HLA-DQ

From Wikipedia, the free encyclopedia

DQ1 binding pocket with ligand
major histocompatibility complex, class II, DQ
Structure
Type	Cell surface receptor
Quartenary	αβ-heterodimer, ligand
Ligand	polypeptides, >9 residues DQ type restricted
MMDB	17145
Identifiers	alpha
Symbol(s)	HLA-DQA1 CELIAC1
Entrez	3117
OMIM	146880
EBI-HLA	DQA1*010101
Identifiers	beta 1
Symbol(s)	HLA-DQB1, IDDM1, CELIAC1
Entrez	3119
OMIM	604305
EBI-HLA	DQB1*050501
Shared data
Locus	chr.6 6p21.31
EBI-HLA	*Allele Alignments*

HLA DQ is a protein/peptide-antigen receptor and graft-versus-host disease antigen that is composed of 2 subunits, DQα and DQβ. DQα and DQβ are encoded by two loci, HLA-DQA1 and HLA-DQB1, that are found in the MHC Class II (or HLA-D) region in the Human Leukocyte Antigen complex on human chromosome 6 (see protein boxes on right for links).

Contents 1 Structure, Functions, Genetics 1.1 Function 1.2 Genetics 1.3 Understanding the Heterdimeric DQ Isoforms 1.4 Science of Understanding DQ Function 2 DQ Selection and Evolution 2.1 Regional Evolution 3 HLA DQ2 3.1 DR3-DQ2 (DQ2.5) 3.2 DR7-DQ2 (DQ2.2) 4 HLA DQ4 4.1 DQ4 and Disease 5 HLA DQ5 6 HLA DQ6 6.1 DQ6.4 7 HLA DQ7 7.1 DQ7 Distribution and Selection 7.2 HLA DQ7 and Human Disease 8 HLA DQ8 9 HLA DQ9 10 Heterozygous DQ Combinations and Disease 10.1 DQ2.5/DQ8 Heterozygotes 10.2 DQ2.2/DQ7.55 Heterozygotes 11 References 12 Links

[edit] Structure, Functions, Genetics

[edit] Function

The name 'HLA DQ' originally describes a transplantation antigen of MHC class II category of the major histocompatibility complex of humans; however, this status is an artifact of the era organ transplantation. HLA DQ functions as a cell surface receptor for foreign or self antigens. The immune system surveys antigens for foreign pathogens when presented by MHC receptors (like HLA DQ). The MHC Class II antigens are found on antigen presenting cells (APC) (macrophages, dendritic cells, and B-lymphocytes). Normally, these APC 'present' class II receptor/antigens to a great many T-cells, each with unique T-cell receptor (TCR) variants. A few TCR variants that recognize these DQ/antigen complexes are on CD4 positive (CD4+) T-cells. These T-cells, called T-helper cells, can promote the amplification of B-cells which, in turn recognize a different portion of the same antigen. Alternatively, macrophages and other megalocytes consume cells by apoptotic signaling and present self-antigens. Self antigens, in the right context, form a suppressor T-cell population that protects self tissues from immune attack or autoimmunity.

[edit] Genetics

A protein variant is called an isoform, and describes proteins that generally vary with respect to primary structure (or amino acid sequence) of the polypeptides that form the protein, or variant post-translational modification of that protein. There are 9 DQ serotypes from DQ1 to DQ9. Each serotype represents an isoform grouping based on the sero-reactivity of the chains (usually beta chain). Each serotype represents several major DQA1:DQB1 multigene haplotypes. An example of one of several haplotypes of DQ2 is DQA1*0501:DQB1*0201 (DQ2.5). This haplotype is genetically linked (or alternatively, is in linkage disequilibrium with alleles at other loci) to alleles at other loci, for example DRB1*0301. This haplotype therefore can be extended for most World’s populations to DRB1*0301:DQA1*0501:DQB1*0201. This haplotype in turns is equivalent to the serologically defined haplotype DR3-DQ2 (DR17-DQ2, R17 is a 'split antigen' of DR3 broad antigen group). It can be further extended in Northwestern Europe to include and HLA A and B allele, A*0101, B*0801 to form on of the most common 5 loci haplotypes A1-B8-DR3-DQ2.5 also known as “Super B8” and one could write this out in its long form A*0101 : B*0801 : DRB1*0301 : DQA1*0501 : DQB1*0201. There are other genes not shown in between these, they also make up a segment of a chromosome that is in strong linkage disequilibrium. What this means is that for unknown reason Super B8 genes have not randomized themselves with alleles at adjacent loci as one expects given time and known rates of inter-locus recombination. The reason could be a strong recent founder affect, such a recent rapid colonization by independent but small groups that only recently mixed, or it could be that the haplotype is under selective pressure to maintain a certain structure or a little of both.

