.I I.
.1...
3: I

. . :4
206M?

:1

. .7 :l. .3
3/! 2774.37!)-

x _
.2; and.

 

 

”3‘9

(; b
#ka \ .

g56f4j33

This is to certify that the
dissertation entitled

MOLECULAR ANALYSIS OF SMITH-MAGENIS SYNDROME:
MAPPING, CANDIDATE GENE IDENTIFICATION, AND
PRELIMINARY CHARACTERIZATION OF RAH

presented by
REBECCA E. SLAGER

has been accepted towards fulfillment
of the requirements for the

PhD. degree in Genetics

CEMM 6%

Major Professor’s Signature
5/0201/0 4

Date

MSU is an Affinnative Action/Equal Opportunity Institution

W
Michigan State I
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

AUG 28 @903

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/CIRC/DateDue.p65-p. 15

MOLECULAR ANALYSIS OF SMITH-MAGENIS SYNDROME: MAPPING,
CANDIDATE GENE IDENTIFICATION, AND PRELIMINARY
CHARACTERIZATION OF RAII
By

Rebecca E. Slager

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Genetics program

2004

ABSTRACT

MOLECULAR ANALYSIS OF SMITH-MAGENIS SYNDROME: MAPPING,
CANDIDATE GENE IDENTIFICATION, AND PRELIIVIINARY
CHARACTERIZATION OF RAII

By4

Rebecca E. Slager

Smith-Magenis syndrome (SMS) is a multiple congenital anomalies, mental
retardation syndrome typically associated with a microdeletion of human chromosome 17
band p112. SMS patients display characteristic physical abnormalities as well as
developmental delay, sleep disturbance, and self-injurious behaviors. The research foCus
in the Elsea laboratory is to understand the molecular basis of SMS and which gene or
genes contribute to the phenotype of this complex physical and neurobehavioral disorder.
Previous to this work, a great deal of effort was put forth to define the deletion interval
and to begin to map genes, ESTs, and genomic markers to the SMS region of 17p. My
research project focused on fine-mapping the smallest deleted region along 17p11.2
which could still reproduce the entire SMS phenotype, or what is termed the SMS critical
interval. Through a combination of in silico analysis of high-throughput genome
sequence and hybridization mapping of genes and expressed sequence tags (ESTs) to
17pl 1.2 genomic clones, a physical and transcription map of the SMS critical region was
created. Mapping of genes and ESTs to the contig was reconciled with naturally-
occurring patient deletions by Southern hybridization to digested DNA from a somatic
cell hybrid panel of overlapping 17p deletions. The physical map of ~1.5 Mb SMS ‘
critical interval consisted of a minimum tiling path of 18 genomic clones and the

transcription map contained 17 known genes, 12 ESTs, and six genomic markers. Recent

work by other members of the Elsea laboratory has re-defined the SMS critical interval as
a ~950 kb region containing ~l8-22 genes.

Several promising candidate genes were identified within the SMS critical
interval and my research focused on the characterization of one particular gene, retinoic
acid induced 1 (R411). This large, novel protein is highly expressed in the brain and
shares some sequence homology with a transcriptional coactivator, though the cellular
role of RAIl is currently unknown. As we did not have the ability to assess the
functional role of RAII, we began a mutation screen of RAII in several patients with an
SMS phenotype but no detectable 17p11.2 deletion. We hypothesized that a mutation in
a single gene in these patients may produce an SMS phenotype. Our studies
subsequently revealed dominant, deleterious RAIl mutations in four patients. These
findings strongly suggest that haploinsufficiency of the RAN protein is sufficient to
produce the craniofacial and neurobehavioral features seen in SMS patients.

In order to assess Rail haploinsufficiency in a live animal model, a gene-targeting
construct was developed to remove Rail expression in mice and embryonic stem cells
were subsequently identified which contained the correctly targeted construct. In
addition, we assessed Rail dosage sensitivity by constructing stable Rail BAC transgenic
lines and performed a preliminary physical assessment of these mice. Preliminary
measurements of weight and total body length suggest that animals with higher Rail

copy numbers may weigh less and have a smaller body length than normal littermates.

I would like to dedicate this work to my grandfather, Dr. Robert E. Lucas, who received
his PhD. from Michigan State University in 194 7.

iv

ACKNOWLEDGMENTS

I would like to thank all of my fi'iends and family who have supported me
throughout my education at Michigan State University. I am especially grateful to my
mentor, Sarah Elsea, for her encouragement and devotion to excellence. My guidance _
committee, Karen Friderici, Patrick Venta, Will Kopachik, and Michael Grotewiel, has
provided useful advice and demonstrated that they always believed in my abilities. And
this work would never have been achieved without the enthusiasm, collaboration, and
dedication of other members of the Elsea laboratory, including Christopher Vlangos,
Dennis Lettau, Tiffany Newton, Catherine Barth, Santhosh Girirajan and Valerie
Vinoverski. Past members of the lab who have become inestimable friends are Ellen
Wilch, Leeyoung Park, and Soumya Korrapati. I would also like to thank Mei Zhu,
Becky Bedilu, Sainan Wei, Heather Prince, Trevor Wagner, and Paolo Struffi for their
friendship and Jon Stoltzfus for always taking time out to help with experiments and to
read drafis of publications. Collaborators Sally Camper, Thom Saunders, Laura Schmidt,
Guy Rouleau, and Pragna Patel have contributed invaluable expertise to this work. And I
would like to recognize the hard work of Jeannine Lee, Genetics program secretary, and
Pappan from Biochemistry, who always have the right answers. Most of all, I want to
thank my husband for his unfailing support and faith in me.

This work was partially funded by a Dissertation Completion Fellowship from the

MSU graduate school.

TABLE OF CONTENTS

LIST OF TABLES ........................................................................... viii
LIST OF FIGURES...............- .......................................................... ix
KEY TO ABBREVIATIONS .............................................................. xi

Chapter I. Introduction

Microdeletion syndromes and genomic disorders ................................. 1
SMS clinical summary ................................................................ 7
Determination of the molecular basis of microdeletion syndromes ........... 10
Animal models of microdeletion syndromes ...................................... 20
Conclusions ............................................................................ 26

Chapter II. Fine-mappinggthe SMS critical interval and identification of SMS
candidate genes

Definition of the SMS critical interval .............................................. 27
Physical and transcription map of the SMS critical interval ...................... 29
Re-definition of the SMS critical interval .......................................... 38
EST characterization and analysis ................................................... 38
Summary and analysis of candidate genes .......................................... 47
Known genes originally mapped to the SMS critical interval .................... 52
Conclusions ............................................................................. 56
Materials and methods ................................................................. 57

Chapter III. RAII as a candidate gene for SMS

Rail/RAH cloning and genomic structure .......................................... 62
Expression pattern ofRAIl/Rail .................................................... 68
Clinical description of putative SMS patients with no cytogenetic deletion...75
FISH experiments ..................................................................... 79
Sequencing of RAII in putative SMS patients with no deletion ................. 86
Other RAII sequence features ........................................................ 95
Towards high-throughput RAIl mutation screening ............................. 104
Conclusions ........................................................................... l 13
Materials and methods ............................................................... 114

vi

Chapter IV. In vivo evaluation of mouse Rail

SMS mouse models .................................................................. 123
BAC DNA isolation .................................................................. 127
Assessment of Rail BAC transgenic founder mice .............................. 129
Copy number determination by Southern analysis ............................... 132
Molecular assessment of Rail BAC transgenic F 1 mice ........................ 137
Cloning of the Rail knockout vector ............................................... 153
Assessment of gene-targeted mice by Southern analysis ....................... 157
Conclusions... ................................................................... 160
Materials and methods ............................................................... 161

Chapter V. Discussion

The role ofRAIl in SMS ............................................................ 172
RAII mouse models .................................................................. 175
RAIl and craniofacial development ................................................ 176
Other possible roles of RAIl/Rail in development .............................. 179
Neurological development, retinoic acid and RA]! .............................. 181
A possible role for RAIl in ADHD and schizophrenia? ......................... 183
Implications ofRAIl mutation screening ......................................... 186
Another SMS locus? ................................................................. 187
Conclusions ............................................................................ 188
APPENDIX A .................................................................................. 190
APPENDIX B .................................................................................. 199
BIBLIOGRAPHY ............................................................................. 205

vii

LIST OF TABLES

Table 1. ESTs mapped by Lucas et a1. within the 2001 ~1.5 Mb critical interval which
have been identified as known genes ....................................................... 43

Table 2. Novel ESTs mapped by the Elsea lab within the 2001 ~1.5 Mb critical
interval .......................................................................................... 44

Table 3. ESTs recently mapped by other groups within the 2001 ~1.5 Mb critical

interval ......................................................................................... 45
Table 4. Phenotype of SMS candidate genes disrupted in mouse ..................... 48
Table 5. Phenotypic characteristics of SMS patients a cytogenetic deletion and putative
SMS patients with deleterious RAll mutations ........................................... 83
Table 6. Phenotypic characteristics of putative SMS patients with no cytogenetic deletion
who harbor no obvious RAII mutation ..................................................... 85
Table 7. RA]! primers used this this study. ............................................... 87 '
Table 8. M11 sequence changes detected within patient pool of putative SMS patients
with no cytogenetic deletion and normal controls ....................................... 99
Table 9. PCR primers used for amplification of mouse Rail ........................... 133

Table 10. Summary of F 1 Rail BAC transgenic breeding from ~August- December

Table 11. Rail BAC transgenic F l offspring 5 week qualitative assessment data... 139
Table 12. Rail BAC transgenic F 1 offspring 5 week quantitative assessment data. 140
Table 13. Rail BAC transgenic F l offspring 10 week qualitative assessment data..141

Table 14. Rail BAC transgenic F1 offspring 10 week quantitative assessment data..142

viii

LIST OF FIGURES

Figure 1. Repetitive elements involved in the SMS common l7p11.2 deletion and

dup(1 7)(p1 1.2p11.2) duplication ............................................................ 6
Figure 2. Transcription map of the ~1.5 Mb SMS critical interval (adapted fi'om Lucas et
al., 2001.) ....................................................................................... 32
Figure 3. Mapping RAII to somatic cell hybrids ......................................... 35
Figure 4. Mapping RAIl to BACs/PACs ................................................... 37
Figure 5. SMS deletion analysis and refined SMS critical interval (adapted from Vlangos
et al., 2003) .................................................................................... 40
Figure 6. Human RAIl northern analysis .................................................. 42
Figure 7. Human and mouse RAIl/Rail genomic structure ............................. 67
Figure 8. Mouse and human Rai/RAIl amino acid sequence alignment .............. 70-2
Figure 9. Mouse Rail embryonic expression pattern ..................................... 77

Figure 10. FISH analysis on putative SMS patients with no detectable deletion... 81

Figure 11. SMS129 RAII mutation analysis .............................................. 89
Figure 12. SMSIS6 RAIl mutation analysis .............................................. 92
Figure 13. SMS159 RAII mutation analysis .............................................. 94
Figure 14. SMS188 RAII mutation analysis ............................................... 97
Figure 15. RA]! sequence changes ..................... . ................................... 101
Figure 16. DHPLC RAII mutation screening results from parental and control
samples .......................................................................................... 103
Figure 17. DHPLC screening results of known RAII mutations ....................... 106-8
Figure 18. TGCE screening results of known RAII mutations ......................... 110-12

Figure 19. Mouse chromsome 11 genomic region syntenic to the SMS deletion region
containing Rail BAC 326M22 ............................................................... 126

ix

Figure 20. Copy standard and PCR evaluation of Rail BAC transgenic mice ........ 131

Figure 21. Southern analysis to determine the integrated Rail BAC copy number in
founder mice ................................................................................... 135

Figure 22. Southern analysis of Rail BAC transgenic founders 755 and 775, and F1

offspring ........................................................................................ 145
Figure 23. Weights of Rail F 1 BAC transgenic animals and normal littermates at 5 and
10 weeks ....... g ................................................................................. 148
Figure 24. Total body lengths of Rail F l BAC transgenic animals and normal littermates
at 5 and 10 weeks .............................................................................. 150
Figure 25. Weight vs. total body length of F1 Rail BAC transgenic mice ........... 152
Figure 26. Rail knockout construct ........................................................ 156
Figure 27. Southern analysis of ES cell DNA containing the targeted Rail knockout ‘
construct ........................................................................................ 159
Figure 28. BAC 326M22 DNA quantitative PCR standard curve ..................... 202

ABI:
ADHD:
AGS:

A IPAFZ:

ASP:
BAC:
BAZIB:
bp:
COPS3:

CREBBP:

Cyln2:
DGS:
DHPLC:
DRGZ:
ELN:
ES:
ng‘B:
FISH:
FLII:
GFP:
GTSF :
HT GS:
IHC:
ISH:

JA Gl :
kb:
LCR:
LIMKI :
LLGLl:
LOD:
M-RIP:

MYOISA:

NAHR:
ng:

N T 5M:
NTF:
NTM:
PAC:
PBS:
PCR:
PEM T2:

KEY TO ABBREVIATIONS

amino acid

Applied Biosystems

attention deficit hyperactivity disorder

Alagille syndrome

ATP synthase mitochondrial F1 complex assembly factor 2
Angelman syndrome

affected sib pair

bacterial artificial chromosome

bromodomain adjacent to zinc finger domain, 13
base pair

COP9 homolog subunit 3

Creb-binding protein

cytoplasmic linker 2

DiGeorge syndrome

denaturing high-performance liquid chromatography
developmentally-regulated GTP-binding protein 2
Elastin

embryonic stem

Fibroblast growth factor 8

fluorescent in situ hybridization

flightless I homolog

green fluorescent protein

Genomics Technology and Support Facility
high-throughput genome sequence
immunohistochemistry

in situ hybridization

Jaggedl

kilobase

low-copy repeat

Lim-kinase 1

lethal giant larvae homolog 1

logarithm of odds .

myosin phosphatase-Rho interacting protein
myosin XVA

non-allelic homologous recombination
nanograms

5',3'-nucleotidase, mitochondrial
non-transgenic female

non-transgenic male

Pl-artificial chromosome

phosphate-buffered saline

polymerase chain reaction
phosphatidylethanolamine N-methyltransferase

xi

P83
PHD:
PWS:
RA:
RAII:
RARE:
RASDI :
REM:
SCA2:
SHMTI:
SMS:
SNP:
SREBFI:
SVAS:
Tbxl :
TCFZO:
TGCE:
TF:

TM:
TOMILZ:
T 0P3A :

U OfM TAMC:

UBE3A:
UTR:
VCFS:
WS:
WSCR:
YAC:

 

picograms

plant homeOdomain
Prader-Willi syndrome

retinoic acid

retinoic acid induced 1

retinoic acid response element
RAS, dexamethasone-induced 1
rapid eye movement
spinocerebellar ataxia type 2
serine hydroxymethyl transferase 1
Smith-Magenis syndrome
single nucleotide polymorphism

' sterol regulatory element binding transcription factor 1

supravalvular aortic stenosis

T-box containing transcription factor 1
Transcription factor 20

temperature gradient capillary electrophoresis
transgenic female

transgenic male

target of mybl-like 2

topoisomerase III alpha

University of Michigan Transgenic Animal Core
E6-AP ubiquitin protein ligase

untranslated region

velo-cardio—facial syndrome

Williams syndrome

Williams syndrome critical interval

yeast artificial chromosome

xii

Chapter I.

' Introduction

Microdeletion syndromes and genomic disorders

Smith-Magenis syndrome (SMS) is a multiple congenital anomalies, mental
retardation syndrome typically associated with a microdeletion of human chromosome 17
hand p11.2 (Smith et a1. 1982; Smith et a1. 1986; Greenberg et al. 1991; Greenberg et al.
1996). Microdeletion syndromes result from small deletions of chromosomal material
and form a subset of chromosomal abnormalities which traditionally account for 6%-10%
of all infants born with severe congenital anomalies (Shapira 1998). Diagnosis of a
particular microdeletion syndrome is typically made by molecular cytogenetic procedures
such as fluorescent in situ hybridization (FISH), as the deletions can be undetectable even
by means such as high-resolution chromosomal banding analysis (prometaphase or
metaphase spreads to 550 band resolution). FISH analysis relies on the hybridization of a
fluorescently labeled probe that is specific for a region of a chromosome; the absence of
one chromosomal copy of the signal in a metaphase spread is diagnostic for a deletion.
FISH can also be useful in the confirmation of a clinical diagnosis and the detection of
Patients with different deletion sizes (Shapira 1998). As a microdeletion generally occurs
’18 nova on only one member of a homologous pair of chromosomes, the mode Of

inheritance of the microdeletion syndrome is considered to be autosomal dominant.

As molecular cytogenetics improves and more complex chromosomal

rearrangements can be detected, a new discipline has emerged to understand the

 

 

   

underlying mechanism of deletions, duplications, translocations, and inversions, Which
fall under the broad term of genomic disorders. Through the completion of the human
genome project; new sequence data indicate that many of the chromosomal regions
throughout the genome that are prone to rearrangement are flanked by chromosome-
specific low-copy repeats (LCRs). The structure of these LCR segments can be simple or
complex and may encompass various genes, gene fragments, pseudogenes, and retroviral
sequences (Ii et a1. 2000; Stankiewicz et a1. 2003). Many of these LCRs are present in
>2-10 copies and tend to be preferentially located in centromeric and telomeric regions,
and the genomic distance between the LCR blocks is typically kilobases to megabases (J i,
Eichler et a1. 2000; Stankiewicz et al. 2002). LCR repeat blocks of ~10-400 kb are
highly similar (>95%) and have the ability to act as substrates for non-allelic homologous
recombination (NAHR) or unequal crossing over during meiosis, which results from the
mis-pairing of the repeated fragments (Stankiewicz, Shaw et a1. 2003). NAHR can occur
between homologous chromosomes, sister chromatids, or within a single chromatid to
produce unequal products of recombination or a “looping out” of genetic material
(Stankiewicz and Lupski 2002). Several genetic disorders, including microdeletion and
duplication syndromes or other disorders caused by large chromosomal rearrangements,
l’esult fi-om NAHR sponsored by the mispairing of flanking LCRs. The origin of many of

these LCR segments is thought to be duplication events which occurred during primate I
s1>eciation approximately 35-50 million years ago (Stankiewicz and Lupski 2002), though

how these sequences became fixed within the human genome is currently unknown.

 

The complex nature of chromosome-specific LCRs has recently been under
intense investigation following the availability of human drafi and finished genomic
sequence. Many genomic regions, including the pericentromic regions of chromosome
17, did not have correct sequence alignment or orientation due to the high similarity
between the LCRs, which caused the misassembly of several bacterial artificial
chromosome (BAC) and P1 artificial chromosome (PAC) clones. This was especially
true of repeated LCR blocks of ~200 kb, which is the size of many sequenced BACs
(Stankiewicz, Shaw et a1. 2003). Many reviews now exist which analyze the structure of
the LCRs flanking known microdeletions as well as the possible mechanism of NAHR
(Ji, Eichler et a1. 2000; Park et a1. 2002; Stankiewicz and Lupski 2002; Stankiewicz,
Shaw et a1. 2003). As our lab is interested primarily in the molecular basis of SMS, this
work will highlight the LCR repeat regions which flank the SMS common deletion
breakpoints within l7pl 1.2, termed the proximal and distal SMS-REPS, or SMS-REPP
and SMS-REPD, as well as a third middle REP (SMS-REPM). The SMS-REPS contain
at least 14 genes and pseudogenes including the keratin (KER) gene cluster, CLP, and
T RE. Many of these genes also map to multiple locations throughout the genome,
including the q arm of chromosome 17 (Park, Stankiewicz et a1. 2002). The. SMS-REPS
are >98% similar repeat blocks of ~200kb which are believed to have arisen from a
progenitor SMS-REP with nearly the same structure as the SMS-REPP, nearly ~40-65
million years old during the divergence of Old and New World monkeys (Park,
Stankiewicz et al. 2002; Stankiewicz and Lupski 2002). While the SMS-REPP and SMS-
REPD occur in the same orientation, various complex rearrangements, including two

terminal deletions, an interstitial deletion, and an inversion which occurred throughout

 

   

the evolution of the SMS-REPS have been proposed to explain the existence of the SMS-
REPM, which occurs in the opposite orientation (Park, Stankiewicz et a1. 2002;
Stankiewicz and Lupski 2002). Interestingly, sequence homology comparisons between
mouse and human genomes show that the SMS-REPS are present at breaks in the 'synteny
of these genomes (Stankiewicz and Lupski 2002). The highly similar restriction pattern
of several sequenced BACs from the human genome project which comprise the SMS-
REPs are shown in the Southern hybridization in Figure 1, panel (a). As shown in the
schematic in Figure 1, panel (b), de novo deletions and duplications within l7pll.2 are
likely due to interchromosomal NAHR events involving the SMS-REPS (Chen et al.
1997). The SMS-REPP and the SMS-REPD, which flank the ~4 Mb SMS common
deletions are in the same orientation and unequal crossing over of these repeat blocks can
sponsor meiotic NAHR which result in the SMS common deletion (Chen, Manian et al.
1997). Less frequently, the inverted SMS-REPM can also act as a substrate for NAHR,
as at least five reported deletions have been analyzed which involve the SMS-REPM
(Park, Stankiewicz et al. 2002; Stankiewicz, Shaw et a1. 2003). Several patients with the
SMS common deletion were identified by the presence of a unique junction fi'agment, the
putative remnant from an unequal crossing over event (Chen, Manian et al. 1997). In-
depth molecular analysis of the recombinant junctions from SMS patients with the
common deletion demonstrate that ~50% of NAHR events occurred within the KER gene
cluster (Bi et al. 2003). Patients with the reciprocal duplication of l7p1 1.2 have also
recently been identified and characterized (Potocki et al. 2000). As SMS patients usually
display a number of physical abnormalities and self-abusive behaviors (Greenberg,

Guzzetta et al. 1991; Smith et al. 1998), the phenotype of these dup(17)(p11.2 p112)

 

 

t9 involved in the SMS common l7pll.2 deletion and

. “‘6 O I
vefi‘ 6‘3 \‘cat‘on
Rafi“: ‘19\\- 6“?
“mg {\ouo“ pattern of the SMS-REPS is represented by Southern
est B AC3 mapping to the proximal REP (b6676C11 and

3 The $.\“\.\\fl ‘N A {tom
(‘1: bfid’t‘lfifio“ D. . \s distal REP (bcl98Hl S, pc48114 and bc219A15 h °
Y me‘“ , ,sownrn
bcA3AD'2, she mam P @cISSMZO, bc58lM24, pc37NO7 and bc340lO, shown in
qu‘e) and ‘3‘ as a B AC containing unique sequence (bc81519, shown in blue) was
$36,325?th EcoRL electrophoresed on a 1% TAE gel, and transferred to a nylon
mimbrane- The Southern blot was probed With a >10 kb fragment from bc198H15,
which hybridized to several bands In BACs from all three of the SMS-REP sequences
and highlights the extremely similar restriction pattern of these BACs. As a ne ativ
control, bc198H15 BAC fragment did not bind to the unique genomic sequerisce ‘6
pc815I9. In-depth sequence analysis reveals that the SMS-REPS >9 ° ' ' m
(Stankiewicz et al. 2002) are 8/0 Identical
(b) Meio .
tic non-all 1' ' ‘

1‘ cp’CSen e no homologous recombination on human chromo '
mmontsd (aeiapted fiom Potocki et al. 2000), which results in two prosgtiEES'lih; 1SIVIISS
REPP (111.1138:er aSSBENell as the reciprocal duplication of the 17p11.2 region. 'The SMS-
.the FLI] ) 1798-REPD (purple) flank unique genomic sequence which includes
identical an REPP 881168. Homologous recombination can occur between the ~95%
an du SMS’ 1 2p] land the SMS-REPD, which produces the SMS common deletion
The S 1307)“) IIVI, Wh:2), both ofwhich can be Identified by unique junction fragments.
MS- p lch occurs in the opposrte orientation as the SMS-REPP and the

S .
dgs‘REPD and may sponsor unusual l7pll.2 deletions, is not represented in this
Stain

Images .1 n this dissertation are presented in color.

  

bc34010
bc219AIS

  

bc581 M24
pc37N07

.8. - a...

i. a, .9
g m

a .r .
..

bun—0:0: mam 90—639:
mamas" necdehkezs

WRENM

25¢!me
5.52
.538 9.323533
NZNQ we Egon:
mam-:2.

 

patients is generally considered to be less severe (Potocki, Chen et al. 2000). The
duplication patients usually have a normal physical appearance, mild mental retardation
and some behavioral abormalities (Potocki, Chen et al. 2000). These dup(17)(p11.2
p11.2) patients were also identified by the presence of a unique junction fragment (Chen,
Manian et al. 1997; Ji, Eichler et al. 2000; Potocki, Chen et al. 2000). The existence of
these reciprocal deletion and duplication syndromes within 17p provides evidence for the
model of NAHR sponsored by complex genome architecture (Chen, Manian et al. 1997;
Potocki, Chen et al. 2000) and also provides a unique opportunity to identify particularly

dosage-sensitive genes within the 17pl 1.2 region.

SMS clinical summary

Smith-Magenis syndrome (SMS) is a multiple congenital anomalies/mental
retardation syndrome associated with an interstitial deletion on chromosome 17 involving
band pll.2 (Smith, L. et al. 1982; Smith, McGavran et al. 1986; Stratton et al. 1986;
Greenberg, Guzzetta et al. 1991). The birth prevalence of SMS is estimated to be
approximately 1225,000 (Greenberg, Guzzetta et al. 1991), although SMS is likely
underdiagnosed due to the fact that it is a recently-described syndrome and its phenotype
can be subtle. Many cases have been identified in the last 10 years as a result of high-
resolution cytogenetic techniques such as FISH. SMS is known to affect individuals of
all ethnic and racial backgrounds. A detailed phenotypic summary of SMS is available
on the world wide web at the GeneReviews website (www.genetests.org), through
executing a search on this website for Smith-Magenis syndrome. This website is

overseen by a board of professionals who update the site regularly and provide accurate,

referenced information about the clinical and molecular aspects of SMS. Most of the

SMS clinical information below has been adapted fiom the SMS entry at GeneReviews:

In infancy, there are very subtle signs that a baby may have SMS. Usually, the
infant is born at term and presents with a normal birth height and weight. As a baby, an
SMS child may display dysmorphic facies, characterized by midface hypoplasia, a short,
up-turned nose, and a “tented” upper lip, although these physical features may be so
subtle as to be overlooked. Infants with SMS are often described as “perfect babies” who
sleep for long periods of time and cry infrequently. Some feeding difficulties are
common in SMS infants, as well as poor suck, and gastroesophogeal reflux. Nearly

100% of SMS infants are hypotonic and lethargic (Greenberg, Lewis et al. 1996).

The physical features of an individual with SMS become more pronounced in
early childhood. In addition to the typical SMS facial appearance noted above, children
with SMS may have ocular abnormalities, including myopia, strabismus, or microcomea.
In general, SMS children also have short stature as well as short hands and feet and some
SMS children have mild scoliosis. Other common features of children with SMS include
otolaryngologic difficulties, otitis media (often severe enough to induce ear tube
placement), hearing loss, and hypercholesterolemia (Smith, McGavran et al. 1986; Smith
et al. 1990; Greenberg, Guzzetta et al. 1991). In general, oral sensory motor dysfunction
is a major concern, especially as many SMS children have a marked hoarse voice and

significant speech delay. However, with the proper intervention, an SMS child can

develop verbal speech by school age. Cardiac, renal anomalies and cleft palate have been

noted to occur in smaller percentage of patients (www.genetests.org).

The sleep disturbances in an SMS child shift fiom prolonged napping in infancy
to a repeated pattern of naps and awakenings (De Leersnyder et al. 2001). As an SMS
child ages, the number of naps increases and sleep at night decreases. In general, there is
an overall decrease in rapid eye movement (REM) sleep. Recently, SMS children have
also been documented to have an inversion of the normal circadian rhythm of melatonin,
which is the most likely cause of their sleep disturbances (Potocki et al. 2000; De
Leersnyder, De Blois et a1. 2001). Little is known about how the melatonin secretion and
distribution pathway is affected in these individuals, though a study of SMS children
given a specific Bj—adrenergic agonist (acetbutolol) as well as an evening dose of
melatonin demonstrated markedly increased hours of sleep and decreased awakenings
(De Leersnyder et al. 2001). Some overall improvement in the inappropriate behaviors of

these children was also noted (De Leersnyder, de Blois et al. 2001).

SMS children are often diagnosed with some form of developmental delay and
mild to moderate mental retardation. During school age, the behavioral aspects of the
SMS phenotype, which include attention deficit and hyperactivity, attention-seeking
behavior, tantrums, aggression and self-injury, become pronounced and may escalate
through puberty (Smith, Dykens et al. 1998). Several self-injurious behaviors such as
self-biting, skin picking, and self-hitting are common in SMS patients. Two behaviors

unique to most SMS patients are termed onychotillomania (pulling out of nails) and

polyemboilokomania (insertion of objects into bodily orifices) (Smith, Dykens” et al.
1998; Finucane et a1. 2001). Several stereotypic behaviors such as self-hugging, teeth
grinding, and body rocking are also common to many SMS patients (Finucane, Dirrigl et

al. 2001).

Following puberty, the physical aspects of SMS, especially the facial features,
such as midface hypoplasia, prognathism, synophrys, and heavy brows tend to beOome
more pronounced. Scoliosis may also become more severe with age. The
neurobehavioral phenotype may worsen during adolescence, though the aggressive and
self-injurious behaviors eventually may lessen through adulthood (Smith, Dykens et a1.
1998). An ongoing life history study of SMS patients conducted at the National Institutes
of Health (NIH) and spearheaded by Ann Smith should provide valuable information

about adult individuals with SMS.

Determination of the molecular basis of microdeletion syndromes

Patients diagnosed with a microdeletion syndrome present with a characteristic
phenotype, which typically results from reduction of functional proteins consequent to
the loss of one copy of a gene or genes that is/are contained within the deleted
chromosomal region. Often this is due to haploinsufficiency for a dosage-sensitive gene
product or products, though it is also possible that hemizygosity for a recessive gene may
produce a phenotype if the remaining c0py is mutated or disrupted in some way. Several
microdeletion syndromes exist whose primary phenotype arises from the disruption of a

single gene, such as Alagille syndrome (AGS), which is caused by dominant mutations or

10

deletion of one copy of the human Jaggedl (JA GI) gene (Li et al. 1997) and Rubenstein-
Taybi syndrome, which is caused by haploinsufficiency of the Crab-binding protein

(CREBBP) (Petrij et al. 1995; Oda et a1. 1997).

For several other microdeletion syndromes, it is possible that haploinsufficiency
of a single gene is not enough to produce the entire phenotype (Shapira 1998). It is likely
that these are contiguous gene syndromes. This term, introduced by Schmickel in 1986,
is used to describe the involvement of multiple functionally-unrelated genes that lie close
to one another on a chromosome (Schmickel 1986) and the complex phenotype that
arises from gene dosage imbalances of several genes associated with chromosomal
deletions or duplications. A true contiguous gene syndrome contains several genes that
may contribute to the phenotype and at least one or more genes that follows a Mendelian
pattern of inheritance which can be traced in the general population (Schmickel 1986).
An example of a contiguous gene syndrome is Williams syndrome (WS), which is
associated with a common deletion of 7ql 1.2 that encompasses the elastin (ELN) locus
(Francke 1999). The ~50 kb ELN maps to the center region of Williams syndrome
critical region (WSCR) and encodes the protein tropoelastin. Haploinsufficency of ELN
alone is sufficient to account for the cardiac and vascular abnormalities that are observed
in WS patients but not for other features associated with the syndrome such as
hypercalcemia, dysmorphic facial features, and a distinct cognitive profile. Patients who
only harbor small deletions or point mutations within the ELN gene present with one of
the two autosomal dominant cardiac disorders (typically supravavular aortic stenosis or

cutis laxa) but not the other physical and neurobehavioral characteristic aspects of WS

11

(Francke 1999; Peoples et a1. 2000). The role of other genes within the WSCR are
currently under investigation in order to understand their contribution to the full WS
phenotype. At this time, it is not yet known whether SMS [del(17)p11.2] is a contiguous
gene syndrome, though data discussed in this work suggest that the majority of the SMS

phenotype may be due to haploinsufiiency of one gene, retinoic acid induced l or RAII .

Determining the primary gene or genes involved in producing the phenotype of a
particular microdeletion syndrome is an extremely complex process and multiple
approaches are combined to identify and analyze promising candidate genes. Several
‘ labs have used techniques similar to those of our own group to identify the genes
involved in AGS and WS. Increasingly (as discussed below), animal models are
becoming an invaluable resources in understanding the in vivo effects of gene dosage and
how haploinsufficency of a single gene can have pleiotropic effects. This work will
discuss AGS and WS as particular models of microdeletion syndrome analysis and
techniques we have applied to understand the molecular basis of SMS. AGS has aspects
of a microdeletion syndrome as well as a single gene disorder. The study of WS is also a
useful model for SMS, as genes other than ELN that map to the WSCR are currently
under investigation in order to understand this contiguous gene disorder. Our lab is using
similar approaches to understand how RAIl haploinsufficiency can produce the multiple
phenotypic effects of SMS and to decipher the role of other hemizygous genes in the

deleted region.

12

AGS (OMIM #118450) is an autosomal dominant disorder now commonly
diagnosed using at least three of the five major clinical characteristics first described by
Alagille in 1975: cardiac disease including pulmonic valvular stenosis and peripheral
pulmonary arterial stenosis, cholestasis, skeletal abnormalities including “butterfly”
vertebrae, ocular abnormalities, and a characteristic facies consisting of a broad forehead,
pointed mandible, deep-set eyes, and an unusual bulbous tip on the nose (Alagille et al.
1975; Krantz et al. 1997 ; Oda, Elkahloun et al. 1997). Many AGS patients present with
neonatal jaundice and cholestasis due to an insufficient number of bile ducts, which are
identified on liver biopsy (Gridley 2003). While ‘some degree of mental retardation has
been reported in a subset of patients, most individuals with AGS have normal intelligence
(Oda, Elkahloun et al. 1997). Many familial cases of AGS exist, both those with normal
karyotypes and those who harbor interstitial deletions of 20p12. The latter finding
strongly suggested that the locus for AGS mapped within this deletion region (Oda,
Elkahloun et al. 1997). Further linkage analysis in an AGS family with no detectable
deletion confirmed the locus mapped to 20p, between the genetic markers D20S59 and
D20865 (Oda, Elkahloun et al. 1997). In depth cytogenetic analysis also revealed that
two AGS families were carrying a balanced translocation with t(2;20)(q21.3;p12) which
helped to define an preliminary centromeric boundary of the AGS critical interval. The
telomeric boundary was defined by FISH analysis with a clone for the SNAP gene (Oda,
Elkahloun et al. 1997). Thus, the critical interval for AGS was defined as an ~1.3 Mb
interval between SNAP and D20S186 (Oda, Elkahloun et al. 1997). A preliminary 3.7
Mb genomic contig containing this chromosomal region was then created using yeast

artificial chromosomes (Y ACs) (Pollet et al. 1995). Using information generated from

13

this contig as well as high-resolution cytogenetic analysis of two AGS patients with
submicroscopic 20p12 deletions, one group reduced the AGS critical region to ~250 kb
and a new overlapping, genomic contig was created using bacterial artificial
chromosomes or BAC clones (Oda, Elkahloun et al. 1997). This same group had also
mapped a cDNA representing the human homolog of the rat Jaggedl gene (JA G1) to this
AGS critical region (Oda et a1. 1997). JAG! is widely expressed and was an attractive
AGS candidate gene because it encodes a ligand for the Notch transmembrane receptor,
which has been reported to be critical for the determination of certain developmental cell
fates (Oda, Elkahloun et al. 1997). The genomic structure of human JAGI was
determined and detrimental dominant mutations (including frameshifis, and splice donor
mutations) were identified within the coding region ofJA G1 in several AGS patients with
no cytogenetic deletion (Oda, Elkahloun et al. 1997). A separate group also
simultaneously identified distinct JA 0] coding mutations in four unrelated patients (Li,
Krantz et a1. 1997). Once these mutations were reported, analysis of the
genotype/phenotype profile of AGS intensified as well as an attempt to further
understand the expression pattern and cellular role of the JAGl protein. A recent
extensive survey of the JA GI mutations in AGS summarized the types and frequencies of
sequence changes found in patients. Briefly, the findings revealed: 72% of the
mutations led to a premature stop codon, 15% were splice site mutations, 13% were
missense mutations and 3-7% of patients carried deletions of the entire JAGI gene
(Spinner et al. 2001). While the JAGl protein does contain several well-conserved
domains, mutations were found to occur throughout the gene and not within a particular

domain (Piccoli et al. 2001). It is also important to note that JAG] mutations have only

14

been found in ~70% of AGS patients (Piccoli and Spinner 2001; Gridley 2003). The
etiology of AGS in these patients is currently unknown, though it is possible that they
may harbor cryptic mutations that may affect the regulation or expression of JA GI .
However, extensive JA G1 mutation screening has revealed no clear genotype/phenotype
correlation. Analysis of AGS is also complicated by the high penetrance but variable
expressivity of individuals carrying JAGl mutations (Gridley 2003). Familial mutations
are common but phenotypic features may vary greatly between patients, and not all
family members harboring presumably deleterious mutations may meet the clinical
criteria of AGS (Gridley 2003). Another complicating factor in JA G1 mutation screening
is the clinical significance of AGS missense mutations. These standard criteria are
typically used to determine whether a missense mutation is disease causing: a.)
determination that the sequence change occurs within an evolutionary conserved amino
acid, b.) segregation within a family with disease phenotype, and c.) absence of the
change within the general population (Piccoli and Spinner 2001). Several studies have
discovered particular JAGl missense mutations that segregate only with particular
cardiac features, notably pulmonic stenosis or tetralogy of Fallot and hypoplatic
pulmonary arteries but not the entire AGS syndromic features (Krantz et al. 1999).
Another family harboring a particular JAGl missense mutation was found to exhibit
hearing loss, vestibular defects, and congenital heart defects (Le Caignec et al. 2002).
These important mutation studies demonstrate that there is much to be learned about the
cellular role of JAGl and that genetic modifiers may greatly affect the clinical phenotype
of AGS. This hypothesis is supported by animal models of AGS (discussed below)

(Gridley 2003).

