Nuclear gems and cajal (coiled) bodies in fetal tissues nucleolar distribution of the spinal muscular atrophy protein, SMN

Experimental Cell Research 265, 252–261 (2001) doi:10.1006/excr.2001.5186, available online at http://www.idealibrary.com on Nuclear Gems and Cajal (Coiled) Bodies in Fetal Tissues: Nucleolar Distribution of the Spinal Muscular Atrophy Protein, SMN P J Young, T T Le,* M Dunckley,† Nguyen thi Man, A H M Burghes,* and G E Morris MRIC Biochemistry Group, North East Wales Institute, Mold Road, Wrexham LL11 2AW, United Kingdom; *Department of Medical Biochemistry, Ohio State University, Columbus, Ohio 43210; and †Neuromuscular Unit, Imperial College School of Medicine, Hammersmith Hospital, London, United Kingdom SMN, the affected protein in spinal muscular atrophy (SMA), is a cytoplasmic protein that also occurs in nuclear structures called “gems” and is involved in snRNP maturation Coilin-p80 is a marker protein for nuclear Cajal bodies (coiled bodies; CBs) which are also involved in snRNP maturation, storage or transport We now show that gems and CBs are present in all fetal tissues, even those that lack gems/CBs in the adult Most gems and CBs occur as separate nuclear structures in fetal tissues, but their colocalization increases with fetal age and is almost complete in the adult In adult tissues, up to half of all gems/CBs are inside the nucleolus, whereas in cultured cells they are almost exclusively nucleoplasmic The nucleolar SMN is often more diffusely distributed, compared with nucleoplasmic gems Up to 30% of cells in fetal tissues have SMN distributed throughout the nucleolus, instead of forming gems in the nucleoplasm The results suggest a function for gems distinct from Cajal bodies in fetal nuclei and a nucleolar function for SMN Spinal cord, the affected tissue in SMA, behaves differently in several respects In both fetal and adult motor neurons, many gems/CBs occur as larger bodies closely associated with the nucleolar perimeter Uniquely in motor neurons, gems/CBs are more numerous in adult than in fetal stages and colocalization of gems and CBs occurs earlier in development These unusual features of motor neurons may relate to their special sensitivity to reduced SMN levels in SMA patients © 2001 Academic Press INTRODUCTION Childhood spinal muscular atrophy (SMA) is an autosomal recessive disorder characterized by the loss of alpha motor neurons of the spinal cord [1] The disease determining protein, SMN (survival of motor neuron), is a 40-kDa ubiquitously expressed protein [2, 3] There are two inverted copies of the SMN gene on chromo1 To whom correspondence should be addressed Fax: 44-1978290008 E-mail: Morrisge@newi.ac.uk 0014-4827/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved some 5q13, SMN1 and SMN2 [3] The two genes differ by only 11 nucleotides, none of which alter the coding sequence [3, 4] However, one of these differences alters a putative exon splicing enhancer in exon 7, ensuring that the main product of SMN2 is a truncated transcript lacking exon [4, 5] The severity of the disease is directly correlated to the levels of functional full-length SMN produced from the SMN2 gene [2, 6] Deletions of the SMN1 gene produce the disease phenotype, but increased copy number of the SMN2 gene can modify the disease severity [3, 7, 8] SMN localizes in the cytoplasm and in subnuclear structures termed gemini of coiled bodies, or “gems” [9] SMN has been shown to play an important role in the cytoplasmic assembly of spliceosomal U snRNPs [10 –12] Pellizzoni et al [13] have suggested that nuclear SMN may be involved in recycling/regeneration of U snRNPs and Meister et al [14] have proposed that SMN may play an active role in pre-mRNA splicing by altering the Sm core protein components of the U snRNPs However, the nuclear role of SMN may not be restricted to splicing with several other possible functions suggested, including regulation of gene transcription [15–17] and nucleocytoplasmic transport [18] The close relationship between nuclear gems and Cajal bodies suggests that they may have related functions Cajal bodies contain snRNA specific transcription factors, PTF and TATA-box binding protein (TBP), and have been shown to accumulate around the U1 and U2 gene clusters on chromosomes and 17, respectively, in interphase cells, leading to the suggestion that they mediate U1 and U2 transcription [19] Inhibition of dephosphorylation using okadaic acid results in the relocation of Cajal bodies to the nucleolus [20] The identification of low levels of coilin-p80 and U snRNPs in cultured HeLa cell nucleoli under normal conditions suggests that U snRNPs may undergo a maturation process within the nucleolus [20 –22], while the phosphorylation state of coilin-p80 may direct U snRNPs to and from the nucleolus [20] Thus, Cajal bodies may be involved in the production, maturation, and storage of U snRNPs [20 –24] Although 252 NUCLEAR GEMS AND CAJAL BODIES IN FETAL TISSUES gems were first described as adjacent to Cajal bodies [9], more recent studies suggested that they may be the same structure [23, 25] and our own studies of SMN and coilin-p80 colocalization in adult tissues could be interpreted as supporting this view [26] By studying fetal tissues, however, we now show that gems and Cajal bodies are predominantly separate early in development and colocalization increases with