Peptides of pHLIP family for targeted intracellular and extracellular delivery of cargo molecules to tumors

Linden C. Wyatt; Anna Moshnikova; Troy Crawford; Donald M. Engelman; Oleg A. Andreev; Yana K. Reshetnyak

doi:10.1073/pnas.1715350115

New Research In

Contributed by Donald M. Engelman, February 4, 2018 (sent for review September 5, 2017; reviewed by Nitin Nitin and William C. Wimley)

Significance

Targeted delivery has been limited by reliance on tumor cell biomarkers. The emergence of the pH (low) insertion peptide (pHLIP) technology provides an alternative by targeting a metabolic marker, tumor cell surface acidity. We report several new pHLIPs, including a new concept, pHLIP bundles, and we evaluate these constructs alongside a new generation of pHLIPs. We also discuss challenges inherent to the design and accurate evaluation of pHLIPs. Our research elucidates the strengths and weaknesses of existing pHLIPs, proposes future peptide modifications that could further improve tumor targeting, and discusses the applicability of this new generation of pHLIPs for specific areas of drug delivery. The principles and new constructs promise to advance applications to tumor therapy.

Abstract

The pH (low) insertion peptides (pHLIPs) target acidity at the surfaces of cancer cells and show utility in a wide range of applications, including tumor imaging and intracellular delivery of therapeutic agents. Here we report pHLIP constructs that significantly improve the targeted delivery of agents into tumor cells. The investigated constructs include pHLIP bundles (conjugates consisting of two or four pHLIP peptides linked by polyethylene glycol) and Var3 pHLIPs containing either the nonstandard amino acid, γ-carboxyglutamic acid, or a glycine−leucine−leucine motif. The performance of the constructs in vitro and in vivo was compared with previous pHLIP variants. A wide range of experiments was performed on nine constructs including (i) biophysical measurements using steady-state and kinetic fluorescence, circular dichroism, and oriented circular dichroism to study the pH-dependent insertion of pHLIP variants across the membrane lipid bilayer; (ii) cell viability assays to gauge the pH-dependent potency of peptide-toxin constructs by assessing the intracellular delivery of the polar, cell-impermeable cargo molecule amanitin at physiological and low pH (pH 7.4 and 6.0, respectively); and (iii) tumor targeting and biodistribution measurements using fluorophore-peptide conjugates in a breast cancer mouse model. The main principles of the design of pHLIP variants for a range of medical applications are discussed.

The targeted delivery of drugs to cancer cells promises to maximize their therapeutic effects while reducing side effects. Although many biomarkers exist that can be exploited to improve tumor targeting and treatment outcomes, such as various receptors overexpressed at the surfaces of some cancer cells, useful markers are not present in all tumors. Further, the heterogeneity of the cancer cell population in an individual tumor and between tumors of various patients limits the effective use of biomarker targeting technologies, and rapid mutation increases the likelihood of the selection of cancer cell phenotypes that do not express high levels of the targeted biomarker. Thus, biomarker targeting can act as a selection method that may lead to the development of drug resistance and poor patient outcomes (1 ⇓–3).

It is well known that acidosis is ubiquitous in tumors, including both primary tumors and metastases, as a consequence of their rapid metabolism (4). The acidic microenvironment is generated by the increased use of glycolysis by cancer cells, and by the abundance of carbonic anhydrase proteins on the cancer cell surfaces. Tumor cells stabilize their cytoplasmic pH by exporting the acidity to the extracellular environment. As a result of the flux and the membrane potential, the extracellular pH is lowest at the surfaces of cancer cells, where it is significantly lower than normal physiological pH and the bulk extracellular pH in tumors (5 ⇓–7). The low pH region persists at the cancer cell surface even in well-perfused tumor areas. The acidity on the surfaces of cancer cells is a targetable characteristic that is not subject to clonal selection, and the level of acidity is a predictor of tumor invasion and aggression, since more rapidly growing tumor cells are more acidic.

The emerging technology based on pH (low) insertion peptides (pHLIPs) comprises a variety of acidity-targeting peptides, each possessing different tumor-targeting characteristics. The pHLIPs can be used in a wide variety of applications, so it is desirable to have a range of options for specific applications. Some examples of these applications include (i) fluorescence imaging (8 ⇓–10) and fluorescence image-guided surgery (11); (ii) nuclear imaging including PET and SPECT (single-photon emission computed tomography) (12, 13); (iii) therapeutic applications such as the targeted delivery of polar toxins that cannot cross cell membranes (14, 15), drug-like molecules that inherently diffuse across cell membranes (16, 17), and gene therapy (18); and (iv) nanotechnology for enhancing the delivery of gold nanoparticles (19, 20) or liposome-encapsulated payloads to cancer cells (21).

The pHLIPs are triggered to insert across the membranes of cancer cells by the acidity at the cancer cell surface. The behavior of peptides in the pHLIP family is typically described in terms of three states: at physiological pH (pH 7.4), peptides exist in equilibrium between a solvated state (state I) and a membrane-adsorbed state (state II); a decrease in pH shifts the equilibrium toward a membrane-inserted state (state III) (22). The mechanism of action of peptides in the pHLIP family is well understood: Protonatable residues, which are interspersed throughout the hydrophobic middle region and the C-terminal, membrane-inserting region of the peptides, are negatively charged at physiological pH but become neutral by protonation with a decrease in pH. The loss of charge and increase in overall hydrophobicity drives pHLIPs to partition across the hydrophobic core of the membrane bilayer to form transmembrane (TM) helix. This helix spans the lipid bilayer, leaving the N terminus in the extracellular space and placing the C terminus in the intracellular space where, due to the more alkaline pH in the cytosol, the C terminus can again become deprotonated and charged, stably anchoring the peptide in the cell membrane.

Following an extensive characterization of wild-type (WT) pHLIP, a first generation of pHLIP variants was created to examine the effects on targeting due to fairly straightforward changes to the WT primary structure such as sequence truncation, the addition and replacement of some protonatable residues with others, and sequence reversal (23 ⇓–25). Importantly, a number of the changes that were investigated had adverse effects on pHLIP properties, suggesting that further studies might reveal design principles and give more useful molecules for targeted therapy. Of these first-generation variants, Variant 3 (Var3) appeared to have the most desirable insertion characteristics, and much research has been focused on the use of Var3 for various applications (10, 11, 26, 27). Lately, new variants have emerged that incorporate more exotic changes to the peptide primary structure; these changes include the use of the nonstandard amino acids γ-carboxyglutamic acid (Gla), a residue with two protonatable carboxyl groups, and α-aminoadipic acid (Aad), a more hydrophobic version of the glutamic acid residue (28), as well as the creation of a pHLIP peptide de novo, ATRAM (29). Here, we examine several new members of the pHLIP family of peptides, including pHLIP bundles, compare their biophysical properties to some of the previous generation variants, and evaluate the utility of nine pHLIPs in drug delivery and tumor imaging applications. These variants significantly expand the useful range of applications in targeted cancer therapy.

Results

The pHLIP Constructs.

We investigated nine pHLIP variants; among them are Var3/Gla (with nonstandard amino acid Gla), Var3/GLL (with glycine−leucine−leucine motif), and pHLIP bundles (Table 1). The pHLIP bundles consist of two- or four-armed polyethylene glycol (PEG) 2-kDa spacers conjugated to the cysteine residue at the N terminus of WT: PEG-2WT and PEG-4WT, respectively (Fig. 1 A and B). Our motivation is to increase both the membrane affinity and the cooperativity of the transition from the membrane-surface state to the membrane-inserted state. Enhancement of affinity is expected to improve targeting, and higher cooperativity should narrow the window of pH that produces TM drug delivery. The information about all pHLIP variants used in the study with additional variations from the addition of single N- or C-terminal cysteine or lysine residues for conjugation purposes is provided in Tables S1 and S2. Nine pHLIP variants can be grouped together in various ways by shared characteristics. A WT-like group contains peptides with two protonatable residues (shown in bold in Table 1) in the putative TM region, multiple protonatable residues in the membrane-inserting C-terminal region, and two tryptophan residues (residue W) both located at the beginning of the helix-forming TM region; this group includes WT, PEG-2WT and PEG-4WT, WT/Gla, and WT/Gla/Aad. A Var3-like group is based on Var3 from the first pHLIP series (25). This group includes Var3, Var3/Gla, and Var3/GLL, each of which have three protonatable residues in the TM region and tryptophan residues located at the beginning and end of the TM region. Considering this scheme, ATRAM, with its multiple glycine and leucine residues and single tryptophan located about two-thirds to the end of its TM part, is in a group of its own. Other subgroups can be considered as well: a subgroup of peptides that incorporate the nonstandard Gla residue, shown in italics in Table 1 (i.e., WT/Gla, WT/Gla/Aad, and Var3/Gla), and another subgroup that includes peptides containing the GLL motif (Var3/GLL and ATRAM). When performing analysis of biophysical measurements, analyzing variants with respect to their group mates becomes important: The very different characteristics of peptides from various groups make it difficult to accurately compare the behavior of all peptides at the same time.

