close

Вход

Забыли?

вход по аккаунту

?

Distribution of immunoglobulin allotypes among local populations of Kenya olive baboons.

код для вставкиСкачать
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 70:29-38(1986)
Distribution of ImmunoglobuIin Allotypes Among Local
Populations of Kenya Olive Baboons
T.J. OLIVIER, D.H. COPPENHAVER, AND A.G. STEINBERG
Anthropology Department, Northwestern University, Euanston, Illinois
60201 (TJ.0.1,
Institute of Primate Research, National Museums of Kenya,
Nairobi, Kenya (TJ 0.1,
Department of Microbiology, University of Texas
Medical Branch at Galveston, Galveston, Texas 77550 0.
H. C.), and
Department of Bwlogy, Case Western Reserve University, Cleveland, Ohio
44106 (A.G.S.)
KEY WORDS
Kenya baboons, Immunoglobulins,Local variations
ABSTRACT
In this paper we report on the distributions of immunoglobulin
allotypes among 564 olive baboons collected at six localities in Kenya. The
sample localities and sizes are 1) Lake Magadi, N = 107; 2) Nanyuki, N = 77;
3) Lake Baringo, N = 55; 4) Mosiro, N = 132; 5) Isiolo, N = 36; 6) Gilgil, N =
157. Gm allotypes 1, 10, 13, 15, and 17 are polymorphic among these samples.
Gm(l1) and Km(3) were present in all samples, and Gm(2,3,5,6,14,16,21,24,26)
and Km(1) were absent from all samples. The proportions of individuals positive for polymorphic allotypes varied substantially between different local
samples, as did the arrays and estimated frequencies of haplotypes. Allotype
frequencies in local samples do not appear to be simply related to either
geographic location or habitat characteristics of the localities. Our data suggest that much of the geographic variability in Kenya olive baboon populations
occurs between populations separated by small geographic distances.
Olive baboons are large, cercopithecoid
monkeys. They occur in Africa, in a broad
geographic area that lies immediately south
of the Sahara Desert. Olive baboons are ecologically generalized; within Kenya alone,
they occur in semidesert, savanna, and forest
habitats. These baboons usually are considered to comprise the species Papio anubis.
However, olive baboons and related forms
found in other regions of Africa have been
classified as a single species, l? cynocephalus
(Buettner-Janusch, 1966).
Studies over the past 15 years have shown
that gene frequency variations often occur
between neighboring social groups in baboon
and other monkey populations (Olivier et al.,
1974; Steinberg et al., 1977; Buettner-Janusch et al., 1983). The complex social subdivisions and dynamics in these populations
are believed to provide much of the basis for
the existence of these microvariations (Buettner-Janusch et al., 1983; Olivier, 1984). However, information on patterns of genetic
variation among baboon populations separated substantially in space and living under
varied environmental conditions are sparse.
0 1986 ALAN R. LISS, INC.
As a result, we have not known how gene
distributions vary between localities within
the baboon populations of large regions. We
have not known whether any such local variations are associated with environmental
variations or whether they might be distributed according to a simple pattern of isolation by distance.
To provide information in these areas, we
have undertaken a survey of biological variation among local populations of Kenya olive
baboons. The goal of this work has been to
provide a profile of genetic variation between
local populations within a larger regional baboon population. The local populations we
have sampled are separated by varied geographic distances and occupy heterogeneous
habitats. Many of the variations in sample
site habitats are attributable to attitude
variations, lower elevation sites being more
arid. We have chosen sampling sites in which
Received December 12, 1983; revision accepted October 29,
1985.
T.J. Olivier is now at the Department of Anthropology, Yale
University, 51 Hillhouse Ave., New Haven, CT 06511.
30
T.J.OLIVIER, D.H. COPPENHAVER, AND A.G. STEINBERG
geographic locations and habitat types vary
somewhat independently.
We have obtained multitroop samples from
each of the sampling localities. The maximum geographic separation between troops
of a single sampling locality in this survey is
about 20 km. Most troops at a single locality
are separated by much smaller distances. The
design of this project has been discussed more
fully elsewhere (Olivier et al., 1979).
