Gal:
galactose, NeuAc: \textit{N}-acetylneuraminic acid; 9Ac-NeuAc:
\textit{N}-acetylneuraminic acid with acetyl group at C9.}
\vspace*{-2pt}
\end{table*}

The most extensively studied viral lectin is hemagglutinin (HA) from
influenza A virus. Sequence variations in HA have resulted in at least
18 subtypes identified across birds and mammals. Importantly, HA
specificity underlies the interspecies transmission barrier: avian
influenza strains preferentially recognize sialic acid on
NeuAc(${\upalpha}$2-3)Gal epitopes, whereas human strains target
NeuAc(${\upalpha}$2-6)Gal receptors~\cite{14}. In coronaviruses, the
N-terminal domain of the spike glycoprotein contains a galectin-like
lectin module. Binding to sialylated glycans is well established in
human coronaviruses (HCoVs)~\cite{15} and in MERS-CoV~\cite{16}, where
it constitutes a primary receptor-recognition mechanism. Although the
lectin domain is structurally conserved in SARS-CoV-2, the sialic
acid-binding activity observed in the early Wuhan strain~\cite{17} has
been lost in subsequent variants~\cite{18}.

Lectins from fungal opportunistic pathogens also play a central role in
host--pathogen interactions. In \textit{Aspergillus fumigatus}, lectins
expressed on both conidia and hyphae bind efficiently to fucosylated
glycans on epithelial cells and mucins~\cite{19}. A closely related
lectin has been characterized in \textit{Scedosporium apiospermum}, an
emerging opportunistic pathogen of growing clinical concern~\cite{20}.
Both fungi are associated with nosocomial infections and display
increasing resistance to antifungal treatments.

Additional structural data are available for lectins from bacterial
pathogens causing respiratory \mbox{diseases,} including \textit{Bordetella
pertussis} (whooping cough), \textit{Yersinia pestis}
(plague)~\cite{21} and \textit{Mycoplasma pneumoniae} (atypical
pneumonia)~\cite{22}. However, the present review focuses on
\textit{Pseudomonas aeruginosa} and species of the \textit{Burkholderia
cepacia} complex. Rather than providing a comprehensive account of
efforts to develop high-affinity ligands against microbial lectins, the
aim here is to illustrate how structural studies have guided the
rational design of active compounds through distinct strategies. This
review also highlights how structural and biophysical investigations
conducted at the CERMAV center for research on plant macromolecules
have fostered a productive dialogue with carbohydrate chemistry groups,
ultimately advancing the synthesis of antibacterial pathoblockers
effective against \textit{P.~aeruginosa} and different species of\break
\textit{Burkholderia}.

\begin{figure*}
\includegraphics{fig01}
{\vspace*{-2pt}}
\caption{\label{fig1}Crystal structures of soluble lectins from
bacterial pathogens in their oligomeric form and in complex with human
glycan epitopes. (A) LecA from \textit{P.~aeruginosa} complexed with iso-Gb3
trisaccharide Gal(${\upalpha}$1-3)Gal(${\upbeta}$1-4)Glc, PDB 2VXJ. (B) LecB
from \textit{P.~aeruginosa} complexed with Lewis a trisaccharide
Gal(${\upbeta}$1-3)[Fuc(${\upalpha}$1-4)]GlcNAc, PDB W8H. (C) BambL from
\textit{B.~ambifaria} complexed with H-type 1 trisaccharide
Fuc(${\upalpha}$1-2)Gal(${\upbeta}$1-3)GlcNAc, PDB 3ZW2. (D) Bc2LC-nt from
\textit{B.~cenocepacia} complexed with H-type 1 trisaccharide
Fuc(${\upalpha}$1-2)Gal(${\upbeta}$1-3)GlcNAc, PDB 6TID. Proteins are
represented by ribbons colored by chains, glycans by sticks, and
calcium ions by green spheres (Pymol, Schr\"{o}dinger). Oligosaccharide
ligands are also represented as cartoons using the Symbol Nomenclature
for Graphical Representations of Glycans~\cite{34} (Gal: yellow circle;
Fuc: red triangle; Glc: blue circle, GlcNAc: blue square).} 
{\vspace*{-4pt}}
\end{figure*}

\subsection{Lectins from \textit{Pseudomonas aeruginosa} and
\textit{Burkholderia cepacia} complex}\label{ssec21}

