Interstrain competition amongst mouse spermatozoa inseminated in various proportions, as affected by the genotype of the Y chromosome

Nonagouti (KP X C57BL)F1 hybrid females were artificially inseminated with a mixture of spermatozoa from males of the KE (nonagouti) and CBA (agouti) strains and the genotype of young was estimated by fur pigmentation. When KE and CBA spermatozoa mixed in the ratios of 1:1, 2:1 and 4:1 were inseminated after ovulation, 87%, 56% and 29% of progeny, respectively, were sired by CBA males, i.e. proportions of CBA progeny were significantly higher than ratios of CBA spermatozoa in the mixture. The surplus of CBA progeny was significantly less in females inseminated before ovulation, which may suggest that more rapid capacitation of CBA spermatozoa is partly responsible for their competitive advantage. In preparations from oviducal flushings of females killed 2-3 h after insemination, CBA spermatozoa (recognized by their shape) were found in similar proportions as in the inseminated mixture. There was therefore no evidence of their preferential selection at the uterotubal junction. No competitive advantage of CBA spermatozoa occurred when they were inseminated with spermatozoa from males of the KE.CBA strain, congenic with KE but with the Y chromosome derived from the CBA strain. This indicates that genetic factors linked with the Y chromosome may influence competitive ability of spermatozoa.     

Changes in germ cell population in young and adult female mice from two inbred strains: CBA/Kw and KE

Female mice from two inbred strains CBA/Kw and KE differ markedly in fertility. The gametes of females from KE strain are of poorer quality than those of CBA/Kw. We analyzed the number of oocytes per ovary in KE and CBA/Kw mice aged 5, 25, 90, 180 and 360 days. The ovaries were dissected and processed according to the routine histological methods. In case of five-day-old females we used a modified distributed point counting method while in order to examine the gonads of older females, the nucleoli counting method was applied. In general, we observed gradual decrease in germ cell number throughout the whole life of females from both strains. The noticeable wave of oocyte loss occurs between 5th and 25th days of life. The mice from KE inbred strain on day 25th (1650 +/- 322 vs. 1140 +/- 210) and 90(th) (1040 +/- 211 vs. 692 +/- 89) days have significantly (p<0.005) more germ cells than the females from CBA/Kw strain. In older females the differences were not statistically significant. Interestingly, CBA/Kw females were found to have more rapid loss of primordial follicles throughout their lives. This can explain their shortened reproductive lifespan which was observed earlier.     

In-vivo and in-vitro maturation rate of oocytes from two strains of mice

Female mice of the KE and CBA strains were used to examine the rate of oocyte maturation in vivo and in vitro. In CBA females killed just before ovulation most preovulatory oocytes were already in the metaphase II stage, while the oocytes of KE mice were arrested at metaphase I until the time of ovulation, and further stages of maturation occurred in the oviduct, reaching the metaphase II stage 3-5 h later. A similar strain difference in oocyte maturation rate was observed from in-vitro culture of cumulus-free oocytes, isolated from the ovaries of PMSG-primed females and intact females killed at the metoestrous phase of the cycle. This indicates that the strain-specific course of maturation is determined in the oocyte by a few days before ovulation. Therefore, if the rate of oocyte maturation is influenced by somatic components of the follicle, this must occur at some earlier stages of follicle development.     

Preferential segregation of marker chromosomes 14 and 18 in mouse recombinant inbred strains derived from the KE and CBA/Kw strains

The segregation pattern of chromosomes 14 and 18 were analyzed in recombinant inbred strains of mice developed from KE and CBA/Kw strains. The analysis was possible owing to the fact that the C-band on chromosomes 14 of the CBA/Kw strain and that of chromosome 18 of the KE strain show size polymorphism: while the CBA/Kw mice have a small sized C-band on chromosome 14, the KE mice show small C-bands on chromosome 18. Chromosomes were identified by G-banding and FISH. The results show that the chromosomes with small centromeric chromatin segregate preferentially.     

