Close
Logo
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected 
Invert selection
Deselect all
Deselect all
 Click here to refresh results
 Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Genetics and the environment: evaluating the role of noncoding RNA in autism spectrum disorder
(USC Thesis Other) 

Genetics and the environment: evaluating the role of noncoding RNA in autism spectrum disorder

doctype icon
play button
PDF
 Download
 Share
 Open document
 Flip pages
 More
 Summarize
 Download a page range
 Download transcript
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content i
 

 
GENETICS
 AND
 THE
 ENVIRONMENT:
 
EVALUATING
 THE
 ROLE
 OF
 NONCODING
 RNA
 IN
 AUTISM
 SPECTRUM
 DISORDER
 

 

 
By
 
Jessica
 Jolynn
 DeWitt
 

 

 
______________________________________________________________________________
 

 

 
A
 Dissertation
 Presented
 to
 the
 
FACULTY
 OF
 THE
 USC
 GRADUATE
 SCHOOL
 
UNIVERSITY
 OF
 SOUTHERN
 CALIFORNIA
 
In
 Partial
 Fulfillment
 of
 the
 
Requirement
 for
 the
 Degree
 
DOCTOR
 OF
 PHILOSOPHY
 
(DEVELOPMENT,
 STEM
 CELL,
 AND
 REGENERATIVE
 MEDICINE)
 

 

 
August
 2017
 

 

 

 

 

   
   
   
   
   
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
   
 

   
 
ii
 

 
DEDICATION
 

 

 
For
 my
 sister,
 who
 was
 always
 so
 proud
 of
 me
 and
 my
 work
 and
 whose
 impact
 on
 my
 life
 
sparked
 my
 interest
 in
 complex
 neurodevelopmental
 disorders,
 and
 her
 son,
 whose
 generation
 
this
 work
 will
 hopefully
 effect.
 
 

 

 

 

 

 

 

 

 

 

 

 

   
 
iii
 

 
ACKNOWLEDGEMENTS
 

 
First
 and
 foremost,
 I
 have
 to
 thank
 my
 mentor,
 Dr.
 Dan
 Campbell,
 for
 developing
 my
 
scientific
 writing,
 pushing
 me
 to
 think
 harder,
 and
 honing
 my
 presentation
 skills.
 I
 was
 Dan’s
 
first
 grad
 student,
 and
 as
 I
 grew
 as
 a
 scientist,
 he
 grew
 as
 a
 mentor.
 In
 the
 beginning,
 he
 
provided
 a
 lot
 of
 hand-­‐holding
 as
 I
 got
 my
 bearings
 in
 the
 new
 world
 of
 graduate
 school.
 He
 was
 
always
 gentle
 and
 kind
 with
 all
 the
 mistakes
 I
 made.
 It
 seemed
 very
 little
 could
 ruffle
 his
 
feathers.
 As
 the
 years
 progressed,
 my
 mistakes
 were
 fewer
 and
 he
 gave
 me
 more
 
independence
 as
 I
 figured
 things
 out
 faster
 and
 tried
 new
 approaches.
 He
 never
 looked
 over
 my
 
shoulder,
 but
 trusted
 me
 to
 do
 the
 work
 and
 come
 to
 him
 if
 I
 needed
 advice.
 He
 showed
 me
 
what
 a
 mentor
 should
 be
 and
 I’m
 very
 thankful
 for
 that
 mentorship.
 
 

  I
 also
 want
 to
 thank
 my
 qualifying
 and
 dissertation
 committees:
 Dr.
 Gerry
 Coatzee,
 Dr.
 
Karen
 Chang,
 Dr.
 Ruchi
 Bajpai,
 Dr.
 Ruth
 Wood,
 and
 Dr.
 Justin
 Ichida.
 Dr.
 Coatzee
 provided
 tough
 
feedback
 during
 my
 qualifying
 exams
 and
 pushed
 me
 to
 really
 think
 hard
 about
 what
 I
 was
 
doing
 and
 why
 I
 was
 doing
 it.
 More
 importantly,
 he
 forced
 me
 to
 justify
 my
 research
 methods
 
as
 opposed
 to
 other
 methods.
 I
 learned
 that,
 as
 scientists,
 we
 must
 understand
 why
 we
 do
 
what
 we
 do,
 as
 well
 as
 why
 we
 are
 not
 doing
 something
 else.
 I
 thank
 Dr.
 Chang
 for
 serving
 as
 
my
 qualifying
 committee
 chair.
 Dr.
 Bajpai
 served
 as
 my
 dissertation
 committee
 chair
 and
 
inspired
 me
 multiple
 times
 when
 I
 had
 existential
 crises.
 Every
 conversation
 I
 had
 with
 her
 left
 
me
 encouraged
 and
 inspired
 to
 dig
 deeper
 and
 make
 scientific
 discoveries.
 She
 provided
 quality
 
feedback
 and
 always
 believed
 that
 I
 could
 and
 would
 be
 successful.
 Dr.
 Wood,
 not
 only
 was
 the
 
outside
 member
 of
 my
 committee,
 but
 also
 allowed
 me
 to
 rotate
 in
 her
 lab
 at
 the
 beginning
 of
 
my
 time
 in
 graduate
 school.
 Her
 false
 modesty
 to
 hide
 her
 brilliance
 taught
 me
 to
 think
 
iv
 

 
carefully
 about
 how
 I
 explained
 my
 research
 to
 someone
 not
 in
 my
 field,
 even
 though
 I’m
 sure
 
she
 knew
 exactly
 what
 I
 was
 talking
 about.
 Lastly,
 I
 thank
 Dr.
 Ichida
 for
 his
 helpful
 insights.
 
 

 
  I
 would
 like
 to
 thank
 Nicole
 Campbell,
 for
 teaching
 me
 laboratory
 techniques
 and
 
helping
 me
 with
 experiments.
 I
 especially
 appreciate
 her
 patience
 and
 mentoring
 guidance
 in
 
non-­‐science
 topics.
 
 
 

  I
 also
 need
 to
 thank
 the
 various
 undergraduates
 who
 came
 through
 our
 lab,
 for
 helping
 
with
 experiments
 and
 making
 grad
 school
 more
 entertaining.
 I
 especially
 owe
 gratitude
 to
 
Sarah
 Danehower,
 who
 really
 got
 me
 through
 second
 year
 when
 I
 had
 familial
 crises.
 
 

  I
 have
 to
 thank
 my
 undergraduate
 mentors,
 Dr.
 Joshua
 Morris,
 who
 sparked
 my
 original
 
interest
 in
 genetics,
 Dr.
 Kevin
 Huang,
 Ryon
 Maland,
 and
 all
 the
 faculty
 and
 students
 at
 my
 alma
 
mater,
 Azusa
 Pacific
 University,
 who
 helped
 shape
 who
 I
 am
 as
 a
 scientist,
 a
 researcher,
 and
 a
 
person.
 
 

  I
 want
 to
 thank
 my
 friends
 and
 family
 members
 for
 their
 support
 throughout
 my
 
graduate
 career.
 Rafael
 Martinez
 held
 me
 when
 I
 cried.
 My
 grandmas,
 Rosie
 and
 Glenda,
 
constantly
 expressed
 enthusiasm
 and
 interest
 in
 what
 I
 was
 doing
 and
 kept
 me
 from
 losing
 sight
 
of
 how
 lucky
 I
 have
 been
 to
 pursue
 something
 unknown
 and
 completely
 new.
 Dr.
 Matthew
 
Barnes
 helped
 me
 integrate
 my
 scientific
 work
 with
 my
 faith.
 Kerrie
 Hawkes
 and
 Dr.
 Jennifer
 
Martin
 remind
 me
 to
 breathe
 and
 provide
 female
 mentorship.
 Thank
 you
 to
 all
 my
 friends
 who
 
kept
 asking
 me
 when
 I
 was
 going
 to
 be
 done
 and
 telling
 me
 to
 hurry
 up
 and
 finish.
 
Lastly,
 I
 want
 to
 thank
 my
 graduate
 school
 friends
 for
 giving
 each
 other
 reality
 checks
 and
 
reminding
 each
 other
 that
 we
 are
 not
 imposters:
 we
 deserve
 to
 be
 here.
 

 
v
 

 
Kimberly
 Babos,
 Stephanie
 Kuwahara,
 Dr.
 Trisha
 Staab,
 Dr.
 Jillian
 Shaw,
 Matthew
 Wong,
 Aaron
 
Wolfe,
 Julia
 Taylor,
 Marisa
 Jones,
 Dominic
 Shayler,
 Dr.
 Coco
 Dong,
 Mary
 Donhoffner,
 Dr.
 Nick
 
Goeden,
 Dr.
 Tara
 Kerin,
 Dr.
 Mallory
 Indi
 Bradley,
 Lisa
 Nguyen,
 Dr.
 Patrick
 Hecht,
 Brent
 
Wilkinson,
 and
 Maria
 Soria.
 
 

 

 

 

 

 

 

 

 

 

   
 
vi
 

 
TABLE
 OF
 CONTENTS
 

 

 

   
 
DEDICATION
  ii
 
 
ACKNOWLEDGEMENTS
  iii
 
LISTS
 OF
 TABLES
  xi
 
LISTS
 OF
 FIGURES
  xiii
 
ABSTRACT
  15
 
CHAPTER
 1:
 Introduction
  16
 
 
1.1
 Autism:
 Diagnosis,
 prevalence,
 impact,
 and
 etiology
 
  16
 

  1.2
 Evidence
 that
 lncRNAs
 Contribute
 to
 ASD
  19
 

  1.3
 What
 is
 a
 ncRNA?
  23
 
1.3.1
 Small
 ncRNA
  24
 
1.3.2
 Long
 ncRNA
 (lncRNA)
  25
 
1.4
 NcRNA
 in
 Clinical
 Trials
  26
 
   
   
   
   
   
  25
 
1.5
 MSNP1AS
 as
 a
 candidate
 gene
 for
 autism
 spectrum
 disorder
  28
 
1.6
 Future
 Directions
 and
 Conclusions
  29
 
CHAPTER
 2:
 Impact
 of
 the
 autism-­‐associated
 noncoding
 RNA
 MSNP1AS
 on
 neuronal
 
  30
 
architecture
 and
 gene
 expression
 in
 human
 neural
 progenitor
 cells.
   
 

  2.1
 Abstract
  30
 
 

  2.2
 Introduction
  31
 

  2.3
 Materials
 and
 Methods
  33
 
2.3.1
 Cell
 Culture
  33
 
   
   
   
   
   
   
  32
 
vii
 

 
2.3.2
 Transfection
 of
 Over-­‐Expression
 Constructs
 
  33
 
2.3.3
 Imaging
  34
 
   
  33
 
2.3.4
 Neural
 Progenitor
 Cell
 Harvest
 for
 RNA
 Purification
  34
 
2.3.5
 RNA
 Purification
  35
 
   
   
   
   
  34
 
2.3.6
 Quantitative
 RT-­‐PCR
 (qRT-­‐PCR)
  35
 
2.3.7
 Construction
 of
 Strand-­‐Specific,
 Ribosomal
 RNA
 Depleted
 RNA
 Sequencing
 
  35
 
Libraries
 
2.3.8
 Data
 Analysis
  37
 

  2.4
 Results
  38
 
2.4.1
 Overexpression
 of
 MSNP1AS
 Decreased
 Neurite
 Number
 and
 Length
 in
 SK-­‐N
  38
 
SH
 and
 ReNcell
 CX
 Human
 Neural
 Progenitor
 Cells
 
2.4.2
 Genome-­‐Wide
 Changes
 in
 Gene
 Expression
 Following
 MSNP1AS
 
  41
 
Overexpression
 in
 Human
 Neural
 Progenitor
 Cells
 
2.4.3
 Transcriptional
 Consequences
 of
 MSNP1AS
 Overexpression
 Are
 Enriched
 
  46
 
in
 Protein
 Synthesis
 and
 Chromatin
 Regulation
 

  2.5
 Discussion
  47
 
2.5.1
 Summary
 of
 Results
  47
 
   
   
  51
 

  2.6
 Conclusion
  55
 

  2.7
 Acknowledgements
  56
 
CHAPTER
 3:
 Transcriptional
 gene
 silencing
 of
 the
 autism-­‐associated
 long
 noncoding
 RNA
 
  56
 

  MSNP1AS
 in
 human
 neural
 progenitor
 cells.
   
 

  3.1
 Abstract
  56
 
viii
 

 

  3.2
 Introduction
  57
 
3.3
 Materials
 and
 Methods
  59
 
3.3.1
 Cell
 culture
  59
 
3.3.2
 Design
 of
 antisense
 RNA
 to
 the
 MSNP1AS
 proximal
 gene
 promoters
  60
 
3.3.3
 Transfection
 of
 small
 interfering
 RNAs
 (sasRNAs)
 
  60
 
3.3.4
 Neural
 progenitor
 cell
 harvest
 
  61
 
3.3.5
 RNA
 purification
  62
 
3.3.6
 Quantitative
 RT-­‐PCR
 (qPCR)
 to
 confirm
 MSNP1AS
 knockdown
  62
 
3.3.7
 Construction
 of
 strand-­‐specific,
 ribosomal
 RNA
 depleted
 RNA
 sequencing
 
  62
 
libraries
 
3.3.8
 Data
 analysis
  63
 
3.3.9
 qPCR
 to
 confirm
 altered
 gene
 expression
 observed
 in
 RNA-­‐seq
  63
 
3.3.10
 Analysis
 of
 Brainspan
 data
  64
 
3.4
 Results
  65
 
3.4.1
 MSNP1AS
 knockdown
 results
 in
 genome-­‐wide
 transcriptional
 changes
 
  65
 
3.4.2
 Transcriptional
 consequences
 of
 MSNP1AS
 knockdown
  66
 
3.4.3
 Expression
 of
 MSNP1AS
 in
 developing
 human
 brain
  66
 
3.5
 Discussion
  71
 
CHAPTER
 4:
 Distinct
 gene
 expression
 patterns
 resulting
 from
 prenatal
 pesticide
 exposure
  76
 

 in
 autism
 spectrum
 disorder.
   
 
4.1
 Abstract
  76
 
4.2
 Introduction
  77
 
ix
 

 
4.3
 Materials
 and
 Methods
  78
 
4.3.1
 RNA
 sequencing
 analysis
  78
 
4.3.2
 Functional
 enrichment
  79
 
4.3.3
 Weighted
 Gene
 Co-­‐Expression
 Network
 Analysis
 (WGCNA)
  79
 
4.3.4
 Module
 Preservation
  80
 
4.4
 Results
  81
 
4.4.1
 All
 differentially
 expressed
 genes:
 exposed
 vs
 nonexposed
  81
 
4.4.2
 TD
 Exposed
 vs
 Nonexposed
  81
 
4.4.3
 DD
 Exposed
 vs
 Nonexposed
  83
 
4.4.4
 ASD
 Exposed
 vs
 Nonexposed
  83
 
4.4.5
 All
 ASD
 vs
 DD
  86
 
4.4.6
 All
 ASD
 vs
 TD
  86
 
4.4.7
 Differentially
 Expressed
 Genes
 Separated
 by
 Age
  86
 
4.4.7.1
 TD
 Exposed
 vs
 Nonexposed
 
  87
 

  4.4.7.2
 DD
 Exposed
 vs
 Nonexposed
 
  87
 

  4.4.7.3
 ASD
 Exposed
 vs
 Nonexposed
 
  88
 

  4.4.7.4
 Exposed
 ASD
 vs
 Exposed
 TD
 
  89
 

  4.4.7.5
 Exposed
 ASD
 vs
 Exposed
 DD
 
  90
 
4.4.8
 Weighted
 gene
 co-­‐expression
 network
 analysis
 
  90
 

  4.5
 Discussion
  91
 
 
CHAPTER
 5:
 Summary
 and
 Future
 Directions
   
  93
 
5.1
 Summary
   
  93
 
x
 

 
 
5.2
 Overexpression
 of
 MSNP1AS
 leads
 to
 changes
 in
 MSN
 protein
 expression,
 
   
  94
 
protein
 synthesis
 and
 chromatin
 organization.
 
5.3
 Knockdown
 of
 MSNP1AS
 reveals
 changes
 in
 genes
 involved
 in
 chromatin
 
  95
 
organization
 and
 immune
 response
 
5.4
 Immune
 response
 is
 observed
 in
 vivo.
  96
 
5.5
 Future
 Directions
  97
 
REFERENCES
  99
 

 

 
xi
 

 
LIST
 OF
 TABLES
 

 

 

 
Table
 2.1.
 Read
 mapping
 of
 RNA
 Sequencing
 libraries.
  36
 
Table
 2.2.
 Relative
 expression
 of
 MSNP1AS
 after
 transfection
 with
 overexpression
 
  37
 
construct
 in
 SK-­‐N-­‐SH
 cells.
 
Table
 2.3.
 Relative
 expression
 of
 MSNP1AS
 after
 transfection
 with
 overexpression
 
  39
 
 
construct
 in
 ReNcell
 CX
 cells.
 
Table
 2.4.
 Top
 thirty
 differentially
 expressed
 genes
 in
 SK-­‐N-­‐SH
 cells
 24
 hrs
 after
 MSNP1AS
 
 
  42
 
overexpression.
 
Table
 2.5.
 Top
 thirty
 differentially
 expressed
 genes
 in
 SK-­‐N-­‐SH
 cells
 72
 hrs
 after
 MSNP1AS
  43
 
 
overexpression.
 
Table
 2.6.
 Top
 thirty
 differentially
 expressed
 genes
 in
 ReNcell
 CX
 cells
 24
 hrs
 after
 
  44
 
MSNP1AS
 overexpression.
 
Table
 2.7.
 Top
 thirty
 differentially
 expressed
 genes
 in
 ReNcell
 CX
 cells
 72
 hrs
 after
 
  45
 
MSNP1AS
 overexpression.
 
Table
 2.8.
 Top
 twenty-­‐five
 significant
 p-­‐values
 of
 gene
 ontology
 categories
 for
 SK-­‐N-­‐SH
 
  48
 
cells
 24
 hrs
 post-­‐transfection.
 
Table
 2.9.
 Top
 twenty-­‐five
 significant
 p-­‐values
 of
 gene
 ontology
 categories
 for
 SK-­‐N-­‐SH
  49
 
 
cells
 72
 hrs
 post-­‐transfection.
 
Table
 2.10.
 Top
 twenty-­‐five
 significant
 p-­‐values
 of
 gene
 ontology
 categories
 for
 ReNcell
 CX
  50
 
 
cells
 24
 hrs
 post-­‐transfection.
 
Table
 2.11.
 Top
 twenty-­‐five
 significant
 p-­‐values
 of
 gene
 ontology
 categories
 for
 ReNcell
 CX
  51
 
 
xii
 

 
cells
 72
 hrs
 post-­‐transfection.
   
 
Table
 3.1.
 MSNP1AS
 Fold
 Change
 After
 Transfection.
 
  61
 
Table
 3.2.
 MSNP1AS
 Stranded
 Cuffdiff
 Tophat
 Read
 Count.
  63
 
Table
 3.3.
 Top
 thirty
 differentially
 expressed
 genes
 following
 knockdown
 of
 MSNP1AS.
  67
 
Table
 3.4.
 Top
 Ten
 Differentially
 Expressed
 Genes
 Following
 Knockdown
 of
 MSNP1AS.
  73
 
Table
 3.5.
 Relative
 Fold
 Change
 of
 MSN
 After
 MSNP1AS
 Knockdown.
  75
 
Table
 4.1.
 Differentially
 Expressed
 Genes
 (Exposed
 vs.
 Not
 Exposed).
  82
 
Table
 4.2.
 Differentially
 Expressed
 Genes
 (TD
 Exposed
 vs.
 Not
 Exposed).
  82
 
Table
 4.3.
 Differentially
 Expressed
 Genes
 (ASD
 Exposed
 vs.
 Not
 Exposed).
  84
 

 

   
 
xiii
 

 
LIST
 OF
 FIGURES
 

 
Chapter
 2
 
2.1.
 Representative
 ReNcell
 CX
 cells
 at
 24
 h
 post-­‐transfection.
  40
 
2.2.
 Overexpression
 of
 MSNP1AS
 decreases
 neurite
 number
 and
 length.
  40
 
2.3.
 GO
 enrichment
 analysis
 on
 all
 differentially
 expressed
 genes
 
  47
 
due
 to
 overexpression
 of
 MSNP1AS.
   
 
2.4.
 Potential
 mechanisms
 by
 which
 MSNP1AS
 alters
 neuronal
 architecture.
  52
 
Chapter
 3
 
3.1.
 Small
 noncoding
 antisense
 RNAs
 (sasRNAs)
 were
 designed
 to
 silence
 
  60
 
transcription
 by
 targeting
 the
 MSNP1AS
 gene
 promoter.
 
 
3.2.
 Quantitation
 of
 MSNP1AS
 knockdown.
 
  68
 
3.3.
 Volcano
 plot
 generated
 to
 show
 the
 differentially
 expressed
 genes
 resulting
 
  69
 
from
 the
 MSNP1AS
 knockdown.
 
3.4.
 Confirmation
 of
 altered
 expression
 of
 genes
 identified
 by
 RNA
 seq.
  70
 
3.5.
 Gene
 Ontology
 enrichment
 analysis
 of
 differentially
 expressed
 genes
 (q<0.05)
 
  71
 
in
 SK-­‐N-­‐SH
 cells
 after
 24
 hours
 of
 MSNP1AS
 knockdown
 revealed
 changes
 in
 
chromatin
 assembly.
 
3.6.
 GO
 enrichment
 analysis
 of
 differentially
 expressed
 genes
 (q<0.05)
 in
 SK-­‐N-­‐SH
 
  72
 
cells
 after
 24
 hours
 of
 MSNP1AS.
 
 
3.7.
 Boxplot
 representing
 the
 expression
 of
 MSNP1AS
 through
 human
 cerebral
 
  73
 
cortex
 development.
 
xiv
 

 

 
Chapter
 4
 
4.1.
 Volcano
 plot
 generated
 to
 show
 the
 differentially
 expressed
 genes
 between
 
  84
 
the
 exposed
 to
 pesticides
 group
 and
 the
 unexposed.
 
4.2.
 GO
 enrichment
 analysis
 of
 differentially
 expressed
 genes
 (q<0.05)
 of
 the
 
  85
 
exposed
 to
 pesticides
 versus
 not
 exposed
 ASD
 children.
 
4.3.
 Venn
 diagrams
 of
 differentially
 expressed
 genes
 across
 age
 groups.
  87
 
4.4.
 GO
 enrichment
 analysis
 of
 differentially
 expressed
 genes
 (p<0.05)
 of
 the
 TD
 
  88
 
children
 exposed
 versus
 not
 exposed
 to
 pesticides
 in
 the
 younger
 (less
 than
 3.5
 
years)
 children.
 
4.5.
 GO
 enrichment
 analysis
 of
 differentially
 expressed
 genes
 (q<0.05)
 of
 the
 ASD
 
  91
 
children
 exposed
 to
 pesticides
 versus
 TD
 children
 exposed
 to
 pesticides
 in
 the
 
younger
 (less
 than
 3.5
 years)
 children.
 
4.6.
 Weighted
 Gene
 Co-­‐Expression
 Network
 Analysis.
  91
 

 

 

 

 

 
15
 

 
ABSTRACT
 
From
 genome-­‐wide
 association
 studies
 (GWAS),
 our
 laboratory
 discovered
 a
 novel
 gene
 
that
 has
 a
 highly
 significant
 association
 with
 ASD.
 This
 gene
 is
 a
 long
 noncoding
 RNA
 (lncRNA)
 
called
 MSNP1AS
 (moesin
 pseudogene
 1,
 antisense).
 MSNP1AS
 has
 increased
 expression
 in
 the
 
cerebral
 cortex
 of
 people
 with
 ASD
 and
 people
 with
 the
 ASD-­‐associated
 genetic
 marker
 
revealed
 by
 the
 GWAS.
 I
 hypothesized
 that
 MSNP1AS
 overexpression
 changes
 the
 expression
 of
 
other
 genes
 that
 contribute
 to
 neuronal
 processes.
 I
 collaborated
 with
 experts
 in
 psychiatry,
 
HIV
 research,
 and
 prostate
 cancer
 research
 to
 uncover
 the
 effect
 of
 ASD-­‐associated
 noncoding
 
RNAs
 (ncRNAs)
 on
 human
 neuronal
 cells.
 
 
First,
 I
 overexpressed
 MSNP1AS
 in
 human
 neuronal
 cell
 lines
 to
 provide
 a
 working
 model
 
of
 the
 ASD
 phenotype.
 I
 used
 RNA-­‐Seq
 to
 find
 transcripts
 with
 altered
 expression
 in
 the
 cells
 
after
 overexpression
 and
 recorded
 the
 impact
 of
 the
 increased
 MSNP1AS
 expression
 on
 cell
 
morphology.
 MSNP1AS
 is
 94%
 identical
 and
 antisense
 to
 moesin,
 a
 protein-­‐coding
 gene.
 Moesin
 
is
 an
 X
 chromosome
 gene
 that
 plays
 a
 role
 in
 neuronal
 processes
 and
 stability
 as
 well
 as
 
immune
 response.
 While
 MSNP1AS
 overexpression
 significantly
 decreased
 moesin
 protein
 
expression,
 moesin
 transcript
 levels
 remained
 unchanged.
 RNA-­‐sequencing
 after
 MSNP1AS
 
overexpression
 revealed
 in
 changes
 in
 gene
 expression
 of
 genes
 related
 to
 chromatin
 
organization
 and
 protein
 synthesis,
 pointing
 to
 a
 pathway
 independent
 of
 the
 MSNP1AS/MSN
 
interaction.
 Moesin
 contributes
 to
 neuron
 growth,
 and
 ASD
 patients
 have
 decreased
 long-­‐
distance
 brain
 connectivity,
 which
 makes
 moesin
 a
 good
 candidate
 for
 study
 in
 this
 disorder.
 
Cell
 growth
 was
 stunted
 by
 MSNP1AS
 overexpression.
 Neurite
 length
 and
 number
 were
 
decreased
 after
 overexpression,
 hinting
 at
 an
 effect
 on
 connectivity.
 
 
16
 

 
Second,
 I
 determined
 the
 best
 way
 to
 effectively
 decrease
 MSNP1AS
 in
 human
 neural
 
progenitor
 cells
 to
 reveal
 transcripts
 with
 altered
 expression
 as
 potential
 therapeutic
 targets.
 
 
These
 results
 validated
 MSNP1AS’s
 effect
 on
 chromatin
 organization
 and
 uncovered
 an
 immune
 
response.
 Within
 ASD
 is
 a
 subset
 of
 patients
 who
 experience
 immune
 dysfunction,
 such
 as
 
autoimmune
 disorders
 and
 gastrointestinal
 upset.
 More
 work
 needs
 to
 be
 done
 to
 evaluate
 
whether
 or
 not
 MSNP1AS
 dysregulation
 is
 specific
 to
 these
 patients.
 
Through
 a
 collaboration
 with
 UC
 Davis,
 I
 was
 able
 to
 validate
 these
 changes
 in
 immune
 
response
 genes’
 expression
 in
 vivo
 and
 point
 to
 a
 potential
 environmental
 contribution.
 The
 
Childhood
 Autism
 Risks
 from
 Genetics
 and
 the
 Environment
 (CHARGE)
 study
 took
 blood
 
samples
 from
 children
 with
 and
 without
 autism
 who
 had
 either
 been
 exposed
 to
 pesticides
 in
 
their
 environment
 or
 not.
 The
 RNA
 from
 the
 blood
 samples
 was
 sequenced
 and
 the
 genes
 
changed
 showed
 pathways
 involved
 in
 the
 immune
 system.
 Taken
 together,
 these
 experiments
 
suggest
 molecular
 mechanisms
 that
 are
 working
 together
 to
 alter
 brain
 development,
 
contributing
 to
 ASD
 pathogenesis.
 
 

 
Chapter
 1:
 Introduction
 

 
1.1 Autism:
 Diagnosis,
 prevalence,
 impact,
 and
 etiology
 
Autism
 spectrum
 disorder
 is
 a
 neurodevelopmental
 disorder
 marked
 by
 two
 core
 
symptoms:
 impairments
 in
 social
 communication
 and
 behavioral
 flexibility.
 Deficits
 in
 social
 
communication
 include
 avoidance
 of
 social
 interaction
 and
 eye
 contact
 and
 an
 inability
 to
 use
 
verbal
 language
 to
 communicate.
 
 Impairments
 in
 behavioral
 flexibility
 include
 repetitive
 
behaviors
 such
 as
 hand
 flapping
 and
 restricted
 interests
 such
 as
 demonstrating
 an
 all-­‐
encompassing
 interest
 in
 specific
 objects
 (1).
 The
 behavioral
 diagnosis
 is
 not
 possible
 until
 the
 
17
 

 
child
 has
 reached
 an
 age
 at
 which
 they
 begin
 to
 develop
 language,
 usually
 between
 two
 and
 
three
 years
 old
 (2).
 
The
 prevalence
 of
 autism
 spectrum
 disorder
 has
 increased
 in
 recent
 years,
 due
 in
 part
 to
 
redefining
 the
 diagnostic
 criteria.
 A
 meta-­‐analysis
 of
 epidemiological
 studies
 spanning
 the
 
United
 States,
 United
 Kingdom,
 Japan
 and
 Sweden
 from
 1996
 to
 2001
 estimated
 the
 prevalence
 
to
 be
 between
 1
 in
 167-­‐334
 (3).
 A
 more
 recent
 study
 from
 2010
 estimated
 autism
 prevalence
 to
 
be
 1
 in
 68
 (4).
 This
 increase
 may
 be
 due
 to
 increased
 awareness,
 surveys,
 public
 interest,
 and
 
changes
 in
 service
 ability
 at
 both
 the
 public
 and
 professional
 level
 (5).
 Quantifying
 prevalence
 is
 
difficult
 and
 pinpointing
 specific
 environmental
 effects
 is
 challenging.
 Further,
 data
 from
 low-­‐
income
 countries
 is
 difficult
 to
 obtain,
 making
 regional
 comparisons
 another
 hurdle
 (6).
 
One
 motivating
 factor
 towards
 studying
 autism
 spectrum
 disorder
 is
 the
 impact
 it
 has
 on
 
the
 caregivers.
 Parents
 of
 autistic
 children
 score
 higher
 on
 stress
 tests
 than
 any
 other
 parental
 
groups.
 These
 parents
 also
 have
 a
 higher
 incidence
 of
 mental
 health
 problems
 (7).
 The
 current
 
treatments
 for
 autism
 spectrum
 disorder
 are
 limited
 (8).
 They
 start
 with
 intensive
 behavioral
 
interventions
 called
 applied
 behavioral
 analysis,
 and
 can
 include
 counseling,
 speech
 therapy,
 
social
 skills
 training,
 occupational
 therapy,
 and
 physical
 therapy
 (9).
 The
 type
 and
 extent
 of
 
therapy
 needed
 is
 determined
 by
 where
 the
 child
 falls
 on
 the
 spectrum;
 accessibility
 of
 
therapies
 depends
 on
 the
 location
 and
 living
 situation
 of
 the
 family.
 Many
 families
 find
 therapy
 
inaccessible
 (10).
 Families
 living
 in
 rural
 areas
 are
 more
 likely
 to
 use
 alternative,
 untested
 
treatments
 due
 to
 reduced
 access
 to
 care
 (11).
 Cost
 can
 also
 be
 a
 prohibitive
 factor
 (12,
 13),
 
and
 costs
 increase
 as
 patients
 get
 older
 (14).
 Often,
 the
 treatments
 available
 are
 not
 
18
 

 
particularly
 effective
 (15)
 because
 they
 only
 apply
 to
 a
 subset
 of
 autism
 patients
 (16-­‐19),
 
further
 reducing
 the
 likelihood
 of
 effective
 treatment.
 
Subtyping
 ASD
 is
 also
 an
 area
 of
 contention.
 Redefining
 diagnostic
 criteria
 has
 aided
 
somewhat
 in
 unifying
 incongruous
 aspects
 of
 the
 disorder,
 but
 heterogeneity
 in
 the
 ASD
 
phenotype
 contributes
 to
 the
 complexity
 of
 etiology.
 While
 an
 individual
 must
 have
 social
 
communication
 deficits
 and
 repetitive,
 restricted-­‐interest
 behaviors
 (RRB),
 the
 types
 of
 deficits
 
in
 those
 categories
 differ
 greatly.
 For
 example,
 a
 subset
 of
 individuals
 with
 ASD
 might
 have
 
varying
 levels
 of
 language
 development.
 One
 group
 might
 have
 trouble
 with
 grammar
 and
 
syntax,
 while
 another
 subset
 might
 not
 speak
 at
 all
 (20).
 
 
The
 variation
 observed
 in
 the
 phenotype
 is
 reflected
 in
 the
 genetic
 variation.
 The
 estimated
 
heritability
 of
 autism
 spectrum
 disorder
 is
 more
 than
 70%
 (21-­‐23).
 Chromosomal
 regions
 and
 
linkage
 regions
 with
 genome-­‐wide
 significant
 association
 with
 ASD
 have
 been
 identified
 using
 
whole
 genome
 methods
 (24-­‐28).
 
 Rare
 mutations
 of
 protein
 coding
 genes
 (29-­‐34)
 and
 rare
 copy
 
number
 variations
 (CNVs)
 (35-­‐37)
 have
 also
 been
 associated,
 but
 these
 powerful
 genome-­‐wide
 
approaches
 have
 not
 produced
 reliable
 candidate
 genes.
 Most
 genes
 that
 have
 been
 identified
 
do
 not
 overlap
 with
 parallel
 studies.
 Of
 the
 70%
 of
 phenotypic
 variability
 that
 can
 be
 
contributed
 to
 genetics,
 half
 of
 heritability
 is
 due
 to
 common
 variants
 (38),
 so
 it
 makes
 sense
 to
 
focus
 on
 the
 common
 variants
 responsible.
 
 
In
 tandem
 with
 heritability,
 it
 is
 suggested
 that
 ASD
 etiology
 can
 be
 attributed
 to
 
environmental
 factors.
 Seasonal
 factors,
 nutrition,
 birth
 order,
 maternal
 stress,
 migration,
 air
 
pollution,
 and
 organic
 toxicants
 have
 all
 been
 implicated
 in
 ASD
 risk
 (39).
 
 

 
19
 

 
1.2
 Evidence
 that
 lncRNAs
 Contribute
 to
 ASD
 
As
 for
 all
 common
 complex
 genetic
 disorders,
 there
 is
 an
 on
 going
 debate
 about
 the
 
relative
 contributions
 of
 rare
 and
 common
 genetic
 variants
 to
 ASD.
 The
 identification
 of
 rare
 de
 
novo
 copy
 number
 variants
 (CNVs)
 on
 chromosomes
 2p,
 7q,
 15q,
 and
 16p
 provide
 support
 for
 
rare
 variants
 and
 important
 biological
 clues
 to
 the
 etiology
 of
 ASD
 (35-­‐37,
 40-­‐42).
 However,
 the
 
contribution
 of
 these
 CNVs
 remains
 at
 <6%
 of
 cases
 (with
 each
 CNV
 observed
 in
 <1%)
 despite
 
extensive
 cohort
 collection
 and
 technological
 advances
 that
 increased
 CNV
 detection
 resolution
 
by
 10-­‐fold
 (42).
 It
 is
 now
 clear
 that
 CNVs
 do
 not
 account
 for
 a
 large
 fraction
 of
 ASD,
 either
 
individually
 or
 collectively
 (43).
 Further,
 the
 implicated
 CNVs
 are
 typically
 large
 (~250
 kb)
 and
 
encompass
 multiple
 protein-­‐coding
 genes,
 making
 it
 difficult
 to
 determine
 which
 gene(s)
 
contribute
 to
 the
 disorder.
 Genome-­‐wide
 association
 studies
 (GWASs)
 have
 proven
 effective
 in
 
identifying
 associated
 genetic
 variants
 in
 other
 complex
 genetic
 disorders,
 such
 as
 type
 2
 
diabetes,
 Crohn’s
 disease,
 and
 prostate
 cancer
 (44-­‐50).
 However,
 application
 of
 GWAS
 to
 ASD
 
yielded
 mixed
 results
 (25-­‐28).
 Because
 GWAS
 is
 the
 definitive
 technique
 for
 identifying
 
associated
 common
 variants
 of
 large
 effect,
 the
 lack
 of
 obvious
 protein-­‐coding
 genes
 implicated
 
by
 GWAS
 pushed
 the
 ASD
 genetics
 fields
 toward
 searching
 for
 rare
 causal
 variants
 (29-­‐34,
 40,
 
42,
 51-­‐53).
 Whole
 exome
 sequencing
 (next-­‐generation
 sequencing
 restricted
 to
 the
 exons
 of
 
protein-­‐coding
 genes)
 has
 proven
 effective
 in
 identifying
 causal
 genetic
 variants
 in
 Mendelian
 
disorders,
 such
 as
 Miller
 syndrome,
 high
 myopia,
 and
 malformations
 of
 cortical
 development
 
(54-­‐56).
 However,
 application
 of
 whole
 exome
 sequencing
 to
 ASD
 also
 yielded
 mixed
 results
 
(29-­‐34).
 Whole
 exome
 sequencing
 approaches
 have
 identified
 de
 novo
 mutations
 that
 may
 be
 
specific
 to
 individuals
 with
 ASD,
 but
 the
 identified
 de
 novo
 mutations
 do
 not
 account
 for
 a
 large
 
20
 

 
fraction
 of
 cases,
 either
 individually
 or
 collectively
 (29-­‐34).
 Thus,
 the
 definitive
 experiments
 
have
 not
 yet
 resolved
 the
 common
 variant
 versus
 rare
 variant
 debate,
 and
 suggest
 that
 both
 
common
 variants
 and
 rare
 variants
 contribute
 to
 ASD
 risk
 (1).
 
 
Based
 on
 a
 variety
 of
 evidence,
 we
 consider
 common
 variants
 likely
 to
 play
 a
 large
 role
 
in
 ASD
 risk.
 First,
 the
 high
 heritability
 of
 ASD
 suggests
 that
 the
 de
 novo
 mutations
 implicated
 by
 
exome
 sequencing
 and
 CNV
 detection
 play
 a
 small
 role
 in
 the
 disorders.
 Second,
 sub-­‐clinical
 
ASD
 traits
 are
 present
 in
 >50%
 of
 family
 members
 of
 affected
 individuals
 (57-­‐60),
 suggesting
 
that
 common
 heritable
 factors
 contribute
 to
 a
 large
 fraction
 of
 cases.
 Third,
 the
 heritability
 of
 
ASD
 traits
 is
 high
 and
 indistinguishable
 between
 extreme
 cases
 and
 the
 general
 population,
 
indicating
 that
 these
 traits
 are
 the
 quantitative
 extreme
 of
 a
 neurodevelopmental
 continuum,
 
not
 a
 mutation-­‐induced
 abnormality
 (61,
 62).
 Fourth,
 the
 conclusions
 from
 the
 exome
 
sequencing
 studies
 that
 no
 single
 gene
 is
 causal
 and
 that
 hundreds
 of
 genes
 will
 increase
 risk
 
(29-­‐34)
 indicate
 that
 evidence
 for
 association
 of
 each
 genetic
 variant
 with
 the
 disorder
 will
 
continue
 to
 be
 of
 utmost
 importance.
 While
 genome-­‐wide
 significant
 association
 of
 common
 
markers
 may
 drift
 with
 additional
 genotyping,
 the
 association
 is
 likely
 to
 remain
 significant.
 In
 
contrast,
 the
 nominally
 significant
 association
 of
 a
 rare
 variant
 can
 disappear
 with
 the
 positive
 
identification
 of
 a
 mutation
 in
 a
 single
 unaffected
 individual.
 Although
 the
 genetic
 effect
 sizes
 of
 
common
 variants
 for
 ASD
 risk
 may
 be
 small,
 similar
 effect
 sizes
 have
 identified
 therapeutic
 
targets
 for
 Crohn’s
 disease
 and
 type
 2
 diabetes
 (1).
 
 
The
 exome
 sequencing
 studies
 (29-­‐34)
 all
 concluded
 that
 no
 single
 gene
 is
 causal
 for
 
ASD
 and
 that
 400–1000
 different
 genes
 may
 increase
 ASD
 risk.
 An
 alternative
 interpretation
 of
 
these
 data
 is
 that
 de
 novo
 mutation
 of
 protein-­‐coding
 genes
 may
 contribute
 little
 to
 ASD.
 
21
 

 
Instead,
 genetic
 variation
 affecting
 long
 noncoding
 RNA
 (lncRNA),
 which
 are
 by
 definition
 
excluded
 from
 exome
 sequencing
 studies,
 may
 contribute
 substantially.
 Three
 lines
 of
 evidence
 
support
 this
 interpretation.
 First,
 while
 individual
 common
 genetic
 variants
 contribute
 small
 risk
 
to
 ASD,
 collectively
 common
 genetic
 variants
 explain
 more
 than
 40%
 of
 genetic
 risk
 for
 ASD
 
(63).
 Second,
 large-­‐scale
 whole
 genome
 sequencing
 for
 high-­‐density
 lipoprotein
 (HDL)
 indicates
 
that
 noncoding
 regions
 contribute
 as
 much
 as
 protein-­‐coding
 genes
 to
 this
 complex
 genetic
 
condition
 (64).
 Third,
 nearly
 two-­‐thirds
 (64%)
 of
 all
 long
 transcripts
 in
 the
 human
 brain
 are
 
lncRNAs
 (65).
 It
 is
 also
 becoming
 increasingly
 clear
 that
 the
 number
 of
 ncRNAs
 encoded
 in
 a
 
genome
 increases
 with
 developmental
 complexity
 (66,
 67),
 and
 that
 genetic
 variation
 within
 
human
 lncRNAs
 contributes
 to
 neurodevelopmental
 disorders
 (1,
 67-­‐69).
 
 
Concurrent
 with
 these
 genetic
 observations,
 the
 contribution
 of
 lncRNAs
 became
 more
 
apparent
 in
 other
 complex
 genetic
 disorders,
 including
 cancer
 (70,
 71)
 and
 schizophrenia
 (72).
 
Emerging
 evidence
 from
 multiple
 approaches
 indicate
 a
 contribution
 of
 lncRNA
 to
 ASD
 (1).
 
 
The
 first
 evidence
 that
 lncRNAs
 may
 contribute
 to
 ASD
 was
 derived
 from
 a
 study
 of
 X-­‐
linked
 protein-­‐coding
 genes.
 Noor
 et
 al.
 found
 that
 0.5%
 of
 individuals
 have
 mutation
 in
 the
 
protein-­‐coding
 PTCHD1
 gene.
 An
 additional
 0.5%
 of
 ASD
 cases
 also
 had
 deletions
 of
 a
 complex
 
noncoding
 RNA
 (ncNRA)
 locus
 (PTCHD1AS1/PTCHD1AS2)
 5ʹ′
 to
 the
 PTCHD1
 locus
 that
 are
 
associated
 with
 ASD
 (55).
 Although
 the
 deletions
 are
 rare
 and
 the
 functional
 mechanisms
 of
 the
 
PTCHD1AS1
 and
 PTCHD1AS2
 ncRNAs
 have
 not
 been
 explored
 extensively,
 they
 are
 presumed
 to
 
regulate
 expression
 of
 the
 protein-­‐coding
 gene
 on
 the
 opposite
 strand,
 PTCHD1
 (1,
 55).
 
The
 second
 report
 indicating
 evidence
 for
 lncRNA
 contribution
 to
 ASD
 was
 based
 on
 a
 
genome-­‐wide
 significant
 association
 in
 a
 protein-­‐coding
 gene-­‐poor
 region
 of
 chromosome
 
22
 

 
5p14.1.
 A
 GWAS
 of
 ASD
 indicated
 the
 most
 significant
 association
 to
 date
 (P
 =
 10
−10
)
 for
 
rs4307059
 on
 chromosome
 5p14.1.7
 The
 same
 rs4307059
 allele
 was
 also
 associated
 with
 social
 
communication
 phenotypes
 in
 a
 general
 population
 sample
 (73).
 However,
 rs4307059
 genotype
 
was
 not
 correlated
 with
 expression
 of
 either
 of
 the
 flanking
 protein-­‐coding
 genes,
 CDH9
 and
 
CDH10
 (26).
 We
 identified
 a
 3.9
 kb
 ncRNA
 that
 is
 transcribed
 directly
 at
 the
 site
 of
 the
 
chromosome
 5p14.1
 ASD
 GWAS
 peak
 (74).
 The
 ncRNA
 is
 encoded
 by
 the
 opposite
 (anti-­‐sense)
 
strand
 of
 moesin
 pseudogene
 1
 (MSNP1),
 and
 is
 thus
 designated
 MSNP1AS
 (moesin
 
pseudogene
 1,
 anti-­‐sense).
 MSNP1AS
 is
 94%
 identical
 and
 anti-­‐sense
 to
 the
 X
 chromosome
 
transcript
 MSN,
 which
 encodes
 a
 protein
 (moesin)
 that
 regulates
 neuronal
 architecture
 and
 
immune
 response.
 Expression
 of
 MSNP1AS
 in
 postmortem
 temporal
 cortex
 is
 increased
 12.7-­‐
fold
 in
 individuals
 with
 ASD
 and
 increased
 22-­‐fold
 in
 individuals
 with
 the
 rs4307059
 risk
 allele
 
(74).
 The
 MSNP1AS
 ncRNA
 binds
 MSN
 and
 its
 over-­‐expression
 in
 cultured
 neurons
 causes
 
significant
 decreases
 in
 moesin
 protein,
 neurite
 number
 and
 neurite
 length.
 Thus,
 our
 discovery
 
reveals
 a
 functional
 lncRNA
 which,
 based
 on
 the
 GWAS
 findings,
 contributes
 to
 ASD
 risk
 (1).
 
 
To
 evaluate
 whether
 or
 not
 changes
 in
 lncRNA
 expression
 are
 part
 of
 ASD
 pathogenesis,
 
33,000
 lncRNAs
 and
 30,000
 mRNA
 transcripts
 from
 ASD
 and
 control
 prefrontal
 cortex
 and
 
cerebellum
 postmortem
 brain
 tissue
 were
 profiled
 using
 microarray.
 Two
 hundred
 lncRNAs
 
were
 differentially
 expressed
 in
 ASD.
 Previously,
 it
 had
 been
 reported
 that
 mRNA
 transcription
 
within
 ASD
 brains
 was
 homogenous
 (75,
 76).
 Another
 study
 also
 found
 increased
 transcriptional
 
homogeneity
 in
 lncRNAs
 in
 ASD
 brains
 (77).
 Although
 the
 conclusions
 of
 this
 study
 implicate
 
lncRNA,
 the
 sample
 size
 was
 small
 (2
 ASD
 cases
 and
 2
 matched
 controls).
 More
 studies
 with
 
larger
 sample
 sizes
 will
 be
 necessary
 to
 determine
 specific
 lncRNAs
 with
 consistently
 altered
 
23
 

 
expression
 in
 ASD
 brains
 (1).
 
 
Therefore,
 accumulating
 evidence
 from
 multiple
 experimental
 approaches
 suggest
 that
 
lncRNAs
 contribute
 to
 ASD.
 The
 promise
 of
 a
 new
 pharmacology
 to
 target
 lncRNA
 suggests
 
possible
 treatment
 options
 (1).
 

 
1.3
 What
 is
 a
 ncRNA?
 
The
 eukaryotic
 genome
 is
 much
 more
 complex
 than
 previously
 thought.
 Protein-­‐coding
 
gene
 exons
 comprise
 <2%
 of
 the
 total
 genomic
 material.
 However,
 analyses
 of
 the
 
transcriptome
 revealed
 that
 the
 human
 genome
 is
 differentially
 and
 dynamically
 transcribed
 to
 
produce
 a
 spectrum
 of
 ncRNA
 of
 all
 shapes
 and
 sizes
 (78).
 NcRNAs
 have
 been
 arbitrarily
 sorted
 
into
 two
 categories
 based
 on
 their
 size:
 long
 ncRNA
 (greater
 than
 200
 nucleotides
 in
 length)
 and
 
short
 ncRNAs
 (less
 than
 200
 nucleotides
 in
 length)
 (79).
 These
 categories
 can
 be
 further
 broken
 
down
 into
 subcategories
 based
 on
 other
 properties.
 Few
 ncRNAs
 have
 been
 characterized
 
functionally.
 However,
 the
 functions
 that
 have
 been
 discovered
 are
 obviously
 an
 important
 part
 
of
 development
 and
 organismal
 maintenance.
 As
 key
 parts
 of
 cellular
 processes,
 these
 ncRNAs
 
leave
 the
 organism
 vulnerable
 to
 disease
 when
 they
 become
 dysregulated.
 Several
 ncRNAs
 
have
 been
 shown
 to
 play
 a
 role
 in
 the
 pathogenesis
 of
 diseases
 such
 as
 cancer,
 HIV,
 heart
 
disease,
 and
 neurological
 disorders.
 Characterizing
 these
 cellular
 components
 is
 important
 to
 
our
 understanding
 of
 these
 diseases
 and
 may
 provide
 insight
 into
 their
 pathogenesis
 and
 
treatment
 (1).
 

 

   
 
24
 

 
1.3.1
 Small
 ncRNA
 
Small
 ncRNA,
 as
 a
 category,
 include
 RNA
 such
 as
 microRNAs
 (miRNAs),
 short
 interfering
 
RNAs
 (siRNAs),
 small
 nucleolar
 RNAs
 (snoRNAs),
 and
 PIWI-­‐interacting
 RNAs
 (piRNAs).
 The
 genes
 
encoding
 these
 RNA
 can
 be
 found
 throughout
 the
 genome
 (80).
 Small
 ncRNA
 can
 also
 be
 
derived
 from
 processing
 lncRNA
 (81,
 82).
 The
 main
 method
 used
 by
 small
 ncRNA
 to
 regulate
 the
 
genome
 is
 by
 binding
 to
 their
 specific
 target
 transcripts
 via
 complementary
 nucleotide
 
sequences,
 altering
 downstream
 events
 to
 influence
 the
 translation
 of
 the
 target.
 To
 read
 more
 
about
 small
 ncRNA
 subclasses,
 biogenesis,
 and
 forms
 of
 genetic
 regulation,
 see
 the
 following
 
reviews
 (1,
 80,
 83-­‐87).
 
MiRNA
 is
 one
 of
 the
 best-­‐studied
 classes
 of
 small
 ncRNAs
 and
 was
 first
 discovered
 in
 
1993
 in
 C.
 elegans
 (88).
 Since
 its
 discovery,
 miRNA
 is
 believed
 to
 be
 relatively
 conserved
 among
 
vertebrates
 (89-­‐91).
 MiRNAs
 can
 regulate
 genes
 post
 translation
 by
 binding
 to
 either
 the
 3’-­‐(92)
 
or
 the
 5’-­‐UTR
 (93)
 of
 a
 target
 transcript
 which
 typically
 inhibits
 translation.
 One
 particular
 
miRNA
 can
 target
 multiple
 transcripts
 and
 one
 transcipt
 can
 be
 regulated
 by
 multiple
 miRNA
 
(94).
 More
 than
 60%
 of
 human
 protein-­‐coding
 genes
 are
 predicted
 targets
 of
 miRNA,
 lending
 
credence
 to
 the
 idea
 that
 miRNAs
 are
 prevalent
 and
 important
 in
 the
 human
 genome
 (95).
 
Regulation
 by
 miRNA
 is
 vital
 to
 mammalian
 developmental
 processes
 and,
 like
 many
 other
 
ncRNAs,
 miRNA
 expression
 can
 be
 specifically
 and
 differentially
 expressed
 throughout
 tissue
 
types
 and
 the
 developmental
 process
 as
 reviewed
 (1,
 96-­‐98).
 
Like
 miRNA,
 siRNA
 also
 binds
 to
 complementary
 sequences
 on
 target
 transcripts
 and
 
regulates
 gene
 expression.
 The
 difference
 is
 that
 siRNA
 complementary
 sequences
 have
 stricter
 
requirements
 in
 that
 they
 must
 have
 close
 to
 perfect
 sequence
 complementation
 (99).
 
 Due
 to
 
25
 

 
this
 difference,
 siRNA
 and
 miRNA
 have
 two
 separate
 mechanisms
 for
 repression.
 SiRNA
 targets
 
with
 very
 close
 complementation
 undergo
 direct
 cleavage,
 and
 miRNA
 targets
 are
 destabilized
 
or
 transcriptionally
 repressed
 (100-­‐102).
 Endogenous
 siRNAs
 (endo-­‐siRNAs)
 are
 purported
 to
 
protect
 against
 transposons
 and
 dsDNA
 viruses
 in
 somatic
 tissues
 in
 Drosophila
 (103-­‐105).
 
