Genomic cloning and tissue-specific expression of lipocortin I

MISSING IMAGE

Material Information

Title:
Genomic cloning and tissue-specific expression of lipocortin I
Physical Description:
viii, 106 leaves : ill. ; 29 cm.
Language:
English
Creator:
Horlick, Kenneth Robert, 1963-
Publication Date:

Subjects

Subjects / Keywords:
Research   ( mesh )
Adrenal Cortex Hormones   ( mesh )
Gene Expression   ( mesh )
Cloning, Molecular   ( mesh )
Annexin I   ( mesh )
Base Sequence   ( mesh )
Molecular Sequence Data   ( mesh )
Genes, Structural -- genetics   ( mesh )
Mice -- genetics   ( mesh )
Rats -- genetics   ( mesh )
Department of Biochemistry and Molecular Biology thesis Ph.D   ( mesh )
Dissertations, Academic -- College of Medicine -- Department of Biochemistry and Molecular Biology -- UF   ( mesh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1990.
Bibliography:
Bibliography: leaves 89-105.
Statement of Responsibility:
by Kenneth Robert Horlick.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001581151
oclc - 24594090
notis - AHK5058
System ID:
AA00004760:00001

Full Text











GENOMIC CLONING AND TISSUE-SPECIFIC
EXPRESSION OF LIPOCORTIN I


















BY

KENNETH ROBERT HORLICK


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


1990















To Marta


Digitized by the Internet Archive
in 2011 with funding from
University of Florida, George A. Smathers Libraries with support from LYRASIS and the Sloan Foundation


http://www.archive.org/details/genomiccloningti00horl














ACKNOWLEDGEMENTS

I would like to thank the members of my committee for

their invaluable guidance and support during my graduate

studies. Special thanks go to Dr. Richard Boyce for both

his guidance and understanding. Also, I would like to thank

my fellow graduate students, Bill Wong, Bill Dougall, Jan-

Ling Hsu, Sally Chesrown, and Jonathan Hurt, as well as Joan

Wilson, Dr. Gary Visner, Dr. Gerald Stephanz, and Dr. John

Valentine, for their helpful discussions, support, and

friendship over the past five years.

In addition, I would like to acknowledge my mentor, Dr.

Harry Nick, for all his guidance and support. My education

and training under his direction were a valuable and

rewarding experience.

Finally, I would like to express my unending gratitude

to my parents, Herbert and Maxine Horlick, whose unfaltering

support and love have enabled me to reach this point.


iii
















TABLE OF CONTENTS

page

ACKNOWLEDGEMENTS ...............................iii

LIST OF FIGURES ..................................... V

ABSTRACT...................................................ii

CHAPTERS

1 INTRODUCTION..... ................................... 1

2 CHARACTERIZATION OF THE MOUSE LIPOCORTIN I GENE..... 8

Introduction....................................... 8
Materials and Methods............................... 10
Results............... ................. ............ 13
Discussion .......................................... 27

3 REGULATION OF LIPOCORTIN I IN RAT MAMMARY TISSUE
DURING GLANDULAR DIFFERENTIATION AND INVOLUTION..... 33

Introduction.......... ......... ..................... 33
Materials and Methods.............................. 37
Results................................................ 40
Discussion .......................................... 50

4 TISSUE-SPECIFIC REGULATION OF LIPOCORTIN I mRNA
LEVELS IN RATS DURING DIABETES AND BY GLUCAGON....... 55

Introduction....................................... 55
Materials and Methods............................... 58
Results ............... .............................. 60
Discussion.......................................... 70

5 CONCLUSIONS AND FURTHER DIRECTIONS................... 79

REFERENCES.... ........................................... 89

BIOGRAPHICAL SKETCH.....................................106














LIST OF FIGURES
Figure page

2-1 Southern blot analysis of mouse genomic DNA
with a lipocortin I cDNA probe.................. 14

2-2 Restriction map and genomic structure of the
mouse lipocortin I gene........................ 18

2-3 Sequence of the mouse lipocortin I gene.......... 20

2-4 Sequence of the mouse lipocortin I promoter
region........................................... 24

2-5 Correlation of lipocortin I genomic structure
with protein domains............................ 26

3-1 Regulation of mammary lipocortin I mRNA during
pregnancy, lactation, and involution............ 41

3-2 Decrease in mammary lipocortin I mRNA levels
with the onset of lactation..................... 43

3-3 Regulation of lipocortin I mRNA and protein
levels in rat mammary tissue during glandular
differentiation and involution.................. 44

3-4 Densitometric analysis of the data in Figure
3-3.............................................. 45

3-5 The role of the suckling stimulus directly
after birth on lipocortin I mRNA levels in
mammary tissue.................................. 47

3-6 Induction of lipocortin I mRNA and protein
expression in mammary tissue of dams following
pup removal at birth or day seven of suckling... 48

3-7 Specificity of the mammary lipocortin I mRNA
regulation following pup withdrawal.............. 49

3-8 Densitometric analysis of the increase in
lipocortin I mRNA levels following pup
withdrawal ...................................... 51









4-1 Expression of lipocortin I mRNA in tissues from
a control and three-day streptozotocin-diabetic
rat.............................................. 61

4-2 Increased levels of lipocortin I mRNA in the
heart, kidney, and hindleg muscle of three-day
streptozotocin-diabetic rats.................... 62

4-3 Heart, kidney, and hindleg muscle lipocortin I
mRNA expression in streptozotocin-injected rats
on and off insulin treatment.................... 64

4-4 Induction of lipocortin I mRNA in the rat heart
after injection of glucagon..................... 66

4-5 Increased levels of lipocortin I mRNA in the
heart after glucagon injection in four animals.. 67

4-6 Time course of the increase in lipocortin I
mRNA in the heart after injection of glucagon... 68

4-7 Lipocortin I mRNA expression in the rat heart
following injection of glucagon and other
hormones......................................... 69

4-8 Lipocortin I mRNA expression in the mouse heart
following injection of glucagon and other
hormones......................................... 71














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

GENOMIC CLONING AND TISSUE-SPECIFIC
EXPRESSION OF LIPOCORTIN I

by

KENNETH ROBERT HORLICK

December, 1990

Chairman: Dr. Harry S. Nick
Major Department: Biochemistry and Molecular Biology

The mouse lipocortin I structural gene was isolated and

characterized by screening two mouse Balb/c liver genomic

libraries with mouse lipocortin I cDNA and oligonucleotide

probes. Restriction enzyme mapping, Southern blotting, and

DNA sequencing were carried out on three overlapping genomic

clones. The lipocortin I gene spans 17 Kbp and is divided

into 13 exons which encode a protein of 346 amino acid

residues. The promoter region of the gene has a TATA box

and a CCAAT box located upstream of the transcription

initiation site at -31 and -76 bp, respectively. Several

potential regulatory sites in the 5' flanking region were

identified by comparison to known protein-binding consensus

sequences. A comparison of the lipocortin I gene structure

with that of the closely related gene lipocortin II suggests

a common evolutionary ancestor for these proteins.


vii








Lipocortin I mRNA and protein levels in the rat mammary

gland were studied during various stages of differentiation.

Both were abundant in the nonpregnant state, and began a

striking decline at midpregnancy when epithelial cells

undergo proliferation and differentiation. In the lactating

gland, no lipocortin I was detected. During involution of

the gland after weaning, lipocortin I mRNA and protein

increased to levels observed in nonpregnant animals. The

repression of lipocortin I mRNA expression at birth and

during lactation is dependent upon the suckling stimulus,

since pup withdrawal resulted in a rapid, mammary-specific

increase in lipocortin I mRNA levels. These results suggest

lipocortin I may play a role in mammary cell growth status

during gland differentiation.

To investigate potential alterations in lipocortin I

expression during diabetes, I injected rats with

streptozotocin and analyzed mRNA levels in several tissues.

In three-day streptozotocin-diabetic rats, lipocortin I mRNA

levels were significantly increased in the heart, kidney,

and hindleg muscle. Glucagon, a hormone highly elevated in

diabetics, caused induction of lipocortin I mRNA in the

heart, but not in the kidney or muscle, four hours after

injection.


viii














CHAPTER 1

INTRODUCTION

Lipocortin I (lipo I), also called annexin I (Crumpton

and Dedman, 1990), is one of the best characterized members

of the annexins, a group of calcium-dependent, phospholipid-

binding proteins which share about 50% sequence homology and

whose physiological role remains unclear. The annexins

differ from previously characterized calcium-binding

proteins in that they lack an EF-hand-type calcium binding

site (Geisow and Walker, 1986). Proteins which are part of

this family include lipocortins (Huang et al., 1986),

calpactins (Glenney, 1986), calelectrins (Geisow et al.,

1986; Geisow and Walker, 1986; Sudhof et al., 1988),

calcimedins (Smith and Dedman, 1986), anticoagulant protein

(Funakoshi et al., 1987), endonexins (Geisow et al., 1986;

Geisow and Walker, 1986; Schlaepfer et al., 1987; Weber et

al., 1987), and chromobindins (Creutz et al., 1987).

Various annexins have been implicated in cellular processes

including inflammation (Flower, 1988), exocytosis

(Burgonyne, 1988; Drust and Creutz, 1988), differentiation

(Violette et al., 1990), blood coagulation (Tait et al.,

1988), the immune response (Hirata and Iwata, 1983), and

membrane-cytoskeletal linkage (Roy-Choudhury et al., 1988).








2

Most recently, lipocortin III was identified as inositol

1,2-cyclic phosphate 2-phosphohydrolase, an enzyme important

in metabolizing the most highly active inositol phosphate

intermediates (Ross et al., 1990). This suggests the

possibility that the other annexins may also play a role in

inositol metabolism.

Lipocortin I has a molecular weight of 35,000 daltons

and has been immunolocalized mainly to the inner surface of

the plasma membrane (Glenney et al., 1987). Most of its

observed biological properties, including binding to

phospholipid and actin, are dependent on calcium levels

(Glenney, 1986; Schlaepfer and Haigler, 1987). Multiple

binding sites for both calcium and phospholipid apparently

exist in the protein (Klee, 1988). Recent studies using

monoclonal antibodies detected both calcium-bound and

calcium-free forms of lipo I associated with membranes in

A431 cells (Hayashi et al., 1990). Early experiments

indicated the protein was secreted, but this observation

remains unexplained because no evidence for a signal peptide

was found (Wallner et al., 1986). Expression is limited to

certain cell and tissue types (Fava et al., 1989; Glenney et

al., 1987; Pepinsky et al., 1986; Wallner et al., 1986), and

in some cases lipo I may account for as much as 0.1 to 0.5%

of total cellular protein (Violette et al., 1990).

Phosphorylation of specific cellular proteins by growth

factor/hormone receptor kinases is believed to be important








3

in intracellular signal transduction (Ullrich and

Schlessinger, 1990). Lipocortin I has been shown to be a

substrate for the epidermal growth factor receptor tyrosine

kinase in intact A-431 human carcinoma cells and human

fibroblasts (Giugni et al., 1985; Sawyer and Cohen, 1985),

and is phosphorylated by the epidermal growth factor

receptor in vitro in the presence of epidermal growth factor

and calcium (De et al., 1986). Karasik et al. (1988) have

demonstrated that lipo I is also a substrate for the insulin

receptor tyrosine kinase in isolated rat hepatocytes. In

addition, both protein kinase C and cyclic-AMP-dependent

protein kinase have been shown to phosphorylate lipo I on

serine/threonine residues (Varticovsky et al., 1988).

Lipocortin II, or p36, is a substrate for the pp60v8

oncogene kinase, further implicating the annexins in

cellular signal transduction (Cooper and Hunter, 1983).

The lipocortins were originally defined as

glucocorticoid-induced, anti-inflammatory proteins which

inhibit phospholipase A2, the enzyme which releases

arachidonic acid from membrane phospholipids and begins the

cascade of enzymatic steps leading to production of

prostaglandins and leukotrienes (Blackwell et al., 1980;

Flower and Blackwell, 1979; Hirata et al., 1980). Several

of the lipocortins were shown to inhibit phospholipase A2,

as well as other phospholipases, in vitro (Hirata, 1981).

The mechanism of this inhibition is now believed to be








4

"substrate depletion," where lipocortins bind to

phospholipids and protect them from cleavage by

phospholipases. This is because no direct interaction

between lipocortins and phospholipases has been detected,

and the inhibition can be overcome by raising the

concentration of phospholipid substrate in the in vitro

assays (Aarsman et al., 1987; Davidson et al., 1987; Haigler

et al., 1987).

The annexins share a common core region consisting of

four 70-amino acid repeating units (eight for

p68/calelectrin), based on predicted secondary structure

considerations (Crompton et al., 1988). In addition, a 17-

amino acid consensus sequence, believed to contain the sites

responsible for calcium/phospholipid binding activity, was

identified within each repeat (Geisow et al., 1986). The N-

termini of the lipocortins are divergent and may be involved

in the regulation of the individual proteins (Crompton et

al., 1988; Huang et al., 1986; Pepinsky et al., 1988; Saris

et al., 1986; Wallner et al., 1986). The N-terminus of lipo

I contains sites for tyrosine and serine/threonine

phosphorylation (Karasik et al., 1988; Sawyer and Cohen,

1985; Varticovski et al., 1988), proteolysis (Ando et al.,

1989; Chuah and Pallen, 1989), and head-to-head dimerization

(Pepinsky et al., 1989). The N-terminus of lipocortin II

also contains phosphorylation sites and is the location of

binding of pll, a protein which forms a tetramer with








5

lipocortin II and binds tightly to the inner membrane

surface (Glenney and Tack, 1985).