View this table:

Table 1.

List of main groups of pHLIP variants

Fig. 1.

Schematic organization of pHLIP bundles: (A) PEG-2WT with 2-kDa two-arm PEG and two WT pHLIPs, and (B) PEG-4WT with 2-kDa four-arm PEG and four WT pHLIPs. Transitions between the three states of PEG-2WT and PEG-4WT in phosphate buffer at pH 8 (state I), in the presence of POPC liposomes at pH 8 (state II) and in the presence of liposomes at pH 4 (state III), were monitored by changes of (C and D) tryptophan fluorescence, (E and F) CD, and (G and H) OCD signals. Normalized pH-dependent steady-state transitions from state II to state III were examined by analyzing the shift in position of fluorescence spectrum maximum of (I) PEG-2WT and (J) PEG-4WT in the presence of physiological concentrations of calcium and magnesium ions. The data were fitted using the Henderson−Hasselbalch equation; the fitting curves and 95% confidence interval are shown by red and blue lines, respectively.

Biophysical Steady-State and Kinetics Studies.

A variety of spectroscopic techniques were employed to probe the interaction between pHLIP variants and 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) phospholipid bilayers in liposomes; these techniques included steady-state fluorescence spectroscopy, circular dichroism (CD), oriented circular dichroism (OCD), and stopped-flow fluorescence measurements. Steady-state fluorescence and CD experiments were conducted in phosphate buffer titrated with hydrochloric acid to drop the pH from pH 8 to pH 4 to ensure consistency with previously published data (25, 29, 30). Steady-state and kinetics fluorescence experiments measuring the pH-dependent transition from state II to state III were carried out in phosphate buffer containing the physiological concentrations of free calcium (1.25 mM) and magnesium (0.65 mM) ions found in blood, since we expected that some of the constructs might bind these ions.

We established that, in solution, PEG-2WT and PEG-4WT most probably exist in compact coil conformations, where tryptophan and other aromatic residues can form stacking structures. The resulting exciton formation was seen as a minimum around 230 nm in the CD spectra of these pHLIP bundles (Fig. 1 E and F). At pH 8, changes in the tryptophan fluorescence show that both constructs interact with the lipid bilayer, and it appears that PEG-4WT exhibits stronger binding than PEG-2WT in state II. With a reduction of pH, both pHLIP bundles inserted into the bilayer to form helices, and the TM orientations of these helices were confirmed by OCD measurements (Fig. 1 G and H). It is important to note that, in state III, the membrane-inserted state, the exciton signal generated by π−π stacking is no longer present, suggesting that the insertion of each pHLIP renders it independent of the other(s). The pK of the transition from state II to state III was shifted to pH 6.6, and, as might be expected, the cooperativity of the transition was increased for PEG-4WT compared with PEG-2WT (Fig. 1 I and J).

We compared the groups consisting of WT, Var3, and ATRAM pHLIP variants to the newly designed Var3/Gla and Var3/GLL pHLIP variants. The HPLC retention times of the peptides indicate increasing hydrophobicity within the groups in the following order, from less to more hydrophobic: WT, WT/Gla, WT/Gla/Aad and Var3, Var/Gla, Var3/GLL, and ATRAM, with ATRAM being the most hydrophobic (Table S2). Both new pHLIP variants, Var3/Gla and Var3/GLL, demonstrated a pH-dependent interaction with the membrane (Fig. S1). Var3/GLL showed a higher percentage of membrane-inserted population at pH 8, which reflects a higher affinity of the peptide for the lipid bilayer both at physiological and high pH due to the increased hydrophobicity of the peptide.

As seen for previous pHLIP designs, a blue shift (or decrease in Stokes shift) resulting from the environmental changes from state I to state II and state III was observed for all peptides (Table S3), indicating partitioning of the peptides into the lipid bilayer. However, we cannot directly compare the positions of fluorescence spectra maxima for peptides belonging to the different groups, since the locations of the tryptophan residues within the peptides varies greatly. With this fact in mind, we can conclude that the peptides had very different conformations in state II at pH 8, and that the highest membrane affinity was exhibited by the PEG-pHLIPs and by the WT/Gla/Aad, Var3/GLL, and ATRAM peptides. The PEG-pHLIPs have multiple binding sites due to the linking of multiple WT peptides within a single construct, which is expected to enhance binding affinity. The WT/Gla/Aad, Var3/GLL, and ATRAM have the most hydrophobic sequences, and thus exhibit strong binding/insertion. We also found that some peptides were especially sensitive to the presence of calcium and magnesium ions, namely WT, variants containing the Gla residue (WT/Gla, WT/Gla/Aad, and Var3/Gla), and ATRAM. This sensitivity was most obviously seen as a decreased Stokes shift (usually 2 nm to 3 nm) in state I and/or state II, and might reflect slight increases in the hydrophobicity of the peptides caused by the coordination of divalent cations resulting from the presence of closely spaced protonatable residues, such as those found in the C-terminal region of WT and, to some degree, in ATRAM, or to the presence of the Gla residue, with its two protonatable carboxyl groups, in the WT/Gla, WT/Gla/Aad, and Var3/Gla peptides. It is known that a Gla residue can form a complex with a calcium ion (31 ⇓–33). The decrease in Stokes shift in state II is likely due to the location of membrane-adsorbed peptides deeper in the lipid membrane (especially for the more hydrophobic pHLIPs: WT/Gla/Aad, Var3/GLL, and ATRAM) and/or a shift in peptide population from the solvated to the membrane-adsorbed state.

In contrast to tryptophan fluorescence changes, which are dependent on the location of tryptophan residues within the peptide sequence, the appearance of helicity is a more general parameter which can be compared among all peptides. In Fig. 2A (and Table S3), we give the ratio of ellipticity at 205 nm to 222 nm, an indicator of the degree of helicity (lower ratios indicate higher helicity), obtained for different peptides in different states. In state I, the lowest ratios were observed for pHLIP bundles, which correlate with the appearance of the exciton signal at 230 nm. In state II, the most structured peptides (ratios < 1.5) were PEG-4WT, WT/Gla/Aad, Var3/GLL, ATRAM, and PEG-2WT, which exhibited a higher affinity to the membrane and an increase in the peptide-inserted population at pH 8. At low pH, all peptides exhibited similar helical content, as expected from the formation of TM helices.

Fig. 2.

(A) Ellipticity ratios of CD signals at 205 nm to 222 nm are shown for pHLIP variants in states I, II, and III. The values of ellipticity ratios are given in Table S3. (B) The therapeutic index (TI) was calculated for different pHLIP-amanitin constructs as the ratio of EC₅₀ at pH 7.4 to EC₅₀ at pH 6.0.

The transitions from state II to state III seen in steady-state and kinetics modes exhibited pK values in the range of pH 5.7 to 6.6 in the presence of physiological concentrations of calcium and magnesium ions, with the highest cooperativity observed for PEG-4WT, and transition times varying from 0.1 s to 37.5 s (Table 2). There are subtleties that affect the comparison and interpretation of the data such as: (i) The peptides are in different starting conditions in state II at pH 8 due to greatly differing overall peptide hydrophobicity; (ii) difference in peptides pK values, which reflect equilibrium between peptides’ membrane-adsorbed and membrane-inserted populations; (iii) characteristic times, which report the movement of tryptophan residues into environments inside the membrane; however, since the tryptophan residues are located in different regions of each pHLIP, their movement into the membrane, as measured via changes in fluorescence parameters, should be expected to be different; and (iv) the cooperativity of the transition is a somewhat unstable parameter in the fitting of experimental pH-dependence data using the Henderson−Hasselbalch equation, especially if slopes are introduced at the initiation and completion of the transition (34). Lower values of cooperativity (n < 1) were observed for the peptides with tryptophan residues located at (Var3 group) or close (ATRAM) to the C terminus, which must be translocated across the cell membrane. ATRAM and Var3/GLL, which are the most hydrophobic pHLIPs and are therefore likely to be located more deeply than others in the membrane at pH 8, demonstrated the fastest times of insertion. As we showed previously, the removal of protonatable residues from the inserting C terminus increases the rate of the transition from state II to state III (24, 25). Thus, the group of Var3-like peptides exhibited fast insertion times (t < 1 s), a characteristic most attributable to the presence of less protonatable residues in the Var3-like peptides as well as a decrease in the number of these residues located at the C-terminal ends of the peptides. In the group of WT peptides, the time of insertion decreased as the hydrophobicity of the peptide increased, with insertion times listed in the following order (from longest to shortest time of insertion): WT, WT/Gla, WT/Gla/Aad, PEG-2WT, and PEG-4WT.

View this table:

Table 2.

The midpoint (pK), cooperativity (n) and SE, and time (s) parameters characterizing the pH-dependent transition of pHLIP variants in the presence of POPC liposomes

Intracellular Delivery of Polar Cargo.

We asked whether the pHLIP bundles could cause any acute cytotoxicity by themselves. HeLa cells were treated with either PEG-2WT or PEG-4WT at physiological pH (pH 7.4) and low pH (pH 6.0) for 2 h. We did not observe any cytotoxic effect at either pH, even when treating with concentrations up to 10 µM (construct concentration is presented as concentration of WT pHLIP).

