Kidney International, Vol. 49 (1996), pp. 261—266

Applications in renal immunopathology of reflection contrast

microscopy, a novel superior light microscopical techniqueFis A. PRINS, JAN A. BRUIJN, and EMILE DE HEER

Department of Patho1oy, University of Leiden, Leiden, The Netherlands

In renal immunopathology subdivision of glomerulopathies isbased upon immunofluorescence patterns on deposition sites ofelectrondense deposits along the glomerular capillary wall and themesangium. However, immunofluorescence has a relatively lim-ited resolution power, while the laborious and expensive tech-nique of electronmicroscopy (EM) allows for inspection of onlyhighly magnified, small areas of tissue. This difference in bothresolution power and sample size calls for a bridging technique,that allows direct comparison of immunohistochemistry at thelight microscopical (LM) level to immuno-EM techniques. Forthese purposes we have applied a novel high-resolution LMtechnique, named reflection contrast microsocopy (RCM), on aseries of experimental models of glomerulopathies. This micro-scope, originally designed by Ploem [1], requires a light micro-scope equipped with an epi-illuminator and specific optics [2, 3],and facilities for plastic embedding and ultrathin sectioningsimilar to those used for EM purposes. In this paper the possibil-ities for application of RCM in renal immunopathology will beoutlined and its perspectives will be discussed.

Methods

Antibodies

The preparation and specificities of antisera against dipeptidylpeptidase Type IV (DPPIV, CD26) and gp330 have been de-scribed previously [4]. Antibodies against rat laminin were ob-tained from a rabbit immunized with rat laminin from the L2sarcoma (Calbiochem, San Diego, CA, USA). Monoclonal anti-body 5D3 (rat MoAb against a distinct epitope on EHS-laminin)has been described elsewere [5—7].

Tissues

Renal tissues were obtained from rats injected with antibodiesagainst gp330, rat L2 laminin, or RTE. Rats with active Heymannnephritis [6] and mice with experimental lupus nephritis inducedby chronic graft-versus-host disease [7] were studied separately(summarized in Table 1).

Received for publication June 14, 1995and in revised form August 21, 1995Accepted for publication August 21, 1995

© 1996 by the International Society of Nephrology

Tissue processing for pre-embedding techniques

Lewis rats were injected intravenously with 0.5 ml polyclonalrabbit anti-gp330 or 0.5 ml anti-laminin sera. After 21 days therats were perfused with PBS and subsequently with periodate-lysine-paraformaldehyde (PLP) for perfusion fixation [8]. Tissueslices of 70 tm thickness were obtained with a Vibratome (OxfordLaboratories, USA) and submerged in PBS. The sections werewashed for 60 minutes with 0.5 M NH4CI in 0.1 M phosphate buffer(to quench free aldehyde groups), and incubated for 16 hours withperoxidase-conjugated swine anti-rabbit IgG, diluted 1:50(DAKO, Glostrup, Denmark). The sections were washed forthree hours in PBS and postfixed in 1% glutaraldehyde in 100 mMcacodylate buffer (1 hr), washed again, and developed withdiaminobenzidine in 5 mrvt Tris HC1 buffer and 0.05% hydrogenperoxide for 30 minutes. The tissue was postflxed for two hours in2% osmium tetra-oxide, dehydrated, and embedded in Epon 812(Fluka Chemie AG, Buchs, Switzerland).

For in vivo detection of HRP-conjugated MoAb, mice withChronic graft-versus-host disease (GvHD) were i.v. injected withMoAb 5D3-HRP [9]. After 24 hours the kidneys were perfusedwith 1% glutaraldehyde in 0.1 M cacodylate buffer. Strips of kidneycortex were incubated on ice for two hours in the same fixativeand 70 m tissue slices were cut with a vibratome. The slices werethen developed for 45 minutes with freshly prepared 0.05%diaminobenzidine and 0.01% hydrogen peroxide at pH 6.0 [10].After the peroxidase reaction, slices were post-fixed with 2%osmium tetra-oxide, dehydrated, and embedded in Epon 812.

