doi:10.1006/prep.2001.1522, available online at http://www.idealibrary.com on
Overexpression, Purification, and Characterization of
Glutaminase-Interacting Protein, a PDZ-Domain
Protein from Human Brain
J. Carlos Aledo, Abel Rosado, Lucı´a Olalla, Jose´ A. Campos, and Javier Ma´rquez
1 Departamento de Biologı´a Molecular y Bioquı´mica, Laboratorio de Quı´mica de Proteı´nas,Facultad de Ciencias, Universidad de Ma´laga, 29071 Ma´laga, Spain
Received April 6, 2001, and in revised form August 1, 2001
Most of these interactions are brought about by
modu-A human brain cDNmodu-A clone coding for a novel PDZ- lar interaction domains, conserved sequence elements domain protein of 124 amino acids has been previously which recognize consensus motifs in their target poly-isolated in our laboratory. The protein was termed GIP
peptides. Among the most studied and
well-character-(glutaminase-interacting protein) because it interacts
ized binding domains are the Src homology (SH22and
with the C-terminal region of the human brain
glutami-SH3) (1), the phosphotyrosine-binding (2), and the
nase L. Here we report the heterologous expression of
pleckstrin homology domains (1, 3).
GIP as a histidine-tagged fusion protein in Escherichia
Recently, a new interaction domain, originally
de-coli cells. The induction conditions (temperature and
nominated GLGF (containing a conserved
Gly-Leu-Gly-isopropyl -D-thiogalactopyranoside concentrations)
Phe sequence element), was reported to be involved in
were optimized in such a way that GIP accounted for
the targeting of proteins to the cytoplasmic side of the
about 20% of the total E. coli protein. A simple and
plasma membrane (4, 5). These domains were initially
rapid procedure for purification was developed, which
found in the postsynaptic density protein PSD95/
yielded 17 mg of purified GIP per liter of bacterial cell
SAP90, the Drosophila lethal discs large-1 tumor
su-culture. The apparent molecular mass of the protein
pressor gene product, and the zonula occludens or tight
by SDS–PAGE was 16 kDa, whereas in native form it
junction protein ZO-1 (4, 6). For this reason, they were
was determined to be 28 kDa, which suggests dimer
renamed PDZ domains, where the acronym represents
formation. The nature and integrity of the recombinant
the initials of these three proteins (7).
protein were verified by mass spectrometry analysis.
PDZ domains are small protein interaction modules
The functionality of the GIP protein was tested with
composed of 80–100 amino acid residues. They have
an in vitro activity assay: after being pulled down with
been shown to mediate the distribution and clustering
glutathione S-transferase–glutaminase, GIP was
revealed by Western blot using anti-GIP antibodies. of proteins in membranes, acting as a scaffold where
Furthermore, the glutaminase activity in crude rat signaling molecules are linked to a multiprotein
com-liver extracts was inhibited by the presence of recombi- plex. The PDZ domain specifically binds to the
C-termi-nant purified GIP protein. 䉷2001 Academic Press
2Abbreviations used: GIP, glutaminase-interacting protein; GST,
glutathione S-transferase; MALDI-TOF, matrix-assisted laser desorption/ionization time-of-flight; MS, mass spectrometry; PDZ, postsynaptic density protein PSD-95/SAP90, Drosophila tumor
sup-Protein–protein interactions mediate the transfer of
pressor protein DLG, and tight junction protein ZO-1; PAG,
phos-information in many signaling systems and the
tar-phate-activated glutaminase; SH2, Src-homology 2 domain; SH3,
geting of proteins to specific membrane localizations. Src-homology 3 domain; SDS–PAGE, sodium dodecyl sulfate–
polyacrylamide gel electrophoresis; FPLC, fast protein liquid chroma-tography; IPTG, isopropyl-D-thiogalactopyranoside; ECL, enhanced chemiluminescence; BSA, bovine serum albumin; TPBS, 0.05%
1To whom correspondence should be addressed. Fax: (34)
952-132000. E-mail: [email protected]. Tween 20 in phosphate-buffered saline; IEF, isoelectric focusing.
