0022-1554/92/$3.30 Vol. 40, NO.10,pp. 1613-1618, 1992 Printed in US.A.
The Journal of Histochemistry and Cytochemistry Copyright 0 1992 by The Histochemical Society, Inc.
Brief Report
Nonspecific Binding of Nucleic Acid Probes to Paneth Cells in the Gastrointestinal Tract with In Situ Hybridization’ KERRYN L. GARRETT,2 MIRANDA D. GROUNDS, and MANFRED W. BEILHARZ Departments of Microbiology (KLG,MWB) and Pathology (MDG), University of Western Australias Nedland!, 6009, Australia. Received for publication January 17, 1992 and in revised form May 4, 1992; accepted May 9, 1992 (2B2564).
Nonspecific binding of a number of unrelated nucleic acid probes to cells in the crypts of Lieberkuhn was observed in the small intestine of mice with the in situ hybridization technique. Hybridization signal was localized to cells which, by virtue of their histological position, represented F’aneth cells. This signal could not be removed by RNAse, DNAse, or proteinase K treatment, and was not removed after high-
Introduction A recent report described the localization of tumor necrosis factor messenger RNA (mRNA) to Paneth cells in the gastrointestinalepithelium by in situ hybridization (1). Similarly, others have used in situ hybridization to analyze gene expression and localized mRNA for lysozyme (2), cryptdin (3), serum amyloid A (4), and al-antitrypsin ( 5 ) to Paneth cells in the crypts of Lieberkuhn. Paneth cells were described over a century ago (6). but their precise physiological role in the intestinal epithelium remains unclear. This communicationreports the results of in situ hybridization experiments that used a variety of nucleic acid probes. All eight probes showed nonspecific binding to Paneth cells. We propose that there may be a general nonspecific binding of nucleic acids to Paneth cells and hence a need to interpret in situ hybridization data on these cells with caution.
Materials and Methods Animals. Adult (6-10 week old) Swiss SJL/J and Balb/c mice and Sprague-Dawley rats were obtained from the Animal Resource Centre, Murdoch University, Perth, Western Australia.
Supported by grants from the National Health & Medical Research Council of Australia (MDG,MWB)and by an Australian PostgraduateResearch Award (KLG). Correspondence to Ms. Kerryn L. Garrett, Dept. of Microbiology, Univ. of Western Australia, Nedlands. Perth, Westem Australia 6009, Australia.
stringency washing conditions. This report indicates that caution must be exercised in the interpretation of in situ hybridization data when looking for nucleic acid sequences in the gastrointestinal tract. ( J Hisrochem Cycochem 40: 1613-1618, 1992) KEY WORDS:F’aneth cells; In situ hybridization; Nonspecific hybrid-
ization; Gastrointestinal tract.
Tissue Preparation. Animals were sacrificed by cervical dislocation under fluothane anesthesia and regions of the gastrointestinaltract were dissected and fixed for approximately 5 hr in 4% paraformaldehyde. Tissues were processed through alcohols and chloroform and embedded in paraffin wax.Sections 5 pm thick were cut and mounted on subbed slides (gelatin 0.5%, chromalum 0 . 0 5 % ) and stored at -20% until used. Other tissues were frozen in liquid nitrogen and stored at - 70%. Frozen sections were cut at 5-8 Bm at - 20°C in a cryostat, collected on silanatcd slides, and allowed to air-dry for 30 min. probes. Probes for the skeletal musde-specific gene myogenin, pBU65#7 (anti-sense)and pBU65#ll (sense), were obtained from Dr. W. E. Wright (University of Texas).These represent a 1486 BP cDNA fragment cloned in both orientations into a modified T3/T7 transcription vector (7). A pBU65#7 probe with the c-myc homology sequence removed was prepared by linearizingthe template with PstI instead ofHindIII. The skeletal musclespecific MyoDl probe was obtained from Dr. R. L. Davis (University of Washington) and is a 1785 BP cDNA fragment cloned into a T3/T7 transcriptionvector (8). Sense and anti-sense riboprobes were transcribed from linearized plasmids in the presence of [35S]-UTPor [lzsI]-CTPusing the above templates and a Promega (Madison, WI) riboprobe labeling kit according to the manufacturer’s specifications. Riboprobes were reduced in length by alkaline hydrolysis to about 150 nucleotides. A 171 BP sequence from the 3’ untranslated region of the human a-cardiacactin gene (9) in the pHP34 vector was obtained from Dr. p. Gunning (Camperdown Children’sHospital, New South wales, Australia) and was nick-translated with [35S]IATPusing an Amersham (hole, UK) nicktranslation kit according to the manufacturer’s specifications. A mouse Y chromosome-specificsequence, I45SCS (10). was obtained from Dr. Y.Nishioka (McGill University, Quebec, Canada) and was labeled with digoxigenin by random priming using a Boehringcr Mannheim (Mannheim, Germany) digoxigenin DNA l a b e h kit accordingto the manufacturer‘s specifications. This probe hybridizes to male but not to female nuclei (11).