15

The molecular analysis of WS (OMIM #194050) has many similarities to AGS as
well as SMS. The main difference between WS and AGS is that haploinsufficiency of a
single gene, JA G], is thought to be responsible for the majority of the AGS phenotype,
and multiple genes are believed to contribute to the complete WS phenotype. WS is
associated with ~1.6 Mb heterozygous deletion of human chromosome 7 band ql 1.23 and
diagnosis for WS is typically made using FISH with cosmid probes for ELN (Francke
1999). Physical features of WS include a flat nasal bridge, a short, up-tumed nose, a
long philtrunr, full lips and lower cheeks and a characteristic small chin as well as a
cardiac abnormalities (most frequently supravavular aortic stenosis or SVAS),
hypercalcemia, constipation, and impaired vision (Osborne 1999). Most WS patients also
have mild mental retardation, but the distinct behavioral and cognitive profile is perhaps
the most intriguing aspect of WS, as it is composed of friendliness and anxiety as well as
relatively high performance in linguistics but severely impaired visual-spatial abilities
(Francke 1999; Osborne 1999). WS patients have motor delay and ~70 % suffer from
attention deficit and hyperactivity disorder (ADHD) (Francke 1999; Osborne 1999). In
order to understand the molecular mechanism of WS, as with SMS, deletions in WS
patients were analyzed to determine the extent of a particular deletion and eventually to
create a physical and transcription map of the deleted region (Francke 1999; Peoples,
Franke et al. 2000). While unusual deletions identified in several SMS patients or those
patients harboring overlapping deletions within l7pll.2 greatly enhanced the genomic
and phenotypic analysis of the SMS critical interval (Bi et al. 2002; Vlangos et al. 2002),

deletion sizes in WS were found to be extremely uniform (Perez Jurado et al. 1996;

16

Francke 1999). Similar to the SMS-REPS flanking the common deletion on 17p, LCRs
of ~320 kb containing several genes, portions of genes, and pseudogenes, including
NCFI and GIFZI, as well as members of the PMSZ mismatch repair family of genes, also
flank the WS deleted region (DeSilva et al. 1999; Francke 1999; Stankiewicz and Lupski
2002). While these repeated segments made assembly of BACs and PACs within the
flanking region of the WS deletion difficult, the creation of fluid, overlapping physical
and transcription maps of the unique sequence within the WSCR was relatively
straightforward, using a combination of the molecular biological techniques and in silica
data from the human genome project (Hockenhull et al. 1999; Peoples, Franke et al.
2000). Once the physical and transcription map of the WSCR had been created and ~19
transcripts were mapped to this region (Karmiloff-Smith et al. 2003), the next major
challenge became to understand the genotype/phenotype correlation between
haploinsufficiency for genes within the WSCR other than ELN. The phenotypic
characteristics associated with deficiency of tropoelastin are cardiac abnormalities,
hernias and possible premature aging (Karmiloff-Smith, Grant et al. 2003). The role of
the other ~18 genes within the WSCR is still unknown, though Lim-kinase 1 (LIMKI)
has been} discussed as a possible candidate gene for the cognitive defects in WS, as this
gene is highly expressed in neurons and has been shown to phosphorylate cofilin, which
may be involved in regulation of actin reorganization (Kanniloff-Smith, Grant et al.
2003). However, several high-functioning SVAS patients with an above-average
cognitive profile and no spatial impairment have been reported to be deleted for a portion
of the WSCR, including LIMKl, so perhaps haploinsufficiency of LIMKI alone is not

enough to produce the mental defects seen in WS patients (Hockenhull, Carette et al.

17

1999; Kmmilofi-Smith, Grant et al. 2003). While to date, no genes other than ELN have
been definitely implicated in the WS phenotype, it is interesting that a novel, putative
transcriptional regulator, WBSCR9, now named BAZIB, or WSTF, has been mapped to
this critical region (Peoples et al. 1998). This novel gene contains numerous structural
motifs, including a bromodomain, putative nuclear localization signals, nuclear receptor
binding motifs, and a plant homeodomain or PHD finger domain (Peoples, Cisco et al.
1998; Franke et al. 1999), which is also predicted to be present in RAIl, the gene we
believe to be responsible for many of the phenotypic characteristics of SMS. Also
similar to RAH, BAZIB is widely expressed throughout development in human and mice
(Peoples, Cisco et al. 1998). The PHD domain has been implicated in protein-protein
interactions and is present in some proteins involved in chromatin-remodeling (Aasland
et al. 1995). The structure of the PHD domain within the BAZIB protein has been
crystallized, revealing a core Cys4-His-Cy53 zinc binding domain with the ability to
coordinate two Zn2+(Pascual et al. 2000). Outside of the core zinc finger are two variable
loops which are thought to confer specificity to binding ligands (Pascual, Martinez-
Yamout et al. 2000). While highly speculative, it would be fascinating to discover a
similar cellular function by BAZIB and RAII and perhaps a similar role in mental or
behavioral development. In both WS and SMS, current assessment of promising
candidate genes is being carried out using mouse models in order to link complex

phenotypes to specific developmental pathways.

To further complicate the phenotypic and molecular analysis of a certain

microdeletion syndrome, genomic imprinting may contribute to the overall phenotype.

18

Genomic imprinting is a normal process wherein a particular gene is “functionally
hemizygous and retain[s] molecular memory of [its] parental origin throughout
embyrogenesis using parental and allele-specific DNA methylation” (Khan et a1. 1999).
A great deal has been discerned about the normal imprinting process from classic studies
of imprinting disorders such as the Angelman syndrome (AS) and Prader-Willi syndrome
(PWS). Both of these disorders can be associated with a similar ~4 Mb deletions on
chromosome 15q1 1-qlB, although the former is typically derived from a maternal
deletion and the latter by a paternal deletion (Mann et al. 1999; Ohta et al. 1999; Tsai et
al. 1999; Butler 2002). Despite a common microdeletion, the phenotype and molecular
basis of AS and PWS appear to be vastly different. AS patients display severe cognitive
impairment, ataxia, and inappropriate laughter (Khan and Wood 1999), which most likely
results from the lack of expression from the maternal allele of E6-AP ubiquitin protein
ligase (UBE3A) (Kishino et al. 1997; Matsuura et al. 1997). UBE3A has been found to be
paternally imprinted only in the brain (Rougeulle et al. 1997; Vu et al. 1997). Deleterious
frameshift mutations in UBE3A identified in AS patients without a deletion indicate that
this is a single-gene disorder (Kishino, Lalande et al. 1997; Matsuura, Sutcliffe et al.
1997), whereas PWS may be a contiguous gene disorder. The phenotype of PWS
includes hyperphagia, severe obesity, hypotonia, short hands and feet, hypogonadism and
a typical facies, and no patients have been identified to .date with a single gene mutation
(Vogels et al. 2002). Currently, several paternally expressed genes in the 15q11-q13
region are under investigation as PWS candidate genes (Cassidy et al. 2000). As SMS
patients have been identified with both maternal and paternal deletions and the

phenotypic characteristics of all patients are similar, imprinted genes are not believed to

19

contribute to the SMS phenotype (Greenberg, Guzzetta et a1. 1991) and will not be

discussed further in this work.

Animal models of microdeletion syndromes .\

The monOallelic microdeletion of chromosome 22q11.2 or del(22)q11 is thought
to be one of the most common genomic disorder(s), estimated to occur in ~l :4000 live
births (Lindsay et al. 1998; Yamagishi 2002). Several multiple congenital anomalies
syndromes have been separately mapped to this region, including the DiGeorge syndrome
(DGS), velo-cardio-facial syndrome (V CF S), as well as conotrucal anomaly face
syndrome. Many of these disorders share overlapping clinical features. As molecular
cytogenetics have improved, it was. determined that many of this spectrum of disorders
shared a common deletion region within 22q11.2 and is now known as 22q11 deletion
syndrome (hereafter referred to as 22ql IDS) (Y amagishi et al. 2003). While the clinical
findings of 22q11DS are variable, approximately 75% of patients have congenital heart
defects, typically of the cardiac outflow tract and aortic arch. Patients also have
characteristic facies, immunodeficiency, hypocalcemia, and developmental and
behavioral problems (Y arnagishi and Srivastava 2003). To date, ~30 genes have been
mapped to the 22ql 1DS critical interval and direct sequencing of candidate genes in DGS
and VCF S patients with no detectable deletion has not revealed deleterious mutations,
suggesting perhaps that haploinsufficiency of multiple genes may produce the phenotypic
features (Y amagishi and Srivastava 2003). As 22q11DS is an extremely complex

genomic disorder, animal models have played a crucial role in identifying the in viva role

of promising candidate genes. Several candidate genes, including HIRA, UFDIL and

20

CRKL, were separately targeted in mice and heterozygotes were not found to be
haploinsufficient, though it remains a possibility that these genes may modify the human
phenotype (Lamour et a1. 1995; Pizzuti et al. 1997; Roberts et a1. 1997; Wilrning et al.
1997; Lindsay and Baldini 1998; Farrell et al. 1999; Yamagishi et al. 1999; Guris et al.
2001; Lindsay et al. 2001; Yamagishi et al. 2003). In order to reproduce the 22q1 1DS
phenotype in an animal model, several groups have applied sophisticated genetic
manipulations to remove the orthologous chromosomal region in mice (V itelli et al.
2002). Briefly, this embryonic stem (ES) cell technology involves Cre-mediated loxP
recombination (V itelli, Lindsay et al. 2002). For example, in one targeting experiment, a
loxP site was engineered into the Hira gene and one was inserted between the Gbplbb
and the Cld5 gene, ~150 kb centromeric to the adjacent loxP site (V itelli, Lindsay et al.
2002). Cre recombinase specifically targets the genomic region between the loxP sites
for deletion (V itelli, Lindsay et al. 2002). Phenotypically, the mice carrying the
heterozygous deletion of the chromosomal region orthologous to 22q11DS displayed
somewhat more subtle features than human 22q11DS patients, though several significant
phenotypic features were reproduced, such as aortic arch abnormalities, and parathyroid
and neurobehavioral defects (Lindsay 2001; Paylor et al. 2001; Taddei et al. 2001). This
same mouse ES cell engineering technology is currently being used in SMS studies by
the Lupski research group at the Baylor College of Medicine to generate mice
hemizygous for the region of mouse chromosome 11 which is orthologous to the SMS
common deletion region (Walz et al. 2002; Walz et al. 2003), as well as to create nested
deletions within-this genomic region, in order to more precisely map phenotypic effects

of certain genes.

21

In parallel with large-scale analysis of the orthologous chromosomal region to the
22q11DS critical region in mice, targeted disruptions of single candidate genes can
provide valuable information about the phenotypic contribution of each particular gene.
Through gene-targeting, disruption of one promising candidate gene, the T-box
containing transcription factor (Tbxl ), was also found to produce aortic arch defects
(Lindsay, Vitelli et al. 2001). Tbxl is a member of a family of proteins containing a T-
box, a domain which has been implicated in DNA-binding as well as protein-protein
interactions (Lindsay, Vitelli et al. 2001; Yamagishi and Srivastava 2003). Mice
hemizygous for Tbxl had aortic arch defects and homozygous mice displayed an even
more severe phenotype, reproducing most of the cardiac abnormalities seen in 22q11DS
as well as craniofacial abnormalities, cleft palate, and hyploplasia of the thymus and
parathyroid glands (Lindsay and Baldini 1998; Guris, Fantes et al. 2001; Jerome et al.
2001; Lindsay, Vitelli et al. 2001; Merscher et al. 2001). While some evidence suggests
that Tbxl is not expressed in cardiac neural crest cells, Tbxl expression is required for
normal aortic arch development, as only a single aortic arch was present in Tbxl
homozygous mutant cells (Lindsay and Baldini 1998). It is possible that Tbxl has the
ability to trigger cellular signalling to induce artery development, affecting several
molecules such as FgB. Again, a genetic link between Tbxl and Fgf8 was demonstrated
through an animal model: mice doubly heterozygous for Tbl and 17ng had significantly
more severe cardiovascular defects than single gene knockouts (V itelli et al. 2002).
While Tbxl may play a role in the 22qllDS, further analysis is needed to fully

understand its contribution to the development of this disorder, as at least one patient has

22

been reported with the full spectrum of cardiac and craniofacial anomalies who carries an
unusual 22q11DS deletion which does not involve TBXI (Y amagishi, Garg et al. 1999).
A search for IBXI mutations in individuals with 22q11DS but with no detectable
deletion or patients with congenital heart disease did not reveal obvious detrimental
mutations, though the significance of several missense sequence changes remains to be
determined (Lindsay, Vitelli et al. 2001). Further animals are also being designed in
order to screen genetic modifiers of Tbxl as well as to better understand embryonic
cardiac development. These critical animal studies of 22q1 1DS, combining large-scale
deletion analysis as well as single candidate gene knockouts, are a model for the overall

in viva analysis of candidate genes in SMS.

Animal studies of AGS involving the mouse Jag! gene have also proved to be
quite complex. As mentioned above, a family carrying a particular missense mutation in
JAGl displayed hearing loss and certain vestibular abnormalities, as well as cardiac
defects (Le Caignec, Lefevre et al. 2002). Vestibular defects were also recapitulated in
two separate mouse lines carrying randomly generated Jag] missense mutations (Kieman
et al. 2001; Tsai et al. 2001). While human JAGl mutation screening supports the
hypothesis that JAGl haploinsufficiency is causative for AGS, gene targeting in mice to
reduce Jag] expression revealed that homozygous gene-targeted (or knockout) mice died
in utera and Jagl/+ heterozygotes displayed some anterior chamber eye defects but
otherwise were a disappointing phenotypic model for the other features of human AGS
(Xue et al. 1999). This result is not unique to AGS, as several animal models for
recessive metabolic disorders exists, which essentially do not recapitulate the human

phenotype (Elsea et al. 2002). However, a separate mouse model was engineered that

23

combined a Jag] mutation as well as a NatchZ (a member of the Notch family of
transmembrane receptors that interacts with ligands such as Jagl) hypomorphic allele.
These mice exhibited several of the phenotypic features similar to AGS patients such as
jaundice, bile duct, heart, eye, aird kidney abormalities (McCright et al. 2001) and
enhanced the understanding of the Notch developmental pathway. Further animal studies
are ongoing to understand the role of other genetic modifiers that may also influence the

variable phenotypic expressivity seen in human AGS patients.

In order to analyze WS candidate genes, mice hemizygous for elastin (Eln) were
also found to be clinically normal, though some histological evidence demonstrated
increased elastin lemellae in the artery walls (Li et al. 1998). Mice homozygous for the
Eln disruption showed increased thickness of the arterial wall due to smooth muscle
accumulation afier embryonic day 17.5 and died soon after birth (Li et a1: 1998; Li, F aury
et al. 1998). Similar to human patients who carry ELN mutations, mice hemizygous for
Eln did not display obvious intellectual impairment (Li, Brooke et a1. 1998). A possible
candidate gene for the cognitive abnonnalies found in WS, Cyln2, which encodes Clip-
115, a microtubule-binding protein, was also removed in mice by gene-targeting
(Hoogenraad et al. 2002). Mice hemizygous for Cyln2 display some characteristics
similar to WS patients, including growth deficiency, motor deficits, and several brain
abnormalities, which were assessed with several sophisticated behavioral tests ”
(Hoogenraad et al. 1998). As discussed in Chapter IV, we also intend to apply several
behavioral analyses to discern subtle mental and behavioral defects in gene-targeted and

transgenic mouse models for SMS. Animal models such as these are very important to

24

the identification of new cellular pathways which underlie complex neurobehavioral

development.

_ Another form of mouse in viva gene analysis, BAC transgenesis, will also be
discussed in Chapter IV, and haw the Elsea laboratory used this technology to determine
the dosage sensitivity of one of the Candidate genes in the SMS deletion region. While
BAC transgenics do not appear to be widely used except for certain complementation
experiments, they can be useful for preliminary transgenic studies without complex
genetic engineering (Heintz 2001). In a classic example, the Camper lab introduced a
BAC containing the unconventional myosin My015 gene into a line of shaker-2 mutant
mice (Probst et al. 1998). The shaker-2 mice, whose phenotype included deafiiess and a
characteristic circling behavior, were found to harbor a mutation within the motor domain
of My015. The transgenic BAC supplied the wild type form of My015 and fully restored
normal hearing and behavior to these mice (Probst, Fridell et al. 1998). In another
example, a group studying DGS syndrome used human BACS overexpressing several
human 22qu candidate genes to assess which genes in the region produced measurable
phenotypic effects when present in multiple copies (F unke et al. 2001; Merscher, Funke
et al. 2001). The transgenic animals displayed severe inner ear defects and hyperactive
behavior (F unke, Epstein et al. 2001). The expression pattern of Tbxl suggests that this
gene may be responsible for the ear disorders in the BAC transgenic mice as well as the
hearing loss present in some DGS patients (Funke, Epstein et al. 2001). As described in

Chapter IV, we used a similar analysis to examine overexpression of mouse Rail.

25

Conclusions

As highlighted in the examples above, there are several issues to consider when
undertaking a genetic assessment in a live animal model such as mice, including strain
difl'erences, the lack of an obvious phenotype, and the role of genetic modifiers. In the
development of mouse models for SMS candidate genes, we are also concerned about
embryonic lethality. Several of the SMS candidate genes targeted in mice were found to
be essential for embryonic development (V langos et al. 2003); we believe there is a
possibility that true knockouts for many of the promising SMS candidate may die in
utera. In that case, more sophisticated chromosomal engineering techniques utilizing the
Cre and loxP recombination system may have to be applied to specifically remove a
particular candidate gene at certain temporal stages, in order to properly assess the effect

of dosage on the developing mouse embryo.

26

Chapter II.
Fine-mapping the SMS critical interval and identification of SMS

candidate genes

The research goal in our lab is to understand the underlying molecular and
biochemical basis of SMS, and the cellular basis of the complex physical and
nembehavioral phenotype manifested by SMS patients. In order to proceed effectively,
we required a reliable physical and transcription map of the smallest deleted region along
17Pll:2 which could still reproduce the entire SMS phenotype, or what is termed the

critical intervalior region. Previous to this work, a great deal of effort was put forth to
ideal“)! the critical interval and to begin to map genes, ESTs, and genomic markers to the
SMS region of 17p. A preliminary genomic contig was also created. The first major
focus of my work was to complete the fine-mapping of the critical interval and enhance
the transcription map, simultaneously identifying the genes and expressed transcripts
which localize to this region. We would then prioritize these genes as possible SMS

candidate genes.

Definition of the SMS critical interval

The definition of the SMS critical interval (published in 1997) was made possible
by advances in FISH detection, which identified SMS patients with smaller or unusual
17p11.2 deletion (Juyal et al. 1996; Elsea et al. 1997). In order to further analyze these
deletions, somatic rodentzhuman hybrid cell lines that retain the deleted copy of

chromosome 17 were developed from 30 patients (Guzzetta et al. 1992; Elsea, Purandare

27

et a1. 1997)- These cell lines were created through a fusion and selection process in
WhiCh an initial, single cell with two nuclei (human and rodent) is formed. After several
1'0st of division, human chromosomes are lost randomly. The hybrid eventually
stabilizes and very few human chromosomes remain, making it possible to analyze the
genes Present on a single chromosome more efficiently (Kao et al. 1976). The SMS
common deletion interval (typical for ~75% of SMS patients carrying deletions) is
bordered by marker D17S58 (proximal) and cosmid cCIl7-498 (distal) (Juyal, Figuera et
al- 1996; Elsea, Purandare et al. 1997). The SMS patient with the smallest deletion
detected to date, HOU142-54O represents the ~1.5 Mb SMS critical interval which lies
between marker D17S29 (centromeric) and cosmid cCIl7-638 (telomeric) (Elsea,
yorandare et al. 1997). Irnmortalized cell lines from 2 non-SMS patients with
oveflaPping deletions along 17p11.2-p12 were also created, HOU92-357 (Hy357-2D)
afld HOU261-765 (Hy765-18D), for further subdivision of the critical interval (Juyal,
Figuera et al. 1996; Elsea, Purandare et al. 1997). Eventually, a 17p mapping panel was
created that also included a partial 17p arm control (88H5), and an iso-17q (LS-1) (Elsea,

Purandare et al. 1997).

Once a somatic cell hybrid mapping panel was established, the lOcalization of
numerous markers along chromosome 17p11-p12 was carried out, including short tandem
repeats (STRs), sequence-tagged sites (STSs), expressed sequence tags (ESTs) and
known genes (Elsea, Purandare et al. 1997). As of 1997, within the ~1.5 Mb critical
interval, 23 genes/ESTs were mapped that were deleted in all patients. This was

accomplished through polymerase chain reaction (PCR) fine mapping with primer sets

28

fi‘Om various markers to DNA from the rodentzhuman somatic cell hybrid paIIEI. The
PWSence or absence of a marker in naturally-occurring deletions along 17p (from the
hybrid panel) also demarked "bins," which divided the common and critical deletion
regions. Additional mapping information was gained from restriction fragment length
Paymorphisms or FISH analysis with cosrnids representing various genomic loci (Elsea,
Pumdare et al.1997).

Physical and transcription map of the SMS critical interval
The definition of the SMS critical interval and a method to efficiently map
market‘s Within this critical region established a basis for the contiguous physical and
“anscfiption map of the SMS critical interval. After difficulty was encountered in
attempting to reproduce a yeast artificial chromosome (Y AC) map of the region (Chen,
Manian et a1. 1997), large insert genomic clones such as bacterial and Pl-artificial
chromosomes (BAC/PACs) were chosen to create the physical contig of the SMS critical
interval. BACs and PACs within the contig were assembled through a combination of
Southern hybridization, PCR, and human genome project sequence data analysis.
Initially, known genes, ESTs, and genomic markers that mapped to the critical interval
were used to screen a gridded hCITB BAC library in order to form a preliminary contig.
Overlapping BACs were identified and grouped according to Ala-PCR fingerprinting and
restriction digests. In order to map markers to the initial contig, known genes and ESTS
were hybridized to EcaRI—digested BACs or PCR-arnplified from BAC DNA. This
preliminary contig had many gaps and could not be completed with the BAC resources

that were available.

29

This work focuses on fine-mapping the SMS critical interval and completing the
genomic contig, which contained several gaps in 1998. As late as 1999, little human
genome Project sequence data were available for 17p1 1.2, most likely due to the low-

COPY repeats that are prominent in this genomic region which can complicate sequence
sequence alignment (Stankiewicz, Shaw et al. 2003). As soon as draft sequence for the
unique, gene-rich regions of the SMS critical interval was available publically through
NCBI’ We were able to take advantage of these data to improve the coverage of our
contig. We used BLAST searches with known l7p1 1.2 gene and EST sequence data to
ideflfifB' high-throughput genomic sequence (HTGS) RPCI BAC/PAC and CIT BAC data

M Were pertinent to the SMS critical interval. We subsequently purchased several of

mese Clones for analysis and utilized drafi sequence alignments from the Golden Path
(pupil/genomeucscedu) and Ensembl (http://www.ensembl.org) databases to organize
our contig. We searched BAC-end sequence information fi'om GenBank to establish a
minimum tiling path of 16 BACS and 2 PACs (Figure 2) (Lucas at al. 2001). Several

overlapping clones were also included in the map to provide greater coverage (Figure 2).

Gene order within the contig was determined through HTGS sequence analysis
and Southern hybridization for the presence or absence of a certain gene or EST in
EcaRI-digested BAC or PAC DNA. Transcription orientation was obtained from the
human genome browser database, cosmid sequencing, and known overlapping segments,
such as the 3' ends of FL]! and LLGLI. Cosmid genomic clones for several of the genes

were identified through hybridization of EST inserts to gridded filters of the Los Alamos

30

Figure 2- Tummfimo

Lucas out, 2001 . " al.,, a, we

on] ' t ”1'5 Mb SMS critical interval (adapted from
MS crltl 111 CFVa] as or
IJCtWee 2001 ' °
pans D (Oflglnan . .
1997) -BEPM) to :9 1W0 Vertical hay, ‘11ow m Juyal et al., 1996 and Elsea
. e ' c marks. approximaxeiy from the middle

llie S s
. ,RBP (SMS dlstal s -
Ms . 0,. an overlapping Co - MS REP (SMS-REPD). The transcription map is

egiOflwet’CmaP ughacob"
d PAC DNA, PCR, and databa: mathn of hybridization to £00”-

cues:
g. BAC an A“ ESTS and genes Were also h salience analysis from the human
ybrrdlzed to the somatic cell hybrid

dlge ' act. S
6 pro) I'ITG Sequence m1 . .
gcflo panel to 1006““ sequence overlaps 3101818 and POSItion within the SMS critical
“filmed the position of clones where

mg n,
Mp AC Genes, ESTS’ and mark .
B fit. al. ers Show“ m blue are localized out Of the

31

 

 

  

 

 

 

 

   

Al.
25%:ng 0 6 4 mamu—fiw—VU
n a m m g
J 1 ..VM M M Im 1v
3 1m 1. s m m 11 3 mm. 3 m.
m s 5n $9 0 ..52111 lilom 12.322510 1
2521 J 83“.). 3 2 4 h3 .um _ 14 F
73FPMFB 3W6MAZODIR3Iwfini—i'7‘snIbPRm/ m 7 00
mmmm MMflH mmémCHmmvw/mummmCDmDMImun Fanrom
I. 1 ..lm .70 GHmHLULMI L S
mmwmmsmm mmmmmummwmmwwmemmwmwnmmemwmm. W.
1 11111 “..1 111111 1.1 1-112.151 1 1:1... 11111 1 1 11....“ _.1....1 1 1 1 V
33538 ...:
comm—EA :0
combos:
“8323+
56x8 b.9051
nogzocafitu $0.3»:
GSQUNN 0;;
0952.? & >mo33m O 0
60383;.
do_:>uo DJDIDJPI} .0...
in?
canvass 3&2: IT???
goufilefllbl
008:; I0:
wouueflwgofifii
$385.0;
0 - wasp—0:8 mam—«am II .325 Wowwmm_jufl+fli__
n
M . “Elam”. I -Eaaasmg €83.83
I n I I . . ;
0 o paw w MEG—8n moneouoo cvwwwwvmw;
. g;
max .025 02—3. :2 Saw—35¢ o coawémiz
cowoq can
. 025:0 Zion 1v - Hana—ungu— it
. . anemowa
£088 82% am”.
._ -59; mg“ flag:
I m3 . €202;
. .623

 

 

flow-50m
from SHR
order. We
b}bfidizin
DNA (Lu
mprcscmi
(Lucas, W

control by?

 

8313mm.
Vlangos e1
“llhin the
MS DN‘:
EDNA clo
membrane
contig, as
PUbh'shed c
Were mum
D1737“), I

flow-sorted chromosome l7-specific cosmid library (Kallioniemi et al. 1994). The region
from SHMTI to FLII is well-represented by overlapping cosmids, which confirmed gene
order. We reconciled human genome project data with naturally-occurring deletions by
hybridizing genes and ESTs to EcoRI-digested rodentzhuman somatic cell hybrid panel
DNA (Lucas, Vlangos et al. 2001). An example is shown in Figure 3, in which an EST
representing the 3’ end of RAII is hybridized to this somatic eell hybrid mapping panel
(Lucas, Vlangos et al. 2001). RA]! is deleted in all SMS patients but is present in the

control hybrid lines, MHZZ-6 and 88H5, as well as human genomic DNA.

As published in 2001, the ~1.5 Mb SMS critical interval is represented by the
SMS patient HOU142-540, who carries the smallest deletion described to date (Lucas,
Vlangos et al. 2001). All ESTs were also hybridized to EcoRI-digested BACs and PACs
within the contig to confirm mapping positions. In Figure 4, the digested BACs and
PACs DNA fiom within the SMS critical interval is shown in (a), and in (b), an ~2.0 kb
cDNA clone containing the 3 ’ end of RAII demonstrates positive hybridization to a
membrane containing the digested BAC/PAC DNA. RAII maps to pc253P07 within the
contig, as predicted fiom human genome project draft sequence. Within this 2001
published contig, 17 known genes and 12 ESTs were identified and six genomic markers
were utilized in the generation of the physical map of the critical region: Dl78258,
D17S740, D17Sl794, D17S620, D17S447 (also called F62), and D17S71 (also called A-

1041) (Lucas, Vlangos et al. 2001).

33

Figure 3. Mapping RAII to somatic cc“ hybrids.

In order to fine-map M1 to a s ecific region withi - from
EST DKFZp434A139QZ was hybridized t0 3 pangl loll? IEICZ fiifigfgme 0‘3
hybridS, carrying deleted chromosomes 17 from SMS patient: (H $41 -20D 113540—10
an new» and WM with We deli—.... 2
(Hf/65481) 331d Hy357-ZD) as well as Chromosome 179 pa sitive controls “27.-6,
human genomic, and Hy88HS), and 17q (LS-l), ham$ter (2123) and mouse (0‘40) cell
line negative °°“"°‘s' A >20 kb RA” ”and,“ evident only ’ in the MHZZ’6a “mm
H yngS’ and HY357-2D lanes. (Present bUt falnt on this exp : sure) demonstrating that
RA]! is deleted m all SMS patients and maps to the central onion, of the SMS critical
interval between D178258 and D17S620- P

34

 

 

«Ii-rem an <—¢PIOZ<

“we“ 4‘
'3' .f - ’

l 3 "Hz; 1*‘M_ I," m V‘.
.‘ . '» - ‘ 9. M. ..I ~
'5 . ..t' “I

p—l

Ill:- 5‘

NIH22-6

ZI

Human genomic
Hy88H5

 

 

 

Hamster
3 Mouse
--E
”3:12!

35

Figure 4 .
. Mapping RAII to B ACs ,1, ACs.

(a) DNA (3
”8) from BAC - .
(10 “3) were digested with 852311313: Si“: critical interval a d . NA
was transferred to a nylon membrane u , e e t 1‘0threSed Ove .0 human genomic D
5mg S andard S Outhemmlght‘ The digested D
b aut ' tec hniques.
hipzpnmggggggspi‘ plgimitctiiei 80‘“th blot i
Red arrows indicate positive RAII use“: which repres 3 Shown, hybridized to the
roject sequence anal , hYbridization in :nts the 3 ’ end of human RAH .
Ysrs) and a faint ban d p120 53P07 (which confirms
sent in the human I
ane.

genome p .
hybridization is seen in several other lanes

human

BAC-end

36

 

    

Human genomic

bc189D22

‘ a Marker

 

» ‘ «lHuman genomic

f‘bc189D22
_ pcl78F10

bc83E23
_‘ ch8 1113

pc253P07
' ' bc80B7

bc124H2
pc962P8

_ bc524Fll
3hc57c3

 

 

V
N
O
F
U'

37

Re-definition of the SMS critical interval
Recent FISH deletion analysis by other members of the lab refined the current
SMS critical interval to a ~950 kb interval, containing 13 known genes, 12 predicted
genes, and 3 ESTs (V langos, Yim et al. 2003). One patient, SMSIBS, was identified who
displayed typical SMS physical features and behaviors but who harbored an unusual
l7pl 1.2 deletion (V langos, Yim et al. 2003). This patient’s proximal deletion breakpoint
occurred within the SMS-REPP but the distal deletion breakpoint occurred within the
~1.5 Mb SMS critical interval, proximal to RASDI (Figures 2 and 5) (V langos, Yim et al.
2003). ' Based on our lab’s analysis of the deletions carried by HOU142-540 and
SMSl35, the current SMS critical interval spans from the SMS-REPM to the genomic

region just proximal to RASDI (Figure 5) (V langos, Yim et al. 2003).

EST characterization and analysis

To define the ~1.5 Mb SMS critical interval, numerous markers were mapped
along chromosome 17p1 1-p12, including short tandem repeats (STRs), sequence-tagged
sites (STSs), expressed sequence tags (ESTs), and 12 identified genes. Binning/fine-
mapping of ESTs along l7pll.2 was accomplished through PCR, FISH, or Southern
hybridization to DNA from a panel of rodentzhuman somatic cell hybrids. In order to
more precisely map these markers to the transcription map, plasmid clones for 13 ESTs
that mapped to the critical interval were obtained commercially, and DNA was isolated.
All clone inserts were sequenced and extensive database searches of the nucleotide
sequence and 6-frame amino acid translation were conducted in order to identify

homologous genes and sequence motifs. We determined the tissue expression pattern

38

Figure 5. SMS deletion anal ‘ fi
ysns and re ned SMS ~ - - ‘ g from
Vlangos et al., 2003). crltlcal nntemn (actin c

The ~l.5 Mb SMS critical interval, which was mapped in Lu t 31 2001 is defined by
the deletion breakpoints from SMS Patient HOU142-540. §::I deletion ,anahls.‘s from
t SMSl35 re-defines the distal end of the ~950 kb SM S critical intervah which

atien . .
Extends from SMS-REPM to the.genom‘° region just Proxirr) al to RASDI. Relevant
genomic markers are included, which map to SMS Critical interval as well as the “'4 Nib
SMS proximal and distal common Interval.

39

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5
M 9 .11.. m.»
g «M m 7
n n 7 m
mam- D D mg. m m mam-
Emu _ ”MES _ _ arm—5
... 3e . . - . I 932.

 

583$—

 

 

 

 

 

z; 2:. l medias
I l mam—um
.lwmtv g @1203—
383.»—

 

wBfiEn— 02:52. 383.». Eafi nan—Be: 383.»—

40

Figure 6. Human RAII northern analysis.

The ~2.0 kb insert from EST DKFZP434A139Q2, rcpresenting the 3’ end of human
RAII , was hybridized to an adult Clontech MTN northern blot (shown in the panels on
the left) and a fetal Clontech northern blot (shown in the panels on the right). An ~8.0 kb

transcript is evident in all adult and fetal tiSSUCS. B~actin is included as a mRNA. loading
control for both blots.

41

"l”V'fl

 

"9°V‘9

 

 

 

 

 

 

4’2

   

 

 

 

 

 

 

 
 

 

Mme. aEE—Bows dingo flint—.832. 38... Ben... 553. «and
Anon—wear 888.95
Essen minus N.» we «88%. a e: Brenden“ ease Beam 3%
name." is E. agape 583
9382 5.38.“ S we 5582 a a. Seaweedae was :23
8852* Eu :2: was: BE a». mafia man—Benn agape—n Enemflmuv
llle
>8???“ mega ab 5. 5:813? m: <=m wan Q>H newsman NGKFN
889% Eudoouv mm. SMS an: memes. Emcee Banana" 23 7:33
maeoou 5 row? can“. an. I
0252:» them 58 Benn—o
UEquAwézucON mmbqwar mb E. 55813 5 w: O>Q Honour sen—awn an:
3352..» 3.5233 3:: amazon 88538 miner 3.5 5283:” ><3N$3
Run w 83:

 

 

 

 

 

  

 

 

Hes—o H. amen 393.9— 5. been" a. 3. 33:: :5 38 ...—.m 2:. 2.32: 383.». £52. .276 and: 32:52— 8
5.2:— no:%. 08% E20 van: 0388 Roxana 8 Eng: 85 385m 588:3 mono mm .838 0:880 03:0 958:
mgm 9.30m— Eaai Ade—8&8 a" a; ~88. 3085 825 Kannada moB ZOE 8:828 moan? manage 83
205 3085688? Elam. mam—Scone.

£858 «30:8 F gunmen a» a? ~89 £9908 Home—8 wear: 8 58¢ 33:8 5 5568.3 o» 3.. ~89
922.53. 83.8 e552 3 98o 83:8 E He 25 fie. $8

43

  

.r..¢C If
. 5:. J
AHMW~ LEG WV. _. . . |11u
. OCZU. v P2: CCZCZSM . .
020$ conga—.3 Qua—mzvtm WPSC~=~C 00:33.3: . .l 4-

0 DZQ Zfl‘w~ UfiOu . awn—03 0W 3.....< VG H.403: . I - <1? 2

93430.05 mrxfiuw . F. a . ..
u: .1. u ..~.:~ta

   

 

 

1% 2 . .
W (<0. . a; $3.56 C

HG 7:25.365 ENC ~

3.. z

 

Pb

    

I: I_ H q— ulh
THCHG»: H303—

 

 

 

 

 

   

 

 

 

 

 

and Eat—.52. dingo dun—8 one—6%....— 38? Son...
0253—» .5888:
323 55.8 988% :8 :2 933 26333: E 35. 9 u; g ..Hzmfl 322
00qu $388 I), 58 on: on: E: Edna 8°38
mammuuo 8 .6 3% 838»
085E #5835 :5 88:8 2.5 no:
3.38m $933.8 PM man ab .9 58°13 E a: and: BE 88. 38:2. £00.:an 55532288
0355.» wOomnuue If $2558 88028 gong—E
KQOqum mm; GAS PA 5. was Ab gong F was: #093. an. BEG—o. WOQAmwm“ @083: mocha 5 <3:an
A0352? wOoooawav 538.. 3388 BE 88— 52 25 75% ab .3 5:5: Ea gamuugaoa @385
Ill] 533 Sb 3:: Ba 8E a . Eu 25 88:8
UWMNuumav‘: 5° mar—Moog Cimouo «898 Mm .9 goat" Ea Mm? Won: 3.x. 32.81? A So an £53535 8
AOoEwgk >Euoow3 853” was: ammcom RES @885 rouge—ow A0332
mamnoov
mas—~83 mmbmqq: Guiana 833. 9.5% 8%. ".88 883. was: :28
035wa Z?» oumoa .

 

 

 

   

 

.358 N. 236— am? Emu—8.. 3. :8 Eng :5 $.31: =8 “2: Lu as 233— E8..<»_. 00:8 328 can: 032
3.985% 8 Ems.— Ea «368m 5&ow8m 98 in «98 mm .08:qu 0:858 om So 952: mg
88% » 3%»ng 93 880 cm 980 538 mmam 8.: 8533 «9588 Suggmaou

m 8.3.8. magi «88», u

. . . Q. B . .
568$ Scan :88 Egg—m8 ".33 205 88028 @088: .8838 BE 202 @8886888 we ca 2255 ”Samoa.

953. £55.38

44

‘ ff UPM- (11"- I-

ll!!!‘

 

 

 
 

  

 

 

 

   

 

 

 

 

    

    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

and 23:82. Canaan .535 Hun—.833. $38... 335 Nona—.258
Dona-E» >33: L
magma mm.» $.ch Pm. Po. ”an 9m 5 530123 K0033” 332.: mega E 9. on a. $88
3052:" E“Amiga: w: 3:: BE moi macaw rm: 3308a 02%.
E
mama 582$ N.» ma u... 5. 533% a w: H8839 32m 355 a E 2 a. 383
A0252? >393th was: 85 won: 328. 33-3-58 no: 98 “.385...
38°53 5 an»: 25 an.
L?
$9.338 572”” mogonmfiwfia noun 9 on wr Auoouv
I\|\|I' ox Summon E w: Emu—Sm
F9383 ”gran” aoaoumnufia son... 9 on E. 988
6252? Bugatnw 8» Emacs E a: nmmcam
mZQK mono Nguyen” aaBoEnana noun 9 a" E. 388
8268* weoummwv 86893: E u: 3:: 23:8
wan— moan
mama 5:338 WHEOW maaonmqfina noun 9 an up 388
8258* 558233 96338: E w: 3:: mum—5m
Ea man;
9:96 :25 ”.7?un non—3338 noun 9 on i. $88
263355 5 was: 9.wa 25
was;
- w may

Has—o u. amen 38:5. 592.2. 3. 3:2. «3:? $9.? Em N9: 1._.m 25513.». ESEE. mo=o§=m 9a

   

 

  
   

,. a... \

 

 

LEW...

22338 cm 9: 35on gm Rummage: 5% E mos Arcomm a" a; Moos. 8331 092 Wosom £08 :36qu
8 "Eu «ammo: 3. 09$ «3:3. 038 :36 go: 2928 30553 8 Ems: Ea msaim Sagan 93 can mono mm
saga Sean 2. an 853 mam 258. 5822 38? a a; 88V. 95%... 53 Egg m

202 88028 8:53 38:80. 03

45

and Iran

Hgnc
flmnsa
RAIIgt

Emcc

 

meShf

and transcript size for each EST by northern analysis of adult human and fetal InRNAs.
Figure 6 shows the multiple tissue northern (MTN) and human fetal tissue blot
expression pattern of EST DKFZp434A139Q2, which represents the 3’ end of the human

RAII gene. An ~8.0 kb transcript is observed in all adult and fetal tissues examined; this

is the expected RAII mRNA size for the major transcript (Toulouse et al. 2003).

Tables 1-3 show the current status of the genes and ESTs which were mapped to
the SMS critical interval in 2001. several transcripts, shaded in grey in each table, have
now been localized outside of the current SMS critical region. Table 1 highlights seven
ESTs that we mapped to the contig in 2001, which represent known genes. Several other
novel ESTs were mapped and characterized by our group; current expression and protein
data about these transcripts is presented in Table 2. Table 3 describes genes that have
been mapped to the ~1.5 Mb SMS critical interval by other research groups. As shown in
Table 2, several of the ESTs that we sequenced were novel or only matched other
unidentified clones in the database. It remains a possibility that some of these ESTs may
be cloning artifacts or may contain genomic contamination. As the human genome
Project and large-scale EST projects continue to deposit more information in the public

databases, a number of transcripts continue to be identified within the SMS critical
interval and expression data for the novel genes and ESTs increases. Experimental
evidence will be needed to determine if any of these novel genes may play a phenotypic
role in SMS. And while we have focused our research efforts on genes and ESTs within

the Critical interval, it remains a possibility that l7pll.2 deletion breakpoints alter

46

regulatory and binding elements, Which can affect the expression of genes outside the

deleted region. This phenomenon could contribute to the SMS phenotype.