Endo-­‐siRNAs
 have
 also
 been
 identified
 in
 mice
 (106-­‐109)
 and
 in
 human
 cells
 (110-­‐112).
 SiRNA
 
has
 been
 extensively
 researched
 for
 use
 in
 gene
 therapy
 and
 may
 be
 able
 to
 correct
 ncRNA
 
dysregulation
 in
 a
 number
 of
 different
 diseases
 (1,
 113).
 

 
1.3.2
 Long
 ncRNA
 (lncRNA)
 
LncRNA
 is
 an
 expansive
 category
 that
 includes
 many
 different
 structural
 features
 and
 
mechanisms
 of
 action.
 LncRNA
 include
 sense
 and
 antisense
 strands
 in
 pseudogenes,
 in
 
intergenic
 regions
 (lincRNA)
 and
 in
 protein-­‐coding
 genes.
 After
 transcription,
 they
 may
 be
 
further
 processed
 through
 splicing
 and/or
 the
 addition
 of
 a
 5’-­‐methyl-­‐guanosine
 cap
 and
 3’
 poly
 
(A)
 tail.
 They
 also
 can
 contain
 localization
 signals
 that
 direct
 them
 to
 the
 cytoplasm
 or
 the
 
nucleus
 (114).
 LncRNAs
 have
 many
 diverse
 functional
 roles
 as
 well,
 such
 as
 employing
 
enhancer-­‐like
 functions,
 participating
 in
 the
 recruitment
 of
 chromatin-­‐modifying
 complexes,
 
modulating
 alternative
 splicing,
 providing
 a
 scaffold
 for
 the
 assembly
 of
 protein
 complexes,
 and
 
acting
 as
 competing
 endogenous
 RNAs
 (ceRNAs)
 (115-­‐117).
 XIST,
 as
 well
 as
 its
 interaction
 with
 
the
 polycomb
 repressive
 complex,
 PRC2,
 is
 the
 most
 thoroughly
 studied
 lncRNA
 (118).
 XIST’s
 
antisense
 gene,
 TSIX,
 was
 discovered
 in
 1999
 (119).
 TSIX
 serves
 as
 a
 regulatory
 element
 of
 XIST.
 
It
 is
 part
 of
 a
 repression
 pathway
 along
 with
 other
 pluripotency
 factors
 (120).
 There
 is
 much
 to
 
be
 discovered
 about
 lncRNA
 before
 it
 can
 be
 used
 to
 treat
 disease
 (1).
 
26
 

 
1.4
 NcRNA
 in
 Clinical
 Trials
 
NcRNAs
 are
 part
 of
 many
 gene
 regulatory
 pathways.
 When
 lncRNAs
 are
 dysregulated
 by
 
genetic,
 epigenetic
 or
 environmental
 factors,
 they
 can
 cause
 dysregulation
 in
 other
 genes
 of
 
gene
 networks
 that
 contribute
 to
 altered
 development.
 It
 would
 be
 difficult
 to
 alter
 these
 
dysregulated
 translated
 genes
 that
 contribute
 to
 clinical
 diagnoses.
 However,
 they
 can
 be
 
manipulated
 through
 the
 use
 of
 their
 regulatory
 ncRNAs.
 The
 ncRNAs
 complementary
 to
 these
 
translated
 disease-­‐causing
 genes
 can
 target
 these
 genes
 that
 are
 over
 or
 under-­‐expressed.
 The
 
primary
 challenge
 is
 to
 transport
 the
 complementary
 ncRNA
 to
 a
 position
 in
 which
 they
 can
 be
 
effective.
 Using
 ncRNA
 as
 therapeutic
 targets,
 especially
 lncRNAs,
 is
 a
 relatively
 new
 field
 of
 
research,
 but
 next-­‐generation
 RNA
 sequencing
 is
 quickly
 being
 developed
 that
 will
 help
 identify
 
the
 relevant
 transcripts
 (1,
 121).
 
 
Several
 types
 of
 ncRNAs
 have
 been
 utilized
 as
 drugs
 to
 suppress
 or
 upregulate
 genes
 
implicated
 in
 disease.
 SiRNA
 and
 RNAi
 are
 the
 main
 candidates
 for
 gene
 suppression
 and
 have
 
successfully
 repressed
 their
 target
 genes
 in
 several
 diseases
 including
 cancer
 as
 mentioned
 
previously.
 In
 addition
 to
 siRNA
 and
 RNAi,
 antisense
 RNAs
 and
 ribozymes
 (RNA
 that
 is
 
catalytically
 active)
 are
 also
 candidates
 for
 use
 in
 clinical
 trials
 (1).
 
 
The
 first
 example
 of
 RNA
 therapeutics
 used
 in
 clinical
 trials
 is
 antisense
 RNA.
 Antisense
 
RNA
 usually
 targets
 mRNA,
 but
 they
 have
 also
 been
 used
 to
 target
 miRNA
 as
 well.
 These
 RNA
 
are
 chemically
 synthesized
 oligonucleotides
 (ON)
 that
 are
 antisense
 to
 miRNA
 (antagomirs)
 and
 
can
 inhibit
 miRNA
 involved
 in
 disease.
 A
 locked
 nucleic
 acid
 (LNA)-­‐modified
 ON,
 called
 
miravirsen,
 inhibits
 miR-­‐122,
 a
 liver
 miRNA
 involved
 in
 Hepatitis
 C
 viral
 (HCV)
 infection.
 
Miravirsen
 has
 completed
 two
 Phase
 1
 clinical
 trials
 and
 is
 safe.
 It
 is
 currently
 in
 Phase
 IIa
 
27
 

 
clinical
 trials
 to
 test
 tolerability
 and
 efficacy
 in
 HCV
 patients
 (122).
 Twenty-­‐two
 different
 siRNA
 
therapies
 have
 now
 reached
 clinical
 testing
 for
 the
 treatment
 of
 16
 disorders
 (1,
 123).
 
SiRNA
 has
 enjoyed
 moderate
 success
 with
 both
 systemic
 and
 local
 delivery
 as
 a
 drug.
 
SiRNA
 can
 specifically
 knockdown
 any
 mRNA
 and
 is
 easily
 designed
 and
 screened.
 Its
 silencing
 
effects
 also
 last
 a
 long
 time
 and
 can
 retain
 its
 catalytic
 ability
 in
 RISC
 for
 long
 periods
 as
 well
 
(123).
 Many
 siRNA
 drugs
 treating
 vision
 loss
 using
 local
 delivery
 have
 initiated
 Phase
 I
 clinical
 
trials
 (124-­‐127).
 There
 are
 several
 siRNA
 and
 shRNA
 drugs
 delivered
 locally
 to
 treat
 diseases
 
from
 vision
 loss
 to
 respiratory
 viruses
 that
 are
 currently
 in
 Phase
 0
 and
 1
 clinical
 trials
 (1,
 128,
 
129).
 
RNA
 aptamers
 are
 single-­‐stranded
 nucleic
 acids
 with
 stable
 three-­‐dimensional
 shapes
 
that
 allow
 them
 to
 bind
 to
 molecular
 targets
 with
 high
 affinity
 and
 specificity
 (129,
 130).
 RNA
 
aptamers
 are
 similar
 to
 antibodies
 in
 affinity
 and
 specificity
 for
 targets,
 but
 they
 also
 have
 
certain
 advantages
 over
 antibodies.
 Since
 aptamers
 are
 formed
 in
 vitro
 with
 SELEX,
 they
 are
 
more
 efficiently,
 accurately,
 and
 economically
 produced,
 which
 makes
 them
 a
 better
 choice
 for
 
clinical
 application.
 Aptamers
 are
 chemically
 modified,
 therefore
 they
 can
 target
 ligands
 
without
 producing
 the
 immune
 response
 present
 with
 antibodies.
 The
 ability
 of
 aptamers
 to
 be
 
modified
 is
 another
 advantage
 they
 hold
 over
 antibodies.
 Aptamers
 can
 be
 joined
 to
 ribozymes
 
to
 make
 riboswitches
 or
 siRNA
 to
 make
 aptamer-­‐siRNA
 chimeras.
 They
 are
 also
 smaller
 than
 
antibodies,
 which
 improves
 tissue
 penetration
 and
 transport.
 At
 least
 six
 RNA
 aptamers
 have
 
been
 clinically
 tested.
 One,
 a
 VEGF-­‐specific
 modified
 RNA
 aptamer,
 treats
 AMD
 and
 is
 now
 an
 
FDA
 approved
 drug
 (1,
 131-­‐133).
 
 
Ribozymes
 are
 RNA
 molecules
 that
 can
 perform
 biochemical
 actions
 similar
 to
 those
 of
 
28
 

 
an
 enzyme.
 Ribozymes
 have
 gone
 under
 clinical
 testing
 as
 gene
 therapy-­‐based
 approaches
 to
 
treating
 HIV
 using
 CD4+
 T
 cells
 or
 CD34+
 hematopoietic
 stem
 cells
 (HSCs).
 CD34+
 HSCs
 
differentiate
 into
 various
 hematopoietic
 lineages
 including
 CD4+
 T
 cells
 (134-­‐137).
 ShRNA
 can
 
work
 in
 conjunction
 with
 ribozymes.
 RNA-­‐based
 treatment
 was
 developed
 for
 HIV-­‐1
 using
 gene-­‐
modified
 autogolous
 CD34+
 hematopoietic
 progenitor
 cells
 (HPCs)
 and
 is
 now
 in
 a
 Phase
 0
 
clinical
 study
 at
 the
 City
 of
 Hope
 (138).
 The
 RNA-­‐based
 treatment
 involved
 three
 RNA
 
components,
 an
 shRNA,
 a
 Trans-­‐activation-­‐responsive
 (TAR)
 decoy,
 and
 a
 hammerhead
 
ribozyme,
 all
 encoded
 in
 the
 lentiviral
 gene
 vector.
 The
 hammerhead
 ribozyme
 cleaves
 the
 
mRNA
 of
 the
 chemokine
 receptor
 5
 (CCR5)
 protein
 (139).
 The
 CCR5
 receptor
 also
 is
 a
 
coreceptor
 for
 HIV-­‐1
 infection
 and
 a
 subset
 of
 CD4+
 T
 cells
 express
 it.
 This
 receptor
 is
 not
 
necessary
 for
 normal
 T
 cell
 function
 and
 is
 not
 prone
 to
 mutational
 escape,
 therefore
 it
 could
 
be
 a
 good
 target
 for
 anti-­‐HIV
 therapy
 (1).
 

 
1.5
 MSNP1AS
 as
 a
 candidate
 gene
 for
 autism
 spectrum
 disorder
 
Wang
 et
 al
 (26)
 found
 several
 common
 variants
 after
 performing
 a
 GWAS
 on
 ASD
 
patients
 (26).
 In
 2012,
 the
 Campbell
 lab
 published
 a
 paper
 on
 a
 ncRNA
 under
 this
 GWAS
 peak
 
(74).
 The
 transcript
 was
 a
 pseudogene
 called
 MSNP1
 (moesin
 pseudogene
 1).
 MSNP1
 is
 not
 
transcribed
 and
 is
 94%
 identical
 to
 the
 protein-­‐coding
 gene
 moesin,
 which
 is
 on
 the
 X
 
chromosome.
 Northern
 blots
 revealed
 a
 transcript
 expressed
 at
 the
 MSNP1
 locus.
 The
 opposite
 
strand
 of
 MSNP1
 is
 transcribed
 to
 produce
 MSNP1AS
 (moesin
 pseudogene
 1
 antisense).
 A
 gene
 
94%
 identical
 to
 a
 gene
 on
 a
 different
 chromosome,
 going
 in
 the
 opposite
 direction,
 is
 under
 
the
 GWAS
 peak.
 There
 is
 a
 significant
 12.7-­‐fold
 increase
 in
 expression
 of
 MSNP1AS
 in
 autism
 
29
 

 
brains
 compared
 to
 control.
 I
 evaluated
 the
 effect
 of
 this
 increase
 and
 developed
 a
 way
 to
 
knockdown
 the
 gene
 in
 the
 hopes
 that
 the
 effects
 would
 be
 negated
 (67,
 69).
 There
 is
 also
 a
 
significant,
 but
 smaller,
 2.4-­‐fold
 increase
 of
 MSN
 expression
 in
 autism
 brains.
 CDH9
 and
 CDH10
 
were
 also
 tested,
 but
 there
 was
 not
 a
 significant
 difference
 between
 the
 control
 brains
 and
 the
 
autism
 brains.
 Preliminary
 data
 suggested
 that
 one
 effect
 of
 this
 increase
 in
 expression
 of
 
MSNP1AS
 is
 altered
 neuronal
 function
 in
 mouse
 primary
 cortical
 neurons.
 When
 transfected
 
with
 MSNP1AS
 overexpression
 constructs,
 the
 neurite
 length
 and
 number
 was
 decreased.
 

 
1.6
 Future
 Directions
 and
 Conclusions
 
Based
 on
 recent
 genetics
 findings,
 I
 anticipate
 that
 ncRNAs
 will
 account
 for
 a
 large
 
proportion
 of
 ASD
 risk.
 My
 own
 published
 and
 unpublished
 data
 indicate
 that
 the
 most
 
significant
 genetic
 associations
 with
 ASD
 implicate
 lncRNAs.
 The
 contributions
 of
 these
 lncRNAs
 
to
 ASD
 etiology
 need
 to
 be
 established
 mechanistically
 before
 antisense
 therapies
 for
 ASD
 can
 
be
 developed.
 Antisense
 therapy
 has
 proven
 effective
 for
 other
 common
 disorders,
 including
 
cancer
 and
 HIV.
 Antisense
 siRNA
 treatments
 for
 HIV
 and
 cancer
 are
 effective
 at
 inhibiting
 the
 
progression
 of
 the
 diseases
 in
 adults.
 Application
 of
 antisense
 siRNA
 to
 treat
 developmental
 
brain
 disorders
 poses
 unique
 challenges,
 both
 medical
 and
 societal.
 The
 medical
 challenges
 
include
 timing
 and
 delivery
 of
 antisense
 therapy
 to
 specific
 brain
 regions
 at
 the
 correct
 
developmental
 stages.
 The
 societal
 challenges
 include
 the
 parental
 decision
 to
 alter
 brain
 
development.
 However,
 given
 the
 burden
 of
 ASD
 on
 both
 families
 and
 society,
 and
 the
 lack
 of
 
effective
 treatment
 options,
 we
 propose
 that
 experiments
 designed
 to
 understand
 potential
 
biological
 interventions
 are
 worthwhile
 (1).
 
30
 

 

 
Chapter
 2:
 Impact
 of
 the
 Autism-­‐Associated
 Long
 Noncoding
 RNA
 MSNP1AS
 on
 Neuronal
 
Architecture
 and
 Gene
 Expression
 in
 Human
 Neural
 Progenitor
 Cells
 
 
This
 paper
 was
 published
 in
 Genes
 (67).
 

 
2.1
 Abstract
 
We
 previously
 identified
 the
 lncRNA,
 MSNP1AS,
 (moesin
 pseudogene
 1,
 antisense)
 as
 a
 
functional
 element
 revealed
 by
 genome
 wide
 significant
 association
 with
 ASD.
 MSNP1AS
 
expression
 was
 increased
 in
 the
 postmortem
 cerebral
 cortex
 of
 individuals
 with
 ASD
 and
 
particularly
 in
 individuals
 with
 the
 ASD-­‐associated
 genetic
 markers
 on
 chromosome
 5p14.1.
 
Here,
 we
 mimicked
 the
 overexpression
 of
 MSNP1AS
 observed
 in
 postmortem
 ASD
 cerebral
 
cortex
 in
 human
 neural
 progenitor
 cell
 lines
 to
 determine
 the
 impact
 on
 neurite
 complexity
 and
 
gene
 expression.
 ReNcell
 CX
 and
 SK-­‐N-­‐SH
 were
 transfected
 with
 an
 overexpression
 vector
 
containing
 full-­‐length
 MSNP1AS.
 Neuronal
 complexity
 was
 determined
 by
 the
 number
 and
 
length
 of
 neuronal
 processes.
 Gene
 expression
 was
 determined
 by
 strand-­‐specific
 RNA
 
sequencing.
 MSNP1AS
 overexpression
 decreased
 neurite
 number
 and
 neurite
 length
 in
 both
 
human
 neural
 progenitor
 cell
 lines.
 RNA
 sequencing
 revealed
 changes
 in
 gene
 expression
 in
 
proteins
 involved
 in
 two
 biological
 processes:
 protein
 synthesis
 and
 chromatin
 remodeling.
 
These
 data
 indicate
 that
 overexpression
 of
 the
 ASD-­‐associated
 lncRNA
 MSNP1AS
 alters
 the
 
number
 and
 length
 of
 neuronal
 processes.
 The
 mechanisms
 by
 which
 MSNP1AS
 overexpression
 
impacts
 neuronal
 differentiation
 may
 involve
 protein
 synthesis
 and
 chromatin
 structure.
 These
 
31
 

 
same
 biological
 processes
 are
 also
 implicated
 by
 rare
 mutations
 associated
 with
 ASD,
 
suggesting
 convergent
 molecular
 mechanisms.
 

 
2.2
 Introduction
 
Autism
 spectrum
 disorder
 (ASD)
 is
 a
 neurodevelopmental
 disorder
 characterized
 by
 deficits
 
in
 social
 communication
 and
 repetitive
 behaviors
 with
 restricted
 interests
 (1,
 140).
 Highly
 
penetrant,
 rare,
 de
 novo
 loss
 of
 function
 mutations
 have
 been
 associated
 with
 ASD,
 and
 many
 
of
 the
 associated
 genes
 converge
 upon
 the
 biological
 processes
 of
 protein
 synthesis
 and
 
chromatin
 structure
 (141,
 142).
 However,
 40%
 of
 the
 heritability
 of
 ASD
 resides
 in
 common
 
genetic
 variants
 (143,
 144).
 The
 first
 reported
 common
 genetic
 variants
 with
 genome-­‐wide
 
significant
 association
 with
 ASD
 mapped
 to
 a
 small
 cluster
 of
 chromosome
 5p14.1,
 including
 the
 
common
 genetic
 variant
 rs4307059
 with
 ASD
 association
 of
 P=10
−10

 (26).
 The
 same
 rs4307059
 
allele
 was
 also
 identified
 as
 a
 predictor
 of
 stereotyped
 conversation
 and
 poorer
 communication
 
skills
 in
 a
 population-­‐based
 sample
 of
 >7,000
 individuals
 (73),
 suggesting
 that
 rs4307059
 may
 
be
 a
 quantitative
 trait
 locus
 for
 social
 communication
 phenotypes.
 We
 identified
 a
 3.9
 kb
 
lncRNA
 that
 is
 transcribed
 directly
 at
 the
 site
 of
 the
 chromosome
 5p14.1
 ASD
 association
 signal
 
(74).
 The
 lncRNA
 is
 encoded
 by
 the
 opposite
 (anti-­‐sense)
 strand
 of
 moesin
 pseudogene
 1
 
(MSNP1),
 and
 is
 thus
 designated
 MSNP1AS
 (moesin
 pseudogene
 1,
 anti-­‐sense).
 Expression
 of
 
MSNP1AS
 in
 the
 postmortem
 temporal
 cortex
 is
 increased
 12.7-­‐fold
 in
 individuals
 with
 ASD
 and
 
increased
 22-­‐fold
 in
 individuals
 with
 the
 rs4307059
 risk
 allele
 (74).
 Thus,
 our
 discovery
 revealed
 
a
 lncRNA,
 which
 based
 on
 the
 highly
 significant
 genetic
 association
 findings
 (26),
 contributes
 to
 
ASD
 risk
 (74).
 
32
 

 
MSNP1AS
 is
 94%
 identical
 and
 anti-­‐sense
 to
 the
 X
 chromosome
 transcript
 MSN,
 which
 
encodes
 a
 protein
 (moesin)
 that
 regulates
 neuronal
 architecture
 (145-­‐149).
 The
 lncRNA
 
MSNP1AS
 binds
 MSN
 and
 its
 over-­‐expression
 in
 cell
 lines
 caused
 significant
 decreases
 in
 moesin
 
protein
 (74),
 which
 influences
 neuronal
 process
 stability
 (147).
 Based
 on
 these
 observations,
 a
 
direct
 hypothesis
 is
 that
 overexpression
 of
 MSNP1AS
 in
 progenitors
 of
 cortical
 projection
 
neurons
 will
 decrease
 neuronal
 complexity,
 consistent
 with
 observations
 of
 decreased
 long
 
distance
 connectivity
 observed
 in
 the
 cerebral
 cortex
 of
 individuals
 with
 ASD
 (150-­‐155).
 We
 
tested
 this
 hypothesis
 by
 transfecting
 differentiating
 human
 neural
 progenitor
 cells
 with
 an
 
overexpression
 vector
 that
 drives
 full-­‐length
 MSNP1AS
 and
 directly
 measuring
 the
 number
 and
 
length
 of
 the
 neurites.
 We
 demonstrated
 previously
 that
 one
 mechanism
 by
 which
 MSNP1AS
 
may
 alter
 neuronal
 complexity
 is
 by
 decreasing
 the
 expression
 of
 moesin
 protein
 (74).
 However,
 
alternative
 hypotheses
 about
 the
 mechanism(s)
 of
 the
 ASD-­‐associated
 lncRNA
 MSNP1AS
 had
 
not
 been
 tested.
 While
 our
 previous
 experiments
 revealed
 that
 MSNP1AS
 over-­‐expression
 
decreased
 expression
 of
 moesin
 protein
 (74),
 it
 was
 not
 yet
 clear
 if
 MSNP1AS
 regulated
 the
 
expression
 of
 the
 MSN
 transcript
 or
 if
 MSNP1AS
 over-­‐expression
 altered
 the
 expression
 of
 
other
 genes.
 Therefore,
 we
 performed
 unbiased
 RNA
 sequencing
 on
 differentiating
 human
 
neural
 progenitor
 cells
 to
 determine
 the
 impact
 of
 increased
 MSNP1AS
 on
 gene
 expression.
 
These
 experiments
 mimic
 the
 increased
 expression
 of
 MSNP1AS
 observed
 in
 postmortem
 
brains
 of
 individuals
 with
 ASD
 and
 reveal
 multiple
 mechanisms
 by
 which
 the
 lncRNA
 MSNP1AS
 
may
 contribute
 to
 altered
 neuronal
 architecture.
 

 

   
 
33
 

 
2.3
 Materials
 and
 Methods
 
2.3.1
 Cell
 Culture
 
The
 human
 neural
 progenitor
 cell
 lines
 SK-­‐N-­‐SH
 cells
 (American
 Type
 Culture
 Collection,
 
Manassas,
 VA,
 USA)
 and
 ReNcell
 CX
 cells
 (Millipore,
 Billerica,
 MA,
 USA)
 were
 cultured
 according
 
to
 the
 manufacturer’s
 protocols
 and
 maintained
 in
 a
 75
 cm
2

 flask
 at
 37
 °C
 and
 5%
 CO
2
.
 When
 
cells
 were
 75%
 confluent,
 the
 human
 neural
 progenitor
 cells
 were
 subcultured
 to
 a
 density
 of
 1
 
×
 10
6

 cells
 per
 75
 cm
2

 flask
 for
 harvests
 24
 h
 post-­‐transfection,
 and
 5
 ×
 10
5

 cells
 per
 75
 cm
2

 flask
 
for
 harvests
 72
 h
 post-­‐transfection.
 
2.3.2
 Transfection
 of
 Over-­‐Expression
 Constructs
 
Full-­‐length
 MSNP1AS
 was
 inserted
 into
 pIRES2-­‐AcGFP
 (Clontech)
 (74),
 a
 mammalian
 
over-­‐expression
 construct.
 Each
 experiment
 was
 made
 up
 of
 one
 pIRES2-­‐AcGFP
 negative
 
control
 transfection
 and
 one
 pIRES2-­‐AcGFP-­‐MSNP1AS
 transfection.
 Cells
 were
 transfected
 using
 
Amaxa
 Nucleofector
 (Lonza,
 Walkersville,
 MD,
 USA)
 technology,
 using
 2
 µg
 of
 vector
 per
 well,
 
and
 subcultured
 into
 6-­‐well
 plates.
 One
 milliliter
 of
 fresh
 prewarmed
 media
 was
 added
 to
 each
 
well
 and
 the
 cells
 were
 centrifuged
 twice
 at
 130×
 g
 for
 10
 min
 for
 the
 SK-­‐N-­‐SH
 cells
 and
 300×
 g
 
for
 5
 min
 for
 the
 ReNcell
 CX
 cells
 with
 PBS
 washes
 of
 7
 mL
 and
 4
 mL,
 respectively.
 The
 cell
 pellet
 
was
 resuspended
 in
 Nucleofector
 solution
 containing
 a
 supplement
 using
 1
 ×
 10
6

 cells
 per
 well.
 
One
 hundred
 microliters
 of
 cell/Nucleofector
 solution
 was
 added
 to
 a
 new
 cuvette
 and
 placed
 
in
 the
 Amaxa
 Nucleofector
 using
 the
 T-­‐16
 program.
 Five
 hundred
 microliters
 of
 fresh
 
prewarmed
 media
 was
 added
 to
 the
 cuvette,
 mixed,
 and
 transferred
 to
 the
 well
 with
 a
 
disposable
 pipet.
 A
 two
 mL
 culture
 medium
 was
 added
 to
 the
 transfected
 cells
 in
 the
 6-­‐well
 
34
 

 
plate.
 The
 cells
 were
 incubated
 at
 37
 °C
 in
 5%
 CO
2

 until
 harvest.
 Each
 experiment
 was
 repeated
 
four
 times.
 
2.3.3
 Imaging
 
At
 24
 and
 72
 h
 post-­‐transfection,
 human
 neural
 progenitors
 were
 viewed
 using
 an
 
Olympus
 CKX41
 inverted
 microscope
 and
 an
 attached
 Q
 Imaging
 QICAM
 Fast
 1394
 Digital
 
Camera
 captured
 digital
 images.
 Randomly
 selected
 fields
 of
 the
 6-­‐well
 plate
 were
 imaged.
 
Three
 randomly
 selected
 fields
 were
 imaged
 in
 each
 experiment.
 Neurite
 length
 and
 number
 of
 
each
 GFP-­‐positive
 cell
 within
 the
 field
 were
 quantified
 using
 Autoneuron
 software.
 Only
 
neurons
 unobstructed
 by
 other
 neurons
 or
 debris
 were
 analyzed.
 Statistical
 significance
 was
 
calculated
 using
 the
 Mann-­‐Whitney
 U
 test.
 
2.3.4
 Neural
 Progenitor
 Cell
 Harvest
 for
 RNA
 Purification
 
After
 24
 h
 and
 72
 h,
 the
 culture
 medium
 was
 removed
 and
 discarded.
 The
 cells
 were
 
washed
 with
 one
 mL
 of
 PBS
 wash
 and
 the
 PBS
 wash
 was
 aspirated.
 One
 mL
 of
 Trypsin/EDTA
 
solution
 for
 SK-­‐N-­‐SH
 cells
 and
 Accutase
 for
 ReNcell
 CX
 cells
 was
 added
 to
 each
 well
 and
 the
 cells
 
were
 incubated
 for
 4
 min.
 Five
 mL
 of
 cell
 type-­‐specific
 medium
 was
 added
 and
 cells
 were
 
triturated
 and
 transferred
 to
 15
 mL
 conical
 tubes.
 SK-­‐N-­‐SH
 cells
 were
 centrifuged
 at
 130×
 g
 for
 
10
 min
 and
 ReNcell
 CX
 cells
 were
 centrifuged
 at
 300×
 g
 for
 5
 min.
 The
 supernatant
 was
 
aspirated
 and
 5
 mL
 of
 PBS
 added.
 The
 cell
 pellet
 was
 triturated
 and
 centrifuged
 at
 130×
 g
 for
 10
 
min
 and
 300×
 g
 for
 5
 min
 respectively.
 The
 PBS
 wash
 was
 aspirated
 and
 1
 mL
 of
 fresh
 PBS
 was
 
added.
 Half
 of
 the
 solution
 was
 transferred
 to
 each
 of
 two
 1.5
 mL
 eppendorf
 tubes.
 The
 tubes
 
were
 centrifuged
 at
 4
 °C
 at
 130×
 g
 for
 10
 min
 for
 SK-­‐N-­‐SH
 cells
 and
 300×
 g
 for
 5
 min
 for
 ReNcell
 
CX
 cells.
 The
 supernatant
 was
 decanted
 and
 the
 cell
 pellets
 were
 frozen
 at
 −80
 °C.
 
35
 

 
2.3.5
 RNA
 Purification
 
The
 Qiagen
 RNEasy
 kit
 was
 used
 to
 isolate
 total
 RNA
 using
 vacuum
 technology
 according
 
to
 the
 manufacturer’s
 protocol
 (Qiagen,
 Valencia,
 CA,
 USA).
 The
 RNA
 was
 eluted
 with
 35
 
microliters
 of
 RNAse-­‐free
 water
 and
 quantified
 using
 the
 NanoDrop
 ND-­‐1000
 
Spectrophotometer
 (v3.1.2;
 Thermo
 Fisher
 Scientific,
 Waltham,
 MA,
 USA).
 The
 RNA
 was
 stored
 
at
 −20
 °C.
 
2.3.6
 Quantitative
 RT-­‐PCR
 (qRT-­‐PCR)
 
To
 confirm
 overexpression
 of
 MSNP1AS,
 consistent
 with
 the
 ~12-­‐fold
 increase
 observed
 
in
 postmortem
 brains
 of
 individuals
 with
 ASD,
 cDNA
 was
 synthesized
 using
 the
 SuperScript
 III
 
First-­‐Strand
 Synthesis
 System
 for
 qRT-­‐PCR
 protocol
 (Invitrogen/Thermo
 Fisher
 Scientific,
 
Waltham,
 MA,
 USA).
 Five
 hundred
 nanograms
 of
 RNA
 were
 used
 to
 make
 one
 and
 a
 half
 
reaction
 volumes
 (35
 μL)
 of
 cDNA.
 The
 cDNA
 was
 stored
 at
 −20
 °C.
 The
 qRT-­‐PCR
 protocol
 
described
 in
 Kerin
 et
 al.
 (74)
 was
 used
 to
 validate
 over-­‐expression
 of
 MSNP1AS.
 
2.3.7
 Construction
 of
 Strand-­‐Specific,
 Ribosomal
 RNA
 Depleted
 RNA
 Sequencing
 Libraries
 
Directional
 RNA-­‐Seq
 libraries
 were
 prepared
 for
 Illumina
 HiSeq
 2000
 sequencing
 using
 
the
 Stranded
 Total
 TruSeq
 RNA
 Sample
 Preparation
 kit
 with
 Ribo-­‐Zero
 Gold
 (Illumina)
 using
 the
 
manufacturer’s
 protocol
 using
 the
 Hamilton
 Starlet
 Liquid
 Handling
 robot.
 One
 nanogram
 of
 
RNA
 was
 used
 for
 each
 sample.
 In
 brief,
 Ribo-­‐Zero
 was
 used
 to
 deplete
 cytoplasmic
 and
 
mitochondrial
 rRNA
 from
 total
 RNA.
 The
 depleted
 RNA
 was
 fragmented
 and
 primed
 with
 
random
 hexamers
 to
 synthesize
 first
 strand
 cDNA
 using
 Superscript
 II
 (Life
 Technologies,
 
Carlsbad,
 CA,
 USA).
 Next,
 the
 second
 strand
 was
 synthesized,
 incorporating
 dUTP
 in
 place
 of
 
dTTP.
 A
 single
 “A”
 base
 was
 added
 to
 the
 3ʹ′
 ends
 of
 the
 fragments
 and
 the
 indexed
 adaptors
 
36
 

 
were
 ligated
 to
 the
 ends
 of
 the
 ds
 cDNA
 to
 prepare
 them
 for
 hybridization
 onto
 the
 flow
 cell.
 
PCR
 was
 used
 to
 selectively
 enrich
 the
 fragments
 with
 ligated
 adapters
 and
 to
 amplify
 the
 
amount
 of
 DNA
 in
 the
 library.
 The
 libraries
 were
 produced
 in
 a
 96-­‐well
 format
 and
 quality
 
controlled
 using
 the
 Agilent
 Technologies
 2200
 TapeStation
 Instrument.
 Libraries
 were
 pooled
 
(four
 samples
 per
 lane)
 and
 sequenced
 on
 Illumina
 HiSeq
 2000
 to
 a
 targeted
 depth,
 generating
 
an
 average
 of
 20
 million
 paired-­‐end
 50-­‐cycle
 reads
 for
 each
 sample
 (Table
 2.1
 and
 2.2).
 
Table
 2.1.
 Read
 mapping
 of
 RNA
 Sequencing
 Libraries
 from
 SK-­‐N-­‐SH
 cells
 

 
  Number
 of
 reads
  Number
 of
 reads
 mapped
  %
 mapped
 
SK-­‐N-­‐SH
 24
 hr
   
 
   
 
   
 
 
control
 biological
 replicate
 1
  11,099,318
  10,746,390
  96.8
 
control
 biological
 replicate
 2
  20,496,459
  19,422,292
  94.8
 
control
 biological
 replicate
 3
  10,824,013
  10,458,919
  96.6
 
control
 biological
 replicate
 4
  31,400,592
  29,709,967
  94.6
 
Average
  18,455,096
  17,584,392
  95.7
 

 
 

   
 

 
 
MSNP1AS
 OE
 biological
 replicate
 1
  19,971,956
  19,337,134
  96.8
 
MSNP1AS
 OE
 biological
 replicate
 2
  13,807,707
  12,881,017
  93.3
 
MSNP1AS
 OE
 biological
 replicate
 3
  18,326,170
  17,660,733
  96.4
 
MSNP1AS
 OE
 biological
 replicate
 4
  10,900,335
  10,366,157
  95.1
 
Average
  15,751,542
  15,061,260
  95.4
 

 
 

   
 

 
 
SK-­‐N-­‐SH
 72
 hr
   
 
   
 
   
 
 
control
 biological
 replicate
 1
  46,378,868
  44,126,735
  95.1
 
control
 biological
 replicate
 2
  22,219,398
  20,950,969
  94.3
 
control
 biological
 replicate
 3
  25,049,815
  24,123,433
  96.3
 
control
 biological
 replicate
 4
  24,021,135
  23,272,061
  96.9
 
Average
  29,417,304
  28,118,300
  95.7
 

 
 

   
 

 
 
MSNP1AS
 OE
 biological
 replicate
 1
  17,844,865
  17,128,866
  96.0
 
MSNP1AS
 OE
 biological
 replicate
 2
  10,176,928
  9,651,721
  94.8
 
MSNP1AS
 OE
 biological
 replicate
 3
  14,363,510
  13,878,004
  96.6
 
MSNP1AS
 OE
 biological
 replicate
 4
  15,219,358
  14,512,024
  95.4
 
Average
  14,401,165
  13,792,654
  95.7
 

 
   
 
37
 

 
Table
 2.2.
 Read
 mapping
 of
 RNA
 Sequencing
 Libraries
 from
 ReNcell
 CX
 cells
 
ReNcell
 CX
 24
 hr
   
 
   
 
   
 
 
control
 biological
 replicate
 1
  27,003,948
  26,267,471
  97.3
 
control
 biological
 replicate
 2
  20,860,192
  19,778,456
  94.8
 
control
 biological
 replicate
 3
  15,548,803
  14,713,657
  94.6
 
control
 biological
 replicate
 4
  26,490,502
  25,268,925
  95.4
 
Average
  22,475,861
  21,507,127
  95.5
 

 
 

   
 

 
 
MSNP1AS
 OE
 biological
 replicate
 1
  22,807,209
  22,258,018
  97.6
 
MSNP1AS
 OE
 biological
 replicate
 2
  17,922,225
  17,200,036
  96.0
 
MSNP1AS
 OE
 biological
 replicate
 3
  13,290,230
  12,659,451
  95.3
 
MSNP1AS
 OE
 biological
 replicate
 4
  16,948,688
  16,375,844
  96.6
 
Average
  17,742,088
  17,123,337
  96.4
 

 
 

   
 

 
 
ReNcell
 CX
 72
 hr
   
 
   
 
   
 
 
control
 biological
 replicate
 1
  20,666,642
  20,014,343
  96.8
 
control
 biological
 replicate
 2
  14,360,503
  13,888,459
  96.7
 
control
 biological
 replicate
 3
  33,262,563
  31,348,111
  94.2
 
control
 biological
 replicate
 4
  35,502,944
  34,507,246
  97.2
 
Average
  25,948,163
  24,939,540
  96.2
 

 
 

   
 

 
 
MSNP1AS
 OE
 biological
 replicate
 1
  11,987,370
  11,670,193
  97.4
 
MSNP1AS
 OE
 biological
 replicate
 2
  12,533,511
  11,987,081
  95.6
 
MSNP1AS
 OE
 biological
 replicate
 3
  25,523,571
  24,097,726
  94.4
 
MSNP1AS
 OE
 biological
 replicate
 4
  22,607,983
  21,685,434
  95.9
 
Average
  18,163,109
  17,360,109
  95.8
 

 
2.3.8
 Data
 Analysis
 
Data
 analysis
 was
 performed
 using
 TopHat
 (version
 2.0.10;
 
https://ccb.jhu.edu/software/tophat/index.shtml)
 (156)
 to
 align
 the
 Illumina
 short
 reads
 
against
 the
 reference
 human
 genome
 ENSEMBL
 GrCH38
 version
 81.
 Sequence
 alignments
 were
 
generated
 as
 BAM
 files
 (157),
 and
 then
 Cuffdiff
 (version
 2.2.1;
 http://cole-­‐trapnell-­‐
lab.github.io/cufflinks/)
 (158)
 was
 used
 to
 summarize
 the
 gene
 expression
 values
 as
 FPKM
 
measures.
 The
 gene
 expression
 of
 samples
 with
 MSNP1AS
 over-­‐expression
 were
 compared
 to
 
the
 gene
 expression
 of
 the
 negative
 control
 experiment
 samples
 to
 find
 other
 differentially-­‐
38
 

 
expressed
 genes.
 Cuffdiff
 was
 also
 used
 to
 calculate
 the
 expression
 fold
 change,
 p-­‐values
 and
 
FDR
 values.
 Genes
 with
 p
 <
 0.05
 were
 used
 as
 the
 input
 for
 the
 Database
 for
 Annotation,
 
Visualization
 and
 Integrated
 Discovery
 (DAVID)
 (version
 6.7;
 Leidos
 Biomedical
 Research,
 Inc.,
 
Frederick,
 MD,
 USA)
 functional
 annotation
 (159,
 160).
 

 
2.4
 Results
 
2.4.1
 Overexpression
 of
 MSNP1AS
 Decreased
 Neurite
 Number
 and
 Length
 in
 SK-­‐N-­‐SH
 and
 
ReNcell
 CX
 Human
 Neural
 Progenitor
 Cells
 
To
 mimic
 the
 overexpression
 of
 MSNP1AS
 observed
 in
 postmortem
 brains
 of
 individuals
 
with
 ASD,
 MSNP1AS
 was
 overexpressed
 ~50-­‐fold
 (Table
 2.3)
 in
 human
 neural
 progenitor
 cells
 
using
 a
 human
 overexpression
 vector
 containing
 the
 full
 length
 MSNP1AS
 transcript.
 Human
 
neural
 progenitor
 cells
 that
 overexpressed
 MSNP1AS
 had
 decreased
 neurite
 length
 compared
 
to
 human
 neural
 progenitor
 cells
 transfected
 with
 the
 empty
 vector
 control
 (Figure
 2.1
 and
 
Figure
 2.2).
 In
 SK-­‐N-­‐SH
 cells,
 neurite
 length
 was
 reduced
 6-­‐fold
 (p
 =
 1.1
 ×
 10
−2
)
 at
 24
 h
 post-­‐
transfection
 and
 5-­‐fold
 (p
 =
 6.0
 ×
 10
−4
)
 at
 72
 h
 post-­‐transfection
 in
 neural
 progenitor
 cells
 
transfected
 with
 the
 MSNP1AS
 over-­‐expression
 vector
 compared
 to
 controls
 (Figure
 2.2A,B).
 
Similarly,
 MSNP1AS
 overexpression
 in
 ReNcell
 CX
 human
 neural
 progenitor
 cells
 caused
 a
 
significant
 decrease
 in
 neurite
 length
 at
 72
 h
 post-­‐transfection
 by
 12.5-­‐fold
 (p
 =
 2.5
 ×
 10
−2
),
 but
 
the
 decreased
 neurite
 length
 observed
 at
 24
 h
 post-­‐transfection
 was
 not
 significant
 (2.3-­‐fold;
 p
 
=
 2.9
 ×
 10
−1
)
 (Figure
 2.2C,D).
 

   
 
39
 

 
Table
 2.3.
 Relative
 expression
 of
 MSNP1AS
 after
 transfection
 with
 overexpression
 construct.
 

 
  MSNP1AS
 Relative
 Expression
 
ReN24
 -­‐
 1
  30.84
   
 
   
 
 
ReN24
 -­‐
 2
  25.58
 

 

 
 
ReN24
 -­‐
 3
  5.17
 

 

 
 
ReN24
 -­‐
 4
  145.82
 

 

 
 

 
   
 
  average
  51.85
 

 
 

   
 

 
 
ReN72
 -­‐
 1
 
  7.29
   
 
   
 
 
ReN72
 -­‐
 2
  29.08
 

 

 
 
ReN72
 -­‐
 3
  37.17
 

 

 
 
ReN72
 -­‐
 4
  39.34
 

 

 
 

 
   
 
  average
  28.22
 

 
 

   
 

 
 
SK24
 -­‐
 1
 
  4855.15
   
 
   
 
 
SK24
 -­‐
 2
  1586.70
 

 

 
 
SK24
 -­‐
 3
  87.60
 

 

 
 
SK24
 -­‐
 4
  1596.98
 

 

 
 

 
   
 
  average
  2031.61
 

 
 

   
 

 
 
SK72
 -­‐
 1
  816.43
   
 
   
 
 
SK72
 -­‐
 2
  1170.18
 

 

 
 
SK72
 -­‐
 3
  6254.52
 

 

 
 
SK72
 -­‐
 4
  142.42
 

 

 
 

 
   
 
  average
  2095.89
 

 
 

   
 

 
 
total
 average
  1051.89
   
 
   
 
 

 

 

   
 
40
 

 

 
Figure
 2.1.
 Representative
 ReNcell
 CX
 cells
 at
 24
 h
 post-­‐transfection.
 The
 human
 neural
 
progenitor
 cells
 were
 transfected
 with
 (A)
 an
 empty
 vector
 control
 or
 (B)
 the
 MSNP1AS
 
overexpression
 vector.
 For
 each
 experiment,
 the
 transfection
 of
 MSNP1AS
 over-­‐expression
 or
 
control
 vector
 was
 repeated
 four
 times.
 Three
 randomly
 selected
 fields
 from
 each
 of
 the
 four
 
replicates
 were
 imaged
 and
 all
 isolated
 GFP-­‐positive
 cells
 within
 the
 imaged
 fields
 were
 
examined
 for
 neurite
 length
 and
 neurite
 number
 using
 Autoneuron
 software.
 

 

 
Figure
 2.2.
 Overexpression
 of
 MSNP1AS
 decreases
 neurite
 number
 and
 length.
 SK-­‐N-­‐SH
 cells
 
that
 overexpressed
 MSNP1AS
 had
 decreased
 neurite
 length
 after
 (A)
 24
 h
 and
 (B)
 72
 h.
 ReNcell
 
CX
 cells
 that
 overexpressed
 MSNP1AS
 also
 had
 decreased
 neurite
 length
 after
 (D)
 72
 h
 but
 the
 
decreased
 neurite
 length
 after
 (C)
 24
 h
 was
 not
 significant.
 SK-­‐N-­‐SH
 cells
 that
 overexpressed
 
MSNP1AS
 had
 fewer
 neurites
 per
 cell
 after
 (E)
 24
 h
 and
 a
 trend
 toward
 reduced
 neurites
 per
 
cell
 after
 (F)
 72
 h.
 ReNcell
 CX
 cells
 that
 overexpressed
 MSNP1AS
 also
 had
 trend
 toward
 fewer
 
neurites
 per
 cell
 after
 (H)
 72
 h,
 but
 the
 neurite
 number
 decrease
 after
 (G)
 24
 h
 was
 not
 
significant
 (*
 p
 <
 0.05,
 Mann-­‐Whitney
 U
 test).
 
MSNP1AS
 overexpression
 also
 caused
 a
 reduction
 in
 the
 number
 of
 neurites
 in
 human
 neural
 
progenitor
 cells.
 In
 SK-­‐N-­‐SH
 cells,
 MSNP1AS
 overexpression
 caused
 an
 8.3-­‐fold
 reduction
 in
 
neurite
 number
 at
 24
 h
 post-­‐transfection
 (p
 =
 2.1
 ×
 10
−2
)
 and
 a
 trend
 toward
 reduced
 neurite
 
number
 at
 72
 h
 post-­‐transfection
 (4.4-­‐fold;
 p
 =
 6.2
 ×
 10
−2
)
 (Figure
 2.2E,F).
 Similarly,
 MSNP1AS
 
overexpression
 in
 ReNcell
 CX
 cells
 caused
 a
 trend
 toward
 fewer
 neurite
 number
 at
 72
 h
 post-­‐
transfection
 (4.6-­‐fold;
 p
 =
 1.0
 ×
 10
−1
),
 but
 the
 neurite
 number
 decrease
 at
 24
 h
 post-­‐
transfection
 was
 not
 significant
 (1.3-­‐fold;
 p
 =
 7.9
 ×
 10
−1
)
 (Figure
 2.2G,H).
 
41
 

 
2.4.2
 Genome-­‐Wide
 Changes
 in
 Gene
 Expression
 Following
 MSNP1AS
 Overexpression
 in
 
Human
 Neural
 Progenitor
 Cells
 
Changes
 in
 the
 transcriptome
 caused
 by
 MSNP1AS
 overexpression
 may
 provide
 insight
 
into
 the
 mechanisms
 by
 which
 an
 increase
 of
 MSNP1AS
 influences
 neuronal
 architecture.
 
Therefore,
 RNA
 sequencing
 was
 used
 to
 perform
 genome-­‐wide
 transcriptome
 profiling
 on
 the
 
two
 human
 neural
 progenitor
 cells
 lines
 SK-­‐N-­‐SH
 and
 ReNcell
 CX
 cells
 with
 and
 without
 
MSNP1AS
 overexpression.
 For
 all
 experiments,
 the
 average
 number
 of
 reads
 ranged
 from
 14–29
 
million
 with
 96%
 of
 the
 reads
 mapping
 to
 the
 genome
 (Table
 2.1
 and
 2.2).
 No
 change
 in
 gene
 
expression
 survived
 Bonferroni
 correction
 for
 multiple
 comparisons
 in
 any
 of
 the
 experiments
 
(Table
 2.4,
 2.5,
 2.6,
 and
 2.7).
 However,
 each
 of
 the
 experiments
 revealed
 100–400
 genes
 with
 
nominally
 significant
 changes
 in
 gene
 expression
 (p
 <
 0.05).
 All
 downstream
 analyses
 were
 
performed
 with
 the
 sets
 of
 genes
 that
 had
 significant
 (p
 <
 0.05)
 changes
 in
 expression
 following
 
MSNP1AS
 over-­‐expression.
 Differential
 gene
 expression
 analysis
 in
 SK-­‐N-­‐SH
 cells
 revealed
 157
 
differentially
 expressed
 genes
 (p
 <
 0.05)
 at
 24
 h
 post-­‐transfection
 and
 351
 genes
 differentially
 
expressed
 (p
 <
 0.05)
 at
 72
 h
 post-­‐transfection
 (Table
 2.4
 and
 2.5).
 In
 ReNcell
 CX
 cells,
 MSNP1AS
 
overexpression
 revealed
 267
 genes
 differentially
 expressed
 at
 24
 h
 post-­‐transfection
 and
 164
 
genes
 differentially
 expressed
 at
 72
 h
 post-­‐transfection
 (Table
 2.6
 and
 2.7).
 No
 single
 gene
 was
 
significantly
 (p
 <
 0.05)
 altered
 in
 each
 of
 the
 four
 experimental
 conditions
 (Table
 2.4,
 2.5,
 2.6,
 
and
 2.7).
 
