Using immunohistochemistry (IHC) to successfully determine the expression and spatial arrangement of different proteins in a tissue fundamentally depends on choosing an appropriate antibody that targets the antigen of interest. Achieving high specificity and sensitivity are obvious considerations when choosing an antigen-detecting protein for IHC. Yet a third and often overlooked factor in this process is reagent reliability. Issues regarding antibody reproducibility in biomedical research are becoming increasingly prevalent.1 Antibody reliability issues are compounded in IHC assays when one considers the fact that acquired images are usually read by eye, an inherently subjective subject. Therefore, rigorous methods, especially regarding the use of antibodies, must be established to minimize variation across individual IHC assays over time. This study aimed to innovate antigen-detecting proteins for IHC to provide assays with high specificity, high sensitivity, and consistency of use.
Traditionally, IHC uses monoclonal or polyclonal immunoglobulin G (IgG) antibodies to detect the antigen of interest. However, monoclonal and polyclonal antibodies often lack in sensitivity and specificity, respectively, and lose stability over time. To update our toolbox of antigen-detecting proteins for IHC, we have exchanged conventional IgG antibodies for nanoantibodies (nAbs). nAbs are obtained from nurse sharks and camelids such as llamas and alpacas. These animals evolved to have a unique family of antibodies that contain only heavy chains, which harbor a highly specific antigen-binding capacity in the absence of a light chain.2 nAbs are only about 15 kDa, making them among the smallest intact antigen-binding fragments available (Fig. 1). Even though nAbs are only about 10% the size of conventional monoclonal and polyclonal antibodies, nAbs possess a longer antigen recognizing region, enhancing their potential specificity for target antigens.3 nAbs can have a much higher antigen affinity than even monoclonal antibodies and can recognize epitopes that are hidden to conventional IgGs.4,5 Unlike IgG-derived Fv fragments, which are unstable and can spontaneously dimerize,6 nAbs are soluble, homogenous, and do not aggregate in nature.7 nAbs are expressed from a single gene and do not require posttranslational modifications; therefore, nAbs can be produced to reach high quantities in Escherichia coli.8 Importantly, nAbs are extremely stable at high temperatures and low and high pH, have a very good shelf life, and do not lose much binding activity over time.5,9–12 Perhaps the most useful application for nAbs is their ability to be genetically engineered; nAbs can be modified to be fused to other molecules such as another VHH domain, enzymes, or fluorescent proteins.9,13
Despite over 20 years of nAb research, the use of nAbs in IHC is poorly characterized. nAbs have been shown to be more sensitive than conventional IgG antibodies in IHC,14 perhaps due to the presence of a longer antigen-binding domain. The unique biophysical properties of nAbs give them an edge in several technologies,13,15 and they outperform conventional antibodies in many assays requiring a protein-binding reagent. The ability of nAbs to be genetically modified allows for variations to be made in the mode of detection for IHC. This is particularly important if one considers that, in addition to choosing the correct antibody, the outcome of an IHC assay also greatly depends on the use of a sensitive detection system when visualizing antigen-antibody reactions.
Here, we have optimized the use of penetrant, highly specific nAbs for use in IHC. Notably, we have modified native nAbs to allow for covalent modification of a highly biotinylated amplifier protein, which eliminates the need for secondary antibodies and further signal amplification. As a proof of principle, we demonstrate the use of 2 nAbs against basic fibroblast growth factor (bFGF) and p16. bFGF expression is highly regulated and is involved in development, tissue regeneration, and wound repair. Dysregulation of bFGF is believed to contribute to the pathogenesis of many types of cancer.16 The cell cycle inhibitor p16 is a sensitive marker for identifying high-risk human papillomavirus tumors, and is often expressed in many other tumor types.17 Therefore, incorporating nAbs to develop robust IHC methods to detect bFGF and p16 not only offers a valuable research tool, but also promises clinical and diagnostic value.
MATERIALS AND METHODS
nAbs were purified by metal affinity chromatography from E. coli periplasmic extract as previously described.18
Tissue blocks of pancreatic adenocarcinoma, spleen, breast carcinoma, ovarian carcinoma, and lung were obtained from BioOption (Brea, CA). Tissue had been formalin-fixed and paraffin processed. All sections were obtained from the same block of tissue.
