By Reed P. Warren, J. Dennis Odell, W. Louise Warren, Roger A. Burger,
Alma Maciulis, Wayne W. Daniels and Anthony R. Torres
The Center for Persons with Disabilities and the Department of Biology,
Utah State University, Logan, Utah 84322.
Address all correspondence to Reed P. Warren, Ph.D., UMC 6895, Utah
State University, Logan, Utah, 84322, USA, Telephone: (801) 797-1924,
FAX: (801) 797-2044, E-mail: Medlab@cc.usu.edu
=======
Summary
=======
We reported that the major histocompatibility complex (MHC) including
the null allele of the C4B gene and the extended haplotype B44-C30-DR4
is associated with autism. We report now that the third hypervariable
region (HVR-3) of certain DRb1 alleles have very strong association with
autism. The HVR-3 of DRb1*0401 or the shared HVR-3 alleles DRb1*0404
and DRb1*0101, was expressed on extended haplotypes in 23 of 50 (46%)
autistic subjects as compared to only 6 of 79 (7.5%) normal subjects.
Another HVR-3 sequence, the DRb1*0701 allele, was carried on extended
haplotypes in 16 (32.0%) of the autistic subjects as compared to 8
(10.1%) of the normal subjects.
============
Introduction
============
Autism is a behaviorally-defined, biologically-based developmental
disability manifesting in early childhood and affecting boys about 4-5
times more frequently than girls (Gillberg and Coleman, 1992). Autism
involves severely impaired development of language and communication,
unusual behaviors, and often mental retardation. The etiology of most
cases of autism is unknown but a genetic component for this disorder is
clearly established (reviewed in Gillberg and Coleman, 1992). Moreover,
a fairly impressive volume of evidence has accumulated suggesting that
many autistic subjects have immune imbalances including numbers and
functions of T cells and their subsets (Warren et al., 1986, Stubbs et
al., 1977, Warren et al., 1990, Yonk et al, 1990), activated T cells
(Plioplys et al., 1994b, Warren et al., 1995a) and depressed natural
killer cell function (Warren et al., 1987).
We have reported that the allelic products of certain genes of the major
histocompatibility complex (MHC) are associated with autism (Warren et
al., 1991, Warren et al., 1992, Daniels, et al. 1995) including the
null allele of the C4B gene (located in the class III region of the
MHC), and the extended haplotype B44-S30- DR4 which consists of the 44
allele of the HLA-B region, the S allele of the BF gene, the 3 allele of
C4A, the 0 or null allele of C4B and the DR4 allele (Figure 1).
Extended haplotypes are observed when certain MHC haplotypes, or
combinations of DRb1 alleles and other polymorphisms of the various MHC
genes, are expressed together far more often than expected considering
the relative distances between their respective genes. An unknown
mechanism prevents crossovers (or other mutational events) from
occurring within this portion of the MHC, resulting in a relative fixity
of DNA sequence seen in extended MHC haplotypes (also termed ancestral
haplotypes) as reviewed (Alper et al., 1986, Degli-Esposti et al.,
1992). There are about a dozen or so extended haplotypes that have been
identified with a gene frequency of 0.0043 or greater. Investigations
have associated one or more of these extended haplotypes with specific
disorders, most of which are known or suspected to be autoimmune in
nature, as reviewed (Degli-Esposti et al., 1992, Fraser et al., 1990).
In attempting to further characterize the possible involvement of the
MHC in autism, we have investigated in particular, one portion of the
extended haplotype, the amino acid sequence motif encoded by the third
hypervariable region (HVR-3) of HLA-DR beta-1 (DRb1). The HVR-3
sequence, along with hypervariable regions 1 and 2, are essential
(Winchester, 1995) for the presentation of peptides to T lymphocytes by
antigen presenting cells (APC). Thus, HVR-3 is necessary for the
development of specific cell- mediated and humoral (antibody) immune
responses to pathogens and other foreign antigens. The multiple alleles
of the HVR-3 sequence result in a diversity of tertiary structures in
one of the binding sites of the class II MHC molecules. This normal
diversity is responsible for the widely ranging magnitude, from very
weak to very strong, in which the MHC molecule binds to antigens. In
theory, the HVR-3 sequence would be unique for each specific extended
MHC haplotype in humans. In reality, several of the different extended
haplotypes carry either identical HVR-3 sequences or sequences differing
by only a single structurally- similar amino acid (Winchester, 1995).
