Nonthreaded Isomers of Sungsanpin and Ulleungdin Lasso Peptides Inhibit H1299 Cancer Cell Migration

Abstract

Lasso peptides are a structurally distinct class of biologically active natural products defined by their short sequences with impressively interlocked tertiary structures. Their characteristic peptide [1]rotaxane motif confers marked proteolytic and thermal resiliency, and reports on their diverse biological functions have been credited to their exceptional sequence variability. Because of these unique properties, taken together with improved technologies for their biosynthetic production, lasso peptides are emerging as a designable scaffold for peptide-based therapeutic discovery and development. Although the defined structure of lasso peptides is recognized for its remarkable properties, the role of the motif in imparting bioactivity is less understood. For example, sungsanpin and ulleungdin are natural lasso peptides that similarly exhibit encouraging cell migration inhibitory activities in A549 lung carcinoma epithelial cells, despite sharing only one-third of the sequence homology. We hypothesized that the shape of the lasso motif is beneficial for the preorganization of the conserved residues, which might be partially retained in variants lacking the threaded structure. Herein, we describe solid-phase peptide synthesis strategies to prepare acyclic, head-to-side chain (branched), and head-to-tail (macrocyclic) cyclic variants based on the sungsanpin (Sun) and ulleungdin (Uln) sequences. Proliferation assays and time-lapse cell motility imaging studies were used to evaluate the cell inhibitory properties of natural Sun compared with the synthetic Sun and Uln isomers. These studies demonstrate that the lasso motif is not a required feature to slow cancer cell migration and more generally show that these nonthreaded isomers can retain similar activity to the natural lasso peptide despite the differences in their overall structures.

Graphical Abstract

INTRODUCTION

Due to their fascinatingly threaded structures, remarkable stability to proteolytic degradation, diverse biological activities (e.g., glucagon receptor antagonism, bacterial RNA polymerase (RNAP) inhibition, cancer cell migration/invasion inhibition), and promise as scaffolds that can be engineered to exhibit novel functions, lasso peptides have drawn considerable attention as structurally distinct ribosomally synthesized and post-translationally modified peptides (RiPPs).¹ Since their discovery, applied site-directed mutagenesis methods have curated sequence-diverse lasso peptide libraries to uncover trends that can help improve and diversify lasso peptide functions.² To accompany the surge of lasso peptide discovery efforts, in silico screening predictions and other bioinformatic approaches have been used to expedite the characterization of previously unknown lasso peptide biosynthetic gene clusters.^3,4 As a result, the development of heterologous expression systems (e.g., Escherichia coli), the advent of lasso peptide cell-free biosynthesis, and residue-selective peptide modification strategies have enabled access to many natural and unnatural lasso peptide variants.^5–8 Therapeutic discovery efforts have focused on the design of sequence-modified lasso peptides to improve and generate novel therapeutic modalities. For example, sequence mutations of the prototypical lasso peptide, microcin J25 1a (MccJ25), showed that ring-localized mutations can decrease bacterial RNAP inhibitory activity, while loop-localized mutations can impart an up to 5-fold increase in the same activity (Figure 1).² Moreover, Marahiel and co-workers showed that an RGD epitope grafted onto the native sequence loop (1a′) of the inactive MccJ25 successfully converted it to a nanomolar integrin inhibitor.⁹

Figure 1.

Bioactive class II lasso peptides, microcin J25 1a, sungsanpin 2a, and ulleungdin 3a, inspire the development of modified lasso peptides (1a′) and stable cyclic peptides (4) that change or retain the biological properties of a parent lasso peptide. Sungsanpin (2a) and ulleungdin (3a) lasso peptides exhibit cell migration inhibitory activities in A549 lung carcinoma epithelial cells, suggesting that their lasso motifs may be a less critical recognition feature for their unknown mechanism(s) of action. The representations of sungsanpin (2a) and ulleungdin (3a) reflect their suggested stopper residues, Trp (W, ref 21) and Tyr (Y, ref 22), as supported by through-space 2-D NMR spectroscopy.

