new dissecting kmt2d missense mutations in kabuki syndrome … · 2018. 9. 14. · introduction...
TRANSCRIPT
O R I G I N A L A R T I C L E
Dissecting KMT2D missense mutations in Kabuki
syndrome patientsDario Cocciadiferro1,2,†, Bartolomeo Augello1, Pasquelena De Nittis3,Jiyuan Zhang4, Barbara Mandriani5, Natascia Malerba1,2, Gabriella M. Squeo1,Alessandro Romano6, Barbara Piccinni7, Tiziano Verri7, Lucia Micale1,Laura Pasqualucci4 and Giuseppe Merla1,*1Division of Medical Genetics, IRCCS Casa Sollievo della Sofferenza Hospital, San Giovanni Rotondo, Italy,2PhD Program in Experimental and Regenerative Medicine, Faculty of Medicine, University of Foggia, Italy,3Center for Integrative Genomics, University of Lausanne, CH-1015 Lausanne, Switzerland, 4Department ofPathology and Cell Biology, Institute for Cancer Genetics, Columbia University, New York, NY, USA,5Telethon Institute of Genetics and Medicine, TIGEM, Pozzuoli, Naples, Italy, 6Institute of ExperimentalNeurology, Division of Neuroscience, San Raffaele Scientific Institute, Milan, Italy and 7Department ofBiological and Environmental Sciences and Technologies, University of Salento, Lecce, Italy
*To whom correspondence should be addressed at: Division of Medical Genetics, IRCCS Casa Sollievo della Sofferenza, viale Cappuccini, 71013, SanGiovanni Rotondo, Foggia, Italy. Tel: þ39 0882416350; Fax: þ39 0882411616; Email: [email protected]
AbstractKabuki syndrome is a rare autosomal dominant condition characterized by facial features, various organs malformations,postnatal growth deficiency and intellectual disability. The discovery of frequent germline mutations in the histonemethyltransferase KMT2D and the demethylase KDM6A revealed a causative role for histone modifiers in this disease.However, the role of missense mutations has remained unexplored. Here, we expanded the mutation spectrum of KMT2Dand KDM6A in KS by identifying 37 new KMT2D sequence variants. Moreover, we functionally dissected 14 KMT2D missensevariants, by investigating their impact on the protein enzymatic activity and the binding to members of the WRAD complex.We demonstrate impaired H3K4 methyltransferase activity in 9 of the 14 mutant alleles and show that this reduced activity isdue in part to disruption of protein complex formation. These findings have relevant implications for diagnostic andcounseling purposes in this disease.
IntroductionKabuki syndrome (KS, MIM #147920, MIM #300867) is a rare au-tosomal dominant condition characterized by striking facialfeatures such as elongated palpebral fissures with eversion ofthe lateral third of the lower eyelid, short columella with
depressed nasal tip, skeletal anomalies, dermatoglyphic abnor-malities, mild to moderate intellectual disability and postnatalgrowth deficiency (1,2). Additional findings include congenitalheart defects, genitourinary anomalies, cleft lip and/or palate,susceptibility to infections, gastrointestinal abnormalities,
†
Present address: Laboratory of Medical Genetics, Bambino Gesu Children’s Hospital, IRCCS, Rome, Italy.Received: April 24, 2018. Revised: May 30, 2018. Accepted: June 21, 2018
VC The Author(s) 2018. Published by Oxford University Press. All rights reserved.For permissions, please email: [email protected]
1
Human Molecular Genetics, 2018, Vol. 0, No. 0 1–18
doi: 10.1093/hmg/ddy241Advance Access Publication Date: 22 June 2018Original Article
Downloaded from https://academic.oup.com/hmg/advance-article-abstract/doi/10.1093/hmg/ddy241/5042898by Ospedale Casa Sollievo della Sofferenza useron 27 August 2018
ophthalmologic defects, ptosis and strabismus, dental anoma-lies (including widely spaced teeth and hypodontia), ear pits,visceral abnormalities and premature thelarche (3). In 2010,whole-exome sequencing successfully identified heterozygousloss of function mutations in the KMT2D gene (MIM #602113,NM_003482.3, also known as MLL2 and MLL4) as the major causeof KS (4). KMT2D encodes a conserved member of the SET1 fam-ily of histone lysine methyltransferases (KMTs), which catalyzesthe methylation of lysine 4 on histone H3 (H3K4), a modificationassociated with active transcription (5–7). The enzymatic func-tion of KMT2D depends on a cluster of conserved C-terminaldomains, including plant homeodomains (PHD), two phenylala-nine and tyrosine (FY)-rich motifs [FY-rich, N-terminal (FYRN)and FY-rich, C-terminal (FYRC)] and a catalytic Su(var)3–9,Enhancer-of-zeste, Trithorax (SET) domain. In mammaliancells, KMT2D functions as a major histone H3K4 mono-/di-/tri-methyltransferase (8–15) that is required for the epigenetic con-trol of active chromatin states in a tissue-specific manner (16).
In addition to KMT2D mutations, a subset of KS individualshas been identified with either point mutations or microdele-tions encompassing the X-linked gene, KDM6A (MIM #300128,NM_021140.3, also known as UTX) (17–19), which encodes for aHistone H3 lysine-27 demethylase. KDM6A plays a crucial rolein chromatin remodeling (20,21) and interacts with KMT2D in aconserved SET1-like complex (16).
More than 600 KMT2D mutations have been identified so farin KS patients; roughly 84% of them are truncating events (22)including non-sense, indels, small duplications, splice-site, andframeshift mutations, while the remaining 16% are representedby missense variants whose pathogenicity has not been investi-gated. As a consequence, the interpretation of missense muta-tions has remained a challenging problem in genetic diagnosisand counseling.
Here we surveyed a cohort of 505 KS patients to expand themutation spectrum of KMT2D and KDM6A, and investigated thefunctional impact of missense mutations in the pathogenesis ofthe disease.
ResultsMutation screening of KMT2D and KDM6A
To expand the spectrum of mutations targeting KMT2D andKDM6A in KS, we integrated our mutation database with muta-tional screening of 202 newly diagnosed KS patients for a totalof 505 cases, by using Sanger sequencing and/or multiplexligation-dependent probe amplification (MLPA) of the KMT2Dcoding sequence, followed by KDM6A analysis in patientsresulted as KMT2D-negative. We identified a total of 208 KMT2Dvariants distributed in 196/505 (39%) patients, including 37 thathave not been described before (Table 1). These included 54non-sense mutations (26%), 59 frameshift mutations (28%), 69missense mutations (33%), 13 splice site variants (6%), 12 indels(6%) and 1 gross deletion (Table 1). Missense mutations weredistributed across the entire length of the KMT2D gene (Fig. 1A)and were represented by 8 missense variants (11%) localizedwithin the PHD 1–6 domains, one change (1%) in the Coiled Coil/Poly Q region, 25 variants (36%) localized within the C-terminalof the protein (amino acid 4507–5537) and 36 (52%) variants lo-calized outside of known domains and/or in uncharacterizedportions of the protein. Among the KMT2D variants identified,66 occurred de novo and 32 were inherited from an apparentlyasymptomatic parent, whereas for the remaining 110 variantswe had no access to parental DNA.
Moreover, we identified 14 KDM6A variants; 12 of those werepredicted to be pathogenic based on publicly available algo-rithms (see Materials and Methods) and 7 were never describedpreviously. Seven of the variants were de novo, while for theremaining ones the inheritance was unknown. Thus, 210/505(41%) KS patients in our cohort carried genetic alterations in oneof these two genes; the underlying event in the remaining 295cases remains unknown (see Discussion).
Pathogenic assessment of missense mutations bybioinformatics tools
Analogous to previous studies, most KMT2D variants in KS areinactivating truncating events that render the protein function-ally defective due to the loss of the catalytic SET activity.However, �30% (69/208) of the mutations found in our data set,and 100 of 621 (16.1%) mutations from a recently publishedmutational analysis review (22), are in-frame amino acidchanges that affect various residues along the KMT2D protein.To begin to elucidate the functional consequences of KMT2Dmissense mutations in KS, we first applied published bioinfor-matics tools to the 58 unique missense variants identified inthe 69 KS patients (see Materials and Methods). These predic-tion algorithms classified 16/58 variants (identified in 20/69patients) as pathogenic or likely pathogenic, while 24/58 var-iants (30/69 patients) were scored as likely benign, and 18/58(19/69 patients) as variants of uncertain significance (VOUS,Table 2).
To further assess how KMT2D missense variants may affectprotein function/activity, a combination of structure-basedmethods was employed. Since KMT2D is a multi-domain pro-tein too large to be accurately modeled using computationalmethods, we aimed to predict and analyze the individual struc-tures of single KTM2D domains. Three-dimensional modelswere obtained for the ZF-7, FYR and SET domains (QMEANscores of 0.63, 0.68 and 0.73, respectively) of the human KMT2D(see Materials and Methods for details). Moreover, a crystalstructure at 1.4 A resolution of the WIN domain was retrievedfrom the Protein Data Bank. Eight missense variants were ana-lyzed to test their effects on protein stability (DDG) and changein total charge (DCharge). The analysis showed that all eightvariants target highly conserved regions/residues within thesedomains (Supplementary Material, Fig. S1) and are expectedto significantly alter their structure (Supplementary Material,Fig. S2). In particular, with one exception (p.H5059P, which fallson the N-terminal residue of the predicted ZF-PHD7 and can bepoorly analyzed), all variants were predicted to alter the totalcharge of the domain and/or the protein stability(Supplementary Material, Fig. S3). These data suggest that themissense variants located in the ZF-PHD7, FYR, WIN and SETdomains affect the normal structure of the KMT2D protein andmay thus potentially alter/impair the protein function.
Missense variants impair KMT2D methyltransferaseactivity
To experimentally test the functional impact of KS-associatedKMT2D missense variants, we generated FLAG-tagged versionsof 14 representative KMT2D mutant alleles that harbor aminoacid changes in both functional N-terminal and C-terminaldomains of the protein including PHD 4–5–6–7, FYRN, WIN andSET domains, using the FUSION–KMT2D construct as template(see Materials and Methods) (Fig. 1B).
2 | Human Molecular Genetics, 2018, Vol. 00, No. 00
Downloaded from https://academic.oup.com/hmg/advance-article-abstract/doi/10.1093/hmg/ddy241/5042898by Ospedale Casa Sollievo della Sofferenza useron 27 August 2018
Tab
le1.
KM
T2D
and
KD
M6A
vari
ants
iden
tifi
edin
ou
rco
ho
rt
IDIn
her
itan
ceEx
on
/in
tro
nV
aria
nt
AA
chan
geR
efer
ence
AC
MG
clas
sifi
cati
on
KM
T2D
Non
-sen
seK
B49
NA
ex5
c.66
9T>
Gp
.(Tyr
223*
)(4
6)P
KB
343
NA
ex8
c.10
16G>
Ap
.(Trp
339*
)T
his
stu
dy
PK
B35
NA
ex10
c.19
21G>
Tp
.(Glu
641*
)(4
6)P
KB
33N
Aex
16c.
4419
G>
Ap
.(Trp
1473
*)(4
6)P
KB
63N
Aex
19c.
4895
del
Cp
.(Ser
1632
*)(4
6)P
KB
317
NA
ex22
c.52
12G>
Tp
.(Glu
1738
*)(1
9)P
KB
336
De
novo
ex22
c.52
69C>
Tp
.(Arg
1757
*)(1
8)P
KB
262
NA
ex26
c.56
74C>
Tp
.(Gln
1892
*)(1
9)P
KB
429
NA
ex26
c.57
07C>
Tp
.(Arg
1903
*)(1
8,22
,73,
74)
PK
B26
NA
ex31
c.62
95C>
Tp
.(Arg
2099
*)(4
,22,
46)
PK
B50
2D
eno
voex
31c.
7228
C>
Tp
.(Arg
2410
*)(9
,73,
75,7
6)P
KB
66N
Aex
31c.
7246
C>
Tp
.(Gln
2416
*)(4
6)P
KB
59N
Aex
31c.
7903
C>
Tp
.(Arg
2635
*)(2
2,46
)P
KB
153
De
novo
ex31
c.79
03C>
Tp
.(Arg
2635
*)(2
2,46
)P
KB
226
De
novo
ex31
c.79
03C>
Tp
.(Arg
2635
*)(2
2,46
)P
KB
338
De
novo
ex31
c.79
33C>
Tp
.(Arg
2645
*)(7
7)P
KB
198
De
novo
ex31
c.79
36G>
Tp
.(Glu
2646
*)(1
9)P
KB
352
NA
ex32
c.82
27C>
Tp
.(Gln
2743
*)T
his
stu
dy
PK
B32
3N
Aex
33c.
8311
C>
Tp
.(Arg
2771
*)(7
6,77
)P
KB
289
NA
ex34
c.87
43C>
Tp
.(Arg
2915
*)(2
2,77
–79)
PK
B42
2D
eno
voex
34c.
9396
C>
Ap
.(Cys
3132
*)T
his
stu
dy
PK
B18
6D
eno
voex
34c.
9961
C>
Tp
.(Arg
3321
*)(4
,75,
76,7
9)P
KB
56D
eno
voex
34c.
1013
5C>
Tp
.(Gln
3379
*)(4
6)P
KB
168
De
novo
ex39
c.10
750C>
Tp
.(Gln
3584
*)(1
9)P
KB
46D
eno
voex
39c.
1084
1C>
Gp
.(Ser
3614
*)(4
6)P
KB
41N
Aex
39c.
