Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Polycomb and polyhomeotic are constituents of a multimeric protein complex in chromatin of Drosophila melanogaster.

Polycomb and polyhomeotic are constituents of a multimeric protein complex in chromatin of... The EMBO Journal vol.1 1 no.8 pp.2941 -2950, 1992 Polycomb and polyhomeotic are constituents of a multimeric protein complex in chromatin of Drosophila melanogaster responsible for maintaining the repressed state of homeotic Axel Franke1, Marco DeCamillis2, genes (reviewed in Paro, 1990). Mutations in members of Daniele Zink1, Niansheng Cheng2, the Pc-G show an ectopic expression of homeotic regulators W.Brock2 and Renato Paro1 3 Hugh (Struhl and Akam, 1985; Weeden et al., 1986; Dura and 'ZMBH, University of Heidelberg, Im Neuenheimer Feld 282, D-6900 Ingham, 1988; McKeon and Brock, 1991; Simon et al., Heidelberg, Germany and 2Department of Zoology, University of 1992), which can result in dramatic transformations of body British Columbia, 6270 University Boulevard, Vancouver, BC, structures (Struhl, 1981; Duncan, 1982; Duncan and Lewis, V6T 1Z4, Canada 1982; Ingham, 1984; Dura et al., 1985; Jurgens, 1985; 3Corresponding author Breen and Duncan, 1986; Adler et al., 1989; Jones and Communicated by H.Jackle 1990). Pc-G genes are not only involved in Gelbart, homeotic gene expression patterns, but have maintaining been found to perform similar roles on other developmental The polycomb group (Pc-G) genes are responsible for regulators (Ingham, 1984; Breen and Duncan, 1986; Dura maintaining the repressed state of homeotic genes during development. It has been suggested that the Pc-G exerts and Ingham, 1988; Smouse et al., 1988; Busturia and its transcriptional control by regulating higher order Morata, 1988; Wu et al., 1989). chromatin structure. In particular, the finding of genetic Two genes of the Pc-G group have been extensively and molecular similarities to components involved in characterized at the genetic and molecular level. mutant embryos of the Polycomb (Pc) gene heterochromatin formation, led to the proposal that Homozygous genes are permanently repressed by mechan- display posterior transformations of all segments, charac- homeotic teristic for a deregulation of all the homeotic genes of the similar to those responsible for heterochromatin isms ANT-C and BX-C (Lewis, 1978; Denell and Frederick, compaction. Because of synergistic effects, Pc-G gene Sato and Denell, 1985; Tiong and Russell, 1990). Pc products are thought to act in a multimeric complex. 1983; found to exert its regulatory role at the transcriptional Using immunoprecipitation we show that two members was level (Weeden et al., 1986; Zink et al., 1991). The Pc of the Pc-G, Polycomb and polyhomeotic, are constituents is a nuclear protein associated with - 100 different of a soluble multimeric protein complex. Size fractiona- protein sites on polytene chromosomes of larval salivary glands tion indicates that a large portion of the two proteins are 1989; Zink, 1990). Reporter genes linked (Zink and Paro, found in a distinct complex of molecular weight 2-5 x homeotic cis-regulatory sequences become ectopically to 106 Da. During embryogenesis the two proteins show the when tested in a background lacking Pc+ distribution. In addition, by double- expressed same spatial and Bienz, 1991; Zink et al., 1991). labelling we can demonstrate that function (Muller immunofluorescence et al. showed that the Pc protein Zink (1991) and polyhomeotic have the same binding Additionally, Polycomb exactly sequences of the Antenna- binds to regulatory on polytene chromosomes of larval salivary specifically patterns gene. We that some Pc-G proteins act in pedia glands. propose of the other well characterized gene, analysis Genetic multimeric complexes to compact the chromatin of stably revealed a high degree of pleiotropy of (ph), like the homeotic regulators. polyhomeotic repressed genes Although the homeotic transforma- mutant phenotype. words: chromatin regulation/Drosophila/homeotic gene the Key classify it as a Pc-G gene, additional defects tions observed group/polytene chromosomes regulation/polycomb death in the ventral epidermis (Dura et al., 1987) like cell central nervous system axons (Smouse of and misrouting and 1990) can be identified. Smouse Perrimon, et al., 1988; Introduction structure and is arranged as a a complex gene ph shows unit. lesions in both units result Only mechanisms that maintain the determined state tandemly duplicated Regulatory alleles, whereas single lesion or hypomorphic are necessary for the proper development of an in null strong of a cell weak phenotypes. The two only hypomorphic alleles display body plan. In Drosophila melanogaster, the organized within a stretch of 25 kb of genomic units contained of the homeotic genes of the genetic activity differential that have certain characteristics similar proteins DNA encode and bithorax-complex (BX-C) Antennapedia- (ANT-C) with DNA (Deatrick et al., 1991; interacting of proteins the various body structures along the anterior- determine al., In addition to four blocks of et 1992). DeCamillis axis 1978; Kaufman et al., 1980). Since posterior (Lewis, the and serine/threonine-rich sequences, glutamine repeats is throughout development, they have their function required a with partial homology to the displays region a model to identify and characterize ph protein been used as system and a C4 zinc finger. Like motif single for the transmission of the determined helix-loop-helix the elements required - 100 chromosome sites and ph to binds polytene which define. Genetic analysis has revealed a Pc, ph states they sequences from the recognizes regulatory specifically distinct class of genes involved in part of this process. The et al., 1992). region (DeCamillis bithoraxoid of the Polycomb group (Pc-G) were found to be genes Press Oxford University A.Franke et al. The we found that our Pc-G genes display some common features described rabbit anti- with previously polyclonal another Pc antibodies distinct class and of Drosophila the modifiers genes, (Zink Paro, 1989) could be used of to position effect variegation in vitro (PEV). These are translated Pc genes as well thought immunoprecipitate as protein to encode structural Pc and constituents from regulatory of nuclear extracts, chromatin endogenous protein embryonic or (reviewed in from nuclear Eissenberg, 1989; extracts of Henikoff, 1990). Some Schneider L-2 tissue Pc- culture cells. To G genes show effects on the PEV (D.Sinclair, perform we N.Clegg, immunoprecipitations, essentially T.Grigliatti followed and the H.W.Brock, submitted) and of et al. who reciprocally, protocol Gay (1988), were able some modifiers to of PEV homeotic the association display of the with transformations identify other engrailed protein (Reuter et nuclear al., 1990; factors. In this G.Reuter, personal method the nuclei are communication). a lysed by Paro and Hogness (1991) have found a molecular shock and the released soluble nuclear relationship hypertonic proteins between the Pc protein and the and heterochromatin-associated concentrated a 30% partially purified ammonium by HP1 protein, which encoded sulfate is The by Su(var)205 redissolved (Eissenberg soluble nuclear precipitation. proteins et al., 1990). were Both proteins share the chromo with anti-Pc domain, antibodies and a immunoprecipitated Protein 48 amino acid motif found to be A for beads responsible Materials specific (see and -Sepharose methods). chromatin To binding (Messmer et Based on if Pc is al., 1992). associated with determine other nuclear these proteins similarities, it has been in the proposed that the Pc-G soluble we genes the repress fraction, separated immunoprecipitate homeotic genes by regionally the SDS -PAGE. compacting these chromatin initial by However, did not experiments (heterochromatinization) and thus eliminating the to a yield visible enough coprecipitated proteins give pattern accessibility ofDNA to in silver diffusible stained transcription factors. To circumvent This the gels. sensitivity problem mechanism could the we imprint order tried to higher structure in of vivo labelled nuclear immunoprecipitate chromatin with the determined state of the cell as from tissue culture cells. defined Pc is proteins endogenously by the differential expression pattern of in these developmental cells et expressed (Messmer Cells were al., 1992). regulators like homeotic genes (Locke et for 3 h al., 1988; Gaunt with labelled and [35S]methionine and [35S]cysteine, and Singh, 1990; Paro, 1990; Reuter et nuclear al., were 1990). with anti-Pc proteins immunoprecipitated Because modifiers of the PEV show a antibodies and very on high gene separated gels. Disappointingly, only very dose sensitivity, Locke et al. weak (1988) proposed could be that observed after signals prolonged exposures heterochromatin is packaged by reiterated not multimeric protein Because the (data shown). Pc immunoprecipitated protein complexes composed of the products of seemed to these have genes. a small amount of incorporated only Heterochromatin apparently can spread we along attributed the the radioactivity, result to the negative chromosome in a cooperative fashion and thus transcrip- of Pc and the other relatively high stability associated tionally inactivate large chromosomal Pc-G regions. Jiirgens have a low proteins. such proteins that might turnover, (1985) has found that the double and a 3 h triple mutant is too short to labelling period incorporate sufficient combination of four of genes the Pc-G, Additional sex we combs have evidence radioactivity. from the Indeed, develop- (Asx), Polycomb-like (Pet), Posterior sex mental of combs (Psc) Pc that the and is stable profile protein very Sex during combs on midleg (Scm) showed a marked enhancement and embryogenesis (B.Zink R.Paro, unpublished results). of the homeotic transformation compared with the single the However, by radioactively after labelling proteins mutations. This synergistic effect of the Pc-G could be we were able to immunoprecipitation, visualize the Pc- explained by a participation of the associated different For Pc-G gene this we used proteins. a purpose modified products in a common regulatory structure, i.e. a Bolton-Hunter multimeric to 35S the reagent couple to precipitated protein complex. Pc-G proteins could form multimeric et proteins see also (Assoian al., Materials 1980; and protein units that are functionally and The mechanistically of this methods). method disadvantage is that the equivalent to heterchromatic complexes, except that antibodies in the are also labelled precipitate and result in particular Pc-G proteins would target a the complexes on the to that can strong signal autoradiogram obscure the euchromatic genes. of other in the same signals molecular proteins weight range. Here we present evidence that two members of shows the Pc-G the result of 1 this Figure In experiment. lane 3, are part of a large multimeric protein 10-15 complex. new DeCamillis Pc-associated are seen to proteins coprecipitate et al. (1992) have shown that the ph and Pc with proteins have with Pc, the compared control lanes 5 and 4, 6 (see many overlapping binding sites on polytene to chromosomes. In lane no legend Figure 1). band 3, major corresponding We show that this in vivo correlation is based on to the a molecular shown in Lane as an Pc-protein, 1 in vitro association produced of the two proteins in a soluble nuclear complex. 35S-labelled is visible. protein, this However, is not Using immunoprecipitations we can demonstrate that Pc and that the surprising, labelling of proteins with considering the ph are found in a multimeric complex. We have furthermore can 35S-labelling reagent SRL, substantially change the refined the correlation of the Pc and ph binding patterns on of in SDS mobility -PAGE proteins et (Assoian al., polytene 1980). chromosomes by using The results double-immunofluorescence shown in Figure clearly indicate that Pc is techniques. We propose that the Pc-G proteins exert associated with their several other in proteins the soluble nuclear repressory function on the chromatin of homeotic genes as fraction. The multimeric protein is complex stable enough multimeric protein units. to be immunoprecipitated with anti-Pc antibodies, providing molecular means of possible characterizing the various Results components. Immunoprecipitation of Pc and associated proteins and Polycomb polyhomeotic are colocalized in the from nuclear extracts same multimeric protein complex We used immunoprecipitations to enrich for proteins that Biochemical identification of the coprecipitated proteins may associate with the Pc protein. In an initial experiment would be very difficult, because only small amounts of 2942 Pc-G genes form a multimeric complex material could be recovered. However, we were able to take advantage of the considerable amount of genetic and molecular data available on the Pc-G system. Obvious candidates for proteins interacting at the molecular level with Pc are other members of the Pc-G. Indeed, we have shown previously that Pc and ph have similar binding patterns on _- Pc - Fc ., LA. ,, - _ow"' Pr, - Sw*fi.D IL - Fig. 1. Immunoprecipitation of Pc-associated proteins from embryonic nuclear extracts. 