Although the cellular function of lipo I is presently

unclear, its implication in the important processes

mentioned above justifies further investigation of this

protein. This manuscript describes the groundwork for

studies designed to investigate the regulation of lipo I

gene expression using rats and mice as models. Chapter 2

involves the isolation and characterization of the mouse

lipo I structural gene and promoter. Knowledge of the gene

structure would enable a comparison with that of lipo II,

recently published (Amiguet et al., 1990), and other

annexins when these are eventually described. Such a

comparison would reveal if the exon-intron patterns of the

genes correlate with the 70-amino acid repeats and 17-amino

acid consensus sequences of the corresponding proteins. In

addition, valuable information regarding the evolution of

the annexins and their relationship to one another would be

gained.

Obtaining the promoter sequence in the 5' flanking

region of lipo I would likely be essential in studying gene

regulation, since both nonspecific and specific protein

binding sites important in gene expression are often located

upstream of the transcription initiation site of genes

(Mitchell and Tjian, 1989). Comparison of the promoter

region of lipo I with previously identified consensus








6

sequences representing regulatory protein binding sites

might locate similar regulatory regions in lipo I. Whether

putative regulatory sites are important in vivo could then

be addressed in future experiments, described in Chapter 5.

Chapters 3 and 4 involve an attempt to identify and

define a systems) which exhibit regulation of lipo I gene

expression so that the molecular basis of this regulation

could later be investigated. Also, exploring changes of

lipo I expression in response to various stimuli might help

elucidate its biological role. Based on a link between

prolactin, an important hormone in lactation, and a lipo I-

like gene in the pigeon cropsac (Horseman, 1989), potential

hormonal control of lipo I in the rat mammary gland is

evaluated in Chapter 3. The mammary gland, like the pigeon

cropsac, differentiates into a nutrient-secreting tissue

during pregnancy and is known to be a target of prolactin

action (Topper and Freeman, 1980). Northern and Western

analyses were carried out to analyze lipo I mRNA and protein

levels in mammary tissue of dams during pregnancy,

lactation, and weaning. During these periods, several

hormones undergo large alterations in plasma concentration,

and gland morphology and cell population change radically

(Borellini and Oka, 1989). In addition, the consequences of

premature pup withdrawal and prevention of the suckling

stimulus on lipo I expression in mammary tissue are

addressed in Chapter 3. These studies revealed a dramatic








7

regulation of lipo I during gland differentiation and

involution, and a hypothesis addressing this phenomenon is

provided.

In Chapter 4, tissue-specific regulation of lipo I mRNA

levels in diabetic rats is shown. These studies were

undertaken to explore a possible connection between lipo I

and the altered lipid metabolism known to occur in some

diabetic tissues. An attempt to identify a particular

hormone involved in regulation of lipo I during diabetes

implicated the peptide hormone glucagon, which is highly

elevated in the diabetic state. These experiments represent

the first report of glucagon influencing lipo I mRNA

expression and of tissue-specific alterations in lipo I in a

disease state.














CHAPTER 2

CHARACTERIZATION OF THE MOUSE
LIPOCORTIN I GENE

Introduction

The studies described in this chapter were carried out

for two main purposes: to determine the genomic structure of

the lipo I gene, and to obtain promoter sequence from the 5'

flanking region of the gene. Information provided by such

experiments would enable future studies aimed at

investigating specific regulatory mechanisms involved in

lipo I gene regulation.

It has been proposed that exons encode functional or

structural domains of proteins (Blake, 1983; Gilbert, 1978).

The lipo I protein consists of a unique N-terminus and a

core region consisting of four 70-amino acid repeating units

(Crompton et al., 1988). In addition, a 17-amino acid

consensus sequence believed to be responsible for

calcium/phospholipid binding was identified within each of

the four 70 amino acid repeats (Geisow et al., 1986). By

characterizing the lipo I gene structure, I could determine

if discrete exons encode protein domains, such as the unique

N-terminus and the internal repeating units.

The only lipocortin structural gene which has been

characterized until now is that of mouse lipo II, which was

8








9

reported to consist of at least 12 exons spanning 22 Kbp and

to be located on chromosome nine (Amiguet et al., 1990).

These authors knew from primer extension analysis (Saris et

al., 1986) that they had not located the first -52 bp of the

mRNA in the gene, and therefore had not identified the first

genomic exon(s), which are noncoding. The intron/exon

structure of the lipo II gene was analyzed in relation to

the known protein domains. It was shown that although no

striking correlation existed with regard to exons and the

70-amino acid repeats, three of the four 17-amino acid

consensus regions were encoded at the end of one exon and

the start of the following exon. In addition, the first two

exons of lipo II were found to code for separate regions

believed to be important in regulation of protein function.

In this chapter, similar analysis regarding exon-protein

domain correlation for lipo I is shown, and a comparison

made between the two gene structures. The results provide

interesting insight into the evolution of the annexin

family.

The region upstream of many genes has been shown to

encode sites for binding of regulatory proteins (Mitchell

and Tjian, 1989). Such sites may influence mRNA expression

in either a positive or negative fashion. Both gene-

specific and ubiquitous transcription regulatory factors

have been identified which bind to certain DNA sequences.

For many of these factors, a consensus binding sequence has








10

been derived from studying the promoter regions of several

genes. Regulatory elements defined so far include those

that mediate gene expression by glucocorticoids (GRE)(Picard

et al., 1988), cAMP (CRE)(Gonzalez et al., 1989), and serum

(SRE)(Norman et al., 1988). Sequencing the 5' upstream

region of lipo I would enable localization of potential

regulatory sites involved in transcription by comparison

with known consensus sequences. This is especially

important because there is evidence that glucocorticoids

induce lipo I mRNA levels (Phillips et al., 1989; Wallner et

al., 1986; Wong et al., submitted) and that the presence or

absence of serum influences lipo I mRNA expression in

several cell types (K. Horlick and W. Wong, unpublished

results).

The work described in this chapter will be important in

future studies aimed at investigating mechanisms of

regulation of mouse lipo I gene expression. A restriction

map and genomic sequence are required for studies such as

chromatin structure analysis (Zaret and Yamamoto, 1984),

deletion analysis (Dean et al., 1983), and in vivo

footprinting (Church and Gilbert, 1984).

Materials and Methods

Library Screening and Isolation of Lipo I Genomic Clones

A mouse Balb/c liver genomic library obtained from Dr.

Tom Caskey, Baylor, and one purchased from Clonetech, Palo

Alto, CA, were screened as described by Sambrook et al.








11

(1982). Typically, 20,000 recombinant phages were plated on

15-cm YT agar plates and incubated at 37*C for 16-18 hr.

Duplicate nitrocellulose lifts of each plate were washed in

a solution containing 50 mM Tris (pH 8.0), 1 M NaCl, 1 mM

EDTA, and 0.1% SDS at 370C for 1 hr. The filters were

prehybridized in 6X SSC (lX SSC contains 0.15 M NaCl, 0.015

M sodium citrate, pH 7.0), 5X Denhardt's solution (IX

Denhardt's solution is composed of 0.1 g each of ficoll,

polyvinylpyrrolidone, and bovine serum albumin), 0.1% SDS,

and 100 Ag/ml denatured salmon sperm DNA for 2 hr at 60C.

A P-labeled lipo I cDNA probe synthesized by random primer

extension (Feinberg and Vogelstein, 1983) was added to the

hybridization solution, and the filters were further

incubated at 60C for an additional 16-18 hr. The filters

were subjected to three 1-hr washes at 60C in lX SSC and

0.1 % SDS, and exposed to Kodak X-OMAT AR film for three

days at -85C with an intensifying screen (lightning-plus,

Du Pont). When a kinased oligonucleotide was used as a

probe, the washes were performed at 50C in 3X SSC and 0.3%

SDS. Each positive clone identified from the primary

screening was further purified to homogeneity, grown up

large scale, and its genomic insert excised and subcloned

into M13mpl8 for DNA sequencing (Messing, 1983).

Probe Preparation

Hybridization probes were synthesized with a-2P dATP

(NEN, 800uCi/mmole) by random primer extension with hexamer








12

primers (Pharmacia) (Feinberg and Vogelstein, 1983) using

all or part of a 900 bp partial mouse lipo I cDNA insert

(clone BWS9, W. Wong) or a 2.7 Kb Xba fragment from genomic

clone Jlipol9. A 61-bp oligonucleotide corresponding to the

first exon was radiolabeled using gamma-32P ATP and T4

polynucleotide kinase (BRL) as described by Sambrook et al.

(1982).

Genomic DNA Isolation and Southern Analysis

Mouse genomic DNA was isolated from confluent A31

Balb/c fibroblast cells by lysis in a solution of 1 mM Tris

pH 8.0, 100 mM EDTA, 1% SDS, and 50 mg/ml proteinase K

(Boehringer Mannheim Biochemicals). Following incubation

for 1 hr at 50C, nucleic acids were ethanol precipitated

and resuspended. After a 1-hr incubation at 370C in the

presence of 100 ug/ml DNase-free RNase A (Sigma), the DNA

was purified by phenol/chloroform extractions. Genomic DNA

was then ethanol precipitated and resuspended at a

concentration of about 0.5 mg/ml. Restriction enzyme-

digested DNA (30 Mg) was run on 1% agarose gels containing

0.04 M Tris-acetate and 1 mM EDTA (pH 8.0) for 14-16 hr at

30 volts. The gel was treated as follows prior to transfer:

Depurinated two times with 0.25 M HCl (10 min each),

denatured with 0.5 M NaOH (two times, 25 min each), and

neutralized with 1.78 M Tris-borate, 1.78 M boric acid, and

0.04 M EDTA, pH 8.0. Genomic DNA was transferred

electrophoretically for 1 hr to a noncharged nylon membrane








13

(GeneScreen, NEN), immobilized on the membrane by UV cross-

linking (Church and Gilbert, 1984), and hybridized with a

specific radiolabeled probe.

Hybridization of Southern Blots

Hybridizations were done as previously described

(Church and Gilbert, 1984). Blots were prehybridized for

15-30 min in a buffer containing 1% bovine serum albumin

(Sigma), 1 mM EDTA, 0.5 M NaHPO, (pH 7.2), and 7% SDS at

60'C. After addition of the labeled probe, the blots were

hybridized at 60C for 12-16 hr. The blots were then

subjected to three 10-min washes in 1 mM EDTA, 40 mM NaHPO4

(pH 7.2), 1% SDS at 65"C, and exposed to Kodak X-OMAT AR

film for various periods of time. When a kinased

oligonucleotide probe was used, the hybridization was done

at 50C for 3-5 hr, and the wash was done in the buffer

described above, but containing a final concentration of

0.45 M Na+ at 50C.

Results

Isolation of Lipo I Genomic Clones

Initially, mouse lipo I cDNA clone BWS9, containing

almost all of the lipo I coding sequence, was used to screen

a Balb/c mouse liver EMBL3 genomic library. This probe was

also used in genomic Southern blot analysis to estimate the

length of genomic DNA which contained lipo I coding sequence

(Fig. 2-1). Clone Jlipol9 was isolated and then

characterized by Southern blot analysis using the cDNA to

















2 3 4


- 23.2

- 9.4

- 6.6


- 4.4


-2.3
-2.0












Figure 2-1. Southern blot analysis of mouse genomic DNA with
a lipocortin I cDNA probe. Genomic DNA isolated from mouse
A31 fibroblast cells was digested with Xba I (lane 1), Sca I
(lane 2), Sst I (lane 3), and Pst I (lane 4). Restricted
DNA (30 pg/lane) was run on 1% agarose gels,
electrotransferred to GeneScreen membrane, and hybridized
with a lipocortin I cDNA probe. The numbers on the right
designate the relative position of molecular size markers in
Kbp.









15

localize exons. Restriction fragments containing exons were

cloned into M13 vectors and sequenced by the dideoxy method.

Intron/exon junctions were determined by comparing genomic

sequence with that of the mouse lipo I cDNA. Jlipol9 had a

19 kb insert which was found to contain 10 lipo I exons, but

not the 5' or 3' ends of the gene. Southern blot analysis

comparing lipo I restriction fragments from the genome to

those present in Jlipol9 (data not shown) revealed that this

clone contained a seven kb fragment at its 5' end which was

not part of the lipo I gene. Thus, the 19 kb insert in this

clone was a combination of two genomic fragments, one

containing part of the lipo I gene and another from

elsewhere in the genome, which were apparently ligated

together during library construction. This phenomenon was

also reported with regard to the cloning of the human

cytosolic aldehyde dehydrogenase (ALDH1) gene, where the

authors note they identified several genomic clones which

contain ALDH1 sequence at one end and other unidentified

sequences at the other end (Hsu et al., 1989). The Southern

analysis and subsequent characterization of another clone

rule out the possibility that Jlipol9 contained a

pseudogene.

To isolate the remainder of the gene, 5'- and 3'-

specific probes were used to screen a second mouse liver

EMBL3 genomic library. A 2.7 kb Xba restriction fragment

from Jlipol9 and an APCR-derived oligonucleotide








16

corresponding to the 5' end of the lipo I mRNA (Horlick et

al., submitted) were used as probes for the 5' end, while a

460 bp fragment corresponding to the 3' end of the cDNA was

utilized to probe for the 3' end. Two new clones, 5'lipol4

and 3'lipol7, were isolated using these probes and

characterized as explained above. Exon one sequence was

obtained from 5'lipol4, while the last two exons were found

in 3'lipol7.

