Interruption of a metabotropic glutamate receptor gene in a Smith-Lemli-Opitz syndrome patient

INTERRUPTION OF A METABOTROPIC GLUTAMATE RECEPTOR GENE IN
A SMITH-LEMLI-OPITZ SYNDROME PATIENT












BY

TIFFANY LEIGH ALLEY


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY



UNIVERSITY OF FLORIDA


1997


















ACKNOWLEDGMENTS


The last-four-and-a-half years have presented me with the most dramatic

changes in my life both academically and personally. I would like to take this

opportunity to personally thank all of the individuals who made this journey a

bit more tolerable and sometimes downright fun.

I would like to begin by thanking the members of my committee who held

on tight while my project took many unexpected turns: Drs. Margaret R.

Wallace, Thomas P. Yang, Harry S. Nick, Daniel L. Purich, Charles M. Allen, and

Roberto T. Zori. Each member contributed a unique perspective to a very

unpredictable project. Their dedication to my graduate success was greatly

appreciated.

I am deeply indebted to Dr. Wallace who served in the capacity of a true

mentor. She provided an academic environment that nurtured my scientific and

personal growth; for this I am grateful. She also showed me that motherhood

and career are not mutually exclusive and the only true sacrifice is that of sleep.

Many thanks go to the past and present members of the Wallace

laboratory who gave both technical and emotional support: Dr. Steve Colman,

Dr. Sonja Rasmussen, Jun Zhang, Jennifer Johnson, Reena Kamath, Sophia

Krkljus, Vu Ho, Corinne Abernathy, and Rachael Trimpert. I'd like to extend

special thanks to Dr. Steve Colman with whom I developed a friendship that I

hope will last a lifetime. Steve was not only my friend, but also my confidant,

my emotional punching bag, and sometimes my sanity. I'll never forget the

inane bantering or the laughs.











Very few individuals have a lab away from lab, but in this respect I was

fortunate. The members of the Nick lab have for years been my surrogate family.

I thank them all for the warmth and love they have shown me through the years.

Much gratitude goes to my surrogate parents, Dr. Harry Nick and Joan Monnier,

for caring about me as if I was their own. And a very special thanks to Dr. Nick

for believing in me even when I didn't believe in myself.

Several laboratories directly contributed to my thesis project and deserve

to be.acknowledged. First, I'd like to acknowledge the laboratory of Drs. Lap-

Chee Tsui and Steve Scherer. Their expertise and gracious contributions to my

project forged its success. I would also like to thank the members of the R.C.

Philips Unit, especially Dr. Roberto Zori and Brian Gray, for their continued

technical support.

On a more personal note, I want to thank my family for their unwavering

support of my educational pursuits (and my other whims). My family's love for

me is only paralleled by my love for them.

Lastly, I would like to acknowledge three individuals who have been

essential in my life. Catherine Golden has given me a friendship that knows no

boundaries. I appreciated her encouragement and love during my graduate

years. I thank Toren Anderson for showing me what love can truly be. He was

the driving force behind the completion of my thesis for it enabled us to be

reunited. And finally I thank my guardian angel, my mother. It is to her that I

dedicate this work, for all her unselfish acts that made this moment possible.



















TABLE OF CONTENTS

page

ACKNOW LEDGMENTS ......................................................................................... ii

A B ST R A C T ............................................................................................................... v i

CHAPTERS

1 IN T R O D U C T IO N .................................................................................................. 1

Clinical Description of Smith-Lemli-Opitz Syndrome (UF53) .................... 1
Discovery of a Biochemical Defect in Cholesterol Synthesis ....................... 6
Case Report of a Severely Affected Smith-Lemli-Opitz Patient ................. 10
Positional Cloning Effort .................................................................................. 14

2 IDENTIFICATION OF A YEAST ARTIFICIAL CHROMOSOME
SPANNING THE TRANSLOCATION BREAKPOINT IN UF53 ....... 17

In tro d u ctio n ......................................................................................................... 17
Materials and Methods ....................................................................................... 21
R e su lts .................................................................................................................... 2 3
D iscu ssio n .............................................................................................................. 28

3 IDENTIFICATION OF A METABOTROPIC GLUTAMATE RECEPTOR
GENE (GMR8) DIRECTLY DISRUPTED BY UF53'S
TRANSLOCATION .................................................................................... 30

In tro d u ctio n ......................................................................................................... 30
Materials and Methods ....................................................................................... 35
R esu lts .................................................................................................................... 3 8
D iscu ssio n .............................................................................................................. 4 4

4 PHYSICAL MAPPING OF THE TRANSLOCATION BREAKPOINT
R E G IO N ..................................................................................................... .. 50

In tro d u ctio n ........................................................................................................... 5 0
Materials and Methods ....................................................................................... 56
R e Sul ts ..................................................................................................................... 5 8
D iscu ssio n ............................................................................................................... 5 9












5 MICROSATELLITE TYPING AND SOUTHERN BLOT ANALYSIS ......... 67

Introduction ........................................................................................................... 67
M aterials and M ethods ....................................................................................... 72
Results ..................................................................................................................... 74
D iscussion ............................................................................................................... 77

6 CONCLUSIONS AND FUTURE DIRECTIONS ........................................... 80

Conclusions ......................................................................................................... 80
Future directions ................................................................................................ 83


REFEREN C ES ............................................................................................................. 92


BIO G RA PH ICA L SKETCH ..................................................................................... 105


















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


INTERRUPTION OF A METABOTROPIC GLUTAMATE RECEPTOR GENE IN
A SMITH-LEMLI-OPITZ SYNDROME PATIENT

By

Tiffany Leigh Alley

May, 1997



Chairman: Dr. Margaret R. Wallace
Co-Chairman: Dr. Thomas P. Yang
Major Department: Biochemistry and Molecular Biology


Smith-Lemli-Opitz syndrome (SLOS) is an autosomal recessive disorder

characterized by facial and limb abnormalities, internal malformations and

mental retardation. SLOS patients have an apparent biochemical defect in

cholesterol metabolism which results in decreased cholesterol concentrations and

a significant increase in 7-dehydrocholesterol (7-DHC), the immediate precursor

to cholesterol. The enzyme proposed to be defective in SLOS is 7-DHC reductase

(7-DHCR), which converts 7-DHC to cholesterol. The lack of nucleic acid and

protein sequence for mammalian 7-DHCR has prevented a definitive link

between 7-DHCR and SLOS.
Furthermore, since no SLOS gene(s) have been identified, a positional

cloning approach was undertaken to identify SLOS candidate gene(s). An SLOS












patient (UF53) with a de novo balanced translocation in 7q32.1 led to the

hypothesis that one SLOS allele was interrupted by the translocation while the

other allele was inactivated by a more subtle mutation.

Using fluorescence in situ hybridization (FISH) with yeast artificial

chromosomes (YACs) from a contig of the 7q32 translocation region, one YAC,

HSC7E1289, was identified, which spanned UF53's translocation, and other

YACs localized proximal and distal to the breakpoint. The contig contains two

CpG islands: one is located at the distal end of the contig while the other lies at

the distal end of HSC7E1289. A partial cDNA of a metabotropic glutamate

receptor (GRM8) mapped to the proximal end of HSC7E1289. Southern blot

analysis of the mouse homolog of GRM8 suggested that the GRM8 gene might be

interrupted by UF53's translocation. FISH analysis of cosmids containing GRM8

exons identified two cosmids, 85b12 and 153a8, which fell directly proximal and

distal to the translocation, respectively, implying that GRM8 was interrupted by

the translocation. PCR analysis demonstrated.that GRM8 exon 1 was contained

in 153a8 while exon 2 lay in 85b12, indicating that the translocation interrupts

GRM8 in intron 1. This was unexpected since there is no known relationship

between GRM8 and cholesterol biosynthesis. However, GRM8 is expressed

throughout the central and peripheral nervous system during embryogenesis

and may play a role in SLOS through a mechanism not yet understood.


















CHAPTER 1
INTRODUCTION


Clinical Description of Smith-Lemli-Opitz Syndrome
In 1964, Smith, Lemli, and Opitz described a multiple congenital

anomalies/mental retardation syndrome in three unrelated males. Originally

termed RSH syndrome, representing the initials of each patient, this disorder is

better known today as Smith-Lemli-Opitz syndrome or SLOS. SLOS is usually

associated with features including (Table 1.1): 1. growth and mental retardation;
2. unusual facies (micrognathia, ptosis, strabismus, anteverted nares, low-set

ears, and cleft palate); 3. microcephaly; 4. limb abnormalities (in particular,

syndactyly of toes 2-3); 5. pseudohermaphroditism in males; and 6.
malformations of the heart, lungs, and gastrointestinal tract (Cherstvoy et al.,

1984; Smith et al., 1964). Clinicians also report that photosensitivity and self-

abusive behavior are common among SLOS patients (Irons et al., 1995). In

addition to clinical manifestations, several studies suggested an underlying

metabolic defect including reports of abnormal serum steroids (Chasalow et al.,

1985) and low maternal serum estriol levels and sex reversal (McKeever and
Young, 1990). Initially it was hypothesized that some SLOS features were due to

the direct effect of a genetic mutation(s) on morphogenesis while other anomalies

were secondary to the primary defect(s), accounting for the broad spectrum of

features (Cherstvoy et al., 1984).
SLOS is inherited as an autosomal recessive trait. This mode of
inheritance is supported by reports of recurrence among same sex and different












TABLE 1.1 Manisfestations of Smith-Lemli-Opitz Syndrome and UF53 Case Report


Clinical findings %
Developmental
Birth asphyxia/cyanosis
Breech presentation
Birth weight <2500 g
Prematurity
Neonatal feeding
difficulties
Neonatal hypotonia
Oligohydramnios
Prenatal growth
deficiency, mild
Psychomotor retardation
Shrill cry
Decreased fetal movement
Face
Eyelid ptosis
Strabismus
Cataracts
Low-set ears
Broad nasal bridge
Anteverted nares
Cleft posterior palate
Broad maxillary
aveolar ridges
Small tongue
Redundant sublingual
tissue
Micrognathia
Short neck
Cranium/CNS
Microcephaly
Narrow high forehead
Seizures
Ventricular dilation
Cerebellar vermis hypoplasia
Corpus callosum hypoplasia
Limbs


Postaxial polydactyly
Positional hand abnormalities
Short thumbs
Transverse/bridged
palmar creases
Fingertip whorls; increased
Short proximal limbs
Dislocated hips
Positional foot anomalies
Svndactvlv. 2-3 toes


Freouencv in literature


UF53


(50)


(50)


+
(60-70) +
+


(60)


(50)


(75-85)


65-95


Freauencv in literature


(80)












TABLE 1.1 continued


Clinical Findings %
Genitalia
Abnormal/ambiquous
in XY males
Hypospadias
Bilateral cryptochidism
Male pseudohermaphroditism
Micropenis and/or chordee
Female phenotype in XY males
Normal genitalia in XX females
Gastrointestinal
Pyloric stenosis
Colon aganglionosis
Gastroesophageal reflux
Skin


Heart


Lungs
P
Ir
Renal
H
C

Pancreas
Is
Liver


Facial hemangioma/
nevus flammeus
Loose skin, especially
nape of neck
Sacral dimple

Malformations, varied types


ulmonary hypoplasia
complete lobulation

ypoplastic or solitary
ystic changes, mostly
obstructive type

let cell hyperplasia


T TI:;' 2


I L ,A.t.jLUI Y J.L~ L.Lt ULI. I JJ-


(100) N/Ab

N/A
N/A
(70) N/A
N/A
N/A
+


15-25


(35)


(50)


(57)


(40-50)


Hepatic dysfunction
Gall bladder anomaly
Additional case manifestations
Loose skin, inner thighs +
Vertical, tibial skin creases +
Frequencies were given when available (Wallace et al., 1994).
Parentheses denote frequency in severely affected SLOS patients.
a information not available.
b not applicable.











sex sibships with SLOS (Curry et al., 1987; Kohler, 1983; Rutledge et al., 1984) and

consanguinity among certain SLOS families (Merrer et al., 1988) in which the

parents failed to display the disorder. The incidence of SLOS is between 1:20,000

and 1:40,000 live births (Lowry and Yong, 1980; Opitz, 1994) with an estimated

carrier frequency of 1-2% (Chasalow et al., 1985). In unpublished studies of the

Swedish population, an SLOS incidence of approximately 1:10,000 was calculated

when both live births and miscarriages were taken into account (John Opitz,

personal communication). These numbers suggest that SLOS is one of the most

common autosomal recessive disorders in the Caucasian population, and its

prevalence currently is likely underestimated.

The factor contributing most to the underestimation of frequency is

misdiagnosis. Misdiagnosis has been common since the severity can vary

dramatically, there is no pathognomonic feature, and several other genetic

disorders shares features with SLOS, such as Meckel syndrome (Lowry et al.,

1983), Pallister-Hall syndrome (Donnai et al., 1987), and lethal acrodysgenital

dwarfism (Merrer et al., 1988). Within the clinical genetics community, the

variability in SLOS and its overlap with Meckel syndrome (MS) brought forth

the question of whether these two syndromes represented different segments of a

broad phenotypic spectrum of a single syndrome. Concordance studies of

affected SLOS and MS sibships did not support this theory, though, and several

important distinctions between MS and SLOS have since been established, as

follows (Lowry et al., 1983). While both MS patients and SLOS patients may

display microcephaly, this feature is common in SLOS patients and rare in MS.

Although MS patients tend to exhibit encephalocele, they occasionally present

with microcephaly in the absence of encephalocele, which is similar to SLOS.
Both SLOS and MS males tend to have hypoplasia of external genitalia and

cryptorchidism; however, in SLOS males, these features are usually











overshadowed by pseudohermaphroditism. A final feature which led to

diagnostic confusion was the presence of orofacial clefts. The distinction,

though, was that MS patients typically have cleft lip while SLOS patients have

cleft palate. Clearly, accurate diagnosis of SLOS based on clinical features is

difficult, leading to misdiagnosis and probable underestimation of the prevalence

of SLOS in the general population.

In 1987, Curry et al. proposed classifying a phenotypically distinct

category of SLOS as Type II. Common features of SLOS Type I included major

structural abnormalities, male pseudohermaphroditism and early lethality.

Diagnosis of SLOS Type II was based on the presence of 3 or more of the

following: congenital heart defect, postaxial hexadactyly, cataracts, severe

genital ambiguity or pseudohermaphroditism in (46, XY) males, cleft palate and

small tongue, early death, or the presence of at least one other specific

malformation (Johnson et al., 1994). Curry's group felt that Type II represented

the extreme end of the SLOS phenotypic spectrum and was far more common

than milder SLOS (Type I). SLOS Type II was also believed to be inherited as an

autosomal recessive trait. However, it was unclear whether the same gene

caused both types. Curry's group reported the existence of intrafamilial

concordance of certain features among their cited Type II sibships as well as

other previously reported sibships. They reported that affected sibships exhibited

either SLOS Type I or Type II with the greatest discordance in longevity. The

same year, Opitz and Lowry argued that SLOS Type I and Type II may exist in

the same sibship, disputing the concordance so strongly favored by Curry's

group (Opitz and Lowry, 1987). The nosology debate over SLOS Type II as a

separate entity would not be resolved until 1993, when two collaborative groups

discovered a biochemical marker for SLOS shared by both Type I and Type II.

Still of debate, though, was whether allelic variation, genetic heterogeneity, or












possibly a contiguous gene syndrome accounted for the broad phenotypic

variation seen among SLOS patients.


Discovery of a Biochemical Defect in Cholesterol Synthesis

Several years ago, an apparent biochemical defect in cholesterol

biosynthesis was identified in SLOS patients (Irons et al., 1993; Tint et al., 1994).

In serum, the defect resulted in a 200- to 2,000-fold increase in the cholesterol

precursor 7-dehydrocholesterol (7-DHC) paralleled by a decrease in cholesterol

concentrations. This abnormality was also detected in erythrocytes, the lens of

the eyes, feces, bile, and cultured fibroblasts from SLOS patients, indicating that

the defect was widespread and most likely affected all organs. The enzyme

proposed to be defective in SLOS patients was 7-DHC reductase (7-DHCR),

which is responsible for the saturation of the C-7 double bond of 7-DHC to

produce cholesterol (Tint et al., 1994). The hypothesized 7-DHCR gene seemed

an attractive candidate gene for SLOS: however, no nucleic acid or protein

sequence is available in mammals (Duvoisin et al., 1995) to establish a definitive

correlation between this enzyme and SLOS.

The neutral sterols from the serum and tissues of fasting SLOS patients

were identified and their levels measured by capillary-column gas

chromatography (Tint et al., 1994). Several abnormal sterols were detected in the

serum and tissues of all SLOS patients studied. The most prominent sterol was

identified as 7-DHC and represented the largest sterol fraction in these patients.

The two other sterols were later identified as 8-dehydrocholesterol (8-DHC), an

(isomer of 7-DHC), and 19-nor-5,7,9(10)-cholestatrien-3 /3- ol (Batta et al., 1995).