Table 1. Experimental models of glomerulonephritis, embedding andstaining techniques

Experimentalmodels Staining technique Development

Passive anti-GBM Pre-embedding EM Peroxidasenephritis

Passive Heymann Pre-embedding EM Peroxidasenephritis

Active Heyniann Post-embedding EM Immuno-gold-silvernephritis

Experimental lupus Post-embedding EM Immunogoldnephritis Post-embedding EM

Post-embedding EMUltracryo EMPost-embedding

Alkaline phosphatasePeroxidasePeroxidaseHematoxylin

261

v\3

B2

4C

262 Prins et a!: Reflection contract microscopy

Prins et a!: Reflection contract microscopy 263

Fig. 1. Linear staining pattern along the GBM of a rat injected with anti-laminin antibodies. Tissue samples were incubated with peroxidase-labeledanti-rabbit IgG and embedded in epon resin. Sequential ultrathin sections were examined by RCM (A, X740) and EM (B, X2040).

Fig. 2. Granular staining pattern of IgG deposition along the GBM of a rat injected with a rabbit antisera against gp330. At five days after injection, tissuesamples of the kidney were incubated with peroxidase-labelcd anti-rabbit IgG and embedded in epon resin. Sequential ultrathin sections were examinedby RCM (A, X740) and EM (B, x2040)

Tissue processing for post-embedding techniquesMouse and rat kidneys were obtained after perfusion fixation

with PLP (periodate-lysine-paraformaldehyde) fixative or with 4%freshly prepared paraformaldehyde and 0.25% glutaraldehyde in0.1 M phosphate buffer, pH 7.4. Dehydration and embedding inLowicryl K4 M (Polysciences, Warrington, PA, USA) were per-formed at 20°C according to Altman, Schneider and Papermaster[11].

For post-embedding immunostaining of ultrathin cryosectionsthe perfusion-fixed pieces of kidney cortex were incubated for 30minutes with 2.3 M sucrose, mounted on a specimen holder, andsnap frozen in liquid nitrogen. Ultrathin sections were cut on aReichert UltracutS/FCS microtome (Leica, Vienna, Austria) ac-cording to Tokuyasu [12].

Sectioning and immunostainingBriefly, ultrathin (60 to 80 nm in thickness) sections were cut

and placed on gelatin-coated or aminosilane pretreated glassslides [2, 3]. For EM examinations the tissue sections werecollected on carbon-coated collodion films on copper grids. Theslides/grids were preincubated in 1% bovine serum albumin(BSA) in PBS, followed by incubation for 16 hours with rabbitanti-rat or rabbit anti-mouse IgG (Nordic, Tilburg, The Nether-lands).

The sections were then washed four times for five minutes withPBS and incubated for two hours with goat anti-rabbit-peroxidase(DAKO), or with goat anti-rabbit ultra-small gold, or with 15 nmgold-conjugated goat anti-rabbit IgG (Aurion, Wageningen, TheNetherlands). Peroxidase was visualized with diaminobenzidine-tetrahydrochloride-H202 for 10 minutes. Silver-enhancement ofultra-small gold conjugates was performed for 10 minutes withR-Gent (Aurion). All slides were washed three times for fiveminutes with distilled water. Some sections were counterstainedwith 0.1% Light Green (LichtgrUn-Gelblich Merck) for 10 min-utes, then washed for one minute with water, and air-dried.

For alkaline phosphatase cytochemistry slides were incubatedfor one hour with goat anti-rabbit-ATPase (Sigma). ATPasedetection was performed as described previously [13]. Briefly, 4mg naphthol-ASMX-phosphate (Sigma) in 5 ml water and 60 mgFast Red TR salt (Sigma) in 5 ml of 0.2 M Tris/HC1 buffer, pH 8.2,were gently mixed and used immediately. The slides were incu-bated with 200 pA of this solution for five minutes at room

temperature, then washed three times for five minutes with PBSand five minutes with water. The slides were overlayed with 10 pAprotein-fixative mixture, containing 20 mg/mI BSA and 2% glu-taraldehyde, and air-dried for 15 minutes. This procedure wasnecessary to prevent dissolution of the ATPase substrate in theimmersion oil [13].

