• Aucun résultat trouvé

Frequency distribution of mRNA and pre-mRNA in growing and differentiated Friend cells

N/A
N/A
Protected

Academic year: 2022

Partager "Frequency distribution of mRNA and pre-mRNA in growing and differentiated Friend cells"

Copied!
21
0
0

Texte intégral

(1)

Article

Reference

Frequency distribution of mRNA and pre-mRNA in growing and differentiated Friend cells

MAURON, Alex, SPOHR, Georges

Abstract

The frequency distribution of poly(A)+-mRNA in growing and in differentiated Friend cells has been measured by mRNA-cDNA hybridization and their differences established by heterologous hybridization of mRNA of one type and cDNA of the other. It was shown that induction of Friend cells involves an increase in abundance of a small number of mRNAs, while no specific pattern of messenger disappearance could be detected. The frequency distribution of pre-mRNA was determined by hybridizing nuclear RNA with the cDNA probes complementary to mRNA. In uninduced Friend cells, it was shown that most precursor messenger sequences are present at a single frequency of about 3 molecules per nucleus, independently of their final frequency in polysomal mRNA. In induced Friend cells, the frequency distribution of pre-mRNA is more heterogeneous and correlated to some extent with the corresponding mRNA frequency distribution.

MAURON, Alex, SPOHR, Georges. Frequency distribution of mRNA and pre-mRNA in growing and differentiated Friend cells. Nucleic Acids Research , 1978, vol. 5, no. 8, p. 3013-3032

PMID : 278967

Available at:

http://archive-ouverte.unige.ch/unige:85373

Disclaimer: layout of this document may differ from the published version.

1 / 1

(2)

Frequency distribution of mRNA and pre-mRNA in growing and differentiated Friend cells

Alex Mauron and Georges Spohr+

Swiss Institute for Experimental Cancer Research, Ch. des Boveresses, 1066 Epalinges, Switzerland

Received 2 May 1978

ABSTRACT

The frequency d i s t r i b u t i o n of poly(A) -mRNA in growing and in d i f f e r e n - t i a t e d Friend cells has been measured by mRNA-cDNA hybridization and t h e i r differences established by heterologous hybridization of mRNA of one type and cDNA of the other. I t was shown that induction of Friend c e l l s involves an increase in abundance of a small number of mRNAs, while no s p e c i f i c pat- tern of messenger disappearance could be detected. The frequency d i s t r i b u - t i o n of pre-mRNA was determined by hybridizing nuclear RNA with the cDNA probes complementary to mRNA. In uninduced Friend c e l l s , i t was shown that most precursor messenger sequences are present at a single frequency of about 3 molecules per nucleus, independently of t h e i r f i n a l frequency in po- lysomal mRNA. In induced Friend c e l l s , the frequency d i s t r i b u t i o n of pre- mRNA is more heterogeneous and correlated to some extent with the correspon- ding mRNA frequency d i s t r i b u t i o n .

INTRODUCTION

The control of gene expression i n eukaryotic c e l l s s t i l l defies a general explanation, p a r t l y because of the m u l t i p l i c i t y of levels a t which molecular regulation devices may operate. In order to appreciate the impor- tance of these various levels of c o n t r o l , i t would be important to know which sets of genes are transcribed i n t o nuclear RNA and i n t o messenger RNA in a given c e l l type.

Many studies have recently been performed i n which the d i v e r s i t y of the sequences in mRNA and i n nuclear RNA of d i f f e r e n t cells has been measu- red. The results of such measurements are expressed in terms of the sequence complexity of an RNA population, i . e . , the t o t a l number of RNA species pre- sent. For instance, the sequence complexity of mRNA from sea urchin gastrula has been determined by hybridizing radioactive non-repetitive DNA with an excess of mRNA ( 1 ) . At s a t u r a t i o n , the proportion of DNA taken i n t o hybrid form i s a measure of the complexity of mRNA. On the other hand, the k i n e t i c s of t h i s hybridization i s also related to the complexity. From both these

3 0 1 3

at Institut universitaire de hautes etudesinternationales - Bibliotheque on June 14, 2016http://nar.oxfordjournals.org/Downloaded from

(3)

Nucleic Acids Research

data taken together, i t was possible to conclude that different mRNA spe- cies occur at widely different frequencies. Another technique for determi- ning sequence complexity is best suited to study the frequency distribu- tion of mRNA populations : polyadenylated mRNA is used as a template in the synthesis of a cDNA probe which is back-hybridized with i t s template.

The kinetics of this RNA-excess hybridization is related to the sequence complexity of the mKNA. I t usually shows a pattern of multiple transitions which suggests that the frequencies of individual messenger species are clustered around 2 to 4 discrete values, rather than taking any value in a broad, continuous range. In other words, the quantitative structure of an mRNA population could be approximately described by dividing i t into 2 to 4 abundance classes (2-4).

In cell differentiation, control mechanisms are thought to operate in order to express the genes necessary for the performance of specialized cell functions. Thus, the sets of sequences found in mRNAs of various cell types presumably differ to some extent. The hybridization techniques just described allow us to observe the effects of cell differentiation at the mRNA level. Comparisons between the mRNA populations of different cell types of various origins have been made by heterologous mRNA-non-repeti- tive-DNA or mRNA-cDNA hybridization (5-13). A more or less extensive over- lap between the mRNA of different tissues is usually found, with some mes- senger sequences being specific for a certain cell type. In addition, sequences common to different cell types may differ in abundance, again stressing the importance of the quantitative, as well as the qualitative composition of mRNA populations.

The sequence complexity of nuclear RNA has also been measured (sea urchin (14), drosophila cells (15), rat tissues (16), mouse brain (17), uninduced Friend cells (18, 19). I t has usually been found to be 5-10 times higher than the complexity of polysomal mRNA.