Such extended haplotypes are associated with identical or very similar
binding patterns to specific antigens.
=====================
Materials and Methods
=====================
Subjects
========
This investigation included 50 randomly-chosen autistic subjects (41
males and 9 females) and all but 6 pairs of their parents. Also
included were 79 normal subjects, 64 males and 15 females. All of the
subjects were of northern european derivation and all but two autistic
subjects were living in northern Utah. Diagnosis of infantile autism
followed DSM-IIIR criteria as established by a psychiatrist or a
psychologist. Blood samples from the subjects were acquired following
informed consent procedures. None of the autistic subjects had an
identifiable cause of their disease and all were living at home at the
time of study.
HLA typing
==========
Peripheral blood lymphocytes isolated by ficoll-hypaque centrifugation
were HLA typed by the technique we previously described (Warren et al.,
1992). B-cells were obtained by use of the Lympho-Kwik B cell Isolation
Reagent (One Lambda, Inc., Canoga Park, CA). The
microlymphocytotoxicity test was a one-stage, complement-dependent
reaction in which the purified lymphocytes were incubated in preloaded
trays with monoclonal antibodies specific for the various HLA
specificities and rabbit serum as a source of complement (One Lambda,
Inc.). A total of 2,000 cells were pipetted into microtest trays
preloaded with HLA antisera. The cells, antibodies and complement were
mixed and incubated at room temperature for 60 min.
Complement typing
=================
The genetic typing for the complement proteins C4A, C4B and BF were
performed by previously described techniques (Warren et al., 1991,
Warren et al., 1992). Briefly, typing for the C4 phenotypes was carried
out by incubating samples with neuraminidase from Clostridium
perfringens overnight at room temperature with continuous dialysis
against 0.1 M phosphate buffer, pH 7.0, containing 0.005 M EDTA-Na2.
The desialated samples were subjected to electrophoresis and
immunofixation with goat anti- human C4 (Atlantic Antibodies,
Scarborough, ME) at 1% and subjected to crossed immunoelectrophoresis.
Some samples, processed as above, were developed with a C4 complement
overlay consisting of antibody-sensitized sheep red blood cells and C4-
deficient guinea pig serum incorporated into a gel and layered onto a C4
agarose gel. The presence of null alleles (Q0) was determined by
inspection of immunofixation patterns or by crossed
immunoelectrophoresis. BF typing was carried out with frozen plasma
that is subjected to electrophoresis in agarose gel and immunofixation
with goat-antiserum to human factor B (Atlantic Antibodies) as
previously described (Warren et al., 1991, Warren et al., 1992).
Typing of DR b1 alleles with the PCR-RFLP technique
---------------------------------------------------
Mononuclear cells (approximately 108-1010) were added to 300l of 75mM
NaCl, 24mM EDTA solution, 20l Pronase E (20 mg/ml) and 12l 20% sodium
dodecyl sulfate. This mixture was incubated at 55oC 12-16 hr with
gentle shaking. The DNA was extracted three times, twice with an equal
volume of 25:24:1 phenol:chloroform: isoamyl alcohol and once with 24:1
chloroform:isoamyl alcohol. The DNA was washed and precipitated twice
before spooling into an appropriate amount of 10mM Tris-HCl and 0.1mM
EDTA. The PCR was carried out according the methods of Inoko and Ota,
(1993) using 200ng genomic DNA, 50mM KCl, 2.5mM MgCl2, 10mM Tris-HCl,
200uM of each dNTP, 1.5 units Taq polymerase and 1mM primers specific
for DRb1 alleles. Mineral oil overlay was used and the samples
subjected to 30 cycles of temperature cycling optimized for each primer
set. RFLP consisted of using allele-specific restriction enzymes for
each PCR product. Nineteen different enzymes were used to characterize
the DRb1 alleles and suballeles. Comparison of the cleavage patterns
with published information enabled identification of the various
alleles.
