To investigate the role of the regions that are most affected structurally by the swapping, we expressed variant proteins by replacing two crucial residues with the corresponding ones fro
Trang 1without exchange As a consequence, the hinge region 16–22
has a different tertiary structure in the two forms In the
native protein, the equilibrium ratio between MxM and
M¼M is about 7 : 3 Kinetic analysis of the swapping
pro-cess for a recombinant sample shows that it folds mainly in
the M¼M form, then undergoes interconversion into the
MxM form, reaching the same 7 : 3equilibrium ratio To
investigate the role of the regions that are most affected
structurally by the swapping, we expressed variant proteins
by replacing two crucial residues with the corresponding
ones from RNase A: Pro19, within the hinge peptide, and
Leu28, located at the interface between subunits We
dimers The P19A mutation slightly increases the thermal stability of the monomer, but it does not alter the swapping tendency of the dimer In contrast, the L28Q mutation sig-nificantly affects both the dimerization and swapping pro-cesses but not the thermal stability of the monomer Overall, these results suggest that the structural determinants that control the exchange of N-terminal arms in BS-RNase may not be located within the hinge peptide, and point to a crucial role of the interface residues
Keywords: bovine seminal ribonuclease; domain swapping; proline; ribonuclease A; site-directed mutagenesis
Bovine seminal ribonuclease (BS-RNase), the only dimeric
protein in the pancreatic-type ribonuclease family, is
characterized in solution by an equilibrium between two
different structures [1]: in the form dubbed MxM, the
N-terminal arms are exchanged, or swapped, between the
two identical subunits, whereas in the form indicated as
M¼M no swapping occurs In the native protein, the
equilibrium ratio between MxM and M¼M is about 7 : 3
The two identical subunits are linked through two disulfide
bridges between Cys31 and 32 of one subunit with Cys32¢
and 31¢, respectively, of the partner subunit Each subunit
has 83% of the amino-acid sequence identical with that of
bovine pancreatic RNase A In particular, both enzymes
exhibit active sites constituted by identical amino-acid
residues in the same sequence position Beside ribonuclease
activity, BS-RNase is endowed with several additional
biological activities, such as allostery [2], cytotoxicity toward
malignant cells [3], immunosuppression and
antispermato-genesis [4] Domain swapping in BS-RNase was found to be
determinant for all of these activities, which may suggest a physiological role for this structural peculiarity
A folded and stable monomeric derivative of BS-RNase can be obtained by selective reduction of the dimeric protein with a moderate excess of dithiothreitol, and stabilized by either alkylation of the exposed thiol groups [5] or reaction with glutathione [6] All monomeric derivatives of BS-RNase are catalytically more active than the native dimeric enzyme, but they do not exhibit any allosteric property and have no detectable special biological action [7]
In a recent paper, we reported an NMR characterization
of the N67D variant of monomeric BS-RNase [8], hence-forth called mBS The mutation avoids sample heterogen-eity arising from the spontaneous deamidation of Asn67 [9], but it does not affect enzymatic activity Comparison of the solution structures, as well as specific NMR relaxation experiments, indicated that the hinge region 16–22 is much more flexible in mBS than in RNase A However, this region shows the greatest sequence difference from RN-ase A: GNSPSSS in BS-RNRN-ase vs STSAASS in RNRN-ase A
As a consequence of its flexibility, the structure of this segment is not well defined in the solution structure of mBS (Fig 1A) Moreover, owing to extensive overlap of diag-nostic signals, we could not unequivocally assign trans isomerism to Pro19 Mutagenic studies have shown that Pro19 and Leu28, which in BS-RNase makes a hydrophobic contact at the interface between the two subunits (Fig 1B), are two crucial residues in inducing dimerization and swapping N-terminal arms in RNase A variants [10,11]
As the first step of a study aimed to investigate, through a
Correspondence to D Picone, Dipartimento di Chimica, Universita`
di Napoli Federico II, Via Cintia, 80126, Napoli, Italy.
Fax: + 39 081 674409, Tel.: + 39 081 674406,
E-mail: picone@chemistry.unina.it
Abbreviations: BS-RNase, bovine seminal ribonuclease; mBS,
mono-meric N67D BS-RNase; RNase A, bovine pancreatic ribonuclease;
DVS, divinyl sulfone.
