4
’.‘
1'
a
.

 

V
y.
.4
‘EJ

V-zu,

 

 

 

 

 

   

   
  

 

 

  
  
   
 
  
   
     
 

‘WV‘1T‘I-‘vw

   
 

J
.1‘; 2-” ,’
'rl‘r-Al;1-u>rll
“)Plu mu
. “W
:1 1.»

 

 

       

, A
“'71:,

V
m

 
 
 

:2? ‘
1;»:

  

1‘;

      

2—“.
an,
A 9,,

 

    
   

 

  

hw-

  

  

 
 

     

     

-.?'
. vhf: .‘
4’, p... _
" " {get if
. "‘12 5
I

<1,
1..
V“

Y

.
. x
g,

    
 

  

     

,‘t
r,
,'

Y
I.
1‘“

      
    

*

  

,5

1,3“
l ; ’-
"'3 :3”? {if 94:?!

       

  
  

 

' n;

   
 
 
 
  
    
   

 
  

hr, 914;; "wa

any, a: 1

 

     

¥
‘

    
  

‘"
33
«£13;

   

~35;
WV},

   
  

   
     
   
      

I"?
“3'21"; Lfig

52%:

 
 

wfhw:fl .
life"

'31 y
if“; 1 r 1%?”
qfifiéfl let-d.

{p32 9‘;
”in“ 7
Qty}? ”Warmly "I

    
 

3T4:

9“?“
a?

 

‘S
«r. .
xix",

  
   
       

   
      
   
    
 
   
  
     

’LBSTIJL“ .13} #111; '1 1‘

       
      
    

    

. 1,
345-
wt“

 

   
   
  

 

         
   
 
  
 
 

 
         
 

  

 
   

    

 

    
    

   
   

  
   
    

  
 
  

   
  

     
  
   

      
   
    

    
    

 

 
    
  

1;"; '7 . "o "'5’”ij 1‘ -"_‘P[r"A_r-n
.‘ . 1: .. n , "f; :3:
,9'14'1'“ *1’ Mr {33:2 ' mail" '1" ‘3‘" , ”'5",-
«11-‘13‘ 14,4; M #1 ,1y1n;15¥§;;11,:15m2;:1§.1~,1 .1. ma “11’
. r - '5" ' 1r: .
‘ ?;”U,M€W' fr??? ‘3' 71‘5523“? ’1??”‘:,:,,r,g,
' “ W3)“ mire"! 3.5%; "171; g— {If E‘Efig
'.1 ”F ‘ I- w ' rm .q- ,
. 1:15,!» $1915 :”,-;5‘,"1'19-f._,a. .31 1am: 4271,51 f 1' 51““?22- ,
.,:. 1,1 2m 4.521;; 515,1 "3'1“: "fif‘krfi "" 3:553 ‘
, 'Z-zy'r ”j' [uprr'ifi‘m ' '6: 5C 5:. h ‘ If?!”
., ,w v 1 ,1.,.1::,1«1‘1W:§:§£ 157%.?an 191:3: $2.3,
i, 11' "6-, M 43 r’" 1 55:71:; ,3" If?" :é."‘:"“*;‘
'~’!};:'5".r ""1"" " 1231,11,, 39%;" 1: $11.12 w 1;»,
11-.” ‘1 ' 'I" ’l " "1;
1 . :32: :99

    
  

  

‘ ,thw
"saga—*1:

r;
rx'rrrfri

   

'
1“."

 

f.) '.r vr
film... ,.. "" 1.”
, ’lflfl' r’rw‘ 4'; "M‘JE‘E"

     
 
  
 

   
     

      
 

     

     

  
  
 
 
  
  

   

   
         
    

3" 0'53?" \ ‘74:" w; ,. - L "w
5!, 1514’; - @213” "I; Tav’b'si' ;:"'(”1'fr”'£","1f;’::m;r’ 71?;

1”” . mac: :,.1.1,1,;1,;-;,,,,;.~.:1, ,., ,1? M12, , ~., , '7'.

‘ 111K ‘ a? "613“"? 1' 4‘ I . g g , ”I, "If... - ,. H 1,1257“
4 1" C'l-flr'U 5m!» 3" “" 1 J7? u? 3.3:;«7 ',.' ,’"':(‘!.4r {£5112} 3,-
I?" M! (pl, I'A'XI'I'L’ ' ”r ’ f“ '1', ff“ I” '14,: fi’:;.“;::£f .1,
. . , Li‘ff
)I’r' [if/1’ L'ffl‘T-h'xiflh mfg.- in": fix}; frcf , "' 3,1,1 1', 1; 'Z.,’_:‘;f;‘-Ar;1;11
' ’ W411: '1913151 1:6 ”1,1355: #3231? 9“,,» X??- 111-1 ,3: 1, 2,131,111,», ,.
' ”.01 o;

I‘ ,- .
up 'JI';, _
,7" '1?“ 1151'1Ww 11:91,” ,3}? ’11,,1,,1:~:r:,~1s.¥‘;:
1" '1; 4;}! 3;, ”x .r....,¢,o,,mr..,,,.,,
, ngf ,119 111,01 1:13;: n :15"

     

(10951“! 3M

l“lllllllllmr

' ”BRARY

Michigan state
l’l‘iv1grsit"

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ll

 

 

This is to certify that the

dissertation entitled

RNA METABOLISM IN
BOVINE PETUITARY CELLS

presented by

Karen Heinlein Friderici

has been accepted towards fulfillment
of the requirements for

Ph . D . degree in _Bi9_Qh_emis_Lry

M

jor professor

Date Feb. 19, 1988

Mun-MAM -' " '7 m, ' ' ' ' 0-1277!

 

 

MSU

LIBRARIES
-—

\v

 

 

RETURNING MATERIALS:
Place in book drop to
remove this checkout from
your record. FINES will
be charged if book is
returned after the date
stamped below.

 

 

 

 

RNA METABOLISM IN
BOVINE PITUITARY CELLS

By

Karen Heinlein Friderici

A THESIS

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

DOCTOR OF PHILOSOPHY

Department of Biochemistry

1988

ABSTRACT

RNA METABOLISM IN
BOVINE PITUITARY CELLS

By

Karen Heinlein Friderici

A system for the culture of bovine pituitary cells was
developed. Substiutution of D-valine for L-valine in the growth
medium resulted in undiminished prolactin synthesis over a 30-day
culture period. In serum-free medium or in medium containing charcoal
treated serum addition of hydrocortisone stimulated growth hormone
synthesis 3-5 fold. In either medium prolactin synthesis was
diminished 50% by hydrocortisone addition. In serum-free medium
bromoergocriptine reduced prolactin secretion 80-90% and prolactin
synthesis by 40-50%. No stimulation of prolactin protein synthesis
was observed using either estrogen or thyroid releasing hormone.

The primary bovine pituitary cell cultures were used in studies
of the half-life of prolactin, growth hormone and tubulin mRNAs.
Under conditions in which total poly A+ RNA decayed 50%, prolactin,
growth hormone and tubulin showed no turnover. This result, coupled
with the slow response of protein synthesis levels to hormonal
effects, suggested a very long half-life for prolactin and growth
hormone mRNA. It is not known how prolactin and growth hormone mRNA
maintain their unusual stability. Studies to compare metabolism of
these stable mRNAs to less stable mRNAs in the same cells were
initiated. A transformed bovine pituitary cell line was characterized
and a cDNA library prepared from these cells was screened for clones

containing mRNAs of varying abundance.

The cloned DNA was characterized. Southern analysis showed that
most cDNA clones represented single copy genes. The mRNAs produced by
these genes varied in size from <700 bp to >2000 bp. Kinetic analysis
of the cloned DNAs representing more abundant mRNAs revealed a 45 min
to 8 hr range of half-lives.

One cDNA clone which hybridized to RNA with a half-life of 45
min was further analyzed. This cloned sequence contained a bovine
short interspersed repetitive element. Sequence analysis of this
element indicated 97% homology with the right monomer of the dimeric
bovine A element. Transcription of this bovine repetitive element was
measured in transformed pituitary cells and in normal kidney cells.
High molecular weight RNA containing this element was observed. No

evidence for small transcription products was found.

To Dan,
whose unwavering encouragement

and faith made this possible

ACKNOWLEDGEMENTS

I would like to express my sincere appreciation to Dr. Fritz
Rottman for all the years of guidence, advice, and encouragement. He was
primarily responsible for my development as a scientist.

I wish to acknowledge the advice and guidance of the people
who have served on my committee: Drs. Jerry Dodgson, Hsing-Jien Kung,
Ed Fritsch, Ron Patterson, Arnold Revsin, and John Wang.

I would like to thank Dr. Ed Fritsch for worthwhile experiences
in his lab and for the opportunity to participate in the Cold Spring
Harbor cloning course. Also fond thanks to Dr. Michele Fluck for
encouraging me to finish my experiments while working for her.

There are many, many friends and coworkers who provided much
needed moral support and scientific assistance through out the years.
Thanks to Sally Camper and Bob Lyons who were always there, helped me
lots and never gave up. Thanks to Nancy Sasavage for steady
encouragement. Thanks to the Giltner Hall bunch; David, Kristen and
Mark, we made the trip together. Thanks to Wynne who always had good
advice and who shared the transition from Biochemistry to Microbiology
along with Bates and Boyer.

I would especially like to thank my Mother and Father for
starting me on the road to higher education. They were the first to

encourage me and help me to open new doors. I will always be grateful.

TABLE OF CONTENTS

LIST OF TABLES ..................................... vii
LIST OF FIGURES ................................... viii
PART I. PRIMARY PITUITARY CELL CULTURE AND KINETIC ANALYSIS OF

PROLACTIN AND GROWTH HORMONE mRNA.

INTRODUCTION ................................. 2
LITERATURE REVIEW ............................ 3
MATERIALS AND METHODS ....................... 23
RESULTS ..................................... 36
DISCUSSION .................................. 75
REFERENCES .................................. 81

PART II. CHARACTERIZATION OF A BOVINE REPETITIVE ELEMENT.

LITERATURE REVIEW ........................... 87
INTRODUCTION ................................ 98
MATERIALS AND METHODS ....................... 99
RESULTS .................................... 102
DISCUSSION ................................. 118
REFERENCES ................................. 123

vi

LIST OF TABLES

TABLE 1. Hormone effects on PRL/GH transcription ............. 12
TABLE 2. Cloning Scheme ...................................... 32
TABLE 3. Characterization of Pituitary cDNA clones ........... 68

vii

PART I

Figure

Figure
Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

10.

ll.

12.

l3.

14.

15.

16.

LIST OF FIGURES

Primary bovine pituitary cell cultures maintained

in D-or L—valine ...................................... 38
Effect of D-Valine on DNA and Protein Synthesis ....... 40
Gel profile of newly synthesized protein .............. 43

Prolactin and growth hormone as a percent of
total protein synthesis ............................... 45

Effect of charcoal treatment of serum on GH and
PRL synthesis ......................................... 46

Prolactin protein synthesis in serum-free medium ...... 49

Time course of accumulation of newly synthesized
prolactin in the culture medium ....................... 51

Protein synthesis in serum-free Dulbeccos medium ...... 53

Protein synthesis in cell cultures from fetal
pituitaries ........................................... 54

PRL and GH synthesis in fetal pituitaries ............. 56

Evaluation of hybridization assay using filter-

bound plasmid ......................................... 59
Pulse-chase of primary pituitary monolayer cultures. .61
Specific mRNA binding to Nitrocellulose filters

during chase .......................................... 63
Northern analysis of RNA corresponding to cDNA

clones ................................................ 69
Southern blot of SAS DNA probed with cDNA clones ...... 70
Decay kinetics of cDNA clones ......................... 73

viii

PART II

Figure 1. Southern blot analysis ............................... 104
Figure 2. Map of pBl/l6 ........................................ 106
Figure 3. Homology of bovine repetitive element with DNA

from other species ................................... 109
Figure 4. Sequencing gel of bovine repetitive element .......... 111

Figure 5. Sequence of pB1/16 in region of repetitive element...ll3
Figure 6. Transcription of bovine repetitive sequences ......... 116

Figure 7. Organization of bovine and caprine repetitive
elements ............................................. 120

PART I

PRIMARY PITUITARY CELL CULTURE

AND KINETIC ANALYSIS OF PROLACTIN

AND GROWTH HORMONE mRNA

INTRODUCTION

My research interests are directed toward studying mRNA turnover
and how mRNA stability might contribute to control of gene expression.
A system in which a specific, high abundance mRNA could be studied,
similar to casein production in mammary tissue or ovalbumin production
in chickens, would be ideal. In this system, the levels of a
particular mRNA could be increased or decreased by appropriate
hormones and the relative contribution of transcription and turnover
to specific mRNA levels could be measured. The research effort in our
laboratory uses the bovine pituitary as a model system for studying
control of expression of prolactin and growth hormone. Initially, the
majority of my effort was directed toward establishing a bovine
primary pituitary cell culture system that maintained proper hormone
production and responded appropriately to other hormones.

After a functioning cell culture system was established,
methods and conditions to label and analyze mRNA turnover were
determined. The results of these studies combined with data from
other laboratories indicated that both PRL and GH mRNAs probably have
an unusually long half-life. How this extreme stability is maintained
is not known. As a first step toward understanding why PRL and GH
mRNA are so stable, a series of other bovine mRNAs were prepared for
comparison with PRL and CH mRNA. These mRNAs were characterized as to

size, abundance and turnover.

LITERATURE REVIEW

mRNA Stability

The proteins produced in a cell are determined primarily by
the composition of its mRNA population. The molecular mechanisms
responsible for controlling mRNA composition operate at a variety of
levels. Initial control must be at the level of gene transcription,
but this event is followed by a number of steps which are all
potential control points. Nuclear RNA processing (capping,
polyadenylation, methylation, splicing), transport to the cytoplasm,
and cytoplasmic stability all contribute to the eventual mRNA content.
The role of mRNA stability in regulation of gene expression has
recently received considerable attention. It has been shown that
stability of individual mRNAs is important in controlling the
expression of certain genes.

Early estimates of mRNA half-life suggested that all mRNA in
the cell turned over at equal, but relatively slow, rates (Greenberg,
1972, Singer and Penman, 1973). It was later found by methodologies
described below that not all the Poly A+ RNA in a cell decayed at
equal rates (Berger and Cooper, 1975, Puckett et al.,l975). Cloning
technology has allowed us to study the half-life of individual mRNA
species using a number of methods. Each has some drawbacks and in a
given situation some methods are not appropriate; all are relatively

tedious.

 

 

 

The most direct measurement of mRNA half-life is pulse-chase
analysis. In this technique, cells are exposed to a pulse of 3H-
uridine for several hours. The chase is usually performed by adding
fresh medium containing unlabeled uridine and cytidine. RNA is
prepared at time intervals during the chase and the amount of
radioactivity specifically hybridizing to an excess of a filter-bound
DNA probe is determined. The time required for the cpm to be reduced
to one half the initial cpm is the half-life. This method is easiest
to use with an mRNA that has a high rate of transcription and a
relatively long half-life. Since nucleotide chases are not efficient,
the specific activity of the UTP pools should be monitored and
factored into the calculations (Shapiro and Brock, 1985).

In an approach to equilibrium assay, cells are labeled
continously for a long period of time (Greenberg,l972). RNA is
prepared at time points throughout the labeling period and cpm
hybridizing to a specific probe are determined. An equation is then
applied which takes into account the initial accumulation rate, the
length of time required to reach equilibrium and the doubling time of
the cells. For this method to be valid, 3H-uridine incorporation into
total cellular RNA must be linear over the entire labeling period.
This method is less direct than the previous one and assumes constant
rates of synthesis and decay.

In a third method transcription is inhibited, usually by a
drug such as actinomycin D. The reduction in mass of the mRNA may

then be followed by northern or S anaylsis using labeled

1

complementary DNA probes. This method works well for species of low

abundance or with low transcription rates since probes can be labeled

to very high specific activity. It has all the disadvantages
associated with drug treatments which is particularly a problem for
mRNAs that have long half-lives. There are some systems where
withdrawal of a hormone or changes in cell physiology result in
complete shut-off of transcription. In these cases northern analysis
of decay can also be used without the necessity of labeling cells
(Shapiro and Brock, 1985).

Until recently, studies on the half-life of specific mRNAs
have focused on viral genes or cellular genes that are tissue-specific
and produce a highly abundant mRNA. Vitellogenin, ovalbumin and
conalbumin are egg proteins produced at high levels in the liver or
oviduct. Their expression is positively controlled by the steroid
estrogen. In all cases, estrogen increases the rate of transcription
and also produces large increases in mRNA stability (Brock and
Shapiro, 1983, McKnight and Palmiter, 1979). Vitellogenin mRNA in
Xenopus liver has a half-life of 3 wk when induced and decays with a
17 hr half-life when estrogen is withdrawn. It has been postulated
that the increased stability may be mediated by binding of estrogen to
a cytoplasmic estrogen receptor (Shapiro and Brock, 1985).

In the mammary gland, casein production is stimulated to high
levels by prolactin, a polypeptide hormone. Similar to estrogen,
prolactin exerts its effect by increasing transcription and
concomitantly stabilizing casein mRNA (Guyette et al., 1979). How
prolactin increases stability of casein mRNA is not known.

Another gene product which needs to acheive high levels in
committed cells is globin. When murine erythroleukemia cells are

induced to differentiate, globin remains very stable with a half-life

 

of 60 hr (Vollach and Housman, 1981) while other mRNAs become less
stable (Darnell, 1982). It has also been found that not all of the
globins have the same stability. Human delta-globin which is only 1 -
1.5 % of total cell globin, has a half-life of 4.5 hrs as compared to
16.5 hrs for human beta-globin (Ross and Pizarro, 1983).

These are examples of genes whose products are controlled, to
a large extent, by posttranscriptional mechanisms. Whether these
genes were exceptional was not known.

Several attempts have been made to study larger numbers of
genes expressed in the same cell. Darnell and coworkers (Harpold et
al., 1979) isolated nine clones from CHO cells and studied their
transcription rates and steady state levels. They found that
differences in transcription rates between individual clones did not
necessarily predict the steady state levels they would attain. The
conclusion, therefore, was that posttranscriptional events were in
part controlling the final mRNA concentration. Derman et al.(l981),
studying a set of liver-specific mouse mRNAs, suggested that control
of these sequences was exclusively at the transcriptional level.
Using mouse L-cells, Carneiro and Schibler (198A) isolated mRNAs
corresponding to housekeeping genes of either low or moderate
abundance. They concluded that, in contrast to the accumulation of
liver specific mRNAs, the accumulation of ubiquitous mRNAs is mainly
posttranscriptionally controlled. It appears, therefore, that
selective stabilization or destabilization may apply to many types of
mRNAs, not only those which reach high levels in specilized cells.

How can stabilization or destabilization of a specific mRNA be

accomplished? Within the last few years some information regarding

this question has been accumulated in a number of systems. So far it
seems that there are a number of signals a cell can use to modulate
mRNA levels. I will describe three reasonably well documented
systems, tubulin, histone, and growth regulated mRNAs, which indicate
the diversity of these control mechanisms.

The principal subunits of microtubules are a- and B-tubulin
which belong to a small evolutionarily conserved multigene family.
Each family contains approximately six genes which are expressed in
specific, complicated programs during differentiation (Cleveland.
1987). Microtubules are formed by the polymerizations of heterodimers
consisting of u- and B-tubulin. Tubulin protein synthesis is self
regulated by the apparent intracellular concentration of heterodimers,
so that an approximately 2-fold elevation in the concentration of
unpolymerized tubulin is accompanied by a specific 5-10 fold
repression of new tubulin synthesis. Reduction of unpolymerized
subunit levels stimulates new tubulin protein synthesis 3-4 fold.