=======
Results
=======
HVR-3 sequences in autism
=========================
HVR-3 sequences arbitrarily designated as 1-5, carried on various
extended haplotypes, are presented in Table 1. The percentages of the
autistic and normal subjects carrying the various HVR-3 sequences are
presented in Table 2. Seventeen of the 50 (34%) autistic subjects had
sequence 1 as compared to only 2 of the 79 (2.5%) normal subjects.
Sequence 3 was represented in 16 (32%) of the autistic subjects as
compared to 8 (10.1%) in the normal subjects. None of the other
sequences were significantly altered in the autistic subjects. In all,
31 of the 50 (62.0%) autistic subjects had either sequence 1 or 3 (two
of these 31 subjects had both sequences). In contrast, only 10 of the
79 (12.6%) normal subjects had either sequence 1 or 3. HVR-3 sequence 2
differs from sequence 1 by only one basic polar hydrophilic amino acid
at position 71 (Winchester, 1995) (also see Table 2). Sequence 2 has
arginine at position 71 while sequence 1 has lysine at this position.
Although sequence 2 was not elevated in autism, if because of the
sequence similarities, the six autistic subjects having sequence 2 are
added to the total number of subjects having sequence 1 or 3, 35 of the
50 (70.0%) autistic subjects would have one of these four HVR-3
sequences (one subject had both sequence 2 and 4 and another had both
sequence 3 and 4). Included in the numbers of subjects (in both groups)
with the various HVR-3 sequences were those who had the class II and
class III fragments of the various extended haplotypes as well as the
entire extended haplotype.
The expression of the HVR-3 sequence motifs was also compared in the
autistic and normal subjects without regard to whether these sequences
were part of an extended haplotype. In this analysis, the total number
of autistic subjects having sequences 1, 2 or 3 was 37 (74%). Two
subjects had two of these sequences. However, 28 (35.4%) of the normal
subjects had sequence 1 and or 2 and 22 additional subjects had sequence
3. These data suggest that the associations of HVR-3 sequences with
autism were much greater when the sequences were part of extended
haplotypes (Table 2).
HVR-3 sequences expressed on chromosomes segregating
to the autistic child as compared to those on
chromosomes not passing to the autistic child
====================================================
An additional statistical analysis was performed to establish a possible
relationship between certain HVR-3 sequences and autism. The number of
sequences segregating to the autistic children were compared to the
number of sequences expressed on parental chromosomes (family normal)
which were not received by the autistic child (Table 3). The use of
these family normal chromosomes as controls are ideal because they
eliminate any bias that is introduced by use of matched normal control
subjects. The same general results were obtained with this analysis as
were seen with the use of unrelated controls as described above. Both
HVR-3 sequences 1 and 3 were increased on autistic chromosomes as
compared to "family normal" chromosomes. However, sequence 1 on
autistic chromosomes was significantly increased only when chromosomes
of the fathers were included in the analysis. In fact, eight mothers
expressed sequence 1 on the chromosome which did not pass to their
autistic child. As will be disscussed below, it is possible that
expression of sequence 1 on either of her chromosomes makes a mother
more likely to give birth to a child who will develop autism.
An analysis of the possible relative contributions of HVR-3
sequences, C4A alleles or C4B alleles to the development of autism
==================================================================
Since our group has previously demonstrated an association of the C4B
null allele with autism (Warren et al. 1991), we performed a relative
risk analysis (Table 4) in order to determine which HVR sequences or C4A
and C4B alleles were associated with the highest risk of developing
autism. The relative risk calculation (see Table 4) approximates how
many times more likely a person is to develop autism if he has the
relative marker than if he does not. HVR-3 sequence 1 had a relative
risk of 19.8 which was clearly the highest; the C4B null allele was next
with a relative risk of 4.6.
