J . Forens. Sci. Soc. (1979), 19, 59
Receizjed 30 November, 1978
ABO, Gm and Km Grouping of Bloodstains on Microtitre Plates M. J. DAVIE Home Ofice Central Research Establishment, Aldermaston, Reading, Berkshire, RG7 4 P N , England The use of microtitre plates for ABO, Gm and Icm typing of bloodstains using an extraction technique is shown to be a cheap and reliable alternatiue to methods currently in use. For stains up to one-month old a minimum of 5mm of stained cotton thread is suficient for ABO grouping with a further IOmm being needed for G m l , 2 , 1 0 and K m l typing. The A and B antigens were detectable in stains up to 24years old but the H antigen could not be detected reliably in stains over twoyears old. G m l and Gm2 were detected in stainr up to oneyear old and K m l in stains up to six-months old. However, GmlO was not detected in stains more than one-month old.
Introduction Kind (1960) and Outteridge (1962) described methods for ABO grouping of bloodstains by absorption-elution, using aqueous stain extracts. Kind and Cleevely (1969) modified Kind's earlier technique by using 5% ammonia as a n extractant. I n parallel with these extractive techniques, single thread methods were developed by Nickolls and Pereira (1962) and Howard and Martin (1969). While the extractive methods are simple and reliable, their main shortcoming would seem to be the need for larger amounts of stain than those used with single thread techniques. Using microtitre plates we have attempted to determine the limits of an ammonia extraction technique with regard to the amount of stain needed and the age of stain for ABO grouping. I n addition, with the increasing importance of Gm and K m typing of bloodstains in forensic serology (Khalap et al., 1976), it was considered worthwhile to undertake studies on Gm/Km typing using similar bloodstain extracts as for the ABO grouping, taking advantage of the convenience and sensitivity of microtitre methods. Parallel ABO and Gm/Km typing were therefore performed on 56 bloodstains. The use of microtitre plates for Gm typing was described by Bore1 et al., (1967). Using a similar technique we have determined the limits of sensitivity for the Gm typing of bloodstains. Materials V-bottom (for Gm/Km) and flat-bottomed (for ABO) microtitre plates were purchased from Dynatech Laboratories (Daux Road, Billingshurst, Sussex). Anti-A and Anti-B sera were purchased from Lorne Laboratories (17 Halkingcroft, Langley, Berkshire) and anti-H was prepared as described by Boorman and Dodd (1966). Anti-Gm and Km reagents were purchased from Molter, West Germany (Dr Molter GMBH, Postfach 10 40 49, 6900 Heidelberg 1). Bloodstains up to 24-years-old were available from the bloodstain collection at the Home Ofice Central Research Establishment. Stains eight-years-old or more were on filter paper, while more recent stains were on clean white cotton cloth. Methods ABO Grouping Varying amounts of stained thread (from 1-20mm) were extracted in four
drops of 5% ammonia (5ml a880 Ammonia, 95ml dist. H,O) in a flat-bottomed microtitre plate well for one hour at room temperature (14-16°C). The thread was then removed and the extract distributed in three wells of the microtitre plate for A, B and H testing. The extract was dried in the wells at 56OC (1-2 hours). Anti-sera were then added to the wells. Various absorption times between three and eighteen hours at 4°C were used. After absorption the anti-sera were removed by inverting the plate and giving it a flick over a sink. The wells were then filled with saline at 4°C and emptied in a similar way three times. The wells were filled a fourth time with saline and left at 4OC for 30 minutes. This step was repeated to give a total wash time of about one hour. After the final wash as much saline as possible was flicked out of the wells and then one drop of 2% suspensions of A,, B and 0 cells was added to the relevant wells. Elution was performed a t 56°C for 15 minutes after which the plates were placed on a "Rotatest" for 30 minutes. Plates were examined microscopically ( x 40) or with a hand-lens. Scoring was as described by Pereira (1971). The effect of stain age on the detection of ABO grouping was investigated in stains up to 24-years-old. For this purpose 20sq mm of stains on filter paper or 7sq mm of stains on cotton cloth were extracted in three drops of 5% ammonia. It was recognized that a great excess of material was being used on the more recent stains on cotton cloth.
