Transcript tpj12346-sup-0001-FigureS1-S7

S1, Expression of the constructs that were used for the functional mapping
of the N- and C-terminal domains of UPF1
C1P1-3A
C1P1-2AP4A
12
C1P1-4A
10 11
C1P3-4A
9
C1P1-2A
NΔNcP1-2A
8
ΔNU1C4
NΔNcP3A
7
ΔNU1C3
NΔNc
6
ΔNU1C2
NΔNn
5
ΔNU1C1
NP1-4A
4
NΔP1-2P3D
NP1-2A4A
3
NΔP1-2P3A
NP3A
2
C-terminal domain mapping constructs
NΔP1-2
U1ΔC
1
NΔNcP1-3A
ΔNΔC
N-terminal domain mapping constructs
13
14
15
16
17
18
19
20
21
22
Anti-HA
Anti-P14
Figure S1, Expression of the constructs that were used for the functional mapping of the N- and C-terminal domains of UPF1.
The constructs were co-agroinfiltrated with P14 into wild-type N. benthamiana leaves, and then the expression of the mutant proteins was
studied at 3 d.p.i by western-blot assays. The constructs were HA-tagged, thus HA antibodies could be used to detect the expression of the
test constructs. P14 was used as a control. Note that the N-terminal domain mapping constructs (lanes 1-13) accumulate to comparable
levels except the NΔP-1-2P3A protein (red letters), which was barely detectable. The C-terminal domain mapping constructs (lanes 14-22)
also accumulate to comparable levels.
S2, Sequence alignment of the N-terminal domains of UPF1
Arabidopsis Nn-region (1-35 a.a.)
Arabidopsis Nc-region (36-109 a.a.)
Figure S2, Sequence alignment of the N-terminal domains of UPF1.
Multiple sequence alignment of the N-terminal domains of Arabidopsis thaliana, Vitis vinifera, Oryza sativa and Homo sapiens
UPF1 homologs. Residues conserved in all homologs are highlighted in black and those, which are conserved only in a subset of
UPF1 proteins are in grey. The alignment was built using T-Coffee and presented using GenDoc.
Red squares mark the S/TQ sites that according to the mutational studies play a role in plant NMD. Grey or black arrows mark the
phosphorylation sites of Arabidopsis UPF1 that were identified in MS analysis. Asterisks indicate the serine tract where single and
double phosphorylations were detected. The black arrow marks the conserved phosphorylated residue T29. In mammals, the
corresponding threonine (T28) is phosphorylated by SMG1 and the phosphorylated threonine is bound by SMG6.
The NMD relevant Nn-region of Arabidopsis UPF1 is marked with a red line, while the Nc-region, which does not play a role in
NMD, is marked with grey line. Note that the Nn-region is highly conserved within Angiosperm and show similarity between plants
and humans, while the Nc-region is divergent even within plants.
Identified
phosphorylated
residues
U1ΔC
Trypsin
digestion
NP1-4A
Trypsin +
pepsin digestion
Trypsin
digestion
S3
1
S3A
S13
95
S13A
T29
1
T29A
S42
5
5
5
T44
3
2
3
S54
1
1
S72
2
2
4
S75-S79
120
6
140
S98
2
S103 - S105
1
S223
1
S268
1
S279
1
T292
1
T297
1
S610-S612
20
S848
1
S105A
1
1
2
10
29
1
Figure S3, Identification of the phosphorylated amino acids of the N-terminal domain of UPF1.
To identify the phosphorylated residues in the N-terminal domain of Arabidopsis UPF1, U1ΔC mutant protein was expressed in N. benthamiana leaves, and the proteins were
immunoprecipitated with HA antibody and separated by SDS/PAGE. Coomassie-Blue-stained bands were excised, in-gel digested with trypsin or with trypsin and pepsin.
The phosphopeptides were enriched on titanium oxide resin and analyzed by LC-MS/MS. Sequence searches were carried out by MASCOT. As a negative control, NP1-4A
mutant, in which all four S/TQ sites were mutated to alanine (red letters), was expressed and similarly analyzed except that only trypsin digested NP1-4A sample was studied.
The number of identified residue corresponds to the protein sequence of UPF1 (AT5G47010). S/TQ potential PIKK kinase target sites are in bold. The numbers in the
columns indicate the numbers of identified phosphopeptides. The abundance of the phosphopeptides can only be compared within an individual experiment.
The first trypsin cleavage site is located at amino acid 62. As a 62 a.a. long fragment is too long for phosphorylation studies, the phosphorylated residues of the Nn region (135 a.a.) can only be identified in the trypsin and pepsine double digested sample.
Note that the Nc region (36-109 a.a.) of the N-terminal domain of UPF1 is heavily phosphorylated, while the CH and helicase domains (109-1011) are barely phosphorylated.
Relevantly, all three S/TQ sites (S3, S13 and T29) of the Nn region (1-35 a.a.) that were functionally important for NMD activity can be phosphorylated.
S4, The 1076 serine of UPF1 plays a role in NMD
a
b
S1013 T1056 S1076 S1085
ΔNU1C1
λN HA CH Helicase
Gc-I + P14
C-term.region1
C1P1-2AP4A
A1013 A1056 A1076 S1085
C1P1-3A
VIGS: U1
λN HA CH Helicase
C-term.region1
A1013 A1056 S1076 A1085
C1P1-2AP4A
ΔNΔC
Gc-I
C1P1-3A
λN HA CH Helicase
-
C-term.region1
λN HA CH Helicase
200nt
GFP
AAA
ΔNU1C1
Ls intron
c
Gc-I + P14
VIGS: U1
Lane
1
ΔNΔC
ΔNΔC
2
ΔNU1C1 C1P1-3A C1P1-2AP4A
3
4
5
Gc-I
P14
1
±0,07
0,81
±0,33
0,35
±0,04
0,7
±0,17
0,54
±0,16
Figure S4, The 1076 serine of UPF1 plays a role in NMD.