\textit{Pseudomonas aeruginosa} is a ubiquitous opportunistic pathogen
associated with severe infections in immunocompromised individuals,
particularly in patients with cystic fibrosis and those under
mechanical ventilation~\cite{23}. It is included in the World Health
Organization's priority list of pathogens due to its critical role in
the development of antimicrobial resistance. \textit{P.~aeruginosa}
produces several soluble factors, including pyocins, i.e., short
peptides with antibacterial activity, and two lectins, LecA and LecB,
which are specific for galactose and fucose, respectively~\cite{24}.
Both lectins are tetrameric and bind monosaccharides through
interactions that involve calcium ions~\cite{25,26} (Figure~\ref{fig1}A
and \ref{fig1}B). LecA binds to host glycolipids, inducing
rearrangements of the plasma membrane that not only exert cytotoxic
effects but also promote cellular internalization of the
bacterium~\cite{27}. LecB has a strong preference for Lewis a
oligosaccharides but binds broadly to fucose residues on glycoproteins
and glycolipids~\cite{28}. Both lectins contribute significantly to
biofilm formation~\cite{29}.

The \textit{Burkholderia cepacia} complex (BCC) comprises at least
twenty distinct \textit{Burkholderia} species, many of which are
clinically important opportunistic pathogens in cystic fibrosis
patients~\cite{30}. \textit{Burkholderia cenocepacia} produces several
lectins with LecB-like domains, showing specificity for fucose and
mannose~\cite{31}. Among them, BC2L-C is notable for its modular
structure: its N-terminal domain resembles LecB, while its C-terminal
domain adopts a tumor necrosis factor (TNF)-${\upalpha}$-like fold, also
with fucose specificity~\cite{32} (Figure~\ref{fig1}D). In contrast,
\textit{Burkholderia ambifaria} produces a different lectin that
assembles into a trimeric ${\upbeta}$-propeller, exposing six
fucose-binding sites~\cite{33} (Figure~\ref{fig1}C).

\begin{figure*}
{\vspace*{-2pt}}
\includegraphics{fig02}
{\vspace*{-2pt}}
\caption{\fontsize{9.9}{12}\selectfont\label{fig2}Details of the glycan
binding sites of bacterial lectins from crystal structures.
(A)~LecA/galactose, PDB 1OKO.  (B)~LecB/fucose, PDB 1GZT. (C)
BambL/fucose, PDB 3ZW0. (D)~BC2LC-nt/SeMeFuc, PDB 3WQ4. Hydrogen bonds
are represented by green dashed lines, calcium ions by green spheres
and water molecules as small red balls. Schematic representation of
ligands is displayed in the right upper corner of each panel
(ChemSketch, ACD/Labs).} 
{\vspace*{-4pt}}
\end{figure*}

\subsection{Forces involved in protein--sugar interaction}
\label{ssec22}

The binding sites of the four lectins discussed in this review are
shown in Figure~\ref{fig2}, and they exemplify the classical
interactions typically observed in protein--carbohydrate recognition.
In all cases, hydroxyl groups of the glycans participate in hydrogen
bonding with amino acid side chains within the protein binding site.
Statistical analyses have shown that among polar amino acids,
asparagine and aspartate are preferred, occurring at approximately
twice the frequency expected by chance~\cite{35}. However, aromatic
amino acids are most strongly favored, with a pronounced prevalence of
tryptophan, followed by tyrosine and then histidine. 

This preference is attributed to CH--${\uppi}$ interactions, whereby
the CH--presenting face of monosaccharides interacts with electron-rich
aromatic side chains~\cite{36}. For example, fucose interacts with
tryptophan in the binding site of BambL (Figure~\ref{fig2}C).
Additional contacts are often mediated by water molecules that are
integral components of the binding site; for instance, water molecule
W1 in LecA is consistently observed in crystal structures and forms a
hydrogen bond with O6 of the glycan within a small pocket of the
protein surface (Figure~\ref{fig2}A). The two \textit{P.~aeruginosa}
lectins \mbox{exhibit} a nonconventional mode of glycan recognition. In these
cases, calcium ions are directly involved in binding: O3 and O4 of
galactose in LecA and O2, O3, and O4 in LecB are coordinated by calcium
(Figure~\ref{fig2}A and~\ref{fig2}B). The presence of two calcium ions
is unique to this latter lectin class, and a recent neutron diffraction
study demonstrated their role in creating low-barrier hydrogen bonds,
which contribute to the unusually high affinity of these
interactions~\cite{37}.

