The Fanconi Anemia Core Complex Forms Four Complexes of Different Sizes in Different Subcellular Compartments*

Fanconi anemia (FA) is an autosomal recessive disease marked by congenital defects, bone marrow failure, and cancer susceptibility. FA cells exhibit a characteristic hypersensitivity to DNA crosslinking agents such as mitomycin C. The molecular mechanism for the disease remains elusive, but at least 6 FA proteins are known to be part of what is termed the FA core complex. We used affinity pulldown of FLAG-FANCA to pull down the FA complex from whole-cell extracts. Mass spectroscopy detected previously reported FA-binding proteins, including FANCA, FANCC, FANCG, cdc2, and GRP94, thus validating the approach. We further describe a method of purification of the FA core complex in an effort to find novel complex components and biochemical activity to define the function of the complex. By using conventional chromatographic fractionation of subcellular preparations, we report: (i) the FA core complex exists in a cytoplasmic form at 500–600 kDa; (ii) a larger, 750-kDa cytoplasmic form is seen only at mitosis; (iii) a nuclear form achieves a size of 2 megaDaltons; and (iv) a distinct 1-megaDalton FA core complex exists bound to chromatin that contains phosphorylated FANCA after undergoing DNA damage. We are continuing our analysis using mass spectroscopy in an effort to characterize novel binding proteins. These data will help define the biochemical role of the FA core complex in normal cell physiology as well as in the development of the FA disease state.

Fanconi anemia (FA) 1 is a genetic disease of cancer susceptibility marked by congenital defects, bone marrow failure, and myeloid leukemia (1)(2)(3)(4)(5). To date, at least 11 complementation groups have been defined (6 -8), and eight genes have been cloned (9 -18). However, none of the gene products resemble any known proteins, and all have few identifiable functional protein motifs.
Cells derived from patients with the disease exhibit characteristic hypersensitivity caused by DNA crosslinking agents and generalized decreased survival (19 -23). In addition, a well described G 2 -phase cell-cycle delay has also been described that is thought to be secondary to a defective S or G 2 checkpoint (24 -26). Others have implicated cytokine signaling, apoptotic and oxidative damage defects (19,(27)(28)(29). However, no defined biochemical mechanism for crosslinker hypersensitivity has been elucidated. Patient and cellular phenotypes across all of the complementation groups are similar, suggesting an interrelatedness or cooperativity between the FA proteins.
This cooperativity has been borne out by work we have done in showing binding of FANCA and FANCC in a protein complex in both nucleus and cytoplasm (30 -32). Subsequent work has found the FANCE, FANCF, FANCG, and FANCL proteins in the complex as well (33)(34)(35)(36)(37)(38). Although reports describing binding partners to the FA core complex, including GRP94, cdc2, STAT1, and FAZF, have been published, little progress has been made in providing a unifying theory of FA protein function (27,39,40).
The core complex of FA described above fails to form in all of the complementation groups except FA-D2. The FANCD2 protein is ubiquitinated in response to DNA damage and in Sphase, which is dependent upon the existence of the FA core complex. In addition, FANCD2 binds to BRCA1, and FANCD1 is BRCA2, providing links to better described pathways of DNA repair and genome surveillance, including homologous recombination (17,41).
FANCD2 ubiquitination is not the only modification that has been shown to be functionally important. Several reports have shown that FANCA phosphorylation is vital to FA pathway function and is missing in all but FA-D1 and FA-D2 cells (32,42,43).
A recent study details the purification of a BLM-containing FA core complex, using FANCA antibody, in which several sizes of complex were noted in a whole-cell extract (44). Helicase activity was also coprecipitated. This approach yielded the latest cloned FA gene, FANCL (18). The coprecipitation of FA proteins with another genomic instability syndrome protein, BLM, puts FA at the nexus of DNA repair syndromes.
In this study, we attempt to validate a more conventional approach in concert with modern methods to purify the FA core complex. We use a series of chromatographic methods coupled with an immunoaffinity step to characterize the FA core complex with respect to size and composition. In doing so, we show a means to identify complex components that not only can shed light upon normal FA function but also has the potential to demonstrate biochemical activity. We also demonstrate that the FA core complex is not simply one complex; it is actually composed of at least four distinct complexes, depending upon its subcellular localization. Purification also allows us to isolate a phosphorylated form of FANCA, which is DNA-damage inducible.