In this survey, various systems have been
examined in sampled animals. We have previously reported dental metric variations in
local populations of Kenya baboons (Olivier
et al., 1979). In this paper, we report on our
findings regarding immunoglobulin allotype
distributions in different local Kenya olive
baboon populations.
Polymorphism occurs in various species for
antigens (allotypes) determined by the primary structure of constant regions of the
chains of Ig molecules. The Gm allotypes occur in the constant (C) domains of the heavy
03) chains of immunoglobulin G (IgG), the
gamma chain. IgG occurs in four subclasses
(called IgGl through IgG4, in their relative
order of decreasing concentration). These differ from each other in the amino acid sequences of their C domains, but resemble
each other more than they do H chains of the
other four classes of Ig. The subclasses are
determined by four closely linked genes. Genetic determinants of Gm antigens have been
found on the C domains of the H chains of
IgG1, IgG2, and IgG3. None have been found
on IgG4.
In humans and in other species where the
mode of inheritance has been established,
the Gm allotypes are transmitted via haplotypes (a unit composed of at least one allotype from IgGl and at least one allotype from
IgG3). The mode of inheritance of allotypes
in baboons has not been established with
pedigree studies. In analyses of phenotypes
in captive baboon colonies, the distribution
of phenotypes in parents and offspring usually is consistent with inheritance by haplotypes, but some exceptions have been
observed (Steinberg and others, unpublished
observations). We provisionally assume that
the baboon allotypes are inherited via haplotypes similar to those established for other
species.
The Gm haplotypes differ among the human races not only in frequency, but also in
that certain haplotypes have been found in
only one race. Furthermore, no two races
thus far studied have the same array of haplotypes. In addition, the haplotypes vary in a
cline among populations within a race (Steinberg and Cook, 1981).
The Km (Inv) allotypes occur in the C domain of the kappa chain. The three known
allotypes are transmitted by three codominant alleles, Kml, Km1>2,and Km3. These
vary in frequency among the races and
among populations within a race (reviewed
in Steinberg and Cook, 1981).
In an earlier study (Steinberg et al., 1977)
samples from 10 troops of olive baboons from
a small area in the Laikipia District of
Kenya, 2 troops from Ethiopia, and animals
of unknown origin from South Africa were
tested for Km(1,2,3) and for Gm(1,2,3,5,6,
11,13,14,15,16,17,21,24).
All the samples were
negative for Gm(2,6,14,16,24). All Kenyan
and Ethiopian samples were positive for
Km(3) and for Gm(l1) and Gm(17). Some degree of polymorphism was found for Km(l),
Km(1,2), and for Gm(l,3,5,11,13,15,21)in at
least one of the sample sets. Interestingly, no
two troops from Kenya had the same estimated array of “haplotypes,” and these
troops differed from the animals from Ethiopia and South Africa. Further, sizeable differences among troops in frequencies of
allotypes were observed, along with differences in allotype frequencies between the
western and eastern clusters of troops within
the Laikipia study area.
MATERIALS AND METHODS
This report is based on the analysis of
serum samples from 564 olive baboons. The
localities sampled, the abbreviations used for
the sites, and the approximate mean altitude
in meters of each sampling locality are 1)
Lake Magadi (LM, alt. = 650 m); 2) Nanyuki
(NK, alt. = 1700 m); 3) Lake Baringo (BG,
alt. = 950 m); 4) Mosiro (MS, alt. = 1300 m);
5) Isiolo (IS,alt. = 1700 m); and 6) Gilgil (GL,
alt. = 2000 m). The approximate geographic
centers of the localities are shown in Figure
1. The locations from which the samples were
collected for the report by Steinberg et al.
(1977) also are shown. The number of troops
0 and number of animals (N) sampled at
each site for the present study are 1)LM (T
= 4, N = 107);2) NK (T = 3, N = 77); 3) BG
CT = 3, N = 55); 4) MS (T = 8, N = 132);5)
IS Cr = multiple-see below, N = 36); 6) GL
Cr = 2, N = 157).
Samples from four sites CLM, NK, BG, MS)
were collected during 1977 in capture safaris
31
IMMUNOGLOBULIN ALLOTYPES IN KENYA BABOONS
36"
I
37O
38O
I
I
V
Rurnuruti
0
AIS
-
0 Nanyuki
0"
-0"
100-
AGL
KIM
50-
1"
- 1"
0-
c
5'
A
2 O
I
38'
gLM
1
37"
350
40'
Fig. 1. Map of collection localities. A , Area sampled for present study; V,area sampled in
previous study (Steinberg et al., 1977).
conducted by T.J.O. and D.H.C. Details of the
collections were presented in a report on
odontometric variation among these baboons
(Olivier et al., 1979).