1613 Downloaded from jhc.sagepub.com at RUTGERS UNIV on June 3, 2015
1614
G m T , GROUNDS, BEILHARTZ
I
Y
Figure 1. Low- and high-power views of nonspecific binding of a radioactive probe to jejunum. Hybridization of 1251-labeledpBU65#7 probe on sections Of @jUnum. (A) Cow-power darkfield view illustrates the accumulation of silver grains in the emulsion over cells at the base of the crypts of Lieberkuhn. (E) Brightfield view at higher power demonstrates that the hybridization is to particular cells within the crypt as shown by arrows. Original magnifications: A x 160; B x 600. Bars: A = 100 pm; E = 25 pm.
A Balblc genomic DNA probe was similarly labeled with digoxigcnin with the same DNA labeling kit.
In situ Hybridization. In situ hybridization was carried out as described in Grounds et al. (12). Paraffin sections used for hybridization with radioactive probes were de-waxed and rehydrated. then post-fixed in 4% paraformddchydc for 30 min. The following pre-treatments were performed at room temperaturewith washes of PBS (pH 7.2)- or DEPC-treated water between each: 0.2N HCI. 20 min; 0.01% Triton X-100,1.5 min; 20 wglml
proteinase K, 15 min (37°C); 4% paraformaldehyde, 30 min; 0.09 M triethanolamine/0.25% acetic anhydride. 10 min. Sections were then dehydrated in a graded series of alcohols and air-dried. RNAsc and DNAsc control sections were treated with 100 pglml RNAse A (Sigma; St Louis, MO) or 50 Ulml DNAse (RNAsc free; Bochringer Mannheim) before the tricthanolaminelacctic anhydride step. Probe was applied to sections at 5 x io4 cpm/pl in hybridization cocktail (50% formami&, 10% dextran sulfate. 20 mM Tris, 5 mM EDTA, 300 mM NaCI, 2 x Denhardt's, 10 mM NaH2P04, 10 mM dithiothreitol, and 250 mglml salmon sperm DNA).
I
Figure 2. Probe binding at various levels of the gastrointestinal tract. Hybridization of '251-labeledpBU65#7 probe on sections of (A) ileum, (E) jejunum, and (C) large intestine ksqder darkfield microscopy. Signal is present in the crypts of Lieberkuhnof the ileum and jejunum but no signal is present in the large intestine. All fields are at the same level of magnification. Original magnification x 160. Bar = 100 pm.
Downloaded from jhc.sagepub.com at RUTGERS UNIV on June 3, 2015
1615
NONSPECIFIC NUCLEI ACID BINDING TO PANETH CELLS
Table 1. Binding of eight diferent nucleic acid probes to various levels of the gastrointestinal tract in mice" Probe
Myogenin (anti-sense) Myogenin (sense) Myogenin (anti-sense minus c-myc homology) MyoDl (anti-sense) MyoDl (sense) Human a-cardiac actin 145SC5 (Y chromosome-specific) Balb/c genomic
Labeling method
Riboprobe 35S, 1251 Riboprobe 35S, I 2 5 I Riboprobc 35S, 12,1 Riboprobc 35S, l2,I Riboprobe 35S, 1251 Nick-translated DNA 35s Random primed DNA (digoxigenin) Random primed DNA (digoxigenin)
Ueum
Jejunum
+ + +
+
Colon
+ + + + + +
+
+
+c
+ , hybridization signal; - lack of hybridization signal. The Y chromosome-specificprobe was analyzed on sections of both male and female ileum. jejunum. and luge intestine. and the signal corresponding to the Paneth cells was present in ileum and jejunum of both male and female tissues. Normal male-specific hybridization was seen at all levels of intestine examined from male mice but not in female mice. Hybridization with the genomic Balb/c DNA probe was present at all levels of the intestines in Balb/c mice; however. the stronger signal corresponding to Pancth cells was observed only in the ileum and jejunum.
Sections were hybridized for about 16 hr at 50°C. Those sections hybridized with 3'S-labeled probes were washed in four changes of 2 x SSC (1 x SSC is 0.15M NaCI, 0.015M trisodium citrate, pH 7.0) containing 0.1% Triton X-100.1 mM EMA. and 5 pM dithiothreitol at G O T over 1 hr, followed by 0.1 x SSC with the same additions for a further 30 min at 60'C. Sections hybridized with '*'I-labeled probes were washed in fivc changes over 5 hr in 0.1 x SSC, 20 mM Tris, 100 pM KI at room temperature, then in 30% formamide, 0.1 x SSC. 20 mM Tris, 100 pM KI at 75'C for 20 min. To remove nonspecifically bound probe, slides were treated with 40 pglml RNAse A (Sigma) for 30 min at 37'C. then washed in six changes of 2 x SSC at G O T over a 1-hr period. Sections were dehydrated in alcohols, air-dried. and coated in NTB-2 emulsion (Eastman Kodak; Rochester, NY)and exposed for 6-7 days before developing. Sections were stained in Giemsa after developing and examined under both brightfield and darkfield microscopy.