We have chosen to study the frequency distribution of mRNA and of pre-mRNA of Friend cells, an established line of mouse erythroleukemic cells transformed by an RNA tumor virus (20). These cells resemble the proerythroblast stage of normal red blood cell differentiation but upon addition of a variety of inducing agents, in particular dimethyisulfoxide (DMSO), they experience a differentiation process whose biochemical para- meters have been extensively characterized. In the presence of the indu-

3014

at Institut universitaire de hautes etudesinternationales - Bibliotheque on June 14, 2016http://nar.oxfordjournals.org/Downloaded from

(4)

cing agent, there i s an accumulation of many proteins s p e c i f i c f o r t e r m i - n a l l y d i f f e r e n t i a t e d e r y t h r o c y t e s , such as oCand (5 giobin ( 2 1 ) , s p e c t r i n ( 2 2 ) , acetylcholinesterase ( 2 3 ) , enzymes involved i n the biosynthesis of heme ( 2 4 ) , e t c . In the case of uninduced Friend c e l l s , the sequence com- p l e x i t y of polyadenylated mRNA had already been determined by B i r n i e e t a l . ( 3 ) . These authors showed t h a t the mRNA contains about 104 d i f f e r e n t sequences divided i n t o 3 abundance classes. Our f i r s t aim was t o measure the sequence d i s t r i b u t i o n of polyadenylated mRNA before and a f t e r d i f f e - r e n t i a t i o n and t o determine the nature and extent of the differences bet- ween the two stages. This was achieved by synthesizing the cDNA probes complementary t o the poly(A)+-mRNA of both stages and performing the homo- logous and heterologous hybridizations w i t h both mRNAs.

In a d d i t i o n , we wanted to gain some i n s i g h t i n t o the o r i g i n of the q u a n t i t a t i v e s t r u c t u r e of mRNA populations, i . e . t h e i r d i s t r i b u t i o n in discrete frequency classes. We asked the question whether the subclass of nuclear RNA containing the messenger sequences t o be expressed as poly- somal mRNA i s organized i n the same q u a n t i t a t i v e s t r u c t u r e as polysomal mRNA. In other words, we wanted t o know whether a sequence which i s abun- dant i n polysomal mRNA i s already abundant in t h i s subclass (pre-mRNA).

We had observed t h a t i n Friend c e l l l i n e F4N, according to the d i f f e r e n - t i a t i o n stage, there i s a c o r r e l a t i o n between the concentration of the globin sequence i n nuclear RNA and the abundance of globin mRNA in the cytoplasm ( t h i s i s d i f f e r e n t i n other c e l l l i n e s ) . In order to determine whether t h i s i s a general property of polyadenylated mRNAs, we hybridized t o t a l nuclear RNA of both types of Friend c e l l s w i t h the cDNA probes com- plementary t o mRNA. From the k i n e t i c s of h y b r i d i z a t i o n , we could e s t a b l i s h the frequency d i s t r i b u t i o n of pre-mRNA in growing and i n d i f f e r e n t i a t e d Friend c e l l s .

MATERIALS AND METHODS 1. Cell c u l t u r e

Friend c e l l s ( l i n e F4N i s o l a t e d by Ostertag and obtained from Dr. H.

Eisen) were grown i n suspension i n Dulbecco's modified Eagle's medium sup- plemented w i t h 15% f e t a l c a l f serum (Gibco) and a doubled concentration of glutamin. The c e l l s were induced by adding 1,25% DMSO and d i l u t i n g every day with DMSO-containing medium.

3015

at Institut universitaire de hautes etudesinternationales - Bibliotheque on June 14, 2016http://nar.oxfordjournals.org/Downloaded from

(5)

Nucleic Acids Research

2. Cell f r a c t i o n a t i o n

A l l steps were carried out at 4 C. Cells were collected by c e n t r i f u - gation at 280 g. The p e l l e t was resuspended in 10 v o l . of Earle's saline and recentrifuged. Washed c e l l s were taken up i n 8 v o l . of l y s i s buffer (10 mM Triethanolamine-HCl pH 7.4, 10 mM NaCl, 1 mM MgCl2> 1 mM Mn C l2) . A f t e r 3 m i n . , i s o t o n i c i t y was restored by adding sucrose (0.25 M). 0.1%

Triton-X-100 was added and the cells lysed with a dounce homogenizer. The lysate was layered on a 10 ml. cushion containing 1M sucrose i n l y s i s b u f f e r adjusted to pH 6.9 and centrifuged for 10 min. at 630 g. The upper layer containing the cytoplasmic material was centrifuged f o r 20 min. at 10000 g. and the supernatant further processed to obtain polysomes. The i n t e r f a c e was discarded and the nuclear p e l l e t was resuspended in 0.25 M sucrose pH 6.9 l y s i s buffer containing 0.1% Triton-X-100. Nuclei were centrifuged f o r 5 min. at 630 g . , the p e l l e t resuspended as before and kept over l i q u i d nitrogen u n t i l further processed.

3. Preparation of Friend c e l l mRNA

The cytoplasmic material was layered on a 10 ml. cushion containing 15% sucrose in l y s i s buffer pH 7.4 and centrifuged in a f i x e d angle rotor (Beckman Ti 60) for 60 min. at 60000 RPM and 4°C. The polysomal p e l l e t was resuspended in 0.25 M sucrose in lysis buffer pH 7.4 and extracted with phenol-chloroform according to Perry e t a l . (25). Total polysomal RNA was subjected t o o l i g o (dT)-cellulose chromatography (26). The RNA was redissolved i n h i g h - s a l t buffer (10 mM Tris-HCl pH 7.5, 0.5 M NaCl, 1 mM EDTA, 0.2%

SDS) and passed three times over an oligo(dT)-cellulose column at room temperature. A f t e r r i n s i n g with 3 column volumes of b u f f e r , the poiy- adenylated RNA was recovered by e l u t i o n with prewarmed low-salt buffer (without NaCl). A f t e r heat denaturation of the eluate (2 min. a t 100°C), the chromatography was repeated and the second poly(A) pool was used in cDNA synthesis and h y b r i d i z a t i o n s .

4. Preparation of Friend c e l l nuclear RNA

The nuclear f r a c t i o n (see section 2) was extracted with hot phenol as described by Scherrer (27). The recovered aqueous phase was precipitated twice with ethanol and treated with DNase ( l O ^ g / m l RNase-free grade, Wor- thington) i n the presence of 2 mM MnCU. The digestion (20 min. at 4 C) was stopped by addition of EDTA (10 mM), SDS (0.2%), and adjusted t o 0,15 M NaCl. The RNA was then subjected to exclusion chromatography on Bio-Gel

3016

at Institut universitaire de hautes etudesinternationales - Bibliotheque on June 14, 2016http://nar.oxfordjournals.org/Downloaded from

(6)

A5m (Bio-Rad) equilibrated with 20 mM Tris-HCl pH 7.5, 0.15 M NaCl, 0.2%

SDS, 1 mM EDTA, 5 U-USP/ml heparin.