developmental age Although SMN and coilin-p80 are ubiquitously expressed, gems and Cajal bodies are absent from some cell types, including smooth and cardiac muscle, endothelial cells, dermal and epidermal cells, myoepithelial cells, stomach parietal and peptic cells, and spleen parenchymal cells [26] Total SMN levels are high in cardiac muscle, so we suggested that, although cytoplasmic SMN levels may influence gem formation, additional cell-specific factors may be required for the formation of gems and Cajal bodies [26] The absence of both structures from several cell types suggests that, although both SMN and coilin-p80 are ubiquitous proteins, visible Cajal bodies and gems are not essential for viability Previous studies have found a drop in total SMN levels between fetal and adult tissues [27] In the present study, we have addressed the question of whether the elevated SMN levels in fetal tissues affect nuclear gem and Cajal body formation and distribution We also demonstrate a novel nucleolar localization of SMN and report several unique characteristics of nuclear gems in motor neruons of the spinal cord that may aid the understanding of the disease pathology EXPERIMENTAL PROCEDURES Immunohistochemistry Human fetal (10 –18 weeks) and adult pig (2 months) tissues were embedded in tissue freezing medium and frozen in isopentane cooled in liquid nitrogen Cryostat sections (5 ␮m) were mounted on microscope slides and fixed for with 50% methanol/50% acetone followed by a PBS wash Previously reported anti-SMN mouse monoclonal antibodies (MANSMA1, -3, and -5 [26]), an anti-p80 rabbit polyclonal (Ab204 [28]), and an anti-fibrillarin human auto-anti-serum [29] were used during this study Mabs were diluted ϩ with PBS (ca ␮g/ml final concn) and protein localization was revealed using FITC-conjugated horse anti-(mouse Ig) Nuclear localization was confirmed using ethidium bromide or DAPI as counterstain For double labeling experiments, fixed sections were incubated with the mouse mAb for h, followed by the rabbit polyclonal antibody (diluted in 250 PBS) for h The mAb was detected with a FITC-conjugated goat anti-(mouse Ig) (Sigma: diluted in 80 PBS) and the polyclonal antibody with a TRITC-conjugated goat anti-(rabbit Ig) (Sigma: diluted in 128 PBS) For triple labeling experiments, sections were exposed to mouse mAb and rabbit polyclonal antibodies as stated above, followed by the human auto-antiserum against fibrillarin (diluted in 1000) for h The human anti-serum was detected using an AMCA-conjugated goat anti-(human Ig) (Chemicon: diluted in 200 PBS) TRITC and FITC (L4) filter sets and a 63X oil immersion objective were used with the Leica DMRB photomicroscope The TRITC filter was also used for ethidium bromide and a UV filter for AMCA and DAPI Images were 253 captured under standard and equivalent conditions using an integrating video camera (JVC) and image grabber (Neotech) RESULTS Distribution of SMN, SIP-1, and Coilin-p80 in Human Fetal Tissues Although SMN and coilin-p80 are ubiquitously expressed, we have shown previously that neither gems nor Cajal bodies are ubiquitous nuclear structures [26], since both are absent from some adult cell types, including smooth and cardiac muscle, endothelial cells, dermal and epidermal cells, myoepithelial cells, stomach parietal and peptic cells, and spleen parenchymal cells [26] When Cajal bodies and gems were present in adult tissues, SMN was always found to colocalize with coilin-p80 [26] Most studies on adult tissues were performed with pig or rabbit tissues because of restricted availability of adult human tissues However, human skeletal muscle biopsy sections gave similar quantitative results to pig muscle and the tissues that lacked Cajal bodies and gems were the same in both pig and rabbit [26] Total SMN levels are higher in fetal tissues than in adult tissues [26, 27] and Burlet et al [27] found nuclear gems in all fetal tissues they examined, including brain, heart, kidney, skeletal muscle, and thymus Studies on cultured skin fibroblasts from SMA patients suggest that nuclear gem formation may be inhibited (or decreased) by reduced SMN levels [2, 26] To determine whether the elevated SMN levels in fetal tissues are associated with an increase in nuclear gem and Cajal body formation, quantitative counts of nuclear gems and Cajal bodies were performed on human fetal tissues using anti-SMN mAbs, MANSMA1 (exon specific) and MANSMA3 (exon specific), and an anti-p80 mAb, 5P10 (Table 1) Nuclear Cajal bodies and gems were present in all tissues tested at all fetal ages (Table 1), including heart, skin, and spleen (Fig 1), which were negative for both structures in adult tissues [26] For all fetal tissues tested, there was an increase in the average number of gems per nucleus with fetal age (Table 1) Cajal bodies were also present in all fetal tissues and were more numerous than in adult tissues, except in motor neurons (Table 1) In general, nuclear gems were more abundant in the CNS than in non-CNS tissues, except that very high gem counts were found in fetal skin (Table 1) This is surprising because nuclear gems are absent from all cell types of adult skin (Table 1; Fig in Ref [13]) The number of nuclear gems was higher in motor neurons of adult spinal cord than in fetal spinal cord (Table 1) This unusual developmental expression pattern may signify a special function of