EM sections were stained for 30 seconds with lead citrate [14].Sections for RCM were not counterstained.

Examination of the preparations

Ultrathin sections were collected on either copper grids or glassslides and examined using respectively a Philips CM1O electronmicroscope, operating at 60 Ky, or a Leitz Orthoplan microscope(Leitz, Wetzlar, Germany), equipped for epi-illumination, andadapted for RCM as described [2, 151. The slides were studiedwith immersion oil (refractive index 1.518) between objective andslide (no coverslips were used). Photomicrographs were taken onKodak TMAX-100, black and white film.

Results

Pre-embedding techniques

Demonstration of deposition patterns of immune complexes in theglomerulus. In order to compare deposition patterns of im-munecomplexes along the glomerular capillary wall both anti-GBM and anti-epithelial antibodies were injected into rats andthe deposition patterns of immunoglobulins were compared by IF,RCM, and EM techniques. Polyclonal antibodies against laminin,gp330, and DPPIV were used for the demonstration of linear orgranular deposition staining patterns, respectively.

The RCM image of renal tissue from a rat, injected withanti-laminin antibodies, as shown in Figure 1A demonstrates aclear linear staining pattern of the GBM with high contrast andclear vision of stained and non-stained tissue. Figure lB shows acomplementary immuno-EM view of a sequential ultrathin sec-tion from the same glomerulus without counterstaining. Thecontrast obtained from immunostaining with DAB was higherusing RCM than by using EM.

The RCM image of a rat, injected with anti-gp330 antibodies,(Fig. 2A) demonstrated the granular localization of IgG depositedalong the GBM. Figure 2B shows a complementary immuno-EMview of a sequential ultrathin section from the same glomerulus

Fig. 3. Distribution of laminin in glomeruli during development of gtomerulosclerosis. Mice with chronic graft-versus-host disease were injected withperoxidase-conjugated monoclonal antibodies. Tissue samples were incubated in diaminobenzidine and embedded in epon resin. In vivo boundantibodies were visualized on sequential ultrathin section by both RCM (A, X740) and EM (B, X4700)

Fig. 4. Demonstration of sub-endothelial deposit formation in the glomerulus by anti -DPP IV antibodies. Rats were injected with goat antiserum againstDPP IV. After 24 hours tissue samples from the kidney were snap frozen. The deposition pattern was visualized by immunofluorescence (A). Fixed tissuesamples were incubated with anti-goat Ig-peroxidase and diaminobenzidine-H202 and subsequently embedded in epon resin. Sequential ultrathinsections were examined by both RCM (B) and EM (C). (A and B, X740; C, X21,000)

afl - -, $ ;-- :- .— -a 'a -- a i.r

-4.;

ii

k 6i:- -

Al

264 Prins et at: Reflection contract microscopy

Prins et a!: Reflection Contract microscopy 265

Fig. 5. Demonstration by post-embedding immuno-gold staining for rat IgG of subepithelial immune deposits in the gloinendar capillaty wall from a rat withactive Heyman nepritis at week 10 after induction. The RCM image of IgG deposits was obtained by silver-enhanced immuno-gold staining (A).Transmission EM image of same material (B) demonstrates identical localization, staining was performed with anti-rat IgG conjugated to 15 nm goldparticles. (A )< 740; B X 17,000)Fig. 6. RCM micrograph of an ultrathin kidney section from a mouse with experimental lupus nephritis. After indirect immunogold staining (15 nm goldparticles without silver enhancement) RCM (A) and EM (B) were performed on sequential sections. (A X740; B X13,200)

without counterstaining. Clear subepithelial deposits of IgG canbe seen along the GBM. Because of the high definition of theRCM technique, almost all deposits along the GBM (visible withimmuno-EM) were also visible with RCM.

Binding of anti-DPP IV antibodies in the glomerulus (Fig. 4A)demonstrated a homogeneous staining pattern by IF. However,examination by RCM clearly demonstrated binding of the anti-bodies bound along the GBM (Fig. 4B). A subendothelial local-ization was confirmed by EM (Fig. 4C).