Rabbit polyclonal bFGF antibodies were obtained from Abcam (Cambridge, MA; ab16828) and Bio-Rad (Hercules, CA; AHP1038) and used at a dilution of 1:100. Mouse monoclonal p16 antibody was obtained from Thermo Scientific (Waltham, MA; MA5-17093) and used at a dilution of 1:100. Biotinylated rabbit IgG (BA-1100), biotinylated mouse IgG (BA-2001), and streptavidin-bound horseradish peroxidase (SA-HRP) (SA-5004) were obtained from Vector Laboratories (Burlingame, CA) and used at a dilution of 1:150. Biotinylated bFGF nAb, bFGF nAb Plus (ABP-NAB-BFGFPB), and p16 nAb Plus (ABP-NAB-P16PB) were obtained from Allele Biotechnology (San Diego, CA). bFGF nAb and nAb Plus were used at 1 μg/mL and p16 nAb Plus was used at 4.5 μg/mL.
Tissue sections were cut at 4 μm and dried overnight at 40°C. Sections were deparaffinized in xylene (IMEB, San Marcos, CA) and then hydrated in descending grades of alcohol. Endogenous peroxidase was blocked with 3% H2O2 for 10 minutes. Heat-induced epitope retrieval was attained using a pressure cooker (Tefal, Rumilly, France) for 10 minutes at high pressure followed by a 20-minute cooling period; Diva Decloaker antigen retrieval solution (Biocare, Concord, CA) was used for bFGF and Borg Decloaker RTU antigen retrieval solution (Biocare) was used for p16. Sections were blocked with serum-free protein blocker (Dako, Carpinteria, CA) for 15 minutes to reduce nonspecific antibody binding. All primary antibodies were incubated for 1 hour at room temperature. For rabbit polyclonal bFGF and mouse monoclonal p16 antibodies, biotinylated anti-rabbit or anti-mouse secondary antibodies were added subsequently after washing the primary antibody off and incubated for 1 hour at room temperature. Then all sections were washed and incubated with SA-HRP conjugate diluted at 1:150 for 45 minutes. For tyramide signal amplification (TSA), sections were incubated with TSA-DNP (PerkinElmer, Waltham, MA; NEL747A001KIT) for 10 minutes, washed, and incubated with anti-DNP-HRP (PerkinElmer; NEL747A001KIT) for 30 minutes. For colorimetric staining, all sections were washed and DAB substrate (Vector Laboratories; SK-4105) was added for 10 minutes at room temperature. Counterstaining was performed using 20% Mayer hematoxylin (IMEB) for 5 minutes. Sections were washed with tap water, dehydrated with ascending grades of alcohol followed by 3 changes of xylene before samples were mounted in xylene-based mounting medium. All washes were carried out with PBS-T unless otherwise stated. Negative control sections were processed by omitting primary antibody.
Images were acquired using a brightfield microscope (Amscope, Irvine, CA) using a ×40 objective.
Monoclonal, polyclonal, and nAbs vary in many biophysical properties, particularly in specificity and sensitivity (Table 1). To assess the performance of nAbs compared with conventional IgG antibodies in IHC, we developed a nAb against the clinically relevant biomarker bFGF. bFGF staining is often heterogeneous and significantly increased in malignant tissues compared with normal tissues. For example, bFGF is known to be upregulated in pancreatic adenocarcinoma.19 To this end, we used tissue isolated from a pancreatic adenocarcinoma to directly compare the staining specificity and sensitivity of 2 bFGF polyclonal antibodies and a bFGF nAb. To ensure an even comparison, we chose quality commercial antibodies that were previously validated and referenced, and subsequently optimized each antibody for our assay individually.