1111
9C>
Tp
.(Arg
3707
*)(4
6)P
KB
44N
Aex
39c.
1111
9C>
Tp
.(Arg
3707
*)(4
6)P
KB
42D
eno
voex
39c.
1126
9C>
Tp
.(Gln
3757
*)(2
2,46
)P
KB
25N
Aex
39c.
1143
4C>
Tp
.(Gln
3812
*)(4
6)P
KB
244
De
novo
ex39
c.11
674
C>
Tp
.(Gln
3892
*)(7
6)P
KB
178
NA
ex39
c.11
704C>
Tp
.(Gln
3902
*)(1
9)P
KB
425
De
novo
ex39
c.11
731C>
Tp
.(Gln
3911
*)T
his
stu
dy
PK
B46
1aN
Aex
39c.
1174
9C>
Tp
.(Gln
3917
*)T
his
stu
dy
PK
B46
3N
Aex
39c.
1184
5C>
Tp
.(Gln
3949
*)T
his
stu
dy
PK
B18
1N
Aex
39c.
1186
9C>
Tp
.(Gln
3957
*)(1
9)P
KB
358
NA
ex39
c.11
944C>
Tp
.(Arg
3982
*)(1
8,22
,77)
PK
B40
NA
ex39
c.12
274C>
Tp
.(Gln
4092
*)(1
8,46
)P
KB
114
De
novo
ex39
c.12
274C>
Tp
.(Gln
4092
*)(1
8,46
)P
KB
65N
Aex
39c.
1207
6C>
Tp
.(Gln
4026
*)(4
6)P
KB
333
NA
ex39
c.12
703C>
Tp
.(Gln
4235
*)(4
)P
KB
410
NA
39c.
1276
0C>
Tp
.(Gln
4254
*)(2
2)P
(co
nti
nu
ed)
3Human Molecular Genetics, 2018, Vol. 00, No. 00 |
Downloaded from https://academic.oup.com/hmg/advance-article-abstract/doi/10.1093/hmg/ddy241/5042898by Ospedale Casa Sollievo della Sofferenza useron 27 August 2018
Tab
le1.
Co
nti
nu
ed
IDIn
her
itan
ceEx
on
/in
tro
nV
aria
nt
AA
chan
geR
efer
ence
AC
MG
clas
sifi
cati
on
KB
82D
eno
voex
39c.
1284
4C>
Tp
.(Arg
4282
*)(1
9)P
KB
350
De
novo
ex39
c.12
844C>
Tp
.(Arg
4282
*)(1
9)P
KB
189
De
novo
ex39
c.12
955A
>T
p.(A
rg43
19*)
(19,
22)
PK
B18
3D
eno
voex
39c.
1345
0C>
Tp
.(Arg
4484
*)(9
,22,
74,7
7)P
KB
450
NA
ex39
c.13
450C>
Tp
.(Arg
4484
*)(9
,22,
74,7
7)P
KB
175
De
novo
ex39
c.13
507C>
Tp
.(Gln
4503
*)(1
9)P
KB
73D
eno
voex
40c.
1366
6A>
Tp
.(Lys
4556
*)(4
6)P
KB
83N
Aex
48c.
1502
2G>
Tp
.(Glu
5008
*)(1
9)P
KB
377
NA
ex48
c.15
061C>
Tp
.(Arg
5021
*)(1
8,76
)P
KB
45N
Aex
48c.
1507
9C>
Tp
.(Arg
5027
*)(2
2,46
,77)
PK
B72
NA
ex48
c.15
079C>
Tp
.(Arg
5027
*)(2
2,46
,77)
PK
B36
2N
Aex
50c.
1601
8C>
Tp
.(Arg
5340
*)(7
7)P
KB
130
NA
ex52
c.16
360C>
Tp
.(Arg
5454
*)(4
,22,
75,7
7)P
Fram
esh
ift
KB
454
NA
ex3
c.23
4_23
5del
GC
p.(G
ln79
Ala
fs*7
)T
his
stu
dy
PK
B46
9N
Aex
3c.
345d
up
Ap
.(Ser
116I
lefs
*7)
Th
isst
ud
yP
KB
337
NA
ex4
c.44
6_44
9del
TA
TG
p.(V
al14
9Gly
fs*5
8)T
his
stu
dy
PK
B75
De
novo
ex4
c.47
2del
Tp
.(Cys
158V
alfs
*50)
(46)
PK
B8
De
novo
ex5
c.58
8del
Cp
.(Cys
197A
lafs
*11)
(45)
PK
B58
NA
ex6
c.70
5del
Ap
.(Glu
237S
erfs
*24)
(46)
PK
B57
NA
ex8
c.10
35_1
036d
elC
Tp
.(Cys
346S
erfs
*17)
(46)
PK
B89
NA
ex10
c.13
45_1
346d
elC
Tp
.(Leu
449V
alfs
*5)
(46,
74)
PK
B15
6D
eno
voex
10c.
1503
du
pT
p.(P
ro50
2Ser
fs*7
)(1
9)P
KB
116
NA
ex10
c.16
34d
elT
p.(L
eu54
5Arg
fs*3
85)
(76)
PK
B34
9N
Aex
10c.
1634
del
Tp
.(Leu
545A
rgfs
*385
)(7
6)P
KB
545
NA
10c.
2091
du
pC
p.(T
hr6
98H
isfs
*6)
Th
isst
ud
yP
KB
369
NA
ex11
c.35
96_3
597d
elp
.(Leu
1199
His
fs*7
)T
his
stu
dy
PK
B48
De
novo
ex11
c.29
93d
up
Cp
.(Met
999T
yrfs
*69)
(46)
PK
B20
3N
Aex
11c.
3161
_317
1del
CG
TT
GA
GT
CC
Cp
.(Pro
1054
His
fs*1
0)(1
8,19
,43)
PK
B30
9N
Aex
11c.
3730
del
Gp
.(Val
1244
Ser
fs*8
6)(1
9)P
KB
142
De
novo
ex13
c.40
21d
elG
p.(V
al13
41Le
ufs
*35)
(19)
PK
B31
1N
Aex
14c.
4135
_413
6del
AT
p.(M
et13
79V
alfs
*52)
(19,
22,8
0)P
KB
524
NA
14c.
4135
_413
6del
AT
p.(M
et13
79V
alfs
*52)
(19,
22,8
0)P
KB
188
De
novo
ex16
c.44
54d
elC
p.(P
ro14
85Le
ufs
*21)
(19)
PK
B15
9N
Aex
19c.
4896
_490
5del
AG
AT
GC
CC
TT
p.(A
sp16
33A
lafs
*86)
(19)
PK
B44
3aD
eno
voex
25c.
5575
del
Gp
.(Asp
1859
Th
rfs*
17)
Th
isst
ud
yP
KB
3N
Aex
26c.
5652
du
pp
.(Lys
1885
Gln
fs*1
8)(4
5)P
KB
84N
Aex
26c.
5779
del
Cp
.(Gln
1927
Lysf
s120
*)(4
6)P
KB
146
De
novo
ex27
c.58
57d
elC
p.(L
eu19
53T
rpfs
*94)
(19)
PK
B20
8N
Aex
28c.
5954
del
Cp
.(Th
r198
5Ly
sfs*
62)
(19)
PK
B22
1D
eno
voex
29c.
6149
_615
0del
GA
p.(A
rg20
50Ly
sfs*
6)(1
9)P
KB
525
NA
30c.
6212
_621
3del
AC
p.(H
is20
71Pr
ofs
*10)
Th
isst
ud
yP
KB
152
De
novo
ex31
c.65
83d
elA
p.(T
hr2
195P
rofs
*69)
(19)
P
(co
nti
nu
ed)
4 | Human Molecular Genetics, 2018, Vol. 00, No. 00
Downloaded from https://academic.oup.com/hmg/advance-article-abstract/doi/10.1093/hmg/ddy241/5042898by Ospedale Casa Sollievo della Sofferenza useron 27 August 2018
Tab
le1.
Co
nti
nu
ed
IDIn
her
itan
ceEx
on
/in
tro
nV
aria
nt
AA
chan
geR
efer
ence
AC
MG
clas
sifi
cati
on
KB
267
NA
ex31
c.65
94d
elC
p.(T
yr21
99Il
efs*
65)
(75)
PK
B79
De
novo
ex31
c.65
95d
elT
p.(T
yr21
99Il
efs*
65)
(4,2
0,22
,45,
46,7
5–77
,81)
PK
B10
2D
eno
voex
31c.
6595
del
Tp
.(Tyr
2199
Ilef
s*65
)(4
,20,
22,4
5,46
,75–
77,8
1)P
KB
342
NA
ex31
c.65
95d
elT
p.(T
yr21
99Il
efs*
65)
(4,2
0,22
,45,
46,7
5–77
,81)
PK
B67
De
novo
ex31
c.66
38_6
641d
elG
CG
Cp
.(Gly
2213
Ala
fs*5
0)(4
6)P
KB
176
NA
ex31
c.67
38d
elA
p.(L
ys22
46A
snfs
*18)
(19)
PK
B25
3N
Aex
31c.
6794
del
Gp
.(Gly
2265
Glu
fs*2
1)(1
9,22
)P
KB
278
NA
ex31
c.74
81d
up
Tp
.(Ala
2496
Serf
s*10
)(1
9,79
)P
KB
313
De
novo
ex32
c.81
96d
elG
p.(S
er27
33V
alf
s*24
)(1
9)P
KB
80N
Aex
33c.
8273
del
Gp
.(Gly
2758
Ala
fs*2
9)(4
6)P
KB
243
De
novo
ex34
c.84
30_8
431i
nsA
Ap
.(Gln
2811
Asn
fs*4
1)(1
9)P
KB
182
NA
ex34
c.92
03d
elA
p.(G
ln30
68G
lyfs
*3)
(19)
PK
B30
NA
ex38
c.10
606d
elC
p.(A
rg35
36A
lafs
*122
)(4
6)P
KB
101
De
novo
ex39
c.11
066_
1107
8del
CT
GG
AT
CC
CT
GG
Cp
.(Ala
3689
Va
lfs*
56)
(46)
P
KB
504
De
novo
ex39
c.11
093d
up
Gp
.(Ph
e369
9Leu
fs*1
4)T
his
stu
dy
PK
B49
5D
eno
voex
39c.
1171
5del
Gp
.(Gln
3905
His
fs*7
4)T
his
stu
dy
PK
B17
2N
Aex
39c.
1264
7del
Cp
.(Pro
4216
Leu
fs*6
2)(1
9)P
KB
192
De
novo
ex39
c.12
966d
elA
p.(G
ln43
22H
isfs
*62)
(19)
PK
B54
NA
ex39
c.13
129d
up
Tp
.(Trp
4377
Leu
fs*3
3)(4
6)P
KB
121
De
novo
ex39
c.13
277d
up
Tp
.(Ala
4428
Serf
s*59
)(1
9)P
KB
540
NA
41c.
1378
0del
Gp
.(Ala
4594
Pro
fs*2
3)(2
2)P
KB
123
De
novo
ex42
c.13
884d
up
Cp
.(Th
r462
9His
f*18
)(1
9)P
KB
481
NA
ex42
c.13
895d
up
Cp
.(Ser
4633
Ilef
s*14
)T
his
stu
dy
PK
B19
7D
eno
voex
47c.
1459
2du
pG
p.(P
ro48
65A
lafs
*48)
(19)
PK
B12
5N
Aex
48c.
1503
1del
Gp
.(Glu
5011
Ser
fs*4
0)(1
9)P
KB
16D
eno
voex
48c.
1537
4du
pT
p.(P
he5
126L
eufs
*12)
(45)
PK
B53
5N
A50
c.16
043_
1604
4del
AC
p.(H
is53
48Le
ufs
*14)
Th
isst
ud
yP
KB
355
NA
ex53
c.16
438_
1644
1del
AA
CT
p.(A
sn54
80V
al*6
)(7
5)P
KB
64N
Aex
53c.
1646
9_16
470d
elA
Ap
.(Lys
5490
Arg
fs*2
1)(4
6)P
KB
533
NA
53c.
1646
9_16
470d
elA
Ap
.(Lys
5490
Arg
fs*2
1)(4
6)P
Mis
sen
seK
B21
ain
her
ited
Mex
3c.
346T>
Cp
.(Ser
116P
ro)
(19)
LBK
B21
a,b
inh
erit
edM
ex4
c.51
0G>
Cp
.Gln
170H
is(1
9,45
)P
KB
256
NA
ex5
c.62
6C>
Tp
.(Th
r209
Ile)
(19)
VO
US
KB
458
NA
ex8
c.10
76G>
Cp
.(Arg
359P
ro)
Th
isst
ud
yV
OU
SK
B26
9in
her
ited
Mex
10c.
1940
C>
Ap
.(Pro
647G
ln)
(19,
45,7
7)LB
KB
126
NA
ex10
c.20
74C>
Ap
.(Pro
692T
hr)
(77)
LBK
B37
4cin
her
ited
Mex
10c.
2074
C>
Ap
.(Pro
692T
hr)
(77)
LBK
B48
7N
Aex
10c.
2654
C>
Tp
.(Pro
885L
eu)
Th
isst
ud
yV
OU
SK
B37
0aIn
her
ited
P–M
ex11
c.28
37C>
Gp
.(Ala
946G
ly)
Th
isst
ud
yLB
KB
215
Inh
erit
edM
ex11
c.33
92C>
Tp
.(Pro
1131
Leu
)(1
9)LB
KB
341
Inh
erit
edM
ex11
c.33
92C>
Tp
.(Pro
1131
Leu
)(1
9)LB
KB
222
Inh
erit
edP
ex11
c.35
72C>
Tp
.(Pro
1191
Leu
)(1
9)LB
(co
nti
nu
ed)
5Human Molecular Genetics, 2018, Vol. 00, No. 00 |
Downloaded from https://academic.oup.com/hmg/advance-article-abstract/doi/10.1093/hmg/ddy241/5042898by Ospedale Casa Sollievo della Sofferenza useron 27 August 2018
Tab
le1.