Summary and analysis of candidate genes
Once our physical and transcription map of the SMS critical interval was
published, we began to prioritize the genes within this region as candidates for the SMS
phenotype. Several characteristics that we considered when identifying SMS candidate
genes were: expression in brain, developmental expression, known protein motifs, DNA-
binding domains, and protein interaction domains. Following the discovery that
haploinsufficiency of the transcription factor Tbxl may play an important role in
22ql 1DS, we were especially interested in putative transcription factors which may have
a global impact on embryonic development, though other signalling molecules could also
prove to be especially dosage—sensitive. However, many of genes that mapped to the
SMS critical interval were novel and showed expression pattern in every tissue analyzed -
(Tables 1-3). We relied on other factors to reduce the number of candidate genes that we
Considered in our studies. To date, genes distal to PEMT 2 are no longer considered part
Of the current SMS critical interval and are therefore a low priority for future analysis as
SMS candidate genes (Figure 2). Several of the known genes that map to the critical
interval including, T 0P3A (Hanai et al. 1996; Fritz et al. 1997; Elsea et al. 1998), FLII
(Chen et al. 1995; Campbell et al. 1997; Campbell et al. 2000; Campbell et al. 2002),
MYOISA (Wang et al. 1998; Liang et al. 1999), SREBFI (Elsea, Purandare et al. 1997),
PEA”? (Walkey et al. 1999), and COPS3 (Elsea ct al. 1999; Potocki et al. 1999; Y an et

31' 2003) have been well-characterized, and we also do not consider them priority

47

 

 

 

 

 

 

 

 

 

 

 

32.8 no.5 .5368.— flananeuwneeu meaouufiosa ”3.2.2.8
era—586a 3.28.55
Set? 20 ageing chQoao FEE C can figmboomv
2E. 2o manogmaom @2583.” $9». EX: 2 a 388
~56: 2o 3:035:28 $383 maugamm Ba $83" on an. 283
fiance—ma @388
hunch Zo 333533 2o mucosa—Sam F wage a» E. 203
bong: 93F...“ thug
oBonEo SEE
was: 2o egoawmaom 2o manage—Eon on ism—rave 2 up. Good
_ 8:53 “:2
093w Zo agogwmnom macQoBo .23: 55 2 &. Goons

 

 

Haw—a A. 5.2336." cm mam gun—Eu? mason Ema-68¢ E 32.8. .35 $65 «<3
3389 n68 Seaman. on E. 383.

48

candidates genes for SMS. As described in Table 4, several genes homologous to SMS
candidate genes, such as Top3a. F liih, My015, Srebfl, Pemt, and Cops3 have been
analyzed in gene-targeted mice and in all cases heterozygous animals appeared normal
(and healthy, which therefore suggests that these genes are not particularly dosage-
sensitive. In the case of MY015A, mutations in this gene. are causative for hereditary
recessive deafness (DFNBB), and this gene was only found to contribute to the deafness
of one SMS patient who harbored a MY015A missense mutation on this individual’s non-
deleted chromosome 17 (Liburd et al. 2001). The SHMTl protein has also been analyzed
in patient lymphoblastoid cell lines and while haploinsufficiency of SHMTl was
detected, serum folate, glycine, and serine levels were normal (Elsea et al. 1995). We
have not completely disregarded SHMTI as an SMS candidate gene, as studies in other

types of patient cell lines or cerebrospinal fluid may be more conclusive.

There are several known genes within the current SMS critical interval which
have not been extensively analyzed and that we consider to be possible SMS candidate
genes based on expression data and their putative cellular role. A summary of these 5

genes, listed proximally to distally, is given below:

LLGLI: LLGLI (GenBank NM_OO4140; Unigene Hs. 95659) is the human homolog of

the Drosophila tumor suppressor gene, D-lethal (2) giant larvae. When disrupted, D—lgl
can PTOduce tumors in imaginal disks and abnormal transformation of the adult Optic
centers in larval brains (Strand et al. 1995). Human LLGLI is expressed in brain, kidney,
and muSCle, and shows an association with the cytoskeleton (Strand, Unger et al. 1995)-

Alltlbodies against LLGLI were able to co-immunoprecipitate LLGLI and n’mUSC‘e

no

49

myosin heavy chain; this interaction is conserved between human and Drosophila
(Strand, Unger et al. 1995). Both Drosophila and human LLGLl are also associated with
a serine kinase, which is able to specifically recognize and phosphorylate serine residues
within the protein (Strand, Unger et al. 1995). The LLGLI cDNA encodes a protein of
1057 aa, with a predicted molecular weight of 115 kDa (Strand, Unger et al. 1995).
LLGLI was mapped to chromosome 17p1 1.2 by FISH and Southern blotting (Strand,
Unger et al. 1995), and was demonstrated to overlap at the 3' end with the FLII gene
(Campbell, Fountain et al. 1997). LLGLI is transcribed in the opposite orientation as
FLII, just telomeric to FLII in the SMS critical interval, and is found on bc347A12,
bc189D22, and pc178F10 (Figure 2). The contribution of this gene to the SMS

phenotype is unknown.

DRGZ: Directly adjacent to MY015A is developmentally-regulated GTP-binding protein
2 (DRGZ) (GenBank NM__001388; Unigene Hs.78582) (Liang et al. 1999). Fine
mapping of this gene, represented by the EST WI-13499, confirms that DRGZ maps to
Up] 1.2 (V langos et al. 2000), within the SMS critical region. DRGZ maps to bc347A12
and bc189D22 (Figure 2), as well as a region of genomic sequence that was generated
through MY015 analysis (GenBank AF051976). The DRG2 protein is a member of a
novel family of GTP-binding proteins that are highly conserved and may play a role in
oncogenesis (Schenker et al. 1994). DRG2 is highly homologous to DRGI, a
developmentally regulated gene that is expressed in embryonic and neoplastic tissue and
down-regulated in adult tissues (Sazuka et a1. 1992; Sazuka et al. 1992). Northern

analysis of DRGZ shows low levels of expression in all adult and fetal tissues examined

50

(Schenker, Lach et a1. 1994; Vlangos, Das et al. 2000). Recently, cloned Xenopus laevis
drgl and drgZ indicates that the recombinant XDRGl and XDRG2 proteins both have in
vitro RNA binding ability (Ishikawa et a1. 2003). The in vivo cellular role of this

extremely well-conserved gene is currently unknown.

AH’AFZ (formerly ATPIZ): ATPAF2 is the human homolog of the yeast nuclear gene
Ath 2p, which is required to mediate the formation of the F10‘ subunit of the
mitochondrial ATPase (Wang et a1. 2001; Bi, Yan et a1. 2002). The cellular role 0f yeast
Atp12p appears to be limited to ATP synthase assembly (Wang, White et al. 2001). We .
showed that the EST D1782021, which represent AH’AFZ, is expressed on a northern
blot in all tissues examined (Table 1); further studies using quantitative reverse-
transcription PCR indicated that expression of mouse Atpafl mRNA/varied up to 30-fold.
The highest mRNA expression was found in brown adipose tissue (Pickova et al. 2003).
The human ATPAF2 sequence contains a mitochodrial targeting fimction and a human
ATPAF2 cDNA has the ability to rescue a yeast atpaf2 mutant, demonstrating an
equivalent cellular function (Wang, White et a1. 2001; Ackerrnan 2002). The effect of
ATPAF2 haploinsufficiency in SMS patients is not yet known, though it is possible that
the hypotonia demonstrated in SMS patients could be related to abnormal mitochondrial

fimction mediated by ATPAF 2 haploinsufficiency.
TOMILZ: Our group mapped the target of mybl-like 2 (TOM1L2) ESTs AOO3A44 and

stSG9692 to the SMS critical interval (Figure 2 and Table l), and the full-length human

and mouse transcripts of TOM1L2/Tom112, were subsequently assembled by the Lupski

51

 

 

 

group and determined to share a similar genomic structure (Bi, Yan et al. 2002). Little is
known about the TOM 1 L2 protein, except that VHS and GAT domains are present,
which have been implicated in intracellular trafficking and sorting (Bi, Yan et al. 2002).
TOMILZ appears to be expressed ubiquitously, though at higher levels in heart, brain,
and skeletal muscle (Table 1). The VHS domain, which is named for three hallmark
proteins which contain this domain, Vp527, Hrs, and STAM (Lohi et al. 1998; Lohi et al.
2002) is found in several protein complexes which can interact to bind ubiquitin and sort
ubiquinated proteins through the endosome (Yamakami et a1. 2003). The GAT domain is
found in proteins which have been implicated in sorting receptors through the trans-Golgi
network (Shiba et a1. 2003; Yamakarni, Yoshimori et al. 2003). T 0M1L2 was named for
its similarity to the T 0M1 gene, a specific target for the v-myb oncogene (Y amakami,
Yoshimori et al. 2003). While Tornl was recently shown to bind directly to ubiquitin and
to interact with the Tollip protein in a yeast two-hybrid screen (Y amakami, Yoshimori et
al. 2003), the specific cellular role of TOM1L2 is unknown. The Elsea lab is currently
performing in vitro and in viva analysis of TOM112/Tom112 in order to understand

whether cellular trafficking may be disrupted in some SMS patients.

RAII: The identification and characterization of RA]! will be discussed at length in the

following chapter.

Known genes originally mapped to the SMS critical interval
Three EST transcripts, NIB1041, WI-11472, and stSG26124 which were mapped
to the ~1.5 Mb SMS critical interval contig have subsequently been recognized by other

groups to be known genes, RASDI, NT 5M, and M-RIP, respectively (Table 1). Though

52

these genes now map out of the current SMS critical interval, the cellular role of these
genes is under investigation by other research groups and may prove to play a role in the

modulation, of the SMS phenotype.

J RASDI (formerly AGSI/DEXRASI): We previously localized the EST marker NIB1041
to the SMS critical interval by PCR to somatic cell hybrids and by Southern blotting to
BAC DNA (Elsea, Purandare et al. 1997). NIB1041 maps to the distal end of the 2001
critical interval and to bc524F11, bc3073J20, and bc40123 in our transcription map
(Figure 2). Sequence analysis in our laboratory has determined that the ~1.5 NIB1041
insert is a 99% match to the cDNA for activator of G-signaling 1 (AGS!) (GenBank
AF069506, Unigene Hs.25829). The 1740 bp cDNA sequence for AGS] also matches
the cDNA sequence for dexamethasone—induced ras-related protein (DEWSI)
(GenBank AF262018, Unigene Hs.25829), although the complete DEHASI cDNA
sequence in GenBank is 5141 bp long. Unigene has grouped AGS] and DER/1S1
together in the same entry, indicating that these sequences represent the same gene.
AGS] (activator of G-protein signaling), was isolated, through a Saccharomyces
cerevisiae screen to identify mammalian nonreceptor activators of G-protein signaling
pathways (Cismowski et a1. 1999). AGSl, a ras-related protein, is able to process
guanosine triphosphate exchange of the heterotrimeric G protein G, subunit (Cismowski, -
Takesono et al. 1999; Cismowski et al. 2000). DexRasI was discovered independently in
mice through differential display analysis to identify genes that were induced by
glucocorticoid hormone (dexamethasone) treatment in AtT-20 cells (Kemppainen et al.

1998). Murine DexRasI is expressed in heart, brain, and liver, and DEHASI mRNA

53

levels in these tissues significantly increased after glucocorticoid treatment (Kemppainen
and Behrend 1998). Human DEXRASI cDNA was isolated through a yeast two-hybrid
screen and subsequently cloned (Tu et al. 1999). The mouse and human DEXRASl
protein are 98% identical at the amino acid level and both are members of the Ras
superfamily of GTPases (Tu and Wu 1999). Expression of human DWASI was also
shown to be strongly stimulated by dexamethasone, in a variety of tissues (Tu and Wu
1999). Northern analysis demonstrated DEXRASI expression in all adult tissues
examined (Tu and Wu 1999). This result is in agreement with our northern analysis of
EST NIBlO41 (Table 1). Recent investigation into the role of rat Dexrasl in signaling
demonstrates an interaction between the neuronal nitric acid (N 0) synthase adaptor
protein CAPON and Dexrasl to enhance NO signaling (Fang et al. 2000). Human
DERASI may also play a role in cell adhesion and extracellular matrix interactions, (Tu
and Wu 1999) though the contribution to the SMS phenotype is unknown. In 2001, the
Human Gene Nomenclature Committee formally changed the name of AGSI/DERASI
to RASDI. Recent work on RASDI has identified a functional glucocorticoid response
element in the 3’ untranslated region of the human RASDI gene (Kemppainen et a1.
2003) and demonstrated that mouse Rasdl is expressed in a circadian manner in the
suprachiasmatic nucleus of the brain (Takahashi et al. 2003). The latter finding may be
relevant to SMS research because of the inverted circadian rhythm seen in SMS patients,
though this gene has recently been mapped out of the SMS critical interval and may not

be responsible for the majority of sleep abnormalities in SMS patients.

54

NT5M (WT-2): Our studies mapped the EST WI-11472 to the distal region of the 1997
critical region, adjacent to COPS3 (Figure 2) (Elsea, Purandare et al. 1997). BLAST
searches revealed that that the sequence of the ~2.1 kb WI-11472 insert is a 99% match
to the cDNA for deoxyribonucleotidase-2 (M—Z/NTSM) (GenBank NM_020201,
Unigene Hs.16614), which codes for a mitochondrial deoxyribonucleotidase (Rampazzo
et al. 2000). Independently, Rampazzo et al. mapped NT 5M to the SMS region based on
sequence analysis of a genomic clone (GenBank AC006071) and the proximity of NT5M
to COPS3, which is transcribed in the opposite orientation (Rampazzo, Gallinam et al.
2000). Deoxyribonucleotidases hydrolyze the monophosphate ester linkages of the 5' or
3' carbon of deoxyribonucleotides, producing an inorganic phosphate and the
corresponding deoxyribonucleoside (Paglia et al. 1984). The full-length cDNA for
NT SM is 1617 bp long and contains a N-terminal mitochondrial leader sequence
(Rampazzo, Gallinaro et al. 2000). Northern analysis of NT 5M revealed high expression
in the heart, brain, and skeletal muscle, a "typical" pattern for a mitochrondrial enzyme
(Rampazzo, Gallinaro et al. 2000). These data are in agreement with our northern
analysis of EST WI-11472 (Table 1). The authors demonstrated that NT5M is targeted to
the mitochondria by GFP-fusion localization and by in vitro incubation of the
recombinant protein with mitochondria (Rampazzo, Gallinaro et al. 2000). Excluding the
mitochondrial leader sequence, NT 5M shares a 52% amino acid identity to dNT -1, a
ubiquitous, cytoplasmic deoxyribonucleotidase (Rampazzo, Gallinaro et al. 2000).
NT5M activity is limited to the dephosphorylation of thymine and uracil
deoxyribonucleotides and may play an important role in modulation of (111? substrate

pools and removal of excess d'ITP fiom the mitochondria (Rampazzo, Gallinaro et a1.

55

2000). A potential role in SMS is yet to be identified, though, similar to ATPAF2,

modulation of mitochondrial function by NT5M may be affected in SMS patients.

M-RH’: The myosin phOSphatase-RhoA interacting protein, or M-RIP was recently
identified through a yeast two-hybrid screen for proteins which interact with myosin
phosphatase and RhoA (Surks et a1. 2003). Our group mapped an EST for M-RH’,
stSG26124, to the very distal end of the 2001 SMS critical interval, near the SMS-REPD
(Figure 2) and demonstrated ubiquitous expression on a northern blot (Table 1). The
human M-RIP is highly homologous to the murine p116RIP3 protein and was localized
to actin myofilaments (Surks, Richards et al. 2003). M-RIP directly binds the myosin-
subunit of myosin phosphatase in smooth muscle. An adjacent domain of M-RIP binds
RhoA, and these proteins may act to regulate myosin phosphatase (Surks, Richards et al.

2003), though currently no role in SMS has been proposed for this gene.

Conclusions

The physical and transcription map of the SMS critical interval and simultaneous
gene identification highlighted many potential SMS candidate genes. Subsequent studies
in our lab will focus on prioritizing the genes which may play a role in the SMS
phenotype as well as in-depth analysis of expression pattern and determination of the in

viva cellular role of each potential candidate gene.

56

Materials and Methods

DNA Preparation:

WW: BAC and PAC clones were obtained from Research
Genetics (now Invitrogen) and BAC/PAC Resources to construct the SMS l7p1 1.2
contig, and DNA was isolated using a modified Qiagen very low copy protocol devised
by our lab. Bacteria containing BACs or PACs were streaked on selective plates. Single
colonies were picked and grown while shaking for 16 to 18 hours at 37°C in 2 ml of
Luria-Bertani Broth (LB) containing selective antibiotic (12.5 ng/ml chlorarnphenicol for
BACs and 100 jig/ml kanamycin for PACs). Starter cultures were then diluted 1/500 in
(250 ml LB containing selective antibiotics and grown for 16 to 18 hours at 37°C while
shaking. Bacteria were pelleted by spinning at 6000 x g for 15 minutes at 4°C. The
supernatant was removed, and the pellet was completely resuspended in 60 m1 of Buffer
P1 (50 mM Tris-HCl, pH 8.0, 10 mM EDTA, pH 8.0, 0.1 mg/ml RNase A). Sixty
milliliters of Buffer P2 (0.2 N NaOH, 1% sodium dodecyl sulfate {SDS}) were then
added to lyse the cells followed by 60 m1 of Buffer P3 (3 M potassium acetate). The tubes
were gently inverted and then placed on ice for 30 minutes to precipitate cellular debris.
Following this incubation, the tubes were centrifuged at 4°C at 2 20,000 x g. Supernatant
containing plasmid DNA was promptly filtered through filter paper (Whatrnan 4).
BAC/PAC DNA was precipitated by adding 0.7 volumes of room temperature
isopropanol and centrifirged at 2 15,000 x g for 30 minutes at 4°C. The pelleted DNA
was allowed to dry slighty (inverted for 5 minutes) and then resuspended in 500 pl sterile
1x TE buffer (10 mM Tris-HCL, pH 8.0, 1 mM EDTA). After the DNA was

resuspended, 5 ml of QBT (0.75 M NaCl and 50 mM MOPS buffer pH adjusted to 7.0,

57

containing 15% isopropanol and 0.15% Triton X-100) was added to the plasmid DNA
and then the DNA was applied to an equilibrated Qiagen Tip 500. The column was
washed 3 times with 10 ml of buffer QC (1 M NaCl, 50 mM MOPS buffer pH 7.0,
containing 15% isopmpanol). The BAC/PAC DNA was eluted from the Qiagen Tip 500
using 15 ml of buffer QF (1.25 M NaCl, 50 mM Tris, pH 8.5, containing 15%
isopropanol) preheated to 65°C. The purified DNA was precipitated by adding 0.7 ml of
room temperature isopropanol and centrifuged at 2 15,000 x g at 4°C for 30 minutes.
The supernatant was decanted and the plasmid pellet was allowed to air dry for 5 to 10
minutes. The pellet was resuspended in 200 nl sterile 1x TE. Purified BAC/PAC DNA
was electrophoresed on a 1% TBE gel, stained with ethidium bromide, visualized on a

transilluminator, and quantified using an Ultrospec 2000 UV spectrophotometer.

W: I.M.A.G.E. EST clones representing markers mapping to

17p] 1.2 were obtained (Research Genetics, Incyte Genomics, or the German Genome
Project) and grown on LB agar plates containing appropriate antibiotic selection.
Individual colonies were then grown in 5 mL LB broth culture containing antibiotic and
DNA was isolated from the broth cultures using the Qiagen miniprep kit according to

manufacturer’s instructions. Briefly, pelleted bacteria] was resuspended in 250 uL of

buffer P1, lysed with 250 uL of buffer P2 and 350 uL of Qiagen buffer N3 (3 M
guanidine-HCl, pH 4.8). Cellular debris was pelleted in a microcentrifilge at maximum
speed and the supernatant was removed to a QIAprep spin column, and centrifuged for
30-60 seconds. The column was subsequently washed with 0.5 mL of buffer PB,

centrifuged for 30-60 seconds, and buffer PE (10 mM Tris-HCl, pH 7.0, 50% ethanol)

58

and centrifuged again for 30-60 seconds. The DNA was eluted with 3a50 pL of distilled
water or 1x TE buffer. Prepared DNA was run on a 1% TBE gel, stained with ethidium
bromide, visualized on a transilluminator, and quantitated using a Ultrospec 2000 UV

spectrophotometer.

Southern analysis: Standard Southern analysis was used to map markers to BACs/PACs
within the contig and to the hybrid mapping panel. BAC or PAC DNA (6 pg) or human
genomic DNA (9 pg) or rodent-human hybrid DNA was digested with 4 U/ug of EcoRI
for ~16-18 hours at 37°C. The human genomic restriction digest also contained 2.5 mM
of spermidine. All digest samples were then electrophoresed in 1% agarose gels in 1x
Tris-acetate—EDTA buffer (TAE; 0.04 M Tris-acetate, 1 mM EDTA) with buffer
recirculation. Gels were depurinated in two gel volumes of 0.25 N HCl and denatured in
two gel volumes of 0.4 N NaOH. DNA was transferred to an Amersham Hybond-N+
nylon membrane in 10x sodium chloride and sodium citrate (SSC; 20x SSC stock
solution is composed of 3 M NaCl, 0.3 M Na—citrate) and UV-crosslinked in a
Stratalinker 1800 to ensure stability. Filters were prehybridized and hybridized in a
solution of 1 M NaCl, 1% SDS, 10% dextran sulfate, and 0.1 mg/mL herring sperm DNA
at 65°C or 1% bovine serum albumin or BSA, 1 mM EDTA, 7% SDS, and 0.5 M
NazHPOa, pH 7.2 at 50°C. All DNA probes (purified PCR products or plasmid inserts)
were labeled with the Amersham Rediprime II kit and unincorporated nucleotides were
removed with a spermine and herring sperm DNA precipitation. During preassociation,
~106 cpm/mL of probe was annealed to 0.25 mg/mL placental DNA to mask repeat

sequences, and then hybridized to the filter for ~16-18 hours. Blots were washed for 20-

59

30 minutes in 0.1x SSC, 0.1% SDS at room temperature, followed by a stringent wash in
preheated 0.1x SSC, 0.1 SDS, at 65°C. The blots were then exposed to X-ray film with

one or two intensifying screens at -80°C for 1-5 days.

Database searches: Extensive information about the genes/ESTs in the region was

obtained through the National Center for Biotechnology Information (N CBI) website

(http://www.ncbi.n1m.nih. gov), including Unigene
(http://www.ncbi.nlm.nih.gov/UniGene), the EST database
(http://www.ncbi.nlm.nih.gov/dbEST) and the STS database

(http://www.ncbi.nlm.nih.gov/dbSTS). Draft assemny of the human genome project data
was found at the human genome browser site (http://genome.ucsc.edu) and through
Project ENSEMBL (http://www.ensembl.org). Supplementary marker information was
derived from the Genome Database (http://www.gdb.org). Sequence alignments and
unfinished high-throughput genome sequence BLAST searches were conducted through
the Baylor Search Launcher or directly through the NCBI MapViewer. Proteomic
analysis was carried out at the Expasy site (http://www.expasy.ch/) and through the
Baylor Search Launcher (http://searchlauncher.bcm.tmc.edu:9331/seq-search/protein-
search.html). Conserved sequence. domains were identified through NCBI

(wwwncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi).
PCR: The polymerase chain reaction (PCR) was also used to map markers to the SMS

critical interval. A standard PCR reaction was performed in 25 pl volumes.

Amplification of ~50-200 ng of template DNA was performed using IU of Taq

60

polymerase in a cocktail consisting of 0.8 nM of each appropriate primer (Table 7), 0.25
mM dNTPs, and 1x PCR bufi‘er (10x buffer contains 100 mM Tris-HCl pH 8.3, 15 mM
MgC12, 500 mM KCl, 0.1% gelatin). Amplification was performed in an MJ Research or

Applied Biosystems (ABI) thermocycler under the following conditions: initial denature
at 94°C for 4 minutes, followed by 30 cycles of 94°C for 1 minute, 55°C for 1 minute,
and 72°C for 3 minutes, and a final extension of 72 °C for 10 minutes. PCR products
were then electrophoresed on a 1% or 2% TBE agarose gels, stained with ethidium

bromide, and visualized on a UV transilluminator.

Northern analysis: The expression pattern of various ESTs was visualized using human
multiple tissue, fetal, and brain H northern blots purchased from Clontech and handled
according to manufacturer's instructions. Briefly, prehybridizations and hybridizations
were performed in 5x SSPE (0.75 M NaCl, 0.05 M NaH2P04, 5 mM EDTA), 10x
Denhardt's solution (0.2% BSA, 0.2% Ficoll, 0.2% Polyvinylpyrrolidone), 100 pg/mL
herring sperm DNA, 2.0% SDS, 50% deionized forrnarnide at 42°C, with ~lo6 cpm/mL
of each probe. All DNA probes (purified PCR products or plasmid inserts) were labeled
with the Amersham Rediprime H kit and unincorporated nucleotides were removed with
a spermine and herring sperm DNA precipitation. Filters were hybridized for ~18-20
hours at 42°C and then washed in 2x SSC, 0.05% SDS for ~30 minutes at room
temperature with 1-2 changes of solution, then transferred to a wash of 0.1x SSC, 0.1%
SDS for ~30-40 rrrinutes at 50°C. The blots were exposed to Amersham Hyperfilm, with

one or two intensifying screens at -80°C for 2-5 days.

61

Chapter III.

RAII as a candidate gene for SMS

The physical and transcription map of the SMS critical interval allowed us to
identify several promising SMS candidate genes. The second major goal of my research
project was to focus on the characterization of one gene, retinoic acid induced 1 or RAII.
Our preliminary RAH expression data indicated that this gene was highly expressed in
brain as well as many other tissues and may play a role in neuronal development. As the
cellular role of this gene is currently unknown, we began a mutation screen of RAII in
several patients with an SMS phenotype but no detectable 17p1 1.2 deletion. We
hypothesized that a mutation in a single gene in these patients may produce an SMS
phenotype. Our studies subsequently revealed dominant, deleterious RAII mutations in
four patients. These findings strongly suggest that haploinsufficiency of the RAM
protein is sufficient to produce the craniofacial and neurobehavioral features seen in SMS

patients.

Rail/RAH cloning and genomic structure

The mouse form of the RaiI gene was originally cloned and named Gt] by a
Japanese group in 1995 (Irnai et a1. 1995) who were primarily interested in identifying
genetic factors affecting neural differentiation. The Gt] transcript was found to be
significantly up-regulated in mouse embryonal carcinoma cells following retinoic acid

(RA) treatment to induce neuronal differentiation (Imai, Suzuki et al. 1995) though it is

62

not known at which biochemical stage th acts within the RA signalling pathway. The
mouse th cDNA cloned in 1995 is 7222 bp long (GenBank NM_009021), encoding a
protein of 1840 amino acids (aa) (Imai, Suzuki et a1. 1995). A discussion of the possible
role. of th/Rail within the RA pathway is provided 'in Chapter V. Preliminary
characterization of Cr], later officially renamed retinoic acid induced 1 or Rail,
demonstrated high expression of this transcript in brain as well as several other adult
mouse tissues measured on a northern blot, including heart, lung, liver, kidney, thymus,
and small intestine. In situ hybridization (ISH) localized the Rail mRNA to several
regions of the adult mouse brain consisting of neuronal populations and
immunohistochemistry (IHC) with an anti-Rail antibody identified the Rail protein
(weakly) in the cytosol and neurites of neurons (Imai, Suzuki et al. 1995). Following this
first report identifying Rail, little more was published on this gene until our group
demonstrated that the human homolog. of Rail (RAII) mapped to PAC RPCI-l pc253P07
within the central region of the SMS critical interval on human chromosome 17 band
p11.2 (Figure 2) (Lucas, Vlangos et al. 2001). As described in Chapter H, we mapped
this gene to the SMS region by hybridization of an RAII clone, EST DKFZp434A139
(GenBank AL133649), to the l7pll.2 somatic cell hybrid panel (Figure 3), then
subsequently to pc253P07 DNA by human genome project sequence analysis and by
Southern hybridization of DKFZp434A139 to EcoRI-digested pc253P07 DNA (Figure 4)
(Lucas, Vlangos et al. 2001). Further information about the genomic structure of RAII
was published in 2001 by Seranski, et al. Using RT-PCR and 5’ RACE fragments, this
group assembled a cDNA of 5667 bp (GenBank M271790) with an 5589 bp open

reading frame encoding 1863 aa. According to their analysis, the S’UTR was not

63

identified and the coding region of RAII consists of 7 exons over ~ 20kb of genomic
sequence (detailed intron-exon sequence was deposited into GenBank under accession
M271791) (Seranski et al. 2001). The Seranski RAIl cDNA sequence and the original
mouse Rail transcript share 76% identity. Small regions of sequence similarity were also
noted between RAII and the transcriptional coactivator SPBP (later renamed TCF20)
.(Rekdal et al. 2000; Seranski, Hofi‘ et al. 2001). At the time of the publication of our
transcription map of the SMS critical interval, we could not discern the cellular role of
the Rail/RAH protein and no other proteins in the public databases showed significant

similarity.

The Seranski group noted that the RA]! protein contains polyglutamine and
polyserine tracts. As polyglutamine tracts (encoded by the CAG codon) have been found
to expand in several neurodegenerative diseases, this group sequenced the CAG repeats
from the non-deleted chromosome from SMS patients. Their results demonstrated that
the repeats ranged from 12-14 and while slightly different sequence variations existed,
the repeats were not found to be expanded or mutated (Seranski, Hoff et al. 2001). The
Rouleau research group at the University of British Columbia was also interested in the
CAG repeats regions within RAII and the possible role of M11 in schizophrenia (Joober
et al. 1999). Joober et al. demonstrated that while no expansions of the CAG repeat
within RAIl (then called hGTl), were identified, the CAG repeat size significantly
differed between groups of schizophrenia patients who responded to conventional
neuroleptic treatment and those who did not. Various statistical analyses demonstrated

that neuroleptic responders had somewhat shorter RAII CAG repeat sizes compared to

controls, possibly indicating that RAII may be involved in the etiology of sOme forms of ‘
schizophrenia (Joober, Benkelfat et al. 1999). Further analysis of the RA]! CAG repeat
was carried out by the same group, who hypothesized that cellular interactions between
the polyglutamine regions of the ataxin-2 and RAIl proteins may modify the age of onset
of spinocerebellar ataxia type 2 (SCA2) (Hayes et al. 2000). Linear regression analysis
demonstrated a possible effect of RAlI on ataxin-2, though no cellular experiments were
done to confirm this hypothesis (Hayes, Turecki et al. 2000). Similar to previous studies,
Hayes and coworkers reported that the number of CAG repeats within RAIl ranged from

10-18 and that the repeats were not expanded (Hayes, Turecki et a1. 2000).

The genomic structure of RAII presented by Seranski and colleagues did not
agree with many of the EST 3 which represent M11 in the NCBI Unigene database
(http://www.ncbi.nlm.nih.gov/UniGene/clust.cgi?ORG=Hs&CID=438904) or those
aligned in the human genome browser (http://genome.ucsc.edu). It is difficult to assess
whether this is due to a tissue-specific alternatively spliced form of M11, as the Seranski
paper does not list the tissue library from which their clones were made. The Rouleau
group subsequently published their own genomic structure of RAII- in 2003 (Toulouse,
Rochefort et al. 2003), which seems to represent a more common form of the gene, as
many ESTs also have the same intron/exon boundaries. Toulouse and colleagues cloned
the RAN cDNA from SKNSH neuroblastoma cells treated with 10'5 M RA and induced
to undergo neuronal differentiation. This group identified 470 bp of 5’ UTR, which was
confirmed by the presence of upstream CpG islands and a putative RA response element

within this regulatory region (Toulouse, Rochefort et al. 2003). They redefined the

65

Figure 7. Human and mouse RAH/Rail genomic structure.

spliced RAIl transcript variants derived from the latest builds of the human and mouse
genome browser (http://genome.ucsc.edu) are shown. The GenBank accession number
and the tissue that each cDNA was isolated from is indicated to the right of each genomic
stfucture, though it should be noted that not all. cDNA transcripts are full-length and may
be based on incomplete EST fi'agments. GenBank accession CL215311 represents
sequence 5' of gene-trapped Rail ES cell line insert from the German Gene Trap

Consortium (httpzllgenetrapgsf. de). CpG islands and a RA—response element were
identified in the RA]! regulatory region by Toulouse et al.

‘Toulouse A et al., Genomics 2003. Aug;82(2): 162-71.
2Seranski P et al., Gene. 2001 May 30;270(1-2):69-76.
3Imai et al., Brain Res Mol Brain Res. 1995 Jul;3 10.2): 1-9.

66

=53»: 3H: mane—H8 Seaman 3.5. ween. 5E3

F3033 _um~.:lu..m. mmmma 03 W.H.._Hw.n.|1.m.n__q. .. Awamzmrc 332mg Nit—Q mmfimmfl
mm: .1 . I " >253»? c3...
NIH HHF _ “ Ill-.1.? ><=§u9 52.3.5. no: .51
I mm: IiiT+ 38.3%. 3::
mad H Trill:- lllllll l_ errata. Eln—5::-
mmH H III >333? ear—53...»
mm H a Taxi. games. as:
mm; H _illTif when—nee. ...—acacEwcmnnoan

 

 

P» «mu—5:8 Own Esau.
0.2.5::

67

32.2.. act «255m 335.com .AOoH. Sea 953
mm. ._ H 1...; .... (.1.. 23.. i s g ......t. t ... a. .1.... f .1 ..T ......m 538:2: I33». 3.9:...
IT. H . H:H...#s+.m. .. ... _ .. . 2233? 5 new. 25er .52.

Em. .. H a... >§ae8.aaaqe=a E.

intr'on/exoll junctions that had been proposed by the Seranski group 35 Wei/as one of the
RAII isoforms that was present in the NCBI database (XM__016259) and predicted 1452
bp of 3’UTR (Toulouse, Rochefort et al. 2003). The recently proposed RAII open
reading frame (GenBank AY172136 and GenProt AAO31738) is 5718 bp and encodes an
1906 a Pmle‘ln (Toulouse, Rochefort et al. 2003). The most common spliced forms of
human and mouse RAH/Rail are shown in Figure 7. Based on our own sequence
analYSiS 35 Well as that of Toulouse and coworkers, the major sequence features present
in the 11931311 RAIl protein are a nuclear localization signal spanning a 1160-1240 and 3
Predicted PHD zinc finger domain from aa 1856-1903, similar to a PHD/LAP domain
identified in TCF20. If the published mouse Rail protein structure (GenProt
NP~033 047) is correct and the entire coding sequence is present in one exon, the latter
motif is unique to the human protein sequence, though there are several mouse ESTs
(Matt-k AK129449, AK013909) that have 3’ intron/exon boundaries which mirror the
human RAH junctions and would contain the PHD domain (Figure .8)‘ An alignment of
the m 1 a sequence published by Rouleau (GenProt AA031738) and the firll-length
predicted mouse Rail protein containing a PHI) domain is shown in Figure 8. In addition

to the peptides used to generate several antibodies for in vitro analysis of RAIl/Rail, this

figure also highlights the putative nuclear localization signal.

EXPl'ession pattern of RAH/Rail
Though there are relatively few publications regarding RAIl, there is detailed
information documenting the expression pattern of this gene. A review of the published

literature as well as our own experiments and information from the public databases

68

Fig”e 3- Mme and human Rani/RAH an in" acid sequence alignment.

(at) The amino acid Sequences of the mouse and buznan Rail/RAH proteins are alignefl
using ClustalW and BOXShade programs. “3 pl'Odlcted unclear localization signal- 18
underlined and the putative PHD zinc finger (Mum! is outlined with a box- Peoude
sequences used to develop antibodies for immUHOhlStochemisu-y and western 311211355 are

shown in bold.

