Site search   
  Home    Publications

Browse Publications
List of Contents

Quantitative Neurogenetics & QTL Mapping

Genetics of Myopia

Control of Neuron Number and Stereology

Growth Cones and Dying Axons

Retina Development and Visual System Mutants

Grant Application

U.S. Patent


Need Help?
Help with Publications
Help with Nervenet
Contact Us

Note to the Reader This is a preprint of a paper in press at Behavior Genetics.
Cite as: Williams RW, Airey DC, Kulkarni A, Zhou G, Lu L (2000) Genetic dissection of the olfactory bulb of mice: QTLs on chromosomes 4, 6, 11, and 17 modulate bulb size. Behavior Genetics: in press.

Genetic dissection of the olfactory bulb of mice:
QTLs on chromosomes 4, 6, 11, and 17 modulate bulb size

Robert W. Williams, David C. Airey, Anand Kulkarni, Guomin Zhou*, Lu Lu
Center for Neuroscience and Department of Anatomy and Neurobiology, University of Tennessee, 855 Monroe Avenue, Memphis, Tennessee 38163 USA
*Current address: Department of Embryology and Histology, Shanghai Medical University, Shanghai, PRC
Email questions and comments to


     Table 1: Bulb Weight in BXD strains
     Table 2: A Correlation Matrix, Bulb Weight and 7 other Traits
     Figure 4: Interval Maps of the Bulb QTLs
Bulb Structure and Behavior
Glomerular Architecture and Bulb Weight
Comparison of QTLs that Modulate, Bulb, Hippocampus, and Cerebellum
Effects of Age on the Size of the Olfactory Bulbs
Candidate Genes for the Bulb QTLs



Olfaction is influenced by a complex mix of environmental factors and genes that modulate the production, migration, and maturation of cells in the olfactory bulbs. In this study we analyzed effects of sex, age, and brain weight on olfactory bulb size in sexually mature mice. Regression corrected values (residuals) were used to map four quantitative trait loci (QTLs) that selectively modulate bulb weight. Traits were mapping using C57BL/6J (B6) and DBA/2J (D2) parental strains, an F2 intercross, and 35 BXD recombinant inbred strains.
     Bilateral bulb weight in adult mice ranges from 10 to 30 mg. Half of this remarkable variation can be predicted from differences in brain weight, sex, body weight, and age. A 100 mg difference in brain weight is associated with a 4.4 mg difference in bulb weight. Bulbs gain ~0.2 mg/week—a 1% increase that continues until at least to 300 days of age.
     Males tend to have slightly larger bulbs than do females. Heritability of variation among BXD and F2 mice is modest, and h2 ranges between 0.2 and 0.3. We identified four QTLs with selective effects on bulb size (genome-wide p < .05). Bulb1 is located on Chr 4 and Bulb2 is located on Chr 6. Alleles inherited from B6 at both loci increase weight by 0.5–1.0 mg. Bulb3 is located on proximal Chr 11 and Bulb4 is located near the centromere on Chr 17. In contrast, B6 alleles at these loci decrease bulb weight by 0.5–1.0 mg. Collectively, these four QTLs account for 20% of the phenotypic variance in bulb weight.



The olfactory bulbs are a highly conserved but highly variable part of the vertebrate forebrain (Allison, 1953; Gittleman, 1991; Royet et al., 1998). They can be identified unambiguously in virtually every vertebrate, and unlike most CNS regions, establishing their homology across diverse vertebrate classes is simple. However, the size of the olfactory bulbs differs enormously across taxa. For example, large-brained apes, including humans, have comparatively small olfactory bulbs, whereas smaller-brained insectivores and rodents have comparatively large bulbs. This is probably the best example of a violation of allometric scaling of CNS components. Among a wide range of mammalian species, the correlation between bulb and brain weight is significantly lower (r = 0.7) than similar correlations for other parts of the CNS (r >0.95), the cerebellum, hippocampus, neocortex, and medulla included (Finlay and Darlington, 1995). This indicates that a subset of genetic and molecular mechanisms that are important in bulb development and growth are likely to be distinct and partly independent from those that modulate other parts of the forebrain. There is already excellent support for this hypothesis at the genetic level: for example, mutations in Pax6, Ncam, Igf1, and Fgf8 all produce drastic effects on the olfactory bulbs but have minimal effects on neocortex, hippocampus, striatum, or thalamus (Dellovade et al., 1998; Tomasiewicz et al., 1993; Cheng et al., 1998, Cremer et al., 1994).
     In this study we have initiated a complex trait analysis of the olfactory bulbs with the goal of mapping and characterizing genes and developmental mechanisms that generate variation in the size and structure of the bulbs. The size of the mouse olfactory bulbs vary at least three-fold, from 10 to 30 mg. This suggests that a set of quantitative trait loci (QTLs) that modulate bulb development may be especially tractable to quantitative genetic and behavioral analysis (Williams, 1998, 2000). It should be feasible to map, clone, and characterize key loci that contribute to variation in weight, glomerular architecture, levels of receptors and transmitters, and responsivity to experimental and environmental manipulations. Functional correlates between allelic variants and phenotypic variants at the glomerular, cellular, and subcellular levels should be comparatively easy to explore in this highly modular and well-laminated part of the brain.
     The bulbs have a fascinating developmental history that has provided additional motivation for this quantitative genetic analysis. There is, for example, a steady increase in numbers of glomeruli over a one-month period after birth (Pomeroy et al., 1990). Bulb maturation and numbers of neurons and glomeruli can be up- and down-regulated by environmental and genetic manipulations (Woo et al., 1987; Royet et al., 1989; Maruniak et al., 1989; Roselli-Austin and Williams, 1990; Brunjes, 1994; LaMantia et al., 2000). Even more remarkable is the retention of developmental processes in the bulbs of sexually mature mammals that are extinguished in almost all other CNS regions shortly after birth. These processes include the constant turnover of the olfactory receptor cell axonal input to the bulb and the steady replenishment of key inhibitory neurons—particularly periglomerular and granule cells—from a residual population of a few thousand stem cells located deep in the subventricular zone (Lois and Alvarez-Buylla, 1994, van der Kooy and Weiss, 2000). These features make the olfactory bulb a superb system in which to explore the genetic basis of neurogenesis and the genetic basis of neuronal plasticity in adult mammals.



Material and Methods

We used two related crosses of mice to map QTLs controlling bulb weight. The first consists of a set of 358 animals belonging to 35 BXD recombinant inbred strains. These strains were generated by crossing C57BL/6J (B6) and DBA/2J (D2) parental strains in the 1970s (BXD1 through 32) and 1990s (BXD33 through 42) (Taylor, 1989; Taylor et al., 1999). The second set of animals consists of reciprocal F1 and F2 intercrosses that were also generated by crossing B6 and D2 strains. The F2 animals were produced at the University of Tennessee by mating B6D2F1 and D2B6F1 mice as described in Zhou and Williams (1999): 103 animals were B6D2F2s and 71 were D2B6F2s (further abbreviated BDF2s and DBF2s). Mice were maintained at 20–24 °C on a 14/10h light-dark cycle in a pathogen-free colony at the University of Tennessee. Animals were fed a 5% fat Agway Prolab 3000 rat and mouse chow. The average age of BXD/Ty animals was 80 days, whereas that of F2 mice was 98 days. Total numbers of males and females are 349 and 321, respectively.
     [For analysis of the effects of age, sex, and brain weight on the bulb, we supplemented our analysis with an additional set of ~600 mice belonging to a wide variety of genotypes that are included in the Mouse Brain Library ( and databases at All breeding stocks were originally from the Jackson Laboratory ( These cases are not described in the print version published in Behavior Genetics, but are included in some of the figures in this web edition.]