MATERIALS AND METHODS
Cell Culture-Cells were grown at 37°C in a 5% CO 2 incubator. HeLa cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, whereas FA-A mutant GM6914 cells were grown in 15% fetal bovine serum. Each cell line was transduced with pMMP-FLAG-FANCA or pMMP vector alone, and the resulting cells were selected with puromycin. Cells were treated for 24 h in 0.1 M mitomycin C (MMC), a dose producing a 90% kill in mutant cells and a 10% kill in wild-type cells.
Synchronization-HeLa cells were treated overnight in the presence of 1 M nocodazole (Sigma). Mitotic cells were collected the next morning by shaking the plate and collecting the cells in suspension. Verification of the cell-cycle state was achieved by using FACScan analysis (BD Biosciences) (1).
Cell Fractionation-Cells were collected similarly and resuspended in 10 mM Hepes, pH 7.4, 40 mM KCl, and 2 mM MgCl 2 . After 20 -30 strokes on ice with a Dounce homogenizer, the cytoplasmic extract was separated from the nuclei by spinning at 1500 rpm. The cytoplasmic extract was clarified by spinning at high speed. The nuclear pellet was further extracted in the lysis buffer using 500 mM NaCl as above.
Chromatin Preparation-Procedures for permeabilization and sequential subnuclear extraction were adapted from the methods of Burtelow et al. (45) and Reyes et al. (46) and are detailed in our previous work (48). In brief, cells pelleted (100 l) from one large plate were resuspended and permeabilized in 5 ml of low salt buffer (10 mM Hepes, pH 7.4, 10 mM KCl, 50 g/ml digitonin) containing protease and phosphatase inhibitors (2 g/ml leupeptin, 1 g/ml pepstatin, 1 g/ml aprotinin, 1 mM phenylmethylsulfonyl fluoride, and 1 mM Na 3 VO 4 ) for 15 min at 4°C. After washing, the nuclei were resuspended in 200 l of permeabilization buffer containing 30 units of DNase I (RNase Free, Roche Applied Science) for 15 min at room temperature and an additional 15 min at 37°C. Chromatin proteins were extracted by adding extraction buffer (1% Triton X-100, 50 mM Hepes, pH 7.4, 150 mM NaCl, 30 mM sodium phosphate, 10 mM NaF, and 1 mM EDTA) containing protease and phosphatase inhibitors for 10 min at 4°C. Supernatant collected at 14,000 rpm in a microfuge for 10 min was termed the chromatin fraction.
SDS-Page and Immunoblotting-SDS-PAGE analysis was followed by gel transfer in 25 mM Tris and 200 mM glycine onto nylon-supported nitrocellulose. Filters were blocked for 1 h in 5% bovine serum albumin in Tris-buffered saline and were incubated in TBS plus Tween 20 containing primary antibody for at least 1 h at room temperature. Filters were then washed in TBS plus Tween 20, incubated with horseradish peroxidase-linked secondary antibody (Amersham Biosciences), washed again, and visualized by enzyme-linked chemiluminescence (30).
Immunoprecipitation-Lysates from FA-A mutant either infected with vector only or with FLAG-FANCA were immunoprecipitated with FLAG-affinity gel. After extensive washing, the beads were incubated with 100 g/ml FLAG peptide in Tris-buffered saline. The resulting eluates were either concentrated and buffer-exchanged into 100 mM NaHCO 3 (for mass spectroscopy) or simply concentrated and kept in TBS for SDS-PAGE (using Centricon 100-kDa cutoff filters).
Mass Spectroscopic Peptide Analysis-Mass spectroscopy was performed as described previously. Briefly, the eluates from the FLAG immunoprecipitations were dialyzed, reduced, and subjected to proteolysis with trypsin. Digestion was stopped by the addition of 5 l of glacial acetic acid, and the samples were stored at Ϫ35 C. Experiments utilized an HPLC (Applied Biosystems or Agilent HPLC) gradient composed of 0.1 M acetic acid and 70% acetonitrile with 0.1 M acetic acid. Tryptic peptides from solution digestion were analyzed by a method described previously, using nanoflow-HPLC/microelectrospray ionization interfaced to a ThermoFinnigan LCQ ion-trap mass spectrometer. Data-dependant analysis was performed in which one full MS scan from m/z of 300 -2000 was performed, followed by five tandem MS scans of the most abundant ions present in the full mass scan. All tandem MS scans were performed with an isolation window of 3 Da (precursor m/z Ϯ 1.5 Da) and with 35% collision energy.