 
MSNP1AS
 overexpression
 did
 not
 alter
 expression
 of
 MSN,
 the
 transcript
 that
 encodes
 
moesin
 protein
 and
 is
 bound
 by
 the
 MSNP1AS
 lncRNA.
 In
 SK-­‐N-­‐SH
 cells,
 MSN
 transcript
 was
 
slightly
 increased
 at
 24
 h
 post-­‐transfection
 (1.17-­‐fold;
 p
 =
 3.0
 ×
 10
−2
)
 and
 at
 72
 h
 post-­‐
 
 
42
 

 
Table
 2.4.
 Top
 thirty
 differentially
 expressed
 genes
 in
 SK-­‐N-­‐SH
 cells
 24
 hrs
 after
 
MSNP1AS
 overexpression.
 
gene
  locus
 
log2
 
(fold_change)
  p_value
  q_value
 
CXCL8
  4:73740505-­‐73743716
  -­‐1.85
  5.00E-­‐05
  1.07E-­‐01
 
ADGRG1
  16:57610651-­‐57665580
  1.08
  5.00E-­‐05
  1.07E-­‐01
 
TXNIP
  1:145992434-­‐145996600
  1.27
  5.00E-­‐05
  1.07E-­‐01
 
BIRC3
  11:102317449-­‐102339403
  1.33
  5.00E-­‐05
  1.07E-­‐01
 
INSM2
  14:35534041-­‐35537054
  1.41
  5.00E-­‐05
  1.07E-­‐01
 
MIR320A
  8:22244961-­‐22245043
  4.40
  5.00E-­‐05
  1.07E-­‐01
 
SPP1
  4:87975649-­‐87983426
  -­‐2.48
  1.00E-­‐04
  1.83E-­‐01
 
CHGA
  14:92923079-­‐92935293
  1.06
  1.50E-­‐04
  2.40E-­‐01
 
FST
  5:53480408-­‐53487134
  -­‐1.37
  3.00E-­‐04
  4.26E-­‐01
 
PPP1R15A
  19:48872391-­‐48876057
  0.96
  4.50E-­‐04
  5.76E-­‐01
 
CBR3-­‐AS1
  21:36131766-­‐36294274
  -­‐1.74
  5.00E-­‐04
  5.81E-­‐01
 
TFPI2
  7:93591572-­‐93911265
  -­‐0.93
  5.50E-­‐04
  5.86E-­‐01
 
NTS
  12:85874294-­‐85882992
  -­‐0.92
  6.50E-­‐04
  6.40E-­‐01
 
NPY
  7:24284162-­‐24291865
  -­‐0.79
  8.00E-­‐04
  7.15E-­‐01
 
GADD45B
  19:2476121-­‐2478259
  1.42
  8.50E-­‐04
  7.15E-­‐01
 
TMEM158
  3:45224465-­‐45226278
  -­‐0.79
  9.00E-­‐04
  7.15E-­‐01
 
SLC5A7
  2:107986522-­‐108013994
  1.13
  9.50E-­‐04
  7.15E-­‐01
 
TRAF1
  9:120902392-­‐120929173
  1.34
  1.30E-­‐03
  8.75E-­‐01
 
ATF3
  1:212565333-­‐212620777
  1.39
  1.30E-­‐03
  8.75E-­‐01
 
KLF10
  8:102648778-­‐102655902
  0.70
  1.55E-­‐03
  9.91E-­‐01
 
OPRD1
  1:28812141-­‐28877336
  0.98
  2.25E-­‐03
  1.00E+00
 
STC1
  8:23841914-­‐23854807
  -­‐1.48
  2.40E-­‐03
  1.00E+00
 
ARRDC4
  15:97960697-­‐97973838
  0.92
  2.40E-­‐03
  1.00E+00
 
EGR1
  5:138465489-­‐138469315
  0.84
  2.85E-­‐03
  1.00E+00
 
AXL
  19:41219202-­‐41261766
  0.68
  2.90E-­‐03
  1.00E+00
 
RP11-­‐69E11.4
  1:39491645-­‐39558707
  2.57
  2.95E-­‐03
  1.00E+00
 
ASS1
  9:130444928-­‐130501274
  -­‐1.52
  3.15E-­‐03
  1.00E+00
 
SULT1E1
  4:69791871-­‐69860152
  1.18
  3.30E-­‐03
  1.00E+00
 
GADD45A
  1:67685060-­‐67688338
  0.70
  3.95E-­‐03
  1.00E+00
 
DPP4
  2:161992240-­‐162075169
  -­‐0.90
  4.70E-­‐03
  1.00E+00
 
PI15
  8:74599774-­‐74866939
  -­‐0.75
  5.05E-­‐03
  1.00E+00
 

 

   
 
43
 

 
Table
 2.5.
 Top
 thirty
 differentially
 expressed
 genes
 in
 SK-­‐N-­‐SH
 cells
 72
 hrs
 after
 
MSNP1AS
 overexpression.
 
gene
  locus
 
log2
 
(fold_change)
  p_value
  q_value
 
Metazoa_SRP
  14:49853615-­‐49853914
  -­‐2.13
  5.00E-­‐05
  1.56E-­‐01
 
RMRP
  9:35657750-­‐35658018
  -­‐1.83
  5.00E-­‐05
  1.56E-­‐01
 
KIFC3
  16:57758216-­‐57863053
  -­‐1.60
  5.00E-­‐05
  1.56E-­‐01
 
PUF60
  8:143816343-­‐143829859
  -­‐1.31
  5.00E-­‐05
  1.56E-­‐01
 
GPR137
  11:64269829-­‐64289500
  -­‐1.72
  1.50E-­‐04
  3.37E-­‐01
 
TMEM134
  11:67461709-­‐67469272
  -­‐1.96
  2.50E-­‐04
  3.37E-­‐01
 
FAM195B
  17:81822360-­‐81833302
  -­‐1.87
  2.50E-­‐04
  3.37E-­‐01
 
YIF1B
  19:38305103-­‐38317273
  -­‐1.67
  2.50E-­‐04
  3.37E-­‐01
 
RPS15
  19:1438357-­‐1440494
  -­‐1.37
  2.50E-­‐04
  3.37E-­‐01
 
PKD1
  16:2039814-­‐2135898
  -­‐1.46
  3.00E-­‐04
  3.37E-­‐01
 
RPL8
  8:144789764-­‐144792587
  -­‐1.08
  3.00E-­‐04
  3.37E-­‐01
 
DCXR
  17:82035135-­‐82037732
  -­‐1.74
  3.50E-­‐04
  3.37E-­‐01
 
RPL36
  19:5674946-­‐5720572
  -­‐1.42
  3.50E-­‐04
  3.37E-­‐01
 
RPS9
  19:54200741-­‐54281184
  -­‐1.31
  4.00E-­‐04
  3.57E-­‐01
 
FKBP2
  11:64241002-­‐64244132
  -­‐1.51
  5.00E-­‐04
  3.91E-­‐01
 
RPPH1
  14:20343047-­‐20357905
  -­‐1.22
  5.00E-­‐04
  3.91E-­‐01
 
RAC3
  17:82031623-­‐82034204
  -­‐2.02
  6.00E-­‐04
  4.17E-­‐01
 
PLA2G4C
  19:48047842-­‐48110817
  -­‐1.51
  6.00E-­‐04
  4.17E-­‐01
 
EEF1A2
  20:63488012-­‐63499315
  -­‐1.38
  6.50E-­‐04
  4.28E-­‐01
 
ACTL6B
  7:100643096-­‐100656461
  -­‐1.85
  7.00E-­‐04
  4.38E-­‐01
 
VGF
  7:101162508-­‐101165593
  -­‐1.38
  7.50E-­‐04
  4.47E-­‐01
 
BCAR1
  16:75226073-­‐75268053
  -­‐1.82
  8.00E-­‐04
  4.55E-­‐01
 
Metazoa_SRP
  14:49862546-­‐49862849
  -­‐1.05
  1.00E-­‐03
  5.44E-­‐01
 
RHBDF1
  16:58058-­‐76355
  -­‐1.79
  1.50E-­‐03
  7.51E-­‐01
 
MRPL55
  1:228106678-­‐228109312
  -­‐1.02
  1.50E-­‐03
  7.51E-­‐01
 
PLP1
  X:103773717-­‐103792619
  -­‐1.20
  1.60E-­‐03
  7.60E-­‐01
 
CCL2
  17:34255217-­‐34257203
  -­‐1.01
  1.65E-­‐03
  7.60E-­‐01
 
HIST1H1E
  6:26156390-­‐26157050
  -­‐1.39
  1.70E-­‐03
  7.60E-­‐01
 
NR4A1
  12:52022831-­‐52059507
  -­‐1.58
  1.80E-­‐03
  7.77E-­‐01
 
TNK2
  3:195863363-­‐195911945
  -­‐1.62
  1.90E-­‐03
  7.92E-­‐01
 

 

   
 
44
 

 
Table
 2.6.
 Top
 thirty
 differentially
 expressed
 genes
 in
 ReNcell
 CX
 cells
 24
 hrs
 
after
 MSNP1AS
 overexpression.
 
gene
  locus
 
log2
 
(fold_change)
  p_value
  q_value
 
MIR663AHG
  20:26186919-­‐26251526
  1.44
  5.00E-­‐05
  6.75E-­‐01
 
CTGF
  6:131948175-­‐132077393
  0.67
  1.50E-­‐04
  8.44E-­‐01
 
TEX101
  19:43401495-­‐43418597
  1.39
  2.00E-­‐04
  8.44E-­‐01
 
SIRT6
  19:4174108-­‐4182604
  1.70
  2.50E-­‐04
  8.44E-­‐01
 
IFI6
  1:27666060-­‐27703063
  -­‐0.65
  3.50E-­‐04
  9.46E-­‐01
 
MX1
  21:41420303-­‐41459214
  -­‐0.86
  5.00E-­‐04
  1.00E+00
 
SLC7A11
  4:138027421-­‐138242349
  0.61
  5.50E-­‐04
  1.00E+00
 
FOSL1
  11:65892048-­‐65900573
  0.99
  6.50E-­‐04
  1.00E+00
 
SNORD3A
  17:19188015-­‐19188714
  -­‐0.54
  8.00E-­‐04
  1.00E+00
 
OLFM1
  9:135075242-­‐135121179
  1.78
  8.00E-­‐04
  1.00E+00
 
NTRK2
  9:84668550-­‐85027070
  0.94
  9.00E-­‐04
  1.00E+00
 
PDE2A
  11:72576140-­‐72674591
  1.28
  1.25E-­‐03
  1.00E+00
 
NFATC2
  20:51386956-­‐51562831
  0.75
  1.80E-­‐03
  1.00E+00
 
EGR1
  5:138465489-­‐138469315
  0.61
  2.15E-­‐03
  1.00E+00
 
MNT
  17:2384059-­‐2401118
  1.04
  2.20E-­‐03
  1.00E+00
 
LGALS3
  14:55124109-­‐55145413
  -­‐0.50
  2.35E-­‐03
  1.00E+00
 
ANKRD52
  12:56230048-­‐56258391
  0.63
  2.35E-­‐03
  1.00E+00
 
EP300
  22:41091785-­‐41197456
  0.53
  2.65E-­‐03
  1.00E+00
 
CLDN1
  3:190305700-­‐190322475
  0.78
  2.90E-­‐03
  1.00E+00
 
C9orf116
  9:135495180-­‐135506447
  0.72
  3.65E-­‐03
  1.00E+00
 
NEXN
  1:77881347-­‐77943895
  0.84
  3.75E-­‐03
  1.00E+00
 
AC010970.2
  Y:10197255-­‐10199103
  0.73
  3.90E-­‐03
  1.00E+00
 
UBR4
  1:19072109-­‐19210276
  0.60
  4.05E-­‐03
  1.00E+00
 
CMIP
  16:81445169-­‐81711762
  0.60
  4.40E-­‐03
  1.00E+00
 
STC2
  5:173314712-­‐173329503
  0.59
  4.45E-­‐03
  1.00E+00
 
CYR61
  1:85580760-­‐85583962
  0.52
  4.75E-­‐03
  1.00E+00
 
IFI44
  1:78649795-­‐78664078
  -­‐0.84
  4.80E-­‐03
  1.00E+00
 
RFX7
  15:56087279-­‐56243266
  0.59
  4.90E-­‐03
  1.00E+00
 
MYC
  8:127735433-­‐127741434
  0.46
  4.95E-­‐03
  1.00E+00
 
TTYH3
  7:2631950-­‐2664802
  0.57
  5.05E-­‐03
  1.00E+00
 

 

   
 
45
 

 
Table
 2.7.
 Top
 thirty
 differentially
 expressed
 genes
 in
 ReNcell
 CX
 cells
 72
 hrs
 after
 
MSNP1AS
 overexpression.
 
gene
  locus
 
log2
 
(fold_change)
  p_value
  q_value
 
SNAP91
  6:83552879-­‐83709691
  -­‐1.80
  3.50E-­‐04
  1.00E+00
 
SCG5
  15:32641675-­‐32697098
  -­‐2.01
  7.00E-­‐04
  1.00E+00
 
DPP7
  9:137110541-­‐137115177
  1.09
  9.50E-­‐04
  1.00E+00
 
NPTX1
  17:80467147-­‐80477843
  -­‐0.92
  1.20E-­‐03
  1.00E+00
 
LIMK1
  7:74082932-­‐74122525
  0.61
  1.20E-­‐03
  1.00E+00
 
EIF4EBP1
  8:38030340-­‐38060365
  0.61
  1.30E-­‐03
  1.00E+00
 
CDR1-­‐AS
  X:140782404-­‐140784871
  -­‐2.25
  1.50E-­‐03
  1.00E+00
 
MTRNR2L8
  11:10507886-­‐10509189
  -­‐0.95
  2.05E-­‐03
  1.00E+00
 
ANKRD1
  10:90912095-­‐90921276
  0.59
  2.30E-­‐03
  1.00E+00
 
ZRANB2-­‐AS2
  1:71081323-­‐72282734
  -­‐1.02
  3.20E-­‐03
  1.00E+00
 
LTBP2
  14:74498169-­‐74612378
  0.46
  3.20E-­‐03
  1.00E+00
 
JUP
  17:41754603-­‐41786931
  0.62
  3.65E-­‐03
  1.00E+00
 
MTRNR2L1
  17:22523110-­‐22525696
  -­‐0.91
  3.90E-­‐03
  1.00E+00
 
CHL1
  3:195757-­‐409417
  -­‐1.87
  4.20E-­‐03
  1.00E+00
 
PTP4A1
  6:63521760-­‐63583587
  -­‐0.53
  4.35E-­‐03
  1.00E+00
 
SHISA4
  1:201888679-­‐201892306
  0.68
  5.75E-­‐03
  1.00E+00
 
PQLC1
  18:79902419-­‐79951664
  0.91
  5.95E-­‐03
  1.00E+00
 
C16orf13
  16:634426-­‐636366
  0.64
  6.05E-­‐03
  1.00E+00
 
PLP1
  X:103773717-­‐103792619
  0.57
  6.35E-­‐03
  1.00E+00
 
NRXN1
  2:49918504-­‐51032561
  -­‐1.26
  8.00E-­‐03
  1.00E+00
 
MYBL2
  20:43667018-­‐43716496
  0.62
  8.00E-­‐03
  1.00E+00
 
ELN
  7:74027788-­‐74069907
  0.55
  8.05E-­‐03
  1.00E+00
 
LRRTM3
  10:65912517-­‐67696169
  -­‐1.85
  8.25E-­‐03
  1.00E+00
 
ZBTB20
  3:114338093-­‐115147271
  0.74
  9.00E-­‐03
  1.00E+00
 
SLC47A2
  17:19678287-­‐19718979
  -­‐0.99
  1.06E-­‐02
  1.00E+00
 
PDCD5
  19:32581067-­‐32587452
  -­‐0.41
  1.08E-­‐02
  1.00E+00
 
BIN1
  2:127048026-­‐127107355
  0.58
  1.18E-­‐02
  1.00E+00
 
UBTD1
  10:97498867-­‐97571209
  0.69
  1.25E-­‐02
  1.00E+00
 
CRIP2
  14:105472961-­‐105480170
  0.80
  1.32E-­‐02
  1.00E+00
 
GIPC1
  19:14477759-­‐14496149
  0.59
  1.39E-­‐02
  1.00E+00
 

 

   
 
46
 

 
transfection
 (1.16-­‐fold;
 p
 =
 3.0
 ×
 10
−1
).
 In
 ReNcell
 CX
 cells,
 a
 significant
 change
 in
 MSN
 transcript
 
expression
 was
 not
 observed
 at
 24
 h
 post-­‐transfection
 (0.97-­‐fold;
 p
 =
 8.0
 ×
 10
−1
)
 and
 at
 72
 h
 
post-­‐transfection
 (0.92-­‐fold;
 p
 =
 5.0
 ×
 10
−1
).
 These
 results
 suggest
 that,
 although
 MSNP1AS
 
regulates
 expression
 of
 moesin
 protein
 (74),
 MSNP1AS
 does
 not
 directly
 regulate
 the
 expression
 
of
 the
 MSN
 transcript.
 
2.4.3
 Transcriptional
 Consequences
 of
 MSNP1AS
 Overexpression
 Are
 Enriched
 in
 Protein
 
Synthesis
 and
 Chromatin
 Regulation
 
Gene
 ontology
 (GO)
 analysis
 using
 the
 DAVID
 web
 server
 for
 differentially
 expressed
 
 
(p
 <
 0.05)
 genes
 revealed
 an
 enrichment
 of
 genes
 involved
 in
 protein
 synthesis
 and
 chromatin
 
regulation
 (Figure
 2.3A;
 Table
 2.8,
 2.9,
 2.10,
 and
 2.11).
 In
 SK-­‐N-­‐SH
 cells
 at
 72
 h
 post-­‐
transfection
 of
 the
 MSNP1AS
 overexpression
 construct,
 the
 351
 genes
 with
 altered
 expression
 
(p
 <
 0.05)
 were
 enriched
 for
 genes
 involved
 in
 translational
 elongation
 (Bonferroni
 corrected
 p
 
=
 4.7
 ×
 10
−11
),
 structural
 constituent
 of
 ribosome
 (Bonferroni
 corrected
 p
 =
 1.2
 ×
 10
−7
),
 and
 
nucleosome
 organization
 (Bonferroni
 corrected
 p
 =
 2.8
 ×
 10
−6
).
 No
 enrichment
 that
 survived
 
Bonferroni
 correction
 was
 observed
 among
 the
 157
 differentially
 expressed
 genes
 in
 SK-­‐N-­‐SH
 
cells
 at
 24
 h
 post-­‐transfection.
 However,
 a
 trend
 toward
 enrichment
 of
 chromatin
 related
 genes
 
was
 observed
 in
 SK-­‐N-­‐SH
 cells
 at
 24
 h
 (uncorrected
 p
 =
 5.8
 ×
 10
−2
)
 with
 some
 of
 the
 same
 genes
 
differentially
 expressed
 as
 SK-­‐N-­‐SH
 at
 72
 h
 post-­‐transfection
 (e.g.,
 HIST1H2AJ,
 (Table
 2.8
 and
 
2.9).
 Similarly,
 in
 ReNcell
 CX
 human
 neural
 progenitor
 cells
 at
 24
 h
 post-­‐transfection,
 
differentially
 expressed
 (p
 <
 0.05)
 genes
 were
 enriched
 for
 translational
 elongation
 (Bonferroni
 
corrected
 p
 =
 3.3
 ×
 10
−23
),
 structural
 constituent
 of
 ribosome
 (Bonferroni
 corrected
 p
 =
 7.3
 ×
 
10
−20
),
 and
 translation
 (Bonferroni
 corrected
 p
 =
 1.7
 ×
 10
−13
)
 (Figure
 2.3B;
 Table
 2.10
 and
 2.11).
 
47
 

 
No
 enrichment
 that
 survived
 Bonferroni
 correction
 was
 observed
 among
 the
 164
 genes
 
differentially
 expressed
 in
 
ReNcell
 CX
 cells
 at
 72
 h
 post-­‐transfection.
 However,
 an
 enrichment
 of
 genes
 involved
 in
 
translation
 (uncorrected
 p
 =
 4.0
 ×
 10
−2
)
 was
 among
 the
 top
 results
 in
 ReNcell
 CX
 cells
 at
 72
 h
 
post-­‐transfection.
 These
 data
 suggest
 that
 genes
 involved
 in
 translation
 are
 altered
 by
 
MSNP1AS
 overexpression.
 
 

 
Figure
 2.3.
 GO
 enrichment
 analysis
 on
 all
 differentially
 expressed
 genes
 due
 to
 overexpression
 
of
 MSNP1AS.
 GO
 enrichment
 analysis
 was
 performed
 on
 (A)
 SK-­‐N-­‐SH
 cells
 after
 72
 h,
 and
 (B)
 
ReNcell
 CX
 cells
 after
 24
 h.
 

 
2.5
 Discussion
 
2.5.1
 Summary
 of
 Results
 
The
 data
 presented
 here
 indicate
 that
 overexpression
 of
 the
 ASD-­‐associated
 lncRNA
 MSNP1AS
 
alters
 neuronal
 architecture
 in
 human
 neural
 progenitor
 cells
 by
 three
 potential
 mechanisms
 
(Figure
 2.4).
 As
 we
 previously
 described
 (74),
 MSNP1AS
 overexpression
 decreases
 expression
 of
 
moesin
 protein,
 which
 is
 known
 to
 decrease
 neuronal
 complexity
 (147).
 Here,
 we
 demonstrate
 
that
 MSNP1AS
 overexpression
 also
 changes
 the
 expression
 of
 genes
 involved
 in
 protein
 
synthesis
 and
 chromatin
 organization.
 Therefore,
 MSNP1AS
 overexpression
 has
 multiple
 
   
 
48
 

 
Table
 2.8.
 Top
 twenty-­‐five
 significant
 p-­‐values
 of
 gene
 ontology
 categories
 for
 SK-­‐N-­‐SH
 cells
 
24
 hrs
 post-­‐transfection.
 
Term
  Pvalue
 
Fold
 
Enrichment
  Bonferroni
  Benjamini
 
GO:0010033~response
 to
 organic
 
substance
  9.27E-­‐05
  3.10
  0.10
  0.10
 
GO:0006915~apoptosis
  1.64E-­‐04
  3.27
  0.17
  0.09
 
GO:0012501~programmed
 cell
 death
  1.91E-­‐04
  3.22
  0.20
  0.07
 
GO:0007167~enzyme
 linked
 receptor
 
protein
 signaling
 pathway
  2.53E-­‐04
  4.22
  0.26
  0.07
 
GO:0008219~cell
 death
  3.07E-­‐04
  2.92
  0.30
  0.07
 
GO:0016265~death
  3.30E-­‐04
  2.90
  0.32
  0.06
 
GO:0044421~extracellular
 region
 part
  4.99E-­‐04
  2.55
  0.09
  0.09
 
GO:0001568~blood
 vessel
 development
  5.23E-­‐04
  4.82
  0.46
  0.08
 
GO:0042060~wound
 healing
  5.98E-­‐04
  5.50
  0.50
  0.08
 
GO:0001944~vasculature
 development
  6.14E-­‐04
  4.71
  0.51
  0.08
 
GO:0050878~regulation
 of
 body
 fluid
 
levels
  6.83E-­‐04
  6.52
  0.55
  0.08
 
GO:0005576~extracellular
 region
  8.23E-­‐04
  1.89
  0.15
  0.08
 
GO:0009719~response
 to
 endogenous
 
stimulus
  9.49E-­‐04
  3.57
  0.67
  0.10
 
GO:0007596~blood
 coagulation
  1.03E-­‐03
  7.73
  0.70
  0.10
 
GO:0050817~coagulation
  1.03E-­‐03
  7.73
  0.70
  0.10
 
GO:0007179~transforming
 growth
 factor
 
beta
 receptor
 signaling
 pathway
  1.07E-­‐03
  10.94
  0.71
  0.09
 
GO:0007178~transmembrane
 receptor
 
protein
 serine/threonine
 kinase
 signaling
 
pathway
  1.08E-­‐03
  7.65
  0.71
  0.09
 
GO:0007599~hemostasis
  1.33E-­‐03
  7.30
  0.79
  0.10
 
GO:0009991~response
 to
 extracellular
 
stimulus
  1.37E-­‐03
  4.78
  0.80
  0.09
 
GO:0009725~response
 to
 hormone
 
stimulus
  1.80E-­‐03
  3.58
  0.88
  0.12
 
GO:0009611~response
 to
 wounding
  2.14E-­‐03
  2.97
  0.92
  0.13
 
GO:0005578~proteinaceous
 extracellular
 
matrix
  2.27E-­‐03
  3.82
  0.35
  0.14
 
GO:0042127~regulation
 of
 cell
 
proliferation
  2.28E-­‐03
  2.50
  0.93
  0.13
 
GO:0042981~regulation
 of
 apoptosis
  2.78E-­‐03
  2.45
  0.96
  0.15
 
GO:0043067~regulation
 of
 programmed
 
cell
 death
  3.04E-­‐03
  2.43
  0.97
  0.15
 

 
49
 

 
Table
 2.9.
 Top
 twenty-­‐five
 significant
 p-­‐values
 of
 gene
 ontology
 categories
 for
 SK-­‐N-­‐SH
 cells
 
72
 hrs
 post-­‐transfection.
 
Term
  PValue
 
Fold
 
Enrichment
  Bonferroni
  Benjamini
 
GO:0006414~translational
 elongation
  2.68E-­‐14
  10.63
  4.72E-­‐11
  4.72E-­‐11
 
GO:0022626~cytosolic
 ribosome
  1.29E-­‐13
  11.69
  3.87E-­‐11
  3.87E-­‐11
 
GO:0033279~ribosomal
 subunit
  3.09E-­‐11
  7.81
  9.31E-­‐09
  4.66E-­‐09
 
GO:0003735~structural
 constituent
 of
 
ribosome
  2.36E-­‐10
  6.47
  1.15E-­‐07
  1.15E-­‐07
 
GO:0044445~cytosolic
 part
  5.60E-­‐10
  6.58
  1.69E-­‐07
  5.62E-­‐08
 
GO:0034728~nucleosome
 organization
  1.61E-­‐09
  8.66
  2.83E-­‐06
  1.42E-­‐06
 
GO:0000786~nucleosome
  1.84E-­‐09
  10.85
  5.52E-­‐07
  1.38E-­‐07
 
GO:0005840~ribosome
  4.45E-­‐09
  5.14
  1.34E-­‐06
  2.68E-­‐07
 
GO:0032993~protein-­‐DNA
 complex
  7.44E-­‐09
  8.56
  2.24E-­‐06
  3.73E-­‐07
 
GO:0006323~DNA
 packaging
  3.35E-­‐08
  6.88
  5.91E-­‐05
  1.97E-­‐05
 
GO:0022627~cytosolic
 small
 ribosomal
 
subunit
  4.42E-­‐08
  13.15
  1.33E-­‐05
  1.90E-­‐06
 
GO:0006334~nucleosome
 assembly
  4.43E-­‐08
  8.31
  7.81E-­‐05
  1.95E-­‐05
 
GO:0006412~translation
  5.51E-­‐08
  3.89
  9.72E-­‐05
  1.94E-­‐05
 
GO:0031497~chromatin
 assembly
  6.62E-­‐08
  8.02
  1.17E-­‐04
  1.95E-­‐05
 
GO:0065004~protein-­‐DNA
 complex
 
assembly
  1.10E-­‐07
  7.67
  1.94E-­‐04
  2.78E-­‐05
 
GO:0006333~chromatin
 assembly
 or
 
disassembly
  6.59E-­‐07
  5.92
  1.16E-­‐03
  1.45E-­‐04
 
GO:0015935~small
 ribosomal
 subunit
  2.62E-­‐06
  8.35
  7.87E-­‐04
  9.85E-­‐05
 
GO:0005829~cytosol
  3.45E-­‐06
  1.98
  1.04E-­‐03
  1.15E-­‐04
 
GO:0006325~chromatin
 organization
  7.78E-­‐06
  3.12
  1.36E-­‐02
  1.52E-­‐03
 
GO:0051276~chromosome
 organization
  1.21E-­‐05
  2.77
  2.12E-­‐02
  2.14E-­‐03
 
GO:0015934~large
 ribosomal
 subunit
  3.66E-­‐05
  7.07
  1.10E-­‐02
  1.10E-­‐03
 
GO:0034621~cellular
 macromolecular
 
complex
 subunit
 organization
  3.83E-­‐05
  3.01
  6.54E-­‐02
  6.13E-­‐03
 
GO:0022625~cytosolic
 large
 ribosomal
 
subunit
  7.19E-­‐05
  9.69
  2.14E-­‐02
  1.97E-­‐03
 
GO:0034622~cellular
 macromolecular
 
complex
 assembly
  9.33E-­‐05
  3.04
  1.52E-­‐01
  1.36E-­‐02
 
GO:0005198~structural
 molecule
 activity
  1.02E-­‐04
  2.31
  4.89E-­‐02
  2.47E-­‐02
 

 

   
 
50
 

 
Table
 2.10.
 Top
 twenty-­‐five
 significant
 p-­‐values
 of
 gene
 ontology
 categories
 for
 ReNcell
 CX
 cells
 
24
 hrs
 post-­‐transfection.
 
Term
  PValue
 
Fold
 
Enrichment
  Bonferroni
  Benjamini
 
GO:0006414~translational
 
elongation
  2.14E-­‐26
  19.44
  3.33E-­‐23
  3.33E-­‐23
 
GO:0003735~structural
 constituent
 
of
 ribosome
  1.97E-­‐22
  12.18
  7.27E-­‐20
  7.27E-­‐20
 
GO:0022626~cytosolic
 ribosome
  3.47E-­‐22
  21.04
  9.29E-­‐20
  9.29E-­‐20
 
GO:0005840~ribosome
  1.35E-­‐20
  10.45
  3.62E-­‐18
  1.81E-­‐18
 
GO:0033279~ribosomal
 subunit
  2.67E-­‐20
  14.53
  7.15E-­‐18
  2.38E-­‐18
 
GO:0006412~translation
  1.17E-­‐16
  6.81
  1.72E-­‐13
  8.62E-­‐14
 
GO:0044445~cytosolic
 part
  3.44E-­‐16
  11.21
  8.93E-­‐14
  2.23E-­‐14
 
GO:0022627~cytosolic
 small
 
ribosomal
 subunit
  5.44E-­‐14
  25.18
  1.46E-­‐11
  2.92E-­‐12
 
GO:0030529~ribonucleoprotein
 
complex
  2.53E-­‐12
  4.66
  6.79E-­‐10
  1.13E-­‐10
 
GO:0005198~structural
 molecule
 
activity
  2.54E-­‐12
  4.01
  9.39E-­‐10
  4.70E-­‐10
 
GO:0005829~cytosol
  3.73E-­‐12
  2.91
  9.99E-­‐10
  1.43E-­‐10
 
GO:0015935~small
 ribosomal
 
subunit
  2.01E-­‐11
  15.99
  5.39E-­‐09
  6.74E-­‐10
 
GO:0015934~large
 ribosomal
 subunit
  1.21E-­‐08
  12.72
  3.25E-­‐06
  3.62E-­‐07
 
GO:0022625~cytosolic
 large
 
ribosomal
 subunit
  2.18E-­‐08
  18.35
  5.85E-­‐06
  5.85E-­‐07
 
GO:0003723~RNA
 binding
  2.63E-­‐07
  2.95
  9.72E-­‐05
  3.24E-­‐05
 
GO:0043228~non-­‐membrane-­‐
bounded
 organelle
  9.08E-­‐07
  1.82
  2.43E-­‐04
  2.21E-­‐05
 
GO:0043232~intracellular
 non-­‐
membrane-­‐bounded
 organelle
  9.08E-­‐07
  1.82
  2.43E-­‐04
  2.21E-­‐05
 
GO:0042981~regulation
 of
 apoptosis
  9.89E-­‐05
  2.35
  1.42E-­‐01
  4.99E-­‐02
 
GO:0043067~regulation
 of
 
programmed
 cell
 death
  1.16E-­‐04
  2.33
  1.64E-­‐01
  4.39E-­‐02
 
GO:0010941~regulation
 of
 cell
 death
  1.23E-­‐04
  2.32
  1.73E-­‐01
  3.73E-­‐02
 
GO:0016563~transcription
 activator
 
activity
  2.27E-­‐04
  2.93
  8.04E-­‐02
  2.07E-­‐02
 
GO:0000123~histone
 
acetyltransferase
 complex
  4.35E-­‐04
  9.30
  1.10E-­‐01
  9.67E-­‐03
 
GO:0003712~transcription
 cofactor
 
activity
  6.31E-­‐04
  2.92
  2.08E-­‐01
  4.55E-­‐02
 
GO:0051726~regulation
 of
 cell
 cycle
  6.37E-­‐04
  3.08
  6.28E-­‐01
  1.52E-­‐01
 
GO:0010604~positive
 regulation
 of
 
macromolecule
 metabolic
 process
  6.40E-­‐04
  2.12
  6.30E-­‐01
  1.32E-­‐01
 
51
 

 
Table
 2.11.
 Top
 twenty-­‐five
 significant
 p-­‐values
 of
 gene
 ontology
 categories
 for
 ReNcell
 CX
 
cells
 72
 hrs
 post-­‐transfection.
 
Term
  PValue
 
Fold
 
Enrichment
  Bonferroni
  Benjamini
 
GO:0043232~intracellular
 non-­‐
membrane-­‐bounded
 organelle
  2.05E-­‐03
  1.63
  3.46E-­‐01
  3.46E-­‐01
 
GO:0043228~non-­‐membrane-­‐bounded
 
organelle
  2.05E-­‐03
  1.63
  3.46E-­‐01
  3.46E-­‐01
 
GO:0005856~cytoskeleton
  2.43E-­‐03
  1.95
  3.96E-­‐01
  2.23E-­‐01
 
GO:0032268~regulation
 of
 cellular
 
protein
 metabolic
 process
  4.58E-­‐03
  2.88
  9.84E-­‐01
  9.84E-­‐01
 
GO:0044430~cytoskeletal
 part
  1.38E-­‐02
  1.97
  9.44E-­‐01
  6.18E-­‐01
 
GO:0044456~synapse
 part
  1.80E-­‐02
  3.34
  9.76E-­‐01
  6.08E-­‐01
 
GO:0050840~extracellular
 matrix
 binding
  1.89E-­‐02
  14.01
  9.97E-­‐01
  9.97E-­‐01
 
GO:0005911~cell-­‐cell
 junction
  2.24E-­‐02
  3.70
  9.91E-­‐01
  6.09E-­‐01
 
GO:0042734~presynaptic
 membrane
  2.48E-­‐02
  12.13
  9.94E-­‐01
  5.79E-­‐01
 
GO:0005198~structural
 molecule
 activity
  2.71E-­‐02
  2.19
  1.00E+00
  9.84E-­‐01
 
GO:0005829~cytosol
  2.95E-­‐02
  1.68
  9.98E-­‐01
  5.88E-­‐01
 
GO:0045202~synapse
  3.08E-­‐02
  2.64
  9.98E-­‐01
  5.54E-­‐01
 
GO:0005913~cell-­‐cell
 adherens
 junction
  3.51E-­‐02
  10.05
  9.99E-­‐01
  5.61E-­‐01
 
GO:0005912~adherens
 junction
  4.23E-­‐02
  3.78
  1.00E+00
  5.91E-­‐01
 
GO:0006446~regulation
 of
 translational
 
initiation
  4.23E-­‐02
  9.08
  1.00E+00
  1.00E+00
 
GO:0016023~cytoplasmic
 membrane-­‐
bounded
 vesicle
  4.30E-­‐02
  2.13
  1.00E+00
  5.63E-­‐01
 
GO:0031988~membrane-­‐bounded
 vesicle
  5.09E-­‐02
  2.06
  1.00E+00
  5.94E-­‐01
 
GO:0005925~focal
 adhesion
  5.52E-­‐02
  4.60
  1.00E+00
  5.95E-­‐01
 
GO:0070161~anchoring
 junction
  5.79E-­‐02
  3.41
  1.00E+00
  5.86E-­‐01
 
GO:0005924~cell-­‐substrate
 adherens
 
junction
  6.06E-­‐02
  4.43
  1.00E+00
  5.78E-­‐01
 
GO:0019226~transmission
 of
 nerve
 
impulse
  6.13E-­‐02
  2.48
  1.00E+00
  1.00E+00
 
GO:0030055~cell-­‐substrate
 junction
  6.91E-­‐02
  4.19
  1.00E+00
  6.04E-­‐01
 
GO:0044449~contractile
 fiber
 part
  7.06E-­‐02
  4.15
  1.00E+00
  5.90E-­‐01
 
GO:0030054~cell
 junction
  7.18E-­‐02
  2.04
  1.00E+00
  5.75E-­‐01
 
GO:0015629~actin
 cytoskeleton
  7.77E-­‐02
  2.62
  1.00E+00
  5.86E-­‐01
 

   
 
52
 

 

 
Figure
 2.4.
 Potential
 mechanisms
 by
 which
 MSNP1AS
 alters
 neuronal
 architecture.
 MSNP1AS
 
overexpression
 decreases
 the
 expression
 of
 moesin
 protein,
 as
 well
 as
 the
 expression
 of
 genes
 
involved
 in
 protein
 synthesis
 and
 chromatin
 organization,
 leading
 to
 a
 decrease
 in
 neurite
 
complexity.
 

 
molecular
 functions:
 MSNP1AS
 alters
 expression
 of
 genes
 that
 contribute
 to
 chromatin
 
organization;
 MSNP1AS
 binds
 MSN
 and
 alters
 the
 translation
 of
 moesin
 protein;
 and
 MSNP1AS
 
alters
 the
 expression
 of
 genes
 that
 regulate
 translation
 more
 globally
 (Figure
 2.4).
 
MSNP1AS
 was
 discovered
 as
 the
 functional
 element
 revealed
 by
 an
 ASD
 GWAS
 (74).
 The
 
allele
 frequencies
 of
 the
 chromosome
 5p14.1
 genetic
 variants
 with
 genome-­‐wide
 significant
 
association
 are
 high:
 greater
 than
 half
 the
 population
 carries
 at
 least
 one
 copy
 of
 the
 risk
 allele
 
(26).
 It
 is
 unclear
 whether
 individuals
 who
 are
 homozygous
 for
 the
 rs4307059
 risk
 allele
 are
 at
 
higher
 risk
 for
 ASD
 than
 those
 heterozygous
 at
 rs4307059.
 However,
 the
 expression
 of
 
MSNP1AS
 is
 increased
 in
 the
 postmortem
 temporal
 cortex
 of
 individuals
 who
 are
 homozygous
 
for
 rs4307059
 compared
 to
 individuals
 who
 are
 heterozygous
 for
 rs4307059
 (74).
 It
 is
 striking
 
that
 the
 biological
 functions
 of
 this
 ASD-­‐associated
 lncRNA—protein
 synthesis
 and
 chromatin
 
organization—are
 matched
 closely
 to
 the
 biological
 functions
 revealed
 by
 genes
 with
 rare
 de
 
novo
 mutations
 associated
 with
 ASD
 (161,
 162).
 These
 results
 suggest
 that
 both
 common
 and
 
53
 

 
rare
 ASD-­‐associated
 variants
 converge
 upon
 the
 common
 molecular
 pathways.
 Further,
 the
 
chromosome
 5p14.1
 genetic
 marker
 with
 the
 most
 significant
 association
 (rs4307059
 with
 ASD
 
association
 p
 =
 10
−10
)
 also
 contributes
 to
 altered
 social
 communication
 in
 a
 general
 population
 
sample
 (73).
 Together,
 these
 data
 suggest
 common
 molecular
 pathways
 that
 contribute
 to
 
social
 communication
 that
 involve
 protein
 translation
 and
 chromatin
 organization.
 
MSNP1AS
 is
 antisense
 to
 moesin
 (membrane-­‐organizing
 extension
 spike
 protein).
 Moesin
 is
 a
 
member
 of
 the
 ERM
 (exrin/radixin/moesin)
 family
 of
 proteins
 that
 link
 the
 actin
 filaments
 to
 
the
 cellular
 membrane.
 Along
 with
 radixin,
 moesin
 protein
 localizes
 to
 growth
 cones
 and
 
filopodia
 that
 emanate
 from
 neurite
 shafts,
 both
 regions
 of
 high
 motility
 and
 growth.
 These
 
proteins
 are
 especially
 important
 during
 development,
 as
 shown
 in
 rat
 cerebral
 cortex.
 In
 rats,
 
ERM
 protein
 expression
 reaches
 a
 maximum
 near
 birth
 and
 gradually
 declines
 through
 
postnatal
 development
 (147),
 which
 is
 during
 synaptic
 maturation.
 In
 the
 intermediate
 zone
 of
 
developing
 rat
 cerebral
 cortex,
 ERM
 proteins
 are
 significantly
 expressed
 in
 neurite
 extensions
 at
 
embryonic
 day
 17
 (146),
 a
 time
 of
 substantial
 synapse
 formation.
 They
 regulate
 adhesion
 
receptors,
 signaling
 molecules
 that
 provide
 spatial
 information
 to
 the
 cell,
 and
 growth
 cone
 
actin
 to
 mediate
 attractive
 growth
 cone
 guidance
 (163).
 When
 antisense
 RNA
 is
 used
 to
 
knockdown
 moesin
 in
 cultured
 hippocampal
 and
 cortical
 neurons,
 specific
 phenotypes
 are
 
observed,
 including
 a
 dramatic
 reduction
 in
 the
 rate
 of
 neurite
 advancement
 (147),
 suppression
 
of
 neurite
 formation
 (145),
 growth
 cone
 collapse
 (147),
 suppression
 of
 an
 increase
 in
 dendritic
 
spine
 formation
 induced
 by
 estrogen
 (149),
 and
 suppression
 of
 an
 increase
 in
 active
 presynaptic
 
boutons
 induced
 by
 glutamate
 (148).
 In
 addition,
 ERM
 proteins
 have
 been
 shown
 to
 mediate
 
neuritogenesis
 (164).
 These
 data
 denote
 that
 the
 function
 of
 moesin
 involves
 both
 regulating
 
54
 

 
axonal
 growth
 cone
 development
 presynaptically
 and
 initiating
 dendritic
 spine
 development
 
postsynaptically.
 Moesin
 knockout
 mice
 have
 not
 been
 evaluated
 for
 behavioral
 phenotypes
 or
 
brain
 development.
 
While
 MSN
 transcripts
 did
 not
 undergo
 a
 significant
 change
 in
 gene
 expression,
 
significant
 alterations
 are
 observed
 in
 neurite
 development.
 As
 we
 previously
 described,
 
overexpressed
 MSNP1AS
 binds
 MSN
 transcript
 and
 prevents
 translation
 of
 moesin
 protein.
 The
 
current
 study
 expands
 our
 knowledge
 of
 MSNP1AS
 overexpression
 in
 two
 ways.
 First,
 the
 
observation
 that
 MSNP1AS
 overexpression
 does
 not
 alter
 expression
 of
 MSN
 transcript
 
indicates
 that
 MSNP1AS
 does
 not
 directly
 contribute
 to
 the
 regulation
 of
 the
 MSN
 transcription.
 
Instead,
 in
 relation
 to
 the
 regulation
 of
 moesin,
 it
 appears
 that
 MSNP1AS
 acts
 only
 to
 inhibit
 the
 
translation
 of
 moesin
 protein.
 Second,
 MSNP1AS
 acts
 more
 globally
 than
 just
 inhibiting
 the
 
translation
 of
 moesin.
 The
 overexpression
 of
 MSNP1AS
 caused
 changes
 in
 the
 expression
 of
 
genes
 involved
 in
 the
 protein
 synthesis
 and
 chromatin
 regulation.
 These
 transcriptional
 changes
 
suggest
 that
 MSNP1AS
 participates
 in
 global
 biological
 processes
 that
 impact
 neuronal
 
differentiation
 beyond
 its
 impact
 on
 moesin.
 Additional
 experiments
 will
 be
 necessary
 to
 
determine
 the
 relative
 contributions
 of
 MSNP1AS
 to
 the
 biological
 processes
 of
 chromatin
 
regulation,
 protein
 synthesis,
 and
 the
 regulation
 of
 moesin.
 It
 will
 be
 important
 to
 determine
 
the
 expression
 of
 moesin
 protein
 and
 the
 MSNP1AS
 ncRNA
 in
 neurons
 derived
 from
 patients
 
with
 the
 ASD-­‐associated
 rs4307059
 allele,
 as
 well
 as
 from
 patients
 with
 Cri-­‐du-­‐chat
 syndrome
 
with
 deletion
 of
 chromosome
 5p14.1,
 as
 these
 experiments
 may
 provide
 further
 evidence
 of
 a
 
contribution
 of
 MSNP1AS
 to
 altered
 brain
 development.
 It
 will
 also
 be
 important
 to
 determine
 
if
 individuals
 with
 ASD
 and
 the
 ASD-­‐associated
 rs4307059
 allele
 exhibit
 common
 comorbid
 
55
 

 
features
 of
 ASD,
 such
 as
 gastrointestinal
 disorders
 or
 epilepsy.
 A
 network
 analysis
 of
 microarray
 
data
 from
 postmortem
 ASD
 brain
 placed
 MSN
 as
 a
 central
 node
 (165).
 However,
 there
 was
 no
 
enrichment
 of
 the
 genes
 in
 the
 MSN
 node
 and
 the
 genes
 with
 altered
 expression
 following
 
MSNP1AS
 over-­‐expression
 identified
 here.
 
In
 this
 study,
 overexpressed
 MSNP1AS
 results
 in
 shortened
 neurites
 and
 fewer
 neurites
 
per
 cell,
 indicating
 that
 neuritogenesis
 and
 extension
 are
 both
 impacted.
 The
 decreased
 neurite
 
number
 and
 growth
 upon
 overexpression
 of
 MSNP1AS
 reinforces
 studies
 that
 suggest
 that
 
altered
 short-­‐
 and
 long-­‐range
 connectivity
 in
 the
 brains
 of
 autism
 patients
 may
 be
 contributing
 
to
 the
 pathogenesis
 of
 the
 disorder
 (166).
 This
 is
 in
 line
 with
 global
 observations
 of
 decreased
 
connectivity
 in
 the
 brains
 of
 individuals
 with
 ASD
 (151-­‐155,
 167),
 although
 there
 is
 evidence
 for
 
increased
 connectivity
 between
 some
 brain
 regions
 in
 ASD
 (154,
 155).
 
NcRNA
 comprise
 90%
 of
 the
 human
 genome
 and
 are
 believed
 to
 contribute
 to
 regulatory
 
function
 as
 enhancers
 and
 promoters
 (168,
 169).
 These
 elements
 differentiate
 across
 space
 and
 
time
 in
 a
 significant
 way,
 but
 the
 exact
 functional
 roles
 of
 most
 lncRNAs
 have
 not
 been
 
quantified
 (170-­‐172).
 The
 contributions
 of
 lncRNAs
 to
 psychiatric
 disorders
 are
 a
 topic
 of
 
ongoing
 research
 (173).
 The
 data
 from
 this
 study
 provide
 further
 evidence
 for
 pathways
 
reported
 to
 influence
 ASD,
 as
 well
 as
 giving
 additional
 insight
 into
 the
 molecular
 function
 of
 the
 
ASD-­‐associated
 lncRNA,
 MSNP1AS.
 

 
2.6
 Conclusion
 
We
 previously
 identified
 the
 lncRNA
 MSNP1AS
 as
 a
 functional
 element
 revealed
 by
 
genome
 wide
 significant
 association
 with
 ASD.
 Using
 a
 candidate
 gene
 approach,
 we
 showed
 
56
 

 
that
 one
 function
 of
 MSNP1AS
 is
 to
 regulate
 the
 expression
 of
 moesin
 protein.
 The
 data
 
presented
 here
 indicate
 that
 MSNP1AS
 over-­‐expression
 does
 not
 significantly
 alter
 the
 
expression
 of
 the
 MSN
 transcript,
 suggesting
 that
 MSNP1AS
 functions
 specifically
 at
 regulating
 
the
 translation
 of
 the
 MSN
 transcript
 to
 moesin
 protein.
 Further,
 the
 unbiased
 RNA
 sequencing
 
data
 revealed
 that
 over-­‐expression
 of
 MSNP1AS
 alters
 the
 expression
 of
 genes
 involved
 in
 
protein
 synthesis
 and
 chromatin
 organization.
 These
 biological
 processes
 are
 also
 implicated
 by
 
rare
 mutations
 associated
 with
 ASD,
 suggesting
 convergent
 molecular
 mechanisms
 that
 
contribute
 to
 the
 altered
 brain
 development
 of
 ASD.
 

 
2.7
 Acknowledgements
 
This
 work
 was
 supported
 by
 National
 Institute
 of
 Mental
 Health
 grants
 R01MH100172
 (to
 
Daniel
 B.
 Campbell)
 and
 R21MH099504
 (to
 Daniel
 B.
 Campbell).
 

 

 
Chapter
 3:
 Transcriptional
 Gene
 Silencing
 of
 the
 Autism-­‐Associated
 Long
 Noncoding
 RNA
 
MSNP1AS
 in
 Human
 Neural
 Progenitor
 Cells
 
These
 data
 were
 published
 in
 Developmental
 Neuroscience
 (69).
 

 
3.1
 Abstract
 

  The
 lncRNA,
 MSNP1AS
 (moesin
 pseudogene
 1,
 antisense),
 is
 a
 functional
 element
 that
 
was
 previously
 associated
 to
 autism
 spectrum
 disorder
 (ASD)
 with
 genome
 wide
 significance.
 
Expression
 of
 MSNP1AS
 was
 increased
 12-­‐fold
 in
 the
 cerebral
 cortex
 of
 individuals
 with
 ASD
 and
 
22-­‐fold
 in
 individuals
 with
 a
 genome-­‐wide
 significantly
 associated
 ASD
 genetic
 marker
 on
 
57
 

 
chromosome
 5p14.1.
 Overexpression
 of
 MSNP1AS
 in
 human
 neuronal
 cells
 caused
 decreased
 
expression
 of
 moesin
 protein,
 which
 is
 involved
 in
 neuronal
 process
 stability.
 In
 this
 study,
 we
 
hypothesize
 that
 MSNP1AS
 knockdown
 impacts
 global
 transcriptome
 levels.
 We
 transfected
 the
 
human
 neural
 progenitor
 cell
 line,
 SK-­‐N-­‐SH,
 with
 constructs
 that
 caused
 a
 50%
 suppression
 of
 
MSNP1AS
 expression.
 After
 24
 hours,
 cells
 were
 harvested
 for
 total
 RNA
 isolation.
 Strand-­‐
specific
 RNA-­‐Seq
 analysis
 indicated
 altered
 expression
 of
 1,352
 genes,
 including
 altered
 
expression
 of
 318
 genes
 following
 correction
 for
 multiple
 comparisons.
 Expression
 of
 the
 OAS2
 
gene
 was
 increased
 >150-­‐fold,
 a
 result
 that
 was
 validated
 by
 quantitative
 PCR.
 Gene
 ontology
 
analysis
 of
 the
 318
 genes
 with
 altered
 expression
 following
 correction
 for
 multiple
 comparisons
 
indicated
 that
 upregulated
 genes
 were
 significantly
 enriched
 for
 genes
 involved
 in
 immune
 
response
 and
 downregulated
 genes
 were
 significantly
 enriched
 for
 genes
 involved
 in
 chromatin
 
remodeling.
 These
 data
 indicate
 multiple
 transcriptional
 and
 translational
 functions
 of
 
MSNP1AS
 that
 impact
 ASD-­‐relevant
 biological
 processes.
 Chromatin
 remodeling
 and
 immune
 
response
 are
 biological
 process
 implicated
 by
 genes
 with
 rare
 mutations
 associated
 with
 ASD.
 
Our
 data
 suggest
 that
 the
 functional
 elements
 implicated
 by
 association
 of
 common
 genetic
 
variants
 impact
 the
 same
 biological
 processes,
 suggesting
 a
 possible
 shared
 common
 molecular
 
pathway
 of
 ASD.
 
 

 
3.2
 Introduction
 
Autism
 spectrum
 disorder
 (ASD)
 is
 a
 neurodevelopmental
 condition
 characterized
 by
 
deficits
 in
 social
 communication
 and
 behavioral
 flexibility
 (140,
 174).
 Recent
 advances
 in
 
genomics
 have
 begun
 to
 identify
 the
 genetic
 variants
 that
 contribute
 to
 ASD
 risk
 (26,
 175,
 176).
 
58
 

 
The
 important
 steps
 are
 to
 translate
 these
 genetic
 findings
 to
 a
 biological
 understanding
 of
 ASD
 
pathogenesis
 and
 to
 development
 biologically
 based
 treatments
 (1).
 Recently,
 we
 identified
 a
 
lncRNA
 that
 contributes
 to
 ASD
 risk
 (74).
 A
 GWAS
 identified
 a
 single
 peak
 on
 chromosome
 
5p14.1
 for
 ASD,
 indicating
 a
 significant
 association
 (P
 =
 10
-­‐10
)
 for
 rs4307059
 (26).
 The
 same
 
rs4307059
 allele
 was
 also
 associated
 with
 social
 communication
 phenotypes
 in
 a
 general
 
population
 sample
 (73).
 Directly
 under
 the
 chromosome
 5p14.1
 ASD
 GWAS
 peak,
 we
 identified
 
a
 lncRNA
 that
 was
 almost
 4
 kb
 (74).
 This
 transcript
 is
 94%
 identical
 and
 anti-­‐sense
 to
 the
 X
 
chromosome
 transcript
 MSN,
 which
 encodes
 a
 neuronal
 architecture
 protein,
 and
 is
 on
 the
 
opposite
 (antisense)
 strand
 of
 moesin
 pseudogene
 1
 (MSNP1)
 (74).
 Therefore,
 we
 named
 the
 
newly
 discovered
 gene
 MSNP1AS
 (moesin
 pseudogene
 1,
 anti-­‐sense)
 (74).
 Expression
 of
 
MSNP1AS
 was
 detected
 in
 5
 brain
 regions
 (frontal
 cerebral
 cortex,
 temporal
 cerebral
 cortex,
 
occipital
 cerebral
 cortex,
 cerebellum
 and
 spinal
 cord)
 of
 the
 adult
 and
 in
 fetal
 frontal
 cerebral
 
cortex
 (74).
 In
 peripheral
 tissues,
 MSNP1AS
 was
 expressed
 at
 high
 levels
 in
 the
 adult
 peripheral
 
blood
 and
 fetal
 heart
 (74).
 These
 data
 indicate
 that
 MSNP1AS
 is
 expressed
 at
 high
 levels
 in
 both
 
the
 nervous
 and
 circulatory
 systems.
 Here
 we
 use
 publicly
 available
 databases
 of
 gene
 
expression
 to
 provide
 additional
 details
 of
 MSNP1AS
 expression.
 MSNP1AS
 expression
 is
 
increased
 12.7
 times
 in
 the
 postmortem
 temporal
 cerebral
 cortex
 of
 individuals
 with
 ASD
 
compared
 to
 controls
 and
 increased
 22
 times
 in
 individuals
 with
 the
 ASD-­‐associated
 rs4307059
 
allele
 (74).
 Therefore,
 the
 highly
 significant
 GWAS
 findings
 led
 to
 identification
 of
 a
 functional
 
lncRNA
 that
 contributes
 to
 ASD
 risk
 (74).
 
 
Jiang
 et
 al
 (177)
 recently
 reported
 rescue
 of
 cellular
 phenotypes
 of
 Down’s
 syndrome
 by
 
transgenic
 expression
 of
 another
 lncRNA,
 XIST
 (X
 inactive
 specific
 transcript).
 When
 XIST,
 which
 
59
 

 
normally
 targets
 the
 “extra”
 X
 chromosome
 of
 females
 for
 inactivation,
 was
 expressed
 on
 
chromosome
 21
 in
 cells
 derived
 from
 Down’s
 syndrome
 patients,
 the
 extra
 copy
 of
 
chromosome
 21
 was
 inactivated
 (177).
 These
 studies
 provide
 a
 clear
 path
 toward
 translation
 of
 
manipulating
 lncRNA
 expression
 to
 rescue
 disorder
 phenotypes.
 Because
 MSNP1AS
 is
 over-­‐
expressed
 in
 postmortem
 cerebral
 cortex
 of
 individuals
 with
 ASD
 (74),
 we
 used
 transcriptional
 
gene
 silencing
 in
 human
 neuronal
 progenitor
 cells
 to
 examine
 the
 potential
 for
 rescue.
 By
 
targeting
 the
 MSNP1AS
 gene
 promoter,
 we
 were
 able
 to
 reduce
 the
 expression
 of
 MSNP1AS
 
transcript
 to
 half
 of
 its
 normal
 expression
 and
 mimic
 a
 potential
 therapeutic
 for
 ASD
 and
 social
 
communication
 deficits.
 RNA
 sequencing
 was
 used
 as
 an
 unbiased
 approach
 to
 identify
 genes
 
with
 altered
 expression
 following
 knockdown
 of
 MSNP1AS.
 
