Cytogcnct. Cell Genet. 15: 276-280 (1975)
Mitotic association patterns of nucleolar organizing chromosomes in the mouse M.A. S pence1-2 and F.W. L uthardt1 Departments of 'Psychiatry and "Biomathematics, University of California, Los Angeles, School of Medicine, Los Angeles, Calif.
A bstract
Following the report of localization of nucleolar organizer regions to chromo some pairs 15, 18, and 19 of the mouse, a study was undertaken to test these chromo somes for association during mitotic metaphase. Cells from seven strains of mice were analyzed, and no strong evidence was found for any significant association patterns.
Supported in part by: University of California at Los Angeles; Mental Retarda tion Program and Child Psychiatry Program, NPI, UCLA, MCH-927; Interdisci plinary Training in Mental Retardation, HD-04612; Mental Retardation Research Center, UCLA, HD-00345; Research Training in Mental Retardation, HD-05615; Developmental Biology in Mental Retardation. Request reprints from: Dr. M. A nne S pence , Division of Medical Genetics, NeuroPsychiatric Institute, The Center for the Health Sciences, 760 Westwood Plaza, Los Angeles, CA 90024 (USA).
Downloaded by: Vanderbilt University Library 129.59.95.115 - 1/13/2019 6:19:49 PM
The nonrandom association of human acrocentric chromosomes in mitosis is well recognized (W arburton et al., 1973). Several biological explanations have been suggested for this phenomena, including the specific morphology of the acrocentric chromosomes (S pence-C ampbell et al., 1972). However, a more intriguing hypothesis is that the partici pation of the acrocentric chromosomes in the nucleolar organizer activity is responsible for the apparent association. This concept is discussed in some detail by C ooke (1972). H enderson et al. (1972) have localized the nucleolar organizing (NO) regions in human acrocentric chromosomes and assessed the relative importance of their heterochromatic regions. As a continuation of this work, they have recently reported similar studies for the mouse, Mus
Spence, L uthardt
277
Mitotic association patterns
musculus (H kndhrson et al., 1974). Utilizing in situ hybridization with 123I-labeled ribosomal RNA, they demonstrated the presence of rDNA sites in the negatively heteropycnotic region (NHR) of chromosomes 15, 18, and 19. Spencb-C ampbeix et al. (1972) examined cells from seven strains of mice and could find no associations among their telocentric chromosomes. These analyses were restricted to testing for association between homo logous chromosomes, since no groupings comparable to the human acrocentrics could be achieved with this set of morphologically similar chromosomes. In addition, at that time no information was yet available concerning the presence of nucleolar organizer activity in the mouse or the localization of any such activity to specific chromosomes. The current study is designed to investigate the association patterns of the mouse chromosomes defined to participate in nucleolar organizer activity. Materíais and methods The metaphase spreads analyzed were prepared from fibroblasts from five chromosomally normal strains, DBA/2J, CBA/J, C3H/HeJ, BALB/cJ, and C57BL/6J; a total of 50 cells, including males and females, were analyzed. Chromosomes were identified by quinacrine mustard staining (F rancke and N esbitt , 1971) and desig nated by number, as recommended by the C ommittee on Standardized G enetic N omenclature for M ice (1972). The sample is described in greater detail by Spence -C ampbell et al. (1972). Distance between chromosomes was computed by recording centromeric position as X, Y coordinates from a transparent graph superimposed over the photograph of the metaphase spread. The distance between chromosomes i and j is given by: 2r(X¡—Xj)( Y j—Yj)
S jy
(Y i-Y j)*
+
1
7^7
where X t, Yi and Xj, Yj are the coordinates for the two chromosomes i and /, Sx and Sy are the standard deviations, and r is the correlation coefficient for X and T. The value A \j is a dimensionless value but subsequently will be referred to as the distance between the two centromeres. The chromosomes of interest, pairs 15, 18, and 19, are treated as a reference group. By computer, the network of lines is selected that connects the reference group members in the shortest possible distance from all possible networks. This network is called the minimum spanning tree (MST). At random, groups of three pairs are drawn from the remaining 34 chromosomes, i.e., five groups, and an MST is computed for each group. The MST of the reference group is ranked among all groups MSTs. The ranks are standardized to mean zero so that cells missing chromo
Downloaded by: Vanderbilt University Library 129.59.95.115 - 1/13/2019 6:19:49 PM
( X i- X p
278
Spence, L uthardt
Mitotic association patterns
somes due to technical artifact may be included in the sample. A similar procedure forming more groups would be done if fewer pairs are present in the reference groups. Under the null hypothesis of no association over the entire sample, the reference group would have equal numbers of observations for each rank, which is tested by chi-square. A subtle shift in the number of observations per each rank might not result in a significant chi-square, so, in addition, a two-tailed f-test is performed to test the null hypothesis that the mean rank for the reference group is zero if there is no association. Simulation studies with this method have demonstrated that it permits detecting very loose association, i.e., restricted to only 85 °/o of the total cell area, and half the time with samples of only 25 cells. For a complete discussion, see Spence et al. (1975). W arburton et al. (1973) suggest that size variation in the human chromosomes may be an important factor contributing to position in the metaphase spread and, therefore, contributing to apparent associations. Since mouse chromosomes are more similar in size, it was felt that the contribution due to this effect could be ignored for the purpose of this analysis.