5. Preparation of globin mRNA

Globin mRNA was prepared from anemic mice as described by Williamson (28) using phenol-chloroform extraction and oligo(dT) cellulose chromato- graphy.

6. Synthesis of cDNA

cDNA was synthesized with reverse transcriptase as described by Imaizumi et al. (29).

7. Hybridization

RNA-DNA hybridizations were performed as described by Imaizumi et a l . (29) with s l i g h t modifications. The hybridization mixture contained the RNA sample, 0.3 M NaCl, 50 mM Tris-HCl pH 7.5, 0.1 mg/ml of E.Coli 4S RNA, 0.1% SDS and 200-300 cpm of cDNA. The mixture was sealed i n microcapillary pipettes i n portions of 10 yul, incubated at 65°C for various times and flushed i n t o 100^1 of SI digestion mixture (30 mM sodium acetate pH 4 . 5 , 25 mM NaCl, 1 mM ZnSO., 6% g l y c e r o l , 6 yug/ml of heat-denatured salmon sperm DNA and 5 units/ml of SI nuclease prepared according to Vogt ( 3 0 ) ) . The incubation was carried out at 45 C f o r 40 min. The proportion of cDNA taken i n hybrid form was calculated using undigested standards and subtrac- t i n g the zero-time value at the lowest RNA concentrate from every time- point (5-7% of t o t a l ) .

8. Data reduction

An RNA-cDNA hybridization i n RNA excess follows f i r s t - o r d e r k i n e t i c s , the pi

given by :

thus the proportion of DNA taken i n t o hybrid form as a function of R t i s

(1) H = l - e 'k Rot

where R is the concentration of RNA i n mole of nucleotide per l i t e r and t the hybridization time i n seconds. The rate constant K is related t o the sequence complexity of RNA and can be computed from the k i n e t i c curve by taking the R t value corresponding to H = 0.5 : Rgti/? = 1 n 2/k.

I f the RNA consists of various sequences clustered i n t o N frequency classes, i . e . i f the frequency of every mRNA type i s one of N values, i t is easy t o show that eq. 1 becomes :

(2) H = 1 - ) pne"knpnRot (Bishop e t a l . , 2) n-1

at Institut universitaire de hautes etudesinternationales - Bibliotheque on June 14, 2016http://nar.oxfordjournals.org/Downloaded from

(7)

Nucleic Acids Research

where p is the r e l a t i v e mass of RNA comprising the mRNA types of the n-th frequency class and k the corresponding rate constant. We assume t h a t the same values p apply to the cDNA. This assumption is sound, as shown by Hastie and Bishop ( 7 ) . The f i t t i n g of eq. (2) to the experimen- t a l data was performed using a non-linear least squares routine on a CDC CYBER 7326 system. Having been provided with the data, the mathematical model and a choice of N, t h i s program computes the values of p and k giving the best f i t and the relevant standard errors and confidence i n t e r - v a l s .

RESULTS

Frequency d i s t r i b u t i o n of uninduced c e l l poly(A) -mRNA

We measured the sequence complexity of poly(A) -mRNA of uninduced and f u l l y induced (4 days i n ~\ .25% DMSO) Friend c e l l s . Total polyriboso- mal poly(A)+-mRNA from uninduced cells was transcribed i n t o H-labeled

4

cDNA ( s p e c i f i c a c t i v i t y about 10 cpm/ng). Fig. 1 shows the hybridization curve of mRNA from uninduced cells with the cDNA transcribed from i t : an excess of mRNA was incubated with cDNA f o r various time-periods and the f r a c - t i o n of cDNA i n hybrid form determined as the proportion of DNA resistant to

i

o u. 1.0- o

£

£ 0.8- 2I

< 0.6- Za

v

r~X10.4-

r 0 2-

ccu. -2 -1 0 1 2 3

log Rot for mRNA

0 1 2 3 4

log Rot for polysomal RNA Fig. 1

Complexity of uninduced c e l l mRNA. Poly(A) -mRNA was p u r i f i e d from unindu- ced Friend c e l l s and hybridized with i t s own cDNA at RNA concentrations of 110 t o 150/jg/ml. The data were f i t t e d t o a two-component model. H y b r i d i - zation with globin cDNA i s also included but since in the l a t t e r case t o - t a l polysomal RNA was used instead of p u r i f i e d mRNA, the RQt scale is s h i f t e d t o account f o r the d i l u t i o n factor of 50 (o o) Hybridization of uninduced c e l l mRNA with the corresponding cDNA. (o o) Hybridiza- t i o n of uninduced c e l l polysomal RNA with globin cDNA.

3018

at Institut universitaire de hautes etudesinternationales - Bibliotheque on June 14, 2016http://nar.oxfordjournals.org/Downloaded from

(8)

the single-stranded specific nuclease SI. In order to make a comparison on the same graph with a hybridization curve obtained with a single kinetic component we show on the same figure the hybridization kinetics of total polysomal RNA (not purified poly(A) -mRNA) of uninduced cells with globin cDNA. The abscissa of the latter curve has been shifted by a factor of 50, expressing the fact that poly(A) -mRNA represents approximately 1/50 of total polysomal RNA. The curve obtained by hybridizing uninduced cell mRNA with i t s cDNA shows two transitions, suggesting two abundance classes.

Therefore, the data have been reduced to a model involving two kinetic components using the non-linear least squares technique. The relevant nume- rical data are summarized in Table 1. In determining the fraction of each abundance class, the assumption was made that the cDNA remaining unhybri- dized at very high R t values does not belong to any particular abundance class. The rate constant for each class provided by the computer calcula- tion was converted to R

Q

t, using the relation : R t , = In 2/k. The value of R t , is proportional to the sequence complexity of the mRNA in a given kinetic component. The standard of complexity is provided by the hybridi- zation of mouse globin mRNA with its cDNA. Reticulocyte 9S globin mRNA is a mixture in equal amounts of the mRNAs coding for the o6and £> -chain.

The summed MW of these two messengers is 4x10 dalton and we obtain a -4 -1

R

o

t, value of 4x10 mole xsxl in our hybridization conditions. Conside- ring that an mRNA of number average size has a molecular weight of 600'000 dalton, the standard R t , value, i . e . the R t , corresponding to a comple- xity of 1 average mRNA, w i l l be

6 0 0

'

0 0

° x 4xlO"

4

= 6 x 10"

4

mole xsxl"

1

400'000

Our results show that poly(A)

+

-mRNA can be ordered into two frequency classes. The f i r s t , low complexity class contains about 160 frequent types of messengers. The other contains the rare messengers, which make up most

4

of the sequence diversity of mRNA which corresponds to 10 different spe- cies. As shown by Fig. 1 and by Table 1, in uninduced cells the globin mRNA is a quite typical member of the latter class.

