Co-Opting Host Receptors for Targeted Delivery of Bioconjugates—From Drugs to Bugs

Bioconjugation has allowed scientists to combine multiple functional elements into one biological or biochemical unit. This assembly can result in the production of constructs that are targeted to a specific site or cell type in order to enhance the response to, or activity of, the conjugated moiety. In the case of cancer treatments, selectively targeting chemotherapies to the cells of interest limit harmful side effects and enhance efficacy. Targeting through conjugation is also advantageous in delivering treatments to difficult-to-reach tissues, such as the brain or infections deep in the lung. Bacterial infections can be more selectively treated by conjugating antibiotics to microbe-specific entities; helping to avoid antibiotic resistance across commensal bacterial species. In the case of vaccine development, conjugation is used to enhance efficacy without compromising safety. In this work, we will review the previously mentioned areas in which bioconjugation has created new possibilities and advanced treatments.


Introduction
The process of conjugation has allowed scientists to improve therapies and vaccinations by linking therapeutic agents or protective antigens to targeting moieties that focus delivery of the cargo. Increasing the specificity of therapies promotes concentration of the agent which consequently reduces the needed systemic dose and minimizes side effects. Targeting of prophylactics and therapies can be achieved by fusing distinct functional moieties; a process often termed bioconjugation. Common methods of bioconjugation include genetic fusion, chemical conjugation, and in vivo bioconjugation; examples of each will be discussed in this review. Bioconjugation is broadly utilized from basic preclinical development to current clinical use and impacts medical disciplines from cancer, to neuropathies, to infectious disease.
In the case of treating cancer patients, systemic administration of chemotherapy regularly causes deleterious side effects and has limited efficacy; however, great strides have been made in limiting side effects and enhancing efficacy by specifically targeting therapies. Through targeted drug delivery, drugs can be concentrated in the desired tissue and specific anti-tumor cells can be activated [1][2][3][4][5]. The targeting concept has also been applied to neurodegenerative diseases with the goal of suppressing pathological immune responses and/or promoting repair pathways [6,7]. The ability to target antibiotics is a promising development as drug resistant pathogens have become increasingly prevalent and indiscriminate application of antibiotics promotes the development and spread of antibiotic resistance [8,9]. Further, conjugated vaccines have also been developed for many pathogenic microbes; the importance of the latter is glaringly highlighted by the COVID-19 pandemic. the rapid bioorthogonal removal of a 3-isocyanopropyl masking group through reaction with tetrazines [31][32][33]. Collectively, this method of selected drug release reduces side effects on healthy cells and increases potency of anti-cancer therapies.