Co
nti
nu
ed
IDIn
her
itan
ceEx
on
/in
tro
nV
aria
nt
AA
chan
geR
efer
ence
AC
MG
clas
sifi
cati
on
KB
32In
her
ited
Pex
11c.
3773
G>
Ap
.(Arg
1258
Gln
)(4
6)LB
KB
307
De
novo
ex14
c.41
71G>
Ap
.(Glu
1391
Lys)
(19,
22)
LPK
B28
aIn
her
ited
Mex
15c.
4249
A>
Gp
.(Met
1417
Val
)(4
6)LB
KB
28a
Inh
erit
edM
ex15
c.42
52C>
Ap
.(Leu
1418
Met
)(4
6)LB
KB
174
Inh
erit
edM
ex15
c.42
83T>
Cp
.(Ile
1428
Th
r)(1
9)LB
KB
138
NA
ex16
c.44
27C>
Gp
.(Ser
1476
Cy
s)(1
9)V
OU
SK
B34
Inh
erit
edP
ex16
c.45
65A>
Gp
.(Gln
1522
Arg
)(4
6)LB
KB
535
NA
25c.
5549
G>
Ap
.(Gly
1850
Asp
)T
his
stu
dy
LBK
B11
9In
her
ited
Mex
31c.
6638
G>
Ap
.(Gly
2213
Asp
)(1
9)LB
KB
204c
Inh
erit
edM
ex31
c.66
38G>
Ap
.(Gly
2213
Asp
)(1
9)LB
KB
330a
NA
ex31
c.67
33C>
Gp
.(Leu
2245
Val
)T
his
stu
dy
VO
US
KB
326c
Inh
erit
edM
ex31
c.68
11C>
Tp
.(Pro
2271
Ser)
(19)
LBK
B10
7aN
Aex
31c.
6970
C>
Ap
.(Pro
2324
Th
r)(1
9)V
OU
SK
B43
0N
Aex
31c.
7378
C>
Tp
.(Arg
2460
Cys
)(7
7)LB
KB
122
Inh
erit
edM
ex31
c.78
29T>
Cp
.(Leu
2610
Pro
)(1
9,73
)LB
KB
287
Inh
erit
edM
ex31
c.78
29T>
Cp
.(Leu
2610
Pro
)(1
9,73
)LB
KB
27N
Aex
34c.
8521
C>
Ap
.(Pro
2841
Th
r)(4
6)V
OU
SK
B33
0aN
Aex
34c.
8774
C>
Tp
.(Ala
2925
Va
l)T
his
stu
dy
VO
US
KB
443a
Inh
erit
edM
ex34
c.99
71G>
Tp
.(Gly
3324
Val
)T
his
stu
dy
LBK
B32
6cIn
her
ited
Pex
34c.
1019
2A>
Gp
.(Met
3398
Val
)(1
9)LB
KB
357
Inh
erit
edP
ex34
c.10
192A
>G
p.(M
et33
98V
al)
(19)
LBK
B29
7N
Aex
35c.
1025
6A>
Gp
.(Asp
3419
Gly
)(7
7)LB
KB
378
NA
ex35
c.10
256A
>G
p.(A
sp34
19G
ly)
(77)
LBK
B29
2D
eno
voex
37c.
1049
9G>
Tp
.(Gly
3500
Val
)(1
9)LP
KB
86a
NA
ex39
c.10
966C>
Tp
.(Arg
3656
Cys
)(1
9)V
OU
SK
B37
4cIn
her
ited
Pex
39c.
1138
0C>
Tp
.(Pro
3794
Ser)
Th
isst
ud
yLB
KB
293
Inh
erit
edP
ex39
c.11
794C>
Gp
.(Gln
3932
Glu
)(1
9)LB
KB
204c
Inh
erit
edP
ex39
c.12
070A
>G
p.(L
ys40
24G
lu)
(19)
LBK
B38
5N
Aex
39c.
1230
2A>
Cp
.(Gln
4101
Pro
)T
his
stu
dy
VO
US
KB
247
NA
ex39
c.12
485G>
Ap
.(Arg
4162
Gln
)(1
9)V
OU
SK
B10
7aN
Aex
39c.
1248
8C>
Tp
.(Pro
4163
Leu
)(1
9)V
OU
SK
B17
0In
her
ited
Mex
39c.
1325
6C>
Tp
.(Pro
4419
Leu
)(1
9)LB
KB
512
NA
45c.
1438
1A>
Gp
.(Lys
4794
Arg
)T
his
stu
dy
VO
US
KB
86a
NA
ex48
c.14
893G>
Ap
.(Ala
4965
Th
r)(1
9)V
OU
SK
B38
aD
eno
voex
48c.
1508
4C>
Gp
.(Asp
5028
Glu
)(4
6)LP
KB
154
NA
ex48
c.15
088C>
Tp
.(Arg
5030
Cys
)(1
8,45
)V
OU
SK
B18
5D
eno
voex
48c.
1508
8C>
Tp
.(Arg
5030
Cys
)(1
8,45
)LP
KB
423
Inh
erit
edP
ex48
c.15
089G>
Ap
.(Arg
5030
His
)T
his
stu
dy
LBK
B38
aD
eno
voex
48c.
1510
0T>
Gp
.(Ph
e503
4Val
)(4
6)LP
KB
76D
eno
voex
48c.
1517
6A>
Cp
.(His
5059
Pro
)(4
6)LP
KB
129
NA
ex48
c.15
292A
>C
p.(T
hr5
098P
ro)
(19)
VO
US
KB
462
De
novo
ex48
c.15
310T>
Cp
.(Cys
5104
Arg
)T
his
stu
dy
LPK
B17
1In
her
ited
Mex
48c.
1556
5G>
Ap
.(Gly
5189
Arg
)(1
8,22
,46)
VO
US
KB
264
NA
ex48
c.15
640C>
Tp
.(Arg
5214
Cys
)(2
2,45
,75,
76)
LP
(co
nti
nu
ed)
6 | Human Molecular Genetics, 2018, Vol. 00, No. 00
Downloaded from https://academic.oup.com/hmg/advance-article-abstract/doi/10.1093/hmg/ddy241/5042898by Ospedale Casa Sollievo della Sofferenza useron 27 August 2018
Tab
le1.
Co
nti
nu
ed
IDIn
her
itan
ceEx
on
/in
tro
nV
aria
nt
AA
chan
geR
efer
ence
AC
MG
clas
sifi
cati
on
KB
376
NA
ex48
c.15
640C>
Tp
.(Arg
5214
Cys
)(2
2,45
,75,
76)
LPK
B24
De
novo
ex48
c.15
641G>
Ap
.(Arg
5214
His
)(4
)LP
KB
219
NA
ex48
c.15
641G>
Ap
.(Arg
5214
His
)(4
,75)
LPK
B40
8D
eno
voex
48c.
1564
1G>
Ap
.(Arg
5214
His
)(4
,75)
LPK
B10
9D
eno
voex
48c.
1564
9T>
Cp
.(Trp
5217
Arg
)(1
9)LP
KB
17D
eno
voex
50c.
1601
9G>
Ap
.(Arg
5340
Gln
)(2
2,46
)LP
KB
169
De
novo
ex51
c.16
273G>
Ap
.(Glu
5425
Lys)
(19,
22,7
8)P
KB
90N
Aex
51c.
1629
5G>
Ap
.(Arg
5432
Gln
)(2
2,82
,83)
LPK
B46
7N
Aex
51c.
1629
5G>
Ap
.(Arg
5432
Gln
)(2
2,82
,83)
LPK
B48
0N
Aex
52c.
1638
5A>
Gp
.(Asp
5462
Gly
)T
his
stu
dy
VO
US
KB
177
De
novo
ex52
c.16
412G>
Tp
.(Arg
5471
Met
)(1
9)LP
KB
489
NA
ex53
c.16
498C>
Tp
.(Arg
5500
Trp
)(7
8)P
KB
120
De
novo
ex54
c.16
528T>
Gp
.(Tyr
5510
Asp
)(1
9)LP
Ind
elK
B40
4In
her
ited
Mex
10c.
2283
_230
9del
p.(A
la76
5_G
ln77
3del
)T
his
stu
dy
LBK
B27
4N
Aex
10c.
2532
_259
1del
p.(A
rg84
5_Pr
o86
4del
)(1
9)V
OU
SK
B37
0dD
eno
voex
14c.
4202
_421
0del
p.(S
er14
01_C
ys14
03d
el)
Th
isst
ud
yLP
KB
384
NA
ex39
c.11
220_
1122
2du
pp
.(Gln
3745
du
p)
Th
isst
ud
yLB
KB
461a
,dN
Aex
39c.
1122
0_11
222d
up
p.(G
ln37
45d
up
)T
his
stu
dy
LBK
B28
1In
her
ited
Mex
39c.
1171
4_11
716d
up
p.(G
ln39
05d
up
)(1
9)LB
KB
71In
her
ited
Mex
39c.
1181
9_11
836d
up
p.(L
eu39
40_G
ln39
45d
up
)(4
6)LB
KB
227
Inh
erit
edP
ex39
c.11
843_
1186
0del
p.(L
3948
_Q39
53d
el)
(19)
LBK
B22
8In
her
ited
Pex
39c.
1185
4_11
874d
up
p.(Q
3952
_Q39
58d
up
)(1
9)LB
KB
77N
Aex
48c.
1516
3_15
168d
up
p.(A
sp50
55_L
eu50
56d
up
)(1
8,46
)V
OU
SK
B40
3D
eno
voex
53c.
1648
9_16
491d
elp
.(Ile
5497
del
)(2
2,46
,75,
76)
LPK
B53
NA
ex53
c.16
489_
1649
1del
p.(I
le54
97d
el)
(22,
46,7
5,76
)LP
Sp
lice
site
KB
286
De
novo
int
2–3
c.17
7-2A
>C
r.17
7_40
0del
224;
p.S
59R
fs*8
6(1
9)P
KB
31N
Ain
t3–
4c.
400þ
1G>
Ar.
177_
400d
el22
4;p
.Ser
59A
rgfs
*86
(46)
PK
B20
De
novo
int
3–5
c.40
1-3
A>
Gr.
400_
401i
nsA
G;p
.Gly
134G
lufs
*75
(46)
PK
B44
2N
Ain
t6–
7c.
840-
6del
Cr.
?T
his
stu
dy
VO
US
KB
519
NA
int
13–1
4c.
4132
-2A>
Gr.
?T
his
stu
dy
PK
B52
9N
Ain
t16
–17
c.45
84-6
C>
Gr.
?T
his
stu
dy
VO
US
KB
210
De
novo
in17
–18
c.46
93þ
1G>
Ar.
4681
_469
3del
13;p
.Val
1561
Arg
fsp
lice
*11
(48)
PK
B29
NA
int
42–4
3c.
1399
9þ
5G>
Ar.
1384
0_13
999d
el16
0;p
.Asn
4614
Ilef
s*5
(46,
77)
PK
B29
0D
eno
voin
t47
–48
c.14
643þ
1G>
Ar.
1464
4_14
875d
el23
2;p
.Glu
4882
Pro
fs*3
6(1
9)P
KB
195
De
novo
int
47–4
8c.
1464
4-3C>
Gr.
1464
4_14
875d
el23
2;p
.Gln
4882
Pro
fs*3
6(1
9)P
KB
7D
eno
voin
t44
–45
c.14
252-
6_14
252-
5in
sGA
AA
r.14
252_
1438
2del
131;
p.V
al47
51_G
lufs
pli
ce*2
2(4
5)P
KB
360
NA
int
47–4
8c.
1464
4-2A
>T
r.?
(77)
PK
B49
6N
Ain
t53
–54
c.16
520_
1652
1þ
1del
AG
Gr.
?T
his
stu
dy
P
(co
nti
nu
ed)
7Human Molecular Genetics, 2018, Vol. 00, No. 00 |
Downloaded from https://academic.oup.com/hmg/advance-article-abstract/doi/10.1093/hmg/ddy241/5042898by Ospedale Casa Sollievo della Sofferenza useron 27 August 2018
Tab
le1.
Co
nti
nu
ed
IDIn
her
itan
ceEx
on
/in
tro
nV
aria
nt
AA
chan
geR
efer
ence
AC
MG
clas
sifi
cati
on
Gro
ssd
elet
ion
KB
43e
NA
ex48
–51
c.15
785-
238_
1616
8del
ins
p.?
(19)
PK
DM
6AN
on-s
ense
KB
215f
De
novo
ex6
c.51
4C>
Tp
.(Arg
172*
)(1
9,22
,84)
PK
B34
1D
eno
voex
6c.
514C>
Tp
.(Arg
172*
)(2
2,84
)P
Fram
esh
ift
KB
39N
Aex
16c.
1846
_184
9del
p.(T
hr6
16ty
rfs*
8)(1
9)P
KB
141
NA
ex17
c.21
18_2
119i
ns
p.(G
707H
fs*1
3)T
his
stu
dy
PK
B43
4N
Aex
17c.
2515
_251
8del
p.(A
sn83
9Val
fs*2
7)(8
5)P
KB
381
NA
ex20
c.30
44d
elC
p.(T
hr1
015M
etfs
*33)
Th
isst
ud
yP
Mis
sen
seK
B41
5N
Aex
16c.