Frequency distribution of induced cell poly (A) -mRNA

In order to determine the frequency distribution of mRNA in induced Friend cells, cDNA transcribed from poly(A)

+

-mRNA of induced cells was hy- bridized to i t s own template. As shown in Fig. 2, the hybridization takes

at Institut universitaire de hautes etudesinternationales - Bibliotheque on June 14, 2016http://nar.oxfordjournals.org/Downloaded from

(9)

TABLE 1

A Abundance class no.

1 2

B

Fraction of cDNA i n hybrid

0.414 0.536

Fraction of cDNA i n frequency class

0.44 0.56

R t , of pure component

0.098 5.9

V j

371

Sequence complexity (number of d i f f e r e n t mRNAs of average size)

163 9870

2proportion of globin sequence i n t o t a l poly(A)+-mRNA

5.4 x 10"5

Number of molecules of each mRNA per ceTl

1620 34

Number of globin mRNA molecules per c e l l

49

Numerical data from the hybridization cruves i n F i g . 1 . A. Hybridization of uninduced c e l l poly(A) -mRNA with i t s own cDNA.

1 . The number of molecules of each type i s calculated i n the following way : uninduced Friend c e l l s contain about 30 pg/cell of t o t a l RNA, which we take as a reasonable estimate of polysomal RNA. We estimate 2% of t h i s amount to be poly(A)+-mRNA, i . e . 30 pg x 0.02 = 0.6 pg or 0.6 x 1 0 " '2g . Since the number average molecular weight of these molecules i s 6 x 10^ g/mole, t h e i r t o t a l number w i l l be equal to (0.6 x 1 0 ' ^2 g x 6 x 102 3 mole"1) / 6 x 10^ g/mole = 600'000 where 6 x 102 3 i s Avogadro's number.

Of these 600'000 molecules, 264'000 belong to the f i r s t abundance class and 336'000 to the second. Dividing these figures by the corresponding complexities, one obtains the number of molecules of each sequence. B. Hybridization of t o t a l uninduced c e l l polyribosomal RNA with globin cDNA. 2. The proportion o f globin sequence in t o t a l polysomal RNA i s given by the r a t i o of Rot i f o r the globin mRNA-globin cDNA hybridization to the Roti observed with polysomal RNA : 4 x 10"q / 371 = 10"6. Since poly(A)+-mRNA i s It of polysomal RNA the corresponding f r a c t i o n of globin sequence i s 10"6 / 0.02 = 5.4 x 10"5. In the same way as i n A, the number of globin mRNA molecules i s given by 0.6 x 10~l z g x 5.4 x 10"5 x 6 x 102 3 mole"1) / 4 x 10= g/mole=49 introducing the molecular weight o f the<\ +/B globin mRNAs instead o f the number average molecular weight o f a l l mRNAs.

at Institut universitaire de hautes etudesinternationales - Bibliotheque on June 14, 2016http://nar.oxfordjournals.org/Downloaded from

(10)

log Rot for polysomal RNA

Fig. 2

Complexity of induced cell mRNA. Poly(A) -mRNA was purified from induced cells and hybridized with i t s own cDNA at mRNA concentrations of 30 to 180/jg/ml. The broad transition in the higher R

Q

t range comprises two first-order components which could not be accurately resolved. This has l i t t l e effect on the parameters for the most abundant class which covers 23-27% of the hybridization and has a complexity of 4-6 messenger sequen- ces, whatever assumptions are made for the niddle and high-complexity class. Since induced cells contain 15 pg of RNA, this corresponds to 13000-19000 molecules per species per c e l l . Hybridization with globin cDNA is also shown and the same comment as in Fig. 1 applies, (e •) Hybridization of induced cell mRNA with the corresponding cDNA.(o o) Hybridization of induced cell polysomal RNA with globin cDNA.

place over a larger range of R t values than in the uninduced situation.

Computer analysis of the data shows that, contrary to the previous case, the f i t of these data to a 2-component model is not satisfactory. In addi- tion, significant hybridization takes place at lower R t values than in Fig. 1, meaning that a few mRNA species must be present at a very high fre- quency. This is confirmed by the calculation, which shows that a three-com- ponent model is appropriate. The f i r s t of the three kinetic components, com- prising 23-27% of the messenger mass corresponds to only 4-6 different se- quences. These sequences are present at 13'000 to 19'000 molecules/cell.

The middle and high complexity class could not be resolved in a reliable way but comparing the data to those in Fig. 1 shows that they are not grossly different in the higher R t range. Thus, the total complexity does not seem to change significantly from the uninduced to the induced stage (this is further confirmed by the heterologous hybridizations described later). As in the previous experiment, induced-cell total polysomal RNA has been hybri- dized with globin cDNA and the same remarks apply for the scale-shift. The

3021

at Institut universitaire de hautes etudesinternationales - Bibliotheque on June 14, 2016http://nar.oxfordjournals.org/Downloaded from

(11)

Nucleic Acids Research

result shows that the globin mRNA is one of the 4-6 mRNAs in the highest frequency class. (Table 2)

Having established the quantitative structure of the mRNA population in both types of c e l l s , we wished to establish the differences between them. Messenger species present in one state could be totally absent in the other; furthermore, quantitative rearrangements could take place during induction such as frequent mRNAs becoming rarer or vice-versa.

In order to investigate these possibilities, we have done a hetero- logous hybridization : mRNA from induced cells has been hybridized with cDNA complementary to uninduced cell mRNA. Fig. 3 shows that the final level of hybridization (90%) is not significantly different from that of the homologous reaction involving the same cDNA (Fig. 1). Moreover, the shape of both curves is indistinguishable but slightly shifted to higher Rot in the heterologous reaction. This is confirmed by the computer curve f i t t i n g which yields similar proportions and similar relative rate cons- tants for both classes. (The uncorrected proportion of the abundant class in the homologous reaction is 0.414 + 0.016 (standard error), as shown in

Table 2

A. Fraction of cDNA i n 1st abundance class

0.23 - 0.27

B.