1046-5928/01 $35.00 411
Copyright䉷2001 by Academic Press All rights of reproduction in any form reserved.
nal residues of its target protein; a consensus binding BamHI–HindIII sites of pQE–31. A pGEX–GIP fusion
construct was also engineered by releasing the GIP peptide X-Ser/Thr-X-Val has been proposed, where Val
is the C-terminal residue (8). Nevertheless, other C- cDNA from pJG4–5GIP with EcoRI. This insert was gel-purified and ligated in-frame in the EcoRI site of terminus motifs distinct from the consensus sequence
have also been identified as PDZ-binding motifs (9). the expression vector pGEX–6p-1. Since the first works reporting the importance of these
homologous domains, the number of PDZ-domain-con- Expression of the Recombinant Protein
taining proteins has increased rapidly and techniques
The 124 residues encoded by the predicted open read-of sequence analyses have shown that PDZ-domain
ho-ing frame of GIP plus 13 additional amino acids at the mologues are present in yeast, plants, and bacteria (10).
NH2-termini (including a 6⫻ His tag) were expressed
We have cloned a cDNA encoding for a novel
PDZ-in Escherichia coli cells. The orientation and sequence domain protein using an in vivo selection screen, the
of the GIP cDNA in the expression vectors were con-yeast two-hybrid system, and the C-terminus of human
firmed by DNA sequencing, as described elsewhere (13). phosphate-activated glutaminase (PAG) as bait (11).
E. coli BL21(DE3) cells (Stratagene) transformed with
This protein was termed GIP (glutaminase-interacting
pQE–GIP were cultured in 240 ml of 2⫻YTA medium protein), because it interacts with the L isoenzyme of
(tryptone 16 g/L, yeast extract 10 g/L, NaCl 5 g/L, and human brain PAG (11). In brain, PAG is responsible
ampicillin 100 g/ml) at 37⬚C with vigorous shaking for the synthesis of neurotransmitter glutamate; two
and adequate aeration, until the optical density reached isoenzymes, named the L and K types, are expressed
1.0 unit at 600 nm. Then, isopropyl--D
-thiogalactopyr-in human bra-thiogalactopyr-in tissue (12). Here we report the
overex-anoside (IPTG) was added to a final concentration of pression and purification of GIP, as well as a first
molec-0.1 mM, and the incubation was continued for an addi-ular characterization of the purified protein.
tional 5-h period at 23⬚C. Afterward, the culture was centrifuged at 10,000 rpm in a Beckman JA-10 rotor
MATERIALS AND METHODS
for 10 min at 4⬚C. The drained pellet can be stored at
Materials ⫺20⬚C for up to 2 weeks or subjected to subsequent
cell disruption. The fast protein liquid chromatography (FPLC)
sys-tem and the chelating Sepharose fast flow, CM
Sepha-Purification of Recombinant GIP
rose fast flow, and Superose 6 chromatography media
were from Amersham-Pharmacia Biotech. Reagents for The harvested cellular pellet was suspended by add-SDS–PAGE and detergents (TX-100, NP-40, and ing 0.5 ml of ice-cold buffer I (20 mM Na
2HPO4, 0.5 M
Tween-20) were from Roche. HPLC ultrapure water, NaCl, 10 mM imidazole, pH 7.4) per milliliter of culture. generated by a Milli-RO 60 coupled to a Milli-Q water This cell suspension was sonicated on ice with four 30-purification system (Millipore), was used to prepare the s pulses separated by 1-min intervals at a setting of aqueous solutions. The expression vectors pQE–31 and 4–8 m from peak to peak. After that, TX-100 was pGEX–6p-1 were purchased from Qiagen and Amers- added to a final concentration of 1% (w/v). The suspen-ham-Pharmacia Biotech, respectively. Custom-made sion was incubated for 30 min at 4⬚C with end-over-single-stranded oligonucleotide primers were obtained end mixing to help in solubilization of the recombinant from Amersham-Pharmacia Biotech. protein. The cell debris was removed by centrifugation; the supernatant was filtered through a 0.2-m filter
