Article Text

Download PDFPDF

Immunohistochemistry of deparaffinised sections using antigen retrieval with microwave combined pressure cooking versus immunofluorescence in the assessment of human renal biopsies
  1. Suozhu Shi1,
  2. Ping Zhang2,
  3. Qingli Cheng3,
  4. Jie Wu1,
  5. Jing Cui1,
  6. Ying Zheng1,
  7. Xue-Yuan Bai1,
  8. Xiangmei Chen1
  1. 1Department of Nephrology, State Key Laboratory of Kidney Diseases, Chinese PLA General Hospital, Beijing, China
  2. 2Department of Pathology, Wang Jing Hospital, Chinese Academy of Chinese Medical Sciences, Beijing, China
  3. 3Department of Geriatric Nephrology, Chinese PLA General Hospital, Beijing, China
  1. Correspondence to Professor Xue-Yuan Bai and Xiangmei Chen, Department of Nephrology, State Key Lab of Kidney Diseases, Chinese PLA General Hospital, 28 Fuxing Road, Beijing 100853, China; xueyuan_bai{at}yahoo.com.cn, xmchen301{at}126.com

Abstract

Background Immunofluorescence of frozen tissue sections (IF-F) is a traditional technique used in renal biopsy. However, IF-F has certain disadvantages, such as a few or even no glomeruli in the section, and limited long-term preservation of the fluorescently labelled samples.

Methods We compared two-step immunohistochemistry (IHC) staining of deparaffinised sections for antigen retrieval with microwave combined high-pressure cooking to IF-F used to detect antigens of IgG, IgA, IgM, C3, C1q, κ and λ in patient renal biopsy samples. The number of glomeruli detected, sensitivity and specificity of positive staining, tissue structure, and location staining of the antigens were determined using the two methods in 285 patients diagnosed with different renal diseases.

Results Concordant observations between IF-F and IHC were 99% for all antigen staining (1969 of 1995 observations) and 100% for IgG, IgA and IgM (all 285 observations). The number of glomeruli in IHC sections was significantly greater compared with IF-F sections (p<0.001). IHC provided clearer images of tissue structure, more precise localisation of positive-staining antigens, and IHC staining allowed simultaneous evaluation of tissue by light microscopy. Correlation between tissue structure and immune deposits are not readily attained by IF-F.

Conclusions IHC is superior to IF-F for immunopathological diagnosis of renal biopsy tissue and is a reliable replacement for the more traditional IF-F method.

  • Antigen Retrieval
  • Diagnosis
  • Immunohistochemistry
  • Kidney
  • Microwave

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction

Several methods are currently used in the pathological examination of renal biopsy tissue. These include light microscopy, immunopathology and electron microscopy. Immunopathological examination is indispensible in the diagnosis of renal diseases, and direct immunofluorescence in frozen tissue sections (IF-F) is one of the more traditional methods used for the detection of immunoglobulin and complement accumulation in renal tissue.

Although IF-F is convenient, rapid and widely available, it has certain inherent shortcomings in clinical practice, including reduced accuracy of pathological diagnosis due to few or absent glomeruli, and difficulty in long-term preservation of fluorescently labelled samples, rendering them unsuitable for retrospective studies.

Recently, antigen retrieval with protease digestion in formalin-fixed and paraffin-embedded (FFPE) tissues has been used for immunopathological analysis. However, samples from renal biopsy are small and the sections are thin, and enzyme digestion often causes tissue destruction and antigen loss, sometimes lowers the positive detections rates, sometimes increases the positive detection rate, and even leads to false positive results.1–4

Bánkfalvi et al5 reported that microwave heating for antigen retrieval was effective for immunohistochemical staining in FFPE tissue sections. Kumada et al6 applied microwave antigen retrieval on FFPE sections for the detection of 178 different routine clinical antigens by immunohistochemical staining. Their results showed enhanced specificity of antigen–antibody reactions and a reduction in non-specific reactions. The immunohistochemical results of Rhodes et al7 on FFPE sections confirmed that optimal staining results relied on microwave antigen retrieval. Cristina et al8 found that antigen retrieval by high-pressure cooking demonstrated a specificity of 100% in the diagnosis of κ-type or λ-type non-Hodgkin's lymphoma. In detection of myosin-V by immunohistochemistry (IHC) using formalin-fixed brain tissue from bees, Calabrin et al9 found that antigen retrieval by high-pressure steam cooking gave satisfactory results.

