GPCR-targeting nanobodies: attractive research tools, diagnostics, and therapeutics

Review

GPCR-targeting nanobodies: attractive research tools, diagnostics, and therapeutics Azra Mujic´-Delic´*, Raymond H. de Wit*, Folkert Verkaar, and Martine J. Smit Amsterdam Institute for Molecules, Medicines and Systems (AIMMS), Division of Medicinal Chemistry, Faculty of Sciences, VU University Amsterdam, de Boelelaan 1083, 1081 HV Amsterdam, The Netherlands

G-protein-coupled receptors (GPCRs) represent a major therapeutic target class. A large proportion of marketed drugs exert their effect through modulation of GPCR function, and GPCRs have been successfully targeted with small molecules. Yet, the number of small new molecular entities targeting GPCRs that has been approved as therapeutics in the past decade has been limited. With new and improved immunization-related technologies and advances in GPCR purification and expression techniques, antibody-based targeting of GPCRs has gained attention. The serendipitous discovery of a unique class of heavy chain antibodies (hcAbs) in the sera of camelids may provide novel GPCR-directed therapies. Antigen-binding fragments of hcAbs, also referred to as nanobodies, combine the advantages of both small molecules (e.g., molecular cavity binding, low production costs) and monoclonal antibodies (e.g., high affinity and specificity). Nanobodies are gaining ground as therapeutics and are also starting to find application as diagnostics and as high-quality tools in GPCR research. Herein, we review recent advances in the use of nanobodies in GPCR research. Towards nanobody-based targeting of GPCRs GPCRs (see Glossary) are seven transmembrane-spanning proteins that detect a vast repertoire of stimuli (ranging from light to large glycoproteins) and activate various intracellular signaling pathways. Upon activation, GPCRs undergo a conformational change that enables the recruitment and activation of proteins that relay signals from the plasma membrane to the cell interior. These include several classes of G proteins that modulate effector proteins and the scaffolding protein b-arrestin, which has been implicated in both termination of G protein-mediated GPCR signaling as well as initiation of distinct GPCRdependent signaling cascades [1]. Corresponding author: Smit, M.J. ([email protected]). Keywords: GPCRs; nanobodies; VHH; single-domain antibody; signaling. * These authors contributed equally to this work. 0165-6147/$ – see front matter ß 2014 Published by Elsevier Ltd. http://dx.doi.org/10.1016/j.tips.2014.03.003

GPCR-induced signaling is a virtually universal means of intercellular communication, and deregulated GPCR signaling has been implicated in a multitude of diseases. An estimated 30% of marketed drugs exert their effect through modulation of GPCR function [2]. Considering that only a small proportion of these receptors are targeted by current therapies [3], the GPCR family provides a large untapped source of potential therapeutic interventions. Not surprisingly, many pharmaceutical companies have active programs aimed at identifying novel GPCR-interacting molecules. Low molecular weight molecules have proven to be very successful as therapeutics. Yet, in the past decade the number of small new molecular entities targeting GPCRs that were approved as therapeutics has Glossary Affinity: a biochemical parameter representing the intrinsic ligand–target binding strength, generally described by the equilibrium constant (Kd) of association (Kon) and dissociation rates (Koff). Avidity: functional affinity, representing the combined binding strength of multiple ligand–target interactions. Complementarity-determining region (CDR): the variable domains of antibodies contain three hypervariable CDR loop regions, which primarily form the antigen-binding paratope. Chemokine receptors: a family of GPCRs that upon activation by chemokines orchestrate directional migration of immune cells in homeostatic and inflammatory conditions. Clearance: a pharmacokinetic parameter depicting the rate at which a substance is excreted from the body. Crystallization chaperone: a molecule that binds the crystallization target and increases the probability of high-quality crystallization potentially through restriction of conformation diversity. Epitope: the antigen surface that is recognized by an antibody (fragment). Formatting: the procedure of multimerization to improve the pharmacokinetic and/or pharmacodynamic properties of an antibody fragment. G protein-coupled receptors (GPCRs): a superfamily of seven transmembrane domain cell surface receptors, which are key modulators of signal transduction and well-established drug targets. Heavy chain antibody (hcAb): a fully functional antibody devoid of light chains naturally expressed in camelids. Nanobody (Nb): an antibody fragment consisting of the recombinant 12– 15 kDa VHH domain. Nbs are the smallest functional antibody-based biologics known to date. Paratope: the interacting surface of the antibody (fragment) that binds the epitope. Positron emission tomography (PET): an indirect high resolution molecular imaging technique using a radiolabeled tracer. Single photon emission computed tomography (SPECT): a direct molecular imaging technique using a radiolabeled tracer. Tissue penetration: a pharmacokinetic parameter describing the propensity of a molecule to penetrate deep into the tissue. Valency: the number of paratopes of an antibody (fragment). VHH: the variable domain of a heavy chain antibody, which is fully capable of binding an antigen.

Trends in Pharmacological Sciences, May 2014, Vol. 35, No. 5

247

Review

Trends in Pharmacological Sciences May 2014, Vol. 35, No. 5

been limited. The high attrition rate in preclinical and clinical trials, ascribed to, for example, toxicity, insufficient efficacy, or inadequate selectivity, is leading to an enormous increase in drug discovery costs [4]. In the meantime, the implementation of biologics as therapeutics has gained considerable attention (Box 1). Monoclonal antibodies (mAbs) are attractive therapeutics as they are highly selective and display extended half-lives compared with small molecules. With new and improved immunizationrelated technologies and advances in expression and/or purification methods of GPCRs, there are ongoing efforts directed at antibody-based targeting of GPCRs. Recently, mogamulizumab, a mAb targeting the chemokine receptor CCR4, was approved for treatment of adult T cell leukemia in Japan [5]. Currently, several therapeutic mAbs targeting GPCRs for the treatment of various indications (e.g., inflammation, oncology) are under investigation in clinical trials [3,6,7]. Although proven effective for various indications, mAbs are large (150 kDa) heteromultimeric glycoproteins, which are associated with the inability to target intracellular components and molecular cavities, limited drug administration routes, and high production costs. Nanobodies (Nbs) represent a potential novel class of antibody-based therapeutics with some favorable characteristics. They have already been successfully developed against several drug targets. Recently, we identified the first Nbs targeting GPCRs in vivo. Furthermore, the Nb platform has led to significant breakthroughs in basic research on GPCR structure and signaling. In this review, we will focus on the exciting progress in the field of GPCR-targeting Nbs. Unique characteristics and therapeutic potential of Nbs Discovery and advantages of Nbs In the early 1990s, Hamers-Casterman et al. discovered that camel sera contain a unique class of antibodies devoid of light chains and composed only of heavy chains [8]. Besides camels, other closely related members from the