1843
C>
Tp
.(Leu
615P
he)
Th
isst
ud
yV
OU
SK
B27
2N
Aex
17c.
2326
G>
Tp
.(Asp
776T
yr)
Th
isst
ud
yV
OU
SK
B13
1D
eno
voex
20c.
2939
A>
Tp
.(Asp
980V
al)
(19)
PK
B38
0D
eno
voex
26c.
3743
A>
Gp
.(Gln
1248
Arg
)T
his
stu
dy
PG
ross
del
etio
ns
KB
11N
Aex
1–2
p.?
Th
isst
ud
yP
KB
50D
eno
voex
5–9
p.?
(17)
PS
pli
cesi
teK
B31
4D
eno
voin
t11
–12
c.97
5-1G>
Ar.
876_
1320
del
;p.C
ys29
3Ile
fsX
26T
his
stu
dy
PK
B12
7D
eno
voin
t22
–23
c.33
84þ
3_33
84þ
6del
r.32
10_3
284d
el;p
.Asn
1070
_Lys
1094
del
(19)
P
aD
etec
ted
toge
ther
wit
hp
ath
oge
nic
mu
tati
on.
bFa
lls
inth
ela
stba
seo
fex
on
,pre
dic
ted
tod
isru
pt
the
do
no
rsp
lice
site
.c Pa
tien
tsw
ith
com
po
un
dh
eter
ozy
gou
sva
rian
ts.
dD
etec
ted
toge
ther
wit
ho
ther
vari
ant.
eId
enti
fied
byM
LPA
.f D
etec
ted
toge
ther
wit
hm
isse
nse
inK
MT
2D.
P,p
ath
oge
nic
;LP,
like
lyp
ath
oge
nic
;LB
,lik
ely
ben
ign
;VO
US
vari
ant
of
un
kno
wn
sign
ifica
nce
.
8 | Human Molecular Genetics, 2018, Vol. 00, No. 00
Downloaded from https://academic.oup.com/hmg/advance-article-abstract/doi/10.1093/hmg/ddy241/5042898by Ospedale Casa Sollievo della Sofferenza useron 27 August 2018
The 14 tested missense variants have been selected as repre-sentative of the entire set of variants identified in our KS cohortthat harbor amino acid changes in both N-terminal and C-ter-minal domains of the protein including PHD 4–5–6–7, FYRN,WIN and SET domains. This construct was selected because theprotein encoded by the FUSION–KMT2D wild-type vector retainsdose-dependent mono, di and tri methyltransferase activityin vitro to the same extent of a full-length wild-type protein(Supplementary Material, Fig. S4), while guaranteeing highertransfection efficiency (not shown).
We measured the ability of the mutated proteins to catalyzemono-, di- and tri-methylation of H3K4 (H3K4me1, me2 andme3) in vitro, using purified nucleosomal histones as substrateand western blot analysis with antibodies directed against thesethree histone marks. In this assay, 9/14 missense variants led toheterogeneous but significant impairment in H3K4 monome-thylation, compared with wild-type KMT2D (range 30–80%),with two additional mutants showing borderline effects (Fig. 2Aand B). Additionally, six of the nine mutations were associatedwith reduced H3K4me2, and 6 with reduced H3K4me3 levels(Fig. 2A and B).
To independently confirm the reduced H3K4 tri-methyltransferase activity, we used a more sensitive epigeneticEGFP reporter system that specifically monitors this modifica-tion (23) (Fig. 2C). Fluorescence data confirmed the significantlyreduced ability to catalyze H3K4me3 for all 6 missense variantstested (on average, �50% compared with the KMT2D wild-typecounterpart) (Fig. 2D). Together, these data indicate that, to vari-ous extents, a subset of KS-associated missense mutations can
inactivate the function of KMT2D, analogous to truncatingmutations.
KMT2D-complex protein interaction
ASH2L and RbBP5 interact with KMT2D co-regulating the ex-pression of target genes (24,25). To determine whether KMT2Dmissense variants affect its function by altering the interactionwith ASH2L and RbBP5, we immunoprecipitated protein lysatesfrom HEK 293T cells co-transfected with Flag-tagged FUSION–KMT2D constructs and vectors expressing either ASH2L orRbBP5. Western blot analysis of immunoprecipitated cell lysatesshowed that the missense variants located at the C-terminalof the protein (p.H5059P, p.T5098P, p.G5189R, p.W5217R,p.R5340Q, p.E5425K p.R5471M and Y5510D) exhibit a weakerinteraction with both ASH2L and RbBP5, while p.E1391K andp.F5034V only distressed the interaction between KMT2Dand ASH2L, when compared with the wild-type construct.The missense variants p.I1428T and p.Q1522R exhibited anincreased interaction with both ASH2L and RbBP5, while thep.M1417V and p.S1476C only increased the interaction ofKMT2D with ASH2L (Fig. 3A).
Combined conformational 3D modeling and free energy in-teraction variations (see Supplementary Material, Fig. S2) indi-cated that the p.R5340Q missense variant—in the WIN motif ofKMT2D—alters the protein domain structure, changing an evo-lutionarily conserved amino acid position expected to tolerateArginine and minimally Methionine (Fig. 3B, left). Since theKMT2D WIN motif (5337–5342) is necessary for the interaction
Figure 1. Missense variants distribution across the entire length or FUSION–KMT2D gene. (A) Schematic representation of the KMT2D protein (PHD, plant homeodo-
main; HMG, high-mobility group; Coiled Coil domain; LXXLL domain, motifs with the consensus sequence L–X–X–L–L motif; Poli Q region, Poli Q reach region identified
in this study; ZF domain, Zinc Finger Domain; FYRN, FY-rich, N-terminal; FYRC, FY-rich, C-terminal; WIN, WDR5 Interaction domain, SET, Su(var)3-9, Enhancer-of-
zeste, Trithorax). In black, missense variants found in our cohort of KS patients. De novo variants are indicated in bold, pathogenic variants with P, likely pathogenic
with LP, likely benign with LP, and VOUS with V. (B) Distribution of the functionally analyzed KMT2D missense variants across the FUSION–KMT2D construct composed
of PHD4–5–6 (amino acids 1358–1572) and ZF-PHD7–FYRN–FYRC–WIN–SET–post-SET domains (amino acids 4507–5537). De novo variants are indicated in bold, patho-
genic variants are indicated with P, likely pathogenic with LP, likely benign with LB and VOUS with V.
9Human Molecular Genetics, 2018, Vol. 00, No. 00 |
Downloaded from https://academic.oup.com/hmg/advance-article-abstract/doi/10.1093/hmg/ddy241/5042898by Ospedale Casa Sollievo della Sofferenza useron 27 August 2018
Tab
le2.
Bio
info
rmat
ics
anal
ysis
of
KM
T2D
mis
sen
seva
rian
ts
Pred
icti
on
soft
war
es
Co
de
Var
ian
tA
Ach
ange
AC
MG
Var
ian
tC
lass
Poly
ph
enA
lign
GV
GD
Pro
vean
SIFT
UM
Dp
red
icto
rM
uta
tio
nT
aste
r
Sco
reR
esu
ltSc
ore
Res
ult
Sco
reR
esu
ltSc
ore
Res
ult
Sco
reR
esu
ltSc
ore
Res
ult
KB
21a
c.34
6T>
Cp
.(Ser
116P
ro)
LBPD
0.99
773
.35
Cla
ssC
65�
3.35
D0.
001
D90
PD
C0,
994
KB
21a
,bc.
510G>
Cp
.Gln
170H
isP
PD1.
0024
.08
Cla
ssC
15�
2.32
N0.
016
D10
0P
DC
0.99
8K
B25
6c.
626C>
Tp
.(Th
r209
Ile)
VO
US
B0.
096
89.2
8C
lass
C65
�2.
23N
0.00
9D
75P
PM0.
980
KB
458
c.10
76G>
Cp
.(Arg
359P
ro)
VO
US
B0.
011
102.
71C
lass
C65
0.04
N0.
2056
T33
PMPM
0.99
9K
B26
9c.
1940
C>
Ap
.(Pro
647G
ln)
LBPD
0.59
375
75.1
4C
lass
C65
�0.
90N
0.07
71T
78P
PM0.
999
KB
126,
KB
374a
c.20
74C>
Ap
.(Pro
692T
hr)
LBB
0.00
837
.56
Cla
ssC
35�
0.92
N0.
011
D66
PPPM
0.99
9K
B48
7c.
2654
C>
Tp
.(Pro
885L
eu)
VO
US
B0.
000
97.7
8C
lass
C65
�1.
53N
0.00
3D
60PP
MPM
0.99
9K
B37
0ac.
2837
C>
Gp
.(Ala
946G
ly)
LBB
0.00
060
.00
Cla
ssC
55�
0.78
N0.
032
D30
PMPM
0.99
9K
B21
5,K
B34
1c.
3392
C>
Tp
.(Pro
1131
Leu
)LB
B0.
196
97.7
8C
lass
C65
�0.
30N
0.00
1D
39PM
PM0.
989
KB
222
c.35
72C>
Tp
.(Pro
1191
Leu
)LB
PD0.
764
97.7
8C
lass
C65
�2.
50D
0.00
1D
35PM
DC
0.99
9K
B32
c.37
73G>
Ap
.(Arg
1258
Gln
)LB
PD0.
997
42.8
1C
lass
C35
�1.
37N
0.00
2D
75P
PM0.
691
KB
307
c.41
71G
>Ap
.(Glu
1391
Lys)
LPPD
1.00
056
.87
Cla
ssC
55�
3.29
D0.
001
D78
PD
C0.
999
KB
28a
c.42
49A>
Gp
.(Met
1417
Val
)LB
PD0.
476
20.5
2C
lass
C15
00:1
9N
0.40
97T
60PP
MD
C0.
559
KB
28a
c.42
52C>
Ap
.(Leu
1418
Met
)LB
PD1.
000
14.3
0C
lass
C0
�1.
71N
0.00
4D
69PP
DC
0.99
9K
B17
4c.
4283
T>
Cp
.(Ile
1428
Th
r)LB
PD0.
889
89.2
8C
lass
C65
01:1
4N
0.58
47T
93P
DC
0.99
6K
B13
8c.
4427
C>
Gp
.(Ser
1476
Cys
)V
OU
SB
0.00
511
1.67
Cla
ssC
650.
05N
0.17
71T
99P
PM0.
885
KB
34c.
4565
A>
Gp
.(Gln
1522
Arg
)LB
PD0.
997
42.8
1C
lass
C35
�3.
43D
0.02
1D
100
PD
C0.
999
KB
535a
c.55
49G>
Ap
.(Gly
1850
Asp
)LB
PD0.
535
93.7
7C
lass
C65
�0.
72N
0.10
3T
87P
PM0.
997
KB
119,
KB
204a
c.66
38G>
Ap
.(Gly
2213
Asp
)LB
PD0.
454
93.7
7C
lass
C65
�0.
70N
0.00
4D
42PM
DC
0.91
7K
B33
0ac.
6733
C>
Gp
.(Leu
2245
Val
)V
OU
SPD
0,69
30.9
2C
lass
C25
�0.
46N
0.04
6D
69PP
PM0.
996
KB
326a
c.68
11C>
Tp
.(Pro
2271
Ser)
LBB
0.02
473
.35
Cla
ssC
65�
2.00
N0.
004
D75
PD
C0.
850
KB
107a
c.69
70C>
Ap
.(Pro
2324
Th
r)V
OU
SPD
0.98
537
.56
Cla
ssC
35�
2.47
N0.
003
D81
PPM
0.99
8K
B43
0c.
7378
C>
Tp
.(Arg
2460
Cys
)LB
PD1.
000
179.
53C
lass
C65
�2.
22N
0.02
3D
96P
DC
0.99
9K
B12
2,K
B28
7c.
7829
T>
Cp
.(Leu
2610
Pro
)LB
B0.
076
97.7
8C
lass
C65
�1.
20N
0.08
19T
72PP
DC
0.99
8K
B27
c.85
21C>
Ap
.(Pro
2841
Th
r)V
OU
SB
0.00
937
.56
Cla
ssC
35�
1.45
N0.
010
D69
PPD
C0.
997
KB
330a
c.87
74C>
Tp
.(Arg
2925
Val
)V
OU
SB
0.00
465
.28
Cla
ssC
65�
1.07
N0.
004
D84
PPM
0.95
8K
B44
3ac.
9971
G>
Tp
.(Gly
3324
Val
)LB
PD0.
988
109.
55C
lass
C65
�2.
24N
0.01
1D
78P
DC
0.99
9K
B32
6a,K
B35
7c.
1019
2A>
Gp
.(Met
3398
Val
)LB
B0.
000
20.5
2C
lass
C15
�2.
84D
0.00
2D
41PM
PM0.
958
KB
297,
KB
378
c.10
256A
>G
p.(A
sp34
19G
ly)
LBPD
1.00
093
.77
Cla
ssC
65�
1.81
N0.
000
D10
0P
DC
0.99
9K
B29
2c.
1049
9G>T
p.(G
ly35
00V
al)
LPPD
0.68
7510
9.55
Cla
ssC
65�
8.18
D0.
000
D10
0P
DC
0.99
9K
B86
ac.
1096
6C>
Tp
.(Arg
3656
Cys
)V
OU
SPD
1.00
017
9.53
Cla
ssC
65�
1.87
N0.
001
D84
PD
C0.
999
KB
374a
c.11
380C>
Tp
.(Pro
3794
Ser)
LBB
0.01
673
.35
Cla
ssC
65�
1.04
N0.
