€
C. Nocker, N. Kaiser, D. Foley et al.
Tetrahedron 87 (2021) 132118
Table 2
linezolid, as well as the selective mitochondrial ribosome inhibitor
chloramphenicol, did not inhibit autophagy in our hands (data not
shown).
List of ribosomal proteins identified as hits in a TPP experiment performed with
Autoxain, including gene name, protein name and
independent experiments.
D
Tm [ꢁC]. Data is mean of three
It must be noted that cycloheximide, which inhibits translation
by targeting the ribosome, was previously reported to inhibit
autophagy by activating mTOR [20], but we were unable to
corroborate this in our studies. To shed light on this, we compared
compound bioactivity in a more general sense using the cell
painting assay (CPA), which monitors morphological changes upon
compound treatment (Fig. 2a) [21]. A significant change compared
to DMSO is annotated by an induction value (see supporting infor-
mation for a detailed description) [7,22]. This enables the generation
of fingerprints which can be compared between compounds to
produce a biological similarity score and identify common modes-
of-action. Cycloheximide, as well as two other translation inhibitors
emetine and anisomycin, displayed clear morphological changes
compared to the DMSO control, and had a high degree of bio-
similarity between each other (Fig. 2b). This is consistent with
previous reports that protein synthesis inhibitors display compa-
rable fingerprints in the CPA [23]. However, Autoxain and inhibitors
of mitochondrial ribosome activity linezolid and chloramphenicol
did not induce morphological changes, which did not enable their
biosimilarity to be calculated (Fig. 2c). As a result, further work is
needed to comprehensively characterize the link between auto-
phagy and the proper function of the cytosolic and mitochondrial
ribosome. To more comprehensively map autoxain’s target profile,
immobilizing the molecule on a solid phase and carrying out a
proteome-wide pull-down experiment may offer new insights.
Based on the existing SAR, the oxazolidinone ring appears to offer
the most viable position for linker functionalization.
Entry
Gene name
Protein name
D
Tm [ꢁC]
1
2
3
4
5
6
7
8
RPL3
RPL4
RPLS9
60S ribosomal protein L3
60S ribosomal protein L4
40S ribosomal protein S9
60S ribosomal protein L10
40S ribosomal protein S13
60S ribosomal protein L17
40S ribosomal protein S18
60S ribosomal protein L18a
60S ribosomal protein L19
60S ribosomal protein L21
60S ribosomal protein L35
60S ribosomal protein L35a
60S ribosomal protein L36
40S ribosomal protein S3a
40S ribosomal protein S6
40S ribosomal protein S14
ꢂ4.2
ꢂ4.3
ꢂ3.6
ꢂ2.7
ꢂ3.6
ꢂ3.2
ꢂ3.4
ꢂ4.4
ꢂ3.8
ꢂ4.3
ꢂ3.1
ꢂ3.8
ꢂ4.3
ꢂ3.4
ꢂ3.5
ꢂ3.6
ꢂ4.6
ꢂ3.0
ꢂ2.5
ꢂ4.4
ꢂ2.3
ꢂ2.9
ꢂ3.1
RPL10
RPS13
RPL17
RPS18
RPL18A
RPL19
RPL21
RPL35
RPL35A
RPL36
RPS3A
RPS6
RPS14
RPS24
MRPS10
MRPS34
MRPS35
MRPS22
MRPL41
MRPL48
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
40S ribosomal protein S24
28S ribosomal protein S10, mitochondrial
28S ribosomal protein S34, mitochondrial
28S ribosomal protein S35, mitochondrial
28S ribosomal protein S22, mitochondrial
39S ribosomal protein L41, mitochondrial
39S ribosomal protein L48, mitochondrial
experiment [24]. Possible hits were identified according to the
following criteria: (1) A protein is consistently stabilized or desta-
bilized in all three replicates, and; (2) The intensity of the TMT label
at the highest temperature must be 40% or less of that at the highest
temperature, and; (3) The shift in Tm (D
Tm) had to be > 2.0 ꢁC in
all three replicates. In total 63 proteins were identified as hits (see
Supplementary Information Table 2 for full list). Strikingly, 23 of
these were ribosomal proteins (Table 2).
3. Conclusion
In summary, we have identified a new class of oxazolidinone
autophagy inhibitors using phenotypic screening. The structural
features responsible for autophagy activity were explored, and
target identification experiments were carried out using thermal
proteome profiling. This resulted in the identification of the ribo-
some and the mitochondrial ribosome as likely targets for 1, which
were significantly destabilized by treatment with 1. This work
highlights the power of thermal proteome profiling towards iden-
tifying the molecular mode-of-action of a bioactive compound.
Crucially, TPP does not require pre-functionalization of an active
compound, significantly reducing the time between hit identifica-
tion and target hypothesis. Furthermore, destabilization of an
entire protein complex, in this case the ribosome, can efficiently be
identified by this technique. Therefore, we anticipate that TPP will
be widely employed for studying the interaction of small molecules
with their target proteins in the cellular environment.
Oxazolidinones are a known chemical class of antibiotics that
inhibit bacterial protein biosynthesis by targeting the prokaryotic
ribosome selectively and include the clinically approved com-
pounds linezolid and tedizolid. Linezolid binds to the peptidyl-
transferase center of the prokaryotic ribosome, where it blocks
the A site pocket and thus interferes with protein synthesis [17].
Although it does not interfere with eukaryotic protein biosynthesis,
linezolid causes serious side effects during long-term treatment,
which was proposed to occur due to inhibition of the mitochondrial
ribosome [18,19]. In the TPP, compound 1 destabilized various
cytosolic and mitochondrial ribosomal protein subunits. This sug-
gests that in analogy to the oxazolidinone antibiotics, 1 targets the
ribosome. In contrast to linezolid, it may not differentiate between
cytosolic and mitochondrial ribosomes, which may be required for
its autophagy inhibitory activity. This is supported by the fact that
5