Alignment of the putative PHD domain from RM] With consensus ue 06 derived
goal the NCBI conserved dowsnagztzbasel CDD. KOG09S4. In this digging?“ identical
amino acids are indicated m f ces fmmunr ar m0 aCids are in blue. In addition. two
(native bipartite NLS sequefi1 Rafi/Ran were identified by similarity to the
P as sequence 10-12 '
00956“

69

1 MQSFRE: RCGFHGKQQNYPQTSQETSRLENYRQPGQAGLSCDRQRLLAKDYYS PQPYTGYE "10"”
1 MQSFRE: RCGFHGKQQNYQQTSQETSRLENYRQPSQAGLSCDRQRLLAKDYYN pgpyps yg human
61 GGTGTE- s GTVATAAADKYHRGS ------ KS LQGRPAFP- SYVQDSSPYPGRYSGEEGLQT
61 GGAGT P s GTAAAVAADKYHRGSKALPTQQGLQGRPAFPGYGVQDSS PYPGRYAGEESLQA
114 WGGPQ P PPPQPQPLPGAVS KYEENLMKKTWPPPNRQYPEQGPQLPFRTHSLHVP-PPQP
12 1 WGAPQP PPPQPQPLPAGVAKYDENLMKKTAVPP-SRQYAEQGAQVP FRTHSLHVQQPPPP
17 3 QQPLAYPKLQRQKPQNDLAS PLPE‘PQGSHFPQHSQS FPTSSTYAPTVQGGGQGAHSYKSC
18 o QQPLAYPKLQRQKLQNDIAS PLPFPQG‘I‘HE‘PQHSQSFPTSSTYSSSVQGGGQGAHSYKSC
2 33 TAP SAQPHDRPMSANANLAPGQRVQNLHAYQPGRLGYEQQQQ ---------- ALQGRHHT
2 4 o TAPTAQPHDRPLTAS s sLAPGQRVQNLHAYQSGRLSYDQQQQQQQQQQQQQQALQS RHHA
2 8 3

3 0 “MWQHYGQQGQGYCPPDTAVRTPEQYYQT FS PS SSHS PARSVGRS PSYS ST
0 QETLHYQNIMYQHYGQQGQGYCQPDAAVRT PEQYYQT FS PS S SHS PARSVGRS P SYS ST

34
3 6: PSPLMPNLEINE‘PYSQQPLSTGAF'PTGITDHSHIS'IVIPLLNPSP'I‘DAASSVDPQAGNCKPLQK
P S PIMPNLENFPYSQQPLSTGAFPAGITDHSHE‘MPLLNPS PTDATSSVDTQAGNCKPLQK
4 O3
4 2 0 EKLPDNLLSEVS LQSLTALTSQVENISNTVQQLLLSKATMPQKKGVKNLVS RTPEQHKSQ
KLPENLLSDLSLQSLTALTLQVENISNTVQQLLLSKAAVPQKKGVKNLVS RTPEQHKSQ
4 63
4 8 0 I‘ICSPEGSGYSAIEIPAGTPLSEPPSSTPQSTHAEPQDTDYLSGSIEIDPLBRS FLYCSQARGS P
I‘3CSPEGSGYSAEZPAGTPLSBPPSSTPQSTHAEIPQE‘.ADYLSGSEDPLEIRS FLYCNQARGSP
52 3
54 0 ARVNSNSKAKPESVSTCSVTSPDDMSTKSDDSFQSLHSTLPLDSFSKFVAGERDCPRLLL
ARVNSNSKAKPESVSTCSVTS PDDMSTKSDDSFQSLHGSLPLDSFSKFVAGERDCPRLLL
5 8 3 SALAQEDLASEI LGLQEAIVEKADKAWAEASSLPKDNGKPPFSLENHGACLDTVAKTSWS
6 0 O SALAQEDLASEI LGLQE‘AI GEKADKAWAEAPSLVKDSSKPPFSLENHSACLDSVAKSAWP
6 4 3 QPGEPETLPEPLQLDKGGSTKDFS PGLFEDPSVAFATTDPKKTS S PLSFGTKPLLGTAT P
60 0 RPGEPEALPDSLQLDKGGNAKDFSPGLFEDPSVAFATPDPKKTTGPLSFGTKPTLGVPAP
'7 03 DPTTAAFDCFPDTPTAS SVDGANPFAWPEENLGDACPRWGLHPGELTKGLEQGAKAS DGV
'7 20 DPTTAAFDCFPDTTAAS SADSANPFAWPEENLGDACPRWGLHPGELTKGLEQGGKASDGI
‘1 63 GKADAHEASACMGFQEDHAI GKPAAALSGDFKQQEAEGVKEEVGGLLQCPEVAKADQWLE
'1 8 0 SKGDTHEASACLGFQEEDP PGEKVASLPGDFKQEEVGGVKEEAGGLLQCPEVAKADRWLE
8 23 ESRHCCSSTDFGDLPLLP P PGRKEDLEAEEEYSSLCELLGS PEQRPSLQDPLS PKAPLMC
8 40 DSRHCCSTADFGDLPLLPPTSRKEDLEAEBEYSSLCELLGSPEQRPGMQDPLS PKAPLIC
8 8 3 TKEEAEEALDTKAGWVS PCHLSGEPAVLLGPSVGAQS KVQSWE‘ES S LSHMKPGEEGPEME
9 0 0 TKEEVEEVLDSKAGWGS PCHLSGESVI LLGPTVGTESKVQSWFES SLSHMKPGEEGPDGE
9 4 3 RAPGSSGTSQGSLAPKPNKPAVPEGPIAKKEPVPRGKSLRS RRVHRGLPEAEDS PCRVPA
9 60 RAPGDSTTSDASLAQKPNKPAVPEAPIAKKEPVPRGKSLRS RRVHRGLPEAEDSPCRAPV
1 0 0 3 LPKDLLLPESCTGPPQGQAEGAGAPGRGLS EGLPRMCTRS LTALS EZPQT PGPPGLTTT PT
1020

LPKDLLLPESCTGP PQGQMEGAGAPGRGAS EGLPRMCT RS LTALS EPRT PGP PGLTTT PA

70

Figure 3, con tinned

1063
1080

1123
1140

1183
1200

1236
1260

1294
1320

1349
1380

1408
1439

1468
1499

1528
1559

1588
1619

1647
1679

1707
1739

1767
1785

1827
1845

1887
1905

PPDKLGGKQRAAFKSGKRVGKPS PKAASS PSNPAALPVASDS S PMGSKTKEPDSPSMPGK
PPDKLGGKQRAAFKSGKRVGKPS PKAASS PSNPAALPVASDS S PMGSKTKETDSPSTPGK

DQRSWLRS RTKPQQVFHAKRRRPSES RI PDCRATKKLPANNHLPTAFKVSSGPQKEGRM
DQRSMI LRS RTKTQEI FHSKRRRPSEGRLPNCRATKKLLDNSHLPATFKVSSSPQKEGRV

SQRVK‘sIPKPGTGNKLS DRPLHTLKRKSAFMAPVPAKKRS LI LRSNN ------- GSGGDGR
SQRARVPKPGAGSKLSDRPLHALKRKSAFMAPVPTKKRNLVLRS RSSSSSNAS GNGGDGK

EERRESS PGLLRRMAS PQRARPRGSG- -EPPPPPPLEPPAACMGLSTQSSLPSAVRTKVL
EERP EGS PTLFKRMS S PKKAKPTKGNGEPATKLPPPBTPDACFKLAS RAAFQGAMKTKVL

PPPJKC-‘JRGLKLEAIVQKITSPGLKKLACRVAGAPPG'I‘PRS PALPERR ----- PGGS PAGAE
PPps1§GRGLKLEZAIVQKITS PSLKKFACKAPGAS PGNPLSPSLSDKDRGLKGAGGS PVGVE

EGLGGMGQMLPAAS G-ADPLCRN PAS RS LKGKLLNS KKLS SAADCPKAEAFMS PETLPSL
E:GLVNVG'I‘GQKLPTSGADPLCRNPTNRSLKGKLMNS KKLS - STDCFKTEAFTS PEALQPG

G'TARAPKKRS RKGRTGTLGPSKGPLEKRPCPGQPLLLAPHDRASSTQGGGEDNS SGGGKK
G'I‘ALAPKKRS RKGRAGAHGLSKGPLEKRPYLGPALLLT PRDRASGTQGASEDNSGGGGKK

P KTEELGPASQPPEGRPCQPQTRAQKQPGQASYSSYS KRKRLSRGRGKTAHAS PCKGRA‘I‘
9’ KMEELGLASQP PEARPCQPQTRAQKQPGHTNYSSYS KRKRLTRGRAKNTTS S PCKGRAK

I\RRQQQVLPLDPAEPEI RLKYI SSCKRLRADS RT PAFS P FVRVEKRDAYTTI CTVVNS PG
RRRQQQVLPLDPAEPEI RLKYI S SCKRLRSDSRTPAFS PFVRVEKRDAFTTI CTWNS PG

DEPKPHWKPSSSAASSSTSS-SSLEPAGASLTTFPGGSVLQQRPSLPLSSTMHLGPWSK
DAPKPHRKPSSSASSSSSSSSFSLDAAGASLATLPGGSI LQPRPSLPLSSTMHLGPWSK

ALSTSCLVCCLCQN PAN FKDLGDLCG PYY PEHCLmlmARLEGTLEEAS LPLER
ALSTSCLVCCLCQN PAN FKDLGDLCGPYYPEHCLPKKKPKLKEKVRPEGTCEEAS LPLER

'I' LKGLECSASTTAAAPTTAT ITT PTALGRLSRPDGPADPAKQGPLRTSARGLSRRLQSCY
T LKGPECAAAATAGKP -------------- PRPDG PAD PAKQGPLRTSARGLSRRLQSCY

CCDGQGDGGEEVAQADKS RKHECSKEAPTEPGGDTQEHWVH EACAVWTSGVYLVAGKLFG
CCDGREDGGEEAAPADKGRKHECSKEAPAEPGGEAQEHWVHEACAVWTGGVYLVAGKLFG

 

LQEAMKVAVDM CTSCHEPGATISCSYKGCIHTYHYPCANDTGCTFIEENFTLKCPK KR
LQEAMKVAVD CSSCQEAGATIGCCHKGCLHTYHYPCASDAGCIFIEENFSLKCPK KR

LPL
LP-

71

Figure 8, continued

Alignmetlt 0f putative Rail PHD domain (a 1856-1903) to consensus PHD from
NCBI conserved domain database CDD KOGO9S4:

RAI 1 MC 5 SC «.3 E—asar [Gate :5 KGCLHT z a Yprras my :C I FI ------- 1‘ ”EN rs LKK spa H
PHD \ICNLC EZVKSGACI Qij‘Siefir-iT-TRTAf-‘f IVT<.:ana.;-;LEMKTILKENDsVK: Ksycs an

TWO putative bipartite NLS (aa 1160-1176 and aa 1223-1239) were identified by BLAST
homolog): Searches. The basic or bipartite NLS is composed of two stretches of basic
ammo acids separated by a spacer of 10-12 amino acids (KRXquKRRKY:

£2131 a 1160-1176: RRRPSEGRLPNCRATKK
13a 1223-1239: KRKSAFMAPVPTKKRNL

1
R c
”has J et al., Cell. 1991 Feb 8;64(3):615-23.

72

demonstrate-9 that Rail/RAJ] mRNA is expressed on a northern blot in 3.111703, all tissues

in the adult animal (Lucas, Vlangos et al. 2001; Seranski, Hoff et a1. 2001; Toulouse,
Rochefort Ct 211. 2003). Following the original cloning of mouse RaiI, Imai et al.
demonstrated Rail expression in several adult mouse tissues on a northern blot as well as
preliminary [SH and IHC results which demonstrate localization to regions of the brain.
Our own lab has performed standard northern blots using Clontech adult tissue and fetal
blots and identified ~8.0 kb RAII mRN A expression in all tissues examined (Figure 6)
(Lucas, V1atlgos et al. 2001). Seranski and coworkers also utilized Clontech blots to
examine RAH expression and identified the common ~8.0 kb transcript in all adult and
fetal tissues as well as several specific neuronal tissues (cerebellum, cerebral cortex,
medulla, bone marrow, occipital pole, frontal lobe, temporal lobe, and putamen). Two
other s1‘31ice variants, a 1.5 kb and a 10 kb transcript, were expressed in most tissues at
somewhat lower levels. The Rouleau group performed a similar northern analysis using
the Clontech MTN IV brain blot containing amydala, caudate nucleus, corpus collosum,
hiPPOCEKI'I-rpus, whole brain, substantia nigra, and thalamus and noted that the ~8.0 RAII
transcript was present in all of these regions of the brain at similar levels except corpus

collosmj, where no expression was found (Toulouse, Rochefort et a1. 2003).

An examination of the current NCBI Unigene database entry for human RAII
(115438904) confirms that there are at least 135 ESTs representing this gene, isolated
from a wide variety of adult and fetal tissues libraries. A large number of these ESTs are
from cancer cell lines, suggesting that RAII may be highly expressed in several different

Carcinomas. It is also interesting to find that RAII mRNA comprises 1.97% of an

73

expression library made fi'om adult colon, suggesting the RA]! may have a very
important regulatory role in the digestive tract
{http://wwwncbi.nlm.nih.gov/UniGene/library.cgi?ORG=Hs&LID=3202). It is possible
that RAH expression in the Colon is very important physiologically, as SMS patients
Ofien have gastrointestinal disorders including constipation. The database of Human
Unidentified Gene-Encoded Large Proteins (HUGE) is another valuable source of
information about RAII expression. On the HUGE website, the Japanese Kazusa Human
CDNA ij ect provides information for entry KIAA1820 (GenBank ABOS8723;
http://WWW.kazusa.or.jp/huge/gfpage/KIAA1 820), which represents one transcript variant
Of M1 (Figure 7). The expression pattern of KIAA1820 was determined by a
combinalzion of RT-PCR cDNA amplification and quantitation by ELISA. The HUGE
”315mg revealed at least mid-level RAII expression in all tissues examined, high
exp re ssiQn in various subsections of the brain (including the corpus collosum, which did
not Show expression on a northern blot), the kidney, spleen, heart and spinal cord, and
very high expression in the whole brain and ovaries. At the current time, it is difficult to
deteflnine if RAIl expression in the reproductive system affects SMS patients, as no

patiems have been reported to have children and the fertility of these individuals is

unknown.

While most of these studies examined the RaiI/RAII mRNA levels, protein
expression studies are also underway. Preliminary IHC experiments have been
perfOImed by the Rouleau lab, using antibodies generated against the human and mouse
Rail peptides (Figure 8). Initial studies in P19, Cos-1 and Skbr—3 cells indicate that the

Rail protein is localized mainly in the nucleus (G. Rouleau, personal communication).

74

Further WGStern blot and IHC studies in mouse tissues demonstrate that the Raj] protein
is expressed as a ~l75 kDa protein in all tissues examined, at espeCially high levels in
neuronal tissues (spinal cord and brain; G. Rouleau, personal communication). Valerie
Vinoverski 30111 the Elsea laboratory has also demonstrated positive nuclear localization
using an Ra.11:GFP fusion protein (data not shown); these preliminary data provide
evidence for true nuclear targeting utilizing the nuclear localization signal and possible
eVidenCC for the role of Rail in transcriptional regulation. Further experiments such as
mutation of the putative NLS in the RailzGFP fusion will provide further evidence that

Rail is truly targeted to the nucleus.

The developmental expression pattern of RaiI/RAII is less well-characterized. As

stated ashove, expression on human fetal northern blot (Figure 6) and various ESTs
suggest that RA]! is expressed during embryonic stages, but a full developmental profile
has ”Qt yet been established. The mouse Unigene entry (Mm.275497) lists fewer ESTs
that that) the human entry, many derived from whole brain, though some have been
isolated from mouse embryo (GenBank AK013909, BQ832285, A1893920), indicating
that Rai 1 may be embryonically expressed. We have purchased a mouse Rm] 3’ end
cDNA clone (IMAGE11211624) from heart and used this radiolabeled cDNA to probe 3
mouse full-stage conceptus northern blot (representing embryonic days E4.5-E18.5). As
shown in Figure 9, an ~8.0 kb Rail transcript is evident at all mouse developmental
stages, with highest expression from embryonic days E9.5-E15.5. A more definitive

d6V310pmental profile of the Rail/RAH mRNA and protein expression pattern is

underway using ISH and IHC and will be necessary for determining whether Rail/RAH

75

Figure 9- Mouse Rail embryonic expreSSion pattern.

The .E.ST WAGE; 1211624, repreSenting th? middle and 3 ’ end of mouse RA”, Was
hybridized to a SeeGene full-stage embryo“!c northern blot- An ~8.0 kb transcript is

evident at all time points and is strongest from 510'5‘51 5 - 5' Gan/I is included as a
mRN A loading control.

76

was

al.-analoo-oO'U

‘I’I 0'1

 

 

 

 

 

I .1”?!

 

- qxss~

«at!!! aw: sea test; . -,

 

 

77

E4.5
E5.5
E6.5
E7.5

, E8.5

E9.5

E10.5
El 1.5
E12.5
E135
E145
E15.5
El6.5
El7.5
E18.5

has a global impact on early develoPment and whether haploinsufficicncy of Rail/[(4411
could cause neuronal as well as craniofacial defects. RAII may be especially important
to the development of higher organisms, as putative RAII EST homologs have been
sequenced from rat (GenBank BU758552), Bos taurus (GenBank BE667697), Sus scrofa
(BQS98394), Gallus gallus (BU138127), and Xenopus laevis (GenBank BG813716); to

date, no orthologs have been identified in Drosophila, C. elegans or yeast.

Clinical description of putative SMS patients with no cytogenetic deletion
As the true cellular function of RAII is unknown and not currently measurable,
0111' laboratory devised another method to assess RAII as a potential candidate gene. We
hYPOthesized that patients with features consistent with the SMS phenotype but in whom
no 17p 11.2 deletion could be detected may harbor a mutation in a single candidate gene
which may be primarily responsible for their phenotype. As described in the Chapter I,
there is precedent for this type of analysis in several microdeletion syndromes, such AGS
in Which haploinsufficiency of a single gene (JA GI) due to mutation or deletion has been
found to be primarily responsible for the AGS phenotype. We began our mutation
screening by identifying an initial pool of 12 patients with many features consistent with
SMS but who did not harbor a 17p11.2 deletion detectable by routine cytogenetics or

FISH analysis (Tables 5, 6).

As shown in shown in Tables 5 and 6, these putative SMS patients with no
cytogenetic deletion display many of the typical features of SMS such as craniofacial
abnormalities, flat midface, sleep disturbances, and characteristic behavioral features

such as self-abusive and explosive episodes, as well as hyperactivity. However, these

78

patients do not display many of the more variable characteristics seen in <50% of typical
SMS patients, such as gross organ abnormalities including heart and renal defects.
SMS 1 29, SMSISG, SMSlS9 (Table 5) were part of an initial evaluation of individuals
placed at the Elwyn Institute residential care facility based on a strong clinical suspicion
of SMS. A detailed clinical summary of SMSlZ9, SMSIS6, and SMSlS9 contributed by
Brenda Fincucane, a genetic counselor who evaluated these patients, is described within
Materials and Methods. SMSI88 was referred to Dr. Elsea by a clinical from Belgium
(Koenraad Devriendt), who suspected this patient had SMS because of his explosive and
self-abusive behavior and developmental delay (Table 5). However, some of the patients
in this preliminary pool (Table 6), such as SMSl 17, who has widely spaced fingers and
toes, and SMSl 19, who was exposed to drugs and alcohol in utero, display inconsistent

features and may not truly have SMS.

FISH experiments
All putative SMS nondeletion patients in our mutation screen were evaluated by
Christopher Vlangos in the Elsea lab for sub-microscopic deletions within Up] 1.2 using
an eMtensive series of FISH probes that span the SMS critical region (Figure ll) and data
not shown). In Figure 10 panel (a), an abbreviated contig of the SMS critical interval
demonstrates several of the BACs, PACs, and cosmids used for our FISH experiments.
Representative FISH results using genomic clones from within the SMS critical interval
are shown for SMS 129 in Figure 10 using BAC probe CITC-40123 (representing the
P EMT 2 gene) in panel (b) and cosmid probe c62F2, which contains the FLII gene, in

pane] (c). While these FISH analyses do not completely eliminate the possibility of a

79

Figure 10. FISH analy sis on putative SMS patients with no detectable deletion.

(a) An abbreviated contig of the SMS critical interval (as of 2003) is shown, highlighting
RAII in blue. Several BACS used for FISH to evaluate SMS ddefions are shown

the contig, as well as various overlapping cosmids (dashed lineS) also used for FISH
probes. BAC arc-40123 (containing PE MTZ) and cosmid 62F2 (containing the FL 1,
gem? are shown in red and were used as probes in the FISH experiments shown in Panels
b an c.

(b). FISH analysis of SMSIZ9 Sl'IOWing no deletion of the green test probe (bc40123,
wdlcated by White arrows) perfortfled by Christopher Vlangos of the Elsea laboratory,
1116 red control probe (RPCIll-314M5) was ”3"" ‘0 'dcm'fy "‘6 9 arm of chromosome

17'

. ' deletion of the green test be '
FISH anal srs of SMSIZ9 showmg {10 . 1.1m cosmid (:6ze
(c) m the chroyr’nosome 17 specific cosmid 11131317) performed by ChflStOPbcr Vlangos of
(fioElsea laboratory. The red comml Probe (RPC111'314M5) was used to identify the q
17. Similar experiments were perfortned on all of the nondeletion

the l
mee . 0 o .
iatrtiscdesgii in Tables 5 and 6 to determine that these mdrvrduals did not harbor

17p1 1 .2 dClctionS-

80

'\'-

‘4' -.

4“ .lm rod [-1343

 

 

8|

.ouoa uni I I'IDJH

 

011N789“! II'IDJX

Loarszod I-Dau

610916“! ”‘13:!!!

WW“ 3.1!.)

e

—f

”3

v!

C.

   
 

L

SMCR8 —v
SHMTI J

O
(D
5

Table 5. Phenotypic channel-mien of SMS patient- without a cytosenefic deletion
and putative SMS patients with deleterious RAH mutations.

Features of an SMS patient with a typical 17p11.2 deletion (SMSI26) and a gm,"
1 71311-2 deletion (HOU142-540) are 0011“th to putative SMS patients With no
cytogenetic deletion who barber deleterious RAII mutations (SMSIZ9, SMSI 56
5M5159fimd 3M8188). Phenotypic information was derived fiom www.genetests.0rg;

+ 7" posmve, - = negative, N= unknown or not evaluated. 111 the interest of spaCe, the
efider and 336(8) 0f evaluation Of each patient are "Sth below:
6111316, age of evaluation 41 years
ale, ages of evaluation 10, 14, 15 years

ale, ages of evaluation 14, 30 years
ofefllale, ages of evaluation 17, 21, 31 years
6, ages of evaluation 13, 19 years

(Male, age of evaluation 12 years

82

_

 

mam—wo-

mosses,

mam—N0.

259%..

2(5— me.

9:22..

 

go»
3.6... 35an F v.33 a. mam wean—.3

 

nausea-ensign.

 

 

gamma.“ 5 ..o . Fe»

 

v3“ .355 8596 8 who

 

Havana c. .2. :.

+++

 

was? 953 mean

+++

 

wan: 323v.

 

mac: mango

 

Cacao 02 58303

 

$8.89 an. . <28

++++++++

+++

+Z+++++++

 

2235;353:331

 

Zgg 883305

+

 

m .803 more

+

 

323 nor.

 

53580 e .583»

 

m5, . Ema—ego."

 

Mafiurcu “Eu $59 sauna"

 

On «HAVE—082:.»

 

war 08620558me

++++

 

mama 3855138 a». .5"

+++++++

 

Ewan can u

+

 

>333: mung"

+++++++++++

+++++++

+++++++

+

+

+++++++ZZ++

 

mom—«Eda 85:5: 3 we I 3.x. 3. mam ”22.8

 

 

magnum 3%

+

 

09:8 men—2.5338

 

 

mNmomme

+++

+++

 

moons;

+++

z++z

 

\‘Illl .
woes—.8 85:5: ... A 8.x. 3. mam 32.8

 

 

 

&o<mmo=_w~ gages

 

‘\O“a\l‘\.||||!
wane goes—z...

 

 

mesa
So» a \ £20

 

 

 

+++

 

 

 

 

 

 

83

Table 6 Phenotypi . .
, ° c character-macs of putative S . .
deletion who harbor no obvious RAII mutation. MS patients wnth no ogenetic
Phenotypic info . .
rmatzon was dCflVed from w -
fl unkn _ .genetestsg ’ + = ' - = .

29mm tamer-mt evaluated In the “new“ 0f Spacefgtiae mail-moi)! phgegan‘ie’

10W: HSlstcnt w‘th SMS: and the age 0f evaluation of each pan, 'ent is Pm
a emale wi . .

F th memally placed fingers and Widely spaced toes age of evajuafio
, n 16

emale With drug and alcohoI ex ‘ ’ e ‘

posure In u m, age OfeValuat
cfefilale, age of evaluation 17 years 1°" 6-5Y6ars
ofemale, age of evaluation 17 years
cfemale, age of evaluation not known

(Female, age of eva\u.ation 21 ycars
gMa‘C, age of cvaluauon 6 years
age of cvahxation 10 yeaxs

hFema‘e,

84'

 

ME :4. mama—m... mam Sun mam—mug mamas. mam—ea. mam—can, mgmsey

 

a:
0 Eng 3552. :— v.33 en mam 552.8

 

wuaonnnmEm—SSE

 

was: a E:

+

 

+

45.58 c Fa»

 

To EB «3358» o

 

+++
++++
I

Hanna : on m

 

was? m ammo mono

 

I
++++++
++++++
++++++

manor awn _

 

mg: “.883

 

++

35%." 08> ofl 8.85:2

+

 

+2+++
++++

+-
zz+z++z++
+2:-

+2-
++++++

moan no <98

 

22:6... man—Eggnog—

 

Zouc: REQmmou

 

m can: an?

 

399. nor.

 

++++

Ewan—o r 203»

 

+++++

m5 minim—boo

85

 

++2+++
++++++

+

mfifirc E 8555 E ,
0: 3035825

 

+++++++

 

++++++++

+
I
l
I

we oaaaowozoafiu

 

  

Egg SE B \ moon «5

SEE Exam
>338: mooEUm

mam—8.6m 8:58: 3 u: I amok. on mam nan—.8

 

 

+++++++ZZ++
+++++++zz++

 

244444
+1
+++
+2
+++
+++

 

 

maam _Omw

 

+2

Doc—ma acuogmaom

mmom m

 

 

 

++++
-+
-+
.+
.

Z++Z

z+++
.

moons;

 

\c"
Mama—.8 8:58: 5 A max. on mam wan—.8

 

Ounaoémnafi macawcam + - - - -

 

22

\X‘ I I I l l
wax; anagram

$35” . + + .
OK»: \ Ewen . - .

 

 

224
I

 

 

 

 

 

 

 

 

 

 

 

 

small, cryptic deletion within 17p11.2, they demonstrate that the individuals Screened for

single gene mutations did not harbor detectable deletions within the SMS critical region.

Sequencing ofRAI] in putative SMS patients with no deletion
Having identified 12 putative SMS patients without detectable deletions (Tables
5-6), we undertook systematic PCR and sequencing of genes within the SMS critical
region using DNA from our non-deletion patient pool, beginning with three genes first
localized to the SMS critical interval in 1997, DRG2, RASDI, and RAII. RAII primer
sequences spanning the entire genes are shown in Table 7. While our initial sequencing
identified Several known SNPs, no deleterious mutations were found in either RASDI or
DRG2 in any of our patients. However, a 29 bp deletion within exon 3 of RAII was
identified on one allele from SMSlZ9 (Figure 11). This deletion was clearly evident on a
2% agarose gel following PCR amplification with exon 3 primers, as two bands
representing the deleted and non-deleted alleles were resolved on the gel (Figure 11,
panel b). We isolated and purified each band from the gel and sequenced them
separately. As shown in the RA]! mRNA coding sequence in Figure 11, panel c., the 29
bP deletion produces a frameshifi that introduces 8 incorrect amino acids, followed by a
Premature stop codon, truncating the protein. This mutation also abolishes a PspOMI
restriction site (Figure 11, panel b). We predict that this deleted allele encodes a
truncated and either abnonnally-functioning or non-functional RAll protein, likely
resulting in haploinsufficiency for RAIL We sequenced the exon 3 amplimer from the
Parents of SMSIZ9, SMslz7 and SMSlZ9 as well as one sibling, SMSI64 and subjected

this amplimer to restriction digest with PspOMI. As shown in Figure 11, panel b-a the 7-9

86

 

 

  

 

 

 

 

  

 

 

 

 

 

  

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

~83»...— 3152. 853 ”9658 e152. Gnu; guns—Em

85 .

onioannQnooanogoo o>nooQon>QoH>oHQoQo 3
oooa>aammao>om>o>ooo nno>oaooa>oon>ao>>o>do AL
no>ao>n>omnn§o>gnfi o> macaronjoannjqnwgo S
ndn>nn>n>nnn>ndn>anbn G>an>nano§ _ Eb? _ _ er 3
0>§OQHQ>QHQO>>OQE one >>GQOO>OQQ>>CCE _ C 0 mo
onooo>nanaoo>oneoo>n anabao>on§oono> (e _ to? a
8>nfinn>>fin>8>8>fififiw0 O>O>Q>OGOOHOGQ>QOH8fiQ as.
l O>O>HO>>OO>OOHO>>O>R CHOO>QOO>OOOHHQQOHO>Q 3
OOEQHQHO>OOQOOOHO>QHQ ono>oammoojaognfiaon am
[ OnQuanEnngoanonQu nn>n>di>nn>¢¢ee _ s _ _en E
noolfl.ono>o p > >n>1ooaoo aa.—danger? _ E _ r c _ c at.
HOG >OHOHOO>>>QOOOOOOH >OQOOOO>>Q HQO>_ (C _ CC oo
GOHGOnQonOHOOGHGQwGQ 0.22115 infindnndnndonswn i
Hoamn>mgonoono>oa >r _ c cn>o>joanono>o> 3
oo>n>o>on>oo>>nnoan>na >>§hoao>aoo>aoano a.
§O>00§Q8oano>>n n>>>an00>>nOHOO>>H>OO i—
QQ>OHQHQ>>QO>QQHQOQ>OQ QO>QHQO>QHQO>OHOHQO>QQ A o
0>OQQHOOHQHQOH>QEAQOO ojo>o>0>0000>>nn>HOHon i
0905838033393 oan>>§>>o>a>n>>no>io A u.
O>QOHOO>H>O>O>O>>HQAO OOOEQHQO>OO>OO>OOO>OO AL
QQHOOQHOOHQQHQO>O>>OQQ oaQQO>O>OOOOEQHOOHOO an
on>8>0>>>oOOQHQOOO>O Q>OHOHQ>>9.00Q>QOHOQHO 3
m>oaooaoéomo>noo>>oo>o ooooooagooocyoontynago S
neon>moooooo>oano jonionoooaoooo>o>o S
noo>a>no>o>o>>>oo> OOQOHQJOQQOSQ mo
o>>>aoao>onoanmooann Q>ooaoobo>ooagoa>o S
Hona>§8aoom>>>oa nooo>anfi5>>>oaoo>>>oo 8

 

 

 

 

Haw—n u. it @152.“ =an :5 "Eu 3:3,. we: 35on £03 8858 $03 :5 morn—EH Eu B 29.
manager on enemas; 3. :5 Emma :8. .553 enaaam oo<a~ 335 "883va 8 So: a venues“ mac—woo
£533. .25 magnum 883388 m9. :8 3:5; a Banana ea game: 0 85:0: 2 H3583 wow
33502 28 3:35? name; 8 =5 2% a? 8 2:35 ages: aaEmam. >5“...an S can 536

.3558 55:33.

87

Figure 11. SMSIZ9 KAII mutation analysis.

(a) SMSIZ9 pictured at age 30.

(b) Pedigree analysis of the SMSIZ9, his parents (Sh/[8127 and SMSl28) and one
sibling (SMSl64). The RA13/ 14 PCR product is present in lanes 1, 3, S, and 7‘, “0‘3
doublet in affected individual SMSIZ9. The PspOMI digestion of the RA13/ l4 PCR
product is shown in lanes 2, 4, 6, and 8; note undigested mutant allele in afiec
individual $431» which is not present in the parents or sibling. smsns was not noted
to carry the 29 bp deletion by PCR, restriction digest, or direct sequencing, though fur‘her
DIIPLC analysis showed ~20°/o mosaicism in this parental sample (see Figure 1 6), The
29 bp mutation was not detected in 102 Caucasian control Saulples which were analyzed

by PCR amplification and ngOMl restriction digestion.

(a) The sequence traCing represents the the 29 bp deletion detected following direct

uencing of the RAB/l4 ampllmer of RAII from SMSlZ9, as well as the “a“ type

uence detected on the other allele; this deletion eliminates a PSpOMI restriction site,

- sincorporates 8 amim acids (4 of which are shown in the diagram) and produces a

“51 stop codon. Other mutation screening results show that SMSIZ9 is
(‘1‘: terozygous (cm for known SNP in exon 4 (dbSNPzrs381 871 7).

88

m. mam—Ne U.

mam—Ma mam—nu

[me

mam—ac

mam—N0

s8 E.

 

me. E.

awe we

 

.3 E.
Q m m

_\ ~ 5 s n _e w _s o
m.->nnnn>oa£ . .

u. m or >t ”23.36;
. TO>OHG>>> rm.

 

Q m w a V k .. 3:55
ur>onon>oa£no>nan>>>oo .u.

. a?

 

mm m
normBm WNW omww
nsnnm Mm Wnnnn

89

bp deletion was not evident in these individuals, though we could not rule Out 10% Ieve 1
mosaicism by PCR and sequencing methods. We did pursue further analysis oftbe exon
3 amplimer from SMSlZ8 by Transgenomic DHPLC mutation screening, Which revealed
that the 29 bp mutation was present at ~20% level within the genome of this
phenotypically normal parent (Figure 16). We also screened >200 Caucasian control

chromosomes by PCR and PspOMI restriction digest and did not detect this particular

mutation (data not shown).

We then examined three additional patients, SMSlS6, SMSlS9, and SMSISS for
mutations in RAII. In SMSIS6, we identified a deletion of a single cytosine on one allele
within exon 3 (Figure 12, panel c), which occurred in a run of 6 CS ending at nucleotide
position 5265 of the RAII mRNA. This 5265delC on the coding strand produces a
subsequent frameshifi, introducing 74 incorrect amino acids, abolishing a BglI

recognition site, and truncating the protein (Figure 12, panel b). As in the case of the
3SdelG mutation in connexin 26 which is implicated in recessive hearing loss (Zelante et
al. 1997), we were unable to determine absolutely which C is deleted within this run of 6
CS in RAII. As shown in Figure 12, panel b., the parents of SMSIS6 (SMSIS4 and
SMS 155) as well as >200 Caucasian control chromosomes screened by PCR and BgII
resu‘iction digest (data not shown) did not carry the 5265delC mutation. Again, we
Predict that this frameshifi mutation produces an abnormal or truncated RAIl protein,

which ultimately creates haploinsufficiency of this crucial developmental molecule.

90

Figure 12. SMSIS6 RAH mutation analysis.

(a) SMSlS6 pictured at age 31.

(b) The sequence tracing from the RA25/26 RAII amplimer reveals the 5265 (MC in
exon 3 on one allele; this mutation eliminates a Bgll restriction site, misincorpm'ates 74
amino acids, eliminates a BgII restriction site, and produces a downstream stop wdm“
Other sequencing results for SMSIS6 reveal that this individual is heterozygous (Cm {0‘
known SNP in exon 4 OT (dbSNP:rs38187l7) and homozygous (G to C) for known
within a non-coding region (dbSNPII'82297508).

(c) Pedigree analysis 0fthe SMSIS6, her parents (SMSlS4 and SMSISS). The RA25/26
PCR product is present in la-fleS 19 3, and 5 and the Bgll digestion of the RA25/26 PCR
l,foduct is shown in lanes 2, 4, 6, and 8 (note undigested mutant band present in sMSI56
whim is not resent m parental samPles). The 5265delC mutation was not detected in
1 O Caucasim control DNA samples WhiCh were analyzed by PCR mplificafion and

13 £11 restriction digestion-

9l

 

m. meamHme .U.

mam—us mama“

r

mam—ma

.. , .. _,.. ..., _ a u a u a
....M a: E. Iv
. as E.
3. E.
emu n:

N. a 5 D

 

Q 5 k b E \— R Q or .. 235::
C. 7. 5 O>ODOQHD>ODQOOO>OOHO>OOOODOO>>DO>OOOO nu.
m.-000>>nnnnnn

>nnnnao>nonnnn>oan>nnnnonn>>nn>onoon-u.
c e w a

E we; a s sh s , LS? em,“ a a s ...:52:
g: a: a; s a )

Sage—n

.mM ..M

m WWW wmmn
n awn amnmw
5. aim m .m w
um. n.) .Wmllm
n mw memm

92

 

Figure 13. SMSIS9 RAH mutation analysis.
(a) SMSIS9 pictured at ages 11 and 19.

(b) The 1449de1C mutation in R3411 exon 3 on one allele is shown, which misincorporfites \
34 (12 of which are Showm 1n the sequence tracing) aInino acids and prodthes a

premature stop codon. As no known restriction site was altered by the 1 449delC

mutation, amplified and sequencéd the amplimer containing the 1449deIC mutation fi‘Om

7100 Caucasian samples (including the parents of SMSISQ), and this mutation was not

detected in this ventilation-

93

 

magma

 
   

ewe 3

U . m. m Q M Q 1 u. x. m E \— Q .. 722:1
M Q NN q M. OOECCC>DODOOH>OHO>COOO>OOCOOO>OOOruc

uronn>en>nqon>nnnn
o>>omo>ononnq>nan>onno>onnnon>non> -u.

Q m M me Q ..s. ...:EE:

E??? f ; ,,._
\

2.3.50

94

Similarly, a deletion of a single cytosine within a run of 4 C’s beginning at nt
position 1 449 of the RAH mRNA sequence (Figure 13) was found on one allele in patient
SMSlS9 and a deletion of a single cytosine within a run of 5 C’s beginning at nt position
3801 was detected in one allele of SMSlSS (Figure 14). These deletions also produce

fiameshifls within the RAII coding sequence, and premature stop codons which can

truncate the protein and cause cellular haploinsufficiency. As no restriction sites were

altered by these deletions, we directly sequenced 100-200 PCR amplimers from
Caucasian control chromosomes as well as parental samples from the family of SMSlS9,
and the mother and two sisters of SMSISS. We did not detect the l449delC or the
3801delC mutation Within this population (data not shown). The deleterious mutations in
RAII we have identified affect all known putative transcripts from this gene with the
exception of KlAAl 820 (GenBank AB058723), as the protein is truncated in each
mutation prior to the regions of alternative splicing (Figures 7 and 15, panel b). The
location of the 2.9 hp deletion, l449delC, 3801de1C, and 5265delC mutations, as well as
several RAII single nucleotide polymorphisms (SNPs) mapped in reference to the RAII
sequence published by Toulouse et al. (GenBank AY172136) is shown in Figure 15. A
schematic of the predicted RAIl protein truncations which may arise from the deleterious

RAIl mutations is also shown in Figure 15, panel b.

Other RAH sequence features

While no other obviously deleterious RAII mutations were identified in our
sequencing p0pulation, several presumably benign RAII single nucleotide SNPs were

identified (Table 8 and Figure 15), which occur in putative SMS patients as well as

95

Figure 14. SMSI83 RAH mutation analysis.

The 3801delC mutation '9 RAII exon 3 on one allele is shown, which misincorporates 46
( 1 0 Of which are shown In the sequence tracing) amino acids and produces a premature
stop 90don. As no known restri ction site was altered by the 380 l delC mutation, we PCR-
ampllfied and sequenced the amplimer containing the 3801delC mutation from >75
CaucaSian samples (“Ending the mother and two sisters of SMSI88), and this mutation
was n0t detected in this Population-

h m, 3 91.2032“:
O>OOOHOHHO>>Q>QQ>HQHOHHOHOOD?» -u.

mRQ>QQQHHOOOO
>OOOHIOHHO>>Q>QQ>HQMOHHOHOOO>>Q -m.
h

a... ... E as? as.

mac—no.0

“359:.

9'7

normal control chromosomes. As shown in Table 8, known SNPs (those found in the
NCBI SNP database: http:l/www.ncbi.nlm.nih.gov/entrez/queryfcgi?db=snp) include a
synonymous change in exon 4 (dbSNPzr538187 17) present in several individuals and a
SNP within a non-coding region of the RAII 3’ UTR (dbSNPzr82297508) (Figure 15).
Our sequencing results have also revealed two unclassified SNPs, including a
synonymous change in exon 3 identified in DNA from patient SMSl96 and a
nonsynonymous change in exon 3, which produces an R to G amino acid change (Table
8). The latter sequence change was isolated by Transgenomic, Inc. denaturing high-
performance liq Hid chromatography (DHPLC) analysis of control and. parental
chromosomes (Figure 1 6). As this normal individual is heterozygous for the SNP, this
amino acid change is presumably benign, though the true cellular effect of this sequence
change (or any missense mutation) on the RAN protein is difficult to assess. We also
analyzed the structure and number of RAII CAG repeats within our patient and control
population. Our studies demonstrated that the number of CAG repeats ranged from 10-
15, which correlates with other published RAII information (Joober, Benkelfat et al.
1999; Hayes, Turecki et al. 2000; Seranski, Hoff et al. 2001). We commonly found

different numbers of CAG repeats on each allele from a single individual (data not

shown), though we did not identify expanded repeats.

One Other interesting sequencing feature involving a possibly polymorphic amino
acid repeat was discovered in a control sample. Following the detection of the 3801delC
mutation from an RAJ 1‘ exon 3 PCR product amplified from SMSISS DNA, we

sequenced the same atnplimer in >100 control chromosomes. Although none of these

98

 

Hue—n a. has «SEES gunman @3889 £an came:
2:. uowfia 8.56.“.

was onuiucxd mam 32.2.8 SE. .5 ofiomgnan

99

 

 

,(a) The most recently {lescribed RAH genomic structure is shown (Toulouse et al., 2003)
”1 this schematic d?nved from the genome browser (http://genome.ucsc.edu) and
munitions identified In 0111' patient population are numbered in agreement with the
nucleotide positions Of this new genomic structure, beginning with the translation start
codon of GenBank accession AY172136 as at 1. Single nucleotide polymorphisms are

also indicated and described in the table below the genomic structure.

0» A schematic is shown of the normal human RAIl protein (N terminal at the far tea
and C tflminal at the far right) and predicted truncated RAIl protein structures which
rmay arise fi-om the dominant, frameshifi RAII mutations identified in SMSIZ9, sMSIS9,
M3156 d SMSl88. Colored bars indicate motifs within the RA“ protem: blue
5 , an . . .
resents the polyglutamine tract, green represents the nuclear localization Signal, f
allow represent the polyserine tract, purple represents the PHI) domfnn, and. red
Y presents incorrect a which are incorporated downstream of the fiameshifi mutations.
A“ mutations are predicted to elimi nate the PHD domain.

100

m2.“v >

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m. A
no an no. aa.—no.0
mag—qU O
n
ma E ...m ms mm mo
T.“.M.~HH. «111 4 in r it 4 .l t l r ll.__...-.n:._..its._.. .Tr
— _ m2."v O
stcncfi $83.0
me > can—fimaoa m2? aosmvfiosvcaocm 255mm m: 88: u 80353 3
m 838— :agag— A05“ W 8 O me 035on
Illllmmwcuwl ~50—swung m2? $503555 Graeme m: 98: w A19 mrv m
mzw n acmenawf 33.“ $33308 035mm 5 $8: a A59
m2—c U acmemewodow coach... 3 u. soséamsm monconoa AQQ
13.5-0 zrm wearm
3. _ ' 22.5»— .EBe: F»: 138:.
_ r mam—mo
mam—me
mam—mm
mam—ma

 

 

,.
!

iii
if
he
ml
m5
sis

101

Figure 16. DHPLC RAH mutation screening results from parental and control
samples.

