Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

A reversible cell penetrating peptide-cargo linkage allows dissection of cell penetrating peptide- and cargo-dependent effects on internalization and identifies new functionalities of putative endolytic peptides

A reversible cell penetrating peptide-cargo linkage allows dissection of cell penetrating... TYPE Original Research PUBLISHED 21 November 2022 DOI 10.3389/fphar.2022.1070464 A reversible cell penetrating peptide-cargo linkage allows OPEN ACCESS EDITED BY Nian-Qiu Shi, dissection of cell penetrating Jilin Medical University, China REVIEWED BY peptide- and cargo-dependent Jing Sun, Geneleap Biotech, Luye Pharm, China Ji Li, effects on internalization and University of Michigan, United States *CORRESPONDENCE identifies new functionalities of Jonathan L. McMurry, jmcmurr1@kennesaw.edu putative endolytic peptides SPECIALTY SECTION This article was submitted to Experimental Pharmacology Daniel P. Morris, Lucy C. Snipes, Stephanie A. Hill, and Drug Discovery, a section of the journal Michael M. Woods, Maria M. Mbugua, Lydia R. Wade and Frontiers in Pharmacology Jonathan L. McMurry* RECEIVED 14 October 2022 ACCEPTED 09 November 2022 Department of Molecular and Cellular Biology, Kennesaw State University, Kennesaw, GA, PUBLISHED 21 November 2022 United States CITATION Morris DP, Snipes LC, Hill SA, Cell penetrating peptides (CPPs) are a promising technology for therapeutic Woods MM, Mbugua MM, Wade LR and McMurry JL (2022), A reversible cell delivery of macromolecular cargos. CPPs have generally used covalent linkages penetrating peptide-cargo linkage to cargo, ensuring a common fate as one molecule. Conversely, our CPP- allows dissection of cell penetrating peptide- and cargo-dependent effects adaptor, TAT-CaM, noncovalently binds calmodulin binding sequence (CBS)- on internalization and identifies new containing cargos in calcium rich media then dissociates in the calcium-poor functionalities of putative endosomal environment following internalization, enhancing endosomal endolytic peptides. Front. Pharmacol. 13:1070464. escape relative to standard CPPs. In this study, we report cell entry of doi: 10.3389/fphar.2022.1070464 positively charged protein cargos that were not increased by TAT-CaM while COPYRIGHT cargos based on the negatively charged maltose binding protein (MBP) © 2022 Morris, Snipes, Hill, Woods, displayed little intrinsic internalization but were internalized by TAT-CaM. In Mbugua, Wade and McMurry. This is an open-access article distributed under addition, association of positively charged proteins with negatively charged the terms of the Creative Commons nucleic acids reduced internalization. This evidence points to the dominant role Attribution License (CC BY). The use, distribution or reproduction in other cargo charge plays in apparent CPP effectiveness. There has been little forums is permitted, provided the systematic investigation as to how interaction between CPPs and cargos original author(s) and the copyright impacts internalization efficiency. Our adaptors provide a tool that allows owner(s) are credited and that the original publication in this journal is combinatorial assays to detect emergent properties. Toward this end we cited, in accordance with accepted added 4 endolytic peptide (EP) sequences between cargo CBS and MBP academic practice. No use, distribution moieties to create 4 new cargos and between TAT and CaM to create or reproduction is permitted which does not comply with these terms. 4 new adaptors. The new cargos were assayed for internalization alone and with a panel of CPP-adaptors to identify combinations that displayed increased internalization efficiency or other properties. Among the most important results, addition of the EP LAH4 improved adaptor performance and provided some CPP capability to cargos. MBP-LAH4-CBS was internalized more effectively by most adaptors, suggesting this sequence has general stimulatory ability. Two other EPs, Aurein 1.2 and HA2, also provided some CPP capability to their MBP cargos but were unexpectedly antagonistic to internalization by most adaptors due to retention of adaptor/cargo complexes Frontiers in Pharmacology 01 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 on the cell surface. We thus identified LAH4 as stimulator of internalization in both adaptors and cargos and uncovered new functionality for Aurein 1.2 and HA2, which may be related to their identification as EPs. Future experiments will test new endolytic capabilities made possible with combinatorial approaches. KEYWORDS cell-penetrating peptides, protein transduction domains (PTD), endolytic peptides, endocytosis, calmodulin (CAM) Introduction cargo proteins in all cell lines that we examined. Subsequent efforts to develop CPP-based applications ensued, some of which Cell-penetrating peptides (CPPs), sometimes called protein were very successful, e.g. we delivered human papillomavirus transduction domains (PTDs), have long held great promise for E2 protein to cervical cancer cells, inducing senescence and thus overcoming failures of biomolecule therapeutic leads due to demonstrating that cargos remained folded and active even after bioavailability issues relating to failure to cross membranes delivery (LeCher et al., 2020). (Fonseca et al., 2009; Johnson et al., 2011; Lönn and Dowdy, In the course of efforts to develop applications, particularly 2015; Derakhshankhah and Jafari, 2018). Though the specific co-delivery of Cas9 and crRNAs to effect CRISPR/Cas genome endocytic mechanisms by which they do so remain a subject of editing by exogenous delivery of the components, we noticed a debate, CPPs are capable of mediating penetration of the plasma correlation between charge of the cargo and efficacy of TAT- membranes of mammalian cells by molecules to which they are CaM-mediated delivery, the more negative the cargo, the worse coupled, allowing delivery of “cargos” to cell interiors, a the delivery. We also observed that positively charged cargo potentially transformative platform technology for drug proteins were taken up themselves, sometimes as well as or even delivery and other applications. However, development of better than when coupled to TAT-CaM. Noting that most CPPs CPP-based delivery systems has been hindered because cargos are positively charged and calmodulin is itself rather extensively are normally coupled to CPPs by covalent or other irreversible negatively charged (net −24, TAT-CaM with its positive CPP and linkages and become entrapped in endosomes after cellular entry linker sequences has a charge of −16), we sought to develop next- rather than reaching the cytoplasm and their subsequent targets generation CPP-adaptors that might better utilize more favorable (Lecher et al., 2017). To address this problem, we have employed charge states to deliver cargos, the first of which incorporated a high affinity, reversible noncovalent coupling strategy, calmodulin from naked mole rat, Heterocephalus glaber. Naked attaching cargos to CPPs via a CPP-containing “adaptor” mole rat CaM (NMR-CaM) has a positively charged domain of protein, calmodulin. Our prototype CPP-adaptor, TAT- unknown function N-terminal to the EF hand domain, which is Calmodulin (TAT-CaM), consists of the cell penetrating invariant from that of human CaM. The resultant adaptor, TAT- moiety from HIV transactivator of transcription and human NMR-CaM, has a lesser net negative charge (−5) and evinced calmodulin (Salerno et al., 2016; Ngwa et al., 2017). TAT-CaM exactly the behavior we predicted, delivering cargos that had binds CaM binding-site (CBS) containing cargos with nM resisted efficient TAT-CaM-mediated delivery, e.g. polyA affinity in the presence of calcium but negligibly in its binding protein PAB-1 (Gentry et al., 2021). absence. The CBS can be located anywhere in the cargo. Although not a basis for early adaptor design, we realized that 2+ TAT-CaM/cargo complexes form spontaneously in the Ca - the CPP-adaptor strategy afforded the ability to directly compare 2+ containing extracellular milieu but, concurrent with Ca flux internalization of cargo alone with that of the CPP-adaptor/cargo during early endocytosis (Albrecht et al., 2015), cargos dissociate complex, an advantage over examinations of cargo charge done from CPP-adaptors within endosomes and are released to the with traditional covalent coupling strategies, e.g (Hymel et al., cytoplasm though the CPP-adaptor remains trapped. 2022). It was the availability of this comparison that led to the Our first reports of the success of our reversible coupling realization that cargo charge dramatically changed the properties strategy (Salerno et al., 2016; Ngwa et al., 2017) were attended by of TAT-CaM/cargo complexes. In support of this idea we report high hopes for a generally utile delivery vehicle adaptable not herein a body of experimental data using positive and negative only to any desired protein but nucleic acids and perhaps other cargos as well as adaptors designed to have a more positive net biomolecules as well. Possible cargos are limited only by the charge. requirement of attaching a CBS to the cargo, which can be Another advantage of the adaptor/cargo model is the ability achieved recombinantly for proteins, with covalent chemistry to change the characteristics of the adaptor and the cargo for nucleic acids, or via a secondary adaptor strategy such as independently. This allows for the improved CPP using a recombinant CBS-biotin binding protein to deliver characteristics of new adaptors without impact on the released biotinylated nucleic acids. In some respects, those hopes were cargos’ ability to leave the endosome. Similarly, adaptor delivery realized as we showed that TAT-CaM worked for multiple model systems can be modified to effect more efficient cytoplasmic Frontiers in Pharmacology 02 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 delivery of cargos by inclusion of endolytic peptide sequences eliminate some extraneous sequences originally designed to (EPs) (also called “endosomolytic peptides” or “endosomal facilitate cloning; however, the His-tag, TAT and CaM escape domains”)(Aqeel and Khan, 2022). Some EPs are sequences are invariate from the original study (Salerno et al., thought to destabilize endosomes in a pH-dependent manner 2016). All references to TAT-CaM in this paper mean TAT- by binding and disrupting membranes during the acidification CaM 2.0. that occurs in early transport, e.g. haemagglutinin-derived peptides (Wharton et al., 1988). Toward this end we added four endolytic peptide (EP) sequences: Aurein 1.2, GALA, Expression, purification, and labeling HA2 and LAH4 (Wadia et al., 2004; Akita et al., 2011; Shahmiri et al., 2015; Moulay et al., 2017) to cargos between Proteins were expressed essentially as described (Salerno the CBS and negatively charged maltose binding protein (MBP) et al., 2016; Ngwa et al., 2017) with minor modifications. to create 4 new cargos and between TAT and CaM to create Briefly, plasmids were freshly transformed into BL21 (DE3) 4 new adaptors to assess internalization as a function of adaptor pLysS. Overnight cultures grown from single colonies were and cargo charge. Because both CaM and MBP have intrinsic subcultured into 1L Luria-Bertani broth and grown with negative charge, neither have significant intrinsic internalization vigorous shaking at 37 C. At OD ~0.4, the temperature was behavior, and the effects of the EPs are manifest without lowered to 30 C, cells were induced with 0.2 mm IPTG and complication from competing behaviors. While EPs are growth continued for 4 hours. The procedure was altered slightly supposed to increase escape, an increase in intracellular for MBP-CBS cargos, the cells for which were grown in Terrific concentration may also be due to advantageous CPP-like Broth (TB) supplemented with 0.25% (w/v) glucose and behavior. In this report we have focused on the clear and induction was conducted at 32 C. Cells were harvested by differential impact these EPs have on internalization. centrifugation at 10,000 × g and frozen at −80 C. Purification was also performed essentially as described via immobilized metal affinity chromatography (Salerno et al., 2016). Materials and methods Briefly, cell pellets were thawed on ice, resuspended in lysis buffer (50 mm Tris pH 8, 500 mm NaCl, 10 mm imidazole, 10% Plasmids, strains, and cell lines glycerol and 6 mm β-mercaptoethanol for disulfide containing proteins). 1 mg/ml DNAse and 0.25 mg/ml lysozyme were added The E. coli strain used in this study, BL21 (DE3)pLysS was during resuspension. For TAT-NMR-CaM only, Halt Protease propagated from purchased cells from EMD Millipore Inhibitor Cocktail (ThermoFisher, Waltham, MA, United States) (Burlington, MA, United States) or other established supplier. was added to 1x per manufacturer’s protocol. Cells were broken Baby hamster kidney (BHK) cells (#CCL-10) and HEK-293 cells via passage through a French press at 20,000 psi and subjected to (CRL-1573) were purchased from ATCC and cultured in centrifugation at ~27,000 × g to pellet unbroken cells and debris. Dulbecco’s Modified Eagles’ Medium with GlutaMAX Clarified lysate was passed over a cobalt affinity column using an Supplement (Gibco, United States) and 5% or 10% fetal FPLC system while monitoring A , washed with wash buffer bovine serum. (equivalent to lysis buffer with 25 mm imidazole instead of Plasmids used in this study were previously described 10 mm) until baseline absorbance was attained, after which (Salerno et al., 2016; Ngwa et al., 2017; Gentry et al., 2021)or protein was eluted in elution buffer (lysis buffer with 250 mm constructed as described in Supplementary Figure S1 from parent imidazole). Protein-containing fractions were pooled, vectors pET19b and pET22b (EMD Millipore, United States), concentrated and exchanged by passage over a desalting pMalc-5x (New England Biolabs, United States) or pCal-N- column into 10 mm HEPES, pH 7.4, 150 mm NaCl, 10% FLAG (Agilent Technologies, United States). For example, glycerol, 1 mm CaCl for fluorescence labeling, biotinylation pJM161, encoding TAT-naked mole rat calmodulin (TAT- or other further use. Quantitation was done using Bradford NMR-CaM) consisted of an E. coli-optimized synthetic gene Assay with bovine serum albumin as standard. For (GeneScript, Piscataway, NJ, United States) cloned into NdeI and fluorescence labeling, DyLight 488, 550 or 650 NHS Esters BamHI sites in pET19b with an in-frame stop codon prior to the (ThermoFisher) were used to introduce fluorescent labels BamHI site. The encoded TAT-NMR-CaM protein consists of under conditions that resulted in incorporation efficiencies the TAT peptide sequence (YGRKKRRQRRR) N-terminally below 0.6 dyes/molecule as determined using the algorithm fused to Heterocephalus glaber (naked mole rat) calmodulin suggested by the manufacturer (ThermoFisher). Dye removal (GenBank: EHB02604.1) (Gentry et al., 2021). A vector- columns were then used to remove unreacted dye as encoded 10xHis tag containing an enterokinase cleavage site recommended by the manufacturer (ThermoFisher). and several spacer residues is N-terminal to TAT. The version of All proteins were assayed by biolayer interferometry to assure TAT-CaM used in this study, initially