Central Nervous System Therapies
Alzheimer's disease is the most common form of dementia in the United States, where there are currently over 5 million people living with the disease [53]. Brain pathology characteristically involves the development of amyloid beta plaques, aggregated tau protein, neurofibrillary tangles, and a variety of neurological defects. As researchers have discovered more about Alzheimer's disease, they have also unveiled promising targets for therapies [6]. However, a major limitation on delivery of neurodegenerative therapies is bioavailability and crossing the blood-brain-barrier (BBB). Fan et al. developed a braintargeted nanoparticle (NP) that increases the delivery of curcumin, an anti-inflammatory compound [36]. Targeting of the curcumin-loaded NPs to the brain was achieved through conjugation of the particles to B6 peptides. B6 is a short peptide that binds the transferrin receptor (TfR) which is expressed on endothelial cells of the BBB and facilitates transcytosis; targeting to this receptor increases BBB-permeability and availability of B6-conjugates in the brain [36,[54][55][56]. Administration of the B6-conjugated curcumin-loaded NP to mice improved their memory and spatial learning which has promise to extend to treatment of humans with neurodegenerative diseases [36].
Parkinson's disease is another common neurodegenerative disorder that is characterized by the slowing of movements and the presence of a resting tremor [7,57]. As with Alzheimer's disease, delivery of therapeutics for Parkinson's disease also requires crossing the BBB. Rusiecka et al. enhanced the delivery of dopamine to the brain through conjugation to the cell-penetrating peptide TP10 [37] which is derived from wasp venom [58]. This heightened cell permeability improves delivery of TP10-conjugates to the brain and in the case of dopamine, delivers it in a form that has a high affinity for its receptor [37]. It should be noted that the lipophilic peptides such as TP10 are not inherently BBB specific and additional BBB-specific targeting modalities might enhance the specificity of this approach.
Additional examples of targeted CNS therapies include those focused on demyelination (a hallmark of multiple sclerosis-MS) and post-stroke, neuro-regeneration [38,39]. The aptamer LJM-3064 is a small (40 nt,~13 kDa) single stranded DNA molecule selected in a screen for myelein-binding activity [59]. Tetramers of LJM-3064 administered to mice in a viral encephalitis model displayed significant remyelination of the spinal cord. The results indicate that the myelin-binding aptamer, when multimerized, has inherent promyelinogenic properties. LJM-3064 has subsequently been used to decorate the surfaces of anti-inflammatory exosomes (derived from mesenchymal stem cells). The aptamerexosome bioconjugate reduced both CNS inflammation and demyelination in a murine experimental autoimmune encephalomyelitis (EAE) model of MS [38].
Neuro-regeneration by adult neural stem cells holds tremendous medical potential but has been limited, particularly in the context of ischemic stroke, by poor survival of migrating neuroblasts. In attempt to activate pro-survival pathways in these cells, the Akt-activating drug SC-79 has been encapsulated in NP decorated with Ab specific for a neural cell adhesion molecule (PSA-NECM). When the drug-loaded conjugates (α-PSA-NECM-NP-SC-79) were instilled into murine lateral ventricles, the authors noted activation of Akt in neuroblasts within in the subventricular zone [39]. Collectively, the ability to target conjugates for enhanced passage across the BBB and/or to specific sites of pathology shows promise for enhancing treatment of brain-specific diseases.

Antimicrobials
Since the discovery of penicillin by Alexander Fleming in the early 1900's, antibiotics have been used to treat countless infections and have greatly improved medical care. However, emerging bacterial resistance to antibiotics is widely recognized and has hindered the effective use of many compounds [60,61]. The emergence of new and more extensively drug-resistant bacteria have encouraged the discovery of novel antibiotic compounds as well as targeted delivery of existing antibiotics which promotes more direct and concentrated administration of a drug [62]. By focusing delivery of the drug to the pathogen which is causing disease, the exposure of commensal microbes to antibiotic agents is re-duced. This is important to prevent development of antibiotic resistance; in many cases the antibiotic activity of a drug is broad and sub-lethal exposure permits bacteria to enhance their drug resistance mechanisms. Therefore, targeting an antibiotic to a specific site or pathogen can both reduce generation of drug resistant microbes while leaving endogenous commensal ("good") bacterial populations intact.
One method of targeting delivery of antibiotics and limiting off-target effects is to conjugate the antibiotic to a pathogen-specific Ab. Lehar et al. conjugated the antibiotic rifalogue to an Ab specific for MRSA (methicillin-resistant Staphylococcus aureus) and showed increased killing of MRSA in a murine bacteremia model. In this case, the antibiotic was not active until the conjugate-opsonized bacteria were engulfed in a phagosome, it was then activated by intracellular proteases and able to kill the bacteria within the host cell [40]. This method not only concentrates the antibiotic at the location of the pathogen, but also reduces exposure of commensal bacteria to the drug which would have the potential to induce antibiotic resistance mechanisms.
Similarly, targeted NPs and virus-like particles (VLPs) have also been used to concentrate antibiotic delivery. NPs enhance bioavailability, solubility, stability, and controlled release of different substances. In addition, due to their size, NPs are taken up by phagocytic mechanisms versus the passive entry of smaller molecules such as free drug [63,64]. Antibiotic-decorated particles have been shown to increase the amount of antibiotic internalized by lung-resident macrophages; therefore, concentrating the drug and potentially increasing effectiveness of treatments against stubborn pulmonary bacterial infections [65]. Gold NPs have also been utilized to enhance efficacy of antibiotics and reduce the necessary dose [66][67][68]. Work by Pornpattananangkul et al. utilized antibiotic-loaded gold NP that would release the drug upon contact with the pore-forming α-toxin of MRSA [41]. Since the particle did not liberate the antibiotic without the presence of toxin, the drug would only be released in sites near to the infection. Again, this concentrates the drug, increasing its activity and reducing off-target effects.