Frozen sections used for hybridization with the digoxigenin-labeled probes were post-fixed in 4% paraformaldehyde for 30 min and the following pre-treatments performed at room temperature with washes of PBS (pH 7.2)- or DEPC-treated water between each: 0.2% Triton X-100.5 min; 0.1 M HCI, 20 min; 2 pg proteinase K. 20 min (37°C); 0.1% glycine in PBS (brief rinse); and 4 % paraformaldehyde, 30 min. Sections were prehybridized in hybridization buffer [SO% formamide, 5 x SSC, 0.1% N-lauroylsarcosinc. 0.02% SDS. 5 % (w/v) blocking reagent from kit, and 100 pglml salmon sperm DNA] for 15 min at 37°C. then hybridized with probe at a final concentration of 0.25-0.5 nglpl at 42'C for approximately 16 hr. Before hybridization both the probe and target DNA in the tissue were denatured by heating on an oven hotplate at 100°Cfor 10 min. After hybridization sections were washed for 15 min in two changes of each of the following solutions: 2 x SSC, 1 x SSC. and 0.1 x SSC. all at room temperature. Hybridized probe was detected with the digoxigenin color
Figure 3. Digoxigenin-labeled Y-chromosome-specific probe on male jejunum. Hybridization of the Y chromosome-specific probe 145SC5 labeledwith non-radioactive digoxigenin marker on jejunum of male Balblc mouse. Signal in the Paneth cells was stronger, and the peroxidase color reaction developed more rapidly that the specific sig nal for Y chromosome-specificsequences (shown faintly in most other cells). Signal with the digoxigenin-labeled genomic probe was similar in appearance to the signal shown here for the 145SC5 probe. Original magnification x 600. Bar = 25 wm.
Downloaded from jhc.sagepub.com at RUTGERS UNIV on June 3, 2015
1616
GARRET, GROUNDS, BEILHARTZ
4c
4D
Figure 4. In situ hybridizationcontrol indicate nonspecific nature of probe binding. Hybridization of 1251-labeledanti-sense MyaDl probe to sections of jejunum treatedwith (A) RNAse. (B) DNAse. (C) proteinase K. or (D) washed at high stringency, photographedwith brightfield microscopy.These treatments failed to remove the apparent hybridization signal. All fields are at the same magnification. Original magnification x 600. Bar = 25 pm.
detection system where sections were rinsed in Buffer 1, incubated with a 1:500dilution of antibody conjugate in Buffer 1 for 60 min at room temperature, then washed in Buffer 1 twice for 15 min each. Sections were equilibrated in Buffer 3 and the nitroblue tetrazolium and X-phosphate color substrate solutions were developed in the dark and monitored microscopically until the reaction was stopped with Buffer 4. Sections were stained with 2% methyl green as a nuclear stain.
Results Each of the eight probes examined in this study bound to cells in the jejunum of the gastrointestinal tract of mice. The position of
.
the in situ hybridization signal at the base of the crypts of Lieberkuhn (Figure 1) correlates with the histological location of Paneth cells (13). Furthermore, signal was observed in both the ileum and jejunum but not in the large intestine (Figure 2) with all of the eight different probes, as summarized in Table 1. The binding exhibited by each probe was similar regardless of probe type (DNA or RNA), labeling technique (radioactive or non-radioactive), tissue preparation (frozen or paraffin), isotope (3% or l*SI), detection method (photographic emulsion or antibody/alkalinephosphatase reaction) or mouse strain (SJL/J or Balbk). Similar hybridization was also observed in the intestinal tract of Sprague-Dawley rats (data not shown).
Downloaded from jhc.sagepub.com at RUTGERS UNIV on June 3, 2015
1617
NONSPECIFIC NUCLEI ACID BINDING TO PANETH CELLS
In situ hybridization signal was observed in sections of gastrointestinal tissue of male mice when probed with digoxigenin-labeled Y chromosome-specific probe 145SC5 (see Figure 3), and signal was also present in Paneth cells of female ileum and jejunum but not in the large intestine or in other female tissues (11). Many tissues of Balblc mice probed with the Balbk genomic DNA probe exhibit signal in’ all cells; however, additional signal was localized to Paneth cells in the ileum and jejunum. With both the Y chromosome and genomic DNA probes, the signal corresponding to the Paneth cells could be clearly differentiated from the normal hybridization signal because the Paneth cell binding was more intense and developed faster than the normal hybridization signal. The nature of the binding of these nucleic acid probes was investigated further by treatment of the sections with RNAse, DNAse, or proteinase K before hybridization. These treatments, however, did not alter the signal from that of the original sections (see Fig ure 4). Experiments on muscle tissue run concurrently with those on intestinal tissues indicated that treatment with DNAse and RNAse A efficiently degraded these nucleic acids, since specific hybridization of these probes to target DNA (11) and RNA (12) sequences was prevented. Proteinase K is a commonly used treatment for in situ hybridization, required to expose nucleic acids by digestion of proteins. Omission or variation of this step commonly results in a lack of positive hybridization signal; however, signal in Paneth cells was seen both with and without proteinase K treatment (Figure 4C). In addition, sections were washed at an increased stringency (0.1 x SSC in 30% formamide at 90’C) to remove reversible signal resulting from nucleic acid hybridization, but this treatment also failed to reduce the signal.