Box 1. Conventional mAbs and fragments The immune system contains a diverse repertoire of tools to protect the host organism against harmful pathogens and xenobiotics. Antibodies, which are glycoprotein complexes secreted by B cells, are one of these tools that provide humoral (i.e., antibody-mediated) immunity by binding and subsequent neutralization of target molecules. The most abundantly expressed antibodies are the immunoglobulin-g (IgG) proteins, which are the origin of conventional mAbs. The overall structure of these 150 kDa molecules is very complex, consisting of four peptides made up from two identical heavy (H) and two identical light (L) chains together assembling in a tetrameric Y-shaped structure (see Figure 1 in main text). Both chains of conventional mAbs consist of several different functional domains. Whereas the heavy chains consist of one variable domain (VH) and three constant domains (CH1–3), the light chains consist of only two domains, one variable (VL) and one constant domain (CL). The constant domains CH1 and CL together with the variable domains comprise the two Fab regions, both able to bind one antigen each. The bivalent structure of an antibody is believed to contribute to increased avidity, thereby conferring high retention times. The remaining part of the antibody complex consisting of the CH2–CH3 homodimer is termed the Fc region [86]. The specific antigen-binding characteristics of a mAb are dependent on the complementarity-determining regions (CDRs) encoded by both variable domains (VH and VL), which together constitute the variable fragment (Fv). The constant domains in the Fc region are not directly involved in binding but are involved in mediating additional immune responses upon target binding. To improve the pharmacokinetics of mAbs, new formats of conventional antibody-based therapeutics were developed. Through removal of the Fc region and multimerization of the antigen-specific (VH and VL) domains, several types of antibodybased fragments of relatively small size were produced. The 55 kDa monovalent Fab fragments consist, as the name suggests, solely of one functional Fab region and thus also contain, besides the variable domains, the CH1 and CL constant domains. This is in contrast to the single chain Fv (scFv) fragments, which are the two variable domains directly connected by a peptide linker resulting in a 30 kDa size. Furthermore, several modulatory approaches are applied to these two functional antibody-based building blocks to generate multivalent constructs [34,87]. With the design of conventional IgG-based antibody fragments, tissue penetration is increased, whereas immunogenicity is decreased; however, this comes with the cost of increased blood clearance [88].

VH

∼2.5 nm

VL

VHH CH1 CL

∼15 nm CH2

CH2

CH3

CH3

Convenonal anbody 150 kDa

Heavy chain anbody 90 kDa

Fab fragment 50 kDa

VHH/Nanobody 15 kDa TRENDS in Pharmacological Sciences

Figure 1. Structural features of different antibodies. Conventional g-immunoglobulins consist of two heavy chains, each containing three invariant (constant) domains (CH1–CH3) and a single variable region (VH), and two light chains composed of a constant region (CL) and a variable region (VL). The VH and VL regions together form the paratope, and it is this antigen-recognizing portion of the antibody that is retained in Fab fragments. By contrast, camelid heavy chain antibodies are composed of a homodimer composed of two heavy chains containing two constant regions (i.e., they are lacking the CH1 domain) and a single variable (VHH) domain harboring three complementarity determining regions (rectangles). Nanobodies are the smallest functional single chain antigen-recognizing polypeptides known to date, with a molecular weight of 12–15 kDa and 1.5–2.5 nm dimensions.

248

Review

Box 2. Structural properties of hcAbs and Nbs In addition to conventional antibodies, camelids naturally express hcAbs. The camelid hcAb is a homodimer composed of H chains that individually fold in three domains, an N terminal variable domain (VHH) and two constant domains (CH2 and CH3) (see Figure 1 in main text). Thus, these distinctive antibodies lack the CH1 domain, which is present in conventional antibodies. Because the CH1 domain is involved with L chain interactions, it is believed that loss of this domain impairs L chain coupling. Owing to the lack of light chains, the antigen-binding properties of these distinctive immunoglobulins are dependent on a single 12–15 kDa variable domain of the heavy chain antibody (VHH) (see Figure 1 in main text). As a result of the simple structural nature of these VHHs, the recombinant equivalents of these natural domains, termed Nbs, retain full antigen-binding characteristics. Similar to a human VH domain, the camelid VHH domain is organized in three hypervariable complementarity-determining regions (CDR1–3) separated by four more conserved framework regions (FR1–4), folding in fourand five-stranded antiparallel b-sheets. The hypervariable CDRs are located within the loops of this structure and cluster at one side of the domain, hence forming a paratope [15,89]. Although the fold of a VHH domain largely overlaps with VH, there are some distinctive differences in CDRs and FRs, both contributing to the unique properties of the VHH domain. To support dimerization with VL, the FR2 of VH contains a number of conserved hydrophobic residues (Val37, Gly44, Leu45, and Trp47 according to Kabat numbering). A hallmark of VHHs is that these residues are often substituted to more hydrophilic or smaller residues, thus abrogating VL interaction and resulting in a strict monomeric antigen-binding state and increased solubility of the domain [14,15,90]. Other critical differences between VH and VHH lie within the CDRs. For example, the N terminal region of CDR1 tends to be more variable in VHH resulting in more conformational diversity of this loop [10]. Moreover, because CDRs of the VHH are not restricted to pair with another domain form a paratope, this allows the variable regions of VHHs to adopt a more flexible and often convex (i.e., protruding) conformation. Overall, the unique flexible paratope structure and small size enables VHHs (and Nbs as their recombinant equivalents) to bind molecular clefts and cavities [21,22].