Figure 1: Olfactory bulb cut

Figure 1. Olfactory bulb cut in the sagittal plane to illustrate the placement of the dissection cuts. The position of a dissection cut is represented by a dark bar flanked by light bars in four of the sections. Surface landmark criteria for making this cut are described in the text. As illustrated in the upper right and lower left micrographs, bulb weight does include a small component of the anterior olfactory nucleus (AON). These serial sections are space approximately 0.3 mm apart. The upper left section is approximately 2.2 mm lateral to the medial surface of the bulb. Scale bar in the lower right is 2 mm.



Fixation and dissection

Mice were anesthetized deeply with Avertin (1.25% 2,2,2-tribromoethanol and 0.8% tert-pentyl alcohol in water, 0.5–1.0 ml ip). Mice were perfused through the heart with 0.1 M phosphate buffered saline followed by 1.25% glutaraldehyde and 1.0% paraformaldehyde in 0.1 M phosphate buffer (2–4 min), and then with 2.5% glutaraldehyde and 2.0% paraformaldehyde in 0.1 M buffer (1–2 min) at an increased flow rate.
     Left and right olfactory bulbs were cut free from the remainder of the forebrain under a dissecting microscope with a small scalpel blade or a razor. The brain was placed down on its dorsal surface, and the edge of the blade was aligned perpendicularly across the ventral midline at the waist of the olfactory peduncle behind the ventral-caudal end of the glomerular surface of the bulb (Fig. 1). The bulbs were cut free along the broad line marked in figure 1, rolled on tissue paper, and immediately weighed individually to the nearest 0.1 mg on a digital balance. This particular level of section cuts through the anterior olfactory nucleus. The bulb weight therefore includes the entirely accessory olfactory bulb (approximately 5% of total bulb volume) and the rostral parts of anterior olfactory nucleus.

Figure 1B: Olfactory bulb ventral view Figure 1B. Image of the ventral surface of the olfactory bulbs that illustrates the sharp demarcation between the bulbs and the remainder of the forebrain. The knife cut was placed at the rostral border of the region of pigmentation.

     All data were recorded directly into a FileMaker Pro relational database. In general, values for right and left bulbs do not differ by more than 0.5 mg (5%). Variation within isogenic strains is also low (coefficients of variation are often under 3%) demonstrating the reliability of the dissection procedure. The following data were obtained for the majority of cases: sex, age, body weight, brain weight just prior to bulb dissection, and weight of the cerebellum and hippocampus. Data from runts (less than 10 g body weight), hydrocephalic animals, cases with incomplete data and statistically defined outliers were excluded (n = 29). QTLs were mapped with and without outliers and none of the main results depend on their exclusion. Original data files are available at


Regression analysis

Regression analysis was carried out using Data Desk 6.01 (Data Description, to explore covariance between bulb weight and other variables. We explored associations between the weight of the olfactory bulbs—expressed as the summed weights of the two bulbs—with variation in brain weight excluding the bulbs, sex, logarithm of age, body weight, and the weights of the cerebellum and hippocampus. Our aim was to protect our QTL analysis against genes that have widespread effects on body and brain weight and to focus attention on the subset of genes that have more intense or selective effects on the olfactory bulbs. Our second purpose was to understand relations among sex, age, and brain weight. Previous work has reported no gain in olfactory bulb weight in adult mice as a function of age (Pomeroy et al., 1990), but in our large data set there is compelling evidence for such a relation. As part of this analysis we also tested for possible interaction effects among variables, for example to determine whether the increase in bulb weight as a function of age varies between males and females. The regression analysis was performed independently for the BXD and F2 cases.

Genotyping and QTL mapping

DNA was extracted from the spleens of F2 animals using a high-salt procedure (Laird et al., 1991; A set of 129 microsatellite loci distributed across all autosomes and the X chromosome were typed in all F2s using a standard PCR protocol (Love et al., 1990, Dietrich et al., 1994) detailed by Lu and colleagues (2000). In brief, each 10 µl PCR reaction contained 1X PCR buffer, 1.92 mM MgCl2, 0.25 units of Taq DNA polymerase, 0.2 mM of each deoxynucleotide, 132 nM of the primers and 50 ng of genomic DNA. The microsatellite primer pairs were purchased from Research Genetics, Huntsville, AL. A loading dye (60% sucrose, 1.0 mM cresol red) was added to the reaction before the PCR (Routman et al., 1994). PCRs were carried out in 96-well microtiter plates. We used a high stringency touchdown protocol in which the annealing temperature was lowered progressively from 60 °C to 50 °C in 2 degree steps over the first 6 cycles. After 30 cycles, PCR products were run on 2.5% Metaphor agarose gels (FMC Inc., Rockland ME), stained with ethidium bromide, and photographed. F2 genotypes were read directly from photographs and entered into Excel 98 and transferred to Map Manager QTb28 for mapping and permutation analysis.
     Genomic DNA from 35 BXD RI strains was obtained from Dr. B. A. Taylor (Jackson Laboratory). We genotyped a set of 380 MIT markers distributed across all autosomes and the X chromosome using the PCR protocol described above. These new genotypes were pooled with a pre-existing set of genotypes for MIT loci (Taylor et al., 1999) downloaded from the Mouse Genome Database and the Portable Dictionary of the Mouse Genome. This high-resolution BXD genotype database includes data for 790 markers of which 640 have been genotyped for both the initial set and the new set of BXD strains. The BXD RI genotype database is available on our web site,
     QTLs were mapped and analyzed using Map Manager QT. This program implements simple and composite interval mapping methods described by Haley and Knott (1992). Genome-wide significance levels for assessing the confidence of the linkage statistics were estimated by comparing the highest likelihood ratio statistic (LRS) of correctly ordered data sets with LRSs computed for 10,000 permutations (Churchill and Doerge, 1994). The permutation analysis of the BXD data demonstrates that LRS scores of 8.9, 15.0, and 22.5 are associated with genome-wide p values of 0.5, 0.05, and 0.001 respectively. LRS scores can be converted to LOD scores by dividing by 4.6. The 2 LOD confidence interval—an approximate estimate of the 95% confidence band in which QTLs are actually located—is the length of the chromosomal interval over which the LOD score of linkage is within 2 units of the peak value.
     BXD RI strains and F2 progeny were initially mapped separately using conventional interval mapping methods. To improve the statistical power of our search for QTLs affecting the olfactory bulbs we eventually pooled mapping results from the F2 cross with those obtained from the initial screen of the BXD strains. The BXD and F2 crosses share the same B and D alleles but were derived using different breeding schemes. As a result, there are several important differences between these data sets, for example, the presence or absence of dominance deviation and the extent of linkage disequilibrium between neighboring loci. However, our method of pooling probabilities is insensitive to particular models of QTL action, and could be easily generalized to multiple crosses segregating for entirely different alleles. The main condition of the pooling procedure is that probabilities (or equivalently, LOD scores or LRS scores) be combined at the same marker locus or intervals defined using a common set of markers. We have used the very well mapped MIT markers for this purpose. This restriction minimizes problems that might arise by combining two or more independently computed maps. Cumulative probabilities for the pooled results at individual marker loci or common map positions were computed as the CHI2 probability of –2(lnP<BXD> + lnP<F2>) with four degrees of freedom. The corresponding formula used in our spreadsheet is "=CHIDIST(–2*(LN(0.001)+LN(0.002)),4)" where 0.001 and 0.002 are examples of point-wise probabilities for BXD and F2 data sets at a given map position. The cumulative p in this example is 0.0000282. Because we have combined point-wise probabilities for both groups of progeny, the QTLs that we have mapped should not be considered confirmed QTLs. However, all four QTLs for which we report data have genome-wide probabilities under 0.05.




