- Email: [email protected]
DELs Inside Cells
TIPS 1690 No. of Pages 3
Trends in Pharmacological Sciences
Spotlight of the putative protein ligands to which cells successfully [4]. A second crucial dethey were attached. sign element was to include in the construct a sulfonyl fluoride crosslinking reagent that Animesh Roy1 and 2, This screening protocol requires purified would result in covalent coupling of the * Thomas Kodadek soluble protein, which is not ideal. There is encoding DNA to the target protein if the DNA-encoded libraries (DELs) com- always the risk that this does not represent construct localizes to the target via the prise large numbers of small mole- the physiologically relevant form of the attached ligand (Figure 1). Thus, ligands cules, each of which is conjugated target. Proteins inside cells are invariably can now be mined from a CPP-conjugated to an encoding DNA. Krusemark associated with other proteins and they DEL by incubating the library with cells that can be post-translationally modified in express a target protein. The cells are then and colleagues recently described myriad ways. Moreover, some proteins lysed and the tagged target is pulled down a method to introduce DELs into are difficult to purify in large quantities or and washed under conditions that denature living cells and recover conjugates are not sufficiently soluble to serve as the protein but not the DNA (Figure 1). The that bind to an intracellular target. ideal targets in this format. Therefore, target protein-tethered encoding tags are This proof-of-principle study sug- there is interest in developing new modali- sequenced to reveal the identities of the gests that it may be feasible to ties for DEL screening. ligands [4].
DELs Inside Cells
screen DELs against protein targets An attractive idea is to screen DELs using The efficacy of this system was demonin their native environment. Most small-molecule probes and drug leads are currently discovered via some type of high-throughput screening (HTS) campaign. DEL technology has emerged as a powerful tool for this purpose [1]. Whereas the cost of HTS conducted in the traditional 384- or 1536-well plate format scales roughly with the number of compounds to be screened, truly vast DELs can be made and screened inexpensively in a single tube. Most DELs are created by solution-phase split and pool synthesis, resulting in a collection of small molecules, each of which is attached physically to a DNA tag whose sequence encodes its structure. DELs can be vast because they are theoretically limited only by the number of molecules in the test tube. Libraries from hundreds of thousands to billions of molecules have been reported. The most common format, by far, for DEL screening is to incubate the library with an immobilized protein of interest and then collect the small-molecule–DNA conjugates that coprecipitate with the protein [2]. These molecules are released, and the process is usually repeated two to three times. Finally, the encoding tags on the coprecipitated molecules are amplified using PCR and the amplicons are deep sequenced, revealing the predicted structure
intact cells so as to evaluate ligand–protein interactions in the native environment of the target. Of course, for intracellular targets this is frustrated by the fact that the DNA-encoding tag attached to each molecule will not pass through a membrane. A second issue is that small-molecule– protein target association is usually driven by using a relatively high protein concentration in a screen because the level of any one molecule in the DEL is very low, certainly well below the KD of any complex likely to be discovered in the screen. This has so far limited the live-cell screening approach to the discovery of ligands for the extracellular domain of an integral membrane receptor [3], and in this case the target was massively overexpressed.
strated in relatively simple model experiment. A small (96-member) DEL was constructed by varying four of the positions in a known peptide ligand for the chromodomain (ChD) of protein CBX7. This library (conjugated to the CPP and the crosslinker) was incubated with cells expressing high levels of a fusion protein composed of the CBX7 ChD fused to the HaloTag protein. Using the steps described above, the team successfully enriched peptides that associate with the ChD. Quantitative analysis revealed that the recovery of the encoding tag tethered to a known ChD ligand was about 0.012%, a level 20-fold above the background defined by the recovery of nonbinding small-molecule–DNA conjugates. Interestingly, when the experiment was repeated using a cell lysate instead of intact cells, a much higher enrichment of 600fold above background was observed, suggesting that only a fraction of the conjugates escape the internalized vesicle and access the cytoplasm. In a separate experiment, the authors also demonstrated successful enrichment (300-fold) of a ligand that recognizes the extracellular domain of an integral membrane protein.