Targeted Vaccinations
Vaccinations have been used for centuries in order to protect populations against disease-causing microorganisms. In fact, the process of intentionally exposing an individual to a less virulent form of a pathogen has led to eradication of the disfiguring and deadly disease, small pox [69]. However, while this vaccine prevented later contraction of small pox, it initially caused mild disease as it involved administration of virulent cow pox virus [70]. Vaccine development has since evolved with an interest in increasing the safety of vaccinations without sacrificing protective efficacy. Inactivation of pathogens such as influenza [71] and cholera [72] greatly increase their safety for use as a vaccines. Safety is even further improved in vaccines against Hepatitis B [73], diptheria, and tetanus [74]; utilizing purified proteins or inactivated toxins in subunit vaccines: HBsAg and DTaP, respectively. Increasing the safety of these vaccines comes at a trade off with reduced efficacy since there is no live infection generating danger signals. In some cases, immune stimulation is restored through the addition of an adjuvant such as alum [75,76]. However, many adjuvants cause off-target effects, including unwanted activation of bystander cells and potential induction of autoimmunity [77,78]. These drawbacks to adjuvants have encouraged the development of adjuvant-independent enhancement of vaccines; this has been achieved through fusion of Ag to stimulatory or targeting molecules. Research aimed at generating a vaccine against a potential biological weapon, Francisella tularensis (Ft), has shown that opsonizing inactivated Ft with α-Ft Abs targets the bacteria to Fc receptors on immune cells. This interaction stimulates uptake and presentation of Ag, activation of immune cells, and ultimately results in production of an enhanced protective immune response against virulent Ft challenge [79][80][81]. Similar methods of targeting to host cells have been achieved through bioconjugation; we have diagrammed several of these in Figure 1.  [44]. (b) A plasmid encoding a peptide carrier and capsular PS synthases is transformed into an oligosachride transferase (OTase)-engineered strain of E. coli which assembles the glycoconjugate which is subsequently purified for use as a vaccine. Modeled after production of a Kp capsular PS-Pa exotoxin A glycoconjugate vaccine [46]. (c) NPs are chemically activated to facilitate conjugation of Ag; loaded NPs are then administered as the vaccine. Modeled after activated amines to conjugate Pfs25 and generate a malaria vaccine [47]. (d) Glycoconjugate vaccines produced by chemically activating a peptide carrier, linking the purified components, and administering the conjugate as a vaccine. Modeled after generation of the meningitis vaccine [45]. (e) A plasmid encoding a targeted protein subunit conjugate that is anchored to an OMP is transformed into bacteria which are administered as a live vaccination. The conjugate protein is trafficked to the OM and cleaved from the OMP anchor. Modeled after CR-targeted Ts Ag, produced by the engineered Salmonella vaccine strain [52]. (f) A plasmid encoding an OMP-anchored conjugate (lacking a cleavage site) is expressed in bacteria, effectively targeting the whole bacterial cell to immune cells when administered as a vaccine. Modeled after targeting Ft to CRs [51].