Recombinant DNAs pore size (Sartorius) and applied to a Ni-chelating
Seph-arose fast flow column (0.5⫻2.5 cm) equilibrated with The 1.3-kb cDNA encoding the GIP was originally
cloned into the pJG4-5 vector (11). The resultant con- buffer I. The column was washed with the same buffer and then eluted with a linear imidazole gradient (0.01– struction, designated as pJG4–5GIP, was used as
tem-plate DNA for PCR. The oligonucleotide primers 0.5 M) in buffer I at a flow rate of 30 ml/h. The total volume of the gradient was 20 ml. Fractions of 0.5 ml designated to amplify the insert corresponding to the
coding region of GIP were primer BamHI (forward): were analyzed by SDS–PAGE; those showing recombi-nant GIP were pooled and concentrated to 0.4 ml by 5⬘
-TGGGCGACCAGAGCAGGGATCCGATGTCCTA-CATCCCGGG-3⬘ and primer BCO2 (reverse): 5⬘-GA- ultrafiltration with a 10-kDa Microsep microconcentra-tor. Afterward, the sample was diluted with 3.6 ml of 50 CAAGCCGACAACCTTGATTGGAG-3⬘. The PCR
prod-uct contained a BamHI site next to the start codon mM Mes buffer, pH 6.2, and applied to a CM–Sepharose column (0.7 ⫻ 2.5 cm), previously equilibrated with ATG, introduced by the forward primer, and a HindIII
restriction site within the 3⬘-untranslated region, 490 buffer II (50 mM NaCl, 50 mM Mes, pH 6.2). The column was washed through with the same medium and then bases downstream from the stop codon TAG. A
high-expression recombinant plasmid, pQE–GIP, was con- eluted with 20 ml of a linear NaCl gradient (50–500 mM) at a flow rate of 60 ml/h; 0.5-ml fractions were structed by inserting the amplified fragment into the
collected. Levels of GIP protein throughout the purifica- SDS–PAGE and Analytical Determinations
tion procedure were estimated by densitometric analy- SDS–PAGE was performed as described by Laemmli ses, either after SDS–PAGE and Coomassie staining (16) using 12% isocratic polyacrylamide gels with a 3% or after Western blotting and processing by enhanced polyacrylamide stacking gel. Apparent molecular chemiluminescence (ECL), using purified and quanti- masses of proteins were determined by using the low-fied GIP as a standard. molecular-weight calibration kit (Amersham-Phar-Glutathione S-transferase (GST) fusion proteins of macia Biotech), containing phosporylase b (94 kDa), GIP and the C-terminal region of PAG (amino acids BSA (67 kDa), ovalbumin (43 kDa), carbonic anhydrase 347–602) were also expressed in BL21(DE3) cells from (30 kDa), trypsin inhibitor (20.1 kDa), and␣ -lactalbu-pGEX–6p-1 (11) and pGEX–4t-1 (13) plasmid con- min (14.4 kDa). The gels were stained with Coomassie structs, respectively. The basic induction and purifica- blue or with silver according to the method of Merril et tion protocols using glutathione–Sepharose beads were al. (17). Samples of purified GIP were also analyzed by followed, as described previously (11, 13). denaturing isoelectric focusing (IEF) in mini slab gels. The 5% polyacrylamide gel contained 1.5% (v/v)
ampho-Antibody Production and Purification line (pH 3.5–10, LKB) and 8 M urea. Samples were
diluted in 20l of sample buffer containing 8 M urea. Purified recombinant GST–GIP fusion protein was
Focusing was carried out at 200 V, 10 mA, for 4 h. The used as an antigen and injected subcutaneously into
high isoelectric point ( pI) markers kit from pH 5–10.5 two New Zealand white rabbits. Polyclonal antibodies
was used (Amersham-Pharmacia Biotech). Gel filtra-were generated as per Segura et al. (14), using 1 mg of
tion chromatography was done with a Superose-6 col-protein in each injection. Decomplemented serum was
umn coupled to the FPLC system and equilibrated with affinity purified against the purified GIP protein bound
a buffer composed of 20 mM NaH2PO4, 0.5 M NaCl,
to BrCN-activated Sepharose (Pharmacia), as
recom-pH 7.4. The molecular mass of the purified GIP was mended by the supplier.