Note on DQ Sertoype Names in Relationship with Allele Names
DQ sertoypes are usually a recognition of Beta chain isoform groups For example, DQ2 = alleles that start with DQB1*02 (*0201, *0202, and *0203).

• DQ2 is encoded by alleles that start with DQB1*02
• DQ4 is encoded by alleles that start with DQB1*04

There two exceptions worth noting:

• DQ1 is 'split' into DQ5 and DQ6, in this case the DQB1 alleles start with
  • DQ5 encoded by by alleles that start with DQB1*05
  • DQ6 encoded by by alleles that start with DQB1*06
• DQ3 is split into DQ7, DQ8, and DQ9, but allele retain DQB1*03.
  • DQ7 = DQB1*0301 (or very similar alleles)
  • DQ8 = DQB1*0302 (or very similar alleles)
  • DQ9 = DQB1*0303 (or very similar alleles)

[edit] Understanding the Heterdimeric DQ Isoforms

DQ(αβ) Isoforms given one materal (m) and one paternal (p) chromosome 6

HLA		DQB1
	allele	(m)	(p)
DQA1	(m)	αmβm (Cis m)	αmβp (Trans)
	(p)	αpβm (Trans)	αpβp (Cis p)
Result: 2 Cis, αmβm & αpβp, isoforms and 2 trans,αmβp & αpβm.

Each combination of DQA1 allele gene product with each combination of DQB1 'gene' product can potentially recombine to produce one isoform. DQ genes are highly variable in the human population. In a typical population there are many DQ alpha and beta. For Example, in the Istanbul area there are 9 alpha and 16 beta isoforms ^[1] allowing for a potential within that population of 144 DQ(αβ) heterodimeric isoforms. Most isoforms are not common.

The more common isoforms result from 'cis' haplotypes (Table to right), for example in the above example there are 20 or so more common haplotypes. These 20 haplotypes will account for at least 50% of the DQ αβ isoforms. The other 120 or so, isoforms result from random 'trans' combinations of haplotypes in individuals as a result of 'trans' paternal/maternal gene product isoforms. See Table to right.

DQ Isoforms from a person with phenotype A1*0201:B1*0202/ A1*0505:B1*0301

HLA		DQB1
	allele	*0202	*0301
DQA1	*0201	α2β2	α2β301
	*0505	α5β2	α5β301

For example, suppose one has the (for simplicity let us remove the HLA DQ prefix) A1*0201:B1*0202 and A1*0505:B1*0301 haplotypes, one 'cis'isoforms that one would share with many other people would be DQα²⁰¹β²⁰² and the other DQα⁵⁰⁵β³⁰¹. In this example, while technically correct has problems. For example, DQA1*0505 gene product is processed but the 'mutation' is on an end of the protein that is not present in the 'mature' gene product and is equal to the mature DQB1*0501 gene product. Therefore:

DQ α⁵⁰⁵ = α⁵⁰¹ = α⁵. There are other α^5xx subunit isoforms, all of the known subunits vary outside of the binding pocket as a result except α⁵⁰² which has a 58D->R mutation that is within 5A of the peptide binding cleft. Therefore the α⁵⁰¹, α⁵⁰³ - α⁵⁰⁹ can be treat as α⁵.

The other point concerns DQ A1*0201 and B1*0201, despite having similar *#### are not in any way closely related isoforms, this is just nomenclature issue. B1*0201, B1*0202, B1*0203 are very closely related and their gene products differ by one amino acid well away from the peptide binding pocket. Most β differ as a result of mutation in peptide contact residues in the binding pocket. As a result many authors conclude that β²⁰¹, β²⁰², and β²⁰³ are random, functionally equivalent variants (at least as far as peptide binding goes), thus can be treated as β². And, if that is not enough reductionism, there is only one DQA1*02 allele, DQA1*0201, and therefore we can consider this subunit DQ α². To obtain sequence comparisons see IMGT.