In this study, IHC of FFPE tissue with antigen retrieval through microwave heating combined with high-pressure cooking was compared with IF-F using 285 renal biopsies obtained from patients with various renal diseases. The sensitivity and specificity of antigen detection for several antibodies, including IgG, IgA, IgM, C3, C1q, κ and λ were determined.

Materials and methods

Renal tissue samples

A total of 285 renal biopsy samples were obtained from the Department of Nephrology, Chinese People's Liberation Army (PLA) General Hospital, Beijing. Eight common renal diseases were evaluated, including 76 cases of membranous nephropathy (MN), 66 of lupus nephritis, 62 of IgA nephropathy (IgAN), 17 of acute postinfectious glomerulonephritis, 18 of membranoproliferative glomerulonephritis (MPGN), 19 of antiglomerular basement membrane (GBM) nephritis and 27 of primary amyloidosis (PA). First we conducted IF test for all patients. The confirmed eight common renal diseases patients by IF were enrolled in this study, and were then compared with IHC results.

Ultrasound-guided renal biopsy samples were collected percutaneously using an automated biopsy gun (BARD, USA). The samples were divided into three parts on a wax plate: (1) frozen tissue, embedded by optimal cutting temperature (OCT) compound (Tissue-Tek; Sakura Finetek Europe BV, Zoeterwude, The Netherlands) and preserved in liquid nitrogen, was sectioned by Cryomicrotome (Leica CM3050S, Germany) and used in immunofluorescence staining; (2) an appropriate number of specimens, fixed in 10% formaldehyde, were evaluated by optical microscopy; and (3) a small amount of cortical tissue (1–2 mm in diameter), fixed in 2.5% glutaraldehyde, was used for transmission electron microscopy examination.

Management of Pathological Analysis

Three pathologists in our department have read the slides of the 285 biopsies. The IF and IHC slides of the same patient were separately read by two of the pathologists. Before reading the IF or the IHC slides, they do not know that the slide is the same patient's. In order to enhance the accuracy rate of diagnosis, these pathologists read the IF or the IHC slides twice blindly. To obtain good reproducibility, we process renal biopsy specimens in strict accordance with standardised procedures at every step of IHC.

Light microscopy

Tissue samples fixed in formaldehyde were dehydrated sequentially using 50%, 75%, 85%, 95% and 100% ethanol and then clarified with chloroform for 30 min. The samples were immersed and embedded in paraffin (Beijing BHKT Clinical Reagent Co., LTD, melting point: 56–58°C). Twenty consecutive sections (1–2 μm thick) were sliced with a paraffin microtome (Leica RM2255, Germany). The sections were incubated at 60°C for later evaluation. The sliced samples were dewaxed by xylene and rehydrated sequentially using 100, 95, 85 and 75% ethanol-water. All renal biopsy samples were stained with H&E, periodic acid-Schiff, periodic acid-methenamine silver and Masson methods. Pathological diagnosis was made according to the WHO pathological standards for glomerular diseases.

IF-F

Fresh renal biopsy samples were embedded in OCT compound (Tissue-Tek; Sakura Finetek Europe) and sliced consecutively by Cryomicrotome (Leica CM3050S, Germany) to a thickness of 3–4 μm. The sections were air-dried at room temperature for 20 min and washed twice with 0.01 M phosphate buffer solution (PBS) (pH 7.2–7.4, 5 min/wash). Fluorescein isothiocyanate (FITC)-labelled rabbit antihuman IgG, IgA, IgM, C3, C1q, κ and λ polyclonal antibodies (1 : 40 dilutions) were added separately on top of the tissue and incubated at room temperature for 30 min, followed by washing twice in PBS (3 min/wash). The samples were mounted in glycerol and examined by fluorescent microscopy (Nikon 80i, Japan). The negative control should consist of replacing the primary antibody with PBS. Controls for negatively stained tissues were obtained from kidneys of living donors.