A proliferation assay was employed to evaluate the ability of pHLIPs to deliver the amanitin toxin, a relatively cell-impermeable, polar cargo molecule (35, 36). For amanitin to induce cytotoxicity, it must be translocated across the cell membrane, be released from the peptide carrier, and reach its target in the nucleus (RNA polymerase II). Amanitin was conjugated via a cleavable disulfide link to the inserting, C termini of the peptides. The translocation capabilities of the pHLIP-amanitin conjugates were measured as the inhibition of proliferation of HeLa cells treated with increasing concentrations (up to 2 µM) of pHLIP-amanitin at either physiological pH (pH 7.4) or low pH (pH 6.0) for 2 h, followed by removal of the constructs, transfer of cells to normal cell culture media, and assessment of cell death at 48 h.

Each of the conjugates demonstrated pH-dependent cytotoxicity (Fig. S2). The calculated EC₂₀, EC₅₀, and EC₈₀ at physiological and low pH are shown in Table 2. At low pH, the most potent constructs were the pHLIP bundles, which exhibited the highest cooperativity of transition from membrane-adsorbed to membrane-inserted states. The least toxic at normal pH among all constructs was Var3. Fig. 2B lists the therapeutic indexes (TIs), defined as the ratio of EC₅₀ at pH 7.4 to EC₅₀ at pH 6.0 for each case. A TI of about 9 was obtained for WT/Gla and Var3, and the TI was around 5.5 for PEG-2WT, Var3/Gla, and ATRAM. It is desirable to have high potency, which is defined as a difference between cell viability at low and physiological pHs at different concentrations of the construct (Fig. 3). All constructs had high potency (60 to 70%) at particular concentrations; however, just a few constructs, namely Var3, Var3/Gla, and WT/Gla, had a high, stable potency over a wide range of concentrations. The pHLIP bundles displayed the highest potency at the lowest concentrations (0.1 μM to 0.2 µM). The potency of ATRAM peaked at concentrations around 0.5 µM and declined sharply at higher concentrations; this decline is most likely associated with the increased hydrophobicity of ATRAM, which results in a high affinity for the cell membrane at normal and high pH and promotes the shift in equilibrium toward the membrane-inserted form that is associated with the translocation of cargo across the cell membrane.

Fig. 3.

The pH-dependent potency was defined as the difference between cancer cell viability when cells were incubated at pH 7.4 and pH 6.0 at varying concentrations of different pHLIP-amanitin constructs. (A) The WT-like group. (B) Var3-like group and ATRAM.

Tumor Targeting.

To evaluate the tumor targeting and biodistribution characteristics of the pHLIP variants, we conjugated the fluorescent dye Alexa Fluor 546 (AF546) to the noninserting, N termini of seven of the peptides. Our previous data indicate excellent tumor targeting by AF546-pHLIPs (9, 27). In the case of pHLIP bundles, AF546 was conjugated to the inserting, C termini of the PEG-2WT and PEG-4WT pHLIPs, as the N termini were occupied by PEG polymers. A well-established mouse model, using implanted cells of acidic 4T1 murine breast tumors, was used in the study; this model is targeted well by pHLIPs (9, 27). Following the development of breast tumors in the mouse flank, each fluorescent construct was introduced by a single tail vein injection. Animals were killed 4 h after the injection of the fluorescent conjugates, and the tumor and major organs (kidney, liver, lungs, spleen, and muscle) were collected and imaged. We selected the 4-h postinjection time point based on previous pharmacokinetics data which show that the highest tumor targeting with pHLIPs is observed 4 h after the injection of construct (9, 27). The mean values of the surface fluorescence intensity of tumors, muscle, and organs are given in Table S4. The normalized tumor fluorescence intensity (normalized by tumor uptake of AF546-WT) for all constructs is shown in Fig. 4A. The highest tumor targeting was observed for the Var3 construct, as well as for Var3/Gla and ATRAM. The tumor uptakes of the WT and Var3/GLL constructs were significantly reduced, by 1.6- and 2.6-fold, respectively, compared with the uptake of Var3. The uptakes of WT/Gla, WT/Gla/Aad, and the pHLIP bundles were reduced even further compared with the uptake of the WT construct. It is possible that the decreased tumor targeting observed in the PEG-pHLIP bundles might be attributed to the fact that the AF546 dye was conjugated to the C terminus, which is translocated into the cytosol. At the same time, the tumor-to-muscle ratio of the WT-like group was in the range of 5.4 to 7.5. The highest tumor-to-muscle ratios were observed for Var3 (T/M = 8.9) and PEG-2WT (T/M = 7.5), and the lowest ratio was observed in Var3/GLL (T/M = 4.0) (Fig. 4B and Table S5). Among all constructs, only Var3/GLL demonstrated a tumor-to-kidney ratio less than 1 (Fig. 4C and Table S5). The highest tumor-to-liver ratio was found in Var3 and Var3/Gla (Fig. 4D and Table S5).

Fig. 4.

Normalized tumor fluorescence intensities of the AF546-pHLIP constructs are shown; (A) the signals were normalized by the tumor intensity of AF546-WT. (B) Tumor-to-muscle (T/M), (C) tumor-to-kidney (T/K), and (D) tumor-to-liver (T/L) fluorescence intensity ratios are provided. Statistically significant differences were determined by two-tailed unpaired Student’s t test; *P ≤ 0.05; and **P ≤ 0.005.

Because PEG-2WT-AF546 and PEG-4WT-AF546 are several times larger than the other pHLIP variants, we expected that they might have slower pharmacokinetics. Therefore, we also performed imaging at the 24-h postinjection time point for the two PEG-pHLIP conjugates; however, we did not observe any significant signal increase in tumors at 24 h postinjection compared with 4 h postinjection (Table S4).

Discussion

To advance cancer therapy using a range of agents with different properties, we have developed new versions of pHLIP variants and pHLIP bundles, and compared their performance to the performance of recently introduced variants with nonstandard amino acids (Gla and Aad) and the hydrophobic GLL motif. Our goal was to correlate the biophysical properties of the membrane interactions of different pHLIPs, including physiological concentrations of free calcium and magnesium ions, to the ability of these pHLIPs to move polar cargo across the cell membrane and to target acidic tumors.

The thermodynamic parameters of pK and cooperativity of pH-dependent transition from State II at pH 8 to State III at pH < 5 can be taken as predictors of the performance of a pHLIP for drug delivery and tumor targeting (17, 28, 29). While pK is a rather stable fitting parameter, the cooperativity parameter (Hill coefficient) might vary over a wide range resulting from different fittings which are within the level of accuracy of the experimental measurements. Moreover, if different binding affinities are assumed, the Hill formulation loses validity. In general, highly cooperative transitions are hard to measure in biological systems with noise, especially when examining relatively short peptides like the class of pHLIP peptides (28). Only if the biological system is approximated to be infinite can a phase transition occur (37). Moreover, transition parameters for different peptides can only be truly compared when both peptides have precisely the same starting and ending states; although this condition is met for the membrane-inserted state (state III) of the peptides, which appears very similar among pHLIP variants, the condition that the initial state (state II) of the peptides be identical is not met. As hydrophobicity varies widely among peptides of the pHLIP family due to the difference in numbers of protonatable, polar, and hydrophobic residues and their location within the peptide sequences, the characteristics of the peptide population in the initial state of the transition also varies as these peptides position themselves at different interaction levels with the hydrophobic/hydrophilic boundary region of a bilayer.

The population percentages of inserted peptide presented in Table S6 were calculated from the pH-dependent transitions of pHLIP variants. The numbers represent the percentage of membrane-inserted peptides at varying pH assuming that, at the beginning of the transition (state II) (i.e., at physiological pH and higher), the population of inserted peptides is close to zero. In reality, close consideration of the interaction between a pHLIP variant and the membrane at pH 8, in conditions more alkaline than physiological conditions where the inserted peptide population should be even less than at physiological conditions, indicates that the most hydrophobic sequences, such as ATRAM and Var3/GLL, and bundled pHLIPs with multiple binding sites within a single construct, exhibit a significant inserted peptide population, as shown by the loss of pH-dependent differences in the translocation of the polar, cell-impermeable cargo amanitin with an increase in construct concentration (i.e., a decrease in potency at higher concentrations). Additionally, as previously shown using the pore-forming peptide melittin, helix formation, membrane binding, and insertion properties are very sensitive to primary structure changes involving glycine and leucine residues (38). Ultimately, due to patient variability, it is highly desirable that potential therapeutic pHLIP constructs are able to discriminate between healthy and tumor tissue over a wide concentration range, meaning that a constant potency is necessary to avoid targeting normal tissue and the resulting significant side effects, suggesting that the properties of these variants may not be well suited for clinical development using agents that require tight targeting.

