By David N. LeffScience Editor

Before siccing a bloodhound on the trail of a wantedperson, you have to prime the animal's olfactory systemwith a whiff of its quarry's body odor. A similar principleapplies in the human cellular immune system. Beforeturning killer T cells loose against tumors, those cytotoxiclymphocytes need to get a sense for their antigenictarget's size and shape.

This immunogenic mating dance begins in a specializedcompartment of the cancerous cell called the endoplasmicreticulum (ER). That organelle is a lacy tracery of tubularmembranes in which the tumor-associated antigen,escorted by ER insertion signaling peptides, meets thereceptor to which it binds. This is a molecule called theClass I Major Histocompatibility Complex (MHC I).

Eventually, the complexed antigen makes rendezvouswith an advance party of T cells, which normally hangout on the tumor cell's surface membrane. It's the firstcontact the CD8-positive cytotoxic lymphocytes havewith their antigenic prey.

That recognition triggers the killer T cell to churn out azillion antigen-specific copies of itself, with which to takeon and kill off the tumor cells. That, at least, is thestrategy being pursued by researchers constructingvaccines based on cytotoxic lymphocytes of the body'scell-mediated immune defenses, rather than the better-known antibodies of the humoral immune system.

That's the natural scenario to which vaccinologistNicholas Restifo and his colleagues at National Institutesof Health's (NIH) National Cancer Institute (NCI) haveapplied genetic engineering strategies to develop T-cell-activating vaccines against cancers, infections andautoimmune diseases.

Earlier this month, the NIH Office of TechnologyTransfer forwarded a notice to the Federal Register,inviting biotechnology and pharmaceutical companies tolicense the key invention that Restifo, its first author, andhis fellow inventors are patenting. Its title describes theirmodus operandi:

"Immunogenic chimeras comprising nucleic acidsequences encoding endoplasmic reticulum signalsequence peptides and at least one other peptide, and theiruses in vaccines and disease treatments."

The key to their approach, Restifo explained to BioWorldToday, is "that in the human cellular immune system, Tlymphocytes don't recognize whole antigens, wholepieces of cancer cells _ or whole viral antigens, for thatmatter. Rather," he added, "they recognize short, epitopicfragments about nine amino acids long."

So he and his team at the NCI's Surgical Branch trimmeddown tumor-associated antigens to their minimalimmunogenic length of nine amino acids. Then, adjoiningthe 27-base-pair cDNA sequence for these mini-epitopes,they ligated DNA sequences encoding the ER insertionsignaling peptides.

They cloned this dual-duty chimeric DNA, whichexpressed both the minimal antigen and its joined-at-the-hip ER signaling chaperone, into recombinant vacciniavirus vectors, and tested this prototype vaccine in miceand in human volunteers.

"In our mouse experiments," Restifo said, "using modeltumor antigens, such vaccine constructs have elicitedkiller T cell immune responses such as cause tumors toregress in humans."

He continued, "We are now attempting to make bettermurine immune responses against a variety of epitopes,including those from influenza and vesicular stomatitisviruses, as well as tumors." Restifo added: "We haven'ttried HIV yet, but that's in our sights as well _ perhapsfor an eventual licensee."

In their preliminary human trials, the vaccinologist wenton, "we took the gene encoding an antigen that we knowis recognized by anti-tumor lymphocytes and gave it,within the last year, to immunize three patients withmalignant melanoma at the NCI clinical center."Although this trial produced no therapeutic response, itdid encourage the group to plan repeating it in the nearfuture with other potentially more powerful antigenicsequences.

"Use of minimal-determinant constructs," Restifoobserved, "may have the advantage of avoiding thepotential dangers of immunizing with proteins of knownor potential oncogenic function, such as ras or p53, ifpresent in full length."

Now the NCI is offering to license its pending U.S. andinternational patents protecting this strategy. Restifo seestwo areas a licensee might profitably pursue: "One is thedesign of recombinant vaccines to treat infectious diseaseor cancer.

"But," he continued, "there's another area that we think itmight be useful for _ the selective shutting off of thecell-mediated immune system."

He explained: "This is just a way to stimulate T cells. Butthey can also be stimulated in a way that will shut themoff. The on/off switch is the same switch. So theinvention may be applicable to autoimmune diseases _for example, multiple sclerosis or diabetes _ as well."

Editor's note: For patent licensing information, consultRobert Benson, NIH office of Technology Transfer, (301)496-7056, extension 267; for technical questionsconcerning this invention, call Nicholas Restifo, NCISurgical Branch, (301) 496-4904. n

(c) 1997 American Health Consultants. All rights reserved.