\vspace*{-1pt}

\section{Strategies for designing high-affinity glycomimetics against
lectins from pathogens}\label{sec3}

\vspace*{-1pt}

The structural information described above served as the basis for the
development of high affinity glycomimetics that are able to compete
with the binding of the lectins to the glycan targets with the aim of
inhibiting the first stage of adhesion, but also biofilm formation.
High-affinity glycomimetics are also a strategy for the vectorization
of antibiotics~\cite{38} and other compounds or for labeling pathogen
or biofilms by fluorescent compounds~\cite{39}.

\vspace*{-1pt}

\subsection{Structure-based design of monovalent glycomimetics}
\label{ssec31}

\vspace*{-1pt}

LecA binds to the terminal ${\upalpha}$-galactose residue present in Gb3,
which belongs to the globosphingolipid group of glycolipids and
consists of the Gal(${\upalpha}$1-4)Gal(${\upbeta}$1-4)Glc trisaccharide
linked to ceramide~\cite{40}. The affinity of LecA for galactose or for
its natural ligand Gal(${\upalpha}$1-4)Gal is relatively low with
dissociation constant ($K_{\mathrm{d}}$) between 50 and 100~$\upmu$M,  but this
limitation is compensated during infection by multivalent effects
induced by the many oligosaccharide heads of glycolipids present on the
membrane. An isomer of the natural ligand, the isogloboside
trisaccharide Gal(${\upalpha}$1-3)Gal(${\upbeta}$1-4)Glc, displays
comparable affinity and the crystal structure of the complex with LecA
has been described (Figure~\ref{fig3}A)~\cite{40}, providing a basis
for structure-guided design of mimetics. In the development of
glycomimetics, ${\upbeta}$-galactosides with aromatic aglycones exhibit
enhanced affinity, largely due to a T-shaped electronic interaction
with His50 in the binding site~\cite{41}. For example,
\textit{p}-nitrophenyl-${\upbeta}$-galactoside (PNP-Gal) binds with a
$K_{\mathrm{d}}$ of 5~$\upmu$M and the particular interaction is illustrated in
Figure~\ref{fig3}B. Other \textit{O}- and \textit{S}-linked aromatic
derivatives have also been investigated~\cite{42}, and the best
monovalent ligand reported to date, a diarylthiogalactoside reaches an
affinity of ${\sim}1~\upmu$M~\cite{43}. This ligand adopts a binding
mode closely resembling that of the natural ligand
(Figure~\ref{fig3}C), underscoring the potential of ligand design
strategies that mimic the structural features of native glycans.

\begin{figure*}[p!]
\includegraphics{fig03}
\caption{\label{fig3}Crystal structures of LecA (surface in beige and
ribbon in blue) complexed with natural and synthetic ligands. (A)
Gal(${\upalpha}$1-3)Gal(${\upbeta}$1-4)Glc, PDB 2VXJ. (B)
$p$-Nitrophenyl-${\upbeta}$-galactoside, PDB 3ZYF. 
(C)~Diarylthiogalactoside, PDB 7Z63. (D) Cyanocatechol, PDB 6YO3. (E)
Phenylbutyryl hydroxamic acid, PDB 7FJH. (F) Difluoro tolcapone
derivative, PDB 9I7Z. Schematic representation of ligands is displayed
in the right upper corner of each panel (ChemSketch, ACD/Labs).}
\end{figure*}

LecB exhibits strong affinity for natural ligands such as fucose
($K_{\mathrm{d}}\approx 3~\upmu$M) and Fuc(${\upalpha}$1-4)GlcNAc-containing
trisaccharides, with a $K_{\mathrm{d}}$ of 210~nM for the Lewis a antigen
(Figure~\ref{fig4}A) \cite{28}. Functionalization at the reducing end
of this disaccharide yielded potent inhibitors with affinities
comparable to those of the natural trisaccharide (Figure~\ref{fig4}B)
\cite{44}. 