Recently, Krusemark et al. reported a clever strategy to circumvent these limitations to some extent [4]. Several laboratories have demonstrated that specific peptides, known as cell-penetrating peptides (CPPs), are taken up by cells using an active transport mechanism such as endocytosis or pinocytosis, and that some of the engulfed molecules can escape the internalized organelle and access the cytosol. A cyclic peptide of this type was fused to a DNA– small-molecule conjugate. The resulting This study is significant in two ways. It construct was shown to be able to enter demonstrates that DNA-encoded small Trends in Pharmacological Sciences, Month 2020, Vol. xx, No. xx
1
Trends in Pharmacological Sciences
Trends in Pharmacological Sciences
Figure 1. Schematic Illustration of the System Developed for Screening DNA-Encoded Libraries (DELs) against Tagged Proteins in Living Cells. A library of potential ligands (blue triangle) are fused to sulfonyl fluoride and a cell-penetrating cyclic peptide (CPP) via DNA strands. The DNA contains a coding sequence for the attached small molecule. When the library is incubated with cells, the CPP mediates uptake of the construct. If the small molecule binds to the target protein, proximitydriven crosslinking takes place (red), resulting in a covalent connection between the encoding tag and the protein. Target protein-associated tags are sequenced after cell lysis and enrichment of the target protein.
molecules can access intracellular targets when tethered to a CPP. It also highlights the utility of the covalent crosslinking strategy for trapping low levels of complexes that would otherwise dissociate during the many washing and processing steps inherent in any screen. This is the first demonstration of a DEL screen, albeit a simple one, against a protein that resides inside living cells, an important milestone. By contrast, this is very much a model study. The ultimate goal is to mine DELs for ligands of a native protein. This study does not demonstrate that capability. Given the relatively modest enrichment levels observed, it may be that target protein overexpression is essential, although 2
this remains to be determined. For the same reason, the size of the library that can be screened is limited because it will be crucial to ensure that the level of enrichment of a ligand relative to the naïve library is sufficient to allow a bona fide ligand to rise above the level of noise, which is more problematic as the size of the library increases [5]. It is unlikely that billions of compounds will be screened in this format soon. In addition, the time and effort required to make CPP-conjugated DELs may limit adoption of this approach by the broader community. Indeed, it is tempting to predict that the protocol from this paper that is most likely to be adopted widely will be to screen the library in a cell
Trends in Pharmacological Sciences, Month 2020, Vol. xx, No. xx
lysate and use the crosslinking strategy to at least partially overcome the limitation of a low target protein concentration. If this could be made efficient, it would remove the requirement for target purification, a significant practical advance. Finally, a much more ambitious goal would be to employ DELs in phenotypic screens [6]. The technology does not yet enable such efforts. Although Krusemark et al. surmount one of the barriers to phenotypic screening, namely the impermeability of small-molecule–DNA conjugates, the DELs they employ are present as an intractable mixture of compounds, as are all DELs created by solution-phase split
Trends in Pharmacological Sciences
and pool synthesis. To carry out a pheno- [10]. This is likely to remain an area of high typic screen, individual compounds must interest for the foreseeable future. be segregated spatially such that each 1 Biotechnologies, Jupiter, FL 33458, USA can be exposed to cells individually. In 2Deluge Department of Chemistry, The Scripps Research Institute, traditional HTS this is accomplished by Jupiter, FL 33458, USA placing cells and one or more compounds in the wells of a microtiter plate. DELs *Correspondence: [email protected] (T. Kodadek). created by solid-phase split and pool syn- https://doi.org/10.1016/j.tips.2020.01.007 thesis [7] may be amenable to this applica© 2020 Elsevier Ltd. All rights reserved. tion because beads can be spatially segregated [8,9], but this remains to be demonstrated. References 1.
