Intracellular Compartmentalization of PDE4 Cyclic AMP-Specific Phosphodiesterases

METHODS: A Companion to Methods in Enzymology 14, 65–79 (1998) Article No. ME970566

Intracellular Compartmentalization of PDE4 Cyclic AMP-Specific Phosphodiesterases G. Scotland, M. Beard, S. Erdogan, E. Huston, F. McCallum, S. J. MacKenzie, A. H. Peden, L. Pooley, N. G. Rena, A. H. Ross, S. J. Yarwood, and M. D. Houslay1 Molecular Pharmacology Group, Division of Biochemistry and Molecular Biology, Davidson Building, Institute of Life and Biomedical Sciences, University of Glasgow, Glasgow G12 8QQ, Scotland, United Kingdom

The PDE4 cyclic AMP-specific phosphodiesterase family comprises a large number of different isoforms encoded by four distinct genes, with additional complexity arising through alternate mRNA splicing. This generates a number of distinct PDE4 isoforms with unique N-terminal regions. The range of such splice variants emanating from the four PDE4 genes appears to be highly conserved across species. One key role for such regions appears to be their potential to target isoforms to specific intracellular sites. Evidence for such a targeting role for these N-terminal regions can be gleaned by a variety of techniques. These include subcellular fractionation, confocal microscopy, binding assays to show association with proteins having src homology 3 (SH3) domains, and generation of chimeric constructs of these Nterminal regions with proteins that are normally expressed in the cytosol. q 1998 Academic Press

The PDE4 cyclic AMP specific phosphodiesterase family comprises a large number of different isoforms encoded by four distinct genes with additional complexity arising through alternate mRNA splicing (1–3). The genes are complex structures with many exons. For example, the human PDE4A gene spans 50 kb, comprises at least 17 exons, and is located 1 To whom correspondence and reprint requests should be addressed. E-mail: [email protected].

on chromosome 19p13.2. The catalytic unit of the various PDE4 family members is encoded by Ç7 exons, the sequences of which show high homology for the four different PDE4 families (4A, 4B, 4C, 4D). However, the extreme 3* exons, which encode the Cterminal tail of the various PDE4 enzymes, are very different and encode amino acid sequences that show no similarity between the C-terminal regions of members of any of the four PDE4 gene families. Nevertheless, within a particular PDE4 gene family, all active isoforms appear to have identical C-terminal regions. This observation has been usefully exploited (4–9) to generate antibodies/antisera that are able to detect all active PDE4 isoforms of a particular PDE4 gene family. Similarly, generic reverse transcription polymerase chain reaction (RT-PCR) primers can also be designed to identify the presence of transcripts for all active members of a particular PDE4 gene family without a priori knowledge of the range of splice variants present (10, 11). This is because alternative mRNA splicing appears, for active species, to be uniquely associated with 5* domain swaps. It should, however, be noted, that regions of sequence within the putative catalytic domain are PDE4 subfamily specific and have also been usefully exploited to generate generic RT-PCR primers aimed, again, at amplifying/detecting transcripts for all active members arising from a particular PDE4 gene subfamily. Thus, both molecular and immunological methods can be employed to detect active members of particular PDE4 subfamilies. 65

1046-2023/98 $25.00 Copyright q 1998 by Academic Press All rights of reproduction in any form reserved.

AID

Methods A 0566

/

671e$$$$81

02-19-98 14:50:46

metaa

AP: Methods A

SCOTLAND ET AL.

66

Alternative mRNA splicing appears to be restricted to producing PDE4 isoforms with distinct N-terminal regions (1–3). This is seen for products of all four PDE4 genes and is also seen for products of the cognate phosphodiesterase (PDE) gene in Drosophila melanogaster. Two distinct splice junctions have been noted that yield so-called ‘‘long’’ and ‘‘short’’ PDE4 isoforms. These splice junctions occur around regions of sequence that are unique to members of the PDE4 gene family and that are highly conserved between the four families. Such regions have been termed upstream conserved regions (UCRs), of which there are two, UCR1 and UCR2. Short linker regions (LRs), which show high heterogeneity between the four PDE4 subclasses, are also identified (3). These are LR1, which links UCR1 to UCR2, and LR2, which links UCR to the putative catalytic region (Fig. 1). Long forms of all four PDE4 genes possess all of these regions together with unique N-terminal regions. The short forms of 4B, 4C, and 4D again have unique extreme N-terminal regions, but while they lack UCR1 they have an intact UCR2. For PDE4A the situation is subtly different for, while again the short isoforms have unique N-terminal regions, the place-

ment of the second splice junction is different such that not only is UCR1 not present in these short PDE4A forms but they exhibit an N-terminally truncated UCR2 region. Such a mechanism of alternative mRNA splicing seems likely to produce approximately 20 isoforms from these four PDE4 genes. The range of splice variants emanating from the four PDE4 genes appears to be highly conserved across species. This suggests that this complexity is not an aberration but, instead, is of fundamental importance to the cAMP signaling system found in a wide variety of species ranging from D. melanogaster to humans. Such observations have prompted an analysis of the putative functional role(s) of these extreme N-terminal regions. Two proposals for which supportive experimental evidence is available are that such regions can serve to target isoforms to specific intracellular sites (4, 5, 8, 12–16) and that they can regulate the functioning of the catalytic unit either through interaction with binding proteins (4, 12) or through phosphorylation (2, 17, 18). Evidence of a targeting role for these N-terminal regions came initially from studies done on the short

FIG. 1. Schematic representation of the domain structure of PDE4 isoforms based on sequence comparison analyses. The size of the catalytic unit has been determined by similarity of the sequence with members of all PDE familes (28). That this region forms the catalytic unit has been proven by mutation and deletion studies (29–33). The upstream conserved regions (UCRs) were defined on the basis of similarity studies done on members of all the PDE4 families (1, 34). The linker regions (LRs) were identified as regions that link the two UCRs together (LR1) and UCR2 to the catalytic unit (LR2) and that are unique to specific PDE4 subfamilies (3).