Results The analyses were performed by testing for association among pairs 15, 18, and 19; 15 and 18; 15 and 19; and 18 and 19. No tests were attempted for pair-wise association since the earlier studies on the same data had indicated that none existed. Results for the chi-square tests and the (-tests are given in table I. The only statistically significant value is for the /-test for reference group 18 and 19, where the P value of 0.04 is just barely significant at the 0.05 level. Analyses for pairs 15, 18, and 19 were also run on two chromosomally abnormal strains. First, 15 cells carrying a single translocation Rb (9.19) 163H (formerly T(9;19)163H) were analyzed, and the results were again nonsignificant for both the chi-square and /-test: P = 0.82 and 0.28, respectively. Second, 38 cells carrying two translocations Rb (9.19) 163H were analyzed; the chi-square gave P = 0.50, and the /-test gaveP = 0.79, neither being significant.
The apparent lack of nonrandom mitotic association patterns for NO-bearing mouse chromosomes suggests that the mere presence of rDNA does not obligate these chromosomes to be involved in formation of
Downloaded by: Vanderbilt University Library 129.59.95.115 - 1/13/2019 6:19:49 PM
Discussion
Spence, L uthardt Mitotic association patterns
279
Table I. Tests for association of mitotic metaphase among chromosome pairs 15, 18, and 19.“ Reference group
n
Chi-square
15, 18, 19
50
3.52, 5 (0.62)
15, 18
50
15, 19
50
18, 19
50
7.20, 9 (0.62) 10.80, 9 (0.29) 13.60, 9 (0.14)
f-test -0.37, 49 (0.72) 1.20, 49 (0.23) 0.62, 49 (0.54) -2.08, 49 (0.04)
a common nucleolus. Recent studies on NO human chromosomes by H enderson et al. (1972) and E vans et al. (1974) have revealed no significant difference in the amount of rDNA labeling between associated or unassociated satellited chromosomes. Furthermore, E vans et al. (1974) demonstrated that mitotic association was not correlated with variations in the amount of rDNA, confirming the report of O rye (1974). They were also able to localize rDNA to the nucleolar constriction (stalk) and not the satellite region. These observations, in addition to the previously reported influence of translocations upon association patterns involving non-nucleolar organizing chromosomes by S pence -C ampbei.l et al. (1972), suggest that other factors must be considered before proposing that the nonrandom mitotic association between NO chromosomes in some species is an ubiquitous phenomenon. In evaluating the failure to detect association in this sample, the single statistically significant test for pairs 18 and 19 has been ignored. This single result is not strong evidence and could be due to chance. The degree of this association, if real, would be considerably less than that reported for the human acrocentric chromosomes. Several strains of normal mice were pooled to obtain the required sample size for analysis, and heterogeneity with respect to the number of rDNA genes or their distribution could preclude detection of a single association pattern for the
Downloaded by: Vanderbilt University Library 129.59.95.115 - 1/13/2019 6:19:49 PM
»The reference group is defined in the text. The sample size is given in column n as number of cells analyzed; all cells were from karyotypically normal mouse strains. Values for the chi-square and f-tests are given, followed by the number of degrees of freedom. P values are given in parentheses.
280
Spence, L uthardt
Mitotic association patterns
rDNA chromosomes. However, no additional evidence for this associa tion was detected in the independent analyses of the two strains carrying translocations.