bridization alone has been the criterion for localizing a given mRNA to Paneth cells, the result should be independently corroborated by (a) antibody localization, (b) RNAse control, and (c) hightemperature removal of hybridization signal. The nature of the nonspecific binding we have observed is enigmatic. A wide range ofprobes were investigated in this study, including sense and anti-sense probes for skeletal muscle-specific regulatory genes (12), both with and without the region of c- myc homology, a general muscle structuralgene (9). a mouse Y chromosome-specific sequence (10,11), and a genomic DNA probe. The equal binding of this diverse range of probes rules out the possibility of sequence homology, and the different labeling procedures used preclude a methodology dependence. In addition, the failure to abrogate the signal with either RNAse, DNAse, proteinase K, or high-stringency washes indicates that the binding is nonspecific. These data indicate that the binding was not a nucleic acid-nucleic acid hybridization, nor was it associatedwith binding to proteins in the tissue. Since the precise nature of the nonspecific binding of nucleic acids to Paneth cells reported here remains unclear, further work is required to define it. Alternatively,the strong nonspecific binding of nucleic acid probes to Paneth cells could in fact be a useful approach for identlfying these cells in situ.
Literature Cited 1. Keshav S, Lawson L, Chung P, Stein M, Perry VH, Gordon S. Tumor
necrosis factor mRNA localized to Paneth cells of normal murine intestinal epithelium by in situ hybridization. J Exp Med 1990;171:327 2. Chung LP, Keshav S, Gordon S.Cloning the human lysozyme cDNA:
Discussion The present work reports the binding of a multitude of different nucleic acid probes, labeled by various techniques, to cells in both the ileum and jejunum but not in the large intestine of two different mouse strains and in Sprague-Dawley rats. The distribution of the positively labeled cells correlates with the localization of Paneth cells, which are restricted almost completely to the small intestine, extending from the gastroduodenal junction to the ileocecal valve and are usually found in the crypts of Lieberkuhn (14). The number of Paneth cells appears to increase towards the ileum, and although they have been reported in both the stomach and the large intestine in diseased states they are usually not present in these areas in normal tissue (14). Previous reports of “A detection localized by in situ hybridization to Paneth cells in the gastrointestinal tract have identified the expression of certain genes likely to have a physiological role in the intestinal tract (1-5). In some of these experiments there was additional independent evidence to support the presence of these specific transcripts in Paneth cells, in agreement with results from the in situ hybridization experiments. For example, the lysozyme protein has also been localized to Paneth cells with immunofluorescence (15), and Northern analysis of RNA from the small bowel has detected signal for cryptdin (3) and tumor necrosis factor (1). Northern analysis, however, represents a large mixture of cells from a tissue sample and is therefore not definitive for a specific cell type such as Paneth cells. The data presented in this report suggest that where in situ hy-
inverted Alu repeat in the mRNA and in situ hybridisation for macrophages and paneth cells. Proc Natl Acad Sci USA 1988;85:6227
3. Ouellette AJ, Greco RM, James M, Frederick D, Naftilan J, Fallon JT. Developmental regulation ofcryptidin, a corticostatin/defensin precursor mRNA in mouse small intestinal crypt epithelium. J Cell Biol 1989;108:1687 4. Meek RL, Eriksen N, Benditt EP. Serum amyloid A in the mouse: sites of uptake and mRNA expression. Am J Pathol 1989;135:411
5. Koopman P, Povey S, Lovell-Badge RH. Widespread expression of human al-antitrypsin in transgenic mice revealed by in situ hybridisation. Genes Dev 1989;3:16 6. Schwable G. Beiuage zur Kenntnis der Driisen in den Darmwandungen ins besondere der Bruner’schen Driisen. Arch Mikrosk Anat 1872;8:92 7. Wright WE, Sassoon DA, Lin VK.Myogenin, a factor regulating myogenesis, has a domain homologous to MyoD. Cell 1989;56:607 8. Davis RL, Weinuaub H, Lassar AB. Expression of a single transfected cDNA converts fibroblasts to myoblasts. Cell 1987;51:987
9. Gunning P, Mohun T, Ng SX,Ponte P, Kedes L. Evolution of the hu-
man sarcomeric-actingenes: evidencefor units of selection within the 3’ unuanslated regions of the “As. J Mol E d 1984;20:202 10. Nishioka Y. Application of a Y-chromosomespecific repetitivesequence to sexing mouse embryos. Teratology 1988;38:181 11. Grounds MD, Lai MC, Fan Y,CodlingJC, Beilhart MW. Transplantation in the mouse model- the use of a Y-chromosome specific DNA clone to identify donor cells in situ. Transplantation 1991;52:1101 12. Grounds MD. Garrett KL, Lai CM, Wright WE, Beilharz MW. Identification of skeletal muscle precursor cells in vivo by use of MyoDl and myogenin probes. Cell Tissue Res 1992;267:99
Downloaded from jhc.sagepub.com at RUTGERS UNIV on June 3, 2015
1618
GARRET, GROUNDS, BEILHARTZ
13. MadaraJL,TrierJS. Functional morphology of the mucosa of the small
15. Erlandsen SL, Parson JA, Taylor TD. Ultrastructural immunocytochem-
intestine. In Johnson U,ed. Physiology of the gastrointestinal tract. 2nd ed. New York Raven Press, 1987:1209
ical localization of lysozyme in the paneth cells of man. J Histochem Cytochem 1974;22:401
14. Creamer B. Paneth cell function. Lancet 1967;ii:314
Downloaded from jhc.sagepub.com at RUTGERS UNIV on June 3, 2015
The Journal of Histochemistry and Cytochemistry Copyright 0 1992 by The Histochemical Society, Inc.
