6.1. INTRODUCTION

Members of the transient receptor potential (TRP) family are membrane proteins; they constitute cation channels that are involved in a vast variety of physiological processes in mammals, flies, and worms. Rather than going into specific functions of single TRP proteins, this contribution describes different procedures used to generate antibodies specific for TRP proteins, the characterization of these antibodies by Western blotting and immunoprecipitation, and common pitfalls that have to be considered when pursuing these applications. The antibodies for TRPC3, TRPC4, TRPM3, and TRPM4, which are described, have been generated in the authors’ laboratory.

Generating a research-grade antibody for a membrane protein is a painstaking effort that occupies months of laborious bench work and needs money for synthesizing, expressing, and purifying the antigen, for hosting the suitable animals to be used for immunization, for generating and selecting hybridoma cells, and for purifying and characterizing the final product. Especially, the validation process, including analysis of the requisite efficacy and/or specificity of the antibody, is crucial and a major ordeal that in a few cases is accomplished by many commercial suppliers that have proliferated and offer for sale antibodies directed at a wide range of proteins. Unfortunately, those commercial antibodies fail in many cases even the most basic tests of activity and/or specificity (e.g., selectively interacting with the target protein in protein lysates prepared from cells overexpressing the target protein versus lysates from cells that do not express the target protein at all), and it appears to be the rule rather than the exception that vendors pass the burden of antibody validation/quality control to the end user. The end user has to decide either to simply use such an antibody as it is for her/his intended research goals, with the assumption that the vendor has performed adequate quality control to demonstrate activity and specificity, or, instead, to spend considerable funds and time-consuming experiments to critically evaluate the antibody before use and eventual publication of results.

In some fields of TRP channel research, the poor quality of commercially available antibody reagents has caused considerable frustration among investigators and has led to publication and perpetuation of erroneous research results (compare¹). A recent issue of Naunyn-Schmiedeberg’s Archives of Pharmacology (volume 379, pages 385–434) deals with the poor quality of commercial antibodies for G-protein coupled receptors; several other important articles and editorials also discuss the use of antibodies and their crucial validation in general.^2–8

In this chapter, we discuss the in-house generation of different antibodies, including polyclonal, monoclonal, and Fabs (Fragments antigen binding) generated against peptides and recombinant fragments derived from TRPC3, TRPC4, TRPM3, and TRPM4. Because excellent recipes for the generation of antibodies do exist (e.g., Refs. 9 and 10), we do not include here step-by-step protocols but rather point to the problems in obtaining specific antibodies especially for TRP proteins.

6.2. GENERATION OF POLYCLONAL ANTISERA FOR TRPM4 IN RABBITS USING SYNTHETIC PEPTIDES AS ANTIGENS

Antibodies that recognize intact proteins can be produced through the use of short (∼10–25 amino acid residues) synthetic peptides derived from the primary structure, without first having to isolate the protein. An antibody produced in response to a simple linear peptide will most likely recognize a linear epitope in a protein. Furthermore, that epitope must be solvent-exposed to be accessible to the antibody. The general features of a protein that correspond to these criteria are turns or loop structures that are generally found on the protein surface connecting other elements of secondary structure and areas of high hydrophilicity, especially those containing charged residues. Accordingly, computer algorithms that predict protein hydrophilicity and tendency to form turns are useful (e.g., http://www.expasy.org/tools/protscale.html). In general, which prediction method to use is not crucial because there tends to be a high level of agreement among them. In addition, hydrophilic protein stretches obeying the above rules are far from abundant in TRP proteins, which, as integral membrane proteins, are rather hydrophobic (Figures 6.1a and 6.2a). None of the prediction methods will identify the one single sequence guaranteed to produce an effective antibody against any given protein. Rather, several sequences will be identified that have a higher than average probability of being an effective antigen.

FIGURE 6.1

Characterization of TRPM4 antibodies. (a) Hydropathy blot of the TRPM4 protein sequence according to Kyte and Doolittle. Predicted transmembrane domains and the positions of peptides (peptides 578, 732, 733, 734, 735, 679) and a protein fragment, which (more...)