evaluated by size-exclusion chromatography. The col-umn was calibrated with the following standards
Western Blot Analysis (Sigma): BSA (67 kDa), ovalbumin (43 kDa),
chymo-trypsinogen A (25 kDa), ribonuclease A (15.6 kDa), and Western blotting was carried out essentially as
de-cytocrome c (12 kDa). Salmon sperm DNA (Sigma) was scribed by Towbin et al. (15). Briefly, samples were
sub-used as a marker of the column void volume. Protein jected to SDS–PAGE and then transferred to
nitrocellu-concentrations were determined by the Bradford lose. After being blocked with 3% (w/v) bovine serum
method (18). Glutaminase was solubilized from crude albumin (BSA) in TPBS buffer (0.05% (v/v) Tween-20 in
homogenates of rat liver with TX-100, according to the PBS), the filter was incubated for 1 h with the indicated
method of Haser et al. (19). Hepatic PAG activity was dilution of the anti-GIP antibody in TPBS containing
measured with the stopped glutamate assay (20) for 10 1% (w/v) BSA. After three 15-min washes in TPBS,
min at 37⬚C, using the following incubation medium: the filter was incubated with a secondary anti-rabbit
10–50 mM glutamine, 10–100 mM potassium phos-IgG–horseradish peroxidase conjugate. Finally, the blot
phate, 0.2 mM EDTA, 2 mM NH4Cl, 20 mM Tris/HCl,
was washed another three times with TPBS and
proc-pH 8.0. Blanks with sample and phosphate omitted essed for enhanced chemiluminiscence with the ECL
were run in parallel. system, as per the manufacturer’s instructions
(Amers-ham). A prestained protein marker kit (broad range,
MALDI-TOF MS Analyses
6–175 kDa) from New England Biolabs was used as a
MALDI-TOF mass spectrometry was performed on control of the protein transfer and for estimation of the
purified samples of recombinant GIP on a Bruker Biflex molecular mass.
TOF mass spectrometer using sinapic acid as a matrix. Spectra were obtained in the linear mode at an
acceler-In Vitro Pull-Down Assays
ating voltage of 20 kV. A nitrogen laser with an emission wavelength of 337 nm was used. The samples of GIP Purified recombinant GIP was incubated, for 2 h at
room temperature, with purified GST–PAG or GST were additionally cleaned and concentrated for the analysis using a C18resin as described by Villanueva et
alone. Afterward, GST fusion proteins were pulled
down using glutathione–Sepharose beads. The beads al. (21). The masses of the peptide ions were calibrated
using myoglobin as an internal standard. were washed four times with binding buffer (50 mM
Tris–HCl, pH 8.0, 150 mM NaCl, 1 mM EDTA, 5 mM
RESULTS AND DISCUSSION
dithiothreitol, 0.1% (w/v) Triton X-100) and suspended
Overproduction of Human Brain GIP in E. coli
in SDS gel loading buffer. After being boiled for 5 min,
the eluted proteins were analyzed by SDS–PAGE and The T5 promoter expression system yielded from 200 to 250 mg of soluble GIP protein per liter of bacterial Western blotting.
were set up for the heterologous expression of recombi-nant human GIP.
Purification of Recombinant GIP
Purification of recombinant GIP is summarized in Table 1, where the yield and enrichment fold through-out the different steps of the purification schedule are indicated. Total protein was determined by the Brad-ford method while the amount of GIP protein, in ex-tracts and partially purified samples, was estimated by densitometric analyses after SDS–PAGE, as described in the previous section.
The soluble E. coli extract containing approximately 67.8 mg of GIP in 365 mg of total protein (Table 1), was immediately submitted to metal-chelate affinity chromatography on an Ni–sepharose column (Fig. 2A),
FIG. 1. SDS–PAGE of soluble E. coli extracts: optimization of GIP synthesis by pQE–GIP. The effect of temperature on the level of GIP soluble protein at optimal (0.1 mM) IPTG concentration was evaluated. Both S (soluble) and P (pellet) fractions of bacterial cul-tures induced at 22 and 37⬚C are shown.