The samples from Isiolo (IS)were provided
in 1977 through the courtesy of M i . Richard
Mann, a commercial trapper. The samples
were obtained from multiple troops, but the
exact number of troops represented and troop
assignments of individual animals at Isiolo
are not available.
The Gilgil samples were collected in 1979
in the course of a genetic-behavioral project
that involved participation of R. Byles, S.
Strum, T.J.O., staff of the Kenya Institute of
Primate Research, and others. Samples
tested in this study are from the behavioral
study focal troops (GLW = Pump House, GLY
T.J. OLNIER, D.H. COPPENHAVER,AND A.G. STEINBERG
32
RESULTS
= Eburru Cliffs). Details of the Gilgil site
K m (Znv)
are presented in other reports (Harding, 1976;
Harding and Strum, 1976; Byles and SandAll of the 564 samples were Km(3) and
ers, 1981).
none were Km(1). In our earlier study (SteinAll samples were absorbed with washed, berg et al., 1977) 10 baboons from the Lakipacked human red blood cells (RBC)until the pia District of Kenya were Km(1,3), and 4
sera no longer caused the RBC to agglutin- from South Africa were Km(1,2,3).Both East
ate. The absorbed samples were diluted one- African and South African populations were
eighth in normal saline and tested for allo- uniformly positive €or Km(3). We now have
types listed in Appendix I, except as indi- tested over 1,200 serum samples from bacated in later tables, by methods previously boons and all have had Km(3). It may be that
reported (Steinberg, 1962).All reagents listed all baboons are Km(3). Km(3) is the most
in Appendix I, except those labeled R39, R42 frequent of the Km allotypes among humans,
(rabbits), and RM (rhesus monkey) are from but no population thus far tested has been
100% Km(3) (Steinberg and Cook, 1981).
human donors.
Gm
TABLE 1. Gm phenotype occurrences in samples from
local olive baboon populations'
Location2
Grnohenotvoe LM
1,10,11,13,17
1,10,11,17
1,11,13,17
1,11,15
1,11,15,17
1,11,17
10,11,13,17
10,11,17
11,13,17
11,15
11,15,17
11,17
Sample
sizes
NK BG MS IS
-
-
_
_
0
3
38
0
0
2
9
1
_
-
_
0
6
14
8
2
_ _
_ _
0
0
1 8 4
46 59
2
2
107
77
0
1
4
22
55
0
0
0
0
0
37
0
0
6
25
0
0
0
0
0
5
0
36
5
0
0
0
0
58
12
132
GL
N
6
10
3
-
110
6
10
3
17
122
38
3
19
2
23
172
148
157
564
4
3
19
2
-
-
'A cell entry of - indicates phenotype involved one or more
antigens not tested for at the given locality. Only phenotypes
found in one or more animals in the materials for this paper are
listed.
'Locations defined in Materials and Methods and Figure 1.
The numbers of animals with each phenotype a t each location sampled are presented
in Table 1. Proportions of phenotypes positive for each allotype at each locality are
presented in Table 2. Gm allotypes 1, 10, 13,
15, and 17 are polymorphic among these samples. Gm(l1) was present in all samples and
Gm(2,3,5,6,14,16,21,24,26)were absent from
all samples.
Proportions of individuals positive for some
allotypes varied substantially among samples from different localities. The proportion
of individuals positive for Gm(1) varied from
a low of 0.121 at Gilgil to a high of 0.861 at
Isiolo. The proportion of individuals positive
for Gm(15)varied from a low of 0.455 a t Lake
Baringo to a high of 1.00 at Isiolo.