Gm/Km Grouping Sensitisation of Cells. Group O R,R, cells were bled by fingerprick for sensitisation. After four saline washes one drop of packed cells and one drop of coating anti-D were incubated for 13 hours at 37OC, then washed four times before use. Cells were used at about 0.1 % suspension strength. Anti-serum Titration. Anti-sera titres were determined by using doubling dilutions of sera, one drop of sensitised cells being added to one drop of each dilution in the wells of a V-bottom micro-titre plate. Dilutions from 114 to 11512 were tested. 19 hours incubation on a Rotatest followed by one minute centrifugation at 1000 rpm made microscopic interpretation of results very easy. Anti-sera were found to have useful working dilutions between 1/64 and 11256. Care must be exercised as the titre in saline may be much higher than in the presence of a stain extract. This problem can be avoided by performing titres in the presence of a stain extract negative for the factor the antiserum is detecting, i.e., each well contains the aritiserum dilution, the stain extract and sensitised red cells. Bloodstain testing. Each stain was tested using varying lengths of thread (2mm, 5mm, lOmm and 20mm). The thread was placed in the wells and one drop of 5% Ammoilia added. The plate was shaken on a Luckhams vibrator for ten minutes and one drop of diluted antiserum added to the wells. The plates were left for 1& hours at 4°C for inhibition to take place and then the threads were removed and one drop of appropriate cell suspension added. The plates were rotatested for 1 hours for agglutination to take place and then centrifuged for one minute at 1000 rpm before reading. Plates were first read microscopically using a low magnification ( x ~ O ) placed , on a sloping fronted (ca. 20" from vertical) light box and read using the streak method described by Bore1 et al., (1967), and then checked again microscopically. Typical positive and negative results are shown in Figure 1.
Results ABO Grouping Size of Stain. Stains up to four-weeks-old were tested. All stains were made on clean white cotton cloth from which blood-stained threads were teased. Threads were cut up to provide 1, 2, 5 , 10 and 20mm lengths and extracted and tested as previously described. Typical results are given in Table 1.
1:igurc 1. Gm Grouping on a microtitre plate. Streak reading method. Typical positive and negative reactions.
It is quite clear from these results that the A and B antigens are readily detectable in l m m of stained thread. This is the equivalent of about 0 . 1 ~ 1of liquid blood. The H antigen on the other hand required 5mm of thread ( 0 . 5 ~ 1 of liquid blood), this length giving good H reactions in A,'s and B's as well. T o be sure of getting reactions for H, 5mm of thread would be the minimum amount that could be used in a practical technique in the laboratory. It may well be necessary to increase the amount of stain extracted in the case of older stains. Age of Stain. The A antigen in A, and A, stains was detected consistently in all stains tested up to five-years-old. Older stains showed more variability, with one 24-year-old A, stain giving a good positivc, while a seven-year-old A, stain gave no reaction. In 16 A, stains tested, none gave any reaction with anti-H while no A, stain over two-years-old gave an H reaction. The B antigen in B stains was clearly detected in stains up to 22-years-old although one ten-yearold stain gave a very weak reaction. The H antigen in B's, as with A,'s was not detected in stains more than two-years-old. The H antigen proved by far the least reliable, as with the previous work on stain size, and while the occasional positive reaction was obtained with 0 ' s up to 20-years-old, no consiste~cycould be achieved in stains more than twoyears-old regardless of ABC) grcup. An increase in the amount of extract in the H wells or an increase in albumin in the 0 cell suspension (Lang, 1976) may TABLE I T H E EFI:ECT OF STAINED THREAD LENGTH ON ABO GROUPING IN MICROTITRE PLATES Slain Croufi
Length ( m m )
-4 I
1 2 51020
Scoring as in Prrcita (1971 ).
1
A, 2 51020
B 1
2 51020
1
0 2 51020
improve matters and will be investigated further. Therefore with the method used as described, the limits for reliable detection of antigens would seem to be five years for A, nine years for B and two years for H, although these limits could probably be extended by increasing the amount of material under test.