X-Y chromosome dissociation in mouse strains difering in efficiency of spermatogenesis: Elevated frequency of univalents in pubertal males

Dissociation of the X-Y chromosome bivalent in diakinesis-metaphase I spermatocytes of adult mice was significantly more frequent in the CBA strain (29%) than in C57, KP, or KE strains (7–11%). Autosome dissociatio (1–5%) involved only the smallest chromosome pairs. Eleyatedfrequency of X-Y dissociation in the CBA strain correlates with significantly lower testes weight and lower yield of spermatogenesis, which suggests that sex bivalent dissociation man be responsible for some loss of spermatogenic cells. However, sperm quality is not affected, the percentage of normal spermatozoa and their fertlizing capacity being higher in CBA thatn in the remaining strains. Two congenic strains, KE and KE. CBA (the latter with the Y chromosome introduced from CBA), had the same level of X-Y dissociatios, suggesting that the Y chromosome plays no rle in the determination of this character. In comparison with adult males pubertal (27–29 day-old) males had twice as hig a frequency of X-Y dissociation in KE an KP strains, and combined frequeicies of dissociated sex and autosome bivalents were significantly higher in pubertal males of all tested strains. Although te level of chromosome dissociation is not sufficient to explain increased mortality of germ cells observed in pubertal males, it could be one of the contributing factors.     

Correlation of centromeric heterochromatin C-band polymorphism with breeding failure in mice

The aim of the present study was to test the hypothesis about the relation between segregation of chromosomes 14 and 18 and the deterioration of mouse fertility and vitality. The analysis was possible because C-banding on chromosome 14 and chromosome 18 of the CBA/Kw and KE strains show size polymorphism. A small sized C-band on chromosome 14 is characteristic for the CBA/Kw mice, while the KE mice show small C-bands on chromosomes 18. Thus, if fertility parameters are affected in a centromere-dependent manner, we should observe non-random inheritance of both chromosome pairs in recombinant inbred (RI) strains. The results showed statistically significant preferential segregation of chromosomes 14 and 18 with small C-bands. Most of the RI strains inherited chromosome 14 from the CBA/Kw strain and chromosome 18 from the KE strain, and did not manifest a deterioration of fertility and vitality. On the contrary, RI strains that inherited chromosomes 14 and 18 from one of the parental strains, particularly the KE strain, stopped breeding or had difficulties in producing the next generation.     

Frequency of X-Y chromosome dissociation in mouse spermatocytes from interstrain crosses,recombinant inbred strains,and chimeras: Possible involvement of paternal genome imprinting

The frequency of dissociation of the X-Y chromosome bivalent in diakinesis-metaphase I spermatocytes differs significantly between two inbred mouse strains, CBA (29%) and KE (7%), that were used to obtain reciprocal F1 hybrids, and to develop recombinant inbred (R1) strains. The level of X-Y dissociation was significantly higher in (KExCBA)F1 hybrids sired by the CBA males (24%) than in reciprocal F1 hybrids (12%), revealing the inheritance after the father. Among 14 RI strains, nine were concordant with KE, one with CBA, and four had intermediate phenotype, significantly different from both progenitor strains. This shows that at least two genes are involved, and their possible linkage with agouti and Trf loci is suggested. The linkage with agouti was confirmed by testing additional 10 CBXE incipient RI strains. There was no significant difference in the level of X-Y dissociation between EXCB RI strains derived from the original cross sired by the CBA males and CBXE RI strains derived from the reciprocal cross. The involvement of the Y chromosome-linked factors was unlikely because it was found earlier (Krzanowska, 1989: Gamete Res 23:357–365) that two congenic strains, KE and KE.CBA, differing with respect to the source of the Y chromosome, had the same level of X-Y dissociation. Thus, the difference obtained between reciprocal F1 hybrids is interpreted in terms of paternal genome imprinting imposed by CBA males and propagated only in the presence of some alleles derived from this strain. Analysis of six KE ? CBA-T6 chimeras, among them three germ line chimeras, points to the conclusion that the tendency to low or high level of X-Y chromosome dissociation is expressed rather autonomously by KE or CBA-T6 spermatocytes (as recognized by a marker chromosome pair), respectively, and was not modified by the presence of somatic cells of the opposite strain. © 1994 Wiley-Liss, Inc.     