Genomic Structure of Lipo I

The three lipo I genomic clones examined, including a

restriction map consistent with genomic Southern analysis,

are shown in Fig. 2-2. The DNA sequence of the mouse lipo I

gene and deduced amino acid sequence from the coding region

are shown in Fig. 2-3. The gene consists of 13 exons

spanning approximately 17 kb. Exon and intron sizes are

shown in Table 2-1. Exons range in size from 59 to 341 bp,

with an average size of 107 bp. Intron lengths vary from

109 bp to 6.0 kb, averaging 1.4 kb. The first exon is

entirely noncoding and represents the true start site of

transcription as determined by primer extension (Horlick et

al., submitted). Exon two contains the first 66 coding

bases and exon 13 has the final 54. In 7 of 11 cases,

introns occur between codons. All of the exon-intron

junctions conform to the consensus sequences established for

intron donor (5'GTA/GAGT3') and acceptor (5'[Py]n-NC/TAG3')

splice signals (Breathnach and Chambon, 1981; Padgett et

al., 1986).













0 4


C ) O II






0 0 H

.H4 0
4HO *.
0 4) 4)H




) fn II
-- 0 U .1%
0 0 M 0
Sr-4 0 C





)0 U) 0 N


0 r-C P

V )0 ) H



rl m
C Q00













*H 0 H ) II
O rO 04
t o Ur 4)
. 0r.
-H 0 c QW 11









0N004
0 04 r -4
e 4 W aO a0










tO 4M -)
4) > M 0
P PC-











( -H U









0 4-i 0 II
3 -H ti0 E
V C V a






N000
C -H ( )
00 Q) 0) 4)
c 0e 4 Eu H




C= -(0 X



r E-4 Q) U)



I - r (
S) 1COw
Q) ,C C 3 -
3 *MO















Cr,

CU
q-g
q-9


Icn


Ij


I-
LU


U) -


'In
Imm

Ecu




















2


cn-


cn


*r4


Un -


Un-j












o ow 4-
41 0 0
0 O 00)
4 Cr. r.1 l
0 c o (1 3 u id 4)


a 0 0) 4) O
U 41 0 0 +
&3M U P4 M
UQ) CP4 0 l
H r-l -H c
0-) ) 4\-

o4 0 > 4 0 r

SC0 *4 4+J u
.0.4 x *00

SU a)fl Ul) .i cr U) 0
O 0 4) 4)
04 0Cr4 94
C W 0 0) C4 rl
* m QM V M
# C US O0

Q-) U r. 8 r 0) >4V0
u --I u -H
0 ON 0 w o 9a
*4rk4. *4J -*k


0 0 ) 4 0
C 0) r ) 0 UC r-0
o r 0 ) *H CH c U
HP .M -4 0 0 U) 0
0 *-I O 1 C 3
OCOS~a 04: OH
0 0 0 -H 0
o C 0 HP 0a) -

S4) C C 8rl-- 4 c -r :


0 <4 J I r0 Cv 0
S4) 1-H 0 C0 143

'O X C4C. -4 0
0 0 0.v 0 04
) 0) 0 4 a 4-

0U) 4 ) U0 0C
C ASM 0 *L) )>

C Q a tmou
0 m 4 4 ON 1*1 *4
0 -4 0 W0 4)40 04
4 C -4. J0o

0 () 4 i 0C -4 a
4 ) 0 4) 0
wOCCCOO -HC
0 () 4) 4) 4O C O 4


rv-4 U u uA.Q 0 QO










51
21





I5
Is




I






US 4
53
4 n
Bi








3!
i i

LA
CS
y5

St



ag
ya
i5
s
WA


MI
[5
SO



8
Sj j
g&
5i
5
^
IS
MA f
IS
Gff
51
a.2l


i



I




SIa
as s
3d
SI

El
51

Eg
li
B^
IS
8S3

5 3
5
El
91
'


03

8
85
5

I




IS




ip

85
4-








U,
El 1
lJ
YS







gg
gg







aEJ
i






B
8
MF l
81




S

M&
yif

81
E1
81
51
il


Ih


BL






3!S


B 8
519,


g ^
I
E





Z i
13!



LiP
8'
IS
E*
5>


eg




I
Sii

5p
3h
gs

a
i
^*l








4--
5 S
3

S k
$5

if
5
s ^
si
8 **
21
214


r5








in
I
g ;



51



s* k
U
J
uj














0
51

'k



Ba
5^


E51

gB


SB


s3:!







&i
'p8

.pD
Ur
- 5


p- u


s

s -s












CI
S
Ei
.a;
t


ggfi
E ."



yi$

Ef
y^t i




yi
3s
is
.e *

***^
SS


g^
3!



&

i



aS
8



GI







3!t
p., u.
p










S-






*40
IS
VI
2i
Ya


4 -i
Lg
gi
s s
if
MS

Ed



5i@
5 t
%
5


gi


88

PI
83'


gi
as
821


I








I



i





























Table 2-1. Mouse lipocortin I intron and exon sizes.


exon length (bp) intron length (bp)


1 61 I 6000
2 80 II 109
3 109 III 600
4 88 IV 1800
5 114 V 600
6 91 VI 1400
7 80 VI 800
8 56 VIII 1300
9 94 IX 1300
10 96 X 1200
11 59 XI 800
12 123 XII 900
13 342








22

About 1400 bp of sequence upstream of the transcription

initiation site is shown in Fig. 2-4. The initiation site

closely resembles the weak consensus sequence 5'-

PyPygAPyPyPyPyPy-3', with transcription beginning at the A,

which has been identified as an independent recognition site

for transcription initiation in several mammalian genes

(Corden et al., 1980). A TATA box and CCAAT box are located

at positions -31 and -77, respectively. In addition, the

following potential regulatory sites were identified based

on reported consensus sequences (Mitchell and Tjian, 1989;

Addison and Kurtz, 1986): AP-1 binding site (-1331), Oct-1

binding site (-1217), cAMP response element (-325),

glucocorticoid response element (-298 and -107), delayed

glucocorticoid response element (-82), serum response

element (-234 and -213), and AP-2 binding site (80 bp into

the first intron, not shown).

Structural Organization of the Lipo I Gene

Figure 2-5 shows the position of the exons relative to

the corresponding amino acid sequence. The four C-terminal

amino acid repeats have been aligned, and the more highly

conserved 17-amino acid units believed to be responsible for

calcium/phospholipid binding are underlined (Geisow et al,

1986; Pepinsky et al., 1988). Although the third 70-amino

acid repeat is encoded by exons 8, 9, and 10, no other

striking correlation between exon structure and the large

repeats can be seen. Three of the four 17-amino acid











o
411




0 *HC
0 H4






P40
E-4) U )





ONt

- Q)> -
H 4
r. a



H 4)
QO c
*d 4)
H r-c


H- 0
O )



0 0



> 0-H4





09 0
O
4J 4J





0wi




* O
) *to <
C4 0 -


a 4. -
4r o A
&M '
CO 10







g^d~







ii i ~
7.
7jI .I Ki
jB g




I I
*I










0



.0
;44-) W





rO ) C r ) (

H 0 -H 0
00 00
*: *r d

CO Q 0-
rH 0 -H 0
o0 c 4 -4
P 4.) -4 0
d 40 4.i -H
*0 0 4)



--4 $4 -4
-) 0 0 S-H 6 3
M Z-H V




.4 9 O -'"
P a O 4-)



SC
0-H CO 3






4 V. Cg 0
(3 c* 4 a, *O
*-I m


0 a) 4 0 "

0 043 ( -rCC

HO 4) 0p
0o n to 3 to
*O c 00


0- 0 0>
iH 0 H ) 14
r-4 0 0 -
-H0 *r( 4




*i fi ro-q *C
- 0 ( 0 43
044) ) 0O 9



0. 4 0 4 -HI

0 U5 3 1 l
in 4) -rH (0 0 0U

-1C 0 0 N
40 torq 04 )

fH43 rd ^ X C
in 3 H ri -H














































































a
O
0














U-
w 0











O











Z
45





.UJ





rU








27

repeats occur at a Lys-Gly junction which is at the end of

one exon and the start of the next. The third 17-amino acid

repeat, unlike the other three, is located entirely within

one exon. This pattern is essentially identical to that

reported for mouse p36, or lipo II, where the first, second,

and fourth 17-amino acid repeats occur at exon boundaries,

while the third is located in one exon. Interestingly, the

second exon encodes the N-terminal 22 amino acids, including

sites for phosphorylation, proteolysis, and head-to-head

dimerization (see intro. for ref.).

Discussion

In this chapter, I have shown the complete mouse lipo I

gene structure and about 1400 bp of upstream flanking

region. This enables an analysis of lipo I gene/protein

structure relationships, as well as a comparison with that

of mouse lipo II. A complete comparison, however, is not

yet possible because the 5' end and promoter region of lipo

II have not been identified (Amiguet et al., 1990). Such an

analysis is useful to determine whether lipo I fits the

hypothesis that exons encode discrete functional or

structural domains of proteins (Blake, 1983; Gilbert, 1978).

The lipo I protein can be divided into two regions, an

N-terminus unique to this protein and a core region

consisting of four 70-amino acid repeats. These structural

repeats are shared among the other lipocortins and are

believed to be necessary for calcium/phospholipid binding.








28

All four are encoded by three exons, and in three of the

four cases exon/exon boundaries show no striking correlation

with amino acid repeat boundaries. However, in the case of

the third 70-amino acid repeat, exons 8, 9, and 10

correspond almost exactly with the repeat boundaries. The

significance of this observation is not clear, although it

suggests that each repeat may have originally been encoded

by a combination of three separate exons whose boundaries

then became less defined during subsequent duplication. The

pattern seen in lipo I is very similar to that of lipo II in

terms of the 70-amino acid repeats. In lipo II, however,

the boundaries of the third repeat do not correspond quite

as well to the exon/exon boundaries.

Correlation of the 17-amino acid consensus sequences

within each repeat (Geisow et al., 1986) to exon structure

supports the idea of a common ancestor exon for the repeat

unit. As shown in Fig. 2-5, three of four of the 17-amino

acid repeats begin at a Lys-Gly which is at the end of one

exon and the start of the next. The third repeat is the

exception, because it is encoded in the central part of one

exon. This pattern is virtually identical to that seen in

lipo II, except that in this case the first repeat starts

two amino acids before the Lys-Gly (Amiguet et al., 1990).

It is interesting to note that the majority of the

unique N-terminus is encoded by exon two. This part of the

protein contains potential sites for tyrosine and serine








29

phosphorylation, proteolysis, and dimerization. Therefore,

it is possible that exon two is a lipo I-specific regulatory

exon responsible for mediating its particular activity.

This correlates with the fact that the putative second and

third exons of lipo II encode regions believed important in

its regulation (Amiguet et al., 1990). Lipo II exon two

encodes the site for binding of a protein called pll, while

exon three encodes phosphorylation and proteolytic cleavage

sites (Glenney and Tack, 1985). Mitsunobu et al. (1989)

have isolated a cDNA from GL-5-JCK human glioblastoma cells

in which the region coding for the first 16 amino acids of

lipo II was fused to the coding region for the 370 C-

terminal amino acids of the c-raf-1 oncogene/kinase.

Expression of this lipo II/c-raf-1 fusion protein resulted

in transformation of NIH3T3 cells. The first 16 amino acids

of lipo II, encoded entirely by the putative second exon,

contain the binding site for pll. Binding of pll to lipo II

results in the formation of a p362pll2 tetramer which is

anchored to the cytoskeleton (Zokas and Glenney, 1987), so

it is possible that aberrant localization of the c-raf-1

kinase to the cytoskeleton, mediated by the pll binding site

encoded by the lipo II second exon, caused the cell

transformation observed by Mitsunobu et al. (1989).

Further analysis between the genomic structures of lipo

I and lipo II leaves little doubt of their common ancestry.

If one assumes that the -52 bp of noncoding sequence








30

(indicated by primer extension) not identified in the lipo

II gene make up its first exon (Amiguet et al., 1990; Saris

et al., 1986), then the two genomic structures are nearly

identical. Both would consist of 13 exons, 6 of which are

exactly the same in length. The shared gene structure

supports the evolution of these two genes from a common 13

exon-containing ancestral gene, rather than the co-evolution

of the genes from a duplication of an ancestral repeat-

encoding exon. This theory would also account for the

origin of calelectrin/p68 from a duplication of the 13 exon

ancestor to form the eight-repeat structure of p68. The

verification of this hypothesis must await the

characterization of gene structures for the remaining

members of the lipocortin family, including p68.

Of the potential regulatory sequences identified in the

1400 bp upstream of the transcription initiation site, the

most interesting are the GREs. Dexamethasone (DEX) has been

shown to induce lipo I mRNA levels in several cell types

(Phillips et al., 1989; Wallner et al., 1986; Wong et al.,

submitted), although it is clear that this response does not

occur in all cells investigated (Bienkowski et al., 1989;

Bronnegard et al., 1988; Hullin et al., 1989; Isacke et al.,

1989; Northup et al., 1988; William et al., 1988). Whether

this specificity is due solely to the presence or absence of

glucocorticoid receptors, or involves other factors, is

presently unclear. Especially intriguing is the region at








31

positions -82 to -42, which resembles the delayed

glucocorticoid response element (sGRE) of the rat alphas-

globulin promoter (Addison and Kurtz, 1986). This 45 bp

region was shown to mediate induction of transcription by

DEX. This induction, unlike those termed rapid primary

responses, required protein synthesis. A response which is

blocked by cycloheximide and usually shows a lag between

administration of the hormone and initiation of the response

has been called an indirect or secondary response (Chen and

Feigelson, 1980; Hess et al., 1990; Klein et al., 1988; Paek

and Axel, 1987; Widman and Chasin, 1982). The induction of

lipo I mRNA in the mouse pre-adipocyte line 3T3-L1 by DEX

both exhibits a lag time of about two hr and is blocked by

cycloheximide (Wong et al., submitted). Therefore, it is

possible that the delayed response to DEX observed in 3T3-

L1 cells is mediated by the putative sGRE in the lipo I

promoter.