Although 8-DHC is not a precursor to cholesterol, it is believed to be formed

from 7-DHC by an unidentified isomerase found in the hepatic microsomes. The

abnormal sterol 19-nor-5,7,9(10)-cholestatrien-3 1-ol is seen in some but not all












affected individuals; its origin is still unknown. Although no correlation

between the severity of SLOS and the levels of 7-DHC and 8-DHC has yet been

established, these sterols appear to be significantly more elevated in severely

affected patients. Conversely, 19-nor-5,7,9(10)-cholestatrien-3 13-ol seems to be in

similar quantities in both mildly and severely affected patients, about 10% of

total sterols, when present. It should be noted that the total sterol content does

not appear to be abnormally low since the elevated 7-DHC and presence of

aberrant sterols compensate for the cholesterol deficiency.

In cholesterol biosynthesis, the pathway becomes branched after the

formation of lanosterol (Figure 1.1) in which either the side chain C-24 double

bond is saturated immediately or as the final step (Clayton, 1965). In the case of

early C-24 saturation, the following cholesterol precursors include 24,25-

dihydrolanosterol, lathosterol, and 7-DHC. The other branch of the pathway

proceeds through a series of intermediates with an unsaturated side chain, C-24,

including cholest-5,7,24-triene-3 /3-ol and desmosterol. It is generally accepted

that these two pathways do not constitute separate and mutually exclusive

pathways, but instead share certain enzymes including 7-DHC reductase and

3 P -hydrosterol A 24-reductase, the enzyme responsible for reducing the C-24

double bond (Steinberg and Avigan, 1960).

In higher eukaryotes, cholesterol constitutes up to one-third of the lipid

membrane and is an essential component of the cell membrane and outer

mitochondrial membrane (Opitz et al., 1994). Cholesterol and its intermediates

are not only responsible for maintaining membrane integrity, but is also involved

in production of steroid hormones, bile acids, and vitamin D; protein

glycosylation, and myelination of the central and peripheral nervous system

(Acosta, 1995). In addition, de novo cholesterol synthesis is intricately involved in

the development of the testis, liver and brain (Ness, 1994). Considering the


















Lathosterol


4:


HO


Cholest-5,7,24-triene-3B-ol


7-Dehydrocholesterol


ORnEu n...


riu HO
Desmosterol Cholesterol



Figure 1.1. Schematic representation of cholesterol biosynthesis from
lanosterol (Tint et al., 1995). The right side of the pathway demonstrates early
saturation of the C-24 bond (7-DHC conversion to cholesterol is penultimate
step) while the left side demonstrates late saturation. In SLOS, it is proposed
that both branches have a block in cholesterol biosynthesis, indicated by the
Xs, at the level of saturation of the C-7 double bond by 7-DHCR.


HO


Lanosterol












ubiquitous nature of cholesterol, it seems plausible that a defect in its

biosynthesis could be the underlying mechanism for the widespread damage

seen in SLOS. However, the proposition of 7-DHCR as an SLOS candidate gene

is based strictly on circumstantial evidence since its involvement in SLOS at a

molecular level has not been corroborated experimentally.

Analysis of the urinary bile acids of SLOS patients by continuous flow fast

atom bombardment mass spectrometry uncovered two abnormalities (Natowicz

and Evans, 1994). First, normal urinary bile acids (cholanoates) were absent or at

subnormal concentrations. Cholanoates are essential in emulsification and

absorption of dietary fats (Acosta, 1995). This deficiency could thus account for

the feeding difficulties seen in 90% of SLOS patients. Second, the bile acids that

were present were abnormal: there was an added unit of unsaturation,

suggesting that 7-DHC may be further processed into bile acids in the presence

of the defect in cholesterol biosynthesis.

Cholesterol synthesis begins in the fetal adrenal glands and liver at about

8 and 10 weeks, respectively (Acosta, 1995). The accumulation of 7-DHC in SLOS

fetuses would presumably begin with the initiation of cholesterol synthesis

during the first trimester. The effects of 7-DHC accumulation on development,

cholesterol biosynthesis regulation, placental transfer of cholesterol, cell

membrane transport, and membrane permeability at this developmental stage

are still unknown.

Although the molecular basis connecting the defect in cholesterol

biosynthesis and SLOS has yet to be proven, both diagnosis and prenatal testing

for SLOS are now possible based on examination of the cholesterol and total

sterol profile. Prior to the biochemical discovery, women at risk for carrying an

SLOS fetus could obtain prenatal diagnosis based only on detection of a gross

anatomical defect, using by transvaginal ultrasonography during the first












trimester (Hobbins et al., 1994). Unfortunately, ultrasonography may fail to

detect milder cases of SLOS. Today, prenatal diagnosis is based on the

measurement of 7-DHC in amniotic fluid and chorionic villi sampling (Dallaire et

al., 1995; Mills et al., 1996). This method allows for detection of the abnormal

cholesterol profile in a fetus as early as the first trimester.

Several other disorders display a defect in cholesterol biosynthesis

reflected by a deficit in serum cholesterol concentrations, such as

hypobetalipoproteinemia (Linton et al., 1993). SLOS, though, is unique in that

patients not only have low levels of circulating cholesterol but also show low

levels in all tissues studied, indicating a disturbance of the endogenous

production of cholesterol. However, it can not be ruled out that this metabolic

defect is secondary to a primary defect (other than in 7-DHCR) not yet

elucidated. Characterization of the molecular defect(s) in SLOS is thus crucial for

further understanding of this disorder, as well as to improve diagnosis.


Case Report of a Severely Affected SLOS Patient, UF53

In 1994, our group reported an affected Caucasian female (UF53)

displaying the severe phenotype of SLOS (Wallace et al., 1994). The patient
presented with the characteristic SLOS facies (micrognathia, cleft palate, ptosis,

anteverted nares, hypoplastic tongue), developmental delays, syndactyly of toes

2-3, limb abnormalities, Hirschsprung disease and microcephaly (refer to Table

1.1 and Figure 1.2). UF53 was the only affected sib of three children born to

normal parents with an unremarkable family history. Prior to the child's death at

5-months of age, lymphoblastoid and skin fibroblast cultures were established,

and DNA was obtained from the parents' leukocytes. No autopsy was

performed.

















































Figure 1.2. Photographs of UF53 displaying the severe form of SLOS (Wallace
et al., 1994). The unusual facies (ptosis, strabismus, micrognathia, cleft palate,
and broad nasal bridge) and microcephaly associated with SLOS are shown in
the top panels. Note also the colostomy bag resulting from surgery for
Hirschsprung disease and the 2-3 toe syndactyly, both commonly seen in the
severe form of the disease.











The defect in cholesterol biosynthesis had yet to be discovered at the time

of UF53's diagnosis, therefore biochemical testing of the patient's serum neutral

sterol concentrations was not possible. It was observed though, that at age 3

weeks and 5 weeks, the patient had serum cholesterol levels of 45 and 55 mg/dL,

respectively. This is in contrast to the serum cholesterol levels of normal 3-

month-old children fed low cholesterol corn-soybean oil-based diets, which are

reported to be 109+9 mg/dL (Hayes et al., 1992). I later confirmed the defect in

cholesterol metabolism in UF53 by analysis of the fibroblast cultures, in

collaboration with Dr. Steve Tint (Alley et al., 1995). To confirm the clinical

diagnosis of SLOS in UF53, fibroblast cultures grown in delipidated medium

(cholesterol-deficient) were sent to Dr. Tint (Veteran's Hospital, East Orange

City, New Jersey) for analysis of cholesterol and 7-DHC levels. His studies

detected the SLOS abnormal sterol profile in UF53. Normal fibroblasts have

cholesterol concentrations of 40-60 mg/2 X 106 cells with a complete absence of

7-DHC. Cholesterol and 7-DHC concentrations in UF53 fibroblasts were 19 and

2.7 mg/2 X 10 6 cells, respectively. The results of the assay clearly demonstrated

that UF53 had a dramatic decrease in cholesterol concentration while 7-DHC

was significantly elevated. Together, the clinical presentation and the presence

of the aberrant cholesterol profile confirmed the diagnosis of SLOS.

Of particular interest in UF53, though, was the presence of a de novo

balanced translocation between the long arms of chromosomes 7 and 20,

[t(7;20)(q32.1;q13.2)], detected by routine cytogenetic studies (Wallace et al.,

1994). The idiogram in Figure 1.3 schematically represents the breakpoint

assignments to 7q32.1 and 20q13.2, which were made at a resolution of 650-800

bands/haploid set of chromosomes. The de novo appearance of both the

translocation and SLOS in UF53 suggested a possible etiological connection.

Therefore, based on this hypothesized connection, UF53's translocation served as






































?q31




7








Figure 1.3. Schematic representation of the karyotype, [t(7;20)(q32.1;q13.2)],
resulting from UF53's translocation between the long arms of chromosome 7
and chromosome 20 at q32.1 and q13.2 (indicated by the arrows), respectively
(Wallace et al., 1994)-










a physical landmark for positional cloning efforts in the elucidation of the

causative SLOS gene.


Positional Cloning Effort
Positional cloning is a method of isolating disease-causing genes based on

the relative chromosomal location of the gene rather than knowledge of the

etiological basis of the disorder (functional cloning). Although SLOS has an

apparent biochemical defect in cholesterol biosynthesis, the lack of

7-DHCR protein sequence eliminates the possibility of functional cloning based

on this proposed defect.
Figure 1.4 illustrates the general approach to positional cloning. Most

positional cloning efforts begin with the genetic linkage of a disorder to a broad

chromosomal location based on its segregation in families with polymorphic

markers. In the case of SLOS, linkage was not a feasible approach for several

reasons: 1. lack of large families; 2. diagnostic difficulty in adults, as abnormal

features lessen with age (de Die Smulders andFryns, 1992); and 3. the possibility

of genetic heterogeneity, which could prevent linkage detection. Hence, the

discovery of UF53's translocation was fortuitous since gross chromosomal

rearrangements act as physical landmarks for positional cloning of disease genes

when linkage is not available. UF53's translocation offered a unique resource in

the search for a putative SLOS gene due to the de novo appearance of the

chromosomal rearrangement and SLOS in this patient. Several human disease

genes have been cloned based on the presence of translocations including

neurofibromatosis type 1 (Wallace et al., 1990), Duchenne muscular dystrophy

(Monaco et al., 1986), choroideremia (Cremers et al., 1990), and adult polycystic

kidney disease (Wunderle et al., 1994).





















eneticiI


Markers






YACs Candidate genes

Figure 1.4. Schematic of the general approach to positional cloning. Initially, family pedigrees are collected in
which the disease gene is segregating. These families are studied with various markers until there is evidence
which links on or more of the markers to the disease. The region is further narrowed by additional linkage.
The existence of a translocation allows for the quick advancement to the cytogenetic level. In order to narrow
the region of interest, one would try to identify large genomic clones, such as YACs in the region. Once the
region of interest has been narrowed to the kilobase level, various methods can be employed to identify
candidate genes in the region.










I chose to begin positional cloning efforts at the 7q32.1 region, rather than

chromosome 20, based on literature review which suggested that an SLOS locus
resided in the distal 7q region. This assignment was based on several reports of

individuals with SLOS and SLOS-like features with chromosomal deletions in the
7q32-qter region (Berry et al., 1989; Bogart et al., 1990; Schwartz et al., 1983;

Young et al., 1984). More convincing, though, was the report of another severely

affected SLOS male with a maternally inherited balanced translocation involving
7q32 (Curry et al., 1987). To fit the autosomal recessive model, I hypothesized

that UF53's translocation disrupts a putative SLOS allele, at 7q32.1, while a more

subtle, inherited, mutation disrupts the other allele.

















CHAPTER 2
IDENTIFICATION OF A YEAST ARTIFICIAL CHROMOSOME SPANNING
UF53'S TRANSLOCATION




Introduction
In positional cloning, once a chromosomal location has been established, a

variety of methods can be employed to successively narrow the region of
interest. To more finely map UJF53's translocation from a cytogenetic level to a

kilobase (kb) level, I intended to pursue two methods: somatic cell hybridization

and fluorescence in situ hybridization (FISH). Both methods have been
extensively used in the elucidation of the human genomic map and would

directly establish position of large clones, such as yeast artificial chromosomes

(YACs) and cosmids, relative to rearrangements such as the UF53 translocation

(Brock, 1993; Rothwell, 1993).

Somatic cell hybridization is a method in which somatic cells from two
different species, usually human and rodent, are fused together in the presence

of polyethylene glycol or Sendai virus. These agents alter the cell membranes in
such a way that fusion is enhanced and cells from the different species form a
single cell with two nuclei called a heterokaryon. These heterokaryons are then

grown in HAT medium which selects for hybrids containing nuclei from both
species, due to complementation of a mutation in the rodent cells. The selected,

newly-formed cells have a tendency to lose the human chromosomes while
maintaining the rodent background. Since this loss is random, each clone differs











in its human chromosome complement. After the initial loss of chromosomes,

stable cell lines from the clones can be established since no additional

chromosome loss rarely occurs.

The goal was to establish individual somatic cell hybrids from UF53's cell

lines:, one containing the derivative 7 chromosome, one having the derivative 20

chromosome, and one having the normal 7 chromosome. With these hybrids, the

location of genomic clones and cDNAs relative to the translocation could be

established based on sequence complementation to this region using Southern

methodology. Fibroblast cell lines were sent to Dr. James P. Evans (University of

Washington, Seattle) to establish somatic cell hybrids, but unfortunately all

attempts at fusion failed. A second attempt was made by Dr. David Cox

(Stanford University, California) to produce hybrids using both the fibroblast

and lymphoblastoid cell lines from UF53. Repeated experiments using both cell

lines -against several rodent backgrounds also resulted in failure. It was

suggested that the inability to establish somatic cell hybrids was possibly due to

the abnormal sterol composition in the cell membranes in UF53, the result of the

cholesterol biosynthesis defect or some other abnormality specific to UF53 or

SLOS (Dr. David Cox, personal communication). Further attempts to establish

somatic cell hybrids were abandoned.

As an alternative, FISH was developed to localize large genomic clones,

such as yeast artificial chromosomes (YACs) and cosmids, relative to the UF53

translocation. Chromosome in situ hybridization is based on the concept that a

labeled genomic clone (probe), when heated, will denature into single-stranded

forms that will anneal to complementary sequence on a denatured chromosome,

producing a hybridization signal which reveals the probe's cytogenetic location

(Figure 2.1). Figure 2.2 represents the three possible outcomes (distal, proximal,

or spanning) resulting from the hybridization of a YAC clone against UF53's












FISH


chromosome 7


dig dig

+ Y "" YAC clone


hybridization


dig dig
YAC clone
ILI~I .III chromosome 7


detection


metaphase spread


i
dig


2e
dig


.. ..-... -YAC clone
chromosome 7




Figure 2.1. Illustration of FISH analysis. The patient's metaphase chromosomes, shown in the circle, are denatured
and hybridized with a biotin- or digoxygenin-labeled genomic clone (i.e. YAC clone). The YAC clone anneals to its
homologous sequence on the chromosome (i.e. chromosome 7). The hybridization is detected by using
fluorescently-tagged antibodies specific to biotin or digoxygenin and then detected by fluorescent microscopy.











chromosome 7
Distal (3TIIII E!U liillil)





Proximal lIEIMIIII1Iiiil)





Spanning cffEffIpCLI IjjjIyUjI
=


derivative 7
Iiii Mnxluzii|m





miII1111l~ll


derivative 20










DOM-


Figure 2.2. Schematic representation of expected FISH results. A genomic clone which hybridizes distal (shown as
a blue box) or proximal (shown as a yellow box) to the breakpoint will have a hybridization signal on the normal
chromosome 7 and on the derivative chromosome 20 or derivative chromosome 7, respectively. A genomic clone
spanning the translocation (shown as a green box) will have hybridization signals on the normal chromosome 7
and both derivative chromosomes.











metaphase spreads. One of three outcomes was expected from each FISH

experiment. In the first, a YAC clone is found to be distal to the breakpoint,

indicated by a hybridization signal on the normal chromosome 7 and derivative

chromosome 20. Secondly, a YAC clone lying proximal to the breakpoint,

indicated by a signal on the normal chromosome 7 and the derivative

chromosome 7. The third outcome would be detection of a YAC clone spanning

the translocation, in which hybridization signals would be detected on the

normal chromosome 7 as well as both derivative chromosomes. Since somatic

cell hybridization was no longer possible, FISH provided the most direct means

of localizing genoniic clones from the 7q32 region relative to UF53's

translocation breakpoint. The ultimate goal of such FISH experiments was

localization of the translocation to one YAC.

As an initial step in my positional cloning effort, as described in this chapter,

I used FISH to position several YAC clones proximal and distal to the 7q32.1

translocation breakpoint in UF53 and successfully identified a YAC clone that

spanned the translocation. These results placed the translocation breakpoint on a

more highly resolved physical map, on the genetic map, and narrowed the

candidate gene region to approximately 1 Mb.


Materials and Methods

YAC clones and DNA markers. The chromosome 7-specific YAC clones

were provided by our collaborators Dr. Lap-Chee Tsui and Dr. Stephen W.