 

 

 
3.3
 Materials
 and
 Methods
 
3.3.1
 Cell
 culture
 
SK-­‐N-­‐SH
 neuroblastoma
 neural
 progenitor
 cells
 were
 demonstrated
 previously
 to
 
express
 levels
 of
 MSNP1AS,
 MSN,
 and
 moesin
 protein
 consistent
 with
 those
 of
 developing
 
human
 brain
 (74).
 The
 human
 neural
 progenitor
 cell
 line
 SK-­‐N-­‐SH
 (American
 Type
 Culture
 
Collection;
 Manassas,
 VA,
 USA)
 was
 cultured
 in
 minimum
 essential
 medium
 supplemented
 with
 
10%
 heat-­‐inactivated
 fetal
 bovine
 serum,
 1%
 penicillin/streptomycin,
 nonessential
 amino
 acids,
 
and
 1.5
 g/L
 sodium
 bicarbonate
 in
 a
 75
 cm
2

 flask
 at
 37
 ̊C
 and
 5%
 CO
2
.
 When
 cells
 were
 75%
 
confluent,
 the
 human
 neuronal
 cells
 were
 subcultured
 to
 a
 density
 of
 1
 x
 10
6

 cells
 per
 well
 in
 a
 6
 
well
 plate
 for
 harvests
 24
 hours
 post-­‐transfection.
 
 

   
 
60
 

 
3.3.2
 Design
 of
 antisense
 RNA
 to
 the
 MSNP1AS
 proximal
 gene
 promoters
 
Small
 noncoding
 antisense
 RNAs
 (sasRNA)
 to
 silence
 gene
 transcription
 were
 designed
 
with
 proprietary
 software
 (178-­‐181).
 The
 program
 was
 used
 to
 identify
 the
 best
 sasRNA
 to
 
silence
 the
 lncRNA
 MSNP1AS
 gene
 transcript
 (Figure
 3.1).
 
 We
 cloned
 sasRNAs
 directed
 to
 the
 
putative
 gene
 promoter
 into
 the
 pCDNA3-­‐U6M2
 expression
 vector
 with
 the
 following
 target
 
sequence
 for
 MSNP1AS
 knockdown
 (KD):
 AATTCTAGAAATGTGCCAACA
 (Figure
 3.1).
 

 

 
Figure
 3.1.
 Small
 noncoding
 antisense
 RNAs
 (sasRNAs)
 were
 designed
 to
 silence
 transcription
 by
 
targeting
 the
 MSNP1AS
 gene
 promoter.
 

 
3.3.3
 Transfection
 of
 small
 interfering
 RNAs
 (sasRNAs)
 
 
Targets
 for
 transcriptional
 gene
 silencing
 (TGS)
 of
 MSNP1AS
 were
 defined
 in
 human
 
neuronal
 cell
 lines.
 Eight
 sasRNAs
 were
 designed
 directed
 to
 MSNP1AS
 TGS
 targets.
 The
 TGS
 
were
 screened
 for
 effectiveness
 in
 U87
 or
 Lan
 6
 cells.
 The
 TGS
 sasRNAs
 were
 screened
 for
 
efficacy
 and
 three
 were
 chosen
 for
 use
 in
 SK-­‐N-­‐SH
 cells.
 Each
 transfection
 set
 was
 considered
 
an
 independent
 experiment.
 The
 three
 sasRNA
 were
 tested
 at
 5
 different
 vector
 concentrations
 
(Table
 3.1).
 Cells
 harvest
 occurred
 24
 hours
 after
 transfection.
 Based
 on
 optimization
 
experiments,
 we
 used
 1
 ug
 of
 sasRNA
 per
 million
 cells
 to
 achieve
 ~50%
 suppression
 of
 
MSNP1AS
 in
 further
 studies.
 
 
 

   
 
61
 

 
Table
 3.1.
 MSNP1AS
 Fold
 Change
 After
 Transfection.
 

 
  0.5
 ug
  1
 ug
  2
 ug
  3
 ug
  4
 ug
 
AS-­‐1
  0.96
  0.17
  0.19
  0.41
  0.78
 
AS-­‐2
  3.04
  0.15
  0.49
  0.52
   
 
 
AS-­‐3
   
 
  0.55
  0.19
  0.77
  0.35
 

 
Cells
 were
 transfected
 using
 Amaxa
 Nucleofector
 (Lonza;
 Walkersville,
 MD,
 USA)
 
technology
 and
 subcultured
 into
 6
 well
 plates.
 One
 ml
 of
 fresh
 prewarmed
 media
 was
 added
 to
 
each
 well
 and
 the
 cells
 were
 centrifuged
 twice
 at
 130xg
 for
 10
 min
 for
 the
 SK-­‐N-­‐SH
 cells
 with
 
PBS
 washes
 of
 7
 ml
 and
 4
 ml,
 respectively.
 The
 cell
 pellet
 was
 resuspended
 in
 Nucleofector
 
solution
 containing
 supplement
 using
 1
 x
 10
6

 cells
 per
 well.
 One
 hundred
 microliters
 of
 
cell/nucleofector
 solution
 was
 added
 to
 a
 new
 cuvette
 and
 placed
 in
 the
 Amaxa
 Nucleofector
 
using
 the
 T-­‐16
 program.
 Five
 hundred
 microliters
 of
 fresh
 prewarmed
 media
 was
 added
 to
 the
 
cuvette,
 mixed,
 and
 transferred
 to
 the
 well
 with
 a
 disposable
 pipet.
 Two
 ml
 culture
 medium
 
was
 added
 to
 the
 transfected
 cells
 in
 the
 6
 well
 plate.
 The
 cells
 were
 incubated
 at
 37
 ̊C
 in
 5%
 
CO
2

 until
 harvest.
 Each
 experiment
 was
 repeated
 4
 times.
 
3.3.4
 Neural
 progenitor
 cell
 harvest
 
 
After
 24
 hours,
 the
 culture
 medium
 was
 removed
 and
 discarded.
 The
 cells
 were
 washed
 
with
 1
 ml
 of
 PBS
 wash
 and
 the
 PBS
 wash
 was
 aspirated.
 One
 ml
 of
 Trypsin/EDTA
 solution
 was
 
added
 to
 each
 well
 and
 the
 cells
 were
 incubated
 for
 4
 min.
 Five
 ml
 of
 cell
 type-­‐specific
 medium
 
was
 added
 and
 cells
 were
 triturated
 and
 transferred
 to
 15
 ml
 conical
 tubes.
 SK-­‐N-­‐SH
 cells
 were
 
centrifuged
 at
 130xg
 for
 10
 min.
 The
 supernatant
 was
 aspirated
 and
 5
 ml
 of
 PBS
 added.
 The
 cell
 
pellet
 was
 triturated
 and
 centrifuged
 at
 130xg
 for
 10
 min.
 The
 PBS
 wash
 was
 aspirated
 and
 1
 ml
 
of
 fresh
 PBS
 was
 added.
 Half
 of
 the
 solution
 was
 transferred
 to
 each
 of
 two
 1.5
 ml
 eppendorf
 
tubes.
 The
 tubes
 were
 centrifuged
 at
 4
 ̊C
 at
 130xg
 for
 10
 min.
 The
 supernatant
 was
 decanted
 
62
 

 
and
 the
 cell
 pellets
 were
 frozen
 at
 -­‐80
 ̊C.
 
 
 
3.3.5
 RNA
 purification
 
The
 Qiagen
 RNEasy
 kit
 was
 used
 to
 isolate
 total
 RNA
 using
 vacuum
 technology
 according
 
to
 the
 manufacturer’s
 protocol
 (Qiagen;
 Valencia,
 CA,
 USA).
 The
 RNA
 was
 eluted
 with
 35
 
microliters
 of
 RNAse-­‐free
 water
 and
 quantified
 using
 the
 NanoDrop
 ND-­‐1000
 
Spectrophotometer
 (v3.1.2)
 (Thermo
 Fisher
 Scientific;
 Waltham,
 MA,
 USA).
 The
 RNA
 was
 stored
 
at
 -­‐80
 ̊C.
 
3.3.6
 Quantitative
 RT-­‐PCR
 (qPCR)
 to
 confirm
 MSNP1AS
 knockdown
 
To
 confirm
 knockdown
 of
 MSNP1AS,
 cDNA
 was
 synthesized
 using
 the
 SuperScript
 III
 
First-­‐Strand
 Synthesis
 System
 for
 qRT-­‐PCR
 protocol
 (Invitrogen/Thermo
 Fisher
 Scientific;
 
Waltham,
 MA,
 USA).
 Five
 hundred
 nanograms
 of
 RNA
 were
 used
 to
 make
 one
 and
 a
 half
 
reaction
 volumes
 (35
 ml)
 of
 cDNA.
 The
 cDNA
 was
 stored
 at
 -­‐20
 ̊C.
 The
 qRT-­‐PCR
 protocol
 
described
 in
 Kerin
 et
 al
 (74)
 was
 used
 to
 validate
 knockdown
 of
 MSNP1AS.
 
 
3.3.7
 Construction
 of
 strand-­‐specific,
 ribosomal
 RNA
 depleted
 RNA
 sequencing
 libraries
 
Directional
 RNA-­‐Seq
 libraries
 were
 prepared
 for
 Illumina
 HiSeq
 2000
 sequencing
 using
 
the
 Stranded
 Total
 TruSeq
 RNA
 Sample
 Preparation
 kit
 with
 Ribo-­‐Zero
 Gold
 (Illumina)
 using
 
manufacturer’s
 protocol
 using
 the
 Hamilton
 Starlet
 Liquid
 Handling
 robot.
 One
 nanogram
 of
 
RNA
 was
 used
 for
 each
 sample.
 Ribo-­‐Zero
 was
 used
 to
 deplete
 cytoplasmic
 and
 mitochondrial
 
rRNA
 from
 total
 RNA.
 The
 depleted
 RNA
 was
 fragmented
 and
 primed
 with
 random
 hexamers
 to
 
synthesize
 first
 strand
 cDNA
 using
 Superscript
 II
 (Life
 Technologies).
 Next,
 the
 second
 strand
 
was
 synthesized,
 incorporating
 dUTP
 in
 place
 of
 dTTP.
 A
 single
 ‘A’
 base
 was
 added
 to
 the
 3’
 
ends
 of
 the
 fragments
 and
 the
 indexed
 adaptors
 were
 ligated
 to
 the
 ends
 of
 the
 ds
 cDNA
 to
 
63
 

 
prepare
 them
 for
 hybridization
 onto
 the
 flow
 cell.
 PCR
 was
 used
 to
 selectively
 enrich
 the
 
fragments
 with
 ligated
 adapters
 and
 to
 amplify
 the
 amount
 of
 DNA
 in
 the
 library.
 The
 libraries
 
were
 produced
 in
 a
 96-­‐well
 format
 and
 quality
 controlled
 using
 the
 Agilent
 Technologies
 2200
 
TapeStation
 Instrument.
 Libraries
 were
 pooled
 (four
 samples
 per
 lane)
 and
 sequenced
 on
 
Illumina
 HiSeq
 2000
 to
 a
 targeted
 depth
 generating
 an
 average
 of
 20
 million
 paired-­‐end
 50-­‐
cycle
 reads
 for
 each
 sample
 (Table
 3.2).
 
Table
 3.2.
 MSNP1AS
 Stranded
 Cuffdiff
 Tophat
 Read
 Count.
 
Sample
  input
  mapped
  %
 mapped
 
MSNP1AS
 KD
 Biological
 Replicate
 1
  30
 943
 117
  29
 969
 022
  96.9
 
MSNP1AS
 KD
 Biological
 Replicate
 2
  16
 186
 531
  15
 361
 602
  94.9
 
MSNP1AS
 KD
 Biological
 Replicate
 3
  17
 890
 746
  16
 932
 628
  94.6
 
MSNP1AS
 KD
 Biological
 Replicate
 4
  28
 801
 262
  27
 527
 759
  95.6
 
Control
 Biological
 Replicate
 1
  11
 099
 318
  10
 746
 390
  96.8
 
Control
 Biological
 Replicate
 2
  20
 496
 459
  19
 422
 292
  94.8
 
Control
 Biological
 Replicate
 3
  10
 824
 013
  10
 458
 919
  96.6
 
Control
 Biological
 Replicate
 4
  31
 400
 592
  29
 709
 967
  94.6
 

 
3.3.8
 Data
 analysis
 
Data
 analysis
 was
 performed
 using
 TopHat
 
 (version
 2.0.10)
 (156)
 to
 align
 the
 Illumina
 
short
 reads
 against
 the
 reference
 human
 genome
 ENSEMBL
 GrCH38
 version
 81.
 Sequence
 
alignments
 were
 generated
 as
 BAM
 files
 (157),
 and
 then
 Cuffdiff
 (version
 2.2.1)
 (158)
 was
 used
 
to
 summarize
 the
 gene
 expression
 values
 as
 FPKM
 measures.
 The
 gene
 expression
 of
 samples
 
with
 MSNP1AS
 knocked
 down
 was
 compared
 to
 the
 gene
 expression
 of
 the
 negative
 control
 
experiment
 samples
 to
 find
 other
 differentially-­‐expressed
 genes.
 Cuffdiff
 was
 also
 used
 to
 
calculate
 the
 expression
 fold
 change,
 P-­‐values
 and
 FDR
 values.
 
 Genes
 with
 FDR
 q<0.05
 were
 
used
 as
 the
 input
 for
 DAVID
 (version
 6.7)
 functional
 annotation
 (159,
 160).
 
3.3.9
 qPCR
 to
 confirm
 altered
 gene
 expression
 observed
 in
 RNA-­‐seq
 
Ten
 RNA
 with
 q<0.05
 in
 the
 RNA-­‐seq
 results
 were
 selected
 to
 validate
 by
 qPCR
 using
 
64
 

 
Taqman
 Gene
 Expression
 assays
 from
 Life
 Technologies:
 ABCA1
 (Assay
 ID
 Hs01059118_m1),
 
ANGPTL4
 (Assay
 ID
 Hs01101127_m1),
 ATF3
 (Assay
 ID
 Hs00231069_m1),
 CXCL8
 (Assay
 ID
 H
 
s00174103_m1),
 GADD45B
 (Assay
 ID
 Hs04188837_g1),
 ID1
 (Assay
 ID
 Hs03676575_s1),
 INSM2
 
(Assay
 ID
 Hs00261625_s1),
 MIR320A
 (Assay
 ID
 Hs04233529_s1),
 OAS2
 (Assay
 ID
 
Hs00942643_m1),
 and
 TMEM215
 (Assay
 ID
 Hs01942498_s1).
 
3.3.10
 Analysis
 of
 Brainspan
 data
 
Normalized
 expression
 data
 representing
 reads
 per
 kilobase
 per
 million
 mapped
 reads
 
(RPKM)
 were
 downloaded
 from
 the
 BrainSpan
 Atlas
 of
 the
 Developing
 Human
 Brain
 database.
 
Details
 on
 the
 tissues,
 materials,
 and
 sequencing
 protocols
 are
 available
 on
 the
 BrainSpan
 
website
 (www.brainspan.org).
 Data
 used
 in
 this
 analysis
 were
 restricted
 to
 neocortical
 brain
 
tissues
 given
 the
 importance
 of
 these
 regions
 in
 ASD
 pathology
 as
 well
 as
 the
 altered
 expression
 
of
 MSNP1AS
 in
 these
 regions
 of
 the
 brain
 (primary
 auditory
 cortex,
 dorsolateral
 prefrontal
 
cortex,
 inferior
 parietal
 cortex,
 inferolateral
 termporal
 cortex,
 primary
 motor
 cortex,
 medial
 
prefrontal
 cortex,
 orbital
 frontal
 cortex,
 primary
 somatosensory
 cortex,
 superior
 temporal
 
cortex,
 primary
 visual
 cortex,
 ventrolateral
 prefrontal
 cortex).
 The
 data
 was
 also
 partitioned
 
into
 10
 age
 groups
 to
 identify
 the
 expression
 pattern
 of
 MSNP1AS
 across
 age.
 Only
 genes
 
showing
 a
 RPKM
 >
 1
 in
 at
 least
 one
 sample
 were
 used
 in
 this
 analysis
 to
 reduce
 the
 amount
 of
 
noise
 generated
 from
 genes
 showing
 extremely
 low
 expression
 values
 in
 all
 samples.
 All
 RPKM
 
values
 were
 then
 transformed
 to
 the
 log2+1
 as
 suggested
 when
 constructing
 a
 weighted
 gene
 
co-­‐expression
 network.
 

 

   
 
65
 

 
3.4
 Results
 
Expression
 of
 the
 ASD-­‐associated
 lncRNA
 MSNP1AS
 is
 increased
 12
 times
 in
 the
 cerebral
 
cortex
 of
 individuals
 with
 ASD
 compared
 to
 controls
 (74).
 We
 used
 transcriptional
 gene
 
silencing
 to
 knock
 down
 MSNP1AS
 and
 thus
 mimic
 a
 potential
 therapeutic
 for
 ASD.
 To
 
quantitatively
 model
 the
 consequences
 of
 MSNP1AS
 knockdown,
 sasRNA-­‐mediated
 knockdown
 
of
 MSNP1AS
 was
 performed
 in
 the
 human
 neural
 progenitor
 cell
 line,
 SK-­‐N-­‐SH.
 The
 knockdown
 
efficiency
 of
 MSNP1AS
 transcript
 levels
 was
 confirmed
 via
 qPCR
 (Figure
 3.2).
 
 
3.4.1
 MSNP1AS
 knockdown
 results
 in
 genome-­‐wide
 transcriptional
 changes
 
 
Changes
 in
 the
 transcriptome
 caused
 by
 MSNP1AS
 knockdown
 may
 be
 able
 to
 reveal
 the
 
impact
 of
 biologically-­‐based
 therapeutic
 approaches
 on
 human
 neural
 progenitor
 cell
 
differentiation.
 The
 transcriptional
 changes
 resulting
 from
 the
 knockdown
 were
 evaluated
 using
 
genome-­‐wide
 transcriptome
 profiling
 of
 human
 neuronal
 progenitor
 cells,
 SK-­‐N-­‐SH,
 via
 RNA-­‐
seq.
 Differential
 gene
 expression
 analysis
 using
 the
 Ensembl
 GrCH38
 genome
 annotation
 
revealed
 1,352
 differentially
 expressed
 genes
 (p<0.05)
 (Table
 3.3).
 When
 a
 more
 stringent
 test
 
for
 multiple
 comparisons
 was
 applied,
 a
 significance
 level
 of
 q<0.05
 showed
 318
 differentially
 
expressed
 genes
 (Table
 3.4,
 Figure
 3.3),
 with
 65%
 of
 the
 genes
 down-­‐regulated.
 Subsequent
 
qPCR
 validation
 was
 performed
 for
 genes
 with
 large
 changes
 in
 expression
 following
 MSNP1AS
 
knockdown.
 Comparison
 of
 gene
 expression
 levels
 measured
 by
 RNA-­‐Seq
 and
 qPCR
 indicated
 a
 
highly
 significant
 Pearson
 correlation
 coefficient
 (Figure
 3.4),
 validating
 the
 magnitude
 of
 gene
 
expression
 changes
 identified
 by
 RNA-­‐seq.
 
 

   
 
66
 

 
3.4.2
 Transcriptional
 consequences
 of
 MSNP1AS
 knockdown
 
 
GO
 analysis
 using
 the
 DAVID
 web
 server
 of
 the
 MSNP1AS
 knockdown
 revealed
 functions
 
important
 in
 chromatin
 assembly
 and
 nucleosome
 organizaton
 (Figure
 3.5).
 We
 also
 performed
 
GO
 analysis
 grouping
 the
 differentially
 expressed
 genes
 (q<0.05)
 by
 their
 direction
 of
 
differential
 expression.
 Genes
 that
 are
 up-­‐regulated
 by
 MSNP1AS
 knockdown
 were
 enriched
 for
 
immune
 response
 (Figure
 3.6A).
 Genes
 that
 are
 down-­‐regulated
 by
 MSNP1AS
 knockdown
 were
 
enriched
 for
 nuclesosome
 assembly
 and
 chromatin
 assembly
 (Figure
 3.6B).
 GO
 analysis
 
indicates
 that
 genes
 encoding
 proteins
 involved
 in
 protein
 synthesis
 and
 chromatin
 regulation
 
were
 altered
 in
 expression
 following
 MSNP1AS
 transcriptional
 gene
 silencing.
 There
 is
 no
 
evidence
 of
 enrichment
 of
 genes
 in
 the
 moesin
 pathway,
 suggesting
 that
 inhibition
 of
 moesin
 
protein
 translation
 is
 independent
 of
 transcriptional
 regulation.
 Comparison
 of
 genes
 with
 
altered
 expression
 following
 MSNP1AS
 transcriptional
 gene
 silencing
 did
 not
 reveal
 enrichment
 
of
 genes
 previously
 implicated
 in
 ASD
 (182).
 
3.4.3
 Expression
 of
 MSNP1AS
 in
 developing
 human
 brain
 
Analysis
 of
 Brainspan
 data
 indicate
 high
 levels
 of
 MSNP1AS
 in
 the
 cerebral
 cortex
 during
 early
 
embryogenesis
 and
 again
 immediately
 after
 birth
 (Figure
 3.7).
 GTex
 indicates
 levels
 of
 
expression
 of
 MSNP1AS
 at
 low
 to
 undetectable
 levels
 in
 12
 regions
 of
 adult
 brain.
 However,
 
GTex
 indicates
 high
 levels
 of
 MSNP1AS
 expression
 in
 peripheral
 blood,
 consistent
 with
 our
 
previous
 report
 (74).
 
 

 

   
 
67
 

 

 
Table
 3.3.
 Top
 thirty
 differentially
 expressed
 genes
 in
 SK-­‐N-­‐SH
 cells
 24
 hrs
 after
 MSNP1AS
 
overexpression.
 
gene
  locus
  log2(fold_change)
  p_value
  q_value
 
OAS2
  12:112906776-­‐113017751
  -­‐7.37
  5.00E-­‐05
  4.12E-­‐03
 
TMEM215
  9:32783498-­‐32787399
  -­‐4.41
  5.00E-­‐05
  4.12E-­‐03
 
IFI44L
  1:78619921-­‐78646145
  -­‐3.87
  5.00E-­‐05
  4.12E-­‐03
 
LRRC55
  11:57181746-­‐57191717
  -­‐3.77
  5.00E-­‐05
  4.12E-­‐03
 
CCL2
  17:34255217-­‐34257203
  -­‐3.49
  5.00E-­‐05
  4.12E-­‐03
 
IFI6
  1:27666060-­‐27703063
  -­‐3.29
  5.00E-­‐05
  4.12E-­‐03
 
CTD-­‐2369P2.8
  19:10251900-­‐10289019
  -­‐3.25
  5.00E-­‐05
  4.12E-­‐03
 
LCP1
  13:46125919-­‐46211871
  -­‐3.08
  5.00E-­‐05
  4.12E-­‐03
 
SOD2
  6:159669056-­‐159789749
  -­‐2.94
  5.00E-­‐05
  4.12E-­‐03
 
MX1
  21:41420303-­‐41459214
  -­‐2.93
  5.00E-­‐05
  4.12E-­‐03
 
C3
  19:6677703-­‐6737603
  -­‐2.92
  5.00E-­‐05
  4.12E-­‐03
 
C1S
  12:6970892-­‐7071032
  -­‐2.79
  5.00E-­‐05
  4.12E-­‐03
 
TLR2
  4:153701499-­‐153705699
  -­‐2.63
  5.00E-­‐05
  4.12E-­‐03
 
ICAM1
  19:10251900-­‐10289019
  -­‐2.60
  5.00E-­‐05
  4.12E-­‐03
 
IL32
  16:3065296-­‐3087100
  -­‐2.52
  5.00E-­‐05
  4.12E-­‐03
 
VCAM1
  1:100719741-­‐100739045
  -­‐2.48
  5.00E-­‐05
  4.12E-­‐03
 
IFI27
  14:94104835-­‐94116698
  -­‐2.45
  5.00E-­‐05
  4.12E-­‐03
 
CMPK2
  2:6840569-­‐6898239
  -­‐2.41
  5.00E-­‐05
  4.12E-­‐03
 
CFB
  6:31897784-­‐31952084
  -­‐2.40
  5.00E-­‐05
  4.12E-­‐03
 
TNFSF18
  1:172775904-­‐173064015
  -­‐2.39
  5.00E-­‐05
  4.12E-­‐03
 
DDX60
  4:168216292-­‐168318807
  -­‐2.36
  5.00E-­‐05
  4.12E-­‐03
 
CTSS
  1:150730195-­‐150765957
  -­‐2.32
  5.00E-­‐05
  4.12E-­‐03
 
IFIH1
  2:162267078-­‐162318703
  -­‐2.32
  5.00E-­‐05
  4.12E-­‐03
 
HERC5
  4:88457116-­‐88506163
  -­‐2.29
  5.00E-­‐05
  4.12E-­‐03
 
MMP12
  11:102862735-­‐102875034
  -­‐2.21
  5.00E-­‐05
  4.12E-­‐03
 
CBLN4
  20:55997439-­‐56005472
  -­‐2.18
  5.00E-­‐05
  4.12E-­‐03
 
PTGS2
  1:186671790-­‐186680427
  -­‐2.12
  5.00E-­‐05
  4.12E-­‐03
 
DDX58
  9:32455704-­‐32526324
  -­‐2.05
  5.00E-­‐05
  4.12E-­‐03
 
CFH
  1:196651877-­‐196747504
  -­‐2.02
  5.00E-­‐05
  4.12E-­‐03
 
IFI44
  1:78649795-­‐78664078
  -­‐2.02
  5.00E-­‐05
  4.12E-­‐03
 

 
68
 

 

 
Figure
 3.2.
 Quantitation
 of
 MSNP1AS
 knockdown.
 Quantitative
 PCR
 showed
 that
 a
 1
 ug
 
transfection
 of
 MSNP1AS
 sasRNA
 resulted
 in
 an
 ~45%
 reduction
 of
 MSNP1AS
 transcript
 
(*p<0.05
 by
 Student
 t-­‐test).
 
0
0.2
0.4
0.6
0.8
1
1.2
Control MSNP1AS
Relative expression
*
69
 

 

 
Figure
 3.3.
 Volcano
 plot
 generated
 to
 show
 the
 differentially
 expressed
 genes
 resulting
 from
 
the
 MSP1AS
 knockdown.
 Each
 dot
 represents
 a
 change
 in
 gene
 expression,
 with
 genes
 above
 
the
 FDR-­‐corrected
 significance
 threshold
 shown
 in
 red.
 
 
 
70
 

 

 
Figure
 3.4.
 Confirmation
 of
 altered
 expression
 of
 genes
 identified
 by
 RNA
 seq.
 Ten
 genes
 with
 
large
 changes
 in
 gene
 expression
 following
 MSNP1AS
 knockdown
 were
 assayed
 via
 qPCR
 to
 
establish
 the
 validity
 of
 the
 RNA-­‐seq
 differential
 expression
 analysis.
 Knockdown
 of
 MSNP1AS
 
was
 also
 validated.
 qPCR
 and
 RNA-­‐seq
 values
 showed
 a
 statistically
 significant
 correlation
 with
 
both
 Pearson
 (P=0.0002)
 and
 Spearman
 (P=0.009)
 methods.
 

 

   
 
OAS2
GADD45B
CXCL8
ABCA1
INSM2
ANGPTL4
ATF3
ID1
MIR320A
TMEM215
MSNP1AS
R² = 0.79821
-­‐6
 
-­‐4
 
-­‐2
 
0
 
2
 
4
 
6
 
8
 
-­‐3
  -­‐2
  -­‐1
  0
  1
  2
  3
  4
  5
  6
 
RNA Seq -log2 fold change
qPCR -DDCt
71
 

 
3.5
 Discussion
 
Our
 findings
 indicate
 that
 MSNP1AS
 knockdown
 disrupted
 the
 expression
 of
 318
 genes
 
(q<0.05),
 many
 of
 which
 are
 involved
 in
 chromatin
 organization
 and
 immune
 response.
 
MSNP1AS
 is
 the
 functional
 element
 revealed
 by
 genome-­‐wide
 significant
 association
 with
 ASD
 
(26,
 74),
 and
 is
 increased
 12.7-­‐fold
 in
 the
 postmortem
 cerebral
 cortex
 of
 individuals
 with
 ASD
 
(74).
 
 We
 test
 the
 effect
 of
 a
 rescue
 of
 this
 phenotype
 by
 transcriptional
 gene
 silencing
 of
 
MSNP1AS
 in
 human
 neurons.
 
 After
 knocking
 down
 the
 lncRNA
 in
 SK-­‐N-­‐SH
 cells,
 we
 evaluated
 
changes
 in
 the
 transcriptome
 using
 RNA-­‐Seq
 to
 elucidate
 the
 consequences
 of
 suppressing
 
MSNP1AS
 expression.
 These
 experiments
 seek
 to
 define
 the
 role
 of
 MSNP1AS
 in
 the
 pathways
 

 
Figure
 3.5.
 Gene
 Ontology
 enrichment
 analysis
 of
 differentially
 expressed
 genes
 (q<0.05)
 in
 SK-­‐
N-­‐SH
 cells
 after
 24
 hours
 of
 MSNP1AS
 knockdown
 revealed
 changes
 in
 chromatin
 assembly.
 
0 1 2 3 4 5 6 7
structural constituent of ribosome
macromolecular complex assembly
structural molecule activity
ribosome
cytosolic ribosome
ribonucleoprotein complex
ribosomal subunit
cellular macromolecular complex assembly
cellular macromolecular complex subunit
chromatin assembly or disassembly
chromatin
DNA packaging
cytosol
translational elongation
nucleosome organization
protein-DNA complex assembly
chromatin assembly
nucleosome assembly
non-membrane-bounded organelle
intracellular non-membrane-bounded
protein-DNA complex
nucleosome
-Log (Bonferroni)
72
 

 

 Figure
 3.6.
 GO
 enrichment
 analysis
 of
 differentially
 expressed
 genes
 (q<0.05)
 in
 SK-­‐N-­‐SH
 cells
 
after
 24
 hours
 of
 MSNP1AS.
 A)
 Analysis
 of
 upregulated
 genes
 revealed
 enrichment
 of
 genes
 
involved
 in
 immune
 response
 following
 knockdown
 of
 MSNP1AS.
 B)
 Analysis
 of
 downregulated
 
genes
 revealed
 changes
 in
 chromatin
 and
 nucleosome
 assembly
 following
 knockdown
 of
 
MSNP1AS.
 
 

   
 
0 2 4 6 8 10 12
positive regulation of immune response
activation of immune response
innate immune response
response to molecule of bacterial origin
humoral immune response
inflammatory response
extracellular region
positive regulation of immune system process
response to wounding
response to virus
defense response
immune response
-Log (Bonferroni)
A
0 1 2 3 4 5 6 7
structural constituent of ribosome
macromolecular complex assembly
structural molecule activity
ribosome
cytosolic ribosome
ribonucleoprotein complex
ribosomal subunit
cellular macromolecular complex assembly
cellular macromolecular complex subunit
chromatin assembly or disassembly
chromatin
DNA packaging
cytosol
translational elongation
nucleosome organization
protein-DNA complex assembly
chromatin assembly
nucleosome assembly
non-membrane-bounded organelle
intracellular non-membrane-bounded organelle
protein-DNA complex
nucleosome
-Log (Bonferroni)
B
73
 

 

 
 
Figure
 3.7.
 Boxplot
 representing
 the
 expression
 of
 MSNP1AS
 through
 human
 cerebral
 cortex
 
development.
 MSNP1AS
 exhibits
 a
 dynamic
 expression
 pattern
 across
 development
 with
 the
 
highest
 expression
 being
 evident
 during
 early
 and
 late
 prenatal
 development.
 
 

 
Table
 3.4.
 Top
 Ten
 Differentially
 Expressed
 Genes
 Following
 Knockdown
 of
 MSNP1AS.
 
Gene
a

  Locus
a

 
Fold
 
Change
 
 
Protein-­‐
 or
 
noncoding
 
P-­‐value
  Q-­‐value
 
OAS2
  12:112906776-­‐113017751
  164.9
  protein-­‐coding
  5.00E-­‐05
  4.12E-­‐03
 
TMEM215
  9:32783498-­‐32787399
  21.24
  protein-­‐coding
  5.00E-­‐05
  4.12E-­‐03
 
IFI44L
  1:78619921-­‐78646145
  14.59
  protein-­‐coding
  5.00E-­‐05
  4.12E-­‐03
 
LRRC55
  11:57181746-­‐57191717
  13.67
  protein-­‐coding
  5.00E-­‐05
  4.12E-­‐03
 
LAMP3
  3:183122212-­‐183163839
  11.53
  protein-­‐coding
  1.00E-­‐04
  7.33E-­‐03
 
ITIH4
  3:52812974-­‐52897596
  -­‐8.11
  protein-­‐coding
  2.00E-­‐04
  1.28E-­‐02
 
CDC42EP1
  22:37560446-­‐37569405
  -­‐10.38
  protein-­‐coding
  2.00E-­‐04
  1.28E-­‐02
 
ANGPTL4
  19:7958578-­‐8489114
  -­‐13.1
  protein-­‐coding
  1.00E-­‐04
  7.33E-­‐03
 
HIST1H4J
  6:27824107-­‐27824480
  -­‐17.32
  protein-­‐coding
  5.00E-­‐05
  4.12E-­‐03
 
MIR320A
  8:22244961-­‐22245043
  -­‐19.42
  noncoding
  5.00E-­‐05
  4.12E-­‐03
 
a

 Refers
 to
 ENSEMBL
 human
 genome
 GrCH38
 build
 81
 

 

   
 
74
 

 
that
 lead
 to
 ASD
 pathogenesis
 and
 suggest
 that
 MSNP1AS
 acts
 at
 the
 transcriptional
 level
 to
 
affect
 immune
 response
 and
 chromatin
 organization.
 
 

  The
 biological
 function
 of
 the
 MSNP1AS
 lncRNA
 revealed
 by
 these
 data,
 chromatin
 
organization,
 is
 similar
 to
 that
 of
 genes
 with
 rare
 de
 novo
 mutations
 associated
 with
 ASD
 (162,
 
183).
 The
 etiology
 of
 ASD
 seems
 to
 be
 effected
 by
 both
 rare
 and
 common
 variants
 in
 similar
 
ways.
 LncRNA
 are
 regulated
 through
 spliceosome
 machinery
 and
 post-­‐translational
 
modifications,
 much
 like
 protein-­‐coding
 transcripts
 (184-­‐187).
 However,
 lncRNA
 are
 more
 
loosely
 regulated,
 are
 produced
 in
 relatively
 lower
 abundance,
 and
 have
 fewer
 exons
 than
 
protein-­‐coding
 genes.
 In
 spite
 of
 this,
 lncRNA
 expression
 is
 extremely
 specific
 to
 cell
 type
 and
 
tissues
 (184,
 186,
 188).
 LncRNAs
 can
 regulate
 proteins
 in
 a
 variety
 of
 ways,
 including
 as
 a
 
framework
 or
 scaffold
 to
 facilitate
 interaction
 with
 other
 proteins
 or
 DNA,
 as
 enhancers
 or
 
silencers
 to
 regulate
 transcription,
 and
 as
 decoys
 to
 bind
 proteins
 or
 other
 RNA
 (185,
 189-­‐197).
 
In
 addition,
 lncRNAs
 are
 being
 evaluated
 for
 epigenetic
 modifications
 and
 chromatin
 
accessibility
 to
 reveal
 their
 contribution
 to
 psychiatric
 disease
 etiology
 (173).
 
Accumulating
 data
 supports
 the
 role
 of
 the
 immune
 system
 in
 the
 pathogenesis
 of
 ASD
 
(198,
 199).
 Our
 previously
 published
 data
 showed
 that
 MSNP1AS
 binds
 moesin
 protein,
 which
 is
 
involved
 in
 neuronal
 process
 stability
 and
 immune
 response
 (74).
 Our
 data
 show
 that
 after
 
MSNP1AS
 is
 knocked
 down,
 MSN
 transcript
 levels
 remain
 unchanged
 (Table
 3.5).
 This
 lack
 of
 
expression
 change
 suggests
 that,
 while
 MSNP1AS
 regulates
 moesin
 protein,
 suppression
 of
 
MSNP1AS
 does
 not
 alter
 MSN
 transcription.
 The
 lack
 of
 MSN
 suppression
 combined
 with
 the
 
decrease
 in
 moesin
 protein
 expression
 shown
 (74)
 suggest
 that
 MSNP1AS
 binds
 to
 the
 MSN
 
transcript
 to
 prevent
 translation.
 Further
 experiments
 will
 be
 necessary
 to
 determine
 the
 
75
 

 
mechanisms
 of
 moesin
 protein
 regulation.
 The
 biological
 function
 revealed
 by
 genes
 
upregulated
 after
 MSNP1AS
 knockdown
 points
 to
 the
 immune
 response
 as
 well.
 Since
 MSN
 
transcript
 levels
 are
 not
 perturbed
 by
 MSNP1AS
 knockdown,
 these
 immune
 response
 genes
 
may
 be
 acting
 independently
 of
 MSN.
 Our
 results
 indicate
 a
 150-­‐fold
 increase
 in
 OAS2
 
expression
 in
 SK-­‐N-­‐SH
 human
 neural
 progenitor
 cells
 following
 MSNP1AS
 transcriptional
 gene
 
silencing.
 OAS2
 is
 one
 of
 three
 genes
 that
 make
 up
 the
 OAS
 gene
 cluster.
 These
 genes
 regulate
 
viral
 infection
 resistance
 and
 are
 involved
 in
 cell
 growth,
 differentiation
 and
 death.
 The
 role
 of
 
the
 OAS
 gene
 cluster
 as
 an
 ASD
 susceptibility
 locus
 is
 currently
 under
 investigation.
 Genetic
 
variants
 in
 OAS2
 are
 associated
 with
 another
 neurodevelopmental
 disorder,
 ADHD
 (200).
 
Table
 3.5.
 Relative
 Fold
 Change
 of
 MSN
 After
 MSNP1AS
 Knockdown
 

 
  MSN
  MSNP1AS
 
 
MSNP1AS
 KD
 Biological
 Replicate
 1
  1.14
  0.62
 
MSNP1AS
 KD
 Biological
 Replicate
 2
  0.5
  0.71
 
MSNP1AS
 KD
 Biological
 Replicate
 3
  1.39
  0.68
 
MSNP1AS
 KD
 Biological
 Replicate
 4
  1.08
  0.17
 

 
Because
 the
 ASD
 phenotype
 is
 highly
 heterogeneous,
 a
 network
 of
 many
 genes
 is
 likely
 
contributing
 to
 abnormal
 regulation.
 A
 variety
 of
 genes
 may
 contribute
 to
 several
 different
 ASD
 
phenotypes;
 however,
 strong
 evidence
 suggests
 that
 many
 of
 the
 pathways
 involved
 may
 
overlap
 to
 form
 an
 interconnected
 network.
 The
 results
 presented
 here
 give
 some
 insight
 into
 
the
 interconnected
 network
 and
 possible
 molecular
 pathology
 of
 ASD.
 Translation
 of
 these
 
findings
 to
 patients
 will
 be
 a
 challenge
 due
 to
 technical
 delivery
 concerns.
 However,
 these
 first
 
steps
 are
 important
 to
 identify
 potential
 biomarkers
 and
 treatment
 targets.
 

 

 
76
 

 
Chapter
 4:
 Distinct
 gene
 expression
 patterns
 resulting
 from
 prenatal
 pesticide
 exposure
 in
 
autism
 spectrum
 disorder.
 
 
In
 preparation.
 

 
4.1.
 Abstract
 
ASD
 Is
 a
 complex
 neurodevelopmental
 disorder
 characterized
 by
 deficits
 in
 social
 
communication
 and
 behavioral
 flexibility.
 The
 degree
 to
 which
 these
 traits
 are
 affected
 is
 
influenced
 by
 both
 genetics
 and
 the
 environment.
 In
 this
 study,
 the
 effect
 of
 pesticide
 exposure
 
in
 relation
 to
 ASD
 diagnosis
 on
 the
 transcriptome
 was
 evaluated
 in
 the
 population-­‐based
 case-­‐
control
 Childhood
 Autism
 Risks
 from
 Genetics
 and
 the
 Environment
 (CHARGE)
 Study.
 Blood
 was
 
taken
 from
 children
 diagnosed
 as
 typically
 developing
 (TD),
 developmentally
 delayed
 (DD),
 or
 
ASD.
 Pesticide
 exposure
 was
 determined
 based
 on
 mapping
 addresses
 from
 the
 pregnancy
 and
 
early
 childhood.
 
 This
 sample
 consisted
 of
 145
 cases
 of
 children
 with
 autism,
 69
 cases
 with
 
developmental
 delay,
 and
 75
 controls
 with
 typical
 development.
 Covariate-­‐controlled
 mixed-­‐
effect
 linear
 models
 were
 used
 to
 identify
 gene
 transcripts
 that
 were
 significantly
 associated
 
with
 diagnostic
 or
 exposure
 status.
 Strand-­‐specific
 RNA
 sequencing
 analysis
 indicated
 altered
 
expression
 of
 many
 genes
 between
 diagnostic
 groups
 and
 following
 comparison
 of
 exposed
 to
 
non-­‐exposed
 samples.
 Children
 with
 ASD
 that
 were
 exposed
 to
 pesticides
 had
 differential
 
expression
 of
 genes
 related
 to
 immune
 response,
 a
 result
 that
 is
 supported
 by
 previous
 studies.
 
TD
 children
 who
 were
 exposed
 to
 pesticides
 had
 differentially
 expressed
 genes
 related
 to
 
protein
 synthesis.
 Protein
 synthesis
 and
 immune
 response
 are
 biological
 processes
 implicated
 
by
 genes
 with
 rare
 mutations
 associated
 with
 ASD.
 Our
 data
 suggest
 that
 the
 functional
 
77
 

 
elements
 implicated
 by
 association
 of
 common
 genetic
 variants
 impact
 the
 same
 biological
 
processes
 that
 are
 influenced
 by
 environmental
 factors.
 This
 study
 reinforces
 the
 need
 to
 
evaluate
 both
 environmental
 and
 genetic
 factors
 when
 considering
 the
 mechanisms
 that
 could
 
be
 converging
 to
 contribute
 to
 ASD
 risk.
 
 

 
4.2.
 Introduction
 
 
Autism
 is
 a
 complex
 neurodevelopmental
 disorder
 that
 effects
 social
 communication
 
and
 behavior.
 As
 recently
 as
 the
 1990s,
 cases
 of
 autism
 increased
 6
 fold
 in
 California
 in
 births
 
counting
 children
 under
 5
 years
 old.
 The
 reason
 for
 the
 recent
 rise
 in
 autism
 is
 to
 some
 degrees
 
unknown.
 Two
 of
 the
 causes
 are
 clear:
 a
 younger
 age
 of
 diagnosis
 accounts
 for
 a
 12%
 increase
 
and
 revision
 of
 diagnostic
 criteria
 accounts
 for
 a
 56%
 increase
 (201).
 As
 it
 stands,
 autism
 affects
 
1
 in
 56
 children
 (202),
 who
 are
 usually
 diagnosed
 between
 the
 ages
 of
 2
 and
 3
 years
 old
 (203).
 
The
 behavioral
 diagnosis
 is
 stringent,
 but
 a
 medical
 diagnosis
 is
 desirable
 due
 to
 the
 wide
 
variance
 in
 phenotype.
 
 
While
 the
 estimated
 heritability
 of
 ASD
 is
 quite
 high,
 around
 90%
 (21-­‐23),
 the
 
contributions
 of
 the
 environment
 are
 not
 negated.
 Evidence
 is
 accumulating
 that
 puts
 the
 
importance
 of
 environmental
 impact
 just
 as
 high
 as
 the
 genetic
 effect
 (204).
 Hundreds
 of
 
genetic
 factors
 may
 be
 contributing
 to
 ASD,
 with
 any
 one
 specific
 variant
 only
 imparting
 a
 tiny
 
percentage
 of
 absolute
 risk
 (205).
 Because
 the
 ASD
 phenotype
 is
 highly
 heterogeneous,
 a
 
network
 of
 many
 genetic
 variants
 (both
 common
 and
 rare)
 is
 likely
 contributing
 to
 abnormal
 
regulation,
 combining
 with
 various
 environmental
 risk
 factors
 in
 a
 multitude
 of
 complex
 ways
 to
 
effect
 specific
 individuals.
 
 
78
 

 
The
 goal
 of
 this
 study
 is
 to
 use
 RNA
 sequencing
 on
 peripheral
 blood
 to
 identify
 
differentially
 expressed
 genes
 in
 children
 exposed
 prenatally
 to
 pesticides
 compared
 to
 children
 
not
 exposed,
 regularly
 exposed
 (6+
 months)
 and
 or
 exposed
 to
 indoor
 pesticides
 (including
 flea
 
collar).
 Three
 groups
 of
 children
 were
 evaluated:
 children
 with
 autism
 spectrum
 disorder
 (ASD),
 
children
 with
 developmental
 delay
 (DD),
 and
 children
 drawn
 from
 the
 general
 population
 of
 
births,
 frequency
 matched
 to
 cases,
 by
 age,
 gender
 &
 geography
 (TD).
 DESeq2
 is
 a
 more
 
conservative
 analysis
 approach,
 but
 is
 able
 to
 reduce
 the
 likelihood
 of
 false
 positives.
 It
 
calculates
 a
 Cook’s
 distance
 statistic
 in
 order
 to
 identify
 genes
 that
 show
 differential
 expression
 
that
 is
 driven
 by
 an
 outlier.
 We
 created
 weighted
 gene
 co-­‐expression
 networks
 to
 examine
 the
 
more
 subtle,
 yet
 significant,
 changes
 in
 the
 transcriptome
 in
 response
 to
 prenatal
 pesticide
 
exposure.
 This
 study
 sought
 to
 examine
 the
 combined
 effect
 of
 genetic
 susceptibility
 for
 ASD,
 
shown
 through
 transcriptomic
 changes,
 and
 pesticide
 exposure
 on
 risk
 of
 ASD
 in
 the
 Childhood
 
Autism
 Risks
 from
 Genetics
 and
 Environment
 (CHARGE)
 study.
 

 
4.3.
 Materials
 and
 Methods
 
4.3.1
 RNA
 sequencing
 analysis
 
Quality
 control
 analyses
 of
 raw
 reads
 from
 RNA-­‐Seq
 was
 carried
 out
 using
 fastqc.
 The
 
files
 that
 passed
 quality
 control
 were
 processed
 using
 Cutdapt
 (version
 1.3)
 to
 remove
 the
 
adapters
 (206).
 The
 8
 replicates
 were
 then
 merged
 to
 create
 one
 "merged
 fastq"
 file.
 Reads
 
were
 then
 aligned
 to
 the
 ENSEMBL
 GrCH38
 version
 81
 transcriptome
 using
 TopHat
 (version
 
2.0.10)
 with
 default
 settings
 (156).
 One
 sample
 with
 read
 counts
 2
 standard
 deviations
 below
 
the
 mean
 was
 removed.
 Accepted_hits.bam
 files
 from
 tophat2
 were
 then
 used
 in
 htseq-­‐count
 
79
 

 
to
 count
 the
 gene
 alignments.
 Htseq-­‐count
 output
 files
 were
 used
 for
 differential
 expression
 
analysis
 using
 the
 R
 program,
 DESeq2,
 using
 default
 options.
 One
 outlier
 (CHG0550)
 was
 
removed.
 Samples
 CHG0863,
 CHG0879,
 CHG0885,
 CHG0907,
 CHG0915,
 CHG0921
 were
 
removed
 due
 to
 apparent
 labeling
 as
 the
 incorrect
 sex.
 The
 data
 was
 adjusted
 for
 age
 and
 sex.
 
Normalized
 expression
 data
 for
 every
 gene
 and
 every
 subject
 were
 then
 extracted
 from
 
DESeq2.
 The
 log2+1
 was
 extracted
 for
 use
 in
 gene
 co-­‐expression
 network
 analyses.
 Gene
 co-­‐
expression
 networks
 were
 then
 formed
 for
 all
 three
 diagnostic
 groups
 using
 WGCNA
 (version
 
1.51).
 
 
 
4.3.2
 Functional
 enrichment
 
DAVID
 (version
 6.7)
 (207)
 was
 used
 to
 functionally
 annotate
 differentially
 expressed
 
genes
 and
 gene
 co-­‐expression
 modules
 of
 interest.
 This
 enables
 identification
 of
 enriched
 gene
 
ontology
 terms.
 Redundant
 terms
 or
 terms
 with
 a
 Bonferroni-­‐corrected
 p-­‐value
 >
 0.05
 were
 
removed.
 The
 analyses
 were
 carried
 out
 using
 default
 options.
 
4.3.3
 Weighted
 Gene
 Co-­‐Expression
 Network
 Analysis
 (WGCNA)
 
Weighted
 gene
 co-­‐expression
 network
 analysis
 (WGCNA)
 uses
 correlation
 networks
 to
 
describe
 networks
 of
 co-­‐expressed
 genes
 that
 are
 related
 to
 particular
 phenotypic
 traits.
 This
 
system’s
 biological
 technique
 reveals
 patterns
 across
 microarray
 or
 RNA-­‐Seq
 samples
 (208).
 The
 
approach
 finds
 groups
 of
 genes
 (clusters),
 summarizes
 them
 using
 a
 hub
 gene
 or
 eigengene,
 
and
 calculates
 measurements
 of
 the
 module
 including
 connection
 strengths.
 The
 connectivity
 
level
 calculations
 determine
 the
 organization
 of
 the
 modules
 based
 on
 gene
 clustering.
 The
 hub
 
gene
 has
 to
 be
 related
 to
 the
 phenotypic
 data
 for
 the
 modules
 and
 the
 trait
 to
 be
 considered
 
related.
 
80
 

 
Normalized
 gene
 expression
 values
 were
 collected
 for
 each
 sample
 using
 Cuffdiff.
 The
 
genes
 used
 in
 the
 coexpression
 network
 were
 restricted
 to
 those
 showing
 a
 minimum
 of
 0.1
 
FPKM
 to
 reduce
 the
 amount
 of
 noise
 in
 the
 network.
 This
 resulted
 in
 27,984
 expressed
 genes.
 
The
 WGCNA
 method
 creates
 an
 adjacency
 matrix
 from
 the
 gene
 correlation
 values
 using
 a
 
power
 function.
 The
 power
 was
 selected
 from
 a
 scale
 free
 topology
 fit
 and
 a
 power
 of
 20
 was
 
chosen
 for
 the
 network.
 The
 ‘blockwiseModules’
 function
 was
 utilized
 for
 network
 
construction.
 To
 facilitate
 the
 formation
 of
 large
 and
 distinct
 modules,
 the
 minimum
 module
 
size
 was
 set
 to
 100
 with
 the
 minimum
 module
 membership
 connectivity
 (kME)
 set
 at
 0.7.
 
Modules
 showing
 a
 correlation
 value
 of
 0.75
 were
 merged.
 The
 eigengene
 of
 the
 formed
 
modules
 was
 then
 extracted
 and
 a
 Pearson’s
 correlation
 was
 calculated
 with
 the
 cell
 collection
 
time.
 These
 calculations
 were
 used
 to
 identify
 biologically
 interesting
 modules
 for
 further
 
analysis.
 Modules
 were
 selected
 for
 additional
 analysis
 if
 the
 correlations
 with
 cell
 collection
 
time
 had
 a
 p-­‐value
 <0.1.
 
4.3.4
 Module
 Preservation
 
The
 differences
 between
 the
 three
 diagnostic
 groups
 are
 examined
 by
 testing
 how
 well
 the
 
modules
 are
 preserved
 in
 the
 three
 networks.
 The
 DD
 and
 ASD
 modules
 were
 tested
 to
 see
 how
 
well
 they
 are
 preserved
 with
 the
 TD
 modules
 by
 calculating
 a
 Z
 statistic.
 This
 helps
 answer
 the
 
question
 “Do
 the
 genes
 in
 the
 test
 module
 match
 the
 control
 module
 significantly
 better
 than
 a
 
random
 sample
 of
 genes?”
 The
 individual
 Z
 scores
 are
 then
 summarized
 into
 a
 composite
 
measure
 called
 a
 Zsummary.
 A
 Zsummary
 >10
 is
 considered
 to
 be
 highly
 preserved.
 