AID

Methods A 0566

/

671e$$$$81

02-19-98 14:50:46

metaa

AP: Methods A

ANALYSES OF PDE4 INTRACELLULAR TARGETING

67

PDE4A form RD1 (4, 5, 13–15). This PDE4A isoIntracellular targeting is not an exclusive propform, when expressed in a variety of cell types and erty of PDE4A isoforms and has been noted for varialso in its native state, is found exclusively as a ous PDE4B and PDE4D isoforms (9). Nevertheless, membrane-associated species. In contrast, an engi- the natural range of partner/interacting proteins neered PDE4A species in which the N-terminal al- and the functional significance of such effects have ternatively spliced domain has been removed to yet to be defined. leave the ‘‘core’’ PDE4A species was found to be loA likely role for the intracellular targeting of PDE4 cated entirely in the cytosol (13). Such an engineered isoforms relates to their possible involvement in deform was fully functional and, indeed, was about fining the compartmentalization of cAMP signaling twice as active as RD1. The notion that the extreme within cells (20). The other pieces of molecular appaN-terminal region could serve as a targeting signal ratus that would be consistent with such a notion are was then proven by forming chimeric constructs (14) the localization of specific adenylate cyclase isoforms between this species and the normally cytosolic, bac- to defined regions of the plasma membrane and the terial protein chloramphenicol acetyltransferase. targeting of a fraction of the intracellular receptors Such a chimera was exclusively membrane-associ- for cAMP, namely, PKA-RII, to specific intracellular ated in a fashion akin to RD1. Indeed, the three- locales by virtue of their interaction with binding prodimensional structure of this N-terminal region and teins called AKAPs (20–22). These species are disthe presumptive targeting domain within it have tinctly located within cells and thus allow for distinct now been identified (19). The structure of this N- spatial and temporal cAMP gradients to be first esterminal alternatively spliced region does not sug- tablished and then subsequently sampled at different gest that it inserts into the membrane bilayer but, points within the three-dimensional matrix defined rather, interacts with an as yet unidentified mem- by a particular cell. The targeting of PDE4 species brane species, the subcellular localization of which might thus serve to manipulate the form of these is thus conferred on RD1. Indeed, in transfected COS gradients, to protect certain protein complexes from cells RD1 appears to be located in punctate vesicle protein kinase A (PKA) action, and to manipulate the structures that are found underlying the plasma threshold at which anchored PKA forms might be membrane, associated with the Golgi, and also in affected by adenylate cyclase activation occurring at dispersed vesicle structures (4). Analysis of human a discrete intracellular locale. thyroid cells transfected so as to express RD1 in a Here we describe various ways that have been stable fashion, however, showed it to be localized used to analyze the intracellular targeting of PDE4 exclusively to the Golgi (15). isoforms. Intracellular targeting has also been noted for the long-form PDE4A species, RPDE6 and PDE46 (5, 12, 16). In this case both in transiently transfected COS cells and when natively expressed in TESTED SEQUENCES FOR RT-PCR brain, these forms partition between a particulate DETECTION OF SPECIFIC PDE4 ISOFORMS (cytoskeletal?) fraction and the cytosol. The proposal has been made that they associate with cyA first step toward understanding the basis of toskeletal structures as these PDE isoforms are re- compartmentalization is to identify which of the possistant to release by detergent treatment. The sible molecular participants are present in the sysmechanism of anchorage has been postulated to be tem under investigation. Reliable means to do this mediated via interaction with src homology 3 (SH3) include Western blotting of proteins and analysis of domain-containing proteins (12). This proposal is RNA populations by Northern blotting, RNase probased on the observation that the extreme N-termi- tection, or RT-PCR. RT-PCR is an attractive technal region of these two PDEs contains motifs of the nique to employ because the necessary probes are form PxxPxxR, which form a consensus for binding easily made and the reactions themselves are robust to SH3 domains. Confirmation that this can occur, and rapid. RNA isolated from tissues is reverse tranin principle, has been demonstrated using in vitro scribed into cDNA, and then a fragment of this cDNA binding analyses with recombinant RPDE6 and is amplified by repeated rounds of specifically both intact src kinase and glutathione S-trans- primed DNA polymerization in the PCR. ferase (GST) fusion proteins formed from the SH3 The primers presented here have been used sucdomains of a wide range of proteins (12). cessfully to detect PDEs from a number of rat tissue

AID

Methods A 0566

/

671e$$$$81

02-19-98 14:50:46

metaa

AP: Methods A

SCOTLAND ET AL.

68

and cell line sources. Though these primers were not in any way designed with phylogenetic conservation in mind, the high conservation of PDE sequence between mammals means that used at low stringency, they are likely to be able to detect the relevant PDE in a range of species, as shown by our ability to identify several mammalian homologues for RD1 (RNPDE4A1) using the rat-specific primers. Design of Primers for Each Published Rat PDE4 Splice Variant We have designed and tested primers for RT-PCR detection of PDE4 splice variants reported in the rat. The longest reported sequences for each PDE4 splice variant were downloaded from the GenBank. These were used in multiple pairwise alignments on the GeneJockey II program (by Biosoft) to delimit how much unique sequence was contained in each splice variant . Primer pairs were designed to these unique sequences for RNPDE4A8 and RNPDE4A5. Even though there was often enough unique sequence for amplifying a good-sized fragment from the unique sequence alone, we designed most of our primers some way in from the 5* end, having found that some primers designed to the extreme 5* ends are poor amplifiers in PCR. Therefore, for RNPDE4A1, RNPDE4B2, RNPDE4B1, RNPDE4D1, and RNPDE4D3, we used a splice variant-specific primer coupled to a gene-specific (but not splice variant-specific) downstream primer. RNPDE4D2 has no unique sequence relative to RNPDE4D1 so primers were designed that allow discrimination from RNPDE4D1 by the size of the fragment (see below). RNPDE4B1 and RNPDE4C1 are likely to represent truncated sequences present in a number of splice variants and so any primers designed to these sequences are probably detecting more than one transcript. Our primers to RNPDE4D3 are also likely to be able to cross-detect the as yet uncloned rat homologues of HSPDE4D4 and HSPDE4D5. Candidate primer pairs are screened against a library of PDE sequences to make sure that they do not anneal to sequences other than those to which they were designed. They are also checked for internal or intermolecular complementarity. Primer sequences of between 18 and 26 bp are chosen and the melting temperature of each pair is matched (Ç657C). Total RNA Isolation and Reverse Transcription The PCR is highly sensitive. All glassware, solutions, Eppendorfs, and pipet tips should be sterilized

AID

Methods A 0566

/

671e$$$$81

by autoclaving. Filter tips reduce the risk of contaminating pipets with DNA. A pipet reserved for prePCR manipulations is also a good precaution. Tips must be discarded after one manipulation. RNA is highly sensitive to degradation by nucleases. Water used in RNA isolation is therefore treated with the RNase inhibitor diethyl pyrocarbonate (DEPC, 0.1%). Total RNA is typically isolated by homogenizing Ç150 mg ‘‘snap frozen’’ tissue in 1 ml Tri-Reagent (Sigma). RNA is degraded rapidly in situ so tissue should be used rapidly after sacrifice or immediately frozen in liquid nitrogen. RNA can be extracted from cell lines by scraping one confluent 80-cm2 flask into 1 ml Tri-Reagent. The RNA is then extracted by phase separation: The homogenate is stored at room temperature for 5 min, then centrifuged at 12,000g for 10 min at 47C. The supernatant is then transferred to a new tube. DNA and protein are removed from the supernatant by chloroform extraction. To each tube 0.2 ml of RNase-free chloroform per milliliter of Tri Reagent originally used is added. The chloroform and aqueous phases are mixed by brief vortexing and stored at room temperature for 3 min. The phases are then separated by centrifugation at 12,000g for 15 min at 47C. The upper aqueous phase is transferred to a clean microcentrifuge tube. RNA is precipitated by the addition of 0.5 ml of isopropanol/ml of Tri Reagent originally used at room temperature for 10 min. This RNA is then pelleted by centrifugation at 12,000g for 10 min at 47C. The RNA pellet is washed with 75% ethanol and stored at 0807C. Prior to reverse transcription, the RNA pellet is dried on the bench-top and resuspended in Ç100 ml DEPC water. Reverse Transcription Reverse transcription of RNA to cDNA is performed using MMLV reverse transcriptase (Firststrand cDNA Synthesis Kit, Pharmacia). Five micrograms of total RNA (as determined by its absorbance at 260 nm) is denatured at 657C for 10 min, then immediately placed on ice for 1 min. The RNA is incubated at 377C for 1 h in a reaction that contains MMLV reverse transcriptase, dithiothreitol (DTT), deoxynucleotide triphosphates (dNTPs), and an oligo(dT) primer which allows priming of cDNA synthesis from the poly(A) tail found in most mature RNAs. PCR Amplification Amplification is performed in 11 PCR buffer (50 mM KCl, 20 mM Tris–HCl, pH 9.0, 0.1% Triton X-

02-19-98 14:50:46

metaa

AP: Methods A

ANALYSES OF PDE4 INTRACELLULAR TARGETING

100) containing 200 mM concentrations of each dNTP, 500 mM concentrations of each primer (sense and antisense), 1.5 mM MgCl2 ; 5 U of Taq DNA polymerase (all reagents were from Promega) is added, just before cycling commences, to give a total volume of 50 ml. The PCR is then carried out on a thermocycler. A typical amplification protocol for primers with a melting temperature of 657C consists of 30 cycles of denaturation for 1 min at 947C, annealing for 1 min at 587C, and extension for 1 min at 727C. An aliquot of each reaction is resolved by electrophoresis on a 2% agarose gel, run with weight standards, and visualized with ethidium bromide under UV light. PCR Optimization The conditions of the PCR may need to be optimized, to strengthen the signal or alternatively to reduce artifactual priming events. A detailed discussion of PCR optimization is beyond the scope of this article. However, we have found two simple alterations that have been consistently useful: (1) Decreasing the temperature of the annealing step can often strengthen the PCR signal without appreciably elevating the background. (2) Hot start, where an essential PCR component is withheld until the first melt, can increase specificity. Our preferred method of doing this is to reversibly inactivate Taq with an antibody (Clontech) that is destroyed in the first melting step. Tested Primers for Splice Variant-Specific RT-PCR in the Rat This is set out below for various rat PDE4 isoforms. Information is given as name of splice variant; accession number; sense primer; antisense primer; starting nucleotide number; ending nucleotide number; fragment size. A more exhaustive list is given in Ref. 11. • RNPDE4A1A (RD1); M26715; ttcttctgcgagacctgctccaagc; caggccccatttgctcaagttctcc; 71; 442; 372. • RNPDE4A5A (RPDE6); L27057; aaggagcctgtctctctctcttccg; ggtaccggtgccgtggaagga; 1425; 1683; 259. • RNPDE4A8A (RPDE39); L36467; gcccagagaggcttggtgatttatcc; atattcgaggcagtgtcagcctcttgc; 45; 259; 215. • RNPDE4B1A (DPD); J04563; aaaccttcacggagcaccgaacaagagg; gccacgttgaagatgttaaggccccatt; 39; 545; 507.