Scherer (Hospital for Sick Children, Toronto). These clones were isolated from a

chromosome 7-specific YAC library (Scherer et al., 1992) by screening with the

microsatellite markers listed in Table 2.1. Information about the microsatellite

markers can be found in the Genome Database (GDB). The unpublished

G~n~thon microsatellite marker AFMa125whl was provided by Dr. Jean











Weissenbach. The Toronto group determined the sizes of the YACs by pulsed-

field gel electrophoresis of the yeast chromosomes followed by blot hybridization

with a vector specific probe (pBR322) and comparison to YPH149 S. cerevesiae

chromosome standards (Scherer and Tsui, 1991).

Alu PCR. Yeast containing chromosome 7 YACs was grown in YPD

medium to a density of 2X107 cells/ml (Scherer et al., 1993). The cells underwent

spheroplasting using the Yeast Cell Lysis Preparation Kit (BIO 101), and high

molecular weight DNA was prepared using the G NOMETm DNA Kit (BIO 101).

The isolated DNA was phenol-chloroform extracted, ethanol precipitated and

dissolved in TE. In order to generate labeled probes specific to the chromosome
7 YAC, 20-30 ng of purified DNA from each YAC clone were subjected to inter-

Alu PCR under conditions described by Tagle and Collins (1992) using

combinations of Alu element and YAC vector primers (Brooks Wilson et al., 1990;

Lengauer et al., 1992; Riley et al., 1990; Tagle and Collins, 1992).

FISH. Five-hundred ng of combined inter-Alu and Alu-vector PCR

products were biotinylated by random priming using the BioPrimeTM DNA

Labeling System (Gibco BRL). Approximately 100-150 ng of biotinylated DNA

were used as a probe for each slide. Peripheral lymphocyte and lymphoblastoid

chromosome spreads were freshly prepared by standard cytogenetic procedures

(Yunis, 1976). The biotinylated YAC probe was simultaneously hybridized to the

slides with digoxygenin-labeled alpha-satellite centromeric probes for

chromosomes 7 and 20 (Oncor), using Cot-1 DNA competition (Pinkel et al.,

1986). Slides were hybridized for approximately 16 hours at 37C. Post

hybridization washes consisted of one 1XSSC (0.01 M NaCl, 0.3 M sodium citrate,

pH 7.0) wash at 72C for 5 minutes followed by three consecutive washes in

1XPBD (Oncor) for 3 minutes at room temperature. Signal detection was attained

by incubation with fluorescein isothiocyanate-conjugated (FITC) avidin and anti-











digoxygenin (Oncor). One round of signal amplification, using anti-

avidin/FITC-avidin, was also carried out. Chromosomes were counterstained

with propidium iodide (Oncor). The hybridization signals were analyzed and

documented with an Olympus BHS fluorescence microscope system and the

CytoVision/CytoProbe computer imaging device (Applied Imaging, Inc.) in

collaboration with Brian A. Gray and Dr. Roberto Zori (Raymond C. Philips Unit,

University of Florida).


Results

Localization of the translocation breakpoint at 7q32.1 using FISH. Alu

PCR products from eleven YAC clones (HSC7E61, HSC7E67, HSC7E451,

HSC7E1351, HSC7E888, HSC7E77, HSC7E81, HSC7Ell7, HSC7E591, HSC7E135,

HSC7E5), from locations throughout the long arm of chromosome 7, were

initially used as probes in FISH mapping experiments to determine their

locations relative to the UF53 translocation breakpoint. Examples of the

hybridization signals are shown in Figure 2.3 and the results are summarized in

Table 2.1.

HSC7E1351 and HSC7E244, which were previously mapped to the 7q32

region (Kunz et al., 1994), were determined by FISH to be proximal (centromeric)

and distal (telomeric), respectively, to the translocation breakpoint (Table 2).

These two clones were also linked within a contig of fourteen YACs anchored by

the ffiicrosatellite markers AFM323wd5 (D7S686), AFMa125wh1, AFM309yfl

(D7S680), AFM218xf1O (D7S514) and AFM206xcl (D7S635) (Figure 2.4). The

contig spans approximately 3 Mb (Figure 2.4) and the genetic distance between

D7S686-(D7S680-D7S514-D7S635) is 1 cM (Gyapay et al., 1994).

To refine the localization of the translocation breakpoint within the YAC

contig, Alu-PCR products from four additional YACs in the contig were tested by


































Figure 2.3. FISH analysis of the SLOS translocation breakpoint region using YAC clones from the 7q32 region.
Biotin-labeled Ali PCR products from the YACs (yellow) were cohybridized with digoxygenin-labeled alpha
satellite centromeric probes for chromosome 7 and 20 (yellow) to metaphase chromosomes, under competitive
conditions. Metaphase spreads of UF53 lymphocytes after hybridization with the YAC clones (a) HSC7E451,
(b) HSC7E1289, and (c) HSC7E261 demonstrate proximal, spanning and distal positioning of probes, respectively.
Metaphase spread of an unaffected individual (d) hybridized with HSC7E1289 serves as a control. Chromosomes
are counterstained with propidium iodide. Arrows point to the centromeric signals specific for chromosomes 7
and 20.






25.

















Table 2.1. Summary of FISH Experiments Using YACs

bLocation Relative
aMap Location on Size to Ihe Translocation
YAC Clones chromosome 7 (kb) DNA Markers Contained within YAC Breakpoint
HSC7E61 7q21.2-q21.3 460 D7S558, D7S646, D7S657, D7S689 Proximal
HSC7E67 7q32 360 D7S648 Proximal
HSC7E451 7q32 590 D7S487, D7S648 Proximal
HSC7E1351 7q32 1,200 D7S686, AFMa125whl Proximal
HSC7E1289 7q32 1,800 D7S686, AFMal25whl, D7S680, D7S514 Spans
HSC7E261 7q32-q33 580 D7S514, D7S635 Distal
HSC7E476 7q32-q33 700 D7S514, D7S635 Distal
HSC7E888 7q32-q33 410 D7S514 Distal
HSC7E244 7q32-q33 680 D7S680, D7S514, D7S635 Distal
HSC7E77 7q33 380 HBNF Distal
OSC7E81 7q33 350 HBNF Distal
HSC7E117 7q33-q34 480 'TCRB Distal
HSC7E591 7q33-q34 580 TCRB Distal
HSC7E135 7q36 1,300 Distal
HSC7E5 7q36 300 D7S 104 Distal
aThe chromosome locations of the (HSC7E)-YACs were determined by FISH in this study or in another study (Kunz et al., 1994).
bproximal = signals observed proximal to the breakpoint (retained on der 7); distal = signals observed distal to breakpoint (transferred
to the chromosome 20 translocation derivative); and spans = probe spans the breakpoint (signals were observed on both translocation
derivatives)
















I SLOS
translocatioh I


HSC7E122


IISC7E 1326


SC7E1351:: : : :: : :: ::


7cen


ItSC7E261

ISC7E244


I1SC71;1289


7qter


I I I I I


Scale: in kb


100 500 1000


YAC proximal

Z YAC spans breakpoint
XYAC distal


Figure 2.4. Tentative map of YACs overlapping HSC7E1289, which is known to span the translocation breakpoint.
YAC locations relative to breakpoint were based on FISH data, according to the key. The YACs represented are
considered nonchimeric, since the marker data are consistent, and the YACs were derived from a chromosome 7
library in which chimerism is unusual (Kunz et al., 1994). Genetic markers used to determined overlaps are listed

below the line. FISH results suggest that the translocation most likely falls between markers AFMa125wh1 and
AFM309yfl, as indicated by the box.


1 HSC7E888
". ." ",.", ". ", ". I SC7E476


I.... ...... ............................. ........... .. .


.. 1 11 1 ''1'- 1 1 1 I I I . ... .


/a, 11 z /


z 11 z z 11











FISH. Three of the YACs (HSC7E888, HSC7E261, HSC7E476) appeared to be

distal to the breakpoint (Table 2 and Figure 2.4). However, HSC7E1289

wasfound to span the breakpoint since in addition to a signal on the normal

chromosome 7, the probe also yielded signals on both translocation derivatives

(Figure 2.3). The combination of the FISH results from the other YACs suggested

that the SLOS breakpoint lies in the telomeric half of HSC7E1289 (Figure 2.4).

The results unequivocally supported the karotyping results that placed the UF53

translocation breakpoint (and presumably SLOS) at 7q32.1.


Discussion

The existence of an SLOS patient with a de novo balanced translocation

suggested a direct association between the chromosomal abnormality and the

syndrome. The UF53 translocation served as a landmark in my initial search for

the SLOS gene. I identified an 1,800 kb YAC, HSC7E1289, that spans the

translocation breakpoint, moving the level of map resolution for this

translocation from the cytogenetic band to the megabase-level.

Due to the large size of HSC7E1289, steps were taken to narrow the region

of the breakpoint. HSC7E1351 and HSC7E244, which overlap HSC7E1289,

mapped proximal and distal, respectively, by FISH. These results tentatively

placed the breakpoint (i.e. SLOS gene) between markers AFMa125whl and

AFM309yf1 (Figure 2.4). It should be noted, however, that the Altu-PCR FISH

approach yielded results of a qualitative nature. Since Alu element distribution

in the genome is nonuniform and was unknown within these YACs, the Alu PCR

probe fragments may not have been representative of the entire YAC; thus, lack

of signal on one or the other derivative chromosome may have been a false

negative signal. To counter this, the probes included vector-Alu PCR products

(present for most of the YACs) to help represent the YAC ends in the probe












mixture, but this does not guarantee observable signal from the ends. Other false

negative signals could have been the result of a microdeletion at the breakpoint,

for which there is currently no evidence. Therefore, I could not yet exclude the

breakpoint from a larger region of HSC7E1289, and the conservative estimate of

the SLOS breakpoint region was approximately 1 Mb.
The wide phenotypic spectrum of SLOS had traditionally caused a

dilemma in clinical diagnosis; it was thought that many cases of SLOS were not

appropriately diagnosed and were categorized under the nonspecific label of

multiple congenital anomalies/mental retardation syndrome. SLOS, previously

viewed as an uncharacterized, difficult-to-diagnose rare autosomal recessive

disorder, was seen in a new light with the discovery of the aberrant sterol profile.

However, the biochemical findings may not be exclusive to SLOS, and there may

be biochemical or genetic heterogeneity. Clearly, for clinical purposes and

development of effective therapies, where correct diagnosis is crucial, it is very

important to elucidate the gene(s) involved in SLOS. Thus, at this point in my

work, identification of the UF53 SLOS gene, presumably at the site of the

translocation, was my goal. Localization of the UF53 translocation breakpoint

was the first significant step toward this goal.

















CHAPTER 3
PHYSICAL MAPPING OF THE TRANSLOCATION BREAKPOINT REGION


Introduction

Prior to my arrival in Dr. Margaret Wallace's laboratory, a collaboration

with Dr. Lap-Chee Tsui at the Hospital for Sick Children (Toronto, Ontario) was
established based on his involvement in the Human Genome Project, mapping

chromosome 7. After narrowing the UF53 translocation breakpoint region to
200-300 kb by FISH, significant steps toward physical mapping of the region and

identification of genes in the region needed to be taken. Dr. Tsui, Dr. Wallace,

and I felt that the techniques and resources available in Dr. Tsui's laboratory
would expedite the identification of candidate cDNAs. His laboratory was kind
enough to provide me the space, resources, and their expertise for a month in

1994. This chapter includes data from experiments which I initiated while

visiting Dr. Tsui's laboratory and completed upon my return to Dr. Wallace's
laboratory.

The following methodologies were executed and the details of each

discussed in this chapter: 1. pulsed-field gel electrophoresis to establish a rare-
restriction site map of the YAC contig; 2. direct selection of cDNAs from

HSC7E1326; and 3. FISH studies using HSC7E1289 cosmids, to continue

narrowing the breakpoint region.
An extensive YAC contig was constructed, including (HSC7E)-YACs as

well as CEPH-G~n~thon YACs, by Dr. Tsui's laboratory prior to my arrival. The

chromosome 7-specific (HSC7E)-YAC library (Scherer et al., 1992) and CEPH-












G~n~thon library were screened with the polymorphic microsatellite markers
specific for the 7q32 region (Kunz et al., 1994). Additional YACs thus identified

were aligned with HSC7E1289 (the spanning YAC) and other YACs in the contig,
based on the polymorphic microsatellite markers contained within each. The

additional YACs would be used in future studies.
Construction of a physical map of the 7q32 region using PFGE established

the locations of rare-restriction endonuclease sites. YAC clone DNA was

digested with rare-restriction endonucleases, and the fragments were separated
by PFGE and Southern blotted. Restriction maps of the YAC clones were

constructed by sequential hybridizations of the PFGE blots to probes derived

from the YAC ends. Autoradiography patterns revealed the positions of the rare-

restriction sites in the overlapping YAC clones. These data simultaneously

revealed the locations of two CpG islands, which serve as potential landmarks

for constitutively expressed genes.
A relatively new methodology, direct selection, was used to identify

cDNAs encoded by the region. Direct selection is particularly useful and

efficient in screening large DNA fragments, even those greater than 1.0 Mb
(Parimoo et al., 1991). As illustrated in Figure 3.1, the selection process is based

on hybridization of cDNA fragments to immobilized cloned genomic DNA and

recovery of the selected cDNAs by PCR (Rommens et al., 1994). Direct selection

begins with the production of cDNAs from reverse transcription of total and
poly-A+ RNA from various tissues, using primers with specific sequence tags for

each tissue. These cDNAs are pooled and expanded by PCR. The YACs are
isolated from pulsed-field gels and transferred to a filter. Occasionally, due to
size overlap, it is impossible to separate the YAC from a yeast chromosome, a

disadvantage overcome by subsequent negative selection with yeast genomic


















Immobilize YAC







-Hybridize Elute specific cONAs
2 Cycles

Amplify and Clone



Expanded Characterization of
cDNA Pool Mini-library of cDNAs

Blocked for Repeat
Normalized
Figure 3.1. Schematic of basic direct selection strategy (Rommens et al., 1994). The pool of expanded cDNAs is
competed with repetitive DNA and hybridized with the immobilized genomic DNA. The membrane is washed
and the cDNA eluted off. These cDNAs are again competed with repetitive DNA and rehybridized to the
membrane. The membrane is once again washed and the cDNAs eluted and amplified by PCR. The PCR products
are subsequently cloned into a plasmid vector and characterized.












DNA. Prior to the hybridization, competition with repetitive DNA is necessary;

competition against either the cDNA pool or the immobilized YAC works

equally well (Lovett, 1994). The cDNA pool is then hybridized with the filter,

after which the filter is washed and the bound cDNAs are eluted. The eluted

cDNAs are again competed and rehybridized to the same filter followed by

elution. These highly-enriched, selected cDNAs are amplified by PCR and

shotgun subcloned into a vector, followed by transformation into competent E.

coli cells. Clones are then analyzed for content of repetitive elements, yeast DNA

or human genomic DNA (i.e. false positive background clones) (Parimoo et al.,

1993). Prior to my arrival in Toronto, direct cDNA selection from HSC7E1326

was initiated by Dr. Joanna Rommens in collaboration with Dr. Tsui's laboratory.

I was responsible for the analysis of the direct selected cDNA clones to determine

whether they contained yeast or human genomic DNA. I was also involved in

localizing the human cDNA clones on the YAC contig using Southern analysis.
Additionally, in a parallel positional cloning strategy, the Genome Data

Base (GDB) and literature were searched for potential candidate genes. Initial

candidate genes included all genes and uncharacterized cDNAs in 7q32, even if

their relationship to cholesterol metabolism was unknown. This "positional

candidate" approach has been used successfully to identify the genes for

hereditary hemorraghic telangiectasia (McAllister et al., 1994) and

achondroplasia (Shiang et al., 1994), among others. This strategy can save

significant effort, although it relies on good fortune. Literature review revealed

that the PAX4 gene maps to same general region as the translocation breakpoint.

PAX4 is one of nine genes belonging to the PAX family of homeobox genes. This

family encodes nuclear transcription factors involved in developmental control

during embryogenesis in both vertebrates and invertebrates (Stuart and Gruss,

1995). PAX4 was originally mapped to chromosome 7 in 1993 (Stapleton et al.,











1993) and then regionally mapped to the chromosome band 7q32 by FISH a year
later (Tamura et al., 1994). The PAX4 gene was later found to reside

approximately 500 kilobases distal to the UF53 breakpoint. Due to its close
proximity to the translocation breakpoint and its ubiquitous involvement in

transcriptional regulation during embryogenesis, investigation into its possible
involvement in SLOS was warranted. Although PAX4 was clearly not directly

disrupted by the translocation, it was conceivable that the expression of this gene

could be influenced by a position effect similar to that described in aniridia
(which involves another family member PAX 6.) (Fantes et al., 1995) and

campomelic dysplasia (Foster et al., 1994; Schafer et al., 1995). In both disorders,

the causative genes are positioned 100-200 kb from translocation breakpoints.
One explanation for positional effect is that the gene rearrangement leads to an

inappropriate chromatin environment for normal gene expression (Foster et al.,
1994). Since position effect was difficult to investigate, I set aside further analysis

of PAX4 while searching for other genes closer to the breakpoint.