designated “TAT-CaM high affinity (low nm-to-high pm) binding between adaptors and 2.0” (Ngwa et al., 2017), is a variant of a prior version modified to CBS-cargos in the presence of calcium and fast dissociation upon Frontiers in Pharmacology 03 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 TABLE 1 Schematic descriptions of proteins used in this study; A, CPP-adaptors and adaptor controls; B, cargo proteins. Additional details including amino acid sequences, spacers and tags are described in Supplementary Information. A Name Description (N-to-C) Parent vector References TAT-CaM (2.0) His-TAT-CaM pET19b Ngwa et al. (2017) CaM Cntl His-CaM (no TAT) pET19b This study TAT-Aur-CaM His-TAT-Aurein 1.2-CaM pET19b This study TAT-GALA-CaM His-TAT-GALA-CaM pET19b This study TAT-LAH4-CaM His-TAT-LAH4-CaM pET19b This study TAT-HA2-CaM His-TAT-HA2-CaM pET19b This study TAT-NMR-CaM His-TAT-naked mole rat CaM pET19b Gentry et al. (2021) pGFP-CaM His-GFP (+36)-CaM pET19b Lawrence et al. (2007), This study CBS-Tam CBS-Tamavidin (C627S)-His pET22b This study MBP-CBS MBP-CBS-His pMalc-5x This study MBP-AUR MBP-Aurein 1.2-CBS-His pMalc-5x This study MBP-GALA MBP-GALA-CBS-His pMalc-5x This study MBP-LAH4 MBP-LAH4-CBS-His pMalc-5x This study MBP-HA2 MBP-HA2-CBS-His pMalc-5x This study Cas9-NLS His-CBS-Cas9-NLS pET19b This study Cys-less Cas9-NLSx2 His-CBS-NLS-Cas9-NLS (C80S, C574S) pET19b This study its removal with EDTA. All evinced kinetics and affinities highly 30 min of use. Following assembly, complexes were transferred to similar to CaM with a natural binding partner (McMurry et al., room temperature for 5–10 min and then diluted into media also 2011) and within the ranges observed for previously reported at room temperature. After 1–3 min complex containing media adaptor-cargo pairings (Salerno et al., 2016; Ngwa et al., 2017; was microcentrifuged at maximum velocity (>15,000 rcf) for Gentry et al., 2021). Table 1 lists all adaptors and cargos used in 1 min to remove precipitates, briefly warmed in a 37 C bath this study and lists the plasmids from which they were expressed. (2–4 min) and transferred onto BHK cells from which growth media had just been removed. Unless otherwise stated, cells were treated with the indicated concentrations of complex at 37 C 2+ Cell culture under 5% CO for 1 h, washed twice with PBS (containing Ca 2+ and Mg ) and analyzed in imaging media usually comprised of BHK cells were maintained in a 37 C, 5% CO environment growth media without phenol red but including NucBlue, Live in growth media consisting of DMEM (GlutaMax, (ThermoFisher). Aliquots of proteins containing cysteine ThermoFisher) containing +4.5 g/L D-glucose and 1.9 mm residues (Cas9-NLS and TAT-NMR-CaM) were thawed after 2+ Ca with no sodium pyruvate) supplemented with 4 mm addition of 1/20th volume of 1 mm DTT in storage buffer, which glutamine and 5% or 10% fetal bovine serum (FBS) as became less than 5 µm during cell treatment. indicated. Cells were replated in coverslip slides in the same Internalization of TAT-CaM alone was assayed using media 20–24 h before use in cell penetration assays. DyLight 488 and 550 labeled adaptor modified with less than 0.3 dyes per protein. TAT-CaM in storage buffer was diluted 10- fold with imaging media and transferred onto BHK cells. TAT- Cell penetration assays CaM internalization of Cas9 used various labels on both proteins, as well as an optional treatment with green CMFDA CellTracker Internalization assays were performed as described (Salerno (Molecular Probes). When CellTracker was used, treatments et al., 2016; Ngwa et al., 2017) with minor modifications. Protein were followed by a PBS wash and a 20 min incubation with aliquots in a common storage buffer (10 mm HEPES, pH 7.4, PBS (mock) or CMFDA CellTracker in PBS, as indicated. TAT- 150 mm NaCl, 10% glycerol, 1 mm CaCl ) were thawed and CaM internalization experiments utilizing the “cysless” CBS- sequentially assembled into complexes in this buffer on ice within Tam were performed with various concentrations of an Frontiers in Pharmacology 04 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 1 By itself, TAT-CaM internalization requires concentrations above 1 µm. BHK cells were incubated with (A,B) Dylight 550 or (C,D) Dylight 448 labeled TAT-CaM, respectively, for 60 or 30 min at 1 µm or 10 μm, in DMEM with 10% FBS under a 5% CO2 atmosphere at 37 C. Cells were then 2+ 2+ washed 3 times with PBS (with Mg and Ca ) and imaged in media without phenol red but containing NucBlue. Dye incorporation averaged less than 0.3 dyes per TAT-CaM molecule. oligonucleotide (CCATCCTGGTCGAGCTGGACGGCGACG) across the cell. Experiments used either a Zeiss LSM 700 or LSM labeled with a 5′ Alexa 488 dye and a 3’ biotin equilibrated 900 confocal microscope generally at 400× with settings adjusted with CBS-Tam prior to complexation with the CPP-adaptor in to best image quality unless otherwise stated. Opti-MEM media (ThermoFisher), both preceded and followed by two washes with PBS and final addition of imaging media. Confocal imaging was initiated as soon as practical (usually Results and discussion 7–15 min following the washes) and continued for up to 1 h in an incubated chamber that maintained a humidified, 37 C, 5% CO The CPP-adaptor technology was devised to facilitate environment. In some experiments a parallel plate was initially dissociation of CPP and cargo, enabling cargo release from treated identically; however, imaging was performed with the endosomally entrapped CPPs (LeCher et al., 2017). However, complex present between 20 and 60 min after addition. Cells with another facet of the CPP-adaptor technology is the simplicity of normal nuclei and monolayer distribution that avoided cell comparative analysis of adaptor, cargo and complex stacking were selected for imaging. Images were taken at a internalization properties. In this regard the properties of our confocal plane just below the nuclear center to minimize slide CPP adaptor, TAT-CaM, are important for understanding TAT- surface background while retaining signal from intracellular CAM/cargo internalization. As observed for TAT-cargos and structures across the cell. When NucBlue was not used, CPPs in general (Lönn and Dowdy, 2015), concentrations above images were taken from a similar confocal plane above slide 1 µm are necessary for effective internalization of TAT-CaM surface background that still included intracellular structures (Figure 1). By itself TAT-CaM is a rather small protein (~26 kDa) Frontiers in Pharmacology 05 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 2 Internalization of Cas9 bound to TAT-CaM occurs readily at very low concentration. Cas9-NLS bound to equimolar TAT-CaM at 100 nm was incubated with BHK cells in DMEM with 10% FBS for 1 h prior to PBS treatment, washing and imaging. Panels show (A) 550-labeled TAT-CaM, (B) 488- labeled Cas9-NLS and (C) Composite including NucBlue. In (D) the enlarged inset (Figure 2D) shows Cas 9-NLS and TAT-CaM are often in separate but juxtaposed endosomes. whose properties may be altered by labeling; however, similar While at the time these results were mystifying, it has long internalization was observed with modification by two amino- been recognized that most CPPs are positively charged and may reactive fluorescent reagents, Dylight 550 (Figures 1A,B) and function by neutralizing negative charge on plasma membranes Dylight 488 (Figures 1C,D). To limit artifacts from labeling, (Skotland et al., 2015), which is believed to aid inward membrane TAT-CaM was modified at low efficiency (<0.3 dyes/molecule) curvature necessary for vesicle internalization (Alves et al., 2010). with reduced concentrations of the reactive dyes below the Nevertheless, we were not aware of the extent to which surface manufacturer recommendations. Indeed, for the same reason, charge of a cargo impacted CPP behavior as studies using other proteins in this report were labeled at <0.6 dye/molecule to covalently bound CPP moieties rarely test intrinsic cargo limit introduction of multiple dyes per protein monomer. internalization ability. Unexpectedly, when labeled Cas9 (the protein component of It is also known that negative charge (Hymel et al., 2022) and CRISPR) containing a CBS tag and a nuclear localization signal nucleic acids (Yokoo et al., 2021) often inhibit CPP-mediated (NLS) was internalized with equimolar 100nM TAT-CaM, the internalization. Concurrent with the Cas9 (CRISPR) TAT-CaM/Cas9 complex internalized readily at a concentration experiments, a companion project involved development of an that was more than 100-fold lower than observed for TAT-CaM adaptor-based method to introduce donor DNA alongside alone (Figure 2). In other words, cargo internalization increased CRISPR. To introduce biotin labeled nucleic acids, we created TAT-CaM internalization. Also note that TAT-CaM and a CBS-tagged a version of avidin based on the expressible fungal Cas9 often appear separated but in nearby vesicular locations homolog, tamavidin (Takakura et al., 2010; Takakura et al., (Figure 2C and enlarged inset 2D). Importantly, under most 2013). With CBS-Tamavidin (Tam), biotin labelled nucleic conditions TAT-CaM does not increase Cas9 internalization, i.e. acids can be delivered into cells as a Tam-oligo cargo complex there is a lack of TAT-CaM dependent specificity (Figures 3A,B). reversibly bound to any adaptor. This version of “cysless” CBS- On the other hand, at Cas9 concentrations below 50 nm where Tam (Tam) did not have an NLS and the lone cysteine was intrinsic internalization becomes modest, high TAT-CaM converted to serine to prevent disulfide-induced precipitation. As concentrations specifically increased Cas9 internalization avidin orthologs including tamavidin form obligate tetramers (Figure 3C). The implication of this experiment is that (Takakura et al., 2009), as many as 4 nucleic acids can be bound Cas9 internalizes more readily when bound to TAT-CaM. to a single tetramer, which will have 4 CBS sites that can bind as However, this positive effect appears insignificant when many as 4 adaptors. intrinsic Cas9 internalization is high. In addition, we have While early attempts to deliver labeled oligos clearly resulted observed other positively charged proteins used as cargos in some internalization of the TAT-CaM/CBS-Tam/oligo displayed similar intrinsic internalization (e.g. polyA binding complex, subsequent experiments showed CBS-Tam tetramers protein, Gentry and McMurry, unpublished). in the presence of a molar excess of oligos hardly enter the cell. Frontiers in Pharmacology 06 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 3 Specific and nonspecific Cas9 internalization with the TAT-CaM. BHK cells in DMEM with 10% FBS were treated for 1 h with Cas9-NLS in the absence (top) or presence (bottom) of excess TAT-CaM prior to washing (PBS x3) and imaging. Cas9-NLS at concentrations >50 nm showed no increase in internalization with inclusion of TAT-CaM as shown in examples (A) 500 nm and (B) 50 nm; however, at low Cas9-NLS concentrations including (C) 10 nm, TAT-CaM specifically increased internalization. This inconsistency was explained by decreasing the average TAT-CaM (Figure 5A). However, TAT-CaM-mediated number of oligos per tetramer, which produced the surprising internalization of MBP-CBS required a return to µm result that decreasing oligo concentration increased oligo concentrations (Figure 5B) characteristic of typical CPP- internalization (Figure 4). Maximal oligo internalization was mediated internalization processes. To test the extent to which observed at an average of 1 oligo per Tam tetramer the negative charge of CaM was inhibitory, the internalization of (Figure 4C), although some entered at 2 oligos per Tam a TAT-less CaM control (CaM-Cntl) in complex with MBP was tetramer (Figure 4B). Beyond the nearly complete block to concurrently examined and found to further reduce MBP-CBS internalization, stoichiometric Tam/oligo concentrations also internalization (Figure 5A, left panel and Figures 8–10 below). reduced complex binding to the glass coverslip. These results Importantly, comparison of TAT-CaM/MBP-CBS to (TAT-less) are consistent with increased negative charge decreasing the CaM-Cntl/MBP-CBS isolates the effect of the TAT sequence on ability of the CPP to internalize cargo. While this explanation complex internalization, a conclusion that is valid because was accepted at the time, the dramatic inhibitory effect of oligo neither CaM nor MBP contribute significantly to the on complex internalization seems to indicate the oligo is forced internalization process. into the proximity of the membrane and coverslip, suggesting Given these results, it might be expected that association of a steric hindrance may play a role. crRNA with Cas9 would decrease intrinsic internalization. To test the idea that a negatively charged cargo would display Indeed, internalization of “cysless” CBS-Cas9-NLSx2 little intrinsic internalization, we created a CBS fusion cargo (Figure 6B) was much greater than observed when it was part based on the negatively charged maltose binding protein (MBP). of the CRISPR complex (Figure 6C). Nevertheless, the CRISPR As predicted, MBP-CBS by itself displayed little tendency to complex still internalized readily (Figure 6C) and TAT-CaM still enter BHK cells, but was induced to internalize when bound to had no ability to detectably increase CRISPR internalization and Frontiers in Pharmacology 07 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 4 CBS-Tam, a cargo made from the fusion of CBS to the fungal avidin, tamavidin, internalizes model oligos but only at stoichiometries below 2 oligos per CBS-Tam tetramer. BHK cells were incubated with 600 nm TAT-CaM/Tam complexes bound to an oligo with a 5′ Alexa 488 dye and a 3′ biotin present at concentrations of (A) 600, (B) 300, (C) 150, or (D) 0 nm oligo for 1 h prior to washing (PBSx3) and imaging. in some experiments seemed to be somewhat inhibitory The strategy of using noncovalent association of adaptor and (Figure 6D). Because the ATTO 550 crRNA duplex displays cargo allowed independent modification of adaptors and cargos to virtually no cell penetration by itself (Figure 6A), essentially all increase internalization. Further, combinatorial inclusion of the crRNA inside the cell entered as part of the CRISPR complex. Endolytic Peptides (EPs) within both adaptors and cargos should Thus, the crRNA provides a measure of CRISPR internalization. increase endosomal escape and aid internalization as the beneficial When Cas9 was not labeled similar amounts of crRNA/ properties of these peptides may be due to both factors. To explore Cas9 internalization were observed (unpublished, D. Morris). thesepossibilities,wechose aset of 4 EPs,Aurein1.2 (AUR), GALA, Although TAT-CaM failed to increase CRISPR internalization, LAH4 and HA2, previously suggested to increase endosomal escape, intrinsic internalization and frequent TAT-CaM-dependent and incorporated these sequences into a set of MBP-EP-CBS cargos effects on the details of intracellular localization (Figures and TAT-EP-CaM adaptors. Problematically, association of 6C,D) suggested some potential for genome modification. negative TAT-CaM with negative MBP produces complexes with Unfortunately, even with CRISPR complexes