(Received 1 August 2003, revised 2 October 2003,
accepted 7 October 2003)
Trang 2systematic mutagenic approach, the role played by the hinge
and interface regions in the swapping process, we prepared
BS-RNase variants by replacing Pro19 and Leu28 with the
corresponding residues from RNase A Here we report a
characterization of monomeric P19A, L28Q and P19A/
L28Q variants of mBS carried out by 2D NMR, CD and
differential scanning microcalorimetry, and an investigation
of the kinetics of swapping of all variant dimers in
comparison with that of the parent protein
Materials and methods
Construction of mBS mutants
Site-directed mutagenesis was performed by a megaprimer
PCR method [12] to produce the mutants coding for
P19A-mBS, L28Q-mBS and P19A/L28Q-P19A-mBS, starting from the
pET-22b(+) plasmid cDNA coding for the wild-type
enzyme which already carries the N67D mutation, to avoid
sample heterogeneity by spontaneous deamidation at the
Asn67 site [8]
PCR amplification was performed with an Eppendorf
Mastercycler amplifier The forward flanking primer
sequence used in these experiments, 5¢-GAGTGCGGCC
GCAAGCTTGGGCTG-3¢, had an estimated Tmof 82C
The reverse flanking primer sequence, 5¢-ATATACA
TATGAAAGAAAG-3¢, had a calculated Tm of 42C
The mutagenic primers for each variant are: P19A (5¢-AGA
GCTGCTAGCAGAGTTG-3¢) and L28Q (5¢-CACAT
CATCCTGGTTGCAA-3¢) (nucleotides that represent
mutations are underlined) For the mutant P19A/L28Q-mBS the mutagenic primer L28Q was used starting from the pET-22b(+) plasmid cDNA coding for the mutant P19A-mBS The amplified, mutated genes were separated, excised, and purified from the agarose gel followed by cloning into pET-22b(+) between the HindIII and NdeI sites
Insertion of the correct mutations was confirmed by DNA sequencing
Recovery of proteins All the proteins were expressed in Escherichia coli and purified in monomeric form, with Cys31 and 32 linked to two glutathione molecules, as described previously [13] Monomers with Cys31 and 32 in the reduced form were obtained by selective reduction of the mixed disulfide bridges with a 5 : 1 molar excess of dithiothreitol for 20 min
at room temperature in 0.1MTris/acetate buffer, pH 8.4 The samples were either carboxyamidomethylated with iodoacetamide [5], to obtain the monomeric proteins used for CD and microcalorimetric analysis, or dialyzed against 0.1M Tris/acetate, pH 8.4, for 20 h at 4C, to obtain dimers The last step of the purification procedure was always a gel filtration on Sephadex G-75 to separate monomers from dimers All dimerization steps were performed at 4C
Recombinant RNase A was obtained and purified as described previously [8]
Protein homogeneity was verified by SDS/PAGE and MALDI-TOF MS, registered at the Sezione di
Fig 1 Ribbon representation of the solution structure of mBS-RNase (A),as derived from heteronuclear NMR data (pdb accession code 1WQW),and the X-ray structure of the MxM form of BS-RNase (B; pdb accession code 1BSR) Pro19 and Leu28 are highlighted The figure was drawn with MOLMOL software [25].
Trang 3Assessing the extent of the N-terminal swap
at equilibrium
Cross-linking experiments were performed using divinyl
sulfone (DVS) as a 10% solution in ethanol The protein
(20 lg) in sodium acetate buffer (100 mM, pH 5, 100 lL)
and DVS (1 lL of the 10% solution) was incubated at
30C [11]; this is 1000-fold excess of sulfone over each
subunit of the protein Aliquots were withdrawn over a
period of 96 h, quenched with 2-mercaptoethanol (final
concentration 200 mM), incubated for 15–30 min at room
temperature, and loaded on gels for reducing SDS/PAGE
The ratio of monomer to cross-linked dimer was estimated
qualitatively by Coomassie blue staining
NMR
NMR measurements were performed on Bruker DRX400
and DRX500 spectrometers All spectra were collected
using the standard Bruker pulse sequence library Protein
concentration was 2 mMin 95% H2O/5% D2O, pH 5.65
CD
The CD spectra were recorded with a Jasco J-715
spectro-polarimeter equipped with a Peltier-type temperature
con-trol system (model PTC-348WI) The instrument was
calibrated with an aqueous solution ofD
-10-camphorsulf-onic acid at 290 nm [14] Molar ellipticity per mean residue,
[h] in degreesÆcm2Ædmol)1, was calculated from the equation
[h]¼ hobsmrw/10lC, where hobsis the ellipticity measured in
degrees, mrw is the mean residue molecular mass (117 Da
[5]), C is the protein concentration in gÆL)1, and l is the
optical path length of the cell in cm A 0.1-cm path length
cell and a protein concentration of 0.3mgÆmL)1in 10 mM
sodium acetate buffer, pH 5.0, were used CD spectra were
recorded at 25C with a time constant of 16 s, a 2-nm band
width, and a scan rate of 5 nmÆmin)1; they were
signal-averaged over at least five scans, and baseline-corrected by
subtracting the buffer spectrum Thermal unfolding curves
were recorded in the temperature scan mode at 222 nm
from 25C up to 85 C with a scan rate of 0.5 KÆmin)1
Scanning calorimetry
Calorimetric measurements were performed on a