Protein Farm software was used as an interface between the raw mass spectrometric information and a database searching program, SEQUEST (35). After deletion of poor quality spectra and conversion to file format, SEQUEST was used to search against a protein data base. An eluant was only identified as containing a protein if it contained three or more peptides, with a cross correlation score above 2; each of those spectra were manually confirmed.
Chromatography-Cytoplasmic extract was filtered and poured over a gravity anion-exchange resin column (Q-Sepharose, Amersham Biosciences) and then eluted with 0.3, 0.6, and 0.9 M NaCl. 50 l was then run on SDS-PAGE and subjected to immunoblotting. The FANCAcontaining fractions were then placed over a MonoQ anion-exchange column (Amersham Biosciences) and eluted using a 100 -500 mM salt gradient. After analysis by immunoblotting, the FANCA peak fractions were pooled and concentrated down to 600 l, filtered, and subjected to gel filtration (Superose 6, Amersham Biosciences). Molecular mass markers (Sigma) were run as standards to determine the molecular masses of complexes within each fraction. FANCA fractions were then immunoprecipitated with FLAG-affinity resin (Sigma). 10% of the beads were probed by immunoblotting with FANCC, FANCG, FANCF (gift of M. Hoatlin, Portland, OR), and FANCE (gift of K. J. Patel, Cambridge, UK) antibodies. The remaining beads were run on gel and silver-stained (Bio-Rad).
Nuclear extract was first extracted in 500 mM NaCl, filtered, and passed over a P11 cation-exchange column (Whatman). The FANCA fraction was then run directly on gel filtration. FANCA fractions were then chromatographed using a heparin-Sepharose CL-6B column (Amersham Biosciences). Finally, the FANCA fractions were immunoprecipitated by FLAG-affinity gel. The beads were analyzed as above.
Chromatin extract was chromatographed as follows: (i) passage by gravity over Q-Sepharose anion-exchange resin; (ii) placement on a MonoQ anion-exchange column followed by elution with a 100 -500 mM gradient; (iii) Superose 6 gel filtration; or (iv) FLAG-affinity immunoprecipitation. All fractionation and chromatographic steps were followed by immunoblotting detection of FANCA.
Phosphatase Reaction-Phosphatase reactions were taken from previous protocols (45). After anti-FLAG immunoprecipitation, the resulting washed beads were incubated with phosphatase (Roche Applied Science) or PP2A (gift of David Brautigan, University of Virginia) for 30 min at 30°C in the supplied reaction buffer ( phosphatase, in a 100-l mix containing 100 units of enzyme in the presence or absence of the phosphatase inhibitors, 1 mM sodium pyrophosphate and 1 mM sodium orthovanadate). All reactions were stopped by the addition of 40 l of 5ϫ sample buffer (62 mM 2% SDS, 2% ␤-mercaptomethanol).

RESULTS
Affinity Purification of the FA Complex-To demonstrate the validity of purification of epitope-tagged FA complex, we performed batch FLAG-affinity precipitation of FLAG-FANCA from GM6914 ϩ FLAG-FANCA and GM6914 ϩ vector control cells in whole-cell extract (500 mM NaCl). Immunoprecipitation was performed with covalently bound anti-FLAG-affinity resin. After extensive washing, the precipitates were eluted with FLAG peptide, and the resulting eluates were concentrated and buffer-replaced with 100 mM NH 4 HCO 3 . Thereafter, these eluates were analyzed by tandem mass time-of-flight ion-spray MS. Peptide sequence was detected by correlating to the FA proteins FANCA, FANCC, and FANCG (Fig. 1A). Additionally, cdc2 and GRP94, which are both known and previously reported binding proteins of FANCA ( Fig. 1B; Refs. 40, 47), were detected.
