Nature | Vol 577 | 30 January 2020 | 685
interacts with TRAF6, we next examined whether Rv0222 is ubiqui-
tinated in host cells. We first detected the possible polyubiquitina-
tion of Rv0222 in HEK293T cells (Extended Data Fig. 3a). Seven lysine
residues in ubiquitin—K6, K11, K27, K29, K33, K48 and K63—have been
reported to form polyubiquitin chains^13 –^17. We therefore co-trans-
fected haemagglutinin (HA)-tagged wild-type ubiquitin (containing
all seven lysine residues) or sequential ubiquitin substitution mutants
(with only one of the seven lysine residues retained as lysine, and
the other six replaced with arginine)with Flag-tagged Rv0222 into
HEK293T cells. A co-immunoprecipitation assay showed that the
K11 ubiquitin mutant markedly increased the polyubiquitination
of Rv0222 (Extended Data Fig. 3b). In addition, we detected endog-
enous K11-linked polyubiquitin conjugates on Rv0222 in macrophages
infected with H37Rv(ΔRv0222 + Rv0222) (Fig. 3a and Extended Data
Fig. 3c). Thus, Rv0222 may undergo K11-linked polyubiquitination
in host cells.
We therefore hypothesized that a host factor is likely to be involved
in its polyubiquitination. ANAPC2, BIRC2 and RNF7 are three known
host E3 ligases with K11-linked polyubiquitination activity^16. Co-
immunoprecipitation showed that ANAPC2 interacted with Rv
(Fig. 3b and Extended Data Fig. 3d), and markedly increased the K11-
linked polyubiquitination of Rv0222 (Fig. 3c). We further performed
biochemical confirmation of ANAPC2-mediated ubiquitination of
Rv0222 in vitro (as described previously^13 ,^17 ), and found that K11-
linked polyubiquitination conjugates of Rv0222 were formed only
in the presence of ANAPC2 (Fig. 3d), suggesting an essential role of
this ligase in mediating K11-linked polyubiquitination of Rv
directly. It has been reported that the anaphase-promoting com-
plex/cyclosome (APC/C) assembles K11/K48-branched chains^14.
Using a K11/K48-bispecific antibody, we did not observe notable
amounts of K11/K48-branched chains on Rv0222 (Extended Data
Fig. 3c, e), suggesting that Rv0222 is mainly ubiquitinated with K11-
linked polyubiquitin.
Ubiquitination of Rv0222 inhibits immunity
To determine the location of the lysine residue on Rv0222 responsible
for the attachment of K11-linked polyubiquitin, we examined the struc-
ture of Rv0222 (Protein Data Bank code 5KJP or 6LDZ) and identified
nine lysine residues (Fig. 3e). Substitution of Rv0222 K76 with alanine
nearly abrogated K11-linked ubiquitination, suggesting that K76 is
essential for such ubiquitination (Fig. 3f). Furthermore, overexpres-
sion of ANAPC2 markedly enhanced the K11-linked ubiquitination of
Rv0222 but not of Rv0222(K76A) in HEK293T cells (Fig. 3f). In addition,
shRNA-mediated silencing of ANAPC2 markedly inhibited K11-linked
polyubiquitination of Rv0222 in HEK293T cells (Fig. 3g). Thus, host
ANAPC2 may enhance the K11-linked polyubiquitin of M. tuberculosis
Rv0222 at K76.
We next investigated the role of K11-linked Rv0222 polyubiquit-
ination in inhibition of TRAF6 signalling. In HEK293T cells, TRAF6-
induced activation of the NF-κB and AP-1 reporter genes was
markedly inhibited by Rv0222, but not by mutant Rv0222(K76A)
(Extended Data Fig. 4a, b). Moreover, Rv0222(K76A) lost its inhibi-
tory effect on K63-linked polyubiquitination of TRAF6 in HEK293T
cells (Extended Data Fig. 4c). Levels of proinflammatory cytokine
mRNAs were much lower in primary peritoneal macrophages infected
with the H37Rv(ΔRv0222 + Rv0222) strain than in those infected
with H37Rv(ΔRv0222 + GFP), but were much higher in macrophages
infected with H37Rv(ΔRv0222 + Rv0222(K76A)) than in those infected
with H37Rv(ΔRv0222 + Rv0222) (Fig. 4a–c). The cytokine levels were
not completely restored when the mutant strain was complemented
with the K76A mutant, suggesting that Rv0222 may regulate the expres-
sion of Il1b and Il6 through an additional K76-independent mechanism.