Anaylsis of tubulin mRNA levels showed that the altered rates
of tubulin protein synthesis could be quantitatively accounted for by
corresponding changes in tubulin mRNA levels. Transcription run-off
experiments (Cleveland and Havercroft, 1983) and experiments with
enucleated cells (Pittenger and Cleveland, 1985) have demonstrated
that changes in tubulin mRNA transcription rates are not responsible
for regulation of mRNA levels. Autoregulation is instead a result of
cytoplasmic events which modulate tubulin mRNA stability. Control of
B-tubulin mRNA stability apparently resides in the first 16 codons of
the mRNA (Gay et al., 1987). Whether autoregulation is due to tubulin

subunits interacting directly with B-tubulin mRNA or with nascent

tubulin protein is not yet known.

A number of mRNAs whose level is controlled during the cell
cycle have also been studied. Messenger RNAs for the replication-
dependent histone proteins required throughout 8 phase are selectively
stabilized during DNA synthesis. These mRNAs do not have a 3' poly A
tail but are terminated instead by a highly conserved stem-loop
structure (Schumperli, 1986). When DNA synthesis ceases, the half-
life of these mRNAs decreases from 40 min to 8 min in Hela cells
(Heintz et al., 1983). Accelerated mRNA turnover does not occur if
protein synthesis is inhibited (Graves and Marzluff, 1984). It has
now been shown that the terminal 30 nucleotides containing the stem-
loop structure are all that is required for regulation of histone mRNA
stability (Pandey and Marzluff, 1987) provided that translation of the
mRNA terminates within 300 bp of this structure (Graves et al., 1987).
These results suggest that the degradation system may recognize the
free 3' end of histone mRNA only after translation has proceded to a
normal termination codon.

Another group of genes whose transcription and stability have
been examined are those which are activated almost immediately after
resting cells are stimulated to divide. Two oncogenes, c—fos and c-
myc, are activated in this way (Greenberg et al., 1986, Piechaczyk et
al., 1985) and ten clones identified by Lau and Nathans (1987) have
been added to this group. In all cases these genes are regulated both
transcriptionally and posttranscriptionally. When shut-off is
required not only does transcription cease but the mRNA is
destabilized to allow an extremely rapid response in gene expression.

Shaw and Kamen (1986) found that an evolutionarily conserved

 

.__.-__..-.._.__ _._. . _..’.__._

AT rich S8 nucleotide sequence located in the 3’ end of the
lymphokine, granulocyte-monocyte colony stimulating factor (GM-CSF),
could confer instability to globin mRNA. These authors propose that
mRNAs containing a repeated AUUU motif are generally involved in the
selective degradation of transiently expressed messengers. Control of
c-myc is not so simple, however, since sequences in exon 1 or the
first intron have been associated with mRNA instability (Piechaczyk et
al., 1985) and at least some of the AUUU motif can remain at the 3'
end in more stable varients of c-myc (Jones and Cole, 1987).

To date there is no clear understanding of how individual
mRNAs may be stabilized or destabilized. We are just beginning to
determine which sequences are important for mRNA stability or
instability, but the mechanisms involved are not known.
Endonucleolytic cleavage at a specific site, followed by
exonucleolytic degradation is the most frequently suggested method for
mRNA degradation. Results from studies using histone mRNA in a cell-
free system support this speculation (Ross and Kobs, 1986). It will
be of interest to see if polyadenylated mRNAs are also specifically
controlled in this system. It seems likely that there will be more

than one system for mRNA degradation.

Prolactin and Growth Hormone Expression

The system we have chosen for studying regulation of gene
expression is the bovine anterior pituitary. In particular, we are
interested in the aspects which control production of two major
products of the pituitary gland, prolactin and growth hormone. One

important tool for these studies was a cell culture system could be

10

used understand which factors control growth hormone and prolactin
production.

The pituitary gland lies at the base of the brain and is
connected to the hypothalamus by the hypophysial stalk.
Physiologically the pituitary is divisible into two distinct portions,
the anterior and posterior pituitary, separated by a zone called the
pars intermedia. Embryologically the anterior pituitary originates
from Rathke's pouch which is an embryonic invagination of the
epithelium of the throat. The posterior pituitary is neural in origin
and is an outgrowth of the hypothalamus.

Six hormones which play major roles in control of metabolic
functions are secreted by the anterior pituitary. There are at least
five types of secretory cells in this gland, each producing a
different hormone. These cells can be distinguished by immunostaining
techniques or on the basis of morphology using electron microscopy.
The cell types and the hormones they produce are: 1) Somatotropes--
growth hormone (GH), 2) Lactotropes--prolactin (PRL), 3)
Corticotropes--Corticotropin (ACTH), 4) Thyrotropes--thyroid
stimulating hormone (TSH), 5) Gonadotropes--luteinizing hormone (LH)
and folicule stimulating hormone (FSH).

Almost all secretion by the pituitary is controlled by either
hormonal or nervous signals from the hypothalamus. This control is
complex and involves a number of factors for each secreted hormone.
Some of these factors interact to influence the effect of other
hormones on the secretion from the cell. The primary mode of
regulation for most of the pituitary hormones is stimulation of

secretion. Control of prolactin appears relatively unique among the

11

hormones secreted by the mammalian pituitary in that the major
regulatory mechanism involves inhibition of PRL production (Meites and
Clemens, 1972). This inhibition is demonstrated by in vivo
experiments where removal of the pituitary from direct control of the
hypothalumus results in an increase in PRL secretion.

Lactotrophs are distinctive in that they contain a mixture of
small and very large secretory granules. PRL is synthesized on
membranes of the rough endoplasmic reticulum as a precursor which is
rapidly processed and transported into the golgi (reviewed by Maurer,
l982d, Dannies and Tam, 1982). Within one hour PRL is found in small
secretory granules and at later times in the large granules. These
large granules appear only when the cells are not secreting at a rapid
rate. PRL is tightly packaged in the mature forms of the secretory
granules and once packaged is very stable. Secretion occurs by fusion
of the granule membrane with the plasma membrane and release of the
contents (exocytosis). PRL in granules can be degraded within the
cells by digestion in lysosomes (Farquhar, 1977).

A large number of hormones or factors have been found to
influence secretion of PRL in vivo. For purposes of this review I
will focus only on those that have been shown to work directly on the
PRL secreting cell (Table 1). In most cases the effect on release is
immediate, followed by much slower changes in levels of protein
synthesis. Although it is possible that a particular hormone can
affect secretion without affecting synthesis of PRL, factors which
alter secretion have also been shown to change the synthesis of PRL
(Laverriere et al., 1983). Most of the studies on PRL and GH release

and synthesis have been performed using rat pituitary systems.

 

12

TABLE 1

Hormone effects on PRL/CH transcription

 

Prolactin Growth Hormone

Dopamine Decrease No effect
Estrogen Increase Increase(slight)
Epidermal Growth

Factor Increase
Thyrotropin

Releasing hormone Increase Decrease
Triiodothyronine(T3) Decrease Increase
Vasoactive Intestinal

Peptide (VIP) Increase
Somatostatin Decrease Decrease
Glucocorticoids Decrease Increase
Growth Hormone

Releasing Factor No effect Increase

Ca++ Increase Increase(slight)

 

l3

Hemipituitary or primary cell cultures have been used but cell lines
derived from rat pituitary tumors (designated collectively as GH
cells) have been most extensively studied. The original cell lines:
6H3 and GHl’ can secrete both PRL and GH. Clonal varients can secrete
varying levels of one or the other of these hormones (Bancroft, 1981).
They resemble normal pituitary cells in that they have retained their
differentiated function since they respond to appropriate stimuli by
secreting and synthesizing PRL or CH. It should be remembered that
they differ from normal cells because they are transformed, do not

have storage granules, and sometimes require concentrations of factors

to affect release which are different from primary cells.

Regulation of ProlactinngNA Levels

Dopamine, secreted by the hypothalamus, is thought to be the
major hormone responsible for the inhibition of PRL production
(MacLeod et al., 1970). Receptors for dopamine and dopamine agonists
(primarily the catecholamines) are found on the plasma membrane of
pituitary cells (Caron et al., 1978, Cronin et al., 1978). Dopamine
has a limited stability in culture so most studies have been performed
using the more stable ergocryptine agonists of dopamine. When these
compounds interact with receptors on the lactotrophs they cause
inhibition of PRL release within minutes, after which PRL accumulates
within the cell for at least 8-16 hrs. Initially, bromocryptine
inhibits release without affecting synthesis and stability of the PRL
protein. Later PRL is degraded in the cell and PRL synthesis
decreases (Dannies and Tam, 1982).

In primary rat pituitary cultures the dopamine agonist,

14

ergocryptine, was shown to decrease PRL synthesis and to also decrease
mRNA levels (Maurer, 1980). Analysis of the time course of
ergocryptine effects showed that PRL synthesis decreased rather
slowly, so that 4-6 days were required for maximal inhibition. The
effect on PRL gene transcription was more immediate with 70-80%
inhibition of transcription within 45 min after addition of
ergocryptine to cultured cells (Maurer, 1981). The rapid decrease in
transcription but delayed effects on PRL mRNA levels and PRL protein
synthesis suggest that PRL mRNA is very stable with an estimated half
life of 48 hours.

The in vivo stimulation of PRL by estrogen has been recognized
for many years. Estrogen affects most aspects of prolactin
production, and the mechanism of its action has been extensively
investigated (Dannies, 1985). One result of short term estrogen
treatment is an increase in the number of PRL producing cells either
in the gland (Gersten and Baker, 1980) or in primary cultures
(Lieberman et al., 1982). In certain strains of rats long term
exposure to estrogen can result in pituitary tumors. All of the
increase in PRL production in estrogen treated cells is not simply a
result of an increased number of secreting cells. In primary cell
culture the increase in PRL synthesis is greater than the increase in
cell number (Lieberman et al., 1982). Further analysis reveals that
estrogen stimulates transcription from the PRL gene. Cells from 6
week old male rats treated with estrogen showed a 6-fold increase in
PRL mRNA transcription at 24 hr after addition of estradiol to culture
medium. The effect of estradiol on cells from female breeder rats was

less. There was no effect on growth hormone transcription (Shull and

15

Gorski, 1986).

With the large number of regulators affecting PRL production
their potential interaction can be very complicated. In the animal,
estradiol greatly decreases the ability of dopamine to inhibit PRL
secretion (Raymond et al., 1978). This antagonism extends to PRL
synthesis in cultured primary cells (Maurer, 1982b). The mechanism of
estrogen reduction of dopamine inhibition remains unknown but it has
been shown that estrogen is not acting at the level of the dopamine
receptor (DiPaolo et al., 1979). In contrast, estradiol has been
shown to increase the number of receptors of thyrotopin releasing
hormone, another hormone which stimulates PRL synthesis and release.

L-triiodothyronine (T3) inhibits PRL synthesis in rat primary
pituitary cell cultures. A 75% reduction in PRL synthesis with 6 days

of treatment by 10 nM T was paralleled by a decrease in PRL mRNA

3
content (Maurer, 1982a). This inhibition was not blocked by
estradiol, unlike the effect of estrogen on ergocryptine inhibition.
Contrasting effects were seen in CH3 cells where 50 nM T3 caused a two

fold increase in PRL production (Perrone et al., 1980) while 0.5 nM T3
reduced PRL synthesis (Stanley and Samuels, 1984).

The peptide hormones, thyrotropin releasing hormone (TRH) and
epidermal growth factor (EGF), have been shown to stimulate PRL
synthesis in rat GC cells (Dannies and Tashjian 1973, Murdoch et al.,
l982a). Transcription of PRL is stimulated within minutes after
addition of either hormone to the culture medium with maximal
stimulation occuring at 30-60 min (Murdoch et al., 1985). This

stimulation of PRL transcription is interesting in that it is

transitory. The initial burst of PRL gene transcription produces a

16

doubling of PRL mRNA synthesis within hours but then falls back to a
new level that is still higher than the uninduced mRNA synthesis.
Under this condition cytoplasmic PRL mRNA levels do not reach their
maximum level until 16-20 hours after addition of the hormone to the
cultures (Murdoch et al., 1985). Using these results the authors
estimate that the half life of PRL mRNA in CH cells is approximately 8
hr. Using pulse-chase analysis Laverriere et al. (1983) report that
TRH increases the half life of PRL mRNA from 17 hr in untreated cells
to 27 hr in treated cells.

Murdoch et al., (1985) have identified a basic nuclear protein
whose phosphorylation parallels the increases in PRL transcription.
TRH and EGF stimulate phosphorylation on 3-5 sites on this protein,
all on serine residues. It is possible that one mechanism by which
peptide hormones affect transmission of their signal from the plasma
membrane receptor to the 5' regulatory DNA sequence is through
phosphorylation of a specific nuclear protein.

A number of other factors have been found to regulate
prolactin synthesis. Vasoactive intestinal peptide (VIP) is found in
nerve termini in the hypothalamus and it has been shown to stimulate
PRL secretion in vivo and in vitro. In CH3 cells VIP increased PRL
mRNA and mRNA precursor levels 25 hours after treatment (Carrillo et
al., 1985). The increase in PRL mRNA levels was not blocked by
concurrent addition of the glucocorticoid, dexamethasone, to the
cultures. The glucocorticoids have been shown to reduce PRL mRNA
synthesis in the GH and GC cell lines. (Potter et a1, 1981, Evans et
al., 1982). However, in primary cultures of rat and ovine pituitary

cells no glucocorticoid effect on PRL synthesis was demonstrated

17

(Lieberman et al., 1978, Vician et al., 1979) leading Shull and Gorski
(1986) to suggest that the glucocorticoid effect could be unique to
pituitary tumor cell lines.

Most of the hormones which affect PRL synthesis and PRL mRNA
transcription are factors which bind to plasma membrane receptors.
Transduction of the signal generated by receptor hormone complexes at
the cell surface to yield a nuclear transcription response is still
not understood. Several cellular processes have been postulated to
mediate the intracellular effects of the hormones. These include
elevation of cytosolic calcium (Ca2+), a stimulation of CAMP levels
and activation of the phosphatidylinositol cycle (Bancroft et al.,
1985, Murdoch et al., 1985).

In serum-free medium, calcium alone has been shown to
specifically stimulate PRL mRNA levels (White and Bancroft, 1983). In
CH cells either TRH or EGF require extracellular calcium to achieve
maximal stimulation of PRL mRNA levels. When calcium is completely
removed from the medium by chelation with EGTA, no response to TRH and
a very small response to EGF is observed (Bancroft et al., 1985).

That these cells were still hormone-responsive in the absence of
calcium was shown by stimulation of CH mRNA by dexamethasone.
Qualitatively similar results have been found using primary rat
cultures, although these cells seem to regulate their intracellular
calcium in such a way that the calcium in the medium must be much
lower to demonstrate the response (Gick and Bancroft, 1985).

Adenosine 3',5’-monophosphate (CAMP) is often considered an
intracellular mediator of hormone action. CAMP may be involved in

regulation of PRL gene expression by certain hormones. Maurer (1982c)

 

18

found that CAMP could reverse the inhibition of egrocryptine on PRL
mRNA levels. In CH4 cells CAMP stimulated PRL transcription (Murdoch
et al, 1982b) as did forskolin, an activator of adenylate cyclase.
Forskolin also increased phosphorylation of the basic chromatin
associated protein that was also phosphorylated in the presence of TRH
(Murdoch et al., l982b). In other studies PRL synthesis was not
affected by CAMP. CH1 cells showed no effect on PRL synthesis in the
presence of forskolin even though it stimulated CAMP levels 400 fold
(Stanley and Samuels, 1984).

In summary, PRL protein syntheis and mRNA levels are
controlled by a wide variety of hormones (Table 1). In the animal the
primary regulation is inhibition of PRL secretion, but interactions of
all the hormones which affect PRL synthesis are undoubtedly important
in the physiological regulation of circulating PRL levels. The
nuclear effects of these hormones may be mediated by intracellular
levels of calcium and/or CAMP.

The half-life of PRL mRNA has been very difficult to
determine. In most of these studies it appears to a very stable mRNA.
In inhibition of PRL by ergocryptine, days are required to reach a new
steady state mRNA level. This implies that PRL mRNA is not rapidly
turned over in treated cells. During induction by estrogen or the
peptide hormones TRH or EGF, transcription rates change within
minutes, but new steady state mRNA levels are not acheived for hours
or days. It is possible that changes in PRL mRNA stability are
occurring during these treatments but probably are not large and are

superimposed on the kinetics of an mRNA which under all circumstances

is relatively stable.

19

Regulation of Growth Hormone Transcription

Most of the hormones which affect PRL synthesis also influence
CH synthesis, frequently in an inverse relationship (Table 1). CH
synthesis and secretion are stimulated primarily by glucorticoid and
thyroid (T3) hormones. The interrelation of these two hormones in
mediating expression of growth hormone is unclear. Studies using
various CH cell lines have shown that glucocorticoid and thyroid
hormone stimulate growth hormone synthesis (reviewed by Bancroft et
al., 1985). Reports by two groups (Martial et al., 1977, and Samuels
et al., 1979) suggested that glucocorticoids cannot induce significant
stimulation of GH synthesis in the absence of thyroid hormone. In

contrast, Dobner et a1. (1981) using CH cells reported significant

3

glucocorticoid stimulation of mRNA levels in the absence of T3. Evans

et a1. (1982) also showed that either dexamethasone (a synthetic

glucocorticoid) g; T could increase mRNA transcription and steady

3

state levels about 2-5 fold in C/C cells. These latter studies were
performed in a serum-substitute induction medium where no hormones

were added and 1% serum supplement was throughly treated to remove T3

and hydrocortisone. Bancroft et a1. (1985) found no synergism between

dexamethasone and T3 in CH3 cells in the induction of mRNA levels.

They consistently found a larger stimulation of CH synthesis by

dexamethasone alone than by T alone. Yaffee and Samuels (1984) and

3
Nyborg et a1. (1984) found that CC cells grown in 10% serum depleted
of thyroid and steroid hormones showed a low and variable response to

dexamethasone but always a synergistic response with T3 plus

dexamethasone. These authors feel that T3 can act as the primary

20

regulator of the CH gene and glucocorticoids further amplify the T3

response. Using CH3 or CO cells in serum-substituted medium Diamond
and Goodman (1985) found that dexamethasone alone produced no
stimulation in CC cells and a slight induction of transcription in CH3
cells. In their studies dexamethasone in combination with T3 induced
higher and faster transcription responses than T3 alone.
Glucocorticoids and T3 are probably the primary physiological
factors which control CH production. Several groups feel that
glucocorticoids potentiate the effects of T3 while others feel that
the two hormones function independently. The CH response to
glucocorticoids has been variable, ranging from no response at all to
production of high levels of CH. These differences may be the result
of varying cell lines used for the studies, growth conditions
(suspension vs. monolayer), or treatment of serum and medium
components. The polypeptides, somatostatin and growth hormone
releasing factor (CRF) have been shown respectively to inhibit and
stimulate CH secretion (Vale et al., 1977, Rivier et al., 1982). CRF

can also stimulate transcription of the CH gene as well as stimulating

CH release (Barinaga et al., 1983).

Turnover of Growth Hormone mRNA

Estimates of the half life of CH mRNA and whether the half-
life in influenced by hormones are also open questions. Evans et al.,
(1984) proposed that regulation of CH gene transcription can entirely
account for the effects of dexamethasone and T3 on CH synthesis. They
did not, however, completely exclude the possibilty of additional

modulation. Diamond and Goodman (1985) measured transcription rates

   

 

21

and Changes in CH mRNA levels in cells induced with T3 plus
dexamethasone. They found a five-fold increase in transcription by
nuclear run-on assay and a 50- fold increase in mRNA levels at 24 hr.

using S analysis. They concluded that the large accumulation of

l
cytoplasmic GH mRNA contrasting with relatively small changes in
transcription was not consistent with a purely transcriptional
mechanism of hormone induction. They estimate 20 hr for the half life
of CH mRNA in the presence of hormones and infer a two hour half life
without hormones present.