Whole-cell Extract Contains Two Different FA Complexes-We and others have demonstrated that the FA complexes exist in both the cytoplasm and nucleus. However, it is unclear what function the FA complexes play and what the complex composition is in each compartment. Thus, isolation and purification of the FA complexes would serve to define them to determine potential functions. To ascertain characteristics of FA protein purification, we prepared whole-cell lysate in 150 mM NaCl from GM6914 ϩ FANCA cells and applied it to a gravity anionexchange column. Immunoblotting for FANCA revealed that most protein was bound and eluted in the 0.3 and 0.6 M fractions (data not shown). These fractions were pooled, rediluted to 150 mM NaCl, and run over an anion-exchange (MonoQ, Amersham Biosciences) column by HPLC. Proteins were eluted from the column using a salt gradient. Subsequent immuno-blotting of fractions revealed two distinct peaks containing FANCA: one eluted with low salt and the other eluted with high salt (Fig. 2A).
Fractions from these two separate peaks were pooled separately and passed over a gel-filtration column (Superose 6, Amersham Biosciences). Immunoblotting for FANCA revealed that the low salt fractions corresponded to a FANCA-containing complex of 600 kDa, whereas the high salt fractions corresponded to a FANCA-containing complex of 2 MDa (Fig. 2B), as compared with a set of molecular mass standards.
Stepwise Purification of FANCA-containing Complexes-Published data from this lab and others (30,48,51) have revealed that FA proteins are contained in both nucleus and cytoplasm. Reasoning that these cellular compartments might contain differently sized and differently composed FA complexes, we performed a stepwise scheme of purification, as detailed in the scheme in Fig. 3. First, we fractionated FA-A mutant GM6914 ϩ FLAG-FANCA or GM6914 ϩ vector cells (negative control), either untreated or treated with 0.1 M MMC into cytoplasmic and nuclear extracts. In all subsequent figures, the vector control was conducted in parallel but is not shown except for the final affinity step. Because of the significant precipitation of proteins in nuclear extracts made with high salt and subsequently lowered to low salt (data not shown), we decided to make nuclear extracts at 500 mM NaCl while maintaining the cytoplasmic extract at 50 mM salt, as called for during the simple fractionation procedure. Second, we passed the extract (20 mg) from the cytoplasmic fraction over a gravity anion-exchange column and eluted with a step gradient of NaCl. FANCA was detected predominantly in the 0.3 and 0.6 M NaCl step elutions (Fig. 4A). These fractions were diluted back to 100 mM, passed over an anion-exchange column, and eluted using a 100 -500 mM NaCl gradient. Immunoblotting showed FANCA in lower salt fractions, which is consistent with the data in Fig. 2 (data not shown). These FANCA fractions were then concentrated and loaded on a gel filtration column. Immunoblotting revealed FANCA in fractions corresponding to a predicted size of ϳ600 kDa (Fig. 4B), which is After immunoprecipitation with anti-FLAG-affinity gel, the FA complexes were eluted with FLAG peptide, and the resulting solution was digested with trypsin and analyzed by MS. Peptides from FANCA, FANCG, and FANCC were detected. B, peptides from previously described binding proteins cdc2, GRP94, and HSP70 were also detected. consistent with the finding of a complex at that size in the whole-cell extract in Fig. 2. Overall, we noted that less FANCA was present in the cytoplasmic extract of MMC-treated cells. Total protein was 200 g, a 100-fold decrease of the starting material.
Similarly, we assayed the nuclear extract, starting with 20 mg of protein. First, the nuclei made in the fractionation were extracted in 500 mM NaCl and passed through a P11 column, where FANCA was predominantly detected in the flow-through (Fig. 5A). The flow-through was loaded directly on a gel filtration column after concentration. Immunoblotting revealed FANCA in a 2-MDa complex (Fig. 5B), indicating that the second peak in the analysis of whole-cell extracts in Fig. 2, B and C likely corresponds to a nuclear form of the FA complex. Previous work in our lab with different tagged FANCA and FANCG constructs co-expressed in FANCA-and FANCG-null mutant cells, respectively, revealed that homodimerization of FANCA or FANCG is unlikely (data not shown). These data indicate that the FA nuclear complex consists of additional proteins added after presumed transport from the cytoplasm to the nucleus. To further purify the complex, we pooled the FANCA fractions and passed them over a heparin column. The FANCA fractions in extract from MMC-treated cells were detected in flow-through and early in the elution (data not shown) but did not bind at all in the untreated group. Total protein at this step was ϳ200 g as well, denoting a 100-fold purification. A, wholecell extract was prepared from HeLa cells in 500 mM NaCl and successively passed over gravity anion-exchange and gradient-anion exchange columns. B, two complexes were noted; each was then subjected to gel filtration. Each step was followed by FANCA immunoblotting. Immunoblotting revealed the presence of 500-to 600-kDa and 2-MDa complexes containing FANCA. wce, whole-cell extract Interestingly, none of the FANCA from non-drug-treated cells appeared to bind to the heparin column at all. Overall, we detected increased FANCA-containing complex in nuclear extract from MMC-treated cells.