Other potential regulatory sites of interest include

the two serum response-like elements. Different lipo I mRNA

expression levels were observed in several tissue culture

cells depending on the presence or absence of serum (K.

Horlick and W. Wong, unpublished data). In agreement with

this, the mRNA for lipo II (calpactin I) was shown to be

induced as a primary response to serum stimulation of

quiescent A31 cells (Keutzer and Hirschhorn, 1990). Also,

as documented in Chapter 4, an induction of lipo I mRNA in








32

the rat heart after injection of glucagon was observed.

Since glucagon action is mediated in many cases by increased

levels of cAMP, it is possible that the cAMP response

element identified in the lipo I promoter mediates this

tissue-specific response. In support of this, lipo I mRNA

levels are altered in 3T3-L1 cells after treatment with

methylisobutyl xanthine, a phosphodiesterase inhibitor which

leads to increased intracellular cAMP levels (Parsons et

al., 1988; W. Wong, unpublished data). Other experiments

shown in Chapter 4, however, do not support a role for cAMP

in the induction of heart lipo I mRNA by glucagon.

In conclusion, I have described the genomic structure

of mouse lipo I and compared it to the known portion of the

lipo II locus. The two closely related genes are unlinked

and located on chromosomes 19 and 9, respectively (Amiguet.

et al., 1990; Horlick et al., submitted). It will be of

interest to determine whether the other members of the

lipocortin family show a similar intron/exon organization,

further implicating a common gene ancestor which gave rise

to these proteins.














CHAPTER 3

REGULATION OF LIPOCORTIN I IN RAT MAMMARY TISSUE
DURING GLANDULAR DIFFERENTIATION AND INVOLUTION

Introduction

The mammary gland in both humans and rodents is a

complex tissue exhibiting dramatic morphologic and

functional changes during pregnancy and lactation in females

(Borellini and Oka, 1989; Hollmann, 1974; Topper and

Freeman, 1980). The gland consists of several cell types,

including fat cells, epithelial cells, and myoepithelial

cells, whose number and morphology change according to

developmental status.

During sexual maturation, epithelial cells grow

extensively to form the tree-shaped ductal system

characteristic of mammary tissue. Upon sexual maturation,

this ductal development stops even though the major hormonal

stimuli for growth are still present. At this stage the

epithelial cells are dormant, arrested in the G, phase of

the cell cycle. Pregnancy causes hormonal changes which

relieve mammary epithelial cells of growth constraints, and

consequently cell replication resumes in the first half of

pregnancy. In middle to late pregnancy, epithelial

proliferation continues, and differentiation into secretary

cells, or alveoli, takes place. As term approaches, alveoli

33








34

begin to accumulate secretary products and increase

cytoplasmic size.

In the lactating gland, secretary cells are highly

polarized with respect to both morphology and function. A

large part of the cytoplasm is occupied by rough endoplasmic

reticulum and a well-developed Golgi apparatus is located in

the apical region. Alveoli manufacture and secrete large

quantities of protein, fat, and carbohydrate.

As suckling decreases during natural weaning,

involution of the gland begins. At this time, secretary

alveoli develop large vacuoles containing casein micelles

and large fat droplets. As weaning progresses, these

vacuoles fuse with hydrolytic lysosomes, leading to

autophagy. The majority of epithelia are destroyed, but

some survive to become resting cells, like those in the

mature virgin gland, which will proliferate and

differentiate upon the next pregnancy.

The hormonal determinants of mammary gland development

are complex and beyond the scope of this introduction. They

include insulin, cortisol, prolactin, estrogen, epidermal

growth factor, transforming growth factor beta, thyroid

hormones, retinoids, and vitamin D3 (Borellini and Oka,

1989). In addition, other factors affecting mammary growth

include calcium, polyamines, and poly(ADP-ribose).

A lipo I-like gene was studied in the pigeon cropsac,

an organ which displays the same hormonal specificity as the








35

mammary gland, but is substantially simpler (Horseman, 1989;

Pukac and Horseman, 1986). In the cropsac, a homogenous

population of pseudostratified squamous epithelia

proliferates and differentiates after prolactin treatment.

As mentioned above, prolactin is very important in

establishing and maintaining lactation in the mammary gland.

The differentiated epithelia of the cropsac synthesize and

store large amounts of lipoprotein globules which are used

to nourish offspring.

Several genes which were induced in the cropsac after

intradermal injection of prolactin were cloned and

characterized (Pukac and Horseman, 1986). The mRNA for the

major prolactin-induced gene was stimulated as early as

three hr after injection of the hormone and peaked at levels

about 70-fold higher than before injection. This mRNA

encoded a polypeptide of molecular weight 35,500. When the

nucleotide sequence of this prolactin-induced cDNA was

determined, it was found to be highly similar to that of

mammalian lipo I (Horseman, 1989). The only region of

substantial sequence difference was a domain in mammalian

lipo I encoding amino acids 20-40, including a tyrosine

phosphorylation site and other residues believed to be

physiologically important. It was concluded that

lipocortins might be regulated by prolactin in other

tissues, and that the difference in sequence near the N-

termini of mammalian and avian lipocortins might represent a

divergence yielding different regulatory mechanisms.








36

Lozano et al. (1989b) studied several mammary gland

calcium-binding proteins in the mouse mammary gland during

development. Some of these proteins had previously been

identified as calelectrins and calpactin I/lipocortin II

(Hom et al., 1988). It was found that the mammary gland

calcium-binding proteins were developmentally regulated:

They were present in large amounts in the mature virgin

gland, but were virtually absent in the lactating gland.

Immunolocalization revealed that ductal epithelial cells in

the virgin gland were rich in calelectrins and calpactin I,

while myoepithelia did not contain them. The secretary

alveolar epithelial cells of the lactating gland, in

contrast to epithelia from the virgin gland, were

essentially devoid of these proteins. Based on these

results, the authors seriously questioned a role for

calelectrin and calpactin I in casein exocytosis. They

previously had proposed that these calcium binding proteins

might be involved in epithelial cell spreading, including

cell motility or cell shape changes (Braslau et al., 1984;

Rocha et al., 1986).

This chapter shows studies designed to investigate lipo

I mRNA and protein expression in the rat mammary gland

during natural differentiation and involution, as well as

after premature termination of lactation by pup removal.

Since two other annexins were shown to be developmentally

regulated in the mouse mammary gland as described above, it








37

would be important to determine if lipo I would exhibit

regulation in mammary tissue. Another question these

studies address is whether or not mammalian lipo I might be

induced by prolactin, similar to the lipo I-like pigeon gene

in the cropsac. The profile of prolactin expression in the

rat during pregnancy and lactation is known, and could be

compared to expression of lipo I.

The results described in this chapter indicate that

both lipo I mRNA and protein levels show striking

alterations in the mammary gland of dams during normal

differentiation and involution, as well as following

premature pup removal. In contrast to the pigeon lipo I-

like gene, rat mammary lipo I does not appear to be induced

by prolactin, and is in fact inversely related to predicted

prolactin levels.

Materials and Methods

Animals

Timed-pregnant Sprague-Dawley rats were obtained from

Charles River Laboratories.

RNA Isolation

Total RNA was isolated by homogenizing tissues in a

guanidinium thiocyanate-sarcosyl solution followed by acid

phenol/chloroform extraction and ethanol precipitation as

described previously (Chomczynski and Sacchi, 1987).

Typically, about 1 g of tissue was lysed in 5 ml of

thiocyanate solution containing 4 M guanidine thiocyanate,








38

25 mM sodium citrate (pH 7.0), 0.5% sarcosyl, 100 mM 2-

mercaptoethanol. To this solution, 300 pl of 2 M sodium

acetate (pH 4.0), 3 ml of H20-saturated phenol, and 1 ml of

chloroform:isoamyl alcohol (24:1) were added in stepwise

fashion. The mixture was centrifuged at 10,000 x g for 10

min to separate phases. The aqueous phase was mixed with 3

ml of isopropanol and total RNA precipitated by incubating

the mixture at -20C for at least 1 hr, followed by

centrifugation at 10,000 x g for 30 min. The RNA pellet was

then resuspended in 600 pl of guanidine thiocyanate solution

and further purified by isopropanol and ethanol

precipitation.

Northern Analysis

An aliquot of 20 pg total RNA was loaded on a 1% MOPS (3-

(N-morpholino)- propanesulphonic acid)-formaldehyde agarose

gel as described previously (Sambrook et al., 1982). The

gel was run for 12-16 hr at 40 volts with constant buffer

recirculation. The gel was then subjected to three 30-min

washes with water, one 45-min wash in 0.05 N NaOH and 0.01 M

NaC1, one 45-min wash in 0.1 M Tris-HCl (pH 8.0), and a 30-

min wash in a solution containing 0.089 M Tris-borate, 0.089

M boric acid, and 0.002 M EDTA, pH 8.0. The RNA was

electrophoretically transferred from the gel to a noncharged

nylon membrane (GeneScreen, NEN) for 1 hr. The RNA was

immobilized on the membrane by cross-linking with UV light

(Church and Gilbert, 1984) and analyzed with a specific

radiolabeled probe.










Probe Synthesis

A mouse lipocortin I cDNA provided by William Wong of

this laboratory, a rat Cu/Zn superoxide dismutase cDNA

provided by Jan-Ling Hsu of this laboratory, and a mouse

histone H4 cDNA (Seiler-Tuyns and Birnstiel, 1981) were used

as probes. Probes labelled with a-"P dATP (NEN,

800uCi/mmol) were synthesized either by random primer

extension, as described in Chapter 2, or by an M13 single-

stranded probe synthesis method using an M13mpl8 clone

containing the mouse lipo I cDNA insert in the appropriate

orientation to produce a probe complementary to mRNA (Church

and Gilbert, 1984). Hybridization was carried out as

described in Chapter 2.

Western Analysis

Mammary tissue was weighed and homogenized in 10 ml of

20 mM Tris-HCl (pH 7.4), 1 mM EDTA, 255 mM sucrose

supplemented with 1% aprotinin. A polytron was used for

homogenization (two times at half-maximal speed for 10 sec

each). After the suspension was spun at 212,000 X g for 70

min at 4C, a Markwell protein assay was performed on the

supernatant. 100 Ag of total protein was run on a 10%

polyacrylamide reducing gel and transferred to

nitrocellulose. The membrane was blocked for 3 hr using 20

mM Tris-HCl (pH 7.4), 0.5 M NaCl (TBS) supplemented with 3%

BSA and 0.05% Tween and incubated for 16 hr at 4C with an

anti-lipocortin I antibody (Huang et al., 1986) at 1:1000









40

dilution in 3% BSA/TBS. The membrane was washed for 1 hr

with 1% BSA/TBS and then incubated with 1 X 107 cpm of I125

labeled protein A for 1 hr. After three 20-min washes in 25

ml TBS, the membrane was air-dried and exposed for 3-5 days

to Kodak X-OMAT film.

Results

To investigate lipo I mRNA levels in the rat mammary

gland during differentiation and involution, I carried out

Northern analysis using a mouse lipo I cDNA probe (Fig. 3-

1). Messenger RNA levels from mature virgin animals were

compared with those from animals which were either pregnant,

lactating, or weaned. The mRNA was very abundant in the

nonpregnant state. At midpregnancy, a large decline was

observed, corresponding to the time when epithelial cells

proliferate and differentiate (Topper and Freeman, 1980).

By day 22 of pregnancy (1-2 days before birth), mRNA levels

decreased at least 50-fold as compared to virgin animals.

At birth a small, but reproducible increase occurs. With

the onset of lactation, lipo I mRNA was undetectable and

remained so for the remainder of the suckling period. After

weaning had occurred (30 days following birth), the mRNA

rose to levels comparable to those seen in virgin animals.

To illustrate the time during pregnancy when epithelial

cells begin proliferating, I also determined histone H4 mRNA

levels, which increase during DNA synthesis and cell

proliferation (Seiler-Tuyns and Birnstiel, 1981). A rise in



















PREGNANT
NP 7 11 15 19 22 B 1 3


LACTATING
7 15 19 22 W


Upocortin I


C4/Zn SOD
















Figure 3-1. Regulation of mammary lipocortin I mRNA during
pregnancy, lactation, and involution. Total RNA was
isolated from mammary tissue of female rats which were
mature virgins (NP), pregnant for various lengths of time
(days), had just given birth (B), had suckled pups for
various lengths of time (days), or after weaning (W, 30 days
after birth). Birth usually occurs on day 23. Northern
blot analysis was carried out with a cDNA probe for mouse
lipocortin I, then the blot was stripped and reprobed with
mouse histone H4 and rat copper/zinc superoxide dismutase
(SOD) cDNAs.


I








42

histone H4 is apparent at day 15 of pregnancy, when mammary

epithelial cells begin proliferation and differentiation,

and also after weaning, during glandular involution.

Copper/zinc superoxide dismutase mRNA levels are illustrated

as an internal control, and show only a small variation over

the same time course.

Figure 3-2 focuses on the decrease in steady-state lipo

I mRNA levels in mammary tissue of dams following the start

of suckling. Two animals are shown to demonstrate

reproducibility. The decline in lipo I mRNA can be compared

to levels of copper/zinc superoxide dismutase, which remain

relatively constant during the experiment.

To investigate whether the lipo I protein followed a

similar expression pattern, extracted mammary tissue

proteins were analyzed by immunoblotting (Fig. 3-3). Also

shown are mammary lipo I mRNA levels from the same animals.