Yaffe and Samuels (1984), on the other hand, determined a half
life of 50 hr for CH mRNA in either induced or uninduced cells using
pulse-chase analysis. The chase employed was unusual in that the
initial time point measured was 27 hr after the start of Chase and
measurements were at 15-25 hr intervals after that. When
transcriptional rates were measured they found that immediately after
addition of T3 and dexamethasone to the medium there was a lO-fold
increase in the rate of transcription which then leveled to a new rate
of about 3-4 times the uninduced level. This would result in a
relatively greater increase in CH mRNA accumulation at early times as
compared to later times during the hormone treatment. Therefore, the
time required to achieve half maximal CH mRNA production would be
shorter than the half life of the mRNA. In other words, approach to
steady state would not be a valid method for measuring mRNA half life.
Nyborg et a1. (1984) also found a burst of transcription, followed by
a decrease to a new level of transcription. The effect using T3 was

immediate while the effect with dexamethasone was slower, on the order

of 30-60 min. When CC cells were removed from T3 and dexamethasone-

 

22

containing medium, CH mRNA levels were reduced 92% in 24 hrs. They
concluded that the half-life of CH mRNA in uninduced cells is 8 hrs.

Paek and Axel 1987, have reported that dexamethasone can
affect the half-life of CH mRNA in a biphasic manner. This study
involved transfection of a human CH cDNA under control of the HSV tk
promoter into mouse fibroblasts. Since the HSV tk promoter is not
glucocorticoid responsive, any changes in mRNA levels in the presence
of dexamethasone must be due to stabilization of the mRNA.
Significantly higher levels of mRNA accumulation were found in the
presence of hormone while transcription rate measurements and approach
to steady state showed no alteration due to hormone exposure. Pulse-
chase analysis, using very low cpm, revealed that while 2/3 of the
mRNA turned over at the same rate in induced and noninduced cells,
about 1/3 did not seem to decay at all.

There is, at present, no consensus on the half-life of CH
mRNA. Some of the variability between experiments may result from
differences in culture conditions which produce different responses to
hormone induction of CH protein synthesis. It seems possible that the
hormones which affect CH synthesis may affect CH mRNA turnover in

different ways.

23

MATERIALS AND METHODS

Primapy pituitagy cell culture.

Fresh bovine pituitary glands were obtained at a slaughterhouse
and transported to the laboratory on ice. The intact glands could be
stored on ice at least six hours without affecting the viability and
yield of the resulting cell cultures. After dissection of the tissue
the cell culture preparation must continue uninterrupted until the
cells can be plated. Working under sterile conditions in a hood, the
encapsulated gland was dipped in 70% ethanol, placed on a sterile
cutting board, cut in half sagittally and the anterior pituitary was
dissected free of surrounding connective tissue. Two to three
hemipituitaries were placed on the platform of a Stadie-Riggs slicer
and the tissue was sliced 1 mm thick. The slices were washed in cell
culture medium (without serum), drained briefly and minced into 1 mm
cubes using a double-edged razor blade. All blades used for tissue
preparation were replaced frequently since dull blades compress tissue
and damage cells adjacent to the cut. The pieces were placed in a
beaker and washed with three to four Changes of cell culture medium.
The tissue was transferred to an Erlenmeyer flask and dissociated with
3 mg/ml collagenase (Worthington CLS) in culture medium with gentle
stirring for two hours at 37°C. Generally five to six glands were
dissociated in 150 m1 collagenase solution. After incubation the cell

suspension was filtered thru several layers of cheese Cloth,

 

24

centrifuged at low speed at room temperature for 5 min. to pellet the
cells. The cells were resuspended in culture medium and repelleted
two more times. The cells were suspended in trypsin solution [0.25%
pancreatin (CIBCO), 50 m1/6 glands] and incubated with gentle stirring
for 15-30 min at 37°C. The cells were centrifuged again, washed three
to four times with very low speed centrifugation to separate red blood
cells from the cell mixture. Cells from five to six glands were
resuspended in 20 ml medium and counted. Generally 2-3 x 108 cells
could be recovered from six glands. Cells were diluted in cell
culture medium containing 10% calf serum (CIBCO) and distributed to
dishes at 1 x 105 cells/cmz. Cultures were maintained at 370 in a 5%
CO2 atmosphere.

To optimize maintainence of lactotrophs and somatotrophs in the
cell culture a variety of media were tested. The specific types of
media and additions to the medium are described in the text and figure
legends. Initial experiments used primarily Swim's S-77 medium which
had been especially formulated without methionine and valine (CIBCO).
Unless otherwise stated all media contained D-valine in place of L-
valine to inhibit overgrowth of the primary cultures with cells which
were not hormone secreting cells (Padmanabhan et al., 1982).

Several types of serum (CIBCO) were tested in our experiments.

We found that serum required dialysis to realize the full effects of
D-valine substitution. We either purchased dialyzed serum or dialyzed
normal serum for 2 days with 3 changes against 0.9 M NaCl in PB.

After dialysis the serum was sterilized through 0.2u filters. In some
experiments serum was treated with activated Charcoal (Norite) to

remove steroid hormones. For this 10 g/ml norite was added to

 

25

dialyzed serum and the suspension was stirred for 60 min at room
temperature. The Charcoal was removed by centrifugation at 10,000 x g
rpm for 20 min and then filter sterilized through a 0.45 micro filter
and then through a 0.2 micron filter to remove all residual Charcoal.
Serum-free cell culture conditions were also developed. Several
basic media along with a variety of additions were tested as described
in the text and figure legends. The best medium was MEM D-Va1(CIBCO)
containing 10 ug/ml insulin and 5 ug/ml transferrin, buffered with 15
mM Hepes pH 7.2. For maintainence of fetal pituitary cultures serum-
free medium was superior to medium containing serum but these cultures
showed an absolute requirement for 40 ng/ml hydrocortisone. For all
serum-free cultures the initial cell attachment required serum. Cell
suspensions were plated in medium containing 10% calf serum, allowed
to attach overnight, after which the medium was replaced with serum-
free medium. A variety of cell attachment factors were tested in the
course of these experiments. Inclusion of fetuin in the medium or
pretreatment of the culture dishes with poly-D-lysine, gelatin or

serum did not significantly enhance attachment.

Protein synthesis assay.

Cells were plated at 3 x 106 cells/60 mm culture dish. At the
times indicated the cell culture medium was aspirated and 1 m1 of
labeling medium was added. Labeling medium generally contained 50
uCi/ml 35S-methionine in Swim's medium containing 10% calf serum and 3
ug/ml methionine (20% of normal). At the end of labeling, medium was
removed and cells were washed three times with 0.9% NaCl. Cells were

lysed in 0.3 m1 of 0.5% SDS, 2.5% 2-mercaptoethanol, and 10 mM TrisCl

26

pH 7.5. Since these samples are very viscous, the DNA was sheared by
aspirating several times with a 1 ml syringe through a 22 gauge
needle. Aliquots for measurement of total incorporation into protein
were removed, and the remainder of the sample was boiled for 5 min.
and stored in the freezer until ready for PACE-gel analysis.

To determine total protein synthesis 25 ul of cell lysate was
spotted in duplicate on 2.5 cm Whatman 3MM discs which had been
saturated with methionine. The discs were air dried, transferred to a
beaker containing 10 ml/disc of cold 10% TCA and incubated in ice 15
min. The discs were then transferred to hot 10% TCA and boiled for 10
min. The samples were rinsed briefly in 5% TCA, several times in EtOH
for 1 min each, dried with a heat lamp and counted (Roberts and
Patterson, 1973).

To quantitate prolactin and growth hormone protein synthesis
labeled cell lysates were analyzed by discontinuous polyacrylamide gel
electrophoresis (Laemmli, 1970) in tube gels. Cel tubes were Cleaned
in Chromic acid and silanized using Siliclad or dichlorodimethyl
silane in CHCl3. The 10 cm separating gel was 12.5% acrylamide
(30:0.8 acrylamide:bisacrylamide). The concentration of bisacrylamide
in the separating gel was crucial for resolution of the small PRL and
CH proteins. Bisacrylamide seemed to of variable purity and needed
either recrystallization or new lots had to be screened. The stacking
gel was 3% acrylamide (0.5 cm). Frozen samples were thawed, 10 ul dye
was added to 40 ul of lysate and 50 ul total was applied per tube.
Electrophoresis was at 2 mA/tube for about 8 hr. The gels were
removed from the tubes using a tube eliminator (BioRad) and stained

overnight in 0.05% Coomassie Blue in 25% isopropanol, 10% acetic acid.

 

27

The gels were destained in 10% acetic acid for several hours and
sliced into 2 mm pieces. One to two slices were digested with 1 ml of
30% H202 at 60°C for 5-15 hours and counted in 10 ml Triton/toluene
based counting fluid. NIH prolactin standard was included to verify
the position of PRL cpm. Slab gels were similar to tube gels, except
that 15-20 ul of lysate was applied per well. These gels were 1.5 mm
thick, were run at 50 V until the dye was thru the stacking gel and
then 100V for 5 hr. Staining and destaining were as described above.

The gels were imbedded with PPO in DMSO or treated with Enhance (New

England Nuclear) before autoradiography.

DNA Sypthesis Assay.

Cell division in primary cultures could not be determined simply
by counting cells. In the mixed cell population some cells were very
refractory to trypsin while others lysed with little provocation.

Cell division was determined, instead, by uptake of 3H-thymidine
(Steck et al., 1979). For each assay point, 1 x 106 cells in 35 mm
dishes were labelled for 3 hr at 37°C with 0.5 ml of culture medium
containing 0.5 uCi of 3H-thymidine. At the end of the labelling
period the medium was removed and the plates were rinsed three times
with PBS. The cells were detached with 0.25% trypsin in PBS-EDTA for
10 min at 37°C. Cells were collected directly onto CF/A filters using
a filtering apparatus, washed six times with PBS and lysed with two
washes of cold 10% TCA, followed by two washes with methanol. Filters

were dried and counted in a Triton-toluene scintillation fluid.

Measurement of RNA Turnover.

 

 

28

Cytoplasmic RNA was prepared by rinsing the cell monolayers two
to three times with cold PBS. The cells were scraped in PBS,
centrifuged at low speed and resuspended in 1 m1 of buffer (25 mM
NaCl, 25 mM TrisCl pH 7.5, 5 mM EDTA, 0.5% Triton) per 108 cells.
Cells were broken using 20-30 strokes with a B pestle in the dounce
homogenizer. Nuclei and cell debris were pelleted at 10,000 x g, 5
min. The supernatant was adjusted to 250 ug/ml Proteinase K, 100
ug/ml tRNA, 0.5% SDS and 0.11 M NaOAc and incubated for 25 min at
370C. The RNA was extracted with two volumes of phenol/CHCI3 followed
by two extractions with two volumes of CHCl3 and precipitated in two

volumes of EtOH overnight.

. 3 . . .
Incorporation of H-uridine into total RNA was measured by

 

labeling 5 x 105 cells in 35 mm culture dishes with 0.3 ml medium
containing 80 uCi/3H-uridine. The cells were rinsed with PBS and
lysed in 0.5% SDS, 2.5% 2-mercaptoethanol, and 10 mM TrisCl pH 7.5.
The DNA was sheared with a syringe and 250 ul aliquots were
precipitated with 5 ml of cold 10% TCA. The precipitate was collected

on Whatman CFC filter discs.

Preparation of filter-bound plasmid for RNA hybridization.

Nitrocellulose (S & S) circles, 7 mm in diameter, were produced
using a paperpunch. These were numbered with black ink to identify
the plasmid they would contain, soaked in 2X SSC and arranged on
blotting paper saturated with 2X SSC. Plasmid for binding was
linearized with EcoRI, extracted with phenol/CHC13, and ethanol
precipitated. The plasmid was resuspended in 2 M NHAOH and denatured
and nicked by boiling 1.5 min. The solution was adjusted to 2 M NaCl

 

29

and the final concentration of the plasmid was 50 ug/ml. Each filter
was spotted with 4 ul of plasmid solution. The filters were air dried
and baked at 80°C for 2 hrs, prewashed and air dried (McKnight and
Palmiter, 1979) Prehybridization was for 10-18 hr in 1.2x
hybridization buffer (McKnight and Palmiter, 1979) containing 0.4
mg/ml Poly(A) and 0.4 mg/ml tRNA. Labeled RNA was added at 10% of the
final volume and hybridization was allowed to proceed for more than 48
hrs. Generally 10 ul of hybridization reaction was allowed for each
filter. Reactions were performed in 10 x 75 mm polypropylene tubes
(Sarstadt) tube with a tight fitting cap in a 45°C oven.

Hybridization was performed in an oven instead of a waterbath so that
it was not necessary to overlay the reaction with mineral oil to
prevent evaporation and condensation in the tube. Filters were washed
6 times in 1X SSC containing 0.5% SDS at 680C for 10 min each wash.

To determine specifically bound RNA each filter was placed in a
scintillation vial and the RNA was eluted by heating to 60°C for 1 hr

in 0.5 ml of water.

End-labeling of RNA

Poly A+ RNA was nicked and end-labeled using polynucleotide
kinase (Spradling et al., 1980). RNA (1.5 ug) was nicked by heating
in 0.05 M Tris pH 9.5 for 10 min. at 90°C (2 ul total volume in sealed

capillary tube). This reaction was added to 8 ul of ATP/Kinase (25

uCi gamma-32P-ATP, 2.5 units T -kinase in 62.5 mM Tris pH 7.5, 12.5 mM

4

MgCl 2.5 mM dithiothreitol) and incubated at 37°C for 45 min. The

2’
reaction was terminated by addition of 10 ul 4M NHaAc. Carrier RNA

(100 ug) and H20 to make 200 ul total volume were added and extracted

30

with phenol/CHCl The labeled RNA was precipitated 2 times from 0.2

3.
M NaAC with 2 volumes of ethanol. The yield was 107 cpm/ug RNA.

CLONING 0F CDNA

Preparation of mRNA

An SV40 transformed pituitary cell line, 5A5, was used as a source
of mRNA for CDNA Cloning. Confluent cell cultures were diluted 1:5
and plated at 5 x 106 cells/100 mm dish 36 hr before harvest for RNA.
Monolayers from 40 dishes (3 x 108 cells) were washed three at a time
in cold PBS and scraped into cold PBS. After collection the cells
were pelleted at low speed in a desk top centrifuge and resuspended in
6 ml of lysis buffer [25 mM NaCl, 25 mM TrisCl pH 7.5, 5 mM EDTA, 0.5%
Triton and 10 mM vanadyl nucleotide complex] Cells were broken in a
dounce homogenizer with 40 passes of the B pestle. Nuclei were
pelleted at 8000 rpm for 10 min in an HB4 roter. The supernatant was
adjusted to 600 ug/ml Proteinase K, 0.075 M NaOAc pH 5.5, 0.5% SDS,
and digested for 10 min at 370C. The digestion mixture was extracted

once with phenol, twice with CHCl and ethanol precipitated overnight

3’
and RNA was collected at 8000 RPM for 30 min in an HB4 roter. The
pellet was dissolved in 1 m1 of water and the RNA concentration

determined by absorbance at A The RNA yield in this case was 1

260'
mg. Yields of cytoplasmic RNA using the above procedure varied widely
(1-5 mg/3 x 108 cells) probably because cell breakage using the dounce
homogenizer was not always complete. Intactness of total RNA was

judged by electrophoresis on a minigel of 1% agarose in TBE stained

with ethidium bromide.

 

31

Poly A+ RNA was prepared using oligo dT cellulose fines (PL
Biochemicals) in a batch elution procedure. For 1 mg of cytoplasmic
RNA 25 mg of fines were swelled overnight in water, treated with 0.1 N
NaOH for 30 min, and washed three times with HSB (0.12 M NaCl, 0.01 M
TrisCl pH 7.6, 0.001 M EDTA, 0.2% SDS). RNA was denatured by heating
at 65°C for 2 min, and an equal volume (1 m1) of 2X HSB was added. To
this solution 25 mg of oligo dT cellulose fines in 0.5 ml HSB were
added and the suspension was mixed gently for 15 min. The cellulose
was pelleted and washed with 2 ml HSB 6 times with 5 min of mixing
each time to remove Poly A- RNA. Poly A+ RNA was collected using 3
elutions with LSB (0.01 M Tris Cl, pH 7.6, 0.001 M EDTA, 0.2% SDS) of
0.2 ml each. The LSB fractions were pooled, adjusted to 0.2 M NaOAc

pH 5.5 and precipitated with two volumes of ethanol.

Prepagation of CDNA Libragy.

To prepare the CDNA library Poly A+ RNA was used as a substrate
for oligo dT primed CDNA synthesis using reverse transcriptase.
Double stranded DNA tailed with oligo dC was annealed to pBR322 which
had oligo dC residues added to the PstI site (Maniatis et al., 1982).
The general cloning scheme and important aspects of the reactions are

detailed in Table 2.

Screening CDNA Library

Colonies from the initial transformation were picked to gridded
agar plates, and onto gridded nitrocellulose disks which were layered
on LB agar (master filter). The master filter was grown for 6 hr at

37°C, and replicated according to Maniatis et al., (1982). The

 

 

 

 

32

TABLE 2

CLONING SCHEME

Poly A+ RNA (10 ug RNA)

Reverse Eganscriptase (Dr. J.W. Beard) 210 units
100 uCi P~dNTP, 100 units RNasin (Promega)
20 ug Oligo dT primer 0

225 ul Rxn, 1.25 hr, 42 C

CDNA (1.25 ug, 4.5 x 105cpm)

Phenol/Chloroform extract

Spun column, ethanol ppt.

0.16 N NaOH, 68°C, 10 min, neutralize, ethanol ppt.
Klenow polymerase (Boehringer Mannheim) 35 units/0.6 ug
\y CDNA O

330 ul Rxn, 13 hr, 16 C

 

dsDNA (470 ng)

Phenol/chloroform, ethanol ppt.
(35K rpm, 45 min, SW 50.1)
S1 nuclease 800 units/375 n dsDNA
0.88 ml rxn, 45 min, 37 C

Blunt end dsDNA

Phenol/chloroform, ethanol ppt.

(35K rpm, 45 min, SW 50.1)
Resuspend and precipitate from 0.15 M NaOAc/EtOH
Resuspend and precipitate from 0.12 M NaCl/EtOH
Resuspend in tailing buffer
Terminal transferase (NEN) 20 units/40 ng dsDNA
W 5 uCi H-dCTP, 2515 pmole dCTP
37°C, 45 sec

 

dsDNA with 30 residue dC tail
Phenol/chloroform, spun column in annealing buffer
PstI cut, dC tailed pBR322 (NEN)
4.2 ng tailed dsDNA o o
100 ul rxn, 10 min 68 C, 2 hr 45 C
Annealed vector and plasmid

¢/ transform frozen competent HBlOl

5 x 106 transformants/ug insert DNA (97% Amps)

 

 

33

replica filters were grown 4 hr and each was replicated to two more
filters which were grown for 11 hr. Filters were prepared for
hybridization to 32F labeled probe by the method of Crunstein-Hogness
(Maniatis et al., 1982)

Probes for screening the CDNA library for abundance classes were
prepared from Poly A+ RNA by two methods. One used reverse
transcriptase primed with random primers prepared from calf thymus DNA
according to the procedure described in Maniatis et a1. (1982). The
other used nicked RNA end-labeled with AT32P and polynucleotide

kinase.
ANALYSIS OF CELLULAR SEQUENCES

Northern Analysis

Poly A+ RNA was prepared for analysis on 0.9% Agarose gels by
heating 1 ug of poly A+ RNA in 50% formamide, 6% formaldehyde, 10%
glycerol in 50 ul at 650C for 5 min. The gel buffer and running

buffer were 0.02 M boriC acid, 0.2 mM Na EDTA pH 8.3, 6% formaldehyde.