Also, during my visit, a partial cDNA of the mGluR8 gene (Duvoisin et al.,

1995) was serendipitously mapped to the human 7q32 region. GRA48, the

human homolog of nGluR8, a novel member of the metabotropic glutamate

receptor gene family, was found to lie on HSC7E1289; however, the genomic

structure, cDNA sequence, and expression pattern of this gene had not been
determined (Scherer et al., 1996). Please note the difference in nomenclature

between the mouse and the human homologs: the mouse gene is referred to as

nGluR8 while the human gene is referred to as GRM8. While the partial GRM8

cDNA probe clearly mapped significantly centromeric to the breakpoint, it was
unknown if its genomic sequence was disrupted by the translocation. Although

this gene was not known to be involved in cholesterol synthesis, it was unclear if

its disruption could contribute to the SLOS phenotype. Thus, GRM8 became the











top candidate gene in the region and was further analyzed as described in
Results.


Materials and Methods

Large scale preparation of YAC PFGE blocks. One-hundred ml of YPD

(1% bacto yeast extract, 2% bacto peptone, 2% dextrose, pH 5.8) media with 100

tg/ml ampicillin was inoculated with a 5-10 il of a YAC glycerol stock and
incubated overnight at 30'C. Cultures were centrifuged at 2000 rpm for 5

minutes at 20'C, supernatant decanted, and the pellet resuspended in 4 ml SCE

(1.0 M sorbital, 0.1 M sodium citrate, 0.06 M EDTA). To this suspension,

2 .l of t9-mercaptoethanol and 50 [Il lyticase (100 units/ml) were added and
mixed. An equal volume of molten (50'C) 1% SeaPlaque GTG agarose (FMC

BioProducts) in 50 mM EDTA was added and mixed. This mixture was poured

onto a glass plate and allowed to cool for 15 minutes at room temperature and 15
minutes at 4C. The agarose was cut into appropriate sized blocks

(approximately 5 mm X 10 mm), transferred to 20 ml SCEM (SCE + 0.2% /3-

mercaptoethanol), and incubated at 37C overnight while gently shaking. The

YAC blocks were washed with lysis buffer (100 mM EDTA, 10 mM Tris-Cl, at pH

7.5) three times to rinse away debris, followed by an overnight incubation in 20
ml lysis buffer with I mg/ml Proteinase K at 37C with gentle shaking. The

blocks were then dialyzed with 50 mM EDTA, 3 times over a period of 1-2 hours.
Blocks were transferred to fresh 50 mM EDTA and stored at 4'C.

Partial digest of YAC blocks with rare-restriction endonucleases.

Individual YAC blocks were placed in eppendorf tubes containing serial

dilutions (1:0, 1:10, 1:50) of the restriction endonucleases BssHII, MluI, or NotI, 10
iil appropriate 10 X buffer, and sterile distilled water in a final volume of 100 [l.
These were incubated on ice for 30 minutes, followed by incubation at the












appropriate temperature (50'C for BssHII and 37*C for Mlul and NotI) for a
minimum of 3 hours. The digestions were stopped with the addition of 50 mM

EDTA.
PFGE of YAC blocks. Digested YAC blocks were loaded into the lanes of

a 1/2 X TBE 1% agarose gel in 1/2 X TBE buffer. The gel was run in a BioRad
CHEF DRII hexagonal field unit, at 4C, ramping from 20 seconds per electrode

to 40 seconds per electrode evenly over 20 hours at 200 volts. DNA was

transferred to a nylon membrane using standard Southern methodology
(Sambrook et al., 1989). The PFGE blot was hybridized with a vector-specific
probe (pBR322) labeled by random-priming with a-32P dCTP (MegaPrime kit,

Amersham). Hybridization patterns were visualized by autoradiography, and

band sizes were determined by comparison to YPH149 yeast chromosome

standards (Scherer and Tsui, 1991).
Rapid DNA preparation from yeast. Five ml of YPD with 100 tg/ml

ampicillin were inoculated with 5-10 p1 YAC glycerol stock and incubated

overnight at 30*C. The cells were pelleted at 2000 rpm for 5 minutes, the

supernatant removed, and the pellet resuspended in 500 tl sterile distilled water

and transferred to an eppendorf tube. The cells were pelleted by

microcentrifugation and the supernatant removed. To the cells, 200 pLl of GDIS
(2.0% Triton-X-100, 1.0% SDS, 100 mM NaCl, 10 mM Tris-Cl (pH 8.0), 1 mM

EDTA), 200 jtl phenol-chloroform:isoamyl alcohol (24:1), and .35 grams of heat-

treated glass beads were added to the cell suspension and vortexed for 2.5

minutes. To this mixture, 200 p1 sterile distilled water were added, mixed, and

then microcentrifuged for 1 minute. The aqueous layer was transferred to a fresh

eppendorf tube containing 6 pLl RNAse A (10 mg/ml) and incubated at room

temperature for 30 minutes. The DNA was then ethanol precipitated and
resuspended in 100 pLl sterile distilled water.











Southern blot analysis of 7q32 YACs with mGluR8, mGluR8 selected

cosmids, and direct selected cDNAs. Purified genomic DNA from the YAC
clones from the chromosome 7-specific YAC library or the CEPH-G~n~thon YAC
library (Figure 3.2) were subjected to restriction endonuclease digestion with

EcoRI. The restriction fragments were separated on a 1% agarose gel and

transferred to a nylon membrane by standard protocol (Sambrook et al., 1989).
The full-length mGluR8 cDNA (kindly provided by Dr. Duvoisin), unique

cosmid sequences (described below), and direct selected cDNAs were labeled by
random-priming with c-32P dCTP (MegaPrime kit, Amersham) for individual

hybridization against YAC blots. Hybridization and wash conditions were based

on a version of the Church and Gilbert SDS/sodium phosphate method (Church

and Gilbert, 1984).

Identification of unique sequence fragments from the mGluR8 cosmids.
The full-length mGluR8 cDNA was used to screen a human chromosome 7-

specific cosmid library to identify cosmids presumably containing exons of the

GRM8 gene. The screening, done in Toronto, resulted in the retrieval of 16
cosmids. The bacterial strains containing the cosmids were grown in 100 ml

Luria broth with 25ug/ml kanamycin overnight in a 37C orbital shaker. The
cosmid DNA was then purified according to the Qiagen Plasmid Midi Kit

protocol. Each cosmid (200 ng) was digested with EcoRI, with the fragments
separated on a 1% agarose gel and transferred to a nylon membrane (Sambrook

et al., 1989). The cosmid blot was hybridized with total human genomic DNA

(50 ng) labeled by the random-priming with x-32P dCTP (MegaPrime kit,
Amersham). Bands were visualized by autoradiography. Fragments containing

unique (non-repetitive) sequence were identified by the absence of hybridization
with total human genomic DNA, and were isolated according to the QIAEX II

protocol (Qiagen).











Double digest of GRM8 cosmids. Cosmid DNA was extracted as

described above. Each cosmid (500 ng) was digested with the restriction
endonuclease EcoRI at 37'C for 1 hour. The digested DNA was ethanol
precipitated and resuspended in IXTE. Three 100 ng aliquots of the EcoRI

digestion were then digested with the rare-restriction endonucleases BssHII,

Mlul, or NotI for an additional hour. The fragments were then separated on a 1%

agarose gel, stained with ethidium bromide and visualized on a UV light box.


Results
Completion of 7q32 the YAC contig. Prior to my arrival in Toronto,

screening of the chromosome 7-specific (HSC7E) YAC library and CEPH-

G~n thon YAC library with the 5 polymorphic microsatellite markers contained
in HSC7E1289 (refer to Figure 2.2) and two new G~n~thon markers, AFMa065zg9

and D7S1801, resulted in the recovery of 16 YACs. These additional YACs were

aligned with HSC7E1289 based on their marker content, resulting in an

expansive 7q32 contig composed of 24 overlapping YACs (Figure 3.2). This

construction of a more complete YAC contig was intended to facilitate more
precise positioning of various clones and cDNAs recovered from other studies.

Establishment of rare-restriction sites in the 7q32 YAC contig by PFGE.

Using restriction endonucleases that recognized rare restriction sites, PFGE

produced a physical map of the YAC clones (HSC7E1289, HSC7E122, c752c8,
HSC7E476, HSC7E261, and HSC7E244) surrounding the translocation breakpoint

(Figure 3.3). A CpG island consisting of two BssI-HI sites, one Mli site and one
NotI site, was identified distal to the translocation on the telomeric end of

HSC7E1289. An additional CpG island was also identified at the distal end of the

contig not contained in HSC7E1289. Of particular interest, though, was the












HSC7E91
c855h10 I I 1 .1


I I I I


c744c2 I Ii I


c769g101
cM3Od41 I


c938c7 |II


l I -F] HSC7E1824

I -?- Izc799c8


M u n


II


I I HSC7338


II I1SCE888


1 0 i


I I


] HSC7E476


HSC7E122 1IZ Z 7Z
HSC7EI326 I I I
I I I I I' II -


I n I
HSC7E261
ZHSC7E244


i m I I1 I


I I II I


I I I


I II I I I I


7 cen


Scale: in kb

I I I I
250 500 750


o's.
N
C


7 qter


1000


Figure 3.2. Expanded map of the 7q32 YAC contig. The (HSC7E)-YACs were obtained from a chromosome 7-
specific YAC library (Scherer et al., 1992) and the CEPH YACs were obtained from the CEPH-G~nthon YAC
library. The five original microsatellite markers, as well as two additional G~n6thon markers, were used to align
the YACs (markers indicated by black vertical bars),. YAC HSC7E1289 found to span the UF53 breakpoint by FISH,
is shaded.


c752h8


.j c948d8
71 c928cl


c784c6


HSC7EI 351


- ---I
c742e10


c805c8


HSC7E31289


J c813f10


I-


SI I II


E m I


I mon "sm" "'mom N\\\% Amim, ommomm's ,, I MOMMOMMEN,


I


I













SLOS Translocation Region


I r


N N
M B BBBN BM
IHSC7E476


BN
HSC7E122 I I ZI I


M B BN
c752h8 11 1 I I I I


I MB
B N M B
BNI M BI HSC7E261
I
MB
HSC7E244
MB
B BN B N


HSC7EI 289


I I


153a8 (5.0 kb)


PAX4


Scale: in kb

I I I I I
250 500 750 1000




Figure 3.3. Physical map of the translocation region at 7q32. The YACs (open and stippled boxes) are drawn to
scale and show the rare-restriction sites established from PFGE analysis: B=BssHII, M=MIuI, and N=NotI. The
cosmid 153a8 (horizontally striped box) and the PAX4 gene (diagonally striped box) were placed on the map by
hybridization against Southern blots of the YACs. The complete distribution of the GRM8 gene on HSC7E1289 is
currently unknown and is thus not indicated on the map. 153a8 and PAX4 are not to scale.


7 cen


7 qter












mapping of a partial GRM8 cDNA to the centromeric end of HSC7E1289.

Southern blot analysis of 7Q32 YACs with full-length GRM8 cDNA.

Hybridization of the 7q32 YAC Southern blot with the full-length mGluR8 cDNA

showed hybridization across a large span of the YAC contig (Figure 3.4). The
lane containing HSC7E1289 DNA showed at least 6 EcoRI bands ranging from 3

to 12 kb in size. More interesting, though, was the detection of bands in YAC

clones known to fall proximal (HSC7E1326 and HSC7E1351) and distal

(HSC7E244 and HSC7E122) to the translocation based on FISH studies. Several

of the CEPH-G~n~thon YACs, both centromeric and telomeric in the contig, had

similar banding patterns as those seen in the YACs from the FISH studies. Two

CEPH-Gn6thon YACs (c963f9 and c897c6) from outside the 7q32 region served

as controls for yeast cross-hybridization, YAC vector hybridization, or

nonspecific hybridization. The Southern analysis suggested that the GRM8 gene

contained sequences located on both sides of UF53's translocation. The mGluR8

cDNA was used to screen a chromosome 7 cosmid library from which 16 cosmids

were retrieved. These GRM8 cosmids were positioned on the contig by Southern

blot analysis.
Localization of unique sequence fragments from GRM8 cosmids on the

7q32 YAC contig. Hybridization of the unique sequence fragments against YAC

contig Southern blots placed all of the cosmids within 1 Mb on the contig.

Relative placement was based on the presence or absence of a hybridization

signal(s) in overlapping YACs (demonstrated in Figure 3.5). A hybridization

signal of the same size, present in all lanes, was concluded to be cross

hybridization of the residual cosmid vector sequences in the probe to the YAC

vector. This crude localization also established that all of the cosmids were

contained within HSC7E1289, the YAC spanning UF53's translocation. One

cosmid, 153a8, was localized near the telomeric end of HSC7E1289 while the



























.f ,qv .XJ / $ ,:"


12 kb -
7 kb -
5 kb -
4 kb -
3 kb -


Figure 3.4. Southern blot analysis of the YACs from the 7q32 contig hybridized
with the full-length inGluR8 cDNA to determine the genomic span of GRM8. The
Southern blot was composed of the (HSC7E)-YACs used in the FISH studies and
CEPH YACs which overlapped with HSC7E1289, with the exception of c963f9
and c897c6 which served as controls. Fragment sizes indicated to the left of the
panel were based on Gibco-BRL's I kb ladder.
































0 -5kb




Figure 3.5. Southern blot analysis of the YACs from the 7q32 contig hybridized
with a 5.0 kb EcoRI fragment from the GRM8 cosmid 153a8. The 5.0 kb fragment
was chosen based on its lack of repetitve sequences. This experiment indicated
the relative location of this cosmid on the YAC contig.












most centromeric cosmids, 13e6 and 192d4, were localized approximately 1 Mb
from 153a8 (Figure 3.6). Since 153a8 was localized to the same region of

HSC7E1289 as the CpG island, it was further analyzed for the presence of a CpG

island.
Identification of a CpG island in the GRM8 cosmid 153a8. The double

digests of the GRM8 cosmids with EcoRI and BssHHI, MluI, or NotI established

that only one cosmid contained a putative CpG island. At least two BssHl sites,

two MluI sites, and one NotI site were found to be contained within cosmid

153a8 (Figure 3.7). The relative placement of 153a8 on the YAC contig and the

existence of a CpG island in this cosmid strongly suggested that the CpG island

on the distal end of HSC7E1289 and the CpG island contained in 153a8 were one

and the same. Furthermore, 153a8's hybridization to the mGluR8 cDNA

suggested that this CpG island likely represented the 5' end of the GRM8 gene.


Discussion
Following the identification of YAC HSC7E1289 spanning UF53's

translocation, experiments were undertaken to further narrow the translocation

breakpoint and to identify genes in the region. Since the rate of progress and

success of any one positional cloning approach was unpredictable, several

approaches were concurrently pursued to insure that cDNAs would ultimately

be identified. The direct selection of cDNAs from HSC7E1326 was an

unsuccessful approach. Approximately 200 clones were recovered from the

dire selection of HSC7E1326. Of these 200, I isolated fifty of the inserts which

were hybridized against the YAC Southern blots, and found that approximately

80% of the clones were false positives containing repetitive sequence (data not

shown). It was predicted that 70-140 unique sequence, direct-selected cDNAs

should have been isolated from HSC7E1326 based on its 700 kb size











HSC7E1289





cen


If) 1
r-4~


I II I


$
cos 192d4
cos 13e6


5
cos 81g2
cos 15317
cos 150h8
cos 245h5
cos 241h6


I I


I
cos 153a8




cos 85b12
cos 85d12
cos 15a5
cos 41e10
cos 41g9


Figure 3.6. Schematic representation of location of thirteen of the GRM8 cosmids along HSC7E1289. The arrows
indicate the relative positions of the cosmids with respect to the DNA markers (denoted by the black vertical bars).
The positions of the cosmids were established by Southern blot analysis of YACs from the 7q32 contig. For
simplicity, only HSC7E1289 is included in this illustration. Cosmid 153a8 cross hybridized with the 5' end of the
mGluR8 cDNA, suggesting that the 5' end of GRM8 lies toward the telomere.


I tel


I



















1 2 3 4 5


1. 1 kb ladder
2. EcoRI
3. EcoRI and BssHII
4. EcoRI and Mlul
5. EcoRI and NotI


Cosmid 153a8


Figure 3.7. Restriction digests of the GRM8 cosmid 153a8. Cosmid 153a8 was
digested with EcoRI alone (lane 2) and double digested with EcoRI and BssHII,
MiuI, or NotI (lanes 3-5). Ethidium bromide-staining of the 1% agarose gel
revealed the loss of several EcoRI bands and the subsequent appearance of new
smalIer bands as a result of the double digestion with the above rare-restriction
endonucleases. The molecular weight marker is Gibco-BRL's 1 kb ladder
(lane 1).











(Rommens et al., 1994). Due to the lack of unique cDNA clones in this

experiment, it was suggested that HSC7E1326 was rich in repetitive sequences

(Dr. Joanna Rommens, Toronto, personal communication). Further analysis of

the few unique clones was abandoned, as well as direct selection using other

7q32 YACs, after the discovery that GRM8 was putatively interrupted in UF53

(discussed below).