\begin{figure*}
\includegraphics{fig04}
\caption{\label{fig4}Crystal structures of LecB (surface in beige)
complexed with natural and synthetic ligand. (A)
Gal(${\upbeta}$1-3)[Fuc(${\upalpha}$1-4)]GlcNAc, PDB W8H. (B)
Fuc(${\upalpha}$1-4)GlcNAc triazole derivative, PDB 2JDK. (C)
${\upbeta}$-Fucopyranosyl-thiophenesulfonamide derivative, PDB 5MAZ. (D)
${\upbeta}$-Fucopyranosyl-biphenyl-3-carboxamide, PDB 8AIY. Schematic
representation of ligands is displayed in the right upper corner of
each panel (ChemSketch, ACD/Labs).} 
{\vspace*{-2pt}}
\end{figure*}

The observation that \textsc{d}-mannose, which is stereochemically
related to \textsc{l}-fucose, is also a ligand for LecB has inspired
the design of carbohydrate ring mimics, in which the axial anomeric
oxygen of fucose is replaced by an equatorial sulfonamide
moiety~\cite{45} (Figure~\ref{fig4}C). Building on this approach,
${\upbeta}$-fucosyl amides, sulfonamides, and thiourea derivatives were
synthesized, among which a ${\upbeta}$-fucosyl amide diaryl derivative
displayed both high affinity and notable stability in plasma~\cite{46}
(Figure~\ref{fig4}D).

BambL from \textit{B.~ambifaria} also interacts with fucose-containing
oligosaccharides, exhibiting high affinity for the fucose
monosaccharide~\cite{33} as well as for fucosylated glycans such as
Lewis and ABO antigens (Figure~\ref{fig5}A). The synthesis of a
conformationally constrained aryl-${\upalpha}$-\textit{O}-fucosyl
analogue produced a compound with binding affinity comparable to that
of native fucose, but with substantially enhanced selectivity, as it is
not recognized by either human fucose-binding lectins or
LecB~\cite{47}. The crystal structure of the BambL--ligand complex
further revealed that the compound acts as an effective glycomimetic of
natural fucosylated glycans, engaging the protein surface at the same
binding site as the blood group B oligosaccharide~\cite{48}
(Figure~\ref{fig5}B).

\begin{figure*}
{\vspace*{3pt}}
\includegraphics{fig05}
{\vspace*{3pt}}
\caption{\label{fig5}Crystal structures of BambL (surface in beige)
complexed with natural and synthetic ligands. (A)
Gal(${\upalpha}$1-3)[Fuc(${\upalpha}$1-2)]Gal(${\upbeta}$1-4)GlcNAc, PDB
3ZWE. (B) Conformationally constrained fucose-based glycomimetics, PDB
6ZFC. Schematic representation of ligands is displayed in the right
upper corner of each panel (ChemSketch, ACD/Labs).}
{\vspace*{6pt}}
\end{figure*}

BC2L-C-nt from \textit{B. cenocepacia} also binds fucose but displays a
marked preference for H-type 1 (Figure~\ref{fig6}A) and H-type 3
epitopes~\cite{49}, albeit with modest affinity. An innovative
fragment-based theoretical approach identified a binding pocket
adjacent to the fucose-binding site, and several potential hits were
subsequently validated using biophysical methods~\cite{50}. From the
results of the {fragment-based} screening, bifunctional ligands were
\mbox{synthesized} \mbox{employing} 
a panel of rationally selected linkers, yielding
an alkene-bound glycomimetic with a tenfold increase in affinity
relative to fucose (Figure~\ref{fig6}B)~\cite{51}. Although
\textit{N}-fucosides were also synthesized, their poor aqueous
solubility limited their application as antagonists. However, a
subsequent \mbox{generation} of derivatives overcame this limitation,
exhibiting both satisfactory affinity and improved solubility
(Figure~\ref{fig6}C)~\cite{52}.

\begin{figure*}
\includegraphics{fig06}
\caption{\label{fig6}Crystal structures of BC2L-C-nt (surface in beige)
complexed with natural and synthetic ligands. (A)
Fuc(${\upalpha}$1-2)Gal${\upbeta}$(1-3)GlcNAc, PDB 6TID. (B) Ethynyl-phenyl
derivative of fucose, PDB 7OLU. (C)~Aminomethyl-phenyl derivative of
fucose, PDB 8BRO. Schematic representation of ligands is displayed in
the right upper corner of each panel (ChemSketch, ACD/Labs).}
{\vspace*{-2pt}}
\end{figure*}

\subsection{Covalent glyco-derived inhibitors}\label{ssec32}

Covalent inhibitors have recently been developed to target growth
factors in cancer and proteases in \mbox{viral} infections. Because
specificity represents the primary challenge in this strategy,
successful design requires precise positioning of the ligand warhead in
close proximity to a reactive amino acid within the lectin binding
site. In the case of LecA, this was achieved by exploiting Cys62,
located at the base of the binding pocket near the secondary hydroxyl
group of galactose (Figure~\ref{fig2}A)~\cite{53}. A
phenyl-${\upbeta}$-galactoside derivative bearing an epoxy group at the
C6 position was synthesized, and formation of the covalent adduct was
confirmed by mass spectrometry. Furthermore, a fluorescein-conjugated
derivative enabled specific staining of \textit{P.~aeruginosa}
biofilms.