IHC

Primary antibodies of rabbit antihuman IgG, IgA, IgM, C3,C1q, κ and λ antibodies (1 : 40 dilutions; Dako) and indirect two-step staining method (with the EnVision Dual Link System-horseradish peroxidase (HRP) reagent, Code K4065, Dako) were used in this study. A rotary paraffin microtome (Leica RM2255, Germany) was used to section the samples (1–2 μm). The sections were collected and placed on polylysine-coated slides and incubated in a 60°C oven for 60 min. After dewaxing with xylene, rehydrating with 100% (5 min, twice), 95, 85 and 75% (5 min each) ethanol, and washing with PBS, the slides were placed in EDTA antigen-repairing solution (0.01 M, pH 8.0). The slides were then placed in a microwave oven (850 W; Sharp, Japan) until the temperature of the antigen-repairing solution reached 100°C for 10 min. The solution was then cooled down to room temperature for 20 min, and the slides were washed with PBS for 5 min. The slides were then placed in citrate antigen-repairing solution (0.01 M, pH 6.0) and heated in a high-pressure cooker until steam arose. The slides were kept inside the cooker for 2 min, and then cooled off at room temperature for 20 min, followed by washing in PBS for 5 min. The activity of endogenous peroxidase was blocked by soaking the slides in peroxidase-blocking solutions (S2023; Dako) at room temperature for 10 min, followed by washing in PBS for 5 min. The slides were incubated with primary antibody at 4°C overnight, and after washing in PBS (5 min), the slides were incubated with the EnVision Dual Link System-HRP (Code K4065, Dako) at room temperature for 30 min. After washing with PBS for 5 min, the slides were transferred to hydrogen peroxide/3-3-diaminobenzidine tetrahydrochloride solution for colouration for 5 min. The slides were counterstained with Mayer haematoxylin for 2 min and mounted with a cover slide. For the negative controls, primary antibody was replaced by PBS. Negatively stained tissues were obtained from kidneys of living donors. Sections were examined by optical microscopy (Nikon 80i, Japan).

Analysis of positive staining area

The positively stained area was determined using image analysis software (NIS-ELEMENTS BR, Nikon, Japan). In each case, the IgG, IgA, IgM, C3, C1q, κ and λ positive areas in five complete glomeruli, as well as the entire area containing the glomeruli, were evaluated. The ratio of the two areas represented the relative content of immunoglobulin, complement, κ and λ in the glomeruli.

Statistical analysis

Statistical analysis was performed using SPSS V.17.0 software. The ratio of glomerular counts and positively stained areas was expressed as mean±SD, and the statistical significance between IF-F and IHC was compared by t-test. In addition, the average difference was given with CIs in addition to the p values. Diagnostic tests were used to determine the concordance, specificity, sensitivity, positive predictive value, negative predictive value and likelihood ratios of IHC.

Results

Number of glomeruli

As shown in table 1, the average number of glomeruli equalled 27.99±2.10 on slides stained by IHC, and 6.40±0.53 on slides stained with IF-F in all glomerular diseases. There is no relationship between the number of glomeruli detected and the type of diagnosis. The results were similar in different glomerular diseases.

Table 1

The numbers of glomeruli detected by IF-F and IHC

Positive staining rates

The IgG, IgA, IgM, C3, C1q, κ and λ positive regions revealed by IHC were consistent with those revealed by IF-F in the renal tissue samples for all eight renal diseases. Concordance for observations between IF-F and IHC was 99% for all antigen staining (1969/1995 observations), 100% for IgG, IgA and IgM (all 285 observations), 98% for C3 (280/285 observations), 96% for C1q (274/285 observations), 98% for κ (279/285 observations) and 99% for λ (281/285 observations). The sensitivity, specificity, positive predictive value, negative predictive value and likelihood ratios of IHC are presented in table 2.

Table 2

Renal Biopsy Specimens Stained for IgG, IgA, IgM, C3, C1q, κ and λ by IF-F and IHC

Our results suggest that IHC (in FFPE sections with antigen retrieval through microwave heating combined with high-pressure cooking) was able to diagnose lupus nephritis, MN, IgAN, acute postinfectious glomerulonephritis, MPGN, anti-GBM glomerulonephritis and PA.