Protein Subunit Vaccine Conjugates
Directing protein Ag to Ag-presenting cells (APCs) such as dendritic cells (DCs), increases both cellular and humoral immune responses. DCs phagocytose Ag and subsequently process and present peptides via both major histocompatibility complex-I (MHCI) and -II (MHCII), allowing for involvement of a variety of downstream cells. One example of targeting Ag to DCs is accomplished through the activating receptor, Clec9A (C-type lectin domain family 9 member A) [48,82,83]. Park et al. accomplished fusion of the influenza Ag M2e to α-Clec9A Ab through genetic engineering of a plasmid. Plasmids encoded the heavy and light chains of α-Clec9A Ab, a flexible glycine linker, and three repeats of the M2e peptide. The plasmids were then transfected into Freestyle 293F cells for expression of the Ag-Ab conjugates. The secreted Ag-Ab fusions were purified via protein G affinity chromatography. Ultimately, administration of the targeted Ags to mice resulted in increased specific immunoglobulin responses and increased protection against lethal influenza infection compared to Ag conjugated to an isotype control Ab [35]. Generation of vaccines via this method is diagramed in Figure 1a. Similar technologies have been used to fuse influenza hemagglutinin (HA) to α-MHC-II Ab, α-CCR1/3/5 Ab [42], or cytokines [43,44]. In the case of cytokines, when HA was conjugated to Xcl1 (a ligand for Xcr1) or Ccl3 (a ligand for CCR1/3/5) the vaccines targeted different subsets of DCs and therefore resulted in increased CD4+ and CD8+ T cell proliferation or only increased CD4+ T cell proliferation, respectively [43]. This phenomenon further emphasizes the ability to tailor targeting of Ags to induce specific immune responses through conjugation to different peptides. Further, these enhanced immune responses discussed above were seen in the absence of conventional adjuvant; when tested, targeted immunogens without adjuvant, were similarly or more efficient than untargeted-Ag administered with traditional adjuvant.
Conjugate vaccines have also been targeted to complement receptors on various immune cells. Pioneering work by Dempsey et al. involved conjugating hen egg lysozyme (HEL) to complement protein C3d, a ligand for complement receptors 2 (CR2) and 3 (CR3). Mice immunized with pure (C3d)3-HEL produced higher α-HEL Ab titers than those immunized with HEL plus Freud's complete adjuvant [84]. Targeting Ag to CRs is a concept that our lab has also explored and applied to whole cell vaccines which will be discussed later in this review. CR2 is primarily expressed on B cells and follicular dendritic cells (FDCs); its engagement facilitates Ag uptake and presentation by FDCs, and activation of B cells. In addition, C3d engagement of CR3 on APCs induces translocation to the lymph node and increased Ag presentation. C3d conjugation has been applied to bacterial [49] as well as viral peptides [50,82] via genetic engineering of plasmids encoding the fusion constructs. This concept has also been extended to the minimum CR2-binding region of C3d, p28, which also has shown promise as a method for increasing Ag-specific immune responses [83,[85][86][87][88].

Glyconjugate Vaccines
In addition to proteins, Ags of interest for vaccination may also be lipoproteins, peptidoglycan, polysaccharides (PS), and other capsular material. Enhancement of the immune response to PS components has been achieved by linking them to carrier or stimulatory peptides [89]. PS are primarily recognized by Toll-Like Receptors 2 (TLR2) and 4 (TLR4) which are expressed on many immune cells [90,91]. Linking PS to stimulatory or targeted peptides increases their recognition by the immune system; especially through interaction with helper T cells [45,[92][93][94]. This technique has been applied to various vaccines through chemical conjugation (Figure 1d), including the approved vaccines against Neisseria meningitidis (Nm) and Streptococcus pneumoniae (Sp) [45].
To simplify and expand chemical conjugation, PS have been conjugated to carrier proteins or lipids through the use of glycosylation processes within bacteria [46,[95][96][97]. In this case, engineered strains of Escherichia coli (Ec) that express oligosaccharyltransferases (OTase) are utilized for vaccine production. OTase transfers PS to carrier proteins and these conjugates can then be harvested and purified from Ec (Figure 1b). An example of this conjugation technique involves Ags of Klebsiella pneumoniae (Kp), a multi-drug-resistant bacteria that produces substantial amounts of PS capsule which shields it from the immune system [98]. Development of a PS capsule-targeting immune response, as opposed to a protein-targeted response would be extremely advantageous for this pathogen. A bivalent vaccine against two Kp serotypes has been engineered by Feldman et al. using both K1 and K2 capsular PS locus clusters [46]. These PS were each produced by Ec which transferred them to exotoxin protein A from Pseudomonas aeruginosa (Pa) as a carrier. These purified K1 and K2 conjugates were then mixed and used to vaccinate mice; the bivalent vaccine induced increased α-Kp Ab levels as well as protection from lethal Kp infection.