Camelidae family (e.g., llamas and alpacas) naturally express the same type of 90 kDa heavy chain antibodies (hcAbs) (Figure 1). Interestingly, another class of evolutionary distinct hcAbs (immunoglobulin new antigen receptor, Ig-NAR) was discovered in cartilaginous fish [9–11]. The antigen-binding properties of hcAbs are solely dependent on a single 12–15 kDa variable domain of the heavy chain antibody (VHH) (Figure 1 and Box 2). The recombinant equivalents of these VHH domains, termed Nbs, retain full antigen-binding characteristics. The biochemical properties of Nbs provide advantages over conventional antibody-based molecules, thus facilitating many biotechnological and therapeutic applications. In particular, their small size, single domain nature, and hydrophilic characteristics contribute greatly to the excellent biochemical and pharmaceutical properties of Nbs, which include stability, production, affinity, and formatting. The straightforward and cost-effective production of Nbs is an important advantage compared with the high production costs of mAb-based biologics. Although microbial production of conventional antibody-based fragments can be achieved, their multidomain structure, glycosylation, and hydrophobic characteristics result in low yields and formation of aggregates. Fortunately, the production of Nbs is straightforward. The monomeric nature, absence of posttranslational modifications and high aqueous solubility

Trends in Pharmacological Sciences May 2014, Vol. 35, No. 5

contribute to the cost-effective production of Nbs in a variety of cell expression systems of bacterial, yeast, plant, or mammalian origin [12]. Nbs have low propensity for aggregation and are highly stable [13–17]. They retain full antigen-binding capacity at harsh thermal (60–808C) and chemical (2–3 M guanidinium chloride) denaturing conditions and furthermore display high refolding capability [16,17]. Moreover, the small proteins can be resistant to proteolytic degradation in the gastrointestinal tract, which potentially permits effective oral administration of Nbs [18]. Major strengths of antibody-based biologics, including Nbs, are their high affinity and specificity for their target. The affinity of Nbs is in the nanomolar–picomolar range, and as such they are highly suited for research, diagnostic, and therapeutic purposes [14,19–21]. The antigen-interacting surface of a conventional antibody has a flat or concave (i.e., hollow) shape, whereas Nbs can adopt a more flexible and often convex (i.e., protruding) conformation. Overall, the unique flexible paratope structure and small size enables Nbs to bind molecular clefts and cavities, for instance, receptor-binding pockets or enzyme active sites. This is supported by X-ray crystallography studies of Nbbound lysozyme or epidermal growth factor receptor (EGFR), revealing buried epitopes that are solely targeted by Nbs and not by other antibody-based molecules [21,22]. The small size also has a major impact on clearance. Because the Nb size is below the renal cut-off size of 50– 60 kDa, the proteins are rapidly cleared from the body. On average, the half-life of a Nb does not exceed 2 h, which is comparable to other antibody fragments, although a lot shorter than mAbs. A short half-life is ideal for a number of purposes, including stem cell mobilization, toxin delivery, and noninvasive in vivo imaging [23–25]. However, for chronic treatment, prolonged presence of Nbs is desired. This can be achieved by formatting the monomeric Nbs, generating multimeric constructs. Half-life extension methods used for conventional antibody fragments, such as PEGylation or fusion to Nbs targeting serum albumin can be used to tailor the half-life of Nbs from <2 h to several weeks [26–29]. This broadens their applicability as therapeutics for both acute and chronic treatments. The Nb platform allows a plug-and-play approach with multiple possibilities not only to tailor half-life but also to increase avidity and to create multispecific Nbs. Through coupling of two identical (bivalent) or distinct Nbs that recognize different epitopes on the same antigen (biparatopic) spectacular, up to 4000-fold, increases in avidity and potency compared with their monovalent counterparts are observed [19,20,27,30]. Diagnostics and therapeutic potential of Nbs Nbs are highly suitable as diagnostic tools and have been evaluated as biosensors and in vivo imaging tracers. Immunoimaging is a valuable tool for noninvasive visualization of biomarkers in vivo. Currently, radiolabeled biomarker-specific mAbs are used in the clinic to image tumors and for patient stratification. With their high affinity and specificity, proper tissue penetration and rapid clearance [26,31], Nbs provide advantages over mAbs and meet most characteristics of an ideal noninvasive molecular imaging probe [32–34]. An array of Nbs targeting the human epidermal 249

Review

Trends in Pharmacological Sciences May 2014, Vol. 35, No. 5

growth factor receptors 1 and 2 (HER1/HER2) [31,35,36], the macrophage mannose receptor [37], and the vascular cell adhesion molecule 1 (VCAM1) [38] have been successfully labeled with radionuclides for positron emission tomography (PET) and/or single photon emission computed tomography (SPECT) imaging purposes, to image tumors, inflammation, and atherosclerotic lesions, respectively. The preclinical results indicate that high signal-to-noise ratio images are obtained within a few hours after administration [31,38]. These results demonstrate the potential of Nb-linked tracers as a new class of diagnostic imaging tools. In addition, Nbs have been described that are able to traverse the blood–brain barrier (BBB) [39–41]. This is a very promising characteristic, because targeted transport of molecules across the BBB remains a major challenge in modern drug and diagnostics development. By coupling of BBB crossing Nbs to therapeutics or imaging probes, a window of opportunity is opened for brain tissue-targeted drug delivery and neuroimaging. Although no Nbs have found their way to the market yet, various humanized and sequence optimized Nbs have entered the clinical phases of drug research. Most advanced are Nbs targeting tumor necrosis factor (TNF)a [27] and interleukin-6 receptor (IL-6R) [42] for the treatment of rheumatoid arthritis (RA). Nbs against TNFa and IL-6R display excellent efficacy in in vitro models and

mouse models of RA and have successfully progressed into Phase IIb clinical testing. Nbs generated and characterized preclinically targeting receptor tyrosine kinases such as EGFR [43], hepatocyte growth factor receptor (HGFR/cMet) [28], and vascular endothelial growth factor receptor 2 (VEGFR2) [44] represent promising therapeutic candidates for the treatment of cancer. Owing to the high sequence similarity of the camelid VHH and the human VH framework regions, it is unlikely that Nbs will be highly immunogenic [14,15], which is supported by recent findings showing no immunogenicity in murine and human studies [15,27,45]. Additionally, to further reduce risk of immunogenicity, Vincke et al. formulated a general strategy for Nb humanization [46]. Hence, it seems unlikely that immunogenicity will pose a major concern in the development of Nbs for diagnostic or therapeutic applications. GPCR-targeting nanobodies Modulation of GPCR function by Nbs The first Nbs targeting GPCRs, namely chemokine receptors and their ligands, modulating GPCR function were recently described [19,20]. Chemokine receptors are expressed on immune cells and their main function is to guide leukocytes towards chemokine gradients upon inflammation [47]. This receptor family is implicated in

(B) MAbs Small molecules Nbs (A)

(C)