In vivo localization ofperoxidase-conjugated monoclonal antibod-ies. In these experiments mice were injected with peroxidase-conjugated monoclonal antibodies against laminin epitopes inorder to study intramolecular changes of laminin during develop-ment of experimentally induced glomeruloscierosis [9].

As shown in Figure 3A, in this study RCM analysis waspreferred over immuno-EM for testing the effectiveness of severalmonoclonal antibodies because a larger area could be studied andmore antibodies could be investigated with less effort and time.

In particular areas of interest the ultrastructural localizationswere investigated in more detail with sequentially cut ultrathinsections by immuno-EM (Fig. 3B). This allows the examination ofparticular cells or capillary wall areas at high magnification.

Visualization of adjacent tissue in Figures 1,2,3, and 4B and 4Cwas induced by post-fixation with osmium, which was inherent tothe pre-embedding EM technique.

Post-embedding techniques

Detection of localization of proteins by immunogold staining.Figure 5A shows an RCM image of IgG deposits along the GBMof a rat with active Heymann nephritis stained with silver-enhanced immunogold. Figure 5B shows the complementaryimmuno-EM view of a part of the capillairy wall stained with 15nm gold-conjugated antibodies only. Deposit formation in theglomerular capillary wall can be seen both by RCM (Fig. 5A) andEM (Fig. 5B) photomicrographs (arrows).

Comparison of immunogold staining with enzymeimmunohistochemistiy in RCM

Figures 6 to 9 are all examples of kidney sections from micewith chronic graft-versus-host disease, a model for lupus nephritis.Each figure represents a different staining technique on Lowicryl-embedded ultrathin sections;

Figure 6A shows a visualization by RCM of mouse immuno-globulin-containing deposits in glomeruli by indirect immunogold

(15 nm) staining [2, 3]. Note that the background of the tissue isbrighter, which is caused by a gelatine coating of the glass slide.Figure 6B shows a corresponding immuno-EM image with thesame immunogold reagents.

Figure 7A shows RCM images of glomeruli from GvH micestained for mouse immunoglobulins by indirect immunoperoxi-dase (diamino-bezidine as substrate, RCM color: white) [16].Figure 7B shows RCM images of glomeruli from GvH micestained for mouse immunoglobulins with using alkaline phos-phatase-conjugated antibodies (fast red TR salt as a substrate,RCM color: yellow) [13].

Figure 8 demonstrates a GvH mouse-derived ultrathin cryosec-tion stained by indirect immunoperoxidase (diamino-bezidine assubstrate, RCM color: white). Figure 9 represents a hematoxylin-stained ultrathin Lowicryl section from a GvH mouse demonstrat-ing the extremely high resolution, by which cellular organelies canbe visualized. Microvilli, mitochondira and lysosomes can bedistiguished.

Discussion

We here presented a summary of our experiences with RCM inrenal immunopathology, which demonstrates the potentials ofRCM as an analytical tool in renal cell biology. The majoradvantages of RCM in comparison to immunofluorescence anddark field microscopy are obvious: (1) the extremely high resolu-tion at low magnification, (2) the superior sensitivity, and (3) thefact that one can see the non-stained tissue without beingconfused between signal and background. The advantage of theEM in comparison to RCM is its better resolution and magnifyingcapacity. RCM, however, is favorite over EM when a large fieldhas to be examined. Also, the staining contrast is generally higherin RCM than in both immuno-EM and bright field microscopy,which facilitates acurate quantitative morphometric analysis. Pre-liminary experiments using image analysis software have demon-strated inter-measurement variation of less than 0.2% betweensequential ultrathin sections (data not shown). Only smallamounts of label are needed for detection with RCM. This is amajor advantage over bright field microscopy, fluorescence- orepi-polarization microscopy (EPM) [15], where a twofold higheramount of label is needed. Compared to fluorescence microscopya further advantage of RCM is the stability of the label (no fading)and the fact that the non-stained tissue remains visible [17]. Wehave been able to re-examine previously stained sections after oneyear without any loss of brightness of signal or resolution.