An important consideration in selecting and utilizing a primary antibody for IHC is whether to use a directly labeled primary or implement a secondary antibody (indirect detection). Owing to flexibility and relatively low cost, most protocols use indirect means of detection (Fig. 2A). Indirect methods generally generate a more intense signal than direct methods, as multiple secondary antibodies are theoretically able to bind to a single primary. However, indirect methods risk nonspecific binding of the secondary antibody and require many stages of incubation, extra controls, and rigorously followed steps to avoid undesirable interactions. An alternative method is to use direct detection by a labeled primary antibody to eliminate any concerns of nonspecific binding of the secondary antibody. To create a streamlined, cost-effective nAb-derived IHC protocol, we chemically biotinylated the bFGF nAb for use in direct labeling (Fig. 2B). To represent a typical amplification system, we used a biotinylated rabbit secondary antibody for signal detection in tissues stained with bFGF rabbit polyclonal antibodies. We compared bFGF staining with tissue stained without primary antibody as a negative control. Our results indicate that with all antibodies tested, bFGF antigen could be detected in pancreatic adenocarcinoma (Fig. 2C). Consistent with other published data,16 bFGF staining is intense in both the nucleus and cytoplasm of pancreatic adenocarcinoma cells. Incorporation of a directly labeled nAb provided several advantages: (1) removal of a potentially background-inducing step; (2) elimination of costs of secondary antibodies and related reagents; (3) a streamlined protocol with increased reproducibility; and (4) a reduction in total handling time.
Although chemically biotinylating a primary antibody removes the need for a secondary antibody, a tertiary method of amplification may still be required. In our hands, signal from both the polyclonal bFGF antibodies and the bFGF nAb was much stronger after TSA,20 a widely used and effective technique for signal enhancement in IHC. TSA has been reported to increase sensitivity up to 100-fold compared with conventional avidin-biotin complex methods,21,22 and is often used for detecting small quantities of antigen or enhancing the performance of low affinity mouse and rabbit antibodies. However, the TSA method spreads the label from the location of the primary antibody, decreasing resolution and making label levels unreliable indicators of the amount of antigen present. Furthermore, any background label that is present before TSA is also amplified, clouding the pathologic interpretation. Finally, the cost of TSA kits is not trivial.
We took a novel approach to locally amplify the signal to eliminate the need for both secondary and tertiary layers of amplification by taking advantage of the fact that nAbs are amenable to genetic engineering. IHC staining intensity is a function of enzyme activity, and sensitivity is achieved by increasing the number of enzymatic molecules bound to the tissue. We hypothesized that if the number of biotin molecules at the site of the protein of interest could be maximized through direct conjugation with the nAb (Fig. 3A), we could achieve a high number of enzyme molecules through SA-HRP without use of a secondary antibody or additional signal amplification processes such as TSA (Fig. 3B).
To amplify signal intensity locally, we genetically engineered a panel of nAbs to be covalently modified with a fusion protein. The resulting nAbs, termed “nAb Plus,” are produced with an amplifier protein that can accommodate high-density covalent labeling of biotin molecules without affecting the binding function of the nAb (Fig. 3A). To determine if direct amplification of the signal on the level of the primary antibody would be sufficient to detect antigen without TSA or another means of amplification, we compared a polyclonal antibody against bFGF to a biotinylated bFGF nAb Plus. We also included a monoclonal p16 antibody and a biotinylated p16 nAb Plus for further comparison. Like our process of selecting antibodies for bFGF, we chose a previously referenced and validated p16 antibody.
We first tested each nAb Plus against negative tissue controls to determine if the addition of biotin molecules would affect nonspecific binding and thus increase background levels (Figs. 3C, D, top row). Little to no background could be detected with bFGF nAb Plus or p16 nAb Plus in their respective tissues, indicating that the nAb Pluses can maintain a high degree of specificity. Furthermore, we found that use of the nAb Plus to detect bFGF in both breast carcinoma and pancreatic adenocarcinoma increased clarity and resolution compared with conventional antibodies and had less background (Fig. 3C). p16 immunostaining was weakly positive in spleen and strongly positive in ovarian carcinoma for both conventional antibodies and p16 nAb Plus (Fig. 3D). Interestingly, the background was remarkably reduced when using the nAb Plus, particularly in the strongly positive ovarian carcinoma tissue. Using nAb Plus bypassed the need for a secondary antibody and accompanying amplification systems such as TSA while reducing background and increasing clarity of immunostaining and antigen localization.
By combining anatomic, immunologic, and biochemical techniques, IHC identifies particular cellular components within the appropriate tissue and disease context. Owing to the importance of choosing the best primary antibody in IHC for optimum assay development, ongoing advancements in new antibodies for use in IHC cannot be undervalued. Our results show that nAbs are a valuable improvement over traditional IgG antibodies for use in IHC. We bypassed the need for a secondary antibody—which is known to introduce background—by using a directly biotinylated nAb. The signal from the directly biotinylated bFGF nAb was comparable to the signal from 2 different commercial polyclonal antibodies used with a biotinylated rabbit secondary antibody (Fig. 2). Next, we used genetically modified nAbs to achieve the affinity and specificity of a single epitope antibody with high sensitivity by incorporating a novel, specific signal amplification method using covalent modifications. Addition of a biotinylated amplifier protein to nAbs enabled high density, covalent labeling of biotin and surpassed traditional amplification methods in terms of sensitivity, convenience, and detection efficiency (Fig. 3). Ultimately, the incorporation of nAb Plus to IHC protocols provides a simpler and more cost-effective protocol, easier standardization, and less procedural variability.