GmlKm Grouping The results for stains positive for factors Gml, 2, 10 and K m l up to one-yearold are given in Table 2. TABLE 2 T H E EFFECT O F VARYING STAINED THREAD LENGTH O N TYPING GM/KM POSITIVE BLOODSTAINS U P T O ONE-YEAR OLD
Correct w i n g s ' 4 Gm2+ GmlO+
Gml+
Kml+
Length of thread (mm)
Number of individual stains tested
9.5
ion
4.7
44
20
8
14
9
Twenty individual G m l + stains were tested at four thread lengths, the 5% (i.e., one stain) which failed to react at the two shorter thread lengths was one-year-old. The eight Gm2f stains gave no false negatives. For GmlO and Kml the amount of stain required for correct grouping increased with age, 2mm only being adequate for stains up to one-month-old. Nine G m l - , six GmlO - and 16 K m l - stains all grouped as negative for all four lengths of thread. In 2 1 Gm2 - stains 16 were negative in all tests but five gave apparent inhibition for a single thread length only (1 a t 2mm and 4 a t 5n:m). Table 3 shows the actual results for a plate with false Gm2 positives TABLE 3 GM2 TYPING RESULTS SHOWING FALSE POSITIVE REACTIONS
Length of thread (mm) 2 5
A
B
*
Donor C D 3 3
-
E 1
-*
F 4
-*
-.
20 Known Gm2 Type * False inhibitions
These false positives may have been due to the presence of large numbers of small fibres shed by the threads in these wells interfering with agglutination. Four hundred and twenty-eight individual tests were performed. Only five falsc positives (ca. 1%) were seen, all being in Gm2 typing. Where stains were known to be positive for any given factor, all but one stain proved completely typeable although older stains required up to 20mm of thread (see Table 2). For stains up to one-month-old, lOmm of thread was adequate for all four markers. The one stain that could not be typed was two years old and failed to react for GmlO and K m l but gave a positive for Gml as was expected.
ABO and GmlKm typing in parallel Fifty-six stains, all about two-months-old, were tested simultaneously on the same extract for ABO and G m l , 2, 10 and K m l . 30mm (i.e., twice the mini-
mum) of stained thread was extracted in 12 drops of 5% ammonia. A breakdown of the ABO results is given in Table 4. All reactions with anti-A were TABLE 4 SYNOPSIS OF ABO GROUPING RESULTS Known Grout,
AI A2 A (subtype unknown) B 0
Number 13 9 4 4 26
Anti-A 13 9 4 0 2
Positive with Anti-B 0 1 0 4 0
Anti-H 0 8
2 4 26
-
56
scored as 4 except the two false positives on the Group 0 ' s which were 2 and 3 respectively. These two and the false B result were on the same plate, the wash time of which had been cut to half an hour. Subsequent plates given one hours wash gave no false positives and this wash procedure must be adhered to. All B reactions were graded 4 while reactions with H were more variable. I t is notable that eight out of nine known A,'s gave reactions with anti-H while none of the 13 A,'s gave any reaction for H , this probably being a property of the batch of anti-H in use. All stains were tested for G m l , 2, 10 and K m l . While G m l , 2 and K m l proved very reliable, GmlO was less so. O n occasions GmlO grouping failed totally but the failures gave no special cause for concern as the controls also failed to react. The lower levels of IgG3 which carries the GmlO marker and the generally lower titre of anti-Gml0 (ca. 1/64) as compared with Gml and 2 (ca. 11256) make testing for this antigen a littlc morc difficult. I n a total of 224 single tests (56 x 4), no false negatives and one false positive were seen.