A cknowledgements We are indebted to Dr. M uriel N esbitt , who kindly provided her sample of mouse cells for analysis. Sue T ideman provided technical support for the study. Computing assistance was obtained from the Mental Retardation Research Center Computing Resources Group and the Health Sciences Computing Facility, UCLA, sponsored by NIH Special Research Resources Grant RR-3.
References C ommittee on Standardized G enetic N omenclature for M ice : Standard karyo
type of the mouse, Mus musculus. J. Hered. 63: 69-72 (1972). C ooke , P.: Patterns of secondary association between the acrocentric autosomes of
man. Chromosoma 36: 221-240 (1972). E vans, H.J.; B uckland, R.A., and P ardue, M.L.: Location of the genes coding for
Manuscript received 4 August 1975; accepted for publication 1 October 1975.
Downloaded by: Vanderbilt University Library 129.59.95.115 - 1/13/2019 6:19:49 PM
18S and 28S ribosomal RNA in the human genome. Chromosoma 48: 405-426 (1974). F rancke, U. and N esbitt , M.: Identification of the mouse chromosomes by quinacrine mustard staining. Cytogenetics 10: 356-366 (1971). H enderson , A.S.; E icher , E.M.; Yu, M.T., and A twood , K.C.: The chromosomal location of ribosomal DNA in the mouse. Chromosoma 49: 155-160 (1974). H enderson , A.S.; W arburton, D. and A twood , K.C.: Location of ribosomal DNA in the human chromosome complement. Proc. natn. Acad. Sei. USA 69: 3394-3398 (1972). O rye, E.: Satellite association and variations in length of the nucleolar constriction of normal and variant human G chromosomes. Humangenetik 22: 299-309 (1974). Spence , M.A.; F orsythe , A.B.; N esbitt , M., and F rancke, U.: Methods for detect ing nonrandom association of metaphase chromosomes. Technical Report No. 16 (Health Sciences Computing Facility, University of California, Los Angeles 1975). S pence -C ampbell , M.A.; F orsythe , A.B., and N esbitt , M.: Metaphase distribution of the mouse chromosomes. Chromosoma 39: 289-295 (1972). W arburton, D.; N aylor, A.F., and W arburton, F.E.: Spatial relations of human chromosomes identified by quinacrine fluorescence at metaphase. I. Mean interchromosomal distances and distances from the cell center. Humangenetik 18: 297-306 (1973).
Mitotic association patterns of nucleolar organizing chromosomes in the mouse M.A. S pence1-2 and F.W. L uthardt1 Departments of 'Psychiatry and "Biomathematics, University of California, Los Angeles, School of Medicine, Los Angeles, Calif.
A bstract
Following the report of localization of nucleolar organizer regions to chromo some pairs 15, 18, and 19 of the mouse, a study was undertaken to test these chromo somes for association during mitotic metaphase. Cells from seven strains of mice were analyzed, and no strong evidence was found for any significant association patterns.
Supported in part by: University of California at Los Angeles; Mental Retarda tion Program and Child Psychiatry Program, NPI, UCLA, MCH-927; Interdisci plinary Training in Mental Retardation, HD-04612; Mental Retardation Research Center, UCLA, HD-00345; Research Training in Mental Retardation, HD-05615; Developmental Biology in Mental Retardation. Request reprints from: Dr. M. A nne S pence , Division of Medical Genetics, NeuroPsychiatric Institute, The Center for the Health Sciences, 760 Westwood Plaza, Los Angeles, CA 90024 (USA).