3
A typical 16 x 16 cm gel used 200 ml agarose and was run at 200 V for

2 hr.
Samples used as controls were stained in 1 ug/ml ethidium bromide

in 0.1 M NHAOAC for 1 hr. The rest of the gel was prepared for

transfer by rinsing 3 times with H O, 5 min each time, then 30 min in

2
0.1M Tris pH 7.5, and 20 min with 1X SSC. Transfer was with 1X SSC to
CeneScreen (NEN) which had been soaked in 1X SSC. Blots were
prehybridized overnight at 45°C in 50% formamide, 5x ssc, 5x

Denhardts, 100 ug/ml salmon sperm DNA, 20 ug/ml Poly A, 25 mM NaPOA,

 

34

1% SDS. For hybridization, 0.25 ml/cm2 of a solution containing 50%
formamide, 5X SSC, 0.5% SDS, 100 ug/ml salmon sperm DNA, 20 ug/ml Poly
A was used to replace the prehybridization solution. This was
incubated at 45°C for two hours and then 106 cpm/ml (2 x 108 cpm/ug)
CDNA containing probe was added and the hybridization continued for 24
hr.

Blots were washed 2 times with 2X SSC at room temperature for 10 min
with shaking. Followed by two changes of 2X SSC/0.1% SDS at 68°C for a

total of 90 min.

Southern Analysis

Total genomic DNA was prepared as described (Friderici, et al.,
1984). Briefly, 5A5 cells were lysed in buffer containing 10 mM Tris
pH 7.6, 10 mM EDTA, 0.5% SDS and 100 ug/ml Proteinase K (3 ml/5 x 107
cells). The lysate was digested overnight at 370C, extracted once

with phenol and twice with CHCl The lysate was ethanol precipated,

3.
resuspended in 0.5ml 10 mM Tris pH 7.6, 1 mM EDTA (TE) and dialyzed
against several changes of TE buffer for 1-2 days. Southern blotting

was performed according to the procedures described in Maniatis et a1.

(1982).

Turnover of Selected mRNA Sequences

5A5 cells were plated at 6 x 106 cells/150 cm2 flask in DME
medium the day before initiation of labeling. Cells were labeled for
2 hr with 0.4 mCi/lO m1 medium/flask. At indicated times one flask
was rinsed with PBS—EDTA, trypsinized, suspended in PBS and pelleted.

Cells were rinsed once with 5 ml PBS containing 4 mM vanadyl complexed

35

nucleotides. Cells were swollen on ice in 1 m1 RSB + 10 mM vanadyl
nucleotides. Cells were dounce homogenized with 20 strokes of the B
pestle, 0.5% Triton X100 was added and cells were rehomogenized.
Nuclei were removed at 5000 x g, 5 min and the supernatant was
adjusted to 100 ug/ml Proteinase K, 0.15 M NaOAc and 0.5% SDS, and
incubated 10 min at 37°C. The cytoplasmic extract was extracted with
phenol and chloroform and ethanol precipitated. The RNA pellet was
resuspended in 50 ul of water.

For hybridization analysis, filters were prehybridized
overnight as described. The prehybridization mixture was removed and
replaced with 90 ul hybridization solution containing 10 ul 3H-uridine
labeled total RNA for 9 filters. Hybridization reactions were
performed in triplicate at 42°C for 3 days. Filters were processed as

described under preparation of filter-bound plasmid.

 

 

36

RESULTS

When primary cell cultures were prepared from pituitary glands
using standard techniques (Vale et al., 1977), a rapid decrease in PRL
and CH production resulted, and the cells exhibited little response to
added hormones. To circumvent this problem we substituted D-valine
for L-valine in the growth medium. Certain cells (primarily
fibroblasts) do not contain the enzymes necessary for conversion of D-
to L-valine and cannot grow and divide in medium without L-valine
(Gilbert and Migeon, 1975). In the pituitary, cells which produce PRL
and CH are of epithelial origin and generally epithelial cells contain
the enzymes necessary for this isomerization.

When primary cell cultures derived from adult animals were
established in L-valine, a mixture of cell types was seen (Figure 1A).
After cells were allowed to grow for 8 days in L-valine containing
medium they produced a confluent monolayer (Figure 1B) dominated by
one or two types of cells. Cells established in D-valine showed less
spreading and cell division 2 days after plating (Figure 1C) and by 8
days, little cell division had occurred as measured by microscopy
(Figure 1D). 3H-thymidine incorporation into TCA precipitable material
(Figure 2A) was very low in cells maintained in D-valine, whereas
cells grown in L-valine showed high levels of incorporation during the
first 7 days of culture and division ceased as the cultures reached

confluence.

37

Figure 1. Primary bovine pituitary cell cultures maintained in D- or
L-valine.
Cells were prepared from bovine pituitary glands as described
in Materials and Methods and plated on cover slips in 35 mm
culture dishes. Growth medium was Swims medium containing 10%
dialyzed calf serum and either L-valine (A and B) or D-valine (C
and D). The cells were stained with trichrome stain at 2 days (A

and C) or 8 days (B and D) after initial plating.

38

 

39

Figure 2. Effect of D-Valine on DNA and Protein synthesis.
Cells were maintained in Swims medium containing 10 % dialyzed
calf serum and either D- or L- valine.
A. DNA synthesis measured by 3H-thymidine incorporation into
TCA precipitable cpm after 3 hr of labeling.
B. Total protein synthesis after 3 hrs of 35S-methionine
incorporation. Hot TCA precipitable cpm were determined for a

portion of the cell lysate.

 

 

 

 

 

 

 

Figure 2

l 7 l i I
A DNA SYNTHESIS
l9 _
9 200- -
X
e
Q
3
LL]
.7:
9
2
>- Ioo- —
I
l—
I
I
'0
L‘VoL
D-VaL
tr i 'l 1* i r" i
3
'5‘
9 PROTEIN SYNTHESIS
x 400—
s
Q
0
DJ
E
Z
9
E
w 200— —
2
J)
3 L-Val
(.4 - ::____________..—oD-Val
1 l J 1 1
IO 20 30

DAYS m CULTURE

 

41

Protein synthesis in D-valine medium remained at a constant
low level. Cells in L-valine showed very high protein synthesis
initially but incorporation decreased as cell division ceased and the
cells became less active metabolically (Figure 2B). Epithelial cells
from adult animals would not necessarily divide but would be expected
to maintain a certain level of protein synthesis.

As a measure of maintenance of differentiated cellular
function, PRL secretion into the medium was measured by
radioimmunoassay (Padmanabhan et a1, 1982). In L-valine medium, PRL
secretion decreased steadily and by day 20 very little secretion
occurred (data not shown). In D-valine medium, however, PRL secretion
begins at the same level but stays high throughout the 30 days these
cells were maintained in primary culture. Cells grown in D-valine
also retained their ability to respond to thyrotropin releasing
hormone by rapidly secreting increased amounts of PRL into the medium.
Higher levels of PRL secretion were observed when cultures were
maintained in 5 ng/ml estradiol. Initially the estrogen response is
about 1.5 fold over controls but decreases with time to D-valine
control levels.

Synthesis of PRL or CH can be used as a measure of the mRNA
level of that protein since there is no translational control of PRL
or CH protein production (Shull and Gorski, 1986). We, therefore,
developed a system for determining the level of de novo protein
synthesis so that optimization of culture conditions and hormone
responsiveness could be evaluated. Cells were labeled with 358-
methionine and lysed and the cell lysate was subjected to disc SDS

polyacrylamide gel eletrophoresis (Figure 3). Even after 28 days in

 

42

Figure 3. Gel profile of newly synthesized protein.
Cells were maintained in either D- or L-Valine Swims medium for
28 days. Cell lysates prepared from cells labeled 3 hr with 353-
methionine were subjected to SDS polyacrylamide gel
electrophoresis. Gels were sliced 2 mm wide and radioactivity

in each slice was determined.

 

43

 

 

 

 

 

 

 

 

 

I l I I I
b LJ'VOL
5 PRL (Leo/O)
w
1
6C; : J. l i l
"’ PRL (l5.0 °/o)
" v
E 0' Vol.
0.
3 1’
Lu
E
2: IOr-
9
I
.—
Lu
2
' GH (1.6 %)
DU)
.0 v
5-
D
I l l J 1
IO 20 30 4O 5O

SLICE NUMBER

Figure 3

44

culture, PRL was the major protein product of cultures maintained in
D-valine medium (Figure 3B). Only a low levels of PRL synthesis was
detected in cultures maintained in L-valine medium. Prolactin
accounted for 10-15% of the total protein synthesis throughout the
culture period (Figure 4A). Growth hormone synthesis also remained
relatively constant in this experiment (Figure 4B). That these major
protein products were, in fact, PRL and CH was verified by comigration
with authentic NIH standard bovine PRL and CH on these gels.
Quantitation of PRL labeling was further analyzed in some lysates by
immune precipitation. It was found that all of the material in the
presumed PRL band could be precipitated with PRL antiserum (data not
shown).

The results obtained for CH synthesis and secretion tended to
be variable. We found that CH production could be maintained only
with certain batches of serum. Serum which originally maintained CH
synthesis lost that ability after Charcoal treatment to remove
steroids. Addition of 40 ng/ml hydrocortisone restored CH production
(Figure 5B). PRL synthesis in these cultures showed very different
results (Figure 5A). Charcoal treatment of serum increased PRL
synthesis, while addition of hydrocotisone depressed PRL synthesis
below control levels. These results can be explained if one considers
endocrine control of the pituitary. PRL production is apparently
primarily under negative control by the hypothalmus. Its primary
inhibitor is dopamine, but hydrocortisone is also known to decrease
PRL production. Hydrocortisone is responsible for increases in CH
synthesis and secretion. Apparently serum from different groups of

animals contained varying levels of hydrocortisone which was removed

45

 

 

 

 

 

 

 

l | l r f I
A.
I5- “
z D- Val.
'—
o —
<1 ..
.4 IO
0
a:
e.
3e 5 - I
L- Vol.
m
2: J
o L i I l :
2 B . f i l I I
2: 5£)r I
:1:
I
i-
; D-Val._
g 25_ a////p—__——-FF_ 9' II
o
o\° L' Vol.
I 3 J : I j E L A 1
IO 20 30

DAYS in CULTURE

Figure 4. Prolactin and growth hormone as a percent of total protein
synthesis.
Cell lysates were analyzed as in Figure 3. Percentage is the
cpm migrating with NIH standard PRL or CH divided by total cpm

recovered from each gel.

46

 

 

 

 

 

 

 

 

 

A I I
15- "
E
[.—
o I0L ‘
4
.1
o
a:
a.
s 5” '
l I
I I
m
z B
o
z —
0: 4r—
0
I
E '5 tr I;
A
3 2— W 4 ..
o
r:
o i #3 4k
g I l ‘0
I0 20

DAYS NICULTURE

Figure 5. Effect of charcoal treatment of serum on CH and PRL
synthesis.

Cells were maintained in D-valine Swims medium containing 10%
of the indicated serum: (-—0—— ) control, untreated serum;
(—EJ-—) Charcoal treated serum; ( —A—) Charcoal treated serum
plus 40 ng/ml hydrocortisone. Conditions for labeling and

quantitation of PRL and CH synthesis were the same as in Figure 3.

47

by the Charcoal treatment. This allowed PRL production to rise due to
lack of repression by hydrocortisone or dopamine, while without these
necessary stimulants CH production decreased.

At this point we had a primary tissue culture system that
could be maintained for three to four weeks with continued PRL and CH
production. However, when I attempted to Change the levels of PRL and
CH protein synthesis by adding hormones to the cultures the magnitude
of the responses was small. The largest effects were seen with
hydrocortisone which increased CH synthesis about five-fold and
decreased PRL synthesis 50% (Figure 5). Results were still somewhat
variable and this was not too surprising when one considers the number
of factors in serum which can affect the synthesis and release of PRL
and CH.

We felt that using a medium that was completely serum-free
might increase the magnitude and reliabibity of these responses.
Serum-free medium has been shown to allow increases in hormonal
response of cells in culture (Barnes and Sato, 1980). A determined
effort was made to establish pituitary cells in serum-free medium.
Several types of commercial media were tested; plates coated with
gelatin, serum, poly-D-lysine, or fibronectin were used to enhance
cell attachment; and fetuin, insulin, transferrin and Hepes were added
to the medium to replace serum components. Since fibroblasts do not
proliferate in serum-free medium, the necessity of using L-Valine was
reexamined. These media, with L-valine, generally supported
considerable cell division for 1-2 weeks followed by a variable
decline in total protein synthesis and a reduced number of cells per

dish. There was a concomitant loss of PRL and CH synthesis. Hams's

48

F12 medium was used under conditions designed to maintain cell
viability in epithelial cells from thyroid tissue (Ambesi-Impiombata
et al., 1980). Using this medium, cell division occurred readily and
viabiliy was maintained but no hormone response was demonstrated and
PRL and CH production declined after 8 days (data not shown).

Serum-free MEM medium containing D-valine (CIBCO) and
supplemented with only transferrin, insulin and HEPES did maintain a
constant level of protein synthyses for more than two weeks. In
serum-free medium, PRL synthesis tends to increase with time in
culture for about five to six days (Figure 6). This probably reflects
release of these cells from the hypothalmic repression that occured in
the intact animal. Dexamethasone, an agonist of hydrocortisone,
depresses PRL synthesis. Growth hormone synthesis showed a more
complete reduction in these cultures than in those maintained in 10%
charcoal treated calf serum (data not shown).

Using this serum-free medium a number of other attempts were
made to produce larger changes in prolactin or growth hormone
synthesis. Bromoergocryptine is a ergot alkaloid that is an agonist
of dopamine. Dopamine is unstable in culture medium but
bromoergocryptine is not. In rat pituitary cell cultures these
compounds greatly reduce the amount of PRL secreted into the medium
and the amount of PRL protein synthesized. When secretion of PRL into
the medium was analyzed (Figure 7) it was found that bromoergocryptine
reduced PRL secretion more than 80%. Hydrocortisone also reduced
secretion of PRL to a lesser extent. While the effect of
bromoergocryptine on PRL secretion is dramatic, the effect on PRL

protein synthesis is less pronounced (Figure 6).

49

 

 

 

 

 

in" T I
'9.
X
e 8"
cg *‘iD
O
z 6"
p.
C)
<1
_J 4"
O
0:
CL
2!—
J I
5 IO
DAYS in CULTURE
Figure 6. Prolactin protein synthesis in serum-free medium.

Primary pituitary cultures maintained in serum-free medium
consisting of D-val MEM (CIBCO) were treated with 10 nM
bromoergocryptine (Ly—fig ), 5 uM dexamethasone (Ck——CI), or
untreated (o-——-o), beginning one day after establishing the
Cell

cultures and continuing throughout the culture period.

cultures were labeled for 1.5 hr. with 35S-methionine.

50

Figure 7. Time course of accumulation of newly synthesized prolactin
in the culture medium.
Cells were maintained in serum-free medium containing 10 nM
bromoergocryptine or 40 ng/ml hydrocortisone for three days prior
to labeling. Aliquots of the medium were denatured with SDS and

analyzed on SDS-polyacrylamide gels.

51

 

l I

0 Control

75 - "

(0pm “0’31

H drocortison
5... v 1
25— ._

PROLAC TIN R E LEASE

 

 

 

_ Bromoergo-
° . cryptine
o o’:“"'
I ‘ l l
5 TO
TIME (hrs)

Figure 7

52

Several other media were also tested. It was suggested
(Mauer, personal communication) that while D-val MEM maintained cell
viability and PRL and CH production, responsiveness to exogeneously
added hormones was diminished. Rat primary cells respond well in
serum-free Dulbecco's medium with insulin, transferrin and HEPES. The
results using this medium with a variety of hormones are shown in Fig.
8. After two days of treatment there has been no effect on PRL levels
and new CH was not synthesized under any condition even though CH
protein was still present in these cells as shown by Coomassie blue
staining of the gel. At seven days there was still no effect on PRL
levels (data not shown). Other methods for preparing cells from the
pituitary gland (Miller et al., 1977) were also evaluated, but had no
noticible effect on hormone responsiveness.

Serum-free pituitary cell cultures maintained in D-valine MEM
responded to hydrocortisone and ergocryptine. The magnitude of the
responses at the level of protein synthesis were not as dramatic as
the secretory response but may reflect what is happening in the intact
animal. Omission of serum from the medium does eliminate concerns
about introducing unknown serum hormones to the cultures.

Fetal pituitary cell cultures were also established in view of
the possibility that hormone responses might be more dramatic in
tissue from animals that had a shorter history of previous exposure to
Circulating hormones. The bovine pituitary differentiates throughout
gestation and mRNA levels change throughout development (Nilson et
al., 1980). Fetal cultures were established using methods similar to
those used for adult tissues and their response to a variety of

hormones was tested (Figure 9 and Figure 10 A-D) It was found that

53

'8; |
Q' a

3‘0..-“

*5!
1 f \
fl a}; ‘

c ca D HC E T3

 

Figure 8. Protein Synthesis in Serum-Free Dulbeccos Medium.
Tissue was prepared in CMF-HBSS (Miller, et al., 1977) and
grown on Poly-D-lysine coated cell culture dishes in Dulbeccos
medium, with insulin and transferrin (C). Additions to the

medium were; lO-8M Bromoergocryptine (CB), 5 x 10‘7 M

Dexamethasone (D), 10'8 M Hydrocortisone (HC), 10.8 M Estrogen

(E), or 3 x 10'8 M Triiodothyronine (T).

54

 

day ( l)( 2 )( 4 )( 6 )( IO )( I4

Figure 9. Protein Synthesis in Cell Cultures From Fetal Pituitaries.
Primary cell Cultures were prepared from bovine fetuses of

mixed gestational ages. Cells were maintained in serum—free D-
valine MEM, supplemented with insulin, transferrin and HEPES.
Treatment with hormones was begun one day after plating cells.
HC-hydrocortisone, 40 ng/ml; CB-CB154(a dopamine agonist), 10 nM;
E-estrogen, 50 ng/ml. At days indicated below the figure, cells
were labeled with 35S-methionine for 90 min. Total cell lysates
were prepared and 12 ul lysate/lane was analyzed directly on SDS-

PAGE (13% acrylamide).

55

Figure 10. PRL and CH synthesis in Fetal Pituitaries.

Fetal cultures maintained in serum-free Swims medium containing
D-valine, 2 ug/ml insulin, 5 ug/ml transferrin, with or without
hydrocortisone (HC).

(A) %PRL = cpm in PRL divided by the total cpm recovered from

the gel.

(B) %CH = cpm in CH divided aby the total cpm recovered from

the gel.

(C) Total Cpm in PRL protein (% PRL x cpm in total protein).

(D) Total cpm in CH protein (% CH x cpm in total protein).

(E) Total 3SS-Methionine incorporation into cellular protein.