AnalySiS by Transgenomic DHPLC of parental and randomly selected Caucasian control
samples detected ~200/0 mosaicism for the 29 bp deletion within the SMSIZ8 parental
sample (mother of SMSIZ9) in the exon 3 amplimer, as well as an unclassified SNP in a
normal Caucasian control (G/C heterozygote which results in a nonsynonymous R to G
33 Change) in the same amplimer.

102

v/N'” _

‘\

 

Fluorescence (mV)
- a gas s, s a s

L A

l

 

 

 

 

 

 

 

 

 

I Hue: u 2593 82.2: 3.
mam—n3 .3an 5933 3.. no
3. no.

I Hue: u 8.56. .8an

ll.- 58: u 02:3. 3.5. E41 382:

 

 

 

 

103

controls contained the 1449delC deletion, one sample did contain a unique
polymorphism: one allele coded for two adjacent glutamic acid moieties at a position

1260 and 1261 and one allele coded for only one glutamic acid (data not shown). The
- effect of this change on the cellular role of the RAM protein, as well as any of the
identified SNPs, is unknown at this time, and awaits the result of future fimctional

studies.

Towards high-throughput RAII mutation screening

In order to facilitate more efficient, high-throughput RAII mutation screening, we
performed preliminary DI-IPLC and temperature gradient capillary electrophoresis
(TGCE) experiments to determine whether these techniques could identify known RAII
mutations alongside known normal controls. Both of these technologies analyze the
retention and elution of crude PCR amplimer from a matrix (Kuklin et al. 1997; Kuklin et
al. 1999; Li et al. 2002). Those amplimers with no sequence changes (SNPs,
insertions/deletions) will theoretically elute as a single homoduplex peak and those with
anomalous sequence features will elute as heteroduplex peaks of differing sizes and
retention times. In future, we intend to tier our mutation screening by first performing
high-throughput DHPLC or TGCE analysis of all PCR products, then sequencing only
those products with unusual results. All PCR amplification for DHPLC experiments
were performed by Trangenomic, Inc., as our lab had difficulty producing enough crude
PCR amplimer for this analysis. As shown in Figures 16 and 17, three known RAII

mutations: the 29 bp deletion, 5265delC, and 1449delC were effectively detected by this

104

Figure 17. DHPLC screening results of known RAII mutations.

Utilizing template DNA and primers from our lab, Transgenomic amplified and analyzed
PCR products containing three known mutations in RA]! :

Panel (a) shows the 29 bp deletion within exon 3, which can be detected at a 1:200
dilution. In panel (b) the 5265delC mutation heteroduplex is compared to two normal
control samples and the 1449delC mutation heteroduplex is compared to two normal
samples in panel (c).

Samples were compared to parental and normal Caucasian control samples using DHPLC
analysis. The results of this analysis demonstrate that known mutations can be
distinguished from control samples by DHPLC, though the amount of PCR product
required for this analysis does not make this method cost effective for our current
research purposes.

 

Absorbance (mV)

.3
G

 

 

C< 538:2.

flog—69.33 365 N93. an. IV

Ill.)

 

 

.BNU-BGONOO

 

 

lII>l

 

 

.lll mam—Ne No 3. go—
| :3 0:58: 2. mam—No

..I _u no: 2.5.2. 3. mZMEc

 

 

 

 

 

~ u ..

....Bo A3559»;

alyzed

as

moral

norm

1

MPH

:an be

3(0335

cuntfi’.

106

39:6 5. 83.2.2.

 

o. .

“3

A.

 

Absorbance (mV)

A
l
M

 

 

322.35.... .89..

 

3 it t.

 

 

 

02:3. meat—a .

Ill. mam—um
30833.“.5... a2. a S. no.

 

02:3. 2.5—.5 N

 

 

 

 

1 . i C 4 1 2 ‘ 1 i ‘ d I ‘ 1 d‘ 4 l i q 1 J 444444

...—so .3553.

107

Enid 3. 83.5.2.

0.

Absorbance (mV)

 

30.22.:on Rex

/

 

 

 

 

 

 

l 02:3. 2:: En .

 

02:3. 2.56.0 N

mam—me
...onanenwmecm ..2. . E. an.

 

 

 

 

.....Bo .3553.

108

Figure 18. TGCE screening results of known RAII mutations.

In collaboration with Dr. Pragna Patel the Baylor College of Medicine, we analyzed
crude PCR products containing three known mutations in RAII using SpectruMedix
TGCE analysis: in panel (a) the 29 bp del is compared to a normal control, in (b)
5265delC is compared to a normal control, and in (c) 1449delC is compared to a normal
control. As shown in TGCE results, known mutations produce a distinctive heteroduplex
peak which can be distinguished from controls.

109

 

m.

«we:

 

3.2.2

. 92:.

3.2:.

3.2...

 

.98:

02:3. 3223

 

mam—Ne" no 3. 2:23:

110

U. 3.8: _ Ems—6 5. 8.55.2—
Page
6.8: w

98..

N.§{£I;L 1 ’b? r .l»

 

 

 

02:3. meat—a

11]

who:

NLS:

mam—m9 gamma—O

 

 

n.

whee

”.25

N38

9.2x.

r3:

9!!

aka:

 

mam—=6 um. 3:35:2—

 

_ ¥

 

02:3. man—Ea

 

$8

mam—me“ Enema—O

112

method, as well as ~20% mosaicism of the 29 bp deletion in the SMSlZQ parental
sample, and a previously undetected SNP from a control sample. While we found
DHPLC to be a powerful method for mutation screening, the amount of PCR product
required was too great to make this a cost-effective RAH screening method by the Elsea

lab.

In contrast, TGCE required less crude PCR product, which Ellen Wilch fi'om the
Elsea lab subsequently diluted 1:5 with distilled water and dried overnight before sending
these PCR products to collaborators at the Baylor College of Medicine for SpectruMedix
TGCE analysis. As shown in Figure 18, our initial results showed that TGCE analysis of
the PCR products was able to distinguish known RAII mutations from normal controls,
though subsequent analysis with unknown samples did not produce usable results.
Possible reasons for this failure are: PCR amplification did not yield enough crude
product for analysis, the dilution factor was too high, or the PCR products were too old
(optimal analysis is performed with amplimers that are less than a week old, though our
initial TGCE samples were at least one month old) (Li, Liu et al. 2002). We h0pe to
optimize our PCR amplification in the future for productive TGCE analysis, though at

this time we continue to directly sequence all PCR products.

Conclusions
We believed that RAII was a promising candidate gene for SMS based on
preliminary evidence that suggested that it is highly expressed in the brain and may have

some functional relationship to a transcriptional coactivator. Though still very little is

113

known about this large, novel protein. Our approach to evaluating RAII as a candidate
gene for SMS was to take advantage of the valuable data supplied by SMS patients with
no detectable l7pl 1.2 deletions. While several clinicians did not truly believe that these
individuals had SMS, we began a sequencing effort to identify single-gene mutations in
these patients and identified four nondeletion patients who harbored deleterious,
frameshifl RAII mutations. We now believe that RAIl haploinsufliciency is responsible
for the majority of the SMS phenotype, including the craniofacial, neurological, and
ultimately, the behavioral abnormalities, though perhaps not the gross organ defects and
cleft palate. Several other SNPs in RAII were found through our mutation screen and at

this time, it is difficult to assess the cellular importance of these sequence changes.
Materials and Methods

Patient ascertainment (referred by Brenda Finucane of the Elwyn Institute, Elwyn,
PA): This study was approved by the Michigan State University Committee on Research
Involving Human Subjects. The phenotypic characteristics discussed here are also
summarized in Table 5. SMSlZ9 is a 30 year old man who was admitted to residential
placement at age 14 because of aggressive and disruptive behaviors which could no
longer be managed at home. He was the product of an uncomplicated firll term
pregnancy, weighing 7 lbs., 1 oz. Motor milestones were normal but speech development
was significantly delayed. From an early age, he exhibited aggressive and self-injurious
behaviors, as well as sleep disturbance and frequent “self-hugging”. He is obese. His

behavior is currently stable, although he continues to require residential placement and

114

psychotropic medications because of aggression. Results of fragile X and chromosome 1
analyses at age 14 were normal. At age 21, he was reevaluated in genetics clinic and felt
to have many behavioral and physical characteristics of SMS. A repeat cytogenetic study
(650 band resolution) was normal, as were results of chromosome analysis on skin
fibroblasts. SMSlS6 is a 32 year old obese, white female. Pregnancy and delivery were
normal; birth weight was 8 pounds. Motor milestones were achieved on time and there
was no history of hypotonia. Speech development was mildly delayed compared to that
of her siblings. The patient has a history of mild mental retardation and emotional
disturbance, including aggressive and defiant behaviors which prompted residential
placement during adolescence. She had self-injurious behaviors, including
onychotillomania and polyembolokoilarnania, as well as significant sleep disturbance.
Her behavior stabilized in the residential setting, and she was able to return home to live
with her parents at age 21. She continues to exhibit skin and nail picking, and when
excited, the “self-hugging” stereotypy typical of people with SMS. Her facial features
are subtly similar to those seen in patients with a 17p11.2 cytogenetic deletion.
Chromosome and fragile X analyses at age 17 were normal, as were results of subsequent
FISH studies for deletion 17p11.2. SMSISQ is a 19 year old patient who has a history of
mild mental‘ retardation, self-injury, and aggressive behaviors which resulted in
residential placement at age 13. Pregnancy and delivery were uncomplicated, birth
weight was 8 lbs. 1 oz. Motor milestones were significantly delayed (walked at 21
months), although he did not have infantile hypotonia. Speech developed within normal
limits. Since early childhood, the patient has exhibited “self-hugging” behavior when

excited. Macrocephaly was noted in infancy, and MRI studies of his head and spine at

115

age 11 revealed mild hydrocephalus, Arnold-Chiari malformation, and spina bifida
occulta. A temporal lobe cyst was also found and surgically removed. On physical exam
at age 13, he was obese and had facial features subtly suggestive of SMS. He had
macroeephaly, normal height, gynecomastia and hypogonadism. Results of fragile X and
cytogenetic studies were normal, including FISH analysis for deletion 17p11.2. DNA
was isolated according to standard protocols from peripheral blood or buccal cells from

these patients as well as parental and sibling controls, and Caucasian control samples.

Genomic DNA preparation: In order to isolate template suitable for PCR amplification,
DNA was isolated from peripheral blood or buccal cells from all putative SMS patients,

parental and sibling controls, and Caucasian control samples.

DNA isolation figm whole blood; If 5300 “L of peripheral blood was provided, genomic

DNA was extracted using the Puregene kit according to manufacturer’s instructions.
Briefly, 300 uL of whole blood was mixed with red blood cell lysis solution and
incubated for 10 minutes at room temperature. This solution was centrifuged for 20
seconds at maximum speed and the supernatant was removed, leaving the white cell
pellet and ~10 pL of residual liquid. The white cell pellet was resuspended and lysed
with Puregene Cell Lysis solution and treated with RNase A. The solution was then
treated with Puregene Protein Precipitation Solution, vortexed, and centrifuged for 3
minutes at maximum speed. The supernatant containing genomic DNA was removed to a
new tube, 300 pL of isopropanol was added, and the mixture was centrifuged at top
speed. The DNA pellet was washed with 70% ethanol, dried, and resuspended in 100 pL

of distilled water or TE buffer. If larger amounts of blood were supplied (~15-50 mL),

116

blood was centrifuged at 2000 x g, plasma was removed, and the remaining cells were
mixed with solution A (0.32 M sucrose, 10 mM Tris, pH 7.5, 5 mM MgC12 and 1%
Triton X-100) and placed on ice for 30 minutes. The mixture was then Centrifuged at
2500 RPM (NEED X G), the supernatant was removed, and 50 mL of solution A were
added to the pellet. This solution was placed on ice for 20 minutes, centrifuged as above,
and the supernatant was removed. The pellet was resuspended in a solution B consisting
of 10 mM Tris, pH7.5, 400 mM NaCl, and 2 mM EDTA, pH 8 and subsequently
digested overnight at 37°C with 100 uL of 20% SDS and 50 uL of 20 mg/mL proteinase
K solution. The following day, 3 mL of saturated phenol pH 8.0 were added to the
solution while rocking, and then the samples were centrifuged at 2000 x g for 15 mins.
The upper phase was removed with a Pasteur pipet and transferred to a new 15 mL
polypropylene tube and 3 mL of chloroformzisoamyl alcohol (24:1) were added to this
aqueous phase with rocking for 15 mins. Following another centrifiigation at 2000 x g,
the DNA upper phase was removed and the DNA was precipitated with 2 volumes of
95% ethanol. The DNA was removed with a Pasteur pipet and allowed to sit 70%
ethanol for 5 mins, then placed in 200 uL of TE, pH 7.5.

DNA isolation from buccal cells: Genomic DNA was isolated from buccal cells by
boiling the cheek brushes in 400 uL of 50 mM NaOH at 95°C for 10 mins. The brush

was then discarded and the sample was placed on ice for 10 minutes. This solution was

neutralized with 40 nL of l M Tris, pH 8.0.

RAIl PCR amplification and sequencing reactions: Analysis of RA]! in patient and

control DNA was performed by PCR amplification and subsequent sequencing and

117

analysis of PCR products. PCR primers covering the entire RAH coding sequence, 5’
and 3’ untranslated regions (UTR) and alternative splice variants (Table 7) were
generously provided by Dr. Laura Schmidt of NCI-Frederick or were designed by this
laboratory and synthesized at the Michigan State University Macromolecular, Structure,
Sequencing, and Synthesis Facility. PCR was performed in a 25 uL volume with 50-200
ng DNA template, essentially as described in Chapter II. PCR arr‘iplification was
performed in an ABI or MJ Research thermocycler with the following conditions (unless
otherwise noted; Table 7): initial denature at 94°C for 4 minutes, 30 cycles of 94°C for 1
minute, 64°C for 1 minute, and 72°C for 1 minutes, and a final extension of 72 °C for 10
minutes. In some cases, Qiagen 5x Q solution or Invitrogen 10x PCR enhancer solution
was added to difficult templates. In order to check each PCR amplification, 5 1.1L of the
reaction was electrophoresed in 2% agarose gels containing ethidium bromide.
Successful reactions were then purified using the Qiagen Gel Extraction Kit according to
manufacturer’s instructions or treated enzymatically in the following manner: 2 pL of
USB shrimp alkaline phosphatase (1 units/uL) and 1 uL USB exonuclease I (10 units/pL)
were added to 5 pL of PCR amplification mixture, the solution was mixed and incubated
at 37°C for 15 minutes in a thermocycler, and then inactivated at 80°C for 15 minutes. A
sequencing reaction containing at least 1040 ng of purified PCR product template in
distilled water and 30 pmol of sequencing primer (the forward or reverse PCR primer or
an internal primer) was then prepared and sequencing was conducted at the Michigan
State University Genomics Technology Support Facility using an ABI PRISM® 3100

Genetic Analyzer or ABI PRISM® 373 0x1 DNA Analyzer.

118

Sequence analysis: Sequence data from every RAII PCR amplimer was directly
compared by BLAST alignments to the published GenBank sequence for RAII mRNA
(GenBank M271790; AY172136) at the National Center for Biotechnology Information
(NCBI) (http://www.ncbi.nlm.nih.gov/BLAST/) as well as genomic sequence data for the
RAH genomic region (GenBank M271791; NT_010718). We searched fer putative
single nucleotide polymorphisms (SNPs) by searching the NCBI SNP database
(http:l/www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=snp) or directly through BLAT

alignments using the human genome browser (http://genome.ucsc.edu) .

Analysis of control samples: In order to determine that specific RAII mutations were not
present in relatives of the patients that we had identified as carrying RAII mutations, or in
the non-affected Caucasian population, we amplified control genomic DNAs with PCR
primers specific to RA]! coding exons and intron/exon boundaries (Table 7). The RA]!
exon 3 29 bp deletion could be detected by digesting the 493 bp exon 3 PCR product with
the restriction enzyme PspOMI, which recognizes the G/GGCC sequence present on the
non—deleted allele at position 361 of the amplimer, and allele and produces fragments of
361 bp and 132 bp (Figure 11). This restriction enzyme does not cut within the allele
carrying the deletion. Following PCR amplification, the samples were digested with 1-2
U of PspOMI at 37°C for 2 hours, then resolved on a 2% agarose gel in 1x TBE
containing ethidium bromide. The resulting fragments were photographed using
Alphalmager v5.5 software. The parents and one sibling of SMSIZ9 (Figure 11) were
analyzed by this method as well as >100 control samples (data not shown), and none

were found to harbor this deletion using this method, though mosaicism was

119

subsequently detected in one parental sample (SMSlZ8) by Transgenomic DHPLC
analysis (Figure 16). In order to assess whether the RA]! 5265de1C mutation found in
patient SMSlS6 was present in the parents of SMSlS6 or within the control Caucasian
population, >100 DNA samples were screened by PCR amplification and restriction
digest with the enzyme BgII, which recognizes the sequence GCCNNNN/NGGC
beginning at position of the present beginning at position 171 in the wild type allele of
the exon 3 amplimer, resulting in fragment sizes of 325 and 171 bp; Bgll does not cut in
the mutated deleted C allele (Figure 12). PCR products were digested with 4-5 U of
enzyme, electrophoresed, and photographed as above. Neither the parents of SMSISG
(Figure 12) nor the control samples was found to carry the 5265de1C mutation. As the
1449delC mutation identified in SMSlS9 and the 3801de1C mutation detected in
SMSl88 do not alter known restriction sites, in order to assess whether these mutations
were polymorphisms, DNA samples fi'om the relatives of SMSlS9 and SMSlS8 as well
as 100-200 Caucasian control chromosomes were PCR-amplified using exon 3 primers,
gel purified or digested with SAP and exonuclease I as described above, and directly
sequenced. Neither the parents of SMSlS9 or the mother and two sisters of SMSlS8 nor

any of the Caucasian control samples were found to harbor these delC deletions.

FISH: FISH was performed by Christopher Vlangos from the Elsea laboratory on
patient metaphase chromosomes isolated and prepared using standard cytogenetic
protocols. FISH probes were created using BAC 'or cosmid DNA by using a
commercially available nick translation kit to incorporate Spectrum Green or Spectrum

Orange dUTP by following manufacturer instructions. Probe DNA (lOOng BAC and

120

180ng cosmid) was precipitated, hybridized to metaphase spreads and washed per
manufacturer recommendations (V ysis Inc., Downers Grove, IL). Slides were
counterstained using Vectashield antifade with DAPI (V ector Labs, Burlingarne, CA).
Analysis of the FISH experiments was carried out on a Zeiss Axioplan2 microscope and
photographed with a Hamamatsu black and white camera. using Zeiss AxioVision

software version 2.0.

Northern analysis: In order to determine the embryonic expression pattern of mouse
Rail ‘, a SeeGene full-stage mouse conceptus northern blot containing total RNA from
mouse embryonic and extra-embryonic tissues was hybridized with ~106 cpm/mL of the
radioactively-labeled Rail EST clone (IMAGE:1211624) according to the standard
northern protocol as described in Chapter 11. As an RNA loading control, the
housekeeping gene Gapdh was subsequently radio-labeled and 106 cpm/mL was

hybridized to the same blot.

Denaturing High-Performance Liquid Chromatography (DHPLC): DNA templates
from SMSlZ9, SMSlS6, and 159, parental samples and randomly selected Caucasian
control samples as well as RAII primer sets were sent to Transgenomic, Inc. All PCR
reactions and subsequent DHPLC analysis were performed on-site by Transgenomic

personnel.

Temperature Gradient Capillary Electrophoresis ( T GCE): Standard PCR reactions
were performed as described above by Ellen Wilch in the Elsea laboratory, to amplify

known RAII mutations from SMSlZ9, SMSlS6, and SMSlS9 template DNA as well as

121

to amplify the same PCR products from control template DNA. Crude PCR reactions
were diluted 1:5 and 1:10 in distilled H20, dried overnight at 65°C, and shipped to the
laboratory of Dr. Pragna I. Patel at the Baylor College of Medicine. TGCE analysis was

performed by Anthony Rohr using Spectrumedix equipment.

122

Chapter IV.

In vivo evaluation of mouse Rail

Based on our identification of deleterious RAII mutations in several SMS patients
with no detectable deletion, we hypothesize that haploinsufficiency of RAIl is
responsible for the majority of the craniofacial and neurobehavioral abnormalities in
SMS. In order to assess Rail haploinsufficiency in a live animal model, we developed a
knockout construct to remove Rail expression. In addition, we assessed Rail dosage
sensitivity by constructing BAC transgenic lines and performing preliminary physical

assessment of these mice.

SMS mouse models

In order to determine whether Rail dosage-sensitivity could be assessed in a
whole animal mouse model, we wished to assess the in vivo effects of removal of the
Rail protein, as well as the effects of overexpression. The research focus of the Elsea
laboratory is to create mouse models for specific SMS candidate genes within the
deletion interval, as Opposed to the large-scale deletion analysis approach of the Lupksi
laboratory at the Baylor College of Medicine. This group has created targeted deletions
and duplications of a large region of mouse chromosome 11 which has homology of
synteny to the SMS common deletion region using Cre-loxP targeted recombination
chromosomal engineering in 129SS/SvaBrd ES cells (Walz, Caratini-Rivera et al.
2003). This targeted orthologous chromosomal region contains ~40 genes, most of which

are also present in the human SMS common deletion region, although synteny has not

123

been completely conserved through evolution (Bi, Yan et a1. 2002). Specific physical
characteristics were evident in the heterozygous deletion [Df(11)1 7/+] and duplication
[Dp(11)l 7/+] mice, which were evaluated in a mixal C57BL/6 Tyf'B'd x
129SSISvaBrd background. Specifically, the Df(l 1 )1 7/+ mice were significantly obese,
had marked craniofacial abnormalities including a shorter snout, were less fertile than
control littermates, and some portion of these mice had seizures (Walz, Caratini-Rivera et
al. 2003). The duplication mice were noted to be significantly underweight compared to
wild type animals (W alz, Caratini-Rivera et al. 2003) Behavioral characterization of
these Df(11)17/+ and Dp(1 I )1 7/+ mice was recently reported (Walz et al. 2004). Open-
field locomotor activity studies indicate that male Df(11)1 7/+ mice are hypoactive and
that male Dp(11)17/+ male mice are hyperactive (W alz, Spencer et a1. 2004). The male
duplication mice also demonstrated impaired fear conditioning specific to a context or
environment and the deletion mice showed significant circadian rhythm period
differences (Walz, Spencer et al. 2004). These Lupski lab mouse studies are an important
introduction to assessment of the SMS and dup(17p11.2) phenotype in mice. Several of
the physical aspects of SMS have been reproduced in these deletion and duplication mice,
demonstrating that haploinsufficiency of the gene or genes which are most important to
the syndrome, such as hyperactivity and circadian rhythm abnormalities, can produce
measurable effects in an animal model. Further behavioral assessment of the mice may
produce more insight into which mental or cognitive pathways have been affected by

gene-dosage.

124

Figure 19. Mouse chromsome 11 genomic region syntenic to the SMS deletion
region containing Rail BAC 326M22.

A portion of genomic region of mouse chromosome 11 syntenic to the SMS deletion
region is shown in this schematic adapted from the mouse genome browser
(http://genome.ucsc.edu). BAC 326M22 (GenBank AC096624), which was used for the
Elsea laboratory BAC transgenesis experiments, is highlighted. This BAC contains the
entire mouse Rail and Srebfl genes and the 3’end of T 0112112.

125

 

 

mam zmnxmnm 03 mmsmdan $39 mmnlmdnos I¢61.Q
mam zwnxmnm _

ZWUM
. _ _ __ : : :
flan-:....

r..4.._,......;., maxilla ..1-24...;

were. a m.

1..... «a .3 .... w m

. _ .
_ .

:. .aeufi .;.. annsrm. $322. $3; names:
we.» ae;:etrmz a Czwmxmm r.e.:
_reflaanw.

Lu .0. FW~J¢H a

......t-H m.....H...,.
mu a .. .34"
mLeihwflr
2.03:3 1.4-..23.......«.. u

. .. ..1 nu..Q. v
HF"... 71“.»... W....*

in.fl 4::is;

HIT-eta - 1*. ...13....?

126

etion
lain!
' .AKL
4‘: 4
the

3! "*3

l
1

While work using larger deletions as reported above shows that certain aspects of
the SMS phenotype can be reproduced in mice, it does not clarify the relationship
between specific genes and the phenotype. In order to further define important genes in
the SMS region, our lab is interested in creating BAC transgenics and single-gene
knockouts in mice. Based on our human RAII mutation screening, we believe that Rail
is most likely responsible for the majority of the physical and behavioral abnormalities
seen in SMS. Therefore, we have chosen to assess Rail dosage sensitivity in mice by
engineering our own gene targeting construct to reduce/eliminate Rail expression in vivo
as well as to create a stable Rail BAC transgenic lines to determine whether under or

overexpression of Rail can reproduce the phenotypic effects of SMS.

BAC DNA isolation

We developed our BAC transgenic mice in collaboration with the University of
Michigan Transgenic Animal Model Core (U of M TAMC). Through this arrangement,
our lab performed the necessary experiments to isolate and purify the BAC DNA and the
TAMC performed the pro-nuclear microinjection, as well as implantation of the
potentially transgenic mouse eggs into donor females. The BAC that we chose for
injection was RPCI-23 bc326M22 (GenBank accession AC096624), which contains the
entire mouse Rail gene as well as the hill-length Srebfl and the 3’ end of T om112. This
BAC is from a genomic library created from DNA from pooled female C57BL/6J mouse
DNA and has been fully sequenced by the mouse genome project. As there is >50 kb of
sequence upstream of the Rail translation start site intact within bc326M22, we believe

that all of the regulatory elements for correct tissue-specific expression of Rail should be

127

present in this BAC, although very little is known about the factors that are necessary for
the developmental expression of this gene. We chose to use a BAC transgenic rather than
a more specific transgenic construct to drive Rail expression because the Rail promoter
has not yet been fully characterized, and we may not be able to provide the correct
sequences to reproduce the cellular expression pattern. A disadvantage of using the BAC
transgenic is that other genes are also present within this relatively large piece (227,682
bp) of genomic DNA. As shown in Figure 19, the entire Srebfl is present in this BAC,
this gene has been very well-studied and overexpression may not have a measurable
phenotypic effect (Shimano et a1. 1997), though it remains a possibility that Srebfl
overexpression may confound our analysis. Since the 5’ end of Tom] 12 is not present in
this BAC (Figure 19), we believe that there will not be any expression from this gene.
No other genes have been identified in the genomic region present in bc326M22 to date,
though small, cryptic genes may exist. It is important to emphasize that the Rail BAC
transgenic is a first step towards understanding Rail overexpression and in the future,
more specific experiments involving Rail transgenesis may have to be performed to

verify our initial results.

Prior to microinjection, our lab isolated bc326M22 BAC DNA using essentially
the same modified Qiagen low—copy DNA isolation protocol that we had used to isolate
BAC DNA for 17p] 1.2 mapping studies (Chapter H Materials and Methods). Minor
adjustments were made to the final steps of washing and resuspending the bc326M22
DNA to ensure that this DNA was very clean, yet remained intact and resistant to

shearing. Following the recommendations of the U of M TAMC, we dissolved the DNA

128

in a standard microinj ection buffer (Schedl et a1. 1993), checked the concentration of the
DNA on a UV spectrophotometer and ran the DNA on a agarose gel to check the DNA
quality. We also linearized l p. g of the DNA with Natl and sent the circular and
linearized forms of bc326M22 DNA to U of M. The TAMC ran the digested and
undigested DNA on a pulsed-field gel to confirm the quality of the DNA prep and
readjusted the concentration of the DNA to 0.5-1.0 ng/uL for injection. The BAC DNA
was injected directly into the fertilized mouse eggs from two separate crosses: [C57BL/6
x (C57BL/6 x SJL)F1] and [(C57BL/6 x SJL)F1 x (C57BL/6 x SJL)F1] and then the eggs

were transferred to recipient mothers.

Assessment of Rail BAC transgenic founder mice

Three weeks following the microinjection of the bc326M22 BAC transgene and
transfer of the eggs to donors, 42 live pups ( 17 females and 25 males) were born. All
mice appeared outwardly “normal” and healthy. As hybrid strains were used for
transgenesis, genes for coat color were allowed to segregate randomly and were not
indicative of transgenic founder mice. Therefore, the 42 mice were agouti, black, white,
or yellow. In order to determine which animals were transgenic, we developed PCR
assays to amplify the specific T7 and Sp6 insert/vector ends of bc326M22. When BAC
DNA integrates into the genome, often a small portion of the bacterial vector sequences
are also integrated into the mouse genome, and these unique insert/vector sequences can
be amplified and distinguished from wild-type mouse genomic sequence. The PCR
reactions we designed could detect the bc326M22 spiked into the mouse genome from a

range of 0.1-100 BAC copy level. An example of the T7 and Sp6 PCR assays are shown

129

Figure 20. Copy standard and PCR evaluation of Rai1 BAC transgenic mice.

(a) and (b) PCR results of the bc326M22 BAC copy standard are demonstrated for the
Sp6 and T7 bc326M22 specific insert/vector ends against a wild-type mouse genomic
background. DNA from bc326M22 was spiked into mouse tail DNA at the
concentrations indicated below for 0.1x-100x BAC copies:

0.1x copy = 3.795 pg BAC DNA spiked into 1 pg tail DNA
1x copy = 37.95 pg BAC DNA spiked into 1 pg tail DNA
10x copy = 379.5 pg BAC DNA spiked into 1 pg tail DNA

100x copy = 3795 pg = 3.795 ng BAC DNA spiked into 1 pg tail DNA

BAC DNA was spiked into 1 pg tail DNA and 200 ng of template was used in the Sp6
and T7 specific BAC-end PCR assays. Wild-type mouse genomic DNA with no BAC
was used as a negative control and straight bc326M22 DNA was used a postive control.
The entire PCR reaction was run on a 2% agarose gel. The PCR reactions shown in (a)
and (b) were quantitative and were performed to demonstrate to the U of M TAMC that
our assays were sensitive enough to routinely distinguish transgenic and non-transgenic
animals.

(e) The T7 BAC end PCR assay was used to detect the presence of the BAC within
mouse genomic DNA from putative founder transgenic mice. DNA from blood samples
were used as template and this PCR was non-quantitative. The results of the T7 PCR
indicate that 754 does not carry this specific end of the BAC transgene and 753, 755, and
760 are transgenic.

(d) The non-quantitative Sp6 bc326M22 end PCR assay was used to screen putative F1

offspring sired by founder 755. F1 mouse 218 is positive by PCR for the integrated
transgenic Rail BAC and 213, 214, 215, 216, and 219 are negative.

130

 

 

s-(essa 33,1 aArrerrruuno

sKesse HDJ aAyerrrunnb-uo N

P"
hdarker

IRC)
Neg. control

100x
10x

PM ZZW9ZS°q L1.

 

§
e-
P
arker
H20
eg control

     
 

55—male founder

213
214

pm ZZW9Z£°¢1 MS

131

(ll

8

U'
'U

  

Blarker
H20

Neg. control
100x

10x

1x

pus zzmzsoq 9ds

0Jx
Pas. control

§

1 .0

,, hdarker
IQC)
PJeg.control

754-negafive

P“a ZZW9Z€°q L1.

753
755
760

in Figure 20, panels (a) and (b). Initially, at two weeks of age, mouse tails were clipped
by the TAMC and shipped to our lab for DNA isolation and identification of the putative
transgenic founders. The original DNA isolated from the tails using a phenol-
choloroform extraction was not able to be amplified by PCR. We attempted to PCR-
amplify the tail DNA with mouse B—globin specific primers, as the TAMC recommends
this template as a positive control for DNA quality, and no positive signals were detected.
As we were not able to identify the putative transgenic founders before the 42 pups were
weaned, all of the mice were transferred to MSU. Peripheral blood was drawn from the
foot of all of the animals, and DNA was isolated from whole blood using an alkaline lysis
protocol. All of the DNA isolated in this manner from the mice was amplifiable using
the mouse B—globin PCR primers. Utilizing the T7 and Sp6 specific primers (Table 9), 9
putative transgenic animals were identified that were positive for the Sp6 bc326M22 end,
or both the T7 and Sp6 BAC end. An example of one of the PCR assays used to identify
BAC transgenic mice is shown in Figure 20, panel (c). As the template DNA was not
quantitated before use in the PCR reactions, the T7 and Sp6 PCR amplifications were not
quantitative for bc326M22 copy number. Those animals that were positive for only the
Sp6 end may be carrying only a portion of bc326M22. According to the U of M TAMC,
it is common to see some portion of BAC. DNA which is not fully integrated to the

genome.
Copy number determination by Southern analysis

Once several putative transgenic founder mice had been identified by PCR, we

attempted to determine the number of copies of bc326M22 that had integrated into the

132

 

$152. 22:...

Heat»:— uonenuno Hutu;

”32.8 Seen—.8 8:3

gnaw—Em

85m. Ann“

 

me: 52:»—
Emma:

80>QOH>O>OO>OQH>HHOO>>

HOOOOH>>QHQO§QQH>HO>

at.

 

we: gain
Wagon"

O>HOOO>OHO§O>OH>HOHO

HO>>Q>Q>>>OQ>HQOHOO>Q>O

mu

 

38833 .3
9.6 one

08H00>OOAAO>O>A4QH>O

8H>>O>>OHOOOO>QHOO

me

 

 

c3338 mum
w>o one

Q>HOOHOOOQ>>30>OH>QHO

HOO>O>F~ HCnL Hr: C>>OQ

mw

 

32.8 Ham—02:4

OOEHQOOHO>O>O>OO>H>O>Q>OQQO>OO

 

OOSO>OQOHQHOO>>QHO>30>800>HOO

oo

 

3.03 a 8 @809
we: was?
SE28
roaowomocm
88523.53

o>>CCFACrCCCCEC>o

OHO>>OO>§OOO>O>>H

mm

 

328 e 8 889
we: moan-
Samoan
nose—omen
8859—558

 

 

QOOH>O>>>HOO>QOHOO>Q

 

H804800H>0>30>jn

 

mu

 

 

Haw—o e. wow ..1an 53 3.. win—58:2. o». 52.8 ~35 .26 $850 manomzam 833888 woe own: ommo 3.: mm

Eggnog.

€139. manages mos C 03$ H.250 A222.SataniganHoEPEB—v

133

Figure 21. Southern analysis to determine the integrated Rail BAC copy number in
founder mice.

In order to assess the copy number of the Rail BAC transgenic putative founder mice,
3.5 pg of genomic DNA from isolated from mouse tail was digested with HindIII and
control mouse genomic standards containing known copy numbers of BAC DNA were ..
also digested as follows: 3.5 pg of control genomic DNA from non-transgenic mice was
spiked with BAC DNA representing 1, 3, 5, or 10 copies of bc326M22, just prior to
digestion with 4U of HindIII. All samples were electrophoresed in 1% agarose gel,
transferred to a nylon membrane, and hybridized to radioactively-labeled probes using
standard protocols. The probes used to determine the bc326M22 copy number were the
3’ end Rail cDNA clone (IMAGE:1211624) in (a), the purified bc326M22 Sp6
insert/vector end PCR product in (b), and the purified T7 insert/vector PCR product in
(c). The .results of copy number calculations for transgenic founder mice from
densitometric analysis is tabulated below the autoradiographs. A standard curve was
generated from the BAC copy standard hybridizations and putative copy numbers derived
from the line equation of the standard curves (and rounded to the next whole number) are
indicated in the table for each transgenic founder. The shading indicates that these
animals were analyzed on a separate set of Southern blots (data not shown).

134

 

 

 

m

P‘

n

 

 

 

 

 

 

 

 

    

  

     

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

O m 2 O m
r2 m i «2 i 2am:
a mwpmw mm»MMe mpemw
rowwmefiflmafi aowwwhflSWZfi owmwm
W N135m77 77 MN13517H7M7 11:51
4 t I l C I a. I ... I, m m
I... . . .. _
:- 1. . ... . no . . . 2582.25
wet Emu. use—x" r I . . , Al— . .
. . . a, . . we .
w>n-uv25a .....we. , . .
O . _ til. i
m8 3.3.33» 2... E .5233» 2...
nee—E2. u>n ...q unaZNN wee uwagnn ere—.3153? Ara—.3552» Arc—.315..."
3383 F90 2:. Pro 2... w>n 33. w>n 83. w>n 2.3.
:...—52:» an” .62... wow 3...: ...—5&2. :...—:2. 5.592.
52.: gonna—H.52— 5. n22.332— fiq mono—.152— man
5.: Man. u>n 2... Fan 2:.
5.33522. r r ... . r p ... .
dw . + H . w
du + + A_ u w
«mo + + . a u
«mm. + + . - .A.
«3 + + A a w
:3 + + _ A
3m + + o A
«3 + + u _c A
.3» + + _w M _w

 

 

135

genome of each mouse. A standard Southern blot hybridization method was developed to
compare the positive hybridization signals of the transgenic mice to a standard curve of
non-integrated BAC DNA spiked into control mouse genomic at 1, 3, 5 and 10 copies
[similar to (Merscher, Funke et a1. 2001)]. All DNA was digested with Hinam, which
cuts the wild-type Rail gene at 2 positions within the coding portion of this gene. For
control samples, bc326M22 DNA was spiked into genomic DNA prior to restriction
digest. We used DNA probes specific to the 3’ end of the endogenous mouse Rail (EST
clone IMAGE: 121 1624), as well as the purified bc326M22 insert/vector T7 and Sp6 PCR
products to determine the quantitative hybridization signal. As shown in Figure 21 ,
hybridization with the T7 and Sp6 PCR products produced two signals, one specific to
the endogenous mouse genomic sequence that is present in the bc326M22 insert and one
unique to the BAC end. As predicted by the PCR reactions, not all of the mice were
positive for the T7 BAC end and all of the transgenic founder mice were positive for the
Sp6 end. Mouse 762, which was positive by PCR for the T7 bc326M22 end, was
determined to be negative by hybridization (Figure 21). It is possible that this sample
was contaminated during PCR amplification. Densitometry with background subtraction
was used to create a standard curve for each Southern blot and the results for were
tabulated and graphed using Microsoft Excel, plotting known BAC copy number vs.
positive spot density. A putative copy number for each transgenic unknown was then
calculated, using the linear equation from the standard curve. Results for the copy
number calculations using Southern blots are shown in Figure 21. While Southern blots
did confirm that the transgenic founders identified by PCR were actually carrying

positive hybridization signal specific to bc326M22, the hybridizations did not produce

136

clear, unambiguous copy number results. The Sp6 PCR product probe hybridization was
particularly difficult to interpret, as this probe contains a mouse-specific repeat and
produced a great deal of background hybridization that artificially increased the spot
density (Figure 21). Therefore the putative copy number results for the Sp6 hybridization
were very high compared to the T7 and Rail hybridizations. However, based on the
hybridization of the Rail EST, which produced the least amount of background signal,
some measurable differences in copy number were evident. Founder animals 760, 767,
and 775 may have ~1-2 copies of the Rail transgenic BAC, animals 753, 755, 762, 765,
and 787 may have ~3-5 copies of bc326M22 and animal 792 may have a very high

integrated copy number of bc326M22 (Figure 21).

