acid amplification schemes that may have poten- 
tial for the design of rapid, automatable diagnostic 
assays. 
The work of Dr. Lizardi is concerned with exploit- 
ing RNA amplification as a tool for the development 
of rapid and sensitive diagnostic assays for patho- 
gens. Dr. Lizardi's work in Mexico is being carried 
out in collaboration with Dr. Fred Kramer (Public 
Health Research Institute, New York City) and Dr. 
Jack Szostak (Massachusetts General Hospital, 
Boston) . 
Amplifiable RNA Probes 
A specific class of RNA molecules known as repli- 
catable RNAs can be amplified exponentially at a 
constant temperature of 37°C, generating millions 
of copies of each molecule. The best-characterized 
RNA amplification reaction, catalyzed by the en- 
zyme Q-beta replicase, is unique because parent 
and daughter RNA single strands are forced apart as 
they are synthesized, in contrast to DNA-dependent 
polymerization reactions, in which the two strands 
remain annealed. 
Assays have been designed in which replicatable 
RNAs are used as reporters for the presence of nu- 
cleic acid targets. In such assays a replicatable RNA 
containing a probe insert is incubated with a biologi- 
cal sample under suitable hybridization conditions 
that allow specific binding of the probe domain to 
its intended target. A subsequent washing step re- 
moves those probes that did not bind to the target. 
Finally, the bound probes are released from their 
target and incubated in the presence of Q-beta repli- 
case. The ensuing replication reaction generates as 
many as 100 million copies of each RNA probe in 
~20 min, and the resulting RNA mass is readily 
quantitated by fluorescence staining. The intensity 
of the fluorescent signal is proportional to the num- 
ber of targets present in the original sample. 
The scheme outlined above, known as RNA probe 
amplification, has been used to detect HIV-1 (hu- 
man immunodeficiency virus type 1) RNA targets in 
human blood, and has been demonstrated to have a 
limit of detection of ~ 10,000 molecules of viral 
nucleic acid. This limit is artificially imposed by the 
presence of replicatable RNA probes that bind non- 
specifically to surfaces in the assay medium. 
RNA Binary Probes That Can Be Amplified 
After Ligation 
It should be possible to improve significantly the 
limit of detection in Q-beta-amplified assays by us- 
ing probes that cannot be amplified by the replicase 
until an additional enzymatic step involving target 
recognition has occurred. This requirement can be 
met by using a binary probe scheme, in which a 
ligase-mediated joining reaction provides additional 
specificity in target recognition. 
Using HIV-1 integrase mRNA as a model target, it 
has recently been possible to implement a binary 
probe system that generates an amplifiable signal 
only in the presence of the viral mRNA target. The 
binary probe system consists of two RNA molecules 
that contain probe segments complementary to the 
target sequence domain, as well as additional se- 
quences derived from a replicatable RNA. Neither of 
the two binary probe molecules is replicatable; how- 
ever, they can become replicatable if joined cova- 
lently to form a single molecule. The RNA probe 
segments are designed to bind to the integrase 
mRNA target so that their 3' and 5' ends are perfectly 
juxtaposed. After binding, the ends can be joined in 
a reaction catalyzed by a novel RNA enzyme known 
as a ribozyme ligase, which was constructed in Dr. 
Szostak's laboratory by modification of the type I 
intron of Tetrahymena. This enzyme catalyzes co- 
valent joining of RNA termini that are perfectly 
aligned by base-pairing on a complementary 
polynucleotide. 
A number of technical problems remain to be 
solved to achieve optimal sensitivity and specificity 
in the binary probe assay. The ligation step involv- 
ing the ribozyme ligase is still relatively inefficient, 
since only 25% of the ligation-competent binary 
probe molecules are joined in a 1-h incubation. An 
actual assay would require a ligation efficiency of 
~75%. Recent progress in the ability to generate 
novel RNA enzymes suggests a possible solution to 
this problem. 
Directed Evolution of More-Efficient 
Ribozymes by in Vitro Selection 
It is now possible to generate novel ribozymes 
with altered catalytic efficiency by directed evolu- 
tion in vitro. Chemical synthesis is used to generate 
a mutant pool of 10'^ molecules of DNA, each con- 
taining a sequence variant of the ribozyme. T7 RNA 
polymerase is then used to generate RNA copies of 
the DNA, thus creating more than 10'^ different ri- 
bozyme species. The ribozymes are incubated for a 
short time in a test tube under RNA ligation condi- 
tions. Those ribozymes that succeed in carrying out 
the ligation reaction are amplified by subsequent 
cycles of reverse transcription followed by PCR, and 
the entire process is repeated several times. After 
four cycles of selection, a relatively small number of 
mutant ribozyme sequences are present in the DNA 
population, instead of the original 10'^. The mole- 
INTERNATIONAL RESEARCH SCHOLARS 
519 
