Análisis Ambiental: Las Matrices De Evaluación

A series o f small scale test experiments were carried out to test the hyb ridisatio n characteristics o f octanucleotides, specifically to obtain quantitative data on the detection limit o f hybrid is at io ns and the discrimination o f full / mis-matched hybr idi sa ti on signals.

Sequenced Ml 3 clones wer e kindly provided by Stephan Beck, that had been generated dur ing a sequ enc ing project o f the human MHC class II region on ch ro m o so m e 6 (Beck et al., 1992). These clones were

approximately 1, 0 00 bp long and there were a total o f 1348 used as controls in the oligo hyb ridisation experiments (there were 1,840 clones o f which 1348 had been sequenced at least partially). The human DNA was cloned into the bacter iop hag e vector m p l8, for which specific

primers were designed to amplify the inserts u s in g the waterbath s yst em described above.

In the first test exp eriment f ou r M l 3 clones wer e amplified by PCR, spotted manually in a dilution series onto H y b o n d N+ (Amersham) and h yb rid is ed with two octamers that contained a complete match in two M l 3 clones and mis-matches in the others. The oligonucleotides were radioactively labelled as described in Materials and M eth o d s using [^^P- YJATF. The se test hyb rid is ati on s were carried out in small ‘lunch b o x e s ’ (18 cm X 12 cm), in a volu me o f 20 ml, at a 3 nM oligonucleotide concentration us ing the conditions described in M aterials and M eth o d s

( h y bri dis ati on conditions w er e adapted from (Drmanac et al., 1990b)). F i g u re 4 - 19a s h o w s an aut or adi og rap h o f a hybridi sa ti on in which the clones B a x 7 . B 4 and B a x 7 . C 9 contain a full match, B a x 7 . D 1 0 a single C- C terminal mis-match and B a x 7 . F 4 no match longer than 4 bp. In this h yb rid is ati on the detection limit was around 20 fmol (20 ng o f 1,500 bp DNA) and the signal ratio for full match hybridi sa ti on to end mis-match

a ) fmol target DNA 2 0.4 200 40 20 4 Bax7.B4 Bax7.B9 Bax7.D10 Bax7.F4

b ) fmol target DNA

200 40 20 4 2 0.4

Bax7.B4

Bax7.B9

Bax7.D10

Bax7.F4

Fig. 4-19 Two oligonucleotide test hybridisations carried out on four s e q u e n c e d Ml 3 clones. A) Hybridisation carried out with oligo Bax7/8M1 (CAGGCCGA). B) Hybridisation carried out with oligo Bax7/8M3 (CACAGGCC). Both hybridisations were carried out in SSarc buffer at 5°C and 3 nM probe concentration for 3 hours. Washing was carried out also in SSarc buffer at 12°C for 30 min. Oligonucleotide probes were labelled, by a

3 2

hy br idisation with an octamer that contains full matches in B a x 7 . B 4 and B a x 7 . C 9 and no matches lo nger than 4 bp in either B a x 7 . D 1 0 or B a x 7 . F 4 , s h o w n in figure 4-1 9b, the detection limit was also around 20 fmol. The spots in these hyb rid is ati ons w er e on average 2 mm in diameter ( 3 . 1 4 mm^), which equates to a detection limit o f the ord er o f 6

fmol m m ‘^ for octamer hybr idi sa ti on s. The results o f these two h yb rid is ati ons indicated that seque nce specific DNA hy bridis ation s with octamers was p o s s i b l e u nd er the conditions described.

A detection limit o f 6 fmol mm'^ is sufficient to allow direct spotting o f

P C R prod uct s after amplification us ing the arraying r ob o ts developed in the lab. The yield o f P CR amplifications, u s in g the waterbath system, vary between app roximately 5 - 50 ng fil’^ fo r th ose 75% o f amplifications that are succe ss ful . Spotting pins o f 0 . 4 mm diameter t ra n sf er around 200 nl each, onto an area o f approximately 0. 13 mm’^ so that for a typical PCR pr odu ct a single tra n sf er deposits DNA at a density o f between 8 and 80 fmol per spot, for an average cDNA of

1, 5 0 0 bp. Since this means that at the lower level the DNA density is close to the detection limit, each P CR pr oduct was tra ns f er r ed onto the same sp ot five times, when using the spotting robots. Us in g the latest generation robots in the lab (Maier et al., 1994a) it is p os sib le to array 2 0 , 7 3 6 P C R products onto 12 replica membranes (22 cm X 22 cm) in 6

hou rs. The membranes wer e p r o ce ss ed as described in Materials and M e th o d s.