 

 

   
 
81
 

 
4.4
 Results
 
4.4.1
 All
 differentially
 expressed
 genes:
 exposed
 vs
 nonexposed
 
When
 controlling
 for
 age,
 sex,
 educational
 level
 of
 the
 mother,
 and
 the
 number
 of
 hours
 
since
 the
 last
 meal
 before
 the
 blood
 draw,
 twenty
 genes
 are
 differentially
 expressed
 regardless
 
of
 diagnosis:
 twelve
 were
 protein-­‐coding
 and
 eight
 were
 ncRNA
 (q<0.05).
 Three
 genes
 had
 an
 
increase
 in
 expression
 and
 17
 had
 a
 decrease
 in
 expression
 (Table
 4.1).
 When
 analyzed
 by
 
DAVID,
 9
 GO
 terms
 achieved
 significance.
 The
 category
 with
 the
 lowest
 p-­‐value
 was
 “defense
 
response
 to
 virus”
 (p=
 1.95E-­‐03).
 
 When
 a
 less
 stringent
 statistical
 test
 was
 applied,
 the
 number
 
of
 differentially
 expressed
 genes
 was
 2,829
 genes
 (p<0.05).
 When
 analyzed
 by
 DAVID,
 10
 GO
 
terms
 achieved
 significance.
 The
 category
 with
 the
 lowest
 Bonferroni-­‐corrected
 p-­‐value
 was
 
“extracellular
 exosome”
 (p=
 7.80E-­‐09).
 
 
4.4.2
 TD
 Exposed
 vs
 Nonexposed
 
In
 the
 TD
 exposed
 vs
 TD
 not
 exposed
 group,
 15
 genes
 were
 differentially
 expressed
 (Table
 4.2,
 
Figure
 4.1)(q
 <
 0.05).
 Two
 genes
 had
 an
 increase
 in
 expression
 in
 exposed
 children:
 both
 
noncoding.
 Thirteen
 genes
 decreased
 in
 expression:
 5
 noncoding
 and
 8
 protein
 coding
 (Table
 
4.2).
 When
 analyzed
 by
 DAVID,
 one
 GO
 term
 achieved
 significance.
 “Anchored
 component
 of
 
membrane”
 had
 a
 p-­‐value
 of
 p=
 3.66E-­‐02.
 When
 a
 less
 stringent
 statistical
 test
 was
 applied,
 the
 
number
 of
 differentially
 expressed
 genes
 was
 2,006
 genes
 (p<0.05).
 When
 analyzed
 by
 the
 
DAVID,
 10
 GO
 terms
 achieved
 significance.
 The
 category
 with
 the
 lowest
 Bonferroni-­‐corrected
 
p-­‐value
 was
 “focal
 adhesion”
 (p=
 2.63E-­‐05).
 

 

   
 
82
 

 
Table
 4.1.
 Differentially
 Expressed
 Genes
 (Exposed
 vs.
 Not
 Exposed)
 
Gene
  Gene
 Type
  Log
 2
 Fold
 Change
  q
 value
 
OR7E115P
  unprocessed
 pseudogene
  -­‐2.16
  4.23E-­‐02
 
MTATP8P2
  processed
 pseudogene
  -­‐2.14
  1.95E-­‐02
 
MTCO2P12
  unprocessed
 pseudogene
  -­‐1.79
  5.59E-­‐11
 
LDLRAD1
  protein
 coding
  -­‐1.53
  9.82E-­‐09
 
CCL8
  protein
 coding
  -­‐1.33
  1.68E-­‐02
 
CXCL10
  protein
 coding
  -­‐1.31
  2.03E-­‐05
 
RP11-­‐5K23.5
  antisense
  -­‐1.14
  8.98E-­‐03
 
RSAD2
  protein
 coding
  -­‐1.08
  4.79E-­‐02
 
ISG15
  protein
 coding
  -­‐0.92
  1.22E-­‐02
 
CPNE4
  protein
 coding
  -­‐0.90
  1.76E-­‐02
 
MTRNR2L12
  protein
 coding
  -­‐0.89
  2.62E-­‐05
 
SEPT4
  protein
 coding
  -­‐0.83
  8.98E-­‐03
 
LAMP3
  protein
 coding
  -­‐0.81
  1.76E-­‐02
 
IGHV3-­‐73
  protein
 coding
  -­‐0.71
  3.32E-­‐04
 
RP11-­‐493L12.6
  processed
 pseudogene
  -­‐0.67
  4.23E-­‐02
 
RPL39L
  protein
 coding
  -­‐0.58
  6.13E-­‐04
 
NT5C3A
  protein
 coding
  -­‐0.33
  4.97E-­‐02
 
ACTG1P20
  processed
 pseudogene
  0.73
  8.98E-­‐03
 
MTND6P4
  processed
 pseudogene
  1.11
  2.90E-­‐02
 
SEPT7P3
  unprocessed
 pseudogene
  1.97
  2.50E-­‐02
 

   
 
Table
 4.2.
 Differentially
 Expressed
 Genes
 (TD
 Exposed
 vs.
 Not
 Exposed)
 
Gene
  Gene
 Type
  Log
 2
 Fold
 Change
  q
 value
 
CD177
  protein
 coding
  -­‐3.62
  4.10E-­‐02
 
RP11-­‐166N6.2
  antisense
  -­‐2.43
  5.75E-­‐03
 
C1orf226
  protein
 coding
  -­‐2.40
  2.11E-­‐03
 
MSLN
  protein
 coding
  -­‐2.39
  1.27E-­‐04
 
MTCO1P40
  processed
 pseudogene
  -­‐2.09
  1.19E-­‐02
 
CPNE4
  protein
 coding
  -­‐1.82
  1.31E-­‐02
 
EIF5AP2
  processed
 pseudogene
  -­‐1.77
  1.37E-­‐02
 
FSTL4
  protein
 coding
  -­‐1.69
  3.94E-­‐02
 
MTRNR2L12
  protein
 coding
  -­‐1.64
  3.94E-­‐02
 
RP11-­‐169F17.1
  processed
 transcript
  -­‐1.53
  3.94E-­‐02
 
RPL39L
  protein
 coding
  -­‐1.47
  8.81E-­‐03
 
TDRD9
  protein
 coding
  -­‐1.21
  4.10E-­‐02
 
XIST
  lincRNA
  1.23
  1.37E-­‐02
 
MTCO3P12
  unprocessed
 pseudogene
  3.08
  3.94E-­‐02
 

 

   
 
83
 

 
4.4.3
 DD
 Exposed
 vs
 Nonexposed
 
In
 the
 DD
 exposed
 vs
 DD
 not
 exposed
 group,
 the
 number
 of
 differentially
 expressed
 
genes
 with
 a
 p-­‐value
 of
 less
 than
 0.05
 was
 2,171
 genes
 (Figure
 4.1).
 When
 analyzed
 by
 the
 
DAVID,
 16
 GO
 terms
 achieved
 Bonferroni-­‐corrected
 significance.
 The
 category
 with
 the
 lowest
 
Bonferroni-­‐corrected
 p-­‐value
 was
 “SRP-­‐dependent
 cotranslational
 protein
 targeting
 to
 
membrane”
 (p=
 4.54E-­‐60).
 
 
4.4.4
 ASD
 Exposed
 vs
 Nonexposed
 
In
 the
 ASD
 exposed
 vs
 ASD
 not
 exposed
 group,
 13
 genes
 were
 differentially
 expressed
 
(Table
 4.3,
 Figure
 4.1).
 Two
 genes
 had
 an
 increase
 in
 expression
 in
 exposed
 children:
 both
 
noncoding.
 Eleven
 decreased
 in
 expression:
 two
 were
 noncoding
 and
 nine
 were
 protein
 coding.
 
When
 analyzed
 by
 DAVID,
 two
 GO
 terms
 achieved
 significance
 (Figure
 4.2).
 “Type
 I
 interferon
 
signaling
 pathway”
 had
 a
 p-­‐value
 of
 p=
 1.52E-­‐02.
 When
 a
 less
 stringent
 statistical
 test
 was
 
applied,
 the
 number
 of
 differentially
 expressed
 genes
 was
 2,132
 genes
 (p<0.05).
 When
 analyzed
 
by
 DAVID,
 14
 GO
 terms
 achieved
 significance.
 The
 category
 with
 the
 lowest
 Bonferroni-­‐
corrected
 p-­‐value
 was
 “defense
 response
 to
 virus”
 (p=
 7.59E-­‐13).
 

 

   
 
84
 

 
Table
 4.3.
 Differentially
 Expressed
 Genes
 (ASD
 Exposed
 vs.
 Not
 Exposed)
 
Gene
  Gene
 Type
  Log
 2
 Fold
 Change
  q
 value
 
MTCO1P42
  processed
 pseudogene
  -­‐3.84
  2.79E-­‐03
 
MTCO2P12
  unprocessed
 pseudogene
  -­‐2.63
  2.50E-­‐17
 
IFI27
  protein
 coding
  -­‐1.88
  9.52E-­‐03
 
CXCL10
  protein
 coding
  -­‐1.45
  2.79E-­‐03
 
ISG15
  protein
 coding
  -­‐1.10
  3.54E-­‐02
 
IGHV3-­‐73
  protein
 coding
  -­‐1.10
  4.47E-­‐05
 
SEPT4
  protein
 coding
  -­‐1.02
  3.33E-­‐02
 
IGLV4-­‐69
  protein
 coding
  -­‐0.93
  4.69E-­‐02
 
IGHV3-­‐49
  protein
 coding
  -­‐0.83
  3.53E-­‐02
 
IGHV3-­‐33
  protein
 coding
  -­‐0.77
  3.55E-­‐02
 
ACOXL
  protein
 coding
  -­‐0.54
  4.69E-­‐02
 
MTCO3P12
  unprocessed
 pseudogene
  2.14
  1.71E-­‐04
 
SEPT7P3
  unprocessed
 pseudogene
  3.18
  7.93E-­‐04
 

 

 

 
Figure
 4.1.
 Volcano
 plot
 generated
 to
 show
 the
 differentially
 expressed
 genes
 between
 the
 
exposed
 to
 pesticides
 group
 and
 the
 unexposed.
 A)
 Typically
 developing.
 B)
 Developmentally
 
Delayed.
 C)
 Autism
 Spectrum
 Disorder.
 Each
 dot
 represents
 a
 change
 in
 gene
 expression,
 with
 
genes
 above
 the
 FDR-­‐corrected
 significance
 threshold
 shown
 in
 red.
 
 
 

 
85
 

 

 

 
Figure
 4.2.
 GO
 enrichment
 analysis
 of
 differentially
 expressed
 genes
 (q<0.05)
 of
 the
 exposed
 to
 
pesticides
 versus
 not
 exposed
 ASD
 children.
 Analysis
 revealed
 enrichment
 of
 genes
 involved
 in
 
defense
 response
 to
 virus.
 

 

   
 
0
  5
  10
 
defense
 response
 to
 virus
 
type
 I
 interferon
 signaling
 pathway
 
response
 to
 virus
 
negative
 regulation
 of
 viral
 genome
 
extracellular
 region
 
inDlammatory
 response
 
chemokine-­‐mediated
 signaling
 pathway
 
extracellular
 space
 
innate
 immune
 response
 
immune
 response
 
endoplasmic
 reticulum
 chaperone
 complex
 
chemokine
 activity
 
heparin
 binding
 
endoplasmic
 reticulum
 lumen
 
-­‐Log
 (Bonferroni-­‐corrected
 p-­‐value)
 
ASD
 Non-­‐exposed
 vs
 Exposed
 
86
 

 
4.4.5
 All
 ASD
 vs
 DD
 
All
 ASD
 genes
 were
 compared
 to
 all
 DD
 genes
 regardless
 of
 exposure
 and
 revealed
 35
 
differentially
 expressed
 genes.
 Ten
 genes
 had
 an
 increase
 in
 expression
 in
 exposed
 children.
 
Twenty-­‐five
 genes
 decreased
 in
 expression.
 When
 analyzed
 by
 the
 DAVID,
 6
 GO
 terms
 achieved
 
significance
 (p<0.05).
 “Extracellular
 space”
 had
 a
 p
 value
 of
 4.80E-­‐3.
 When
 a
 less
 stringent
 
statistical
 test
 was
 applied,
 the
 number
 of
 differentially
 expressed
 genes
 was
 2,989
 genes
 
(p<0.05).
 When
 analyzed
 by
 DAVID,
 32
 GO
 terms
 achieved
 significance.
 The
 category
 with
 the
 
lowest
 Bonferroni-­‐corrected
 p-­‐value
 was
 “SRP-­‐dependent
 cotranslational
 protein
 targeting
 to
 
membrane”
 (p=
 1.26E-­‐56).
 
4.4.6
 All
 ASD
 vs
 TD
 
All
 ASD
 genes
 were
 compared
 to
 all
 TD
 genes
 regardless
 of
 exposure
 and
 revealed
 14
 
differentially
 expressed
 genes
 with
 a
 q-­‐value
 of
 less
 than
 0.05.
 Three
 genes
 had
 an
 increase
 in
 
expression
 in
 exposed
 children.
 Eleven
 decreased
 in
 expression.
 When
 a
 less
 stringent
 statistical
 
test
 was
 applied,
 the
 number
 of
 differentially
 expressed
 genes
 was
 3,171
 genes
 (p<0.05).
 When
 
analyzed
 by
 DAVID,
 5
 GO
 terms
 achieved
 significance.
 The
 category
 with
 the
 lowest
 Bonferroni-­‐
corrected
 p-­‐value
 was
 “poly(A)
 RNA
 binding”
 (p=
 4.34E-­‐12).
 
4.4.7
 Differentially
 Expressed
 Genes
 Separated
 by
 Age
 
There
 was
 a
 significant
 difference
 in
 response
 to
 exposure
 between
 age
 groups.
 The
 
children
 were
 split
 into
 two
 groups:
 3.5
 years
 and
 younger
 and
 3.6
 years
 and
 older.
 The
 
exposed
 and
 not
 exposed
 children
 were
 then
 compared
 in
 each
 diagnostic
 group.
 
 

 
87
 

 

 
Figure
 4.3.
 Venn
 diagrams
 of
 differentially
 expressed
 genes
 across
 age
 groups.
 TD
 children
 
under
 3.5
 years
 had
 33
 differentially
 expressed
 gene
 in
 common
 with
 the
 children
 over
 3.5
 
years.
 DD
 children
 under
 3.5
 years
 had
 34
 differentially
 expressed
 gene
 in
 common
 with
 the
 
children
 over
 3.5
 years.
 ASD
 children
 under
 and
 the
 children
 over
 3.5
 years
 had
 59
 differentially
 
expressed
 genes
 in
 common
 (q<0.05).
 
 

 
4.4.7.1
 TD
 Exposed
 vs
 Nonexposed
 
In
 the
 TD
 exposed
 vs
 TD
 not
 exposed
 group,
 the
 younger
 children
 had
 56
 differentially
 
expressed
 genes.
 In
 older
 children,
 100
 genes
 were
 differentially
 expressed.
 Thirty-­‐three
 of
 
these
 genes
 were
 the
 same
 between
 age
 groups
 (Figure
 4.3).
 The
 genes
 with
 a
 q
 value
 of
 less
 
than
 0.05
 were
 analyzed
 by
 DAVID.
 Eleven
 GO
 terms
 resulting
 from
 the
 older
 children
 achieved
 
a
 significance
 of
 p<0.05.
 The
 category
 with
 the
 highest
 significance
 was
 “histone
 demethylase
 
activity”
 with
 a
 Bonferroni-­‐corrected
 p-­‐value
 of
 9.60E-­‐03.
 Thirteen
 GO
 terms
 resulting
 from
 the
 
younger
 children
 achieved
 a
 significance
 of
 p<0.05.
 The
 category
 with
 the
 highest
 significance
 
was
 “translational
 initiation”
 with
 a
 Bonferroni-­‐corrected
 p-­‐value
 of
 4.02E-­‐04.
 
4.4.7.2
 DD
 Exposed
 vs
 Nonexposed
 
In
 the
 DD
 exposed
 vs
 DD
 not
 exposed
 group,
 the
 younger
 children
 had
 149
 differentially
 
expressed
 genes.
 
 In
 the
 older
 children,
 86
 genes
 were
 differentially
 expressed.
 Thirty-­‐four
 of
 
these
 genes
 were
 the
 same
 between
 age
 groups
 (Figure
 4.3).
 The
 genes
 with
 a
 q
 value
 of
 less
 
than
 0.05
 were
 analyzed
 by
 DAVID.
 Nine
 GO
 terms
 resulting
 from
 the
 older
 children
 achieved
 a
 
significance
 of
 p<0.05.
 The
 category
 with
 the
 highest
 significance
 was
 “histone
 demethylase
 
88
 

 
activity”
 with
 a
 Bonferroni-­‐corrected
 p-­‐value
 of
 1.26
 E-­‐02.
 Eleven
 GO
 terms
 resulting
 from
 the
 
younger
 children
 achieved
 a
 significance
 of
 p<0.05.
 The
 category
 with
 the
 highest
 significance
 
was
 “translational
 initiation”
 with
 a
 Bonferroni-­‐corrected
 p-­‐value
 of
 2.95E-­‐04
 (Figure
 4.4).
 

 
Figure
 4.4.
 GO
 enrichment
 analysis
 of
 differentially
 expressed
 genes
 (p<0.05)
 of
 the
 TD
 children
 
exposed
 versus
 not
 exposed
 to
 pesticides
 in
 the
 younger
 (less
 than
 3.5
 years)
 children.
 Analysis
 
revealed
 enrichment
 of
 genes
 involved
 in
 translational
 initiation,
 rRNA
 binding
 and
 cytosolic
 
small
 ribosomal
 subunit.
 
 

 

 4.4.7.3
 ASD
 Exposed
 vs
 Nonexposed
 
In
 the
 ASD
 exposed
 vs
 ASD
 not
 exposed
 group,
 the
 younger
 children
 had
 152
 
differentially
 expressed
 genes.
 
 In
 the
 older
 children,
 152
 genes
 were
 differentially
 expressed.
 
Fifty-­‐nine
 of
 these
 genes
 were
 the
 same
 between
 age
 groups
 (Figure
 4.3).
 The
 genes
 with
 a
 q
 
value
 of
 less
 than
 0.05
 were
 analyzed
 by
 DAVID.
 Seventeen
 GO
 terms
 resulting
 from
 the
 older
 
children
 achieved
 a
 significance
 of
 p<0.05.
 The
 category
 with
 the
 highest
 significance
 was
 
0
  2
  4
  6
 
translational
 initiation
 
histone
 demethylase
 activity
 
rRNA
 binding
 
cytosolic
 small
 ribosomal
 subunit
 
translation
 initiation
 factor
 activity
 
histone
 H3-­‐K27
 demethylation
 
histone
 demethylase
 activity
 (H3-­‐K27
 
histone
 demethylase
 activity
 (H3-­‐K4
 speciDic)
 
histone
 H3-­‐K4
 demethylation
 
oxidoreductase
 activity,
 acting
 on
 paired
 
dioxygenase
 activity
 
cytoplasmic
 ribonucleoprotein
 granule
 
-­‐Log
 (Bonferroni-­‐corrected
 p-­‐value)
 
TD
 Younger
 Non-­‐exposed
 vs
 Exposed
 
89
 

 
“histone
 demethylase
 activity”
 with
 a
 Bonferroni-­‐corrected
 p-­‐value
 of
 4.39
 E-­‐03.
 Fifteen
 GO
 
terms
 resulting
 from
 the
 younger
 children
 achieved
 a
 significance
 of
 p<0.05.
 The
 category
 with
 
the
 highest
 significance
 was
 also
 “histone
 demethylase
 activity”
 with
 a
 Bonferroni-­‐corrected
 p-­‐
value
 of
 2.74E-­‐03.
 
4.4.7.4
 Exposed
 ASD
 vs
 Exposed
 TD
 
The
 genes
 of
 children
 with
 ASD
 who
 were
 exposed
 to
 pesticides
 were
 compared
 to
 the
 
genes
 of
 DD
 children
 exposed
 to
 pesticides
 separated
 by
 age
 and
 revealed
 5
 differentially
 
expressed
 genes
 (q<0.05).
 When
 less
 stringent
 significance
 requirements
 were
 applied,
 3,399
 
genes
 were
 differentially
 expressed
 (p<0.05).
 When
 the
 genes
 with
 a
 significant
 p-­‐value
 were
 
analyzed
 by
 DAVID,
 thirty
 GO
 terms
 achieved
 Bonferroni-­‐corrected
 significance
 in
 the
 older
 
children.
 The
 category
 with
 the
 highest
 Bonferroni-­‐corrected
 significance
 was
 “nucleosome“
 
(q=2.71E-­‐31).
 When
 the
 younger
 children
 were
 compared,
 3
 genes
 with
 a
 significant
 q-­‐value
 
were
 differentially
 expressed.
 When
 less
 stringent
 significance
 requirements
 were
 applied,
 770
 
genes
 were
 differentially
 expressed
 (p<0.05).
 When
 the
 genes
 with
 a
 significant
 p-­‐value
 were
 
analyzed
 by
 DAVID,
 eight
 GO
 terms
 achieved
 Bonferroni-­‐corrected
 significance
 in
 the
 younger
 
children.
 The
 category
 with
 the
 highest
 Bonferroni-­‐corrected
 significance
 was
 “type
 I
 interferon
 
signaling
 pathway”
 (q=
 1.77E-­‐13)
 (Figure
 4.5).
 
4.4.7.5
 Exposed
 ASD
 vs
 Exposed
 DD
 
The
 genes
 of
 children
 with
 ASD
 who
 were
 exposed
 to
 pesticides
 were
 compared
 to
 the
 
genes
 of
 DD
 children
 exposed
 to
 pesticides
 separated
 by
 age
 and
 revealed
 differentially
 
expressed
 1,242
 genes
 (p<0.05).
 When
 the
 genes
 with
 a
 significant
 p-­‐value
 were
 analyzed
 by
 
DAVID,
 82
 GO
 terms
 achieved
 significance
 in
 the
 older
 children
 (p<0.05).
 The
 category
 with
 the
 
90
 

 
highest
 Bonferroni-­‐corrected
 significance
 was
 “respiratory
 chain
 “
 (q=
 0.012763523).
 When
 the
 
younger
 children
 were
 compared,
 2
 genes
 with
 a
 significant
 q-­‐value
 were
 differentially
 
expressed.
 When
 less
 stringent
 significance
 requirements
 were
 applied,
 1,635
 genes
 were
 
differentially
 expressed
 (p<0.05).
 When
 the
 genes
 with
 a
 significant
 p-­‐value
 were
 analyzed
 by
 
DAVID,
 no
 GO
 terms
 achieved
 Bonferroni-­‐corrected
 significance
 in
 the
 younger
 children,
 but
 52
 
GO
 terms
 had
 a
 significant
 p-­‐value
 
 (p<0.05).
 The
 category
 with
 the
 highest
 Bonferroni-­‐
corrected
 significance
 was
 “calcium
 ion
 binding”
 (q=
 0.085).
 
 
4.4.8
 Weighted
 gene
 co-­‐expression
 network
 analysis
 
Sorting
 and
 identifying
 transcripts
 involved
 in
 the
 biological
 regulatory
 pathways
 is
 a
 
daunting
 task
 due
 to
 the
 volume
 of
 RNAs
 that
 remain
 uncharacterized.
 Protein-­‐coding
 and
 
noncoding
 RNAs
 activated
 as
 a
 response
 to
 pesticide
 exposure
 were
 sorted
 based
 on
 
developmental
 category
 and
 a
 weighted
 gene
 co-­‐expression
 network
 was
 constructed.
 The
 
WGCNA
 formed
 modules
 of
 transcripts
 that
 have
 highly
 correlated
 expression
 patterns.
 This
 
narrowed
 the
 list
 of
 RNAs
 possibly
 regulating
 autism
 spectrum
 disorder
 and
 allowed
 for
 the
 
identification
 of
 transcriptional
 networks
 involved
 (Figure
 4.6).
 

   
 
91
 

 

 
Figure
 4.5.
 GO
 enrichment
 analysis
 of
 differentially
 expressed
 genes
 (q<0.05)
 of
 the
 ASD
 
children
 exposed
 to
 pesticides
 versus
 TD
 children
 exposed
 to
 pesticides
 in
 the
 younger
 (less
 
than
 3.5
 years)
 children.
 Analysis
 revealed
 enrichment
 of
 genes
 involved
 in
 type
 1
 interferon
 
signaling
 pathway,
 a
 pathway
 involved
 in
 transcription,
 response
 to
 virus,
 and
 the
 immune
 
response.
 
 

 

   
   
 
Figure
 4.6.
 Weighted
 Gene
 Co-­‐Expression
 Network
 Analysis.
 (A)
 Cluster
 gene
 dendrogram
 of
 
co-­‐expressed
 genes
 revealed
 25
 modules
 of
 co-­‐expressed
 genes
 not
 including
 the
 gray
 module.
 
 
4.5
 Discussion
 
In
 ASD
 children,
 prenatal
 exposure
 to
 regular
 use
 of
 household
 pesticides
 resulted
 in
 
about
 3
 times
 as
 many
 differentially
 expressed
 genes
 compared
 to
 exposed
 TD
 and
 DD
 children.
 
Genes
 that
 were
 identified
 as
 differentially
 expressed
 in
 exposed
 children
 with
 ASD
 were
 
enriched
 in
 gene
 ontology
 terms
 related
 to
 the
 immune
 response.
 
Prenatal
 pesticide
 exposure
 significantly
 changes
 the
 expression
 of
 gene
 modules
 
important
 in
 immunity
 response.
 Children
 with
 ASD
 do
 not
 show
 the
 same
 expression
 pattern
 
0.00
  5.00
  10.00
  15.00
  20.00
 
type
 I
 interferon
 signaling
 pathway
 
defense
 response
 to
 virus
 
response
 to
 virus
 
negative
 regulation
 of
 viral
 genome
 
interferon-­‐gamma-­‐mediated
 signaling
 
immune
 response
 
2'-­‐5'-­‐oligoadenylate
 synthetase
 activity
 
double-­‐stranded
 RNA
 binding
 
ASD
 vs
 TD
 Exposed
 Younger
 
92
 

 
in
 these
 modules
 as
 seen
 in
 exposed
 TD
 and
 DD
 children.
 The
 GO
 category,
 immune
 response,
 
is
 a
 top
 category
 in
 3
 comparisons:
 All
 Exposed
 vs
 Unexposed
 ASD,
 Younger
 ASD
 exposed
 vs
 DD
 
exposed,
 and
 Younger
 ASD
 exposed
 vs
 TD
 exposed.
 Multiple
 studies
 have
 pointed
 to
 the
 role
 of
 
the
 immune
 system
 in
 the
 etiology
 of
 ASD
 (198,
 199).
 Our
 previously
 published
 data
 showed
 
that
 MSNP1AS,
 a
 long
 non-­‐coding
 RNA
 associated
 with
 ASD,
 binds
 moesin
 protein,
 which
 is
 
involved
 in
 neuronal
 process
 stability
 and
 immune
 response
 (67,
 74).
 When
 MSNP1AS
 is
 
dysregulated,
 the
 expression
 of
 gene
 transcripts
 involved
 in
 the
 immune
 response
 is
 changed
 
(69).
 Immune
 response
 pathways
 seem
 to
 be
 activated
 in
 ASD
 children
 in
 response
 to
 pesticide
 
exposure.
 
 

  The
 biological
 pathways
 impacted
 by
 pesticide
 exposure
 in
 TD
 children
 are
 associated
 
with
 protein
 translation,
 While
 these
 pathways
 are
 similar
 to
 that
 of
 processes
 associated
 with
 
ASD
 (209,
 210),
 they
 were
 activated
 in
 TD
 children
 without
 resulting
 in
 ASD.
 It
 is
 possible
 that
 
the
 appropriate
 response
 to
 pesticide
 exposure
 is
 a
 change
 in
 protein
 translation
 pathways,
 
while
 an
 inappropriate
 reaction
 is
 an
 immune
 response,
 which
 potentially
 triggers
 ASD.
 Another
 
interpretation
 would
 be
 that
 children
 with
 ASD
 activate
 immune
 response
 genes
 instead
 of
 
protein
 translation
 genes.
 It
 is
 unclear
 whether
 the
 immune
 response
 precedes
 or
 comes
 after
 
the
 development
 of
 ASD
 or
 is
 caused
 by
 the
 ASD.
 
 
While
 the
 disruption
 of
 protein
 translation
 processes
 as
 a
 result
 of
 pesticide
 exposure
 
may
 not
 be
 enough
 on
 its
 own
 to
 be
 causative,
 the
 changes
 in
 these
 processes
 reflect
 suspected
 
etiology
 of
 ASD.
 The
 gene
 networks
 affected
 by
 pesticide
 exposure
 and
 contributing
 to
 genetic
 
risk
 are
 likely
 working
 together.
 Given
 that
 the
 ASD
 phenotype
 is
 highly
 heterogeneous,
 a
 
plethora
 of
 genes
 and
 environmental
 risk
 factors
 may
 contribute
 to
 several
 different
 ASD
 
93
 

 
phenotypes;
 however,
 strong
 evidence
 suggests
 that
 many
 of
 the
 pathways,
 including
 immune
 
response
 and
 protein
 translation,
 may
 be
 overlapping
 to
 form
 an
 interconnected
 network.
 
These
 pathways
 may
 be
 working
 together
 to
 contribute
 to
 the
 disease
 pathway
 for
 a
 subset
 of
 
ASD
 patients.
 
 

 
Chapter
 5:
 Summary
 and
 Future
 Directions
 

 
5.1
 Summary
 
Examining
 ncRNAs
 is
 a
 novel
 approach
 to
 understanding
 the
 molecular
 mechanisms
 involved
 in
 
autism.
 Even
 more
 novel
 is
 the
 idea
 that
 an
 antisense
 RNA
 could
 be
 having
 an
 impact
 beyond
 
gene
 silencing.
 Here,
 I
 have
 begun
 this
 work
 by
 focusing
 on
 an
 antisense
 lncRNA
 associated
 with
 
ASD.
 Through
 the
 use
 of
 RNA
 sequencing,
 I
 have
 revealed
 transcriptomic
 changes
 brought
 
about
 by
 dysregulation
 of
 MSNP1AS
 that
 provide
 insight
 into
 the
 etiology
 of
 ASD.
 I
 have
 
described
 the
 consequences
 of
 both
 upregulation
 of
 MSNP1AS
 (67),
 which
 is
 seen
 in
 brains
 of
 
autism
 patients,
 as
 well
 as
 the
 downregulation
 of
 the
 gene
 (69),
 which
 could
 be
 the
 answer
 to
 
amelioration
 of
 the
 disorder.
 Overexpression
 of
 MSNP1AS
 revealed
 changes
 in
 gene
 expression
 
in
 areas
 related
 to
 chromatin
 organization
 and
 protein
 synthesis,
 in
 addition
 to
 changes
 in
 cell
 
morphology.
 Stunted
 cell
 growth
 was
 demonstrated
 by
 an
 observed
 decrease
 in
 neurite
 length
 
and
 number.
 Knockdown
 of
 MSNP1AS
 corroborated
 and
 expanded
 the
 story
 by
 showing
 
changes
 in
 expression
 of
 genes
 not
 only
 involved
 in
 chromatin
 organization,
 but
 also
 immune
 
response.
 
 
94
 

 

  These
 changes
 are
 all
 observed
 in
 immortal
 neuronal
 cell
 lines,
 which,
 while
 relevant
 to
 
human
 health,
 cannot
 be
 said
 to
 exactly
 mirror
 what
 is
 observed
 in
 vivo.
 The
 CHARGE
 study
 
takes
 directly
 from
 the
 source,
 by
 using
 RNA
 derived
 from
 the
 blood
 samples
 of
 children.
 With
 
these
 data,
 I
 showed
 that
 pathways
 impacting
 chromatin
 disorganization,
 protein
 translation,
 
and
 immune
 response
 are
 also
 observed
 in
 ASD
 in
 vivo.
 
 

  While
 a
 specific
 pathway
 was
 not
 revealed
 by
 these
 studies,
 I
 have
 begun
 to
 narrow
 it
 
down
 by
 showing
 that
 the
 in
 vitro
 data
 accurately
 reflects
 the
 changes
 also
 observed
 in
 vivo.
 
Interestingly,
 changes
 in
 MSNP1AS
 are
 not
 seen
 in
 the
 CHARGE
 study
 data.
 This
 is
 not
 entirely
 
unexpected,
 as
 it
 is
 unknown
 at
 what
 point
 MSNP1AS
 becomes
 relevant
 in
 the
 development
 of
 
the
 autism
 brain.
 However,
 the
 gene
 ontology
 categories
 impacted
 by
 MSNP1AS
 dysregulation
 
are
 similar
 to
 the
 ones
 changed
 in
 the
 CHARGE
 study,
 giving
 a
 more
 solid
 direction
 to
 ASD
 
transcriptomic
 studies.
 
 
 

  Secondly,
 as
 was
 discussed
 previously
 in
 this
 dissertation,
 immune
 response,
 chromatin
 
organization,
 and
 protein
 translation
 cannot
 be
 solely
 responsible
 for
 the
 onset
 of
 ASD.
 It
 is
 
possible
 that
 the
 mechanism
 linking
 these
 pathways
 to
 sub-­‐phenotypes
 of
 ASD
 is
 further
 
removed
 from
 the
 initial
 dysregulation
 of
 MSNP1AS.
 It
 is
 also
 possible
 that
 the
 subset
 of
 
children
 with
 ASD
 that
 were
 a
 part
 of
 the
 CHARGE
 study
 are
 not
 in
 the
 same
 subset
 of
 people
 
with
 ASD
 that
 have
 the
 rs4307059
 SNP
 or
 dysregulated
 MSNP1AS.
 
 

 

   
 
95
 

 
5.2
 Overexpression
 of
 MSNP1AS
 leads
 to
 changes
 in
 MSN
 protein
 expression,
 protein
 
synthesis
 and
 chromatin
 organization.
 

  MSNP1AS
 is
 an
 antisense
 RNA.
 Its
 most
 anticipated
 function
 would
 be
 to
 silence
 the
 
element
 to
 which
 it
 is
 antisense.
 MSNP1
 is
 not
 transcribed,
 therefore,
 it
 makes
 sense
 to
 look
 to
 
MSN
 for
 evidence
 of
 MSNP1AS’s
 dysregulation.
 While
 an
 effect
 on
 MSN
 protein
 is
 observed,
 
changes
 in
 MSN
 transcript
 are
 not
 seen,
 indicating
 that
 MSNP1AS
 prevents
 MSN
 translation,
 but
 
not
 transcription.
 MSN
 protein
 is
 a
 part
 of
 the
 ezrin/radixin/moesin
 family,
 which
 is
 a
 family
 of
 
scaffolding
 proteins.
 The
 story
 could
 end
 there,
 but
 I
 found
 that
 there
 was
 more
 to
 it.
 MNSP1AS
 
does
 not
 affect
 MSN
 transcript
 levels,
 but
 it
 does
 bring
 about
 changes
 in
 expression
 of
 many
 
other
 transcripts
 in
 a
 much
 more
 dramatic
 way.
 
 
 
These
 transcripts
 are
 involved
 in
 a
 variety
 of
 processes,
 but
 most
 notably,
 I
 saw
 changes
 in
 
chromatin
 organization
 and
 protein
 synthesis.
 These
 processes,
 while
 not
 pointing
 to
 a
 highly
 
specific
 pathway,
 have
 been
 implicated
 in
 the
 pathogenesis
 of
 ASD.
 In
 addition,
 the
 expression
 
of
 many
 ncRNAs
 is
 affected
 by
 the
 overexpression
 of
 MSNP1AS.
 This
 sets
 the
 stage
 for
 the
 
establishment
 of
 ncRNA
 involvement
 in
 ASD
 as
 well
 as
 a
 more
 broad
 evaluation
 of
 noncoding
 
RNA
 as
 regulatory
 elements
 in
 general.
 From
 a
 more
 basic
 science
 view,
 we
 can
 learn
 how
 
ncRNA
 are
 contributing
 to
 cell
 growth,
 as
 well
 as
 how
 small
 changes
 in
 those
 processes
 
influence
 the
 progression
 of
 neurodevelopmental
 disorders
 like
 ASD.
 Through
 this,
 we
 can
 
move
 away
 from
 searching
 for
 specific
 genes
 that
 have
 a
 large
 impact,
 to
 regulatory
 elements
 
that
 can
 have
 an
 additive
 effect
 to
 stimulate
 disorder
 development.
96
 

 
5.3
 Knockdown
 of
 MSNP1AS
 reveals
 changes
 in
 genes
 involved
 in
 chromatin
 
organization
 and
 immune
 response
 

  The
 overexpression
 of
 MSNP1AS
 in
 my
 first
 paper
 mimicked
 what
 is
 seen
 in
 
autism
 brains.
 To
 amend
 this
 response,
 I
 knocked
 down
 MSNP1AS
 in
 neuronal
 cell
 lines.
 
Through
 the
 development
 of
 an
 sasRNA
 developed
 by
 Dr.
 Kevin
 Morris,
 I
 was
 able
 to
 
knockdown
 the
 antisense
 RNA,
 MSNP1AS.
 This
 alone
 was
 a
 challenge
 as
 the
 
conventional
 method
 of
 knocking
 down
 a
 gene
 involves
 using
 the
 antisense
 of
 that
 
gene.
 Since
 MSNP1AS
 is
 94%
 identical
 to
 the
 protein-­‐coding
 gene,
 MSN,
 I
 had
 to
 use
 a
 
different
 approach.
 The
 sasRNAs
 designed
 by
 Dr.
 Morris’s
 lab
 targeted
 the
 promoter
 
sequence
 of
 MSNP1AS.
 Using
 this
 method,
 I
 was
 able
 to
 successfully
 knockdown
 
MSNP1AS
 by
 about
 45%.
 

  Upon
 knockdown
 and
 subsequent
 RNA
 sequencing,
 many
 changes
 in
 gene
 
expression
 were
 observed.
 These
 changes
 validated
 the
 results
 seen
 after
 the
 
overexpression.
 Genes
 changed
 upon
 overexpression
 of
 MSNP1AS
 were
 expected
 to
 
have
 an
 inverse
 response,
 meaning
 that
 genes
 that
 were
 upregulated
 upon
 
overexpression
 of
 MSNP1AS
 would
 be
 downregulated
 upon
 MSNP1AS
 knockdown.
 
Many
 genes
 confirmed
 this
 inverse
 response.
 The
 genes
 changed
 in
 expression
 also
 had
 
ontology
 related
 to
 chromatin
 organization.
 In
 addition
 to
 affirming
 the
 overexpression
 
results,
 the
 knockdown
 revealed
 another
 pathway
 affected
 by
 dysregulation.
 This
 
lncRNA
 appears
 to
 regulate
 the
 immune
 response.
 These
 gene
 ontology
 categories
 are
 
in
 line
 with
 what
 is
 observed
 in
 ASD.
 Reports
 of
 chromatin
 organization
 at
 the
 molecular
 
level
 in
 ASD
 are
 a
 continuing
 point
 of
 interest
 (211).
 A
 subset
 of
 ASD
 patients
 
97
 

 
experiences
 dysregulation
 of
 immune
 response
 such
 as
 gastrointestinal
 upset
 and
 
autoimmune
 disorders.
 These
 pathways
 may
 be
 working
 together
 to
 contribute
 to
 the
 
disease
 pathway
 for
 a
 subset
 of
 ASD
 patients.
 
 

 
5.4
 Immune
 response
 is
 observed
 in
 vivo.
 

  While
 it
 is
 impossible
 to
 completely
 recapitulate
 complex
 and
 heterogeneous
 
disorders
 like
 ASD
 in
 cell
 models,
 these
 models
 give
 us
 insight
 into
 what
 might
 be
 
occurring
 in
 humans.
 When
 possible,
 it
 is
 always
 better
 to
 confirm
 these
 insights
 in
 the
 
subject
 of
 interest,
 rather
 than
 a
 model.
 The
 CHARGE
 study
 gave
 me
 the
 opportunity
 to
 
make
 these
 confirmations.
 One
 alternative
 explanation
 of
 immune
 response
 pathways
 
being
 present
 in
 cell
 models
 is
 that
 transfection
 can
 sometimes
 trigger
 an
 immune
 
response.
 When
 data
 is
 derived
 from
 the
 blood
 samples
 of
 children,
 such
 explanations
 
are
 no
 longer
 valid.
 The
 CHARGE
 study
 data
 suggests
 that
 an
 immune
 response
 is
 truly
 
present
 in
 autism
 patients
 compared
 to
 controls,
 in
 which
 it
 is
 not
 found.
 These
 results
 
validate
 the
 cellular
 model
 studies.
 
 

  The
 CHARGE
 study
 also
 points
 to
 an
 environmental
 trigger.
 TD
 children
 exposed
 
to
 pesticides
 also
 experience
 changes
 in
 immune
 system
 gene
 pathways.
 While
 
MSNP1AS
 did
 not
 appear
 to
 be
 a
 key
 player
 in
 the
 CHARGE
 data,
 both
 studies
 overlap
 in
 
ways
 that
 contribute
 to
 defining
 a
 more
 concrete
 pathway
 for
 autism
 etiology.
 
 

 

   
 
98
 

 
5.5
 Future
 Directions
 
These
 data
 linked
 MSNP1AS
 dysregulation
 to
 chromatin
 regulation,
 protein
 
synthesis,
 immune
 response,
 and
 the
 regulation
 of
 moesin.
 The
 role
 of
 these
 pathways
 
in
 ASD
 etiology
 is
 supported
 by
 the
 data
 resulting
 from
 the
 CHARGE
 study,
 as
 does
 
accumulating
 evidence
 points
 to
 those
 pathways
 (198,
 199).
 While
 the
 evidence
 points
 
to
 MSNP1AS
 involvement,
 the
 specific
 molecular
 mechanism,
 beyond
 binding
 of
 moesin
 
protein,
 has
 yet
 to
 be
 elucidated.
 The
 transcript
 levels
 of
 MSN
 are
 not
 effected
 by
 
MSNP1AS
 dyregulation
 and
 do
 not
 appear
 to
 be
 changed
 in
 the
 blood
 samples
 from
 
children,
 leading
 to
 the
 conclusion
 that
 these
 immune
 system
 genes
 are
 operating
 in
 a
 
pathway
 separate
 from
 the
 MSN/MSNP1AS
 pathway.
 
 
Additional
 experiments
 will
 be
 needed
 to
 evaluate
 the
 levels
 of
 MSNP1AS
 and
 
moesin
 protein
 in
 neurons
 derived
 from
 patients
 with
 the
 ASD-­‐associated
 rs4307059
 
allele.
 Another
 direction
 would
 be
 to
 determine
 MSNP1AS
 contribution
 to
 altered
 brain
 
development
 using
 neurons
 derived
 from
 Cri-­‐du-­‐chat
 patients
 with
 deletion
 of
 
chromosome
 5p14.1.
 Patients
 with
 ASD-­‐associated
 rs4307059
 have
 not
 been
 
phenotyped.
 It
 will
 be
 important
 to
 determine
 if
 these
 patients
 have
 features
 in
 
common
 with
 ASD
 patients
 such
 as
 gastrointestinal
 distress
 or
 autoimmune
 disorders.
 
 

 

   
 
99
 

 
References
 

 
1.
  DeWitt
 J,
 Campbell,
 DB.
 Targeting
 Noncoding
 RNA
 for
 Treatment
 of
 Autism
 
Spectrum
 Disorders.
 In:
 Hu
 V,
 editor.
 Frontiers
 in
 Autism
 Research,
 Diagnosis,
 and
 
Treatment.
 Singapore:
 World
 Scientific
 Publishing
 Co.
 Pte.
 Ltd;
 2014.
 p.
 203-­‐28.
 
2.
  Battle
 DE.
 Diagnostic
 and
 Statistical
 Manual
 of
 Mental
 Disorders
 (DSM).
 
Codas.
 2013;25(2):191-­‐2.
 PubMed
 PMID:
 24413388.
 
3.
  Rutter
 M.
 Incidence
 of
 autism
 spectrum
 disorders:
 changes
 over
 time
 and
 
their
 meaning.
 Acta
 Paediatr.
 2005;94(1):2-­‐15.
 PubMed
 PMID:
 15858952.
 
4.
  CDC.
 Prevalence
 of
 autism
 spectrum
 disorder
 among
 children
 aged
 8
 years–
autism
 and
 developmental
 disabilities
 monitoring
 network,
 11
 sites,
 United
 States,
 
2010.
 .
 2014.
 
5.
  Fombonne
 E.
 Epidemiology
 of
 pervasive
 developmental
 disorders.
 Pediatr
 
Res.
 2009;65(6):591-­‐8.
 doi:
 10.1203/PDR.0b013e31819e7203.
 PubMed
 PMID:
 
19218885.
 
6.
  Elsabbagh
 M,
 Divan
 G,
 Koh
 YJ,
 Kim
 YS,
 Kauchali
 S,
 Marcin
 C,
 Montiel-­‐Nava
 C,
 
Patel
 V,
 Paula
 CS,
 Wang
 C,
 Yasamy
 MT,
 Fombonne
 E.
 Global
 prevalence
 of
 autism
 and
 
other
 pervasive
 developmental
 disorders.
 Autism
 Res.
 2012;5(3):160-­‐79.
 doi:
 
10.1002/aur.239.
 PubMed
 PMID:
 22495912;
 PMCID:
 PMC3763210.
 
7.
  Hastings
 RP,
 Kovshoff
 H,
 Ward
 NJ,
 degli
 Espinosa
 F,
 Brown
 T,
 Remington
 B.
 
Systems
 analysis
 of
 stress
 and
 positive
 perceptions
 in
 mothers
 and
 fathers
 of
 pre-­‐
school
 children
 with
 autism.
 J
 Autism
 Dev
 Disord.
 2005;35(5):635-­‐44.
 doi:
 
10.1007/s10803-­‐005-­‐0007-­‐8.
 PubMed
 PMID:
 16177837.
 
8.
  Klein
 N,
 Kemper
 KJ.
 Integrative
 Approaches
 to
 Caring
 for
 Children
 with
 
Autism.
 Curr
 Probl
 Pediatr
 Adolesc
 Health
 Care.
 2016;46(6):195-­‐201.
 doi:
 
10.1016/j.cppeds.2015.12.004.
 PubMed
 PMID:
 26776326.
 
9.
  Choueiri
 RN,
 Zimmerman
 AW.
 New
 Assessments
 and
 Treatments
 in
 ASD.
 
Curr
 Treat
 Options
 Neurol.
 2017;19(2):6.
 doi:
 10.1007/s11940-­‐017-­‐0443-­‐8.
 
PubMed
 PMID:
 28251588.
 
10.
  Benevides
 TW,
 Carretta
 HJ,
 Lane
 SJ.
 Unmet
 Need
 for
 Therapy
 Among
 Children
 
with
 Autism
 Spectrum
 Disorder:
 Results
 from
 the
 2005-­‐2006
 and
 2009-­‐2010
 
National
 Survey
 of
 Children
 with
 Special
 Health
 Care
 Needs.
 Matern
 Child
 Health
 J.
 
2016;20(4):878-­‐88.
 doi:
 10.1007/s10995-­‐015-­‐1876-­‐x.
 PubMed
 PMID:
 26662279.
 
11.
  Elder
 JH,
 Brasher
 S,
 Alexander
 B.
 Identifying
 the
 Barriers
 to
 Early
 Diagnosis
 
and
 Treatment
 in
 Underserved
 Individuals
 with
 Autism
 Spectrum
 Disorders
 (ASD)
 
and
 Their
 Families:
 A
 Qualitative
 Study.
 Issues
 Ment
 Health
 Nurs.
 2016;37(6):412-­‐
20.
 doi:
 10.3109/01612840.2016.1153174.
 PubMed
 PMID:
 27070190.
 
12.
  Lavelle
 TA,
 Weinstein
 MC,
 Newhouse
 JP,
 Munir
 K,
 Kuhlthau
 KA,
 Prosser
 LA.
 
Economic
 burden
 of
 childhood
 autism
 spectrum
 disorders.
 Pediatrics.
 
2014;133(3):e520-­‐9.
 doi:
 10.1542/peds.2013-­‐0763.
 PubMed
 PMID:
 24515505.
 
13.
  Ganz
 ML.
 The
 lifetime
 distribution
 of
 the
 incremental
 societal
 costs
 of
 autism.
 
Arch
 Pediatr
 Adolesc
 Med.
 2007;161(4):343-­‐9.
 doi:
 10.1001/archpedi.161.4.343.
 
PubMed
 PMID:
 17404130.
 
100
 

 
14.
  MacDonald
 R,
 Parry-­‐Cruwys
 D,
 Dupere
 S,
 Ahearn
 W.
 Assessing
 progress
 and
 
outcome
 of
 early
 intensive
 behavioral
 intervention
 for
 toddlers
 with
 autism.
 Res
 
Dev
 Disabil.
 2014;35(12):3632-­‐44.
 doi:
 10.1016/j.ridd.2014.08.036.
 PubMed
 PMID:
 
25241118.
 
15.
  Steinhausen
 HC,
 Mohr
 Jensen
 C,
 Lauritsen
 MB.
 A
 systematic
 review
 and
 meta-­‐
analysis
 of
 the
 long-­‐term
 overall
 outcome
 of
 autism
 spectrum
 disorders
 in
 
adolescence
 and
 adulthood.
 Acta
 Psychiatr
 Scand.
 2016;133(6):445-­‐52.
 doi:
 
10.1111/acps.12559.
 PubMed
 PMID:
 26763353.
 
16.
  Eikeseth
 S,
 Smith
 T,
 Jahr
 E,
 Eldevik
 S.
 Intensive
 behavioral
 treatment
 at
 
school
 for
 4-­‐
 to
 7-­‐year-­‐old
 children
 with
 autism.
 A
 1-­‐year
 comparison
 controlled
 
study.
 Behav
 Modif.
 2002;26(1):49-­‐68.
 doi:
 10.1177/0145445502026001004.
 
PubMed
 PMID:
 11799654.
 
17.
  Goldstein
 H.
 Communication
 intervention
 for
 children
 with
 autism:
 a
 review
 
of
 treatment
 efficacy.
 J
 Autism
 Dev
 Disord.
 2002;32(5):373-­‐96.
 PubMed
 PMID:
 
12463516.
 
18.
  Case-­‐Smith
 J,
 Arbesman
 M.
 Evidence-­‐based
 review
 of
 interventions
 for
 
autism
 used
 in
 or
 of
 relevance
 to
 occupational
 therapy.
 Am
 J
 Occup
 Ther.
 
2008;62(4):416-­‐29.
 PubMed
 PMID:
 18712004.
 
19.
  Bitterman
 A,
 Daley
 TC,
 Misra
 S,
 Carlson
 E,
 Markowitz
 J.
 A
 national
 sample
 of
 
preschoolers
 with
 autism
 spectrum
 disorders:
 special
 education
 services
 and
 parent
 
satisfaction.
 J
 Autism
 Dev
 Disord.
 2008;38(8):1509-­‐17.
 doi:
 10.1007/s10803-­‐007-­‐
0531-­‐9.
 PubMed
 PMID:
 18228122.
 
20.
  Boucher
 J.
 Research
 review:
 structural
 language
 in
 autistic
 spectrum
 disorder
 
-­‐
 characteristics
 and
 causes.
 J
 Child
 Psychol
 Psychiatry.
 2012;53(3):219-­‐33.
 doi:
 
10.1111/j.1469-­‐7610.2011.02508.x.
 PubMed
 PMID:
 22188468.
 
21.
  Folstein
 S,
 Rutter
 M.
 Genetic
 influences
 and
 infantile
 autism.
 Nature.
 
1977;265(5596):726-­‐8.
 PubMed
 PMID:
 558516.
 
22.
  Ritvo
 ER,
 Freeman
 BJ,
 Mason-­‐Brothers
 A,
 Mo
 A,
 Ritvo
 AM.
 Concordance
 for
 
the
 syndrome
 of
 autism
 in
 40
 pairs
 of
 afflicted
 twins.
 Am
 J
 Psychiatry.
 
1985;142(1):74-­‐7.
 doi:
 10.1176/ajp.142.1.74.
 PubMed
 PMID:
 4038442.
 
23.
  Bailey
 A,
 Le
 Couteur
 A,
 Gottesman
 I,
 Bolton
 P,
 Simonoff
 E,
 Yuzda
 E,
 Rutter
 M.
 
Autism
 as
 a
 strongly
 genetic
 disorder:
 evidence
 from
 a
 British
 twin
 study.
 Psychol
 
Med.
 1995;25(1):63-­‐77.
 PubMed
 PMID:
 7792363.
 