The lipo I-specific antibody revealed abundant levels of

lipo I protein in the virgin rat gland. Like the mRNA, the

protein declined during pregnancy, was nearly undetected

during lactation, and increased at weaning. Densitometric

analysis of protein and mRNA levels shows a close

correspondence between the two (Fig. 3-4).

Since the lipo I mRNA increased rapidly upon weaning,

the role of the suckling stimulus in lipo I regulation was

investigated. Messenger RNA levels were determined in

mammary tissue of dams which had pups withdrawn at birth.
























0 0 1 1 2 2 3 3


Upo




Cu/Zn SOD


Figure 3-2. Decrease in mammary lipocortin I mRNA levels
with the onset of lactation. Total RNA was isolated from
mammary tissue of animals which had just given birth or had
suckled pups for one, two, or three days. Each
determination was performed in duplicate. Northern analysis
was carried out using mouse lipocortin I and rat Cu/Zn SOD
cDNA probes.


doys after birth





















NP 15P 22P B 1L 7L 15. W


mRNA





protein


Figure 3-3. Regulation of lipocortin I mRNA and protein
levels in rat mammary tissue during glandular
differentiation and involution. Total RNA and protein were
isolated from mammary tissue of rats as described in Figure
3-1. P, pregnant; L, lactating. Northern analysis and
immunoblotting were then carried out using a mouse
lipocortin I cDNA probe and anti-lipocortin I antibody,
respectively.








45







70 24
S22
60-- ..20

< 50- 0 18 z
n -16z W
E 740- 14 o
z/ -.12
* 30-- -10 o
T 8 1
* 20 6 o 0


0 0
NP 15P 22P B 1L 7L 15L W






















Figure 3-4. Densitometric analysis of the data in Figure 3-
3. A Bio Image Visage 60 image processor was used for
quantitation of lipocortin I mRNA and protein levels in
Figure 3-3.








46

Figure 3-5 shows that as early as one day after pups were

removed, lipo I mRNA levels dramatically increased over the

level at birth. Actin mRNA rose to a lesser degree, while

levels of copper/zinc superoxide dismutase mRNA were

relatively constant during the experiment.

To further investigate the role of the suckling

stimulus, lipo I mRNA levels were analyzed in mammary tissue

of dams following pup withdrawal at birth and after seven

days of suckling (Fig. 3-6). Also, time points prior to 24

hr after withdrawal were examined in order to determine how

fast the induction of lipo I mRNA was. If pups are removed

at birth, completely preventing the suckling stimulus,

mammary lipo I mRNA levels increased significantly in the

dams as early as 6 hr after removal, reaching a plateau by

12 hr. If pups were withdrawn after seven days of suckling,

the increase in lipo I mRNA was delayed by several hours and

peaked by 24 hr. In contrast, lipo I protein levels were

low to undetectable for more than 24 hr after pups were

removed either at birth or after seven days of suckling, but

by five days after removal the protein had reached maximal

levels.

To examine the specificity of lipo I regulation

following pup withdrawal, lipo I mRNA levels were next

measured in the mammary gland and several other tissues from

animals whose pups were withdrawn 15 days following birth

(Fig. 3-7). This experiment demonstrated that lipo I mRNA

























0 0 1 3 5 7 9


lipo I





actin



Cu/Zn SOD


Figure 3-5. The role of the suckling stimulus directly after
birth on lipocortin I mRNA levels in mammary tissue. Pups
were removed at birth and total RNA was isolated from
mammary tissue of dams at 0, 1, 3, 5, 7, and 9 days
following pup withdrawal. Northern blot analysis was done
using lipocortin I, actin, and Cu/Zn SOD cDNA probes as
described previously.


days after pup removal






















pups removed at:
time after removal:


mRNA



protein


Bith 7 dqy
0 6 12 24 Sd 0 6 12 24 5d


Figure 3-6. Induction of lipocortin I mRNA and protein
expression in mammary tissue of dams following pup removal
at birth or day seven of suckling. Pups were removed from
the dams at birth or following seven days of suckling, and
total RNA and protein were isolated from mammary tissue of
dams at 0, 6, 12, 24, and 120 hr/5 days following pup
withdrawal. Northern analysis and immunoblotting were
carried out as described in the legend to Figure 3-3.






























mammary tum
time after
removal: 0 6 12 24 5d


HWart KMy Muscle
0 6 12 24 Sd 0 6 12 24 Sd 0 6 12 24 Sd


Figure 3-7. Specificity of the mammary lipocortin I mRNA
regulation following pup withdrawal. Total RNA was isolated
from various tissues of dams whose pups were removed
following 15 days of suckling. RNA was prepared 0, 6, 12,
24, and 120 hr/5 days after pup withdrawal. Northern
analysis was done using a mouse lipocortin I cDNA probe.








50

levels were constant in the heart, kidney, and hindleg

muscle of dams, while there was a significant rise in

mammary tissue expression over the five-day withdrawal

period. Thus, the regulation of lipo I mRNA in the mammary

gland after pup removal was a tissue-specific phenomenon.

The increase in lipo mRNA levels over the five-day period

following pup withdrawal, shown in Fig. 3-6 and 3-7, was

between 60- and 80-fold (Fig. 3-8).

Discussion

It is likely that hormonal changes during pregnancy and

lactation influence the pattern of mammary lipo I mRNA and

protein expression. Based on the awareness of a prolactin-

inducible, lipo I-like gene in the pigeon cropsac, I

explored the effects of prolactin on mammalian lipo I gene

expression. These experiments involved the use of the

midpregnancy-derived mammary epithelial cell line HC11,

which had been clonally selected for its responsiveness to

prolactin (Ball et al., 1988). Treatment of these cells

with prolactin, however, caused no change in lipo I mRNA

expression, which was also the case when mammary explants

were treated with prolactin in culture (data not shown).

Furthermore, in the rat, circulating levels of prolactin are

low until one day before birth, when they rise dramatically

(Rosenblatt et al., 1979). Prolactin levels remain high

during suckling and gradually decline as weaning approaches.

In contrast, lipo I mRNA and protein levels are high in the















80

O-

60-



40-



20-



0-


6 12 24 5d


0 6 12 24 5d


FIGURE 3-8. Densitometric analysis of the increase in
lipocortin I mRNA levels following pup withdrawal.
Quantitation of the mRNA data in Figures 3-6 and 3-7 was
done using a Bio Image Visage 60 image processor. The first
five bars (open) represent pup removal at birth and show a
30-fold mRNA increase from birth to five days after pup
removal. The next five bars (diagonal lines) represent pup
removal after seven days of suckling and exhibit a 120-fold
mRNA increase over the five-day period after pup withdrawal.
The last five bars (X's), which represent pup removal after
15 days of suckling, show a mRNA induction of 320-fold over
the time course.


0 6 12 24 5d


r








52

virgin gland and during the first half of pregnancy, and low

to undetectable at birth and during lactation. This

indicates that if prolactin is involved in regulating

mammary lipo I expression, it would do so by down-

regulation, which I did not observe in HCll cells or mammary

explants. Future studies investigating mammary epithelial

cell lines for hormonal effects on lipo I expression may aid

in understanding the regulation I observed in the mammary

gland during differentiation and involution.

Although prolactin did not affect expression of lipo I

mRNA in the mouse mammary epithelial cell line HC11, I did

observe high levels of the message (data not shown). Thus,

it is likely that the resting ductal epithelial cells, which

give rise to secretary cells during pregnancy, are

responsible for the high lipo I expression observed in the

virgin gland. Immunolocalization of two other annexins,

annexin II and calelectrin, in virgin mammary epithelial

cells by Rocha and coworkers supports this hypothesis (Hom

et al., 1988; Lozano et al., 1989a; Lozano et al., 1989b).

In contrast, secretary cells of the lactating gland were

essentially devoid of these proteins (Lozano et al., 1989b).

The ductal cells, which are dormant in the mature virgin

animal and reported to be arrested in the G, phase of the

cell cycle, begin proliferation and differentiation into

secretary cells around midpregnancy (Topper and Freeman,

1980). The striking decline in lipo I mRNA and protein








53

levels that I observed in the rat mammary gland during

pregnancy corresponds to the onset of these changes in

epithelial cell morphology. I believe that the high level

of lipo I in the virgin mammary gland may be involved in

inhibiting epithelial cell proliferation and

differentiation. This leads to a consideration of what

regulates lipo I. It is well known that epidermal growth

factor (EGF) is important in signaling mammary gland growth

during pregnancy (Taketani and Oka, 1983). The EGF

receptor, which is known to phosphorylate lipo I (Sawyer and

Cohen, 1985), reaches its highest levels in the mammary

gland around midpregnancy (Edery et al., 1985). Perhaps

phosphorylation of lipo I during pregnancy alters its

activity, which is followed by a subsequent down-regulation

of lipo I expression. Although it has been demonstrated

that serine/threonine phosphorylation of lipo I blocks its

phospholipase-inhibitory activity and alters its

calcium/phospholipid binding properties (Ando et al., 1985;

Hirata, 1981; Hirata et al., 1984; Touqui et al., 1986), it

is yet to be investigated whether tyrosine phosphorylation,

too, is inhibitory. Nevertheless, this event may be

necessary to initiate epithelial cell proliferation and

differentiation.

Soon after natural weaning or premature termination of

the suckling stimulus by pup removal, a subpopulation of the

mammary secretary epithelial cells reverts back to the








54

resting state, while the remainder of the cells die

(Borellini and Oka, 1989; Hollmann, 1974). No irreversible

changes in the gland occur immediately after cessation of

suckling, because lactation can recommence several days

after weaning (Silver, 1956). The subset of epithelial

cells which reverts to the resting state may be responsible

for the rapid increase in mammary lipo I mRNA after removal

of the suckling stimulus. These cells may receive a signal

from the sympathetic nervous system when suckling ceases

which causes a rapid transcriptional activation of lipo I in

preparation for return to the resting state. As shown in

Fig. 3-6, translation of lipo I mRNA appears to be delayed

and requires more than 24 hours after cessation of suckling.

This delay may correspond to the time when glandular

involution becomes a committed event, encompassing

epithelial cell death. This appears to occur approximately

five days after weaning (Silver, 1956), and therefore

correlates with the time at which maximal levels of lipo I

are reached. I believe that the high lipo I levels observed

in the involuted gland following natural weaning (Fig. 3-1)

indicate the presence of new resting epithelial cells,

which, like those in the mature virgin gland, may undergo

proliferation and differentiation with the onset of a new

pregnancy.














CHAPTER 4

TISSUE-SPECIFIC REGULATION OF LIPOCORTIN I mRNA
LEVELS IN RATS DURING DIABETES AND BY GLUCAGON

Introduction

Diabetes mellitis, a serious and prevalent disease

among humans, is known to result in a wide variety of

physiological alterations which often may result in death

(Warren et al., 1966). Numerous qualitative and

quantitative alterations take place in the pancreas, the

organ in which the insulin-producing islet cells are

located. Cardiovascular disease is one of the leading

causes of death in diabetics, with both heart dysfunction

and atherosclerosis frequently developing. Renal

alterations include a marked hypertrophy with corresponding

increase in kidney weight, as well as an increased

glomerular filtration rate. Blindness may result from

retinopathy, such as microaneurysms. Nervous system damage,

most notably in peripheral nerves, results in symptoms

including pain and loss of sense in the feet and legs.

Gangrene in the extremities, bone and skin lesions, muscle

atrophy, liver damage, and gastrointestinal problems often

occur in diabetics.

Many studies regarding diabetes have relied upon animal

models, including rats (Fein et al., 1980), rabbits (Bhimji

55








56

et al., 1985), and dogs (Regan et al., 1981). Administration

of alloxan (Dunn et al., 1943) or the antibiotic

streptozotocin (STZ)(Junod et al., 1967) has been shown to

cause impaired f-cell function and hypoinsulinemia in young

animals, thus simulating the diabetic state. In addition,

there are Chinese hamster (Yerganian et al., 1957), mouse

(Bielschowski and Bielschowski, 1956), and rat (Rodrigues

and McNeill, 1990) inbred strains which develop spontaneous,

hereditary diabetes.

Several biochemical alterations have been observed in

tissues from diabetic animals. Fatty acid compositions of

phospholipids of heart, kidney, aorta, and serum undergo

changes in both human (Chase et al., 1979) and animal (Faas

and Carter, 1980) diabetes. Specifically, linoleic acid was

found to be increased and arachidonic acid to be decreased

in most tissues as compared to nondiabetic controls (Faas

and Carter, 1980; Holman et al., 1983; Schrade et al.,

1963). Physical properties of cell membranes, such as

fluidity, may be altered in diabetic tissues (Cheung et al.,

1980; Shinitzky, 1980). Recent studies have focused on the

diminished arachidonic acid levels and altered arachidonate

metabolism in heart muscle of STZ-diabetic rats

(Gudbjarnason et al., 1987; Pieper, 1990). Since

alterations in arachidonate-derived products such as

prostaglandins and leukotrienes have been implicated in

heart dysfunction in diabetic rats (Johnson et al., 1984),








57

it is possible that changes in arachidonate metabolism could

play a critical role in diabetic heart disease.

Many other biochemical changes in the diabetic heart

have been documented. The P-adrenergic receptor number in

cardiac membranes was lowered and their coupling to

adenylate cyclase was altered in STZ-diabetic rats (Atkins

et al., 1985). The 1,2-diacylglycerol content in myocardium

from STZ-diabetic rats was highly elevated (Okumura et al.,

1988). Significant reductions in calcium binding and uptake

have been observed in sarcoplasmic reticulum sections from

diabetic rat hearts (Bergh et al., 1988; Bielefeld et al.,

1983; Fein et al., 1980; Ganguly et al., 1983; Penpargkul et

al., 1979). Both phosphoinositide synthesis and release

from membrane phospholipids were greatly reduced in diabetic

rat hearts (Bergh et al., 1988). In addition to the heart,

reduced inositol content and turnover have been observed in

diabetic nerve, lens, and kidney (Craven and DeRubertis,

1989; Winegrad, 1987).