culture. Only the soluble fraction of the E. coli extract was used for purification of GIP. Figure 1 shows Coo-massie-stained SDS–PAGE of GIP in bacterial extracts transformed with the expression vector pQE–GIP. By densitometric analyses of the gels, we estimated that GIP accounts for 27.5% of the total soluble E. coli pro-tein for the pQE–GIP construct. The expression of re-combinant GIP was optimized in terms of IPTG concen-trations and temperature of the induction. At 22⬚C, no clear effect was seen by varying the IPTG concentra-tions from 0.1 to 10 mM. However, at 37⬚C, the amount of GIP expressed at 0.1 mM reached a maximum and did not increase further at higher IPTG levels (results not shown). We then checked the effect of temperature on the amount of soluble GIP synthesized. A clear effect was seen after separation of soluble and pellet fractions of total bacterial extracts: the amount of soluble GIP increased about three times at 37⬚C with regard to the levels found at 22⬚C (Fig. 1). In agreement with these results, 37⬚C and 0.1 mM IPTG induction conditions
TABLE 1
FIG. 2. (A) SDS–PAGE analysis after metal-chelate affinity
chro-Purification of Recombinant GIP from E. coli BL21 matography on an Ni–Sepharose column. Fractions from the
Ni-(DE3)/pQE–GIP Cells column containing about 10–25g of protein were analyzed by SDS– PAGE, as described under Materials and Methods. The column was Total Fraction Enrich- eluted with an imidazol gradient from 10 to 500 mM. The gel was protein GIP GIP/total Yield ment stained with Coomassie blue R-250. Lane 1, molecular mass markers, Step (mg) (mg) prot. (%) (fold) with the masses indicated on the left side of the gel; lane 2, crude bacterial extract; lane 3, flow-through fraction; lanes 4–10, fractions Crude extract 365 67.8 0.185 100 1
of the imidazole gradient. (B) SDS–PAGE pattern of the final purified Soluble extract 210 57.9 0.275 85.4 1.4
preparation of GIP. Lane 1, molecular mass markers, the same as in Affinity Ni-column 12.7 7.8 0.614 11.5 3.3
A; lane 2, pooled fractions after cation-exchange chromatography. Cation-exchange 4.2 4.2 1.000 6.2 5.4
taking advantage of the 6⫻His tag added to GIP in the purification schedule was cation-exchange chromatog-raphy on carboxymethyl Sepharose Fast Flow. We first NH2-termini. GIP was bound to the gel and eluted with
a gradient of imidazol buffer: representative fractions concentrated the sample by ultrafiltration and changed the buffer to eliminate the high NaCl concentration (0.5 corresponding to different imidazol concentrations of
the gradient are shown in lanes 4–10 of the SDS–PAGE M) present in the sample. GIP was strongly bound to the resin at pH 6.2, in agreement with its highly basic gel (Fig. 2A). As can be seen, gradient elution allows
for an effective GIP purification. Although GIP starts pI of 9.3, determined by denaturing IEF of the purified
protein. After the unbound anionic contaminant pro-to elute at about 0.15 M imidazole (fraction 12, lane 4
of the gel), the peak is mainly concentrated through teins were washed, GIP eluted around 0.3 M of NaCl as a homogeneous protein. From the approximate 67.8 fractions 28 (0.35 M, lane 7) to 40 (0.5 M, lane 10);
these fractions were pooled for the next chromato- mg of recombinant GIP present in the initial 365 mg of crude bacterial extracts, we finally isolated 4.2 mg graphic step. The high level of purification achieved at
the end of the gradient would allow us to end purifica- of pure GIP accounting for 6.2% of the yield (Table 1). Although at first this may seem a low level of purified tion by selecting only these fractions. However, to
im-prove the yields of the protocol in terms of protein GIP, we should note that about 4 mg of protein was purified from a 240-ml culture; the schedule can be amount, fractions with lower levels of purification, but
still having a high GIP content, were also pooled and easily scaled up allowing the isolation of preparative amounts of GIP for structural analyses. We usually subjected to further purification.
After affinity chromatography, the last step of the pooled five 0.5-ml fractions; thus, the volume of the final
FIG. 3. MALDI mass spectra of recombinant human GIP. The two major peaks correspond to the singly (m/z⫽15,304) and doubly charged (m/z⫽7654) peptide ions. Components with higher mass could be ascribed to protein/matrix adducts. Y-axis: a.i., absolute intensity (arbitrary units); X-axis: m/z, mass/charge.
preparation was about 2.5 ml, which gives a protein concentration of approximately 1.7 mg/ml. The silver-stained SDS gel of this purified protein preparation is shown in Fig. 2B. A clear single band corresponding to an apparent molecular mass of about 16 kDa was observed for the purified GIP, as determined from a plot of log molecular mass vs Rg, the relative mobility. This result is in good agreement with the predicted Mr
of 15,302.4 calculated for the 137-residue polypeptide.