In the example 'cis'isoforms are DQ α²β² and α⁵β³⁰¹ and two 'trans'isoforms would be α⁵β² and DQ α²β³⁰¹. DQ α⁵β² also happens to be the cis-isoform of the DQA1*0501:B1*0201 (DQ2.5) haplotype. See Table-Upper Left for Illustration. Therefore, a trans-isoform may also be common in a population, but generally are not. Confusing? The illustration on the right shows the binding pocket of DQ α⁵β², this binding pocket represents a minority of amino acids in either DQ α⁵ or DQ β² subunits. Only the amino acids that are in the binding pocket and also have side chains that point in the direction of the peptide (yellow, in this case wheat gliadin peptide) may directly affect binding. Those that stick out of either DQ α⁵ or DQ β² subunits but not interfering with the binding of peptides can be altered without altering binding. These changes might alter T-cell recognition or have more basal forms of functional changes.

DQ Isoforms from a person with phenotype A1*0201:B1*0202/ A1*0501:B1*0201

HLA		DQB1
	allele	*0202	*0201
DQA1	*0201	α2β2	α2β2
	*0501	α5β2	α5β2
Result: 2 Isoforms: 50% α5β2 and 50% α2β2

Therefore, DQ2.5 and DQ2.2 produce an DQ β² chain that, in the illustration to the left would not markedly alter the arrangement of blue atoms. Like wise DQ2.5 and DQ7.55 produce the same mature protein and as a result same arrangement of red atoms. Changes in the gene sequence that changes amino acids side chains inside the binding pocket, or indirectly change the space under the binding pocket can alter binding, many other changes have no effect.

DQ Isoforms from a DQ2.5 homozygote

HLA		DQB1
	allele	*0201	*0201
DQA1	*0501	α5β2	α5β2
	*0501	α5β2	α5β2
Result: 1 Isoform: 100% α5β2

In general, most people have 2 haplotypes and most people have two alleles at each loci, so that they generally can produce 4 isoforms. There are instance also where the subunits can only combine to produce 2 DQ isoforms(above-left) or 1 (double homozygote, right table). In the example above-left the two haplotypes DQ A1*0501: B1*0201/ A1*0201: B1*0202 produce 50% DQ α⁵β² and 50% α²β²

To trans or not to trans
There is some controversy in the literature whether trans-isoforms are relevant. Recent genetic studies into coeliac disease have revealed that that the DQA1*0505/DQB1*0201 gene products explain disease not linked to the haplotype that produces DQ8 and DQ2.5, strongly suggesting the transisoforms can be involved in disease. But, in this example, it is known that the transproduct is almost identical to a know cis-'isoform' produced by DQ2.5. There is other evidence that some haplotypes are linked to disease but show neutral linkage with other particular haplotypes are present. At present, the bias of relative isoform frequency toward cis pairing is unknown, it is known that some trans-isoforms occur.

[edit] Science of Understanding DQ Function

HLA D (-P,-Q,-R) genes are members of the Major Histocompatibility Complex (MHC) gene family and have analogs in other mammalian species. In mice the MHC locus known as IA is homologous to human HLA DQ. Several autoimmune diseases that occur in humans that are mediated by DQ also can be induced in mice and are mediated through IA. Myasthenia Gravis is an example of one such disease^[3]. Linking specific sites on autoantigens is more difficult in humans due to the complex variation of heterologous humans, but subtle differences in T-cell stimulation associated with DQ-types has been observed^[4].

[edit] DQ Selection and Evolution

The table below represents the most frequent 3 loci DR-DQ haplotypes in European Americans^[5]. The top 25 of these cover the vast majority of haplotypes encountered in Europe and North America but also are highly represented in Asia, the Indigeonous American and African populations, where there are gaps they will be mentioned. The table illustrates that there is considerable [linkage disequilbrium] in Caucasian.

25 (of 75) Most common DR-DQ haplotypes in Caucasian Americans

DQ	DR-DQ	DR	DQ		Freq
Serotype	haplotype	B1	A1	B1	%[5]		rank
DQ2	DR17-DQ2	0301	0501	0201	13.	1	2
	DR7-DQ2	0701	0201	0202	11.	1	3
DQ4	DR8-DQ4	0801	0401	0402	2.	2	12
DQ5	DR1-DQ5	0101	0101	0501	9.	1	4
		0102	0101	0501	1.	4	14
		0103	0101	0501	0.	5	25
	DR10-DQ5	1001	0104	0501	0.	7	21
	DR16-DQ5	1601	0102	0502	1.	0	17
	DR14-DQ5	1401	0104	0503	2.	0	13
DQ6	DR15-DQ6	1502	0103	0601	0.	7	23
		1501	0102	0602	14.	2	1
	DR13-DQ6	1301	0103	0603	5.	6	6
		1302	0102	0604	0.	7	10
		1302	0102	0609	3.	4	22
DQ7	DR11-DQ7	1101	0505	0301	5.	6	5
		1104	0505	0301	2.	7	11
	DR12-DQ7	1201	0505	0301	1.	1	15
	DR13-DQ7	1303	0505	0301	0.	7	20
	DR4-DQ7	0401	0303	0301	5.	3	7
		0407	0303	0301	0.	9	18
DQ8	DR4-DQ8	0402	0301	0302	1.	0	16
		0404	0301	0302	4.	2	8
		0401	0302	0302	0.	7	24
DQ9	DR7-DQ9	0701	0201	0303	3.	7	9
	DR9-DQ9	0901	0302	0303	0.	8	19