Sex-dependent frequency and type of autosomal univalency at the first meiotic metaphase in mouse germ cells

Univalents at the first meiotic metaphase in mouse spermatocytes occur mainly in the XY pair, making it difficult to compare the amounts of univalency in males and females. In this study, the amounts of autosomal univalency in male and female meiosis were compared using the model strain CBA-T6, in which univalency of the small marker autosome pair T6 has been shown to occur very frequently in spermatocytes. Mice from inbred CBA and DBA strains were also analysed. The total frequencies of univalency (sex chromosomes plus autosomes) in metaphase I spermatocytes were 45.6% in CBA, 36.9% in CBA-T6, and 37.3% in DBA males. The aneuploidy in metaphase II spermatocytes ranged from 1.4 to 3% in these strains, which was in agreement with previous findings that most primary spermatocytes with abnormal chromosome configurations are arrested in their development before metaphase II. In the CBA-T6 strain, autosomal univalency at metaphase I mostly involved chromosome pair T6; however, its frequency differed significantly between the sexes, amounting to 18.9% in spermatocytes and 4.3% in oocytes. In the CBA strain, autosomal univalents at metaphase I were seen in 7.7% of the spermatocytes and 1.4% of the oocytes and, in DBA mice, in 4.9% of the spermatocytes and 3.8% of the oocytes. However, in DBA oocytes, when univalency occurred it usually concerned a greater number of bivalents in one cell (range: 2-19 disjoined bivalents), a phenomenon very rare in males of this strain. This study shows that univalent formation differs between the male and female types of meiosis.     

Phenotype and fertilizing capacity of spermatozoa of chimaeric mice produced from two strains that differ in sperm quality

Spermatozoa of males from the inbred mouse strains, KE (albino) and CBA (agouti), are distinguishable by head shape and differ in quality: CBA spermatozoa show a lower percentage of abnormal heads and higher efficiency of fertilization. Aggregation chimaeras were produced to investigate whether these differences are intrinsic or extrinsic to spermatogenic cells. Among 24 overt chimaeras, 14 were males: 1 was sterile, 8 produced either KE or CBA spermatozoa, as recognized by shape and by progeny testing, and 5 (21%) were germ-line chimaeras. Both the level of abnormal sperm heads and the efficiency of fertilization of CBA and KE spermatozoa produced by chimaeric males (except two subfertile ones) were within the range characteristic for the respective strain. All 5 germ-line chimaeras, irrespective of their coat colour composition, produced about 98% KE and only 2% CBA spermatozoa, which indicated strong selection against CBA germ cells. However, mature CBA spermatozoa showed high competitive ability, because the proportion of agouti progeny (from the CBA component) sired by those males was significantly higher than the proportion of CBA spermatozoa, estimated from vaginal plug preparations after every mating. The fact that this difference was particularly striking for one chimaera with a preponderance of CBA somatic component may suggest some influences extrinsic to spermatogenic cells. We conclude from this study that sperm head shape, the level of sperm abnormalities and fertilizing capacity are determined largely autonomously by genes acting in the germ cells. The internal environment created by foreign somatic cells exerts only minor modifications, unless there has been deterioration beyond the range of 'normality'.     