In contrast to the finding in our earlier
study in which all samples were Gm(17),
some baboons from four of the localities in
the present study (LM,NK, BG, and IS) are
Gm(-17). Indeed, Gm(17) has a lower fre-
TABLE 2. Freouencies of Gm allotvnes in local samnles'
Location'
Gm
LM
NK
BG
1
10
11
13
15
17
Sample
sizes
0.383
0.156
0.509
-
-
0.470
0
1.00
0
0.961
0.922
1.00
0
0.455
0.872
0
0.720
1.00
-
1.00
0
0.981
0.804
107
77
55
MS
1.00
132
IS
GL
0.861
0
1.00
0
1.00
0.833
0.121
0.242
1.00
0.0892
36
-
1.00
157
'An entry of - indicates local samples not tested for a particular antigen. Only antigens positive in some animals analyzed in this report
are listed.
'Locations defined in Materials and Methods and Figure 1.
IMMUNOGLOBULIN A L L O W E S IN KENYA BABOONS
(a)
Gml
Gm 15
33
(b)
Fig. 2. Local sample allotype frequencies. a-c. Local variations in Gml, Gm15, and Gm17
allotypes. d. Three-dimensional projection of allotype frequencies in the five local samples in
which all three of these allotypes were analyzed.
quency than Gm(15) in three of the four
localities.
The most polymorphic systems in this survey are Gm allotypes 1, 15, and 17. Maps of
the proportions of individuals positive for
these allotypes in local samples are presented in Figure 2a-c. Inspection of these
maps indicates that these allotype frequencies seem to vary irregularly.
Figure 2d presents a three-dimensional
projection of the axes representing frequencies in local samples for Gm allotypes 1, 15,
and 17. Positions of local samples in which
all three of the above allotypes have been
determined also are plotted. The projection
shows that the most extensive variations lie
along the Gm(1) and Gm(15)axes.
In Figure 2d, the BG, IS, and NK sites
occupy the most extreme locations; MS is
relatively central. The NK-IS pairing has a
relatively high Gm frequency separation despite their geographic proximity and their
similar habitats. The site pairing with the
least Gm frequency distance is LM-NK, despite the sizeable geographic and attitude
differences between the sites. In fact, distances between site locations in this Gm frequency projection show small negative
correlations both with geographic distances
separating sites and with altitude differences
between site pairs.
It was assumed that allotypes are transmitted via haplotypes, as in humans. Various
sets of haplotypes were postulated to explain
34
T.J. OLIVIER, D.H. COPPENHAVER, AND A.G. STEINBERG
TABLE 3. Frequencies of postulated haplotypes for localities LM, NK,BG, MS,and IS
LM
BG
NK
IS
MS
Haplotype
Freq.
SE
Freq.
SE
Freq.
SE
Freq.
SE
Freq.
SE
1,11,15
1,11,15,17
1,11,17
11.15
11,15,17
11,17
0.028
0.187
0.016
0.030
0.074
0.022
0.052
0.076
0.057
0.204
0.103
0.087
0.606
0.042
0.036
0.031
-
0.029
0.035
0.035
0.514
-
0.059
-
0.488
-
0.297
0.038
-
0.035
-
0.215
0.547
0.163
TABLE 4. Frequencies of postulated haplotypes
for troop GLY
Haplotype
Frequency
SD
1,10,11,17
1,11,13,17
1,11,17
10,11,13,17
lOJlJ7
11,13,17
11,17
0.139
0.064
0.071
0.052
0.123
0.039
0.512
0.041
0.029
0.035
0.028
0.044
0.027
0.060
-
0.031
0.266
0.460
0.274
-
0.486
-
-
0.059
-
Hardy-Weinberg distribution (data not presented). Since troop memberships of animals
in the Isiolo samples are not available, fwther analysis by troop cannot be done at Isiolo. Possibly there were unusual features in
the population structures of these sites, or
perhaps more haplotypes than we have postulated were present.
Eight phenotypes were present in troop
GLY but only two were present in GLW. The
samples from the two troops were tested at
the same time and with the same reagents.
Note that although the samples from three
regions (MS,IS, and GL) were tested for
Gm(lO), only those from the GL region
showed Gm(l0). In addition, only samples
from troop GLY from the GL region show
Gm(1) and Gm(13). While it is true that a
different anti-Gm(l3) reagent (Wal) was used
for samples from the GL area, anti-Gm(l3)
Ter (used for samples from all other localities
in this study) gave Gm(l3)-positivereactions
with some of the samples reported on by
Steinberg et al. (1977).It is unlikely that the
differences are artifacts. An estimate of haplotypes present in troop GLY is presented in
Table 4. Byles and Sanders (1981) have reported strong differentiation in ABO antigen
frequencies in troops from the Gilgil study
site. A separate report (Coppenhaver and
Olivier, 1986) will consider distribution of
allotypes among troops in this survey in more
detail.
the observed distributions of phenotypes.