that both cleaved modest internalization, which probably limits endolytic outcomes, target DNA in vitro and modified genomic DNA when nevertheless, these modified adaptors and cargos produced clear and introduced using CRISPR-Max, no detectable genome differential effects on complex internalization. modification occurred when CRISPR was introduced with or Initially, intrinsic internalization of each cargo alone was without TAT-CaM in HEK-293 (Supplementary Figure S2)or compared to that observed with TAT-CaM, four TAT-EP-CaMs BHK cells (Morris and McMurry, unpublished). and CaM-Cntl to obtain a profile of adaptor effectiveness. To Frontiers in Pharmacology 08 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 5 TAT-CaM specifically induces MBP-CBS internalization but requires µm concentrations. (A) Internalization of MBP-CBS alone (left panels), with equimolar with TAT-CaM (center panels) or with “TAT-less” CaM control (right panels) following 1 h incubation with BHK cells in growth media. Cells were washed (PBS x2) and imaged immediately (top panels) or later (bottom panels) in imaging media with NucBlue. (B) Internalization of equimolar MBP-CBS and TAT-CaM at 4 concentrations following a 2 h incubation, washing (PBS x2) and imaging in media with NucBlue. limit the time differential between imaging treatments, GALA alone, TAT-CaM modestly increased experiments were performed in sets of 4, each containing the internalization (Figure 7C), while TAT-LAH4-CaM no adaptor control as a reference. Multiple rounds of images were remained a very effective adaptor (Figure 7D)and CaM- taken as soon as possible after complex removal and PBS Cntl was again inhibitory (Figure 7D). washing. As shown above, relative to MBP-CBS alone, The MBP-LAH4 cargo displayed strong internalization association with TAT-CaM specifically increased across the adaptor set (Figure 8) and reduced detection internalization (Figure 7A). Internalization of MBP-CBS by sensitivity was required to produce non-saturating signal. As TAT-LAH4-CaM was also very efficient (Figure 7B), while shown more directly below, the LAH4 sequence in MBP-LAH4 TAT-AUR-CaM was marginally effective (Figure 7A) and itself acted as an effective CPP. Against this background of higher CaM-Cntl depressed internalization (Figure 7B). internalization, TAT-CaM increased MBP-LAH4 internalization With MBP-GALA as cargo, the adaptors displayed modestly (Figure 8A) while TAT-LAH4-CaM was still very similar activity as with MBP-CBS. Compared to MBP- effective (Figure 8B). Unexpectedly, TAT-AUR-CaM, TAT- Frontiers in Pharmacology 09 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 6 Internalization of CRISPR components and CRISPR compared to the TAT-CaM/CRISPR complex. BHK cells in DMEM with 5% FBS were treated for 1 h with (A) crRNA, (B) Cys-less-Cas9-NLSx2, (C) CRISPR or (D) the TAT-CaM/CRISPR prior to washing (PBS x2) and imaging in media with NucBlue. The Cys-less Cas9-NLSx2 concentrations were 100 nm; however, the crRNA and TAT-CaM were used at 150 nm to ensure available Cas9 was fully bound. GALA-CaM and TAT-HA2-CaM were inhibitory and decreased efficacy with MBP-AUR and MBP-HA2. Further, the MBP-LAH4 internalization much as CaM-Cntl. reciprocal exchange of LAH4 (into cargo) and Aurein and Cargos MBP-AUR and MBP-HA2 displayed similar but HA2 (into adaptor) produced poor internalization in both distinctly alternative behaviors (Figure 9). Both EPs appeared experiments, suggesting this negative interaction is to have some CPP capability of their own (see below) but also independent of context. More broadly, the comparison of bound the coverslip surface complicating analysis. In addition, adaptor effectiveness confirmed the utility of the TAT and both cargos were resistant to increased internalization by other LAH4 sequences and demonstrated that inclusion of GALA, adaptors, although TAT-CaM showed some efficacy. Most AUR and HA2 sequences in the adaptor did not increase strikingly, TAT-LAH-CaM had modest impact on MBP-AUR internalization. (Figure 9B) and no impact on MBP-HA2 internalization despite While beneficial EP effects on cargo internalization are this adaptor’s effectiveness with the other cargos (Figure 9D). implied by reduced sensitivity used in detection, direct Also noteworthy was the loss of MBP-HA2 internalization when comparison provided a more quantitative measure. To complexed with TAT-HA2-CaM to levels below that of CaM compare these EP effects, all 5 cargos were simultaneously control (Figure 9D). analyzed first alone and then with selected adaptors. Although too large for presentation in the body of this report, Comparison of the intrinsic internalization tendencies of the the overall internalization characteristics of the MBP-EP-CBS EP-cargos showed LAH4-MBP internalized readily, MBP-AUR cargos and TAT-EP-CaM adaptors are most apparent in a and MBP-HA2 to some extent and MBP-CBS and MBP-GALA noncolor matrix simultaneously displaying all the very little (Figure 10A). Stated alternatively, when fused into an representative images (Supplementary Figure S3). The MBP-CBS cargo, LAH4 is a relatively good CPP, Aurein, and reproducibility of the pattern is shown in a nearly complete HA2 have CPP characteristics and GALA appears to have no second matrix performed at half concentration (0.5 µm) that is such property. missing only the CaM control (Supplementary Figure S4). As shown above, TAT-CaM has some ability to increase Significantly, both data sets demonstrate the ability of TAT- internalization of all 5 cargos. Nevertheless, internalization of LAH4-CaM to dramatically increase internalization of MBP- MBP-EP-CBSs by TAT-CaM (Figure 10B) correlated with the CBS, MBP-GALA and MBP-LAH4, while having minimal intrinsic internalization observed for the cargos themselves. Frontiers in Pharmacology 10 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 7 550- MBP-CBS and 550- MBP-GALA internalized with adaptor set including controls. Internalization of equimolar adaptor/cargo complexes (1 µm) in growth media were analyzed as sets of four conditions, each containing the no adaptor cargo as reference. All complexes were internalized for 1 h prior to washing (PBS x2) and imaging in media containing NucBlue. Internalization of CBS-MBP in the upper panel (A) used: no adaptor, TAT- CaM, TAT-AUR-CaM, or TAT-GALA-C and in the lower panel (B) used: no adaptor, “TAT-less” CaM-Ctrl, TAT-LAH4-CaM, and TAT-HA2-CaM. Internalization of MBP-GALA used the same adaptor sets in the upper (C) and lower (D) panels. For presentation, intensities within each set were adjusted identically: however, for these cargos no-adaptor reference are not presented at equal intensity in upper and lower panels to avoid oversaturating internalization with TAT-LAH4-CaM. Frontiers in Pharmacology 11 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 8 550-MBP-LAH4 internalized with adaptor set including controls as in the prior figure. Equimolar adaptor/cargo complexes (1 µm) were internalized in growth media for 1 h prior to washing (PBS x2) and imaging in media containing NucBlue. Internalization of MBP-LAH4 with adaptors as above in both the upper panel (A) and the lower panel (B). For presentation, intensities within each set adjusted identically but with the no-adaptor reference at approximately equal intensity for both panels. Indeed, internalization of the adaptor/cargo complexes internalization, it was clear that more effective adaptors maintained almost the same rank order as observed for the were needed. Several approaches were used to increase the cargos alone (MBP-LAH4>MBP-AUR > MBP-HA2>MBP- internalization potential of these next-generation adaptors CBS = MBP-GALA) except that TAT-CaM-induced and one of the most successful involved use of CaM from internalization of MBP-CBS became visible (MBP-CBS > naked mole rat to create a TAT-NMR-CaM adaptor (Gentry MBP-GALA). These results showed that substrate EPs still et al., 2021). A second approach used an engineered positive played a major role in boosting the internalization process, version of GFP (pGFP), which has been shown (Lawrence but the overall efficiency was dependent on both cargo EPs et al., 2007; Thompson et al., 2012) to be a powerful cell and TAT-CaM. penetrating protein (also ‘CPP’), to create a pGFP-CaM On the other hand, when the TAT-LAH4-CaM was used as adaptor. Although full characterization of these very an adaptor, the internalization properties of the complexes different adaptors is beyond the scope of this report, it changed qualitatively (MBP-LAH4>MBP-CBS > MBP-AUR > seemed valuable to see how these more effective adaptors MBP-HA2 = MBP-GALA) (Figure 10C). While MBP-GALA interacted with the MBP-EP-CBS cargos. internalization remained low, the relative signal became similar When TAT-NMR-CaM was used to internalize the EP-cargo to HA2-MBP, in agreement with above data showing TAT- set, cargos having the least intrinsic internalization ability (MBP- LAH4-CaM adaptor dramatically increased internalization of CBS and MBP-GALA) were induced relative to those with strong MBP-GALA (Figure 7D). These data show that internalization intrinsic internalization (Figure 11A). While all cargos stimulated by the adaptor was more significant to the internalized similarly, the cargo without an EP (MBP-CBS) internalization process than the CPP effects of the cargo-EP internalized most readily. MBP-AUR internalized the least peptides. when non-background signal was considered. Clearly with the Given the beneficial role of positive charge during the TAT-NMR-CaM adaptor, specific internalization had become internalization process and recognition that the CaM robust enough that intrinsic internalization was no longer a portion of TAT-CaM was acidic and inhibited major factor. Frontiers in Pharmacology 12 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 9 550-MBP-AUR and 550-MBP-HA2 internalized with adaptor set including controls as in prior 2 figures. Equimolar adaptor/cargo complexes (1 µm) were internalized in growth media for 1 h prior to washing (PBSx2) and imaging in media containing NucBlue. Internalization of MBP-AUR and MBP-HA2 with the adaptors as above in both the upper panels (A,C) and the lower panels (B,D). For presentation, intensities within each set adjusted identically but with the no-adaptor reference at approximately equal intensity for all panels. Frontiers in Pharmacology 13 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 10 Direct comparison of internalization efficiencies for five 550-labeled MBP-EP-CBS cargos: MBP-CBS, MBP-AUR, MBP-GALA, MBP-LAH4, and MBP-HA2. Cargo alone and adaptor/cargo complexes (1 µm) were internalized in growth media for 1 h prior to washing (PBS x2) and imaging in media containing NucBlue. Internalization of the 5 substrates was concurrently analyzed in (A) with no adaptor present, in (B) with TAT-CaM and in (C) with TAT-LAH4-CaM. The individual experiments represented by each column were independently adjusted to accent pattern differences within each column (for direct comparison of adaptors see Figures 7–9). Frontiers in Pharmacology 14 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 11 Direct Comparison of internalization efficiencies for the MBP-EP-CBS cargos internalized by two next generation adaptors. (A) Internalization of the five 550-labeled MBP-EP-CBS cargos by TAT-NMR-CaM (A) and pGFP-CaM (B) under the same conditions used with TAT-EP-CaM adapters. Internalization of these 5 cargos with (C) TAT-NMR-CaM and (D) pGFP-CaM analyzed in the later part of a 1 h internalization incubation when the complex is present in solution. Frontiers in Pharmacology 15 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 the only factor driving internalization, then MBP-AUR and MBP-HA2 should have increased as well. However, relative internalization was clearly lower. Additional evidence more dramatically demonstrated the unusual behavior of the MBP-AUR and MBP-HA2 cargos. When EP-cargo internalization by TAT-NMR-CaM (Figure 11C) and pGFP-CaM (Figure 11D) was analyzed with adaptor/cargo complexes still present, internalization patterns of MBP-AUR and MBP-HA2 were strikingly different from other cargos. Particularly for MBP-AUR, little intracellular cargo was apparent and there was a black space in the cell center. Indeed, the lack of intracellular MBP-AUR internalized by pGFP-CaM during treatment (Figure Fig11D) did not appear consistent with that present after washing (Figure 11B) in these concurrent experiments, suggesting a wave of internalization associated with the washing event. In contrast to the other adaptors, the fluorescence of the pGFP-CaM allowed simultaneous analysis of adaptor and cargo internalization. When imaged with adaptor/cargo complexes present at a concentration of 1 μm, colocalization of pGFP- CaM and cargos again broke into two groups with MBP-AUR and MBP-HA2 producing much lower levels of internalized cargo apparent as green cells (Figure 12A). The cell surface staining observed with MBP-CBS is a characteristic of this cargo that occurred at high complex concentration, a situation that also appeared to produce a wave of internalization associated with the washing event. However, when the concentration of pGFP-CaM/cargo complexes was reduced 5-fold (Figure 12B), the MBP-CBS pattern both increased in intensity and assumed a similar appearance to that of MBP-GALA and MBP-LAH4. In contrast, both MBP-AUR and MBP-HA2 retained their distinctive green appearance, although modest internalization of pGFP-CaM/MBP-AUR became more apparent. Nevertheless, the relative ineffectiveness of both EP and next generation adaptors at inducing MBP-AUR and MBP-HA2 internalization strongly suggests that intrinsic properties of the EP sequences are responsible for the differences. Conclusion FIGURE 12 Colocalization of MBP-EP-CBS cargos and pGFP-CaM The primary conclusion of these studies is that cargo during comparison of the internalization of the five MBP-EP-CBS characteristics often play an important role in CPP associated cargos at two concentrations. Simultaneous internalization of the five 550-labeled MBP-EP-CBS cargos with equimolar pGFP- internalization. Further, net charge may be a dominating CaM at complex concentrations of (A) 1000 nm or (B) 200 nm. As characteristic governing CPP-adaptor mediated internalization in the prior 5 figures, equimolar adaptor/cargo complexes were internalized in growth media for 1 h prior to washing (PBS x2) and efficiency. Our recent experience suggests that cargo-induced imaging in media containing NucBlue. increases in complex internalization represent the typical behavior observed with cargo proteins having a net charge ranging from positive to neutral. A dramatic example of the ability of charge to increase intrinsic cargo internalization can be When pGFP-CaM was used to internalize the EP-cargo set, seen in supercharged GFP, which was created by mutating MBP-CBS and MBP-GALA joined MBP-LAH4 as the cargos that surface residues to R or K and produced a powerful cell internalized most readily (Figure 11B). If adaptor efficacy were penetrating protein capable of delivering a covalently attached Frontiers in Pharmacology 16 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 cargo (Lawrence et al., 2007; McNaughton et al., 2009). This