The results are divided into two major sections. The first deals with normal variation in the size of the olfactory bulbs as a function of sex, age, brain and body weights. This systematic description of sources of variance in bulb weight is interesting in it own right, but is also a prerequisite to the second section of the results in which we summarize QTL mapping results. The rationale for the statistical treatment of phenotype data, as well as a more complete explanation of the linear regression methods, is provided in Williams (2000). In brief, we are attempting to ensure that the QTLs that we have mapped have specific or intense effects on the size of the mouse olfactory bulb.

Variation in bulb size in normal mice

Variation in the weight of the olfactory bulbs among normal sexually mature mice is substantial. Bilateral bulb weight ranges from 10 to 30 mg (Figs. 2, 3). Variation is substantial even among isogenic groups of mice. For example, among our sample of C57BL/6J animals between 30 and 300 days of age, bilateral bulb weight varied from 14.3 mg to 29.1 mg. After correcting for sex, age, body weight, and brain weight, differences within sets of isogenic animals are greatly reduced. The within-strain coefficient of variation averages 3% (Table 1). This matches the coefficient of variation in other parts of the mouse CNS (Williams et al., 1996, 1998; Williams 2000; Lu et al. 2000).
     Olfactory bulbs of the parental strains, B6 and D2, weigh 22.7 ± 0.31 and 19.8 ± 0.32 mg), respectively (Table 1). This 13% difference is highly significant (p < 0.001). Brains of B6 mice are typically 20% heavier than those of D2 mice, 496 ± 5.6 mg versus 415 ± 3.9 mg, and body weight is also greater, 24.7 versus 19.3 gm at 75 days. When bulb weights of the two strains are adjusted to account for differences in brain weight, body weight, sex, and age, the original 3-mg difference in olfactory bulb weight is reduced tenfold to 0.3 mg: 21.9 ± 0.4 versus 21.6 ± 0.3 mg (Table 1). This fact illustrates why it is so critical to assess multiple variables prior to mapping system-specific QTLs.
     The olfactory bulbs of the reciprocal F1 hybrids—BDF1 and DBF1—weigh 23.3 ± 0.3 and 24.3 ± 0.5 mg, respectively, an insignificant difference. Corresponding weights for the two subsets of F2 animals (BDF2 and DBF2) are 24.7 ± 0.13 and 24.9 ± 0.16 mg, respectively. The fact that both F1 and F2 progeny have bulbs that are larger than either parental strain suggests substantial heterosis for this trait. Heterosis is usually less pronounced in F2 than in F1 progeny (Falconer and Mackay, 1996). In this case, the bulbs of F2 progeny are slightly larger than those of F1 progeny, a result that may be due to maternal effect (superiority of the F1 mothers). Males of the two F2 subsets differ in their sex chromosomes, and both sexes differ in mitochondrial genotype. However, we were unable to detect any significant phenotypic differences between subsets of F2s.
     Olfactory bulbs of the parental strains, B6 and D2, weigh 22.7 ± 0.31(n = 50) and 19.8 ± 0.32 mg (n = 24), respectively (Table 1). This 13% difference is highly significant (p < 0.001). Brains of B6 mice are typically 20% heavier than those of D2 mice, 496 ± 5.6 mg versus 415 ± 3.9 mg, and body weight is also greater, 24.7 versus 19.3 gm at 75 days. When bulb weights of the two strains are adjusted to account for differences in brain weight, body weight, sex, and age, the original 3-mg difference in olfactory bulb weight is reduced tenfold to 0.3 mg: 21.9 ± 0.4 versus 21.6 ± 0.3 mg (Table 1, column 3). This small difference is insignificant, but as we show below these two strains still carry different alleles at four loci that have reciprocal effects on bulb weight.
Table 1. Olfactory bulb weight data for BXD and related strains


Bulb weight ± SE adjusted (mg)*


Bulb weight original (mg)¶



Brain weight (mg)

Body weight (g)


20.7 ± 0.2







21.1 ± 0.7







18.3 ± 0.5







22.0 ± 0.4







19.3 ± 0.2







20.0 ± 0.4







18.7 ± 0.4







20.0 ± 0.5







19.4 ± 0.2







19.9 ± 0.5







18.0 ± 0.5







19.7 ± 0.4







21.0 ± 0.4







18.6 ± 0.6







20.9 ± 0.5







17.3 ± 0.5







19.8 ± 0.7







18.7 ± 0.5







18.6 ± 0.7







18.3 ± 0.5







20.3 ± 0.3







19.7 ± 0.4







20.1 ± 0.6







20.7 ± 0.6







21.4 ± 0.6







20.1 ± 0.3







20.3 ± 0.4







23.2 ± 0.4







23.6 ± 1.0







18.1 ± 0.5







21.2 ± 2.0







22.5 ± 0.7







20.4 ± 0.3







20.3 ± 0.7







19.3 ± 0.1







21.9 ± 0.4







21.6 ± 0.3







23.3 ± 0.3







24.3 ± 0.5







20.3 ± 0.5






* Bilateral olfactory bulb weight after correction by multiple regression for differences in logarithm of age, sex, body weight, and brain minus olfactory bulb weight. All cases were adjusted to that of 75-day-old 22g female mice with a brain weight –bulb weight of 420 mg.
¶ Bilateral olfactory bulb weight after correction only for logarithm of age and sex. These values are very close to means of the original values. Data are adjusted to those expected of 75-day-old female mice.
‡ This strain is now apparently extinct.


Figure 2: Correlation of brain weight and bulb weight

Figure 2. Relation between bilateral olfactory bulb weight and weight of the brain. The brain was reweighed just before dissecting both olfactory bulbs for this set of 853 mice (BXD, F2, and several groups of inbred strains not used for QTL analysis). Crosses and the upper regression line represent males; circles and the lower regression line represent females. The regression line for the both sexes combined is bulb = 0.06(brain–bulb) – 6.0. Subtracting bulb weight from that of the whole brain and recomputing the regression does not change the slope significantly but does change the y-intercept to –4.6 mg. The average bilateral sex difference in bulb weight is about 0.5 mg.

Brain weight and olfactory bulb weight

Variation in brain weight is the single most important predictor of variation in bulb weight among genetically diverse groups of adult mice. Fifty percent of the variation in bulb weight can be predicted given data on brain weight (Fig. 2, Table 2). For the BXD mice treated as individuals rather than as strain means, the line of best fit is olfactory bulb weight (mg) = 0.054(brain weight in mg) – 3.51. Brain weight naturally includes the weight of both bulbs; therefore, to compare structurally independent variables, we subtracted olfactory bulb weight from brain weight and recomputed regression equations. Although the slope is the same, the amount of variance explained is reduced to 45%. Correlations between the olfactory bulbs and other CNS components are surprisingly variable; the correlation between bulb weight and hippocampal weight is only 0.35, whereas that with the cerebellum is 0.70 (Table 2).
     Among the heterogeneous population of F2 mice, variation in brain weight is also the most important predictor of variation in bulb weight: 44% of variance is accounted for by total brain weight, whereas 38% is accounted for by the weight of the brain minus the weight of both bulbs.


Figure 3: Increase in bulb weight with age
Figure 3. Progressive increase in olfactory bulb weight as a function age. Crosses represent a sample of C57BL/6J mice, circles represent a sample of A/J mice. The overall regression relation (central line labeled all) is based on a 1162 cases.