nuclear SMN in motor neurons In contrast to nuclear 254 YOUNG ET AL TABLE Quantitative Nuclear Gem and Cajal Body Counts for Fetal Human Tissues Nuclear gems CNS Brain (crebral cortex—molecular neurons) 12.8 weeks 13.5 weeks Adult rabbit Spinal cord (anterior horn—motor neurons) 10.0 weeks 14.8 weeks Adult rabbbit Non-CNS Skeletal muscle (muscle nuclei) 10.3 weeks 13.3 weeks Adult rabbit Heart (myocardial cells) 11.0 weeks 12.8 weeks Adult rabbit Liver (parenchymal cells) 12.0 weeks 13.3 weeks Adult rabbit Lung (myoepithelial cells) 11.8 weeks 14.8 weeks Adult rabbit Skin (epidermis) 11.8 weeks 14.8 weeks Adult rabbit Spleen (parenchymal cells) 13.0 weeks 18.2 weeks Adult rabbit Cajal bodies MANSMA1 MANSMA3 MANSIP1A 5P10 0.87 (0–3, 167) 1.01 (0–4, 200) 1.12 (1–3, 145) 0.79 (0–2, 100) 0.96 (0–2, 100) 1.01 (0–3, 137) 0.79 (0–2, 150) 0.92 (0–2, 217) 1.03 (0–3, 100) 1.71 (0–6, 200) 1.64 (0–6, 209) 1.14 (1–3, 150) 0.85 (0–2, 130) 1.15 (0–2, 175) 2.07 (1–4, 54) 0.96 (0–2, 175) 1.21 (0–2, 100) 2.25 (0–6, 45) 0.93 (0–2, 156) 1.26 (0–2, 134) 2.03 (0–6, 57) 1.18 (0–3, 150) 1.21 (0–3, 160) 2.15 (1–4, 61) 0.48 (0–2, 150) 0.89 (0–2, 140) 0.44 (0–2, 71) 0.45 (0–2, 100) 0.96 (0–2, 154) 0.41 (0–2, 57) 0.47 (0–1, 76) 0.86 (0–2, 132) 0.39 (0–2, 59) 1.00 (0–2, 160) 1.30 (0–3, 130) 0.46 (0–2, 50) 0.39 (0–2, 201) 0.43 (0–2, 167) (200 nuclei) 0.34 (0–2, 115) 0.39 (0–2, 95) (150 nuclei) 0.29 (0–1, 89) 0.41 (0–2, 145) (276 nuclei) 3.56 (1–8, 250) 2.51 (1–4, 150) (145 nuclei) 0.31 (0–2, 125) 0.77 (0–2, 200) 0.89 (0–2, 75) 0.42 (0–2, 150) 0.82 (0–3, 100) 0.95 (0–3, 60) 0.37 (0–1, 90) 0.76 (0–2, 145) 0.97 (0–2, 58) 1.54 (0–3, 150) 1.98 (0–3, 170) 0.85 (0–2, 121) 0.45 (0–2, 200) 0.81 (0–3, 140) (200 nuclei) 0.52 (0–2, 215) 0.87 (0–2, 200) (150 nuclei) 0.47 (0–2, 175) 0.79 (0–2, 186) (150 nuclei) 2.70 (0–6, 150) 1.03 (0–3, 117) (210 nuclei) 1.31 (0–2, 165) 1.77 (0–3, 115) (200 nuclei) 1.18 (0–3, 100) 1.59 (0–3, 150) (200 nuclei) 1.24 (0–3, 135) 1.62 (0–3, 164) (100 nuclei) 2.93 (0–7, 217) 2.01 (0–3, 126) (200 nuclei) 0.25 (0–1, 100) 0.65 (0–2, 85) (150 nuclei) 0.31 (0–1, 150) 0.59 (0–1, 120) (200 nuclei) 0.34 (0–1, 160) 0.64 (0–2, 154) (215 nuclei) 0.56 (0–1, 125) 1.04 (0–2, 67) (200 nuclei) Note Data from adult rabbit tissues are shown for comparison Nuclear gems were recognized using anti-SMN monoclonal antibodies MANSMA1 (exon specific) and MANSMA3 (exon specific) and anti-SIP1A mAb MANSIP1A Cajal bodies were identified by the anti-coilin-p80 mAb 5P10 The average number of gems per section is followed, in parentheses, by the range and the number of nuclear bodies counted (when nuclear bodies were absent, the number of nuclei counted is shown) gems, the highest levels of Cajal bodies were found outside the central nervous system, in heart, lung, and skin, none of which contain Cajal bodies in adult tissues (Table 1) To determine the exact relationship between nuclear SMN and coilin-p80 in fetal tissues, double labeling experiments were performed using MANSMA1 and Ab204 [28] Figure shows that in fetal cerebral cortex, Cajal bodies and nuclear gems occurred as both independent and colocalized structures, an expression pattern more akin to primary skin fibroblast cultures [26] than adult mammalian tissues In all fetal tissues, colocalization of SMN and coilin-p80 within nuclear bodies increased with increasing fetal age (Table 2) Motor neurons of the spinal cord displayed the highest degree of colocalization, with 50% of nuclear bodies at the age of 14.8 weeks containing both SMN and coilinp80 (Table 2) SMN and coilin-p80 in fetal motor neurons were often localized in larger bodies around the nucleolus (Fig 3B) as previously described in adult pig motor neurons [18, 26, 30, 31] In non-CNS tissues, although the number of independent gems increased with age, the majority of gems were independent of Cajal bodies at all ages, with fewer than 25% of identified bodies containing both proteins (Table 2) Previous reports have identified cytoplasmic dotstructures containing SMN in fetal skeletal muscle [27] As well as in skeletal muscle (10.3 and 13.3 weeks), we observed similar structures in fetal brain (12.8 and 13.5 weeks), lung (11.8 and 14.8 weeks), liver (12.0 and 13.3 weeks), and spleen (13.0 and 18.2 weeks) These cytoplasmic bodies were not seen in the NUCLEAR GEMS AND CAJAL BODIES IN FETAL TISSUES 255 tion with another nucleolar protein (results not shown) Fibrillarin was used as a nucleolar marker and triple labeling was performed on fetal human and adult pig tissues, using the mouse mAb MANSMA1, anti-coilinp80 rabbit sera 204 Ab, and human anti-fibrillarin sera (Fig 3) In fetal liver (13.0 weeks), skeletal muscle (13.3 weeks, Fig 3A), spinal cord (14.8 weeks, Fig 3B), skin (11.8 weeks), and spleen (18.2 weeks), fibrillarin was predominantly localized in the nucleolus (Table 3) In all tissues, the majority of nucleoli stained only for fibrillarin (66 –70%), but SMN colocalized with fibrillarin in 24 –31% of the nucleoli (Table 3, Fig 3A) In all tissues except spinal cord, those cells with SMN throughout the nucleolus did not express nuclear gems (Fig 3A) and vice versa (results not shown) Motor neurons of the spinal cord were the only cell type that sometimes displayed both nucleoplasmic gems and SMN colocalized with fibrillarin (Fig 3B) Cajal bodies, with or without SMN, were observed in less than 3% of nucleoli in fetal tissues (Table 