In summary, this report by Krusemark et al. joins a small, but growing, body of literature that is beginning to add to the repertoire of assays by which DELs can be screened for bioactive compounds
2.
3.
Neri, D. and Lerner, R.A. (2018) DNA-encoded chemical libraries: a selection system based on endowing organic compounds with amplifiable information. Annu. Rev. Biochem. 87, 479–502 Clark, M.A. et al. (2009) Design, synthesis and selection of DNA-encoded small-molecule libraries. Nat. Chem. Biol. 5, 647–654 Wu, Z. et al. (2015) Cell-based selection expands the utility of DNA-encoded small-molecule library technology to cell
surface drug targets: identification of novel antagonists of the NK3 tachykinin receptor. ACS Comb. Sci. 17, 722–731 4. Cai, B. et al. (2019) Selection of DNA-encoded libraries to protein targets within and on living cells. J. Am. Chem. Soc. 141, 17057–17061 5. Satz, A.L. et al. (2017) Analysis of current DNA encoded library screening data indicates higher false negative rates for numerically larger libraries. ACS Comb. Sci. 19, 234–238 6. Moffat, J.G. et al. (2017) Opportunities and challenges in phenotypic drug discovery: an industry perspective. Nat. Rev. Drug Discov. 16, 531–543 7. MacConnell, A.B. et al. (2015) DNA-encoded solid-phase synthesis: encoding language design and complex oligomer library synthesis. ACS Comb. Sci. 17, 518–534 8. Borchardt, A. et al. (1997) Small molecule-dependent genetic selection in stochastic nanodroplets as a means of detecting protein–ligand interactions on a large scale. Chem. Biol. 4, 961–968 9. Price, A.K. et al. (2016) hvSABR: photochemical dose– response bead screening. Anal. Chem. 88, 2904–2911 10. Kodadek, T. et al. (2019) Beyond protein binding: recent advances in screening DNA-encoded libraries. Chem. Commun. 55, 13330–13341
Trends in Pharmacological Sciences, Month 2020, Vol. xx, No. xx
3
Trends in Pharmacological Sciences
Spotlight of the putative protein ligands to which cells successfully [4]. A second crucial dethey were attached. sign element was to include in the construct a sulfonyl fluoride crosslinking reagent that Animesh Roy1 and 2, This screening protocol requires purified would result in covalent coupling of the * Thomas Kodadek soluble protein, which is not ideal. There is encoding DNA to the target protein if the DNA-encoded libraries (DELs) com- always the risk that this does not represent construct localizes to the target via the prise large numbers of small mole- the physiologically relevant form of the attached ligand (Figure 1). Thus, ligands cules, each of which is conjugated target. Proteins inside cells are invariably can now be mined from a CPP-conjugated to an encoding DNA. Krusemark associated with other proteins and they DEL by incubating the library with cells that can be post-translationally modified in express a target protein. The cells are then and colleagues recently described myriad ways. Moreover, some proteins lysed and the tagged target is pulled down a method to introduce DELs into are difficult to purify in large quantities or and washed under conditions that denature living cells and recover conjugates are not sufficiently soluble to serve as the protein but not the DNA (Figure 1). The that bind to an intracellular target. ideal targets in this format. Therefore, target protein-tethered encoding tags are This proof-of-principle study sug- there is interest in developing new modali- sequenced to reveal the identities of the gests that it may be feasible to ties for DEL screening. ligands [4].