second-generation Setaram Micro-DSC A scanning rate of
about 15 mgÆper L culture Each of these variants retains a catalytic activity against yeast RNA comparable with that
of parent mBS, indicating that a native conformation is present A further indication of the similarity of their global fold to that of the parent protein is provided by the 1D
1H-NMR spectra (data not shown), which display all the characteristic signals in almost identical positions The similarity was confirmed by CD measurements (Fig 2) The estimation of secondary-structure content, performed
by the neural network-based procedure implemented in the programK2D[16,17], yielded very similar values for all the protein samples (28% a-helix, 36% b-sheet and 40% random coil); these values are also in good agreement with the secondary structure derived from the NMR structure of mBS [8]
To allow a more accurate evaluation of the effect of single point mutations on the solution structure of monomeric derivatives, we analysed the 2D NMR spectra of the different variants of mBS Figure 3shows the expanded regions of TOCSY spectra of L28Q-mBS (panel L28Q), P19A/L28Q-mBS (panel PALQ) and P19A-mBS (panel P19A), in comparison with the same region of the parent mBS (panel mBS) The new signal at 8.40–1.40 p.p.m., which appears in the spectra of P19A-mBS and
Fig 2 Far-UV spectra of mBS-RNase (solid curve),P19A-mBS-RNase (dashed curve),L28Q-mBS-RNase (dotted/dashed curve) and P19A/ L28Q-mBS-RNase (dotted curve) in 10 m M sodium acetate buffer,
pH 5.0,25 C The horizontal dotted line indicates the zero value of the ellipticity.
Trang 4P19A/L28Q-mBS, has been attributed to Ala19 in spite of
the difference with the assignment made for RNase A in
very similar experimental conditions [18] Moreover, in the
spectra of L28Q-mBS and P19A/L28Q-mBS, the
correla-tions that belong to Leu28 are missing, and are replaced by
a set of new signals tentatively attributed to Gln28 (data not
shown) Apart from these expected differences, the high
similarity among the spectra of the variants and the parent
protein provides additional, strong evidence of an essentially
identical fold of all variant proteins However, the
NH-CbH3cross-peak of Ala19 is significantly broader than
other cross-peaks; this may reflect an equilibrium between
different conformations of the 16–22 hinge region, occurring
at a rate comparable to the NMR chemical shift time scale
A thermodynamic characterization of all monomeric
proteins was performed by differential scanning
calori-metry The results, shown in Table 1, agreed with the
temperature-induced unfolding curves obtained by CD
measurements (data not shown) No sizeable differences
were found among the monomeric variants tested, and only
for P19A-mBS a significant thermal stabilization occurs
The introduction of Ala for Pro19 makes mBS more similar
to RNase A, which has a thermal stability higher than that
of mBS However, P19A/L28Q-mBS has the same thermal stability of the parent protein, in spite of greater similarity to RNase A
To directly characterize the process of exchange of the N-termini, we prepared dimers of all variants The mono-meric proteins were submitted to mild reduction, which selectively removes the glutathione molecules linked to Cys31 and 32, followed by air oxidation and gel filtration on Sephadex G-75 to obtain the corresponding dimers We obtained 80% of dimer for both BS-RNase and P19A-BS-RNase, whereas in the case of L28Q and P19A/L28Q variants the yield of dimer was significantly lower, 50% The recombinant dimers were obtained predominantly in their nonswapped forms; the extent of exchange of the N-termini, as evaluated by selective reduction of the interchain disulfide bridges followed by gel filtration (vide supra), was initially 15% The dimers were then incubated
at 37C for a week, to allow equilibration between the interconverting forms A qualitative evaluation of the extent
of swapping at equilibrium was performed by cross-linking experiments with DVS, followed by SDS/PAGE analysis under reducing conditions DVS covalently joins the two His residues of the active site (His12 and His119) [11], which belong to the same subunit in M¼M and to different subunits in MxM Thus, upon reduction and denaturation, either a product of molecular mass 27 000 Da (from MxM)
or a product of molecular mass 13500 Da (from M¼M) is obtained Figure 4 depicts the time-dependent course of this reaction in the equilibrium mixtures of dimers After 24 h of reaction, all proteins show almost the same relative proportion of exchanging and nonexchanging forms However, after 48 h, a slight prevalence of the MxM form can be seen in the BS-RNase and P19A variant, whereas the L28Q and P19A/L28Q variants still contain comparable amounts of MxM and M¼M This becomes even more
Fig 3 Regions of 500 MHz 2D TOCSY spectra of L28Q-mBS-RNase (panel L28Q), P19A/L28Q-mBS-RNase (panel PALQ), P19A-mBS-RNase (panel P19A),and mBS-RNase (panel mBS) The N1H/b-methyl cross-peak of Ala19 is boxed in the P19A and PALQ panels.