Throughout the human genome, the cytosines in CpG dinucleotides are
typically methylated, with the exception of a small percentage of unmethylated

CpG's which occur in discrete "islands" (Bird et al., 1987). These islands, now

known as CpG islands, are defined as regions of DNA, usually more than 200 bp
long, with a GC-content greater than 50% (Larsen et al., 1992). The number of

CpG islands in the haploid human genome is estimated to be approximately

30,000 (Bird, 1986). In studies by Larsen et al. in 1992, 50% of the genes analyzed

were associated with CpG islands, with a majority of this fraction representing

housekeeping and other ubiquitously expressed genes. The CpG islands
associated with these particular genes began 5' to the transcription initiation site,

included one or more exons and introns, and usually included the promoters.

Additionally, 40% of tissue-specific expressed genes and other genes with limited

expression were associated with CpG islands. In these cases, CpG islands were
not biased to the 5' end of the transcription unit. Some CpG islands were shown
to encompass bi-directional promoters, implying that a single CpG island could

be associated with two genes (Poschl et al., 1988).
Physical mapping of the YAC contig using PFGE and rare-restriction

endonucleases established the presence of two CpG islands. The first CpG island

was located on the telomeric end of HSC7E1289, while the other was found
approximately 0.5 Mb distal on the contig. As shown in Figure 3.3, the second

CpG island corresponded to the PAX4 gene, which was placed on the map by












hybridization of YAC Southern blots with a mouse cDNA clone, provided by

Peter Gruss, Max Planck Institute, Germany (Dr. Steve Scherer, personal

communication). Of greater importance, though, was the presence of the CpG

island on the telomeric end of HSC7E1289, which lay distal to the breakpoint.

Since CpG islands are almost always associated with genes, the presence of the

CpG island on distal HSC7E1289 implied that a gene was in close proximity.

The most pivotal experiment performed during my visit to Dr. Tsui's

laboratory was the Southern blot analysis of the 7q32 YACs with the full-length

mouse mGluR8 cDNA. Initially, GRM8 was not viewed as a potential candidate

gene based on the fact that it had no apparent connection with cholesterol

metabolism and appeared to map quite proximal to the translocation breakpoint.

But since the full span of the gene was not known, I convinced the Tsui

laboratory that we needed to pursue full mapping of this gene. Thus, I

hybridized the full-length mGluR8 cDNA against the YAC Southern blots to test

this gene's distribution on the contig. Fortuitously, the rnGluR8 cDNA

hybridized to YACs known to fall proximal and distal to UF53's translocation.

The Southern blot patterns observed in HSC7E1289 and the overlapping YACs

(HSC7E1351, HSC7E1326, HSC7E122, HSC7E244, c784a6, c805c8, c830d4, c744c2,

c813f 10, c938c7, c752h8 and c855h10) indicated that the GRM8 gene was

distributed along the central portion of the contig, and could possibly span the

breakpoint. Although these original data were not completely conclusive,

further investigation into the relative position of the GRM8 gene to the

translocation was warranted.

At the time of these studies, no cDNA or genomic sequence information

existed for the human GRM8 gene. Further experimentation required human

GRM8 cDNA or genomic clones. Thus, as mentioned previously, a human

chromosome 7 cosmid library was screened with the full-length nzGluR8 cDNA












and 16 cosmids retrieved. To quickly place the cosmids on the physical map,
unique sequences were isolated from each cosmid and individually hybridized

against the YAC Southern blots (Figure 3.5). The relative locations of the GRM8
cosmids to HSC7E1289 are represented in Figure 3.6. The cosmids were localized

along a 700 kb section of HSC7E1289 between the markers AFM323wd5 and
AFM218xf1O. The physical mapping of the cosmids, coupled with my earlier

FISH studies tentatively placing UF53's translocation between markers

AFMa125whl and AFM309yfl, strengthened the argument for further

investigation of the GRM8 gene.
Rare restriction site analysis of the GRM8 cosmids shed further light on

the relative location of one cosmid on the physical map, establishing the

existence of a CpG island in the cosmid 153a8 (Figure 3.7). Southern blot analysis

of this particular cosmid placed it on the YAC contig at the same location as the

CpG island established by PFGE (Figure 3.3 and Figure 3.6). Collectively, these

results confirmed that the CpG island found in the YAC contig near the distal

end of HSC7E1289 was also contained in the GRM8 cosmid 153a8. These
preliminary data suggested that the GRM8 gene was most likely disrupted by the

translocation in UF53. Confirmation of these preliminary data would come from

FISH studies with the GRM8 cosmids.

















CHAPTER 4
IDENTIFICATION OF A METABOTROPIC GLUTAMATE RECEPTOR GENE
(GRM8) DIRECTLY DISRUPTED BY UF53'S TRANSLOCATION



Introduction
Since GMR8 became the key candidate gene in the translocation region, it

was important to review the known roles of glutamate receptors in order to

better evaluate GRM8 as a possible SLOS gene. Glutamate not only serves as the

major neurotransmitter in the central nervous system (CNS) of all mammals but

also plays a vital role in neuronal plasticity, as exemplified by long-term

potentiation (LTP) and long-term depression (LTD), and neuronal toxicity

(Nakanishi, 1992; Tanabe et al., 1992). Glutamatergic dysfunction has been

implicated in neuronal cell death following ischemia, hypoglycemia, and anoxia

(Choi, 1992), in epilepsy (Rogers et al., 1994), and in neurodegenerative disorders

(Thomas, 1995).
Two distinct groups of receptors are responsible for mediating the actions

of glutamate: the ionotropic and metabotropic glutamate receptors. Activation of

the ionotropic glutamate receptors, which are coupled to ion channels in the

membrane, results in the entry of calcium and sodium ions into the cell leading

to polarization of the membrane (MacDermott et al., 1986; Murphy et al., 1987;

Thomas, 1995). The ionotropic glutamate receptors are divided into distinct

subtypes based on their pharmacological and electrophysiological selectivity for

N-methyl-D-aspartate (NMDA), alpha-amino-3-hydroxy-5-methyl-isoxasole-4-

propionate (AMPA) and kainate (Monaghan et al., 1989). The metabotropic

glutamate receptors are G-protein-coupled receptors which indirectly mediate


CC%'











intracellular signaling through secondary signaling pathways (Knopfel et al.,

1995; Scherer et al., 1996). To date, eight mGluRs have been identified and

subdivided into three subfamilies based on sequence homology and

pharmacological properties (Duvoisin et al., 1995).
mGluRs were first postulated to exist based on studies demonstrating that

glutamate stimulation resulted in the activation of phospholipase C (PLC) via a
G-protein leading to the accumulation of inositol triphosphate (IP3) in the CNS

(Sladeczek et al., 1985; Sugiyama et al., 1987). The first mGluR was cloned using

a functional expression screening process (Houamed et al., 1991; Masu et al.,
1991). In brief, this screening procedure made use of the secondary signaling

pathway in Xenopus oocytes that link G-protein activation with chloride channel

currents that could be electrophsiologically measured. The oocytes were injected

with pools of total RNA from rat cerebellum and tested for glutamate stimulation
reflected by an oscillatory chloride current response. These pools were then

successively subdivided until a single clone eliciting a response, called mGluR1,
was identified. The remaining family members and several splice variants were

identified by using mGluR1 as a probe or using its sequence to design degenerate

PCR primers to screen additional cDNA libraries (Abe et al., 1992; Duvoisin et al.,
1995; Minakami et al., 1993; Nakajima et al., 1993; Okamoto et al., 1994; Pin et al.,

1992; Tanabe et al., 1992). As mentioned above, the mGluRs are subdivided into

three subgroups based on their sequence homology (Pin and Duvoisin, 1995).
mGluR1 and mGluR5, as well as their splice isoforms, compromise Group-I,

mGluR2 and mGluR3, Group-i, and mGluR4, mGluR6-mGluR8, Group-II.

Within any given group, amino acid sequence homology is around 70%, where

as between groups the homology falls to around 40% (Figure 4.1).



















40 50 60 70 80 90
I I I I I I


mGluR1

mGIuR5

mGluR2

mGluR3

mGluR4

mGluR7

mGluR8


mGIuR6


Figure 4.1. Dendrogram of the mGluR family members (Pin and Duvoisin, 1995).
The number at the top indicates % amino acid identity between members of the
family.











Although the mGluRs have differences in sequence, there are important

predicted structural features which are consistent among the members of this

family. All of the glutamate receptors have an extracellular amino terminus

with a putative signal peptide, seven hydrophobic segments forming seven

transmembrane domains, and an intracellular carboxy terminus as predicted by

hydrophobicity analysis (Sudzak et al., 1994). The conserved regions, illustrated

in Figure 4.2 using mGluRla as an example, are evenly distributed throughout

the length of the receptor. This differentiates the mGluRs from other G-protein-

coupled receptors which show high conversation in the transmembrane regions.

There are twenty-one cysteine residues conserved among the mGluRs; nineteen

of these are located in the extracellular domain and extracellular loops in the

putative ligand binding domain. The third intracellular loop, which is
implicated in the functional coupling of the other receptors to G-proteins

(Luttrell et al., 1993; Maggio et al., 1993; Pin and Duvoisin, 1995), has a highly

conserved amino acid sequence between the different mGluR subtypes. This was

an unexpected observation based on the coupling of these receptors to different

secondary-signaling pathways.

Localization of the mGluR mRNA in the CNS has been strictly based on in

situ hybridization using cDNA clones for the different receptor subtypes.

Hybridization with the mGluR probes demonstrated that the cells in most of the

different brain regions are positive for multiple mGluR subtypes (Sudzak et al.,

1994). For example, hybridization of probes for Group-I, Group-Il, and mGluR4

and mGluR7 (from Group-I), all showed positive mRNA expression in the

cerebral neocortex. In addition to expression in the CNS, mGluR1, mGluR6 and

mGluR8 have detectable expression patterns in the retina (Duvoisin et al., 1995;

Nakanishi, 1994). The heterogeneous mGluR expression distribution in the CNS

suggests that different intracellular responses are generated as a result of





































525


415
360
250 O

i40 195











855


910


1020-


-965

-1075


1185
1199 c xi


Figure 4.2. Schematic representation of the structure of mGluR1 (Huoamed et al.,
1991; Masu et al., 1991). The conserved cysteine residues are represented by

filled circles and the signal peptide is represented by light gray circles. mGluR1
is the largest of all described family members.


C




















0






0


470











the differential expression of mGluR subtypes.
mGluRs have been shown to be directly or indirectly coupled to multiple

secondary-signaling pathways involving the activation of PLC, mobilization of

intracellular stores of calcium, stimulation and inhibition of adenylate cyclase,

and presynaptic inhibition of glutamate release (Nakanishi, 1994; Pin and
Duvoisin, 1995). Group-I receptors directly activate PLC which in turn cleaves
phosphotidylinositol-4,5-bis -phosphate (PIP2) into inositol-1,4,5-triphosphate

(IP3) and diacylglycerol (DAG). These then act as secondary messengers (Jeffrey

Conn et al., 1994), with II'3 causing the release of calcium from intracellular

stores. The calcium ions are then free to interact with a variety of proteins which

stimulate or inhibit various cellular activities. DAG, on the other hand, activates

protein kinase C (PKC) in the presence of calcium. Once activated, PKC affects a

variety of cellular responses involved in cell growth and metabolism. Also, the

accumulation of DAG causes the release of arachodonic acid by a DAG-specific
phospholipase. Both cyclooxygenase and lipoxygenase use arachodonic acid as a

substrate for the production of prostaglandins and leukotrienes, respectively,
which have a variety of cellular effects including the inflammatory cascade.

Glutamate stimulation of Group-il and -III mGluRs induces G-protein-mediated
inhibition of adenylate cyclase (Knopfel et al., 1995). Inhibition of adenylate

cyclase results in the decreased production of cyclic adenosine monophosphate

(cAMP). cAMP plays an essential role in modulations of ion channels, regulation
of synaptic transmission, regulation of gene expression and a multitude of
metabolic functions. Thus, inhibition of cAMP can have a significant impact on

the cellular activities in the CNS.
My preliminary studies implicated a metabotropic glutamate receptor

gene, GRM8, as the putative candidate gene interrupted by the translocation in

UF53. mGluR8 was the newest member of the metabotropic glutamate receptor











family (Duvoisin et al., 1995). The receptor was identified by screening a mouse

retina cDNA library with degenerate primers derived from two conserved amino

acid sequences in the sixth transmembrane region and two PCR fragments

encoding the seventh transmembrane region from mGluR1 and mGluR3

(Houamed et al., 1991; Tanabe et al., 1992). The deduced amino acid sequence

revealed that mGluR8 had the same structural features seen in the other family

members (described above). Assignment of this receptor to the Group-III

subfamily was based on sequence homology to mGluR4, mGluR6, and mGluR7.
In situ hybridization studies in mouse revealed that wide-spread rnGluR8 mRNA

expression was observable throughout the CNS, retina and peripheral nervous

system (PNS) at embryonic day 16. In the adult mouse, expression was limited to

the olfactory bulb in the mitral/tufted cells, which are responsible for

presynaptic regulation of glutamate release, and to the retina, in cell types not yet

determined.

As described previously, the full-length mGluR8 cDNA was used to screen

a human chromosome 7-specific cosmid library (Duvoisin et al., 1995) and the

retrieved cosmids were used in FISH experiments with UF53's metaphase

spreads. The desired goal of this set of experiments was the discovery of cosmid

clones falling both proximal and distal to the translocation, to confirm that the

translocation interrupts GRM8. As described in this chapter, I identified two

such cosmids (one proximal and one distal to the breakpoint) indicating that the

GRM8 gene spans a large genomic region and is interrupted by UF53's

translocation.


Materials and Methods
FISH. The bacterial strains carrying the chromosome 7-specific cosmids

were grown in Luria Broth supplemented with 25 tg/ml kanamycin for 16











hours. The cosmid DNA was purified (Qiagen, Maxi-Prep), and 1 Ig was nick

translated with digoxygenin-11-dUTP according to the Large Fragment Probe

Labeling Kit protocol (Oncor). The digoxygenin-labeled DNA (50 ng) was

ethanol precipitated with 50 ng human Cot-1 DNA (Gibco BRL) and

resuspended in 10 pJ Hybrisol VI (Oncor).

UF53 lymphoblastoid chromosome spreads were freshly prepared by

standard cytogenetic procedures (Yunis, 1976). The labeled cosmid probe DNA

was denatured at 72C for 5 minutes and allowed to reanneal at 37C for 15

minutes prior to hybridization. The chromosomes on the slides were denatured

and hybridized with the probe at 37C for at least 16 hours in a humidifying

chamber. Post-hybridization washes consisted of 1XSSC at 72C for 5 minutes

followed by 1XPBD at room temperature. The chromosomes were incubated

with rhodamine-conjugated rabbit anti-digoxygenin IgG (Boehringer Mannheim)

for 15 minutes at 37C, followed by three 2 minute 1XPBD washes at room

temperature, and counterstained with DAPI/antifade (Oncor) to allow signal

detection and chromosome identification. Subsequent analysis and

documentation of hybridization signals employed an Olympus BHS fluorescence

system and the CytoVision/CytoProbe computer imaging device (Applied

Imaging) with the technical support of Brain Gray and Dr. Roberto Zori at the

R.C. Philips Unit.

PCR Amplification of Exons 1 and 2. Dr. Stephen W. Scherer and Dr.

Lap-Chee Tsui provided the preliminary sequence and primers for GRM8 exons

1 and 2. PCR primers included the 5' end

(5'-ATGTATGCGAGGGAAAG-3') and the 3' end

(5'-ATGTTAGCAAACCATGAT-3') of human GRM8 exon 1, and the 5' end (5'-

ATACCTCAAATTAGGCTACA-3') and the 3' end

(5'-TCACCCCTCGAGAACTGGGTGAAG-3') of GRM8 exon 2. Amplification












was performed with 100 ng of cosmid DNA, 200 ng of each primer, 0.2 mM
dNTPs, 1.5 mM MgCl2, and 0.8 U Taq polymerase in a 50 pfl reaction, under

standard buffer conditions. PCR conditions consisted of 35 cycles of 1 minute at

94C, 1 minute at 65C and 1 minute at 72C, followed by a final 30 minute

extension at 72*C. PCR products were analyzed on a 1% agarose gel and

detected with ethidium bromide staining and UV illumination.

Southern blot analysis of GRM8 cosmids with exon 2. A Southern blot of

the GRM8 cosmids was prepared by digestion of 500 ng of each cosmid with

EcoRI, separation by gel electrophoresis on a 1% agarose gel, and transfer to

Hybond N (Amersham). Twenty-five ng of the 214 bp exon 2 PCR product was

labeled by random-priming with a-32P dCTP (MegaPrime kit, Amersham) and

hybridized against the GRM8 cosmid blot. Hybridization and wash conditions

were based on the Church and Gilbert SDS/sodium phosphate method (Church

and Gilbert, 1984). Bands were visualized by autoradiography.