Signaling-deficient chemokine mutants

MAbs

HN

Heparin GAG-binding-deficient chemokine mutants

HN N

HN

N

NH

NH NH

Nbs

Small molecules

Glycosaminoglycan (GAG)

Chemokine Chemokine receptor

Plasma membrane Gαi TRENDS in Pharmacological Sciences

Figure 2. Potential therapeutic strategies for targeting of the chemokine system. (A) Much effort from both industrial and academic laboratories has centered around the development of antibody- and small molecule-based inhibition of chemokine receptors [6,47]. Additionally, several nanobodies have been described that inhibit chemokine binding to the CXCR4 and CXCR7 chemokine receptors [19,20]. Lastly, chemokine binding can be competitively inhibited by N terminal truncation mutants of chemokines that have retained chemokine receptor binding but are deficient in receptor activation [91]. (B) Although chemokine-binding antibodies and nanobodies form the mainstay of potential therapeutic options targeting chemokines [6,59], there are also examples of small molecules that bind CXCL12 and prevent binding to its cognate receptor CXCR4 [92,93]. (C) Finally, chemokine function can be modulated through disruption of chemokine gradient formation. Such gradients are pivotal for the directed migration of leukocytes and require the interaction of chemokines to cell-associated glycosaminoglycans (GAGs) [94]. Chemokine gradient formation can be altered by heparin (a highly sulfated soluble GAG subtype) [95] and by chemokine mutants that are defective in GAG binding, but can still bind and activate chemokine receptors [96–98].

250

Review

Trends in Pharmacological Sciences May 2014, Vol. 35, No. 5

several pathologies, such as inflammatory diseases, cancer, and HIV [48–50]. Targeting of these receptors and their ligands has been intensively investigated in the past years (Figure 2), leading to the FDA approval of two small molecule drugs specifically targeting this receptor family. Maraviroc, a CCR5 antagonist, is successfully used for several years now in the treatment of CCR5 tropic HIV1 infection, and Plerixafor (AMD3100), a CXCR4 antagonist, has recently been approved for hematopoietic stem cell mobilization [51,52]. One of the most-investigated chemokine–chemokine receptor pairs is the CXCL12/CXCR4 axis, because of its role in cancer and HIV entry [53,54]. Using a time-efficient whole cell immunization of llamas with intact CXCR4expressing HEK293T cells, followed by phage display and counterselection, Nbs against CXCR4 were identified [19]. The reported CXCR4–Nbs bind with high affinity and specificity to the extracellular loop 2 (ECL2) and potently inhibit CXCL12-induced signaling and chemotaxis (Figure 3A). Biparatopic CXCR4–Nbs, which have distinct and partially overlapping epitopes, show enhanced affinities and potencies (Figure 3A). Interestingly, these biparatopic Nbs act as inverse agonists on CXCR4, whereas AMD3100 acts a neutral antagonist [19]. Because CXCR4 is overexpressed in a large number of tumors [55], which is Key:

Monovalent CXCR4 Nb 1 Monovalent CXCR4 Nb 2 Biparatopic CXCR4 Nb (1+2)

(B)

100

120

CD34+ cells/ μl blood

Specific 125I–CXCL12 binding (%)

(A)

often associated with increased levels of basal activity, the use of inverse agonistic CXCR4 biparatopic Nbs could be beneficial. The two identified Nbs targeting CXCR4 are highly selective for human but not murine CXCR4, indicating that minor changes (five additional amino acids) in ECL2 are sufficient to affect Nb affinity for CXCR4. This underlines the exquisite selectivity that can be achieved by Nbs. Of therapeutic interest, CXCR4–Nbs were demonstrated to inhibit HIV-1 virus replication in vitro as well as to dose-dependently induce stem cell mobilization in cynomolgus monkeys, which underscores their therapeutic potential (Figure 3B) [19]. Similar to CXCR4, CXCR7 binds CXCL12 and plays an important role in proliferation, angiogenesis, and metastasis [56]. However, CXCR7 is an atypical chemokine receptor as it does not signal through G proteins but specifically activates signaling pathways through recruitment of b-arrestin [57,58]. Specific Nbs binding to this receptor were recently developed and characterized [20]. CXCR7–Nbs display high affinity and antagonistic properties, as they are able to inhibit CXCL12-induced recruitment of b-arrestin 2 to CXCR7 [20]. Moreover, CXCR7– Nbs were shown to inhibit head and neck cancer tumor growth in a mouse xenograft model (Figure 3C) through inhibition of angiogenesis [20]. The use of CXCR7–Nbs

100 80 60 40 20 0

Key:

80 60 40 20 0

–12 –11 –10

–9

–8

–7

6

0

–6

log [Nb]

(D)

PBS Biparatopic CXCR7 Nb

200

*

150 100 50 0 –10

Key:

120

Migrated cells (%)

250

12

18

24

Time post–injecon (h)

Key:

(C)

Tumor size (mm3)

AMD3100 Biparatopic CXCR4 Nb

Monovalent CXCL12 Nb

100 80 60 40 20 0

0

10

20

30

40

Treatment duraon (days)

50

–10

–9

–8

–7

–6

log [Nb] TRENDS in Pharmacological Sciences

Figure 3. Nanobodies targeting the chemokine system effectively inhibit chemokine function in vitro and in vivo. (A) A biparatopic nanobody (Nb) targeting the CXCR4 chemokine receptor displays increased binding affinity in a CXCL12 radioligand displacement assay compared with the individual monovalent Nbs from which it is composed. Adapted from [19]. (B) The biparatopic CXCR4-targeting Nb induces the mobilization of CD34-positive stem cells in the blood of macaque monkeys to a similar extent as the small molecule CXCR4 antagonist AMD3100. Adapted from [19]. (C) A biparatopic CXCR7-targeting Nb inhibits growth in mice xenografted with the head-andneck cancer cell line UM-SCC-22A. Adapted from [20]. (D) A CXCL12-binding Nb concentration-dependently inhibits CXCL12-induced chemotaxis of murine pre-B L1.2 cells in a transwell migration assay. Adapted from [59].