Table 1 Thermodynamic parameters of thermal denaturation of mBS,
P19A-mBS,L28Q-mBS,P19A/L28Q-mBS and RNase A Estimated
error on Td, DdH and DdCp was 0.2 C, 5% and 10%, respectively.
Td (C)
DdH(Td) (kJÆmol)1)
DdCp (kJÆmol)1ÆK)1)
P19A/L28Q-mBS 53.0 387 5.0
Trang 5form is present in native BS-RNase and P19A variant.
A more detailed comparison of the swapping process
among the different BS-RNase variants has been performed
by a different method, based on the observation that, after
selective reduction of the interchain disulfide bridges, the
MxM form retains a dimeric structure, whereas the M¼M
form dissociates into two monomers upon gel filtration [1]
We observed (Fig 5A) that, at 37C, the dislocation of the
N-terminal arms for the recombinant BS-RNase is very
similar to that reported for the native enzyme [1], reaching
the same 7 : 3equilibrium in 5 days This result also
provides indirect evidence of correct pairing of the
inter-chain disulfide bridges in the recombinant sample [11] The
kinetics of the N-terminal arms swapping in the
P19A-BS-RNase is comparable to that of BS-P19A-BS-RNase (Fig 5B)
Interestingly, the amount of the MxM form at the plateau
as evaluated by integrating the peaks obtained on gel
filtration is 70%, i.e in complete agreement with that
obtained for the native enzyme and also in agreement with
the qualitative estimation based on the DVS reaction In
contrast, Fig 5C,D clearly shows that replacement of
Leu28 with Gln affects in a significant way the kinetics
and equilibrium of swapping in both the L28Q-BS-RNase
and P19A/L28Q-BS-RNase variants Furthermore, the
plateau corresponds to 50% of MxM for both proteins, in
agreement with the DVS cross-linking results
As reported previously [1], at 4C the interconversion of
the native protein was obviously slowed down For instance,
after 500 h only 25% of MxM is present in BS-RNase, and
20% in P19A-BS-RNase, i.e the process is still far from
equilibrium To assess this hypothesis, we monitored the
reverse process, namely the interconversion of MxM into
M¼M, for P19A-BS-RNase To isolate the MxM dimer,
the equilibrium mixture, containing both forms in the usual
7 : 3ratio, was reduced selectively with dithiothreitol and
gel filtered The fractions containing the reduced dimer were
pooled and air oxidized The amount of the MxM form in
the isolated sample at this stage was 85% Aliquots were
then separately incubated at 4C and 3 7 C, and the extent
of the swapping was assessed as previously described
Figure 6 shows the amount of the MxM form present as a
function of incubation time P19A-BS-RNase behaves like
the native sample [1], reaching the typical equilibrium
mixture in 180 h at 37 C, whereas at 4 C the process
seems kinetically frozen, and the content of MxM is
constant over several weeks
of two paradigmatic residues involved in the dislocation
of BS-RNase N-terminal arms, i.e Pro19 and Leu28, located in the hinge peptide and in the intersubunit interface, respectively CD and NMR spectra suggest a close similarity in the global fold and in the structural properties of all monomeric variants to those of the parent protein These results are in agreement with the solution structure of mBS [8], which indicates that both residues are solvent exposed and make only a limited number of van der Waals interactions with the rest of the protein The only difference related to the substitution of Pro with Ala is a slight increase in the thermal stability, which becomes more similar to that of RNase A This result is in agreement with theoretical calculations [19], and CD and calorimetric studies on the A19P-RNase A variant [20], which both support a destabilizing effect resulting from the introduction of a Pro at position 19 in RNase A
Surprisingly, the substitution of Pro19 by Ala does not affect the dislocation of N-termini in BS-RNase A Pro residue is often found in crucial regions of other domain swapped proteins [21], and cis–trans Pro isomerization has been regarded as a key event in protein folding (for a recent overview, see Wedemeyer et al [22]) This observation led to the definition of hinge Pro residues as quaternary structure helpers The similarity of kinetic behaviour between BS-RNase and P19A-BS-RNase seems to rule out the involvement of a cis–trans Pro19 isomerization as a crucial step in the swapping mechanism of BS-RNase In the crystallographic structure of BS-RNase MxM form [23], the peptide bond between residues 18 and 19 has a trans conformation, and it seems reasonable to assume that the same holds for the P19A-BS-RNase MxM form If Pro19 were cis in the M¼M form, the P19A variant would hardly