Number of chromocentres in the nuclei of mouse Sertoli cells in relation to the strain and age of males from puberty to senescence

Nuclei of mouse Sertoli cells were examined on air-dried toluidine blue-stained preparations to analyse factors influencing the aggregation of heterochromatin into chromocentres. For the CBA strain males tested at 1.3, 2, 3, 6, 9 and 12 months of age, mean numbers of heterochromatin bodies were 4.8, 2.4, 2.1, 1.8, 1.6 and 1. 4, respectively; two chromocentres predominated from 2 to 9 months of age. In the KE strain, heterochromatin aggregation was significantly accelerated; nuclei containing only one chromocentre were predominant from 6 months. The number of chromocentres did not change with the stages of the seminiferous cycle, and after 1 month of cryptorchid condition. Cryptorchidism resulted in disruption of spermatogenesis and Sertoli cell dysfunction, as demonstrated by the lack of immunohistochemically detectable androgen receptors. The difference in the number of chromocentres between KE and CBA Sertoli cells persisted after 3 days of in vitro culture, but unidentified cells with numerous chromatin bodies were also observed. Testing recombinant inbred strains indicates that at least two genes are involved in the difference in the number of chromocentres between progenitor KE and CBA strains; however, no correlations were found with 15 marker loci or with parameters linked to reproduction. Of the eight strains tested, AKR and C3H showed a 'CBA-like' chromocentre pattern; C57BL, B10.BR, B10.BR-Y(del) and KP were 'KE-like'; and BALB/c and DBA/2 were intermediate. The results showed that centromere aggregation in the Sertoli cell progresses throughout the life of a male in a strain specific manner; however, its functional significance remains unknown.     

Ribosomal genes in inbred mouse strains: interstrain and intrastrain variations of copy number and extent of methylation

Quantitative dot hybridization was used to estimate the rDNA copy number in brain tissues of five inbred mouse strains (AKR/JY, NZB/B1OrlY, CBA/CaLacY, 101/HY, and 129/JY), which were obtained from the collection of the Research Center of Biomedical Technologies (Y). In each strain, 9-12 mice aged 1-2 months were examined. The rDNA copy number per diploid genome in strains AKR (range 105-181, mean +/- SD 136 +/- 27) and NZB (129-169, 148 +/- 12) was significantly lower than in strains CBA (172-267, 209 +/- 31), 101 (179-270, 217 +/- 30), and 129 (215-310, 264 +/- 33). Mice of strain NZB were relatively homogeneous in this trait (CV = 8.1%). Strains AKR, CBA, 101, and 129 displayed significant between-group differences, CV varying from 12.5 to 19.9%. The same DNA specimens were digested with MspI or HpaII and used to estimate the extent of methylation of the 28S rDNA region. Regardless of the strain, all mice could be classed into two groups. One group (20 mice) had a methylated fraction accounting for less than 8% of rDNA and included all nine mice of strain NZB, seven out of nine mice of strain 101, and three out of ten mice of strain 129. In the other group (29 mice), the methylated fraction varied from 18 to 38%. A possible role of methylation and the genome dosage of ribosomal genes in phenotypic variation (quantitative trait variation) of inbred mouse strains is discussed.     

Influence of genetic factors on the fertilization of mouse ova in vitro

In-vitro fertilization was studied in F1 hybrids, 4 inbred and outbred strains of mice. Experiments were performed with intact ova (Series I) and denuded ova (Series II). In Series I the highest percentages of fertilization were obtained with gametes of F1 hybrids (86.7%) and of the CBA strain (87.5%). The KP, KE, and particularly the C57 strains, which in vivo give less than 100% fertilization, also gave low indices in vitro (42.9, 26.0 and 8.6% respectively). Testing gametes of these inbred strains with F1 gametes showed that spermatozoa and ova are responsible for the low percentages of fertilization. The rate of fertilization was mainly dependent on the genotype of the spermatozoon. In Series II, high percentages of fertilization (97-100%) were regularly obtained in all groups, indicating that differences between strains pertained mainly to binding of spermatozoa with the zona pellucida. The incidence of polyspermy, which did not exceed 5% in Series I and reached 58% in Series II, was dependent on the genotype of both gametes.     