029
D66
PPPM
0.99
3K
B29
3c.
1179
4C>
Gp
.(Gln
3932
Glu
)LB
B0.
002
29.2
7C
lass
C25
�0.
68N
0.04
7D
45PM
PM0.
999
KB
204a
c.12
070A
>G
p.(L
ys40
24G
lu)
LBB
0.29
656
.87
Cla
ssC
5500
:08
N0.
000
D63
PPM
PM0.
997
KB
385
c.12
302A
>C
p.(G
ln41
01Pr
o)
VO
US
B0.
000
75.1
4C
lass
C65
�0.
03N
0.20
14T
66PP
PM0.
939
KB
247
c.12
485G>
Ap
.(Arg
4162
Gln
)V
OU
SB
0.00
342
.81
Cla
ssC
3500
:30
N0.
000
D72
PPPM
0.99
5K
B10
7ac.
1248
8C>
Tp
.(Pro
4163
Leu
)V
OU
SPD
0.99
197
.78
Cla
ssC
65�
2.29
N0.
000
D72
PPD
C0.
983
KB
170
c.13
256C>
Tp
.(Pro
4419
Leu
)LB
PD1.
000
97.7
8C
lass
C65
�4.
16D
0.00
0D
87P
DC
0.99
9K
B51
2c.
1438
1A>
Gp
.(Lys
4794
Arg
)V
OU
SPD
0.99
926
.00
Cla
ssC
25�
1.43
N0.
005
D69
PPD
C0.
999
KB
86a
c.14
893G>
Ap
.(Ala
4965
Th
r)V
OU
SPD
0.94
158
.02
Cla
ssC
55�
1.17
N0.
0847
T75
PD
C0.
997
(co
nti
nu
ed)
10 | Human Molecular Genetics, 2018, Vol. 00, No. 00
Downloaded from https://academic.oup.com/hmg/advance-article-abstract/doi/10.1093/hmg/ddy241/5042898by Ospedale Casa Sollievo della Sofferenza useron 27 August 2018
Tab
le2.
Co
nti
nu
ed
Pred
icti
on
soft
war
es
Co
de
Var
ian
tA
Ach
ange
AC
MG
Var
ian
tC
lass
Poly
ph
enA
lign
GV
GD
Pro
vean
SIFT
UM
Dp
red
icto
rM
uta
tio
nT
aste
r
Sco
reR
esu
ltSc
ore
Res
ult
Sco
reR
esu
ltSc
ore
Res
ult
Sco
reR
esu
ltSc
ore
Res
ult
KB
38a
c.15
084C
>Gp
.(Asp
5028
Glu
)LP
PD0.
999
44.6
0C
lass
C35
�3.
83D
0.00
1D
66PP
DC
0.99
9K
B15
4,K
B18
5c.
1508
8C>T
p.(A
rg50
30C
ys)
VO
US
PD1.
000
179.
53C
lass
C65
�7.
56D
0.00
0D
99P
DC
0.99
9K
B42
3c.
1508
9G>
Ap
.(Arg
5030
His
)LB
PD1.
000
28.8
2C
lass
C25
�4.
79D
0.00
0D
81P
DC
0.99
9K
B38
ac.
1510
0T>G
p.(P
he5
034V
al)
LPPD
0.99
948
.95
Cla
ssC
45�
6.17
D0.
001
D84
PD
C0.
999
KB
76c.
1517
6A>C
p.(H
is50
59Pr
o)
LPPD
1.00
076
.28
Cla
ssC
65�
9.60
D0.
001
D93
PD
C0.
999
KB
129
c.15
292A
>C
p.(T
hr5
098P
ro)
VO
US
PD0.
992
37.5
6C
lass
C35
�4.
21D
0.10
35T
93P
DC
0.99
9K
B46
2c.
1531
0T>C
p.(C
ys51
04A
rg)
LPPD
1.00
017
9.53
Cla
ssC
65�
11.5
1D
0.00
0D
96P
DC
0.99
9K
B17
1c.
1556
5G>
Ap
.(Gly
5189
Arg
)V
OU
SPD
1.00
012
5.13
Cla
ssC
65�
7.68
D0.
000
D10
0P
DC
0.99
9K
B26
4,K
B37
6c.
1564
0C>
Tp
.(Arg
5214
Cys
)LP
PD1.
000
179.
53C
lass
C65
�7.
68D
0.00
0D
96P
DC
0.99
9K
B24
,KB
219,
KB
408
c.15
641G
>Ap
.(Arg
5214
His
)LP
PD1.
000
28.8
2C
lass
C25
�4.
80D
0.00
0D
78P
DC
0.99
9K
B10
9c.
1564
9T>C
p.(T
rp52
17A
rg)
LPPD
1.00
010
1.29
Cla
ssC
65�
13.4
3D
0.00
0D
96P
DC
0.99
9K
B17
c.16
019G
>Ap
.(Arg
5340
Gln
)LP
PD1.
000
42.8
1C
lass
C35
�3.
84D
0.00
0D
81P
DC
0.99
9K
B16
9c.
1627
3G>A
p.(G
lu54
25Ly
s)P
PD1.
000
56.8
7C
lass
C55
�3.
70D
0.00
0D
75P
DC
0.99
9K
B90
,KB
467
c.16
295G>
Ap
.(Arg
5432
Gln
)LP
PD1.
000
42.8
1C
lass
C35
�3.
70D
0.00
0D
84P
DC
0.99
9K
B48
0c.
1638
5A>
Gp
.(Asp
5462
Gly
)V
OU
SPD
1.00
093
.77
Cla
ssC
65�
6.69
D0.
000
D90
PD
C0.
999
KB
177
c.16
412G
>Tp
.(Arg
5471
Met
)LP
PD1.
000
91.6
4C
lass
C65
�5.
72D
0.00
0D
100
PD
C09
99K
B48
9c.
1649
8C>T
p.(A
rg55
00T
rp)
PPD
1.00
010
1.29
Cla
ssC
65�
7.44
D0.
000
D99
PD
C0.
999
KB
120
c.16
528T>
Gp
.(Tyr
5510
Asp
)LP
PD1.
000
159.
94C
lass
C65
�9.
52D
0.00
0D
96P
DC
0.99
9
Bo
ldva
lues
ind
icat
ede
novo
vari
ants
.aD
etec
ted
toge
ther
wit
ho
ther
vari
ant.
bV
aria
nt
that
fall
sin
the
last
base
of
the
exon
.
P,p
ath
oge
nic
;LP,
like
lyp
ath
oge
nic
;LB
,lik
ely
ben
ign
;VO
US,
vari
ant
of
un
kno
wn
sign
ifica
nce
;PD
,pro
babl
yd
amag
ing;
B,b
enig
n;D
,del
eter
iou
s/d
amag
ing;
N,n
eutr
al;T
,to
lera
ted
;PP,
pro
babl
yp
ath
oge
nic
;PPM
,pro
babl
ep
oly
mo
r-
ph
ism
;PM
,po
lym
orp
his
m;D
C,d
isea
seca
usi
ng.
11Human Molecular Genetics, 2018, Vol. 00, No. 00 |
Downloaded from https://academic.oup.com/hmg/advance-article-abstract/doi/10.1093/hmg/ddy241/5042898by Ospedale Casa Sollievo della Sofferenza useron 27 August 2018
with the WDR5 protein (26), we used transient transfection/co-immunoprecipitation assays to assess whether p.R5340Q alters theability of KMT2D [expressed as an Myc-tagged green fluorescentprotein (GFP) fusion] to physically bind a Flag-tagged WDR5 protein.
As shown in Figure 3B, western blot analysis of FLAG-immunoprecipitates using anti-Myc antibodies showed efficientco-immunoprecipitation of the wild-type KMT2D, but not of theR5340 mutant, with WDR5. This was not due to differences inexpression levels/protein stability between the two KMT2D pro-teins, as documented by western blot analysis of EGFP in the to-tal cell lysates, indicating that this variant completely abrogatesthe interaction with WDR5 (Fig. 3B).
DiscussionHere we report the mutation pattern of KMT2D and KDM6A in acohort of 505 patients clinically diagnosed as KS-affected, anddocument the functional role of a subset of KMT2D missensemutations in impairing the protein enzymatic activity.
Overall, the mutation detection rate for KMT2D and KDM6Ain our cohort (41%) was lower than that reported in a recent sur-vey (22). This does not seem to be due to differences in sensitiv-ity as the same Sanger Sequencing approach was used, andmay be derived from the involvement of other causative genes,epigenetics mechanisms or yet unrecognized promoter or deep
Figure 2. KMT2D missense variants are associated with defective methyltransferase activity and diminished H3K4 methylation. (A) H3K4 Lysine methyltransferase ac-
tivity of mutated FLAG-KMT2D proteins on HeLa nucleosomes. (B) Relative H3K4 methyltransferase activity of semi-purified FLAG-KMT2D proteins on HeLa nucleo-
somes (mean6SD; n¼two independent biological replicates). Data are expressed as fold differences compared with the activity of the wild-type control, arbitrarily set
as 1. Pathogenic variants are indicated with P, likely pathogenic with LP, likely benign with LB, and VOUS with V. Student’s T-test is indicated as *P<0.05; **P<0.01;
***P<0.001. (C) Schematic representation of the epigenetic reporter allele encoding the two halves of GFP separated by a flexible linker region with a histone tail (H3)
and a histone reader (TAF3 PHD) at the N and C termini, respectively (23). Modification of the histone tail by (KMT2D-mediated) trimethylation leads to reconstitution
of the GFP structure and function, which can be measured by fluorescence. (D) The H3K4me3 indicator demonstrated a decreased H3K4me3 activity for the 6 tested
missense variants compared with the WT. Pathogenic variants are indicated with P, likely pathogenic with LP, likely benign with LB, and VOUS with V. Student’s t-test
is indicated as *P<0.001.
12 | Human Molecular Genetics, 2018, Vol. 00, No. 00
Downloaded from https://academic.oup.com/hmg/advance-article-abstract/doi/10.1093/hmg/ddy241/5042898by Ospedale Casa Sollievo della Sofferenza useron 27 August 2018
intronic variants affecting normal splicing. Moreover, we cannotexclude that some patients might have been clinically misdiag-nosed and present conditions partially overlapping with KS. Inview of multiple studies highlighting the clinical and molecularoverlap of KS (27–31), these samples should be interrogated byusing NGS targeted-genes panels focused on components of thehistone methylation machinery that are associated to diseasesin differential diagnosis with KS.
Our screening revealed 58 different missense variants dis-tributed across the entire length of the KMT2D gene. Indeed, 9 ofthe 14 representative mutations that were tested in our studyhad variably reduced histone methylation activity in vitro, in-cluding H3K4 mono-, di- and/or tri-methylation. This effect isconsistent with the known function of KMT2D as a majormono/dimethyltransferase, which is enriched at enhancerregions and is required for enhancer activation during cell dif-ferentiation, as reported in brown adipose tissue and skeletal
muscle development (11,32), or in mature B cells (33,34). Eventhough recent publications focus on the role of KMT2D as majormonomethyltransferase, at least during carcinogenesis (35),KMT2D is also required during oogenesis and early develop-ment for bulk histone H3 lysine 4 trimethylation (36). The abilityto modulate H3K4(me3), at active promoters/transcription startsites is important for the regulation of actively transcribedgenes (14,15), and may reflect in part the formation ofpromoter-enhancer loops. As a matter of fact, six of the nine KSmissense variants with reduced histone methylation activityshowed a flawed H3K4(me3) activity in two different assays,emphasizing the role of KMT2D in modulating histoneH3K4(me3) in vitro.
Of note, all nine functionally defective missense mutationswere localized within the FYRN, WIN and SET domains, andmay contribute to the KS phenotypes by interfering with its en-zymatic activity in a manner analogous to C-terminal
Figure 3. Interaction of KMT2D missense mutants with ASH2L, RbBP5 and WDR5. (A) On the left, immunoblot analysis of KMT2D, ASH2L and RbBP5 in HEK293T cell
line transfected with wild-type KMT2D or KMT2D missense mutants followed by immunoprecipitation with anti-Flag antibody. On the right, relative interaction of the
indicated KMT2D missense mutants with ASH2L (top) and RbBP5 (bottom) (mean6SD, n¼2 independent replicates). Pathogenic variants are indicated with P, likely
pathogenic with LP, likely benign with LB, and VOUS with V. Student’s T-test is indicated as *P<0.05, **P<0.01 and ***P<0.001. (B) On the left, the predicted amino acid
tolerance for KMT2D p.5340: the presence of Methionine (M) is very slightly tolerated. On the right, immunoblot analysis of KMT2D and WDR5 in HEK293 cell line trans-
fected with wild-type WDR5, wild-type C-Ter KMT2D or the p.R5340Q KMT2D missense mutant before (input) and after immunoprecipitation with the anti-Flag anti-
body. EV, empty vector.
13Human Molecular Genetics, 2018, Vol. 00, No. 00 |
Downloaded from https://academic.oup.com/hmg/advance-article-abstract/doi/10.1093/hmg/ddy241/5042898by Ospedale Casa Sollievo della Sofferenza useron 27 August 2018
truncating mutations. In contrast, amino acid changes in theN-terminal domain of the protein had minimal effects on H3K4methylation. Indeed, according to ACMG, most of the aminoacid changes in the N-terminal domain of the protein are classi-fied as Likely Benign or Variants Of Unknown Significance(Table 2), suggesting that they may act through different mech-anisms or represent neutral events.