25 11 (protein concentration 15 mg/mil) of an embryonic nuclear extract was immunoprecipitated in a total volume of 500 using anti-Pc antibodies. The precipitate was labelled with Al 35SLR (see Materials and methods) and fractionated on a 8% SDS-PAGE gel. The gel was dried on a piece of Whatman paper and exposed for 24 h. Lane 1, in vitro translated 35S-labelled Pc protein as a size reference; lane 2, 14C protein molecular mass standards with the on corresponding sizes given the left, lane 3, immunoprecipitations .. F- S.h:... with anti-Pc antibodies and protein A-Sepharose; lane 4, the same 4bl antibodies; immunoprecipitation as in lane 3 but omitting the anti-Pc lane 5, mock-immunoprecipitation without nuclear extracts. Note the strong signal from the immunoglobulin chains; lane 6, mock- Westerns were incubated with alkaline phosphatase coupled anti-rabbit immunoprecipitation without nuclear extract and without antibodies. the new bands in lane 1 are antibody as secondary antibodies. Thus, not due to with the The bands cross-reactivity secondary antibody. denoted with Fc in lanes 1 and 3 are caused the of by cross-reactivity Fig. 2. The Pc and ph protein are coimmunoprecipitated in the same the with the anti-Pc antibodies. secondary antibodies (C) multimeric protein complex. Panels A and B show an of Pc with nuclear Coprecipitation anti-ph antibody. Embryonic immunoprecipitation of 100 embryonic nuclear extract (protein extracts were with The immunoprecipitated anti-ph antibody. Al concentration 15 mg/ml) with anti-Pc antibodies in a total volume of precipitates were separated by SDS-PAGE, transferred to examined Western for the of 1.5 ml. One-third of the precipitate in A and two-thirds of the nitrocellulose and using blotting presence Western blots with antibodies to precipitate in B were separated on a 8% (A) and 6% (B) SDS-PAGE ph and Pc. Lanes 1-5 show probed with anti-Pc lanes 6-7 show western blots with antibodies to Pc. and transferred to nitrocellulose. Blot A was ph and probed probed Pc Blot Lane Western blot of an nuclear extract before antibodies and shows the of the 1, embryo specific precipitation protein. of lane of the to show B was probed with anti-ph antibodies to show the co-purification immunoprecipitation mobility ph products; 2, with Some Lane of nuclear extract with anti- of the same extract ph protein. 1, immunoprecipitation immunoprecipitation ph antibody. Lane Pc antibodies and Protein lane mock- of occurs. A-Sepharose; 2, degradation ph 3, mock-immunoprecipitation lacking without lane mock- Protein lane immunoprecipitation primary antibodies; 3, A-Sepharose; 4, mock-immunoprecipitation lacking nuclear without nuclear lane mock- lane extract; immunoprecipitation extract; 4, antibody; 5, mock-immunoprecipitation lacking shown in Pc reacted with the same without nuclear extract and lane lane 6, antibody immunoprecipitate immunoprecipitation primary antibody; with lane of the Pc and lane 2 to show that Pc 7, 5 nuclear extract to show the coprecipitates anti-ph antibody; 5, mobility ph proteins, I1 extract before forms. At least blot of an nuclear The encodes Western embryo immunoprecipitation respectively. ph gene multiple protein molecular bands above the in of Pc. The four different can be detected to show mobility higher weight ph proteins unstaged embryos with the used Due to the Pc are due to and in protein cross-reactivity secondary antibody (M.DeCamillis W.H.Brock, preparation). instability seems to have occurred. Both in this of the ph protein, some degradation particular experiment. 2943 A-Franke et at. polytene chromosomes of of the salivary glands (DeCamillis et presence some al., ph protein. Although degradation seems 1992). We to therefore tested have ifph copurifies in the most of the immuno- occurred, are ph products found in the precipitation with anti-Pc antibodies. In results lane The are the same 6, shown precipitate. is immunoprecipitate probed in Figure 2A and for 2B. the of of the Immunoprecipitations Pc The embryonic presence protein. comparison with the nuclear extracts were nuclear performed, on SDS extract in lane separated 7 shows that -PAGE Pc can also be and blotted onto nitrocellulose filters. In Figure 2A, antibodies. the precipitated by using anti-ph We find that in Western blot was probed with anti-Pc antibodies. Lane with 1, general immunoprecipitation antibodies is anti-ph less which contains the normal efficient than immunoprecipitate with the clearly anti-Pc shows antibodies. This could reflect the enrichment a of the Pc reduced when protein, with of the compared the accessibility in the ph protein or complex, mock-immunoprecipitations in the control lanes 3 difference in the and 2, 4 protein of the composition for complex (see legend to Figure the two 2A). The same set of the immuno- proteins. Alternatively, could be due discrepancy precipitations were tested for the to a presence of the different ph of the two protein specificity antibodies used, although in Figure 2B. Lane 1 identifies as one this seems ph of the as the two proteins quite unlikely antibodies gave that can be copurified with Pc, that Pc and results on indicating blots. ph comparable Western are part of the same multimeric protein The complex. of the two coprecipitation could also be proteins To prove the specificity of the interaction and a to explained of the by anti-Pc and cross-reactivity anti-ph demonstrate the of reproducibility the antibodies. in immunoprecipitation However, blots with Western probed anti-Pc protocol, we performed the reciprocal we experiment. never antibodies, detected a Using band at the cross-reacting specific anti-ph antibodies (DeCamillis et size of al., the 1992), we This is expected also the ph protein. case for the were able to coimmunoprecipitate the Pc protein with from reciprocal antibodies experiment also anti-ph (see embryonic nuclear extracts. Figure 2C, lane In order 2, shows to the 2). exclude Figure this completely result of possibility, the immunoprecipitation we have analysed used a first for more the In vitro specific translated approach. Pc and were 35S-labelled ph proteins immunoprecipitated with and anti-Pc anti-ph antibodies, respectively 3). (Figure In both cases we found no evidence for a cross-reactivity of the two This antibodies. that the proves of copurification the Pc and the in the ph proteins are due immunoprecipitates to the of these two presence in the same proteins multimeric complex. We tested the anti-Pc for the immunoprecipitate presence of additional that could a in proteins the play part regulatory and for which we had system, antibodies. Histone specific has an role in HI the of important inactive compaction chromatin in (reviewed Grunstein, antibodies 1990). Using Histone against Drosophila HI (kindly provided by S.Elgin, we have found that St. Louis) does not HI with coprecipitate Pc and thus is not a constituent of the major multimeric not protein A similar complex (data result shown). negative Fig. 3. Control of cross-reactivity of the anti-ph antibody with Pc and of the anti-Pc antibody with ph. (A) Immunoprecipitation of 35S- labelled Pc protein produced in a reticulocyte lysate. 5 jig of the translation mixture was incubated in a total volume of 500 ,ul. The precipitate was fractionated on an 8% SDS-PAGE gel and exposed overnight. Lane 1, immunoprecipitation with anti-ph antibodies and Protein A-Sepharose; lane 2, immunoprecipitation with anti-Pc antibodies and Protein A-Sepharose; lane 3, mock-immuno- precipitation without antibody; lane M, 14C protein marker as size standards that are given on the right. (B) Control for cross-reactivity 4. Fig. of of nuclear anti-Pc antibody Immunoprecipitations extracts with ph from protein. ph proteins that had been staged with anti-Pc antibodies. embryos The translated in a were reticulocyte lysate were precipitates separated immunoprecipitated with SDS-PAGE by and transferred to antibodies to ph and nitrocellulose. Pc. The The filter immunoprecipitates were upper was separated by incubated with antibodies and the SDS-PAGE and anti-ph lower filter exposed to with autoradiography. anti-Pc Lane 1, ph mRNA antibodies. At translated in the a every reticulocyte lysate stage analysed concentration containing was protein [35S]methionine; lane normalized to have a 2, constant amount of immunoprecipitation of ph translation products with Pc anti-ph immunoprecipitated antibody; On of the lane protein. lanes the 3, top individual immunoprecipitation of ph translation are products with embryonic anti-Pc stages denoted the time from antibodies. Note by the (h) Both absence of filters egg were Lane laying. incubated immunoprecipitation. with alkaline 4, immunoprecipitation using ph antibody of a phosphatase anti-rabbit lysate that did coupled antibodies as not contain antibodies. The band below secondary ph the Pc message; lane 5, strong is due to the mock-immunoprecipitation of signal cross- lysate containing ph with the chains translation reactivity of the products and protein heavy A, but antibodies. lacking ph primary Molecular antibody. standards are indicated weight on the left. 2944 Pc-G genes form a multimeric complex Size measurements of the multimeric protein complex was obtained when the immunoprecipitate was probed with The data we have presented suggest that Pc is associated antibodies against the homeotic proteins Antennapedia with 10-15 other nuclear proteins and show that the Pc-G (kindly provided by W.Gehring, Basel) and Ultrabithorax member ph is one of the constituents of this soluble (kindly provided by R.White, Cambridge). This indicates multimenc complex. Using gel filtration and density gradient that Pc-G proteins do not directly interact with homeotic centrifugation methods, we have determined the approximate proteins to exert their negative regulatory role. This molecular weight of the complex. As starting material we conclusion had also been suggested by genetic experiments used the same embryonic nuclear extracts and buffers as for (Castelli-Gair and Garcia-Bellido, 1990). We found no the immunoprecipitations, since under these conditions the evidence of a coprecipitation of the engrailed protein, complex was shown to be stable. The best separation of indicating that the engrailed-associated multimeric protein nuclear extract was achieved by gel filtration chromatography complexes identified by Gay et al. (1988) are distinct from Sephacryl S400 HR column (see Materials and on a the Pc-ph complexes identified here. methods). Different fractions were collected, separated by SDS-PAGE, blotted onto nitrocellulose and tested for the Developmental profile of the Pc - pH interaction presence of Pc and ph protein, respectively. Figure 5 shows The activity of the Pc-G is needed after germ band elonga- that significant amounts of Pc and ph proteins copurify as tion, to take over and maintain the homeotic expression a high molecular weight complex. Most of the Pc-ph pattern initiated by the early maternal and segmentation genes over a narrow molecular weight range. elutes protein (Struhl and Akam, 1985; Weeden et al., 1986; Kuziora and Extrapolating from the molecular weight markers indicated McGinnis, 1988; Celniker et al., 1990; McKeon and Brock, to Figure 5), we have estimated that the weight legend (see 1991; Simcox et al., 1991; Simon et al., 1992). We wanted of the Pc-ph complex is -2-5 x 103 kDa. A more to test whether this activity is regulated at the level of weight could not be obtained because of the absence accurate complex formation. In particular, we have tried to identify a suitable molecular weight marker. There is a clear upper of at what embryonic stages Pc is associated with ph. We corresponding to the size of the complex (Figure 5, band collected nuclear extracts from staged embryos, performed 8), but there is a smear of smaller complexes decreasing lane immunoprecipitations with anti-Pc antibodies and analysed molecular weight to the size of the monomers. Molecular in for the copurification of ph. We normalized the mass of of these smaller complexes were estimated using weights immunoprecipitated proteins to give a constant amount of S-300 HR column that separates molecules in a Sephacryl Pc protein in each lane (Figure 4). Our data suggest that of 10-1500 kDa (data not shown). It is not clear the range Pc and ph can be coprecipitated over the entire period of the smear is due to partial dissociation of the whether embryogenesis. Although the ratio between the coprecip- during preparation, or whether the two proteins complexes itated ph and Pc proteins clearly increases during with other proteins, resulting in multimeric are associated embryogenesis, there is no stage at which the two proteins of different composition and thus of different complexes are not associated. This shows that the Pc-ph complexes sizes. form early in development during the initiation stage of comparable results by separating the have obtained We homeotic gene regulation, suggesting that Pc-G repression on sucrose gradients (data not shown). nuclear complexes homeotic expression is most probably controlled by a of most of the Pc and ph is found in the After centrifugation, means than formation of the Pc-G complex. different Fig. 5. Cofractionation of Pc and ph in multimeric protein complexes separated by gel filtration. Nuclear protein extracts of wild type embryos (200 jd, 15 mg/ml) was applied to a column of Sephacryl S-400 HR (90 cm long, 2 cm wide). Fractions were collected after the expected elution of catalase (240 kDa, see Materials and methods). 12 x 10 ml fractions were collected until the exclusion volume was reached. Fractions were concentrated and applied to a 7.5% SDS-PAGE gel. The electrophoretically separated proteins were transferred to nitrocellulose. The upper half of the filter was probed with anti-ph antibodies and the lower half with anti-Pc antibodies. The Western blot was developed using alkaline phosphatase coupled anti-rabbit antibodies as secondary antibodies. Lane NE, 5 A1 (15 mg/ml) embryonic nuclear extract to indicate the expected size of Pc and ph; lanes I -11, different fractions analysed. The column was equilibrated separately with marker proteins indicated at the bottom. 