Molecular assessment of Rail BAC transgenic Fl mice

‘Mature male founder Rail BAC transgenic mice (755, 760, 762, 767, and 775)
were mated to control female C57BL/6 mice purchased fi'om the Jackson Laboratory.
The founder males Were able to breed normally with the female mice and generally
produced litters of ~4—6 F1 generation pups. We began the evaluation of these mice by
taking tail biopsies fiom the 2—3 week old Fl mice and isolating DNA from the tails using
the 7U of M TAMC standard protocol. The tail DNA was first verified using the mouse
B—globinePCR reaction, then amplified with the bc326M22 T7 and Sp6 BAC end specific
primers (Figure 20, panel (1). Three of the transgenic founder males, 755, 760, and 775,
produced transgenic F l offspring; male founders 762 and 767 were not transmitting.
Southern analysis of DNA from F 1 animals compared to founder DNA (using the same

hybridization probes as in Figure 21) demonstrated that the offspring of 760 and 775

137

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

22.: w>n 2532. H3:— 2:_:a2. a. 2:552. S. 2:332. a. 25. 2:52. a. $2.83»? 3.
555220 3 :30; 55:2. 2. :o:- unsung? 5 5582.3 3 3552.? E "Samoan
35:2. :3: E engunaan Ban 2481:.“ .25.» a3:1:m 2:30 3.3.15
emu—.15 m.— Baou cam—.1:“
on... 1:
d: A no A: M m a 3e:
:2 u 3 MW A q o .3...
3”. N a A. A: . A: See

 

 

 

Haw—o 3. m:e::=:.w. 3. m.— afig 990 n

:2 3:88:20: m: "Em 820.

v
Jeannie we.

an:m:m 36:: 1.95:2 I 608532. nocu. Kan 2.95:2
:18 £08 :58: 8 0255: $528 2.: 9o eamcfiefi 3 3:8 208 BESS: 3. wow 8 :2333 3:2.
03:35 :88 2:.ng 3a Fawn—8: he: w>0. Ham 82o £8 506:8 3:69.: 2:55; 2:: we? 22 :5:
2537. was. :5: 2:: 8:89:33 :08 :2 80:88? $222 :5 2552 cm ”Sign m: .322 2.: 2E0: :53
88:8: wee 2:228: 2:: c2532:— m:o::m:om. HS: :56 35:03 220 :oségmamafi 9:: :5: 03.3% :8

138

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

23.553010 :— awning? E 23-30.05.030 5 .:...0::n0::0 E 005200
50:00 50.00 “0:00:00
05030 0000003032 0.1.2 :u:~ :u»: :«3
00:08: 03050: 0300000000 203:0: 203.0: 203:0: 203:0:
:00: . :.E. 00:: :02: .
ammo 83080:: 403:0: 403:0: 403:0: 403:0:
HE. m 403:0: 403:0: 403:0: 403:0:
mosh: 0:03:30: 403:0: 403:0: 403:0: 403:0:
830:8: 80:00:00 0:: 003 4030.: 403:0: 42.50: 403:0:
3:8:
:00: 8:33.30: 0:: 30:0: 203:0: 203:0: 203:0: 203:0:
50: . SEW
many? morn—m. :::0:0:. 203:0: 203:0: 203:0: 203:0:
80800.0:
(5:: Em. 03 0800::m 203:0: 203:0: 203:0: 203:0:
00080.0: 3:0: 203.0: 2085: 203:0: 203:0:
50:00:00 8 00:5 20:00: :0 3. 2030: 203:0: 203:0: 203.80:
8::
00:: 000000800: 403:0: 40:00: 203:: 403:0:
200: :0:::&:m 403:0: 403:0: 203:0: 403:0:
.950 80: 40050: 403:: 203:0: 403:0:
05:: 350::m. @0035 03 20:0 20:0 20:0 20:0
:::m:0: 8308030.:

 

 

 

 

 

 

H020 Z. 22.: w>n 30:30:? :0: 0:40:15 m €00: 0:050:20 000000503 :09. m0: 0:: 05:30:20
0000003030. :08. :.00 80:0 0:: 380:0 :.gmmoao 0300::m 0:, 30:0 300:0m0::0 :.0:::0a 2030 08:30: 0::
0030000: 8 :08 m2: 0:: :0:-:0:mm0::0 80:0 0:: :.030:0 03.00::m A00 00:30: :08. SEO: 0000003030 200.0
0:00 0030: 0:: 0: 90 :o:-=0:0m0::0 03015 0:. 2:0 :0:-30:m::3::m 30:00:. 20 0030:: 2:008:00: ::
30:0m0::0 0:: :0:.30:mm0::0 03:15 £030 :0:0:.

139

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2.00.390; .... ...—6389.0 .... 209.82.30.00 .... ...—.9898 ....
L 9.08 90.8 0090.8 890.8
953.890 8.8.9098“. an... an.» an». an...
0058.. g ...m H Sb 0... H80 .0.0LH0.u .0.~ H .ub
30850303.»...
0.000.088.8280 3... Hr... .....H ..0 uo.qH ..0 SQH ...
89%.; 00.0 H No 0... H N... 00.0 H Na .80 H w...
908.80 ..39 098.. 0.00 .... H 0.0 .... H 0.3 8.0 H 0.0 .0... Ho...
b.0898 ..39 0:80 8: 39H .5 N... H N. 8.0 H ...m .....H 0.0
roam... 0.. ....09. .0... .9... ...09 80 8 .98” ..39 80 8 .98. .88 80 8 .98” ....09 80 8 .98.
.0... H Nu .0... H .... .... H N. .0. H N
0089880...” 38.9880...” 0.08.9880...” 909.9880...”
...0Ho.m ...mHob .....H... ....Ho.m
gfigangg 039808.98. 089808.98. 009808.98” 009808.98”
.00 H u... ...... H .... ...u H u .0... H u
.39 .98 8 ...... ....09 .98 .0 ...... ..89 .98 8 ...... ..89 .98 8 ......
...0Ho... ...mH ..w ...wH. .....H ....
gags»? ...—09.88.88” 009808.98” 068808.88” 008808.98”
...q H ... .... H 0.. ...... H .5 .00 H .
....09 .98 8 ...... 0.09 .98 8 ...... .88 .98 8 ...... 0.09.98 8 ......
.u.. H P. ....m H ..q ....0 H ..0 .0. H ....
899.. 0280.. an... ..30 ..38 80 8 .98” ..38 .8 8 .98. .38 80 8 .98” 0.08 80 8 .98” .u
....0 H . .00 H 0.0 ...... H 0.0 H 0.0
..39 .98 8 ...... ..39 .98 8 .0... ....09 .98 8 ...... ....09 .98 8 ......
.00 H “.0 ....m H .... 2.0 H ..u .0. H ..u

 

 

...—20 .N. .8... Pro 39309.0 .... 030...: u 3.00.. ...—09.525 308080... 00.0. ...0. 0... 0:09.805
0000009098. 203% $.00... ..0. 90.0 .50 0080.0 59089.0 03.2.3 0.. 59099.0 00.5008 $0.0
0090.80 .50 008.880 .0 30.80 8.08 ...09 0.. 909-...90m0s8 90.0 .50 890.0 0.8.8.3 A... 009.8.
008. 8...... 0000009008 £08 .000 00.9.00 0:. 09 .00 009898098 ammunam 0.. ...0 900-...900.....=m
90.00.. U08..00 90.5.00 0.. £05... 90 .08. 000.. .0090 0.0 0.000080 .9 mica. N. .50 N... ......0 .8.
0.08 000000900. .0 :0. 80.0000 ... ...... 8.0.0 00 0080 .0. 0m001=m 20.0 9.00.... 90000.00 0. 300 50.000

0.. 09.0. 20 0.2.08 05.000908 .9 80.3890 090 90:-..090m030 03.0.8.5 20.0 90.00.

140

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

141

 

gin-3:85 3 :...-3:35 3 Zena—.8535 a 5.2.62.5 5
3:8 3:5. 55:8 55:5.
E358 38332.9“ an: :1» an»: :18
0358: 25:8— 5??? 22.5: 22.5: 2035: 20:2:
Ear—E :5 8:5. 89:
9mm 59583: 2035— 2035— 223: . 22.3:
8% 22.8: 22:5: 2035: 2035:
no.5: 3.85:8 225.: 20:8: 2035: 22.5:
2:55 55:88 E: «a. 3:5: 29.8:— 2035. 22.8: 2085.
36: 89.588 8.: :EEB 22.52 2085— 2035: 22.8:
5:55me
9.5.5:. :23? 53:8. 20:2: 2035: 2035: 2085.
858:8
«5:: E 2035: 22.5: 22.—E: 2085.
was...»— 3:9. 2°35: 2035. 22.8»— 2°35—
wavonuo 8 can: :55: 5 3. 2035— 2035: 2035— 29.5:—
8:
on: 88382: 22.8: 20:8: 29.8: 29.5:
28: EE 22.8». 22.8: 2035: 22.8»—
Hcco 83 29.5: 2035— 2035: 2085—
€:: 253.5. :85 o: 5:95: 225 225 225 225
896888
283%. 588.38.:
282555: 3:8 :53 3:85. 7:8 :55 3:85. 7:8 Sana 2:8 :55
83:3. ran—:5: on. 83:6. 95:5: on. 3:85. 59:5. 3:85 83:5.
58:9. 9. 0:555: 98:2. 9. 0:539: aa.—:5: on. 833. SEA—o: on. 98:3.
:5 :8: 5:53: :5 :8: 885:: 9. $525: :5 =8: 2 853:: :5 :8:
3:53: 8:51:

 

 

 

 

 

 

:.qu 5. 33.: Pro :...—3:85 3 cauaam 8 3.8.» :...—.595 888353 :98. 38332: :58 won
53:55:.“ 033:: 0:. :5 :55 Susannah“ 85:2 855 £53 8835: ms: 88:58: 8 :98 :..88 a: so?
:85:an emu—ism Ann 838. :35. 5%.: P4585855; $53 :8 8.3.5: 9: o: :5 soaégmmoao ommgam
om :5 35.525333 8:89. 20 3395 3:58:08 .538: ”38:35 2.: soaédnmmaao ammunam $53
508:. .155 8:5 3.: £2 :5: 50:50: 5 :5 825. mm Edam o». <39; E25583 $53 55: man 95
ago—can“: «$555853.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

23.385585 a: :.:—888:5 q: 2085:8835 a: :.:—5:585 m:
85:55 82:5. 88:8 88:8
8:895 588.8583 an: pl: 8": 3::
H0: 038:8 30:8: ”:.:Hob ~:H: ~:.:H :.: no.:H:.:
roam? 0:58: ::.:H 95 :N.: H 9: 3.: H 9» 3.: H 5.:
Q5888 :68 0:855:55 ::.cHo.: ::.eHo.u ::.:Ho.: ::.uHo.:
uwsuoogofiflga 8.3“: -.:H: ~:.uH:.: 35H :.:
gown—68:55:80 908858855“ , 38858855” #0885888” #0885888”
2.: H :.: :5: H: :9: H :.: :9: H :.u
:68 :88 8 :.::x :.88 :88 8 :.::H :68 :855 8 :8“ :68 :855 8 :8”
:9: H. : :9» H oh :9... H ob :9: H :.:
gong—588:8: :68858859 :.08858659 :68—858559 38858855”
:9: H 9: :5: H o.: :...... H : 2.: H :.u
:68 :88 8 Eu” :.38 :855 8 :8” :68 :855 8 :8“ :.08 :855 8 Eu”
:9: Ho: :9» H 0.: :9: H :.: :9: H ch
:88: 0255: :5: :8: :.08 85 8 :88” :08 85 8 :88” :68 85 8 :88” :68 85 8 :88”
:9: H 0.... :9: H 95 :5: H 0.» :5: H o:
:68 :855 8 :8” :68 :855 8 Ev” :68 :855 8 :8” :68 .855 8 :8”
:9: H :.~ :9mH :.: :9» H :.: :9: H :.u
Egg—0:68:18; :.08858555” #0885888” :.08858859 :68858859 :u
:9: Ho: :9: Ho: :9: Ho.» Hob
:.08 :855 8 :8" :68 :855 8 :8” :68 :855 8 :8” :.08 :88 8 Eu”
:9: H :.w :95 H :.: :90 H : :9: H :.:

 

 

:...—:5 :.:. :8: Pro :...—5:585 m: 03:18: 8 $.55: 2:55:56 8858858: :.:». :..0: 5::
988535 5555558585. 858:5 5:55 :.0: 85:5 8:: 5885 8:55:58,” 03:18: 0:. 885 385585
88:58 258 889:5: 8: 88:88: :0 358:5 $855 $08 5:: 808588585 855 8:: 88:5
0381:: A8 5083: :58. 8:8: 355558585 258 5:50 5535: 0:: 0:: :5 88:85:85 033:: 0:.
:5 80853858: 85:55:. U388: 585555 0:. 855:: 8:: 88: 50::. 588:. 55 955585: :8 3:85:

N: 58: NA. 20 03:08.: :5538555 8 885:585 8:: 808885885 03:13 25:6 808:.

142

consistently displayed the same restriction patterns as 760 or 775 (Figure 22),
respectively. However several of the offspring of 7 55 showed hybridization bands not
consistent with those in 755 (Figure 22). This suggests that 760 and 775 have one
bc3 28M22 chromosomal integration sites and that these males are stably transmitting the
BAC to their offspring (Table 10). In contrast, founder 755 may have multiple
integration sites and may not be stably transmitting this BAC to the next generation.

Table 10 summarizes our breeding experiments to date.

All of the F l offspring of all founders appeared normal and healthy. We
developed a battery of fairly simple and reproducible physical and behavioral
assessments to evaluate the health of F1 bc326M22 offspring. Most of the tests were
suggested by the text, What ’3 wrong with my mouse?: Behavioral phenooping of
transgenic and knockout mice (Crawley 2000) and a copy of our health assessment form
is supplied in Appendix A Health assessments were performed at 5 and 10 weeks of age
and will be ongoing until the F 1 mice reach one year of age. Preliminary qualitative and
quantitative data are presented in Tables ll-l4. All transgenic offspring were evaluated
alongside their wild type littermates. The only physical or behavioral differences that
were apparent between the bc3 26M22 transgenic mice and their normal siblings were
small differences in weight and total body length (measured from tip of nose to tip of tail)
and these data are presented in detail in Figures 23 and 24. The error bars on each graph
indicate one standard deviation above or below the mean. We were especially interested
in determining whether the BAC transgenic mice might be significantly underweight, as

this particular characteristic was noted in the Dp(1 [)1 7/+

143

Figure 22. Southern analysis of Rail BAC transgenic founders 755 and 77S and F1
offspring.

Southern analysis similar to the experiment depicted in Figure 21 was carried out to
determine if each founder male (755, 760, and 775) had the same restriction pattern as
offspring sired by each respective male. In order to analyze the restriction patterns, 3.5
ug of genomic DNA fiom each founder male or F1 offspring was digested with Hindfll,
electrophoresed in 1% agarose gel, transferred to a nylon membrane, and hybridized to
radioactively-labeled probes using standard protocols. The control lane included 3.5 ug
DNA from non-transgenic mice spiked with 3 copies of bc326M22 just prior to digestion
with HindIII. In the autoradiograph shown, the Rail 3’ end cDNA clone
(IMAGE: 121 1624) hybridized to three fiagments of DNA from founders 7 55 and 775
(indicated by red arrows), while the black arrows indicate several inconsistent bands in
the 755 F1 ofl‘spring 218, 257 and 261. Two of the bands seen in the 755 F1 offspring
are also present in the control DNA containing 3 copies of bc326M22, and these may
have arisen from partially digested DNA However, 218, 257, and 261 have multiple
inconsistent bands not seen in 755 or the control sample, suggesting that 755 may have
multiple BAC integration sites and may not be transmitting the Rail BAC transgene in a
stable manner. The 775 F1 offspring 252 shows the same restriction pattern as 775.
Hybridization patterns fi'om Fl offspring sired by 760 were also similar to 760 (data not
shown), suggesting that founder males 760 and 775 were stably transmitting the
integrated bc326M22.

144

UI
I-
5'

 

145

mice (Walz, Caratini-Rivera et al. 2003), though these mice were engineered in a
different genetic background. We measured the weights of all transgenic and non-
transgenic F l offspring of 755, 760, and 775 at S and 10 weeks. Gender played a
significant factor in the weight distribution and females were always statistically
underweight compared to males. However, as shown in Figure 23, the effect of the BAC
transgene on weight distribution is not entirely clear, though the transgenic offspring of
7 55 were significantly underweight (using two-way ANOVA) at 5 weeks compared to
their normal littermates. There did not appear to be significant interaction between the
BAC transgene and gender to affect weight distribution. The transgenic offspring of 755
also had a smaller total body length, though this difference was not significant and there
was a high amount of deviation, especially in the female mice (Figure 24). These results
may reflect the fact that the estimated BAC copy number in 755 is higher than 760 and
775 or this animal may have mutiple integration sites (Figure 21). Perhaps a higher
dosage of Rat] does contributeto reduced overall body weight and perhaps body length.
However, our animal numbers are low (especially the numbers of transgenic male
offspring of 755) and it is difficult to make a statistically significant assessment at this
time. When weight vs. body length was plotted in the 10 week old mice (Figure 25), the
755 F1 transgenic and non-transgenic offspring were distributed linearly, while the 760
transgenic and non-transgenic offspring showed similar patterns. It was difficult to
assess the 775 offspring as no non-transgenic females were born to this founder. We
continue to breed founder males 760 and 77S and will pursue the physical assessment of

greater numbers of F 1 Rail BAC transgenic offspring.

146

Figure 23. Weights of Rail Fl BAC transgenic animals and normal littermates at 5
and 10 weeks.

(a) The weights of the F 1 offspring of each "transgenic founder male are shown,
comparing non-transgenic males (dark blue) and females (cream; NTM and NTF,
respectively) to transgenic males (burgundy) and females (light blue; TM and TF) at 5
weeks. Error bars indicate one standard deviation above or below the mean for each data
set. An asterisk (*) indicates statistically significant results as described below (p< 0.05).

Statistical analysis of data from F1 offspring of 755: two—way ANOVA, p=0.009 for
gender, p=0.003l for the effect of transgenic vs. non-transgenic, and p=0.25 . for

interaction between gender and the transgene. Bonferroni posttests showed p<0.05 for
males“ and p>0.0S for females; n=2-8.

Statistical analysis of data from F1 offspring of 760: two-way ANOVA, p=0.003“ for
gender, p=0.85 for the effect of transgenic vs. non-transgenic, and p=0.10 for interaction
between gender and the transgene. Bonferroni posttests showed p>0.0S for both males
and females; n=6-8.

Statistical analysis of data from F1 offspring of 775 (only males could be analyzed as
there were no non-transgenic female offspring of 775): unpaired t test, p=0.22 for the
effect of transgenic vs. non-transgenic; n=4-6.

(b) The weights at 10 weeks of F1 offspring of each transgenic founder male are shown.
Total numbers for each subset are indicated and error bars indicate one standard deviation
below or above the mean for each data set. An asterisk (*) indicates statistically
significant results as described below (p< 0.05).

Statistical analysis of data from F1 offspring of 755: two-way ANOVA, p=0.0014* for
gender, p=0.099 for the effect of transgenic vs. non-transgenic, and p=0.89 for interaction
between gender and the transgene. Bonferroni posttests showed p>0.05 for both males
and females; n=2-8.

Statistical analysis of data from F1 offspring of 760: two-way ANOVA, p<0.0001* for
gender, p=O.75 for the effect of transgenic vs. non-transgenic, and p=0.25 for interaction
between gender and the transgene. Bonferroni posttests showed p>0.0S for both males
and females; n=6-8.

Statistical analysis of data from F 1 offspring of 775 (only males could be analyzed as

there were no non-transgenic female offspring of 775): unpaired t test, p=0.35 for the
effect of transgenic vs. non-transgenic; n=4-6.

147

_M.w~:_.n... <

 

r.

 

 

an: m2." 85:32.5 a £55: 536:8 :

 

 

»
1

43:82.5 5:: 30:85:85.5 855

l I. III if L

 

 

:ll'll lull..lll

he: as 85:32.5 3 £55: 566:8

 

Average weight in

 

 

 

 

 

 

for
for
for
for
M
lrs
E
It
am
or
{Mr
la
13‘
5

 

 

 

 

ti
5
b

:35

ml"

148

Figure 24. Average total body lengths of Rail Fl BAC transgenic animals and
normal littermates at 5 and 10 weeks.

(a) The total body lengths (measured from the tip of the nose to the tip of tail) of the F 1
offspring of each transgenic founder male are shown, comparing non-transgenic males
(NTM, dark blue) and females (NTF, cream) to transgenic males (TM, burgundy) and
females (TF, light blue) at 5 weeks. Total numbers for each subset are indicated and error
bars indicate one standard deviation below or above the mean for each data set. An
asterisk (*) indicates statistically significant results as described below (p< 0.05).

Statistical analysis of data from F1 offspring of 755: two-way ANOVA, p=0.l4 for
gender, p=0.034* for the effect of transgenic vs. non-transgenic, and p=0.95 for
interaction between gender and the transgene. Bonferroni posttests showed p>0.05 for
males and females; n=2-8.

Statistical analysis of data from F l offspring of 760: two-way ANOVA, p<0.0001* for
gender, p=0.22 for the effect of transgenic vs. non-transgenic, and p=0.91 for interaction
between gender and the transgene. Bonferroni posttests showed p>0.05 for males and
females; n=6—8.

Statistical analysis of data from F1 offspring of 775 (only males could be analyzed as
there were no non-transgenic female offspring of 77S): unpaired t test, p=O.77 for the
effect of transgenic vs. non-transgenic; n=4-6.

(b) The total body length at 10 weeks of F1 offspring of each transgenic founder male are
shown. Total numbers for each subset are indicated and error bars indicate one standard
deviation below or above the mean for each data set. An asterisk (*) indicates statistically
significant results as described below (p< 0.05).

Statistical analysis of data from F1 offspring of 755: two-way ANOVA, p=0.0032“ for
gender, p=0.ll for the effect of transgenic vs. non-transgenic, and p=0.92 for interaction
between gender and the transgene. Bonferroni posttests showed p>0.05 for males and
females; n=2-8.

Statistical analysis of data from F1 offspring of 760: two-way ANOVA, p=0.0084“ for
gender, p=O.77 for the effect of transgenic vs. non-transgenic, and p=0.97 for interaction
between gender and the transgene. Bonferroni posttests showed p>0.05 for males and
females; n=6-8.

Statistical analysis Of data from F1 offspring of 775 (only males could be analyzed as

there were no non-transgenic female offspring of 775): unpaired t test, p=0.71 for the
effect of transgenic vs. non-transgenic; n=4-6.

149

llld

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

”- mm: w>o 3:33.." a See: 33. 62... 83:. l—
.
m “Mm , u I,» I , t _
m. flmm r|l |l\‘r l ‘ {film/l H,
h .8 ‘
Wm 2m
0 Io
u m. m
m. a , k
m :... . a 2.42.
:a ,H I 42.
43:50:... 2... :o:.n3:ene:.o :..oe _
D 24.—.1
U _ _ D 4:
o _ mm... m>o "3:33... 3 .23.. 88. can? .035 a
. a”
m .8 ‘111 lit 1‘1
n .3 ‘ iii
w. .8 w l- r
a? +,
.m Io
. mm ‘ ,
m. 3... r l .
u ..no I- , w
M 2...
«mm 8.. _
43:83.0 3:: :o:..3:eue:.o 3.2. L
Hanna. mmm mmm mun awn Mama martin», MW

150

Figure 25. Weight vs. total body length of F1 10 week old Rail BAC transgenic
mice.

The average 10 weight of F1 offspring is plotted vs. average body lengths for each
transgenic founder male are shown. The offspring of 755 shown in (a), the offspring of
760 are shown in (b) and the offspring of 775 are shown in panel (c). Each symbol
representing a particular dataset is labeled with the genotype (transgenic or non-
transgenic) and sex of the animal. There were no non-transgenic female offspring of 775.

151

 

my.

 

 

 

2a 3 33:3 3 $8.. 5...... 5. 8% as... U.

59’:
OO

Avengebodylengthln
mm
s 5i
0 O

 

 

 

 

 

 

 

 

>3..er £06...» ... n33»

 

Average body length
In mm

 

 

a.

 

43 3 232.3 3 too.“ .206... <9 02...

$35

 

390
390

 

 

390
Bob

£7.sz

 

oo 24..

 

39°
30.0

 

 

 

8 we 8
>333 .206... _: :33...

50

 

 

 

 

44m 4... cam—.126 .3 seer <56... <9 can?

390
30.0
3.6.0
30.0
.390
..Aob

Average length In mm

.265

 

MES

 

e2;

 

 

o

 

 

 

 

4

>330.“ £06.: .: @333

..o no mo

 

 

152

At this time, all of the Rail BAC transgenic animals appear to be normal and healthy
(Tables 11-14), and only a few animals displayed the characteristic lower body, weight
that was seen in the Dp(11)17/+ mice (Walz, Caratini-Rivera et al. 2003). None of the
mice appeared to display phenotypes similar to individuals who harbor a duplication of
the entire SMS region (Potocki, Chen et a1. 2000), though these features can be subtle.
However, while we have determined that bc326M22 is present within the genome of
these transgenic animals (Figure 21), we have not truly resolved the copy number of each
founder or F1 BAC transgenic animals and, most importantly, we have not determined
whether there truly is expression from this BAC. Real-time PCR assays were designed to
assess copy number, but remain to be optimized and successfully applied using genomic
DNA. A smnmary of these experiments is presented in Appendix B. And, future Rail
expression studies may show physiological expression from bc326M22. Also, though we
did not observe any outward differences in the F1 mice, we intend to sacrifice ~5
transgenic and non-transgenic animals (males and females) to examine their organs for

any gross abnormalities and to perform skeletal evaluations.

Cloning of the Rail knockout vector

In order to create a mouse model of reduced or absent RaiI expression, we
developed our own specific Rail knockout vector. We cloned two separate targeting
aims of homologous sequence from the mouse Rail genomic region into the plasmid
targeting vector, pNZTKz (this plasmid contains a aeomycin resistance cassette for
positive selection, the lac_Z_ gene, and a thymidine l_(inase gene for negative selection).

The targeting arms are regions of sequence homologous to the endogenous Rail gene to

153

allow for correcting targeting and replacement of a large portion of Rail, including the
translation start site, with the lacZ gene and the neo cassette (Figure 26). Similar to the
BAC transgenic animal model described above, we performed the molecular biology to
create the Rail knockout vector and the U of M TAMC carried out the cell culture and
electroporation of the targeting plasmid DNA into Bruce4 C57BL/6 ES cells. The
C57BL/6 genetic background will be retained throughout our gene-targeting experiment
so we will not have to backcross the animals to a pure genetic background‘for evaluation.
The C57BL/6 genetic background has been used extensively for behavioral evaluations
and several standardized studies have been published using this inbred strain. The
targeting vector pNZTKz was chosen for our studies because it contains the B—
galactosidase lacZ gene, which should replace the endogenous Rail gene (along with the
neo cassette) if targeting occurs correctly. A map of the pNZTKz knockout vector and
the cloned Rail fragments is shown in Figure 26. Each arm of the targeting vector, the
2.4 kb intronic fragment and the 5.4 kb internal fragment, was amplified by PCR from
bc326M22 template using high-fidelity polymerase. Each PCR amplimer was then
cloned separately into an Invitrogen TOPO cloning vector, sequenced to confirm, then
cut out of the original cloning vectors and ligated into the pNZTKz cloning vector. A
detailed explanation of the cloning steps is provided in the Chapter V. Materials and
Methods section. All putative pNZTKz clones containing both of the Rail gene
fragments were sequenced to confirm the correct sequence and orientation. In
preparation for microinjection by the U of M TAMC, DNA from the pNZTKz knockout
vector containing both the Rail. intronic and internal fragments was isolated using the

Qiagen endo-free kit to reduce contamination from endotoxin, and then linearized

154

Figure 26. Rail knockout construct.

A 2.4 kb Rail intronic region fragment was cloned immediately upstream of the start site
of Rail translation behind the lacz site in the pNZTKZ vector and a 5.4 kb internal
sequence representing the 3’ end of Rail and the UTR, was cloned into the second
polylinker site of the pNZTKz vector. ApaLI (A) and EcoRI (E) restriction sites within
the Rail genomic region and the pNZTKz vector are depicted, as well as the probes (a
and b) which will be used in Southern analysis to detect the presence of the endogenous
and targeted Rail genes. Following genomic DNA digestion with ApaLI and EcoRI,
probe a should anneal to a 5.5 kb fragment in the wild type Rail and 2.8 fragments in the
properly targeted gene. Probe b should anneal to a 9.2 kb fragment in the endogenous
gene and a 7.3 fragments in the recombinant Rail.

155

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Ta... ....r a.» .6 _
gnome—2... hat fll_ wag .... A

a _ 6.3:... .838: 8 5.8.8.: a Ah .9 98:3. u a... E. w

u... 5.
IX ...... . .. .... j. .......... ....

:2... 425.5 a. ........... «Rana—s..- ‘ as...” we...»
gunman—i <82...

Mm. a iv _ .2. ....

 

 

 

 

 

 

 

vl

156

utilizing the unique AscI restriction site contained within the vector. The linearized DNA
was purified by phenol-chloroform extraction and 200 pg of the RailszZTKz targeting
plasmid DNA was sent to the U of M TAMC for electroporation into Bruce4 C57BL/6

ES cells.

Assessment of gene-targeted mice by Southern analysis

As shown in the diagram depicting the Rail knockout vector targeting of the
endogenous Rail gene (Figure 26), there are 3 Alw441/ApaLI (these enzymes are
isoschizomers) within this genomic region. One specific ApaLI site lies within the
portion of Rail that should be replaced by the lacZ and the neo cassette from pNZTKz if
the gene-targeting within the ES cells has occurred correctly. Therefore, this ApaLI site
should be eliminated. A double restriction digest of ApaLI and EcoRI was used to
identify the ES DNA which contains one copy of the endogenous Rail and one copy of
the lacZ and neo gene-targeting construct. The new, introduced EcoRI sites from the
vector polylinker should be present only in gene-targeted DNA. Probes a (5’) and b (3’),
which lie outside of the targeting arms have been chosen as probes in Southern blot

hybridization (Figures 26, 27).

Following injection of the linearized pNZTKzzRaiI targeting construct into the
Bruce4 C57BL/6 ES cells, the U of M TAMC shipped 5, 96-well plates containing ES
cell DNA for screening by our lab to identify homologous recombinants. We digested
the ES cell DNA with ApaLI and EcoRI restriction enzymes and created Southern blots,

which were subsequently hybridized with probes a and b. As shown in the

157

Figure 27. Southern analysis of ES cell DNA containing the targeted Rail knockout
construct.

(a) In the 5’ probe a Southern hybridization, the endogenous 5.2 kb fragment is
visualized only in the normal ES cell DNA (represented by plate 3 well E10) and in the
targeted ES cell DNA from plate 3 well E13, plate 1 well C12, plate 1 well D9, plate 2
well A3, and plate 3 well H12, both the endogenous 5.5 kb fragment and the targeted 2.8
kb fragment are resolved.

(b) In the 3’ probe b Southern hybridization, the endogenous 9.2 kb fragment is
visualized only in the normal ES cell DNA from plate 3 well E10 and in the targeted ES
cell DNA samples from plate 3 well E13, plate 1 well C12, plate 2 well A3, and plate 3
well H12, both the endogenous 9.2 kb fragment and the targeted 7.3 kb fragment are
resolved. The probe b hybridization from plate 1 well D9 was too faint to discern and is
not shown.

158

m
E
3
9. m
P

l
1
E
3
......
P

  
 

. All 83%..

gnome—:...." nachos—5"

... Al 93...... l saga
533 a 05 5.8 _ g

«...—engeeulV iv «gouge—6'7 ..l

All 3328 . ,H. All 8822.
330 u in
...-8 u .5

m. ....er a vaulfiusae:

 

«2.82.2.1 _.,

22.82.25

       

., Al 3822.
:30 a 05

«5.825

3:3 N >u 5:8 u :5

u. Educ u 5.2.3.332.

159

autoradiographs in Figure 27, probes a and b hybridized to the expected fragments sizes
in the endogenous and targeted ES cell DNA. Five putative homologous recombinants
were identified by Southern hybridization: plate 1 well D9, plate 3 well E11, plate 1 well
C12, plate 2 well A3, and plate 3 well H12. Several other possible homologous
recombinants were identified which showed the recombinant fi'agment with probe a but
the probe b hybridization was ambiguous (often too faint or too much background to
clearly distinguish). We also used a probe for the full-length Rail gene to detect any
rearragements but this probe lies within the targeting vector, so correct targeting to mouse
chromosome 11 would not be distinguished from random integration. The ES cells from
all possible homologous recombinants will be expanded. More DNA will be harvested
from the expansions and the Southern hybridizations with probes a and b will be repeated

for confirmation before the ES cells are injected into albino C57BL/6"2j/"2j blastocysts.

Conclusions

The majority of the molecular biology experiments have been completed to
establish stable Rail BAC trangenic lines and to inject Rail gene-targeted ES cells into
blastocysts to produce chimeric mice. The preliminary assessment of the BAC transgenic
mice revealed that perhaps higher Rail copy number can contribute to a smaller overall
body length and less total weight, though the copy number of the these mice needs to
confirmed and greater number of animals are required for statistically significance. No
other obvious physical or behavioral abnormalities were noted. At least four putative

homologous recombinant ES cell lines were identified by Southern hybridization which

160

carried the targeted Rail construct. These will be confirmed with more ES cell DNA

before injection into mouse blastocysts.

Materials and Methods

' Mouse copy number calculations: Copy number calculations were determined for
RPCI-23 bc326M22 (GenBank Accesssion AC096624) using the copy standards for PCR
genotyping protocol devised by the U of M TAMC
(http://www.med.umich.edu/tamc(spike.html). For these calculations, the haploid mouse
genome is assumed to be 3 x 109 base pairs (bp) and the size of the bc326M22 BAC

insert is 227,682 bp. To determine the copy standard for 1 ug of genomic DNA, half of

this amount (0.5 pg) is used in the calculations, as transgenic mice are hemizygous:

 

mass of transgene DNA = N bp transgene DNA

0.5 ug genomic DNA 3 x 109 bp genomic DNA

mass of transgene DNA = 227.682 bp insert of transgene DNA

0.5 ug genomic DNA 3 x 109 bp genomic DNA

mass of transgene DNA = (0.5 u enomic DNA) (227.682 bp DNA)

 

 

3 x 10 bp genomic DNA
= 0.00003 795 ug transgene DNA

= 37.95 pg

The copy standard calculations calculated that 1 copy of bc326M22 = 37.95 pg added to
1 ug of genomic DNA. Following this standard:

0.1 copies bc326M22: 3.795 pg spiked into 1 ug genomic DNA
1 copy bc326M22: 37.95 pg spiked into 1 pg genomic DNA
10 copies bc326M22: 379.5 pg spiked into 1 ug genomic DNA

100 copies bc326M22: 3.795 ng spiked into 1 ug genomic DNA

161

PCR:

Amplification of the mouse Rail genomic reghm for cloning into the knockout vector:

 

 

Template DNA from RPCI—23 bc326M22 was amplified with primers from the genomic
region immediately 5’ upstream of the Rail (T able 9) translation start site (hereafter
referred to as the intronic fragment) as well as with primers specific for the central and 3’
end of Rail (hereafter referred to as the internal fragment). These. fragments were cloned
into the Rail knockout vector, pNZTKg, (generously provided to the Elsea laboratory by
Dr. Richard Palmiter at the University of Washington) as targeting arms to direct
integration of the pNZTKz lacZ gene and neomycin resistance (neo') cassette into the
mouse genome. The intronic and internal fragments were amplified by standard PCR
protocol (see Chapter 11 Materials and Methods) using the high-fidelity Accuprime Pfx

polymerase (Invitrogen).

T7 and S96 PCR: In order to type putative Rail BAC transgenic founder mice, PCR
assays were developed to specifically amplify the RPCI-23 bc326M22 Sp6 and T7
insert/vector BAC ends. As these specific BAC ends are often‘ incorporated into the
mouse genome, the PCR assays were designed to show that we could detect the BAC at
various copy levels within the wild type mouse genomic background (see above for copy
number calculations). Prior to PCR amplification using standard conditions, the
appropriate amount of BAC DNA for 0.1, 1, 10, and 100 bc326M22 copies was spiked
into 1 ug control mouse genomic DNA. Sp6 and T7 bc326M22 end PCR amplification
was carried out using a 200 ng template, utilizing the genomic DNA spiked with BAC

DNA (Figure 20). The entire PCR reaction was electrophoresed on a 2% agarose gel in

162

1x TBE and visualized. Once we had established a sensitive T7 and Sp6 PCR assay to
detect the BAC ends, we assessed the putative Rail BAC transgenic founder mice
utilizing template DNA isolated from whole blood using crude alkaline lysis. One uL of
DNA from the putative founders was first cycled using primers specific for mouse [3-

globin to ascertain the DNA quality, then analyzed with bc326M22 insert/vector specific

Sp6 and T7 primers.

Cloning of the Rail knockout vector: Following amplification of the Rail intronic and
internal fragments with Invitrogen Pfx high-fidelity polymerase, 4 uL of the PCR
reaction to amplify the internal fragment and 2 uL of the intronic PCR product were
separately cloned into the Invitrogen pCR-XL-topo or the pCR—blunt II-TOPO vector,
respectively, following manufacturer’s instructions. Ligation was allowed to progress for
5 minutes, then 2 uL of the cloning mixture was immediately transformed into TOPO
OneShot chemically competent E. coli cells and plated onto LB agar containing
kanamycin. Colonies formed after overnight grth at 37°C and putative transformants
were picked and grown in LB and kanamycin broth culture. These potential
transformants were screened by PCR to amplify the polylinker site of the pCR-XL-topo
or the pCR—blunt II-TOPO, using -20M13 Universal and M13 reverse primers. Positive
TOPO transformants were identified that contained either the Rail 2.4 kb intronic or the

5.4 kb internal fragment.

Once the Rail intron and internal fragments had been amplified and cloned

separately into a TOPO vector, plasmid DNA was isolated from the positive clones and

163

then digested with specific enzymes prior to successive cloning into the pNZTKz
knockout vector. Approximately 0.5 pg of DNA from the Rail internal TOPO clone and
1 pg of DNA from pNZTKz was digested with SpeI and NotI. The linearized Rail
internal fragment (released from the TOPO vector) and the linearized pNZTKg DNA
were then electrophoresed onto DE-81 ion-exchange paper and purified using the
“tombstoning” method. The paper segments containing DNA were microcentrifuged in a
0.65 mL eppendorf tube with a slit in the bottom (placed within a 1.5 mL eppendorf tube
to trap the flow) in order to release excess gel running buffer. The 0.65 mL eppendorf
tube containing the ion-exchange paper was placed within a fresh 1.5 mL' eppendorf tube,
and the paper was incubated for 3 minutes in 100 pL high tombstone buffer (1 M LiCl,
10 mM Tris-HCl, pH 7.6, 1 mM EDTA, and 20% ethanol), then centrifuged at room
temperature for 3 minutes. Following centrifugation, the eluent was retained; the
washing and centrifirgation steps were repeated three times. DNA was precipitated from
the combined 400 pL of high tombstone buffer by adding 1 mL of 95% ethanol,
incubating at —80°C for 15 minutes and centrifuging the mixture at maximum speed for
10 minutes. The purified DNA from pNZTKz and the Rail internal fragment was
resuspended was resuspended in TE buffer. Each fragment was electrophoresed on a 1%
agarose gel and the concentration of the DNA was estimated using the Invitrogen 1 kb
ladder DNA marker. The internal fiagment and the knockout vector were ligated
together using T4 DNA ligase at 2:1 insertzvector ratios. The ligation reaction was
allowed to proceed for 1 hour at room temperature and the ligation mixture was checked
on gel. The ligation then continued overnight at 16°C in a thermocycler. The ligation

mixture was diluted 1:5 in 1x TE buffer prior to transformation into Invitrogen DHSa

164

high-efficiency E. coli cells. The transformed cells were then plated on LB agar
containing ampicillin. Several individual colonies were picked and grown overnight at
37°C in LB and ampicillin broth cultures. Putative Rail intemalszZTthransformants
were screened by PCR using primers specific for pNZTKz polylinker 2, using 2 pL of
overnight bacterial broth culture as template. One colony was isolated that contained the
5.4 kb Rail internal fragment cloned into pNZTKz in the correct orientation; the insert

was also confirmed by direct. sequencing.