Age and olfactory bulb weight

One of the more important results of this study is that the olfactory bulbs grow substantially even in sexually mature mice. In retrospect, given the constant influx of neurons via the rostral migratory stream, this gain might be expected. Among our sample of BXD animals, which range in age from 30 to 300 days, the slope of this increase is approximately 6.5 mg for a 10-fold increase in age. Mice are sexually mature by 50 days of age, and over the next 200 days the summed weight of the two bulbs increases by 6.0 ± 1.0 mg. This amounts to 0.2 mg/week, a 1.0% increase. Variation in age among genetically diverse BXD mice accounts for 20% of the variance in olfactory bulb weight. In the isogenic parental strain, C57BL/6J, a remarkable 54% of the variance in bulb weight is accounted for by age (Fig. 3, upper line and crossed points, Table 2). We know that brain weight increases in adult mice as a function of age (Lu et al., 2000) and therefore some increase in bulb weight is expected. However, there is a significantly greater proportional increase in bulb weight. After controlling for increases in brain weight as a function of age, the olfactory bulbs still gain 4.6 mg between 50 and 250 days of age.
     In our set of F2 progeny there is also a significant upward trend in olfactory bulb weight between 75 and 150 days that amounts to approximately 0.14 mg/week (p = 0.002). Correcting for the gain in brain weight reduces the slope for the bulb to 0.09 mg/week (p = 0.012), a value that is significantly higher (proportionally) than that of other brain regions we have examined—hippocampus included (Lu et al., 2000).


Sex and olfactory bulb weight

On average, olfactory bulbs of males weigh 0.35 mg more each than those of females (Fig. 2, p <0.0001). In the sample of BXD mice, total brain weight of females at dissection averaged 8 mg more than that of males—432.3 ± 2.8 versus 424.2 ± 3.5 mg (p of 0.17 when brains weights are corrected for variation in age). After compensation for differences in brain weight and age, the adjusted sex difference in bulb weight between male and female BXD mice amounts to 0.7 mg bilaterally (p < 0.006). The same 0.7 mg difference was also noted among F2 intercross progeny. Although statistically significant and reliable across data sets, the 3.5% mean difference in olfactory bulb weight between the sexes is modest. While any statistically significant sex difference is often, and incorrectly, referred to as a sexual dimorphism, the overlap between sexes is very substantial and certainly does not produce two modes. Sex accounts for only 0.3% of the total variance in olfactory bulb weight.

Body weight and olfactory bulb weight

Body weight has a significant correlation of 0.56 with olfactory bulb weight in BXD mice (Table 2). An increase in body weight of 1 g is associated with an increase in bulb weight of 0.17 ± 0.01 mg per side. There is still a highly significant association between the weight of the body and that of the olfactory bulb, even after we control for variation in brain weight and sex. However, strong association between body and bulb is almost purely due to a common primary effect of age. After the addition of either age or the logarithm of age to a multiple regression analysis, body weight becomes an insignificant predictor of bulb weight (p = 0.21). In F2 mice body weight correlates only weakly with olfactory bulb weight (r = 0.26). Variation in body weight thus accounts for only ~7% of the variation in bulb weight. An increase in body weight of 1 gram is associated with an increase in olfactory bulb weight of 0.12 ± 0.035 mg.  


Table 2. Correlation matrix for 8 traits among individual BXD animals









log age

olfactory bulb
















































sex (F=1)








log age








The lower left half of this table lists correlations between variables. All correlations were computed using individual values from BXD animals (n = 254 to 351). In the upper right half lists coefficients of determination (R2).

Correlations greater than 0.138 are significant (P <0.05). Correlations greater than 0.181 are significant (P <0.01).

Abbreviations: OB: olfactory bulbs; CB: cerebellum; HP: hippocampus; F=1 means that for purposes of regression analysis we assigned females a value of 1 and males a value of 0.


Comparison of right and left bulbs

Measured differences in weights of right and left bulbs are due to biological differences and technical error. The mean difference between the two sides averages 0.43 mg. Following a correction for small n, this corresponds to a right-left coefficient of variation of 4.0% (Gurland and Tripathi correction; Sokal and Rohlf, 1995). This value sets an upper limit on variation generated by developmental noise and the magnitude of error introduced by fixation and dissection.
     The mean weights of right and left bulbs across all cases differ by merely 0.039 mg in F2 mice and by 0.054 mg in BXD mice. The right side has a miniscule weight advantage (0.3–0.5%) that does not reach the threshold of statistical significance (p = 0.2 in F2 and p = 0.1 in BXD).

Multiple regression analysis

To map QTLs that have specific or relatively intense effects on the olfactory bulbs, we corrected all bulb weight data by multiple linear regression using the function

olfactory bulb weight (bilateral, fixed in mg) = 4.0 (logarithm of age in days)+ 0.059 (body weight in gm) + 0.046 (brain weight – olfactory bulb weight in mg) – 0.59 (if female) – 7.24

This linear model accounts for 58% of the variance among a heterogeneous set of just over 600 animals. The predicted olfactory bulb weight of a 75-day-old 22-gm female with a brain weight (minus the bulbs) of 430 mg is 19.86 mg. The difference between the actual weight and the predicted bulb weight is the residual value. These residuals (or the equivalent regression adjusted olfactory bulb weights) were used for mapping.
     A similar multiple linear regression model was used to create a set of adjusted values for the BDF2 mice:

olfactory bulb weight (bilateral, fixed in mg) = 4.59 (logarithm of age in days) – 0.0067 (body weight in gm) + 0.051 (brain weight – olfactory bulb weight in mg) – 0.63 (if female) –6.99

This equation accounts for approximately 42% of the variance among the F2 sample (n = 177). The mean of the adjusted weights for the F2 mice is 24.8 ± 0.1 mg (SE). The cumulative probability density plot for the F2 sample extends from 21 mg to 28 mg and has a prominent mode at approximately 24.5 mg and a minor mode at 26 mg.

Strain differences in bulb weight

At 75 days of age, the weight of the olfactory bulbs ranges from a low of 18.1 mg in BXD36 to a high of 23.6 mg in BXD35 (Table 1). Values adjusted only for age and sex average 20.5 ± 2.1 mg (SD) across the BXD strains, and range from a low of 17.3 mg in BXD24 to a high of 24.4 mg in BXD35. The range greatly exceeds the difference between D2 and B6 parental strains (19.8 versus 22.7 mg unadjusted; 21.6 versus 21.9 mg adjusted).
     Strain averages for olfactory bulb weight of BXD strains are characterized by two modes centered at approximately 18.5 and 21.0 mg (Fig. 3A). An unexpectedly large number of strains (n = 9) have weights less than that of the low parental strain (Table 1). A probability density plot of BXD strain averages after correction for differences in brain weight and other parameters almost eliminates the bimodality but increases the excess of strains with smaller than expected olfactory bulb weights (Fig. 4B). Both parental strains and the reciprocal F1s all have appreciably larger bulbs than all but one or two of the BXD strains.
     The size of the bulbs as a fraction of total brain weight averages 4.7% in BXD strains. The range extends from 4.0% in BXD21 to 5.5% in BXD35 at 75 days of age. These values shift upward very slightly as a function of age.

QTL analysis of variance in bulb weight

An analysis of variance of BXD mice indicates that the broad-sense heritability of olfactory bulb weight ranges from 20% when computations are performed using completely unadjusted data to 27% for fully corrected values (Williams et al., 1996). The corresponding estimates of broad sense heritability derived by comparing variance of F1 and F2 progeny is ~33%. As noted above, after compensating for differences in brain weight there is essentially no difference in the weight of the bulbs between the parental strains (Table 1). In this situation it is not useful to apply the Castle-Wright equation to estimate the minimum number loci that modulate a quantitative trait (Wright, 1978). The output of the QTL analysis itself is far more germane in estimating numbers of modulatory loci (Grisel et al., 1997).


