Tải bản đầy đủ - 0trang
5 Cross-Linking Protocols Based on Biological Systems
General Approaches for Chemical Cross-Linking
target proteins. Since intermolecular cross-linking would also occur, dilute solutions should be used
to avoid close encounter of the macromolecules. For intermolecular conjugation of two proteins,
such as the preparations of immunotoxins or enzyme-antibody conjugates, a two-step procedure is
desired as discussed above under Section 9.2.2, with examples presented therein. Three-step and
multistep procedures are also applicable as shown in Sections 9.2.3 and 9.2.4.
22.214.171.124 Cross-Linking Multisubunit Complexes
Similar to single subunit proteins, dilute solutions of strongly associated, stable multisubunit complexes should be used to avoid intermolecular cross-linking as discussed in Section 9.3.3. In most
cases, a one-step procedure can be used for intracomplex subunit cross-linking as showed above in
Section 9.2.1 for ribosome protein complexes. Various reagents, either homofunctional or heterofunctional, may be used depending on the aims of the experiments. These are illustrated in Section
9.4 for reverse transcriptase (Section 126.96.36.199.1), ATP synthase (Sections 188.8.131.52.2 and 184.108.40.206.1),
aldolase (Section 220.127.116.11.1), and UvsY protein hexamer (Section 18.104.22.168).
Investigation of protein–protein interactions of readily dissociable protein complexes is more challenging. Some proteins interact with several other proteins transiently for regulatory or enzymatic
purposes. In these cases, a condition that causes the formation of the complex must be achieved during the cross-linking experiment, either by inclusion of substrates or cofactors that induce the interaction or by increasing the concentrations of the interacting subunits. The most commonly used reagents
are photoactivatable cross-linkers, although oxidative cross-linking such as hexahistidine-mediated
cross-linking methodology has been evolving.65,66 Any of the photoactivatable reagents listed in
Appendix E may be used. In the first step, the reagent is covalently attached to lysine or cysteine side
chains of the prey protein. The derivatized prey protein is then incubated with an interacting protein
and exposed to UV light to induce cross-linking between the prey protein and the interacting proteins.
For label transfer experiments, cleavable reagents, particularly those containing disulfide bonds, are
used. After photoactivation, the reagent is cleaved, thereby transferring the label to the interacting
proteins. The transferred label enables the identification or isolation of the interacting proteins.
For hexahistidine-cross-linking, a hexahistidine tag or a NH2-Gly-Gly-His tag is first genetically
introduced into the prey protein.66–68 On incubation of the tagged prey protein with other interacting
protein in the presence of nickel, fast and efficient cross-linking of proteins is achieved when oxidized by peracids like magnesium monoperoxyphthalic acid or KHSO5. The cross-linked adducts
can be identified immunologically. Klein et al.69 have used this method to study the transcriptional
activation domain–coactivator complexes.
Advances in genetic cloning have facilitated the development of a new methodology of
cross-linking to study protein–protein interaction, particularly in living cells. Photoactivatable
amino acid analogues such as p-benzoyl-l-phenylalanine (pBpa), p-azido-l-phenylalanine, and
4′-[3-(trifluoromethyl)-3H-diazirin-3-yl]-l-phenylalanine are site-specifically incorporated into
interested proteins by means of heterologous amber suppressor tRNA/aminoacyl-tRNA synthetase
pairs that recognize the unnatural amino acids.70 After association of the modified protein with target macromolecules, the complex is photoactivated to initiate cross-linking. Lee et al.71 have used
pBpa to study the interaction of Escherichia coli catabolite activator protein (CAP) with DNA. pBpa
was genetically incorporated into CAP in bacteria in response to an amber nonsense codon using
an orthogonal tRNA/aminoacyl-tRNA synthetase pair. On binding with DNA and after UV irradiation, SDS–PAGE analysis showed that the mutant CAP containing pBpa formed a covalent complex
with a DNA fragment containing the consensus operator sequence.