FIGURE 6.2

Characterization of TRPC3 antibodies. (a) Hydropathy blot of the TRPC3 protein sequence according to Kyte and Doolittle. Predicted transmembrane domains and the positions of peptides (peptides 1055, 400, 1070, 1071, 1072, 399, 704, 1138, 398) and the (more...)

Figure 6.1a shows a Kyte and Doolittle plot¹¹ for the mouse TRPM4 protein. The hydrophobic stretches (positive values) predicted to represent transmembrane domains are indicated, as well as the positions of a protein fragment used for generating monoclonal antibodies and the peptides 578, 732, 733, 734, 735, and 679 likely to be antigenic according to the criteria listed above. After designing these six peptides, they were synthesized, coupled to a carrier protein, and used for immunizing six rabbits.

Rabbits are the usual animal of choice because they are genetically divergent from the mouse (and human) sources of the proteins studied; they provide as much as 25 mL of serum from each bleed without significant harmful effects. One has to consider that even in genetically identical animals, a single preparation of antigen will elicit different antibodies. Because most laboratory rabbits are outbred, these differences are more pronounced.⁹ The synthesized peptides, the haptens, have to be coupled to a carrier protein, a relatively large molecule capable of stimulating an immune response independently. Haptens themselves are too small and cannot elicit antibody responses on their own because they cannot cross link B-cell receptors and they cannot recruit T-cell help. When coupled to a carrier protein, however, they become immunogenic because the protein will carry multiple hapten groups that can now cross link B-cell receptors. In addition, T-cell-dependent responses are possible because T cells can be primed to respond to peptides derived from the protein. The most commonly used carrier protein is keyhole limpet hemocyanin, which is usually preferred over bovine or rabbit serum albumin because it tends to elicit a stronger immune response and is evolutionarily more remote from mammalian proteins.

The speed of developing a specific antibody depends on priming and boosting immunizations, but the actual amounts of specific antibody produced will vary considerably, depending on the immunogenicity of the antigen. Booster immunizations are started 4–8 weeks after the priming immunization and continued at 2- to 3-week intervals. Prior to the priming immunization and following the primary and each booster immunization, blood is taken and serum prepared. The pre-immune serum—from blood taken prior to immunization—is a critical control to ensure that the antibody activity detected in later bleeds is due to the immunization.

The presence of specific antibodies will then be determined using an appropriate technique such as Western blot of protein samples containing the target protein. In the case of TRPM4, we used protein lysates from COS cells that do not express TRPM4 (“C” in Figure 6.1b, c, d, f) and COS cells that are transfected with the mTRPM4 cDNA (“C-M4” in Figure 6.1b, c, d, f). Figure 6.1b and c show the corresponding Western blots using the sera obtained from bleeds 1, 2, and 3 after immunizations with peptides 578 and 679. A number of protein bands are recognized by these sera in lysates from both control COS cells (C) and TRPM4-expressing COS cells (C-M4), indicating that they are not related to TRPM4. However, the ∼135-kDa TRPM4 protein (arrowhead in Figure 6.1b and c) is clearly recognized by sera from the second and third bleeds of rabbit 578 and serum from the third bleed of rabbit 679. After affinity purification, the purified ab 578 nicely recognizes only TRPM4, whereas the ab 679 recognizes TRPM4 and at least five additional proteins both in lysates from COS controls (C), as well as TRPM4-expressing COS cells (C-M4) (Figure 6.1d). Although the affinity purification was repeated, the specificity of ab 679 was not improved.

Immunization of rabbits with the remaining four TRPM4-derived peptides did not yield any antibodies at all (Table 6.1), showing that only two out of the six immunizations yielded antibodies capable to decorate the TRPM4 protein overexpressed in COS cells. In another series of Western blots (Figure 6.1e) using microsomal protein fractions from mouse kidney, only ab 578 recognized the ∼135-kDa TRPM4 protein (Figure 6.1e, wt, wild type), whereas ab 679 did not. As a control, the protein fractions from the same type of tissues from TRPM4-deficient mice¹² were used (Figure 6.1e, M4^–/–).