240 kDa, catalase; 750 kDa, mouse proteasome. The exclusion volume was determined by specifications given by the supplier (Pharmacia). Dextran blue 2000 (2 x 106 Da) could not be used as a size standard since it seemed to interact with the matrix, resulting in a broad elution peak. 2945 A.Franke et al. in The 6. of the Pc and distribution of the two were the same at all As Fig. Distribution ph protein embryos. spatial proteins embryonic stages. an and after at band band retraction at 16 are Note the intense of the CNS example embryos germ elongation (top) germ stage depicted. labelling in both latter cases. of the exceeds pellet, suggesting that the from different and different and weight complex glands chromosomes, our the separation of this method. The elution experience has shown that a of so capacity profiles comparison many binding exclusion in the case of Pc and - can be of Pc and were the same or sites, ph 100, very difficult. ph using gel in These results that we have found that different can centrifugation sucrose gradients. suggest Furthermore, preparations of the Pc and show intensities due to the independent techniques used, ph participate different signal complexity of the in We estimate To avoid these the same high molecular weight complexes. fixation/immunostaining procedure. problems - in that 25 % of all Pc protein found the different fractions we have isolated mouse anti-Pc antibodies. These polyclonal in seen in lanes 5-7 is localized the large distinct complex us to the of have allowed superimpose binding patterns Pc in Figure 5. and on the same ph preparation by using double-immuno- fluorescence In the two techniques. Figure 7, binding in A Pc and ph show the same distribution patterns are compared. Panel shows the distribution of the on and bind to same the Pc protein visualized by using specific mouse anti-Pc embryogenesis target genes polytene chromosomes antibodies and DTAF-conjugated anti-mouse antibodies as If in we Pc and ph associate the same multimeric complex, secondary antibodies. In is visualized panel B, ph protein would expect equivalent, completely on overlapping patterns the same chromosome set by using the rabbit polyclonal of expression during embryogenesis. 6 shows Figure anti-ph antibodies and rhodamine-conjugated anti-rabbit embryos at two different developmental times stained for antibodies as secondary antibodies. It is clear that the two the presence of the Pc (left Panel) and ph (right panel) are patterns completely overlapping. We found no evidence of proteins. At the stage the germ band extension (stages for separate binding sites of Pc or ph in many preparations as well as after band 7-9), germ retraction (stage 16), the from different glands of different developmental stages. distribution of the two is proteins identical. Detailed We have excluded that the overlapping pattern is caused of the of descriptions developmental profile each protein will by an insufficient filtering of the two fluorescence signals. be presented elsewhere. At no embryonic stage have we seen We could not observe signals when filters for rhodamine a difference when the two major proteins were compared, were used in chromosomes stained for Pc protein alone and small differences in although particular subsets of tissues reciprocally we could not observe a signal when filters for might have gone undetected due to the relatively fluorescein were used in chromosomes stained for the ph homogeneous distribution of the two proteins. protein alone (see Materials and methods for filter specifica- By analysing the distribution of Pc and ph on the polytene tions). Additionally, we have compared the pattern of Pc chromosomes of salivary glands, we could demonstrate that with that of the nuclear protein modulo (Krejci et al., 1989; the biochemically identified Pc-ph complexes are also antibodies kindly provided by J.Pradel, Marseille) in double- present in vivo on their site of action. DeCamillis et al. immunofluorescence staining. Modulo shows a distinct (1992) have presented evidence that Pc and ph have pattern from Pc. The two patterns could clearly be restrained overlapping patterns on polytene chromosomes, but these to the individual fluorescence signal by using the appropriate comparisons were difficult to interpret because only rabbit filters (data not shown). Our results show that Pc and ph polyclonal antibodies were available for each protein. Thus target the same set of genes on polytene chromosomes. the cytological locations of Pc and ph had to be compared Taken together, the signals visible on the chromosomes most 2946 Pc-G genes form a multimeric complex grounds, Jurgens (1985) estimated that there should be around 30-40 different Pc-G genes. Most probably only a subset of them will directly take part in the multimeric complex. Others might be involved in modifying functions or in indirectly controlling the Pc-G/homeotic regulatory system. The complex might also contain proteins that fulfil a more general, structural function and thus may not be readily identified as Pc-G genes. Modifiers of PEV, which were also found to exhibit some homeotic phenotypes, could be such elements. The recently characterized protein BJ1 (Frasch, 1991), which shows homology to the vertebrate gene RCC 1 (Regulator of Chromosome Condensation; Ohtsubo et al., 1991), could be another constituent. Both BJI and the Pc-ph complexes are localized in the condensed bands in polytene chromosomes and have partially over- lapping patterns. Most probably, the multimeric complex is structured by such basic building units that are used to compact many other chromatin domains. The addition of particular Pc-G gene products could subsequently render the complexes more specific for chromatin of homeotic genes or other developmental regulators. Our data from the gel filtration and velocity gradient ultracentrifugation experiments suggest that a large portion of Pc and ph proteins are colocalized in a discrete complex of a molecular mass of 2-5 x 106 Da. However, some of two proteins are also found in lower molecular weight these Though this could be explained by a partial complexes. of the complex during the preparation, it could dissociation also indicate a much more heterogeneous composition of the multimeric units. Tartof and colleagues (Locke et al., 1988; Tartof and Bremer, 1990) have suggested a combinatorial model for the formation of heterochromatin. They interpreted sensitivity of the modifiers of PEV with the law the dose mass action. Multimeric protein complexes, sensitive to of of each individual component, could compact the dose in a cooperative way. Each individual heterochromatin 7. Double-immunofluorescence of polytene chromosomes to Fig. of heterochromatin would be composed of different domain visualize the distribution of the Pc and ph proteins. (A) Staining of of proteins some of which may be commonly chromosomes with polyclonal mouse anti-Pc antibodies and DTAF- combinations anti-mouse antibodies. (B) Staining of chromosomes with conjugated Here we present for the first time molecular evidence shared. rabbit anti-ph antibodies and rhodamine-conjugated anti- polyclonal Pc-G proteins could use similar multimeric that the rabbit antibodies. The distribution of the two proteins is identical. to perform their regulatory role. complexes Chromosomes were photographed with an epifluorescence equipped Zeiss Axiophot using the appropriate filters (see Materials and regulate chromatin by being part of a methods). Pc and ph protein complex represent the binding of the Pc-ph multimeric suggests that some Pc-G genes can interact probably Genetic analysis to their respective target sequences. 1985; Kennison and Russel, 1987). complexes synergistically (Jiirgens, that Pc and ph participate in a multimeric The finding some or all of this synergism. complex could explain Discussion with the individual wild-type protein distributions, Compared no difference of ph protein distribution in we have found were together because of their similar Pc-G genes grouped or of Pc protein distribution in ph- embryos Pc- embryos, phenotype. This suggests that they repress the homeotic M.DeCamillis and H.W.Brock, unpublished 1991; (Franke, genes using a common mechanism. Here we show homeotic This suggests that the synergistic interactions results). members of the Pc-G, Polycomb and polyhomeotic, that two these two (Dura et al., 1985; R.Campbell between proteins constituents of a large multimeric nuclear protein are are not at the transcriptional H.W.Brock, unpublished) and The soluble complex was stable enough to be complex. more to be protein -protein interactions. but are likely level, by immunoprecipitation. Formally however, we purified of homeotic genes can be first detected Pc-G repression have only proven that the Pc and ph proteins are part of the band even when the maternal of elongation, at the stage germ The coprecipitation of 10-15 additional proteins complex. is eliminated (Struhl and Akam, 1985; contribution Pc-G in 1 is suggestive evidence that other proteins Figure depicted Pc and ph can be coprecipitated Gelbart, 1990). Jones and associated with these two Pc-G gene proteins. are potentially in including 0-5 h post- embryogenesis, at all stages could be the identity of these other proteins? Obviously What when the homeotic genes are not repressed in fertilization, members of the Pc-G will be good candidates for other it appears that Pc-ph complex mutant embryos. Thus, Pc-G in the same molecular structure. Based on genetic interacting 2947 A.Franke et al. Zink et al. and DeCamillis et al. for homeotic formation is not sufficient gene complex. (1991) (1992) derepression. of Pc and to have shown that for this It There could be several explanations finding. binding may ph polytene DNA in but it chromosomes is be that the complex is active early development, However, sequence-dependent. neither Pc nor to bind DNA in vitro in does not bind to the homeotic genes early embryogenesis, ph protein appears and and because the chromatin of the homeotic genes is (M.Oed R.Paro, H.W.Brock, N.Cheng unpublished perhaps One candidate for the DNA is the Posterior Mutations in Additional sex inaccessible to the complex. results). binding sex combs which contains DNA combs expression of at the ectopic even-skipped (Psc) protein presumptive permit motifs et van Lohuizen et blastoderm et the stage (Sinclair al., 1992), supporting binding (Brunk al., 1991; al., Another is whether the multimeric that a complex containing Pc-G could be genes 1991a). question suggestion complex the binds a limited number of sites at and acts to active in development. Alternatively, complex early may target genes lack or modifications that allow recognition of the formation of chromatin without a structural components regulate being or whether the is a constituent homeotic which in turn implies that the complex could genes, component itself, complex of chromatin that is itself to alter chromatin have different compositions or properties at different directly required structure effects over chromosomal stages. developmental through cooperative long distances. Potential of Pc-G have been found in Pc and have the same distribution ph during homologues genes mammals. the Pc chromo domain as a similar and a completely overlapping embryogenesis binding Using probe, have been identified in mice and humans on polytene chromosomes pattern proteins (Singh DeCamillis et al. (1992) suggested that most ph binding sites et Pearce et the Psc al., 1991; al., 1992). Similarly, protein is A murine was on chromosomes overlapped with Pc binding conserved. related polytene sites, highly protein, bmi-J but that 1O % of the sites appeared to differ. found to increase the incidence of B-cell However, lymphomagenesis when with it is difficult to compare sites on separate overexpressed the polytene together myc protein Lohuizen et It will also be to chromosome preparations that may differ in age, puffing (van al., 1991b). interesting see whether vertebrates use similar amount of squashing, or intensity of the immuno- multimeric stage, complexes to Using double-label immunofluorescence, we now staining. stably repress developmental regulators by regulating order chromatin structure. demonstrate that all of the target sites detectable using these higher are the same for Pc and ph. We analysed glands reagents of different ages and the - 100 sites always overlapped Materials and methods making it unlikely that there are developmental completely, differences of binding specificity of the two proteins in this of Preparation nuclear extracts tissue. Nuclear were proteins prepared by the method of Gay et al. (1988), with some minor modifications. All The completely overlapping pattern of Pc and ph binding steps were performed at 4°C. As starting 3-6 dechorionated material, g wild type embryos from an overnight to chromosomes poses a paradox, because muta- egg polytene lay were suspended in buffer B (15 mM HEPES pH 7.6, 10 mM KCI, tions in each gene have different phenotypes. Three 5 mM 0.1 mM 0.5 mM 1 mM 1 MgCl2, EDTA, EGTA, mM PMSF, cannot be ruled out. First, minor polytene explanations 2 2 DTT, Leupeptin, and 2.5 ,ug/ml Pepstatin Aprotinin; 4 jig/ml ml/g jg/ml binding sites that have not been detected in our procedure embryos). Embryos were homogenized with 10-15 strokes in a Kontes glass homogenizer using a type A pestle. The homogenate differ for each protein and account for the differences was might pre-cleared by passage through two layers of Miracloth (Calbiochem) and subsequently in Secondly, binding of Pc and ph to polytene phenotype. at 2000 for centrifuged 10 min. The g pellet was resuspended in buffer chromosomes might differ from binding to chromosomes (2 ml/g embryos). The solution was overlayed on a cushion of the same of diploid tissues or in other polytene tissues. Thirdly, volume of B buffer plus 0.8 M sucrose in a centrifuge tube and spun in containing Pc and ph may bind in the same swing-out rotor at 1000 g for 10 min. The nuclear complexes pellet was resuspended in B- buffer 150 mM KCI (0.7 ml/g embryos). Nuclei were by lysed the regions at cytological resolution, but have different binding addition of 4M to (NH4)2SO4 a final concentration of 0.4 M and further sites at the DNA level. Notwithstanding the arguments incubation for 15 min on ice. The lysate was cleared by centrifugation of above, the simplest interpretation of our results is that binding 120 000 1 g for h. The soluble nuclear proteins in the supematant were to a particular site on polytene chromosomes might not precipitated by slow addition of (NH4)2SO4 to a final concentration of 30% (w/v), centrifuged at 30 000 g for 15 min and carefully resuspended necessarily mean that the complex is functional. It may be in buffer C mM (25 HEPES pH 100 mM 7.6, KCl, 5 mM MgCI2, 0.1 mM EDTA, that the Pc-ph complexes are always associated with all 1 mM 1 mM PMSF, DTT, 2 Leupeptin, 2 Pepstatin, 2.5 jig/ml 1g/mi their potential target genes and that additional components itg/ml and Aprotinin 10% glycerol; 100 embryos). 