consistent with the nearly complete block of MBP-AUR and supercharged GFP was of course used in the design of pGFP- pGFP-CaM internalization apparent in images taken while the CaM and produced an adaptor that appears to be a powerful CPP complex was present in the medium. This suggests that a wave of with most cargos. internalization accompanies washing, a phenomenon that The experiments presented here represent a fraction of the appears fairly common. We have avoided experiments, usually conditions, cargos and methodologies used during our efforts to involving high complex concentrations, where adaptor/cargo understand why some cargos including Cas9 displayed no TAT- complexes strongly display this behavior because this issue is CaM-specific increase in internalization. While the ability of beyond the scope of the current study. Significantly, a reduction positive charge to enable CPP activity is well established, the in pGFP-CaM/MBP-AUR concentration increased dominating effect cargo charge can have on this process seems to internalization during the incubation period. However, for this be an understudied topic. Between that space where the CPP cargo, inhibition of entry remains readily apparent. dominates the internalization process and where the adaptor has The TAT-CaM adaptor was obviously not designed based on marginal impact on strong intrinsic cargo internalization may lie the fact that CaM was a very negative protein that could isolate the general case where both adaptor (CPP) and cargo are CPP sequence efficacy. As the effect the negative CaM was having positively impacting membrane association and on internalization became apparent (see data on the TAT-less internalization. This is not particularly surprising and CaM control), this characteristic was regarded as an unequivocal represents a specific case of what has been termed the A-B negative. However, recognition that CPP-EP-CaM fusions site problem (Jencks, 1981). When two binding sites can provided adaptors whose internalization ability was based independently associate with a target, two relatively weak sites almost entirely on the CPP characteristics has proved can generate binding beyond that expected from addition of their valuable. In combination with cargos based on intrinsically individual free energy because the first binding event pays much negative MBP, which also provides little propensity for of the entropic price of association, increasing the enthalpic internalization, complexes can be created in which CPP and benefit of the second site association. Clearly, cargo EP sequences can be combinatorially tested to improve characteristics including charge seriously complicate internalization and probably cargo escape. In this study, we investigation of CPP-adaptor function and behavior. This is focused entirely on effects the 4 EP sequences had on not a flaw in the adaptor concept but rather a characteristic of internalization because it was amenable to confocal analysis. CPP/cargo internalization events. Indeed, a CPP-less control The surprising benefits of the LAH4 as a CPP, apparent in both cargo is an important control even for fused CPP-cargo the TAT-LAH4-CaM adaptor and the LAH-MBP cargo, suggest proteins. Clearly, cargo-induced internalization complicates the sequence will have general utility in stimulating cell interpretation of the existing CPP literature based on covalent penetration. Clearly, adding the LAH4 sequence to TAT- association of CPPs and cargos. NMR-CaM and GFP-CaM provides one potential avenue for establishing the utility of this sequence in creating even more There are practical considerations that are unique to and essential for the interpretability of CPP-adaptor/CBS-cargo powerful adaptors. experiments. In general, the primary goal of CPP The negative consequences the AUR and HA2 sequences on internalization is to increase cargo entry. However, when most cargo internalization may not recommend their use as CPPs but of the basis of cargo entry resides in characteristics of the cargo does not preclude their utility as EPs with the right CPP-adaptor. (e.g., positive charge), unexpected outcomes can occur. These Indeed, given the striking phenotype associated with inclusion of include not only an inability of the adaptor to increase adaptor/ these relatively short sequences in the MBP cargos, it is tempting cargo internalization, but even adaptor inhibition of the to speculate that their previously identified EP properties are internalization process. When the adaptor inhibits complex responsible. Because the TAT-EP-CaM/MBP-EP model internalization, as occurred in some situations with TAT- produced relatively low internalization and because endosomal CaM, the adaptor was used in excess of cargo, so that all escape is not easily captured with confocal microscopy, internalizing cargo was bound by the inhibitory adaptor. Of investigation of combinatorial endolytic properties of EP course, the reciprocal is also the case, when the adaptor drives sequences in adaptors and cargos has not been addressed internalization of inhibitory cargo, the cargo needs to be in excess here. Future experiments designed to measure endosomal for interpretation of adaptor internalization to be interpretable. escape using EPs in the adaptor context will require These factors were a consideration during CRISPR incorporation of a reporter that directly measures escape, e.g. internalization and an excess of both the crRNA duplex and (Milech et al., 2015; Teo et al., 2021). While the ability of the TAT-CaM were used, so that essentially all Cas9 entering the cell TAT-EP-CaM/MBP-EP system to measure endosomal escape did so as part of a complete complex. The ability of the MBP- remains to be established, our adaptor/cargo strategy clearly has AUR cargo to block pGFP-CaM internalization suggests pGFP- utility in investigating the roles of CPP and EP sequences in the CaM was not in excess during that experiment. As noted, the internalization process. This system may prove quite valuable in amount of MBP-AUR apparent inside washed cells was not pursuit of even more effective CPPs that can deliver large Frontiers in Pharmacology 17 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 macromolecular cargoes with all the therapeutic potential State University College of Science and Mathematics. MW was implicit in CPP capability. supported by the PHS grant R25GM111655. Data availability statement Acknowledgments The datasets presented in this study can be found in online We are grateful to Drs. Carol Chrestensen and Scott Nowak repositories. The names of the repository/repositories and for helpful discussions and technical assistance. accession number(s) can be found in the article/ Supplementary Material. Conflict of interest Author contributions The authors declare that the research was conducted in the absence of any commercial or financial relationships that could DM designed experiments, performed most of the be construed as a potential conflict of interest. experimentation, analyzed the data and wrote large portions of the manuscript. LS and SH performed extensive experimentation and analyzed data for several of the Publisher’s note constructs. MW developed the EP work, designed and produced the EP-containing adaptors and cargos. MM All claims expressed in this article are solely those of the authors developed the tamavidin work, designed and produced the and do not necessarily represent those of their affiliated organizations, Tam constructs and performed the nucleic acid delivery or those of the publisher, the editors and the reviewers. Any product experiments. JM is principal investigator of the group, that may be evaluated in this article, or claim that may be made by its oversaw all research, helped analyze and interpret the data, manufacturer, is not guaranteed or endorsed by the publisher. and wrote much of the manuscript. Supplementary material Funding The Supplementary Material for this article can be found This work was supported by the U.S. PHS grants to JM online at: https://www.frontiersin.org/articles/10.3389/fphar. (R15EB028609) and DM (R15GM129711) and the Kennesaw 2022.1070464/full#supplementary-material References Akita, H.,Masuda,T., Nishio,T., Niikura, K.,Ijiro, K., andHarashima,H.(2011). Cargo complexes: A study on the effect of net cargo charge and length. Cells 11 (7), Improving in vivo hepatic transfection activity by controlling intracellular trafficking: the 1195. doi:10.3390/cells11071195 function of GALA and maltotriose. Mol. Pharm. 8 (4), 1436–1442. doi:10.1021/mp200189s Jencks, W. P. (1981). On the attribution and additivity of binding energies. Proc. Albrecht, T., Zhao, Y., Nguyen, T. H., Campbell, R. E., and Johnson, J. D. (2015). Natl. Acad. Sci. U. S. A. 78 (7), 4046–4050. doi:10.1073/pnas.78.7.4046 Fluorescent biosensors illuminate calcium levels within defined beta-cell endosome Johnson, R. M., Harrison, S. D., and Maclean, D. (2011). Therapeutic applications subpopulations. Cell Calcium 57 (4), 263–274. doi:10.1016/j.ceca.2015.01.008 of cell-penetrating peptides. Methods Mol. Biol. 683, 535–551. doi:10.1007/978-1- Alves, I. D., Jiao, C.-Y., Aubry, S., Aussedat, B., Burlina, F., Chassaing, G., et al. 60761-919-2_38 (2010). Cell biology meets biophysics to unveil the different mechanisms of Lawrence, M. S., Phillips, K. J., and Liu, D. R. (2007). Supercharging proteins can penetratin internalization in cells. Biochim. Biophys. Acta 1798 (12), 2231–2239. impart unusual resilience. J. Am. Chem. Soc. 129 (33), 10110–10112. doi:10.1021/ doi:10.1016/j.bbamem.2010.02.009 ja071641y Aqeel, A., and Khan, J. M. (2022). pH-sensitive endosomolytic peptides in gene Lecher, J. C., Nowak, S. J., and McMurry, J. L. (2017). Breaking in and busting out: and drug delivery: Endosomal escape and current challenges. J. Drug Deliv. Sci. Cell-penetrating peptides and the endosomal escape problem. Biomol. Concepts 8 Technol. 76, 2231–103786. doi:10.1016/j.jddst.2022.103786 (3-4), 131–141. doi:10.1515/bmc-2017-0023 Derakhshankhah, H., and Jafari, S. (2018). Cell penetrating peptides: A concise LeCher, J. C., Didier, H. L., Dickson, R. L., Slaughter, L. R., Bejerano, J. C., Ho, S., review with emphasis on biomedical applications. Biomed. Pharmacother. 108, et al. (2020). Utilization of a cell-penetrating peptide-adaptor for delivery of HPV 1090–1096. doi:10.1016/j.biopha.2018.09.097 protein E2 into cervical cancer cells to arrest cell growth and promote cell death. Fonseca, S. B., Pereira, M. P., and Kelley, S. O. (2009). Recent advances in the use biorxiv 8 (3-4), 131–141. doi:10.1101/2020.01.31.928358 of cell-penetrating peptides for medical and biological applications. Adv. Drug Lönn, P., and Dowdy, S. F. (2015). Cationic PTD/CPP-mediated macromolecular Deliv. Rev. 61 (11), 953–964. doi:10.1016/j.addr.2009.06.001 delivery: Charging into the cell. Expert Opin. Drug Deliv. 12 (10), 1627–1636. doi:10. Gentry, S. B., Nowak, S. J., Ni, X., Hill, S. A., Wade, L. R., Clark, W. R., et al. 1517/17425247.2015.1046431 (2021). A real-time assay for cell-penetrating peptide-mediated delivery of McMurry, J. L., Chrestensen, C. A., Scott, I. M., Lee, E. W., Rahn, A. M., Johansen, molecular cargos. PLOS ONE 16 (9), e0254468. doi:10.1371/journal.pone.0254468 A. M., et al. (2011). Rate, affinity and calcium dependence of CaM binding to eNOS Hymel, H. C., Rahnama, A., Sanchez, O. M., Liu, D., Gauthier, T. J., and Melvin, and nNOS: Effects of phosphorylation. FEBS J. 278 (24), 4943–4954. doi:10.1111/j. A. T. (2022). How cargo identity alters the uptake of cell-penetrating peptide (CPP)/ 1742-4658.2011.08395.x Frontiers in Pharmacology 18 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 McNaughton, B. R., Cronican, J. J., Thompson, D. B., and Liu, D. R. (2009). Takakura, Y., Tsunashima, M., Suzuki, J., Usami, S., Kakuta, Y., Okino, N., et al. Mammalian cell penetration, siRNA transfection, and DNA transfection by (2009). Tamavidins--novel avidin-like biotin-binding proteins from the supercharged proteins. Proc. Natl. Acad. Sci. U. S. A. 106 (15), 6111–6116. Tamogitake mushroom. Febs J. 276 (5), 1383–1397. doi:10.1111/j.1742-4658. doi:10.1073/pnas.0807883106 2009.06879.x Milech, N., Longville, B. A., Cunningham, P. T., Scobie, M. N., Bogdawa, H. M., Takakura, Y., Oka, N., Kajiwara, H., Tsunashima, M., Usami, S., Tsukamoto, H., Winslow, S., et al. (2015). GFP-complementation assay to detect functional CPP et al. (2010). Tamavidin, a versatile affinity tag for protein purification and and protein delivery into living cells. Sci. Rep. 5, 18329. doi:10.1038/srep18329 immobilization. J. Biotechnol. 145 (4), 317–322. doi:10.1016/j.jbiotec.2009.12.012 Moulay, G., Leborgne, C., Mason, A. J., Aisenbrey, C., Kichler, A., and Bechinger, Takakura, Y., Sofuku, K., and Tsunashima, M. (2013). Tamavidin 2-REV: an B. (2017). Histidine-rich designer peptides of the LAH4 family promote cell delivery engineered tamavidin with reversible biotin-binding capability. J. Biotechnol. 164 of a multitude of cargo. J. Pept. Sci. 23 (4), 320–328. doi:10.1002/psc.2955 (1), 19–25. doi:10.1016/j.jbiotec.2013.01.006 Ngwa, V. M., Axford, D. S., Healey, A. N., Nowak, S. J., Chrestensen, C. A., and Teo, S. L. Y., Rennick, J. J., Yuen, D., Al-Wassiti, H., Johnston, A. P. R., and McMurry, J. L. (2017). A versatile cell-penetrating peptide-adaptor system for Pouton, C. W. (2021). Unravelling cytosolic delivery of cell penetrating peptides efficient delivery of molecular cargos to subcellular destinations. PloS one 12 (5), with a quantitative endosomal escape assay. Nat. Commun. 12 (1), 3721. doi:10. e0178648. doi:10.1371/journal.pone.0178648 1038/s41467-021-23997-x Salerno, J. C., Ngwa, V. M., Nowak, S. J., Chrestensen, C. A., Healey, A. N., and Thompson, D. B., Villaseñor, R., Dorr, B. M., Zerial, M., and Liu, D. R. (2012). McMurry, J. L. (2016). Novel cell-penetrating peptide-adaptors effect intracellular Cellular uptake mechanisms and endosomal trafficking of supercharged proteins. delivery and endosomal escape of protein cargos. J. Cell Sci. 129 (5), 893–897. doi:10. Chem. Biol. 19 (7), 831–843. doi:10.1016/j.chembiol.2012.06.014 1242/jcs.182113 Wharton, S. A., Martin, S. R., Ruigrok, R. W., Skehel, J. J., and Wiley, D. C. (1988). Shahmiri, M., Enciso, M., and Mechler, A. (2015). Controls and constrains of the Membrane fusion by peptide analogues of influenza virus haemagglutinin. J. Gen. membrane disrupting action of Aurein 1.2. Sci. Rep. 5, 16378. doi:10.1038/srep16378 Virol. 69 (Pt 8), 1847–1857. doi:10.1099/0022-1317-69-8-1847 Skotland, T., Iversen, T. G., Torgersen, M. L., and Sandvig, K. (2015). Cell- Yokoo, H., Oba, M., and Uchida, S. (2021). Cell-penetrating peptides: Emerging penetrating peptides: Possibilities and challenges for drug delivery in vitro and in tools for mRNA delivery. Pharmaceutics 14 (1), 78. doi:10.3390/ vivo. Molecules 20 (7), 13313–13323. doi:10.3390/molecules200713313 pharmaceutics14010078 Frontiers in Pharmacology 19 frontiersin.org http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Frontiers in Pharmacology Pubmed Central

A reversible cell penetrating peptide-cargo linkage allows dissection of cell penetrating peptide- and cargo-dependent effects on internalization and identifies new functionalities of putative endolytic peptides

Loading next page...