In addition to the steady-state experiments, it is important to probe tumor targeting and to examine the biodistribution of the constructs when injected into the high-flow-rate blood stream, since targeted delivery is always opposed by clearance from the blood. The best tumor targeting was shown by faster-inserting pHLIP constructs. Thus, in the design of new pHLIP variants, the biophysical kinetics parameters need to be considered in addition to the more traditionally prioritized steady-state properties. These kinetics parameters might be especially critical for the delivery and translocation of a cargo across a membrane, since we have shown that charges and the presence of cargo at the inserting end of a pHLIP can slow the process of insertion (24). Different cargoes linked to a pHLIP alter biodistribution and tumor targeting (27). Less polar pHLIP variants conjugated with hydrophobic cargoes might have a higher tendency toward targeting normal tissue and hepatic clearance. On the other hand, the size of links in pHLIP bundles could be used to tune biodistribution and redirect clearance from renal to hepatic.

Among the pHLIP variants we investigated, Var3 demonstrated excellent performance in vitro (the most stable potency over a wide range of concentrations) and high tumor targeting. Variants containing the Gla residue, especially the WT/Gla construct, showed an increase in the cooperativity of the membrane insertion transition as previously reported (28), and showed an improved TI. However, the tumor targeting of WT/Gla was lower compared with the tumor targeting of WT.

The γ-carboxyglutamic acid is not naturally encoded in the human genome, but is introduced into proteins through the posttranslational carboxylation modification of glutamic acid, resulting in an amino acid with two carboxyl groups. Several proteins are known to have Gla-rich domains, including many coagulation factors, which coordinate calcium ions, inducing conformational changes in the proteins that enhance the hydrophobicity and affinity of the proteins to the cell membrane bilayer (39). Calcium complex formation by a pHLIP increases the hydrophobicity of the peptide and alters the interaction between peptide and membrane; this fact, along with the fact that the cost of synthesizing a Gla-containing peptide is very high (Gla is one of the most expensive amino acids) might somewhat reduce enthusiasm in using the Gla residue, but, if there were sufficient advantages in a specific case, the cost might be justified. While considering peptide synthesis, it is worthwhile to note that very hydrophobic pHLIP sequences (like ATRAM), especially when coupled with even moderately hydrophobic cargoes, might be challenging to produce in the large quantities needed for clinical translation.

There is no single recipe for the best pHLIP: The peptide will need to be tailored to each specific medical application. For example, kidney clearance might be preferred to liver clearance for PET-pHLIP imaging constructs (13). High tumor-to-normal tissue fluorescence intensity ratios will be the key in fluorescence-guided surgical applications (11). Delivery of highly toxic molecules, such as amanitin, would require minimal off-targeting; thus high potency and TI will be critical. However, for the delivery of polar peptide nucleic acids or other highly specific inhibitors of particular pathways in cancer cells, neither of which are associated with toxicity in normal cells, the requirement to reduce off-targeting might be much lower, and the emphasis would be shifted toward the efficiency of delivery, the goal being to translocate as much cargo as possible (18, 40). The pHLIP bundles might yield excellent results in these types of applications, supported by the observation that PEG-4WT is the most efficient at delivering the polar molecule amanitin to the intracellular space. We believe that bundling multiple Var3 pHLIPs, in the same fashion that we linked two or four WT pHLIPs, might be more advantageous. Var3 demonstrates membrane insertion rates orders of magnitude faster than the insertion rates of WT; with the knowledge that faster insertion rates observed in biophysical experiments correlate to better tumor targeting in vivo, it stands to reason that potential PEG-Var3 constructs might demonstrate better tumor targeting still.

In drug delivery applications, pHLIP peptides are best designed for the delivery of polar, cell-impermeable molecules (14, 35, 41, 42). The intracellular delivery of a polar cargo could be further tuned by altering the link connecting the cargo to pHLIP and/or by attaching modulator molecules to the inserting end of the peptide (14, 15, 18, 35). Additionally, pHLIP could be used for the targeted delivery of cell-permeable, drug-like molecules since it can significantly increase the time of retention in blood, positively alter the biodistribution of drugs that typically rely on passive diffusion, and enhance tumor targeting, all of which would lead to an increase in TI (16). More-polar pHLIP variants are expected to be better suited to applications involving the intracellular delivery of cell-permeable cargoes.

We have now established a set of properties for a number of pHLIPs, which can be selected as starting points for clinical development in different circumstances. This body of work, with the prior studies, opens pathways for targeted delivery using a range of imaging and therapeutic agents in the fight against cancer.

Materials and Methods

The pHLIP Characterization and pHLIP Bundle Synthesis.

All peptides were purchased from CS Bio Co. Peptides were characterized by reversed phase HPLC (RP-HPLC) using Zorbax SB-C18 and Zorbax SB-C8, 4.6 × 250 mm 5-μm columns (Agilent Technology). For biophysical measurements, PEG-2WT and PEG-4WT were made by conjugating either 2-kDa bifunctional maleimide-PEG-maleimide or 2-kDa four-arm PEG-maleimide (Creative PEGWorks) to Cys-WT via an N-terminal cysteine residue. Purification of the PEG-pHLIP constructs was conducted using RP-HPLC. Peptide concentration was calculated by absorbance at 280 nm, where, for WT, WT/Gla, and WT/Gla/Aad, ε₂₈₀ = 13,940 M⁻¹⋅cm⁻¹; for Var3, Var3/Gla, and Var3/GLL, ε₂₈₀ = 12,660 M⁻¹⋅cm⁻¹; and, for ATRAM, ε₂₈₀ = 5,690 M⁻¹⋅cm⁻¹. PEG construct concentration was presented in terms of peptide concentration, not molecular concentration.

Liposome Preparation.

Small unilamellar vesicles were used as model membranes and were prepared by extrusion. POPC (Avanti Polar Lipids) was dissolved in chloroform at a concentration of 12.5 mg/mL, then desolvated by rotary evaporation for 2 h under vacuum. The resulting POPC film was rehydrated in 10 mM phosphate buffer at pH 8, either with ions (1.25 mM calcium and 0.65 mM magnesium) or without ions, vortexed, and extruded 15 times through a membrane with a pore size of 50 nm.

Steady-State Fluorescence Measurements.

Steady-state fluorescence spectra were measured using a PC1 spectrofluorometer (ISS) with temperature control set to 25.0 °C. The tryptophan fluorescence was excited using an excitation wavelength of 295 nm. Excitation and emission slits were set to 8 nm, and excitation and emission polarizers were set to 54.7° and 0.0°, respectively. Sample preparation was conducted 24 h before experiments to allow for state II equilibration. A buffer-only sample was used as a baseline for state I, and a buffer-with-POPC-only sample was used as a baseline for states II and III.

The pH Dependence Measurements.

Measurements of pH dependence were taken with the PC1 spectrofluorometer by using the shift in the position of maximum of peptide fluorescence as an indication of changes of the peptide environment at varying pH. All pH dependence measurements were conducted using blood physiological concentrations of free calcium and magnesium ions (1.25 and 0.65 mM, respectively). After the addition of hydrochloric acid, the pH of solutions containing 5 µM peptide and 1 mM POPC were measured using an Orion PerHecT ROSS Combination pH Micro Electrode and an Orion Dual Star pH and ISE Benchtop Meter (Thermo Fisher Scientific) before and after spectrum measurement to ensure equilibration. The tryptophan fluorescence spectrum at each pH was recorded, and the spectra were analyzed using the Protein Fluorescence and Structural Toolkit to determine the positions of spectral maxima (λmax). The position of λmax was plotted as a function of pH and normalized, such that position of spectral maximum in state II was set to 1 and λmaxfinal− position of spectral maximum in state III was set to 0. The normalized pH dependence was fit with the Henderson−Hasselbalch equation (using OriginLab software) to determine the cooperativity (n) and transition midpoint (pK) of transition of the peptide population from state II to state III,normalized pH dependence=11+10n(pH−pK).[1]

Steady-State CD and OCD Measurements.

Steady-state CD was measured using an MOS-450 spectrometer (Bio-Logic Science Instruments) in the range of 190 nm to 260 nm with a step size of 1 nm, and with temperature control set to 25.0 °C. Samples were prepared 24 h before experiments to allow for state II equilibration. A buffer-only sample was used as baseline for state I, and a buffer-with-POPC-only sample was used as baseline for states II and III.