RotJ of pure component

0.0022-0.0038

0.79

Sequence complexity (number of d i f f e -

rent mRNAs of average size)

3.6 - 6.3

Proportion of globin sequence i n t o t a l poly(A) -mRNA

0.025

Number of molecules of each mRNA per c e l l

13000-19000

Number of globin molecules per c e l l

11300

Numerical data from the hybridization in Fig. 2. A. Hybridization of induced cell poly(A) -mRNA with its own cDNA. Close inspection of the computer fitting reveals that the parameters pertaining to the 2nd and 3rd abundance class are not uniquely definded by the data (in other words, these two kinetic components are not well separated). But it also shows that this uncertainty barely affects the parameters for the 1st class, which are comprised in the range indicated. The calculations are similar to the ones in Table 1, taking 15 pg as the amount of RNA per induced cell. B. Hybridization of total induced cell polysomal RNA with globin cDNA.

3022

at Institut universitaire de hautes etudesinternationales - Bibliotheque on June 14, 2016http://nar.oxfordjournals.org/Downloaded from

(12)

Fig. 3

Heterologous hybridization : induced cell mRNA with uninduced cell cDNA.

Poly(A)

+

-mRNA obtained from induced cells was hybridized with cDNA transcri- bed from uninduced cell mRNA. The data were f i t t e d to a two component model.

For the purpose of comparison, the curve f i t t i n g of the homologous reaction is also shown. (• •) Heterologous reaction : data points and curve f i t - ting. ( ) : uninduced cell mRNA with uninduced cell cDNA, curve f i t t i n g from Fig. 1.

Table 1 while the corresponding value in the heterologous reaction is 0.448 + 0.013). This result is consistent with the interpretation that, except for the 4 to 6 mRNAs which are greatly increased, the vast majority of the messenger RNAs which are present in the uninduced state are still expressed in the induced cell and present at the same relative frequency.

Moreover, since the heterologous hybridization is essentially complete, there must be little or no qualitative differences between the two popu- lations of messengers.

This point was confirmed by performing the opposite heterologous hybridization : mRNA from uninduced cells has been hybridized with cDNA complementary to induced cell mRNA. In this experiment, we probe the mRNA population of uninduced cells for the existence and frequency of sequences characteristic of the differentiated state. As shown in Fig. 4 this hetero- logous reaction also goes to completion. As compared to the homologous reaction in the induced case shown in the same figure, this heterologous hybridization does not show the fast-reacting component. Therefore, most messengers abundant in the induced cell are already present in the uninduced stage, but they hybridize at higher R

Q

t, which means that they are present at a lower frequency.

In summary, it appears that at the level of the population of mRNA

at Institut universitaire de hautes etudesinternationales - Bibliotheque on June 14, 2016http://nar.oxfordjournals.org/Downloaded from

(13)

Nucleic Acids Research

log Rot

Fig. 4

Heteroiogous hybridization : uninduced-cell mRNA with induced cell cDNA Poly(A)

+

-mRNA obtained from uninduced cells was hybridized with cDNA trans- cribed from induced-cell mRNA. The data were f i t t e d to a two-component model.

The curve f i t t i n g of the homologous reaction in the induced case is also shown.

(• •) Heterologous reaction : data points and curve f i t t i n g . ( ) Homo- logous reaction : induced-cell mRNA with induced-cell cDNA curve f i t t i n g from Fig. 2.

sequences, the differentiation of Friend cells involves mainly an increase in abundance of a small subset of messenger sequences already present before induction. As a result of this increase, a few mRNA species reach concentrations which are not found for any messenger type in the undif- ferentiated cells. On the other hand, based on the heterologous hybridi- zation in Fig. 3, we are unable to detect any specific pattern of messenger disappearance linked to cell differentiation.

Quantitative distribution of messenger sequences in pre-mRNA of uninduced cells.

The sequence complexity of total poly(A)

+

- nuclear RNA of uninduced Friend cells had been measured using a cDNA probe transcribed from i t (Getz et a l . 18). I t has been found to be 5 times higher than the comple- xity of poly(A)

+

-mRNA, indicating that only a fraction of poly(A)

+

-nuclear RNA is pre-mRNA. Furthermore, the population of polyadenylated nuclear RNA can be ordered into 2 frequency classes. Our aim is to know whether at the level of nuclear RNA, the messenger sequences are distributed with the same relative frequencies as in mRNA i t s e l f . In other words, we shall consider the frequency distribution not of hnRNA as a whole, but of the subset of hnRNA sequences which consist of messenger sequences (pre-mRNA).

I t is important to realize that the result of Getz et a l . on total poly(A)

3024

at Institut universitaire de hautes etudesinternationales - Bibliotheque on June 14, 2016http://nar.oxfordjournals.org/Downloaded from

(14)

nuclear RNA does not allow one to draw conclusions about the frequency d i s t r i b u t i o n of pre-mRNA : pre-mRNA i s only a subset o f t o t a l p o l y ( A )+ nuclear RNA, comprising 20^ o f i t s complexity. Therefore, i f pre-mRNA consisted o f m u l t i p l e frequency classes, they might not be resolved i n a h y b r i d i z a t i o n involving the whole o f polyadenylated nuclear RNA.

Total nuclear RNA of uninduced c e l l s has been hybridized with cDNA complementary to mRNA o f uninduced c e l l s . I f the r e l a t i v e abundance of messenger sequences i n mRNA and pre-mRNA do correspond, the Rot curve should be p a r a l l e l to the curve i n F i g . 1 and s h i f t e d to the r i g h t by a f a c t o r expressing the d i l u t i o n o f the reacting messenger sequences by unrelated sequences. The r e s u l t ( F i g . 5) shows that t h i s i s not the case.

Rather than being biphasic, the curve looks p a r a l l e l to the single-compo- nent hybridization curves i n v o l v i n g globin cDNA. Therefore, w i t h i n hnRNA, most o f the messenger sequences hybridize as a single abundance class whose RotJ equals 690. Since the complexity of mRNA i s about 10^f average messenger species and the corresponding RotJ equal to 6 mole xsxl

(Table 1 ) , the proportion o f nuclear RNA d r i v i n g the r e a c t i o n , i . e . the r e l a t i v e mass of the pre-messenger sequences i n t o t a l nuclear RNA i s 6/690 = 0.87%. The amount o f t o t a l nuclear RNA has been determined by measuring the RNA and DNA content o f nuclear preparations with o r c i n o l and diphenylamine respectively (31) and found to be about 3 pg/nucleus.