Positively stained areas

Areas positively stained for IgG, IgA, IgM, C3, C1q, κ and λ were calculated and analysed by image analysis software for all eight renal diseases. The average number of positively stained areas did not differ significantly between the two methods in lgAN, MPGN and PA disease (p>0.05) (table 3).

Table 3

The positive distribution areas of IgG, IgA, IgM and C3 by IF-F and IHC

Antigen location

IHC staining was visualised by light microscopy. In IgAN, IgA primarily accumulated in the mesangial and paramesangial regions. In MN, IgG accumulated on the glomerular capillary walls. In lupus nephritis, immunoglobulin and complement were deposited mainly on glomerular capillary walls. In anti-GBM nephritis, IgG linear deposition was found in the GBM.

Antigens were fixed and preserved effectively by fixation, dehydration, clarification and paraffin embedding. In addition, the sections were thin (1–2 μm thick). As compared with IF-F, IHC of FFPE sections provided clearer tissue structure and more accurate antigen location, and at the same time, light microscopic observation was performed as a reference. For IF-F staining, sections were observed by fluorescent microscopy. Due to the lack of preprocessing of frozen tissue and the sections being thicker than paraffin sections (3–4 μm), the tissue structure was unclear and the antigen location was diffused. Thus, the tissue structure and antigen location revealed by IHC were superior to those of IF-F (figures 1 and 2).

Figure 1

Representative renal tissue images from light microscope with periodic acid-Schiff stain (PAS) staining and immunohistochemistry (IHC) staining of microwave pressure cooker retrieval of formaldehyde-fixed and paraffin-embedded renal tissues and immunofluorescence staining of frozen tissue sections (IF-F). (A1): PAS staining of a patient with membranous nephropathy (MN); (A2) IHC of the same patient, with IgG granules (brown) at the capillary loops of the glomeruli; (A3) IF-F of the same patient, with IgG granules (green) at the capillary loops of the glomeruli. A1 and A2: consecutive sections; ×400. (B1) PAS staining of a patient with IgA nephropathy (IgAN); (B2): IHC of the same patient, with patchy IgA granules (brown) lateral to the mesangial region; (B3) IF-F of the same patient, with patchy IgA granules (green) lateral to the mesangial region. B1 and B2: consecutive sections; ×400. (C1): PAS staining of a patient with primary amyloidosis (PA); (C2): IHC of the same patient, mesangial deposits of κ (brown) in a patient with PA, amyloid protein, light chain derived (AL) κ- type; (C3): IF-F of the same patient, mesangial deposits of κ (green) in a patient with PA, AL κ type. C1 and C2: consecutive sections; ×400. (D1): PAS staining of a patient with lupus nephritis (LN, type IV+V); (D2): IHC of the same patient, with IgG granules (brown) at the capillary loops of the glomeruli; (D3): IF-F of the same patient, with IgG granules (green) at the capillary loops of the glomeruli. D1 and D2: consecutive sections; ×400.

Figure 2

Representative renal tissue images from light microscope with PAS staining and immunohistochemistry (IHC) staining of microwave pressure cooker retrieval of formaldehyde-fixed and paraffin-embedded renal tissues and immunofluorescence staining of frozen tissue sections (IF-F). (a1): PAS staining of a patient with membranous nephropathy (MN); (a2) IHC of the same patient, with C3 granules (brown) at the capillary loops of the glomeruli; (a3) Negative control for IHC staining of the same patient; (a4) IF-F of the same patient, with C3 granules (green) at the capillary loops of the glomeruli. a1,a2 and a3: consecutive sections; ×400. (b1) PAS staining of a patient with antiglomerular basement membrane (GBM); (b2): IHC of the same patient, with IgG linear (brown) deposition was found in the GBM; (b3) Negative control for IHC staining of the same patient; (b4) IF-F of the same patient, with IgG linear (green) deposition was found in the GBM. b1,b2 and b3: consecutive sections; ×400. (c1): PAS staining of a patient with lupus nephritis (LN, type IV+V); (c2): IHC of the same patient, with C3 granules (brown) at the capillary loops of the glomeruli; (c3) Negative control for IHC staining of the same patient; (c4): IF-F of the same patient, with C3 granules (green) at the capillary loops of the glomeruli. c1, c2 and c3: consecutive sections; ×400.