The promising potential of lasso peptides has led to considerable effort toward general access to these scaffolds through chemical synthesis—a challenge that has yet to be realized. Direct attempts to synthesize lasso peptides by the cyclization of their acyclic precursors have been unsuccessful, yielding only head-to-side-chain cyclic lasso peptide isomers. The only reported approach for accessing a native lasso peptide, the disulfide-containing BI-32169, incorporated a cryptand-imidazolium host–guest complex linked to side chains in the ring to template the key lasso-forming macrolactamization event.¹⁰ However, this elaborate strategy required sequence-dependent anchoring steps, thereby limiting its general adoption. As an alternative approach, the Bode and Evans groups have independently developed synthetic [1]-rotaxanes utilizing cryptand-templated rings and peptide-based loops.^11,12 Bode and co-workers showed that various peptide sequences can be grafted into the common rotaxane structure that importantly exhibited lasso peptide-like thermal and protease stabilities, which highlights their potential to serve as lasso peptide mimics.¹¹

On the other hand, cases of nonthreaded lasso peptide derivatives have been shown to possess varying levels of biological activity. Interestingly, head-to-side chain 1b and head-to-tail cyclic 1c constitutional isomers (nonthreaded) of natural MccJ25 do not exhibit bacterial RNAP inhibitory activities.¹³ However, such drastic structural deviations are not always met with a loss of activity. Kaur and co-workers showed that disulfide-containing bicyclic and branched variants of the MccJ25 sequence, although reduced relative to the natural lasso, retained significant antibacterial activity against Salmonella Newport despite being nonthreaded side-chain-to-side-chain cyclic peptides.¹⁴ Fliss and co-workers compared the activity of similar cyclic disulfide MccJ25 variants against various Gram-negative and Gram-positive bacterial strains, demonstrating their ability to decrease bacterial RNAP activity.¹⁵ The beneficial role of this lasso motif can be seen in the crystal structure of MccJ25 bound within the secondary channel of RNAP determined by Darst, Link, and co-workers.¹⁶ MccJ25 binds deep in this channel, with the ring and tail facing the interior of the active site, which agrees with the effect of mutations on its activity.¹⁶ Still, carefully designed nonthreaded peptides can improve activity relative to the lasso peptide. RES-701–1 is a potent endothelin B receptor (ET_B)-selective antagonist, whereas both the acyclic and head-to-side-chain isomers of RES-701–1 have significantly reduced binding activities.¹⁷ However, Yamasaki and co-workers showed that a head-to-side-chain hybrid peptide combining the ring segment of RES-701–1 with a C-terminal fragment of endothelin-1 gave improved ET_B binding activity, which improved further with the acyclic analog.^18,19 The findings from these studies support future efforts to evaluate the properties of nonthreaded acyclic and cyclic isomers of lasso peptides given their ease of access through chemical synthesis.

Cancer metastasis is a major cause of death in cancer patients that is initiated by migration of the cancer cells away from the primary tumor, and more efficacious treatments that inhibit this process are needed.²⁰ Toward this end, we became interested in sungsanpin 2a (isolated from Streptomyces sp. SNJ013)²¹ and ulleungdin 3a (isolated from Streptomyces sp. KCB13F003),²² which exhibit similar cell migration inhibitory activities in A549 lung carcinoma epithelial cells despite sharing only one-third sequence homology (Figure 2). This observation suggests that their similar lasso motifs may be a critical preorganization element for the conserved residues to impart their unknown mechanism(s) of action. We hypothesized that, although the lasso motif is beneficial, the arrangement of their sequence could be a more essential determinant for their activity. To examine this further, we designed a series of synthetically accessible sungsanpin (Sun-2b–2d) and ulleungdin (Uln-3b–3d) isomers to evaluate the spatial relationship of these sequences toward activity using time-lapse cell motility imaging assays with H1299 human nonsmall cell lung carcinoma cells. Specifically, we posited that partial conservation of the conserved sequence arrangement in these lasso peptides could bestow some comparable biological responses across the series. The head-to-side-chain cyclic isomers, Sun-2b and Uln-3b, were anticipated to be most like their respective threaded counterparts, sungsanpin 2a and ulleungdin 3a, whereas the acyclic peptides, Sun-2d and Uln-3d, were thought to have more disordered conformations and would therefore be the least active. Similarly, Tulla-Puche and co-workers prepared a sungsanpin-inspired bicyclic peptide 4 (nonthreaded) that shows encouraging lasso peptide-like thermal and protease stability properties.²³ Like the reported bioactivity of sungsanpin 2a, the bicyclic variant 4 (at 100 μM concentration) decreased cancer cell motility, as evaluated by transwell tumor cell migration and invasion assays using A549 lung carcinoma epithelial cells. However, comparison to sungsanpin 2a and its nonthreaded isomer, Sun-2b, was not conducted. Recently, Zou and co-workers compared the A549 cell invasion inhibitory activities of Sun-2b with the acyclic and head-to-tail cyclic peptides constructed from the loop–tail regions of sungsanpin 2a.²⁴ These results were unknown at the onset of our studies and do not include any comparison to natural lasso sungsanpin 2a. We reasoned that time-lapse cell motility imaging assays with H1299 human nonsmall cell lung carcinoma cells, shown by one of our groups to exhibit faster in vitro motion than A549 cells, could better inform the relationship of sequence arrangement to activity for the lasso motif compared to a series of Sun-2b/2c and Uln-3b/3c lasso peptide isomers.²⁵

