29 May 1988 .
Dr. Joshua Lederberg MAY 27' 1938
The Rockefeller University Coy 6 4
123@ York Avenue ce WWF
New York, NY 10021-6399 CF THE PRES.

Dear Dr. Lederberg:

I was pleased to receive your letter and to learn that there is still
interest in the time-honored Weil-Felix reaction. Yes, a carbohydrate is
involved in the cross-reaction between rickettsiae and Proteus antigens. No,
there is no evidence that the two bacteria are otherwise related. Not all the
evidence for these assertions has been published. Described below is sore
detail of the experiments which allow us to cone to these conclusions.

In 1985 Robert Goldwasser spent sone time in our lah. He brought with him
a fair number of patients’ sera from Israel, which we tested for Legionella
antibodies. To our surprise, quite a few reacted with L. bozemanii, a species
which has not been isolated very often in Israel. It turned out that these
sera, but not other sera, reacted also with Rickettsia typhi. In other words,
emdemic typhus infection, which occurs in Israel with moderate frequency,
elicits antibodies not only to rickettsiae but also to L. bozemanii. These
results have been confirmed by appropriate cross-absorption tests. (See Isr J
Med Sci 22: 131-138, 1986.) Several of my colleagues proceeded to identify by
wodern technology the carbohydrate moiety of the LPS responsible for the
cross-reaction between rickettsiae, L. bozemanii, Proteus OX19, and other
bacteria. This work is now being submitted for publication. Still unknown is
the mechanism by which in evolutionary history, the same glycosyl transferase
has appeared in (or transferred to) so many different bacteria.

Brenner et al (J. Bacteriol. 98: 637-650, 1969) showed that there is sone
similarity in polynucleotide sequence between E. coli and Proteus mirabilis and
P. vulgaris. This was recently confirmed in the laboratory of Carl R. Woese
(see his review in Microbiol. Rey. 51: 221-271, 1987), who with his associates
has studied 165 rRNA nucleotide sequences in a great number of bacteria.
Although not specifically stated in the review, Proteus, as other enterics,
belongs in the gamme-subdivision of the purple bacteria. Weisburg and Woese
have extended the 165 rRNA analysis to the rickettsiae with some of the nucleic
acids that my colleagues and I have furnished. Some of this work has been
published (Science 23@:556-558, 1985), the rest is being prepared for
publication now. Members of the genera Rochalimaea and Rickettsia clearly
belong in the alpha-subdivision of the purple bacteria. Thus, Proteus and
Rickettsia are not related.

 

I hope that this will ansver your query. Do not hesitate to call on me if
there should be any further questions. With best regards,

Sincerely,
< ,* Lv "
bs Why
Emilio Weiss

2612 Raymond Street
Chevy Chase, MD 20815