Figure 4: BXD and F2 Probability Densities

Figure 4. Probability density plots of olfactory bulb weights in 35 BXD strains. A. Olfactory bulb weight data adjusted only for variation in sex and age and prior to compensation by linear regression for variation in brain weight. The small distributions labeled D, B, and F1 are for the two parental strains and the two reciprocal F1 hybrids computed using the group average and the standard error of the mean (see Williams et al., 1996 for details). The parental and F1 distributions have been reduced in size relative to the probability density of the BXD set (bold curve). The fine normal distribution was computed using the average and standard deviation for all BXD strain averages. B. Similar probability density plot of olfactory bulb weight, but now with compensation for variation in brain weight. The bimodality evident in A is greatly reduced. The distribution in B was used for mapping and is based on values in the second column of Table 2.



Mapping olfactory bulb QTLs in BXD mice

After factoring out much of the variability in bulb weight associated with brain weight, sex, body weight, and age, we were able to map four QTLs. We have used both simple and composite interval mapping to define chromosomal regions associated with differences in bulb weight at a genome-wide significance level of 0.05. All data on the centimorgan positions of QTLs and microsatellite marker loci (the MIT loci) in this section are taken from online mouse chromosome committee reports maintained as part of the Mouse Genome Database (Mouse Genome Database). [For sake a legability, locus symbols in this web edition have often not be set in an italic font.]
     Among the BXD strains we detected a particularly good match between variation in bulb weight and the distribution of B and D alleles in the interval between the markers D17Mit267 and D17Mit133 on the proximal part of Chr 17 (3 to 10 cM). As assessed by simple interval mapping, the association between differences in weight and alleles on Chr 17 has a likelihood ratio statistic (LRS) of 13.3 at the marker D17Mit113, equivalent to a LOD score of 2.9 and a point-wise p of 0.00026 (Fig. 5G). We estimated the genome-wide probability of linkage by permutation and found that 12% of 10,000 permutations had peak LRS scores as high as the correctly ordered data. This indicates that the genome-wide probability associated with a putative QTL on Chr 17 (or equivalently, the type 1 error) is 0.12 when BXD strains alone are considered. An equivalent analysis of the F2 progeny (5H), described in more detail below, provided a clear validation of this interval on proximal Chr 17, subsequently called the Bulb4 locus.
     The average weight for 18 BXD strains with a BB genotype at the marker D17Mit113 was 19.26 ± 0.23 mg, whereas that for 16 strains with a DD genotype was 20.86 ± 0.34 mg, a 1.6 mg difference. A single D allele in this interval therefore has an additive effect of +0.8 mg on bulb weight. The correlation between olfactory bulb weight and alleles at D17Mit113 is 0.48, suggesting that as much as 25% of the genetic variance and 5–10% of the total phenotypic variance is generated by this putative QTL.
     The BXD data highlighted three other intervals with comparatively high LRS scores using simply interval mapping (Fig. 5A, C, and E). All three intervals were subsequently verified in the analysis of F2 intercross progeny:


  1. Chr 4 between D4Mit322 and D4Mit332 (36–54 cM) with an LRS that peaks at 9.3 (p = 0.0025) at the marker D4Mit186 (44 cM according to the Chr 4 Committee Report). B alleles have an additive effect of 0.7 mg each (Fig. 5A).

  2. Chr 6 between D6Mit16 and D6Mit29 (30–36 cM) with an LRS that peaks at 10 at the marker D6Mit19 at 34 cM (p = 0.0016). B alleles at this marker are associated with an additive effect of 0.7 mg (Fig. 5C).

  3. Chr 11 between D11Mit51 and D11Mit20 (18–20 cM) with an LRS that peaks at 13.2 (point-wise p = 0.00028). D alleles have an additive effect of 0.7 mg at the marker D11Mit231 (17.5 cM). In figure 5E, the D allele effect is plotted as a negative B allele effect.  

Bulb1 BXD

Figure 5A Interval maps of Bulb1 on Chr 4 mapped using BXD strains. The thick black line corresponds to the BXD LRS values. The LRS = 4.6xLOD score and was computed at 1-cM intervals. The thin black line and the right axis indicates the estimated mean effect of replacing a single D allele with a B allele. The x-axis is the approximate positions of marker loci along the chromosomes measured in centimorgans (1 cM = approximately 2 million base pairs of DNA). The x-axis have been corrected to the chromosome committee report consensus values.


Bulb1 F2

Figure 5B Interval maps of Bulb1 on Chr 4 mapped using F2 progeny. The thick black line corresponds to the LRS values. The LRS = 4.6xLOD score and was computed at 1-cM intervals. The thin black line and the right axis indicates the estimated mean effect of replacing a single D allele with a B allele.



Bulb2 BXD

Figure 5C Interval maps of Bulb2 on Chr 6 mapped using BXD strains. There is a good possibility that Bulb2 represents two linked QTLs. Other conventions as in Fig. 5A above.


Bulb2 F2

Figure 5D Interval maps of Bulb2 on Chr 6 mapped using F2 progeny.



Bulb3 BXD

Figure 5E Interval maps of Bulb3 on Chr 11 mapped using BXD strains.


Bulb3 F2

Figure 5F Interval maps of Bulb3 on Chr 11 mapped using F2 progeny.



Bulb4 BXD

Figure 5G Interval maps of Bulb4 on Chr 17 mapped using BXD strains.


Bulb4 F2

Figure 5H Interval maps of Bulb4 on Chr 17 mapped using F2 progeny. Other conventions as in Fig 5A above.


     We controlled for variation associated with the two intervals with the highest LRS scores (Chrs 11 and 17) using the composite interval mapping technique and searched for secondary QTLs affecting olfactory bulb weight. This procedure revealed an unusually high LRS on Chr 10 between D10Mit79 and D10Mit3 (5–21 cM). The LRS peaks at 15.9 (point-wise p = 0.00007). A set of 10,000 permutation tests was run with control for D11Mit231 and D17Mit113 and this Chr 10 interval almost reaches the genome-wide significance threshold (p = 0.06). However, there is no support at all for this interval in the F2 cross.


Mapping olfactory bulb QTLs using the F2 intercross

Analysis of the F2 data set provides strong support for all four primary intervals identified using the BXD set (Fig 5B, D, F, H). In each case the polarity of effects of B and D alleles is the same in the F2 as in the BXD set.
     Bulb1 (Fig. 5B). The LRS peaks at 6.8 in the same interval between D4Mit178 (30.6 cM) and D4Mit332 (54 cM) thereby supporting the same region highlighted in the analysis of BXD strains. The LRS rises to 11.4 when we control for the other three intervals on chromosomes 6, 11, and 17, represented respectively by the markers D6Mit291, D11Mit51, and D17Mit135. The pooled point-wise p of this Chr 4 locus using BXD and F2 data sets is 0.0003, equivalent to an LRS of ~16.0 in BXD and F2 data sets at the marker D4Mit186, and significant at a genome-wide probability of <0.05. The 2-LOD confidence interval of this QTL extends from approximately 25 to 60 cM. A single B allele contributes an average of about 0.4 mg bilaterally to bulb weight.
     Bulb2 (Fig. 5D). Strong linkage was also verified on Chr 6, with a peak LRS of 16.0 just proximal to the marker D6Mit291 (55 cM). The LRS for the F2 data has a genome-wide p less than 0.05 and defines a second QTL, Bulb2, even without the support of the BXD data. The 2-LOD confidence interval of this QTL is also broad and extends from 20 to 65 cM. When data from BXD and F2 sets are combined, Bulb2 is most likely to map between 50 and 65 cM on distal Chr 6. The cumulative genome-wide probability supporting Bulb2 is well under 0.01 (cumulative point-wise p is <0.00001). There is a good possibility that Bulb2 represents two linked QTLs; one located proximally between 20 and 25 cM (close to D6Mit16 and D6Mit188) and a second located more distally as described above. This would be consistent with the broad LRS map in the F2 and the bimodal map generated using the BXD set. Individuals with BB, BD, and DD genotypes at D6Mit291 have mean adjusted olfactory bulb weights of 25.3 ± 0.17 mg, 24.7 ± 0.16 mg, and 24.4 ± 0.17 mg, respectively. Bulb2 is responsible for ~7% of the total phenotypic variance in olfactory bulb weight of an age-adjusted population of F2 animals. The additive effect of a single B-to-D allele substitution is ~0.5 mg.  