Chromatin Extract Contains a 1-MDa Form of the FA Complex-We have previously reported that a subset of the FA proteins resides in a protein fraction liberated by DNase treatment and is thus termed chromatin extract (48). We also have reported that increased FA proteins localize to chromatin after MMC treatment (48). To characterize the FA complex residing on chromatin, we first extracted chromatin by DNase treatment of FA-A mutant GM6914 ϩ FLAG-FANCA or ϩ vector cells with or without MMC treatment after permeabilization and removal of cytoplasmic and soluble nuclear proteins. We used approximately the same number of cells as in the production of the nuclear and cytoplasmic extracts above. The result-ing extract consisted of 5 mg of protein in 200 mM salt. As in the previous sections, we passed the chromatin extract successively over a gravity anion-exchange column (Fig. 6A), an anion-exchange chromatography column (data not shown), and gel filtration (Fig. 6B), detecting the FA complex by FANCA immunoblotting. The chromatin FA complex was detected as a 1-MDa complex, a form smaller than that seen in nuclear extract but larger than the cytoplasmic complex. A much more pronounced amount of FA proteins was seen throughout chromatin fraction chromatography on extract from cells treated with 0.1 M MMC, which is consistent with our published data demonstrating greater chromatin localization upon DNA damage (48). Interestingly, a higher mobility form of FANCA was noted only in the fractions from chromatin made from MMCtreated cells. The higher molecular mass forms of the complex in fractions 12-14 do not contain this isoform, suggesting they may be part of the higher molecular mass nuclear complex.
Cytoplasmic Extract from Mitotic Cells Contains both Small and Large FA Complex-We have previously shown that the FA complex cannot be detected in a nuclear extract at mitosis and that the FANCG protein, although still part of the complex, becomes phosphorylated (48). To define the complex in which phosphorylated FANCG resides, we purified the FA complex as above from cytoplasmic extract made from HeLa ϩ FLAG-FANCG or ϩ vector cells arrested at mitosis after a 16-h incubation in 1 M nocodazole versus that made from an asynchronous cell population. All purification steps were followed by FANCA immunoblotting. The mitotic extract displayed a similar behavior to asynchronous extract over gravity (Fig. 7A) and column anion-exchange chromatography (data not shown). These FANCA fractions were then passed over a gel filtration column. The cytoplasmic FA core complex in HeLa cells was the same as in the FA-A-corrected cells in Fig. 4 (GM6914 ϩ FLAG-FANCA). Compared with cytoplasmic extract from asynchronous cells displaying the 600-kDa form of the FA complex, we observed a form existing in higher molecular mass fractions in the mitotic extract corresponding to ϳ750 kDa (Fig. 7B). The higher mobility forms of FANCG seen in mitosis were also apparent (see Fig. 8D), confirming that the complex is intact at the time of egress from chromatin and suggesting that the Affinity Purification and Silver-staining of FA Complexes-Based on our experience with this scheme of chromatography, three column purification steps still result in complex mixtures. To further purify the different preparations, we incubated pooled fractions from chromatographic preparations in the earlier sections containing the FA proteins with anti-FLAG-affinity gel. The result of this immunoprecipitation was run by SDS-PAGE for immunoblotting. In Fig. 8, A-D are shown cytoplasmic, nuclear, chromatin, and mitotic preparations, respectively. All four compartments contained FA core complex consisting of at least FANCC, FANCF, and FANCG coprecipitating with FLAG-FANCA expressed from pMMP-FLAG-FANCA. In the cytoplasmic preparation, a clear decrease in FA core complex was seen after MMC treatment (Fig.  8A). However, no FANCE coprecipitation was detected in the cytoplasm. In contrast, a clear increase in magnitude was evident in the nuclear fraction of MMC-treated cells (Fig. 8B), and FANCE coprecipitated with FLAG-FANCA. Consistent with these data, the chromatin pulldown also displayed increased FA core complex (Fig. 8C), which is in agreement with our published data (48). FANCC and FANCE coprecipitation with FLAG-FANCA in chromatin is shown in Fig. 9. A comparison of mitotic versus asynchronous extracts shows that the complex was intact at the time of egress from the nucleus at mitosis (Fig. 8D). Again, FANCE is not detectable in either the mitotic or asynchronous FA cytoplasmic core complex. FANCE is presented separately for clarity, as it tends to be difficult to resolve from FANCG on SDS-PAGE. Overall, these data are consistent with the idea that an inducible shift occurs from cytoplasm to nucleus after DNA damage and back again after mitosis. Although we have not shown if the increase in nuclear/chromatin FA core complex levels represents active nuclear import or increased nuclear/diminished cytoplasmic FA protein stability, the net result is increased nuclear/chromatin FA proteins.