(a) Schematic, non-proportional representation of the C1-region mutant constructs that were used in this experiment. To generate C1P13A and C1P1-2AP4A constructs the C1P1 (S1013), C1P2 (T1056) and C1P3 (S1076) or the C1P1, C1P2 and C1P4 (S1085) S/TQ sites of
the ΔNU1C1construct were mutated to alanine. (b-c) NMD competency tests of the C1-region UPF1 mutant constructs. UPF1-silenced
leaves (VIGS:U1) were infiltrated with Gc-I NMD reporter + P14 and with C1P1-3A or C1P1-2AP4A test constructs. As controls,
ΔNU1C1 positive and ΔNΔC negative control constructs were also co-infiltrated with Gc-I+P14. Note that the C1P1-2AP4A construct
that retained the S1076 site could complement the NMD deficiency of UPF1-silenced leaves more efficiently than the C1P1-3A construct,
which retained the S1085 site. These data suggest that the S1076 potential PIKK kinase target site is more important for NMD function
than the S1085 site.
S5, Sequence alignment of the C1-regions of UPF1.
Figure S5, Sequence alignment of the C1-regions of UPF1.
Multiple sequence alignment of the C1-regions of the C-terminal domains of Arabidopsis thaliana, Vitis vinifera ,Oryza sativa and
Homo sapiens UPF1 homologs. Residues conserved in all homologs are highlighted in black and those, which are conserved only in a
subset of UPF1 proteins are in grey. The alignment was built using T-Coffee and presented using GenDoc.
Red square marks the 1076 S/TQ site that according to the NMD competency studies plays an important role in plant NMD. Green
square marks the 1085 S/TQ site that might play a minor role in NMD. Asterisks indicate the phosphorylation sites of Arabidopsis
UPF1 that were suggested by preliminary MS analysis. Note that the C1-region is highly conserved within plants, while the H.
sapiens UPF1 region cannot be aligned except the two SQ sites and the neighboring amino acids.
S6, Phosphomimetic of T29 does not function in NMD
S3 S13 T29
U1ΔC
λN HA
T29
NΔP1-2
λN HA
NΔP1-2P3A
λN HA
NΔP1-2P3D
λN HA
λN HA
36
NΔP1-2P3D
CH Helicase
S105
NΔP1-2
CH Helicase
NΔP1-2
CH Helicase
Lane
e
NΔP1-2
1
Gc-I + P14
Gc-I + P14
NΔNn NΔP1-2P3A NΔP1-2P3D
2
3
4
S105
Nc-region
CH Helicase
109
NΔNn
NΔNn
c
VIGS: U1
S105
Nc-region
26 35
NΔP1-2P3D
S105
Nc-region
26 35
D29
NΔP1-2P3A
Nc region (36-109 a.a.)
Nc-region
26 35
A29
NΔP1-2P3A
CH Helicase
Gc-I + P14
d
S105
Nn-r. N-term. Nc-r.
Nn (1-35)
NΔNn
VIGS: U1
Gc-I + P14
b
a
1
±0,33
6,05
±1,1
5,31
±1,07
6,33
±0,61
Lane
ΔNΔP1-2
1
NΔNn NΔP1-2P3A NΔP1-2P3D
2
3
4
Gc-I
Gc-I
P14
P14
1
±0,14
3, 54
±0,47
3,29
±0,68
3,43
±0,57
Figure S6, Phosphomimetic of T29 does not function in NMD.
(a) Schematic, non-proportional representation of the N-terminal mutant constructs that were used in this experiment. The extreme Nterminal segment (1-26 a.a.) that contains the P1 and P2 potential PIKK target sites (S3 and S13) was deleted from the Nn region of
U1ΔC to generate NΔP1-2 construct. T29 of the NΔP1-2 construct was changed to alanine or to the phosphomimic aspartate (NΔP12P3A and NΔP1-2P3D). NΔNn construct was generated form U1ΔC by deleting the whole Nn-region. (b-c) NMD competency tests of
the N-terminal UPF1 deletion mutant constructs. UPF1-silenced leaves (VIGS:U1) were infiltrated with Gc-I NMD reporter + P14 and
with one of the N-terminal mutants. Note that mutation of T29 of the NΔP1-2 construct to either alanine or aspartate impaired the NMD
fucntion. (d-e) The UPF1 mutants that were NMD incompetent in the NMD competency assay act in a dominant-negative manner. Gc-I +
P14 were co-infiltrated into wild-type leaves with one of the UPF1 N-terminal mutants. Note that although NΔP1-2P3A protein
accumulated to barely detectable levels (S2) it acted as a dominant-negative mutant.
S7, SMG7 and UPF1 co–localize with the P-body marker DCP1
a
CFP
S7-C
S3 S13 T29
U1-Y
D1-R
λN HA
N-term.
N-term.
S105
CH Helicase
C-term.
YFP
RFP
b
c
Figure S7, SMG7 and UPF1 co–localize with the P-body marker DCP1.
(a) Schematic, non-proportional representation of the used constructs. Co-localization of (b) SMG7-CFP (S7-C, left, cyan,) and (c) UPF1YFP (U1-Y, left, yellow) with (b,c) DCP1-RFP (D1-R, middle, red) in Arabidopsis mesophyll protoplasts. Right panel shows an overlay
of both signals and fluorescence of chloroplasts.