3), but some coilin-p80 in diffuse form was detectable in the SMN-positive nucleoli of fetal tissues (Figs 3A and 3B) Gems and Cajal Bodies in the Nucleolus in Adult Tissues FIG Distribution of SMN and coilin-p80 in fetal human and adult pig tissues Visible gems and Cajal bodies were not detected in adult heart, spleen, and skin, but were detected in the corresponding fetal tissues SMN was identified using MANSMA1 and a FITCconjugated horse anti-(mouse Ig) Coilin-p80 was identified using Ab204 and a FITC-conjugated goat anti-(rabbit Ig) Nuclei were counterstained with ethidium bromide The bar represents 30 ␮m corresponding adult tissues, suggesting a real difference between fetal and adult tissues SMN and SIP-1 colocalized in both nuclear and cytoplasmic bodies in fetal cells from all ages (results not shown) SMN Colocalizes with Fibrillarin in the Nucleolus in Fetal Tissues SMN was also observed in nucleoli in all fetal tissues examined The nucleolar localization was obtained using three monoclonal antibodies against different SMN epitopes (MANSMA1, -3, and -4) and anti-SMN rabbit serum, thus ruling out the possibility of a cross-reac- Double and triple labeling with the fibrillarin autoantibody revealed that nuclear gems and Cajal bodies in adult tissues were commonly found as discrete structures within nucleoli, as well as in the nucleoplasm (Fig 3A and Table 4) Fibrillarin staining was restricted to the nucleolus in adult tissues (Figs 3A and 3B) This was confirmed by viewing labeled cells under phase contrast (results not shown) Colocalization of SMN and coilin-p80 was not mediated by fibrillarin, since most gems/CBs containing both proteins were outside the nucleolus and were not stained by the anti-fibrillarin antibody (Table 4; Fig 3B) Those adult pig tissues that lack both Cajal bodies and gems also showed no nucleolar staining with either MANSMA1 or 204 Ab (results not shown) As previously reported [26], the majority of Cajal bodies in adult pig tissues also contain SMN (Figs and 3, Table 4) This increase in colocalization appears to be linked with an increase in nucleolar association of Cajal bodies, with 38 – 46% of Cajal bodies in pig cerebral cortex, skeletal muscle (Figs 3A and 3C), and liver (Fig 3C) localizing within areas of intense fibrillarin staining (Table 4) The possibility that the CBs are lying on top of, rather than inside, the nucleoli can only be ruled out completely by 3D reconstruction experiments In motor neurons, however, of 48 Cajal bodies closely associated with the nucleolus, only was within the nucleolar boundary (Fig 3B, Table 4) In adult 256 YOUNG ET AL FIG Distribution of SMN and coilin-p80 by double labeling on fetal human and adult pig cerebral cortex SMN (green) was identified using MANSMA1 and a FITC-conjugated horse anti-(mouse Ig) Coilin-p80 (red) was identified using Ab204 and a TRITC-conjugated goat anti-(rabbit Ig) Nuclei were counterstained with DAPI (blue) Independent nuclear gems (white arrow), independent Cajal bodies (yellow arrows), and colocalized SMN and coilin-p80 (green arrows) are indicated The bar represents 30 ␮m tissues, SMN distribution was sometimes intermediate between the diffuse nucleolar staining pattern characteristic of fetal tissues (Figs 3A–3C) and the concentration into sharply defined nucleoplasmic gems typical of cultured cell lines This may best be seen in adult liver parenchymal cells which often contain large patches of SMN staining within larger, fibrillarin-positive nucleoli (Fig 3C) Within these patches of nucleolar SMN staining, coilin-p80 staining is usually more concentrated in structures of similar size to nucleoplasmic Cajal bodies (Fig 3C) These were counted as “p80/ SMN” bodies in Table DISCUSSION A number of novel observations have emerged from the present study First, the association between gems and CBs is developmentally regulated, since they occur mainly as separate structures in fetal tissues but are almost invariably colocalized in adult tissues Second, the strong SMN staining throughout the nucleolus in fetal tissues has not been reported previously, although nucleolar staining by anti-SMN antisera has been reported in the adult central nervous system [31] Third, we found that, in most adult tissues examined, colocalized gems/CBs occur very frequently inside nucleoli, as well as in the nucleoplasm Finally, spinal cord motor neurons were found to differ from other cell types in several ways that may be related to the specific pathogenic effects on this tissue when SMN levels are reduced in SMA patients It was known from earlier studies [26, 27] that SMN levels are higher in fetal tissues compared with adult tissues and our quantitative studies have now con- firmed that this is reflected in higher numbers of gems and CBs in the nucleus (Table 1) Lafarga et al [32] also showed that CBs in cerebellar Purkinje cell nuclei decreased from 1.33 in newborn rats to 0.47 in adults Evidence that CB numbers are higher in dividing cells and metabolically active cells has been available for some time [33–38] and the higher CB numbers in fetal tissues are consistent with this In an earlier study of adult tissues [26], however, we noted that many tissues had no visible gems or CBs and these were not always tissues perceived as nondividing or “inactive” (e.g., epidermis and spleen) This suggested that additional tissue-specific factors