Discussion Microtitre plates have been in use in microbiology and other fields for over ten years. Their use in blood group serology has been rather less well documented. I n forensic serology the only published work is by Alfultis (1965) on a technique for ABO grouping bloodstains using an inhibition method. In 1967 Bore1 et al. described a microtitre method for Gm/Km typing of serum which, with minor modifications, has proved equally effective for bloodstain extracts. There seemed to be no reason why ABO grouping of similar extracts by an elution technique should not be equally successful and the results given here have confirmed this. The limits of sensitivity in terms of amount of stain and age of stain have been explored for ABO and Gm/Km typing, somewhere between 5,ul and lop1 of whole blood as stain being required for both tests on stains up to one month old. For older stains an increase in the amount of stain extracted will help. The greatest problem encountered in this work was the comparatively poor performance for detection of GmlO. This marker is on the y3 heavy chain of IgG3 which represents only 7 4 % of the total IgG (Grubb, 1970) in normal plasma. O n the other hand, Gml and Gm2 are IgGl markers present on 70-80% of the IgG. This near 10-fold difference may, in part, explain the relative difficulty of detecting GmlO, especially in older stains. I n Caucasians and Mongoloids the marker Gm4 (= Gm3) has the same distribution as GmlO and is carried on the IgG 1 molecules. Preliminary experiments with this marker indicate that it may well be detectable in much the same way as Gml and Gm2, that is, in stains up to one-year-old. A problem with Gm4 is its
lack of correlation with GmlO in Negro bloods, most (> 90%) negroes being G m l + 4 - lo+. Although GmlO may give problems, its use with Gm4 can provide useful indications of the ethnic origin of a bloodstain (Ropartz et al., 1966: Grubb. 1970). In terms of economics, microtitre would seem to have some advantages. Each 96-well microtitre plate can be used for 32 ABO determinations or 12 Gm/Km tests using G m l , 2, 10 and K m l and full controls. They compare well in price with cellulose acetate or polycarbonate sheet, as used in the technique of Howard and Martin (1969), one microtitre platc bcing the equivalent of 2-3 cellulose acetate sheets in terms of the number of tests which can be performed. Two other areas where savings can be made are in the fact that moist chambers for incubation are unnecessary, as used plates can be used as lids for plates under test, and large beakers of cold saline are unnecessary as each well can be filled with saline for the washing step. However, perhaps the biggest day-to-day saving can be made by the use of anti-Gm and K m reagents at high dilutions. Our current batches of antisera are diluted as follows : Gml, 2 and K m l 1/64 and GmlO 1/32. These dilutions will of course vary from batch to batch and more markedly between batches from different suppliers, one batch tested being usable at 1/100. I n conclusion, the use of microtitre plates for ABO, Gm and K m typing of bloodstains would seem to be a cheap and reliable alternative to methods currently in use. We have not undertaken to determine the usefulncss of these techniques as applied to the more varied substrates and conditions of stains normally met with in forensic science, although, so far there have been no problems with fresh stains on glass, tissue, newspaper and a wide variety of man-made and synthetic fibres (King, 1978). References ALFULTIS, H. M., 1965, J. Forens. Sci., 10, 319. BOORMAN, K. E. and DODD,B. E., 1966, An Introduction to Blood Group Serology, Churchill, London. BOREL,H., PRYCE,S. and ALLEN,F. H., 1967, Vox Sang., 12, 319. GRUBB, R., 1970, Genetic Markers of Human Immunoglobulins, Chapman and Hall, London. HOWARD, H. D. and MARTIN,P. D., 1969, J. Forens. Sci. Soc., 9, 28. KHALAP,S., PEREIRA, M., and RAND,S., 1976, Med. Sci. Law, 16, ( I ) , 40. KIND,S. S., 1960, Nature, 185, 397. KIND,S. S. and CLEEVELY, R . M., 1969, J . Forens. Sci. Soc., 9, 131. KING,L., 1978, Personal Communication. L A ~ T B. G , G., 1976, J. Forens. Sci. Soc., 16, 61. NICKOLLS, L. D. and PEREIRA, M., 1962, Med. Sci. Lam, 2, 172. OUTTERIDGE, R. A., 1962, Nature, 194, 385. PEREIRA,M., iri 1971, "The Examination and Typing of Bloodstains in the Crime Laboratory", by B. J. Culliford, U.S. Dept of Justice Law Enforcement Assistance Administration, U.S. Government Printing Office, Washington C.C. 20402. ROPARTZ,C., GOLD,E. R., RIVAT,L., ROUSSEAU, P. Y., 1966, Transfusion (Paris), 9, 293.