Figure 2.

Motivating question: Is the shape, the lasso motif, more important than sequence arrangement for the biochemical activities of sungsanpin and ulleungdin lasso peptides?

RESULTS AND DISCUSSION

Chemical Synthesis of Sun and Uln-Lasso Peptide Isomers.

To evaluate these relationships across a series of Sun/Uln-lasso peptide isomers, we developed solid-phase peptide synthesis (SPPS) strategies to access acyclic (2d/3d, control peptides) and branched cyclic (2b/3b) and macrocyclic (2c/3c) variants. We recognized that some key synthetic residues would enable modular access to the cyclic variants in this series (Figure 3). All target peptides were prepared by automated Fmoc-based solid-phase peptide synthesis techniques using a 2-chlorotrityl chloride (2-CTC)-based linker and solid support to attach the peptide to a carboxylic acid moiety. Synthetic resin-bound peptides were cleaved and globally deprotected by using acidic conditions. Crude peptides were purified by using reverse-phase high-performance liquid chromatography (rpHPLC).

Figure 3.

On-resin strategies for the synthesis of acyclic (Sun-2d and Uln-3d), branched cyclic (Sun-2b and Uln-3b), and macrocyclic variants (Sun-2c and Uln-3c) of sungsanpin and ulleungdin lasso peptides.

Synthesis of Acyclic Peptides Sun-2d and Uln-2d.

Initial efforts to prepare acyclic peptides, Sun-2d and Uln-3d, using the standard incorporation of Fmoc-Asp(Ot-Bu)–OH at respective Asp8 positions were challenged by aspartimide formation, which gave inseparable mixtures. This frequently encountered issue prompted the application of strategies to avoid aspartimide formation. To prepare Sun-2d, we incorporated a pseudoproline dipeptide at Asp8Ser9 using commercial Fmoc-Asp(Ot-Bu)-Ser[ψ(Me,Me)Pro]–OH (Figure 3A and SI pg. S12). Fortunately, this strategy also improved subsequent coupling efficiencies and provided Sun-2d in a 40% overall yield following purification by rpHPLC. Because target Uln-3d lacks a Ser9 or Thr9 residue needed to use the same pseudoproline dipeptide approach, we applied the cyanosulfurylide (CSY) strategy recently developed by Bode and co-workers.²⁶ This strategy temporarily masks the requisite carboxylic acid functionality at Asp8, and unlike conventionally used sterically encumbered ester derivatives (e.g., Asp(Ompe)), it avoids the potential for aspartimide formation altogether. As described, we found that Fmoc-Asp(CSY)–OH S-3 is readily prepared (SI Scheme S1) and incorporated using Fmoc-based SPPS to access an Asp8(CSY)-containing peptide (Figure 3A and SI pg. S14) following acid-promoted cleavage and deprotection. Treatment with N-chlorosuccinimide (NCS) readily converts the CSY moiety at position 8 into a carboxylic acid, providing target Uln-3d. Although we observed a few overoxidation products that likely derive from oxidation at Trp5, we were able to obtain Uln-3d in a 14% overall yield following purification by rpHPLC. Next, we set out to prepare the branched cyclic lasso peptide isomers, Sun-2b and Uln-3b.

Synthesis of Branched Cyclic Lasso Peptide Isomers, Sun-2b and Uln-3b.