Studies reported in this chapter were carried out to

investigate the possible role of lipo I in the altered

metabolism associated with diabetes. Lipocortin I has been

implicated as a regulator of arachidonic acid metabolism

through modulation of phospholipase A2 activity (Hirata et

al., 1980). As noted above, levels of arachidonic acid and

some of its metabolites were shown to be altered in several

tissues of diabetic animals. The phospholipid-binding








58

activity of lipo I is known to be calcium-dependent

(Glenney, 1986), and significant changes in calcium levels

in some diabetic tissues have been documented. Finally, the

closely related protein lipocortin III was recently

identified as a hydrolase involved in regulation of

phosphoinositide turnover, suggesting the function of lipo I

may also be to regulate this important pathway. As

described above, alterations in phosphoinositide metabolism

have been documented in several diabetic tissues.

After demonstrating tissue-specific induction of lipo I

in STZ-diabetic rats, certain hormones were investistigated

to determine which ones might play a role in lipo I

induction. It was found that glucagon, which is highly

elevated in diabetic animals (Unger et al., 1963), caused

induction of lipo I in the heart as early as 4 hr after

injection. Lipocortin I mRNA expression in two other

tissues in which induction was seen in diabetic animals was

not affected by glucagon up to 24 hr after injection.

Possible reasons for the relatively rapid induction of lipo

I mRNA expression in the heart by glucagon, and induction in

heart and other tissues in diabetic animals, will be

discussed. These studies are the first to link glucagon and

diabetes with any of the annexins.

Materials and Methods

STZ-Diabetes and Insulin Treatment

Male Sprague-Dawley rats (100-150 grams) were injected

intraperitoneally with 65 mg/Kg streptozotocin (Aldrich),









59

which was dissolved in 0.05 M sodium citrate buffer, pH 4.5.

After 24 hr, urine was analyzed using Boehringer Mannheim

Chemstrips to detect glucosuria and ketosis. Animals

exhibiting these symptoms, as well as loss of body weight,

were sacrificed after three-four days and tissues isolated

for RNA extraction. Another group of animals, made diabetic

as described above, was injected once daily for seven days

with 2 U of NPH beef insulin (Squibb). Urine glucose and

ketone levels returned to normal during insulin treatment.

Two animals maintained on insulin were sacrificed following

the final injection, while two others were taken off insulin

for three days to allow diabetic symptoms to again develop.

The normal state of insulin-treated rats, and diabetic state

of rats removed from insulin treatment, were confirmed by

both urine analysis and blood glucose determination (Sigma).

Hormones

Glucagon, a kind gift of Lilly Research Laboratories,

was used at a concentration of 1 mg/Kg body weight.

Dibutyral cAMP (Sigma) was used at 50 mg/Kg, dexamethasone

phosphate (Quad) at 10 mg/Kg, epinephrine (Sigma) at 1

mg/Kg, and insulin at 2 mg/Kg. Glucagon, epinephrine, and

insulin were dissolved at a concentration of 1 mg/ml in a

buffer containing 1% bovine serum albumin and 1.6% glycerol,

pH 2.5-3.0.









60

RNA Isolation. Northern Analysis, Probe Preparation, and
Hybridization

These procedures are described in Materials and Methods

in Chapters 2 and 3. A mouse P-actin cDNA was kindly

provided by Larry Kedes.

Results

To determine if changes in lipo I mRNA expression occur

in tissues of diabetic rats, I injected rats with STZ to

induce diabetes and compared tissue lipo I mRNA levels with

those of controls. Figure 4-1 shows Northern analysis

comparing lipo I mRNA levels in several tissues from a

diabetic and control rat. While levels in the brain,

intestine, lung, and thymus were similar, a significant

increase was observed in the heart, kidney, and hindleg

muscle. Actin mRNA levels are shown as an internal control.

Note that the mouse P-actin cDNA probe used hybridized to

two mRNA species of different sizes in heart and hindleg

muscle.

To demonstrate the reproducibility of the induction of

lipo I mRNA in the heart, kidney, and hindleg muscle, I

repeated the above experiment using two other STZ-diabetic

rats. Figure 4-2 further illustrates this induction, and

also shows that lipo I mRNA levels in the liver and

intestine are not affected by acute diabetes. Again, actin

mRNA levels are shown for comparison.

To investigate whether the induction of lipo I mRNA in

the heart, kidney, and hindleg muscle was truly a result of

















1 /


'O ^* ;


SDCDCD /
D C D C D


Figure 4-1. Expression of lipocortin I mRNA in tissues from
a control and three-day streptozotocin-diabetic rat. Rats
were injected intraperitoneally with buffer only or 65 mg/kg
streptozotocin and tissue RNA was isolated and subject to
Northern analysis after three days. A mouse P-actin probe
was used as an internal control.


CD


D
D


lipo I



actin

























Heart Kidney
C D1 D2 C D1 D2 C


*_ Intestine
02 C D1 D2


lipo I



actin





















Figure 4-2. Increased levels of lipocortin I mRNA in the
heart, kidney, and hindleg muscle of three-day
streptozotocin-diabetic rats. Northern analysis was carried
out as described in the Figure 4-1 legend.








63

the diabetic state and not caused by another effect of STZ,

I carried out an experiment in which insulin was used to

control the disease (Fig. 4-3). Rats were injected with STZ

to induce diabetes. After three days, when both urine

glucose and ketones were very high, animals were subject to

daily injections of 2 U NPH beef insulin for the following

seven days. One day after insulin treatment, urine glucose

and ketone levels approached normal and remained so until

the end of the treatment. Following the seven-day insulin

therapy, two animals were sacrificed and total RNA was

isolated from tissues. Two other rats were removed from

insulin treatment, and as a consequence urine glucose and

ketones rose within a day. Three days after stopping

insulin treatment, animals were sacrificed and total RNA was

isolated from tissues. Northern analysis (Fig. 4-3a) showed

that in the heart and kidney, lipo I mRNA levels increased

when animals were removed from insulin treatment and

exhibited diabetic symptoms. In contrast, levels in hindleg

muscle remained the same. To confirm diabetes and show that

insulin treatment was effective, blood samples were taken

and plasma glucose levels determined (Fig. 4-3b). While

blood glucose levels in the insulin-treated rats were only

slightly above the normal range, levels in the animals taken

off insulin were clearly indicative of diabetes.

I next investigated hormones which might play a role in

the altered lipo I expression in diabetic rat tissues. A









64










heart kidney muscle
Insulin on off on off on off

a .lipo IO















On i^ '
Ionsun lictin

300

400--




S100.

0-
On Off
Insulin Insulin










Figure 4-3. Heart, kidney, and hindleg muscle lipocortin I
mRNA expression in streptozotocin-injected rats on and off
insulin treatment. a) Three days after intraperitoneal
injection of 65 mg/kg streptozotocin, rats were given daily
injections of 2 U NPH beef insulin for seven days.
Following the final insulin injection, two rats were
sacrificed and tissue RNA isolated. Two other rats were
removed from the insulin treatment for three days before
isolating tissue RNA. Northern analysis shows lipocortin I
and actin mRNA levels. b) Serum glucose levels in the rats
represented above were determined using a Sigma kit and are
shown graphically. The normal range is around 150-200
mg/dL.








65

preliminary study using animals which had been injected with

glucagon, insulin, and dexamethasone revealed that lipo I

mRNA was induced in the heart 4 hr after injection of

glucagon (Jan-Ling Hsu and K. Horlick, unpublished results).

Glucagon, which is known to mediate cellular effects via

increased cAMP levels and calcium mobilization (Wakelam et

al., 1986), is highly elevated during diabetes (Unger et

al., 1963). To further study a potential link between

glucagon and lipo I induction, I injected rats with glucagon

and assayed lipo I mRNA levels in several tissues after 4

hr. Figure 4-4 demonstrates that glucagon injection caused

an increase in lipo I mRNA levels in the heart, but not in

kidney or muscle, after 4 and 24 hr. Figure 4-5 shows

glucagon induction of lipo I mRNA levels in the heart of

four animals, representing an increase of between two- and

five-fold compared to the control animal. Quantitation was

done using actin mRNA levels to normalize lipo I data. To

determine the approximate time required for induction of

heart lipo I mRNA by glucagon, a time course experiment was

undertaken (Fig. 4-6). The results showed that induction

required about 4 hr and could still be seen 24 hr after

injection.

Next, I tested other hormones to help define the

mechanism by which lipo I mRNA expression is elevated in the

heart (Fig. 4-7). Dexamethasone, a synthetic

glucocorticoid, and insulin, had no effect. Epinephrine, a












Glucagon





4


Glucagon





24


Heart
- +


Kidney
- +


Muscle
- +


lipo I





actin


Figure 4-4. Induction of lipocortin I mRNA in the rat heart
after injection of glucagon. Four or 24 hr after
intravenous tail vein injection of 1 mg/kg glucagon, total
RNA was isolated from various tissues and subject to
Northern analysis. Actin mRNA levels are shown as an
internal control.

















Glucagon


- + + + +


Figure 4-5. Increased levels of lipocortin I mRNA in the
heart after glucagon injection in four animals. Heart RNA
was isolated 4 hr after injection of 1 mg/kg glucagon or
buffer only and subject to Northern analysis. One of the
two actin mRNA species is also shown.


lipo I


ia ctin



















0 0.5 1


2 4 8 12 24


Figure 4-6. Time course of the increase in lipocortin I mRNA
in the heart after injection of glucagon. Rats were
injected intravenously with 1 mg/kg glucagon and heart RNA
was isolated after various times (hr). Northern analysis
also shows actin mRNA levels.


lipo I




actin



























S n i .lipo I




actin













Figure 4-7. Lipocortin I mRNA expression in the rat heart
following injection of glucagon and other hormones. Rats
were injected intravenously with saline, glucagon,
epinephrine, dibutyral cAMP, dexamethasone, or insulin and
heart RNA was isolated after 4 hr. Northern analysis shows
lipocortin I and actin mRNA levels.








70

hormone which acts through increased cAMP levels and effects

heart function (Regan et al., 1964), was also without

effect. In addition, injection of cAMP did not change heart

lipo I mRNA expression. The induction of heart lipo I mRNA

by glucagon and lack of effect of the other hormones and

cAMP was also observed in mice (Fig. 4-8).

Discussion

Studies in this chapter demonstrate that tissue-

specific regulation of lipo I mRNA levels occurs as a result

of diabetes, and that glucagon may play a part in this

regulation. These experiments are the first to report a

link between a member of the annexin family and diabetes,

and also to show alterations in expression of an annexin in

response to glucagon.

Of the STZ-diabetic tissues studied, the heart, kidney,

and hindleg muscle showed a significant increase in steady-

state lipo I mRNA levels. To gain possible insight into the

role of lipo I in these tissues, the known effects of

diabetes on them must be considered. In the case of the

heart, cardiovascular disease is a major cause of death in

diabetes, responsible for about 70% of diabetic mortality

(Christlieb, 1973). While atherosclerosis is an important

component, studies in the rat (Fein et al., 1980), rabbit

(Fein et al., 1985), dog (Regan et al., 1974), and monkey

(Haider et al., 1981) show evidence for primary

cardiomyopathy, or dysfunction in the heart itself. A



























lipo I





actin












Figure 4-8. Lipocortin I mRNA expression in the mouse heart
following injection of glucagon and other hormones. Mice
were injected intravenously with saline, glucagon,
epinephrine, dibutyral cAMP, dexamethason, or insulin and
heart RNA was isolated after 4 hr. Northern analysis shows
lipocortin I and actin mRNA levels.









72

variety of alterations in heart rate have been described in

diabetic rats (Pfaffman, 1980). In addition, structural

changes in diabetic rat myocardium have been documented

(Thompson, 1988). In that study, alterations in all three

primary components of the heart, the cardiocytes, connective

tissue matrix, and microvasculature, were observed, most of

which were reversed after insulin treatment. Interestingly,

the effects of diabetes on the heart appeared to be focal,

with groups of normal and abnormal cardiocytes interspersed

throughout the muscle tissue.

A multitude of biochemical alterations have been

studied in diabetic hearts. The normal flux of metabolites

through glycolysis and the Krebs cycle is disrupted by the

large amounts of ketone bodies which accumulate and are the

main source of energy for heart respiration in diabetes

(Taegtmeyer and Passmore, 1985). A decreased ability of the

sarcoplasmic reticulum to take up calcium, depressed myosin

and actomyosin ATPase activities, and lowered Na /K+ ATPase

were observed in diabetic rat hearts (Tahiliani and McNeill,

1986). The P-adrenergic receptor number in cardiac

membranes was lowered and their coupling to adenylate

cyclase altered in diabetic rat cardiac membranes (Atkins et

al., 1985).

Most interesting with regard to lipo I are the changes

in lipid metabolism documented in the diabetic heart. Fatty

acid compositions of phospholipids of the heart are altered,








73

including a decline in arachidonic acid levels (Faas and

Carter, 1980; Holman et al., 1983; Schrade et al., 1963).

Some metabolites of arachidonic acid, such as prostacyclin

and leukotrienes, are also affected (Johnson et al., 1984).

Since lipo I has been implicated as an inhibitor of

arachidonic acid metabolism through modulation of

phospholipase A2 (Hirata et al., 1980), it is possible that

the elevated levels of lipo I mRNA observed in the diabetic

rat heart may play a role in reduced arachidonic acid levels

and regulation of arachidonic acid-derived products.