==========
Discussion
==========
Our current investigation, if replicable, provides a novel and
potentially very important advance in the understanding of the
pathophysiology of autism. These data provide a reasonable basis for
the development of this disorder based upon a possible anomaly in
function or regulation of the immune system which could be associated
with immune deficiency and/or an autoimmune mechanism.
The basis for the strongest MHC associations in this and previous
studies appears to be the extended haplotype. The extended MHC
haplotype comprises a relatively constant sequence of DNA over a large
multi-region of the MHC inclusive of, but not limited to, the HLA-B
through the HLA-DR regions (Figure 1). Therefore, it is possible that
any gene(s) within this extended sequence is associated with autism.
However, the association of fragments of B44-SC30-DR4 with autism
(Daniels et al., 1995) suggests that the relevant gene (or genes) is
(are) found within the class II or class III regions of the extended
haplotype).
The relevance of the class II DRb1 HVR-3 sequences to the development of
autism is not known. However, the tertiary structures resulting from
the primary amino acid arrangement of these sequences may not bind (or
bind with very low affinity) to a certain specific virus particle or a
peptide of some other pathogen. Thus, the pathogen would not be
presented to the T cells and no specific immune response would be
generated. The pathogen might be able to persist and spread to the
central nervous system infecting tissues and stimulating autoimmune
mechanisms. The presence of these sequences in the mothers of affected
fetuses may have allowed a pathogen to be present during the time that
she was carrying the child, perhaps permitting damage to the developing
brain of the fetus with manifestations of symptoms becoming apparent
some time after birth. That nine of the mothers of the autistic
subjects (but none of the fathers) had sequence 1 as part of an extended
haplotype that did not segregate to their affected child, supports the
possibility of an inadequate maternal immune response taking place
during the pregnancy. With this concept, the the fetus (although not
having sequence 1) may not have been capable of protecting itself in the
absence of a normal maternal immune response.
An alternative possibility to account for the association of the HVR-3
sequences with autism is that the tertiary structures resulting from
these sequences bind an unknown self antigen with strong affinity,
sufficiently strong as to trigger an autoimmune response that targets
the central nervous system. The presence of these sequences in the
mothers may have caused her to mount an immune attack on the developing
central nervous system of the child while having no effect on her own
mature central nervous system.
The relatively greater enhanced expression of the DRb1 sequences in
autism when carried on extended haplotypes appears to implicate, in
addition to DRb1, another gene within the extended haplotype in the
development of symptoms of autism. In this disorder we have detected
decreased levels of the C4B protein (Warren et al., 1994) and have found
that about 20% of autistic subjects have decreased levels of
immunoglobulin A in their serum (Warren et al., 1995b). The C4B gene is
known to be located in the class III region of the MHC and there is
evidence (Shaffer et al., 1989, Volanakis et al., 1992) that an unknown
gene regulating immunoglobulin A production is also located in the class
III region of the MHC.
The HVR-3 sequences we refer to as 1, 2 and 3 have been the subject of
other studies. Interestingly, these sequences are clearly associated
with rheumatoid arthritis (reviewed in Winchester, 1995). The basis for
the association with rheumatoid arthritis appears due to the near
identity of their DRb1 HVR-3 sequences. We should point out that none
of our control subjects had rheumatoid arthritis or any other known
autoimmune disorder.
These data on the HVR-3 sequences 1 and 2 in autism, and their
similarity to those seen in rheumatoid arthritis, support the
possibility that a significant number of cases of autism are autoimmune
in nature or arise from a maternal immune attack against fetal tissue.