The entire KMT2D C-terminal domain (aa 4507–5537) isinvolved in the interaction with the protein of the WRAD com-plex (25), albeit the specific amino acid interaction region is notfully characterized yet. Our study showed that missense var-iants localized in the C-terminal domains impaired the interac-tion of KMT2D with other components of its multi-subunitcomplex, including WDR5, RbBP5 and ASH2L, whereas most ofthe variants localized within the PHD 4–6 do not affect the inter-action. The same C-terminal missense variants also showedaltered protein domain structure and free energy interactionlevels, suggesting that structural changes occurring in KMT2Dmay reduce the interaction with the proteins of the WRADcomplex.
These data corroborate the hypothesis that the reducedmethyltransferase activity is a direct consequence of the lack ofmulti-protein WRAD complex formation. Consistently, the mis-sense variant R5340Q, localized within the WIN domain, fullyabrogated the interaction with WDR5, resulting in a significantlyreduced methyltransferase activity.
The classification and interpretation of KMT2D missensevariants is a significant challenge in molecular diagnostics andgenetic counseling. A number of prediction tools have beenimplemented in the last years, mainly as support for NGS data.However, in silico analyses should not be considered as conclu-sive evidence in the assessment of variants of unknown clinicalsignificance. The American College of Medical Genetics andGenomics and the Association for Molecular PathologyStandards and Guidelines for the interpretation of sequencevariants (37) recommended that these predictions should beused as support to additional findings, including functional evi-dence that can prove the pathogenicity of the variants found.Recently, different DNA methylation profiling studies showedthat KS patients present a highly specific and univocalDNA methylation signature that has the potential to be used asa diagnostic method for differentiating samples with patho-genic mutations from those with benign variants, and thereforeenabling the functional assessment of genetic variants ofunknown clinical significance (31,38–41).
Although additional studies will be required to dissect theprecise mechanisms underlying the pathogenic role of the mis-sense variants in KS, our data indicate that a subset of themmay influence the disease status by affecting: (i) the methyl-transferase activity of the protein and (ii) the interaction withthe WRAD complex.
The biochemical approaches presented here may providethe basis for the development of diagnostic assays that could beof support to in silico prediction tools, with the advantage ofaddressing the functional effect of the variant.
ConclusionsThis study expands the number of KMT2D and KDM6A muta-tions that are associated with KS and provides evidence that anumber of naturally occurring missense mutations in KMT2Deffectively impact KMT2D interaction and H3K4 methylation ac-tivity. These data have direct relevance for diagnostic andcounseling purposes.
Materials and MethodsPatients and samples preparation
Our study cohort comprised a total of 505 index patients clini-cally diagnosed as affected by KS (Table 1), including 202 newcases that were referred to our institution between 2014 and2017, and 303 patients from previous studies (19,42–48).
Patients were enrolled after obtaining appropriate informedconsent by the physicians in charge, and approval by the localethics committees. KS patients were included following theinclusions criteria as reported (46). Genomic DNA was extractedfrom fresh and/or frozen peripheral blood leukocytes of patientsand their available family members using an automated DNAextractor and commercial DNA extraction Kits (Qiagen,Germany). Total RNA was extracted from peripheral blood leu-kocytes using TRIzol reagent (ThermoFisher Scientific, USA) andreverse transcribed using the QuantiTect Transcription kit(Qiagen), according to the manufacturer’s instructions.
Sequencing and MLPA of KMT2D and KDM6A
Mutation screening of all 54 coding exons of the KMT2D (MIM#602113, NM_003482.3) gene and 29 coding exons of the KDM6A(MIM #300128, NM_021140.3) gene was performed by PCR ampli-fication and direct sequencing as reported (46). MLPA analysiswas performed as in (47).
In silico analysis of KMT2D and KDM6A variants
The putative causal and functional effect of missense variantswas estimated by using the following in silico prediction tools:Polyphen-2 version 2.2.2 (http://genetics.bwh.harvard.edu/pph;date last accessed 5 July 2018) (49), Align GVGD (http://agvgd.hci.utah.edu; date last accessed 5 July 2018) (50), PROVEAN v1.1(http://provean.jcvi.org/index.php; date last accessed 5 July 2018)(51), SIFT v1.03 (http://sift.jcvi.org/; date last accessed 5 July 2018)(52), UMD-predictor (http://www.umd.be/; date last accessed 5July 2018) (53) and Mutation Taster (http://www.mutationtaster.org/; date last accessed 5 July 2018) (54) using default parameters.Splice-site variants were evaluated for putative alteration of reg-ulatory process at the transcriptional or splicing level withHuman Splice Finder (http://www.umd.be/HSF3/; date lastaccessed 5 July 2018) (55), NNSPLICE (http://www.fruitfly.org/seq_tools/splice.html; date last accessed 5 July 2018) (56) andNetGene2 (http://www.cbs.dtu.dk/services/NetGene2; date lastaccessed 5 July 2018) (57). Frequency variants were checked ondbSNP (http://www.ncbi.nlm.nih.gov/projects/SNP/; date lastaccessed 5 July 2018), 1000 Genomes Project (http://www.internationalgenome.org/; date last accessed 5 July 2018), EVS (http://evs.gs.washington.edu/EVS/; date last accessed 5 July 2018) and ExAC(http://exac.broadinstitute.org/; date last accessed 5 July 2018).
Protein modeling
Three-dimensional models of the ZF-PHD7 (residues 5059–5137), FYRN (residues 5180–5327) and SET (residues 5398–5537)domains of the human KMT2D protein were obtained employ-ing a combination of threading and homology methods. Foreach domain, different models were generated using publiclyavailable web servers (58–61). Final KMT2D model domainswere obtained using the Modeller program and, as templates,the server models (62). A total of 200 unique models were gener-ated by Modeller for each KMT2D domain. Qmean server (63)
14 | Human Molecular Genetics, 2018, Vol. 00, No. 00
Downloaded from https://academic.oup.com/hmg/advance-article-abstract/doi/10.1093/hmg/ddy241/5042898by Ospedale Casa Sollievo della Sofferenza useron 27 August 2018
was used to assess the quality of the predicted models andKoBaMIN web server (64) was used to carry out the protein en-ergy minimization. The crystal structure of the WIN domainbound to WDR5 was obtained from the RCSB Protein Data Bank(PDB code: 3UVK; residues 5337–5347) (65). The effects of mis-sense variants on KMT2D domains structure and stability werepredicted by the Rosetta Backrub server (66) and FoldX program(67). Total charge of domains was calculated using the PROPKAprogram (68) and the electrostatic surface potentials were com-puted using the Adaptive Poisson-Boltzmann Solver (APBS) soft-ware (69). Models’ inspection and model figures were performedusing UCSF Chimera (version 1.11) (70).
Plasmids, cell lines and transfection assays
The pFlag-CMV2 FUSION–KMT2D vector [cDNA spanningthe PHD4–5–6 domains (amino acids 1358–1572)] and ZF-PHD7-FYRN-FYRC-WIN-SET-post SET domains (amino acids4507–5537) was a gift of Professor Min Gyu Lee, Department ofMolecular and Cellular Oncology, The University of Texas (25).We sub-cloned this partial KMT2D sequence into the p3XFlag-CMV14 vector (Sigma) using standard procedures. Fourteen rep-resentative KMT2D variants identified in our KS cohort that har-bor amino acid changes in both N-terminal and C-terminaldomains of the protein, including PHD 4–5–6–7, FYRN, WIN andSET domains, have been selected for functional assays:p.E1391K (Likely Pathogenic), p.M1417V (Likely Benign), p.I1428T(Likely Benign), p.S1476C (VOUS), p.Q1522R (Likely Benign),p.F5034V (Likely pathogenic), p.H5059P (Likely Pathogenic),p.T5098P (VOUS), p.G5189R (VOUS), p.W5217R (Likely patho-genic), p.R5340Q (Likely Pathogenic), p.E5425K (Pathogenic),p.R5471M (Likely Pathogenic) and p.Y5510D (Pathogenic).Expression plasmids harboring patient-derived KMT2D mis-sense variants were generated by site-directed mutagenesisaccording to standard methods.
HEK 293 and 293T cells were cultured in Dulbecco’s ModifiedEagle Medium with 10% FBS, penicillin (100 U/ml) and streptomy-cin (100 lg/ml) (Life Technologies). The KMT2D C-terminal ORFwas assembled into the pcDNA3-Myc-EGFP vector (71) by PCR sitedirected amplification using human cDNA and the pFlag-CMV2FUSION–KMT2D vector as templates respectively. The WDR5 fulllength ORF was assembled into the p3XFlag-CMV14 vector (Sigma)by RT-PCR amplification reactions using cDNA from HEK 293Tcells as template. HEK 293T cells were transiently transfected us-ing the polyethylenimine method, following the published proto-cols (72). Cells were harvested 48 h after transfection and used forprotein extraction and histone methyltransferase (HMT) assay.
In vitro HMT assay and epigenetic reporter allele
Partially purified FLAG-KMT2D wild-type and mutant derivativeproteins were extracted from transfected HEK 293T cells by co-IPbuffer (50 mM Tris, pH 7.5, 250 mM NaCl, 1% Triton X-100, 1 mMEDTA), followed by overnight incubation with EZviewTM Red Anti-FLAG Affinity Gel (Sigma) at 4�C and final elution in BC100 buffer(20 mM Tris, pH 7.5, 10% glycerol, 0.2 mM EDTA, 1% Triton X-100,100 mM NaCl) containing FLAG peptide (Sigma). KMT2D proteinamounts were quantified by Coomassie staining and immunoblotanalysis using mouse monoclonal FLAG antibodies. Enzymatic ac-tivity against HeLa nucleosomes was measured following a pub-lished method (25). Briefly, equal amounts of wild-type or mutantFLAG-KMT2D proteins were incubated at 37�C for 4 h with HeLanucleosomes (Reaction Biology) in KMT buffer (50 mM Tris, pH 8.5,
100 mM KCl, 5 mM MgCl2, 10% glycerol, 4 mM DTT) supplementedwith S-adenosyl methionine (New England BioLabs). Reactionswere stopped by adding equal volumes of 2� Laemmli buffer andheated at 100�C for 5 min before loading onto Tris–glycine 4–20%gradient gels. All assays were performed at least two timesindependently.
The epigenetic reporter allele, a kind gift from Dr HansT. Bjornsson (McKusick-Nathans Institute of Genetic Medicine,Johns Hopkins University, Baltimore), was used as described (23).
Co-immunoprecipitation assay and western blottinganalysis
Co-immunoprecipitations were performed using Dynabeadsmagnetic beads (ThermoFisher Scientific) and EZviewTM RedAnti-FLAG Affinity Gel (Sigma) following the manufacturer’sinstructions. Complexes were analyzed by Western blot using theindicated antibodies. Protein extracts were resolved on NuPAGETris–acetate 3–8% gels (for KMT2D) or Tris–glycine 4–20% gels (forhistone H3) (ThermoFisher Scientific) and transferred to nitrocel-lulose membranes (GE Healthcare) according to the manufac-turer’s instructions. Antibodies used were mouse monoclonalantibody to a-tubulin (clone DM1A, Sigma), rabbit polyclonal anti-H3K4me1 (Abcam), anti-H3K4me2 (Active Motif), anti-H3K4me3(Abcam), rabbit monoclonal anti-Histone H3 (clone D1H2, CellSignaling Technology), mouse monoclonal anti-Flag (Sigma cat#F3165), rabbit monoclonal anti GFP (Santa Cruz), rat monoclonalanti-HA (Roche), mouse monoclonal anti Myc (Roche), rabbit poly-clonal anti Ash2 (Bethyl) and rabbit polyclonal anti RbBP5(Bethyl). Horseradish peroxidase conjugated anti-mouse (SantaCruz) or anti-rabbit (Santa Cruz) antibodies, and the ECL chemilu-minescence system (GE Healthcare) were used for detection.ImageJ software (http://imagej.nih.gov/ij/; date last accessed 5July 2018) was used to quantify band signal intensity. Values areexpressed as fold differences relative to the wild-type proteinsample, set at 1, after normalization for the loading control.
Supplementary MaterialSupplementary Material is available at HMG online.
Acknowledgements
We acknowledge the family that agreed to participate and madethis study possible. We also acknowledge Professor Hans T.Bjornsson for providing Epigenetic reporter allele vector. We aregrateful to the Genomic Disorder Biobank and TelethonNetwork of Genetic Biobanks (Telethon Italy Grant GTB12001G)for biobanking biospecimens.
Conflict of Interest statement: The authors declare no conflicts ofinterest with the exception of G.M. who is a consultant forTakeda Pharmaceutical Company.
FundingN.I.H. Grant RO1-CA172492 (to L.P.), in part; Lymphoma ResearchFoundation post-doctoral fellowship to J.Z. Telethon - Italy(Grant no. GGP13231), Italian Ministry of Health, Jerome LejeuneFoundation, Daunia Plast, Circolo Unione Apricena, FidapaApricena, and Associazione Italiana Sindrome Kabuki (AISK) toG.M. The funders had no role in study design, data collection andanalysis, decision to publish or preparation of the manuscript.
15Human Molecular Genetics, 2018, Vol. 00, No. 00 |
Downloaded from https://academic.oup.com/hmg/advance-article-abstract/doi/10.1093/hmg/ddy241/5042898by Ospedale Casa Sollievo della Sofferenza useron 27 August 2018
References1. Kuroki, Y., Suzuki, Y., Chyo, H., Hata, A. and Matsui, I. (1981)
A new malformation syndrome of long palpebral fissures,large ears, depressed nasal tip, and skeletal anomalies asso-ciated with postnatal dwarfism and mental retardation. J.Pediatr., 99, 570–573.