Results
Identification of an Interrupted Glutamate Receptor Gene using FISH.

Nine cosmid clones (13e6, 15a5, 41e10, 85b12, 109c4, 150h8, 153a8, 153f7, 192d4)

which hybridized to portions of the full-length mouse cDNA were chosen as

probes in FISH experiments to determine their location relative to UF53's

translocation breakpoint. Prior to FISH analysis, the cosmids were positioned

relative to the full-length mGluR8 cDNA, by our collaborators in Toronto, based

on the hybridization of the cosmids to one of three cDNA fragments produced by

digestion of the cDNA with EcoRI, KpnI, and Miul. Of the sixteen cosmids, only

one, 153a8, showed specificity for the 5' end of the nGluR8 cDNA while the

others contained sequences from the middle of the cDNA or the 3' end.

FISH analysis of the 9 cosmids demonstrated that none spanned the











translocation breakpoint; all of the cosmids were localized proximal to the

breakpoint with the exception of 153a8 (summarized in Table 4.1). This cosmid,

specific for the 5' end of the cDNA and known to contain a CpG island (see
previous chapter), was found to lie directly distal to the breakpoint (Figure 4.3).

These results indicated that the GRM8 gene was disrupted by UF53's

translocation. Also, combined with previous mapping of the 3' end of GRM8 to

the centromeric end of HSC7E1289 and physical mapping of the region (Alley et

al., 1996), my results indicated that the GRM8 gene spans at least 700 kb in the

geno-ne.
PCR of Exons I and 2. To localize the site of the GRM8 disruption, PCR

amplification of individual exons from cosniid DNA established that exons 1 and

2 were contained within cosmids 153a8 and 85b12 respectively (Figure 4.4).
These results confirmed the interruption of the GRM8 gene and established that

the translocation breakpoint was in intron 1. Based on the previous experiments

which physically mapped these cosmids on the YAC contig (Alley et al., 1996),

the intron was estimated to be at least 100 kb in size. Additionally, Southern blot

analysis of the GRM8 cosmids confirmed that exon 2 was contained in 85b12 as

well as four other cosmids (15a5, 85b12, 41e10, and 41h9). These four cosmids,

which had previously mapped to the same location in the 7q32 contig (Figure

3.5), were confirmed to overlap 85b12.


Discussion
GRM8 was one of two members of the gene family encoding the
metabotropic glutamate receptors which had been mapped to human

chromosome 7; GRM3 had mapped to the 7q21.1-7q21.2 region (Scherer et al.,

1996). Originally, a partial cDNA of GRM8 was mapped to the centromeric end

of the YAC HSC7E1289 (Scherer et al., 1996) that spans the UF53 breakpoint































Table 4.1 Summary of FISH Experiments with GRM8 Cosmids


Cosmid clones
13e6
15a5
41el0
85b12
109c4
150h8
153a8
153f7
19284


Location relative to
the translocation breakpoint
proximal
proximal
proximal
proximal
proximal
proximal
distal
proximal
nroximal




















































Figue 4.3. FISH analysis of the UF53 translocation region using cosmid clones
containing the GRM8 coding sequence. Digoxygenin-labeled cosmids (pink)
were hybridized to metaphase chromosomes under competition with human
Cot-1. UF53 metaphase spreads after hybridization with the cosmid clones (a)
153a8 and (b) 85b12 demonstrated distal and proximal positioning of probes,
respectively. Chromosomes were counterstained with DAPI.



























cs.I -f





510 bp -2 4 b



Exon 1 Exon 2



Figure 4.4. PCR amplification of GRM8 exons 1 and 2. Cosmid clones 153a8 and
85b12 and the full-length rnGlitR8 cDNA were tested for the presence of exon 1
(left-side panel) or exon 2 (right-side panel) by PCR amplification. PCR primers
were specific for the 5' and the 3' ends of exons 1 and 2. The sizes of the products
are indicated at the side of each panel.












(Alley et al., 1995). With my subsequent experiments, I demonstrated that GRM8

is directly interrupted by UF53's translocation and that the breakpoint lies within

intron 1. To fit the autosomal recessive inheritance of SLOS and provide further

evidence for GRM8 as an SLOS candidate gene, demonstration of a mutation in

the other UF53 GRM8 allele would have been required. However, mutational

analysis of individual exons was not feasible at the time since the gene structure

had not been fully elucidated, and mRNA from appropriate tissue (brain and

retina) was not available for mutation screening since UF53 was deceased and

only skin fibroblasts and lymphoblasts were available.

The CpG island I discovered in HSC7E1289 and the cosmid 153a8

corresponds to the 5' end of GRM8. Physical mapping of GRM8 suggests that the

genomic size of this gene is 700-1300 kb. Figure 4.5 illustrates the relative

position and orientation of the GRM8 gene with respect to HSC7E1289.

Preliminary data suggests that the GRM8 transcript is approximately 2.7 kb in

size, indicating that the vast majority of the genomic sequence is represented by

the introns (Dr. Stephen Scherer, personal communication). Intron 1 is estimated

to be at least 100 kb in length, and therefore the interval containing the

translocation breakpoint location remains quite large. Because there is no known

connection between the glutamate receptor and cholesterol biosynthesis, and

further mutational studies in UF53 are not yet possible, the role of GRM8 in

SLOS, if any, is not clear. I propose several hypotheses about the involvement of

GRM8 in SLOS. First, GRM8 may be the affected SLOS gene in our patient

through a biochemical mechanism not yet understood, with a mutation existing

on the other allele. Since genetic heterogeneity might play a role in SLOS, GRM8

may or may not be the affected gene in other SLOS families. Secondly, due to the

large size of the intron 1, one could speculate that there are other functional

genes embedded within this GRM8 intron, as seen in NF1 and the choline




















HSC7E1289


31


exon 2 exon I
=zzq-qiZz4 5' GRM8


Translocation
Region





Figure 4.5. Illustration of the relative position and orientation of the GRM8 gene on YAC HSC7E1289. Exons 1 and
2 were positioned based on experiments which placed 153a8 and 85b12 distal and proximal, respectively, to the
DNA marker AFM309yfl. The sizes of the untranslated 5' and 3' ends and sizes of the introns are not known. The
estimated minimal distance between AFM218xf1O and AFMa065zg9 is 700 kb, with maximal size estimated at 1,100
kb, based on PFGE and YAC overlap data.


I I I I I Ia I I










transferase gene (Bejanin et al., 1994; Cawthon et al., 1990); one of these could be
the actual SLOS gene functionally disrupted by the translocation. Third, it is

possible that GRM8 is not involved in SLOS and was coincidentally interrupted
by the translocation in UF53. Forth, UF53 may have SLOS due to mutations at

another locus, with no mutation on the other GRM8 allele. However, part of her
clinical picture may be attributed to her being functionally hemizygous at GRM8

- the effect of this genotype is unknown. Further analysis of UF53 and other

SLOS patients will be necessary to test each of these hypotheses.

Of additional interest is the linkage of an autosomal dominant form of
retinitis pigmentosa (RP) to the 7q32 region. RP collectively refers to a group of

degenerative retinal disorders that affect the photoreceptors and retinal pigment

epithelium, resulting in night blindness and loss of peripheral vision of early
adulthood (Heckenlively et al., 1988). RP is a genetically heterogeneous disease

with forms including autosomal dominant, autosomal recessive, and X-linked
modes of inheritance (Humphries, 1993). The autosomal dominant form had

been associated with eight different loci including the RPIO locus linked to 7q in

two large unrelated RP families (Jordan et al., 1993; McGuire et al., 1995). The
linkage data from the two families established that the RPIO gene was flanked by

the markers D7S686 and D7S530 which are separated by 5 cM (McGuire et al.,

1996)'. These two markers were mapped on a 5 Mb interval in a large YAC
contig, containing the 7q32 SLOS YAC contig. The highest combined lod score,

13.08, was associated with marker D7S514. This is the same marker flanked by

the two GRM8 cosmids, 153a8 and 85b12, which flank the translocation. GRM8

seems a particularly interesting candidate gene for RP1O based on its reported

expression in the retina of adult mice (Duvoisin et al., 1995). I propose that one
theory is that a defect in GRM8 is responsible for both RP10 and SLOS. Although

there are no reports of SLOS patients with RP, it is interesting to speculate that








66


the loss of a single GRM8 allele results in RP10, an autosomal dominant disorder,

while the loss of both GRM8 alleles results in SLOS, an autosomal recessive

disorder. Given that many SLOS patients do not live beyond early childhood
and therefore would not develop RP (regardless of gene mutations), this theory

and the role of GRM8 in RP10 and SLOS merits further investigation.

















CHAPTER 5
MICROSATELLITE TYPING AND SOUTHERN BLOT ANALYSIS


Introduction

The interruption of GRM8 by UF53's translocation is the first discovery of

a gene disrupted in an SLOS patient. Unfortunately, without the full GRM8 gene

structure, it is impossible to test the other allele in UF53 for mutations. To
further investigate the possible relationship between GRM8 and SLOS,

microsatellite genotyping was performed on a large SLOS family with

rmicrosatellite markers in the 7q32 region. In addition, to try to more closely map

the translocation breakpoint, probes from exon I and 2 were designed for

Southern blot analysis of UF53's genomic DNA.
Microsatellite Genotyping. The variable expression seen in both the

clinical and biochemical presentation of SLOS suggests genetic heterogeneity.

Genetic heterogeneity refers to the notion that different mutations can cause
similar phenotypes. Genetic heterogeneity is subdivided into allelic

heterogeneity (different mutations at the same locus) and locus heterogeneity

(mutations at different loci). Several reports support the theory of locus

heterogeneity in SLOS. Berry et al. reported a family with both SLOS and Miller-
Dieker syndrome segregating independently; no cholesterol findings were
reported (Berry et al., 1989). The children all carried unbalanced karyotypes

which arose from meiosis of one parent's balanced translocation

[t(7,17)(q34;p13.1)1, implying that an SLOS locus may lie in the 7q34-qter region

or in the 17p13.1-pter region. An individual with severe SLOS features, low











serum cholesterol, and normal concentrations of 7-DHC, was reported with an

interstitial deletion at 17p13 related to a balanced paracentric inversion in the

father [inv(17)(pll.2p13)], implicating proximal chromosome 17p in non-typical

SLOS (Yang et al., 1994). Additionally, there was a report of three sisters having
SLOS-like features with a small terminal deletion of chromosome 4 (Hill et al.,
1991). Although these patients did not have the classic SLOS clinical and/or

biochemical phenotype (in some cases the tests were not done), is it possible that

recessive mutations at several genetic loci can independently result in SLOS?
To address the question of heterogeneity, I chose to genotype the

microsatellite markers in the 7q32 translocation region to determine if there was

any co-inheritance between the loci and SLOS in other families. SLOS families

with one or more affected siblings were collected through various clinical
collaborations. The most extensive SLOS family (Figure 5.4), consisting of
multiple affected siblings as well as affected first cousins, was provided by Dr.

Ngozi Nwokoro (Cleft Palate Craniofacial Center, University of Pittsburgh).

Microsatellite analysis had two possible outcomes: 1. identical genotypes among

affected siblings with no unaffected siblings having the same genotype, thus

indicating consistent inheritance of SLOS and this region, or 2. a variation of the

above rules, thereby ruling out that the SLOS locus in that family resides with the

GRM8 region. If all of the markers were consistent with an SLOS locus in that

region, further investigation, such as GRM8 mutational analysis, would be

warranted in these families.
Microsatellites are defined as a group of short tandemly repeated
sequences that are found throughout the human genome (Hughes, 1993).
Commonly, microsatellite repeats are composed of the dinucleotide (CA)n,

which is particularly abundant in the human genome. It is estimated that as
many as 50,000 microsatellites are found throughout the genome with the unit of











repeat containing from 10-60 copies of the dinucleotide sequence (Jeffreys et al.,

1990% Weber and May, 1989). These highly polymorphic repeats serve as

extremely useful markers for the mapping of disease genes and overall genetic

mapping of the genome. Microsatellite markers are typed by amplification of
the repeat unit using PCR primers specific for the microsatellite of interest, and

analysis of the PCR products on a denaturing polyacrylamide sequencing gel as

illustrated in Figure 5.1.

Southern blot analysis. Since the translocation lay in such a large intron, I

could not rule out disruption of an embedded gene other than GRM8 as the SLOS
associated event. Localization of the breakpoint in close proximity to either exon

1 or 2 would lend more support for GRM8's role than an internal gene. In an

attempt to detect the translocation by Southern blot analysis, I generated PCR
fragments from exons I and 2 to use as probes against Southern blots of UF53

genomic DNA. Repeated failure of these small probes to hybridize led to my

characterization of additional sequence outside of the open reading frame by

direct sequencing of cosmids. Sequencing directly from cosmids is a challenging
task since success is dependent upon the integrity of the large cosmid DNA

molecule. To insure that the cosmid DNA was greater than 85% supercoiled, I
purified the DNA with the NucleobondTM AX 500 Kit (Macherey-Nagel). I

designed internal sequencing primers from exon 1 and 2 which would allow for

the generation of sequence 5' to each of the exons. I obtained approximately 300
additional bp of sequence in the 5' untranslated region of exon 1, but no
sequence was obtained 5' of exon 2 despite numerous attempts. From the new

sequence in exon 1, 1 designed primers which amplify a 613 bp fragment
containing 5' untranslated sequence and open reading frame (Figure 5.2).

Unfortunately, hybridization of the 613 bp fragment against the UF53 Southern
blot resulted in nonspecific repetitive hybridization which I could not eliminate












Microsatellite. Typing


Mother

n=60
n=35 -.


Father
.o separation 60
-mm n =50
50
n =60
35
Child
Lk -17- n=35

IEEE n=60



Figure 5.1. Schematic of microsatellite typing. Microsatellite analysis is performed using PCR primers specific for
each microsatellite marker. One of the two primers (indicated by arrows on the left panel) is end-labeled with
y32p dATP using kinase (indicated by the *). The patient's DNA is then subjected to PCR using these primers and
the PCR products are separated on a 6% denaturing polyacrylamide gel. The genotype of each individual can then
be determined from autoradiography of the gel (right-side panel). In the example illustrated here, the mother has
one allele with 60 repeats and the other with 35 repeats, while the father has alleles of 50 and 60 repeats. Thus, the
child inherited the allele with 35 repeats from the mother and the allele with 60 repeats from the father.













5'-caggaatat tctgctacaaggctgatttcaaggacatgaattgttgacc

tcatc cca a catca gaacc t ca gatgttc ta a tttttgc acc attcc a ggc a

a gttg at ctta ta a gga aat a a a attg a a c ctta gg g g tct gat gga a att

cactgtgacattcaaatcaagaaaacttgctaatgcccacagagccttttcc

ccatgg g c c ctgatggta g c c t c ca g aa ggt g c a g c c tc ag gtggtgc c c

tttc ttct gtgg ca a ga a ta a a ctttgg g tct tgg a ttg c a a t ac ca c c tg t

ggagaaaATGGTATGCGAGGGAAAGCGATCAGCCTCTT

GCCCTTGTTTCTTCCTCTTGACCGCCAAGTTCTACTGG

ATCCTCACAATGATGCAAAGAACTCACAGCCAGAGTA

TGCCCATTCCATACGGGTGGATGGGGA CATTATTTTG

GGGGGTCTCTTCCCTGTCCACGCAAAGGGAGAGAGAG

GGGTGCCTTGTGGGGAG*CTGAAGAAGGAAAAGGGAT

TCACAGACTGGAGGCCATGCTTTATGCAATTGACCAG

TTAACAAGGCCCTGTCTCCTTTTCCAACATCACTCTGG

GTGTCCGCATCCTCGACACGTGCTTCTAGGGACACC
ATGCTTTGGAGCAGTCACTAACAATCGTGCAGGCATT

AATAGAGAAAGATGCTTCGGATGTGAAGTGTGCTAAT

GGAGATCCACCCATTATTCACCAAGCCCGACAAGATT

TCTGGCGTCATAGGTGCTGCAGCAAGCTCGTGTCCAT

CATGGTT GC TAACAT TTTAAGACTTTTTAAGATAC-3'




Figure 5.2. Partial nucleotide sequence of GRM8 exon 1. The 5' untranslated
region and open reading frame are shown in lower-case and upper-case,
respectively. PCR primers were designed from the underlined sequence (arrows
indicate orientation). The translational start site, ATG, is italicized and the AluI
restriction site is shown in bold type with the asterisk representing the cleavage
site.











with Cot-I competition. I searched the sequence for restriction sites to generate

smaller fragments that might have a less repetitive nature. An AluI restriction

site was located toward the 3' end of the fragment, which produced two

fragments of 460 bp and 153 bp in size (Figure 5.2). The 460 bp fragment

successfully hybridized against the Southern blot of UF53, and no gross DNA

abnormalities were detected.