24.
  Autism
 Genome
 Project
 C,
 Szatmari
 P,
 Paterson
 AD,
 Zwaigenbaum
 L,
 Roberts
 
W,
 Brian
 J,
 Liu
 XQ,
 Vincent
 JB,
 Skaug
 JL,
 Thompson
 AP,
 Senman
 L,
 Feuk
 L,
 Qian
 C,
 
Bryson
 SE,
 Jones
 MB,
 Marshall
 CR,
 Scherer
 SW,
 Vieland
 VJ,
 Bartlett
 C,
 Mangin
 LV,
 
Goedken
 R,
 Segre
 A,
 Pericak-­‐Vance
 MA,
 Cuccaro
 ML,
 Gilbert
 JR,
 Wright
 HH,
 
Abramson
 RK,
 Betancur
 C,
 Bourgeron
 T,
 Gillberg
 C,
 Leboyer
 M,
 Buxbaum
 JD,
 Davis
 
KL,
 Hollander
 E,
 Silverman
 JM,
 Hallmayer
 J,
 Lotspeich
 L,
 Sutcliffe
 JS,
 Haines
 JL,
 
Folstein
 SE,
 Piven
 J,
 Wassink
 TH,
 Sheffield
 V,
 Geschwind
 DH,
 Bucan
 M,
 Brown
 WT,
 
Cantor
 RM,
 Constantino
 JN,
 Gilliam
 TC,
 Herbert
 M,
 Lajonchere
 C,
 Ledbetter
 DH,
 Lese-­‐
Martin
 C,
 Miller
 J,
 Nelson
 S,
 Samango-­‐Sprouse
 CA,
 Spence
 S,
 State
 M,
 Tanzi
 RE,
 Coon
 
H,
 Dawson
 G,
 Devlin
 B,
 Estes
 A,
 Flodman
 P,
 Klei
 L,
 McMahon
 WM,
 Minshew
 N,
 
Munson
 J,
 Korvatska
 E,
 Rodier
 PM,
 Schellenberg
 GD,
 Smith
 M,
 Spence
 MA,
 Stodgell
 C,
 
Tepper
 PG,
 Wijsman
 EM,
 Yu
 CE,
 Roge
 B,
 Mantoulan
 C,
 Wittemeyer
 K,
 Poustka
 A,
 
Felder
 B,
 Klauck
 SM,
 Schuster
 C,
 Poustka
 F,
 Bolte
 S,
 Feineis-­‐Matthews
 S,
 Herbrecht
 E,
 
101
 

 
Schmotzer
 G,
 Tsiantis
 J,
 Papanikolaou
 K,
 Maestrini
 E,
 Bacchelli
 E,
 Blasi
 F,
 Carone
 S,
 
Toma
 C,
 Van
 Engeland
 H,
 de
 Jonge
 M,
 Kemner
 C,
 Koop
 F,
 Langemeijer
 M,
 Hijmans
 C,
 
Staal
 WG,
 Baird
 G,
 Bolton
 PF,
 Rutter
 ML,
 Weisblatt
 E,
 Green
 J,
 Aldred
 C,
 Wilkinson
 JA,
 
Pickles
 A,
 Le
 Couteur
 A,
 Berney
 T,
 McConachie
 H,
 Bailey
 AJ,
 Francis
 K,
 Honeyman
 G,
 
Hutchinson
 A,
 Parr
 JR,
 Wallace
 S,
 Monaco
 AP,
 Barnby
 G,
 Kobayashi
 K,
 Lamb
 JA,
 Sousa
 
I,
 Sykes
 N,
 Cook
 EH,
 Guter
 SJ,
 Leventhal
 BL,
 Salt
 J,
 Lord
 C,
 Corsello
 C,
 Hus
 V,
 Weeks
 
DE,
 Volkmar
 F,
 Tauber
 M,
 Fombonne
 E,
 Shih
 A,
 Meyer
 KJ.
 Mapping
 autism
 risk
 loci
 
using
 genetic
 linkage
 and
 chromosomal
 rearrangements.
 Nat
 Genet.
 
2007;39(3):319-­‐28.
 doi:
 10.1038/ng1985.
 PubMed
 PMID:
 17322880;
 PMCID:
 
PMC4867008.
 
25.
  Weiss
 LA,
 Arking
 DE,
 Gene
 Discovery
 Project
 of
 Johns
 H,
 the
 Autism
 C,
 Daly
 
MJ,
 Chakravarti
 A.
 A
 genome-­‐wide
 linkage
 and
 association
 scan
 reveals
 novel
 loci
 for
 
autism.
 Nature.
 2009;461(7265):802-­‐8.
 doi:
 10.1038/nature08490.
 PubMed
 PMID:
 
19812673;
 PMCID:
 PMC2772655.
 
26.
  Wang
 K,
 Zhang
 H,
 Ma
 D,
 Bucan
 M,
 Glessner
 JT,
 Abrahams
 BS,
 Salyakina
 D,
 
Imielinski
 M,
 Bradfield
 JP,
 Sleiman
 PM,
 Kim
 CE,
 Hou
 C,
 Frackelton
 E,
 Chiavacci
 R,
 
Takahashi
 N,
 Sakurai
 T,
 Rappaport
 E,
 Lajonchere
 CM,
 Munson
 J,
 Estes
 A,
 Korvatska
 
O,
 Piven
 J,
 Sonnenblick
 LI,
 Alvarez
 Retuerto
 AI,
 Herman
 EI,
 Dong
 H,
 Hutman
 T,
 
Sigman
 M,
 Ozonoff
 S,
 Klin
 A,
 Owley
 T,
 Sweeney
 JA,
 Brune
 CW,
 Cantor
 RM,
 Bernier
 R,
 
Gilbert
 JR,
 Cuccaro
 ML,
 McMahon
 WM,
 Miller
 J,
 State
 MW,
 Wassink
 TH,
 Coon
 H,
 Levy
 
SE,
 Schultz
 RT,
 Nurnberger
 JI,
 Haines
 JL,
 Sutcliffe
 JS,
 Cook
 EH,
 Minshew
 NJ,
 Buxbaum
 
JD,
 Dawson
 G,
 Grant
 SF,
 Geschwind
 DH,
 Pericak-­‐Vance
 MA,
 Schellenberg
 GD,
 
Hakonarson
 H.
 Common
 genetic
 variants
 on
 5p14.1
 associate
 with
 autism
 spectrum
 
disorders.
 Nature.
 2009;459(7246):528-­‐33.
 doi:
 10.1038/nature07999.
 PubMed
 
PMID:
 19404256;
 PMCID:
 PMC2943511.
 
27.
  Ma
 D,
 Salyakina
 D,
 Jaworski
 JM,
 Konidari
 I,
 Whitehead
 PL,
 Andersen
 AN,
 
Hoffman
 JD,
 Slifer
 SH,
 Hedges
 DJ,
 Cukier
 HN,
 Griswold
 AJ,
 McCauley
 JL,
 Beecham
 GW,
 
Wright
 HH,
 Abramson
 RK,
 Martin
 ER,
 Hussman
 JP,
 Gilbert
 JR,
 Cuccaro
 ML,
 Haines
 JL,
 
Pericak-­‐Vance
 MA.
 A
 genome-­‐wide
 association
 study
 of
 autism
 reveals
 a
 common
 
novel
 risk
 locus
 at
 5p14.1.
 Ann
 Hum
 Genet.
 2009;73(Pt
 3):263-­‐73.
 doi:
 
10.1111/j.1469-­‐1809.2009.00523.x.
 PubMed
 PMID:
 19456320;
 PMCID:
 
PMC2918410.
 
28.
  Anney
 R,
 Klei
 L,
 Pinto
 D,
 Regan
 R,
 Conroy
 J,
 Magalhaes
 TR,
 Correia
 C,
 
Abrahams
 BS,
 Sykes
 N,
 Pagnamenta
 AT,
 Almeida
 J,
 Bacchelli
 E,
 Bailey
 AJ,
 Baird
 G,
 
Battaglia
 A,
 Berney
 T,
 Bolshakova
 N,
 Bolte
 S,
 Bolton
 PF,
 Bourgeron
 T,
 Brennan
 S,
 
Brian
 J,
 Carson
 AR,
 Casallo
 G,
 Casey
 J,
 Chu
 SH,
 Cochrane
 L,
 Corsello
 C,
 Crawford
 EL,
 
Crossett
 A,
 Dawson
 G,
 de
 Jonge
 M,
 Delorme
 R,
 Drmic
 I,
 Duketis
 E,
 Duque
 F,
 Estes
 A,
 
Farrar
 P,
 Fernandez
 BA,
 Folstein
 SE,
 Fombonne
 E,
 Freitag
 CM,
 Gilbert
 J,
 Gillberg
 C,
 
Glessner
 JT,
 Goldberg
 J,
 Green
 J,
 Guter
 SJ,
 Hakonarson
 H,
 Heron
 EA,
 Hill
 M,
 Holt
 R,
 
Howe
 JL,
 Hughes
 G,
 Hus
 V,
 Igliozzi
 R,
 Kim
 C,
 Klauck
 SM,
 Kolevzon
 A,
 Korvatska
 O,
 
Kustanovich
 V,
 Lajonchere
 CM,
 Lamb
 JA,
 Laskawiec
 M,
 Leboyer
 M,
 Le
 Couteur
 A,
 
Leventhal
 BL,
 Lionel
 AC,
 Liu
 XQ,
 Lord
 C,
 Lotspeich
 L,
 Lund
 SC,
 Maestrini
 E,
 Mahoney
 
W,
 Mantoulan
 C,
 Marshall
 CR,
 McConachie
 H,
 McDougle
 CJ,
 McGrath
 J,
 McMahon
 WM,
 
Melhem
 NM,
 Merikangas
 A,
 Migita
 O,
 Minshew
 NJ,
 Mirza
 GK,
 Munson
 J,
 Nelson
 SF,
 
Noakes
 C,
 Noor
 A,
 Nygren
 G,
 Oliveira
 G,
 Papanikolaou
 K,
 Parr
 JR,
 Parrini
 B,
 Paton
 T,
 
Pickles
 A,
 Piven
 J,
 Posey
 DJ,
 Poustka
 A,
 Poustka
 F,
 Prasad
 A,
 Ragoussis
 J,
 Renshaw
 K,
 
102
 

 
Rickaby
 J,
 Roberts
 W,
 Roeder
 K,
 Roge
 B,
 Rutter
 ML,
 Bierut
 LJ,
 Rice
 JP,
 Salt
 J,
 Sansom
 
K,
 Sato
 D,
 Segurado
 R,
 Senman
 L,
 Shah
 N,
 Sheffield
 VC,
 Soorya
 L,
 Sousa
 I,
 Stoppioni
 V,
 
Strawbridge
 C,
 Tancredi
 R,
 Tansey
 K,
 Thiruvahindrapduram
 B,
 Thompson
 AP,
 
Thomson
 S,
 Tryfon
 A,
 Tsiantis
 J,
 Van
 Engeland
 H,
 Vincent
 JB,
 Volkmar
 F,
 Wallace
 S,
 
Wang
 K,
 Wang
 Z,
 Wassink
 TH,
 Wing
 K,
 Wittemeyer
 K,
 Wood
 S,
 Yaspan
 BL,
 
Zurawiecki
 D,
 Zwaigenbaum
 L,
 Betancur
 C,
 Buxbaum
 JD,
 Cantor
 RM,
 Cook
 EH,
 Coon
 
H,
 Cuccaro
 ML,
 Gallagher
 L,
 Geschwind
 DH,
 Gill
 M,
 Haines
 JL,
 Miller
 J,
 Monaco
 AP,
 
Nurnberger
 JI,
 Jr.,
 Paterson
 AD,
 Pericak-­‐Vance
 MA,
 Schellenberg
 GD,
 Scherer
 SW,
 
Sutcliffe
 JS,
 Szatmari
 P,
 Vicente
 AM,
 Vieland
 VJ,
 Wijsman
 EM,
 Devlin
 B,
 Ennis
 S,
 
Hallmayer
 J.
 A
 genome-­‐wide
 scan
 for
 common
 alleles
 affecting
 risk
 for
 autism.
 Hum
 
Mol
 Genet.
 2010;19(20):4072-­‐82.
 doi:
 10.1093/hmg/ddq307.
 PubMed
 PMID:
 
20663923;
 PMCID:
 PMC2947401.
 
29.
  Chahrour
 MH,
 Yu
 TW,
 Lim
 ET,
 Ataman
 B,
 Coulter
 ME,
 Hill
 RS,
 Stevens
 CR,
 
Schubert
 CR,
 Collaboration
 AAS,
 Greenberg
 ME,
 Gabriel
 SB,
 Walsh
 CA.
 Whole-­‐exome
 
sequencing
 and
 homozygosity
 analysis
 implicate
 depolarization-­‐regulated
 neuronal
 
genes
 in
 autism.
 PLoS
 Genet.
 2012;8(4):e1002635.
 doi:
 
10.1371/journal.pgen.1002635.
 PubMed
 PMID:
 22511880;
 PMCID:
 PMC3325173.
 
30.
  Iossifov
 I,
 Ronemus
 M,
 Levy
 D,
 Wang
 Z,
 Hakker
 I,
 Rosenbaum
 J,
 Yamrom
 B,
 
Lee
 YH,
 Narzisi
 G,
 Leotta
 A,
 Kendall
 J,
 Grabowska
 E,
 Ma
 B,
 Marks
 S,
 Rodgers
 L,
 
Stepansky
 A,
 Troge
 J,
 Andrews
 P,
 Bekritsky
 M,
 Pradhan
 K,
 Ghiban
 E,
 Kramer
 M,
 Parla
 
J,
 Demeter
 R,
 Fulton
 LL,
 Fulton
 RS,
 Magrini
 VJ,
 Ye
 K,
 Darnell
 JC,
 Darnell
 RB,
 Mardis
 
ER,
 Wilson
 RK,
 Schatz
 MC,
 McCombie
 WR,
 Wigler
 M.
 De
 novo
 gene
 disruptions
 in
 
children
 on
 the
 autistic
 spectrum.
 Neuron.
 2012;74(2):285-­‐99.
 doi:
 
10.1016/j.neuron.2012.04.009.
 PubMed
 PMID:
 22542183;
 PMCID:
 PMC3619976.
 
31.
  O'Roak
 BJ,
 Deriziotis
 P,
 Lee
 C,
 Vives
 L,
 Schwartz
 JJ,
 Girirajan
 S,
 Karakoc
 E,
 
Mackenzie
 AP,
 Ng
 SB,
 Baker
 C,
 Rieder
 MJ,
 Nickerson
 DA,
 Bernier
 R,
 Fisher
 SE,
 
Shendure
 J,
 Eichler
 EE.
 Exome
 sequencing
 in
 sporadic
 autism
 spectrum
 disorders
 
identifies
 severe
 de
 novo
 mutations.
 Nat
 Genet.
 2011;43(6):585-­‐9.
 doi:
 
10.1038/ng.835.
 PubMed
 PMID:
 21572417;
 PMCID:
 PMC3115696.
 
32.
  O'Roak
 BJ,
 Vives
 L,
 Girirajan
 S,
 Karakoc
 E,
 Krumm
 N,
 Coe
 BP,
 Levy
 R,
 Ko
 A,
 
Lee
 C,
 Smith
 JD,
 Turner
 EH,
 Stanaway
 IB,
 Vernot
 B,
 Malig
 M,
 Baker
 C,
 Reilly
 B,
 Akey
 
JM,
 Borenstein
 E,
 Rieder
 MJ,
 Nickerson
 DA,
 Bernier
 R,
 Shendure
 J,
 Eichler
 EE.
 
Sporadic
 autism
 exomes
 reveal
 a
 highly
 interconnected
 protein
 network
 of
 de
 novo
 
mutations.
 Nature.
 2012;485(7397):246-­‐50.
 doi:
 10.1038/nature10989.
 PubMed
 
PMID:
 22495309;
 PMCID:
 PMC3350576.
 
33.
  Sanders
 SJ,
 Murtha
 MT,
 Gupta
 AR,
 Murdoch
 JD,
 Raubeson
 MJ,
 Willsey
 AJ,
 
Ercan-­‐Sencicek
 AG,
 DiLullo
 NM,
 Parikshak
 NN,
 Stein
 JL,
 Walker
 MF,
 Ober
 GT,
 Teran
 
NA,
 Song
 Y,
 El-­‐Fishawy
 P,
 Murtha
 RC,
 Choi
 M,
 Overton
 JD,
 Bjornson
 RD,
 Carriero
 NJ,
 
Meyer
 KA,
 Bilguvar
 K,
 Mane
 SM,
 Sestan
 N,
 Lifton
 RP,
 Gunel
 M,
 Roeder
 K,
 Geschwind
 
DH,
 Devlin
 B,
 State
 MW.
 De
 novo
 mutations
 revealed
 by
 whole-­‐exome
 sequencing
 
are
 strongly
 associated
 with
 autism.
 Nature.
 2012;485(7397):237-­‐41.
 doi:
 
10.1038/nature10945.
 PubMed
 PMID:
 22495306;
 PMCID:
 PMC3667984.
 
34.
  Neale
 BM,
 Kou
 Y,
 Liu
 L,
 Ma'ayan
 A,
 Samocha
 KE,
 Sabo
 A,
 Lin
 CF,
 Stevens
 C,
 
Wang
 LS,
 Makarov
 V,
 Polak
 P,
 Yoon
 S,
 Maguire
 J,
 Crawford
 EL,
 Campbell
 NG,
 Geller
 
ET,
 Valladares
 O,
 Schafer
 C,
 Liu
 H,
 Zhao
 T,
 Cai
 G,
 Lihm
 J,
 Dannenfelser
 R,
 Jabado
 O,
 
Peralta
 Z,
 Nagaswamy
 U,
 Muzny
 D,
 Reid
 JG,
 Newsham
 I,
 Wu
 Y,
 Lewis
 L,
 Han
 Y,
 Voight
 
103
 

 
BF,
 Lim
 E,
 Rossin
 E,
 Kirby
 A,
 Flannick
 J,
 Fromer
 M,
 Shakir
 K,
 Fennell
 T,
 Garimella
 K,
 
Banks
 E,
 Poplin
 R,
 Gabriel
 S,
 DePristo
 M,
 Wimbish
 JR,
 Boone
 BE,
 Levy
 SE,
 Betancur
 C,
 
Sunyaev
 S,
 Boerwinkle
 E,
 Buxbaum
 JD,
 Cook
 EH,
 Jr.,
 Devlin
 B,
 Gibbs
 RA,
 Roeder
 K,
 
Schellenberg
 GD,
 Sutcliffe
 JS,
 Daly
 MJ.
 Patterns
 and
 rates
 of
 exonic
 de
 novo
 
mutations
 in
 autism
 spectrum
 disorders.
 Nature.
 2012;485(7397):242-­‐5.
 doi:
 
10.1038/nature11011.
 PubMed
 PMID:
 22495311;
 PMCID:
 PMC3613847.
 
35.
  Glessner
 JT,
 Wang
 K,
 Cai
 G,
 Korvatska
 O,
 Kim
 CE,
 Wood
 S,
 Zhang
 H,
 Estes
 A,
 
Brune
 CW,
 Bradfield
 JP,
 Imielinski
 M,
 Frackelton
 EC,
 Reichert
 J,
 Crawford
 EL,
 
Munson
 J,
 Sleiman
 PM,
 Chiavacci
 R,
 Annaiah
 K,
 Thomas
 K,
 Hou
 C,
 Glaberson
 W,
 Flory
 
J,
 Otieno
 F,
 Garris
 M,
 Soorya
 L,
 Klei
 L,
 Piven
 J,
 Meyer
 KJ,
 Anagnostou
 E,
 Sakurai
 T,
 
Game
 RM,
 Rudd
 DS,
 Zurawiecki
 D,
 McDougle
 CJ,
 Davis
 LK,
 Miller
 J,
 Posey
 DJ,
 Michaels
 
S,
 Kolevzon
 A,
 Silverman
 JM,
 Bernier
 R,
 Levy
 SE,
 Schultz
 RT,
 Dawson
 G,
 Owley
 T,
 
McMahon
 WM,
 Wassink
 TH,
 Sweeney
 JA,
 Nurnberger
 JI,
 Coon
 H,
 Sutcliffe
 JS,
 
Minshew
 NJ,
 Grant
 SF,
 Bucan
 M,
 Cook
 EH,
 Buxbaum
 JD,
 Devlin
 B,
 Schellenberg
 GD,
 
Hakonarson
 H.
 Autism
 genome-­‐wide
 copy
 number
 variation
 reveals
 ubiquitin
 and
 
neuronal
 genes.
 Nature.
 2009;459(7246):569-­‐73.
 doi:
 10.1038/nature07953.
 
PubMed
 PMID:
 19404257;
 PMCID:
 PMC2925224.
 
36.
  Weiss
 LA,
 Shen
 Y,
 Korn
 JM,
 Arking
 DE,
 Miller
 DT,
 Fossdal
 R,
 Saemundsen
 E,
 
Stefansson
 H,
 Ferreira
 MA,
 Green
 T,
 Platt
 OS,
 Ruderfer
 DM,
 Walsh
 CA,
 Altshuler
 D,
 
Chakravarti
 A,
 Tanzi
 RE,
 Stefansson
 K,
 Santangelo
 SL,
 Gusella
 JF,
 Sklar
 P,
 Wu
 BL,
 
Daly
 MJ,
 Autism
 C.
 Association
 between
 microdeletion
 and
 microduplication
 at
 
16p11.2
 and
 autism.
 N
 Engl
 J
 Med.
 2008;358(7):667-­‐75.
 doi:
 
10.1056/NEJMoa075974.
 PubMed
 PMID:
 18184952.
 
37.
  Bucan
 M,
 Abrahams
 BS,
 Wang
 K,
 Glessner
 JT,
 Herman
 EI,
 Sonnenblick
 LI,
 
Alvarez
 Retuerto
 AI,
 Imielinski
 M,
 Hadley
 D,
 Bradfield
 JP,
 Kim
 C,
 Gidaya
 NB,
 
Lindquist
 I,
 Hutman
 T,
 Sigman
 M,
 Kustanovich
 V,
 Lajonchere
 CM,
 Singleton
 A,
 Kim
 J,
 
Wassink
 TH,
 McMahon
 WM,
 Owley
 T,
 Sweeney
 JA,
 Coon
 H,
 Nurnberger
 JI,
 Li
 M,
 
Cantor
 RM,
 Minshew
 NJ,
 Sutcliffe
 JS,
 Cook
 EH,
 Dawson
 G,
 Buxbaum
 JD,
 Grant
 SF,
 
Schellenberg
 GD,
 Geschwind
 DH,
 Hakonarson
 H.
 Genome-­‐wide
 analyses
 of
 exonic
 
copy
 number
 variants
 in
 a
 family-­‐based
 study
 point
 to
 novel
 autism
 susceptibility
 
genes.
 PLoS
 Genet.
 2009;5(6):e1000536.
 doi:
 10.1371/journal.pgen.1000536.
 
PubMed
 PMID:
 19557195;
 PMCID:
 PMC2695001.
 
38.
  Gaugler
 T,
 Klei
 L,
 Sanders
 SJ,
 Bodea
 CA,
 Goldberg
 AP,
 Lee
 AB,
 Mahajan
 M,
 
Manaa
 D,
 Pawitan
 Y,
 Reichert
 J,
 Ripke
 S,
 Sandin
 S,
 Sklar
 P,
 Svantesson
 O,
 
Reichenberg
 A,
 Hultman
 CM,
 Devlin
 B,
 Roeder
 K,
 Buxbaum
 JD.
 Most
 genetic
 risk
 for
 
autism
 resides
 with
 common
 variation.
 Nat
 Genet.
 2014;46(8):881-­‐5.
 doi:
 
10.1038/ng.3039.
 PubMed
 PMID:
 25038753;
 PMCID:
 PMC4137411.
 
39.
  Liu
 L,
 Zhang
 D,
 Rodzinka-­‐Pasko
 JK,
 Li
 YM.
 Environmental
 risk
 factors
 for
 
autism
 spectrum
 disorders.
 Nervenarzt.
 2016;87(Suppl
 2):55-­‐61.
 doi:
 
10.1007/s00115-­‐016-­‐0172-­‐3.
 PubMed
 PMID:
 27531204.
 
40.
  Pinto
 D,
 Pagnamenta
 AT,
 Klei
 L,
 Anney
 R,
 Merico
 D,
 Regan
 R,
 Conroy
 J,
 
Magalhaes
 TR,
 Correia
 C,
 Abrahams
 BS,
 Almeida
 J,
 Bacchelli
 E,
 Bader
 GD,
 Bailey
 AJ,
 
Baird
 G,
 Battaglia
 A,
 Berney
 T,
 Bolshakova
 N,
 Bolte
 S,
 Bolton
 PF,
 Bourgeron
 T,
 
Brennan
 S,
 Brian
 J,
 Bryson
 SE,
 Carson
 AR,
 Casallo
 G,
 Casey
 J,
 Chung
 BH,
 Cochrane
 L,
 
Corsello
 C,
 Crawford
 EL,
 Crossett
 A,
 Cytrynbaum
 C,
 Dawson
 G,
 de
 Jonge
 M,
 Delorme
 
R,
 Drmic
 I,
 Duketis
 E,
 Duque
 F,
 Estes
 A,
 Farrar
 P,
 Fernandez
 BA,
 Folstein
 SE,
 
104
 

 
Fombonne
 E,
 Freitag
 CM,
 Gilbert
 J,
 Gillberg
 C,
 Glessner
 JT,
 Goldberg
 J,
 Green
 A,
 Green
 
J,
 Guter
 SJ,
 Hakonarson
 H,
 Heron
 EA,
 Hill
 M,
 Holt
 R,
 Howe
 JL,
 Hughes
 G,
 Hus
 V,
 
Igliozzi
 R,
 Kim
 C,
 Klauck
 SM,
 Kolevzon
 A,
 Korvatska
 O,
 Kustanovich
 V,
 Lajonchere
 
CM,
 Lamb
 JA,
 Laskawiec
 M,
 Leboyer
 M,
 Le
 Couteur
 A,
 Leventhal
 BL,
 Lionel
 AC,
 Liu
 
XQ,
 Lord
 C,
 Lotspeich
 L,
 Lund
 SC,
 Maestrini
 E,
 Mahoney
 W,
 Mantoulan
 C,
 Marshall
 
CR,
 McConachie
 H,
 McDougle
 CJ,
 McGrath
 J,
 McMahon
 WM,
 Merikangas
 A,
 Migita
 O,
 
Minshew
 NJ,
 Mirza
 GK,
 Munson
 J,
 Nelson
 SF,
 Noakes
 C,
 Noor
 A,
 Nygren
 G,
 Oliveira
 G,
 
Papanikolaou
 K,
 Parr
 JR,
 Parrini
 B,
 Paton
 T,
 Pickles
 A,
 Pilorge
 M,
 Piven
 J,
 Ponting
 CP,
 
Posey
 DJ,
 Poustka
 A,
 Poustka
 F,
 Prasad
 A,
 Ragoussis
 J,
 Renshaw
 K,
 Rickaby
 J,
 Roberts
 
W,
 Roeder
 K,
 Roge
 B,
 Rutter
 ML,
 Bierut
 LJ,
 Rice
 JP,
 Salt
 J,
 Sansom
 K,
 Sato
 D,
 Segurado
 
R,
 Sequeira
 AF,
 Senman
 L,
 Shah
 N,
 Sheffield
 VC,
 Soorya
 L,
 Sousa
 I,
 Stein
 O,
 Sykes
 N,
 
Stoppioni
 V,
 Strawbridge
 C,
 Tancredi
 R,
 Tansey
 K,
 Thiruvahindrapduram
 B,
 
Thompson
 AP,
 Thomson
 S,
 Tryfon
 A,
 Tsiantis
 J,
 Van
 Engeland
 H,
 Vincent
 JB,
 Volkmar
 
F,
 Wallace
 S,
 Wang
 K,
 Wang
 Z,
 Wassink
 TH,
 Webber
 C,
 Weksberg
 R,
 Wing
 K,
 
Wittemeyer
 K,
 Wood
 S,
 Wu
 J,
 Yaspan
 BL,
 Zurawiecki
 D,
 Zwaigenbaum
 L,
 Buxbaum
 
JD,
 Cantor
 RM,
 Cook
 EH,
 Coon
 H,
 Cuccaro
 ML,
 Devlin
 B,
 Ennis
 S,
 Gallagher
 L,
 
Geschwind
 DH,
 Gill
 M,
 Haines
 JL,
 Hallmayer
 J,
 Miller
 J,
 Monaco
 AP,
 Nurnberger
 JI,
 Jr.,
 
Paterson
 AD,
 Pericak-­‐Vance
 MA,
 Schellenberg
 GD,
 Szatmari
 P,
 Vicente
 AM,
 Vieland
 
VJ,
 Wijsman
 EM,
 Scherer
 SW,
 Sutcliffe
 JS,
 Betancur
 C.
 Functional
 impact
 of
 global
 
rare
 copy
 number
 variation
 in
 autism
 spectrum
 disorders.
 Nature.
 
2010;466(7304):368-­‐72.
 doi:
 10.1038/nature09146.
 PubMed
 PMID:
 20531469;
 
PMCID:
 PMC3021798.
 
41.
  Gilman
 SR,
 Iossifov
 I,
 Levy
 D,
 Ronemus
 M,
 Wigler
 M,
 Vitkup
 D.
 Rare
 de
 novo
 
variants
 associated
 with
 autism
 implicate
 a
 large
 functional
 network
 of
 genes
 
involved
 in
 formation
 and
 function
 of
 synapses.
 Neuron.
 2011;70(5):898-­‐907.
 doi:
 
10.1016/j.neuron.2011.05.021.
 PubMed
 PMID:
 21658583;
 PMCID:
 PMC3607702.
 
42.
  Sanders
 SJ,
 Ercan-­‐Sencicek
 AG,
 Hus
 V,
 Luo
 R,
 Murtha
 MT,
 Moreno-­‐De-­‐Luca
 D,
 
Chu
 SH,
 Moreau
 MP,
 Gupta
 AR,
 Thomson
 SA,
 Mason
 CE,
 Bilguvar
 K,
 Celestino-­‐Soper
 
PB,
 Choi
 M,
 Crawford
 EL,
 Davis
 L,
 Wright
 NR,
 Dhodapkar
 RM,
 DiCola
 M,
 DiLullo
 NM,
 
Fernandez
 TV,
 Fielding-­‐Singh
 V,
 Fishman
 DO,
 Frahm
 S,
 Garagaloyan
 R,
 Goh
 GS,
 
Kammela
 S,
 Klei
 L,
 Lowe
 JK,
 Lund
 SC,
 McGrew
 AD,
 Meyer
 KA,
 Moffat
 WJ,
 Murdoch
 JD,
 
O'Roak
 BJ,
 Ober
 GT,
 Pottenger
 RS,
 Raubeson
 MJ,
 Song
 Y,
 Wang
 Q,
 Yaspan
 BL,
 Yu
 TW,
 
Yurkiewicz
 IR,
 Beaudet
 AL,
 Cantor
 RM,
 Curland
 M,
 Grice
 DE,
 Gunel
 M,
 Lifton
 RP,
 
Mane
 SM,
 Martin
 DM,
 Shaw
 CA,
 Sheldon
 M,
 Tischfield
 JA,
 Walsh
 CA,
 Morrow
 EM,
 
Ledbetter
 DH,
 Fombonne
 E,
 Lord
 C,
 Martin
 CL,
 Brooks
 AI,
 Sutcliffe
 JS,
 Cook
 EH,
 Jr.,
 
Geschwind
 D,
 Roeder
 K,
 Devlin
 B,
 State
 MW.
 Multiple
 recurrent
 de
 novo
 CNVs,
 
including
 duplications
 of
 the
 7q11.23
 Williams
 syndrome
 region,
 are
 strongly
 
associated
 with
 autism.
 Neuron.
 2011;70(5):863-­‐85.
 doi:
 
10.1016/j.neuron.2011.05.002.
 PubMed
 PMID:
 21658581;
 PMCID:
 PMC3939065.
 
43.
  Schaaf
 CP,
 Zoghbi
 HY.
 Solving
 the
 autism
 puzzle
 a
 few
 pieces
 at
 a
 time.
 
Neuron.
 2011;70(5):806-­‐8.
 doi:
 10.1016/j.neuron.2011.05.025.
 PubMed
 PMID:
 
21658575.
 
44.
  Scott
 LJ,
 Mohlke
 KL,
 Bonnycastle
 LL,
 Willer
 CJ,
 Li
 Y,
 Duren
 WL,
 Erdos
 MR,
 
Stringham
 HM,
 Chines
 PS,
 Jackson
 AU,
 Prokunina-­‐Olsson
 L,
 Ding
 CJ,
 Swift
 AJ,
 Narisu
 
N,
 Hu
 T,
 Pruim
 R,
 Xiao
 R,
 Li
 XY,
 Conneely
 KN,
 Riebow
 NL,
 Sprau
 AG,
 Tong
 M,
 White
 
PP,
 Hetrick
 KN,
 Barnhart
 MW,
 Bark
 CW,
 Goldstein
 JL,
 Watkins
 L,
 Xiang
 F,
 Saramies
 J,
 
105
 

 
Buchanan
 TA,
 Watanabe
 RM,
 Valle
 TT,
 Kinnunen
 L,
 Abecasis
 GR,
 Pugh
 EW,
 Doheny
 
KF,
 Bergman
 RN,
 Tuomilehto
 J,
 Collins
 FS,
 Boehnke
 M.
 A
 genome-­‐wide
 association
 
study
 of
 type
 2
 diabetes
 in
 Finns
 detects
 multiple
 susceptibility
 variants.
 Science.
 
2007;316(5829):1341-­‐5.
 doi:
 10.1126/science.1142382.
 PubMed
 PMID:
 17463248;
 
PMCID:
 PMC3214617.
 
45.
  Zeggini
 E,
 Weedon
 MN,
 Lindgren
 CM,
 Frayling
 TM,
 Elliott
 KS,
 Lango
 H,
 
Timpson
 NJ,
 Perry
 JR,
 Rayner
 NW,
 Freathy
 RM,
 Barrett
 JC,
 Shields
 B,
 Morris
 AP,
 
Ellard
 S,
 Groves
 CJ,
 Harries
 LW,
 Marchini
 JL,
 Owen
 KR,
 Knight
 B,
 Cardon
 LR,
 Walker
 
M,
 Hitman
 GA,
 Morris
 AD,
 Doney
 AS,
 Wellcome
 Trust
 Case
 Control
 C,
 McCarthy
 MI,
 
Hattersley
 AT.
 Replication
 of
 genome-­‐wide
 association
 signals
 in
 UK
 samples
 reveals
 
risk
 loci
 for
 type
 2
 diabetes.
 Science.
 2007;316(5829):1336-­‐41.
 doi:
 
10.1126/science.1142364.
 PubMed
 PMID:
 17463249;
 PMCID:
 PMC3772310.
 
46.
  Hakonarson
 H,
 Grant
 SF.
 Genome-­‐wide
 association
 studies
 (GWAS):
 impact
 
on
 elucidating
 the
 aetiology
 of
 diabetes.
 Diabetes
 Metab
 Res
 Rev.
 2011;27(7):685-­‐
96.
 doi:
 10.1002/dmrr.1221.
 PubMed
 PMID:
 21630414.
 
47.
  Franke
 A,
 Balschun
 T,
 Karlsen
 TH,
 Hedderich
 J,
 May
 S,
 Lu
 T,
 Schuldt
 D,
 
Nikolaus
 S,
 Rosenstiel
 P,
 Krawczak
 M,
 Schreiber
 S.
 Replication
 of
 signals
 from
 recent
 
studies
 of
 Crohn's
 disease
 identifies
 previously
 unknown
 disease
 loci
 for
 ulcerative
 
colitis.
 Nat
 Genet.
 2008;40(6):713-­‐5.
 doi:
 10.1038/ng.148.
 PubMed
 PMID:
 
18438405.
 
48.
  Franke
 A,
 McGovern
 DP,
 Barrett
 JC,
 Wang
 K,
 Radford-­‐Smith
 GL,
 Ahmad
 T,
 
Lees
 CW,
 Balschun
 T,
 Lee
 J,
 Roberts
 R,
 Anderson
 CA,
 Bis
 JC,
 Bumpstead
 S,
 Ellinghaus
 
D,
 Festen
 EM,
 Georges
 M,
 Green
 T,
 Haritunians
 T,
 Jostins
 L,
 Latiano
 A,
 Mathew
 CG,
 
Montgomery
 GW,
 Prescott
 NJ,
 Raychaudhuri
 S,
 Rotter
 JI,
 Schumm
 P,
 Sharma
 Y,
 
Simms
 LA,
 Taylor
 KD,
 Whiteman
 D,
 Wijmenga
 C,
 Baldassano
 RN,
 Barclay
 M,
 Bayless
 
TM,
 Brand
 S,
 Buning
 C,
 Cohen
 A,
 Colombel
 JF,
 Cottone
 M,
 Stronati
 L,
 Denson
 T,
 De
 
Vos
 M,
 D'Inca
 R,
 Dubinsky
 M,
 Edwards
 C,
 Florin
 T,
 Franchimont
 D,
 Gearry
 R,
 Glas
 J,
 
Van
 Gossum
 A,
 Guthery
 SL,
 Halfvarson
 J,
 Verspaget
 HW,
 Hugot
 JP,
 Karban
 A,
 Laukens
 
D,
 Lawrance
 I,
 Lemann
 M,
 Levine
 A,
 Libioulle
 C,
 Louis
 E,
 Mowat
 C,
 Newman
 W,
 Panes
 
J,
 Phillips
 A,
 Proctor
 DD,
 Regueiro
 M,
 Russell
 R,
 Rutgeerts
 P,
 Sanderson
 J,
 Sans
 M,
 
Seibold
 F,
 Steinhart
 AH,
 Stokkers
 PC,
 Torkvist
 L,
 Kullak-­‐Ublick
 G,
 Wilson
 D,
 Walters
 
T,
 Targan
 SR,
 Brant
 SR,
 Rioux
 JD,
 D'Amato
 M,
 Weersma
 RK,
 Kugathasan
 S,
 Griffiths
 
AM,
 Mansfield
 JC,
 Vermeire
 S,
 Duerr
 RH,
 Silverberg
 MS,
 Satsangi
 J,
 Schreiber
 S,
 Cho
 
JH,
 Annese
 V,
 Hakonarson
 H,
 Daly
 MJ,
 Parkes
 M.
 Genome-­‐wide
 meta-­‐analysis
 
increases
 to
 71
 the
 number
 of
 confirmed
 Crohn's
 disease
 susceptibility
 loci.
 Nat
 
Genet.
 2010;42(12):1118-­‐25.
 doi:
 10.1038/ng.717.
 PubMed
 PMID:
 21102463;
 
PMCID:
 PMC3299551.
 
49.
  Haiman
 CA,
 Le
 Marchand
 L,
 Yamamato
 J,
 Stram
 DO,
 Sheng
 X,
 Kolonel
 LN,
 Wu
 
AH,
 Reich
 D,
 Henderson
 BE.
 A
 common
 genetic
 risk
 factor
 for
 colorectal
 and
 prostate
 
cancer.
 Nat
 Genet.
 2007;39(8):954-­‐6.
 doi:
 10.1038/ng2098.
 PubMed
 PMID:
 
17618282;
 PMCID:
 PMC2391283.
 
50.
  Haiman
 CA,
 Patterson
 N,
 Freedman
 ML,
 Myers
 SR,
 Pike
 MC,
 Waliszewska
 A,
 
Neubauer
 J,
 Tandon
 A,
 Schirmer
 C,
 McDonald
 GJ,
 Greenway
 SC,
 Stram
 DO,
 Le
 
Marchand
 L,
 Kolonel
 LN,
 Frasco
 M,
 Wong
 D,
 Pooler
 LC,
 Ardlie
 K,
 Oakley-­‐Girvan
 I,
 
Whittemore
 AS,
 Cooney
 KA,
 John
 EM,
 Ingles
 SA,
 Altshuler
 D,
 Henderson
 BE,
 Reich
 D.
 
Multiple
 regions
 within
 8q24
 independently
 affect
 risk
 for
 prostate
 cancer.
 Nat
 
106
 

 
Genet.
 2007;39(5):638-­‐44.
 doi:
 10.1038/ng2015.
 PubMed
 PMID:
 17401364;
 PMCID:
 
PMC2638766.
 
51.
  Gillis
 RF,
 Rouleau
 GA.
 The
 ongoing
 dissection
 of
 the
 genetic
 architecture
 of
 
autistic
 spectrum
 disorder.
 Mol
 Autism.
 2011;2(1):12.
 doi:
 10.1186/2040-­‐2392-­‐2-­‐
12.
 PubMed
 PMID:
 21740537;
 PMCID:
 PMC3156724.
 
52.
  Levy
 D,
 Ronemus
 M,
 Yamrom
 B,
 Lee
 YH,
 Leotta
 A,
 Kendall
 J,
 Marks
 S,
 Lakshmi
 
B,
 Pai
 D,
 Ye
 K,
 Buja
 A,
 Krieger
 A,
 Yoon
 S,
 Troge
 J,
 Rodgers
 L,
 Iossifov
 I,
 Wigler
 M.
 Rare
 
de
 novo
 and
 transmitted
 copy-­‐number
 variation
 in
 autistic
 spectrum
 disorders.
 
Neuron.
 2011;70(5):886-­‐97.
 doi:
 10.1016/j.neuron.2011.05.015.
 PubMed
 PMID:
 
21658582.
 
53.
  Scherer
 SW,
 Dawson
 G.
 Risk
 factors
 for
 autism:
 translating
 genomic
 
discoveries
 into
 diagnostics.
 Hum
 Genet.
 2011;130(1):123-­‐48.
 doi:
 
10.1007/s00439-­‐011-­‐1037-­‐2.
 PubMed
 PMID:
 21701786.
 
54.
  Bilguvar
 K,
 Ozturk
 AK,
 Louvi
 A,
 Kwan
 KY,
 Choi
 M,
 Tatli
 B,
 Yalnizoglu
 D,
 
Tuysuz
 B,
 Caglayan
 AO,
 Gokben
 S,
 Kaymakcalan
 H,
 Barak
 T,
 Bakircioglu
 M,
 Yasuno
 K,
 
Ho
 W,
 Sanders
 S,
 Zhu
 Y,
 Yilmaz
 S,
 Dincer
 A,
 Johnson
 MH,
 Bronen
 RA,
 Kocer
 N,
 Per
 H,
 
Mane
 S,
 Pamir
 MN,
 Yalcinkaya
 C,
 Kumandas
 S,
 Topcu
 M,
 Ozmen
 M,
 Sestan
 N,
 Lifton
 
RP,
 State
 MW,
 Gunel
 M.
 Whole-­‐exome
 sequencing
 identifies
 recessive
 WDR62
 
mutations
 in
 severe
 brain
 malformations.
 Nature.
 2010;467(7312):207-­‐10.
 doi:
 
10.1038/nature09327.
 PubMed
 PMID:
 20729831;
 PMCID:
 PMC3129007.
 
55.
  Noor
 A,
 Whibley
 A,
 Marshall
 CR,
 Gianakopoulos
 PJ,
 Piton
 A,
 Carson
 AR,
 Orlic-­‐
Milacic
 M,
 Lionel
 AC,
 Sato
 D,
 Pinto
 D,
 Drmic
 I,
 Noakes
 C,
 Senman
 L,
 Zhang
 X,
 Mo
 R,
 
Gauthier
 J,
 Crosbie
 J,
 Pagnamenta
 AT,
 Munson
 J,
 Estes
 AM,
 Fiebig
 A,
 Franke
 A,
 
Schreiber
 S,
 Stewart
 AF,
 Roberts
 R,
 McPherson
 R,
 Guter
 SJ,
 Cook
 EH,
 Jr.,
 Dawson
 G,
 
Schellenberg
 GD,
 Battaglia
 A,
 Maestrini
 E,
 Autism
 Genome
 Project
 C,
 Jeng
 L,
 
Hutchison
 T,
 Rajcan-­‐Separovic
 E,
 Chudley
 AE,
 Lewis
 SM,
 Liu
 X,
 Holden
 JJ,
 Fernandez
 
B,
 Zwaigenbaum
 L,
 Bryson
 SE,
 Roberts
 W,
 Szatmari
 P,
 Gallagher
 L,
 Stratton
 MR,
 Gecz
 
J,
 Brady
 AF,
 Schwartz
 CE,
 Schachar
 RJ,
 Monaco
 AP,
 Rouleau
 GA,
 Hui
 CC,
 Lucy
 
Raymond
 F,
 Scherer
 SW,
 Vincent
 JB.
 Disruption
 at
 the
 PTCHD1
 Locus
 on
 Xp22.11
 in
 
Autism
 spectrum
 disorder
 and
 intellectual
 disability.
 Sci
 Transl
 Med.
 
2010;2(49):49ra68.
 doi:
 10.1126/scitranslmed.3001267.
 PubMed
 PMID:
 20844286;
 
PMCID:
 PMC2987731.
 
56.
  Shi
 Y,
 Li
 Y,
 Zhang
 D,
 Zhang
 H,
 Li
 Y,
 Lu
 F,
 Liu
 X,
 He
 F,
 Gong
 B,
 Cai
 L,
 Li
 R,
 Liao
 S,
 
Ma
 S,
 Lin
 H,
 Cheng
 J,
 Zheng
 H,
 Shan
 Y,
 Chen
 B,
 Hu
 J,
 Jin
 X,
 Zhao
 P,
 Chen
 Y,
 Zhang
 Y,
 Lin
 
Y,
 Li
 X,
 Fan
 Y,
 Yang
 H,
 Wang
 J,
 Yang
 Z.
 Exome
 sequencing
 identifies
 ZNF644
 
mutations
 in
 high
 myopia.
 PLoS
 Genet.
 2011;7(6):e1002084.
 doi:
 
10.1371/journal.pgen.1002084.
 PubMed
 PMID:
 21695231;
 PMCID:
 PMC3111487.
 
57.
  Losh
 M,
 Adolphs
 R,
 Poe
 MD,
 Couture
 S,
 Penn
 D,
 Baranek
 GT,
 Piven
 J.
 
Neuropsychological
 profile
 of
 autism
 and
 the
 broad
 autism
 phenotype.
 Arch
 Gen
 
Psychiatry.
 2009;66(5):518-­‐26.
 doi:
 10.1001/archgenpsychiatry.2009.34.
 PubMed
 
PMID:
 19414711;
 PMCID:
 PMC2699548.
 
58.
  Losh
 M,
 Childress
 D,
 Lam
 K,
 Piven
 J.
 Defining
 key
 features
 of
 the
 broad
 autism
 
phenotype:
 a
 comparison
 across
 parents
 of
 multiple-­‐
 and
 single-­‐incidence
 autism
 
families.
 Am
 J
 Med
 Genet
 B
 Neuropsychiatr
 Genet.
 2008;147B(4):424-­‐33.
 doi:
 
10.1002/ajmg.b.30612.
 PubMed
 PMID:
 17948871;
 PMCID:
 PMC2746421.
 
107
 

 
59.
  Almasy
 L,
 Gur
 RC,
 Haack
 K,
 Cole
 SA,
 Calkins
 ME,
 Peralta
 JM,
 Hare
 E,
 Prasad
 K,
 
Pogue-­‐Geile
 MF,
 Nimgaonkar
 V,
 Gur
 RE.
 A
 genome
 screen
 for
 quantitative
 trait
 loci
 
influencing
 schizophrenia
 and
 neurocognitive
 phenotypes.
 Am
 J
 Psychiatry.
 
2008;165(9):1185-­‐92.
 doi:
 10.1176/appi.ajp.2008.07121869.
 PubMed
 PMID:
 
18628350;
 PMCID:
 PMC2644284.
 
60.
  Egan
 MF,
 Goldberg
 TE,
 Gscheidle
 T,
 Weirich
 M,
 Rawlings
 R,
 Hyde
 TM,
 Bigelow
 
L,
 Weinberger
 DR.
 Relative
 risk
 for
 cognitive
 impairments
 in
 siblings
 of
 patients
 
with
 schizophrenia.
 Biol
 Psychiatry.
 2001;50(2):98-­‐107.
 PubMed
 PMID:
 11527000.
 
61.
  Robinson
 EB,
 Koenen
 KC,
 McCormick
 MC,
 Munir
 K,
 Hallett
 V,
 Happe
 F,
 Plomin
 
R,
 Ronald
 A.
 Evidence
 that
 autistic
 traits
 show
 the
 same
 etiology
 in
 the
 general
 
population
 and
 at
 the
 quantitative
 extremes
 (5%,
 2.5%,
 and
 1%).
 Arch
 Gen
 
Psychiatry.
 2011;68(11):1113-­‐21.
 doi:
 10.1001/archgenpsychiatry.2011.119.
 
PubMed
 PMID:
 22065527;
 PMCID:
 PMC3708488.
 
62.
  Lundstrom
 S,
 Chang
 Z,
 Rastam
 M,
 Gillberg
 C,
 Larsson
 H,
 Anckarsater
 H,
 
Lichtenstein
 P.
 Autism
 spectrum
 disorders
 and
 autistic
 like
 traits:
 similar
 etiology
 in
 
the
 extreme
 end
 and
 the
 normal
 variation.
 Arch
 Gen
 Psychiatry.
 2012;69(1):46-­‐52.
 
doi:
 10.1001/archgenpsychiatry.2011.144.
 PubMed
 PMID:
 22213788.
 
63.
  Devlin
 B,
 Melhem
 N,
 Roeder
 K.
 Do
 common
 variants
 play
 a
 role
 in
 risk
 for
 
autism?
 Evidence
 and
 theoretical
 musings.
 Brain
 Res.
 2011;1380:78-­‐84.
 doi:
 
10.1016/j.brainres.2010.11.026.
 PubMed
 PMID:
 21078308;
 PMCID:
 PMC3638767.
 
64.
  Morrison
 AC,
 Voorman
 A,
 Johnson
 AD,
 Liu
 X,
 Yu
 J,
 Li
 A,
 Muzny
 D,
 Yu
 F,
 Rice
 K,
 
Zhu
 C,
 Bis
 J,
 Heiss
 G,
 O'Donnell
 CJ,
 Psaty
 BM,
 Cupples
 LA,
 Gibbs
 R,
 Boerwinkle
 E,
 
Cohorts
 for
 H,
 Aging
 Research
 in
 Genetic
 Epidemiology
 C.
 Whole-­‐genome
 sequence-­‐
based
 analysis
 of
 high-­‐density
 lipoprotein
 cholesterol.
 Nat
 Genet.
 2013;45(8):899-­‐
901.
 doi:
 10.1038/ng.2671.
 PubMed
 PMID:
 23770607;
 PMCID:
 PMC4030301.
 
65.
  Kapranov
 P,
 St
 Laurent
 G,
 Raz
 T,
 Ozsolak
 F,
 Reynolds
 CP,
 Sorensen
 PH,
 
Reaman
 G,
 Milos
 P,
 Arceci
 RJ,
 Thompson
 JF,
 Triche
 TJ.
 The
 majority
 of
 total
 nuclear-­‐
encoded
 non-­‐ribosomal
 RNA
 in
 a
 human
 cell
 is
 'dark
 matter'
 un-­‐annotated
 RNA.
 
BMC
 Biol.
 2010;8:149.
 doi:
 10.1186/1741-­‐7007-­‐8-­‐149.
 PubMed
 PMID:
 21176148;
 
PMCID:
 PMC3022773.
 
66.
  Mattick
 JS.
 The
 genetic
 signatures
 of
 noncoding
 RNAs.
 PLoS
 Genet.
 
2009;5(4):e1000459.
 doi:
 10.1371/journal.pgen.1000459.
 PubMed
 PMID:
 
19390609;
 PMCID:
 PMC2667263.
 
67.
  DeWitt
 JJ,
 Grepo
 N,
 Wilkinson
 B,
 Evgrafov
 OV,
 Knowles
 JA,
 Campbell
 DB.
 
Impact
 of
 the
 Autism-­‐Associated
 Long
 Noncoding
 RNA
 MSNP1AS
 on
 Neuronal
 
Architecture
 and
 Gene
 Expression
 in
 Human
 Neural
 Progenitor
 Cells.
 Genes
 (Basel).
 
2016;7(10).
 doi:
 10.3390/genes7100076.
 PubMed
 PMID:
 27690106;
 PMCID:
 
PMC5083915.
 
68.
  Qureshi
 IA,
 Mattick
 JS,
 Mehler
 MF.
 Long
 non-­‐coding
 RNAs
 in
 nervous
 system
 
function
 and
 disease.
 Brain
 Res.
 2010;1338:20-­‐35.
 doi:
 
10.1016/j.brainres.2010.03.110.
 PubMed
 PMID:
 20380817;
 PMCID:
 PMC2883659.
 