Receizjed 30 November, 1978
ABO, Gm and Km Grouping of Bloodstains on Microtitre Plates M. J. DAVIE Home Ofice Central Research Establishment, Aldermaston, Reading, Berkshire, RG7 4 P N , England The use of microtitre plates for ABO, Gm and Icm typing of bloodstains using an extraction technique is shown to be a cheap and reliable alternatiue to methods currently in use. For stains up to one-month old a minimum of 5mm of stained cotton thread is suficient for ABO grouping with a further IOmm being needed for G m l , 2 , 1 0 and K m l typing. The A and B antigens were detectable in stains up to 24years old but the H antigen could not be detected reliably in stains over twoyears old. G m l and Gm2 were detected in stainr up to oneyear old and K m l in stains up to six-months old. However, GmlO was not detected in stains more than one-month old.
Introduction Kind (1960) and Outteridge (1962) described methods for ABO grouping of bloodstains by absorption-elution, using aqueous stain extracts. Kind and Cleevely (1969) modified Kind's earlier technique by using 5% ammonia as a n extractant. I n parallel with these extractive techniques, single thread methods were developed by Nickolls and Pereira (1962) and Howard and Martin (1969). While the extractive methods are simple and reliable, their main shortcoming would seem to be the need for larger amounts of stain than those used with single thread techniques. Using microtitre plates we have attempted to determine the limits of an ammonia extraction technique with regard to the amount of stain needed and the age of stain for ABO grouping. I n addition, with the increasing importance of Gm and K m typing of bloodstains in forensic serology (Khalap et al., 1976), it was considered worthwhile to undertake studies on Gm/Km typing using similar bloodstain extracts as for the ABO grouping, taking advantage of the convenience and sensitivity of microtitre methods. Parallel ABO and Gm/Km typing were therefore performed on 56 bloodstains. The use of microtitre plates for Gm typing was described by Bore1 et al., (1967). Using a similar technique we have determined the limits of sensitivity for the Gm typing of bloodstains. Materials V-bottom (for Gm/Km) and flat-bottomed (for ABO) microtitre plates were purchased from Dynatech Laboratories (Daux Road, Billingshurst, Sussex). Anti-A and Anti-B sera were purchased from Lorne Laboratories (17 Halkingcroft, Langley, Berkshire) and anti-H was prepared as described by Boorman and Dodd (1966). Anti-Gm and Km reagents were purchased from Molter, West Germany (Dr Molter GMBH, Postfach 10 40 49, 6900 Heidelberg 1). Bloodstains up to 24-years-old were available from the bloodstain collection at the Home Ofice Central Research Establishment. Stains eight-years-old or more were on filter paper, while more recent stains were on clean white cotton cloth. Methods ABO Grouping Varying amounts of stained thread (from 1-20mm) were extracted in four
drops of 5% ammonia (5ml a880 Ammonia, 95ml dist. H,O) in a flat-bottomed microtitre plate well for one hour at room temperature (14-16°C). The thread was then removed and the extract distributed in three wells of the microtitre plate for A, B and H testing. The extract was dried in the wells at 56OC (1-2 hours). Anti-sera were then added to the wells. Various absorption times between three and eighteen hours at 4°C were used. After absorption the anti-sera were removed by inverting the plate and giving it a flick over a sink. The wells were then filled with saline at 4°C and emptied in a similar way three times. The wells were filled a fourth time with saline and left at 4OC for 30 minutes. This step was repeated to give a total wash time of about one hour. After the final wash as much saline as possible was flicked out of the wells and then one drop of 2% suspensions of A,, B and 0 cells was added to the relevant wells. Elution was performed a t 56°C for 15 minutes after which the plates were placed on a "Rotatest" for 30 minutes. Plates were examined microscopically ( x 40) or with a hand-lens. Scoring was as described by Pereira (1971). The effect of stain age on the detection of ABO grouping was investigated in stains up to 24-years-old. For this purpose 20sq mm of stains on filter paper or 7sq mm of stains on cotton cloth were extracted in three drops of 5% ammonia. It was recognized that a great excess of material was being used on the more recent stains on cotton cloth.