AID

Methods A 0566

/

671e$$$$81

69

• RNPDE4B2A (pde4); M25350; ttggtagatcactgacacctcatcccg; gccacgttgaagatgttaaggccccatt; 20; 686; 667. • RNPDE4C1B (RPDE36); L27061; ccatggcccagatcactgggctg; ggacttcctggtccgcattaggctaat; 459; 1708; 1250. • RNPDE4D1A (pde3.1); M25349; tccggtgaagcgcttaagaactgagtcc; cctggttgccagaccgactcatttca; 237; 463; 227 and also the primer pairs acgtcaagctggagcatctcggc; cctggttgccagaccgactcatttca; 66; 463; 398. • RNPDE4D2A (pde3.2); U09456; acgtcaagctggagcatctcggc; cctggttgccagaccgactcatttca; 1; 321; 321. • RNPDE4D3A (pde3.3); U09457; ctaatttgcaagatcgcgcacccagc; cctggttgccagaccgactcatttca; 323; 577; 255.

TRANSIENT EXPRESSION OF PDE4 ISOFORMS IN COS CELLS Ultimately, analyses of the distribution and functioning of PDE4 isoforms need to be made in cells where they are expressed natively. However, the complex patterns of expression of various PDE isoforms in cells can often militate against definitive analyses of specific isoforms. To provide a test bed for comparative studies and for the analysis of mutant PDE4 forms, we have exploited the transient expression of PDE4 isoforms in COS cells. To date, analyses of the intracellular location of various PDE4 isoforms in COS cells have closely paralleled the situation seen for the natively expressed forms (23, 24). However, it is important to recognize the limitations of such a system and that PDE4 isoforms may also show different patterns of intracellular distribution, even among distinct cell types where they are expressed natively. The COS cell lines (COS-1, -3, and -7) are simian (African green monkey) kidney cell lines derived from the CV-1 line by transfection with an origindefective mutant of the Simian virus 40 (SV40) virus. The cells contain integrated copies of the early region of SV40 encoding the SV40 T antigen. This facilitates overexpression of proteins in these lines by transient transfection. The presence of the SV40 T antigen releases plasmids containing an SV40 origin of replication from normal copy number control. This allows the plasmid to be replicated more than once per cell cycle, resulting in the accumulation of multiple copies of the transfected cDNA and, hence, very high levels of expression.

02-19-98 14:50:46

metaa

AP: Methods A

SCOTLAND ET AL.

70

Maintenance of COS Cells COS cell lines are maintained in continuous culture as a monolayer at 377C in an atmosphere of 5% CO2 . The growth medium is a standard Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 2 mM L-glutamine, 10% (v/v) fetal calf serum, 1 unit/ml penicillin, and l mg/ml streptomycin. These lines have a fibroblast-like morphology and will shed cells on reaching confluency. Cells are passaged at about 90% confluency, and then split 1:3 to 1:6 every 2–4 days. Due to the willingness of COS cells to support growth of viruses and the possibility of their containing a complete SV40 region these lines should be regarded as potentially biohazardous material and appropriate precautions taken. DEAE Dextran Transfection TE buffer: 10 mM Tris–HCl, pH 7.2, 1 mM EDTA. Transfection medium: [DMEM (Sigma) supplemented with 2 mM L-glutamine, 10% (v/v) Nuserum, 1 unit/ml penicillin, 1 mg/ml streptomycin, 0.l mM chloroquine]. Shock buffer: 10% (v/v) dimethyl sulfoxide (DMSO) in phosphate-buffered saline (PBS). DEAE dextran. The cells are passaged 24 h before transfection and seeded onto new plates at about 50% confluency. We routinely transfect with 10 mg DNA per 79 cm2 plated cells. Prior to transfection the DNA is diluted to 40 ng/ml in sterile TE buffer, then further diluted 5:9 in sterile DEAE dextran (10 mg/ml in PBS). This mixture is incubated for 15 min at room temperature to allow formation of a DNA–DEAE complex. The growth medium is aspirated from the cells and replaced with transfection medium. We add 5 ml transfection medium for every 79 cm2 plated cells. The DNA–DEAE dextran mixture is dropped onto the cells and mixed by swirling. The cells are then incubated at 377C in an atmosphere of 5% CO2 for 3–4 h. Following this incubation the cells are shocked with DMSO; transfection medium is removed by aspiration and replaced with an equal volume of shock buffer, then incubated for exactly 2 min at room temperature. The cells are then washed once in PBS, transferred to growth medium, and incubated at 377C in an atmosphere of 5% CO2 for about 72 h before harvesting. Mock transfections are performed by treating the cells as described above but omitting DNA from the mixtures. Alternatively, mock transfections may be performed by using native vector in place of recombinant vector plus cDNA.

AID

Methods A 0566

/

671e$$$$81

Separation of Transfected COS Cells to P1, P2, and S Fractions KHEM buffer, pH 7.2: 50 mM KCl, 50 mM Hepes, pH 7.4, 10 mM EGTA, 1.92 mM MgCl2 , 1 mM DTT. TEA/KCl buffer, pH 7.2: 10 mM triethanolamine, 150 mM KCl. Protease inhibitor cocktail: 156 mg/ml benzamidine, 40 mg/ml phenylmethylsulfonyl fluoride (PMSF), 1 mg/ ml each of aprotinin, leupeptin, pepstatin A, and antipain. Prepare as a 10001 stock in 100% DMSO and store at 0207C. After transfection and the expression of PDE4 isoforms, cells are harvested and disrupted prior to the generation of subcellular fractions for analysis. Routinely we achieve ú96% cell breakage and thus generate a low-speed pellet (P1) containing nuclear and cytoskeletal components; a high-speed pellet (P2) containing plasma membranes, Golgi, endoplasmic reticulum, lysosomes, and endosomes; and a highspeed supernatant (S) fraction. This provides a rapid and reproducible means of determining whether PDE4 isoforms have the propensity to interact with subcellular structures in COS cells. We have also devised (13) a gradient subcellular fractionation technique that can be employed to determine association with specific intracellular organelles using biochemical procedures. COS7 cells are prepared for homogenization 72 h after transient transfection. The growth medium is aspirated and the cells are incubated in 5 ml (per 75cm2 flask) ice-cold complete KHEM buffer (KHEM / protease inhibitor cocktail) for 45 min at 47C. The cells are then washed with 5 ml of TEA/KCl at 47C for 10 min (longer incubation causes lifting of cells), the buffer is aspirated, and cells are washed with 5 ml of KHEM incomplete buffer. Finally, cells are washed with 1 ml of complete KHEM for 2 min and then the buffer is drained. The cells are harvested by scraping and collected by centrifugation at 6000gav before being finally resuspended into a volume of about 100 ml per 75-cm2 plate. This cell suspension is disrupted with 20 strokes of the glass pestle in a glass Dounce homogenizer. The postnuclear (P1) particulate fraction is generated by centrifuging the cell homogenate for 10 min at 1000gav at 47C. The P1 pellet is washed twice in complete KHEM buffer. High-speed centrifugation (for 1 h at 100,000gav at 47C) of remaining supernatant yields a second pellet fraction (P2) and a highspeed supernatant (S) termed the cytosolic fraction. The P2 pellet is washed in complete KHEM and re-

02-19-98 14:50:46

metaa

AP: Methods A

ANALYSES OF PDE4 INTRACELLULAR TARGETING

centrifuged as before. Finally, the P1 and P2 pellets are resuspended in KHEM buffer so as to normalize them both to the same volume as the S fraction. Samples are made into aliquots and snap-frozen in liquid nitrogen before being stored at 0807C.

