The "background theory"
TOPICS: |1| 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23
FIGURES: | 1 | 2 | 3 | 4 | 5 | 6 |7 | 8 | 9 | 10 | 11 |TABLES: | 1 | 2 | 3 | 4 | 5 | 6 | 7
Background noise in FISH can originate from slide preparation (Fig. 8c-f, Fig. 9a), from improperly prepared or used DNA probes (Fig. 8 and Fig. 9b-d) or from antibodies (Fig. 9e). At least four different types of background can be differentiated:
Type 1. Background located primarily or exclusively on the nuclei and chromosomes (Fig. 9b).
Type 2. Background located primarily among and around the chromosomes and the nuclei but with a relatively "clean" slide surface otherwise (Fig. 9c and d).
Type 3. Background covering the "free" areas of the slide but with relatively clean areas among the chromosomes (Fig. 9e).
Type 4. Background characteristic to aged but non-pretreated slides (Fig. 8d and 8e).
Type 1 background
(on nuclei and chromosomes).
As with any other DNA hybridization assay, the main factors influencing background are: (1) the amount of repetitive sequences of the probe, and the extent to which they are blocked by Cot1. (2) Hybridization temperature (lowering it increases nonspecific binding of the repetitive sequences). (3) The balance between hybridization time and amount of DNA probe. (4) The stringency of the post-hybridization washes. A simple experiment, easy to reproduce in any laboratory, is to label a larger probe (BAC, cosmid) and hybridize it onto several identical slides, then vary some of the washing parameters. Insufficient Cot1 DNA, high amount of probe, low temperature washes or high salt concentration in the wash buffer, all result in an increased background (data not shown). Some of these factors are discussed and illustrated below.
We performed a more complex test, to differentiate some of the factors influencing hybridization efficiency and background appearance. Three different chromosome 1-specific cosmid probes, one labeled with biotin, one with digoxigenin and one with dinitrophenyl were hybridized on several slides. For each hybridization, 1 ?l each labeled cosmid (10-20 ng) were mixed with 200ng human Cot1 DNA (some slides) and 7 ?l hybridization buffer. The mix was denatured 5 minutes at 75° C, then reannealed 30 minutes at 37° C. This very low amount of competitor DNA was chosen so as to allow repetitive sequences to hybridize. Prior to use, slides were aged for 8 hours at 65° C, then pretreated 10 minutes in pepsin (some slides) and denatured 2 minutes at 75° C (some slides). Posthybridization washes were done at 45° C (first wash) and 65° C (second wash), three times 5 minutes each. Antibody detection was done according to the triple color detection scheme detailed in Table 7.
- The first slide was pepsin pretreated and hybridization took place at room temperature, overnight (Fig. 8a, 8f), in the presence of 200ng Cot1. This was the "reference" hybridization, and its results were compared with all other hybridizations.
- To assess the importance of BSA(bovine serum albumin) in blocking unspecific binding of the antibodies during detection, another slide (Fig. 8b) was incubated 15 minutes in 3% BSA prior to antibody detection. 3% BSA was also mixed in every antibody solution in 4xSSC. When compared with Fig. 8a, the background level on nuclei was lower but the signals were also somewhat reduced. It looks as if BSA blocks some of the unspecific binding of the antibodies to the glass, but may also partially block antibodies from reaching the labeled DNA. This and other similar experiments showed that a blocking reagent like BSA was not necessary for regular FISH. Other ways to decrease background were found more effective than BSA (see below).On another slide (Fig. 8c) hybridization was performed at 37 C overnight. Results were indistinguishable from hybridization at room temperature (Fig. 8a).
- Another slides were prepared similarly to the reference, but with no pepsin pretreatment (Fig. 8d and 8e). Hybridization was far superior on the pretreated slide (Fig. 8a, 8f). Non-pretreated slides, after aging, show a relatively strong natural fluorescence of the nuclei and chromosomes, especially when exposed to the blue (DAPI) or green (FITC) filters. A "film" of biological residua (probably cytoplasmic in origin) covers the entire slide (Fig. 9a) and can be best visualized in areas of the slide where it is "cracked" or "scratched". On non-pretreated slides, this film impairs probe accessibility to the nuclei and chromosomes, decreasing hybridization efficiency.
- In another experiment (Fig. 8g), we tested whether a longer denaturing time (6 minutes) influenced hybridization or background. Results did not differ from the reference hybridization. However, when several chromosome painting probes were hybridized onto extensively aged slides (with dry heat plus chemical aging), results were far superior when slide denaturing was done 8 minutes at 94° C, rather than 2 minutes at 75° C (data not shown). On extensively aged slides, hybridization was never optimal.
- In another experiment (Fig. 8h) a shorter hybridization time was used (only three hours at room temperature). There are no visible differences from the reference hybridization (Fig. 8a), performed overnight. This indicates that 10-20 ng of a cosmid probe provides a high enough density of labeled DNA to allow hybridization in just a few hours, and overnight hybridization does not really improve the signal.
- In another experiment (Fig. 8i) a different, variable-temperaturehybridization technique was used: the slide was incubated on the metal block of a thermocycler and subjected to 15 cycles of temperature variation, 10 minutes at 60° C and 10 minutes at 15° C. The low temperature was used in order to allow efficient hybridization whereas the high temperature was used so that it helps "remove" the repetitive DNA or partially matched DNA fragments. This type of hybridization provided best results when compared with any other hybridization shown in Fig. 8. Cosmid signals were strong and background was very low. However, performing such hybridizations routinely can prove to be difficult, as they require a thermocycler. Other factors can be changed to optimize hybridization and yield the same good results.