9.5.2 Membrane-Bound Proteins
Membrane-bound proteins present a special challenge for studying protein–protein interactions.
These proteins consist of transmembrane helices that are hydrophobic in nature and an extramembraneous region exposed to the aqueous media. The choice of cross-linking reagent depends on
Chemistry of Protein and Nucleic Acid Cross-Linking and Conjugation
objectives of the experiment. Generally, the exposed section of the protein is of interest, such as
the membrane surface receptors. In this case, the use of membrane-impermeable cross-linkers will
ensure cell-surface specific cross-linking.35 Compounds containing formal charges such as imidates, hydroxyl and amino groups, polyethylene glycol, and sulfonyl groups such as sulfo-NHSesters are water-soluble, membrane-impermeable, and nonreactive with inner-membrane proteins.
Thus, bis-imidoesters, sulfo-NHS-esters, and sulfonated photoreactive cross-linkers are good
choices. Cross-linkers with spacer arms formed from polyethylene glycol provide the added benefit of transferring their hydrophilic spacer to the cross-linked complex, decreasing the potential
for aggregation and precipitation of the cross-linked adduct. For these reagents, reaction time and
concentration are less critical. If water-insoluble cross-linkers are used, the amount of reagent and
reaction must be well controlled to reduce membrane penetration and reaction with inner membrane
proteins. Numerous examples of studies on membrane surface proteins exist in the literature.66,72,73
For example, Patzke et al.74 investigated the structural organization of the coxsackievirus–adenovirus receptor (CAR) using homobifunctional bis(sulfosuccinimidyl) suberate (BS3) cross-linker.
CAR is a type I transmembrane protein composed of two Ig domains, a membrane distal D1 and
a membrane proximal D2. Cross-linking was started by the addition of BS3 to extracellular CAR
domains to a final concentration of 1 mM and incubated on ice for 1–2 h. Protein concentrations
were chosen such that one of the two putative binding partners was used in up to 20-fold molar
excess. The reaction was quenched with 50 mM Tris-HCl with subsequent heating. The cross-linked
product was analyzed by SDS–PAGE and Western blot. The results showed the homophilic and
heterophilic binding activities of D1 and D2. Western blot analysis identified monomeric, homodimeric, trimeric, and tetrameric complexes of CAR-D1D2. An additional cross-linked species, with
an apparent mass of 40 kDa, indicates that the D1 monomer also binds D2.
To study the inner membrane proteins, reagents of greater hydrophobicity for membrane penetration are required. NHS-esters and photoactivatable phenyl compounds are useful. Although imidates
are water-soluble, they can still penetrate membranes and may be used under certain circumstances.
Water-insoluble dicyclohexylcarbodiimide can also provide valuable information. Various crosslinkers with differing spacer arm lengths can be used to determine the distance between molecules
located in the membrane. Successful cross-linking with shorter cross-linkers is a strong indication
that the molecules are in close approximation and may be interacting. Failure to obtain cross-linking with short cross-linkers, while obtaining conjugation with reagents with longer spacer arms,
would indicate that the molecules are located in the same region of the membrane but not interacting. Wu et al.75 investigated the position of β4 transmembrane helices in the BK potassium channel
by disulfide cross-linking. BK channels are composed of α-subunits and four types of β-subunits.
The locations of the two β4 transmembrane (TM) helices, TM1 and TM2, relative to the seven
αTM helices, S0–S6, were analyzed from the extent of disulfide bond formation between cysteines
substituted in the extracellular flanks of these TM helices. Disulfide cross-linking was effected by
oxidation of reduced sulfhydryl groups with 4,4′-(azodicarbonyl)-bis-[1,1-dimethylpiperazinium,
diiodide]. From the highly cross-linked cysteine pairs, the authors inferred that β4 TM2 is close to
αS0 and that β4 TM1 is close to both αS1 and S2.