might be needed for gem/CB formation SMN and coilin-p80 are always present in some form in all tissues, but this is not established for all the other known components of gems and CBs Cytoplasmic dot-like structures containing SMN have previously been reported in fetal skeletal muscle [27] In addition to fetal skeletal muscle, we have seen similar structures in fetal brain, lung, spleen, and liver, but not in the corresponding adult pig tissues or human skeletal muscle (results not shown) This signifies a real difference between adult and fetal tissues with regard to the cytoplasmic accumulation of SMN We have shown that gems and CBs occur mainly as separate structures in fetal cells but show increasing colocalization during fetal development (Table 2), although only a limited period of 10 –18 weeks gestation could be studied with available tissues One possible explanation is that a linker protein is up-regulated during development Fibrillarin was considered as a candidate, since it interacts with SMN in the yeast two-hybrid system [9] and has been detected in Cajal NUCLEAR GEMS AND CAJAL BODIES IN FETAL TISSUES 257 258 YOUNG ET AL TABLE Quantitative Counts of SMN and Coilin-p80 Colocalization in Nuclear Bodies in Fetal Human and Adult Pig Tissues Tissue/age (weeks) Brain (cerebral cortex—molecular neurons) 12.8 weeks 13.5 weeks Adult Spinal cord (anterior horn—motor neurons) 10.0 weeks 14.8 weeks Adult Skeletal muscle (muscle nuclei) 10.3 weeks 13.3 weeks Adult Adult human Heart (myocardial cells) 11.0 weeks 12.8 weeks Adult Lung (myoepithelial cells) 11.8 weeks 14.8 weeks Adult Liver (parenchymal cells) 12.0 weeks 13.0 weeks Adult Skin 11.8 weeks 16.1 weeks Adult Nuclear bodies/ total cells p80 SMN (%) p80 (%) SMN (%) 350/152 411/209 233/208 67 (19.1) 159 (38.6) 217 (93.1) 207 (59.1) 201 (48.9) 16 (6.9) 76 (21.8) 51 (12.5) (0) 178/129 172/156 212/94 71 (39.8) 86 (50.0) 212 (100) 74 (41.5) 62 (36.1) (0) 33 (18.7) 24 (13.9) (0) 163/140 232/122 68/146 51/117 19 (11.6) 39 (16.8) 62 (91.2) 48 (92.3) 106 (65.0) 118 (50.8) (8.8) (7.7) 38 (23.4) 75 (32.4) (0) (0) 277/102 250/95 0/250 12 (4.3) 52 (20.8) n/a 241 (87.0) 120 (48.0) n/a 24 (8.7) 78 (31.2) n/a 579/204 187/118 0/300 27 (4.7) 19 (10.2) n/a 497 (85.8) 100 (53.4) n/a 55 (9.5) 68 (36.4) n/a 288/155 316/128 137/144 53 (18.4) 79 (25.0) 129 (94.2) 208 (72.2) 149 (47.1) (5.8) 27 (9.4) 88 (27.9) (0) 295/80 76/24 0/150 34 (11.5) 16 (21.1) n/a 194 (65.7) 32 (42.1) n/a 67 (22.8) 28 (36.8) n/a Note Counts were also performed on adult human skeletal muscle to determine whether the adult human and pig counts were comparable SMN was identified with MANSMA1 and coilin-p80 was identified with Ab204 n/a, not applicable bodies [9], but this was ruled out by its absence from many colocalized gems/CBs in our study (Fig 3B) Another possibility is that some posttranslational modification (e.g., phosphorylation) of a component of gems or CBs enables their colocalization Colocalization of fibrillarin with SMN was only observed when SMN (but not coilin-p80) occurred throughout the nucleolus in 25–30% of most fetal cells studied, instead of forming nucleoplasmic gems (Table 3) It is not yet clear whether these two different distributions of SMN cor- respond to different cell types or different metabolic states, but the cells with nucleolar SMN staining did not appear to form any recognizable structure in the fetal tissues Although SMN dispersed throughout nucleoli was not observed in adult tissues as a specific staining pattern, SMN was present throughout the cytoplasm and nucleus (including nucleoli) in motor neurons (Fig 3B), so our data are not inconsistent with other reports of SMN in adult motor neuron nucleoli [31] A high proportion of gems/CBs in adult tissues FIG Distribution of SMN, coilin-p80, and fibrillarin by triple labeling on (A) human fetal and adult pig skeletal muscle and (B) human fetal and adult pig spinal cord SMN (red) was identified using MANSMA1 and a TRITC-conjugated goat anti-(mouse Ig) Coilin-p80 (green) was identified using Ab204 and a FITC-conjugated goat anti-(rabbit Ig) Fibrillarin (blue) was identified using a human auto-antiserum and an AMCA-conjugated goat anti-(human Ig) (A) In fetal skeletal muscle, SMN colocalized with fibrillarin within the nucleolus (yellow arrows) Independent Cajal bodies (white arrows) are indicated In adult skeletal muscle, a Cajal body containing both SMN and coilin-p80 within the nucleolus (broken green arrows) is shown (B) Cajal bodies containing both SMN and coilin-p80 adjacent to, but not within, the nucleolus (green arrows) and within the nucleoplasm (broken white arrows) are indicated (C) Expanded views of nucleoli in adult pig tissues (a) The bright SMN staining (red) within the nucleoli (blue) is larger than typical gems but does not fill the nucleolus as in fetal tissue, while coilin-p80 staining (green outside nucleoli, or white when colocalized with SMN) is less dispersed (b) SMN staining within nucleoli occurs in muscle as well as liver (c) Three Cajal bodies, two inside a nucleolus and one in the nucleoplasm The white bars in A and B represent 30 ␮m and the black bar in C represents 10 ␮m 259 NUCLEAR GEMS AND CAJAL BODIES IN FETAL TISSUES TABLE Localization of SMN, Coilin-p80, and Fibrillarin in the Nucleolus of Fetal Tissues Nucleoli Tissue (cell) Liver 13.3 weeks (parenchyma) Skeletal Muscle 13.3 weeks Spinal cord 