-1? 23 -1 5 This corresponds to 0.0087 x 3 x 10 g x 6 x 10 mole / 6.10 g/mole

= 26100 messenger-sized sequences or 26100/10000 = 3 molecules o f each species approximately. The r e s u l t shows t h a t about 3 molecules o f every messenger sequence expressed i n the uninduced c e l l are present i n i t s nucleus, which agrees with the abundance o f the complex class o f t o t a l poly(A) -hnRNA found by Getz e t a l . (6 molecules/nucleus).

Quantitative d i s t r i b u t i o n o f messenger sequences i n pre-mRNA o f induced cells

In order to see whether in induced c e l l s , a similar difference between the frequency distribution of mRNA and pre-mRNA can be observed, induced cell nuclear RNA has been hybridized with cDNA transcribed from induced cell poly(A) -mRNA. Relative to the uninduced cell situation, the hybridization curve is shifted toward the lower Rot values, indicating a relative enrichment in precursor messenger sequences. In addition, the curve in Fig. 6 extends over a s l i g h t l y broader range of Rot values, i n d i -

at Institut universitaire de hautes etudesinternationales - Bibliotheque on June 14, 2016http://nar.oxfordjournals.org/Downloaded from

(15)

Nucleic Acids Research

log Rot

Fig. 5

Frequency of messenger sequences in nuclear RNA from uninduced cells. Total nuclear RNA prepared from uninduced cells was hybridized to cDNA complemen- tary to uninduced cell poly(A)

+

-mRNA. The RNA concentration was 4.4 mg/ml.

A similar hybridization has been performed with globin cDNA. (• - •) Hybri- dization of uninduced cell nuclear RNA with cDNA transcribed from uninduced cell mRNA. ( o ) Hybridization of uninduced cell nuclear RNA with globin nDNA.

log Rot

Fig. 6

Frequency of messenger sequences in nuclear RNA from induced cells. Total nuclear RNA has been prepared from induced cells and hybridized to cDNA complementary to induced cell poly(A)

+

-mRNA at an RNA concentration of 3.3 mg/ml. A similar preparation has been hybridized with globin cDNA. (• - •) Hybridization of induced cell nuclear RNA with cDNA transcribed from indu- ced cell mRNA. (o - o) Hybridization of induced cell nuclear RNA with glo- bin cDNA.

eating kinetic heterogeneity. Therefore, the pre-mRNA of induced cells seems to be more heterogeneous in i t s frequency distribution, and precursor messenger sequences are enriched to varying extents as compared to the uninduced situation. As shown by hybridizing the same RNA with globin cDNA, this enrichment is higher for the globin sequence which is increased

3026

at Institut universitaire de hautes etudesinternationales - Bibliotheque on June 14, 2016http://nar.oxfordjournals.org/Downloaded from

(16)

100-fold in induced nuclear RNA relative to uninduced nuclear RNA.

The possibility s t i l l exists that the globin sequence is an exception and that for most sequences the relative increase observed after induction is non-specific and due, for instance, to a decrease in precursor ribosomal RNA. Alternatively, this increase could be specific for the sequences enriched in induced-cell mRNA. This could be tested by a heterologous hybridization in which induced-cell nuclear RNA is reacted with the cDNA probe for uninduced mRNA. We have hybridized a single induced-cell nuclear RNA preparation with both cDNA probes, that i s , we repeated the homolo- gous reaction (induced cDNA) and made the heterologous reaction using the same material. Fig. 7 shows in dotted lines the two homologous reactions taken from Fig. 5 and 6.

The f i r s t set of data points is a repetition of the homologous reac- tion in the induced situation and therefore the points match the curve taken from Fig. 6. The second corresponds to the heterologous reaction of the induced-cell nuclear RNA with the cDNA complementary to uninduced- cell mRNA; this data set can be seen to match the homologous reaction curve for the uninduced case. Therefore,the higher frequency found for induced-cell pre-mRNA is not due to a general enrichment in pre-messenger sequences but is specific for those mRNAs which are more abundant in the cytoplasm after induction. Induced-cell pre-mRNA is more heterogeneous in i t s frequency distribution which shows some correlation with the abundances of the cytoplasmic mRNAs.

DISCUSSION

In these experiments, we have dealt with uninduced and induced cells as if they were homogeneous populations. This is very nearly so in the un- induced state, since one finds less than 0,1% of spontaneously differentia- ted cells. Cell heterogeneity may be more of a problem in the induced stage : due to the stochastic nature of the commitment to induction (32), no pure population of completely differentiated cells can ever be obtained.

In the extreme, one could argue that the heterologous hybridization in Fig. 4 would be due to the undifferentiated cells contaminating the induced population. These cells represent not more than 10% of the total, but they contain twice as much polysomal RNA. This 20% contamination is insufficient to explain the data in Fig. 3 but it may influence them by making the

at Institut universitaire de hautes etudesinternationales - Bibliotheque on June 14, 2016http://nar.oxfordjournals.org/Downloaded from

(17)

Nucleic Acids Research

cr0

9 i.o- a.a

\ 0.8-

z

° 0.6-

t

^T* 0.4- oz

o: 02-

3r

S v

• - - — ' ' '

— o - '

-1 0 1 2 3 4

log Rot

Fig. 7

Homologous and heterologous reaction involving induced-cell nuclear RNA.

A single induced-cell nuclear RNA preparation was used in homologous and heterologous hybridization (cDNA transcribed from induced and from unin- cU-ced-cell mRNA respectively). The RNA concentration in both reaction .'Mixtures was 3.3 mg/ml. (• - •) Homologous reaction, (o - o) Heterologous reaction. ( — ) Homologous reaction;; taken from Fig. 4 and 5.

heterologous curve more similar to the homologous one. Therefore, apart from the very drastic increase of 4 to 6 mRNAs, other frequency changes may occur and go undetected, either because of their low amplitude or because they involve only a minority of the messenger species in a frequen- cy class. Furthermore, the fact that no specific pattern of mRNA

disappearance can be detected must also be considered in view of the limited resolution of the hybridization technique. Such experiments do not allow, in most cases, to detect changes in single mRNAs. However, Peterson and McConkey have established protein patterns of uninduced and induced Friend cells by two-dimensional electrophoresis : 98% of the 500 proteins that they are able to resolve are similar in quantity and quality in both types of cells and they observe the decrease or disappearance of only four chromatin proteins and two cytoplasmic proteins. These findings confirm that there is a high degree of similarity between the two mRNA populations.