TABLE 6.1

Peptides Derived from the Mouse TRPM4 Sequence Used as Antigens to Immunize Rabbits

A monoclonal antibody (mab VF7D7E10), prepared in parallel using the indicated N-terminal TRPM4 fragment for immunization (Figure 6.1a), also nicely recognizes the TRPM4 protein overexpressed in COS cells (Figure 6.1f, left panel, C-M4 versus C), as well as the endogenous protein present in protein fractions from wild-type mice (Figure 6.1f, right panel). In contrast to ab 578 (Figure 6.1e), it also recognizes additional proteins of ∼120 kDa, which are also present in the fractions obtained from TRPM4-deficient mice¹² (Figure 6.1f, M4^–/–) and therefore are not related to TRPM4. This example demonstrates the importance of testing the antibody on native tissues and using samples from the gene knockout mice, as without the TRPM4 knockout mice being available, it would have been difficult to decide which of the proteins recognized by this mab in protein fractions from wild-type animals is, in fact, TRPM4.

In summary, our efforts on producing TRPM4 antibodies have yielded one sensitive and specific polyclonal anti-TRPM4 antibody out of six immunizations using different potential antigenic peptides (ab 578, Table 6.1) and, in addition, one sensitive but less specific monoclonal antibody (mab VF7D7E10) from a purified N-terminal fragment.

6.3. GENERATION OF POLYCLONAL ANTISERA FOR TRPC3 IN RABBITS USING RECOMBINANT PROTEIN FRAGMENTS AS ANTIGENS

TRPC3, together with TRPC6 and TRPC7, constitutes a structurally related subgroup within the TRPC subfamily of proteins.¹³ Overall, amino acid sequence identity among the three proteins is 69.4% with TRPC3 being more closely related to TRPC7 (81.0% identity) than to TRPC6 (71.2% identity). The sequence similarities among the three proteins add further problems to the generation of specific antibodies as shown below. In order to make TRPC3-specific antibodies, we first immunized rabbits with synthetic peptides derived from the TRPC3 primary structure (Figure 6.2a, peptides 398, 399, 400, 704, 1055, 1070, 1071, 1072, and 1138) as described above. However, only two of the nine immunizations using nine rabbits yielded antibodies (ab 1070 and 1071) that specifically recognized the TRPC3 protein overexpressed in COS or human embryonic kidney (HEK) 293 cells. However, neither of them recognized the TRPC3 protein in Western blots using protein fractions from mice.

We therefore prepared His-tagged TRPC3 fusion proteins (fp) 1 to 7 and a TRPC3-maltose binding protein (MBP) fusion protein (fp306)¹⁴ (Figure 6.2a) using fragments derived from the N- and C-terminal sequences of the TRPC3 protein. The fusion constructs were expressed as recombinant proteins in Escherichia coli, affinity-purified, and used for immunization. Then the sera obtained from consecutive bleedings were tested for the presence of antibodies against TRPC3. Only three out of eight immunized rabbits produced antibodies that recognize TRPC3 expressed in COS cells: fp306 (Figure 6.2b, left panel) and fp1 (Figure 6.2c, left panel), both encompassing parts of the TRPC3 N-terminus, and fp7 (Figure 6.2c, right panel), covering the C-terminal end. The antibodies were affinity-purified using the respective TRPC3-protein fragments fused to GST (fp1, fp7) or, in the case of ab fp306, by chromatography on MBP-sepharose followed by a second purification step on TRPC3-MBP-sepharose to remove the anti-His-/anti-MBP-antibodies.¹⁴ The antibody fp306 nicely recognized both the mouse (Figure 6.2B left panel, mC3) and the human TRPC3 (Figure 6.2b, left panel, hC3) expressed in COS cells, but not in COS cells that had not been transfected with the respective cDNA (Figure 6.2b, “C”). As an additional control, an HA-tagged hTRPC3 cDNA was expressed (Figure 6.2b, C-hC3-HA) to demonstrate that the antibodies for the HA-tag (mab HA3F10, Figure 6.2b, right panel) and for TRPC3 (fp 306, Figure 6.2b, left panel) decorate proteins of the same molecular weight.