Gm/Km Grouping Sensitisation of Cells. Group O R,R, cells were bled by fingerprick for sensitisation. After four saline washes one drop of packed cells and one drop of coating anti-D were incubated for 13 hours at 37OC, then washed four times before use. Cells were used at about 0.1 % suspension strength. Anti-serum Titration. Anti-sera titres were determined by using doubling dilutions of sera, one drop of sensitised cells being added to one drop of each dilution in the wells of a V-bottom micro-titre plate. Dilutions from 114 to 11512 were tested. 19 hours incubation on a Rotatest followed by one minute centrifugation at 1000 rpm made microscopic interpretation of results very easy. Anti-sera were found to have useful working dilutions between 1/64 and 11256. Care must be exercised as the titre in saline may be much higher than in the presence of a stain extract. This problem can be avoided by performing titres in the presence of a stain extract negative for the factor the antiserum is detecting, i.e., each well contains the aritiserum dilution, the stain extract and sensitised red cells. Bloodstain testing. Each stain was tested using varying lengths of thread (2mm, 5mm, lOmm and 20mm). The thread was placed in the wells and one drop of 5% Ammoilia added. The plate was shaken on a Luckhams vibrator for ten minutes and one drop of diluted antiserum added to the wells. The plates were left for 1& hours at 4°C for inhibition to take place and then the threads were removed and one drop of appropriate cell suspension added. The plates were rotatested for 1 hours for agglutination to take place and then centrifuged for one minute at 1000 rpm before reading. Plates were first read microscopically using a low magnification ( x ~ O ) placed , on a sloping fronted (ca. 20" from vertical) light box and read using the streak method described by Bore1 et al., (1967), and then checked again microscopically. Typical positive and negative results are shown in Figure 1.
Results ABO Grouping Size of Stain. Stains up to four-weeks-old were tested. All stains were made on clean white cotton cloth from which blood-stained threads were teased. Threads were cut up to provide 1, 2, 5 , 10 and 20mm lengths and extracted and tested as previously described. Typical results are given in Table 1.
1:igurc 1. Gm Grouping on a microtitre plate. Streak reading method. Typical positive and negative reactions.
It is quite clear from these results that the A and B antigens are readily detectable in l m m of stained thread. This is the equivalent of about 0 . 1 ~ 1of liquid blood. The H antigen on the other hand required 5mm of thread ( 0 . 5 ~ 1 of liquid blood), this length giving good H reactions in A,'s and B's as well. T o be sure of getting reactions for H, 5mm of thread would be the minimum amount that could be used in a practical technique in the laboratory. It may well be necessary to increase the amount of stain extracted in the case of older stains. Age of Stain. The A antigen in A, and A, stains was detected consistently in all stains tested up to five-years-old. Older stains showed more variability, with one 24-year-old A, stain giving a good positivc, while a seven-year-old A, stain gave no reaction. In 16 A, stains tested, none gave any reaction with anti-H while no A, stain over two-years-old gave an H reaction. The B antigen in B stains was clearly detected in stains up to 22-years-old although one ten-yearold stain gave a very weak reaction. The H antigen in B's, as with A,'s was not detected in stains more than two-years-old. The H antigen proved by far the least reliable, as with the previous work on stain size, and while the occasional positive reaction was obtained with 0 ' s up to 20-years-old, no consiste~cycould be achieved in stains more than twoyears-old regardless of ABC) grcup. An increase in the amount of extract in the H wells or an increase in albumin in the 0 cell suspension (Lang, 1976) may TABLE I T H E EFI:ECT OF STAINED THREAD LENGTH ON ABO GROUPING IN MICROTITRE PLATES Slain Croufi
Length ( m m )
-4 I
1 2 51020
Scoring as in Prrcita (1971 ).
1
A, 2 51020
B 1
2 51020
1
0 2 51020
improve matters and will be investigated further. Therefore with the method used as described, the limits for reliable detection of antigens would seem to be five years for A, nine years for B and two years for H, although these limits could probably be extended by increasing the amount of material under test.