A common question asked by many patients affected by autoimmunity, after being typed for various HLA or cytokines, is the meaning of the genetic typing. This page is designed to help understand the nature of the MHC Class II types including their distribution and linked diseases. The HLA-DR and DQ loci are associated with probably the greatest number of different diseases relative to any other loci. This is due to the complex nature of immunity and the great variation at these loci. Most of these diseases are low in frequency, some, like Type 1 diabetes and Celiac Disease are uncommon but not rare.

One common belief about 'gene tests' is that because a gene is linked to a disease it is therefore a mutation. With many disease associated genes, this is not true, there is frequently a suspect environmental causes, and that cause with the other genes and HLA that actually triggers disease in a complex and not well understood process. These environmental influences or selection change over time. And the frequencies of haplotypes lag or are in disequilibrium.

Gluten-sensitive enteropathy is an example. In the case of Coeliac disease, wheat is a factor that changed for some groups 25000 years ago and for other groups, 7500 to 5500 years ago as Near Eastern/Danubian Neolithic culture spread from SE anatolia into susceptible populations of NW Europe. There are other factors, such as the child's age when wheat is first introduced, whether or not one is exposed to a chronic upper GI infection, whether or not one eats wheat during the infection, etc. Molecular genetics and Archaeology has offered a window into the past. For example the molecular genetics tells us that Homo sapiens redistributed about >95,000 years ago^[6] and at that time many of the common DR-DQ types had already formed. The archaeology reveals that hunter gatherers early on primarily were coastal foragers that moved inland, and as they moved inland they became more effective seed gatherers and finally seed cultivators. Before the neolithization in many parts of the world, human activities were frequently associated with shell middens, nut harvesting , or, alternatively, hunting camps. The seeds that archaeologist have found from the last Ice Age tend to be more diverse, but in places like the Levant seed utilization had already increased. Of the seeds used from the tribe Triticeae the greatest diversity used was in the region from Anatolia to Iran to Arabia, these seeds were small and difficult to harvest and were probably only eaten periodically. Coeliac disease would have been episodic in small children and late onset cases probably non-existent. But as cultivation began, selection also began reducing associated haplotypes, as a result the combined HLA types, DQ2.5 and DQ8 are relatively low in areas where wheat has been cultivated the longest. In Ireland, for instance where Neolithization began late and proceeded slowly, wheat was not a common crop until after WWII, barley was often fermented to beer, and DQ2.5 is at a very high relative frequency. Therefore DQ that confer susceptibility cannot be considered negatively selective until intense cultivation appears. HLA DQ can be positively or negatively selective depending on the circumstances. This is called variable selection and all of the HLA factors are under variable selection as a consequence of how diseases appear and because of non-disease based selection on HLA. While variable selection is common in evolution, the forces acting on the HLA loci may be a stochastic example of the process. Other examples of stressing factors may include microbial exposure and immune education (Type 1 diabetes).

[edit] Regional Evolution

HLA DQ- subunit alleles, mature chains, contact variants

Known	HLA-DQ		Potential
Locus:	A1	B1	Combinations
Alleles	33	78	2475
Subunit:	α	β	isoforms
Mature Chains	24	58	1392
Contact Variants*	~9	40	360
Caucasian (USA)
Contact Variants (CV)	7	12	84
CV-haplotypes			30
*Subunits vary within 9Â of peptide in DQ2.5 or DQ8