71

plete KHEM buffer alone (control) or in complete KHEM buffer containing increasing concentrations of NaCl (0.5, 0.75, 1, 1.5, and 2 M). Samples are then incubated on ice for 30 min with occasional agitation, before recentrifuging at either 1000gav for 10 min (for P1 pellets) or 100,000gav for 1 h (for P2 pellets). Solubilization of PDE4 Isoforms Using Salt and Both supernatant and pellet fractions are retained Detergent Washing and then analyzed for PDE activity and for the reVarious PDE4 isoforms have been demonstrated lease of specific isoforms as determined by sodium to undergo distinct patterns of intracellular tar- dodecyl sulfate–polyacrylamide gel electrophoresis geting (3, 23, 24). To date, it has been shown that (SDS–PAGE) and immunoblotting. this appears to occur through protein–protein interA similar procedure can be adopted using a range action with membrane-associated proteins. The na- of detergent concentrations to determine susceptibilture of these interactions varies within each PDE ity to release. Routinely, detergent washing of P1 and gene family and would appear to be dependent on P2 fractions is done essentially as described for the the unique N-terminal regions of splice variants. salt treatment described above, but the concentraThe character of this interaction is of great impor- tions of Triton X-100 used are usefully varied from tance in the study of PDE compartmentalization and 0.01 to 5%. Certain integral proteins appear to recould be the key to understanding the nature and quire the addition of high salt concentrations as well function of differential targeting. as detergent to effect their release. To address this, Insight into the type of association of PDEs with in separate experiments we also treated the pellet membranes can be ascertained using nonionic deter- fractions with a range of Triton X-100 concentrations gent extraction to reveal integral membrane pro- in the presence of 1 M NaCl as detailed above. teins and elevated ionic strength to reveal peripheral Failure to release PDEs by any of these methods proteins adhering through ionic interactions. It may indicate that the enzyme is interacting with should be noted that the release of PDEs by these cytoskeletal components. various treatments could be because the PDE itself forms an integral or peripheral protein. However, an alternative explanation could be that the anchor species is either an integral or peripheral protein. ANALYSIS OF MEMBRANE LOCALIZATION Additionally, detergent or salt may disrupt the pro- OF cAMP-SPECIFIC PHOSPHODIESTERASE tein–protein interaction between the specific PDE4 RD1 (RNPDE4A1) BY CELL-FREE COUPLED isoform and its anchor. It is likely that the various PDE4 isoforms may provide a full spectrum of exam- TRANSCRIPTION/TRANSLATION SYSTEM ples of these possibilities. Those PDEs associated with membranes via peThe ability to generate proteins from their correripheral interactions are easily released by exposure sponding RNAs in cell-free environments originally to solutions of varying ionic strengths. This disrupts reported by Pelham (25), has been exploited by many electrostatic interactions without interfering with researchers working in a diverse range of biological the lipid bilayer of the membrane. It is important fields. This procedure requires that the gene of interwhen following this method to ensure that mem- est be cloned into a plasmid vector prior to the generbranes are prepared under isotonic conditions to ation of run-off RNA transcripts, which are subseavoid premature solubilization of peripheral pro- quently used to program the protein synthesis teins or nonspecific association of soluble proteins. ‘‘machinery’’ in cell lysates previously depleted of enSamples (200 mg each) of the cell particulate P1 and dogenous RNA and DNA by incubation in the presP2 fractions are washed twice by rehomogenizing in ence of Ca2/ dependent micrococcal nuclease. 200 ml of complete KHEM. These are then centrifuged Both the transcription and translation steps can be either at 1000gav for 10 min to recover P1 or for 1 h combined in a single reaction that makes use of douat 100,000gav to recover the P2 fraction. This is an ble-stranded plasmid DNA as template, thus simpliimportant step as any contaminating cytosolic enzyme fying experimental design. We have used (14, 19) a can seriously undermine interpretation. commercially available coupled transcription /transEach sample is then rehomogenized either in com- lation system in conjunction with a specially con-

AID

Methods A 0566

/

671e$$$$81

02-19-98 14:50:46

metaa

AP: Methods A

SCOTLAND ET AL.

72

structed plasmid vector to investigate elements responsible for membrane association of the cAMPspecific phosphodiesterase RD1 (RNPDE4A1). However, such a strategy could be readily adapted to analyze any stretch of sequence that is thought to contain a putative membrane/cytoskeletal targeting motif. Generation of Plasmid pGS7 Plasmid pGS7 was constructed as described by us in some detail (14) to generate an in-frame fusion between the N-terminal 1–100 amino acid fragment of RD1 and the bacterial enzyme chloramphenicol acetyltransferase (CAT). We chose CAT as our reporter system specifically for the reasons that this enzyme is exclusively expressed in the cytosol and is absent from mammalian cells. Generation of N-terminal RD1 Deletion Mutants Various mutations were generated within the 25amino-acid N-terminal region of RD1 by a modification of the method of overlapping extension mutagenesis which makes use of a complementary pair of primers containing the mutation of interest to create point mutations, deletions and insertions. The primer pairs used follow: 5*-GGTCGACTCTAGAATGTGCGAGACCTGC-3* and 5*-GCAGGTCTCGCACATTCTAGAGTCGACC3* to generate the deletion DP2-F7 (inclusive); 5*-GACTTCTTCCCCTGGCTGG-3* and 5*-CCAGCCAGGGGAAGAAGTC-3* to generate the deletion DC8-K13 (inclusive); 5*-CCTGCTCCAAGGACCAGTTCAAAAGG-3 * and 5*-CCTTTTGAACTGGTCCTTGGAGCAGG-3* to generate the deletion DP14-W20 (inclusive); 5*-GCTGGTGGATGCTGAACC-3* and 5*-GGTTCAGCATCCACCAGC-3* to generate the deletion DD21-R25 (inclusive). The above primers were used in conjunction with the following sense and antisense primers containing recognition sites for XbaI and XhoI (underlined), respectively, to generate partial PCR fragments with complementary 3* ends: 5*-GCGAGGGAATTCTAGAATGCCTCTGGTT-3*, 5*-GGCTCCTCGAGCTTCCAGTGTGT-3*. Under the reaction conditions used, the following components are mixed in a 0.5-ml microtube. Template DNA 5* primer

AID

Methods A 0566

1 mg 25 pmol

/

671e$$$$81

3* primer 101 Taq buffer 2 mM dNTPs 25 mM MgCl2 (See Note 6) Taq polymerase (5 units/ml) Distilled water

25 pmol 5 ml 5 ml 3 ml 0.5 ml To 50 ml

The reaction mixture is overlaid with 40 ml mineral oil or, alternatively, a thermal cycler that has a heated lid is used. The first cycle of PCR was designed when the addition of restriction enzyme recognition sites to a DNA fragment is required for sub cloning. PCR conditions follow: Denaturation Annealing Extension Denaturation Annealing Extension

947C 377C 727C 947C 507C 727C

1 min 1 min 1.5 min 1 min 1 min 1.5 min

1 cycle

30 cycles

On completion of thermal cycling PCR products were analyzed by low-melting-point agarose gel electrophoresis, and the bands identified and excised. A second round of PCR was carried out as before using Ç5 ml of each gel slice combined with the primers containing the XbaI and XhoI recognition sites. The recombinant PCR fragment containing the deletion was then purified as above and digested by XbaI and XhoI as described. Mutagenic fragments were ligated into the previously digested pGS7 in place of the wild-type fragment. Generation of Mature RD1–CAT Chimeras by Coupled Transcription/Translation By virtue of the bacterial SP6 promoter situated upstream of the ATG start codon of the RD1 sequence of pGS7, mature RD1–CAT chimeric protein species were generated using the TnT Coupled Transcription/ Translation according to the following protocol, Lysate TnT buffer SP6 polymerase Amino acid mix (minus methionine) [35S]methionine (1000 Ci/mmol) Plasmid DNA Distilled water