GmlKm Grouping The results for stains positive for factors Gml, 2, 10 and K m l up to one-yearold are given in Table 2. TABLE 2 T H E EFFECT O F VARYING STAINED THREAD LENGTH O N TYPING GM/KM POSITIVE BLOODSTAINS U P T O ONE-YEAR OLD
Correct w i n g s ' 4 Gm2+ GmlO+
Gml+
Kml+
Length of thread (mm)
Number of individual stains tested
9.5
ion
4.7
44
20
8
14
9
Twenty individual G m l + stains were tested at four thread lengths, the 5% (i.e., one stain) which failed to react at the two shorter thread lengths was one-year-old. The eight Gm2f stains gave no false negatives. For GmlO and Kml the amount of stain required for correct grouping increased with age, 2mm only being adequate for stains up to one-month-old. Nine G m l - , six GmlO - and 16 K m l - stains all grouped as negative for all four lengths of thread. In 2 1 Gm2 - stains 16 were negative in all tests but five gave apparent inhibition for a single thread length only (1 a t 2mm and 4 a t 5n:m). Table 3 shows the actual results for a plate with false Gm2 positives TABLE 3 GM2 TYPING RESULTS SHOWING FALSE POSITIVE REACTIONS
Length of thread (mm) 2 5
A
B
*
Donor C D 3 3
-
E 1
-*
F 4
-*
-.
20 Known Gm2 Type * False inhibitions
These false positives may have been due to the presence of large numbers of small fibres shed by the threads in these wells interfering with agglutination. Four hundred and twenty-eight individual tests were performed. Only five falsc positives (ca. 1%) were seen, all being in Gm2 typing. Where stains were known to be positive for any given factor, all but one stain proved completely typeable although older stains required up to 20mm of thread (see Table 2). For stains up to one-month-old, lOmm of thread was adequate for all four markers. The one stain that could not be typed was two years old and failed to react for GmlO and K m l but gave a positive for Gml as was expected.
ABO and GmlKm typing in parallel Fifty-six stains, all about two-months-old, were tested simultaneously on the same extract for ABO and G m l , 2, 10 and K m l . 30mm (i.e., twice the mini-
mum) of stained thread was extracted in 12 drops of 5% ammonia. A breakdown of the ABO results is given in Table 4. All reactions with anti-A were TABLE 4 SYNOPSIS OF ABO GROUPING RESULTS Known Grout,
AI A2 A (subtype unknown) B 0
Number 13 9 4 4 26
Anti-A 13 9 4 0 2
Positive with Anti-B 0 1 0 4 0
Anti-H 0 8
2 4 26
-
56
scored as 4 except the two false positives on the Group 0 ' s which were 2 and 3 respectively. These two and the false B result were on the same plate, the wash time of which had been cut to half an hour. Subsequent plates given one hours wash gave no false positives and this wash procedure must be adhered to. All B reactions were graded 4 while reactions with H were more variable. I t is notable that eight out of nine known A,'s gave reactions with anti-H while none of the 13 A,'s gave any reaction for H , this probably being a property of the batch of anti-H in use. All stains were tested for G m l , 2, 10 and K m l . While G m l , 2 and K m l proved very reliable, GmlO was less so. O n occasions GmlO grouping failed totally but the failures gave no special cause for concern as the controls also failed to react. The lower levels of IgG3 which carries the GmlO marker and the generally lower titre of anti-Gml0 (ca. 1/64) as compared with Gml and 2 (ca. 11256) make testing for this antigen a littlc morc difficult. I n a total of 224 single tests (56 x 4), no false negatives and one false positive were seen.