To see whether additional proteins become part of the FA core complex after MMC treatment, we ran chromatin-immunoprecipitated proteins from the affinity step above on SDS-PAGE and silver-stained the resulting gel. All proteins in the complex were markedly higher in magnitude after MMC treatment (marked by an *), and several additional binding proteins can be seen (Fig. 8E, ➜). FA-A mutant GM6914 cells both untreated and treated with MMC were included as negative controls and run in parallel. Any band from both the negative controls and in the FLAG-FANCA-containing cells was consid-ered nonspecific. Immunoblotting on a parallel gel revealed the relative positions of the coprecipitating FA proteins, as indicated.
Protein extracts were made from ϳ20 plates of cells. Nuclear and cytoplasmic extracts each totaled 20 mg, whereas chromatin extracts totaled ϳ7.5 mg. Purification by chromatography resulted in a 100-fold diminution in protein amount for all three types of extracts. Affinity-purification resulted in another 2-fold purification, resulting in an ϳ200-fold overall.
FANCA Is a Chromatin-bound Phosphoprotein in Response to MMC Treatment-FANCA immunoblotting of chromatin extract from MMC-treated cells detected a higher mobility FANCA form (Fig. 6B). This result was confirmed as an isoform of FANCA, after immunoprecipitation with anti-FLAG-affinity gel and FANCA immunoblotting (Fig. 8C). The two FANCA isoforms are not due to dimerization of FANCA, because the mutant FA-A cell line GM6914 used in this experiment contained no endogenous FANCA; thus, the only FANCA species was FLAG-FANCA. Past reports have shown that FANCA is phosphorylated and that mutant cells display a lack of FANCA phosphorylation (32,49). To demonstrate that FANCA is a phosphoprotein, we performed phosphatase reactions with immunoprecipitated FANCA from chromatin extract made from MMC-treated cells (as in the previous section). Both and PP2A phosphatase reactions caused the high mobility form of FANCA, which was inhibited by phosphatase inhibitors (Fig.  9), to disappear. Incubation in reaction buffer at 30°C had no effect upon the two FANCA isoforms. The phosphatase reaction did not affect the coprecipitation of either FANCC or FANCE. DISCUSSION In this paper we describe the purification of the FA core complex in its four forms: asynchronous cytoplasmic, mitotic, chromatin, and nuclear. In addition, we show evidence to confirm our earlier findings that increased FA core complex is present in chromatin after MMC treatment and that it displays egress from the nucleus at mitosis. These data suggest that the complex is translocated into the nucleus where it becomes much larger and presumably becomes activated for a particular biochemical function. MMC treatment results in a number of changes in the FA complex, including a higher mobility FANCA form on chromatin, an increase in the amount of nuclear complex, and a decrease in the amount of cytoplasmic complex. These data are summarized in Table I. FANCA has been reported to be phosphorylated in a functionally important manner (49). However, an additional iso- form has not been noted, and a cytoplasmic kinase has been implicated (42,43). We did not see the higher mobility isoform of FANCA unless we performed chromatography. Even then, the phosphorylated isoform was noted only in chromatin preparations after DNA damage. This finding does not rule out that FANCA is also phosphorylated in the cytoplasm, producing an isoform with identical mobility to unphosphorylated FANCA. Importantly, we have described a rational approach for purification of the native complex and its binding partners that will allow identification of proteins and functional motifs and which can subsequently be tested biochemically using these extracts. This approach has already shown promise, as we have confirmed the ability to detect the pulldown protein and the known FA-binding proteins (FANCA, FANCC, FANCG, cdc2, and GRP94) by tandem mass time-of-flight ion-spray MS (27,39,40). We have also found by MS at least two novel binding proteins with obvious functional motifs, 2 the binding of which we are currently confirming by immunoprecipitation, immuno-blotting experiments, and biochemical testing. In addition, we show that cytoplasmic extract displays less FA protein after MMC treatment, an observation we believe is consistent with our data showing localization of FA core complex in the cytoplasm only in early G 1 phase, followed by exclusive nuclear localization from the border between G 1 -S until mitosis. 3 This would explain the increase in chromatin binding seen after MMC treatment, whereby more cells are arrested during S or G 2 /M phase of the cell cycle. In fact, because MMC treatment will induce an increase in S and G 2 percentages even in wildtype cells, the possibility exists that the primary effect on the FA core complex is upon the cell cycle. We are currently investigating this possibility.