251

Review might thus be of therapeutic interest in this cancer type, but also other cancers showing high expression of CXCR7 might be effectively targeted by CXCR7–Nbs. Finally, generation of Nbs targeting chemokines have also been recently reported [59]. Inflammatory chemokines CCL2 (binding to CCR2), CCL5 (binding to CCR1, CCR3, and CCR5), and CXCL11 (binding to CXCR3 and CXCR7), as well as the homeostatic chemokine CXCL12 (binding to CXCR4 and CXCR7), were successfully targeted by chemokine-specific Nbs. All these Nbs inhibit chemokine binding to their cognate chemokine receptor (Figure 2). Furthermore, the CXCL11- and CXCL12-specific Nbs inhibit Gai signaling and chemotaxis through CXCR3 and CXCR4, respectively (Figure 3D). All chemokine–chemokine receptor pairs mentioned above have been implicated in cancer [49,60,61]. Additionally, inflammatory diseases such as RA [62,63] and multiple sclerosis (MS) [64,65] are dependent on CXCL11–CXCR3- and CCL2–CCR2-induced cellular signaling. The chemokine-targeting Nbs might prove therapeutically beneficial for these indications. Collectively, Nbs against CXCR4 and CXCR7, as well as against a panel of chemokines, show great promise as therapeutics and research tools to further characterize the chemokine receptor system. GPCR-targeting Nbs as research tools Although Nbs have received a great deal of attention for their therapeutic potential, their use as biomolecular tools in basic GPCR research is also gaining momentum. One application of Nbs in biomolecular research is their use as crystallization chaperones. A major hurdle for successful crystallization of GPCRs has been their hydrophobic and thermolabile nature. Through in-frame insertion of hydrophilic crystallization chaperones (such as T4 lysozyme), addition of high-affinity inverse agonists, or antagonists, with long residence times or through mutations that increase thermal stability, several structures of GPCRs in their inactive state have now been published [66–69]. Crystallization of the agonist-bound active state of a GPCR has provided an even bigger challenge, because this receptor conformation appears to be intrinsically less stable than the inactive state [70]. Crystallization of such energetically unfavorable protein conformations has been achieved through co-crystallization with antibodies, antibody fragments, or Nbs [71]. Indeed, co-crystallization of the b2 adrenergic receptor (b2AR) and more recently the M2 muscarinic acetylcholine receptor (M2R) with Nbs that specifically recognize the active receptor conformation led to the unveiling of the first active state GPCR structures [72,73]. These GPCR-targeting Nbs act as so-called ‘G protein mimetics’, enhancing agonist affinities indicating that they stabilize active states of the receptor. In addition, a Nb directed against the Gas heterotrimer was used to stabilize a co-crystal composed of the active b2AR in complex with its G protein [74]. Another important asset of GPCR-targeting Nbs in GPCR research is their use as conformation-specific biosensors. Fluorescently tagged Nbs provide a means of tracking proteins in live cells with unprecedented resolution [75,76]. Recent efforts [77] harnessed this knowledge to study the spatial localization of b2AR signaling using Nbs that recognized activated b2AR [72] and nucleotide-free (active) Gas 252

Trends in Pharmacological Sciences May 2014, Vol. 35, No. 5

[78], respectively. By genetically fusing these Nbs to GFP and expressing them in b2AR-expressing cells, followed by microscopic examination of GFP localization following b2AR activation, direct evidence for G protein activation from internalized GPCRs was provided [77], a phenomenon that has been suggested to occur for multiple GPCRs but was never directly ascertained [79–81]. In addition, these intracellularly expressed Nbs, also termed intrabodies, can also be utilized to modulate receptor-dependent signaling. A recent study showed that conformational-specific b2AR intrabodies inhibit downstream signal transduction, including G protein activation, G protein-coupled receptor kinase (GRK)-mediated receptor phosphorylation, b-arrestin recruitment, and receptor internalization [82]. These findings illustrate the potential of Nbs as research tools to study receptor-specific GPCR signaling. Future perspectives Conformation-selective Nbs Conformation-selective Nbs, that target the intracellular domain of GPCRs, recognize and stabilize conformationspecific states, such as the active state of b2AR and M2R [72,73]. Nbs that target the extracellular domain of GPCRs and effectively inhibit GPCR function bind to the ECL epitopes, in particular ECL2 [19,20], indicating that these Nbs also recognize conformation-dependent structures. Immunization of camelids with intact cells or membranes overexpressing GPCRs allow recognition of GPCR extracellular domains in their native conformation with appropriate post-translational modifications. This is of particular interest in view of the emerging concept that one can selectively modulate GPCR-induced signaling pathways by means of biased ligands [1]. By binding to distinct GPCR conformations, biased ligands may selectively modulate downstream signaling pathways. Biased ligands may prove to be superior as therapeutics, minimizing activation of pathways that cause adverse side effects in some diseases [1]. To obtain biased conformation-specific Nbs, one can use genetically engineered GPCRs favoring (in)activation of a given signal transduction pathway [e.g., constitutively (in)active GPCRs, stabilized receptors in specific state (StaRs)] overexpressed in cells or viral particles, or purified GPCRs (e.g., in nanodiscs) in the presence of high-affinity compounds with enhanced residence time (slow off-rate) in immunization and/or selection strategies [70]. These biased conformation-selective Nbs provide unique opportunities to selectively modulate and examine specific GPCR signaling pathways. In addition, the Nbs targeting the b2AR–Gs complex have proven to be excellent biosensors to monitor signaling of GPCRs from endosomes with high spatiotemporal resolution in living cells [77]. The nucleotide-free Gas-binding Nb may be a generic tool to study signaling through any Gscoupled GPCR [77]. Similar Nb-based conformation-specific biosensors are likely to be pivotal tools for the advancement of GPCR research. Multimerization Multimeric Nb constructs can be readily generated by linking multiple monovalent Nbs. This is an appealing