As shown in Figure 6.3c, the affinity-purified ab fp1 only recognizes TRPC3 (calculated Mr 95,672), whereas ab fp7 recognizes TRPC3 and, in addition, the slightly larger TRPC6 (calculated Mr 106,733) and TRPC7 (calculated Mr 99,475), indicating that the epitope recognized by ab fp7 is common to all three proteins. Accordingly, we used ab fp1 to identify the TRPC3 protein in microsomal membrane protein fractions prepared from various mouse tissues. It appeared that TRPC3 was detected in protein fractions prepared from cerebellum (Figure 6.2d, wild type), corresponding well to our previous findings that TRPC3 is needed for mGluR-dependent signaling in mouse cerebellar Purkinje cells.¹⁵ As a control, we used protein fractions of the same tissue from TRPC3-deficient mice¹⁵ (Figure 6.2d, TRPC3^–/–).

FIGURE 6.3

Characterization of TRPC4 and TRPM3 Fab fragments. (a) The Fab fragment consists of the variable (V) and constant (C) domains of the immunoglobulin light (L) and heavy (H) chains linked by a disulfide bridge. (b) Coomassie-stained 6.5%/15% SDS polyacrylamide (more...)

6.4. GENERATION OF RECOMBINANT FRAGMENT ANTIGEN BINDING (FAB FRAGMENTS) FOR TRPC4 AND TRPM3 USING SYNTHETIC PEPTIDES AS ANTIGENS

The large size of standard immunoglobulins comprising two heavy protein chains and two light chains that are intricately folded may impose practical limitations, especially when the antibody should gain access to hard-to-reach regions of the target protein, for example, to the pore region of ion channels. For such a condition, simpler and smaller proteins might perform better than full-size immunoglobulins. Especially for the cryptic epitopes of peptides displayed in an MHC class I context, antibody fragments were created by chopping off the stem of the Y-shaped immunoglobulin leaving just one “hand” to perform the chemical duty of the antibody:¹⁶ These antibody fragments, so called Fragments antigen binding or Fabs (∼50 kDa), comprise a complete light chain paired with the V_H and C_H1 domains of a heavy chain (Figure 6.3a). These fragments cannot recruit other effector molecules and cells in the same way as the full-size antibodies (∼150 kDa) do because they lack the protein stem that performs such task, but they might be able to sneak into domains of an ion channel protein, which line the ion-conducting pore or which are critical for channel gating and thereby might interfere with the channel function.

For isolation and selection of Fabs specific for mouse TRPC4 and mouse TRPM3, a phagemid library expressing a nonimmune, semisynthetic human Fab repertoire of 3.7 × 10¹⁰ independent fragments¹⁷ was screened with peptides (13–57 amino acid residues in length) derived from mouse TRPC4 and mouse TRPM3. The peptides were immobilized and incubated with the phages. Specifically bound phages were eluted after several washing steps and used to infect E. coli to obtain recombinant Fabs. These recombinant Fabs were purified from bacteria, and their authenticity was confirmed by Coomassie Blue-staining, which yielded one band of 50 kDa under nonreducing conditions and two bands of ∼23 kDa under reducing conditions (Figure 6.3b). The specificity of the recombinant Fabs was determined by ELISA using the respective peptides/protein fragments as antigens. Sixty-three clones out of 96 were specific for TRPC4 and 50 out of 96 for TRPM3. DNA sequencing of those positive clones yielded 9 and 12 independent Fabs for TRPC4 and TRPM3, respectively.