50-100 aliquots were ltl/g IL or modifications of the complex determine it gene-specific frozen in liquid nitrogen and stored for later use at -80°C. Protein Constitutive binding is a commonly used mechanism concentration was activity. measured by the Bradford method (Bradford, 1976). for gene regulation. The recruitment of particular accessory Immunoprecipitation and radioactive labelling of nuclear proteins or a specific modification of the regulatory structure proteins seems to be necessary to invoke the transcriptional activity. Protein A-Sepharose CL-4B (PAS: Pharmacia LKB) was equilibrated in Examples for this are the c-fos induction by the serum buffer CBX (buffer C plus 10 mg/ml BSA and 0.1% Triton X-100) for 30 min. Buffer response factor (Herrera et al., 1989; Dalton and Treisman, CBX was added to give a 50% Protein A-Sepharose suspension. Embryonic - nuclear extract ( 25-50 dl) was mixed with 16 vol 1992) in mammalian cells or the phosphorylation of the heat of buffer CBX. An appropriate amount of antiserum was added to the mixture shock factor in yeast (Sorger and Pelham, 1988). In this and incubated by slow rotation for 30 min at 40C. The optimal amount view, Pc or ph mutations have different phenotypes because of antiserum needs determined to be for each antibody individually. Two complexes lacking each protein are differentially susceptible volumes of the Protein A-Sepharose suspension were added and the incubation continued for 30 minat 4°C. to modification or activation. The complexes were precipitated at 2000 g for 1minin an Eppendorf centrifuge. The pellet was washed We show that two Pc-G are constituents of a large genes five times for 1minin buffer C plus 0.1% Triton X-100 and 0.05% SDS multimeric Additional will be protein experiments and three times in buffer C only. Between complex. each wash the complexes were needed to determine what defines the site of precipitated as above. After the last wash the pellet was either the resuspended binding 2948 Pc-G genes form a multimeric complex in 3-4 vol of SDS-PAGE loading buffer for direct analysis or resuspended anti-rabbit-antibodies diluted 1:40. (Jackson) rhodamine-conjugated protein, fluorochromes were reduced Slides in 100 of 0.1M Na-borate buffer pH 8.5 for radioactive labelling of All performed under light. jtl steps involving the incubation with the The the proteins. were washed as after primary antibody. washed in PBS and embedded in Moviol-2.5% DABCO were Immunoprecipitated proteins were labelled using the 35SLR (Amersham preparations Chromosomes were with International). 50 (1mCi/ml, specific activity >800 Ci/mmol) photographed (1,4-diazobicyclo-(2,2,2)-octane). ACi31SLR Zeiss the filters were dried in a speed vac. The resuspended immunoprecipitate was added, an Axiophot using appropriate equipped epifluorescence filter number 48 79 filter number 48 79 09) mixed and incubated for 30 minon ice. After the reaction unincorporated 15; DTAF, using (rhodamine, 35SLR was neutralized by the addition of an equal volume of 0.2 M KodakT-Max 400. glycine in Na-borate buffer for Smin. SDS and ,B-mercaptoethanol were added to a final concentration of 2 and 5%, respectively and the protein-Sepharose complex dissociated for 5 minat 60°C. Labelled proteins were separated Acknowledgements from the 35SLR-glycine filtration over a small column by gel (Bio-Gel We would like to thank H.P.Blaschkowski for advice in P 10; Biorad). The pooled protein fractions were concentrated by precipitation velocity gradient ultra and filtration R.Paro would like to thank with 2 vol of ethanol for 30 minon ice and a centrifugation for 15 min gel analysis. centrifugation Solf for technical assistance and Annette Schwarz for the at 30 000 g. The protein pellet was resuspended in SDS-PAGE loading Christine typing The work of R.P. was from the buffer. by supported grants manuscript. Bundesministerium fur und and the (0316001A) by Forschung Technologie Deutsche The work of H.W.B. was Gel filtration chromatography and sucrose gradient supported Forschungsgemeinschaft. a from the National Science and Research and Medical ultracentrifugation by Engineering grant Research Councils. For gel filtration we used two media with different separation capacities. Sephacryl HR (Pharmacia) molecules in the between S-300 separates range 10- 1500 kDa, Sephacryl HR in the between 20-1 x 104 S-400 range kDa. The material was suspended in running buffer (25 mM HEPES pH 7.6, References 100mMKCI, 5 mMMgC12, 0.1 mMEDTA and 10% glycerol) to reach and Brunk.B. Dev. Genet., 249-260. 10, a final ratio of 2/3 Sephacryl and 1/3 buffer. The suspension was loaded (1989) Adler,P.N., Charlton,J. and Anal. (1980) into a glass column that was 90 cm long and 2 cm wide. The column was Rubenstein,A.H. Assoian,R.K., Blix,P.M., Tager,H.S. 70-76. calibrated separately using proteins of known molecular weight. For the Biochem., 103, Anal. 248-254. Biochem., 72, Sephacryl S-300 HR column: cytochrome C (13.3 kDa, Boehringer Bradford,M.M. (1976) Dev. 442-456. and 118, Mannheim), BSA (66 kDa, Merck), catalase (240 kDa, Boehringer Duncan,I.M. (1986) Breen,T.R. Biol., and Nature, 353, 351-353. Mannheim) and for the Sephacryl 400-HR; catalase (240 kDa, Boehringer Adler,P.N. (1991) Brunk,B.P., Martin,E.C. 104, 713-720. and Mannheim) and mouse proteasome (750 kDa, Rivett,A.J., 1985, kindly Morata,G. (1988) Development, Busturia,A. and EMBO J., 9, 4267 -4275. Garcia-Bellido,A. (1990) provided by M.Gernold and P.Kloetzel). The exclusion volume of the Castelli-Gair,J.E. and EMBO J., 9, Sephacryl 300-HR column was determined with Blue dextran 2000 Keelan,D.J. Lewis,E.B. (1990) Celniker,S.E., 4277-4286. (Pharmacia). 200 of embryonic nuclear extracts were mixed with 1.5 ml Al 597-612. and Cell, 68, running buffer and loaded on the column. 5 ml fractions were collected. Treisman,R. (1992) Dalton,S. and Gene, 105, Brock,H.W. (1991) 2 of BSA were added as carrier and the fractions mixed with an equal Randsholt,N.B. ytg Deatrick,J., Daly,M., 185- 195. volume of 20% TCA-50% acetone. The were by proteins precipitated and Genes Dev., Brock,H.W. (1992) incubation on ice for 30 at 10 000 for 10 min and Pierre,D. min, centrifuged r.p.m. DeCamillis,M., Cheng,N., 223-232. resuspended in SDS-PAGE buffer. The were neutralized loading samples 6, Dev. 97, 34-47. and by of and stored at For the sucrose Frederick,R.D. (1983) the addition Na2CO3 crystals -20°C. Denell,R.E. Biol., 49-70. Genetics, 102, gradient centrifugation a SW 60 Ti rotor (Beckman Instruments) was used. (1982) Duncan,I.M. In and Green,P.B. (eds), and A 5-20% sucrose gradient in buffer C (25 mM HEPES pH 7.6, 100mM Lewis,E.B. (1982) Subtelny,S. Duncan,I.M. and Liss, New York, Order: Its pp. 5 mM MgCl2, 0.1 mM EDTA, 1 mM PMSF, 1 mM DTT, 2 Origin Regulation. KCI, ytg/ml Developmental 533-554. Leupeptin, 2 Pepstatin and 2.5 1tg/ml Aprotinin) was prepared in tLg/ml Mol. Gen. Genet., 198, and Santamaria,P. (1985) a 4 ml centrifuge tube. 40 of embryonic nuclear extracts were mixed Dura,J.-M., Brock,H.W. Al 213-220. with 160 buffer C and loaded on top of the gradient. Centrifugation was Al 103, 733-741. and Dura,J.-M. (1988) Development, at 50 000 r.p.m. for 12 h at 40C. 20 x 200 1l fractions were collected Ingham,P. Santamaria,P., Deatrick,J., Erk,I., with an ISCO Model 640. Proteins were precipitated with 2 vol ethanol Dura,J.M., Randsholt,N.B., and Brock,H.W. (1987) Cell, Weddell,D. Freeman,S.J., and resuspended in SDS-PAGE loading buffer. The gradient was calibrated Freeman,J.D., 829-839. 51, in separate tubes using different enzymes of known molecular weight (all 11, 14-17. (1989) Bioessays, obtained from Boehringer Mannheim): horse radish peroxidase (44 kDa, Eissenberg,J.C. Foster-Hartnett,D.M., Hartnett,T., Ngan,V. 1 mg/gradient), lactate kDa, 0.5 fumarase James,T.C., dehydrogenase (140 mg/gradient), Eissenberg,J.C., Acad. Sci. USA, 87, 9923-9927. Proc. and (194 kDa, 0.2 mg/gradient), catalase (240 kDa, 0.4 mg/gradient). The (1990) Natl. Elgin,S. of PhD Germany. thesis, Heidelberg, of the in the different fractions was determined by the University localization proteins Franke,A. (1991) EMBO J., 10, 1225-1236. reaction. corresponding enzyme Frasch,M. (1991) Trends Genet., 6, 208-212. and (1990) Gaunt,S.J. Singh,P.B. EMBO J., 7, 4291-4297. and (1988) Poole,S. Kornberg,T. Immunolocalization of proteins on embryos and on polytene Gay,N.J., Trends Genet., 6, 395-400. Grunstein,M. (1990) chromosomes spreads Trends Genet., 6, 422-426. Henikoff,S. (1990) Immunostaining of embryos was based on the procedure described by Nature, 340, 68-70. and Nordheim,A. (1989) Shaw,P.E. Herrera,R.E., MacDonald and Struhl (1986). Fixation and spreading of the chromosomes 815-823. Cell, 37, (1984) essentially followed the protocol of Zink and Paro (1989) and Zink et al. Ingham,P.W. Genetics, 126, 185-199. and Gelbart,W.M. (1990) Jones,R.S. (1991). The chromosome immunostaining protocol was modified slightly Nature, 316, 153-155. (1985) in order to reduce background staining. After washing slides in PBS, they Jurgens,G. and Genetics, 115-133. Wakimoto,B. 94, (1980) Kaufman,T.C., Lewis,R. were transferred to blocking solution (10% BSA, 10% dry milk, 0.2% NP-40 Genetics, 116, 75-86. and Russell,M.A. (1987) Kennison,J.A. and 0.2% Tween 20-80 in PBS) and incubated for 1 h at 370C. Although and (1989) Nucleic Bennani,N. Pradel,J. Garzino,V., Mary,C., Krejci,E., the blocking solution was cloudy, it did not appear to interfere with very 8101-8115. Acids Res., 17, the immunostaining reactions. Slides were briefly washed in PBS and then EMBO J., 7 3233-3244. and McGinnis,W. (1988) Kuziora,M.A. the chromosomes were incubated with the primary antibody at the appropriate (1978) Nature, 276, 565-570. Lewis,E.B. concentration in blocking solution for 1 h at 370C. We used the purified and 181-198. Tartof,K.D. (1988) Genetics, 120, Kotarski,M.A. Locke,J., anti-Pc and rabbit anti-ph antibodies at a 1:100 mouse polyclonal polyclonal Nature, 324, 537-545. and (1986) Struhl,G. MacDonald,P.M. dilution. Slides were briefly washed in PBS and then vigorously shaken and Genes Dev., in press. Paro,R. (1992) Franke,A. Messmer,S., for 15 min in 300 mM NaCl, 0.2% NP-40 and 0.2% Tween 20-80 in PBS, and Brock,H.W. (1991) Rouxs Arch. Dev. BioL. 199, 387-396. McKeonj. and for additional 15 min in 400 mM NaCl, 0.2% NP-40 and 0.2% Tween Muller,J. and Bienz,M. (1991) EMBO J., 10, 3147-3155. in PBS. Slides were briefly washed in PBS and blocked in 20-80 again Ohtsubo,M., Yoshida,T., Seino,H., Nishitani,H., Clark,K.L., Sprague,G., blocking solution as above. The incubation with the secondary antiy Frasch,M. and Nishmoto,T. (1991) EMBO J., 10, 1265-1273. for 1 h at in solution. For the Pc protein, we used DTAF- was 37°C blocking Paro,R. (1990) Trends Genet., 6, 416-421. anti-mouse antibodies (Jackson) diluted 1:100 and for the ph conjugated 2949 A.Franke et al. and Hogness,D.S. (1991) Proc. Natl. Acad. Sci. USA, 88, Paro,R. 263-267. Singh,P.B. and Gaunt,S.J. (1992) Development, 114, Pearce,J.J.H., 921-930. Reuter,G., Giarre,M., Farah,J., Gausz,J., Spierer,A. and Spierer,P. (1990) Nature, 344, 219-223. (1985) J. Bio. Chem., 260, 12600-12606. Rivett,A.J. and Denell,R.E. (1985) Dev. Biol., 110, 53-63. Sato,T. Simcox,A.A., Hersperger,E., Shearn,E., Whittle,J.R.S. and Cohen,S.M. 34, 11-20. (1991) Mech. Dev., and (1992) Development, 114, in press. Simon,J., Chiang,A. Bender,W. Niichools,F., Slade,E. and Brock,H.W. Sinclair,D.A., Campbell,R.B., 130, 817-825. (1992) Genetics, Miller,J.R., Pearce,J., Kothary,R., Burton,R.D., Paro,R., Singh,P.B., and Gaunt,S. (1991) Nucleic Acids Res., 19, 789-794. James,T.C. Smouse,D. and Perrimon,N. (1990) Dev. Biol., 139, 169-185. Smouse,D., Goodman,C., Mahowald,A. and Perrimon,N. (1988) Genes 830-842. Dev., 2, (1988) Cell, 54, 855-864. Sorger,P.K. and Pelham,H.R.B. Nature, 293, 36-41. Struhl,G. (1981) Struhl,G. and Akam,M. (1985) EMBO J., 4, 3259-3264. Tartof,K.D. and Bremer,M. (1990) Development, Suppl., 35-45. Tiong,S.Y.K. and Russell,M.A. (1990) Dev. Biol., 141, 306-318. van Lohuizen,M., Frasch,M., Wientjens,E. and Berns,A. (1991a) Nature, 353-355. 353, van Lohuizen,M., Verbeek,S., Scheijen,B.W., van der Gulden,H. and Cell, 65, 737-752. Berns,A. (1991b) Wedeen,C., Harding,K. and Levine,M. (1986) Cell, 44, 739-748. Wu,C.-T., Jones,R.S., Lasko,P.F. and Gelbart,W.M. (1989) Mol. Gen. Genet., 218, 499-564. Zink,B. (1990) PhD thesis, University of Heidelberg, Germany. Zink,B. and Paro,R. (1989) Nature, 337, 468-471. Zink,B., Engstrom,Y., Gehring,W.J. and Paro,R. (1991) EMBO J., 10, 153-162. Received on April 1, 1992; revised on May 11, 1992 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The EMBO Journal Springer Journals

Polycomb and polyhomeotic are constituents of a multimeric protein complex in chromatin of Drosophila melanogaster.

Download PDF
10 pages

Loading next page...