 
/lp/pubmed-central/a-reversible-cell-penetrating-peptide-cargo-linkage-allows-dissection-YhiUNBFAI0

References (58)

Publisher
Pubmed Central
Copyright
Copyright © 2022 Morris, Snipes, Hill, Woods, Mbugua, Wade and McMurry.
eISSN
1663-9812
DOI
10.3389/fphar.2022.1070464
Publisher site
See Article on Publisher Site

Abstract

TYPE Original Research PUBLISHED 21 November 2022 DOI 10.3389/fphar.2022.1070464 A reversible cell penetrating peptide-cargo linkage allows OPEN ACCESS EDITED BY Nian-Qiu Shi, dissection of cell penetrating Jilin Medical University, China REVIEWED BY peptide- and cargo-dependent Jing Sun, Geneleap Biotech, Luye Pharm, China Ji Li, effects on internalization and University of Michigan, United States *CORRESPONDENCE identifies new functionalities of Jonathan L. McMurry, jmcmurr1@kennesaw.edu putative endolytic peptides SPECIALTY SECTION This article was submitted to Experimental Pharmacology Daniel P. Morris, Lucy C. Snipes, Stephanie A. Hill, and Drug Discovery, a section of the journal Michael M. Woods, Maria M. Mbugua, Lydia R. Wade and Frontiers in Pharmacology Jonathan L. McMurry* RECEIVED 14 October 2022 ACCEPTED 09 November 2022 Department of Molecular and Cellular Biology, Kennesaw State University, Kennesaw, GA, PUBLISHED 21 November 2022 United States CITATION Morris DP, Snipes LC, Hill SA, Cell penetrating peptides (CPPs) are a promising technology for therapeutic Woods MM, Mbugua MM, Wade LR and McMurry JL (2022), A reversible cell delivery of macromolecular cargos. CPPs have generally used covalent linkages penetrating peptide-cargo linkage to cargo, ensuring a common fate as one molecule. Conversely, our CPP- allows dissection of cell penetrating peptide- and cargo-dependent effects adaptor, TAT-CaM, noncovalently binds calmodulin binding sequence (CBS)- on internalization and identifies new containing cargos in calcium rich media then dissociates in the calcium-poor functionalities of putative endosomal environment following internalization, enhancing endosomal endolytic peptides. Front. Pharmacol. 13:1070464. escape relative to standard CPPs. In this study, we report cell entry of doi: 10.3389/fphar.2022.1070464 positively charged protein cargos that were not increased by TAT-CaM while COPYRIGHT cargos based on the negatively charged maltose binding protein (MBP) © 2022 Morris, Snipes, Hill, Woods, displayed little intrinsic internalization but were internalized by TAT-CaM. In Mbugua, Wade and McMurry. This is an open-access article distributed under addition, association of positively charged proteins with negatively charged the terms of the Creative Commons nucleic acids reduced internalization. This evidence points to the dominant role Attribution License (CC BY). The use, distribution or reproduction in other cargo charge plays in apparent CPP effectiveness. There has been little forums is permitted, provided the systematic investigation as to how interaction between CPPs and cargos original author(s) and the copyright impacts internalization efficiency. Our adaptors provide a tool that allows owner(s) are credited and that the original publication in this journal is combinatorial assays to detect emergent properties. Toward this end we cited, in accordance with accepted added 4 endolytic peptide (EP) sequences between cargo CBS and MBP academic practice. No use, distribution moieties to create 4 new cargos and between TAT and CaM to create or reproduction is permitted which does not comply with these terms. 4 new adaptors. The new cargos were assayed for internalization alone and with a panel of CPP-adaptors to identify combinations that displayed increased internalization efficiency or other properties. Among the most important results, addition of the EP LAH4 improved adaptor performance and provided some CPP capability to cargos. MBP-LAH4-CBS was internalized more effectively by most adaptors, suggesting this sequence has general stimulatory ability. Two other EPs, Aurein 1.2 and HA2, also provided some CPP capability to their MBP cargos but were unexpectedly antagonistic to internalization by most adaptors due to retention of adaptor/cargo complexes Frontiers in Pharmacology 01 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 on the cell surface. We thus identified LAH4 as stimulator of internalization in both adaptors and cargos and uncovered new functionality for Aurein 1.2 and HA2, which may be related to their identification as EPs. Future experiments will test new endolytic capabilities made possible with combinatorial approaches. KEYWORDS cell-penetrating peptides, protein transduction domains (PTD), endolytic peptides, endocytosis, calmodulin (CAM) Introduction cargo proteins in all cell lines that we examined. Subsequent efforts to develop CPP-based applications ensued, some of which Cell-penetrating peptides (CPPs), sometimes called protein were very successful, e.g. we delivered human papillomavirus transduction domains (PTDs), have long held great promise for E2 protein to cervical cancer cells, inducing senescence and thus overcoming failures of biomolecule therapeutic leads due to demonstrating that cargos remained folded and active even after bioavailability issues relating to failure to cross membranes delivery (LeCher et al., 2020). (Fonseca et al., 2009; Johnson et al., 2011; Lönn and Dowdy, In the course of efforts to develop applications, particularly 2015; Derakhshankhah and Jafari, 2018). Though the specific co-delivery of Cas9 and crRNAs to effect CRISPR/Cas genome endocytic mechanisms by which they do so remain a subject of editing by exogenous delivery of the components, we noticed a debate, CPPs are capable of mediating penetration of the plasma correlation between charge of the cargo and efficacy of TAT- membranes of mammalian cells by molecules to which they are CaM-mediated delivery, the more negative the cargo, the worse coupled, allowing delivery of “cargos” to cell interiors, a the delivery. We also observed that positively charged cargo potentially transformative platform technology for drug proteins were taken up themselves, sometimes as well as or even delivery and other applications. However, development of better than when coupled to TAT-CaM. Noting that most CPPs CPP-based delivery systems has been hindered because cargos are positively charged and calmodulin is itself rather extensively are normally coupled to CPPs by covalent or other irreversible negatively charged (net −24, TAT-CaM with its positive CPP and linkages and become entrapped in endosomes after cellular entry linker sequences has a charge of −16), we sought to develop next- rather than reaching the cytoplasm and their subsequent targets generation CPP-adaptors that might better utilize more favorable (Lecher et al., 2017). To address this problem, we have employed charge states to deliver cargos, the first of which incorporated a high affinity, reversible noncovalent coupling strategy, calmodulin from naked mole rat, Heterocephalus glaber. Naked attaching cargos to CPPs via a CPP-containing “adaptor” mole rat CaM (NMR-CaM) has a positively charged domain of protein, calmodulin. Our prototype CPP-adaptor, TAT- unknown function N-terminal to the EF hand domain, which is Calmodulin (TAT-CaM), consists of the cell penetrating invariant from that of human CaM. The resultant adaptor, TAT- moiety from HIV transactivator of transcription and human NMR-CaM, has a lesser net negative charge (−5) and evinced calmodulin (Salerno et al., 2016; Ngwa et al., 2017). TAT-CaM exactly the behavior we predicted, delivering cargos that had binds CaM binding-site (CBS) containing cargos with nM resisted efficient TAT-CaM-mediated delivery, e.g. polyA affinity in the presence of calcium but negligibly in its binding protein PAB-1 (Gentry et al., 2021). absence. The CBS can be located anywhere in the cargo. Although not a basis for early adaptor design, we realized that 2+ TAT-CaM/cargo complexes form spontaneously in the Ca - the CPP-adaptor strategy afforded the ability to directly compare 2+ containing extracellular milieu but, concurrent with Ca flux internalization of cargo alone with that of the CPP-adaptor/cargo during early endocytosis (Albrecht et al., 2015), cargos dissociate complex, an advantage over examinations of cargo charge done from CPP-adaptors within endosomes and are released to the with traditional covalent coupling strategies, e.g (Hymel et al., cytoplasm though the CPP-adaptor remains trapped. 2022). It was the availability of this comparison that led to the Our first reports of the success of our reversible coupling realization that cargo charge dramatically changed the properties strategy (Salerno et al., 2016; Ngwa et al., 2017) were attended by of TAT-CaM/cargo complexes. In support of this idea we report high hopes for a generally utile delivery vehicle adaptable not herein a body of experimental data using positive and negative only to any desired protein but nucleic acids and perhaps other cargos as well as adaptors designed to have a more positive net biomolecules as well. Possible cargos are limited only by the charge. requirement of attaching a CBS to the cargo, which can be Another advantage of the adaptor/cargo model is the ability achieved recombinantly for proteins, with covalent chemistry to change the characteristics of the adaptor and the cargo for nucleic acids, or via a secondary adaptor strategy such as independently. This allows for the improved CPP using a recombinant CBS-biotin binding protein to deliver characteristics of new adaptors without impact on the released biotinylated nucleic acids. In some respects, those hopes were cargos’ ability to leave the endosome. Similarly, adaptor delivery realized as we showed that TAT-CaM worked for multiple model systems can be modified to effect more efficient cytoplasmic Frontiers in Pharmacology 02 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 delivery of cargos by inclusion of endolytic peptide sequences eliminate some extraneous sequences originally designed to (EPs) (also called “endosomolytic peptides” or “endosomal facilitate cloning; however, the His-tag, TAT and CaM escape domains”)(Aqeel and Khan, 2022). Some EPs are sequences are invariate from the original study (Salerno et al., thought to destabilize endosomes in a pH-dependent manner 2016). All references to TAT-CaM in this paper mean TAT- by binding and disrupting membranes during the acidification CaM 2.0. that occurs in early transport, e.g. haemagglutinin-derived peptides (Wharton et al., 1988). Toward this end we added four endolytic peptide (EP) sequences: Aurein 1.2, GALA, Expression, purification, and labeling HA2 and LAH4 (Wadia et al., 2004; Akita et al., 2011; Shahmiri et al., 2015; Moulay et al., 2017) to cargos between Proteins were expressed essentially as described (Salerno the CBS and negatively charged maltose binding protein (MBP) et al., 2016; Ngwa et al., 2017) with minor modifications. to create 4 new cargos and between TAT and CaM to create Briefly, plasmids were freshly transformed into BL21 (DE3) 4 new adaptors to assess internalization as a function of adaptor pLysS. Overnight cultures grown from single colonies were and cargo charge. Because both CaM and MBP have intrinsic subcultured into 1L Luria-Bertani broth and grown with negative charge, neither have significant intrinsic internalization vigorous shaking at 37 C. At OD ~0.4, the temperature was behavior, and the effects of the EPs are manifest without lowered to 30 C, cells were induced with 0.2 mm IPTG and complication from competing behaviors. While EPs are growth continued for 4 hours. The procedure was altered slightly supposed to increase escape, an increase in intracellular for MBP-CBS cargos, the cells for which were grown in Terrific concentration may also be due to advantageous CPP-like Broth (TB) supplemented with 0.25% (w/v) glucose and behavior. In this report we have focused on the clear and induction was conducted at 32 C. Cells were harvested by differential impact these EPs have on internalization. centrifugation at 10,000 × g and frozen at −80 C. Purification was also performed essentially as described via immobilized metal affinity chromatography (Salerno et al., 2016). Materials and methods Briefly, cell pellets were thawed on ice, resuspended in lysis buffer (50 mm Tris pH 8, 500 mm NaCl, 10 mm imidazole, 10% Plasmids, strains, and cell lines glycerol and 6 mm β-mercaptoethanol for disulfide containing proteins). 1 mg/ml DNAse and 0.25 mg/ml lysozyme were added The E. coli strain used in this study, BL21 (DE3)pLysS was during resuspension. For TAT-NMR-CaM only, Halt Protease propagated from purchased cells from EMD Millipore Inhibitor Cocktail (ThermoFisher, Waltham, MA, United States) (Burlington, MA, United States) or other established supplier. was added to 1x per manufacturer’s protocol. Cells were broken Baby hamster kidney (BHK) cells (#CCL-10) and HEK-293 cells via passage through a French press at 20,000 psi and subjected to (CRL-1573) were purchased from ATCC and cultured in centrifugation at ~27,000 × g to pellet unbroken cells and debris. Dulbecco’s Modified Eagles’ Medium with GlutaMAX Clarified lysate was passed over a cobalt affinity column using an Supplement (Gibco, United States) and 5% or 10% fetal FPLC system while monitoring A , washed with wash buffer bovine serum. (equivalent to lysis buffer with 25 mm imidazole instead of Plasmids used in this study were previously described 10 mm) until baseline absorbance was attained, after which (Salerno et al., 2016; Ngwa et al., 2017; Gentry et al., 2021)or protein was eluted in elution buffer (lysis buffer with 250 mm constructed as described in Supplementary Figure S1 from parent imidazole). Protein-containing fractions were pooled, vectors pET19b and pET22b (EMD Millipore, United States), concentrated and exchanged by passage over a desalting pMalc-5x (New England Biolabs, United States) or pCal-N- column into 10 mm HEPES, pH 7.4, 150 mm NaCl, 10% FLAG (Agilent Technologies, United States). For example, glycerol, 1 mm CaCl for fluorescence labeling, biotinylation pJM161, encoding TAT-naked mole rat calmodulin (TAT- or other further use. Quantitation was done using Bradford NMR-CaM) consisted of an E. coli-optimized synthetic gene Assay with bovine serum albumin as standard. For (GeneScript, Piscataway, NJ, United States) cloned into NdeI and fluorescence labeling, DyLight 488, 550 or 650 NHS Esters BamHI sites in pET19b with an in-frame stop codon prior to the (ThermoFisher) were used to introduce fluorescent labels BamHI site. The encoded TAT-NMR-CaM protein consists of under conditions that resulted in incorporation efficiencies the TAT peptide sequence (YGRKKRRQRRR) N-terminally below 0.6 dyes/molecule as determined using the algorithm fused to Heterocephalus glaber (naked mole rat) calmodulin suggested by the manufacturer (ThermoFisher). Dye removal (GenBank: EHB02604.1) (Gentry et al., 2021). A vector- columns were then used to remove unreacted dye as encoded 10xHis tag containing an enterokinase cleavage site recommended by the manufacturer (ThermoFisher). and several spacer residues is N-terminal to TAT. The version of All proteins were assayed by biolayer interferometry to assure TAT-CaM