56

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

T ‘1 FT T’ 1 ’1
A
+HC
‘5- - IS
A
10,... 'HCO _. ‘0 A/
/
3' / x
n.‘ / 0
R a
’0’ O
5.. ‘,-’ _ 5 ‘\
2 \
+HC \ -HC
‘\0 .',,..o
1 l ‘l‘ l I l
4 9 l4 4 9 14
DAYS DAYS
C I l I] l T
6 A
+HC
A 2- //0\ d A/
‘2 A /' ;‘ Ad”,,”’
2. 0" \ §§4
: \ -HC I
I- -
2 52 O
a a x
u u \
“\
+HCA' \ 'HC
1 l l I Y- - —-o
4 9 l4 4 9 14
DAYS DAYS
5‘ l I I
'0 E
C A
E o.u___ ‘\‘\\\‘;:c
2 \
a \
a20— \ -'
S \
2 -HCb
E 1 J I
V 9 14
DAYS

Figure 10

 

 

57

the fetal pituitary synthesizes much more CH than PRL which is
consistent with the very high level of growth during fetal development
(Figure 9). In these cultures hydrocortisone had the effect of
maintaining the viability of the culture [as measured by continued
protein synthesis, (Figure 10E)] as well as maintaining CH synthesis.
Synthesis of GH in cultures containing hydrocortisone was very high
(18% of total protein at 14 days) and increased steadily throughout
the culture period (Figure 10D). While the effects of hydrocortisone
on PRL synthesis were dramatic (Figure 10A), the effect of
hydrocortisone on PRL synthesis could not be studied in this system
since the cultures did not survive well without it (Figure 10C and E).
Neither estrogen nor bromoergocryptine had any effect on the synthesis
of PRL or CH (Figure 9) in these cultures.

To study the stability of specific mRNAs in cultures, a method
to quantitate small amounts of radiolabeled individual mRNAs using
plasmid DNA bound to nitrocellulose filters was Characterized.
Hybridization and wash conditions were tested and adjusted using
pituitary polysomal Poly A+ RNA that was labeled with 32P using
polynucleotide kinase (Figure 11A). The assay shows that specific
binding increases linearly with increased amounts of RNA indicating
that the filters are not saturated. This was an important control
since prolactin and growth hormone represent a large percentage of the
Poly A+ RNA in the pituitary.

To test this assay for sequences present at much lower levels
a different gene product was used. It was essential to use another
bovine sequence for this work since it had been determined that

stringent wash conditions were necessary to produce reliable results

58

Figure 11. Evaluation of Hybridization Assay Using Filter-bound
Plasmid.

A. Pituitary polysomal poly A+ RNA was labeled with 32P using
polynucleotide kinase. The labeled RNA was hybridized to filters
containing plasmid DNA; pPRL( O ), pCH(,A,) and (,A,) duplicate
hybridization reactions, and pBR322( D ).

B. & C. Plasmid bound filter hybridization using in vivo 3H-
uridine labeled RNA. Transformed cells were labeled in culture
for two hours, cytoplasmic poly A+ RNA was prepared and hybridized

to filters containing a-tubulin ( O ), pCH or pBR322 ( C] ).

59

 

 

I

 

 

 

 

 

 

 

 

l
<
z
m
‘1 w 3
J 4 é) U I A
“—3 9: 0. 8 9
uunqni % wdo punog JGHL-j
I 1
n
l l
o o
S S
mdo ONn08 Halflhd

5

IO
ng RNA

.5

Figure 11

60

with this assay. A plasmid containing Chicken a-tubulin was used to
screen a bovine CDNA library made from pituitary gland Poly A+ RNA. A
bovine -tubulin clone was identified which contained an 1100 bp.
insert. This clone was presumed to be authentic a-tubulin because
the initial screen used washes at 0.2 X SSC at 55°C, but the insert
was not sequenced. In CHO cells a-tubulin represents approximately 1%
of the mRNA and has a turnover of 2-10 hr, depending on the levels of
unpolymerized tubulin in the cells (Cleveland et al., 1981). The
filter hybridization assay was repeated using Poly A+ and Poly A- RNA
from 3H-uridine labeled transformed fetal bovine cells (discussed
below). Of the Poly A+ RNA, 0.1% bound (Figure 11B) while 0.01% of
the Poly A- RNA bound to the a-tubulin plasmid. Filters containing
CH or pBR322 plasmid showed background binding (0.0006%). This assay
was also linear over several RNA concentrations (Figure 11C)

To determine the turnover of PRL, CH, and (x-tubulin in
primary pituitary cultures pulse-chase analysis was attempted. Cells
were labeled for 3.5 hrs. with 3H-uridine and a chase was performed
with medium containing cold uridine and cytidine. Cytoplasmic RNA was
prepared from cultures at time intervals after the start of the Chase
(Figure 12). Although preliminary experiments had indicated that this
chase procedure was adequate, it is clear that the chase did not
become effective until after 15 hr. During the time period between 15
and 27 hour incorporation into total RNA was constant whereas the
amount of label in Poly A+ RNA declined by nearly 50%. In contrast,
PRL and CH mRNA showed a 1.2 and 1.4 fold increase, respectively
(Figure 13) during this interval. Tubulin mRNA was also stable in

this experiment but did not accumulate to the level of PRL mRNA even

61

 

 

 

 

 

0
_a l
‘0 .
I2 %
x -
X
520‘ 0 “ICE
° C.
J 0
<
Z 4
(I Z
I
0
— +
12 id
0)
<1 IO'- - 5 >.
.J .J
0- C)
0 CL
'—
).
0
l l
IO 20
CHASE TIME HHS)
Figure 12. Pulse-Chase of Primary Pituitary Monolayer Cultures.

Cells were labeled for 3.5 hr. with 240 uCi 3H-uridine/S x 610
cells. At the times indicated, cytoplasmic RNA was extracted from

107 cells. An aliquot from each time point was hybridized to

oligo dT cellulose and the amount of poly A+ RNA determined.

62

Figure 13. Specific mRNA Binding to Nitrocellulose Filters During
Chase.
Total RNA from Figure 12 was hybridized to filters. Filters

were washed and eluted as described in Materials and Methods.

63

 

APEV mzfr ww<Io

hum

7| _

h..\\.

c==nzh .

4”?“

n:
4

ma muswfim

J OON

.Joov

 

 

81:90 on x S'I/ w d0

9

64

though tubulin attained a higher level during the 2 hour pulse.
Interestingly, this preliminary evidence showed that PRL and CH had
lower turnover rates than total RNA which would be primarily rRNA.
These data indicated that the mRNA for PRL and CH was very stable and
even under effective Chase conditions would be very difficult to
measure in these non-dividing primary cultures. This is in agreement
with the observations that the levels of PRL and CH mRNA, as measured
by protein synthesis, Change very little and rather slowly in response
to hormones which control their secretion. In the bovine pituitary
cell any control of PRL and CH mRNA levels by alteration of mRNA
stablility does not result in dramatic changes and would be very
difficult to quantitate.

The unusual stability of PRL and CH mRNAs in primary cell
cultures was intriguing. The extreme stability of these mRNAs could
be due in part to the metabolism of the non-dividing lactotrophs and
somatotrophs which synthesize and secrete these proteins.
Alternatively, the stability of these mRNAs could be a result of a
basic sequence difference from other mRNAs. Therefore, I decided to
identify other mRNAs of varying stabilities in cells of bovine
pituitary origin for comparison with PRL and CH mRNAs.

Studies on less stable mRNAs are not possible in these primary
cell cultures since the cell population is so heterogeneous. If mRNAs
other than PRL or CH are studied they could be produced by a subset of
cells that are behaving differently than the lactotrophs and
somatotrophs for which these cultures were optimized. Conditions for
the maintenance of other cell types may not be optimal. Cell death in

a small subset of cells may result in an alteration of the apparent

65

half-life of any mRNA they synthesize. To approach this problem we
attempted to develop SV40 transformed bovine cells which were capable
of producing PRL or CH.

We needed to use fetal cell cultures for SV40 transformation
since this virus requires cell division to establish transformation.
In the adult gland cell division and differentiation has ceased, but
in the fetus the pituitary cells are still proliferating.

Somatotrophs were kept viable and functional by including
hydrocortisone in the growth medium during establishment of the
transformation.

Dr. Robert Lyons established several Clonal cell lines
following SV40 transformation. As expected, many of these cell lines
show very different phenotypes. Papova virus transformation can lead
to a varity of phenotypes but since the original tissue had a variety
of cell types also, it was hoped that the morphological variation
might be due to differences in the parental cells. Several cell lines
were Chosen for further studies. One cell line had the morphology of
an epithelial cell, and appeared to contain secretory granules. There
was preliminary evidence for CH mRNA production in these cells. We
expected to use these lines to study regulation of CH gene expression
in a homogeneous cell population.

To obtain mRNAs from various stability classes I constructed a
CDNA library from the cell line described above. From this library I
hoped to identify a number of bovine mRNAs with varying stabilities to
compare with the very stable CH mRNA. The primary objective of this
approach was to obtain a number of bovine mRNA Clones in each of

several stability classes. The metabolism of these mRNAs could be

66

compared with that of the CH mRNA. Following identification of clones
representing different stability Classes these bovine sequences could
ten be examined to determine how they differ from the PRL and CH
mRNAs. The CDNA library would also be used to identify Clones
representing mRNAs which accumulated to different levels in two
transformed cell lines which showed very different mophologies. The
problem we planned to address was: if two mRNAs reach different
steady-state levels in two cell types, is this accomplished primarily
through transcriptional control or by stability differences.

A CDNA library was constructed using mRNA from an SV40
transformed cell line named 5A5, as described in Materials and
Methods. The transformation efficiency was 9 x 106 colonies/ug insert
DNA or 3 x 106 colonies/ug insert + vector DNA. Approximately 600
colonies were transferred to gridded plates and to duplicate filters.
Preliminary screening was attempted using mRNA from cells labeled in
culture with 32PO4 or 3H-uridine for short and long periods of time.
Messenger RNA from each was prepared and used for hybridization to
duplicate filters containing Clones from the CDNA library. Phosphate
labeling always resulted in very high backgrounds, and 3H-uridine
produced too little signal to be useful. I decided that I would have
to assume that mRNA abundance was sometimes a reflection of RNA
stability and screen the library for abundant, moderately abundant,
and rare mRNAs.

To screen for sequences of varying abundance poly A+ RNA from
SAS cells and a morphologically different cell line, 8A3, was used to
make labeled CDNA using reverse transcriptase with random primers.

Clones representing highly abundant mRNAs were easily identified, and

67

were the same in 8A3 and 5A5 mRNA. A few Clones seemed to represent
higher levels of mRNA in one or the other cell line. These were
Chosen for further study along with some which showed low levels of
hybridization to the mRNA from either cell line.

Approximately 30 Clones were chosen for further study. These
clones had an average insert size of 350 nt with a range of 50-1300
nt. The clones with larger inserts were chosen for further analysis
(Table 3).

Northern analysis was used to determine the sizes of mRNA
these cDNAs represented and to give some indication of their relative
abundance in a transformed cell line (8A3) and polysomal Poly A+ RNA
prepared from intact pituitary glands (Figure 14). The RNA from 8A3
cells was partially degraded resulting in the ladder effect seen in
these RNA samples. RNA from cell line 5A5 was also analyzed. It was
more degraded than 8A3 RNA but showed a similar intensity of signal.
Several of the CDNA clones represented mRNAs which were very abundant
(1/112, 1/105, 3/10, 2/71) as judged by their intensity relative to a-
tubulin. Some were moderately abundant (5/14, 5/38) and some were not
detectable on Northern blots (1/44, 5/48, 5/90). Most were present in
RNA from all three cell types. Clone 5/14 showed no hybridization to
pituitary gland RNA. The sizes of the mRNAs varied from 700-2000 bp
(Table 3). Clone 1/16 is an interesting exception and will be
discussed in Part II. Two Clones, 3/10 and 1/105, showed the same size
and pattern on Northern analysis and were subsequently shown to
contain the same sequence.

Southern analysis showed that most of these CDNA Clones

represent single copy genes (Fig. 15; Table 3). There were two

68

ownmuoouon uoz .o.z

 

 

 

 

 

.tc m cm N. SN ;m_;-uoe ooo~ ooo_ .SSS-a

N mN 20. .o.z so. 00; om\ma

s :. zo_ .o.z m_me_m so. omm w:\ma

: m oumtmnos 0mm. o_mc_m oumtonoe ooM mm\ma

.t; m.w am AN om Sentence om_. oumtonoe oom s_\ma

sol .o.z noe-zo_ ow. Ao\ma

.t; m.@ mNN om G_N ;m_; oom— m_me_m ;m_; om; o.\ma

.c; w mm SN mm cos-em_; ooo. o_m:_m «Coronas osm _~\~a

.t; m.s we GN NA ;m_; owe o_mc_m 3mm; owN ~__\_a

;m_; oom_ o.9;m £m_; osu mo_\_a

N o 30— .o.z umaoc nos 30— own ::\_a

.c_e ms m. mm mam mumtmeos ooom sweat >_;m_; so. omN. o_\_a

mu .t; N .L; p .L; N mocmvcznm Annvo~_m a >Qou oocmvcanm AaAvo~_m oco—u
m__oo m<m >tmE_LmI, m_m>_mc< econucoz ctmcuaom cmoLom <zoo atone.

 

A<zm .mu0u Co emmv mmmx_mc< u_uoc_x

m mmm<b

 

mono—u <zoo >Lmu_:u_m mo co_umN_Louomcmzu

69

Tub 11112 217] MO Illb
’TP__T‘P‘TP,TI_’ TP
—4500 bp
. O .. —2000bp
-I20bp

Figure 14. Northern Analysis of RNA Corresponding to cDNA Clones.
Plasmid DNA from the indicated Clones was nick-translated to a
specific activity of 1-2 x 108 cpm/ug DNA and used to probe
northern blots of Poly A+ RNA (1 ug/lane) from; T, Cytoplasmic
Poly A+ RNA from SV40 transformed 8A3 cells or P, Pituitary

polysomal poly A+ RNA.

70

   

   

.33? '1 1 i

“I...“

Figure 15. Southern Blot of 5A5 DNA Probed with CDNA Clones.

Ten micrograms of high molecular weight DNA from SAS cells was
digested with EcoRI (E), HindIII (H), or Bam HI (B) and resolved
on 0.7% agarose in TAE, blotted, and hybridized three days at
65°C in 6X SSC/2X Denhardts with 2 x 106 cpm/ml nick-translated

Cloned DNA.

71

exceptions; clone 1/44 probably contains a sequence which is
moderately repetitive and Clone 1/16 contains a sequence which is
highly repetitive (data not shown).

I attempted to study the turnover of these mRNAs in primary
cell cultures and in transformed cells. In one experiment with
primary pituitary cell cultures, analysis of the RNA by filter
hybridization showed they did not synthesize detectable levels of PRL
and CH (data not shown). A two hour pulse revealed that these cells
did show appreciable synthesis of some of mRNA corresponding to the
other CDNA clones tested (Table 3).

5A5 cells were used to study the metabolism of mRNA
corresponding to a number of the CDNA clones. To investigate mRNA
turnover in these cultures pulse-chase analysis was performed. At two
hours of pulse the various mRNAs displayed a wide range of
accumulations (Table 3). A preliminary comparison of RNA from
cultures labeled for short time periods shows a very similar pattern
of abundance in the primary cell culture and transformed cells.

The half-life of the clones which showed the highest levels of
label was determined (Fig 16, Table 3). All Clones except p1/16
showed moderately long half-lives of 4-8 hrs. Clone p3/10 which was
very abundant according to the initial library screen, has the highest
transcription rate of the Clones studied and a relatively rapid
turnover (6.5 hr).

Clone p1/16 contains sequences which hybridize to a single
copy, nonrepetitive gene. But it also contains a repetitive element
which is frequently found in introns in hnRNA and to a lesser extent

in the 3' untranslated regions of mRNA. RNA hybridizing to this

72

Figure 16. Decay Kinetics of cDNA Clones.

Cell cultures were pulsed for 2 hrs with 3H-uridine and chased
in medium containing unlabeled uridine and cytidine. Each time
point was analyzed, in triplicate, for binding to each cloned
cDNA. Filter-bound pBR322 DNA was used a background and was

subtracted prior to determination of parts per million (ppm)

bound.

73

 

200

I00

3
50

2O

 

 

 

Figure 16

I l T
fi
q
.I
'1
A
q
D -:
J l l
IO 20 30

TIME (Hours)

 

74

sequence turns over very rapidly (45 min.).

75

DISCUSSION

A bovine primary cell culture system for the study of PRL and
CH synthesis was established and optimized. To maintain long—term PRL
and CH synthesis, D-valine had to be substituted for L-valine in all
the growth media tested. This replacement had the effect of
suppressing cell division in these cultures. When growth was allowed
to procede, the cultures were dominated by a single cell type within
5-7 days of initial plating. PRL and CH production declined as soon
as confluence was reached. Although fibroblast overgrowth is
frequently a problem in primary cultures, the cells which proliferated
in these cultures were probably not fibroblasts. Serum-free medium
will not support fibroblast growth without a number of additional
factors, including fibroblast growth factor. In our serum-free
cultures with L-valine, cell division occurred readily with the
concomitant loss of PRL and CH production seen in serum containing
medium with L-valine. The cells which overgrew these primary cultures
grew in a ‘cobblestone' pattern characteristic of epithelial cells.
Lactotroph and somatotrophs probably do not divide in primary cell
cultures from adult tissue but do remain viable and responsive for at
least 30 days.

Our primary cell culture system is consistently sensitive to
hydrocortisone or dexamethasone when serum is pretreated with charcoal
or when serum-free medium is used. In serum-free medium a more

complete shut-off of growth hormone production was observed in the

76

absence of hydrocortisone. Charcoal treatment may not remove all the
glucocorticoids and does not remove T3. Addition of hydrocortisone to
these cells resulted in a 3-5 fold increase in CH synthesis in adult
and fetal cultures. There are conflicting reports concerning the
effects of hydrocortisone on CH production. In some systems
hydrocortisone by itself produces little or variable response, while
in others it has been shown to induce 2-5 fold increases in CH
production (reviewed by Bancroft et al., 1985). In rat cells CH

production responds to T stimulation but these bovine pituitary cells

3
never exhibited this response.

PRL synthesis in this primary culture system was reliably
inhibited by hydrocortisone. This is in agreement with results using
transformed rat CH cell lines (Evans et al., 1982). In previous
experiments with primary cell cultures from either rats (Lieberman et
al., 1978) or sheep (Vician et al., 1979) no hydrocortisone repression
of PRL synthesis was detected. Shull and Gorski (1986) have
postulated that only transformed cells display an inhibition of PRL
synthesis in the presence of hydrocortisone. We now describe a
primary cell culture system in which PRL synthesis is inhibited by
corticosteroids. It is possible that differences in culture
conditions and media may account for the varying observations.
Alternatively, bovine cells may respond to hormonal stimulation
differently than rat or ovine cells.

Using ovine primary cultures Vician et a1. (1979) found that
addition of estrogen doubled PRL synthesis. We were not able to

demonstrate an estrogen stimulation of PRL synthesis in our bovine

primary cultures. These cells did, however, respond to estrogen with

77

a 70% increase in secretion of PRL (Padmanabhan et al., 1982).
Similar results were obtained using bromoergocryptine, a dopamine
agonist. This drug very effectively blocked secretion of PRL from
primary pituitary cells. Using similar conditions, decreases in PRL
synthesis were not always observed.

Bovine primary pituitary cells respond predictably to hormonal
stimuli at the level of secretion. Changes in protein synthesis were
observed primarily in response to hydrocortisone but not estrogen or
bromoergocryptine. It is possible that dissociation of the pituitary
gland could lead to changes in differentiated function of these cells
and that they respond slightly differently when in association with
other cells (Clayton, et al., 1985). It is also possible that bovine
cells respond differently to hormones than do the rat cells on which
most of the mRNA synthesis studies have been performed.