Toward Sun-2b and Uln-3b, we developed on-resin cyclization strategies to minimize required purification events (Figure 3B). Generally, strategies to prepare branched cyclic peptides leverage orthogonal protection strategies to permit N-terminal-selective cyclization at a carboxylic acid-containing residue side chain. Informed by the potential for aspartimide formation at Asp8 in each sequence, we avoided the conventional strategy that would use an Asp8(Oall) residue. We recognized that most lasso peptides, including sungsanpin and ulleungdin, have a Xaa#(Gly1) isopeptide linkage, where Xaa = Asp or Glu, and # = 7–9.³ Accordingly, we designed the novel dipeptide 5 (Fmoc-Asp(Gly(Oall))–OH) that incorporates the Asp8-(Gly1) linkage at an early stage. The coupling of H-Gly-Oall S-5 and Fmoc-Asp(Opfp)-OtBu S-6 (SI Scheme S2), followed by deprotection of the resulting dipeptide provides Fmoc-Asp(Gly(Oall))–OH 5 in a 94% isolated yield (SI pg. S10). Dipeptide 5 is then readily incorporated using standard Fmoc-based SPPS techniques. Sequence elongation, on-resin deallylation, and cyclization, followed by resin cleavage and global deprotection, provide Sun-2b and Uln-3b in 15 and 6% overall yields, respectively. We found this strategy to circumvent aspartimide formation and enable streamlined access to branched cyclic lasso peptide isomers Sun-2b and Uln-3b. Given the efficient preparation and utility of dipeptide 5 for on-resin peptide cyclization, other branched cyclic lasso peptide isomers should be accessible with this method. Next, we set out to prepare the macrocyclic lasso peptide isomers Sun-2c and Uln-3c.

Synthesis of Macrocyclic Lasso Peptide Isomers, Sun-2c and Uln-2c.

In general, head-to-tail cyclization of peptides can be especially challenging. Large ring-forming cyclizations incur many entropic penalties during backbone preorganization.^27,28 Solution-phase cyclization strategies often leverage high dilution or pseudo-high dilution conditions to minimize oligomerization. However, such approaches can be met with epimerization and hydrolysis at the activated residue. We first considered chemoselective ligation approaches, such as the use of an Ala or native chemical ligation strategy, which has been demonstrated for the preparation of MccJ25-1c.^13,29 However, both sungsanpin and ulleungdin sequences lack Ala residues, and we sought a modular strategy that could be applied toward both Sun-2c and Uln-3c. We recognized that the Asp8 residue common to targets Sun-2c and Uln-3c could serve as a side-chain anchoring point (Figure 3C). Unsurprisingly, attempts to cyclize side-chain-anchored peptides were unsuccessful for both sequences. In response, we incorporated N-(2,4-dimethoxy)benzyl (Dmb) groups in hopes of inducing backbone preorganization and minimizing solid-phase aggregation to favor macrocyclization.³⁰ In each series, we found that the incorporation of Dmb groups, using Fmoc-Leu15-(Dmb)Gly1-OH for Sun-2c and Fmoc-Gly15(Dmb)Gly1-OH for Uln-3c (Figure 3C and SI pgs. S20 and S22, respectively), enabled successful macrocyclizations to afford Sun-2c and Uln-3c in 2 and 6% overall yields, respectively, following resin cleavage, global deprotection, and purification by rpHPLC.

Characterization Studies.

Characterization Using MS/MS.

To evaluate purity and identify, we first characterized synthetic variants in the series using analytical LC-MS monitored by UV with λ_max = 220 nm (see the SI). Next, each variant was characterized using tandem mass spectrometry, MS/MS. Interestingly, while sungsanpin and ulleungdin peptides share only one-third sequence homology, their molecular weights are almost identical. For example, the synthetic acyclic peptides, Sun-2d and Uln-3d, have calculated [M + H⁺]⁺ ions = 1610.8 m/z. Although we know sequence because of how the peptides were prepared, MS/MS analysis provides data to confirm connectivity. As expected, both Sun-2d and Uln-3d displayed the indicative b and y ions associated with the fragmentation of their selected [M + 2H⁺]²⁺ ions (m/z = 806). For branched cyclic isomers, Sun-2b and Uln-3b, we observed fragmentation patterns common to previously reported branched cyclic peptides.^31,32 The MS/MS spectra for Sun-2b and Uln-3b, with selection for [M + 2H⁺]²⁺ parent ions (m/z = 797), showed key tail-localized fragmentation resulting in daughter ions with the preservation of the branched cyclic (ring) features. The MS/MS spectra for macrocyclic isomers Sun-2c and Uln-3c, with selection for [M + 2H⁺]²⁺ parent ions (m/z = 797), showed fragmentation patterns that importantly confirmed the synthetically installed Asp8Ser9 and Asp8Ile9 linkages, respectively.