used in this study, initially designated “TAT-CaM high affinity (low nm-to-high pm) binding between adaptors and 2.0” (Ngwa et al., 2017), is a variant of a prior version modified to CBS-cargos in the presence of calcium and fast dissociation upon Frontiers in Pharmacology 03 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 TABLE 1 Schematic descriptions of proteins used in this study; A, CPP-adaptors and adaptor controls; B, cargo proteins. Additional details including amino acid sequences, spacers and tags are described in Supplementary Information. A Name Description (N-to-C) Parent vector References TAT-CaM (2.0) His-TAT-CaM pET19b Ngwa et al. (2017) CaM Cntl His-CaM (no TAT) pET19b This study TAT-Aur-CaM His-TAT-Aurein 1.2-CaM pET19b This study TAT-GALA-CaM His-TAT-GALA-CaM pET19b This study TAT-LAH4-CaM His-TAT-LAH4-CaM pET19b This study TAT-HA2-CaM His-TAT-HA2-CaM pET19b This study TAT-NMR-CaM His-TAT-naked mole rat CaM pET19b Gentry et al. (2021) pGFP-CaM His-GFP (+36)-CaM pET19b Lawrence et al. (2007), This study CBS-Tam CBS-Tamavidin (C627S)-His pET22b This study MBP-CBS MBP-CBS-His pMalc-5x This study MBP-AUR MBP-Aurein 1.2-CBS-His pMalc-5x This study MBP-GALA MBP-GALA-CBS-His pMalc-5x This study MBP-LAH4 MBP-LAH4-CBS-His pMalc-5x This study MBP-HA2 MBP-HA2-CBS-His pMalc-5x This study Cas9-NLS His-CBS-Cas9-NLS pET19b This study Cys-less Cas9-NLSx2 His-CBS-NLS-Cas9-NLS (C80S, C574S) pET19b This study its removal with EDTA. All evinced kinetics and affinities highly 30 min of use. Following assembly, complexes were transferred to similar to CaM with a natural binding partner (McMurry et al., room temperature for 5–10 min and then diluted into media also 2011) and within the ranges observed for previously reported at room temperature. After 1–3 min complex containing media adaptor-cargo pairings (Salerno et al., 2016; Ngwa et al., 2017; was microcentrifuged at maximum velocity (>15,000 rcf) for Gentry et al., 2021). Table 1 lists all adaptors and cargos used in 1 min to remove precipitates, briefly warmed in a 37 C bath this study and lists the plasmids from which they were expressed. (2–4 min) and transferred onto BHK cells from which growth media had just been removed. Unless otherwise stated, cells were treated with the indicated concentrations of complex at 37 C 2+ Cell culture under 5% CO for 1 h, washed twice with PBS (containing Ca 2+ and Mg ) and analyzed in imaging media usually comprised of BHK cells were maintained in a 37 C, 5% CO environment growth media without phenol red but including NucBlue, Live in growth media consisting of DMEM (GlutaMax, (ThermoFisher). Aliquots of proteins containing cysteine ThermoFisher) containing +4.5 g/L D-glucose and 1.9 mm residues (Cas9-NLS and TAT-NMR-CaM) were thawed after 2+ Ca with no sodium pyruvate) supplemented with 4 mm addition of 1/20th volume of 1 mm DTT in storage buffer, which glutamine and 5% or 10% fetal bovine serum (FBS) as became less than 5 µm during cell treatment. indicated. Cells were replated in coverslip slides in the same Internalization of TAT-CaM alone was assayed using media 20–24 h before use in cell penetration assays. DyLight 488 and 550 labeled adaptor modified with less than 0.3 dyes per protein. TAT-CaM in storage buffer was diluted 10- fold with imaging media and transferred onto BHK cells. TAT- Cell penetration assays CaM internalization of Cas9 used various labels on both proteins, as well as an optional treatment with green CMFDA CellTracker Internalization assays were performed as described (Salerno (Molecular Probes). When CellTracker was used, treatments et al., 2016; Ngwa et al., 2017) with minor modifications. Protein were followed by a PBS wash and a 20 min incubation with aliquots in a common storage buffer (10 mm HEPES, pH 7.4, PBS (mock) or CMFDA CellTracker in PBS, as indicated. TAT- 150 mm NaCl, 10% glycerol, 1 mm CaCl ) were thawed and CaM internalization experiments utilizing the “cysless” CBS- sequentially assembled into complexes in this buffer on ice within Tam were performed with various concentrations of an Frontiers in Pharmacology 04 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 1 By itself, TAT-CaM internalization requires concentrations above 1 µm. BHK cells were incubated with (A,B) Dylight 550 or (C,D) Dylight 448 labeled TAT-CaM, respectively, for 60 or 30 min at 1 µm or 10 μm, in DMEM with 10% FBS under a 5% CO2 atmosphere at 37 C. Cells were then 2+ 2+ washed 3 times with PBS (with Mg and Ca ) and imaged in media without phenol red but containing NucBlue. Dye incorporation averaged less than 0.3 dyes per TAT-CaM molecule. oligonucleotide (CCATCCTGGTCGAGCTGGACGGCGACG) across the cell. Experiments used either a Zeiss LSM 700 or LSM labeled with a 5′ Alexa 488 dye and a 3’ biotin equilibrated 900 confocal microscope generally at 400× with settings adjusted with CBS-Tam prior to complexation with the CPP-adaptor in to best image quality unless otherwise stated. Opti-MEM media (ThermoFisher), both preceded and followed by two washes with PBS and final addition of imaging media. Confocal imaging was initiated as soon as practical (usually Results and discussion 7–15 min following the washes) and continued for up to 1 h in an incubated chamber that maintained a humidified, 37 C, 5% CO The CPP-adaptor technology was devised to facilitate environment. In some experiments a parallel plate was initially dissociation of CPP and cargo, enabling cargo release from treated identically; however, imaging was performed with the endosomally entrapped CPPs (LeCher et al., 2017). However, complex present between 20 and 60 min after addition. Cells with another facet of the CPP-adaptor technology is the simplicity of normal nuclei and monolayer distribution that avoided cell comparative analysis of adaptor, cargo and complex stacking were selected for imaging. Images were taken at a internalization properties. In this regard the properties of our confocal plane just below the nuclear center to minimize slide CPP adaptor, TAT-CaM, are important for understanding TAT- surface background while retaining signal from intracellular CAM/cargo internalization. As observed for TAT-cargos and structures across the cell. When NucBlue was not used, CPPs in general (Lönn and Dowdy, 2015), concentrations above images were taken from a similar confocal plane above slide 1 µm are necessary for effective internalization of TAT-CaM surface background that still included intracellular structures (Figure 1). By itself TAT-CaM is a rather small protein (~26 kDa) Frontiers in Pharmacology 05 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 2 Internalization of Cas9 bound to TAT-CaM occurs readily at very low concentration. Cas9-NLS bound to equimolar TAT-CaM at 100 nm was incubated with BHK cells in DMEM with 10% FBS for 1 h prior to PBS treatment, washing and imaging. Panels show (A) 550-labeled TAT-CaM, (B) 488- labeled Cas9-NLS and (C) Composite including NucBlue. In (D) the enlarged inset (Figure 2D) shows Cas 9-NLS and TAT-CaM are often in separate but juxtaposed endosomes. whose properties may be altered by labeling; however, similar While at the time these results were mystifying, it has long internalization was observed with modification by two amino- been recognized that most CPPs are positively charged and may reactive fluorescent reagents, Dylight 550 (Figures 1A,B) and function by neutralizing negative charge on plasma membranes Dylight 488 (Figures 1C,D). To limit artifacts from labeling, (Skotland et al., 2015), which is believed to aid inward membrane TAT-CaM was modified at low efficiency (<0.3 dyes/molecule) curvature necessary for vesicle internalization (Alves et al., 2010). with reduced concentrations of the reactive dyes below the Nevertheless, we were not aware of the extent to which surface manufacturer recommendations. Indeed, for the same reason, charge of a cargo impacted CPP behavior as studies using other proteins in this report were labeled at <0.6 dye/molecule to covalently bound CPP moieties rarely test intrinsic cargo limit introduction of multiple dyes per protein monomer. internalization ability. Unexpectedly, when labeled Cas9 (the protein component of It is also known that negative charge (Hymel et al., 2022) and CRISPR) containing a CBS tag and a nuclear localization signal nucleic acids (Yokoo et al., 2021) often inhibit CPP-mediated (NLS) was internalized with equimolar 100nM TAT-CaM, the internalization. Concurrent with the Cas9 (CRISPR) TAT-CaM/Cas9 complex internalized readily at a concentration experiments, a companion project involved development of an that was more than 100-fold lower than observed for TAT-CaM adaptor-based method to introduce donor DNA alongside alone (Figure 2). In other words, cargo internalization increased CRISPR. To introduce biotin labeled nucleic acids, we created TAT-CaM internalization. Also note that TAT-CaM and a CBS-tagged a version of avidin based on the expressible fungal Cas9 often appear separated but in nearby vesicular locations homolog, tamavidin (Takakura et al., 2010; Takakura et al., (Figure 2C and enlarged inset 2D). Importantly, under most 2013). With CBS-Tamavidin (Tam), biotin labelled nucleic conditions TAT-CaM does not increase Cas9 internalization, i.e. acids can be delivered into cells as a Tam-oligo cargo complex there is a lack of TAT-CaM dependent specificity (Figures 3A,B). reversibly bound to any adaptor. This version of “cysless” CBS- On the other hand, at Cas9 concentrations below 50 nm where Tam (Tam) did not have an NLS and the lone cysteine was intrinsic internalization becomes modest, high TAT-CaM converted to serine to prevent disulfide-induced precipitation. As concentrations specifically increased Cas9 internalization avidin orthologs including tamavidin form obligate tetramers (Figure 3C). The implication of this experiment is that (Takakura et al., 2009), as many as 4 nucleic acids can be bound Cas9 internalizes more readily when bound to TAT-CaM. to a single tetramer, which will have 4 CBS sites that can bind as However, this positive effect appears insignificant when many as 4 adaptors. intrinsic Cas9 internalization is high. In addition, we have While early attempts to deliver labeled oligos clearly resulted observed other positively charged proteins used as cargos in some internalization of the TAT-CaM/CBS-Tam/oligo displayed similar intrinsic internalization (e.g. polyA binding complex, subsequent experiments showed CBS-Tam tetramers protein, Gentry and McMurry, unpublished). in the presence of a molar excess of oligos hardly enter the cell. Frontiers in Pharmacology 06 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 3 Specific and nonspecific Cas9 internalization with the TAT-CaM. BHK cells in DMEM with 10% FBS were treated for 1 h with Cas9-NLS in the absence (top) or presence (bottom) of excess TAT-CaM prior to washing (PBS x3) and imaging. Cas9-NLS at concentrations >50 nm showed no increase in internalization with inclusion of TAT-CaM as shown in examples (A) 500 nm and (B) 50 nm; however, at low Cas9-NLS concentrations including (C) 10 nm, TAT-CaM specifically increased internalization. This inconsistency was explained by decreasing the average TAT-CaM (Figure 5A). However, TAT-CaM-mediated number of oligos per tetramer, which produced the surprising internalization of MBP-CBS required a return to µm result that decreasing oligo concentration increased oligo concentrations (Figure 5B) characteristic of typical CPP- internalization (Figure 4). Maximal oligo internalization was mediated internalization processes. To test the extent to which observed at an average of 1 oligo per Tam tetramer the negative charge of CaM was inhibitory, the internalization of (Figure 4C), although some entered at 2 oligos per Tam a TAT-less CaM control (CaM-Cntl) in complex with MBP was tetramer (Figure 4B). Beyond the nearly complete block to concurrently examined and found to further reduce MBP-CBS internalization, stoichiometric Tam/oligo concentrations also internalization (Figure 5A, left panel and Figures 8–10 below). reduced complex binding to the glass coverslip. These results Importantly, comparison of TAT-CaM/MBP-CBS to (TAT-less) are consistent with increased negative charge decreasing the CaM-Cntl/MBP-CBS isolates the effect of the TAT sequence on ability of the CPP to internalize cargo. While this explanation complex internalization, a conclusion that is valid because was accepted at the time, the dramatic inhibitory effect of oligo neither CaM nor MBP contribute significantly to the on complex internalization seems to indicate the oligo is forced internalization process. into the proximity of the membrane and coverslip, suggesting Given these results, it might be expected that association of a steric hindrance may play a role. crRNA with Cas9 would decrease intrinsic internalization. To test the idea that a negatively charged cargo would display Indeed, internalization of “cysless” CBS-Cas9-NLSx2 little intrinsic internalization, we created a CBS fusion cargo (Figure 6B) was much greater than observed when it was part based on the negatively charged maltose binding protein (MBP). of the CRISPR complex (Figure 6C). Nevertheless, the CRISPR As predicted, MBP-CBS by itself displayed little tendency to complex still internalized readily (Figure 6C) and TAT-CaM still enter BHK cells, but was induced to internalize when bound to had no ability to detectably increase CRISPR internalization and Frontiers in Pharmacology 07 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 4 CBS-Tam, a cargo made from the fusion of CBS to the fungal avidin, tamavidin, internalizes model oligos but only at stoichiometries below 2 oligos per CBS-Tam tetramer. BHK cells were incubated with 600 nm TAT-CaM/Tam complexes bound to an oligo with a 5′ Alexa 488 dye and a 3′ biotin present at concentrations of (A) 600, (B) 300, (C) 150, or (D) 0 nm oligo for 1 h prior to washing (PBSx3) and imaging. in some experiments seemed to be somewhat inhibitory The strategy of using noncovalent association of adaptor and (Figure 6D). Because the ATTO 550 crRNA duplex displays cargo allowed independent modification of adaptors and cargos to virtually no cell penetration by itself (Figure 6A), essentially all increase internalization. Further, combinatorial inclusion of the crRNA inside the cell entered as part of the CRISPR complex. Endolytic Peptides (EPs) within both adaptors and cargos should Thus, the crRNA provides a measure of CRISPR internalization. increase endosomal escape and aid internalization as the beneficial When Cas9 was not labeled similar amounts of crRNA/ properties of these peptides may be due to both factors. To explore Cas9 internalization were observed (unpublished, D. Morris). thesepossibilities,wechose aset of 4 EPs,Aurein1.2 (AUR), GALA, Although TAT-CaM failed to increase CRISPR internalization, LAH4 and HA2, previously suggested to increase endosomal escape, intrinsic internalization and frequent TAT-CaM-dependent and incorporated these sequences into a set of MBP-EP-CBS cargos effects on the details of intracellular localization (Figures and TAT-EP-CaM adaptors. Problematically, association of 6C,D) suggested some potential for genome modification. negative