Review procedure, because bivalent and biparatopic Nbs targeting the extracellular domain of GPCRs show enhanced affinities and potencies compared with monovalent Nbs [19,20]. It is tempting to speculate that these linked Nbs target GPCR dimers, although additional experiments are required to further substantiate this. The generation of multispecific constructs offers a potential solution to multifactorial diseases where targeting multiple targets is therapeutically beneficial. For example, there are instances where targeting GPCR heterodimers may prove a viable therapeutic option [83]. With increased awareness of the role of GPCRs in, for example, cancer [84], the use of multimeric Nbs targeting and inhibiting both GPCR and RTK signaling networks that dictate proliferation, metastasis, or angiogenesis at once may be favorable. Diagnostics and therapeutics In view of their high affinity and selectivity, GPCR-targeting Nbs may serve as excellent tools as diagnostics and, potentially, as therapeutics. With advances in microscope and imaging technologies, GPCR-targeting Nbs may be ideal to monitor GPCR expression at high resolution and in vivo, respectively. As diagnostics and as modulators of GPCR function, GPCR-targeting Nbs will serve as important tools to validate the role of GPCRs in pathology. GPCR-targeting Nbs could also be used as targeting devices for toxic enzymes or engineered to incorporate effector functions such as antibody-dependent cellular cytotoxicity (ADCC) to eradicate tumors. For instance, the mAb targeting CCR4 in the treatment of T cell leukemia has been effectively altered by glycoengineering (defucosylation/Potelligent1 technology) to increase its ability to induce ADCC [85]. Concluding remarks Taken together, the aforementioned studies clearly demonstrate the potential value of GPCR-targeting Nbs. They combine the advantages of both small molecules (e.g., cavity binding, low production costs) and monoclonal antibodies (e.g., high affinity and specificity). GPCR-targeting Nbs play a prominent role as crystallization chaperones and G protein mimics to obtain structural information on the active state of GPCRs and use as conformation-specific biosensors. In view of their high affinity and selectivity, GPCRtargeting Nbs serve as attractive tools as diagnostics and, potentially, as therapeutics. They can modulate GPCR function, through binding to extracellular domains or intracellular domains. Identification of biased conformationselective Nbs will allow selective modulation of specific GPCR signaling pathways. Through engineering methods, Nbs can be formatted to increase their potencies, target multiple GPCRs (distinct or homo/heterodimers), or tailor their half-life, broadening their applicability as therapeutics for both acute and chronic treatments. Hence, the emergence of this novel class of antibodies as high-quality research tools, diagnostics, and therapeutics is likely to herald exciting advancements in GPCR research. Acknowledgments The authors are supported by The Netherlands Organization for Scientific Research (NWO) and the Dutch Technology Foundation STW. M.J.S participates in the European COST action CM1207 (GLISTEN).

Trends in Pharmacological Sciences May 2014, Vol. 35, No. 5

References 1 Rajagopal, S. et al. (2010) Teaching old receptors new tricks: biasing seven-transmembrane receptors. Nat. Rev. Drug Discov. 9, 373–386 2 Overington, J.P. et al. (2006) How many drug targets are there? Nat. Rev. Drug Discov. 5, 993–996 3 Hutchings, C.J. et al. (2010) Therapeutic antibodies directed at G protein-coupled receptors. MAbs 2, 594–606 4 Kola, I. and Landis, J. (2004) Can the pharmaceutical industry reduce attrition rates? Nat. Rev. Drug Discov. 3, 711–715 5 Beck, A. and Reichert, J.M. (2012) Marketing approval of mogamulizumab: a triumph for glyco-engineering. MAbs 4, 419–425 6 Klarenbeek, A. et al. (2012) Targeting chemokines and chemokine receptors with antibodies. Drug Discov. Today Technol. 9, e237–e244 7 Webb, D.R. et al. (2013) Opportunities for functional selectivity in GPCR antibodies. Biochem. Pharmacol. 85, 147–152 8 Hamers-Casterman, C. et al. (1993) Naturally occurring antibodies devoid of light chains. Nature 363, 446–448 9 Greenberg, A.S. et al. (1995) A new antigen receptor gene family that undergoes rearrangement and extensive somatic diversification in sharks. Nature 374, 168–173 10 Conrath, K.E. et al. (2003) Emergence and evolution of functional heavy-chain antibodies in Camelidae. Dev. Comp. Immunol. 27, 87– 103 11 Flajnik, M.F. et al. (2011) A case of convergence: why did a simple alternative to canonical antibodies arise in sharks and camels? PLoS Biol. 9, e1001120 12 Frenzel, A. et al. (2013) Expression of recombinant antibodies. Front. Immunol. 4, 217 13 Arbabi-Ghahroudi, M. et al. (2005) Prokaryotic expression of antibodies. Cancer Metastasis Rev. 24, 501–519 14 Harmsen, M.M. and De Haard, H.J. (2007) Properties, production, and applications of camelid single-domain antibody fragments. Appl. Microbiol. Biotechnol. 77, 13–22 15 Muyldermans, S. (2013) Nanobodies: natural single-domain antibodies. Annu. Rev. Biochem. 82, 775–797 16 Dumoulin, M. et al. (2002) Single-domain antibody fragments with high conformational stability. Protein Sci. 11, 500–515 17 Ewert, S. et al. (2002) Biophysical properties of camelid V(HH) domains compared to those of human V(H)3 domains. Biochemistry 41, 3628– 3636 18 Harmsen, M.M. et al. (2006) Selection and optimization of proteolytically stable llama single-domain antibody fragments for oral immunotherapy. Appl. Microbiol. Biotechnol. 72, 544–551 19 Jahnichen, S. et al. (2010) CXCR4 nanobodies (VHH-based single variable domains) potently inhibit chemotaxis and HIV-1 replication and mobilize stem cells. Proc. Natl. Acad. Sci. U.S.A. 107, 20565–20570 20 Maussang, D. et al. (2013) Llama-derived single variable domains (nanobodies) directed against CXCR7 reduce head and neck cancer cell growth in vivo. J. Biol. Chem. 288, 29562–29572 21 Schmitz, K.R. et al. (2013) Structural evaluation of EGFR inhibition mechanisms for nanobodies/VHH domains. Structure 21, 1214–1224 22 De Genst, E. et al. (2006) Molecular basis for the preferential cleft recognition by dromedary heavy-chain antibodies. Proc. Natl. Acad. Sci. U.S.A. 103, 4586–4591 23 Cortez-Retamozo, V. et al. (2004) Efficient cancer therapy with a nanobody-based conjugate. Cancer Res. 64, 2853–2857 24 Harrison, R.A. et al. (2006) Neutralisation of venom-induced haemorrhage by IgG from camels and llamas immunised with viper venom and also by endogenous, non-IgG components in camelid sera. Toxicon 47, 364–368 25 Movahedi, K. et al. (2012) Nanobody-based targeting of the macrophage mannose receptor for effective in vivo imaging of tumor-associated macrophages. Cancer Res. 72, 4165–4177 26 Tijink, B.M. et al. (2008) Improved tumor targeting of anti-epidermal growth factor receptor nanobodies through albumin binding: taking advantage of modular nanobody technology. Mol. Cancer Ther. 7, 2288– 2297 27 Coppieters, K. et al. (2006) Formatted anti-tumor necrosis factor alpha VHH proteins derived from camelids show superior potency and targeting to inflamed joints in a murine model of collagen-induced arthritis. Arthritis Rheum. 54, 1856–1866 28 Vosjan, M.J. et al. (2012) Nanobodies targeting the hepatocyte growth factor: potential new drugs for molecular cancer therapy. Mol. Cancer Ther. 11, 1017–1025 253