Brief Report
Nonspecific Binding of Nucleic Acid Probes to Paneth Cells in the Gastrointestinal Tract with In Situ Hybridization’ KERRYN L. GARRETT,2 MIRANDA D. GROUNDS, and MANFRED W. BEILHARZ Departments of Microbiology (KLG,MWB) and Pathology (MDG), University of Western Australias Nedland!, 6009, Australia. Received for publication January 17, 1992 and in revised form May 4, 1992; accepted May 9, 1992 (2B2564).
Nonspecific binding of a number of unrelated nucleic acid probes to cells in the crypts of Lieberkuhn was observed in the small intestine of mice with the in situ hybridization technique. Hybridization signal was localized to cells which, by virtue of their histological position, represented F’aneth cells. This signal could not be removed by RNAse, DNAse, or proteinase K treatment, and was not removed after high-
Introduction A recent report described the localization of tumor necrosis factor messenger RNA (mRNA) to Paneth cells in the gastrointestinalepithelium by in situ hybridization (1). Similarly, others have used in situ hybridization to analyze gene expression and localized mRNA for lysozyme (2), cryptdin (3), serum amyloid A (4), and al-antitrypsin ( 5 ) to Paneth cells in the crypts of Lieberkuhn. Paneth cells were described over a century ago (6). but their precise physiological role in the intestinal epithelium remains unclear. This communicationreports the results of in situ hybridization experiments that used a variety of nucleic acid probes. All eight probes showed nonspecific binding to Paneth cells. We propose that there may be a general nonspecific binding of nucleic acids to Paneth cells and hence a need to interpret in situ hybridization data on these cells with caution.
Materials and Methods Animals. Adult (6-10 week old) Swiss SJL/J and Balb/c mice and Sprague-Dawley rats were obtained from the Animal Resource Centre, Murdoch University, Perth, Western Australia.
Supported by grants from the National Health & Medical Research Council of Australia (MDG,MWB)and by an Australian PostgraduateResearch Award (KLG). Correspondence to Ms. Kerryn L. Garrett, Dept. of Microbiology, Univ. of Western Australia, Nedlands. Perth, Westem Australia 6009, Australia.
stringency washing conditions. This report indicates that caution must be exercised in the interpretation of in situ hybridization data when looking for nucleic acid sequences in the gastrointestinal tract. ( J Hisrochem Cycochem 40: 1613-1618, 1992) KEY WORDS:F’aneth cells; In situ hybridization; Nonspecific hybrid-
ization; Gastrointestinal tract.
Tissue Preparation. Animals were sacrificed by cervical dislocation under fluothane anesthesia and regions of the gastrointestinaltract were dissected and fixed for approximately 5 hr in 4% paraformaldehyde. Tissues were processed through alcohols and chloroform and embedded in paraffin wax.Sections 5 pm thick were cut and mounted on subbed slides (gelatin 0.5%, chromalum 0 . 0 5 % ) and stored at -20% until used. Other tissues were frozen in liquid nitrogen and stored at - 70%. Frozen sections were cut at 5-8 Bm at - 20°C in a cryostat, collected on silanatcd slides, and allowed to air-dry for 30 min. probes. Probes for the skeletal musde-specific gene myogenin, pBU65#7 (anti-sense)and pBU65#ll (sense), were obtained from Dr. W. E. Wright (University of Texas).These represent a 1486 BP cDNA fragment cloned in both orientations into a modified T3/T7 transcription vector (7). A pBU65#7 probe with the c-myc homology sequence removed was prepared by linearizingthe template with PstI instead ofHindIII. The skeletal musclespecific MyoDl probe was obtained from Dr. R. L. Davis (University of Washington) and is a 1785 BP cDNA fragment cloned into a T3/T7 transcriptionvector (8). Sense and anti-sense riboprobes were transcribed from linearized plasmids in the presence of [35S]-UTPor [lzsI]-CTPusing the above templates and a Promega (Madison, WI) riboprobe labeling kit according to the manufacturer’s specifications. Riboprobes were reduced in length by alkaline hydrolysis to about 150 nucleotides. A 171 BP sequence from the 3’ untranslated region of the human a-cardiacactin gene (9) in the pHP34 vector was obtained from Dr. p. Gunning (Camperdown Children’sHospital, New South wales, Australia) and was nick-translated with [35S]IATPusing an Amersham (hole, UK) nicktranslation kit according to the manufacturer’s specifications. A mouse Y chromosome-specificsequence, I45SCS (10). was obtained from Dr. Y.Nishioka (McGill University, Quebec, Canada) and was labeled with digoxigenin by random priming using a Boehringcr Mannheim (Mannheim, Germany) digoxigenin DNA l a b e h kit accordingto the manufacturer‘s specifications. This probe hybridizes to male but not to female nuclei (11).