OCD was measured using supported planar POPC bilayers prepared using a Langmuir−Blodgett system (KSV Nima). Fourteen quartz slides with 0.2-mm spacers were used; after sonicating the slides in 5% cuvette cleaner (Contrad 70; Decon Labs) in deionized water (≥18.2 MΩ cm at 25 °C; Milli-Q Type 1 Ultrapure Water System, EMD Millipore) for 15 min and rinsing with deionized water, the slides were immersed and sonicated for 10 min in 2-propanol, sonicated again for 10 min in acetone, sonicated a final time in 2-propanol for 10 min, and rinsed thoroughly with deionized water. Lastly, the slides were immersed in a 3:1 solution of sulfuric acid to hydrogen peroxide for 5 min and rinsed three times in deionized water. The slides were stored in deionized water until they were used. POPC bilayers were deposited on the 14 slides using a Langmuir−Blodgett minitrough: a 2.5 mg/mL solution of POPC in chloroform was spread on the subphase (deionized water), and the chloroform was allowed to evaporate for 15 min, after which the POPC monolayer was compressed to 32 mN/m. A lipid monolayer was deposited on the slides by retrieving them from the subphase, after which a solution of 10 µM peptide and 500 µM of 50-nm POPC liposomes at pH 4 was added to the slides, resulting in the creation of the supported bilayer by fusion between the monolayer on the slides and the peptide-laden lipid vesicles. After incubation for 6 h at 100% humidity, the slides were rinsed with buffer solution to remove excess liposomes, and the spaces between the cuvettes were filled with buffer at pH 4. Measurements were taken at three points during the experiment: immediately after the addition of the peptide/lipid solution (0 h), after the slides were rinsed to remove excess liposomes following the 6-h incubation time (6 h), and after an additional 12-h incubation time and rinse with buffer (18 h); these measurements were recorded on the MOS-450 spectrometer with sampling times of 2 s at each wavelength. Control measurements were conducted using a peptide solution between slides without supported bilayers and in the presence of POPC liposomes.

Kinetics Measurements.

Stopped-flow fluorescence measurements were made using an SFM-300 mixing system (Bio-Logic Science Instruments) in conjunction with the MOS-450 spectrometer. All solutions were degassed for 15 min before loading into the stopped-flow system. The pHLIP variants were incubated with POPC for 24 h before the experiment to reach state II equilibrium, and insertion was induced by mixing equal volumes of pHLIP/POPC solutions with hydrochloric acid diluted to ensure a pH drop from pH 8 to pH 4. Kinetics data were fit by one-, two-, three-, or four-state exponential models in OriginLab.

Amanitin pHLIP Conjugates.

Τηε α-amanitin (Sigma-Aldrich) was conjugated to succinimidyl 3-(2-piridyldithio)propionate) (SPDP; Thermo Fisher Scientific), followed by purification and conjugation of the SPDP-amanitin to the C-terminal cysteine residues of pHLIP peptides. For synthesis of PEG-2WT-amanitin and PEG-4WT-amanitin, Lys-WT-Cys with N-terminal lysine and C-terminal cysteine residues was used, and the Lys-WT-SPDP-amanitin was conjugated to dibenzocyclooctyne-sulfo-N-hydroxysuccinimidyl ester (DBCO-NHS ester; Sigma-Aldrich), resulting in DBCO-WT-SPDP-amanitin. Finally, two-arm or four-arm PEG-azide (Creative PEGWorks) was conjugated to DBCO-WT-SPDP-amanitin, resulting in PEG-DBCO-WT-SPDP-amanitin, with a cleavable disulfide bond present in SPDP, between the peptide and amanitin cargo. Construct concentration was calculated by absorbance at 310 nm, where, for α-amanitin, ε₃₁₀ = 13,000 M⁻¹⋅cm⁻¹. Construct concentration was presented in terms of peptide/amanitin concentration. Purification was conducted using RP-HPLC. Zorbax SB-C18 columns (9.4 × 250 mm, 5 µm; Agilent Technologies) were used for all peptide-amanitin conjugates other than ATRAM-amanitin, PEG-2WT-amanitin, and PEG-4WT-amanitin, for which Zorbax SB-C8 columns (9.4 × 250 mm, 5 µm; Agilent Technologies) were used.

Cell Proliferation Assay.

Human cervix adenocarcinoma cells (HeLa; American Type Culture Collection) were authenticated, stored according to the instructions of the supplier, and used within 3 mo of frozen aliquot resuscitation. Cells were cultured in DMEM (Sigma-Aldrich) at pH 7.4 with 4.5 g/L d-glucose, supplemented with 10% heat-inactivated FBS (Sigma-Aldrich) and 10 µg/mL ciprofloxacin (Sigma-Aldrich), in a humidified atmosphere of 5% CO₂ and 95% air at 37 °C. The pH 6.0 medium was prepared by mixing 13.3 g of dry DMEM in 1 L of deionized water. HeLa cells were loaded in the wells of 96-well plates (5,000 cells/well) and incubated overnight. The standard growth medium was replaced with medium without FBS, at pH 6.0 or 7.4, containing increasing amounts of pHLIP-amanitin conjugates (from 0 to 2.0 µM). Treatment with amanitin alone for 2 h and at concentrations up to 2 µM does not induce cell death. After 2-h incubation with the pHLIP-amanitin conjugates, the constructs were removed and replaced with standard growth medium. Cell viability was assessed after 48 h using colorimetric assay: the CellTiter 96 AQ_ueous One Solution Cell Proliferation Assay (Promega); the colorimetric reagent was added to cells for 1 h, followed by absorption measurement at 490 nm. All samples were prepared in triplicate, and each experiment was repeated between three and six times. All obtained cell proliferation data were normalized by corresponding controls (nontreated cells). There was no difference in the viability of cells incubated with media, without construct, at pH 7.4 and pH 6.0; therefore, the role of pH was excluded from consideration. Normalized cell viability data obtained in different experiments were averaged and presented in terms of the logarithm of dose of pHLIP-amanitin constructs. The dose–response function was used for fitting the obtained data (Fig. S2) (OriginLab),cell viability=Ab+At−Ab1+10p(log⁡x0−x),[2]

where Ab and At are the bottom and the top asymptotes, respectively. The top asymptote was set as constant, 100%, while for bottom asymptote we allowed small variations in the range of 0 to 10%; p is the slope (cooperativity parameter), and log⁡x0 is the center of the transition, the concentration for half response, which is used to calculate the EC₂₀, EC₅₀, EC₈₀ values,EC20=10(log⁡x0+log⁡0.25/p)[3]EC50=10log⁡x0[4]EC80=10(log⁡x0+log⁡4/p).[5]

Therapeutic index (TI) was calculated according to the equationTI=EC50pH 7.4EC50pH 6.0.[6]

Additionally, the cytotoxicity of the PEG-2WT and PEG-4WT constructs without amanitin was tested: These experiments demonstrated no cytotoxicity at physiological or low pH at treatment concentrations up to 10 µM.

Fluorescent pHLIP Conjugates.

Alexa Fluor 546 (AF546) C₅ maleimide (Thermo Fisher Scientific) was conjugated to the N-terminal cysteine residues of WT, Var3, Var3/Gla, and ATRAM. AF546 NHS Ester (Thermo Fisher Scientific) was conjugated to the N-terminal lysine residues of WT/Gla, WT/Gla/Aad, and Var3/GLL. For PEG-2WT and PEG-4WT, Cys-WT-Lys, with N-terminal cysteine and C-terminal lysine residues, was used, and was first conjugated to two-arm maleimide-PEG-maleimide or four-arm PEG-maleimide resulting in PEG-WT-Lys. Then, AF546 NHS Ester was conjugated to the C-terminal lysine residue, resulting in two-arm and four-arm PEG-pHLIP constructs with C-terminal AF546 fluorophores. Construct concentration was calculated by absorbance at 554 nm, where, for AF546, ε₅₅₄ = 93,000 M⁻¹⋅cm⁻¹. Construct concentration was presented in terms of AF546/peptide concentration, not molecular concentration. Purification was conducted using RP-HPLC for all peptides other than PEG-4WT-AF546, which was purified via Amicon Ultra MWCO 10-kDa centrifugal filter (Sigma-Aldrich). Zorbax SB-C18 columns (9.4 × 250 mm, 5 µm; Agilent Technologies) were used for all AF546-peptide conjugates except AF546-ATRAM and PEG-2WT-AF546, for which Zorbax SB-C8 columns (9.4 × 250 mm, 5 µm; Agilent Technologies) were used.

Ex Vivo Imaging.

All animal studies were conducted according to the animal protocol AN04-12-011 approved by the Institutional Animal Care and Use Committee at the University of Rhode Island, in compliance with the principles and procedures outlined by the National Institutes of Health for the care and use of animals. Mouse mammary cells (4T1; American Type Culture Collection) were s.c. implanted in the right flank (8 × 10⁵ cells/0.1 mL/flank) of adult female BALB/cAnNHsd mice (Envigo). When tumors reached ∼5 cm to 6 mm in diameter, single tail vein injections of 100 µL of 40 µM fluorophore-pHLIP solutions in PBS were performed. Mice were killed 4 h or 24 h after injection, and necropsy was immediately performed. Tumors and major organs were cut in half and imaged using an FX Kodak in-vivo image station connected to an Andor CCD camera. Mean surface fluorescence intensity of tumor, tissue, and organs was obtained via analysis of fluorescence images in ImageJ (National Institutes of Health). The corresponding autofluorescence signal was subtracted to obtain the net fluorescence intensities used in the study. Autofluorescence was calculated after imaging tumors, tissue, and organs collected from mice with no injection of fluorescent pHLIP constructs.

