Predicting clinical response to anticancer drugs using an ex vivo platform that captures tumour heterogeneity.

Predicting clinical response to anticancer drugs remains a major challenge in cancer treatment. Emerging reports indicate that the tumour microenvironment and heterogeneity can limit the predictive power of current biomarker-guided strategies for chemotherapy. Here we report the engineering of personalized tumour ecosystems that contextually conserve the tumour heterogeneity, and phenocopy the tumour microenvironment using tumour explants maintained in defined tumour grade-matched matrix support and autologous patient serum. The functional response of tumour ecosystems, engineered from 109 patients, to anticancer drugs, together with the corresponding clinical outcomes, is used to train a machine learning algorithm; the learned model is then applied to predict the clinical response in an independent validation group of 55 patients, where we achieve 100% sensitivity in predictions while keeping specificity in a desired high range. The tumour ecosystem and algorithm, together termed the CANScript technology, can emerge as a powerful platform for enabling personalized medicine.

Structural effects of hypermodified nucleosides in the Escherichia coli and human tRNALys anticodon loop: the effect of nucleosides s2U, mcm5U, mcm5s2U, mnm5s2U, t6A, and ms2t6A.

Previous nuclear magnetic resonance (NMR) studies of unmodified and pseudouridine39-modified tRNA(Lys) anticodon stem loops (ASLs) show that significant structural rearrangements must occur to attain a canonical anticodon loop conformation. The Escherichia coli tRNA(Lys) modifications mnm(5)s(2)U34 and t(6)A37 have indeed been shown to remodel the anticodon loop, although significant dynamic flexibility remains within the weakly stacked U35 and U36 anticodon residues. The present study examines the individual effects of mnm(5)s(2)U34, s(2)U34, t(6)A37, and Mg(2+) on tRNA(Lys) ASLs to decipher how the E. coli modifications accomplish the noncanonical to canonical structural transition. We also investigated the effects of the corresponding human tRNA(Lys,3) versions of the E. coli modifications, using NMR to analyze tRNA ASLs containing the nucleosides mcm(5)U34, mcm(5)s(2)U34, and ms(2)t(6)A37. The human wobble modification has a less dramatic loop remodeling effect, presumably because of the absence of a positive charge on the mcm(5) side chain. Nonspecific magnesium effects appear to play an important role in promoting anticodon stacking. Paradoxically, both t(6)A37 and ms(2)t(6)A37 actually decrease anticodon stacking compared to A37 by promoting U36 bulging. Rather than stack with U36, the t(6)A37 nucleotide in the free tRNAs is prepositioned to form a cross-strand stack with the first codon nucleotide as seen in the recent crystal structures of tRNA(Lys) ASLs bound to the 30S ribosomal subunit. Wobble modifications, t(6)A37, and magnesium each make unique contributions toward promoting canonical tRNA structure in the fundamentally dynamic tRNA(Lys)(UUU) anticodon.

An RNA complex of the HIV-1 A-loop and tRNA(Lys,3) is stabilized by nucleoside modifications.

The HIV transcription initiation complex involves a putative interaction between the primer tRNA anticodon and a conserved A-rich loop in the HIV genome. Surface plasmon resonance was used to demonstrate that the hypermodified nucleosides in the tRNA anticodon stem loop (ASL) stabilize RNA-RNA interactions in a model for the anticodon/A-loop complex. tRNA ASL hairpins with the modifications of Escherchia coli tRNALys and human tRNALys,3 each form stable complexes. Partially modified tRNA ASLs bind the A-loop hairpin with lesser affinity, and it was found that the modifications of the bacterial and mammalian tRNAs make distinct contributions toward stabilizing the RNA complex. One model for the anticodon/A-loop RNA complex that is consistent with the known modification effects on tRNA structure and function is that of complementary tRNAs, as seen for the published crystal structure of tRNAAsp.

Structural features of tRNALys favored by anticodon nuclease as inferred from reactivities of anticodon stem and loop substrate analogs.

The bacterial tRNA(Lys)-specific PrrC-anticodon nuclease efficiently cleaved an anticodon stem-loop (ASL) oligoribonucleotide containing the natural modified bases, suggesting this region harbors the specificity determinants. Assays of ASL analogs indicated that the 6-threonylcarbamoyl adenosine modification (t(6)A37) enhances the reactivity. The side chain of the modified wobble base 5-methylaminomethyl-2-thiouridine (mnm(5)s(2)U34) has a weaker positive effect depending on the context of other modifications. The s(2)U34 modification apparently has none and the pseudouridine (psi39) was inhibitory in most modification contexts. GC-rich but not IC-rich stems abolished the activity. Correlating the reported structural effects of the base modifications with their effects on anticodon nuclease activity suggests preference for substrates where the anticodon nucleotides assume a stacked A-RNA conformation and base pairing interactions in the stem are destabilized. Moreover, the proposal that PrrC residue Asp(287) contacts mnm(5)s(2)U34 was reinforced by the observations that the mammalian tRNA(Lys-3) wobble base 5-methoxycarbonyl methyl-2-thiouridine (mcm(5)s(2)U) is inhibitory and that the D287H mutant favors tRNA(Lys-3) over Escherichia coli tRNA(Lys). The detection of this mutation and ability of PrrC to cleave the isolated ASL suggest that anticodon nuclease may be used to cleave tRNA(Lys-3) primer molecules annealed to the genomic RNA template of the human immunodeficiency virus.