In order to clone the 2.4 kb Rail intron fragment into the knockout vector, 1 pg of
DNA from Rail intemalszZTKz transformant was then cleaved by the Mid restriction
enzyme within polylinker 1 and 1 pg DNA from the TOPO cloned intronic fragment was
digested with NheI and XbaI. The intronic fragment would be correctly ligated to the
NheI ends of pNZTKg as the ends of NheI and XbaI are cohesive. The Mid and XbaI
digest released the intronic fragment from the TOPO vector. This fragment was
electrophoresed on a 1% gel, cut out of the gel, and purified according to manufacturer’s
instructions using the Qiagen gel extraction kit. The NheI-digested DNA from the Rail
internalszZTKz transformant was electrophoresed through a 1% gel onto DE-81 ion
exchange paper, and purified using the “tombstoning” method described above. Similar
to the ligation of the larger internal fragment, the linearized intronic and Rail
internal:pNZTK2 fragments were electrophoresed on a 1% agarose gel and the
concentration of the DNA was estimated using the Invitrogen 1 kb ladder DNA marker.
The intronic fragment and the intemalszZTKz DNA were ligated together using T4
DNA ligase at 3:1 and 2:1. insertzvector ratios. The ligation reaction progressed for 1

hour at room temperature, after which the ligation mixture was checked on gel. The

165

ligation then continued overnight at 16°C in a thermocycler. The ligation mixture was
subsequently diluted 1:5 in 1x TE buffer before transformation into Invitrogen DHSa
high-efficiency E. coli cells. The transformed cells were then plated on LB agar
containing ampicillin. Several individual colonies were picked and grown overnight at
37°C in LB broth cultures containing ampillicin. Putative pNZTKz transformants
containing the 2.4 kb Rail intron fragment ligated into the NheI site within polylinker 1 .
were identified by PCR using primers specific for pNZTKz polylinker 1, amplified using
2 pL overnight bacterial broth culture as template. In total, five colonies was isolated
that contained the 2.4 kb Rail intron fragment cloned into pNZTKz in the correct
orientation (as well as the previously cloned internal fragment); All inserts were

confirmed by direct sequencing.

In preparation for microinjection by the U of M TAMC, DNA from the pNZTKz
knockout vector containing both the Rail intronic and internal fragments was isolated
from 250 mL LB and ampicillin broth culture using the Qiagen endo-free kit (in order to
reduce contamination from endotoxin), and then linearized utilizing the unique AscI
restriction site contained within the vector. The linearized DNA was purified by 1:1
phenol-chloroform extraction, followed by 24:1 chloroform: isoamyl alcohol extraction,
and precipitation with l M NaCl and 2 volumes of absolute ethanol. The purified DNA

was resuspended in 200 pL of sterile 1x TE buffer at a concentration of 1 pg/pL and

approximately 200 pg of DNA was sent to the TAMC.

166

DNA isolation:

BAC DNA: In preparation for microinjection into the fertilized eggs from two separate
crosses: [CS7BL/6 x (C57BL/6 x SJL)F1] and [(C57BL/6 x SJL)F1 x (C57BL/6 x
SJL)F1] for the creation of the Rail BAC transgenic mice, RPCI—23 bc326M22 BAC
DNA was isolated as previously described (Chapter 11 Materials and Methods), except for
the following minor modifications to prevent shearing of the BAC DNA After the
elution of the BAC DNA with buffer QF from the Qiagen purification column, the BAC
DNA was precipitated with 0.7 volumes of room temperature isopropanol and then
centrifuged at _>. 15,000 x g at 4° C for 30 minutes. The supernatant was then decanted
and washed twice with 5 mL of 70% ethanol. The DNA pellet was recovered after
centrifirgation at 15,000 x g at 4°C for 5 minutes. The ethanol was then removed and
the DNA pellet was allowed to dry for 30-60 minutes. We resuspended the pellet in 200
pL of microinjection buffer (10 mM Tris-HCL, 0.1 mM EDTA, pH 8.0, 100 mM NaCl,
and a spermine/spermidine polyamine mix), visualized the DNA on. a 0.7% agarose gel,
stained with ethidium bromide, and quantitated the DNA using an Ultrospec 2000 UV
spectrophotometer. Prior to shipping the purified BAC DNA to the U of M TAMC), 1
pg of the bc326M22 DNA was restricted with NotI. The U of M TAMC verified the
concentration of bc326M22 DNA and checked the quality of the digested and undigested

DNA on a pulsed-field gel.
Plasmid DNA isolation: Plasmid DNA from pNZTKz, and Invitrogen pCR-XL-topo or
the pCR-blunt II-TOPO was isolated utilizing either the Qiagen mini spin prep or the

modified Qiagen midi-prep DNA isolation for greater quanitities of DNA. Plasmid DNA

167

was isolated from a 5 mL bacterial culture (LB containing appropriate antibiotic), using
the Qiagen mini prep kit according to manufacturer’s instructions. The protocol for the
modified midi-prep was the same as for the cosmid DNA isolation (Chapter II Material
and Methods). All plasmid DNA was electrophoresed on a 1% agarose gel and

quantitated using an Ultrospec 2000 UV spectrophotometer.

Mouse tail DNA isolation: DNA was isolated from mouse tail biopsies of 0.5-1.0 cm
using a standard protocol modified from the U of M TAMC website
(www.med.umich.edu/tDNA.html). The tail tips were resuspended in 600 pL of TNES
lysis buffer (10 mM Tris, pH 7.5, 400 mM NaCl, 10 mM EDTA, and 0.6% SDS; and 35
pL of 10 mg/mL proteinase K solution and incubated overnight at 55°C. The digested
solution was then treated with 166.7 pL of 6 M saturated NaCl, vortexed vigorously for
15 seconds and centrifirged at 14,000 x g for 5 minutes at room temperature. The
supernatant was removed to a new tube and the DNA was precipitated with 1 volume of
95% ice cold ethanol. The DNA was pelleted by centrifirgation at maximum speed for 10
minutes and dried for 15-30 minutes. The DNA pellet was then resuspended in 100-200

pL of TE (10 mM Tris, pH 8.0, 1 mM EDTA) and allowed to dissolve at room

temperature or at 65°C for 10 minutes.

DNA isolation from whole blood for PCR: In order to isolate genomic DNA from whole
blood, 500 pL of 1x red blood cell (RBC) lysis buffer was added to 100 pL of whole
blood, mixed well and placed on ice for 10 minutes. The solution was then

microcentrifuged for 30 seconds at maximum speed and. the supernatant was removed

168

and discarded. The pellet was resuspended in 500 pL of 1x RBC lysis buffer and spun
again at maximum speed in a microcentrifuge for 30 seconds. The resulting supernatant
was carefirlly removed and discarded and the RBC lysis was repeated if the pellet
remained red. Subsequently 400 pL of 0.05 N NaOH was added to the cell pellet, which
was then heated at 95°C for 10 minutes and chilled on ice for 10 minutes. The solution
was neutralized with 40 pL of 1M Tris, pH 8.0. This DNA isolation produced DNA
appropriate for PCR template though generally did not produce clean enough DNA for

Southern analysis.

Sequencing: Direct dye terminator sequencing using ABI PRISM® 3100 Genetic
Analyzer, the Applied Biosystems 3730xl DNA Analyzer, or the ABI PRISM® 3700
DNA Analyzer was performed by the Michigan State University Genomics Technology
Support Facility (GTSF) to confirm all cloning reactions. Sequencing reactions were
prepared according to GTSF requirements
(http://genomics.msu.edu/protocols/Custom_DNA_Amounts.html) and consisted of 1 pg
of double-stranded template plasmid DNA and 30 pmol of sequencing primer in a

volume of 12 pL.

Southern analysis:

Analysis of Rail BAC transgenic mouse copy number: In order to assess the copy
number of the Rail BAC transgenic putative founder mice, 3.5 pg of genomic DNA from
isolated from mouse tail was digested with 4 U/pg of HindIII for ~16-18 hours at 37°C.

These genomic DNA digests contained 2.5 mM of spermidine. Standards containing

169

 

known copy numbers of BAC DNA were also digested as follows: 3.5 pg of control
genomic DNA from non-transgenic mice was spiked with insert pg representing 1, 3, ‘5,
or 10 copies (see above for copy number calculations) of b0326M22 just prior to
digestion with 4U of HindIII and 2.5 mM of spermidine. All samples were
electrophoresed in 1% agarose gel in 1x TAE buffer, transferred to a Hybond—N+ nylon
membrane, and hybridized to radioactively-labeled probes using the standard protocol
described in Chapter 2. The probes used to determine the b0326M22 copy number were
the 3’ end Rail EST (MAGE21211624), and the purified Sp6 and T7 PCR products

representing the specific insert/vector ends of bc326M22.

Aim/gs of Rail knockout homologous recombinants: In order to identify Bruce4
C57BL/6 ES cell DNA containing the correctly targeted RailszZTKz construct
integrated within the mouse genome, 3 pg of DNA was digested with ApaLI and EcoRI
restriction enzymes overnight at 37°C. The digested genomic DNA was electrophoresed
on a 1% agarose gel in 1x TAE, transferred to Hybond-N+ membrane using standard
techniques, and hybridized to radioactively-labeled probes a and b (Figure 26) as well as
a plasmid containing the full-length Rail gene, using the standard protocol described in

Chapter II.

170

Chapter V.

Discussion

The role of RAII in SMS

As this work has described, our molecular analysis of the basis of SMS has
produced several promising lines of research. The generation of the physical and
transcription map of the critical interval and subsequent EST analysis identified many
novel transcripts and several possible candidate genes for the phenotype of SMS.
Continued deletion and breakpoint analysis from SMS patients with rare or unusual
deletions has enabled our group to filrther delineate the SMS critical interval. It was
especially important for us to take advantage of the valuable information provided by
SMS patients with no deletion. While many clinicians doubted that these individuals
truly had SMS, we trusted the diagnosis of our collaborators, especially Brenda Finucane
of the Elwyn Institute, who had a great deal of practical experience dealing with SMS
patients. We were convinced that several of these non-deletion patients displayed
characteristic features of SMS, especially the self-abusive and stereotypic behaviors.
Utilizing DNA from these putative SMS patients, our lab undertook a brute force
mutation screening effort to discover single-gene mutations. We were ultimately
rewarded when dominant, deleterious mutations were identified in the RA]! in four
patients (Table 5, Figures 11-14). This result recasts SMS as a possible single-gene
disorder, similar to AGS. Based on the phenotype of those individuals harboring RAIl
mutations, we believe that RAIl haploinsufficiency is likely responsible for the physical

and neurobehavioral aspects of SMS, which are present in >75% of patients (Table 5).

171

Other characteristics of SMS which are less common, including large organ abnormalities
and cleft palate (Table 5), are likely due to the contribution of other genes within the
deletion interval. In regards to the developmental role of other possible SMS candidate
genes, our lab is especially interested in DRGZ, which is extremely conserved through
evolution and T OMILZ, which may be implicated in membrane. trafficking. There are
several noteworthy genes that map to the SMS deletion interval within 17p11.2 whose

cellular role is currently under investigation.

The true cellular and biochemical role and any possible interacting molecules of
the RAN protein is also unknown at this time. The RAII amino acid sequence contains
few clues about possible protein-protein or DNA interactions, except for the putative
PHD domain, a nuclear localization site, predicted glycosylation sites, and short regions
of sequence similarity to the transcription factor T CF20 (Rekdal, Sjottem et al. 2000).
While the evolutionary relationship between TCF20 and RAIl proteins is not well-
understood, aa analysis demonstrates that both proteins contain PHD (or PHD-like
domains in the case of TCF20) at their C-terminal ends (Rekdal, Sjottem et al. 2000)
(Figure 8). TCF20 contains an ~80 a long cysteine and histidine-rich zinc finger domain
at the C-terminal end of this protein termed a ZNF2 or a PHD/LAP domain, which is
closely related to classic PHD domain (Lyngso et al. 2000; Rekdal, Sjottem et al. 2000).
Although PHD domains are known to occur in several transcriptional regulators such as
trithorax (Aasland, Gibson et al. 1995), their function is poorly understood. Inna recent
publication by Lyngso et al., this group used different fragments of TCF20 as bait in a

yeast two-hybrid experiment and characterized one of the interacting proteins as ring

172

finger 4 (RNF4). However, only TCF20 constructs not containing the PHD/LAP domain
showed positive interaction (Lyngso, Bouteiller et al. 2000). Deletion of this single
domain within the C-terminus restored interaction with RNF4 (Lyngso, Bouteiller et al.
2000). This group concludes that the PHD/LAP domain is capable of sponsoring intra-
chain interactions within the TCF 20 protein and may compete for external zinc finger
binding, and therefore acts as a negative regulator of cofactor interaction. Possibly this
function occurs in RAIl and other PHD containing proteins (Lyngso, Bouteiller et al.
2000). It is also interesting to note that TCF20 is not grouped with any of the traditional
family of transcription factors (i.e. those with a consensus DNA binding domain), yet it
enhances the activity of several well-characterized transcription factors such as c-Jun,
Sp1, and Pax6 (Rekdal, Sjottem et a1. 2000). Perhaps RAII and TCF20 are related

members of a family of transactivation proteins.

Future research directions in the Elsea laboratory are focused on in vitro and in
vivo studies of [MI] and the mouse homolog Rail, recapitulating some of the TCF20
functional studies (Rekdal, Sjottem et al. 2000). Promising preliminary experiments from
the Elsea laboratory as well as IHC experiments performed by collaborators suggest that
RAIl/Rail is present in the nucleus of mammalian cells (G. Rouleau, personal
communication). Subsequent studies are now underway to identify Rail/RAH
interacting molecules. Interestingly, the only protein that has shown positive interaction
with the mouse Rail protein was the zinc finger region of ch20, which bound a portion
of Rail in a yeast two-hybrid experiment (Rekdal, Sjottem et a1. 2000). In our research

lab, the full-length Rail protein will be used as bait to attempt to pull-down other

173

 

interacting proteins. It will be useful to also focus on the PHD domain and proteins that
may specifically bind this motif, as well as to know if the PHI) can sponsor RAIl/Rail

protein dimerization.

RAIl mouse models

The in vitro analysis described above should provide clues about Rail
biochemical pathways. At the same time, detailed phenotypic analysis of the Rail
knockout mice may implicate known developmental signalling pathways. This work has
focused on the molecular experiments necessary to develop the Rail knockout.
Subsequent injection of the recombinant Bruce4 C57BL/6 ES cells (Lemckert et al. 1997)
by the U of M TAMC into albino C57BL/6°2j/°2j blastocysts which lack the tyrosinase
locus (Schuster-Gossler et al. 2001) should produce Rail gerrnline chimeric animals.
The chimeras will be carefully monitored for any obvious abnormal physical or
behavioral features and then bred to C57BL/6 females to monitor transmission of the
targeted Rail gene. Mice hemizygous and homozygous for Rail disruption will be
rigorously assessed with physical and behavioral measurements similar to the battery of
tests run on the Rail BAC transgenic mice (Appendix A) and will also be sacrificed and
examined internally for gross organ defects. More sophisticated tests such as Morris
water maze, open field assessment, curiosity tests, circadian rhythm measurements, and
conditioned fear assessments can be applied if the mice appear severely affected, either
physically or behaviorally (Crawley 2000). Based on the Lupski lab’s targeted
disruption of mouse chromosome 11 orthologous to the SMS common deletion interval,

Df(1 l)l7/+ mice heterozygous for this deletion demonstrated marked physical and

174

 

behavioral features that mirrored some of those seen in patients with SMS (W alz,
Caratini-Rivera et al. 2003; Walz, Spencer et al. 2004). A similar set of experiments to
our Rail knockout mice may be performed on mice developed from purchased, targeted
mouse ES cells from the German Genetrap Consortium (GGTC). According to the GGTC
website (http://genetrap.gsf.de), one 129S2 ES cell line (MO73F06; GenBank CL215311)
developed by their consortium contains a genetrap insertion between noncoding exon 2
and coding exon 3 (Figure 7), which is designed to sponsor alternative gene splicing into
the “trap” instead of the normal Rail mRN A transcript. If genetrap splicing is successful,
this Rail genetrap ES cell may create a disruption of the entire gene, though it is possible
that leaky splicing “around” the gene trap vector can at low levels (estimated to be ~10%
from the GGTC website: genetrap.gsfde/info/faqs.html#1) may produce an Rail
hypomorph. We intend to purchase this ES cell clone for injection into C57BL/6
blastocysts. Mice derived from this method will be in a mixed 129 Sv and C57BL/6
background and may display a slightly different phenotype than our original C57BL/6
knockouts. A disadvantage of using this cell line is the amount of backcrossing required

to produce a pure C57BL/6 line for behavioral analysis.

RAM and craniofacial development

Craniofacial development is largely controlled by the migration, placement, and
expansion of cells derived from the cranial neural crest (Gilbert 2000). Most of the face,
including the jaw, teeth, and facial cartilage is a product of neural crest cells. These
unique cells, the only neural crest cells to produce cartilage and bone, originate in the
hindbrain, which is arranged into compartments called rhombomeres. A gradient of

overlapping Hox gene expression delineates the rhombomeres and specifies the migration

175

pattern and fate of the neural crest cells. Cells from different rhombomeres take different
migration pathways and are fated to form different tissues; this migration is spatially and
temporally controlled (Gilbert 2000). Cells from rhombomeres 1 and 2 travel to the first
pharyngeal arch and form the jawbones and the malleus bones of the ear, as well as the
bones of the face through a phenomenon. of expansion called the frontonasal process
(Gilbert 2000). Cells from rhombomere 4 then migrate to the second pharyngeal arch,
followed by cells fi'om rhombomere 6, which travel to the third and fourth pharyngeal
arch as well as the pouches which will go on to form the thymus, thyroid glands and the
parathyroid. The neural crest cells in rhombomeres 3 and 5 join migrating cells which
travel on either side of the rhombomere (Gilbert 2000). Other neural crest cells from the
fore-and midbrain play a role in the formation of the frontonasal process, palate, and
mesenchyme in the first pharyngeal arch. Other cranial neural crest cells originate the

mesenchyme that produces the neck bones and muscles (Gilbert 2000).

In the Rail hemizygous and homozygous mutant mice as well our BAC
transgenics, it will be important to assess in the mouse embryos whether or not the
rhombomere structures and functions are intact. Whole mount immunostaining and in
situ hybridization will be used to evaluate the gene expression appropriate to various
regions of the central nervous system. Several well-characterized markers will be used to
visualize the neuroectoderrn (0tx2), the midbrain/hindbrain boundary (En-2),
rhombomeres 3 and 5 (Krox20), the notochord and floor plate (Shh), and rhombomere 4
(Hoxb-l) (Alavizadeh et al. 2001). In most cases, an antibody against the appropriate

markers can be purchased and used for immunostaining during embryonic days E85-

176

E10.5. Briefly, this procedure involves fixing the embryos overnight in a mixture of
methanol/DMSO, then a bleaching step followed by rehydration in phosphate-buffered
saline (PBS). The embryos are then treated with a blocking step before the primary
. antibody is added. Following several washes, the secondary antibody is added and then
detected with the appropriate substrate. If the immunostaining yields unsatisfactory
results, whole mount in situ using DIG-labeled cDNA probes will be used to identify the
CNS structures (Alavizadeh, Kieman et al. 2001). In addition to the assessment of
rhombomere structure, we intend to determine if there are any gross abnormalities in the
patterning of cranial nerves in the Rail gene-targeted and transgenic mice. Whole mount
immunohistochemical analysis with an anti-neurofilament antibody, using the basic
protocol described above, will allow for the visualization of cranial nerves (Alavizadeh,
Kieman et al. 2001). In all experiments, normal wild type littermates will be used as

controls.

In addition to in-depth embryonic analysis, we will perform craniofacial
measurements on newborn and three-month old animals. In their experiments with the
heterozygous Df(1 l)l 7/+ mice, the Lupski group found that while newborn mice did not
display obvious craniofacial defects, adult mice demonstrated significantly shorter skulls
as well as broader and shorter snouts and nasal bones compared to wild type mice (Walz,
Caratini-Rivera et al. 2003). We intend to conduct similar measurements for various
cranial points, such as the nasal bone, the anterior notch on the frontal process, the
intersection of the parietal and intraparietal bones, and the intersection of the interparietal

and occipital bones along the midline (Walz, Caratini-Rivera et al. 2003). Measurements

177

 

will be conducted on live and sacrificed animals (Walz, Caratini-Rivera et al. 2003).
Skeletons will be prepared similarly to Walz et al. Briefly, cartilage will be stained with
alcian blue and acetic acid, fixed in ethanol, and treated in 2% KOH until clear. The
skeletons will be stained in alizarin red and KOH and cleared with a 1:1 mixture of
ethanol and glycerol. All gene-targeted and transgenic mice will be compared to age-

matched wild type littermates.

Other possible roles of RAIl/Rail in development

Another significant physical finding demonstrated in the Df(ll)l7/+ mice was
marked obesity (Walz, Caratini-Rivera et al. 2003). As the heterozygous mice harboring
a duplication of the similar chromosomal region [Dp(l 1)] 7/+] were found to be
significantly underweight, this result suggests that gene dosage effects may influence this
trait (Walz, Caratini-Rivera et a1. 2003). We have preliminary evidence that RAIl may
influence obesity, as three of the patients harboring deleterious RAIl mutations who have
been evaluated by our collaborators (SMS129, SMSlS6, and SMS159), are all
significantly overweight. Though we currently have a small number of patients, RAII
screening continues in order to identify other individuals with RA]! mutations whose
physical traits we can evaluate. As mentioned in Chapter I, the ongoing SMS life history
study at the NIH should provide valuable information about typical body weight change
in SMS patients through adult life. We will also perform fi'equent weight measurements
on Rail knockout mice to determine whether obesity is a trait manifested in these
animals. The Lupski group noted that the both the Df(l l)! 7/+ and the Dp(1 [)1 7/+ mice

were significantly underweight during the first month of life compared to wild type mice

178

and that the Dp(1 1)] 7/+ mice remain so throughout their entire lives (Walz, Caratini-
Rivera et al. 2003). In contrast, beginning at four months of age, the Df(11)1 7/+ mice
weigh significantly more that the control animals (Walz, Caratini-Rivera et al. 2003).
This finding strongly suggests that RAII or other dosage-sensitive genes within the SMS
deletion interval is playing a role in regulating some level of the metabolism controlling

overall body weight.

Perhaps the most striking behavioral defects measured in the Df(l 1)] 7/+ mice
were the circadian rhythm defects demonstrated by significant period length differences
(Walz, Spencer et al. 2004). This rare biological phenomenon directly correlates with the
inverted circadian rhythm of melatonin and lack of REM sleep demonstrated by SMS
patients (Potocki, Glaze et al. 2000; De Leersnyder, De Blois et a1. 2001). Preliminary
data also suggests that patient SMSlS9, who carries a de novo RAll deletion (Figure 13),
also has a measurable inversion of the circadian rhythm of melatonin (A. Smith, personal
communication). We hypothesize that RAIl haploinsufficiency has multiple, pleiotropic
effects and perhaps may have a global influence on development not only of the cranial
neural crest, a metabolic pathway affecting overall body weight, but also the synthesis or
secretion of melatonin. We intend to collaborate with Dr. Maya Bucan in the circadian
rhythm evaluation of our Rail mouse models. Dr. Bucan’s lab has produced elegant
studies of the activity and rest cycles of various mutant mice and has devised the proper
equipment for these studies (Pickard et a1. 1995; Nolan et al. 1997; Kapfhamer et al.

2002). Most of these behavioral studies were completed in the C57BL/6 mouse genetic

179

background, which is one of the most important reasons that we wished to use C57BL/6

Bruce4 ES cells (Lemckert, Sedgwick et a1. 1997) for our Rail targeting project.

Neurological development, retinoic acid and RAII

RAII signalling may also have a developmental effect on the neurological system,
which could be extremely sensitive to gene dosage at critical embryonic time points. It is
tempting to link Rail/RAH to neuronal development simply because of its strong
upregulation following treatment with retinoic acid (RA) (Imai, Suzuki et a1. 1995). The
Rouleau group also analyzed RAII cDNA fi'om SKHSH neuroblastoma cells induced
with RA and identified the S’UTR of RAII (Toulouse, Rochefort et al. 2003). Upstream
of RA]! non-coding exons 1 and 2 are significant stretches of CpG dinucleotides, as well
as a RA-responsive element (RARE), which demonstrates that RAIl may be directly

inducible by RA (Toulouse, Rochefort et al. 2003).

RA is low-molecular weight, lipophilic substance derived from vitamin A.
Generally, animals take in these retinoid compounds from meat or B-carotene in plants.
Cells acquire RA from the blood system, where RA exists bound to a retinol-binding
protein (Maden et a1. 2003). Once inside the cell, retinol is coverted to retinal and then to
RA by retinal dehydrogenases (Maden and Hind 2003). Separate isomers of RA exist,
which subsequently act through different receptors. RA synthesized and modified by the
cell can enter the nucleus and activate or repress gene activity by binding to ligand-
activated nuclear transcription factors (Maden and Hind 2003). In humans as well as

other mammals, there are classes of these transcription factors: retinoic acid receptors

180

and retinoid X receptors; each class contains several isoforrns. These receptors then
heterodimerize and recognize RAREs in the promoter regions of RA-responsive genes
(Maden and Hind 2003) similar to the sequence found in the 5’ upstream region of RA]!
(Toulouse, Rochefort et al. 2003). It has not been determined whether the RARE in the
RAN promoter region has functional significance, though a gel shift assay containing this
sequence and various retinoic acid receptors could determine in vitro if certain receptors

bind to this RARE.

It has been known since the 1950’s that RA is highly teratogenic when applied to
pregnant animals and can affect multiple organ systems, including the central nervous
system (CNS) (Maden 2002). It was subsequently disCovered that RA added to cultured
embryonal carcinoma cells induced the differentiation of these cells into neurons and
glia. RA has the ability to regulate in vivo neuronal differentiation through the action of
several hundreds of genes, including transcription factors, enzymes, cell-surface
receptors, enzymes, and growth factors (Maden 2002). Recent experiments show that
RA has an active role in the overall patterning of neurons along the anteroposterior and
dorsovental axes (Maden 2002). Signalling by RA seems to be especially crucial for the
embryonic development of the hindbrain and anterior spinal cord, where RA acts to
position posterior rhombomeres and generate appropriate boundaries (Maden 2002).
Another interesting role for RA was demonstrated by Lee et al., who showed that RA and
bone morphogenetic proteins have the ability to specify the frontonasal mass and
maxillary prominences of chicken embryos. RA levels specifically determine whether

certain regions of the face form maxillary or frontonasal forms (Lee et al. 2001). It seems

181

likely that RA]! could play a crucial role in the RA signalling pathway both in developing
CNS as well as in craniofacial patterning. It will be very important to examine the
hindbrain of Rail heterozygotes and knockout mice for intact expression of various
rhombomere markers mentioned above and to assess overall cranial neural patterning. It
will also be useful to assess the learning capacity of Rail mutant mice, as mice lacking
RA receptors often have severely impaired abilities in the Morris water maze, which
measures spatial learning and memory (Maden 2002). Also, very little is known about
RA signalling in the adult animal and whether certain adult-onset neurological diseases
may be related to dysfunctional RA signalling (Maden 2002). It has been proposed that
the etiology of some forms of schizophrenia may be related to RA. Dopamine receptors
are regulated by RA and- the dopaminergic system is often abnormal in schizophrenia
patients (Viggiano et al. 2003). It is interesting that RAIl has also been implicated in the
etiology of schizophrenia and could be the link between the RA signalling pathway and

this complex mental disease.

A possible role for RAIl in ADHD and schizophrenia?

As the developmental significance of RAM is still unknown, we can speculate that
perhaps other neurological disorders besides SMS that may be caused by perturbations in
the amount or quality of the RAN protein. A recent article by Ogdie et al., has reported
positive linkage for attention deficit/hyperactivity disorder (ADHD) to a portion of
chromosome 17p11 near marker D1781857 (which maps to a genomic region just distal
to the SMS-REPD), using data from 204 nuclear families comprising 270 affected sib

pairs (ASPs) (Ogdie et al. 2003). The maximum multipoint LOD score for the 17p11

I82

genomic region was 2.98, which is strongly suggestive for linkage, though the candidate
gene region was actually very broad and encompassed 17pll through 17q11 (Ogdie,
Macphie et al. 2003). While not definitively implicating RAII in this genomewide scan,
RAII would be a very good ADI-1D candidate gene based on the hyperactivity and
attention-seeking behavior displayed by the four patients analyzed who carry RAII
mutations, though the serotonin transporter gene (5H TI) maps to 17q11, which several
studies have also linked to ADHD (Ogdie, Macphie et al. 2003). It may be possible to
screen individuals with ADHD for RAII mutations, though it is difficult to assess the
biological significance of missense mutations within the RA]! coding sequence, as well
as sequence changes within non-coding regions of this gene. As we do not know the true
cellular role of the RAI] protein or the structure of the wild type protein, we cannot
determine at this time which changes may have biological significance and which are
essentially silent. We could speculate that some form of ADHD may be influenced or
caused by hypomorphic RAII mutations or perhaps even a dominant negative effect that
an incorrectly functioning RAII protein may enact on the normal copy. It is also likely
that promoter mutations or those that alter the regulation or the temporal expression of
RA]! could produce a form of ADHD or even the entire SMS phenotype if the mutation
was severe enough. Of course, the genetic background of the individual also plays a
crucial role and may further complicate the interpretation of missense mutations. A great
deal of research is currently ongoing into genetic factors that contribute to ADHD and
most of the emphasis has been on the genes within the dopamine pathway (Sonuga-Barke
2003; Viggiano, Ruocco et al. 2003). RAM may provide insight into an alternative

biological pathway of neurological and ultimately behavioral development. This

183

pathway may contain other particularly dosage sensitive genes which, when altered, can

produce ADI-II) or possibly even SMS.

As discussed in Chapter 11], prior to our identification of RAII mutations in SMS
patients, many of the publications from'the Rouleau group focused on M11 as a possible
candidate gene for schizophrenia or perhaps other neurological disorders (Joober,
Benkelfat et al. 1999; Hayes, Turecki et al. 2000; Toulouse, Rochefort et al. 2003).
These studies were not particularly compelling because they were largely theoretical or
statistical and this group offered no biological evidence to support their hypotheses. The
Rouleau group was especially interested in the CAG repeats within RAM and whether
they may affect the age of onset of SCA2 (Hayes, Turecki et al. 2000) or the association
of these repeat regions with responsiveness of certain schizophrenics to conventional
neuroleptic treatment (J oober, Benkelfat et al. 1999). No biological confirmation of these
conjectures was forthcoming by the Rouleau group or others. But a recent genomewide
linkage study involving both 353 ASPs as well as a single pedigree strongly suggests that
a locus or loci within l7p1 1.2-q25 does contribute to schizophrenia (Williams et al.
2003). The maximum LOD score using ASPs for thiS'genomic region was 3.35, though,
this candidate region also includes the 5H 77‘ serotonin transporter gene. However, RAIl
may also be a priority candidate for detailed sequencing or perhaps haplotype and SNP
analysis in this patient population in order to determine whether there may be an

association with schizophrenia.

184

Implications of RAII mutation screening

To date, only 4 of the 12 putative SMS patients we have sequenced were found to
harbor deleterious RAII mutations. We cannot definitively determine that all of these
non-deletion patients truly have SMS, especially SMS] 17 and SMS] 19,, who display
distinctive, non-SMS characteristics (Table 6). Since our RAII mutation screening has
concentrated on the protein coding region and the 3’UTR, it is possible that many of
these individuals have deleterious mutations within the RAN promoter, regulatory, or
enhancer regions. It would be useful to measure the RM] protein levels in all putative
SMS patients with no deletion compared to parental or wild type sample, to determine if
RAIl is truly haploinsufficient in these individuals. Western analysis of RM] prior to
mutation screening may be a way to prioritize which nondeletion patients to screen for
sequence changes, though we may only be able to easily obtain lymphoblasts for this
analysis, which may not show RAII expression. But it is also possible that some
nondeletion patients may carry a mutation or mutations within a gene in the same
developmental pathway that RAIl plays a role in, or perhaps in a direct signaling target
gene of RAII. Mutations within a single biochemical pathway can produce a similar
phenotype in some cases. It is also important to recall that within the cohort of AGS
patients without 20p12 deletions, only ~70% were found to cany known JAG] mutations
(Piccoli and Spinner 2001). Though our pool of SMS patients without 17p] 1.2 deletions
is much smaller than the AGS patients studied, we also have a significant percentage
without identified mutations. In both AGS and SMS, it will take time to firlly understand
which gene or genes and ultimately, which developmental pathways that have been

disrupted in these particular individuals. Following our identification of RAII mutations

185

(Slager et al. 2003), many more nondeletion patients with an SMS phenotype were
referred to our group. By sequencing RA]! in these individuals, we will obtain a clearer

picture of the type of RA] 1 mutations that exist within this population.

Another SMS locus?

In conclusion, we can speculate on the existence of another SMS locus
somewhere in the genome. As our cohort of patients with SMS features but no detectable
l7p1 1.2 deletion grows, it is likely that several may be found to carry mutations in RAII,
though we may still be left with a number of individuals with no obvious RAII mutation.
Perhaps there exists another, etiologically-distinct, genetic syndrome with an extremely
similar phenotype to SMS. Or perhaps there is another locus for SMS, which may be
associated with a separate gene mutation. As suggested above, once the biological role of
RA]! is known, other genes in the same biochemical pathway can be screened for
mutations. Dominant mutations in these genes may have the same developmental efi‘ect
as those within RAII. Potentially, T CF20, which seems to be evolutionarily related to
RAII, could be screened for mutations as well. It is interesting to note that human T CF 20
maps to chromosome 22q13 (Rekdal, Sjottem et al. 2000). Within this particular
chromosomal region are at least three known genes which are homologous to genes
within the SMS deletion interval. While most of the genes within 22ql3 do not have
homologous counterparts to those in 17p11.2, RASDZ (GenBank NM_014310; formerly
termed Rhes) was localized to 22ql3 by sequence analysis, but the original rat Rhes was
cloned based on its similarity to RASDI, which maps to the distal end of the current SMS

critical interval (Figure 2). TOMILZ, which maps to the middle region of the SMS

186

critical interval, shares several similar domains with T 0M1 (NM_005488), which maps
to 22q13, though T OMILZ mRNA expression on a northern blot (Table 1) suggests that
this cDNA is at least twice as long as T 0M1 . As sequencing analysis continues, other
genes may be identified within 22q13, which may prove to be homologous to SMS
candidate genes. We do not have any data in the Elsea laboratory about the possible
evolutionary relationship between the 17p] 1.2 and 22q13 genomic regions and no direct
sequence comparisons have been published. As the biological fiJnction of these genes is
just beginning to emerge, we also cannot determine at this time whether these
homologous genes have a truly similar cellular function. It may be unlikely that 22q13
harbors another SMS locus, but as the molecular components of SMS are just beginning
to be understood, at this time we cannot rule out any hypothesis. It may also not be a
coincidence that positive linkage for bipolar disorder and schizophrenia (Liang et al.

2002) has been mapped to this genomic region as well as 17p11.2.

Conclusions

At this time, the biological implications of identifying mutations in a single gene,
RAII, in SMS patients are seemingly limitless. Since so little is known about the cellular
role of this gene and whether it is truly involved in RA signaling or mental and
craniofacial development, that we can speculate about the various embryonic pathways
that could rely on RAII for proper evolution. The global effect of RAII
haploinsufficienCy could lead to SMS and perhaps missense or hypomorphic mutations
could produce ADHD or even schizophrenia. Knowledge of the true biochemical role of

RA]! awaits the outcome of both in vivo and in vitro experiments, including the physical

187

 

and behavioral assessment of RaiI mutant mice. We have helped to advance the
molecular analysis of SMS from mapping and gene identification to mutation screening

and mice.

188

 

 

 

APPENDIX A

Elsea lab physical and behavioral assessment for transgenic/gene-
targeted mice

Please note that this form usually has entries for mice up to 52 weeks of age, but in the

interest of space, only weeks 5-6, 10-12 and 20-22 are shown.

Name of investigator:
Mouse number:
DOB:

Strain:

Sex:

Coat color:
Genotype:

*Please note any physical or behavioral abnonnalies, even if not specifically

indicated on this form“

Directions for performing most of the physical and behavioral tests are given at the

end of this form

On any given day, please perform the assays requiring anesthesia last!

 

 

 

 

 

 

 

 

 

 

 

 

Physical assessment and simple reflexes 5-6 wks. 10—12 wks. 20—22 wks.

Date assessed

Weight (grams)

Whisker appearance

Skin or fur abnormalities? yes/no yes/no yes/no

If so, where?

Bald patches? yes/no yes/no yes/no

If so, where?

Condition of genital/rectal areas

Condition of nails

Condition of teeth

Cage movement? yes/no yes/no yes/no

Briefly describe

Righting? Did mouse Did mouse Did mouse
right itself? right itself? right itself?

yes/no yes/no yes/no

time in secs: time in secs: time in secs:

 

 

 

 

189

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sound orienting? Right: Right: Right:
yes/no yes/no yes/no
Left: Left: Left:
. yes/no yes/no yes/no
Pupil constriction/dilation? Right eye: Right eye: Right eye:
Constriction? Constriction? Constriction?
yes/no yes/no yes/no
Dilation? Dilation? Dilation?
yedno yes/no yes/no
Left eye: Left eye: Lefi eye:
Constriction? Constriction? Constriction?
yes/no yes/no yes/no
Dilation? Dilation? Dilation?
yes/no yes/no yes/no
Whisker response? Normal/ Normal] Normal/
Normal/abnormal abnormal abnormal abnormal
If abnormal, please comment
Eye blink? Right eye: Right eye: Right eye:
If no, please comment yes/no yes/no yes/no
Lefi eye: Lefi eye: Lefi eye:
yes/no yes/no yes/no
Ear Mitch? Right ear. Right ear. Right car.
If no, please comment yes/no yes/no yes/no
Left ear: Left ear”. Left ear.
yes/no yes/no yes/no
Bodily measurements
Distance from outer ear to outer ear (mm)
Distance from outer eye to outer eye (mm)
Distance from top of head to tip of nose (mm)
Distance from tip of nose to tip of tail (mm)
ngth of trunk from mandible (mm)
Length of limb -front left (m) From toe to From toe to From toe to
knee: knee: knee:
From knee to From knee to From knee to
hip: hip: hip:
Length of limb-front right (m) From toe to From toe to From toe to
knee: knee: knee:
From knee to From knee to From knee to
hip: hip: hip:

 

 

 

 

190

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Length of limb-back left (m) From toe to From toe to From toe
knee: knee: knee:
From knee to From knee to From bee to
hip: hip: hip:
Length of limb-back right.(mm) From toe to From toe to From toe
knee: knee: knee:
From knee to From knee to From knee to
thigh: thigh: thigh:
General behavior-observational assessment
Wild running? yes/no yes/no yes/no
If so, please comment
Freezing? yes/no yes/no yes/no
If so, please comment
Snifiing? yes/no yes/no yes/no
Licking? yes/no Elm yes/no
Rearing? yes/no Jes/no yes/no
Defecation? yes/no yes/no yes/no
Urination? gym yes/no yes/no
Movement around entire cage? yes/no yes/no yes/no
Sensorimotor reflexes and strength
Postural reflex Cage shake Cage shake
Was animal able to stay upright for 10s test? from side to from side to
If no, please comment side: side:
Yes/no Yes/no
Cage shake up Cage shake “P
and down: and down:
Yes/no Yes/no
Response to being picked up by tail Normal] Normal!
Normal/abnormal abnormal abnormal
If abnormal, please comment
Cage top hang test (seconds) Test 1: Test 1:
Test three times (max 60 seconds)
Test 2: Test 2:
Test 3: Test 3:

 

 

 

 

191

 

 

 

 

Gaiting test
Normal/abnormal
Ifunusual gait, please comment

Normal/
abnormal

Normal]
abnormal

 

 

Hot plate test (seconds)

Note time in seconds and appropriate
assessment (using letters) of mouse’s
response:

a jump

b. raise paw

paw lick

paw shake

other (please comment)

sap-.0

Maxtrialtimeis30seconds

Comments:

 

Time:

Response:

 

 

Time:

Response:

 

 

192

Neurological assessment

 

 

Nest-building assessment

At each time point, chose appropriate
assessment (using letters) for each nest (more
than one ean be chosen):

mouse huddled on nestlet

mouse huddled ofl‘ nestlet

mouse not huddled

mouse moving

mouse chewing nestlet

nestlet chewed and flufiy

nestlet chewed and flat

nestlet partially chewed and flat
nestlet partially chewed and flufl’y
other (please comment)

v-r-p-qu nap-99:9

At 120 minutes, measure and record the depth
ofthe thickest part ofthe nest

Comments:

 

 

30 min:

60min:

90min:

120 min:

Depth of nest:

 

193

 

 

 

 

 

Tabetest
Notethemousemnnbersforeachanimalused
in test and which one is dominant/submissive
aswellaswhedtertheanimalsaresiblingsor
iftheyhavebeenhousedinthesamecage

 

Test 1:
numbers:

Dominant:

Submissive:

Animals in
same cage?
yes/no

Test 2:
Animal
numbers:

Dominant:

Submissive:

Animals in
same cage?
yes/no

Test 3:
Animal
numbers:

Dominant:
Submissive:
Animals in

same cage?
yes/no

 

 

Test 1:

numbers:

Submissive:

Animals in
same cage?
yes/no
Test 2:
nmnbers:
Dominant:
Submissive:
Animals in
same cage?
yes/no
Test 3:
numbers:
Dominant:
Submissive:
Animals in

same cage?
yes/no

 

Comments:

194

 

Has this animal displayed normal breeding behavior?
# of litters?