Review 29 Schmidt Slordahl, T. et al. (2013) Anti-c-MET nanobody1 – a new potential drug in multiple myeloma treatment. Eur. J. Haematol. 91, 399–410 30 Kolkman, J.A. and Law, D.A. (2010) Nanobodies – from llamas to therapeutic proteins. Drug Discov. Today Technol. 7, e95–e146 31 Xavier, C. et al. (2013) Synthesis, preclinical validation, dosimetry, and toxicity of 68Ga-NOTA-anti-HER2 nanobodies for iPET imaging of HER2 receptor expression in cancer. J. Nucl. Med. 54, 776–784 32 Vaneycken, I. et al. (2011) Immuno-imaging using nanobodies. Curr. Opin. Biotechnol. 22, 877–881 33 Oliveira, S. et al. (2013) Targeting tumors with nanobodies for cancer imaging and therapy. J. Control. Release 172, 607–617 34 Cuesta, A.M. et al. (2010) Multivalent antibodies: when design surpasses evolution. Trends Biotechnol. 28, 355–362 35 Vaneycken, I. et al. (2011) Preclinical screening of anti-HER2 nanobodies for molecular imaging of breast cancer. FASEB J. 25, 2433–2446 36 Vosjan, M.J. et al. (2011) Facile labelling of an anti-epidermal growth factor receptor nanobody with 68Ga via a novel bifunctional desferal chelate for immuno-PET. Eur. J. Nucl. Med. Mol. Imaging 38, 753–763 37 Put, S. et al. (2013) SPECT imaging of joint inflammation with nanobodies targeting the macrophage mannose receptor in a mouse model for rheumatoid arthritis. J. Nucl. Med. 54, 807–814 38 Broisat, A. et al. (2012) Nanobodies targeting mouse/human VCAM1 for the nuclear imaging of atherosclerotic lesions. Circ. Res. 110, 927– 937 39 Li, T. et al. (2012) Cell-penetrating anti-GFAP VHH and corresponding fluorescent fusion protein VHH–GFP spontaneously cross the blood– brain barrier and specifically recognize astrocytes: application to brain imaging. FASEB J. 26, 3969–3979 40 Rutgers, K.S. et al. (2011) Transmigration of b amyloid specific heavy chain antibody fragments across the in vitro blood–brain barrier. Neuroscience 190, 37–42 41 Haqqani, A.S. et al. (2012) Multiplexed evaluation of serum and CSF pharmacokinetics of brain-targeting single-domain antibodies using a NanoLC–SRM-ILIS method. Mol. Pharm. 10, 1542–1556 42 Mima, T. and Nishimoto, N. (2009) Clinical value of blocking IL-6 receptor. Curr. Opin. Rheumatol. 21, 224–230 43 Roovers, R.C. et al. (2011) A biparatopic anti-EGFR nanobody efficiently inhibits solid tumour growth. Int. J. Cancer 129, 2013–2024 44 Behdani, M. et al. (2012) Generation and characterization of a functional nanobody against the vascular endothelial growth factor receptor-2; angiogenesis cell receptor. Mol. Immunol. 50, 35–41 45 Sarker, S.A. et al. (2013) Anti-rotavirus protein reduces stool output in infants with diarrhea: a randomized placebo-controlled trial. Gastroenterology 145, 740–748 46 Vincke, C. et al. (2009) General strategy to humanize a camelid singledomain antibody and identification of a universal humanized nanobody scaffold. J. Biol. Chem. 284, 3273–3284 47 Scholten, D.J. et al. (2012) Pharmacological modulation of chemokine receptor function. Br. J. Pharmacol. 165, 1617–1643 48 Chatterjee, A. et al. (2012) Chemokines and chemokine receptors in susceptibility to HIV-1 infection and progression to AIDS. Dis. Markers 32, 143–151 49 Vandercappellen, J. et al. (2008) The role of CXC chemokines and their receptors in cancer. Cancer Lett. 267, 226–244 50 White, G.E. et al. (2013) CC chemokine receptors and chronic inflammation –therapeutic opportunities and pharmacological challenges. Pharmacol. Rev. 65, 47–89 51 Perry, C.M. (2010) Maraviroc: a review of its use in the management of CCR5-tropic HIV-1 infection. Drugs 70, 1189–1213 52 DiPersio, J.F. et al. (2009) Plerixafor. Nat. Rev. Drug Discov. 8, 105–106 53 Furusato, B. et al. (2010) CXCR4 and cancer. Pathol. Int. 60, 497–505 54 Weiss, R.A. (2001) Gulliver’s travels in HIVland. Nature 410, 963–967 55 Zlotnik, A. et al. (2011) Homeostatic chemokine receptors and organspecific metastasis. Nat. Rev. Immunol. 11, 597–606 56 Duda, D.G. et al. (2011) CXCL12 (SDF1a)–CXCR4/CXCR7 pathway inhibition: an emerging sensitizer for anticancer therapies? Clin. Cancer Res. 17, 2074–2080 57 Burns, J.M. et al. (2006) A novel chemokine receptor for SDF-1 and ITAC involved in cell survival, cell adhesion, and tumor development. J. Exp. Med. 203, 2201–2213