DELs Inside Cells
screen DELs against protein targets An attractive idea is to screen DELs using The efficacy of this system was demonin their native environment. Most small-molecule probes and drug leads are currently discovered via some type of high-throughput screening (HTS) campaign. DEL technology has emerged as a powerful tool for this purpose [1]. Whereas the cost of HTS conducted in the traditional 384- or 1536-well plate format scales roughly with the number of compounds to be screened, truly vast DELs can be made and screened inexpensively in a single tube. Most DELs are created by solution-phase split and pool synthesis, resulting in a collection of small molecules, each of which is attached physically to a DNA tag whose sequence encodes its structure. DELs can be vast because they are theoretically limited only by the number of molecules in the test tube. Libraries from hundreds of thousands to billions of molecules have been reported. The most common format, by far, for DEL screening is to incubate the library with an immobilized protein of interest and then collect the small-molecule–DNA conjugates that coprecipitate with the protein [2]. These molecules are released, and the process is usually repeated two to three times. Finally, the encoding tags on the coprecipitated molecules are amplified using PCR and the amplicons are deep sequenced, revealing the predicted structure
intact cells so as to evaluate ligand–protein interactions in the native environment of the target. Of course, for intracellular targets this is frustrated by the fact that the DNA-encoding tag attached to each molecule will not pass through a membrane. A second issue is that small-molecule– protein target association is usually driven by using a relatively high protein concentration in a screen because the level of any one molecule in the DEL is very low, certainly well below the KD of any complex likely to be discovered in the screen. This has so far limited the live-cell screening approach to the discovery of ligands for the extracellular domain of an integral membrane receptor [3], and in this case the target was massively overexpressed.
strated in relatively simple model experiment. A small (96-member) DEL was constructed by varying four of the positions in a known peptide ligand for the chromodomain (ChD) of protein CBX7. This library (conjugated to the CPP and the crosslinker) was incubated with cells expressing high levels of a fusion protein composed of the CBX7 ChD fused to the HaloTag protein. Using the steps described above, the team successfully enriched peptides that associate with the ChD. Quantitative analysis revealed that the recovery of the encoding tag tethered to a known ChD ligand was about 0.012%, a level 20-fold above the background defined by the recovery of nonbinding small-molecule–DNA conjugates. Interestingly, when the experiment was repeated using a cell lysate instead of intact cells, a much higher enrichment of 600fold above background was observed, suggesting that only a fraction of the conjugates escape the internalized vesicle and access the cytoplasm. In a separate experiment, the authors also demonstrated successful enrichment (300-fold) of a ligand that recognizes the extracellular domain of an integral membrane protein.
Recently, Krusemark et al. reported a clever strategy to circumvent these limitations to some extent [4]. Several laboratories have demonstrated that specific peptides, known as cell-penetrating peptides (CPPs), are taken up by cells using an active transport mechanism such as endocytosis or pinocytosis, and that some of the engulfed molecules can escape the internalized organelle and access the cytosol. A cyclic peptide of this type was fused to a DNA– small-molecule conjugate. The resulting This study is significant in two ways. It construct was shown to be able to enter demonstrates that DNA-encoded small Trends in Pharmacological Sciences, Month 2020, Vol. xx, No. xx
1
Trends in Pharmacological Sciences
Trends in Pharmacological Sciences
Figure 1. Schematic Illustration of the System Developed for Screening DNA-Encoded Libraries (DELs) against Tagged Proteins in Living Cells. A library of potential ligands (blue triangle) are fused to sulfonyl fluoride and a cell-penetrating cyclic peptide (CPP) via DNA strands. The DNA contains a coding sequence for the attached small molecule. When the library is incubated with cells, the CPP mediates uptake of the construct. If the small molecule binds to the target protein, proximitydriven crosslinking takes place (red), resulting in a covalent connection between the encoding tag and the protein. Target protein-associated tags are sequenced after cell lysis and enrichment of the target protein.