Another warhead was used for covalent inhibition of BC2L-C-nt by
fucosides connected to salicylaldehyde groups by a variety of linkers
with the aim of targeting Lys108 in the binding site of the
lectin~\cite{54}. Mass spectrometry conducted on the whole lectin and
on digestion peptides confirmed the formation of the covalent adduct.

\subsection{Non-carbohydrate glycomimetics}\label{ssec33}

All of the inhibitors described above retain the principle of a glycan
residue bound in the primary binding site of the lectin. In contrast,
non-carbohydrate glycomimetics offer the potential for more
straightforward synthetic routes, broader chemical diversity, and
improved pharmacokinetic properties. Using a virtual screening
strategy, more than 1500 compounds from the National Cancer Institute
(NCI) Diversity set IV were docked into LecA, leading to the
identification of promising candidates~\cite{55}. Many of these
compounds contained a catechol moiety, raising the possibility of false
positives due to nonspecific reactivity with the protein surface.
Nevertheless, screening of catechol libraries with multiple \mbox{biophysical}
methods confirmed that several electron-deficient catechols bind
specifically to LecA. Although their affinities were only in the
millimolar range, the crystal structure of LecA in complex with
cyanocatechol demonstrated that its oxygen atoms mimic the roles of
galactose O3 and O4 in coordinating calcium within the binding site
(Figure~\ref{fig3}D)~\cite{55}. Building on this concept of
metal-binding pharmacophores as novel scaffolds for inhibiting
Ca$^{2+}$-dependent lectins, both virtual and NMR-based screening of
general and specialized compound libraries were undertaken~\cite{56}.
Several hydroxamates were identified as specific LecA ligands, again
with affinities in the low millimolar range, and crystal structures
confirmed their calcium-coordinating interactions
(Figure~\ref{fig3}E)~\cite{56}. By contrast, malonates displayed
broader activity, binding not only to LecA and LecB but also to human
C-type lectins. To further develop the catechol-based scaffold, 3267
unique catechol derivatives from the Hoffmann-La Roche compound library
were tested for LecA binding~\cite{57}. Among these, tolcapone, a drug
used in the treatment of Parkinson's disease, and several structural
analogues, emerged as potent ligands with affinities around
10~$\upmu$M---approximately 5- to 10-fold stronger than galactose. The
crystal structure of one representative compound (Figure~\ref{fig3}F)
revealed that these ligands not only coordinate calcium but also form
hydrogen bonds with conserved water molecules and engage in hydrophobic
interactions with the protein surface.

\subsection{Structure-based multivalent ligands}\label{ssec34}

Structural information on lectins can also be used for the design of
multivalent ligands with appropriate linkers to reach two or more
binding sites of the lectins, resulting in a chelating effect that can
result in several-fold increases in affinity~\cite{7}. 

A substantial body of literature has described the synthesis of
glycoclusters, glycopolymers, and other multivalent ligands targeting
LecA and LecB (see Refs.~\cite{58,59,60}). In the case of 
\textit{P.~aeruginosa}, the present article focuses on selected
examples of structure-based design of divalent ligands for LecA. The
architecture of the LecA tetramer (Figure~\ref{fig1}A), which presents
two galactose-binding sites separated by approximately 29~\AA{} on the
same face, provides an optimal framework for the rational design of
divalent ligands. Flexible linkers, such as those based on polyethylene
glycol (PEG), proved ineffective in promoting efficient chelation due
to the entropic penalty arising from their conformational flexibility.
In contrast, introducing additional interactions at the protein
surface, for instance, through linkers identified from a large
heteroglycoconjugate library generated by nucleic-acid-encoded peptide
synthesis, yielded high-affinity binders with dissociation constants as
low as 82~nM~\cite{61}. Rigid spacers proved even more effective at
eliciting strong avidity effects, provided they positioned the two
galactose moieties at the appropriate distance while retaining adequate
aqueous solubility. Subsequent optimization, involving the
incorporation of alternating glucose and triazole units, resulted in a
ligand with an affinity of 28~nM, corresponding to an approximately
100-fold improvement relative to the monovalent ligand~\cite{62}.
Remarkably, this compound was the only divalent ligand successfully
crystallized with LecA (Figure~\ref{fig7}A), as the high protein
concentrations required for crystallization typically favor ligand
cross-bridging rather than intramolecular chelation. Additional potent
divalent galactosides, with dissociation constants in the range of
10--40~nM, were later reported by Titz and colleagues, who also
demonstrated their improved solubility and effectiveness in reducing
\textit{P.~aeruginosa} invasiveness in host cells
(Figure~\ref{fig7}B)~\cite{39}. \looseness=1