PRL and CH mRNA synthesis and turnover were studied in these
cultures using pulse-Chase analysis. In these primary cells the chase
did not become effective until 15 hrs after its initiation. It is not
unusual for incorporation into cytoplasmic RNA to increase after
initiation of chase due to continued transport of nuclear sequences to
the cytoplasm. When this occurs, however, it is usually on the order
of only 1-3 hr. It is possible that cells which are not dividing do
not respond rapidly to typical Chase conditions. Between 15 and 27
hours of Chase, label in poly A+ RNA decreased nearly 50% while 3H-
uridine in total RNA remained constant. During this time when poly A+
RNA as a whole decayed 50%, PRL and CH mRNA did not decay but instead
increased slightly. The mRNAs for these proteins showed kinetics very

similar to total or rRNA in these primary cultures indicating that

78

they are very stable messenger RNAs. This correlates well with
evidence from changes in levels of synthesis of these proteins. Under
conditions of stimulation or inhibition, new steady-state protein
levels are not acheived for several days. This would be expected of
mRNAs whose half-lives are very long.

Since the half-lives of PRL and CH were unusually extended in
these cells it was of interest to determine what factors contributed
to this extreme stability. One approach to this question was to
compare CH and PRL mRNA to other bovine mRNAs of various stability
Classes. To begin this study, a CDNA library was prepared from a
transformed bovine pituitary cell line which initially appeared to
produce CH.

Results from CDNA Clones of bovine mRNAs are basically
descriptive. The majority (75%) of the clones examined by Southern
analysis represented unique genes (Table 3). One had a highly
repetitive short interspersed sequence and another contained a
moderately repetitive sequence.

Northern analysis detected only abundant or moderately
abundant sequences. The mRNA represented by these clones varied in
size from less than 700 bp to 2000 bp (Table 3). The insert sizes
(Table 3) in these plasmids are approximately 1/4 to 1/2 the full
length of the mRNA they represent. Most clones probably contain the
3'end of the messenger RNA since the average size of the first strand
produced by reverse transcriptase in the CDNA cloning was 500-600 bp.
Second strand synthesis went to completion with an average mass of
1200 to 1400 bp (data not shown). Therefore, the reverse

transcriptase copied part of the mRNA starting at the poly A tail, and

79

then the Klenow polymerase made a complete copy of that fragment.

Preliminary kinetic analysis of some of these Clones was
performed in primary pituitary cell cultures and in transformed bovine
pituitary cells (Table 3). Clones which were determined to be of low
abundance on either Northern analysis, or in the initial library
screen using 32P-CDNA copied from transformed cell mRNA, were labeled
to very low levels in the 1 hr or 2 hr pulse. Interestingly, the
levels of labeling attained in the primary pituitary cells at two hr.
were very similar to those found in the transformed cells at two hr.

In the transformed cells, the most abundant mRNAs incorporated
enough label during the pulse to allow determination of their turnover
rates. The half-lives of these mRNAs varied from 4.5 to 8.5 hr.

One Clone, p3/10, was very abundant. It hybridized very
strongly on the initial CDNA library screens. Of the approximately 20
clones analyzed in more detail two were found to contain homologous
sequences (p3/10 and pl/105). This mRNA probably has a very high
transcription rate since it contained 2-3 X more 3H-uridine than any
of the other mRNAs at 2 hours in both primary and transformed cells.
On the other hand, it has an intermediate turnover rate of 6.5 hr
(Table 3). This mRNA could be interesting to compare to PRL and CH
since it accumulates to high levels in these cells but seems to do so
by a very different mode than do CH and PRL mRNA.

RNA which hybridized to Clone p1/16 had a short half-life of
45 min. This clone contained a short interspersed repetitive element
(Characterized in Part 11) These repetitive elements are frequently
found as part of Pol II transcripts, usually in intron sequences and

occasionally in the 3' untranslated end of mature mRNAs (Jelinek and

80

Schmid, 1982). The labeled RNA sequences hybridizing to clone pl/16
could represent hnRNA which may have contaminated the cytoplasmic RNA
preparation. This would be consistent with the larger size of the RNA
sequences hybridizing to this clone in a Northern analysis.
Alternatively, mature mRNA containing a repetitive element in an

untranslated region may be turned over very rapidly in these cells.

81

REFERENCES

Ambesi-Impiombato, C., Parks, A.B. and Coon, H. 1981. Proc. Natl.
Acad. Sci. USA 77:3455-3459.

Bancroft, F.C., 1981. In "Functionally Differentiated Cell Lines"
(Sato, C. ed.) 47-60. Alan R. Liss. InC., New York.

Bancroft, C., CiCk, C.C., Johnson H.E. and White B.A. 1985.
Biochemical Action of Hormones 12:173-213.

Barinaga, M., Yamamoto, C. Revier, C. Vale, W., Evans R., and
Rosenfeld M.F. 1983. Nature 306:84-85.

Barnes, D. and Sato, C. 1980. Anal. Biochem. 1022255—270.

Berger, S.L., and Cooper, H.L. 1975. Proc. Nat. Acad. Sci. USA.
72:3873-3877.

Brock, M.L., and Shapiro, D.J. 1983. Cell 34:207-214.

Carneiro, M., and Schibler, U. 1984. J. Mol. Biol. 178:869-880.

Caron, M.C., Beaulieu, M., Raymond, V., Cagne, B., Prouin, J.,
Lefkowitz, R.J. and Labrie F. 1978. J. Biol. Chem. 253 2244-
2253.

Carrillo, A.J., Pool, T.B., and Sharp, Z.D. 1985. Endocrinology
116:202-206.

Clayton, D.F., Harrelson, A.L., and Darnell, J.E. 1985. Mol. Cell.
Biol. 5:2623-2632.

Cleveland, D.W., Lopata, M.A., Sherline, P., and Kirschener, M.W.
1981. Cell 25:537-546.

Cleveland, D.W. and Havercroft, J.C. 1983. J. Cell Biol. 91:919-
924.

Cleveland, D.W. 1987. J. Cell Biol. 104:381-383.

Cronin, M.J., Roberts, J.M., and Weiner, R.I. 1978. Endocrinology
103:302-309.

Dannies, P.S., and Tam, S.W. 1982. In "Cellular Regulation of
Secretion and Release" (Conn, P.M. ed.) 529-568 Academic Press,

82

New York.
Dannies, P.S. and Tashjian 1973. J. Biol. Chem. 248:6174-6179.
Dannies, P.S. 1985. Biochemical Action of Hormones 12:289-310
Darnell, J.E. 1982. Nature 297:365-371

Derman, E., Krauter, K., Walling, L., Weinberger, C., Ray, M., and
Darnell, J.E. 1981. Cell 23:731-739.

DiPasolo, T., Carmichael, R., Labrie, F., and Raynaud, J.P. 1979.
Mol. Cell. Endocrinol 16:99-112.

Dobner, P.R., Kawasaki, E.S., Yu, L.-Y., and Bancroft, F.C. 1981.
Proc. Natl. Acad. Sci. USA 78:2230-2234.

Diamond, D.J. and Goodman, H.M. 1985. J. Mol. Biol. 181 41-62.

Evans, R.M., Birnberg, N.C., and Rosenfeld, M.C. 1982. Proc. Natl.
Acad. Sci. USA 79:7659-7663.

Farquhar, M.C., 1977. In "Comparative Endocrinology of Prolactin"
(H.D. Dellman, J.A. Johnson, and D.M. Klachko, eds.) 37-94

Plenum, New York.

Friderici, K., Oh, S-Y., Ellis, R., Cuacci, V., and Fluck, M.M.. 1984.
Virology 137:67-78.

Cay, D.A., Yen, T.J., Lau, J.T.Y., and Cleveland, D.W. 1987. Cell
50:671-679.

Cick, C.C. and Bancroft, C. 1985. J. Biol. Chem. 16017614-7618.
Cersten, B.E. and Baker, B.L. 1970. Am. J. Anat. 12811-20.
Gilbert, S.F. and Migeon, B.R. 1975. Cell 5 11-17.

Graves, R.A., Pandey, N B., Chadchoy, N., and Marzluff, W.F. 1987.
Cell 48:615-626.

Graves, R.A. and Marzluff, W.F. 1984. Mol. Cell. Biol. 42351-354.
Creenberg, J.R. 1972. Nature 240:102-104.

Creenberg, M.E., Hermanowaki, A.L., and Ziff, E.B. 1986. Mol. Cell.
Biol. 6:1050-1057.

Cuyette, W.A., Matuski, R.J., Rosen, J.M. 1979. Cell 17:1013-1023.

Harpold, M.M., Evans, R.M., Salditt-Ceorgieff, M., and Darnell, J.E.
1979. Cell 17:1025-1035.

 

83
Heintz, N., Sive, H.L., and Roeder, R.C. 1983. Mol. Cell. Biol.
3:539-550.
Jelinck, W.R. and Schmid, C.W. 1982. Ann. Rev. Biochem. 51:813-844.
Jones, T.R., and Cole, M.D. 1987. Mol. Cell. Biol. 7:4513-4521.
Laemmli, U. 1970. Nature 227:680-686.

Lau, L.F. and Nathans, D. 1987. Proc. Natl. Acad. Sci. USA 84:1182-
1186.

Laverriere, J.N., Morin, A., Tixier-Vidal, A., Truong, A.T., Courdji,
D., and Martial, J.A. 1983. EMBO J. 2:1493-1499.

Lieberman, M.E., Maurer, R.A., and Gorski, J. 1978. Proc. Natl.
Acad. Sci. USA 75:5946-5949.

Lieberman, M.E., Maurer, R.A., Claude, P., and Gorski, J. 1982.
Mol. Cell. Endocrinol. 25:277-294.

MacLeod, R.M., Fontham, E.H., and Lehmeyer, J.E. 1970.
Neuroendocrinology 6:283-294.

Maniatis, T., Fritsch, E.F., and Sambrook, J. 1982. "Molecular
Cloning -- A Laboratory Manual". Cold Spring Harbor Laboratory
N.Y.

Martial, J.A., Baxter, J.D., Goodman, H.M., and Seeburg, P.H. 1977.
Proc. Natl. Acad. Sci. USA 74:1816-1820.

Maurer, R.A., 1980. J. Biol. Chem. 255:8092-8097.
Maurer, R.A., 1981. Nature 294:94-97.

Maurer, R.A., l982a. Endocrinology 110:1507-1514.
Maurer, R.A., l982b. Endocrinology 11011515-1520.
Maurer, R.A., 1982C. Endocrinology 110:1957-1963.

Maurer, R.A., 1982d. in "Cellular Regulation of Secretion and
Release" (Conn, P.M. ed.) 267-292 Academic Press, New York.

McKnight, C.S. and Palmiter, R.D. 1979. J. Biol. Chem. 254:9050-
9058.

Meites, J. and Clemens, J.A , 1972. Recent Prog. Horm. Res. 28:165-
221.

Miller, W.L., Knight, M.M., Crinek, H.J., and Gorski, J. 1977.
Endocrinology 100:1306-1316.

84

Murdoch, C.H., Potter, E., Necolaisen, A.K., Evans, R.M., and
Rosenfeld, M.C. 1982. Nature 300:192-194.

Murdoch, C.H., Rosenfeld, M.C., and Evans, R.M. 1982b. Science
218:1315-1317.

Murdoch, C.H., Evans, R.M., and Rosenfeld, M.C., 1985. Biochemical
Action of Hormones 12:35-68.

Nilson, J.H., Barringer, K.J., Convey, E.M., Friderici, K.A., and
Rottman, P.M. 1980, J. Biol. Chem. 255:5871-5885.

Nyborg, J.R., Nguyen, A.P., and Spindler, S.R., 1984. J. Biol. Chem.

Padmanabhan, V., Friderici, K.H., Convey, E.M., and Rottman, P.M.
1982. Mol. Cell. Endocr. 28:613-626.

Pandey, N.B. and Marzluff, W.F. 1987. Mol. Cell. Biol. 7:4557-4559.
Paek, I. and Axel, R. 1987. Mol. Cell. Biol. 7:1496-1507.

Perrone, M.H., Greer, T.L., and Hinkle, P.M. 1980. Endocrinology
106:600-605.

Piechaczyk, M., Yang, J.Q., Yang, Blanchard, J.M., Jeanteur, P., and
Marcu, K.B. 1985. Cell 42:589-597.

Pittenger, M.R. and Cleveland, D.W. 1985. J. Cell Biol. 101:1941-
1952.

Potter, B., Nicolaisen, A.K., Ong, E.S., Evans, R.M., and Rosenfeld,
M.C. 1981. Proc. Natl. Acad. Sci. USA 78:6662-6666.

Puckett, L., Chambers, 8., and Darnell, J.E. 1975. Proc. Natl. Acad.
Sci. USA 72:389-393.

Raymond, V., Beulieu, M., Labrie, F., and Boissur, J. 1978.
Science 200:1173-1175.

Rivier, J., Speiss, J., Thorner, M., and Vale, W. 1982. Nature
300:276-278.

Roberts, B.E. and Patterson, B.M. 1973. Proc. Nat. Acad. Sci. USA
70:2330-2334.

Ross, J., and Kobs, C. 1986. J. M01. Biol. 188:579-593.
Ross, J., and Pizarro, A. 1983. J. Mol. Biol. 167:607-617.

Samuels, H.H., Stanley, F., and Shapiro, L.E. 1979. Biochemistry
18:715-721.

Schumperli, D. 1986. Cell 45 471-472.

85

Shapiro, D.J., and Brock, M.L. 1985. "Biochemical Actions of
Hormones" Vol XII. Academic Press, Inc.

Shaw, C., and Kamen, R. 1986. Cell 46:659-677.

Shull, J.D., and Gorski, J. 1986. Vitamins and Hormones 43, 197-
249.

Singer, R.H., and Penman, S. 1973. J. Mol. Biol 78:321-334.

Spradling, A.C., Digan, M.E., Mahowold, A.P., Scot, M., and Craig, E.A.
1980. Cell 19:905-915.

Stanley, F., and Samuels, H.H. 1984. J. Biol. Chem. 259:9768-9775.

Steck, P.A., Voss, F.C., and Wang, J.L., 1979. J. Cell. Biol.
83:562-575.

Vale, W., Rivier, C., and Brown, M.A. 1977. Ann. Rev. Physiol
39:473-525.

Vician, L., Shupnik, M.A., and Gorski, J. 1979. Endocrinology
104:736-743.

Vollach, V., and Housman, D. 1981. Cell 23:509-514.

White, B.A., and Bancroft, F.C. 1983. J. Biol. Chem. 258:4618-
4622.

Yaffe, B.M., and Samuels, H.H. 1984. J. Biol. Chem. 259:6284-6291.

PART II

CHARACTERIZATION OF A

BOVINE REPETITIVE ELEMENT

86

87

LITERATURE REVIEW

The eukaryotic genome is a mixture of single copy sequences
interspersed with sequences that are repeated many times. In mammals
these repetitive sequences may account for 10-30% of the genomic DNA
(Laird, 1971). Repetitive sequences can be broadly Classified as
highly repetitive ( >106 copies per genome) or moderately repetitive
(103 - 105 copies per genome) (Britten and Kohn, 1968).

Highly repetitive DNA consists of tandemly repeated short
sequences. These segments frequently have buoyant densities in cesium
chloride gradients that differ from the bulk of an organism's DNA so
that they form satellite bands and are present mainly at centromeric
and telemeric positions on the Chromosome (Jelinek and Schmid, 1982).
No transcriptional activity has been described from this type of
repetitive DNA.

Moderately repetitive DNA is widely distributed throughout the
genome, and is interspersed in unique sequences. These repetitive
sequences are divided into two Classes according to their length; long
(6-7 kb) and short (<500 kb). Mammalian genomes contain a long
interspersed element known as the LINE (L1) family (Singer, 1982),
whose best studied members are the Kpn family in primates and BamHI
(MIFl) family in the mouse. Only a single LINE family has been found
in each mammalian species studied (Weiner et al., 1986). The LINE

families of different species are very homologous over a portion of

88

their length, somewhat homologous in other portions and show species
specific divergence over the rest. Family members are heterogeneous
in size, frequently showing a variable truncation of what has been
designated the "5'" end. There can also be internal rearrangements,
including deletions and inversions. Several human and mouse segments
contain open reading frames over 600 bp long near the 3' end, in the
region of mouse and human homology (Singer and Skowronski, 1985). The
extreme 3' end of the elements in the L1 family is variable in length
and rich in A-residues. These elements are apparently mobile in the
genome and make target site duplications at the site of insertion.

Little is known about the function(s) of long interspersed
repeats. Some may contain functional genes capable of producing
proteins. They could also influence the expression of neighboring
genes either by enhancing (Leuders et al., 1984) or silencing (Laimins
et al., 1986) their transcription. It has also been suggested that L1
units could foster rearrangements both within Ll elements and within
adjacent genomic regions (Singer and Skowronski, 1985).

Short dispersed repeats (SINES) consist of families of
sequences approximately 70-300 bp that are present at over 100,000
copies per genome. Each family is a set of elements that are Closely
similar in sequence and length but are not identical. The first
sequence of this type was described in the human genome and named the
"alu family" (Houck et al., 1979). This is the major family of
repetitive elements in humans and typifies the short interspersed
repetitive elements found in many species.

In general the alu sequences are 300 bp in length and consist

of an imperfect dimer of two directly repeated, approximately 130 bp

89

units with a 31 bp insertion in the second monomer. The nucleotide
sequences of different members of this family tend to differ from a
consensus sequence by about 14% (Schmid and Shen, 1986). The alu
sequence is highly defined at the 5' end and contains a variable
length A-rich sequence interrupted by other nucleotides at the 3' end.
In the genome this whole structure is flanked by direct repeats of 8
to 20 base pairs that are unique for each alu insert.

Alu family members are fairly evenly dispersed throughout the
human genome, and occur on the average every 5000-8000 bp (Jelinek
and Schmid, 1982). They have been found both within RNA polymerase II
transcription units and between transcription units (Fritsch et al.,
1980). Alu sequences are heavily represented in hnRNA molecules and
comprise as much as 18-25% of the mass of HeLa hnRNA (Federoff et al.,
1977, Jelinek et al., 1978). Poly A+, polyribosome-associated
cytoplasmic HeLa cell RNA also contains these sequences, but at a
lower frequency per molecule than hnRNA (Jelinek et al., 1978).
Apparently alu units can occur both within introns and in the 5’ or 3'
noncoding regions of messenger RNAs.

Alu family sequences also contain the A and B block promoter
elements necessary for RNA polymerase III transcription (Duncan et
al., 1981, Fowlkes and Shenk, 1980). One or more A blocks lie Close
to the 5' end of the element (positions 5 and 31) and a B block is
found between position 70 and 100. Alu sequences can be transcribed
in vitro or in vivo by RNA polymerase III to yield low molecular

weight RNAs (Duncan et al., 1979, Allan and Paul, 1984).

90

A summary of the salient features of the alu family are

diagrammed below.

d.r. : A L U S E Q U E N C E : A-rich sequence : d.r.
41::lst monomer -—————- 2nd monomer
A & B blocks 3lbp insert

All mammalian species contain short interspersed repetitive
elements which show varying homologies or structural similarities to
the human alu family. In mice there are two major SINE families
designated B1 and 82 which occur at about 105 copies per genome
(Kramerov et al., 1977, 1979). The 130 bp Bl repeat is homologous to
the first momomer of human Alu, and shares homology with the polyoma
viral replication origin (Karayev et al., 1982). B2 resembles B1 in
its general structure and organization but has a different concensus
sequence. Both have a split Pol III promoter, 3' poly A sequence and
are flanked by direct repeats (Krayev et al., 1982). B1 and B2 are
found in both hnRNA and mRNA as part of RNA Pol II transcription units
(Kramerov et al., 1982, Ryskov et al., 1983).

Another repetitive sequence found in rodents is the rat ID
sequence (Sutcliffe et al., 1982). This is an abundant (105
COpies/genome) SINE in rats, while the homologous sequence in mice is
present at 104 copies/genome (Sapienza and StJacques, 1986). ID
sequences are 82 bp long, have a split Pol III promoter and a 3' poly
A tail of variable length.