Materials and Methods

Isolation of genomic DNA from patient leukocytes or fibroblasts. Patient

blood or fibroblast cell line samples were acquired from the following clinicians:

Dr. Mira Irons (New England Medical Center, Boston), Dr. Ngozi Nwokoro

(Cleft Palate Craniofacial Center, University of Pittsburgh), Dr. Laura Keppen (

University of South Dakota School of Medicine, Sioux Falls), Dr. Richard Kelley

(Kennedy Krieger Institute, Baltimore) and Dr. Dianne Abuelo (Rhode Island

Hospital, Providence). DNA was isolated from the leukocytes following basic

procedures as follows (Madisen et al., 1987). Whole blood samples were

separated by centrifugation in EDTA vacutainer tubes at 3000 rpm for 10

minutes. The buffy coat containing the leukocytes was transferred to a sterile 50

cc conical tube. Forty-five ml of RBC lysis solution (0.15 M ammonium chloride,

0.02 M Tris, pH 7.65) was added, mixed, and incubated at 37C for several
minutes to lyse red blood cells. Leukocytes were pelleted by centrifugation at

2000 rpm for 5 minutes. The supernatant was poured off and the pellet gently

washed with 1XPBS (phosphate-buffered saline). The pellet was resuspended in

5 ml high TE (100 mM Tris, pH 8.0; 40 mM EDTA, pH 8.0) and 5 ml Lysis

Solution (high TE with 0.2% SDS) were added to lyse the leukocytes. The DNA

was phenol/chloroform extracted, isopropanol precipitated, and resuspended in

approximately 1 ml of TE. DNA concentrations were determined by











fluorometry. For fibroblasts, these were grown in culture under our standard

conditions (Alley et al., 1995) and the cells harvested by trypsinization. The cells

were washed in PBS and lysed as the leukocytes above, with the rest of the

procedure being the same.

Microsatellite genotvping of patient samples. Three microsatellite
markers (AFM218xf10, AFM309fyl, and AFMa125whl) from the translocation

region were used to study the SLOS families. One primer of each set was end-

labeled with y-32P dATP using T4 polynucleotide kinase (Sambrook et al., 1989).

The PCR reactions were performed with 25-50 ng genomic DNA, 200 mM of each

dNTP, 40 ng of cold primer (forward and reverse), 6 ng of end-labeled primer,

1.0 units of Taq polymerase, using standard buffer conditions in a total volume of
20 jil. PCR was performed using a Perkin Elmer Cetus 480 thermal cycler with

the following conditions: 2 minutes hot start at 94C, then 30 cycles of 94C for 40

seconds, 62C for 40 seconds, and 72*C for 40 seconds followed by an extension at
72C for 5 minutes. A 10 1 sample of each reaction plus 10 i of stop buffer was

heat denatured and electrophoresed on a 6% denaturing polyacrylamide gel.

The bands were visualized by autoradiography.

Sequence of 5' untranslated region of exon 1. Cosmid 153a8 was grown as

previously described and the DNA purified according to the protocol for the

NucleobondT AX 500 Kit (Macherey-Nagel) to insure that greater than 85% of
the recovered DNA was in its supercoiled configuration. One jig of 153a8 cosmid
DNA was prepared for sequence analysis according to the protocol for the Taq

DyeDeoxyTm Terminator Cycle Sequencing Kit (Applied Biosystems). The

sequencing primer (5'-TGCATCATTGTGAGGATCCAGTAG-3') was designed
from internal sequence in the open reading frame of exon 1. The samples were

then subjected to cycle sequencing using the Perkin Elmer Cetus model 480

thermal cycler with the following conditions: hot start at 96C followed by 25











cycles of 30 seconds at 96C, 15 seconds at 50'C, and 4 minutes at 601C. The

sequencing reactions were ethanol precipitated, dried, and the DNA

resuspended. Sequence was analyzed on the ABI Automated Sequencer 373A

through the Center for Mammalian Genetics.
Southern blot analysis of UF53. A 613 bp genomic fragment from GRM8

exon 1 was amplified from cosmid 153a8 using the above conditions. The
primers used were (5'-TCATCCCAACATCAGAACCT-3'), in the 5' untranslated

region, and (5'-GTTAGTGACTGCTCCAAAGCATA-3'), in the open reading

frame (Figure 5.2). The 613 bp product, containing 278 bp of the 5' untranslated
region, was then subjected to digestion with the restriction endonuclease Alul, to

produce a 460 bp and 153 bp fragment. The 460 bp fragment (the 5' end) was

purified according to the QiaQuick protocol (Qiagen,) and labeled by random-

priming with alpha-32P dCTP (Mega Prime kit, Amersham). For the Southern

blot analysis, UF53's DNA was independently digested with the restriction

enzymes EcoRI, HindII, and PstI, subjected to agarose gel electrophoresis,

transferred to Hybond N (Amersham), and hybridized as previously described

(Wallace et al., 1990). The wash conditions were based the Church and Gilbert
SDS/sodium phosphate method at 60'C (Church and Gilbert, 1984). Signals

were detected with autoradiography.


Results
Microsatellite genotyping. To demonstrate coinheritance of SLOS and the
7q32 markers in an SLOS family, affected siblings would have to display

identical genotypes while unaffected siblings must have genotypes unlike their

affected siblings. The results of the microsatellite genotyping revealed that the

Nwokoro family was inconsistent for coinheritance of the 7q32 markers and

SLOS (Figure 5.3 and 5.4). UF339, affected, and UF448, unaffected, are sisters in




























AFM309yfl




1 S -1 IS

3






Figure 5.3. Autoradiograph of microsatellite typing of AFM309yf I in the SLOS
patients/family members. The numbers indicate the arbitrary allele name
assignments (exact sizes undetermined).





















UF320


4,4 1,4 1,4
2,3 3,4 1,3




UF339 UF448 UF44b IUF44A



1,2 1,1 2,2 1,1
4,4 4,4 1,1 1,3
2,3 2,3 -,- 1,3


Figure 5.4. Pedigree of SLOS family from Dr. Ngozi Nwokoro. DNA has been obtained from individuals with UF
numbers. Half-shaded symbols represent obligate carriers, fully-shaded symbols represent affected individuals.
Genotypes determined by microsatellite typing of AFMa125whl, AFM309yfl, and AFM218fx1O are shown below
each individual. It is unknown which individual in the top generation was the carrier.


UF337


UF338


UF321











the same nuclear family displaying the same genotypes for the markers

AFM.309yfl and AFM218xf1O. Also, UF443 and UF444, who are both affected

and are first cousins to UF339 and UF448, display different genotypes for the

markers AFMaI25whl and AFM309yfl. This clearly rules out the GRM8 region

as having the Nwokoro family SLOS gene, with virtual certainty.

Southern blot analysis. Southern blot analysis of UF53 was performed

using a genomic probe derived from exon 1 to determine if the translocation

breakpoint lay near exon 1, and to assay for any additional gross DNA

rearrangements. The 460 bp probe detected 8-10 kb EcoRI and HindII fragments

and a 6.0 kb PstI fragment (Figure 5.5). No abnormalities in band size or

intensity were apparent with EcoRI and HindIlI. Although the UF53 PstI band

appears to be less intense than the control, ethidium bromide staining of the gel

revealed that this lane was under-loaded with respect to the control (data not

shown). Therefore, the results suggest that the breakpoint does not lie within the

vicinity of exon 1 and no obvious rearrangements lie in this region on the other

allele.

Discussion

Microsatellite genotyping. Microsatellite genotyping of the three markers,

AFMa125whI, AFM309yfl, and AFM218xf1O was extended to all patient samples

including the Nwokoro family, UF53 and parents, and various relatives and

patients of other SLOS families (Figure 5.3). However, since most of the families

were incomplete or I only had DNA from one child, there were no conclusive

results from the genotyping except for the Nwokoro family. The genotyping of

UF53's family, while not pertinent to this study, was important for future genetic

studies since it showed that the genotypes were consistent with correct paternity.

The microsatellite genotyping demonstrated that the Nwokoro family did not fit

the model for coinheritance of SLOS and the 7q32 markers, and I can reasonably


























HindIII
o V
4Ax


10 kb- o-
8 kb -
6 kb -


Figure 5.5. Autoradiograph of Southern blot analysis of EcoRI digests of UF53
and control DNA with 460 bp PCR fragment probe from GRM8 exon 1.
The sizes, indicated to the left, were estimated from a 1 kb molecular
weight marker.


PstI
C9Ox


EcoRI











conclude that GRM8 is not the SLOS gene in the Nwokoro family. If GRM8 is

ultimately confirmed as the SLOS gene in the UF53 family, the lack of association

between SLOS in the Nwokoro family and the 7q32 locus would support the

theory of genetic locus heterogeneity.
Southern blot analysis. Southern blot analysis with exon 1 did not show

any abnormal banding patterns in UF53, ruling out rearrangements at the 5' end

of intron 1. Therefore, the translocation region is still considerably large and

further mapping of the intron would be necessary to determine its precise

location. This would be pertinent to pursue the possibility of an embedded gene

playing a role.

To determine if any of the rare-restriction sites contained within cosmid

153a8 were in the known sequence of exon I of GRM8, a computer search for

Bss-II, MluI, and NotI restriction sites in the sequence was conducted (Alley et

al., 1996). The search yielded no rare-restriction sites for the endonucleases,

suggesting that the CpG island is outside this region, most likely further

upstream based on the CpG island literature.

















CHAPTER 6
CONCLUSIONS AND FUTURE DIRECTIONS

Conclusions. The aim of my studies was to examine the UF53

translocation breakpoint region to determine if any genes were disrupted,

thereby identifying potential candidate genes for SLOS. My work has lead to the

identification of the first candidate SLOS gene disrupted at the molecular level.

As described in Chapter 2, using FISH analysis of YACs, I identified a
1,800 kb YAC, called HSC7E1289, that spans UF53's translocation breakpoint.
The translocation, which was originally mapped to the cytogenetic band of

7q32.1, was resolved to a megabase level with the localization of this YAC. I

further narrowed this region by FISH analysis of YACs known to overlap
HSC7E1289. The results mapped HSC7E1351-and HSC7E244 proximal and

distal to the translocation, respectively, placing the breakpoint tentatively
between markers AFMa125whl and AFM309yf1. Although these markers are

separated by approximately 200-300 kb, the translocation breakpoint could not

be excluded from a larger region of HSC7E1289, since the probes generated from

the YACs were fragments derived from Alu-PCR. The distribution and

orientation of the Alu elements was not known for the YACs and the fragments

may not have represented the entire YAC. Therefore, a conservative estimate of
1 Mb for the translocation breakpoint region was proposed.

In Chapter 3, 1 discussed the construction of a physical map of the YAC
clones (HSC7E1289, HSC7E122, c752c8, HSC7E476, HSC7E261 and HSC7E244)

surrounding the translocation breakpoint, by PFGE using restriction











endonucleases that recognized rare-restriction sites. The rare-restriction site

mapping established the presence of two CpG islands in the contig. One CpG

island was located on the telomeric end of HSC7E1289, and the other was

approximately 500 kb distal. The latter CpG island was found to correspond to

the PAX4 gene, which had previously been mapped to the same region by

Southern blot analysis of the YAC clones with a PAX4 mouse cDNA clone. An

intriguing aspect of PAX4 is that this gene (and all other members of the PAX

gene family) are monoallelically expressed in the absence of genomic imprinting

(Dr. Shirley Tilghman, personal communication). It is possible, though perhaps

unlikely that the translocation of the active PAX4 allele resulted in the

suppression of its expression.

In Chapters 3 and 4, a series of studies led to the conclusion that the GRM8

gene was interrupted by the translocation in UF53. The mapping of a partial

GRM8 cDNA to the centromeric end of HSC7E1289 was the turning point in my

positional cloning effort. Hybridization of the full-length mGluR8 cDNA against

Southern blots of the YAC clones revealed that the gene was distributed across

the contig. YAC clones known to lie proximal and distal to the breakpoint, as

well as HSC7E1289, had hybridization patterns which suggested that the GRM8

gene was possibly interrupted by the translocation. The full-length mGluR8

cDNA was also used to screen a human chromosome 7-specific cosmid library

which resulted in the retrieval of 16 cosmids. These cosmids were localized

along a 700 kb section of the contig by hybridization of unique sequences from

each cosmid against YAC Southern blots. One cosmid, 153a8, was localized to the

same region of the CpG island found in HSC7E1289, and was found to contain

that CpG island. FISH analysis identified two cosmids, 153a8 and 85b12, which

lie distal and proximal to the translocation, respectively. 153a8 was known to be

specific for the 5' end of the mGluR8 cDNA and contained a CpG island,











implying that the 5' end of the GRM8 gene falls distal to the breakpoint. In order

to localize the GRM8 disruption, PCR amplification of individual exons from the

cosmids 153a8 and 85b12 was performed. Exon 1 was found to be contained

within 153a8 while exon 2 was identified in 85b12. These results confirmed the

interruption of the GRM8 gene and established that the breakpoint was in intron
1, which is estimated to be approximately 100 kb.

The translocation breakpoint could not be detected at the 5' end of intron 1

by Southern blot analysis as discussed in Chapter 5. Hybridization of a 460 bp
exon 1 probe against a Southern blot of UF53 revealed no aberrations with the

three restriction endonucleases.

Chapter 5 also addressed whether SLOS was genetically heterogeneous in

nature. Three microsatellite markers from the 7q32 region, (AFMa125whl,
AFM309yfl, and AFM218xfI0) were genotyped in SLOS families including the

large Nwokoro family, UF53's family, and various relatives and patients of other

SLOS families. Due to the lack of complete families, conclusive results were

obtained only from the Nwokoro family. Microsatellite genotyping of this family

showed that with a high probability, the aforementioned markers and SLOS were
not coinherited based on the genotypes of affected and unaffected siblings. In

order to fit the model of an SLOS gene lying in the GRM8 region, affected
individuals would have to share the same genotype while unaffected siblings

would have genotypes distinct from their affected siblings. This was not the case

in the Nwokoro family, indicating that the SLOS locus in this family is elsewhere
in the genome. The elimination of the Nwokoro family's SLOS locus from the
7q32 region implies there are most likely at least two loci involved in SLOS,

assuming that UF53's SLOS locus is GRM8 or another gene in the region. Locus
heterogeneity in SLOS is also plausible since it is logical that defects in other

enzymes/proteins of the cholesterol pathway besides 7-DHCR could lead to











cholesterol and/or 7-DHC level abnormalities. Recently, Dr. Kelley from the

Kennedy Krieger Institute contacted us about a mildly affected SLOS patient

having "a defect of sterol transport rather than biosynthesis", determined by

radioactive tracer studies (Dr. Richard Kelley, personal communication). The

patient had a cholesterol profile indistinguishable from other SLOS patients,

supporting the notion that a different genetic defects can cause SLOS.

Future directions. I had put forth the hypothesis that UF53's translocation

breakpoint, at the 7q32.1 region, directly disrupts one allele of a putative SLOS

gene .while a more subtle mutation disrupts the other allele. However, without

the GRM8 genomic sequence, I was unable to determine if another mutation

interrupted the other GRM8 allele. I feel that once the entire cDNA sequence and

exon structure is made available, mutational analysis of the other allele is

essential. Heteroduplex analysis and single-strand conformation polymorphism

(SSCP) analysis can be used in concert to screen for mutations found in or around

the GRM8 exons. Heteroduplex and SSCP analysis are both PCR based tests

which can detect subtle mutations, such as base substitutions and small

insertions or deletions, demonstrated by aberrant migration on polyacrylamide

gels relative to controls (Hayashi and Yandell, 1993).

Heteroduplex analysis begins with PCR amplification of the region of

interest (White et al., 1992). The PCR products, containing both alleles, are

denatured by heat and then slowly reannealed. The DNA will reanneal, forming

perfectly matched duplex DNA fragments (homoduplexes) and duplex DNA

fragments in which the strands differ slightly, called heteroduplexes, if the two

alleles differ. The heteroduplex is a result of a base substitution, an insertion or a

deletion on one allele relative to the other. The detection of the heteroduplex is

based on abnormal migration on a native polyacrylamide gel with respect to the

normal homoduplexes.











SSCP analysis, like heteroduplex analysis, begins with the PCR

amplification of the region of interest such as an exon (Hayashi and Yandell,

1993). The PCR products are heated to produce single strands to be separated on

a non-denaturing polyacrylamide gel. The single strands of DNA will fold upon

themselves in a unique confirmation determined by their primary sequence

under the non-denaturing conditions. The presence of a mutation often alters the

migration of the single-stranded product. If heteroduplex and SSCP analysis do

not detect any mutations in UF53's DNA, the ultimate strategy would be to

sequence all of the exons. If a mutation in UF53's other GRM8 allele was still not
detected it would cease to be a candidate gene for SLOS. Since SLOS is an

autosomal recessive disorder, both alleles of any putative SLOS gene would have

to be disrupted in such a way that the function of the encoded protein would be
predicted to be altered in some fashion.