To characterise the ol ig onucleotide hyb rid is ati on s a furthe r experiment was carried out to estimate the p r o por ti on o f target sites filled during an oligo nu cle ot id e hybridi sa ti on . Three separate mem bra ne strips ( H yb o nd N + , Am ersham) wer e spotted manually with: 1) a serial dilution o f k n o w n amount s o f alkali denatured target DNA alo n g sid e a negative control DNA. 2) purified alkali denatured radioactively labelled DNA of k n o w n concentration. 3) a dilution series o f [^^P-y]ATP. The radioactively labelled DNA was spotted onto tw o strips o f membrane.

a 'm o c k ' hybridisation, that is, an incubation in hybridisation buffer w ithout probe, followed by the same w as h in g as the actual h yb rid is ati ons . The membrane containing the serial dilution of target DNA and negative control, was hybridised with an octanucleotide, w ho s e specific activity had been measured, and washed as described above. The percentage DNA binding to the membrane was calculated as the fraction o f radioactivity remaining on the filter with the radioacively labelled DNA. After hybridisation and wa sh in g the filter containing the target DNA was exposed on one screen together with the strip on which the dilution series o f [^^P-y]ATP was spotted. After scanning the screen using a p ho s ph o ri m ag er (Molecular Dynamics, Sunnyville CA), the ‘Image Quant v3.2* softw ar e form Molecular Dynamics was used to quantify the signal on each of the spots. Figure 4-20 shows a plot of the moles o f [^^P-y]ATP against the pixel counts measured by the ph o s p ho rim ag er . This plot was used to estimate the moles of probe molecules detected at each hybr idi si ng spot. The plot also confirms that the detection o f the p h os ho ri m ag er is linear over 5 orders o f magnitude, as claimed by the manufacturers. In this experiment the percent binding o f the DNA to the nylon membranes was estimated at 50%. At an oligonucleotide concentration o f 5 nM it was estimated that there was approximately one pr obe molecule for every 7 target molecules bo und to each spot (i.e. 14%). This implies that the equilibrium constant (Kcq) for the formation of DNA duplex (probe + target <=> duplex), is small for an octamer hybridisation at 5 nM in 1 M Na"*^ at 5°C and that the equ ilibriu m lies to the left, assumi ng that the reaction has reached eq uilib rium during the hyb ridisation, which in terms o f kinetics is likely in the time given (Wetmur , 1991).

Expe rie nce o f many oligonucleotide hybridisations indicates that the pr o p o rt i o n o f target sites filled varies greatly between oligonucleotides, a conclu sion derived from the variable hybridisation signals obtained from different oligonucleotides hybridised under identical conditions to the same target DNAs.

o

g

I

logCmoles ATP) vs.Jbg(p£aJLVQliim£)L

I I 8 7 6 5 4 - 1 7 - 1 6 - 1 5 - 1 4 log(moles ATP) 3'i Fig. 4 - 2 0 : A pl ot of logio pi xel c o u n t s v e r s u s logio mo l e s of [ ‘ P - y ]

33

ATP. A serial dilution of { P -y ] ATP was sp ot te d onto a nylon membrane (Hybond N+, Amersham) and expos ed to a p h o s p h o r storage screen an room temperature for 1 hour. The image was s c an n ed using a ph o sp h o r imager (Molecular Dynamics) and the signal quantitated using the ImageQuant v3.2 software. An ellipse was drawn around each signal and the signal calculated by integration of the volume of the ellipse.

4.3 Softw are tools for the autom ated analysis o f