Further characterization of the recombinant antibodies included immunoblot analyses (1) of peptides (10- to 19mers) representing the antigenic TRP protein sequences and that were spotted on nitrocellulose membranes (dot blots, Figure 6.3c) and (2) of lysates of cells expressing TRPM3 and TRPC4, respectively (Western blots, Figure 6.3d). The epitope mapping by the dot blot of seven independent Fabs for TRPC4 is shown at the bottom panel of Figure 6.3c. Each spot, 1 to 24, represents a peptide decamer immobilized on the nitrocellulose membrane. The 24 peptides cover amino acid residues (aa) 489 to 521 of the TRPC4 primary structure, with sequences shifted by one amino acid residue per spot (Figure 6.3c, top panel). The Fab P3B5 (Figure 6.3c, bottom, bold) decorates spots 12 to 16 (Figure 6.3c, bottom panel), representing aa 500 to 513 (Figure 6.3c, top panel, bold). The smallest epitope present in all five spots is highlighted in gray and most probably represents the minimal epitope recognized by this Fab. Similar approaches revealed that the shortest and longest sequences recognized by the recombinant Fabs cover 4 and 10 aa, respectively. Interestingly, the 6-aa epitope recognized by Fab P3B5 is present not only within the 33-aa TRPC4 protein fragment used to screen the phagemid library but also in all other TRPCs except TRPC1 and TRPC2 (Figure 6.3c).

For detecting the full-length proteins in Western blots by the Fabs, TRPM3 and TRPC4 were initially enriched by immunoprecipitation, and then Fabs were used at concentrations up to 100 μg/mL (which corresponds to ∼2 μM) to detect the TRP proteins in cell lysates. The Western blots in Figure 6.3D summarize the results obtained with the TRPM3 Fab P1C10. It recognizes the TRPM3 protein in cell lysate (Figure 6.3d, lane 4), as well as among the proteins retained after immunoprecipitation with ab 695, an anti-peptide antibody generated for TRPM3 (Figure 6.3d, lane 6, arrowhead). As a control, we used the monoclonal antibody 9F6G8 (Figure 6.3d, lanes 1 to 3) generated against a TRPM3 fragment.

Recombinant Fabs are efficiently synthesized in bacteria, for example, P1C10 can be easily tailored to produce a concentration of at least 1.2 mg/L—one might rig them to tow fluorescent proteins—and because of their smaller size compared to full-length immunoglobulins, they might be suitable tools for immunocytochemistry and for interfering with the TRP protein function.

6.5. GENERAL COMMENTS ON THE VALIDATION OF ANTIBODIES

As shown above, the specificity controls are crucial before antibodies should be used. From our experience, the following approaches appear to be essential to evaluate the quality of any antibody:

• 1. Western blot: Selective decoration of the target protein in protein lysates prepared from cells overexpressing the target protein versus lysates from cells that do not express the target protein at all (compare Figures 6.1 and 6.2). Ideally, if different antibodies raised against different epitopes of the same target protein are available, they should label the same protein, as in the example shown in Figure 6.2b for ab fp306 and the anti-HA-antibody.
• 2. Immuncytochemistry: Selective staining of cells overexpressing the target protein versus those that do not express the target protein at all.

One should consider that HEK293 cells, like COS cells and Chinese hamster ovary (CHO) cells, have originally been generated and selected by many investigators because they very efficiently overexpress foreign proteins.¹⁸ Antibodies recognizing proteins overexpressed in these cells do not necessarily have the sensitivity to recognize the same proteins expressed under their endogenous promotors in primary cells and tissues. Therefore, the most rigorous approaches include comparing results obtained using the equivalent cells/tissue sections or protein fractions from wild-type animals and knock out animals in which the target antigen has been genetically deleted. This deletion should be demonstrated by independent techniques such as Southern blots and PCR.

Reduction of antibody staining intensity upon knock-down approaches such as siRNA might be useful as long as independent controls—i.e., controls that do not use the antibody to be validated—have demonstrated that the expression of the protein or mRNA is really knocked down.

Finally, one should be aware that the widely used “absorption control,” the competition of antibodies generated against synthetic peptides with excess peptide, determines only the specificity of the antibody for the incubating peptide but does not prove the specificity of the antibody for the target protein in the tissue.