\begin{figure}
\includegraphics{fig07}
\caption{\label{fig7}Some multivalent glycosides with high affinity for
bacterial lectins. (A) Divalent galactoside with
triazolyl--glucose-based linker in complex with LecA (surface in beige,
crystal structure PDB 4YWA). (B) Divalent galactoside with
aromatic-based linker also binding to LecA. (C) Pentavalent fucoside
with high affinity for BambL (model superimposed on crystal structure,
PDB 6ZFC).}
\end{figure}

The spherical architecture of LecB, with its fucose-binding sites
positioned far apart, is not favorable for the design of divalent
chelating ligands. Consequently, only limited improvements in binding
affinity have been achieved relative to fucose itself~\cite{60}. In
contrast, the propeller-shaped architecture of BambL is more amenable
to multivalent ligand design, as it allows simultaneous engagement of
two or more of the six fucose-binding sites, each separated by
approximately 20~\AA{}. The glycomimetic structure depicted in
Figure~\ref{fig5}B was assembled into a hexavalent ligand using a
peptide raft scaffold~\cite{63}. The resulting construct exhibited the
appropriate shape and size to interact with multiple sites of the BambL
propeller, as demonstrated by the structural superposition shown in
Figure~\ref{fig7}C, and achieved a $K_{\mathrm{d}}$ of 14~nM. An alternative
strategy exploited ligand cooperativity through PNA--fucose conjugates,
which contained only a single fucose residue but were capable of
hybridizing via the PNA (peptide nucleic acid) backbone in the presence
of the protein~\cite{64}. This cooperative assembly resulted in
high-affinity binding, with a\break $K_{\mathrm{d}}$ of 11~nM.

\section{Prospectives: identification of new targets}\label{sec4}

This article highlights a subset of bacterial species implicated in
lung infections, yet the principles discussed are broadly applicable to
other carbohydrate-dependent pathogens. Indeed, comparable strategies
have already proven effective against uropathogenic \textit{Escherichia
coli}, which use FimH adhesin to bind to mannosylated epitopes on
urothelial cells, a crucial step in urinary tract infections~\cite{65}.
Looking forward, the identification of novel lectin targets is likely
to expand considerably, as systematic data mining of genomic resources
now enables the prediction of putative lectins across bacterial, viral,
and fungal species~\cite{66}. Dedicated tools such as LectomeXPlore
will accelerate this process, providing a framework for the rational
prioritization of targets~\cite{67}. In parallel, the rapid advances of
carbohydrate-based therapeutics in oncology and vaccinology underscore
their untapped potential in infectious disease management. Positioned
as a complementary strategy to antibiotics, these approaches could play
a decisive role in addressing the dual challenges of antimicrobial
resistance and the emergence of novel pathogens driven by climate
change.\looseness=-1

\section*{Acknowledgments}

Pathogens in the graphical abstract are designed by Freepik
(\href{http://www.freepik.com/}{www.freepik.com}).

\section*{Declaration of interests}

The author does not work for, advise, own shares in, or receive funds
from any organization that could benefit from this article, and has
declared no affiliations other than their research organization.

\section*{Funding}

Support from CNRS, CBH-EUR-GS (ANR-17-EURE-0003) and Glyco@Alps
(ANR-15-IDEX-02) is acknowledged.

\CDRGrant[ANR]{ANR-17-EURE-0003}
\CDRGrant[ANR]{ANR-15-IDEX-02}

\back{}

\printbibliography
\refinput{crchim20250737-reference.tex}

\end{document}