- In another experiment (data not shown) two commercially available cosmid probes were hybridized overnight onto four slides, using various temperatures: room temperature, 37° C, 50° C and 60° C. Hybridization signals and background were similar but slide morphology deteriorated as the hybridization temperatures were increased. Therefore, hybridization at precisely 37° C is not necessary, and can be replaced with hybridization at any temperature between 20-45 ° C with similar results.
- In the last experiment shown in Fig 8, three areas of the same slide were hybridized with the same probe, but using different amounts of competitor DNA. 1 ?g human Cot1 DNA was used in Fig. 8a, 8j, 10 ?g salmon sperm DNA in Fig. 8k and 1 ?g Cot1 + 10 ?g salmon sperm in Fig. 8l. Results show that 1 ?g Cot1 DNA worked better than 200 ng (Fig. 8a, reference) or 10 ?g of salmon sperm DNA (Fig. 8k). This experiment indicated that Cot1 DNA was a superior blocker of repetitive sequence hybridization compared to salmon sperm DNA. At the same time, the combination of 1?g Cot1 and 10 ?g salmon sperm (Fig. 8l) competed out repetitive sequences even better, indicating that further increase in the total amount of competitor DNA should be beneficial.
Type 1 background prevention
The amount of competitor DNA should be adjusted according to the amount of labeled DNA and the type of probe used (see Table with guidlines). Insufficient amount of Cot1 DNA will increase type I background, by allowing repetitive sequences to bind (Fig. 9b).
Posthybridization washing conditions (especially the stringent SSC wash) should be adjusted, so as to allow the removal of repetitive sequences but preserve the specific signals. Beside the stringency of the SSC buffer, the washing temperature and the duration of the wash are also important.
Type 2 background
(among and around nuclei and chromosomes).
This background is located primarily around chromosomes and nuclei, but with a relatively clean slide surface otherwise (Fig. 9c and 9d). Two main contributing factors are:
- The labeled DNA probe fragments are too long. Long fragments will not completely penetrate the proteinaceous residua surrounding the chromosomes and nuclei and will not be accessible to the target DNA (Fig. 9c). This is shown in another experiment (Fig. 9d) in which two probes, one optimally prepared and the other one composed of long DNA fragments were simultaneously hybridized. The optimally cut probe (red color) hybridized to its target, whereas the other one (green) yielded type 2 background. Long DNA probe fragments will merely "rim" the chromosomes and nuclei (probably binding to loops of DNA protruding out of the chromosomal mass) (Fig. 5i); and
- (2)Poor cell suspension. When a cell suspension is not prepared correctly (usually too short hypotonic pretreatment), there is a lot of residual material surrounding nuclei and chromosomes. Even a good protease pretreatment of the slide will not eliminate this material. During hybridization, fragments of DNA probe will remain "trapped" within this material.
Type 2 background prevention
Proper hypotonic treatment of the cell suspension and a cutting the DNA probe fragments to an average of 200-300 bp will always improve results. Slide preparations should not be excessively aged and should be protease pretreated.
Type 3 background
(on "free" glass areas, not on nuclei or chromosomes)
This type of background may appear on any slide, and is not dependent upon the quality of the cellular suspension or even slide preparation process (Fig. 9e). It is primarily due to an insufficient posthybridization wash (lack of complete removal of the thick hybridization buffer) prior to adding the first antibody, or to a very high antibody concentration. This sometimes happens when small antibody aliquots lose water through evaporation during long storage, so more labeled protein is used by mistake during the detection. In any case, antibodies seem to adhere primarily to the "free" slide surfaces, not covered by metaphases or nuclei (Fig. 9e). As hybridization to the chromosomes and nuclei is not necessarily affected, FISH signals can also be distinguished.
Type 3 background prevention.
Posthybridization washes should be properly done. It is very important to use three or more changes of each washing buffer and, perhaps, place the jar with the slides on a shaker, or to manually shake the jar from time to time, to make sure that slides are properly rinsed. Do not use too high antibody concentrations during detection.
Type 4 background
(on aged, non-pretreated slides).
Two factors contribute to type 4 background:
- (1) Extensively aged, nonpretreated slides. The background is relatively evenly distributed on the entire surface of the slide, with no predisposition for any area of the slide. On such slides, DNA hybridization is less efficient (Fig. 8d, 8e, 9a) because the slide surface is covered by the hardened cytoplasmic residua. As detailed above, probe DNA fragments can be trapped in this proteinaceous residua all over the slide at random, and will be detected by the labeled antibody. However, the non-specific binding of the antibodies to this hardened material covering the slide is usually reduced. The result using non-pretreated slides, will be specs of signals distributed almost evenly all over the slide, at a low density. If the density is high, an additional poor posthybridization washing technique should be suspected. Hybridization signals are reduced in size when compared with pretreated slides (Fig. 8d and 8e show the smallest signals)
- (2): A much heavier background all over the slide may be seen when there is a cross reaction between two antibodies mixed in the same solution during detection. In such cases, the antibodies will precipitate on the slide and will be visualized as a large number of fluorescent specks of different sizes all over the slide (data not shown).
Type 4 background prevention
Slides should not be excessively aged and should be pretreated with a protease.
Attention should be given to proper posthybridization and postantibody washes. If more than one antibody is used in a solution, check for antibody-antibody cross-reactions.
Questions, comments, ideas?
Back to Tavi'sHome Page
Last modified on: Feb12, 2001