6.6. COMPARING PROTEIN EXPRESSION DATA WITH RNA EXPRESSION DATA

To study whether a given gene is expressed in specific cells or tissues, mRNA/ transcript expression can be analyzed by Northern blots and RT-PCR. By pursuing these approaches, one only measures the expression levels of the mRNA, and the results do not substitute for protein expression data. Although extensive work on RT-PCR techniques is going on, little attention appears to have been paid to the relation between the mRNA expression level and corresponding protein abundance in eukaryotes. The few studies that addressed this question yield amazing results:^19–22 In order to obtain an estimate of the overall relationship between mRNA and protein abundances in human liver, Anderson and Seilhamer²⁰ found a correlation coefficient of 0.48 between them. These results, which are halfway between a perfect correlation and no correlation at all, were confirmed by studies comparing protein and mRNA abundances for one gene product across 60 human cell lines¹⁹ or for more than 150 gene products in the yeast Saccharomyces cerevisiae²³ Figure 6.4 shows extracted data from the latter study,²³ which concluded that the correlation between mRNA and protein levels is insufficient to predict protein expression levels from quantitative mRNA data. It was shown for some genes that while the mRNA levels were almost the same, the protein levels varied by more than 20-fold (Figure 6.4, data taken from table 1 of Ref. 23).

FIGURE 6.4

Correlation between protein expression and mRNA abundance. Data of the 11 yeast proteins are taken from table 1 from Ref. 23. The mRNA abundance (copies per cell, open squares) ranges from 8.9 copies (URA1, BAT2, CYS3, VMA2) to 11.9 (SSUP45), and the (more...)

Surely, antibodies are not always available, nor are knock out mice or knock-down approaches that can be readily applied to evaluate an antibody. The results summarized above reveal that, for analysis of gene expression, simple deduction from mRNA transcript analysis is insufficient and that approaches using antibodies or antibody-based enrichment of the proteins of interest, combined with state-of-the-art mass spectrometry-based proteomics using stable isotope labeling, comprise the currently available approaches for quantitative analysis of proteins present in a given tissue or cell.

6.7. CONCLUSION

Antibody-based strategies for characterization, localization, and isolation of target proteins are among the most crucial and widely used techniques in molecular and cellular science. Although generation of antibodies is assumed by many merely to represent a routine procedure, evaluation and appropriate use of antibodies remain a sophisticated and challenging ordeal. This applies especially for low abundance membrane proteins like the TRPs.

ACKNOWLEDGMENTS

The authors thank Michael X. Zhu for the cDNA of the HA-tagged human TRPC3; Christine Wesely, Ute Soltek, Christin Matka, and Kerstin Fischer for their valuable assistance; and Sabine Pelvay, Ramona Gölzer, and Elisabeth Ludes for synthesizing peptides, immunizing, and bleeding rabbits. This work was supported by Homburger Forschungsförderungsprogramm (HOMFOR) (M.M., S.L., M.J., S.E.P., G.H., R.Z., and V.F.), by Forschungskommission der Universität des Saarlandes (S.E.P., V.F.), by Deutsche Forschungsgemeinschaft (M.J., S.E.P., R.Z., and V.F.), and by Fonds der Chemischen Industrie (V.F.).

REFERENCES

1.

Flockerzi V, Jung C, Aberle T, et al. Specific detection and semi-quantitative analysis of TRPC4 protein expression by antibodies. Pflügers Archiv. 2005;451:81–86. [PubMed: 15965705]

2.

Swaab D.F, Pool C.W, Van Leeuwen F.W. Can specificity ever be proved in immunocytochemical staining. Journal of Histochemistry and Cytochemistry. 1977;25:388–391. [PubMed: 325124]

3.

Willingham M.C. Conditional epitopes. Is your antibody always specific? Journal of Histochemistry and Cytochemistry. 1999;47:1233–1236. [PubMed: 10490451]

4.

Burry R.W. Specificity controls for immunocytochemical methods. Journal of Histochemistry and Cytochemistry. 2000;48:163–166. [PubMed: 10639482]

5.

Saper C.B, Sawchenko P.E. Magic peptides, magic antibodies: guidelines for appropriate controls for immunohistochemistry. Journal of Comparative Neurology. 2003;465:161–163. [PubMed: 12949777]

6.

Gown A.M. Unmasking the mysteries of antigen or epitope retrieval and formalin fixation. American Journal of Clinical Pathology. 2004;121:172–174. [PubMed: 14983929]

7.