     Bulb3 and Bulb4 (Fig. 5F and H). The F2 genome-wide analysis also highlights the Chr 11 (Bulb3) and Chr 17 (Bulb4) intervals. With control for the first two QTLs, the LRS on Chr 11 peaks at 11.8 near D11Mit51 (15 cM) whereas the LRS on proximal Chr 17 peaks at 14.6 near D17Mit135 (6.6 cM). The 2-LOD support intervals for these loci are ~20 cM and 10 cM, respectively. Both intervals match up nicely with intervals identified in the BXD set. When F2 and BXD data are combined, both intervals have cumulative point-wise probabilities less than 0.0003, equivalent to genome-wide p values of less than 0.01. Bulb3 and Bulb4 have effect polarities opposite to those of Bulb1 and Bulb2. Alleles inherited from DBA/2 are associated with larger bulbs, and in both cases the effect amounts to ~0.5 mg per allele in the F2 cross.
     Composite interval mapping in which we control for three of the QTL intervals and remap the fourth interval supports, and in one case (Bulb2) significantly improves, the linkage statistics. The LRS associated with Bulb2 is boosted from 15.5 to 21.0
     When all four of these QTLs are considered together, as much as 20% of the total phenotypic variance and nearly half of the genotypic variance can be explained among the F2 sample. The cumulative LRS score for all four intervals rises to 45, indicating relative independence and additivity of the Bulb loci. These same four intervals account for as much as 38% of the variance among the BXD strains.




We tested for pair-wise epistatic interactions among the four intervals with the largest effects in the BXD setÑthose on chromosomes 4, 6, 11, and 17. No significant nonlinear interactions were uncovered. The effects of the Chr 4 and Chr 6 intervals add together and the B/B two-locus genotype has a bulb weight 1.9 mg greater than that of the D/D genotype. The corresponding difference among the F2 animals with these genotypes is very closely matched at 1.7 mg. The double heterozygotes (H/H) had an intermediate bulb weight. Conversely, the effects of the Chr 11 and Chr 17 intervals add together and the B/B genotype has a bulb weight 2.4 mg less than that of the D/D genotype. In the F2 the difference between these two genotypes is 1.9 mg; again a fairly close match between data sets. The cumulative modulation of all four intervals acting together is not as great as expected, indicating possible saturation of effects on the bulb, weak epistasis, or overestimation of single QTL effects. Eight BXD strains had the four-locus genotype expected to have the largest bulb size (B/B/D/D) and five strains had the opposite four-locus genotype expected to have the smallest bulb size (D/D/B/B). These two extreme groups had average bulb weights that differed by 3.25 mg, not the predicted 5.5 mg difference expected if allelic effects at all four loci added linearly. It is likely that single QTL effects are more modest than our initial estimates.




The four QTLs that we have mapped modulate the weight of the olfactory bulbs over a range of 3–5 mg, equivalent to a 20% gain or loss in weight. These QTLs are the first members of what is probably a larger set of normally polymorphic genes that influence bulb development. The difference in bulb size among the BXD recombinant inbred strains is sufficiently large to motivate a search for underlying anatomical effects on cell populations and numbers of glomeruli. This may even be a large enough difference to motivate a search for corresponding behavioral variation. The four bulb-specific QTLs identified in this cross between C57BL/6J and DBA/2J can be divided into pairs in which alleles inherited from the parental strains that have opposing effects of 0.5 to 1.0 mg each.


Relations between structure and behavior

Research in behavioral genetics typically progresses directly from allelic variants to behavioral variants without the intermediate analysis of transmitter systems, cells, or CNS nuclei. There are good reasons for side-stepping the neuroanatomy. Quantitative analysis of structural variation is often difficult and may lack sufficient power or precision to demonstrate functional relations with behavior. But studies of behavioral variants eventually need to return to the neuroanatomical and molecular biology of the brain. In the present study we have reversed the sequence and have started by establishing linkage between allelic variants and structural variants (Williams, 2000). The 30–40% difference in bulb weight between animals that we have exploited to map QTL is an excellent target for correlated behavioral analysis. For example, Gheusi and colleagues (2000) have recently shown that a 30–40% reduction in bulb size associated with blocking the movement of cells along the rostral migratory stream in mice leads to a significant loss in the ability to discriminate odors. It will be interesting to determine if normal variants that match experimental variants in terms of magnitude of the phenotypic effect will show similar behavioral differences.
     From a technical perspective a major advantage of the olfactory system is that the bulbs are neatly defined. Input and output relations of major cell classes—mitral cells, tufted cells, granule cells, and periglomerular cells—are comparatively well understood at the synaptic level (Shepherd and Greer, 1998). A major advantage is the division of the outer third of the olfactory bulb into a mosaic of approximately 1500–2000 glomeruli that receive restricted input from subsets of receptor cells. It will be interesting to compare data on weight with data on numbers of glomeruli and individual cell types. Weight among BXD mice varies from 17 mg in BXD24 to 24 mg in BXD35. Does this variation correlate with the area of the olfactory receptor epithelium, variation in numbers or size of glomeruli, or possibly, differences in the flow of neurons from the rostral migratory stream? From a behavioral standpoint, the olfactory system of mice is particularly amenable to quantitative assays of chemosensory sensitivity and discrimination ability (Coopersmith and Leon, 1984; Gheusi et al., 2000) using a variety of well-characterized laboratory tests. It will be feasible to score some of the most extreme strains to assess the prospects of a correlative behavior analysis of mice with large and small olfactory bulbs.


Weight and Number of Glomeruli

Figure 6. Tight correlation between the size of the olfactory bulb and the number of glomeruli. Data were extracted from a developmental analysis of the CF1 strain (Pomeroy et al., 1990; their figures 3C and 4). Bulb size, reexpressed here in milligrams, is an estimate based on the volume measured in histological sections. The equation of the regression line is n glomeruli = 244 + 200(bulb weight in mg).


Relation between glomerular architecture and bulb weight

The weight of the bulb increases from 0.8 mg at birth to 8.0 mg at 80 days (Pomeroy et al., 1990). Over this period the number of glomeruli increases approximately 5-fold, from 400 to 2000, by a process in which small new glomeruli intercalate between large old glomeruli (LaMantia et al., 1992). The number of synapses per glomerulus increases nearly 30-fold, from 7000 to 200,000 (Pomeroy et al., 1990). Correlations among these variables are high. Our reanalysis of figures reproduced in the paper by Pomeroy and colleagues (1990) demonstrates that the correlation between the logarithm of age and number of glomeruli is 0.95, whereas that between bulb volume and number of glomeruli is 0.99 (Fig 5). For each 1 mg increase in the weight of the bulb, numbers of glomeruli increase by 200. We do not yet know whether these strong correlations apply with equal force across different genotypes, but if they do, then BXD35 mice should have an average of 700 more glomeruli per bulb than do BXD24 mice (Table 1). There are no technical impediments to determining cellular (and glomerular) correlates of differences in bulb weight other than the requirement to apply relatively high-throughput quantitative morphometric methods. Even this problem has been partly solved: high-resolution images (5 µm/pixel) of serial Nissl-stained sections cut through the bulbs of all strains included in this study are now available online as part of the Mouse Brain Library collection (Rosen et al., 2000; Williams and Rosen, It should be practical to rapidly assess the degree to which different layers and parts of the olfactory bulb (and accessory olfactory nucleus) vary as a function of bulb weight using the stereological method detailed so carefully by Royet and colleague (1989, 1998).
     In contrast to the variation in bulb weight and glomerular number, the number of olfactory receptor genes is a fixed feature of the mouse genome (Buck and Axel, 1991; Axel, 1995; Buck 1996). Variation in weight may be associated with marked differences in patterns of receptor axon convergence onto individual glomeruli (Royet et al., 1989; 1998, Vassar et al., 1994; Ressler et al., 1994).