Acknowledgments

We are grateful to Dr. Dhammika Weerakkody for his assistance and helpful discussions. The research reported in this publication was supported, in part, by the National Institute of General Medical Sciences of the National Institutes of Health under Award R01GM073857 (to O.A.A., Y.K.R., and D.M.E.), and, in part, by the Institutional Development Award Network for Biomedical Research Excellence from the National Institute of General Medical Sciences of the National Institutes of Health under Grant P20GM103430.

Footnotes

↵¹To whom correspondence may be addressed. Email: donald.engelman{at}yale.edu, andreev{at}uri.edu, or reshetnyak{at}uri.edu.

Author contributions: D.M.E., O.A.A., and Y.K.R. designed research; L.C.W., A.M., and T.C. performed research; L.C.W., A.M., and Y.K.R. analyzed data; and L.C.W., D.M.E., O.A.A., and Y.K.R. wrote the paper.
Reviewers: N.N., University of California, Davis; and W.C.W., Tulane University School of Medicine.
Conflict of interest statement: D.M.E., O.A.A., and Y.K.R. are founders of pHLIP, Inc. They have shares in the company, but the company did not fund any part of the work reported in the paper, which was done in their academic laboratories.
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1715350115/-/DCSupplemental.

Published under the PNAS license.

View Abstract

References

↵
1. Marusyk A,
2. Polyak K
(2010) Tumor heterogeneity: Causes and consequences. Biochim Biophys Acta 1805:105–117.
OpenUrl CrossRef PubMed
↵
1. Gillies RJ,
2. Verduzco D,
3. Gatenby RA
(2012) Evolutionary dynamics of carcinogenesis and why targeted therapy does not work. Nat Rev Cancer 12:487–493.
OpenUrl CrossRef PubMed
↵
1. Lloyd MC, et al.
(2016) Darwinian dynamics of intratumoral heterogeneity: Not solely random mutations but also variable environmental selection forces. Cancer Res 76:3136–3144.
OpenUrl Abstract/FREE Full Text
↵
1. Estrella V, et al.
(2013) Acidity generated by the tumor microenvironment drives local invasion. Cancer Res 73:1524–1535.
OpenUrl Abstract/FREE Full Text
↵
1. Zhang X,
2. Lin Y,
3. Gillies RJ
(2010) Tumor pH and its measurement. J Nucl Med 51:1167–1170.
OpenUrl Abstract/FREE Full Text
↵
1. Hashim AI,
2. Zhang X,
3. Wojtkowiak JW,
4. Martinez GV,
5. Gillies RJ
(2011) Imaging pH and metastasis. NMR Biomed 24:582–591.
OpenUrl CrossRef PubMed
↵
1. Anderson M,
2. Moshnikova A,
3. Engelman DM,
4. Reshetnyak YK,
5. Andreev OA
(2016) Probe for the measurement of cell surface pH in vivo and ex vivo. Proc Natl Acad Sci USA 113:8177–8181.
OpenUrl Abstract/FREE Full Text
↵
1. Reshetnyak YK, et al.
(2011) Measuring tumor aggressiveness and targeting metastatic lesions with fluorescent pHLIP. Mol Imaging Biol 13:1146–1156.
OpenUrl CrossRef PubMed
↵
1. Adochite RC, et al.
(2014) Targeting breast tumors with pH (low) insertion peptides. Mol Pharm 11:2896–2905.
OpenUrl CrossRef
↵
1. Tapmeier TT, et al.
(2015) The pH low insertion peptide pHLIP variant 3 as a novel marker of acidic malignant lesions. Proc Natl Acad Sci USA 112:9710–9715.
OpenUrl Abstract/FREE Full Text
↵
1. Golijanin J, et al.
(2016) Targeted imaging of urothelium carcinoma in human bladders by an ICG pHLIP peptide ex vivo. Proc Natl Acad Sci USA 113:11829–11834.
OpenUrl Abstract/FREE Full Text
↵
1. Macholl S, et al.
(2012) In vivo pH imaging with (99m)Tc-pHLIP. Mol Imaging Biol 14:725–734.
OpenUrl CrossRef PubMed
↵
1. Demoin DW, et al.
(2016) PET imaging of extracellular pH in tumors with 64Cu- and 18F-labeled pHLIP peptides: A structure-activity optimization study. Bioconjug Chem 27:2014–2023.
OpenUrl
↵
1. An M,
2. Wijesinghe D,
3. Andreev OA,
4. Reshetnyak YK,
5. Engelman DM
(2010) pH-(low)-insertion-peptide (pHLIP) translocation of membrane impermeable phalloidin toxin inhibits cancer cell proliferation. Proc Natl Acad Sci USA 107:20246–20250.
OpenUrl Abstract/FREE Full Text
↵
1. Wijesinghe D,
2. Engelman DM,
3. Andreev OA,
4. Reshetnyak YK
(2011) Tuning a polar molecule for selective cytoplasmic delivery by a pH (Low) insertion peptide. Biochemistry 50:10215–10222.
OpenUrl CrossRef PubMed
↵
1. Burns KE,
2. Robinson MK,
3. Thévenin D
(2015) Inhibition of cancer cell proliferation and breast tumor targeting of pHLIP-monomethyl auristatin E conjugates. Mol Pharm 12:1250–1258.
OpenUrl
↵
1. Burns KE,
2. Hensley H,
3. Robinson MK,
4. Thévenin D
(2017) Therapeutic efficacy of a family of pHLIP-MMAF conjugates in cancer cells and mouse models. Mol Pharm 14:415–422.
OpenUrl
↵
1. Cheng CJ, et al.
(2015) MicroRNA silencing for cancer therapy targeted to the tumour microenvironment. Nature 518:107–110.
OpenUrl CrossRef PubMed
↵
1. Yao L, et al.
(2013) pHLIP peptide targets nanogold particles to tumors. Proc Natl Acad Sci USA 110:465–470, and erratum (2013) 110:3651.
OpenUrl Abstract/FREE Full Text
↵
1. Daniels JL,
2. Crawford TM,
3. Andreev OA,
4. Reshetnyak YK
(2017) Synthesis and characterization of pHLIP® coated gold nanoparticles. Biochem Biophys Rep 10:62–69.
OpenUrl
↵
1. Yao L,
2. Daniels J,
3. Wijesinghe D,
4. Andreev OA,
5. Reshetnyak YK
(2013) pHLIP®-mediated delivery of PEGylated liposomes to cancer cells. J Control Release 167:228–237.
OpenUrl CrossRef PubMed
↵
1. Reshetnyak YK,
2. Segala M,
3. Andreev OA,
4. Engelman DM
(2007) A monomeric membrane peptide that lives in three worlds: In solution, attached to, and inserted across lipid bilayers. Biophys J 93:2363–2372.
OpenUrl CrossRef PubMed
↵
1. Reshetnyak YK,
2. Andreev OA,
3. Segala M,
4. Markin VS,
5. Engelman DM
(2008) Energetics of peptide (pHLIP) binding to and folding across a lipid bilayer membrane. Proc Natl Acad Sci USA 105:15340–15345.
OpenUrl Abstract/FREE Full Text
↵
1. Karabadzhak AG, et al.
(2012) Modulation of the pHLIP transmembrane helix insertion pathway. Biophys J 102:1846–1855.
OpenUrl CrossRef PubMed
↵
1. Weerakkody D, et al.
(2013) Family of pH (low) insertion peptides for tumor targeting. Proc Natl Acad Sci USA 110:5834–5839.
OpenUrl Abstract/FREE Full Text
↵
1. Cruz-Monserrate Z, et al.
(2014) Targeting pancreatic ductal adenocarcinoma acidic microenvironment. Sci Rep 4:4410.
OpenUrl PubMed
↵
1. Adochite RC, et al.
(2016) Comparative study of tumor targeting and biodistribution of pH (low) insertion peptides (pHLIP peptides) conjugated with different fluorescent dyes. Mol Imaging Biol 18:686–696.
OpenUrl
↵
1. Onyango JO, et al.
(2015) Noncanonical amino acids to improve the pH response of pHLIP insertion at tumor acidity. Angew Chem Int Ed Engl 54:3658–3663.
OpenUrl
↵
1. Nguyen VP,
2. Alves DS,
3. Scott HL,
4. Davis FL,
5. Barrera FN
(2015) A novel soluble peptide with pH-responsive membrane insertion. Biochemistry 54:6567–6575.
OpenUrl
↵
1. Hunt JF,
2. Rath P,
3. Rothschild KJ,
4. Engelman DM
(1997) Spontaneous, pH-dependent membrane insertion of a transbilayer alpha-helix. Biochemistry 36:15177–15192.
OpenUrl CrossRef PubMed
↵
1. Cabaniss SE,
2. Pugh KC,
3. Pedersen LG,
4. Hiskey RG
(1991) Proton, calcium, and magnesium binding by peptides containing gamma-carboxyglutamic acid. Int J Pept Protein Res 37:33–38.
OpenUrl PubMed
↵
1. Shikamoto Y,
2. Morita T,
3. Fujimoto Z,
4. Mizuno H
(2003) Crystal structure of Mg2+- and Ca2+-bound Gla domain of factor IX complexed with binding protein. J Biol Chem 278:24090–24094.
OpenUrl Abstract/FREE Full Text
↵
1. Huang M,
2. Furie BC,
3. Furie B
(2004) Crystal structure of the calcium-stabilized human factor IX Gla domain bound to a conformation-specific anti-factor IX antibody. J Biol Chem 279:14338–14346.
OpenUrl Abstract/FREE Full Text
↵
1. Barrera FN, et al.
(2011) Roles of carboxyl groups in the transmembrane insertion of peptides. J Mol Biol 413:359–371.
OpenUrl CrossRef PubMed
↵
1. Moshnikova A,
2. Moshnikova V,
3. Andreev OA,
4. Reshetnyak YK
(2013) Antiproliferative effect of pHLIP-amanitin. Biochemistry 52:1171–1178.
OpenUrl CrossRef PubMed
↵
1. Weerakkody D, et al.
(2016) Novel pH-sensitive cyclic peptides. Sci Rep 6:31322.
OpenUrl
↵
1. Sharma GP,
2. Reshetnyak YK,
3. Andreev OA,
4. Karbach M,
5. Müller G
(2015) Coil-helix transition of polypeptide at water-lipid interface. J Stat Mech-Theory Exp, P01034.
↵
1. Krauson AJ, et al.
(2015) Conformational fine-tuning of pore-forming peptide potency and selectivity. J Am Chem Soc 137:16144–16152.
OpenUrl
↵
1. Kalafatis M,
2. Egan JO,
3. van’t Veer C,
4. Mann KG
(1996) Regulation and regulatory role of gamma-carboxyglutamic acid containing clotting factors. Crit Rev Eukaryot Gene Expr 6:87–101.
OpenUrl PubMed
↵
1. Reshetnyak YK,
2. Andreev OA,
3. Lehnert U,
4. Engelman DM
(2006) Translocation of molecules into cells by pH-dependent insertion of a transmembrane helix. Proc Natl Acad Sci USA 103:6460–6465.
OpenUrl Abstract/FREE Full Text
↵
1. Burns KE,
2. Thévenin D
(2015) Down-regulation of PAR1 activity with a pHLIP-based allosteric antagonist induces cancer cell death. Biochem J 472:287–295.
OpenUrl Abstract/FREE Full Text
↵
1. Burns KE,
2. McCleerey TP,
3. Thévenin D
(2016) pH-selective cytotoxicity of pHLIP-antimicrobial peptide conjugates. Sci Rep 6:28465.
OpenUrl