The activity of alkaline phosphatase in the sera of DBA/2NCrj and CBA/BrA mice with calcified tongue lesions

Spontaneously occurring calcified lesions were found in the tongues of DBA/2NCrj and CBA/BrA mice. In the DBA/2NCrj strain, the frequency of the lesion was 80% (males) and 88% (females). The youngest age of a mouse with this lesion was 18 days after birth, and 3-4 lesions were found in the tongue of 6- to 8-week-old mice. In CBA/BrA mice, 20% of females and 48% of males had the lesions. No significant differences were found in the serum calcium concentrations in high and low lesion-developing strains, but the alkaline phosphatase activities in the high-developing DBA/2NCrj, DBA/LiA, and CBA/BrA strains were higher than in strains with no calcified lesions.     

Sperm penetration in vitro into ovarian and tubal oocytes from mice of the inbred KE and C57 strains

The method of in vitro fertilization was applied to test a previous suggestion that the lowered fertilizability of the tubal oocytes of female KE strain mice and the high resistance of their zona pellucida to proteolytic enzymes, are due to the premature cortical reaction taking place near the time of ovulation. Therefore higher fertilizability of ovarian oocytes is expected. The effectiveness of F₁ hybrid sperm penetration into ovarian and tubal KE oocytes confirmed these assumptions. The ovarian KE oocytes recovered 9–10 hours after administration of human chorionic gonadotropin (HCG) showed significantly higher penetrability (70–83%) than did the tubal oocytes recovered 12 hours after HCG (about 50%) and 14–16 hours after HCG (20%). Similar results were obtained with C57 oocytes. Sperm penetration into ovarian oocytes (10 hours after HCG) was much more effective (67%) than into tubal oocytes (18%); this finding correlated with more rapid zona dissolution by chymotrypsin. On the basis of these results one might speculate that premature cortical reaction takes place also in the C57 strain.     

Effect of time of ovulation on fertilization after intrabursal transfer of spermatozoa (ITS): improvement of a new method for artificial insemination in mice

The timing of AI in relation to ovulation was examined to improve intrabursal transfer of spermatozoa (ITS) in mice, a new method of AI that involves transfer of spermatozoa into a space near the infundibulum. Two microliters of fresh epididymal B6C3F1 spermatozoa (containing 2 x 10(5) spermatozoa) were inseminated 1, 7, 12, or 17 h after hCG administration. At 1.7 days after ITS, normal cleaving embryos were recovered at rates ranging from 6 to 50% (21.5 +/- 15.8%; mean +/- S.D.), 40-100% (75.2 +/- 20.2%), 33-100% (60.1 +/- 19.3%), and 6-47% (22.7 +/- 13.3%), respectively. The rate obtained by ITS 7h after hCG administration was comparable (P > 0.05) to that (90.5 +/- 6.3%) for embryos obtained after natural mating (control), but rates at all other times were significantly less than control. To examine whether in vivo fertilization rate differs when spermatozoa from various mouse strains are used, B6C3F1 females were inseminated with spermatozoa from ICR, C57BL/6N and C3H/HeN mice 7 h after hCG administration. There were strain differences (P < 0.01 for ICR and B6C3F1 versus C57BL/6N and C3H/HeN) for in vivo fertilization rates (83.9 +/- 10.3%, 75.2 +/- 20.2%, 33.6 +/- 24.5% and 25.6 +/- 16.1% for ICR, B6C3F1, C57BL/6N and C3H/HeN, respectively). Similar rates (72.9 +/- 7.3% and 27.5 +/- 46.2% for ICR and C57BL/6N, respectively) were also obtained when oocytes were inseminated with spermatozoa of the same strain. In addition, females (B6C3F1) inseminated by ITS of fresh B6C3F1 spermatozoa 7 h after hCG administration yielded normal mid-gestational fetuses with an average litter size of 7.0 +/- 4.9, which seemed much higher than the previously reported litter size of 3.2. In conclusion, the timing of AI was considered a key factor affecting in vivo fertilization efficiency.     