be expected to display such similar thermodynamic and kinetic behaviour in the interconversion process, as a cis peptide bond is significantly less stable for Ala than for Pro
In other words, in the P19A M¼M form, Ala19 would either be forced into an unfavourable cis conformation or would have already assumed the intrinsically preferred trans conformation; in both cases, an effect on the swapping process would be expected
Even from a structural point of view, the role of Pro19 is still ambiguous All structural studies on BS-RNase report poor definition of the hinge region around Pro19, and this abnormality has often been taken as an indication of high
Trang 6intrinsic flexibility of this tract of the protein Yet, in the
crystallographic structure of a dimeric variant of human
pancreatic ribonuclease, called PM8 [24], in which the
N-terminal region and the whole hinge loop are identical with those of BS-RNase, the hinge assumes a 310 helix structure This result suggests that the flexibility of the corresponding region in BS-RNase cannot be ascribed to the hinge loop itself, but probably arises from interactions with other parts of the molecule
Overall, our data indicate that Pro19 is not crucial in the swapping process, and suggest that the structural determi-nants for this process are located in regions different from the hinge loop 16–22
In the search for the region primarily involved in the swapping, we focused our attention on the interface between the subunits In a previous study [8] we found that, in mBS, helix 2 (residues 24–32) is more disordered than in RNase A This disorder may be attributed to the proximity
of the hinge peptide (residues 16–22), or possibly to an unfavourable solvent exposure of the Leu28 side chain (residue 28 is Gln in RNase A) As far as the monomer is concerned, introduction of Gln for Leu at position 28 does not increase the stability or significantly affect the structure,
in spite of the greater hydrophilicity and helical propensity
of Gln However, the presence of Leu28 in BS-RNase seems
to favour the dimerization process: for both L28Q-BS-RNase and P19A/L28Q-BS-L28Q-BS-RNase, the yield of dimer obtained on oxidation of the monomeric, reduced protein is only 50%, which is significantly lower than for BS-RNase and P19A-BS-RNase (at least 80%) Interestingly, literature data for some variants of RNase A also report a higher yield of dimer when Gln28 is substituted with Leu11 The influence of the Leu side chain is also evident during the swapping process, and again the two mutants in which Leu28 had been substituted display closely matching kinetic behaviours and at equilibrium contain a smaller percentage
of MxM (50%) Thus, several data indicate that Leu28 is crucial not only for the interaction between the subunits, but even for the swapping process Finally, the close similarity between the swapping behaviour of the N-terminal arms in L28Q-BS-RNase and P19A/L28Q-BS-RNase confirms that Pro19 does not represent a key residue in this process From a different perspective, our results provide further evidence of the evolutionary significance of the A19P and Q28L substitutions in the path leading from RNase A to
Fig 6 Kinetic analysis of the MxM to M=M conversion for P19A-BS-RNase The percentage of MxM form as a function of the incubation time at 4 C (r) and 3 7 C (d) is reported.
Fig 5 Kinetic analysis of the M=M to MxM conversion for BS-RNase
(A),P19A-BS-RNase (B),L28Q-BS-RNase (C) and
P19A/L28Q-BS-RNase (D) The percentage of the MxM form as a function of the
incubation time at 4 C (r) and 3 7 C (d) is reported for each dimeric
protein.
Trang 7Parente, A & D’Alessio, G (1992) The dual-mode
quater-nary structure of seminal RNase Proc Natl Acad Sci USA 89,
1870–1874.
2 Piccoli, R., Di Donato, A & D’Alessio, G (1988) Co-operativity
in seminal ribonuclease function Biochem J 253, 329–336.