Furthermore, the expression of proinflammatory cytokines in lung
tissue from both H37Rv∆Rv0222-infected C57BL/6 and severe com-
bined immunodeficient (SCID) mice^18 was substantially higher than
in H37Rv-infected mice, confirming the inhibitory effect of Rv0222 on
cytokine expression in vivo. However, infection with H37RvΔRv
HA–K11-Ub
ANAPC
b c
e
a d
f
E1 + E
E
Rv
IB: K11-Ub K11-Ub
+
+
+
+
+
IB: Rv0222 Rv
g
Flag–Rv
IP: Flag
IB: HA
Lysate
IB: Flag
Myc–ANAPC
K11-Ub
Rv
IP: Flag
IB: Flag Rv
Lysate
IB: Myc
+
+
+
- –
IP: Flag
IB: K11-Ub
ΔRv0222 + Rv
036
IP: Flag
IB: Flag
Flag–Rv
Myc–ANAPC
IP: Flag
IB: Myc
IP: Flag
IB: Flag
Lysate
IB: Myc
++–
- ++
Lysate
IB: Flag
Flag–Rv
HA–K11-Ub
Myc–ANAPC
IB: HA
IB: Myc
IB: Flag
IB: Flag
–++
+
+–+
+
++K
–+^
K11-Ub
ANAPC
+
Rv
Rv
Time (h)
K
K
K
K
K138K
K
K
K
K
Myc–Rv
IB: HA
IB: Myc
HA–K11-Ub
shRNA
K11-Ub
Rv
IB: ANAPC2 ANAPC
IB: Myc
IB: GAPDH
Rv
GAPDH
K76 +
K
K
K
++
Scramble
–– +
++
ANAPC
––++
IP: Myc
Lysate
IP: Flag
Lysate
K
Fig. 3 | Rv0222 is K11-ubiquitinated at K76 by host ANAPC2. a, Immunoblot and
immunoprecipitation of lysates from peritoneal macrophages infected with
H37Rv(ΔRv0222 + Rv0222) for the indicated times (MOI = 5). b, Immunoblot
and immunoprecipitation of lysates of HEK293T cells transfected with
plasmids encoding Flag–Rv0222 and Myc–ANAPC2. c, Immunoblot and
immunoprecipitation of lysates of HEK293T cells transfected with plasmids
encoding HA-tagged K11-linked ubiquitin (K11-Ub), Flag-tagged Rv0222 and Myc-
tagged ANAPC2. d, The E3 ligase APC/C produces K11-linked ubiquitin chains on
Rv0222 in vitro. e, Locations of lysine residues (red) on the Rv0222 structure
(Protein Data Bank code 5KJP or 6LDZ; structure generated with PyMOL,
version 2.0). f, Immunoblot and immunoprecipitation of lysates of HEK293T
cells transfected with plasmids encoding HA-tagged K11-Ub, Myc-tagged
ANAPC2 and Flag-tagged Rv0222 or Flag-tagged Rv0222(K76A).
g, Immunoblot and immunoprecipitation of lysates of control (scramble) or
ANAPC2-knockdown HEK293T cells transfected with plasmids encoding HA-
tagged K11-Ub and Myc-tagged Rv0222 or Myc-tagged Rv0222(K76A). Data in
a–d, f, g are representative of one experiment with at least three independent
biological replicates. For gel source data, see Supplementary Fig. 1.