L. Baltzer et al. / Bioorg. Med. Chem. 7 (1999) 83±91
89
structure of MNRR (Fig. 4) it is very unlikely that Arg-
26 can bind residues in an acyl intermediate at the side
chain of His-15, suggesting that only Arg-30 binds in
the MNIR26R30 catalyzed reaction. Since the magni-
tudes of the rate enhancements for the MNIR26R30, the
MNIR30R34 and the MNIR34R37 catalyzed reactions
are approximately equal, and roughly equal to that of
half the rate enhancement of MNRR it seems likely that
only one arginine binds in all of these peptides. The
second-order rate constants of the peptide catalyzed
reactions are therefore compatible with a model where
Arg-30 may bind the anionic transition state of the
reaction catalyzed by HisH+-15±His-19, and Arg-34
may bind the transition state of the reaction catalyzed
by HisH+-11±His-15, but where Arg-30 is not in a
position to bind in the transition state of the MNI cat-
alyzed reaction and Arg-34 does not bind in the MNII
catalyzed one. Unfortunately, there is also no extra
binding energy provided by the ¯anking arginines in
positions 26 and 37 in the MNI and MNII catalyzed
reactions.
The same experiments were carried out for the JNII
series of peptides where it was shown that residues in
positions 8 and 19 contribute little to transition state
binding in the catalysis of acyl-transfer reactions. JNII
was therefore supplemented with homoarginines, argi-
nines, lysines, and ornithines in positions 11 and 15; the
second-order rate constants are given in Figure 3. Sys-
tematic variation of the residues shows interesting
optima. Arg-11 has been combined with Arg-15, Lys-15
and Orn-15 and the maximum reactivity is observed
for Lys-15 suggesting that Lys is the optimum residue
in position 15 for the hydrolysis of I. When Lys-11 is
supplemented by homoarginine-15 and Arg-15 the
reactivity increases as the residue in position 15 is
`shortened', in agreement with the hypothesis that
Lys-15 may provide the maximum transition state
binding. Only a single catalyst with an even shorter
side chain in position 11, that of Orn-11, has been stu-
died. JNIIOrn11R15 shows a rate constant that is
1
0.099 M 1s and a rate enhancement that is expected
to arise from the substitution of Arg-15 by Lys-15
would bring its reactivity to within the range of that
with Arg-11 Lys-15. These results suggest that Lys
should be the optimum residue in position 11, too. The
series of peptides that are based on Arg-15 do indeed
show a maximum for Lys-11, whereas the catalysts with
Arg-11 or Orn-11 are less reactive. Also, the two Lys-15
based catalysts containing Arg-11 or hArg-11 are more
reactive the shorter the side chain in position 11 in good
agreement with the conclusion that Lys-11 provides
optimum stabilization.
Reactivity as a function of the length of the side chains
of the binding residues
It was shown previously that the reactivity of the MN-
42 based motifs depended strongly on the nature of the
side chain of the binding residues.11 Two arginines bind
less well than one arginine and one lysine, and the same
holds for two lysines. The best binding occurs for unlike
amino acids, perhaps due to repulsion of residues of
equal charge and length. The largest rate constant
for the peptides in the MN series was obtained by
introducing Lys-30±Arg-34 (k2=0.135M 1 s 1), whereas
Arg-30 and Lys-34 were slightly less reactive
(k2=0.110M 1 s 1). In order to probe whether the rate
constants of the MN series could be enhanced further
Arg was replaced by homo arginine (hR) in MNKhR,
because homoarginine has a longer side chain that
can bind remote substrates better. The second-order
rate constant has now been measured and it is
0.090 M 1 s 1, which is lower than that of MNKR. It
appears that in the catalysis of acyl-transfer reactions of
mono-p-nitrophenyl fumarate the optimum binding
distance for a residue in the 34 position in the best pos-
sible conformation is obtained if an arginine residue is
used. MNKR is therefore the most ecient catalyst in
the MN series of peptides and the rate enhancement
accomplished by the introduction of Lys-30 and Arg-34
can be estimated for each of the two-residue sites.
According to the arguments discussed above, the
reactivity of MNIKR, i.e., the MN-42 motif with His-
11, His-15, Lys-30 and Arg-34, should be approximately
half of that of MNKR which catalyzes the hydrolysis of
I with a second order rate constant of 0.135 M 1 s 1. It
The binding distances appear to be such that the natu-
rally occurring residues Arg and Lys provide the best
transition state stabilization possible as longer and
shorter side chains give rise to less ecient catalysts.
The Lys-11 Lys-15 combination will perhaps prove to
be the most ecient although the lesson from earlier
studies is that the combination of two identical residues
is less reactive than non-identical ones.
The magnitude of the second-order rate constant for the
JNIIR11K15 catalyzed reaction is the largest obtained
for a single well-de®ned reactive site based on the reac-
tivity of the HisH+±His pair. The rate enhancement
over that of the JNII catalyzed reaction approaches a
factor of three. It is, however, for the catalyst
JNIIK11R15 that has a slightly lower reactivity, that the
most pronounced discrimination between I and II is
observed, a factor of 2.5.
The rate constant maxima for both the JN and the MN
series of peptides occur for the same type of binding
residues, Arg and Lys, although in dierent order. In
the MN series Lys has the lower number in the sequence
whereas in the JN series Arg has the lower number. In
the MN series the Lys residue is therefore located
opposite to the nucleophilic histidine, whereas in the JN
series it is opposite to the protonated one, suggesting
that the reactive sites may bind the transition states in
dierent geometries, something that may turn out to
have interesting stereochemical consequences. The fact
that Arg and Lys can bind cooperatively in the
1
should therefore be approximately 0.07 M 1 s which
corresponds to a reactivity enhancement of more than a
factor of six, and a ÁÁG of approximately 1.1 kcal/mol,
over that of the MNI catalyzed reaction. The observed
stabilization is approximately a factor of two larger
than what is obtained from a single salt bridge,16 sug-
gesting that both Lys-30 and Arg-34 bind in the MNKR
catalyzed reaction.