TAT-CaM with negative MBP produces complexes with Unfortunately, even with CRISPR complexes that both cleaved modest internalization, which probably limits endolytic outcomes, target DNA in vitro and modified genomic DNA when nevertheless, these modified adaptors and cargos produced clear and introduced using CRISPR-Max, no detectable genome differential effects on complex internalization. modification occurred when CRISPR was introduced with or Initially, intrinsic internalization of each cargo alone was without TAT-CaM in HEK-293 (Supplementary Figure S2)or compared to that observed with TAT-CaM, four TAT-EP-CaMs BHK cells (Morris and McMurry, unpublished). and CaM-Cntl to obtain a profile of adaptor effectiveness. To Frontiers in Pharmacology 08 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 5 TAT-CaM specifically induces MBP-CBS internalization but requires µm concentrations. (A) Internalization of MBP-CBS alone (left panels), with equimolar with TAT-CaM (center panels) or with “TAT-less” CaM control (right panels) following 1 h incubation with BHK cells in growth media. Cells were washed (PBS x2) and imaged immediately (top panels) or later (bottom panels) in imaging media with NucBlue. (B) Internalization of equimolar MBP-CBS and TAT-CaM at 4 concentrations following a 2 h incubation, washing (PBS x2) and imaging in media with NucBlue. limit the time differential between imaging treatments, GALA alone, TAT-CaM modestly increased experiments were performed in sets of 4, each containing the internalization (Figure 7C), while TAT-LAH4-CaM no adaptor control as a reference. Multiple rounds of images were remained a very effective adaptor (Figure 7D)and CaM- taken as soon as possible after complex removal and PBS Cntl was again inhibitory (Figure 7D). washing. As shown above, relative to MBP-CBS alone, The MBP-LAH4 cargo displayed strong internalization association with TAT-CaM specifically increased across the adaptor set (Figure 8) and reduced detection internalization (Figure 7A). Internalization of MBP-CBS by sensitivity was required to produce non-saturating signal. As TAT-LAH4-CaM was also very efficient (Figure 7B), while shown more directly below, the LAH4 sequence in MBP-LAH4 TAT-AUR-CaM was marginally effective (Figure 7A) and itself acted as an effective CPP. Against this background of higher CaM-Cntl depressed internalization (Figure 7B). internalization, TAT-CaM increased MBP-LAH4 internalization With MBP-GALA as cargo, the adaptors displayed modestly (Figure 8A) while TAT-LAH4-CaM was still very similar activity as with MBP-CBS. Compared to MBP- effective (Figure 8B). Unexpectedly, TAT-AUR-CaM, TAT- Frontiers in Pharmacology 09 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 6 Internalization of CRISPR components and CRISPR compared to the TAT-CaM/CRISPR complex. BHK cells in DMEM with 5% FBS were treated for 1 h with (A) crRNA, (B) Cys-less-Cas9-NLSx2, (C) CRISPR or (D) the TAT-CaM/CRISPR prior to washing (PBS x2) and imaging in media with NucBlue. The Cys-less Cas9-NLSx2 concentrations were 100 nm; however, the crRNA and TAT-CaM were used at 150 nm to ensure available Cas9 was fully bound. GALA-CaM and TAT-HA2-CaM were inhibitory and decreased efficacy with MBP-AUR and MBP-HA2. Further, the MBP-LAH4 internalization much as CaM-Cntl. reciprocal exchange of LAH4 (into cargo) and Aurein and Cargos MBP-AUR and MBP-HA2 displayed similar but HA2 (into adaptor) produced poor internalization in both distinctly alternative behaviors (Figure 9). Both EPs appeared experiments, suggesting this negative interaction is to have some CPP capability of their own (see below) but also independent of context. More broadly, the comparison of bound the coverslip surface complicating analysis. In addition, adaptor effectiveness confirmed the utility of the TAT and both cargos were resistant to increased internalization by other LAH4 sequences and demonstrated that inclusion of GALA, adaptors, although TAT-CaM showed some efficacy. Most AUR and HA2 sequences in the adaptor did not increase strikingly, TAT-LAH-CaM had modest impact on MBP-AUR internalization. (Figure 9B) and no impact on MBP-HA2 internalization despite While beneficial EP effects on cargo internalization are this adaptor’s effectiveness with the other cargos (Figure 9D). implied by reduced sensitivity used in detection, direct Also noteworthy was the loss of MBP-HA2 internalization when comparison provided a more quantitative measure. To complexed with TAT-HA2-CaM to levels below that of CaM compare these EP effects, all 5 cargos were simultaneously control (Figure 9D). analyzed first alone and then with selected adaptors. Although too large for presentation in the body of this report, Comparison of the intrinsic internalization tendencies of the the overall internalization characteristics of the MBP-EP-CBS EP-cargos showed LAH4-MBP internalized readily, MBP-AUR cargos and TAT-EP-CaM adaptors are most apparent in a and MBP-HA2 to some extent and MBP-CBS and MBP-GALA noncolor matrix simultaneously displaying all the very little (Figure 10A). Stated alternatively, when fused into an representative images (Supplementary Figure S3). The MBP-CBS cargo, LAH4 is a relatively good CPP, Aurein, and reproducibility of the pattern is shown in a nearly complete HA2 have CPP characteristics and GALA appears to have no second matrix performed at half concentration (0.5 µm) that is such property. missing only the CaM control (Supplementary Figure S4). As shown above, TAT-CaM has some ability to increase Significantly, both data sets demonstrate the ability of TAT- internalization of all 5 cargos. Nevertheless, internalization of LAH4-CaM to dramatically increase internalization of MBP- MBP-EP-CBSs by TAT-CaM (Figure 10B) correlated with the CBS, MBP-GALA and MBP-LAH4, while having minimal intrinsic internalization observed for the cargos themselves. Frontiers in Pharmacology 10 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 7 550- MBP-CBS and 550- MBP-GALA internalized with adaptor set including controls. Internalization of equimolar adaptor/cargo complexes (1 µm) in growth media were analyzed as sets of four conditions, each containing the no adaptor cargo as reference. All complexes were internalized for 1 h prior to washing (PBS x2) and imaging in media containing NucBlue. Internalization of CBS-MBP in the upper panel (A) used: no adaptor, TAT- CaM, TAT-AUR-CaM, or TAT-GALA-C and in the lower panel (B) used: no adaptor, “TAT-less” CaM-Ctrl, TAT-LAH4-CaM, and TAT-HA2-CaM. Internalization of MBP-GALA used the same adaptor sets in the upper (C) and lower (D) panels. For presentation, intensities within each set were adjusted identically: however, for these cargos no-adaptor reference are not presented at equal intensity in upper and lower panels to avoid oversaturating internalization with TAT-LAH4-CaM. Frontiers in Pharmacology 11 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 8 550-MBP-LAH4 internalized with adaptor set including controls as in the prior figure. Equimolar adaptor/cargo complexes (1 µm) were internalized in growth media for 1 h prior to washing (PBS x2) and imaging in media containing NucBlue. Internalization of MBP-LAH4 with adaptors as above in both the upper panel (A) and the lower panel (B). For presentation, intensities within each set adjusted identically but with the no-adaptor reference at approximately equal intensity for both panels. Indeed, internalization of the adaptor/cargo complexes internalization, it was clear that more effective adaptors maintained almost the same rank order as observed for the were needed. Several approaches were used to increase the cargos alone (MBP-LAH4>MBP-AUR > MBP-HA2>MBP- internalization potential of these next-generation adaptors CBS = MBP-GALA) except that TAT-CaM-induced and one of the most successful involved use of CaM from internalization of MBP-CBS became visible (MBP-CBS > naked mole rat to create a TAT-NMR-CaM adaptor (Gentry MBP-GALA). These results showed that substrate EPs still et al., 2021). A second approach used an engineered positive played a major role in boosting the internalization process, version of GFP (pGFP), which has been shown (Lawrence but the overall efficiency was dependent on both cargo EPs et al., 2007; Thompson et al., 2012) to be a powerful cell and TAT-CaM. penetrating protein (also ‘CPP’), to create a pGFP-CaM On the other hand, when the TAT-LAH4-CaM was used as adaptor. Although full characterization of these very an adaptor, the internalization properties of the complexes different adaptors is beyond the scope of this report, it changed qualitatively (MBP-LAH4>MBP-CBS > MBP-AUR > seemed valuable to see how these more effective adaptors MBP-HA2 = MBP-GALA) (Figure 10C). While MBP-GALA interacted with the MBP-EP-CBS cargos. internalization remained low, the relative signal became similar When TAT-NMR-CaM was used to internalize the EP-cargo to HA2-MBP, in agreement with above data showing TAT- set, cargos having the least intrinsic internalization ability (MBP- LAH4-CaM adaptor dramatically increased internalization of CBS and MBP-GALA) were induced relative to those with strong MBP-GALA (Figure 7D). These data show that internalization intrinsic internalization (Figure 11A). While all cargos stimulated by the adaptor was more significant to the internalized similarly, the cargo without an EP (MBP-CBS) internalization process than the CPP effects of the cargo-EP internalized most readily. MBP-AUR internalized the least peptides. when non-background signal was considered. Clearly with the Given the beneficial role of positive charge during the TAT-NMR-CaM adaptor, specific internalization had become internalization process and recognition that the CaM robust enough that intrinsic internalization was no longer a portion of TAT-CaM was acidic and inhibited major factor. Frontiers in Pharmacology 12 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 9 550-MBP-AUR and 550-MBP-HA2 internalized with adaptor set including controls as in prior 2 figures. Equimolar adaptor/cargo complexes (1 µm) were internalized in growth media for 1 h prior to washing (PBSx2) and imaging in media containing NucBlue. Internalization of MBP-AUR and MBP-HA2 with the adaptors as above in both the upper panels (A,C) and the lower panels (B,D). For presentation, intensities within each set adjusted identically but with the no-adaptor reference at approximately equal intensity for all panels. Frontiers in Pharmacology 13 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 10 Direct comparison of internalization efficiencies for five 550-labeled MBP-EP-CBS cargos: MBP-CBS, MBP-AUR, MBP-GALA, MBP-LAH4, and MBP-HA2. Cargo alone and adaptor/cargo complexes (1 µm) were internalized in growth media for 1 h prior to washing (PBS x2) and imaging in media containing NucBlue. Internalization of the 5 substrates was concurrently analyzed in (A) with no adaptor present, in (B) with TAT-CaM and in (C) with TAT-LAH4-CaM. The individual experiments represented by each column were independently adjusted to accent pattern differences within each column (for direct comparison of adaptors see Figures 7–9). Frontiers in Pharmacology 14 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 FIGURE 11 Direct Comparison of internalization efficiencies for the MBP-EP-CBS cargos internalized by two next generation adaptors. (A) Internalization of the five 550-labeled MBP-EP-CBS cargos by TAT-NMR-CaM (A) and pGFP-CaM (B) under the same conditions used with TAT-EP-CaM adapters. Internalization of these 5 cargos with (C) TAT-NMR-CaM and (D) pGFP-CaM analyzed in the later part of a 1 h internalization incubation when the complex is present in solution. Frontiers in Pharmacology 15 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 the only factor driving internalization, then MBP-AUR and MBP-HA2 should have increased as well. However, relative internalization was clearly lower. Additional evidence more dramatically demonstrated the unusual behavior of the MBP-AUR and MBP-HA2 cargos. When EP-cargo internalization by TAT-NMR-CaM (Figure 11C) and pGFP-CaM (Figure 11D) was analyzed with adaptor/cargo complexes still present, internalization patterns of MBP-AUR and MBP-HA2 were strikingly different from other cargos. Particularly for MBP-AUR, little intracellular cargo was apparent and there was a black space in the cell center. Indeed, the lack of intracellular MBP-AUR internalized by pGFP-CaM during treatment (Figure Fig11D) did not appear consistent with that present after washing (Figure 11B) in these concurrent experiments, suggesting a wave of internalization associated with the washing event. In contrast to the other adaptors, the fluorescence of the pGFP-CaM allowed simultaneous analysis of adaptor and cargo internalization. When imaged with adaptor/cargo complexes present at a concentration of 1 μm, colocalization of pGFP- CaM and cargos again broke into two groups with MBP-AUR and MBP-HA2 producing much lower levels of internalized cargo apparent as green cells (Figure 12A). The cell surface staining observed with MBP-CBS is a characteristic of this cargo that occurred at high complex concentration, a situation that also appeared to produce a wave of internalization associated with the washing event. However, when the concentration of pGFP-CaM/cargo complexes was reduced 5-fold (Figure 12B), the MBP-CBS pattern both increased in intensity and assumed a similar appearance to that of MBP-GALA and MBP-LAH4. In contrast, both MBP-AUR and MBP-HA2 retained their distinctive green appearance, although modest internalization of pGFP-CaM/MBP-AUR became more apparent. Nevertheless, the relative ineffectiveness of both EP and next generation adaptors at inducing MBP-AUR and MBP-HA2 internalization strongly suggests that intrinsic properties of the EP sequences are responsible for the differences. Conclusion FIGURE 12 Colocalization of MBP-EP-CBS cargos and pGFP-CaM The primary conclusion of these studies is that cargo during comparison of the internalization of the five MBP-EP-CBS characteristics often play an important role in CPP associated cargos at two concentrations. Simultaneous internalization of the five 550-labeled MBP-EP-CBS cargos with equimolar pGFP- internalization. Further, net charge may be a dominating CaM at complex concentrations of (A) 1000 nm or (B) 200 nm. As characteristic governing CPP-adaptor mediated internalization in the prior 5 figures, equimolar adaptor/cargo complexes were internalized in growth media for 1 h prior to washing (PBS x2) and efficiency. Our recent experience suggests that cargo-induced imaging in media containing NucBlue. increases in complex internalization represent the typical behavior observed with cargo proteins having a net charge ranging from positive to neutral. A dramatic example of the ability of charge to increase intrinsic cargo internalization can be When pGFP-CaM was used to internalize the EP-cargo set, seen in supercharged GFP, which was created by mutating MBP-CBS and MBP-GALA joined MBP-LAH4 as the cargos that surface residues to R or K and produced a powerful cell internalized most readily (Figure 11B). If adaptor efficacy were penetrating protein capable of delivering a covalently