2. Niikawa, N., Matsuura, N., Fukushima, Y., Ohsawa, T. andKajii, T. (1981) Kabuki make-up syndrome: a syndrome ofmental retardation, unusual facies, large and protrudingears, and postnatal growth deficiency. J. Pediatr., 99, 565–569.
3. Cheon, C.K. and Ko, J.M. (2015) Kabuki syndrome: clinicaland molecular characteristics. Korean J. Pediatr., 58, 317–324.
4. Ng, S.B., Bigham, A.W., Buckingham, K.J., Hannibal, M.C.,McMillin, M.J., Gildersleeve, H.I., Beck, A.E., Tabor, H.K.,Cooper, G.M., Mefford, H.C. et al. (2010) Exome sequencingidentifies MLL2 mutations as a cause of Kabuki syndrome.Nat. Genet., 42, 790–793.
5. Shilatifard, A. (2012) The COMPASS family of histone H3K4methylases: mechanisms of regulation in development anddisease pathogenesis. Annu. Rev. Biochem., 81, 65–95.
6. Kouzarides, T. (2007) Chromatin modifications and theirfunction. Cell, 128, 693–705.
7. Li, B., Carey, M. and Workman, J.L. (2007) The role of chroma-tin during transcription. Cell, 128, 707–719.
8. Ang, S.Y., Uebersohn, A., Spencer, C.I., Huang, Y., Lee, J.E.,Ge, K. and Bruneau, B.G. (2016) KMT2D regulates specific pro-grams in heart development via histone H3 lysine 4 di-me-thylation. Development, 143, 810–821.
9. Herz, H.M., Mohan, M., Garruss, A.S., Liang, K., Takahashi,Y.H., Mickey, K., Voets, O., Verrijzer, C.P. and Shilatifard, A.(2012) Enhancer-associated H3K4 monomethylation byTrithorax-related, the Drosophila homolog of mammalianMll3/Mll4. Genes. Dev., 26, 2604–2620.
10. Hu, D., Gao, X., Morgan, M.A., Herz, H.M., Smith, E.R. andShilatifard, A. (2013) The MLL3/MLL4 branches of theCOMPASS family function as major histone H3K4 monome-thylases at enhancers. Mol. Cell. Biol., 33, 4745–4754.
11. Lee, J.E., Wang, C., Xu, S., Cho, Y.W., Wang, L., Feng, X.,Baldridge, A., Sartorelli, V., Zhuang, L., Peng, W. et al. (2013)H3K4 mono- and di-methyltransferase MLL4 is required forenhancer activation during cell differentiation. Elife(Cambridge), 2, e01503.
12. Guo, C., Chen, L.H., Huang, Y., Chang, C.C., Wang, P., Pirozzi,C.J., Qin, X., Bao, X., Greer, P.K., McLendon, R.E. et al. (2013)KMT2D maintains neoplastic cell proliferation and globalhistone H3 lysine 4 monomethylation. Oncotarget, 4,2144–2153.
13. Kaikkonen, M.U., Spann, N.J., Heinz, S., Romanoski, C.E.,Allison, K.A., Stender, J.D., Chun, H.B., Tough, D.F., Prinjha,R.K., Benner, C. et al. (2013) Remodeling of the enhancerlandscape during macrophage activation is coupled to en-hancer transcription. Mol. Cell, 51, 310–325.
14. Heintzman, N.D., Hon, G.C., Hawkins, R.D., Kheradpour, P.,Stark, A., Harp, L.F., Ye, Z., Lee, L.K., Stuart, R.K., Ching, C.W.et al. (2009) Histone modifications at human enhancers re-flect global cell-type-specific gene expression. Nature, 459,108–112.
15. Denissov, S., Hofemeister, H., Marks, H., Kranz, A., Ciotta, G.,Singh, S., Anastassiadis, K., Stunnenberg, H.G. and Stewart,A.F. (2014) Mll2 is required for H3K4 trimethylation on biva-lent promoters in embryonic stem cells, whereas Mll1 is re-dundant. Development, 141, 526–537.
16. Issaeva, I., Zonis, Y., Rozovskaia, T., Orlovsky, K., Croce, C.M.,Nakamura, T., Mazo, A., Eisenbach, L. and Canaani, E. (2007)Knockdown of ALR (MLL2) reveals ALR target genes andleads to alterations in cell adhesion and growth. Mol. Cell.Biol., 27, 1889–1903.
17. Lederer, D., Grisart, B., Digilio, M.C., Benoit, V., Crespin, M.,Ghariani, S.C., Maystadt, I., Dallapiccola, B. and Verellen-Dumoulin, C. (2012) Deletion of KDM6A, a histone demethy-lase interacting with MLL2, in three patients with Kabukisyndrome. Am. J. Hum. Genet., 90, 119–124.
18. Miyake, N., Mizuno, S., Okamoto, N., Ohashi, H., Shiina, M.,Ogata, K., Tsurusaki, Y., Nakashima, M., Saitsu, H., Niikawa,N. et al. (2013) KDM6A point mutations cause Kabuki syn-drome. Hum. Mutat., 34, 108–110.
19. Micale, L., Augello, B., Maffeo, C., Selicorni, A., Zucchetti, F.,Fusco, C., De Nittis, P., Pellico, M.T., Mandriani, B., Fischetto,R. et al. (2014) Molecular analysis, pathogenic mechanisms,and readthrough therapy on a large cohort of Kabuki syn-drome patients. Hum. Mutat., 35, 841–850.
20. Hong, S., Cho, Y.W., Yu, L.R., Yu, H., Veenstra, T.D. and Ge, K.(2007) Identification of JmjC domain-containing UTX andJMJD3 as histone H3 lysine 27 demethylases. Proc. Natl. Acad.Sci. U.S.A., 104, 18439–18444.
21. Lan, F., Bayliss, P.E., Rinn, J.L., Whetstine, J.R., Wang, J.K.,Chen, S., Iwase, S., Alpatov, R., Issaeva, I., Canaani, E. et al.(2007) A histone H3 lysine 27 demethylase regulates animalposterior development. Nature, 449, 689–694.
22. Bogershausen, N., Gatinois, V., Riehmer, V., Kayserili, H.,Becker, J., Thoenes, M., Simsek-Kiper, P.O., Barat-Houari, M.,Elcioglu, N.H., Wieczorek, D. et al. (2016) Mutation update forKabuki syndrome genes KMT2D and KDM6A and further de-lineation of X-linked Kabuki syndrome subtype 2. Hum.Mutat, 37, 847–864.
23. Bjornsson, H.T., Benjamin, J.S., Zhang, L., Weissman, J.,Gerber, E.E., Chen, Y.C., Vaurio, R.G., Potter, M.C., Hansen,K.D. and Dietz, H.C. (2014) Histone deacetylase inhibitionrescues structural and functional brain deficits in a mousemodel of Kabuki syndrome. Sci. Transl. Med., 6, 256ra135.
24. Kim, J.H., Sharma, A., Dhar, S.S., Lee, S.H., Gu, B., Chan, C.H.,Lin, H.K. and Lee, M.G. (2014) UTX and MLL4 coordinatelyregulate transcriptional programs for cell proliferation andinvasiveness in breast cancer cells. Cancer Res., 74,1705–1717.
25. Dhar, S.S., Lee, S.H., Kan, P.Y., Voigt, P., Ma, L., Shi, X.,Reinberg, D. and Lee, M.G. (2012) Trans-tail regulation ofMLL4-catalyzed H3K4 methylation by H4R3 symmetricdimethylation is mediated by a tandem PHD of MLL4. GenesDev., 26, 2749–2762.
26. Shinsky, S.A., Hu, M., Vought, V.E., Ng, S.B., Bamshad, M.J.,Shendure, J. and Cosgrove, M.S. (2014) A non-active-site SETdomain surface crucial for the interaction of MLL1 and theRbBP5/Ash2L heterodimer within MLL family core com-plexes. J. Mol. Biol., 426, 2283–2299.
27. Au, P.Y.B., You, J., Caluseriu, O., Schwartzentruber, J.,Majewski, J., Bernier, F.P., Ferguson, M., Valle, D.,Parboosingh, J.S., Sobreira, N. et al. (2015) GeneMatcher aidsin the identification of a new malformation syndrome withintellectual disability, unique facial dysmorphisms, andskeletal and connective tissue abnormalities caused by denovo variants in HNRNPK. Hum. Mutat., 36, 1009–1014.
28. Lange, L., Pagnamenta, A.T., Lise, S., Clasper, S., Stewart, H.,Akha, E.S., Quaghebeur, G., Knight, S.J., Keays, D.A., Taylor,J.C. et al. (2016) A de novo frameshift in HNRNPK causing a
16 | Human Molecular Genetics, 2018, Vol. 00, No. 00
Downloaded from https://academic.oup.com/hmg/advance-article-abstract/doi/10.1093/hmg/ddy241/5042898by Ospedale Casa Sollievo della Sofferenza useron 27 August 2018
Kabuki-like syndrome with nodular heterotopia. Clin. Genet.,90, 258–262.
29. Yuan, B., Pehlivan, D., Karaca, E., Patel, N., Charng, W.L.,Gambin, T., Gonzaga-Jauregui, C., Sutton, V.R., Yesil, G.,Bozdogan, S.T. et al. (2015) Global transcriptional disturban-ces underlie Cornelia de Lange syndrome and related phe-notypes. J. Clin. Invest., 125, 636–651.
30. Parenti, I., Teresa-Rodrigo, M.E., Pozojevic, J., Ruiz Gil, S.,Bader, I., Braunholz, D., Bramswig, N.C., Gervasini, C.,Larizza, L., Pfeiffer, L. et al. (2017) Mutations in chromatinregulators functionally link Cornelia de Lange syndromeand clinically overlapping phenotypes. Hum. Genet., 136,307–320.
31. Sobreira, N., Brucato, M., Zhang, L., Ladd-Acosta, C., Ongaco,C., Romm, J., Doheny, K.F., Mingroni-Netto, R.C., Bertola, D.,Kim, C.A. et al. (2017) Patients with a Kabuki syndrome phe-notype demonstrate DNA methylation abnormalities. Eur. J.Hum. Genet, 25, 1335–1344.
32. Guenther, M.G., Jenner, R.G., Chevalier, B., Nakamura, T.,Croce, C.M., Canaani, E. and Young, R.A. (2005) Global andHox-specific roles for the MLL1 methyltransferase. Proc. Natl.Acad. Sci. U.S.A., 102, 8603–8608.
33. Zhang, J., Dominguez-Sola, D., Hussein, S., Lee, J.E., Holmes,A.B., Bansal, M., Vlasevska, S., Mo, T., Tang, H., Basso, K. et al.(2015) Disruption of KMT2D perturbs germinal center B celldevelopment and promotes lymphomagenesis. Nat. Med.,21, 1190–1198.
34. Lindsley, A.W., Saal, H.M., Burrow, T.A., Hopkin, R.J.,Shchelochkov, O., Khandelwal, P., Xie, C., Bleesing, J.,Filipovich, L., Risma, K. et al. (2016) Defects of B-cell terminaldifferentiation in patients with type-1 Kabuki syndrome. J.Allergy Clin. Immunol., 137, 179–187.
35. Lv, S., Ji, L., Chen, B., Liu, S., Lei, C., Liu, X., Qi, X., Wang, Y.,Lai-Han Leung, E., Wang, H. et al. (2018) Histone methyl-transferase KMT2D sustains prostate carcinogenesis andmetastasis via epigenetically activating LIFR and KLF4.Oncogene, 37, 1354–1368.
36. Andreu-Vieyra, C.V., Chen, R., Agno, J.E., Glaser, S.,Anastassiadis, K., Stewart, A.F. and Matzuk, M.M. (2010)MLL2 is required in oocytes for bulk histone 3 lysine 4 trime-thylation and transcriptional silencing. PLoS Biol., 8,e1000453.
37. Richards, S., Aziz, N., Bale, S., Bick, D., Das, S., Gastier-Foster,J., Grody, W.W., Hegde, M., Lyon, E., Spector, E. et al. (2015)Standards and guidelines for the interpretation of sequencevariants: a joint consensus recommendation of theAmerican College of Medical Genetics and Genomics andthe Association for Molecular Pathology. Genet. Med., 17,405–424.
38. Butcher, D.T., Cytrynbaum, C., Turinsky, A.L., Siu, M.T.,Inbar-Feigenberg, M., Mendoza-Londono, R., Chitayat, D.,Walker, S., Machado, J., Caluseriu, O. et al. (2017) CHARGEand Kabuki syndromes: gene-specific DNA methylation sig-natures identify epigenetic mechanisms linking these clini-cally overlapping conditions. Am. J. Hum. Genet., 100,773–788.
39. Choufani, S., Cytrynbaum, C., Chung, B.H., Turinsky, A.L.,Grafodatskaya, D., Chen, Y.A., Cohen, A.S., Dupuis, L.,Butcher, D.T., Siu, M.T. et al. (2015) NSD1 mutations generatea genome-wide DNA methylation signature. Nat. Commun.,6, 10207.
40. Grafodatskaya, D., Chung, B.H., Butcher, D.T., Turinsky, A.L.,Goodman, S.J., Choufani, S., Chen, Y.A., Lou, Y., Zhao, C.,Rajendram, R. et al. (2013) Multilocus loss of DNA
methylation in individuals with mutations in the histone H3lysine 4 demethylase KDM5C. BMC Med. Genomics, 6, 1.