25 ml 2 ml 1 ml 1 ml 4 ml 0.5–2 mg To 50 ml

Reactions were incubated at 307C for 60–120 min and then stored at 0207C. Membrane Association Assay The assay buffer referred to in these experiments was TEN (40 mM Tris–HCl, pH 7.5, 10 mM EDTA,

02-19-98 14:50:46

metaa

AP: Methods A

ANALYSES OF PDE4 INTRACELLULAR TARGETING

120 mM NaCl), and the cell membranes used were COS7 cell P2 membranes and as such are referred to here only as an example. Duplicate samples of the above reaction mixtures are incubated with an equal volume of assay buffer either without or with cell membrane fractions for 30 min at 377C. Following incubation, the samples are subjected to centrifugation at 100,000g to generate pellet (P) and supernatant (S) fractions. The pellet fractions are washed by resuspension in an equal volume of assay buffer to minimize contamination by the reticulocyte lysate reaction mixture. The washed pellets are centrifuged as above and the resultant supernatants pooled with the supernatants from the initial spin. Pellets are resuspended in assay buffer, and all volumes and concentrations in the pellet and supernatant fractions are normalized accordingly. When the reactions were analyzed with a PhosphorImager following electrophoresis through a 10% denaturing polyacrylamide gel, multiple bands were observed in each lane (14). These bands arise as a result of multiple initiations from internal methionine codons downstream from the initiating methionine. Such events can occur when the initiating methionine falls within a suboptimal Kozak sequence particularly using cell-free systems which may lead to overexpression of recombinant proteins. While this may be regarded as a disadvantage from the point of some researchers, we were able to exploit this potential promiscuity in our studies of the mechanism of membrane association of RD1. Despite the fact that multiple truncated protein species are generated within this system, it was noted that in the lanes corresponding to the wild-type pGS7 only the fulllength 1–100 RD1–CAT chimera was able to associate with added membranes, whereas the truncated forms remained exclusively in the supernatant fraction (14). Such an ability to produce both the fulllength chimera and the N-terminally truncated forms within the same assay thus allows a panel of species to be ‘‘offered’’ to membranes. From this the species able to associate with membranes can readily be identified, with the others serving as internal controls. When developing such a methodology for use in analyzing other putative targeting sequences using chimeric species it is likely to be essential to also produce labeled CAT itself so that this can be added to the membrane binding assays so as to provide an internal control of a nonbinding species. This approach has been used to demonstrate that localization of the cAMP-specific PDE RD1 (RNPDE4A1) to the plasma membrane involves spe-

AID

Methods A 0566

/

671e$$$$81

73

cific amino acid residues contained within its extreme N terminus (14, 19). Undoubtedly, the success of this method in this instance is due to the generation of multiple protein species which serve as internal controls for membrane association. While it is recognized that many proteins expressed in this system would provide but a single protein species in instances where the initiation methionine was located within an optimal Kozak sequence, it is possible to generate native soluble CAT in the cell-free coupled transcription/translation system from a plasmid lacking the RD1-specific sequence, such as pGS8 described by us previously (14).

BINDING OR PULL-DOWN ASSAYS The recruitment of molecules into signaling complexes by regulated protein–protein interactions is recognized as an important feature of many signal transduction cascades. Such interactions are frequently mediated by the association of well-defined, independently folding structures, typified by SH2 and SH3 domains, present in one of the binding partners with short, peptide-like motifs in the target molecules (23, 26, 27). It is widely accepted that these domains can be produced as correctly folded structures, retaining their native binding characteristics when expressed in Escherichia coli as in-frame fusion proteins with GST. We have used (12) such approaches to investigate the potential interactions of certain members of the PDE4 cAMP-specific, rolipram-sensitive family of phosphodiesterases using SH3 and other protein binding domains. Construction of SH3-Domain GST Fusion Proteins The generation of GST fusion proteins is the first requirement for expression and purification of recombinant proteins. The pGEX vectors from Pharmacia Biotech contain a tac promoter for high-level inducible expression in E.coli hosts and require mild elution conditions for releasing the fusion protein from the affinity matrix, thus permitting both biochemical and kinetic analyses. The sequences of many proteins containing SH3 domains can be found in the GenBank database. Here we describe the generation of a fusion protein containing the SH3 domain of fodrin (nonerythroid spectrin), a cytoskeletal protein. Alignment of fodrin with the tyrosyl kinase Src indicated that the SH3 domain

02-19-98 14:50:46

metaa

AP: Methods A

SCOTLAND ET AL.

74

lay between amino acid residues 974 and 1021. The sequence of the fragment was examined for the absence of any restriction sites located in the multiple cloning site of the vector. The set of pGEX vectors has a range of available restriction sites and a choice of either thrombin or factor Xa cleavage sites. We have successfully used several of these vectors. In our case, neither BamHI nor EcoRI was present in the fodrin SH3 sequence and could be used in the design of PCR primers. The addition of these sites (underlined below) to the ends of the primers permitted inframe directional cloning of the SH3 domain to the N terminus of GST in pGEX-5X-1. Each restriction site was protected with six additional bases 5* to the end to facilitate restriction enzyme binding and cleavage. The primers used were PR1: 5*-CTG GTC GGA TCC CTC TAC GAC TAT CAG GAG AAG-3*, PR2: 5*-GGG TCC GAA TTC TTC ACG TAC GCA GCC GGC AC-3*. Templates for PCR can be either the plasmidcloned gene or cDNA derived from cell lines or tissue. We isolated RNA from Jurkat cells, a human T-cell line. Cells are homogenized in TRI reagent (Sigma) at a density of 5–10 1 106 cells ml01 and incubated at room temperature for 5 min. The homogenate is extracted with chloroform twice and the RNA precipitated with ethanol as per standard techniques. At this point it is convenient to freeze aliquots of RNA for latter use. For long-term storage, RNA should be kept under ethanol at 0807C. Five micrograms of total RNA is processed through to cDNA in the Pharmacia Biotech First-Strand cDNA Synthesis Kit. Using the method outlined by the manufacturer, which suggests a final resuspension volume of 33 ml, we have routinely found that 3 ml of this is sufficient to yield enough product in PCR to facilitate cloning. Typically, PCR conditions would be as follows: 30 cycles at 947C for 1 min, 507C for 1 min, and 727C for 1 min. Of course, the annealing temperature is determined by the sequence of the primers and needs to be determined for each specific primer set. With the primers described above, a 157-bp fragment is produced on PCR. The DNA is purified by passage through a low-melting-temperature agarose gel, subjected to a double digest with BamHI and EcoRI in a compatible buffer, and again gel purified as per standard techniques. At the same time, the vector is prepared in a similar manner. The two pieces are ligated overnight at 167C and transformed into E. coli JM109.

AID

Methods A 0566

/

671e$$$$81

All fusion proteins generated by cloning PCR fragments should be restriction mapped and sequenced on both strands to verify the generation of in-frame fusions. It is also important to authenticate the sequence by comparison with that lodged in GenBank and to affirm the absence of any PCR-generated point mutations. The constructs can now be used to generate small-scale inductions so as to confirm production of a protein of expected molecular weight. Induction of Fusion Proteins Modified PBS: PBS (Sigma) containing l mM DTT and a cocktail of protease inhibitors (40 mg/ml PMSF, 156 mg/ml benzamidine, 1 mg/ml apoprotinin, 1 mg/ml antipain, l mg/ml leupeptin, 1 mg/ml pepstatin, dissolved in DMSO). Isopropyl-b-D-thiogalactoside. Luria broth containing 100 mg/ml ampicillin. Cultures of E. coli (JM109), transformed with either the plasmid pGEX-5X-1 (for the production of GST) or recombinant pGEX-2T containing an SH3 or other protein binding domain as an in-frame fusion, are grown overnight at 377C, with agitation, in Luria broth containing 100 mg/ml ampicillin. Overnight cultures are used to inoculate larger (400 ml) cultures in the same medium and these are grown at 377C, with agitation, until the absorbance at 600 nm reaches 0.600–1.000 OD units. Expression of the fusion protein is then induced by adding isopropyl-b-Dthiogalactoside to a final concentration of 0.l mM and growth is continued for a further 4–6 h at 377C with agitation. The bacteria are harvested by centrifugation at 2500gav for 5 min in a refrigerated centrifuge, then resuspended in 20 ml modified PBS. The resuspended bacteria can be stored at 0207C in conveniently sized aliquots. Freezing prevents degradation of the fusion protein during storage and also weakens the bacteria, thus facilitating subsequent lysis. Purification of Fusion Proteins Modified PBS: see above. Glutathione agarose beads. Frozen aliquots of the bacteria are rapidly thawed at room temperature. The volume used is dependent on the level of expression achieved, which can vary widely between constructs. In our hands 4 ml of resuspended bacteria yields between 0.4 and 3 mg of most fusion proteins used. The defrosted bacteria are held on ice and sonicated in 20-s pulses, separated by 30-s intervals, until cell lysis is complete.