9.5.3 Nucleic Acids and Nucleic Acid –Protein Complexes
Nucleic acid cross-linking, either DNA–DNA, RNA–RNA, or DNA–RNA coupling, can be achieved
chemically or photochemically. Nucleic acid–specific homobifunctional, heterobifunctional, and
photosensitive cross-linkers are described in Chapters 5 and 6. Many of these reagents are a result
of the search for tumor-therapeutic drugs and may be used for chemical studies of nucleic acid
interactions. In some cases, the cross-linked nucleic acids are a result of carcinogenic oxidative
or UV-induced damages. Since thymine is photosensitive, it can be activated by UV-irradiation to
form pyrimidine-pyrimidine cross-links.76 Such a mechanism can be used to prepare cross-linked
DNA simply by exposing the nucleic acid to 250–270 nm UV light. Photocross-linking can also
General Approaches for Chemical Cross-Linking
be effected in the presence of metal complexes and other photosensitive compounds as described
in Chapters 5 and 6. In addition to chemical and photochemical approaches, disulfide cross-links
have been engineered into DNA and RNA.77,78 This method of cross-linking has been used to probe
solution structures, to monitor dynamic motion and thermodynamics, and to study the process of
tertiary structure folding and function of DNA and RNA.
Of more interest in the study of nucleic acids are their interactions with proteins. Numerous studies on protein regulation of gene express, DNA damage repair, chromosome structure, and general
nucleic acid–protein interactions have been published.79–82 Reagents described in previous chapters,
especially in Chapters 5 and 6, have been used for these studies. In addition, photo-cross-linking of
DNA and protein have also been explored. For example, Neher et al.83 used the method to study the
interaction of Xeroderma pigmentosum group C (XPC)-Rad23B complex and cisplatin-damaged
DNA strand. XPC–Rad23B is a protein complex involved in the recognition of damaged bulky
DNA adducts and initiates the global genomic nucleotide excision repair pathway. When a mixture
of XPC–RAD23B and cisplatin damaged DNA on ice was photolyzed by UV irradiation using
General Electric-15 Watt bulbs, which emitted a wavelength of 254 nm, cross-linking between the
protein complex and DNA was achieved. Analysis of the adducts by SDS–PAGE revealed that the
XPC–Rad23B complex makes direct contact with the cisplatin-damaged DNA strand. Using denaturation and immunoprecipitation analysis, it was found that the XPC subunit was shown to directly
bind with the damaged DNA, while the Rad23B–DNA interaction was largely indirect via its interaction with XPC. The power of photo-cross-linking may be realized from these experiments.
9.6 CONDITIONS FOR CLEAVAGE OF CROSS-LINKED COMPLEXES
Cleavage of the cross-linkers requires specific conditions depending on the type of bonding in the
reagents. A general approach is given below for some of the common linkages.
9.6.1 Disulfide Linkages
Cleavage of disulfide bonds can be easily achieved by incubating with sulfhydryl compounds such
as β-mercaptoethanol, dithiothreitol, or dithioerythritol at concentrations of about 10–100 mM,
between pH 7 and 9 at 25°C–37°C for 10–30 min. Occasionally, concentrations of reducing reagents
up to 0.4 M may be used.84 Common buffers such as Tris and phosphate, as well as detergents
such as Triton X-100 and sodium dodecylsulfate (SDS) do not interfere with the cleavage reaction.
Disulfide bonds can also be conveniently cleaved during electrophoresis by addition of a reducing
agent to the electrophoresis buffer.
9.6.2 Glycol Bonds
Vincinal glycol bonds can be cleaved by 15 mM sodium periodate, pH 7–7.5, for 4–5 h at 25°C.85
Buffers such as triethanolamine and phosphate, as well as SDS, do not interfere. But Tris cannot be
used since it reacts with sodium periodate.