Many HLA DQ were under positive selection of 10,000s potentially 100,000s of years in some regions. As people moved they have tend to loose haplotypes and in the process loose allelic diversity. On the other hand, on arrival at new distal locations, selection would offer unknown selective forces that would have initially favored diversity in arrivals. By an unknown process, rapid evolution occurs, as has been seen in South Americas indigeonous population (Parham and Ohta, 1996, Watkins 1995), and new alleles rapidly appear. This process may be of immediate benefit of being positively selective in that new environment, but these new alleles might also be 'sloppy' in a selective perspective, having side effects if selection changed. The table to the left demonstrates how absolute diversity at the global level translates into relative diversity at the regional level, even among heterogenous Caucasian Americans, in these cases we can see that via random αβ pairing one can arrive at 80 isoforms in the US, this is substantially less than the total number of isoforms potential in the global population, and looking at the haplotypes there are about 30 relevant cis-isoforms, with 10 that would dominate in the population. The 50 potential trans-isoforms would generally represent no more than a 25% contribution in any given individual. Some uniquely 'trans'-sioforms, (e.g) α¹β² could be as high as 15%, others such as α⁶β⁶⁰¹ randomly occur at less than 0.0015%. Of course the proportion of cis and transisoforms is not random as is evident in the linkage studies; A1 alleles do not randomly pair with B1 alleles on chromosomes. However, the per α-subunit/per β-subunit bias in pairing at the mature polypeptide level is virtually unknown.

What is the relevance of this pairing? The discussion of disease association with HLA is based on statistical argument. Part of the statistical argument is called the power of the determination and usually reflects the N of the total population under survey or the n of an individual subset (for example a given Haplotype or isoform). For a disease like Coeliac disease which is common enough to have millions affected individuals world-wide and for haplotypes like DQ2.5 that are at frequencies of >20% (phenotype frequencies close to 40%) is some populations, assessing the significance of the association is rather easy. Even assessing the significance of 1 'trans'-haplotypes (DQ2.2/DQ7.55 may not be the only trans-haplotype) has been done for Coeliac disease. Most other autoimmune and DQ mediated conditions are rare. This page does not go into detail about negative associations with disease, but a survey of pubmed will reveal a number of common autoimmune diseases. For example, the haplotype A1*0103:B1*0603 is negatively associated with myasthenia gravis, even in populations where no positive association can be recognized. More complex situations exist, for example, in some rare cases they show that a specific haplotype pair is associated with this negative or positive association. The meaning of these complex negative associations is not clear, but may indicate effects of pairing and relative isoform frequencies.

Effects of modernization on DQ mediated diseases One such effector that acts on HLA-DQ is modernization. Modernization has markedly changed culture in many areas of the world and this modernization is associated with a rise of autoimmune diseases. One reasoning as to how HLA behave in different ways across the human population is that humans expanded from a relatively small population in Africa that underwent numerous expansions from intermediate, small populations, along the way. This process is most evident in Austronesia, Eskimoes and in peoples of Northern Russia. As mentioned above, rather than re-inventing the diversity lost in migration, evolution modifies as need existing alleles or the frequencies drifted away from negatively selective to positively selected alleles. The problem is that what is selective for a largely marine-food eating culture on the arctic circle is not also selective for a largely bread/beef/dairy eating culture in say Wisconson. For example, In Sardinia, may have been settled by a relatively small group of Africans before the neolithic, small number of settlers has resulted in a very high frequency of one type A30-Cw5-B18-DR3-DQ2.5, the highest freqeuncy haplotype in 'Europe' (Sardinia is actually in the Mediterranean but is considered part of Italy). One has to assume that this haplotype was adapted or neutral to that location or the frequency of the haplotype would have declined over time through balancing selection (Heterozygous sexual selection). Myasthenia gravis was very low >50 years ago, however the rate of Myasthenia gravis has increased almost 10 fold and is approaching the rate seen in other parts of Europe. The reason for the rate change is unclear. That science is aware of the environmental factors, particularly diet, activities and stress, does not change the fact that science knows relatively little about how haplotypes were selected in the past or how certain haplotypes have come under modern negative selection with regard to some autoimmune diseases. There is no cohesive theory that explains autoimmune associations.

[edit] HLA DQ2

DQ2 is encoded by DQB1*02 alleles in combination with other alpha alleles. The two most common DQ2 β chains are very similar.

• DQB1*0201
  • DQA1*0501 linked
  • DRB1*0301 linked, (DR17)
  • DRB1*0301: DQA1*0501 : DQB1*0201, "DR3-DQ2"
  • Most common isoform α₅β₂
  • common names "DQ2" (ambiguous), "DQ2.5", "DR3-DQ2", "Super B8"
• DQB1*0202
  • DRB1*0701 linked
    • DQA1*0201 linked
      • Most common isoform α₂β₂
      • common names "DQ2" (ambiguous), "DR7-DQ2" (ambiguous), "DR7-DQ2.2"
    • DQA1*0303 linked
      • Most common isoform α₃β₂
      • common name "DR7-DQ2" (ambiguous), "DR7-DQ2.33"
• DQB1*0203 (rare)

[edit] DR3-DQ2 (DQ2.5)

The HLA DR3-DQ2 haplotype has its own page.