Cryopreservation of wild mouse spermatozoa

Spermatozoa of wild mice from China, Czechoslovakia, Denmark, India, Japan and Switzerland were frozen and stored at -196 degrees C. After thawing, intact oocytes were inseminated in vitro with relatively high motility frozen-thawed mouse spermatozoa from Czechoslovakia, Denmark and India, while oocytes with a partially dissected zona were inseminated with low motility frozen-thawed spermatozoa from China, Japan and Switzerland. Embryos developing to the 2-cell stage from oocytes fertilized with frozen-thawed spermatozoa were transferred to the oviducts of female recipients on the first day of pseudopregnancy (day when a vaginal plug was confirmed). Successful embryo development to the 2-cell stage was 46 to 67%. Offspring resulted from 17 to 51% of these transferred 2-cell embryos.     

Ribosomal genes in inbred mouse strains: Interstrain and intrastrain variation of copy number and extent of methylation

Quantitative dot hybridization was used to estimate the rDNA copy number in brain tissues of five inbred mouse strains (AKR/JY, NZB/B1OrlY, CBA/CaLacY, 101/HY, and 129/JY), which were obtained from the collection of the Research Center of Biomedical Technologies (Y). In each strain, 9–12 mice aged 1–2 months were examined. The rDNA copy number per diploid genome in strains AKR (range 105–181, mean ± SD 136 ± 27) and NZB (129–169, 148 ± 12) was significantly lower than in strains CBA (172–267, 209 ± 31), 101 (179–270, 217 ± 30), and 129 (215–310, 264 ± 33). Mice of strain NZB were relatively homogeneous in this trait (CV = 8.1%). Strains AKR, CBA, 101, and 129 displayed significant between-group differences, CV varying from 12.5 to 19.9%. The same DNA specimens were digested with MspI or HpaII and used to estimate the extent of methylation of the 28S rDNA region. Regardless of the strain, all mice could be classed into two groups. One group (20 mice) had a methylated fraction accounting for less than 8% of rDNA and included all nine mice of strain NZB, seven out of nine mice of strain 101, and three out of ten mice of strain 129. In the other group (29 mice of strains AKR, CBA, 101, and 109), the methylated fraction varied from 18 to 38%. A possible role of methylation and the genome dosage of ribosomal genes in phenotypic variation (quantitative trait variation) of inbred mouse strains is discussed.     

Immune responses of different mouse strains after challenge with equivalent lethal doses of Toxoplasma gondii

Most immunological studies that utilize different strains of inbred mice following T. gondii infection fail to compensate for differences in host susceptibility to the size of the parasite innoculum. To address this concern, susceptible C57BL/6 and resistant CBA/J mice were orally infected with either an equivalent 50% lethal dose (LD50) of brain cysts of the 76K strain of T. gondii (15 cysts in C57BL/6, 400 cysts in CBA/J) or the same dose of parasites in each mouse strain. C57BL/6 mice receiving 400 cysts (LD50 of CBA/J mice) died post infection, whereas CBA/J mice that received 15 cysts (LD50 of C57BL/6 mice) survived. Parasite loads in the brains and serum Toxoplasma-specific IgG1 titers of LD50-infected C57BL/6 mice were significantly higher than those in LD50- or 15 cysts-infected CBA/J mice, whereas splenocyte proliferation to Toxoplasma antigen and the percentage of CD8 alpha+ T cells were reduced in LD50-infected C57BL/6 mice. In contrast, serum IgG2a and IgM titers, the percentage of gamma delta T cells and IFN-gamma expression of spleen of LD50-infected CBA/J mice were higher than those of either 15 cysts-infected CBA/J mice or LD50-infected C57BL/6 mice. These observations demonstrate that the immune response between LD50-infected C57BL/6 and CBA/J mice was more prominent when compared to C57BL/6 or CBA/J mice receiving the same parasite inoculum. These observations would suggest that caution must be excersized in the planning and interpretation of data when the size of the parasite inoculum has not been adjusted for mouse strain.     