02-19-98 14:50:46

metaa

AP: Methods A

ANALYSES OF PDE4 INTRACELLULAR TARGETING

The intervals are included to allow heat generated during pulses of sonication to be dissipated and so to avoid excessive heating of the sample. The sonicate is then subjected to centrifugation for 1 min at high speed in a bench-top centrifuge to pellet bacterial debris and the supernatant is transferred to fresh tubes. Glutathione–Sepharose beads equilibrated in modified PBS containing 1 mM DTT and protease inhibitor cocktail (see above) are added to the sample and incubated end-over-end for either l h at room temperature or 47C overnight. We routinely add a 100-ml bed volume of glutathione agarose beads, which is sufficient to bind about 3 mg fusion protein. Following incubation, the beads are collected by centrifugation for 5 s at high speed in a bench-top centrifuge and the supernatant is discarded. At this point residual bacterial debris is occasionally present, occurring as a deposit on the walls of the tube slightly above the top of the pelleted beads. Where this occurs the debris is carefully removed and discarded with the supernatant. The beads are held on ice and washed three times over a 15-min period with l ml modified PBS containing 1 mM DTT and protease inhibitor cocktail. Washed beads are resuspended as a 25% slurry and assayed for protein concentration. Binding Assay Modified KHEM buffer: 50 mM KCl, 50 mM Hepes–KOH (pH 7.2), l0 mM EGTA, 1.92 mM MgCl2 containing 1 mM DTT, and a protease inhibitor cocktail (40 mg/ml PMSF, 156 mg/ml benzamidine, l mg/ ml apoprotinin, l mg/ml antipain, l mg/ml leupeptin, l mg/ml pepstatin, dissolved in DMSO). Elution buffer: 10 mM glutathione, 50 mM Tris– HCl, pH 8.0. Volumes of slurry containing 400 mg immobilized fusion protein are pelleted and the supernatants discarded. Within each assay the volumes taken are equalized with washed beads. The pellets are resuspended in crude cytosol diluted to a final volume of 200 ml in modified KHEM buffer. When performing pull-down assays on PDEs overexpressed in COS-7 cells we aim to use cytosol containing 30 to 40 EU PDE activity; this is typically 20 mg total protein. The immobilized fusion protein and cytosol are incubated together for 10 min end-over-end at 47C. The beads are then collected by centrifugation for 5 s at high speed in a bench-top centrifuge, and the supernatant is retained as the unbound fraction. The beads are held on ice and washed three times over

AID

Methods A 0566

/

671e$$$$81

75

a 15-min period with 400 ml KHEM containing 1 mM DTT and protease inhibitor cocktail. These washes are pooled along with the unbound fraction, and aliquots are taken for PDE assay and Western blotting. Bound PDE is eluted from the beads by incubating three times in 100 ml elution buffer, for at least 10 min each, end-over-end at 47C. The eluted fractions are pooled and aliquots taken for PDE assay and Western blotting. As PDEs are labile molecules, care should be taken to minimize opportunity for mechanical disruption of the enzyme at all stages of the assay. To this end, samples in which PDE is present should be mixed by gentle inversion, not vortexing, and should be maintained on ice whenever possible.

USE OF CONFOCAL MICROSCOPY TO ANALYZE SUBCELLULAR LOCALIZATION OF PDE4 ISOFORMS IN CULTURED CELL LINES Confocal microscopy has been used to analyze the subcellular localization of PDE4 isoforms transiently expressed in the COS cell system and in cells stably expressing transfected PDE species (4, 15, 16). Briefly, cells expressing the PDE of interest are grown and fixed to coverslips and the PDE4 isoform is then labeled using an antibody labeled with a fluorophore. The PDE is then localized by analyzing the cell under the laser scanning confocal microscope. In the microscope a laser light source is used to excite the fluorophore. Emitted light is then detected by photodetectors positioned behind a pinhole aperture. The pinhole aperture allows only light that is in focus to pass through while blocking the path of unfocused light. Coupled with the shallow depth of field (0.5–1.5 mm), the overall effect of this is that the microscope collects only light from a well-defined optical section within the specimen, rather than from most of the specimen as in conventional light microscopy. Elimination of the ‘‘out-of-focus’’ fluorescence results in an image of greater contrast, clarity, and detection. By taking a succession of optical sections throughout the specimen at successive focal planes it is possible to reconstruct a three-dimensional image of the cell. This allows visualization of the precise localization and arrangement of the PDE4 isoforms with respect to subcellular structures and compartments. The following methods detail the preparation of

02-19-98 14:50:46

metaa

AP: Methods A

SCOTLAND ET AL.

76

PDE-expressing cultured cell lines for analysis of the subcellular localization of the PDE species under the confocal microscope. The microscope used in this laboratory is a Zeiss laser-scanning confocal Axiovert 100 microscope. Although the methods are generally applicable to sample preparation, individual microscopes may vary in their mode of operation and detection and this should be considered when preparing samples for analysis. Preparation of COS-7 Cells Transiently Expressing PDE4 Isoforms Medium and solutions used for cell culture and transfection were: Complete DMEM: Sterile distilled water, 288 ml; DMEM 1 10, 40 ml; 10,000 units/ml penicillin per 10,000 g/ml streptomycin, 8 ml; glutamine (200 mM), 4 ml; sodium bicarbonate (7.5%), 20 ml; fetal calf serum, 40 ml. PBS. Trypsin/EDTA. 10 mg/ml DEAE dextran (Sigma) in PBS: Make fresh on day of use and filter-sterilized through a 0.45-mm filter. 100 mM Chloroquine (Sigma): Filter-sterilize and store at 47C. DMEM / 10% Nuserum (not fetal calf serum). 10% DMSO in PBS: Make fresh each time. 10 mM Tris/1 mM EDTA, pH 7.5. Sterilized. COS-7 cells are grown and maintained in complete DMEM in 75-cm2 tissue culture flasks at 377C in an atmosphere of 95% O2 :5% CO2 . Cells are fed on alternate days with fresh medium. Twenty-four hours prior to transfection, the cells are split onto 10-cm plates at 50% density as follows. Spent culture medium is removed and the monolayer of cells is washed once with 10 ml of PBS. After removal of the PBS, 5 ml of trypsin/EDTA is added at 377C for 5 min. The flask contents are transferred to a sterile universal bottle and spun at 3000gav for 5 min. The supernatant is discarded, the cell pellet resuspended in fresh culture medium at the required density, plated onto 10 cm plates, and returned to the incubator for a further 24 h. On the day of transfection, for each 10-cm plate, 5–10 mg of the DNA is diluted to 250 ml using Tris/EDTA in a sterile bijou bottle and 200 ml of DEAE dextran is added. The DNA/dextran mixture is left at room temperature for 15 min. In a separate tube, to 5 ml of DMEM/10% Nuserum, 5 ml chloroquine is added. The old growth medium