[edit] DR7-DQ2 (DQ2.2)

Several haplotypes in Iberia and Africa. Most common is DRB1*07:DQA1*0201:DQB1*0202

Distribution
DR7-DQ2 (DR7-DQ2.2) is found at high frequencies in the Mediterranean and Western Africa. The Eurasian geographic distribution of DQ2.2 is slightly greater than DQ2.5. Compared to DQ2.5 the freqeuncy in Sardinia is low, but in Iberia it is high reach a maximum frequency of ~30% in Northern Iberia, half that in the British Ilses. It extends along the Mediterranean and Africa at relatively high frequency and is found in high freqeucies in some Central Asian, Mongolians, and Han. It does not appear to have an indigenous presence in the West Pacific rim or the New World and DQ2.2 presence in SE Asia and Indonesia is likely the result of gene flow from India and China in post-Neolithic times. The haplotype shows considerable diversity in Africa and this has translated to Iberia with 2 addition haplotypes, DQA1*0303:DQB1*0202 and DR7:DQA1*0201:DQB1*0303. The expansion of DQ2.2 into Europe appears to have been slightly later or biased by some constriction between Iberia and the rest of the continent.

Associated Diseases
DR7-DQ2 is associated with GSE in rare cases and requires the heterozygous haplotype pair DR7-DQ2/DR11-DQ7 (or DR7-DQ2/DR12-DQ7). The disease association is through a trans-chromosomal pairing of the DQA1 and DQB1 gene products.

[edit] HLA DQ4

20 of the most common DR-DQ haplotypes in Japanese

DQ	DR-DQ	DR	DQ		Freq
Serotype	haplotype	B1	A1	B1	%[7]		rank
DQ4	DR4-DQ4	0405	0303	0401	14.	3	2
		0410	0303	0402	1.	9	11
	DR8-DQ4	0802	0401	0402	1.	3	15
DQ5	DR1-DQ5	0101	0101	0501	6.	4	5
	DR14-DQ5	1401	0104	0502	1.	1	17
		1401	0104	0503	0.	8	18
		1405	0104	0503	1.	6	13
DQ6	DR8-DQ6	0803	0103	0601	9.	3	4
	DR13-DQ6	1302	0102	0604	6.	0	6
	DR15-DQ6	1502	0103	0601	12.	9	3
		1501	0102	0602	6.	0	7
DQ7	DR11-DQ7	1101	0505	0301	1.	8	12
	DR12-DQ7	1201	0505	0301	2.	7	9
		1202	0601	0301	1.	3	15
	DR14-DQ7	1403	0503	0301	1.	4	14
DQ8	DR4-DQ8	0403	0301	0302	1.	9	10
		0406	0301	0302	3.	4	8
	DR8-DQ8	0802	0401	0302	0.	7	19
DQ9	DR9-DQ9	0901	0302	0303	16.	0	1

DQ4 distribution
The table to the left shows the values of Japanese^[7](values converted from phenotype frequencies to haplotype frequencies for sake of consistency) DR-DQ types. This table is presented here because of the diversity of DQ4 types in the Japanese population not seen elsewhere. DQ4 is typically rare most of the world but where it shows of frequently is something of interest. The node of DQ4 is with the DQA1*0401:DQB1*0402 (DQ4.24 for this page) haplotype in Northwestern Mexico and the highland region of western South America reaching 40% haplotype frequencies in that area. Outside of the Indigenous American population DQ4.24 is elevated at 10% in the Ainu of Hokkaidō, Japan. There are a number of other A-B haplotypes that suggest a connection between the Ainu and the Meso-American and Andean populations as well as Lakota Souix all have DQ4 levels higher than the Ainu. The linkage of DQ4 in Asia appears to be heaviest with DR8 (DR*0801, DR*0802, DR*0804) for DQ4.24 and the frequency is elevated from the Ryūkyū Islands to Okhotsk, Ulchi, Negidal, Tofalar at approximately 10% falling off in the Mansi at 4% and punctate levels in between. Haplotype diversity of DQB1*0402 appears to be centered around the Amur River/Japanese Island Chain, and diversity of DQB1*0401 very roughly follows a similar pattern. DQ4.24 is also high in the Swedes however this may be due to east to west gene flow tracable at other HLA loci.