Size of litters?

Comments:

Cause of death?
Euthanasia or other
Date of death:
Age at death:

 

Followingeuthanasia:

 

Skeletal assessment

 

Nasal distance (mm)

 

Distance between anterior notch on frontal processes (mm)

 

Intersection of parietal and interparietal bones (mm)

 

Intersection of interparietal and occipital bones (mm)

 

Bregma (mm)

 

Intersection of maxilla and sphenoid on inferior alveolar
ridge (mm)

 

Gross examination of internal organs

 

 

Abnormalities?

 

195

 

 

Description of physical tests

Righting: turn the mouse onto its back and measure time in seconds
for mouse to right itself; comment if mouse is unable to
right itself or unusual response is noted

Sound orienting: make brief, sharp sound to the right and lefi of mouse; note
if mouse turns head in appropriate direction

Pupil constriction/dilation: with a pen-light, shine a beam of light in the direction of
the mouse’s eye; constriction should occur when light is
shone and dilation should occur when light is removed

Whisker response: lightly brush the whiskers of animal with a small paint
brush and note response (normal mice will stop moving
their whiskers and may turn head) '

Eye blink: approach the eye with the tip of a clean cotton swab and
note if eye blinks (test both eyes)
Ear twitch: touch with car with the tip of a clean cotton swab and note

if ear twitches (test both ears)

Bodily measurements ,

In a lime hood, use ~ 300 uL of isofluorane to anesthetize the mouse and wait until the
animal is fully asleep. Measure the animal according to the diagram and record
measurements in mm. Monitor the animal until he/she wakes up and is functioning
normally.

Sensorimotor and reflexes

Postural reflex: place mouse in empty cage shake from side to side and up
and down for 10 seconds each; note mouse’s ability to
maintain upright position

Response to being picked up: note response when mouse is picked up by tail for 10 s; a
normal response is to raise head, extend extremities and
reach for ground when lowered

Cage top hang test: hang mouse in empty cage with a cage top; measure time in
seconds for mouse to remain hanging and note crawling
movement along cage top; repeat test three times (max 60
seconds)

Gaiting test: gaiting is measured by coloring each mouse’s foot with one
color (using non-toxic materials) and walking mouse
through a small tunnel on white paper

Hot-plate test: place mouse on analgesic hot plate set at 60°C and measure
time in seconds for mouse to display a common response
(jump, raise paw, paw lick or paw shake) or an unusual
response (note and comment); max trial time is 30 secs.

196

Behavioral tests
Nest building assessment:

Tube test:

nesting material is placed in the cage with the mouse and
the quality and depth of the nest over various time periods
(30, 60, 90, and 120 minutes) is assessed (mouse huddled
on nestlet, mouse huddled off nestlet, mouse not huddled,
mouse moving, mouse chewing nestlet, nestlet chewed and
fluffy, nestlet chewed and flat, nestlet partially chewed and
flat, nestlet partially chewed and fluffy); at 120 min., the
thickest part of the nest is measured

two mice (note mouse numbers and whether animals are
siblings as well if they have been housed in same cage) are
placed in 30 cm x 3.5 cm tube and monitored for
dominance (animal pushes the other mouse out of the tube)
and submission. The test is repeated with transgenic/non-
transgenic mice and with mice from different cages.

197

 

APPENDIX B
Preliminary real-time PCR experiments to establish Rail BAC transgenic copy
number

As the genomic Southern hybridization shown in Figure 21 did not produce
unambiguous Rail BAC copy number results, an alternative method to determine BAC
copy number number in our male founder mice as well as the F1 offspring was devised
by the Elsea laboratory utilizing Sybr Green real-time quantitative PCR Our research
plan was to compare the PCR amplification of transgenic animals to non-transgenic
littermates and to determine DNA copy number by measuring the quantitative PCR
results of transgenic animals against a standard curve of known BAC copy number
spiked into mouse genomic DNA, similar to the standard curve in Figure 21. The
fluorescent molecule Sybr Green binds to double-stranded DNA and is provided as part
of a 2x ABI Master Mix solution. The appropriate primers, template DNA and water are
added to this mix before real-time amplification. The ABI 7700 (available through the
MSU GTSF) detects fluorescent incorporation of Sybr Green during the logarithmic
phase of PCR Three sets of primers were designed for our experiments by the ABI
Primer Express program, which should be amplifiable under real-time PCR cycling
conditions: 95°C for 5 minutes, and 40 cycles of 95°C for 15 sec and 60°C for 1 minute.
Three primer sets were utilized to confirm that the copy number results were the same for
each set and to distinguish animals who may not have an intact BAC transgene. The
primers (listed below) were designed to amplify a portion of the mouse Rail gene, as

well as genomic regions near the bc326M22 Sp6 and T7 BAC ends (note that these

198

primers contain unique BAC vector sequences in contrast to the T7 and Sp6 primers
listed in Table 9):

Rail F primer
5 ’-GCCTGAAATCCGACTCAAATAC AT-3 ’

Rail R primer
5’-TGGTGTAAGCATCTCGCTTCT C-3 ’

Genomic region near Sp6 BAC end F primer
5’-GGTAACCTGGAACCGAGACAT C-3’

Genomic region near Sp6 BAC end R primer
5 ’-AAGGAAACAAGCCCCAGAAAC-3 ’

Genomic region near T7 BAC end F primer
5 ’-TCCCACTAAGGGAGCTCTTTATACC-3 ’

Genomic region near T7 BAC end R primer
5’-TCCAAATGTCCTTTAGCCCTTTC-3’

The primers were optimized on the Elsea laboratory ABI standard thermocycler,
using the same cycling conditions as the ABI 7700 real-time thermocycler provided by
the MSU GTSF. BAC 326M22 DNA was used to optimize the primers, as genomic
DNA isolated fi'om mouse tails did not provide consistent amplification. All three of the
primer sets amplified the purified BAC DNA. In order to test the Sp6 primers in the real-
time thermocycler, a standard curve of BAC DNA ranging from 37.9 pg-158 ng of BAC
DNA was serially diluted. The reaction was run with a Sp6 final concentration of 300
nM, which was recommended by Dr. Annette Thelen of the GTSF and all samples were
run in triplicate along with a no template control. The results of the standard curve are
shown in Figure 28 which plots the log of the BAC DNA concentration against the
threshold value (CT). The threshold value is an experimentally determined value
measured during the logarithmic cycle of PCR approximately 10 standard deviations

above a baseline fluorescent value. In the experiment depicted in Figure 28, the threshold

199

Figure 28. BAC 326M22 DNA quantitative PCR standard curve.

A standard curve of 37.9 pg-158 ng of BAC 326M22 DNA was amplified with primers
covering the genomic region near the Sp6 bc326M22 end using the ABI 7700 real-time
PCR thermocycler under standard conditions. For this reaction, the baseline value was
measured from cycles 3 to 8 and threshold values (CT) were measured at 0.534. In the
graph the log quantity is plotted against CT. Two of the replicates did not amplify due to
human error. The line equation for the standard curve y = -2.8143x + 15.972 and the R2
value was 0.9814. This experiment demonstrated that quantitative PCR did work in the
laboratory on purified BAC DNA, though subsequent experiments with mouse genomic
DNA failed.

200

 

 

 

 

990 3353 62> mas—53¢ 256
v. n .2391 3.8»

wa u 0.8:

CT

 

.N , .H o H N w
Eon 055:3.

 

 

prim

alum
due “5
y (W.
duel
r1 mi!
2603

201

was set at 0.534 and the baseline was measured from cycles 3 to 8. Due to human error,
two of the replicates produced no amplification and were not plotted on the graph. The
line equation and the R2 value for the standard curve are displayed on the graph in Figure
28.

The standard curve with BAC demonstrated that the real-time PCR did work with
at least one set of real-time primers designed in the Elsea laboratory and subsequent
experiments with the Rail and T7 PCR sets also shown positive amplification with BAC
DNA at a primer concentration of 300 nM (data not shown). However, none of the
primers sets were able to amplify genomic DNA isolated from mouse tails (using the
method described in Chapter IV Materials and Methods). This protocol for isolation of
DNA fiom tails was appropriate for other PCR reactions, as demonstrated in Figure 20.
DNA from control and transgenic animals was then purified with a standard phenol-
choloform extraction, precipitated with ethanoL and resuspended in TE buffer. This
purified DNA was tested with the real-time primers in the Elsea lab ABI thermocycler
and amplification was noted, though not consistently in all samples. BAC 326M22 DNA
used as a control in these reactions was amplified. When 50 or 100 ng of purified mouse
genomic DNA were tested under real-time cycling conditions in the ABI 7700, there was
either no amplification above baseline or all of the replicates produced different CT
values. There was no consistent amplification that could be analyzed. Possible reasons
for this failure are: the original concentration of the genomic DNA was not correct, a
contaminant such as salt fiom the DNA extraction was inhibiting the PCR reaction, or
perhaps a great deal of error was created within the replicate samples when diluting the

DNA. The genomic DNA was tested twice with the Sp6 BAC end primers and while the

202

BAC DNA controls were amplified, the genomic DNA samples failed. Perhaps a
different method of isolating genomic DNA fiom the mice would provide better results.
We have had variable success in the Elsea laboratory isolating usable DNA fi'om mouse
whole blood, probably due to blood collection tubes containing heparin, which can inhibit
several of the standard molecular procedures that are routinely performed in the lab such
as PCR and restriction digests. In the firture, if a viable blood collection method is
devised without heparin, genomic DNA extracted from mouse blood may provide better

real-time PCR amplification and resolve the integrated Rail BAC copy number.

203

 

BIBLIOGRAPHY

Aasland, R., Gibson, T. J. and Stewart, A. F. (1995). "The PHD finger: implications for
chromatin-mediated transcriptional regulation." Trends Biochem Sci 20(2): 56-9.

Ackerman, S. H. (2002). "Atpl 1p and Atp12p are chaperones for F (l)-ATPase
biogenesis in mitochondria." Biochim Biophys Acta 1555(1-3): 101-5.

Alagille, D., Odievre, M., Gautier, M. and Dommergues, J. P. (1975). "Hepatic ductular
hypoplasia associated with characteristic facies, vertebral malformations, retarded
physical, mental, and sexual development, and cardiac murmur." [Mag 86(1):
63-71.

Alavizadeh, A., Kieman, A. E., et al. (2001). "The Wheels mutation in the mouse causes
vascular, hindbrain, and inner ear defects." Dev Biol 234(1): 244-60.

Bi, W., Park, S. S., et a]. (2003). "Reciprocal crossovers and a positional preference for
strand exchange in recombination events resulting in deletion or duplication of
chromosome 17p11.2." Am J Hum Genet 73(6): 1302-15.

Bi, W., Yan, J., et al. (2002). "Genes in a refined Smith-Magenis syndrome critical
deletion interval on chromosome Up] 1.2 and the syntenic region of the mouse."

Genome Res 12(5): 713-28.

Butler, M. G. (2002). "Imprinting disorders: non-Mendelian mechanisms affecting
growth." 1 Pediatr Endocrinol Metab 15 Suppl 5: 1279-88.

Campbell, H. D., Fountain, S., et al. (2002). "Fliih, a gelsolin-related cytoskeletal
regulator essential for early mammalian embryonic development." Mol Cell Biol
22(10): 3518-26.

Campbell, H. D., Fountain, S., et al. (1997). "Genomic structure, evolution, and
expression of human FLII, a gelsolin and leucine-rich-repeat family member:
overlap with LLGL." Genomics 42(1): 46-54.

Campbell, H. D., Fountain, S., et a1. (2000). "Fliih, the murine homologue of the
Drosophila melanogaster flightless I gene: nucleotide sequence, chromosomal
mapping and overlap with nglh." DNA SQ 11(1-2): 29-40.

Cassidy, S. B., Dykens, E. and Williams, C. A. (2000). "Prader-Willi and Angelman
syndromes: sister imprinted disorders." Am J Med Genet 97(2): 136-46.

Chen, K. S., Gunaratne, P. H., et al. (1995). "The human homologue of the Drosophila

melanogaster flightless-I gene (flil) maps within the Smith-Magenis microdeletion
critical region in 17p11.2." Am J Hum Genet 56(1): 175-82.

204

Chen, K. S., Manian, P., et al. (1997). "Homologous recombination of a flanking repeat
gene cluster is a mechanism for a common contiguous gene deletion syndrome."

M 17(2): 154-63.

Cismowski, M. J ., Ma, C., et al. (2000). "Activation of heterotrimeric G-protein signaling
by a ras-related protein. Implications for signal integration." ,1 Biol Ghem 275(31):
23421-4.

Cismowski, M. J ., Takesono, A., et al. (1999). "Genetic screens in yeast to identify
mammalian nonreceptor modulators of G-protein signaling." Nat Biotghngl
‘ 17(9): 878-83.

Crawley, J. N. (2000). What's wron wi mouse?: Behavioral heno

Eugenie and knockout mice, John Wiley and Sons, Inc.

De Leersnyder, H., De Blois, M. C., et al. (2001). "Inversion of the circadian rhythm of
melatonin in the Smith-Magenis syndrome." J Pfliatr 139(1): 111-6.

De Leersnyder, H., de Blois, M. C., et al. (2001). "beta(l)-adrenergic antagonists improve
sleep and behavioural disturbances in a circadian disorder, Smith-Magenis

syndrome." 1 Mg Cmet 38(9): 586-90.

DeSilva, U., Massa, H., Trask, B. J. and Green, E. D. (1999). "Comparative mapping of
the region of human chromosome 7 deleted in Williams syndrome." Genome Res
9(5): 428-36.

Elsea, S. H., Fritz, E., Schoener-Scott, R., Meyn, M. S. and Patel, P. I. (1998). "Gene for
topoisomerase HI maps within the Smith-Magenis syndrome critical region:
analysis of cell-cycle distribution and radiation sensitivity." Am J Med Genet
75(1): 104-8.

Elsea, S. H., Juyal, R. C., et al. (1995). "Haploinsufficiency of cytosolic serine
hydroxymethyltransferase in the Smith-Magenis syndrome." Am J Hum Genet
57(6): 1342-50.

Elsea, S. H. and Lucas, R. E. (2002). "The mousetrap: what we can learn when the mouse

model does not mimic the human disease." ILARJ 43(2): 66-79.

Elsea, S. H., Mykytyn, K., et a1. (1999). "Hemizygosity for the COP9 signalosome
subunit gene, SGN3, in the Smith- Magenis syndrome." Am J Med Genet 87(4):
342-8.

Elsea, S. H., Purandare, S. M., et al. (1997). "Definition of the critical interval for Smith-
Magenis syndrome." Gflogenet Cell Genet 79(3-4): 276-81.

205

 

Fang, M., Jaffrey, S. R., et al. (2000). "Dexrasl: a G protein specifically coupled to
neuronal nitric oxide synthase via CAPO ." Neurog 28(1): 183-93. '

Farrell, M. J., Stadt, H., et al. (1999). "HIRA, a DiGeorge syndrome candidate gene, is
required for cardiac outflow tract septation." Girc Res 84(2): '1 27-35.

Finucane, B., Dirrigl, K. H. and Simon, E. W. (2001). "Characterization of self-injurious
behaviors in children and adults with Smith-Magenis syndrome." & J Ment
Retard 106(1): 52-8. ‘

Francke, U. (1999). "Williams-Beuren syndrome: genes and mechanisms." Hum Mol
(get 8(10): 1947-54.

Franke, Y., Peoples, R. J. and Francke, U. (1999). "Identification of GTFZIRDI, a
putative transcription factor within the Williams-Beuren syndrome deletion at

7q11.23." Gflogenet Gel! Genet 86(3-4): 296-304.

Fritz, E., Elsea, S. H., Patel, P. I. and Meyn, M. S. (1997). "Overexpression of a tnmcated
human topoisomerase III partially corrects multiple aspects of the ataxia-

telangiectasia phenotype." me Natl Acfl Sci [_J S A 94(9): 4538-42.

Funke, B., Epstein, J. A., et al. (2001). "Mice overexpressing genes fi'om the 22qll
region deleted in velo-cardio-facial syndrome/DiGeorge syndrome have middle
and inner ear defects." Here Mol Geget 10(22): 2549-56.

Gilbert, S. (2000). Developmental Biology, Sinauer Associates, Inc.

Greenberg, F ., Guzzetta, V., et a1. (1991). "Molecular analysis of the Smith-Magenis
syndrome: a possible contiguous- gene syndrome associated with del(17)(p11.2)."

Am] Hm Genet 49(6): 1207-18.

Greenberg, F., Lewis, R. A., et a1. (1996). "Multi-disciplinary clinical study of Smith-
Magenis syndrome (deletion 17p11.2)." Am 1 Med Genet 62(3): 247-54.

Gridley, T. (2003). "Notch signaling and inherited disease syndromes." Hum Mel Genet
12 Spec No 1: R9-13.

Guris, D. L., Fantes, J., Tara, D., Druker, B. J. and Irnamoto, A. (2001). "Mice lacking
the homologue of the human 22q11.2 gene CRKL phenocopy neurocristopathies
of DiGeorge syndrome." Nat Genet 27(3): 293-8.

Guzzetta, V., Franco, B., et al. (1992). "Somatic cell hybrids, sequence-tagged sites,

simple repeat polymorphisms, and yeast artificial chromosomes for physical and
genetic mapping of proximal 17p." Genomics 13(3): 551-9.

206

Hanai, R., Caron, P. R. and Wang, J. C. (1996). "Human TOP3: a single-copy gene
encoding DNA t0poisomerase 111." m Net] Acfl Sci [J S A 93(8): 3653-7.

Hayes, 8., Turecki, G., et a1. (2000). "CAG repeat length in RAM is associated with age
at onset variability in spinocerebellar ataxia type 2 (SCA2)." Him; Mol Cget
9(12): 1753-8. '

Heintz, N. (2001). "BAC to the future: the use of bac transgenic mice‘for neuroscience
research." NeLReLNeereeej 2(12): 861-70.

Hockenhull, E. L., Carette, M. J ., et al. (1999). "A complete physical contig and partial
transcript map of the Williams syndrome critical region." ngomics 58(2): 138-
45.

Hoogenraad, C. C., Eussen, B. H., et al. (1998). "The murine CYLN2 gene: genomic
organization, chromosome localization; and comparison to the human gene that is
located within the 7q11.23 Williams syndrome critical region." Genomics 53(3):
348-58.

Hoogenraad, C. C., Koekkoek, B., et al. (2002). "Targeted mutation of Cyln2 in the
Williams syndrome critical region links CLIP-115 haploinsufficiency to
neurodevelopmental abnormalities in mice." Net Genet 32(1): 116-27.

Irnai, Y., Suzuki, Y., et al. (1995). "Cloning of a retinoic acid-induced gene, GT1, in the
embryonal carcinoma cell line P19: neuron-specific expression in the mouse
brain." Brain Res Mol Brain Res 31(1-2): 1-9.

Ishikawa, K., Azuma, S., et al. (2003). "Cloning and characterization of Xenopus laevis
drg2, a member of the developmentally regulated GTP-binding protein
subfamily." Gene 322: 105-12.

Jerome, L. A. and Papaioannou, V. E. (2001). "DiGeorge syndrome phenotype in mice
mutant for the T-box gene, Tbxl." Nat Genet 27(3): 286-91.

J i, Y., Eichler, E. E., Schwartz, S. and Nicholls, R. D. (2000). "Structure of chromosomal
duplicons and their role in mediating human genomic disorders." Genome Res
10(5): 597-610.

Joober, R., Benkelfat, C., et al. (1999). "Analysis of 14 CAG repeat-containing genes in
schizophrenia." Am ,1 Med Genet 88(6): 694-9.

Juyal, R. C., F iguera, L. E., et al. (1996). "Molecular analyses of 17p11.2 deletions in 62
Smith-Magenis syndrome patients." Am J Hum Genet 58(5): 998-1007.

207

 

Kallioniemi, O. P., Kallioniemi, A., et a1. (1994). "Physical mapping of chromosome 17
cosmids by fluorescence in situ hybridization and digital image analysis."
55% 20(1): 125-8.

Kao,_F., Jones, C. and Puck, T. T. (1976). "Genetics of somatic mammalian cells:
genetic, immunologic, and biochemical analysis with Chinese hamster cell
hybrids containing selected human chromosomes." PNAS 73: 193-97.

Kapfhamer, D., Valladares, O., et al. (2002). "Mutations in Rab3a alter circadian period
and homeostatic response to sleep loss in the mouse." Nat Geeet 32(2): 290-5.

Karmiloff-Smith, A., Grant, J ., et a1. (2003). "Using case study comparisons to explore
genotype-phenotype correlations in Williams-Beuren syndrome." 1 Mg fleet
. 40(2): 136-40.

Kemppainen, R. J. and Behrend, E. N. (1998). "Dexamethasone rapidly induces a novel
ras superfanrily member-related gene in AtT-20 cells." 1 Biol Ghem 273(6): 3129-
31.

Kemppainen, R. J ., Cox, E., Behrend, E. N., Brogan, M. D. and Ammons, J. M. (2003).
"Identification of a glucocorticoid response element in the 3'-flanking region of
the human Dexrasl gene." Biocht'm Bioehys Acta 1627(2-3): 85-9.

Khan, N. L. and Wood, N. W. (1999). "Prader-Willi and Angehnan syndromes: update
on genetic mechanisms and diagnostic complexities." Gurr min Neurol 12(2):
149-54. ~

Kieman, A. E., Ahituv, N., et al. (2001). "The Notch ligand Jaggedl is required for inner
ear sensory development." Prg Netl Aeag Sci [J S A 98(7): 3873-8.

Kishino, T., Lalande, M. and Wagstaff, J. (1997). "UBE3A/E6-AP mutations cause
Angehnan syndrome." Nat Genet 15(1): 70-3.

Krantz, I. D., Piccoli, D. A. and Spinner, N. B. (1997). "Alagille syndrome." J Med Genet
34(2): 152-7. '

Krantz, I. D., Smith, R., et al. (1999). "Jaggedl mutations in patients ascertained with
isolated congenital heart defects." Am J Med Genet 84(1): 56-60.

Kuklin, A., Davis, A. P., Hecker, K. H., Gjerde, D. T. and Taylor, P. D. (1999). "A novel
technique for rapid automated genotyping of DNA polymorphisms in the mouse."

Mol Gell Probes 13(3): 239-42.
Kuklin, A., Munson, K., Gjerde, D., Haefele, R. and Taylor, P. (1997). "Detection of

single-nucleotide polymorphisms with the WAVE DNA fragment analysis
system." Genet Test 1(3): 201-6.

208

Lamont, V., Lecluse, Y., et al. (1995). "A human homolog of the S. cerevisiae HIRl and
HIR2 transcriptional repressors cloned from the DiGeorge syndrome critical

region." Helm Mel Genet 4(5): 791-9.

Le Caignec, C., Lefevre, M., et al. (2002). "Familial deafness, congenital heart defects,
and posterior embryotoxon caused by cysteine substitution in the first epidermal-
growth-factor-like domain of jagged l." AmiflmGenet 71(1): 180-6.

Lee, S. H., Fu, K. K., Hui, J. N. and Richman, J. M. (2001). "Noggin and retinoic acid
‘ transform the identity of avian facial prominences." Nature 414(6866): 909-12.

Lemckert, F. A., Sedgwick, J. D. and Korner, H. (1997). "Gene targeting in C57BL/6 ES
cells. Successful germ line transmission using recipient BALB/c blastocysts
developmentally matured in vitro." Nueleic Acids Res 25(4): 917-8.

Li, D. Y., Brooke, B., et a1. (1998). "Elastin is an essential determinant of arterial
morphogenesis." Nature 393(6682): 276-80.

Li, D. Y., Faury, G., et a1. (1998). "Novel arterial pathology in mice and humans
hemizygous for elastin." ,1 Glie Invest 102(10): 1783-7.

Li, L., Krantz, I. D., et al. (1997). "Alagille syndrome is caused by mutations in human
J aggedl, which encodes a ligand for Notchl [see comments]." Nat Genet 16(3):
243-51.

Li, Q., Liu, Z., Monroe, H. and Culiat, C. T. (2002). "Integrated platform for detection of
DNA sequence variants using capillary array electrophoresis." Eleegqphoresie
23(10): 1499-511.

Liang, S. G., Sadovnick, A. D., et al. (2002). "A linkage disequilibrium study of bipolar
disorder and microsatellite markers on 22q13." Beychiatt Genet 12(4): 231-5.

Liang, Y., Wang, A., et al. (1999). "Characterization of the Human and Mouse
Unconventional Myosin XV Genes Responsible for Hereditary Deafness DFNB3
and Shaker 2." Genomics 61(3): 243-58.

Liang, Y., Wang, A., et al. (1999). "Characterization of the human and mouse
unconventional myosin XV genes responsible for hereditary deafness DFNB3 and
shaker 2." Gem 61(3): 243-58.

Liburd, N., Ghosh, M., et al. (2001). "Novel mutations of MY 015A associated with
profound deafness in consanguineous families and moderately severe hearing loss
in a patient with Smith-Magenis syndrome." Hum Genet 109(5): 535-41.

209

 

Lindsay, E. A. (2001). "Chromosomal microdeletions: dissecting de122qll syndrome."
W 2(1 1): 853-68-

Lindsay, E. A. and Baldini, A. (1998). "Congenital heart defects and 22q11 deletions:
which genes count?" Mol Med Today 4(8): 350-7.

Lindsay, E. A., Vitelli, F., et al. (2001). "Tbxl haploinsufficieny in the DiGeorge
syndrome region causes aortic arch defects in mice." Ngge 410(6824): 97-101.

Lohi, O. and Lehto, V. P. (1998). "VHS domain marks a group of proteins involved in
endocytosis and vesicular trafficking." F EBS Lett 440(3): 25 5-7.

Lohi, O., Poussu, A., Mac, Y., Quiocho, F. and Lehto, V. P. (2002). "VHS domain -- a
longshoreman of vesicle lines." F BBS Lett 513(1): 19-23.

Lucas, R. E., Vlangos, C. N., Das, R, Patel, P. I. and Elsea, S. H. (2001). "Genomic
organisation of the approximately 1.5 Mb Smith-Magenis syndrome critical
interval: transcription map, genomic contig, and candidate gene analysis." Eur J

Hum Geeet 9(12): 892-902.

Lyngso, C., Bouteiller, G., et al. (2000). "Interaction between the transcription factor
SPBP and the positive cofactor RNF4. An interplay between protein binding zinc
fingers." ,1 Biol Gleam 275(34): 26144-9.

Maden, M. (2002). "Retinoid signalling in the development of the central nervous
system." Nat Rev Neuroeci 3: 843-53.

Maden, M. and Hind, M. (2003). "Retinoic acid, a regeneration-inducing molecule." _D_ey
_Dle 226(2): 237-44.

Mann, M. R. and Bartolomei, M. S. (1999). "Towards a molecular understanding of
Prader-Willi and Angelman syndromes." Hum Mel Genet 8(10): 1867-73.

Matsuura, T ., Sutcliffe, J. S., et a1. (1997). "De novo truncating mutations in E6-AP
ubiquitin-protein ligase gene (UBE3A) in Angehnan syndrome." Nat Genet 15(1):
74-7.

McCright, B., Gao, X., et al. (2001). "Defects in development of the kidney, heart and
eye vasculature in mice homozygous for a hypomorphic Notch2 mutation."

Develeemeet 128(4): 491-502.

Merscher, S., Funke, B., et al. (2001). "TBXl is responsible for cardiovascular defects in
velo-cardio- facial/DiGeorge syndrome." _(fl 104(4): 619-29.

Nolan, P. M., Kapfl'rarner, D. and Bucan, M. (1997). "Random mutagenesis screen for
dominant behavioral mutations in mice." Methods 13(4): 379-95.

210

Oda, T., Elkahloun, A. G., Meltzer, P. S. and Chandrasekharappa, S. C. (1997).
"Identification and cloning of the human homolog (JAGl) of the rat J aggedl gene
from the Alagille syndrome critical region at 20p12." Cmomics 43(3): 376-9.

Oda, T., Elkahloun, A. G., et a1. (1997). "Mutations in the human Jaggedl gene are
responsible for Alagille syndrome [see comments]." Nat Gget 16(3): 235-42.

Ogdie, M. N., Macphie, I. L., et al. (2003). "A genomewide scan for attention-
deficit/hyperactivity disorder in an extended sample: suggestive linkage on
17p11." Am 1 HQ; Genet 72(5): 1268-79.

Ohta, T., Buiting, K., et al. (1999). "Molecular mechanism of angelrnan syndrome in two
large families involves an imprinting mutation." Am 1 Hum Genet 64(2): 385-96.

Osborne, L. R. (1999). "Williams-Beuren syndrome: unraveling the mysteries of a
microdeletion disorder." Mel fleet Metee 67(1): 1-10.

Paglia, D. E., Valentine, W. N. and Brockway, R. A. (1984). "Identification of thymidine
nucleotidase and deoxyribonucleotidase activities among normal isozymes of 5'-

nucleotidase in human erythrocytes." Proc Natl Acfl Sci G S A 81(2): 588-92.

Park, S. S., Stankiewicz, P., et al. (2002). "Structure and evolution of the Smith-Magenis
syndrome repeat gene clusters, SMS-REPS." Genome Res 12(5): 729-38.

Pascual, J., Martinez-Yarnout, M., Dyson, H. J. and Wright, P. E. (2000). "Structure of
the PHI) zinc finger from human Williams-Beuren syndrome transcription
factor." 1 Mel Biol 304(5): 723-9.

Paylor, R., McIlwain, K. L., et al. (2001). "Mice deleted for the
DiGeorge/velocardiofacial syndrome region show abnormal sensorimotor gating
and learning and memory impairments." Hum Mol Genet 10(23): 2645-50.

Peoples, R., Franke, Y., et al. (2000). "A physical map, including a BAC/PAC clone
contig, of the Williarns- Beuren syndrome-deletion region at 7q11.23." m
Bug Genet 66(1): 47-68.

Peoples, R. J., Cisco, M. J., Kaplan, P. and Francke, U. (1998). "Identification of the
WBSCR9 gene, encoding a novel transcriptional regulator, in the Williams-
Beuren syndrome deletion at 7q11.23." Gytogenet Gell Genet 82(3-4): 238-46.

Perez Jurado, L. A., Peoples, R., Kaplan, P., Hamel, B. C. and Francke, U. (1996).

"Molecular definition of the chromosome 7 deletion in Williams syndrome and
parent-of-origin effects on growth." Am J Hum Genet 59(4): 781-92.

211

 

Petrij, F., Giles, R. H., et al. (1995). "Rubinstein-Taybi syndrome caused by mutations in
the transcriptional co-activator CBP." Name 376(653 8): 348-51.

Piccoli, D. A and Spinner, N. B. (2001). "Alagille syndrome and the Jaggedl gene."
mamas 21(4): 525-34-

Pickard, G. 13., Sollars, P. J., Rinchik, E. M., Nolan, P. M. and Bucan, M. (1995).
"Mutagenesis and behavioral screening for altered circadian activity identifies the
mouse mutant, Wheels." m 705(1-2): 255-66.

Pickova, A., Paul, J ., Petruzzella, V. and Houstek, J. (2003). "Differential expression of
ATPAF 1 and ATPAF 2 genes encoding F(1)-ATPase assembly proteins in mouse
tissues." FBBS Lett 551(1-3): 42-6.

Pizzuti, A., Novelli, G., et al. (1997). "UFDIL, a developmentally expressed
ubiquitination gene, is deleted in CATCH 22 syndrome." Hum Mel Gflet 6(2):
259-65.

Pollet, N., Dhome-Pollet, S., et al. (1995). "Construction of a 3.7-Mb physical map
within human chromosome 20p12 ordering 18 markers in the Alagille syndrome
locus." finemics 27(3): 467-74.

Potocki, L., Chen, K. S. and Lupski, J. R. (1999). "Subunit 3 of the COP9 signal
transduction complex is conserved from plants to humans and maps within the
smith-magenis syndrome critical region in 17p11.2." Genemice 57(1): 180-2.

Potocki, L., Chen, K. S., et al. (2000). "Molecular mechanism for duplication l7pll.2-
the homologous recombination reciprocal of the Smith-Magenis microdeletion."

Nat Genet 24(1): 84-7.

Potocki, L., Glaze, D., et a1. (2000). "Circadian rhythm abnormalities of melatonin in
Smith-Magenis syndrome." ,1 Med Genet 37(6): 428-33.

Probst, F. J ., Fridell, R. A., et al. (1998). "Correction of deafitess in shaker-2 rrrice by an
unconventional myosin in a BAC transgene." Science 280(5368): 1444-7.

Rampazzo, C., Gallinaro, L., et al. (2000). "A deoxyribonucleotidase in mitochondria:
involvement in regulation of dNT P pools and possible link to genetic disease."

Btee Netl Acae Sci 11 s A 97(15): 8239-44.

Rekdal, C., Sjottem, E. and Johansen, T. (2000). "The nuclear factor SPBP contains
different fimctional domains and stimulates the activity of various transcriptional
activators." 1 Biol Ghem 275(51): 40288-300.

212

Roberts, C., Daw, S. C., Halford, S. and Scambler, P. J. (1997). "Cloning and
developmental expression analysis of chick Hira (Chira), a candidate gene for
DiGeorge syndrome." Burn Mol Genet 6(2): 237-45.

Rougeulle, C., Glatt, H. and Lalande, M. (1997). "The Angelman syndrome candidate
gene, UBE3A/E6-AP, is imprinted in brain." Nat Genet 17(1): 14-5.

Sazuka, T., Kinoshita, M., et a1. (1992). "Expression of DRG during murine embryonic
development." Biechem Bioehys Res Gemfl 189(1): 371-7.

Sazuka, T., Tomooka, Y., Ikawa, Y., Noda, M. and Kumar, S. (1992). "DRG: a novel
developmentally regulated GTP-binding protein." Biochem Bioehys Res

Cemetrte 189(1): 363-70.

Schedl, A., Larin, Z., et al. (1993). "A method for the generation of YAC transgenic mice
by pronuclear microinjection." NeeLeteAejeeBee 21(20): 4783-7.

Schenker, T., Lach, C., Kessler, B., Calderara, S. and Trueb, B. (1994). "A novel GTP-
binding protein which is selectively repressed in SV4O transformed fibroblasts." ,1
Biel Ghem 269(41): 25447-53.

Schmickel, R. D. (1986). "Contiguous gene syndromes: a component of recognizable
syndromes." ,1 Pediag 109: 231-41.

Schuster-Gossler, K., Lee, A. W., et a1. (2001). "Use of coisogenic host blastocysts for
efficient establishment of germline chimeras with C57BL/6J ES cell lines."

Biotechnigees 31(5): 1022-4, 26.

Seranski, P., Hoff, C., et a1. (2001). "RAIl is a novel polyglutamine encoding gene that is
deleted in Smith-Magenis syndrome patients." Gene 270(1-2): 69-76.

Shapira, S. K. (1998). "An update on chromosome deletion and microdeletion
syndromes." Gurr Gein Bediatr 10(6): 622-7.

Shiba, Y., Katoh, Y., et al. (2003). "GAT (GGA and Toml) domain responsible for
ubiquitin binding and ubiquitination." Meg.

Shimano, H., Shimomura, 1., et a1. (1997). "Elevated levels of SREBP-2 and cholesterol
synthesis in livers of mice homozygous for a targeted disruption of the SREBP-l

gene." ,1 Glin levest 100(8): 2115-24.

Slager, R. E., Newton, T. L., Vlangos, C. N., Finucane, B. and Elsea, S. H. (2003).
"Mutations in RAIl associated with Smith-Magenis syndrome." Nat Genet 33(4):
466-8.

213

 

Smith, A. C., Dykens, E. and Greenberg, F. (1998). "Behavioral phenotype of Smith-
' Magenis syndrome (del 17p11.2)." Am I1 Med Genet 81(2): 179-85.

Smith, A. C., McGavran, L., et al. (1986). "Interstitial deletion of (l7)(p11.2p11.2) in
nine patients." Am ,1 MM Genet 24(3): 393-414.

Smith, A. C. M., L., M., Waldstein, G. and Robinson, J. (1982). "Deletion of the 17 short
arm in two patients with facial clefts and congenital heart disease." M
m 34: A410:

Smith, J. R., Osborne, T. F., Goldstein, J. L. and Brown, M. S. (1990). "Identification of
nucleotides responsible for enhancer activity of sterol regulatory element in low
density lipoprotein receptor gene." ,1 Biol Ghem 265(4): 2306-10.

Sonuga—Barke, E. J. (2003). "The dual pathway model of AD/HD: an elaboration of
neuro-developmental characteristics." Nemsei Biobehav Rev 27(7): 593-604.

Spinner, N. B., Colliton, R. P., et a1. (2001). "Jaggedl mutations in alagille syndrome."
Hum Mutat 17(1): 18-33.

Stankiewicz, P. and Lupski, J. R. (2002). "Molecular-evolutionary mechanisms for
genomic disorders." Gurr Gein Genet Dev 12(3): 312-9.

Stankiewicz, P., Shaw, C. J., et a1. (2003). "Genome architecture catalyzes nonrecurrent
chromosomal rearrangements. "Am Am,1Hum Genet 72(5): 1101- 16.

Strand, D., Unger, S., et al. (1995). "A human homologue of the Drosophila tumour
suppressor gene l(2)gl maps to 17p11.2-12 and codes for a cytoskeletal protein
that associates with nonmuscle myosin H heavy chain." Gecogene 11(2): 291-
301.

Stratton, R. F., Dobyns, W. B., et al. (1986). "Interstitial deletion of (l7)(p11.2p11.2):
report of six additional patients with a new chromosome deletion syndrome." _A_m

,1 Meg Genet 24(3): 421-32.

Surks, H. K., Richards, C. T. and Mendelsohn, M. E. (2003). "Myosin phosphatase-Rho
interacting protein. A new member of the myosin phosphatase complex that
directly binds RhoA." ,1 Biol Chem 278(51): 51484-93.

Taddei, 1., Morishima, M., Huynh, T. and Lindsay, E. A. (2001). "Genetic factors are
major determinants of phenotypic variability in a mouse model of the
DiGeorge/de122q1'l syndromes." Proc Netl Acad Sci U S A 98(20): 11428-31.

Takahashi, H. Umeda, N., et a1. (2003). "Mouse dexarnethasone-induced RAS protein 1
gene is expressed m a circadian rhythmic manner in the suprachiasmatic nucleus"

Brein Ree Mol Brain Res 110(1): 1 -6.

214

 

   

l11111111311111le116111111l