9.6.3 Azo Bonds
Azo linkages can be cleaved by reduction with 0.1 M sodium dithionite in 0.15 M NaCl, buffered at
pH 8 with 0.1 M NaHCO3 for 25 min.86 Disulfide reducing agents do not interfere with this process.
9.6.4 Sulfone Linkages
Sulfone bonds can be easily hydrolyzed in 100 mM sodium phosphate adjusted to pH 11–12 with
Tris, 6 M urea, 0.1% SDS, and 2 mM dithiothreitol for 2 h at 37°C.87 The presence of dithiothreitol
is not absolutely necessary and the denaturants may not be needed.
Chemistry of Protein and Nucleic Acid Cross-Linking and Conjugation
9.6.5 Ester and Thioester Bonds
Theoretically esters and thioesters can be hydrolyzed under both acidic and alkaline conditions.
They are most conveniently cleaved by 1 M hydroxylamine, in 50 mM Tris, pH 7.5–8.5, 25 mM
CaCl2 and 1 mM benzamidine for 3–6 h 25°C–37°C.88
9.6.6 Acetals, Ketals, and Orthoesters
These acetal, ketal, and orthoester bonds are acid labile and base stable.89 Cross-linked products are
stable at basic conditions and can be manipulated at pH 8–9.90 They are approximately 100 times
more stable at pH 7.4 and 1000 times more stable at pH of 8.4. They can be cleaved at pH 5.4 in
minutes to hours, orthoesters being the fastest, acetals the slowest, and ketals intermediate.
9.7 REACTION COMPLICATIONS
9.7.1 General Considerations
Cross-linking of two different proteins with a bifunctional reagent can give rise to intramolecular and intermolecular products. Intermolecularly, a range of products including the desired
1:1-conjugate and the undesired polyconjugates and polymers of each of the reactant proteins
can occur. Other possible products include aggregates of the various newly formed dimers,
oligomers, and polymers. These side reactions may lead to the loss of catalytic or immunologic
activities of the original reactants. Intrachain cross-linking may give rise to a variety of complications, including change of structure, which may lead to different mobility on polyacrylamide
gel electrophoresis. To address some of these problems, cleavable reagents are particularly
helpful. Upon cleavage, the molecule should return to its original molecular state, including
electrophoretic mobility. In most of the applications of cross-linking reagents a parallel experiment using the analogous monofunctional reagent should be carried out. These experiments
may elucidate whether chemical modification itself is the cause of the problem, such as loss of
the biological activity or immunogenicity. If such is the case a different reagent may be chosen.
Alternatively, protection of enzyme-active sites or immunological activities may be carried out
with inhibitors,91 substrates,92 or antigens.93,94
While sulfhydryl cleavable reagents have many advantages, there are also several disadvantages. Major advantages include (1) rapid cleavage of the disulfide bond under mild conditions,
(2) quantitative completion of the cleavage reaction, (3) ability to be cleaved both before and after
electrophoresis, and (4) the specificity of the reduction reaction. The disadvantages of the use of
these reagents are the following: (1) they are susceptible to disulfide exchange with the possibility
of linking noninteracting molecules; (2) their use precludes the application of reducing agents for
the isolation of cross-linked complexes; and (3) they cannot be used in a system that is sensitive to
oxidation and which would normally be kept under reducing conditions. Disulfide exchange usually involves the presence of free sulfhydryl groups that must be present in significant excess over
disulfides.95 This reaction can be decreased by lowering the pH of the reaction below the pKa of
For other cleavable reagents, there are also advantages and disadvantages. The major disadvantages with the use of glycol reagents include (1) the reduced rate of cleavage relative to
that achieved with disulfides, (2) the difficulty in achieving complete cleavage, (3) the lack of
specificity of the cleavage reaction, namely, the carbohydrate portions of glycoproteins can also
be disrupted, and (4) the oxidative side reaction of the carbohydrates may lead to potential formation of Schiff bases with the protein amino groups. A cross-linking reagent will have to be
carefully chosen to suit the particular system under investigation to reduce the complications to