Since DQA1*0401:DQB1*0402 is found in the !Kung, one reasonably assumes it evolved in Africa and exited with one of several potential waves. Tracing the migration route is excessively difficult, but it appears that a possible origin in Central Asia and not the West Pacific Rim/Austro-Indic route postulated as the early human distribution. The most common haplotypes in the !Kung (for example Cw-B) that also appear in Eurasia appear to have been associated with the earliest migration, and is suggestive of a coastal migration; however the relatively high frequencies in the Ainu^[8] and Amur basin^[9] suggest a migration through the Transbaikal that is consistent with archaeology from about 18 kya. One expects with such a route that Korean would be higher than Japanese and Japanese higher the Ryukuans still higher than Tiawan aboriginals. From the west gene frequencies in the Levant and Black Sea region are at 'diffusive' levels whereas there are pockets of increased frequency in the Zoroastrians of Yadz region (DQA1*0401 and DQB1*0402). Thus the DR8-DQ4.24 haplotype is probably one of western origin.

The DR4-DQA1*0303:DQB1*040X can be found at high frequencies in PNG highland groups ^[10] but not DQ4.24. The DR*0405 and DR*410 are found specifically associated with these DQ types and there is some haplotype diversity. So that it appears the presence of the DQA1*03:DQB1*04 is of West Pacific Rim origins in Japanese and proximal Siberians, but unfortuantly there is no current typing of these haplotypes in the Taiwan aboriginal population. The presence in Indonesia may be the result of retrograde gene flow that can be established by other HLA types as well as mtDNA.

[edit] DQ4 and Disease

DQ4 is associated with:

• juvenile diabetic retinopathy ^[11][12]

The DR8-DQ4 haplotype is associated with

• papillary thyroid carcinomas^[13]
• juvenile idiopathic arthritis^[14] via DR8 allele ^[15],
• chronic chlamydia infection^[16],
• possibly pemphigus ^[17]

DR4-DQ4(DRB1*0405:DQB1*0401) is associated with:

• autoimmune pancreatitis ^[18],
• juvenile type 1 diabetes ^[19],
• rheumatoid arthritis^[20] in Japanese.

The DQA1*0303:DQB1*04 haplotype is associated with:

• Crohn's disease ^[21][22] in Japanese.

Other diseases mentioned are high altitude pulmonary edema, Vogt-Koyanagi-Harada syndrome (DRB1*0405, see above table for Japanese), HIV resistance in the US, and haemophilia A(anti-FVIII inhibitor response).

[edit] HLA DQ5

[Section in Progress] DR4-DQ8 and DR3/7 DQ2 are associated with Idiopathic nephrotic syndrome in Polish children (PMID 16967287) negatively associated with DQ5. A study on the relationship between HLA-DR, DQ antigen, and intracranial aneurysm in the Han nationality show DQ5 more likely (PMID 16904993). MuSK antibody-positive myasthenia gravis HLA-DR14-DQ5 (PMID 16769963). (PMID 15144476), (PMID 14635465), (PMID 12137599), (PMID 11836294), (PMID 10833064).

[edit] HLA DQ6

DQ6 is composed of several common haplotypes

• A1*0103:B1*0601 (DQ6.1)
• A1*0102:B1*0602 (DQ6.2)
• A1*0103:B1*0602
• A1*0103:B1*0603 (DQ6.3)
• A1*0102:B1*0603
• A1*0102:B1*0604 (DQ6.4)
• A1*0102:B1*0605 (DQ6.5)
• A1*0102:B1*0609

[edit] DQ6.4

A1*0102:B1*0604 is associated with tymoma induced myastenia gravis^[23].

[edit] HLA DQ7

DQ7 represents several DQA1 DQB1*0301 combinations, and is linked to several DRB1* alleles.