The target of the autoimmune response in rheumatoid arthritis is
unknown, but reasonably may be collagen. In the behavior disorders, the
target might be a brain antigen such as myelin basic protein and/or
neurofilament proteins. Both cell mediated (Weizman et al., 1982) and
serum antibodies to myelin basic protein (Singh et al., 1993) have been
reported to be present in autism. Moreover, a study (Plioplys et al.,
1994a) has detected antibodies reactive with neurofilament proteins in
serum of autistic subjects. That these HVR-3 sequences result in
rheumatoid arthritis in some subjects and autism in others, may be
related to a modifying influence of another gene, related or unrelated
to the immune system, and/or to an environmental factor.
Clearly, additional clarification of the role of the MHC, pathogens and
autoimmune processes in the development of autism is warranted.
===============
Acknowledgments
===============
Supported by Grant MH42119 from the National Institute of Mental Health
and a grant from the Willard L. Eccles Charitable Foundation. The
authors appreciate the assistance of Peggy L. Dahle in the preparation
of this manuscript.
==========
References
==========
Alper, C.A., Awdeh, Z.L. and Yunis, E.J. (1986) Complotypes,
extended haplotypes, male segregation distortion and disease
markers. Human Immunol. 15:366-373.
Daniels, W.W., Warren, R.P., Odell, J.D., Maciulis, A., Burger,
R.A., Warren, W.L. and Torres, A.R. (1995) Increased frequency
of the extended or ancestral Haplotype B44-SC30-DR4 in autism.
Neuropsychobiology 32:120-123.
Degli-Esposti, M., Leaver, A.L., Christiansen, F.T., Witt, C.S.,
Abrahma, L.J. and Dawkins, R.L.(1992) Ancestral
haplotypes:conserved population MHC haplotypes. Human Immunol.
34:242-252.
Fraser, P.A., Stern, S., Larson, M.G., Marcus-Bagley, D., Awdeh,
Z., Glass, D.N. and Alper, C.A. (1990) HLA extended haplotypes
in childhood and adult onset HLA-DR4-associated arthropathies.
Tissue Antigens 35:56-59.
Gillberg, C. and Coleman, M. (1992) The Biology of the Autistic
Syndromes-2nd Edition. Mac Keith Press, Oxford, Blackwell
Scientific Publications Ltd., London, England, 317 pp.
Inoki, H. and Ota, M. (1993) PCR-RFLP. In: Handbook of HLA Typing
Techniques. K.M. Hui and J.L. Bidwell (Eds) CRC Press, Boca
Raton, 9-70.
Plioplys, A.V., Greaves, A., Kazemi, K. and Silverman, E. (1994a)
Immunoglobulin reactivity in autism and RettUs syndrome.
Develop. Brain Dys. 7:12-16.
Plioplys, A.V., Greaves, A., Kazemi, K. and Silverman, E. (1994b)
Lymphocyte function in autism and Rett Syndrome.
Neuropsychobiology 7:12-16.
Shaffer, F.M., Palermos, J., Zhu, Z.B., Barger, B.O., Cooper, M.D.
and Volankis, J.E. (1989) Individuals with IgA and common
variable immunodeficiency share polymorphisms of major
histocompatibility complex class III genes. PNAS 86:8015-8019.
Singh, V.K., Warren, R.P., Odell, J.D., Warren, W.L. and Cole, P.
(1993) Antibodies to myelin basic protein in children with
autistic behavior. Brain, Behav. And Immun. 7:97-103.
Stubbs, E.G., Crawford, M.L., Burger, D.R. and Vandenbark, A.A.
(1977) Depressed lymphocyte responsiveness in autistic children.
J. Autism Child. Schizophr. 7:49-55.
Volanakis, J.E., Zhu, A-B, Schaffer, F.M., Macon, K.J., Palermos,
J., Barger, B.O., Go, R., Campbell, R.D., Schroeder, H.W. Jr.
and Cooper, M.D. (1992) Major histocompatibility complex class
III genes and susceptibility to Immunoglobulin A deficiency and
common variable immunodeficiency. J. Clinic. Invest. 89:1914-
1922.
Warren, R.P., Burger, R.A., Odell, J.D., Torres, A.R. and Warren,
W.L. (1994) Decreased plasma concentrations of the C4B
complement protein in autism. Arch. Pediatr. Adolesc. Med.