14.8 weeks (Alpha-motor neurons) (Nonneuronal cells) Spleen 18.2 weeks (parenchyma) Skin 11.8 weeks (epidermis) Cajal bodies in the nucleolus Fib (%) SMN/Fib (%) p80 (%) SMN/p80 (%) 87 (66.0) 69 (72.6) 39 (29.5) 24 (25.3) (1.5) (3.0) (2.1) 98 (85.2) 103 (66.4) 106 (69.3) 134 (70.2) 17 (14.8) 48 (31.0) 45 (29.3) 51 (26.7) (1.3) (0.7) (2.1) (1.3) (0.7) (1.0) Note The number of nucleoli containing fibrillarin alone (Fib), SMN colocalized with fibrillarin (SMN/Fib), Cajal bodies lacking SMN (p80), and Cajal bodies containing SMN (SMN/p80) is indicated Coilin-p80 was only present in nucleoli in the form of Cajal bodies Independent gems were not seen in nucleoli were inside nucleoli (Table 4) but the nucleolar SMN was sometimes more diffuse than is typical of gems (Fig 3C), although less diffuse than in the fetal nucleoli Nucleolar CBs have been reported in cultured cells but only as 1–2% of the total, although the proportion was greatly increased by okadaic acid, which inhibits protein dephosphorylation, or by transfection with mutants which mimic phosphorylated coilin-p80 [20] This suggests that nucleolar versus nucleoplasmic localization of proteins can be influenced by their state of phosphorylation Whether the localization of SMN is affected by SMN phosphorylation has not been established A nucleolar localization has also been established for another protein component of isolated gems, gemin4 [39] This protein interacts with SMN only indirectly through gemin3 (dp103) [15, 40, 41] and it is TABLE Localization of SMN and Coilin-p80 in the Nucleoplasm and Nucleolus of Adult Pig Tissues Nucleoplasm Tissue (cell) Spinal cord (motor neurons) Cerebral cortex (molecular neurons) Liver (parenchyma) Skeletal muscle Human skeletal muscle Nucleolus p80 (%) p80/SMN (%) p80 (%) p80/SMN (%) 48 (98.0) (2.0) (1.0) (3.5) (8.5) (10.0) 49 (52.0) 29 (50.0) 25 (53.2) 38 (54.3) (1.0) (1.0) (1.4) 43 (46.0) 26 (44.8) 18 (38.3) 24 (34.3) Note The number of Cajal bodies lacking SMN (p80) and containing both SMN and coilin-p80 (SMN/p80) in the nucleoplasm and nucleolus is indicated Counts were also performed on adult human skeletal muscle to determine whether the adult human and pig counts were comparable An anti-fibrillarin human autoantibody with an AMCA-conjugated goat anti-(human Ig) secondary was used to reveal nucleoli not clear whether it has any role in regulating nucleolar distribution of SMN The observed distribution of SMN (cytoplasmic, nucleolar, gems/CBs in nucleoplasm or nucleolus, nucleoplasmic gems without coilin-p80) may be stages in one pathway, some of which only become highly populated when a bottleneck arises Gall et al [21], have suggested that gems/CBs may be part of a complex pathway for assembly of the RNA transcription and processing machinery SMN is believed to be involved in the maturation of snRNPs in the cytoplasm and their transport into the nucleus [10 –13] Both gems and CBs have been proposed to be involved in the assembly, storage, or transport of components of the RNA splicing machinery [13, 23] It has been suggested that U snRNPs may undergo an additional maturation step within the nucleolus [20 –22] Bauer and Gall [42] suggested that U snRNPs and coilin-p80 are codependent on each other for CB targeting The nucleolar protein Nopp140 also appears to be essential for both this targeting process and the structural integrity of CBs and may pass through CBs before accumulating in the nucleolus [43] Passage through CBs on the way to the nucleolus has also been reported for U snRNPs [21] and small nucleolar RNAs (snoRNAs) [22] Low levels of coilin-p80 and U snRNPs have been identified within HeLa cell nucleoli under normal conditions and uptake of CBs into the nucleolus may be controlled by the phosphorylation state of coilin-p80 [20] It is conceivable that some components of gems/CBs, including SMN, may function in ribosomal RNA transcription/ processing in the nucleolus, as well as mRNA transcription/processing in the nucleoplasm [21, 44] It would be of interest for the pathogenesis of SMA if SMN could be shown to be different in motor neurons in some specific way that is obviously relevant to their survival We have found, however, that motor neurons differ from other cell types in many ways They are the 260 YOUNG ET AL only cell type in which we found more gems/CBs in adult than in fetal stages (Table 1) Colocalization of SMN with coilin-p80 was much greater than in other cell types at all stages of development (Table 2; Fig 3B) The relationship of SMN to the nucleolus was also different in motor neurons (Tables and 4) We noted in an earlier study that over 20% of gems/CBs in adult motor neurons were larger and adjacent to the nucleolus [26] It is not clear how these are related to “perinucleolar caps” which have been reported in neuronal and other cell types [32] These “caps” at the nucleolar periphery contain coilin-p80 which appears to exchange with CBs in response to transcriptional activity, “caps” occurring under low transcription conditions, and CBs under high transcription conditions [37] Apart from the perinucleolar bodies, gems/CBs in the interior of the nucleolus were hardly ever observed in adult or fetal motor neurons and the dispersed nucleolar distribution of SMN was less common in fetal motor neurons than in other fetal cell types (Table 3; Fig 3B) It is unclear how these quantitative