Thank you for sending us a copy of the appraisal of our application for renewal of contract #71-2147. Needless to say, we
were pleased by the kind words about our past performance and future promise, and appreciative of your support. We feel constrained,
however, since you have solicited our comments, to make some rejoinders to the remarks about our use of reassociation kinetics
to detect viral sequences in cells.
First, the design of these experiments is to observe the effect of unlabeled cell DMA upon the renaturation kinetics of labeled
viral polymerase product, not the reverse as was stated by the reviewer. We have found about 10-15 copies of DNA representing
at least 25% of avian tumor virus 70S genomes in normal chick cells, without a detectable increase In transformed cells. As
you know, Gelb et al. have very similar results with a murine system. It is impossible to compare these numbers, as was attempted,
with the results of Baluda and Nayak for several reasons: (1) by their own admission, "the actual number of copies remains
undetermined" in their experiments; and (2) their claim that four additional viral genomes are present in leukemic chick
cells is subject to considerable skepticism since (a) the number is computed from the weight of RNA annealed, which may not
represent a random sampling of 70S sequences; (b) a very small fraction of the incubated RNA is annealed, again suggesting
that only a mall selected population of sequences is annealed; (c) the available DNA sites are not saturated, implying an
underestimate of the amount of DNA capable of annealing; and, importantly, (d) the hybridization reactions are performed at
relatively low CoT values (not precisely calculable, of course, with filter hybridization techniques) at which only highly
reiterated sequences in the cell genome would be expected to anneal. Most of these criticisms apply to all of the published
DNA-RNA hybridization work in this field; we believe that the only satisfactory approach to the hybridization of 70S RNA and
cell DNA is with the technique of RNA-DNA hybridization in solution at high CoT values as recently described by Melli and
Bishop. Our laboratory and others are currently attempting to apply this technique to the problem of detecting RNA tumor virus
sequences, as outlined in our renewal application.
Our estimates of copy number are jeopardized principally by some uncertainty regarding the precise complexity of the DNA probe,
and, of course, the results are likely somewhat limited because the double-stranded probes may not be representative of the
entire 70S genome. Experiments of the type originally described by Duesberg and Cannani with unfractionated enzymatic produce
are now being performed in our laboratory to test directly the representation of viral sequences in the rapidly and slowly
reannealing fractions of double stranded product. However, the general technique of measuring copies of viral sequences in
cells has been validated in several ways: with the reconstruction experiments using SV-40 DNA, performed by Gelb et al; with
single-strand specific nuclease as a test for duplex formation; with melting curves of cell-product hybrids; with experiments
employing BUdR labeled cell DNA to demonstrate cell-product density hybrids; and with experiments altering cell/probe ratios
to alter the degree of acceleration of reannealing. in addition, reannealing of product in the presence of several heterologous
cell DNA's does not augment the renaturation, providing excellent controls against which to measure copy numbers. The
ability to follow a reaction to its completion assures us that all the DNA is participating in the annealing reaction, particularly
when essayed with single-strand specific nuclease, as well as by elution from hydroxyapatite. Conformation of the data to
theoretical expectations of second-order kinetics further substantiates claims for the validity of this technique.
The reviewer raises the possibility that transcription of viral RNA of cellular origin, particularly free tRNA or 4S RNA in
the 70S complex, may by a "source of error" in our experiments. There are several objections to this point: (1) the
DNA product anneals to purified 70S RNA; (2) it does not hybridize to RNA extracted from normal chick cells; (3) its pattern
of reassociation is too rapid for it to be copied from a heterogeneous population of! cellular RNA (or DNA) molecules (although
the CoT l/2 alone would not exclude transcription from several tRNA species); and (4) it reassociates with cell DNA at high
cell CoT values consistent with ten-fold representation in chick cells, far below the reiteration frequency demonstrated for
The reviewer suggests that we should synthesize our probes with purified polymerase and purified 70S or 35S RNA template.
Aside from the expense of time, money, and materials incurred by this approach, there is no good evidence it would offer any
advantages. In fact, we have recently demonstrated that the product of the "purified" reaction has similar reassociation
kinetics and is homologous to the product of the "crude" reaction. Moreover, we are unable at present to obtain
with the purified reaction that small fraction of slowly reassociating double stranded product, found in the crude reaction,
which presumably allows us to detect a five-fold larger fraction of the genome. Using dT as a primer, as the reviewer suggests,
with native or melted 70S RNA as template, purified polymerase synthesizes product against homologous to and now less complex
than the principal product of the crude reaction.
We want to repeat our thanks for your interest in our research program and hope you will feel free to direct to us any suggestions
or questions you may have.