254

Trends in Pharmacological Sciences May 2014, Vol. 35, No. 5

58 Rajagopal, S. et al. (2010) Beta-arrestin- but not G protein-mediated signaling by the ‘‘decoy’’ receptor CXCR7. Proc. Natl. Acad. Sci. U.S.A. 107, 628–632 59 Blanchetot, C. et al. (2013) Neutralizing nanobodies targeting diverse chemokines effectively inhibit chemokine function. J. Biol. Chem. 288, 25173–25182 60 Craig, M.J. and Loberg, R.D. (2006) CCL2 (monocyte chemoattractant protein-1) in cancer bone metastases. Cancer Metastasis Rev. 25, 611– 619 61 Allen, S.J. et al. (2007) Chemokine: receptor structure, interactions, and antagonism. Annu. Rev. Immunol. 25, 787–820 62 Koch, A.E. et al. (1992) Enhanced production of monocyte chemoattractant protein-1 in rheumatoid arthritis. J. Clin. Invest. 90, 772–779 63 Qin, S. et al. (1998) The chemokine receptors CXCR3 and CCR5 mark subsets of T cells associated with certain inflammatory reactions. J. Clin. Invest. 101, 746–754 64 Mahad, D.J. and Ransohoff, R.M. (2003) The role of MCP-1 (CCL2) and CCR2 in multiple sclerosis and experimental autoimmune encephalomyelitis (EAE). Semin. Immunol. 15, 23–32 65 Sorensen, T.L. et al. (1999) Expression of specific chemokines and chemokine receptors in the central nervous system of multiple sclerosis patients. J. Clin. Invest. 103, 807–815 66 Rasmussen, S.G. et al. (2007) Crystal structure of the human b2 adrenergic G-protein-coupled receptor. Nature 450, 383–387 67 Warne, T. et al. (2008) Structure of a b1-adrenergic G-protein-coupled receptor. Nature 454, 486–491 68 Shimamura, T. et al. (2011) Structure of the human histamine H1 receptor complex with doxepin. Nature 475, 65–70 69 Hollenstein, K. et al. (2013) Structure of class B GPCR corticotropinreleasing factor receptor 1. Nature 499, 438–443 70 Steyaert, J. and Kobilka, B.K. (2011) Nanobody stabilization of G protein-coupled receptor conformational states. Curr. Opin. Struct. Biol. 21, 567–572 71 Koide, S. (2009) Engineering of recombinant crystallization chaperones. Curr. Opin. Struct. Biol. 19, 449–457 72 Rasmussen, S.G. et al. (2011) Structure of a nanobody-stabilized active state of the b2 adrenoceptor. Nature 469, 175–180 73 Kruse, A.C. et al. (2013) Activation and allosteric modulation of a muscarinic acetylcholine receptor. Nature 504, 101–106 74 Rasmussen, S.G. et al. (2011) Crystal structure of the b2 adrenergic receptor–Gs protein complex. Nature 477, 549–555 75 Ries, J. et al. (2012) A simple, versatile method for GFP-based superresolution microscopy via nanobodies. Nat. Methods 9, 582–584 76 Rothbauer, U. et al. (2006) Targeting and tracing antigens in live cells with fluorescent nanobodies. Nat. Methods 3, 887–889 77 Irannejad, R. et al. (2013) Conformational biosensors reveal GPCR signalling from endosomes. Nature 495, 534–538 78 Westfield, G.H. et al. (2011) Structural flexibility of the Gas a-helical domain in the b2-adrenoceptor Gs complex. Proc. Natl. Acad. Sci. U.S.A. 108, 16086–16091 79 Ferrandon, S. et al. (2009) Sustained cyclic AMP production by parathyroid hormone receptor endocytosis. Nat. Chem. Biol. 5, 734– 742 80 Calebiro, D. et al. (2010) Imaging of persistent cAMP signaling by internalized G protein-coupled receptors. J. Mol. Endocrinol. 45, 1–8 81 Mullershausen, F. et al. (2009) Persistent signaling induced by FTY720-phosphate is mediated by internalized S1P1 receptors. Nat. Chem. Biol. 5, 428–434 82 Staus, D.P. et al. (2014) Regulation of b2-adrenergic receptor function by conformationally selective single-domain intrabodies. Mol. Pharmacol. 85, 472–481 83 Panetta, R. and Greenwood, M.T. (2008) Physiological relevance of GPCR oligomerization and its impact on drug discovery. Drug Discov. Today 13, 1059–1066 84 O’Hayre, M. et al. (2013) The emerging mutational landscape of G proteins and G-protein-coupled receptors in cancer. Nat. Rev. Cancer 13, 412–424 85 Ishida, T. et al. (2012) Defucosylated anti-CCR4 monoclonal antibody (KW-0761) for relapsed adult T-cell leukemia-lymphoma: a multicenter phase II study. J. Clin. Oncol. 30, 837–842 86 George, A.J. (2000) The antibody molecule. Methods Mol. Med. 40, 1–21

Review 87 Chan, A.C. and Carter, P.J. (2010) Therapeutic antibodies for autoimmunity and inflammation. Nat. Rev. Immunol. 10, 301–316 88 Sanz, L. et al. (2004) Antibodies and gene therapy: teaching old ‘magic bullets’ new tricks. Trends Immunol. 25, 85–91 89 Chothia, C. et al. (1992) Structural repertoire of the human VH segments. J. Mol. Biol. 227, 799–817 90 Vu, K.B. et al. (1997) Comparison of llama VH sequences from conventional and heavy chain antibodies. Mol. Immunol. 34, 1121– 1131 91 Mortier, A. et al. (2008) Regulation of chemokine activity by posttranslational modification. Pharmacol. Ther. 120, 197–217 92 Ziarek, J.J. et al. (2012) Fragment-based optimization of small molecule CXCL12 inhibitors for antagonizing the CXCL12/CXCR4 interaction. Curr. Top. Med. Chem. 12, 2727–2740 93 Hachet-Haas, M. et al. (2008) Small neutralizing molecules to inhibit actions of the chemokine CXCL12. J. Biol. Chem. 283, 23189–23199

Trends in Pharmacological Sciences May 2014, Vol. 35, No. 5

94 Proudfoot, A.E. et al. (2003) Glycosaminoglycan binding and oligomerization are essential for the in vivo activity of certain chemokines. Proc. Natl. Acad. Sci. U.S.A. 100, 1885–1890 95 Ziarek, J.J. et al. (2013) Heparin oligosaccharides inhibit chemokine (CXC motif) ligand 12 (CXCL12) cardioprotection by binding orthogonal to the dimerization interface, promoting oligomerization, and competing with the chemokine (CXC motif) receptor 4 (CXCR4) N terminus. J. Biol. Chem. 288, 737–746 96 Johnson, Z. et al. (2004) Interference with heparin binding and oligomerization creates a novel anti-inflammatory strategy targeting the chemokine system. J. Immunol. 173, 5776–5785 97 Severin, I.C. et al. (2010) Characterization of the chemokine CXCL11– heparin interaction suggests two different affinities for glycosaminoglycans. J. Biol. Chem. 285, 17713–17724 98 Nellen, A. et al. (2012) Interference with oligomerization and glycosaminoglycan binding of the chemokine CCL5 improves experimental liver injury. PLoS ONE 7, e36614

255