differences in motor neurons could render them more susceptible to low SMN concentrations The possibility that SMN may have an additional role in motor neurons, outside RNA splicing, has been raised [16, 17] However, the observation that most of the missense mutations that cause SMA are found in regions of SMN involved in binding Sm core proteins of snRNPs [10] is consistent with involvement of a snRNP maturation defect in SMA pathogenesis This work was supported by a research grant from the Muscular Dystrophy Association (U.S.A.) We thank Dr Angus Lamond (University of Dundee) for p80-coilin antibodies and Dr Michael Pollard (Scripps Research Institute, La Jolla, CA) for fibrillarin antibodies 10 11 12 13 14 15 16 REFERENCES Munsat, T L., and Davies, K E (1990) Genetic homogeneity between acute and chronic form of spinal muscular atrophy Nature 345, 823– 825 Coovert, D D., Le, T T., McAndrew, P E., Strasswimmwer, J., Crawford, T O., Mendell, J., Coulson, S E., Androphy, E J., Prior, T W., and Burghes, A H M (1997) The survival of motor neuron protein in spinal muscular atrophy Hum Mol Genet 6, 1205–1224 Lefebvre, S., Burglen, L., Reboullet, S., Clermont, C., Burlet, P., Viollet, L., Benichou, B., Cruaud, C., Millasseau, P., Zeniani, M., Lepaslier, D., Frezal, J., Cohen, D., Weissenbach, J., Munnich, A., and Melki, J (1995) Identification of a spinal muscular atrophy-determining gene Cell 80, 155–165 Monani, U R., Lorson, C L., Parsons, D W., Prior, T W., Androphy, E J., Burghes, A H M., and McPherson, J D (1999) A single nucleotide difference that alters splicing patterns distinguishing the SMA gene and SMN1 from the copy gene SMN2 Hum Mol Genet 8, 1177–1183 Lorson, C L., Hahnen, E., Androphy, E., and Wirth, B (1999) A single nucleotide in the SMN gene regulates splicing and is responsible for spinal muscular atrophy Proc Natl Acad Sci USA 96, 6307– 6311 17 18 19 20 21 Lefebvre, S., Burlet, P., Liu, Q., Bertrandy, S., Clermont, O., Munnich, A., Dreyfuss, G., and Melki, J (1997) Correlation between severity and SMN protein level in spinal muscular atrophy Nat Genet 16, 265–269 Hsieh-Li, H M., Chang, J G., Jong, Y J., Wu, M H., Wang, N M., Tsai, C H., and Li, H (2000) A mouse model for spinal muscular atrophy Nat Genet 24, 66 –70 Monani, U R., Sendtner, M., Coovert, D D., Parsons, D W., Andreassi, C., Le, T T., Jablonka, S., Rossol, W., Prior, T W., Morris, G E., and Burghes, A H M (2000) The human centromeric survival of motor neuron gene (SMN2) rescues embryonic lethality in SMN Ϫ/Ϫ mice and results in a mouse with spinal muscular atrophy Hum Mol Genet 9, 333–339 Liu, Q., and Dreyfuss, G (1996) A novel nuclear structure containing the survival of motor neuron protein EMBO J 15, 3555–3565 Buhler, D., Raker, V., Luhrmann, R., and Fischer, U (1999) Essential role for the tudor domain of the SMN in spliceosomal U snRNP assembly: Implications for spinal muscular atrophy Hum Mol Genet 8, 2351–2357 Fischer, U., Liu, Q., and Dreyfuss, G (1997) The SMN–SIP1 complex has an essential role in spliceosomal snRNP biogenesis Cell 90, 1023–1029 Liu, Q., Fischer, U., Wang, F., and Dreyfuss, G (1997) The spinal muscular atrophy disease product, SMN, and its associated protein SIP1 are in complex with spliceosomal snRNP proteins Cell 90, 1013–1022 Pellizzoni, L., Kataoka, N., Charroux, B., and Dreyfuss, G (1998) A novel function for SMN, the spinal muscular atrophy disease gene product, in pre-mRNA splicing Cell 95, 615– 624 Meister, G., Buhler, D., Laggernauer, B., Zobawa, M., Lottspeich, F., and Fischer, U (2000) Characterization of a nuclear 20S complex containing the survival of motor neurons (SMN) protein and a specific subset of spliceosomal Sm proteins Hum Mol Genet 9, 1977–1986 Campbell, L., Hunter, K M D., Mohaghegh, P., Tinsley, J M., Brasch, M A., and Davies, K E (2000) Direct interaction of SMN with dp103, a putative RNA helicase: A role for SMN in transcription regulation Hum Mol Genet 9, 1093–1100 Strasswimmer, J., Lorson, C L., Breiding, D E., Chen, J J., Le, T T., Burghes, A H M., and Androphy, E J (1999) Identification of survival motor neuron as a transcriptional activatorbinding protein Hum Mol Genet 8, 1229 –1226 Williams, B Y., Hamilton, S L., and Sarkar, H K (2000) The survival of motor neuron protein interacts with the transactivator FUSE binding protein from human fetal brain FEBS Lett 470, 207–220 Bechade, C., Rostaing, P., Cisterni, C., Kalisch, R., La Bella, V., Pettmann, B., and Triller, A (1999) Subcellular distribution of survival of motor neuron (SMN) protein: Possible involvement in nucleocytoplasmic and dendritic transport Eur J Neurosci 11, 293–304 Schul, W., van Driel, R., and de Jong, L (1998) Coiled bodies and U2 snRNA genes adjacent to coiled bodies are enriched in factors required for snRNA transcription Mol Cell Biol 9, 1025–1036 Lyon, C E., Bohmann, K., Sleeman, J., and Lamond, A I (1997) Inhibition of protein dephosphorylation results in the accumulation of splicing snRNPs and coiled bodies within the nucleolus Exp Cell Res 230, 84 –93 Gall, J G., Bellini, M., Wu, Z., and Murphy, C (1999) Assembly of the nuclear transcription and processing machinery: Cajal bodies (coiled bodies) and transcriptosomes Mol Biol Cell 10, 4385– 4402 NUCLEAR GEMS AND CAJAL BODIES IN FETAL TISSUES 22 23 24 25 26 27 28 29 30 31 32 Narayanan, A., Speckmann, W., Terns, R., and Terns, M P (1999) Role of the box C/D