1613 Downloaded from jhc.sagepub.com at RUTGERS UNIV on June 3, 2015
1614
G m T , GROUNDS, BEILHARTZ
I
Y
Figure 1. Low- and high-power views of nonspecific binding of a radioactive probe to jejunum. Hybridization of 1251-labeledpBU65#7 probe on sections Of @jUnum. (A) Cow-power darkfield view illustrates the accumulation of silver grains in the emulsion over cells at the base of the crypts of Lieberkuhn. (E) Brightfield view at higher power demonstrates that the hybridization is to particular cells within the crypt as shown by arrows. Original magnifications: A x 160; B x 600. Bars: A = 100 pm; E = 25 pm.
A Balblc genomic DNA probe was similarly labeled with digoxigcnin with the same DNA labeling kit.
In situ Hybridization. In situ hybridization was carried out as described in Grounds et al. (12). Paraffin sections used for hybridization with radioactive probes were de-waxed and rehydrated. then post-fixed in 4% paraformddchydc for 30 min. The following pre-treatments were performed at room temperaturewith washes of PBS (pH 7.2)- or DEPC-treated water between each: 0.2N HCI. 20 min; 0.01% Triton X-100,1.5 min; 20 wglml
proteinase K, 15 min (37°C); 4% paraformaldehyde, 30 min; 0.09 M triethanolamine/0.25% acetic anhydride. 10 min. Sections were then dehydrated in a graded series of alcohols and air-dried. RNAsc and DNAsc control sections were treated with 100 pglml RNAse A (Sigma; St Louis, MO) or 50 Ulml DNAse (RNAsc free; Bochringer Mannheim) before the tricthanolaminelacctic anhydride step. Probe was applied to sections at 5 x io4 cpm/pl in hybridization cocktail (50% formami&, 10% dextran sulfate. 20 mM Tris, 5 mM EDTA, 300 mM NaCI, 2 x Denhardt's, 10 mM NaH2P04, 10 mM dithiothreitol, and 250 mglml salmon sperm DNA).
I
Figure 2. Probe binding at various levels of the gastrointestinal tract. Hybridization of '251-labeledpBU65#7 probe on sections of (A) ileum, (E) jejunum, and (C) large intestine ksqder darkfield microscopy. Signal is present in the crypts of Lieberkuhnof the ileum and jejunum but no signal is present in the large intestine. All fields are at the same level of magnification. Original magnification x 160. Bar = 100 pm.
Downloaded from jhc.sagepub.com at RUTGERS UNIV on June 3, 2015
1615
NONSPECIFIC NUCLEI ACID BINDING TO PANETH CELLS
Table 1. Binding of eight diferent nucleic acid probes to various levels of the gastrointestinal tract in mice" Probe
Myogenin (anti-sense) Myogenin (sense) Myogenin (anti-sense minus c-myc homology) MyoDl (anti-sense) MyoDl (sense) Human a-cardiac actin 145SC5 (Y chromosome-specific) Balb/c genomic
Labeling method
Riboprobe 35S, 1251 Riboprobe 35S, I 2 5 I Riboprobc 35S, 12,1 Riboprobc 35S, l2,I Riboprobe 35S, 1251 Nick-translated DNA 35s Random primed DNA (digoxigenin) Random primed DNA (digoxigenin)
Ueum
Jejunum
+ + +
+
Colon
+ + + + + +
+
+
+c
+ , hybridization signal; - lack of hybridization signal. The Y chromosome-specificprobe was analyzed on sections of both male and female ileum. jejunum. and luge intestine. and the signal corresponding to the Paneth cells was present in ileum and jejunum of both male and female tissues. Normal male-specific hybridization was seen at all levels of intestine examined from male mice but not in female mice. Hybridization with the genomic Balb/c DNA probe was present at all levels of the intestines in Balb/c mice; however. the stronger signal corresponding to Pancth cells was observed only in the ileum and jejunum.
Sections were hybridized for about 16 hr at 50°C. Those sections hybridized with 3'S-labeled probes were washed in four changes of 2 x SSC (1 x SSC is 0.15M NaCI, 0.015M trisodium citrate, pH 7.0) containing 0.1% Triton X-100.1 mM EMA. and 5 pM dithiothreitol at G O T over 1 hr, followed by 0.1 x SSC with the same additions for a further 30 min at 60'C. Sections hybridized with '*'I-labeled probes were washed in fivc changes over 5 hr in 0.1 x SSC, 20 mM Tris, 100 pM KI at room temperature, then in 30% formamide, 0.1 x SSC. 20 mM Tris, 100 pM KI at 75'C for 20 min. To remove nonspecifically bound probe, slides were treated with 40 pglml RNAse A (Sigma) for 30 min at 37'C. then washed in six changes of 2 x SSC at G O T over a 1-hr period. Sections were dehydrated in alcohols, air-dried. and coated in NTB-2 emulsion (Eastman Kodak; Rochester, NY)and exposed for 6-7 days before developing. Sections were stained in Giemsa after developing and examined under both brightfield and darkfield microscopy.