Downloaded by: Vanderbilt University Library 129.59.95.115 - 1/13/2019 6:19:49 PM
The nonrandom association of human acrocentric chromosomes in mitosis is well recognized (W arburton et al., 1973). Several biological explanations have been suggested for this phenomena, including the specific morphology of the acrocentric chromosomes (S pence-C ampbell et al., 1972). However, a more intriguing hypothesis is that the partici pation of the acrocentric chromosomes in the nucleolar organizer activity is responsible for the apparent association. This concept is discussed in some detail by C ooke (1972). H enderson et al. (1972) have localized the nucleolar organizing (NO) regions in human acrocentric chromosomes and assessed the relative importance of their heterochromatic regions. As a continuation of this work, they have recently reported similar studies for the mouse, Mus
Spence, L uthardt
277
Mitotic association patterns
musculus (H kndhrson et al., 1974). Utilizing in situ hybridization with 123I-labeled ribosomal RNA, they demonstrated the presence of rDNA sites in the negatively heteropycnotic region (NHR) of chromosomes 15, 18, and 19. Spencb-C ampbeix et al. (1972) examined cells from seven strains of mice and could find no associations among their telocentric chromosomes. These analyses were restricted to testing for association between homo logous chromosomes, since no groupings comparable to the human acrocentrics could be achieved with this set of morphologically similar chromosomes. In addition, at that time no information was yet available concerning the presence of nucleolar organizer activity in the mouse or the localization of any such activity to specific chromosomes. The current study is designed to investigate the association patterns of the mouse chromosomes defined to participate in nucleolar organizer activity. Materíais and methods The metaphase spreads analyzed were prepared from fibroblasts from five chromosomally normal strains, DBA/2J, CBA/J, C3H/HeJ, BALB/cJ, and C57BL/6J; a total of 50 cells, including males and females, were analyzed. Chromosomes were identified by quinacrine mustard staining (F rancke and N esbitt , 1971) and desig nated by number, as recommended by the C ommittee on Standardized G enetic N omenclature for M ice (1972). The sample is described in greater detail by Spence -C ampbell et al. (1972). Distance between chromosomes was computed by recording centromeric position as X, Y coordinates from a transparent graph superimposed over the photograph of the metaphase spread. The distance between chromosomes i and j is given by: 2r(X¡—Xj)( Y j—Yj)
S jy
(Y i-Y j)*
+
1
7^7
where X t, Yi and Xj, Yj are the coordinates for the two chromosomes i and /, Sx and Sy are the standard deviations, and r is the correlation coefficient for X and T. The value A \j is a dimensionless value but subsequently will be referred to as the distance between the two centromeres. The chromosomes of interest, pairs 15, 18, and 19, are treated as a reference group. By computer, the network of lines is selected that connects the reference group members in the shortest possible distance from all possible networks. This network is called the minimum spanning tree (MST). At random, groups of three pairs are drawn from the remaining 34 chromosomes, i.e., five groups, and an MST is computed for each group. The MST of the reference group is ranked among all groups MSTs. The ranks are standardized to mean zero so that cells missing chromo
Downloaded by: Vanderbilt University Library 129.59.95.115 - 1/13/2019 6:19:49 PM
( X i- X p
278
Spence, L uthardt
Mitotic association patterns
somes due to technical artifact may be included in the sample. A similar procedure forming more groups would be done if fewer pairs are present in the reference groups. Under the null hypothesis of no association over the entire sample, the reference group would have equal numbers of observations for each rank, which is tested by chi-square. A subtle shift in the number of observations per each rank might not result in a significant chi-square, so, in addition, a two-tailed f-test is performed to test the null hypothesis that the mean rank for the reference group is zero if there is no association. Simulation studies with this method have demonstrated that it permits detecting very loose association, i.e., restricted to only 85 °/o of the total cell area, and half the time with samples of only 25 cells. For a complete discussion, see Spence et al. (1975). W arburton et al. (1973) suggest that size variation in the human chromosomes may be an important factor contributing to position in the metaphase spread and, therefore, contributing to apparent associations. Since mouse chromosomes are more similar in size, it was felt that the contribution due to this effect could be ignored for the purpose of this analysis.
Results The analyses were performed by testing for association among pairs 15, 18, and 19; 15 and 18; 15 and 19; and 18 and 19. No tests were attempted for pair-wise association since the earlier studies on the same data had indicated that none existed. Results for the chi-square tests and the (-tests are given in table I. The only statistically significant value is for the /-test for reference group 18 and 19, where the P value of 0.04 is just barely significant at the 0.05 level. Analyses for pairs 15, 18, and 19 were also run on two chromosomally abnormal strains. First, 15 cells carrying a single translocation Rb (9.19) 163H (formerly T(9;19)163H) were analyzed, and the results were again nonsignificant for both the chi-square and /-test: P = 0.82 and 0.28, respectively. Second, 38 cells carrying two translocations Rb (9.19) 163H were analyzed; the chi-square gave P = 0.50, and the /-test gaveP = 0.79, neither being significant.