motif in localization of small nucleolar RNAs to coiled bodies and nucleoli Mol Biol Cell 10, 2131–2147 Matera, G (1999) Nuclear bodies: Multifaceted subdomains of the interchromatin space Trends Cell Biol 9, 302–309 Sleeman, J E., and Lamond, A I (1999) Newly assembled snRNPs associate with coiled bodies before speckles, suggesting a nuclear snRNP maturation pathway Curr Biol 9, 1065– 1074 Carvalho, T., Almeida, F., Calapez, A., Lafarga, M., Berciano, M T., and Carmo-Fonseca, M (1999) The spinal muscular atrophy disease gene product, SMN: A link between snRNP biogenesis and the Cajal (coiled) body J Cell Biol 147, 715– 728 Young, P J., Le, T T., Nguyen, thi Man, Burghes, A H M., and Morris, G E (2000) The relationship between SMN, the spinal muscular atrophy protein, and nuclear coiled bodies in differentiated tissues and cultured cells Exp Cell Res 256, 365–374 Burlet, P., Huber, C., Bertrandy, S., Ludosky, M A., Zwaenepole, I., Clermont, O., Roume, J., Delezoide, A L., Cartaud, J., Munnich, A., and Lefebvre, S (1998) The distribution of SMN protein complex in human fetal tissues and its alteration in spinal muscular atrophy Hum Mol Genet 7, 1927–1933 Bohmann, K., Ferreira, J A., and Lamond, A I (1995) Mutational analysis of p80 coilin indicates a functional interaction between coiled bodies and the nucleolus J Cell Biol 131, 817– 831 Takeuchi, K., Turley, S J., Tan, E M., and Pollard, K M (1995) Analysis of the autoantibody response to fibrillarin in human disease and murine models of autoimmunity J Immunol 154, 961–971 Battaglia, G., Princivalle, A., Forti, F., Lizier, C., and Zeviani, M (1997) Expression of the SMN gene, the spinal muscular atrophy determining gene, in the mammalian central nervous system Hum Mol Genet 6, 1961–1971 Francis, J W., Sandrock, A W., Bhide, P G., Vonsattel, J P., and Brown, R H., Jr (1998) Heterogeneity of subcellular localization and electrophoretic mobility of survival motor neuron (SMN) protein in mammalian neural cells and tissues Proc Natl Acad Sci USA 95, 6492– 6497 Lafarga, M., Andres, M A., Fernandez-Viadero, C., Villegas, J., and Berciano, M T (1995) Number of nucleoli and coiled bodies and distribution of fibrillar centres in differentiating Purkinje neurons of chick and rat cerebellum Anat Embryol 191, 359 –367 Received October 19, 2000 Revised version received January 19, 2001 33 34 35 36 37 38 39 40 41 42 43 44 261 Carmo-Fonseca, M., Pepperkok, R., Carvalho, M T., and Lamond, A I (1992) Transcription-dependent colocalization of the U1, U2, U4/U6 and U5 snRNPs in coiled bodies J Cell Biol 117, 1–14 Carmo-Fonseca, M., Ferreira, J., and Lamond, A I (1993) Assembly of the snRNP-containing coiled bodies is regulated in interphase and mitosis: Evidence that the coiled body is a kinetic nuclear structure J Cell Biol 120, 841– 852 Ferreira, J., and Carmo-Fonseca, M (1995) The biogenesis of the coiled body during early mouse development Development 121, 601– 612 Janevski, J., Park, P C., and De Boni, U (1997) Changes in morphology and spatial position of coiled bodies during NGFinduced neuronal differentiation of PC12 cells J Histochem Cytochem 45, 1523–1531 Lafarga, M., Berciano, M T., Garcia-Sergura, L M., Andreas, M A., and Cormo-Fonseca (1998) Acute osmotic/stress stimuli induce a transient decrease of transcriptional activity in the neurosecretory neurons of supraoptic nuclei J Neurocytol 27, 205–217 Santama, N., Dotti, C G., and Lamond, A L (1996) Neuronal differentiation in the rat hippocampus involves a stage-specific reorganization of subnuclear structures both in vivo and in vitro Neuroscience 8, 892–905 Charroux, B., Pellizzoni, L., Perkinson, R A., Yong, J., Shevchenko, A., Mann, M., and Dreyfuss, G (2000) Gemin4: A novel component of the SMN complex that is found in both gems and the nucleoli J Cell Biol 148, 1177–1186 Charroux, B., Pellizzoni, L., Perkinson, R A., Shevchenko, A., Mann, M., and Dreyfuss, G (1999) Gemin3: A novel DEAD box protein that interacts with SMN, the spinal muscular atrophy gene product, and is a component of gems J Cell Biol 147, 1181–1193 Grundhoff, A T., Kremmer, E., Tureci, O., Glieden, A., Gindorf, C., Atz, J., Mueller-Lantzsch, N., Schubach, W H., and Grasser, F A (1999) Characterisation of DP103, a novel DEAD box protein that binds to the Epstein–Barr virus nuclear proteins EBNA2 and EBNA3C J Biol Chem 274, 19136 –19144 Bauer, D W., and Gall, J G (1997) Coiled bodies without coilin Mol Biol Cell 8, 73– 82 Isaac, C., Yang, Y., and Meier, T (1998) Nopp140 functions as a molecular link between the nucleolus and the coiled bodies J Cell Biol 142, 319 –329 Platani, M., Goldberg, I., Swedlow, J R., and Lamond, A I (2000) In vivo analysis of Cajal body movement, separation, and joining in live human cells J Cell Biol 151, 1561–1574  ... nucleoli of fetal tissues (Figs 3A and 3B) Gems and Cajal Bodies in the Nucleolus in Adult Tissues FIG Distribution of SMN and coilin-p80 in fetal human and adult pig tissues Visible gems and Cajal bodies. .. Note The number of nucleoli containing fibrillarin alone (Fib), SMN colocalized with fibrillarin (SMN/ Fib), Cajal bodies lacking SMN (p80), and Cajal bodies containing SMN (SMN/ p80) is indicated... all fetal tissues, colocalization of SMN and coilin-p80 within nuclear bodies increased with increasing fetal age (Table 2) Motor neurons of the spinal cord displayed the highest degree of colocalization,