CHARACTERIZATION OF ANTIBODIES DIRECTED AGAINST THE ANkRD2 HUMAN MUSCLE PROTEIN

In order to study the function of the Ankrd2 protein, for which commercial antibodies are not available, we report the production and analysis of polyclonal antibodies to full-length Ankrd2 and its C-terminal and N-terminal regions, as well as a monoclonal antibody to the C-terminus of the protein� Epitope mapping making use of recombinant deletion mutants showed that an epitope located in region 323-333 aa of Ankrd2 is detected by the monoclonal antibody� The high specificity of all four anti-Ankrd2 antibodies for recombinant and endogenous Ankrd2 protein is also demonstrated�


INTRODUCTION
Three closely related proteins -CARP (cardiac ankyrin repeat protein; Ankrd1 or MARP1), Ankrd2 (ankyrin repeat domain 2; Arpp or MARP2), and DARP (diabetes-related ankyrin repeat protein; MARP3) -have been allocated to the muscle ankyrin repeat protein (MARP) family (Miller et al�, 2��3)� Miller et al�, 2��3)� Muscle ankyrin repeat proteins are characteri�ed by ankyrin repeat proteins are characteri�ed by are characteri�ed by modular structure, muscle-specific expression, and dual cellular locali�ation (nuclear and cytoplasmic)� They can play important structural and signalling roles and link the elastic I-band region as a passive stretch sensor to the control of transcription� Kemp et al� (2���) identified Ankrd2 as a novel identified Ankrd2 as a novel Ankrd2 as a novel as a novel transcript expressed in fast tibialis anterior muscles after 7 days of passive stretch immobili�ation in vivo� In addition to this, Ankrd2 was also dis-discovered through systematic sequencing of human skeletal muscle 3´-expressed sequence tags (ESTs) (�anfranchi et al�, 1���), showing �3�� identity at �anfranchi et al�, 1���), showing �3�� identity at ), showing �3�� identity at the amino acid level to CARP, a nuclear protein expressed in heart (Zou et al�, 1��7), endothelial cells (Chu et al�, 1��5) and (to a lesser extent) in skeletal muscle (Baumeister et al�, 1��7)� The human Ankrd2 gene encodes a 37-kDa protein that has a nuclear locali�ation signal (N�S), two PEST protein-destabili�ation motifs, and four ankyrin-repeat domains (Kemp et al�, 2���; Moriyama et al�, 2��1; Pallavicini et al�, 2��1)� Ankyrin repeats are ubiquitous motifs involved in macromolecular interactions, whereas PEST sequences (rich in Pro, Glu, Asp, Ser, and Thr) serve as signals that target proteins for rapid destruction� Although Ankrd2 is predominantly expressed in skeletal muscle, in the I band of the sarcomere (Tsukamoto et al�, 2��2), it is also found in adult heart (Pallavicini et al�, 2��1; Ishiguro et al�, 2��2)� It becomes strongly upregu-It becomes strongly upregulated in myocytes upon stretch (Kemp et al�, 2���), with denervation of skeletal muscle (Tsukamoto et al�, 2��2), and during eccentric exercise (Barash et al�,2���), suggesting its role in muscle cell response to acute stress� It has been shown that Ankrd2 can interact with the Z-disk proteins titin (Miller et al�, 2��3) and telethonin (Kojic et al�, 2���), as well as with three transcription factors: YB1, PM�, and p53 (Kojic et al�, 2���)� The observation that Ankrd2 can interact with proteins located both in the nucleus and in the Z-disk supports the hypothesis of its potential role S� KOJIĆ ET A�� �8� as a molecular messenger shuttling between the nucleus and the cytoplasm� Not only does Ankrd2 bind p53, both in vitro and in vivo, it also enhances upregulation of the p21 WAFI/CIPI promoter by p53 (Kojic et al�, 2���)� These findings suggest that Ankrd2 may be involved in sensing stress signals and linking these to muscle gene regulation� In this paper, we describe a preparation of four anti-Ankrd2 antibodies (three polyclonal and one monoclonal) and precise mapping of their epitopes in order to define their specificity for the Ankrd2 protein� In view of the dual locali�ation of Ankrd2 in muscle cells and the presence of a common motif such as ankyrin repeats, it is important to determine that anti-Ankrd2 antibodies specifically recogni�e the Ankrd2 protein and no other protein containing ankyrin repeats in muscle and heart tissue�

Expression and purification of His-tagged recombinant proteins
his-tagged recombinant proteins were produced using the QIAexpress system (QIAGEN)� The Ankrd2 constructs in pQE3� were transformed in E. coli strain M15 carrying pREP�� Bacteria were grown in �B containing 1�� μg/ml ampicillin and 3� μg/ml kanamycin� Expression from the pQE3� vector was induced by adding isopropyl-β-D-thi- Mice were immuni�ed with �� μg of purified proteins [diluted in PBS and mixed with Freund's incomplete adjuvant (SIGMA)] by intra-peritoneal injection� Every three weeks mice were injected with proteins� After three injections, a blood sample was taken; thereafter boosts were given every three weeks and samples taken once every three weeks� Sera were prepared by allowing the blood to coagulate (37°C for 3� min) and then centrifuging at 12���g at �°C for 3� min� To avoid the growth of bacteria and fungi, sodium a�ide was added to the sera to a final concentration of ���2��� Samples were aliquoted and stored at -2�°C� All collected sera were tested by Western blot analysis for immunoreactivity against human striated muscle proteins�

Monoclonal antibody production
A modified version of the protocol of harlow and �ane (harlow and �ane, 1��8) was used for production of monoclonal antibodies� here BA�B/c female mice were immuni�ed as already described for the production of polyclonal antibodies� When a good immune response was obtained, the animals were sacrificed� The spleen was taken and the cells disso-ciated and then fused with NP1 myeloma cells in the presence of polyethylene glycol� After fusion, hybridoma cells were selected using RPMI medium supplemented with sodium hypoxanthine aminopterin and thymidine (hAT, SIGMA)� Clones were tested by en�yme-linked-immunosorbent assay (E�ISA) en�yme-linked-immunosorbent assay (E�ISA) (E�ISA) against the Ankrd2 protein� In order to obtain monoclonal cell lines, positive clones were subcloned, tested by E�ISA, and then grown in DMEM supplemented with hAT� These monoclonal cell lines were tested by Western blot analysis for immunoreactivity against striated muscle proteins�

Expression and purification of GST-tagged recombinant proteins
Production of GST-tagged recombinant proteins and their purification were accomplished using the Glutathione-S-transferase Gene Fusion System (GE healthcare)� The Ankrd2 constructs in pGEX-�P-3 expression vector were transformed in E. coli strain B�21(DE3)-p�ys S (Promega) and grown in �B with ampicillin (1�� μg/ml)� The GST-recombinant protein expression was induced by adding IPTG to a final concentration of ��5-1 mM and the bacteria were grown for an additional 3 h at room temperature� Cells were pelleted by

Antigen and antibody preparation
In order to produce antibodies to the Ankrd2 protein, the full-length (5-333 aa) protein and two fragments of Ankrd2, N1-Ankrd2 (5-173 aa) and C1-Ankrd2 (2�1-333 aa), were used as antigens� Polyclonal antibodies were raised against the whole protein (anti-F�-Ankrd2), as well as the N-(anti-N-Ankrd2) and C-terminal (anti-C-Ankrd2) regions of Ankrd2� Monoclonal antibody was prepared against the C-terminus of Ankrd2� The antigens were expressed in E. coli as histagged recombinant proteins and used for immu-ni�ation of mice� Spleen cells isolated from mice immuni�ed with his-tagged C-terminal Ankrd2 were fused with NP1 myeloma cells and selected for hybridoma cells secreting Ankrd2 monoclonal antibody� Epitope mapping was performed for clone 2F1��

Mapping antigenic epitope for monoclonal anti-Ankrd2 antibody
Since the C-terminal region of Ankrd2 (2�1-333 aa) was used for the production of monoclonal antibod- For Western blot analysis, proteins (1/5� of that used in panel A) were separated by SDS-PAGE, blotted, and probed with the monoclonal anti-Ankrd2 antibody, clone 2F1� (Fig� 3, panel C), as well as the anti-F�-Ankrd2 antibody (Fig� 3, panel B) as a positive control and the pre-immune mouse serum (Fig� 3, panel E) as a negative control� The monoclonal antibody raised against the C-terminus of Ankrd2 interacted only with recombinant proteins containing the last 1� aa of the Cterminus, vi��, C (27�-333 aa), AC (12�-333 aa), and SC (317-333 aa), but not with the Ankrd2 fragments UC (5-31� aa) and DC ( 5  Here GST-Ankrd2 (FL, aa 5-333) and its fragments tagged with GST (DC, aa 5-323; UC, aa 5-316; AC, aa 120-333; N, aa 5-120; C, aa and SC, were expressed in E. coli and purified under native conditions.The GST protein was used as a negative control.Equal amounts of purified proteins were separated by SDS-PAGE and the proteins were visualized by staining with Coomassie Brilliant Blue (panel A).For Western blot analysis, 1/50 of that used for visualization was separated by SDS-PAGE, blotted, and probed with anti-FL (panel B), monoclonal anti-Ankrd2, clone 2F10 (panel C), and anti-N-Ankrd2 (panel D) antibodies, as well as pre-immune mouse serum (panel E) as a negative control.The C-terminal region of Ankrd2 with ankyrin repeats (AC) was not detected by the N-terminal polyclonal antibody (Fig� 3, panel D), suggesting that this antibody is specific for the N-terminal region of Ankrd2 and does not cross react with the ankyrin repeats� Polyclonal antibody raised against the C-terminus of Ankrd2 reacted with the C-terminus of the Ankrd2 protein, but not with the N-terminus and ankyrin repeats (Fig� �), thus demonstrating that this antibody is also specific for the Ankrd2 protein�

Anti-Ankrd2 antibodies recognize endogenous Ankrd2 in human heart and skeletal muscle
The specificity of all anti-Ankrd2 antibodies was tested by Western blot analysis of human heart and skeletal muscle proteins� The proteins were separated by SDS-PAGE and blotted� Endogenous Ankrd2 was detected by all four anti-Ankrd2 antibodies� Figure 5 shows a single �3-kDa band corresponding to endogenous Ankrd2, detected in both human heart and skeletal muscle protein extracts� DISCUSSION For investigation of newly discovered proteins, ones for which commercial antibodies are not available, it is helpful to produce specific antibodies as an alternative to tags and tag-specific antibodies� here we present the production and analysis of three mouse polyclonal and one monoclonal antibody raised against the full-length protein, the N-terminus and, the C-terminus of the Ankrd2 protein� The choice of antigens was based on modular structure of the Ankrd2 protein, the N-terminus (1-12� aa), ankyrin repeats in the middle portion of the protein (121-278 aa), and the C-terminus (27�-333 aa)� Recombinant his-tagged proteins purified under denatured conditions were used for immuni�ation of mice� Although all anti-Ankrd2 antibodies were generated against denatured antigens, they recogni�ed denatured Ankrd2 in human heart and skeletal muscle extracts (Fig� 5), as well as native protein in muscle tissue (as demonstrated by immunohistochemistry) (Pallavicini et al�, 2��1)� We experimentally confirmed the theoretical prediction of potential epitopes represented by amino acid sequences that may adopt an exposed conformation based on ProtScale analysis of Ankrd2 C-terminus� Using a panel of deletant mutants, we precisely mapped the epitope specifically recogni�ed by the monoclonal anti-Ankrd2 antibody (clone 2F1�) to the region of the last 1� aa of Ankrd2 (323-333 aa)� This region may represent a linear epitope, to judge from its si�e� Epitope si�e is defined by the binding site of the antibody and limited by complementary binding between antigen and antibody� Epitopes can be linear or conformational� Typically, �-7 amino acids or sugars can fit into the deep pocket structures of linear epitope binding sites; for example, the � his tag is only six amino acids but can be seen by its specific antibody� On the other hand, conformational epitopes of globular proteins cover much greater space on flatter surface binding sites of antibody and may consist of 15-22 amino acids (Kindt et al�, 2���)� As the anti-N-and anti-C-Ankrd2 antibodies were made using the corresponding regions and a part of ankyrin repeats, we took into account the fact that the ankyrin repeats represent a common motif composed of 33 aa that mediate protein-protein interactions (Michaely and Bennett, 1��2)� We here confirm that the anti-N-terminal and anti-C-terminal Ankrd2 antibodies are specific for the N-and C-termini of Ankrd2 and do not recogni�e Ankrd2 ankyrin repeats; they are therefore unlikely to cross react with the large number of ankyrin repeat proteins in cells� Western blot analysis of human heart and skeletal muscle proteins showed the existence of a single band that corresponds to the endogenous Ankrd2 protein, thus demonstrating the specificity of these antibodies raised against the human striated muscle protein Ankrd2� It has already been seen that Ankrd2 is more abundantly expressed in skeletal muscle than in the adult heart (Moriyama et al�, 2��1), as shown in Fig� �� It would appear that the polyclonal anti-C-Ankrd2 antibody demonstrated the lowest affinity for endogenous Ankrd2� In conclusion, we have characteri�ed a set of highly specific and defined antibodies produced in order to study a MARP protein family member, Ankrd2� Carp, a , a a cardiac ankyrin-repeated protein, and its new homolog, ankyrin-repeated protein, and its new homolog, ankyrin-repeated protein, and its new homolog, -repeated protein, and its new homolog, repeated protein, and its new homolog, protein, and its new homolog, protein, and its new homolog  Kemp, T. �., Sadusky, T. �., Saltisi, �., Carey, N., Moss, �., �ang, S. �., Sasson, D. A., Goldspink, G.,

Fig. 1 .
Fig. 1.ProtScale hydrophobicity graph for the Ankrd2 C-terminal region from amino acids 241 to 333.Arrows designate the peaks with minimal scores.The sequence of the region containing the second, third, and fourth potential epitope (310-333 aa) is given under the position axis.Sequences of the potential second, third, and fourth epitopes are underlined (the third and fourth regions are merged).Length of the interval (window size) used for profile computation was 9.
Fig. 2. Schematic representation of human Ankrd2 and its deletant mutants used to map antigenic epitopes.NLS, nuclear localization signal; ANK, ankyrin repeat domains.

Fig. 3 .
Fig. 3. Mapping of anti-Ankrd2 epitopes by Western blot analysis.Here GST-Ankrd2 (FL, aa 5-333) and its fragments tagged with GST (DC, aa 5-323; UC, aa 5-316; AC, aa 120-333; N, aa 5-120; C, aa 279-333; and SC, aa 317-333) were expressed in E. coli and purified under native conditions.The GST protein was used as a negative control.Equal amounts of purified proteins were separated by SDS-PAGE and the proteins were visualized by staining with Coomassie Brilliant Blue (panel A).For Western blot analysis, 1/50 of that used for visualization was separated by SDS-PAGE, blotted, and probed with anti-FL (panel B), monoclonal anti-Ankrd2, clone 2F10 (panel C), and anti-N-Ankrd2 (panel D) antibodies, as well as pre-immune mouse serum (panel E) as a negative control.

Fig. 4 .
Fig. 4. Specificity of polyclonal anti-C-Ankrd2 antibody for the C-terminus of Ankrd2.Western blot analysis of GST-tagged Ankrd2 (FL, aa 5-333), the N-terminus with ankyrin repeats (NA, aa 5-278), and the C-terminus (C, aa 279-333).Purified recombinant proteins were separated by SDS PAGE, blotted, and probed with anti-GST and anti-C-Ankrd2 antibodies.The GST protein was used as a negative control.