Characterization of Recombinant GIP
Although the engineered cDNA clone used in this work to direct the synthesis of GIP has been checked by DNA sequencing, the recombinant protein product
FIG. 4. Glutaminase binding assay. Purified GIP was evaluated by
was also carefully characterized. By MALDI-TOF mass
its capacity to interact with the C-terminal region of
phosphate-spectrometry, recombinant GIP purified according to activated glutaminase (PAG) by using an in vitro pull-down assay the method described here gave a sharp major peak (11). GST–PAG or GST alone was expressed in E. coli and purified with a mass of 15,304 ⫾ 1 Da (Fig. 3), corresponding by glutathione–Sepharose. The affinity-purified GST proteins immo-bilized on glutathione–Sepharose beads were incubated with purified
to the molecular ion [(M⫹H)+], almost identical to the
recombinant GIP. Bound proteins were analyzed by SDS–PAGE and
predicted mass for the singly charged ion of 15,303.4
Western blotting, revealed with anti-GIP antibodies, and visualized
Da. A second peak corresponding to the doubly charged by fluorography. Left lane, control lane with only recombinant GST; [(M ⫹ 2H)2+] peptide ion also appeared at a m/z ratio
right lane, assay in the presence of GST–PAG. The relative positions
of half of the major peak (Fig. 3). Thus, the nature and of the prestained molecular mass markers are indicated on the right side of the film.
integrity of the recombinant expressed protein were verified. Moreover, no differences in the MS spectra were observed after reduction of the protein with 20
mM dithiothreitol (results not shown), which indicates Student test) than control values (100%) in the absence of exogenous GIP. Thus, purified GIP seems to be able the absence of oxidized methionine residues. The MS
analysis also confirmed the methionine residue at the to recognize PAG even in crude tissue extracts. Taken together, the pull-down interaction assay and the en-N-terminal, which is thus not removed in vivo from the
N-terminus of the neosynthesized protein in E. coli zyme activity assay argue in favor of a fully functional recombinant purified GIP. Furthermore, concerning the cells.
Being a PDZ protein, GIP activity should be evalu- physiological role of GIP, we are tempted to speculate whether it functions to destabilize the glutaminase ated by testing its capacity to interact with its target
C-termini motif present in human brain L-PAG. The multimer, a role similar to that ascribed to the similarly sized protein inhibitor of neuronal nitric oxide synthase functionality of recombinant GIP has been studied by
using it in a pull-down assay against recombinant (22), although in this case the interaction is not medi-ated by PDZ domains (23). The stoichiometry of native GST–PAG. The assay was a variant of that previously
reported employing in vitro translated [35S]GIP (11). As human L-PAG enzyme is unknown, and the possibility
of aggregation to highly active polymeric forms has not shown in Fig. 4, the purified recombinant GIP was able
to interact properly with the GST–PAG: it associates been studied either, although it was early demonstrated
in vitro for the rat K-PAG enzyme in the presence of the
specifically with GST–PAG, but not with the control
GST alone. The presence of pulled-down GIP was re- activator inorganic phosphate (24). Therefore, whether human L-PAG self-associates to active multimeric vealed by Western blot using anti-GIP antibodies (Fig.
4). Thus, the PDZ domain in the recombinant purified forms, as well as the possibility that binding of GIP to PAG destabilizes the structure of PAG and thereby GIP was shown to be functional.
In order to assess the effect of purified GIP on the inhibits the enzyme activity, a mechanism analogous to that described for neuronal nitric oxide synthase, catalytic activity of PAG, we developed PAG assays in
the absence or presence (0.4 mg/ml) of added GIP. As remains to be established.
Finally, the molecular mass of the purified GIP was a source of PAG activity, rat liver extracts were used,
taking advantage of the high degree of similarity be- determined by FPLC in order to assess the stoichiome-try of the native protein. Gel filtration chromatography tween human brain L-PAG and the rat liver enzyme
(13). For five different experiments, the mean PAG ac- on a Superose 6 column gave an apparent molecular mass of 28 kDa (Fig. 5), indicating that the protein is tivity value in the presence of added GIP was 72.2 ⫾
FIG. 5. Determination of the molecular mass of native GIP by size-exclusion chromatography. Purified recombinant GIP was applied to a gel filtration Superose 6 column and eluted to a flow rate of 24 ml/h. The column was calibrated with the molecular mass standard proteins indicated in the figure. Salmon sperm DNA was used as a marker of the column void volume. A molecular mass of 28,000 Da was calculated for GIP from its partition coefficient, Kav.