Frequencies of postulated haplotypes that
produced the best fits to the Hardy-Weinberg
model were estimated by the program
MAXIM (Kurczynski and Steinberg, 19671,
as modified and rewritten in APL by one of
us (A.G.S.). If a sufficiently large number of
haplotypes are postulated, any observed set
of phenotypes can be generated by some set
of haplotype frequencies. "he data were analyzed with the minimum number of haplotypes required used to explain the observed
phenotypes, with the limit that the number
of haplotypes postulated must be less than
the number of phenotypes observed. By coincidence, each comparison has one degree of
freedom.
The haplotype arrays and frequencies that
give the best fits to the Hardy-Weinbergdistribution are shown in Tables 3 and 4. Using
DISCUSSION
the estimated haplotype frequencies, the
Although variation in immunoglobulin dissamples from NK, BG, MS, and GL all gave
reasonable fits to the Hardy-Weinberg distri- tributions in olive baboon populations living
bution (0.1 < P < 0.2; data not shown). The close together is common, there is considersample from LM gave a poor fit to the Hardy- able similarity in the physical appearance of
Weinberg distribution (chi-square 15.3, P < baboons from different areas within the spe0.001), as did the Isiolo sample (chi-square cies' range. Our allotype data offer some clues
5.48, P = 0.019). The data for the four LM as to how minor variations in the genetic
troops were examined separately. Only the composition of different regional baboon popdata for troop B gave a satisfactory fit to the ulations might exist with substantial genetic
IMMUNOGLOBULIN ALLOTYPES IN KENYA BABOONS
microvariation within each regional population.
It is possible to aggregate data from localities in our current survey into southern and
northern clusters of sites. LM and MS provide a southern cluster based on 12 troops
and 239 animals. NK,BG, and IS form a
northern cluster based on at least 8 troops
and 168 animals. Gilgil is excluded because
of its intermediate position.
In such an arrangement, the frequencies of
the Gm(1) allotype in the southern and
northern clusters are 43% and 42%, respectively. The frequencies of Gm(15) are 84%
and 80%.The frequencies of Gm(17) are 91%
and 89%. Compared to intertroop differences
reported in Steinberg et al. (1977) or compared to differencesbetween local samples in
this report, these southern-northern regional
differences are quite small.
It may be that generally there is little variation in allotype frequencies of olive baboon
populations of different large regions of
Kenya (i.e., the two regional populations represented by our northern and southern clusters of sampling sites). Instead there may be
relatively larger variations in allotype frequencies among local populations within
each region (i.e., the local populations represented by animals in the IS, NK, or BG sample sets) and additional variations among
troops at each locality (i.e., the different
troops at the LM sampling area or the different troops in the NK sampling area). This
interpretation is consistent with the irregular variations in Gm allotype frequencies
among our local samples.
However, in interpreting the frequencies of
Gm allotypes in our local samples, we must
take care to consider possible sampling effects stemming from the small numbers of
troops represented in most of our local samples. Our previous study of over 500 olive
baboons belonging to 10 troops in the Laikipia District of Kenya revealed major variations between neighboring troops in Gm
allotype frequencies (Steinberg et al., 1977).
For example, the frequency of Gm(1) varied
from 28% to 89% in the different troops in
this earlier study. The proportion of Gm(15)
ranged from 24% to 100% among the different troops. Although Gm(17) is polymorphic
in most of the local samples in the present
study, it was present in all animals in our
earlier Laikipia troop study. Hence, we have
no range of variation of Gm(17) frequencies
in previous Laikipia samples.
35
In the present survey, the larger geographic scope compared to our earlier expedition and practical limitations inherent in a
larger survey in the limited time available
forced us to sample a smaller number of
troops within each locality. Given the large
intertroop variation observed in our earlier
study, it seems possible that local sample
frequencies may be highly affeded by which
few troops one happens to capture and include in one's local samples. Under this hypothesis, the variation in allotype frequencies
in our local samples (Table 2) may be due to
the choice of troops for capture, rather than
differences between all of the troops in different local populations. If this is correct, then
one might argue that there is little local geographic variation in allotype frequencies
within large regions, only variations between troops.