Discussion

In the early 1950s, the IF-F technique was applied to the diagnosis of renal disease by renal pathologists using renal biopsy tissue. Since then, IF-F has been the gold standard in examining renal biopsies, and results have been satisfactory.10 However, due to the drawback of absent or low numbers of glomeruli in samples examined using IF-F, it often fails to meet the requirements of a pathological diagnosis.

Although satisfactory results are obtained using IHC on FFPE sections following protease digestion without pretreatment of frozen tissues,11 antigen retrieved by IHC methods with enzyme digestion results in high background, false-positive staining, tissue destruction and antigen loss, which are yet to be resolved. Howie et al3 and Furness and Boyd12 reported that the degree of protease digestion is difficult to control. Insufficient digestion results in inadequate antigen exposure, which lowers positivity rates; in contrast, overdigestion leads to the destruction of tissue integrity and antigen loss. Howat et al13 demonstrated that the detection of IgG linear accumulation in renal biopsies of anti-GBM nephritis by protease digestion was inaccurate and/or defective. Molne et al4 showed the total consistency was only 70–80% when comparing the IHC method combined with enzymatic digestion antigen retrieval with IF-F, and the sensitivity was only 73–86%. In five cases of anti-GBM nephritis, IgG linear accumulation was detected in two cases by the IHC method, and the other three cases were read as negative.

Other studies showed negative results or lower abundance of C3 antigen when the immunofluorescence technique was applied to FFPE sections, compared with frozen tissue sections.14 ,15 Thus, it is a traditional point of view that FFPE sections are traditionally regarded as being unsuitable for immunofluorescence staining due to formation of a tight complex between protein and calcium and other divalent ions during formaldehyde fixation, blocking antigen epitopes.16–18 Thus, it was necessary to develop a method that could overcome the shortcomings of IF-F and IHC methods with enzyme digestion in FFPE sections.

In recent decades, antigen retrieval techniques have improved greatly. In particular, microwave oven heating and high-pressure cooking have been proved to be effective in repairing FFPE tissues.5–9 Shi et al19 showed that microwave heating significantly increased antigen staining in paraffin sections, which might be due to the enhanced exposure of intracellular antigens and recovery of cell viability. Studies have shown that microwave-radiation treatment of paraffin-embedded sections for immunohistochemical staining provides good tissue morphology, and effectively preserves the antigens that might be easily disrupted during the paraffin embedding process.20–26 Weakly stained antigens by routine antigen retrieval techniques show strong staining using this method, which significantly enhances the detection rate.27–29

In the present study, the IgG, IgA, IgM, C3, C1q, κ and λ positive regions revealed by novel IHC were consistent with those revealed by IF-F in the renal tissue samples for all eight renal diseases. Concordance between observations with IF-F and IHC were 99% for all antigen staining (1969/1995 observations), 100% for IgG, IgA and IgM (all 285 observations), 98% for C3 (280/285 observations), 96% for C1q (274/285 observations), 98% for κ (279/285 observations) and 99% for λ (281/285 observations). The consistency rate was 100% for the detection of IgG and IgA in MN, lupus nephritis and IgAN. For the detection of IgG, C3 and C1q in acute postinfectious glomerulonephritis, MPGN, lupus nephritis and anti-GBM nephropathy, results from IHC were also consistent with the positive staining in IF-F, which suggested that the sensitivities of IHC for C3 and C1q detection were comparable with that of IF-F. In the 19 cases of anti-GBM nephropathy, IHC staining showed linear IgG accumulation, whereas Molne et al4 and Howat et al13 reported linear IgG accumulation in few cases of anti-GBM nephropathy using an IHC method with enzyme-digestion antigen retrieval.