that the predicted theoretical molecular mass of GIP is basis for these interactions is unknown. In conclusion, here we have reported a bacterial system for overex-15,302 Da, and the molecular mass in SDS gels has
been shown to be approximately 16 kDa, the estimation pression of GIP and a convenient scheme for purifica-tion to homogeneity of large quantities of active protein, by size-exclusion chromatography is consistent with
GIP being a dimer formed by two identical monomers. useful for further biochemical and biophysical studies. This observation agrees with the behavior shown by
native GIP in ultrafiltration experiments with 30-kDa
ACKNOWLEDGMENTS
cutoff microconcentrators: most of the GIP protein
re-mained in the concentrated solution. Furthermore, GIP This study was supported by Research Grant SAF 98-0153 of the Spanish CICYT. We thank the Fundacio´n Cientı´fica Ramo´n Areces
is mainly composed of just one PDZ domain and these
for a predoctoral fellowship to JAC. We are indebted to Professor I.
have been previously shown to be essentially globular
Nu´n˜ez de Castro for critical reading of the manuscript and to F. Canals
domains (25, 26); therefore, it cannot be argued that the
and the Servei de Proteo`mica i Bioinforma¨tica from the Universitat
value obtained by gel filtration has been overestimated, Auto´noma of Barcelona for the MALDI-TOF MS analysis of the due to a departure of the overall shape from a sphere. protein.
On the other hand, although single PDZ domains also appear to be functional, the members of the PDZ family
REFERENCES
usually have two or more copies of this motif, in
agree-ment with their role of scaffold proteins that allow the 1. Pawson, T. (1995) Protein modules and signalling networks. Na-anchor and ensemble of multiprotein complexes (4, 6). ture 373, 573–580.
As far as we are aware, GIP is the shortest member of 2. Bork, P., and Margolis, B. (1995) A phosphotyrosine interaction
this family, containing only one PDZ domain that al- domain. Cell 80, 693–694.
most represents its whole primary structure. This may 3. Saraste, M., and Hyvonen, M. (1995) Pleckstrin homology do-mains: A fact file. Curr. Opin. Struct. Biol. 5, 403–408.
explain why GIP needs self-association to fulfill its role
4. Ponting, C. P., Phillips, C., Davies, K. E., and Blake, D. J. (1997)
as an adaptor to couple interacting proteins.
Aggrega-PDZ domains: Targeting signalling molecules to
sub-membra-tion of PDZ-domain proteins has been reported for
nous sites. BioEssays 19, 469–479.
Homer proteins, where the coiled coil C-terminal region
5. Kim, S. K. (1995) Tight junctions, membrane-associated
guanyl-seems to be essential for oligomerization (27). In addi- ate kinases and cell signaling. Curr. Opin. Cell Biol. 7, 641–649. tion, it has been reported that some PDZ domains can 6. Fanning, A. S., and Anderson, J. M. (1999) PDZ domains: Funda-dimerize while simultaneously binding to the C-termi- mental building blocks in the organization of protein complexes
at the plasma membrane. J. Clin. Invest. 103, 767–772.
7. Kennedy, M. B. (1995) Origin of PDZ (DHR, GLGF) domains. quantitation of microgram quantities of protein utilizing the prin-ciple of protein-dye binding. Anal. Biochem. 72, 248–254. Trends Biochem. Sci. 20, 350.
19. Haser, W. G., Shapiro, R. A., and Curthoys, N. P. (1985) Compari-8. Kornau, H. C., Schenker, L. T., Kennedy, M. B., and Seeburg, P.
son of the phosphate-dependent glutaminase obtained from rat H. (1995) Domain interaction between NMDA receptor subunits
brain and kidney. Biochem. J. 229, 399–408. and the postsynaptic density protein PSD-95. Science 269, 1737–
20. Lund, P. (1985)L-Glutamine andL-glutamate. UV-method with 1740.
glutaminase and glutamate dehydrogenase, in “Methods of Enzy-9. Songyang, Z., Fanning, A. S., Fu, C., Xu, J., Marfatia, S. M.,
matic Analysis” (H. U. Bergmeyer, Ed.), Vol. 8, pp. 357–363, Chishti, A. H., Crompton, A., Chan, A. C., Anderson, J. M., and
Verlag Chemie, Weinheim. Cantley, L. C. (1997) Recognition of unique carboxyl-terminal
21. Villanueva, J., Canals, F., Querol, E., and Avile´s, F. X. (2001) motifs by distinct PDZ domains. Science 275, 73–77.