Traveling Wave Ion Mobility Spectrometry–Mass Spectrometry.

Ion mobility spectrometry–mass spectrometry is used to separate isobaric ions based on their size, shape, and charge. We turned to traveling wave ion mobility spectrometry (TWIMS) performed on the commercially available Waters Cyclic IMS-MS platform to learn more about the shape of synthetic isomers in the sungsanpin and ulleungdin series (Figure 4).^31,32 We selected the [M + 2H⁺]⁺ ion across each series, and the relative intensity was normalized based on the highest acquired intensity measurement.

Figure 4.

Eight m (A) and 8 m (B) cyclic ion mobility spectrometry–mass spectrometry cIMS–MS separations differentiate lassoed, branched, and macrocyclic peptide isomers.

As anticipated, natural sungsanpin 2a exhibited the earliest arrival time (t_a = 81.0 ms, highest mobility), which can be attributed to its compact threaded shape (Figure 4A). Interestingly, 2a also showed a wider peak shape relative to branched Sun-2b (t_a = 82.8 ms) and macrocyclic Sun-2c (t_a = 84.3 ms), which suggests that additional conformations may be accessible in the gas phase. For reference, the acyclic Sun-2d isomer (t_a = 86.6 ms) had the latest arrival, consistent with its higher degree of rotational freedom relative to lassoed and cyclic peptide variants. Interestingly, the study reported by Tulla-Puche and co-workers suggested that accessible side-chain-based electrostatic interactions may further stabilize (constrain) the dominant conformation for sungsanpin 2a in the gas phase.²³ The ability of branched Sun-2b and bicyclic 4 variants to access similar side-chain-based electrostatic interactions may explain why their arrival time distributions are more lasso-like (closest in arrival time) than the macrocyclic isomer, Sun-2c. In the ulleungdin series, the macrocyclic Uln-3c isomer (t_a = 54.0 ms) arrives before the branched Uln-3b isomer (t_a = 55.5 ms), followed by the acyclic isomer, Uln-3d (t_a = 59.4 ms) (Figure 4B). This surprising swap in arrival time suggests that Uln-3c (macrocycle) can adopt a conformation that is more compact than that of Ulb-3b (branched). The comparison of arrival times within a series showed informative trends. As anticipated, nonthreaded cyclic peptides arrive later than their lassoed isomer. The analysis of [M + 2H⁺]⁺ ions across series shows a difference in average arrival time (Sun series: avg t_a = ~83 ms vs Uln series: avg t_a = ~ 56 ms) attributable to sequence differences that shift their respective mobilities (i.e., size/shape).

Biological Activity.

Previous reports from one of our groups,²¹ and the Ahn group,²² independently showed that sungsanpin 2a and ulleungdin 3b at a 50 μM concentration inhibit 47 and 48% of A549 lung carcinoma epithelial cells, respectively, from invading the lower chamber in a transwell invasion assay. Based on the hypothesis that the native lasso motif provides beneficial preorganization to the conserved residues in the sequence, we evaluated the series of synthetic accessible Sun and Uln isomers using time-lapse cell imaging-based cancer cell migration assays to track individual cells and globally determine the relationship of sequence arrangement to this activity.²⁵ Prior to the migration assays, proliferation assays were run at 25 and 50 μM concentrations for each peptide over a 24 h period to confirm cell viability and proliferation relative to the untreated control (DMSO) (SI Figure S58). The cells were labeled with Janus green dye to measure the number of viable cells (absorbance at 620 nm) present after 24 h. When compared against cells treated with synthetic isomers, the cells treated with sungsanpin 2a exhibited the lowest number of proliferated cells, at 25 and 50 μM. More cells were observed for the 25 μM 2a vs 50 μM 2a treatments. Therefore, we chose the 25 μM concentration for subsequent time-lapse cell imaging studies. Imaging the cells over the course of 15 h and again after 24 h of 25 μM 2a showed that the cells are viable with the treatment (SI Figures S59–S64).