Conjugated Nanoparticle-Based Vaccines
Nanoparticles are larger than subunit vaccines and have been shown to promote trafficking of Ag to lymph nodes [99]. NPs can be formed using a variety of materials including: self-assembling lipids (liposomes), sucrose polymers, gold, and proteins [100].
Peptide NPs may be loaded by using genetic techniques, leaving their self-assembling domains intact [101]. In addition, NPs may be decorated via chemical conjugation (Figure 1c). This process involves activation of the NP to reveal active residues that are then available for bond formation with moieties of interest [102]. For example, N-hydroxy-succinimide is used to activate primary amines in lysine residues [102,103]. This strategy was utilized to display Plasmodium falciparum protein Pfs25 on NPs to be tested as a malaria vaccine [47,104]. It was found that chemical conjugation of the Ag to the NP resulted in generation of more Abs, compared to genetic or SpyTag conjugation where "tag" and "catcher" peptides are added to the desired conjugates. In some cases, the conjugated cargo may be released from the NPs in a regulated manner. Schudel et al. recently utilized NPs to concentrate delivery at the lymph nodes where cargo was then released in phases due to the use of linkers with different half-lives [105]. Delaying the release of some cargo from NPs may allow for engagement of different lymphatic cell populations and has the potential to enhance immune responses and/or drug effects. In addition to targeting the lymph nodes, loaded NPs have also been further decorated with mannose [48] and anti-DEC-205 mAbs [106]; targeting them to mannose receptor on APCs or DEC-205 on DCs, respectively.

Targeting and Conjugation Involving Whole-Cell Vaccines
Pompa-Mera et al. created a plasmid which encoded a fusion protein of p28, a Trichinella spiralis (Ts) peptide, and a Salmonella outer membrane protein (OMP) with a proteolytic site between the OMP and p28-Ts peptide fusion [52]. The plasmid was then transformed into Salmonella, resulting in a strain that produced the OM-located fusion protein, from which the targeted Ts peptide was liberated. This strain was administered as a live vaccine to mice where it induced increased Ts-specific Ab levels and conferred protection against subsequent Ts challenge. This method of vaccine generation is diagramed in Figure 1e and combines the targeting ability of p28 with the generation of danger signals from a live infection.
Our lab similarly produced a CR-targeted vaccine via generation of a plasmid encoding a CR-targeting fusion protein that was transformed into bacteria which were then administered as a whole cell vaccine. In this case, we engineered a plasmid encoding C3d conjugated to YadA, a well-characterized Yersinia OMP without a cleavage site between the proteins. Therefore, transformation of bacteria with the plasmid resulted in expression of OM-localized, surface-displayed C3d. This promotes targeting of the entire bacterial immunogen to CRs on immune cells and enhances protective responses; we have demonstrated this utilizing murine pulmonary Francisella tularensis infection as a model [51]. This technique is diagramed in Figure 1f.

Summary
The ability to combine therapies or prophylactics with another element that facilitates delivery to their target has advanced a wide range of fields; including the treatment of cancers, dementia, and bacterial infections, as well as enhancement of vaccinations. Targeting has been achieved through bioconjugation of two or more moieties that embody different functions. Generally, one subunit will provide the targeting or enhancement function, whereas the other performs the desired treatment. This not only enhances treatment efficacy but also limits unwanted side effects; appealing traits to virtually all patient care. It is important to note however, that the potential for side effects may not be absolutely eliminated, even with targeting to distinct sites or cells. Another potential drawback in developing targeted treatments is the delay in approval for new treatments. In the case of cancer treatments, a broadly administered chemotherapy agent may be more readily available for patient care. Further, targeting therapies and prophylactic treatments to specific cell types may impede involvement of other cells whose benefits are not yet fully realized. These facts underscore the need for continued research and advancement of targeted therapies. Funding: This review was funded by grants from the National Institutes of Health (RO1AI123129) and the Department of Defense (W81XH1910661).