148:180-183.
Warren, R.P., Foster, A. and Margaretten, N.C. (1987) Reduced
natural killer cell activity in autism. J. Amer. Acad. Child
Psych. 26:333-335.
Warren, R.P., Margaretten, N.C., Pace, N.C. and Foster, A. (1986).
Immune abnormalities in patients with autism. J. Autism Dev.
Disord. 16:189-197.
Warren, R.P., Odell, J.D., Warren, W.L., Burger, R.A., Maciulis,
A., Daniels, W.W. and Torres, A.R. (1995a) Immunoglobulin A
deficiency in a subset of autistic subjects. Journal of Autism
and Developmental Disorders (in press).
Warren, R.P., Singh, V.K., Cole, P., Odell, J.D., Pingree, C.B.,
Warren, W.L., DeWitt, C.W. and McCullough, M. (1992) Possible
association of the extended MHC haplotype B44-SC30-DR4 in
autism. Immunogenetics 36:203-207.
Warren, R.P., Singh, V.K., Cole, P., Odell, J.D., Pingree, C.B.,
Warren, W.L. and White, E.(1991) Increased frequency of the null
allele at the complement C4B locus in autism. Exper. Immunol.
83:438-440.
Warren, R.P., Yonk, L.J., Burger, R.A., Cole, P., Odell, J.D.,
Warren, W.L., White, E. and Singh, V.K. (1990) Deficiency of
suppressor-inducer (CD4+CD45R+) T cells in autism. Immunol.
Invest. 19:245-252.
Warren, R.P., Yonk, J., Burger, R.A., Odell, J.D. and Warren, W.L.
(1995b). DR positive T cells in autism: Association with
decreased plasma levels of the complement C4B protein.
Neuropsychobiology 31:53-57.
Weizman, A., Weizman, R., Szekely, G.A., Wijsenbeek, H. and
Livini, E. (1982). Abnormal immune response to brain tissue
antigen in the syndrome of autism. Am. J. Psy. 139:1462-1465.
Winchester, R. (1995) Rheumatoid arthritis. In: SamterUs
Immunologic Diseases, Fifth Edition. M.M. Frank, K.V. Austen,
H.N. Claman, and E.R. Unanue (Eds) Little, Brown and Co.,
Boston, 699-757.
Yonk, L.J., Warren, R.P., Burger, R.A., Cole, P., Odell, J.D.,
Warren, W.L., White, E. and Singh, V.K. (1990) CD4+ helper T
cell depression in autism. Immunol. Lett. 25:341-346.
-=-=-=-=-=-=-=-=-=-
HLA-B C4B HLA-DR
-------|------//---------|-----|-|-----//------|--|----
BF C4A a b
Figure 1. The human extended major histocompatibility complex
--------- (MHC) on the short arm of chromosome six consisting of from
left to right: the HLA-B region; three genes for complement
proteins, BF, C4A and C4B; and the HLA-DR regions consisting
of the genes for the alpha (a) and beta (b) chains. Spacing
of genes in the illustration approximately correspond to the
spacing of the genes on chromosome 6 except that the distance
between HLA-B and BF and between C4B and HLA-DR are greater
than shown.
-=-=-=-=-=-=-=-=-=-
Table 1. Specific amino acid sequences (arbitrarily numbered 1-5),
-------- of DRb1 alleles carried on extended haplotypes in subjects
with autism and normal subjects.
Sequences* DRb1 Extended Amino acid positions in
allele Haplotype DRb1 chain hypervariable
regions of HVR-3
57 60 67 70 71 73 74 78
____________________________________________________________________
1 0401 B44-SC30-DR4 D Y L Q K A A Y
0401 B62-SC33-DR4
2 0404 B60-SC31-DR4 D Y L Q R A A Y
0404 B35-SC31-DR4
0101 B35-FC(3,2)0-DR1
0101 B35-SC30-DR1
0101 B65-SC2(2,1)-DR1
3 0701 B44-FC31-DR7 V S I D R G Q V
0701 B57-S631-DR7
0701 B13-SC31-DR7
4 1501 B7-SC31-DR15 D Y I Q A A A Y
5 0301 B8-SC01-DR3 D Y I D E A A Y
___________________________________________________________________
*The HVR-3 region of sequences 1 and 2 differ only at amino acid
position number 71. The extended haplotypes B35-FC(3,2)0-DR1 and
B65-SC2(2,1)-DR1 carry a heteroduplication of the C4A and C4B genes,
respectively.