Previously, one of our groups showed that H1299 human nonsmall cell lung carcinoma cells exhibit faster in vitro motion than A549 cells, and accordingly, migration assays using H1299 cells can provide a better measure of substrate inhibitory effects.²⁵ Therefore, we chose to use H1299 cells in time-lapse cell imaging experiments. For all experiments, cell nuclei were labeled with DRAQ5 for automated tracking via a MATLAB workflow using custom software based on u-track multiple particle tracking. Only cells that displayed a persistent random walk were included in the velocity calculations. Velocity was calculated as the mean-squared displacement (MSD). In our results, we applied a p-value of 0.1 as significant, meaning that there is a 10% chance the samples are drawn from the same distribution. These measured median velocities are summarized in SI Table S1. The median velocity of cells treated with sungsanpin 2a was lower than control cells (Figure 5 and SI Table S1; lasso 2a median velocity = 0.23 μm/min vs DMSO velocity = 0.28 μm/min, p = 0.08). Similarly, the velocity of cells treated with branched Sun-2b was lower than control cells (branched Sun-2b median velocity = 0.21 μm/min vs DMSO velocity = 0.28 μm/min, p = 0.08). Thus, cell migration may be inhibited by treatments, 2a or Sun-2b, in H1299 cancer cells. Branched Uln-3b (MSD velocity = 0.23 μm/min) and macrocyclic Uln-3c (MSD velocity = 0.22 μm/min) isomers also exhibit a lower trending MSD velocity in comparison with control cells (DMSO treatment, MSD velocity = 0.28 μm/min, p-values = 0.15). Finally, at p-values greater than 0.40, we find that treatment with acyclic Sun-2d (MSD velocity = 0.27 μm/min, p = 0.43), acyclic Uln-3d (MSD velocity = 0.23 μm/min, p = 0.42), and macrocyclic Sun-2c (MSD velocity = 0.25 μm/min, p = 0.60) did not slow migration.

Figure 5.

Natural lasso, branched cyclic, and macrocyclic peptides exhibit better migration inhibition than acyclic or Sun-2c macropeptides at 25 μM. The horizontal yellow line represents the median velocity of natural Sun-2a (0.23 μm/min).

We suggest that three of the four constrained synthetic isomers, branched Sun-2b, branched Uln-3b, and macrocyclic Uln-3c isomers, can slow the migration of H1299 cancer cells, as interpreted by analyses of their MSD velocities. The macrocyclic Sun-3c isomer, and as anticipated, the more flexible acyclic variants, Sun-2d and Uln-3d, did not slow cell migration. Interestingly, the relative trend of median velocities in the sungsanpin series (2a ≳ 2b > 2c > 2d) qualitatively tracks well with the TWIMS arrival times (Figure 4A). This was also observed for the ulleungdin series (3c > 3b > 3d), where the Uln-3c and Uln-3b swaps are seen in both the TWIMS arrival time and median velocities (Figure 4B). This suggests that the more compact isomers could exhibit a higher degree of preorganization. However, more in-depth experiments would be needed to determine the extent of solution-based conformational similarity between the conserved residues for each series to corroborate this observation. Additionally, future stability experiments could decouple the lack of observed activity and the susceptibility of each isomer toward proteolytic degradation.²³ This might be especially informative for the macrocyclic Sun-2c, which was less compact despite being constrained and could make it more susceptible to degradation leading to no indication of slowed migration in H1299 cells. Taken together, we observe a complex interplay between the conserved sequence and shape using time-lapse cell imaging assays to monitor H1299 cell displacement as an improved readout for cancer cell migration inhibition.

CONCLUSIONS

The appreciation for lasso peptides as stable and highly constrained scaffolds with promising biological activities inspired our efforts to evaluate the relationship between conserved residue arrangement and activity across synthetically accessible lasso peptide isomers of sungsanpin 2a and ulleungdin 3a. We identified Asp8 as a common residue that can be strategically modified to access acyclic, branched, and macrocyclic isomers in both the sungsanpin and ulleungdin series. Access to these isomers enabled a systematic and broad understanding of the lasso peptide structure as it relates to the inhibition of cancer cell migration for these peptide sequences. Our results suggest that constrained lasso peptide isomers, Sun-2b, Uln-3b, and Uln-3c, exhibit an ability to slow H1299 cell migration comparable to that of natural sungsanpin 2a. Our results, and those reported by Zou and co-workers,²⁴ suggest that the lasso motif is not a required feature to slow cancer cell migration. These lasso peptide isomers Sun-2b, Uln-3b, and Uln-3c may serve as useful synthetically accessible lead compounds for projected modeling, target identification, and optimization studies, which could inform the future design of these constrained peptides.