Physical properties of cell membranes, including fluidity,

appear to be altered in the diabetic heart (Cheung et al.,

1980; Shinitzky, 1980). The phospholipid-binding activity

of lipo I, its localization to the inner surface of the

plasma membrane, and the implication of some of the annexins

in membrane-cytoskeletal linkage, offer the possibility that

elevated lipo I could be involved in membrane alterations

seen in the diabetic heart. Finally, elevated levels of

lipo I in the diabetic rat heart may play a role in reduced

phosphoinositol synthesis and release (Bergh et al., 1988).

If the function of lipo I is similar to that of lipo III, a

hydrolase which metabolizes cyclic phosphoinositol

intermediates (Ross et al., 1990), a connection between

elevated lipo I expression and altered phosphoinositol

turnover in the diabetic heart is possible.








74

The two most notable alterations in the diabetic kidney

are an increase in glomerular filtration rate and a marked

hypertrophy (Cortes et al., 1987; Seyer-Hanson, 1983). A

long list of biochemical changes have been observed in the

diabetic kidney, including both increases and decreases in

enzymes involved in glycoprotein metabolism, increased

thickness of glomerular basement membrane, decreased

guanylate cyclase activity, increased gluconeogensis,

increased glycogen content, increase in RNA synthesis, and

changes in pentose phosphate metabolites (Cortes et al.,

1983; Steer et al., 1985). Most of these alterations were

corrected after insulin treatment. As was noted in the

heart, a deficiency in arachidonic acid was also observed in

the diabetic kidney (Faas and Carter, 1980; Holman et al.,

1983). In addition, the glomerular inositol content and the

turnover of polyphosphoinositides was reduced by almost 60%

in streptozotocin-diabetic rats (Craven and DeRubertis,

1989), similar to what was observed in the diabetic heart.

It is intriguing to think that the observed induction in

heart and kidney lipo I mRNA I observed might be involved in

these changes in arachidonic acid or phosphoinositol

metabolism in the diabetic.

Diabetes has profound effects on the biochemical,

morphological, and contractile properties of skeletal muscle

as well as that of cardiac muscle. Amino acid transport is

reduced (Manchester, 1970), Na-K -ATPase activity is








75

depressed (Moore et al., 1983), and increased proteolysis

and a reduction of protein synthesis lead to loss of muscle

strength and mass (Smith et al., 1989). Insulin treatment

was found to reverse the increased proteolysis observed in

streptozotocin-diabetic rat skeletal muscle. To the best of

my knowledge, arachidonic acid and phosphoinositol levels

have not been studied in diabetic skeletal muscle. If

levels of these were found to decrease compared to control

animals, as is the case in heart and kidney, a role for lipo

I in their down-regulation would be strengthened.

In contrast to what was observed in the heart and

kidney, lipo I mRNA was not increased in the hindleg muscle

of STZ-injected animals removed from insulin treatment

relative to those which were on insulin (Fig. 4-3). In

light of the observation that hindleg muscle lipo I mRNA was

increased after STZ-injection (Fig. 4-1 and 4-2) relative to

normal animals, several possibilities exist. One is that

the increase in lipo I mRNA in hindleg muscle after STZ-

injection was a direct, transient effect of STZ on that

muscle, rather that a consequence of diabetes. Or, perhaps

the increase in hindleg muscle lipo I mRNA only occurs

transiently during the first few days of acute diabetes, and

cannot recur during a second incidence of severe diabetes

following insulin treatment.

Having established that lipo I mRNA levels increase in

the diabetic heart, kidney, and skeletal muscle, I next








76

investigated several hormones which might mediate this

effect. Of the hormones tested, only glucagon, a hormone

which is highly elevated during diabetes (Unger et al.,

1963), consistently altered lipo I mRNA expression. Known

target tissues for glucagon action include the liver,

adipose tissue, heart, kidney, intestine, and skeletal

muscle (Lefebvre and Luyckx, 1979). The effect I observed,

an induction of three- to five-fold was observed in the

heart, but not in the kidney or skeletal muscle, starting 4

hr after injection. No changes were observed in kidney or

skeletal muscle up to 24 hr after glucagon administration

(Fig. 4-4).

These results suggest that elevated glucagon levels may

mediate a relatively rapid induction of lipo I mRNA in the

diabetic heart, while induction in the diabetic kidney and

skeletal muscle may require longer exposure to high glucagon

levels, or not involve glucagon. In support of a more long

term effect of glucagon in the kidney, glucagon infusion for

seven days promoted renal growth in the rat, a phenomenon

known to occur during diabetes (Logan and Lee, 1988).

The positive chronotropic and inotropic actions of

glucagon on the heart (increase in heart rate and

contractility) have been well documented in many species,

including dog, rat, and guinea pig (Farah and Tuttle, 1960).

Alterations in intracellular calcium levels by glucagon are

important in these effects on the rat heart (Chernow et al.,








77

1987), but the role of cAMP is unclear (Busuttil et al.,

1976; LaRaia et al., 1968; Rodgers et al., 1981). In the

adult mouse heart, evidence was shown that cAMP was not

altered during the chronotropic and inotropic actions of

glucagon (Clark et al., 1976).

The induction of lipo I mRNA in the rat heart by

glucagon may not involve changes in cAMP levels. Increased

lipo I mRNA levels were not observed in either rats or mice

after injection of epinephrine, a hormone which elicits

heart effects similar to those of glucagon and acts via cAMP

(Regan et al., 1964), or after injection of dibutyral cAMP

(Fig. 5 and 6). I believe that induced levels of lipo I

mRNA in the heart may be due to glucagon-mediated, cAMP-

independent changes in calcium levels (Petersen and Bear,

1986; Wakelam et al., 1986).

In summary, lipo I mRNA levels are induced in the

heart, kidney, and hindleg muscle of streptozotocin-diabetic

rats. Glucagon injection leads to increased lipo I mRNA

levels in the heart, but not in the kidney or hindleg

muscle, as early as 4 hr following administration. These

results suggest that glucagon may play a primary role in

heart lipo I mRNA induction during diabetes, and that

induction in the diabetic kidney and skeletal muscle

involves factors other than, or is independent of, glucagon.

Induction of lipo I mRNA during diabetes may involve

alterations in arachidonic acid or phosphoinositol








78

metabolism (Bergh et al., 1988; Craven and DeRubertis, 1989;

Faas and Carter, 1980; Gubjarnason et al., 1987; Holman et

al., 1983; Johnson et al., 1984; Pieper, 1990; Schrade et

al., 1963; Winegrad, 1987). Perhaps these changes occur

rapidly in the heart in response to elevated glucagon, and

more gradually in kidney and skeletal muscle as the disease

progresses and tissue function deteriorates.














CHAPTER 5

CONCLUSIONS AND FURTHER DIRECTIONS

Studies in this dissertation provide the basis for

future experiments to investigate the mechanisms governing

regulation of the lipo I gene. In Chapter 2, the mouse lipo

I gene structure was determined. An interesting finding was

that three of the four 17-amino acid consensus sequences

believed to be important in lipo I phospholipid/calcium

binding activity were encoded by the last codon of one exon

and the following 16 codons of the next exon. This is

highly suggestive of an exon duplication event early in the

evolution of lipo I. There is a striking similarity in the

gene structures of lipo I and lipo II (Amiguet et al.,

1990), indicating that these two annexins are derived from a

common ancestor. It will be very interesting to compare the

gene structures of lipo I and II with those of the other

annexins, including the 67 kDa species, when these are

determined. Such analysis may provide insight regarding the

relationship between the annexins and their evolution.

The lipo I promoter region was shown to contain several

potential regulatory sites which matched or resembled known

consensus sequences. There is indirect evidence that some

of these sites may be functional. Lipocortin I expression








80

in the mouse mammary epithelial cell line HC11 is effected

by serum (K. Horlick, unpublished results), and two

potential serum response elements were identified in the

lipo I promoter. The sites identified which are most likely

to be involved in lipo I regulation are the putative

glucocorticoid response elements, particularly the delayed

glucocorticoid response element. Lipo I mRNA has been shown

to be induced by dexamethasone in macrophages (Wallner et

al., 1986) and fibroblasts (Phillips et al., 1989; Wong et

al., submitted), although this does not occur in all cell

types examined (Bienkowski et al., 1989; Bronnegard et al.,

1988; Hullin et al., 1989; Isacke et al., 1989; Northup et

al., 1988; William et al., 1988). Studies outlined later in

this chapter are designed to determine whether the sites

identified in the lipo I promoter are indeed involved in lipo

I mRNA regulation, and to possibly identify others.

Chapters 3 and 4 involve tissue-specific regulation of

lipo I mRNA in two very different rat systems: The

differentiating mammary gland, and streptozotocin-induced

diabetes. Studies in Chapter 3 demonstrate that lipo I

exhibits a striking regulation as the mammary gland

undergoes normal differentiation and involution, and that

the suckling stimulus plays a role in this regulation.

While there are several different cell types in mammary

tissue, it is likely that epithelial cells are the dominant

factor in the lipo I regulation I observed.








81

Immunolocalization studies using mammary tissue sections

from various stages of development, similar to those carried

out for lipocortin II and calelectrin in the mouse mammary

gland (Lozano et al., 1989a), could confirm this hypothesis.

The fact that large changes in lipo I expression coincide

with periods when epithelial cells change growth and

differentiative status suggests the two may be related. In

light of this, I have proposed that high levels of lipo I

may be associated with the dormant epithelial cells of the

mature virgin and involuted glands. A hormone or growth

factor-mediated event during pregnancy, which leads to

decreased levels of lipo I, may be necessary for dormant

epithelial cells to proliferate and differentiate into

secretary cells.

One of the most interesting findings in these studies

is that some type of translational control of lipo I

apparently occurs in mammary tissue. After premature pup

withdrawal, lipo I mRNA increased rapidly and reached peak

levels around 12-24 hr following removal. In contrast,

little or no protein was detected until some time after 24

hr following pup withdrawal, and peak levels were observed

by five days following withdrawal. This indicates that

although a large pool of lipo I mRNA accumulates soon after

cessation of suckling, translation is significantly delayed.

I believe that a rapid signal sent to the mammary gland

after pup withdrawal, mediated by the sympathetic nervous








82

system, triggers a rapid increase in lipo I mRNA expression

in secretary epithelial cells destined to revert to the

resting state during involution. In situ hybridization to

localize lipo I mRNA could confirm this theory. The delay

in translation of lipo I mRNA may reflect the time necessary

for the gland, and epithelial cells in particular, to commit

to involution. Lipocortin I mRNA may accumulate in certain

epithelial cells immediately after cessation of suckling in

preparation for reversion to the resting state, but

translation could be delayed for several days in case

suckling must resume. At a time believed to be about five

days following pup removal (Silver, 1956), lactation is no

longer possible and epithelial cells are committed to death

or reversion to dormancy (Hollmann, 1974). The expression

pattern of mammary lipo I during both natural development

and after premature cessation of suckling is consistent with

an important role for lipo I in the resting epithelial

cells. While this role cannot be currently determined, I

have suggested lipo I may possibly be necessary to prevent

phosphoinositol turnover in these cells, keeping them from

proliferating until the appropriate time during pregnancy.

The in vivo mammary gland studies, particularly those

involving removal of the suckling stimulus, indicate that

lipo I mRNA is subject to regulation at the levels of both

mRNA expression and translation. The former type of

regulation could be either due to transcriptional control,








83

mRNA stability changes, or a combination of both. Due to

the complexity of the mammary gland and its multiple cell

types, an in vitro system will likely be necessary to

further explore mammary lipo I regulation. While indirect

evidence suggests a mammary epithelial cell line would be

the best to investigate, immunolocalization and in situ

hybridization studies would be useful in confirming this. I

have carried out preliminary studies on lipo I mRNA

expression in the mouse mammary epithelial line HC11, which

has been shown to be responsive to both prolactin and

dexamethasone (Ball et al., 1988; Doppler et al., 1989).

These cells, derived from midpregnant mouse mammary tissue,

express high levels of lipo I mRNA (data not shown).

Therefore, it could be a useful system in determining

factors that down-regulate lipo I during pregnancy.

Accordingly, I investigated the effects of several hormones

on expression of lipo I mRNA in HCll cells. Neither

prolactin, estrogen, progesterone, epidermal growth factor,

nor insulin had an effect (data not shown), but two

noteworthy observations were made. Dexamethasone caused a

reproducible increase in lipo I mRNA levels, and serum had

an inhibitory influence on mRNA expression. More thorough

studies using HC11, or another of the available mammary

epithelial cell lines, may be useful in further defining the

hormonal signals which regulate lipo I in the mammary gland.

Experiments utilizing mammary explants in culture may also








84

prove useful (Matusik and Rosen, 1978), but would still have

the problem of multiple cell types.

In Chapter 4, it was shown that lipo I mRNA expression

was increased in certain tissues of three-day

streptozotocin-diabetic rats, and that glucagon may play a

role in lipo I mRNA induction in the diabetic heart.

Diabetes is a complex disease resulting in many hormonal

alterations, so again an in vitro system will be required to

further study the mechanism of lipo I regulation. An

important question to address is what particular hormones,

or aspects of altered cellular metabolism, are factors in

regulating lipo I expression.

Experiments should focus on cell lines derived from the

heart, kidney, and skeletal muscle, because these tissues

showed altered lipo I expression in diabetic animals. Heart

cells would be most interesting due to the ability of

glucagon to alter lipo I mRNA levels in the heart. As

mentioned regarding the mammary gland studies, in situ

hybridization would be helpful in indicating which

particular cell types in tissues express lipo I mRNA, and

therefore which cell lines might be the best candidates for

study.