[edit] DQ7 Distribution and Selection

HLA DQ7 is commonly represented in the human population, particularly in the near eastern population, by high frequencies of the haplotype DR11-DQA1*0505-DRB1*0301. In older HLA studies DQA1*0501 was recognized however Pena discovered that all DRB1*0301 associated DQ1*0501 in Europeans was actually DQA1*0505. This makes analysis of the allele frequencies in Europeans complicated. The processed alpha subunit of DQA1*0505 is identical to that of DQA1*0501, but some slight differences in the association with autoimmune disease are observed. DQ7 is found spread globally and is found in high frequencies in the Near East, Greece, Italy and France and is an indicator of the effects of the neolithic migrations. The frequency in Lebanese is approximately 40% and should be considered under positive selection at this level. It is also high in frequency in the new world, but with DR types less commonly encountered in the old world. DQA1*05 allele is not clear in the new world. More rarely DQ7 is found with DQA1*0303:DQB1*0301 and other DQA1* alleles. There is a rather large degree of disequilibration about DQA1*0301 suggesting that this is one of the older and more established HLA DQB1* alleles in eurasia. The intron structure of DQB1 suggest that DQB1*0301 DQB1*0302/*0303 split occurred before DQB1*0302/*0303, the distribution of *03 in Africa suggest that recombination DQA1*03:DQB1*0301 are primarily the result of recombination events that have occurred in Africa. A recent study of Myastenia Gravis in Houston confirms the presence of A*0505:B*0301 in Nigeria. B1*0301 and A1*03 haplotypes are found at relatively high frequencies in SE Asia and Austronesia, also indicating that it is well established in the exoafrican population. DQB1*0301 may be under current positive selection in the human population, at least in areas where DQ2.5 and DQ8 are high, as it confers resistance to type 1 diabetes and coeliac disease (except in areas where DQ2.2 is high). A high frequency of B1*0301 in the near east may be explained by a selective sweep of DQ2 and DQ8 from the population as a result of >25,000 years of Triticeae use and >10,000 years of triticeae cultivation. This allele may be associated with a preferential (or linked sweep) expansion into the Paris basin by Italian, near eastern and Armenian class I haplotypes, that show a marked displacement of France's pre-neolithic settlers as populations on all sides exhibit high levels of A1-B8...DQ2.5, A2-B44, A3-B7. The Belgian population shows specific class I links to the Armenian or Eastern Black Sea popuations as well as Anatolians, Syrians and Lebanese, that shows little east/west spread and Germans shows minor influences from Paris Basin enriched types and direct influences of DQB1*0301 enriched peoples from the Carpathian/Danube/Balkan region, with the Netherlands intermediate between German and French links to the black sea and eastern mediterranea. Gene frequencies of DQ types appear to align with the boundary of LBK culture during the early neolithic period, except in SE.Briton where it appears a pastoral (cattle) culture migrated from Northern France (links to Belgium and Paris Basin) during France's early neolithic.

[edit] HLA DQ7 and Human Disease

DQB1*0301 appears to be more associated with early onset myasthenia gravis in Japanese than DQ8, and was also found along with DQB1*0304 to be associated with Chinese MG. DQ7 or asscoiated DR types may play a role in rheumatoid arthritis. In celiac disease the DQ7 (A*0505/1) can mediate celiac disease when HLA DQ2.2 is also present. HLA DQB1*0301 in Turks is associated with Thymoma but the risk may be associated with HLA class I loci. With regard to Type 1 Diabetese, Hepatitis types B & C, HLA DQ7 may be protective. DQA1*0505, DQB1*0301 appear to increase the risk for melanoma in the Spanish population however this may have a linkage to more recent fair skinned migrants. DQB1*0301 is also associated with Allergic fungal sinisitus, human papillomavirus (HPV) induced warts, limited cutaneous systemic schlerosis in Africans, primary sclerosing cholangitis in Southern Europeans. DQB1*0301 does not to play a role in any frequently occurring autoimmune disease and its presence in the near east and suppressed frequencies of coeliac disease and Type 1 diabetes in these regions is suggestive that it has a positive selection in post-mesolithic cereal based societies in the Western Eurasia.

[edit] HLA DQ8

DQ8 has been moved to a new page

[edit] HLA DQ9

[Section in Progress]

[edit] Heterozygous DQ Combinations and Disease

population	DQ2.5	DQ8	DQ2.5/8
-----------------.	----%----.	----%----.	----%----.
Sweden	15.9	18.7	5.9
Jalisco	11.4	22.8	5.2
England	12.4	16.8	4.2
Khazak	13.1	11	2.9
Uygur	12.6	11.4	2.9
Finland	9	15.7	2.8
Poland	10.7	9.9	2.1

[edit] DQ2.5/DQ8 Heterozygotes

The distribution of this phenotype is largely the result of admixtures between peoples of eastern or central Asian origin and peoples of western or central Asian origin. The highest frequencies, by random mating, are expected in Sweden, but pockets of high levels also occur in Mexico, and a larger range risk exists in Central Asia.

Diseases that appear to be increased in Heterozygotes are Type 1 Diabetes. New evidence is showing an increased risk for late onset Type 1 diabetes in Heterozygotes (which includes ambiguous Type I/Type II diabetes. Celiac Disease may have a slightly increased risk with a more severe course of disease.

[edit] DQ2.2/DQ7.55 Heterozygotes

[In Progress]

[edit] References

[edit] Links

Molecular Anthropology Yahoo Group
HLA Allele and Haplotype Frequency Database
IMGT-HLA Database