Simulations were carried out to examine
whether variations in allotype frequencies
similar to those observed among our local
samples might have arisen by sampling errors, in particular sampling errors stemming
from examination of only a few troops at
each locality. For these simulations we assumed that there were no variations between
localities in allotype frequencies. Further, allotype frequencies for the ten troops in our
1977 Laikipia sample were assumed to exhibit microvariations in allotype frequencies
characteristic of those found within localities
throughout Kenya.
For one series of simulations, an array containing Gm(1) allotype frequencies identical
to those observed for the ten 1977 Laikipia
troops was established. An unweighted mean
for these ten troop frequencies also was calculated. In a set of simulation runs for each
locality, troop Gm(1) allotype frequencies
were drawn from this array by means of
pseudorandom sampling with replacement.
The number of frequencies drawn in one run
equal the number of troops sampled from
that locality in our survey. The exact number
of troops represented in our Isiolo sample is
not known. For these simulations, the number was set to two; this maximizes the sampling errors in the simulated drawings of the
Isiolo samples. An unweighted mean for the
frequencies of allotypes drawn in one run
was calculated.
Five hundred simulated multitroop samples for each locality were drawn by this
simulation process. A count was kept of the
number of runs in which allotype frequency
36
T.J. OLIVIER, D.H. COPPENHAVER, AND A.G. STEINBERG
deviations of the simulated local samples
were comparable to those in the real local
sample. A deviation was considered comparable if 1)the deviation of the mean of the
troop frequencies in the simulated sample
lay in the same direction as the deviation of
the real local sample allotype frequency from
the mean of the array of Laikipia troop frequencies; and 2) the magnitude of the simulated sample deviation from the Laikipia
sample mean equaled or exceeded the magnitude of the deviation of the real local sample from the mean of the Laikipia troop array.
Similar simulations were carried out for
Gm(15). Simulations for Gm(17)were not carried out since this allotype was present in all
animals in our earlier study.
Deviations in the simulated samples comparable to those observed in the real survey
samples occurred in 25%or more of the simulations of Gm(1) sampling at localities LM,
BG, and MS and of Gm(15) sampling at MS.
However, comparable deviations were observed in only 5%to 1%of the simulations of
Gm(1)sampling at IS and Gm(15)at NK, BG,
and IS. Further, comparable deviations were
observed in simulations in less than 1%of
the runs for Gm(1)at NK and GL, and Gm(15)
at LM.
The occurrence of so many simulation runs
in which comparable sample deviations were
rare indicates that many of the variations in
our local sample allotype frequencies do reflect geographic variation in allotype frequencies in local olive baboon populations.
However, somo of the variations in local sample allotype frequencies undoubtedly are due
to sampling errors stemming from our capture procedures.
Although our results indicate that local
geographic variation occurs within the
Kenya olive baboon population, it is not immediately clear over just what geographic
distance scales these variations take place.
In some studies of human populations, a
rapid decay of kinship between population
members separated by geographic distances
of 20 km or less has been reported (Friedlaender, 1971). Some of our observations lead us
to speculate that many of the spatial variations in genetic compositions of baboon populations occur at a microgeographic level,
between sets of neighboring troops occupying
small areas separated by a few kilometers.
Further examination of data from our previous study of Laikipia District troops and
the NK site in the current survey reveals
some observations on microgeographic variation in allotype frequencies.
The troops reported on in Steinberg et al.
(1977)were captured in two geographic clusters (Fig. 1). The western cluster contained
four troops; the eastern, six. Given the number of troops in each cluster, their spa&al
compactness, and the large numbers of animals sampled in each troop, it seems likely
that allotype frequencies determined from
these clusters are close to the correct ones for
the small areas covered by each cluster. For
Gm(1) the western and eastern clusters had
frequencies of 55%and 38%,respectively. For
Gm(15), the frequencies were 86% and 68%.
These variations between nearby troop clusters are as large as or larger than many of
the variations between localities in this survey (Table 2).