69.
  DeWitt
 JJ,
 Hecht
 PM,
 Grepo
 N,
 Wilkinson
 B,
 Evgrafov
 OV,
 Morris
 KV,
 Knowles
 
JA,
 Campbell
 DB.
 Transcriptional
 Gene
 Silencing
 of
 the
 Autism-­‐Associated
 Long
 
Noncoding
 RNA
 MSNP1AS
 in
 Human
 Neural
 Progenitor
 Cells.
 Dev
 Neurosci.
 
2016;38(5):375-­‐83.
 doi:
 10.1159/000453258.
 PubMed
 PMID:
 28030860.
 
108
 

 
70.
  Lin
 M,
 Pedrosa
 E,
 Shah
 A,
 Hrabovsky
 A,
 Maqbool
 S,
 Zheng
 D,
 Lachman
 HM.
 
RNA-­‐Seq
 of
 human
 neurons
 derived
 from
 iPS
 cells
 reveals
 candidate
 long
 non-­‐
coding
 RNAs
 involved
 in
 neurogenesis
 and
 neuropsychiatric
 disorders.
 PLoS
 One.
 
2011;6(9):e23356.
 doi:
 10.1371/journal.pone.0023356.
 PubMed
 PMID:
 21915259;
 
PMCID:
 PMC3168439.
 
71.
  Jin
 G,
 Sun
 J,
 Isaacs
 SD,
 Wiley
 KE,
 Kim
 ST,
 Chu
 LW,
 Zhang
 Z,
 Zhao
 H,
 Zheng
 SL,
 
Isaacs
 WB,
 Xu
 J.
 Human
 polymorphisms
 at
 long
 non-­‐coding
 RNAs
 (lncRNAs)
 and
 
association
 with
 prostate
 cancer
 risk.
 Carcinogenesis.
 2011;32(11):1655-­‐9.
 doi:
 
10.1093/carcin/bgr187.
 PubMed
 PMID:
 21856995;
 PMCID:
 PMC3204347.
 
72.
  Ripke
 S,
 O'Dushlaine
 C,
 Chambert
 K,
 Moran
 JL,
 Kahler
 AK,
 Akterin
 S,
 Bergen
 
SE,
 Collins
 AL,
 Crowley
 JJ,
 Fromer
 M,
 Kim
 Y,
 Lee
 SH,
 Magnusson
 PK,
 Sanchez
 N,
 Stahl
 
EA,
 Williams
 S,
 Wray
 NR,
 Xia
 K,
 Bettella
 F,
 Borglum
 AD,
 Bulik-­‐Sullivan
 BK,
 Cormican
 
P,
 Craddock
 N,
 de
 Leeuw
 C,
 Durmishi
 N,
 Gill
 M,
 Golimbet
 V,
 Hamshere
 ML,
 Holmans
 
P,
 Hougaard
 DM,
 Kendler
 KS,
 Lin
 K,
 Morris
 DW,
 Mors
 O,
 Mortensen
 PB,
 Neale
 BM,
 
O'Neill
 FA,
 Owen
 MJ,
 Milovancevic
 MP,
 Posthuma
 D,
 Powell
 J,
 Richards
 AL,
 Riley
 BP,
 
Ruderfer
 D,
 Rujescu
 D,
 Sigurdsson
 E,
 Silagadze
 T,
 Smit
 AB,
 Stefansson
 H,
 Steinberg
 S,
 
Suvisaari
 J,
 Tosato
 S,
 Verhage
 M,
 Walters
 JT,
 Multicenter
 Genetic
 Studies
 of
 
Schizophrenia
 C,
 Levinson
 DF,
 Gejman
 PV,
 Kendler
 KS,
 Laurent
 C,
 Mowry
 BJ,
 
O'Donovan
 MC,
 Owen
 MJ,
 Pulver
 AE,
 Riley
 BP,
 Schwab
 SG,
 Wildenauer
 DB,
 
Dudbridge
 F,
 Holmans
 P,
 Shi
 J,
 Albus
 M,
 Alexander
 M,
 Campion
 D,
 Cohen
 D,
 Dikeos
 D,
 
Duan
 J,
 Eichhammer
 P,
 Godard
 S,
 Hansen
 M,
 Lerer
 FB,
 Liang
 KY,
 Maier
 W,
 Mallet
 J,
 
Nertney
 DA,
 Nestadt
 G,
 Norton
 N,
 O'Neill
 FA,
 Papadimitriou
 GN,
 Ribble
 R,
 Sanders
 
AR,
 Silverman
 JM,
 Walsh
 D,
 Williams
 NM,
 Wormley
 B,
 Psychosis
 Endophenotypes
 
International
 C,
 Arranz
 MJ,
 Bakker
 S,
 Bender
 S,
 Bramon
 E,
 Collier
 D,
 Crespo-­‐Facorro
 
B,
 Hall
 J,
 Iyegbe
 C,
 Jablensky
 A,
 Kahn
 RS,
 Kalaydjieva
 L,
 Lawrie
 S,
 Lewis
 CM,
 Lin
 K,
 
Linszen
 DH,
 Mata
 I,
 McIntosh
 A,
 Murray
 RM,
 Ophoff
 RA,
 Powell
 J,
 Rujescu
 D,
 Van
 Os
 J,
 
Walshe
 M,
 Weisbrod
 M,
 Wiersma
 D,
 Wellcome
 Trust
 Case
 Control
 C,
 Donnelly
 P,
 
Barroso
 I,
 Blackwell
 JM,
 Bramon
 E,
 Brown
 MA,
 Casas
 JP,
 Corvin
 AP,
 Deloukas
 P,
 
Duncanson
 A,
 Jankowski
 J,
 Markus
 HS,
 Mathew
 CG,
 Palmer
 CN,
 Plomin
 R,
 Rautanen
 
A,
 Sawcer
 SJ,
 Trembath
 RC,
 Viswanathan
 AC,
 Wood
 NW,
 Spencer
 CC,
 Band
 G,
 
Bellenguez
 C,
 Freeman
 C,
 Hellenthal
 G,
 Giannoulatou
 E,
 Pirinen
 M,
 Pearson
 RD,
 
Strange
 A,
 Su
 Z,
 Vukcevic
 D,
 Donnelly
 P,
 Langford
 C,
 Hunt
 SE,
 Edkins
 S,
 Gwilliam
 R,
 
Blackburn
 H,
 Bumpstead
 SJ,
 Dronov
 S,
 Gillman
 M,
 Gray
 E,
 Hammond
 N,
 Jayakumar
 A,
 
McCann
 OT,
 Liddle
 J,
 Potter
 SC,
 Ravindrarajah
 R,
 Ricketts
 M,
 Tashakkori-­‐Ghanbaria
 
A,
 Waller
 MJ,
 Weston
 P,
 Widaa
 S,
 Whittaker
 P,
 Barroso
 I,
 Deloukas
 P,
 Mathew
 CG,
 
Blackwell
 JM,
 Brown
 MA,
 Corvin
 AP,
 McCarthy
 MI,
 Spencer
 CC,
 Bramon
 E,
 Corvin
 AP,
 
O'Donovan
 MC,
 Stefansson
 K,
 Scolnick
 E,
 Purcell
 S,
 McCarroll
 SA,
 Sklar
 P,
 Hultman
 
CM,
 Sullivan
 PF.
 Genome-­‐wide
 association
 analysis
 identifies
 13
 new
 risk
 loci
 for
 
schizophrenia.
 Nat
 Genet.
 2013;45(10):1150-­‐9.
 doi:
 10.1038/ng.2742.
 PubMed
 
PMID:
 23974872;
 PMCID:
 PMC3827979.
 
73.
  St
 Pourcain
 B,
 Wang
 K,
 Glessner
 JT,
 Golding
 J,
 Steer
 C,
 Ring
 SM,
 Skuse
 DH,
 
Grant
 SF,
 Hakonarson
 H,
 Davey
 Smith
 G.
 Association
 between
 a
 high-­‐risk
 autism
 
locus
 on
 5p14
 and
 social
 communication
 spectrum
 phenotypes
 in
 the
 general
 
population.
 Am
 J
 Psychiatry.
 2010;167(11):1364-­‐72.
 doi:
 
10.1176/appi.ajp.2010.09121789.
 PubMed
 PMID:
 20634369;
 PMCID:
 PMC3008767.
 
109
 

 
74.
  Kerin
 T,
 Ramanathan
 A,
 Rivas
 K,
 Grepo
 N,
 Coetzee
 GA,
 Campbell
 DB.
 A
 
noncoding
 RNA
 antisense
 to
 moesin
 at
 5p14.1
 in
 autism.
 Sci
 Transl
 Med.
 
2012;4(128):128ra40.
 doi:
 10.1126/scitranslmed.3003479.
 PubMed
 PMID:
 
22491950.
 
75.
  Lintas
 C,
 Sacco
 R,
 Persico
 AM.
 Genome-­‐wide
 expression
 studies
 in
 autism
 
spectrum
 disorder,
 Rett
 syndrome,
 and
 Down
 syndrome.
 Neurobiol
 Dis.
 
2012;45(1):57-­‐68.
 doi:
 10.1016/j.nbd.2010.11.010.
 PubMed
 PMID:
 21130877.
 
76.
  Voineagu
 I.
 Gene
 expression
 studies
 in
 autism:
 moving
 from
 the
 genome
 to
 
the
 transcriptome
 and
 beyond.
 Neurobiol
 Dis.
 2012;45(1):69-­‐75.
 doi:
 
10.1016/j.nbd.2011.07.017.
 PubMed
 PMID:
 21839838.
 
77.
  Ziats
 MN,
 Rennert
 OM.
 Aberrant
 expression
 of
 long
 noncoding
 RNAs
 in
 
autistic
 brain.
 J
 Mol
 Neurosci.
 2013;49(3):589-­‐93.
 doi:
 10.1007/s12031-­‐012-­‐9880-­‐
8.
 PubMed
 PMID:
 22949041;
 PMCID:
 PMC3566384.
 
78.
  Carninci
 P,
 Kasukawa
 T,
 Katayama
 S,
 Gough
 J,
 Frith
 MC,
 Maeda
 N,
 Oyama
 R,
 
Ravasi
 T,
 Lenhard
 B,
 Wells
 C,
 Kodzius
 R,
 Shimokawa
 K,
 Bajic
 VB,
 Brenner
 SE,
 Batalov
 
S,
 Forrest
 AR,
 Zavolan
 M,
 Davis
 MJ,
 Wilming
 LG,
 Aidinis
 V,
 Allen
 JE,
 Ambesi-­‐
Impiombato
 A,
 Apweiler
 R,
 Aturaliya
 RN,
 Bailey
 TL,
 Bansal
 M,
 Baxter
 L,
 Beisel
 KW,
 
Bersano
 T,
 Bono
 H,
 Chalk
 AM,
 Chiu
 KP,
 Choudhary
 V,
 Christoffels
 A,
 Clutterbuck
 DR,
 
Crowe
 ML,
 Dalla
 E,
 Dalrymple
 BP,
 de
 Bono
 B,
 Della
 Gatta
 G,
 di
 Bernardo
 D,
 Down
 T,
 
Engstrom
 P,
 Fagiolini
 M,
 Faulkner
 G,
 Fletcher
 CF,
 Fukushima
 T,
 Furuno
 M,
 Futaki
 S,
 
Gariboldi
 M,
 Georgii-­‐Hemming
 P,
 Gingeras
 TR,
 Gojobori
 T,
 Green
 RE,
 Gustincich
 S,
 
Harbers
 M,
 Hayashi
 Y,
 Hensch
 TK,
 Hirokawa
 N,
 Hill
 D,
 Huminiecki
 L,
 Iacono
 M,
 Ikeo
 
K,
 Iwama
 A,
 Ishikawa
 T,
 Jakt
 M,
 Kanapin
 A,
 Katoh
 M,
 Kawasawa
 Y,
 Kelso
 J,
 Kitamura
 
H,
 Kitano
 H,
 Kollias
 G,
 Krishnan
 SP,
 Kruger
 A,
 Kummerfeld
 SK,
 Kurochkin
 IV,
 Lareau
 
LF,
 Lazarevic
 D,
 Lipovich
 L,
 Liu
 J,
 Liuni
 S,
 McWilliam
 S,
 Madan
 Babu
 M,
 Madera
 M,
 
Marchionni
 L,
 Matsuda
 H,
 Matsuzawa
 S,
 Miki
 H,
 Mignone
 F,
 Miyake
 S,
 Morris
 K,
 
Mottagui-­‐Tabar
 S,
 Mulder
 N,
 Nakano
 N,
 Nakauchi
 H,
 Ng
 P,
 Nilsson
 R,
 Nishiguchi
 S,
 
Nishikawa
 S,
 Nori
 F,
 Ohara
 O,
 Okazaki
 Y,
 Orlando
 V,
 Pang
 KC,
 Pavan
 WJ,
 Pavesi
 G,
 
Pesole
 G,
 Petrovsky
 N,
 Piazza
 S,
 Reed
 J,
 Reid
 JF,
 Ring
 BZ,
 Ringwald
 M,
 Rost
 B,
 Ruan
 Y,
 
Salzberg
 SL,
 Sandelin
 A,
 Schneider
 C,
 Schonbach
 C,
 Sekiguchi
 K,
 Semple
 CA,
 Seno
 S,
 
Sessa
 L,
 Sheng
 Y,
 Shibata
 Y,
 Shimada
 H,
 Shimada
 K,
 Silva
 D,
 Sinclair
 B,
 Sperling
 S,
 
Stupka
 E,
 Sugiura
 K,
 Sultana
 R,
 Takenaka
 Y,
 Taki
 K,
 Tammoja
 K,
 Tan
 SL,
 Tang
 S,
 
Taylor
 MS,
 Tegner
 J,
 Teichmann
 SA,
 Ueda
 HR,
 van
 Nimwegen
 E,
 Verardo
 R,
 Wei
 CL,
 
Yagi
 K,
 Yamanishi
 H,
 Zabarovsky
 E,
 Zhu
 S,
 Zimmer
 A,
 Hide
 W,
 Bult
 C,
 Grimmond
 SM,
 
Teasdale
 RD,
 Liu
 ET,
 Brusic
 V,
 Quackenbush
 J,
 Wahlestedt
 C,
 Mattick
 JS,
 Hume
 DA,
 
Kai
 C,
 Sasaki
 D,
 Tomaru
 Y,
 Fukuda
 S,
 Kanamori-­‐Katayama
 M,
 Suzuki
 M,
 Aoki
 J,
 
Arakawa
 T,
 Iida
 J,
 Imamura
 K,
 Itoh
 M,
 Kato
 T,
 Kawaji
 H,
 Kawagashira
 N,
 Kawashima
 
T,
 Kojima
 M,
 Kondo
 S,
 Konno
 H,
 Nakano
 K,
 Ninomiya
 N,
 Nishio
 T,
 Okada
 M,
 Plessy
 C,
 
Shibata
 K,
 Shiraki
 T,
 Suzuki
 S,
 Tagami
 M,
 Waki
 K,
 Watahiki
 A,
 Okamura-­‐Oho
 Y,
 
Suzuki
 H,
 Kawai
 J,
 Hayashizaki
 Y,
 Consortium
 F,
 Group
 RGER,
 Genome
 Science
 G.
 The
 
transcriptional
 landscape
 of
 the
 mammalian
 genome.
 Science.
 
2005;309(5740):1559-­‐63.
 doi:
 10.1126/science.1112014.
 PubMed
 PMID:
 
16141072.
 
79.
  Kapranov
 P,
 Cheng
 J,
 Dike
 S,
 Nix
 DA,
 Duttagupta
 R,
 Willingham
 AT,
 Stadler
 PF,
 
Hertel
 J,
 Hackermuller
 J,
 Hofacker
 IL,
 Bell
 I,
 Cheung
 E,
 Drenkow
 J,
 Dumais
 E,
 Patel
 S,
 
Helt
 G,
 Ganesh
 M,
 Ghosh
 S,
 Piccolboni
 A,
 Sementchenko
 V,
 Tammana
 H,
 Gingeras
 TR.
 
110
 

 
RNA
 maps
 reveal
 new
 RNA
 classes
 and
 a
 possible
 function
 for
 pervasive
 
transcription.
 Science.
 2007;316(5830):1484-­‐8.
 doi:
 10.1126/science.1138341.
 
PubMed
 PMID:
 17510325.
 
80.
  Mattick
 JS,
 Makunin
 IV.
 Small
 regulatory
 RNAs
 in
 mammals.
 Hum
 Mol
 Genet.
 
2005;14
 Spec
 No
 1:R121-­‐32.
 doi:
 10.1093/hmg/ddi101.
 PubMed
 PMID:
 15809264.
 
81.
  Jalali
 S,
 Jayaraj
 GG,
 Scaria
 V.
 Integrative
 transcriptome
 analysis
 suggest
 
processing
 of
 a
 subset
 of
 long
 non-­‐coding
 RNAs
 to
 small
 RNAs.
 Biol
 Direct.
 
2012;7:25.
 doi:
 10.1186/1745-­‐6150-­‐7-­‐25.
 PubMed
 PMID:
 22871084;
 PMCID:
 
PMC3477000.
 
82.
  Keniry
 A,
 Oxley
 D,
 Monnier
 P,
 Kyba
 M,
 Dandolo
 L,
 Smits
 G,
 Reik
 W.
 The
 H19
 
lincRNA
 is
 a
 developmental
 reservoir
 of
 miR-­‐675
 that
 suppresses
 growth
 and
 Igf1r.
 
Nat
 Cell
 Biol.
 2012;14(7):659-­‐65.
 doi:
 10.1038/ncb2521.
 PubMed
 PMID:
 22684254;
 
PMCID:
 PMC3389517.
 
83.
  Winter
 J,
 Jung
 S,
 Keller
 S,
 Gregory
 RI,
 Diederichs
 S.
 Many
 roads
 to
 maturity:
 
microRNA
 biogenesis
 pathways
 and
 their
 regulation.
 Nat
 Cell
 Biol.
 2009;11(3):228-­‐
34.
 doi:
 10.1038/ncb0309-­‐228.
 PubMed
 PMID:
 19255566.
 
84.
  Carthew
 RW,
 Sontheimer
 EJ.
 Origins
 and
 Mechanisms
 of
 miRNAs
 and
 siRNAs.
 
Cell.
 2009;136(4):642-­‐55.
 doi:
 10.1016/j.cell.2009.01.035.
 PubMed
 PMID:
 
19239886;
 PMCID:
 PMC2675692.
 
85.
  Matera
 AG,
 Terns
 RM,
 Terns
 MP.
 Non-­‐coding
 RNAs:
 lessons
 from
 the
 small
 
nuclear
 and
 small
 nucleolar
 RNAs.
 Nat
 Rev
 Mol
 Cell
 Biol.
 2007;8(3):209-­‐20.
 doi:
 
10.1038/nrm2124.
 PubMed
 PMID:
 17318225.
 
86.
  Luteijn
 MJ,
 Ketting
 RF.
 PIWI-­‐interacting
 RNAs:
 from
 generation
 to
 
transgenerational
 epigenetics.
 Nat
 Rev
 Genet.
 2013;14(8):523-­‐34.
 doi:
 
10.1038/nrg3495.
 PubMed
 PMID:
 23797853.
 
87.
  Kim
 VN,
 Han
 J,
 Siomi
 MC.
 Biogenesis
 of
 small
 RNAs
 in
 animals.
 Nat
 Rev
 Mol
 
Cell
 Biol.
 2009;10(2):126-­‐39.
 doi:
 10.1038/nrm2632.
 PubMed
 PMID:
 19165215.
 
88.
  Lee
 RC,
 Feinbaum
 RL,
 Ambros
 V.
 The
 C.
 elegans
 heterochronic
 gene
 lin-­‐4
 
encodes
 small
 RNAs
 with
 antisense
 complementarity
 to
 lin-­‐14.
 Cell.
 
1993;75(5):843-­‐54.
 PubMed
 PMID:
 8252621.
 
89.
  Berezikov
 E,
 Guryev
 V,
 van
 de
 Belt
 J,
 Wienholds
 E,
 Plasterk
 RH,
 Cuppen
 E.
 
Phylogenetic
 shadowing
 and
 computational
 identification
 of
 human
 microRNA
 
genes.
 Cell.
 2005;120(1):21-­‐4.
 doi:
 10.1016/j.cell.2004.12.031.
 PubMed
 PMID:
 
15652478.
 
90.
  Altuvia
 Y,
 Landgraf
 P,
 Lithwick
 G,
 Elefant
 N,
 Pfeffer
 S,
 Aravin
 A,
 Brownstein
 
MJ,
 Tuschl
 T,
 Margalit
 H.
 Clustering
 and
 conservation
 patterns
 of
 human
 microRNAs.
 
Nucleic
 Acids
 Res.
 2005;33(8):2697-­‐706.
 doi:
 10.1093/nar/gki567.
 PubMed
 PMID:
 
15891114;
 PMCID:
 PMC1110742.
 
91.
  Ibanez-­‐Ventoso
 C,
 Vora
 M,
 Driscoll
 M.
 Sequence
 relationships
 among
 C.
 
elegans,
 D.
 melanogaster
 and
 human
 microRNAs
 highlight
 the
 extensive
 
conservation
 of
 microRNAs
 in
 biology.
 PLoS
 One.
 2008;3(7):e2818.
 doi:
 
10.1371/journal.pone.0002818.
 PubMed
 PMID:
 18665242;
 PMCID:
 PMC2486268.
 
92.
  Lai
 EC.
 Micro
 RNAs
 are
 complementary
 to
 3'
 UTR
 sequence
 motifs
 that
 
mediate
 negative
 post-­‐transcriptional
 regulation.
 Nat
 Genet.
 2002;30(4):363-­‐4.
 doi:
 
10.1038/ng865.
 PubMed
 PMID:
 11896390.
 
111
 

 
93.
  Lytle
 JR,
 Yario
 TA,
 Steitz
 JA.
 Target
 mRNAs
 are
 repressed
 as
 efficiently
 by
 
microRNA-­‐binding
 sites
 in
 the
 5'
 UTR
 as
 in
 the
 3'
 UTR.
 Proc
 Natl
 Acad
 Sci
 U
 S
 A.
 
2007;104(23):9667-­‐72.
 doi:
 10.1073/pnas.0703820104.
 PubMed
 PMID:
 17535905;
 
PMCID:
 PMC1887587.
 
94.
  Krek
 A,
 Grun
 D,
 Poy
 MN,
 Wolf
 R,
 Rosenberg
 L,
 Epstein
 EJ,
 MacMenamin
 P,
 da
 
Piedade
 I,
 Gunsalus
 KC,
 Stoffel
 M,
 Rajewsky
 N.
 Combinatorial
 microRNA
 target
 
predictions.
 Nat
 Genet.
 2005;37(5):495-­‐500.
 doi:
 10.1038/ng1536.
 PubMed
 PMID:
 
15806104.
 
95.
  Friedman
 RC,
 Farh
 KK,
 Burge
 CB,
 Bartel
 DP.
 Most
 mammalian
 mRNAs
 are
 
conserved
 targets
 of
 microRNAs.
 Genome
 Res.
 2009;19(1):92-­‐105.
 doi:
 
10.1101/gr.082701.108.
 PubMed
 PMID:
 18955434;
 PMCID:
 PMC2612969.
 
96.
  Sayed
 D,
 Abdellatif
 M.
 MicroRNAs
 in
 development
 and
 disease.
 Physiol
 Rev.
 
2011;91(3):827-­‐87.
 doi:
 10.1152/physrev.00006.2010.
 PubMed
 PMID:
 21742789.
 
97.
  Wienholds
 E,
 Plasterk
 RH.
 MicroRNA
 function
 in
 animal
 development.
 FEBS
 
Lett.
 2005;579(26):5911-­‐22.
 doi:
 10.1016/j.febslet.2005.07.070.
 PubMed
 PMID:
 
16111679.
 
98.
  Alvarez-­‐Garcia
 I,
 Miska
 EA.
 MicroRNA
 functions
 in
 animal
 development
 and
 
human
 disease.
 Development.
 2005;132(21):4653-­‐62.
 doi:
 10.1242/dev.02073.
 
PubMed
 PMID:
 16224045.
 
99.
  Elbashir
 SM,
 Martinez
 J,
 Patkaniowska
 A,
 Lendeckel
 W,
 Tuschl
 T.
 Functional
 
anatomy
 of
 siRNAs
 for
 mediating
 efficient
 RNAi
 in
 Drosophila
 melanogaster
 embryo
 
lysate.
 EMBO
 J.
 2001;20(23):6877-­‐88.
 doi:
 10.1093/emboj/20.23.6877.
 PubMed
 
PMID:
 11726523;
 PMCID:
 PMC125328.
 
100.
  Guo
 H,
 Ingolia
 NT,
 Weissman
 JS,
 Bartel
 DP.
 Mammalian
 microRNAs
 
predominantly
 act
 to
 decrease
 target
 mRNA
 levels.
 Nature.
 2010;466(7308):835-­‐40.
 
doi:
 10.1038/nature09267.
 PubMed
 PMID:
 20703300;
 PMCID:
 PMC2990499.
 
101.
  Martinez
 J,
 Patkaniowska
 A,
 Urlaub
 H,
 Luhrmann
 R,
 Tuschl
 T.
 Single-­‐stranded
 
antisense
 siRNAs
 guide
 target
 RNA
 cleavage
 in
 RNAi.
 Cell.
 2002;110(5):563-­‐74.
 
PubMed
 PMID:
 12230974.
 
102.
  Hutvagner
 G,
 Zamore
 PD.
 A
 microRNA
 in
 a
 multiple-­‐turnover
 RNAi
 enzyme
 
complex.
 Science.
 2002;297(5589):2056-­‐60.
 doi:
 10.1126/science.1073827.
 
PubMed
 PMID:
 12154197.
 
103.
  Wang
 XH,
 Aliyari
 R,
 Li
 WX,
 Li
 HW,
 Kim
 K,
 Carthew
 R,
 Atkinson
 P,
 Ding
 SW.
 
RNA
 interference
 directs
 innate
 immunity
 against
 viruses
 in
 adult
 Drosophila.
 
Science.
 2006;312(5772):452-­‐4.
 doi:
 10.1126/science.1125694.
 PubMed
 PMID:
 
16556799;
 PMCID:
 PMC1509097.
 
104.
  Li
 H,
 Li
 WX,
 Ding
 SW.
 Induction
 and
 suppression
 of
 RNA
 silencing
 by
 an
 
animal
 virus.
 Science.
 2002;296(5571):1319-­‐21.
 doi:
 10.1126/science.1070948.
 
PubMed
 PMID:
 12016316.
 
105.
  Chung
 WJ,
 Okamura
 K,
 Martin
 R,
 Lai
 EC.
 Endogenous
 RNA
 interference
 
provides
 a
 somatic
 defense
 against
 Drosophila
 transposons.
 Curr
 Biol.
 
2008;18(11):795-­‐802.
 doi:
 10.1016/j.cub.2008.05.006.
 PubMed
 PMID:
 18501606;
 
PMCID:
 PMC2812477.
 
106.
  Babiarz
 JE,
 Ruby
 JG,
 Wang
 Y,
 Bartel
 DP,
 Blelloch
 R.
 Mouse
 ES
 cells
 express
 
endogenous
 shRNAs,
 siRNAs,
 and
 other
 Microprocessor-­‐independent,
 Dicer-­‐
112
 

 
dependent
 small
 RNAs.
 Genes
 Dev.
 2008;22(20):2773-­‐85.
 doi:
 
10.1101/gad.1705308.
 PubMed
 PMID:
 18923076;
 PMCID:
 PMC2569885.
 
107.
  Watanabe
 T,
 Totoki
 Y,
 Toyoda
 A,
 Kaneda
 M,
 Kuramochi-­‐Miyagawa
 S,
 Obata
 Y,
 
Chiba
 H,
 Kohara
 Y,
 Kono
 T,
 Nakano
 T,
 Surani
 MA,
 Sakaki
 Y,
 Sasaki
 H.
 Endogenous
 
siRNAs
 from
 naturally
 formed
 dsRNAs
 regulate
 transcripts
 in
 mouse
 oocytes.
 
Nature.
 2008;453(7194):539-­‐43.
 doi:
 10.1038/nature06908.
 PubMed
 PMID:
 
18404146.
 
108.
  Tam
 OH,
 Aravin
 AA,
 Stein
 P,
 Girard
 A,
 Murchison
 EP,
 Cheloufi
 S,
 Hodges
 E,
 
Anger
 M,
 Sachidanandam
 R,
 Schultz
 RM,
 Hannon
 GJ.
 Pseudogene-­‐derived
 small
 
interfering
 RNAs
 regulate
 gene
 expression
 in
 mouse
 oocytes.
 Nature.
 
2008;453(7194):534-­‐8.
 doi:
 10.1038/nature06904.
 PubMed
 PMID:
 18404147;
 
PMCID:
 PMC2981145.
 
109.
  Song
 R,
 Hennig
 GW,
 Wu
 Q,
 Jose
 C,
 Zheng
 H,
 Yan
 W.
 Male
 germ
 cells
 express
 
abundant
 endogenous
 siRNAs.
 Proc
 Natl
 Acad
 Sci
 U
 S
 A.
 2011;108(32):13159-­‐64.
 
doi:
 10.1073/pnas.1108567108.
 PubMed
 PMID:
 21788498;
 PMCID:
 PMC3156200.
 
110.
  Yang
 N,
 Kazazian
 HH,
 Jr.
 L1
 retrotransposition
 is
 suppressed
 by
 
endogenously
 encoded
 small
 interfering
 RNAs
 in
 human
 cultured
 cells.
 Nat
 Struct
 
Mol
 Biol.
 2006;13(9):763-­‐71.
 doi:
 10.1038/nsmb1141.
 PubMed
 PMID:
 16936727.
 
111.
  Xia
 J,
 Joyce
 CE,
 Bowcock
 AM,
 Zhang
 W.
 Noncanonical
 microRNAs
 and
 
endogenous
 siRNAs
 in
 normal
 and
 psoriatic
 human
 skin.
 Hum
 Mol
 Genet.
 
2013;22(4):737-­‐48.
 doi:
 10.1093/hmg/dds481.
 PubMed
 PMID:
 23175445;
 PMCID:
 
PMC3554200.
 
112.
  Chan
 WL,
 Yuo
 CY,
 Yang
 WK,
 Hung
 SY,
 Chang
 YS,
 Chiu
 CC,
 Yeh
 KT,
 Huang
 HD,
 
Chang
 JG.
 Transcribed
 pseudogene
 psiPPM1K
 generates
 endogenous
 siRNA
 to
 
suppress
 oncogenic
 cell
 growth
 in
 hepatocellular
 carcinoma.
 Nucleic
 Acids
 Res.
 
2013;41(6):3734-­‐47.
 doi:
 10.1093/nar/gkt047.
 PubMed
 PMID:
 23376929;
 PMCID:
 
PMC3616710.
 
113.
  Burnett
 JC,
 Rossi
 JJ.
 RNA-­‐based
 therapeutics:
 current
 progress
 and
 future
 
prospects.
 Chem
 Biol.
 2012;19(1):60-­‐71.
 doi:
 10.1016/j.chembiol.2011.12.008.
 
PubMed
 PMID:
 22284355;
 PMCID:
 PMC3269031.
 
114.
  Ponting
 CP,
 Oliver
 PL,
 Reik
 W.
 Evolution
 and
 functions
 of
 long
 noncoding
 
RNAs.
 Cell.
 2009;136(4):629-­‐41.
 doi:
 10.1016/j.cell.2009.02.006.
 PubMed
 PMID:
 
19239885.
 
115.
  Nagano
 T,
 Fraser
 P.
 No-­‐nonsense
 functions
 for
 long
 noncoding
 RNAs.
 Cell.
 
2011;145(2):178-­‐81.
 doi:
 10.1016/j.cell.2011.03.014.
 PubMed
 PMID:
 21496640.
 
116.
  Rinn
 JL,
 Chang
 HY.
 Genome
 regulation
 by
 long
 noncoding
 RNAs.
 Annu
 Rev
 
Biochem.
 2012;81:145-­‐66.
 doi:
 10.1146/annurev-­‐biochem-­‐051410-­‐092902.
 
PubMed
 PMID:
 22663078;
 PMCID:
 PMC3858397.
 
117.
  Wang
 KC,
 Chang
 HY.
 Molecular
 mechanisms
 of
 long
 noncoding
 RNAs.
 Mol
 
Cell.
 2011;43(6):904-­‐14.
 doi:
 10.1016/j.molcel.2011.08.018.
 PubMed
 PMID:
 
21925379;
 PMCID:
 PMC3199020.
 
118.
  Brockdorff
 N.
 Noncoding
 RNA
 and
 Polycomb
 recruitment.
 RNA.
 
2013;19(4):429-­‐42.
 doi:
 10.1261/rna.037598.112.
 PubMed
 PMID:
 23431328;
 
PMCID:
 PMC3677253.
 
113
 

 
119.
  Lee
 JT,
 Davidow
 LS,
 Warshawsky
 D.
 Tsix,
 a
 gene
 antisense
 to
 Xist
 at
 the
 X-­‐
inactivation
 centre.
 Nat
 Genet.
 1999;21(4):400-­‐4.
 doi:
 10.1038/7734.
 PubMed
 
PMID:
 10192391.
 
120.
  Nesterova
 TB,
 Senner
 CE,
 Schneider
 J,
 Alcayna-­‐Stevens
 T,
 Tattermusch
 A,
 
Hemberger
 M,
 Brockdorff
 N.
 Pluripotency
 factor
 binding
 and
 Tsix
 expression
 act
 
synergistically
 to
 repress
 Xist
 in
 undifferentiated
 embryonic
 stem
 cells.
 Epigenetics
 
Chromatin.
 2011;4(1):17.
 doi:
 10.1186/1756-­‐8935-­‐4-­‐17.
 PubMed
 PMID:
 21982142;
 
PMCID:
 PMC3197471.
 
121.
  Hu
 VW.
 The
 expanding
 genomic
 landscape
 of
 autism:
 discovering
 the
 'forest'
 
beyond
 the
 'trees'.
 Future
 Neurol.
 2013;8(1):29-­‐42.
 doi:
 10.2217/fnl.12.83.
 PubMed
 
PMID:
 23637569;
 PMCID:
 PMC3637978.
 
122.
  Lanford
 RE,
 Hildebrandt-­‐Eriksen
 ES,
 Petri
 A,
 Persson
 R,
 Lindow
 M,
 Munk
 ME,
 
Kauppinen
 S,
 Orum
 H.
 Therapeutic
 silencing
 of
 microRNA-­‐122
 in
 primates
 with
 
chronic
 hepatitis
 C
 virus
 infection.
 Science.
 2010;327(5962):198-­‐201.
 doi:
 
10.1126/science.1178178.
 PubMed
 PMID:
 19965718;
 PMCID:
 PMC3436126.
 
123.
  Castanotto
 D,
 Rossi
 JJ.
 The
 promises
 and
 pitfalls
 of
 RNA-­‐interference-­‐based
 
therapeutics.
 Nature.
 2009;457(7228):426-­‐33.
 doi:
 10.1038/nature07758.
 PubMed
 
PMID:
 19158789;
 PMCID:
 PMC2702667.
 
124.
  Dejneka
 NS,
 Wan
 S,
 Bond
 OS,
 Kornbrust
 DJ,
 Reich
 SJ.
 Ocular
 biodistribution
 of
 
bevasiranib
 following
 a
 single
 intravitreal
 injection
 to
 rabbit
 eyes.
 Mol
 Vis.
 
2008;14:997-­‐1005.
 PubMed
 PMID:
 18523657;
 PMCID:
 PMC2405815.
 
125.
  Cho
 WG,
 Albuquerque
 RJ,
 Kleinman
 ME,
 Tarallo
 V,
 Greco
 A,
 Nozaki
 M,
 Green
 
MG,
 Baffi
 JZ,
 Ambati
 BK,
 De
 Falco
 M,
 Alexander
 JS,
 Brunetti
 A,
 De
 Falco
 S,
 Ambati
 J.
 
Small
 interfering
 RNA-­‐induced
 TLR3
 activation
 inhibits
 blood
 and
 lymphatic
 vessel
 
growth.
 Proc
 Natl
 Acad
 Sci
 U
 S
 A.
 2009;106(17):7137-­‐42.
 doi:
 
10.1073/pnas.0812317106.
 PubMed
 PMID:
 19359485;
 PMCID:
 PMC2678451.
 
126.
  Kleinman
 ME,
 Yamada
 K,
 Takeda
 A,
 Chandrasekaran
 V,
 Nozaki
 M,
 Baffi
 JZ,
 
Albuquerque
 RJ,
 Yamasaki
 S,
 Itaya
 M,
 Pan
 Y,
 Appukuttan
 B,
 Gibbs
 D,
 Yang
 Z,
 Kariko
 
K,
 Ambati
 BK,
 Wilgus
 TA,
 DiPietro
 LA,
 Sakurai
 E,
 Zhang
 K,
 Smith
 JR,
 Taylor
 EW,
 
Ambati
 J.
 Sequence-­‐
 and
 target-­‐independent
 angiogenesis
 suppression
 by
 siRNA
 via
 
TLR3.
 Nature.
 2008;452(7187):591-­‐7.
 doi:
 10.1038/nature06765.
 PubMed
 PMID:
 
18368052;
 PMCID:
 PMC2642938.
 
127.
  Samuel-­‐Abraham
 S,
 Leonard
 JN.
 Staying
 on
 message:
 design
 principles
 for
 
controlling
 nonspecific
 responses
 to
 siRNA.
 FEBS
 J.
 2010;277(23):4828-­‐36.
 doi:
 
10.1111/j.1742-­‐4658.2010.07905.x.
 PubMed
 PMID:
 21078117.
 
128.
  Guo
 Y,
 Srinivasula
 SM,
 Druilhe
 A,
 Fernandes-­‐Alnemri
 T,
 Alnemri
 ES.
 Caspase-­‐
2
 induces
 apoptosis
 by
 releasing
 proapoptotic
 proteins
 from
 mitochondria.
 J
 Biol
 
Chem.
 2002;277(16):13430-­‐7.
 doi:
 10.1074/jbc.M108029200.
 PubMed
 PMID:
 
11832478.
 
129.
  Alvarez
 R,
 Elbashir
 S,
 Borland
 T,
 Toudjarska
 I,
 Hadwiger
 P,
 John
 M,
 Roehl
 I,
 
Morskaya
 SS,
 Martinello
 R,
 Kahn
 J,
 Van
 Ranst
 M,
 Tripp
 RA,
 DeVincenzo
 JP,
 Pandey
 R,
 
Maier
 M,
 Nechev
 L,
 Manoharan
 M,
 Kotelianski
 V,
 Meyers
 R.
 RNA
 interference-­‐
mediated
 silencing
 of
 the
 respiratory
 syncytial
 virus
 nucleocapsid
 defines
 a
 potent
 
antiviral
 strategy.
 Antimicrob
 Agents
 Chemother.
 2009;53(9):3952-­‐62.
 doi:
 
10.1128/AAC.00014-­‐09.
 PubMed
 PMID:
 19506055;
 PMCID:
 PMC2737834.
 
114
 

 
130.
  Bouchard
 PR,
 Hutabarat
 RM,
 Thompson
 KM.
 Discovery
 and
 development
 of
 
therapeutic
 aptamers.
 Annu
 Rev
 Pharmacol
 Toxicol.
 2010;50:237-­‐57.
 doi:
 
10.1146/annurev.pharmtox.010909.105547.
 PubMed
 PMID:
 20055704.
 
131.
  Keefe
 AD,
 Pai
 S,
 Ellington
 A.
 Aptamers
 as
 therapeutics.
 Nat
 Rev
 Drug
 Discov.
 
2010;9(7):537-­‐50.
 doi:
 10.1038/nrd3141.
 PubMed
 PMID:
 20592747.
 
132.
  Sanghvi
 YS.
 A
 status
 update
 of
 modified
 oligonucleotides
 for
 
chemotherapeutics
 applications.
 Curr
 Protoc
 Nucleic
 Acid
 Chem.
 2011;Chapter
 
4:Unit
 4
 1
 -­‐22.
 doi:
 10.1002/0471142700.nc0401s46.
 PubMed
 PMID:
 21901670.
 
133.
  Thiel
 KW,
 Giangrande
 PH.
 Therapeutic
 applications
 of
 DNA
 and
 RNA
 
aptamers.
 Oligonucleotides.
 2009;19(3):209-­‐22.
 doi:
 10.1089/oli.2009.0199.
 
PubMed
 PMID:
 19653880.
 
134.
  Amado
 RG,
 Mitsuyasu
 RT,
 Rosenblatt
 JD,
 Ngok
 FK,
 Bakker
 A,
 Cole
 S,
 Chorn
 N,
 
Lin
 LS,
 Bristol
 G,
 Boyd
 MP,
 MacPherson
 JL,
 Fanning
 GC,
 Todd
 AV,
 Ely
 JA,
 Zack
 JA,
 
Symonds
 GP.
 Anti-­‐human
 immunodeficiency
 virus
 hematopoietic
 progenitor
 cell-­‐
delivered
 ribozyme
 in
 a
 phase
 I
 study:
 myeloid
 and
 lymphoid
 reconstitution
 in
 
human
 immunodeficiency
 virus
 type-­‐1-­‐infected
 patients.
 Hum
 Gene
 Ther.
 
2004;15(3):251-­‐62.
 doi:
 10.1089/104303404322886101.
 PubMed
 PMID:
 
15018734.
 
135.
  Macpherson
 JL,
 Boyd
 MP,
 Arndt
 AJ,
 Todd
 AV,
 Fanning
 GC,
 Ely
 JA,
 Elliott
 F,
 
Knop
 A,
 Raponi
 M,
 Murray
 J,
 Gerlach
 W,
 Sun
 LQ,
 Penny
 R,
 Symonds
 GP,
 Carr
 A,
 
Cooper
 DA.
 Long-­‐term
 survival
 and
 concomitant
 gene
 expression
 of
 ribozyme-­‐
transduced
 CD4+
 T-­‐lymphocytes
 in
 HIV-­‐infected
 patients.
 J
 Gene
 Med.
 
2005;7(5):552-­‐64.
 doi:
 10.1002/jgm.705.
 PubMed
 PMID:
 15655805.
 
136.
  Michienzi
 A,
 Castanotto
 D,
 Lee
 N,
 Li
 S,
 Zaia
 JA,
 Rossi
 JJ.
 RNA-­‐mediated
 
inhibition
 of
 HIV
 in
 a
 gene
 therapy
 setting.
 Ann
 N
 Y
 Acad
 Sci.
 2003;1002:63-­‐71.
 
PubMed
 PMID:
 14751823.
 
137.
  Wong-­‐Staal
 F,
 Poeschla
 EM,
 Looney
 DJ.
 A
 controlled,
 Phase
 1
 clinical
 trial
 to
 
evaluate
 the
 safety
 and
 effects
 in
 HIV-­‐1
 infected
 humans
 of
 autologous
 lymphocytes
 
transduced
 with
 a
 ribozyme
 that
 cleaves
 HIV-­‐1
 RNA.
 Hum
 Gene
 Ther.
 
1998;9(16):2407-­‐25.
 doi:
 10.1089/hum.1998.9.16-­‐2407.
 PubMed
 PMID:
 9829540.
 
138.
  DiGiusto
 DL,
 Krishnan
 A,
 Li
 L,
 Li
 H,
 Li
 S,
 Rao
 A,
 Mi
 S,
 Yam
 P,
 Stinson
 S,
 Kalos
 M,
 
Alvarnas
 J,
 Lacey
 SF,
 Yee
 JK,
 Li
 M,
 Couture
 L,
 Hsu
 D,
 Forman
 SJ,
 Rossi
 JJ,
 Zaia
 JA.
 RNA-­‐
based
 gene
 therapy
 for
 HIV
 with
 lentiviral
 vector-­‐modified
 CD34(+)
 cells
 in
 patients
 
undergoing
 transplantation
 for
 AIDS-­‐related
 lymphoma.
 Sci
 Transl
 Med.
 
2010;2(36):36ra43.
 doi:
 10.1126/scitranslmed.3000931.
 PubMed
 PMID:
 20555022;
 
PMCID:
 PMC3130552.
 
139.
  Li
 MJ,
 Kim
 J,
 Li
 S,
 Zaia
 J,
 Yee
 JK,
 Anderson
 J,
 Akkina
 R,
 Rossi
 JJ.
 Long-­‐term
 
inhibition
 of
 HIV-­‐1
 infection
 in
 primary
 hematopoietic
 cells
 by
 lentiviral
 vector
 
delivery
 of
 a
 triple
 combination
 of
 anti-­‐HIV
 shRNA,
 anti-­‐CCR5
 ribozyme,
 and
 a
 
nucleolar-­‐localizing
 TAR
 decoy.
 Mol
 Ther.
 2005;12(5):900-­‐9.
 doi:
 
10.1016/j.ymthe.2005.07.524.
 PubMed
 PMID:
 16115802.
 
140.
  Levitt
 P,
 Campbell
 DB.
 The
 genetic
 and
 neurobiologic
 compass
 points
 toward
 
common
 signaling
 dysfunctions
 in
 autism
 spectrum
 disorders.
 J
 Clin
 Invest.
 
2009;119(4):747-­‐54.
 doi:
 10.1172/JCI37934.
 PubMed
 PMID:
 19339766;
 PMCID:
 
PMC2662567.
 
115
 

 
141.
  Sahin
 M,
 Sur
 M.
 Genes,
 circuits,
 and
 precision
 therapies
 for
 autism
 and
 
related
 neurodevelopmental
 disorders.
 Science.
 2015;350(6263).
 doi:
 
10.1126/science.aab3897.
 PubMed
 PMID:
 26472761.
 
142.
  Geschwind
 DH,
 State
 MW.
 Gene
 hunting
 in
 autism
 spectrum
 disorder:
 on
 the
 
path
 to
 precision
 medicine.
 Lancet
 Neurol.
 2015;14(11):1109-­‐20.
 doi:
 
10.1016/S1474-­‐4422(15)00044-­‐7.
 PubMed
 PMID:
 25891009.
 
143.
  Klei
 L,
 Sanders
 SJ,
 Murtha
 MT,
 Hus
 V,
 Lowe
 JK,
 Willsey
 AJ,
 Moreno-­‐De-­‐Luca
 D,
 
Yu
 TW,
 Fombonne
 E,
 Geschwind
 D,
 Grice
 DE,
 Ledbetter
 DH,
 Lord
 C,
 Mane
 SM,
 Martin
 
CL,
 Martin
 DM,
 Morrow
 EM,
 Walsh
 CA,
 Melhem
 NM,
 Chaste
 P,
 Sutcliffe
 JS,
 State
 MW,
 
Cook
 EH,
 Jr.,
 Roeder
 K,
 Devlin
 B.
 Common
 genetic
 variants,
 acting
 additively,
 are
 a
 
major
 source
 of
 risk
 for
 autism.
 Mol
 Autism.
 2012;3(1):9.
 doi:
 10.1186/2040-­‐2392-­‐
3-­‐9.
 PubMed
 PMID:
 23067556;
 PMCID:
 PMC3579743.
 
144.
  Chen
 JA,
 Penagarikano
 O,
 Belgard
 TG,
 Swarup
 V,
 Geschwind
 DH.
 The
 
emerging
 picture
 of
 autism
 spectrum
 disorder:
 genetics
 and
 pathology.
 Annu
 Rev
 
Pathol.
 2015;10:111-­‐44.
 doi:
 10.1146/annurev-­‐pathol-­‐012414-­‐040405.
 PubMed
 
PMID:
 25621659.
 
145.
  Furutani
 Y,
 Matsuno
 H,
 Kawasaki
 M,
 Sasaki
 T,
 Mori
 K,
 Yoshihara
 Y.
 
Interaction
 between
 telencephalin
 and
 ERM
 family
 proteins
 mediates
 dendritic
 
filopodia
 formation.
 J
 Neurosci.
 2007;27(33):8866-­‐76.
 doi:
 
10.1523/JNEUROSCI.1047-­‐07.2007.
 PubMed
 PMID:
 17699668.
 
146.
  Mintz
 CD,
 Dickson
 TC,
 Gripp
 ML,
 Salton
 SR,
 Benson
 DL.
 ERMs
 colocalize
 
transiently
 with
 L1
 during
 neocortical
 axon
 outgrowth.
 J
 Comp
 Neurol.
 
2003;464(4):438-­‐48.
 doi:
 10.1002/cne.10809.
 PubMed
 PMID:
 12900915.
 
147.
  Paglini
 G,
 Kunda
 P,
 Quiroga
 S,
 Kosik
 K,
 Caceres
 A.
 Suppression
 of
 radixin
 and
 
moesin
 alters
 growth
 cone
 morphology,
 motility,
 and
 process
 formation
 in
 primary
 
cultured
 neurons.
 J
 Cell
 Biol.
 1998;143(2):443-­‐55.
 PubMed
 PMID:
 9786954;
 PMCID:
 
PMC2132841.
 
148.
  Kim
 HS,
 Bae
 CD,
 Park
 J.
 Glutamate
 receptor-­‐mediated
 phosphorylation
 of
 
ezrin/radixin/moesin
 proteins
 is
 implicated
 in
 filopodial
 protrusion
 of
 primary
 
cultured
 hippocampal
 neuronal
 cells.
 J
 Neurochem.
 2010;113(6):1565-­‐76.
 doi:
 
10.1111/j.1471-­‐4159.2010.06713.x.
 PubMed
 PMID:
 20367752.
 
149.
  Sanchez
 AM,
 Flamini
 MI,
 Fu
 XD,
 Mannella
 P,
 Giretti
 MS,
 Goglia
 L,
 Genazzani
 
AR,
 Simoncini
 T.
 Rapid
 signaling
 of
 estrogen
 to
 WAVE1
 and
 moesin
 controls
 
neuronal
 spine
 formation
 via
 the
 actin
 cytoskeleton.
 Mol
 Endocrinol.
 
2009;23(8):1193-­‐202.
 doi:
 10.1210/me.2008-­‐0408.
 PubMed
 PMID:
 19460862.
 
150.
  Courchesne
 E,
 Pierce
 K.
 Why
 the
 frontal
 cortex
 in
 autism
 might
 be
 talking
 
only
 to
 itself:
 local
 over-­‐connectivity
 but
 long-­‐distance
 disconnection.
 Curr
 Opin
 
Neurobiol.
 2005;15(2):225-­‐30.
 doi:
 10.1016/j.conb.2005.03.001.
 PubMed
 PMID:
 
15831407.
 
151.
  Hardan
 AY,
 Pabalan
 M,
 Gupta
 N,
 Bansal
 R,
 Melhem
 NM,
 Fedorov
 S,
 Keshavan
 
MS,
 Minshew
 NJ.
 Corpus
 callosum
 volume
 in
 children
 with
 autism.
 Psychiatry
 Res.
 
2009;174(1):57-­‐61.
 doi:
 10.1016/j.pscychresns.2009.03.005.
 PubMed
 PMID:
 
19781917;
 PMCID:
 PMC2761427.
 
152.
  Frazier
 TW,
 Hardan
 AY.
 A
 meta-­‐analysis
 of
 the
 corpus
 callosum
 in
 autism.
 
Biol
 Psychiatry.
 2009;66(10):935-­‐41.
 doi:
 10.1016/j.biopsych.2009.07.022.
 PubMed
 
PMID:
 19748080;
 PMCID:
 PMC2783565.
 
116
 

 
153.
  Brito
 AR,
 Vasconcelos
 MM,
 Domingues
 RC,
 Hygino
 da
 Cruz
 LC,
 Jr.,
 Rodrigues
 
Lde
 S,
 Gasparetto
 EL,
 Calcada
 CA.
 Diffusion
 tensor
 imaging
 findings
 in
 school-­‐aged
 
autistic
 children.
 J
 Neuroimaging.
 2009;19(4):337-­‐43.
 doi:
 10.1111/j.1552-­‐
6569.2009.00366.x.
 PubMed
 PMID:
 19490374.
 
154.
  Monk
 CS,
 Peltier
 SJ,
 Wiggins
 JL,
 Weng
 SJ,
 Carrasco
 M,
 Risi
 S,
 Lord
 C.
 
Abnormalities
 of
 intrinsic
 functional
 connectivity
 in
 autism
 spectrum
 disorders.
 
Neuroimage.
 2009;47(2):764-­‐72.
 doi:
 10.1016/j.neuroimage.2009.04.069.
 PubMed
 
PMID:
 19409498;
 PMCID:
 PMC2731579.
 