Currently, there are no universally accepted standardization guidelines for determining the applicability of particular antibodies for IHC. The polyclonal antibodies that we selected in these studies were previously referenced, reviewed, and validated for IHC. Indeed, when combined with a tertiary amplification method such as TSA, we achieved reasonably low background and high signal detection with the selected polyclonal antibodies. Nevertheless, individual validation of every lot of a polyclonal antibody is imperative, as batches of serum from immunized animals never contain the exact same combination of antibodies. Even monoclonal antibodies produced from hybridomas have the possibility of dying off or losing their antibody genes. In terms of production, nAbs are more consistent across lots than conventional antibodies as they are produced in E. coli from a single gene. One step toward standardizing IHC across disparate labs and clinics would be to incorporate nAbs, as the means of manufacture, stability, and shelf life of nAbs make them more reliable across individual assays. In addition to antibody choice, differences in how specimens are fixed and processed, how antigen retrieval is performed, and how signal is ultimately detected can all undoubtedly affect IHC outcome.23 Adopting standards for as many of these steps as possible is crucial for consistency across individual assays. Therefore, another step toward standardizing individual IHC assays could also be achieved by using a biotinylated nAb Plus to eliminate variations caused by different signal amplification and detection methods.
Clinically, performing IHC to detect cancer biomarkers plays an important role in the prevention and early diagnosis of cancer. In some cases of disease, intracellular localization of a protein is indicative of tissue health. For example, bFGF is normally expressed in the nuclei of normal mammary tissues, but is often detected in the cytoplasm of breast cancer tissues.16 With polyclonal antibodies, the signal may not be specific enough to clearly decipher between nuclear versus cytoplasmic staining. Developing specific, robust antibodies such as nAbs for IHC to clearly define the precise intracellular position of an antigen could be revolutionary in regards to biomarkers used for cancer diagnoses. Furthermore, with the advent of personalized health care, it may be desirable to perform several different diagnostic IHC tests on patients before selecting a targeted therapy. In this sense, testing a single biomarker for a single therapeutic target is inefficient. Our method of creating nAb Plus with covalent amplifier proteins allows for easy replacement of biotin with fluorescent probes, enzymes, or virtually any compound that can be used in a biochemical assay. This technology of adding various labels to nAbs enables several biomarkers to be multiplexed in IHC. In fact, a nAb Plus can theoretically be used in virtually any diagnostic test that requires antigen detection.
For a given antigen, there may be hundreds of conventional monoclonal and polyclonal IgG antibodies that might work in a given assay. Although it seems beneficial to have a plethora of antibodies to choose from, the researcher enters the paradox of choice: how to select the most relevant antibodies and test them all in a cost-effective manner. Furthermore, commercial antibodies are notorious for lacking reproducibility.1 Even with the correct antibody, storage at the wrong temperature, microbial contamination, or damage from repeated freeze/thaw cycles can adversely affect antibody function. Therefore, despite the fact that antibodies are the most widely used protein-binding reagent in basic and clinical research studies,24 many commercial antibodies are simply not reliable. Here, we provide a realistic solution to these problems by providing a specific, highly sensitive, stable, reproducible, and functional nAb. However, while nAbs are superior to traditional IHC antibodies in many aspects, few nAbs are currently available. This is in part due to the long, laborious, and costly process of nAb development. Nevertheless, we believe that we and others have shown the superior benefit of using nAbs in many protein-detecting assays; it is our hope that more nAbs against cancer biomarkers, stem cell markers, and other relevant antigens will be available soon.
The authors thank J. Sebastian Gomez Cavazos and Tiffany Phan for editing and proofreading as well as the entire Allele Biotechnology nAb team for discussions and technical support.
1. Bradbury A, Pluckthun A. Reproducibility: standardize antibodies
used in research. Nature. 2015;518:27–29.
