Picture this bad-dream script:
You're a newspaper photographer, sent out to cover a bigfootball game. You have only one camera, and the onlylens-focus option you can set on it is the close-up, so youcan follow only one player.
You'd have to pick that solitary player from amongmaybe 100 players between both benches or the field. Hemay or may not be in the game. Even if he is, youwouldn't get a very good picture of what was happeningon the gridiron.
You might try to get 100 cameras to follow everything,but that's a lot of work, and it's awfully complex.
Molecular oncologist Kenneth Kinzler draws this analogyto the way gene expression in tissues is analyzed today,and the technology he and his associates have invented toanalyze that gene expression tomorrow.
"Like that make-believe football scenario," Kinzler toldBioWorld Today, "right now, you can look at genes oneat a time pretty readily. What our new technology does isprovide you with the equivalent of a wide-angle lens. Soyou can back off and capture the whole big picture inyour view-finder, see all the players at once, figure outwhich are stars, and which are not."
And, Kinzler pointed out, "As in football, so with geneanalysis, to get a good feel for what's going on, you wantto look at a lot of different games. Thus, we want to lookat a lot of different tissues, normal tissue vs. diseasedtissue. And being able to watch all the players at once,over a lot of different games, we can get a good feel forwhat's going on."
The name of Kinzler's game is SAGE _ Serial Analysisof Gene Expression. That's the title of his paper in thecurrent issue of Science, dated Oct. 20. Kinzler directsthe molecular genetics laboratory at the Program inHuman Genetics at Johns Hopkins University inBaltimore.
It's also the title of a pending patent, which Hopkins hasjust licensed exclusively to PharmaGenics Inc. (PGI), ofAllandale, N.J. Its lead inventors are Kinzler andoncologist Bert Vogelstein, a co-author of the Sciencepaper. Both are members of PGI's scientific advisoryboard, and inventors of a previously licensed patentcovering restoration of the p53 tumor suppressor gene'santi-cancer function.
SAGE's Game Plan: Analyzing Genes Wholesale
"SAGE provides us with a new tool," Kinzler explained,"to examine cancers in a different way. We hope that bycomparing a series of normal tissues to a series of tumors,we may identify genes _ call them players _ who aregoing to be useful for diagnostics or therapeutics."
He added, "Obviously, with about 150,000 genes in thehuman genome, doing the cancer candidates one at a timeis hard. We hope this technology will give us a chance totell the guys apart."
Kinzler defends his 150,000-gene approximation thisway: "When I recently talked to people in the genomefield, and mentioned the customary estimate of 80,000,they pounced on me and said it's more like 150,000, atleast."
To present SAGE in Science, Kinzler and his co-authorschose human pancreatic tissues, healthy and cancerous.Why pancreas? "We wanted something," he explained,"that when we saw the gene expression pattern out to theend, we could be comfortable it was working. And wechose human pancreas," he added, "because, if we couldget it to work in human tissues, then we're in good shapefor everything else."
He defines "everything else" as "any disease where youmight suspect that there could be a difference in patternof expression. For instance, the body's response toinfection involves changes in gene expression, if youassay the right tissue. SAGE could apply to almostanything _ even how people respond to diet."
Nine Base Pairs: 262,144 Separate Sequences
Putting the technology to work begins with RNAtranscripts expressed in the cells of target tissues. "Wedevised a way," Kinzler said, "to extract from the cDNAof each transcript an extremely short, unique, identifyingtag, comparable to a Social Security number." Theyligated these end to end, and cloned them for sequencing.
They excised these brief sequences, each nine base pairslong, of which they analyzed 840 in the case of pancreas,from the same known location on each transcript. "If youcan remove nine base pairs from exactly the same spot inevery transcript," he observed, "then theoretically eachone, with four base possibilities (ACTG) at each position,can distinguish 262,144 disparate sequences."
PharmaGenics' CEO, Michael Sherman, co-invented amethod for screening drugs to rehabilitate p53, alsolicensed from Hopkins. He described to BioWorld Todaya generic example of SAGE at work:
"Let's say you take 10 lung cancer tissue samples, andhalf a dozen normal samples, do the SAGE analysis, thenchoose those genes as targets for diagnosis, therapy ordrug development, that were consistently differentiallyexpressed between both batches.
"That's a perfect example of a tumor suppressor gene,which would be expressed in normal tissues but notmalignant ones. Based on that, you would think aboutsome sort of gene therapy or other approach.
"Conversely," Sherman added, "you might have a gene,like a growth factor, that's expressed in a tumor tissue ata much higher level than in normal. There, you wouldwant to inhibit, either with a small molecule, or antisense,or something like that.
"On an automated sequencer," he said, "SAGE can allowyou to look at 100,000 transcripts, at the rate of 3,000 perperson per sequencer per day."
Armed with Hopkins' latest license, Sherman is "indiscussions with parties interested in a collaborativemode of applying the SAGE procedure."
A kindred paper in this week's Science bears the title:"Quantitative Monitoring of Gene Expression Patternswith a Complementary DNA Microarray." Its seniorauthor is biochemist Patrick Brown of StanfordUniversity's Howard Hughes Medical Institute.
Kinzler sees it as "similar but complementary to SAGE."He explained that "their technique started with clonedgenes, which is a very useful way. The one difference inour situations is that SAGE doesn't have to have thegenes cloned to look at them." n
-- David N. Leff Science Editor
(c) 1997 American Health Consultants. All rights reserved.