Lipman N.S, Jackson L.R, Trudel L.J, Weis-Garcia F. Monoclonal versus polyclonal antibodies: distinguishing characteristics, applications, and information resources. ILAR Journal. 2005;46:258–268. [PubMed: 15953833]

8.

Rhodes K.J, Trimmer J.S. Antibody-based validation of CNS ion channel drug targets. Journal of General Physiology. 2008;131:407–413. [PMC free article: PMC2346570] [PubMed: 18411328]

9.

Harlow E, Lane D. Antibodies a Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; 1988.

10.

Harlow E, Lane D. Using Antibodies. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; 1999.

11.

Kyte J, Doolittle R.F. A simple method for displaying the hydropathic character of a protein. Journal of Molecular Biology. 1982;157:105–132. [PubMed: 7108955]

12.

Vennekens R, Olausson J, Meissner M, et al. Increased IgE-dependent mast cell activation and anaphylactic responses in mice lacking the calcium-activated nonselective cation channel TRPM4. Nature Immunology. 2007;8:312–320. [PubMed: 17293867]

13.

Flockerzi V. An introduction on TRP channels. Handbook of Experimental Pharmacology. 2007;2007:1–19. [PubMed: 17217048]

14.

Philipp S, Strauss B, Hirnet D, et al. TRPC3 mediates T-cell receptor-dependent calcium entry in human T-lymphocytes. Journal of Biological Chemistry. 2003;278:26,629–26,638. [PubMed: 12736256]

15.

Hartmann J, Dragicevic E, Adelsberger H, et al. TRPC3 channels are required for synaptic transmission and motor coordination. Neuron. 2008;59:392–398. [PMC free article: PMC2643468] [PubMed: 18701065]

16.

Held G, Matsuo M, Epel M, et al. Dissecting cytotoxic T cell responses towards the NY-ESO-1 protein by peptide/MHC-specific antibody fragments. European Journal of Immunology. 2004;34:2919–2929. [PubMed: 15368308]

17.

de Haard H.J, van Neer N, Reurs A, et al. A large non-immunized human Fab fragment phage library that permits rapid isolation and kinetic analysis of high affinity antibodies. Journal of Biological Chemistry. 1999;274:18,218–18,230. [PubMed: 10373423]

18.

Eaton D.L, Wood W.I, Eaton D, et al. Construction and characterization of an active factor VIII variant lacking the central one-third of the molecule. Biochemistry. 1986;25:8343–8347. [PubMed: 3030393]

19.

Tew K.D, Monks A, Barone L, et al. Glutathione-associated enzymes in the human cell lines of the National Cancer Institute Drug Screening Program. Molecular Pharmacology. 1996;50:149–159. [PubMed: 8700107]

20.

Anderson L, Seilhamer J. A comparison of selected mRNA and protein abundances in human liver. Electrophoresis. 1997;18:533–537. [PubMed: 9150937]

21.

Anderson N.L, Anderson N.G. Proteome and proteomics: new technologies, new concepts, and new words. Electrophoresis. 1998;19:1853–1861. [PubMed: 9740045]

22.

Pradet-Balade B, Boulme F, Beug H, Mullner E.W, Garcia-Sanz J.A. Translation control: bridging the gap between genomics and proteomics? Trends in Biochemical Science. 2001;26:225–229. [PubMed: 11295554]

23.

Gygi S.P, Rochon Y, Franza B.R, Aebersold R. Correlation between protein and mRNA abundance in yeast. Molecular Cell Biology. 1999;19:1720–1730. [PMC free article: PMC83965] [PubMed: 10022859]

24.

Link S, Meissner M, Held B, et al. Diversity and developmental expression of L-type calcium channel β2 proteins and their influence on calcium current in murine heart. Journal of Biological Chemistry. 2009;284:30129–30137. [PMC free article: PMC2781568] [PubMed: 19723630]

25.

Wagner T.F, Loch S, Lambert S, et al. Transient receptor potential M3 channels are ionotropic steroid receptors in pancreatic beta cells. Nature Cell Biology. 2008;10:1421–1430. [PubMed: 18978782]