Fig. 7. RCM micrographs of a mouse kidney with experimental lupus nephritis after (A) staining with indirect immunoperoxidase, or (B) alkalinephosphatase.conjugated anti-mouse !gG (substrate: fast red TR salt, counterstaining with light-green; x740).Fig. 8. RCM micrograph of an ulrrarhin clyosection of a mouse with experimental lupus nephritis stained for mouse IgG with indirect peroxidase,counterstained with light-green (X 740).Fig. 9. Ultrathin Lowictyl-embedded section of a mouse kidney with experimental lupus nephritis, stained with hematoxilin. Ultrastructural localization ofcellular organelles (microvilli, mitochondria, and lysosomes) can be distinguished (X 1480)

266 Prins et al: Reflection contract microscopy

A significant improvement over bright field microscopy can beobtained in RCM when ultrathin sections are used. This results ina higher image definition. It visualizes the residual tissue andthereby allows a more precise localization of small structures.Small cellular parts and organelles (such as lysosomes, mitochon-dna and microvilli) can be observed (Fig. 9).

Sequential ultrathin sectioning of the same tissue allows thedetection of multiple antigens or epitopes in one particularglomerulus, or even on the same cells or cluster of cells [9, 161.

Sequential sectioning also bridges LM and electron microscopy.After localization of an interesting site, one can perform furtherexaminations on the same cell at higher magnification using EM(Figs. 1 and 2). RCM can be used as a less laborious alternativefor EM. Although the tissue embedding and ultrathin sectioningare not essentally different in labor intensity, the microscopicalexamination in an electron microscope is usually more timeconsuming. Secondly, both the purchase and the maintenance ofan electron microscope are far more expensive than accessoriesfor the adaptation of an epi-illuminator from an fluorescencemicroscope. Every laboratory that has a modern epi-fluorescencemicroscope available can adapt this instrument for RCM whilemaintaining its full use for fluorescence microscopy. Only a fewadditional parts [3] must be acquired (at a moderate cost).Especially in those laboratories where EM facilities are common,RCM can help to bridge LM and EM. This alternative use isjustified after comparison of the RCM image with the EM image(see for example Fig. 3).

Different immunocytochemical labeling techniques can be usedwith RCM, for instance immunogold staining (Figs. 5 and 6).Virtually every enzyme substrate that results in an electron denseproduct in the transmission electron microscope can also bevisualized by RCM (Figs. 7 and 8). Counterstaining of all theseultrathin sections is possible. This results in a clear visibility ofnonimmuno-stained adjacent tissue.

RCM can also be used as an alternative for ultracryo im-muno-EM (Fig. 8). Ultrathin cryo sections are treated as inpost-embedding techniques using Lowicryl-embedded tissues. Incombination with immunoperoxidase there may be some diffusionof substrate in surrounding tissue (Fig. 8). The ultrastructure iscomparable to that of post-embedded material.

For autoradiography using silver emulsions on tissue sectionsafter radioactive in situ hybridization, the silver grains can bevisualized by RCM [18]. Silver grains in emulsions gave surpris-ingly superior images as compared to regular light microscopy(where one can see the tissue, but barely the grains), or dark fieldmicroscopy (where one can see the grain, but not the tissue).

So, in conclusion the applications of this technique are numer-ous, of which several are largely unexplored.

Acknowledgments

Part of these studies were performed with financial support by theDutch Kidney Foundation. The authors thank Dr. SO. Warnaar and Dr.B. Ploem for critical reading of the manuscript, and Drs. C. Kootstra, 0.Veldhuizen and E.H.G. van Leer for supplying renal tissues.

Reprint requests to FransA. Pnns, Faculty of Medicine, Academic HospitalLeiden, Department of Pathology, P.O. Box 9600, L1-Q, 2300 RC Leiden,The Netherlands.