molecules can access intracellular targets when tethered to a CPP. It also highlights the utility of the covalent crosslinking strategy for trapping low levels of complexes that would otherwise dissociate during the many washing and processing steps inherent in any screen. This is the first demonstration of a DEL screen, albeit a simple one, against a protein that resides inside living cells, an important milestone. By contrast, this is very much a model study. The ultimate goal is to mine DELs for ligands of a native protein. This study does not demonstrate that capability. Given the relatively modest enrichment levels observed, it may be that target protein overexpression is essential, although 2
this remains to be determined. For the same reason, the size of the library that can be screened is limited because it will be crucial to ensure that the level of enrichment of a ligand relative to the naïve library is sufficient to allow a bona fide ligand to rise above the level of noise, which is more problematic as the size of the library increases [5]. It is unlikely that billions of compounds will be screened in this format soon. In addition, the time and effort required to make CPP-conjugated DELs may limit adoption of this approach by the broader community. Indeed, it is tempting to predict that the protocol from this paper that is most likely to be adopted widely will be to screen the library in a cell
Trends in Pharmacological Sciences, Month 2020, Vol. xx, No. xx
lysate and use the crosslinking strategy to at least partially overcome the limitation of a low target protein concentration. If this could be made efficient, it would remove the requirement for target purification, a significant practical advance. Finally, a much more ambitious goal would be to employ DELs in phenotypic screens [6]. The technology does not yet enable such efforts. Although Krusemark et al. surmount one of the barriers to phenotypic screening, namely the impermeability of small-molecule–DNA conjugates, the DELs they employ are present as an intractable mixture of compounds, as are all DELs created by solution-phase split
Trends in Pharmacological Sciences
and pool synthesis. To carry out a pheno- [10]. This is likely to remain an area of high typic screen, individual compounds must interest for the foreseeable future. be segregated spatially such that each 1 Biotechnologies, Jupiter, FL 33458, USA can be exposed to cells individually. In 2Deluge Department of Chemistry, The Scripps Research Institute, traditional HTS this is accomplished by Jupiter, FL 33458, USA placing cells and one or more compounds in the wells of a microtiter plate. DELs *Correspondence: [email protected] (T. Kodadek). created by solid-phase split and pool syn- https://doi.org/10.1016/j.tips.2020.01.007 thesis [7] may be amenable to this applica© 2020 Elsevier Ltd. All rights reserved. tion because beads can be spatially segregated [8,9], but this remains to be demonstrated. References 1.
In summary, this report by Krusemark et al. joins a small, but growing, body of literature that is beginning to add to the repertoire of assays by which DELs can be screened for bioactive compounds
2.
3.
Neri, D. and Lerner, R.A. (2018) DNA-encoded chemical libraries: a selection system based on endowing organic compounds with amplifiable information. Annu. Rev. Biochem. 87, 479–502 Clark, M.A. et al. (2009) Design, synthesis and selection of DNA-encoded small-molecule libraries. Nat. Chem. Biol. 5, 647–654 Wu, Z. et al. (2015) Cell-based selection expands the utility of DNA-encoded small-molecule library technology to cell
surface drug targets: identification of novel antagonists of the NK3 tachykinin receptor. ACS Comb. Sci. 17, 722–731 4. Cai, B. et al. (2019) Selection of DNA-encoded libraries to protein targets within and on living cells. J. Am. Chem. Soc. 141, 17057–17061 5. Satz, A.L. et al. (2017) Analysis of current DNA encoded library screening data indicates higher false negative rates for numerically larger libraries. ACS Comb. Sci. 19, 234–238 6. Moffat, J.G. et al. (2017) Opportunities and challenges in phenotypic drug discovery: an industry perspective. Nat. Rev. Drug Discov. 16, 531–543 7. MacConnell, A.B. et al. (2015) DNA-encoded solid-phase synthesis: encoding language design and complex oligomer library synthesis. ACS Comb. Sci. 17, 518–534 8. Borchardt, A. et al. (1997) Small molecule-dependent genetic selection in stochastic nanodroplets as a means of detecting protein–ligand interactions on a large scale. Chem. Biol. 4, 961–968 9. Price, A.K. et al. (2016) hvSABR: photochemical dose– response bead screening. Anal. Chem. 88, 2904–2911 10. Kodadek, T. et al. (2019) Beyond protein binding: recent advances in screening DNA-encoded libraries. Chem. Commun. 55, 13330–13341
Trends in Pharmacological Sciences, Month 2020, Vol. xx, No. xx
3