ID sequences were originally thought to be part of pol II

transcription units that were only expressed in brain nuclei

91

(Sutcliffe et al., 1984b, Milner et al., 1984), and it was speculated
that they could possibly function as some sort of tissue specific
control signal for neural specific genes. Subsequent studies have
shown, however, that ID sequences are present in the hnRNA synthesized
in other tissues at similar levels to that found in brain hnRNA (Owens
et al., 1985, Sapienza and St-Jacques, 1986).

In addition to being expressed as part of larger Pol II
transcription units, short interspersed repetitive elements from
humans (Duncan et al., 1979), rodents (Haynes and Jelinek, 1981),
cows (Furth and Su, 1986) and lower vertebrates (Matsumoto et al.,
1983, Endoh and Okada, 1986) can all function as templates for RNA

polymerase III in ip vitro transcription systems. In these systems the

 

RNA products are usually longer than the repetitive sequence being
examined. One end of the transcript corresponds exactly to the 5’ end
of the repetitive unit but the 3' end generally terminates outside the
element yielding an RNA transcript longer than the actual repetitive
unit.

To date, Pol III trancription from repetitive elements in vivo
has been demonstrated primarily in rodent cells (Haynes and Jelinek,
1981, Scott et al., 1983, Singh et al., 1985, Sutcliff et al., 1984a),
where expression of these elements shows an intriguing tissue specific
expression. B01 and BC2 are the Pol III transcription products of rat
ID sequences (Sutcliffe,l984a, Owens et al., 1985). These small
cytoplasmic RNAs are found pply in brain and neural tissue and in
certain established cell lines (McKinnon et al.,l987). Their
expression begins in rat brain immediately after birth and is found in

large amounts 5-10 days postnatally.

 

 

92

Pol III transcription of the mouse repetitive B1 element has
not been shown in vivo but the B2 element is transcribed under certain
Circumstances. B2 transcription in mouse cells is induced by
transformation with SV40 (Scott et al., 1983, Singh et al., 1985),
other transforming viruses, and carcinogens (Scott et al., 1983). It
is elevated in early embryos and undifferentiated embryonal carcinoma
(EC) cells and decreases as development proceeds or upon in vitro
differentiation of EC cells (Murphy et al., 1983,). These B2
sequences are found to be induced by serum in secondary cultures of
mouse embryo fibroblasts (Edwards et al., 1985). It appears,
therefore that the mouse 82 pol III product is induced in response to
growth stimulation of cells.

Short interspersed repetitive elements in the ruminant genome
share some but not all of the structural features seen in other
species studied to date. In goats and cattle there are about 10
copies/genome of a 120 bp repetitive unit (named A or BCS) which shows
limited homology to the replication origin of papova viruses. These
elements are flanked by direct repeats and have a sharply defined 5'
end but do not have an A-rich 3' tail, possessing instead short tandem
repeats of variable length in the 3' region. There has been no
evidence of a Pol III split promoter in this A sequence (Watanabe et
al., 1982, Rogers, 1985). The ruminant genome also contains an
additional repeat of 73 bp called the C repeat (Schon et al.,l981).
This unit has the typical Pol III promoter but does not have a poly A
tail. To date the C element has only been found paired with A
elements.

How alu type sequences may have dispersed throughout the

93

eukaryotic genome is suggested by the flanking direct repeats at their
borders and by their general structure, including the pol III promoter
and Poly A sequences. These features are typical of retroposons
(Rogers 1985, Weiner et al., 1986) where RNA sequences transcribed
from short repeated sequences are reverse transcribed and reinserted
into the genome at another site.

The alu/Bl family is thought to derive from 7SL RNA by one or
more internal deletions of the 7SL RNA sequence followed in primates
by dimerization (Weiner et al., 1986). The 78L RNA itself is highly
conserved throughout evolution and is a component of the RNP required
for cotranslational secretion of membrane and secretory proteins. In
rodents there is another RNA (4.58) which is also homologous to a
portion of the B1 repetitive sequence. The function of this RNA is
not known but it is found hydrogen bonded to mRNA and hnRNA in mouse
cells (Schoeniger and Jelinek, 1986).

Although the Alu/Bl family arose from 7SL RNA, many other
families of SINEs are apparently derived from tRNAs. There is, of
course, homology between the internal Pol III promoters of SINEs and
tRNA. For some families of repetitive elements homology to tRNA
extends beyond the promoters and into the adjoining region. Such
homology was first noted between the goat C family and a cysteine tRNA
(Rogers, 1985). The rat ID, mouse B2, rabbit C and bovine 73 bp
repeats were also found to display homologies to tRNA, although
sequence divergence within the repetitive elements has led to
disagreement on which is the exact parental tRNA species (Daniels and
Deninger 1985, Lawrence et al., 1985, Sakamoto and Okada 1985). In

all cases the tRNA homology lies at the 5' end of the element as

94

expected for an internal Pol III promoter. In the case of elements
which are longer than a typical tRNA sequence the 3' portion may have
originated as a read-thru transcript from the tRNA gene.

There are also interesting composites of repetitive elements.
The prosimian, Calagos, has a repetitive family called monomer which
is 68% homologous to met tRNA and is frequently found in combination
with the right monomer of the alu family. The three repetitive
families in Galagos; alu, alu/monomer, and monomer are all present in
this prosimian genome in approximately equal portions (Daniels and
Deninger 1985). Coats and cattle have a similar situation in which
the C repeat is found linked to the 5' end of the A sequence (Schon et
al.,l98l, Watanabe et al., 1982). Alternatively, A sequences are
linked to each other other with the sequence (CACTTT)3CATCCATT
repeated between them (Watanabe et a1, 1982, Skowronski et al., 1984).

Before discussing the possible functions of SINE sequences it
is instructive to consider how they may have evolved. SINE sequences
are relatively homologous within each species but are often
characteristically different between species. To account for this
phenomenon it has been suggested that the expansion of these
repetitive sequences within the genome began quite recently in
evolutionary time and may be continuing today (Weiner et al., 1986).
Mutations occurring when there were a few copies in the genome
(founder sequences) would alter the consensus sequence somewhat
between two species. Once the copy number increased it would be
difficult for subsequent mutations to alter the consensus. It is not
known whether the expansion is exponential (each new element can

retro-transpose as easily as its parent) or linear (a smaller number

95

of elements functioning as active retroposons).

The possible functions of these small but numerous inhabitants
of the vertebrate genome have been the focus of a great deal of
speculation. Early suggestions involved control of gene expression or
origins of DNA replication (Britten and Davidson, 1969). It is
difficult to reconcile an overriding function such as DNA replication,
where the cellular machinery is well conserved, with sequences which
are so heterogeneous between species. On the other hand having
dispersed Pol III promoters in the genome could conceivably serve some
generalized function (Daniels and Deninger, 1985).

The possibility that repetitive sequences function in the
tissue or differentiation stage specific control of gene expression
has been eagerly pursued. Particularily tantalizing is the fact that
SINEs are found as part of Pol II transcription products, both as
hnRNA and mRNA. The rat brain ID sequence was proposed as an example
of tissue specific expression of a repetitive element (Sutcliffe et
al, 1984b), but it now appears that the specificity lies not in the
expression of Pol II products (Owens et al., 1985) but of Pol III
products to yield low molecular weight RNAs found only in brain and
certain cell lines (McKinnon et al., 1987, Owens et al., 1985).
Initial observations in mouse systems gave indications that B2
sequences might be associated with genes expressed only in transformed
cells (Scott, et a1. 1983, Murphy, et a1. 1983) Again most of the
increase in cellular levels of these sequences was due to an increase
in Pol III transcription of the BZ repetitive element (Singh et a1.
1985). The significance of this Pol III stimulation in rodent cells

is not known. Recently Clemens (1987) observed that a small region of

96

the B2 consensus sequence is complementary to the AU rich sequence at
the 3' end of certain inducible mRNAs. This AU sequence has been
shown to be responsible for the short half-life of some transiently
expressed mRNAs in human and animal cells (Shaw and Kamen, 1986).
Clemens (1987) proposes that Changes in the composition of the B2 RNA
population might differentially regulate the susceptibility of
individual mRNAs to nucleolytic degradation in the cytoplasm.

The main reason for doubting a functional role for alu-like

 

sequences is their extreme lability in evolution. Related species can
have divergent consensus sequences and different mammalian orders have
entirely different consensuses (Rogers, 1985). Apparently SINEs
evolved after many of the basic cellular control systems were
determined. The individual locations of repetitive elements in
expressed mRNAs are also not conserved between species, i.e. rat
growth hormone has a SINE in an intron, human does not (Malissen et
al., 1982). The mouse MAK gene has been found to have a B1 element in
its 3' untranslated region in some but not all strains of mice. The
repetitive unit does not appear to alter the expression of this gene
in any way (King et al., 1986).

It is possible that repetitive sequences have no physiological
function but act as selfish DNA (Orgel and Crick, 1980, Doolittle and
Sapeinza, 1980). In this model, sequences which have Pol III
promoters are able to retrotranspose and have been selected only for
their efficiency to do so. It is possible that repetitive elements
are found more frequently in genes that are expressed in an
undifferentiated cell since these may be the ones more likely to be

expressed in the germ line, and active chromatin in the germline may

97

be more accessible to retroposition (Weiner el al., 1986).

Somewhere between these extremes is a suggestion by Hess et
a1. (1983) that one consequence of insertion of a repetitive element
may be blockage of gene conversion in stretches of homologous DNA. So
far, careful examination of the globin family in a number of species
has supported this proposal (Schimenti and Duncan, 1984, Spence et
al., 1985, Brunner et al., 1986). Insertion of a repetitive element
between duplicated genes might disrupt the gene correction processes
and allow genes to drift independently, thus permitting modified or
divergent function in what was originally one gene.

Retroposition may help maintain the complexity and fluidity of
eukaryotic genomes (Weiner et al., 1986). The abundance of
transposable elements suggests that the excess DNA is not subject to
strong negative selection. The wealth of genetic variation may serve
as raw material for positive selection, as well as for negative

selection and neutral drift (Rogers, 1985).

 

98

INTRODUCTION

A CDNA library was constructed from an SV40-transformed bovine
pituitary cell line. From this library a Clone (pB1/6) was isolated
which contained a repetitive sequence. Since ruminant repetitive
elements which have been described show some unusual structural features
this clone was Characterized to determine the nature of its repetitive
sequence. Transcription of repetitive elements from a number of species
has been demonstrated under certain conditions of cell growth or
differentiation. The expression of this bovine repetitive element was

examined in both normal and transformed cells.

99

MATERIALS AND METHODS

Construction of CDNA library and DNA analysis.
Construction of the CDNA library, preparation of genomic
cellular DNA and Southern blot analysis were described in part one of

this thesis.

Dot blot analysis of species homology.

DNA from various vertebrates was prepared by phenol/chloroform
extraction and ethanol precipitation. Concentrations of DNA solutions
were determined by absorbance at 260 nm. Hamster and human DNA was a
gift from Dr. Paul Bates. Serial dilutions of DNA ranging from 2 ug to
0.06 ug were prepared in 100 ul final volume. To each sample 100 ul of
NaI solution (5 g/2 ml hot water) was added, the samples were incubated
at 95°C for 20 min., and applied directly to nitrocellulose paper which
had been soaked in 6X SSC using a dot blot apparatus. The filter was
washed three times for 5 min in water, 3 X 5 min in 70% Ethanol and 10
min in acetic anhydride solution (0.25 ml/100 ml 0.1 M triethanolamine).
The filter was prehybridized for 1 hr in 6X SSC/5X Denhardts and
hybridized 14 hr in 6X SSC/5X Denhardts containing 2x 106 cpm/ml of
nick translated pBl/l6 plasmid. A low stringency wash (1X SSC/0.1% SDS
at 65°C for 1 hr.) was performed first and the filter was exposed to
film for 2 hr and 11 hr. Then a high stringency wash (0.4X SSC/0.1% SDS
at 65°C for 1 hr) was performed and the filter was reexposed to film for

2 hr.

100

Sequencing.

pB1/16 plasmid DNA (10 ug) was digested for 3 hrs. using 20
units Msp I. The resulting fragments were end labeled by adding 40 uCi
dCT32P (Amersham 3000 Ci/mMole) and 2 units Klenow polymerase to the
restriction reaction and allowing the reaction to proceed 15 min at room
temperature. 2 nmoles of each dNTP was added and the reaction was
continued for 5 min. The labeled fragments were ethanol precipitated,
digested for 3 hrs with 20 units of PstI and resolved on a 1.2% agarose
gel in TAE buffer. An 840 bp fragment corresponding to the large PstI
to MspI portion of the insert was eluted according to the method of
Cirvitz (Maniatis et al., 1982).

Chemical sequencing was performed according to Maxam and Gilbert

(1980).

Northern analysis.

RNA was prepared from the SAS cell line (described in part I)
and MDBK cells (Madin-Darby bovine kidney). Total Poly A+ RNA was
prepared by lysing newly confluent 100 mm dishes in 2 m1 binding buffer
(0.5M NaCl, 10 mM Tris pH 7.5, 1 mm EDTA, 0.2% SDS) plus 200 ug/ml
Proteinase K. Viscosity was reduced by aspirating the lysate several
times thru a syringe fitted with an 18 gauge needle and the samples were
digested at 37°C for 45 min. Oligo dT cellulose (BRL) was added at a
final concentration of 5 mg/ml and the suspension was agitated at room
temperature for 1 hr. The oligo dT cellulose was pelleted by
centrifugation, resuspended in binding buffer, transferred to a quick
sep column (Isolabs) and washed 1 time with binding buffer, followed by

one wash with binding buffer containing 0.1 M NaCl. Poly A+ RNA was

101

eluted with two washes of 10 mM Tris pH 7.6/1 mM EDTA. RNA was
precipitated with ethanol in the presence of carrier tRNA and
resuspended in 50 ul northern sample buffer per 100 mm dish of cells.

Total RNA was prepared from these cell lines by lysis with
Quanidinium isothiocyanate and pelleting in CsCl (Maniatis et al., 1982,
Chirgwin et al., 1979,).

Gel analysis was performed on a 1.1% agarose gel as described in
part 1 of this thesis. RNA was transferred to nitrocellulose, the
filter was baked, prehybridized in 2.5X Denhardts, 3X SSC, 50 mM NaPOA,
50 % formamide, 50 ug/ml salmon sperm DNA at 42°C and hybridized for 20
hr in the same buffer containing 5% dextran sulfate, 0.1% SDS and 106

cpm/ml nick translated pBl/l6. The filter was washed at 65°C with 2X

SSC/0.5% SDS.

Cell Cultures.
Cell line 5A5 was derived by SV40 transformation of bovine
primary pituitary cells and was the gift of Dr. Robert Lyons. MDBK

cells are nontransformed cells derived from bovine kidney.

102

RESULTS

 

Southern blot analysis and mapping of pBlzl6

Genomic DNA prepared from cell line 5A5 was digested with
several restriction enzymes and analyzed by Southern blot analysis using
nick translated plasmid pB1/16 as a probe (Figure 1A). The results
indicated that pB1/16 contained a repetitive element which was highly
interspersed in the bovine genome.

Plasmid pBl/l6 was mapped using a variety of restriction enzymes
to determine the size and location of the repetitive sequence. Total
plasmid DNA or the purified PstI insert fragment was digested, resolved
on agarose gel and blotted. The blot was then probed with nick-
translated 5A5 genomic bovine DNA to determine which fragments contained
sequences which were highly represented in the bovine DNA (Figure 2A,
B). The resulting map (Figure 2C) revealed that there was one small
(<200 bp) repetitive sequence and that it was contained entirely within
the pBl/16 insert.

To determine the nature of the rest of pBl/16 the 700 bp Hinfl
fragment which did not contain the highly repetitive sequence was used
to probe a Southern blot of SAS DNA (Figure 18). The Hian fragment
apparently represents a single or low copy sequence in the bovine
genome.

Homology of the repetitive sequence to sequences in other

species was examined. DNA from a variety of organisms ranging from fish

103

Figure 1. Southern Blot analysis.

Ten micrograms of DNA prepared from 5A5 cells was digested
with EcoRI (lane R), BamHI (lane B) or HindIII (lane H), resolved on a
0.7% agarose gel and blotted. Panel A was probed with nick-translated
pBl/16 plasmid DNA (108Cpm/ug, 107 cpm in 5 ml 2X SSC/1X Denhardts).
Panel B was probed with nick-translated purified PstI-Hian fragment

from pBl/l6 insert DNA (see Figure 2).

104

 

B
R B H
-
9.5- "'
6.7-
44-
an
2.3-
2.0-
pBl/lb PsII-Hinfl

Figure 1

Figure 2. Map of pBl/l6.

Plasmid DNA (0.5 ug/lane) was digested with a variety of
restriction enzymes, resolved on a 1.5% agarose gel in TAE buffer,
stained with ethidium bromide (Panel A) and transferred to
nitrocellulose. (Panel B) The blot was probed with nick-translated
5A5 DNA (1.5 x 108 cpm/ug, 106 cpm/ml, 2X SSC/1X Denhardts, filters
were not blocked with carrier DNA). Enzymes used in this example were
1; BglI, 2; PvuII, 3; PstI, 4: PstI + PvuII, 5; BglI + PvuII, 6; BglI
+ PstI, 7; BglI + PstI + PvuII, 8; Ddel, 9; DdeI + PvuII, 10; DdeI +
BglI.

Bottom panel is restriction map of pBl/l6 insert and location
and size of the repetitive element (hatched region). Enzymes which do

not cut within the pBl/l6 insert are; AvaI, BamHI, ClaI, HaeIII,

HindIII, RsaI, Sau3A, StuI, Tan, XbaI, XhoI.

 

 

106

12345678910

   

 

 

 

 

 

 

 

PsII BgII MSPI PvuII PstI
I
I I I
200 600 1000
AIuI L l l J
Dde I l #1
Hini I L g P I
EcoRlI l I I—1 g4
Fnu4Hl L l J g
Hho I l——_1___l

Figure 2

107

to human was applied to nitrocellulose in a dot blot apparatus and
probed with nick-translated pB1/16 plasmid DNA (Figure 3). The filter
was washed first at low stringency (Figure 3A,B), exposed to film, and
rewashed at higher stringency (Figure 3C). Of the species examined, all
showed some homology at low stringency after a long exposure (Figure
3A). After short exposure times chicken was undetectable, but salmon
and rat showed some homology. A small amount of hybridization to rodent
DNA remained after the high stringency wash, while no hybridization to
fish or human DNA was detected. Hybridization to bovine cell line DNA's
was unchanged between high and low stringency washes. Similar results
were obtained using Eco RI digested DNA analysed by Southern blot, where
hybridization was performed in high and low stringency conditions (data

not shown).

Sequence of the repetitive element in DB1/l6

The repetitive region in pB1/16 was sequenced from the MspI site
at the right side of the element using end filling by Klenow polymerase
and Maxam-Gilbert sequencing. A sequencing gel is shown in Figure 4 and
the sequence is shown in Figure 5. The repetitive region of pBl/16 is
compared to the bovine repetitive A element consensus sequence
determined by Watanabe et a1. (1982) and Skowronski et a1. (1984). The
repetitive element in pB1/16 shows 4 Changes for this consensus sequence
(97% homology) and ends with an ACC repeat which has been shown to occur
on the 5' side of certain bovine and goat repetitive sequences (Watanabe
et a1. 1982, Skowronski et a1. 1984).

The homology with the A element ends at position 95 in Figure 5

and no homology to the 73 bp C element described by Schon et a1. is

108

Figure 3. Homology of bovine repetitive element with DNA from other
species.