The three troops in the NK sample in the
present survey were captured along a 5-km
segment of the Ewaso Nanyuki river. This
capture area lay slightly south and east of
the eastern cluster of troops studied previously. The GmU) frequency for the threetroop NK sample lies outside the range of
single-troop Gm(1) frequencies in our previous Laikipia study; hence, there were no
deviations in the simulated samples for
Gm(1) at the NK site comparable to those in
the real NK sample. About 4% of the runs in
the Gm(15) NK simulations gave deviations
comparable to those observed in the real NK
data. The existence of microgeographic variations in allotype frequencies may explain
why the allotype frequencies in the recent
NK sample are so unlikely to have been
formed from a population with troop frequencies like those in our earlier study. A general
occurrence of sizeable microgeographic variations in allotype frequencies might also explain how the IS sample differs so much from
the NK sample.
Mechanisms that could contribute to microgeographic variations in gene frequencies
would include simple restriction of emigration distances by male baboons that leave
their natal troops. Further, Cheney and Seyfarth (1983) note that in vervet and other
cercopithecoid monkey populations, there is
some tendency for related males to emigrate
to the same groups. These patterns might
generate social and genetic relationships between groups occupying very small geographic areas. Fix (1978) notes that where
IMMUNOGLOBULINALLOTYPES IN KENYA BABOONS
migration is kin-structured, high levels of
intergroup differentiation can exist, despite
high rates of intergroup migration.
McCluskey et al. (1977) examined the distributions of genetic variations in immunoglobulin molecules in the rabbit population
of Australia. They found variations in immunoglobulin allele frequencies between
populations living near the west and east
coasts of Australia, between different local
populations in Western Australia, and between separate warren systems at a single
locality. In many respects, these findings parallel the results of our study of baboon immunoglobulins. However, one must assume
that the divergences among Australian rabbit populations have developed quickly, since
rabbits were introduced into Australia during the last century.
Further work on the inheritance of Ig phenotypes in baboons is needed, as are further
data on allotype distributions in different baboon populations. If it is assumed that the
haplotypes are genetically transmitted and
that the allotypic markers are carried by the
Ig molecules of baboons, we have evidence
for the early origin of many of the Gm allotypes and for all three of the Km allotypes.
The present data also serve to underscore
our earlier conclusion that extensive polymorphism for the Ig allotypes exists among
baboons. The first study established that
troops from a limited region (less than 50
square miles) may differ markedly in Gm
composition. The present report reveals differences between localities of a single country, as well as statistical associations of the
different antigenic factors. We speculate that
there is little variation in allotype frequencies of olive baboon populations of large regions of Kenya. Our data suggest instead
that microgeographic variations in allotype
frequencies within large regions may be
common.
ACKNOWLEDGMENTS
The authors thank the Oflice of the President and the Ministry of Tourism and Wildlife of the government of Kenya for permission to conduct this research. We thank Simon Taiti of the latter ministry and Joseph
Mungai of the Department of Human Anatomy of the University of Nairobi for their
cooperation in arranging this research. We
also thank Jim Else, Joseph Tesot, and other
members of the staff of the Institute of Pri-
37
mate Research for their cooperation and assistance, especially in the collection of the
Gilgil materials. The contribution of the Isiolo materials by Mr. Richard Mann is much
appreciated. The support of the U.S. National Science Foundation (grant numbers
BNS 76-17123, DEB 78-06914) and U.S. National Institutes of Health (grant number 1
RO 1 GM24407-01) are gratefully acknowledged. T.J.O. and D.H.C. wish to thank their
wives, Wren Olivier and Anne Coppenhaver,
for their help and support in various stages
of this work. We also thank members of our
capture team, Bensen Mativo, Ndumbuthi
Nzuo, Kasakolo Ngolana, and Muli Ndambuki for their many months of help in the
bush. Tim Williams is thanked for his help
and good company. The hospitality of Jack
Fairhall during our Nanyuki safari was a
great help. Members of the Gilgil behavioralgenetic project, including R.H. Byles, S.C.
Strum, and E. Hunt are thanked for their
efforts. The authors also thank J. BuettnerJanusch and Carole Ober for their cooperation in logistic requirements.
LITERATURE CITED
BuettnerJanusch, J (1966)A problem in evolutionary
systematics: Nomenclature and classification of baboons, genus Pupio.Folia Primatol. 4:288-308.