41. Kernohan, K.D., Cigana Schenkel, L., Huang, L., Smith, A.,Pare, G., Ainsworth, P., Boycott, K.M., Warman-Chardon, J.and Sadikovic, B. (2016) Identification of a methylation pro-file for DNMT1-associated autosomal dominant cerebellarataxia, deafness, and narcolepsy. Clin. Epigenetics, 8, 91.
42. Andersen, M.S., Menazzi, S., Brun, P., Cocah, C., Merla, G. andSolari, A. (2014) Clinical diagnosis of Kabuki syndrome: phe-notype and associated abnormalities in two new cases. Arch.Argent. Pediatr., 112, 26–32.
43. Cappuccio, G., Rossi, A., Fontana, P., Acampora, E., Avolio, V.,Merla, G., Zelante, L., Secinaro, A., Andria, G. and Melis, D.(2014) Bronchial isomerism in a Kabuki syndrome patientwith a novel mutation in MLL2 gene. BMC Med. Genet., 15, 15,
44. Ejarque, I., Uliana, V., Forzano, F., Marciano, C., Merla, G.,Zelante, L., Di Maria, E. and Faravelli, F. (2011) Is Hardikarsyndrome distinct from Kabuki (Niikawa-Kuroki) syndrome?Clin. Genet., 80, 493–496.
45. Makrythanasis, P., van Bon, B.W., Steehouwer, M.,Rodriguez-Santiago, B., Simpson, M., Dias, P., Anderlid, B.M.,Arts, P., Bhat, M., Augello, B. et al. (2013) MLL2 mutation de-tection in 86 patients with Kabuki syndrome: agenotype-phenotype study. Clin. Genet., 84, 539–545.
46. Micale, L., Augello, B., Fusco, C., Selicorni, A., Loviglio, M.N.,Silengo, M.C., Reymond, A., Gumiero, B., Zucchetti, F.,D’Addetta, E.V. et al. (2011) Mutation spectrum of MLL2 in a co-hort of Kabuki syndrome patients. Orphanet J. Rare Dis., 6, 38.
47. Priolo, M., Micale, L., Augello, B., Fusco, C., Zucchetti, F.,Prontera, P., Paduano, V., Biamino, E., Selicorni, A., Mammi,C. et al. (2012) Absence of deletion and duplication of MLL2and KDM6A genes in a large cohort of patients with Kabukisyndrome. Mol. Genet. Metab., 107, 627–629.
48. Ratbi, I., Fejjal, N., Micale, L., Augello, B., Fusco, C., Lyahyai, J.,Merla, G. and Sefiani, A. (2013) Report of the first clinical caseof a Moroccan Kabuki patient with a novel MLL2 mutation.Mol. Syndromol., 4, 152–156.
49. Adzhubei, I.A., Schmidt, S., Peshkin, L., Ramensky, V.E.,Gerasimova, A., Bork, P., Kondrashov, A.S. and Sunyaev, S.R.(2010) A method and server for predicting damaging mis-sense mutations. Nat. Methods, 7, 248–249.
50. Tavtigian, S.V., Deffenbaugh, A.M., Yin, L., Judkins, T., Scholl,T., Samollow, P.B., de Silva, D., Zharkikh, A. and Thomas, A.(2005) Comprehensive statistical study of 452 BRCA1 mis-sense substitutions with classification of eight recurrentsubstitutions as neutral. J. Med. Genet., 43, 295–305.
51. Choi, Y., Sims, G.E., Murphy, S., Miller, J.R. and Chan, A.P.(2012) Predicting the functional effect of amino acid substi-tutions and indels. PLoS One, 7, e46688.
52. Kumar, P., Henikoff, S. and Ng, P.C. (2009) Predicting theeffects of coding non-synonymous variants on protein func-tion using the SIFT algorithm. Nat. Protoc., 4, 1073–1081.
53. Frederic, M.Y., Lalande, M., Boileau, C., Hamroun, D.,Claustres, M., Beroud, C. and Collod-Beroud, G. (2009)UMD-predictor, a new prediction tool for nucleotide substi-tution pathogenicity – application to four genes: fBN1, FBN2,TGFBR1, and TGFBR2. Hum. Mutat., 30, 952–959.
54. Schwarz, J.M., Cooper, D.N., Schuelke, M. and Seelow, D.(2014) MutationTaster2: mutation prediction for thedeep-sequencing age. Nat. Methods, 11, 361–362.
55. Desmet, F.O., Hamroun, D., Lalande, M., Collod-Beroud, G.,Claustres, M. and Beroud, C. (2009) Human splicing finder:an online bioinformatics tool to predict splicing signals.Nucleic Acids Res., 37, e67.
17Human Molecular Genetics, 2018, Vol. 00, No. 00 |
Downloaded from https://academic.oup.com/hmg/advance-article-abstract/doi/10.1093/hmg/ddy241/5042898by Ospedale Casa Sollievo della Sofferenza useron 27 August 2018
56. Reese, M.G., Eeckman, F.H., Kulp, D. and Haussler, D. (1997)Improved splice site detection in Genie. J. Comput. Biol., 4,311–323.
57. Hebsgaard, S.M., Korning, P.G., Tolstrup, N., Engelbrecht,J., Rouze, P. and Brunak, S. (1996) Splice site prediction inArabidopsis thaliana pre-mRNA by combining local andglobal sequence information. Nucleic Acids Res., 24,3439–3452.
58. Zhang, Y. (2008) I-TASSER server for protein 3D structureprediction. BMC Bioinformatics, 9, 40.
59. Kelley, L.A., Mezulis, S., Yates, C.M., Wass, M.N. andSternberg, M.J. (2015) The Phyre2 web portal for proteinmodeling, prediction and analysis. Nat. Protoc., 10, 845–858.
60. Li, J., Deng, X., Eickholt, J. and Cheng, J. (2013) Designing andbenchmarking the MULTICOM protein structure predictionsystem. BMC Struct. Biol., 13, 2.
61. Kallberg, M., Wang, H., Wang, S., Peng, J., Wang, Z., Lu, H.and Xu, J. (2012) Template-based protein structure modelingusing the RaptorX web server. Nat. Protoc., 7, 1511–1522.
62. Sali, A. and Blundell, T.L. (1993) Comparative protein model-ling by satisfaction of spatial restraints. J. Mol. Biol., 234,779–815.
63. Benkert, P., Kunzli, M. and Schwede, T. (2009) QMEAN serverfor protein model quality estimation. Nucleic Acids Res., 37,W510–W514.
64. Rodrigues, J.P., Levitt, M. and Chopra, G. (2012) KoBaMIN: aknowledge-based minimization web server for proteinstructure refinement. Nucleic Acids Res., 40, W323–W328.
65. Zhang, P., Lee, H., Brunzelle, J.S. and Couture, J.F. (2012) Theplasticity of WDR5 peptide-binding cleft enables the bindingof the SET1 family of histone methyltransferases. NucleicAcids Res., 40, 4237–4246.
66. Lauck, F., Smith, C.A., Friedland, G.F., Humphris, E.L. andKortemme, T. (2010) RosettaBackrub—a web server for flexi-ble backbone protein structure modeling and design. NucleicAcids Res., 38, W569–W575.
67. Guerois, R., Nielsen, J.E. and Serrano, L. (2002) Predictingchanges in the stability of proteins and protein com-plexes: a study of more than 1000 mutations. J. Mol. Biol.,320, 369–387.
68. Rostkowski, M., Olsson, M.H., Sondergaard, C.R. and Jensen,J.H. (2011) Graphical analysis of pH-dependent properties ofproteins predicted using PROPKA. BMC Struct. Biol., 11, 6.
69. Baker, N.A., Sept, D., Joseph, S., Holst, M.J. and McCammon,J.A. (2001) Electrostatics of nanosystems: application tomicrotubules and the ribosome. Proc. Natl. Acad. Sci. U.S.A.,98, 10037–10041.
70. Pettersen, E.F., Goddard, T.D., Huang, C.C., Couch, G.S.,Greenblatt, D.M., Meng, E.C. and Ferrin, T.E. (2004) UCSFChimera–a visualization system for exploratory researchand analysis. J. Comput. Chem., 25, 1605–1612.
71. Reymond, A., Meroni, G., Fantozzi, A., Merla, G., Cairo, S.,Luzi, L., Riganelli, D., Zanaria, E., Messali, S., Cainarca, S. et al.(2001) The tripartite motif family identifies cell compart-ments. EMBO J., 20, 2140–2151.
72. Thomas, M., Lu, J.J., Ge, Q., Zhang, C., Chen, J. and Klibanov,A.M. (2005) Full deacylation of polyethylenimine dramati-cally boosts its gene delivery efficiency and specificity tomouse lung. Proc. Natl. Acad. Sci. U.S.A., 102, 5679–5684.
73. Pasqualucci, L., Trifonov, V., Fabbri, G., Ma, J., Rossi, D.,Chiarenza, A., Wells, V.A., Grunn, A., Messina, M., Elliot, O.
et al. (2011) Analysis of the coding genome of diffuse largeB-cell lymphoma. Nat. Genet., 43, 830–837.
74. Dentici, M.L., Di Pede, A., Lepri, F.R., Gnazzo, M., Lombardi,M.H., Auriti, C., Petrocchi, S., Pisaneschi, E., Bellacchio, E.,Capolino, R. et al. (2015) Kabuki syndrome: clinical and mo-lecular diagnosis in the first year of life. Arch. Dis. Child, 100,158–164.
75. Hannibal, M.C., Buckingham, K.J., Ng, S.B., Ming, J.E., Beck,A.E., McMillin, M.J., Gildersleeve, H.I., Bigham, A.W., Tabor,H.K., Mefford, H.C. et al. (2011) Spectrum of MLL2 (ALR) muta-tions in 110 cases of Kabuki syndrome. Am. J. Med. Genet. A,155A, 1511–1516.
76. Banka, S., Veeramachaneni, R., Reardon, W., Howard, E.,Bunstone, S., Ragge, N., Parker, M.J., Crow, Y.J., Kerr, B.,Kingston, H. et al. (2012) How genetically heterogeneous isKabuki syndrome? MLL2 testing in 116 patients, review andanalyses of mutation and phenotypic spectrum. Eur. J. Hum.Genet., 20, 381–388.
77. Paulussen, A.D., Stegmann, A.P., Blok, M.J., Tserpelis, D.,Posma-Velter, C., Detisch, Y., Smeets, E.E., Wagemans, A.,Schrander, J.J., van den Boogaard, M.J. et al. (2011) MLL2 mu-tation spectrum in 45 patients with Kabuki syndrome. HumMutat., 32, E2018–E2025.
78. Lin, J.L., Lee, W.I., Huang, J.L., Chen, P.K., Chan, K.C., Lo, L.J.,You, Y.J., Shih, Y.F., Tseng, T.Y. and Wu, M.C. (2015)Immunologic assessment and KMT2D mutation detection inKabuki syndrome. Clin. Genet., 88, 255–260.
79. Van Laarhoven, P.M., Neitzel, L.R., Quintana, A.M., Geiger,E.A., Zackai, E.H., Clouthier, D.E., Artinger, K.B., Ming, J.E. andShaikh, T.H. (2015) Kabuki syndrome genes KMT2D andKDM6A: functional analyses demonstrate critical roles incraniofacial, heart and brain development. Hum. Mol. Genet.,24, 4443–4453.
80. Cheon, C.K., Sohn, Y.B., Ko, J.M., Lee, Y.J., Song, J.S., Moon,J.W., Yang, B.K., Ha, I.S., Bae, E.J., Jin, H.S. et al. (2014)Identification of KMT2D and KDM6A mutations by exomesequencing in Korean patients with Kabuki syndrome.J. Hum. Genet., 59, 321–325.
81. Morgan, A.T., Mei, C., Da Costa, A., Fifer, J., Lederer, D.,Benoit, V., McMillin, M.J., Buckingham, K.J., Bamshad, M.J.,Pope, K. et al. (2015) Speech and language in a genotyped co-hort of individuals with Kabuki syndrome. Am. J. Med. Genet.A, 167, 1483–1492.
82. Kokitsu-Nakata, N.M., Petrin, A.L., Heard, J.P., Vendramini-Pittoli, S., Henkle, L.E., dos Santos, D.V., Murray, J.C. andRichieri-Costa, A. (2012) Analysis of MLL2 gene in the firstBrazilian family with Kabuki syndrome. Am. J. Med. Genet. A,158A, 2003–2008.
83. Liu, S., Hong, X., Shen, C., Shi, Q., Wang, J., Xiong, F. and Qiu,Z. (2015) Kabuki syndrome: a Chinese case series and sys-tematic review of the spectrum of mutations. BMC Med.Genet., 16, 26.
84. Banka, S., Lederer, D., Benoit, V., Jenkins, E., Howard, E.,Bunstone, S., Kerr, B., McKee, S., Chris Lloyd, I., Shears, D.et al. (2015) Novel KDM6A (UTX) mutations and a clinical andmolecular review of the X-linked Kabuki syndrome (KS2).Clin. Genet, 87, 252–258.
85. Lederer, D., Shears, D., Benoit, V., Verellen-Dumoulin, C. andMaystadt, I. (2014) A three generation X-linked family withKabuki syndrome phenotype and a frameshift mutation inKDM6A. Am. J. Med. Genet. A, 164, 1289–1292.
18 | Human Molecular Genetics, 2018, Vol. 00, No. 00
Downloaded from https://academic.oup.com/hmg/advance-article-abstract/doi/10.1093/hmg/ddy241/5042898by Ospedale Casa Sollievo della Sofferenza useron 27 August 2018