AID

Methods A 0566

/

671e$$$$81

from the COS cells is added, as is the DMEM/Nuserum/chloroquine mixture. The DNA/DEAE dextran mixture is dropped onto the cells. The mixture is gently swirled and incubated at 377C for 3–4 h. Cells are shocked with 10% DMSO: The transfection medium is aspirated and replaced with 5 ml of 10% DMSO in PBS. It is left for 2 min at room temperature and aspirated immediately. The cells are washed once with PBS and then normal growth medium is added. Incubation is at 377C for a further 48 h. After the 48-h incubation, the cells should have recovered sufficiently to be transferred to coverslips. The cells are trypsinized from the plates as before and plated onto coverslips (see below) to give a final cell density of 60–80%. The cells are allowed to grow for 24 h and then fixed (see below). Preparation of Cells Stably Expressing Transfected PDE Species The precise details of culture medium and conditions are obviously determined by the parental cell line concerned and conditions used to obtain and maintain stable transfection of the PDE species. For the purpose of analysis under the confocal microscope, growing cells should be transferred to coverslips 24 h before fixation (see below) to give a final cell density of about 60–80% at the time of fixation. In theory, there is no reason why the method cannot be applied for the visualization of PDEs occurring naturally in cultured cell lines in a way analogous to fluorescence studies on tissue mounts. In practice, however, detection and image capture are limited by the relative abundance of the species of interest and this will probably be limiting in most native cell lines. Preparation of Coverslips and Fixation Protocol We routinely use 18 1 18-mm coverslips, placed in six-well tissue culture plates (one slip per well). The culture plates containing the coverslips must be sterilized before use. The coverslips can be autoclaved to sterilize them and then transferred to the culture plates using sterile forceps in the cell culture hood. Alternatively, the coverslips can be placed into the wells in a nonsterile environment, the culture plate wrapped in cling film, and the whole plate sterilized by microwaving on high power for 6 min. Once the plates containing the coverslips have been prepared, the cells can be transferred onto the coverslips at the appropriate time and density (see above). The final density of the cells is crucial for ensuring easy detection and capture of good images.

02-19-98 14:50:46

metaa

AP: Methods A

ANALYSES OF PDE4 INTRACELLULAR TARGETING

In our hands, we typically obtain a transfection efficiency of about 30% in the COS cell system, so only a relatively small number of cells are actually expressing the PDE of interest. Therefore, when using transiently transfected cells, it is best to aim for slightly higher densities of cells. When using stable transfectants though, each cell is expressing the PDE of interest. In this case it is better to aim for slightly lower densities so that individual cells can be visualized under the microscope without interference from neighboring, overlapping cells. Twentyfour hours after transfer, the cells are ready to be fixed. The growth medium is removed and each well is washed three times with 2 ml of PBS containing 1 mM CaCl2 and 1 mM MgCl2 . Care should be taken when washing the cells to apply and remove the PBS gently such that cells are not dislodged from the coverslip. Two milliliters of formaldehyde solution is added to each well and left to fix for 20 min. The formaldehyde solution is prepared by adding 15 g of paraformaldehyde (Sigma) to 500 ml of PBS that has been preheated to 807C in the fume hood. Once the paraformaldehyde has dissolved, the solution is allowed to cool to room temperature and 0.5 ml of 1 M CaCl2 and 0.5 ml of 1 M MgCl2 are added. The solution is filtered through a 0.45-mm nitrocellulose filter and stored in small aliquots at 0207C. After the cells have been treated with formaldehyde, it is crucial that they should not be allowed to dry out. The formaldehyde solution is removed and each well washed again three times with 2 ml of PBS containing calcium and magnesium. The cells are now fixed to the coverslips. They may be stored like this at 47C in PBS for up to a month before labeling or labeled immediately. If storing, the most effective way is to fill each well with PBS, seal the culture plate, with Parafilm and leave in the refrigerator. Staining Procedures 50 mM Ammonium Chloride in PBS Dissolve 0.156 g of NH4Cl in 50 ml PBS. Store at room temperature. PBS/0.2% Fish Skin Gelatin/0.1% Goat Serum A quantity of 2.2 g of 45% fish skin gelatin (Sigma) is weighed in a 500-ml glass beaker and 5 ml of goat serum and 500 ml of PBS are added. The mixture is stirred until the gelatin has dissolved and then the solution is filtered through nitrocellulose. The solution is stored at 47C for up to 7 days and filtered just before use.

AID

Methods A 0566

/

671e$$$$81

77

0.1% Triton X-100 in PBS Five hundred microliters of 10% Triton X-100 is added to 50 ml of PBS. This solution is made fresh each time. Mowiol Mounting Medium Six grams of analytical grade glycerol is placed in a 50-ml conical centrifuge tube and then 2.4 g of Mowiol (Calbiochem) is added. This is stirred thoroughly to mix, 6 ml of distilled water is added, and the solution is left for 2 h at room temperature. Twelve milliliters of 0.2 M Tris–HCl, pH 8.5, is added and the solution incubated at 507C for 10 min with occasional stirring to dissolve the Mowiol. The mixture is clarified by centrifugation at 5000g for 15 min, then filtered through nitrocellulose before use. The mixture is stored in a glass vial at room temperature. Method for Indirect Immunofluorescent Staining with Triton 1. Remove PBS from wells and add 2 ml of 50 mM NH4Cl/PBS for 10 min. 2. Wash three times with 2 ml of PBS. 3. Place coverslips in 2 ml of 0.1% (w/v) Triton X100 in PBS for 4 min. 4. Wash three times with 2 ml of PBS. 5. Wash three times with 2 ml of PBS/0.2% gelatin/0.1% goat serum over 5 min. 6. Wash three times with 2 ml of PBS over 5 min. 7. Prepare the anti-PDE antibody (primary antibody): Centrifuge the antibody in a microfuge on full speed for 10 min to remove any particulate debris. Dilute the antibody in PBS/gelatin/serum to the final working concentration and store on ice. The dilution of antibody used depends on the antibody and must be determined for each individual cell line and antibody combination. Commercial antibodies are normally supplied with the relevant data. For in-house sera, we usually use the same dilutions as determined for Western blotting procedures. 8. For each coverslip, place 100 ml of the antibody solution onto a strip of Parafilm. Carefully lay the coverslip on top of this droplet with the cell-covered side facing into the antibody solution. Try to avoid trapping air bubbles to ensure even contact between the solution and cell surface. Leave for 2 h. A pair of watchmakers forceps with curved ends is particularly useful for handling the coverslips. 9. Wash the coverslips three times in PBS/gelatin/ serum over 5 min. 10. Wash twice with 2 ml of PBS over 5 min.

02-19-98 14:50:46

metaa

AP: Methods A

SCOTLAND ET AL.

78

11. Stain with the secondary antibody as described in steps 7 and 8 for 2 h. The secondary antibody is an anti-IgG fluorescent conjugate. The antiIgG moiety is chosen with reference to the primary antibody. The fluorophore is chosen dependent on the capacities of the microscope available. Anti-IgG fluorophores are available from a number of sources such as TCS Biologicals and Molecular Probes. To maintain the fluorophore in as good condition as possible, this secondary labeling is best performed away from direct light. 12. Wash the cells twice with 2 ml of PBS/gelatin over 5 min. 13. Wash twice with 2 ml of PBS over 5 min. 14. Mount the coverslips onto microscope slides. Wash the coverslips with distilled water. Dry by blotting onto filter paper, and mount the coverslips in 12 ml of Mowiol with the cells facing into the Mowiol. 15. The slides may be stored at 47C in a foilwrapped container for up to a year or until the fluorescence fades. Alternative Staining Procedures and Controls Direct fluorescent labeling. It is sometimes desirable to analyze naturally occurring structures within the cell of interest such as the Golgi complex or nucleus. For this purpose, some commercial antibodies have been generated to specifically recognize such subcellular structures and are supplied prelabeled with the fluorophore, omitting the need for a two-stage antibody procedure. In such cases, steps 8–10 of the procedure outlined above are omitted. Dual labeling. If the microscope is sophisticated enough to deal with two fluorophores, dual labeling can be used to observe two structures within the same cell, at the same time. This can be particularly useful for ascertaining colocalization of molecules of interest. The labeling procedure is essentially as outlined above except that in steps 7–9, there are now two primary antibodies. These should be derived from different hosts, i.e., one mouse and one rabbit, such that they can be distinguished by the use of specific secondary antibodies. The secondary antibodies are then selected to have different fluorophores. To illustrate, one could analyze the PDE using a polyclonal antiserum raised in rabbit visualized with an anti-rabbit IgG–rhodamine conjugate and, in the same cell, look at actin using a mouse monoclonal anti-actin antibody visualized with an anti-mouse IgG-conjugated fluorescein isothiocyanate (FITC) label.