Article Alerts

Email Article

Citation Tools

Request Permissions

Proceedings of the National Academy of Sciences: 115 (12)

Table of Contents

Submit

Article Classifications

Biological Sciences
Medical Sciences

[1] ↵
Marusyk A,
Polyak K
(2010) Tumor heterogeneity: Causes and consequences. Biochim Biophys Acta 1805:105–117.
OpenUrl CrossRef PubMed

[2] Marusyk A,

[3] Polyak K

[4] ↵
Gillies RJ,
Verduzco D,
Gatenby RA
(2012) Evolutionary dynamics of carcinogenesis and why targeted therapy does not work. Nat Rev Cancer 12:487–493.
OpenUrl CrossRef PubMed

[5] Gillies RJ,

[6] Verduzco D,

[7] Gatenby RA

[8] ↵
Lloyd MC, et al.
(2016) Darwinian dynamics of intratumoral heterogeneity: Not solely random mutations but also variable environmental selection forces. Cancer Res 76:3136–3144.
OpenUrl Abstract/FREE Full Text

[9] Lloyd MC, et al.

[10] ↵
Estrella V, et al.
(2013) Acidity generated by the tumor microenvironment drives local invasion. Cancer Res 73:1524–1535.
OpenUrl Abstract/FREE Full Text

[11] Estrella V, et al.

[12] ↵
Zhang X,
Lin Y,
Gillies RJ
(2010) Tumor pH and its measurement. J Nucl Med 51:1167–1170.
OpenUrl Abstract/FREE Full Text

[13] Zhang X,

[14] Lin Y,

[15] Gillies RJ

[16] ↵
Hashim AI,
Zhang X,
Wojtkowiak JW,
Martinez GV,
Gillies RJ
(2011) Imaging pH and metastasis. NMR Biomed 24:582–591.
OpenUrl CrossRef PubMed

[17] Hashim AI,

[18] Zhang X,

[19] Wojtkowiak JW,

[20] Martinez GV,

[21] Gillies RJ

[22] ↵
Anderson M,
Moshnikova A,
Engelman DM,
Reshetnyak YK,
Andreev OA
(2016) Probe for the measurement of cell surface pH in vivo and ex vivo. Proc Natl Acad Sci USA 113:8177–8181.
OpenUrl Abstract/FREE Full Text

[23] Anderson M,

[24] Moshnikova A,

[25] Engelman DM,

[26] Reshetnyak YK,

[27] Andreev OA

[28] ↵
Reshetnyak YK, et al.
(2011) Measuring tumor aggressiveness and targeting metastatic lesions with fluorescent pHLIP. Mol Imaging Biol 13:1146–1156.
OpenUrl CrossRef PubMed

[29] Reshetnyak YK, et al.

[30] ↵
Adochite RC, et al.
(2014) Targeting breast tumors with pH (low) insertion peptides. Mol Pharm 11:2896–2905.
OpenUrl CrossRef

[31] Adochite RC, et al.

[32] ↵
Tapmeier TT, et al.
(2015) The pH low insertion peptide pHLIP variant 3 as a novel marker of acidic malignant lesions. Proc Natl Acad Sci USA 112:9710–9715.
OpenUrl Abstract/FREE Full Text

[33] Tapmeier TT, et al.

[34] ↵
Golijanin J, et al.
(2016) Targeted imaging of urothelium carcinoma in human bladders by an ICG pHLIP peptide ex vivo. Proc Natl Acad Sci USA 113:11829–11834.
OpenUrl Abstract/FREE Full Text

[35] Golijanin J, et al.

[36] ↵
Macholl S, et al.
(2012) In vivo pH imaging with (99m)Tc-pHLIP. Mol Imaging Biol 14:725–734.
OpenUrl CrossRef PubMed

[37] Macholl S, et al.

[38] ↵
Demoin DW, et al.
(2016) PET imaging of extracellular pH in tumors with 64Cu- and 18F-labeled pHLIP peptides: A structure-activity optimization study. Bioconjug Chem 27:2014–2023.
OpenUrl

[39] Demoin DW, et al.

[40] ↵
An M,
Wijesinghe D,
Andreev OA,
Reshetnyak YK,
Engelman DM
(2010) pH-(low)-insertion-peptide (pHLIP) translocation of membrane impermeable phalloidin toxin inhibits cancer cell proliferation. Proc Natl Acad Sci USA 107:20246–20250.
OpenUrl Abstract/FREE Full Text

[41] An M,

[42] Wijesinghe D,

[43] Andreev OA,

[44] Reshetnyak YK,

[45] Engelman DM

[46] ↵
Wijesinghe D,
Engelman DM,
Andreev OA,
Reshetnyak YK
(2011) Tuning a polar molecule for selective cytoplasmic delivery by a pH (Low) insertion peptide. Biochemistry 50:10215–10222.
OpenUrl CrossRef PubMed

[47] Wijesinghe D,

[48] Engelman DM,

[49] Andreev OA,

[50] Reshetnyak YK

[51] ↵
Burns KE,
Robinson MK,
Thévenin D
(2015) Inhibition of cancer cell proliferation and breast tumor targeting of pHLIP-monomethyl auristatin E conjugates. Mol Pharm 12:1250–1258.
OpenUrl

[52] Burns KE,

[53] Robinson MK,

[54] Thévenin D

[55] ↵
Burns KE,
Hensley H,
Robinson MK,
Thévenin D
(2017) Therapeutic efficacy of a family of pHLIP-MMAF conjugates in cancer cells and mouse models. Mol Pharm 14:415–422.
OpenUrl

[56] Burns KE,

[57] Hensley H,

[58] Robinson MK,

[59] Thévenin D

[60] ↵
Cheng CJ, et al.
(2015) MicroRNA silencing for cancer therapy targeted to the tumour microenvironment. Nature 518:107–110.
OpenUrl CrossRef PubMed

[61] Cheng CJ, et al.

[62] ↵
Yao L, et al.
(2013) pHLIP peptide targets nanogold particles to tumors. Proc Natl Acad Sci USA 110:465–470, and erratum (2013) 110:3651.
OpenUrl Abstract/FREE Full Text

[63] Yao L, et al.