3 Youle, R.J & D’Alessio, G (1997) Ribonucleases: Structures and
Functions (D’Alessio, G & Riordan, J.F., eds), pp 491–514.
Academic Press, New York.
4 D’Alessio, G., Di Donato, A., Parente, A & Piccoli, R (1991)
Seminal RNase: a unique member of the ribonuclease superfamily.
Trends Biochem Sci 16, 104–106.
5 D’Alessio, G., Malorni, M.C & Parente, A (1975) Dissociation of
bovine seminal ribonuclease into catalytically active monomers by
selective reduction and alkylation of the intersubunit disulfide
bridges Biochemistry 14, 1116–1122.
6 Smith, G.K & Schaffer, S.W (1979) Selective reduction of seminal
ribonuclease by glutathione Arch Biochem Biophys 196, 102–
108.
7 D’Alessio, G., Di Donato, A., Mazzarella, L & Piccoli, R (1997)
Seminal ribonuclease: the importance of diversity Ribonucleases:
Structures and Functions (D’Alessio, G & Riordan, J.F., eds), pp.
383–423 Academic Press, New York.
8 Avitabile, F., Alfano, C., Spadaccini, R., Crescenzi, O., D’Ursi,
A.M., D’Alessio, G., Tancredi, T & Picone, D (2003) The
swapping of terminal arms in ribonucleases: comparison of the
solution structure of monomeric bovine seminal and pancreatic
ribonucleases Biochemistry 42, 8704–8711.
9 Di Donato, A & D’Alessio, G (1981) Heterogeneity of bovine
seminal ribonuclease Biochemistry 20, 7232–7237.
10 Di Donato, A., Cafaro, V & D’Alessio, G (1994) Ribonuclease A
can be transformed into a dimeric ribonuclease with antitumor
activity J Biol Chem 269, 17394–17396.
11 Ciglic, M.I., Jackson, P.J., Raillard, S.I., Haugg, M., Jermann,
T.M., Opitz, J.G., Trabesinger-Ruf, N & Benner, S.A (1998)
Origin of dimeric structure in the ribonuclease superfamily.
Biochemistry 37, 4008–4022.
dichroism using an unsupervised learning neural network Protein Eng 6, 383–390.
17 Lobley, A., Whitmore, L & Wallace, B.A (2002) DICHROWEB:
an interactive website for the analysis of protein secondary structure from circular dichroism spectra Bioinformatics 18, 211–212.
18 Shimotakahara, S., Rios, C.B., Laity, J.H., Zimmerman, D.E., Scheraga, H.A & Montelione, G.T (1997) NMR structural analysis of an analog of an intermediate formed in the rate-determining step of one pathway in the oxidative folding of bovine pancreatic ribonuclease A: automated analysis of 1 H, 13 C, and 15 N resonance assignments for wild-type and [C65S, C72S] mutant forms Biochemistry 36, 6915–6929.
19 Mazzarella, L., Vitagliano, L & Zagari, A (1995) Swapping structural determinants of ribonucleases: an energetic analysis of the hinge peptide 16–22 Proc Natl Acad Sci USA 92, 3799–3803.
20 Catanzano, F., Graziano, G., Cafaro, V., D’Alessio, G., Di Donato, A & Barone, G (1997) From ribonuclease A toward bovine seminal ribonuclease: a step by step thermodynamic ana-lysis Biochemistry 36, 14403–14408.
21 Bergdoll, M., Remy, M.H., Cagnon, C., Masson, J.M & Dumas,
P (1997) Proline-dependent oligomerization with arm exchange Structure 5, 391–401.
22 Wedemeyer, W., Welker, E & Scheraga, H (2002) Proline cis-trans isomerization and protein folding Biochemistry 41, 14637– 14644.
23 Mazzarella, L., Capasso, S., Demasi, D., Di Lorenzo, G., Mattia, C.A & Zagari, A (1993) Bovine seminal ribonuclease: structure at 1.9 angstrom resolution Acta Crystallogr D49, 389–402.
24 Canals, A., Pous, J., Guasch, A., Benito, A., Ribo, M., Vilanova,
M & Coll, M (2001) The structure of an engineered domain-swapped ribonuclease dimer and its implications for the evolution
of proteins toward oligomerization Structure 9, 967–976.
25 Koradi, R., Billeter, M & Wuthrich, K (1996) MOLMOL: a program for display and analysis of macromolecular structures.
J Mol Graph 14, 29–32.