DNA from the species shown in the figure was found to
nitrocellulose and hybridized as described in Materials and Methods.
DNA levels were Lane 1; 2 ug, lane 2; 0.35 ug, lane 3; 0.06 ug. The
blot was washed first with 1x SSC/0.1% SDS for 1 hr at 650C. Panel A
shows 11 hr. exposure to film, Panel B; 2hr. The blot was rewashed in
0.4X SSC/0.1% SDS for 1 hr. at 650C and exposed to film for 2 hr;

Panel C.

109

SoImon

Chicken

Hamster

Mouse

Rat

Bovine 5A5

5C6

Human

Figure 3

 

”n

110

Figure 4. Sequencing Cel of Bovine repetitive element.
pBl/16 plasmid was labeled with 32P-dCTP and prepared as
described in Materials and Methods. The chemical sequencing reactions

were resolved on a 32 x 37 cm 6% acrylamide-urea gel at 50 W. The gel

was dried and exposed to film for 15 hr. with an intensifing screen.

111

 

C .|
CA GAc GAAGCGACTTAGCAG c A

I...‘ ..~._ .. 2
LP... 3
It“. 2.5 2.. ..
E; .3... ._
‘85: . .

GGCG
GGGAGGGCTGTGCCGGGTGG
C

Elate ..
EH! ENE = a. :1
”u; a“

TC

7|

ATGG GGG

.3

T ThurcA C A .u AnuT;LG

.IGT
r»
Turk
_” .GA
..oe

Figure 4

112

Figure 5. Sequence of pBl/16 in region of repetitive element.
Consensus sequence of bovine element AR as compiled by
Skowronski et a1. is shown above pBl/16 sequence. Bases which differ

in the AL consensus sequence are shown above the AR sequence.
Restriction sites are indicated beneath the pBl/16 sequence. Homology

to other species has been ascribed to the region between 128 and 171

in this map (Skowronski et al., 1984).

 

113

A A C A T C T G T T C T G G N N G T C C G T G T T C C G C A T C T G C T

50

C T T T G T C A C A T C C C T C T C T C C G C T T A C A C G C T C T T T

HhaI

100

C C A G A A G G A A A T C C

G G C A C C A T T T C C A G T G A G T C C T G G A G A A C C A A A T C C

Hian EcoRII

C C A C T C C A G T A T T C T T C C C T C C A G A A T C C C A G
C C A C T C C A G T C T T C T T C C C T G C A G A A T C C C A C

C A A C
C A A -

 

Hian EcoRII

EcoRII

150

Fnu4H

 

 

C G A C C C C C C A C C C T G C T G C C C T C C C C T C T A T G C C C T
Q C A C C G G G C A G C C T G C T C G G C T C C C G T C T A T G C G G T

EcoRII BglI

200

C G C A C A C A G T C C G A C A C C A C T G A A G C G A C T T

T G C A C A G A C T C C C A C A C G A C T G A A G C G A C T T A G C A G

 

Fnu4H

DdeI

Hian

C A C C A G C A C C A G C C G C

Fnu4H Fnu4H Fnu4H Fnu4H

Figure 5

114

found. There is no evidence of a total or partial repeat of the A
element upstream of the 3' end and there are no CACTT sequences which
have been shown to link some bovine and caprine repetitive sequences

(Watanabe et al., 1982, Skowronski et al., 1984,).

Transcription of bovine repetitive sequences

 

Transcription of repetitive elements in other species has been
demonstrated under certain conditions of cell growth, differentiation,
or transformation. The cell line 5A5, which was derived by SV40
transformation of bovine pituitary cells, was used to study transcription
of the repetitive element found in pB1/16. A nontransformed cell line
derived from bovine kidney (MBDK) was also examined.

Total RNA was prepared from cell lines 5A5 and MDBK by the
method of Chirgwin et a1. (1979). Total Poly A+ RNA was also prepared
as described in materials and methods. Purified RNA was resolved
according to size by denaturing formaldehyde gel electrophoresis,
blotted and probed with nick-translated pB1/16 DNA (Figure 6). No
discrete species of RNA was found which contained homology to the
pB1/16 repetitive element, but a smear of high molecular weight RNA
was evident in all the RNA samples. This probably represents RNA
still carrying introns with repetitive sequences or mRNAs with
repetitive elements in the 3' untranslated region. Similar analysis
with cytoplasmic Poly A+ RNA also showed a smear of RNA but the size
range was slightly smaller (data not shown). The RNA from transformed
cell line 5A5 and the nontransformed MDBK cells had similar levels of

repetitive sequences.

115

Figure 6. Transcription of Bovine repetitive sequences.

Northern blot analysis of RNA from cell lines 5A5 and MDBK.
Lanes labeled A+; Poly A+ RNA prepared as described in material and
methods. Each lane contains half the RNA from a confluent 100 mm dish
of cells. Lanes labeled Total; 10 ug/lane of total RNA from 5A5 and
MDBK cells. Lane labeled Pit. polys. A+; Polysomal Poly A+ RNA from

pituitary glands (1 ug).

.185
_tRNA

+<I

 

(O

u +< .m>_oo_ In

 

s<hohn

Figure 6

117

Repetitive sequences in some other species have been shown to be
homologous to RNA polymerase III transcripts, producing RNA species in
the 100 to 600 bp range. There is no evidence for small transcripts in

the northern blot analysis presented here.

118

DISCUSSION

Previous analyses of the predominant moderately repetitive DNA
sequences in ruminants have led to a number of generalizations
regarding their nature. The repetitive elements in genomic DNA have
been found in two types of arrangements. Dimeric sequences composed
of two related 117 bp A monomers have been observed in five cases in
cows and two in goats (Figure 7). In all instances the right and left
monomer differed reproducibly from each other at certain positions
(Skowronski et al., 1984) (Figure 5). In addition the left and right
halves of the dimer are separated by a short repeating unit consisting
of the sequence (CACTTT)3CATCCATT. A second type of element is the A
unit preceded by the 73 bp C sequence (Figure 7). To date, only one
example of an isolated A unit has been described (Watanabe et al.,
1982) and it has been suggested that this sequence arose by
recombination between right and left monomers in a dimer unit
(Skowronski et al., 1984). This conclusion was based on the
observation that a portion of the sequence fit the consensus for a
left monomer while the right portion fits the consensus for the right
monomer.

The repetitive element described in this paper is 97.4%
homologous with the right monomer and contains only right monomer
bases at the sites where systematic differences have been reported

(Figure 5). The repetitive element in pBl/16 is terminated at the 3’

 

119

Figure 7. Organization of Bovine and Caprine repetitive elements.
Prototype elements are given in the top line. AR; A element,

right half; sequence same as that shown in Figure 5. A A element,

L;
left half, differences from AR are shown in Figure 5. AR and AL are
linked by the sequence (CACTTT)3CATCCATT in all cases except the goat
BF which is (CACTTT)2CATCCATT. C; 73 bp C element; concensus sequence
is : TCCCTCACACTCTCCCTGCAGTCCCCCACACCTCCCTTCNATCCCTCCCTTCCCAACATCCCCT.
Data were obtained from the following sources. Coat globin;
Spence et al., 1985, Schon et al., 1981, Shapiro et al., 1983. Bovine
globin; Schimenti and Duncan, 1984. Bovine satellite; Skowronski et

al., 1984. Bovine repetitive; Richardson et al., 1986. Bovine

corticotropine-b-lipotropin precursor introns; Wantanabe et al., 1982.

.120

 

 

 

U

 

 

 

— U

 

 

 

 

 

_ pen a

 

 

 

 

w

cflao»uogwAImucHaouuooHuuou mzH>om

SSC FIVOIIII g: \\\\_
sfiuo<vnnnnnHHU\
ofioo<vnwu. _. ..m1.
pea Sufi—II EN NNKNN ..... ETLCGSF a. . 0F
. £85 a a. . .D
mow :NNNNNIEQ ESSSS HUS

 

Nagoficflq E

 

Amman
m>~e~emamz mzH>om

mm

 

 

 

 

Ema ~m\fim
ma~game<m mz~>om

 

 

 

 

.14 r

 

 

u

sz040 mzH>om

 

 

JIIE >

 

 

 

 

 

 

 

 

 

 

 

u m

m
ou\mm
u

mszOAO h<ou

 

 

 

«a mfiuo<vnww I... .nlI
ofioo<cmw. .4. ..~
0508? ANN N NE
an NACG<CHI IIm.. .nl
m :4 Hour
mafiaeeo<ov mo NACG<CHM u.. .nI
«a
an 69: F EEK NE an
an =AGG<VHHH. u GQNNNNKNNMGmo mo cAou<Cplr II_..._I
< o

mm>HOHomm

121

end with a repeating (ACC)6 associated with right monomers. There is
no evidence of another repetitive sequence to the left of the element
or of the typical CACTTT repeat found in the spacer region of the
dimer. The region to the left of the repetitive element also shows no
homology to the 73bp tRNA-like repeat. The repetitive element in
pBl/16 is apparently an example of a third possible arrangement for
repetitive elements in the bovine genome. Namely, an A unit
unaccompanied by any other repetitive unit.

Repetitive elements can be transcribed by either of two RNA
polymerases. Transcription by RNA Pol II can occur when the element
is contained within the Pol II transcription unit. Alternatively
repetitive elements can be transcribed from internal promoters by RNA
Pol III. In vitro transcription of repetitive elements by Pol III has
been demonstrated for a wide variety of species using either cloned
DNA segments or total DNA (Matsumoto et al., 1986, Furth and Su, 1986,
Kramerov et al., 1982, Duncan et al., 1979). In vivo Pol III
transcription has been most Clearly demonstrated in rodent cells. An
RNA corresponding to the B1 sequence is transcribed when mouse cells
are transformed by SV40 (Murphy et al., 1983, Singh et al., 1985). A
species of cytoplasmic RNA (BC) homologous to the rat ID repetitive
element is transcribed specifically in rat brain and pituitary cells
(McKinnon et al., 1987). An alu repetitive sequence near the e-globin
gene has been shown to produce Pol III transcripts only in the nuclei
of a cell line which expressed e-globin (Allan and Paul, 1984).

Clearly, repetitive elements can be transcribed and their
transcription is regulated in some way by the cell so that the RNA is

expressed only under certain conditions. Bovine repetitive sequences

 

122

can be transcribed by Pol III in vitro (Furth and Su, 1986) to yield
two species of RNA, a 120 bp sequence and a heterologous length length
fraction 180-650 bp. Both showed homology to the A repetitive unit.

In vivo transcription of these sequences was not tested by Furth and

 

Su.

In the present study in vivo transcription of bovine
repetitive elements was examined. Both a normal (MDBK) and
transformed cell line were tested. RNA sequences of high molecular
weight were found to contain repetitive elements (Figure 6) presumably
as part of Pol II transcription products. No difference in the level
of expression of HMW RNA species harboring repetitive sequences was
observed between normal kidney and transformed pituitary cells. I
found no evidence for a unique Pol III transcription product in bovine
pituitary tissue, transformed pituitary cells or normal cells.
Expression of the A sequence in cattle shows no analogy to expression

of ID or B2 repetitive elements in rodents.

123

REFERENCES

Ackerman, E.J. 1983. EMBO 2:1417-1422.

Allan, M., and Paul, J. 1984. NuCl. Acids. Res. 12:1193-1200.
Britten, R.J., and Davidson, E.M. 1969. Science 165:349-358.
Britten, R.J., and Kohn, D.E. 1968. Science 161:529-540.

Brunner, A.M., Schimenti, J.C., and Duncan, C.H. 1986. Biochem.
25:5028-5036.

Bennett, 1984.
Daniels, C.R., and Deininger, P.L. 1985. Nature 317:819-822.
Clemens, M.J. 1987. Cell 49:157-158.

Chirgwin, J.M., Przybyla, A.E., MacDonald, R.F.J., and Rutter, W.J.
1979. Biochemistry 18:5294-5299.

Doolittle, W., and Sapienza, C. 1980. Nature 284:601-603.

Duncan, C., Biro, P.A., Choudary, P.V., Elder, J.T., Wang, R.R.C.,
Forget, B.C., DeRiel, J.K., Weissman, S.M. 1979. Proc. Natl.
Acad. Sci. USA 76:5095-5099.

Duncan, C.H., Jagadeeswaran, P., Wang, R.R.C., and Weissmann, S.M.
1981. Gene 13:185-196.

Edwards, D.R., Parfett, C.L.J., and Denhardt, D.T. 1985. Mol. Cell.
Biol. 5:3280-3288.

Endoh, H., and Okada, N. 1986. Proc. Natl. Acad. Sci. USA 83 251-
255.

Federoff, N.B., Wellauer, P.R., and Wall, R. 1977. Cell 10:597-610.
Fowlkes, D.M., and Shenk, T. 1980. Cell 22:405-413.
Fritsch, E.F., Lawn, R.M., and Maniatis, T. 1980. Cell 19:959-972.

Furth, J.J. and Su C-Y. 1986. Biochem. J. 237:827-835.

124

Haynes, S.R., and Jelinek, W.R. 1981. Proc. Natl. Acad. Sci. USA
78:6130-6134.

Hess, J., Fox, M., Schmid, C., and Shen, C-K.J. 1983. Proc. Natl.
Acad. Sci. USA 80:5970-5974.

Houck, C.M., Rinehart, F.P., and Schmid, C.W. 1979. J. Mol. Biol.
132:289-306.

Jelinek, W.R., Evans, R., Wilson, M., Salditt-Ceorgieff, M., Darnell,
J.E. 1978. Biochemistry 17:2776-2783.

Jelinek, W.R., and Schmid, C.W. 1982. Ann. Rev. Biochem. 51:813-844.

King, D., Snider, L.D., and Lingrel, J.B. 1986. M01. Cell. Biol.
6:209-217.

Kramerov, D.A., Ryskov, A P., and Ceorgiev, C.P. 1977. Biochem.
Biophys. Acta 475:461-475.

Kramerov, D.A., Crigoryan, A.A., Ryskov, A.P., and Ceorgiev, C.P.
1979. Nucl. Acids Res. 6:697-713.

Kramerov, D.A., Lekakh, I.V., Samarina, O.P. and Ryskov, A.P. 1982.
Nucl. Acids Res. 10:7477-7491.

Krayev, A.S., Markusheva, T.V., Kramerov, D.A., Ryskov, A.P., Skryabin,
K.C., Bayev, A.A., and Ceorgiev, C.P. 1982. Nucl. Acids Res.
10:7461-7474.

Laimins, L., Holmgren-Konig, M., and Khoury,C. 1986. Proc. Natl.
Acad. Sci. 83:3151-3155.

Laird, C.D. 1971. Chromosoma 32:378-406.

Lawrence, C.B., McDonnell, D.P., and Ramsey, W.J. 1985. Nucl. Acids
Res. 13:4239-4251.

Lueders, K.K., Fewell, J.W., Kuff, E.L., and Koch, T. 1984. Mol.
Cell. Biol. 4:2128-2135.

McKinnon, R.D., Danielson, P., Brow, M.A.D., Bloom, F.E., and
Sutcliffe, J.C. 1987. Mol. Cell. Biol. 7:2148-2154.

Malissen, B., Malissen, B., and Jordan, B.R. 1982. Proc. Natl. Acad.
Sci. USA 79:893-897.

Maniatis, T., Fritsch, E.F., and Sambrook, J. 1982. "Molecular
Cloning-A Laboratory Manual". Cold Spring Harbor Laboratory.
New York.

125

Matsumoto, K-I., Murakami, K., and Okada, N. 1986. Proc. Natl. Acad.
Sci. USA 83:3156-3160.

Matsumoto, K-I., Murakami, 1., and Okada, N. 1984. Biochem. Biophys.
Res. Comm. 124:514-522.

Maxam, A.M. and Gilbert, W. 1980. Methods Enzymol. 65:499.

Milner, R.J., Bloom, F.E., Lai, C., Lerner, R.A., and Sutcliffe, J.C.
1984. Proc. Natl. Acad. Sci. USA 81:713-717.

Murphey, D., Brickell, P.M., Latchman, D.S., Willison, K., and Rigby,
P.W.J. 1983. Cell 35:865-871.

Owens, C.P., Chaudhari, N., and Hahn, W.E. 1985. Science 229:1263-
1265.

Orgel, L., and Crick, F.H.C. 1980. Nature, 284:604-607.

Richardson, K.K., Crosby, R.M., Good, P.J., Rosen, N.L., and Mayfield,
J.E. 1986. Eur. J. Biochem. 154:349-354.

Rogers, J. 1985. Int. Rev. Ctyol. 93:187-279.

Ryskov, A.P., Ivanov, P.L., Kramerov, D.A., and Ceorgiev, C.P. 1983.
Nucl. Acids Res. 11:6541-6558.

Sakamoto, K., and Okada, N. 1985. J. Mol. Evol. 22:134-150.
Sapienza, C., and St-Jacques, B. 1986. Nature 319:418-420.
Schimenti, J., and Duncan, G. 1984. Nucl. Acids Res. 12:1641-1655.

Schon, B.A., Cleary, M.L., Haynes, J.R., and Lingrel, J.B. 1981. Cell
27:359-369.

Schmid, C.W., and Shen, C-K.J. 1986. In "Molecular Evolutionary
Genetics" (MacIntyre, R.J., ed.) p323-354. Plenum. New York.

Schoeniger, L.O., and Jelinek, W.R. 1986. Mol. Cell. Biol. 6:1508-
1519.

Scott, M.R.D., Westphal, K-H., and Rigby, P.W.J. 1983. Cell 34 557—
567.

Shapiro, S.C., Schon, E.A., Townes, T.M., and Lingrel, J.B. 1983. J.
Mol. Biol. 169:31-52.

Shaw, C., and Kamen, R. 1986. Cell 46:659-667.

126

Singer, M.F. 1982. Cell 28:433-434.

Singer, M.C. and Skowronski, J1 1985. Trends Biochem. Sci. 10:119-
122.

Singh, K., Carey, M., Saragosti, S., and Botchan, M. 1985. Nature
314:553-556.

Skowronski, J., Plucienniczak, A., Bednarek, A., and Jaworski, J.
1984. J. Mol. Biol. 177:399-416.

Spence, S.E., Young, R.M., Garner, K.J. and Lingrel, J.B. 1985.
Nucl. Acids Res. 13:2171-2186.

Sutcliffe, J.C., Milner, R.J., Bloom, F.E., and Lerner, R.A. 1982.
Proc. Natl. Acad. Sci. USA 79:4942-4946.

Sutcliffe, J.C., Milner, R.J., Cattesfeld, J.M., and Lerner, R.A.
1984a. Nature 308:237-241.

Sutcliffe, J.C., Milner, R.J., Cattesfeld, J.M., and Reynold, W.
1984b. Science 225:1308-1315.

Watanabe, Y., Tsukada,_T., Matake, M., Nakanishi, S., and Numa, S.
1982. Nucl. Acids Res. 10:1459-1469.

Weiner, A.M., Deininger, P.L., and Efstratiadis, A. 1986. Ann. Rev.
Biochem. 55:631-661.

 

 

 

flywj‘.’y,’]),__ee4n w... .- .47- .4 “2“" - at- ..... .
.55 '

   

   

 

->\v>r\~-
”mun
\

 

 

 

s...“
..-..\.w._._"~.t~...-....
H l- | MN...

 

 

 

.....~. vu...
'.......-A'....

t...” .1.
“Human-- ,..

. ....‘.c...
--r ‘M‘\\~~1

 
 
 

 

   

 

   
  
    
         
     

nIcHIanS TE UNIV LIBRARIES

III I III

293 0537

   

 

 

 

 

 

 

M ... .... .
... . -~ . ..
'.. .. _ I
.. .v. 7

I
I