Frozen sections used for hybridization with the digoxigenin-labeled probes were post-fixed in 4% paraformaldehyde for 30 min and the following pre-treatments performed at room temperature with washes of PBS (pH 7.2)- or DEPC-treated water between each: 0.2% Triton X-100.5 min; 0.1 M HCI, 20 min; 2 pg proteinase K. 20 min (37°C); 0.1% glycine in PBS (brief rinse); and 4 % paraformaldehyde, 30 min. Sections were prehybridized in hybridization buffer [SO% formamide, 5 x SSC, 0.1% N-lauroylsarcosinc. 0.02% SDS. 5 % (w/v) blocking reagent from kit, and 100 pglml salmon sperm DNA] for 15 min at 37°C. then hybridized with probe at a final concentration of 0.25-0.5 nglpl at 42'C for approximately 16 hr. Before hybridization both the probe and target DNA in the tissue were denatured by heating on an oven hotplate at 100°Cfor 10 min. After hybridization sections were washed for 15 min in two changes of each of the following solutions: 2 x SSC, 1 x SSC. and 0.1 x SSC. all at room temperature. Hybridized probe was detected with the digoxigenin color
Figure 3. Digoxigenin-labeled Y-chromosome-specific probe on male jejunum. Hybridization of the Y chromosome-specific probe 145SC5 labeledwith non-radioactive digoxigenin marker on jejunum of male Balblc mouse. Signal in the Paneth cells was stronger, and the peroxidase color reaction developed more rapidly that the specific sig nal for Y chromosome-specificsequences (shown faintly in most other cells). Signal with the digoxigenin-labeled genomic probe was similar in appearance to the signal shown here for the 145SC5 probe. Original magnification x 600. Bar = 25 wm.
Downloaded from jhc.sagepub.com at RUTGERS UNIV on June 3, 2015
1616
GARRET, GROUNDS, BEILHARTZ
4c
4D
Figure 4. In situ hybridizationcontrol indicate nonspecific nature of probe binding. Hybridization of 1251-labeledanti-sense MyaDl probe to sections of jejunum treatedwith (A) RNAse. (B) DNAse. (C) proteinase K. or (D) washed at high stringency, photographedwith brightfield microscopy.These treatments failed to remove the apparent hybridization signal. All fields are at the same magnification. Original magnification x 600. Bar = 25 pm.
detection system where sections were rinsed in Buffer 1, incubated with a 1:500dilution of antibody conjugate in Buffer 1 for 60 min at room temperature, then washed in Buffer 1 twice for 15 min each. Sections were equilibrated in Buffer 3 and the nitroblue tetrazolium and X-phosphate color substrate solutions were developed in the dark and monitored microscopically until the reaction was stopped with Buffer 4. Sections were stained with 2% methyl green as a nuclear stain.
Results Each of the eight probes examined in this study bound to cells in the jejunum of the gastrointestinal tract of mice. The position of
.
the in situ hybridization signal at the base of the crypts of Lieberkuhn (Figure 1) correlates with the histological location of Paneth cells (13). Furthermore, signal was observed in both the ileum and jejunum but not in the large intestine (Figure 2) with all of the eight different probes, as summarized in Table 1. The binding exhibited by each probe was similar regardless of probe type (DNA or RNA), labeling technique (radioactive or non-radioactive), tissue preparation (frozen or paraffin), isotope (3% or l*SI), detection method (photographic emulsion or antibody/alkalinephosphatase reaction) or mouse strain (SJL/J or Balbk). Similar hybridization was also observed in the intestinal tract of Sprague-Dawley rats (data not shown).
Downloaded from jhc.sagepub.com at RUTGERS UNIV on June 3, 2015
1617
NONSPECIFIC NUCLEI ACID BINDING TO PANETH CELLS
In situ hybridization signal was observed in sections of gastrointestinal tissue of male mice when probed with digoxigenin-labeled Y chromosome-specific probe 145SC5 (see Figure 3), and signal was also present in Paneth cells of female ileum and jejunum but not in the large intestine or in other female tissues (11). Many tissues of Balblc mice probed with the Balbk genomic DNA probe exhibit signal in’ all cells; however, additional signal was localized to Paneth cells in the ileum and jejunum. With both the Y chromosome and genomic DNA probes, the signal corresponding to the Paneth cells could be clearly differentiated from the normal hybridization signal because the Paneth cell binding was more intense and developed faster than the normal hybridization signal. The nature of the binding of these nucleic acid probes was investigated further by treatment of the sections with RNAse, DNAse, or proteinase K before hybridization. These treatments, however, did not alter the signal from that of the original sections (see Fig ure 4). Experiments on muscle tissue run concurrently with those on intestinal tissues indicated that treatment with DNAse and RNAse A efficiently degraded these nucleic acids, since specific hybridization of these probes to target DNA (11) and RNA (12) sequences was prevented. Proteinase K is a commonly used treatment for in situ hybridization, required to expose nucleic acids by digestion of proteins. Omission or variation of this step commonly results in a lack of positive hybridization signal; however, signal in Paneth cells was seen both with and without proteinase K treatment (Figure 4C). In addition, sections were washed at an increased stringency (0.1 x SSC in 30% formamide at 90’C) to remove reversible signal resulting from nucleic acid hybridization, but this treatment also failed to reduce the signal.