-=-=-=-=-=-=-=-=-=-
Table 2. Frequency of HVR-3 sequences in subjects with autism
--------
Number of subjects with arbitrarily
designated sequence and percentage
______________________________________________
HVR-3 Autism Normal
sequence n = 50 n = 79
____________________________________________________________
1 17 (34.0)* 2 (2.5)
2 6 (12.0) 4 (5.0)
3 16 (32.0)* 8 (10.1)
4 5 (10.0) 8 (10.1)
5 14 (28.0) 19 (24.1)
____________________________________________________________
*As compared to the normal subjects, the frequency of sequence 1 was
significantly increased (x**2 = 20.1 (with Yates correction); P LT
0.001) in the autistic subjects. Sequence 3 was also significantly
increased (x**2 = 7.9 (Yates); P lt 0.02). All P values were determined
after correction (multiplying the P value by 6 to allow for the number
of common sequences observed in the autistic and normal subjects).
-=-=-=-=-=-=-=-=-=-
Table 3. HVR-3 sequences segregating from parental chromosomes*a
--------
Maternal chromosomes Paternal chromosomes
_____________________________ _____________________________
Segregating Not segregating Segregating Not segregating
to autistic to autistic to autistic to autistic
HVR-3 child child child child
sequence*b (n = 44) (n = 44) (n = 44) (n = 44)
________________________________________________________________________
1 9*c 8 7*c 0
2 2 2 4 1
3 9*d 2 5*d 4
4 4 3 1 3
5 7 3 7 3
________________________________________________________________________
*a Only autistic subjects whose parents were studied are included in
this analysis.
*b Arbitrarily designated HVR-3 sequences on extended haplotypes (from
Table 1).
*c Number of chromosomes carrying sequence 1 segregating to the autistic
child is significantly (p LT 0.01) increased compared to that of
paternal sequences not segregating to the autistic child.
*d Number of chromosomes carrying sequence 3 segregating to the autistic
child is significantly increased (p = 0.05) compared to that of
parental sequences not segregating to the autistic child.
-=-=-=-=-=-=-=-=-=-
Table 4. Relative risk of HVR-3 and C4 alleles in autism
Frequency in subjects (percentage):
____________________________________
Marker Autistic (n = 50) Normal (n = 79) Relative Risk*a
___________________________________________________________________
HVR-3*b
1 34.0 2.5 19.8
2 12.0 5.0 2.6
3 32.0 10.1 4.2
4 10.0 10.1 1.0
5 28.0 24.1 1.2
C4A
2 10.0 3.8 2.8
3 92.0 87.3 1.7
4 2.0 5.1 0.5
6 10.0 3.8 2.8
0 (null) 42.0 26.5 2.0
C4B
1 92.0 94.9 0.6
2 6.0 5.1 1.2
3 6.0 2.5 2.6
0 (null) 54.0 20.2 4.6
___________________________________________________________________
*a Relative risk approximates how many times more likely a person is to
develop a disorder if he has a marker than if he does not. Relative
risk = (number of patients with a marker x number of contols without
a marker)/(number patients without the marker x number of controls
with the marker). In this calculation homozygotes and heterozygotes
each count as 1.
*b Arbitrarily designated HVR-3 sequences carried on extended haplotypes
(from Table 1).
The purpose of this copyright is to protect your right to make free copies of this paper for your friends and colleagues, to prevent publishers from using it for commercial advantage, and to prevent ill-meaning people from altering the meaning of the document by changing or removing a few paragraphs.