If GRM8 is eliminated as a candidate gene for SLOS in UF53, based on the

failure to identify another mutation in the other allele, I feel the next logical step
in ruling out the region would be to search intron 1 for embedded genes. As I

had eluded in Chapter 4, it is plausible that a gene is embedded within the 100-

kilobase intron 1, whose disruption results in SLOS. A methodology well suited

for this search is exon trapping, an established technique which recovers exons in

a genomic region (Church et al., 1994; Niu and Crouse, 1993). In exon trapping,

target genomic DNA is shotgun subcloned into an exon trap vector (i.e. pSPL3)
having a multiple cloning site flanked by functional splice donor and acceptor
sites. The recombinant plasmids are transfected into COS-7 (monkey kidney)

cells where splicing will occur if the cloned DNA contains an exon in the proper

orientation. The RNA produced by the splicing event is extracted and used for
synthesis of first strand cDNA which is then amplified using primers

complementary to the splicing vector. The recovered PCR products are cloned











and analyzed for inserts that represent putative exons. Due to the large size of

the GRM8 intron 1, several cosmids representing the region would have to be

used for exon trapping to maximize the chance that all possible genes embedded

in the intron would be recovered (Monaco, 1994). Alternatively, P-1 derived

artificial chromosomes (PACs), which have sizes of approximately 100 kb, could

be used for exon trapping if one or more is found to contain intron 1. A distinct
disadvantage of exon trapping is the failure to detect intronless or single intron

genes (Monaco, 1994). Therefore if exon trapping fails to detect any exons, cDNA

direct selection could be pursued to further screen the intron (discussed in

Chapter 3).

Another approach to studying the involvement of GRM8, or other
candidate SLOS genes, in SLOS would be the creation of a knockout mouse for

the candidate gene (Joyner, 1993). Briefly, a mutation in the candidate gene

(usually a major disruption via insertion) is introduced in a mouse embryonic

stem (ES) cell line by homologous recombination. The mutation is selected and

confffrmed, and the mutation-bearing ES cells are injected into blastocysts and

implanted into pseudo-pregnant mice. Progeny are mosaic and are bred to

create a line of fully heterozygous mice (one normal and one mutant allele).

These mice can then be crossed to produce homozygous mutant pups, whose

phenotype would be compared to SLOS. If the knockout gene is involved in
SLOS, one would expect that knockout homozygotes would produce a

phenotype similar to that of SLOS and/or display early lethality. This is one way

to functionally analyze a candidate gene, although, since knockout mouse lines

do not always produce the expected phenotype, this method may not yield

useful data. However, if the mouse phenotype recapitulated SLOS, it would

provide a model for understanding development of SLOS features throughout











embryogenesis, and would provide a model for testing genetic or medical

therapies.

In addition to the continued investigation of GRM8 and its relation to

SLOS, I feel it is imperative to examine other genes which may also be potentially

involved in the disorder. As I mentioned in Chapter 5, SLOS is believed to be

genetically heterogeneous, implying that more than one gene may be responsible

for the SLOS phenotype. Literature review revealed several interesting genes

involved in the regulation of cholesterol synthesis and cholesterol transport.

These include the genes for sterol regulatory element binding proteins 1 and 2

(SREBP-1 and-2) and sonic hedgehog (SHH).

SREBP-1 and SREBP-2, the protein products of the SREBF-1 and SREBF-2

genes, are structurally related proteins involved in the regulation of cholesterol

homeostasis (Hua et al., 1993; Wang et al., 1993). SREBP-1 and SREBP-2 bind to

the sterol regulatory element-1 (SRE-1) in the promoters of the low density

lipoprotein receptor gene (Yokoyama et al., 1993) and the 3-hydroxy-3-

methylglutaryl CoA (HMG-CoA) synthase, HMG-CoA reductase, and farnesyl

diphosphate synthase genes (Ericsson et al., 1996; Vallett et al., 1996; Wang et al.,

1993), activating transcription. Additionally, the SREBPs activate several of the

genes involved in fatty acid biosynthesis (Hua et al., 1996).
The SREBPs have a tripartite structure consisting of an NE2-terrninal

segment (the transcription factor), a middle segment composed of two membrane

spanning regions separated by a hydrophilic domain, and a carboxy-terminal

segment (Hua et al., 1996). Both proteins are bound to the endoplasmic

reticulum and nuclear envelope, with their NH2-terminal and carboxy-terminal

segments projecting into the cytosol (Hua et al., 1995). In the absence of

cholesterol and other sterols, a two step proteolytic process releases the NH2-

terminal segment which enters the nucleus and activates transcription of the











aforementioned genes (Sakai et al., 1996). The first cleavage, resulting in the

cleavage of the hydrophilic domain between the transmembrane spanning

regions, is dependent upon sterol concentrations. The second cleavage is not

sterol dependent and is responsible for the release of the NH2-terminal segment

containing the transcription factor. Recently, a SREBP cleavage activating

protein (SCAP), responsible for regulation of cholesterol biosynthesis by

stimulating the first cleavage step of SREBP-1 and -2, was isolated (Hua et al.,

1996). SCAP, whose activity is sterol dependent, is hypothesized to regulate the

protease which is involved in the first cleavage of the SREBPs by direct protein-

protein interaction.

SREBF-1 and SREBF-2 have been localized to chromosomes 17p11.2 and

22q13, respectively (Hua et al., 1995). Although neither of the SREBFs map to a

region which has been implicated in SLOS, they are particularly interesting genes

based on their extensive involvement in cholesterol and fatty acid synthesis.

Based on the known biochemistry, it feasible that homozygous disruption of

either gene could result in decreased serum and tissue cholesterol levels. Since

the locations of both SREBP genes are known, microsatellite genotyping of our

SLOS families with markers in and around these genes is feasible. Furthermore,

sequences of the intron/exon boundaries of the SREBF-1 gene are available

which would allow mutational analysis in SLOS patients (Hua et al., 1995). Since

the location and cDNA sequence of SCAP have not been elucidated, investigation

of its involvement with SLOS is not yet possible.

SHH, unlike the SREBPs, has a more direct link to SLOS. The SHH gene

encodes a secreted protein responsible for early embryo patterning in the

vertebrate CNS, anterior-posterior limb axis and somites (Echelard et al., 1993;

Fan and Tessier Lavigne, 1994; Riddle et al., 1993). Hh, the SI-ill homolog in

Drosophila, undergoes autocatalytic processing mediated by its carboxy-terminal











domain, producing a lipid modified amino-terminal fragment (Echelard et al.,

1993; Fan and Tessier Lavigne, 1994; Porter et al., 1996; Riddle et al., 1993). The

modified amino-terminal fragment, responsible for signaling activity for

embryonic patterning, was found to be covalently attached to cholesterol (Porter

et al., 1996). The covalent modification with cholesterol appears to be essential

for regional localization and concentration of Hh in various organizing centers in

the developing Drosophila embryo (Porter et al., 1996).
SHH was recently identified as a gene for holoprosencephaly (HIPE), a

genetically and phenotypically heterogeneous malformation sequence involving

the forebrain and midface (Belloni et al., 1996). HIPE has been assigned to four

distinct loci based on linkage and the analysis of chromosomal rearrangements in

HPE patients (Gurrieri et al., 1993). Both SHH and HPE were mapped to the 7q36

region (Belloni et al., 1995; Belloni et al., 1996). Mutations in SHH were identified

in patients with autosomal dominantly inherited HPE, indicating that the

absence of one SHH allele is sufficient to produce the HPE phenotype (Roessler et

al., 1996).
A casual connection between SLOS and HIPE was made in 1991 when

Verloes et al. described a group of patients with I-E and polydactyly that had

other features resembling SLOS. Recently, four patients with clinically and

biochemically confirmed SLOS and malformations characteristic of the HPE

sequence were reported (Kelley et al., 1996). SLOS features included syndactyly

of toes 2-3, micrognathia, mental retardation, cleft palate, and ambiguous

genitalia. The incomplete HPE sequence was estimated to occur in the SLOS

population at a frequency of 4% (Kelley et al., 1996). Thus, there can be clinical

overlap between SLOS and HPE raising the question of genetic overlap. Another

possible connection between these two disorders is based on the defect in

cholesterol metabolism in SLOS and the improper signaling of the SHH gene in











HIFE. SHH is also interesting since it resides at 7q36, a region suggested to be

involved in SLOS based on the patient who was monosomic at 7q34-qter due to

unbalanced translocation segregation (Berry et al., 1989). Kelley et al. (1996)
proposed two mechanisms in which the defect in cholesterol metabolism in SLOS

individuals might affect the function of the SHH protein. First, a deficiency in

cholesterol could lead to inefficient processing of the SHH protein resulting in

deficient signaling. Second, competition between 7-DHC and cholesterol for

covalent linkage to the NH2-terminal SHH fragment may result in an increased

amount of 7-DHC-modified SHH protein which could subsequently result in

inefficient or absent signaling. These two theories suggest that abnormal

processing of SHH may be secondary to the abnormal cholesterol metabolism

seen in SLOS patients, resulting in HFE features as well. Alternatively, I propose

that SHH could be an SLOS gene. Inefficient SHH signaling due to certain

mutations on both SHH alleles (possibly preventing binding to cholesterol) could

lead to improper sterol metabolism. However, this theory assumes that some

SHH mutations do not result in HPE, and that having mutations on both alleles is

not necessarily a lethal condition. Mutational analysis of the SHH gene in our

SLOS patient panel is currently underway.

Once an SLOS gene has been found and confirmed to contain mutations in

SLOS patients, several issues can be addressed such as genetic heterogeneity,

genotype/phenotype relationship, gene expression and regulation and general

protein function. In order to efficiently test SLOS patients for mutations in SLOS

gene(s), good mutational screening methods must be developed. The genomic

structure of the gene will ultimately determine which strategies which will

provide the most reliable and efficient mutation screening of the gene. The

presence or absence of mutations in an SLOS gene would be the best evidence for

or against genetic heterogeneity. If different loci were identified with mutations











in different SLOS patients, locus heterogeneity would be confirmed. Predictions

about the effect of different mutations on protein function and how these lead to

the clinical and biochemical profile seen in SLOS would also begin with

mutational analysis. These data could in turn be used to develop pharmaceutical

or other interventions early in embryogenesis.

Mutational analysis would also benefit known SLOS families by allowing

prenatal diagnosis at the earliest possible stage (8-9 weeks via CVS) or even

preimplantation. Genetic screening would be more powerful than the current

biochemical screening since the biochemical analysis is not always consistent in

SLOS and could be inaccurate due to maternal contamination of a fetal sample.

Also it is unclear if carriers can always be biochemically distinguished from

affected individuals at that stage in development. Early identification might be

crucial if any therapies are developed since the defects begin early in

embryogenesis.

The discovery of different mutations may eventually allow a correlation

between genotype and phenotype to be made. Several questions can be

addressed based on such studies. Do missense and in-frame mutations cause

result in mild SLOS cases? Do nonsense and frameshift mutations and major

disruptions cause severe SLOS? What is the phenotype of a person who has both

a "mild" and a "severe" allele?

Elucidation of an SLOS gene will warrant studies addressing expression

and regulation of the gene to build a sufficient foundation of knowledge to better

understand its protein's normal function and pathogenesis of SLOS when it is

defective. Protein studies will also contribute to the basic understanding of the

functional or enzymatic activities of the gene product and their action of other

genes and/or proteins.











I would like to conclude with a quote from Drs. Opitz and de la Cruz

(1994) in which the clinicians and researchers involved in SLOS "agreed

unanimously that research in this field be given highest priority in order to better

understand cholesterol synthesis in the mammalian brain, cholesterol transport

from mother to embryo to fetus, pre- and postnatal metabolic compensation in

cholesterol synthesis, the nature of the blood-brain barrier for cholesterol,

treatment of affected infants, children and adults, structure and genetic

specification of a 7-DHC reductase enzyme (which has never been purified!) and

its eVolution, the variability of the syndrome and whether it is genetically homo-

and heterogeneous, the population genetics of the RSH syndrome, possible
selective advantages (or disadvantages) of heterozygotes, and means of newborn

screening, carrier detection, and prenatal diagnosis."

















REFERENCES


Abe, T., Sugihara, H., Nawa, H., Shigemoto, R., Mizuno, N., and Nakanishi, S.
(1992). Molecular characterization of a novel metabotropic glutamate receptor
mGluR5 coupled to inositol phosphate/Ca2+ signal transduction. J. Biol. Chem.
267, 13361-8.

Acosta, P. B. (1995). Theoretical and practical aspects of dietary treatment of
Smith-Lemli-Opitz syndrome. Int. Pediatr. 10, 37-40.

Alley, T. L., Gray, B. A., Lee, S. H., Scherer, S. W., Tsui, L. C., Tint, G. S., Williams,
C. A., Zori, R., and Wallace, M. R. (1995). Identification of a yeast artificial
chromosome clone spanning a translocation breakpoint at 7q32.1 in a Smith-
Lemli-Opitz syndrome patient. Am. J. Hum. Genet. 56, 1411-6.

Alley, T. L., Scherer, S. W., Huzienga, J. J., Tsui, L.-C., and Wallace, M. R. (1996).
Physical mapping of the chromosome 7 breakpoint region in an SLOS patient
with t(7;20)(q32.1;qt3.2). Am. J. Med. Genet. 68, 279-281.

Batta, A. K., Salen, G., Tint, G. S., and Shefer, S. (1995). Identification of 19-nor-
5,7,9(10)-cholestatrien-3 beta-ol in patients with Smith-Lemli-Opitz syndrome. J.
Lipid Res. 36, 2413-8.

Bejanin, S., Cervini, R., Mallet, J., and Berrard, S. (1994). A unique gene
organization for two cholinergic markers, choline acetyltransferase and a
putative vesicular transporter of acetylcholine. J. Biol. Chem. 269, 21944-7.

Belloni, E., Frumkin, A., Scherer, S. W., Rovet, J., Mitchell, H., Bamoski, B., Hing,
A., Donis-Keller, H., L.-C., T., and Muenke, M. (1995). Characterization of
holoprosencephaly minimal critical region in 7q36. Chromsome 7 workshop
1994. Cytogenet. Cell. Genet. 71, 31.

Belloni, E., Muenke, M., Roessler, E., Traverso, G., Siegel-Bartelt, J., Frumkin, A.,
Mitchell, H. F., Donis-Keller, H., Helms, C., Hing, A. V., Heng, H. H. G., Koop, B.,
Martindale, D., Rommens, J. M., Tsui, L.-C., and Scherer, S. W. (1996).
Identification of Sonic hedgehog as a candidate gene responsible for
holoprosencephaly. Nat. Genet. 14,353-356.











Berry, R., Wilson, H., Robinson, J., Sandlin, C., Tyson, W., Campbell, J., Porreco,
R., and Manchester, D. (1989). Apparent Smith-Lemli-Opitz syndrome and
Miller-Dieker syndrome in a family with segregating translocation
t(7;17)(q34;p13.1). Am. J. Med. Genet. 34,358-65.

Bird, A. P. (1986). CpG-rich islands and the function of DNA methylation. Nature
321, 209-13.

Bird, A. P., Taggart, M. H., Nicholls, R. D., and Higgs, D. R. (1987). Non-
methylated CpG-rich islands at the human alpha-globin locus: implications for
evolution of the alpha-globin pseudogene. EMBO f. 6, 999-1004.

Bogart, M. H., Cunniff, C., Bradshaw, C., Jones, K. L., and Jones, 0. W. (1990).
Terminal deletions of the long arm of chromosome 7: five new cases. Am. J. Med.
Genet. 36, 53-5.

Brock, D. J. H. (1993). Molecular genetics for the clinician (Cambridge: University
Press).

Brooks Wilson, A. R., Goodfellow, P. N., Povey, S., Nevanlinna, H. A., de Jong, P.
J., and Goodfellow, P. J. (1990). Rapid cloning and characterization of new
chromosome 10 DNA markers by Alu element-mediated PCR. Genomics 7, 614-
20.

Cawthon, R. M., Weiss, R., Xu, G. F., Viskochil, D., Culver, M., Stevens, J.,
Robertson, M., Dunn, D., Gesteland, R., O'Connell, P., and et al. (1990). A major
segment of the neurofibromatosis type 1 gene: cDNA sequence, genomic
structure, and point mutations [published erratum appears in Cell 1990 Aug
10;62(3):following 608]. Cell 62, 193-201.

Chasalow, F. I., Blethen, S. L., and Taysi, K. (1985). Possible abnormalities of
steroid secretion in children with Smith-Lemli-Opitz syndrome and their parents.
Steroids 46, 827-43.

Cherstvoy, E. D., Lazjuk, G. I., Ostrovskaya, T. I., Shved, I. A., Kravtzova, G. I.,
Lurie, I. W., and Gerasimovich, A. I. (1984). The Smith-Lemli-Opitz syndrome. A
detailed pathological study as a clue to a etiological heterogeneity. Virchows
Arch. A. Pathol. Anat. Histopathol. 404, 413-25.

Choi, D. W. (1992). Excitotoxic cell death. J. Neurobiol. 23, 1261-76.

Church, D. M., Stotler, C. J., Rutter, J. L., Murrell, J. R., Trofatter, J. A., and
Buckler, A. J. (1994). Isolation of genes from complex sources of mammalian
genomic DNA using exon amplification. Nat. Genet. 6,98-105.

xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EKZBIULBD_L7FIVQ INGEST_TIME 2014-07-10T21:52:19Z PACKAGE AA00022839_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES