Gene, 141 (1994) 53-61

0 1994 Elsevier Science B.V. All rights reserved. 0378-1119/94/$07.00 53

GENE 07739

The use of arbitrary primers and the RADES method for the rapid identification of developmentally regulated genes in trypanosomes

(Fingerprints; differentiation; cDNA; randomly amplified polymorphic DNA; PCR; RAPD)

Noel B. Murphy and Roger Pellk

International Lahoratory,for Research on Animal Diseases (ILRAD), Nairobi, Kenya

Received by P.F.G. Sims: 19 July 1993; Revised/Accepted: 11 October/l November 1993; Received at publishers: 2 December 1993

SUMMARY

Biological processes, such as the cell-division cycle, differentiation and development, are driven by changes in gene

expression. Short oligodeoxyribonucleotide primers (lo-mers) of arbitrary sequence are currently used in the polymerase

chain reaction (PCR) to generate genomic fingerprints (RAPDs) for the characterisation and differentiation of organisms

and for mapping loci of interest. Since the products of such reactions are generally less than 1 kb in size, the use of

arbitrary primers on cDNA should generate RAPDs which are characteristic of expressed genes. To assess this possibility,

two model systems were employed; one in which actively dividing Trypanosoma brucei brucei bloodstream forms

differentiate to non-dividing forms, and the second in which non-dividing metacyclic forms of T. congolense differentiate

to actively dividing bloodstream forms. In the technique herein, mRNA from each differentiated form was reverse

transcribed into cDNA which was then used as the template in the PCR. The resultant products were examined by

agarose-gel electrophoresis. As few as lo3 trypanosomes were sufficient for the generation of a RAPD print after first

amplifying the total cDNA through exploitation of the fixed 3’ and 5’ ends of trypanosome nuclear mRNAs. Differences

in RAPD patterns between the differentiated forms examined were mainly due to differences in gene expression. The

technique can rapidly identify genes expressed at very low levels and which are up- or down-regulated in the different

forms examined. PCR products of interest are easily purified from the agarose gels for direct cloning and complete

sequence determination due to their relatively small size (0.1-l kb).

INTRODUCTION

A principal focus of molecular biology research is to

identify and analyse developmentally regulated genes that

control cellular differentiation, because this information

furthers our understanding of how such genes influence

biological processes. Classical methods for isolating de-

velopmentally regulated genes involve differential screen-

Corrrspondence to: Dr. N.B. Murphy, ILRAD, P.O. Box 30709, Nairobi,

Kenya. Tel. (254-2) 632-311; Fax (254-2) 631-499;

e-mail: murphynb@ilrad.msm.cgnet.com

Abbreviations: AP, arbitrary primer; bp, base pair(s); BSF, bloodstream

form(s); dNTP, deoxyribonucleoside triphosphate; ds, double

strand(ed); EtdBr, ethidium bromide; GAPDH, glyceraldehyde-3-phos-

SSDI 0378-I 119(93)E0735-V

ing of cDNA libraries or subtractive enrichment

methodologies (see Sambrook et al., 1989). Differential

screening is relatively insensitive and can identify only

moderately to highly expressed genes. Subtractive enrich-

ment methodologies are more sensitive but are techni-

cally difficult and require large quantities of mRNA

(Welcher et al., 1986). Recent developments, including

the use of the polymerase chain reaction (PCR), have

phate dehydrogenase; ILTat, International Laboratory Trypanozoon

antigen type: Int, intermediate; kb, kilobase or 1000 bp; LS, long slen-

der; nt, nucleotide(s); oligo, oiigodeoxyribonucleotide; PCR, polymerase

chain reaction; RADES, random amplified differentially expressed se-

quences; RAPD, random amplified polymorphic DNA; ss, single

strand(ed); SS, short stumpy; T., Trypanosoma; T.h.b., T. hrucei brucei;

TX.. T. congolense; Tub, gene encoding B-tubulin of T.h.b.

54

improved the latter technique (Hara et al., 1991; Houge,

1993; Lee et al., 1991; Nedivi et al., 1993; Wang and

Brown, 1991). However, the methodology remains tech-

nically difficult and is subject to artefacts and problems

in analysis, particularly for many important genes tran-

scribed at low levels.

Our interest in differential gene expression is to identify

developmentally regulated genes in African trypano-

somes, particularly those genes involved in the onset and

arrest of proliferation at different life-cycle stages.

Trypanosomes that undergo complete cyclical develop-

ment can control their growth rates at different points

during their life-cycle (Ashcroft, 19.57; Balber, 1972; Black

et al., 1985; Seed and Sechelski, 1989; Shapiro et al., 1984;

Vickerman, 1971; 1985; Wijers et al., 1959). In the tsetse

fly vector, mature metacyclic forms, which are infective

for the mammalian host, are non-dividing and arrested

in G, (Vickerman, 1985). When an infected fly takes a

blood-meal from a mammalian host, the infective meta-

cyclic forms pass into the host and begin to differentiate

into actively dividing bloodstream forms (BSF). These

trypanosomes proliferate in the vascular system, resulting

in relatively large numbers of organisms. Studies of

Trypanosoma brucei brucei (T.b.b.) in rodents have shown

that towards the peak of an infection, the trypanosomes

are capable of differentiating to non-dividing forms ar-

rested in G, (Ashcroft, 1957; Shapiro et al., 1984;

Vickerman, 1971). These forms can be cleared by the host

immune system through recognition of the surface coat

of the parasite (Sendashonga et al., 1982). Clarifying the

regulatory mechanisms and factors involved in the con-

trol of trypanosome cell division will give a better under-

standing of host-parasite interactions, and how these

influence parasite proliferation and host pathogenesis.

Use of arbitrary primers on genomic DNA in the PCR,

termed ‘AP-PCR’, generates reproducible fingerprint pat-

terns characteristic of the DNA source and the primer

used (Welsh and McClelland, 1990; Williams et al., 1990).

Recent reports demonstrate the usefulness of AP-PCR in

differentiating trypanosome species and subspecies (Dirie

et al., 1993; Tibayrenc et al., 1993; Waitumbi and Murphy,

1993). Because the products of AP-PCR, known as

‘random amplified polymorphic DNAs’ (RAPDs), gen-

erally range in size between 0.1 and 1 kb, cDNA should

also act as an appropriate template for such reactions.

Furthermore, the resultant RAPDs should reflect the

population of genes which are expressed in the cells from

which the cDNA has been generated. Two recent reports

have exploited AP-PCR to identify differentially ex-

pressed genes in mammalian tissues. In both reports, arbi-

trary primers were used to generate cDNA from a subset

of either total RNA (Welsh et al., 1992) or polyadenylated

RNA molecules (Liang and Pardee, 1992) and were de-

pendent on the analysis of radiolabelled PCR products

on polyacrylamide sequencing gels. Further characterisa-

tion of products of interest required their purification

from dried sequencing gels followed by additional PCR

amplifications, prior to cloning or Northern blot analysis.

Although these represent significant improvements over

previous methods, the purification steps necessitate re-

moval of urea and elution of the single stranded (ss) DNA

of interest from the gel, during which contaminating

DNA can be introduced that can act as a template in the

additional rounds of amplification required for analysing

the products of interest. In addition, radiochemicals

and polyacrylamide used require careful handling.

Subsequent exposure of the gel to X-ray film and the

re-amplification steps also lengthen the time of analysis.

Herein we describe a PCR-based method termed

‘random amplified differentially expressed sequences’

(RADES), which is rapid, requires only a few organisms,

generates reproducible results and can be used to identify

developmentally expressed genes with minimum effort

and time. The RADES products are visualised on agarose

gels from which DNA fragments of interest can be puri-

fied for further analysis, without the need for additional

rounds of the PCR. Confirmation that the differentially

amplified PCR products are from developmentally ex-

pressed genes is quick and easy, and rapid sequence

analysis can be carried out following cloning of the pro-

ducts in plasmid vectors. We have tested the RADES

technique in trypanosome model systems and have iden-

tified many, and mostly as yet uncharacterised, develop-

mentally regulated genes during both the life-and cell-

division cycles. By amplifying total cDNA prior to the

RADES-PCR, we show that as few as lo3 organisms can

generate a RAPD pattern. The technique, in combination

with anchor PCR methodologies, could be extended to

the analysis of differential gene expression in single cells.

RESULTS AND DISCUSSION

(a) Amplification of cDNA from different forms of T-6.

hcei ILTatl.1 with arbitrary primers

Pleomorphic isolates of T.b.b. display several distinc-

tive morphological stages during a parasitaemic wave in

an infected rodent host. Early in infection, actively divid-

ing long slender (LS) forms, whose flagellum is partly

free, predominate during the rising parasitaemia.

Towards the peak of a parasitaemic wave, the LS forms

shorten and differentiate through an intermediate (Int)

stage to a non-dividing short stumpy (SS) form whose

flagellum is completely covered by the undulating mem-

brane. The SS form is a terminal stage, arrested in G, of

the cell-division cycle (Shapiro et al., 1984) and quickly

55

1 2 3 4 5 6 7 8 9 10

12345678 9 10

6

:

*

I 2 3 4 5 6 7 6 9 10 II-r I

C

Fig. I. Fingerprinting of cDNA generated from different life-cycle

stages of T.b.b. ILTatl.1 by the RADES technique. Each set of RADES-

PCRs includes a single primer and 4 separate reactions using LS, Int.

SS and procylic ILTatl.1 cDNA, respectively, as the target source. Oligo

primers were synthesised by the phosphoramidite method on an DNA

synthesizer (Model 381A, Applied Biosystems, Foster City, CA,

USA). Primers used in the examples shown above were I, ILO508,

5’-CGGCCCCTGT; 2, ILO523, S-CGAGCCGCGG: 3, IL0525,

5’-CGGACGTCGC; 4, lLO526, 5’-GCCGTCCGAG; 5, ILO543,

5’-GTGTCCGGCG; 6, ILO867, 5’-CGTTCCCCGC; 7, IL0868

5’-CAGCCTCGGC: 8, ILO872, 5’-CCCGCCATCT; 9, ILO873,

5’CATGTGCAGG; 10, ILO874, 5’-GAGGTGGCGC. (A) An EtdBr-

stained 1.5”/0 agarose gel of RADES fingerprints from LS, Int, SS and

procyclic cDNAs, left to right respectively, with primers i-i0 as above.

Size markers (hashed lines) represent the position of concatamers of

the 123-bp ladder (GIBCO-BRL, Uxbridge, UK). Square brackets high-

light products hybridising to labelled cDNA in panels B and C and

which are discussed in section c. Following transfer to a nylon filter,

hybridisation with labelled LS cDNA (B) and SS cDNA (C) was carried

out. Filled arrowheads point to detectable specific hybridisation signals

of products preferentially amplified in (B) LS or (C) SS forms. The

asterisk in (B) highlights a specific hybridisation signal to a LS product,

despite products of the same size being generated from the other forms

and is discussed in section c. The open circle in (B) highlights a product

which is not detectable in the EtdBr-stained gel, but which hybridises

to labelled LS cDNA. The open arrowhead in(C) highlights a hybridisa-

tion signal to an apparent LS-specific product with labelled SS cDNA

alone. Open boxes in(C) show amplified procylic products which prefer-

entially hybridise to the labelled SS cDNA. Methods: The different BSF

of T.b.6. tLTat1.1, LS, Int and SS were generated and purified as de-

scribed (Pelle and Murphy, 1993a). Procyclic culture form trypano-

dies unless transferred to in vitro or in vivo (tsetse fly

vector) conditions in which the insect procyclic form can

develop. Relatively pure populations of these different

forms of T.b.b. ILTatl.1 can be generated and purified

for biochemical and genetic analysis (Pelle and

Murphy, 1993b).

To test whether short oligodeoxyribonucleotide (oligo)

primers (lo-mers) of arbitrary sequence used on total try-

panosome cDNA could generate a detectable pattern of

products when analysed by agarose-gel electrophoresis,

mRNA from populations of LS, Int, SS and procyclic

forms was reverse transcribed and double-stranded (ds)

cDNA generated. The oligo(dT) primer, dNTPs and

buffer for cDNA synthesis were removed by filtration

through a Centricon 30 column (Amicon, Beverly, MA,

USA) prior to the RADES-PCR. For AP-PCR of geno-

mic DNA, a target DNA concentration of 1 ng/ul gener-

ates reproducible RAPDs which can be visualised on

polyacrylamide and agarose gels following simple ethid-

ium bromide (EtdBr) staining (Dirie et al., 1993;

Waitumbi and Murphy, 1993). The cDNA was, therefore,

adjusted to this concentration in the PCR reaction mix-

ture and 40 rounds of amplification were carried out. As

with RAPDs from genomic DNA, RADES products from

cDNA were easily detected in agarose gels (Fig. 1A) and

most primers generated consistent fingerprints, although

the number of bands detected were usually fewer than

with genomic DNA (Dirie et. al., 1993; Waitumbi and

somes were propagated in vitro as described (Bienen et al., 1991). Total

RNA was isolated by the acid guanidinium-thiocyanate-phenol-

chloroform method (Chomczynski and Sacchi, 1987) and enriched for

polyadenylated transcripts by oligo(dT) chromatography (Sambrook

et al., 1989). The poly(A)-enriched RNA was reverse transcribed and

converted to ds cDNA using a cDNA synthesis kit according to the

manufacturers instructions (GIBCO-BRL). Removal of the oligo(dT)

primer and dNTPs was achieved by filtration through a Centricon 30

column (Amicon). The ds cDNA was diluted to 20 ng/ul and used as

the template for the RADES-PCR. PCRs contained 20 ng of target

DNA/10 mM TrisHCl (pH 8.3)/50 mM KC]/3 mM MgCl,/O.OS% (v/v)

NP40/0.05% (v/v) Tween-20/200 uM of dNTPs/0.6 pM primer/O.5 units

of Taq (Thermus aquaticus) DNA polymerase (Promega, Madison, WI,

USA) in a volume of 20 ul, and were performed on a programmable

thermal cycler (MJ Research, Watertown, MA, USA). Cycling condi-

tions were 94°C for 45 s, 37’C for I min and 72’C for I min for 40

cycles followed by a 5 min extension at 72’C. Amplification products

were analysed by electrophoresis in 1.5% agarose gels and detected by

UV illumination following staining with EtdBr. DNA within the gels

was denatured and transferred to nylon filters (Nytran,

Schleicher&Schuell. Dassel, Germany) by standard methods (Sambrook

et al., 1989). Total cDNA from LS and SS forms was labelled with

[a-32P]dCTP to high specific activity by the random priming method

(Feinberg and Vogelstein, 1983) using a Multiprime labelling kit

(Amersham, Aylesbury, UK). Hybridisations and stringent washing of

the filters were performed by standard procedures (Sambrook et al.,

1989). Probe removal for additional hybridisations was achieved by

treatment of the filter for 10 min with 0.5 M NaOH at room temper-

ature. followed by extensive washing of the filter in H20.

56

Murphy, 1993). The fingerprints within each set of four

PCRs, representing the different forms examined (LS, Int,

SS and procyclic) were similar with 6 out of 10 primers

(Fig. lA, primers 1 to 6). Some primers generated highly

variable fingerprints in which no common bands were

evident between the products from the different trypano-

some forms tested (e.g., Fig. lA, primer 7). Such primers

do not usually generate reproducible fingerprints in

repeat experiments, even with the same template cDNA

and under apparently identical conditions (data not

shown). Other primers showed numerous polymorphisms

between the different cDNAs, but also contained at least

one common band for all forms tested (Fig. lA, primers

8, 9 and 10). These primers usually produced consistent

fingerprints in repeat experiments (data not shown).

Finally, in some reactions, differences in the overall inten-

sity of bands, rather than the intensity of individual

bands, were apparent (e.g., in Fig. lA, lane 2 of primer 6

and lane 1 of primer 10, there is very weak amplification).

These differences were generally accounted for by slight

pipetting differences of target cDNA or oligo primers, or

by the position of the tube in the heating block of the

thermal cycler. These anomalies could be rectified in

repeat experiments. The majority of the primers tested

generated similar and reproducible fingerprints for all the

forms (Fig. lA, primers l-6 and 8). However, a variable

number of differences between cDNAs of the different

forms were observed, dependent on the primer used, and

initial analysis suggest that as many as 70% of primers

yielded at least one band which varied in the different

forms. These differences are of particular interest, since

they should represent RADES products generated from

differentially expressed sequences.

(b) Verification of the specificity of RADES-PCR

products

It was possible that the products of the RADES-PCR

were generated from the most highly expressed genes

from each trypanosome form and therefore the technique

would not be suitable for the identification of genes ex-

pressed at low levels. In addition, the differences between

the fingerprints generated from cDNA of the different

forms may have been due to artefacts of the reaction. To

verify the specificity of the products, a Southern blot of

the agarose gel shown in Fig. 1A was generated and

probed with 32P-labelled cDNA from LS forms (Fig. 1B)

and SS forms (Fig. 1C). It is evident from this experiment

that the majority of the products visualised in the EtdBr-

stained gel do not generate strong hybridisation signals

with 32P-labelled cDNA. Therefore, the technique does

not select for only highly expressed genes. Further, the

pattern of hybridising fragments confirms the specificity

of the PCR reaction, since products specifically amplified

in one form preferentially hybridise to the labelled cDNA

from that form (Fig. 1B and C, filled arrowheads). In

some cases, hybridisation was observed to fragments not

visible in the EtdBr-stained gel (Fig. lB, primer 10, lane

1, open circle); this may be due to weak priming and poor

amplification of a relatively abundant cDNA.

Although the majority of fragments specifically ampli-

fied in one form have been found to have a higher level

of expression in that form, there were a small number of

false positive products. This is illustrated by a hybridisa-

tion signal with 32P-labelled cDNA from SS forms

(Fig. lC, open arrow, primer 9, lane 1) to a product which

appears to be specific for LS forms in the EtdBr-stained

gel (Fig 1 A, square bracket, primer 9, lane 1 ), but which

fails to show a detectable hybridisation signal with 32P-

labelled cDNA from LS forms.

As stated earlier, the number of amplified products for

each primer appears to be relatively few. However, some

bands in the EtdBr-stained gel were found to contain

more than one amplified product. For example, primer 8

generated a product of approx. 300 bp from each form

(Fig. 1 A, primer 8, lanes l&4), yet the hybridisation signal

with 32P-labelled cDNA from both LS and SS forms is

only evident for this sized product generated from the LS

form cDNA (Fig. 1 B, primer 8, lane 1, asterisk). The use

of high resolution gels, such as higher percentage

MetaphorTM agarose (FMC, Rockland, ME, USA) or

polyacrylamide, should improve the separation of such

co-migrating fragments.

It is of interest that some products specifically ampli-

fied in procyclic forms hybridise to 32P-labelled cDNA

from SS but not LS forms (compare Fig. 1B and lC,

primers 7 and 9, lane 4, open boxes). There are several

biochemical properties of SS forms which would suggest

that they are pre-adapted for transformation to procyclic

forms (Turner et al., 1988) and our results would appear

to support these conclusions.

(c) Northern blot and sequence analysis of RADES

products

Further verification that a RADES product specific for

one form is due to expression or up-regulation of a tran-

script in that form only was obtained by Northern blot

analysis of poly(A)-enriched RNA from the different

forms. Stage-specific RADES products were purified from

the agarose gels and 32P-labelled to high specific activity.

Fig. 2A shows a Northern blot of LS, SS and procyclic

forms (lanes 1-3) hybridised with a LS-specific product

(panel I), a SS product, (panel II) and a Tub-gene probe

(panel III), the latter as a control for the amount of RNA

loaded. Fig. 2B (panel I) shows a second Northern blot

hybridised with an Int-specific RADES product.

Although the level of stable transcripts for this gene is

A

II III

123 123

57

B

I II

1234 1234

C

I II

12 12

I

123

Fig. 2. Northern blot analysis of mRNA from the different forms of T.h.h. ILTatl.1 hybridised with differentially amplified RADES products. RADES

products identified as being specifically amplified only from cDNA of one form of T.b.b. lLTatl.l were purified from the 1.5% agarose gels using a

GenecleanTM kit (BIO 101, La Jolla, CA, USA), and labelled to high specific activity, as outlined in the legend of Fig. 1, for hybridisation to Northern

blots prepared as previously described (Pelle and Murphy, 1993a; 1.4% agarose gel). (A) Hybrisation to a Northern blot of 1, LS; 2, SS; and 3,

procylic poly(A)-enriched RNA, of RADES products specifically amplified from (I) LS cDNA and (II) SS cDNA. The blot was hybridised with Tub

(panel III), to control for the amount of mRNA loaded. Hashed lines show the positions of an RNA size marker (GIBCO-BRL) of 9.5. 7.5, 4.4, 2.4,

1.4 and 0.24 kb. (B) Hybridisation to a Northern blot of 1, LS; 2, Int; 3, SS; and 4 procyclic mRNA with (I) an Int-specific RADES product and (II)

Tub. (C) Northern blot hybridisation with (I) a procyclic-specific RADES product and (II) Tub to 1, BSF mRNA and 2. procyclic mRNA.

relatively low in all forms, there is clear up-regulation in

Int forms, suggesting that there may be transient, but

specific, up-regulation of some genes in the switch from

LS to SS forms. Finally, an example of the level of stable

transcripts for a procyclic-specific RADES product in

BSF (lane l), and procyclic forms (lane 2) is given in

Fig. 2C. Although transcripts for this gene are easily de-

tected in BSF, the level of stable transcripts is approx.

five-fold higher in procyclic forms (compare the Tub

probe to the RADES probe) and two transcript size

classes are evident in procyclic forms. Five additional

products have been tested, but generated only faint or

undetectable signals on Northern blots (data not shown).

Verification that these products are also differentially ex-

pressed will require more sensitive RNase protection or

quantitative PCR experiments. However, the relative

specificity of the signals obtained from the hybridisation

experiments with labelled cDNA (Fig. 1A and 1B) would

suggest that the number of false positive products are low.

To determine whether any of the products of

differentially expressed genes we have identified have pre-

viously been characterised, DNA sequence analysis of the

LS, SS and procyclic products used to probe the

Northern blots shown in Fig. 2 was undertaken,

following cloning of these products into the EcoRV site

of pBluescript (Stratagene, La Jolla, CA, USA). A search

of both the GenBank database (release 77) and the

EMBL database (release 35) identified the LS-specific

gene as glyceraldehyde-3-phosphate dehydrogenase

(GAPDH) (Michels et al., 1986); no homologues of the

other products were found. A more detailed description

of these, and other developmentally regulated genes iden-

tified in this work, will be reported elsewhere. It is interes-

ting that the GAPDH gene is down-regulated both in SS

and procyclic forms (see Fig. 2A, panel I), whereas a pre-

vious report has identified this gene as BSF specific

(Bakalara et al., 1991). This provides further evidence for

SS forms being pre-adapted for transformation to the

insect procyclic form.

(d) RADES-PCR of total amplified cDNA from small

numbers of organisms at various time-points during

semi-synchronous cell division

In the preceding sections, the RADES products were

generated from different forms of T.b.b. which can be

isolated in relatively large numbers and, therefore, are

easily analysed in this system. In many cases, however,

there is a need to analyse differential gene expression in

systems where limited numbers of cells are available. For

example, some insect stages of trypanosomes, namely epi-

mastigote and metacyclic forms, are difficult to propagate

in sufficient numbers for molecular studies, and therefore,

very little information is available about these forms.

Systems for the propagation of epimastigote and meta-

58

cyclic forms of T. congolense (T.c.) IL3000 have pre-

viously been described (Bienen et al., 1991; Fish et al.,

1989) and in vitro cultivation of these parasites as BSF

can now be achieved in the absence of feeder-cell layers

(Hirumi and Hirumi, 1991). Mature metacyclic forms of

T.c. IL3000 derived in vitro were isolated by diethylamin-

oethyl-cellulose (DE-52, Beckman) column chromatogra-

phy (Lanham and Godfrey, 1970), incubated in insect

form media (Bienen et al., 1991; Fish et al., 1989) at 27°C

overnight and then transferred to an axenic culture

system for T.c. BSF (Hirumi and Hirumi, 1991) at 34°C.

Giemsa and acridine orange staining of parasites taken

at different time points after the media and temperature

change indicated that they were dividing in a semi-

synchronous manner for the first cell division, since more

than 70% of cells displayed two nuclei at 10 h, whereas

only 5% displayed two nuclei at 12 h. In actively dividing

control cultures approx. 20% of cells had two nuclei.

Samples of 2 x lo6 trypanosomes were taken at 0, 2,4, 6,

10 and 12 h following initiation of the bloodstream-form

cultures. Three additional samples were taken from cul-

tures initiated with a different batch of metacyclic forms,

at 0, 4 and 8 h following initiation of the BSF cultures,

in order to test the reproducibility of results obtained

between different samples. Total RNA was purified from

lo6 parasites isolated at each time-point and mRNA was

reverse transcribed into ss cDNA using an oligo(dT)

primer containing an AscI restriction site (IL01487).

Since trypanosome mRNAs contain a common 5’ se-

quence, called the mini-exon or spliced leader sequence

(De Lange et al., 1984; Parsons et al., 1984), and the

different species of African trypanosomes in which this

gene has been characterised have complete homology for

the 25 nt at the 3’ end of this sequence (Cook and

Donelson, 1987; De Lange et al., 1984; Parsons et al.,

1984) a primer (ILO1488), also containing an AscI re-

striction site, was designed which contains nt 16 to 39 of

the mini-exon sequence. Half the ss cDNA generated

from trypanosomes isolated at different time points

following initiation of BSF cultures was used as a tem-

plate in the PCR; IL01487 and IL01488 were used as

primers to amplify the total cDNA from these samples.

The design of these primers facilitates digestion of the

total amplified cDNA with the rare-cutting restriction

enzyme AscI to generate ends compatible for cloning of

the PCR products into plasmid vectors such as pNEB193

(New England Biolabs, Beverly, MA, USA). The primers

and dNTPs were removed by filtration through

Centricon 30 columns (Amicon) and the amplified cDNA

was analysed by agarose-gel electrophoresis. The major-

ity of products ranged in size between 0.5 and 1 kb and

approx. 10 ug of total ds cDNA was generated from each

time point (data not shown). Aliquots of this cDNA were

diluted to 10 ng/ul and used as templates in the RADES-

PCR. Products of the reactions were then analysed by

agarose-gel electrophoresis as before. The RADES fin-

gerprints from parasites isolated on different days and

from different starting material were reproducible and

directly comparable (Fig. 3A, B and C, compare finger-

prints of lanes 1 and 2 and of lanes 4 and 5). Of particular

interest is the fact that small alterations are evident in

the RADES products from parasites collected at different

time points, indicating temporal regulation of some genes

at specific points during this first cell-division cycle. For

primer IL0508 (Fig. 3A) a product of approx. 260 bp

appeared at 6 h and further increased up to the 10 and

12 h time points. A second, but less abundant, product

of approx. 350 bp was also evident at the 10 and 12 h

time points. Amplification of the cDNA at the 8 h time

point failed in this particular experiment, but was success-

ful in repeat experiments (data not shown). For primer

ILO868, the RADES products of all time points are iden-

tical, except that an additional product is observed at 10

h representing temporal up-regulation of a gene at this

time point (Fig. 3B). Primer IL0872 (Fig. 3C) generated

the most polymorphic set of RADES prints in this series

of experiments, with loss of a product of approx. 300 bp

at the 6 and 8 h time points (Fig 3C, upper band of the

doublet is not amplified), and a significant decrease in a

product of approx. 500 bp only at the 10 h time point.

The transient increase and decrease of some RADES pro-

ducts at specific time points suggests that the majority of

the differences observed reflect cell-cycle differences

rather than differences between the two life-cycle stages.

If subtractive enrichment procedures had been used to

identify genes which are specifically expressed at each

time point in these experiments, a total of nine subtractive

enrichments between each individual time point, and all

the others combined, would have been required.

Furthermore, if this method had been used for a direct

comparison between all of the time points to identify

differentially expressed genes, a total of 72 standard

subtractive enrichments would have been required. Even

if there was no limit to the amount of material available,

such an undertaking would clearly be difficult since each

successful subtractive enrichment experiment can take

approx. 2 months before clones of interest are identified.

With the RADES technique, as many as 12 different

samples, each primed with 8 different oligos, can be ac-

commodated in a 96-well PCR machine. Gel analysis of

the resultant products can be carried out on the same

day. The purification and ligation of products of interest

into an appropriate vector can also be carried out on the

same day, and sequence analysis completed within 1

week. In addition, the requirement for large numbers of

cells is no longer a limiting factor as each profile in this

59

Fig. 3. RADES-PCR on amplified total cDNA generated from small numbers of TX. IL3000 parasites isolated at different time-points during the

first cell-division cycle, when transforming from metacyclic forms to BSF. Total amplified cDNA from TX. IL3000 parasites collected at 0, 2, 4. 6, 10

and 12 h (lanes 1, 3, 4, 6, 8, 9) following transfer of one batch of metacyclic forms to BSF conditions and 0, 4 and 8 h (lanes 2, 5 and 7) following

transfer of a second batch. Primers used in the above examples were: (A) ILO508, (B) IL0868 and (C) ILO872. Size markers (123 bp ladder, GIBCO-

BRL; hashed lines), PCR and electrophoresis conditions were the same as for Fig. 1. Methods: Total RNA was isolated from cell pellets in 100 pl of

lysis buffer as outlined in Fig. 1. Polyadenylated transcripts were reverse transcribed to ss cDNA with M-MLV reverse transcriptase (GIBCO-BRL)

and IL01487 (S-TAGGCGCGCC[T],,) as primer, under standard conditions (Sambrook et al., 1989). Total cDNA was amplified with IL01487

and a mini-exon primer, IL01488 (5’-TAGGCGCGCCTAGAACAGTTTCTGTACTATATTG), in the same PCR buffer described in Fig. 1. except

the volume was 100 ul and cycling conditions were 94°C for 1 min, 55’C for 1 min and 72°C for 2 min for 40 cycles followed by a 5 min extension

at 72°C. Both the oligo(dT) and miniexon primers contain an AscI site (5’-GGCGCGCC), preceeded by TA so that amplified cDNA could be

restricted by AscI and cloned into pNEBl93. The mini-exon primer is universal for African trypanosomes and, in addition to the AscI site, contains

nt 16 to 39 of the mini-exon sequence which is identical for all the African parasites characterised to date. Following removal of the primers, dNTPs

and buffer, cDNA was diluted to 10 ng/ul and the RADES-PCR conditions were as given in Fig. 1.

study represents the products generated from approx. lo3

trypanosomes only. Additional rounds of amplification

could reduce this number even further. Since the mRNA

content of mammalian cells is ten times greater than try-

panosomes, even fewer mammalian cells would be re-

quired to generate a RADES fingerprint. Finally, we

believe that the ease with which we have identified devel-

opmentally regulated genes through the RADES tech-

nique will encourage its exploitation for studies on

differential gene expression, not only in trypanosomatids,

but also in other eukaryotes. Further use of the RADES-

PCR technique should lead to modifications, such as

more simple cloning procedures of the products (Boyd,

1993) which will extend its application and simplify

analyses of the products generated.

(e) Conclusions

(I ) RADES-PCR is a quick and cost-effective alterna-

tive to subtractive enrichment methods for the isolation

of developmentally regulated sequences and can be ap-

plied simultaneously to many samples. Although some

false-positive products can be generated, our experience

is that the technique is reproducible and less prone to

artefacts than subtractive enrichement methods, provided

care is taken to replicate the times between temperature

shifts in the PCR (Penner et al., 1993; data not shown).

(2) There are an almost limitless number of primers

and combinations of primers, including oligo(dT), which

can be used to generate RADES fingerprints. Thus, prim-

ers which do not generate fingerprints, or which show

high variability, can be quickly discarded.

(3) Since RADES is a PCR-based method and gener-

ates reproducible results, even after amplification of total

cDNA from small numbers of organisms, it is particularly

applicable to systems where cell numbers are limiting.

Theoretically it could be performed on cDNA generated

from single cells (Belyavski et al., 1989; Froussard, 1993).

(4) RADES-PCR, unlike differential screening and

subtractive enrichment methods, is not selective for mod-

erately to highly expressed sequences. Most products,

even those clearly abundant in the EtdBr-stained gels,

showed weak or no hybridisation to labelled cDNA. In

addition, RADES can differentially amplify sequences

which are up-regulated in one cell type, and is not limited

to sequences present in one cell type and absent in an-

other. The method is therefore applicable in systems

where lowly abundant, differentially-expressed sequences

are involved in the control of the differentiated state.

(5) Like other AP-PCR methods, RADES primers can

be used for any organism without prior sequence infor-

mation of the products to be amplified.

(6) RADES allows for rapid cloning and sequence

analysis of differentially expressed sequences which

should lead to the facile generation of databases of genes

differentially expressed during both the cell-division cycle

and in differentiated cells.

(7) The RADES method is relatively simple. Many

laboratories now routinely use PCR methods, whereas

60

subtractive enrichment methods are labour-intensive and

require a greater level of experience in molecular biology

techniques and, therefore, are not as easily applied.

(8) Verification that differentially amplified products

are derived from differentially expressed sequences can

be generated by rapid Northern blotting (Pellt and

Murphy, 1993a), or hybridisation with total labelled

cDNA which can reveal additional polymorphisms.

(9) Surprisingly, our results show that the variability

in gene expression among the different BSF of T.b.b. was

as great as between bloodstream and procyclic (insect)

forms.

ACKNOWLEDGEMENTS

The authors are indebted to Drs. W.R. Fish, A.S.

Peregrine and R. Bishop for their critical reading of and

suggestions on this manuscript. Our thanks to Dr. and

Mrs. Hirumi for in vitro propagated parasites and to Mr.

S. Wasike for provision of T.b. brucei parasites. This is

ILRAD publication number 1206.

REFERENCES

Ashcroft, M.T.: The polymorphism of Tr_rpanosomu brucei and

Trypunosoma rhodesiense, its relationship to relapses and remissions

of infections in white rats, and the effect of cortisone. Ann. Trop.

Med. Parasitol. 31 (1957) 301-312.

Bakalara, N., Kendall, G., Michels, P.A.M. and Opperdoes, F.R.:

Trypanosoma brucei glyceraldehyde-3-phosphate dehydrogenase

genes are stage-regulated at the transcriptional level. EMBO J. 10

(1991) 3861-3868.

Balber, A.E.: Trypunosoma hrucei: fluxes of the morphological variants

in intact and x-irradiated mice. Exp. Parasitol. 31 (1972) 3077319.

Belyavsky, A.. Vinogradova, T. and Rajewsky, K.: PCR-based cDNA

library construction: general cDNA libraries at the level of a few

cells. Nucleic Acids Res. 17 (1989) 2919-2932.

Bienen, E.J., Webster, P. and Fish, W.R.: Trypanosomu (Nunnomomas)

congolense: Changes in respiratory metabolism during the life cycle.

Exp. Parasitol. 73 ( 1991) 403-4 12.

Black, S.J., Sendashonga, C.N., O’Brien, C., Borowy. N.K., Naessens.

J., Webster, P. and Murray, M.: Regulation of parasitaemia in mice

infected with Trypnnosoma brucei. Curr. Top. Microbial. Immunol.

117(1985)933118.

Boyd. A.C.: Turbo cloning: a fast, efficient method for cloning PCR

products and other blunt-ended DNA fragments into plasmids.

Nucleic Acids Res. 21 (1993) 817~821.

Chomczynski, P. and Sacchi, N.: Single-step method of RNA isolation

by acid guanidinium thiocyanate-phenol-chloroform extraction.

Anal. Biochem. 162 ( 1987) 156-l 59.

Cook, G.A. and Donelson. J.E.: Mini-exon gene repeats of Trypunosomu (Nwmomonus) congolense have internal repeats of 190 base pairs.

Mol. Biochem. Parasitol. 25 (1987) 113-122.

De Lange, T., Berkvens, T.M., Veerman, H.J., Frasch, A.C., Barry, J.D.

and Borst, P.: Comparison of the genes coding for the common 5’

terminal sequence of messenger RNAs in three trypanosome species.

Nucleic Acids Res. 12 (1984) 4431-4443.

Dirie, M.F., Murphy, N.B. and Gardiner, P.R.: DNA fingerprinting of

Trypunosomn Cux isolates rapidly identifies intraspecific relation-

ships. J. Euk. Microbial. 40 (1993) 1322134.

Feinberg, A.P. and Vogelstein, B.: A technique for radiolabeling DNA

restriction endonuclease fragments to high specific activity. Anal.

Biochem. 132 (1983) 6613.

Fish, W.R., Muriuki, C., Muthiani, A.M., Grab, D.J. and Lonsdale-

Eccles, J.D.: Disulfide bond involvement in the maintenance of the

cryptic nature of the cross-reacting determinant of metacyclic forms

of Trypanosoma congolense. Biochemistry 28 (1989) 541555421.

Froussard, P.: A random-PCR method (rPCR) to construct whole

cDNA library from low amounts of RNA. Nucleic Acids Res. 20

(1993) 2900.

Hara, E.. Kato, T., Nakada, S., Sekiya, S. and Oda, K.: Subtractive

cDNA cloning using oligo(dT),,-latex and PCR: isolation of cDNA

clones specific to undifferentiated human embryonal carcinoma

cells. Nucleic Acids Res. 19 (1991) 709777104.

Hirumi, H. and Hirumi, K.: In vitro cultivation of Trypunosomu con-

golense bloodstream forms in the absence of feeder cell layers,

Parasitology 102 (1991) 2255236.

Houge, G.: Simplified construction of a subtracted cDNA library using

asymmetric PCR. PCR Methods Appl. 2 (1993) 2044209.

Lanham, S.M. and Godfrey, D.G.: Isolation of salivarian trypanosomes

from man and other mammals using DEAE-cellulose. Exp.

Parasitol. 28 (1970) 521-534.

Lee, SW.. Tomasetto, C. and Sager, R.: Positive selection of tumor-

suppressor genes by subtractive hybridization. Proc. Natl. Acad. Sci.

USA 88 (1991) 282552829.

Liang, P. and Pardee. A.B.: Differential display of eukaryotic RNA by

means of the polymerase chain reaction. Science 257 ( 1992) 9677971.

Michels, P.A.M.. Poliszczak, A., Osinga, K.A., Misset, 0.. Van Beeumen,

J.. Wierenga. R.K.. Borst, P. and Opperdoes, F.R.: Two tandemly

linked identical genes code for the glycosomal glyceraldehyde-

phosphate dehydrogenase in Trypanosornc~ brucei. EMBO J. 5 ( 1986)

104991056.

Nedivi, E., Hevroni, D., Naot, D., Israeli, D. and Citri, Y.: Numerous

candidate plasticity-related genes revealed by differential cDNA

cloning. Nature 363 (1993) 718-722.

Parsons, M., Nelson, R.G., Watkins, K.P. and Agabian, N.:

Trypanosome mRNAs share a common 5’ spliced leader sequence.

Cell 38 ( 1984) 3099316.

Pellt, R. and Murphy, N.B.: Northern hybridization: rapid and simple

electrophoretic conditions. Nucleic Acids Res. 21 (1993a)

278332784.

PellC, R. and Murphy, N.B.: Stage-specific differential polyadenylation

of mini-exon derived RNA in African trypanosomes. Mol. Biochem.

Parasitol. 59 (1993b) 2777286.

Penner, G.A., Bush, A., Wise, R.. Kim, W., Domier, L., Kasha, K.,

Laroche, A., Stoles, G., Molnar, S.J. and Fedak, G.: Reproducibility

of random amplified polymorphic DNA (RAPD) analysis among

laboratories. PCR Methods Appl. 2 (1993) 341-345.

Sambrook. J.. Fritsch, E.F. and Maniatis, T.: Molecular Cloning: A

Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold

Spring Harbor, NY, 1989.

Seed, J.R. and Sechelski, J.B.: Mechanism of long slender (LS) to short

stumpy (SS) transformation in the African trypanosomes.

J. Protozool. 36 (1989) 572-577.

Sendashonga, C.N. and Black, S.J.: Humoral responses against

Trypanosornu hrucri variable surface antigen are induced by degen-

erating parasites. Parasite Immunol. 4 (1982) 2455257.

Shapiro, S.Z., Naessens, J., Leisegang, B., Moloo, S.K. and Magondu,

J.: Analysis by flow cytometry of DNA synthesis during the life cycle

of African trypanosomes. Acta Trop. 41 ( 1984) 3 13-323.

Tibayrenc, M., Neubauer, K., Barnabe, C., Guerrini, F., Skarecky, D.

and Ayala, F.J.: Genetic characterization of six parasitic protozoa:

61

parity between random-primer DNA typing and multilocus enzyme

electrophoresis. Proc. Natl. Acad. Sci. USA 90 (1993) 133551339.

Turner, C.M.R., Barry, J.D. and Vickerman, K.: Loss of variable antigen

during transformation of Trypanosoma brucei rhodesiense from

bloodstream to procyclic forms in the tsetse fly. Parasitol. Res. 74

(1988) 507-511.

Vickerman, K.: Developmental cycles and biology of pathogenic trypa-

nosomes. Br. Med. Bull. 41 (1985) 1055114.

Vickerman, K.: Morphological and physiological considerations of ex-

tracellular blood protozoa. In: Fallis, A.M. (Ed.), Ecology and

Physiology of Parasites. University of Toronto Press. Toronto,

1971, pp. 58-91.

Waitumbi, J.N. and Murphy, N.B.: Inter- and intra-species differentia-

tion of trypanosomes by genomic fingerprinting with arbitrary prim-

ers. Mol. Biochem. Parasitol. 58 (1993) 181-186.

Wang, Z. and Brown, D.D.: A gene expression screen. Proc. Nat]. Acad.

Sci. USA 88 (1991) 11505511509.

Welcher, A.A., Torres, A.R. and Ward, D.C.: Selective enrichment of

specific DNA, cDNA and RNA sequences using biotinylated probes,

avidin and copper-chelate agarose. Nucleic Acids Res. 14 (1986)

10027710044.

Welsh, J., Chada, K., Dalai, S.S., Cheng, R., Ralph, D. and McClelland,

M.: Arbitrarily primed PCR fingerprinting of RNA. Nucleic Acids

Res. 20 (1992) 496554970.

Welsh, J. and McClelland, M.: Fingerprinting genomes using PCR with

arbitrary primers. Nucleic Acids Res. 18 ( 1990) 721337218.

Wijers, D.J.B.: Polymorphism in Trypumsomu gm~hiensr and

Trypunosomu rhodrsiense and the significance of intermediate forms.

Ann. Trop. Med. Parasitol. 53 (1959) 59-65.

Williams, J.G.K., Kubelik, A.R., Livak. K.J., Rafalski. J.A. and Tingey.

S.V.: DNA polymorphisms amplified by arbitrary primers are useful

as genetic markers. Nucleic Acids Res. 18 ( 1990) 6531-6535.