3.5 | Intrinsically disordered microproteins
Microproteins are much shorter than annotated proteins, and they tend to
exhibit limited conservation to protein domains of known function. As a
result, it is challenging to perform bioinformatic analyses, for example
of predicted structure or intrinsic disorder, of microproteins with
confidence, particularly because many of these predictive algorithms
rely, at least in part, on homology to structures of known, larger
proteins on which they are trained. Nonetheless, some studies have
suggested that microproteins may be enriched in intrinsic disorder
relative to canonical proteins (though an alternative analysis suggests
that evolutionarily young microproteins are de-enriched in intrinsic
disorder), which, if true, suggests that some microproteins could carry
out cellular functions associated with intrinsically disordered
proteins, such as regulating signaling and other processes by binding to
protein partners via short linear interaction motifs (SLIMs). In this
section we discuss two human microproteins that have been experimentally
confirmed to be predominantly intrinsically disordered.
MRI (Modulator of retroviral infection) was first identified in a cDNA
library screen for host proteins that could complement resistance to
retroviral infection of human cells, but it remained annotated as a
predicted or uncharacterized protein-coding gene (C7ORF49 ) in the
early 2010s. While the long isoform of MRI (MRI-1 hereafter) is 157
amino acids long and therefore not a microprotein, a 2013 peptidomics
study identified an unannotated, sORF encoded isoform (MRI-2) of 69
amino acids. Follow-up work demonstrated that the long MRI-1 and short
MRI-2 proteins could interact with a complex of proteins essential for
the non-homologous end joining pathway (NHEJ), which is essential for
repairing DNA double strand breaks in G1 phase of the cell cycle, as
well as for B and T cell receptor gene diversification via V(D)J
recombination. Specifically, MRI-1 interacts with the double-strand
break binding adaptor proteins Ku70/80 (Ku) and DNA-PKcs (DNA-dependent
protein kinase catalytic subunit), while MRI-2 binds to Ku. Both of
these MRI isoforms contain an N-terminal Ku-binding motif, explaining
their association with Ku, while MRI-1 also contains a C-terminal
XLF-like motif (XLM) that associates with additional, distinct NHEJ
factors. The XLM of MRI-1 is absent in the frameshifted, truncated MRI-2
isoform. One study suggests that MRI inhibits aberrant NHEJ at telomeres
during S phase, while two studies to date are consistent with a positive
role for MRI in NHEJ during most phases of the cell cycle, suggesting
that the activity of MRI may be context-dependent. Purified MRI-2 was
shown to promote NHEJ in vitro. However, abrogating all isoforms via
knockout of the MRI gene in vivo and in pre-B cells increases
sensitivity to ionizing radiation and inhibits NHEJ when coupled with
knockout of the NHEJ “sentinel” gene XLF. Purified MRI-1 was shown to
be predominantly intrinsically disordered via hydrogen-deuterium
exchange; while MRI-2 was not directly investigated in this study, it is
likely to have a similar degree of intrinsic disorder because these
proteins share substantial sequence identity until the frameshift that
truncates MRI-2. Interestingly, the N-terminal and C-terminal motifs of
MRI-1 alone can nucleate separate complexes of NHEJ factors, and MRI-1
can recruit NHEJ factors to chromatin in the presence of DNA double
strand breaks. It is interesting to speculate MRI-2 may therefore be
able to serve the same nucleating function in NHEJ via its Ku-binding
motif even in the absence of the C-terminal XLM. Sleckman and colleagues
proposed that MRI-1 serves as an adaptor protein for NHEJ, promoting
stable association of active NHEJ complexes at sites of double strand
breaks as a result of its (1) intrinsic disorder, (2) independent linear
interaction motifs, and (3) its potential to multimerize. While better
understanding of the contributions of individual MRI isoforms to their
function in vivo is required, MRI-1 and MRI-2 appear to be paradigmatic
examples of intrinsically disordered (micro)proteins that promote
assembly of a functional protein interaction network.
Another example of an experimentally validated, intrinsically disordered
microprotein is NBDY. NBDY is a 68-amino acid microprotein expressed
from a previously misannotated lncRNA (LOC550643 ). NBDY
associates with members of the cytoplasmic mRNA decapping complex. The
interaction partners of NBDY, EDC4 and DCP1A, are coactivators required
for allosteric activation of DCP2, which catalyzes the first step in
5′-to-3′ mRNA decay (removal of the 7-methylguanosine cap), thus
regulating the stability of thousands of specific mRNA substrates .
Genetic ablation or silencing of NBDY stabilizes a majority of DCP2
substrates, consistent with the requirement of NBDY for their effective
decapping, including transcripts encoding proteins involved in immune
responses – a pathway previously reported to be regulated by DCP2.
However, at the same time, a number of DCP2 substrates are destabilized
by NBDY ablation, suggesting that the microprotein may act as a
specificity factor for recruitment of mRNA targets to the decapping
complex. In particular, in the presence of NBDY, DCP2 substrate mRNAs
with shorter 5′UTRs decay more rapidly, suggesting that there may be a
requirement for NBDY for efficient recognition of transcripts with short
leader sequences by DCP2. While the molecular mechanism by which NBDY
regulates the mRNA decapping complex is not yet known, mRNA decapping
proteins have previously been reported to associate via SLIMs within
disordered regions, and it is likely that NBDY participates in this
network. NMR experiments indicated that NBDY is largely intrinsically
disordered in solution, consistent with its ability to phase-separate in
the presence of RNA to form liquid droplets in vitro. Within the
intrinsically disordered NBDY sequence, two independent SLIMs interact
with the WD40 domain of EDC4 and the EVH1 domain of DCP1A. The
interaction between EDC4 and NBDY appears to be more important for NBDY
function in mutagenesis experiments, but, given the relatively low
affinity of NBDY for EDC4 (KD ~ 1
micromolar), the interaction with DCP1A could speculatively be important
for increasing avidity of NBDY for the mRNA decapping complex, retaining
it at interaction sites. Importantly, NBDY also partially localizes to
and regulates phase-separated RNA granules termed P-bodies in cells,
consistent with a role for intrinsically disordered microproteins in
biological phase separation. NBDY is phosphorylated downstream of EGFR
and cyclin-dependent kinase signaling, and this phosphorylation is
required for dissociation of P-bodies – likely via electrostatic
repulsion of negatively charged P-body components that promotes
liquid-phase remixing and cell proliferation. Taken together, NBDY’s
intrinsic disorder enables its SLIM-mediated protein-protein
interactions, phase separation and regulation of P-bodies, providing a
well-defined example of the functional significance of intrinsic
disorder in a microprotein.