 
/lp/springer-journals/polycomb-and-polyhomeotic-are-constituents-of-a-multimeric-protein-Peaf3StS2K

References (1)

Publisher
Springer Journals
Copyright
Copyright © European Molecular Biology Organization 1992
ISSN
0261-4189
eISSN
1460-2075
DOI
10.1002/j.1460-2075.1992.tb05364.x
Publisher site
See Article on Publisher Site

Abstract

The EMBO Journal vol.1 1 no.8 pp.2941 -2950, 1992 Polycomb and polyhomeotic are constituents of a multimeric protein complex in chromatin of Drosophila melanogaster responsible for maintaining the repressed state of homeotic Axel Franke1, Marco DeCamillis2, genes (reviewed in Paro, 1990). Mutations in members of Daniele Zink1, Niansheng Cheng2, the Pc-G show an ectopic expression of homeotic regulators W.Brock2 and Renato Paro1 3 Hugh (Struhl and Akam, 1985; Weeden et al., 1986; Dura and 'ZMBH, University of Heidelberg, Im Neuenheimer Feld 282, D-6900 Ingham, 1988; McKeon and Brock, 1991; Simon et al., Heidelberg, Germany and 2Department of Zoology, University of 1992), which can result in dramatic transformations of body British Columbia, 6270 University Boulevard, Vancouver, BC, structures (Struhl, 1981; Duncan, 1982; Duncan and Lewis, V6T 1Z4, Canada 1982; Ingham, 1984; Dura et al., 1985; Jurgens, 1985; 3Corresponding author Breen and Duncan, 1986; Adler et al., 1989; Jones and Communicated by H.Jackle 1990). Pc-G genes are not only involved in Gelbart, homeotic gene expression patterns, but have maintaining been found to perform similar roles on other developmental The polycomb group (Pc-G) genes are responsible for regulators (Ingham, 1984; Breen and Duncan, 1986; Dura maintaining the repressed state of homeotic genes during development. It has been suggested that the Pc-G exerts and Ingham, 1988; Smouse et al., 1988; Busturia and its transcriptional control by regulating higher order Morata, 1988; Wu et al., 1989). chromatin structure. In particular, the finding of genetic Two genes of the Pc-G group have been extensively and molecular similarities to components involved in characterized at the genetic and molecular level. mutant embryos of the Polycomb (Pc) gene heterochromatin formation, led to the proposal that Homozygous genes are permanently repressed by mechan- display posterior transformations of all segments, charac- homeotic teristic for a deregulation of all the homeotic genes of the similar to those responsible for heterochromatin isms ANT-C and BX-C (Lewis, 1978; Denell and Frederick, compaction. Because of synergistic effects, Pc-G gene Sato and Denell, 1985; Tiong and Russell, 1990). Pc products are thought to act in a multimeric complex. 1983; found to exert its regulatory role at the transcriptional Using immunoprecipitation we show that two members was level (Weeden et al., 1986; Zink et al., 1991). The Pc of the Pc-G, Polycomb and polyhomeotic, are constituents is a nuclear protein associated with - 100 different of a soluble multimeric protein complex. Size fractiona- protein sites on polytene chromosomes of larval salivary glands tion indicates that a large portion of the two proteins are 1989; Zink, 1990). Reporter genes linked (Zink and Paro, found in a distinct complex of molecular weight 2-5 x homeotic cis-regulatory sequences become ectopically to 106 Da. During embryogenesis the two proteins show the when tested in a background lacking Pc+ distribution. In addition, by double- expressed same spatial and Bienz, 1991; Zink et al., 1991). labelling we can demonstrate that function (Muller immunofluorescence et al. showed that the Pc protein Zink (1991) and polyhomeotic have the same binding Additionally, Polycomb exactly sequences of the Antenna- binds to regulatory on polytene chromosomes of larval salivary specifically patterns gene. We that some Pc-G proteins act in pedia glands. propose of the other well characterized gene, analysis Genetic multimeric complexes to compact the chromatin of stably revealed a high degree of pleiotropy of (ph), like the homeotic regulators. polyhomeotic repressed genes Although the homeotic transforma- mutant phenotype. words: chromatin regulation/Drosophila/homeotic gene the Key classify it as a Pc-G gene, additional defects tions observed group/polytene chromosomes regulation/polycomb death in the ventral epidermis (Dura et al., 1987) like cell central nervous system axons (Smouse of and misrouting and 1990) can be identified. Smouse Perrimon, et al., 1988; Introduction structure and is arranged as a a complex gene ph shows unit. lesions in both units result Only mechanisms that maintain the determined state tandemly duplicated Regulatory alleles, whereas single lesion or hypomorphic are necessary for the proper development of an in null strong of a cell weak phenotypes. The two only hypomorphic alleles display body plan. In Drosophila melanogaster, the organized within a stretch of 25 kb of genomic units contained of the homeotic genes of the genetic activity differential that have certain characteristics similar proteins DNA encode and bithorax-complex (BX-C) Antennapedia- (ANT-C) with DNA (Deatrick et al., 1991; interacting of proteins the various body structures along the anterior- determine al., In addition to four blocks of et 1992). DeCamillis axis 1978; Kaufman et al., 1980). Since posterior (Lewis, the and serine/threonine-rich sequences, glutamine repeats is throughout development, they have their function required a with partial homology to the displays region a model to identify and characterize ph protein been used as system and a C4 zinc finger. Like motif single for the transmission of the determined helix-loop-helix the elements required - 100 chromosome sites and ph to binds polytene which define. Genetic analysis has revealed a Pc, ph states they sequences from the recognizes regulatory specifically distinct class of genes involved in part of this process. The et al., 1992). region (DeCamillis bithoraxoid of the Polycomb group (Pc-G) were found to be genes Press Oxford University A.Franke et al. The we found that our Pc-G genes display some common features described rabbit anti- with previously polyclonal another Pc antibodies distinct class and of Drosophila the modifiers genes, (Zink Paro, 1989) could be used of to position effect variegation in vitro (PEV). These are translated Pc genes as well thought immunoprecipitate as protein to encode structural Pc and constituents from regulatory of nuclear extracts, chromatin endogenous protein embryonic or (reviewed in from nuclear Eissenberg, 1989; extracts of Henikoff, 1990). Some Schneider L-2 tissue Pc- culture cells. To G genes show effects on the PEV (D.Sinclair, perform we N.Clegg, immunoprecipitations, essentially T.Grigliatti followed and the H.W.Brock, submitted) and of et al. who reciprocally, protocol Gay (1988), were able some modifiers to of PEV homeotic the association display of the with transformations identify other engrailed protein (Reuter et nuclear al., 1990; factors. In this G.Reuter, personal method the nuclei are communication). a lysed by Paro and Hogness (1991) have found a molecular shock and the released soluble nuclear relationship hypertonic proteins between the Pc protein and the and heterochromatin-associated concentrated a 30% partially purified ammonium by HP1 protein, which encoded sulfate is The by Su(var)205 redissolved (Eissenberg soluble nuclear precipitation. proteins et al., 1990). were Both proteins share the chromo with anti-Pc domain, antibodies and a immunoprecipitated Protein 48 amino acid motif found to be A for beads responsible Materials specific (see and -Sepharose methods). chromatin To binding (Messmer et Based on if Pc is al., 1992). associated with determine other nuclear these proteins similarities, it has been in the proposed that the Pc-G soluble we genes the repress fraction, separated immunoprecipitate homeotic genes by regionally the SDS -PAGE. compacting these chromatin initial by However, did not experiments (heterochromatinization) and thus eliminating the to a yield visible enough coprecipitated proteins give pattern accessibility ofDNA to in silver diffusible stained transcription factors. To circumvent This the gels. sensitivity problem mechanism could the we imprint order tried to higher structure in of vivo labelled nuclear immunoprecipitate chromatin with the determined state of the cell as from tissue culture cells. defined Pc is proteins endogenously by the differential expression pattern of in these developmental cells et expressed (Messmer Cells were al., 1992). regulators like homeotic genes (Locke et for 3 h al., 1988; Gaunt with labelled and [35S]methionine and [35S]cysteine, and Singh, 1990; Paro, 1990; Reuter et nuclear al., were 1990). with anti-Pc proteins immunoprecipitated Because modifiers of the PEV show a antibodies and very on high gene separated gels. Disappointingly, only very dose sensitivity, Locke et al. weak (1988) proposed could be that observed after signals prolonged exposures heterochromatin is packaged by reiterated not multimeric protein Because the (data shown). Pc immunoprecipitated protein complexes composed of the products of seemed to these have genes. a small amount of incorporated only Heterochromatin apparently can spread we along attributed the the radioactivity, result to the negative chromosome in a cooperative fashion and thus transcrip- of Pc and the other relatively high stability associated tionally inactivate large chromosomal Pc-G regions. Jiirgens have a low proteins. such proteins that might turnover, (1985) has found that the double and a 3 h triple mutant is too short to labelling period incorporate sufficient combination of four of genes the Pc-G, Additional sex we combs have evidence radioactivity. from the Indeed, develop- (Asx), Polycomb-like (Pet), Posterior sex mental of combs (Psc) Pc that the and is stable profile protein very Sex during combs on midleg (Scm) showed a marked enhancement and embryogenesis (B.Zink R.Paro, unpublished results). of the homeotic transformation compared with the single the However, by radioactively after labelling proteins mutations. This synergistic effect of the Pc-G could be we were able to immunoprecipitation, visualize the Pc- explained by a participation of the associated different For Pc-G gene this we used proteins. a purpose modified products in a common regulatory structure, i.e. a Bolton-Hunter multimeric to 35S the reagent couple to precipitated protein complex. Pc-G proteins could form multimeric et proteins see also (Assoian al., Materials 1980; and protein units that are functionally and The mechanistically of this methods). method disadvantage is that the equivalent to heterchromatic complexes, except that antibodies in the are also labelled precipitate and result in particular Pc-G proteins would target a the complexes on the to that can strong signal autoradiogram obscure the euchromatic genes. of other in the same signals molecular proteins weight range. Here we present evidence that two members of shows the Pc-G the result of 1 this Figure In experiment. lane 3, are part of a large multimeric protein 10-15 complex. new DeCamillis Pc-associated are seen to proteins coprecipitate et al. (1992) have shown that the ph and Pc with proteins have with Pc, the compared control lanes 5 and 4, 6 (see many overlapping binding sites on polytene to chromosomes. In lane no legend Figure 1). band 3, major corresponding We show that this in vivo correlation is based on to the a molecular shown in Lane as an Pc-protein, 1 in vitro association produced of the two proteins in a soluble nuclear complex. 35S-labelled is visible. protein, this However, is not Using immunoprecipitations we can demonstrate that Pc and that the surprising, labelling of proteins with considering the ph are found in a multimeric complex. We have furthermore can 35S-labelling reagent SRL, substantially change the refined the correlation of the Pc and ph binding patterns on of in SDS mobility -PAGE proteins et (Assoian al., polytene 1980). chromosomes by using The results double-immunofluorescence shown in Figure clearly indicate that Pc is techniques. We propose that the Pc-G proteins exert associated with their several other in proteins the soluble nuclear repressory function on the chromatin of homeotic genes as fraction. The multimeric protein is complex stable enough multimeric protein units. to be immunoprecipitated with anti-Pc antibodies, providing molecular means of possible characterizing the various Results components. Immunoprecipitation of Pc and associated proteins and Polycomb polyhomeotic are colocalized in the from nuclear extracts same multimeric protein complex We used immunoprecipitations to enrich for proteins that Biochemical identification of the coprecipitated proteins may associate with the Pc protein. In an initial experiment would be very difficult, because only small amounts of 2942 Pc-G genes form a multimeric complex material could be recovered. However, we were able to take advantage of the considerable amount of genetic and molecular data available on the Pc-G system. Obvious candidates for proteins interacting at the molecular level with Pc are other members of the Pc-G. Indeed, we have shown previously that Pc and ph have similar binding patterns on _- Pc - Fc ., LA. ,, - _ow"' Pr, - Sw*fi.D IL - Fig. 1. Immunoprecipitation of Pc-associated proteins from embryonic nuclear extracts. 