The apparent lack of nonrandom mitotic association patterns for NO-bearing mouse chromosomes suggests that the mere presence of rDNA does not obligate these chromosomes to be involved in formation of
Downloaded by: Vanderbilt University Library 129.59.95.115 - 1/13/2019 6:19:49 PM
Discussion
Spence, L uthardt Mitotic association patterns
279
Table I. Tests for association of mitotic metaphase among chromosome pairs 15, 18, and 19.“ Reference group
n
Chi-square
15, 18, 19
50
3.52, 5 (0.62)
15, 18
50
15, 19
50
18, 19
50
7.20, 9 (0.62) 10.80, 9 (0.29) 13.60, 9 (0.14)
f-test -0.37, 49 (0.72) 1.20, 49 (0.23) 0.62, 49 (0.54) -2.08, 49 (0.04)
a common nucleolus. Recent studies on NO human chromosomes by H enderson et al. (1972) and E vans et al. (1974) have revealed no significant difference in the amount of rDNA labeling between associated or unassociated satellited chromosomes. Furthermore, E vans et al. (1974) demonstrated that mitotic association was not correlated with variations in the amount of rDNA, confirming the report of O rye (1974). They were also able to localize rDNA to the nucleolar constriction (stalk) and not the satellite region. These observations, in addition to the previously reported influence of translocations upon association patterns involving non-nucleolar organizing chromosomes by S pence -C ampbei.l et al. (1972), suggest that other factors must be considered before proposing that the nonrandom mitotic association between NO chromosomes in some species is an ubiquitous phenomenon. In evaluating the failure to detect association in this sample, the single statistically significant test for pairs 18 and 19 has been ignored. This single result is not strong evidence and could be due to chance. The degree of this association, if real, would be considerably less than that reported for the human acrocentric chromosomes. Several strains of normal mice were pooled to obtain the required sample size for analysis, and heterogeneity with respect to the number of rDNA genes or their distribution could preclude detection of a single association pattern for the
Downloaded by: Vanderbilt University Library 129.59.95.115 - 1/13/2019 6:19:49 PM
»The reference group is defined in the text. The sample size is given in column n as number of cells analyzed; all cells were from karyotypically normal mouse strains. Values for the chi-square and f-tests are given, followed by the number of degrees of freedom. P values are given in parentheses.
280
Spence, L uthardt
Mitotic association patterns
rDNA chromosomes. However, no additional evidence for this associa tion was detected in the independent analyses of the two strains carrying translocations.
A cknowledgements We are indebted to Dr. M uriel N esbitt , who kindly provided her sample of mouse cells for analysis. Sue T ideman provided technical support for the study. Computing assistance was obtained from the Mental Retardation Research Center Computing Resources Group and the Health Sciences Computing Facility, UCLA, sponsored by NIH Special Research Resources Grant RR-3.
References C ommittee on Standardized G enetic N omenclature for M ice : Standard karyo
type of the mouse, Mus musculus. J. Hered. 63: 69-72 (1972). C ooke , P.: Patterns of secondary association between the acrocentric autosomes of
man. Chromosoma 36: 221-240 (1972). E vans, H.J.; B uckland, R.A., and P ardue, M.L.: Location of the genes coding for
Manuscript received 4 August 1975; accepted for publication 1 October 1975.
Downloaded by: Vanderbilt University Library 129.59.95.115 - 1/13/2019 6:19:49 PM
18S and 28S ribosomal RNA in the human genome. Chromosoma 48: 405-426 (1974). F rancke, U. and N esbitt , M.: Identification of the mouse chromosomes by quinacrine mustard staining. Cytogenetics 10: 356-366 (1971). H enderson , A.S.; E icher , E.M.; Yu, M.T., and A twood , K.C.: The chromosomal location of ribosomal DNA in the mouse. Chromosoma 49: 155-160 (1974). H enderson , A.S.; W arburton, D. and A twood , K.C.: Location of ribosomal DNA in the human chromosome complement. Proc. natn. Acad. Sei. USA 69: 3394-3398 (1972). O rye, E.: Satellite association and variations in length of the nucleolar constriction of normal and variant human G chromosomes. Humangenetik 22: 299-309 (1974). Spence , M.A.; F orsythe , A.B.; N esbitt , M., and F rancke, U.: Methods for detect ing nonrandom association of metaphase chromosomes. Technical Report No. 16 (Health Sciences Computing Facility, University of California, Los Angeles 1975). S pence -C ampbell , M.A.; F orsythe , A.B., and N esbitt , M.: Metaphase distribution of the mouse chromosomes. Chromosoma 39: 289-295 (1972). W arburton, D.; N aylor, A.F., and W arburton, F.E.: Spatial relations of human chromosomes identified by quinacrine fluorescence at metaphase. I. Mean interchromosomal distances and distances from the cell center. Humangenetik 18: 297-306 (1973).