We have found that the mRNA of uninduced Friend cells is distributed into 2 frequency classes. Similar experiments had already been performed by Birnie et a l . (3) leading to a similar estimate of the total complexity but with a distribution of mRNA in 3 classes. In order to appreciate f u l l y the difference between their results and ours, one should bear in mind the use of more kinetic components is bound to improve the f i t between the data

3028

at Institut universitaire de hautes etudesinternationales - Bibliotheque on June 14, 2016http://nar.oxfordjournals.org/Downloaded from

(18)

and the model. The element of arbitrariness lies in the choice of a signi- ficance threshold and we consider a 2-component model to be satisfactory.

Fortunately, the number of components affects very l i t t l e the overall

A

complexity found (10 messenger sequences)

Our experiments with nuclear RNA from uninduced cells show that

pre-mRNA is organized into a single abundance class and corresponds approxi- mately to 3 molecules per nucleus. Comparison of this result with the expe- riment by Getz et al. already shows that pre-mRNA sequences are not enriched over other poly(A) hnRNA sequences which do not correspond to messenger species found in the uninduced Friend cell. It also shows that the more abundant poly(A)

+

-hnRNA sequences found by Getz et al. are apparently not precursors to the abundant mRNAs, at least not the majority of them.

The same authors have measured the complexity of uninduced cell total

nuclear RNA by hybridizing it with single-copy DNA (19) : nuclear RNA trans- cribed from single-copy genes is four times more complex than mRNA and comprises 3% of total nuclear RNA. Interestingly, we find that pre-mRNA amounts to 0.8% of total nuclear RNA, i.e. about 1/4 of the complex class of nuclear RNA (0.8/3 = 1/4) : the ratio of the amounts matches the ratio of the complexities, further supporting the view that pre-mRNA sequences are not enriched over other transcripts not to be exported into the cyto- plasm.

In conclusions, our results with uninduced Friend cells suggest that posttranscriptional events are important on a quantitative level, since most prospective messenger sequences have the same abundance in the nucleus and are not more abundant than other transcripts from unique sequences.

The situation is somewhat different in the induced cells where pre- mRNA is enriched to varying extents in sequences found to be abundant in the cytoplasm. This could mean that for induced mRNAs, transcriptional activation is more important. On the other hand, this result may be due to contamination of our nuclear preparation by cytoplasmic material.

Considering the concentration of the globin sequence in polysomal and in nuclear RNA of induced cells, this contamination should amount to at least 10% in order to explain our results. However, in uninduced cells such a level of cytoplasmic contamination would lead to a 15-fold enrich- ment of abundant messenger sequences over the amount found for total

at Institut universitaire de hautes etudesinternationales - Bibliotheque on June 14, 2016http://nar.oxfordjournals.org/Downloaded from

(19)

Nucleic Acids Research

sequences. Since this would certainly be detectable in Fig. 5, cytoplasmic contamination is not a likely explanation for our result with the nuclear globin sequence.

Contamination of polysomes by nuclear material would have an effect on our experiments since it would contribute RNA of high complexity to the polysomal fraction used to prepare the cDNA. However, the lysis procedure had been devised to purify polysomes which tested by centrifugation in a CsCl gradient, appear to be free of nuclear RNPs (Spohr et al., 34) Nevertheless, even if we allow for a contamination of 10% of poly(A) -mRNA by poly(A)

+

nuclear RNA, this would contribute a maximum of 10%

of cDNA sequences of nuclear origin (probably less, since on a weight basis, poly(A) nuclear RNA is reverse-transcribed less than poly(A) -mRNA).The complex component of polyadenylated nuclear RNA is about 5 times more complex than poly(A)

+

-mRNA (18) and would therefore hybridize with its cDNA at a R

o

t \ of 30 in our conditions. In the experiment of Fig. 1, this contaminant of 10% would therefore hybridize at a R

o

t \ of 300 or more; it would be present as an additional minor complexity class, which

is not observed, rather than merging into the complex class of mRNA.

An approach similar to ours was used by Sippel et al. (35) and Jacquet et al. (36) on rat liver and embryonal carcinoma cells respective- ly. In both studies, the frequency distribution of messenger sequences in nuclear RNA was found to be narrower than in mRNA. Kinetic studies involving the globin messenger sequence in Friend cells were also performed. Aviv et al. (37) have demonstrated differential turn-over of globin mRNA as compared to total poly(A)

+

-mRNA while Orkin et al. (38) have shown an increase of the transcriptional rate of the globin gene during differentiation. Similarly, our results suggest that the generation of the quantitative structure of the mRNA population in a cell cannot be ascribed to a single mechanism. Post-transcriptional modulation probably does occur, while transcriptional activation might be most important for the few very predominant mRNAs which appear in terminally differentiated cells.

Acknowledgements

We thank Raymonde Cornuz for excellent technical assistance. We are deeply indebted to Bernhard Hirt for his continuous support and advice.

3030

at Institut universitaire de hautes etudesinternationales - Bibliotheque on June 14, 2016http://nar.oxfordjournals.org/Downloaded from

(20)

We acknowledge the g i f t of reverse-transcriptase by Heidi Diggelmann and Joseph Beard, The Office of Program Resources and L o g i s t i c s , Viral Cancer Program, National Cancer I n s t i t u t e , Bethesda, MD. We thank Doris

Chiapparelli, Sophie Cherpillod and Pierre Dubied f o r help in preparation of the manuscript and David Hughes f o r c r i t i c a l l y reading i t . We also thank David Appleby and Yves Depeursinge for assistance i n the computer work, which was performed at the Computer Center of the Ecole Polytechnique Federale de Lausanne. Supported by the Swiss National Science Foundation (grant No. 3.537.75)

+ To whom requests f o r reprints should be sent at his present address : Dgpartement de Biologie Animale de 1'Universite de Geneve, route de Malagnou 154, CH- 1224 Chene-Bougeries, Switzerland

References

1. Galau, G.A., B r i t t e n , R.J. and Davidson, E.H. (1974) Cell 2, 9-20 2. Bishop, J . O . , Morton, J . G . , Rosbash, M. and M e l v i l l e Richardson (1974)