attached Frontiers in Pharmacology 16 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 cargo (Lawrence et al., 2007; McNaughton et al., 2009). This consistent with the nearly complete block of MBP-AUR and supercharged GFP was of course used in the design of pGFP- pGFP-CaM internalization apparent in images taken while the CaM and produced an adaptor that appears to be a powerful CPP complex was present in the medium. This suggests that a wave of with most cargos. internalization accompanies washing, a phenomenon that The experiments presented here represent a fraction of the appears fairly common. We have avoided experiments, usually conditions, cargos and methodologies used during our efforts to involving high complex concentrations, where adaptor/cargo understand why some cargos including Cas9 displayed no TAT- complexes strongly display this behavior because this issue is CaM-specific increase in internalization. While the ability of beyond the scope of the current study. Significantly, a reduction positive charge to enable CPP activity is well established, the in pGFP-CaM/MBP-AUR concentration increased dominating effect cargo charge can have on this process seems to internalization during the incubation period. However, for this be an understudied topic. Between that space where the CPP cargo, inhibition of entry remains readily apparent. dominates the internalization process and where the adaptor has The TAT-CaM adaptor was obviously not designed based on marginal impact on strong intrinsic cargo internalization may lie the fact that CaM was a very negative protein that could isolate the general case where both adaptor (CPP) and cargo are CPP sequence efficacy. As the effect the negative CaM was having positively impacting membrane association and on internalization became apparent (see data on the TAT-less internalization. This is not particularly surprising and CaM control), this characteristic was regarded as an unequivocal represents a specific case of what has been termed the A-B negative. However, recognition that CPP-EP-CaM fusions site problem (Jencks, 1981). When two binding sites can provided adaptors whose internalization ability was based independently associate with a target, two relatively weak sites almost entirely on the CPP characteristics has proved can generate binding beyond that expected from addition of their valuable. In combination with cargos based on intrinsically individual free energy because the first binding event pays much negative MBP, which also provides little propensity for of the entropic price of association, increasing the enthalpic internalization, complexes can be created in which CPP and benefit of the second site association. Clearly, cargo EP sequences can be combinatorially tested to improve characteristics including charge seriously complicate internalization and probably cargo escape. In this study, we investigation of CPP-adaptor function and behavior. This is focused entirely on effects the 4 EP sequences had on not a flaw in the adaptor concept but rather a characteristic of internalization because it was amenable to confocal analysis. CPP/cargo internalization events. Indeed, a CPP-less control The surprising benefits of the LAH4 as a CPP, apparent in both cargo is an important control even for fused CPP-cargo the TAT-LAH4-CaM adaptor and the LAH-MBP cargo, suggest proteins. Clearly, cargo-induced internalization complicates the sequence will have general utility in stimulating cell interpretation of the existing CPP literature based on covalent penetration. Clearly, adding the LAH4 sequence to TAT- association of CPPs and cargos. NMR-CaM and GFP-CaM provides one potential avenue for establishing the utility of this sequence in creating even more There are practical considerations that are unique to and essential for the interpretability of CPP-adaptor/CBS-cargo powerful adaptors. experiments. In general, the primary goal of CPP The negative consequences the AUR and HA2 sequences on internalization is to increase cargo entry. However, when most cargo internalization may not recommend their use as CPPs but of the basis of cargo entry resides in characteristics of the cargo does not preclude their utility as EPs with the right CPP-adaptor. (e.g., positive charge), unexpected outcomes can occur. These Indeed, given the striking phenotype associated with inclusion of include not only an inability of the adaptor to increase adaptor/ these relatively short sequences in the MBP cargos, it is tempting cargo internalization, but even adaptor inhibition of the to speculate that their previously identified EP properties are internalization process. When the adaptor inhibits complex responsible. Because the TAT-EP-CaM/MBP-EP model internalization, as occurred in some situations with TAT- produced relatively low internalization and because endosomal CaM, the adaptor was used in excess of cargo, so that all escape is not easily captured with confocal microscopy, internalizing cargo was bound by the inhibitory adaptor. Of investigation of combinatorial endolytic properties of EP course, the reciprocal is also the case, when the adaptor drives sequences in adaptors and cargos has not been addressed internalization of inhibitory cargo, the cargo needs to be in excess here. Future experiments designed to measure endosomal for interpretation of adaptor internalization to be interpretable. escape using EPs in the adaptor context will require These factors were a consideration during CRISPR incorporation of a reporter that directly measures escape, e.g. internalization and an excess of both the crRNA duplex and (Milech et al., 2015; Teo et al., 2021). While the ability of the TAT-CaM were used, so that essentially all Cas9 entering the cell TAT-EP-CaM/MBP-EP system to measure endosomal escape did so as part of a complete complex. The ability of the MBP- remains to be established, our adaptor/cargo strategy clearly has AUR cargo to block pGFP-CaM internalization suggests pGFP- utility in investigating the roles of CPP and EP sequences in the CaM was not in excess during that experiment. As noted, the internalization process. This system may prove quite valuable in amount of MBP-AUR apparent inside washed cells was not pursuit of even more effective CPPs that can deliver large Frontiers in Pharmacology 17 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 macromolecular cargoes with all the therapeutic potential State University College of Science and Mathematics. MW was implicit in CPP capability. supported by the PHS grant R25GM111655. Data availability statement Acknowledgments The datasets presented in this study can be found in online We are grateful to Drs. Carol Chrestensen and Scott Nowak repositories. The names of the repository/repositories and for helpful discussions and technical assistance. accession number(s) can be found in the article/ Supplementary Material. Conflict of interest Author contributions The authors declare that the research was conducted in the absence of any commercial or financial relationships that could DM designed experiments, performed most of the be construed as a potential conflict of interest. experimentation, analyzed the data and wrote large portions of the manuscript. LS and SH performed extensive experimentation and analyzed data for several of the Publisher’s note constructs. MW developed the EP work, designed and produced the EP-containing adaptors and cargos. MM All claims expressed in this article are solely those of the authors developed the tamavidin work, designed and produced the and do not necessarily represent those of their affiliated organizations, Tam constructs and performed the nucleic acid delivery or those of the publisher, the editors and the reviewers. Any product experiments. JM is principal investigator of the group, that may be evaluated in this article, or claim that may be made by its oversaw all research, helped analyze and interpret the data, manufacturer, is not guaranteed or endorsed by the publisher. and wrote much of the manuscript. Supplementary material Funding The Supplementary Material for this article can be found This work was supported by the U.S. PHS grants to JM online at: https://www.frontiersin.org/articles/10.3389/fphar. (R15EB028609) and DM (R15GM129711) and the Kennesaw 2022.1070464/full#supplementary-material References Akita, H.,Masuda,T., Nishio,T., Niikura, K.,Ijiro, K., andHarashima,H.(2011). Cargo complexes: A study on the effect of net cargo charge and length. Cells 11 (7), Improving in vivo hepatic transfection activity by controlling intracellular trafficking: the 1195. doi:10.3390/cells11071195 function of GALA and maltotriose. Mol. Pharm. 8 (4), 1436–1442. doi:10.1021/mp200189s Jencks, W. P. (1981). On the attribution and additivity of binding energies. Proc. Albrecht, T., Zhao, Y., Nguyen, T. H., Campbell, R. E., and Johnson, J. D. (2015). Natl. Acad. Sci. U. S. A. 78 (7), 4046–4050. doi:10.1073/pnas.78.7.4046 Fluorescent biosensors illuminate calcium levels within defined beta-cell endosome Johnson, R. M., Harrison, S. D., and Maclean, D. (2011). Therapeutic applications subpopulations. Cell Calcium 57 (4), 263–274. doi:10.1016/j.ceca.2015.01.008 of cell-penetrating peptides. Methods Mol. Biol. 683, 535–551. doi:10.1007/978-1- Alves, I. D., Jiao, C.-Y., Aubry, S., Aussedat, B., Burlina, F., Chassaing, G., et al. 60761-919-2_38 (2010). Cell biology meets biophysics to unveil the different mechanisms of Lawrence, M. S., Phillips, K. J., and Liu, D. R. (2007). Supercharging proteins can penetratin internalization in cells. Biochim. Biophys. Acta 1798 (12), 2231–2239. impart unusual resilience. J. Am. Chem. Soc. 129 (33), 10110–10112. doi:10.1021/ doi:10.1016/j.bbamem.2010.02.009 ja071641y Aqeel, A., and Khan, J. M. (2022). pH-sensitive endosomolytic peptides in gene Lecher, J. C., Nowak, S. J., and McMurry, J. L. (2017). Breaking in and busting out: and drug delivery: Endosomal escape and current challenges. J. Drug Deliv. Sci. Cell-penetrating peptides and the endosomal escape problem. Biomol. Concepts 8 Technol. 76, 2231–103786. doi:10.1016/j.jddst.2022.103786 (3-4), 131–141. doi:10.1515/bmc-2017-0023 Derakhshankhah, H., and Jafari, S. (2018). Cell penetrating peptides: A concise LeCher, J. C., Didier, H. L., Dickson, R. L., Slaughter, L. R., Bejerano, J. C., Ho, S., review with emphasis on biomedical applications. Biomed. Pharmacother. 108, et al. (2020). Utilization of a cell-penetrating peptide-adaptor for delivery of HPV 1090–1096. doi:10.1016/j.biopha.2018.09.097 protein E2 into cervical cancer cells to arrest cell growth and promote cell death. Fonseca, S. B., Pereira, M. P., and Kelley, S. O. (2009). Recent advances in the use biorxiv 8 (3-4), 131–141. doi:10.1101/2020.01.31.928358 of cell-penetrating peptides for medical and biological applications. Adv. Drug Lönn, P., and Dowdy, S. F. (2015). Cationic PTD/CPP-mediated macromolecular Deliv. Rev. 61 (11), 953–964. doi:10.1016/j.addr.2009.06.001 delivery: Charging into the cell. Expert Opin. Drug Deliv. 12 (10), 1627–1636. doi:10. Gentry, S. B., Nowak, S. J., Ni, X., Hill, S. A., Wade, L. R., Clark, W. R., et al. 1517/17425247.2015.1046431 (2021). A real-time assay for cell-penetrating peptide-mediated delivery of McMurry, J. L., Chrestensen, C. A., Scott, I. M., Lee, E. W., Rahn, A. M., Johansen, molecular cargos. PLOS ONE 16 (9), e0254468. doi:10.1371/journal.pone.0254468 A. M., et al. (2011). Rate, affinity and calcium dependence of CaM binding to eNOS Hymel, H. C., Rahnama, A., Sanchez, O. M., Liu, D., Gauthier, T. J., and Melvin, and nNOS: Effects of phosphorylation. FEBS J. 278 (24), 4943–4954. doi:10.1111/j. A. T. (2022). How cargo identity alters the uptake of cell-penetrating peptide (CPP)/ 1742-4658.2011.08395.x Frontiers in Pharmacology 18 frontiersin.org Morris et al. 10.3389/fphar.2022.1070464 McNaughton, B. R., Cronican, J. J., Thompson, D. B., and Liu, D. R. (2009). Takakura, Y., Tsunashima, M., Suzuki, J., Usami, S., Kakuta, Y., Okino, N., et al. Mammalian cell penetration, siRNA transfection, and DNA transfection by (2009). Tamavidins--novel avidin-like biotin-binding proteins from the supercharged proteins. Proc. Natl. Acad. Sci. U. S. A. 106 (15), 6111–6116. Tamogitake mushroom. Febs J. 276 (5), 1383–1397. doi:10.1111/j.1742-4658. doi:10.1073/pnas.0807883106 2009.06879.x Milech, N., Longville, B. A., Cunningham, P. T., Scobie, M. N., Bogdawa, H. M., Takakura, Y., Oka, N., Kajiwara, H., Tsunashima, M., Usami, S., Tsukamoto, H., Winslow, S., et al. (2015). GFP-complementation assay to detect functional CPP et al. (2010). Tamavidin, a versatile affinity tag for protein purification and and protein delivery into living cells. Sci. Rep. 5, 18329. doi:10.1038/srep18329 immobilization. J. Biotechnol. 145 (4), 317–322. doi:10.1016/j.jbiotec.2009.12.012 Moulay, G., Leborgne, C., Mason, A. J., Aisenbrey, C., Kichler, A., and Bechinger, Takakura, Y., Sofuku, K., and Tsunashima, M. (2013). Tamavidin 2-REV: an B. (2017). Histidine-rich designer peptides of the LAH4 family promote cell delivery engineered tamavidin with reversible biotin-binding capability. J. Biotechnol. 164 of a multitude of cargo. J. Pept. Sci. 23 (4), 320–328. doi:10.1002/psc.2955 (1), 19–25. doi:10.1016/j.jbiotec.2013.01.006 Ngwa, V. M., Axford, D. S., Healey, A. N., Nowak, S. J., Chrestensen, C. A., and Teo, S. L. Y., Rennick, J. J., Yuen, D., Al-Wassiti, H., Johnston, A. P. R., and McMurry, J. L. (2017). A versatile cell-penetrating peptide-adaptor system for Pouton, C. W. (2021). Unravelling cytosolic delivery of cell penetrating peptides efficient delivery of molecular cargos to subcellular destinations. PloS one 12 (5), with a quantitative endosomal escape assay. Nat. Commun. 12 (1), 3721. doi:10. e0178648. doi:10.1371/journal.pone.0178648 1038/s41467-021-23997-x Salerno, J. C., Ngwa, V. M., Nowak, S. J., Chrestensen, C. A., Healey, A. N., and Thompson, D. B., Villaseñor, R., Dorr, B. M., Zerial, M., and Liu, D. R. (2012). McMurry, J. L. (2016). Novel cell-penetrating peptide-adaptors effect intracellular Cellular uptake mechanisms and endosomal trafficking of supercharged proteins. delivery and endosomal escape of protein cargos. J. Cell Sci. 129 (5), 893–897. doi:10. Chem. Biol. 19 (7), 831–843. doi:10.1016/j.chembiol.2012.06.014 1242/jcs.182113 Wharton, S. A., Martin, S. R., Ruigrok, R. W., Skehel, J. J., and Wiley, D. C. (1988). Shahmiri, M., Enciso, M., and Mechler, A. (2015). Controls and constrains of the Membrane fusion by peptide analogues of influenza virus haemagglutinin. J. Gen. membrane disrupting action of Aurein 1.2. Sci. Rep. 5, 16378. doi:10.1038/srep16378 Virol. 69 (Pt 8), 1847–1857. doi:10.1099/0022-1317-69-8-1847 Skotland, T., Iversen, T. G., Torgersen, M. L., and Sandvig, K. (2015). Cell- Yokoo, H., Oba, M., and Uchida, S. (2021). Cell-penetrating peptides: Emerging penetrating peptides: Possibilities and challenges for drug delivery in vitro and in tools for mRNA delivery. Pharmaceutics 14 (1), 78. doi:10.3390/ vivo. Molecules 20 (7), 13313–13323. doi:10.3390/molecules200713313 pharmaceutics14010078 Frontiers in Pharmacology 19 frontiersin.org

Journal

Frontiers in PharmacologyPubmed Central

Published: Nov 21, 2022

There are no references for this article.