BuettnerJanusch, J, Olivier, TJ, Ober, CL, and ChepkoSade, BD (1983)Models for lineal effects in rhesus
group fissions. Am. J. Phys. Anthropol. 61:347-353.
Byles, RH, and Sanders, MF (1981)Intertroop variation
in the frequencies of ABO alleles in a population of
olive baboons. Int. J. Primatol. 2:35-46.
Cheney, DL, and Seyfarth, RM (1983)Nonrandom dispersal in free-ranging vervet monkeys: Social and genetic consequences. Am. Nat. 122:392-412.
Coppenhaver, DH, and Olivier, TJ (1986)Immunoglobulin allotypes of Kenyan olive baboons: Troop frequencies, linkage disequilibria and comparisons with other
studies. Int. J. Primatol. (in press).
Fix AG (1978)The role of kin structured migration in
genetic microdifferentiation. Ann. Hum. Genet.
41:321-339.
Friedlaender, JS (1971)Thepopulation structure of southcentral Bougainville. Am. J. Phys. Anthropol. 35:1326.
Harding, RSO (1976)Ranging patterns of a troop of baboons (Papio unubis) in Kenya. Folia Primatol. 25:143185.
Harding, RSO, and Strum, SC (1976)The predatory baboons of Kekopey. Nat. Hist. 85:46-53.
Kurezynski, TW, and Steinberg, AG (1967)A general
program for maximum likelihood estimation of gene
frequencies. Am. J. Hum. Genet. 19~178-179.
McCluskey, J, Olivier, TJ, and Curtain, C (1977)Immunoglobulin allotypes in some Australian wild rabbit
populations. Aust. J. Zool. 25%-74.
Olivier, TJ (1984)Some issues in the abstraction of ge-
38
T.J. OLIVIER, D.H.COPPENHAVER, AND A.G. STEINBERG
determined human y-globulin types (the Gm and Inv
netic structures of monkey societies. J. Social Biol.
groups). Prog. Med. Genet. 2:l-33.
Struct. 73-73,
Olivier, TJ, BuettnerJanusch, J, and BuettnerJanusch, Steinberg, AG, and Cook, CE (1981)The Distribution of
V (1974)Carbonic anyhydrase isoenzymes in nine
the Human Immunoglobulin Allotypes. Oxford Oxtroops of Kenya baboons, Papio cynocephalus Ginford University Press.
naeus 1766).Am. J. Phys. Anthropol. 41:175-190.
Steinberg, AG, Olivier, TJ, and Buettner-Janusch, J
Olivier, TJ, Freedman, L, and Coppenhaver, D (1979)
(1977)Gm and Inv studies on baboons, PapM cynoceOdontometric variation among local populations of
phalw: Analysis of serum samples from Kenya, EthioKenyan olive baboons. J. Hum. Evol. 8:603-614.
pia and South Africa. Am J. Phys. Anthropol. 47:2130.
Steinberg, AG (1962)Progress in the study of genetically
Appendix I. Reagents Used to Determine the Indicated Human Allotypes in Baboon Sera
Molecules
(in humans)
IgGl
IgG3
Kappa
Allotypes
Numer.
Alphabet.
Antiallotype
Anti-D
Gm 1
2
3
17
5
6
10
11
13
14
15
16
21
24
26
Inv 1
3
a
Mor
Han or Rya*
Coa or How
R39,R42+
Pay or Dra
Har or Wil
Pla
BrY
Ter or Wal
Bur
Gai
Bar or Tai
Bar or Tai
Car
Jol
Jol
Kre
Jol
Car
Jol
Car
Puh
Puh or Vai
Bar
Con or Kre
Kre
Roe
Ham
X
f
2
b'
c3
b5
bo
b3
b4
S
t
g
c5
U
1
3
*The second reagent in each case was used for the GL samples.
+R39and R42 obtained from rabbits.
= Rhesus monkey.
'RM
RM'
Cli
Cur or Cum
Whi
D' An
Nee
~~
Документ
Категория
Без категории
Просмотров
3
Размер файла
758 Кб
Теги
local, population, distributions, baboons, allotypes, kenya, oliver, immunoglobulin, among
1/--страниц
Пожаловаться на содержимое документа