AID

Methods A 0566

/

671e$$$$81

Controls. To ascertain the accuracy of the localization data obtained, a number of control slides should also be prepared: i. Untransfected parental cells should be labeled as above to ensure that the PDE antiserum is specific for the PDE species of interest. There should be no reaction. ii. Cells should be stained with the secondary antibody only to ensure that it is the anti-PDE antibody that is responsible for the localization images and not nonspecific binding of the secondary antibody. Again, no fluorescence should be seen. iii. Cells should be stained with preimmune serum for the primary antibody again to ascertain the specificity of the primary antibody. iv. Cells not permeabilized by Triton should also be labeled. This checks that intracellular staining is really the true location of the PDE rather than fortuitous nonspecific intracellular staining or redistribution of the PDE due to the permeabilization procedure. Image Capture and Optimization Image capture and optimization depend on the type of microscope used, the thickness of the cells, and the resolution demanded. As a routine, 0.25-mm sections using a 160 objective can be captured and analyzed. Taking a stack of sections through the entire cell and then using reconstruction software are recommended with the derivation of xy, zx, and zy sections for analysis as well as full visualization. This approach can, for example, readily distinguish between a plasma membrane-associated enzyme and one showing cortical localization which otherwise may be confused in single xy-section analyses (16). Antibodies for 4A-Specific Splice Variants Such procedures obviously depend on the availability of antibodies specific for the PDE species to be examined. The approach of this laboratory has been to generate antisera using either peptides or GST fusion proteins that reflect the C-terminal regions which are unique to all known active proteins encoded by each of the four PDE4 gene families (4, 6, 9, 16). This clearly does not discriminate between the various isoforms produced by alternative mRNA splicing of each of the PDE4 gene families. These, however, generate isoforms with distinct N-terminal ends, and we have been successful in generating antisera for certain specific PDE4 isoforms using pep-

02-19-98 14:50:46

metaa

AP: Methods A

ANALYSES OF PDE4 INTRACELLULAR TARGETING

tides and GST fusion proteins that reflect the unique parts of these regions.

ACKNOWLEDGMENT This work was supported by a grant from the Medical Research Council to M.D.H.

REFERENCES 1. Bolger, G. (1994) Cell. Signal. 6, 851–859. 2. Conti, M., Nemoz, G., Sette, C., and Vicini, E. (1995) Endocrine Rev. 16, 370–389. 3. Houslay, M. D., Sullivan, M., and Bolger, G. B. (1998) Adv. Pharmacol. 44, 225–342. 4. Shakur, Y., Wilson, M., Pooley, L., Lobban, M., Griffiths, S. L., Campbell, A. M., Beattie, J., Daly, C., and Houslay, M. D. (1995) Biochem. J. 306, 801–809. 5. McPhee, I., Pooley, L., Lobban, M., Bolger, G., and Houslay, M. D. (1995) Biochem. J. 310, 965–974. 6. Lobban, M., Shakur, Y., Beattie, J., and Houslay, M. D. (1994) Biochem. J. 304, 399–406. 7. Huston, E., Lumb, S., Russell, A., Catterall, C., Ross, A. H., Steele, M. R., Bolger, G. B., Perry, M., Owens, R., and Houslay, M. D. (1997) Biochem. J. 328, 549–558. 8. Bolger, G., McPhee, I., and Houslay, M. D. (1996) J. Biol. Chem. 271, 1065–1071. 9. Bolger, G. B., Erdogan, S., Jones, R. E., Loughney, K., Wilkinson, I., Farrell, C., and Houslay, M. D. (1997) Biochem. J. 328, 539–548. 10. Erdogan, S., and Houslay, M. D. (1997) Biochem. J. 321, 165– 175. 11. Kostic, M. M., Erdogan, S., Rena, G., Borchert, G., Hoch, B., Bartel, S., Scotland, G., Huston, E., Houslay, M. D., and Krause, E. G. (1997) J. Mol. Cell Cardiol. 29, 3135–3146. 12. O’Connell, J. C., McCallum, J. F., McPhee, I., Wakefield, J., Houslay, E. S., Wishart, W., Bolger, G., Frame, M., and Houslay, M. D. (1996) Biochem. J. 318, 255–262. 13. Shakur, Y., Pryde, J. G., and Houslay, M. D. (1993) Biochem. J. 292, 677–686.

AID

Methods A 0566

/

671e$$$$81

79

14. Scotland, G., and Houslay, M. D. (1995) Biochem. J. 308, 673–681. 15. Pooley, L., Shakur, Y., Rena, G., and Houslay, M. D. (1997) Biochem. J. 271, 177–185. 16. Huston, E., Pooley, L., Julien, J., Scotland, G., McPhee, I., Sullivan, M., Bolger, G., and Houslay, M. D. (1996) J. Biol. Chem. 271, 31334–31344. 17. Sette, C., and Conti, M. (1996) J. Biol. Chem. 271, 16526– 16534. 18. Sette, C., Vicini, E., and Conti, M. (1994) J. Biol. Chem. 269, 18271–18274. 19. Smith, K. J., Scotland, G., Beattie, J., Trayer, I. P., and Houslay, M. D. (1996) J. Biol. Chem. 271, 16703–16711. 20. Houslay, M. D., and Milligan, G. (1997) Trends Biochem. Sci. 22, 217–224. 21. Klauck, T. M., and Scott, J. D. (1995) Cell. Signal. 7, 747– 757. 22. Coghlan, V. M., Hausken, Z. E., and Scott, J. D. (1995) Biochem. Soc. Trans. 23, 592–596. 23. Houslay, M. D., Scotland, G., Erdogan, S., Huston, E., Mackenzie, S., McCallum, J. F., McPhee, I., Pooley, L., Rena, G., Ross, A., Beard, M., Peder, A., Begg, F., Wilkinson, I., Yarwood, S., Ackerman, C., Houslay, E. S., Hoffman, R., Engels, P., Sullivan, M., and Bolger, G. (1997) Biochem. Soc. Trans. 25, 374–381. 24. Houslay, M. D. (1996) Biochem. Soc. Trans. 24, 980–986. 25. Pelham, H. R. B. (1990) Trends Biochem. Sci. 15, 483–486. 26. Mayer, B. J., and Baltimore, D. (1993) Trends Cell Biol. 3, 8–13. 27. Pawson, T., and Gish, G. D. (1992) Cell 71, 359–362. 28. Charbonneau, H. (1990) in Cyclic Nucleotide Phosphodiesterases (Beavo, J., and Houslay, M. D., Eds.), Vol. 2, pp. 267– 298, Wiley, Chichester. 29. Jin, S. L. C., Swinnen, J. V., and Conti, M. (1992) J. Biol. Chem. 267, 18929–18939. 30. Jacobitz, S., Ryan, M. D., McLaughlin, M. M., Livi, G. P., DeWolf, W. E., and Torphy, T. J. (1997) Mol. Pharmacol. 50, 891–899. 31. Jacobitz, S., McLaughlin, M. M., Livi, G. P., Burman, M., and Torphy, T. J. (1996) Mol. Pharmacol. 50, 891–899. 32. Horton, Y. M., Sullivan, M., and Houslay, M. D. (1995) Biochem. J. 308, 683–691. 33. Owens, R. J., Catterall, C., Batty, D., Jappy, J., Russell, A., Smith, B., O’Connell, J., and Perry, M. (1997) Biochem. J., in press. 34. Bolger, G., Michaeli, T., Martins, T., St John, T., Steiner, B., Rodgers, L., Riggs, M., Wigler, M., and Ferguson, K. (1993) Mol. Cell. Biol. 13, 6558–6571.

02-19-98 14:50:46

metaa

AP: Methods A