A comparison with QTLs that modulate size of hippocampus and cerebellum

In recent work we have used the same groups of mice to map QTLs that modulate the size of the eye (Zhou and Williams, 1999), the brain (Strom, 1999), the dorsal lateral geniculate nucleus (Kulkarni et al., 2000), the caudate (Rosen and Williams, 2000), the hippocampus (Lu et al., 2000), and the cerebellum (Airey et al. 2000). Two of these areas— the cerebellum and hippocampus—are of particular interest. Like the olfactory bulbs, both possess large populations of late-generated granule cells. Polymorphic genes that influence late stages of neurogenesis might be expected to have common effects on all three structures (Finlay and Darlington, 1995). The correlations between weights of these three structures given in Table 2, indicates a much closer coupling between bulb and cerebellum (R2 = 0.49, than between bulb and hippocampus (R2 = 0.13). QTLs for cerebellum map to chromosomes 1, 8, 14, and 19, whereas those for hippocampus map to distal Chr 1 and the proximal Chr 5 (Lu et al., 2000). In contrast, the four Bulb loci map to chromosomes 4, 6, 11, and 17. Thus the Bulb loci do not overlap any of hippocampus- or cerebellum-specific QTLs. The apparent lack of shared genetic control does not mean that cell proliferation and growth in the bulb is autonomous from that in the hippocampus and cerebellum. In fact, we know that genes such as Zipro1 (Yang et al., 1999) are expressed in all three of these granule-cell-rich regions. We have not yet detected QTLs with widespread effects primarily because the regression techniques intentionally highlights QTLs with selective effects.


Effects of age on the size of the olfactory bulbs

We have shown that the olfactory bulb grows significantly in adult mice. This finding amplifies earlier work that unequivocally demonstrated the addition of new glomeruli in young but sexually mature mice (Pomeroy and et al., 1990). In contrast to previous work (Maruniak et al., 1989; LaMantia et al., 1992; Baker et al., 1995), our results also demonstrate that olfactory bulb size does not reach an adult asymptote until very late in life—probably after 300 days of age in strains such as C57BL/6J and A/J. This continued growth appears close to linear when weight data are plotted against the logarithm of age (Fig. 3, note for example the circular data points from strain A/J). Naturally, on a conventional linear time scale, the pace of olfactory bulb growth decelerates at a neat logarithmic rate. This is consistent with the reduction in numbers of neurons delivered to the olfactory bulbs over the rostral migratory stream as a function of age (Tropepe et al., 1997). [To explore the limits of bulb growth, and of glomerular addition and adult neuron proliferation, will require access to genetically uniform mice that are significantly older than 300 days. Such mice are available from the mouse colony supported by the National Institute of Aging (see]
     Effects of age on bulb size and architecture may vary among species. The contrast between mature mice and humans is striking. A preliminary analysis of humans (n = 8) suggests that the weight of the bulb decreases with age—from 50 mg in young adults to less than 40 mg in very old humans (Bhatnagar et al., 1987). Mitral cell number apparently decreases even more steeply than does bulb weight. If these findings were replicated in a larger sample the aging process in humans would provide an interesting developmental and ethological counterpoint to that of rodents.


Candidate genes for the Bulb QTLs

Once the chromosomal positions of the QTLs have been determined to a precision of 1–3 cM, it will be appropriate to assess merits of candidate genes (Rikke and Johnson, 1998). To achieve this level of precision we have generated a large set tenth-generation advanced intercross progeny (Darvasi, 1998; G. Zhou and R. Williams, unpublished) using as starting material the F2 animals mapped in the present study. We are also developing a new technique for high-resolution (subcentimorgan) QTL mapping that like RI strain mapping, obviates the need to genotype any animals (Williams et al., 2000). Given the complexity of bulb development, candidates within QTL intervals will include genes known to be expressed in the olfactory epithelium, the cephalic neural crest, the forebrain, the subventricular zone, the rostral migratory stream, and of course, the olfactory bulb itself.



This research project was support by a grant from the National Institute of Neurological Disorders and Stroke (R01 NS35485). The authors thank Drs. Jing Gu and Xiyun Peng for their assistance in generating, processing, and genotyping F2 and BXD mice. We thank Ms. Leslie Jones for her help generating data for inbred strains. We thank Kathryn Graehl for editing this paper.



Airey DC, Lu L, Williams RW (2000) Gene loci for mouse cerebellum and internal granule layer size. Soc Neurosci Abst 26: in press.

Allison AC (1953) The morphology of the olfactory system in vertebrates. Biol Rev 28:195–244.

Axel R (1995) The molecular logic of smell. Sci Am 1273:154–159.

Baker H, Franzen L, Stone D, Cho JY, Margolis FL (1995) Expression of tyrosine hydroxylase in the aging, rodent olfactory system. Neurobiol Aging 16:119-128.

Benson TE, Ryugo DK, Hinds JW (1984) Effects of sensory deprivation on the developing mouse olfactory system: a light and electron microscopic, morphometric analysis. J Neurosci 4:638–653.

Bhatnagar KP, Kennedy RC, Baron G, Greenberg R (1987) Number of mitral cells and bulb volume in the aging human olfactory bulb: a quantitative morphological study. Anat Rec 218:73–87.

Brunjes PC (1994) Unilateral naris closure and olfactory system development. Brain Res Rev 19:146–160.

Buck L, Axel R (1991) A novel multigene family may encode odorant recognition. Cell 65:175–187.

Buck LB (1996) Information coding in the vertebrate olfactory system. Annu Rev Neurosci 19:517–544.

Cheng CM, Joncas G, Reinhardt RR, Farrer R, Quarles R, Janssen J, McDonald MP, Crawley JN, Powell-Braxton L, Bondy CA (1998) Biochemical and morphometric analyses show that myelination in the insulin-like growth factor 1 null brain is proportionate to its neuronal composition. J Neurosci 18:5673–5681.

Churchill GA, Doerge RW (1994) Empirical threshold values for quantitative trait mapping. Genetics 138:963–971.

Coopersmith R, Leon M (1984) Enhanced neural response to familiar olfactory cues. Science 225:849–851.

Cremer H, Lange R, Christoph A, Plomann M, Vopper G, Roes J, Brown R, Baldwin S, Kraemer P, Scheff S, et al (1994) Inactivation of the N-CAM gene in mice results in size reduction of the olfactory bulb and deficits in spatial learning. Nature 367:455–459.

Darvasi (1998) Experimental strategies for the genetic dissection of complex traits in animals. Nat Gen 18:19–24.

Dellovade TL, Pfaff DW, Schwanzel-Fukuda M (1998) Olfactory bulb development is altered in small-eye (Sey) mice. J Comp Neurol 402:402–418.