Discussion Microtitre plates have been in use in microbiology and other fields for over ten years. Their use in blood group serology has been rather less well documented. I n forensic serology the only published work is by Alfultis (1965) on a technique for ABO grouping bloodstains using an inhibition method. In 1967 Bore1 et al. described a microtitre method for Gm/Km typing of serum which, with minor modifications, has proved equally effective for bloodstain extracts. There seemed to be no reason why ABO grouping of similar extracts by an elution technique should not be equally successful and the results given here have confirmed this. The limits of sensitivity in terms of amount of stain and age of stain have been explored for ABO and Gm/Km typing, somewhere between 5,ul and lop1 of whole blood as stain being required for both tests on stains up to one month old. For older stains an increase in the amount of stain extracted will help. The greatest problem encountered in this work was the comparatively poor performance for detection of GmlO. This marker is on the y3 heavy chain of IgG3 which represents only 7 4 % of the total IgG (Grubb, 1970) in normal plasma. O n the other hand, Gml and Gm2 are IgGl markers present on 70-80% of the IgG. This near 10-fold difference may, in part, explain the relative difficulty of detecting GmlO, especially in older stains. I n Caucasians and Mongoloids the marker Gm4 (= Gm3) has the same distribution as GmlO and is carried on the IgG 1 molecules. Preliminary experiments with this marker indicate that it may well be detectable in much the same way as Gml and Gm2, that is, in stains up to one-year-old. A problem with Gm4 is its
lack of correlation with GmlO in Negro bloods, most (> 90%) negroes being G m l + 4 - lo+. Although GmlO may give problems, its use with Gm4 can provide useful indications of the ethnic origin of a bloodstain (Ropartz et al., 1966: Grubb. 1970). In terms of economics, microtitre would seem to have some advantages. Each 96-well microtitre plate can be used for 32 ABO determinations or 12 Gm/Km tests using G m l , 2, 10 and K m l and full controls. They compare well in price with cellulose acetate or polycarbonate sheet, as used in the technique of Howard and Martin (1969), one microtitre platc bcing the equivalent of 2-3 cellulose acetate sheets in terms of the number of tests which can be performed. Two other areas where savings can be made are in the fact that moist chambers for incubation are unnecessary, as used plates can be used as lids for plates under test, and large beakers of cold saline are unnecessary as each well can be filled with saline for the washing step. However, perhaps the biggest day-to-day saving can be made by the use of anti-Gm and K m reagents at high dilutions. Our current batches of antisera are diluted as follows : Gml, 2 and K m l 1/64 and GmlO 1/32. These dilutions will of course vary from batch to batch and more markedly between batches from different suppliers, one batch tested being usable at 1/100. I n conclusion, the use of microtitre plates for ABO, Gm and K m typing of bloodstains would seem to be a cheap and reliable alternative to methods currently in use. We have not undertaken to determine the usefulncss of these techniques as applied to the more varied substrates and conditions of stains normally met with in forensic science, although, so far there have been no problems with fresh stains on glass, tissue, newspaper and a wide variety of man-made and synthetic fibres (King, 1978). References ALFULTIS, H. M., 1965, J. Forens. Sci., 10, 319. BOORMAN, K. E. and DODD,B. E., 1966, An Introduction to Blood Group Serology, Churchill, London. BOREL,H., PRYCE,S. and ALLEN,F. H., 1967, Vox Sang., 12, 319. GRUBB, R., 1970, Genetic Markers of Human Immunoglobulins, Chapman and Hall, London. HOWARD, H. D. and MARTIN,P. D., 1969, J. Forens. Sci. Soc., 9, 28. KHALAP,S., PEREIRA, M., and RAND,S., 1976, Med. Sci. Law, 16, ( I ) , 40. KIND,S. S., 1960, Nature, 185, 397. KIND,S. S. and CLEEVELY, R . M., 1969, J . Forens. Sci. Soc., 9, 131. KING,L., 1978, Personal Communication. L A ~ T B. G , G., 1976, J. Forens. Sci. Soc., 16, 61. NICKOLLS, L. D. and PEREIRA, M., 1962, Med. Sci. Lam, 2, 172. OUTTERIDGE, R. A., 1962, Nature, 194, 385. PEREIRA,M., iri 1971, "The Examination and Typing of Bloodstains in the Crime Laboratory", by B. J. Culliford, U.S. Dept of Justice Law Enforcement Assistance Administration, U.S. Government Printing Office, Washington C.C. 20402. ROPARTZ,C., GOLD,E. R., RIVAT,L., ROUSSEAU, P. Y., 1966, Transfusion (Paris), 9, 293.