Our results suggested that the antigen retrieval method of combined microwave heating and high-pressure cooking enhanced the specificity of the antigen–antibody reaction, and lowered the non-specific expression level. In addition, we found that microwave heating used for primary antigen retrieval in paraffin sections, followed by further antigen retrieval using a high-pressure cooker, could overcome the shortcoming related to uneven heating by microwaves, and thus allowed for optimal antigen exposure. Moreover, besides antigen retrieval, microwave radiation of tissue sections also resulted in further fixation of the tissue, therefore providing maximum preservation of the antigen present. As a result, the detection rate using the combined heating antigen-retrieval method was much higher than that of antigen retrieval by protease digestion.

Studies have shown that microwave radiation treatment of FFPE tissue sections for enzyme-labelled immunohistochemical staining reduced the staining time, and more importantly, enhanced its quality.30–35 Due to the dual effects of fixation and antigen retrieval of microwave radiation on tissue sections, plus the homogeneous heating by high-pressure cooker, the two retrieval methods avoided the tissue destruction and reduced detection efficacy caused by enzymatic digestion. Therefore, the combination of microwave heating and high-pressure cooking for antigen retrieval of FFPE sections demonstrated the combined advantages of IF-F and IHC; that is, the high sensitivity and low background staining associated with IF-F and the clear tissue structure and accurate antigen location associated with IHC. About the scientific proof that the combined method improves antigen retrieval, this is because during the process of antigen retrieval, the mechanism of microwave heating is from inside to outside and the mechanism of pressure cooking is outside to inside. The combined heating for the samples with the two methods may play a complementary action, overcome an uneven defect emerging from microwave heating, and is helpful for sufficient antigen exposure and increased positive rate.

Compared with IF-F, IHC has the advantages of thinner sections, clearer tissue structure and more accurate antigen location, and is less limited by sample volume. In addition, the light microscopy and IHC results for the same tissue section are comparable. FFPE tissue samples can also be archived for further retrospective evaluation. Moreover, IHC does not require expensive cryomicrotomy and fluorescent microscopy, which reduces the cost significantly. Microwaves combined with high-pressure cooking for antigen retrieval enhances specific staining and reduces background. Thus, IHC with the two combined heating techniques is an important breakthrough in antigen retrieval methods.

Pathological diagnosis of renal biopsy tissue is based on the pathological changes in a certain number of glomeruli therein. That is, the diagnosis is made by inferring the condition of the entire kidney based on the observation of a small, localised area. Thus, the renal samples used in pathological diagnosis must have a sufficient number of glomeruli. Previous studies have shown that the accuracy in determining the pathological status of all glomeruli using samples containing five glomeruli was only 65%; however, the accuracy reached 95% using samples containing 15 glomeruli.36 Thus, it is best to have at least 15 glomeruli in the samples for examination by optical microscopy and immunofluorescence. In the present study, the average number of glomeruli was 27.99±2.10 for the IHC method, whereas the IF-F method produced only 6.40±0.53 glomeruli. Thus, the IHC method provides more comprehensive information, and thus, is more suitable for accurate diagnosis of renal pathology. In addition, long-term preservation of IHC-staining results is possible, which facilitates retrospective analyses. The fluorescence in IF-F staining is easily quenched and thus unsuitable for long-term preservation or retrospective studies.

In conclusion, our results indicate that the detection of immunoglobulin, complement, κ and λ accumulation by novel IHC (with combined microwave heating and high-pressure cooking for antigen retrieval of FFPE sections) may be used as a complement for immunopathological diagnostic technique in renal biopsy samples.

Acknowledgments

The English in this manuscript has been checked and edited by at least two professional editors, both native speakers of English. For a certificate, please see: http://www.textcheck.com/certificate/VoGRY3.

References

Footnotes

  • SS and PZ contributed equally.

  • Contributors All authors contributed significantly to the paper, and all authors are in agreement with the content of the manuscript.

  • Funding This work was supported by a grant (No. 2011CBA01003) from the National Basic Research Program of China (973 Program) to XYB and a grant (No. 2011CB964904) from the National Key Scientific Program of China to XYB

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval The Medical Ethics Committee of Chinese PLA General Hospital.

  • Provenance and peer review Not commissioned; externally peer reviewed.