bridization alone has been the criterion for localizing a given mRNA to Paneth cells, the result should be independently corroborated by (a) antibody localization, (b) RNAse control, and (c) hightemperature removal of hybridization signal. The nature of the nonspecific binding we have observed is enigmatic. A wide range ofprobes were investigated in this study, including sense and anti-sense probes for skeletal muscle-specific regulatory genes (12), both with and without the region of c- myc homology, a general muscle structuralgene (9). a mouse Y chromosome-specific sequence (10,11), and a genomic DNA probe. The equal binding of this diverse range of probes rules out the possibility of sequence homology, and the different labeling procedures used preclude a methodology dependence. In addition, the failure to abrogate the signal with either RNAse, DNAse, proteinase K, or high-stringency washes indicates that the binding is nonspecific. These data indicate that the binding was not a nucleic acid-nucleic acid hybridization, nor was it associatedwith binding to proteins in the tissue. Since the precise nature of the nonspecific binding of nucleic acids to Paneth cells reported here remains unclear, further work is required to define it. Alternatively,the strong nonspecific binding of nucleic acid probes to Paneth cells could in fact be a useful approach for identlfying these cells in situ.
Literature Cited 1. Keshav S, Lawson L, Chung P, Stein M, Perry VH, Gordon S. Tumor
necrosis factor mRNA localized to Paneth cells of normal murine intestinal epithelium by in situ hybridization. J Exp Med 1990;171:327 2. Chung LP, Keshav S, Gordon S.Cloning the human lysozyme cDNA:
Discussion The present work reports the binding of a multitude of different nucleic acid probes, labeled by various techniques, to cells in both the ileum and jejunum but not in the large intestine of two different mouse strains and in Sprague-Dawley rats. The distribution of the positively labeled cells correlates with the localization of Paneth cells, which are restricted almost completely to the small intestine, extending from the gastroduodenal junction to the ileocecal valve and are usually found in the crypts of Lieberkuhn (14). The number of Paneth cells appears to increase towards the ileum, and although they have been reported in both the stomach and the large intestine in diseased states they are usually not present in these areas in normal tissue (14). Previous reports of “A detection localized by in situ hybridization to Paneth cells in the gastrointestinal tract have identified the expression of certain genes likely to have a physiological role in the intestinal tract (1-5). In some of these experiments there was additional independent evidence to support the presence of these specific transcripts in Paneth cells, in agreement with results from the in situ hybridization experiments. For example, the lysozyme protein has also been localized to Paneth cells with immunofluorescence (15), and Northern analysis of RNA from the small bowel has detected signal for cryptdin (3) and tumor necrosis factor (1). Northern analysis, however, represents a large mixture of cells from a tissue sample and is therefore not definitive for a specific cell type such as Paneth cells. The data presented in this report suggest that where in situ hy-
inverted Alu repeat in the mRNA and in situ hybridisation for macrophages and paneth cells. Proc Natl Acad Sci USA 1988;85:6227
3. Ouellette AJ, Greco RM, James M, Frederick D, Naftilan J, Fallon JT. Developmental regulation ofcryptidin, a corticostatin/defensin precursor mRNA in mouse small intestinal crypt epithelium. J Cell Biol 1989;108:1687 4. Meek RL, Eriksen N, Benditt EP. Serum amyloid A in the mouse: sites of uptake and mRNA expression. Am J Pathol 1989;135:411
5. Koopman P, Povey S, Lovell-Badge RH. Widespread expression of human al-antitrypsin in transgenic mice revealed by in situ hybridisation. Genes Dev 1989;3:16 6. Schwable G. Beiuage zur Kenntnis der Driisen in den Darmwandungen ins besondere der Bruner’schen Driisen. Arch Mikrosk Anat 1872;8:92 7. Wright WE, Sassoon DA, Lin VK.Myogenin, a factor regulating myogenesis, has a domain homologous to MyoD. Cell 1989;56:607 8. Davis RL, Weinuaub H, Lassar AB. Expression of a single transfected cDNA converts fibroblasts to myoblasts. Cell 1987;51:987
9. Gunning P, Mohun T, Ng SX,Ponte P, Kedes L. Evolution of the hu-
man sarcomeric-actingenes: evidencefor units of selection within the 3’ unuanslated regions of the “As. J Mol E d 1984;20:202 10. Nishioka Y. Application of a Y-chromosomespecific repetitivesequence to sexing mouse embryos. Teratology 1988;38:181 11. Grounds MD, Lai MC, Fan Y,CodlingJC, Beilhart MW. Transplantation in the mouse model- the use of a Y-chromosome specific DNA clone to identify donor cells in situ. Transplantation 1991;52:1101 12. Grounds MD. Garrett KL, Lai CM, Wright WE, Beilharz MW. Identification of skeletal muscle precursor cells in vivo by use of MyoDl and myogenin probes. Cell Tissue Res 1992;267:99
Downloaded from jhc.sagepub.com at RUTGERS UNIV on June 3, 2015
1618
GARRET, GROUNDS, BEILHARTZ
13. MadaraJL,TrierJS. Functional morphology of the mucosa of the small
15. Erlandsen SL, Parson JA, Taylor TD. Ultrastructural immunocytochem-
intestine. In Johnson U,ed. Physiology of the gastrointestinal tract. 2nd ed. New York Raven Press, 1987:1209
ical localization of lysozyme in the paneth cells of man. J Histochem Cytochem 1974;22:401
14. Creamer B. Paneth cell function. Lancet 1967;ii:314
Downloaded from jhc.sagepub.com at RUTGERS UNIV on June 3, 2015