To date, eight FA genes have been cloned, but their gene products have yielded relatively little information. Advances have been made in the recent elucidation that FANCD2 binds to BRCA1 and that BRCA2 equals FANCD1 (17,41,50). Also, the FANCD2 protein has been shown to be ubiquitinated in a manner that is in part dependent upon the wild-type presence of the FA core complex. It is the FA core complex about which little is known. In fact, none of the members of the core complex (FANCA, FANCC, FANCF, FANCG, FANCE, FANCL) has lower eukaryotic homologs, and few functional motifs exist, although FANCL has apparent ubiquitin ligase activity (18). Therefore, it is critical that all complex components be identified, as these may have functional motifs whose biochemical activities can be tested in the setting of the FA core complex. Thus, methods to purify the FA complex are crucial for identifying these proteins as well as providing material for testing function. A recent report by Wang and co-workers (44) details binding of the BLM helicase to the core complex in a 2-MDa complex, as well as the existence of two smaller FANCA-containing complexes. This work suggests the presence of both large and small complexes in a 200 mM NaCl-extracted nuclear extract. Our data both confirms the report of Wang and coworkers and separates the complexes: the small form is cytoplasmic, the large form is nuclear, and the intermediate form binds to chromatin. On the other hand, using our conditions, we were unable to detect BLM in our preparations.
The cytoplasmic form of the FA core complex is likely to be critical, as past work shows that forced, exclusive expression of FANCC in the nucleus of mutant FA-C cells fails to correct the cells to wild-type MMC sensitivity (51). In addition, the 1-MDa chromatin complex, which is localized to chromatin in a DNAdamage inducible fashion, is only found when extracted using DNase (48). Finally, a fourth unique complex is found at mitotic egress from condensed chromosomes. These findings are suggestive of a very complicated FA pathway, with roles and signaling in cytoplasm, soluble nuclear, and chromatin compartments. The FA core complex seems to be tied closely to cell-cycle regulation.
As presented in this paper, we have performed simple affinity purification and have indeed identified several known specific binding proteins by MS. It is important to establish con-2 G. M. Kupfer, unpublished data. 3 J. Mi, submitted for publication.
FIG. 9. FANCA is a chromatin-bound phosphoprotein in response to MMC treatment. Beads from FLAG-affinity pulldown performed on purified chromatin were incubated in or PP2A phosphatase reactions with or without phosphatase inhibitors (1 mM sodium pyrophosphate, 1 mM sodium orthovanadate). Phosphatase removed the higher FANCA isoform and diminished the overall FANCA protein level. No higher isoform was detectable on crude chromatin extract that was not purified by chromatography. FANCC and FANCE coprecipitation was unaffected by phosphatase reaction. PC, positive control (FA-A mutant GM6914 cells ϩ FLAG-FANCA); NC, negative control (FA-A mutant GM6914 cells ϩ vector); WCE, whole-cell extract. ditions that will enable us to preserve the complex in its native state and purify it cleanly. The silver-stained gel in Fig. 8 suggests that this is indeed the case. We are currently in the process of further analyzing these preparations so that we can perform a complete analysis of the complexes. We also are analyzing further the FANCA isoforms to map phosphorylation sites.