155.
  Minshew
 NJ,
 Keller
 TA.
 The
 nature
 of
 brain
 dysfunction
 in
 autism:
 functional
 
brain
 imaging
 studies.
 Curr
 Opin
 Neurol.
 2010;23(2):124-­‐30.
 doi:
 
10.1097/WCO.0b013e32833782d4.
 PubMed
 PMID:
 20154614;
 PMCID:
 
PMC2975255.
 
156.
  Trapnell
 C,
 Pachter
 L,
 Salzberg
 SL.
 TopHat:
 discovering
 splice
 junctions
 with
 
RNA-­‐Seq.
 Bioinformatics.
 2009;25(9):1105-­‐11.
 doi:
 
10.1093/bioinformatics/btp120.
 PubMed
 PMID:
 19289445;
 PMCID:
 PMC2672628.
 
157.
  Li
 H,
 Handsaker
 B,
 Wysoker
 A,
 Fennell
 T,
 Ruan
 J,
 Homer
 N,
 Marth
 G,
 Abecasis
 
G,
 Durbin
 R,
 Genome
 Project
 Data
 Processing
 S.
 The
 Sequence
 Alignment/Map
 
format
 and
 SAMtools.
 Bioinformatics.
 2009;25(16):2078-­‐9.
 doi:
 
10.1093/bioinformatics/btp352.
 PubMed
 PMID:
 19505943;
 PMCID:
 PMC2723002.
 
158.
  Trapnell
 C,
 Williams
 BA,
 Pertea
 G,
 Mortazavi
 A,
 Kwan
 G,
 van
 Baren
 MJ,
 
Salzberg
 SL,
 Wold
 BJ,
 Pachter
 L.
 Transcript
 assembly
 and
 quantification
 by
 RNA-­‐Seq
 
reveals
 unannotated
 transcripts
 and
 isoform
 switching
 during
 cell
 differentiation.
 
Nat
 Biotechnol.
 2010;28(5):511-­‐5.
 doi:
 10.1038/nbt.1621.
 PubMed
 PMID:
 
20436464;
 PMCID:
 PMC3146043.
 
159.
  Huang
 da
 W,
 Sherman
 BT,
 Lempicki
 RA.
 Systematic
 and
 integrative
 analysis
 
of
 large
 gene
 lists
 using
 DAVID
 bioinformatics
 resources.
 Nat
 Protoc.
 2009;4(1):44-­‐
57.
 doi:
 10.1038/nprot.2008.211.
 PubMed
 PMID:
 19131956.
 
160.
  Huang
 da
 W,
 Sherman
 BT,
 Lempicki
 RA.
 Bioinformatics
 enrichment
 tools:
 
paths
 toward
 the
 comprehensive
 functional
 analysis
 of
 large
 gene
 lists.
 Nucleic
 
Acids
 Res.
 2009;37(1):1-­‐13.
 doi:
 10.1093/nar/gkn923.
 PubMed
 PMID:
 19033363;
 
PMCID:
 PMC2615629.
 
161.
  Huguet
 G,
 Ey
 E,
 Bourgeron
 T.
 The
 genetic
 landscapes
 of
 autism
 spectrum
 
disorders.
 Annu
 Rev
 Genomics
 Hum
 Genet.
 2013;14:191-­‐213.
 doi:
 
10.1146/annurev-­‐genom-­‐091212-­‐153431.
 PubMed
 PMID:
 23875794.
 
162.
  De
 Rubeis
 S,
 He
 X,
 Goldberg
 AP,
 Poultney
 CS,
 Samocha
 K,
 Cicek
 AE,
 Kou
 Y,
 Liu
 
L,
 Fromer
 M,
 Walker
 S,
 Singh
 T,
 Klei
 L,
 Kosmicki
 J,
 Shih-­‐Chen
 F,
 Aleksic
 B,
 Biscaldi
 M,
 
Bolton
 PF,
 Brownfeld
 JM,
 Cai
 J,
 Campbell
 NG,
 Carracedo
 A,
 Chahrour
 MH,
 Chiocchetti
 
AG,
 Coon
 H,
 Crawford
 EL,
 Curran
 SR,
 Dawson
 G,
 Duketis
 E,
 Fernandez
 BA,
 Gallagher
 
L,
 Geller
 E,
 Guter
 SJ,
 Hill
 RS,
 Ionita-­‐Laza
 J,
 Jimenz
 Gonzalez
 P,
 Kilpinen
 H,
 Klauck
 SM,
 
Kolevzon
 A,
 Lee
 I,
 Lei
 I,
 Lei
 J,
 Lehtimaki
 T,
 Lin
 CF,
 Ma'ayan
 A,
 Marshall
 CR,
 McInnes
 
AL,
 Neale
 B,
 Owen
 MJ,
 Ozaki
 N,
 Parellada
 M,
 Parr
 JR,
 Purcell
 S,
 Puura
 K,
 Rajagopalan
 
D,
 Rehnstrom
 K,
 Reichenberg
 A,
 Sabo
 A,
 Sachse
 M,
 Sanders
 SJ,
 Schafer
 C,
 Schulte-­‐
Ruther
 M,
 Skuse
 D,
 Stevens
 C,
 Szatmari
 P,
 Tammimies
 K,
 Valladares
 O,
 Voran
 A,
 Li-­‐
San
 W,
 Weiss
 LA,
 Willsey
 AJ,
 Yu
 TW,
 Yuen
 RK,
 Study
 DDD,
 Homozygosity
 Mapping
 
Collaborative
 for
 A,
 Consortium
 UK,
 Cook
 EH,
 Freitag
 CM,
 Gill
 M,
 Hultman
 CM,
 
Lehner
 T,
 Palotie
 A,
 Schellenberg
 GD,
 Sklar
 P,
 State
 MW,
 Sutcliffe
 JS,
 Walsh
 CA,
 
117
 

 
Scherer
 SW,
 Zwick
 ME,
 Barett
 JC,
 Cutler
 DJ,
 Roeder
 K,
 Devlin
 B,
 Daly
 MJ,
 Buxbaum
 JD.
 
Synaptic,
 transcriptional
 and
 chromatin
 genes
 disrupted
 in
 autism.
 Nature.
 
2014;515(7526):209-­‐15.
 doi:
 10.1038/nature13772.
 PubMed
 PMID:
 25363760;
 
PMCID:
 PMC4402723.
 
163.
  Marsick
 BM,
 San
 Miguel-­‐Ruiz
 JE,
 Letourneau
 PC.
 Activation
 of
 
ezrin/radixin/moesin
 mediates
 attractive
 growth
 cone
 guidance
 through
 regulation
 
of
 growth
 cone
 actin
 and
 adhesion
 receptors.
 J
 Neurosci.
 2012;32(1):282-­‐96.
 doi:
 
10.1523/JNEUROSCI.4794-­‐11.2012.
 PubMed
 PMID:
 22219290;
 PMCID:
 
PMC3306234.
 
164.
  Matsumoto
 Y,
 Inden
 M,
 Tamura
 A,
 Hatano
 R,
 Tsukita
 S,
 Asano
 S.
 Ezrin
 
mediates
 neuritogenesis
 via
 down-­‐regulation
 of
 RhoA
 activity
 in
 cultured
 cortical
 
neurons.
 PLoS
 One.
 2014;9(8):e105435.
 doi:
 10.1371/journal.pone.0105435.
 
PubMed
 PMID:
 25144196;
 PMCID:
 PMC4140760.
 
165.
  Voineagu
 I,
 Wang
 X,
 Johnston
 P,
 Lowe
 JK,
 Tian
 Y,
 Horvath
 S,
 Mill
 J,
 Cantor
 RM,
 
Blencowe
 BJ,
 Geschwind
 DH.
 Transcriptomic
 analysis
 of
 autistic
 brain
 reveals
 
convergent
 molecular
 pathology.
 Nature.
 2011;474(7351):380-­‐4.
 doi:
 
10.1038/nature10110.
 PubMed
 PMID:
 21614001;
 PMCID:
 PMC3607626.
 
166.
  Belmonte
 MK,
 Allen
 G,
 Beckel-­‐Mitchener
 A,
 Boulanger
 LM,
 Carper
 RA,
 Webb
 
SJ.
 Autism
 and
 abnormal
 development
 of
 brain
 connectivity.
 J
 Neurosci.
 
2004;24(42):9228-­‐31.
 doi:
 10.1523/JNEUROSCI.3340-­‐04.2004.
 PubMed
 PMID:
 
15496656.
 
167.
  Carper
 RA,
 Courchesne
 E.
 Localized
 enlargement
 of
 the
 frontal
 cortex
 in
 early
 
autism.
 Biol
 Psychiatry.
 2005;57(2):126-­‐33.
 doi:
 10.1016/j.biopsych.2004.11.005.
 
PubMed
 PMID:
 15652870.
 
168.
  Lander
 ES,
 Linton
 LM,
 Birren
 B,
 Nusbaum
 C,
 Zody
 MC,
 Baldwin
 J,
 Devon
 K,
 
Dewar
 K,
 Doyle
 M,
 FitzHugh
 W,
 Funke
 R,
 Gage
 D,
 Harris
 K,
 Heaford
 A,
 Howland
 J,
 
Kann
 L,
 Lehoczky
 J,
 LeVine
 R,
 McEwan
 P,
 McKernan
 K,
 Meldrim
 J,
 Mesirov
 JP,
 
Miranda
 C,
 Morris
 W,
 Naylor
 J,
 Raymond
 C,
 Rosetti
 M,
 Santos
 R,
 Sheridan
 A,
 Sougnez
 
C,
 Stange-­‐Thomann
 Y,
 Stojanovic
 N,
 Subramanian
 A,
 Wyman
 D,
 Rogers
 J,
 Sulston
 J,
 
Ainscough
 R,
 Beck
 S,
 Bentley
 D,
 Burton
 J,
 Clee
 C,
 Carter
 N,
 Coulson
 A,
 Deadman
 R,
 
Deloukas
 P,
 Dunham
 A,
 Dunham
 I,
 Durbin
 R,
 French
 L,
 Grafham
 D,
 Gregory
 S,
 
Hubbard
 T,
 Humphray
 S,
 Hunt
 A,
 Jones
 M,
 Lloyd
 C,
 McMurray
 A,
 Matthews
 L,
 Mercer
 
S,
 Milne
 S,
 Mullikin
 JC,
 Mungall
 A,
 Plumb
 R,
 Ross
 M,
 Shownkeen
 R,
 Sims
 S,
 Waterston
 
RH,
 Wilson
 RK,
 Hillier
 LW,
 McPherson
 JD,
 Marra
 MA,
 Mardis
 ER,
 Fulton
 LA,
 
Chinwalla
 AT,
 Pepin
 KH,
 Gish
 WR,
 Chissoe
 SL,
 Wendl
 MC,
 Delehaunty
 KD,
 Miner
 TL,
 
Delehaunty
 A,
 Kramer
 JB,
 Cook
 LL,
 Fulton
 RS,
 Johnson
 DL,
 Minx
 PJ,
 Clifton
 SW,
 
Hawkins
 T,
 Branscomb
 E,
 Predki
 P,
 Richardson
 P,
 Wenning
 S,
 Slezak
 T,
 Doggett
 N,
 
Cheng
 JF,
 Olsen
 A,
 Lucas
 S,
 Elkin
 C,
 Uberbacher
 E,
 Frazier
 M,
 Gibbs
 RA,
 Muzny
 DM,
 
Scherer
 SE,
 Bouck
 JB,
 Sodergren
 EJ,
 Worley
 KC,
 Rives
 CM,
 Gorrell
 JH,
 Metzker
 ML,
 
Naylor
 SL,
 Kucherlapati
 RS,
 Nelson
 DL,
 Weinstock
 GM,
 Sakaki
 Y,
 Fujiyama
 A,
 Hattori
 
M,
 Yada
 T,
 Toyoda
 A,
 Itoh
 T,
 Kawagoe
 C,
 Watanabe
 H,
 Totoki
 Y,
 Taylor
 T,
 
Weissenbach
 J,
 Heilig
 R,
 Saurin
 W,
 Artiguenave
 F,
 Brottier
 P,
 Bruls
 T,
 Pelletier
 E,
 
Robert
 C,
 Wincker
 P,
 Smith
 DR,
 Doucette-­‐Stamm
 L,
 Rubenfield
 M,
 Weinstock
 K,
 Lee
 
HM,
 Dubois
 J,
 Rosenthal
 A,
 Platzer
 M,
 Nyakatura
 G,
 Taudien
 S,
 Rump
 A,
 Yang
 H,
 Yu
 J,
 
Wang
 J,
 Huang
 G,
 Gu
 J,
 Hood
 L,
 Rowen
 L,
 Madan
 A,
 Qin
 S,
 Davis
 RW,
 Federspiel
 NA,
 
Abola
 AP,
 Proctor
 MJ,
 Myers
 RM,
 Schmutz
 J,
 Dickson
 M,
 Grimwood
 J,
 Cox
 DR,
 Olson
 
118
 

 
MV,
 Kaul
 R,
 Raymond
 C,
 Shimizu
 N,
 Kawasaki
 K,
 Minoshima
 S,
 Evans
 GA,
 Athanasiou
 
M,
 Schultz
 R,
 Roe
 BA,
 Chen
 F,
 Pan
 H,
 Ramser
 J,
 Lehrach
 H,
 Reinhardt
 R,
 McCombie
 
WR,
 de
 la
 Bastide
 M,
 Dedhia
 N,
 Blocker
 H,
 Hornischer
 K,
 Nordsiek
 G,
 Agarwala
 R,
 
Aravind
 L,
 Bailey
 JA,
 Bateman
 A,
 Batzoglou
 S,
 Birney
 E,
 Bork
 P,
 Brown
 DG,
 Burge
 CB,
 
Cerutti
 L,
 Chen
 HC,
 Church
 D,
 Clamp
 M,
 Copley
 RR,
 Doerks
 T,
 Eddy
 SR,
 Eichler
 EE,
 
Furey
 TS,
 Galagan
 J,
 Gilbert
 JG,
 Harmon
 C,
 Hayashizaki
 Y,
 Haussler
 D,
 Hermjakob
 H,
 
Hokamp
 K,
 Jang
 W,
 Johnson
 LS,
 Jones
 TA,
 Kasif
 S,
 Kaspryzk
 A,
 Kennedy
 S,
 Kent
 WJ,
 
Kitts
 P,
 Koonin
 EV,
 Korf
 I,
 Kulp
 D,
 Lancet
 D,
 Lowe
 TM,
 McLysaght
 A,
 Mikkelsen
 T,
 
Moran
 JV,
 Mulder
 N,
 Pollara
 VJ,
 Ponting
 CP,
 Schuler
 G,
 Schultz
 J,
 Slater
 G,
 Smit
 AF,
 
Stupka
 E,
 Szustakowki
 J,
 Thierry-­‐Mieg
 D,
 Thierry-­‐Mieg
 J,
 Wagner
 L,
 Wallis
 J,
 Wheeler
 
R,
 Williams
 A,
 Wolf
 YI,
 Wolfe
 KH,
 Yang
 SP,
 Yeh
 RF,
 Collins
 F,
 Guyer
 MS,
 Peterson
 J,
 
Felsenfeld
 A,
 Wetterstrand
 KA,
 Patrinos
 A,
 Morgan
 MJ,
 de
 Jong
 P,
 Catanese
 JJ,
 
Osoegawa
 K,
 Shizuya
 H,
 Choi
 S,
 Chen
 YJ,
 Szustakowki
 J,
 International
 Human
 
Genome
 Sequencing
 C.
 Initial
 sequencing
 and
 analysis
 of
 the
 human
 genome.
 Nature.
 
2001;409(6822):860-­‐921.
 doi:
 10.1038/35057062.
 PubMed
 PMID:
 11237011.
 
169.
  Venter
 JC,
 Adams
 MD,
 Myers
 EW,
 Li
 PW,
 Mural
 RJ,
 Sutton
 GG,
 Smith
 HO,
 
Yandell
 M,
 Evans
 CA,
 Holt
 RA,
 Gocayne
 JD,
 Amanatides
 P,
 Ballew
 RM,
 Huson
 DH,
 
Wortman
 JR,
 Zhang
 Q,
 Kodira
 CD,
 Zheng
 XH,
 Chen
 L,
 Skupski
 M,
 Subramanian
 G,
 
Thomas
 PD,
 Zhang
 J,
 Gabor
 Miklos
 GL,
 Nelson
 C,
 Broder
 S,
 Clark
 AG,
 Nadeau
 J,
 
McKusick
 VA,
 Zinder
 N,
 Levine
 AJ,
 Roberts
 RJ,
 Simon
 M,
 Slayman
 C,
 Hunkapiller
 M,
 
Bolanos
 R,
 Delcher
 A,
 Dew
 I,
 Fasulo
 D,
 Flanigan
 M,
 Florea
 L,
 Halpern
 A,
 Hannenhalli
 
S,
 Kravitz
 S,
 Levy
 S,
 Mobarry
 C,
 Reinert
 K,
 Remington
 K,
 Abu-­‐Threideh
 J,
 Beasley
 E,
 
Biddick
 K,
 Bonazzi
 V,
 Brandon
 R,
 Cargill
 M,
 Chandramouliswaran
 I,
 Charlab
 R,
 
Chaturvedi
 K,
 Deng
 Z,
 Di
 Francesco
 V,
 Dunn
 P,
 Eilbeck
 K,
 Evangelista
 C,
 Gabrielian
 
AE,
 Gan
 W,
 Ge
 W,
 Gong
 F,
 Gu
 Z,
 Guan
 P,
 Heiman
 TJ,
 Higgins
 ME,
 Ji
 RR,
 Ke
 Z,
 Ketchum
 
KA,
 Lai
 Z,
 Lei
 Y,
 Li
 Z,
 Li
 J,
 Liang
 Y,
 Lin
 X,
 Lu
 F,
 Merkulov
 GV,
 Milshina
 N,
 Moore
 HM,
 
Naik
 AK,
 Narayan
 VA,
 Neelam
 B,
 Nusskern
 D,
 Rusch
 DB,
 Salzberg
 S,
 Shao
 W,
 Shue
 B,
 
Sun
 J,
 Wang
 Z,
 Wang
 A,
 Wang
 X,
 Wang
 J,
 Wei
 M,
 Wides
 R,
 Xiao
 C,
 Yan
 C,
 Yao
 A,
 Ye
 J,
 
Zhan
 M,
 Zhang
 W,
 Zhang
 H,
 Zhao
 Q,
 Zheng
 L,
 Zhong
 F,
 Zhong
 W,
 Zhu
 S,
 Zhao
 S,
 
Gilbert
 D,
 Baumhueter
 S,
 Spier
 G,
 Carter
 C,
 Cravchik
 A,
 Woodage
 T,
 Ali
 F,
 An
 H,
 Awe
 
A,
 Baldwin
 D,
 Baden
 H,
 Barnstead
 M,
 Barrow
 I,
 Beeson
 K,
 Busam
 D,
 Carver
 A,
 Center
 
A,
 Cheng
 ML,
 Curry
 L,
 Danaher
 S,
 Davenport
 L,
 Desilets
 R,
 Dietz
 S,
 Dodson
 K,
 Doup
 L,
 
Ferriera
 S,
 Garg
 N,
 Gluecksmann
 A,
 Hart
 B,
 Haynes
 J,
 Haynes
 C,
 Heiner
 C,
 Hladun
 S,
 
Hostin
 D,
 Houck
 J,
 Howland
 T,
 Ibegwam
 C,
 Johnson
 J,
 Kalush
 F,
 Kline
 L,
 Koduru
 S,
 
Love
 A,
 Mann
 F,
 May
 D,
 McCawley
 S,
 McIntosh
 T,
 McMullen
 I,
 Moy
 M,
 Moy
 L,
 Murphy
 
B,
 Nelson
 K,
 Pfannkoch
 C,
 Pratts
 E,
 Puri
 V,
 Qureshi
 H,
 Reardon
 M,
 Rodriguez
 R,
 
Rogers
 YH,
 Romblad
 D,
 Ruhfel
 B,
 Scott
 R,
 Sitter
 C,
 Smallwood
 M,
 Stewart
 E,
 Strong
 R,
 
Suh
 E,
 Thomas
 R,
 Tint
 NN,
 Tse
 S,
 Vech
 C,
 Wang
 G,
 Wetter
 J,
 Williams
 S,
 Williams
 M,
 
Windsor
 S,
 Winn-­‐Deen
 E,
 Wolfe
 K,
 Zaveri
 J,
 Zaveri
 K,
 Abril
 JF,
 Guigo
 R,
 Campbell
 MJ,
 
Sjolander
 KV,
 Karlak
 B,
 Kejariwal
 A,
 Mi
 H,
 Lazareva
 B,
 Hatton
 T,
 Narechania
 A,
 
Diemer
 K,
 Muruganujan
 A,
 Guo
 N,
 Sato
 S,
 Bafna
 V,
 Istrail
 S,
 Lippert
 R,
 Schwartz
 R,
 
Walenz
 B,
 Yooseph
 S,
 Allen
 D,
 Basu
 A,
 Baxendale
 J,
 Blick
 L,
 Caminha
 M,
 Carnes-­‐Stine
 
J,
 Caulk
 P,
 Chiang
 YH,
 Coyne
 M,
 Dahlke
 C,
 Mays
 A,
 Dombroski
 M,
 Donnelly
 M,
 Ely
 D,
 
Esparham
 S,
 Fosler
 C,
 Gire
 H,
 Glanowski
 S,
 Glasser
 K,
 Glodek
 A,
 Gorokhov
 M,
 Graham
 
K,
 Gropman
 B,
 Harris
 M,
 Heil
 J,
 Henderson
 S,
 Hoover
 J,
 Jennings
 D,
 Jordan
 C,
 Jordan
 J,
 
Kasha
 J,
 Kagan
 L,
 Kraft
 C,
 Levitsky
 A,
 Lewis
 M,
 Liu
 X,
 Lopez
 J,
 Ma
 D,
 Majoros
 W,
 
119
 

 
McDaniel
 J,
 Murphy
 S,
 Newman
 M,
 Nguyen
 T,
 Nguyen
 N,
 Nodell
 M,
 Pan
 S,
 Peck
 J,
 
Peterson
 M,
 Rowe
 W,
 Sanders
 R,
 Scott
 J,
 Simpson
 M,
 Smith
 T,
 Sprague
 A,
 Stockwell
 T,
 
Turner
 R,
 Venter
 E,
 Wang
 M,
 Wen
 M,
 Wu
 D,
 Wu
 M,
 Xia
 A,
 Zandieh
 A,
 Zhu
 X.
 The
 
sequence
 of
 the
 human
 genome.
 Science.
 2001;291(5507):1304-­‐51.
 doi:
 
10.1126/science.1058040.
 PubMed
 PMID:
 11181995.
 
170.
  Levine
 M,
 Davidson
 EH.
 Gene
 regulatory
 networks
 for
 development.
 Proc
 
Natl
 Acad
 Sci
 U
 S
 A.
 2005;102(14):4936-­‐42.
 doi:
 10.1073/pnas.0408031102.
 
PubMed
 PMID:
 15788537;
 PMCID:
 PMC555974.
 
171.
  Lee
 TI,
 Young
 RA.
 Transcriptional
 regulation
 and
 its
 misregulation
 in
 disease.
 
Cell.
 2013;152(6):1237-­‐51.
 doi:
 10.1016/j.cell.2013.02.014.
 PubMed
 PMID:
 
23498934;
 PMCID:
 PMC3640494.
 
172.
  Visel
 A,
 Rubin
 EM,
 Pennacchio
 LA.
 Genomic
 views
 of
 distant-­‐acting
 
enhancers.
 Nature.
 2009;461(7261):199-­‐205.
 doi:
 10.1038/nature08451.
 PubMed
 
PMID:
 19741700;
 PMCID:
 PMC2923221.
 
173.
  Akbarian
 S,
 Liu
 C,
 Knowles
 JA,
 Vaccarino
 FM,
 Farnham
 PJ,
 Crawford
 GE,
 Jaffe
 
AE,
 Pinto
 D,
 Dracheva
 S,
 Geschwind
 DH,
 Mill
 J,
 Nairn
 AC,
 Abyzov
 A,
 Pochareddy
 S,
 
Prabhakar
 S,
 Weissman
 S,
 Sullivan
 PF,
 State
 MW,
 Weng
 Z,
 Peters
 MA,
 White
 KP,
 
Gerstein
 MB,
 Amiri
 A,
 Armoskus
 C,
 Ashley-­‐Koch
 AE,
 Bae
 T,
 Beckel-­‐Mitchener
 A,
 
Berman
 BP,
 Coetzee
 GA,
 Coppola
 G,
 Francoeur
 N,
 Fromer
 M,
 Gao
 R,
 Grennan
 K,
 
Herstein
 J,
 Kavanagh
 DH,
 Ivanov
 NA,
 Jiang
 Y,
 Kitchen
 RR,
 Kozlenkov
 A,
 Kundakovic
 
M,
 Li
 M,
 Li
 Z,
 Liu
 S,
 Mangravite
 LM,
 Mattei
 E,
 Markenscoff-­‐Papadimitriou
 E,
 Navarro
 
FC,
 North
 N,
 Omberg
 L,
 Panchision
 D,
 Parikshak
 N,
 Poschmann
 J,
 Price
 AJ,
 Purcaro
 M,
 
Reddy
 TE,
 Roussos
 P,
 Schreiner
 S,
 Scuderi
 S,
 Sebra
 R,
 Shibata
 M,
 Shieh
 AW,
 Skarica
 
M,
 Sun
 W,
 Swarup
 V,
 Thomas
 A,
 Tsuji
 J,
 van
 Bakel
 H,
 Wang
 D,
 Wang
 Y,
 Wang
 K,
 
Werling
 DM,
 Willsey
 AJ,
 Witt
 H,
 Won
 H,
 Wong
 CC,
 Wray
 GA,
 Wu
 EY,
 Xu
 X,
 Yao
 L,
 
Senthil
 G,
 Lehner
 T,
 Sklar
 P,
 Sestan
 N.
 The
 PsychENCODE
 project.
 Nat
 Neurosci.
 
2015;18(12):1707-­‐12.
 doi:
 10.1038/nn.4156.
 PubMed
 PMID:
 26605881.
 
174.
  Wilkinson
 B,
 Campbell
 DB.
 Contribution
 of
 long
 noncoding
 RNAs
 to
 autism
 
spectrum
 disorder
 risk.
 Int
 Rev
 Neurobiol.
 2013;113:35-­‐59.
 doi:
 10.1016/B978-­‐0-­‐
12-­‐418700-­‐9.00002-­‐2.
 PubMed
 PMID:
 24290382.
 
175.
  Barnard
 RA,
 Pomaville
 MB,
 O'Roak
 BJ.
 Mutations
 and
 Modeling
 of
 the
 
Chromatin
 Remodeler
 CHD8
 Define
 an
 Emerging
 Autism
 Etiology.
 Front
 Neurosci.
 
2015;9:477.
 doi:
 10.3389/fnins.2015.00477.
 PubMed
 PMID:
 26733790;
 PMCID:
 
PMC4681771.
 
176.
  Wilkinson
 B,
 Grepo
 N,
 Thompson
 BL,
 Kim
 J,
 Wang
 K,
 Evgrafov
 OV,
 Lu
 W,
 
Knowles
 JA,
 Campbell
 DB.
 The
 autism-­‐associated
 gene
 chromodomain
 helicase
 DNA-­‐
binding
 protein
 8
 (CHD8)
 regulates
 noncoding
 RNAs
 and
 autism-­‐related
 genes.
 
Transl
 Psychiatry.
 2015;5:e568.
 doi:
 10.1038/tp.2015.62.
 PubMed
 PMID:
 25989142;
 
PMCID:
 PMC4471293.
 
177.
  Jiang
 J,
 Jing
 Y,
 Cost
 GJ,
 Chiang
 JC,
 Kolpa
 HJ,
 Cotton
 AM,
 Carone
 DM,
 Carone
 BR,
 
Shivak
 DA,
 Guschin
 DY,
 Pearl
 JR,
 Rebar
 EJ,
 Byron
 M,
 Gregory
 PD,
 Brown
 CJ,
 Urnov
 
FD,
 Hall
 LL,
 Lawrence
 JB.
 Translating
 dosage
 compensation
 to
 trisomy
 21.
 Nature.
 
2013;500(7462):296-­‐300.
 doi:
 10.1038/nature12394.
 PubMed
 PMID:
 23863942;
 
PMCID:
 PMC3848249.
 
178.
  Turner
 AM,
 Ackley
 AM,
 Matrone
 MA,
 Morris
 KV.
 Characterization
 of
 an
 HIV-­‐
targeted
 transcriptional
 gene-­‐silencing
 RNA
 in
 primary
 cells.
 Hum
 Gene
 Ther.
 
120
 

 
2012;23(5):473-­‐83.
 doi:
 10.1089/hum.2011.165.
 PubMed
 PMID:
 22122263;
 PMCID:
 
PMC3360501.
 
179.
  Hawkins
 PG,
 Morris
 KV.
 Transcriptional
 regulation
 of
 Oct4
 by
 a
 long
 non-­‐
coding
 RNA
 antisense
 to
 Oct4-­‐pseudogene
 5.
 Transcription.
 2010;1(3):165-­‐75.
 doi:
 
10.4161/trns.1.3.13332.
 PubMed
 PMID:
 21151833;
 PMCID:
 PMC2999937.
 
180.
  Hawkins
 PG,
 Santoso
 S,
 Adams
 C,
 Anest
 V,
 Morris
 KV.
 Promoter
 targeted
 
small
 RNAs
 induce
 long-­‐term
 transcriptional
 gene
 silencing
 in
 human
 cells.
 Nucleic
 
Acids
 Res.
 2009;37(9):2984-­‐95.
 doi:
 10.1093/nar/gkp127.
 PubMed
 PMID:
 
19304753;
 PMCID:
 PMC2685082.
 
181.
  Knowling
 S,
 Morris
 KV.
 Non-­‐coding
 RNA
 and
 antisense
 RNA.
 Nature's
 trash
 or
 
treasure?
 Biochimie.
 2011;93(11):1922-­‐7.
 doi:
 10.1016/j.biochi.2011.07.031.
 
PubMed
 PMID:
 21843589;
 PMCID:
 PMC3195647.
 
182.
  Abrahams
 BS,
 Arking
 DE,
 Campbell
 DB,
 Mefford
 HC,
 Morrow
 EM,
 Weiss
 LA,
 
Menashe
 I,
 Wadkins
 T,
 Banerjee-­‐Basu
 S,
 Packer
 A.
 SFARI
 Gene
 2.0:
 a
 community-­‐
driven
 knowledgebase
 for
 the
 autism
 spectrum
 disorders
 (ASDs).
 Mol
 Autism.
 
2013;4(1):36.
 doi:
 10.1186/2040-­‐2392-­‐4-­‐36.
 PubMed
 PMID:
 24090431;
 PMCID:
 
PMC3851189.
 
183.
  Lee
 H,
 Lin
 MC,
 Kornblum
 HI,
 Papazian
 DM,
 Nelson
 SF.
 Exome
 sequencing
 
identifies
 de
 novo
 gain
 of
 function
 missense
 mutation
 in
 KCND2
 in
 identical
 twins
 
with
 autism
 and
 seizures
 that
 slows
 potassium
 channel
 inactivation.
 Hum
 Mol
 
Genet.
 2014;23(13):3481-­‐9.
 doi:
 10.1093/hmg/ddu056.
 PubMed
 PMID:
 24501278;
 
PMCID:
 PMC4049306.
 
184.
  Cabili
 MN,
 Trapnell
 C,
 Goff
 L,
 Koziol
 M,
 Tazon-­‐Vega
 B,
 Regev
 A,
 Rinn
 JL.
 
Integrative
 annotation
 of
 human
 large
 intergenic
 noncoding
 RNAs
 reveals
 global
 
properties
 and
 specific
 subclasses.
 Genes
 Dev.
 2011;25(18):1915-­‐27.
 doi:
 
10.1101/gad.17446611.
 PubMed
 PMID:
 21890647;
 PMCID:
 PMC3185964.
 
185.
  Guttman
 M,
 Rinn
 JL.
 Modular
 regulatory
 principles
 of
 large
 non-­‐coding
 RNAs.
 
Nature.
 2012;482(7385):339-­‐46.
 doi:
 10.1038/nature10887.
 PubMed
 PMID:
 
22337053;
 PMCID:
 PMC4197003.
 
186.
  Ravasi
 T,
 Suzuki
 H,
 Pang
 KC,
 Katayama
 S,
 Furuno
 M,
 Okunishi
 R,
 Fukuda
 S,
 Ru
 
K,
 Frith
 MC,
 Gongora
 MM,
 Grimmond
 SM,
 Hume
 DA,
 Hayashizaki
 Y,
 Mattick
 JS.
 
Experimental
 validation
 of
 the
 regulated
 expression
 of
 large
 numbers
 of
 non-­‐coding
 
RNAs
 from
 the
 mouse
 genome.
 Genome
 Res.
 2006;16(1):11-­‐9.
 doi:
 
10.1101/gr.4200206.
 PubMed
 PMID:
 16344565;
 PMCID:
 PMC1356124.
 
187.
  Derrien
 T,
 Johnson
 R,
 Bussotti
 G,
 Tanzer
 A,
 Djebali
 S,
 Tilgner
 H,
 Guernec
 G,
 
Martin
 D,
 Merkel
 A,
 Knowles
 DG,
 Lagarde
 J,
 Veeravalli
 L,
 Ruan
 X,
 Ruan
 Y,
 Lassmann
 
T,
 Carninci
 P,
 Brown
 JB,
 Lipovich
 L,
 Gonzalez
 JM,
 Thomas
 M,
 Davis
 CA,
 Shiekhattar
 R,
 
Gingeras
 TR,
 Hubbard
 TJ,
 Notredame
 C,
 Harrow
 J,
 Guigo
 R.
 The
 GENCODE
 v7
 catalog
 
of
 human
 long
 noncoding
 RNAs:
 analysis
 of
 their
 gene
 structure,
 evolution,
 and
 
expression.
 Genome
 Res.
 2012;22(9):1775-­‐89.
 doi:
 10.1101/gr.132159.111.
 
PubMed
 PMID:
 22955988;
 PMCID:
 PMC3431493.
 
188.
  Mercer
 TR,
 Dinger
 ME,
 Sunkin
 SM,
 Mehler
 MF,
 Mattick
 JS.
 Specific
 expression
 
of
 long
 noncoding
 RNAs
 in
 the
 mouse
 brain.
 Proc
 Natl
 Acad
 Sci
 U
 S
 A.
 
2008;105(2):716-­‐21.
 doi:
 10.1073/pnas.0706729105.
 PubMed
 PMID:
 18184812;
 
PMCID:
 PMC2206602.
 
121
 

 
189.
  Batista
 PJ,
 Chang
 HY.
 Long
 noncoding
 RNAs:
 cellular
 address
 codes
 in
 
development
 and
 disease.
 Cell.
 2013;152(6):1298-­‐307.
 doi:
 
10.1016/j.cell.2013.02.012.
 PubMed
 PMID:
 23498938;
 PMCID:
 PMC3651923.
 
190.
  Karreth
 FA,
 Pandolfi
 PP.
 ceRNA
 cross-­‐talk
 in
 cancer:
 when
 ce-­‐bling
 rivalries
 
go
 awry.
 Cancer
 Discov.
 2013;3(10):1113-­‐21.
 doi:
 10.1158/2159-­‐8290.CD-­‐13-­‐0202.
 
PubMed
 PMID:
 24072616;
 PMCID:
 PMC3801300.
 
191.
  Lee
 JT.
 Epigenetic
 regulation
 by
 long
 noncoding
 RNAs.
 Science.
 
2012;338(6113):1435-­‐9.
 doi:
 10.1126/science.1231776.
 PubMed
 PMID:
 23239728.
 
192.
  Mattick
 JS,
 Rinn
 JL.
 Discovery
 and
 annotation
 of
 long
 noncoding
 RNAs.
 Nat
 
Struct
 Mol
 Biol.
 2015;22(1):5-­‐7.
 doi:
 10.1038/nsmb.2942.
 PubMed
 PMID:
 
25565026.
 
193.
  Mercer
 TR,
 Dinger
 ME,
 Mattick
 JS.
 Long
 non-­‐coding
 RNAs:
 insights
 into
 
functions.
 Nat
 Rev
 Genet.
 2009;10(3):155-­‐9.
 doi:
 10.1038/nrg2521.
 PubMed
 PMID:
 
19188922.
 
194.
  Morris
 KV,
 Mattick
 JS.
 The
 rise
 of
 regulatory
 RNA.
 Nat
 Rev
 Genet.
 
2014;15(6):423-­‐37.
 doi:
 10.1038/nrg3722.
 PubMed
 PMID:
 24776770;
 PMCID:
 
PMC4314111.
 
195.
  Orom
 UA,
 Shiekhattar
 R.
 Long
 noncoding
 RNAs
 usher
 in
 a
 new
 era
 in
 the
 
biology
 of
 enhancers.
 Cell.
 2013;154(6):1190-­‐3.
 doi:
 10.1016/j.cell.2013.08.028.
 
PubMed
 PMID:
 24034243;
 PMCID:
 PMC4108076.
 
196.
  Prensner
 JR,
 Chinnaiyan
 AM.
 The
 emergence
 of
 lncRNAs
 in
 cancer
 biology.
 
Cancer
 Discov.
 2011;1(5):391-­‐407.
 doi:
 10.1158/2159-­‐8290.CD-­‐11-­‐0209.
 PubMed
 
PMID:
 22096659;
 PMCID:
 PMC3215093.
 
197.
  Ulitsky
 I,
 Bartel
 DP.
 lincRNAs:
 genomics,
 evolution,
 and
 mechanisms.
 Cell.
 
2013;154(1):26-­‐46.
 doi:
 10.1016/j.cell.2013.06.020.
 PubMed
 PMID:
 23827673;
 
PMCID:
 PMC3924787.
 
198.
  Bourgeron
 T.
 From
 the
 genetic
 architecture
 to
 synaptic
 plasticity
 in
 autism
 
spectrum
 disorder.
 Nat
 Rev
 Neurosci.
 2015;16(9):551-­‐63.
 doi:
 10.1038/nrn3992.
 
PubMed
 PMID:
 26289574.
 
199.
  Ashwood
 P,
 Wills
 S,
 Van
 de
 Water
 J.
 The
 immune
 response
 in
 autism:
 a
 new
 
frontier
 for
 autism
 research.
 J
 Leukoc
 Biol.
 2006;80(1):1-­‐15.
 doi:
 
10.1189/jlb.1205707.
 PubMed
 PMID:
 16698940.
 
200.
  de
 Jong
 S,
 Newhouse
 SJ,
 Patel
 H,
 Lee
 S,
 Dempster
 D,
 Curtis
 C,
 Paya-­‐Cano
 J,
 
Murphy
 D,
 Wilson
 CE,
 Horder
 J,
 Mendez
 MA,
 Asherson
 P,
 Rivera
 M,
 Costello
 H,
 
Maltezos
 S,
 Whitwell
 S,
 Pitts
 M,
 Tye
 C,
 Ashwood
 KL,
 Bolton
 P,
 Curran
 S,
 McGuffin
 P,
 
Dobson
 R,
 Breen
 G.
 Immune
 signatures
 and
 disorder-­‐specific
 patterns
 in
 a
 cross-­‐
disorder
 gene
 expression
 analysis.
 Br
 J
 Psychiatry.
 2016;209(3):202-­‐8.
 doi:
 
10.1192/bjp.bp.115.175471.
 PubMed
 PMID:
 27151072;
 PMCID:
 PMC5007452.
 
201.
  Hertz-­‐Picciotto
 I,
 Delwiche
 L.
 The
 rise
 in
 autism
 and
 the
 role
 of
 age
 at
 
diagnosis.
 Epidemiology.
 2009;20(1):84-­‐90.
 doi:
 10.1097/EDE.0b013e3181902d15.
 
PubMed
 PMID:
 19234401;
 PMCID:
 PMC4113600.
 
202.
  Kim
 D,
 Volk
 H,
 Girirajan
 S,
 Pendergrass
 S,
 Hall
 MA,
 Verma
 SS,
 Schmidt
 RJ,
 
Hansen
 RL,
 Ghosh
 D,
 Ludena-­‐Rodriguez
 Y,
 Kim
 K,
 Ritchie
 MD,
 Hertz-­‐Picciotto
 I,
 
Selleck
 SB.
 The
 joint
 effect
 of
 air
 pollution
 exposure
 and
 copy
 number
 variation
 on
 
risk
 for
 autism.
 Autism
 Res.
 2017.
 doi:
 10.1002/aur.1799.
 PubMed
 PMID:
 28448694.
 
122
 

 
203.
  Autism:
 Fitting
 the
 Pieces
 Together.
 EBioMedicine.
 2015;2(4):273.
 doi:
 
10.1016/j.ebiom.2015.03.014.
 PubMed
 PMID:
 26137565;
 PMCID:
 PMC4485907.
 
204.
  Sandin
 S,
 Lichtenstein
 P,
 Kuja-­‐Halkola
 R,
 Larsson
 H,
 Hultman
 CM,
 
Reichenberg
 A.
 The
 familial
 risk
 of
 autism.
 JAMA.
 2014;311(17):1770-­‐7.
 doi:
 
10.1001/jama.2014.4144.
 PubMed
 PMID:
 24794370;
 PMCID:
 PMC4381277.
 
205.
  Tylee
 DS,
 Hess
 JL,
 Quinn
 TP,
 Barve
 R,
 Huang
 H,
 Zhang-­‐James
 Y,
 Chang
 J,
 
Stamova
 BS,
 Sharp
 FR,
 Hertz-­‐Picciotto
 I,
 Faraone
 SV,
 Kong
 SW,
 Glatt
 SJ.
 Blood
 
transcriptomic
 comparison
 of
 individuals
 with
 and
 without
 autism
 spectrum
 
disorder:
 A
 combined-­‐samples
 mega-­‐analysis.
 Am
 J
 Med
 Genet
 B
 Neuropsychiatr
 
Genet.
 2017;174(3):181-­‐201.
 doi:
 10.1002/ajmg.b.32511.
 PubMed
 PMID:
 
27862943.
 
206.
  Martin
 M.
 Cutadapt
 removes
 adapter
 sequences
 from
 high-­‐throughput
 
sequencing
 reads.
 EMBnetjournal
 2011;17:10-­‐2.
 doi:
 10.14806/ej.17.1.200.
 
207.
  Dennis
 G,
 Jr.,
 Sherman
 BT,
 Hosack
 DA,
 Yang
 J,
 Gao
 W,
 Lane
 HC,
 Lempicki
 RA.
 
DAVID:
 Database
 for
 Annotation,
 Visualization,
 and
 Integrated
 Discovery.
 Genome
 
Biol.
 2003;4(5):P3.
 PubMed
 PMID:
 12734009.
 
208.
  Langfelder
 P,
 Horvath
 S.
 WGCNA:
 an
 R
 package
 for
 weighted
 correlation
 
network
 analysis.
 BMC
 Bioinformatics.
 2008;9:559.
 doi:
 10.1186/1471-­‐2105-­‐9-­‐559.
 
PubMed
 PMID:
 19114008;
 PMCID:
 PMC2631488.
 
209.
  Louros
 SR,
 Osterweil
 EK.
 Perturbed
 proteostasis
 in
 autism
 spectrum
 
disorders.
 J
 Neurochem.
 2016;139(6):1081-­‐92.
 doi:
 10.1111/jnc.13723.
 PubMed
 
PMID:
 27365114;
 PMCID:
 PMC5215415.
 
210.
  Klein
 ME,
 Monday
 H,
 Jordan
 BA.
 Proteostasis
 and
 RNA
 Binding
 Proteins
 in
 
Synaptic
 Plasticity
 and
 in
 the
 Pathogenesis
 of
 Neuropsychiatric
 Disorders.
 Neural
 
Plast.
 2016;2016:3857934.
 doi:
 10.1155/2016/3857934.
 PubMed
 PMID:
 26904297;
 
PMCID:
 PMC4745388.
 
211.
  Manning
 BJ,
 Yusufzai
 T.
 The
 ATP-­‐dependent
 Chromatin
 Remodeling
 Enzymes
 
CHD6,
 CHD7,
 and
 CHD8
 Exhibit
 Distinct
 Nucleosome
 Binding
 and
 Remodeling
 
Activities.
 J
 Biol
 Chem.
 2017.
 doi:
 10.1074/jbc.M117.779470.
 PubMed
 PMID:
 
28533432. 
Abstract (if available)
Abstract From genome-wide association studies (GWAS), our laboratory discovered a novel gene that has a highly significant association with ASD. This gene is a long noncoding RNA (lncRNA) called MSNP1AS (moesin pseudogene 1, antisense). MSNP1AS has increased expression in the cerebral cortex of people with ASD and people with the ASD-associated genetic marker revealed by the GWAS. I hypothesized that MSNP1AS overexpression changes the expression of other genes that contribute to neuronal processes. I collaborated with experts in psychiatry, HIV research, and prostate cancer research to uncover the effect of ASD-associated noncoding RNAs (ncRNAs) on human neuronal cells. ❧ First, I overexpressed MSNP1AS in human neuronal cell lines to provide a working model of the ASD phenotype. I used RNA-Seq to find transcripts with altered expression in the cells after overexpression and recorded the impact of the increased MSNP1AS expression on cell morphology. MSNP1AS is 94% identical and antisense to moesin, a protein-coding gene. Moesin is an X chromosome gene that plays a role in neuronal processes and stability as well as immune response. While MSNP1AS overexpression significantly decreased moesin protein expression, moesin transcript levels remained unchanged. RNA-sequencing after MSNP1AS overexpression revealed in changes in gene expression of genes related to chromatin organization and protein synthesis, pointing to a pathway independent of the MSNP1AS/MSN interaction. Moesin contributes to neuron growth, and ASD patients have decreased long-distance brain connectivity, which makes moesin a good candidate for study in this disorder. Cell growth was stunted by MSNP1AS overexpression. Neurite length and number were decreased after overexpression, hinting at an effect on connectivity. ❧ Second, I determined the best way to effectively decrease MSNP1AS in human neural progenitor cells to reveal transcripts with altered expression as potential therapeutic targets. These results validated MSNP1AS’s effect on chromatin organization and uncovered an immune response. Within ASD is a subset of patients who experience immune dysfunction, such as autoimmune disorders and gastrointestinal upset. More work needs to be done to evaluate whether or not MSNP1AS dysregulation is specific to these patients. ❧ Through a collaboration with UC Davis, I was able to validate these changes in immune response genes’ expression in vivo and point to a potential environmental contribution. The Childhood Autism Risks from Genetics and the Environment (CHARGE) study took blood samples from children with and without autism who had either been exposed to pesticides in their environment or not. The RNA from the blood samples was sequenced and the genes changed showed pathways involved in the immune system. Taken together, these experiments suggest molecular mechanisms that are working together to alter brain development, contributing to ASD pathogenesis. 
Linked assets
University of Southern California Dissertations and Theses
doctype icon
University of Southern California Dissertations and Theses 
Conceptually similar
The autism-associated gene SYNGAP1 regulates human cortical neurogenesis
PDF
The autism-associated gene SYNGAP1 regulates human cortical neurogenesis 
Placenta growth factor-miRNAs-lncRNAs axis in the regulation of ET-1 gene involved in pulmonary hypertension in sickle cell disease
PDF
Placenta growth factor-miRNAs-lncRNAs axis in the regulation of ET-1 gene involved in pulmonary hypertension in sickle cell disease 
The role of Rho-GEF signaling in synapse function and autism-related disorders
PDF
The role of Rho-GEF signaling in synapse function and autism-related disorders 
LINC00261 induces a G2/M cell cycle arrest and activation of the DNA damage response in lung adenocarcinoma
PDF
LINC00261 induces a G2/M cell cycle arrest and activation of the DNA damage response in lung adenocarcinoma 
The role of Ryk and Smek in neurogenesis; Mechanisms of CBP/β-catenin signaling inhibitor and IL-6 mediators in head and neck cancer
PDF
The role of Ryk and Smek in neurogenesis; Mechanisms of CBP/β-catenin signaling inhibitor and IL-6 mediators in head and neck cancer 
Novel roles for Maf1 in embryonic stem cell differentiation and adipogenesis
PDF
Novel roles for Maf1 in embryonic stem cell differentiation and adipogenesis 
Novel long non-coding RNA, LINC00824, contributes to stemness in hepatocytes
PDF
Novel long non-coding RNA, LINC00824, contributes to stemness in hepatocytes 
The AIDEN acronym: increasing nurse anesthetists' knowledge of preoperative care for children with autism spectrum disorder
PDF
The AIDEN acronym: increasing nurse anesthetists' knowledge of preoperative care for children with autism spectrum disorder 
The role of Hic-5 in glucocorticoid receptor binding to chromatin
PDF
The role of Hic-5 in glucocorticoid receptor binding to chromatin 
RNA methylation in cancer plasticity and drug resistance
PDF
RNA methylation in cancer plasticity and drug resistance 
Modeling SynGAP1 truncating mutations in neurodevelopmental disease using iPSC-derived neurons
PDF
Modeling SynGAP1 truncating mutations in neurodevelopmental disease using iPSC-derived neurons 
DNA methylation inhibitors and epigenetic regulation of microRNA expression
PDF
DNA methylation inhibitors and epigenetic regulation of microRNA expression 
Understanding the role of APP and DYRK1A in human brain pericytes
PDF
Understanding the role of APP and DYRK1A in human brain pericytes 
Identification of CBP/FOXM1 as a molecular target in triple negative breast cancer
PDF
Identification of CBP/FOXM1 as a molecular target in triple negative breast cancer 
A functional genomic approach based on shRNA-mediated gene silencing to delineate the role of NF-κB and cell death proteins in the survival and proliferation of KSHV associated primary effusion l...
PDF
A functional genomic approach based on shRNA-mediated gene silencing to delineate the role of NF-κB and cell death proteins in the survival and proliferation of KSHV associated primary effusion l... 
The AIDEN acronym: Increasing nurse anesthetists’ knowledge of preoperative care for children with autism spectrum disorder
PDF
The AIDEN acronym: Increasing nurse anesthetists’ knowledge of preoperative care for children with autism spectrum disorder 
Functional study of C9ORF72 and its implication in the pathogenesis of amyotrophic lateral sclerosis
PDF
Functional study of C9ORF72 and its implication in the pathogenesis of amyotrophic lateral sclerosis 
Elucidating the functional role of CHD7 associated nuclear PDH complex and other associated proteins on neural crest development
PDF
Elucidating the functional role of CHD7 associated nuclear PDH complex and other associated proteins on neural crest development 
Using novel small molecule modulators as a tool to elucidate the role of the Myocyte Enhancer Factor 2 (MEF2) family of transcription factors in leukemia
PDF
Using novel small molecule modulators as a tool to elucidate the role of the Myocyte Enhancer Factor 2 (MEF2) family of transcription factors in leukemia 
Development of targeted therapies for peroxisome biogenesis disorders
PDF
Development of targeted therapies for peroxisome biogenesis disorders 
Asset Metadata
Creator DeWitt, Jessica  Jolynn (author) 
Core Title Genetics and the environment: evaluating the role of noncoding RNA in autism spectrum disorder 
School Keck School of Medicine 
Degree Doctor of Philosophy 
Degree Program Development, Stem Cells and Regenerative Medicine 
Publication Date 07/21/2017 
Defense Date 06/06/2017 
Publisher University of Southern California (original), University of Southern California. Libraries (digital) 
Tag antisense RNA,autism spectrum disorder,epigenetics,genes,genetics,lncRNA,long noncoding RNA,Molecular Biology,ncRNA,neurodevelopment,neurodevelopmental disorder,OAI-PMH Harvest,psychiatric disorder,psychiatry,siRNA, genomics 
Language English
Contributor Electronically uploaded by the author (provenance) 
Advisor Bajpai, Ruchi (committee chair), Campbell, Daniel B. (committee member), Ichida, Justin (committee member), Wood, Ruth I. (committee member) 
Creator Email jdewitt@apu.edu,jjdewitt@usc.edu 
Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-408002
Unique identifier UC11264140 
Identifier etd-DeWittJess-5577.pdf (filename),usctheses-c40-408002 (legacy record id) 
Legacy Identifier etd-DeWittJess-5577.pdf 
Dmrecord 408002 
Document Type Dissertation 
Rights DeWitt, Jessica  Jolynn 
Type texts
Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection) 
Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law.  Electronic access is being provided by the USC Libraries in agreement with the a... 
Repository Name University of Southern California Digital Library
Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
antisense RNA
autism spectrum disorder
epigenetics
genes
genetics
lncRNA
long noncoding RNA
ncRNA
neurodevelopment
neurodevelopmental disorder
psychiatric disorder
psychiatry
siRNA, genomics