References

1. PLOEM IS: Reflection-contrast microscopy as a tool for investigationof the attachment of living cells to glass suface, in MononuclearPhagocytes in Immunity, Infection and Pathology, edited by VAN FORTHR, Oxford, Blackwell, 1975, pp 405—421

2. CORNELESE-TEN VELDE I, PRINS FA: New sensitive light microscopicaldetection of colloidal gold on ultrathin sections by reflection contrastmicroscopy. Histochemist,y 94:61—71, 1990

3. PRINS FA, VAN DIEMEN-STEENVOORDE R, BONNET J, CORNELESE-TENVELDE 1: Reflection contrast microscopy of ultrathin sections inimmunocytochemical localization studies: A versatile technique bridg-ing electron microscopy with light microscopy. Histochemistmy 99:417—425, 1993

4. VAN LEER EHG, DE Roo GM, HOEDEMAEKER PHJ, DE HEER E:Synergistic effects of anti-gp330 and anti-DPP IV antibodies in theinduction of glomerular damage. Exp Nephrol 1:292—300, 1993

5. ABRAHAMSON DR, IRWIN MH, ST. JOHN PL, PERRY EW, AccAvirnMA, HECK LW, COUCHMAN JR: Selective immunoreactivities ofkidney basement membranes to monoclonal antibodies against lami-nm: Localization of the end of the long and short arms to discretemicrodomains. J Cell Biol 5:3477—3491, 1989

6. HOGENDOORN PCW, BRUIJN JA, VAN DEN BROEK LJCM, DE HEER E,FOIDART JM, HOEDEMAEKER PJ, FLEUREN GJ: Antibodies to purifiedrenal tubular epithelial, antigens contain activity against laminin,fibronectin, and type IV collagen. Lab Invest 58:278—286, 1988

7. BRtJIJN JA, BERGIJK EC, DEl-JEER E, FLEUREN GJ, HOEDEMAEKER PJ:

Induction and progression of experimental lupus nephritis: Explora-tion of a pathogenetic pathway. Kidney lot 41:5—13, 1992

8. MCLEAN 1W, NAKANE PK: Periodate-lysine-paraformaldehyde fixa-tive. A new fixative for immunoelectron microscopy. J HistochemCytochem 22:1077—1083, 1974

9. KOOTsTRA CI, BERGIJK EC, VENINGA A, PRINS FA, De HEER E,ABRAHAMSON DR, BRUIJN JA: Qualitative alterations in lamininexpression in experimental lupus nephritis. Am J Pathol 147:476—488,1995

10. GRAHAM RC, KARNOVSKY MJ: The early stages of absorption ofinjected horseradish peroxidase in the proximal tubulus of mousekidney. Ultrastructural cytochemistry by a new technique. J HistochemCyrochem 14:291, 1966

11. ALTMAN LG, SCHNEIDER BG, PAPERMASTER DS: Rapid embedding oftissue in Lowicryl K4M for immunoelectron microscopy. J HistochemCytochem 32:1217—1223, 1984

12. TOKUYASU KT: Application of cryoultramicrotomy to immunocyto-chemistry. J Microsc 143:139, 1986

13. SPEEL EJM, KRAMPS M, BONNET J, RAMAEKERS CS, HOPMAN AHN:Multicolour preparations for in situ hybridization using precipitatingcytochemistry in combination with reflection contrast microscopy.Histochemistry 100:357—366, 1993

14. REYNOLDS ES: The use of lead citrate at high pH as an electron-opaque stain in electron microscopy. J Cell Biol 111:208, 1963

15. CORNELESE-TEN VELDE I, BONNET J, TANKE HJ, PLOEM JS: Reflection

contrast microscopy performed om epi-illumination microscopestands: Comparison of reflection contrast- and epi-polarization mi-croscopy. J Microscopy 159:1—13, 1990

16. BERGIJK EC, MUNAUT C, BAELDE JJ, PRINS F, FOIDART JM, HOEDE-MAEKER PJ, BRUUN JA: A histologic study of the extracellular matrixduring the development of glomerulosclerosis in murine chronicgraft-versus-host disease. Am J Pathol 140:1147—1156, 1992

17. VAN DER PLOEG M, DUIJN P: Reflection versus fluorescence. Histo-chemistry 62:227—232, 1979

18. BERGIJK EC, BAELDE JJ, DE HEER E, KILLEN PD, BRUIJN JA: The roleof the extracellular matrix in the development of glomerulosclerosis inexperimental chronic serum sickness. Exp Nephrol (in press)