Nature 250, 199-204

3. B i r n i e , G.D., McPhail, E., Young, B.D. Getz, M.J. and Paul, J . (1974) Cell D i f f e r e n t i a t i o n 3, 221-232

4. Levy, W., B. and McCarthy, B.J. (1975) Biochemistry 14, 2440-2446 5. Galau, G.A., K l e i n , W.H., Davis, M.M. Wold, B . J . , B r i t t e n , R.J. and

Davidson, E.H. (1976), Cell 7, 487-505'

6. R y f f e l , G.U. and McCarthy, B.J. (1975) Biochemistry 14, 1379-1385 7. Hastie, N.D., and Bishop, J.O. (1976) Cell 9, 761-774

8. Young, B.D., B i r n i e , G.D. and Paul, J . (1976) Biochemistry 15, 2823- 2829

9. Paterson, B.M. and Bishop, J.O. (1977) Cell 12, 751-765

10. A f f a r a , N.A., Jaquet, M., Jakob, H., Jacob, F. and Gros, F. (1977) Cell 12, 509-520

11. Hynes, N.E., Groner, B., Sippel, A . E . , Chi Nguyen-Huu, M. and Schiitz.G.

(1977) Cell 1 1 , 923-932

12. Parker, M.G., Mainwaring, W. and Ian, P. (1977) Cell 12, 401-407 13. Axel, R., Feigelson, P. and Schutz, G. (1976) Cell 7, 247-254 14. Hough, B.R., Smith, M.J., B r i t t e n , R. J . and Davidson, E. H. (1975)

Cell 5, 291-299

15. Levy, W., B. and McCarthy, B.J. (1976) Biochemistry 15, 2415-2419 16. Chikaraishi, D.M., Deeb, S.S. and Sueoka, N. (1978) Cell 13, 111-120 17. Bantle, J.A. and Hahn, W.E. (1976) Cell 8, 139-150

18. Getz, M.J., B i r n i e , G.D., Young, B.D., McPhail, E. and Paul, J . (1975) Cell 4 , 121-129

19. Kleiman, L . , B i r n i e , G.D., Young, B.D. and Paul, J . (1977) Biochemis- t r y 16. 1218-1223

20. Friend, C , Patuleia, M.C. and de Harven, E. (1966) N a t l . Cancer I n s t . monogr. 22, 505-522

21. Friend, C , Scher, W., Holland, J.G. and Sato, T. (1971) Proc. N a t l . Acad. S c i . USA 68, 378-382

22. Eisen, H., Bach, R. and Emery, R. (1977) Proc. N a t l . Acad. S c i . USA 74, 3898-3902

at Institut universitaire de hautes etudesinternationales - Bibliotheque on June 14, 2016http://nar.oxfordjournals.org/Downloaded from

(21)

Nucleic Acids Research

23. Conscience, J . F . , M i l l e r , R.A., Henry, J . and Ruddle, F.H. (1977) Exp.

Cell Res. 105, 401-412

24. Sassa, S. (1976) J . Exp. Med. 143, 305-315

25. Perry, R.P., La Torre, J . , Kelley, D.E. and Greenberg, J.R. (1972) Biochim. Biophys. Acta 262, 220-226

26. Aviv, H. and Leder, P. (1972) Proc. Natl. Acad. S c i . USA 69, 1408-1412 27. Scherrer, K. (1969) i n Fundamental Techniques i n Virology, Habel and

Salzman Eds., pp. 413-432 Academic Press, New York

28. Williamson, R., Morrison, M., Lanyon, G., Eason, R. and Paul, J . (1971) Biochemistry 10, 3014-3021

29. Imaizumi, T . , Diggelmann, H. and Scherrer, K. (1973) Proc. N a t l . Acad.

S c i . USA 70, 1122-1126

30. Vogt, V.M. (1973) Eur. J . Biochem. 33, 192-200

3 1 . McMaster, G. and Modak, S.P. (1977) D i f f e r e n t i a t i o n 8 , 145-152 32. Gusella, J . , G n l l e r , R., Clarke, B., Weeks, V. and Housman, D. (1976)

Cell 9, 221-229

33. Peterson, J . L . and McConkey, E.H. (1976) J . B i o l . Chem.251, 555-558 34. Spohr, G., Granboulan, N., Morel, C. and Scherrer, K. (1970)

Eur. J . Biochem. 17, 296-318

35. Sippel, A . E . , Hynes, N., Groner, B. and Schiitz, G. (1977) Eur. J . Biochem. 77, 141-151

36. Jacquet, M. A f f a r a , N.A., Robert, B., Jakob, H., Jacob, F. and Gros, F. (1978) Biochemistry 17, 69-79

37. Aviv, H., Voloch, Z . , Bastos, R. and Levy, S. (1976) Cell 8, 495-503 38. Orkin, S.H. and Swerdlow, P.S. (1977) Proc. N a t l . Acad. S c i . USA 74,

2475-2479

3032

at Institut universitaire de hautes etudesinternationales - Bibliotheque on June 14, 2016http://nar.oxfordjournals.org/Downloaded from

Références

Documents relatifs

Therefore, the present study was designed to characterize the expression of plasma membrane nucleoside transporters, i.e., ENT1, ENT2, CNT1, CNT2 and CNT3, in HepaRG cells, as well

To address how the combination of elevated antigen expression levels with reduced inflammatory responses upon use of N1mψ nucleoside modified mRNA impacts the

(Note that if one com- pares optimizing agents’ performance to the performance of a separate system composed entirely of non-optimizing agents, there are regimes in m for which

Fin octobre 2017, en France métropolitaine, parmi les personnes inscrites à Pôle emploi et tenues de rechercher un emploi (catégories A, B, C), 3 483 600 sont sans emploi (catégorie

Die  Analyse  hat  gezeigt,  dass  es  sich  bei  der  Hauptzielgruppe  für  E‐Bikes  in  der  Region  Zürichsee  um  erwachsene,  arbeitstätige 

TABLE 4 (Continued) Fire Endurance- of Load-Bearing Structural Tile Walls Thickness ENDURANCE (in.) Tile face Units in Cells in Minimum Solid Nc Combustible

Aucune importance Une légère importance Une certaine importance Une moyenne importance Une importance considérable Une énorme importance Une extrême importance 0 1

In fact, we show that deciding the existence of a solution for Total Dominating Set with no con- flicts, Connected Vertex Cover with no conflicts and Steiner Tree with no conflicts