2. Hamers-Casterman C, Atarhouch T, Muyldermans S, et al. Naturally occurring antibodies
devoid of light chains. Nature. 1993;363:446–448.
3. Muyldermans S, Atarhouch T, Saldanha J, et al. Sequence and structure of VH domain from naturally occurring camel heavy chain immunoglobulins lacking light chains. Protein Eng. 1994;7:1129–1135.
4. Muyldermans S. Nanobodies: natural single-domain antibodies
. Annu Rev Biochem. 2013;82:775–797.
5. van der Linden RH, Frenken LG, de Geus B, et al. Comparison of physical chemical properties of llama VHH
antibody fragments and mouse monoclonal antibodies
. Biochim Biophys Acta. 1999;1431:37–46.
6. Arndt KM, Muller KM, Pluckthun A. Factors influencing the dimer to monomer transition of an antibody single-chain Fv fragment. Biochemistry. 1998;37:12918–12926.
7. Muyldermans S. Single domain camel antibodies
: current status. Rev Mol Biotechnol. 2001;74:277–302.
8. Arbabi-Ghahroudi M, Tanha J, MacKenzie R. Prokaryotic expression of antibodies
. Cancer Metastasis Rev. 2005;24:501–519.
9. De Meyer T, Muyldermans S, Depicker A. Nanobody-based products as research and diagnostic tools. Trends Biotechnol. 2014;32:263–270.
10. Fridy PC, Li Y, Keegan S, et al. A robust pipeline for rapid production of versatile nanobody repertoires. Nat Methods. 2014;11:1253–1260.
11. De Genst E, Saerens D, Muyldermans S, et al. Antibody repertoire development in camelids. Dev Comp Immunol. 2006;30:187–198.
12. Perez JM, Renisio JG, Prompers JJ, et al. Thermal unfolding of a llama antibody fragment: a two-state reversible process. Biochemistry. 2001;40:74–83.
13. Hassanzadeh-Ghassabeh G, Devoogdt N, De Pauw P, et al. Nanobodies and their potential applications. Nanomedicine (Lond). 2013;8:1013–1026.
14. Omidfar K, Moinfar Z, Sohi AN, et al. Expression of EGFRvIII in thyroid carcinoma: immunohistochemical study by camel antibodies
. Immunol Invest. 2009;38:165–180.
15. Baral TN, MacKenzie R, Arbabi Ghahroudi M. Single-domain antibodies
and their utility. Current Protocols in Immunology
. 2013;103:Unit 2.17.
16. Akl MR, Nagpal P, Ayoub NM, et al. Molecular and clinical significance of fibroblast growth factor 2 (FGF2/bFGF
) in malignancies of solid and hematological cancers for personalized therapies. Oncotarget. 2016;7:44735–44762.
17. Mahajan A. Practical issues in the application of p16 immunohistochemistry
in diagnostic pathology. Hum Pathol. 2016;51:64–74.
18. Skottrup PD, Leonard P, Kaczmarek JZ, et al. Diagnostic evaluation of a nanobody with picomolar affinity toward the protease RgpB from Porphyromonas gingivalis
. Anal Biochem. 2011;415:158–167.
19. Hezel AF, Kimmelman AC, Stanger BZ, et al. Genetics and biology of pancreatic ductal adenocarcinoma. Genes Dev. 2006;20:1218–1249.
20. Bobrow MN, Harris TD, Shaughnessy KJ, et al. Catalyzed reporter deposition, a novel method of signal amplification application to immunoassays. J Immunol Methods. 1989;125(1–2):279–285.
21. Toda Y, Kono K, Abiru H, et al. Application of tyramide signal amplification system to immunohistochemistry
: a potent method to localize antigens that are not detectable by ordinary method. Pathol Int. 1999;49:479–483.
22. von Wasielewski R, Mengel M, Gignac S, et al. Tyramine amplification technique in routine immunohistochemistry
. J Histochem Cytochem. 1997;45:1455–1459.
23. O’Hurley G, Sjostedt E, Rahman A, et al. Garbage in, garbage out: a critical evaluation of strategies used for validation of immunohistochemical biomarkers. Mol Oncol. 2014;8:783–798.
24. Bordeaux J, Welsh A, Agarwal S, et al. Antibody validation. BioTechniques. 2010;48:197–209.