25 11 (protein concentration 15 mg/mil) of an embryonic nuclear extract was immunoprecipitated in a total volume of 500 using anti-Pc antibodies. The precipitate was labelled with Al 35SLR (see Materials and methods) and fractionated on a 8% SDS-PAGE gel. The gel was dried on a piece of Whatman paper and exposed for 24 h. Lane 1, in vitro translated 35S-labelled Pc protein as a size reference; lane 2, 14C protein molecular mass standards with the on corresponding sizes given the left, lane 3, immunoprecipitations .. F- S.h:... with anti-Pc antibodies and protein A-Sepharose; lane 4, the same 4bl antibodies; immunoprecipitation as in lane 3 but omitting the anti-Pc lane 5, mock-immunoprecipitation without nuclear extracts. Note the strong signal from the immunoglobulin chains; lane 6, mock- Westerns were incubated with alkaline phosphatase coupled anti-rabbit immunoprecipitation without nuclear extract and without antibodies. the new bands in lane 1 are antibody as secondary antibodies. Thus, not due to with the The bands cross-reactivity secondary antibody. denoted with Fc in lanes 1 and 3 are caused the of by cross-reactivity Fig. 2. The Pc and ph protein are coimmunoprecipitated in the same the with the anti-Pc antibodies. secondary antibodies (C) multimeric protein complex. Panels A and B show an of Pc with nuclear Coprecipitation anti-ph antibody. Embryonic immunoprecipitation of 100 embryonic nuclear extract (protein extracts were with The immunoprecipitated anti-ph antibody. Al concentration 15 mg/ml) with anti-Pc antibodies in a total volume of precipitates were separated by SDS-PAGE, transferred to examined Western for the of 1.5 ml. One-third of the precipitate in A and two-thirds of the nitrocellulose and using blotting presence Western blots with antibodies to precipitate in B were separated on a 8% (A) and 6% (B) SDS-PAGE ph and Pc. Lanes 1-5 show probed with anti-Pc lanes 6-7 show western blots with antibodies to Pc. and transferred to nitrocellulose. Blot A was ph and probed probed Pc Blot Lane Western blot of an nuclear extract before antibodies and shows the of the 1, embryo specific precipitation protein. of lane of the to show B was probed with anti-ph antibodies to show the co-purification immunoprecipitation mobility ph products; 2, with Some Lane of nuclear extract with anti- of the same extract ph protein. 1, immunoprecipitation immunoprecipitation ph antibody. Lane Pc antibodies and Protein lane mock- of occurs. A-Sepharose; 2, degradation ph 3, mock-immunoprecipitation lacking without lane mock- Protein lane immunoprecipitation primary antibodies; 3, A-Sepharose; 4, mock-immunoprecipitation lacking nuclear without nuclear lane mock- lane extract; immunoprecipitation extract; 4, antibody; 5, mock-immunoprecipitation lacking shown in Pc reacted with the same without nuclear extract and lane lane 6, antibody immunoprecipitate immunoprecipitation primary antibody; with lane of the Pc and lane 2 to show that Pc 7, 5 nuclear extract to show the coprecipitates anti-ph antibody; 5, mobility ph proteins, I1 extract before forms. At least blot of an nuclear The encodes Western embryo immunoprecipitation respectively. ph gene multiple protein molecular bands above the in of Pc. The four different can be detected to show mobility higher weight ph proteins unstaged embryos with the used Due to the Pc are due to and in protein cross-reactivity secondary antibody (M.DeCamillis W.H.Brock, preparation). instability seems to have occurred. Both in this of the ph protein, some degradation particular experiment. 2943 A-Franke et at. polytene chromosomes of of the salivary glands (DeCamillis et presence some al., ph protein. Although degradation seems 1992). We to therefore tested have ifph copurifies in the most of the immuno- occurred, are ph products found in the precipitation with anti-Pc antibodies. In results lane The are the same 6, shown precipitate. is immunoprecipitate probed in Figure 2A and for 2B. the of of the Immunoprecipitations Pc The embryonic presence protein. comparison with the nuclear extracts were nuclear performed, on SDS extract in lane separated 7 shows that -PAGE Pc can also be and blotted onto nitrocellulose filters. In Figure 2A, antibodies. the precipitated by using anti-ph We find that in Western blot was probed with anti-Pc antibodies. Lane with 1, general immunoprecipitation antibodies is anti-ph less which contains the normal efficient than immunoprecipitate with the clearly anti-Pc shows antibodies. This could reflect the enrichment a of the Pc reduced when protein, with of the compared the accessibility in the ph protein or complex, mock-immunoprecipitations in the control lanes 3 difference in the and 2, 4 protein of the composition for complex (see legend to Figure the two 2A). The same set of the immuno- proteins. Alternatively, could be due discrepancy precipitations were tested for the to a presence of the different ph of the two protein specificity antibodies used, although in Figure 2B. Lane 1 identifies as one this seems ph of the as the two proteins quite unlikely antibodies gave that can be copurified with Pc, that Pc and results on indicating blots. ph comparable Western are part of the same multimeric protein The complex. of the two coprecipitation could also be proteins To prove the specificity of the interaction and a to explained of the by anti-Pc and cross-reactivity anti-ph demonstrate the of reproducibility the antibodies. in immunoprecipitation However, blots with Western probed anti-Pc protocol, we performed the reciprocal we experiment. never antibodies, detected a Using band at the cross-reacting specific anti-ph antibodies (DeCamillis et size of al., the 1992), we This is expected also the ph protein. case for the were able to coimmunoprecipitate the Pc protein with from reciprocal antibodies experiment also anti-ph (see embryonic nuclear extracts. Figure 2C, lane In order 2, shows to the 2). exclude Figure this completely result of possibility, the immunoprecipitation we have analysed used a first for more the In vitro specific translated approach. Pc and were 35S-labelled ph proteins immunoprecipitated with and anti-Pc anti-ph antibodies, respectively 3). (Figure In both cases we found no evidence for a cross-reactivity of the two This antibodies. that the proves of copurification the Pc and the in the ph proteins are due immunoprecipitates to the of these two presence in the same proteins multimeric complex. We tested the anti-Pc for the immunoprecipitate presence of additional that could a in proteins the play part regulatory and for which we had system, antibodies. Histone specific has an role in HI the of important inactive compaction chromatin in (reviewed Grunstein, antibodies 1990). Using Histone against Drosophila HI (kindly provided by S.Elgin, we have found that St. Louis) does not HI with coprecipitate Pc and thus is not a constituent of the major multimeric not protein A similar complex (data result shown). negative Fig. 3. Control of cross-reactivity of the anti-ph antibody with Pc and of the anti-Pc antibody with ph. (A) Immunoprecipitation of 35S- labelled Pc protein produced in a reticulocyte lysate. 5 jig of the translation mixture was incubated in a total volume of 500 ,ul. The precipitate was fractionated on an 8% SDS-PAGE gel and exposed overnight. Lane 1, immunoprecipitation with anti-ph antibodies and Protein A-Sepharose; lane 2, immunoprecipitation with anti-Pc antibodies and Protein A-Sepharose; lane 3, mock-immuno- precipitation without antibody; lane M, 14C protein marker as size standards that are given on the right. (B) Control for cross-reactivity 4. Fig. of of nuclear anti-Pc antibody Immunoprecipitations extracts with ph from protein. ph proteins that had been staged with anti-Pc antibodies. embryos The translated in a were reticulocyte lysate were precipitates separated immunoprecipitated with SDS-PAGE by and transferred to antibodies to ph and nitrocellulose. Pc. The The filter immunoprecipitates were upper was separated by incubated with antibodies and the SDS-PAGE and anti-ph lower filter exposed to with autoradiography. anti-Pc Lane 1, ph mRNA antibodies. At translated in the a every reticulocyte lysate stage analysed concentration containing was protein [35S]methionine; lane normalized to have a 2, constant amount of immunoprecipitation of ph translation products with Pc anti-ph immunoprecipitated antibody; On of the lane protein. lanes the 3, top individual immunoprecipitation of ph translation are products with embryonic anti-Pc stages denoted the time from antibodies. Note by the (h) Both absence of filters egg were Lane laying. incubated immunoprecipitation. with alkaline 4, immunoprecipitation using ph antibody of a phosphatase anti-rabbit lysate that did coupled antibodies as not contain antibodies. The band below secondary ph the Pc message; lane 5, strong is due to the mock-immunoprecipitation of signal cross- lysate containing ph with the chains translation reactivity of the products and protein heavy A, but antibodies. lacking ph primary Molecular antibody. standards are indicated weight on the left. 2944 Pc-G genes form a multimeric complex Size measurements of the multimeric protein complex was obtained when the immunoprecipitate was probed with The data we have presented suggest that Pc is associated antibodies against the homeotic proteins Antennapedia with 10-15 other nuclear proteins and show that the Pc-G (kindly provided by W.Gehring, Basel) and Ultrabithorax member ph is one of the constituents of this soluble (kindly provided by R.White, Cambridge). This indicates multimenc complex. Using gel filtration and density gradient that Pc-G proteins do not directly interact with homeotic centrifugation methods, we have determined the approximate proteins to exert their negative regulatory role. This molecular weight of the complex. As starting material we conclusion had also been suggested by genetic experiments used the same embryonic nuclear extracts and buffers as for (Castelli-Gair and Garcia-Bellido, 1990). We found no the immunoprecipitations, since under these conditions the evidence of a coprecipitation of the engrailed protein, complex was shown to be stable. The best separation of indicating that the engrailed-associated multimeric protein nuclear extract was achieved by gel filtration chromatography complexes identified by Gay et al. (1988) are distinct from Sephacryl S400 HR column (see Materials and on a the Pc-ph complexes identified here. methods). Different fractions were collected, separated by SDS-PAGE, blotted onto nitrocellulose and tested for the Developmental profile of the Pc - pH interaction presence of Pc and ph protein, respectively. Figure 5 shows The activity of the Pc-G is needed after germ band elonga- that significant amounts of Pc and ph proteins copurify as tion, to take over and maintain the homeotic expression a high molecular weight complex. Most of the Pc-ph pattern initiated by the early maternal and segmentation genes over a narrow molecular weight range. elutes protein (Struhl and Akam, 1985; Weeden et al., 1986; Kuziora and Extrapolating from the molecular weight markers indicated McGinnis, 1988; Celniker et al., 1990; McKeon and Brock, to Figure 5), we have estimated that the weight legend (see 1991; Simcox et al., 1991; Simon et al., 1992). We wanted of the Pc-ph complex is -2-5 x 103 kDa. A more to test whether this activity is regulated at the level of weight could not be obtained because of the absence accurate complex formation. In particular, we have tried to identify a suitable molecular weight marker. There is a clear upper of at what embryonic stages Pc is associated with ph. We corresponding to the size of the complex (Figure 5, band collected nuclear extracts from staged embryos, performed 8), but there is a smear of smaller complexes decreasing lane immunoprecipitations with anti-Pc antibodies and analysed molecular weight to the size of the monomers. Molecular in for the copurification of ph. We normalized the mass of of these smaller complexes were estimated using weights immunoprecipitated proteins to give a constant amount of S-300 HR column that separates molecules in a Sephacryl Pc protein in each lane (Figure 4). Our data suggest that of 10-1500 kDa (data not shown). It is not clear the range Pc and ph can be coprecipitated over the entire period of the smear is due to partial dissociation of the whether embryogenesis. Although the ratio between the coprecip- during preparation, or whether the two proteins complexes itated ph and Pc proteins clearly increases during with other proteins, resulting in multimeric are associated embryogenesis, there is no stage at which the two proteins of different composition and thus of different complexes are not associated. This shows that the Pc-ph complexes sizes. form early in development during the initiation stage of comparable results by separating the have obtained We homeotic gene regulation, suggesting that Pc-G repression on sucrose gradients (data not shown). nuclear complexes homeotic expression is most probably controlled by a of most of the Pc and ph is found in the After centrifugation, means than formation of the Pc-G complex. different Fig. 5. Cofractionation of Pc and ph in multimeric protein complexes separated by gel filtration. Nuclear protein extracts of wild type embryos (200 jd, 15 mg/ml) was applied to a column of Sephacryl S-400 HR (90 cm long, 2 cm wide). Fractions were collected after the expected elution of catalase (240 kDa, see Materials and methods). 12 x 10 ml fractions were collected until the exclusion volume was reached. Fractions were concentrated and applied to a 7.5% SDS-PAGE gel. The electrophoretically separated proteins were transferred to nitrocellulose. The upper half of the filter was probed with anti-ph antibodies and the lower half with anti-Pc antibodies. The Western blot was developed using alkaline phosphatase coupled anti-rabbit antibodies as secondary antibodies. Lane NE, 5 A1 (15 mg/ml) embryonic nuclear extract to indicate the expected size of Pc and ph; lanes I -11, different fractions analysed. The column was equilibrated separately with marker proteins indicated at the bottom. 