I carried out preliminary studies to attempt to

identify a glucagon- or dexamethasone-responsive cell line

of kidney or muscle origin. Lipocortin I mRNA levels were

not affected by these hormones in rat mesangial cells, which








85

are smooth muscle-like cells from the kidney, or in the G8

mouse skeletal muscle line. Chick heart atrial cells were

also examined, but insufficient cross-hybridization with our

mouse cDNA probe was a problem. If heart myocytes were

shown by in situ hybridization to express lipo I mRNA,

further investigation of these cells might prove worthwhile.

It is possible to isolate and maintain beating heart

myocytes from adult mice or rats in culture (Jacobson and

Piper, 1986), so potential hormonal effects on lipo I

expression in these cells can be evaluated.

Providing that future studies based on the information

in Chapters 3 and 4 will define a cell line which exhibits

lipo I mRNA regulation in response to a hormonal stimulus or

change in growth status, experiments to probe the mechanism

of this regulation can be undertaken. Multiple cell types

would make such studies in whole tissues unlikely to succeed

and difficult to interpret. Initially, the mechansim by

which a change in mRNA levels occurs would be investigated.

As stated before, this could result from transcriptional

alterations, or changes in mRNA stability or some other

post-transcriptional means. Experiments incorporating the

transcriptional inhibitor actinomycin D, and the protein

synthesis inhibitor cycloheximide, would help answer such

questions. The nuclear run-on assay, which measures the

amount of nascent transcripts in isolated nuclei, could be

used to investigate transcriptional changes (McKnight and

Palmiter, 1979).








86

After obtaining some idea of the type(s) of mechanisms

involved in regulation of lipo I gene expression, particular

sites in the gene and its promoter mediating this regulation

can be investigated. Two types of experiments, deletion

analysis and chromatin structure analysis, may be undertaken

to identify potential regulatory regions. The first

involves transfection of promoter deletion mutants into cell

lines and assaying for transcription levels of a reporter

gene. This can be done either by selecting for stable

transfectants (Buetti and Diggelmann, 1983) or through

analysis of the entire population of transfected cells,

referred to as transient expression (Dean et al., 1983).

Chromatin structure analysis is based on the observation

that pancreatic deoxyribonuclease, or DNase I,

preferentially digests genes that are or have been

transcriptionally active (Weintraub and Groudine, 1976).

Most of the cleavage sites for DNase I, referred to as DNase

I hypersensitive sites, have been localized to the 5'

flanking region of genes (Elgin, 1981). In many cases,

DNase I hypersensitive sites were shown to occur close to

important regulatory elements, such as enhancers (Zaret and

Yamamoto, 1984). A combination of deletion and chromatin

structure analysis should be useful in providing a low

resolution view of important regulatory regions within the

lipo I gene.








87

If DNase I hypersensitive sites are found using

conventional methods, they may be more precisely located by

a recently developed, high resolution technique. This

involves electrophoresing genomic DNA on native

polyacrylamide gels prior to electroblotting and

hybridization, and has been utilized to map DNAse I

hypersensitive sites in the human histone H4 gene (Pauli et

al., 1988).

Information provided by the above techniques may allow

studies of lipo I gene regulation in vivo at nucleotide

resolution. Intact cells are treated with a chemical probe

for DNA structure, dimethylsulfate, and DNA-protein

interactions are visualized on DNA sequencing gels as G-

residues with altered reactivity to DMS (Church and Gilbert,

1984). This in vivo analysis has the advantage of avoiding

artifacts which might be obtained in in vitro studies such

as gel retardation assays (Fried and Crothers, 1981), but is

limited to revealing interactions at only G-residues.

In conclusion, studies described here make an important

contribution to the annexin field in several ways. The

results in Chapter 2 concerning lipo I, along with what is

known about the lipocortin II gene, strongly suggest a

common ancestor for these proteins and will be important in

elucidating the evolution of the annexins when the remaining

genes are characterized. In addition, sequence information

from the promoter now enables regulatory studies to be








88

designed. Experiments shown in Chapter 3 represent the most

dramatic examples of lipo I mRNA and protein regulation

currently known, and will hopefully lead to a tissue culture

system in which this regulation can be further defined and

investigated. Also, the potential link between lipo I and

mammary cell growth status deserves further investigation,

and may indicate that lipo I, like lipo III, is involved in

an important biochemical pathway such as inositol

metabolism. Finally, studies in Chapter 4 show for the

first time a link between glucagon and induced expression of

an annexin, and offer the possibility that lipo I is

involved in the altered lipid metabolism known to occur

during diabetes.














REFERENCES


Aarsman, A. J., Mynbeek, G., Van Den Bosch, H., Rothhut, B.,
Prieur, B., Comera, C., Jordan, L., and Russo-Marie, F.
(1987). Lipocortin inhibition of extracellular and
intracellular phospholipases A2 is substrate
concentration dependent. FEBS Lett. 219: 176-180.

Addison, W. R., and Kurtz, D. T. (1986). Nucleotide
sequences required for the regulation of a rat a2^
globulin gene by glucocorticoids. Mol, Cell. Biol. 6:
2334-2346.

Amiguet, P., D'Eustachio, P., Kristensen, T., Wetsel, R. A.,
Saris, C. J. M., Hunter, T., Chaplin, D. D., and Tack,
B. F. (1990). Structure and chromosome assignment of
the murine p36 (calpactin I heavy chain) gene.
Biochemistry 29: 1226-1232.

Ando, Y., Imamura, S., Hong, Y. -M., Owada, M. K., Kakunaga,
T., and Kannagi, R. (1989). Enhancement of calcium
sensitivity of lipocortin I in phospholipid binding
induced by limited proteolysis and phosphorylation at
the amino terminus as analyzed by phospholipid affinity
column chromatography. J. Biol. Chem. 264: 6948-6955.

Atkins, F. L., Dowell, R. T., and Love, S. (1985). 3-
adrenergic receptors, adenylate cyclase activity, and
cardiac dysfunction in the diabetic rat. J. Cardiovasc.
Pharmacol. 7: 66-70.

Ball, R. K., Friis, R. R., Schoenenberger, C. A., Doppoler,
W., and Groner, B. (1988). Prolactin regulation of P-
casein gene expression and of a cytosolic 120-kd
protein in a cloned mouse mammary epithelial cell line.
EMBO J. 7: 2089-2095.

Bergh, C.-H., Hjalmarson, A., Sjogren, K.-G., and Jacobsson,
B. (1988). The effect of diabetes on
phosphatidylinositol turnover and calcium influx in
myocardium. Horm. Metabol. Res. 20: 381-386.

Bhimji, S., Godin, D. V., and McNeill, J. H. (1985).
Biochemical and functional changes in hearts from
rabbits with diabetes. Diabetologia 28: 452-457.










Bielefeld, D. R., Pace, C. S., and Boshell, B. R. (1983).
Altered sensitivity of chronic diabetic rat heart to
calcium. Am. J. Physiol. 245: E560-E567.

Bielschowski, M., and Bielschowski, F. (1956). The New
Zealand strain of obese mice. Australian J. Exp. Biol.
Med. 34: 181-198.

Bienkowski, M.J., Petro, M.A., and Robinson, L.J. (1989).
Inhibition of thromboxane A synthesis in U937 cells by
glucocorticoids. J. Biol. Chem. 264: 6536-6544.

Blackwell, G.J., Carnuccio, R., Di Rosa, M., Flower, R.J.,
Parente, L., and Persico, P. (1980). Macrocortin: a
polypeptide causing the anti-phospholipase effect of
glucocorticoids. Nature 287: 147-149.

Blake, C. (1983). Exons-present from the beginning? Nature
306: 535-537.

Borellini, F., and Oka, T. (1989). Growth control and
differentiation in mammary epithelial cells. Env.
Health Perspec. 80: 85-99.

Braslau, D. L., Ringo, D. L., and Rocha, V. (1984).
Synthesis of novel calcium-dependent proteins
associated with mammary epithelial cell migration and
differentiation. Exp. Cell Res. 155: 213-221.

Breathnach, R., and Chambon, P. (1981). Organization and
expression of eucaryotic split genes coding for
proteins. Annu. Rev. Biochem. 50: 349-383.

Bronnegard, M., Andersson, O., Edwall, D., Lund, J.,
Norstedt, G., and Caristedt-Duke, J. (1988). Human
calpactin II (lipocortin I) messenger ribonucleic acid
is not induced by glucocorticoids. Mol. Endocrinol. 2:
732-739.

Buetti, E., and Diggelmann, H. (1983). Glucocorticoid
regulation of mouse mammary tumor virus: identification
of a short essential DNA region. EMBO J. 2: 1423-1429.

Burgonyne, R. D. (1988). Calpactin in exocytosis? Nature
331: 20.

Busuttil, R. W., Paddock, R. J., Fisher, J. W., and George,
W. J. (1976). Changes in cyclic nucleotide levels and
contractile force in the isolated hypoxic rat heart
during perfusion with glucagon. Circ. Res. 38: 162-
167.










Chase, H. P., Williams, R. L., and Dupont, J. (1979).
Increased prostaglandin synthesis in childhood diabetes
mellitus. J. Pediatr. 94: 185-189.

Chen, C.-L. C., and Feigelson. P. (1980). Hormonal control
of a2u globulin synthesis and its mRNA in isolated
hepatocytes. Ann. NY Acad Sci. 349: 28-45.

Chernow, B., Zaloga, G. P., Malcolm, D., Willey, S. C.,
Clapper, M., and Holaday, J. W. (1987). Glucagon's
chronotropic action is calcium dependent. J. Pharmacol.
Exp. Ther. 241: 833-837.

Cheung, H. C., Almira, E. C., Kansal, P. C., and Reddy, W.
J. (1980). A membrane abnormality in lymphocytes from
diabetic subjects. Endocr. Res. Commun. 7: 145-156.

Chomczynski, P., and Sacchi, N. (1987). Single-step method
of RNA isolation by acid guanidinium thiocyanate-
phenol-chloroform extraction. Anal. Biochem. 162: 156-
159.

Christlieb, A. R. (1973). Diabetes and hypertensive vascular
disease: mechanisms and treatment. Am. J. Cardiol. 28:
27-41.

Chuah, S. Y. and Pallen, C. J. (1989). Calcium-dependent and
phosphorylation-stimulated proteolysis of lipocortin I
by an endogenous A431 cell membrane protease. J. Biol.
Chem. 264: 21160-21166.

Church, G.M., and Gilbert, W. (1984). Genomic sequencing.
Proc. Natl. Acad. Sci. USA 81: 1991-1995.

Clark, C. M., JR., Waller, D., Kohalmi, D., Gardner, R.,
Clark, J., Levey, G. S., Wildenthal, K., and Allen, D.
(1976). Evidence that cyclic AMP is not involved in the
chronotropic action of glucagon in the adult mouse
heart. Endocrin. 99: 23-29.

Cooper, J. A., and Hunter, T. (1983). Identification and
characterization of cellular targets for tyrosine
protein kinases. J. Biol. Chem. 258: 1108-1115.

Corden, J., Wasylyk, B., Buchwalder, A., Sassone-Corsi, P.,
Kedinger, C., and Chambon, P. (1980). Promoter
sequences of eukaryotic protein-coding genes. Science
209: 1405-1414.










Cortes, P., Dulmer, F., Goldman, J., and Levin, N. W.
(1987). Relationship between renal function and
metabolic alterations in early streptozotocin-induced
diabetes in rats. Diabetes 36: 80-87.

Cortes, P., Dulmer, F., Venkatachalam, K., and Levin, N. W.
(1983). Effects of diabetes mellitus on renal
metabolism. Mineral Electrolyte Metab. 9: 306-316.

Craven, P. A., and DeRubertis, F. R. (1989). Protein kinase
C is activated in glomeruli from streptozotocin
diabetic rats. J. Clin. Invest. 83: 1667-1675.

Creutz, C. E., Zakas, W. J., Hamman, H. C., Crane, S.,
Martin, W. H., Gould, K. L., Oddie, K. M., and Parsons,
S. J. (1987). Identification of chromaffin granule-
binding proteins. J. Biol. Chem. 262: 1860-1868.

Crompton, M. R., Moss, S. E., and Crumpton, M. J. (1988).
Diversity in the calpactin/lipocortin family. Cell 55:
1-3.

Crumpton, M. J., and Dedman, J. R. (1990). Protein
terminology tangle. Nature 345: 12.

Davidson, F. F., Dennis, E. A., Powell, M., and Glenney, J.
(1987). Inhibition of phospholipase A2 by "lipocortins"
and calpactins: an effect of binding to substrate
phospholipids. J. Biol. Chem. 262: 1698-1705.

De, B. K., Misono, K. S., Lukas, T. J., Mroczkowski, B., and
Cohen, S. (1986). A calcium-dependent 35-kilodalton
substrate for epidermal growth factor receptor/kinase
isolated from normal tissue. J. Biol. Chem. 261: 13784-
13792.

Dean, D. C., Knoll, B. J., Riser, M. E., and O'Malley, B. W.
(1983). A 5'-flanking sequence essential for
progesterone regulation of an ovalbumin fusion gene.
Nature 305: 551-554.

Doppoler, W., Groner, B., and Ball, R. K. (1989). Prolactin
and glucocorticoid hormones synergistically induce
expression of transfected rat 3-casein gene promoter
constructs in a mammary epithelial cell line. Proc.
Natl. Acad. Sci. USA 86: 104-108.

Drust, D., and Creutz, C. (1988). Aggregation of chromaffin
granules by calpactin at micromolar levels of calcium.
Nature 331: 88-91.