[64] ↵
Daniels JL,
Crawford TM,
Andreev OA,
Reshetnyak YK
(2017) Synthesis and characterization of pHLIP® coated gold nanoparticles. Biochem Biophys Rep 10:62–69.
OpenUrl

[65] Daniels JL,

[66] Crawford TM,

[67] Andreev OA,

[68] Reshetnyak YK

[69] ↵
Yao L,
Daniels J,
Wijesinghe D,
Andreev OA,
Reshetnyak YK
(2013) pHLIP®-mediated delivery of PEGylated liposomes to cancer cells. J Control Release 167:228–237.
OpenUrl CrossRef PubMed

[70] Yao L,

[71] Daniels J,

[72] Wijesinghe D,

[73] Andreev OA,

[74] Reshetnyak YK

[75] ↵
Reshetnyak YK,
Segala M,
Andreev OA,
Engelman DM
(2007) A monomeric membrane peptide that lives in three worlds: In solution, attached to, and inserted across lipid bilayers. Biophys J 93:2363–2372.
OpenUrl CrossRef PubMed

[76] Reshetnyak YK,

[77] Segala M,

[78] Andreev OA,

[79] Engelman DM

[80] ↵
Reshetnyak YK,
Andreev OA,
Segala M,
Markin VS,
Engelman DM
(2008) Energetics of peptide (pHLIP) binding to and folding across a lipid bilayer membrane. Proc Natl Acad Sci USA 105:15340–15345.
OpenUrl Abstract/FREE Full Text

[81] Reshetnyak YK,

[82] Andreev OA,

[83] Segala M,

[84] Markin VS,

[85] Engelman DM

[86] ↵
Karabadzhak AG, et al.
(2012) Modulation of the pHLIP transmembrane helix insertion pathway. Biophys J 102:1846–1855.
OpenUrl CrossRef PubMed

[87] Karabadzhak AG, et al.

[88] ↵
Weerakkody D, et al.
(2013) Family of pH (low) insertion peptides for tumor targeting. Proc Natl Acad Sci USA 110:5834–5839.
OpenUrl Abstract/FREE Full Text

[89] Weerakkody D, et al.

[90] ↵
Cruz-Monserrate Z, et al.
(2014) Targeting pancreatic ductal adenocarcinoma acidic microenvironment. Sci Rep 4:4410.
OpenUrl PubMed

[91] Cruz-Monserrate Z, et al.

[92] ↵
Adochite RC, et al.
(2016) Comparative study of tumor targeting and biodistribution of pH (low) insertion peptides (pHLIP peptides) conjugated with different fluorescent dyes. Mol Imaging Biol 18:686–696.
OpenUrl

[93] Adochite RC, et al.

[94] ↵
Onyango JO, et al.
(2015) Noncanonical amino acids to improve the pH response of pHLIP insertion at tumor acidity. Angew Chem Int Ed Engl 54:3658–3663.
OpenUrl

[95] Onyango JO, et al.

[96] ↵
Nguyen VP,
Alves DS,
Scott HL,
Davis FL,
Barrera FN
(2015) A novel soluble peptide with pH-responsive membrane insertion. Biochemistry 54:6567–6575.
OpenUrl

[97] Nguyen VP,

[98] Alves DS,

[99] Scott HL,

[100] Davis FL,

[101] Barrera FN

[102] ↵
Hunt JF,
Rath P,
Rothschild KJ,
Engelman DM
(1997) Spontaneous, pH-dependent membrane insertion of a transbilayer alpha-helix. Biochemistry 36:15177–15192.
OpenUrl CrossRef PubMed

[103] Hunt JF,

[104] Rath P,

[105] Rothschild KJ,

[106] Engelman DM

[107] ↵
Cabaniss SE,
Pugh KC,
Pedersen LG,
Hiskey RG
(1991) Proton, calcium, and magnesium binding by peptides containing gamma-carboxyglutamic acid. Int J Pept Protein Res 37:33–38.
OpenUrl PubMed

[108] Cabaniss SE,

[109] Pugh KC,

[110] Pedersen LG,

[111] Hiskey RG

[112] ↵
Shikamoto Y,
Morita T,
Fujimoto Z,
Mizuno H
(2003) Crystal structure of Mg2+- and Ca2+-bound Gla domain of factor IX complexed with binding protein. J Biol Chem 278:24090–24094.
OpenUrl Abstract/FREE Full Text

[113] Shikamoto Y,

[114] Morita T,

[115] Fujimoto Z,

[116] Mizuno H

[117] ↵
Huang M,
Furie BC,
Furie B
(2004) Crystal structure of the calcium-stabilized human factor IX Gla domain bound to a conformation-specific anti-factor IX antibody. J Biol Chem 279:14338–14346.
OpenUrl Abstract/FREE Full Text

[118] Huang M,

[119] Furie BC,

[120] Furie B

[121] ↵
Barrera FN, et al.
(2011) Roles of carboxyl groups in the transmembrane insertion of peptides. J Mol Biol 413:359–371.
OpenUrl CrossRef PubMed

[122] Barrera FN, et al.

[123] ↵
Moshnikova A,
Moshnikova V,
Andreev OA,
Reshetnyak YK
(2013) Antiproliferative effect of pHLIP-amanitin. Biochemistry 52:1171–1178.
OpenUrl CrossRef PubMed

[124] Moshnikova A,

[125] Moshnikova V,

[126] Andreev OA,

[127] Reshetnyak YK

[128] ↵
Weerakkody D, et al.
(2016) Novel pH-sensitive cyclic peptides. Sci Rep 6:31322.
OpenUrl

[129] Weerakkody D, et al.

[130] ↵
Sharma GP,
Reshetnyak YK,
Andreev OA,
Karbach M,
Müller G
(2015) Coil-helix transition of polypeptide at water-lipid interface. J Stat Mech-Theory Exp, P01034.

[131] Sharma GP,

[132] Reshetnyak YK,

[133] Andreev OA,

[134] Karbach M,

[135] Müller G

[136] ↵
Krauson AJ, et al.
(2015) Conformational fine-tuning of pore-forming peptide potency and selectivity. J Am Chem Soc 137:16144–16152.
OpenUrl

[137] Krauson AJ, et al.

[138] ↵
Kalafatis M,
Egan JO,
van’t Veer C,
Mann KG
(1996) Regulation and regulatory role of gamma-carboxyglutamic acid containing clotting factors. Crit Rev Eukaryot Gene Expr 6:87–101.
OpenUrl PubMed

[139] Kalafatis M,

[140] Egan JO,

[141] van’t Veer C,

[142] Mann KG

[143] ↵
Reshetnyak YK,
Andreev OA,
Lehnert U,
Engelman DM
(2006) Translocation of molecules into cells by pH-dependent insertion of a transmembrane helix. Proc Natl Acad Sci USA 103:6460–6465.
OpenUrl Abstract/FREE Full Text

[144] Reshetnyak YK,

[145] Andreev OA,

[146] Lehnert U,

[147] Engelman DM

[148] ↵
Burns KE,
Thévenin D
(2015) Down-regulation of PAR1 activity with a pHLIP-based allosteric antagonist induces cancer cell death. Biochem J 472:287–295.
OpenUrl Abstract/FREE Full Text

[149] Burns KE,

[150] Thévenin D

[151] ↵
Burns KE,
McCleerey TP,
Thévenin D
(2016) pH-selective cytotoxicity of pHLIP-antimicrobial peptide conjugates. Sci Rep 6:28465.
OpenUrl

[152] Burns KE,

[153] McCleerey TP,

[154] Thévenin D

Main menu

User menu

Search

New Research In

Peptides of pHLIP family for targeted intracellular and extracellular delivery of cargo molecules to tumors

Significance

Abstract

Results

The pHLIP Constructs.

Biophysical Steady-State and Kinetics Studies.

Intracellular Delivery of Polar Cargo.

Tumor Targeting.

Discussion

Materials and Methods

The pHLIP Characterization and pHLIP Bundle Synthesis.

Liposome Preparation.

Steady-State Fluorescence Measurements.

The pH Dependence Measurements.

Steady-State CD and OCD Measurements.

Kinetics Measurements.

Amanitin pHLIP Conjugates.

Cell Proliferation Assay.

Fluorescent pHLIP Conjugates.

Ex Vivo Imaging.

Acknowledgments

Footnotes

References

Citation Manager Formats

Article Classifications

You May Also be Interested in

Similar Articles

Articles

PNAS Portals

Information

Main menu

User menu

Search

New Research In

Physical Sciences

Featured Portals

Articles by Topic

Social Sciences

Featured Portals

Articles by Topic

Biological Sciences

Featured Portals

Articles by Topic

Peptides of pHLIP family for targeted intracellular and extracellular delivery of cargo molecules to tumors

Significance

Abstract

Results

The pHLIP Constructs.

Biophysical Steady-State and Kinetics Studies.

Intracellular Delivery of Polar Cargo.

Tumor Targeting.

Discussion

Materials and Methods

The pHLIP Characterization and pHLIP Bundle Synthesis.

Liposome Preparation.

Steady-State Fluorescence Measurements.

The pH Dependence Measurements.

Steady-State CD and OCD Measurements.

Kinetics Measurements.

Amanitin pHLIP Conjugates.

Cell Proliferation Assay.

Fluorescent pHLIP Conjugates.

Ex Vivo Imaging.

Acknowledgments

Footnotes

References

Citation Manager Formats

Sign up for Article Alerts

Article Classifications

Jump to section

You May Also be Interested in

Similar Articles

Articles

PNAS Portals

Information