240 kDa, catalase; 750 kDa, mouse proteasome. The exclusion volume was determined by specifications given by the supplier (Pharmacia). Dextran blue 2000 (2 x 106 Da) could not be used as a size standard since it seemed to interact with the matrix, resulting in a broad elution peak. 2945 A.Franke et al. in The 6. of the Pc and distribution of the two were the same at all As Fig. Distribution ph protein embryos. spatial proteins embryonic stages. an and after at band band retraction at 16 are Note the intense of the CNS example embryos germ elongation (top) germ stage depicted. labelling in both latter cases. of the exceeds pellet, suggesting that the from different and different and weight complex glands chromosomes, our the separation of this method. The elution experience has shown that a of so capacity profiles comparison many binding exclusion in the case of Pc and - can be of Pc and were the same or sites, ph 100, very difficult. ph using gel in These results that we have found that different can centrifugation sucrose gradients. suggest Furthermore, preparations of the Pc and show intensities due to the independent techniques used, ph participate different signal complexity of the in We estimate To avoid these the same high molecular weight complexes. fixation/immunostaining procedure. problems - in that 25 % of all Pc protein found the different fractions we have isolated mouse anti-Pc antibodies. These polyclonal in seen in lanes 5-7 is localized the large distinct complex us to the of have allowed superimpose binding patterns Pc in Figure 5. and on the same ph preparation by using double-immuno- fluorescence In the two techniques. Figure 7, binding in A Pc and ph show the same distribution patterns are compared. Panel shows the distribution of the on and bind to same the Pc protein visualized by using specific mouse anti-Pc embryogenesis target genes polytene chromosomes antibodies and DTAF-conjugated anti-mouse antibodies as If in we Pc and ph associate the same multimeric complex, secondary antibodies. In is visualized panel B, ph protein would expect equivalent, completely on overlapping patterns the same chromosome set by using the rabbit polyclonal of expression during embryogenesis. 6 shows Figure anti-ph antibodies and rhodamine-conjugated anti-rabbit embryos at two different developmental times stained for antibodies as secondary antibodies. It is clear that the two the presence of the Pc (left Panel) and ph (right panel) are patterns completely overlapping. We found no evidence of proteins. At the stage the germ band extension (stages for separate binding sites of Pc or ph in many preparations as well as after band 7-9), germ retraction (stage 16), the from different glands of different developmental stages. distribution of the two is proteins identical. Detailed We have excluded that the overlapping pattern is caused of the of descriptions developmental profile each protein will by an insufficient filtering of the two fluorescence signals. be presented elsewhere. At no embryonic stage have we seen We could not observe signals when filters for rhodamine a difference when the two major proteins were compared, were used in chromosomes stained for Pc protein alone and small differences in although particular subsets of tissues reciprocally we could not observe a signal when filters for might have gone undetected due to the relatively fluorescein were used in chromosomes stained for the ph homogeneous distribution of the two proteins. protein alone (see Materials and methods for filter specifica- By analysing the distribution of Pc and ph on the polytene tions). Additionally, we have compared the pattern of Pc chromosomes of salivary glands, we could demonstrate that with that of the nuclear protein modulo (Krejci et al., 1989; the biochemically identified Pc-ph complexes are also antibodies kindly provided by J.Pradel, Marseille) in double- present in vivo on their site of action. DeCamillis et al. immunofluorescence staining. Modulo shows a distinct (1992) have presented evidence that Pc and ph have pattern from Pc. The two patterns could clearly be restrained overlapping patterns on polytene chromosomes, but these to the individual fluorescence signal by using the appropriate comparisons were difficult to interpret because only rabbit filters (data not shown). Our results show that Pc and ph polyclonal antibodies were available for each protein. Thus target the same set of genes on polytene chromosomes. the cytological locations of Pc and ph had to be compared Taken together, the signals visible on the chromosomes most 2946 Pc-G genes form a multimeric complex grounds, Jurgens (1985) estimated that there should be around 30-40 different Pc-G genes. Most probably only a subset of them will directly take part in the multimeric complex. Others might be involved in modifying functions or in indirectly controlling the Pc-G/homeotic regulatory system. The complex might also contain proteins that fulfil a more general, structural function and thus may not be readily identified as Pc-G genes. Modifiers of PEV, which were also found to exhibit some homeotic phenotypes, could be such elements. The recently characterized protein BJ1 (Frasch, 1991), which shows homology to the vertebrate gene RCC 1 (Regulator of Chromosome Condensation; Ohtsubo et al., 1991), could be another constituent. Both BJI and the Pc-ph complexes are localized in the condensed bands in polytene chromosomes and have partially over- lapping patterns. Most probably, the multimeric complex is structured by such basic building units that are used to compact many other chromatin domains. The addition of particular Pc-G gene products could subsequently render the complexes more specific for chromatin of homeotic genes or other developmental regulators. Our data from the gel filtration and velocity gradient ultracentrifugation experiments suggest that a large portion of Pc and ph proteins are colocalized in a discrete complex of a molecular mass of 2-5 x 106 Da. However, some of two proteins are also found in lower molecular weight these Though this could be explained by a partial complexes. of the complex during the preparation, it could dissociation also indicate a much more heterogeneous composition of the multimeric units. Tartof and colleagues (Locke et al., 1988; Tartof and Bremer, 1990) have suggested a combinatorial model for the formation of heterochromatin. They interpreted sensitivity of the modifiers of PEV with the law the dose mass action. Multimeric protein complexes, sensitive to of of each individual component, could compact the dose in a cooperative way. Each individual heterochromatin 7. Double-immunofluorescence of polytene chromosomes to Fig. of heterochromatin would be composed of different domain visualize the distribution of the Pc and ph proteins. (A) Staining of of proteins some of which may be commonly chromosomes with polyclonal mouse anti-Pc antibodies and DTAF- combinations anti-mouse antibodies. (B) Staining of chromosomes with conjugated Here we present for the first time molecular evidence shared. rabbit anti-ph antibodies and rhodamine-conjugated anti- polyclonal Pc-G proteins could use similar multimeric that the rabbit antibodies. The distribution of the two proteins is identical. to perform their regulatory role. complexes Chromosomes were photographed with an epifluorescence equipped Zeiss Axiophot using the appropriate filters (see Materials and regulate chromatin by being part of a methods). Pc and ph protein complex represent the binding of the Pc-ph multimeric suggests that some Pc-G genes can interact probably Genetic analysis to their respective target sequences. 1985; Kennison and Russel, 1987). complexes synergistically (Jiirgens, that Pc and ph participate in a multimeric The finding some or all of this synergism. complex could explain Discussion with the individual wild-type protein distributions, Compared no difference of ph protein distribution in we have found were together because of their similar Pc-G genes grouped or of Pc protein distribution in ph- embryos Pc- embryos, phenotype. This suggests that they repress the homeotic M.DeCamillis and H.W.Brock, unpublished 1991; (Franke, genes using a common mechanism. Here we show homeotic This suggests that the synergistic interactions results). members of the Pc-G, Polycomb and polyhomeotic, that two these two (Dura et al., 1985; R.Campbell between proteins constituents of a large multimeric nuclear protein are are not at the transcriptional H.W.Brock, unpublished) and The soluble complex was stable enough to be complex. more to be protein -protein interactions. but are likely level, by immunoprecipitation. Formally however, we purified of homeotic genes can be first detected Pc-G repression have only proven that the Pc and ph proteins are part of the band even when the maternal of elongation, at the stage germ The coprecipitation of 10-15 additional proteins complex. is eliminated (Struhl and Akam, 1985; contribution Pc-G in 1 is suggestive evidence that other proteins Figure depicted Pc and ph can be coprecipitated Gelbart, 1990). Jones and associated with these two Pc-G gene proteins. are potentially in including 0-5 h post- embryogenesis, at all stages could be the identity of these other proteins? Obviously What when the homeotic genes are not repressed in fertilization, members of the Pc-G will be good candidates for other it appears that Pc-ph complex mutant embryos. Thus, Pc-G in the same molecular structure. Based on genetic interacting 2947 A.Franke et al. Zink et al. and DeCamillis et al. for homeotic formation is not sufficient gene complex. (1991) (1992) derepression. of Pc and to have shown that for this It There could be several explanations finding. binding may ph polytene DNA in but it chromosomes is be that the complex is active early development, However, sequence-dependent. neither Pc nor to bind DNA in vitro in does not bind to the homeotic genes early embryogenesis, ph protein appears and and because the chromatin of the homeotic genes is (M.Oed R.Paro, H.W.Brock, N.Cheng unpublished perhaps One candidate for the DNA is the Posterior Mutations in Additional sex inaccessible to the complex. results). binding sex combs which contains DNA combs expression of at the ectopic even-skipped (Psc) protein presumptive permit motifs et van Lohuizen et blastoderm et the stage (Sinclair al., 1992), supporting binding (Brunk al., 1991; al., Another is whether the multimeric that a complex containing Pc-G could be genes 1991a). question suggestion complex the binds a limited number of sites at and acts to active in development. Alternatively, complex early may target genes lack or modifications that allow recognition of the formation of chromatin without a structural components regulate being or whether the is a constituent homeotic which in turn implies that the complex could genes, component itself, complex of chromatin that is itself to alter chromatin have different compositions or properties at different directly required structure effects over chromosomal stages. developmental through cooperative long distances. Potential of Pc-G have been found in Pc and have the same distribution ph during homologues genes mammals. the Pc chromo domain as a similar and a completely overlapping embryogenesis binding Using probe, have been identified in mice and humans on polytene chromosomes pattern proteins (Singh DeCamillis et al. (1992) suggested that most ph binding sites et Pearce et the Psc al., 1991; al., 1992). Similarly, protein is A murine was on chromosomes overlapped with Pc binding conserved. related polytene sites, highly protein, bmi-J but that 1O % of the sites appeared to differ. found to increase the incidence of B-cell However, lymphomagenesis when with it is difficult to compare sites on separate overexpressed the polytene together myc protein Lohuizen et It will also be to chromosome preparations that may differ in age, puffing (van al., 1991b). interesting see whether vertebrates use similar amount of squashing, or intensity of the immuno- multimeric stage, complexes to Using double-label immunofluorescence, we now staining. stably repress developmental regulators by regulating order chromatin structure. demonstrate that all of the target sites detectable using these higher are the same for Pc and ph. We analysed glands reagents of different ages and the - 100 sites always overlapped Materials and methods making it unlikely that there are developmental completely, differences of binding specificity of the two proteins in this of Preparation nuclear extracts tissue. Nuclear were proteins prepared by the method of Gay et al. (1988), with some minor modifications. All The completely overlapping pattern of Pc and ph binding steps were performed at 4°C. As starting 3-6 dechorionated material, g wild type embryos from an overnight to chromosomes poses a paradox, because muta- egg polytene lay were suspended in buffer B (15 mM HEPES pH 7.6, 10 mM KCI, tions in each gene have different phenotypes. Three 5 mM 0.1 mM 0.5 mM 1 mM 1 MgCl2, EDTA, EGTA, mM PMSF, cannot be ruled out. First, minor polytene explanations 2 2 DTT, Leupeptin, and 2.5 ,ug/ml Pepstatin Aprotinin; 4 jig/ml ml/g jg/ml binding sites that have not been detected in our procedure embryos). Embryos were homogenized with 10-15 strokes in a Kontes glass homogenizer using a type A pestle. The homogenate differ for each protein and account for the differences was might pre-cleared by passage through two layers of Miracloth (Calbiochem) and subsequently in Secondly, binding of Pc and ph to polytene phenotype. at 2000 for centrifuged 10 min. The g pellet was resuspended in buffer chromosomes might differ from binding to chromosomes (2 ml/g embryos). The solution was overlayed on a cushion of the same of diploid tissues or in other polytene tissues. Thirdly, volume of B buffer plus 0.8 M sucrose in a centrifuge tube and spun in containing Pc and ph may bind in the same swing-out rotor at 1000 g for 10 min. The nuclear complexes pellet was resuspended in B- buffer 150 mM KCI (0.7 ml/g embryos). Nuclei were by lysed the regions at cytological resolution, but have different binding addition of 4M to (NH4)2SO4 a final concentration of 0.4 M and further sites at the DNA level. Notwithstanding the arguments incubation for 15 min on ice. The lysate was cleared by centrifugation of above, the simplest interpretation of our results is that binding 120 000 1 g for h. The soluble nuclear proteins in the supematant were to a particular site on polytene chromosomes might not precipitated by slow addition of (NH4)2SO4 to a final concentration of 30% (w/v), centrifuged at 30 000 g for 15 min and carefully resuspended necessarily mean that the complex is functional. It may be in buffer C mM (25 HEPES pH 100 mM 7.6, KCl, 5 mM MgCI2, 0.1 mM EDTA, that the Pc-ph complexes are always associated with all 1 mM 1 mM PMSF, DTT, 2 Leupeptin, 2 Pepstatin, 2.5 jig/ml 1g/mi their potential target genes and that additional components itg/ml and Aprotinin 10% glycerol; 100 embryos). 