Monitoring the expression and purification of recombinant pro-10. Ponting, C. P. (1997) Evidence for PDZ domains in bacteria,
teins by MALDI-TOF mass spectrometry. Enzyme Microbial yeasts, and plants. Protein Sci. 6, 464–468.
Technol. 29, 99–103.
11. Olalla, L., Aledo, J. C., Bannenberg, G., and Ma´rquez, J. (2001) 22. Jaffrey, S. R., and Snyder, S. H. (1996) PIN: An associated protein The C-terminus of human glutaminase L mediates association inhibitor of neuronal nitric oxide synthase. Science 274, 774–777. with PDZ domain-containing proteins. FEBS Lett. 488, 116–122. 23. Fan, J-S., Zhang, Q., Li, M., Tochio, H., Yamazaki, T., Shimizu, 12. Aledo, J. C., Go´mez-Fabre, P. M., Olalla, L., and Ma´rquez, J. M., and Zhang, M. (1998) Protein inhibitor of neuronal nitric-(2000) Identification of two human glutaminase loci and tissue- oxide synthase, PIN, binds to a 17-amino acid residue fragment specific expression of the two related genes. Mammal. Genome of the enzyme. J. Biol. Chem. 273, 33472–33481.
11, 1107–1110. 24. Godfrey, S., Kuhlenschmidt T., and Curthoys, N. P. (1977)
Corre-13. Go´mez-Fabre, P. M., Aledo, J. C., del Castillo-Olivares, A., Alonso, lation between activation and dimer formation of rat renal phos-F. J., Nu´n˜ez de Castro, I., Campos, J.A., and Ma´rquez, J. (2000) phate-dependent glutaminase. J. Biol. Chem. 252, 1927–1931. Molecular cloning, sequencing and expression studies of the hu- 25. Doyle, D. A., Lee, A., Lewis, J., Kim, E., Sheng, M., and MacKin-man breast cancer cell glutaminase. Biochem. J. 345, 365–375. non, R. (1996) Crystal structures of a complexed and peptide-free membrane protein-binding domain: Molecular basis of peptide 14. Segura, J. A., Aledo, J. C., Go´mez-Biedma, S., Nu´n˜ez de Castro, I.,
recognition by PDZ. Cell 85, 1067–1076. and Ma´rquez, J. (1995) Tumor glutaminase purification. Protein
26. Tochio, H., Hung, F., Li, M., Bredt, D. S., and Zhang, M. (2000) Express. Purif. 6, 343–351.
Solution structure and backbone dynamics of the second PDZ 15. Towbin, H., Staehelin, T., and Gordon, J. (1979) Electrophoretic
domain of postsynaptic density-95. J. Mol. Biol. 295, 225–237. transfer of proteins from polyacrylamide gels to nitrocellulose
27. Kato, A., Ozawa, F., Saitoh, Y., Fukazawa, Y., Sugiyama, H., and sheets: Procedure and some applications. Proc. Natl. Acad. Sci.
Inokuchi, K. (1998) Novel members of the Ves1/Homer family of USA 76, 4350–4354.
PDZ proteins that bind metabotropic glutamate receptors. 16. Laemmli, U. K. (1970) Cleavage of structural proteins during the J. Biol. Chem. 273, 23969–23975.
assembly of the head of bacteriophage T4. Nature 227, 680–685.
28. Xu, X. Z., Choudhury, A., Li, X., and Montell, C. (1998) Coordina-17. Merril, C. R., Goldman, D., and Van Keuren, M. L. (1984) Gel tion of an array of signaling proteins through homo- and hetero-protein stains: Silver stain. Methods Enzymol. 104, 441–447. meric interactions between PDZ domains and target proteins.
J. Cell Biol. 142, 545–555. 18. Bradford, M. M. (1976) A rapid and sensitive method for the