Apoptosis in the preimplantation mouse embryo: effect of strain difference and in vitro culture.

Cell death by apoptosis occurs predominantly in the inner cell mass (ICM) of the blastocyst, the cell population which carries the germ line and gives rise to the foetus. The frequency of apoptosis in blastocysts varies widely within outbred species such as human and cow. We have addressed the basis of this variation by examining the relative influence of strain difference and in vitro culture conditions on apoptosis, using embryos from two different strains of mice (MF1 and C57BL6/CBA) in two different culture media (M16 and kSOM). In both strains and all crosses apoptosis was first detected by nuclear fragmentation or TUNEL [Terminal deoxynucleotidyl transferase mediated d-UTP nick end-labelling] labelling at the early blastocyst stage. This was true for embryos which had developed in vivo, and in vitro in both M16 and kSOM. The apoptotic index in blastocysts was found to be significantly different between both media and strain (P < 0.0001). Blastocysts from MF1 x MF1 at equivalent stages had an apoptotic index of 32.4% in M16 and 20.3% in kSOM. Blastocysts from C57BL6/CBA x C57BL6/CBA had an apoptotic index of 19.3% in M16 and 14.4% in kSOM. When embryos of similar cell number were compared, a significantly greater apoptotic index was found for cultured MF1 x MF1 embryos with a cell number between 40 and 59 compared to similar directly flushed C57BL6/CBA embryos (P = 0.001), and MF1 embryos (P < 0.0005). MF1 x MF1 embryos and C57BL6/CBA x MF1 embryos of 60-79 cells had a greater apoptotic index in M16 than kSOM (P < 0.0005) but the difference between media was not significant for C57BL6/CBA x C57BL6/CBA. When strain was compared MF1 x MF1 embryos of 60-79 cells had a significantly greater apoptotic index than C57BL6/CBA x MF1 in both media (P < 0.0005 M16; P = 0.002 kSOM) and than C57BL6/CBA x C57BL6/CBA in M16 (P = 0.019). Our data suggest that genetic make-up and the chemical composition of simple medium are equally important in determining the level of apoptosis.     

Increased resorptions in CBA mice exposed to low-frequency magnetic fields: An attempt to replicate earlier observations

This paper has two aims. First, it reports the findings of a study on the effects of low-frequency magnetic fields on reproduction. Second, it serves as an example of an attempt to replicate the results of an experimental study in an independent laboratory and discusses some of the problems of replication studies. To try to replicate the findings of a study reporting increased resorptions (fetal loss) in mice exposed to 20 kHz magnetic fields with sawtooth waveform and to study the possible effects of 50 Hz sinusoidal fields, pregnant mice were exposed to magnetic fields from day 0 to 18 of pregnancy, 24 h per day. The flux densities of the vertical magnetic fields were 15 μT (peak-to-peak) at 20 kHz and 13 or 130 μT (root mean square) at 50 Hz. Two strains of animals were used: CBA/S mice imported from the laboratory reporting the original observations, and a closely related strain CBA/Ca. The CBA/S mice were cleaned of pathogenic microbes and parasites before they were imported into our laboratory. The magnetic field exposures did not affect resorption rate in CBA/Ca mice. In CBA/S, the frequency of resorptions was higher in the exposed mice than in the control group. However, the increase was not significantly different from either the no-effect hypothesis or the results of the original study we were attempting to replicate. Differences between the two studies and difficulties in interpreting the results are discussed. It is concluded that the results tend more to support than argue against increased resorptions in CBA/S mice exposed to the 20 kHz magnetic field. The results demonstrate that animal strain is an important variable in bioelectromagnetics research: even closely related strains may show different responses to magnetic field exposure. Bioelectromagnetics 18:410–417, 1997. © 1997 Wiley-Liss, Inc.