50-100 aliquots were ltl/g IL or modifications of the complex determine it gene-specific frozen in liquid nitrogen and stored for later use at -80°C. Protein Constitutive binding is a commonly used mechanism concentration was activity. measured by the Bradford method (Bradford, 1976). for gene regulation. The recruitment of particular accessory Immunoprecipitation and radioactive labelling of nuclear proteins or a specific modification of the regulatory structure proteins seems to be necessary to invoke the transcriptional activity. Protein A-Sepharose CL-4B (PAS: Pharmacia LKB) was equilibrated in Examples for this are the c-fos induction by the serum buffer CBX (buffer C plus 10 mg/ml BSA and 0.1% Triton X-100) for 30 min. Buffer response factor (Herrera et al., 1989; Dalton and Treisman, CBX was added to give a 50% Protein A-Sepharose suspension. Embryonic - nuclear extract ( 25-50 dl) was mixed with 16 vol 1992) in mammalian cells or the phosphorylation of the heat of buffer CBX. An appropriate amount of antiserum was added to the mixture shock factor in yeast (Sorger and Pelham, 1988). In this and incubated by slow rotation for 30 min at 40C. The optimal amount view, Pc or ph mutations have different phenotypes because of antiserum needs determined to be for each antibody individually. Two complexes lacking each protein are differentially susceptible volumes of the Protein A-Sepharose suspension were added and the incubation continued for 30 minat 4°C. to modification or activation. The complexes were precipitated at 2000 g for 1minin an Eppendorf centrifuge. The pellet was washed We show that two Pc-G are constituents of a large genes five times for 1minin buffer C plus 0.1% Triton X-100 and 0.05% SDS multimeric Additional will be protein experiments and three times in buffer C only. Between complex. each wash the complexes were needed to determine what defines the site of precipitated as above. After the last wash the pellet was either the resuspended binding 2948 Pc-G genes form a multimeric complex in 3-4 vol of SDS-PAGE loading buffer for direct analysis or resuspended anti-rabbit-antibodies diluted 1:40. (Jackson) rhodamine-conjugated protein, fluorochromes were reduced Slides in 100 of 0.1M Na-borate buffer pH 8.5 for radioactive labelling of All performed under light. jtl steps involving the incubation with the The the proteins. were washed as after primary antibody. washed in PBS and embedded in Moviol-2.5% DABCO were Immunoprecipitated proteins were labelled using the 35SLR (Amersham preparations Chromosomes were with International). 50 (1mCi/ml, specific activity >800 Ci/mmol) photographed (1,4-diazobicyclo-(2,2,2)-octane). ACi31SLR Zeiss the filters were dried in a speed vac. The resuspended immunoprecipitate was added, an Axiophot using appropriate equipped epifluorescence filter number 48 79 filter number 48 79 09) mixed and incubated for 30 minon ice. After the reaction unincorporated 15; DTAF, using (rhodamine, 35SLR was neutralized by the addition of an equal volume of 0.2 M KodakT-Max 400. glycine in Na-borate buffer for Smin. SDS and ,B-mercaptoethanol were added to a final concentration of 2 and 5%, respectively and the protein-Sepharose complex dissociated for 5 minat 60°C. Labelled proteins were separated Acknowledgements from the 35SLR-glycine filtration over a small column by gel (Bio-Gel We would like to thank H.P.Blaschkowski for advice in P 10; Biorad). The pooled protein fractions were concentrated by precipitation velocity gradient ultra and filtration R.Paro would like to thank with 2 vol of ethanol for 30 minon ice and a centrifugation for 15 min gel analysis. centrifugation Solf for technical assistance and Annette Schwarz for the at 30 000 g. The protein pellet was resuspended in SDS-PAGE loading Christine typing The work of R.P. was from the buffer. by supported grants manuscript. Bundesministerium fur und and the (0316001A) by Forschung Technologie Deutsche The work of H.W.B. was Gel filtration chromatography and sucrose gradient supported Forschungsgemeinschaft. a from the National Science and Research and Medical ultracentrifugation by Engineering grant Research Councils. For gel filtration we used two media with different separation capacities. Sephacryl HR (Pharmacia) molecules in the between S-300 separates range 10- 1500 kDa, Sephacryl HR in the between 20-1 x 104 S-400 range kDa. The material was suspended in running buffer (25 mM HEPES pH 7.6, References 100mMKCI, 5 mMMgC12, 0.1 mMEDTA and 10% glycerol) to reach and Brunk.B. Dev. Genet., 249-260. 10, a final ratio of 2/3 Sephacryl and 1/3 buffer. The suspension was loaded (1989) Adler,P.N., Charlton,J. and Anal. (1980) into a glass column that was 90 cm long and 2 cm wide. The column was Rubenstein,A.H. Assoian,R.K., Blix,P.M., Tager,H.S. 70-76. calibrated separately using proteins of known molecular weight. For the Biochem., 103, Anal. 248-254. Biochem., 72, Sephacryl S-300 HR column: cytochrome C (13.3 kDa, Boehringer Bradford,M.M. (1976) Dev. 442-456. and 118, Mannheim), BSA (66 kDa, Merck), catalase (240 kDa, Boehringer Duncan,I.M. (1986) Breen,T.R. Biol., and Nature, 353, 351-353. Mannheim) and for the Sephacryl 400-HR; catalase (240 kDa, Boehringer Adler,P.N. (1991) Brunk,B.P., Martin,E.C. 104, 713-720. and Mannheim) and mouse proteasome (750 kDa, Rivett,A.J., 1985, kindly Morata,G. (1988) Development, Busturia,A. and EMBO J., 9, 4267 -4275. Garcia-Bellido,A. (1990) provided by M.Gernold and P.Kloetzel). The exclusion volume of the Castelli-Gair,J.E. and EMBO J., 9, Sephacryl 300-HR column was determined with Blue dextran 2000 Keelan,D.J. Lewis,E.B. (1990) Celniker,S.E., 4277-4286. (Pharmacia). 200 of embryonic nuclear extracts were mixed with 1.5 ml Al 597-612. and Cell, 68, running buffer and loaded on the column. 5 ml fractions were collected. Treisman,R. (1992) Dalton,S. and Gene, 105, Brock,H.W. (1991) 2 of BSA were added as carrier and the fractions mixed with an equal Randsholt,N.B. ytg Deatrick,J., Daly,M., 185- 195. volume of 20% TCA-50% acetone. The were by proteins precipitated and Genes Dev., Brock,H.W. (1992) incubation on ice for 30 at 10 000 for 10 min and Pierre,D. min, centrifuged r.p.m. DeCamillis,M., Cheng,N., 223-232. resuspended in SDS-PAGE buffer. The were neutralized loading samples 6, Dev. 97, 34-47. and by of and stored at For the sucrose Frederick,R.D. (1983) the addition Na2CO3 crystals -20°C. Denell,R.E. Biol., 49-70. Genetics, 102, gradient centrifugation a SW 60 Ti rotor (Beckman Instruments) was used. (1982) Duncan,I.M. In and Green,P.B. (eds), and A 5-20% sucrose gradient in buffer C (25 mM HEPES pH 7.6, 100mM Lewis,E.B. (1982) Subtelny,S. Duncan,I.M. and Liss, New York, Order: Its pp. 5 mM MgCl2, 0.1 mM EDTA, 1 mM PMSF, 1 mM DTT, 2 Origin Regulation. KCI, ytg/ml Developmental 533-554. Leupeptin, 2 Pepstatin and 2.5 1tg/ml Aprotinin) was prepared in tLg/ml Mol. Gen. Genet., 198, and Santamaria,P. (1985) a 4 ml centrifuge tube. 40 of embryonic nuclear extracts were mixed Dura,J.-M., Brock,H.W. Al 213-220. with 160 buffer C and loaded on top of the gradient. Centrifugation was Al 103, 733-741. and Dura,J.-M. (1988) Development, at 50 000 r.p.m. for 12 h at 40C. 20 x 200 1l fractions were collected Ingham,P. Santamaria,P., Deatrick,J., Erk,I., with an ISCO Model 640. Proteins were precipitated with 2 vol ethanol Dura,J.M., Randsholt,N.B., and Brock,H.W. (1987) Cell, Weddell,D. Freeman,S.J., and resuspended in SDS-PAGE loading buffer. The gradient was calibrated Freeman,J.D., 829-839. 51, in separate tubes using different enzymes of known molecular weight (all 11, 14-17. (1989) Bioessays, obtained from Boehringer Mannheim): horse radish peroxidase (44 kDa, Eissenberg,J.C. Foster-Hartnett,D.M., Hartnett,T., Ngan,V. 1 mg/gradient), lactate kDa, 0.5 fumarase James,T.C., dehydrogenase (140 mg/gradient), Eissenberg,J.C., Acad. Sci. USA, 87, 9923-9927. Proc. and (194 kDa, 0.2 mg/gradient), catalase (240 kDa, 0.4 mg/gradient). The (1990) Natl. Elgin,S. of PhD Germany. thesis, Heidelberg, of the in the different fractions was determined by the University localization proteins Franke,A. (1991) EMBO J., 10, 1225-1236. reaction. corresponding enzyme Frasch,M. (1991) Trends Genet., 6, 208-212. and (1990) Gaunt,S.J. Singh,P.B. EMBO J., 7, 4291-4297. and (1988) Poole,S. Kornberg,T. Immunolocalization of proteins on embryos and on polytene Gay,N.J., Trends Genet., 6, 395-400. Grunstein,M. (1990) chromosomes spreads Trends Genet., 6, 422-426. Henikoff,S. (1990) Immunostaining of embryos was based on the procedure described by Nature, 340, 68-70. and Nordheim,A. (1989) Shaw,P.E. Herrera,R.E., MacDonald and Struhl (1986). Fixation and spreading of the chromosomes 815-823. Cell, 37, (1984) essentially followed the protocol of Zink and Paro (1989) and Zink et al. Ingham,P.W. Genetics, 126, 185-199. and Gelbart,W.M. (1990) Jones,R.S. (1991). The chromosome immunostaining protocol was modified slightly Nature, 316, 153-155. (1985) in order to reduce background staining. After washing slides in PBS, they Jurgens,G. and Genetics, 115-133. Wakimoto,B. 94, (1980) Kaufman,T.C., Lewis,R. were transferred to blocking solution (10% BSA, 10% dry milk, 0.2% NP-40 Genetics, 116, 75-86. and Russell,M.A. (1987) Kennison,J.A. and 0.2% Tween 20-80 in PBS) and incubated for 1 h at 370C. Although and (1989) Nucleic Bennani,N. Pradel,J. Garzino,V., Mary,C., Krejci,E., the blocking solution was cloudy, it did not appear to interfere with very 8101-8115. Acids Res., 17, the immunostaining reactions. Slides were briefly washed in PBS and then EMBO J., 7 3233-3244. and McGinnis,W. (1988) Kuziora,M.A. the chromosomes were incubated with the primary antibody at the appropriate (1978) Nature, 276, 565-570. Lewis,E.B. concentration in blocking solution for 1 h at 370C. We used the purified and 181-198. Tartof,K.D. (1988) Genetics, 120, Kotarski,M.A. Locke,J., anti-Pc and rabbit anti-ph antibodies at a 1:100 mouse polyclonal polyclonal Nature, 324, 537-545. and (1986) Struhl,G. MacDonald,P.M. dilution. Slides were briefly washed in PBS and then vigorously shaken and Genes Dev., in press. Paro,R. (1992) Franke,A. Messmer,S., for 15 min in 300 mM NaCl, 0.2% NP-40 and 0.2% Tween 20-80 in PBS, and Brock,H.W. (1991) Rouxs Arch. Dev. BioL. 199, 387-396. McKeonj. and for additional 15 min in 400 mM NaCl, 0.2% NP-40 and 0.2% Tween Muller,J. and Bienz,M. (1991) EMBO J., 10, 3147-3155. in PBS. Slides were briefly washed in PBS and blocked in 20-80 again Ohtsubo,M., Yoshida,T., Seino,H., Nishitani,H., Clark,K.L., Sprague,G., blocking solution as above. The incubation with the secondary antiy Frasch,M. and Nishmoto,T. (1991) EMBO J., 10, 1265-1273. for 1 h at in solution. For the Pc protein, we used DTAF- was 37°C blocking Paro,R. (1990) Trends Genet., 6, 416-421. anti-mouse antibodies (Jackson) diluted 1:100 and for the ph conjugated 2949 A.Franke et al. and Hogness,D.S. (1991) Proc. Natl. Acad. Sci. USA, 88, Paro,R. 263-267. Singh,P.B. and Gaunt,S.J. (1992) Development, 114, Pearce,J.J.H., 921-930. Reuter,G., Giarre,M., Farah,J., Gausz,J., Spierer,A. and Spierer,P. (1990) Nature, 344, 219-223. (1985) J. Bio. Chem., 260, 12600-12606. Rivett,A.J. and Denell,R.E. (1985) Dev. Biol., 110, 53-63. Sato,T. Simcox,A.A., Hersperger,E., Shearn,E., Whittle,J.R.S. and Cohen,S.M. 34, 11-20. (1991) Mech. Dev., and (1992) Development, 114, in press. Simon,J., Chiang,A. Bender,W. Niichools,F., Slade,E. and Brock,H.W. Sinclair,D.A., Campbell,R.B., 130, 817-825. (1992) Genetics, Miller,J.R., Pearce,J., Kothary,R., Burton,R.D., Paro,R., Singh,P.B., and Gaunt,S. (1991) Nucleic Acids Res., 19, 789-794. James,T.C. Smouse,D. and Perrimon,N. (1990) Dev. Biol., 139, 169-185. Smouse,D., Goodman,C., Mahowald,A. and Perrimon,N. (1988) Genes 830-842. Dev., 2, (1988) Cell, 54, 855-864. Sorger,P.K. and Pelham,H.R.B. Nature, 293, 36-41. Struhl,G. (1981) Struhl,G. and Akam,M. (1985) EMBO J., 4, 3259-3264. Tartof,K.D. and Bremer,M. (1990) Development, Suppl., 35-45. Tiong,S.Y.K. and Russell,M.A. (1990) Dev. Biol., 141, 306-318. van Lohuizen,M., Frasch,M., Wientjens,E. and Berns,A. (1991a) Nature, 353-355. 353, van Lohuizen,M., Verbeek,S., Scheijen,B.W., van der Gulden,H. and Cell, 65, 737-752. Berns,A. (1991b) Wedeen,C., Harding,K. and Levine,M. (1986) Cell, 44, 739-748. Wu,C.-T., Jones,R.S., Lasko,P.F. and Gelbart,W.M. (1989) Mol. Gen. Genet., 218, 499-564. Zink,B. (1990) PhD thesis, University of Heidelberg, Germany. Zink,B. and Paro,R. (1989) Nature, 337, 468-471. Zink,B., Engstrom,Y., Gehring,W.J. and Paro,R. (1991) EMBO J., 10, 153-162. Received on April 1, 1992; revised on May 11, 1992

Journal

The EMBO Journal – Springer Journals

Published: Aug 1, 1992

There are no references for this article.