Dietrich WF, Miller JC, Steen RG, Merchant M, Damron D, Nahf R, Gross A, Joyce DC, Wessel M, Dredge RD, et al. (1994) A genetic map of the mouse with 4,006 simple sequence length polymorphisms. Nat Genet. 7:220–245.

Finlay BL, Darlington RB (1995) Linked regularities in the development and evolution of the mammalian brain. Science 268:1578–1584.

Gheusi G, Cremer H, McLean H, Chazal G, Vincent JD, Lledo PM (2000) Importance of newly generated neurons in the adult olfactory bulb for odor discrimination. Proc Natl Acad Sci USA 97:1823–1828.

Gittleman JL (1991) Carnivore olfactory bulb size: allometry, phylogeny and ecology. J Zool Lond 225:253–272.

Grisel JE, Belknap JK, O'Toole LA, Wenger CD, Crabbe JC (1997) Quantitative trait loci affecting methamphetamine responses in BXD recombinant inbred mouse strains. J Neurosci 17:745–754.

Haley CS and Knott SA (1992) A simple regression method for mapping quantitative trait loci in line crosses using flanking markers. Heredity 69:315–324.

Kulkarni A, Airey DC, Williams RW (2000) Genetic architecture of the mouse retinogeniculate system: a QTL analysis of numerical matching. Soc Neurosci Abst 26: in press.

Laird PW, Zijderveld A, Linders K, Rudnicki M, Jaenisch R, Berns A (1991) Simplified mammalian DNA isolation procedure. Nucleic Acids Res 19:4293.

LaMantia AS; Pomeroy SL; Purves D (1992) Vital imaging of glomeruli in the mouse olfactory bulb. J Neurosci 12:976–988.

LaMantia AS; Bhasin N; Rhodes K; Heemskerk J (2000) Mesenchymal/epithelial induction mediates olfactory pathway formation. Neuron in press.

Lois C, Alvarez-Buylla A (1994) Long-distance neuronal migration in the adult mammalian brain. Science 264:1145–1148.

Love JM, Knight AM, McAleer MA, Todd JA (1990) Towards construction of a high resolution map of the mouse genome using PCR-analyzed microsatellites. Nucleic Acids Res 18:4123–4130.

Lu L, Airey DC, Williams RW (2000) Genetic architecture of the mouse hippocampus: Identification of loci with specific effects on hippocampal size. J Neurosci, in rev, <>

Maruniak JA; Taylor JA; Henegar JR; Williams MB (1989) Unilateral naris closure in adult mice: atrophy of the deprived-side olfactory bulbs. Brain Res Dev 47:27–33.

Pennisi E (2000) Mouse sequencers take up the shotgun. Science 287:1179–1181.

Pomeroy SL, LaMantia AS; Purves D (1990) Postnatal construction of neural circuitry in the mouse olfactory bulb. J Neurosci 10:1952–1966.

Ressler KJ, Sullivan Sl, Buck LB (1994) Information coding in the olfactory system: evidence for a stereotyped and highly organized epitope map in the olfactory bulb. Cell 79:1245–1255.

Rikke BA, Johnson TE (1998) Towards the cloning of genes underlying murine QTLs. Mamm Genome 9:963–968.

Rosen GD, Williams RW (2000) Stereological and quantitative genetic analysis of the mouse caudate nucleus. Soc Neurosci Abst 26, in press.

Rosselli-Austin L, Williams J (1990) Enriched neonatal odor exposure leads to increased numbers of olfacotry bulb mitral and granule cells. Dev Brain Res 51:135–137.

Royet JP, Jourdan F, Ploye H, Souchier C (1989) Morphometric modifications associated with early sensory experience in the rat olfactory bulb: II. Stereological study of the population of olfactory glomeruli. J Comp Neurol 289:594–609.

Royet JP, Distel H, Hudson R, Gervais (1998) A re-estimation of the number of glomeruli and mitral cells in the olfactory bulb of rabbit. Brain Res 788:35–42.

Royet JP; Souchier C; Jourdan F, Ploye H (1988) Morphometric study of the glomerular population in the mouse olfactory bulb: numerical density and size distribution along the rostrocaudal axis. J Comp Neurol 270:559–568.

Shepherd GM, Greer CA(1998) Olfactory bulb. In: Shepherd GM, ed. The synaptic organization of the brain, 4th ed. Oxford UP, New York.

Sokal RR, Rohlf FJ (1995) Biometry: the principles and practice of statistics in biological research. 3rd Ed. New York: WH Freeman.

Strom RC (1999) Genetic analysis of variation in neuron number.Dissertation, University of Tennessee, Memphis.

Tomasiewicz H, Ono K, Yee D, Thompson C, Gioridis C, Rutishauser U, Magnuson T (1993) Genetic deletion of a neural cell adhesion molecule variant (N-CAM-180) produces distinct defects in the central nervous system. Neuron 11: 1163–1174.

Taylor BA (1989) Recombinant inbred strains. In: Genetic variants and strains of the laboratory mouse 2nd Ed (Lyon ML, Searle AG, eds), pp 773–796. Oxford: Oxford UP.

Taylor BA, Wnek C, Kotlus BS, Roemer N, MacTaggart T, Phillip (1999) Genotyping new BXD recombinant inbred mouse strains and comparison of BXD and consensus maps. Mamm Genome 10:335–348.

Tropepe V, Craig CG, Morshead DM , van der Kooy D (1997) Transforming growth factor- null and senescent mice show decreased neural progenitor cell proliferation in the forebrain subependmya. J Neurosci 17:7850–7859.

van der Kooy D, Weiss S (2000) Why stem cells. Science 287:1439–1442.

Vassar R, Chou Sk, Sitcheran R, Nunez JM, Vosshall GL, Axel R (1994) Topographic organization of sensory projections to the olfactory bulb. Cell 74-309–318.

Williams RW (1998) Neuroscience meets quantitative genetics: Using morphometric data to map genes that modulate CNS architecture. In: Morrison J, Hof P (eds) Short course in quantitative neuroanatomy. Society of Neuroscience, Washington DC, pp 66—78.

Williams RW (2000) Mapping genes that modulate mouse brain development: a quantitative genetic approach. In: Mouse brain development. (Goffinet A, Rakic P, eds), pp 21–49. Berlin: Springer.

Williams RW, Strom RC, Rice DS, Goldowitz D (1996) Genetic and environmental control of retinal ganglion cell number in mice. J Neurosci 16: 7193–7205.

Williams RW, Strom RC, Goldowitz D (1998) Natural variation in neuron number in mice is linked to a major quantitative trait locus on Chr 11. J Neurosci 18:138–146.

Wright S (1978) Evolution and the genetics of populations. Univ Chicago Press.

Woo CC, Coopersmith R, Leon M (1987) Localized changes in olfactory bulb morphology associated with early olfactory learning. J Comp Neurol 263:113–125.

Yang XW, Wynder C, Doughty ML, Heintz N (1999) BAC-mediated gene-dosage analysis reveals a role for Zipro1 (Ru49/Zfp38) in progenitor cell proliferation in cerebellum and skin. Nat Genet 22:327–335.

Zhou G, Williams RW (1999) Eye1 and Eye2: Gene loci that modulate eye size, lens weight, and retinal area in mouse. Invest Ophthalmol Vis Sci 40:817–825.

Since 29 April 2000; Revised Aug 4, 2000


Neurogenetics at University of Tennessee Health Science Center

Text Only Top of Page

Home Page  |  Genome DBs  |  Phenome DBs  |  Publications  |  People & Associates
Mouse Brain Library  |  Related Sites  |  |   MBL.ORG

Robert W. Williams | Alex Williams © 2002, modify this page